diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,113865 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 4064,
+  "global_step": 16256,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.151574803149606e-05,
+      "grad_norm": 5.482029914855957,
+      "learning_rate": 2e-05,
+      "loss": 13.6418,
+      "step": 1
+    },
+    {
+      "epoch": 6.151574803149606e-05,
+      "eval_loss": 14.886382102966309,
+      "eval_runtime": 1039.1143,
+      "eval_samples_per_second": 3.69,
+      "eval_steps_per_second": 3.69,
+      "step": 1
+    },
+    {
+      "epoch": 0.00012303149606299212,
+      "grad_norm": 5.984772205352783,
+      "learning_rate": 4e-05,
+      "loss": 15.0598,
+      "step": 2
+    },
+    {
+      "epoch": 0.00018454724409448819,
+      "grad_norm": 5.734090805053711,
+      "learning_rate": 6e-05,
+      "loss": 16.4218,
+      "step": 3
+    },
+    {
+      "epoch": 0.00024606299212598425,
+      "grad_norm": 5.444360256195068,
+      "learning_rate": 8e-05,
+      "loss": 12.9511,
+      "step": 4
+    },
+    {
+      "epoch": 0.0003075787401574803,
+      "grad_norm": 5.957120418548584,
+      "learning_rate": 0.0001,
+      "loss": 15.2019,
+      "step": 5
+    },
+    {
+      "epoch": 0.00036909448818897637,
+      "grad_norm": 6.127737045288086,
+      "learning_rate": 0.00012,
+      "loss": 16.6303,
+      "step": 6
+    },
+    {
+      "epoch": 0.00043061023622047243,
+      "grad_norm": 6.821583271026611,
+      "learning_rate": 0.00014,
+      "loss": 13.9561,
+      "step": 7
+    },
+    {
+      "epoch": 0.0004921259842519685,
+      "grad_norm": 6.55456018447876,
+      "learning_rate": 0.00016,
+      "loss": 10.5351,
+      "step": 8
+    },
+    {
+      "epoch": 0.0005536417322834646,
+      "grad_norm": 8.733551025390625,
+      "learning_rate": 0.00018,
+      "loss": 12.0014,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006151574803149606,
+      "grad_norm": 5.629597187042236,
+      "learning_rate": 0.0002,
+      "loss": 10.9767,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006766732283464567,
+      "grad_norm": 9.613995552062988,
+      "learning_rate": 0.00019999999813027883,
+      "loss": 9.8536,
+      "step": 11
+    },
+    {
+      "epoch": 0.0007381889763779527,
+      "grad_norm": 5.925160884857178,
+      "learning_rate": 0.0001999999925211153,
+      "loss": 8.3939,
+      "step": 12
+    },
+    {
+      "epoch": 0.0007997047244094488,
+      "grad_norm": 5.521446704864502,
+      "learning_rate": 0.0001999999831725097,
+      "loss": 8.6053,
+      "step": 13
+    },
+    {
+      "epoch": 0.0008612204724409449,
+      "grad_norm": 4.685490131378174,
+      "learning_rate": 0.00019999997008446232,
+      "loss": 7.699,
+      "step": 14
+    },
+    {
+      "epoch": 0.0009227362204724409,
+      "grad_norm": 3.1345713138580322,
+      "learning_rate": 0.0001999999532569737,
+      "loss": 7.0154,
+      "step": 15
+    },
+    {
+      "epoch": 0.000984251968503937,
+      "grad_norm": 2.5838406085968018,
+      "learning_rate": 0.0001999999326900444,
+      "loss": 4.7423,
+      "step": 16
+    },
+    {
+      "epoch": 0.0010457677165354332,
+      "grad_norm": 2.0108108520507812,
+      "learning_rate": 0.0001999999083836753,
+      "loss": 5.5067,
+      "step": 17
+    },
+    {
+      "epoch": 0.0011072834645669291,
+      "grad_norm": 2.3351149559020996,
+      "learning_rate": 0.0001999998803378672,
+      "loss": 5.7271,
+      "step": 18
+    },
+    {
+      "epoch": 0.0011687992125984253,
+      "grad_norm": 2.063504219055176,
+      "learning_rate": 0.0001999998485526212,
+      "loss": 5.4698,
+      "step": 19
+    },
+    {
+      "epoch": 0.0012303149606299212,
+      "grad_norm": 2.36728572845459,
+      "learning_rate": 0.00019999981302793845,
+      "loss": 5.7157,
+      "step": 20
+    },
+    {
+      "epoch": 0.0012918307086614174,
+      "grad_norm": 1.942466139793396,
+      "learning_rate": 0.00019999977376382032,
+      "loss": 5.8405,
+      "step": 21
+    },
+    {
+      "epoch": 0.0013533464566929134,
+      "grad_norm": 2.3639914989471436,
+      "learning_rate": 0.0001999997307602683,
+      "loss": 4.7815,
+      "step": 22
+    },
+    {
+      "epoch": 0.0014148622047244095,
+      "grad_norm": 2.7135772705078125,
+      "learning_rate": 0.00019999968401728392,
+      "loss": 5.0877,
+      "step": 23
+    },
+    {
+      "epoch": 0.0014763779527559055,
+      "grad_norm": 4.7281622886657715,
+      "learning_rate": 0.00019999963353486898,
+      "loss": 6.4028,
+      "step": 24
+    },
+    {
+      "epoch": 0.0015378937007874017,
+      "grad_norm": 2.477782964706421,
+      "learning_rate": 0.0001999995793130254,
+      "loss": 5.158,
+      "step": 25
+    },
+    {
+      "epoch": 0.0015994094488188976,
+      "grad_norm": 1.902949333190918,
+      "learning_rate": 0.0001999995213517551,
+      "loss": 4.7101,
+      "step": 26
+    },
+    {
+      "epoch": 0.0016609251968503938,
+      "grad_norm": 2.387880563735962,
+      "learning_rate": 0.00019999945965106037,
+      "loss": 5.9719,
+      "step": 27
+    },
+    {
+      "epoch": 0.0017224409448818897,
+      "grad_norm": 1.3084725141525269,
+      "learning_rate": 0.00019999939421094345,
+      "loss": 5.4829,
+      "step": 28
+    },
+    {
+      "epoch": 0.001783956692913386,
+      "grad_norm": 1.1015849113464355,
+      "learning_rate": 0.00019999932503140676,
+      "loss": 5.2135,
+      "step": 29
+    },
+    {
+      "epoch": 0.0018454724409448819,
+      "grad_norm": 0.9842422008514404,
+      "learning_rate": 0.00019999925211245295,
+      "loss": 5.1587,
+      "step": 30
+    },
+    {
+      "epoch": 0.001906988188976378,
+      "grad_norm": 1.2972447872161865,
+      "learning_rate": 0.00019999917545408474,
+      "loss": 5.0864,
+      "step": 31
+    },
+    {
+      "epoch": 0.001968503937007874,
+      "grad_norm": 1.1920126676559448,
+      "learning_rate": 0.00019999909505630497,
+      "loss": 4.7727,
+      "step": 32
+    },
+    {
+      "epoch": 0.00203001968503937,
+      "grad_norm": 1.1871001720428467,
+      "learning_rate": 0.00019999901091911662,
+      "loss": 4.4437,
+      "step": 33
+    },
+    {
+      "epoch": 0.0020915354330708663,
+      "grad_norm": 1.1514976024627686,
+      "learning_rate": 0.00019999892304252293,
+      "loss": 5.1533,
+      "step": 34
+    },
+    {
+      "epoch": 0.0021530511811023623,
+      "grad_norm": 0.9348739385604858,
+      "learning_rate": 0.00019999883142652708,
+      "loss": 4.5989,
+      "step": 35
+    },
+    {
+      "epoch": 0.0022145669291338582,
+      "grad_norm": 1.83991277217865,
+      "learning_rate": 0.00019999873607113257,
+      "loss": 4.7957,
+      "step": 36
+    },
+    {
+      "epoch": 0.002276082677165354,
+      "grad_norm": 0.9639754891395569,
+      "learning_rate": 0.00019999863697634293,
+      "loss": 5.1886,
+      "step": 37
+    },
+    {
+      "epoch": 0.0023375984251968506,
+      "grad_norm": 0.8030492067337036,
+      "learning_rate": 0.00019999853414216188,
+      "loss": 4.9723,
+      "step": 38
+    },
+    {
+      "epoch": 0.0023991141732283465,
+      "grad_norm": 0.7356982231140137,
+      "learning_rate": 0.00019999842756859325,
+      "loss": 4.5846,
+      "step": 39
+    },
+    {
+      "epoch": 0.0024606299212598425,
+      "grad_norm": 0.6708055734634399,
+      "learning_rate": 0.00019999831725564103,
+      "loss": 4.3259,
+      "step": 40
+    },
+    {
+      "epoch": 0.0025221456692913384,
+      "grad_norm": 0.8197684288024902,
+      "learning_rate": 0.00019999820320330938,
+      "loss": 4.4742,
+      "step": 41
+    },
+    {
+      "epoch": 0.002583661417322835,
+      "grad_norm": 0.8065810203552246,
+      "learning_rate": 0.00019999808541160252,
+      "loss": 4.4442,
+      "step": 42
+    },
+    {
+      "epoch": 0.0026451771653543308,
+      "grad_norm": 1.2500178813934326,
+      "learning_rate": 0.00019999796388052486,
+      "loss": 4.7131,
+      "step": 43
+    },
+    {
+      "epoch": 0.0027066929133858267,
+      "grad_norm": 0.6776506304740906,
+      "learning_rate": 0.00019999783861008097,
+      "loss": 4.4275,
+      "step": 44
+    },
+    {
+      "epoch": 0.0027682086614173227,
+      "grad_norm": 0.8938648700714111,
+      "learning_rate": 0.0001999977096002755,
+      "loss": 2.8587,
+      "step": 45
+    },
+    {
+      "epoch": 0.002829724409448819,
+      "grad_norm": 0.8551689982414246,
+      "learning_rate": 0.0001999975768511133,
+      "loss": 5.0646,
+      "step": 46
+    },
+    {
+      "epoch": 0.002891240157480315,
+      "grad_norm": 1.5305277109146118,
+      "learning_rate": 0.00019999744036259935,
+      "loss": 6.1155,
+      "step": 47
+    },
+    {
+      "epoch": 0.002952755905511811,
+      "grad_norm": 1.9128812551498413,
+      "learning_rate": 0.0001999973001347387,
+      "loss": 5.0263,
+      "step": 48
+    },
+    {
+      "epoch": 0.003014271653543307,
+      "grad_norm": 1.0714994668960571,
+      "learning_rate": 0.00019999715616753666,
+      "loss": 4.6472,
+      "step": 49
+    },
+    {
+      "epoch": 0.0030757874015748033,
+      "grad_norm": 0.7470815777778625,
+      "learning_rate": 0.00019999700846099855,
+      "loss": 4.1215,
+      "step": 50
+    },
+    {
+      "epoch": 0.0031373031496062993,
+      "grad_norm": 1.3089985847473145,
+      "learning_rate": 0.00019999685701512992,
+      "loss": 4.2695,
+      "step": 51
+    },
+    {
+      "epoch": 0.0031988188976377952,
+      "grad_norm": 1.1406008005142212,
+      "learning_rate": 0.00019999670182993645,
+      "loss": 4.2568,
+      "step": 52
+    },
+    {
+      "epoch": 0.003260334645669291,
+      "grad_norm": 1.4149187803268433,
+      "learning_rate": 0.00019999654290542393,
+      "loss": 4.1756,
+      "step": 53
+    },
+    {
+      "epoch": 0.0033218503937007876,
+      "grad_norm": 0.8637690544128418,
+      "learning_rate": 0.00019999638024159827,
+      "loss": 4.0287,
+      "step": 54
+    },
+    {
+      "epoch": 0.0033833661417322835,
+      "grad_norm": 0.6743874549865723,
+      "learning_rate": 0.00019999621383846563,
+      "loss": 4.4588,
+      "step": 55
+    },
+    {
+      "epoch": 0.0034448818897637795,
+      "grad_norm": 0.9613650441169739,
+      "learning_rate": 0.00019999604369603214,
+      "loss": 3.479,
+      "step": 56
+    },
+    {
+      "epoch": 0.0035063976377952754,
+      "grad_norm": 0.8676753640174866,
+      "learning_rate": 0.00019999586981430425,
+      "loss": 4.1257,
+      "step": 57
+    },
+    {
+      "epoch": 0.003567913385826772,
+      "grad_norm": 0.7760907411575317,
+      "learning_rate": 0.00019999569219328843,
+      "loss": 4.505,
+      "step": 58
+    },
+    {
+      "epoch": 0.0036294291338582678,
+      "grad_norm": 0.887509822845459,
+      "learning_rate": 0.0001999955108329913,
+      "loss": 4.8,
+      "step": 59
+    },
+    {
+      "epoch": 0.0036909448818897637,
+      "grad_norm": 0.7134016752243042,
+      "learning_rate": 0.00019999532573341963,
+      "loss": 4.7897,
+      "step": 60
+    },
+    {
+      "epoch": 0.0037524606299212597,
+      "grad_norm": 0.8391125202178955,
+      "learning_rate": 0.00019999513689458035,
+      "loss": 4.3249,
+      "step": 61
+    },
+    {
+      "epoch": 0.003813976377952756,
+      "grad_norm": 1.0558634996414185,
+      "learning_rate": 0.00019999494431648058,
+      "loss": 4.2079,
+      "step": 62
+    },
+    {
+      "epoch": 0.003875492125984252,
+      "grad_norm": 0.7980897426605225,
+      "learning_rate": 0.0001999947479991275,
+      "loss": 4.0956,
+      "step": 63
+    },
+    {
+      "epoch": 0.003937007874015748,
+      "grad_norm": 0.7305802702903748,
+      "learning_rate": 0.0001999945479425284,
+      "loss": 4.4303,
+      "step": 64
+    },
+    {
+      "epoch": 0.003998523622047244,
+      "grad_norm": 0.7589819431304932,
+      "learning_rate": 0.00019999434414669077,
+      "loss": 4.7075,
+      "step": 65
+    },
+    {
+      "epoch": 0.00406003937007874,
+      "grad_norm": 0.7469328045845032,
+      "learning_rate": 0.0001999941366116223,
+      "loss": 4.2832,
+      "step": 66
+    },
+    {
+      "epoch": 0.004121555118110236,
+      "grad_norm": 0.9335817694664001,
+      "learning_rate": 0.00019999392533733067,
+      "loss": 4.0655,
+      "step": 67
+    },
+    {
+      "epoch": 0.004183070866141733,
+      "grad_norm": 0.745487630367279,
+      "learning_rate": 0.0001999937103238238,
+      "loss": 4.3128,
+      "step": 68
+    },
+    {
+      "epoch": 0.004244586614173229,
+      "grad_norm": 1.0865602493286133,
+      "learning_rate": 0.00019999349157110978,
+      "loss": 4.8683,
+      "step": 69
+    },
+    {
+      "epoch": 0.004306102362204725,
+      "grad_norm": 0.6161433458328247,
+      "learning_rate": 0.00019999326907919673,
+      "loss": 4.3485,
+      "step": 70
+    },
+    {
+      "epoch": 0.0043676181102362205,
+      "grad_norm": 0.5696821808815002,
+      "learning_rate": 0.00019999304284809303,
+      "loss": 3.9096,
+      "step": 71
+    },
+    {
+      "epoch": 0.0044291338582677165,
+      "grad_norm": 0.7559783458709717,
+      "learning_rate": 0.00019999281287780708,
+      "loss": 3.8548,
+      "step": 72
+    },
+    {
+      "epoch": 0.004490649606299212,
+      "grad_norm": 0.7913506627082825,
+      "learning_rate": 0.0001999925791683475,
+      "loss": 4.0762,
+      "step": 73
+    },
+    {
+      "epoch": 0.004552165354330708,
+      "grad_norm": 0.7013502717018127,
+      "learning_rate": 0.000199992341719723,
+      "loss": 4.4647,
+      "step": 74
+    },
+    {
+      "epoch": 0.004613681102362204,
+      "grad_norm": 0.7977866530418396,
+      "learning_rate": 0.00019999210053194254,
+      "loss": 3.5262,
+      "step": 75
+    },
+    {
+      "epoch": 0.004675196850393701,
+      "grad_norm": 0.7070642113685608,
+      "learning_rate": 0.0001999918556050151,
+      "loss": 3.6275,
+      "step": 76
+    },
+    {
+      "epoch": 0.004736712598425197,
+      "grad_norm": 0.7864336371421814,
+      "learning_rate": 0.0001999916069389498,
+      "loss": 4.7123,
+      "step": 77
+    },
+    {
+      "epoch": 0.004798228346456693,
+      "grad_norm": 0.7045543789863586,
+      "learning_rate": 0.00019999135453375598,
+      "loss": 4.161,
+      "step": 78
+    },
+    {
+      "epoch": 0.004859744094488189,
+      "grad_norm": 0.9334936738014221,
+      "learning_rate": 0.00019999109838944305,
+      "loss": 3.781,
+      "step": 79
+    },
+    {
+      "epoch": 0.004921259842519685,
+      "grad_norm": 0.7052764892578125,
+      "learning_rate": 0.00019999083850602063,
+      "loss": 3.9291,
+      "step": 80
+    },
+    {
+      "epoch": 0.004982775590551181,
+      "grad_norm": 0.7392017841339111,
+      "learning_rate": 0.0001999905748834984,
+      "loss": 4.3447,
+      "step": 81
+    },
+    {
+      "epoch": 0.005044291338582677,
+      "grad_norm": 0.615789532661438,
+      "learning_rate": 0.00019999030752188625,
+      "loss": 3.9038,
+      "step": 82
+    },
+    {
+      "epoch": 0.005105807086614173,
+      "grad_norm": 0.6859673857688904,
+      "learning_rate": 0.00019999003642119415,
+      "loss": 4.9793,
+      "step": 83
+    },
+    {
+      "epoch": 0.00516732283464567,
+      "grad_norm": 0.8041189908981323,
+      "learning_rate": 0.00019998976158143222,
+      "loss": 3.9467,
+      "step": 84
+    },
+    {
+      "epoch": 0.005228838582677166,
+      "grad_norm": 1.1325663328170776,
+      "learning_rate": 0.0001999894830026108,
+      "loss": 4.4563,
+      "step": 85
+    },
+    {
+      "epoch": 0.0052903543307086616,
+      "grad_norm": 1.0158467292785645,
+      "learning_rate": 0.00019998920068474028,
+      "loss": 4.7595,
+      "step": 86
+    },
+    {
+      "epoch": 0.0053518700787401575,
+      "grad_norm": 0.9293975234031677,
+      "learning_rate": 0.0001999889146278312,
+      "loss": 4.3069,
+      "step": 87
+    },
+    {
+      "epoch": 0.0054133858267716535,
+      "grad_norm": 0.9527814984321594,
+      "learning_rate": 0.00019998862483189424,
+      "loss": 3.879,
+      "step": 88
+    },
+    {
+      "epoch": 0.005474901574803149,
+      "grad_norm": 0.8251259326934814,
+      "learning_rate": 0.00019998833129694028,
+      "loss": 4.5881,
+      "step": 89
+    },
+    {
+      "epoch": 0.005536417322834645,
+      "grad_norm": 1.355383276939392,
+      "learning_rate": 0.00019998803402298027,
+      "loss": 4.1265,
+      "step": 90
+    },
+    {
+      "epoch": 0.005597933070866141,
+      "grad_norm": 0.8191584944725037,
+      "learning_rate": 0.00019998773301002537,
+      "loss": 3.8795,
+      "step": 91
+    },
+    {
+      "epoch": 0.005659448818897638,
+      "grad_norm": 0.8508920073509216,
+      "learning_rate": 0.00019998742825808676,
+      "loss": 3.6558,
+      "step": 92
+    },
+    {
+      "epoch": 0.005720964566929134,
+      "grad_norm": 0.9597457051277161,
+      "learning_rate": 0.0001999871197671759,
+      "loss": 4.2718,
+      "step": 93
+    },
+    {
+      "epoch": 0.00578248031496063,
+      "grad_norm": 1.0464812517166138,
+      "learning_rate": 0.0001999868075373043,
+      "loss": 4.6338,
+      "step": 94
+    },
+    {
+      "epoch": 0.005843996062992126,
+      "grad_norm": 1.2142198085784912,
+      "learning_rate": 0.00019998649156848365,
+      "loss": 3.724,
+      "step": 95
+    },
+    {
+      "epoch": 0.005905511811023622,
+      "grad_norm": 1.9712352752685547,
+      "learning_rate": 0.00019998617186072574,
+      "loss": 2.6908,
+      "step": 96
+    },
+    {
+      "epoch": 0.005967027559055118,
+      "grad_norm": 0.8344887495040894,
+      "learning_rate": 0.00019998584841404254,
+      "loss": 4.0843,
+      "step": 97
+    },
+    {
+      "epoch": 0.006028543307086614,
+      "grad_norm": 0.8581575751304626,
+      "learning_rate": 0.00019998552122844616,
+      "loss": 4.2033,
+      "step": 98
+    },
+    {
+      "epoch": 0.00609005905511811,
+      "grad_norm": 0.841177225112915,
+      "learning_rate": 0.00019998519030394883,
+      "loss": 3.5895,
+      "step": 99
+    },
+    {
+      "epoch": 0.006151574803149607,
+      "grad_norm": 0.8409016132354736,
+      "learning_rate": 0.00019998485564056288,
+      "loss": 3.6816,
+      "step": 100
+    },
+    {
+      "epoch": 0.006213090551181103,
+      "grad_norm": 1.1098588705062866,
+      "learning_rate": 0.00019998451723830088,
+      "loss": 3.6792,
+      "step": 101
+    },
+    {
+      "epoch": 0.0062746062992125986,
+      "grad_norm": 0.9511058330535889,
+      "learning_rate": 0.00019998417509717548,
+      "loss": 4.0614,
+      "step": 102
+    },
+    {
+      "epoch": 0.0063361220472440945,
+      "grad_norm": 0.6864252090454102,
+      "learning_rate": 0.00019998382921719942,
+      "loss": 3.8433,
+      "step": 103
+    },
+    {
+      "epoch": 0.0063976377952755905,
+      "grad_norm": 0.8367637395858765,
+      "learning_rate": 0.0001999834795983857,
+      "loss": 4.4264,
+      "step": 104
+    },
+    {
+      "epoch": 0.006459153543307086,
+      "grad_norm": 0.8895636796951294,
+      "learning_rate": 0.00019998312624074736,
+      "loss": 3.8196,
+      "step": 105
+    },
+    {
+      "epoch": 0.006520669291338582,
+      "grad_norm": 1.2029048204421997,
+      "learning_rate": 0.00019998276914429762,
+      "loss": 4.642,
+      "step": 106
+    },
+    {
+      "epoch": 0.006582185039370078,
+      "grad_norm": 0.8605619668960571,
+      "learning_rate": 0.00019998240830904984,
+      "loss": 3.525,
+      "step": 107
+    },
+    {
+      "epoch": 0.006643700787401575,
+      "grad_norm": 1.1288233995437622,
+      "learning_rate": 0.0001999820437350175,
+      "loss": 4.3789,
+      "step": 108
+    },
+    {
+      "epoch": 0.006705216535433071,
+      "grad_norm": 0.8526942729949951,
+      "learning_rate": 0.00019998167542221423,
+      "loss": 3.7825,
+      "step": 109
+    },
+    {
+      "epoch": 0.006766732283464567,
+      "grad_norm": 1.278733491897583,
+      "learning_rate": 0.00019998130337065383,
+      "loss": 3.7498,
+      "step": 110
+    },
+    {
+      "epoch": 0.006828248031496063,
+      "grad_norm": 0.7181769609451294,
+      "learning_rate": 0.0001999809275803502,
+      "loss": 4.2836,
+      "step": 111
+    },
+    {
+      "epoch": 0.006889763779527559,
+      "grad_norm": 0.9225865006446838,
+      "learning_rate": 0.00019998054805131737,
+      "loss": 3.9188,
+      "step": 112
+    },
+    {
+      "epoch": 0.006951279527559055,
+      "grad_norm": 0.9506409168243408,
+      "learning_rate": 0.00019998016478356957,
+      "loss": 3.8086,
+      "step": 113
+    },
+    {
+      "epoch": 0.007012795275590551,
+      "grad_norm": 1.0083173513412476,
+      "learning_rate": 0.00019997977777712106,
+      "loss": 4.0807,
+      "step": 114
+    },
+    {
+      "epoch": 0.007074311023622047,
+      "grad_norm": 0.8565378189086914,
+      "learning_rate": 0.00019997938703198643,
+      "loss": 3.7325,
+      "step": 115
+    },
+    {
+      "epoch": 0.007135826771653544,
+      "grad_norm": 0.7908998131752014,
+      "learning_rate": 0.0001999789925481802,
+      "loss": 4.0611,
+      "step": 116
+    },
+    {
+      "epoch": 0.00719734251968504,
+      "grad_norm": 0.7109431028366089,
+      "learning_rate": 0.00019997859432571713,
+      "loss": 3.9249,
+      "step": 117
+    },
+    {
+      "epoch": 0.0072588582677165355,
+      "grad_norm": 0.7850291728973389,
+      "learning_rate": 0.00019997819236461213,
+      "loss": 3.7864,
+      "step": 118
+    },
+    {
+      "epoch": 0.0073203740157480315,
+      "grad_norm": 1.1817116737365723,
+      "learning_rate": 0.00019997778666488024,
+      "loss": 4.5303,
+      "step": 119
+    },
+    {
+      "epoch": 0.0073818897637795275,
+      "grad_norm": 1.043411135673523,
+      "learning_rate": 0.00019997737722653663,
+      "loss": 3.7504,
+      "step": 120
+    },
+    {
+      "epoch": 0.007443405511811023,
+      "grad_norm": 0.8229930996894836,
+      "learning_rate": 0.0001999769640495966,
+      "loss": 4.1244,
+      "step": 121
+    },
+    {
+      "epoch": 0.007504921259842519,
+      "grad_norm": 0.825101912021637,
+      "learning_rate": 0.00019997654713407555,
+      "loss": 3.8745,
+      "step": 122
+    },
+    {
+      "epoch": 0.007566437007874015,
+      "grad_norm": 0.9613546133041382,
+      "learning_rate": 0.00019997612647998917,
+      "loss": 4.0794,
+      "step": 123
+    },
+    {
+      "epoch": 0.007627952755905512,
+      "grad_norm": 1.0202386379241943,
+      "learning_rate": 0.00019997570208735315,
+      "loss": 4.1264,
+      "step": 124
+    },
+    {
+      "epoch": 0.007689468503937008,
+      "grad_norm": 1.0154190063476562,
+      "learning_rate": 0.0001999752739561833,
+      "loss": 4.2101,
+      "step": 125
+    },
+    {
+      "epoch": 0.007750984251968504,
+      "grad_norm": 1.521721363067627,
+      "learning_rate": 0.00019997484208649576,
+      "loss": 3.69,
+      "step": 126
+    },
+    {
+      "epoch": 0.0078125,
+      "grad_norm": 1.2424317598342896,
+      "learning_rate": 0.00019997440647830653,
+      "loss": 3.8559,
+      "step": 127
+    },
+    {
+      "epoch": 0.007874015748031496,
+      "grad_norm": 0.9340006113052368,
+      "learning_rate": 0.000199973967131632,
+      "loss": 4.1619,
+      "step": 128
+    },
+    {
+      "epoch": 0.007935531496062992,
+      "grad_norm": 1.0604355335235596,
+      "learning_rate": 0.00019997352404648857,
+      "loss": 3.4408,
+      "step": 129
+    },
+    {
+      "epoch": 0.007997047244094488,
+      "grad_norm": 1.0056222677230835,
+      "learning_rate": 0.0001999730772228928,
+      "loss": 3.3851,
+      "step": 130
+    },
+    {
+      "epoch": 0.008058562992125984,
+      "grad_norm": 0.7513048648834229,
+      "learning_rate": 0.00019997262666086142,
+      "loss": 4.2672,
+      "step": 131
+    },
+    {
+      "epoch": 0.00812007874015748,
+      "grad_norm": 1.3518322706222534,
+      "learning_rate": 0.0001999721723604113,
+      "loss": 4.161,
+      "step": 132
+    },
+    {
+      "epoch": 0.008181594488188976,
+      "grad_norm": 1.1235636472702026,
+      "learning_rate": 0.00019997171432155933,
+      "loss": 3.9851,
+      "step": 133
+    },
+    {
+      "epoch": 0.008243110236220472,
+      "grad_norm": 1.1121013164520264,
+      "learning_rate": 0.0001999712525443227,
+      "loss": 3.3369,
+      "step": 134
+    },
+    {
+      "epoch": 0.00830462598425197,
+      "grad_norm": 1.1015123128890991,
+      "learning_rate": 0.00019997078702871872,
+      "loss": 4.1392,
+      "step": 135
+    },
+    {
+      "epoch": 0.008366141732283465,
+      "grad_norm": 1.2108583450317383,
+      "learning_rate": 0.00019997031777476475,
+      "loss": 3.5958,
+      "step": 136
+    },
+    {
+      "epoch": 0.008427657480314961,
+      "grad_norm": 1.0681123733520508,
+      "learning_rate": 0.00019996984478247833,
+      "loss": 4.0924,
+      "step": 137
+    },
+    {
+      "epoch": 0.008489173228346457,
+      "grad_norm": 1.279208779335022,
+      "learning_rate": 0.00019996936805187718,
+      "loss": 3.7196,
+      "step": 138
+    },
+    {
+      "epoch": 0.008550688976377953,
+      "grad_norm": 1.3705270290374756,
+      "learning_rate": 0.00019996888758297909,
+      "loss": 4.3487,
+      "step": 139
+    },
+    {
+      "epoch": 0.00861220472440945,
+      "grad_norm": 0.8328034281730652,
+      "learning_rate": 0.00019996840337580207,
+      "loss": 3.9334,
+      "step": 140
+    },
+    {
+      "epoch": 0.008673720472440945,
+      "grad_norm": 1.0824898481369019,
+      "learning_rate": 0.0001999679154303642,
+      "loss": 3.9965,
+      "step": 141
+    },
+    {
+      "epoch": 0.008735236220472441,
+      "grad_norm": 1.020643711090088,
+      "learning_rate": 0.00019996742374668372,
+      "loss": 4.4019,
+      "step": 142
+    },
+    {
+      "epoch": 0.008796751968503937,
+      "grad_norm": 0.9669337868690491,
+      "learning_rate": 0.00019996692832477902,
+      "loss": 1.996,
+      "step": 143
+    },
+    {
+      "epoch": 0.008858267716535433,
+      "grad_norm": 0.9339452981948853,
+      "learning_rate": 0.0001999664291646686,
+      "loss": 4.0075,
+      "step": 144
+    },
+    {
+      "epoch": 0.008919783464566929,
+      "grad_norm": 0.9043839573860168,
+      "learning_rate": 0.0001999659262663712,
+      "loss": 3.9553,
+      "step": 145
+    },
+    {
+      "epoch": 0.008981299212598425,
+      "grad_norm": 0.9271431565284729,
+      "learning_rate": 0.00019996541962990559,
+      "loss": 3.8066,
+      "step": 146
+    },
+    {
+      "epoch": 0.00904281496062992,
+      "grad_norm": 1.2157026529312134,
+      "learning_rate": 0.0001999649092552907,
+      "loss": 4.9701,
+      "step": 147
+    },
+    {
+      "epoch": 0.009104330708661417,
+      "grad_norm": 0.9345269203186035,
+      "learning_rate": 0.00019996439514254562,
+      "loss": 3.8809,
+      "step": 148
+    },
+    {
+      "epoch": 0.009165846456692913,
+      "grad_norm": 1.0654510259628296,
+      "learning_rate": 0.00019996387729168954,
+      "loss": 3.9988,
+      "step": 149
+    },
+    {
+      "epoch": 0.009227362204724409,
+      "grad_norm": 1.3758782148361206,
+      "learning_rate": 0.00019996335570274192,
+      "loss": 4.3286,
+      "step": 150
+    },
+    {
+      "epoch": 0.009288877952755906,
+      "grad_norm": 1.23570716381073,
+      "learning_rate": 0.00019996283037572216,
+      "loss": 4.0428,
+      "step": 151
+    },
+    {
+      "epoch": 0.009350393700787402,
+      "grad_norm": 0.9689831733703613,
+      "learning_rate": 0.00019996230131065002,
+      "loss": 4.009,
+      "step": 152
+    },
+    {
+      "epoch": 0.009411909448818898,
+      "grad_norm": 0.9608110189437866,
+      "learning_rate": 0.00019996176850754516,
+      "loss": 3.4991,
+      "step": 153
+    },
+    {
+      "epoch": 0.009473425196850394,
+      "grad_norm": 1.1149580478668213,
+      "learning_rate": 0.00019996123196642756,
+      "loss": 3.2724,
+      "step": 154
+    },
+    {
+      "epoch": 0.00953494094488189,
+      "grad_norm": 1.0518015623092651,
+      "learning_rate": 0.00019996069168731732,
+      "loss": 3.8297,
+      "step": 155
+    },
+    {
+      "epoch": 0.009596456692913386,
+      "grad_norm": 1.4498533010482788,
+      "learning_rate": 0.00019996014767023458,
+      "loss": 4.0394,
+      "step": 156
+    },
+    {
+      "epoch": 0.009657972440944882,
+      "grad_norm": 0.9268041253089905,
+      "learning_rate": 0.0001999595999151997,
+      "loss": 3.5148,
+      "step": 157
+    },
+    {
+      "epoch": 0.009719488188976378,
+      "grad_norm": 0.9180638194084167,
+      "learning_rate": 0.0001999590484222332,
+      "loss": 3.4739,
+      "step": 158
+    },
+    {
+      "epoch": 0.009781003937007874,
+      "grad_norm": 1.0417743921279907,
+      "learning_rate": 0.00019995849319135567,
+      "loss": 4.1415,
+      "step": 159
+    },
+    {
+      "epoch": 0.00984251968503937,
+      "grad_norm": 0.8904036283493042,
+      "learning_rate": 0.00019995793422258786,
+      "loss": 4.0283,
+      "step": 160
+    },
+    {
+      "epoch": 0.009904035433070866,
+      "grad_norm": 1.0201314687728882,
+      "learning_rate": 0.00019995737151595072,
+      "loss": 4.0576,
+      "step": 161
+    },
+    {
+      "epoch": 0.009965551181102362,
+      "grad_norm": 1.109643578529358,
+      "learning_rate": 0.00019995680507146522,
+      "loss": 3.6832,
+      "step": 162
+    },
+    {
+      "epoch": 0.010027066929133858,
+      "grad_norm": 0.8940156102180481,
+      "learning_rate": 0.0001999562348891526,
+      "loss": 3.6519,
+      "step": 163
+    },
+    {
+      "epoch": 0.010088582677165354,
+      "grad_norm": 1.2682384252548218,
+      "learning_rate": 0.0001999556609690342,
+      "loss": 4.2614,
+      "step": 164
+    },
+    {
+      "epoch": 0.01015009842519685,
+      "grad_norm": 1.113534688949585,
+      "learning_rate": 0.0001999550833111314,
+      "loss": 3.2479,
+      "step": 165
+    },
+    {
+      "epoch": 0.010211614173228346,
+      "grad_norm": 1.5006425380706787,
+      "learning_rate": 0.00019995450191546588,
+      "loss": 3.6487,
+      "step": 166
+    },
+    {
+      "epoch": 0.010273129921259843,
+      "grad_norm": 1.1574208736419678,
+      "learning_rate": 0.00019995391678205934,
+      "loss": 3.6985,
+      "step": 167
+    },
+    {
+      "epoch": 0.01033464566929134,
+      "grad_norm": 0.9824999570846558,
+      "learning_rate": 0.0001999533279109337,
+      "loss": 3.7214,
+      "step": 168
+    },
+    {
+      "epoch": 0.010396161417322835,
+      "grad_norm": 0.9681886434555054,
+      "learning_rate": 0.0001999527353021109,
+      "loss": 3.8614,
+      "step": 169
+    },
+    {
+      "epoch": 0.010457677165354331,
+      "grad_norm": 1.0207579135894775,
+      "learning_rate": 0.0001999521389556132,
+      "loss": 3.7751,
+      "step": 170
+    },
+    {
+      "epoch": 0.010519192913385827,
+      "grad_norm": 1.0156910419464111,
+      "learning_rate": 0.00019995153887146283,
+      "loss": 3.5665,
+      "step": 171
+    },
+    {
+      "epoch": 0.010580708661417323,
+      "grad_norm": 1.0935839414596558,
+      "learning_rate": 0.00019995093504968225,
+      "loss": 3.7934,
+      "step": 172
+    },
+    {
+      "epoch": 0.010642224409448819,
+      "grad_norm": 1.0409120321273804,
+      "learning_rate": 0.00019995032749029403,
+      "loss": 3.5342,
+      "step": 173
+    },
+    {
+      "epoch": 0.010703740157480315,
+      "grad_norm": 1.3472939729690552,
+      "learning_rate": 0.0001999497161933209,
+      "loss": 3.4779,
+      "step": 174
+    },
+    {
+      "epoch": 0.010765255905511811,
+      "grad_norm": 1.0455843210220337,
+      "learning_rate": 0.00019994910115878574,
+      "loss": 3.9905,
+      "step": 175
+    },
+    {
+      "epoch": 0.010826771653543307,
+      "grad_norm": 1.5592814683914185,
+      "learning_rate": 0.00019994848238671153,
+      "loss": 3.5426,
+      "step": 176
+    },
+    {
+      "epoch": 0.010888287401574803,
+      "grad_norm": 1.3635982275009155,
+      "learning_rate": 0.0001999478598771214,
+      "loss": 3.264,
+      "step": 177
+    },
+    {
+      "epoch": 0.010949803149606299,
+      "grad_norm": 1.2004342079162598,
+      "learning_rate": 0.00019994723363003862,
+      "loss": 3.6888,
+      "step": 178
+    },
+    {
+      "epoch": 0.011011318897637795,
+      "grad_norm": 0.8708282113075256,
+      "learning_rate": 0.00019994660364548662,
+      "loss": 4.0702,
+      "step": 179
+    },
+    {
+      "epoch": 0.01107283464566929,
+      "grad_norm": 1.0471171140670776,
+      "learning_rate": 0.000199945969923489,
+      "loss": 3.7733,
+      "step": 180
+    },
+    {
+      "epoch": 0.011134350393700787,
+      "grad_norm": 1.2412266731262207,
+      "learning_rate": 0.0001999453324640694,
+      "loss": 3.9466,
+      "step": 181
+    },
+    {
+      "epoch": 0.011195866141732283,
+      "grad_norm": 1.2793896198272705,
+      "learning_rate": 0.00019994469126725166,
+      "loss": 3.7525,
+      "step": 182
+    },
+    {
+      "epoch": 0.01125738188976378,
+      "grad_norm": 1.0514830350875854,
+      "learning_rate": 0.00019994404633305978,
+      "loss": 3.9435,
+      "step": 183
+    },
+    {
+      "epoch": 0.011318897637795276,
+      "grad_norm": 1.147952675819397,
+      "learning_rate": 0.00019994339766151787,
+      "loss": 3.918,
+      "step": 184
+    },
+    {
+      "epoch": 0.011380413385826772,
+      "grad_norm": 0.9058355689048767,
+      "learning_rate": 0.0001999427452526502,
+      "loss": 3.9508,
+      "step": 185
+    },
+    {
+      "epoch": 0.011441929133858268,
+      "grad_norm": 1.121683955192566,
+      "learning_rate": 0.00019994208910648114,
+      "loss": 3.957,
+      "step": 186
+    },
+    {
+      "epoch": 0.011503444881889764,
+      "grad_norm": 1.3012406826019287,
+      "learning_rate": 0.00019994142922303524,
+      "loss": 3.0095,
+      "step": 187
+    },
+    {
+      "epoch": 0.01156496062992126,
+      "grad_norm": 1.4108035564422607,
+      "learning_rate": 0.00019994076560233718,
+      "loss": 3.6609,
+      "step": 188
+    },
+    {
+      "epoch": 0.011626476377952756,
+      "grad_norm": 0.9696645736694336,
+      "learning_rate": 0.00019994009824441178,
+      "loss": 3.6165,
+      "step": 189
+    },
+    {
+      "epoch": 0.011687992125984252,
+      "grad_norm": 1.013260006904602,
+      "learning_rate": 0.00019993942714928396,
+      "loss": 3.7161,
+      "step": 190
+    },
+    {
+      "epoch": 0.011749507874015748,
+      "grad_norm": 1.1869155168533325,
+      "learning_rate": 0.00019993875231697888,
+      "loss": 2.2821,
+      "step": 191
+    },
+    {
+      "epoch": 0.011811023622047244,
+      "grad_norm": 0.9423443675041199,
+      "learning_rate": 0.0001999380737475217,
+      "loss": 3.1269,
+      "step": 192
+    },
+    {
+      "epoch": 0.01187253937007874,
+      "grad_norm": 1.3786706924438477,
+      "learning_rate": 0.00019993739144093787,
+      "loss": 3.5943,
+      "step": 193
+    },
+    {
+      "epoch": 0.011934055118110236,
+      "grad_norm": 1.2996402978897095,
+      "learning_rate": 0.00019993670539725282,
+      "loss": 4.1292,
+      "step": 194
+    },
+    {
+      "epoch": 0.011995570866141732,
+      "grad_norm": 0.8906294703483582,
+      "learning_rate": 0.0001999360156164923,
+      "loss": 3.7931,
+      "step": 195
+    },
+    {
+      "epoch": 0.012057086614173228,
+      "grad_norm": 0.9668493866920471,
+      "learning_rate": 0.00019993532209868203,
+      "loss": 3.7264,
+      "step": 196
+    },
+    {
+      "epoch": 0.012118602362204724,
+      "grad_norm": 1.087526798248291,
+      "learning_rate": 0.00019993462484384797,
+      "loss": 3.5555,
+      "step": 197
+    },
+    {
+      "epoch": 0.01218011811023622,
+      "grad_norm": 1.9559763669967651,
+      "learning_rate": 0.0001999339238520162,
+      "loss": 3.7755,
+      "step": 198
+    },
+    {
+      "epoch": 0.012241633858267717,
+      "grad_norm": 0.9076279401779175,
+      "learning_rate": 0.00019993321912321289,
+      "loss": 3.5687,
+      "step": 199
+    },
+    {
+      "epoch": 0.012303149606299213,
+      "grad_norm": 1.1451313495635986,
+      "learning_rate": 0.00019993251065746444,
+      "loss": 3.5538,
+      "step": 200
+    },
+    {
+      "epoch": 0.01236466535433071,
+      "grad_norm": 0.9248700141906738,
+      "learning_rate": 0.00019993179845479735,
+      "loss": 3.4582,
+      "step": 201
+    },
+    {
+      "epoch": 0.012426181102362205,
+      "grad_norm": 2.2651188373565674,
+      "learning_rate": 0.00019993108251523825,
+      "loss": 3.0879,
+      "step": 202
+    },
+    {
+      "epoch": 0.012487696850393701,
+      "grad_norm": 0.9576929211616516,
+      "learning_rate": 0.00019993036283881386,
+      "loss": 3.2448,
+      "step": 203
+    },
+    {
+      "epoch": 0.012549212598425197,
+      "grad_norm": 1.1414614915847778,
+      "learning_rate": 0.0001999296394255511,
+      "loss": 4.0959,
+      "step": 204
+    },
+    {
+      "epoch": 0.012610728346456693,
+      "grad_norm": 0.7668119072914124,
+      "learning_rate": 0.00019992891227547713,
+      "loss": 3.7599,
+      "step": 205
+    },
+    {
+      "epoch": 0.012672244094488189,
+      "grad_norm": 0.9080313444137573,
+      "learning_rate": 0.00019992818138861898,
+      "loss": 4.0564,
+      "step": 206
+    },
+    {
+      "epoch": 0.012733759842519685,
+      "grad_norm": 0.868075430393219,
+      "learning_rate": 0.00019992744676500411,
+      "loss": 3.4342,
+      "step": 207
+    },
+    {
+      "epoch": 0.012795275590551181,
+      "grad_norm": 1.5771456956863403,
+      "learning_rate": 0.00019992670840465995,
+      "loss": 3.7856,
+      "step": 208
+    },
+    {
+      "epoch": 0.012856791338582677,
+      "grad_norm": 1.1196974515914917,
+      "learning_rate": 0.00019992596630761405,
+      "loss": 3.8565,
+      "step": 209
+    },
+    {
+      "epoch": 0.012918307086614173,
+      "grad_norm": 1.1127508878707886,
+      "learning_rate": 0.00019992522047389426,
+      "loss": 4.0551,
+      "step": 210
+    },
+    {
+      "epoch": 0.012979822834645669,
+      "grad_norm": 1.225761890411377,
+      "learning_rate": 0.00019992447090352842,
+      "loss": 3.6294,
+      "step": 211
+    },
+    {
+      "epoch": 0.013041338582677165,
+      "grad_norm": 0.9076347947120667,
+      "learning_rate": 0.00019992371759654454,
+      "loss": 3.8761,
+      "step": 212
+    },
+    {
+      "epoch": 0.01310285433070866,
+      "grad_norm": 1.4209147691726685,
+      "learning_rate": 0.00019992296055297082,
+      "loss": 3.6056,
+      "step": 213
+    },
+    {
+      "epoch": 0.013164370078740157,
+      "grad_norm": 1.5463988780975342,
+      "learning_rate": 0.00019992219977283554,
+      "loss": 3.3891,
+      "step": 214
+    },
+    {
+      "epoch": 0.013225885826771654,
+      "grad_norm": 1.8705159425735474,
+      "learning_rate": 0.00019992143525616722,
+      "loss": 4.2649,
+      "step": 215
+    },
+    {
+      "epoch": 0.01328740157480315,
+      "grad_norm": 1.8354461193084717,
+      "learning_rate": 0.00019992066700299436,
+      "loss": 3.3537,
+      "step": 216
+    },
+    {
+      "epoch": 0.013348917322834646,
+      "grad_norm": 1.284125804901123,
+      "learning_rate": 0.00019991989501334573,
+      "loss": 4.0981,
+      "step": 217
+    },
+    {
+      "epoch": 0.013410433070866142,
+      "grad_norm": 0.8678949475288391,
+      "learning_rate": 0.00019991911928725022,
+      "loss": 3.6141,
+      "step": 218
+    },
+    {
+      "epoch": 0.013471948818897638,
+      "grad_norm": 0.9543637037277222,
+      "learning_rate": 0.0001999183398247368,
+      "loss": 3.8767,
+      "step": 219
+    },
+    {
+      "epoch": 0.013533464566929134,
+      "grad_norm": 1.0744081735610962,
+      "learning_rate": 0.00019991755662583462,
+      "loss": 3.6454,
+      "step": 220
+    },
+    {
+      "epoch": 0.01359498031496063,
+      "grad_norm": 1.1265289783477783,
+      "learning_rate": 0.00019991676969057298,
+      "loss": 3.771,
+      "step": 221
+    },
+    {
+      "epoch": 0.013656496062992126,
+      "grad_norm": 0.8677342534065247,
+      "learning_rate": 0.00019991597901898128,
+      "loss": 4.0227,
+      "step": 222
+    },
+    {
+      "epoch": 0.013718011811023622,
+      "grad_norm": 0.6366098523139954,
+      "learning_rate": 0.00019991518461108916,
+      "loss": 3.8226,
+      "step": 223
+    },
+    {
+      "epoch": 0.013779527559055118,
+      "grad_norm": 0.9632430076599121,
+      "learning_rate": 0.00019991438646692626,
+      "loss": 3.6245,
+      "step": 224
+    },
+    {
+      "epoch": 0.013841043307086614,
+      "grad_norm": 0.8582671880722046,
+      "learning_rate": 0.00019991358458652244,
+      "loss": 3.3309,
+      "step": 225
+    },
+    {
+      "epoch": 0.01390255905511811,
+      "grad_norm": 0.880537211894989,
+      "learning_rate": 0.0001999127789699077,
+      "loss": 3.7726,
+      "step": 226
+    },
+    {
+      "epoch": 0.013964074803149606,
+      "grad_norm": 0.8238226175308228,
+      "learning_rate": 0.00019991196961711215,
+      "loss": 3.7752,
+      "step": 227
+    },
+    {
+      "epoch": 0.014025590551181102,
+      "grad_norm": 1.1190179586410522,
+      "learning_rate": 0.00019991115652816604,
+      "loss": 3.5584,
+      "step": 228
+    },
+    {
+      "epoch": 0.014087106299212598,
+      "grad_norm": 0.866181492805481,
+      "learning_rate": 0.00019991033970309983,
+      "loss": 3.4659,
+      "step": 229
+    },
+    {
+      "epoch": 0.014148622047244094,
+      "grad_norm": 0.858160138130188,
+      "learning_rate": 0.000199909519141944,
+      "loss": 4.1523,
+      "step": 230
+    },
+    {
+      "epoch": 0.014210137795275591,
+      "grad_norm": 1.3982963562011719,
+      "learning_rate": 0.0001999086948447293,
+      "loss": 3.3695,
+      "step": 231
+    },
+    {
+      "epoch": 0.014271653543307087,
+      "grad_norm": 0.8196374773979187,
+      "learning_rate": 0.00019990786681148648,
+      "loss": 3.5243,
+      "step": 232
+    },
+    {
+      "epoch": 0.014333169291338583,
+      "grad_norm": 1.0363407135009766,
+      "learning_rate": 0.00019990703504224654,
+      "loss": 4.0894,
+      "step": 233
+    },
+    {
+      "epoch": 0.01439468503937008,
+      "grad_norm": 0.8233461976051331,
+      "learning_rate": 0.00019990619953704064,
+      "loss": 3.8317,
+      "step": 234
+    },
+    {
+      "epoch": 0.014456200787401575,
+      "grad_norm": 1.2458850145339966,
+      "learning_rate": 0.00019990536029589992,
+      "loss": 3.5802,
+      "step": 235
+    },
+    {
+      "epoch": 0.014517716535433071,
+      "grad_norm": 1.1212568283081055,
+      "learning_rate": 0.00019990451731885585,
+      "loss": 4.3651,
+      "step": 236
+    },
+    {
+      "epoch": 0.014579232283464567,
+      "grad_norm": 1.1830307245254517,
+      "learning_rate": 0.0001999036706059399,
+      "loss": 3.7283,
+      "step": 237
+    },
+    {
+      "epoch": 0.014640748031496063,
+      "grad_norm": 0.9320050477981567,
+      "learning_rate": 0.00019990282015718372,
+      "loss": 3.5635,
+      "step": 238
+    },
+    {
+      "epoch": 0.014702263779527559,
+      "grad_norm": 0.7344046831130981,
+      "learning_rate": 0.00019990196597261916,
+      "loss": 3.3528,
+      "step": 239
+    },
+    {
+      "epoch": 0.014763779527559055,
+      "grad_norm": 0.9713966846466064,
+      "learning_rate": 0.00019990110805227814,
+      "loss": 3.745,
+      "step": 240
+    },
+    {
+      "epoch": 0.01482529527559055,
+      "grad_norm": 1.123129963874817,
+      "learning_rate": 0.00019990024639619276,
+      "loss": 3.2692,
+      "step": 241
+    },
+    {
+      "epoch": 0.014886811023622047,
+      "grad_norm": 0.890461266040802,
+      "learning_rate": 0.0001998993810043952,
+      "loss": 3.54,
+      "step": 242
+    },
+    {
+      "epoch": 0.014948326771653543,
+      "grad_norm": 1.2220909595489502,
+      "learning_rate": 0.00019989851187691784,
+      "loss": 3.6403,
+      "step": 243
+    },
+    {
+      "epoch": 0.015009842519685039,
+      "grad_norm": 0.7764825224876404,
+      "learning_rate": 0.0001998976390137932,
+      "loss": 3.6712,
+      "step": 244
+    },
+    {
+      "epoch": 0.015071358267716535,
+      "grad_norm": 1.2578058242797852,
+      "learning_rate": 0.0001998967624150539,
+      "loss": 3.4161,
+      "step": 245
+    },
+    {
+      "epoch": 0.01513287401574803,
+      "grad_norm": 0.8289483189582825,
+      "learning_rate": 0.00019989588208073274,
+      "loss": 3.8457,
+      "step": 246
+    },
+    {
+      "epoch": 0.015194389763779528,
+      "grad_norm": 1.3260377645492554,
+      "learning_rate": 0.0001998949980108626,
+      "loss": 4.1059,
+      "step": 247
+    },
+    {
+      "epoch": 0.015255905511811024,
+      "grad_norm": 0.9355643391609192,
+      "learning_rate": 0.0001998941102054766,
+      "loss": 3.7215,
+      "step": 248
+    },
+    {
+      "epoch": 0.01531742125984252,
+      "grad_norm": 0.8024367094039917,
+      "learning_rate": 0.00019989321866460786,
+      "loss": 3.727,
+      "step": 249
+    },
+    {
+      "epoch": 0.015378937007874016,
+      "grad_norm": 1.3206759691238403,
+      "learning_rate": 0.0001998923233882898,
+      "loss": 3.7687,
+      "step": 250
+    },
+    {
+      "epoch": 0.015440452755905512,
+      "grad_norm": 0.9962854981422424,
+      "learning_rate": 0.00019989142437655584,
+      "loss": 3.9778,
+      "step": 251
+    },
+    {
+      "epoch": 0.015501968503937008,
+      "grad_norm": 1.0723034143447876,
+      "learning_rate": 0.00019989052162943963,
+      "loss": 3.8162,
+      "step": 252
+    },
+    {
+      "epoch": 0.015563484251968504,
+      "grad_norm": 0.8389244079589844,
+      "learning_rate": 0.0001998896151469749,
+      "loss": 3.7489,
+      "step": 253
+    },
+    {
+      "epoch": 0.015625,
+      "grad_norm": 1.3018784523010254,
+      "learning_rate": 0.0001998887049291956,
+      "loss": 3.7989,
+      "step": 254
+    },
+    {
+      "epoch": 0.015686515748031496,
+      "grad_norm": 0.907413899898529,
+      "learning_rate": 0.00019988779097613568,
+      "loss": 3.3219,
+      "step": 255
+    },
+    {
+      "epoch": 0.015748031496062992,
+      "grad_norm": 1.1340125799179077,
+      "learning_rate": 0.00019988687328782937,
+      "loss": 3.2498,
+      "step": 256
+    },
+    {
+      "epoch": 0.015809547244094488,
+      "grad_norm": 0.7439151406288147,
+      "learning_rate": 0.000199885951864311,
+      "loss": 3.3099,
+      "step": 257
+    },
+    {
+      "epoch": 0.015871062992125984,
+      "grad_norm": 0.9544551968574524,
+      "learning_rate": 0.00019988502670561503,
+      "loss": 3.4564,
+      "step": 258
+    },
+    {
+      "epoch": 0.01593257874015748,
+      "grad_norm": 1.8259236812591553,
+      "learning_rate": 0.000199884097811776,
+      "loss": 2.8979,
+      "step": 259
+    },
+    {
+      "epoch": 0.015994094488188976,
+      "grad_norm": 1.1474432945251465,
+      "learning_rate": 0.0001998831651828287,
+      "loss": 3.522,
+      "step": 260
+    },
+    {
+      "epoch": 0.01605561023622047,
+      "grad_norm": 0.9711191058158875,
+      "learning_rate": 0.00019988222881880797,
+      "loss": 3.7704,
+      "step": 261
+    },
+    {
+      "epoch": 0.016117125984251968,
+      "grad_norm": 1.0646536350250244,
+      "learning_rate": 0.00019988128871974888,
+      "loss": 4.2837,
+      "step": 262
+    },
+    {
+      "epoch": 0.016178641732283464,
+      "grad_norm": 0.8561193943023682,
+      "learning_rate": 0.00019988034488568651,
+      "loss": 3.5426,
+      "step": 263
+    },
+    {
+      "epoch": 0.01624015748031496,
+      "grad_norm": 0.9443851709365845,
+      "learning_rate": 0.0001998793973166562,
+      "loss": 4.1207,
+      "step": 264
+    },
+    {
+      "epoch": 0.016301673228346455,
+      "grad_norm": 1.593692660331726,
+      "learning_rate": 0.00019987844601269337,
+      "loss": 2.5418,
+      "step": 265
+    },
+    {
+      "epoch": 0.01636318897637795,
+      "grad_norm": 25.340497970581055,
+      "learning_rate": 0.00019987749097383358,
+      "loss": 3.4535,
+      "step": 266
+    },
+    {
+      "epoch": 0.016424704724409447,
+      "grad_norm": 0.8485792279243469,
+      "learning_rate": 0.0001998765322001126,
+      "loss": 3.8567,
+      "step": 267
+    },
+    {
+      "epoch": 0.016486220472440943,
+      "grad_norm": 1.0135265588760376,
+      "learning_rate": 0.0001998755696915662,
+      "loss": 4.1648,
+      "step": 268
+    },
+    {
+      "epoch": 0.01654773622047244,
+      "grad_norm": 0.9357320666313171,
+      "learning_rate": 0.00019987460344823044,
+      "loss": 4.1427,
+      "step": 269
+    },
+    {
+      "epoch": 0.01660925196850394,
+      "grad_norm": 1.4145742654800415,
+      "learning_rate": 0.00019987363347014143,
+      "loss": 3.8124,
+      "step": 270
+    },
+    {
+      "epoch": 0.016670767716535435,
+      "grad_norm": 0.7826442718505859,
+      "learning_rate": 0.00019987265975733543,
+      "loss": 4.1073,
+      "step": 271
+    },
+    {
+      "epoch": 0.01673228346456693,
+      "grad_norm": 0.8924596905708313,
+      "learning_rate": 0.00019987168230984885,
+      "loss": 3.3501,
+      "step": 272
+    },
+    {
+      "epoch": 0.016793799212598427,
+      "grad_norm": 1.1438568830490112,
+      "learning_rate": 0.0001998707011277183,
+      "loss": 3.3166,
+      "step": 273
+    },
+    {
+      "epoch": 0.016855314960629923,
+      "grad_norm": 0.774922788143158,
+      "learning_rate": 0.00019986971621098037,
+      "loss": 3.6741,
+      "step": 274
+    },
+    {
+      "epoch": 0.01691683070866142,
+      "grad_norm": 1.11615788936615,
+      "learning_rate": 0.00019986872755967195,
+      "loss": 3.7729,
+      "step": 275
+    },
+    {
+      "epoch": 0.016978346456692914,
+      "grad_norm": 1.7372264862060547,
+      "learning_rate": 0.00019986773517383002,
+      "loss": 3.377,
+      "step": 276
+    },
+    {
+      "epoch": 0.01703986220472441,
+      "grad_norm": 1.5627361536026,
+      "learning_rate": 0.00019986673905349168,
+      "loss": 3.2416,
+      "step": 277
+    },
+    {
+      "epoch": 0.017101377952755906,
+      "grad_norm": 1.200598955154419,
+      "learning_rate": 0.00019986573919869414,
+      "loss": 3.7952,
+      "step": 278
+    },
+    {
+      "epoch": 0.017162893700787402,
+      "grad_norm": 1.037914514541626,
+      "learning_rate": 0.00019986473560947482,
+      "loss": 2.9243,
+      "step": 279
+    },
+    {
+      "epoch": 0.0172244094488189,
+      "grad_norm": 1.046506404876709,
+      "learning_rate": 0.00019986372828587125,
+      "loss": 3.8008,
+      "step": 280
+    },
+    {
+      "epoch": 0.017285925196850394,
+      "grad_norm": 1.192258596420288,
+      "learning_rate": 0.00019986271722792113,
+      "loss": 3.2582,
+      "step": 281
+    },
+    {
+      "epoch": 0.01734744094488189,
+      "grad_norm": 0.8760639429092407,
+      "learning_rate": 0.00019986170243566222,
+      "loss": 3.3292,
+      "step": 282
+    },
+    {
+      "epoch": 0.017408956692913386,
+      "grad_norm": 1.1775727272033691,
+      "learning_rate": 0.00019986068390913248,
+      "loss": 3.4438,
+      "step": 283
+    },
+    {
+      "epoch": 0.017470472440944882,
+      "grad_norm": 1.1048696041107178,
+      "learning_rate": 0.00019985966164837,
+      "loss": 4.091,
+      "step": 284
+    },
+    {
+      "epoch": 0.017531988188976378,
+      "grad_norm": 1.123031497001648,
+      "learning_rate": 0.000199858635653413,
+      "loss": 3.7858,
+      "step": 285
+    },
+    {
+      "epoch": 0.017593503937007874,
+      "grad_norm": 1.103630542755127,
+      "learning_rate": 0.00019985760592429986,
+      "loss": 3.1255,
+      "step": 286
+    },
+    {
+      "epoch": 0.01765501968503937,
+      "grad_norm": 1.1980113983154297,
+      "learning_rate": 0.00019985657246106907,
+      "loss": 3.2937,
+      "step": 287
+    },
+    {
+      "epoch": 0.017716535433070866,
+      "grad_norm": 1.1340224742889404,
+      "learning_rate": 0.00019985553526375932,
+      "loss": 4.2223,
+      "step": 288
+    },
+    {
+      "epoch": 0.017778051181102362,
+      "grad_norm": 0.8204386234283447,
+      "learning_rate": 0.00019985449433240936,
+      "loss": 3.5508,
+      "step": 289
+    },
+    {
+      "epoch": 0.017839566929133858,
+      "grad_norm": 0.9163397550582886,
+      "learning_rate": 0.00019985344966705806,
+      "loss": 3.9192,
+      "step": 290
+    },
+    {
+      "epoch": 0.017901082677165354,
+      "grad_norm": 1.526183843612671,
+      "learning_rate": 0.00019985240126774457,
+      "loss": 3.8117,
+      "step": 291
+    },
+    {
+      "epoch": 0.01796259842519685,
+      "grad_norm": 1.1046873331069946,
+      "learning_rate": 0.00019985134913450808,
+      "loss": 3.5415,
+      "step": 292
+    },
+    {
+      "epoch": 0.018024114173228346,
+      "grad_norm": 1.26083505153656,
+      "learning_rate": 0.0001998502932673879,
+      "loss": 3.5163,
+      "step": 293
+    },
+    {
+      "epoch": 0.01808562992125984,
+      "grad_norm": 1.722540259361267,
+      "learning_rate": 0.00019984923366642355,
+      "loss": 2.9612,
+      "step": 294
+    },
+    {
+      "epoch": 0.018147145669291338,
+      "grad_norm": 0.95418781042099,
+      "learning_rate": 0.00019984817033165463,
+      "loss": 3.6415,
+      "step": 295
+    },
+    {
+      "epoch": 0.018208661417322834,
+      "grad_norm": 1.107298731803894,
+      "learning_rate": 0.00019984710326312089,
+      "loss": 3.7632,
+      "step": 296
+    },
+    {
+      "epoch": 0.01827017716535433,
+      "grad_norm": 0.9547476768493652,
+      "learning_rate": 0.00019984603246086226,
+      "loss": 4.0912,
+      "step": 297
+    },
+    {
+      "epoch": 0.018331692913385825,
+      "grad_norm": 0.9447423219680786,
+      "learning_rate": 0.00019984495792491877,
+      "loss": 3.4904,
+      "step": 298
+    },
+    {
+      "epoch": 0.01839320866141732,
+      "grad_norm": 0.9499867558479309,
+      "learning_rate": 0.00019984387965533062,
+      "loss": 3.532,
+      "step": 299
+    },
+    {
+      "epoch": 0.018454724409448817,
+      "grad_norm": 0.9343771934509277,
+      "learning_rate": 0.00019984279765213812,
+      "loss": 4.0632,
+      "step": 300
+    },
+    {
+      "epoch": 0.018516240157480313,
+      "grad_norm": 0.9678671360015869,
+      "learning_rate": 0.0001998417119153817,
+      "loss": 3.5156,
+      "step": 301
+    },
+    {
+      "epoch": 0.018577755905511813,
+      "grad_norm": 1.4102838039398193,
+      "learning_rate": 0.000199840622445102,
+      "loss": 3.9054,
+      "step": 302
+    },
+    {
+      "epoch": 0.01863927165354331,
+      "grad_norm": 0.9135202169418335,
+      "learning_rate": 0.00019983952924133974,
+      "loss": 3.8946,
+      "step": 303
+    },
+    {
+      "epoch": 0.018700787401574805,
+      "grad_norm": 0.9663935303688049,
+      "learning_rate": 0.00019983843230413583,
+      "loss": 3.794,
+      "step": 304
+    },
+    {
+      "epoch": 0.0187623031496063,
+      "grad_norm": 1.2183406352996826,
+      "learning_rate": 0.00019983733163353126,
+      "loss": 3.2445,
+      "step": 305
+    },
+    {
+      "epoch": 0.018823818897637797,
+      "grad_norm": 1.4783687591552734,
+      "learning_rate": 0.00019983622722956717,
+      "loss": 3.4738,
+      "step": 306
+    },
+    {
+      "epoch": 0.018885334645669292,
+      "grad_norm": 1.125974178314209,
+      "learning_rate": 0.00019983511909228493,
+      "loss": 3.6785,
+      "step": 307
+    },
+    {
+      "epoch": 0.01894685039370079,
+      "grad_norm": 0.9093533158302307,
+      "learning_rate": 0.00019983400722172587,
+      "loss": 4.0304,
+      "step": 308
+    },
+    {
+      "epoch": 0.019008366141732284,
+      "grad_norm": 0.9175082445144653,
+      "learning_rate": 0.00019983289161793166,
+      "loss": 4.0229,
+      "step": 309
+    },
+    {
+      "epoch": 0.01906988188976378,
+      "grad_norm": 1.046862244606018,
+      "learning_rate": 0.00019983177228094398,
+      "loss": 3.573,
+      "step": 310
+    },
+    {
+      "epoch": 0.019131397637795276,
+      "grad_norm": 1.2958662509918213,
+      "learning_rate": 0.00019983064921080473,
+      "loss": 4.438,
+      "step": 311
+    },
+    {
+      "epoch": 0.019192913385826772,
+      "grad_norm": 0.9069321751594543,
+      "learning_rate": 0.00019982952240755583,
+      "loss": 3.9017,
+      "step": 312
+    },
+    {
+      "epoch": 0.019254429133858268,
+      "grad_norm": 0.9694398045539856,
+      "learning_rate": 0.00019982839187123946,
+      "loss": 3.6537,
+      "step": 313
+    },
+    {
+      "epoch": 0.019315944881889764,
+      "grad_norm": 0.8946678042411804,
+      "learning_rate": 0.0001998272576018979,
+      "loss": 3.2072,
+      "step": 314
+    },
+    {
+      "epoch": 0.01937746062992126,
+      "grad_norm": 0.789013147354126,
+      "learning_rate": 0.00019982611959957355,
+      "loss": 3.6613,
+      "step": 315
+    },
+    {
+      "epoch": 0.019438976377952756,
+      "grad_norm": 1.3650214672088623,
+      "learning_rate": 0.000199824977864309,
+      "loss": 3.5059,
+      "step": 316
+    },
+    {
+      "epoch": 0.019500492125984252,
+      "grad_norm": 1.4598431587219238,
+      "learning_rate": 0.0001998238323961469,
+      "loss": 3.9713,
+      "step": 317
+    },
+    {
+      "epoch": 0.019562007874015748,
+      "grad_norm": 0.9690242409706116,
+      "learning_rate": 0.00019982268319513008,
+      "loss": 3.3819,
+      "step": 318
+    },
+    {
+      "epoch": 0.019623523622047244,
+      "grad_norm": 1.1014673709869385,
+      "learning_rate": 0.00019982153026130157,
+      "loss": 3.7626,
+      "step": 319
+    },
+    {
+      "epoch": 0.01968503937007874,
+      "grad_norm": 1.0226904153823853,
+      "learning_rate": 0.00019982037359470446,
+      "loss": 3.4185,
+      "step": 320
+    },
+    {
+      "epoch": 0.019746555118110236,
+      "grad_norm": 0.8130924105644226,
+      "learning_rate": 0.00019981921319538195,
+      "loss": 3.9947,
+      "step": 321
+    },
+    {
+      "epoch": 0.019808070866141732,
+      "grad_norm": 1.3799322843551636,
+      "learning_rate": 0.00019981804906337746,
+      "loss": 4.3647,
+      "step": 322
+    },
+    {
+      "epoch": 0.019869586614173228,
+      "grad_norm": 0.7105659246444702,
+      "learning_rate": 0.00019981688119873457,
+      "loss": 3.5173,
+      "step": 323
+    },
+    {
+      "epoch": 0.019931102362204724,
+      "grad_norm": 1.022532343864441,
+      "learning_rate": 0.00019981570960149692,
+      "loss": 3.4015,
+      "step": 324
+    },
+    {
+      "epoch": 0.01999261811023622,
+      "grad_norm": 0.7684879302978516,
+      "learning_rate": 0.00019981453427170832,
+      "loss": 3.6504,
+      "step": 325
+    },
+    {
+      "epoch": 0.020054133858267716,
+      "grad_norm": 0.9662853479385376,
+      "learning_rate": 0.0001998133552094127,
+      "loss": 3.4114,
+      "step": 326
+    },
+    {
+      "epoch": 0.02011564960629921,
+      "grad_norm": 1.00538969039917,
+      "learning_rate": 0.00019981217241465417,
+      "loss": 4.0655,
+      "step": 327
+    },
+    {
+      "epoch": 0.020177165354330708,
+      "grad_norm": 0.9725746512413025,
+      "learning_rate": 0.000199810985887477,
+      "loss": 3.6332,
+      "step": 328
+    },
+    {
+      "epoch": 0.020238681102362203,
+      "grad_norm": 0.9656527638435364,
+      "learning_rate": 0.00019980979562792545,
+      "loss": 3.7718,
+      "step": 329
+    },
+    {
+      "epoch": 0.0203001968503937,
+      "grad_norm": 1.1181968450546265,
+      "learning_rate": 0.00019980860163604415,
+      "loss": 2.9625,
+      "step": 330
+    },
+    {
+      "epoch": 0.020361712598425195,
+      "grad_norm": 1.0991274118423462,
+      "learning_rate": 0.00019980740391187767,
+      "loss": 3.7054,
+      "step": 331
+    },
+    {
+      "epoch": 0.02042322834645669,
+      "grad_norm": 1.1337770223617554,
+      "learning_rate": 0.00019980620245547086,
+      "loss": 3.8945,
+      "step": 332
+    },
+    {
+      "epoch": 0.020484744094488187,
+      "grad_norm": 1.1334925889968872,
+      "learning_rate": 0.00019980499726686857,
+      "loss": 3.3509,
+      "step": 333
+    },
+    {
+      "epoch": 0.020546259842519687,
+      "grad_norm": 0.7692652940750122,
+      "learning_rate": 0.00019980378834611595,
+      "loss": 3.383,
+      "step": 334
+    },
+    {
+      "epoch": 0.020607775590551183,
+      "grad_norm": 0.9062781929969788,
+      "learning_rate": 0.00019980257569325813,
+      "loss": 4.0356,
+      "step": 335
+    },
+    {
+      "epoch": 0.02066929133858268,
+      "grad_norm": 0.8026053309440613,
+      "learning_rate": 0.0001998013593083405,
+      "loss": 3.7517,
+      "step": 336
+    },
+    {
+      "epoch": 0.020730807086614175,
+      "grad_norm": 0.9178938269615173,
+      "learning_rate": 0.00019980013919140857,
+      "loss": 3.6045,
+      "step": 337
+    },
+    {
+      "epoch": 0.02079232283464567,
+      "grad_norm": 0.6971200108528137,
+      "learning_rate": 0.00019979891534250793,
+      "loss": 3.275,
+      "step": 338
+    },
+    {
+      "epoch": 0.020853838582677166,
+      "grad_norm": 0.8424512147903442,
+      "learning_rate": 0.00019979768776168434,
+      "loss": 3.4437,
+      "step": 339
+    },
+    {
+      "epoch": 0.020915354330708662,
+      "grad_norm": 0.9663639068603516,
+      "learning_rate": 0.0001997964564489837,
+      "loss": 3.714,
+      "step": 340
+    },
+    {
+      "epoch": 0.02097687007874016,
+      "grad_norm": 0.9225979447364807,
+      "learning_rate": 0.00019979522140445207,
+      "loss": 3.7337,
+      "step": 341
+    },
+    {
+      "epoch": 0.021038385826771654,
+      "grad_norm": 0.8167860507965088,
+      "learning_rate": 0.00019979398262813565,
+      "loss": 3.2983,
+      "step": 342
+    },
+    {
+      "epoch": 0.02109990157480315,
+      "grad_norm": 1.0638161897659302,
+      "learning_rate": 0.00019979274012008074,
+      "loss": 3.2327,
+      "step": 343
+    },
+    {
+      "epoch": 0.021161417322834646,
+      "grad_norm": 2.666646718978882,
+      "learning_rate": 0.0001997914938803338,
+      "loss": 3.4651,
+      "step": 344
+    },
+    {
+      "epoch": 0.021222933070866142,
+      "grad_norm": 1.051708698272705,
+      "learning_rate": 0.00019979024390894145,
+      "loss": 3.3001,
+      "step": 345
+    },
+    {
+      "epoch": 0.021284448818897638,
+      "grad_norm": 1.2075786590576172,
+      "learning_rate": 0.00019978899020595044,
+      "loss": 3.0755,
+      "step": 346
+    },
+    {
+      "epoch": 0.021345964566929134,
+      "grad_norm": 0.9660154581069946,
+      "learning_rate": 0.00019978773277140762,
+      "loss": 3.6987,
+      "step": 347
+    },
+    {
+      "epoch": 0.02140748031496063,
+      "grad_norm": 1.011460542678833,
+      "learning_rate": 0.00019978647160536002,
+      "loss": 3.1498,
+      "step": 348
+    },
+    {
+      "epoch": 0.021468996062992126,
+      "grad_norm": 0.7530460357666016,
+      "learning_rate": 0.0001997852067078548,
+      "loss": 3.5957,
+      "step": 349
+    },
+    {
+      "epoch": 0.021530511811023622,
+      "grad_norm": 0.7773799896240234,
+      "learning_rate": 0.0001997839380789393,
+      "loss": 3.7692,
+      "step": 350
+    },
+    {
+      "epoch": 0.021592027559055118,
+      "grad_norm": 1.2352975606918335,
+      "learning_rate": 0.00019978266571866085,
+      "loss": 3.45,
+      "step": 351
+    },
+    {
+      "epoch": 0.021653543307086614,
+      "grad_norm": 0.9332252740859985,
+      "learning_rate": 0.00019978138962706719,
+      "loss": 3.843,
+      "step": 352
+    },
+    {
+      "epoch": 0.02171505905511811,
+      "grad_norm": 0.9313061237335205,
+      "learning_rate": 0.00019978010980420594,
+      "loss": 3.4785,
+      "step": 353
+    },
+    {
+      "epoch": 0.021776574803149606,
+      "grad_norm": 1.235709309577942,
+      "learning_rate": 0.00019977882625012494,
+      "loss": 3.1405,
+      "step": 354
+    },
+    {
+      "epoch": 0.0218380905511811,
+      "grad_norm": 1.6501309871673584,
+      "learning_rate": 0.00019977753896487224,
+      "loss": 2.7059,
+      "step": 355
+    },
+    {
+      "epoch": 0.021899606299212598,
+      "grad_norm": 0.91188645362854,
+      "learning_rate": 0.00019977624794849597,
+      "loss": 3.5843,
+      "step": 356
+    },
+    {
+      "epoch": 0.021961122047244094,
+      "grad_norm": 0.954845130443573,
+      "learning_rate": 0.00019977495320104436,
+      "loss": 3.5038,
+      "step": 357
+    },
+    {
+      "epoch": 0.02202263779527559,
+      "grad_norm": 0.875336229801178,
+      "learning_rate": 0.0001997736547225659,
+      "loss": 3.6924,
+      "step": 358
+    },
+    {
+      "epoch": 0.022084153543307086,
+      "grad_norm": 0.9856054782867432,
+      "learning_rate": 0.00019977235251310906,
+      "loss": 3.5056,
+      "step": 359
+    },
+    {
+      "epoch": 0.02214566929133858,
+      "grad_norm": 0.9993917346000671,
+      "learning_rate": 0.0001997710465727226,
+      "loss": 4.3623,
+      "step": 360
+    },
+    {
+      "epoch": 0.022207185039370077,
+      "grad_norm": 0.9177888631820679,
+      "learning_rate": 0.00019976973690145536,
+      "loss": 3.7133,
+      "step": 361
+    },
+    {
+      "epoch": 0.022268700787401573,
+      "grad_norm": 0.9861302375793457,
+      "learning_rate": 0.00019976842349935626,
+      "loss": 3.7547,
+      "step": 362
+    },
+    {
+      "epoch": 0.02233021653543307,
+      "grad_norm": 0.8839965462684631,
+      "learning_rate": 0.00019976710636647447,
+      "loss": 3.8602,
+      "step": 363
+    },
+    {
+      "epoch": 0.022391732283464565,
+      "grad_norm": 0.6840670108795166,
+      "learning_rate": 0.0001997657855028592,
+      "loss": 4.074,
+      "step": 364
+    },
+    {
+      "epoch": 0.02245324803149606,
+      "grad_norm": 1.1680982112884521,
+      "learning_rate": 0.00019976446090855986,
+      "loss": 3.8069,
+      "step": 365
+    },
+    {
+      "epoch": 0.02251476377952756,
+      "grad_norm": 1.023189902305603,
+      "learning_rate": 0.000199763132583626,
+      "loss": 3.3319,
+      "step": 366
+    },
+    {
+      "epoch": 0.022576279527559057,
+      "grad_norm": 0.847127377986908,
+      "learning_rate": 0.00019976180052810728,
+      "loss": 3.7336,
+      "step": 367
+    },
+    {
+      "epoch": 0.022637795275590553,
+      "grad_norm": 1.0050266981124878,
+      "learning_rate": 0.0001997604647420535,
+      "loss": 3.4195,
+      "step": 368
+    },
+    {
+      "epoch": 0.02269931102362205,
+      "grad_norm": 1.1433039903640747,
+      "learning_rate": 0.00019975912522551462,
+      "loss": 3.5512,
+      "step": 369
+    },
+    {
+      "epoch": 0.022760826771653545,
+      "grad_norm": 0.9511215090751648,
+      "learning_rate": 0.00019975778197854074,
+      "loss": 3.5853,
+      "step": 370
+    },
+    {
+      "epoch": 0.02282234251968504,
+      "grad_norm": 1.236533522605896,
+      "learning_rate": 0.00019975643500118204,
+      "loss": 3.954,
+      "step": 371
+    },
+    {
+      "epoch": 0.022883858267716536,
+      "grad_norm": 0.891208291053772,
+      "learning_rate": 0.00019975508429348894,
+      "loss": 4.1779,
+      "step": 372
+    },
+    {
+      "epoch": 0.022945374015748032,
+      "grad_norm": 1.31670081615448,
+      "learning_rate": 0.00019975372985551196,
+      "loss": 3.4949,
+      "step": 373
+    },
+    {
+      "epoch": 0.02300688976377953,
+      "grad_norm": 0.7102341651916504,
+      "learning_rate": 0.0001997523716873017,
+      "loss": 2.8214,
+      "step": 374
+    },
+    {
+      "epoch": 0.023068405511811024,
+      "grad_norm": 0.823548436164856,
+      "learning_rate": 0.000199751009788909,
+      "loss": 3.883,
+      "step": 375
+    },
+    {
+      "epoch": 0.02312992125984252,
+      "grad_norm": 0.95878666639328,
+      "learning_rate": 0.00019974964416038474,
+      "loss": 3.4145,
+      "step": 376
+    },
+    {
+      "epoch": 0.023191437007874016,
+      "grad_norm": 0.9223459362983704,
+      "learning_rate": 0.00019974827480178002,
+      "loss": 2.991,
+      "step": 377
+    },
+    {
+      "epoch": 0.023252952755905512,
+      "grad_norm": 1.0298465490341187,
+      "learning_rate": 0.00019974690171314604,
+      "loss": 3.8479,
+      "step": 378
+    },
+    {
+      "epoch": 0.023314468503937008,
+      "grad_norm": 0.8228957653045654,
+      "learning_rate": 0.00019974552489453412,
+      "loss": 3.5001,
+      "step": 379
+    },
+    {
+      "epoch": 0.023375984251968504,
+      "grad_norm": 1.0585010051727295,
+      "learning_rate": 0.00019974414434599579,
+      "loss": 3.5123,
+      "step": 380
+    },
+    {
+      "epoch": 0.0234375,
+      "grad_norm": 0.9446009993553162,
+      "learning_rate": 0.00019974276006758262,
+      "loss": 4.298,
+      "step": 381
+    },
+    {
+      "epoch": 0.023499015748031496,
+      "grad_norm": 1.0764799118041992,
+      "learning_rate": 0.0001997413720593464,
+      "loss": 3.554,
+      "step": 382
+    },
+    {
+      "epoch": 0.023560531496062992,
+      "grad_norm": 1.1019024848937988,
+      "learning_rate": 0.00019973998032133907,
+      "loss": 3.4765,
+      "step": 383
+    },
+    {
+      "epoch": 0.023622047244094488,
+      "grad_norm": 1.0713202953338623,
+      "learning_rate": 0.00019973858485361262,
+      "loss": 3.4515,
+      "step": 384
+    },
+    {
+      "epoch": 0.023683562992125984,
+      "grad_norm": 1.5042023658752441,
+      "learning_rate": 0.00019973718565621928,
+      "loss": 3.4442,
+      "step": 385
+    },
+    {
+      "epoch": 0.02374507874015748,
+      "grad_norm": 1.3373968601226807,
+      "learning_rate": 0.00019973578272921132,
+      "loss": 3.0431,
+      "step": 386
+    },
+    {
+      "epoch": 0.023806594488188976,
+      "grad_norm": 0.94234299659729,
+      "learning_rate": 0.00019973437607264126,
+      "loss": 3.8428,
+      "step": 387
+    },
+    {
+      "epoch": 0.02386811023622047,
+      "grad_norm": 0.948725163936615,
+      "learning_rate": 0.00019973296568656163,
+      "loss": 4.3029,
+      "step": 388
+    },
+    {
+      "epoch": 0.023929625984251968,
+      "grad_norm": 0.9045296311378479,
+      "learning_rate": 0.00019973155157102524,
+      "loss": 3.6163,
+      "step": 389
+    },
+    {
+      "epoch": 0.023991141732283464,
+      "grad_norm": 1.085138201713562,
+      "learning_rate": 0.0001997301337260849,
+      "loss": 3.4926,
+      "step": 390
+    },
+    {
+      "epoch": 0.02405265748031496,
+      "grad_norm": 0.8840259313583374,
+      "learning_rate": 0.0001997287121517937,
+      "loss": 3.3733,
+      "step": 391
+    },
+    {
+      "epoch": 0.024114173228346455,
+      "grad_norm": 1.6046147346496582,
+      "learning_rate": 0.0001997272868482048,
+      "loss": 4.6655,
+      "step": 392
+    },
+    {
+      "epoch": 0.02417568897637795,
+      "grad_norm": 1.0255811214447021,
+      "learning_rate": 0.00019972585781537138,
+      "loss": 4.0904,
+      "step": 393
+    },
+    {
+      "epoch": 0.024237204724409447,
+      "grad_norm": 1.0616693496704102,
+      "learning_rate": 0.00019972442505334704,
+      "loss": 3.5304,
+      "step": 394
+    },
+    {
+      "epoch": 0.024298720472440943,
+      "grad_norm": 0.8027012348175049,
+      "learning_rate": 0.00019972298856218523,
+      "loss": 3.6549,
+      "step": 395
+    },
+    {
+      "epoch": 0.02436023622047244,
+      "grad_norm": 1.0294640064239502,
+      "learning_rate": 0.00019972154834193976,
+      "loss": 3.5414,
+      "step": 396
+    },
+    {
+      "epoch": 0.02442175196850394,
+      "grad_norm": 1.3136768341064453,
+      "learning_rate": 0.00019972010439266443,
+      "loss": 3.9144,
+      "step": 397
+    },
+    {
+      "epoch": 0.024483267716535435,
+      "grad_norm": 0.9314261078834534,
+      "learning_rate": 0.00019971865671441322,
+      "loss": 3.2424,
+      "step": 398
+    },
+    {
+      "epoch": 0.02454478346456693,
+      "grad_norm": 0.8915948271751404,
+      "learning_rate": 0.00019971720530724034,
+      "loss": 3.7625,
+      "step": 399
+    },
+    {
+      "epoch": 0.024606299212598427,
+      "grad_norm": 1.0966439247131348,
+      "learning_rate": 0.0001997157501712,
+      "loss": 3.4188,
+      "step": 400
+    },
+    {
+      "epoch": 0.024667814960629923,
+      "grad_norm": 1.0084360837936401,
+      "learning_rate": 0.00019971429130634665,
+      "loss": 3.6151,
+      "step": 401
+    },
+    {
+      "epoch": 0.02472933070866142,
+      "grad_norm": 0.8437523245811462,
+      "learning_rate": 0.00019971282871273483,
+      "loss": 3.585,
+      "step": 402
+    },
+    {
+      "epoch": 0.024790846456692914,
+      "grad_norm": 0.9898659586906433,
+      "learning_rate": 0.0001997113623904192,
+      "loss": 3.8255,
+      "step": 403
+    },
+    {
+      "epoch": 0.02485236220472441,
+      "grad_norm": 0.8994322419166565,
+      "learning_rate": 0.00019970989233945464,
+      "loss": 3.7063,
+      "step": 404
+    },
+    {
+      "epoch": 0.024913877952755906,
+      "grad_norm": 0.8431469202041626,
+      "learning_rate": 0.00019970841855989607,
+      "loss": 3.6806,
+      "step": 405
+    },
+    {
+      "epoch": 0.024975393700787402,
+      "grad_norm": 0.9890526533126831,
+      "learning_rate": 0.0001997069410517987,
+      "loss": 3.2926,
+      "step": 406
+    },
+    {
+      "epoch": 0.0250369094488189,
+      "grad_norm": 0.9792404770851135,
+      "learning_rate": 0.00019970545981521767,
+      "loss": 3.2136,
+      "step": 407
+    },
+    {
+      "epoch": 0.025098425196850394,
+      "grad_norm": 0.9487980604171753,
+      "learning_rate": 0.00019970397485020845,
+      "loss": 3.469,
+      "step": 408
+    },
+    {
+      "epoch": 0.02515994094488189,
+      "grad_norm": 0.7471100091934204,
+      "learning_rate": 0.0001997024861568265,
+      "loss": 3.8003,
+      "step": 409
+    },
+    {
+      "epoch": 0.025221456692913386,
+      "grad_norm": 1.0453441143035889,
+      "learning_rate": 0.00019970099373512753,
+      "loss": 3.3392,
+      "step": 410
+    },
+    {
+      "epoch": 0.025282972440944882,
+      "grad_norm": 1.076411485671997,
+      "learning_rate": 0.00019969949758516734,
+      "loss": 3.2007,
+      "step": 411
+    },
+    {
+      "epoch": 0.025344488188976378,
+      "grad_norm": 1.2227201461791992,
+      "learning_rate": 0.0001996979977070019,
+      "loss": 3.2594,
+      "step": 412
+    },
+    {
+      "epoch": 0.025406003937007874,
+      "grad_norm": 1.1443345546722412,
+      "learning_rate": 0.00019969649410068728,
+      "loss": 3.2293,
+      "step": 413
+    },
+    {
+      "epoch": 0.02546751968503937,
+      "grad_norm": 0.9436380863189697,
+      "learning_rate": 0.00019969498676627968,
+      "loss": 2.9817,
+      "step": 414
+    },
+    {
+      "epoch": 0.025529035433070866,
+      "grad_norm": 0.9178913235664368,
+      "learning_rate": 0.0001996934757038355,
+      "loss": 3.9258,
+      "step": 415
+    },
+    {
+      "epoch": 0.025590551181102362,
+      "grad_norm": 0.9677742123603821,
+      "learning_rate": 0.00019969196091341123,
+      "loss": 3.692,
+      "step": 416
+    },
+    {
+      "epoch": 0.025652066929133858,
+      "grad_norm": 1.1985435485839844,
+      "learning_rate": 0.00019969044239506354,
+      "loss": 3.9301,
+      "step": 417
+    },
+    {
+      "epoch": 0.025713582677165354,
+      "grad_norm": 0.8541794419288635,
+      "learning_rate": 0.00019968892014884918,
+      "loss": 3.5045,
+      "step": 418
+    },
+    {
+      "epoch": 0.02577509842519685,
+      "grad_norm": 0.9312351942062378,
+      "learning_rate": 0.00019968739417482504,
+      "loss": 3.4468,
+      "step": 419
+    },
+    {
+      "epoch": 0.025836614173228346,
+      "grad_norm": 0.922974169254303,
+      "learning_rate": 0.00019968586447304828,
+      "loss": 3.6001,
+      "step": 420
+    },
+    {
+      "epoch": 0.02589812992125984,
+      "grad_norm": 0.9945812225341797,
+      "learning_rate": 0.00019968433104357606,
+      "loss": 4.1477,
+      "step": 421
+    },
+    {
+      "epoch": 0.025959645669291338,
+      "grad_norm": 1.732353925704956,
+      "learning_rate": 0.0001996827938864657,
+      "loss": 3.318,
+      "step": 422
+    },
+    {
+      "epoch": 0.026021161417322834,
+      "grad_norm": 1.0672188997268677,
+      "learning_rate": 0.0001996812530017747,
+      "loss": 3.6703,
+      "step": 423
+    },
+    {
+      "epoch": 0.02608267716535433,
+      "grad_norm": 0.8847132921218872,
+      "learning_rate": 0.00019967970838956066,
+      "loss": 3.5199,
+      "step": 424
+    },
+    {
+      "epoch": 0.026144192913385825,
+      "grad_norm": 1.3853017091751099,
+      "learning_rate": 0.00019967816004988137,
+      "loss": 3.9526,
+      "step": 425
+    },
+    {
+      "epoch": 0.02620570866141732,
+      "grad_norm": 1.160353183746338,
+      "learning_rate": 0.0001996766079827947,
+      "loss": 3.105,
+      "step": 426
+    },
+    {
+      "epoch": 0.026267224409448817,
+      "grad_norm": 1.0820118188858032,
+      "learning_rate": 0.00019967505218835874,
+      "loss": 3.6239,
+      "step": 427
+    },
+    {
+      "epoch": 0.026328740157480313,
+      "grad_norm": 1.7962613105773926,
+      "learning_rate": 0.0001996734926666316,
+      "loss": 3.8743,
+      "step": 428
+    },
+    {
+      "epoch": 0.026390255905511813,
+      "grad_norm": 0.882472813129425,
+      "learning_rate": 0.00019967192941767166,
+      "loss": 3.5656,
+      "step": 429
+    },
+    {
+      "epoch": 0.02645177165354331,
+      "grad_norm": 0.8858088254928589,
+      "learning_rate": 0.0001996703624415373,
+      "loss": 3.9026,
+      "step": 430
+    },
+    {
+      "epoch": 0.026513287401574805,
+      "grad_norm": 0.850070595741272,
+      "learning_rate": 0.0001996687917382872,
+      "loss": 3.6702,
+      "step": 431
+    },
+    {
+      "epoch": 0.0265748031496063,
+      "grad_norm": 0.9542394280433655,
+      "learning_rate": 0.00019966721730798,
+      "loss": 3.6396,
+      "step": 432
+    },
+    {
+      "epoch": 0.026636318897637797,
+      "grad_norm": 1.092016339302063,
+      "learning_rate": 0.00019966563915067468,
+      "loss": 1.6945,
+      "step": 433
+    },
+    {
+      "epoch": 0.026697834645669292,
+      "grad_norm": 1.0792089700698853,
+      "learning_rate": 0.0001996640572664302,
+      "loss": 3.3796,
+      "step": 434
+    },
+    {
+      "epoch": 0.02675935039370079,
+      "grad_norm": 0.8353781700134277,
+      "learning_rate": 0.00019966247165530574,
+      "loss": 3.0346,
+      "step": 435
+    },
+    {
+      "epoch": 0.026820866141732284,
+      "grad_norm": 1.0181708335876465,
+      "learning_rate": 0.00019966088231736055,
+      "loss": 3.4942,
+      "step": 436
+    },
+    {
+      "epoch": 0.02688238188976378,
+      "grad_norm": 1.1239922046661377,
+      "learning_rate": 0.0001996592892526541,
+      "loss": 3.4765,
+      "step": 437
+    },
+    {
+      "epoch": 0.026943897637795276,
+      "grad_norm": 0.979616641998291,
+      "learning_rate": 0.00019965769246124593,
+      "loss": 3.3377,
+      "step": 438
+    },
+    {
+      "epoch": 0.027005413385826772,
+      "grad_norm": 0.890673041343689,
+      "learning_rate": 0.00019965609194319577,
+      "loss": 3.3583,
+      "step": 439
+    },
+    {
+      "epoch": 0.027066929133858268,
+      "grad_norm": 1.0648504495620728,
+      "learning_rate": 0.00019965448769856348,
+      "loss": 3.9946,
+      "step": 440
+    },
+    {
+      "epoch": 0.027128444881889764,
+      "grad_norm": 1.067805290222168,
+      "learning_rate": 0.00019965287972740904,
+      "loss": 3.6387,
+      "step": 441
+    },
+    {
+      "epoch": 0.02718996062992126,
+      "grad_norm": 0.73021399974823,
+      "learning_rate": 0.00019965126802979255,
+      "loss": 3.7621,
+      "step": 442
+    },
+    {
+      "epoch": 0.027251476377952756,
+      "grad_norm": 1.0201114416122437,
+      "learning_rate": 0.00019964965260577434,
+      "loss": 3.4987,
+      "step": 443
+    },
+    {
+      "epoch": 0.027312992125984252,
+      "grad_norm": 1.0614031553268433,
+      "learning_rate": 0.00019964803345541476,
+      "loss": 2.6148,
+      "step": 444
+    },
+    {
+      "epoch": 0.027374507874015748,
+      "grad_norm": 0.9333505630493164,
+      "learning_rate": 0.0001996464105787744,
+      "loss": 3.5059,
+      "step": 445
+    },
+    {
+      "epoch": 0.027436023622047244,
+      "grad_norm": 0.8443762063980103,
+      "learning_rate": 0.0001996447839759139,
+      "loss": 3.6411,
+      "step": 446
+    },
+    {
+      "epoch": 0.02749753937007874,
+      "grad_norm": 0.9799911975860596,
+      "learning_rate": 0.00019964315364689414,
+      "loss": 3.5626,
+      "step": 447
+    },
+    {
+      "epoch": 0.027559055118110236,
+      "grad_norm": 1.5019322633743286,
+      "learning_rate": 0.00019964151959177607,
+      "loss": 3.0615,
+      "step": 448
+    },
+    {
+      "epoch": 0.027620570866141732,
+      "grad_norm": 0.9467806220054626,
+      "learning_rate": 0.00019963988181062074,
+      "loss": 4.1327,
+      "step": 449
+    },
+    {
+      "epoch": 0.027682086614173228,
+      "grad_norm": 0.9385546445846558,
+      "learning_rate": 0.00019963824030348948,
+      "loss": 3.8365,
+      "step": 450
+    },
+    {
+      "epoch": 0.027743602362204724,
+      "grad_norm": 1.4333058595657349,
+      "learning_rate": 0.0001996365950704436,
+      "loss": 3.235,
+      "step": 451
+    },
+    {
+      "epoch": 0.02780511811023622,
+      "grad_norm": 0.9187132120132446,
+      "learning_rate": 0.00019963494611154468,
+      "loss": 3.5599,
+      "step": 452
+    },
+    {
+      "epoch": 0.027866633858267716,
+      "grad_norm": 1.0347806215286255,
+      "learning_rate": 0.00019963329342685435,
+      "loss": 3.5368,
+      "step": 453
+    },
+    {
+      "epoch": 0.02792814960629921,
+      "grad_norm": 1.369787573814392,
+      "learning_rate": 0.0001996316370164344,
+      "loss": 3.4808,
+      "step": 454
+    },
+    {
+      "epoch": 0.027989665354330708,
+      "grad_norm": 1.2370449304580688,
+      "learning_rate": 0.00019962997688034682,
+      "loss": 3.276,
+      "step": 455
+    },
+    {
+      "epoch": 0.028051181102362203,
+      "grad_norm": 1.083688735961914,
+      "learning_rate": 0.00019962831301865362,
+      "loss": 4.3932,
+      "step": 456
+    },
+    {
+      "epoch": 0.0281126968503937,
+      "grad_norm": 1.1103483438491821,
+      "learning_rate": 0.00019962664543141708,
+      "loss": 3.2411,
+      "step": 457
+    },
+    {
+      "epoch": 0.028174212598425195,
+      "grad_norm": 1.0022279024124146,
+      "learning_rate": 0.00019962497411869952,
+      "loss": 3.4561,
+      "step": 458
+    },
+    {
+      "epoch": 0.02823572834645669,
+      "grad_norm": 1.0995935201644897,
+      "learning_rate": 0.00019962329908056347,
+      "loss": 3.7295,
+      "step": 459
+    },
+    {
+      "epoch": 0.028297244094488187,
+      "grad_norm": 1.0070875883102417,
+      "learning_rate": 0.0001996216203170715,
+      "loss": 2.7909,
+      "step": 460
+    },
+    {
+      "epoch": 0.028358759842519687,
+      "grad_norm": 0.9351927042007446,
+      "learning_rate": 0.00019961993782828647,
+      "loss": 3.0528,
+      "step": 461
+    },
+    {
+      "epoch": 0.028420275590551183,
+      "grad_norm": 0.8965405821800232,
+      "learning_rate": 0.0001996182516142713,
+      "loss": 3.7522,
+      "step": 462
+    },
+    {
+      "epoch": 0.02848179133858268,
+      "grad_norm": 0.973232626914978,
+      "learning_rate": 0.00019961656167508893,
+      "loss": 3.7628,
+      "step": 463
+    },
+    {
+      "epoch": 0.028543307086614175,
+      "grad_norm": 1.1065330505371094,
+      "learning_rate": 0.00019961486801080267,
+      "loss": 3.3106,
+      "step": 464
+    },
+    {
+      "epoch": 0.02860482283464567,
+      "grad_norm": 1.243147611618042,
+      "learning_rate": 0.0001996131706214758,
+      "loss": 2.6794,
+      "step": 465
+    },
+    {
+      "epoch": 0.028666338582677166,
+      "grad_norm": 0.9631106853485107,
+      "learning_rate": 0.00019961146950717184,
+      "loss": 3.2503,
+      "step": 466
+    },
+    {
+      "epoch": 0.028727854330708662,
+      "grad_norm": 0.9638842940330505,
+      "learning_rate": 0.00019960976466795437,
+      "loss": 3.341,
+      "step": 467
+    },
+    {
+      "epoch": 0.02878937007874016,
+      "grad_norm": 0.9085739254951477,
+      "learning_rate": 0.0001996080561038871,
+      "loss": 3.1952,
+      "step": 468
+    },
+    {
+      "epoch": 0.028850885826771654,
+      "grad_norm": 0.8209070563316345,
+      "learning_rate": 0.00019960634381503396,
+      "loss": 3.4946,
+      "step": 469
+    },
+    {
+      "epoch": 0.02891240157480315,
+      "grad_norm": 0.9326809644699097,
+      "learning_rate": 0.00019960462780145902,
+      "loss": 3.7444,
+      "step": 470
+    },
+    {
+      "epoch": 0.028973917322834646,
+      "grad_norm": 0.9342178106307983,
+      "learning_rate": 0.0001996029080632264,
+      "loss": 4.1316,
+      "step": 471
+    },
+    {
+      "epoch": 0.029035433070866142,
+      "grad_norm": 1.1912188529968262,
+      "learning_rate": 0.0001996011846004004,
+      "loss": 2.5342,
+      "step": 472
+    },
+    {
+      "epoch": 0.029096948818897638,
+      "grad_norm": 1.049298882484436,
+      "learning_rate": 0.0001995994574130455,
+      "loss": 3.4524,
+      "step": 473
+    },
+    {
+      "epoch": 0.029158464566929134,
+      "grad_norm": 1.0711467266082764,
+      "learning_rate": 0.0001995977265012263,
+      "loss": 3.8034,
+      "step": 474
+    },
+    {
+      "epoch": 0.02921998031496063,
+      "grad_norm": 0.975959300994873,
+      "learning_rate": 0.00019959599186500745,
+      "loss": 3.205,
+      "step": 475
+    },
+    {
+      "epoch": 0.029281496062992126,
+      "grad_norm": 1.1209800243377686,
+      "learning_rate": 0.00019959425350445388,
+      "loss": 3.4684,
+      "step": 476
+    },
+    {
+      "epoch": 0.029343011811023622,
+      "grad_norm": 0.8269549012184143,
+      "learning_rate": 0.0001995925114196306,
+      "loss": 3.3476,
+      "step": 477
+    },
+    {
+      "epoch": 0.029404527559055118,
+      "grad_norm": 0.8775315880775452,
+      "learning_rate": 0.00019959076561060273,
+      "loss": 3.2371,
+      "step": 478
+    },
+    {
+      "epoch": 0.029466043307086614,
+      "grad_norm": 1.2209237813949585,
+      "learning_rate": 0.00019958901607743557,
+      "loss": 2.8734,
+      "step": 479
+    },
+    {
+      "epoch": 0.02952755905511811,
+      "grad_norm": 1.089030146598816,
+      "learning_rate": 0.00019958726282019452,
+      "loss": 3.4011,
+      "step": 480
+    },
+    {
+      "epoch": 0.029589074803149606,
+      "grad_norm": 1.0936998128890991,
+      "learning_rate": 0.00019958550583894515,
+      "loss": 3.243,
+      "step": 481
+    },
+    {
+      "epoch": 0.0296505905511811,
+      "grad_norm": 0.8202275037765503,
+      "learning_rate": 0.00019958374513375318,
+      "loss": 3.1344,
+      "step": 482
+    },
+    {
+      "epoch": 0.029712106299212598,
+      "grad_norm": 0.9325072765350342,
+      "learning_rate": 0.00019958198070468443,
+      "loss": 3.1298,
+      "step": 483
+    },
+    {
+      "epoch": 0.029773622047244094,
+      "grad_norm": 1.0305737257003784,
+      "learning_rate": 0.0001995802125518049,
+      "loss": 3.7783,
+      "step": 484
+    },
+    {
+      "epoch": 0.02983513779527559,
+      "grad_norm": 1.07395339012146,
+      "learning_rate": 0.00019957844067518065,
+      "loss": 3.5047,
+      "step": 485
+    },
+    {
+      "epoch": 0.029896653543307086,
+      "grad_norm": 1.4039740562438965,
+      "learning_rate": 0.000199576665074878,
+      "loss": 3.342,
+      "step": 486
+    },
+    {
+      "epoch": 0.02995816929133858,
+      "grad_norm": 1.080363392829895,
+      "learning_rate": 0.00019957488575096335,
+      "loss": 4.0152,
+      "step": 487
+    },
+    {
+      "epoch": 0.030019685039370077,
+      "grad_norm": 1.056164264678955,
+      "learning_rate": 0.00019957310270350324,
+      "loss": 3.7356,
+      "step": 488
+    },
+    {
+      "epoch": 0.030081200787401573,
+      "grad_norm": 0.8646132349967957,
+      "learning_rate": 0.0001995713159325643,
+      "loss": 3.54,
+      "step": 489
+    },
+    {
+      "epoch": 0.03014271653543307,
+      "grad_norm": 1.3020541667938232,
+      "learning_rate": 0.0001995695254382133,
+      "loss": 4.4182,
+      "step": 490
+    },
+    {
+      "epoch": 0.030204232283464565,
+      "grad_norm": 0.763482391834259,
+      "learning_rate": 0.00019956773122051735,
+      "loss": 3.1194,
+      "step": 491
+    },
+    {
+      "epoch": 0.03026574803149606,
+      "grad_norm": 1.2351323366165161,
+      "learning_rate": 0.00019956593327954343,
+      "loss": 3.3543,
+      "step": 492
+    },
+    {
+      "epoch": 0.03032726377952756,
+      "grad_norm": 0.9430343508720398,
+      "learning_rate": 0.00019956413161535883,
+      "loss": 3.7546,
+      "step": 493
+    },
+    {
+      "epoch": 0.030388779527559057,
+      "grad_norm": 0.8940910696983337,
+      "learning_rate": 0.00019956232622803083,
+      "loss": 3.6072,
+      "step": 494
+    },
+    {
+      "epoch": 0.030450295275590553,
+      "grad_norm": 1.052917718887329,
+      "learning_rate": 0.00019956051711762706,
+      "loss": 3.8007,
+      "step": 495
+    },
+    {
+      "epoch": 0.03051181102362205,
+      "grad_norm": 1.0221415758132935,
+      "learning_rate": 0.0001995587042842151,
+      "loss": 3.2055,
+      "step": 496
+    },
+    {
+      "epoch": 0.030573326771653545,
+      "grad_norm": 0.8283601999282837,
+      "learning_rate": 0.00019955688772786278,
+      "loss": 3.1504,
+      "step": 497
+    },
+    {
+      "epoch": 0.03063484251968504,
+      "grad_norm": 1.0525838136672974,
+      "learning_rate": 0.00019955506744863795,
+      "loss": 4.0667,
+      "step": 498
+    },
+    {
+      "epoch": 0.030696358267716536,
+      "grad_norm": 1.818143606185913,
+      "learning_rate": 0.00019955324344660876,
+      "loss": 2.8654,
+      "step": 499
+    },
+    {
+      "epoch": 0.030757874015748032,
+      "grad_norm": 0.801935613155365,
+      "learning_rate": 0.00019955141572184338,
+      "loss": 3.5129,
+      "step": 500
+    },
+    {
+      "epoch": 0.03081938976377953,
+      "grad_norm": 0.9545162916183472,
+      "learning_rate": 0.00019954958427441017,
+      "loss": 2.9721,
+      "step": 501
+    },
+    {
+      "epoch": 0.030880905511811024,
+      "grad_norm": 1.7371021509170532,
+      "learning_rate": 0.0001995477491043776,
+      "loss": 3.1016,
+      "step": 502
+    },
+    {
+      "epoch": 0.03094242125984252,
+      "grad_norm": 0.9924920201301575,
+      "learning_rate": 0.00019954591021181431,
+      "loss": 3.6612,
+      "step": 503
+    },
+    {
+      "epoch": 0.031003937007874016,
+      "grad_norm": 0.895978569984436,
+      "learning_rate": 0.00019954406759678907,
+      "loss": 3.531,
+      "step": 504
+    },
+    {
+      "epoch": 0.031065452755905512,
+      "grad_norm": 1.4292244911193848,
+      "learning_rate": 0.00019954222125937078,
+      "loss": 3.3464,
+      "step": 505
+    },
+    {
+      "epoch": 0.031126968503937008,
+      "grad_norm": 1.0023441314697266,
+      "learning_rate": 0.00019954037119962847,
+      "loss": 3.2339,
+      "step": 506
+    },
+    {
+      "epoch": 0.031188484251968504,
+      "grad_norm": 1.214519739151001,
+      "learning_rate": 0.0001995385174176313,
+      "loss": 3.3887,
+      "step": 507
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 0.9598400592803955,
+      "learning_rate": 0.00019953665991344867,
+      "loss": 3.3914,
+      "step": 508
+    },
+    {
+      "epoch": 0.031311515748031496,
+      "grad_norm": 0.9039584994316101,
+      "learning_rate": 0.00019953479868715,
+      "loss": 3.9318,
+      "step": 509
+    },
+    {
+      "epoch": 0.03137303149606299,
+      "grad_norm": 0.9331154227256775,
+      "learning_rate": 0.00019953293373880485,
+      "loss": 3.6219,
+      "step": 510
+    },
+    {
+      "epoch": 0.03143454724409449,
+      "grad_norm": 1.092786431312561,
+      "learning_rate": 0.000199531065068483,
+      "loss": 3.5251,
+      "step": 511
+    },
+    {
+      "epoch": 0.031496062992125984,
+      "grad_norm": 1.3036303520202637,
+      "learning_rate": 0.00019952919267625428,
+      "loss": 3.6585,
+      "step": 512
+    },
+    {
+      "epoch": 0.03155757874015748,
+      "grad_norm": 1.07334566116333,
+      "learning_rate": 0.0001995273165621888,
+      "loss": 4.0631,
+      "step": 513
+    },
+    {
+      "epoch": 0.031619094488188976,
+      "grad_norm": 0.9736854434013367,
+      "learning_rate": 0.00019952543672635661,
+      "loss": 3.5591,
+      "step": 514
+    },
+    {
+      "epoch": 0.03168061023622047,
+      "grad_norm": 0.9235316514968872,
+      "learning_rate": 0.00019952355316882808,
+      "loss": 3.2985,
+      "step": 515
+    },
+    {
+      "epoch": 0.03174212598425197,
+      "grad_norm": 0.8185683488845825,
+      "learning_rate": 0.00019952166588967362,
+      "loss": 3.1134,
+      "step": 516
+    },
+    {
+      "epoch": 0.031803641732283464,
+      "grad_norm": 0.8729057312011719,
+      "learning_rate": 0.0001995197748889638,
+      "loss": 4.0634,
+      "step": 517
+    },
+    {
+      "epoch": 0.03186515748031496,
+      "grad_norm": 1.627616286277771,
+      "learning_rate": 0.00019951788016676932,
+      "loss": 3.4362,
+      "step": 518
+    },
+    {
+      "epoch": 0.031926673228346455,
+      "grad_norm": 1.158582329750061,
+      "learning_rate": 0.00019951598172316111,
+      "loss": 3.3043,
+      "step": 519
+    },
+    {
+      "epoch": 0.03198818897637795,
+      "grad_norm": 0.8520174026489258,
+      "learning_rate": 0.00019951407955821004,
+      "loss": 3.9587,
+      "step": 520
+    },
+    {
+      "epoch": 0.03204970472440945,
+      "grad_norm": 1.195931315422058,
+      "learning_rate": 0.00019951217367198732,
+      "loss": 3.464,
+      "step": 521
+    },
+    {
+      "epoch": 0.03211122047244094,
+      "grad_norm": 0.7030872702598572,
+      "learning_rate": 0.0001995102640645642,
+      "loss": 3.5095,
+      "step": 522
+    },
+    {
+      "epoch": 0.03217273622047244,
+      "grad_norm": 0.8625534176826477,
+      "learning_rate": 0.00019950835073601208,
+      "loss": 3.6112,
+      "step": 523
+    },
+    {
+      "epoch": 0.032234251968503935,
+      "grad_norm": 1.4536610841751099,
+      "learning_rate": 0.00019950643368640253,
+      "loss": 3.0904,
+      "step": 524
+    },
+    {
+      "epoch": 0.03229576771653543,
+      "grad_norm": 1.856756329536438,
+      "learning_rate": 0.00019950451291580722,
+      "loss": 3.17,
+      "step": 525
+    },
+    {
+      "epoch": 0.03235728346456693,
+      "grad_norm": 1.1508023738861084,
+      "learning_rate": 0.00019950258842429798,
+      "loss": 3.4621,
+      "step": 526
+    },
+    {
+      "epoch": 0.03241879921259842,
+      "grad_norm": 1.1247261762619019,
+      "learning_rate": 0.00019950066021194674,
+      "loss": 2.9779,
+      "step": 527
+    },
+    {
+      "epoch": 0.03248031496062992,
+      "grad_norm": 1.0827573537826538,
+      "learning_rate": 0.0001994987282788257,
+      "loss": 3.4118,
+      "step": 528
+    },
+    {
+      "epoch": 0.032541830708661415,
+      "grad_norm": 0.9420340061187744,
+      "learning_rate": 0.000199496792625007,
+      "loss": 3.6135,
+      "step": 529
+    },
+    {
+      "epoch": 0.03260334645669291,
+      "grad_norm": 1.12226140499115,
+      "learning_rate": 0.00019949485325056307,
+      "loss": 3.8459,
+      "step": 530
+    },
+    {
+      "epoch": 0.03266486220472441,
+      "grad_norm": 1.1019643545150757,
+      "learning_rate": 0.00019949291015556643,
+      "loss": 3.5305,
+      "step": 531
+    },
+    {
+      "epoch": 0.0327263779527559,
+      "grad_norm": 0.9187541604042053,
+      "learning_rate": 0.00019949096334008975,
+      "loss": 3.455,
+      "step": 532
+    },
+    {
+      "epoch": 0.0327878937007874,
+      "grad_norm": 1.2530937194824219,
+      "learning_rate": 0.00019948901280420582,
+      "loss": 3.2316,
+      "step": 533
+    },
+    {
+      "epoch": 0.032849409448818895,
+      "grad_norm": 1.5688128471374512,
+      "learning_rate": 0.00019948705854798754,
+      "loss": 3.6195,
+      "step": 534
+    },
+    {
+      "epoch": 0.03291092519685039,
+      "grad_norm": 1.0098553895950317,
+      "learning_rate": 0.00019948510057150808,
+      "loss": 3.8563,
+      "step": 535
+    },
+    {
+      "epoch": 0.03297244094488189,
+      "grad_norm": 1.1792165040969849,
+      "learning_rate": 0.00019948313887484056,
+      "loss": 3.4581,
+      "step": 536
+    },
+    {
+      "epoch": 0.03303395669291338,
+      "grad_norm": 0.9664503931999207,
+      "learning_rate": 0.00019948117345805838,
+      "loss": 3.247,
+      "step": 537
+    },
+    {
+      "epoch": 0.03309547244094488,
+      "grad_norm": 1.1041991710662842,
+      "learning_rate": 0.00019947920432123505,
+      "loss": 3.7112,
+      "step": 538
+    },
+    {
+      "epoch": 0.033156988188976375,
+      "grad_norm": 1.2854499816894531,
+      "learning_rate": 0.0001994772314644442,
+      "loss": 2.9991,
+      "step": 539
+    },
+    {
+      "epoch": 0.03321850393700788,
+      "grad_norm": 1.0636075735092163,
+      "learning_rate": 0.0001994752548877596,
+      "loss": 3.4327,
+      "step": 540
+    },
+    {
+      "epoch": 0.03328001968503937,
+      "grad_norm": 1.0861881971359253,
+      "learning_rate": 0.00019947327459125512,
+      "loss": 3.7654,
+      "step": 541
+    },
+    {
+      "epoch": 0.03334153543307087,
+      "grad_norm": 1.067850112915039,
+      "learning_rate": 0.00019947129057500487,
+      "loss": 4.3648,
+      "step": 542
+    },
+    {
+      "epoch": 0.033403051181102365,
+      "grad_norm": 1.035908818244934,
+      "learning_rate": 0.00019946930283908302,
+      "loss": 2.8183,
+      "step": 543
+    },
+    {
+      "epoch": 0.03346456692913386,
+      "grad_norm": 0.9550888538360596,
+      "learning_rate": 0.0001994673113835639,
+      "loss": 3.2119,
+      "step": 544
+    },
+    {
+      "epoch": 0.03352608267716536,
+      "grad_norm": 1.2650779485702515,
+      "learning_rate": 0.00019946531620852195,
+      "loss": 2.8043,
+      "step": 545
+    },
+    {
+      "epoch": 0.03358759842519685,
+      "grad_norm": 1.06884765625,
+      "learning_rate": 0.00019946331731403186,
+      "loss": 3.6802,
+      "step": 546
+    },
+    {
+      "epoch": 0.03364911417322835,
+      "grad_norm": 0.9702146649360657,
+      "learning_rate": 0.00019946131470016828,
+      "loss": 3.6915,
+      "step": 547
+    },
+    {
+      "epoch": 0.033710629921259845,
+      "grad_norm": 1.0026342868804932,
+      "learning_rate": 0.00019945930836700617,
+      "loss": 3.4466,
+      "step": 548
+    },
+    {
+      "epoch": 0.03377214566929134,
+      "grad_norm": 0.948315441608429,
+      "learning_rate": 0.0001994572983146205,
+      "loss": 3.797,
+      "step": 549
+    },
+    {
+      "epoch": 0.03383366141732284,
+      "grad_norm": 0.9876758456230164,
+      "learning_rate": 0.0001994552845430865,
+      "loss": 3.3743,
+      "step": 550
+    },
+    {
+      "epoch": 0.03389517716535433,
+      "grad_norm": 1.9588466882705688,
+      "learning_rate": 0.0001994532670524794,
+      "loss": 2.9741,
+      "step": 551
+    },
+    {
+      "epoch": 0.03395669291338583,
+      "grad_norm": 1.0065449476242065,
+      "learning_rate": 0.0001994512458428747,
+      "loss": 3.2762,
+      "step": 552
+    },
+    {
+      "epoch": 0.034018208661417325,
+      "grad_norm": 1.0300318002700806,
+      "learning_rate": 0.00019944922091434794,
+      "loss": 3.1437,
+      "step": 553
+    },
+    {
+      "epoch": 0.03407972440944882,
+      "grad_norm": 1.2331676483154297,
+      "learning_rate": 0.0001994471922669749,
+      "loss": 4.0851,
+      "step": 554
+    },
+    {
+      "epoch": 0.03414124015748032,
+      "grad_norm": 0.8178260326385498,
+      "learning_rate": 0.00019944515990083134,
+      "loss": 3.5223,
+      "step": 555
+    },
+    {
+      "epoch": 0.03420275590551181,
+      "grad_norm": 1.142243504524231,
+      "learning_rate": 0.00019944312381599337,
+      "loss": 3.38,
+      "step": 556
+    },
+    {
+      "epoch": 0.03426427165354331,
+      "grad_norm": 1.2814786434173584,
+      "learning_rate": 0.0001994410840125371,
+      "loss": 3.3817,
+      "step": 557
+    },
+    {
+      "epoch": 0.034325787401574805,
+      "grad_norm": 1.5520546436309814,
+      "learning_rate": 0.00019943904049053872,
+      "loss": 3.2098,
+      "step": 558
+    },
+    {
+      "epoch": 0.0343873031496063,
+      "grad_norm": 1.2928251028060913,
+      "learning_rate": 0.00019943699325007476,
+      "loss": 3.5189,
+      "step": 559
+    },
+    {
+      "epoch": 0.0344488188976378,
+      "grad_norm": 1.0989892482757568,
+      "learning_rate": 0.0001994349422912217,
+      "loss": 3.5601,
+      "step": 560
+    },
+    {
+      "epoch": 0.03451033464566929,
+      "grad_norm": 1.6099013090133667,
+      "learning_rate": 0.0001994328876140563,
+      "loss": 3.3032,
+      "step": 561
+    },
+    {
+      "epoch": 0.03457185039370079,
+      "grad_norm": 1.1221139430999756,
+      "learning_rate": 0.00019943082921865534,
+      "loss": 3.338,
+      "step": 562
+    },
+    {
+      "epoch": 0.034633366141732284,
+      "grad_norm": 0.9317722916603088,
+      "learning_rate": 0.00019942876710509578,
+      "loss": 3.1798,
+      "step": 563
+    },
+    {
+      "epoch": 0.03469488188976378,
+      "grad_norm": 0.9587872624397278,
+      "learning_rate": 0.0001994267012734548,
+      "loss": 3.3094,
+      "step": 564
+    },
+    {
+      "epoch": 0.034756397637795276,
+      "grad_norm": 0.7804529666900635,
+      "learning_rate": 0.0001994246317238096,
+      "loss": 3.6158,
+      "step": 565
+    },
+    {
+      "epoch": 0.03481791338582677,
+      "grad_norm": 1.6920984983444214,
+      "learning_rate": 0.00019942255845623757,
+      "loss": 2.9784,
+      "step": 566
+    },
+    {
+      "epoch": 0.03487942913385827,
+      "grad_norm": 1.120915412902832,
+      "learning_rate": 0.00019942048147081627,
+      "loss": 3.3734,
+      "step": 567
+    },
+    {
+      "epoch": 0.034940944881889764,
+      "grad_norm": 0.9846224784851074,
+      "learning_rate": 0.00019941840076762333,
+      "loss": 3.8598,
+      "step": 568
+    },
+    {
+      "epoch": 0.03500246062992126,
+      "grad_norm": 1.3900213241577148,
+      "learning_rate": 0.00019941631634673657,
+      "loss": 3.9169,
+      "step": 569
+    },
+    {
+      "epoch": 0.035063976377952756,
+      "grad_norm": 1.881759524345398,
+      "learning_rate": 0.00019941422820823394,
+      "loss": 3.6119,
+      "step": 570
+    },
+    {
+      "epoch": 0.03512549212598425,
+      "grad_norm": 0.9645740389823914,
+      "learning_rate": 0.00019941213635219353,
+      "loss": 3.9157,
+      "step": 571
+    },
+    {
+      "epoch": 0.03518700787401575,
+      "grad_norm": 0.9994139671325684,
+      "learning_rate": 0.00019941004077869354,
+      "loss": 3.6442,
+      "step": 572
+    },
+    {
+      "epoch": 0.035248523622047244,
+      "grad_norm": 0.8723158836364746,
+      "learning_rate": 0.00019940794148781238,
+      "loss": 3.5086,
+      "step": 573
+    },
+    {
+      "epoch": 0.03531003937007874,
+      "grad_norm": 0.8435621857643127,
+      "learning_rate": 0.0001994058384796285,
+      "loss": 3.5677,
+      "step": 574
+    },
+    {
+      "epoch": 0.035371555118110236,
+      "grad_norm": 1.8502157926559448,
+      "learning_rate": 0.00019940373175422057,
+      "loss": 2.3756,
+      "step": 575
+    },
+    {
+      "epoch": 0.03543307086614173,
+      "grad_norm": 1.033142328262329,
+      "learning_rate": 0.00019940162131166738,
+      "loss": 3.186,
+      "step": 576
+    },
+    {
+      "epoch": 0.03549458661417323,
+      "grad_norm": 0.9541724920272827,
+      "learning_rate": 0.0001993995071520478,
+      "loss": 3.2706,
+      "step": 577
+    },
+    {
+      "epoch": 0.035556102362204724,
+      "grad_norm": 0.8491504788398743,
+      "learning_rate": 0.00019939738927544096,
+      "loss": 3.6939,
+      "step": 578
+    },
+    {
+      "epoch": 0.03561761811023622,
+      "grad_norm": 1.10914146900177,
+      "learning_rate": 0.00019939526768192597,
+      "loss": 2.5326,
+      "step": 579
+    },
+    {
+      "epoch": 0.035679133858267716,
+      "grad_norm": 1.1318038702011108,
+      "learning_rate": 0.00019939314237158224,
+      "loss": 3.4916,
+      "step": 580
+    },
+    {
+      "epoch": 0.03574064960629921,
+      "grad_norm": 0.9064112901687622,
+      "learning_rate": 0.0001993910133444892,
+      "loss": 3.1066,
+      "step": 581
+    },
+    {
+      "epoch": 0.03580216535433071,
+      "grad_norm": 0.9380522966384888,
+      "learning_rate": 0.0001993888806007265,
+      "loss": 3.3594,
+      "step": 582
+    },
+    {
+      "epoch": 0.0358636811023622,
+      "grad_norm": 0.9482637643814087,
+      "learning_rate": 0.00019938674414037388,
+      "loss": 3.9135,
+      "step": 583
+    },
+    {
+      "epoch": 0.0359251968503937,
+      "grad_norm": 1.0766505002975464,
+      "learning_rate": 0.0001993846039635112,
+      "loss": 2.9252,
+      "step": 584
+    },
+    {
+      "epoch": 0.035986712598425195,
+      "grad_norm": 1.118417501449585,
+      "learning_rate": 0.00019938246007021853,
+      "loss": 3.4088,
+      "step": 585
+    },
+    {
+      "epoch": 0.03604822834645669,
+      "grad_norm": 1.071599006652832,
+      "learning_rate": 0.00019938031246057603,
+      "loss": 3.3055,
+      "step": 586
+    },
+    {
+      "epoch": 0.03610974409448819,
+      "grad_norm": 0.8757104277610779,
+      "learning_rate": 0.000199378161134664,
+      "loss": 3.3405,
+      "step": 587
+    },
+    {
+      "epoch": 0.03617125984251968,
+      "grad_norm": 1.0440924167633057,
+      "learning_rate": 0.00019937600609256287,
+      "loss": 3.2052,
+      "step": 588
+    },
+    {
+      "epoch": 0.03623277559055118,
+      "grad_norm": 1.0404928922653198,
+      "learning_rate": 0.00019937384733435324,
+      "loss": 3.7384,
+      "step": 589
+    },
+    {
+      "epoch": 0.036294291338582675,
+      "grad_norm": 1.5233558416366577,
+      "learning_rate": 0.00019937168486011588,
+      "loss": 3.1728,
+      "step": 590
+    },
+    {
+      "epoch": 0.03635580708661417,
+      "grad_norm": 0.8011497855186462,
+      "learning_rate": 0.0001993695186699316,
+      "loss": 3.8443,
+      "step": 591
+    },
+    {
+      "epoch": 0.03641732283464567,
+      "grad_norm": 1.0814523696899414,
+      "learning_rate": 0.00019936734876388143,
+      "loss": 3.6467,
+      "step": 592
+    },
+    {
+      "epoch": 0.03647883858267716,
+      "grad_norm": 0.8444287180900574,
+      "learning_rate": 0.0001993651751420465,
+      "loss": 3.6069,
+      "step": 593
+    },
+    {
+      "epoch": 0.03654035433070866,
+      "grad_norm": 1.2145229578018188,
+      "learning_rate": 0.0001993629978045081,
+      "loss": 2.7559,
+      "step": 594
+    },
+    {
+      "epoch": 0.036601870078740155,
+      "grad_norm": 1.7142959833145142,
+      "learning_rate": 0.00019936081675134764,
+      "loss": 2.6239,
+      "step": 595
+    },
+    {
+      "epoch": 0.03666338582677165,
+      "grad_norm": 0.9558281302452087,
+      "learning_rate": 0.00019935863198264667,
+      "loss": 3.744,
+      "step": 596
+    },
+    {
+      "epoch": 0.03672490157480315,
+      "grad_norm": 1.3867777585983276,
+      "learning_rate": 0.0001993564434984869,
+      "loss": 2.9216,
+      "step": 597
+    },
+    {
+      "epoch": 0.03678641732283464,
+      "grad_norm": 0.8331800699234009,
+      "learning_rate": 0.0001993542512989502,
+      "loss": 3.3235,
+      "step": 598
+    },
+    {
+      "epoch": 0.03684793307086614,
+      "grad_norm": 0.8074973225593567,
+      "learning_rate": 0.0001993520553841185,
+      "loss": 3.5989,
+      "step": 599
+    },
+    {
+      "epoch": 0.036909448818897635,
+      "grad_norm": 1.0215924978256226,
+      "learning_rate": 0.00019934985575407394,
+      "loss": 2.5994,
+      "step": 600
+    },
+    {
+      "epoch": 0.03697096456692913,
+      "grad_norm": 1.126510500907898,
+      "learning_rate": 0.00019934765240889877,
+      "loss": 3.3636,
+      "step": 601
+    },
+    {
+      "epoch": 0.03703248031496063,
+      "grad_norm": 1.4110605716705322,
+      "learning_rate": 0.00019934544534867536,
+      "loss": 3.3863,
+      "step": 602
+    },
+    {
+      "epoch": 0.03709399606299212,
+      "grad_norm": 1.2283625602722168,
+      "learning_rate": 0.00019934323457348628,
+      "loss": 3.6643,
+      "step": 603
+    },
+    {
+      "epoch": 0.037155511811023625,
+      "grad_norm": 1.350063681602478,
+      "learning_rate": 0.00019934102008341417,
+      "loss": 3.1657,
+      "step": 604
+    },
+    {
+      "epoch": 0.03721702755905512,
+      "grad_norm": 0.9155988693237305,
+      "learning_rate": 0.0001993388018785418,
+      "loss": 3.8764,
+      "step": 605
+    },
+    {
+      "epoch": 0.03727854330708662,
+      "grad_norm": 0.7571907043457031,
+      "learning_rate": 0.00019933657995895224,
+      "loss": 3.7319,
+      "step": 606
+    },
+    {
+      "epoch": 0.03734005905511811,
+      "grad_norm": 1.08095383644104,
+      "learning_rate": 0.00019933435432472848,
+      "loss": 3.4114,
+      "step": 607
+    },
+    {
+      "epoch": 0.03740157480314961,
+      "grad_norm": 1.051662564277649,
+      "learning_rate": 0.00019933212497595378,
+      "loss": 3.5098,
+      "step": 608
+    },
+    {
+      "epoch": 0.037463090551181105,
+      "grad_norm": 0.868108332157135,
+      "learning_rate": 0.00019932989191271148,
+      "loss": 3.1834,
+      "step": 609
+    },
+    {
+      "epoch": 0.0375246062992126,
+      "grad_norm": 0.7951144576072693,
+      "learning_rate": 0.0001993276551350851,
+      "loss": 3.5424,
+      "step": 610
+    },
+    {
+      "epoch": 0.0375861220472441,
+      "grad_norm": 0.9019333124160767,
+      "learning_rate": 0.0001993254146431583,
+      "loss": 3.8545,
+      "step": 611
+    },
+    {
+      "epoch": 0.03764763779527559,
+      "grad_norm": 0.8380983471870422,
+      "learning_rate": 0.00019932317043701482,
+      "loss": 3.3224,
+      "step": 612
+    },
+    {
+      "epoch": 0.03770915354330709,
+      "grad_norm": 1.1253536939620972,
+      "learning_rate": 0.00019932092251673863,
+      "loss": 3.0683,
+      "step": 613
+    },
+    {
+      "epoch": 0.037770669291338585,
+      "grad_norm": 0.751939594745636,
+      "learning_rate": 0.00019931867088241378,
+      "loss": 3.5026,
+      "step": 614
+    },
+    {
+      "epoch": 0.03783218503937008,
+      "grad_norm": 0.9006920456886292,
+      "learning_rate": 0.00019931641553412444,
+      "loss": 3.1451,
+      "step": 615
+    },
+    {
+      "epoch": 0.03789370078740158,
+      "grad_norm": 0.8925879597663879,
+      "learning_rate": 0.00019931415647195494,
+      "loss": 3.5349,
+      "step": 616
+    },
+    {
+      "epoch": 0.03795521653543307,
+      "grad_norm": 1.1201438903808594,
+      "learning_rate": 0.00019931189369598983,
+      "loss": 3.4925,
+      "step": 617
+    },
+    {
+      "epoch": 0.03801673228346457,
+      "grad_norm": 1.0332320928573608,
+      "learning_rate": 0.0001993096272063136,
+      "loss": 3.1053,
+      "step": 618
+    },
+    {
+      "epoch": 0.038078248031496065,
+      "grad_norm": 0.9431754946708679,
+      "learning_rate": 0.00019930735700301116,
+      "loss": 3.0415,
+      "step": 619
+    },
+    {
+      "epoch": 0.03813976377952756,
+      "grad_norm": 0.9468442797660828,
+      "learning_rate": 0.00019930508308616728,
+      "loss": 3.2477,
+      "step": 620
+    },
+    {
+      "epoch": 0.03820127952755906,
+      "grad_norm": 0.747909426689148,
+      "learning_rate": 0.00019930280545586704,
+      "loss": 3.5828,
+      "step": 621
+    },
+    {
+      "epoch": 0.03826279527559055,
+      "grad_norm": 0.9899452924728394,
+      "learning_rate": 0.00019930052411219558,
+      "loss": 3.871,
+      "step": 622
+    },
+    {
+      "epoch": 0.03832431102362205,
+      "grad_norm": 0.9079659581184387,
+      "learning_rate": 0.00019929823905523827,
+      "loss": 3.4413,
+      "step": 623
+    },
+    {
+      "epoch": 0.038385826771653545,
+      "grad_norm": 0.9257645606994629,
+      "learning_rate": 0.0001992959502850805,
+      "loss": 3.7049,
+      "step": 624
+    },
+    {
+      "epoch": 0.03844734251968504,
+      "grad_norm": 0.7245174646377563,
+      "learning_rate": 0.00019929365780180786,
+      "loss": 3.502,
+      "step": 625
+    },
+    {
+      "epoch": 0.038508858267716536,
+      "grad_norm": 1.024824857711792,
+      "learning_rate": 0.00019929136160550613,
+      "loss": 3.563,
+      "step": 626
+    },
+    {
+      "epoch": 0.03857037401574803,
+      "grad_norm": 1.0008302927017212,
+      "learning_rate": 0.00019928906169626115,
+      "loss": 3.7503,
+      "step": 627
+    },
+    {
+      "epoch": 0.03863188976377953,
+      "grad_norm": 1.0077368021011353,
+      "learning_rate": 0.00019928675807415886,
+      "loss": 3.5193,
+      "step": 628
+    },
+    {
+      "epoch": 0.038693405511811024,
+      "grad_norm": 1.2073721885681152,
+      "learning_rate": 0.00019928445073928548,
+      "loss": 3.7818,
+      "step": 629
+    },
+    {
+      "epoch": 0.03875492125984252,
+      "grad_norm": 1.098296880722046,
+      "learning_rate": 0.00019928213969172727,
+      "loss": 3.5358,
+      "step": 630
+    },
+    {
+      "epoch": 0.038816437007874016,
+      "grad_norm": 1.0128798484802246,
+      "learning_rate": 0.00019927982493157066,
+      "loss": 3.5823,
+      "step": 631
+    },
+    {
+      "epoch": 0.03887795275590551,
+      "grad_norm": 1.2079495191574097,
+      "learning_rate": 0.00019927750645890217,
+      "loss": 3.1283,
+      "step": 632
+    },
+    {
+      "epoch": 0.03893946850393701,
+      "grad_norm": 1.0149550437927246,
+      "learning_rate": 0.0001992751842738086,
+      "loss": 3.761,
+      "step": 633
+    },
+    {
+      "epoch": 0.039000984251968504,
+      "grad_norm": 1.210645079612732,
+      "learning_rate": 0.00019927285837637664,
+      "loss": 3.7968,
+      "step": 634
+    },
+    {
+      "epoch": 0.0390625,
+      "grad_norm": 1.5798389911651611,
+      "learning_rate": 0.00019927052876669333,
+      "loss": 3.2541,
+      "step": 635
+    },
+    {
+      "epoch": 0.039124015748031496,
+      "grad_norm": 1.1927201747894287,
+      "learning_rate": 0.00019926819544484583,
+      "loss": 1.7775,
+      "step": 636
+    },
+    {
+      "epoch": 0.03918553149606299,
+      "grad_norm": 0.9827061891555786,
+      "learning_rate": 0.00019926585841092132,
+      "loss": 3.2568,
+      "step": 637
+    },
+    {
+      "epoch": 0.03924704724409449,
+      "grad_norm": 1.1455357074737549,
+      "learning_rate": 0.00019926351766500725,
+      "loss": 3.9129,
+      "step": 638
+    },
+    {
+      "epoch": 0.039308562992125984,
+      "grad_norm": 0.9932321906089783,
+      "learning_rate": 0.00019926117320719113,
+      "loss": 4.396,
+      "step": 639
+    },
+    {
+      "epoch": 0.03937007874015748,
+      "grad_norm": 1.0183807611465454,
+      "learning_rate": 0.0001992588250375606,
+      "loss": 2.4888,
+      "step": 640
+    },
+    {
+      "epoch": 0.039431594488188976,
+      "grad_norm": 0.986886739730835,
+      "learning_rate": 0.00019925647315620354,
+      "loss": 3.7306,
+      "step": 641
+    },
+    {
+      "epoch": 0.03949311023622047,
+      "grad_norm": 1.146675944328308,
+      "learning_rate": 0.00019925411756320784,
+      "loss": 3.777,
+      "step": 642
+    },
+    {
+      "epoch": 0.03955462598425197,
+      "grad_norm": 0.9272331595420837,
+      "learning_rate": 0.00019925175825866157,
+      "loss": 3.4764,
+      "step": 643
+    },
+    {
+      "epoch": 0.039616141732283464,
+      "grad_norm": 1.1497118473052979,
+      "learning_rate": 0.00019924939524265303,
+      "loss": 3.7044,
+      "step": 644
+    },
+    {
+      "epoch": 0.03967765748031496,
+      "grad_norm": 1.1471151113510132,
+      "learning_rate": 0.00019924702851527052,
+      "loss": 3.0837,
+      "step": 645
+    },
+    {
+      "epoch": 0.039739173228346455,
+      "grad_norm": 0.945509135723114,
+      "learning_rate": 0.00019924465807660252,
+      "loss": 3.4802,
+      "step": 646
+    },
+    {
+      "epoch": 0.03980068897637795,
+      "grad_norm": 1.1797497272491455,
+      "learning_rate": 0.00019924228392673773,
+      "loss": 3.828,
+      "step": 647
+    },
+    {
+      "epoch": 0.03986220472440945,
+      "grad_norm": 0.9096627831459045,
+      "learning_rate": 0.00019923990606576496,
+      "loss": 3.4434,
+      "step": 648
+    },
+    {
+      "epoch": 0.03992372047244094,
+      "grad_norm": 0.9126327037811279,
+      "learning_rate": 0.00019923752449377302,
+      "loss": 3.4173,
+      "step": 649
+    },
+    {
+      "epoch": 0.03998523622047244,
+      "grad_norm": 1.218599796295166,
+      "learning_rate": 0.00019923513921085107,
+      "loss": 3.4591,
+      "step": 650
+    },
+    {
+      "epoch": 0.040046751968503935,
+      "grad_norm": 0.9062703847885132,
+      "learning_rate": 0.00019923275021708825,
+      "loss": 3.5227,
+      "step": 651
+    },
+    {
+      "epoch": 0.04010826771653543,
+      "grad_norm": 1.948494791984558,
+      "learning_rate": 0.00019923035751257388,
+      "loss": 3.9037,
+      "step": 652
+    },
+    {
+      "epoch": 0.04016978346456693,
+      "grad_norm": 1.0055313110351562,
+      "learning_rate": 0.0001992279610973975,
+      "loss": 3.2988,
+      "step": 653
+    },
+    {
+      "epoch": 0.04023129921259842,
+      "grad_norm": 1.0900397300720215,
+      "learning_rate": 0.0001992255609716487,
+      "loss": 3.7328,
+      "step": 654
+    },
+    {
+      "epoch": 0.04029281496062992,
+      "grad_norm": 0.7427022457122803,
+      "learning_rate": 0.00019922315713541717,
+      "loss": 3.2298,
+      "step": 655
+    },
+    {
+      "epoch": 0.040354330708661415,
+      "grad_norm": 0.9899792075157166,
+      "learning_rate": 0.00019922074958879287,
+      "loss": 3.2036,
+      "step": 656
+    },
+    {
+      "epoch": 0.04041584645669291,
+      "grad_norm": 1.4752737283706665,
+      "learning_rate": 0.0001992183383318658,
+      "loss": 2.7682,
+      "step": 657
+    },
+    {
+      "epoch": 0.04047736220472441,
+      "grad_norm": 1.0303484201431274,
+      "learning_rate": 0.00019921592336472613,
+      "loss": 3.2061,
+      "step": 658
+    },
+    {
+      "epoch": 0.0405388779527559,
+      "grad_norm": 0.8618601560592651,
+      "learning_rate": 0.0001992135046874642,
+      "loss": 3.7312,
+      "step": 659
+    },
+    {
+      "epoch": 0.0406003937007874,
+      "grad_norm": 1.1078382730484009,
+      "learning_rate": 0.00019921108230017037,
+      "loss": 3.5266,
+      "step": 660
+    },
+    {
+      "epoch": 0.040661909448818895,
+      "grad_norm": 0.86662757396698,
+      "learning_rate": 0.0001992086562029353,
+      "loss": 3.82,
+      "step": 661
+    },
+    {
+      "epoch": 0.04072342519685039,
+      "grad_norm": 1.2535635232925415,
+      "learning_rate": 0.00019920622639584972,
+      "loss": 4.0284,
+      "step": 662
+    },
+    {
+      "epoch": 0.04078494094488189,
+      "grad_norm": 0.9014574885368347,
+      "learning_rate": 0.00019920379287900443,
+      "loss": 3.8245,
+      "step": 663
+    },
+    {
+      "epoch": 0.04084645669291338,
+      "grad_norm": 1.6962729692459106,
+      "learning_rate": 0.00019920135565249048,
+      "loss": 2.7894,
+      "step": 664
+    },
+    {
+      "epoch": 0.04090797244094488,
+      "grad_norm": 1.0608925819396973,
+      "learning_rate": 0.00019919891471639897,
+      "loss": 3.8036,
+      "step": 665
+    },
+    {
+      "epoch": 0.040969488188976375,
+      "grad_norm": 1.1027709245681763,
+      "learning_rate": 0.00019919647007082123,
+      "loss": 3.385,
+      "step": 666
+    },
+    {
+      "epoch": 0.04103100393700788,
+      "grad_norm": 1.2238614559173584,
+      "learning_rate": 0.0001991940217158486,
+      "loss": 3.4921,
+      "step": 667
+    },
+    {
+      "epoch": 0.04109251968503937,
+      "grad_norm": 0.8270302414894104,
+      "learning_rate": 0.0001991915696515727,
+      "loss": 3.4117,
+      "step": 668
+    },
+    {
+      "epoch": 0.04115403543307087,
+      "grad_norm": 0.9727601408958435,
+      "learning_rate": 0.00019918911387808522,
+      "loss": 3.2665,
+      "step": 669
+    },
+    {
+      "epoch": 0.041215551181102365,
+      "grad_norm": 1.5286084413528442,
+      "learning_rate": 0.00019918665439547795,
+      "loss": 3.4874,
+      "step": 670
+    },
+    {
+      "epoch": 0.04127706692913386,
+      "grad_norm": 1.2479138374328613,
+      "learning_rate": 0.0001991841912038429,
+      "loss": 2.8135,
+      "step": 671
+    },
+    {
+      "epoch": 0.04133858267716536,
+      "grad_norm": 1.5520782470703125,
+      "learning_rate": 0.00019918172430327216,
+      "loss": 3.2049,
+      "step": 672
+    },
+    {
+      "epoch": 0.04140009842519685,
+      "grad_norm": 0.9960653185844421,
+      "learning_rate": 0.00019917925369385798,
+      "loss": 3.8693,
+      "step": 673
+    },
+    {
+      "epoch": 0.04146161417322835,
+      "grad_norm": 1.1939448118209839,
+      "learning_rate": 0.00019917677937569277,
+      "loss": 3.0799,
+      "step": 674
+    },
+    {
+      "epoch": 0.041523129921259845,
+      "grad_norm": 1.0221312046051025,
+      "learning_rate": 0.000199174301348869,
+      "loss": 3.3923,
+      "step": 675
+    },
+    {
+      "epoch": 0.04158464566929134,
+      "grad_norm": 1.2140830755233765,
+      "learning_rate": 0.0001991718196134794,
+      "loss": 3.2849,
+      "step": 676
+    },
+    {
+      "epoch": 0.04164616141732284,
+      "grad_norm": 1.0457992553710938,
+      "learning_rate": 0.00019916933416961672,
+      "loss": 3.9432,
+      "step": 677
+    },
+    {
+      "epoch": 0.04170767716535433,
+      "grad_norm": 0.9691226482391357,
+      "learning_rate": 0.00019916684501737394,
+      "loss": 3.1696,
+      "step": 678
+    },
+    {
+      "epoch": 0.04176919291338583,
+      "grad_norm": 1.0362213850021362,
+      "learning_rate": 0.00019916435215684412,
+      "loss": 3.449,
+      "step": 679
+    },
+    {
+      "epoch": 0.041830708661417325,
+      "grad_norm": 1.5489195585250854,
+      "learning_rate": 0.0001991618555881205,
+      "loss": 3.7896,
+      "step": 680
+    },
+    {
+      "epoch": 0.04189222440944882,
+      "grad_norm": 0.8549931049346924,
+      "learning_rate": 0.00019915935531129644,
+      "loss": 3.6495,
+      "step": 681
+    },
+    {
+      "epoch": 0.04195374015748032,
+      "grad_norm": 0.8777815699577332,
+      "learning_rate": 0.00019915685132646538,
+      "loss": 3.8542,
+      "step": 682
+    },
+    {
+      "epoch": 0.04201525590551181,
+      "grad_norm": 1.2030718326568604,
+      "learning_rate": 0.000199154343633721,
+      "loss": 3.6662,
+      "step": 683
+    },
+    {
+      "epoch": 0.04207677165354331,
+      "grad_norm": 0.8463363647460938,
+      "learning_rate": 0.00019915183223315707,
+      "loss": 3.5805,
+      "step": 684
+    },
+    {
+      "epoch": 0.042138287401574805,
+      "grad_norm": 0.9015154838562012,
+      "learning_rate": 0.00019914931712486752,
+      "loss": 3.5249,
+      "step": 685
+    },
+    {
+      "epoch": 0.0421998031496063,
+      "grad_norm": 0.918318510055542,
+      "learning_rate": 0.00019914679830894635,
+      "loss": 3.4087,
+      "step": 686
+    },
+    {
+      "epoch": 0.0422613188976378,
+      "grad_norm": 0.9999191164970398,
+      "learning_rate": 0.00019914427578548782,
+      "loss": 3.2918,
+      "step": 687
+    },
+    {
+      "epoch": 0.04232283464566929,
+      "grad_norm": 1.3478078842163086,
+      "learning_rate": 0.0001991417495545862,
+      "loss": 3.0957,
+      "step": 688
+    },
+    {
+      "epoch": 0.04238435039370079,
+      "grad_norm": 0.8632199764251709,
+      "learning_rate": 0.00019913921961633597,
+      "loss": 3.8523,
+      "step": 689
+    },
+    {
+      "epoch": 0.042445866141732284,
+      "grad_norm": 0.7963219881057739,
+      "learning_rate": 0.00019913668597083172,
+      "loss": 3.7499,
+      "step": 690
+    },
+    {
+      "epoch": 0.04250738188976378,
+      "grad_norm": 1.2585116624832153,
+      "learning_rate": 0.00019913414861816824,
+      "loss": 3.3049,
+      "step": 691
+    },
+    {
+      "epoch": 0.042568897637795276,
+      "grad_norm": 0.8669255375862122,
+      "learning_rate": 0.0001991316075584404,
+      "loss": 3.9059,
+      "step": 692
+    },
+    {
+      "epoch": 0.04263041338582677,
+      "grad_norm": 0.9403530359268188,
+      "learning_rate": 0.00019912906279174322,
+      "loss": 2.3578,
+      "step": 693
+    },
+    {
+      "epoch": 0.04269192913385827,
+      "grad_norm": 1.0563639402389526,
+      "learning_rate": 0.0001991265143181718,
+      "loss": 3.474,
+      "step": 694
+    },
+    {
+      "epoch": 0.042753444881889764,
+      "grad_norm": 1.5698355436325073,
+      "learning_rate": 0.00019912396213782153,
+      "loss": 3.1837,
+      "step": 695
+    },
+    {
+      "epoch": 0.04281496062992126,
+      "grad_norm": 1.4085670709609985,
+      "learning_rate": 0.00019912140625078778,
+      "loss": 2.8907,
+      "step": 696
+    },
+    {
+      "epoch": 0.042876476377952756,
+      "grad_norm": 0.972629964351654,
+      "learning_rate": 0.00019911884665716617,
+      "loss": 3.3064,
+      "step": 697
+    },
+    {
+      "epoch": 0.04293799212598425,
+      "grad_norm": 0.8210000395774841,
+      "learning_rate": 0.00019911628335705242,
+      "loss": 3.652,
+      "step": 698
+    },
+    {
+      "epoch": 0.04299950787401575,
+      "grad_norm": 1.0592423677444458,
+      "learning_rate": 0.0001991137163505423,
+      "loss": 3.3931,
+      "step": 699
+    },
+    {
+      "epoch": 0.043061023622047244,
+      "grad_norm": 1.0382250547409058,
+      "learning_rate": 0.00019911114563773193,
+      "loss": 2.7117,
+      "step": 700
+    },
+    {
+      "epoch": 0.04312253937007874,
+      "grad_norm": 0.7692904472351074,
+      "learning_rate": 0.00019910857121871732,
+      "loss": 3.3651,
+      "step": 701
+    },
+    {
+      "epoch": 0.043184055118110236,
+      "grad_norm": 1.1790968179702759,
+      "learning_rate": 0.00019910599309359484,
+      "loss": 3.5922,
+      "step": 702
+    },
+    {
+      "epoch": 0.04324557086614173,
+      "grad_norm": 1.1439062356948853,
+      "learning_rate": 0.00019910341126246078,
+      "loss": 3.0696,
+      "step": 703
+    },
+    {
+      "epoch": 0.04330708661417323,
+      "grad_norm": 0.9323583841323853,
+      "learning_rate": 0.0001991008257254118,
+      "loss": 3.8051,
+      "step": 704
+    },
+    {
+      "epoch": 0.043368602362204724,
+      "grad_norm": 0.9486002326011658,
+      "learning_rate": 0.00019909823648254452,
+      "loss": 3.1799,
+      "step": 705
+    },
+    {
+      "epoch": 0.04343011811023622,
+      "grad_norm": 1.1196471452713013,
+      "learning_rate": 0.00019909564353395582,
+      "loss": 3.81,
+      "step": 706
+    },
+    {
+      "epoch": 0.043491633858267716,
+      "grad_norm": 0.9311375617980957,
+      "learning_rate": 0.0001990930468797426,
+      "loss": 2.0111,
+      "step": 707
+    },
+    {
+      "epoch": 0.04355314960629921,
+      "grad_norm": 1.080230712890625,
+      "learning_rate": 0.00019909044652000198,
+      "loss": 3.2013,
+      "step": 708
+    },
+    {
+      "epoch": 0.04361466535433071,
+      "grad_norm": 0.9653453230857849,
+      "learning_rate": 0.0001990878424548312,
+      "loss": 2.9156,
+      "step": 709
+    },
+    {
+      "epoch": 0.0436761811023622,
+      "grad_norm": 0.8740338087081909,
+      "learning_rate": 0.00019908523468432766,
+      "loss": 3.005,
+      "step": 710
+    },
+    {
+      "epoch": 0.0437376968503937,
+      "grad_norm": 1.1873654127120972,
+      "learning_rate": 0.00019908262320858885,
+      "loss": 2.7852,
+      "step": 711
+    },
+    {
+      "epoch": 0.043799212598425195,
+      "grad_norm": 1.3769781589508057,
+      "learning_rate": 0.00019908000802771242,
+      "loss": 2.4786,
+      "step": 712
+    },
+    {
+      "epoch": 0.04386072834645669,
+      "grad_norm": 1.0197645425796509,
+      "learning_rate": 0.00019907738914179616,
+      "loss": 3.5079,
+      "step": 713
+    },
+    {
+      "epoch": 0.04392224409448819,
+      "grad_norm": 1.0652419328689575,
+      "learning_rate": 0.00019907476655093805,
+      "loss": 3.6079,
+      "step": 714
+    },
+    {
+      "epoch": 0.04398375984251968,
+      "grad_norm": 1.560143232345581,
+      "learning_rate": 0.00019907214025523612,
+      "loss": 3.249,
+      "step": 715
+    },
+    {
+      "epoch": 0.04404527559055118,
+      "grad_norm": 1.395227313041687,
+      "learning_rate": 0.0001990695102547886,
+      "loss": 3.2603,
+      "step": 716
+    },
+    {
+      "epoch": 0.044106791338582675,
+      "grad_norm": 0.834183931350708,
+      "learning_rate": 0.00019906687654969377,
+      "loss": 3.2682,
+      "step": 717
+    },
+    {
+      "epoch": 0.04416830708661417,
+      "grad_norm": 0.9044603705406189,
+      "learning_rate": 0.0001990642391400502,
+      "loss": 3.4532,
+      "step": 718
+    },
+    {
+      "epoch": 0.04422982283464567,
+      "grad_norm": 0.6512154340744019,
+      "learning_rate": 0.00019906159802595652,
+      "loss": 3.6596,
+      "step": 719
+    },
+    {
+      "epoch": 0.04429133858267716,
+      "grad_norm": 1.0617462396621704,
+      "learning_rate": 0.00019905895320751143,
+      "loss": 3.2036,
+      "step": 720
+    },
+    {
+      "epoch": 0.04435285433070866,
+      "grad_norm": 0.8350668549537659,
+      "learning_rate": 0.00019905630468481384,
+      "loss": 3.3692,
+      "step": 721
+    },
+    {
+      "epoch": 0.044414370078740155,
+      "grad_norm": 1.4038469791412354,
+      "learning_rate": 0.00019905365245796283,
+      "loss": 3.9514,
+      "step": 722
+    },
+    {
+      "epoch": 0.04447588582677165,
+      "grad_norm": 1.0058015584945679,
+      "learning_rate": 0.00019905099652705756,
+      "loss": 3.0317,
+      "step": 723
+    },
+    {
+      "epoch": 0.04453740157480315,
+      "grad_norm": 1.2147352695465088,
+      "learning_rate": 0.0001990483368921973,
+      "loss": 3.7601,
+      "step": 724
+    },
+    {
+      "epoch": 0.04459891732283464,
+      "grad_norm": 1.2208256721496582,
+      "learning_rate": 0.0001990456735534816,
+      "loss": 3.5044,
+      "step": 725
+    },
+    {
+      "epoch": 0.04466043307086614,
+      "grad_norm": 0.8608539700508118,
+      "learning_rate": 0.00019904300651100995,
+      "loss": 2.9741,
+      "step": 726
+    },
+    {
+      "epoch": 0.044721948818897635,
+      "grad_norm": 1.1605843305587769,
+      "learning_rate": 0.0001990403357648822,
+      "loss": 2.6462,
+      "step": 727
+    },
+    {
+      "epoch": 0.04478346456692913,
+      "grad_norm": 1.0630160570144653,
+      "learning_rate": 0.00019903766131519813,
+      "loss": 2.8357,
+      "step": 728
+    },
+    {
+      "epoch": 0.04484498031496063,
+      "grad_norm": 0.9385619759559631,
+      "learning_rate": 0.00019903498316205777,
+      "loss": 2.9686,
+      "step": 729
+    },
+    {
+      "epoch": 0.04490649606299212,
+      "grad_norm": 0.804070234298706,
+      "learning_rate": 0.00019903230130556126,
+      "loss": 3.6256,
+      "step": 730
+    },
+    {
+      "epoch": 0.044968011811023625,
+      "grad_norm": 1.0888487100601196,
+      "learning_rate": 0.00019902961574580892,
+      "loss": 3.2784,
+      "step": 731
+    },
+    {
+      "epoch": 0.04502952755905512,
+      "grad_norm": 1.000158429145813,
+      "learning_rate": 0.00019902692648290113,
+      "loss": 3.3393,
+      "step": 732
+    },
+    {
+      "epoch": 0.04509104330708662,
+      "grad_norm": 0.8283302783966064,
+      "learning_rate": 0.00019902423351693852,
+      "loss": 4.0251,
+      "step": 733
+    },
+    {
+      "epoch": 0.04515255905511811,
+      "grad_norm": 1.0357900857925415,
+      "learning_rate": 0.00019902153684802173,
+      "loss": 3.1526,
+      "step": 734
+    },
+    {
+      "epoch": 0.04521407480314961,
+      "grad_norm": 1.0264791250228882,
+      "learning_rate": 0.00019901883647625162,
+      "loss": 2.9619,
+      "step": 735
+    },
+    {
+      "epoch": 0.045275590551181105,
+      "grad_norm": 0.8597986698150635,
+      "learning_rate": 0.00019901613240172914,
+      "loss": 3.3135,
+      "step": 736
+    },
+    {
+      "epoch": 0.0453371062992126,
+      "grad_norm": 0.9346214532852173,
+      "learning_rate": 0.0001990134246245555,
+      "loss": 3.0705,
+      "step": 737
+    },
+    {
+      "epoch": 0.0453986220472441,
+      "grad_norm": 0.6926522850990295,
+      "learning_rate": 0.00019901071314483185,
+      "loss": 3.2362,
+      "step": 738
+    },
+    {
+      "epoch": 0.04546013779527559,
+      "grad_norm": 1.4668114185333252,
+      "learning_rate": 0.0001990079979626596,
+      "loss": 4.098,
+      "step": 739
+    },
+    {
+      "epoch": 0.04552165354330709,
+      "grad_norm": 0.8899300694465637,
+      "learning_rate": 0.00019900527907814038,
+      "loss": 3.5201,
+      "step": 740
+    },
+    {
+      "epoch": 0.045583169291338585,
+      "grad_norm": 1.0781344175338745,
+      "learning_rate": 0.00019900255649137573,
+      "loss": 2.6374,
+      "step": 741
+    },
+    {
+      "epoch": 0.04564468503937008,
+      "grad_norm": 0.9585447907447815,
+      "learning_rate": 0.00019899983020246756,
+      "loss": 3.3033,
+      "step": 742
+    },
+    {
+      "epoch": 0.04570620078740158,
+      "grad_norm": 1.1942914724349976,
+      "learning_rate": 0.00019899710021151776,
+      "loss": 3.2966,
+      "step": 743
+    },
+    {
+      "epoch": 0.04576771653543307,
+      "grad_norm": 1.4991570711135864,
+      "learning_rate": 0.00019899436651862846,
+      "loss": 3.7169,
+      "step": 744
+    },
+    {
+      "epoch": 0.04582923228346457,
+      "grad_norm": 1.5413933992385864,
+      "learning_rate": 0.00019899162912390182,
+      "loss": 3.5206,
+      "step": 745
+    },
+    {
+      "epoch": 0.045890748031496065,
+      "grad_norm": 1.7879767417907715,
+      "learning_rate": 0.00019898888802744028,
+      "loss": 2.5892,
+      "step": 746
+    },
+    {
+      "epoch": 0.04595226377952756,
+      "grad_norm": 1.1545370817184448,
+      "learning_rate": 0.0001989861432293463,
+      "loss": 3.0143,
+      "step": 747
+    },
+    {
+      "epoch": 0.04601377952755906,
+      "grad_norm": 1.102148175239563,
+      "learning_rate": 0.00019898339472972253,
+      "loss": 3.2997,
+      "step": 748
+    },
+    {
+      "epoch": 0.04607529527559055,
+      "grad_norm": 1.679919958114624,
+      "learning_rate": 0.00019898064252867173,
+      "loss": 2.7599,
+      "step": 749
+    },
+    {
+      "epoch": 0.04613681102362205,
+      "grad_norm": 0.8177815675735474,
+      "learning_rate": 0.00019897788662629682,
+      "loss": 3.8027,
+      "step": 750
+    },
+    {
+      "epoch": 0.046198326771653545,
+      "grad_norm": 1.5025945901870728,
+      "learning_rate": 0.00019897512702270087,
+      "loss": 2.4238,
+      "step": 751
+    },
+    {
+      "epoch": 0.04625984251968504,
+      "grad_norm": 1.8025085926055908,
+      "learning_rate": 0.00019897236371798708,
+      "loss": 1.9056,
+      "step": 752
+    },
+    {
+      "epoch": 0.046321358267716536,
+      "grad_norm": 1.1328459978103638,
+      "learning_rate": 0.00019896959671225877,
+      "loss": 3.3643,
+      "step": 753
+    },
+    {
+      "epoch": 0.04638287401574803,
+      "grad_norm": 2.4293572902679443,
+      "learning_rate": 0.0001989668260056194,
+      "loss": 2.6224,
+      "step": 754
+    },
+    {
+      "epoch": 0.04644438976377953,
+      "grad_norm": 1.1772873401641846,
+      "learning_rate": 0.0001989640515981726,
+      "loss": 3.6542,
+      "step": 755
+    },
+    {
+      "epoch": 0.046505905511811024,
+      "grad_norm": 1.0882432460784912,
+      "learning_rate": 0.00019896127349002213,
+      "loss": 3.2897,
+      "step": 756
+    },
+    {
+      "epoch": 0.04656742125984252,
+      "grad_norm": 0.8625539541244507,
+      "learning_rate": 0.00019895849168127182,
+      "loss": 3.673,
+      "step": 757
+    },
+    {
+      "epoch": 0.046628937007874016,
+      "grad_norm": 1.0152686834335327,
+      "learning_rate": 0.00019895570617202576,
+      "loss": 3.6141,
+      "step": 758
+    },
+    {
+      "epoch": 0.04669045275590551,
+      "grad_norm": 0.9116948246955872,
+      "learning_rate": 0.00019895291696238803,
+      "loss": 2.3906,
+      "step": 759
+    },
+    {
+      "epoch": 0.04675196850393701,
+      "grad_norm": 1.0996073484420776,
+      "learning_rate": 0.00019895012405246303,
+      "loss": 3.7847,
+      "step": 760
+    },
+    {
+      "epoch": 0.046813484251968504,
+      "grad_norm": 1.1193512678146362,
+      "learning_rate": 0.00019894732744235512,
+      "loss": 3.2236,
+      "step": 761
+    },
+    {
+      "epoch": 0.046875,
+      "grad_norm": 0.9301583766937256,
+      "learning_rate": 0.00019894452713216897,
+      "loss": 3.3093,
+      "step": 762
+    },
+    {
+      "epoch": 0.046936515748031496,
+      "grad_norm": 1.0054373741149902,
+      "learning_rate": 0.0001989417231220092,
+      "loss": 2.9063,
+      "step": 763
+    },
+    {
+      "epoch": 0.04699803149606299,
+      "grad_norm": 0.8930043578147888,
+      "learning_rate": 0.00019893891541198069,
+      "loss": 3.5456,
+      "step": 764
+    },
+    {
+      "epoch": 0.04705954724409449,
+      "grad_norm": 1.1790488958358765,
+      "learning_rate": 0.00019893610400218843,
+      "loss": 3.4562,
+      "step": 765
+    },
+    {
+      "epoch": 0.047121062992125984,
+      "grad_norm": 1.081071138381958,
+      "learning_rate": 0.0001989332888927376,
+      "loss": 3.7093,
+      "step": 766
+    },
+    {
+      "epoch": 0.04718257874015748,
+      "grad_norm": 1.0062880516052246,
+      "learning_rate": 0.00019893047008373338,
+      "loss": 3.2428,
+      "step": 767
+    },
+    {
+      "epoch": 0.047244094488188976,
+      "grad_norm": 1.0124411582946777,
+      "learning_rate": 0.00019892764757528127,
+      "loss": 3.3101,
+      "step": 768
+    },
+    {
+      "epoch": 0.04730561023622047,
+      "grad_norm": 0.8614514470100403,
+      "learning_rate": 0.00019892482136748676,
+      "loss": 4.3242,
+      "step": 769
+    },
+    {
+      "epoch": 0.04736712598425197,
+      "grad_norm": 1.3130532503128052,
+      "learning_rate": 0.00019892199146045555,
+      "loss": 2.7005,
+      "step": 770
+    },
+    {
+      "epoch": 0.047428641732283464,
+      "grad_norm": 1.038522720336914,
+      "learning_rate": 0.0001989191578542935,
+      "loss": 2.9556,
+      "step": 771
+    },
+    {
+      "epoch": 0.04749015748031496,
+      "grad_norm": 1.010750651359558,
+      "learning_rate": 0.00019891632054910648,
+      "loss": 3.5658,
+      "step": 772
+    },
+    {
+      "epoch": 0.047551673228346455,
+      "grad_norm": 1.1879174709320068,
+      "learning_rate": 0.00019891347954500068,
+      "loss": 3.0988,
+      "step": 773
+    },
+    {
+      "epoch": 0.04761318897637795,
+      "grad_norm": 0.996985912322998,
+      "learning_rate": 0.0001989106348420823,
+      "loss": 3.9607,
+      "step": 774
+    },
+    {
+      "epoch": 0.04767470472440945,
+      "grad_norm": 0.7668408751487732,
+      "learning_rate": 0.00019890778644045772,
+      "loss": 3.4647,
+      "step": 775
+    },
+    {
+      "epoch": 0.04773622047244094,
+      "grad_norm": 0.9622384905815125,
+      "learning_rate": 0.00019890493434023343,
+      "loss": 3.105,
+      "step": 776
+    },
+    {
+      "epoch": 0.04779773622047244,
+      "grad_norm": 1.4763271808624268,
+      "learning_rate": 0.0001989020785415161,
+      "loss": 3.6981,
+      "step": 777
+    },
+    {
+      "epoch": 0.047859251968503935,
+      "grad_norm": 1.1392112970352173,
+      "learning_rate": 0.00019889921904441257,
+      "loss": 3.2574,
+      "step": 778
+    },
+    {
+      "epoch": 0.04792076771653543,
+      "grad_norm": 1.0541492700576782,
+      "learning_rate": 0.0001988963558490297,
+      "loss": 3.2582,
+      "step": 779
+    },
+    {
+      "epoch": 0.04798228346456693,
+      "grad_norm": 1.136851191520691,
+      "learning_rate": 0.00019889348895547456,
+      "loss": 3.2251,
+      "step": 780
+    },
+    {
+      "epoch": 0.04804379921259842,
+      "grad_norm": 1.1861759424209595,
+      "learning_rate": 0.0001988906183638544,
+      "loss": 3.225,
+      "step": 781
+    },
+    {
+      "epoch": 0.04810531496062992,
+      "grad_norm": 0.9999700784683228,
+      "learning_rate": 0.00019888774407427655,
+      "loss": 3.6837,
+      "step": 782
+    },
+    {
+      "epoch": 0.048166830708661415,
+      "grad_norm": 0.9262111186981201,
+      "learning_rate": 0.00019888486608684845,
+      "loss": 3.4239,
+      "step": 783
+    },
+    {
+      "epoch": 0.04822834645669291,
+      "grad_norm": 1.036895513534546,
+      "learning_rate": 0.0001988819844016778,
+      "loss": 3.4572,
+      "step": 784
+    },
+    {
+      "epoch": 0.04828986220472441,
+      "grad_norm": 1.0169262886047363,
+      "learning_rate": 0.0001988790990188723,
+      "loss": 3.7527,
+      "step": 785
+    },
+    {
+      "epoch": 0.0483513779527559,
+      "grad_norm": 0.973526656627655,
+      "learning_rate": 0.00019887620993853987,
+      "loss": 3.5371,
+      "step": 786
+    },
+    {
+      "epoch": 0.0484128937007874,
+      "grad_norm": 0.8566667437553406,
+      "learning_rate": 0.0001988733171607885,
+      "loss": 3.8931,
+      "step": 787
+    },
+    {
+      "epoch": 0.048474409448818895,
+      "grad_norm": 1.3148434162139893,
+      "learning_rate": 0.0001988704206857264,
+      "loss": 3.6132,
+      "step": 788
+    },
+    {
+      "epoch": 0.04853592519685039,
+      "grad_norm": 0.8923270106315613,
+      "learning_rate": 0.00019886752051346193,
+      "loss": 3.7041,
+      "step": 789
+    },
+    {
+      "epoch": 0.04859744094488189,
+      "grad_norm": 1.2122759819030762,
+      "learning_rate": 0.00019886461664410347,
+      "loss": 3.2535,
+      "step": 790
+    },
+    {
+      "epoch": 0.04865895669291338,
+      "grad_norm": 1.268984317779541,
+      "learning_rate": 0.00019886170907775964,
+      "loss": 3.0127,
+      "step": 791
+    },
+    {
+      "epoch": 0.04872047244094488,
+      "grad_norm": 1.2826154232025146,
+      "learning_rate": 0.00019885879781453912,
+      "loss": 3.593,
+      "step": 792
+    },
+    {
+      "epoch": 0.048781988188976375,
+      "grad_norm": 1.115561604499817,
+      "learning_rate": 0.0001988558828545509,
+      "loss": 3.3061,
+      "step": 793
+    },
+    {
+      "epoch": 0.04884350393700788,
+      "grad_norm": 0.8783797025680542,
+      "learning_rate": 0.00019885296419790382,
+      "loss": 3.5265,
+      "step": 794
+    },
+    {
+      "epoch": 0.04890501968503937,
+      "grad_norm": 0.8827127814292908,
+      "learning_rate": 0.0001988500418447071,
+      "loss": 3.3861,
+      "step": 795
+    },
+    {
+      "epoch": 0.04896653543307087,
+      "grad_norm": 1.7411853075027466,
+      "learning_rate": 0.00019884711579507005,
+      "loss": 3.7788,
+      "step": 796
+    },
+    {
+      "epoch": 0.049028051181102365,
+      "grad_norm": 1.0998561382293701,
+      "learning_rate": 0.00019884418604910204,
+      "loss": 3.4207,
+      "step": 797
+    },
+    {
+      "epoch": 0.04908956692913386,
+      "grad_norm": 1.1624531745910645,
+      "learning_rate": 0.00019884125260691263,
+      "loss": 3.167,
+      "step": 798
+    },
+    {
+      "epoch": 0.04915108267716536,
+      "grad_norm": 1.0572528839111328,
+      "learning_rate": 0.00019883831546861158,
+      "loss": 3.9809,
+      "step": 799
+    },
+    {
+      "epoch": 0.04921259842519685,
+      "grad_norm": 0.9654624462127686,
+      "learning_rate": 0.00019883537463430862,
+      "loss": 3.4965,
+      "step": 800
+    },
+    {
+      "epoch": 0.04927411417322835,
+      "grad_norm": 1.170025110244751,
+      "learning_rate": 0.00019883243010411378,
+      "loss": 2.8778,
+      "step": 801
+    },
+    {
+      "epoch": 0.049335629921259845,
+      "grad_norm": 0.8353591561317444,
+      "learning_rate": 0.00019882948187813714,
+      "loss": 3.3182,
+      "step": 802
+    },
+    {
+      "epoch": 0.04939714566929134,
+      "grad_norm": 0.9837372899055481,
+      "learning_rate": 0.000198826529956489,
+      "loss": 3.6289,
+      "step": 803
+    },
+    {
+      "epoch": 0.04945866141732284,
+      "grad_norm": 1.003989815711975,
+      "learning_rate": 0.00019882357433927968,
+      "loss": 3.2469,
+      "step": 804
+    },
+    {
+      "epoch": 0.04952017716535433,
+      "grad_norm": 0.790375828742981,
+      "learning_rate": 0.00019882061502661977,
+      "loss": 3.0491,
+      "step": 805
+    },
+    {
+      "epoch": 0.04958169291338583,
+      "grad_norm": 1.0488121509552002,
+      "learning_rate": 0.00019881765201861987,
+      "loss": 3.0511,
+      "step": 806
+    },
+    {
+      "epoch": 0.049643208661417325,
+      "grad_norm": 0.8292686343193054,
+      "learning_rate": 0.00019881468531539082,
+      "loss": 3.6685,
+      "step": 807
+    },
+    {
+      "epoch": 0.04970472440944882,
+      "grad_norm": 1.3160494565963745,
+      "learning_rate": 0.00019881171491704355,
+      "loss": 3.1075,
+      "step": 808
+    },
+    {
+      "epoch": 0.04976624015748032,
+      "grad_norm": 1.0498439073562622,
+      "learning_rate": 0.00019880874082368912,
+      "loss": 3.4956,
+      "step": 809
+    },
+    {
+      "epoch": 0.04982775590551181,
+      "grad_norm": 0.9402647018432617,
+      "learning_rate": 0.00019880576303543874,
+      "loss": 3.3499,
+      "step": 810
+    },
+    {
+      "epoch": 0.04988927165354331,
+      "grad_norm": 1.084028720855713,
+      "learning_rate": 0.0001988027815524038,
+      "loss": 3.2556,
+      "step": 811
+    },
+    {
+      "epoch": 0.049950787401574805,
+      "grad_norm": 1.1181734800338745,
+      "learning_rate": 0.00019879979637469578,
+      "loss": 3.2092,
+      "step": 812
+    },
+    {
+      "epoch": 0.0500123031496063,
+      "grad_norm": 0.8333975076675415,
+      "learning_rate": 0.00019879680750242628,
+      "loss": 3.7273,
+      "step": 813
+    },
+    {
+      "epoch": 0.0500738188976378,
+      "grad_norm": 1.2503973245620728,
+      "learning_rate": 0.0001987938149357071,
+      "loss": 3.7761,
+      "step": 814
+    },
+    {
+      "epoch": 0.05013533464566929,
+      "grad_norm": 0.9153504371643066,
+      "learning_rate": 0.0001987908186746501,
+      "loss": 3.2633,
+      "step": 815
+    },
+    {
+      "epoch": 0.05019685039370079,
+      "grad_norm": 0.9418584108352661,
+      "learning_rate": 0.00019878781871936739,
+      "loss": 2.8659,
+      "step": 816
+    },
+    {
+      "epoch": 0.050258366141732284,
+      "grad_norm": 1.429984450340271,
+      "learning_rate": 0.0001987848150699711,
+      "loss": 3.3354,
+      "step": 817
+    },
+    {
+      "epoch": 0.05031988188976378,
+      "grad_norm": 0.9990026354789734,
+      "learning_rate": 0.00019878180772657357,
+      "loss": 3.7051,
+      "step": 818
+    },
+    {
+      "epoch": 0.050381397637795276,
+      "grad_norm": 0.902008593082428,
+      "learning_rate": 0.00019877879668928725,
+      "loss": 3.286,
+      "step": 819
+    },
+    {
+      "epoch": 0.05044291338582677,
+      "grad_norm": 0.8928101062774658,
+      "learning_rate": 0.00019877578195822473,
+      "loss": 3.4377,
+      "step": 820
+    },
+    {
+      "epoch": 0.05050442913385827,
+      "grad_norm": 1.1649953126907349,
+      "learning_rate": 0.0001987727635334988,
+      "loss": 3.6478,
+      "step": 821
+    },
+    {
+      "epoch": 0.050565944881889764,
+      "grad_norm": 1.1050071716308594,
+      "learning_rate": 0.00019876974141522222,
+      "loss": 3.3032,
+      "step": 822
+    },
+    {
+      "epoch": 0.05062746062992126,
+      "grad_norm": 0.8179242014884949,
+      "learning_rate": 0.0001987667156035081,
+      "loss": 3.7137,
+      "step": 823
+    },
+    {
+      "epoch": 0.050688976377952756,
+      "grad_norm": 1.07375967502594,
+      "learning_rate": 0.00019876368609846954,
+      "loss": 3.6215,
+      "step": 824
+    },
+    {
+      "epoch": 0.05075049212598425,
+      "grad_norm": 1.2247053384780884,
+      "learning_rate": 0.00019876065290021987,
+      "loss": 3.8571,
+      "step": 825
+    },
+    {
+      "epoch": 0.05081200787401575,
+      "grad_norm": 1.3312256336212158,
+      "learning_rate": 0.00019875761600887245,
+      "loss": 3.7873,
+      "step": 826
+    },
+    {
+      "epoch": 0.050873523622047244,
+      "grad_norm": 0.8953959941864014,
+      "learning_rate": 0.0001987545754245409,
+      "loss": 3.7905,
+      "step": 827
+    },
+    {
+      "epoch": 0.05093503937007874,
+      "grad_norm": 1.047165870666504,
+      "learning_rate": 0.00019875153114733892,
+      "loss": 3.6548,
+      "step": 828
+    },
+    {
+      "epoch": 0.050996555118110236,
+      "grad_norm": 1.0614525079727173,
+      "learning_rate": 0.0001987484831773803,
+      "loss": 3.2494,
+      "step": 829
+    },
+    {
+      "epoch": 0.05105807086614173,
+      "grad_norm": 0.8585118055343628,
+      "learning_rate": 0.00019874543151477902,
+      "loss": 3.5713,
+      "step": 830
+    },
+    {
+      "epoch": 0.05111958661417323,
+      "grad_norm": 0.8474485874176025,
+      "learning_rate": 0.0001987423761596493,
+      "loss": 3.2901,
+      "step": 831
+    },
+    {
+      "epoch": 0.051181102362204724,
+      "grad_norm": 0.9042104482650757,
+      "learning_rate": 0.00019873931711210523,
+      "loss": 3.6511,
+      "step": 832
+    },
+    {
+      "epoch": 0.05124261811023622,
+      "grad_norm": 0.7890754342079163,
+      "learning_rate": 0.00019873625437226132,
+      "loss": 3.5091,
+      "step": 833
+    },
+    {
+      "epoch": 0.051304133858267716,
+      "grad_norm": 0.9915399551391602,
+      "learning_rate": 0.00019873318794023204,
+      "loss": 3.4043,
+      "step": 834
+    },
+    {
+      "epoch": 0.05136564960629921,
+      "grad_norm": 0.9173625111579895,
+      "learning_rate": 0.00019873011781613214,
+      "loss": 3.3737,
+      "step": 835
+    },
+    {
+      "epoch": 0.05142716535433071,
+      "grad_norm": 0.951904833316803,
+      "learning_rate": 0.00019872704400007634,
+      "loss": 3.5029,
+      "step": 836
+    },
+    {
+      "epoch": 0.0514886811023622,
+      "grad_norm": 1.28164541721344,
+      "learning_rate": 0.00019872396649217962,
+      "loss": 3.1841,
+      "step": 837
+    },
+    {
+      "epoch": 0.0515501968503937,
+      "grad_norm": 1.246439814567566,
+      "learning_rate": 0.000198720885292557,
+      "loss": 2.8562,
+      "step": 838
+    },
+    {
+      "epoch": 0.051611712598425195,
+      "grad_norm": 0.8655402660369873,
+      "learning_rate": 0.0001987178004013238,
+      "loss": 3.6548,
+      "step": 839
+    },
+    {
+      "epoch": 0.05167322834645669,
+      "grad_norm": 1.0161879062652588,
+      "learning_rate": 0.00019871471181859534,
+      "loss": 3.5581,
+      "step": 840
+    },
+    {
+      "epoch": 0.05173474409448819,
+      "grad_norm": 1.45500910282135,
+      "learning_rate": 0.0001987116195444871,
+      "loss": 3.6157,
+      "step": 841
+    },
+    {
+      "epoch": 0.05179625984251968,
+      "grad_norm": 0.864393413066864,
+      "learning_rate": 0.0001987085235791147,
+      "loss": 2.9813,
+      "step": 842
+    },
+    {
+      "epoch": 0.05185777559055118,
+      "grad_norm": 0.7716377377510071,
+      "learning_rate": 0.00019870542392259394,
+      "loss": 3.2567,
+      "step": 843
+    },
+    {
+      "epoch": 0.051919291338582675,
+      "grad_norm": 1.4415135383605957,
+      "learning_rate": 0.00019870232057504073,
+      "loss": 2.2901,
+      "step": 844
+    },
+    {
+      "epoch": 0.05198080708661417,
+      "grad_norm": 1.4320045709609985,
+      "learning_rate": 0.0001986992135365711,
+      "loss": 2.9525,
+      "step": 845
+    },
+    {
+      "epoch": 0.05204232283464567,
+      "grad_norm": 1.3689614534378052,
+      "learning_rate": 0.00019869610280730124,
+      "loss": 2.9506,
+      "step": 846
+    },
+    {
+      "epoch": 0.05210383858267716,
+      "grad_norm": 0.8497210144996643,
+      "learning_rate": 0.00019869298838734749,
+      "loss": 3.6556,
+      "step": 847
+    },
+    {
+      "epoch": 0.05216535433070866,
+      "grad_norm": 1.1966667175292969,
+      "learning_rate": 0.00019868987027682632,
+      "loss": 3.3893,
+      "step": 848
+    },
+    {
+      "epoch": 0.052226870078740155,
+      "grad_norm": 0.8147649168968201,
+      "learning_rate": 0.0001986867484758543,
+      "loss": 3.7478,
+      "step": 849
+    },
+    {
+      "epoch": 0.05228838582677165,
+      "grad_norm": 0.8276150226593018,
+      "learning_rate": 0.00019868362298454818,
+      "loss": 3.6995,
+      "step": 850
+    },
+    {
+      "epoch": 0.05234990157480315,
+      "grad_norm": 1.0400936603546143,
+      "learning_rate": 0.0001986804938030248,
+      "loss": 3.0143,
+      "step": 851
+    },
+    {
+      "epoch": 0.05241141732283464,
+      "grad_norm": 0.9328922629356384,
+      "learning_rate": 0.00019867736093140124,
+      "loss": 3.1922,
+      "step": 852
+    },
+    {
+      "epoch": 0.05247293307086614,
+      "grad_norm": 1.0343034267425537,
+      "learning_rate": 0.0001986742243697946,
+      "loss": 3.8028,
+      "step": 853
+    },
+    {
+      "epoch": 0.052534448818897635,
+      "grad_norm": 0.9345069527626038,
+      "learning_rate": 0.00019867108411832223,
+      "loss": 3.6094,
+      "step": 854
+    },
+    {
+      "epoch": 0.05259596456692913,
+      "grad_norm": 0.7672252058982849,
+      "learning_rate": 0.00019866794017710148,
+      "loss": 3.5104,
+      "step": 855
+    },
+    {
+      "epoch": 0.05265748031496063,
+      "grad_norm": 0.8580348491668701,
+      "learning_rate": 0.00019866479254625,
+      "loss": 3.4198,
+      "step": 856
+    },
+    {
+      "epoch": 0.05271899606299212,
+      "grad_norm": 0.8594534397125244,
+      "learning_rate": 0.0001986616412258854,
+      "loss": 3.4208,
+      "step": 857
+    },
+    {
+      "epoch": 0.052780511811023625,
+      "grad_norm": 0.9932296276092529,
+      "learning_rate": 0.0001986584862161256,
+      "loss": 3.4437,
+      "step": 858
+    },
+    {
+      "epoch": 0.05284202755905512,
+      "grad_norm": 1.294346570968628,
+      "learning_rate": 0.00019865532751708856,
+      "loss": 3.2218,
+      "step": 859
+    },
+    {
+      "epoch": 0.05290354330708662,
+      "grad_norm": 0.791111171245575,
+      "learning_rate": 0.00019865216512889237,
+      "loss": 2.5387,
+      "step": 860
+    },
+    {
+      "epoch": 0.05296505905511811,
+      "grad_norm": 1.0598771572113037,
+      "learning_rate": 0.00019864899905165531,
+      "loss": 3.2056,
+      "step": 861
+    },
+    {
+      "epoch": 0.05302657480314961,
+      "grad_norm": 1.0186821222305298,
+      "learning_rate": 0.00019864582928549575,
+      "loss": 3.1455,
+      "step": 862
+    },
+    {
+      "epoch": 0.053088090551181105,
+      "grad_norm": 0.7923797369003296,
+      "learning_rate": 0.00019864265583053223,
+      "loss": 3.624,
+      "step": 863
+    },
+    {
+      "epoch": 0.0531496062992126,
+      "grad_norm": 1.0614458322525024,
+      "learning_rate": 0.00019863947868688347,
+      "loss": 3.5062,
+      "step": 864
+    },
+    {
+      "epoch": 0.0532111220472441,
+      "grad_norm": 0.9026811718940735,
+      "learning_rate": 0.00019863629785466823,
+      "loss": 3.5426,
+      "step": 865
+    },
+    {
+      "epoch": 0.05327263779527559,
+      "grad_norm": 1.142218828201294,
+      "learning_rate": 0.0001986331133340054,
+      "loss": 3.5197,
+      "step": 866
+    },
+    {
+      "epoch": 0.05333415354330709,
+      "grad_norm": 0.9204780459403992,
+      "learning_rate": 0.0001986299251250142,
+      "loss": 3.3555,
+      "step": 867
+    },
+    {
+      "epoch": 0.053395669291338585,
+      "grad_norm": 1.0302813053131104,
+      "learning_rate": 0.00019862673322781373,
+      "loss": 3.1022,
+      "step": 868
+    },
+    {
+      "epoch": 0.05345718503937008,
+      "grad_norm": 0.9494417905807495,
+      "learning_rate": 0.00019862353764252342,
+      "loss": 2.8424,
+      "step": 869
+    },
+    {
+      "epoch": 0.05351870078740158,
+      "grad_norm": 1.065163254737854,
+      "learning_rate": 0.00019862033836926272,
+      "loss": 3.2597,
+      "step": 870
+    },
+    {
+      "epoch": 0.05358021653543307,
+      "grad_norm": 1.1206036806106567,
+      "learning_rate": 0.00019861713540815133,
+      "loss": 3.4829,
+      "step": 871
+    },
+    {
+      "epoch": 0.05364173228346457,
+      "grad_norm": 0.9656254053115845,
+      "learning_rate": 0.00019861392875930893,
+      "loss": 3.3788,
+      "step": 872
+    },
+    {
+      "epoch": 0.053703248031496065,
+      "grad_norm": 1.5548181533813477,
+      "learning_rate": 0.00019861071842285552,
+      "loss": 3.2119,
+      "step": 873
+    },
+    {
+      "epoch": 0.05376476377952756,
+      "grad_norm": 1.0688672065734863,
+      "learning_rate": 0.00019860750439891107,
+      "loss": 3.6767,
+      "step": 874
+    },
+    {
+      "epoch": 0.05382627952755906,
+      "grad_norm": 1.1465378999710083,
+      "learning_rate": 0.00019860428668759585,
+      "loss": 3.4529,
+      "step": 875
+    },
+    {
+      "epoch": 0.05388779527559055,
+      "grad_norm": 0.9086549282073975,
+      "learning_rate": 0.00019860106528903011,
+      "loss": 3.5119,
+      "step": 876
+    },
+    {
+      "epoch": 0.05394931102362205,
+      "grad_norm": 0.9865842461585999,
+      "learning_rate": 0.00019859784020333435,
+      "loss": 3.4105,
+      "step": 877
+    },
+    {
+      "epoch": 0.054010826771653545,
+      "grad_norm": 1.0100080966949463,
+      "learning_rate": 0.00019859461143062915,
+      "loss": 3.2203,
+      "step": 878
+    },
+    {
+      "epoch": 0.05407234251968504,
+      "grad_norm": 1.4716598987579346,
+      "learning_rate": 0.00019859137897103527,
+      "loss": 2.3451,
+      "step": 879
+    },
+    {
+      "epoch": 0.054133858267716536,
+      "grad_norm": 0.8393481373786926,
+      "learning_rate": 0.0001985881428246736,
+      "loss": 3.9337,
+      "step": 880
+    },
+    {
+      "epoch": 0.05419537401574803,
+      "grad_norm": 0.8785212635993958,
+      "learning_rate": 0.0001985849029916651,
+      "loss": 3.0711,
+      "step": 881
+    },
+    {
+      "epoch": 0.05425688976377953,
+      "grad_norm": 1.2293334007263184,
+      "learning_rate": 0.00019858165947213099,
+      "loss": 3.7111,
+      "step": 882
+    },
+    {
+      "epoch": 0.054318405511811024,
+      "grad_norm": 0.9070557355880737,
+      "learning_rate": 0.0001985784122661925,
+      "loss": 2.4599,
+      "step": 883
+    },
+    {
+      "epoch": 0.05437992125984252,
+      "grad_norm": 1.1132659912109375,
+      "learning_rate": 0.00019857516137397106,
+      "loss": 2.7006,
+      "step": 884
+    },
+    {
+      "epoch": 0.054441437007874016,
+      "grad_norm": 0.8325349688529968,
+      "learning_rate": 0.00019857190679558828,
+      "loss": 2.6242,
+      "step": 885
+    },
+    {
+      "epoch": 0.05450295275590551,
+      "grad_norm": 0.9353339076042175,
+      "learning_rate": 0.00019856864853116582,
+      "loss": 3.1282,
+      "step": 886
+    },
+    {
+      "epoch": 0.05456446850393701,
+      "grad_norm": 0.9308416247367859,
+      "learning_rate": 0.00019856538658082556,
+      "loss": 2.8726,
+      "step": 887
+    },
+    {
+      "epoch": 0.054625984251968504,
+      "grad_norm": 0.9620156288146973,
+      "learning_rate": 0.00019856212094468943,
+      "loss": 3.9249,
+      "step": 888
+    },
+    {
+      "epoch": 0.0546875,
+      "grad_norm": 0.8439517021179199,
+      "learning_rate": 0.0001985588516228796,
+      "loss": 3.2106,
+      "step": 889
+    },
+    {
+      "epoch": 0.054749015748031496,
+      "grad_norm": 1.04251229763031,
+      "learning_rate": 0.0001985555786155183,
+      "loss": 3.1149,
+      "step": 890
+    },
+    {
+      "epoch": 0.05481053149606299,
+      "grad_norm": 1.1975336074829102,
+      "learning_rate": 0.0001985523019227279,
+      "loss": 3.4565,
+      "step": 891
+    },
+    {
+      "epoch": 0.05487204724409449,
+      "grad_norm": 1.0040483474731445,
+      "learning_rate": 0.00019854902154463095,
+      "loss": 2.4647,
+      "step": 892
+    },
+    {
+      "epoch": 0.054933562992125984,
+      "grad_norm": 0.981143593788147,
+      "learning_rate": 0.00019854573748135014,
+      "loss": 3.2005,
+      "step": 893
+    },
+    {
+      "epoch": 0.05499507874015748,
+      "grad_norm": 1.070953369140625,
+      "learning_rate": 0.00019854244973300822,
+      "loss": 3.7618,
+      "step": 894
+    },
+    {
+      "epoch": 0.055056594488188976,
+      "grad_norm": 1.4045289754867554,
+      "learning_rate": 0.0001985391582997282,
+      "loss": 3.2438,
+      "step": 895
+    },
+    {
+      "epoch": 0.05511811023622047,
+      "grad_norm": 1.3213962316513062,
+      "learning_rate": 0.0001985358631816331,
+      "loss": 3.7315,
+      "step": 896
+    },
+    {
+      "epoch": 0.05517962598425197,
+      "grad_norm": 0.9182040691375732,
+      "learning_rate": 0.00019853256437884622,
+      "loss": 3.2869,
+      "step": 897
+    },
+    {
+      "epoch": 0.055241141732283464,
+      "grad_norm": 1.22479248046875,
+      "learning_rate": 0.0001985292618914908,
+      "loss": 2.8179,
+      "step": 898
+    },
+    {
+      "epoch": 0.05530265748031496,
+      "grad_norm": 1.116674542427063,
+      "learning_rate": 0.00019852595571969044,
+      "loss": 3.9339,
+      "step": 899
+    },
+    {
+      "epoch": 0.055364173228346455,
+      "grad_norm": 0.9990929961204529,
+      "learning_rate": 0.00019852264586356873,
+      "loss": 2.9938,
+      "step": 900
+    },
+    {
+      "epoch": 0.05542568897637795,
+      "grad_norm": 1.1124221086502075,
+      "learning_rate": 0.00019851933232324943,
+      "loss": 3.5173,
+      "step": 901
+    },
+    {
+      "epoch": 0.05548720472440945,
+      "grad_norm": 1.175569772720337,
+      "learning_rate": 0.0001985160150988565,
+      "loss": 2.5636,
+      "step": 902
+    },
+    {
+      "epoch": 0.05554872047244094,
+      "grad_norm": 1.083282470703125,
+      "learning_rate": 0.0001985126941905139,
+      "loss": 2.9933,
+      "step": 903
+    },
+    {
+      "epoch": 0.05561023622047244,
+      "grad_norm": 0.9268473386764526,
+      "learning_rate": 0.00019850936959834586,
+      "loss": 2.7795,
+      "step": 904
+    },
+    {
+      "epoch": 0.055671751968503935,
+      "grad_norm": 1.045573115348816,
+      "learning_rate": 0.00019850604132247673,
+      "loss": 3.4612,
+      "step": 905
+    },
+    {
+      "epoch": 0.05573326771653543,
+      "grad_norm": 1.0998960733413696,
+      "learning_rate": 0.0001985027093630309,
+      "loss": 3.4725,
+      "step": 906
+    },
+    {
+      "epoch": 0.05579478346456693,
+      "grad_norm": 0.9468164443969727,
+      "learning_rate": 0.00019849937372013303,
+      "loss": 3.7601,
+      "step": 907
+    },
+    {
+      "epoch": 0.05585629921259842,
+      "grad_norm": 0.6764987707138062,
+      "learning_rate": 0.00019849603439390782,
+      "loss": 3.081,
+      "step": 908
+    },
+    {
+      "epoch": 0.05591781496062992,
+      "grad_norm": 0.9524703025817871,
+      "learning_rate": 0.00019849269138448017,
+      "loss": 3.6626,
+      "step": 909
+    },
+    {
+      "epoch": 0.055979330708661415,
+      "grad_norm": 0.9894661903381348,
+      "learning_rate": 0.00019848934469197507,
+      "loss": 3.814,
+      "step": 910
+    },
+    {
+      "epoch": 0.05604084645669291,
+      "grad_norm": 1.1476695537567139,
+      "learning_rate": 0.00019848599431651764,
+      "loss": 3.0634,
+      "step": 911
+    },
+    {
+      "epoch": 0.05610236220472441,
+      "grad_norm": 1.4115115404129028,
+      "learning_rate": 0.00019848264025823325,
+      "loss": 2.6633,
+      "step": 912
+    },
+    {
+      "epoch": 0.0561638779527559,
+      "grad_norm": 0.9321492910385132,
+      "learning_rate": 0.00019847928251724722,
+      "loss": 3.2959,
+      "step": 913
+    },
+    {
+      "epoch": 0.0562253937007874,
+      "grad_norm": 1.4686555862426758,
+      "learning_rate": 0.00019847592109368517,
+      "loss": 3.0162,
+      "step": 914
+    },
+    {
+      "epoch": 0.056286909448818895,
+      "grad_norm": 0.9581645131111145,
+      "learning_rate": 0.00019847255598767278,
+      "loss": 3.392,
+      "step": 915
+    },
+    {
+      "epoch": 0.05634842519685039,
+      "grad_norm": 0.8637402653694153,
+      "learning_rate": 0.00019846918719933592,
+      "loss": 3.4464,
+      "step": 916
+    },
+    {
+      "epoch": 0.05640994094488189,
+      "grad_norm": 0.9568231701850891,
+      "learning_rate": 0.00019846581472880053,
+      "loss": 3.121,
+      "step": 917
+    },
+    {
+      "epoch": 0.05647145669291338,
+      "grad_norm": 1.0893813371658325,
+      "learning_rate": 0.00019846243857619272,
+      "loss": 2.4713,
+      "step": 918
+    },
+    {
+      "epoch": 0.05653297244094488,
+      "grad_norm": 1.1776000261306763,
+      "learning_rate": 0.00019845905874163876,
+      "loss": 3.1396,
+      "step": 919
+    },
+    {
+      "epoch": 0.056594488188976375,
+      "grad_norm": 1.2880522012710571,
+      "learning_rate": 0.000198455675225265,
+      "loss": 2.8627,
+      "step": 920
+    },
+    {
+      "epoch": 0.05665600393700788,
+      "grad_norm": 0.8591059446334839,
+      "learning_rate": 0.00019845228802719804,
+      "loss": 3.5525,
+      "step": 921
+    },
+    {
+      "epoch": 0.05671751968503937,
+      "grad_norm": 0.9692816138267517,
+      "learning_rate": 0.00019844889714756446,
+      "loss": 3.2476,
+      "step": 922
+    },
+    {
+      "epoch": 0.05677903543307087,
+      "grad_norm": 1.0331110954284668,
+      "learning_rate": 0.0001984455025864911,
+      "loss": 3.172,
+      "step": 923
+    },
+    {
+      "epoch": 0.056840551181102365,
+      "grad_norm": 1.006731390953064,
+      "learning_rate": 0.00019844210434410487,
+      "loss": 3.1024,
+      "step": 924
+    },
+    {
+      "epoch": 0.05690206692913386,
+      "grad_norm": 0.9818705916404724,
+      "learning_rate": 0.00019843870242053287,
+      "loss": 3.7138,
+      "step": 925
+    },
+    {
+      "epoch": 0.05696358267716536,
+      "grad_norm": 0.8925381302833557,
+      "learning_rate": 0.00019843529681590235,
+      "loss": 3.6507,
+      "step": 926
+    },
+    {
+      "epoch": 0.05702509842519685,
+      "grad_norm": 1.1379152536392212,
+      "learning_rate": 0.0001984318875303406,
+      "loss": 3.8501,
+      "step": 927
+    },
+    {
+      "epoch": 0.05708661417322835,
+      "grad_norm": 0.9514843225479126,
+      "learning_rate": 0.00019842847456397512,
+      "loss": 3.4861,
+      "step": 928
+    },
+    {
+      "epoch": 0.057148129921259845,
+      "grad_norm": 0.8861183524131775,
+      "learning_rate": 0.00019842505791693352,
+      "loss": 3.4651,
+      "step": 929
+    },
+    {
+      "epoch": 0.05720964566929134,
+      "grad_norm": 1.1168620586395264,
+      "learning_rate": 0.0001984216375893436,
+      "loss": 2.9993,
+      "step": 930
+    },
+    {
+      "epoch": 0.05727116141732284,
+      "grad_norm": 1.152671217918396,
+      "learning_rate": 0.00019841821358133327,
+      "loss": 3.676,
+      "step": 931
+    },
+    {
+      "epoch": 0.05733267716535433,
+      "grad_norm": 0.9899214506149292,
+      "learning_rate": 0.00019841478589303053,
+      "loss": 3.3338,
+      "step": 932
+    },
+    {
+      "epoch": 0.05739419291338583,
+      "grad_norm": 0.9296076893806458,
+      "learning_rate": 0.00019841135452456358,
+      "loss": 3.0115,
+      "step": 933
+    },
+    {
+      "epoch": 0.057455708661417325,
+      "grad_norm": 1.2188668251037598,
+      "learning_rate": 0.00019840791947606072,
+      "loss": 3.0072,
+      "step": 934
+    },
+    {
+      "epoch": 0.05751722440944882,
+      "grad_norm": 0.9961307644844055,
+      "learning_rate": 0.0001984044807476504,
+      "loss": 3.2954,
+      "step": 935
+    },
+    {
+      "epoch": 0.05757874015748032,
+      "grad_norm": 1.175564169883728,
+      "learning_rate": 0.00019840103833946123,
+      "loss": 3.4741,
+      "step": 936
+    },
+    {
+      "epoch": 0.05764025590551181,
+      "grad_norm": 0.9977335333824158,
+      "learning_rate": 0.00019839759225162196,
+      "loss": 3.2178,
+      "step": 937
+    },
+    {
+      "epoch": 0.05770177165354331,
+      "grad_norm": 0.6963675618171692,
+      "learning_rate": 0.00019839414248426138,
+      "loss": 3.4635,
+      "step": 938
+    },
+    {
+      "epoch": 0.057763287401574805,
+      "grad_norm": 1.2218871116638184,
+      "learning_rate": 0.00019839068903750853,
+      "loss": 2.8218,
+      "step": 939
+    },
+    {
+      "epoch": 0.0578248031496063,
+      "grad_norm": 0.9182488322257996,
+      "learning_rate": 0.00019838723191149254,
+      "loss": 3.6051,
+      "step": 940
+    },
+    {
+      "epoch": 0.0578863188976378,
+      "grad_norm": 0.899132251739502,
+      "learning_rate": 0.00019838377110634273,
+      "loss": 3.7403,
+      "step": 941
+    },
+    {
+      "epoch": 0.05794783464566929,
+      "grad_norm": 0.9607721567153931,
+      "learning_rate": 0.0001983803066221885,
+      "loss": 3.4934,
+      "step": 942
+    },
+    {
+      "epoch": 0.05800935039370079,
+      "grad_norm": 1.061888575553894,
+      "learning_rate": 0.00019837683845915933,
+      "loss": 3.1566,
+      "step": 943
+    },
+    {
+      "epoch": 0.058070866141732284,
+      "grad_norm": 1.0391288995742798,
+      "learning_rate": 0.000198373366617385,
+      "loss": 3.9206,
+      "step": 944
+    },
+    {
+      "epoch": 0.05813238188976378,
+      "grad_norm": 1.0898751020431519,
+      "learning_rate": 0.00019836989109699532,
+      "loss": 3.3508,
+      "step": 945
+    },
+    {
+      "epoch": 0.058193897637795276,
+      "grad_norm": 1.2978448867797852,
+      "learning_rate": 0.0001983664118981202,
+      "loss": 3.5898,
+      "step": 946
+    },
+    {
+      "epoch": 0.05825541338582677,
+      "grad_norm": 1.1777387857437134,
+      "learning_rate": 0.0001983629290208898,
+      "loss": 3.1728,
+      "step": 947
+    },
+    {
+      "epoch": 0.05831692913385827,
+      "grad_norm": 1.2441579103469849,
+      "learning_rate": 0.00019835944246543433,
+      "loss": 3.45,
+      "step": 948
+    },
+    {
+      "epoch": 0.058378444881889764,
+      "grad_norm": 1.194693922996521,
+      "learning_rate": 0.0001983559522318842,
+      "loss": 3.4311,
+      "step": 949
+    },
+    {
+      "epoch": 0.05843996062992126,
+      "grad_norm": 1.1722536087036133,
+      "learning_rate": 0.00019835245832036988,
+      "loss": 3.1792,
+      "step": 950
+    },
+    {
+      "epoch": 0.058501476377952756,
+      "grad_norm": 1.0798572301864624,
+      "learning_rate": 0.00019834896073102207,
+      "loss": 3.2054,
+      "step": 951
+    },
+    {
+      "epoch": 0.05856299212598425,
+      "grad_norm": 1.0314178466796875,
+      "learning_rate": 0.00019834545946397153,
+      "loss": 3.2794,
+      "step": 952
+    },
+    {
+      "epoch": 0.05862450787401575,
+      "grad_norm": 0.9989464282989502,
+      "learning_rate": 0.00019834195451934919,
+      "loss": 3.109,
+      "step": 953
+    },
+    {
+      "epoch": 0.058686023622047244,
+      "grad_norm": 0.8434916734695435,
+      "learning_rate": 0.0001983384458972861,
+      "loss": 3.4538,
+      "step": 954
+    },
+    {
+      "epoch": 0.05874753937007874,
+      "grad_norm": 0.898003101348877,
+      "learning_rate": 0.00019833493359791354,
+      "loss": 3.5138,
+      "step": 955
+    },
+    {
+      "epoch": 0.058809055118110236,
+      "grad_norm": 1.1175003051757812,
+      "learning_rate": 0.00019833141762136275,
+      "loss": 3.0778,
+      "step": 956
+    },
+    {
+      "epoch": 0.05887057086614173,
+      "grad_norm": 1.2328459024429321,
+      "learning_rate": 0.00019832789796776526,
+      "loss": 3.3396,
+      "step": 957
+    },
+    {
+      "epoch": 0.05893208661417323,
+      "grad_norm": 0.9740937948226929,
+      "learning_rate": 0.0001983243746372527,
+      "loss": 3.9751,
+      "step": 958
+    },
+    {
+      "epoch": 0.058993602362204724,
+      "grad_norm": 0.8782704472541809,
+      "learning_rate": 0.00019832084762995677,
+      "loss": 3.0197,
+      "step": 959
+    },
+    {
+      "epoch": 0.05905511811023622,
+      "grad_norm": 0.9233783483505249,
+      "learning_rate": 0.00019831731694600941,
+      "loss": 3.2042,
+      "step": 960
+    },
+    {
+      "epoch": 0.059116633858267716,
+      "grad_norm": 1.603043794631958,
+      "learning_rate": 0.0001983137825855426,
+      "loss": 2.9229,
+      "step": 961
+    },
+    {
+      "epoch": 0.05917814960629921,
+      "grad_norm": 1.0953506231307983,
+      "learning_rate": 0.00019831024454868858,
+      "loss": 3.3832,
+      "step": 962
+    },
+    {
+      "epoch": 0.05923966535433071,
+      "grad_norm": 0.9482942819595337,
+      "learning_rate": 0.00019830670283557957,
+      "loss": 3.9581,
+      "step": 963
+    },
+    {
+      "epoch": 0.0593011811023622,
+      "grad_norm": 1.1698737144470215,
+      "learning_rate": 0.00019830315744634808,
+      "loss": 3.3931,
+      "step": 964
+    },
+    {
+      "epoch": 0.0593626968503937,
+      "grad_norm": 0.8511584401130676,
+      "learning_rate": 0.00019829960838112662,
+      "loss": 3.3951,
+      "step": 965
+    },
+    {
+      "epoch": 0.059424212598425195,
+      "grad_norm": 1.2059438228607178,
+      "learning_rate": 0.00019829605564004795,
+      "loss": 2.8508,
+      "step": 966
+    },
+    {
+      "epoch": 0.05948572834645669,
+      "grad_norm": 4.7224531173706055,
+      "learning_rate": 0.00019829249922324494,
+      "loss": 2.5098,
+      "step": 967
+    },
+    {
+      "epoch": 0.05954724409448819,
+      "grad_norm": 1.6116784811019897,
+      "learning_rate": 0.0001982889391308505,
+      "loss": 3.1721,
+      "step": 968
+    },
+    {
+      "epoch": 0.05960875984251968,
+      "grad_norm": 2.660712242126465,
+      "learning_rate": 0.00019828537536299785,
+      "loss": 2.8704,
+      "step": 969
+    },
+    {
+      "epoch": 0.05967027559055118,
+      "grad_norm": 1.5035836696624756,
+      "learning_rate": 0.0001982818079198202,
+      "loss": 3.5831,
+      "step": 970
+    },
+    {
+      "epoch": 0.059731791338582675,
+      "grad_norm": 1.2460198402404785,
+      "learning_rate": 0.00019827823680145098,
+      "loss": 2.6681,
+      "step": 971
+    },
+    {
+      "epoch": 0.05979330708661417,
+      "grad_norm": 1.1423448324203491,
+      "learning_rate": 0.00019827466200802372,
+      "loss": 3.4077,
+      "step": 972
+    },
+    {
+      "epoch": 0.05985482283464567,
+      "grad_norm": 1.2081148624420166,
+      "learning_rate": 0.0001982710835396721,
+      "loss": 2.9125,
+      "step": 973
+    },
+    {
+      "epoch": 0.05991633858267716,
+      "grad_norm": 0.8486927151679993,
+      "learning_rate": 0.0001982675013965299,
+      "loss": 3.4384,
+      "step": 974
+    },
+    {
+      "epoch": 0.05997785433070866,
+      "grad_norm": 1.0742801427841187,
+      "learning_rate": 0.0001982639155787311,
+      "loss": 3.2577,
+      "step": 975
+    },
+    {
+      "epoch": 0.060039370078740155,
+      "grad_norm": 0.9942173361778259,
+      "learning_rate": 0.00019826032608640984,
+      "loss": 3.8237,
+      "step": 976
+    },
+    {
+      "epoch": 0.06010088582677165,
+      "grad_norm": 1.2189127206802368,
+      "learning_rate": 0.00019825673291970025,
+      "loss": 3.6131,
+      "step": 977
+    },
+    {
+      "epoch": 0.06016240157480315,
+      "grad_norm": 1.2017916440963745,
+      "learning_rate": 0.00019825313607873678,
+      "loss": 3.727,
+      "step": 978
+    },
+    {
+      "epoch": 0.06022391732283464,
+      "grad_norm": 0.9627209305763245,
+      "learning_rate": 0.00019824953556365387,
+      "loss": 3.2606,
+      "step": 979
+    },
+    {
+      "epoch": 0.06028543307086614,
+      "grad_norm": 1.012366533279419,
+      "learning_rate": 0.0001982459313745862,
+      "loss": 3.1793,
+      "step": 980
+    },
+    {
+      "epoch": 0.060346948818897635,
+      "grad_norm": 1.1381317377090454,
+      "learning_rate": 0.00019824232351166852,
+      "loss": 2.994,
+      "step": 981
+    },
+    {
+      "epoch": 0.06040846456692913,
+      "grad_norm": 1.0519928932189941,
+      "learning_rate": 0.00019823871197503575,
+      "loss": 3.0899,
+      "step": 982
+    },
+    {
+      "epoch": 0.06046998031496063,
+      "grad_norm": 1.1207796335220337,
+      "learning_rate": 0.00019823509676482296,
+      "loss": 2.4573,
+      "step": 983
+    },
+    {
+      "epoch": 0.06053149606299212,
+      "grad_norm": 0.8672242164611816,
+      "learning_rate": 0.00019823147788116535,
+      "loss": 3.0614,
+      "step": 984
+    },
+    {
+      "epoch": 0.060593011811023625,
+      "grad_norm": 1.3230596780776978,
+      "learning_rate": 0.00019822785532419817,
+      "loss": 3.3611,
+      "step": 985
+    },
+    {
+      "epoch": 0.06065452755905512,
+      "grad_norm": 0.9628525972366333,
+      "learning_rate": 0.00019822422909405698,
+      "loss": 3.7742,
+      "step": 986
+    },
+    {
+      "epoch": 0.06071604330708662,
+      "grad_norm": 0.9291071891784668,
+      "learning_rate": 0.0001982205991908773,
+      "loss": 3.5919,
+      "step": 987
+    },
+    {
+      "epoch": 0.06077755905511811,
+      "grad_norm": 1.0526254177093506,
+      "learning_rate": 0.00019821696561479494,
+      "loss": 2.785,
+      "step": 988
+    },
+    {
+      "epoch": 0.06083907480314961,
+      "grad_norm": 1.2061399221420288,
+      "learning_rate": 0.0001982133283659457,
+      "loss": 3.0,
+      "step": 989
+    },
+    {
+      "epoch": 0.060900590551181105,
+      "grad_norm": 0.991559624671936,
+      "learning_rate": 0.0001982096874444656,
+      "loss": 3.6813,
+      "step": 990
+    },
+    {
+      "epoch": 0.0609621062992126,
+      "grad_norm": 1.056264042854309,
+      "learning_rate": 0.0001982060428504909,
+      "loss": 3.5731,
+      "step": 991
+    },
+    {
+      "epoch": 0.0610236220472441,
+      "grad_norm": 1.1567199230194092,
+      "learning_rate": 0.0001982023945841578,
+      "loss": 3.0633,
+      "step": 992
+    },
+    {
+      "epoch": 0.06108513779527559,
+      "grad_norm": 0.9102746248245239,
+      "learning_rate": 0.00019819874264560266,
+      "loss": 3.1717,
+      "step": 993
+    },
+    {
+      "epoch": 0.06114665354330709,
+      "grad_norm": 1.2600808143615723,
+      "learning_rate": 0.00019819508703496219,
+      "loss": 3.0247,
+      "step": 994
+    },
+    {
+      "epoch": 0.061208169291338585,
+      "grad_norm": 1.0843584537506104,
+      "learning_rate": 0.000198191427752373,
+      "loss": 2.7995,
+      "step": 995
+    },
+    {
+      "epoch": 0.06126968503937008,
+      "grad_norm": 1.315144658088684,
+      "learning_rate": 0.00019818776479797192,
+      "loss": 3.0143,
+      "step": 996
+    },
+    {
+      "epoch": 0.06133120078740158,
+      "grad_norm": 1.1260204315185547,
+      "learning_rate": 0.00019818409817189597,
+      "loss": 2.7639,
+      "step": 997
+    },
+    {
+      "epoch": 0.06139271653543307,
+      "grad_norm": 1.2079100608825684,
+      "learning_rate": 0.00019818042787428223,
+      "loss": 3.4243,
+      "step": 998
+    },
+    {
+      "epoch": 0.06145423228346457,
+      "grad_norm": 0.9242305159568787,
+      "learning_rate": 0.00019817675390526796,
+      "loss": 3.9301,
+      "step": 999
+    },
+    {
+      "epoch": 0.061515748031496065,
+      "grad_norm": 1.0740801095962524,
+      "learning_rate": 0.00019817307626499053,
+      "loss": 2.6378,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06157726377952756,
+      "grad_norm": 0.798739492893219,
+      "learning_rate": 0.0001981693949535875,
+      "loss": 3.1644,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06163877952755906,
+      "grad_norm": 0.7899714708328247,
+      "learning_rate": 0.00019816570997119648,
+      "loss": 3.8991,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06170029527559055,
+      "grad_norm": 0.8012131452560425,
+      "learning_rate": 0.0001981620213179553,
+      "loss": 3.2173,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06176181102362205,
+      "grad_norm": 0.8830896615982056,
+      "learning_rate": 0.00019815832899400192,
+      "loss": 3.3369,
+      "step": 1004
+    },
+    {
+      "epoch": 0.061823326771653545,
+      "grad_norm": 0.8521789312362671,
+      "learning_rate": 0.00019815463299947434,
+      "loss": 3.2835,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06188484251968504,
+      "grad_norm": 1.3044441938400269,
+      "learning_rate": 0.00019815093333451082,
+      "loss": 3.1866,
+      "step": 1006
+    },
+    {
+      "epoch": 0.061946358267716536,
+      "grad_norm": 1.2284024953842163,
+      "learning_rate": 0.0001981472299992497,
+      "loss": 2.8092,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06200787401574803,
+      "grad_norm": 0.9837426543235779,
+      "learning_rate": 0.00019814352299382947,
+      "loss": 2.6048,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06206938976377953,
+      "grad_norm": 0.9988184571266174,
+      "learning_rate": 0.00019813981231838875,
+      "loss": 2.8404,
+      "step": 1009
+    },
+    {
+      "epoch": 0.062130905511811024,
+      "grad_norm": 0.9144999384880066,
+      "learning_rate": 0.00019813609797306628,
+      "loss": 3.8141,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06219242125984252,
+      "grad_norm": 0.9978460073471069,
+      "learning_rate": 0.00019813237995800097,
+      "loss": 3.1841,
+      "step": 1011
+    },
+    {
+      "epoch": 0.062253937007874016,
+      "grad_norm": 0.772580623626709,
+      "learning_rate": 0.00019812865827333184,
+      "loss": 3.2944,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06231545275590551,
+      "grad_norm": 1.079771876335144,
+      "learning_rate": 0.0001981249329191981,
+      "loss": 3.1956,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06237696850393701,
+      "grad_norm": 0.9598129987716675,
+      "learning_rate": 0.000198121203895739,
+      "loss": 3.1414,
+      "step": 1014
+    },
+    {
+      "epoch": 0.062438484251968504,
+      "grad_norm": 1.3985854387283325,
+      "learning_rate": 0.00019811747120309403,
+      "loss": 2.9467,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 1.0030540227890015,
+      "learning_rate": 0.00019811373484140274,
+      "loss": 3.0136,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0625615157480315,
+      "grad_norm": 1.0207799673080444,
+      "learning_rate": 0.00019810999481080488,
+      "loss": 3.6262,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06262303149606299,
+      "grad_norm": 1.0925010442733765,
+      "learning_rate": 0.00019810625111144027,
+      "loss": 3.4664,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06268454724409449,
+      "grad_norm": 1.9295374155044556,
+      "learning_rate": 0.00019810250374344894,
+      "loss": 3.0317,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06274606299212598,
+      "grad_norm": 1.2210547924041748,
+      "learning_rate": 0.000198098752706971,
+      "loss": 3.3672,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06280757874015748,
+      "grad_norm": 0.9135864973068237,
+      "learning_rate": 0.00019809499800214672,
+      "loss": 3.1184,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06286909448818898,
+      "grad_norm": 0.882534921169281,
+      "learning_rate": 0.00019809123962911653,
+      "loss": 3.4553,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06293061023622047,
+      "grad_norm": 1.4641104936599731,
+      "learning_rate": 0.00019808747758802092,
+      "loss": 3.6551,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06299212598425197,
+      "grad_norm": 0.9129678010940552,
+      "learning_rate": 0.00019808371187900062,
+      "loss": 4.0676,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06305364173228346,
+      "grad_norm": 1.361633539199829,
+      "learning_rate": 0.00019807994250219642,
+      "loss": 2.3021,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06311515748031496,
+      "grad_norm": 0.8523028492927551,
+      "learning_rate": 0.00019807616945774928,
+      "loss": 3.1719,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06317667322834646,
+      "grad_norm": 1.1481187343597412,
+      "learning_rate": 0.00019807239274580027,
+      "loss": 2.7347,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06323818897637795,
+      "grad_norm": 1.2273963689804077,
+      "learning_rate": 0.00019806861236649063,
+      "loss": 3.2407,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06329970472440945,
+      "grad_norm": 0.9298061728477478,
+      "learning_rate": 0.00019806482831996177,
+      "loss": 3.3864,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06336122047244094,
+      "grad_norm": 0.9620625972747803,
+      "learning_rate": 0.00019806104060635514,
+      "loss": 3.0182,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06342273622047244,
+      "grad_norm": 1.2778711318969727,
+      "learning_rate": 0.0001980572492258124,
+      "loss": 2.5435,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06348425196850394,
+      "grad_norm": 0.919080913066864,
+      "learning_rate": 0.00019805345417847533,
+      "loss": 3.0963,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06354576771653543,
+      "grad_norm": 1.149660587310791,
+      "learning_rate": 0.0001980496554644858,
+      "loss": 3.3355,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06360728346456693,
+      "grad_norm": 0.8131619691848755,
+      "learning_rate": 0.00019804585308398593,
+      "loss": 3.5464,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06366879921259842,
+      "grad_norm": 1.0452358722686768,
+      "learning_rate": 0.00019804204703711785,
+      "loss": 3.3464,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06373031496062992,
+      "grad_norm": 1.077369213104248,
+      "learning_rate": 0.00019803823732402392,
+      "loss": 3.5848,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06379183070866142,
+      "grad_norm": 1.1828263998031616,
+      "learning_rate": 0.00019803442394484658,
+      "loss": 3.0844,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06385334645669291,
+      "grad_norm": 1.0518134832382202,
+      "learning_rate": 0.00019803060689972844,
+      "loss": 3.6265,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0639148622047244,
+      "grad_norm": 0.8143224120140076,
+      "learning_rate": 0.00019802678618881226,
+      "loss": 3.2606,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0639763779527559,
+      "grad_norm": 1.0337432622909546,
+      "learning_rate": 0.00019802296181224086,
+      "loss": 2.9868,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0640378937007874,
+      "grad_norm": 1.346516489982605,
+      "learning_rate": 0.0001980191337701573,
+      "loss": 3.0273,
+      "step": 1041
+    },
+    {
+      "epoch": 0.0640994094488189,
+      "grad_norm": 1.728590488433838,
+      "learning_rate": 0.0001980153020627047,
+      "loss": 2.9352,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06416092519685039,
+      "grad_norm": 1.178573489189148,
+      "learning_rate": 0.00019801146669002636,
+      "loss": 3.5579,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06422244094488189,
+      "grad_norm": 0.904731810092926,
+      "learning_rate": 0.00019800762765226568,
+      "loss": 3.5084,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06428395669291338,
+      "grad_norm": 13.051922798156738,
+      "learning_rate": 0.0001980037849495662,
+      "loss": 3.3322,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06434547244094488,
+      "grad_norm": 1.6329126358032227,
+      "learning_rate": 0.0001979999385820717,
+      "loss": 3.2473,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06440698818897637,
+      "grad_norm": 0.994452714920044,
+      "learning_rate": 0.00019799608854992594,
+      "loss": 3.285,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06446850393700787,
+      "grad_norm": 0.9409752488136292,
+      "learning_rate": 0.00019799223485327292,
+      "loss": 2.995,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06453001968503937,
+      "grad_norm": 1.0003300905227661,
+      "learning_rate": 0.0001979883774922567,
+      "loss": 2.9417,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06459153543307086,
+      "grad_norm": 1.2203540802001953,
+      "learning_rate": 0.00019798451646702158,
+      "loss": 3.3724,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06465305118110236,
+      "grad_norm": 0.8677574396133423,
+      "learning_rate": 0.00019798065177771193,
+      "loss": 2.6556,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06471456692913385,
+      "grad_norm": 0.9182823300361633,
+      "learning_rate": 0.00019797678342447226,
+      "loss": 2.6354,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06477608267716535,
+      "grad_norm": 1.6130425930023193,
+      "learning_rate": 0.00019797291140744724,
+      "loss": 2.1997,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06483759842519685,
+      "grad_norm": 0.8843978643417358,
+      "learning_rate": 0.0001979690357267816,
+      "loss": 3.2877,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06489911417322834,
+      "grad_norm": 1.032376766204834,
+      "learning_rate": 0.00019796515638262037,
+      "loss": 3.2439,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06496062992125984,
+      "grad_norm": 1.0993318557739258,
+      "learning_rate": 0.00019796127337510852,
+      "loss": 3.6614,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06502214566929133,
+      "grad_norm": 1.2545503377914429,
+      "learning_rate": 0.0001979573867043913,
+      "loss": 2.8598,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06508366141732283,
+      "grad_norm": 0.9436438679695129,
+      "learning_rate": 0.00019795349637061405,
+      "loss": 2.5698,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06514517716535433,
+      "grad_norm": 1.1767387390136719,
+      "learning_rate": 0.00019794960237392223,
+      "loss": 3.918,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06520669291338582,
+      "grad_norm": 1.0511192083358765,
+      "learning_rate": 0.00019794570471446144,
+      "loss": 2.4237,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06526820866141732,
+      "grad_norm": 0.8265025019645691,
+      "learning_rate": 0.00019794180339237748,
+      "loss": 2.9187,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06532972440944881,
+      "grad_norm": 1.21897554397583,
+      "learning_rate": 0.0001979378984078162,
+      "loss": 2.0608,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06539124015748031,
+      "grad_norm": 0.9720959067344666,
+      "learning_rate": 0.00019793398976092368,
+      "loss": 3.026,
+      "step": 1063
+    },
+    {
+      "epoch": 0.0654527559055118,
+      "grad_norm": 1.0120106935501099,
+      "learning_rate": 0.000197930077451846,
+      "loss": 3.4939,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0655142716535433,
+      "grad_norm": 1.1378899812698364,
+      "learning_rate": 0.00019792616148072952,
+      "loss": 3.3913,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0655757874015748,
+      "grad_norm": 0.9509864449501038,
+      "learning_rate": 0.00019792224184772062,
+      "loss": 3.2626,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0656373031496063,
+      "grad_norm": 0.9258959293365479,
+      "learning_rate": 0.00019791831855296593,
+      "loss": 3.0919,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06569881889763779,
+      "grad_norm": 0.9419499635696411,
+      "learning_rate": 0.0001979143915966121,
+      "loss": 3.8297,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06576033464566929,
+      "grad_norm": 1.3132987022399902,
+      "learning_rate": 0.00019791046097880602,
+      "loss": 3.0868,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06582185039370078,
+      "grad_norm": 0.8415327072143555,
+      "learning_rate": 0.00019790652669969468,
+      "loss": 3.3283,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06588336614173228,
+      "grad_norm": 0.9655864834785461,
+      "learning_rate": 0.0001979025887594252,
+      "loss": 3.249,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06594488188976377,
+      "grad_norm": 1.4652599096298218,
+      "learning_rate": 0.0001978986471581448,
+      "loss": 3.4353,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06600639763779527,
+      "grad_norm": 2.1550068855285645,
+      "learning_rate": 0.00019789470189600087,
+      "loss": 3.3348,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06606791338582677,
+      "grad_norm": 0.8739051818847656,
+      "learning_rate": 0.000197890752973141,
+      "loss": 3.4334,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06612942913385826,
+      "grad_norm": 1.1301087141036987,
+      "learning_rate": 0.0001978868003897128,
+      "loss": 2.9242,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06619094488188976,
+      "grad_norm": 0.9550288915634155,
+      "learning_rate": 0.0001978828441458641,
+      "loss": 2.9047,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06625246062992125,
+      "grad_norm": 1.0125771760940552,
+      "learning_rate": 0.00019787888424174286,
+      "loss": 3.0966,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06631397637795275,
+      "grad_norm": 1.071850299835205,
+      "learning_rate": 0.00019787492067749711,
+      "loss": 3.4618,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06637549212598425,
+      "grad_norm": 1.0010225772857666,
+      "learning_rate": 0.0001978709534532751,
+      "loss": 3.4773,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06643700787401575,
+      "grad_norm": 0.9636658430099487,
+      "learning_rate": 0.0001978669825692252,
+      "loss": 3.1843,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06649852362204725,
+      "grad_norm": 1.1304068565368652,
+      "learning_rate": 0.00019786300802549585,
+      "loss": 4.0714,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06656003937007875,
+      "grad_norm": 1.07185959815979,
+      "learning_rate": 0.0001978590298222357,
+      "loss": 3.0443,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06662155511811024,
+      "grad_norm": 0.9814765453338623,
+      "learning_rate": 0.0001978550479595935,
+      "loss": 3.4776,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06668307086614174,
+      "grad_norm": 0.9072887897491455,
+      "learning_rate": 0.0001978510624377182,
+      "loss": 3.1141,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06674458661417323,
+      "grad_norm": 0.9005089402198792,
+      "learning_rate": 0.00019784707325675875,
+      "loss": 3.5219,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06680610236220473,
+      "grad_norm": 1.4247026443481445,
+      "learning_rate": 0.0001978430804168644,
+      "loss": 3.3085,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06686761811023623,
+      "grad_norm": 0.8715266585350037,
+      "learning_rate": 0.00019783908391818441,
+      "loss": 2.4987,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06692913385826772,
+      "grad_norm": 1.1338250637054443,
+      "learning_rate": 0.00019783508376086827,
+      "loss": 4.1039,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06699064960629922,
+      "grad_norm": 0.910824716091156,
+      "learning_rate": 0.00019783107994506552,
+      "loss": 3.5396,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06705216535433071,
+      "grad_norm": 1.1239176988601685,
+      "learning_rate": 0.0001978270724709259,
+      "loss": 3.0343,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06711368110236221,
+      "grad_norm": 0.9525589942932129,
+      "learning_rate": 0.00019782306133859929,
+      "loss": 3.2055,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0671751968503937,
+      "grad_norm": 1.3444684743881226,
+      "learning_rate": 0.00019781904654823565,
+      "loss": 3.4342,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0672367125984252,
+      "grad_norm": 1.0572463274002075,
+      "learning_rate": 0.00019781502809998512,
+      "loss": 2.7476,
+      "step": 1093
+    },
+    {
+      "epoch": 0.0672982283464567,
+      "grad_norm": 1.079888105392456,
+      "learning_rate": 0.000197811005993998,
+      "loss": 3.5172,
+      "step": 1094
+    },
+    {
+      "epoch": 0.0673597440944882,
+      "grad_norm": 1.080395221710205,
+      "learning_rate": 0.00019780698023042463,
+      "loss": 3.2525,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06742125984251969,
+      "grad_norm": 1.1758768558502197,
+      "learning_rate": 0.0001978029508094156,
+      "loss": 3.0852,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06748277559055119,
+      "grad_norm": 0.8595554828643799,
+      "learning_rate": 0.00019779891773112154,
+      "loss": 2.885,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06754429133858268,
+      "grad_norm": 1.134384274482727,
+      "learning_rate": 0.00019779488099569333,
+      "loss": 3.3432,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06760580708661418,
+      "grad_norm": 1.028795838356018,
+      "learning_rate": 0.00019779084060328193,
+      "loss": 3.0771,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06766732283464567,
+      "grad_norm": 1.1206470727920532,
+      "learning_rate": 0.0001977867965540383,
+      "loss": 3.5492,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06772883858267717,
+      "grad_norm": 0.969641923904419,
+      "learning_rate": 0.00019778274884811383,
+      "loss": 3.2237,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06779035433070867,
+      "grad_norm": 1.362514853477478,
+      "learning_rate": 0.0001977786974856598,
+      "loss": 3.1885,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06785187007874016,
+      "grad_norm": 0.9830102324485779,
+      "learning_rate": 0.00019777464246682763,
+      "loss": 3.854,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06791338582677166,
+      "grad_norm": 0.9564452767372131,
+      "learning_rate": 0.0001977705837917691,
+      "loss": 3.1298,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06797490157480315,
+      "grad_norm": 1.2482502460479736,
+      "learning_rate": 0.00019776652146063595,
+      "loss": 2.8028,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06803641732283465,
+      "grad_norm": 0.871826708316803,
+      "learning_rate": 0.00019776245547358,
+      "loss": 3.5334,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06809793307086615,
+      "grad_norm": 1.0426543951034546,
+      "learning_rate": 0.0001977583858307534,
+      "loss": 2.7153,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06815944881889764,
+      "grad_norm": 0.9698026776313782,
+      "learning_rate": 0.00019775431253230824,
+      "loss": 3.0922,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06822096456692914,
+      "grad_norm": 1.1073685884475708,
+      "learning_rate": 0.00019775023557839691,
+      "loss": 3.0282,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06828248031496063,
+      "grad_norm": 1.0168137550354004,
+      "learning_rate": 0.00019774615496917183,
+      "loss": 3.4088,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06834399606299213,
+      "grad_norm": 1.176485538482666,
+      "learning_rate": 0.00019774207070478563,
+      "loss": 3.2009,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06840551181102363,
+      "grad_norm": 0.9220025539398193,
+      "learning_rate": 0.00019773798278539099,
+      "loss": 2.9431,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06846702755905512,
+      "grad_norm": 1.1774674654006958,
+      "learning_rate": 0.0001977338912111408,
+      "loss": 3.1239,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06852854330708662,
+      "grad_norm": 1.6612805128097534,
+      "learning_rate": 0.00019772979598218804,
+      "loss": 3.8263,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06859005905511811,
+      "grad_norm": 1.0600665807724,
+      "learning_rate": 0.00019772569709868587,
+      "loss": 2.9515,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06865157480314961,
+      "grad_norm": 1.0937385559082031,
+      "learning_rate": 0.00019772159456078758,
+      "loss": 3.1924,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0687130905511811,
+      "grad_norm": 1.523350477218628,
+      "learning_rate": 0.00019771748836864655,
+      "loss": 2.924,
+      "step": 1117
+    },
+    {
+      "epoch": 0.0687746062992126,
+      "grad_norm": 1.162803292274475,
+      "learning_rate": 0.00019771337852241634,
+      "loss": 2.8772,
+      "step": 1118
+    },
+    {
+      "epoch": 0.0688361220472441,
+      "grad_norm": 1.0304698944091797,
+      "learning_rate": 0.00019770926502225063,
+      "loss": 3.5835,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0688976377952756,
+      "grad_norm": 1.1542965173721313,
+      "learning_rate": 0.0001977051478683033,
+      "loss": 2.9369,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06895915354330709,
+      "grad_norm": 1.0439494848251343,
+      "learning_rate": 0.00019770102706072822,
+      "loss": 3.2568,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06902066929133858,
+      "grad_norm": 1.36103355884552,
+      "learning_rate": 0.0001976969025996795,
+      "loss": 3.0059,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06908218503937008,
+      "grad_norm": 2.0454258918762207,
+      "learning_rate": 0.00019769277448531146,
+      "loss": 2.1963,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06914370078740158,
+      "grad_norm": 0.8615776896476746,
+      "learning_rate": 0.00019768864271777837,
+      "loss": 3.3005,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06920521653543307,
+      "grad_norm": 0.9526301622390747,
+      "learning_rate": 0.00019768450729723476,
+      "loss": 3.5507,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06926673228346457,
+      "grad_norm": 1.2104014158248901,
+      "learning_rate": 0.00019768036822383528,
+      "loss": 3.0812,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06932824803149606,
+      "grad_norm": 1.0440771579742432,
+      "learning_rate": 0.00019767622549773473,
+      "loss": 3.2073,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06938976377952756,
+      "grad_norm": 1.0222456455230713,
+      "learning_rate": 0.000197672079119088,
+      "loss": 2.5662,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06945127952755906,
+      "grad_norm": 0.9526201486587524,
+      "learning_rate": 0.00019766792908805012,
+      "loss": 3.0497,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06951279527559055,
+      "grad_norm": 0.9623120427131653,
+      "learning_rate": 0.00019766377540477632,
+      "loss": 3.4099,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06957431102362205,
+      "grad_norm": 1.3337408304214478,
+      "learning_rate": 0.0001976596180694219,
+      "loss": 2.9482,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06963582677165354,
+      "grad_norm": 0.8709504008293152,
+      "learning_rate": 0.00019765545708214232,
+      "loss": 3.6915,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06969734251968504,
+      "grad_norm": 0.9993990659713745,
+      "learning_rate": 0.00019765129244309323,
+      "loss": 3.2308,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06975885826771654,
+      "grad_norm": 1.3776171207427979,
+      "learning_rate": 0.0001976471241524303,
+      "loss": 2.7394,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06982037401574803,
+      "grad_norm": 0.8330480456352234,
+      "learning_rate": 0.00019764295221030942,
+      "loss": 3.3101,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06988188976377953,
+      "grad_norm": 1.4244885444641113,
+      "learning_rate": 0.00019763877661688656,
+      "loss": 3.1532,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06994340551181102,
+      "grad_norm": 1.044395089149475,
+      "learning_rate": 0.00019763459737231798,
+      "loss": 2.5765,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07000492125984252,
+      "grad_norm": 1.118833303451538,
+      "learning_rate": 0.00019763041447675982,
+      "loss": 3.1069,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07006643700787402,
+      "grad_norm": 0.8039726614952087,
+      "learning_rate": 0.00019762622793036857,
+      "loss": 3.6354,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07012795275590551,
+      "grad_norm": 4.032402038574219,
+      "learning_rate": 0.0001976220377333008,
+      "loss": 2.9925,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07018946850393701,
+      "grad_norm": 1.0094858407974243,
+      "learning_rate": 0.00019761784388571316,
+      "loss": 3.4238,
+      "step": 1141
+    },
+    {
+      "epoch": 0.0702509842519685,
+      "grad_norm": 1.1758644580841064,
+      "learning_rate": 0.0001976136463877625,
+      "loss": 2.186,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0703125,
+      "grad_norm": 1.02078115940094,
+      "learning_rate": 0.00019760944523960578,
+      "loss": 2.3868,
+      "step": 1143
+    },
+    {
+      "epoch": 0.0703740157480315,
+      "grad_norm": 1.0140984058380127,
+      "learning_rate": 0.00019760524044140008,
+      "loss": 2.9906,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07043553149606299,
+      "grad_norm": 0.9056522250175476,
+      "learning_rate": 0.00019760103199330264,
+      "loss": 3.4797,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07049704724409449,
+      "grad_norm": 0.8150445818901062,
+      "learning_rate": 0.00019759681989547085,
+      "loss": 3.3307,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07055856299212598,
+      "grad_norm": 1.070558786392212,
+      "learning_rate": 0.00019759260414806219,
+      "loss": 3.7987,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07062007874015748,
+      "grad_norm": 1.0654360055923462,
+      "learning_rate": 0.00019758838475123435,
+      "loss": 3.3437,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07068159448818898,
+      "grad_norm": 2.1453068256378174,
+      "learning_rate": 0.0001975841617051451,
+      "loss": 3.9364,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07074311023622047,
+      "grad_norm": 1.3946276903152466,
+      "learning_rate": 0.00019757993500995233,
+      "loss": 3.3681,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07080462598425197,
+      "grad_norm": 0.9625020027160645,
+      "learning_rate": 0.0001975757046658141,
+      "loss": 3.5384,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07086614173228346,
+      "grad_norm": 2.0756032466888428,
+      "learning_rate": 0.00019757147067288867,
+      "loss": 3.5641,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07092765748031496,
+      "grad_norm": 1.2127567529678345,
+      "learning_rate": 0.00019756723303133425,
+      "loss": 3.3553,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07098917322834646,
+      "grad_norm": 1.0816690921783447,
+      "learning_rate": 0.00019756299174130936,
+      "loss": 3.6792,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07105068897637795,
+      "grad_norm": 1.1645292043685913,
+      "learning_rate": 0.00019755874680297265,
+      "loss": 2.813,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07111220472440945,
+      "grad_norm": 1.1952012777328491,
+      "learning_rate": 0.00019755449821648279,
+      "loss": 2.6498,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07117372047244094,
+      "grad_norm": 1.1442910432815552,
+      "learning_rate": 0.00019755024598199866,
+      "loss": 3.7574,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07123523622047244,
+      "grad_norm": 1.287163496017456,
+      "learning_rate": 0.0001975459900996793,
+      "loss": 3.5046,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07129675196850394,
+      "grad_norm": 0.8789089918136597,
+      "learning_rate": 0.00019754173056968385,
+      "loss": 3.0898,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07135826771653543,
+      "grad_norm": 0.9765499830245972,
+      "learning_rate": 0.00019753746739217156,
+      "loss": 2.9197,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07141978346456693,
+      "grad_norm": 1.289778232574463,
+      "learning_rate": 0.0001975332005673019,
+      "loss": 2.9454,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07148129921259842,
+      "grad_norm": 0.9475486278533936,
+      "learning_rate": 0.00019752893009523437,
+      "loss": 3.3064,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07154281496062992,
+      "grad_norm": 1.169242024421692,
+      "learning_rate": 0.0001975246559761287,
+      "loss": 3.7468,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07160433070866142,
+      "grad_norm": 1.188733696937561,
+      "learning_rate": 0.00019752037821014473,
+      "loss": 2.7302,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07166584645669291,
+      "grad_norm": 1.4911333322525024,
+      "learning_rate": 0.00019751609679744241,
+      "loss": 3.4195,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0717273622047244,
+      "grad_norm": 1.190866470336914,
+      "learning_rate": 0.00019751181173818179,
+      "loss": 3.0739,
+      "step": 1166
+    },
+    {
+      "epoch": 0.0717888779527559,
+      "grad_norm": 0.8461530804634094,
+      "learning_rate": 0.00019750752303252317,
+      "loss": 3.0228,
+      "step": 1167
+    },
+    {
+      "epoch": 0.0718503937007874,
+      "grad_norm": 1.0839513540267944,
+      "learning_rate": 0.00019750323068062694,
+      "loss": 3.6351,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0719119094488189,
+      "grad_norm": 1.1197946071624756,
+      "learning_rate": 0.00019749893468265354,
+      "loss": 2.4923,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07197342519685039,
+      "grad_norm": 0.9816911220550537,
+      "learning_rate": 0.00019749463503876364,
+      "loss": 3.2467,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07203494094488189,
+      "grad_norm": 1.2147952318191528,
+      "learning_rate": 0.00019749033174911807,
+      "loss": 2.8686,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07209645669291338,
+      "grad_norm": 1.2108619213104248,
+      "learning_rate": 0.00019748602481387768,
+      "loss": 3.9729,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07215797244094488,
+      "grad_norm": 1.0494470596313477,
+      "learning_rate": 0.0001974817142332036,
+      "loss": 3.1591,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07221948818897637,
+      "grad_norm": 0.9098981618881226,
+      "learning_rate": 0.00019747740000725694,
+      "loss": 3.6201,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07228100393700787,
+      "grad_norm": 1.2088602781295776,
+      "learning_rate": 0.00019747308213619909,
+      "loss": 3.1739,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07234251968503937,
+      "grad_norm": 0.9592734575271606,
+      "learning_rate": 0.0001974687606201915,
+      "loss": 3.3692,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07240403543307086,
+      "grad_norm": 0.9151120781898499,
+      "learning_rate": 0.00019746443545939573,
+      "loss": 3.6323,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07246555118110236,
+      "grad_norm": 1.1089199781417847,
+      "learning_rate": 0.0001974601066539736,
+      "loss": 2.9169,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07252706692913385,
+      "grad_norm": 0.9782551527023315,
+      "learning_rate": 0.00019745577420408688,
+      "loss": 3.3159,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07258858267716535,
+      "grad_norm": 1.1399468183517456,
+      "learning_rate": 0.00019745143810989766,
+      "loss": 3.641,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07265009842519685,
+      "grad_norm": 0.9663099050521851,
+      "learning_rate": 0.00019744709837156806,
+      "loss": 3.1417,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07271161417322834,
+      "grad_norm": 1.1374493837356567,
+      "learning_rate": 0.00019744275498926036,
+      "loss": 3.4291,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07277312992125984,
+      "grad_norm": 0.8848555088043213,
+      "learning_rate": 0.00019743840796313696,
+      "loss": 3.5888,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07283464566929133,
+      "grad_norm": 1.1005027294158936,
+      "learning_rate": 0.00019743405729336045,
+      "loss": 3.4772,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07289616141732283,
+      "grad_norm": 0.9828979969024658,
+      "learning_rate": 0.0001974297029800935,
+      "loss": 3.248,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07295767716535433,
+      "grad_norm": 0.761803150177002,
+      "learning_rate": 0.000197425345023499,
+      "loss": 3.2247,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07301919291338582,
+      "grad_norm": 0.9564911127090454,
+      "learning_rate": 0.00019742098342373976,
+      "loss": 2.3259,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07308070866141732,
+      "grad_norm": 1.450649619102478,
+      "learning_rate": 0.00019741661818097902,
+      "loss": 3.0417,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07314222440944881,
+      "grad_norm": 1.08987295627594,
+      "learning_rate": 0.00019741224929537997,
+      "loss": 3.0634,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07320374015748031,
+      "grad_norm": 1.0567649602890015,
+      "learning_rate": 0.00019740787676710597,
+      "loss": 3.1551,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0732652559055118,
+      "grad_norm": 0.8759318590164185,
+      "learning_rate": 0.00019740350059632055,
+      "loss": 3.3998,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0733267716535433,
+      "grad_norm": 1.1017776727676392,
+      "learning_rate": 0.00019739912078318734,
+      "loss": 3.2515,
+      "step": 1192
+    },
+    {
+      "epoch": 0.0733882874015748,
+      "grad_norm": 1.0879738330841064,
+      "learning_rate": 0.00019739473732787013,
+      "loss": 2.7966,
+      "step": 1193
+    },
+    {
+      "epoch": 0.0734498031496063,
+      "grad_norm": 0.9958343505859375,
+      "learning_rate": 0.00019739035023053284,
+      "loss": 2.9168,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07351131889763779,
+      "grad_norm": 1.2158308029174805,
+      "learning_rate": 0.00019738595949133952,
+      "loss": 3.2332,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07357283464566929,
+      "grad_norm": 1.0124412775039673,
+      "learning_rate": 0.00019738156511045433,
+      "loss": 3.1578,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07363435039370078,
+      "grad_norm": 1.380028486251831,
+      "learning_rate": 0.0001973771670880416,
+      "loss": 3.3147,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07369586614173228,
+      "grad_norm": 0.9996014833450317,
+      "learning_rate": 0.00019737276542426585,
+      "loss": 3.2707,
+      "step": 1198
+    },
+    {
+      "epoch": 0.07375738188976377,
+      "grad_norm": 1.0125384330749512,
+      "learning_rate": 0.00019736836011929163,
+      "loss": 3.4038,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07381889763779527,
+      "grad_norm": 0.9751991629600525,
+      "learning_rate": 0.0001973639511732837,
+      "loss": 2.9513,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07388041338582677,
+      "grad_norm": 0.8861411213874817,
+      "learning_rate": 0.00019735953858640687,
+      "loss": 3.1389,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07394192913385826,
+      "grad_norm": 0.9469157457351685,
+      "learning_rate": 0.00019735512235882622,
+      "loss": 3.4496,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07400344488188976,
+      "grad_norm": 1.0882503986358643,
+      "learning_rate": 0.00019735070249070685,
+      "loss": 4.1906,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07406496062992125,
+      "grad_norm": 1.1564371585845947,
+      "learning_rate": 0.00019734627898221406,
+      "loss": 3.267,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07412647637795275,
+      "grad_norm": 1.0798039436340332,
+      "learning_rate": 0.00019734185183351323,
+      "loss": 3.0737,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07418799212598425,
+      "grad_norm": 1.061444878578186,
+      "learning_rate": 0.00019733742104476994,
+      "loss": 3.138,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07424950787401575,
+      "grad_norm": 1.0783135890960693,
+      "learning_rate": 0.0001973329866161499,
+      "loss": 3.0749,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07431102362204725,
+      "grad_norm": 1.1539024114608765,
+      "learning_rate": 0.00019732854854781888,
+      "loss": 2.5826,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07437253937007875,
+      "grad_norm": 0.9650754928588867,
+      "learning_rate": 0.00019732410683994287,
+      "loss": 3.8456,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07443405511811024,
+      "grad_norm": 1.120947003364563,
+      "learning_rate": 0.00019731966149268794,
+      "loss": 3.1112,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07449557086614174,
+      "grad_norm": 0.9343875646591187,
+      "learning_rate": 0.00019731521250622035,
+      "loss": 3.2879,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07455708661417323,
+      "grad_norm": 1.1807644367218018,
+      "learning_rate": 0.00019731075988070647,
+      "loss": 2.2689,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07461860236220473,
+      "grad_norm": 1.0370194911956787,
+      "learning_rate": 0.0001973063036163128,
+      "loss": 2.9196,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07468011811023623,
+      "grad_norm": 1.0175275802612305,
+      "learning_rate": 0.00019730184371320592,
+      "loss": 3.1546,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07474163385826772,
+      "grad_norm": 1.0456432104110718,
+      "learning_rate": 0.0001972973801715527,
+      "loss": 3.6504,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07480314960629922,
+      "grad_norm": 1.6512155532836914,
+      "learning_rate": 0.00019729291299151999,
+      "loss": 2.3841,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07486466535433071,
+      "grad_norm": 1.0278805494308472,
+      "learning_rate": 0.00019728844217327485,
+      "loss": 3.7794,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07492618110236221,
+      "grad_norm": 0.9069507122039795,
+      "learning_rate": 0.0001972839677169845,
+      "loss": 3.6032,
+      "step": 1218
+    },
+    {
+      "epoch": 0.0749876968503937,
+      "grad_norm": 0.9090246558189392,
+      "learning_rate": 0.0001972794896228162,
+      "loss": 3.8303,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0750492125984252,
+      "grad_norm": 1.1226094961166382,
+      "learning_rate": 0.00019727500789093743,
+      "loss": 3.1762,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0751107283464567,
+      "grad_norm": 1.1112005710601807,
+      "learning_rate": 0.00019727052252151582,
+      "loss": 3.0906,
+      "step": 1221
+    },
+    {
+      "epoch": 0.0751722440944882,
+      "grad_norm": 0.9997491836547852,
+      "learning_rate": 0.00019726603351471903,
+      "loss": 3.3517,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07523375984251969,
+      "grad_norm": 0.872468888759613,
+      "learning_rate": 0.00019726154087071492,
+      "loss": 2.8968,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07529527559055119,
+      "grad_norm": 1.2907177209854126,
+      "learning_rate": 0.00019725704458967158,
+      "loss": 2.8734,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07535679133858268,
+      "grad_norm": 0.8261265158653259,
+      "learning_rate": 0.00019725254467175708,
+      "loss": 3.4899,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07541830708661418,
+      "grad_norm": 1.0345170497894287,
+      "learning_rate": 0.00019724804111713971,
+      "loss": 3.0783,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07547982283464567,
+      "grad_norm": 1.3573065996170044,
+      "learning_rate": 0.0001972435339259879,
+      "loss": 3.0607,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07554133858267717,
+      "grad_norm": 0.8318873047828674,
+      "learning_rate": 0.00019723902309847012,
+      "loss": 3.4777,
+      "step": 1228
+    },
+    {
+      "epoch": 0.07560285433070867,
+      "grad_norm": 0.9925003051757812,
+      "learning_rate": 0.0001972345086347551,
+      "loss": 3.0145,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07566437007874016,
+      "grad_norm": 1.055325984954834,
+      "learning_rate": 0.00019722999053501164,
+      "loss": 2.9415,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07572588582677166,
+      "grad_norm": 1.1499285697937012,
+      "learning_rate": 0.00019722546879940875,
+      "loss": 3.0004,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07578740157480315,
+      "grad_norm": 0.9536684155464172,
+      "learning_rate": 0.00019722094342811543,
+      "loss": 3.0304,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07584891732283465,
+      "grad_norm": 0.9196504950523376,
+      "learning_rate": 0.00019721641442130096,
+      "loss": 3.1249,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07591043307086615,
+      "grad_norm": 1.6329963207244873,
+      "learning_rate": 0.0001972118817791347,
+      "loss": 3.2965,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07597194881889764,
+      "grad_norm": 1.507034182548523,
+      "learning_rate": 0.00019720734550178611,
+      "loss": 2.2394,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07603346456692914,
+      "grad_norm": 1.1256660223007202,
+      "learning_rate": 0.00019720280558942484,
+      "loss": 3.0675,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07609498031496063,
+      "grad_norm": 0.9212415814399719,
+      "learning_rate": 0.00019719826204222068,
+      "loss": 3.2328,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07615649606299213,
+      "grad_norm": 0.925105094909668,
+      "learning_rate": 0.0001971937148603435,
+      "loss": 3.4261,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07621801181102363,
+      "grad_norm": 0.9655049443244934,
+      "learning_rate": 0.00019718916404396332,
+      "loss": 3.1175,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07627952755905512,
+      "grad_norm": 1.3411082029342651,
+      "learning_rate": 0.00019718460959325038,
+      "loss": 3.098,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07634104330708662,
+      "grad_norm": 1.1496034860610962,
+      "learning_rate": 0.00019718005150837496,
+      "loss": 3.2257,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07640255905511811,
+      "grad_norm": 1.0347356796264648,
+      "learning_rate": 0.0001971754897895075,
+      "loss": 2.976,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07646407480314961,
+      "grad_norm": 0.8227728605270386,
+      "learning_rate": 0.00019717092443681858,
+      "loss": 3.1517,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0765255905511811,
+      "grad_norm": 0.9568734765052795,
+      "learning_rate": 0.00019716635545047892,
+      "loss": 3.1957,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0765871062992126,
+      "grad_norm": 0.8990029692649841,
+      "learning_rate": 0.00019716178283065938,
+      "loss": 3.7377,
+      "step": 1245
+    },
+    {
+      "epoch": 0.0766486220472441,
+      "grad_norm": 1.024296522140503,
+      "learning_rate": 0.00019715720657753096,
+      "loss": 3.004,
+      "step": 1246
+    },
+    {
+      "epoch": 0.0767101377952756,
+      "grad_norm": 1.13951575756073,
+      "learning_rate": 0.0001971526266912648,
+      "loss": 2.6067,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07677165354330709,
+      "grad_norm": 0.9492692351341248,
+      "learning_rate": 0.0001971480431720321,
+      "loss": 3.618,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07683316929133858,
+      "grad_norm": 1.3595378398895264,
+      "learning_rate": 0.00019714345602000433,
+      "loss": 2.7691,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07689468503937008,
+      "grad_norm": 1.0271995067596436,
+      "learning_rate": 0.00019713886523535298,
+      "loss": 2.6366,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07695620078740158,
+      "grad_norm": 0.9336234331130981,
+      "learning_rate": 0.0001971342708182497,
+      "loss": 3.2161,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07701771653543307,
+      "grad_norm": 1.0632413625717163,
+      "learning_rate": 0.00019712967276886638,
+      "loss": 3.0576,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07707923228346457,
+      "grad_norm": 0.8085705637931824,
+      "learning_rate": 0.00019712507108737487,
+      "loss": 3.5667,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07714074803149606,
+      "grad_norm": 1.0306705236434937,
+      "learning_rate": 0.00019712046577394728,
+      "loss": 2.6184,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07720226377952756,
+      "grad_norm": 1.1383439302444458,
+      "learning_rate": 0.00019711585682875585,
+      "loss": 3.2678,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07726377952755906,
+      "grad_norm": 1.139816403388977,
+      "learning_rate": 0.00019711124425197293,
+      "loss": 3.4245,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07732529527559055,
+      "grad_norm": 0.8280664682388306,
+      "learning_rate": 0.00019710662804377097,
+      "loss": 3.3584,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07738681102362205,
+      "grad_norm": 1.0665096044540405,
+      "learning_rate": 0.00019710200820432255,
+      "loss": 3.1119,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07744832677165354,
+      "grad_norm": 1.0723211765289307,
+      "learning_rate": 0.00019709738473380054,
+      "loss": 3.334,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07750984251968504,
+      "grad_norm": 0.8967238068580627,
+      "learning_rate": 0.00019709275763237774,
+      "loss": 3.1555,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07757135826771654,
+      "grad_norm": 1.0865565538406372,
+      "learning_rate": 0.00019708812690022722,
+      "loss": 3.4624,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07763287401574803,
+      "grad_norm": 0.9117671847343445,
+      "learning_rate": 0.00019708349253752213,
+      "loss": 2.9219,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07769438976377953,
+      "grad_norm": 1.030794620513916,
+      "learning_rate": 0.00019707885454443578,
+      "loss": 3.0161,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07775590551181102,
+      "grad_norm": 0.8990916609764099,
+      "learning_rate": 0.0001970742129211416,
+      "loss": 3.0266,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07781742125984252,
+      "grad_norm": 1.3270927667617798,
+      "learning_rate": 0.00019706956766781314,
+      "loss": 3.4503,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07787893700787402,
+      "grad_norm": 1.04781174659729,
+      "learning_rate": 0.00019706491878462412,
+      "loss": 3.3996,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07794045275590551,
+      "grad_norm": 0.9803470969200134,
+      "learning_rate": 0.0001970602662717484,
+      "loss": 3.5688,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07800196850393701,
+      "grad_norm": 1.1457735300064087,
+      "learning_rate": 0.00019705561012935995,
+      "loss": 2.8462,
+      "step": 1268
+    },
+    {
+      "epoch": 0.0780634842519685,
+      "grad_norm": 1.1132951974868774,
+      "learning_rate": 0.00019705095035763286,
+      "loss": 2.6997,
+      "step": 1269
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 1.0262717008590698,
+      "learning_rate": 0.00019704628695674142,
+      "loss": 2.6823,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0781865157480315,
+      "grad_norm": 1.093435525894165,
+      "learning_rate": 0.00019704161992685997,
+      "loss": 3.2871,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07824803149606299,
+      "grad_norm": 0.9329583048820496,
+      "learning_rate": 0.00019703694926816306,
+      "loss": 3.075,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07830954724409449,
+      "grad_norm": 0.909116804599762,
+      "learning_rate": 0.00019703227498082533,
+      "loss": 3.2848,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07837106299212598,
+      "grad_norm": 0.8814778327941895,
+      "learning_rate": 0.0001970275970650216,
+      "loss": 3.1154,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07843257874015748,
+      "grad_norm": 0.9098310470581055,
+      "learning_rate": 0.00019702291552092674,
+      "loss": 3.3295,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07849409448818898,
+      "grad_norm": 1.0212005376815796,
+      "learning_rate": 0.0001970182303487159,
+      "loss": 2.9134,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07855561023622047,
+      "grad_norm": 1.4569711685180664,
+      "learning_rate": 0.00019701354154856424,
+      "loss": 2.5074,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07861712598425197,
+      "grad_norm": 1.0095500946044922,
+      "learning_rate": 0.00019700884912064707,
+      "loss": 3.0746,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07867864173228346,
+      "grad_norm": 0.9042649269104004,
+      "learning_rate": 0.00019700415306513988,
+      "loss": 2.9573,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07874015748031496,
+      "grad_norm": 1.1612781286239624,
+      "learning_rate": 0.00019699945338221827,
+      "loss": 2.9529,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07880167322834646,
+      "grad_norm": 1.0675075054168701,
+      "learning_rate": 0.00019699475007205802,
+      "loss": 3.2982,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07886318897637795,
+      "grad_norm": 1.2616989612579346,
+      "learning_rate": 0.00019699004313483492,
+      "loss": 3.745,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07892470472440945,
+      "grad_norm": 0.7779290676116943,
+      "learning_rate": 0.00019698533257072508,
+      "loss": 3.2685,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07898622047244094,
+      "grad_norm": 0.9927670359611511,
+      "learning_rate": 0.0001969806183799046,
+      "loss": 3.6978,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07904773622047244,
+      "grad_norm": 0.9485732913017273,
+      "learning_rate": 0.00019697590056254975,
+      "loss": 2.8425,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07910925196850394,
+      "grad_norm": 0.8856753706932068,
+      "learning_rate": 0.00019697117911883697,
+      "loss": 3.5123,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07917076771653543,
+      "grad_norm": 0.9311007857322693,
+      "learning_rate": 0.00019696645404894284,
+      "loss": 3.0121,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07923228346456693,
+      "grad_norm": 1.7257113456726074,
+      "learning_rate": 0.000196961725353044,
+      "loss": 4.8605,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07929379921259842,
+      "grad_norm": 1.2023181915283203,
+      "learning_rate": 0.00019695699303131736,
+      "loss": 3.2529,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07935531496062992,
+      "grad_norm": 1.1551321744918823,
+      "learning_rate": 0.0001969522570839398,
+      "loss": 4.0385,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07941683070866142,
+      "grad_norm": 1.096747875213623,
+      "learning_rate": 0.00019694751751108842,
+      "loss": 3.0005,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07947834645669291,
+      "grad_norm": 1.0798461437225342,
+      "learning_rate": 0.0001969427743129405,
+      "loss": 3.3204,
+      "step": 1292
+    },
+    {
+      "epoch": 0.0795398622047244,
+      "grad_norm": 0.7967138290405273,
+      "learning_rate": 0.0001969380274896734,
+      "loss": 3.6562,
+      "step": 1293
+    },
+    {
+      "epoch": 0.0796013779527559,
+      "grad_norm": 0.9088689684867859,
+      "learning_rate": 0.0001969332770414646,
+      "loss": 3.8532,
+      "step": 1294
+    },
+    {
+      "epoch": 0.0796628937007874,
+      "grad_norm": 1.1157240867614746,
+      "learning_rate": 0.00019692852296849175,
+      "loss": 2.8592,
+      "step": 1295
+    },
+    {
+      "epoch": 0.0797244094488189,
+      "grad_norm": 1.2917592525482178,
+      "learning_rate": 0.00019692376527093264,
+      "loss": 3.5996,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07978592519685039,
+      "grad_norm": 1.2678303718566895,
+      "learning_rate": 0.00019691900394896514,
+      "loss": 2.9334,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07984744094488189,
+      "grad_norm": 1.2033008337020874,
+      "learning_rate": 0.00019691423900276736,
+      "loss": 3.4641,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07990895669291338,
+      "grad_norm": 0.8370565176010132,
+      "learning_rate": 0.00019690947043251745,
+      "loss": 3.3616,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07997047244094488,
+      "grad_norm": 1.1619607210159302,
+      "learning_rate": 0.0001969046982383937,
+      "loss": 3.3225,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08003198818897637,
+      "grad_norm": 1.096163272857666,
+      "learning_rate": 0.00019689992242057463,
+      "loss": 3.3486,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08009350393700787,
+      "grad_norm": 0.949455201625824,
+      "learning_rate": 0.00019689514297923876,
+      "loss": 2.8513,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08015501968503937,
+      "grad_norm": 1.120547890663147,
+      "learning_rate": 0.00019689035991456486,
+      "loss": 3.0374,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08021653543307086,
+      "grad_norm": 0.9276220798492432,
+      "learning_rate": 0.00019688557322673176,
+      "loss": 3.3295,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08027805118110236,
+      "grad_norm": 0.9893090128898621,
+      "learning_rate": 0.00019688078291591848,
+      "loss": 2.6664,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08033956692913385,
+      "grad_norm": 0.9218969345092773,
+      "learning_rate": 0.00019687598898230415,
+      "loss": 3.208,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08040108267716535,
+      "grad_norm": 0.8743588328361511,
+      "learning_rate": 0.00019687119142606802,
+      "loss": 3.5618,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08046259842519685,
+      "grad_norm": 0.8350561857223511,
+      "learning_rate": 0.00019686639024738952,
+      "loss": 3.1773,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08052411417322834,
+      "grad_norm": 0.9574966430664062,
+      "learning_rate": 0.00019686158544644813,
+      "loss": 3.8303,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08058562992125984,
+      "grad_norm": 1.0151900053024292,
+      "learning_rate": 0.00019685677702342358,
+      "loss": 3.392,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08064714566929133,
+      "grad_norm": 1.0215753316879272,
+      "learning_rate": 0.00019685196497849563,
+      "loss": 3.2794,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08070866141732283,
+      "grad_norm": 0.9206169843673706,
+      "learning_rate": 0.00019684714931184427,
+      "loss": 3.5907,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08077017716535433,
+      "grad_norm": 1.0579005479812622,
+      "learning_rate": 0.00019684233002364957,
+      "loss": 3.0488,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08083169291338582,
+      "grad_norm": 0.8552363514900208,
+      "learning_rate": 0.0001968375071140917,
+      "loss": 3.5624,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08089320866141732,
+      "grad_norm": 1.1183719635009766,
+      "learning_rate": 0.0001968326805833511,
+      "loss": 2.3487,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08095472440944881,
+      "grad_norm": 1.2261903285980225,
+      "learning_rate": 0.00019682785043160817,
+      "loss": 2.8473,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08101624015748031,
+      "grad_norm": 1.2035305500030518,
+      "learning_rate": 0.00019682301665904358,
+      "loss": 3.4083,
+      "step": 1317
+    },
+    {
+      "epoch": 0.0810777559055118,
+      "grad_norm": 1.3929685354232788,
+      "learning_rate": 0.00019681817926583804,
+      "loss": 2.4405,
+      "step": 1318
+    },
+    {
+      "epoch": 0.0811392716535433,
+      "grad_norm": 1.0156587362289429,
+      "learning_rate": 0.00019681333825217247,
+      "loss": 2.847,
+      "step": 1319
+    },
+    {
+      "epoch": 0.0812007874015748,
+      "grad_norm": 1.1210336685180664,
+      "learning_rate": 0.00019680849361822792,
+      "loss": 2.6486,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0812623031496063,
+      "grad_norm": 1.0509225130081177,
+      "learning_rate": 0.0001968036453641855,
+      "loss": 3.3208,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08132381889763779,
+      "grad_norm": 1.0451550483703613,
+      "learning_rate": 0.00019679879349022655,
+      "loss": 3.2675,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08138533464566929,
+      "grad_norm": 1.7379437685012817,
+      "learning_rate": 0.00019679393799653248,
+      "loss": 2.6172,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08144685039370078,
+      "grad_norm": 1.3596729040145874,
+      "learning_rate": 0.00019678907888328486,
+      "loss": 2.7276,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08150836614173228,
+      "grad_norm": 1.1472870111465454,
+      "learning_rate": 0.00019678421615066543,
+      "loss": 3.1638,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08156988188976377,
+      "grad_norm": 1.8374762535095215,
+      "learning_rate": 0.00019677934979885596,
+      "loss": 2.7668,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08163139763779527,
+      "grad_norm": 1.2466073036193848,
+      "learning_rate": 0.0001967744798280385,
+      "loss": 3.6797,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08169291338582677,
+      "grad_norm": 1.8974699974060059,
+      "learning_rate": 0.0001967696062383951,
+      "loss": 2.7537,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08175442913385826,
+      "grad_norm": 1.0753202438354492,
+      "learning_rate": 0.00019676472903010805,
+      "loss": 2.6295,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08181594488188976,
+      "grad_norm": 1.1112030744552612,
+      "learning_rate": 0.0001967598482033597,
+      "loss": 2.8862,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08187746062992125,
+      "grad_norm": 0.923079788684845,
+      "learning_rate": 0.00019675496375833257,
+      "loss": 3.6348,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08193897637795275,
+      "grad_norm": 1.09978449344635,
+      "learning_rate": 0.0001967500756952093,
+      "loss": 3.2186,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08200049212598425,
+      "grad_norm": 1.2486450672149658,
+      "learning_rate": 0.00019674518401417273,
+      "loss": 3.1095,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08206200787401575,
+      "grad_norm": 1.103110432624817,
+      "learning_rate": 0.00019674028871540574,
+      "loss": 3.6064,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08212352362204725,
+      "grad_norm": 1.7665225267410278,
+      "learning_rate": 0.00019673538979909135,
+      "loss": 2.6223,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08218503937007875,
+      "grad_norm": 1.5211071968078613,
+      "learning_rate": 0.0001967304872654128,
+      "loss": 3.2859,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08224655511811024,
+      "grad_norm": 0.9834161996841431,
+      "learning_rate": 0.00019672558111455345,
+      "loss": 3.2443,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08230807086614174,
+      "grad_norm": 1.0985209941864014,
+      "learning_rate": 0.0001967206713466967,
+      "loss": 3.0769,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08236958661417323,
+      "grad_norm": 0.8664349317550659,
+      "learning_rate": 0.00019671575796202618,
+      "loss": 3.1721,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08243110236220473,
+      "grad_norm": 1.3253631591796875,
+      "learning_rate": 0.0001967108409607256,
+      "loss": 2.8215,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08249261811023623,
+      "grad_norm": 1.0592551231384277,
+      "learning_rate": 0.00019670592034297884,
+      "loss": 3.4444,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08255413385826772,
+      "grad_norm": 0.7541211247444153,
+      "learning_rate": 0.0001967009961089699,
+      "loss": 3.2264,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08261564960629922,
+      "grad_norm": 1.1770085096359253,
+      "learning_rate": 0.00019669606825888295,
+      "loss": 3.6192,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08267716535433071,
+      "grad_norm": 1.4309608936309814,
+      "learning_rate": 0.00019669113679290223,
+      "loss": 2.2675,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08273868110236221,
+      "grad_norm": 0.9962587952613831,
+      "learning_rate": 0.00019668620171121214,
+      "loss": 3.1349,
+      "step": 1345
+    },
+    {
+      "epoch": 0.0828001968503937,
+      "grad_norm": 1.1290347576141357,
+      "learning_rate": 0.00019668126301399728,
+      "loss": 2.0975,
+      "step": 1346
+    },
+    {
+      "epoch": 0.0828617125984252,
+      "grad_norm": 1.149431824684143,
+      "learning_rate": 0.00019667632070144228,
+      "loss": 3.0944,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0829232283464567,
+      "grad_norm": 1.0089634656906128,
+      "learning_rate": 0.00019667137477373197,
+      "loss": 2.4707,
+      "step": 1348
+    },
+    {
+      "epoch": 0.0829847440944882,
+      "grad_norm": 1.1351467370986938,
+      "learning_rate": 0.0001966664252310513,
+      "loss": 2.6507,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08304625984251969,
+      "grad_norm": 0.9589190483093262,
+      "learning_rate": 0.00019666147207358532,
+      "loss": 3.2852,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08310777559055119,
+      "grad_norm": 1.0837535858154297,
+      "learning_rate": 0.0001966565153015193,
+      "loss": 3.7485,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08316929133858268,
+      "grad_norm": 0.8898327350616455,
+      "learning_rate": 0.0001966515549150386,
+      "loss": 3.397,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08323080708661418,
+      "grad_norm": 1.1231168508529663,
+      "learning_rate": 0.00019664659091432865,
+      "loss": 3.4022,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08329232283464567,
+      "grad_norm": 0.9638388156890869,
+      "learning_rate": 0.00019664162329957515,
+      "loss": 3.4847,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08335383858267717,
+      "grad_norm": 2.1260786056518555,
+      "learning_rate": 0.0001966366520709638,
+      "loss": 2.0011,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08341535433070867,
+      "grad_norm": 1.090923547744751,
+      "learning_rate": 0.00019663167722868056,
+      "loss": 2.4336,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08347687007874016,
+      "grad_norm": 1.0605480670928955,
+      "learning_rate": 0.0001966266987729114,
+      "loss": 3.3111,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08353838582677166,
+      "grad_norm": 1.1333407163619995,
+      "learning_rate": 0.00019662171670384253,
+      "loss": 2.2937,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08359990157480315,
+      "grad_norm": 1.0106078386306763,
+      "learning_rate": 0.00019661673102166021,
+      "loss": 3.2915,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08366141732283465,
+      "grad_norm": 1.0970988273620605,
+      "learning_rate": 0.0001966117417265509,
+      "loss": 3.1645,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08372293307086615,
+      "grad_norm": 0.9035822153091431,
+      "learning_rate": 0.0001966067488187012,
+      "loss": 3.3083,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08378444881889764,
+      "grad_norm": 1.011337161064148,
+      "learning_rate": 0.00019660175229829774,
+      "loss": 2.9783,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08384596456692914,
+      "grad_norm": 0.8051189184188843,
+      "learning_rate": 0.00019659675216552745,
+      "loss": 3.6098,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08390748031496063,
+      "grad_norm": 1.0578702688217163,
+      "learning_rate": 0.00019659174842057725,
+      "loss": 3.1471,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08396899606299213,
+      "grad_norm": 1.2057911157608032,
+      "learning_rate": 0.00019658674106363426,
+      "loss": 2.8978,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08403051181102363,
+      "grad_norm": 1.082565188407898,
+      "learning_rate": 0.0001965817300948858,
+      "loss": 2.9704,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08409202755905512,
+      "grad_norm": 0.8882172107696533,
+      "learning_rate": 0.00019657671551451912,
+      "loss": 2.5763,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08415354330708662,
+      "grad_norm": 1.1878503561019897,
+      "learning_rate": 0.0001965716973227218,
+      "loss": 3.0944,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08421505905511811,
+      "grad_norm": 0.9717600345611572,
+      "learning_rate": 0.00019656667551968155,
+      "loss": 2.6846,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08427657480314961,
+      "grad_norm": 0.8617116212844849,
+      "learning_rate": 0.00019656165010558608,
+      "loss": 3.338,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0843380905511811,
+      "grad_norm": 0.971933901309967,
+      "learning_rate": 0.00019655662108062334,
+      "loss": 3.0648,
+      "step": 1371
+    },
+    {
+      "epoch": 0.0843996062992126,
+      "grad_norm": 0.9987403750419617,
+      "learning_rate": 0.00019655158844498138,
+      "loss": 3.0314,
+      "step": 1372
+    },
+    {
+      "epoch": 0.0844611220472441,
+      "grad_norm": 1.3490629196166992,
+      "learning_rate": 0.0001965465521988484,
+      "loss": 2.6057,
+      "step": 1373
+    },
+    {
+      "epoch": 0.0845226377952756,
+      "grad_norm": 0.9411892294883728,
+      "learning_rate": 0.00019654151234241275,
+      "loss": 3.2607,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08458415354330709,
+      "grad_norm": 1.0720961093902588,
+      "learning_rate": 0.00019653646887586284,
+      "loss": 2.7976,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08464566929133858,
+      "grad_norm": 0.8423062562942505,
+      "learning_rate": 0.0001965314217993873,
+      "loss": 3.3774,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08470718503937008,
+      "grad_norm": 1.1439422369003296,
+      "learning_rate": 0.00019652637111317487,
+      "loss": 3.1959,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08476870078740158,
+      "grad_norm": 1.0007914304733276,
+      "learning_rate": 0.00019652131681741438,
+      "loss": 3.4142,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08483021653543307,
+      "grad_norm": 0.9195958971977234,
+      "learning_rate": 0.0001965162589122949,
+      "loss": 3.4117,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08489173228346457,
+      "grad_norm": 1.494600772857666,
+      "learning_rate": 0.0001965111973980055,
+      "loss": 2.664,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08495324803149606,
+      "grad_norm": 1.2629050016403198,
+      "learning_rate": 0.00019650613227473548,
+      "loss": 3.3006,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08501476377952756,
+      "grad_norm": 1.1010055541992188,
+      "learning_rate": 0.00019650106354267426,
+      "loss": 3.5817,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08507627952755906,
+      "grad_norm": 0.8953258395195007,
+      "learning_rate": 0.00019649599120201137,
+      "loss": 3.7872,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08513779527559055,
+      "grad_norm": 1.8579021692276,
+      "learning_rate": 0.00019649091525293647,
+      "loss": 1.5985,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08519931102362205,
+      "grad_norm": 1.2099645137786865,
+      "learning_rate": 0.00019648583569563939,
+      "loss": 3.22,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08526082677165354,
+      "grad_norm": 0.9755405187606812,
+      "learning_rate": 0.00019648075253031006,
+      "loss": 3.2617,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08532234251968504,
+      "grad_norm": 0.9205663204193115,
+      "learning_rate": 0.0001964756657571386,
+      "loss": 2.9102,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08538385826771654,
+      "grad_norm": 0.9286817312240601,
+      "learning_rate": 0.0001964705753763152,
+      "loss": 3.3893,
+      "step": 1388
+    },
+    {
+      "epoch": 0.08544537401574803,
+      "grad_norm": 0.9016866087913513,
+      "learning_rate": 0.0001964654813880302,
+      "loss": 3.0866,
+      "step": 1389
+    },
+    {
+      "epoch": 0.08550688976377953,
+      "grad_norm": 0.8915243744850159,
+      "learning_rate": 0.00019646038379247412,
+      "loss": 3.3216,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08556840551181102,
+      "grad_norm": 1.2385574579238892,
+      "learning_rate": 0.00019645528258983755,
+      "loss": 2.6611,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08562992125984252,
+      "grad_norm": 0.8684089779853821,
+      "learning_rate": 0.00019645017778031128,
+      "loss": 3.9031,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08569143700787402,
+      "grad_norm": 0.9566271901130676,
+      "learning_rate": 0.00019644506936408615,
+      "loss": 3.8327,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08575295275590551,
+      "grad_norm": 1.1909900903701782,
+      "learning_rate": 0.00019643995734135327,
+      "loss": 3.1189,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08581446850393701,
+      "grad_norm": 0.9863600134849548,
+      "learning_rate": 0.00019643484171230372,
+      "loss": 3.4075,
+      "step": 1395
+    },
+    {
+      "epoch": 0.0858759842519685,
+      "grad_norm": 1.0813329219818115,
+      "learning_rate": 0.0001964297224771288,
+      "loss": 3.0629,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0859375,
+      "grad_norm": 0.998145341873169,
+      "learning_rate": 0.00019642459963602,
+      "loss": 2.4295,
+      "step": 1397
+    },
+    {
+      "epoch": 0.0859990157480315,
+      "grad_norm": 0.9107404351234436,
+      "learning_rate": 0.00019641947318916881,
+      "loss": 3.3199,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08606053149606299,
+      "grad_norm": 1.0151100158691406,
+      "learning_rate": 0.00019641434313676703,
+      "loss": 3.1641,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08612204724409449,
+      "grad_norm": 1.0510371923446655,
+      "learning_rate": 0.00019640920947900642,
+      "loss": 3.1301,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08618356299212598,
+      "grad_norm": 0.980666995048523,
+      "learning_rate": 0.00019640407221607894,
+      "loss": 2.7292,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08624507874015748,
+      "grad_norm": 0.9929638504981995,
+      "learning_rate": 0.00019639893134817675,
+      "loss": 3.4336,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08630659448818898,
+      "grad_norm": 1.1567655801773071,
+      "learning_rate": 0.00019639378687549207,
+      "loss": 3.0931,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08636811023622047,
+      "grad_norm": 1.074410080909729,
+      "learning_rate": 0.00019638863879821726,
+      "loss": 3.5554,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08642962598425197,
+      "grad_norm": 0.9855176210403442,
+      "learning_rate": 0.0001963834871165448,
+      "loss": 3.4209,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08649114173228346,
+      "grad_norm": 1.1772996187210083,
+      "learning_rate": 0.00019637833183066742,
+      "loss": 3.2106,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08655265748031496,
+      "grad_norm": 1.0148838758468628,
+      "learning_rate": 0.00019637317294077784,
+      "loss": 2.8693,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08661417322834646,
+      "grad_norm": 1.0662477016448975,
+      "learning_rate": 0.00019636801044706895,
+      "loss": 3.1811,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08667568897637795,
+      "grad_norm": 0.9249662160873413,
+      "learning_rate": 0.00019636284434973388,
+      "loss": 3.2733,
+      "step": 1409
+    },
+    {
+      "epoch": 0.08673720472440945,
+      "grad_norm": 1.0854074954986572,
+      "learning_rate": 0.00019635767464896574,
+      "loss": 3.0124,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08679872047244094,
+      "grad_norm": 0.9868936538696289,
+      "learning_rate": 0.00019635250134495785,
+      "loss": 2.8762,
+      "step": 1411
+    },
+    {
+      "epoch": 0.08686023622047244,
+      "grad_norm": 1.1523535251617432,
+      "learning_rate": 0.00019634732443790372,
+      "loss": 3.1841,
+      "step": 1412
+    },
+    {
+      "epoch": 0.08692175196850394,
+      "grad_norm": 1.4375278949737549,
+      "learning_rate": 0.0001963421439279969,
+      "loss": 2.9363,
+      "step": 1413
+    },
+    {
+      "epoch": 0.08698326771653543,
+      "grad_norm": 2.4087910652160645,
+      "learning_rate": 0.00019633695981543113,
+      "loss": 1.61,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08704478346456693,
+      "grad_norm": 1.0860693454742432,
+      "learning_rate": 0.00019633177210040022,
+      "loss": 3.4902,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08710629921259842,
+      "grad_norm": 1.1409573554992676,
+      "learning_rate": 0.0001963265807830982,
+      "loss": 3.0425,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08716781496062992,
+      "grad_norm": 1.2634928226470947,
+      "learning_rate": 0.0001963213858637192,
+      "loss": 3.2292,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08722933070866142,
+      "grad_norm": 1.1671264171600342,
+      "learning_rate": 0.00019631618734245747,
+      "loss": 2.757,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08729084645669291,
+      "grad_norm": 1.1523488759994507,
+      "learning_rate": 0.0001963109852195074,
+      "loss": 3.1191,
+      "step": 1419
+    },
+    {
+      "epoch": 0.0873523622047244,
+      "grad_norm": 1.1186935901641846,
+      "learning_rate": 0.00019630577949506352,
+      "loss": 3.0183,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0874138779527559,
+      "grad_norm": 1.0734483003616333,
+      "learning_rate": 0.00019630057016932053,
+      "loss": 3.6442,
+      "step": 1421
+    },
+    {
+      "epoch": 0.0874753937007874,
+      "grad_norm": 0.9546296000480652,
+      "learning_rate": 0.00019629535724247318,
+      "loss": 3.0013,
+      "step": 1422
+    },
+    {
+      "epoch": 0.0875369094488189,
+      "grad_norm": 1.0877541303634644,
+      "learning_rate": 0.00019629014071471644,
+      "loss": 3.1036,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08759842519685039,
+      "grad_norm": 1.0291708707809448,
+      "learning_rate": 0.00019628492058624536,
+      "loss": 2.1671,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08765994094488189,
+      "grad_norm": 1.4959622621536255,
+      "learning_rate": 0.00019627969685725516,
+      "loss": 2.7909,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08772145669291338,
+      "grad_norm": 0.8676504492759705,
+      "learning_rate": 0.00019627446952794116,
+      "loss": 3.671,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08778297244094488,
+      "grad_norm": 1.3619816303253174,
+      "learning_rate": 0.00019626923859849882,
+      "loss": 2.8606,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08784448818897637,
+      "grad_norm": 2.0160131454467773,
+      "learning_rate": 0.00019626400406912378,
+      "loss": 2.7188,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08790600393700787,
+      "grad_norm": 1.283474087715149,
+      "learning_rate": 0.0001962587659400118,
+      "loss": 3.0917,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08796751968503937,
+      "grad_norm": 1.2024790048599243,
+      "learning_rate": 0.0001962535242113587,
+      "loss": 3.0914,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08802903543307086,
+      "grad_norm": 1.0086761713027954,
+      "learning_rate": 0.00019624827888336052,
+      "loss": 3.5813,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08809055118110236,
+      "grad_norm": 1.122427225112915,
+      "learning_rate": 0.00019624302995621342,
+      "loss": 2.6986,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08815206692913385,
+      "grad_norm": 0.9505204558372498,
+      "learning_rate": 0.00019623777743011365,
+      "loss": 3.3105,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08821358267716535,
+      "grad_norm": 0.9694671034812927,
+      "learning_rate": 0.00019623252130525766,
+      "loss": 3.4018,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08827509842519685,
+      "grad_norm": 1.1842195987701416,
+      "learning_rate": 0.00019622726158184195,
+      "loss": 3.2071,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08833661417322834,
+      "grad_norm": 1.1237988471984863,
+      "learning_rate": 0.00019622199826006327,
+      "loss": 2.9975,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08839812992125984,
+      "grad_norm": 1.1798346042633057,
+      "learning_rate": 0.00019621673134011838,
+      "loss": 3.0506,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08845964566929133,
+      "grad_norm": 0.9569427371025085,
+      "learning_rate": 0.00019621146082220424,
+      "loss": 3.4038,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08852116141732283,
+      "grad_norm": 1.1948624849319458,
+      "learning_rate": 0.000196206186706518,
+      "loss": 3.7471,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08858267716535433,
+      "grad_norm": 0.9794732928276062,
+      "learning_rate": 0.00019620090899325678,
+      "loss": 3.1977,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08864419291338582,
+      "grad_norm": 1.146437644958496,
+      "learning_rate": 0.00019619562768261807,
+      "loss": 3.2498,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08870570866141732,
+      "grad_norm": 1.0142436027526855,
+      "learning_rate": 0.00019619034277479922,
+      "loss": 2.9694,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08876722440944881,
+      "grad_norm": 1.2073291540145874,
+      "learning_rate": 0.00019618505426999793,
+      "loss": 3.7414,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08882874015748031,
+      "grad_norm": 0.8664063811302185,
+      "learning_rate": 0.00019617976216841198,
+      "loss": 3.5719,
+      "step": 1444
+    },
+    {
+      "epoch": 0.0888902559055118,
+      "grad_norm": 1.1509360074996948,
+      "learning_rate": 0.00019617446647023922,
+      "loss": 2.6193,
+      "step": 1445
+    },
+    {
+      "epoch": 0.0889517716535433,
+      "grad_norm": 1.211380958557129,
+      "learning_rate": 0.00019616916717567773,
+      "loss": 3.1556,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0890132874015748,
+      "grad_norm": 1.1797057390213013,
+      "learning_rate": 0.0001961638642849256,
+      "loss": 2.9891,
+      "step": 1447
+    },
+    {
+      "epoch": 0.0890748031496063,
+      "grad_norm": 0.9932965636253357,
+      "learning_rate": 0.0001961585577981812,
+      "loss": 2.636,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08913631889763779,
+      "grad_norm": 0.9464065432548523,
+      "learning_rate": 0.0001961532477156429,
+      "loss": 3.3524,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08919783464566929,
+      "grad_norm": 0.9587006568908691,
+      "learning_rate": 0.00019614793403750934,
+      "loss": 3.2017,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08925935039370078,
+      "grad_norm": 1.1871472597122192,
+      "learning_rate": 0.00019614261676397917,
+      "loss": 3.5612,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08932086614173228,
+      "grad_norm": 1.178821086883545,
+      "learning_rate": 0.00019613729589525124,
+      "loss": 3.6466,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08938238188976377,
+      "grad_norm": 1.070365309715271,
+      "learning_rate": 0.0001961319714315245,
+      "loss": 2.8685,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08944389763779527,
+      "grad_norm": 0.9462019801139832,
+      "learning_rate": 0.00019612664337299808,
+      "loss": 3.5174,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08950541338582677,
+      "grad_norm": 1.1890416145324707,
+      "learning_rate": 0.00019612131171987122,
+      "loss": 2.7169,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08956692913385826,
+      "grad_norm": 0.9393964409828186,
+      "learning_rate": 0.0001961159764723433,
+      "loss": 2.9709,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08962844488188976,
+      "grad_norm": 1.3590035438537598,
+      "learning_rate": 0.0001961106376306138,
+      "loss": 2.8818,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08968996062992125,
+      "grad_norm": 1.2976783514022827,
+      "learning_rate": 0.0001961052951948824,
+      "loss": 2.929,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08975147637795275,
+      "grad_norm": 1.0005627870559692,
+      "learning_rate": 0.00019609994916534884,
+      "loss": 3.1515,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08981299212598425,
+      "grad_norm": 1.4387190341949463,
+      "learning_rate": 0.00019609459954221306,
+      "loss": 3.0162,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08987450787401575,
+      "grad_norm": 1.0599027872085571,
+      "learning_rate": 0.00019608924632567507,
+      "loss": 3.7184,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08993602362204725,
+      "grad_norm": 1.027766227722168,
+      "learning_rate": 0.0001960838895159351,
+      "loss": 2.7544,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08999753937007875,
+      "grad_norm": 0.9313614964485168,
+      "learning_rate": 0.00019607852911319345,
+      "loss": 3.0446,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09005905511811024,
+      "grad_norm": 0.996875524520874,
+      "learning_rate": 0.00019607316511765054,
+      "loss": 2.9104,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09012057086614174,
+      "grad_norm": 1.2142434120178223,
+      "learning_rate": 0.00019606779752950697,
+      "loss": 3.5917,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09018208661417323,
+      "grad_norm": 1.1188429594039917,
+      "learning_rate": 0.00019606242634896348,
+      "loss": 2.797,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09024360236220473,
+      "grad_norm": 1.080042839050293,
+      "learning_rate": 0.00019605705157622088,
+      "loss": 2.9507,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09030511811023623,
+      "grad_norm": 0.7989991903305054,
+      "learning_rate": 0.00019605167321148023,
+      "loss": 3.0977,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09036663385826772,
+      "grad_norm": 1.1107628345489502,
+      "learning_rate": 0.00019604629125494257,
+      "loss": 3.4693,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09042814960629922,
+      "grad_norm": 1.1107494831085205,
+      "learning_rate": 0.00019604090570680917,
+      "loss": 3.1678,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09048966535433071,
+      "grad_norm": 0.8681801557540894,
+      "learning_rate": 0.00019603551656728144,
+      "loss": 3.4391,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09055118110236221,
+      "grad_norm": 1.097855806350708,
+      "learning_rate": 0.00019603012383656088,
+      "loss": 3.4255,
+      "step": 1472
+    },
+    {
+      "epoch": 0.0906126968503937,
+      "grad_norm": 1.010108470916748,
+      "learning_rate": 0.00019602472751484922,
+      "loss": 3.147,
+      "step": 1473
+    },
+    {
+      "epoch": 0.0906742125984252,
+      "grad_norm": 1.2888110876083374,
+      "learning_rate": 0.00019601932760234818,
+      "loss": 3.2368,
+      "step": 1474
+    },
+    {
+      "epoch": 0.0907357283464567,
+      "grad_norm": 1.0912327766418457,
+      "learning_rate": 0.0001960139240992597,
+      "loss": 3.3175,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0907972440944882,
+      "grad_norm": 1.0597463846206665,
+      "learning_rate": 0.00019600851700578585,
+      "loss": 2.8034,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09085875984251969,
+      "grad_norm": 1.0188754796981812,
+      "learning_rate": 0.0001960031063221288,
+      "loss": 3.4744,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09092027559055119,
+      "grad_norm": 0.9478357434272766,
+      "learning_rate": 0.0001959976920484909,
+      "loss": 3.1236,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09098179133858268,
+      "grad_norm": 1.0317485332489014,
+      "learning_rate": 0.00019599227418507463,
+      "loss": 3.6597,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09104330708661418,
+      "grad_norm": 1.259272575378418,
+      "learning_rate": 0.00019598685273208258,
+      "loss": 3.219,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09110482283464567,
+      "grad_norm": 1.1214454174041748,
+      "learning_rate": 0.00019598142768971748,
+      "loss": 3.1982,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09116633858267717,
+      "grad_norm": 1.0140573978424072,
+      "learning_rate": 0.0001959759990581822,
+      "loss": 2.8569,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09122785433070867,
+      "grad_norm": 1.019126057624817,
+      "learning_rate": 0.00019597056683767972,
+      "loss": 3.1137,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09128937007874016,
+      "grad_norm": 0.9729040861129761,
+      "learning_rate": 0.00019596513102841317,
+      "loss": 3.1398,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09135088582677166,
+      "grad_norm": 1.0077964067459106,
+      "learning_rate": 0.00019595969163058586,
+      "loss": 2.7765,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09141240157480315,
+      "grad_norm": 1.007948398590088,
+      "learning_rate": 0.00019595424864440113,
+      "loss": 3.5558,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09147391732283465,
+      "grad_norm": 1.2501946687698364,
+      "learning_rate": 0.00019594880207006262,
+      "loss": 3.3725,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09153543307086615,
+      "grad_norm": 0.9611783027648926,
+      "learning_rate": 0.00019594335190777388,
+      "loss": 3.1685,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09159694881889764,
+      "grad_norm": 0.9896363615989685,
+      "learning_rate": 0.0001959378981577388,
+      "loss": 2.3247,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09165846456692914,
+      "grad_norm": 1.208739161491394,
+      "learning_rate": 0.00019593244082016128,
+      "loss": 2.9438,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09171998031496063,
+      "grad_norm": 1.1538095474243164,
+      "learning_rate": 0.00019592697989524542,
+      "loss": 2.6877,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09178149606299213,
+      "grad_norm": 1.0899155139923096,
+      "learning_rate": 0.0001959215153831954,
+      "loss": 3.5134,
+      "step": 1492
+    },
+    {
+      "epoch": 0.09184301181102363,
+      "grad_norm": 1.082905650138855,
+      "learning_rate": 0.0001959160472842156,
+      "loss": 3.2477,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09190452755905512,
+      "grad_norm": 1.5146031379699707,
+      "learning_rate": 0.00019591057559851044,
+      "loss": 3.0407,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09196604330708662,
+      "grad_norm": 1.7728873491287231,
+      "learning_rate": 0.00019590510032628462,
+      "loss": 4.109,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09202755905511811,
+      "grad_norm": 0.9761096239089966,
+      "learning_rate": 0.00019589962146774278,
+      "loss": 2.7116,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09208907480314961,
+      "grad_norm": 1.137981653213501,
+      "learning_rate": 0.00019589413902308987,
+      "loss": 3.2007,
+      "step": 1497
+    },
+    {
+      "epoch": 0.0921505905511811,
+      "grad_norm": 1.0118353366851807,
+      "learning_rate": 0.00019588865299253086,
+      "loss": 3.5068,
+      "step": 1498
+    },
+    {
+      "epoch": 0.0922121062992126,
+      "grad_norm": 1.3862146139144897,
+      "learning_rate": 0.00019588316337627095,
+      "loss": 2.2654,
+      "step": 1499
+    },
+    {
+      "epoch": 0.0922736220472441,
+      "grad_norm": 1.4300364255905151,
+      "learning_rate": 0.00019587767017451536,
+      "loss": 2.167,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0923351377952756,
+      "grad_norm": 1.1539158821105957,
+      "learning_rate": 0.00019587217338746954,
+      "loss": 2.9174,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09239665354330709,
+      "grad_norm": 0.9289186596870422,
+      "learning_rate": 0.00019586667301533904,
+      "loss": 3.5571,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09245816929133858,
+      "grad_norm": 1.217172384262085,
+      "learning_rate": 0.00019586116905832954,
+      "loss": 2.8177,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09251968503937008,
+      "grad_norm": 0.9307460188865662,
+      "learning_rate": 0.00019585566151664687,
+      "loss": 3.378,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09258120078740158,
+      "grad_norm": 1.2297452688217163,
+      "learning_rate": 0.00019585015039049695,
+      "loss": 3.2952,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09264271653543307,
+      "grad_norm": 1.2930610179901123,
+      "learning_rate": 0.00019584463568008586,
+      "loss": 2.7581,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09270423228346457,
+      "grad_norm": 1.1169315576553345,
+      "learning_rate": 0.00019583911738561987,
+      "loss": 1.9515,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09276574803149606,
+      "grad_norm": 1.2458691596984863,
+      "learning_rate": 0.00019583359550730527,
+      "loss": 2.4913,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09282726377952756,
+      "grad_norm": 0.8675018548965454,
+      "learning_rate": 0.00019582807004534859,
+      "loss": 3.3297,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09288877952755906,
+      "grad_norm": 0.8584277629852295,
+      "learning_rate": 0.00019582254099995646,
+      "loss": 2.8296,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09295029527559055,
+      "grad_norm": 0.9365121722221375,
+      "learning_rate": 0.00019581700837133562,
+      "loss": 3.0675,
+      "step": 1511
+    },
+    {
+      "epoch": 0.09301181102362205,
+      "grad_norm": 1.0517319440841675,
+      "learning_rate": 0.00019581147215969296,
+      "loss": 2.9374,
+      "step": 1512
+    },
+    {
+      "epoch": 0.09307332677165354,
+      "grad_norm": 1.052337408065796,
+      "learning_rate": 0.0001958059323652355,
+      "loss": 2.9738,
+      "step": 1513
+    },
+    {
+      "epoch": 0.09313484251968504,
+      "grad_norm": 1.467888593673706,
+      "learning_rate": 0.00019580038898817036,
+      "loss": 2.6466,
+      "step": 1514
+    },
+    {
+      "epoch": 0.09319635826771654,
+      "grad_norm": 1.6226106882095337,
+      "learning_rate": 0.00019579484202870488,
+      "loss": 2.9099,
+      "step": 1515
+    },
+    {
+      "epoch": 0.09325787401574803,
+      "grad_norm": 0.9807246327400208,
+      "learning_rate": 0.0001957892914870465,
+      "loss": 3.2896,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09331938976377953,
+      "grad_norm": 0.8917344808578491,
+      "learning_rate": 0.00019578373736340275,
+      "loss": 3.738,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09338090551181102,
+      "grad_norm": 0.9444370269775391,
+      "learning_rate": 0.0001957781796579813,
+      "loss": 3.4388,
+      "step": 1518
+    },
+    {
+      "epoch": 0.09344242125984252,
+      "grad_norm": 1.2648184299468994,
+      "learning_rate": 0.00019577261837099003,
+      "loss": 3.1261,
+      "step": 1519
+    },
+    {
+      "epoch": 0.09350393700787402,
+      "grad_norm": 0.9885039329528809,
+      "learning_rate": 0.00019576705350263688,
+      "loss": 3.0583,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09356545275590551,
+      "grad_norm": 1.2773464918136597,
+      "learning_rate": 0.00019576148505312993,
+      "loss": 3.2277,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09362696850393701,
+      "grad_norm": 1.024250864982605,
+      "learning_rate": 0.0001957559130226774,
+      "loss": 3.5921,
+      "step": 1522
+    },
+    {
+      "epoch": 0.0936884842519685,
+      "grad_norm": 1.1208009719848633,
+      "learning_rate": 0.0001957503374114877,
+      "loss": 3.2404,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 1.0200251340866089,
+      "learning_rate": 0.0001957447582197693,
+      "loss": 2.8882,
+      "step": 1524
+    },
+    {
+      "epoch": 0.0938115157480315,
+      "grad_norm": 0.849252462387085,
+      "learning_rate": 0.00019573917544773084,
+      "loss": 3.1611,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09387303149606299,
+      "grad_norm": 1.0108615159988403,
+      "learning_rate": 0.00019573358909558106,
+      "loss": 3.3659,
+      "step": 1526
+    },
+    {
+      "epoch": 0.09393454724409449,
+      "grad_norm": 0.8963862061500549,
+      "learning_rate": 0.00019572799916352886,
+      "loss": 3.0139,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09399606299212598,
+      "grad_norm": 1.0505740642547607,
+      "learning_rate": 0.0001957224056517833,
+      "loss": 2.9468,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09405757874015748,
+      "grad_norm": 1.1642138957977295,
+      "learning_rate": 0.00019571680856055355,
+      "loss": 3.3887,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09411909448818898,
+      "grad_norm": 1.4759758710861206,
+      "learning_rate": 0.00019571120789004884,
+      "loss": 3.3533,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09418061023622047,
+      "grad_norm": 0.9786637425422668,
+      "learning_rate": 0.0001957056036404787,
+      "loss": 3.2813,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09424212598425197,
+      "grad_norm": 0.9126378297805786,
+      "learning_rate": 0.00019569999581205263,
+      "loss": 3.5421,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09430364173228346,
+      "grad_norm": 0.9580281376838684,
+      "learning_rate": 0.00019569438440498037,
+      "loss": 3.4481,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09436515748031496,
+      "grad_norm": 0.9448336362838745,
+      "learning_rate": 0.00019568876941947173,
+      "loss": 2.8467,
+      "step": 1534
+    },
+    {
+      "epoch": 0.09442667322834646,
+      "grad_norm": 1.4474658966064453,
+      "learning_rate": 0.00019568315085573667,
+      "loss": 2.8208,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09448818897637795,
+      "grad_norm": 1.0772044658660889,
+      "learning_rate": 0.00019567752871398536,
+      "loss": 2.8978,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09454970472440945,
+      "grad_norm": 1.0101995468139648,
+      "learning_rate": 0.00019567190299442795,
+      "loss": 2.9986,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09461122047244094,
+      "grad_norm": 1.1409443616867065,
+      "learning_rate": 0.00019566627369727485,
+      "loss": 2.9239,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09467273622047244,
+      "grad_norm": 0.8053903579711914,
+      "learning_rate": 0.0001956606408227366,
+      "loss": 3.658,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09473425196850394,
+      "grad_norm": 1.1719447374343872,
+      "learning_rate": 0.00019565500437102376,
+      "loss": 2.7841,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09479576771653543,
+      "grad_norm": 0.9472525119781494,
+      "learning_rate": 0.00019564936434234714,
+      "loss": 3.5594,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09485728346456693,
+      "grad_norm": 1.1015777587890625,
+      "learning_rate": 0.00019564372073691767,
+      "loss": 3.2486,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09491879921259842,
+      "grad_norm": 0.8852831721305847,
+      "learning_rate": 0.00019563807355494638,
+      "loss": 2.8559,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09498031496062992,
+      "grad_norm": 1.1422040462493896,
+      "learning_rate": 0.00019563242279664443,
+      "loss": 3.1645,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09504183070866142,
+      "grad_norm": 0.945395827293396,
+      "learning_rate": 0.0001956267684622231,
+      "loss": 2.9031,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09510334645669291,
+      "grad_norm": 1.7652941942214966,
+      "learning_rate": 0.00019562111055189388,
+      "loss": 2.8053,
+      "step": 1546
+    },
+    {
+      "epoch": 0.0951648622047244,
+      "grad_norm": 1.8170160055160522,
+      "learning_rate": 0.00019561544906586832,
+      "loss": 1.2782,
+      "step": 1547
+    },
+    {
+      "epoch": 0.0952263779527559,
+      "grad_norm": 1.0966241359710693,
+      "learning_rate": 0.00019560978400435811,
+      "loss": 3.3009,
+      "step": 1548
+    },
+    {
+      "epoch": 0.0952878937007874,
+      "grad_norm": 1.6659092903137207,
+      "learning_rate": 0.00019560411536757516,
+      "loss": 3.8671,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0953494094488189,
+      "grad_norm": 1.776224136352539,
+      "learning_rate": 0.00019559844315573136,
+      "loss": 2.758,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09541092519685039,
+      "grad_norm": 1.0678651332855225,
+      "learning_rate": 0.00019559276736903886,
+      "loss": 3.0063,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09547244094488189,
+      "grad_norm": 1.1212856769561768,
+      "learning_rate": 0.00019558708800770992,
+      "loss": 3.414,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09553395669291338,
+      "grad_norm": 0.8908339738845825,
+      "learning_rate": 0.0001955814050719569,
+      "loss": 3.3948,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09559547244094488,
+      "grad_norm": 0.8657857775688171,
+      "learning_rate": 0.0001955757185619923,
+      "loss": 3.2802,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09565698818897637,
+      "grad_norm": 1.052506923675537,
+      "learning_rate": 0.00019557002847802875,
+      "loss": 3.5039,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09571850393700787,
+      "grad_norm": 0.9430912137031555,
+      "learning_rate": 0.00019556433482027905,
+      "loss": 3.5561,
+      "step": 1556
+    },
+    {
+      "epoch": 0.09578001968503937,
+      "grad_norm": 1.0767494440078735,
+      "learning_rate": 0.0001955586375889561,
+      "loss": 3.5439,
+      "step": 1557
+    },
+    {
+      "epoch": 0.09584153543307086,
+      "grad_norm": 1.0406498908996582,
+      "learning_rate": 0.000195552936784273,
+      "loss": 3.1426,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09590305118110236,
+      "grad_norm": 1.1747924089431763,
+      "learning_rate": 0.00019554723240644285,
+      "loss": 3.1453,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09596456692913385,
+      "grad_norm": 1.148873209953308,
+      "learning_rate": 0.000195541524455679,
+      "loss": 2.2426,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09602608267716535,
+      "grad_norm": 1.3935390710830688,
+      "learning_rate": 0.00019553581293219489,
+      "loss": 2.809,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09608759842519685,
+      "grad_norm": 1.16655433177948,
+      "learning_rate": 0.0001955300978362041,
+      "loss": 2.9086,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09614911417322834,
+      "grad_norm": 1.288094401359558,
+      "learning_rate": 0.00019552437916792035,
+      "loss": 2.9677,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09621062992125984,
+      "grad_norm": 1.1570204496383667,
+      "learning_rate": 0.00019551865692755746,
+      "loss": 3.0677,
+      "step": 1564
+    },
+    {
+      "epoch": 0.09627214566929133,
+      "grad_norm": 1.1229149103164673,
+      "learning_rate": 0.00019551293111532943,
+      "loss": 3.1842,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09633366141732283,
+      "grad_norm": 0.9438603520393372,
+      "learning_rate": 0.00019550720173145038,
+      "loss": 3.2214,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09639517716535433,
+      "grad_norm": 0.9971490502357483,
+      "learning_rate": 0.00019550146877613455,
+      "loss": 2.4115,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09645669291338582,
+      "grad_norm": 1.0847529172897339,
+      "learning_rate": 0.0001954957322495963,
+      "loss": 2.5792,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09651820866141732,
+      "grad_norm": 1.110504388809204,
+      "learning_rate": 0.00019548999215205018,
+      "loss": 2.4232,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09657972440944881,
+      "grad_norm": 1.050128698348999,
+      "learning_rate": 0.0001954842484837108,
+      "loss": 3.3132,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09664124015748031,
+      "grad_norm": 1.0960785150527954,
+      "learning_rate": 0.000195478501244793,
+      "loss": 3.2288,
+      "step": 1571
+    },
+    {
+      "epoch": 0.0967027559055118,
+      "grad_norm": 1.0412962436676025,
+      "learning_rate": 0.0001954727504355116,
+      "loss": 3.0765,
+      "step": 1572
+    },
+    {
+      "epoch": 0.0967642716535433,
+      "grad_norm": 1.296726942062378,
+      "learning_rate": 0.00019546699605608176,
+      "loss": 3.6713,
+      "step": 1573
+    },
+    {
+      "epoch": 0.0968257874015748,
+      "grad_norm": 0.7591187357902527,
+      "learning_rate": 0.0001954612381067186,
+      "loss": 2.5546,
+      "step": 1574
+    },
+    {
+      "epoch": 0.0968873031496063,
+      "grad_norm": 1.389554500579834,
+      "learning_rate": 0.00019545547658763742,
+      "loss": 3.087,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09694881889763779,
+      "grad_norm": 0.9995411038398743,
+      "learning_rate": 0.00019544971149905368,
+      "loss": 3.1613,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09701033464566929,
+      "grad_norm": 1.25492525100708,
+      "learning_rate": 0.00019544394284118298,
+      "loss": 2.6681,
+      "step": 1577
+    },
+    {
+      "epoch": 0.09707185039370078,
+      "grad_norm": 1.2039169073104858,
+      "learning_rate": 0.00019543817061424106,
+      "loss": 3.4563,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09713336614173228,
+      "grad_norm": 0.9585862159729004,
+      "learning_rate": 0.0001954323948184437,
+      "loss": 2.5711,
+      "step": 1579
+    },
+    {
+      "epoch": 0.09719488188976377,
+      "grad_norm": 0.9256117343902588,
+      "learning_rate": 0.00019542661545400694,
+      "loss": 3.6488,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09725639763779527,
+      "grad_norm": 1.179465651512146,
+      "learning_rate": 0.00019542083252114685,
+      "loss": 3.3639,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09731791338582677,
+      "grad_norm": 1.063880443572998,
+      "learning_rate": 0.00019541504602007973,
+      "loss": 3.7541,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09737942913385826,
+      "grad_norm": 1.0672614574432373,
+      "learning_rate": 0.00019540925595102192,
+      "loss": 3.2562,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09744094488188976,
+      "grad_norm": 1.158829689025879,
+      "learning_rate": 0.00019540346231418997,
+      "loss": 3.2088,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09750246062992125,
+      "grad_norm": 1.065653920173645,
+      "learning_rate": 0.0001953976651098005,
+      "loss": 2.9853,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09756397637795275,
+      "grad_norm": 1.3315045833587646,
+      "learning_rate": 0.00019539186433807033,
+      "loss": 2.9937,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09762549212598425,
+      "grad_norm": 0.9102661609649658,
+      "learning_rate": 0.00019538605999921632,
+      "loss": 3.4724,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09768700787401575,
+      "grad_norm": 1.4954041242599487,
+      "learning_rate": 0.00019538025209345556,
+      "loss": 2.547,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09774852362204725,
+      "grad_norm": 1.0284899473190308,
+      "learning_rate": 0.00019537444062100526,
+      "loss": 2.6437,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09781003937007875,
+      "grad_norm": 1.0256600379943848,
+      "learning_rate": 0.00019536862558208265,
+      "loss": 3.0651,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09787155511811024,
+      "grad_norm": 1.1532219648361206,
+      "learning_rate": 0.00019536280697690528,
+      "loss": 3.1175,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09793307086614174,
+      "grad_norm": 1.2762154340744019,
+      "learning_rate": 0.00019535698480569068,
+      "loss": 3.3809,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09799458661417323,
+      "grad_norm": 1.0738164186477661,
+      "learning_rate": 0.00019535115906865654,
+      "loss": 2.9758,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09805610236220473,
+      "grad_norm": 1.085594892501831,
+      "learning_rate": 0.00019534532976602079,
+      "loss": 2.8898,
+      "step": 1594
+    },
+    {
+      "epoch": 0.09811761811023623,
+      "grad_norm": 1.274678111076355,
+      "learning_rate": 0.00019533949689800134,
+      "loss": 3.7984,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09817913385826772,
+      "grad_norm": 1.0148051977157593,
+      "learning_rate": 0.00019533366046481634,
+      "loss": 3.0284,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09824064960629922,
+      "grad_norm": 0.9475074410438538,
+      "learning_rate": 0.000195327820466684,
+      "loss": 2.6387,
+      "step": 1597
+    },
+    {
+      "epoch": 0.09830216535433071,
+      "grad_norm": 1.1150304079055786,
+      "learning_rate": 0.0001953219769038228,
+      "loss": 3.2744,
+      "step": 1598
+    },
+    {
+      "epoch": 0.09836368110236221,
+      "grad_norm": 1.08745276927948,
+      "learning_rate": 0.00019531612977645115,
+      "loss": 2.9336,
+      "step": 1599
+    },
+    {
+      "epoch": 0.0984251968503937,
+      "grad_norm": 1.0528756380081177,
+      "learning_rate": 0.00019531027908478775,
+      "loss": 3.1246,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0984867125984252,
+      "grad_norm": 0.9901379942893982,
+      "learning_rate": 0.00019530442482905137,
+      "loss": 3.2291,
+      "step": 1601
+    },
+    {
+      "epoch": 0.0985482283464567,
+      "grad_norm": 0.9817074537277222,
+      "learning_rate": 0.00019529856700946098,
+      "loss": 2.9696,
+      "step": 1602
+    },
+    {
+      "epoch": 0.0986097440944882,
+      "grad_norm": 2.3268849849700928,
+      "learning_rate": 0.00019529270562623554,
+      "loss": 3.2358,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09867125984251969,
+      "grad_norm": 2.6864376068115234,
+      "learning_rate": 0.0001952868406795943,
+      "loss": 1.9611,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09873277559055119,
+      "grad_norm": 1.2535426616668701,
+      "learning_rate": 0.00019528097216975652,
+      "loss": 3.1706,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09879429133858268,
+      "grad_norm": 1.2260714769363403,
+      "learning_rate": 0.0001952751000969417,
+      "loss": 2.7446,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09885580708661418,
+      "grad_norm": 1.0448212623596191,
+      "learning_rate": 0.00019526922446136942,
+      "loss": 3.4677,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09891732283464567,
+      "grad_norm": 1.0969465970993042,
+      "learning_rate": 0.00019526334526325932,
+      "loss": 3.284,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09897883858267717,
+      "grad_norm": 1.1073904037475586,
+      "learning_rate": 0.0001952574625028314,
+      "loss": 3.255,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09904035433070867,
+      "grad_norm": 1.130324363708496,
+      "learning_rate": 0.0001952515761803055,
+      "loss": 2.9681,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09910187007874016,
+      "grad_norm": 0.9500828981399536,
+      "learning_rate": 0.0001952456862959018,
+      "loss": 2.9925,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09916338582677166,
+      "grad_norm": 0.9715516567230225,
+      "learning_rate": 0.00019523979284984054,
+      "loss": 3.3005,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09922490157480315,
+      "grad_norm": 0.9619452953338623,
+      "learning_rate": 0.00019523389584234207,
+      "loss": 2.5175,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09928641732283465,
+      "grad_norm": 1.2279270887374878,
+      "learning_rate": 0.00019522799527362696,
+      "loss": 2.8925,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09934793307086615,
+      "grad_norm": 0.8944621086120605,
+      "learning_rate": 0.00019522209114391584,
+      "loss": 3.0658,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09940944881889764,
+      "grad_norm": 0.9578334093093872,
+      "learning_rate": 0.00019521618345342946,
+      "loss": 2.9098,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09947096456692914,
+      "grad_norm": 0.9918736219406128,
+      "learning_rate": 0.0001952102722023888,
+      "loss": 2.81,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09953248031496063,
+      "grad_norm": 1.1263360977172852,
+      "learning_rate": 0.00019520435739101482,
+      "loss": 3.5851,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09959399606299213,
+      "grad_norm": 1.233177661895752,
+      "learning_rate": 0.00019519843901952876,
+      "loss": 2.886,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09965551181102363,
+      "grad_norm": 1.061827301979065,
+      "learning_rate": 0.00019519251708815193,
+      "loss": 3.5626,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09971702755905512,
+      "grad_norm": 1.0694383382797241,
+      "learning_rate": 0.00019518659159710577,
+      "loss": 3.082,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09977854330708662,
+      "grad_norm": 1.1629586219787598,
+      "learning_rate": 0.00019518066254661185,
+      "loss": 2.5987,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09984005905511811,
+      "grad_norm": 1.177497148513794,
+      "learning_rate": 0.0001951747299368919,
+      "loss": 3.0607,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09990157480314961,
+      "grad_norm": 0.8664459586143494,
+      "learning_rate": 0.00019516879376816777,
+      "loss": 3.2824,
+      "step": 1624
+    },
+    {
+      "epoch": 0.0999630905511811,
+      "grad_norm": 0.9970129728317261,
+      "learning_rate": 0.0001951628540406614,
+      "loss": 3.4968,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1000246062992126,
+      "grad_norm": 1.0182820558547974,
+      "learning_rate": 0.00019515691075459496,
+      "loss": 3.4926,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1000861220472441,
+      "grad_norm": 1.0905342102050781,
+      "learning_rate": 0.00019515096391019065,
+      "loss": 3.2311,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1001476377952756,
+      "grad_norm": 1.009549617767334,
+      "learning_rate": 0.0001951450135076709,
+      "loss": 2.8212,
+      "step": 1628
+    },
+    {
+      "epoch": 0.10020915354330709,
+      "grad_norm": 1.2819478511810303,
+      "learning_rate": 0.00019513905954725815,
+      "loss": 3.3437,
+      "step": 1629
+    },
+    {
+      "epoch": 0.10027066929133858,
+      "grad_norm": 0.9420043230056763,
+      "learning_rate": 0.00019513310202917507,
+      "loss": 3.177,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10033218503937008,
+      "grad_norm": 0.8553197979927063,
+      "learning_rate": 0.00019512714095364447,
+      "loss": 1.7112,
+      "step": 1631
+    },
+    {
+      "epoch": 0.10039370078740158,
+      "grad_norm": 0.9895346164703369,
+      "learning_rate": 0.00019512117632088926,
+      "loss": 3.6127,
+      "step": 1632
+    },
+    {
+      "epoch": 0.10045521653543307,
+      "grad_norm": 1.0643768310546875,
+      "learning_rate": 0.00019511520813113244,
+      "loss": 3.2078,
+      "step": 1633
+    },
+    {
+      "epoch": 0.10051673228346457,
+      "grad_norm": 1.0129084587097168,
+      "learning_rate": 0.00019510923638459723,
+      "loss": 3.1349,
+      "step": 1634
+    },
+    {
+      "epoch": 0.10057824803149606,
+      "grad_norm": 1.1877909898757935,
+      "learning_rate": 0.00019510326108150694,
+      "loss": 2.459,
+      "step": 1635
+    },
+    {
+      "epoch": 0.10063976377952756,
+      "grad_norm": 0.9873893857002258,
+      "learning_rate": 0.00019509728222208495,
+      "loss": 3.7624,
+      "step": 1636
+    },
+    {
+      "epoch": 0.10070127952755906,
+      "grad_norm": 1.0100799798965454,
+      "learning_rate": 0.00019509129980655493,
+      "loss": 3.231,
+      "step": 1637
+    },
+    {
+      "epoch": 0.10076279527559055,
+      "grad_norm": 0.9446800351142883,
+      "learning_rate": 0.0001950853138351405,
+      "loss": 2.9844,
+      "step": 1638
+    },
+    {
+      "epoch": 0.10082431102362205,
+      "grad_norm": 1.1726150512695312,
+      "learning_rate": 0.00019507932430806556,
+      "loss": 2.5454,
+      "step": 1639
+    },
+    {
+      "epoch": 0.10088582677165354,
+      "grad_norm": 1.2232469320297241,
+      "learning_rate": 0.00019507333122555407,
+      "loss": 2.9398,
+      "step": 1640
+    },
+    {
+      "epoch": 0.10094734251968504,
+      "grad_norm": 0.9530431628227234,
+      "learning_rate": 0.00019506733458783012,
+      "loss": 3.4314,
+      "step": 1641
+    },
+    {
+      "epoch": 0.10100885826771654,
+      "grad_norm": 1.0935062170028687,
+      "learning_rate": 0.00019506133439511797,
+      "loss": 2.9391,
+      "step": 1642
+    },
+    {
+      "epoch": 0.10107037401574803,
+      "grad_norm": 1.042476773262024,
+      "learning_rate": 0.00019505533064764202,
+      "loss": 3.3979,
+      "step": 1643
+    },
+    {
+      "epoch": 0.10113188976377953,
+      "grad_norm": 0.9291995167732239,
+      "learning_rate": 0.0001950493233456267,
+      "loss": 3.2358,
+      "step": 1644
+    },
+    {
+      "epoch": 0.10119340551181102,
+      "grad_norm": 1.1159831285476685,
+      "learning_rate": 0.00019504331248929675,
+      "loss": 2.6986,
+      "step": 1645
+    },
+    {
+      "epoch": 0.10125492125984252,
+      "grad_norm": 1.1931830644607544,
+      "learning_rate": 0.00019503729807887686,
+      "loss": 3.3752,
+      "step": 1646
+    },
+    {
+      "epoch": 0.10131643700787402,
+      "grad_norm": 1.0106078386306763,
+      "learning_rate": 0.00019503128011459197,
+      "loss": 3.2072,
+      "step": 1647
+    },
+    {
+      "epoch": 0.10137795275590551,
+      "grad_norm": 1.1162959337234497,
+      "learning_rate": 0.0001950252585966671,
+      "loss": 3.3587,
+      "step": 1648
+    },
+    {
+      "epoch": 0.10143946850393701,
+      "grad_norm": 1.3275763988494873,
+      "learning_rate": 0.00019501923352532744,
+      "loss": 2.9318,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1015009842519685,
+      "grad_norm": 1.265354037284851,
+      "learning_rate": 0.0001950132049007983,
+      "loss": 2.4244,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1015625,
+      "grad_norm": 0.8305611610412598,
+      "learning_rate": 0.00019500717272330508,
+      "loss": 2.9663,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1016240157480315,
+      "grad_norm": 0.8450983166694641,
+      "learning_rate": 0.00019500113699307342,
+      "loss": 3.7105,
+      "step": 1652
+    },
+    {
+      "epoch": 0.10168553149606299,
+      "grad_norm": 1.6029587984085083,
+      "learning_rate": 0.00019499509771032892,
+      "loss": 2.8706,
+      "step": 1653
+    },
+    {
+      "epoch": 0.10174704724409449,
+      "grad_norm": 1.29170560836792,
+      "learning_rate": 0.00019498905487529752,
+      "loss": 3.3245,
+      "step": 1654
+    },
+    {
+      "epoch": 0.10180856299212598,
+      "grad_norm": 1.4038050174713135,
+      "learning_rate": 0.0001949830084882051,
+      "loss": 1.9678,
+      "step": 1655
+    },
+    {
+      "epoch": 0.10187007874015748,
+      "grad_norm": 1.164819598197937,
+      "learning_rate": 0.00019497695854927782,
+      "loss": 2.6726,
+      "step": 1656
+    },
+    {
+      "epoch": 0.10193159448818898,
+      "grad_norm": 1.019464135169983,
+      "learning_rate": 0.0001949709050587419,
+      "loss": 3.0533,
+      "step": 1657
+    },
+    {
+      "epoch": 0.10199311023622047,
+      "grad_norm": 0.8062020540237427,
+      "learning_rate": 0.0001949648480168237,
+      "loss": 3.6954,
+      "step": 1658
+    },
+    {
+      "epoch": 0.10205462598425197,
+      "grad_norm": 0.9138056635856628,
+      "learning_rate": 0.00019495878742374972,
+      "loss": 3.1838,
+      "step": 1659
+    },
+    {
+      "epoch": 0.10211614173228346,
+      "grad_norm": 1.251105546951294,
+      "learning_rate": 0.00019495272327974661,
+      "loss": 3.3729,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10217765748031496,
+      "grad_norm": 1.175068974494934,
+      "learning_rate": 0.00019494665558504113,
+      "loss": 2.3575,
+      "step": 1661
+    },
+    {
+      "epoch": 0.10223917322834646,
+      "grad_norm": 0.8688063621520996,
+      "learning_rate": 0.00019494058433986014,
+      "loss": 3.2017,
+      "step": 1662
+    },
+    {
+      "epoch": 0.10230068897637795,
+      "grad_norm": 1.357554316520691,
+      "learning_rate": 0.0001949345095444307,
+      "loss": 3.137,
+      "step": 1663
+    },
+    {
+      "epoch": 0.10236220472440945,
+      "grad_norm": 1.1580027341842651,
+      "learning_rate": 0.00019492843119897998,
+      "loss": 2.9359,
+      "step": 1664
+    },
+    {
+      "epoch": 0.10242372047244094,
+      "grad_norm": 1.1359632015228271,
+      "learning_rate": 0.00019492234930373528,
+      "loss": 3.0755,
+      "step": 1665
+    },
+    {
+      "epoch": 0.10248523622047244,
+      "grad_norm": 1.156449317932129,
+      "learning_rate": 0.000194916263858924,
+      "loss": 2.9045,
+      "step": 1666
+    },
+    {
+      "epoch": 0.10254675196850394,
+      "grad_norm": 1.692869782447815,
+      "learning_rate": 0.00019491017486477372,
+      "loss": 2.6495,
+      "step": 1667
+    },
+    {
+      "epoch": 0.10260826771653543,
+      "grad_norm": 1.647037148475647,
+      "learning_rate": 0.00019490408232151213,
+      "loss": 3.7131,
+      "step": 1668
+    },
+    {
+      "epoch": 0.10266978346456693,
+      "grad_norm": 2.0870652198791504,
+      "learning_rate": 0.00019489798622936708,
+      "loss": 2.5929,
+      "step": 1669
+    },
+    {
+      "epoch": 0.10273129921259842,
+      "grad_norm": 1.051170825958252,
+      "learning_rate": 0.0001948918865885665,
+      "loss": 3.3908,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10279281496062992,
+      "grad_norm": 1.2347742319107056,
+      "learning_rate": 0.00019488578339933848,
+      "loss": 2.7816,
+      "step": 1671
+    },
+    {
+      "epoch": 0.10285433070866142,
+      "grad_norm": 0.9716838002204895,
+      "learning_rate": 0.00019487967666191128,
+      "loss": 2.1251,
+      "step": 1672
+    },
+    {
+      "epoch": 0.10291584645669291,
+      "grad_norm": 1.0054136514663696,
+      "learning_rate": 0.00019487356637651324,
+      "loss": 3.1206,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1029773622047244,
+      "grad_norm": 0.9073772430419922,
+      "learning_rate": 0.00019486745254337283,
+      "loss": 2.8122,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1030388779527559,
+      "grad_norm": 0.9047666788101196,
+      "learning_rate": 0.00019486133516271868,
+      "loss": 3.0322,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1031003937007874,
+      "grad_norm": 0.8488050103187561,
+      "learning_rate": 0.00019485521423477956,
+      "loss": 2.9376,
+      "step": 1676
+    },
+    {
+      "epoch": 0.1031619094488189,
+      "grad_norm": 1.0432411432266235,
+      "learning_rate": 0.00019484908975978438,
+      "loss": 2.7306,
+      "step": 1677
+    },
+    {
+      "epoch": 0.10322342519685039,
+      "grad_norm": 1.569118857383728,
+      "learning_rate": 0.00019484296173796214,
+      "loss": 2.8799,
+      "step": 1678
+    },
+    {
+      "epoch": 0.10328494094488189,
+      "grad_norm": 1.0753849744796753,
+      "learning_rate": 0.00019483683016954199,
+      "loss": 3.0225,
+      "step": 1679
+    },
+    {
+      "epoch": 0.10334645669291338,
+      "grad_norm": 0.9457221031188965,
+      "learning_rate": 0.00019483069505475318,
+      "loss": 3.7484,
+      "step": 1680
+    },
+    {
+      "epoch": 0.10340797244094488,
+      "grad_norm": 0.9616056084632874,
+      "learning_rate": 0.00019482455639382518,
+      "loss": 3.9435,
+      "step": 1681
+    },
+    {
+      "epoch": 0.10346948818897637,
+      "grad_norm": 1.1589272022247314,
+      "learning_rate": 0.00019481841418698755,
+      "loss": 3.3804,
+      "step": 1682
+    },
+    {
+      "epoch": 0.10353100393700787,
+      "grad_norm": 1.0445996522903442,
+      "learning_rate": 0.0001948122684344699,
+      "loss": 2.9374,
+      "step": 1683
+    },
+    {
+      "epoch": 0.10359251968503937,
+      "grad_norm": 1.1105830669403076,
+      "learning_rate": 0.00019480611913650213,
+      "loss": 3.3434,
+      "step": 1684
+    },
+    {
+      "epoch": 0.10365403543307086,
+      "grad_norm": 1.0911287069320679,
+      "learning_rate": 0.00019479996629331417,
+      "loss": 2.4294,
+      "step": 1685
+    },
+    {
+      "epoch": 0.10371555118110236,
+      "grad_norm": 1.1628919839859009,
+      "learning_rate": 0.00019479380990513608,
+      "loss": 2.8994,
+      "step": 1686
+    },
+    {
+      "epoch": 0.10377706692913385,
+      "grad_norm": 1.0417524576187134,
+      "learning_rate": 0.00019478764997219808,
+      "loss": 3.1673,
+      "step": 1687
+    },
+    {
+      "epoch": 0.10383858267716535,
+      "grad_norm": 1.0237410068511963,
+      "learning_rate": 0.00019478148649473053,
+      "loss": 3.3739,
+      "step": 1688
+    },
+    {
+      "epoch": 0.10390009842519685,
+      "grad_norm": 0.9229444861412048,
+      "learning_rate": 0.0001947753194729639,
+      "loss": 2.9329,
+      "step": 1689
+    },
+    {
+      "epoch": 0.10396161417322834,
+      "grad_norm": 1.081826090812683,
+      "learning_rate": 0.00019476914890712878,
+      "loss": 2.8595,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10402312992125984,
+      "grad_norm": 0.8234174251556396,
+      "learning_rate": 0.000194762974797456,
+      "loss": 3.0372,
+      "step": 1691
+    },
+    {
+      "epoch": 0.10408464566929133,
+      "grad_norm": 1.3565490245819092,
+      "learning_rate": 0.0001947567971441763,
+      "loss": 3.244,
+      "step": 1692
+    },
+    {
+      "epoch": 0.10414616141732283,
+      "grad_norm": 1.4011037349700928,
+      "learning_rate": 0.0001947506159475208,
+      "loss": 2.4778,
+      "step": 1693
+    },
+    {
+      "epoch": 0.10420767716535433,
+      "grad_norm": 1.1216230392456055,
+      "learning_rate": 0.0001947444312077206,
+      "loss": 2.852,
+      "step": 1694
+    },
+    {
+      "epoch": 0.10426919291338582,
+      "grad_norm": 1.3527147769927979,
+      "learning_rate": 0.000194738242925007,
+      "loss": 2.8901,
+      "step": 1695
+    },
+    {
+      "epoch": 0.10433070866141732,
+      "grad_norm": 0.8644570112228394,
+      "learning_rate": 0.0001947320510996114,
+      "loss": 3.3332,
+      "step": 1696
+    },
+    {
+      "epoch": 0.10439222440944881,
+      "grad_norm": 1.21994149684906,
+      "learning_rate": 0.00019472585573176526,
+      "loss": 2.2921,
+      "step": 1697
+    },
+    {
+      "epoch": 0.10445374015748031,
+      "grad_norm": 1.5412436723709106,
+      "learning_rate": 0.0001947196568217004,
+      "loss": 2.4385,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1045152559055118,
+      "grad_norm": 1.1868019104003906,
+      "learning_rate": 0.0001947134543696485,
+      "loss": 1.4761,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1045767716535433,
+      "grad_norm": 1.0659979581832886,
+      "learning_rate": 0.00019470724837584156,
+      "loss": 3.1094,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1046382874015748,
+      "grad_norm": 1.002234697341919,
+      "learning_rate": 0.00019470103884051162,
+      "loss": 3.6337,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1046998031496063,
+      "grad_norm": 0.8678863048553467,
+      "learning_rate": 0.00019469482576389094,
+      "loss": 2.7829,
+      "step": 1702
+    },
+    {
+      "epoch": 0.10476131889763779,
+      "grad_norm": 1.163622260093689,
+      "learning_rate": 0.00019468860914621176,
+      "loss": 2.6073,
+      "step": 1703
+    },
+    {
+      "epoch": 0.10482283464566929,
+      "grad_norm": 1.120011329650879,
+      "learning_rate": 0.0001946823889877066,
+      "loss": 2.4049,
+      "step": 1704
+    },
+    {
+      "epoch": 0.10488435039370078,
+      "grad_norm": 0.8137839436531067,
+      "learning_rate": 0.00019467616528860807,
+      "loss": 2.4458,
+      "step": 1705
+    },
+    {
+      "epoch": 0.10494586614173228,
+      "grad_norm": 1.0645921230316162,
+      "learning_rate": 0.00019466993804914887,
+      "loss": 3.3232,
+      "step": 1706
+    },
+    {
+      "epoch": 0.10500738188976377,
+      "grad_norm": 1.010878324508667,
+      "learning_rate": 0.0001946637072695619,
+      "loss": 2.843,
+      "step": 1707
+    },
+    {
+      "epoch": 0.10506889763779527,
+      "grad_norm": 0.8992338180541992,
+      "learning_rate": 0.00019465747295008013,
+      "loss": 3.2017,
+      "step": 1708
+    },
+    {
+      "epoch": 0.10513041338582677,
+      "grad_norm": 0.9514185786247253,
+      "learning_rate": 0.0001946512350909367,
+      "loss": 3.168,
+      "step": 1709
+    },
+    {
+      "epoch": 0.10519192913385826,
+      "grad_norm": 1.0187352895736694,
+      "learning_rate": 0.00019464499369236486,
+      "loss": 3.2999,
+      "step": 1710
+    },
+    {
+      "epoch": 0.10525344488188976,
+      "grad_norm": 0.9995301365852356,
+      "learning_rate": 0.00019463874875459803,
+      "loss": 3.0195,
+      "step": 1711
+    },
+    {
+      "epoch": 0.10531496062992125,
+      "grad_norm": 0.9492829442024231,
+      "learning_rate": 0.00019463250027786967,
+      "loss": 3.9931,
+      "step": 1712
+    },
+    {
+      "epoch": 0.10537647637795275,
+      "grad_norm": 1.0764844417572021,
+      "learning_rate": 0.00019462624826241355,
+      "loss": 2.87,
+      "step": 1713
+    },
+    {
+      "epoch": 0.10543799212598425,
+      "grad_norm": 1.213992953300476,
+      "learning_rate": 0.00019461999270846335,
+      "loss": 3.4166,
+      "step": 1714
+    },
+    {
+      "epoch": 0.10549950787401575,
+      "grad_norm": 1.03646719455719,
+      "learning_rate": 0.00019461373361625306,
+      "loss": 2.9662,
+      "step": 1715
+    },
+    {
+      "epoch": 0.10556102362204725,
+      "grad_norm": 1.04861319065094,
+      "learning_rate": 0.0001946074709860167,
+      "loss": 3.0677,
+      "step": 1716
+    },
+    {
+      "epoch": 0.10562253937007875,
+      "grad_norm": 0.900629997253418,
+      "learning_rate": 0.00019460120481798847,
+      "loss": 3.5091,
+      "step": 1717
+    },
+    {
+      "epoch": 0.10568405511811024,
+      "grad_norm": 1.3641128540039062,
+      "learning_rate": 0.0001945949351124027,
+      "loss": 2.8289,
+      "step": 1718
+    },
+    {
+      "epoch": 0.10574557086614174,
+      "grad_norm": 1.3361507654190063,
+      "learning_rate": 0.00019458866186949382,
+      "loss": 2.4248,
+      "step": 1719
+    },
+    {
+      "epoch": 0.10580708661417323,
+      "grad_norm": 0.84483402967453,
+      "learning_rate": 0.00019458238508949645,
+      "loss": 3.4238,
+      "step": 1720
+    },
+    {
+      "epoch": 0.10586860236220473,
+      "grad_norm": 1.3066824674606323,
+      "learning_rate": 0.00019457610477264527,
+      "loss": 2.8645,
+      "step": 1721
+    },
+    {
+      "epoch": 0.10593011811023623,
+      "grad_norm": 1.225714087486267,
+      "learning_rate": 0.00019456982091917513,
+      "loss": 2.9806,
+      "step": 1722
+    },
+    {
+      "epoch": 0.10599163385826772,
+      "grad_norm": 0.9972168207168579,
+      "learning_rate": 0.00019456353352932104,
+      "loss": 2.7965,
+      "step": 1723
+    },
+    {
+      "epoch": 0.10605314960629922,
+      "grad_norm": 1.390915036201477,
+      "learning_rate": 0.0001945572426033181,
+      "loss": 2.94,
+      "step": 1724
+    },
+    {
+      "epoch": 0.10611466535433071,
+      "grad_norm": 1.392124056816101,
+      "learning_rate": 0.00019455094814140153,
+      "loss": 2.6231,
+      "step": 1725
+    },
+    {
+      "epoch": 0.10617618110236221,
+      "grad_norm": 1.256435513496399,
+      "learning_rate": 0.00019454465014380676,
+      "loss": 2.9481,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1062376968503937,
+      "grad_norm": 0.97029048204422,
+      "learning_rate": 0.00019453834861076925,
+      "loss": 3.6915,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1062992125984252,
+      "grad_norm": 0.7937756776809692,
+      "learning_rate": 0.00019453204354252465,
+      "loss": 3.8029,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1063607283464567,
+      "grad_norm": 0.9084977507591248,
+      "learning_rate": 0.00019452573493930876,
+      "loss": 3.3589,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1064222440944882,
+      "grad_norm": 1.0813615322113037,
+      "learning_rate": 0.0001945194228013575,
+      "loss": 3.7171,
+      "step": 1730
+    },
+    {
+      "epoch": 0.10648375984251969,
+      "grad_norm": 0.8266887068748474,
+      "learning_rate": 0.00019451310712890685,
+      "loss": 3.4271,
+      "step": 1731
+    },
+    {
+      "epoch": 0.10654527559055119,
+      "grad_norm": 1.131083607673645,
+      "learning_rate": 0.000194506787922193,
+      "loss": 2.7078,
+      "step": 1732
+    },
+    {
+      "epoch": 0.10660679133858268,
+      "grad_norm": 0.8802652955055237,
+      "learning_rate": 0.0001945004651814523,
+      "loss": 2.9153,
+      "step": 1733
+    },
+    {
+      "epoch": 0.10666830708661418,
+      "grad_norm": 0.9421111941337585,
+      "learning_rate": 0.00019449413890692115,
+      "loss": 2.882,
+      "step": 1734
+    },
+    {
+      "epoch": 0.10672982283464567,
+      "grad_norm": 0.8748010396957397,
+      "learning_rate": 0.0001944878090988361,
+      "loss": 3.6763,
+      "step": 1735
+    },
+    {
+      "epoch": 0.10679133858267717,
+      "grad_norm": 0.9864575862884521,
+      "learning_rate": 0.00019448147575743387,
+      "loss": 4.329,
+      "step": 1736
+    },
+    {
+      "epoch": 0.10685285433070867,
+      "grad_norm": 1.439359426498413,
+      "learning_rate": 0.00019447513888295128,
+      "loss": 2.9092,
+      "step": 1737
+    },
+    {
+      "epoch": 0.10691437007874016,
+      "grad_norm": 1.0684739351272583,
+      "learning_rate": 0.00019446879847562532,
+      "loss": 2.8544,
+      "step": 1738
+    },
+    {
+      "epoch": 0.10697588582677166,
+      "grad_norm": 0.9625200629234314,
+      "learning_rate": 0.00019446245453569308,
+      "loss": 3.6116,
+      "step": 1739
+    },
+    {
+      "epoch": 0.10703740157480315,
+      "grad_norm": 2.280651569366455,
+      "learning_rate": 0.00019445610706339176,
+      "loss": 1.9588,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10709891732283465,
+      "grad_norm": 1.1945204734802246,
+      "learning_rate": 0.00019444975605895873,
+      "loss": 3.1934,
+      "step": 1741
+    },
+    {
+      "epoch": 0.10716043307086615,
+      "grad_norm": 1.2523274421691895,
+      "learning_rate": 0.00019444340152263153,
+      "loss": 2.6571,
+      "step": 1742
+    },
+    {
+      "epoch": 0.10722194881889764,
+      "grad_norm": 1.1880015134811401,
+      "learning_rate": 0.0001944370434546477,
+      "loss": 2.4056,
+      "step": 1743
+    },
+    {
+      "epoch": 0.10728346456692914,
+      "grad_norm": 0.9964622259140015,
+      "learning_rate": 0.00019443068185524506,
+      "loss": 3.0852,
+      "step": 1744
+    },
+    {
+      "epoch": 0.10734498031496063,
+      "grad_norm": 1.0852049589157104,
+      "learning_rate": 0.00019442431672466148,
+      "loss": 2.804,
+      "step": 1745
+    },
+    {
+      "epoch": 0.10740649606299213,
+      "grad_norm": 1.0638277530670166,
+      "learning_rate": 0.00019441794806313495,
+      "loss": 3.6221,
+      "step": 1746
+    },
+    {
+      "epoch": 0.10746801181102363,
+      "grad_norm": 1.1972006559371948,
+      "learning_rate": 0.00019441157587090372,
+      "loss": 3.0124,
+      "step": 1747
+    },
+    {
+      "epoch": 0.10752952755905512,
+      "grad_norm": 1.0422728061676025,
+      "learning_rate": 0.00019440520014820593,
+      "loss": 3.7905,
+      "step": 1748
+    },
+    {
+      "epoch": 0.10759104330708662,
+      "grad_norm": 1.0039513111114502,
+      "learning_rate": 0.0001943988208952801,
+      "loss": 2.6413,
+      "step": 1749
+    },
+    {
+      "epoch": 0.10765255905511811,
+      "grad_norm": 1.2659525871276855,
+      "learning_rate": 0.00019439243811236476,
+      "loss": 3.7679,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10771407480314961,
+      "grad_norm": 1.762182354927063,
+      "learning_rate": 0.00019438605179969857,
+      "loss": 2.3868,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1077755905511811,
+      "grad_norm": 1.3598774671554565,
+      "learning_rate": 0.00019437966195752035,
+      "loss": 3.2889,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1078371062992126,
+      "grad_norm": 1.3704978227615356,
+      "learning_rate": 0.00019437326858606907,
+      "loss": 2.3959,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1078986220472441,
+      "grad_norm": 1.1049766540527344,
+      "learning_rate": 0.00019436687168558376,
+      "loss": 2.7217,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1079601377952756,
+      "grad_norm": 1.0465492010116577,
+      "learning_rate": 0.00019436047125630368,
+      "loss": 3.1625,
+      "step": 1755
+    },
+    {
+      "epoch": 0.10802165354330709,
+      "grad_norm": 1.3700177669525146,
+      "learning_rate": 0.00019435406729846814,
+      "loss": 3.7369,
+      "step": 1756
+    },
+    {
+      "epoch": 0.10808316929133858,
+      "grad_norm": 0.9528889656066895,
+      "learning_rate": 0.0001943476598123166,
+      "loss": 2.568,
+      "step": 1757
+    },
+    {
+      "epoch": 0.10814468503937008,
+      "grad_norm": 1.1348025798797607,
+      "learning_rate": 0.0001943412487980887,
+      "loss": 2.9355,
+      "step": 1758
+    },
+    {
+      "epoch": 0.10820620078740158,
+      "grad_norm": 0.9928226470947266,
+      "learning_rate": 0.0001943348342560241,
+      "loss": 3.2097,
+      "step": 1759
+    },
+    {
+      "epoch": 0.10826771653543307,
+      "grad_norm": 1.0274388790130615,
+      "learning_rate": 0.00019432841618636278,
+      "loss": 3.3927,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10832923228346457,
+      "grad_norm": 1.0941137075424194,
+      "learning_rate": 0.00019432199458934468,
+      "loss": 2.8594,
+      "step": 1761
+    },
+    {
+      "epoch": 0.10839074803149606,
+      "grad_norm": 1.2455545663833618,
+      "learning_rate": 0.0001943155694652099,
+      "loss": 3.0901,
+      "step": 1762
+    },
+    {
+      "epoch": 0.10845226377952756,
+      "grad_norm": 1.415635347366333,
+      "learning_rate": 0.00019430914081419877,
+      "loss": 2.5559,
+      "step": 1763
+    },
+    {
+      "epoch": 0.10851377952755906,
+      "grad_norm": 0.8957375288009644,
+      "learning_rate": 0.00019430270863655166,
+      "loss": 3.8697,
+      "step": 1764
+    },
+    {
+      "epoch": 0.10857529527559055,
+      "grad_norm": 1.0922456979751587,
+      "learning_rate": 0.00019429627293250908,
+      "loss": 2.9707,
+      "step": 1765
+    },
+    {
+      "epoch": 0.10863681102362205,
+      "grad_norm": 1.2994139194488525,
+      "learning_rate": 0.00019428983370231172,
+      "loss": 3.1401,
+      "step": 1766
+    },
+    {
+      "epoch": 0.10869832677165354,
+      "grad_norm": 1.011724591255188,
+      "learning_rate": 0.00019428339094620034,
+      "loss": 3.1441,
+      "step": 1767
+    },
+    {
+      "epoch": 0.10875984251968504,
+      "grad_norm": 0.9900513887405396,
+      "learning_rate": 0.00019427694466441587,
+      "loss": 3.07,
+      "step": 1768
+    },
+    {
+      "epoch": 0.10882135826771654,
+      "grad_norm": 0.9562970995903015,
+      "learning_rate": 0.0001942704948571994,
+      "loss": 3.291,
+      "step": 1769
+    },
+    {
+      "epoch": 0.10888287401574803,
+      "grad_norm": 0.9604894518852234,
+      "learning_rate": 0.00019426404152479205,
+      "loss": 3.3372,
+      "step": 1770
+    },
+    {
+      "epoch": 0.10894438976377953,
+      "grad_norm": 0.9098861813545227,
+      "learning_rate": 0.0001942575846674352,
+      "loss": 3.0334,
+      "step": 1771
+    },
+    {
+      "epoch": 0.10900590551181102,
+      "grad_norm": 0.9318273067474365,
+      "learning_rate": 0.00019425112428537027,
+      "loss": 3.5359,
+      "step": 1772
+    },
+    {
+      "epoch": 0.10906742125984252,
+      "grad_norm": 1.01503324508667,
+      "learning_rate": 0.00019424466037883887,
+      "loss": 3.0018,
+      "step": 1773
+    },
+    {
+      "epoch": 0.10912893700787402,
+      "grad_norm": 0.9280456900596619,
+      "learning_rate": 0.00019423819294808267,
+      "loss": 3.1553,
+      "step": 1774
+    },
+    {
+      "epoch": 0.10919045275590551,
+      "grad_norm": 1.0694295167922974,
+      "learning_rate": 0.00019423172199334356,
+      "loss": 3.0465,
+      "step": 1775
+    },
+    {
+      "epoch": 0.10925196850393701,
+      "grad_norm": 0.8418107032775879,
+      "learning_rate": 0.0001942252475148635,
+      "loss": 2.9736,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1093134842519685,
+      "grad_norm": 1.2237215042114258,
+      "learning_rate": 0.00019421876951288458,
+      "loss": 3.2805,
+      "step": 1777
+    },
+    {
+      "epoch": 0.109375,
+      "grad_norm": 1.063139796257019,
+      "learning_rate": 0.00019421228798764906,
+      "loss": 2.724,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1094365157480315,
+      "grad_norm": 0.9280394911766052,
+      "learning_rate": 0.0001942058029393993,
+      "loss": 2.7929,
+      "step": 1779
+    },
+    {
+      "epoch": 0.10949803149606299,
+      "grad_norm": 1.2287083864212036,
+      "learning_rate": 0.00019419931436837783,
+      "loss": 3.3492,
+      "step": 1780
+    },
+    {
+      "epoch": 0.10955954724409449,
+      "grad_norm": 1.0722891092300415,
+      "learning_rate": 0.00019419282227482726,
+      "loss": 2.8415,
+      "step": 1781
+    },
+    {
+      "epoch": 0.10962106299212598,
+      "grad_norm": 1.0751558542251587,
+      "learning_rate": 0.0001941863266589904,
+      "loss": 2.9072,
+      "step": 1782
+    },
+    {
+      "epoch": 0.10968257874015748,
+      "grad_norm": 1.2583298683166504,
+      "learning_rate": 0.0001941798275211101,
+      "loss": 3.2026,
+      "step": 1783
+    },
+    {
+      "epoch": 0.10974409448818898,
+      "grad_norm": 0.9692583084106445,
+      "learning_rate": 0.00019417332486142944,
+      "loss": 3.2885,
+      "step": 1784
+    },
+    {
+      "epoch": 0.10980561023622047,
+      "grad_norm": 1.273679494857788,
+      "learning_rate": 0.00019416681868019152,
+      "loss": 3.194,
+      "step": 1785
+    },
+    {
+      "epoch": 0.10986712598425197,
+      "grad_norm": 1.147007942199707,
+      "learning_rate": 0.0001941603089776397,
+      "loss": 2.6257,
+      "step": 1786
+    },
+    {
+      "epoch": 0.10992864173228346,
+      "grad_norm": 1.042861819267273,
+      "learning_rate": 0.00019415379575401734,
+      "loss": 3.3032,
+      "step": 1787
+    },
+    {
+      "epoch": 0.10999015748031496,
+      "grad_norm": 0.9743854999542236,
+      "learning_rate": 0.00019414727900956806,
+      "loss": 3.2307,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11005167322834646,
+      "grad_norm": 1.0706549882888794,
+      "learning_rate": 0.0001941407587445355,
+      "loss": 2.9525,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11011318897637795,
+      "grad_norm": 0.9730565547943115,
+      "learning_rate": 0.00019413423495916353,
+      "loss": 3.0389,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11017470472440945,
+      "grad_norm": 0.8800128698348999,
+      "learning_rate": 0.00019412770765369608,
+      "loss": 2.95,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11023622047244094,
+      "grad_norm": 0.8032440543174744,
+      "learning_rate": 0.00019412117682837722,
+      "loss": 3.5403,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11029773622047244,
+      "grad_norm": 0.9636474251747131,
+      "learning_rate": 0.00019411464248345118,
+      "loss": 3.0564,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11035925196850394,
+      "grad_norm": 1.060186743736267,
+      "learning_rate": 0.00019410810461916234,
+      "loss": 3.207,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11042076771653543,
+      "grad_norm": 1.686438798904419,
+      "learning_rate": 0.0001941015632357551,
+      "loss": 3.5904,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11048228346456693,
+      "grad_norm": 1.1623353958129883,
+      "learning_rate": 0.00019409501833347417,
+      "loss": 3.4062,
+      "step": 1796
+    },
+    {
+      "epoch": 0.11054379921259842,
+      "grad_norm": 1.3794076442718506,
+      "learning_rate": 0.00019408846991256419,
+      "loss": 2.6702,
+      "step": 1797
+    },
+    {
+      "epoch": 0.11060531496062992,
+      "grad_norm": 0.9549499750137329,
+      "learning_rate": 0.00019408191797327013,
+      "loss": 3.4371,
+      "step": 1798
+    },
+    {
+      "epoch": 0.11066683070866142,
+      "grad_norm": 1.039560317993164,
+      "learning_rate": 0.00019407536251583694,
+      "loss": 2.975,
+      "step": 1799
+    },
+    {
+      "epoch": 0.11072834645669291,
+      "grad_norm": 1.0493844747543335,
+      "learning_rate": 0.00019406880354050978,
+      "loss": 2.9151,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1107898622047244,
+      "grad_norm": 1.222562551498413,
+      "learning_rate": 0.0001940622410475339,
+      "loss": 3.2494,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1108513779527559,
+      "grad_norm": 1.3852157592773438,
+      "learning_rate": 0.0001940556750371547,
+      "loss": 3.126,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1109128937007874,
+      "grad_norm": 1.1538420915603638,
+      "learning_rate": 0.0001940491055096177,
+      "loss": 2.2998,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1109744094488189,
+      "grad_norm": 0.9417049288749695,
+      "learning_rate": 0.00019404253246516865,
+      "loss": 2.8875,
+      "step": 1804
+    },
+    {
+      "epoch": 0.11103592519685039,
+      "grad_norm": 1.0907907485961914,
+      "learning_rate": 0.00019403595590405327,
+      "loss": 2.9281,
+      "step": 1805
+    },
+    {
+      "epoch": 0.11109744094488189,
+      "grad_norm": 1.018187165260315,
+      "learning_rate": 0.00019402937582651744,
+      "loss": 2.6168,
+      "step": 1806
+    },
+    {
+      "epoch": 0.11115895669291338,
+      "grad_norm": 1.3872551918029785,
+      "learning_rate": 0.00019402279223280732,
+      "loss": 2.2898,
+      "step": 1807
+    },
+    {
+      "epoch": 0.11122047244094488,
+      "grad_norm": 1.008433222770691,
+      "learning_rate": 0.00019401620512316905,
+      "loss": 3.8154,
+      "step": 1808
+    },
+    {
+      "epoch": 0.11128198818897637,
+      "grad_norm": 1.228466272354126,
+      "learning_rate": 0.00019400961449784896,
+      "loss": 2.8087,
+      "step": 1809
+    },
+    {
+      "epoch": 0.11134350393700787,
+      "grad_norm": 0.9024825096130371,
+      "learning_rate": 0.0001940030203570935,
+      "loss": 2.5619,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11140501968503937,
+      "grad_norm": 1.064486026763916,
+      "learning_rate": 0.00019399642270114925,
+      "loss": 3.4263,
+      "step": 1811
+    },
+    {
+      "epoch": 0.11146653543307086,
+      "grad_norm": 0.9295080900192261,
+      "learning_rate": 0.00019398982153026291,
+      "loss": 2.8931,
+      "step": 1812
+    },
+    {
+      "epoch": 0.11152805118110236,
+      "grad_norm": 1.1003059148788452,
+      "learning_rate": 0.00019398321684468137,
+      "loss": 2.5103,
+      "step": 1813
+    },
+    {
+      "epoch": 0.11158956692913385,
+      "grad_norm": 1.1258125305175781,
+      "learning_rate": 0.00019397660864465158,
+      "loss": 3.1056,
+      "step": 1814
+    },
+    {
+      "epoch": 0.11165108267716535,
+      "grad_norm": 0.8974843621253967,
+      "learning_rate": 0.00019396999693042065,
+      "loss": 3.1396,
+      "step": 1815
+    },
+    {
+      "epoch": 0.11171259842519685,
+      "grad_norm": 1.0095648765563965,
+      "learning_rate": 0.00019396338170223586,
+      "loss": 3.216,
+      "step": 1816
+    },
+    {
+      "epoch": 0.11177411417322834,
+      "grad_norm": 1.057336688041687,
+      "learning_rate": 0.0001939567629603445,
+      "loss": 2.9565,
+      "step": 1817
+    },
+    {
+      "epoch": 0.11183562992125984,
+      "grad_norm": 1.1198409795761108,
+      "learning_rate": 0.00019395014070499412,
+      "loss": 3.5234,
+      "step": 1818
+    },
+    {
+      "epoch": 0.11189714566929133,
+      "grad_norm": 0.9906701445579529,
+      "learning_rate": 0.00019394351493643238,
+      "loss": 2.6891,
+      "step": 1819
+    },
+    {
+      "epoch": 0.11195866141732283,
+      "grad_norm": 1.815333604812622,
+      "learning_rate": 0.00019393688565490705,
+      "loss": 2.5404,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11202017716535433,
+      "grad_norm": 1.1773436069488525,
+      "learning_rate": 0.00019393025286066595,
+      "loss": 3.154,
+      "step": 1821
+    },
+    {
+      "epoch": 0.11208169291338582,
+      "grad_norm": 1.18867027759552,
+      "learning_rate": 0.0001939236165539572,
+      "loss": 2.9258,
+      "step": 1822
+    },
+    {
+      "epoch": 0.11214320866141732,
+      "grad_norm": 1.3936090469360352,
+      "learning_rate": 0.0001939169767350289,
+      "loss": 2.5971,
+      "step": 1823
+    },
+    {
+      "epoch": 0.11220472440944881,
+      "grad_norm": 1.1651310920715332,
+      "learning_rate": 0.0001939103334041294,
+      "loss": 3.4599,
+      "step": 1824
+    },
+    {
+      "epoch": 0.11226624015748031,
+      "grad_norm": 1.1387213468551636,
+      "learning_rate": 0.00019390368656150707,
+      "loss": 2.4935,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1123277559055118,
+      "grad_norm": 1.1077392101287842,
+      "learning_rate": 0.0001938970362074105,
+      "loss": 3.5107,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1123892716535433,
+      "grad_norm": 1.1750246286392212,
+      "learning_rate": 0.00019389038234208833,
+      "loss": 2.9477,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1124507874015748,
+      "grad_norm": 1.0942513942718506,
+      "learning_rate": 0.00019388372496578939,
+      "loss": 3.3621,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1125123031496063,
+      "grad_norm": 1.245656967163086,
+      "learning_rate": 0.00019387706407876268,
+      "loss": 3.2021,
+      "step": 1829
+    },
+    {
+      "epoch": 0.11257381889763779,
+      "grad_norm": 0.9476128220558167,
+      "learning_rate": 0.00019387039968125723,
+      "loss": 2.883,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11263533464566929,
+      "grad_norm": 0.9918729066848755,
+      "learning_rate": 0.0001938637317735223,
+      "loss": 3.0262,
+      "step": 1831
+    },
+    {
+      "epoch": 0.11269685039370078,
+      "grad_norm": 1.037253737449646,
+      "learning_rate": 0.00019385706035580718,
+      "loss": 3.1729,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11275836614173228,
+      "grad_norm": 1.2221964597702026,
+      "learning_rate": 0.00019385038542836137,
+      "loss": 2.5179,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11281988188976377,
+      "grad_norm": 0.9292991161346436,
+      "learning_rate": 0.00019384370699143447,
+      "loss": 2.9621,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11288139763779527,
+      "grad_norm": 1.2171398401260376,
+      "learning_rate": 0.0001938370250452762,
+      "loss": 2.8357,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11294291338582677,
+      "grad_norm": 1.1688063144683838,
+      "learning_rate": 0.00019383033959013645,
+      "loss": 2.7733,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11300442913385826,
+      "grad_norm": 1.1196861267089844,
+      "learning_rate": 0.0001938236506262652,
+      "loss": 3.0078,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11306594488188976,
+      "grad_norm": 1.190688967704773,
+      "learning_rate": 0.0001938169581539126,
+      "loss": 3.3646,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11312746062992125,
+      "grad_norm": 1.1056948900222778,
+      "learning_rate": 0.0001938102621733289,
+      "loss": 3.2136,
+      "step": 1839
+    },
+    {
+      "epoch": 0.11318897637795275,
+      "grad_norm": 1.331734538078308,
+      "learning_rate": 0.00019380356268476452,
+      "loss": 3.1795,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11325049212598425,
+      "grad_norm": 0.9417758584022522,
+      "learning_rate": 0.00019379685968846992,
+      "loss": 3.2332,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11331200787401575,
+      "grad_norm": 0.9451195597648621,
+      "learning_rate": 0.00019379015318469583,
+      "loss": 3.101,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11337352362204725,
+      "grad_norm": 1.0004974603652954,
+      "learning_rate": 0.000193783443173693,
+      "loss": 2.8831,
+      "step": 1843
+    },
+    {
+      "epoch": 0.11343503937007875,
+      "grad_norm": 0.954467236995697,
+      "learning_rate": 0.00019377672965571228,
+      "loss": 3.2737,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11349655511811024,
+      "grad_norm": 1.2548714876174927,
+      "learning_rate": 0.00019377001263100487,
+      "loss": 3.0435,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11355807086614174,
+      "grad_norm": 1.0386886596679688,
+      "learning_rate": 0.0001937632920998218,
+      "loss": 2.5878,
+      "step": 1846
+    },
+    {
+      "epoch": 0.11361958661417323,
+      "grad_norm": 1.1594774723052979,
+      "learning_rate": 0.0001937565680624145,
+      "loss": 3.0068,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11368110236220473,
+      "grad_norm": 1.0956149101257324,
+      "learning_rate": 0.00019374984051903431,
+      "loss": 3.4821,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11374261811023623,
+      "grad_norm": 1.1208339929580688,
+      "learning_rate": 0.00019374310946993287,
+      "loss": 2.8652,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11380413385826772,
+      "grad_norm": 1.056100845336914,
+      "learning_rate": 0.00019373637491536185,
+      "loss": 3.495,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11386564960629922,
+      "grad_norm": 0.9148539304733276,
+      "learning_rate": 0.0001937296368555731,
+      "loss": 3.0664,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11392716535433071,
+      "grad_norm": 0.7838106155395508,
+      "learning_rate": 0.0001937228952908186,
+      "loss": 3.0263,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11398868110236221,
+      "grad_norm": 1.2357759475708008,
+      "learning_rate": 0.0001937161502213504,
+      "loss": 3.2273,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1140501968503937,
+      "grad_norm": 0.8642683625221252,
+      "learning_rate": 0.00019370940164742078,
+      "loss": 3.1885,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1141117125984252,
+      "grad_norm": 1.0933315753936768,
+      "learning_rate": 0.00019370264956928207,
+      "loss": 2.853,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1141732283464567,
+      "grad_norm": 0.9847070574760437,
+      "learning_rate": 0.00019369589398718676,
+      "loss": 2.4578,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1142347440944882,
+      "grad_norm": 1.2235275506973267,
+      "learning_rate": 0.0001936891349013875,
+      "loss": 2.9122,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11429625984251969,
+      "grad_norm": 1.1270467042922974,
+      "learning_rate": 0.00019368237231213696,
+      "loss": 2.6848,
+      "step": 1858
+    },
+    {
+      "epoch": 0.11435777559055119,
+      "grad_norm": 1.1999391317367554,
+      "learning_rate": 0.00019367560621968814,
+      "loss": 3.5604,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11441929133858268,
+      "grad_norm": 0.8759294152259827,
+      "learning_rate": 0.000193668836624294,
+      "loss": 2.2802,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11448080708661418,
+      "grad_norm": 0.9458711743354797,
+      "learning_rate": 0.00019366206352620764,
+      "loss": 3.1399,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11454232283464567,
+      "grad_norm": 0.9431505799293518,
+      "learning_rate": 0.00019365528692568242,
+      "loss": 3.203,
+      "step": 1862
+    },
+    {
+      "epoch": 0.11460383858267717,
+      "grad_norm": 1.2803919315338135,
+      "learning_rate": 0.00019364850682297168,
+      "loss": 3.2883,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11466535433070867,
+      "grad_norm": 1.0069122314453125,
+      "learning_rate": 0.00019364172321832899,
+      "loss": 2.8669,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11472687007874016,
+      "grad_norm": 1.4869080781936646,
+      "learning_rate": 0.00019363493611200802,
+      "loss": 3.0662,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11478838582677166,
+      "grad_norm": 0.9315390586853027,
+      "learning_rate": 0.00019362814550426253,
+      "loss": 3.8392,
+      "step": 1866
+    },
+    {
+      "epoch": 0.11484990157480315,
+      "grad_norm": 0.8289329409599304,
+      "learning_rate": 0.00019362135139534652,
+      "loss": 3.8234,
+      "step": 1867
+    },
+    {
+      "epoch": 0.11491141732283465,
+      "grad_norm": 1.0998977422714233,
+      "learning_rate": 0.000193614553785514,
+      "loss": 3.414,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11497293307086615,
+      "grad_norm": 0.8537095189094543,
+      "learning_rate": 0.00019360775267501917,
+      "loss": 3.8041,
+      "step": 1869
+    },
+    {
+      "epoch": 0.11503444881889764,
+      "grad_norm": 0.9429873824119568,
+      "learning_rate": 0.00019360094806411638,
+      "loss": 3.0808,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11509596456692914,
+      "grad_norm": 1.2331935167312622,
+      "learning_rate": 0.00019359413995306005,
+      "loss": 1.9134,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11515748031496063,
+      "grad_norm": 0.9669841527938843,
+      "learning_rate": 0.0001935873283421048,
+      "loss": 3.4024,
+      "step": 1872
+    },
+    {
+      "epoch": 0.11521899606299213,
+      "grad_norm": 1.2340986728668213,
+      "learning_rate": 0.0001935805132315053,
+      "loss": 2.0516,
+      "step": 1873
+    },
+    {
+      "epoch": 0.11528051181102363,
+      "grad_norm": 0.9023285508155823,
+      "learning_rate": 0.00019357369462151645,
+      "loss": 3.1417,
+      "step": 1874
+    },
+    {
+      "epoch": 0.11534202755905512,
+      "grad_norm": 1.0570188760757446,
+      "learning_rate": 0.00019356687251239317,
+      "loss": 2.6255,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11540354330708662,
+      "grad_norm": 1.3824774026870728,
+      "learning_rate": 0.00019356004690439065,
+      "loss": 2.3289,
+      "step": 1876
+    },
+    {
+      "epoch": 0.11546505905511811,
+      "grad_norm": 1.2685729265213013,
+      "learning_rate": 0.00019355321779776402,
+      "loss": 1.8533,
+      "step": 1877
+    },
+    {
+      "epoch": 0.11552657480314961,
+      "grad_norm": 1.1483112573623657,
+      "learning_rate": 0.00019354638519276875,
+      "loss": 3.1437,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1155880905511811,
+      "grad_norm": 0.9710890650749207,
+      "learning_rate": 0.0001935395490896603,
+      "loss": 2.8468,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1156496062992126,
+      "grad_norm": 1.075580358505249,
+      "learning_rate": 0.0001935327094886943,
+      "loss": 2.7485,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1157111220472441,
+      "grad_norm": 1.1986594200134277,
+      "learning_rate": 0.00019352586639012653,
+      "loss": 2.5096,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1157726377952756,
+      "grad_norm": 1.0973596572875977,
+      "learning_rate": 0.00019351901979421286,
+      "loss": 3.1665,
+      "step": 1882
+    },
+    {
+      "epoch": 0.11583415354330709,
+      "grad_norm": 1.0129361152648926,
+      "learning_rate": 0.00019351216970120932,
+      "loss": 3.3586,
+      "step": 1883
+    },
+    {
+      "epoch": 0.11589566929133858,
+      "grad_norm": 0.9978131651878357,
+      "learning_rate": 0.0001935053161113721,
+      "loss": 3.1051,
+      "step": 1884
+    },
+    {
+      "epoch": 0.11595718503937008,
+      "grad_norm": 0.8945294618606567,
+      "learning_rate": 0.00019349845902495743,
+      "loss": 3.2763,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11601870078740158,
+      "grad_norm": 1.0221529006958008,
+      "learning_rate": 0.00019349159844222173,
+      "loss": 2.6348,
+      "step": 1886
+    },
+    {
+      "epoch": 0.11608021653543307,
+      "grad_norm": 1.092896580696106,
+      "learning_rate": 0.00019348473436342161,
+      "loss": 3.1745,
+      "step": 1887
+    },
+    {
+      "epoch": 0.11614173228346457,
+      "grad_norm": 1.030393123626709,
+      "learning_rate": 0.0001934778667888137,
+      "loss": 3.6818,
+      "step": 1888
+    },
+    {
+      "epoch": 0.11620324803149606,
+      "grad_norm": 1.2027442455291748,
+      "learning_rate": 0.0001934709957186548,
+      "loss": 2.7114,
+      "step": 1889
+    },
+    {
+      "epoch": 0.11626476377952756,
+      "grad_norm": 1.4980840682983398,
+      "learning_rate": 0.0001934641211532019,
+      "loss": 2.981,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11632627952755906,
+      "grad_norm": 0.8777542114257812,
+      "learning_rate": 0.00019345724309271204,
+      "loss": 2.8134,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11638779527559055,
+      "grad_norm": 0.8980687856674194,
+      "learning_rate": 0.00019345036153744244,
+      "loss": 3.6804,
+      "step": 1892
+    },
+    {
+      "epoch": 0.11644931102362205,
+      "grad_norm": 0.9111537337303162,
+      "learning_rate": 0.00019344347648765035,
+      "loss": 3.4572,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11651082677165354,
+      "grad_norm": 1.2011817693710327,
+      "learning_rate": 0.00019343658794359332,
+      "loss": 2.7335,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11657234251968504,
+      "grad_norm": 1.2333368062973022,
+      "learning_rate": 0.00019342969590552893,
+      "loss": 2.913,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11663385826771654,
+      "grad_norm": 1.009445071220398,
+      "learning_rate": 0.0001934228003737149,
+      "loss": 2.9155,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11669537401574803,
+      "grad_norm": 1.2909005880355835,
+      "learning_rate": 0.00019341590134840905,
+      "loss": 2.901,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11675688976377953,
+      "grad_norm": 1.5580658912658691,
+      "learning_rate": 0.0001934089988298694,
+      "loss": 2.3302,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11681840551181102,
+      "grad_norm": 0.8886326551437378,
+      "learning_rate": 0.00019340209281835407,
+      "loss": 2.0511,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11687992125984252,
+      "grad_norm": 1.1103955507278442,
+      "learning_rate": 0.00019339518331412125,
+      "loss": 3.2081,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11694143700787402,
+      "grad_norm": 1.0729321241378784,
+      "learning_rate": 0.00019338827031742942,
+      "loss": 2.9555,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11700295275590551,
+      "grad_norm": 1.2396705150604248,
+      "learning_rate": 0.00019338135382853698,
+      "loss": 3.6013,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11706446850393701,
+      "grad_norm": 0.8740068674087524,
+      "learning_rate": 0.00019337443384770265,
+      "loss": 3.1413,
+      "step": 1903
+    },
+    {
+      "epoch": 0.1171259842519685,
+      "grad_norm": 0.9985756874084473,
+      "learning_rate": 0.00019336751037518513,
+      "loss": 3.4027,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1171875,
+      "grad_norm": 1.119163155555725,
+      "learning_rate": 0.0001933605834112434,
+      "loss": 3.2367,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1172490157480315,
+      "grad_norm": 0.8043966889381409,
+      "learning_rate": 0.0001933536529561364,
+      "loss": 3.2738,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11731053149606299,
+      "grad_norm": 1.037419080734253,
+      "learning_rate": 0.00019334671901012333,
+      "loss": 3.3971,
+      "step": 1907
+    },
+    {
+      "epoch": 0.11737204724409449,
+      "grad_norm": 0.9374710917472839,
+      "learning_rate": 0.0001933397815734635,
+      "loss": 3.8379,
+      "step": 1908
+    },
+    {
+      "epoch": 0.11743356299212598,
+      "grad_norm": 0.9358224272727966,
+      "learning_rate": 0.00019333284064641633,
+      "loss": 2.8599,
+      "step": 1909
+    },
+    {
+      "epoch": 0.11749507874015748,
+      "grad_norm": 0.9508259892463684,
+      "learning_rate": 0.00019332589622924134,
+      "loss": 3.1503,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11755659448818898,
+      "grad_norm": 0.9231706857681274,
+      "learning_rate": 0.00019331894832219827,
+      "loss": 2.9855,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11761811023622047,
+      "grad_norm": 0.9757208824157715,
+      "learning_rate": 0.00019331199692554685,
+      "loss": 2.8431,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11767962598425197,
+      "grad_norm": 1.1157352924346924,
+      "learning_rate": 0.00019330504203954706,
+      "loss": 2.5955,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11774114173228346,
+      "grad_norm": 1.2429035902023315,
+      "learning_rate": 0.00019329808366445902,
+      "loss": 2.5485,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11780265748031496,
+      "grad_norm": 1.2138863801956177,
+      "learning_rate": 0.00019329112180054285,
+      "loss": 3.358,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11786417322834646,
+      "grad_norm": 1.0206093788146973,
+      "learning_rate": 0.00019328415644805896,
+      "loss": 3.5534,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11792568897637795,
+      "grad_norm": 1.2162081003189087,
+      "learning_rate": 0.00019327718760726777,
+      "loss": 2.6049,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11798720472440945,
+      "grad_norm": 1.3472238779067993,
+      "learning_rate": 0.0001932702152784299,
+      "loss": 2.5852,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11804872047244094,
+      "grad_norm": 1.2386008501052856,
+      "learning_rate": 0.0001932632394618061,
+      "loss": 3.2353,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11811023622047244,
+      "grad_norm": 1.152139663696289,
+      "learning_rate": 0.00019325626015765713,
+      "loss": 2.9827,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11817175196850394,
+      "grad_norm": 0.9423478245735168,
+      "learning_rate": 0.0001932492773662441,
+      "loss": 3.3842,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11823326771653543,
+      "grad_norm": 0.9991691708564758,
+      "learning_rate": 0.00019324229108782804,
+      "loss": 3.206,
+      "step": 1922
+    },
+    {
+      "epoch": 0.11829478346456693,
+      "grad_norm": 1.1851075887680054,
+      "learning_rate": 0.0001932353013226702,
+      "loss": 3.4302,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11835629921259842,
+      "grad_norm": 1.0934733152389526,
+      "learning_rate": 0.00019322830807103202,
+      "loss": 3.1216,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11841781496062992,
+      "grad_norm": 0.8792948722839355,
+      "learning_rate": 0.000193221311333175,
+      "loss": 2.9769,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11847933070866142,
+      "grad_norm": 1.1932494640350342,
+      "learning_rate": 0.0001932143111093607,
+      "loss": 2.5325,
+      "step": 1926
+    },
+    {
+      "epoch": 0.11854084645669291,
+      "grad_norm": 1.098118782043457,
+      "learning_rate": 0.00019320730739985094,
+      "loss": 3.2966,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1186023622047244,
+      "grad_norm": 1.9875314235687256,
+      "learning_rate": 0.00019320030020490765,
+      "loss": 2.2994,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1186638779527559,
+      "grad_norm": 1.1281403303146362,
+      "learning_rate": 0.00019319328952479284,
+      "loss": 3.3972,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1187253937007874,
+      "grad_norm": 1.0175753831863403,
+      "learning_rate": 0.00019318627535976863,
+      "loss": 2.8879,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1187869094488189,
+      "grad_norm": 1.117751955986023,
+      "learning_rate": 0.00019317925771009738,
+      "loss": 3.1918,
+      "step": 1931
+    },
+    {
+      "epoch": 0.11884842519685039,
+      "grad_norm": 4.968902587890625,
+      "learning_rate": 0.00019317223657604144,
+      "loss": 3.1994,
+      "step": 1932
+    },
+    {
+      "epoch": 0.11890994094488189,
+      "grad_norm": 0.9675689935684204,
+      "learning_rate": 0.00019316521195786342,
+      "loss": 3.32,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11897145669291338,
+      "grad_norm": 1.145846962928772,
+      "learning_rate": 0.00019315818385582596,
+      "loss": 2.9063,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11903297244094488,
+      "grad_norm": 0.9326683282852173,
+      "learning_rate": 0.00019315115227019187,
+      "loss": 3.0531,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11909448818897637,
+      "grad_norm": 0.8474321961402893,
+      "learning_rate": 0.00019314411720122414,
+      "loss": 3.2123,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11915600393700787,
+      "grad_norm": 1.0700311660766602,
+      "learning_rate": 0.00019313707864918581,
+      "loss": 3.3403,
+      "step": 1937
+    },
+    {
+      "epoch": 0.11921751968503937,
+      "grad_norm": 0.9991259574890137,
+      "learning_rate": 0.00019313003661434008,
+      "loss": 3.0001,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11927903543307086,
+      "grad_norm": 1.024038553237915,
+      "learning_rate": 0.0001931229910969503,
+      "loss": 2.8374,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11934055118110236,
+      "grad_norm": 1.1284666061401367,
+      "learning_rate": 0.0001931159420972799,
+      "loss": 3.3028,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11940206692913385,
+      "grad_norm": 1.13485848903656,
+      "learning_rate": 0.0001931088896155925,
+      "loss": 2.9937,
+      "step": 1941
+    },
+    {
+      "epoch": 0.11946358267716535,
+      "grad_norm": 1.1891289949417114,
+      "learning_rate": 0.0001931018336521518,
+      "loss": 2.7371,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11952509842519685,
+      "grad_norm": 1.0075335502624512,
+      "learning_rate": 0.0001930947742072217,
+      "loss": 2.9744,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11958661417322834,
+      "grad_norm": 1.1116664409637451,
+      "learning_rate": 0.00019308771128106615,
+      "loss": 3.5589,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11964812992125984,
+      "grad_norm": 1.012611985206604,
+      "learning_rate": 0.00019308064487394929,
+      "loss": 3.0197,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11970964566929133,
+      "grad_norm": 0.8892108201980591,
+      "learning_rate": 0.0001930735749861353,
+      "loss": 2.741,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11977116141732283,
+      "grad_norm": 1.0437053442001343,
+      "learning_rate": 0.00019306650161788862,
+      "loss": 3.2356,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11983267716535433,
+      "grad_norm": 1.3619134426116943,
+      "learning_rate": 0.00019305942476947373,
+      "loss": 3.0375,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11989419291338582,
+      "grad_norm": 1.0049139261245728,
+      "learning_rate": 0.00019305234444115527,
+      "loss": 3.4435,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11995570866141732,
+      "grad_norm": 1.030515432357788,
+      "learning_rate": 0.000193045260633198,
+      "loss": 2.0563,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12001722440944881,
+      "grad_norm": 0.966373085975647,
+      "learning_rate": 0.00019303817334586683,
+      "loss": 2.7599,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12007874015748031,
+      "grad_norm": 1.0807048082351685,
+      "learning_rate": 0.00019303108257942675,
+      "loss": 2.8119,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1201402559055118,
+      "grad_norm": 1.0771872997283936,
+      "learning_rate": 0.00019302398833414292,
+      "loss": 3.476,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1202017716535433,
+      "grad_norm": 1.080277681350708,
+      "learning_rate": 0.00019301689061028068,
+      "loss": 3.1659,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1202632874015748,
+      "grad_norm": 1.2034735679626465,
+      "learning_rate": 0.0001930097894081054,
+      "loss": 3.2906,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1203248031496063,
+      "grad_norm": 1.1539957523345947,
+      "learning_rate": 0.00019300268472788264,
+      "loss": 3.4053,
+      "step": 1956
+    },
+    {
+      "epoch": 0.12038631889763779,
+      "grad_norm": 1.1620454788208008,
+      "learning_rate": 0.00019299557656987805,
+      "loss": 3.5379,
+      "step": 1957
+    },
+    {
+      "epoch": 0.12044783464566929,
+      "grad_norm": 1.1060287952423096,
+      "learning_rate": 0.00019298846493435746,
+      "loss": 3.5576,
+      "step": 1958
+    },
+    {
+      "epoch": 0.12050935039370078,
+      "grad_norm": 0.9905508756637573,
+      "learning_rate": 0.0001929813498215868,
+      "loss": 2.5405,
+      "step": 1959
+    },
+    {
+      "epoch": 0.12057086614173228,
+      "grad_norm": 1.0543802976608276,
+      "learning_rate": 0.00019297423123183213,
+      "loss": 2.8233,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12063238188976377,
+      "grad_norm": 1.1114733219146729,
+      "learning_rate": 0.00019296710916535966,
+      "loss": 2.9207,
+      "step": 1961
+    },
+    {
+      "epoch": 0.12069389763779527,
+      "grad_norm": 0.9835840463638306,
+      "learning_rate": 0.0001929599836224357,
+      "loss": 3.3403,
+      "step": 1962
+    },
+    {
+      "epoch": 0.12075541338582677,
+      "grad_norm": 1.4010146856307983,
+      "learning_rate": 0.0001929528546033267,
+      "loss": 2.5493,
+      "step": 1963
+    },
+    {
+      "epoch": 0.12081692913385826,
+      "grad_norm": 1.1950501203536987,
+      "learning_rate": 0.0001929457221082993,
+      "loss": 3.3134,
+      "step": 1964
+    },
+    {
+      "epoch": 0.12087844488188976,
+      "grad_norm": 1.1537760496139526,
+      "learning_rate": 0.00019293858613762013,
+      "loss": 3.1727,
+      "step": 1965
+    },
+    {
+      "epoch": 0.12093996062992125,
+      "grad_norm": 1.004812479019165,
+      "learning_rate": 0.00019293144669155609,
+      "loss": 2.7182,
+      "step": 1966
+    },
+    {
+      "epoch": 0.12100147637795275,
+      "grad_norm": 0.9878165125846863,
+      "learning_rate": 0.00019292430377037417,
+      "loss": 2.6199,
+      "step": 1967
+    },
+    {
+      "epoch": 0.12106299212598425,
+      "grad_norm": 1.1104050874710083,
+      "learning_rate": 0.0001929171573743414,
+      "loss": 3.4746,
+      "step": 1968
+    },
+    {
+      "epoch": 0.12112450787401575,
+      "grad_norm": 1.1120890378952026,
+      "learning_rate": 0.00019291000750372507,
+      "loss": 3.0427,
+      "step": 1969
+    },
+    {
+      "epoch": 0.12118602362204725,
+      "grad_norm": 1.1429128646850586,
+      "learning_rate": 0.00019290285415879258,
+      "loss": 2.3224,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12124753937007875,
+      "grad_norm": 0.9408043622970581,
+      "learning_rate": 0.00019289569733981136,
+      "loss": 3.2631,
+      "step": 1971
+    },
+    {
+      "epoch": 0.12130905511811024,
+      "grad_norm": 0.961268961429596,
+      "learning_rate": 0.00019288853704704906,
+      "loss": 3.0818,
+      "step": 1972
+    },
+    {
+      "epoch": 0.12137057086614174,
+      "grad_norm": 1.0820953845977783,
+      "learning_rate": 0.00019288137328077345,
+      "loss": 2.2901,
+      "step": 1973
+    },
+    {
+      "epoch": 0.12143208661417323,
+      "grad_norm": 0.8896895051002502,
+      "learning_rate": 0.00019287420604125236,
+      "loss": 3.4551,
+      "step": 1974
+    },
+    {
+      "epoch": 0.12149360236220473,
+      "grad_norm": 1.3226232528686523,
+      "learning_rate": 0.0001928670353287539,
+      "loss": 3.6938,
+      "step": 1975
+    },
+    {
+      "epoch": 0.12155511811023623,
+      "grad_norm": 0.9423373937606812,
+      "learning_rate": 0.00019285986114354612,
+      "loss": 2.9898,
+      "step": 1976
+    },
+    {
+      "epoch": 0.12161663385826772,
+      "grad_norm": 1.5197635889053345,
+      "learning_rate": 0.00019285268348589736,
+      "loss": 2.2062,
+      "step": 1977
+    },
+    {
+      "epoch": 0.12167814960629922,
+      "grad_norm": 1.4292954206466675,
+      "learning_rate": 0.00019284550235607598,
+      "loss": 3.0772,
+      "step": 1978
+    },
+    {
+      "epoch": 0.12173966535433071,
+      "grad_norm": 1.1244150400161743,
+      "learning_rate": 0.00019283831775435053,
+      "loss": 2.9702,
+      "step": 1979
+    },
+    {
+      "epoch": 0.12180118110236221,
+      "grad_norm": 1.2020808458328247,
+      "learning_rate": 0.00019283112968098969,
+      "loss": 2.8698,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1218626968503937,
+      "grad_norm": 1.0522679090499878,
+      "learning_rate": 0.00019282393813626223,
+      "loss": 2.9079,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1219242125984252,
+      "grad_norm": 1.6054037809371948,
+      "learning_rate": 0.0001928167431204371,
+      "loss": 2.8356,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1219857283464567,
+      "grad_norm": 0.8645476698875427,
+      "learning_rate": 0.00019280954463378333,
+      "loss": 3.0112,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1220472440944882,
+      "grad_norm": 1.017970323562622,
+      "learning_rate": 0.00019280234267657012,
+      "loss": 2.9778,
+      "step": 1984
+    },
+    {
+      "epoch": 0.12210875984251969,
+      "grad_norm": 1.374708652496338,
+      "learning_rate": 0.00019279513724906674,
+      "loss": 3.141,
+      "step": 1985
+    },
+    {
+      "epoch": 0.12217027559055119,
+      "grad_norm": 1.1559653282165527,
+      "learning_rate": 0.0001927879283515427,
+      "loss": 2.7693,
+      "step": 1986
+    },
+    {
+      "epoch": 0.12223179133858268,
+      "grad_norm": 0.9424875378608704,
+      "learning_rate": 0.00019278071598426751,
+      "loss": 2.975,
+      "step": 1987
+    },
+    {
+      "epoch": 0.12229330708661418,
+      "grad_norm": 1.1510764360427856,
+      "learning_rate": 0.00019277350014751092,
+      "loss": 1.9126,
+      "step": 1988
+    },
+    {
+      "epoch": 0.12235482283464567,
+      "grad_norm": 1.071962833404541,
+      "learning_rate": 0.00019276628084154273,
+      "loss": 3.0347,
+      "step": 1989
+    },
+    {
+      "epoch": 0.12241633858267717,
+      "grad_norm": 1.1727583408355713,
+      "learning_rate": 0.00019275905806663294,
+      "loss": 2.9212,
+      "step": 1990
+    },
+    {
+      "epoch": 0.12247785433070867,
+      "grad_norm": 0.8645524382591248,
+      "learning_rate": 0.0001927518318230516,
+      "loss": 1.8655,
+      "step": 1991
+    },
+    {
+      "epoch": 0.12253937007874016,
+      "grad_norm": 1.3134901523590088,
+      "learning_rate": 0.00019274460211106894,
+      "loss": 3.5627,
+      "step": 1992
+    },
+    {
+      "epoch": 0.12260088582677166,
+      "grad_norm": 1.3302184343338013,
+      "learning_rate": 0.00019273736893095537,
+      "loss": 2.6431,
+      "step": 1993
+    },
+    {
+      "epoch": 0.12266240157480315,
+      "grad_norm": 0.9993799328804016,
+      "learning_rate": 0.00019273013228298126,
+      "loss": 2.9495,
+      "step": 1994
+    },
+    {
+      "epoch": 0.12272391732283465,
+      "grad_norm": 0.9303006529808044,
+      "learning_rate": 0.0001927228921674173,
+      "loss": 3.0721,
+      "step": 1995
+    },
+    {
+      "epoch": 0.12278543307086615,
+      "grad_norm": 1.1113699674606323,
+      "learning_rate": 0.00019271564858453424,
+      "loss": 2.716,
+      "step": 1996
+    },
+    {
+      "epoch": 0.12284694881889764,
+      "grad_norm": 1.1459795236587524,
+      "learning_rate": 0.0001927084015346029,
+      "loss": 3.4385,
+      "step": 1997
+    },
+    {
+      "epoch": 0.12290846456692914,
+      "grad_norm": 1.0301477909088135,
+      "learning_rate": 0.00019270115101789425,
+      "loss": 2.8199,
+      "step": 1998
+    },
+    {
+      "epoch": 0.12296998031496063,
+      "grad_norm": 1.053978443145752,
+      "learning_rate": 0.00019269389703467952,
+      "loss": 3.1793,
+      "step": 1999
+    },
+    {
+      "epoch": 0.12303149606299213,
+      "grad_norm": 1.1204826831817627,
+      "learning_rate": 0.0001926866395852299,
+      "loss": 3.0828,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12309301181102363,
+      "grad_norm": 1.1153919696807861,
+      "learning_rate": 0.0001926793786698168,
+      "loss": 2.607,
+      "step": 2001
+    },
+    {
+      "epoch": 0.12315452755905512,
+      "grad_norm": 0.9735493063926697,
+      "learning_rate": 0.00019267211428871176,
+      "loss": 3.1594,
+      "step": 2002
+    },
+    {
+      "epoch": 0.12321604330708662,
+      "grad_norm": 0.9745255708694458,
+      "learning_rate": 0.00019266484644218633,
+      "loss": 3.1539,
+      "step": 2003
+    },
+    {
+      "epoch": 0.12327755905511811,
+      "grad_norm": 1.1295251846313477,
+      "learning_rate": 0.0001926575751305124,
+      "loss": 2.7086,
+      "step": 2004
+    },
+    {
+      "epoch": 0.12333907480314961,
+      "grad_norm": 1.1029890775680542,
+      "learning_rate": 0.00019265030035396186,
+      "loss": 2.6804,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1234005905511811,
+      "grad_norm": 1.1840564012527466,
+      "learning_rate": 0.0001926430221128067,
+      "loss": 3.0049,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1234621062992126,
+      "grad_norm": 1.1415008306503296,
+      "learning_rate": 0.0001926357404073191,
+      "loss": 2.7391,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1235236220472441,
+      "grad_norm": 1.3447836637496948,
+      "learning_rate": 0.00019262845523777133,
+      "loss": 2.4628,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1235851377952756,
+      "grad_norm": 1.0658546686172485,
+      "learning_rate": 0.00019262116660443587,
+      "loss": 2.8659,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12364665354330709,
+      "grad_norm": 0.9677819609642029,
+      "learning_rate": 0.00019261387450758525,
+      "loss": 2.7863,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12370816929133858,
+      "grad_norm": 1.2354575395584106,
+      "learning_rate": 0.00019260657894749214,
+      "loss": 3.2747,
+      "step": 2011
+    },
+    {
+      "epoch": 0.12376968503937008,
+      "grad_norm": 0.976366400718689,
+      "learning_rate": 0.0001925992799244294,
+      "loss": 3.0628,
+      "step": 2012
+    },
+    {
+      "epoch": 0.12383120078740158,
+      "grad_norm": 1.3427525758743286,
+      "learning_rate": 0.0001925919774386699,
+      "loss": 2.8257,
+      "step": 2013
+    },
+    {
+      "epoch": 0.12389271653543307,
+      "grad_norm": 1.20229172706604,
+      "learning_rate": 0.00019258467149048676,
+      "loss": 3.012,
+      "step": 2014
+    },
+    {
+      "epoch": 0.12395423228346457,
+      "grad_norm": 1.5831975936889648,
+      "learning_rate": 0.00019257736208015317,
+      "loss": 2.6716,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12401574803149606,
+      "grad_norm": 0.8937756419181824,
+      "learning_rate": 0.00019257004920794247,
+      "loss": 3.3825,
+      "step": 2016
+    },
+    {
+      "epoch": 0.12407726377952756,
+      "grad_norm": 1.0420842170715332,
+      "learning_rate": 0.00019256273287412813,
+      "loss": 3.2827,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12413877952755906,
+      "grad_norm": 0.8844890594482422,
+      "learning_rate": 0.0001925554130789837,
+      "loss": 3.5221,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12420029527559055,
+      "grad_norm": 1.0559093952178955,
+      "learning_rate": 0.00019254808982278293,
+      "loss": 3.3996,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12426181102362205,
+      "grad_norm": 1.0992286205291748,
+      "learning_rate": 0.00019254076310579968,
+      "loss": 3.0289,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12432332677165354,
+      "grad_norm": 1.1364412307739258,
+      "learning_rate": 0.0001925334329283079,
+      "loss": 3.1433,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12438484251968504,
+      "grad_norm": 1.4686455726623535,
+      "learning_rate": 0.0001925260992905817,
+      "loss": 2.3161,
+      "step": 2022
+    },
+    {
+      "epoch": 0.12444635826771654,
+      "grad_norm": 0.9552106857299805,
+      "learning_rate": 0.00019251876219289533,
+      "loss": 3.5077,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12450787401574803,
+      "grad_norm": 1.074429988861084,
+      "learning_rate": 0.00019251142163552315,
+      "loss": 3.2433,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12456938976377953,
+      "grad_norm": 1.048322081565857,
+      "learning_rate": 0.00019250407761873967,
+      "loss": 2.7491,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12463090551181102,
+      "grad_norm": 1.1786361932754517,
+      "learning_rate": 0.0001924967301428195,
+      "loss": 2.9393,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12469242125984252,
+      "grad_norm": 1.1528972387313843,
+      "learning_rate": 0.0001924893792080374,
+      "loss": 3.0483,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12475393700787402,
+      "grad_norm": 1.0103627443313599,
+      "learning_rate": 0.00019248202481466823,
+      "loss": 3.2859,
+      "step": 2028
+    },
+    {
+      "epoch": 0.12481545275590551,
+      "grad_norm": 0.9548351764678955,
+      "learning_rate": 0.00019247466696298705,
+      "loss": 3.5399,
+      "step": 2029
+    },
+    {
+      "epoch": 0.12487696850393701,
+      "grad_norm": 1.4157570600509644,
+      "learning_rate": 0.000192467305653269,
+      "loss": 3.5233,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1249384842519685,
+      "grad_norm": 1.0106545686721802,
+      "learning_rate": 0.00019245994088578928,
+      "loss": 3.0132,
+      "step": 2031
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 1.1340914964675903,
+      "learning_rate": 0.00019245257266082338,
+      "loss": 2.855,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1250615157480315,
+      "grad_norm": 0.955986499786377,
+      "learning_rate": 0.00019244520097864678,
+      "loss": 3.2766,
+      "step": 2033
+    },
+    {
+      "epoch": 0.125123031496063,
+      "grad_norm": 1.252394199371338,
+      "learning_rate": 0.00019243782583953513,
+      "loss": 3.0718,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1251845472440945,
+      "grad_norm": 0.9618849754333496,
+      "learning_rate": 0.00019243044724376427,
+      "loss": 2.5504,
+      "step": 2035
+    },
+    {
+      "epoch": 0.12524606299212598,
+      "grad_norm": 0.9519174098968506,
+      "learning_rate": 0.00019242306519161008,
+      "loss": 2.9266,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1253075787401575,
+      "grad_norm": 0.8883724212646484,
+      "learning_rate": 0.00019241567968334865,
+      "loss": 3.3918,
+      "step": 2037
+    },
+    {
+      "epoch": 0.12536909448818898,
+      "grad_norm": 0.9389039874076843,
+      "learning_rate": 0.00019240829071925605,
+      "loss": 3.6418,
+      "step": 2038
+    },
+    {
+      "epoch": 0.12543061023622049,
+      "grad_norm": 1.0757887363433838,
+      "learning_rate": 0.00019240089829960873,
+      "loss": 2.5337,
+      "step": 2039
+    },
+    {
+      "epoch": 0.12549212598425197,
+      "grad_norm": 1.1163545846939087,
+      "learning_rate": 0.000192393502424683,
+      "loss": 2.7254,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12555364173228348,
+      "grad_norm": 1.4660543203353882,
+      "learning_rate": 0.00019238610309475554,
+      "loss": 2.6065,
+      "step": 2041
+    },
+    {
+      "epoch": 0.12561515748031496,
+      "grad_norm": 1.1205170154571533,
+      "learning_rate": 0.00019237870031010292,
+      "loss": 3.0165,
+      "step": 2042
+    },
+    {
+      "epoch": 0.12567667322834647,
+      "grad_norm": 1.2783455848693848,
+      "learning_rate": 0.00019237129407100208,
+      "loss": 2.9457,
+      "step": 2043
+    },
+    {
+      "epoch": 0.12573818897637795,
+      "grad_norm": 1.0539286136627197,
+      "learning_rate": 0.00019236388437772988,
+      "loss": 3.3152,
+      "step": 2044
+    },
+    {
+      "epoch": 0.12579970472440946,
+      "grad_norm": 0.8988949656486511,
+      "learning_rate": 0.00019235647123056347,
+      "loss": 3.0199,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12586122047244094,
+      "grad_norm": 0.9836733937263489,
+      "learning_rate": 0.00019234905462978,
+      "loss": 3.4567,
+      "step": 2046
+    },
+    {
+      "epoch": 0.12592273622047245,
+      "grad_norm": 1.0852404832839966,
+      "learning_rate": 0.00019234163457565684,
+      "loss": 2.9914,
+      "step": 2047
+    },
+    {
+      "epoch": 0.12598425196850394,
+      "grad_norm": 0.9840605854988098,
+      "learning_rate": 0.00019233421106847146,
+      "loss": 3.2367,
+      "step": 2048
+    },
+    {
+      "epoch": 0.12604576771653545,
+      "grad_norm": 1.1919362545013428,
+      "learning_rate": 0.00019232678410850148,
+      "loss": 3.0249,
+      "step": 2049
+    },
+    {
+      "epoch": 0.12610728346456693,
+      "grad_norm": 1.1188746690750122,
+      "learning_rate": 0.0001923193536960246,
+      "loss": 3.3021,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12616879921259844,
+      "grad_norm": 1.157455325126648,
+      "learning_rate": 0.00019231191983131867,
+      "loss": 2.6523,
+      "step": 2051
+    },
+    {
+      "epoch": 0.12623031496062992,
+      "grad_norm": 0.8539159893989563,
+      "learning_rate": 0.00019230448251466168,
+      "loss": 3.3593,
+      "step": 2052
+    },
+    {
+      "epoch": 0.12629183070866143,
+      "grad_norm": 1.0869812965393066,
+      "learning_rate": 0.00019229704174633175,
+      "loss": 2.5258,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1263533464566929,
+      "grad_norm": 1.1299452781677246,
+      "learning_rate": 0.0001922895975266071,
+      "loss": 3.3694,
+      "step": 2054
+    },
+    {
+      "epoch": 0.12641486220472442,
+      "grad_norm": 1.1932940483093262,
+      "learning_rate": 0.00019228214985576617,
+      "loss": 2.7718,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1264763779527559,
+      "grad_norm": 1.1455442905426025,
+      "learning_rate": 0.00019227469873408736,
+      "loss": 3.4052,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1265378937007874,
+      "grad_norm": 0.9336668848991394,
+      "learning_rate": 0.00019226724416184938,
+      "loss": 1.5774,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1265994094488189,
+      "grad_norm": 1.0019400119781494,
+      "learning_rate": 0.00019225978613933098,
+      "loss": 2.8272,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1266609251968504,
+      "grad_norm": 1.2102813720703125,
+      "learning_rate": 0.00019225232466681103,
+      "loss": 2.9432,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1267224409448819,
+      "grad_norm": 1.0820071697235107,
+      "learning_rate": 0.00019224485974456853,
+      "loss": 2.8997,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1267839566929134,
+      "grad_norm": 1.0744823217391968,
+      "learning_rate": 0.0001922373913728827,
+      "loss": 3.4986,
+      "step": 2061
+    },
+    {
+      "epoch": 0.12684547244094488,
+      "grad_norm": 0.8637837171554565,
+      "learning_rate": 0.0001922299195520327,
+      "loss": 3.3075,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1269069881889764,
+      "grad_norm": 1.0995596647262573,
+      "learning_rate": 0.00019222244428229802,
+      "loss": 2.7902,
+      "step": 2063
+    },
+    {
+      "epoch": 0.12696850393700787,
+      "grad_norm": 1.1715309619903564,
+      "learning_rate": 0.00019221496556395815,
+      "loss": 2.8722,
+      "step": 2064
+    },
+    {
+      "epoch": 0.12703001968503938,
+      "grad_norm": 1.0623703002929688,
+      "learning_rate": 0.0001922074833972928,
+      "loss": 2.7005,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12709153543307086,
+      "grad_norm": 1.330858826637268,
+      "learning_rate": 0.00019219999778258173,
+      "loss": 3.0989,
+      "step": 2066
+    },
+    {
+      "epoch": 0.12715305118110237,
+      "grad_norm": 1.4439423084259033,
+      "learning_rate": 0.00019219250872010485,
+      "loss": 2.5306,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12721456692913385,
+      "grad_norm": 1.4483182430267334,
+      "learning_rate": 0.00019218501621014221,
+      "loss": 3.2994,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12727608267716536,
+      "grad_norm": 0.9107292294502258,
+      "learning_rate": 0.00019217752025297403,
+      "loss": 3.173,
+      "step": 2069
+    },
+    {
+      "epoch": 0.12733759842519685,
+      "grad_norm": 1.405809760093689,
+      "learning_rate": 0.0001921700208488806,
+      "loss": 2.0781,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12739911417322836,
+      "grad_norm": 1.1390111446380615,
+      "learning_rate": 0.00019216251799814233,
+      "loss": 3.4742,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12746062992125984,
+      "grad_norm": 1.2062963247299194,
+      "learning_rate": 0.00019215501170103977,
+      "loss": 2.877,
+      "step": 2072
+    },
+    {
+      "epoch": 0.12752214566929135,
+      "grad_norm": 1.130778193473816,
+      "learning_rate": 0.00019214750195785366,
+      "loss": 3.4687,
+      "step": 2073
+    },
+    {
+      "epoch": 0.12758366141732283,
+      "grad_norm": 0.9238161444664001,
+      "learning_rate": 0.00019213998876886485,
+      "loss": 3.2269,
+      "step": 2074
+    },
+    {
+      "epoch": 0.12764517716535434,
+      "grad_norm": 1.0989537239074707,
+      "learning_rate": 0.0001921324721343542,
+      "loss": 3.2164,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12770669291338582,
+      "grad_norm": 0.8730928897857666,
+      "learning_rate": 0.00019212495205460286,
+      "loss": 3.0204,
+      "step": 2076
+    },
+    {
+      "epoch": 0.12776820866141733,
+      "grad_norm": 0.9903955459594727,
+      "learning_rate": 0.000192117428529892,
+      "loss": 3.3564,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1278297244094488,
+      "grad_norm": 1.3668349981307983,
+      "learning_rate": 0.00019210990156050297,
+      "loss": 2.5606,
+      "step": 2078
+    },
+    {
+      "epoch": 0.12789124015748032,
+      "grad_norm": 1.0942784547805786,
+      "learning_rate": 0.00019210237114671724,
+      "loss": 2.8923,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1279527559055118,
+      "grad_norm": 1.1084980964660645,
+      "learning_rate": 0.0001920948372888164,
+      "loss": 3.1938,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12801427165354332,
+      "grad_norm": 1.0743461847305298,
+      "learning_rate": 0.0001920872999870822,
+      "loss": 3.276,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1280757874015748,
+      "grad_norm": 0.8713526725769043,
+      "learning_rate": 0.0001920797592417965,
+      "loss": 3.5541,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1281373031496063,
+      "grad_norm": 1.0032566785812378,
+      "learning_rate": 0.0001920722150532412,
+      "loss": 3.4922,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1281988188976378,
+      "grad_norm": 0.9197350144386292,
+      "learning_rate": 0.0001920646674216985,
+      "loss": 3.0177,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1282603346456693,
+      "grad_norm": 1.5101362466812134,
+      "learning_rate": 0.0001920571163474506,
+      "loss": 1.8607,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12832185039370078,
+      "grad_norm": 0.9373806118965149,
+      "learning_rate": 0.00019204956183077986,
+      "loss": 3.2096,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1283833661417323,
+      "grad_norm": 0.9915403127670288,
+      "learning_rate": 0.0001920420038719688,
+      "loss": 2.9811,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12844488188976377,
+      "grad_norm": 1.5326875448226929,
+      "learning_rate": 0.00019203444247130002,
+      "loss": 3.2675,
+      "step": 2088
+    },
+    {
+      "epoch": 0.12850639763779528,
+      "grad_norm": 1.4326406717300415,
+      "learning_rate": 0.00019202687762905632,
+      "loss": 2.7264,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12856791338582677,
+      "grad_norm": 0.9657040238380432,
+      "learning_rate": 0.00019201930934552053,
+      "loss": 3.4602,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12862942913385828,
+      "grad_norm": 1.1502094268798828,
+      "learning_rate": 0.00019201173762097565,
+      "loss": 2.7877,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12869094488188976,
+      "grad_norm": 0.9525432586669922,
+      "learning_rate": 0.0001920041624557049,
+      "loss": 2.9059,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12875246062992127,
+      "grad_norm": 1.0436054468154907,
+      "learning_rate": 0.00019199658384999152,
+      "loss": 3.347,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12881397637795275,
+      "grad_norm": 1.1475242376327515,
+      "learning_rate": 0.00019198900180411886,
+      "loss": 3.1498,
+      "step": 2094
+    },
+    {
+      "epoch": 0.12887549212598426,
+      "grad_norm": 0.9711646437644958,
+      "learning_rate": 0.0001919814163183705,
+      "loss": 2.9347,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12893700787401574,
+      "grad_norm": 1.1273972988128662,
+      "learning_rate": 0.00019197382739303005,
+      "loss": 2.9197,
+      "step": 2096
+    },
+    {
+      "epoch": 0.12899852362204725,
+      "grad_norm": 0.9748035669326782,
+      "learning_rate": 0.00019196623502838134,
+      "loss": 3.6823,
+      "step": 2097
+    },
+    {
+      "epoch": 0.12906003937007873,
+      "grad_norm": 1.1704456806182861,
+      "learning_rate": 0.00019195863922470823,
+      "loss": 2.9897,
+      "step": 2098
+    },
+    {
+      "epoch": 0.12912155511811024,
+      "grad_norm": 1.1298121213912964,
+      "learning_rate": 0.0001919510399822948,
+      "loss": 2.9197,
+      "step": 2099
+    },
+    {
+      "epoch": 0.12918307086614172,
+      "grad_norm": 1.082767367362976,
+      "learning_rate": 0.0001919434373014252,
+      "loss": 3.6334,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12924458661417323,
+      "grad_norm": 0.9302771687507629,
+      "learning_rate": 0.00019193583118238373,
+      "loss": 3.3479,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12930610236220472,
+      "grad_norm": 0.9574851989746094,
+      "learning_rate": 0.00019192822162545484,
+      "loss": 3.0599,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12936761811023623,
+      "grad_norm": 0.867185652256012,
+      "learning_rate": 0.00019192060863092305,
+      "loss": 3.5135,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1294291338582677,
+      "grad_norm": 1.0860754251480103,
+      "learning_rate": 0.0001919129921990731,
+      "loss": 3.3289,
+      "step": 2104
+    },
+    {
+      "epoch": 0.12949064960629922,
+      "grad_norm": 0.8683757185935974,
+      "learning_rate": 0.0001919053723301897,
+      "loss": 3.3389,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1295521653543307,
+      "grad_norm": 1.0162105560302734,
+      "learning_rate": 0.0001918977490245579,
+      "loss": 3.6858,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1296136811023622,
+      "grad_norm": 1.0574015378952026,
+      "learning_rate": 0.00019189012228246273,
+      "loss": 3.2458,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1296751968503937,
+      "grad_norm": 0.9541876912117004,
+      "learning_rate": 0.00019188249210418934,
+      "loss": 3.6721,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1297367125984252,
+      "grad_norm": 1.1270397901535034,
+      "learning_rate": 0.00019187485849002313,
+      "loss": 2.8496,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12979822834645668,
+      "grad_norm": 1.2268378734588623,
+      "learning_rate": 0.0001918672214402495,
+      "loss": 3.098,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1298597440944882,
+      "grad_norm": 0.959633469581604,
+      "learning_rate": 0.00019185958095515405,
+      "loss": 3.3525,
+      "step": 2111
+    },
+    {
+      "epoch": 0.12992125984251968,
+      "grad_norm": 0.9318539500236511,
+      "learning_rate": 0.0001918519370350225,
+      "loss": 3.3669,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1299827755905512,
+      "grad_norm": 0.9936181306838989,
+      "learning_rate": 0.00019184428968014073,
+      "loss": 3.1807,
+      "step": 2113
+    },
+    {
+      "epoch": 0.13004429133858267,
+      "grad_norm": 1.0346472263336182,
+      "learning_rate": 0.00019183663889079462,
+      "loss": 2.5442,
+      "step": 2114
+    },
+    {
+      "epoch": 0.13010580708661418,
+      "grad_norm": 1.0084853172302246,
+      "learning_rate": 0.0001918289846672703,
+      "loss": 2.8472,
+      "step": 2115
+    },
+    {
+      "epoch": 0.13016732283464566,
+      "grad_norm": 0.9196690917015076,
+      "learning_rate": 0.00019182132700985402,
+      "loss": 3.3333,
+      "step": 2116
+    },
+    {
+      "epoch": 0.13022883858267717,
+      "grad_norm": 1.067084550857544,
+      "learning_rate": 0.00019181366591883214,
+      "loss": 2.3181,
+      "step": 2117
+    },
+    {
+      "epoch": 0.13029035433070865,
+      "grad_norm": 0.9955731630325317,
+      "learning_rate": 0.00019180600139449106,
+      "loss": 3.6479,
+      "step": 2118
+    },
+    {
+      "epoch": 0.13035187007874016,
+      "grad_norm": 0.9681401252746582,
+      "learning_rate": 0.00019179833343711754,
+      "loss": 2.822,
+      "step": 2119
+    },
+    {
+      "epoch": 0.13041338582677164,
+      "grad_norm": 0.8227283954620361,
+      "learning_rate": 0.00019179066204699815,
+      "loss": 3.3467,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13047490157480315,
+      "grad_norm": 1.3003356456756592,
+      "learning_rate": 0.0001917829872244199,
+      "loss": 2.148,
+      "step": 2121
+    },
+    {
+      "epoch": 0.13053641732283464,
+      "grad_norm": 1.0744545459747314,
+      "learning_rate": 0.00019177530896966969,
+      "loss": 3.1897,
+      "step": 2122
+    },
+    {
+      "epoch": 0.13059793307086615,
+      "grad_norm": 1.087950348854065,
+      "learning_rate": 0.0001917676272830347,
+      "loss": 3.0748,
+      "step": 2123
+    },
+    {
+      "epoch": 0.13065944881889763,
+      "grad_norm": 0.9239485859870911,
+      "learning_rate": 0.00019175994216480213,
+      "loss": 3.3674,
+      "step": 2124
+    },
+    {
+      "epoch": 0.13072096456692914,
+      "grad_norm": 1.120559811592102,
+      "learning_rate": 0.0001917522536152594,
+      "loss": 3.3656,
+      "step": 2125
+    },
+    {
+      "epoch": 0.13078248031496062,
+      "grad_norm": 0.8917050957679749,
+      "learning_rate": 0.000191744561634694,
+      "loss": 3.0813,
+      "step": 2126
+    },
+    {
+      "epoch": 0.13084399606299213,
+      "grad_norm": 1.1524932384490967,
+      "learning_rate": 0.00019173686622339363,
+      "loss": 2.9802,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1309055118110236,
+      "grad_norm": 1.4267041683197021,
+      "learning_rate": 0.00019172916738164594,
+      "loss": 1.9177,
+      "step": 2128
+    },
+    {
+      "epoch": 0.13096702755905512,
+      "grad_norm": 1.139277458190918,
+      "learning_rate": 0.00019172146510973892,
+      "loss": 2.9783,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1310285433070866,
+      "grad_norm": 0.9189260005950928,
+      "learning_rate": 0.00019171375940796055,
+      "loss": 3.1213,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1310900590551181,
+      "grad_norm": 1.0305603742599487,
+      "learning_rate": 0.00019170605027659898,
+      "loss": 2.291,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1311515748031496,
+      "grad_norm": 1.0102895498275757,
+      "learning_rate": 0.00019169833771594253,
+      "loss": 2.8039,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1312130905511811,
+      "grad_norm": 1.037963628768921,
+      "learning_rate": 0.00019169062172627954,
+      "loss": 2.6809,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1312746062992126,
+      "grad_norm": 1.8688286542892456,
+      "learning_rate": 0.00019168290230789858,
+      "loss": 2.4093,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1313361220472441,
+      "grad_norm": 1.1464343070983887,
+      "learning_rate": 0.0001916751794610883,
+      "loss": 3.8907,
+      "step": 2135
+    },
+    {
+      "epoch": 0.13139763779527558,
+      "grad_norm": 1.102041482925415,
+      "learning_rate": 0.00019166745318613754,
+      "loss": 3.3837,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1314591535433071,
+      "grad_norm": 1.0424846410751343,
+      "learning_rate": 0.00019165972348333515,
+      "loss": 2.9332,
+      "step": 2137
+    },
+    {
+      "epoch": 0.13152066929133857,
+      "grad_norm": 1.3418943881988525,
+      "learning_rate": 0.00019165199035297024,
+      "loss": 3.4042,
+      "step": 2138
+    },
+    {
+      "epoch": 0.13158218503937008,
+      "grad_norm": 0.9681660532951355,
+      "learning_rate": 0.00019164425379533193,
+      "loss": 3.7981,
+      "step": 2139
+    },
+    {
+      "epoch": 0.13164370078740156,
+      "grad_norm": 1.3115614652633667,
+      "learning_rate": 0.00019163651381070955,
+      "loss": 3.1689,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13170521653543307,
+      "grad_norm": 1.4256389141082764,
+      "learning_rate": 0.00019162877039939253,
+      "loss": 2.5058,
+      "step": 2141
+    },
+    {
+      "epoch": 0.13176673228346455,
+      "grad_norm": 0.9928346872329712,
+      "learning_rate": 0.00019162102356167044,
+      "loss": 2.9792,
+      "step": 2142
+    },
+    {
+      "epoch": 0.13182824803149606,
+      "grad_norm": 1.1770774126052856,
+      "learning_rate": 0.000191613273297833,
+      "loss": 2.7296,
+      "step": 2143
+    },
+    {
+      "epoch": 0.13188976377952755,
+      "grad_norm": 1.1950336694717407,
+      "learning_rate": 0.00019160551960816993,
+      "loss": 2.2688,
+      "step": 2144
+    },
+    {
+      "epoch": 0.13195127952755906,
+      "grad_norm": 1.3628133535385132,
+      "learning_rate": 0.00019159776249297124,
+      "loss": 3.2527,
+      "step": 2145
+    },
+    {
+      "epoch": 0.13201279527559054,
+      "grad_norm": 1.0751935243606567,
+      "learning_rate": 0.00019159000195252703,
+      "loss": 2.7886,
+      "step": 2146
+    },
+    {
+      "epoch": 0.13207431102362205,
+      "grad_norm": 0.995201587677002,
+      "learning_rate": 0.00019158223798712744,
+      "loss": 3.4188,
+      "step": 2147
+    },
+    {
+      "epoch": 0.13213582677165353,
+      "grad_norm": 1.1440390348434448,
+      "learning_rate": 0.00019157447059706283,
+      "loss": 2.5546,
+      "step": 2148
+    },
+    {
+      "epoch": 0.13219734251968504,
+      "grad_norm": 1.9894284009933472,
+      "learning_rate": 0.00019156669978262367,
+      "loss": 2.0269,
+      "step": 2149
+    },
+    {
+      "epoch": 0.13225885826771652,
+      "grad_norm": 1.0044523477554321,
+      "learning_rate": 0.00019155892554410052,
+      "loss": 3.3013,
+      "step": 2150
+    },
+    {
+      "epoch": 0.13232037401574803,
+      "grad_norm": 1.1525055170059204,
+      "learning_rate": 0.0001915511478817841,
+      "loss": 3.3194,
+      "step": 2151
+    },
+    {
+      "epoch": 0.13238188976377951,
+      "grad_norm": 1.064759373664856,
+      "learning_rate": 0.00019154336679596526,
+      "loss": 3.1015,
+      "step": 2152
+    },
+    {
+      "epoch": 0.13244340551181102,
+      "grad_norm": 1.5603506565093994,
+      "learning_rate": 0.00019153558228693495,
+      "loss": 2.5329,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1325049212598425,
+      "grad_norm": 1.1099865436553955,
+      "learning_rate": 0.00019152779435498428,
+      "loss": 3.1607,
+      "step": 2154
+    },
+    {
+      "epoch": 0.13256643700787402,
+      "grad_norm": 1.4951767921447754,
+      "learning_rate": 0.00019152000300040446,
+      "loss": 2.3043,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1326279527559055,
+      "grad_norm": 1.5682573318481445,
+      "learning_rate": 0.00019151220822348687,
+      "loss": 3.4735,
+      "step": 2156
+    },
+    {
+      "epoch": 0.132689468503937,
+      "grad_norm": 1.2266223430633545,
+      "learning_rate": 0.000191504410024523,
+      "loss": 2.6343,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1327509842519685,
+      "grad_norm": 0.8419822454452515,
+      "learning_rate": 0.00019149660840380443,
+      "loss": 2.9909,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1328125,
+      "grad_norm": 0.9956502318382263,
+      "learning_rate": 0.00019148880336162288,
+      "loss": 2.923,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1328740157480315,
+      "grad_norm": 1.0171998739242554,
+      "learning_rate": 0.00019148099489827024,
+      "loss": 3.5303,
+      "step": 2160
+    },
+    {
+      "epoch": 0.132935531496063,
+      "grad_norm": 1.0490880012512207,
+      "learning_rate": 0.00019147318301403852,
+      "loss": 3.3975,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1329970472440945,
+      "grad_norm": 1.0135270357131958,
+      "learning_rate": 0.00019146536770921983,
+      "loss": 3.159,
+      "step": 2162
+    },
+    {
+      "epoch": 0.13305856299212598,
+      "grad_norm": 1.2536871433258057,
+      "learning_rate": 0.0001914575489841064,
+      "loss": 2.4913,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1331200787401575,
+      "grad_norm": 0.9797085523605347,
+      "learning_rate": 0.00019144972683899064,
+      "loss": 2.6618,
+      "step": 2164
+    },
+    {
+      "epoch": 0.13318159448818898,
+      "grad_norm": 0.9926231503486633,
+      "learning_rate": 0.00019144190127416499,
+      "loss": 3.6253,
+      "step": 2165
+    },
+    {
+      "epoch": 0.13324311023622049,
+      "grad_norm": 1.1673362255096436,
+      "learning_rate": 0.00019143407228992216,
+      "loss": 3.2796,
+      "step": 2166
+    },
+    {
+      "epoch": 0.13330462598425197,
+      "grad_norm": 1.1229745149612427,
+      "learning_rate": 0.00019142623988655487,
+      "loss": 2.5509,
+      "step": 2167
+    },
+    {
+      "epoch": 0.13336614173228348,
+      "grad_norm": 1.0733563899993896,
+      "learning_rate": 0.000191418404064356,
+      "loss": 3.2107,
+      "step": 2168
+    },
+    {
+      "epoch": 0.13342765748031496,
+      "grad_norm": 1.1235971450805664,
+      "learning_rate": 0.00019141056482361858,
+      "loss": 2.4676,
+      "step": 2169
+    },
+    {
+      "epoch": 0.13348917322834647,
+      "grad_norm": 1.1499614715576172,
+      "learning_rate": 0.00019140272216463578,
+      "loss": 3.5562,
+      "step": 2170
+    },
+    {
+      "epoch": 0.13355068897637795,
+      "grad_norm": 1.2973686456680298,
+      "learning_rate": 0.00019139487608770087,
+      "loss": 2.9609,
+      "step": 2171
+    },
+    {
+      "epoch": 0.13361220472440946,
+      "grad_norm": 1.265174388885498,
+      "learning_rate": 0.00019138702659310717,
+      "loss": 1.8579,
+      "step": 2172
+    },
+    {
+      "epoch": 0.13367372047244094,
+      "grad_norm": 1.1295901536941528,
+      "learning_rate": 0.00019137917368114828,
+      "loss": 3.2393,
+      "step": 2173
+    },
+    {
+      "epoch": 0.13373523622047245,
+      "grad_norm": 1.0476328134536743,
+      "learning_rate": 0.00019137131735211783,
+      "loss": 2.6893,
+      "step": 2174
+    },
+    {
+      "epoch": 0.13379675196850394,
+      "grad_norm": 1.0524578094482422,
+      "learning_rate": 0.00019136345760630965,
+      "loss": 3.0721,
+      "step": 2175
+    },
+    {
+      "epoch": 0.13385826771653545,
+      "grad_norm": 1.4952958822250366,
+      "learning_rate": 0.00019135559444401757,
+      "loss": 3.0647,
+      "step": 2176
+    },
+    {
+      "epoch": 0.13391978346456693,
+      "grad_norm": 0.9847830533981323,
+      "learning_rate": 0.0001913477278655357,
+      "loss": 3.2457,
+      "step": 2177
+    },
+    {
+      "epoch": 0.13398129921259844,
+      "grad_norm": 1.1340028047561646,
+      "learning_rate": 0.00019133985787115818,
+      "loss": 3.0013,
+      "step": 2178
+    },
+    {
+      "epoch": 0.13404281496062992,
+      "grad_norm": 1.2363734245300293,
+      "learning_rate": 0.00019133198446117926,
+      "loss": 3.1031,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13410433070866143,
+      "grad_norm": 1.037198543548584,
+      "learning_rate": 0.00019132410763589343,
+      "loss": 3.199,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1341658464566929,
+      "grad_norm": 0.9830179810523987,
+      "learning_rate": 0.00019131622739559522,
+      "loss": 3.1668,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13422736220472442,
+      "grad_norm": 1.25962495803833,
+      "learning_rate": 0.0001913083437405793,
+      "loss": 2.937,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1342888779527559,
+      "grad_norm": 1.3005261421203613,
+      "learning_rate": 0.00019130045667114045,
+      "loss": 3.1793,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1343503937007874,
+      "grad_norm": 1.0775269269943237,
+      "learning_rate": 0.00019129256618757363,
+      "loss": 3.3065,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1344119094488189,
+      "grad_norm": 1.1574277877807617,
+      "learning_rate": 0.0001912846722901739,
+      "loss": 3.1933,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1344734251968504,
+      "grad_norm": 1.239256739616394,
+      "learning_rate": 0.00019127677497923643,
+      "loss": 2.7756,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1345349409448819,
+      "grad_norm": 1.051406741142273,
+      "learning_rate": 0.00019126887425505659,
+      "loss": 2.5486,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1345964566929134,
+      "grad_norm": 1.035353183746338,
+      "learning_rate": 0.00019126097011792976,
+      "loss": 2.7817,
+      "step": 2188
+    },
+    {
+      "epoch": 0.13465797244094488,
+      "grad_norm": 1.3355222940444946,
+      "learning_rate": 0.00019125306256815154,
+      "loss": 2.7027,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1347194881889764,
+      "grad_norm": 1.0844676494598389,
+      "learning_rate": 0.00019124515160601758,
+      "loss": 3.2608,
+      "step": 2190
+    },
+    {
+      "epoch": 0.13478100393700787,
+      "grad_norm": 0.948212742805481,
+      "learning_rate": 0.0001912372372318238,
+      "loss": 3.5898,
+      "step": 2191
+    },
+    {
+      "epoch": 0.13484251968503938,
+      "grad_norm": 0.9655611515045166,
+      "learning_rate": 0.0001912293194458661,
+      "loss": 3.1785,
+      "step": 2192
+    },
+    {
+      "epoch": 0.13490403543307086,
+      "grad_norm": 1.2892543077468872,
+      "learning_rate": 0.0001912213982484405,
+      "loss": 3.1006,
+      "step": 2193
+    },
+    {
+      "epoch": 0.13496555118110237,
+      "grad_norm": 0.8357173204421997,
+      "learning_rate": 0.00019121347363984336,
+      "loss": 3.7352,
+      "step": 2194
+    },
+    {
+      "epoch": 0.13502706692913385,
+      "grad_norm": 1.3466960191726685,
+      "learning_rate": 0.00019120554562037083,
+      "loss": 2.8096,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13508858267716536,
+      "grad_norm": 1.289368987083435,
+      "learning_rate": 0.00019119761419031953,
+      "loss": 2.7014,
+      "step": 2196
+    },
+    {
+      "epoch": 0.13515009842519685,
+      "grad_norm": 1.2319343090057373,
+      "learning_rate": 0.00019118967934998596,
+      "loss": 2.9915,
+      "step": 2197
+    },
+    {
+      "epoch": 0.13521161417322836,
+      "grad_norm": 1.1050667762756348,
+      "learning_rate": 0.0001911817410996669,
+      "loss": 2.9736,
+      "step": 2198
+    },
+    {
+      "epoch": 0.13527312992125984,
+      "grad_norm": 1.1322691440582275,
+      "learning_rate": 0.00019117379943965916,
+      "loss": 2.7692,
+      "step": 2199
+    },
+    {
+      "epoch": 0.13533464566929135,
+      "grad_norm": 1.1215848922729492,
+      "learning_rate": 0.0001911658543702597,
+      "loss": 3.3358,
+      "step": 2200
+    },
+    {
+      "epoch": 0.13539616141732283,
+      "grad_norm": 0.907228946685791,
+      "learning_rate": 0.00019115790589176567,
+      "loss": 2.7941,
+      "step": 2201
+    },
+    {
+      "epoch": 0.13545767716535434,
+      "grad_norm": 0.9893189668655396,
+      "learning_rate": 0.00019114995400447424,
+      "loss": 3.0545,
+      "step": 2202
+    },
+    {
+      "epoch": 0.13551919291338582,
+      "grad_norm": 1.3493099212646484,
+      "learning_rate": 0.00019114199870868284,
+      "loss": 3.8359,
+      "step": 2203
+    },
+    {
+      "epoch": 0.13558070866141733,
+      "grad_norm": 0.9223587512969971,
+      "learning_rate": 0.00019113404000468886,
+      "loss": 2.8744,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1356422244094488,
+      "grad_norm": 0.9883958101272583,
+      "learning_rate": 0.00019112607789278997,
+      "loss": 3.3223,
+      "step": 2205
+    },
+    {
+      "epoch": 0.13570374015748032,
+      "grad_norm": 0.9635325074195862,
+      "learning_rate": 0.00019111811237328393,
+      "loss": 2.5272,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1357652559055118,
+      "grad_norm": 1.406569242477417,
+      "learning_rate": 0.00019111014344646855,
+      "loss": 2.9835,
+      "step": 2207
+    },
+    {
+      "epoch": 0.13582677165354332,
+      "grad_norm": 1.0633811950683594,
+      "learning_rate": 0.00019110217111264187,
+      "loss": 3.0102,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1358882874015748,
+      "grad_norm": 0.817923367023468,
+      "learning_rate": 0.00019109419537210198,
+      "loss": 2.992,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1359498031496063,
+      "grad_norm": 0.8710224628448486,
+      "learning_rate": 0.00019108621622514714,
+      "loss": 3.3439,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1360113188976378,
+      "grad_norm": 1.0938348770141602,
+      "learning_rate": 0.00019107823367207568,
+      "loss": 2.9815,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1360728346456693,
+      "grad_norm": 1.132415771484375,
+      "learning_rate": 0.00019107024771318622,
+      "loss": 3.3152,
+      "step": 2212
+    },
+    {
+      "epoch": 0.13613435039370078,
+      "grad_norm": 1.0283442735671997,
+      "learning_rate": 0.00019106225834877728,
+      "loss": 2.8692,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1361958661417323,
+      "grad_norm": 1.0942211151123047,
+      "learning_rate": 0.00019105426557914764,
+      "loss": 2.8515,
+      "step": 2214
+    },
+    {
+      "epoch": 0.13625738188976377,
+      "grad_norm": 1.0305266380310059,
+      "learning_rate": 0.00019104626940459623,
+      "loss": 3.7,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13631889763779528,
+      "grad_norm": 1.2924418449401855,
+      "learning_rate": 0.000191038269825422,
+      "loss": 3.4524,
+      "step": 2216
+    },
+    {
+      "epoch": 0.13638041338582677,
+      "grad_norm": 0.9356006979942322,
+      "learning_rate": 0.00019103026684192415,
+      "loss": 3.2669,
+      "step": 2217
+    },
+    {
+      "epoch": 0.13644192913385828,
+      "grad_norm": 0.9871873259544373,
+      "learning_rate": 0.00019102226045440192,
+      "loss": 3.144,
+      "step": 2218
+    },
+    {
+      "epoch": 0.13650344488188976,
+      "grad_norm": 1.0207502841949463,
+      "learning_rate": 0.0001910142506631547,
+      "loss": 3.5336,
+      "step": 2219
+    },
+    {
+      "epoch": 0.13656496062992127,
+      "grad_norm": 1.0752569437026978,
+      "learning_rate": 0.00019100623746848202,
+      "loss": 3.2811,
+      "step": 2220
+    },
+    {
+      "epoch": 0.13662647637795275,
+      "grad_norm": 1.2241435050964355,
+      "learning_rate": 0.00019099822087068348,
+      "loss": 2.4628,
+      "step": 2221
+    },
+    {
+      "epoch": 0.13668799212598426,
+      "grad_norm": 1.189210057258606,
+      "learning_rate": 0.00019099020087005893,
+      "loss": 2.7402,
+      "step": 2222
+    },
+    {
+      "epoch": 0.13674950787401574,
+      "grad_norm": 1.6900004148483276,
+      "learning_rate": 0.00019098217746690824,
+      "loss": 2.3356,
+      "step": 2223
+    },
+    {
+      "epoch": 0.13681102362204725,
+      "grad_norm": 1.2660161256790161,
+      "learning_rate": 0.00019097415066153146,
+      "loss": 3.2904,
+      "step": 2224
+    },
+    {
+      "epoch": 0.13687253937007873,
+      "grad_norm": 1.000206470489502,
+      "learning_rate": 0.00019096612045422872,
+      "loss": 3.1005,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13693405511811024,
+      "grad_norm": 1.1480591297149658,
+      "learning_rate": 0.00019095808684530032,
+      "loss": 2.9426,
+      "step": 2226
+    },
+    {
+      "epoch": 0.13699557086614172,
+      "grad_norm": 1.0795291662216187,
+      "learning_rate": 0.00019095004983504669,
+      "loss": 2.9706,
+      "step": 2227
+    },
+    {
+      "epoch": 0.13705708661417323,
+      "grad_norm": 1.0067152976989746,
+      "learning_rate": 0.0001909420094237683,
+      "loss": 3.2742,
+      "step": 2228
+    },
+    {
+      "epoch": 0.13711860236220472,
+      "grad_norm": 1.3154914379119873,
+      "learning_rate": 0.00019093396561176588,
+      "loss": 2.2371,
+      "step": 2229
+    },
+    {
+      "epoch": 0.13718011811023623,
+      "grad_norm": 1.277273416519165,
+      "learning_rate": 0.00019092591839934019,
+      "loss": 3.2151,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1372416338582677,
+      "grad_norm": 1.1501091718673706,
+      "learning_rate": 0.00019091786778679221,
+      "loss": 2.5487,
+      "step": 2231
+    },
+    {
+      "epoch": 0.13730314960629922,
+      "grad_norm": 0.964309573173523,
+      "learning_rate": 0.0001909098137744229,
+      "loss": 3.5667,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1373646653543307,
+      "grad_norm": 1.465368390083313,
+      "learning_rate": 0.00019090175636253353,
+      "loss": 3.2643,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1374261811023622,
+      "grad_norm": 0.9411139488220215,
+      "learning_rate": 0.00019089369555142532,
+      "loss": 3.764,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1374876968503937,
+      "grad_norm": 1.0967278480529785,
+      "learning_rate": 0.00019088563134139977,
+      "loss": 3.2719,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1375492125984252,
+      "grad_norm": 1.075393795967102,
+      "learning_rate": 0.00019087756373275836,
+      "loss": 2.8014,
+      "step": 2236
+    },
+    {
+      "epoch": 0.13761072834645668,
+      "grad_norm": 1.129799485206604,
+      "learning_rate": 0.00019086949272580284,
+      "loss": 3.2005,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1376722440944882,
+      "grad_norm": 1.0808391571044922,
+      "learning_rate": 0.000190861418320835,
+      "loss": 3.1418,
+      "step": 2238
+    },
+    {
+      "epoch": 0.13773375984251968,
+      "grad_norm": 1.4293396472930908,
+      "learning_rate": 0.00019085334051815672,
+      "loss": 3.7516,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1377952755905512,
+      "grad_norm": 1.0245357751846313,
+      "learning_rate": 0.00019084525931807017,
+      "loss": 2.6716,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13785679133858267,
+      "grad_norm": 1.23155677318573,
+      "learning_rate": 0.00019083717472087748,
+      "loss": 2.5286,
+      "step": 2241
+    },
+    {
+      "epoch": 0.13791830708661418,
+      "grad_norm": 0.959392249584198,
+      "learning_rate": 0.00019082908672688097,
+      "loss": 3.1431,
+      "step": 2242
+    },
+    {
+      "epoch": 0.13797982283464566,
+      "grad_norm": 1.230921983718872,
+      "learning_rate": 0.0001908209953363831,
+      "loss": 3.4691,
+      "step": 2243
+    },
+    {
+      "epoch": 0.13804133858267717,
+      "grad_norm": 0.9128425121307373,
+      "learning_rate": 0.00019081290054968642,
+      "loss": 3.4079,
+      "step": 2244
+    },
+    {
+      "epoch": 0.13810285433070865,
+      "grad_norm": 1.1109095811843872,
+      "learning_rate": 0.00019080480236709365,
+      "loss": 3.2824,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13816437007874016,
+      "grad_norm": 1.0398370027542114,
+      "learning_rate": 0.00019079670078890764,
+      "loss": 2.1785,
+      "step": 2246
+    },
+    {
+      "epoch": 0.13822588582677164,
+      "grad_norm": 1.0483040809631348,
+      "learning_rate": 0.00019078859581543126,
+      "loss": 2.8166,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13828740157480315,
+      "grad_norm": 1.0008594989776611,
+      "learning_rate": 0.00019078048744696768,
+      "loss": 3.2393,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13834891732283464,
+      "grad_norm": 1.2194232940673828,
+      "learning_rate": 0.0001907723756838201,
+      "loss": 2.923,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13841043307086615,
+      "grad_norm": 0.8554781675338745,
+      "learning_rate": 0.0001907642605262918,
+      "loss": 3.6347,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13847194881889763,
+      "grad_norm": 1.0500693321228027,
+      "learning_rate": 0.00019075614197468628,
+      "loss": 2.8629,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13853346456692914,
+      "grad_norm": 1.1821551322937012,
+      "learning_rate": 0.00019074802002930712,
+      "loss": 3.2055,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13859498031496062,
+      "grad_norm": 0.9844227433204651,
+      "learning_rate": 0.00019073989469045802,
+      "loss": 2.3517,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13865649606299213,
+      "grad_norm": 1.1263227462768555,
+      "learning_rate": 0.00019073176595844286,
+      "loss": 2.5478,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1387180118110236,
+      "grad_norm": 1.8891668319702148,
+      "learning_rate": 0.00019072363383356559,
+      "loss": 2.2956,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13877952755905512,
+      "grad_norm": 1.19582998752594,
+      "learning_rate": 0.0001907154983161303,
+      "loss": 2.6697,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1388410433070866,
+      "grad_norm": 1.1946008205413818,
+      "learning_rate": 0.0001907073594064412,
+      "loss": 2.8669,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1389025590551181,
+      "grad_norm": 1.4822584390640259,
+      "learning_rate": 0.00019069921710480268,
+      "loss": 2.967,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1389640748031496,
+      "grad_norm": 0.8892099857330322,
+      "learning_rate": 0.00019069107141151916,
+      "loss": 3.357,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1390255905511811,
+      "grad_norm": 1.1192351579666138,
+      "learning_rate": 0.0001906829223268953,
+      "loss": 2.4346,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1390871062992126,
+      "grad_norm": 1.2806509733200073,
+      "learning_rate": 0.0001906747698512358,
+      "loss": 3.1898,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1391486220472441,
+      "grad_norm": 1.3799160718917847,
+      "learning_rate": 0.00019066661398484553,
+      "loss": 2.5949,
+      "step": 2262
+    },
+    {
+      "epoch": 0.13921013779527558,
+      "grad_norm": 1.0725916624069214,
+      "learning_rate": 0.00019065845472802944,
+      "loss": 1.9716,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1392716535433071,
+      "grad_norm": 0.9174157977104187,
+      "learning_rate": 0.00019065029208109274,
+      "loss": 3.3039,
+      "step": 2264
+    },
+    {
+      "epoch": 0.13933316929133857,
+      "grad_norm": 1.1072068214416504,
+      "learning_rate": 0.00019064212604434053,
+      "loss": 2.8258,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13939468503937008,
+      "grad_norm": 1.3751025199890137,
+      "learning_rate": 0.00019063395661807827,
+      "loss": 3.3314,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13945620078740156,
+      "grad_norm": 1.1950440406799316,
+      "learning_rate": 0.0001906257838026114,
+      "loss": 2.9484,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13951771653543307,
+      "grad_norm": 1.0542000532150269,
+      "learning_rate": 0.00019061760759824558,
+      "loss": 3.1632,
+      "step": 2268
+    },
+    {
+      "epoch": 0.13957923228346455,
+      "grad_norm": 1.0068243741989136,
+      "learning_rate": 0.00019060942800528652,
+      "loss": 3.076,
+      "step": 2269
+    },
+    {
+      "epoch": 0.13964074803149606,
+      "grad_norm": 1.282339334487915,
+      "learning_rate": 0.0001906012450240401,
+      "loss": 2.6288,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13970226377952755,
+      "grad_norm": 1.2379690408706665,
+      "learning_rate": 0.0001905930586548123,
+      "loss": 2.6386,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13976377952755906,
+      "grad_norm": 1.1236129999160767,
+      "learning_rate": 0.00019058486889790931,
+      "loss": 3.6571,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13982529527559054,
+      "grad_norm": 0.9708451628684998,
+      "learning_rate": 0.00019057667575363735,
+      "loss": 3.453,
+      "step": 2273
+    },
+    {
+      "epoch": 0.13988681102362205,
+      "grad_norm": 1.043823003768921,
+      "learning_rate": 0.00019056847922230277,
+      "loss": 2.8158,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13994832677165353,
+      "grad_norm": 1.3115283250808716,
+      "learning_rate": 0.00019056027930421207,
+      "loss": 3.2963,
+      "step": 2275
+    },
+    {
+      "epoch": 0.14000984251968504,
+      "grad_norm": 1.0105721950531006,
+      "learning_rate": 0.00019055207599967192,
+      "loss": 3.1878,
+      "step": 2276
+    },
+    {
+      "epoch": 0.14007135826771652,
+      "grad_norm": 1.538059949874878,
+      "learning_rate": 0.00019054386930898907,
+      "loss": 3.4132,
+      "step": 2277
+    },
+    {
+      "epoch": 0.14013287401574803,
+      "grad_norm": 20.41179656982422,
+      "learning_rate": 0.00019053565923247037,
+      "loss": 3.3529,
+      "step": 2278
+    },
+    {
+      "epoch": 0.14019438976377951,
+      "grad_norm": 1.3564420938491821,
+      "learning_rate": 0.00019052744577042284,
+      "loss": 2.8479,
+      "step": 2279
+    },
+    {
+      "epoch": 0.14025590551181102,
+      "grad_norm": 1.141400694847107,
+      "learning_rate": 0.00019051922892315368,
+      "loss": 2.5381,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1403174212598425,
+      "grad_norm": 1.0545505285263062,
+      "learning_rate": 0.0001905110086909701,
+      "loss": 2.9627,
+      "step": 2281
+    },
+    {
+      "epoch": 0.14037893700787402,
+      "grad_norm": 1.2095601558685303,
+      "learning_rate": 0.00019050278507417944,
+      "loss": 2.9454,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1404404527559055,
+      "grad_norm": 0.9952242374420166,
+      "learning_rate": 0.00019049455807308932,
+      "loss": 3.3249,
+      "step": 2283
+    },
+    {
+      "epoch": 0.140501968503937,
+      "grad_norm": 0.9094886183738708,
+      "learning_rate": 0.00019048632768800736,
+      "loss": 3.2085,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1405634842519685,
+      "grad_norm": 0.9985274076461792,
+      "learning_rate": 0.0001904780939192413,
+      "loss": 2.8943,
+      "step": 2285
+    },
+    {
+      "epoch": 0.140625,
+      "grad_norm": 0.9220712780952454,
+      "learning_rate": 0.00019046985676709902,
+      "loss": 3.54,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1406865157480315,
+      "grad_norm": 1.0323412418365479,
+      "learning_rate": 0.00019046161623188857,
+      "loss": 2.9035,
+      "step": 2287
+    },
+    {
+      "epoch": 0.140748031496063,
+      "grad_norm": 1.088555932044983,
+      "learning_rate": 0.00019045337231391813,
+      "loss": 2.8438,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1408095472440945,
+      "grad_norm": 0.9906744360923767,
+      "learning_rate": 0.00019044512501349594,
+      "loss": 3.1622,
+      "step": 2289
+    },
+    {
+      "epoch": 0.14087106299212598,
+      "grad_norm": 1.2523082494735718,
+      "learning_rate": 0.0001904368743309304,
+      "loss": 3.3648,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1409325787401575,
+      "grad_norm": 0.8926820158958435,
+      "learning_rate": 0.00019042862026653006,
+      "loss": 3.0401,
+      "step": 2291
+    },
+    {
+      "epoch": 0.14099409448818898,
+      "grad_norm": 1.1335440874099731,
+      "learning_rate": 0.00019042036282060357,
+      "loss": 2.6867,
+      "step": 2292
+    },
+    {
+      "epoch": 0.14105561023622049,
+      "grad_norm": 0.9188825488090515,
+      "learning_rate": 0.00019041210199345968,
+      "loss": 3.1473,
+      "step": 2293
+    },
+    {
+      "epoch": 0.14111712598425197,
+      "grad_norm": 0.9841790199279785,
+      "learning_rate": 0.00019040383778540736,
+      "loss": 2.5193,
+      "step": 2294
+    },
+    {
+      "epoch": 0.14117864173228348,
+      "grad_norm": 1.1800410747528076,
+      "learning_rate": 0.0001903955701967556,
+      "loss": 3.2378,
+      "step": 2295
+    },
+    {
+      "epoch": 0.14124015748031496,
+      "grad_norm": 1.0771677494049072,
+      "learning_rate": 0.00019038729922781357,
+      "loss": 3.3596,
+      "step": 2296
+    },
+    {
+      "epoch": 0.14130167322834647,
+      "grad_norm": 0.9990601539611816,
+      "learning_rate": 0.00019037902487889055,
+      "loss": 2.7752,
+      "step": 2297
+    },
+    {
+      "epoch": 0.14136318897637795,
+      "grad_norm": 1.1652987003326416,
+      "learning_rate": 0.000190370747150296,
+      "loss": 2.7985,
+      "step": 2298
+    },
+    {
+      "epoch": 0.14142470472440946,
+      "grad_norm": 1.2661709785461426,
+      "learning_rate": 0.0001903624660423394,
+      "loss": 3.3274,
+      "step": 2299
+    },
+    {
+      "epoch": 0.14148622047244094,
+      "grad_norm": 1.2306725978851318,
+      "learning_rate": 0.00019035418155533048,
+      "loss": 2.2388,
+      "step": 2300
+    },
+    {
+      "epoch": 0.14154773622047245,
+      "grad_norm": 2.470457077026367,
+      "learning_rate": 0.00019034589368957896,
+      "loss": 2.6229,
+      "step": 2301
+    },
+    {
+      "epoch": 0.14160925196850394,
+      "grad_norm": 1.0007377862930298,
+      "learning_rate": 0.0001903376024453948,
+      "loss": 3.1514,
+      "step": 2302
+    },
+    {
+      "epoch": 0.14167076771653545,
+      "grad_norm": 1.001059889793396,
+      "learning_rate": 0.00019032930782308808,
+      "loss": 3.033,
+      "step": 2303
+    },
+    {
+      "epoch": 0.14173228346456693,
+      "grad_norm": 0.9958366751670837,
+      "learning_rate": 0.0001903210098229689,
+      "loss": 2.7826,
+      "step": 2304
+    },
+    {
+      "epoch": 0.14179379921259844,
+      "grad_norm": 1.2183269262313843,
+      "learning_rate": 0.0001903127084453476,
+      "loss": 2.7098,
+      "step": 2305
+    },
+    {
+      "epoch": 0.14185531496062992,
+      "grad_norm": 1.005039930343628,
+      "learning_rate": 0.00019030440369053459,
+      "loss": 3.2985,
+      "step": 2306
+    },
+    {
+      "epoch": 0.14191683070866143,
+      "grad_norm": 0.8526681065559387,
+      "learning_rate": 0.00019029609555884047,
+      "loss": 3.0658,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1419783464566929,
+      "grad_norm": 1.026831865310669,
+      "learning_rate": 0.00019028778405057585,
+      "loss": 2.5164,
+      "step": 2308
+    },
+    {
+      "epoch": 0.14203986220472442,
+      "grad_norm": 1.047093152999878,
+      "learning_rate": 0.0001902794691660516,
+      "loss": 3.1854,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1421013779527559,
+      "grad_norm": 0.9196159243583679,
+      "learning_rate": 0.00019027115090557858,
+      "loss": 3.2534,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1421628937007874,
+      "grad_norm": 0.9077377319335938,
+      "learning_rate": 0.00019026282926946792,
+      "loss": 3.2193,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1422244094488189,
+      "grad_norm": 1.2973326444625854,
+      "learning_rate": 0.00019025450425803073,
+      "loss": 2.3427,
+      "step": 2312
+    },
+    {
+      "epoch": 0.1422859251968504,
+      "grad_norm": 0.9583438634872437,
+      "learning_rate": 0.00019024617587157838,
+      "loss": 3.2241,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1423474409448819,
+      "grad_norm": 1.148821473121643,
+      "learning_rate": 0.00019023784411042227,
+      "loss": 2.4698,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1424089566929134,
+      "grad_norm": 0.9410596489906311,
+      "learning_rate": 0.00019022950897487395,
+      "loss": 3.3292,
+      "step": 2315
+    },
+    {
+      "epoch": 0.14247047244094488,
+      "grad_norm": 1.1321419477462769,
+      "learning_rate": 0.00019022117046524516,
+      "loss": 3.3119,
+      "step": 2316
+    },
+    {
+      "epoch": 0.1425319881889764,
+      "grad_norm": 1.0096148252487183,
+      "learning_rate": 0.00019021282858184763,
+      "loss": 3.5015,
+      "step": 2317
+    },
+    {
+      "epoch": 0.14259350393700787,
+      "grad_norm": 1.0742924213409424,
+      "learning_rate": 0.00019020448332499341,
+      "loss": 3.2159,
+      "step": 2318
+    },
+    {
+      "epoch": 0.14265501968503938,
+      "grad_norm": 1.168948769569397,
+      "learning_rate": 0.00019019613469499448,
+      "loss": 3.1299,
+      "step": 2319
+    },
+    {
+      "epoch": 0.14271653543307086,
+      "grad_norm": 1.2413246631622314,
+      "learning_rate": 0.00019018778269216306,
+      "loss": 3.0736,
+      "step": 2320
+    },
+    {
+      "epoch": 0.14277805118110237,
+      "grad_norm": 1.813258171081543,
+      "learning_rate": 0.00019017942731681148,
+      "loss": 3.1674,
+      "step": 2321
+    },
+    {
+      "epoch": 0.14283956692913385,
+      "grad_norm": 1.198988437652588,
+      "learning_rate": 0.00019017106856925216,
+      "loss": 2.9882,
+      "step": 2322
+    },
+    {
+      "epoch": 0.14290108267716536,
+      "grad_norm": 1.601291298866272,
+      "learning_rate": 0.00019016270644979768,
+      "loss": 2.825,
+      "step": 2323
+    },
+    {
+      "epoch": 0.14296259842519685,
+      "grad_norm": 1.0884203910827637,
+      "learning_rate": 0.00019015434095876076,
+      "loss": 2.5524,
+      "step": 2324
+    },
+    {
+      "epoch": 0.14302411417322836,
+      "grad_norm": 1.0770457983016968,
+      "learning_rate": 0.00019014597209645418,
+      "loss": 2.5154,
+      "step": 2325
+    },
+    {
+      "epoch": 0.14308562992125984,
+      "grad_norm": 1.2785227298736572,
+      "learning_rate": 0.0001901375998631909,
+      "loss": 2.8953,
+      "step": 2326
+    },
+    {
+      "epoch": 0.14314714566929135,
+      "grad_norm": 0.9463295340538025,
+      "learning_rate": 0.00019012922425928403,
+      "loss": 3.2088,
+      "step": 2327
+    },
+    {
+      "epoch": 0.14320866141732283,
+      "grad_norm": 1.3348140716552734,
+      "learning_rate": 0.00019012084528504677,
+      "loss": 2.8686,
+      "step": 2328
+    },
+    {
+      "epoch": 0.14327017716535434,
+      "grad_norm": 1.1870275735855103,
+      "learning_rate": 0.0001901124629407924,
+      "loss": 2.8089,
+      "step": 2329
+    },
+    {
+      "epoch": 0.14333169291338582,
+      "grad_norm": 0.9575624465942383,
+      "learning_rate": 0.00019010407722683438,
+      "loss": 3.2573,
+      "step": 2330
+    },
+    {
+      "epoch": 0.14339320866141733,
+      "grad_norm": 1.2884124517440796,
+      "learning_rate": 0.00019009568814348633,
+      "loss": 3.4437,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1434547244094488,
+      "grad_norm": 0.9902863502502441,
+      "learning_rate": 0.0001900872956910619,
+      "loss": 3.5685,
+      "step": 2332
+    },
+    {
+      "epoch": 0.14351624015748032,
+      "grad_norm": 0.9380813837051392,
+      "learning_rate": 0.000190078899869875,
+      "loss": 3.4962,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1435777559055118,
+      "grad_norm": 0.9535052180290222,
+      "learning_rate": 0.0001900705006802395,
+      "loss": 3.9666,
+      "step": 2334
+    },
+    {
+      "epoch": 0.14363927165354332,
+      "grad_norm": 1.104731559753418,
+      "learning_rate": 0.00019006209812246952,
+      "loss": 3.0412,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1437007874015748,
+      "grad_norm": 1.03782057762146,
+      "learning_rate": 0.00019005369219687933,
+      "loss": 3.5872,
+      "step": 2336
+    },
+    {
+      "epoch": 0.1437623031496063,
+      "grad_norm": 0.9733284115791321,
+      "learning_rate": 0.00019004528290378315,
+      "loss": 2.8488,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1438238188976378,
+      "grad_norm": 1.0814661979675293,
+      "learning_rate": 0.0001900368702434955,
+      "loss": 3.0892,
+      "step": 2338
+    },
+    {
+      "epoch": 0.1438853346456693,
+      "grad_norm": 0.9839680790901184,
+      "learning_rate": 0.00019002845421633096,
+      "loss": 2.8826,
+      "step": 2339
+    },
+    {
+      "epoch": 0.14394685039370078,
+      "grad_norm": 1.101294994354248,
+      "learning_rate": 0.00019002003482260427,
+      "loss": 3.1865,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1440083661417323,
+      "grad_norm": 1.2929205894470215,
+      "learning_rate": 0.0001900116120626302,
+      "loss": 2.7967,
+      "step": 2341
+    },
+    {
+      "epoch": 0.14406988188976377,
+      "grad_norm": 1.1540355682373047,
+      "learning_rate": 0.00019000318593672376,
+      "loss": 3.2342,
+      "step": 2342
+    },
+    {
+      "epoch": 0.14413139763779528,
+      "grad_norm": 1.623927354812622,
+      "learning_rate": 0.0001899947564452001,
+      "loss": 2.4769,
+      "step": 2343
+    },
+    {
+      "epoch": 0.14419291338582677,
+      "grad_norm": 1.1608070135116577,
+      "learning_rate": 0.0001899863235883743,
+      "loss": 3.0246,
+      "step": 2344
+    },
+    {
+      "epoch": 0.14425442913385828,
+      "grad_norm": 1.1241041421890259,
+      "learning_rate": 0.00018997788736656182,
+      "loss": 4.1595,
+      "step": 2345
+    },
+    {
+      "epoch": 0.14431594488188976,
+      "grad_norm": 1.146047830581665,
+      "learning_rate": 0.00018996944778007806,
+      "loss": 2.6302,
+      "step": 2346
+    },
+    {
+      "epoch": 0.14437746062992127,
+      "grad_norm": 1.4306036233901978,
+      "learning_rate": 0.00018996100482923867,
+      "loss": 3.1984,
+      "step": 2347
+    },
+    {
+      "epoch": 0.14443897637795275,
+      "grad_norm": 1.1225018501281738,
+      "learning_rate": 0.0001899525585143593,
+      "loss": 3.607,
+      "step": 2348
+    },
+    {
+      "epoch": 0.14450049212598426,
+      "grad_norm": 0.9958668351173401,
+      "learning_rate": 0.00018994410883575586,
+      "loss": 3.2379,
+      "step": 2349
+    },
+    {
+      "epoch": 0.14456200787401574,
+      "grad_norm": 1.3906610012054443,
+      "learning_rate": 0.00018993565579374425,
+      "loss": 2.6991,
+      "step": 2350
+    },
+    {
+      "epoch": 0.14462352362204725,
+      "grad_norm": 0.9641963243484497,
+      "learning_rate": 0.00018992719938864066,
+      "loss": 2.5479,
+      "step": 2351
+    },
+    {
+      "epoch": 0.14468503937007873,
+      "grad_norm": 1.018489956855774,
+      "learning_rate": 0.00018991873962076122,
+      "loss": 3.2731,
+      "step": 2352
+    },
+    {
+      "epoch": 0.14474655511811024,
+      "grad_norm": 0.9933114051818848,
+      "learning_rate": 0.00018991027649042233,
+      "loss": 2.7953,
+      "step": 2353
+    },
+    {
+      "epoch": 0.14480807086614172,
+      "grad_norm": 1.277253270149231,
+      "learning_rate": 0.00018990180999794046,
+      "loss": 2.9857,
+      "step": 2354
+    },
+    {
+      "epoch": 0.14486958661417323,
+      "grad_norm": 0.8776305913925171,
+      "learning_rate": 0.0001898933401436322,
+      "loss": 3.5892,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14493110236220472,
+      "grad_norm": 0.97346431016922,
+      "learning_rate": 0.00018988486692781429,
+      "loss": 2.9466,
+      "step": 2356
+    },
+    {
+      "epoch": 0.14499261811023623,
+      "grad_norm": 1.1422523260116577,
+      "learning_rate": 0.00018987639035080353,
+      "loss": 2.2167,
+      "step": 2357
+    },
+    {
+      "epoch": 0.1450541338582677,
+      "grad_norm": 0.944307267665863,
+      "learning_rate": 0.000189867910412917,
+      "loss": 3.2159,
+      "step": 2358
+    },
+    {
+      "epoch": 0.14511564960629922,
+      "grad_norm": 1.2940914630889893,
+      "learning_rate": 0.00018985942711447168,
+      "loss": 3.1034,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1451771653543307,
+      "grad_norm": 1.1384620666503906,
+      "learning_rate": 0.00018985094045578487,
+      "loss": 3.659,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1452386811023622,
+      "grad_norm": 1.210005521774292,
+      "learning_rate": 0.0001898424504371739,
+      "loss": 2.1531,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1453001968503937,
+      "grad_norm": 0.9526607394218445,
+      "learning_rate": 0.0001898339570589563,
+      "loss": 3.0873,
+      "step": 2362
+    },
+    {
+      "epoch": 0.1453617125984252,
+      "grad_norm": 1.1597949266433716,
+      "learning_rate": 0.0001898254603214496,
+      "loss": 3.0228,
+      "step": 2363
+    },
+    {
+      "epoch": 0.14542322834645668,
+      "grad_norm": 0.9485487341880798,
+      "learning_rate": 0.00018981696022497157,
+      "loss": 3.0265,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1454847440944882,
+      "grad_norm": 1.4213520288467407,
+      "learning_rate": 0.00018980845676984005,
+      "loss": 2.9746,
+      "step": 2365
+    },
+    {
+      "epoch": 0.14554625984251968,
+      "grad_norm": 1.7192801237106323,
+      "learning_rate": 0.00018979994995637305,
+      "loss": 3.3038,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1456077755905512,
+      "grad_norm": 1.2029234170913696,
+      "learning_rate": 0.00018979143978488865,
+      "loss": 2.6652,
+      "step": 2367
+    },
+    {
+      "epoch": 0.14566929133858267,
+      "grad_norm": 0.9722371697425842,
+      "learning_rate": 0.0001897829262557051,
+      "loss": 2.96,
+      "step": 2368
+    },
+    {
+      "epoch": 0.14573080708661418,
+      "grad_norm": 1.1632102727890015,
+      "learning_rate": 0.00018977440936914075,
+      "loss": 2.8978,
+      "step": 2369
+    },
+    {
+      "epoch": 0.14579232283464566,
+      "grad_norm": 1.304142951965332,
+      "learning_rate": 0.00018976588912551407,
+      "loss": 2.6716,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14585383858267717,
+      "grad_norm": 1.0567264556884766,
+      "learning_rate": 0.00018975736552514371,
+      "loss": 3.166,
+      "step": 2371
+    },
+    {
+      "epoch": 0.14591535433070865,
+      "grad_norm": 0.985423743724823,
+      "learning_rate": 0.00018974883856834836,
+      "loss": 3.153,
+      "step": 2372
+    },
+    {
+      "epoch": 0.14597687007874016,
+      "grad_norm": 1.113980770111084,
+      "learning_rate": 0.00018974030825544692,
+      "loss": 3.2573,
+      "step": 2373
+    },
+    {
+      "epoch": 0.14603838582677164,
+      "grad_norm": 1.0734823942184448,
+      "learning_rate": 0.00018973177458675834,
+      "loss": 2.8667,
+      "step": 2374
+    },
+    {
+      "epoch": 0.14609990157480315,
+      "grad_norm": 0.9141244888305664,
+      "learning_rate": 0.00018972323756260176,
+      "loss": 3.0624,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14616141732283464,
+      "grad_norm": 1.1814314126968384,
+      "learning_rate": 0.0001897146971832964,
+      "loss": 2.7761,
+      "step": 2376
+    },
+    {
+      "epoch": 0.14622293307086615,
+      "grad_norm": 1.0851624011993408,
+      "learning_rate": 0.00018970615344916162,
+      "loss": 3.38,
+      "step": 2377
+    },
+    {
+      "epoch": 0.14628444881889763,
+      "grad_norm": 1.1052711009979248,
+      "learning_rate": 0.00018969760636051693,
+      "loss": 3.0557,
+      "step": 2378
+    },
+    {
+      "epoch": 0.14634596456692914,
+      "grad_norm": 1.1383310556411743,
+      "learning_rate": 0.0001896890559176819,
+      "loss": 3.2634,
+      "step": 2379
+    },
+    {
+      "epoch": 0.14640748031496062,
+      "grad_norm": 1.2794373035430908,
+      "learning_rate": 0.00018968050212097635,
+      "loss": 3.0211,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14646899606299213,
+      "grad_norm": 1.0865808725357056,
+      "learning_rate": 0.00018967194497072006,
+      "loss": 3.0244,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1465305118110236,
+      "grad_norm": 1.071244478225708,
+      "learning_rate": 0.00018966338446723305,
+      "loss": 2.5696,
+      "step": 2382
+    },
+    {
+      "epoch": 0.14659202755905512,
+      "grad_norm": 1.4831295013427734,
+      "learning_rate": 0.00018965482061083545,
+      "loss": 3.4781,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1466535433070866,
+      "grad_norm": 1.1738158464431763,
+      "learning_rate": 0.00018964625340184749,
+      "loss": 2.8121,
+      "step": 2384
+    },
+    {
+      "epoch": 0.1467150590551181,
+      "grad_norm": 1.009455919265747,
+      "learning_rate": 0.0001896376828405895,
+      "loss": 2.1034,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1467765748031496,
+      "grad_norm": 1.6041427850723267,
+      "learning_rate": 0.00018962910892738206,
+      "loss": 2.5137,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1468380905511811,
+      "grad_norm": 1.4728931188583374,
+      "learning_rate": 0.0001896205316625457,
+      "loss": 2.7837,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1468996062992126,
+      "grad_norm": 1.088486909866333,
+      "learning_rate": 0.0001896119510464012,
+      "loss": 2.5483,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1469611220472441,
+      "grad_norm": 1.1723324060440063,
+      "learning_rate": 0.00018960336707926943,
+      "loss": 2.8899,
+      "step": 2389
+    },
+    {
+      "epoch": 0.14702263779527558,
+      "grad_norm": 1.2283164262771606,
+      "learning_rate": 0.00018959477976147136,
+      "loss": 2.6406,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1470841535433071,
+      "grad_norm": 1.0358058214187622,
+      "learning_rate": 0.00018958618909332812,
+      "loss": 3.1627,
+      "step": 2391
+    },
+    {
+      "epoch": 0.14714566929133857,
+      "grad_norm": 1.180301547050476,
+      "learning_rate": 0.00018957759507516096,
+      "loss": 3.3849,
+      "step": 2392
+    },
+    {
+      "epoch": 0.14720718503937008,
+      "grad_norm": 0.9963234663009644,
+      "learning_rate": 0.00018956899770729125,
+      "loss": 3.1398,
+      "step": 2393
+    },
+    {
+      "epoch": 0.14726870078740156,
+      "grad_norm": 1.1118930578231812,
+      "learning_rate": 0.00018956039699004047,
+      "loss": 3.3189,
+      "step": 2394
+    },
+    {
+      "epoch": 0.14733021653543307,
+      "grad_norm": 1.0580776929855347,
+      "learning_rate": 0.00018955179292373022,
+      "loss": 2.6606,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14739173228346455,
+      "grad_norm": 1.1711106300354004,
+      "learning_rate": 0.0001895431855086823,
+      "loss": 2.573,
+      "step": 2396
+    },
+    {
+      "epoch": 0.14745324803149606,
+      "grad_norm": 1.0724512338638306,
+      "learning_rate": 0.00018953457474521853,
+      "loss": 2.8524,
+      "step": 2397
+    },
+    {
+      "epoch": 0.14751476377952755,
+      "grad_norm": 1.279122233390808,
+      "learning_rate": 0.00018952596063366096,
+      "loss": 3.4181,
+      "step": 2398
+    },
+    {
+      "epoch": 0.14757627952755906,
+      "grad_norm": 1.0756093263626099,
+      "learning_rate": 0.00018951734317433164,
+      "loss": 2.1862,
+      "step": 2399
+    },
+    {
+      "epoch": 0.14763779527559054,
+      "grad_norm": 1.4737112522125244,
+      "learning_rate": 0.00018950872236755286,
+      "loss": 2.9704,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14769931102362205,
+      "grad_norm": 1.276285171508789,
+      "learning_rate": 0.00018950009821364696,
+      "loss": 3.0188,
+      "step": 2401
+    },
+    {
+      "epoch": 0.14776082677165353,
+      "grad_norm": 1.1166036128997803,
+      "learning_rate": 0.00018949147071293645,
+      "loss": 2.5852,
+      "step": 2402
+    },
+    {
+      "epoch": 0.14782234251968504,
+      "grad_norm": 1.1757376194000244,
+      "learning_rate": 0.00018948283986574397,
+      "loss": 3.5226,
+      "step": 2403
+    },
+    {
+      "epoch": 0.14788385826771652,
+      "grad_norm": 1.3590823411941528,
+      "learning_rate": 0.00018947420567239225,
+      "loss": 2.562,
+      "step": 2404
+    },
+    {
+      "epoch": 0.14794537401574803,
+      "grad_norm": 1.3869441747665405,
+      "learning_rate": 0.00018946556813320418,
+      "loss": 2.0725,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14800688976377951,
+      "grad_norm": 1.0407825708389282,
+      "learning_rate": 0.00018945692724850268,
+      "loss": 3.041,
+      "step": 2406
+    },
+    {
+      "epoch": 0.14806840551181102,
+      "grad_norm": 1.1123112440109253,
+      "learning_rate": 0.00018944828301861098,
+      "loss": 3.2717,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1481299212598425,
+      "grad_norm": 1.2548041343688965,
+      "learning_rate": 0.00018943963544385223,
+      "loss": 2.9347,
+      "step": 2408
+    },
+    {
+      "epoch": 0.14819143700787402,
+      "grad_norm": 1.0621354579925537,
+      "learning_rate": 0.0001894309845245499,
+      "loss": 3.3581,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1482529527559055,
+      "grad_norm": 1.0142029523849487,
+      "learning_rate": 0.00018942233026102736,
+      "loss": 3.1013,
+      "step": 2410
+    },
+    {
+      "epoch": 0.148314468503937,
+      "grad_norm": 1.3148351907730103,
+      "learning_rate": 0.00018941367265360833,
+      "loss": 1.852,
+      "step": 2411
+    },
+    {
+      "epoch": 0.1483759842519685,
+      "grad_norm": 1.1304574012756348,
+      "learning_rate": 0.00018940501170261654,
+      "loss": 3.6758,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1484375,
+      "grad_norm": 1.1277211904525757,
+      "learning_rate": 0.00018939634740837584,
+      "loss": 3.1139,
+      "step": 2413
+    },
+    {
+      "epoch": 0.1484990157480315,
+      "grad_norm": 1.0786327123641968,
+      "learning_rate": 0.00018938767977121022,
+      "loss": 3.0209,
+      "step": 2414
+    },
+    {
+      "epoch": 0.148560531496063,
+      "grad_norm": 1.0200387239456177,
+      "learning_rate": 0.00018937900879144385,
+      "loss": 3.4007,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1486220472440945,
+      "grad_norm": 1.0331919193267822,
+      "learning_rate": 0.00018937033446940094,
+      "loss": 3.4037,
+      "step": 2416
+    },
+    {
+      "epoch": 0.14868356299212598,
+      "grad_norm": 1.2546762228012085,
+      "learning_rate": 0.00018936165680540583,
+      "loss": 3.8149,
+      "step": 2417
+    },
+    {
+      "epoch": 0.1487450787401575,
+      "grad_norm": 1.228763222694397,
+      "learning_rate": 0.00018935297579978307,
+      "loss": 2.7726,
+      "step": 2418
+    },
+    {
+      "epoch": 0.14880659448818898,
+      "grad_norm": 1.0119963884353638,
+      "learning_rate": 0.00018934429145285725,
+      "loss": 2.8846,
+      "step": 2419
+    },
+    {
+      "epoch": 0.14886811023622049,
+      "grad_norm": 1.0856642723083496,
+      "learning_rate": 0.00018933560376495313,
+      "loss": 2.4362,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14892962598425197,
+      "grad_norm": 1.0420386791229248,
+      "learning_rate": 0.0001893269127363956,
+      "loss": 3.0934,
+      "step": 2421
+    },
+    {
+      "epoch": 0.14899114173228348,
+      "grad_norm": 1.0476382970809937,
+      "learning_rate": 0.0001893182183675096,
+      "loss": 2.7801,
+      "step": 2422
+    },
+    {
+      "epoch": 0.14905265748031496,
+      "grad_norm": 0.9357470273971558,
+      "learning_rate": 0.0001893095206586203,
+      "loss": 3.0591,
+      "step": 2423
+    },
+    {
+      "epoch": 0.14911417322834647,
+      "grad_norm": 1.0776855945587158,
+      "learning_rate": 0.00018930081961005292,
+      "loss": 3.3324,
+      "step": 2424
+    },
+    {
+      "epoch": 0.14917568897637795,
+      "grad_norm": 0.9032332897186279,
+      "learning_rate": 0.0001892921152221329,
+      "loss": 2.2427,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14923720472440946,
+      "grad_norm": 0.9162856340408325,
+      "learning_rate": 0.0001892834074951856,
+      "loss": 2.8925,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14929872047244094,
+      "grad_norm": 0.9112592935562134,
+      "learning_rate": 0.00018927469642953675,
+      "loss": 2.5155,
+      "step": 2427
+    },
+    {
+      "epoch": 0.14936023622047245,
+      "grad_norm": 1.1308245658874512,
+      "learning_rate": 0.00018926598202551205,
+      "loss": 2.8622,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14942175196850394,
+      "grad_norm": 1.0939406156539917,
+      "learning_rate": 0.00018925726428343741,
+      "loss": 2.8406,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14948326771653545,
+      "grad_norm": 1.2654662132263184,
+      "learning_rate": 0.0001892485432036388,
+      "loss": 3.2263,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14954478346456693,
+      "grad_norm": 1.1614259481430054,
+      "learning_rate": 0.0001892398187864423,
+      "loss": 3.2203,
+      "step": 2431
+    },
+    {
+      "epoch": 0.14960629921259844,
+      "grad_norm": 1.2707462310791016,
+      "learning_rate": 0.00018923109103217422,
+      "loss": 2.615,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14966781496062992,
+      "grad_norm": 1.0664026737213135,
+      "learning_rate": 0.0001892223599411609,
+      "loss": 3.1636,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14972933070866143,
+      "grad_norm": 1.0653530359268188,
+      "learning_rate": 0.00018921362551372882,
+      "loss": 2.7969,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1497908464566929,
+      "grad_norm": 1.1676580905914307,
+      "learning_rate": 0.00018920488775020462,
+      "loss": 2.9289,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14985236220472442,
+      "grad_norm": 1.281968116760254,
+      "learning_rate": 0.00018919614665091503,
+      "loss": 3.2239,
+      "step": 2436
+    },
+    {
+      "epoch": 0.1499138779527559,
+      "grad_norm": 1.8822236061096191,
+      "learning_rate": 0.00018918740221618693,
+      "loss": 1.8324,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1499753937007874,
+      "grad_norm": 1.2456446886062622,
+      "learning_rate": 0.00018917865444634732,
+      "loss": 3.2012,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1500369094488189,
+      "grad_norm": 0.938025176525116,
+      "learning_rate": 0.00018916990334172333,
+      "loss": 3.601,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1500984251968504,
+      "grad_norm": 1.164123296737671,
+      "learning_rate": 0.00018916114890264214,
+      "loss": 2.2346,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1501599409448819,
+      "grad_norm": 0.9138572812080383,
+      "learning_rate": 0.0001891523911294312,
+      "loss": 3.1167,
+      "step": 2441
+    },
+    {
+      "epoch": 0.1502214566929134,
+      "grad_norm": 1.2211456298828125,
+      "learning_rate": 0.0001891436300224179,
+      "loss": 2.3602,
+      "step": 2442
+    },
+    {
+      "epoch": 0.15028297244094488,
+      "grad_norm": 1.2358133792877197,
+      "learning_rate": 0.00018913486558192995,
+      "loss": 2.8189,
+      "step": 2443
+    },
+    {
+      "epoch": 0.1503444881889764,
+      "grad_norm": 1.0589985847473145,
+      "learning_rate": 0.00018912609780829506,
+      "loss": 3.1811,
+      "step": 2444
+    },
+    {
+      "epoch": 0.15040600393700787,
+      "grad_norm": 1.179186224937439,
+      "learning_rate": 0.0001891173267018411,
+      "loss": 2.4432,
+      "step": 2445
+    },
+    {
+      "epoch": 0.15046751968503938,
+      "grad_norm": 1.2260984182357788,
+      "learning_rate": 0.00018910855226289606,
+      "loss": 3.1854,
+      "step": 2446
+    },
+    {
+      "epoch": 0.15052903543307086,
+      "grad_norm": 0.8841168284416199,
+      "learning_rate": 0.00018909977449178803,
+      "loss": 3.0336,
+      "step": 2447
+    },
+    {
+      "epoch": 0.15059055118110237,
+      "grad_norm": 0.985716700553894,
+      "learning_rate": 0.00018909099338884527,
+      "loss": 1.9374,
+      "step": 2448
+    },
+    {
+      "epoch": 0.15065206692913385,
+      "grad_norm": 1.074365258216858,
+      "learning_rate": 0.00018908220895439613,
+      "loss": 3.4698,
+      "step": 2449
+    },
+    {
+      "epoch": 0.15071358267716536,
+      "grad_norm": 1.0309442281723022,
+      "learning_rate": 0.0001890734211887691,
+      "loss": 2.9151,
+      "step": 2450
+    },
+    {
+      "epoch": 0.15077509842519685,
+      "grad_norm": 1.409875512123108,
+      "learning_rate": 0.00018906463009229284,
+      "loss": 2.9296,
+      "step": 2451
+    },
+    {
+      "epoch": 0.15083661417322836,
+      "grad_norm": 1.2563586235046387,
+      "learning_rate": 0.00018905583566529603,
+      "loss": 2.7689,
+      "step": 2452
+    },
+    {
+      "epoch": 0.15089812992125984,
+      "grad_norm": 1.2409751415252686,
+      "learning_rate": 0.00018904703790810756,
+      "loss": 2.9884,
+      "step": 2453
+    },
+    {
+      "epoch": 0.15095964566929135,
+      "grad_norm": 1.0985852479934692,
+      "learning_rate": 0.0001890382368210564,
+      "loss": 2.9373,
+      "step": 2454
+    },
+    {
+      "epoch": 0.15102116141732283,
+      "grad_norm": 1.1309326887130737,
+      "learning_rate": 0.00018902943240447172,
+      "loss": 2.895,
+      "step": 2455
+    },
+    {
+      "epoch": 0.15108267716535434,
+      "grad_norm": 0.9509692192077637,
+      "learning_rate": 0.00018902062465868267,
+      "loss": 3.2362,
+      "step": 2456
+    },
+    {
+      "epoch": 0.15114419291338582,
+      "grad_norm": 1.2865314483642578,
+      "learning_rate": 0.00018901181358401865,
+      "loss": 3.3974,
+      "step": 2457
+    },
+    {
+      "epoch": 0.15120570866141733,
+      "grad_norm": 1.0868713855743408,
+      "learning_rate": 0.00018900299918080915,
+      "loss": 2.7351,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1512672244094488,
+      "grad_norm": 1.1678780317306519,
+      "learning_rate": 0.00018899418144938378,
+      "loss": 2.9368,
+      "step": 2459
+    },
+    {
+      "epoch": 0.15132874015748032,
+      "grad_norm": 1.233700156211853,
+      "learning_rate": 0.0001889853603900723,
+      "loss": 3.1413,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1513902559055118,
+      "grad_norm": 0.8344703912734985,
+      "learning_rate": 0.00018897653600320448,
+      "loss": 3.318,
+      "step": 2461
+    },
+    {
+      "epoch": 0.15145177165354332,
+      "grad_norm": 1.25886070728302,
+      "learning_rate": 0.0001889677082891104,
+      "loss": 2.6765,
+      "step": 2462
+    },
+    {
+      "epoch": 0.1515132874015748,
+      "grad_norm": 1.5516760349273682,
+      "learning_rate": 0.00018895887724812012,
+      "loss": 3.8068,
+      "step": 2463
+    },
+    {
+      "epoch": 0.1515748031496063,
+      "grad_norm": 1.0216679573059082,
+      "learning_rate": 0.00018895004288056386,
+      "loss": 2.0399,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1516363188976378,
+      "grad_norm": 1.1078599691390991,
+      "learning_rate": 0.000188941205186772,
+      "loss": 2.3883,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1516978346456693,
+      "grad_norm": 0.8893657326698303,
+      "learning_rate": 0.00018893236416707505,
+      "loss": 2.4679,
+      "step": 2466
+    },
+    {
+      "epoch": 0.15175935039370078,
+      "grad_norm": 1.0233559608459473,
+      "learning_rate": 0.00018892351982180354,
+      "loss": 3.4308,
+      "step": 2467
+    },
+    {
+      "epoch": 0.1518208661417323,
+      "grad_norm": 0.9810418486595154,
+      "learning_rate": 0.00018891467215128826,
+      "loss": 3.2923,
+      "step": 2468
+    },
+    {
+      "epoch": 0.15188238188976377,
+      "grad_norm": 1.0622791051864624,
+      "learning_rate": 0.00018890582115586003,
+      "loss": 2.4507,
+      "step": 2469
+    },
+    {
+      "epoch": 0.15194389763779528,
+      "grad_norm": 1.2799025774002075,
+      "learning_rate": 0.00018889696683584985,
+      "loss": 2.6078,
+      "step": 2470
+    },
+    {
+      "epoch": 0.15200541338582677,
+      "grad_norm": 1.3285022974014282,
+      "learning_rate": 0.0001888881091915888,
+      "loss": 2.8204,
+      "step": 2471
+    },
+    {
+      "epoch": 0.15206692913385828,
+      "grad_norm": 1.136087417602539,
+      "learning_rate": 0.00018887924822340814,
+      "loss": 2.7682,
+      "step": 2472
+    },
+    {
+      "epoch": 0.15212844488188976,
+      "grad_norm": 0.9461902976036072,
+      "learning_rate": 0.0001888703839316392,
+      "loss": 3.6224,
+      "step": 2473
+    },
+    {
+      "epoch": 0.15218996062992127,
+      "grad_norm": 1.1105965375900269,
+      "learning_rate": 0.00018886151631661345,
+      "loss": 2.7496,
+      "step": 2474
+    },
+    {
+      "epoch": 0.15225147637795275,
+      "grad_norm": 1.108556866645813,
+      "learning_rate": 0.0001888526453786625,
+      "loss": 3.0452,
+      "step": 2475
+    },
+    {
+      "epoch": 0.15231299212598426,
+      "grad_norm": 0.9500887393951416,
+      "learning_rate": 0.00018884377111811804,
+      "loss": 3.1868,
+      "step": 2476
+    },
+    {
+      "epoch": 0.15237450787401574,
+      "grad_norm": 1.1857069730758667,
+      "learning_rate": 0.00018883489353531199,
+      "loss": 2.9484,
+      "step": 2477
+    },
+    {
+      "epoch": 0.15243602362204725,
+      "grad_norm": 1.29253351688385,
+      "learning_rate": 0.00018882601263057626,
+      "loss": 3.249,
+      "step": 2478
+    },
+    {
+      "epoch": 0.15249753937007873,
+      "grad_norm": 1.2845990657806396,
+      "learning_rate": 0.00018881712840424295,
+      "loss": 2.9015,
+      "step": 2479
+    },
+    {
+      "epoch": 0.15255905511811024,
+      "grad_norm": 1.096891164779663,
+      "learning_rate": 0.00018880824085664433,
+      "loss": 3.0108,
+      "step": 2480
+    },
+    {
+      "epoch": 0.15262057086614172,
+      "grad_norm": 1.402283787727356,
+      "learning_rate": 0.00018879934998811265,
+      "loss": 2.8077,
+      "step": 2481
+    },
+    {
+      "epoch": 0.15268208661417323,
+      "grad_norm": 1.0419037342071533,
+      "learning_rate": 0.0001887904557989805,
+      "loss": 3.3282,
+      "step": 2482
+    },
+    {
+      "epoch": 0.15274360236220472,
+      "grad_norm": 1.116349220275879,
+      "learning_rate": 0.0001887815582895804,
+      "loss": 2.6899,
+      "step": 2483
+    },
+    {
+      "epoch": 0.15280511811023623,
+      "grad_norm": 1.1809985637664795,
+      "learning_rate": 0.00018877265746024507,
+      "loss": 2.0705,
+      "step": 2484
+    },
+    {
+      "epoch": 0.1528666338582677,
+      "grad_norm": 1.0850871801376343,
+      "learning_rate": 0.00018876375331130734,
+      "loss": 2.4094,
+      "step": 2485
+    },
+    {
+      "epoch": 0.15292814960629922,
+      "grad_norm": 1.195765733718872,
+      "learning_rate": 0.00018875484584310025,
+      "loss": 2.8479,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1529896653543307,
+      "grad_norm": 1.0443087816238403,
+      "learning_rate": 0.0001887459350559568,
+      "loss": 3.4191,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1530511811023622,
+      "grad_norm": 1.2100789546966553,
+      "learning_rate": 0.00018873702095021023,
+      "loss": 3.0054,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1531126968503937,
+      "grad_norm": 1.0217691659927368,
+      "learning_rate": 0.0001887281035261939,
+      "loss": 3.6064,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1531742125984252,
+      "grad_norm": 1.0724537372589111,
+      "learning_rate": 0.00018871918278424125,
+      "loss": 2.4732,
+      "step": 2490
+    },
+    {
+      "epoch": 0.15323572834645668,
+      "grad_norm": 1.0309791564941406,
+      "learning_rate": 0.00018871025872468588,
+      "loss": 3.0747,
+      "step": 2491
+    },
+    {
+      "epoch": 0.1532972440944882,
+      "grad_norm": 1.0429619550704956,
+      "learning_rate": 0.00018870133134786152,
+      "loss": 2.642,
+      "step": 2492
+    },
+    {
+      "epoch": 0.15335875984251968,
+      "grad_norm": 1.255984902381897,
+      "learning_rate": 0.00018869240065410195,
+      "loss": 3.0802,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1534202755905512,
+      "grad_norm": 1.6746951341629028,
+      "learning_rate": 0.00018868346664374118,
+      "loss": 2.5153,
+      "step": 2494
+    },
+    {
+      "epoch": 0.15348179133858267,
+      "grad_norm": 0.966847836971283,
+      "learning_rate": 0.00018867452931711326,
+      "loss": 3.3085,
+      "step": 2495
+    },
+    {
+      "epoch": 0.15354330708661418,
+      "grad_norm": 1.048776626586914,
+      "learning_rate": 0.0001886655886745524,
+      "loss": 3.29,
+      "step": 2496
+    },
+    {
+      "epoch": 0.15360482283464566,
+      "grad_norm": 1.1062849760055542,
+      "learning_rate": 0.00018865664471639292,
+      "loss": 3.2607,
+      "step": 2497
+    },
+    {
+      "epoch": 0.15366633858267717,
+      "grad_norm": 0.9726073741912842,
+      "learning_rate": 0.00018864769744296936,
+      "loss": 3.1538,
+      "step": 2498
+    },
+    {
+      "epoch": 0.15372785433070865,
+      "grad_norm": 1.7598234415054321,
+      "learning_rate": 0.00018863874685461616,
+      "loss": 2.5281,
+      "step": 2499
+    },
+    {
+      "epoch": 0.15378937007874016,
+      "grad_norm": 1.1470472812652588,
+      "learning_rate": 0.00018862979295166809,
+      "loss": 3.2594,
+      "step": 2500
+    },
+    {
+      "epoch": 0.15385088582677164,
+      "grad_norm": 1.4921915531158447,
+      "learning_rate": 0.00018862083573446,
+      "loss": 2.418,
+      "step": 2501
+    },
+    {
+      "epoch": 0.15391240157480315,
+      "grad_norm": 1.1320366859436035,
+      "learning_rate": 0.00018861187520332684,
+      "loss": 2.8594,
+      "step": 2502
+    },
+    {
+      "epoch": 0.15397391732283464,
+      "grad_norm": 1.0782604217529297,
+      "learning_rate": 0.00018860291135860364,
+      "loss": 2.7213,
+      "step": 2503
+    },
+    {
+      "epoch": 0.15403543307086615,
+      "grad_norm": 1.564475178718567,
+      "learning_rate": 0.00018859394420062562,
+      "loss": 3.3364,
+      "step": 2504
+    },
+    {
+      "epoch": 0.15409694881889763,
+      "grad_norm": 1.4476138353347778,
+      "learning_rate": 0.0001885849737297281,
+      "loss": 2.5129,
+      "step": 2505
+    },
+    {
+      "epoch": 0.15415846456692914,
+      "grad_norm": 1.1947323083877563,
+      "learning_rate": 0.0001885759999462465,
+      "loss": 3.5789,
+      "step": 2506
+    },
+    {
+      "epoch": 0.15421998031496062,
+      "grad_norm": 1.0495460033416748,
+      "learning_rate": 0.00018856702285051643,
+      "loss": 2.8559,
+      "step": 2507
+    },
+    {
+      "epoch": 0.15428149606299213,
+      "grad_norm": 1.069668173789978,
+      "learning_rate": 0.0001885580424428736,
+      "loss": 3.1556,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1543430118110236,
+      "grad_norm": 1.0805362462997437,
+      "learning_rate": 0.00018854905872365375,
+      "loss": 3.3929,
+      "step": 2509
+    },
+    {
+      "epoch": 0.15440452755905512,
+      "grad_norm": 0.8520386815071106,
+      "learning_rate": 0.0001885400716931929,
+      "loss": 3.5758,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1544660433070866,
+      "grad_norm": 1.0585097074508667,
+      "learning_rate": 0.00018853108135182708,
+      "loss": 2.6767,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1545275590551181,
+      "grad_norm": 1.2130242586135864,
+      "learning_rate": 0.00018852208769989246,
+      "loss": 2.554,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1545890748031496,
+      "grad_norm": 1.1863908767700195,
+      "learning_rate": 0.00018851309073772537,
+      "loss": 2.5405,
+      "step": 2513
+    },
+    {
+      "epoch": 0.1546505905511811,
+      "grad_norm": 0.9951097369194031,
+      "learning_rate": 0.00018850409046566226,
+      "loss": 3.1073,
+      "step": 2514
+    },
+    {
+      "epoch": 0.1547121062992126,
+      "grad_norm": 0.9675514698028564,
+      "learning_rate": 0.0001884950868840397,
+      "loss": 2.7888,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1547736220472441,
+      "grad_norm": 1.0467935800552368,
+      "learning_rate": 0.00018848607999319433,
+      "loss": 3.0279,
+      "step": 2516
+    },
+    {
+      "epoch": 0.15483513779527558,
+      "grad_norm": 1.1126136779785156,
+      "learning_rate": 0.000188477069793463,
+      "loss": 3.068,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1548966535433071,
+      "grad_norm": 1.115752935409546,
+      "learning_rate": 0.0001884680562851826,
+      "loss": 3.2403,
+      "step": 2518
+    },
+    {
+      "epoch": 0.15495816929133857,
+      "grad_norm": 1.1166210174560547,
+      "learning_rate": 0.0001884590394686902,
+      "loss": 2.8397,
+      "step": 2519
+    },
+    {
+      "epoch": 0.15501968503937008,
+      "grad_norm": 0.9159601330757141,
+      "learning_rate": 0.000188450019344323,
+      "loss": 3.0539,
+      "step": 2520
+    },
+    {
+      "epoch": 0.15508120078740156,
+      "grad_norm": 1.0771663188934326,
+      "learning_rate": 0.0001884409959124183,
+      "loss": 2.5781,
+      "step": 2521
+    },
+    {
+      "epoch": 0.15514271653543307,
+      "grad_norm": 1.0342469215393066,
+      "learning_rate": 0.0001884319691733135,
+      "loss": 3.0049,
+      "step": 2522
+    },
+    {
+      "epoch": 0.15520423228346455,
+      "grad_norm": 1.189313292503357,
+      "learning_rate": 0.00018842293912734614,
+      "loss": 3.5678,
+      "step": 2523
+    },
+    {
+      "epoch": 0.15526574803149606,
+      "grad_norm": 1.291243553161621,
+      "learning_rate": 0.00018841390577485392,
+      "loss": 3.4108,
+      "step": 2524
+    },
+    {
+      "epoch": 0.15532726377952755,
+      "grad_norm": 0.9366981983184814,
+      "learning_rate": 0.00018840486911617465,
+      "loss": 3.5328,
+      "step": 2525
+    },
+    {
+      "epoch": 0.15538877952755906,
+      "grad_norm": 1.0716155767440796,
+      "learning_rate": 0.00018839582915164624,
+      "loss": 3.1368,
+      "step": 2526
+    },
+    {
+      "epoch": 0.15545029527559054,
+      "grad_norm": 0.8434606194496155,
+      "learning_rate": 0.00018838678588160672,
+      "loss": 3.3813,
+      "step": 2527
+    },
+    {
+      "epoch": 0.15551181102362205,
+      "grad_norm": 1.0792235136032104,
+      "learning_rate": 0.00018837773930639428,
+      "loss": 2.7557,
+      "step": 2528
+    },
+    {
+      "epoch": 0.15557332677165353,
+      "grad_norm": 1.2072800397872925,
+      "learning_rate": 0.00018836868942634718,
+      "loss": 3.31,
+      "step": 2529
+    },
+    {
+      "epoch": 0.15563484251968504,
+      "grad_norm": 1.0140318870544434,
+      "learning_rate": 0.00018835963624180385,
+      "loss": 3.1315,
+      "step": 2530
+    },
+    {
+      "epoch": 0.15569635826771652,
+      "grad_norm": 2.079460859298706,
+      "learning_rate": 0.00018835057975310286,
+      "loss": 3.2048,
+      "step": 2531
+    },
+    {
+      "epoch": 0.15575787401574803,
+      "grad_norm": 1.06719172000885,
+      "learning_rate": 0.00018834151996058283,
+      "loss": 3.038,
+      "step": 2532
+    },
+    {
+      "epoch": 0.15581938976377951,
+      "grad_norm": 1.0795239210128784,
+      "learning_rate": 0.00018833245686458255,
+      "loss": 2.7969,
+      "step": 2533
+    },
+    {
+      "epoch": 0.15588090551181102,
+      "grad_norm": 1.2871780395507812,
+      "learning_rate": 0.00018832339046544094,
+      "loss": 2.7956,
+      "step": 2534
+    },
+    {
+      "epoch": 0.1559424212598425,
+      "grad_norm": 1.2736550569534302,
+      "learning_rate": 0.000188314320763497,
+      "loss": 3.0828,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15600393700787402,
+      "grad_norm": 1.4442577362060547,
+      "learning_rate": 0.00018830524775908998,
+      "loss": 2.3721,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1560654527559055,
+      "grad_norm": 0.8568169474601746,
+      "learning_rate": 0.00018829617145255906,
+      "loss": 3.3877,
+      "step": 2537
+    },
+    {
+      "epoch": 0.156126968503937,
+      "grad_norm": 1.1411930322647095,
+      "learning_rate": 0.0001882870918442437,
+      "loss": 3.3135,
+      "step": 2538
+    },
+    {
+      "epoch": 0.1561884842519685,
+      "grad_norm": 1.0186865329742432,
+      "learning_rate": 0.0001882780089344834,
+      "loss": 3.0841,
+      "step": 2539
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 1.0038394927978516,
+      "learning_rate": 0.0001882689227236178,
+      "loss": 3.4625,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1563115157480315,
+      "grad_norm": 1.003989577293396,
+      "learning_rate": 0.0001882598332119867,
+      "loss": 2.9372,
+      "step": 2541
+    },
+    {
+      "epoch": 0.156373031496063,
+      "grad_norm": 1.1325857639312744,
+      "learning_rate": 0.00018825074039993003,
+      "loss": 2.7898,
+      "step": 2542
+    },
+    {
+      "epoch": 0.1564345472440945,
+      "grad_norm": 1.1955678462982178,
+      "learning_rate": 0.00018824164428778771,
+      "loss": 2.4494,
+      "step": 2543
+    },
+    {
+      "epoch": 0.15649606299212598,
+      "grad_norm": 1.0180859565734863,
+      "learning_rate": 0.00018823254487589997,
+      "loss": 3.4086,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1565575787401575,
+      "grad_norm": 1.2471017837524414,
+      "learning_rate": 0.00018822344216460704,
+      "loss": 3.2772,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15661909448818898,
+      "grad_norm": 1.466684103012085,
+      "learning_rate": 0.00018821433615424932,
+      "loss": 2.845,
+      "step": 2546
+    },
+    {
+      "epoch": 0.15668061023622049,
+      "grad_norm": 1.0858548879623413,
+      "learning_rate": 0.00018820522684516732,
+      "loss": 2.5548,
+      "step": 2547
+    },
+    {
+      "epoch": 0.15674212598425197,
+      "grad_norm": 1.124759316444397,
+      "learning_rate": 0.00018819611423770172,
+      "loss": 3.5497,
+      "step": 2548
+    },
+    {
+      "epoch": 0.15680364173228348,
+      "grad_norm": 1.2228254079818726,
+      "learning_rate": 0.0001881869983321932,
+      "loss": 3.6576,
+      "step": 2549
+    },
+    {
+      "epoch": 0.15686515748031496,
+      "grad_norm": 1.219349980354309,
+      "learning_rate": 0.0001881778791289827,
+      "loss": 3.3582,
+      "step": 2550
+    },
+    {
+      "epoch": 0.15692667322834647,
+      "grad_norm": 1.2225314378738403,
+      "learning_rate": 0.00018816875662841122,
+      "loss": 2.9545,
+      "step": 2551
+    },
+    {
+      "epoch": 0.15698818897637795,
+      "grad_norm": 1.0091525316238403,
+      "learning_rate": 0.00018815963083081987,
+      "loss": 3.3279,
+      "step": 2552
+    },
+    {
+      "epoch": 0.15704970472440946,
+      "grad_norm": 1.0566068887710571,
+      "learning_rate": 0.00018815050173654992,
+      "loss": 2.8859,
+      "step": 2553
+    },
+    {
+      "epoch": 0.15711122047244094,
+      "grad_norm": 0.9198455214500427,
+      "learning_rate": 0.00018814136934594276,
+      "loss": 3.3401,
+      "step": 2554
+    },
+    {
+      "epoch": 0.15717273622047245,
+      "grad_norm": 0.9795255661010742,
+      "learning_rate": 0.00018813223365933988,
+      "loss": 3.0024,
+      "step": 2555
+    },
+    {
+      "epoch": 0.15723425196850394,
+      "grad_norm": 1.1123073101043701,
+      "learning_rate": 0.00018812309467708288,
+      "loss": 2.7677,
+      "step": 2556
+    },
+    {
+      "epoch": 0.15729576771653545,
+      "grad_norm": 1.1557021141052246,
+      "learning_rate": 0.00018811395239951354,
+      "loss": 2.9502,
+      "step": 2557
+    },
+    {
+      "epoch": 0.15735728346456693,
+      "grad_norm": 1.1244525909423828,
+      "learning_rate": 0.00018810480682697372,
+      "loss": 2.5127,
+      "step": 2558
+    },
+    {
+      "epoch": 0.15741879921259844,
+      "grad_norm": 1.349189281463623,
+      "learning_rate": 0.0001880956579598054,
+      "loss": 2.7031,
+      "step": 2559
+    },
+    {
+      "epoch": 0.15748031496062992,
+      "grad_norm": 1.2928032875061035,
+      "learning_rate": 0.0001880865057983507,
+      "loss": 3.2406,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15754183070866143,
+      "grad_norm": 0.8902611136436462,
+      "learning_rate": 0.0001880773503429519,
+      "loss": 3.3896,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1576033464566929,
+      "grad_norm": 0.9754341244697571,
+      "learning_rate": 0.0001880681915939513,
+      "loss": 3.0796,
+      "step": 2562
+    },
+    {
+      "epoch": 0.15766486220472442,
+      "grad_norm": 1.0898959636688232,
+      "learning_rate": 0.00018805902955169144,
+      "loss": 2.6633,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1577263779527559,
+      "grad_norm": 2.141345262527466,
+      "learning_rate": 0.00018804986421651487,
+      "loss": 3.5195,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1577878937007874,
+      "grad_norm": 1.607120394706726,
+      "learning_rate": 0.00018804069558876437,
+      "loss": 2.4416,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1578494094488189,
+      "grad_norm": 1.2776708602905273,
+      "learning_rate": 0.0001880315236687828,
+      "loss": 2.7084,
+      "step": 2566
+    },
+    {
+      "epoch": 0.1579109251968504,
+      "grad_norm": 1.2368637323379517,
+      "learning_rate": 0.0001880223484569131,
+      "loss": 2.0289,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1579724409448819,
+      "grad_norm": 0.9923470616340637,
+      "learning_rate": 0.00018801316995349844,
+      "loss": 2.6391,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1580339566929134,
+      "grad_norm": 1.1390302181243896,
+      "learning_rate": 0.00018800398815888193,
+      "loss": 3.0906,
+      "step": 2569
+    },
+    {
+      "epoch": 0.15809547244094488,
+      "grad_norm": 1.6636477708816528,
+      "learning_rate": 0.000187994803073407,
+      "loss": 2.2798,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1581569881889764,
+      "grad_norm": 1.4182257652282715,
+      "learning_rate": 0.00018798561469741715,
+      "loss": 2.8945,
+      "step": 2571
+    },
+    {
+      "epoch": 0.15821850393700787,
+      "grad_norm": 1.1042269468307495,
+      "learning_rate": 0.0001879764230312559,
+      "loss": 2.9305,
+      "step": 2572
+    },
+    {
+      "epoch": 0.15828001968503938,
+      "grad_norm": 1.934494972229004,
+      "learning_rate": 0.00018796722807526704,
+      "loss": 2.6461,
+      "step": 2573
+    },
+    {
+      "epoch": 0.15834153543307086,
+      "grad_norm": 1.2007883787155151,
+      "learning_rate": 0.00018795802982979435,
+      "loss": 3.6248,
+      "step": 2574
+    },
+    {
+      "epoch": 0.15840305118110237,
+      "grad_norm": 1.0006004571914673,
+      "learning_rate": 0.0001879488282951818,
+      "loss": 2.779,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15846456692913385,
+      "grad_norm": 1.0485436916351318,
+      "learning_rate": 0.00018793962347177348,
+      "loss": 2.7623,
+      "step": 2576
+    },
+    {
+      "epoch": 0.15852608267716536,
+      "grad_norm": 0.9739270806312561,
+      "learning_rate": 0.0001879304153599136,
+      "loss": 2.8914,
+      "step": 2577
+    },
+    {
+      "epoch": 0.15858759842519685,
+      "grad_norm": 1.3080217838287354,
+      "learning_rate": 0.0001879212039599465,
+      "loss": 2.8672,
+      "step": 2578
+    },
+    {
+      "epoch": 0.15864911417322836,
+      "grad_norm": 0.915787935256958,
+      "learning_rate": 0.00018791198927221666,
+      "loss": 3.2149,
+      "step": 2579
+    },
+    {
+      "epoch": 0.15871062992125984,
+      "grad_norm": 1.8920954465866089,
+      "learning_rate": 0.00018790277129706863,
+      "loss": 3.146,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15877214566929135,
+      "grad_norm": 1.8903926610946655,
+      "learning_rate": 0.0001878935500348471,
+      "loss": 0.9337,
+      "step": 2581
+    },
+    {
+      "epoch": 0.15883366141732283,
+      "grad_norm": 1.1665873527526855,
+      "learning_rate": 0.0001878843254858969,
+      "loss": 3.088,
+      "step": 2582
+    },
+    {
+      "epoch": 0.15889517716535434,
+      "grad_norm": 1.4480130672454834,
+      "learning_rate": 0.00018787509765056298,
+      "loss": 3.1926,
+      "step": 2583
+    },
+    {
+      "epoch": 0.15895669291338582,
+      "grad_norm": 0.9289477467536926,
+      "learning_rate": 0.00018786586652919045,
+      "loss": 3.4044,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15901820866141733,
+      "grad_norm": 1.2378056049346924,
+      "learning_rate": 0.00018785663212212445,
+      "loss": 3.0402,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1590797244094488,
+      "grad_norm": 1.0635478496551514,
+      "learning_rate": 0.0001878473944297103,
+      "loss": 3.1505,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15914124015748032,
+      "grad_norm": 1.1688838005065918,
+      "learning_rate": 0.00018783815345229342,
+      "loss": 2.9389,
+      "step": 2587
+    },
+    {
+      "epoch": 0.1592027559055118,
+      "grad_norm": 1.1395739316940308,
+      "learning_rate": 0.00018782890919021944,
+      "loss": 2.5709,
+      "step": 2588
+    },
+    {
+      "epoch": 0.15926427165354332,
+      "grad_norm": 1.0229625701904297,
+      "learning_rate": 0.000187819661643834,
+      "loss": 2.9451,
+      "step": 2589
+    },
+    {
+      "epoch": 0.1593257874015748,
+      "grad_norm": 1.1295069456100464,
+      "learning_rate": 0.0001878104108134829,
+      "loss": 3.0261,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1593873031496063,
+      "grad_norm": 1.0428454875946045,
+      "learning_rate": 0.0001878011566995121,
+      "loss": 2.8369,
+      "step": 2591
+    },
+    {
+      "epoch": 0.1594488188976378,
+      "grad_norm": 1.1803697347640991,
+      "learning_rate": 0.0001877918993022676,
+      "loss": 2.8142,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1595103346456693,
+      "grad_norm": 1.389388918876648,
+      "learning_rate": 0.0001877826386220956,
+      "loss": 3.5393,
+      "step": 2593
+    },
+    {
+      "epoch": 0.15957185039370078,
+      "grad_norm": 0.9003773927688599,
+      "learning_rate": 0.00018777337465934247,
+      "loss": 3.6719,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1596333661417323,
+      "grad_norm": 1.111184000968933,
+      "learning_rate": 0.0001877641074143545,
+      "loss": 2.119,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15969488188976377,
+      "grad_norm": 1.2031677961349487,
+      "learning_rate": 0.00018775483688747834,
+      "loss": 3.1333,
+      "step": 2596
+    },
+    {
+      "epoch": 0.15975639763779528,
+      "grad_norm": 1.5476621389389038,
+      "learning_rate": 0.0001877455630790606,
+      "loss": 2.7622,
+      "step": 2597
+    },
+    {
+      "epoch": 0.15981791338582677,
+      "grad_norm": 1.1391420364379883,
+      "learning_rate": 0.00018773628598944806,
+      "loss": 2.8103,
+      "step": 2598
+    },
+    {
+      "epoch": 0.15987942913385828,
+      "grad_norm": 1.1086400747299194,
+      "learning_rate": 0.00018772700561898767,
+      "loss": 2.8245,
+      "step": 2599
+    },
+    {
+      "epoch": 0.15994094488188976,
+      "grad_norm": 1.244855284690857,
+      "learning_rate": 0.00018771772196802646,
+      "loss": 3.28,
+      "step": 2600
+    },
+    {
+      "epoch": 0.16000246062992127,
+      "grad_norm": 1.147638201713562,
+      "learning_rate": 0.00018770843503691155,
+      "loss": 2.9945,
+      "step": 2601
+    },
+    {
+      "epoch": 0.16006397637795275,
+      "grad_norm": 1.2616757154464722,
+      "learning_rate": 0.0001876991448259903,
+      "loss": 2.8287,
+      "step": 2602
+    },
+    {
+      "epoch": 0.16012549212598426,
+      "grad_norm": 1.0226805210113525,
+      "learning_rate": 0.00018768985133561003,
+      "loss": 2.994,
+      "step": 2603
+    },
+    {
+      "epoch": 0.16018700787401574,
+      "grad_norm": 1.4731080532073975,
+      "learning_rate": 0.00018768055456611825,
+      "loss": 2.5461,
+      "step": 2604
+    },
+    {
+      "epoch": 0.16024852362204725,
+      "grad_norm": 0.9844657778739929,
+      "learning_rate": 0.0001876712545178627,
+      "loss": 3.1467,
+      "step": 2605
+    },
+    {
+      "epoch": 0.16031003937007873,
+      "grad_norm": 1.2950778007507324,
+      "learning_rate": 0.00018766195119119107,
+      "loss": 2.6336,
+      "step": 2606
+    },
+    {
+      "epoch": 0.16037155511811024,
+      "grad_norm": 0.9095386266708374,
+      "learning_rate": 0.0001876526445864513,
+      "loss": 3.4778,
+      "step": 2607
+    },
+    {
+      "epoch": 0.16043307086614172,
+      "grad_norm": 2.104261636734009,
+      "learning_rate": 0.0001876433347039914,
+      "loss": 2.5709,
+      "step": 2608
+    },
+    {
+      "epoch": 0.16049458661417323,
+      "grad_norm": 1.088032841682434,
+      "learning_rate": 0.00018763402154415947,
+      "loss": 2.933,
+      "step": 2609
+    },
+    {
+      "epoch": 0.16055610236220472,
+      "grad_norm": 2.364264488220215,
+      "learning_rate": 0.0001876247051073038,
+      "loss": 1.5948,
+      "step": 2610
+    },
+    {
+      "epoch": 0.16061761811023623,
+      "grad_norm": 1.1147501468658447,
+      "learning_rate": 0.00018761538539377277,
+      "loss": 2.4606,
+      "step": 2611
+    },
+    {
+      "epoch": 0.1606791338582677,
+      "grad_norm": 1.052539587020874,
+      "learning_rate": 0.00018760606240391487,
+      "loss": 3.0705,
+      "step": 2612
+    },
+    {
+      "epoch": 0.16074064960629922,
+      "grad_norm": 1.5677969455718994,
+      "learning_rate": 0.00018759673613807877,
+      "loss": 3.3733,
+      "step": 2613
+    },
+    {
+      "epoch": 0.1608021653543307,
+      "grad_norm": 1.8305739164352417,
+      "learning_rate": 0.00018758740659661317,
+      "loss": 2.7302,
+      "step": 2614
+    },
+    {
+      "epoch": 0.1608636811023622,
+      "grad_norm": 1.2105153799057007,
+      "learning_rate": 0.00018757807377986697,
+      "loss": 2.7543,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1609251968503937,
+      "grad_norm": 0.9061440229415894,
+      "learning_rate": 0.00018756873768818917,
+      "loss": 3.0881,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1609867125984252,
+      "grad_norm": 1.1745527982711792,
+      "learning_rate": 0.00018755939832192886,
+      "loss": 2.5559,
+      "step": 2617
+    },
+    {
+      "epoch": 0.16104822834645668,
+      "grad_norm": 1.2450575828552246,
+      "learning_rate": 0.0001875500556814353,
+      "loss": 2.4531,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1611097440944882,
+      "grad_norm": 1.066078782081604,
+      "learning_rate": 0.00018754070976705785,
+      "loss": 2.6734,
+      "step": 2619
+    },
+    {
+      "epoch": 0.16117125984251968,
+      "grad_norm": 1.095771312713623,
+      "learning_rate": 0.000187531360579146,
+      "loss": 3.002,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1612327755905512,
+      "grad_norm": 1.0121896266937256,
+      "learning_rate": 0.00018752200811804934,
+      "loss": 3.3758,
+      "step": 2621
+    },
+    {
+      "epoch": 0.16129429133858267,
+      "grad_norm": 1.2014013528823853,
+      "learning_rate": 0.00018751265238411762,
+      "loss": 2.9551,
+      "step": 2622
+    },
+    {
+      "epoch": 0.16135580708661418,
+      "grad_norm": 0.9879522919654846,
+      "learning_rate": 0.00018750329337770068,
+      "loss": 3.1237,
+      "step": 2623
+    },
+    {
+      "epoch": 0.16141732283464566,
+      "grad_norm": 1.1767950057983398,
+      "learning_rate": 0.00018749393109914852,
+      "loss": 2.6582,
+      "step": 2624
+    },
+    {
+      "epoch": 0.16147883858267717,
+      "grad_norm": 1.197123408317566,
+      "learning_rate": 0.0001874845655488112,
+      "loss": 2.6736,
+      "step": 2625
+    },
+    {
+      "epoch": 0.16154035433070865,
+      "grad_norm": 1.1238656044006348,
+      "learning_rate": 0.00018747519672703897,
+      "loss": 2.6599,
+      "step": 2626
+    },
+    {
+      "epoch": 0.16160187007874016,
+      "grad_norm": 0.9763971567153931,
+      "learning_rate": 0.00018746582463418214,
+      "loss": 2.9596,
+      "step": 2627
+    },
+    {
+      "epoch": 0.16166338582677164,
+      "grad_norm": 1.7362912893295288,
+      "learning_rate": 0.0001874564492705912,
+      "loss": 2.8321,
+      "step": 2628
+    },
+    {
+      "epoch": 0.16172490157480315,
+      "grad_norm": 0.9568847417831421,
+      "learning_rate": 0.00018744707063661674,
+      "loss": 3.3541,
+      "step": 2629
+    },
+    {
+      "epoch": 0.16178641732283464,
+      "grad_norm": 0.8894321918487549,
+      "learning_rate": 0.00018743768873260942,
+      "loss": 3.199,
+      "step": 2630
+    },
+    {
+      "epoch": 0.16184793307086615,
+      "grad_norm": 0.991981029510498,
+      "learning_rate": 0.00018742830355892017,
+      "loss": 3.2845,
+      "step": 2631
+    },
+    {
+      "epoch": 0.16190944881889763,
+      "grad_norm": 1.0731241703033447,
+      "learning_rate": 0.00018741891511589982,
+      "loss": 2.8841,
+      "step": 2632
+    },
+    {
+      "epoch": 0.16197096456692914,
+      "grad_norm": 1.0435407161712646,
+      "learning_rate": 0.00018740952340389952,
+      "loss": 3.1084,
+      "step": 2633
+    },
+    {
+      "epoch": 0.16203248031496062,
+      "grad_norm": 1.2547341585159302,
+      "learning_rate": 0.00018740012842327047,
+      "loss": 2.9678,
+      "step": 2634
+    },
+    {
+      "epoch": 0.16209399606299213,
+      "grad_norm": 1.0487385988235474,
+      "learning_rate": 0.00018739073017436396,
+      "loss": 2.894,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1621555118110236,
+      "grad_norm": 1.0344139337539673,
+      "learning_rate": 0.00018738132865753145,
+      "loss": 3.3298,
+      "step": 2636
+    },
+    {
+      "epoch": 0.16221702755905512,
+      "grad_norm": 1.0462045669555664,
+      "learning_rate": 0.0001873719238731245,
+      "loss": 2.8605,
+      "step": 2637
+    },
+    {
+      "epoch": 0.1622785433070866,
+      "grad_norm": 1.3319488763809204,
+      "learning_rate": 0.0001873625158214948,
+      "loss": 3.5064,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1623400590551181,
+      "grad_norm": 1.2549830675125122,
+      "learning_rate": 0.00018735310450299414,
+      "loss": 3.0987,
+      "step": 2639
+    },
+    {
+      "epoch": 0.1624015748031496,
+      "grad_norm": 1.4882997274398804,
+      "learning_rate": 0.00018734368991797447,
+      "loss": 2.9156,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1624630905511811,
+      "grad_norm": 1.173740267753601,
+      "learning_rate": 0.00018733427206678785,
+      "loss": 2.494,
+      "step": 2641
+    },
+    {
+      "epoch": 0.1625246062992126,
+      "grad_norm": 1.0030162334442139,
+      "learning_rate": 0.00018732485094978644,
+      "loss": 2.5028,
+      "step": 2642
+    },
+    {
+      "epoch": 0.1625861220472441,
+      "grad_norm": 1.0753371715545654,
+      "learning_rate": 0.00018731542656732253,
+      "loss": 3.0388,
+      "step": 2643
+    },
+    {
+      "epoch": 0.16264763779527558,
+      "grad_norm": 1.0951024293899536,
+      "learning_rate": 0.0001873059989197486,
+      "loss": 2.8356,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1627091535433071,
+      "grad_norm": 1.1477794647216797,
+      "learning_rate": 0.00018729656800741705,
+      "loss": 2.8893,
+      "step": 2645
+    },
+    {
+      "epoch": 0.16277066929133857,
+      "grad_norm": 1.0408469438552856,
+      "learning_rate": 0.0001872871338306807,
+      "loss": 3.082,
+      "step": 2646
+    },
+    {
+      "epoch": 0.16283218503937008,
+      "grad_norm": 1.5073015689849854,
+      "learning_rate": 0.00018727769638989225,
+      "loss": 3.9883,
+      "step": 2647
+    },
+    {
+      "epoch": 0.16289370078740156,
+      "grad_norm": 0.9847761392593384,
+      "learning_rate": 0.00018726825568540466,
+      "loss": 3.0453,
+      "step": 2648
+    },
+    {
+      "epoch": 0.16295521653543307,
+      "grad_norm": 1.3377817869186401,
+      "learning_rate": 0.00018725881171757091,
+      "loss": 1.8737,
+      "step": 2649
+    },
+    {
+      "epoch": 0.16301673228346455,
+      "grad_norm": 0.9624447226524353,
+      "learning_rate": 0.00018724936448674418,
+      "loss": 3.2776,
+      "step": 2650
+    },
+    {
+      "epoch": 0.16307824803149606,
+      "grad_norm": 1.1767011880874634,
+      "learning_rate": 0.00018723991399327772,
+      "loss": 3.3244,
+      "step": 2651
+    },
+    {
+      "epoch": 0.16313976377952755,
+      "grad_norm": 1.2417418956756592,
+      "learning_rate": 0.0001872304602375249,
+      "loss": 3.0315,
+      "step": 2652
+    },
+    {
+      "epoch": 0.16320127952755906,
+      "grad_norm": 0.9537089467048645,
+      "learning_rate": 0.00018722100321983933,
+      "loss": 3.2985,
+      "step": 2653
+    },
+    {
+      "epoch": 0.16326279527559054,
+      "grad_norm": 1.2169328927993774,
+      "learning_rate": 0.00018721154294057461,
+      "loss": 2.7398,
+      "step": 2654
+    },
+    {
+      "epoch": 0.16332431102362205,
+      "grad_norm": 1.2499926090240479,
+      "learning_rate": 0.00018720207940008446,
+      "loss": 2.6416,
+      "step": 2655
+    },
+    {
+      "epoch": 0.16338582677165353,
+      "grad_norm": 1.1985441446304321,
+      "learning_rate": 0.00018719261259872279,
+      "loss": 2.334,
+      "step": 2656
+    },
+    {
+      "epoch": 0.16344734251968504,
+      "grad_norm": 0.9895098805427551,
+      "learning_rate": 0.0001871831425368436,
+      "loss": 3.0414,
+      "step": 2657
+    },
+    {
+      "epoch": 0.16350885826771652,
+      "grad_norm": 1.035962462425232,
+      "learning_rate": 0.00018717366921480104,
+      "loss": 3.6356,
+      "step": 2658
+    },
+    {
+      "epoch": 0.16357037401574803,
+      "grad_norm": 1.2370784282684326,
+      "learning_rate": 0.00018716419263294936,
+      "loss": 3.0138,
+      "step": 2659
+    },
+    {
+      "epoch": 0.16363188976377951,
+      "grad_norm": 1.1424927711486816,
+      "learning_rate": 0.0001871547127916429,
+      "loss": 2.4701,
+      "step": 2660
+    },
+    {
+      "epoch": 0.16369340551181102,
+      "grad_norm": 1.4885913133621216,
+      "learning_rate": 0.00018714522969123616,
+      "loss": 2.9626,
+      "step": 2661
+    },
+    {
+      "epoch": 0.1637549212598425,
+      "grad_norm": 1.002933382987976,
+      "learning_rate": 0.0001871357433320838,
+      "loss": 3.3379,
+      "step": 2662
+    },
+    {
+      "epoch": 0.16381643700787402,
+      "grad_norm": 1.3244662284851074,
+      "learning_rate": 0.00018712625371454048,
+      "loss": 3.1773,
+      "step": 2663
+    },
+    {
+      "epoch": 0.1638779527559055,
+      "grad_norm": 1.4265782833099365,
+      "learning_rate": 0.00018711676083896112,
+      "loss": 2.6082,
+      "step": 2664
+    },
+    {
+      "epoch": 0.163939468503937,
+      "grad_norm": 1.154798150062561,
+      "learning_rate": 0.0001871072647057007,
+      "loss": 2.8762,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1640009842519685,
+      "grad_norm": 1.4291200637817383,
+      "learning_rate": 0.00018709776531511426,
+      "loss": 3.0576,
+      "step": 2666
+    },
+    {
+      "epoch": 0.1640625,
+      "grad_norm": 1.4053013324737549,
+      "learning_rate": 0.0001870882626675571,
+      "loss": 2.934,
+      "step": 2667
+    },
+    {
+      "epoch": 0.1641240157480315,
+      "grad_norm": 1.0630061626434326,
+      "learning_rate": 0.00018707875676338454,
+      "loss": 2.9174,
+      "step": 2668
+    },
+    {
+      "epoch": 0.164185531496063,
+      "grad_norm": 1.2252604961395264,
+      "learning_rate": 0.000187069247602952,
+      "loss": 2.9414,
+      "step": 2669
+    },
+    {
+      "epoch": 0.1642470472440945,
+      "grad_norm": 0.9744952917098999,
+      "learning_rate": 0.00018705973518661514,
+      "loss": 3.3217,
+      "step": 2670
+    },
+    {
+      "epoch": 0.16430856299212598,
+      "grad_norm": 0.9772462248802185,
+      "learning_rate": 0.00018705021951472963,
+      "loss": 3.0856,
+      "step": 2671
+    },
+    {
+      "epoch": 0.1643700787401575,
+      "grad_norm": 0.8932956457138062,
+      "learning_rate": 0.00018704070058765134,
+      "loss": 3.3233,
+      "step": 2672
+    },
+    {
+      "epoch": 0.16443159448818898,
+      "grad_norm": 1.1857789754867554,
+      "learning_rate": 0.0001870311784057362,
+      "loss": 3.2127,
+      "step": 2673
+    },
+    {
+      "epoch": 0.16449311023622049,
+      "grad_norm": 1.0652039051055908,
+      "learning_rate": 0.00018702165296934027,
+      "loss": 2.5137,
+      "step": 2674
+    },
+    {
+      "epoch": 0.16455462598425197,
+      "grad_norm": 1.6336426734924316,
+      "learning_rate": 0.00018701212427881973,
+      "loss": 2.5006,
+      "step": 2675
+    },
+    {
+      "epoch": 0.16461614173228348,
+      "grad_norm": 1.032503604888916,
+      "learning_rate": 0.000187002592334531,
+      "loss": 3.004,
+      "step": 2676
+    },
+    {
+      "epoch": 0.16467765748031496,
+      "grad_norm": 1.0305896997451782,
+      "learning_rate": 0.0001869930571368304,
+      "loss": 3.7062,
+      "step": 2677
+    },
+    {
+      "epoch": 0.16473917322834647,
+      "grad_norm": 1.2351287603378296,
+      "learning_rate": 0.0001869835186860746,
+      "loss": 2.8354,
+      "step": 2678
+    },
+    {
+      "epoch": 0.16480068897637795,
+      "grad_norm": 1.0796245336532593,
+      "learning_rate": 0.0001869739769826202,
+      "loss": 3.1304,
+      "step": 2679
+    },
+    {
+      "epoch": 0.16486220472440946,
+      "grad_norm": 1.0404778718948364,
+      "learning_rate": 0.00018696443202682406,
+      "loss": 2.8857,
+      "step": 2680
+    },
+    {
+      "epoch": 0.16492372047244094,
+      "grad_norm": 1.0031142234802246,
+      "learning_rate": 0.0001869548838190431,
+      "loss": 3.0542,
+      "step": 2681
+    },
+    {
+      "epoch": 0.16498523622047245,
+      "grad_norm": 1.2063114643096924,
+      "learning_rate": 0.00018694533235963433,
+      "loss": 2.6176,
+      "step": 2682
+    },
+    {
+      "epoch": 0.16504675196850394,
+      "grad_norm": 1.4449584484100342,
+      "learning_rate": 0.00018693577764895497,
+      "loss": 3.0237,
+      "step": 2683
+    },
+    {
+      "epoch": 0.16510826771653545,
+      "grad_norm": 1.9799611568450928,
+      "learning_rate": 0.0001869262196873623,
+      "loss": 1.4179,
+      "step": 2684
+    },
+    {
+      "epoch": 0.16516978346456693,
+      "grad_norm": 1.9806879758834839,
+      "learning_rate": 0.0001869166584752137,
+      "loss": 2.2803,
+      "step": 2685
+    },
+    {
+      "epoch": 0.16523129921259844,
+      "grad_norm": 1.8004262447357178,
+      "learning_rate": 0.00018690709401286674,
+      "loss": 1.5653,
+      "step": 2686
+    },
+    {
+      "epoch": 0.16529281496062992,
+      "grad_norm": 1.2225496768951416,
+      "learning_rate": 0.00018689752630067906,
+      "loss": 3.1655,
+      "step": 2687
+    },
+    {
+      "epoch": 0.16535433070866143,
+      "grad_norm": 1.2399488687515259,
+      "learning_rate": 0.00018688795533900848,
+      "loss": 2.853,
+      "step": 2688
+    },
+    {
+      "epoch": 0.1654158464566929,
+      "grad_norm": 1.375830888748169,
+      "learning_rate": 0.00018687838112821286,
+      "loss": 3.4032,
+      "step": 2689
+    },
+    {
+      "epoch": 0.16547736220472442,
+      "grad_norm": 1.243855357170105,
+      "learning_rate": 0.00018686880366865024,
+      "loss": 2.9293,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1655388779527559,
+      "grad_norm": 1.0390574932098389,
+      "learning_rate": 0.00018685922296067875,
+      "loss": 2.8457,
+      "step": 2691
+    },
+    {
+      "epoch": 0.1656003937007874,
+      "grad_norm": 1.2467584609985352,
+      "learning_rate": 0.00018684963900465664,
+      "loss": 2.3363,
+      "step": 2692
+    },
+    {
+      "epoch": 0.1656619094488189,
+      "grad_norm": 0.9641879200935364,
+      "learning_rate": 0.00018684005180094233,
+      "loss": 2.8004,
+      "step": 2693
+    },
+    {
+      "epoch": 0.1657234251968504,
+      "grad_norm": 1.182455062866211,
+      "learning_rate": 0.00018683046134989435,
+      "loss": 3.6177,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1657849409448819,
+      "grad_norm": 1.1010624170303345,
+      "learning_rate": 0.00018682086765187125,
+      "loss": 3.3501,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1658464566929134,
+      "grad_norm": 1.040527582168579,
+      "learning_rate": 0.00018681127070723186,
+      "loss": 3.0552,
+      "step": 2696
+    },
+    {
+      "epoch": 0.16590797244094488,
+      "grad_norm": 1.476844310760498,
+      "learning_rate": 0.000186801670516335,
+      "loss": 2.2896,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1659694881889764,
+      "grad_norm": 1.1581664085388184,
+      "learning_rate": 0.0001867920670795397,
+      "loss": 2.9406,
+      "step": 2698
+    },
+    {
+      "epoch": 0.16603100393700787,
+      "grad_norm": 1.5145245790481567,
+      "learning_rate": 0.000186782460397205,
+      "loss": 2.7758,
+      "step": 2699
+    },
+    {
+      "epoch": 0.16609251968503938,
+      "grad_norm": 1.2854723930358887,
+      "learning_rate": 0.00018677285046969026,
+      "loss": 2.4492,
+      "step": 2700
+    },
+    {
+      "epoch": 0.16615403543307086,
+      "grad_norm": 1.2777475118637085,
+      "learning_rate": 0.00018676323729735475,
+      "loss": 3.0013,
+      "step": 2701
+    },
+    {
+      "epoch": 0.16621555118110237,
+      "grad_norm": 1.8208692073822021,
+      "learning_rate": 0.00018675362088055794,
+      "loss": 2.5882,
+      "step": 2702
+    },
+    {
+      "epoch": 0.16627706692913385,
+      "grad_norm": 1.038618564605713,
+      "learning_rate": 0.00018674400121965952,
+      "loss": 2.7986,
+      "step": 2703
+    },
+    {
+      "epoch": 0.16633858267716536,
+      "grad_norm": 1.2081722021102905,
+      "learning_rate": 0.0001867343783150191,
+      "loss": 3.2389,
+      "step": 2704
+    },
+    {
+      "epoch": 0.16640009842519685,
+      "grad_norm": 1.1211960315704346,
+      "learning_rate": 0.0001867247521669966,
+      "loss": 2.7704,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16646161417322836,
+      "grad_norm": 1.2080992460250854,
+      "learning_rate": 0.00018671512277595195,
+      "loss": 3.0934,
+      "step": 2706
+    },
+    {
+      "epoch": 0.16652312992125984,
+      "grad_norm": 1.7616084814071655,
+      "learning_rate": 0.00018670549014224524,
+      "loss": 2.1852,
+      "step": 2707
+    },
+    {
+      "epoch": 0.16658464566929135,
+      "grad_norm": 0.9149419069290161,
+      "learning_rate": 0.0001866958542662367,
+      "loss": 3.2312,
+      "step": 2708
+    },
+    {
+      "epoch": 0.16664616141732283,
+      "grad_norm": 0.9722538590431213,
+      "learning_rate": 0.00018668621514828665,
+      "loss": 2.9538,
+      "step": 2709
+    },
+    {
+      "epoch": 0.16670767716535434,
+      "grad_norm": 1.086599349975586,
+      "learning_rate": 0.0001866765727887555,
+      "loss": 3.1159,
+      "step": 2710
+    },
+    {
+      "epoch": 0.16676919291338582,
+      "grad_norm": 1.0480951070785522,
+      "learning_rate": 0.00018666692718800386,
+      "loss": 3.5495,
+      "step": 2711
+    },
+    {
+      "epoch": 0.16683070866141733,
+      "grad_norm": 1.4742212295532227,
+      "learning_rate": 0.00018665727834639242,
+      "loss": 3.1489,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1668922244094488,
+      "grad_norm": 1.0193161964416504,
+      "learning_rate": 0.00018664762626428196,
+      "loss": 3.3482,
+      "step": 2713
+    },
+    {
+      "epoch": 0.16695374015748032,
+      "grad_norm": 1.1660687923431396,
+      "learning_rate": 0.00018663797094203345,
+      "loss": 2.4853,
+      "step": 2714
+    },
+    {
+      "epoch": 0.1670152559055118,
+      "grad_norm": 1.1057393550872803,
+      "learning_rate": 0.00018662831238000792,
+      "loss": 3.0851,
+      "step": 2715
+    },
+    {
+      "epoch": 0.16707677165354332,
+      "grad_norm": 1.1379483938217163,
+      "learning_rate": 0.0001866186505785666,
+      "loss": 3.2729,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1671382874015748,
+      "grad_norm": 1.1868369579315186,
+      "learning_rate": 0.00018660898553807075,
+      "loss": 3.0991,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1671998031496063,
+      "grad_norm": 0.9943464994430542,
+      "learning_rate": 0.00018659931725888173,
+      "loss": 2.8795,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1672613188976378,
+      "grad_norm": 1.1517245769500732,
+      "learning_rate": 0.00018658964574136118,
+      "loss": 1.9903,
+      "step": 2719
+    },
+    {
+      "epoch": 0.1673228346456693,
+      "grad_norm": 1.2196877002716064,
+      "learning_rate": 0.0001865799709858707,
+      "loss": 3.3413,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16738435039370078,
+      "grad_norm": 1.1712861061096191,
+      "learning_rate": 0.00018657029299277208,
+      "loss": 2.8985,
+      "step": 2721
+    },
+    {
+      "epoch": 0.1674458661417323,
+      "grad_norm": 1.503159523010254,
+      "learning_rate": 0.00018656061176242723,
+      "loss": 3.2331,
+      "step": 2722
+    },
+    {
+      "epoch": 0.16750738188976377,
+      "grad_norm": 1.1463723182678223,
+      "learning_rate": 0.00018655092729519818,
+      "loss": 3.2052,
+      "step": 2723
+    },
+    {
+      "epoch": 0.16756889763779528,
+      "grad_norm": 1.0839956998825073,
+      "learning_rate": 0.00018654123959144708,
+      "loss": 3.0505,
+      "step": 2724
+    },
+    {
+      "epoch": 0.16763041338582677,
+      "grad_norm": 1.135030746459961,
+      "learning_rate": 0.0001865315486515362,
+      "loss": 3.5878,
+      "step": 2725
+    },
+    {
+      "epoch": 0.16769192913385828,
+      "grad_norm": 0.9340260624885559,
+      "learning_rate": 0.00018652185447582792,
+      "loss": 3.0226,
+      "step": 2726
+    },
+    {
+      "epoch": 0.16775344488188976,
+      "grad_norm": 0.9414381384849548,
+      "learning_rate": 0.0001865121570646847,
+      "loss": 3.195,
+      "step": 2727
+    },
+    {
+      "epoch": 0.16781496062992127,
+      "grad_norm": 0.9118775129318237,
+      "learning_rate": 0.00018650245641846922,
+      "loss": 2.9963,
+      "step": 2728
+    },
+    {
+      "epoch": 0.16787647637795275,
+      "grad_norm": 1.0968468189239502,
+      "learning_rate": 0.00018649275253754423,
+      "loss": 2.6284,
+      "step": 2729
+    },
+    {
+      "epoch": 0.16793799212598426,
+      "grad_norm": 1.1011162996292114,
+      "learning_rate": 0.0001864830454222726,
+      "loss": 2.9739,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16799950787401574,
+      "grad_norm": 1.2122737169265747,
+      "learning_rate": 0.0001864733350730173,
+      "loss": 2.4948,
+      "step": 2731
+    },
+    {
+      "epoch": 0.16806102362204725,
+      "grad_norm": 1.1605031490325928,
+      "learning_rate": 0.00018646362149014146,
+      "loss": 2.1354,
+      "step": 2732
+    },
+    {
+      "epoch": 0.16812253937007873,
+      "grad_norm": 0.9926984310150146,
+      "learning_rate": 0.0001864539046740083,
+      "loss": 2.6612,
+      "step": 2733
+    },
+    {
+      "epoch": 0.16818405511811024,
+      "grad_norm": 1.195613145828247,
+      "learning_rate": 0.0001864441846249812,
+      "loss": 2.8171,
+      "step": 2734
+    },
+    {
+      "epoch": 0.16824557086614172,
+      "grad_norm": 1.3065422773361206,
+      "learning_rate": 0.00018643446134342364,
+      "loss": 3.3367,
+      "step": 2735
+    },
+    {
+      "epoch": 0.16830708661417323,
+      "grad_norm": 1.0238103866577148,
+      "learning_rate": 0.00018642473482969915,
+      "loss": 3.0942,
+      "step": 2736
+    },
+    {
+      "epoch": 0.16836860236220472,
+      "grad_norm": 1.0642913579940796,
+      "learning_rate": 0.00018641500508417152,
+      "loss": 2.7351,
+      "step": 2737
+    },
+    {
+      "epoch": 0.16843011811023623,
+      "grad_norm": 1.224284052848816,
+      "learning_rate": 0.00018640527210720458,
+      "loss": 3.0255,
+      "step": 2738
+    },
+    {
+      "epoch": 0.1684916338582677,
+      "grad_norm": 1.3287800550460815,
+      "learning_rate": 0.00018639553589916227,
+      "loss": 2.7871,
+      "step": 2739
+    },
+    {
+      "epoch": 0.16855314960629922,
+      "grad_norm": 1.121480941772461,
+      "learning_rate": 0.00018638579646040865,
+      "loss": 3.2881,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1686146653543307,
+      "grad_norm": 1.0270907878875732,
+      "learning_rate": 0.00018637605379130795,
+      "loss": 2.6824,
+      "step": 2741
+    },
+    {
+      "epoch": 0.1686761811023622,
+      "grad_norm": 1.2137147188186646,
+      "learning_rate": 0.00018636630789222448,
+      "loss": 3.1453,
+      "step": 2742
+    },
+    {
+      "epoch": 0.1687376968503937,
+      "grad_norm": 1.23988938331604,
+      "learning_rate": 0.00018635655876352272,
+      "loss": 2.8014,
+      "step": 2743
+    },
+    {
+      "epoch": 0.1687992125984252,
+      "grad_norm": 1.5585132837295532,
+      "learning_rate": 0.00018634680640556715,
+      "loss": 2.578,
+      "step": 2744
+    },
+    {
+      "epoch": 0.16886072834645668,
+      "grad_norm": 1.2602171897888184,
+      "learning_rate": 0.00018633705081872254,
+      "loss": 2.2595,
+      "step": 2745
+    },
+    {
+      "epoch": 0.1689222440944882,
+      "grad_norm": 1.4583719968795776,
+      "learning_rate": 0.00018632729200335365,
+      "loss": 2.9049,
+      "step": 2746
+    },
+    {
+      "epoch": 0.16898375984251968,
+      "grad_norm": 1.076823115348816,
+      "learning_rate": 0.0001863175299598254,
+      "loss": 2.6531,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1690452755905512,
+      "grad_norm": 1.0033206939697266,
+      "learning_rate": 0.00018630776468850284,
+      "loss": 3.3896,
+      "step": 2748
+    },
+    {
+      "epoch": 0.16910679133858267,
+      "grad_norm": 1.155337929725647,
+      "learning_rate": 0.00018629799618975117,
+      "loss": 2.8379,
+      "step": 2749
+    },
+    {
+      "epoch": 0.16916830708661418,
+      "grad_norm": 1.149537444114685,
+      "learning_rate": 0.00018628822446393564,
+      "loss": 2.8538,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16922982283464566,
+      "grad_norm": 1.0222467184066772,
+      "learning_rate": 0.0001862784495114217,
+      "loss": 3.3429,
+      "step": 2751
+    },
+    {
+      "epoch": 0.16929133858267717,
+      "grad_norm": 1.03061044216156,
+      "learning_rate": 0.0001862686713325748,
+      "loss": 2.8768,
+      "step": 2752
+    },
+    {
+      "epoch": 0.16935285433070865,
+      "grad_norm": 0.9499995708465576,
+      "learning_rate": 0.00018625888992776068,
+      "loss": 3.2179,
+      "step": 2753
+    },
+    {
+      "epoch": 0.16941437007874016,
+      "grad_norm": 1.1791208982467651,
+      "learning_rate": 0.00018624910529734505,
+      "loss": 3.1014,
+      "step": 2754
+    },
+    {
+      "epoch": 0.16947588582677164,
+      "grad_norm": 1.065549612045288,
+      "learning_rate": 0.00018623931744169383,
+      "loss": 3.0047,
+      "step": 2755
+    },
+    {
+      "epoch": 0.16953740157480315,
+      "grad_norm": 0.9729413390159607,
+      "learning_rate": 0.000186229526361173,
+      "loss": 3.0117,
+      "step": 2756
+    },
+    {
+      "epoch": 0.16959891732283464,
+      "grad_norm": 1.6614437103271484,
+      "learning_rate": 0.00018621973205614872,
+      "loss": 2.5657,
+      "step": 2757
+    },
+    {
+      "epoch": 0.16966043307086615,
+      "grad_norm": 0.8782540559768677,
+      "learning_rate": 0.00018620993452698727,
+      "loss": 2.8439,
+      "step": 2758
+    },
+    {
+      "epoch": 0.16972194881889763,
+      "grad_norm": 1.7177263498306274,
+      "learning_rate": 0.00018620013377405496,
+      "loss": 2.7479,
+      "step": 2759
+    },
+    {
+      "epoch": 0.16978346456692914,
+      "grad_norm": 1.280905842781067,
+      "learning_rate": 0.00018619032979771833,
+      "loss": 2.512,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16984498031496062,
+      "grad_norm": 1.134282112121582,
+      "learning_rate": 0.00018618052259834395,
+      "loss": 2.7053,
+      "step": 2761
+    },
+    {
+      "epoch": 0.16990649606299213,
+      "grad_norm": 1.0151124000549316,
+      "learning_rate": 0.0001861707121762986,
+      "loss": 2.9071,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1699680118110236,
+      "grad_norm": 1.0354288816452026,
+      "learning_rate": 0.00018616089853194911,
+      "loss": 3.3586,
+      "step": 2763
+    },
+    {
+      "epoch": 0.17002952755905512,
+      "grad_norm": 1.2481003999710083,
+      "learning_rate": 0.00018615108166566247,
+      "loss": 2.8409,
+      "step": 2764
+    },
+    {
+      "epoch": 0.1700910433070866,
+      "grad_norm": 1.1704663038253784,
+      "learning_rate": 0.00018614126157780577,
+      "loss": 2.6837,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1701525590551181,
+      "grad_norm": 1.1585263013839722,
+      "learning_rate": 0.0001861314382687462,
+      "loss": 3.32,
+      "step": 2766
+    },
+    {
+      "epoch": 0.1702140748031496,
+      "grad_norm": 0.8884790539741516,
+      "learning_rate": 0.00018612161173885112,
+      "loss": 3.4036,
+      "step": 2767
+    },
+    {
+      "epoch": 0.1702755905511811,
+      "grad_norm": 0.95437091588974,
+      "learning_rate": 0.000186111781988488,
+      "loss": 2.8512,
+      "step": 2768
+    },
+    {
+      "epoch": 0.1703371062992126,
+      "grad_norm": 0.9761068224906921,
+      "learning_rate": 0.00018610194901802442,
+      "loss": 2.6117,
+      "step": 2769
+    },
+    {
+      "epoch": 0.1703986220472441,
+      "grad_norm": 0.9973070621490479,
+      "learning_rate": 0.00018609211282782803,
+      "loss": 3.1282,
+      "step": 2770
+    },
+    {
+      "epoch": 0.17046013779527558,
+      "grad_norm": 1.330583930015564,
+      "learning_rate": 0.0001860822734182667,
+      "loss": 1.8082,
+      "step": 2771
+    },
+    {
+      "epoch": 0.1705216535433071,
+      "grad_norm": 0.9660400152206421,
+      "learning_rate": 0.00018607243078970837,
+      "loss": 2.8352,
+      "step": 2772
+    },
+    {
+      "epoch": 0.17058316929133857,
+      "grad_norm": 1.3950637578964233,
+      "learning_rate": 0.00018606258494252106,
+      "loss": 2.4804,
+      "step": 2773
+    },
+    {
+      "epoch": 0.17064468503937008,
+      "grad_norm": 1.1968629360198975,
+      "learning_rate": 0.00018605273587707298,
+      "loss": 2.8224,
+      "step": 2774
+    },
+    {
+      "epoch": 0.17070620078740156,
+      "grad_norm": 1.1280027627944946,
+      "learning_rate": 0.00018604288359373242,
+      "loss": 3.2326,
+      "step": 2775
+    },
+    {
+      "epoch": 0.17076771653543307,
+      "grad_norm": 1.3074382543563843,
+      "learning_rate": 0.0001860330280928678,
+      "loss": 3.001,
+      "step": 2776
+    },
+    {
+      "epoch": 0.17082923228346455,
+      "grad_norm": 1.1901464462280273,
+      "learning_rate": 0.00018602316937484767,
+      "loss": 2.1507,
+      "step": 2777
+    },
+    {
+      "epoch": 0.17089074803149606,
+      "grad_norm": 1.409936785697937,
+      "learning_rate": 0.0001860133074400407,
+      "loss": 2.8693,
+      "step": 2778
+    },
+    {
+      "epoch": 0.17095226377952755,
+      "grad_norm": 1.0880374908447266,
+      "learning_rate": 0.0001860034422888156,
+      "loss": 2.4987,
+      "step": 2779
+    },
+    {
+      "epoch": 0.17101377952755906,
+      "grad_norm": 1.4239530563354492,
+      "learning_rate": 0.0001859935739215414,
+      "loss": 2.5481,
+      "step": 2780
+    },
+    {
+      "epoch": 0.17107529527559054,
+      "grad_norm": 1.17850923538208,
+      "learning_rate": 0.000185983702338587,
+      "loss": 2.7345,
+      "step": 2781
+    },
+    {
+      "epoch": 0.17113681102362205,
+      "grad_norm": 1.0997556447982788,
+      "learning_rate": 0.00018597382754032156,
+      "loss": 2.5725,
+      "step": 2782
+    },
+    {
+      "epoch": 0.17119832677165353,
+      "grad_norm": 1.0634764432907104,
+      "learning_rate": 0.00018596394952711441,
+      "loss": 2.8999,
+      "step": 2783
+    },
+    {
+      "epoch": 0.17125984251968504,
+      "grad_norm": 0.9793528318405151,
+      "learning_rate": 0.00018595406829933494,
+      "loss": 3.1254,
+      "step": 2784
+    },
+    {
+      "epoch": 0.17132135826771652,
+      "grad_norm": 1.1364848613739014,
+      "learning_rate": 0.00018594418385735255,
+      "loss": 2.4896,
+      "step": 2785
+    },
+    {
+      "epoch": 0.17138287401574803,
+      "grad_norm": 1.0198535919189453,
+      "learning_rate": 0.00018593429620153694,
+      "loss": 2.89,
+      "step": 2786
+    },
+    {
+      "epoch": 0.17144438976377951,
+      "grad_norm": 0.9916906952857971,
+      "learning_rate": 0.00018592440533225785,
+      "loss": 2.9899,
+      "step": 2787
+    },
+    {
+      "epoch": 0.17150590551181102,
+      "grad_norm": 1.1045984029769897,
+      "learning_rate": 0.00018591451124988511,
+      "loss": 2.0063,
+      "step": 2788
+    },
+    {
+      "epoch": 0.1715674212598425,
+      "grad_norm": 1.1302490234375,
+      "learning_rate": 0.0001859046139547887,
+      "loss": 2.3641,
+      "step": 2789
+    },
+    {
+      "epoch": 0.17162893700787402,
+      "grad_norm": 0.9899945855140686,
+      "learning_rate": 0.00018589471344733878,
+      "loss": 2.9594,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1716904527559055,
+      "grad_norm": 1.1547448635101318,
+      "learning_rate": 0.0001858848097279055,
+      "loss": 2.8485,
+      "step": 2791
+    },
+    {
+      "epoch": 0.171751968503937,
+      "grad_norm": 1.0877596139907837,
+      "learning_rate": 0.00018587490279685929,
+      "loss": 2.9513,
+      "step": 2792
+    },
+    {
+      "epoch": 0.1718134842519685,
+      "grad_norm": 1.1007336378097534,
+      "learning_rate": 0.00018586499265457057,
+      "loss": 3.1355,
+      "step": 2793
+    },
+    {
+      "epoch": 0.171875,
+      "grad_norm": 1.2335039377212524,
+      "learning_rate": 0.00018585507930140988,
+      "loss": 2.3669,
+      "step": 2794
+    },
+    {
+      "epoch": 0.1719365157480315,
+      "grad_norm": 1.007286787033081,
+      "learning_rate": 0.000185845162737748,
+      "loss": 2.7609,
+      "step": 2795
+    },
+    {
+      "epoch": 0.171998031496063,
+      "grad_norm": 1.0199494361877441,
+      "learning_rate": 0.0001858352429639557,
+      "loss": 2.2996,
+      "step": 2796
+    },
+    {
+      "epoch": 0.1720595472440945,
+      "grad_norm": 1.3494491577148438,
+      "learning_rate": 0.00018582531998040397,
+      "loss": 2.272,
+      "step": 2797
+    },
+    {
+      "epoch": 0.17212106299212598,
+      "grad_norm": 1.0779451131820679,
+      "learning_rate": 0.0001858153937874638,
+      "loss": 2.8335,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1721825787401575,
+      "grad_norm": 1.0373282432556152,
+      "learning_rate": 0.00018580546438550646,
+      "loss": 2.6204,
+      "step": 2799
+    },
+    {
+      "epoch": 0.17224409448818898,
+      "grad_norm": 1.079702377319336,
+      "learning_rate": 0.00018579553177490318,
+      "loss": 2.5977,
+      "step": 2800
+    },
+    {
+      "epoch": 0.17230561023622049,
+      "grad_norm": 1.07167387008667,
+      "learning_rate": 0.00018578559595602544,
+      "loss": 2.6819,
+      "step": 2801
+    },
+    {
+      "epoch": 0.17236712598425197,
+      "grad_norm": 1.2057318687438965,
+      "learning_rate": 0.00018577565692924476,
+      "loss": 2.9955,
+      "step": 2802
+    },
+    {
+      "epoch": 0.17242864173228348,
+      "grad_norm": 1.1469814777374268,
+      "learning_rate": 0.00018576571469493285,
+      "loss": 3.6092,
+      "step": 2803
+    },
+    {
+      "epoch": 0.17249015748031496,
+      "grad_norm": 1.0826716423034668,
+      "learning_rate": 0.0001857557692534614,
+      "loss": 3.0771,
+      "step": 2804
+    },
+    {
+      "epoch": 0.17255167322834647,
+      "grad_norm": 1.1636910438537598,
+      "learning_rate": 0.0001857458206052024,
+      "loss": 2.7871,
+      "step": 2805
+    },
+    {
+      "epoch": 0.17261318897637795,
+      "grad_norm": 1.164202094078064,
+      "learning_rate": 0.0001857358687505278,
+      "loss": 2.7703,
+      "step": 2806
+    },
+    {
+      "epoch": 0.17267470472440946,
+      "grad_norm": 1.240195393562317,
+      "learning_rate": 0.00018572591368980982,
+      "loss": 2.4327,
+      "step": 2807
+    },
+    {
+      "epoch": 0.17273622047244094,
+      "grad_norm": 1.0924466848373413,
+      "learning_rate": 0.0001857159554234207,
+      "loss": 3.2046,
+      "step": 2808
+    },
+    {
+      "epoch": 0.17279773622047245,
+      "grad_norm": 1.3040008544921875,
+      "learning_rate": 0.0001857059939517328,
+      "loss": 2.8157,
+      "step": 2809
+    },
+    {
+      "epoch": 0.17285925196850394,
+      "grad_norm": 0.9632295370101929,
+      "learning_rate": 0.00018569602927511863,
+      "loss": 2.8537,
+      "step": 2810
+    },
+    {
+      "epoch": 0.17292076771653545,
+      "grad_norm": 1.1435433626174927,
+      "learning_rate": 0.0001856860613939508,
+      "loss": 2.8402,
+      "step": 2811
+    },
+    {
+      "epoch": 0.17298228346456693,
+      "grad_norm": 1.3415615558624268,
+      "learning_rate": 0.0001856760903086021,
+      "loss": 2.7601,
+      "step": 2812
+    },
+    {
+      "epoch": 0.17304379921259844,
+      "grad_norm": 0.9818133115768433,
+      "learning_rate": 0.00018566611601944537,
+      "loss": 3.4938,
+      "step": 2813
+    },
+    {
+      "epoch": 0.17310531496062992,
+      "grad_norm": 0.9568825960159302,
+      "learning_rate": 0.0001856561385268536,
+      "loss": 2.9053,
+      "step": 2814
+    },
+    {
+      "epoch": 0.17316683070866143,
+      "grad_norm": 1.4131356477737427,
+      "learning_rate": 0.00018564615783119986,
+      "loss": 2.6526,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1732283464566929,
+      "grad_norm": 1.1081582307815552,
+      "learning_rate": 0.00018563617393285738,
+      "loss": 3.0818,
+      "step": 2816
+    },
+    {
+      "epoch": 0.17328986220472442,
+      "grad_norm": 1.0180065631866455,
+      "learning_rate": 0.00018562618683219952,
+      "loss": 3.1819,
+      "step": 2817
+    },
+    {
+      "epoch": 0.1733513779527559,
+      "grad_norm": 1.222563624382019,
+      "learning_rate": 0.00018561619652959977,
+      "loss": 3.122,
+      "step": 2818
+    },
+    {
+      "epoch": 0.1734128937007874,
+      "grad_norm": 1.127455234527588,
+      "learning_rate": 0.00018560620302543162,
+      "loss": 3.1648,
+      "step": 2819
+    },
+    {
+      "epoch": 0.1734744094488189,
+      "grad_norm": 0.9851447939872742,
+      "learning_rate": 0.00018559620632006886,
+      "loss": 1.7192,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1735359251968504,
+      "grad_norm": 2.0624682903289795,
+      "learning_rate": 0.00018558620641388531,
+      "loss": 2.985,
+      "step": 2821
+    },
+    {
+      "epoch": 0.1735974409448819,
+      "grad_norm": 1.1579904556274414,
+      "learning_rate": 0.00018557620330725483,
+      "loss": 3.1637,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1736589566929134,
+      "grad_norm": 1.4565489292144775,
+      "learning_rate": 0.00018556619700055155,
+      "loss": 2.4914,
+      "step": 2823
+    },
+    {
+      "epoch": 0.17372047244094488,
+      "grad_norm": 1.1226341724395752,
+      "learning_rate": 0.00018555618749414965,
+      "loss": 3.0001,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1737819881889764,
+      "grad_norm": 1.0648829936981201,
+      "learning_rate": 0.00018554617478842343,
+      "loss": 2.7491,
+      "step": 2825
+    },
+    {
+      "epoch": 0.17384350393700787,
+      "grad_norm": 0.9068941473960876,
+      "learning_rate": 0.00018553615888374726,
+      "loss": 3.4331,
+      "step": 2826
+    },
+    {
+      "epoch": 0.17390501968503938,
+      "grad_norm": 1.1168879270553589,
+      "learning_rate": 0.0001855261397804957,
+      "loss": 3.3704,
+      "step": 2827
+    },
+    {
+      "epoch": 0.17396653543307086,
+      "grad_norm": 1.126010775566101,
+      "learning_rate": 0.00018551611747904344,
+      "loss": 2.9785,
+      "step": 2828
+    },
+    {
+      "epoch": 0.17402805118110237,
+      "grad_norm": 1.184661865234375,
+      "learning_rate": 0.00018550609197976522,
+      "loss": 2.6603,
+      "step": 2829
+    },
+    {
+      "epoch": 0.17408956692913385,
+      "grad_norm": 0.9783070683479309,
+      "learning_rate": 0.00018549606328303596,
+      "loss": 3.3254,
+      "step": 2830
+    },
+    {
+      "epoch": 0.17415108267716536,
+      "grad_norm": 1.2079834938049316,
+      "learning_rate": 0.00018548603138923068,
+      "loss": 2.3793,
+      "step": 2831
+    },
+    {
+      "epoch": 0.17421259842519685,
+      "grad_norm": 1.5435289144515991,
+      "learning_rate": 0.00018547599629872453,
+      "loss": 2.3666,
+      "step": 2832
+    },
+    {
+      "epoch": 0.17427411417322836,
+      "grad_norm": 1.0477808713912964,
+      "learning_rate": 0.00018546595801189272,
+      "loss": 2.8846,
+      "step": 2833
+    },
+    {
+      "epoch": 0.17433562992125984,
+      "grad_norm": 0.8984761834144592,
+      "learning_rate": 0.00018545591652911064,
+      "loss": 3.2279,
+      "step": 2834
+    },
+    {
+      "epoch": 0.17439714566929135,
+      "grad_norm": 1.069601058959961,
+      "learning_rate": 0.00018544587185075382,
+      "loss": 3.1667,
+      "step": 2835
+    },
+    {
+      "epoch": 0.17445866141732283,
+      "grad_norm": 1.0154985189437866,
+      "learning_rate": 0.00018543582397719782,
+      "loss": 2.1107,
+      "step": 2836
+    },
+    {
+      "epoch": 0.17452017716535434,
+      "grad_norm": 0.8551943302154541,
+      "learning_rate": 0.00018542577290881846,
+      "loss": 3.096,
+      "step": 2837
+    },
+    {
+      "epoch": 0.17458169291338582,
+      "grad_norm": 1.2940270900726318,
+      "learning_rate": 0.0001854157186459915,
+      "loss": 3.0131,
+      "step": 2838
+    },
+    {
+      "epoch": 0.17464320866141733,
+      "grad_norm": 1.0264731645584106,
+      "learning_rate": 0.00018540566118909296,
+      "loss": 2.9411,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1747047244094488,
+      "grad_norm": 1.300167441368103,
+      "learning_rate": 0.00018539560053849893,
+      "loss": 2.9254,
+      "step": 2840
+    },
+    {
+      "epoch": 0.17476624015748032,
+      "grad_norm": 0.9648934602737427,
+      "learning_rate": 0.00018538553669458563,
+      "loss": 3.0534,
+      "step": 2841
+    },
+    {
+      "epoch": 0.1748277559055118,
+      "grad_norm": 0.8517898917198181,
+      "learning_rate": 0.00018537546965772937,
+      "loss": 3.1278,
+      "step": 2842
+    },
+    {
+      "epoch": 0.17488927165354332,
+      "grad_norm": 1.0680350065231323,
+      "learning_rate": 0.0001853653994283066,
+      "loss": 3.5517,
+      "step": 2843
+    },
+    {
+      "epoch": 0.1749507874015748,
+      "grad_norm": 1.1566479206085205,
+      "learning_rate": 0.00018535532600669392,
+      "loss": 3.2086,
+      "step": 2844
+    },
+    {
+      "epoch": 0.1750123031496063,
+      "grad_norm": 1.010288953781128,
+      "learning_rate": 0.000185345249393268,
+      "loss": 2.6997,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1750738188976378,
+      "grad_norm": 1.0967224836349487,
+      "learning_rate": 0.00018533516958840562,
+      "loss": 2.8852,
+      "step": 2846
+    },
+    {
+      "epoch": 0.1751353346456693,
+      "grad_norm": 1.1099971532821655,
+      "learning_rate": 0.00018532508659248377,
+      "loss": 2.6655,
+      "step": 2847
+    },
+    {
+      "epoch": 0.17519685039370078,
+      "grad_norm": 1.116829752922058,
+      "learning_rate": 0.00018531500040587946,
+      "loss": 2.2264,
+      "step": 2848
+    },
+    {
+      "epoch": 0.1752583661417323,
+      "grad_norm": 1.3192884922027588,
+      "learning_rate": 0.0001853049110289699,
+      "loss": 2.9471,
+      "step": 2849
+    },
+    {
+      "epoch": 0.17531988188976377,
+      "grad_norm": 1.0889003276824951,
+      "learning_rate": 0.0001852948184621323,
+      "loss": 2.7823,
+      "step": 2850
+    },
+    {
+      "epoch": 0.17538139763779528,
+      "grad_norm": 1.0434668064117432,
+      "learning_rate": 0.0001852847227057441,
+      "loss": 3.1126,
+      "step": 2851
+    },
+    {
+      "epoch": 0.17544291338582677,
+      "grad_norm": 0.9788315296173096,
+      "learning_rate": 0.00018527462376018286,
+      "loss": 3.3315,
+      "step": 2852
+    },
+    {
+      "epoch": 0.17550442913385828,
+      "grad_norm": 0.9277727007865906,
+      "learning_rate": 0.00018526452162582617,
+      "loss": 3.303,
+      "step": 2853
+    },
+    {
+      "epoch": 0.17556594488188976,
+      "grad_norm": 1.11817467212677,
+      "learning_rate": 0.00018525441630305186,
+      "loss": 3.4043,
+      "step": 2854
+    },
+    {
+      "epoch": 0.17562746062992127,
+      "grad_norm": 1.7669360637664795,
+      "learning_rate": 0.0001852443077922377,
+      "loss": 2.1953,
+      "step": 2855
+    },
+    {
+      "epoch": 0.17568897637795275,
+      "grad_norm": 1.3561985492706299,
+      "learning_rate": 0.00018523419609376183,
+      "loss": 2.7444,
+      "step": 2856
+    },
+    {
+      "epoch": 0.17575049212598426,
+      "grad_norm": 1.0123497247695923,
+      "learning_rate": 0.00018522408120800226,
+      "loss": 3.3491,
+      "step": 2857
+    },
+    {
+      "epoch": 0.17581200787401574,
+      "grad_norm": 1.0606507062911987,
+      "learning_rate": 0.0001852139631353373,
+      "loss": 3.32,
+      "step": 2858
+    },
+    {
+      "epoch": 0.17587352362204725,
+      "grad_norm": 1.3128842115402222,
+      "learning_rate": 0.0001852038418761453,
+      "loss": 2.45,
+      "step": 2859
+    },
+    {
+      "epoch": 0.17593503937007873,
+      "grad_norm": 0.9172083139419556,
+      "learning_rate": 0.00018519371743080468,
+      "loss": 2.9034,
+      "step": 2860
+    },
+    {
+      "epoch": 0.17599655511811024,
+      "grad_norm": 1.2387657165527344,
+      "learning_rate": 0.00018518358979969412,
+      "loss": 2.5456,
+      "step": 2861
+    },
+    {
+      "epoch": 0.17605807086614172,
+      "grad_norm": 1.0548374652862549,
+      "learning_rate": 0.00018517345898319228,
+      "loss": 2.8032,
+      "step": 2862
+    },
+    {
+      "epoch": 0.17611958661417323,
+      "grad_norm": 1.0954030752182007,
+      "learning_rate": 0.00018516332498167803,
+      "loss": 3.406,
+      "step": 2863
+    },
+    {
+      "epoch": 0.17618110236220472,
+      "grad_norm": 0.8940899968147278,
+      "learning_rate": 0.0001851531877955303,
+      "loss": 3.0392,
+      "step": 2864
+    },
+    {
+      "epoch": 0.17624261811023623,
+      "grad_norm": 1.05564546585083,
+      "learning_rate": 0.00018514304742512817,
+      "loss": 3.2299,
+      "step": 2865
+    },
+    {
+      "epoch": 0.1763041338582677,
+      "grad_norm": 1.0954720973968506,
+      "learning_rate": 0.00018513290387085082,
+      "loss": 2.8279,
+      "step": 2866
+    },
+    {
+      "epoch": 0.17636564960629922,
+      "grad_norm": 1.1944141387939453,
+      "learning_rate": 0.0001851227571330776,
+      "loss": 3.0915,
+      "step": 2867
+    },
+    {
+      "epoch": 0.1764271653543307,
+      "grad_norm": 1.3696914911270142,
+      "learning_rate": 0.00018511260721218792,
+      "loss": 2.6339,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1764886811023622,
+      "grad_norm": 0.9730051159858704,
+      "learning_rate": 0.00018510245410856136,
+      "loss": 3.2137,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1765501968503937,
+      "grad_norm": 1.1804161071777344,
+      "learning_rate": 0.0001850922978225775,
+      "loss": 3.6051,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1766117125984252,
+      "grad_norm": 1.1746288537979126,
+      "learning_rate": 0.00018508213835461622,
+      "loss": 2.6873,
+      "step": 2871
+    },
+    {
+      "epoch": 0.17667322834645668,
+      "grad_norm": 1.3410485982894897,
+      "learning_rate": 0.00018507197570505742,
+      "loss": 3.3789,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1767347440944882,
+      "grad_norm": 1.3174092769622803,
+      "learning_rate": 0.00018506180987428106,
+      "loss": 2.8925,
+      "step": 2873
+    },
+    {
+      "epoch": 0.17679625984251968,
+      "grad_norm": 1.135407567024231,
+      "learning_rate": 0.00018505164086266736,
+      "loss": 2.6138,
+      "step": 2874
+    },
+    {
+      "epoch": 0.1768577755905512,
+      "grad_norm": 1.997048258781433,
+      "learning_rate": 0.00018504146867059652,
+      "loss": 3.0405,
+      "step": 2875
+    },
+    {
+      "epoch": 0.17691929133858267,
+      "grad_norm": 1.2101266384124756,
+      "learning_rate": 0.000185031293298449,
+      "loss": 1.562,
+      "step": 2876
+    },
+    {
+      "epoch": 0.17698080708661418,
+      "grad_norm": 1.0764319896697998,
+      "learning_rate": 0.0001850211147466052,
+      "loss": 3.0703,
+      "step": 2877
+    },
+    {
+      "epoch": 0.17704232283464566,
+      "grad_norm": 1.3972926139831543,
+      "learning_rate": 0.00018501093301544585,
+      "loss": 2.5931,
+      "step": 2878
+    },
+    {
+      "epoch": 0.17710383858267717,
+      "grad_norm": 1.1193838119506836,
+      "learning_rate": 0.00018500074810535164,
+      "loss": 2.999,
+      "step": 2879
+    },
+    {
+      "epoch": 0.17716535433070865,
+      "grad_norm": 1.0568618774414062,
+      "learning_rate": 0.0001849905600167034,
+      "loss": 2.7561,
+      "step": 2880
+    },
+    {
+      "epoch": 0.17722687007874016,
+      "grad_norm": 1.1318861246109009,
+      "learning_rate": 0.00018498036874988215,
+      "loss": 2.8063,
+      "step": 2881
+    },
+    {
+      "epoch": 0.17728838582677164,
+      "grad_norm": 1.0891456604003906,
+      "learning_rate": 0.00018497017430526896,
+      "loss": 2.5312,
+      "step": 2882
+    },
+    {
+      "epoch": 0.17734990157480315,
+      "grad_norm": 1.1694058179855347,
+      "learning_rate": 0.00018495997668324505,
+      "loss": 2.8105,
+      "step": 2883
+    },
+    {
+      "epoch": 0.17741141732283464,
+      "grad_norm": 1.129694938659668,
+      "learning_rate": 0.00018494977588419178,
+      "loss": 2.3462,
+      "step": 2884
+    },
+    {
+      "epoch": 0.17747293307086615,
+      "grad_norm": 1.1720085144042969,
+      "learning_rate": 0.00018493957190849058,
+      "loss": 3.2521,
+      "step": 2885
+    },
+    {
+      "epoch": 0.17753444881889763,
+      "grad_norm": 0.982786238193512,
+      "learning_rate": 0.00018492936475652302,
+      "loss": 3.3271,
+      "step": 2886
+    },
+    {
+      "epoch": 0.17759596456692914,
+      "grad_norm": 1.1270408630371094,
+      "learning_rate": 0.00018491915442867086,
+      "loss": 2.6575,
+      "step": 2887
+    },
+    {
+      "epoch": 0.17765748031496062,
+      "grad_norm": 0.9588005542755127,
+      "learning_rate": 0.00018490894092531576,
+      "loss": 3.0686,
+      "step": 2888
+    },
+    {
+      "epoch": 0.17771899606299213,
+      "grad_norm": 1.1300872564315796,
+      "learning_rate": 0.00018489872424683976,
+      "loss": 3.2458,
+      "step": 2889
+    },
+    {
+      "epoch": 0.1777805118110236,
+      "grad_norm": 0.9897785782814026,
+      "learning_rate": 0.0001848885043936249,
+      "loss": 3.6513,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17784202755905512,
+      "grad_norm": 0.9911032915115356,
+      "learning_rate": 0.00018487828136605331,
+      "loss": 2.9246,
+      "step": 2891
+    },
+    {
+      "epoch": 0.1779035433070866,
+      "grad_norm": 1.0441985130310059,
+      "learning_rate": 0.00018486805516450733,
+      "loss": 3.3429,
+      "step": 2892
+    },
+    {
+      "epoch": 0.1779650590551181,
+      "grad_norm": 1.248113751411438,
+      "learning_rate": 0.00018485782578936928,
+      "loss": 2.7709,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1780265748031496,
+      "grad_norm": 1.037109375,
+      "learning_rate": 0.0001848475932410217,
+      "loss": 3.21,
+      "step": 2894
+    },
+    {
+      "epoch": 0.1780880905511811,
+      "grad_norm": 1.0076333284378052,
+      "learning_rate": 0.0001848373575198473,
+      "loss": 3.0321,
+      "step": 2895
+    },
+    {
+      "epoch": 0.1781496062992126,
+      "grad_norm": 1.010467290878296,
+      "learning_rate": 0.0001848271186262288,
+      "loss": 3.4241,
+      "step": 2896
+    },
+    {
+      "epoch": 0.1782111220472441,
+      "grad_norm": 1.0540573596954346,
+      "learning_rate": 0.00018481687656054908,
+      "loss": 3.075,
+      "step": 2897
+    },
+    {
+      "epoch": 0.17827263779527558,
+      "grad_norm": 1.097795009613037,
+      "learning_rate": 0.0001848066313231911,
+      "loss": 2.7237,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1783341535433071,
+      "grad_norm": 1.2217146158218384,
+      "learning_rate": 0.000184796382914538,
+      "loss": 2.8402,
+      "step": 2899
+    },
+    {
+      "epoch": 0.17839566929133857,
+      "grad_norm": 0.9728935956954956,
+      "learning_rate": 0.00018478613133497302,
+      "loss": 3.0878,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17845718503937008,
+      "grad_norm": 1.3414571285247803,
+      "learning_rate": 0.00018477587658487953,
+      "loss": 3.0636,
+      "step": 2901
+    },
+    {
+      "epoch": 0.17851870078740156,
+      "grad_norm": 1.212882399559021,
+      "learning_rate": 0.00018476561866464098,
+      "loss": 2.8864,
+      "step": 2902
+    },
+    {
+      "epoch": 0.17858021653543307,
+      "grad_norm": 1.096276044845581,
+      "learning_rate": 0.00018475535757464092,
+      "loss": 3.2268,
+      "step": 2903
+    },
+    {
+      "epoch": 0.17864173228346455,
+      "grad_norm": 1.0507240295410156,
+      "learning_rate": 0.00018474509331526314,
+      "loss": 2.9792,
+      "step": 2904
+    },
+    {
+      "epoch": 0.17870324803149606,
+      "grad_norm": 1.004919171333313,
+      "learning_rate": 0.0001847348258868914,
+      "loss": 2.8526,
+      "step": 2905
+    },
+    {
+      "epoch": 0.17876476377952755,
+      "grad_norm": 1.0350449085235596,
+      "learning_rate": 0.0001847245552899097,
+      "loss": 3.0216,
+      "step": 2906
+    },
+    {
+      "epoch": 0.17882627952755906,
+      "grad_norm": 0.9788486957550049,
+      "learning_rate": 0.000184714281524702,
+      "loss": 2.9235,
+      "step": 2907
+    },
+    {
+      "epoch": 0.17888779527559054,
+      "grad_norm": 1.0412235260009766,
+      "learning_rate": 0.0001847040045916526,
+      "loss": 2.5885,
+      "step": 2908
+    },
+    {
+      "epoch": 0.17894931102362205,
+      "grad_norm": 0.9957424402236938,
+      "learning_rate": 0.00018469372449114575,
+      "loss": 3.4598,
+      "step": 2909
+    },
+    {
+      "epoch": 0.17901082677165353,
+      "grad_norm": 1.1774275302886963,
+      "learning_rate": 0.00018468344122356584,
+      "loss": 2.6291,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17907234251968504,
+      "grad_norm": 0.9376084208488464,
+      "learning_rate": 0.00018467315478929745,
+      "loss": 2.9041,
+      "step": 2911
+    },
+    {
+      "epoch": 0.17913385826771652,
+      "grad_norm": 1.161729097366333,
+      "learning_rate": 0.00018466286518872523,
+      "loss": 2.8864,
+      "step": 2912
+    },
+    {
+      "epoch": 0.17919537401574803,
+      "grad_norm": 1.2143610715866089,
+      "learning_rate": 0.00018465257242223393,
+      "loss": 3.0935,
+      "step": 2913
+    },
+    {
+      "epoch": 0.17925688976377951,
+      "grad_norm": 1.0992376804351807,
+      "learning_rate": 0.0001846422764902085,
+      "loss": 2.9161,
+      "step": 2914
+    },
+    {
+      "epoch": 0.17931840551181102,
+      "grad_norm": 1.2431981563568115,
+      "learning_rate": 0.00018463197739303385,
+      "loss": 3.3876,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1793799212598425,
+      "grad_norm": 1.3986999988555908,
+      "learning_rate": 0.00018462167513109517,
+      "loss": 2.3047,
+      "step": 2916
+    },
+    {
+      "epoch": 0.17944143700787402,
+      "grad_norm": 1.1481038331985474,
+      "learning_rate": 0.00018461136970477775,
+      "loss": 3.4154,
+      "step": 2917
+    },
+    {
+      "epoch": 0.1795029527559055,
+      "grad_norm": 1.0479485988616943,
+      "learning_rate": 0.0001846010611144669,
+      "loss": 2.8128,
+      "step": 2918
+    },
+    {
+      "epoch": 0.179564468503937,
+      "grad_norm": 0.9980430006980896,
+      "learning_rate": 0.00018459074936054808,
+      "loss": 2.7069,
+      "step": 2919
+    },
+    {
+      "epoch": 0.1796259842519685,
+      "grad_norm": 0.9258325099945068,
+      "learning_rate": 0.00018458043444340695,
+      "loss": 2.8645,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1796875,
+      "grad_norm": 1.2205681800842285,
+      "learning_rate": 0.00018457011636342918,
+      "loss": 3.5522,
+      "step": 2921
+    },
+    {
+      "epoch": 0.1797490157480315,
+      "grad_norm": 1.0883231163024902,
+      "learning_rate": 0.00018455979512100066,
+      "loss": 2.0491,
+      "step": 2922
+    },
+    {
+      "epoch": 0.179810531496063,
+      "grad_norm": 0.9156938195228577,
+      "learning_rate": 0.00018454947071650733,
+      "loss": 3.1423,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1798720472440945,
+      "grad_norm": 0.9804782271385193,
+      "learning_rate": 0.0001845391431503352,
+      "loss": 2.9035,
+      "step": 2924
+    },
+    {
+      "epoch": 0.17993356299212598,
+      "grad_norm": 1.0590972900390625,
+      "learning_rate": 0.00018452881242287056,
+      "loss": 2.8182,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1799950787401575,
+      "grad_norm": 1.0150052309036255,
+      "learning_rate": 0.00018451847853449968,
+      "loss": 2.4926,
+      "step": 2926
+    },
+    {
+      "epoch": 0.18005659448818898,
+      "grad_norm": 1.2389357089996338,
+      "learning_rate": 0.000184508141485609,
+      "loss": 3.634,
+      "step": 2927
+    },
+    {
+      "epoch": 0.18011811023622049,
+      "grad_norm": 1.0588340759277344,
+      "learning_rate": 0.00018449780127658506,
+      "loss": 2.4754,
+      "step": 2928
+    },
+    {
+      "epoch": 0.18017962598425197,
+      "grad_norm": 1.0445685386657715,
+      "learning_rate": 0.0001844874579078145,
+      "loss": 3.5574,
+      "step": 2929
+    },
+    {
+      "epoch": 0.18024114173228348,
+      "grad_norm": 1.296515941619873,
+      "learning_rate": 0.00018447711137968415,
+      "loss": 3.1018,
+      "step": 2930
+    },
+    {
+      "epoch": 0.18030265748031496,
+      "grad_norm": 1.0268439054489136,
+      "learning_rate": 0.00018446676169258085,
+      "loss": 2.6423,
+      "step": 2931
+    },
+    {
+      "epoch": 0.18036417322834647,
+      "grad_norm": 1.206844449043274,
+      "learning_rate": 0.00018445640884689172,
+      "loss": 3.5034,
+      "step": 2932
+    },
+    {
+      "epoch": 0.18042568897637795,
+      "grad_norm": 1.0231800079345703,
+      "learning_rate": 0.00018444605284300382,
+      "loss": 2.8998,
+      "step": 2933
+    },
+    {
+      "epoch": 0.18048720472440946,
+      "grad_norm": 1.1329983472824097,
+      "learning_rate": 0.0001844356936813044,
+      "loss": 2.893,
+      "step": 2934
+    },
+    {
+      "epoch": 0.18054872047244094,
+      "grad_norm": 1.1636980772018433,
+      "learning_rate": 0.0001844253313621809,
+      "loss": 2.7921,
+      "step": 2935
+    },
+    {
+      "epoch": 0.18061023622047245,
+      "grad_norm": 1.3780182600021362,
+      "learning_rate": 0.00018441496588602076,
+      "loss": 2.7474,
+      "step": 2936
+    },
+    {
+      "epoch": 0.18067175196850394,
+      "grad_norm": 1.200681209564209,
+      "learning_rate": 0.00018440459725321161,
+      "loss": 2.2794,
+      "step": 2937
+    },
+    {
+      "epoch": 0.18073326771653545,
+      "grad_norm": 0.9521216154098511,
+      "learning_rate": 0.00018439422546414117,
+      "loss": 3.1492,
+      "step": 2938
+    },
+    {
+      "epoch": 0.18079478346456693,
+      "grad_norm": 1.0252766609191895,
+      "learning_rate": 0.00018438385051919732,
+      "loss": 2.2905,
+      "step": 2939
+    },
+    {
+      "epoch": 0.18085629921259844,
+      "grad_norm": 1.138851284980774,
+      "learning_rate": 0.00018437347241876797,
+      "loss": 2.9196,
+      "step": 2940
+    },
+    {
+      "epoch": 0.18091781496062992,
+      "grad_norm": 1.0616998672485352,
+      "learning_rate": 0.00018436309116324123,
+      "loss": 3.083,
+      "step": 2941
+    },
+    {
+      "epoch": 0.18097933070866143,
+      "grad_norm": 1.1267565488815308,
+      "learning_rate": 0.0001843527067530053,
+      "loss": 3.0686,
+      "step": 2942
+    },
+    {
+      "epoch": 0.1810408464566929,
+      "grad_norm": 1.040378451347351,
+      "learning_rate": 0.0001843423191884485,
+      "loss": 3.522,
+      "step": 2943
+    },
+    {
+      "epoch": 0.18110236220472442,
+      "grad_norm": 1.1794068813323975,
+      "learning_rate": 0.0001843319284699593,
+      "loss": 3.1299,
+      "step": 2944
+    },
+    {
+      "epoch": 0.1811638779527559,
+      "grad_norm": 1.0584663152694702,
+      "learning_rate": 0.00018432153459792622,
+      "loss": 3.1928,
+      "step": 2945
+    },
+    {
+      "epoch": 0.1812253937007874,
+      "grad_norm": 1.1733847856521606,
+      "learning_rate": 0.0001843111375727379,
+      "loss": 3.0753,
+      "step": 2946
+    },
+    {
+      "epoch": 0.1812869094488189,
+      "grad_norm": 0.8866992592811584,
+      "learning_rate": 0.00018430073739478315,
+      "loss": 3.1612,
+      "step": 2947
+    },
+    {
+      "epoch": 0.1813484251968504,
+      "grad_norm": 1.3738411664962769,
+      "learning_rate": 0.00018429033406445092,
+      "loss": 2.6393,
+      "step": 2948
+    },
+    {
+      "epoch": 0.1814099409448819,
+      "grad_norm": 1.0152157545089722,
+      "learning_rate": 0.00018427992758213025,
+      "loss": 3.0609,
+      "step": 2949
+    },
+    {
+      "epoch": 0.1814714566929134,
+      "grad_norm": 1.0058236122131348,
+      "learning_rate": 0.0001842695179482102,
+      "loss": 3.0905,
+      "step": 2950
+    },
+    {
+      "epoch": 0.18153297244094488,
+      "grad_norm": 0.952386200428009,
+      "learning_rate": 0.0001842591051630801,
+      "loss": 3.7145,
+      "step": 2951
+    },
+    {
+      "epoch": 0.1815944881889764,
+      "grad_norm": 1.4700262546539307,
+      "learning_rate": 0.0001842486892271293,
+      "loss": 3.1802,
+      "step": 2952
+    },
+    {
+      "epoch": 0.18165600393700787,
+      "grad_norm": 1.5509454011917114,
+      "learning_rate": 0.00018423827014074732,
+      "loss": 2.446,
+      "step": 2953
+    },
+    {
+      "epoch": 0.18171751968503938,
+      "grad_norm": 1.240302562713623,
+      "learning_rate": 0.00018422784790432374,
+      "loss": 3.2766,
+      "step": 2954
+    },
+    {
+      "epoch": 0.18177903543307086,
+      "grad_norm": 1.246774435043335,
+      "learning_rate": 0.00018421742251824834,
+      "loss": 3.3425,
+      "step": 2955
+    },
+    {
+      "epoch": 0.18184055118110237,
+      "grad_norm": 1.0675222873687744,
+      "learning_rate": 0.00018420699398291092,
+      "loss": 2.7784,
+      "step": 2956
+    },
+    {
+      "epoch": 0.18190206692913385,
+      "grad_norm": 1.0406126976013184,
+      "learning_rate": 0.0001841965622987015,
+      "loss": 2.7582,
+      "step": 2957
+    },
+    {
+      "epoch": 0.18196358267716536,
+      "grad_norm": 1.2244365215301514,
+      "learning_rate": 0.0001841861274660101,
+      "loss": 2.6498,
+      "step": 2958
+    },
+    {
+      "epoch": 0.18202509842519685,
+      "grad_norm": 1.4448777437210083,
+      "learning_rate": 0.00018417568948522704,
+      "loss": 2.5945,
+      "step": 2959
+    },
+    {
+      "epoch": 0.18208661417322836,
+      "grad_norm": 1.1902190446853638,
+      "learning_rate": 0.0001841652483567425,
+      "loss": 2.8037,
+      "step": 2960
+    },
+    {
+      "epoch": 0.18214812992125984,
+      "grad_norm": 1.0696686506271362,
+      "learning_rate": 0.00018415480408094705,
+      "loss": 3.145,
+      "step": 2961
+    },
+    {
+      "epoch": 0.18220964566929135,
+      "grad_norm": 1.2468657493591309,
+      "learning_rate": 0.00018414435665823112,
+      "loss": 3.045,
+      "step": 2962
+    },
+    {
+      "epoch": 0.18227116141732283,
+      "grad_norm": 1.0326061248779297,
+      "learning_rate": 0.0001841339060889855,
+      "loss": 2.8396,
+      "step": 2963
+    },
+    {
+      "epoch": 0.18233267716535434,
+      "grad_norm": 1.2029377222061157,
+      "learning_rate": 0.00018412345237360094,
+      "loss": 3.072,
+      "step": 2964
+    },
+    {
+      "epoch": 0.18239419291338582,
+      "grad_norm": 1.2612524032592773,
+      "learning_rate": 0.00018411299551246833,
+      "loss": 3.095,
+      "step": 2965
+    },
+    {
+      "epoch": 0.18245570866141733,
+      "grad_norm": 1.4041829109191895,
+      "learning_rate": 0.0001841025355059787,
+      "loss": 3.1327,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1825172244094488,
+      "grad_norm": 0.9118495583534241,
+      "learning_rate": 0.00018409207235452324,
+      "loss": 3.3206,
+      "step": 2967
+    },
+    {
+      "epoch": 0.18257874015748032,
+      "grad_norm": 1.0807090997695923,
+      "learning_rate": 0.0001840816060584932,
+      "loss": 2.8646,
+      "step": 2968
+    },
+    {
+      "epoch": 0.1826402559055118,
+      "grad_norm": 1.1355419158935547,
+      "learning_rate": 0.0001840711366182799,
+      "loss": 2.6892,
+      "step": 2969
+    },
+    {
+      "epoch": 0.18270177165354332,
+      "grad_norm": 1.080436110496521,
+      "learning_rate": 0.0001840606640342749,
+      "loss": 2.9004,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1827632874015748,
+      "grad_norm": 0.9702497124671936,
+      "learning_rate": 0.0001840501883068698,
+      "loss": 2.646,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1828248031496063,
+      "grad_norm": 1.5088410377502441,
+      "learning_rate": 0.00018403970943645635,
+      "loss": 3.0654,
+      "step": 2972
+    },
+    {
+      "epoch": 0.1828863188976378,
+      "grad_norm": 2.0557334423065186,
+      "learning_rate": 0.00018402922742342636,
+      "loss": 3.5693,
+      "step": 2973
+    },
+    {
+      "epoch": 0.1829478346456693,
+      "grad_norm": 1.9035929441452026,
+      "learning_rate": 0.00018401874226817185,
+      "loss": 2.6399,
+      "step": 2974
+    },
+    {
+      "epoch": 0.18300935039370078,
+      "grad_norm": 1.2460986375808716,
+      "learning_rate": 0.00018400825397108485,
+      "loss": 3.3273,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1830708661417323,
+      "grad_norm": 1.1875721216201782,
+      "learning_rate": 0.0001839977625325576,
+      "loss": 3.1005,
+      "step": 2976
+    },
+    {
+      "epoch": 0.18313238188976377,
+      "grad_norm": 1.0150827169418335,
+      "learning_rate": 0.00018398726795298243,
+      "loss": 2.7646,
+      "step": 2977
+    },
+    {
+      "epoch": 0.18319389763779528,
+      "grad_norm": 1.151157021522522,
+      "learning_rate": 0.00018397677023275177,
+      "loss": 3.5178,
+      "step": 2978
+    },
+    {
+      "epoch": 0.18325541338582677,
+      "grad_norm": 1.3720941543579102,
+      "learning_rate": 0.00018396626937225815,
+      "loss": 2.9391,
+      "step": 2979
+    },
+    {
+      "epoch": 0.18331692913385828,
+      "grad_norm": 1.2879160642623901,
+      "learning_rate": 0.00018395576537189428,
+      "loss": 2.5229,
+      "step": 2980
+    },
+    {
+      "epoch": 0.18337844488188976,
+      "grad_norm": 1.0771362781524658,
+      "learning_rate": 0.00018394525823205293,
+      "loss": 2.6539,
+      "step": 2981
+    },
+    {
+      "epoch": 0.18343996062992127,
+      "grad_norm": 1.0939757823944092,
+      "learning_rate": 0.00018393474795312702,
+      "loss": 3.0742,
+      "step": 2982
+    },
+    {
+      "epoch": 0.18350147637795275,
+      "grad_norm": 1.1103620529174805,
+      "learning_rate": 0.00018392423453550958,
+      "loss": 3.1332,
+      "step": 2983
+    },
+    {
+      "epoch": 0.18356299212598426,
+      "grad_norm": 1.0912883281707764,
+      "learning_rate": 0.0001839137179795937,
+      "loss": 2.9129,
+      "step": 2984
+    },
+    {
+      "epoch": 0.18362450787401574,
+      "grad_norm": 1.1464272737503052,
+      "learning_rate": 0.00018390319828577272,
+      "loss": 3.0656,
+      "step": 2985
+    },
+    {
+      "epoch": 0.18368602362204725,
+      "grad_norm": 1.238051176071167,
+      "learning_rate": 0.00018389267545443997,
+      "loss": 2.4759,
+      "step": 2986
+    },
+    {
+      "epoch": 0.18374753937007873,
+      "grad_norm": 0.956144392490387,
+      "learning_rate": 0.00018388214948598895,
+      "loss": 2.8499,
+      "step": 2987
+    },
+    {
+      "epoch": 0.18380905511811024,
+      "grad_norm": 0.9345887899398804,
+      "learning_rate": 0.0001838716203808133,
+      "loss": 2.6497,
+      "step": 2988
+    },
+    {
+      "epoch": 0.18387057086614172,
+      "grad_norm": 1.1850191354751587,
+      "learning_rate": 0.0001838610881393067,
+      "loss": 2.2171,
+      "step": 2989
+    },
+    {
+      "epoch": 0.18393208661417323,
+      "grad_norm": 1.4400103092193604,
+      "learning_rate": 0.00018385055276186305,
+      "loss": 2.6551,
+      "step": 2990
+    },
+    {
+      "epoch": 0.18399360236220472,
+      "grad_norm": 1.3012548685073853,
+      "learning_rate": 0.00018384001424887628,
+      "loss": 2.9764,
+      "step": 2991
+    },
+    {
+      "epoch": 0.18405511811023623,
+      "grad_norm": 1.1697193384170532,
+      "learning_rate": 0.0001838294726007405,
+      "loss": 3.0343,
+      "step": 2992
+    },
+    {
+      "epoch": 0.1841166338582677,
+      "grad_norm": 1.8563634157180786,
+      "learning_rate": 0.00018381892781784985,
+      "loss": 2.7118,
+      "step": 2993
+    },
+    {
+      "epoch": 0.18417814960629922,
+      "grad_norm": 1.304090976715088,
+      "learning_rate": 0.0001838083799005987,
+      "loss": 3.1914,
+      "step": 2994
+    },
+    {
+      "epoch": 0.1842396653543307,
+      "grad_norm": 1.0712743997573853,
+      "learning_rate": 0.0001837978288493815,
+      "loss": 2.641,
+      "step": 2995
+    },
+    {
+      "epoch": 0.1843011811023622,
+      "grad_norm": 0.9369164705276489,
+      "learning_rate": 0.00018378727466459273,
+      "loss": 3.0756,
+      "step": 2996
+    },
+    {
+      "epoch": 0.1843626968503937,
+      "grad_norm": 1.0552769899368286,
+      "learning_rate": 0.00018377671734662714,
+      "loss": 3.4769,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1844242125984252,
+      "grad_norm": 1.6638706922531128,
+      "learning_rate": 0.00018376615689587946,
+      "loss": 3.2336,
+      "step": 2998
+    },
+    {
+      "epoch": 0.18448572834645668,
+      "grad_norm": 1.5234099626541138,
+      "learning_rate": 0.0001837555933127446,
+      "loss": 2.1025,
+      "step": 2999
+    },
+    {
+      "epoch": 0.1845472440944882,
+      "grad_norm": 1.011968731880188,
+      "learning_rate": 0.00018374502659761755,
+      "loss": 3.0046,
+      "step": 3000
+    },
+    {
+      "epoch": 0.18460875984251968,
+      "grad_norm": 1.0500223636627197,
+      "learning_rate": 0.00018373445675089352,
+      "loss": 3.0212,
+      "step": 3001
+    },
+    {
+      "epoch": 0.1846702755905512,
+      "grad_norm": 1.1380966901779175,
+      "learning_rate": 0.0001837238837729677,
+      "loss": 2.719,
+      "step": 3002
+    },
+    {
+      "epoch": 0.18473179133858267,
+      "grad_norm": 1.392674446105957,
+      "learning_rate": 0.00018371330766423554,
+      "loss": 3.1892,
+      "step": 3003
+    },
+    {
+      "epoch": 0.18479330708661418,
+      "grad_norm": 1.2562915086746216,
+      "learning_rate": 0.00018370272842509243,
+      "loss": 3.0588,
+      "step": 3004
+    },
+    {
+      "epoch": 0.18485482283464566,
+      "grad_norm": 0.9984306693077087,
+      "learning_rate": 0.00018369214605593398,
+      "loss": 3.0526,
+      "step": 3005
+    },
+    {
+      "epoch": 0.18491633858267717,
+      "grad_norm": 0.9991692304611206,
+      "learning_rate": 0.00018368156055715602,
+      "loss": 2.5125,
+      "step": 3006
+    },
+    {
+      "epoch": 0.18497785433070865,
+      "grad_norm": 1.188226342201233,
+      "learning_rate": 0.00018367097192915425,
+      "loss": 3.5862,
+      "step": 3007
+    },
+    {
+      "epoch": 0.18503937007874016,
+      "grad_norm": 1.1143920421600342,
+      "learning_rate": 0.00018366038017232473,
+      "loss": 3.2779,
+      "step": 3008
+    },
+    {
+      "epoch": 0.18510088582677164,
+      "grad_norm": 0.9553026556968689,
+      "learning_rate": 0.0001836497852870635,
+      "loss": 2.8514,
+      "step": 3009
+    },
+    {
+      "epoch": 0.18516240157480315,
+      "grad_norm": 1.2449262142181396,
+      "learning_rate": 0.00018363918727376673,
+      "loss": 2.4882,
+      "step": 3010
+    },
+    {
+      "epoch": 0.18522391732283464,
+      "grad_norm": 1.2078685760498047,
+      "learning_rate": 0.00018362858613283075,
+      "loss": 3.094,
+      "step": 3011
+    },
+    {
+      "epoch": 0.18528543307086615,
+      "grad_norm": 1.1165739297866821,
+      "learning_rate": 0.00018361798186465196,
+      "loss": 2.8023,
+      "step": 3012
+    },
+    {
+      "epoch": 0.18534694881889763,
+      "grad_norm": 1.1698477268218994,
+      "learning_rate": 0.0001836073744696269,
+      "loss": 3.5701,
+      "step": 3013
+    },
+    {
+      "epoch": 0.18540846456692914,
+      "grad_norm": 1.298369288444519,
+      "learning_rate": 0.00018359676394815228,
+      "loss": 1.9174,
+      "step": 3014
+    },
+    {
+      "epoch": 0.18546998031496062,
+      "grad_norm": 1.154202938079834,
+      "learning_rate": 0.00018358615030062482,
+      "loss": 2.7685,
+      "step": 3015
+    },
+    {
+      "epoch": 0.18553149606299213,
+      "grad_norm": 1.0447369813919067,
+      "learning_rate": 0.00018357553352744144,
+      "loss": 3.4277,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1855930118110236,
+      "grad_norm": 1.1771539449691772,
+      "learning_rate": 0.0001835649136289991,
+      "loss": 2.5921,
+      "step": 3017
+    },
+    {
+      "epoch": 0.18565452755905512,
+      "grad_norm": 1.190240740776062,
+      "learning_rate": 0.000183554290605695,
+      "loss": 3.146,
+      "step": 3018
+    },
+    {
+      "epoch": 0.1857160433070866,
+      "grad_norm": 0.9530576467514038,
+      "learning_rate": 0.00018354366445792636,
+      "loss": 3.2184,
+      "step": 3019
+    },
+    {
+      "epoch": 0.1857775590551181,
+      "grad_norm": 1.1643997430801392,
+      "learning_rate": 0.00018353303518609047,
+      "loss": 2.4643,
+      "step": 3020
+    },
+    {
+      "epoch": 0.1858390748031496,
+      "grad_norm": 1.3893699645996094,
+      "learning_rate": 0.00018352240279058486,
+      "loss": 2.4978,
+      "step": 3021
+    },
+    {
+      "epoch": 0.1859005905511811,
+      "grad_norm": 1.1915732622146606,
+      "learning_rate": 0.00018351176727180715,
+      "loss": 2.9179,
+      "step": 3022
+    },
+    {
+      "epoch": 0.1859621062992126,
+      "grad_norm": 1.0331627130508423,
+      "learning_rate": 0.00018350112863015498,
+      "loss": 3.1606,
+      "step": 3023
+    },
+    {
+      "epoch": 0.1860236220472441,
+      "grad_norm": 1.0317456722259521,
+      "learning_rate": 0.00018349048686602625,
+      "loss": 3.0172,
+      "step": 3024
+    },
+    {
+      "epoch": 0.18608513779527558,
+      "grad_norm": 1.1215744018554688,
+      "learning_rate": 0.00018347984197981886,
+      "loss": 3.0437,
+      "step": 3025
+    },
+    {
+      "epoch": 0.1861466535433071,
+      "grad_norm": 1.0742173194885254,
+      "learning_rate": 0.00018346919397193087,
+      "loss": 2.9614,
+      "step": 3026
+    },
+    {
+      "epoch": 0.18620816929133857,
+      "grad_norm": 1.2000253200531006,
+      "learning_rate": 0.00018345854284276047,
+      "loss": 3.1934,
+      "step": 3027
+    },
+    {
+      "epoch": 0.18626968503937008,
+      "grad_norm": 1.1511261463165283,
+      "learning_rate": 0.00018344788859270595,
+      "loss": 2.5015,
+      "step": 3028
+    },
+    {
+      "epoch": 0.18633120078740156,
+      "grad_norm": 1.3722708225250244,
+      "learning_rate": 0.0001834372312221657,
+      "loss": 2.9504,
+      "step": 3029
+    },
+    {
+      "epoch": 0.18639271653543307,
+      "grad_norm": 0.9201224446296692,
+      "learning_rate": 0.00018342657073153827,
+      "loss": 3.4844,
+      "step": 3030
+    },
+    {
+      "epoch": 0.18645423228346455,
+      "grad_norm": 0.9884334802627563,
+      "learning_rate": 0.0001834159071212223,
+      "loss": 2.5886,
+      "step": 3031
+    },
+    {
+      "epoch": 0.18651574803149606,
+      "grad_norm": 1.0891361236572266,
+      "learning_rate": 0.00018340524039161653,
+      "loss": 2.8732,
+      "step": 3032
+    },
+    {
+      "epoch": 0.18657726377952755,
+      "grad_norm": 1.1983774900436401,
+      "learning_rate": 0.00018339457054311983,
+      "loss": 3.7466,
+      "step": 3033
+    },
+    {
+      "epoch": 0.18663877952755906,
+      "grad_norm": 1.0760478973388672,
+      "learning_rate": 0.00018338389757613125,
+      "loss": 3.2306,
+      "step": 3034
+    },
+    {
+      "epoch": 0.18670029527559054,
+      "grad_norm": 1.1159639358520508,
+      "learning_rate": 0.00018337322149104986,
+      "loss": 2.5739,
+      "step": 3035
+    },
+    {
+      "epoch": 0.18676181102362205,
+      "grad_norm": 1.2417421340942383,
+      "learning_rate": 0.00018336254228827484,
+      "loss": 2.9805,
+      "step": 3036
+    },
+    {
+      "epoch": 0.18682332677165353,
+      "grad_norm": 1.1989775896072388,
+      "learning_rate": 0.00018335185996820564,
+      "loss": 2.9938,
+      "step": 3037
+    },
+    {
+      "epoch": 0.18688484251968504,
+      "grad_norm": 1.2163358926773071,
+      "learning_rate": 0.00018334117453124162,
+      "loss": 3.3261,
+      "step": 3038
+    },
+    {
+      "epoch": 0.18694635826771652,
+      "grad_norm": 1.024832010269165,
+      "learning_rate": 0.00018333048597778242,
+      "loss": 2.7109,
+      "step": 3039
+    },
+    {
+      "epoch": 0.18700787401574803,
+      "grad_norm": 1.0746228694915771,
+      "learning_rate": 0.0001833197943082277,
+      "loss": 3.101,
+      "step": 3040
+    },
+    {
+      "epoch": 0.18706938976377951,
+      "grad_norm": 0.9560264348983765,
+      "learning_rate": 0.00018330909952297728,
+      "loss": 2.5183,
+      "step": 3041
+    },
+    {
+      "epoch": 0.18713090551181102,
+      "grad_norm": 1.8093063831329346,
+      "learning_rate": 0.0001832984016224311,
+      "loss": 3.0613,
+      "step": 3042
+    },
+    {
+      "epoch": 0.1871924212598425,
+      "grad_norm": 1.1308261156082153,
+      "learning_rate": 0.00018328770060698914,
+      "loss": 3.5462,
+      "step": 3043
+    },
+    {
+      "epoch": 0.18725393700787402,
+      "grad_norm": 1.1930819749832153,
+      "learning_rate": 0.00018327699647705163,
+      "loss": 2.7396,
+      "step": 3044
+    },
+    {
+      "epoch": 0.1873154527559055,
+      "grad_norm": 1.103249192237854,
+      "learning_rate": 0.00018326628923301882,
+      "loss": 2.0779,
+      "step": 3045
+    },
+    {
+      "epoch": 0.187376968503937,
+      "grad_norm": 0.9838024377822876,
+      "learning_rate": 0.00018325557887529108,
+      "loss": 3.0339,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1874384842519685,
+      "grad_norm": 0.8826131820678711,
+      "learning_rate": 0.00018324486540426894,
+      "loss": 2.8649,
+      "step": 3047
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 1.0327280759811401,
+      "learning_rate": 0.00018323414882035304,
+      "loss": 2.8945,
+      "step": 3048
+    },
+    {
+      "epoch": 0.1875615157480315,
+      "grad_norm": 1.1037249565124512,
+      "learning_rate": 0.00018322342912394408,
+      "loss": 2.9033,
+      "step": 3049
+    },
+    {
+      "epoch": 0.187623031496063,
+      "grad_norm": 1.0303810834884644,
+      "learning_rate": 0.00018321270631544293,
+      "loss": 3.507,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1876845472440945,
+      "grad_norm": 1.1055032014846802,
+      "learning_rate": 0.0001832019803952506,
+      "loss": 2.735,
+      "step": 3051
+    },
+    {
+      "epoch": 0.18774606299212598,
+      "grad_norm": 1.1260355710983276,
+      "learning_rate": 0.00018319125136376813,
+      "loss": 3.0353,
+      "step": 3052
+    },
+    {
+      "epoch": 0.1878075787401575,
+      "grad_norm": 1.5674724578857422,
+      "learning_rate": 0.00018318051922139677,
+      "loss": 2.4425,
+      "step": 3053
+    },
+    {
+      "epoch": 0.18786909448818898,
+      "grad_norm": 0.9683563113212585,
+      "learning_rate": 0.0001831697839685378,
+      "loss": 3.3272,
+      "step": 3054
+    },
+    {
+      "epoch": 0.18793061023622049,
+      "grad_norm": 1.0325061082839966,
+      "learning_rate": 0.00018315904560559267,
+      "loss": 3.3879,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18799212598425197,
+      "grad_norm": 1.175681233406067,
+      "learning_rate": 0.00018314830413296296,
+      "loss": 2.7779,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18805364173228348,
+      "grad_norm": 1.2323721647262573,
+      "learning_rate": 0.0001831375595510503,
+      "loss": 2.8852,
+      "step": 3057
+    },
+    {
+      "epoch": 0.18811515748031496,
+      "grad_norm": 1.0832608938217163,
+      "learning_rate": 0.00018312681186025652,
+      "loss": 2.7863,
+      "step": 3058
+    },
+    {
+      "epoch": 0.18817667322834647,
+      "grad_norm": 1.0584874153137207,
+      "learning_rate": 0.00018311606106098349,
+      "loss": 2.9237,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18823818897637795,
+      "grad_norm": 1.0021460056304932,
+      "learning_rate": 0.00018310530715363323,
+      "loss": 2.764,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18829970472440946,
+      "grad_norm": 1.0465071201324463,
+      "learning_rate": 0.00018309455013860792,
+      "loss": 3.2562,
+      "step": 3061
+    },
+    {
+      "epoch": 0.18836122047244094,
+      "grad_norm": 1.0747777223587036,
+      "learning_rate": 0.00018308379001630979,
+      "loss": 3.0076,
+      "step": 3062
+    },
+    {
+      "epoch": 0.18842273622047245,
+      "grad_norm": 1.0633777379989624,
+      "learning_rate": 0.00018307302678714118,
+      "loss": 2.9929,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18848425196850394,
+      "grad_norm": 0.9886115193367004,
+      "learning_rate": 0.0001830622604515046,
+      "loss": 3.087,
+      "step": 3064
+    },
+    {
+      "epoch": 0.18854576771653545,
+      "grad_norm": 1.0991543531417847,
+      "learning_rate": 0.00018305149100980263,
+      "loss": 3.0287,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18860728346456693,
+      "grad_norm": 2.0461297035217285,
+      "learning_rate": 0.00018304071846243804,
+      "loss": 2.0251,
+      "step": 3066
+    },
+    {
+      "epoch": 0.18866879921259844,
+      "grad_norm": 1.1459364891052246,
+      "learning_rate": 0.00018302994280981362,
+      "loss": 3.2141,
+      "step": 3067
+    },
+    {
+      "epoch": 0.18873031496062992,
+      "grad_norm": 1.10838782787323,
+      "learning_rate": 0.0001830191640523323,
+      "loss": 3.1557,
+      "step": 3068
+    },
+    {
+      "epoch": 0.18879183070866143,
+      "grad_norm": 1.2815150022506714,
+      "learning_rate": 0.0001830083821903972,
+      "loss": 2.9951,
+      "step": 3069
+    },
+    {
+      "epoch": 0.1888533464566929,
+      "grad_norm": 0.96079021692276,
+      "learning_rate": 0.00018299759722441146,
+      "loss": 3.4039,
+      "step": 3070
+    },
+    {
+      "epoch": 0.18891486220472442,
+      "grad_norm": 1.3475425243377686,
+      "learning_rate": 0.0001829868091547784,
+      "loss": 2.1434,
+      "step": 3071
+    },
+    {
+      "epoch": 0.1889763779527559,
+      "grad_norm": 1.1631135940551758,
+      "learning_rate": 0.00018297601798190137,
+      "loss": 2.9713,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1890378937007874,
+      "grad_norm": 1.3237996101379395,
+      "learning_rate": 0.00018296522370618402,
+      "loss": 2.2829,
+      "step": 3073
+    },
+    {
+      "epoch": 0.1890994094488189,
+      "grad_norm": 1.0038962364196777,
+      "learning_rate": 0.00018295442632802994,
+      "loss": 2.2761,
+      "step": 3074
+    },
+    {
+      "epoch": 0.1891609251968504,
+      "grad_norm": 1.0697776079177856,
+      "learning_rate": 0.00018294362584784284,
+      "loss": 2.5937,
+      "step": 3075
+    },
+    {
+      "epoch": 0.1892224409448819,
+      "grad_norm": 1.1121735572814941,
+      "learning_rate": 0.00018293282226602666,
+      "loss": 3.4037,
+      "step": 3076
+    },
+    {
+      "epoch": 0.1892839566929134,
+      "grad_norm": 0.9752725958824158,
+      "learning_rate": 0.00018292201558298534,
+      "loss": 3.1354,
+      "step": 3077
+    },
+    {
+      "epoch": 0.18934547244094488,
+      "grad_norm": 1.2082483768463135,
+      "learning_rate": 0.00018291120579912306,
+      "loss": 2.6571,
+      "step": 3078
+    },
+    {
+      "epoch": 0.1894069881889764,
+      "grad_norm": 2.0657386779785156,
+      "learning_rate": 0.000182900392914844,
+      "loss": 2.2221,
+      "step": 3079
+    },
+    {
+      "epoch": 0.18946850393700787,
+      "grad_norm": 1.0368775129318237,
+      "learning_rate": 0.0001828895769305525,
+      "loss": 3.4436,
+      "step": 3080
+    },
+    {
+      "epoch": 0.18953001968503938,
+      "grad_norm": 0.9076508283615112,
+      "learning_rate": 0.00018287875784665306,
+      "loss": 3.261,
+      "step": 3081
+    },
+    {
+      "epoch": 0.18959153543307086,
+      "grad_norm": 1.0816810131072998,
+      "learning_rate": 0.0001828679356635502,
+      "loss": 2.7674,
+      "step": 3082
+    },
+    {
+      "epoch": 0.18965305118110237,
+      "grad_norm": 0.9364808201789856,
+      "learning_rate": 0.00018285711038164862,
+      "loss": 3.1344,
+      "step": 3083
+    },
+    {
+      "epoch": 0.18971456692913385,
+      "grad_norm": 0.9905068278312683,
+      "learning_rate": 0.00018284628200135314,
+      "loss": 3.0023,
+      "step": 3084
+    },
+    {
+      "epoch": 0.18977608267716536,
+      "grad_norm": 1.1968648433685303,
+      "learning_rate": 0.00018283545052306867,
+      "loss": 2.1028,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18983759842519685,
+      "grad_norm": 1.0816971063613892,
+      "learning_rate": 0.00018282461594720028,
+      "loss": 2.7386,
+      "step": 3086
+    },
+    {
+      "epoch": 0.18989911417322836,
+      "grad_norm": 1.1540392637252808,
+      "learning_rate": 0.00018281377827415306,
+      "loss": 3.0553,
+      "step": 3087
+    },
+    {
+      "epoch": 0.18996062992125984,
+      "grad_norm": 1.1510194540023804,
+      "learning_rate": 0.00018280293750433236,
+      "loss": 3.3231,
+      "step": 3088
+    },
+    {
+      "epoch": 0.19002214566929135,
+      "grad_norm": 1.140428900718689,
+      "learning_rate": 0.0001827920936381435,
+      "loss": 3.0831,
+      "step": 3089
+    },
+    {
+      "epoch": 0.19008366141732283,
+      "grad_norm": 1.1693488359451294,
+      "learning_rate": 0.000182781246675992,
+      "loss": 2.5561,
+      "step": 3090
+    },
+    {
+      "epoch": 0.19014517716535434,
+      "grad_norm": 1.0507588386535645,
+      "learning_rate": 0.00018277039661828342,
+      "loss": 2.9753,
+      "step": 3091
+    },
+    {
+      "epoch": 0.19020669291338582,
+      "grad_norm": 1.1123806238174438,
+      "learning_rate": 0.00018275954346542363,
+      "loss": 3.0475,
+      "step": 3092
+    },
+    {
+      "epoch": 0.19026820866141733,
+      "grad_norm": 0.993868350982666,
+      "learning_rate": 0.00018274868721781834,
+      "loss": 2.875,
+      "step": 3093
+    },
+    {
+      "epoch": 0.1903297244094488,
+      "grad_norm": 1.3841694593429565,
+      "learning_rate": 0.0001827378278758736,
+      "loss": 2.494,
+      "step": 3094
+    },
+    {
+      "epoch": 0.19039124015748032,
+      "grad_norm": 1.1994870901107788,
+      "learning_rate": 0.00018272696543999544,
+      "loss": 3.2951,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1904527559055118,
+      "grad_norm": 1.3634471893310547,
+      "learning_rate": 0.00018271609991059006,
+      "loss": 2.2066,
+      "step": 3096
+    },
+    {
+      "epoch": 0.19051427165354332,
+      "grad_norm": 1.4088542461395264,
+      "learning_rate": 0.0001827052312880638,
+      "loss": 3.6766,
+      "step": 3097
+    },
+    {
+      "epoch": 0.1905757874015748,
+      "grad_norm": 1.4708058834075928,
+      "learning_rate": 0.00018269435957282308,
+      "loss": 2.7028,
+      "step": 3098
+    },
+    {
+      "epoch": 0.1906373031496063,
+      "grad_norm": 1.077402114868164,
+      "learning_rate": 0.0001826834847652744,
+      "loss": 2.7498,
+      "step": 3099
+    },
+    {
+      "epoch": 0.1906988188976378,
+      "grad_norm": 0.9887830018997192,
+      "learning_rate": 0.00018267260686582445,
+      "loss": 3.0061,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1907603346456693,
+      "grad_norm": 0.9788404107093811,
+      "learning_rate": 0.00018266172587487998,
+      "loss": 2.7412,
+      "step": 3101
+    },
+    {
+      "epoch": 0.19082185039370078,
+      "grad_norm": 1.204798698425293,
+      "learning_rate": 0.00018265084179284796,
+      "loss": 2.83,
+      "step": 3102
+    },
+    {
+      "epoch": 0.1908833661417323,
+      "grad_norm": 0.9261961579322815,
+      "learning_rate": 0.00018263995462013528,
+      "loss": 3.0365,
+      "step": 3103
+    },
+    {
+      "epoch": 0.19094488188976377,
+      "grad_norm": 1.065915584564209,
+      "learning_rate": 0.00018262906435714913,
+      "loss": 2.5456,
+      "step": 3104
+    },
+    {
+      "epoch": 0.19100639763779528,
+      "grad_norm": 1.1161985397338867,
+      "learning_rate": 0.00018261817100429672,
+      "loss": 3.059,
+      "step": 3105
+    },
+    {
+      "epoch": 0.19106791338582677,
+      "grad_norm": 1.074973702430725,
+      "learning_rate": 0.0001826072745619854,
+      "loss": 3.1236,
+      "step": 3106
+    },
+    {
+      "epoch": 0.19112942913385828,
+      "grad_norm": 0.8935304284095764,
+      "learning_rate": 0.00018259637503062264,
+      "loss": 3.207,
+      "step": 3107
+    },
+    {
+      "epoch": 0.19119094488188976,
+      "grad_norm": 1.0382492542266846,
+      "learning_rate": 0.00018258547241061603,
+      "loss": 2.8725,
+      "step": 3108
+    },
+    {
+      "epoch": 0.19125246062992127,
+      "grad_norm": 1.1471459865570068,
+      "learning_rate": 0.00018257456670237328,
+      "loss": 2.9443,
+      "step": 3109
+    },
+    {
+      "epoch": 0.19131397637795275,
+      "grad_norm": 1.1652765274047852,
+      "learning_rate": 0.00018256365790630214,
+      "loss": 3.4587,
+      "step": 3110
+    },
+    {
+      "epoch": 0.19137549212598426,
+      "grad_norm": 1.0849864482879639,
+      "learning_rate": 0.0001825527460228106,
+      "loss": 2.965,
+      "step": 3111
+    },
+    {
+      "epoch": 0.19143700787401574,
+      "grad_norm": 1.2349507808685303,
+      "learning_rate": 0.00018254183105230668,
+      "loss": 2.8753,
+      "step": 3112
+    },
+    {
+      "epoch": 0.19149852362204725,
+      "grad_norm": 1.1886721849441528,
+      "learning_rate": 0.00018253091299519857,
+      "loss": 2.7095,
+      "step": 3113
+    },
+    {
+      "epoch": 0.19156003937007873,
+      "grad_norm": 1.093546986579895,
+      "learning_rate": 0.0001825199918518945,
+      "loss": 2.6218,
+      "step": 3114
+    },
+    {
+      "epoch": 0.19162155511811024,
+      "grad_norm": 1.4060821533203125,
+      "learning_rate": 0.00018250906762280292,
+      "loss": 3.1029,
+      "step": 3115
+    },
+    {
+      "epoch": 0.19168307086614172,
+      "grad_norm": 0.9792125225067139,
+      "learning_rate": 0.00018249814030833224,
+      "loss": 2.0299,
+      "step": 3116
+    },
+    {
+      "epoch": 0.19174458661417323,
+      "grad_norm": 1.4139330387115479,
+      "learning_rate": 0.00018248720990889113,
+      "loss": 3.281,
+      "step": 3117
+    },
+    {
+      "epoch": 0.19180610236220472,
+      "grad_norm": 1.2385259866714478,
+      "learning_rate": 0.00018247627642488837,
+      "loss": 2.48,
+      "step": 3118
+    },
+    {
+      "epoch": 0.19186761811023623,
+      "grad_norm": 1.1222373247146606,
+      "learning_rate": 0.00018246533985673274,
+      "loss": 3.1437,
+      "step": 3119
+    },
+    {
+      "epoch": 0.1919291338582677,
+      "grad_norm": 1.0526341199874878,
+      "learning_rate": 0.00018245440020483327,
+      "loss": 2.8074,
+      "step": 3120
+    },
+    {
+      "epoch": 0.19199064960629922,
+      "grad_norm": 1.4522666931152344,
+      "learning_rate": 0.000182443457469599,
+      "loss": 2.12,
+      "step": 3121
+    },
+    {
+      "epoch": 0.1920521653543307,
+      "grad_norm": 1.36923348903656,
+      "learning_rate": 0.0001824325116514391,
+      "loss": 2.6789,
+      "step": 3122
+    },
+    {
+      "epoch": 0.1921136811023622,
+      "grad_norm": 1.0022773742675781,
+      "learning_rate": 0.00018242156275076298,
+      "loss": 1.9909,
+      "step": 3123
+    },
+    {
+      "epoch": 0.1921751968503937,
+      "grad_norm": 1.2759724855422974,
+      "learning_rate": 0.00018241061076797998,
+      "loss": 2.6541,
+      "step": 3124
+    },
+    {
+      "epoch": 0.1922367125984252,
+      "grad_norm": 1.2863909006118774,
+      "learning_rate": 0.00018239965570349966,
+      "loss": 2.8588,
+      "step": 3125
+    },
+    {
+      "epoch": 0.19229822834645668,
+      "grad_norm": 1.0554441213607788,
+      "learning_rate": 0.00018238869755773173,
+      "loss": 2.0962,
+      "step": 3126
+    },
+    {
+      "epoch": 0.1923597440944882,
+      "grad_norm": 1.3428821563720703,
+      "learning_rate": 0.00018237773633108587,
+      "loss": 2.8345,
+      "step": 3127
+    },
+    {
+      "epoch": 0.19242125984251968,
+      "grad_norm": 1.1681900024414062,
+      "learning_rate": 0.00018236677202397206,
+      "loss": 2.9616,
+      "step": 3128
+    },
+    {
+      "epoch": 0.1924827755905512,
+      "grad_norm": 1.1992582082748413,
+      "learning_rate": 0.00018235580463680026,
+      "loss": 3.2463,
+      "step": 3129
+    },
+    {
+      "epoch": 0.19254429133858267,
+      "grad_norm": 1.1473764181137085,
+      "learning_rate": 0.0001823448341699806,
+      "loss": 2.6987,
+      "step": 3130
+    },
+    {
+      "epoch": 0.19260580708661418,
+      "grad_norm": 1.196133017539978,
+      "learning_rate": 0.00018233386062392327,
+      "loss": 2.8515,
+      "step": 3131
+    },
+    {
+      "epoch": 0.19266732283464566,
+      "grad_norm": 1.069113850593567,
+      "learning_rate": 0.0001823228839990387,
+      "loss": 2.5714,
+      "step": 3132
+    },
+    {
+      "epoch": 0.19272883858267717,
+      "grad_norm": 1.1088560819625854,
+      "learning_rate": 0.0001823119042957373,
+      "loss": 2.7761,
+      "step": 3133
+    },
+    {
+      "epoch": 0.19279035433070865,
+      "grad_norm": 1.1109532117843628,
+      "learning_rate": 0.00018230092151442966,
+      "loss": 2.9441,
+      "step": 3134
+    },
+    {
+      "epoch": 0.19285187007874016,
+      "grad_norm": 1.254149317741394,
+      "learning_rate": 0.0001822899356555265,
+      "loss": 2.1634,
+      "step": 3135
+    },
+    {
+      "epoch": 0.19291338582677164,
+      "grad_norm": 1.3581055402755737,
+      "learning_rate": 0.00018227894671943857,
+      "loss": 2.602,
+      "step": 3136
+    },
+    {
+      "epoch": 0.19297490157480315,
+      "grad_norm": 1.3165230751037598,
+      "learning_rate": 0.00018226795470657688,
+      "loss": 2.495,
+      "step": 3137
+    },
+    {
+      "epoch": 0.19303641732283464,
+      "grad_norm": 0.9455828070640564,
+      "learning_rate": 0.00018225695961735238,
+      "loss": 3.4071,
+      "step": 3138
+    },
+    {
+      "epoch": 0.19309793307086615,
+      "grad_norm": 1.1337190866470337,
+      "learning_rate": 0.00018224596145217627,
+      "loss": 2.9486,
+      "step": 3139
+    },
+    {
+      "epoch": 0.19315944881889763,
+      "grad_norm": 0.8634281158447266,
+      "learning_rate": 0.00018223496021145983,
+      "loss": 2.5807,
+      "step": 3140
+    },
+    {
+      "epoch": 0.19322096456692914,
+      "grad_norm": 1.1702800989151,
+      "learning_rate": 0.00018222395589561442,
+      "loss": 2.6867,
+      "step": 3141
+    },
+    {
+      "epoch": 0.19328248031496062,
+      "grad_norm": 1.0138970613479614,
+      "learning_rate": 0.00018221294850505158,
+      "loss": 2.9248,
+      "step": 3142
+    },
+    {
+      "epoch": 0.19334399606299213,
+      "grad_norm": 0.8498350381851196,
+      "learning_rate": 0.00018220193804018288,
+      "loss": 3.1573,
+      "step": 3143
+    },
+    {
+      "epoch": 0.1934055118110236,
+      "grad_norm": 1.7482917308807373,
+      "learning_rate": 0.00018219092450142007,
+      "loss": 2.8368,
+      "step": 3144
+    },
+    {
+      "epoch": 0.19346702755905512,
+      "grad_norm": 1.2370675802230835,
+      "learning_rate": 0.000182179907889175,
+      "loss": 2.7589,
+      "step": 3145
+    },
+    {
+      "epoch": 0.1935285433070866,
+      "grad_norm": 1.2472766637802124,
+      "learning_rate": 0.00018216888820385963,
+      "loss": 2.2208,
+      "step": 3146
+    },
+    {
+      "epoch": 0.1935900590551181,
+      "grad_norm": 1.0099173784255981,
+      "learning_rate": 0.00018215786544588603,
+      "loss": 1.9741,
+      "step": 3147
+    },
+    {
+      "epoch": 0.1936515748031496,
+      "grad_norm": 1.129630208015442,
+      "learning_rate": 0.00018214683961566635,
+      "loss": 3.0138,
+      "step": 3148
+    },
+    {
+      "epoch": 0.1937130905511811,
+      "grad_norm": 1.1409252882003784,
+      "learning_rate": 0.00018213581071361293,
+      "loss": 2.7303,
+      "step": 3149
+    },
+    {
+      "epoch": 0.1937746062992126,
+      "grad_norm": 1.6557958126068115,
+      "learning_rate": 0.00018212477874013824,
+      "loss": 2.0966,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1938361220472441,
+      "grad_norm": 1.273929476737976,
+      "learning_rate": 0.00018211374369565472,
+      "loss": 2.8796,
+      "step": 3151
+    },
+    {
+      "epoch": 0.19389763779527558,
+      "grad_norm": 1.6584779024124146,
+      "learning_rate": 0.00018210270558057513,
+      "loss": 3.7045,
+      "step": 3152
+    },
+    {
+      "epoch": 0.1939591535433071,
+      "grad_norm": 1.1770477294921875,
+      "learning_rate": 0.00018209166439531212,
+      "loss": 3.3272,
+      "step": 3153
+    },
+    {
+      "epoch": 0.19402066929133857,
+      "grad_norm": 1.3618650436401367,
+      "learning_rate": 0.0001820806201402786,
+      "loss": 2.4141,
+      "step": 3154
+    },
+    {
+      "epoch": 0.19408218503937008,
+      "grad_norm": 1.268057107925415,
+      "learning_rate": 0.0001820695728158876,
+      "loss": 2.4074,
+      "step": 3155
+    },
+    {
+      "epoch": 0.19414370078740156,
+      "grad_norm": 0.9918739795684814,
+      "learning_rate": 0.0001820585224225522,
+      "loss": 3.4577,
+      "step": 3156
+    },
+    {
+      "epoch": 0.19420521653543307,
+      "grad_norm": 1.0962458848953247,
+      "learning_rate": 0.00018204746896068565,
+      "loss": 2.4701,
+      "step": 3157
+    },
+    {
+      "epoch": 0.19426673228346455,
+      "grad_norm": 1.1156789064407349,
+      "learning_rate": 0.00018203641243070128,
+      "loss": 2.8075,
+      "step": 3158
+    },
+    {
+      "epoch": 0.19432824803149606,
+      "grad_norm": 1.3286867141723633,
+      "learning_rate": 0.0001820253528330125,
+      "loss": 2.8658,
+      "step": 3159
+    },
+    {
+      "epoch": 0.19438976377952755,
+      "grad_norm": 1.0462429523468018,
+      "learning_rate": 0.00018201429016803292,
+      "loss": 3.0497,
+      "step": 3160
+    },
+    {
+      "epoch": 0.19445127952755906,
+      "grad_norm": 1.0499671697616577,
+      "learning_rate": 0.00018200322443617624,
+      "loss": 3.3019,
+      "step": 3161
+    },
+    {
+      "epoch": 0.19451279527559054,
+      "grad_norm": 0.9896591305732727,
+      "learning_rate": 0.0001819921556378562,
+      "loss": 3.4629,
+      "step": 3162
+    },
+    {
+      "epoch": 0.19457431102362205,
+      "grad_norm": 1.0455214977264404,
+      "learning_rate": 0.00018198108377348671,
+      "loss": 3.5655,
+      "step": 3163
+    },
+    {
+      "epoch": 0.19463582677165353,
+      "grad_norm": 0.9840503931045532,
+      "learning_rate": 0.0001819700088434819,
+      "loss": 1.9882,
+      "step": 3164
+    },
+    {
+      "epoch": 0.19469734251968504,
+      "grad_norm": 1.0672205686569214,
+      "learning_rate": 0.00018195893084825578,
+      "loss": 3.0716,
+      "step": 3165
+    },
+    {
+      "epoch": 0.19475885826771652,
+      "grad_norm": 1.1981617212295532,
+      "learning_rate": 0.00018194784978822268,
+      "loss": 3.3598,
+      "step": 3166
+    },
+    {
+      "epoch": 0.19482037401574803,
+      "grad_norm": 0.9700261950492859,
+      "learning_rate": 0.00018193676566379696,
+      "loss": 1.9622,
+      "step": 3167
+    },
+    {
+      "epoch": 0.19488188976377951,
+      "grad_norm": 1.1809697151184082,
+      "learning_rate": 0.0001819256784753931,
+      "loss": 2.6971,
+      "step": 3168
+    },
+    {
+      "epoch": 0.19494340551181102,
+      "grad_norm": 1.1023879051208496,
+      "learning_rate": 0.00018191458822342568,
+      "loss": 3.271,
+      "step": 3169
+    },
+    {
+      "epoch": 0.1950049212598425,
+      "grad_norm": 0.9346036911010742,
+      "learning_rate": 0.00018190349490830943,
+      "loss": 3.6673,
+      "step": 3170
+    },
+    {
+      "epoch": 0.19506643700787402,
+      "grad_norm": 1.2652223110198975,
+      "learning_rate": 0.00018189239853045916,
+      "loss": 2.1028,
+      "step": 3171
+    },
+    {
+      "epoch": 0.1951279527559055,
+      "grad_norm": 1.0646331310272217,
+      "learning_rate": 0.00018188129909028986,
+      "loss": 2.8697,
+      "step": 3172
+    },
+    {
+      "epoch": 0.195189468503937,
+      "grad_norm": 1.2980716228485107,
+      "learning_rate": 0.00018187019658821654,
+      "loss": 2.5874,
+      "step": 3173
+    },
+    {
+      "epoch": 0.1952509842519685,
+      "grad_norm": 1.0645416975021362,
+      "learning_rate": 0.0001818590910246544,
+      "loss": 3.0364,
+      "step": 3174
+    },
+    {
+      "epoch": 0.1953125,
+      "grad_norm": 1.1339023113250732,
+      "learning_rate": 0.0001818479824000187,
+      "loss": 2.5531,
+      "step": 3175
+    },
+    {
+      "epoch": 0.1953740157480315,
+      "grad_norm": 1.0823580026626587,
+      "learning_rate": 0.00018183687071472487,
+      "loss": 2.5827,
+      "step": 3176
+    },
+    {
+      "epoch": 0.195435531496063,
+      "grad_norm": 1.0640984773635864,
+      "learning_rate": 0.00018182575596918841,
+      "loss": 3.1239,
+      "step": 3177
+    },
+    {
+      "epoch": 0.1954970472440945,
+      "grad_norm": 1.0301458835601807,
+      "learning_rate": 0.00018181463816382493,
+      "loss": 3.4394,
+      "step": 3178
+    },
+    {
+      "epoch": 0.19555856299212598,
+      "grad_norm": 1.0261757373809814,
+      "learning_rate": 0.00018180351729905021,
+      "loss": 2.7984,
+      "step": 3179
+    },
+    {
+      "epoch": 0.1956200787401575,
+      "grad_norm": 1.2251148223876953,
+      "learning_rate": 0.0001817923933752801,
+      "loss": 2.7822,
+      "step": 3180
+    },
+    {
+      "epoch": 0.19568159448818898,
+      "grad_norm": 1.1537253856658936,
+      "learning_rate": 0.00018178126639293053,
+      "loss": 2.9258,
+      "step": 3181
+    },
+    {
+      "epoch": 0.19574311023622049,
+      "grad_norm": 1.195117473602295,
+      "learning_rate": 0.00018177013635241765,
+      "loss": 2.2224,
+      "step": 3182
+    },
+    {
+      "epoch": 0.19580462598425197,
+      "grad_norm": 1.1156411170959473,
+      "learning_rate": 0.00018175900325415764,
+      "loss": 2.9005,
+      "step": 3183
+    },
+    {
+      "epoch": 0.19586614173228348,
+      "grad_norm": 2.1394522190093994,
+      "learning_rate": 0.00018174786709856678,
+      "loss": 2.7426,
+      "step": 3184
+    },
+    {
+      "epoch": 0.19592765748031496,
+      "grad_norm": 1.71308171749115,
+      "learning_rate": 0.00018173672788606156,
+      "loss": 2.4126,
+      "step": 3185
+    },
+    {
+      "epoch": 0.19598917322834647,
+      "grad_norm": 1.1050273180007935,
+      "learning_rate": 0.0001817255856170585,
+      "loss": 3.2011,
+      "step": 3186
+    },
+    {
+      "epoch": 0.19605068897637795,
+      "grad_norm": 1.0014677047729492,
+      "learning_rate": 0.0001817144402919742,
+      "loss": 2.9657,
+      "step": 3187
+    },
+    {
+      "epoch": 0.19611220472440946,
+      "grad_norm": 1.0543326139450073,
+      "learning_rate": 0.00018170329191122554,
+      "loss": 3.336,
+      "step": 3188
+    },
+    {
+      "epoch": 0.19617372047244094,
+      "grad_norm": 0.972935676574707,
+      "learning_rate": 0.00018169214047522933,
+      "loss": 3.0759,
+      "step": 3189
+    },
+    {
+      "epoch": 0.19623523622047245,
+      "grad_norm": 1.3471523523330688,
+      "learning_rate": 0.00018168098598440259,
+      "loss": 3.1529,
+      "step": 3190
+    },
+    {
+      "epoch": 0.19629675196850394,
+      "grad_norm": 1.1401716470718384,
+      "learning_rate": 0.00018166982843916243,
+      "loss": 3.2176,
+      "step": 3191
+    },
+    {
+      "epoch": 0.19635826771653545,
+      "grad_norm": 1.1750662326812744,
+      "learning_rate": 0.00018165866783992612,
+      "loss": 2.5941,
+      "step": 3192
+    },
+    {
+      "epoch": 0.19641978346456693,
+      "grad_norm": 1.0571677684783936,
+      "learning_rate": 0.00018164750418711093,
+      "loss": 3.0154,
+      "step": 3193
+    },
+    {
+      "epoch": 0.19648129921259844,
+      "grad_norm": 1.02812922000885,
+      "learning_rate": 0.00018163633748113442,
+      "loss": 3.1872,
+      "step": 3194
+    },
+    {
+      "epoch": 0.19654281496062992,
+      "grad_norm": 1.1107584238052368,
+      "learning_rate": 0.00018162516772241407,
+      "loss": 2.9905,
+      "step": 3195
+    },
+    {
+      "epoch": 0.19660433070866143,
+      "grad_norm": 1.2876391410827637,
+      "learning_rate": 0.0001816139949113676,
+      "loss": 3.0473,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1966658464566929,
+      "grad_norm": 0.9676750898361206,
+      "learning_rate": 0.00018160281904841283,
+      "loss": 3.0369,
+      "step": 3197
+    },
+    {
+      "epoch": 0.19672736220472442,
+      "grad_norm": 0.9619333744049072,
+      "learning_rate": 0.00018159164013396763,
+      "loss": 3.2879,
+      "step": 3198
+    },
+    {
+      "epoch": 0.1967888779527559,
+      "grad_norm": 1.0813863277435303,
+      "learning_rate": 0.00018158045816845007,
+      "loss": 3.5412,
+      "step": 3199
+    },
+    {
+      "epoch": 0.1968503937007874,
+      "grad_norm": 1.4956492185592651,
+      "learning_rate": 0.0001815692731522783,
+      "loss": 2.7816,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1969119094488189,
+      "grad_norm": 1.1554566621780396,
+      "learning_rate": 0.00018155808508587054,
+      "loss": 2.4857,
+      "step": 3201
+    },
+    {
+      "epoch": 0.1969734251968504,
+      "grad_norm": 1.0991082191467285,
+      "learning_rate": 0.00018154689396964518,
+      "loss": 2.5876,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1970349409448819,
+      "grad_norm": 1.2308779954910278,
+      "learning_rate": 0.0001815356998040207,
+      "loss": 2.5769,
+      "step": 3203
+    },
+    {
+      "epoch": 0.1970964566929134,
+      "grad_norm": 1.1450248956680298,
+      "learning_rate": 0.0001815245025894157,
+      "loss": 3.2105,
+      "step": 3204
+    },
+    {
+      "epoch": 0.19715797244094488,
+      "grad_norm": 1.3415850400924683,
+      "learning_rate": 0.00018151330232624887,
+      "loss": 2.6375,
+      "step": 3205
+    },
+    {
+      "epoch": 0.1972194881889764,
+      "grad_norm": 1.555673360824585,
+      "learning_rate": 0.00018150209901493912,
+      "loss": 2.1084,
+      "step": 3206
+    },
+    {
+      "epoch": 0.19728100393700787,
+      "grad_norm": 1.1794610023498535,
+      "learning_rate": 0.0001814908926559053,
+      "loss": 2.6396,
+      "step": 3207
+    },
+    {
+      "epoch": 0.19734251968503938,
+      "grad_norm": 1.2137082815170288,
+      "learning_rate": 0.00018147968324956653,
+      "loss": 3.1953,
+      "step": 3208
+    },
+    {
+      "epoch": 0.19740403543307086,
+      "grad_norm": 1.0985503196716309,
+      "learning_rate": 0.00018146847079634194,
+      "loss": 3.0949,
+      "step": 3209
+    },
+    {
+      "epoch": 0.19746555118110237,
+      "grad_norm": 0.9899457693099976,
+      "learning_rate": 0.00018145725529665078,
+      "loss": 2.9369,
+      "step": 3210
+    },
+    {
+      "epoch": 0.19752706692913385,
+      "grad_norm": 1.1252667903900146,
+      "learning_rate": 0.00018144603675091252,
+      "loss": 3.1612,
+      "step": 3211
+    },
+    {
+      "epoch": 0.19758858267716536,
+      "grad_norm": 1.1265114545822144,
+      "learning_rate": 0.00018143481515954665,
+      "loss": 2.5912,
+      "step": 3212
+    },
+    {
+      "epoch": 0.19765009842519685,
+      "grad_norm": 1.0304052829742432,
+      "learning_rate": 0.00018142359052297278,
+      "loss": 2.5821,
+      "step": 3213
+    },
+    {
+      "epoch": 0.19771161417322836,
+      "grad_norm": 1.1099830865859985,
+      "learning_rate": 0.00018141236284161064,
+      "loss": 2.7662,
+      "step": 3214
+    },
+    {
+      "epoch": 0.19777312992125984,
+      "grad_norm": 1.0551482439041138,
+      "learning_rate": 0.00018140113211588012,
+      "loss": 2.8477,
+      "step": 3215
+    },
+    {
+      "epoch": 0.19783464566929135,
+      "grad_norm": 1.0917601585388184,
+      "learning_rate": 0.00018138989834620115,
+      "loss": 2.9089,
+      "step": 3216
+    },
+    {
+      "epoch": 0.19789616141732283,
+      "grad_norm": 0.9330020546913147,
+      "learning_rate": 0.00018137866153299385,
+      "loss": 2.5652,
+      "step": 3217
+    },
+    {
+      "epoch": 0.19795767716535434,
+      "grad_norm": 1.214663028717041,
+      "learning_rate": 0.00018136742167667834,
+      "loss": 3.372,
+      "step": 3218
+    },
+    {
+      "epoch": 0.19801919291338582,
+      "grad_norm": 1.0847883224487305,
+      "learning_rate": 0.000181356178777675,
+      "loss": 2.9408,
+      "step": 3219
+    },
+    {
+      "epoch": 0.19808070866141733,
+      "grad_norm": 1.0513365268707275,
+      "learning_rate": 0.00018134493283640426,
+      "loss": 2.3355,
+      "step": 3220
+    },
+    {
+      "epoch": 0.1981422244094488,
+      "grad_norm": 1.118646502494812,
+      "learning_rate": 0.0001813336838532866,
+      "loss": 2.4445,
+      "step": 3221
+    },
+    {
+      "epoch": 0.19820374015748032,
+      "grad_norm": 1.113264560699463,
+      "learning_rate": 0.00018132243182874267,
+      "loss": 2.9423,
+      "step": 3222
+    },
+    {
+      "epoch": 0.1982652559055118,
+      "grad_norm": 1.466409683227539,
+      "learning_rate": 0.0001813111767631933,
+      "loss": 2.4967,
+      "step": 3223
+    },
+    {
+      "epoch": 0.19832677165354332,
+      "grad_norm": 1.221403956413269,
+      "learning_rate": 0.00018129991865705928,
+      "loss": 2.891,
+      "step": 3224
+    },
+    {
+      "epoch": 0.1983882874015748,
+      "grad_norm": 1.112019658088684,
+      "learning_rate": 0.00018128865751076165,
+      "loss": 2.6986,
+      "step": 3225
+    },
+    {
+      "epoch": 0.1984498031496063,
+      "grad_norm": 0.9426658749580383,
+      "learning_rate": 0.0001812773933247215,
+      "loss": 2.3357,
+      "step": 3226
+    },
+    {
+      "epoch": 0.1985113188976378,
+      "grad_norm": 1.2692878246307373,
+      "learning_rate": 0.0001812661260993601,
+      "loss": 2.7986,
+      "step": 3227
+    },
+    {
+      "epoch": 0.1985728346456693,
+      "grad_norm": 1.1908998489379883,
+      "learning_rate": 0.0001812548558350987,
+      "loss": 2.9896,
+      "step": 3228
+    },
+    {
+      "epoch": 0.19863435039370078,
+      "grad_norm": 1.0685088634490967,
+      "learning_rate": 0.0001812435825323588,
+      "loss": 3.6005,
+      "step": 3229
+    },
+    {
+      "epoch": 0.1986958661417323,
+      "grad_norm": 1.4616525173187256,
+      "learning_rate": 0.00018123230619156193,
+      "loss": 2.5647,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19875738188976377,
+      "grad_norm": 1.181075930595398,
+      "learning_rate": 0.00018122102681312975,
+      "loss": 3.2808,
+      "step": 3231
+    },
+    {
+      "epoch": 0.19881889763779528,
+      "grad_norm": 1.2040804624557495,
+      "learning_rate": 0.0001812097443974841,
+      "loss": 3.1057,
+      "step": 3232
+    },
+    {
+      "epoch": 0.19888041338582677,
+      "grad_norm": 1.18716299533844,
+      "learning_rate": 0.00018119845894504682,
+      "loss": 2.8845,
+      "step": 3233
+    },
+    {
+      "epoch": 0.19894192913385828,
+      "grad_norm": 1.0061001777648926,
+      "learning_rate": 0.00018118717045624,
+      "loss": 3.267,
+      "step": 3234
+    },
+    {
+      "epoch": 0.19900344488188976,
+      "grad_norm": 1.0472499132156372,
+      "learning_rate": 0.0001811758789314857,
+      "loss": 3.0861,
+      "step": 3235
+    },
+    {
+      "epoch": 0.19906496062992127,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 0.00018116458437120617,
+      "loss": 2.7465,
+      "step": 3236
+    },
+    {
+      "epoch": 0.19912647637795275,
+      "grad_norm": 0.9453393220901489,
+      "learning_rate": 0.00018115328677582374,
+      "loss": 3.0583,
+      "step": 3237
+    },
+    {
+      "epoch": 0.19918799212598426,
+      "grad_norm": 1.1621795892715454,
+      "learning_rate": 0.00018114198614576092,
+      "loss": 2.5692,
+      "step": 3238
+    },
+    {
+      "epoch": 0.19924950787401574,
+      "grad_norm": 0.9127908945083618,
+      "learning_rate": 0.0001811306824814403,
+      "loss": 3.286,
+      "step": 3239
+    },
+    {
+      "epoch": 0.19931102362204725,
+      "grad_norm": 1.1290059089660645,
+      "learning_rate": 0.00018111937578328452,
+      "loss": 3.1061,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19937253937007873,
+      "grad_norm": 0.9475670456886292,
+      "learning_rate": 0.00018110806605171645,
+      "loss": 2.5766,
+      "step": 3241
+    },
+    {
+      "epoch": 0.19943405511811024,
+      "grad_norm": 0.9196003079414368,
+      "learning_rate": 0.00018109675328715897,
+      "loss": 3.0021,
+      "step": 3242
+    },
+    {
+      "epoch": 0.19949557086614172,
+      "grad_norm": 1.1052747964859009,
+      "learning_rate": 0.00018108543749003513,
+      "loss": 3.1475,
+      "step": 3243
+    },
+    {
+      "epoch": 0.19955708661417323,
+      "grad_norm": 1.4407484531402588,
+      "learning_rate": 0.00018107411866076807,
+      "loss": 2.4872,
+      "step": 3244
+    },
+    {
+      "epoch": 0.19961860236220472,
+      "grad_norm": 1.0845506191253662,
+      "learning_rate": 0.00018106279679978106,
+      "loss": 2.8045,
+      "step": 3245
+    },
+    {
+      "epoch": 0.19968011811023623,
+      "grad_norm": 1.0183476209640503,
+      "learning_rate": 0.00018105147190749748,
+      "loss": 2.7473,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1997416338582677,
+      "grad_norm": 1.146312952041626,
+      "learning_rate": 0.00018104014398434077,
+      "loss": 3.5209,
+      "step": 3247
+    },
+    {
+      "epoch": 0.19980314960629922,
+      "grad_norm": 0.9749007821083069,
+      "learning_rate": 0.00018102881303073458,
+      "loss": 2.6773,
+      "step": 3248
+    },
+    {
+      "epoch": 0.1998646653543307,
+      "grad_norm": 1.0480376482009888,
+      "learning_rate": 0.0001810174790471026,
+      "loss": 3.0158,
+      "step": 3249
+    },
+    {
+      "epoch": 0.1999261811023622,
+      "grad_norm": 1.3534146547317505,
+      "learning_rate": 0.0001810061420338687,
+      "loss": 2.7491,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1999876968503937,
+      "grad_norm": 1.0750564336776733,
+      "learning_rate": 0.00018099480199145677,
+      "loss": 1.972,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2000492125984252,
+      "grad_norm": 0.9159018397331238,
+      "learning_rate": 0.00018098345892029091,
+      "loss": 3.1137,
+      "step": 3252
+    },
+    {
+      "epoch": 0.20011072834645668,
+      "grad_norm": 1.1984827518463135,
+      "learning_rate": 0.00018097211282079522,
+      "loss": 2.5464,
+      "step": 3253
+    },
+    {
+      "epoch": 0.2001722440944882,
+      "grad_norm": 1.3134299516677856,
+      "learning_rate": 0.00018096076369339406,
+      "loss": 3.0872,
+      "step": 3254
+    },
+    {
+      "epoch": 0.20023375984251968,
+      "grad_norm": 1.0354126691818237,
+      "learning_rate": 0.00018094941153851178,
+      "loss": 2.4053,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2002952755905512,
+      "grad_norm": 1.1897698640823364,
+      "learning_rate": 0.00018093805635657287,
+      "loss": 3.7567,
+      "step": 3256
+    },
+    {
+      "epoch": 0.20035679133858267,
+      "grad_norm": 1.2458997964859009,
+      "learning_rate": 0.000180926698148002,
+      "loss": 2.9601,
+      "step": 3257
+    },
+    {
+      "epoch": 0.20041830708661418,
+      "grad_norm": 1.127963662147522,
+      "learning_rate": 0.00018091533691322388,
+      "loss": 2.5985,
+      "step": 3258
+    },
+    {
+      "epoch": 0.20047982283464566,
+      "grad_norm": 1.0027151107788086,
+      "learning_rate": 0.00018090397265266335,
+      "loss": 3.1939,
+      "step": 3259
+    },
+    {
+      "epoch": 0.20054133858267717,
+      "grad_norm": 1.1506452560424805,
+      "learning_rate": 0.00018089260536674538,
+      "loss": 3.048,
+      "step": 3260
+    },
+    {
+      "epoch": 0.20060285433070865,
+      "grad_norm": 1.2634047269821167,
+      "learning_rate": 0.00018088123505589503,
+      "loss": 2.4124,
+      "step": 3261
+    },
+    {
+      "epoch": 0.20066437007874016,
+      "grad_norm": 1.1463994979858398,
+      "learning_rate": 0.00018086986172053752,
+      "loss": 2.8146,
+      "step": 3262
+    },
+    {
+      "epoch": 0.20072588582677164,
+      "grad_norm": 1.2002328634262085,
+      "learning_rate": 0.0001808584853610981,
+      "loss": 2.7957,
+      "step": 3263
+    },
+    {
+      "epoch": 0.20078740157480315,
+      "grad_norm": 1.116618275642395,
+      "learning_rate": 0.00018084710597800218,
+      "loss": 2.6031,
+      "step": 3264
+    },
+    {
+      "epoch": 0.20084891732283464,
+      "grad_norm": 0.9262930750846863,
+      "learning_rate": 0.00018083572357167535,
+      "loss": 3.0008,
+      "step": 3265
+    },
+    {
+      "epoch": 0.20091043307086615,
+      "grad_norm": 1.065913438796997,
+      "learning_rate": 0.0001808243381425432,
+      "loss": 2.9487,
+      "step": 3266
+    },
+    {
+      "epoch": 0.20097194881889763,
+      "grad_norm": 1.2757937908172607,
+      "learning_rate": 0.0001808129496910315,
+      "loss": 3.3911,
+      "step": 3267
+    },
+    {
+      "epoch": 0.20103346456692914,
+      "grad_norm": 1.1909842491149902,
+      "learning_rate": 0.0001808015582175661,
+      "loss": 3.1195,
+      "step": 3268
+    },
+    {
+      "epoch": 0.20109498031496062,
+      "grad_norm": 1.2262521982192993,
+      "learning_rate": 0.00018079016372257297,
+      "loss": 2.9362,
+      "step": 3269
+    },
+    {
+      "epoch": 0.20115649606299213,
+      "grad_norm": 0.9925058484077454,
+      "learning_rate": 0.00018077876620647824,
+      "loss": 2.7128,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2012180118110236,
+      "grad_norm": 1.0415962934494019,
+      "learning_rate": 0.00018076736566970808,
+      "loss": 2.5366,
+      "step": 3271
+    },
+    {
+      "epoch": 0.20127952755905512,
+      "grad_norm": 1.1746773719787598,
+      "learning_rate": 0.0001807559621126888,
+      "loss": 2.701,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2013410433070866,
+      "grad_norm": 1.065649151802063,
+      "learning_rate": 0.00018074455553584691,
+      "loss": 2.5899,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2014025590551181,
+      "grad_norm": 1.289330005645752,
+      "learning_rate": 0.00018073314593960882,
+      "loss": 2.1928,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2014640748031496,
+      "grad_norm": 1.0905553102493286,
+      "learning_rate": 0.00018072173332440126,
+      "loss": 2.6286,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2015255905511811,
+      "grad_norm": 1.453299641609192,
+      "learning_rate": 0.00018071031769065102,
+      "loss": 3.657,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2015871062992126,
+      "grad_norm": 1.1465785503387451,
+      "learning_rate": 0.00018069889903878493,
+      "loss": 2.4908,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2016486220472441,
+      "grad_norm": 0.9344906806945801,
+      "learning_rate": 0.00018068747736923,
+      "loss": 3.2417,
+      "step": 3278
+    },
+    {
+      "epoch": 0.20171013779527558,
+      "grad_norm": 1.1864597797393799,
+      "learning_rate": 0.00018067605268241334,
+      "loss": 2.7949,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2017716535433071,
+      "grad_norm": 1.4899767637252808,
+      "learning_rate": 0.0001806646249787622,
+      "loss": 2.3663,
+      "step": 3280
+    },
+    {
+      "epoch": 0.20183316929133857,
+      "grad_norm": 1.642462968826294,
+      "learning_rate": 0.00018065319425870384,
+      "loss": 2.1027,
+      "step": 3281
+    },
+    {
+      "epoch": 0.20189468503937008,
+      "grad_norm": 1.2261964082717896,
+      "learning_rate": 0.00018064176052266578,
+      "loss": 3.4302,
+      "step": 3282
+    },
+    {
+      "epoch": 0.20195620078740156,
+      "grad_norm": 1.3315712213516235,
+      "learning_rate": 0.00018063032377107553,
+      "loss": 2.6929,
+      "step": 3283
+    },
+    {
+      "epoch": 0.20201771653543307,
+      "grad_norm": 1.3472989797592163,
+      "learning_rate": 0.0001806188840043608,
+      "loss": 3.0442,
+      "step": 3284
+    },
+    {
+      "epoch": 0.20207923228346455,
+      "grad_norm": 1.207293152809143,
+      "learning_rate": 0.00018060744122294934,
+      "loss": 2.5405,
+      "step": 3285
+    },
+    {
+      "epoch": 0.20214074803149606,
+      "grad_norm": 1.344175934791565,
+      "learning_rate": 0.00018059599542726907,
+      "loss": 2.5853,
+      "step": 3286
+    },
+    {
+      "epoch": 0.20220226377952755,
+      "grad_norm": 1.410900354385376,
+      "learning_rate": 0.00018058454661774796,
+      "loss": 2.7846,
+      "step": 3287
+    },
+    {
+      "epoch": 0.20226377952755906,
+      "grad_norm": 1.0212764739990234,
+      "learning_rate": 0.0001805730947948142,
+      "loss": 2.8529,
+      "step": 3288
+    },
+    {
+      "epoch": 0.20232529527559054,
+      "grad_norm": 1.1240665912628174,
+      "learning_rate": 0.00018056163995889594,
+      "loss": 3.0015,
+      "step": 3289
+    },
+    {
+      "epoch": 0.20238681102362205,
+      "grad_norm": 1.1479837894439697,
+      "learning_rate": 0.00018055018211042158,
+      "loss": 3.2383,
+      "step": 3290
+    },
+    {
+      "epoch": 0.20244832677165353,
+      "grad_norm": 1.699271559715271,
+      "learning_rate": 0.00018053872124981955,
+      "loss": 1.8265,
+      "step": 3291
+    },
+    {
+      "epoch": 0.20250984251968504,
+      "grad_norm": 1.1884534358978271,
+      "learning_rate": 0.00018052725737751848,
+      "loss": 2.5979,
+      "step": 3292
+    },
+    {
+      "epoch": 0.20257135826771652,
+      "grad_norm": 1.1911803483963013,
+      "learning_rate": 0.000180515790493947,
+      "loss": 2.3272,
+      "step": 3293
+    },
+    {
+      "epoch": 0.20263287401574803,
+      "grad_norm": 1.1129341125488281,
+      "learning_rate": 0.00018050432059953396,
+      "loss": 2.9051,
+      "step": 3294
+    },
+    {
+      "epoch": 0.20269438976377951,
+      "grad_norm": 1.1152687072753906,
+      "learning_rate": 0.0001804928476947082,
+      "loss": 3.1709,
+      "step": 3295
+    },
+    {
+      "epoch": 0.20275590551181102,
+      "grad_norm": 1.6190673112869263,
+      "learning_rate": 0.00018048137177989877,
+      "loss": 2.4449,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2028174212598425,
+      "grad_norm": 0.988452136516571,
+      "learning_rate": 0.00018046989285553478,
+      "loss": 2.8364,
+      "step": 3297
+    },
+    {
+      "epoch": 0.20287893700787402,
+      "grad_norm": 1.0552514791488647,
+      "learning_rate": 0.00018045841092204556,
+      "loss": 2.8378,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2029404527559055,
+      "grad_norm": 1.1308726072311401,
+      "learning_rate": 0.0001804469259798604,
+      "loss": 2.992,
+      "step": 3299
+    },
+    {
+      "epoch": 0.203001968503937,
+      "grad_norm": 1.1873024702072144,
+      "learning_rate": 0.00018043543802940884,
+      "loss": 2.6703,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2030634842519685,
+      "grad_norm": 1.3032499551773071,
+      "learning_rate": 0.00018042394707112037,
+      "loss": 3.3674,
+      "step": 3301
+    },
+    {
+      "epoch": 0.203125,
+      "grad_norm": 1.5055042505264282,
+      "learning_rate": 0.00018041245310542473,
+      "loss": 2.3086,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2031865157480315,
+      "grad_norm": 0.9456162452697754,
+      "learning_rate": 0.00018040095613275177,
+      "loss": 2.8847,
+      "step": 3303
+    },
+    {
+      "epoch": 0.203248031496063,
+      "grad_norm": 1.0955500602722168,
+      "learning_rate": 0.00018038945615353133,
+      "loss": 3.1321,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2033095472440945,
+      "grad_norm": 1.6517293453216553,
+      "learning_rate": 0.00018037795316819355,
+      "loss": 2.1573,
+      "step": 3305
+    },
+    {
+      "epoch": 0.20337106299212598,
+      "grad_norm": 1.2305465936660767,
+      "learning_rate": 0.00018036644717716848,
+      "loss": 2.9788,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2034325787401575,
+      "grad_norm": 1.2862086296081543,
+      "learning_rate": 0.00018035493818088641,
+      "loss": 2.8725,
+      "step": 3307
+    },
+    {
+      "epoch": 0.20349409448818898,
+      "grad_norm": 1.9170730113983154,
+      "learning_rate": 0.00018034342617977777,
+      "loss": 2.6547,
+      "step": 3308
+    },
+    {
+      "epoch": 0.20355561023622049,
+      "grad_norm": 1.5010713338851929,
+      "learning_rate": 0.000180331911174273,
+      "loss": 2.9466,
+      "step": 3309
+    },
+    {
+      "epoch": 0.20361712598425197,
+      "grad_norm": 1.0765050649642944,
+      "learning_rate": 0.00018032039316480264,
+      "loss": 3.4386,
+      "step": 3310
+    },
+    {
+      "epoch": 0.20367864173228348,
+      "grad_norm": 1.1432244777679443,
+      "learning_rate": 0.00018030887215179751,
+      "loss": 2.9588,
+      "step": 3311
+    },
+    {
+      "epoch": 0.20374015748031496,
+      "grad_norm": 1.1695033311843872,
+      "learning_rate": 0.00018029734813568835,
+      "loss": 2.1606,
+      "step": 3312
+    },
+    {
+      "epoch": 0.20380167322834647,
+      "grad_norm": 1.130391001701355,
+      "learning_rate": 0.0001802858211169061,
+      "loss": 3.733,
+      "step": 3313
+    },
+    {
+      "epoch": 0.20386318897637795,
+      "grad_norm": 1.8390133380889893,
+      "learning_rate": 0.00018027429109588184,
+      "loss": 2.3565,
+      "step": 3314
+    },
+    {
+      "epoch": 0.20392470472440946,
+      "grad_norm": 1.0521575212478638,
+      "learning_rate": 0.00018026275807304676,
+      "loss": 2.7267,
+      "step": 3315
+    },
+    {
+      "epoch": 0.20398622047244094,
+      "grad_norm": 1.1940370798110962,
+      "learning_rate": 0.00018025122204883202,
+      "loss": 2.7801,
+      "step": 3316
+    },
+    {
+      "epoch": 0.20404773622047245,
+      "grad_norm": 1.099413275718689,
+      "learning_rate": 0.00018023968302366912,
+      "loss": 3.0459,
+      "step": 3317
+    },
+    {
+      "epoch": 0.20410925196850394,
+      "grad_norm": 1.2373571395874023,
+      "learning_rate": 0.00018022814099798948,
+      "loss": 2.7356,
+      "step": 3318
+    },
+    {
+      "epoch": 0.20417076771653545,
+      "grad_norm": 1.5459368228912354,
+      "learning_rate": 0.00018021659597222472,
+      "loss": 2.8296,
+      "step": 3319
+    },
+    {
+      "epoch": 0.20423228346456693,
+      "grad_norm": 1.026990294456482,
+      "learning_rate": 0.00018020504794680656,
+      "loss": 3.3203,
+      "step": 3320
+    },
+    {
+      "epoch": 0.20429379921259844,
+      "grad_norm": 1.056489109992981,
+      "learning_rate": 0.0001801934969221669,
+      "loss": 2.7754,
+      "step": 3321
+    },
+    {
+      "epoch": 0.20435531496062992,
+      "grad_norm": 1.1702975034713745,
+      "learning_rate": 0.0001801819428987376,
+      "loss": 2.1621,
+      "step": 3322
+    },
+    {
+      "epoch": 0.20441683070866143,
+      "grad_norm": 1.234944462776184,
+      "learning_rate": 0.0001801703858769507,
+      "loss": 2.5674,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2044783464566929,
+      "grad_norm": 1.5940791368484497,
+      "learning_rate": 0.00018015882585723848,
+      "loss": 2.5287,
+      "step": 3324
+    },
+    {
+      "epoch": 0.20453986220472442,
+      "grad_norm": 1.2259079217910767,
+      "learning_rate": 0.0001801472628400331,
+      "loss": 2.6286,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2046013779527559,
+      "grad_norm": 1.5038774013519287,
+      "learning_rate": 0.000180135696825767,
+      "loss": 3.0829,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2046628937007874,
+      "grad_norm": 1.2765671014785767,
+      "learning_rate": 0.00018012412781487274,
+      "loss": 3.28,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2047244094488189,
+      "grad_norm": 1.0880296230316162,
+      "learning_rate": 0.00018011255580778284,
+      "loss": 3.1299,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2047859251968504,
+      "grad_norm": 1.218890905380249,
+      "learning_rate": 0.00018010098080493007,
+      "loss": 2.3586,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2048474409448819,
+      "grad_norm": 1.1345899105072021,
+      "learning_rate": 0.0001800894028067473,
+      "loss": 2.905,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2049089566929134,
+      "grad_norm": 1.0350764989852905,
+      "learning_rate": 0.00018007782181366744,
+      "loss": 3.9845,
+      "step": 3331
+    },
+    {
+      "epoch": 0.20497047244094488,
+      "grad_norm": 1.0182693004608154,
+      "learning_rate": 0.00018006623782612356,
+      "loss": 2.5339,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2050319881889764,
+      "grad_norm": 1.3549962043762207,
+      "learning_rate": 0.00018005465084454888,
+      "loss": 2.4729,
+      "step": 3333
+    },
+    {
+      "epoch": 0.20509350393700787,
+      "grad_norm": 1.0325040817260742,
+      "learning_rate": 0.0001800430608693766,
+      "loss": 2.9965,
+      "step": 3334
+    },
+    {
+      "epoch": 0.20515501968503938,
+      "grad_norm": 1.2258576154708862,
+      "learning_rate": 0.00018003146790104024,
+      "loss": 3.2905,
+      "step": 3335
+    },
+    {
+      "epoch": 0.20521653543307086,
+      "grad_norm": 1.550382137298584,
+      "learning_rate": 0.0001800198719399732,
+      "loss": 2.3742,
+      "step": 3336
+    },
+    {
+      "epoch": 0.20527805118110237,
+      "grad_norm": 1.8174108266830444,
+      "learning_rate": 0.00018000827298660918,
+      "loss": 1.4092,
+      "step": 3337
+    },
+    {
+      "epoch": 0.20533956692913385,
+      "grad_norm": 1.0951390266418457,
+      "learning_rate": 0.0001799966710413819,
+      "loss": 2.6863,
+      "step": 3338
+    },
+    {
+      "epoch": 0.20540108267716536,
+      "grad_norm": 1.3479450941085815,
+      "learning_rate": 0.00017998506610472515,
+      "loss": 2.6611,
+      "step": 3339
+    },
+    {
+      "epoch": 0.20546259842519685,
+      "grad_norm": 1.0600186586380005,
+      "learning_rate": 0.00017997345817707296,
+      "loss": 2.4784,
+      "step": 3340
+    },
+    {
+      "epoch": 0.20552411417322836,
+      "grad_norm": 1.4579213857650757,
+      "learning_rate": 0.0001799618472588594,
+      "loss": 2.3377,
+      "step": 3341
+    },
+    {
+      "epoch": 0.20558562992125984,
+      "grad_norm": 1.4230811595916748,
+      "learning_rate": 0.0001799502333505186,
+      "loss": 2.5903,
+      "step": 3342
+    },
+    {
+      "epoch": 0.20564714566929135,
+      "grad_norm": 1.3540979623794556,
+      "learning_rate": 0.0001799386164524849,
+      "loss": 3.1642,
+      "step": 3343
+    },
+    {
+      "epoch": 0.20570866141732283,
+      "grad_norm": 1.136122226715088,
+      "learning_rate": 0.00017992699656519268,
+      "loss": 3.0934,
+      "step": 3344
+    },
+    {
+      "epoch": 0.20577017716535434,
+      "grad_norm": 0.937139630317688,
+      "learning_rate": 0.00017991537368907648,
+      "loss": 2.9353,
+      "step": 3345
+    },
+    {
+      "epoch": 0.20583169291338582,
+      "grad_norm": 1.2418447732925415,
+      "learning_rate": 0.00017990374782457096,
+      "loss": 2.5945,
+      "step": 3346
+    },
+    {
+      "epoch": 0.20589320866141733,
+      "grad_norm": 0.9432669281959534,
+      "learning_rate": 0.0001798921189721108,
+      "loss": 2.9354,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2059547244094488,
+      "grad_norm": 0.9790157675743103,
+      "learning_rate": 0.00017988048713213088,
+      "loss": 1.8784,
+      "step": 3348
+    },
+    {
+      "epoch": 0.20601624015748032,
+      "grad_norm": 0.99591463804245,
+      "learning_rate": 0.00017986885230506617,
+      "loss": 3.2512,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2060777559055118,
+      "grad_norm": 0.9492722749710083,
+      "learning_rate": 0.00017985721449135175,
+      "loss": 3.2783,
+      "step": 3350
+    },
+    {
+      "epoch": 0.20613927165354332,
+      "grad_norm": 1.0633811950683594,
+      "learning_rate": 0.0001798455736914228,
+      "loss": 2.9268,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2062007874015748,
+      "grad_norm": 1.2120282649993896,
+      "learning_rate": 0.00017983392990571462,
+      "loss": 3.2469,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2062623031496063,
+      "grad_norm": 1.115908145904541,
+      "learning_rate": 0.00017982228313466263,
+      "loss": 2.8525,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2063238188976378,
+      "grad_norm": 1.3602335453033447,
+      "learning_rate": 0.00017981063337870235,
+      "loss": 2.4639,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2063853346456693,
+      "grad_norm": 1.0842880010604858,
+      "learning_rate": 0.0001797989806382694,
+      "loss": 3.0721,
+      "step": 3355
+    },
+    {
+      "epoch": 0.20644685039370078,
+      "grad_norm": 1.2288116216659546,
+      "learning_rate": 0.00017978732491379958,
+      "loss": 2.3359,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2065083661417323,
+      "grad_norm": 1.0860278606414795,
+      "learning_rate": 0.0001797756662057287,
+      "loss": 3.5149,
+      "step": 3357
+    },
+    {
+      "epoch": 0.20656988188976377,
+      "grad_norm": 1.340308666229248,
+      "learning_rate": 0.00017976400451449274,
+      "loss": 2.5872,
+      "step": 3358
+    },
+    {
+      "epoch": 0.20663139763779528,
+      "grad_norm": 1.2503540515899658,
+      "learning_rate": 0.0001797523398405278,
+      "loss": 2.6938,
+      "step": 3359
+    },
+    {
+      "epoch": 0.20669291338582677,
+      "grad_norm": 1.2182432413101196,
+      "learning_rate": 0.00017974067218427003,
+      "loss": 3.2749,
+      "step": 3360
+    },
+    {
+      "epoch": 0.20675442913385828,
+      "grad_norm": 1.2690958976745605,
+      "learning_rate": 0.0001797290015461558,
+      "loss": 2.8509,
+      "step": 3361
+    },
+    {
+      "epoch": 0.20681594488188976,
+      "grad_norm": 1.3701444864273071,
+      "learning_rate": 0.0001797173279266215,
+      "loss": 2.7761,
+      "step": 3362
+    },
+    {
+      "epoch": 0.20687746062992127,
+      "grad_norm": 1.3623021841049194,
+      "learning_rate": 0.00017970565132610365,
+      "loss": 2.4911,
+      "step": 3363
+    },
+    {
+      "epoch": 0.20693897637795275,
+      "grad_norm": 1.2444992065429688,
+      "learning_rate": 0.00017969397174503885,
+      "loss": 2.3709,
+      "step": 3364
+    },
+    {
+      "epoch": 0.20700049212598426,
+      "grad_norm": 1.1051099300384521,
+      "learning_rate": 0.0001796822891838639,
+      "loss": 3.0436,
+      "step": 3365
+    },
+    {
+      "epoch": 0.20706200787401574,
+      "grad_norm": 0.9171050786972046,
+      "learning_rate": 0.00017967060364301572,
+      "loss": 3.2158,
+      "step": 3366
+    },
+    {
+      "epoch": 0.20712352362204725,
+      "grad_norm": 1.2564854621887207,
+      "learning_rate": 0.00017965891512293115,
+      "loss": 2.9092,
+      "step": 3367
+    },
+    {
+      "epoch": 0.20718503937007873,
+      "grad_norm": 1.0708986520767212,
+      "learning_rate": 0.00017964722362404735,
+      "loss": 2.7713,
+      "step": 3368
+    },
+    {
+      "epoch": 0.20724655511811024,
+      "grad_norm": 1.1321907043457031,
+      "learning_rate": 0.00017963552914680155,
+      "loss": 2.262,
+      "step": 3369
+    },
+    {
+      "epoch": 0.20730807086614172,
+      "grad_norm": 1.6474391222000122,
+      "learning_rate": 0.000179623831691631,
+      "loss": 2.9039,
+      "step": 3370
+    },
+    {
+      "epoch": 0.20736958661417323,
+      "grad_norm": 1.1167891025543213,
+      "learning_rate": 0.00017961213125897315,
+      "loss": 2.7181,
+      "step": 3371
+    },
+    {
+      "epoch": 0.20743110236220472,
+      "grad_norm": 1.2050490379333496,
+      "learning_rate": 0.00017960042784926549,
+      "loss": 1.6173,
+      "step": 3372
+    },
+    {
+      "epoch": 0.20749261811023623,
+      "grad_norm": 0.9094334840774536,
+      "learning_rate": 0.00017958872146294573,
+      "loss": 2.6179,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2075541338582677,
+      "grad_norm": 1.0480268001556396,
+      "learning_rate": 0.00017957701210045155,
+      "loss": 2.674,
+      "step": 3374
+    },
+    {
+      "epoch": 0.20761564960629922,
+      "grad_norm": 0.9275608658790588,
+      "learning_rate": 0.00017956529976222087,
+      "loss": 3.0159,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2076771653543307,
+      "grad_norm": 1.253602147102356,
+      "learning_rate": 0.00017955358444869167,
+      "loss": 2.8091,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2077386811023622,
+      "grad_norm": 1.105879306793213,
+      "learning_rate": 0.00017954186616030198,
+      "loss": 2.1894,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2078001968503937,
+      "grad_norm": 0.9905300736427307,
+      "learning_rate": 0.00017953014489749,
+      "loss": 3.5553,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2078617125984252,
+      "grad_norm": 1.0873483419418335,
+      "learning_rate": 0.00017951842066069414,
+      "loss": 3.2292,
+      "step": 3379
+    },
+    {
+      "epoch": 0.20792322834645668,
+      "grad_norm": 1.1304577589035034,
+      "learning_rate": 0.00017950669345035274,
+      "loss": 2.8525,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2079847440944882,
+      "grad_norm": 1.0262571573257446,
+      "learning_rate": 0.00017949496326690432,
+      "loss": 2.4539,
+      "step": 3381
+    },
+    {
+      "epoch": 0.20804625984251968,
+      "grad_norm": 1.3065330982208252,
+      "learning_rate": 0.00017948323011078758,
+      "loss": 2.3995,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2081077755905512,
+      "grad_norm": 1.1711301803588867,
+      "learning_rate": 0.00017947149398244125,
+      "loss": 2.2418,
+      "step": 3383
+    },
+    {
+      "epoch": 0.20816929133858267,
+      "grad_norm": 1.046294093132019,
+      "learning_rate": 0.00017945975488230419,
+      "loss": 3.2252,
+      "step": 3384
+    },
+    {
+      "epoch": 0.20823080708661418,
+      "grad_norm": 1.0089293718338013,
+      "learning_rate": 0.00017944801281081534,
+      "loss": 2.8079,
+      "step": 3385
+    },
+    {
+      "epoch": 0.20829232283464566,
+      "grad_norm": 0.9528879523277283,
+      "learning_rate": 0.00017943626776841383,
+      "loss": 3.3244,
+      "step": 3386
+    },
+    {
+      "epoch": 0.20835383858267717,
+      "grad_norm": 1.319229245185852,
+      "learning_rate": 0.00017942451975553887,
+      "loss": 3.0301,
+      "step": 3387
+    },
+    {
+      "epoch": 0.20841535433070865,
+      "grad_norm": 1.083268165588379,
+      "learning_rate": 0.00017941276877262975,
+      "loss": 2.6156,
+      "step": 3388
+    },
+    {
+      "epoch": 0.20847687007874016,
+      "grad_norm": 0.9578465223312378,
+      "learning_rate": 0.0001794010148201259,
+      "loss": 2.6482,
+      "step": 3389
+    },
+    {
+      "epoch": 0.20853838582677164,
+      "grad_norm": 1.0800272226333618,
+      "learning_rate": 0.00017938925789846686,
+      "loss": 3.2841,
+      "step": 3390
+    },
+    {
+      "epoch": 0.20859990157480315,
+      "grad_norm": 1.1026525497436523,
+      "learning_rate": 0.00017937749800809224,
+      "loss": 2.7018,
+      "step": 3391
+    },
+    {
+      "epoch": 0.20866141732283464,
+      "grad_norm": 1.100376844406128,
+      "learning_rate": 0.00017936573514944182,
+      "loss": 2.7881,
+      "step": 3392
+    },
+    {
+      "epoch": 0.20872293307086615,
+      "grad_norm": 1.0984567403793335,
+      "learning_rate": 0.00017935396932295547,
+      "loss": 3.467,
+      "step": 3393
+    },
+    {
+      "epoch": 0.20878444881889763,
+      "grad_norm": 1.0686777830123901,
+      "learning_rate": 0.00017934220052907317,
+      "loss": 3.0696,
+      "step": 3394
+    },
+    {
+      "epoch": 0.20884596456692914,
+      "grad_norm": 1.4114290475845337,
+      "learning_rate": 0.00017933042876823495,
+      "loss": 2.8598,
+      "step": 3395
+    },
+    {
+      "epoch": 0.20890748031496062,
+      "grad_norm": 1.1470184326171875,
+      "learning_rate": 0.00017931865404088109,
+      "loss": 3.0333,
+      "step": 3396
+    },
+    {
+      "epoch": 0.20896899606299213,
+      "grad_norm": 1.0517842769622803,
+      "learning_rate": 0.00017930687634745183,
+      "loss": 3.0324,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2090305118110236,
+      "grad_norm": 1.0261609554290771,
+      "learning_rate": 0.00017929509568838766,
+      "loss": 2.4071,
+      "step": 3398
+    },
+    {
+      "epoch": 0.20909202755905512,
+      "grad_norm": 1.1048933267593384,
+      "learning_rate": 0.00017928331206412904,
+      "loss": 3.0177,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2091535433070866,
+      "grad_norm": 1.1941622495651245,
+      "learning_rate": 0.00017927152547511668,
+      "loss": 2.6485,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2092150590551181,
+      "grad_norm": 1.2509962320327759,
+      "learning_rate": 0.00017925973592179125,
+      "loss": 2.6225,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2092765748031496,
+      "grad_norm": 1.4118572473526,
+      "learning_rate": 0.00017924794340459373,
+      "loss": 2.8167,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2093380905511811,
+      "grad_norm": 1.3405729532241821,
+      "learning_rate": 0.00017923614792396495,
+      "loss": 2.7406,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2093996062992126,
+      "grad_norm": 1.0172096490859985,
+      "learning_rate": 0.0001792243494803461,
+      "loss": 3.2097,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2094611220472441,
+      "grad_norm": 0.9867218732833862,
+      "learning_rate": 0.00017921254807417837,
+      "loss": 2.8052,
+      "step": 3405
+    },
+    {
+      "epoch": 0.20952263779527558,
+      "grad_norm": 1.3202648162841797,
+      "learning_rate": 0.00017920074370590303,
+      "loss": 2.8115,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2095841535433071,
+      "grad_norm": 1.034397006034851,
+      "learning_rate": 0.0001791889363759615,
+      "loss": 2.2164,
+      "step": 3407
+    },
+    {
+      "epoch": 0.20964566929133857,
+      "grad_norm": 1.2713404893875122,
+      "learning_rate": 0.00017917712608479534,
+      "loss": 3.1042,
+      "step": 3408
+    },
+    {
+      "epoch": 0.20970718503937008,
+      "grad_norm": 1.1758406162261963,
+      "learning_rate": 0.00017916531283284614,
+      "loss": 3.0783,
+      "step": 3409
+    },
+    {
+      "epoch": 0.20976870078740156,
+      "grad_norm": 1.32837975025177,
+      "learning_rate": 0.0001791534966205557,
+      "loss": 2.6893,
+      "step": 3410
+    },
+    {
+      "epoch": 0.20983021653543307,
+      "grad_norm": 1.1039366722106934,
+      "learning_rate": 0.00017914167744836584,
+      "loss": 3.0583,
+      "step": 3411
+    },
+    {
+      "epoch": 0.20989173228346455,
+      "grad_norm": 1.3231396675109863,
+      "learning_rate": 0.00017912985531671853,
+      "loss": 2.6919,
+      "step": 3412
+    },
+    {
+      "epoch": 0.20995324803149606,
+      "grad_norm": 1.431348443031311,
+      "learning_rate": 0.0001791180302260559,
+      "loss": 2.7032,
+      "step": 3413
+    },
+    {
+      "epoch": 0.21001476377952755,
+      "grad_norm": 1.185782790184021,
+      "learning_rate": 0.00017910620217682013,
+      "loss": 2.7076,
+      "step": 3414
+    },
+    {
+      "epoch": 0.21007627952755906,
+      "grad_norm": 1.1262843608856201,
+      "learning_rate": 0.0001790943711694535,
+      "loss": 2.568,
+      "step": 3415
+    },
+    {
+      "epoch": 0.21013779527559054,
+      "grad_norm": 1.0105348825454712,
+      "learning_rate": 0.0001790825372043984,
+      "loss": 3.4487,
+      "step": 3416
+    },
+    {
+      "epoch": 0.21019931102362205,
+      "grad_norm": 0.930840253829956,
+      "learning_rate": 0.0001790707002820974,
+      "loss": 2.8886,
+      "step": 3417
+    },
+    {
+      "epoch": 0.21026082677165353,
+      "grad_norm": 1.2147191762924194,
+      "learning_rate": 0.00017905886040299316,
+      "loss": 2.9837,
+      "step": 3418
+    },
+    {
+      "epoch": 0.21032234251968504,
+      "grad_norm": 1.2073414325714111,
+      "learning_rate": 0.00017904701756752832,
+      "loss": 2.3313,
+      "step": 3419
+    },
+    {
+      "epoch": 0.21038385826771652,
+      "grad_norm": 1.1040332317352295,
+      "learning_rate": 0.00017903517177614585,
+      "loss": 3.008,
+      "step": 3420
+    },
+    {
+      "epoch": 0.21044537401574803,
+      "grad_norm": 1.9768810272216797,
+      "learning_rate": 0.00017902332302928864,
+      "loss": 2.2577,
+      "step": 3421
+    },
+    {
+      "epoch": 0.21050688976377951,
+      "grad_norm": 1.2044358253479004,
+      "learning_rate": 0.00017901147132739982,
+      "loss": 2.5778,
+      "step": 3422
+    },
+    {
+      "epoch": 0.21056840551181102,
+      "grad_norm": 0.9705337285995483,
+      "learning_rate": 0.00017899961667092252,
+      "loss": 2.6738,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2106299212598425,
+      "grad_norm": 1.6072447299957275,
+      "learning_rate": 0.0001789877590603001,
+      "loss": 3.031,
+      "step": 3424
+    },
+    {
+      "epoch": 0.21069143700787402,
+      "grad_norm": 1.1735402345657349,
+      "learning_rate": 0.00017897589849597592,
+      "loss": 2.6221,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2107529527559055,
+      "grad_norm": 1.3670343160629272,
+      "learning_rate": 0.00017896403497839352,
+      "loss": 2.7113,
+      "step": 3426
+    },
+    {
+      "epoch": 0.210814468503937,
+      "grad_norm": 1.0459520816802979,
+      "learning_rate": 0.00017895216850799655,
+      "loss": 2.8124,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2108759842519685,
+      "grad_norm": 1.0658642053604126,
+      "learning_rate": 0.0001789402990852287,
+      "loss": 2.9743,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2109375,
+      "grad_norm": 0.9472208619117737,
+      "learning_rate": 0.00017892842671053384,
+      "loss": 3.5488,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2109990157480315,
+      "grad_norm": 1.050059199333191,
+      "learning_rate": 0.00017891655138435594,
+      "loss": 3.3999,
+      "step": 3430
+    },
+    {
+      "epoch": 0.211060531496063,
+      "grad_norm": 0.9536513686180115,
+      "learning_rate": 0.0001789046731071391,
+      "loss": 3.1896,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2111220472440945,
+      "grad_norm": 1.2177181243896484,
+      "learning_rate": 0.00017889279187932746,
+      "loss": 2.9294,
+      "step": 3432
+    },
+    {
+      "epoch": 0.21118356299212598,
+      "grad_norm": 1.0772007703781128,
+      "learning_rate": 0.0001788809077013653,
+      "loss": 3.014,
+      "step": 3433
+    },
+    {
+      "epoch": 0.2112450787401575,
+      "grad_norm": 1.0787955522537231,
+      "learning_rate": 0.00017886902057369705,
+      "loss": 2.8754,
+      "step": 3434
+    },
+    {
+      "epoch": 0.21130659448818898,
+      "grad_norm": 1.083646297454834,
+      "learning_rate": 0.00017885713049676724,
+      "loss": 2.6002,
+      "step": 3435
+    },
+    {
+      "epoch": 0.21136811023622049,
+      "grad_norm": 0.9147472977638245,
+      "learning_rate": 0.00017884523747102045,
+      "loss": 3.3956,
+      "step": 3436
+    },
+    {
+      "epoch": 0.21142962598425197,
+      "grad_norm": 0.9900703430175781,
+      "learning_rate": 0.00017883334149690146,
+      "loss": 2.8203,
+      "step": 3437
+    },
+    {
+      "epoch": 0.21149114173228348,
+      "grad_norm": 0.981787919998169,
+      "learning_rate": 0.00017882144257485503,
+      "loss": 2.7862,
+      "step": 3438
+    },
+    {
+      "epoch": 0.21155265748031496,
+      "grad_norm": 0.9202532768249512,
+      "learning_rate": 0.0001788095407053262,
+      "loss": 3.1484,
+      "step": 3439
+    },
+    {
+      "epoch": 0.21161417322834647,
+      "grad_norm": 1.2718428373336792,
+      "learning_rate": 0.00017879763588876,
+      "loss": 2.9679,
+      "step": 3440
+    },
+    {
+      "epoch": 0.21167568897637795,
+      "grad_norm": 1.2176235914230347,
+      "learning_rate": 0.0001787857281256016,
+      "loss": 2.6884,
+      "step": 3441
+    },
+    {
+      "epoch": 0.21173720472440946,
+      "grad_norm": 1.8733848333358765,
+      "learning_rate": 0.00017877381741629632,
+      "loss": 2.5039,
+      "step": 3442
+    },
+    {
+      "epoch": 0.21179872047244094,
+      "grad_norm": 1.2976387739181519,
+      "learning_rate": 0.00017876190376128948,
+      "loss": 2.9595,
+      "step": 3443
+    },
+    {
+      "epoch": 0.21186023622047245,
+      "grad_norm": 1.0426265001296997,
+      "learning_rate": 0.00017874998716102666,
+      "loss": 3.2486,
+      "step": 3444
+    },
+    {
+      "epoch": 0.21192175196850394,
+      "grad_norm": 1.1980743408203125,
+      "learning_rate": 0.00017873806761595345,
+      "loss": 2.4985,
+      "step": 3445
+    },
+    {
+      "epoch": 0.21198326771653545,
+      "grad_norm": 1.0863741636276245,
+      "learning_rate": 0.0001787261451265155,
+      "loss": 2.9577,
+      "step": 3446
+    },
+    {
+      "epoch": 0.21204478346456693,
+      "grad_norm": 1.8684556484222412,
+      "learning_rate": 0.00017871421969315878,
+      "loss": 2.0643,
+      "step": 3447
+    },
+    {
+      "epoch": 0.21210629921259844,
+      "grad_norm": 1.0975627899169922,
+      "learning_rate": 0.00017870229131632914,
+      "loss": 3.0545,
+      "step": 3448
+    },
+    {
+      "epoch": 0.21216781496062992,
+      "grad_norm": 1.1611520051956177,
+      "learning_rate": 0.00017869035999647267,
+      "loss": 2.7832,
+      "step": 3449
+    },
+    {
+      "epoch": 0.21222933070866143,
+      "grad_norm": 1.1123793125152588,
+      "learning_rate": 0.00017867842573403552,
+      "loss": 3.1826,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2122908464566929,
+      "grad_norm": 1.0625782012939453,
+      "learning_rate": 0.00017866648852946394,
+      "loss": 2.675,
+      "step": 3451
+    },
+    {
+      "epoch": 0.21235236220472442,
+      "grad_norm": 1.0446938276290894,
+      "learning_rate": 0.0001786545483832044,
+      "loss": 2.7626,
+      "step": 3452
+    },
+    {
+      "epoch": 0.2124138779527559,
+      "grad_norm": 1.0759763717651367,
+      "learning_rate": 0.0001786426052957033,
+      "loss": 2.4259,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2124753937007874,
+      "grad_norm": 2.3185644149780273,
+      "learning_rate": 0.0001786306592674073,
+      "loss": 2.3038,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2125369094488189,
+      "grad_norm": 1.304491400718689,
+      "learning_rate": 0.0001786187102987631,
+      "loss": 2.9729,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2125984251968504,
+      "grad_norm": 1.0392402410507202,
+      "learning_rate": 0.00017860675839021753,
+      "loss": 2.8849,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2126599409448819,
+      "grad_norm": 1.20667564868927,
+      "learning_rate": 0.00017859480354221754,
+      "loss": 2.8866,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2127214566929134,
+      "grad_norm": 1.4392369985580444,
+      "learning_rate": 0.0001785828457552101,
+      "loss": 2.8588,
+      "step": 3458
+    },
+    {
+      "epoch": 0.21278297244094488,
+      "grad_norm": 0.9108123183250427,
+      "learning_rate": 0.00017857088502964245,
+      "loss": 3.1609,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2128444881889764,
+      "grad_norm": 1.0792829990386963,
+      "learning_rate": 0.00017855892136596185,
+      "loss": 2.5765,
+      "step": 3460
+    },
+    {
+      "epoch": 0.21290600393700787,
+      "grad_norm": 1.32925283908844,
+      "learning_rate": 0.00017854695476461562,
+      "loss": 2.475,
+      "step": 3461
+    },
+    {
+      "epoch": 0.21296751968503938,
+      "grad_norm": 1.8663039207458496,
+      "learning_rate": 0.00017853498522605126,
+      "loss": 1.7537,
+      "step": 3462
+    },
+    {
+      "epoch": 0.21302903543307086,
+      "grad_norm": 1.0351333618164062,
+      "learning_rate": 0.00017852301275071637,
+      "loss": 2.9399,
+      "step": 3463
+    },
+    {
+      "epoch": 0.21309055118110237,
+      "grad_norm": 1.1798427104949951,
+      "learning_rate": 0.0001785110373390587,
+      "loss": 3.5201,
+      "step": 3464
+    },
+    {
+      "epoch": 0.21315206692913385,
+      "grad_norm": 1.0265551805496216,
+      "learning_rate": 0.00017849905899152598,
+      "loss": 2.767,
+      "step": 3465
+    },
+    {
+      "epoch": 0.21321358267716536,
+      "grad_norm": 1.379037857055664,
+      "learning_rate": 0.00017848707770856619,
+      "loss": 3.1726,
+      "step": 3466
+    },
+    {
+      "epoch": 0.21327509842519685,
+      "grad_norm": 1.1058387756347656,
+      "learning_rate": 0.00017847509349062736,
+      "loss": 2.7013,
+      "step": 3467
+    },
+    {
+      "epoch": 0.21333661417322836,
+      "grad_norm": 1.1397385597229004,
+      "learning_rate": 0.00017846310633815761,
+      "loss": 2.6472,
+      "step": 3468
+    },
+    {
+      "epoch": 0.21339812992125984,
+      "grad_norm": 1.2276667356491089,
+      "learning_rate": 0.0001784511162516052,
+      "loss": 2.7063,
+      "step": 3469
+    },
+    {
+      "epoch": 0.21345964566929135,
+      "grad_norm": 1.1488730907440186,
+      "learning_rate": 0.00017843912323141852,
+      "loss": 3.2972,
+      "step": 3470
+    },
+    {
+      "epoch": 0.21352116141732283,
+      "grad_norm": 1.3565131425857544,
+      "learning_rate": 0.000178427127278046,
+      "loss": 1.8406,
+      "step": 3471
+    },
+    {
+      "epoch": 0.21358267716535434,
+      "grad_norm": 1.2061078548431396,
+      "learning_rate": 0.00017841512839193626,
+      "loss": 2.7805,
+      "step": 3472
+    },
+    {
+      "epoch": 0.21364419291338582,
+      "grad_norm": 1.0344724655151367,
+      "learning_rate": 0.00017840312657353797,
+      "loss": 2.8319,
+      "step": 3473
+    },
+    {
+      "epoch": 0.21370570866141733,
+      "grad_norm": 1.2724344730377197,
+      "learning_rate": 0.00017839112182329992,
+      "loss": 3.3758,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2137672244094488,
+      "grad_norm": 1.1280310153961182,
+      "learning_rate": 0.00017837911414167104,
+      "loss": 3.0109,
+      "step": 3475
+    },
+    {
+      "epoch": 0.21382874015748032,
+      "grad_norm": 1.0856051445007324,
+      "learning_rate": 0.00017836710352910035,
+      "loss": 2.336,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2138902559055118,
+      "grad_norm": 1.1103307008743286,
+      "learning_rate": 0.00017835508998603698,
+      "loss": 2.7529,
+      "step": 3477
+    },
+    {
+      "epoch": 0.21395177165354332,
+      "grad_norm": 1.1494650840759277,
+      "learning_rate": 0.00017834307351293017,
+      "loss": 2.6987,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2140132874015748,
+      "grad_norm": 1.6910741329193115,
+      "learning_rate": 0.0001783310541102292,
+      "loss": 2.1812,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2140748031496063,
+      "grad_norm": 1.0534039735794067,
+      "learning_rate": 0.00017831903177838366,
+      "loss": 3.5454,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2141363188976378,
+      "grad_norm": 1.224324107170105,
+      "learning_rate": 0.00017830700651784304,
+      "loss": 2.1664,
+      "step": 3481
+    },
+    {
+      "epoch": 0.2141978346456693,
+      "grad_norm": 1.3119641542434692,
+      "learning_rate": 0.00017829497832905702,
+      "loss": 2.6302,
+      "step": 3482
+    },
+    {
+      "epoch": 0.21425935039370078,
+      "grad_norm": 1.2114183902740479,
+      "learning_rate": 0.0001782829472124754,
+      "loss": 2.7141,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2143208661417323,
+      "grad_norm": 1.1626266241073608,
+      "learning_rate": 0.00017827091316854806,
+      "loss": 2.7036,
+      "step": 3484
+    },
+    {
+      "epoch": 0.21438238188976377,
+      "grad_norm": 1.0681374073028564,
+      "learning_rate": 0.00017825887619772504,
+      "loss": 1.8832,
+      "step": 3485
+    },
+    {
+      "epoch": 0.21444389763779528,
+      "grad_norm": 1.1081438064575195,
+      "learning_rate": 0.0001782468363004564,
+      "loss": 2.9676,
+      "step": 3486
+    },
+    {
+      "epoch": 0.21450541338582677,
+      "grad_norm": 1.2804137468338013,
+      "learning_rate": 0.00017823479347719244,
+      "loss": 2.0807,
+      "step": 3487
+    },
+    {
+      "epoch": 0.21456692913385828,
+      "grad_norm": 1.106576681137085,
+      "learning_rate": 0.0001782227477283834,
+      "loss": 2.3049,
+      "step": 3488
+    },
+    {
+      "epoch": 0.21462844488188976,
+      "grad_norm": 1.5100988149642944,
+      "learning_rate": 0.00017821069905447982,
+      "loss": 2.4464,
+      "step": 3489
+    },
+    {
+      "epoch": 0.21468996062992127,
+      "grad_norm": 1.0160789489746094,
+      "learning_rate": 0.0001781986474559322,
+      "loss": 3.3639,
+      "step": 3490
+    },
+    {
+      "epoch": 0.21475147637795275,
+      "grad_norm": 1.2841819524765015,
+      "learning_rate": 0.00017818659293319123,
+      "loss": 3.5131,
+      "step": 3491
+    },
+    {
+      "epoch": 0.21481299212598426,
+      "grad_norm": 1.0503064393997192,
+      "learning_rate": 0.00017817453548670762,
+      "loss": 3.3547,
+      "step": 3492
+    },
+    {
+      "epoch": 0.21487450787401574,
+      "grad_norm": 1.1251400709152222,
+      "learning_rate": 0.00017816247511693236,
+      "loss": 2.1102,
+      "step": 3493
+    },
+    {
+      "epoch": 0.21493602362204725,
+      "grad_norm": 1.3862707614898682,
+      "learning_rate": 0.00017815041182431634,
+      "loss": 3.4792,
+      "step": 3494
+    },
+    {
+      "epoch": 0.21499753937007873,
+      "grad_norm": 1.7073233127593994,
+      "learning_rate": 0.0001781383456093107,
+      "loss": 3.2256,
+      "step": 3495
+    },
+    {
+      "epoch": 0.21505905511811024,
+      "grad_norm": 1.3322699069976807,
+      "learning_rate": 0.00017812627647236662,
+      "loss": 2.7801,
+      "step": 3496
+    },
+    {
+      "epoch": 0.21512057086614172,
+      "grad_norm": 1.0148906707763672,
+      "learning_rate": 0.00017811420441393545,
+      "loss": 2.8192,
+      "step": 3497
+    },
+    {
+      "epoch": 0.21518208661417323,
+      "grad_norm": 1.1913801431655884,
+      "learning_rate": 0.00017810212943446865,
+      "loss": 2.7179,
+      "step": 3498
+    },
+    {
+      "epoch": 0.21524360236220472,
+      "grad_norm": 1.0370930433273315,
+      "learning_rate": 0.00017809005153441768,
+      "loss": 3.23,
+      "step": 3499
+    },
+    {
+      "epoch": 0.21530511811023623,
+      "grad_norm": 1.0929335355758667,
+      "learning_rate": 0.00017807797071423424,
+      "loss": 2.9295,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2153666338582677,
+      "grad_norm": 1.6249855756759644,
+      "learning_rate": 0.00017806588697437005,
+      "loss": 3.0322,
+      "step": 3501
+    },
+    {
+      "epoch": 0.21542814960629922,
+      "grad_norm": 1.173597812652588,
+      "learning_rate": 0.00017805380031527703,
+      "loss": 2.6207,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2154896653543307,
+      "grad_norm": 1.1055972576141357,
+      "learning_rate": 0.0001780417107374071,
+      "loss": 2.9746,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2155511811023622,
+      "grad_norm": 1.0831269025802612,
+      "learning_rate": 0.00017802961824121236,
+      "loss": 2.5899,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2156126968503937,
+      "grad_norm": 1.1632916927337646,
+      "learning_rate": 0.000178017522827145,
+      "loss": 2.4211,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2156742125984252,
+      "grad_norm": 1.219599723815918,
+      "learning_rate": 0.00017800542449565736,
+      "loss": 2.7511,
+      "step": 3506
+    },
+    {
+      "epoch": 0.21573572834645668,
+      "grad_norm": 1.2358546257019043,
+      "learning_rate": 0.00017799332324720175,
+      "loss": 2.6675,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2157972440944882,
+      "grad_norm": 1.7749149799346924,
+      "learning_rate": 0.0001779812190822308,
+      "loss": 2.7744,
+      "step": 3508
+    },
+    {
+      "epoch": 0.21585875984251968,
+      "grad_norm": 1.2864423990249634,
+      "learning_rate": 0.00017796911200119707,
+      "loss": 3.4823,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2159202755905512,
+      "grad_norm": 1.0928022861480713,
+      "learning_rate": 0.00017795700200455333,
+      "loss": 2.4408,
+      "step": 3510
+    },
+    {
+      "epoch": 0.21598179133858267,
+      "grad_norm": 1.5277315378189087,
+      "learning_rate": 0.0001779448890927524,
+      "loss": 3.1357,
+      "step": 3511
+    },
+    {
+      "epoch": 0.21604330708661418,
+      "grad_norm": 0.8891454339027405,
+      "learning_rate": 0.0001779327732662473,
+      "loss": 2.8459,
+      "step": 3512
+    },
+    {
+      "epoch": 0.21610482283464566,
+      "grad_norm": 1.0679755210876465,
+      "learning_rate": 0.000177920654525491,
+      "loss": 2.4824,
+      "step": 3513
+    },
+    {
+      "epoch": 0.21616633858267717,
+      "grad_norm": 1.2992061376571655,
+      "learning_rate": 0.00017790853287093673,
+      "loss": 1.8322,
+      "step": 3514
+    },
+    {
+      "epoch": 0.21622785433070865,
+      "grad_norm": 1.3357772827148438,
+      "learning_rate": 0.00017789640830303776,
+      "loss": 2.8454,
+      "step": 3515
+    },
+    {
+      "epoch": 0.21628937007874016,
+      "grad_norm": 1.2166681289672852,
+      "learning_rate": 0.00017788428082224746,
+      "loss": 2.4548,
+      "step": 3516
+    },
+    {
+      "epoch": 0.21635088582677164,
+      "grad_norm": 1.403112769126892,
+      "learning_rate": 0.00017787215042901937,
+      "loss": 0.9259,
+      "step": 3517
+    },
+    {
+      "epoch": 0.21641240157480315,
+      "grad_norm": 1.220599889755249,
+      "learning_rate": 0.0001778600171238071,
+      "loss": 3.0058,
+      "step": 3518
+    },
+    {
+      "epoch": 0.21647391732283464,
+      "grad_norm": 0.9902350902557373,
+      "learning_rate": 0.00017784788090706433,
+      "loss": 3.3722,
+      "step": 3519
+    },
+    {
+      "epoch": 0.21653543307086615,
+      "grad_norm": 1.402869462966919,
+      "learning_rate": 0.0001778357417792449,
+      "loss": 2.5322,
+      "step": 3520
+    },
+    {
+      "epoch": 0.21659694881889763,
+      "grad_norm": 1.2562041282653809,
+      "learning_rate": 0.00017782359974080275,
+      "loss": 2.3598,
+      "step": 3521
+    },
+    {
+      "epoch": 0.21665846456692914,
+      "grad_norm": 1.1125564575195312,
+      "learning_rate": 0.00017781145479219194,
+      "loss": 3.0521,
+      "step": 3522
+    },
+    {
+      "epoch": 0.21671998031496062,
+      "grad_norm": 1.1387441158294678,
+      "learning_rate": 0.00017779930693386658,
+      "loss": 2.4959,
+      "step": 3523
+    },
+    {
+      "epoch": 0.21678149606299213,
+      "grad_norm": 1.0169867277145386,
+      "learning_rate": 0.000177787156166281,
+      "loss": 2.9807,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2168430118110236,
+      "grad_norm": 1.2449568510055542,
+      "learning_rate": 0.00017777500248988953,
+      "loss": 3.5936,
+      "step": 3525
+    },
+    {
+      "epoch": 0.21690452755905512,
+      "grad_norm": 1.238136887550354,
+      "learning_rate": 0.0001777628459051466,
+      "loss": 2.1391,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2169660433070866,
+      "grad_norm": 1.0268182754516602,
+      "learning_rate": 0.00017775068641250692,
+      "loss": 2.514,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2170275590551181,
+      "grad_norm": 1.2687807083129883,
+      "learning_rate": 0.00017773852401242507,
+      "loss": 3.1189,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2170890748031496,
+      "grad_norm": 1.0575512647628784,
+      "learning_rate": 0.00017772635870535593,
+      "loss": 2.7932,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2171505905511811,
+      "grad_norm": 1.1654179096221924,
+      "learning_rate": 0.00017771419049175438,
+      "loss": 2.9429,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2172121062992126,
+      "grad_norm": 1.8106739521026611,
+      "learning_rate": 0.00017770201937207544,
+      "loss": 2.9047,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2172736220472441,
+      "grad_norm": 1.2780232429504395,
+      "learning_rate": 0.0001776898453467743,
+      "loss": 2.856,
+      "step": 3532
+    },
+    {
+      "epoch": 0.21733513779527558,
+      "grad_norm": 1.083888292312622,
+      "learning_rate": 0.0001776776684163061,
+      "loss": 2.4329,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2173966535433071,
+      "grad_norm": 0.97296541929245,
+      "learning_rate": 0.0001776654885811263,
+      "loss": 3.4473,
+      "step": 3534
+    },
+    {
+      "epoch": 0.21745816929133857,
+      "grad_norm": 1.371084451675415,
+      "learning_rate": 0.00017765330584169022,
+      "loss": 2.3877,
+      "step": 3535
+    },
+    {
+      "epoch": 0.21751968503937008,
+      "grad_norm": 1.2002251148223877,
+      "learning_rate": 0.00017764112019845357,
+      "loss": 2.7074,
+      "step": 3536
+    },
+    {
+      "epoch": 0.21758120078740156,
+      "grad_norm": 1.093685269355774,
+      "learning_rate": 0.00017762893165187196,
+      "loss": 2.1601,
+      "step": 3537
+    },
+    {
+      "epoch": 0.21764271653543307,
+      "grad_norm": 1.1541621685028076,
+      "learning_rate": 0.00017761674020240115,
+      "loss": 2.6815,
+      "step": 3538
+    },
+    {
+      "epoch": 0.21770423228346455,
+      "grad_norm": 1.1688827276229858,
+      "learning_rate": 0.0001776045458504971,
+      "loss": 2.7517,
+      "step": 3539
+    },
+    {
+      "epoch": 0.21776574803149606,
+      "grad_norm": 1.101555347442627,
+      "learning_rate": 0.0001775923485966157,
+      "loss": 2.9295,
+      "step": 3540
+    },
+    {
+      "epoch": 0.21782726377952755,
+      "grad_norm": 1.1689693927764893,
+      "learning_rate": 0.00017758014844121317,
+      "loss": 2.9603,
+      "step": 3541
+    },
+    {
+      "epoch": 0.21788877952755906,
+      "grad_norm": 1.0624934434890747,
+      "learning_rate": 0.0001775679453847457,
+      "loss": 2.7323,
+      "step": 3542
+    },
+    {
+      "epoch": 0.21795029527559054,
+      "grad_norm": 1.0984797477722168,
+      "learning_rate": 0.0001775557394276696,
+      "loss": 3.2039,
+      "step": 3543
+    },
+    {
+      "epoch": 0.21801181102362205,
+      "grad_norm": 0.9518739581108093,
+      "learning_rate": 0.0001775435305704413,
+      "loss": 2.9458,
+      "step": 3544
+    },
+    {
+      "epoch": 0.21807332677165353,
+      "grad_norm": 1.0952554941177368,
+      "learning_rate": 0.00017753131881351734,
+      "loss": 2.6599,
+      "step": 3545
+    },
+    {
+      "epoch": 0.21813484251968504,
+      "grad_norm": 1.4689362049102783,
+      "learning_rate": 0.00017751910415735443,
+      "loss": 2.7213,
+      "step": 3546
+    },
+    {
+      "epoch": 0.21819635826771652,
+      "grad_norm": 1.0249031782150269,
+      "learning_rate": 0.0001775068866024092,
+      "loss": 3.307,
+      "step": 3547
+    },
+    {
+      "epoch": 0.21825787401574803,
+      "grad_norm": 1.1485462188720703,
+      "learning_rate": 0.0001774946661491387,
+      "loss": 2.9454,
+      "step": 3548
+    },
+    {
+      "epoch": 0.21831938976377951,
+      "grad_norm": 1.032033920288086,
+      "learning_rate": 0.00017748244279799976,
+      "loss": 2.9812,
+      "step": 3549
+    },
+    {
+      "epoch": 0.21838090551181102,
+      "grad_norm": 1.1632360219955444,
+      "learning_rate": 0.00017747021654944948,
+      "loss": 2.4667,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2184424212598425,
+      "grad_norm": 1.0317723751068115,
+      "learning_rate": 0.00017745798740394513,
+      "loss": 2.9419,
+      "step": 3551
+    },
+    {
+      "epoch": 0.21850393700787402,
+      "grad_norm": 1.1212997436523438,
+      "learning_rate": 0.00017744575536194395,
+      "loss": 2.7854,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2185654527559055,
+      "grad_norm": 1.0436707735061646,
+      "learning_rate": 0.0001774335204239034,
+      "loss": 2.4326,
+      "step": 3553
+    },
+    {
+      "epoch": 0.218626968503937,
+      "grad_norm": 1.0879672765731812,
+      "learning_rate": 0.00017742128259028094,
+      "loss": 3.2347,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2186884842519685,
+      "grad_norm": 1.038681149482727,
+      "learning_rate": 0.0001774090418615342,
+      "loss": 2.8196,
+      "step": 3555
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 1.232180118560791,
+      "learning_rate": 0.00017739679823812097,
+      "loss": 2.6483,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2188115157480315,
+      "grad_norm": 1.3412179946899414,
+      "learning_rate": 0.00017738455172049906,
+      "loss": 2.7886,
+      "step": 3557
+    },
+    {
+      "epoch": 0.218873031496063,
+      "grad_norm": 1.6009882688522339,
+      "learning_rate": 0.0001773723023091264,
+      "loss": 2.6273,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2189345472440945,
+      "grad_norm": 1.4327484369277954,
+      "learning_rate": 0.00017736005000446113,
+      "loss": 2.8457,
+      "step": 3559
+    },
+    {
+      "epoch": 0.21899606299212598,
+      "grad_norm": 1.2265633344650269,
+      "learning_rate": 0.0001773477948069613,
+      "loss": 2.5978,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2190575787401575,
+      "grad_norm": 0.8623350262641907,
+      "learning_rate": 0.00017733553671708528,
+      "loss": 1.6124,
+      "step": 3561
+    },
+    {
+      "epoch": 0.21911909448818898,
+      "grad_norm": 1.0167338848114014,
+      "learning_rate": 0.0001773232757352914,
+      "loss": 2.9215,
+      "step": 3562
+    },
+    {
+      "epoch": 0.21918061023622049,
+      "grad_norm": 1.2758290767669678,
+      "learning_rate": 0.0001773110118620382,
+      "loss": 2.4334,
+      "step": 3563
+    },
+    {
+      "epoch": 0.21924212598425197,
+      "grad_norm": 1.1662333011627197,
+      "learning_rate": 0.00017729874509778425,
+      "loss": 2.1102,
+      "step": 3564
+    },
+    {
+      "epoch": 0.21930364173228348,
+      "grad_norm": 1.1713961362838745,
+      "learning_rate": 0.00017728647544298825,
+      "loss": 3.0083,
+      "step": 3565
+    },
+    {
+      "epoch": 0.21936515748031496,
+      "grad_norm": 1.1610684394836426,
+      "learning_rate": 0.00017727420289810906,
+      "loss": 3.061,
+      "step": 3566
+    },
+    {
+      "epoch": 0.21942667322834647,
+      "grad_norm": 1.2702356576919556,
+      "learning_rate": 0.00017726192746360555,
+      "loss": 2.7268,
+      "step": 3567
+    },
+    {
+      "epoch": 0.21948818897637795,
+      "grad_norm": 1.3367252349853516,
+      "learning_rate": 0.00017724964913993677,
+      "loss": 2.9664,
+      "step": 3568
+    },
+    {
+      "epoch": 0.21954970472440946,
+      "grad_norm": 1.2168015241622925,
+      "learning_rate": 0.00017723736792756188,
+      "loss": 2.6723,
+      "step": 3569
+    },
+    {
+      "epoch": 0.21961122047244094,
+      "grad_norm": 1.2044399976730347,
+      "learning_rate": 0.00017722508382694008,
+      "loss": 2.9462,
+      "step": 3570
+    },
+    {
+      "epoch": 0.21967273622047245,
+      "grad_norm": 1.280274510383606,
+      "learning_rate": 0.00017721279683853084,
+      "loss": 2.764,
+      "step": 3571
+    },
+    {
+      "epoch": 0.21973425196850394,
+      "grad_norm": 1.2335952520370483,
+      "learning_rate": 0.00017720050696279347,
+      "loss": 2.1551,
+      "step": 3572
+    },
+    {
+      "epoch": 0.21979576771653545,
+      "grad_norm": 1.3460063934326172,
+      "learning_rate": 0.00017718821420018767,
+      "loss": 2.9436,
+      "step": 3573
+    },
+    {
+      "epoch": 0.21985728346456693,
+      "grad_norm": 1.2745500802993774,
+      "learning_rate": 0.00017717591855117307,
+      "loss": 1.6406,
+      "step": 3574
+    },
+    {
+      "epoch": 0.21991879921259844,
+      "grad_norm": 1.1861491203308105,
+      "learning_rate": 0.00017716362001620942,
+      "loss": 2.9532,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21998031496062992,
+      "grad_norm": 1.0230655670166016,
+      "learning_rate": 0.00017715131859575672,
+      "loss": 2.9587,
+      "step": 3576
+    },
+    {
+      "epoch": 0.22004183070866143,
+      "grad_norm": 1.6169592142105103,
+      "learning_rate": 0.00017713901429027487,
+      "loss": 3.543,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2201033464566929,
+      "grad_norm": 1.0943502187728882,
+      "learning_rate": 0.00017712670710022402,
+      "loss": 2.87,
+      "step": 3578
+    },
+    {
+      "epoch": 0.22016486220472442,
+      "grad_norm": 1.0527218580245972,
+      "learning_rate": 0.00017711439702606442,
+      "loss": 2.9282,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2202263779527559,
+      "grad_norm": 1.126314401626587,
+      "learning_rate": 0.00017710208406825637,
+      "loss": 2.7035,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2202878937007874,
+      "grad_norm": 0.945110023021698,
+      "learning_rate": 0.0001770897682272603,
+      "loss": 2.8308,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2203494094488189,
+      "grad_norm": 1.0020866394042969,
+      "learning_rate": 0.00017707744950353673,
+      "loss": 2.882,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2204109251968504,
+      "grad_norm": 1.185280203819275,
+      "learning_rate": 0.00017706512789754638,
+      "loss": 2.4782,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2204724409448819,
+      "grad_norm": 1.2268877029418945,
+      "learning_rate": 0.00017705280340975,
+      "loss": 2.6714,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2205339566929134,
+      "grad_norm": 1.3012269735336304,
+      "learning_rate": 0.0001770404760406084,
+      "loss": 2.4749,
+      "step": 3585
+    },
+    {
+      "epoch": 0.22059547244094488,
+      "grad_norm": 1.5876147747039795,
+      "learning_rate": 0.00017702814579058257,
+      "loss": 3.4616,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2206569881889764,
+      "grad_norm": 1.2405167818069458,
+      "learning_rate": 0.00017701581266013362,
+      "loss": 2.9692,
+      "step": 3587
+    },
+    {
+      "epoch": 0.22071850393700787,
+      "grad_norm": 1.0537197589874268,
+      "learning_rate": 0.00017700347664972274,
+      "loss": 3.0627,
+      "step": 3588
+    },
+    {
+      "epoch": 0.22078001968503938,
+      "grad_norm": 1.1566479206085205,
+      "learning_rate": 0.0001769911377598112,
+      "loss": 2.8135,
+      "step": 3589
+    },
+    {
+      "epoch": 0.22084153543307086,
+      "grad_norm": 1.0738964080810547,
+      "learning_rate": 0.00017697879599086044,
+      "loss": 3.4994,
+      "step": 3590
+    },
+    {
+      "epoch": 0.22090305118110237,
+      "grad_norm": 0.9701862335205078,
+      "learning_rate": 0.00017696645134333196,
+      "loss": 2.2822,
+      "step": 3591
+    },
+    {
+      "epoch": 0.22096456692913385,
+      "grad_norm": 1.1575413942337036,
+      "learning_rate": 0.00017695410381768736,
+      "loss": 2.801,
+      "step": 3592
+    },
+    {
+      "epoch": 0.22102608267716536,
+      "grad_norm": 1.2251673936843872,
+      "learning_rate": 0.0001769417534143884,
+      "loss": 3.0494,
+      "step": 3593
+    },
+    {
+      "epoch": 0.22108759842519685,
+      "grad_norm": 1.1632572412490845,
+      "learning_rate": 0.00017692940013389687,
+      "loss": 2.2947,
+      "step": 3594
+    },
+    {
+      "epoch": 0.22114911417322836,
+      "grad_norm": 1.2278437614440918,
+      "learning_rate": 0.00017691704397667477,
+      "loss": 3.2421,
+      "step": 3595
+    },
+    {
+      "epoch": 0.22121062992125984,
+      "grad_norm": 1.478830337524414,
+      "learning_rate": 0.00017690468494318412,
+      "loss": 3.0893,
+      "step": 3596
+    },
+    {
+      "epoch": 0.22127214566929135,
+      "grad_norm": 1.083123803138733,
+      "learning_rate": 0.00017689232303388707,
+      "loss": 2.0624,
+      "step": 3597
+    },
+    {
+      "epoch": 0.22133366141732283,
+      "grad_norm": 1.13132643699646,
+      "learning_rate": 0.00017687995824924592,
+      "loss": 2.6317,
+      "step": 3598
+    },
+    {
+      "epoch": 0.22139517716535434,
+      "grad_norm": 1.0814337730407715,
+      "learning_rate": 0.00017686759058972305,
+      "loss": 2.7627,
+      "step": 3599
+    },
+    {
+      "epoch": 0.22145669291338582,
+      "grad_norm": 1.0861002206802368,
+      "learning_rate": 0.00017685522005578088,
+      "loss": 2.5922,
+      "step": 3600
+    },
+    {
+      "epoch": 0.22151820866141733,
+      "grad_norm": 1.2223539352416992,
+      "learning_rate": 0.00017684284664788206,
+      "loss": 2.8235,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2215797244094488,
+      "grad_norm": 1.4076106548309326,
+      "learning_rate": 0.00017683047036648922,
+      "loss": 3.5307,
+      "step": 3602
+    },
+    {
+      "epoch": 0.22164124015748032,
+      "grad_norm": 1.679412603378296,
+      "learning_rate": 0.00017681809121206527,
+      "loss": 2.5386,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2217027559055118,
+      "grad_norm": 0.9967997670173645,
+      "learning_rate": 0.000176805709185073,
+      "loss": 2.9586,
+      "step": 3604
+    },
+    {
+      "epoch": 0.22176427165354332,
+      "grad_norm": 1.0507670640945435,
+      "learning_rate": 0.00017679332428597553,
+      "loss": 2.2351,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2218257874015748,
+      "grad_norm": 0.984281063079834,
+      "learning_rate": 0.00017678093651523594,
+      "loss": 2.5899,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2218873031496063,
+      "grad_norm": 1.0039386749267578,
+      "learning_rate": 0.00017676854587331744,
+      "loss": 3.1045,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2219488188976378,
+      "grad_norm": 1.2371011972427368,
+      "learning_rate": 0.00017675615236068342,
+      "loss": 3.111,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2220103346456693,
+      "grad_norm": 1.42075514793396,
+      "learning_rate": 0.00017674375597779727,
+      "loss": 2.1325,
+      "step": 3609
+    },
+    {
+      "epoch": 0.22207185039370078,
+      "grad_norm": 1.2394332885742188,
+      "learning_rate": 0.00017673135672512263,
+      "loss": 2.8919,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2221333661417323,
+      "grad_norm": 1.0538228750228882,
+      "learning_rate": 0.00017671895460312308,
+      "loss": 3.1477,
+      "step": 3611
+    },
+    {
+      "epoch": 0.22219488188976377,
+      "grad_norm": 1.0822559595108032,
+      "learning_rate": 0.00017670654961226244,
+      "loss": 2.9843,
+      "step": 3612
+    },
+    {
+      "epoch": 0.22225639763779528,
+      "grad_norm": 1.0310653448104858,
+      "learning_rate": 0.00017669414175300457,
+      "loss": 2.9401,
+      "step": 3613
+    },
+    {
+      "epoch": 0.22231791338582677,
+      "grad_norm": 1.087699055671692,
+      "learning_rate": 0.00017668173102581344,
+      "loss": 3.1694,
+      "step": 3614
+    },
+    {
+      "epoch": 0.22237942913385828,
+      "grad_norm": 1.0068763494491577,
+      "learning_rate": 0.0001766693174311532,
+      "loss": 2.9006,
+      "step": 3615
+    },
+    {
+      "epoch": 0.22244094488188976,
+      "grad_norm": 1.2553200721740723,
+      "learning_rate": 0.00017665690096948798,
+      "loss": 2.8578,
+      "step": 3616
+    },
+    {
+      "epoch": 0.22250246062992127,
+      "grad_norm": 1.0279422998428345,
+      "learning_rate": 0.00017664448164128209,
+      "loss": 2.6279,
+      "step": 3617
+    },
+    {
+      "epoch": 0.22256397637795275,
+      "grad_norm": 1.0819759368896484,
+      "learning_rate": 0.000176632059447,
+      "loss": 3.3382,
+      "step": 3618
+    },
+    {
+      "epoch": 0.22262549212598426,
+      "grad_norm": 1.0260863304138184,
+      "learning_rate": 0.0001766196343871062,
+      "loss": 3.351,
+      "step": 3619
+    },
+    {
+      "epoch": 0.22268700787401574,
+      "grad_norm": 1.0701364278793335,
+      "learning_rate": 0.0001766072064620653,
+      "loss": 3.0093,
+      "step": 3620
+    },
+    {
+      "epoch": 0.22274852362204725,
+      "grad_norm": 1.3157987594604492,
+      "learning_rate": 0.00017659477567234205,
+      "loss": 3.1318,
+      "step": 3621
+    },
+    {
+      "epoch": 0.22281003937007873,
+      "grad_norm": 1.0302443504333496,
+      "learning_rate": 0.00017658234201840134,
+      "loss": 2.8989,
+      "step": 3622
+    },
+    {
+      "epoch": 0.22287155511811024,
+      "grad_norm": 1.1028457880020142,
+      "learning_rate": 0.00017656990550070803,
+      "loss": 2.7069,
+      "step": 3623
+    },
+    {
+      "epoch": 0.22293307086614172,
+      "grad_norm": 1.2673828601837158,
+      "learning_rate": 0.00017655746611972722,
+      "loss": 2.8904,
+      "step": 3624
+    },
+    {
+      "epoch": 0.22299458661417323,
+      "grad_norm": 1.192366361618042,
+      "learning_rate": 0.0001765450238759241,
+      "loss": 2.4213,
+      "step": 3625
+    },
+    {
+      "epoch": 0.22305610236220472,
+      "grad_norm": 1.231314778327942,
+      "learning_rate": 0.00017653257876976388,
+      "loss": 2.7711,
+      "step": 3626
+    },
+    {
+      "epoch": 0.22311761811023623,
+      "grad_norm": 1.2061980962753296,
+      "learning_rate": 0.000176520130801712,
+      "loss": 3.1378,
+      "step": 3627
+    },
+    {
+      "epoch": 0.2231791338582677,
+      "grad_norm": 1.1256341934204102,
+      "learning_rate": 0.0001765076799722339,
+      "loss": 2.7832,
+      "step": 3628
+    },
+    {
+      "epoch": 0.22324064960629922,
+      "grad_norm": 1.2465423345565796,
+      "learning_rate": 0.00017649522628179518,
+      "loss": 2.916,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2233021653543307,
+      "grad_norm": 1.9543424844741821,
+      "learning_rate": 0.00017648276973086156,
+      "loss": 4.2254,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2233636811023622,
+      "grad_norm": 0.9809346199035645,
+      "learning_rate": 0.00017647031031989882,
+      "loss": 3.1264,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2234251968503937,
+      "grad_norm": 1.067562460899353,
+      "learning_rate": 0.00017645784804937292,
+      "loss": 3.4299,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2234867125984252,
+      "grad_norm": 1.488023281097412,
+      "learning_rate": 0.0001764453829197498,
+      "loss": 2.6635,
+      "step": 3633
+    },
+    {
+      "epoch": 0.22354822834645668,
+      "grad_norm": 1.1169947385787964,
+      "learning_rate": 0.00017643291493149566,
+      "loss": 2.4074,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2236097440944882,
+      "grad_norm": 1.0975677967071533,
+      "learning_rate": 0.00017642044408507665,
+      "loss": 3.1156,
+      "step": 3635
+    },
+    {
+      "epoch": 0.22367125984251968,
+      "grad_norm": 1.111701250076294,
+      "learning_rate": 0.00017640797038095921,
+      "loss": 2.7421,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2237327755905512,
+      "grad_norm": 1.2099716663360596,
+      "learning_rate": 0.00017639549381960972,
+      "loss": 2.7177,
+      "step": 3637
+    },
+    {
+      "epoch": 0.22379429133858267,
+      "grad_norm": 1.1738662719726562,
+      "learning_rate": 0.0001763830144014948,
+      "loss": 3.1963,
+      "step": 3638
+    },
+    {
+      "epoch": 0.22385580708661418,
+      "grad_norm": 1.0411376953125,
+      "learning_rate": 0.000176370532127081,
+      "loss": 3.0628,
+      "step": 3639
+    },
+    {
+      "epoch": 0.22391732283464566,
+      "grad_norm": 1.128570795059204,
+      "learning_rate": 0.0001763580469968352,
+      "loss": 3.2806,
+      "step": 3640
+    },
+    {
+      "epoch": 0.22397883858267717,
+      "grad_norm": 1.7050232887268066,
+      "learning_rate": 0.00017634555901122423,
+      "loss": 2.7839,
+      "step": 3641
+    },
+    {
+      "epoch": 0.22404035433070865,
+      "grad_norm": 1.187578558921814,
+      "learning_rate": 0.000176333068170715,
+      "loss": 3.3408,
+      "step": 3642
+    },
+    {
+      "epoch": 0.22410187007874016,
+      "grad_norm": 1.1802122592926025,
+      "learning_rate": 0.00017632057447577475,
+      "loss": 2.7795,
+      "step": 3643
+    },
+    {
+      "epoch": 0.22416338582677164,
+      "grad_norm": 1.5711643695831299,
+      "learning_rate": 0.00017630807792687057,
+      "loss": 2.4431,
+      "step": 3644
+    },
+    {
+      "epoch": 0.22422490157480315,
+      "grad_norm": 1.1380305290222168,
+      "learning_rate": 0.00017629557852446974,
+      "loss": 2.3258,
+      "step": 3645
+    },
+    {
+      "epoch": 0.22428641732283464,
+      "grad_norm": 1.336440920829773,
+      "learning_rate": 0.00017628307626903973,
+      "loss": 3.1594,
+      "step": 3646
+    },
+    {
+      "epoch": 0.22434793307086615,
+      "grad_norm": 0.8767797946929932,
+      "learning_rate": 0.00017627057116104803,
+      "loss": 3.2454,
+      "step": 3647
+    },
+    {
+      "epoch": 0.22440944881889763,
+      "grad_norm": 1.0208879709243774,
+      "learning_rate": 0.00017625806320096228,
+      "loss": 2.973,
+      "step": 3648
+    },
+    {
+      "epoch": 0.22447096456692914,
+      "grad_norm": 1.1276984214782715,
+      "learning_rate": 0.0001762455523892502,
+      "loss": 2.2494,
+      "step": 3649
+    },
+    {
+      "epoch": 0.22453248031496062,
+      "grad_norm": 1.199828863143921,
+      "learning_rate": 0.00017623303872637958,
+      "loss": 3.0042,
+      "step": 3650
+    },
+    {
+      "epoch": 0.22459399606299213,
+      "grad_norm": 1.1619387865066528,
+      "learning_rate": 0.00017622052221281842,
+      "loss": 2.7108,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2246555118110236,
+      "grad_norm": 1.1692456007003784,
+      "learning_rate": 0.00017620800284903475,
+      "loss": 2.7941,
+      "step": 3652
+    },
+    {
+      "epoch": 0.22471702755905512,
+      "grad_norm": 1.0326683521270752,
+      "learning_rate": 0.0001761954806354967,
+      "loss": 3.2131,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2247785433070866,
+      "grad_norm": 1.042405366897583,
+      "learning_rate": 0.0001761829555726726,
+      "loss": 2.9144,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2248400590551181,
+      "grad_norm": 1.151916265487671,
+      "learning_rate": 0.0001761704276610307,
+      "loss": 3.1975,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2249015748031496,
+      "grad_norm": 1.2298741340637207,
+      "learning_rate": 0.0001761578969010396,
+      "loss": 2.8387,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2249630905511811,
+      "grad_norm": 0.9201094508171082,
+      "learning_rate": 0.00017614536329316782,
+      "loss": 2.9877,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2250246062992126,
+      "grad_norm": 0.949965238571167,
+      "learning_rate": 0.00017613282683788404,
+      "loss": 2.8957,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2250861220472441,
+      "grad_norm": 1.2644236087799072,
+      "learning_rate": 0.00017612028753565706,
+      "loss": 2.6644,
+      "step": 3659
+    },
+    {
+      "epoch": 0.22514763779527558,
+      "grad_norm": 1.0027226209640503,
+      "learning_rate": 0.00017610774538695577,
+      "loss": 3.1165,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2252091535433071,
+      "grad_norm": 1.1485493183135986,
+      "learning_rate": 0.00017609520039224924,
+      "loss": 2.8985,
+      "step": 3661
+    },
+    {
+      "epoch": 0.22527066929133857,
+      "grad_norm": 1.0704894065856934,
+      "learning_rate": 0.00017608265255200648,
+      "loss": 2.8227,
+      "step": 3662
+    },
+    {
+      "epoch": 0.22533218503937008,
+      "grad_norm": 1.011855959892273,
+      "learning_rate": 0.00017607010186669679,
+      "loss": 3.3534,
+      "step": 3663
+    },
+    {
+      "epoch": 0.22539370078740156,
+      "grad_norm": 1.3245257139205933,
+      "learning_rate": 0.00017605754833678944,
+      "loss": 2.5868,
+      "step": 3664
+    },
+    {
+      "epoch": 0.22545521653543307,
+      "grad_norm": 0.9737985730171204,
+      "learning_rate": 0.00017604499196275395,
+      "loss": 3.3953,
+      "step": 3665
+    },
+    {
+      "epoch": 0.22551673228346455,
+      "grad_norm": 1.6033354997634888,
+      "learning_rate": 0.00017603243274505977,
+      "loss": 2.8311,
+      "step": 3666
+    },
+    {
+      "epoch": 0.22557824803149606,
+      "grad_norm": 1.2319281101226807,
+      "learning_rate": 0.00017601987068417655,
+      "loss": 2.4443,
+      "step": 3667
+    },
+    {
+      "epoch": 0.22563976377952755,
+      "grad_norm": 1.1386133432388306,
+      "learning_rate": 0.00017600730578057406,
+      "loss": 2.7005,
+      "step": 3668
+    },
+    {
+      "epoch": 0.22570127952755906,
+      "grad_norm": 1.0562220811843872,
+      "learning_rate": 0.00017599473803472219,
+      "loss": 2.6689,
+      "step": 3669
+    },
+    {
+      "epoch": 0.22576279527559054,
+      "grad_norm": 1.1012005805969238,
+      "learning_rate": 0.00017598216744709087,
+      "loss": 3.0262,
+      "step": 3670
+    },
+    {
+      "epoch": 0.22582431102362205,
+      "grad_norm": 1.1715999841690063,
+      "learning_rate": 0.00017596959401815017,
+      "loss": 2.8593,
+      "step": 3671
+    },
+    {
+      "epoch": 0.22588582677165353,
+      "grad_norm": 1.3767801523208618,
+      "learning_rate": 0.0001759570177483703,
+      "loss": 2.7913,
+      "step": 3672
+    },
+    {
+      "epoch": 0.22594734251968504,
+      "grad_norm": 1.0412696599960327,
+      "learning_rate": 0.00017594443863822148,
+      "loss": 1.9473,
+      "step": 3673
+    },
+    {
+      "epoch": 0.22600885826771652,
+      "grad_norm": 1.0217747688293457,
+      "learning_rate": 0.00017593185668817417,
+      "loss": 3.0315,
+      "step": 3674
+    },
+    {
+      "epoch": 0.22607037401574803,
+      "grad_norm": 1.4994243383407593,
+      "learning_rate": 0.00017591927189869878,
+      "loss": 2.6946,
+      "step": 3675
+    },
+    {
+      "epoch": 0.22613188976377951,
+      "grad_norm": 1.2026333808898926,
+      "learning_rate": 0.00017590668427026597,
+      "loss": 2.737,
+      "step": 3676
+    },
+    {
+      "epoch": 0.22619340551181102,
+      "grad_norm": 1.0400476455688477,
+      "learning_rate": 0.00017589409380334644,
+      "loss": 2.8859,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2262549212598425,
+      "grad_norm": 1.3049904108047485,
+      "learning_rate": 0.00017588150049841105,
+      "loss": 2.7335,
+      "step": 3678
+    },
+    {
+      "epoch": 0.22631643700787402,
+      "grad_norm": 1.6036887168884277,
+      "learning_rate": 0.0001758689043559306,
+      "loss": 2.4113,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2263779527559055,
+      "grad_norm": 1.1859341859817505,
+      "learning_rate": 0.0001758563053763762,
+      "loss": 1.5738,
+      "step": 3680
+    },
+    {
+      "epoch": 0.226439468503937,
+      "grad_norm": 1.183452844619751,
+      "learning_rate": 0.00017584370356021894,
+      "loss": 2.6933,
+      "step": 3681
+    },
+    {
+      "epoch": 0.2265009842519685,
+      "grad_norm": 1.0901578664779663,
+      "learning_rate": 0.00017583109890793015,
+      "loss": 2.549,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2265625,
+      "grad_norm": 1.1938402652740479,
+      "learning_rate": 0.00017581849141998105,
+      "loss": 3.2845,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2266240157480315,
+      "grad_norm": 1.2321113348007202,
+      "learning_rate": 0.00017580588109684318,
+      "loss": 2.6008,
+      "step": 3684
+    },
+    {
+      "epoch": 0.226685531496063,
+      "grad_norm": 1.2794272899627686,
+      "learning_rate": 0.00017579326793898804,
+      "loss": 3.2422,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2267470472440945,
+      "grad_norm": 1.3434908390045166,
+      "learning_rate": 0.00017578065194688732,
+      "loss": 2.6831,
+      "step": 3686
+    },
+    {
+      "epoch": 0.22680856299212598,
+      "grad_norm": 1.0106337070465088,
+      "learning_rate": 0.0001757680331210128,
+      "loss": 3.3216,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2268700787401575,
+      "grad_norm": 1.1304069757461548,
+      "learning_rate": 0.00017575541146183632,
+      "loss": 2.0226,
+      "step": 3688
+    },
+    {
+      "epoch": 0.22693159448818898,
+      "grad_norm": 1.0878394842147827,
+      "learning_rate": 0.00017574278696982988,
+      "loss": 2.8568,
+      "step": 3689
+    },
+    {
+      "epoch": 0.22699311023622049,
+      "grad_norm": 1.156894564628601,
+      "learning_rate": 0.00017573015964546555,
+      "loss": 2.73,
+      "step": 3690
+    },
+    {
+      "epoch": 0.22705462598425197,
+      "grad_norm": 1.1808675527572632,
+      "learning_rate": 0.00017571752948921553,
+      "loss": 2.9328,
+      "step": 3691
+    },
+    {
+      "epoch": 0.22711614173228348,
+      "grad_norm": 1.1969925165176392,
+      "learning_rate": 0.00017570489650155214,
+      "loss": 2.9761,
+      "step": 3692
+    },
+    {
+      "epoch": 0.22717765748031496,
+      "grad_norm": 1.4279606342315674,
+      "learning_rate": 0.00017569226068294772,
+      "loss": 2.9156,
+      "step": 3693
+    },
+    {
+      "epoch": 0.22723917322834647,
+      "grad_norm": 1.4941632747650146,
+      "learning_rate": 0.00017567962203387486,
+      "loss": 2.2871,
+      "step": 3694
+    },
+    {
+      "epoch": 0.22730068897637795,
+      "grad_norm": 1.067026972770691,
+      "learning_rate": 0.00017566698055480616,
+      "loss": 2.8285,
+      "step": 3695
+    },
+    {
+      "epoch": 0.22736220472440946,
+      "grad_norm": 1.3120826482772827,
+      "learning_rate": 0.00017565433624621427,
+      "loss": 2.446,
+      "step": 3696
+    },
+    {
+      "epoch": 0.22742372047244094,
+      "grad_norm": 1.2473511695861816,
+      "learning_rate": 0.0001756416891085721,
+      "loss": 2.8983,
+      "step": 3697
+    },
+    {
+      "epoch": 0.22748523622047245,
+      "grad_norm": 1.2621344327926636,
+      "learning_rate": 0.0001756290391423525,
+      "loss": 2.9882,
+      "step": 3698
+    },
+    {
+      "epoch": 0.22754675196850394,
+      "grad_norm": 1.1293426752090454,
+      "learning_rate": 0.0001756163863480286,
+      "loss": 2.8477,
+      "step": 3699
+    },
+    {
+      "epoch": 0.22760826771653545,
+      "grad_norm": 1.0445326566696167,
+      "learning_rate": 0.0001756037307260735,
+      "loss": 3.2392,
+      "step": 3700
+    },
+    {
+      "epoch": 0.22766978346456693,
+      "grad_norm": 1.09107506275177,
+      "learning_rate": 0.00017559107227696044,
+      "loss": 2.5936,
+      "step": 3701
+    },
+    {
+      "epoch": 0.22773129921259844,
+      "grad_norm": 1.1414376497268677,
+      "learning_rate": 0.00017557841100116276,
+      "loss": 2.8414,
+      "step": 3702
+    },
+    {
+      "epoch": 0.22779281496062992,
+      "grad_norm": 1.2529025077819824,
+      "learning_rate": 0.00017556574689915397,
+      "loss": 2.655,
+      "step": 3703
+    },
+    {
+      "epoch": 0.22785433070866143,
+      "grad_norm": 1.066414713859558,
+      "learning_rate": 0.0001755530799714076,
+      "loss": 3.1948,
+      "step": 3704
+    },
+    {
+      "epoch": 0.2279158464566929,
+      "grad_norm": 0.9215474128723145,
+      "learning_rate": 0.00017554041021839733,
+      "loss": 2.6344,
+      "step": 3705
+    },
+    {
+      "epoch": 0.22797736220472442,
+      "grad_norm": 1.3794848918914795,
+      "learning_rate": 0.00017552773764059696,
+      "loss": 2.2302,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2280388779527559,
+      "grad_norm": 1.1183263063430786,
+      "learning_rate": 0.00017551506223848032,
+      "loss": 1.9117,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2281003937007874,
+      "grad_norm": 1.2618422508239746,
+      "learning_rate": 0.00017550238401252148,
+      "loss": 2.4655,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2281619094488189,
+      "grad_norm": 1.3443912267684937,
+      "learning_rate": 0.00017548970296319446,
+      "loss": 2.6363,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2282234251968504,
+      "grad_norm": 1.0671381950378418,
+      "learning_rate": 0.00017547701909097354,
+      "loss": 2.6954,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2282849409448819,
+      "grad_norm": 1.037082314491272,
+      "learning_rate": 0.00017546433239633294,
+      "loss": 3.2835,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2283464566929134,
+      "grad_norm": 1.0716443061828613,
+      "learning_rate": 0.0001754516428797471,
+      "loss": 2.6875,
+      "step": 3712
+    },
+    {
+      "epoch": 0.22840797244094488,
+      "grad_norm": 1.091546893119812,
+      "learning_rate": 0.00017543895054169056,
+      "loss": 3.0892,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2284694881889764,
+      "grad_norm": 1.1520270109176636,
+      "learning_rate": 0.00017542625538263793,
+      "loss": 2.6977,
+      "step": 3714
+    },
+    {
+      "epoch": 0.22853100393700787,
+      "grad_norm": 1.3160027265548706,
+      "learning_rate": 0.00017541355740306394,
+      "loss": 2.7912,
+      "step": 3715
+    },
+    {
+      "epoch": 0.22859251968503938,
+      "grad_norm": 1.4559348821640015,
+      "learning_rate": 0.00017540085660344343,
+      "loss": 2.5114,
+      "step": 3716
+    },
+    {
+      "epoch": 0.22865403543307086,
+      "grad_norm": 1.3259963989257812,
+      "learning_rate": 0.00017538815298425128,
+      "loss": 2.03,
+      "step": 3717
+    },
+    {
+      "epoch": 0.22871555118110237,
+      "grad_norm": 1.1183807849884033,
+      "learning_rate": 0.00017537544654596264,
+      "loss": 2.6118,
+      "step": 3718
+    },
+    {
+      "epoch": 0.22877706692913385,
+      "grad_norm": 1.1659603118896484,
+      "learning_rate": 0.0001753627372890526,
+      "loss": 2.8927,
+      "step": 3719
+    },
+    {
+      "epoch": 0.22883858267716536,
+      "grad_norm": 1.1039680242538452,
+      "learning_rate": 0.0001753500252139964,
+      "loss": 3.2034,
+      "step": 3720
+    },
+    {
+      "epoch": 0.22890009842519685,
+      "grad_norm": 1.2592283487319946,
+      "learning_rate": 0.0001753373103212694,
+      "loss": 3.7559,
+      "step": 3721
+    },
+    {
+      "epoch": 0.22896161417322836,
+      "grad_norm": 1.1704174280166626,
+      "learning_rate": 0.0001753245926113471,
+      "loss": 3.1842,
+      "step": 3722
+    },
+    {
+      "epoch": 0.22902312992125984,
+      "grad_norm": 1.1396660804748535,
+      "learning_rate": 0.00017531187208470508,
+      "loss": 2.8898,
+      "step": 3723
+    },
+    {
+      "epoch": 0.22908464566929135,
+      "grad_norm": 0.9814751744270325,
+      "learning_rate": 0.00017529914874181898,
+      "loss": 3.2531,
+      "step": 3724
+    },
+    {
+      "epoch": 0.22914616141732283,
+      "grad_norm": 1.3537319898605347,
+      "learning_rate": 0.00017528642258316462,
+      "loss": 2.9696,
+      "step": 3725
+    },
+    {
+      "epoch": 0.22920767716535434,
+      "grad_norm": 1.1924139261245728,
+      "learning_rate": 0.00017527369360921782,
+      "loss": 2.299,
+      "step": 3726
+    },
+    {
+      "epoch": 0.22926919291338582,
+      "grad_norm": 1.317351222038269,
+      "learning_rate": 0.00017526096182045463,
+      "loss": 2.8631,
+      "step": 3727
+    },
+    {
+      "epoch": 0.22933070866141733,
+      "grad_norm": 1.1944226026535034,
+      "learning_rate": 0.00017524822721735114,
+      "loss": 3.2731,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2293922244094488,
+      "grad_norm": 1.2070667743682861,
+      "learning_rate": 0.00017523548980038359,
+      "loss": 2.7539,
+      "step": 3729
+    },
+    {
+      "epoch": 0.22945374015748032,
+      "grad_norm": 0.9627958536148071,
+      "learning_rate": 0.00017522274957002817,
+      "loss": 2.7097,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2295152559055118,
+      "grad_norm": 1.2237780094146729,
+      "learning_rate": 0.00017521000652676142,
+      "loss": 3.03,
+      "step": 3731
+    },
+    {
+      "epoch": 0.22957677165354332,
+      "grad_norm": 1.1353424787521362,
+      "learning_rate": 0.0001751972606710598,
+      "loss": 3.1628,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2296382874015748,
+      "grad_norm": 0.810170590877533,
+      "learning_rate": 0.00017518451200339993,
+      "loss": 3.3805,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2296998031496063,
+      "grad_norm": 1.0843451023101807,
+      "learning_rate": 0.00017517176052425857,
+      "loss": 2.8072,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2297613188976378,
+      "grad_norm": 0.956553041934967,
+      "learning_rate": 0.00017515900623411252,
+      "loss": 2.8872,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2298228346456693,
+      "grad_norm": 1.1895673274993896,
+      "learning_rate": 0.00017514624913343875,
+      "loss": 3.059,
+      "step": 3736
+    },
+    {
+      "epoch": 0.22988435039370078,
+      "grad_norm": 0.9109532833099365,
+      "learning_rate": 0.0001751334892227143,
+      "loss": 2.3856,
+      "step": 3737
+    },
+    {
+      "epoch": 0.2299458661417323,
+      "grad_norm": 1.2860714197158813,
+      "learning_rate": 0.00017512072650241627,
+      "loss": 2.6125,
+      "step": 3738
+    },
+    {
+      "epoch": 0.23000738188976377,
+      "grad_norm": 1.327744960784912,
+      "learning_rate": 0.00017510796097302194,
+      "loss": 2.7513,
+      "step": 3739
+    },
+    {
+      "epoch": 0.23006889763779528,
+      "grad_norm": 1.4397276639938354,
+      "learning_rate": 0.00017509519263500872,
+      "loss": 2.8275,
+      "step": 3740
+    },
+    {
+      "epoch": 0.23013041338582677,
+      "grad_norm": 1.2907383441925049,
+      "learning_rate": 0.00017508242148885406,
+      "loss": 2.4498,
+      "step": 3741
+    },
+    {
+      "epoch": 0.23019192913385828,
+      "grad_norm": 1.0868699550628662,
+      "learning_rate": 0.00017506964753503547,
+      "loss": 2.2843,
+      "step": 3742
+    },
+    {
+      "epoch": 0.23025344488188976,
+      "grad_norm": 1.2957431077957153,
+      "learning_rate": 0.00017505687077403067,
+      "loss": 2.591,
+      "step": 3743
+    },
+    {
+      "epoch": 0.23031496062992127,
+      "grad_norm": 1.1553418636322021,
+      "learning_rate": 0.00017504409120631741,
+      "loss": 2.8762,
+      "step": 3744
+    },
+    {
+      "epoch": 0.23037647637795275,
+      "grad_norm": 0.991338849067688,
+      "learning_rate": 0.00017503130883237365,
+      "loss": 2.7298,
+      "step": 3745
+    },
+    {
+      "epoch": 0.23043799212598426,
+      "grad_norm": 1.0265297889709473,
+      "learning_rate": 0.00017501852365267727,
+      "loss": 2.6056,
+      "step": 3746
+    },
+    {
+      "epoch": 0.23049950787401574,
+      "grad_norm": 1.1538071632385254,
+      "learning_rate": 0.00017500573566770646,
+      "loss": 2.7841,
+      "step": 3747
+    },
+    {
+      "epoch": 0.23056102362204725,
+      "grad_norm": 1.0075684785842896,
+      "learning_rate": 0.00017499294487793935,
+      "loss": 3.4946,
+      "step": 3748
+    },
+    {
+      "epoch": 0.23062253937007873,
+      "grad_norm": 1.1396087408065796,
+      "learning_rate": 0.0001749801512838543,
+      "loss": 3.3438,
+      "step": 3749
+    },
+    {
+      "epoch": 0.23068405511811024,
+      "grad_norm": 1.1778677701950073,
+      "learning_rate": 0.0001749673548859297,
+      "loss": 3.1174,
+      "step": 3750
+    },
+    {
+      "epoch": 0.23074557086614172,
+      "grad_norm": 1.2535191774368286,
+      "learning_rate": 0.00017495455568464403,
+      "loss": 2.782,
+      "step": 3751
+    },
+    {
+      "epoch": 0.23080708661417323,
+      "grad_norm": 1.1459214687347412,
+      "learning_rate": 0.00017494175368047593,
+      "loss": 2.6288,
+      "step": 3752
+    },
+    {
+      "epoch": 0.23086860236220472,
+      "grad_norm": 1.3283677101135254,
+      "learning_rate": 0.00017492894887390415,
+      "loss": 2.853,
+      "step": 3753
+    },
+    {
+      "epoch": 0.23093011811023623,
+      "grad_norm": 1.2726435661315918,
+      "learning_rate": 0.0001749161412654075,
+      "loss": 2.9713,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2309916338582677,
+      "grad_norm": 1.1838185787200928,
+      "learning_rate": 0.0001749033308554649,
+      "loss": 2.9957,
+      "step": 3755
+    },
+    {
+      "epoch": 0.23105314960629922,
+      "grad_norm": 1.2085423469543457,
+      "learning_rate": 0.0001748905176445554,
+      "loss": 2.9104,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2311146653543307,
+      "grad_norm": 1.3370356559753418,
+      "learning_rate": 0.00017487770163315813,
+      "loss": 2.1587,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2311761811023622,
+      "grad_norm": 0.9232663512229919,
+      "learning_rate": 0.00017486488282175239,
+      "loss": 2.8693,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2312376968503937,
+      "grad_norm": 1.3708996772766113,
+      "learning_rate": 0.00017485206121081742,
+      "loss": 1.9994,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2312992125984252,
+      "grad_norm": 2.0268423557281494,
+      "learning_rate": 0.00017483923680083278,
+      "loss": 2.5413,
+      "step": 3760
+    },
+    {
+      "epoch": 0.23136072834645668,
+      "grad_norm": 1.2214792966842651,
+      "learning_rate": 0.00017482640959227803,
+      "loss": 2.7398,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2314222440944882,
+      "grad_norm": 1.1904405355453491,
+      "learning_rate": 0.00017481357958563275,
+      "loss": 2.5413,
+      "step": 3762
+    },
+    {
+      "epoch": 0.23148375984251968,
+      "grad_norm": 1.6216009855270386,
+      "learning_rate": 0.0001748007467813768,
+      "loss": 2.9419,
+      "step": 3763
+    },
+    {
+      "epoch": 0.2315452755905512,
+      "grad_norm": 1.6152395009994507,
+      "learning_rate": 0.00017478791117999003,
+      "loss": 2.8294,
+      "step": 3764
+    },
+    {
+      "epoch": 0.23160679133858267,
+      "grad_norm": 1.2298498153686523,
+      "learning_rate": 0.00017477507278195237,
+      "loss": 2.8284,
+      "step": 3765
+    },
+    {
+      "epoch": 0.23166830708661418,
+      "grad_norm": 1.3657490015029907,
+      "learning_rate": 0.000174762231587744,
+      "loss": 2.7979,
+      "step": 3766
+    },
+    {
+      "epoch": 0.23172982283464566,
+      "grad_norm": 1.1195157766342163,
+      "learning_rate": 0.00017474938759784502,
+      "loss": 2.2544,
+      "step": 3767
+    },
+    {
+      "epoch": 0.23179133858267717,
+      "grad_norm": 1.3707011938095093,
+      "learning_rate": 0.00017473654081273576,
+      "loss": 3.176,
+      "step": 3768
+    },
+    {
+      "epoch": 0.23185285433070865,
+      "grad_norm": 1.2653429508209229,
+      "learning_rate": 0.00017472369123289663,
+      "loss": 3.4962,
+      "step": 3769
+    },
+    {
+      "epoch": 0.23191437007874016,
+      "grad_norm": 1.2117109298706055,
+      "learning_rate": 0.0001747108388588081,
+      "loss": 2.9967,
+      "step": 3770
+    },
+    {
+      "epoch": 0.23197588582677164,
+      "grad_norm": 1.107993245124817,
+      "learning_rate": 0.0001746979836909508,
+      "loss": 2.3225,
+      "step": 3771
+    },
+    {
+      "epoch": 0.23203740157480315,
+      "grad_norm": 1.1096861362457275,
+      "learning_rate": 0.00017468512572980544,
+      "loss": 2.8404,
+      "step": 3772
+    },
+    {
+      "epoch": 0.23209891732283464,
+      "grad_norm": 1.1601274013519287,
+      "learning_rate": 0.00017467226497585285,
+      "loss": 2.938,
+      "step": 3773
+    },
+    {
+      "epoch": 0.23216043307086615,
+      "grad_norm": 1.0041624307632446,
+      "learning_rate": 0.0001746594014295739,
+      "loss": 2.9412,
+      "step": 3774
+    },
+    {
+      "epoch": 0.23222194881889763,
+      "grad_norm": 1.1234991550445557,
+      "learning_rate": 0.00017464653509144967,
+      "loss": 2.8784,
+      "step": 3775
+    },
+    {
+      "epoch": 0.23228346456692914,
+      "grad_norm": 1.13557767868042,
+      "learning_rate": 0.00017463366596196127,
+      "loss": 2.8467,
+      "step": 3776
+    },
+    {
+      "epoch": 0.23234498031496062,
+      "grad_norm": 0.9366655349731445,
+      "learning_rate": 0.00017462079404158994,
+      "loss": 2.9236,
+      "step": 3777
+    },
+    {
+      "epoch": 0.23240649606299213,
+      "grad_norm": 1.0853173732757568,
+      "learning_rate": 0.000174607919330817,
+      "loss": 3.1166,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2324680118110236,
+      "grad_norm": 1.2606000900268555,
+      "learning_rate": 0.00017459504183012388,
+      "loss": 3.7738,
+      "step": 3779
+    },
+    {
+      "epoch": 0.23252952755905512,
+      "grad_norm": 1.2832881212234497,
+      "learning_rate": 0.0001745821615399922,
+      "loss": 3.3779,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2325910433070866,
+      "grad_norm": 0.9062511324882507,
+      "learning_rate": 0.00017456927846090352,
+      "loss": 3.2293,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2326525590551181,
+      "grad_norm": 1.0660808086395264,
+      "learning_rate": 0.00017455639259333965,
+      "loss": 3.0381,
+      "step": 3782
+    },
+    {
+      "epoch": 0.2327140748031496,
+      "grad_norm": 1.2825214862823486,
+      "learning_rate": 0.0001745435039377824,
+      "loss": 3.163,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2327755905511811,
+      "grad_norm": 1.1964495182037354,
+      "learning_rate": 0.0001745306124947138,
+      "loss": 3.2127,
+      "step": 3784
+    },
+    {
+      "epoch": 0.2328371062992126,
+      "grad_norm": 1.4827734231948853,
+      "learning_rate": 0.0001745177182646159,
+      "loss": 2.216,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2328986220472441,
+      "grad_norm": 1.005636215209961,
+      "learning_rate": 0.00017450482124797083,
+      "loss": 2.9762,
+      "step": 3786
+    },
+    {
+      "epoch": 0.23296013779527558,
+      "grad_norm": 1.225252628326416,
+      "learning_rate": 0.0001744919214452609,
+      "loss": 2.9279,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2330216535433071,
+      "grad_norm": 1.259799599647522,
+      "learning_rate": 0.0001744790188569685,
+      "loss": 2.4794,
+      "step": 3788
+    },
+    {
+      "epoch": 0.23308316929133857,
+      "grad_norm": 1.2545338869094849,
+      "learning_rate": 0.0001744661134835761,
+      "loss": 2.2247,
+      "step": 3789
+    },
+    {
+      "epoch": 0.23314468503937008,
+      "grad_norm": 1.0598258972167969,
+      "learning_rate": 0.00017445320532556631,
+      "loss": 3.6491,
+      "step": 3790
+    },
+    {
+      "epoch": 0.23320620078740156,
+      "grad_norm": 0.9520148634910583,
+      "learning_rate": 0.00017444029438342176,
+      "loss": 3.2349,
+      "step": 3791
+    },
+    {
+      "epoch": 0.23326771653543307,
+      "grad_norm": 1.1631144285202026,
+      "learning_rate": 0.00017442738065762532,
+      "loss": 2.5538,
+      "step": 3792
+    },
+    {
+      "epoch": 0.23332923228346455,
+      "grad_norm": 0.9799225926399231,
+      "learning_rate": 0.00017441446414865984,
+      "loss": 2.9092,
+      "step": 3793
+    },
+    {
+      "epoch": 0.23339074803149606,
+      "grad_norm": 1.1735042333602905,
+      "learning_rate": 0.00017440154485700836,
+      "loss": 2.4265,
+      "step": 3794
+    },
+    {
+      "epoch": 0.23345226377952755,
+      "grad_norm": 1.1168063879013062,
+      "learning_rate": 0.00017438862278315397,
+      "loss": 2.7924,
+      "step": 3795
+    },
+    {
+      "epoch": 0.23351377952755906,
+      "grad_norm": 1.0778769254684448,
+      "learning_rate": 0.0001743756979275799,
+      "loss": 3.1028,
+      "step": 3796
+    },
+    {
+      "epoch": 0.23357529527559054,
+      "grad_norm": 1.1338492631912231,
+      "learning_rate": 0.00017436277029076944,
+      "loss": 3.0434,
+      "step": 3797
+    },
+    {
+      "epoch": 0.23363681102362205,
+      "grad_norm": 0.975369930267334,
+      "learning_rate": 0.00017434983987320605,
+      "loss": 3.004,
+      "step": 3798
+    },
+    {
+      "epoch": 0.23369832677165353,
+      "grad_norm": 1.1171035766601562,
+      "learning_rate": 0.00017433690667537324,
+      "loss": 2.9072,
+      "step": 3799
+    },
+    {
+      "epoch": 0.23375984251968504,
+      "grad_norm": 1.1739280223846436,
+      "learning_rate": 0.00017432397069775462,
+      "loss": 2.983,
+      "step": 3800
+    },
+    {
+      "epoch": 0.23382135826771652,
+      "grad_norm": 1.12468421459198,
+      "learning_rate": 0.00017431103194083395,
+      "loss": 2.5478,
+      "step": 3801
+    },
+    {
+      "epoch": 0.23388287401574803,
+      "grad_norm": 1.2627280950546265,
+      "learning_rate": 0.00017429809040509504,
+      "loss": 2.8651,
+      "step": 3802
+    },
+    {
+      "epoch": 0.23394438976377951,
+      "grad_norm": 0.9702225923538208,
+      "learning_rate": 0.00017428514609102184,
+      "loss": 3.5127,
+      "step": 3803
+    },
+    {
+      "epoch": 0.23400590551181102,
+      "grad_norm": 1.1090306043624878,
+      "learning_rate": 0.00017427219899909842,
+      "loss": 2.6016,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2340674212598425,
+      "grad_norm": 1.0322195291519165,
+      "learning_rate": 0.00017425924912980892,
+      "loss": 3.0573,
+      "step": 3805
+    },
+    {
+      "epoch": 0.23412893700787402,
+      "grad_norm": 1.055976390838623,
+      "learning_rate": 0.00017424629648363755,
+      "loss": 2.5978,
+      "step": 3806
+    },
+    {
+      "epoch": 0.2341904527559055,
+      "grad_norm": 1.0089236497879028,
+      "learning_rate": 0.00017423334106106872,
+      "loss": 2.5146,
+      "step": 3807
+    },
+    {
+      "epoch": 0.234251968503937,
+      "grad_norm": 1.089057445526123,
+      "learning_rate": 0.00017422038286258687,
+      "loss": 2.8009,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2343134842519685,
+      "grad_norm": 1.0490245819091797,
+      "learning_rate": 0.00017420742188867656,
+      "loss": 2.9392,
+      "step": 3809
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 1.164972186088562,
+      "learning_rate": 0.00017419445813982246,
+      "loss": 2.9869,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2344365157480315,
+      "grad_norm": 1.4297113418579102,
+      "learning_rate": 0.00017418149161650933,
+      "loss": 2.2598,
+      "step": 3811
+    },
+    {
+      "epoch": 0.234498031496063,
+      "grad_norm": 1.112574815750122,
+      "learning_rate": 0.00017416852231922212,
+      "loss": 2.5175,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2345595472440945,
+      "grad_norm": 1.1398168802261353,
+      "learning_rate": 0.0001741555502484457,
+      "loss": 2.8348,
+      "step": 3813
+    },
+    {
+      "epoch": 0.23462106299212598,
+      "grad_norm": 1.489856243133545,
+      "learning_rate": 0.0001741425754046652,
+      "loss": 2.9201,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2346825787401575,
+      "grad_norm": 1.1739065647125244,
+      "learning_rate": 0.00017412959778836585,
+      "loss": 3.2669,
+      "step": 3815
+    },
+    {
+      "epoch": 0.23474409448818898,
+      "grad_norm": 1.0695561170578003,
+      "learning_rate": 0.00017411661740003286,
+      "loss": 3.0884,
+      "step": 3816
+    },
+    {
+      "epoch": 0.23480561023622049,
+      "grad_norm": 1.3243050575256348,
+      "learning_rate": 0.0001741036342401517,
+      "loss": 2.7113,
+      "step": 3817
+    },
+    {
+      "epoch": 0.23486712598425197,
+      "grad_norm": 1.3018125295639038,
+      "learning_rate": 0.0001740906483092078,
+      "loss": 2.6289,
+      "step": 3818
+    },
+    {
+      "epoch": 0.23492864173228348,
+      "grad_norm": 1.1006697416305542,
+      "learning_rate": 0.00017407765960768682,
+      "loss": 2.8368,
+      "step": 3819
+    },
+    {
+      "epoch": 0.23499015748031496,
+      "grad_norm": 1.0473068952560425,
+      "learning_rate": 0.00017406466813607444,
+      "loss": 2.9398,
+      "step": 3820
+    },
+    {
+      "epoch": 0.23505167322834647,
+      "grad_norm": 1.1519701480865479,
+      "learning_rate": 0.00017405167389485644,
+      "loss": 2.5606,
+      "step": 3821
+    },
+    {
+      "epoch": 0.23511318897637795,
+      "grad_norm": 3.245788097381592,
+      "learning_rate": 0.00017403867688451876,
+      "loss": 2.6538,
+      "step": 3822
+    },
+    {
+      "epoch": 0.23517470472440946,
+      "grad_norm": 1.1662770509719849,
+      "learning_rate": 0.00017402567710554745,
+      "loss": 2.9425,
+      "step": 3823
+    },
+    {
+      "epoch": 0.23523622047244094,
+      "grad_norm": 1.5857065916061401,
+      "learning_rate": 0.00017401267455842857,
+      "loss": 3.276,
+      "step": 3824
+    },
+    {
+      "epoch": 0.23529773622047245,
+      "grad_norm": 1.0846441984176636,
+      "learning_rate": 0.00017399966924364836,
+      "loss": 3.5085,
+      "step": 3825
+    },
+    {
+      "epoch": 0.23535925196850394,
+      "grad_norm": 1.150951862335205,
+      "learning_rate": 0.00017398666116169318,
+      "loss": 2.7221,
+      "step": 3826
+    },
+    {
+      "epoch": 0.23542076771653545,
+      "grad_norm": 1.2920191287994385,
+      "learning_rate": 0.0001739736503130494,
+      "loss": 2.2656,
+      "step": 3827
+    },
+    {
+      "epoch": 0.23548228346456693,
+      "grad_norm": 1.03311288356781,
+      "learning_rate": 0.00017396063669820362,
+      "loss": 2.7767,
+      "step": 3828
+    },
+    {
+      "epoch": 0.23554379921259844,
+      "grad_norm": 1.1635220050811768,
+      "learning_rate": 0.00017394762031764245,
+      "loss": 3.006,
+      "step": 3829
+    },
+    {
+      "epoch": 0.23560531496062992,
+      "grad_norm": 1.190101981163025,
+      "learning_rate": 0.00017393460117185258,
+      "loss": 2.9042,
+      "step": 3830
+    },
+    {
+      "epoch": 0.23566683070866143,
+      "grad_norm": 1.8367972373962402,
+      "learning_rate": 0.00017392157926132093,
+      "loss": 2.4121,
+      "step": 3831
+    },
+    {
+      "epoch": 0.2357283464566929,
+      "grad_norm": 1.0097359418869019,
+      "learning_rate": 0.0001739085545865344,
+      "loss": 2.9193,
+      "step": 3832
+    },
+    {
+      "epoch": 0.23578986220472442,
+      "grad_norm": 1.0181254148483276,
+      "learning_rate": 0.00017389552714798008,
+      "loss": 2.4067,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2358513779527559,
+      "grad_norm": 0.9553365111351013,
+      "learning_rate": 0.00017388249694614505,
+      "loss": 2.5772,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2359128937007874,
+      "grad_norm": 1.2399671077728271,
+      "learning_rate": 0.00017386946398151664,
+      "loss": 2.4939,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2359744094488189,
+      "grad_norm": 1.0828347206115723,
+      "learning_rate": 0.0001738564282545822,
+      "loss": 3.0778,
+      "step": 3836
+    },
+    {
+      "epoch": 0.2360359251968504,
+      "grad_norm": 0.9783479571342468,
+      "learning_rate": 0.00017384338976582915,
+      "loss": 3.5883,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2360974409448819,
+      "grad_norm": 1.319039225578308,
+      "learning_rate": 0.00017383034851574508,
+      "loss": 2.8245,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2361589566929134,
+      "grad_norm": 1.4048709869384766,
+      "learning_rate": 0.0001738173045048177,
+      "loss": 1.9127,
+      "step": 3839
+    },
+    {
+      "epoch": 0.23622047244094488,
+      "grad_norm": 1.446419358253479,
+      "learning_rate": 0.00017380425773353471,
+      "loss": 3.1895,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2362819881889764,
+      "grad_norm": 1.3408511877059937,
+      "learning_rate": 0.00017379120820238406,
+      "loss": 2.968,
+      "step": 3841
+    },
+    {
+      "epoch": 0.23634350393700787,
+      "grad_norm": 1.2285147905349731,
+      "learning_rate": 0.0001737781559118537,
+      "loss": 2.6825,
+      "step": 3842
+    },
+    {
+      "epoch": 0.23640501968503938,
+      "grad_norm": 1.507662296295166,
+      "learning_rate": 0.00017376510086243165,
+      "loss": 2.6425,
+      "step": 3843
+    },
+    {
+      "epoch": 0.23646653543307086,
+      "grad_norm": 1.4916882514953613,
+      "learning_rate": 0.00017375204305460618,
+      "loss": 3.1482,
+      "step": 3844
+    },
+    {
+      "epoch": 0.23652805118110237,
+      "grad_norm": 1.2314701080322266,
+      "learning_rate": 0.0001737389824888656,
+      "loss": 3.0269,
+      "step": 3845
+    },
+    {
+      "epoch": 0.23658956692913385,
+      "grad_norm": 1.053454041481018,
+      "learning_rate": 0.0001737259191656982,
+      "loss": 2.2989,
+      "step": 3846
+    },
+    {
+      "epoch": 0.23665108267716536,
+      "grad_norm": 1.1313753128051758,
+      "learning_rate": 0.00017371285308559256,
+      "loss": 2.3592,
+      "step": 3847
+    },
+    {
+      "epoch": 0.23671259842519685,
+      "grad_norm": 1.104814887046814,
+      "learning_rate": 0.00017369978424903724,
+      "loss": 2.8522,
+      "step": 3848
+    },
+    {
+      "epoch": 0.23677411417322836,
+      "grad_norm": 0.9668189883232117,
+      "learning_rate": 0.00017368671265652092,
+      "loss": 2.9893,
+      "step": 3849
+    },
+    {
+      "epoch": 0.23683562992125984,
+      "grad_norm": 1.0854421854019165,
+      "learning_rate": 0.00017367363830853247,
+      "loss": 2.836,
+      "step": 3850
+    },
+    {
+      "epoch": 0.23689714566929135,
+      "grad_norm": 0.9753137230873108,
+      "learning_rate": 0.00017366056120556077,
+      "loss": 3.1315,
+      "step": 3851
+    },
+    {
+      "epoch": 0.23695866141732283,
+      "grad_norm": 1.3163528442382812,
+      "learning_rate": 0.0001736474813480948,
+      "loss": 2.1612,
+      "step": 3852
+    },
+    {
+      "epoch": 0.23702017716535434,
+      "grad_norm": 0.9655376076698303,
+      "learning_rate": 0.0001736343987366237,
+      "loss": 3.0419,
+      "step": 3853
+    },
+    {
+      "epoch": 0.23708169291338582,
+      "grad_norm": 1.1722867488861084,
+      "learning_rate": 0.00017362131337163666,
+      "loss": 2.0622,
+      "step": 3854
+    },
+    {
+      "epoch": 0.23714320866141733,
+      "grad_norm": 1.0025014877319336,
+      "learning_rate": 0.00017360822525362308,
+      "loss": 2.8856,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2372047244094488,
+      "grad_norm": 1.1465651988983154,
+      "learning_rate": 0.0001735951343830723,
+      "loss": 2.7915,
+      "step": 3856
+    },
+    {
+      "epoch": 0.23726624015748032,
+      "grad_norm": 1.0640383958816528,
+      "learning_rate": 0.00017358204076047386,
+      "loss": 3.1604,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2373277559055118,
+      "grad_norm": 1.170831561088562,
+      "learning_rate": 0.00017356894438631741,
+      "loss": 3.4101,
+      "step": 3858
+    },
+    {
+      "epoch": 0.23738927165354332,
+      "grad_norm": 1.3083736896514893,
+      "learning_rate": 0.0001735558452610927,
+      "loss": 2.9896,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2374507874015748,
+      "grad_norm": 1.206196665763855,
+      "learning_rate": 0.00017354274338528947,
+      "loss": 2.1499,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2375123031496063,
+      "grad_norm": 1.5404391288757324,
+      "learning_rate": 0.0001735296387593978,
+      "loss": 2.8423,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2375738188976378,
+      "grad_norm": 1.40328848361969,
+      "learning_rate": 0.00017351653138390758,
+      "loss": 2.4654,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2376353346456693,
+      "grad_norm": 1.0830888748168945,
+      "learning_rate": 0.00017350342125930905,
+      "loss": 3.221,
+      "step": 3863
+    },
+    {
+      "epoch": 0.23769685039370078,
+      "grad_norm": 1.372516393661499,
+      "learning_rate": 0.00017349030838609245,
+      "loss": 2.8886,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2377583661417323,
+      "grad_norm": 1.1456910371780396,
+      "learning_rate": 0.0001734771927647481,
+      "loss": 3.5551,
+      "step": 3865
+    },
+    {
+      "epoch": 0.23781988188976377,
+      "grad_norm": 0.9507797360420227,
+      "learning_rate": 0.00017346407439576648,
+      "loss": 2.7396,
+      "step": 3866
+    },
+    {
+      "epoch": 0.23788139763779528,
+      "grad_norm": 1.0963488817214966,
+      "learning_rate": 0.0001734509532796381,
+      "loss": 3.0061,
+      "step": 3867
+    },
+    {
+      "epoch": 0.23794291338582677,
+      "grad_norm": 1.0494672060012817,
+      "learning_rate": 0.00017343782941685364,
+      "loss": 3.2679,
+      "step": 3868
+    },
+    {
+      "epoch": 0.23800442913385828,
+      "grad_norm": 1.2747398614883423,
+      "learning_rate": 0.00017342470280790387,
+      "loss": 3.3128,
+      "step": 3869
+    },
+    {
+      "epoch": 0.23806594488188976,
+      "grad_norm": 0.9694467186927795,
+      "learning_rate": 0.0001734115734532796,
+      "loss": 3.051,
+      "step": 3870
+    },
+    {
+      "epoch": 0.23812746062992127,
+      "grad_norm": 1.1747195720672607,
+      "learning_rate": 0.0001733984413534719,
+      "loss": 2.3757,
+      "step": 3871
+    },
+    {
+      "epoch": 0.23818897637795275,
+      "grad_norm": 1.2206737995147705,
+      "learning_rate": 0.00017338530650897176,
+      "loss": 2.222,
+      "step": 3872
+    },
+    {
+      "epoch": 0.23825049212598426,
+      "grad_norm": 1.190680980682373,
+      "learning_rate": 0.00017337216892027036,
+      "loss": 2.9398,
+      "step": 3873
+    },
+    {
+      "epoch": 0.23831200787401574,
+      "grad_norm": 1.1967698335647583,
+      "learning_rate": 0.00017335902858785896,
+      "loss": 2.9069,
+      "step": 3874
+    },
+    {
+      "epoch": 0.23837352362204725,
+      "grad_norm": 1.0439010858535767,
+      "learning_rate": 0.00017334588551222896,
+      "loss": 2.5811,
+      "step": 3875
+    },
+    {
+      "epoch": 0.23843503937007873,
+      "grad_norm": 1.0994012355804443,
+      "learning_rate": 0.00017333273969387183,
+      "loss": 2.798,
+      "step": 3876
+    },
+    {
+      "epoch": 0.23849655511811024,
+      "grad_norm": 1.195949912071228,
+      "learning_rate": 0.00017331959113327914,
+      "loss": 2.5519,
+      "step": 3877
+    },
+    {
+      "epoch": 0.23855807086614172,
+      "grad_norm": 1.2671645879745483,
+      "learning_rate": 0.0001733064398309426,
+      "loss": 2.6705,
+      "step": 3878
+    },
+    {
+      "epoch": 0.23861958661417323,
+      "grad_norm": 1.2045211791992188,
+      "learning_rate": 0.00017329328578735397,
+      "loss": 2.2359,
+      "step": 3879
+    },
+    {
+      "epoch": 0.23868110236220472,
+      "grad_norm": 1.3304388523101807,
+      "learning_rate": 0.00017328012900300516,
+      "loss": 3.3732,
+      "step": 3880
+    },
+    {
+      "epoch": 0.23874261811023623,
+      "grad_norm": 1.1989601850509644,
+      "learning_rate": 0.00017326696947838813,
+      "loss": 2.7726,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2388041338582677,
+      "grad_norm": 1.1342605352401733,
+      "learning_rate": 0.000173253807213995,
+      "loss": 3.3676,
+      "step": 3882
+    },
+    {
+      "epoch": 0.23886564960629922,
+      "grad_norm": 1.0714805126190186,
+      "learning_rate": 0.00017324064221031795,
+      "loss": 3.1474,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2389271653543307,
+      "grad_norm": 1.1899442672729492,
+      "learning_rate": 0.0001732274744678493,
+      "loss": 2.7029,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2389886811023622,
+      "grad_norm": 1.1104164123535156,
+      "learning_rate": 0.0001732143039870814,
+      "loss": 3.1564,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2390501968503937,
+      "grad_norm": 1.139034390449524,
+      "learning_rate": 0.0001732011307685068,
+      "loss": 2.8415,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2391117125984252,
+      "grad_norm": 0.9677647948265076,
+      "learning_rate": 0.0001731879548126181,
+      "loss": 3.2023,
+      "step": 3887
+    },
+    {
+      "epoch": 0.23917322834645668,
+      "grad_norm": 0.9688490033149719,
+      "learning_rate": 0.00017317477611990798,
+      "loss": 3.0824,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2392347440944882,
+      "grad_norm": 1.228772759437561,
+      "learning_rate": 0.00017316159469086927,
+      "loss": 3.56,
+      "step": 3889
+    },
+    {
+      "epoch": 0.23929625984251968,
+      "grad_norm": 1.233986258506775,
+      "learning_rate": 0.00017314841052599487,
+      "loss": 2.5321,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2393577755905512,
+      "grad_norm": 1.364211916923523,
+      "learning_rate": 0.00017313522362577782,
+      "loss": 2.7778,
+      "step": 3891
+    },
+    {
+      "epoch": 0.23941929133858267,
+      "grad_norm": 1.221645474433899,
+      "learning_rate": 0.00017312203399071123,
+      "loss": 2.9828,
+      "step": 3892
+    },
+    {
+      "epoch": 0.23948080708661418,
+      "grad_norm": 1.0963560342788696,
+      "learning_rate": 0.00017310884162128828,
+      "loss": 2.9919,
+      "step": 3893
+    },
+    {
+      "epoch": 0.23954232283464566,
+      "grad_norm": 1.2607675790786743,
+      "learning_rate": 0.00017309564651800234,
+      "loss": 3.122,
+      "step": 3894
+    },
+    {
+      "epoch": 0.23960383858267717,
+      "grad_norm": 1.1378735303878784,
+      "learning_rate": 0.0001730824486813468,
+      "loss": 2.3645,
+      "step": 3895
+    },
+    {
+      "epoch": 0.23966535433070865,
+      "grad_norm": 1.2281450033187866,
+      "learning_rate": 0.0001730692481118152,
+      "loss": 2.1803,
+      "step": 3896
+    },
+    {
+      "epoch": 0.23972687007874016,
+      "grad_norm": 0.9839072823524475,
+      "learning_rate": 0.0001730560448099012,
+      "loss": 3.1375,
+      "step": 3897
+    },
+    {
+      "epoch": 0.23978838582677164,
+      "grad_norm": 1.0918517112731934,
+      "learning_rate": 0.00017304283877609845,
+      "loss": 2.3702,
+      "step": 3898
+    },
+    {
+      "epoch": 0.23984990157480315,
+      "grad_norm": 1.1033376455307007,
+      "learning_rate": 0.00017302963001090084,
+      "loss": 2.8341,
+      "step": 3899
+    },
+    {
+      "epoch": 0.23991141732283464,
+      "grad_norm": 0.9550749659538269,
+      "learning_rate": 0.0001730164185148023,
+      "loss": 2.8286,
+      "step": 3900
+    },
+    {
+      "epoch": 0.23997293307086615,
+      "grad_norm": 1.2246708869934082,
+      "learning_rate": 0.00017300320428829686,
+      "loss": 2.4587,
+      "step": 3901
+    },
+    {
+      "epoch": 0.24003444881889763,
+      "grad_norm": 1.1597360372543335,
+      "learning_rate": 0.00017298998733187863,
+      "loss": 3.509,
+      "step": 3902
+    },
+    {
+      "epoch": 0.24009596456692914,
+      "grad_norm": 1.219992756843567,
+      "learning_rate": 0.00017297676764604193,
+      "loss": 2.9096,
+      "step": 3903
+    },
+    {
+      "epoch": 0.24015748031496062,
+      "grad_norm": 1.1194401979446411,
+      "learning_rate": 0.00017296354523128099,
+      "loss": 3.2316,
+      "step": 3904
+    },
+    {
+      "epoch": 0.24021899606299213,
+      "grad_norm": 0.986416757106781,
+      "learning_rate": 0.00017295032008809037,
+      "loss": 2.8399,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2402805118110236,
+      "grad_norm": 1.443367600440979,
+      "learning_rate": 0.0001729370922169645,
+      "loss": 2.7909,
+      "step": 3906
+    },
+    {
+      "epoch": 0.24034202755905512,
+      "grad_norm": 1.6693475246429443,
+      "learning_rate": 0.00017292386161839814,
+      "loss": 2.1845,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2404035433070866,
+      "grad_norm": 1.2144622802734375,
+      "learning_rate": 0.00017291062829288598,
+      "loss": 2.1925,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2404650590551181,
+      "grad_norm": 1.3287628889083862,
+      "learning_rate": 0.00017289739224092286,
+      "loss": 2.1455,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2405265748031496,
+      "grad_norm": 1.2872116565704346,
+      "learning_rate": 0.00017288415346300377,
+      "loss": 2.7266,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2405880905511811,
+      "grad_norm": 1.9660875797271729,
+      "learning_rate": 0.00017287091195962377,
+      "loss": 2.6664,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2406496062992126,
+      "grad_norm": 1.2259751558303833,
+      "learning_rate": 0.000172857667731278,
+      "loss": 2.986,
+      "step": 3912
+    },
+    {
+      "epoch": 0.2407111220472441,
+      "grad_norm": 1.2165358066558838,
+      "learning_rate": 0.00017284442077846168,
+      "loss": 2.7261,
+      "step": 3913
+    },
+    {
+      "epoch": 0.24077263779527558,
+      "grad_norm": 0.986617386341095,
+      "learning_rate": 0.00017283117110167027,
+      "loss": 3.0437,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2408341535433071,
+      "grad_norm": 1.1070950031280518,
+      "learning_rate": 0.00017281791870139913,
+      "loss": 3.0431,
+      "step": 3915
+    },
+    {
+      "epoch": 0.24089566929133857,
+      "grad_norm": 1.4656431674957275,
+      "learning_rate": 0.0001728046635781439,
+      "loss": 2.6687,
+      "step": 3916
+    },
+    {
+      "epoch": 0.24095718503937008,
+      "grad_norm": 1.0634243488311768,
+      "learning_rate": 0.00017279140573240023,
+      "loss": 3.1993,
+      "step": 3917
+    },
+    {
+      "epoch": 0.24101870078740156,
+      "grad_norm": 3.746154546737671,
+      "learning_rate": 0.00017277814516466385,
+      "loss": 2.2579,
+      "step": 3918
+    },
+    {
+      "epoch": 0.24108021653543307,
+      "grad_norm": 1.5117517709732056,
+      "learning_rate": 0.0001727648818754307,
+      "loss": 2.7819,
+      "step": 3919
+    },
+    {
+      "epoch": 0.24114173228346455,
+      "grad_norm": 1.1398324966430664,
+      "learning_rate": 0.0001727516158651967,
+      "loss": 2.7083,
+      "step": 3920
+    },
+    {
+      "epoch": 0.24120324803149606,
+      "grad_norm": 1.0014721155166626,
+      "learning_rate": 0.00017273834713445795,
+      "loss": 3.1758,
+      "step": 3921
+    },
+    {
+      "epoch": 0.24126476377952755,
+      "grad_norm": 0.979610025882721,
+      "learning_rate": 0.0001727250756837106,
+      "loss": 3.0983,
+      "step": 3922
+    },
+    {
+      "epoch": 0.24132627952755906,
+      "grad_norm": 1.0020484924316406,
+      "learning_rate": 0.00017271180151345097,
+      "loss": 2.7634,
+      "step": 3923
+    },
+    {
+      "epoch": 0.24138779527559054,
+      "grad_norm": 1.125357747077942,
+      "learning_rate": 0.00017269852462417537,
+      "loss": 2.3817,
+      "step": 3924
+    },
+    {
+      "epoch": 0.24144931102362205,
+      "grad_norm": 1.0147311687469482,
+      "learning_rate": 0.00017268524501638037,
+      "loss": 3.0484,
+      "step": 3925
+    },
+    {
+      "epoch": 0.24151082677165353,
+      "grad_norm": 1.000040054321289,
+      "learning_rate": 0.0001726719626905625,
+      "loss": 2.7874,
+      "step": 3926
+    },
+    {
+      "epoch": 0.24157234251968504,
+      "grad_norm": 1.1558159589767456,
+      "learning_rate": 0.00017265867764721846,
+      "loss": 3.1676,
+      "step": 3927
+    },
+    {
+      "epoch": 0.24163385826771652,
+      "grad_norm": 1.125652551651001,
+      "learning_rate": 0.00017264538988684503,
+      "loss": 2.8484,
+      "step": 3928
+    },
+    {
+      "epoch": 0.24169537401574803,
+      "grad_norm": 1.6047139167785645,
+      "learning_rate": 0.00017263209940993909,
+      "loss": 3.1686,
+      "step": 3929
+    },
+    {
+      "epoch": 0.24175688976377951,
+      "grad_norm": 1.1161556243896484,
+      "learning_rate": 0.00017261880621699766,
+      "loss": 2.6331,
+      "step": 3930
+    },
+    {
+      "epoch": 0.24181840551181102,
+      "grad_norm": 1.4770761728286743,
+      "learning_rate": 0.00017260551030851776,
+      "loss": 2.094,
+      "step": 3931
+    },
+    {
+      "epoch": 0.2418799212598425,
+      "grad_norm": 1.9441763162612915,
+      "learning_rate": 0.00017259221168499668,
+      "loss": 2.6591,
+      "step": 3932
+    },
+    {
+      "epoch": 0.24194143700787402,
+      "grad_norm": 1.165327548980713,
+      "learning_rate": 0.00017257891034693165,
+      "loss": 2.3921,
+      "step": 3933
+    },
+    {
+      "epoch": 0.2420029527559055,
+      "grad_norm": 1.2461069822311401,
+      "learning_rate": 0.0001725656062948201,
+      "loss": 2.8951,
+      "step": 3934
+    },
+    {
+      "epoch": 0.242064468503937,
+      "grad_norm": 1.2031822204589844,
+      "learning_rate": 0.0001725522995291595,
+      "loss": 2.7088,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2421259842519685,
+      "grad_norm": 2.0765573978424072,
+      "learning_rate": 0.00017253899005044746,
+      "loss": 2.093,
+      "step": 3936
+    },
+    {
+      "epoch": 0.2421875,
+      "grad_norm": 1.1281946897506714,
+      "learning_rate": 0.0001725256778591817,
+      "loss": 2.6323,
+      "step": 3937
+    },
+    {
+      "epoch": 0.2422490157480315,
+      "grad_norm": 1.1387556791305542,
+      "learning_rate": 0.00017251236295585997,
+      "loss": 3.5187,
+      "step": 3938
+    },
+    {
+      "epoch": 0.242310531496063,
+      "grad_norm": 1.0921579599380493,
+      "learning_rate": 0.00017249904534098021,
+      "loss": 3.2498,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2423720472440945,
+      "grad_norm": 1.1916468143463135,
+      "learning_rate": 0.00017248572501504046,
+      "loss": 2.7419,
+      "step": 3940
+    },
+    {
+      "epoch": 0.24243356299212598,
+      "grad_norm": 1.2354893684387207,
+      "learning_rate": 0.00017247240197853874,
+      "loss": 3.1297,
+      "step": 3941
+    },
+    {
+      "epoch": 0.2424950787401575,
+      "grad_norm": 1.4427175521850586,
+      "learning_rate": 0.00017245907623197334,
+      "loss": 2.3002,
+      "step": 3942
+    },
+    {
+      "epoch": 0.24255659448818898,
+      "grad_norm": 1.0316396951675415,
+      "learning_rate": 0.00017244574777584251,
+      "loss": 2.3946,
+      "step": 3943
+    },
+    {
+      "epoch": 0.24261811023622049,
+      "grad_norm": 1.0896556377410889,
+      "learning_rate": 0.00017243241661064467,
+      "loss": 3.2019,
+      "step": 3944
+    },
+    {
+      "epoch": 0.24267962598425197,
+      "grad_norm": 1.077627420425415,
+      "learning_rate": 0.00017241908273687835,
+      "loss": 2.5319,
+      "step": 3945
+    },
+    {
+      "epoch": 0.24274114173228348,
+      "grad_norm": 1.4864741563796997,
+      "learning_rate": 0.00017240574615504217,
+      "loss": 2.7858,
+      "step": 3946
+    },
+    {
+      "epoch": 0.24280265748031496,
+      "grad_norm": 1.3671191930770874,
+      "learning_rate": 0.0001723924068656348,
+      "loss": 2.5321,
+      "step": 3947
+    },
+    {
+      "epoch": 0.24286417322834647,
+      "grad_norm": 1.1925252676010132,
+      "learning_rate": 0.00017237906486915514,
+      "loss": 2.3597,
+      "step": 3948
+    },
+    {
+      "epoch": 0.24292568897637795,
+      "grad_norm": 1.0208948850631714,
+      "learning_rate": 0.00017236572016610203,
+      "loss": 3.2005,
+      "step": 3949
+    },
+    {
+      "epoch": 0.24298720472440946,
+      "grad_norm": 1.0741597414016724,
+      "learning_rate": 0.00017235237275697448,
+      "loss": 3.0759,
+      "step": 3950
+    },
+    {
+      "epoch": 0.24304872047244094,
+      "grad_norm": 1.2639814615249634,
+      "learning_rate": 0.00017233902264227163,
+      "loss": 2.6329,
+      "step": 3951
+    },
+    {
+      "epoch": 0.24311023622047245,
+      "grad_norm": 1.3073880672454834,
+      "learning_rate": 0.00017232566982249275,
+      "loss": 2.4687,
+      "step": 3952
+    },
+    {
+      "epoch": 0.24317175196850394,
+      "grad_norm": 1.0714483261108398,
+      "learning_rate": 0.0001723123142981371,
+      "loss": 2.7898,
+      "step": 3953
+    },
+    {
+      "epoch": 0.24323326771653545,
+      "grad_norm": 1.0281496047973633,
+      "learning_rate": 0.0001722989560697041,
+      "loss": 2.625,
+      "step": 3954
+    },
+    {
+      "epoch": 0.24329478346456693,
+      "grad_norm": 1.1682318449020386,
+      "learning_rate": 0.0001722855951376933,
+      "loss": 2.9939,
+      "step": 3955
+    },
+    {
+      "epoch": 0.24335629921259844,
+      "grad_norm": 1.034995436668396,
+      "learning_rate": 0.00017227223150260434,
+      "loss": 3.0972,
+      "step": 3956
+    },
+    {
+      "epoch": 0.24341781496062992,
+      "grad_norm": 1.3524872064590454,
+      "learning_rate": 0.00017225886516493693,
+      "loss": 2.4204,
+      "step": 3957
+    },
+    {
+      "epoch": 0.24347933070866143,
+      "grad_norm": 1.0742861032485962,
+      "learning_rate": 0.00017224549612519083,
+      "loss": 3.5283,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2435408464566929,
+      "grad_norm": 1.0905797481536865,
+      "learning_rate": 0.00017223212438386608,
+      "loss": 3.1961,
+      "step": 3959
+    },
+    {
+      "epoch": 0.24360236220472442,
+      "grad_norm": 1.1702771186828613,
+      "learning_rate": 0.00017221874994146266,
+      "loss": 2.1172,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2436638779527559,
+      "grad_norm": 1.171758770942688,
+      "learning_rate": 0.00017220537279848065,
+      "loss": 2.5502,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2437253937007874,
+      "grad_norm": 1.1672751903533936,
+      "learning_rate": 0.00017219199295542037,
+      "loss": 2.8632,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2437869094488189,
+      "grad_norm": 1.1456882953643799,
+      "learning_rate": 0.0001721786104127821,
+      "loss": 1.9034,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2438484251968504,
+      "grad_norm": 1.4111616611480713,
+      "learning_rate": 0.00017216522517106623,
+      "loss": 2.9812,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2439099409448819,
+      "grad_norm": 1.1875907182693481,
+      "learning_rate": 0.0001721518372307734,
+      "loss": 2.2768,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2439714566929134,
+      "grad_norm": 1.1189239025115967,
+      "learning_rate": 0.00017213844659240415,
+      "loss": 2.8074,
+      "step": 3966
+    },
+    {
+      "epoch": 0.24403297244094488,
+      "grad_norm": 1.3005532026290894,
+      "learning_rate": 0.0001721250532564593,
+      "loss": 2.5781,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2440944881889764,
+      "grad_norm": 1.5200008153915405,
+      "learning_rate": 0.0001721116572234396,
+      "loss": 2.808,
+      "step": 3968
+    },
+    {
+      "epoch": 0.24415600393700787,
+      "grad_norm": 1.7213349342346191,
+      "learning_rate": 0.00017209825849384602,
+      "loss": 2.3994,
+      "step": 3969
+    },
+    {
+      "epoch": 0.24421751968503938,
+      "grad_norm": 1.2872523069381714,
+      "learning_rate": 0.00017208485706817964,
+      "loss": 2.6462,
+      "step": 3970
+    },
+    {
+      "epoch": 0.24427903543307086,
+      "grad_norm": 1.2290619611740112,
+      "learning_rate": 0.00017207145294694152,
+      "loss": 3.3751,
+      "step": 3971
+    },
+    {
+      "epoch": 0.24434055118110237,
+      "grad_norm": 1.116658329963684,
+      "learning_rate": 0.00017205804613063298,
+      "loss": 2.8811,
+      "step": 3972
+    },
+    {
+      "epoch": 0.24440206692913385,
+      "grad_norm": 1.0720453262329102,
+      "learning_rate": 0.0001720446366197553,
+      "loss": 2.3373,
+      "step": 3973
+    },
+    {
+      "epoch": 0.24446358267716536,
+      "grad_norm": 1.1494511365890503,
+      "learning_rate": 0.00017203122441480996,
+      "loss": 2.6229,
+      "step": 3974
+    },
+    {
+      "epoch": 0.24452509842519685,
+      "grad_norm": 1.305883526802063,
+      "learning_rate": 0.00017201780951629845,
+      "loss": 2.977,
+      "step": 3975
+    },
+    {
+      "epoch": 0.24458661417322836,
+      "grad_norm": 1.1699001789093018,
+      "learning_rate": 0.00017200439192472245,
+      "loss": 2.6592,
+      "step": 3976
+    },
+    {
+      "epoch": 0.24464812992125984,
+      "grad_norm": 1.0400241613388062,
+      "learning_rate": 0.00017199097164058373,
+      "loss": 2.8299,
+      "step": 3977
+    },
+    {
+      "epoch": 0.24470964566929135,
+      "grad_norm": 1.4576330184936523,
+      "learning_rate": 0.00017197754866438408,
+      "loss": 3.1113,
+      "step": 3978
+    },
+    {
+      "epoch": 0.24477116141732283,
+      "grad_norm": 1.012998104095459,
+      "learning_rate": 0.00017196412299662548,
+      "loss": 2.7545,
+      "step": 3979
+    },
+    {
+      "epoch": 0.24483267716535434,
+      "grad_norm": 0.8893311023712158,
+      "learning_rate": 0.00017195069463780995,
+      "loss": 3.3965,
+      "step": 3980
+    },
+    {
+      "epoch": 0.24489419291338582,
+      "grad_norm": 1.195068597793579,
+      "learning_rate": 0.00017193726358843967,
+      "loss": 2.2298,
+      "step": 3981
+    },
+    {
+      "epoch": 0.24495570866141733,
+      "grad_norm": 0.8879215717315674,
+      "learning_rate": 0.00017192382984901686,
+      "loss": 2.737,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2450172244094488,
+      "grad_norm": 0.9551054239273071,
+      "learning_rate": 0.00017191039342004386,
+      "loss": 3.073,
+      "step": 3983
+    },
+    {
+      "epoch": 0.24507874015748032,
+      "grad_norm": 1.0818164348602295,
+      "learning_rate": 0.00017189695430202311,
+      "loss": 3.1757,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2451402559055118,
+      "grad_norm": 1.1997551918029785,
+      "learning_rate": 0.0001718835124954572,
+      "loss": 2.8165,
+      "step": 3985
+    },
+    {
+      "epoch": 0.24520177165354332,
+      "grad_norm": 1.084244966506958,
+      "learning_rate": 0.00017187006800084876,
+      "loss": 2.9663,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2452632874015748,
+      "grad_norm": 0.9372528195381165,
+      "learning_rate": 0.00017185662081870055,
+      "loss": 3.1333,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2453248031496063,
+      "grad_norm": 1.1664632558822632,
+      "learning_rate": 0.00017184317094951537,
+      "loss": 2.8321,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2453863188976378,
+      "grad_norm": 1.2676154375076294,
+      "learning_rate": 0.00017182971839379624,
+      "loss": 2.9091,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2454478346456693,
+      "grad_norm": 1.2897026538848877,
+      "learning_rate": 0.00017181626315204614,
+      "loss": 3.351,
+      "step": 3990
+    },
+    {
+      "epoch": 0.24550935039370078,
+      "grad_norm": 1.1546610593795776,
+      "learning_rate": 0.00017180280522476827,
+      "loss": 3.5054,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2455708661417323,
+      "grad_norm": 0.9925970435142517,
+      "learning_rate": 0.0001717893446124659,
+      "loss": 3.1606,
+      "step": 3992
+    },
+    {
+      "epoch": 0.24563238188976377,
+      "grad_norm": 1.1613878011703491,
+      "learning_rate": 0.0001717758813156423,
+      "loss": 3.0439,
+      "step": 3993
+    },
+    {
+      "epoch": 0.24569389763779528,
+      "grad_norm": 1.278088092803955,
+      "learning_rate": 0.00017176241533480103,
+      "loss": 2.5769,
+      "step": 3994
+    },
+    {
+      "epoch": 0.24575541338582677,
+      "grad_norm": 1.3188815116882324,
+      "learning_rate": 0.0001717489466704455,
+      "loss": 3.2838,
+      "step": 3995
+    },
+    {
+      "epoch": 0.24581692913385828,
+      "grad_norm": 1.1745742559432983,
+      "learning_rate": 0.00017173547532307952,
+      "loss": 2.7287,
+      "step": 3996
+    },
+    {
+      "epoch": 0.24587844488188976,
+      "grad_norm": 1.1159677505493164,
+      "learning_rate": 0.00017172200129320675,
+      "loss": 2.6938,
+      "step": 3997
+    },
+    {
+      "epoch": 0.24593996062992127,
+      "grad_norm": 1.135483980178833,
+      "learning_rate": 0.00017170852458133105,
+      "loss": 2.8629,
+      "step": 3998
+    },
+    {
+      "epoch": 0.24600147637795275,
+      "grad_norm": 1.334519863128662,
+      "learning_rate": 0.0001716950451879564,
+      "loss": 2.2523,
+      "step": 3999
+    },
+    {
+      "epoch": 0.24606299212598426,
+      "grad_norm": 1.0743759870529175,
+      "learning_rate": 0.00017168156311358685,
+      "loss": 3.0543,
+      "step": 4000
+    },
+    {
+      "epoch": 0.24612450787401574,
+      "grad_norm": 1.0468536615371704,
+      "learning_rate": 0.00017166807835872653,
+      "loss": 2.6719,
+      "step": 4001
+    },
+    {
+      "epoch": 0.24618602362204725,
+      "grad_norm": 1.0416256189346313,
+      "learning_rate": 0.00017165459092387971,
+      "loss": 2.2493,
+      "step": 4002
+    },
+    {
+      "epoch": 0.24624753937007873,
+      "grad_norm": 0.9472552537918091,
+      "learning_rate": 0.00017164110080955078,
+      "loss": 2.6996,
+      "step": 4003
+    },
+    {
+      "epoch": 0.24630905511811024,
+      "grad_norm": 1.3395686149597168,
+      "learning_rate": 0.00017162760801624414,
+      "loss": 2.3531,
+      "step": 4004
+    },
+    {
+      "epoch": 0.24637057086614172,
+      "grad_norm": 0.8829546570777893,
+      "learning_rate": 0.00017161411254446436,
+      "loss": 2.9291,
+      "step": 4005
+    },
+    {
+      "epoch": 0.24643208661417323,
+      "grad_norm": 1.1062439680099487,
+      "learning_rate": 0.00017160061439471612,
+      "loss": 2.3875,
+      "step": 4006
+    },
+    {
+      "epoch": 0.24649360236220472,
+      "grad_norm": 1.3008427619934082,
+      "learning_rate": 0.00017158711356750415,
+      "loss": 2.9205,
+      "step": 4007
+    },
+    {
+      "epoch": 0.24655511811023623,
+      "grad_norm": 1.0422676801681519,
+      "learning_rate": 0.0001715736100633333,
+      "loss": 3.3334,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2466166338582677,
+      "grad_norm": 1.0070180892944336,
+      "learning_rate": 0.00017156010388270855,
+      "loss": 2.9762,
+      "step": 4009
+    },
+    {
+      "epoch": 0.24667814960629922,
+      "grad_norm": 1.2347776889801025,
+      "learning_rate": 0.00017154659502613493,
+      "loss": 2.6861,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2467396653543307,
+      "grad_norm": 1.4553788900375366,
+      "learning_rate": 0.00017153308349411762,
+      "loss": 3.4058,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2468011811023622,
+      "grad_norm": 1.333693504333496,
+      "learning_rate": 0.0001715195692871619,
+      "loss": 3.855,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2468626968503937,
+      "grad_norm": 0.9432052969932556,
+      "learning_rate": 0.00017150605240577303,
+      "loss": 3.5243,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2469242125984252,
+      "grad_norm": 1.1108343601226807,
+      "learning_rate": 0.00017149253285045658,
+      "loss": 2.9356,
+      "step": 4014
+    },
+    {
+      "epoch": 0.24698572834645668,
+      "grad_norm": 1.249048113822937,
+      "learning_rate": 0.00017147901062171804,
+      "loss": 2.3456,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2470472440944882,
+      "grad_norm": 1.139425277709961,
+      "learning_rate": 0.00017146548572006306,
+      "loss": 2.3739,
+      "step": 4016
+    },
+    {
+      "epoch": 0.24710875984251968,
+      "grad_norm": 1.1063225269317627,
+      "learning_rate": 0.00017145195814599745,
+      "loss": 3.3263,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2471702755905512,
+      "grad_norm": 1.147321105003357,
+      "learning_rate": 0.000171438427900027,
+      "loss": 3.3311,
+      "step": 4018
+    },
+    {
+      "epoch": 0.24723179133858267,
+      "grad_norm": 1.2345256805419922,
+      "learning_rate": 0.00017142489498265772,
+      "loss": 2.4616,
+      "step": 4019
+    },
+    {
+      "epoch": 0.24729330708661418,
+      "grad_norm": 1.087196946144104,
+      "learning_rate": 0.00017141135939439562,
+      "loss": 2.8321,
+      "step": 4020
+    },
+    {
+      "epoch": 0.24735482283464566,
+      "grad_norm": 1.2062158584594727,
+      "learning_rate": 0.0001713978211357469,
+      "loss": 2.1642,
+      "step": 4021
+    },
+    {
+      "epoch": 0.24741633858267717,
+      "grad_norm": 1.6171399354934692,
+      "learning_rate": 0.00017138428020721782,
+      "loss": 1.6165,
+      "step": 4022
+    },
+    {
+      "epoch": 0.24747785433070865,
+      "grad_norm": 1.3036489486694336,
+      "learning_rate": 0.00017137073660931468,
+      "loss": 3.0217,
+      "step": 4023
+    },
+    {
+      "epoch": 0.24753937007874016,
+      "grad_norm": 1.169500470161438,
+      "learning_rate": 0.00017135719034254394,
+      "loss": 2.9086,
+      "step": 4024
+    },
+    {
+      "epoch": 0.24760088582677164,
+      "grad_norm": 1.3229453563690186,
+      "learning_rate": 0.00017134364140741223,
+      "loss": 3.0569,
+      "step": 4025
+    },
+    {
+      "epoch": 0.24766240157480315,
+      "grad_norm": 1.3041561841964722,
+      "learning_rate": 0.0001713300898044261,
+      "loss": 2.8027,
+      "step": 4026
+    },
+    {
+      "epoch": 0.24772391732283464,
+      "grad_norm": 1.1685458421707153,
+      "learning_rate": 0.0001713165355340924,
+      "loss": 3.5307,
+      "step": 4027
+    },
+    {
+      "epoch": 0.24778543307086615,
+      "grad_norm": 1.18390691280365,
+      "learning_rate": 0.00017130297859691793,
+      "loss": 3.3076,
+      "step": 4028
+    },
+    {
+      "epoch": 0.24784694881889763,
+      "grad_norm": 1.1530523300170898,
+      "learning_rate": 0.0001712894189934097,
+      "loss": 2.33,
+      "step": 4029
+    },
+    {
+      "epoch": 0.24790846456692914,
+      "grad_norm": 0.9363012313842773,
+      "learning_rate": 0.00017127585672407463,
+      "loss": 2.7425,
+      "step": 4030
+    },
+    {
+      "epoch": 0.24796998031496062,
+      "grad_norm": 1.0206855535507202,
+      "learning_rate": 0.00017126229178942003,
+      "loss": 3.1002,
+      "step": 4031
+    },
+    {
+      "epoch": 0.24803149606299213,
+      "grad_norm": 1.0562562942504883,
+      "learning_rate": 0.00017124872418995308,
+      "loss": 2.8861,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2480930118110236,
+      "grad_norm": 0.9037251472473145,
+      "learning_rate": 0.00017123515392618112,
+      "loss": 3.2142,
+      "step": 4033
+    },
+    {
+      "epoch": 0.24815452755905512,
+      "grad_norm": 1.2449413537979126,
+      "learning_rate": 0.00017122158099861163,
+      "loss": 2.712,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2482160433070866,
+      "grad_norm": 0.8742977380752563,
+      "learning_rate": 0.00017120800540775216,
+      "loss": 3.0839,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2482775590551181,
+      "grad_norm": 1.3541055917739868,
+      "learning_rate": 0.00017119442715411034,
+      "loss": 2.8262,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2483390748031496,
+      "grad_norm": 1.2947888374328613,
+      "learning_rate": 0.00017118084623819394,
+      "loss": 3.0909,
+      "step": 4037
+    },
+    {
+      "epoch": 0.2484005905511811,
+      "grad_norm": 1.0870224237442017,
+      "learning_rate": 0.0001711672626605108,
+      "loss": 2.636,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2484621062992126,
+      "grad_norm": 2.628187417984009,
+      "learning_rate": 0.0001711536764215689,
+      "loss": 2.4367,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2485236220472441,
+      "grad_norm": 1.149507999420166,
+      "learning_rate": 0.00017114008752187627,
+      "loss": 3.2509,
+      "step": 4040
+    },
+    {
+      "epoch": 0.24858513779527558,
+      "grad_norm": 1.0537090301513672,
+      "learning_rate": 0.00017112649596194102,
+      "loss": 2.9926,
+      "step": 4041
+    },
+    {
+      "epoch": 0.2486466535433071,
+      "grad_norm": 1.2696714401245117,
+      "learning_rate": 0.00017111290174227147,
+      "loss": 2.766,
+      "step": 4042
+    },
+    {
+      "epoch": 0.24870816929133857,
+      "grad_norm": 1.096143126487732,
+      "learning_rate": 0.0001710993048633759,
+      "loss": 2.9269,
+      "step": 4043
+    },
+    {
+      "epoch": 0.24876968503937008,
+      "grad_norm": 1.3115710020065308,
+      "learning_rate": 0.0001710857053257628,
+      "loss": 2.9828,
+      "step": 4044
+    },
+    {
+      "epoch": 0.24883120078740156,
+      "grad_norm": 1.2399767637252808,
+      "learning_rate": 0.0001710721031299407,
+      "loss": 3.0257,
+      "step": 4045
+    },
+    {
+      "epoch": 0.24889271653543307,
+      "grad_norm": 1.237003207206726,
+      "learning_rate": 0.00017105849827641826,
+      "loss": 2.5767,
+      "step": 4046
+    },
+    {
+      "epoch": 0.24895423228346455,
+      "grad_norm": 1.1193419694900513,
+      "learning_rate": 0.0001710448907657042,
+      "loss": 2.6499,
+      "step": 4047
+    },
+    {
+      "epoch": 0.24901574803149606,
+      "grad_norm": 1.476219654083252,
+      "learning_rate": 0.0001710312805983074,
+      "loss": 2.1775,
+      "step": 4048
+    },
+    {
+      "epoch": 0.24907726377952755,
+      "grad_norm": 1.1738646030426025,
+      "learning_rate": 0.00017101766777473677,
+      "loss": 2.1577,
+      "step": 4049
+    },
+    {
+      "epoch": 0.24913877952755906,
+      "grad_norm": 1.2151941061019897,
+      "learning_rate": 0.0001710040522955014,
+      "loss": 2.8923,
+      "step": 4050
+    },
+    {
+      "epoch": 0.24920029527559054,
+      "grad_norm": 1.0081156492233276,
+      "learning_rate": 0.00017099043416111037,
+      "loss": 2.5435,
+      "step": 4051
+    },
+    {
+      "epoch": 0.24926181102362205,
+      "grad_norm": 1.6084312200546265,
+      "learning_rate": 0.000170976813372073,
+      "loss": 3.3161,
+      "step": 4052
+    },
+    {
+      "epoch": 0.24932332677165353,
+      "grad_norm": 0.9634644985198975,
+      "learning_rate": 0.00017096318992889856,
+      "loss": 3.0413,
+      "step": 4053
+    },
+    {
+      "epoch": 0.24938484251968504,
+      "grad_norm": 1.1880184412002563,
+      "learning_rate": 0.0001709495638320965,
+      "loss": 2.8884,
+      "step": 4054
+    },
+    {
+      "epoch": 0.24944635826771652,
+      "grad_norm": 1.5809603929519653,
+      "learning_rate": 0.0001709359350821764,
+      "loss": 2.5113,
+      "step": 4055
+    },
+    {
+      "epoch": 0.24950787401574803,
+      "grad_norm": 1.1552515029907227,
+      "learning_rate": 0.0001709223036796479,
+      "loss": 2.1435,
+      "step": 4056
+    },
+    {
+      "epoch": 0.24956938976377951,
+      "grad_norm": 1.1125680208206177,
+      "learning_rate": 0.00017090866962502066,
+      "loss": 3.061,
+      "step": 4057
+    },
+    {
+      "epoch": 0.24963090551181102,
+      "grad_norm": 1.1932053565979004,
+      "learning_rate": 0.0001708950329188046,
+      "loss": 2.8902,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2496924212598425,
+      "grad_norm": 1.0116554498672485,
+      "learning_rate": 0.00017088139356150964,
+      "loss": 2.9975,
+      "step": 4059
+    },
+    {
+      "epoch": 0.24975393700787402,
+      "grad_norm": 1.0070563554763794,
+      "learning_rate": 0.00017086775155364582,
+      "loss": 2.6949,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2498154527559055,
+      "grad_norm": 1.2800427675247192,
+      "learning_rate": 0.00017085410689572324,
+      "loss": 2.7662,
+      "step": 4061
+    },
+    {
+      "epoch": 0.249876968503937,
+      "grad_norm": 1.2634629011154175,
+      "learning_rate": 0.00017084045958825215,
+      "loss": 2.7259,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2499384842519685,
+      "grad_norm": 1.0858054161071777,
+      "learning_rate": 0.0001708268096317429,
+      "loss": 2.4251,
+      "step": 4063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1078839302062988,
+      "learning_rate": 0.0001708131570267059,
+      "loss": 2.6104,
+      "step": 4064
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.7858633995056152,
+      "eval_runtime": 1035.4061,
+      "eval_samples_per_second": 3.703,
+      "eval_steps_per_second": 3.703,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2500615157480315,
+      "grad_norm": 1.1881117820739746,
+      "learning_rate": 0.00017079950177365172,
+      "loss": 2.9199,
+      "step": 4065
+    },
+    {
+      "epoch": 0.250123031496063,
+      "grad_norm": 0.983165442943573,
+      "learning_rate": 0.00017078584387309094,
+      "loss": 2.3476,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2501845472440945,
+      "grad_norm": 1.0057117938995361,
+      "learning_rate": 0.00017077218332553428,
+      "loss": 2.8972,
+      "step": 4067
+    },
+    {
+      "epoch": 0.250246062992126,
+      "grad_norm": 1.1742733716964722,
+      "learning_rate": 0.00017075852013149264,
+      "loss": 2.7851,
+      "step": 4068
+    },
+    {
+      "epoch": 0.25030757874015747,
+      "grad_norm": 1.2744694948196411,
+      "learning_rate": 0.0001707448542914769,
+      "loss": 2.862,
+      "step": 4069
+    },
+    {
+      "epoch": 0.250369094488189,
+      "grad_norm": 1.2118338346481323,
+      "learning_rate": 0.0001707311858059981,
+      "loss": 2.4374,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2504306102362205,
+      "grad_norm": 1.2856041193008423,
+      "learning_rate": 0.00017071751467556735,
+      "loss": 2.9866,
+      "step": 4071
+    },
+    {
+      "epoch": 0.25049212598425197,
+      "grad_norm": 1.2575985193252563,
+      "learning_rate": 0.00017070384090069589,
+      "loss": 2.1254,
+      "step": 4072
+    },
+    {
+      "epoch": 0.25055364173228345,
+      "grad_norm": 1.3118568658828735,
+      "learning_rate": 0.00017069016448189504,
+      "loss": 2.1898,
+      "step": 4073
+    },
+    {
+      "epoch": 0.250615157480315,
+      "grad_norm": 1.3466334342956543,
+      "learning_rate": 0.0001706764854196762,
+      "loss": 2.2876,
+      "step": 4074
+    },
+    {
+      "epoch": 0.25067667322834647,
+      "grad_norm": 0.9743967056274414,
+      "learning_rate": 0.0001706628037145509,
+      "loss": 2.8637,
+      "step": 4075
+    },
+    {
+      "epoch": 0.25073818897637795,
+      "grad_norm": 1.0102006196975708,
+      "learning_rate": 0.0001706491193670308,
+      "loss": 2.9113,
+      "step": 4076
+    },
+    {
+      "epoch": 0.25079970472440943,
+      "grad_norm": 1.2817350625991821,
+      "learning_rate": 0.00017063543237762758,
+      "loss": 2.59,
+      "step": 4077
+    },
+    {
+      "epoch": 0.25086122047244097,
+      "grad_norm": 1.4130219221115112,
+      "learning_rate": 0.00017062174274685304,
+      "loss": 2.9045,
+      "step": 4078
+    },
+    {
+      "epoch": 0.25092273622047245,
+      "grad_norm": 1.4422401189804077,
+      "learning_rate": 0.00017060805047521916,
+      "loss": 1.606,
+      "step": 4079
+    },
+    {
+      "epoch": 0.25098425196850394,
+      "grad_norm": 1.0518760681152344,
+      "learning_rate": 0.00017059435556323788,
+      "loss": 2.9586,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2510457677165354,
+      "grad_norm": 1.5292439460754395,
+      "learning_rate": 0.00017058065801142135,
+      "loss": 3.0851,
+      "step": 4081
+    },
+    {
+      "epoch": 0.25110728346456695,
+      "grad_norm": 1.270909070968628,
+      "learning_rate": 0.0001705669578202818,
+      "loss": 2.3259,
+      "step": 4082
+    },
+    {
+      "epoch": 0.25116879921259844,
+      "grad_norm": 1.0167604684829712,
+      "learning_rate": 0.00017055325499033147,
+      "loss": 3.1083,
+      "step": 4083
+    },
+    {
+      "epoch": 0.2512303149606299,
+      "grad_norm": 1.5400890111923218,
+      "learning_rate": 0.00017053954952208285,
+      "loss": 2.2793,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2512918307086614,
+      "grad_norm": 1.200124979019165,
+      "learning_rate": 0.00017052584141604845,
+      "loss": 2.7798,
+      "step": 4085
+    },
+    {
+      "epoch": 0.25135334645669294,
+      "grad_norm": 0.9745237827301025,
+      "learning_rate": 0.0001705121306727408,
+      "loss": 2.5439,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2514148622047244,
+      "grad_norm": 1.3755744695663452,
+      "learning_rate": 0.00017049841729267266,
+      "loss": 2.5398,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2514763779527559,
+      "grad_norm": 1.0205726623535156,
+      "learning_rate": 0.00017048470127635683,
+      "loss": 2.9709,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2515378937007874,
+      "grad_norm": 1.144499659538269,
+      "learning_rate": 0.00017047098262430618,
+      "loss": 2.6125,
+      "step": 4089
+    },
+    {
+      "epoch": 0.2515994094488189,
+      "grad_norm": 0.9498146176338196,
+      "learning_rate": 0.0001704572613370338,
+      "loss": 2.9811,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2516609251968504,
+      "grad_norm": 1.081459879875183,
+      "learning_rate": 0.00017044353741505267,
+      "loss": 3.2504,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2517224409448819,
+      "grad_norm": 1.0736784934997559,
+      "learning_rate": 0.00017042981085887605,
+      "loss": 2.4556,
+      "step": 4092
+    },
+    {
+      "epoch": 0.25178395669291337,
+      "grad_norm": 1.1806200742721558,
+      "learning_rate": 0.00017041608166901723,
+      "loss": 2.9882,
+      "step": 4093
+    },
+    {
+      "epoch": 0.2518454724409449,
+      "grad_norm": 1.0927183628082275,
+      "learning_rate": 0.00017040234984598963,
+      "loss": 2.4372,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2519069881889764,
+      "grad_norm": 1.2574747800827026,
+      "learning_rate": 0.00017038861539030672,
+      "loss": 2.5432,
+      "step": 4095
+    },
+    {
+      "epoch": 0.25196850393700787,
+      "grad_norm": 1.3413653373718262,
+      "learning_rate": 0.00017037487830248206,
+      "loss": 2.9393,
+      "step": 4096
+    },
+    {
+      "epoch": 0.25203001968503935,
+      "grad_norm": 1.115294337272644,
+      "learning_rate": 0.0001703611385830294,
+      "loss": 2.5092,
+      "step": 4097
+    },
+    {
+      "epoch": 0.2520915354330709,
+      "grad_norm": 0.8539628386497498,
+      "learning_rate": 0.00017034739623246248,
+      "loss": 3.2437,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2521530511811024,
+      "grad_norm": 1.1116410493850708,
+      "learning_rate": 0.00017033365125129524,
+      "loss": 2.8885,
+      "step": 4099
+    },
+    {
+      "epoch": 0.25221456692913385,
+      "grad_norm": 1.2089887857437134,
+      "learning_rate": 0.0001703199036400416,
+      "loss": 2.3789,
+      "step": 4100
+    },
+    {
+      "epoch": 0.25227608267716534,
+      "grad_norm": 1.0900084972381592,
+      "learning_rate": 0.0001703061533992157,
+      "loss": 3.07,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2523375984251969,
+      "grad_norm": 1.2203164100646973,
+      "learning_rate": 0.00017029240052933168,
+      "loss": 2.6822,
+      "step": 4102
+    },
+    {
+      "epoch": 0.25239911417322836,
+      "grad_norm": 1.0756914615631104,
+      "learning_rate": 0.00017027864503090388,
+      "loss": 3.1879,
+      "step": 4103
+    },
+    {
+      "epoch": 0.25246062992125984,
+      "grad_norm": 1.077362298965454,
+      "learning_rate": 0.0001702648869044466,
+      "loss": 2.6023,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2525221456692913,
+      "grad_norm": 1.212341070175171,
+      "learning_rate": 0.00017025112615047435,
+      "loss": 2.7379,
+      "step": 4105
+    },
+    {
+      "epoch": 0.25258366141732286,
+      "grad_norm": 1.0732667446136475,
+      "learning_rate": 0.00017023736276950175,
+      "loss": 2.9749,
+      "step": 4106
+    },
+    {
+      "epoch": 0.25264517716535434,
+      "grad_norm": 1.0316119194030762,
+      "learning_rate": 0.00017022359676204345,
+      "loss": 3.0848,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2527066929133858,
+      "grad_norm": 1.2868725061416626,
+      "learning_rate": 0.00017020982812861415,
+      "loss": 2.6431,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2527682086614173,
+      "grad_norm": 1.5501422882080078,
+      "learning_rate": 0.00017019605686972882,
+      "loss": 2.2441,
+      "step": 4109
+    },
+    {
+      "epoch": 0.25282972440944884,
+      "grad_norm": 0.9072484374046326,
+      "learning_rate": 0.0001701822829859024,
+      "loss": 3.0137,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2528912401574803,
+      "grad_norm": 1.0456891059875488,
+      "learning_rate": 0.00017016850647764991,
+      "loss": 2.0821,
+      "step": 4111
+    },
+    {
+      "epoch": 0.2529527559055118,
+      "grad_norm": 1.143022060394287,
+      "learning_rate": 0.00017015472734548657,
+      "loss": 2.4211,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2530142716535433,
+      "grad_norm": 1.2274023294448853,
+      "learning_rate": 0.00017014094558992761,
+      "loss": 2.4947,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2530757874015748,
+      "grad_norm": 1.19190514087677,
+      "learning_rate": 0.0001701271612114884,
+      "loss": 2.951,
+      "step": 4114
+    },
+    {
+      "epoch": 0.2531373031496063,
+      "grad_norm": 0.9448344111442566,
+      "learning_rate": 0.00017011337421068441,
+      "loss": 3.2507,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2531988188976378,
+      "grad_norm": 1.2354605197906494,
+      "learning_rate": 0.0001700995845880312,
+      "loss": 2.58,
+      "step": 4116
+    },
+    {
+      "epoch": 0.25326033464566927,
+      "grad_norm": 1.057706356048584,
+      "learning_rate": 0.00017008579234404435,
+      "loss": 3.0128,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2533218503937008,
+      "grad_norm": 1.4711167812347412,
+      "learning_rate": 0.00017007199747923974,
+      "loss": 2.888,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2533833661417323,
+      "grad_norm": 1.0500775575637817,
+      "learning_rate": 0.00017005819999413313,
+      "loss": 3.1337,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2534448818897638,
+      "grad_norm": 1.199537992477417,
+      "learning_rate": 0.0001700443998892405,
+      "loss": 2.7023,
+      "step": 4120
+    },
+    {
+      "epoch": 0.25350639763779526,
+      "grad_norm": 1.035539150238037,
+      "learning_rate": 0.00017003059716507788,
+      "loss": 2.2437,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2535679133858268,
+      "grad_norm": 1.1303415298461914,
+      "learning_rate": 0.00017001679182216143,
+      "loss": 3.0277,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2536294291338583,
+      "grad_norm": 1.0673182010650635,
+      "learning_rate": 0.0001700029838610074,
+      "loss": 3.3913,
+      "step": 4123
+    },
+    {
+      "epoch": 0.25369094488188976,
+      "grad_norm": 1.110093355178833,
+      "learning_rate": 0.00016998917328213216,
+      "loss": 2.6662,
+      "step": 4124
+    },
+    {
+      "epoch": 0.25375246062992124,
+      "grad_norm": 0.9972797632217407,
+      "learning_rate": 0.00016997536008605204,
+      "loss": 3.1898,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2538139763779528,
+      "grad_norm": 1.1463209390640259,
+      "learning_rate": 0.0001699615442732837,
+      "loss": 2.7446,
+      "step": 4126
+    },
+    {
+      "epoch": 0.25387549212598426,
+      "grad_norm": 1.3149138689041138,
+      "learning_rate": 0.00016994772584434369,
+      "loss": 3.2964,
+      "step": 4127
+    },
+    {
+      "epoch": 0.25393700787401574,
+      "grad_norm": 1.406433343887329,
+      "learning_rate": 0.00016993390479974878,
+      "loss": 2.727,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2539985236220472,
+      "grad_norm": 1.146575689315796,
+      "learning_rate": 0.0001699200811400158,
+      "loss": 2.6396,
+      "step": 4129
+    },
+    {
+      "epoch": 0.25406003937007876,
+      "grad_norm": 1.0424184799194336,
+      "learning_rate": 0.00016990625486566164,
+      "loss": 3.2596,
+      "step": 4130
+    },
+    {
+      "epoch": 0.25412155511811024,
+      "grad_norm": 1.3194386959075928,
+      "learning_rate": 0.00016989242597720337,
+      "loss": 2.967,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2541830708661417,
+      "grad_norm": 1.2052289247512817,
+      "learning_rate": 0.0001698785944751581,
+      "loss": 2.9443,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2542445866141732,
+      "grad_norm": 1.0689449310302734,
+      "learning_rate": 0.0001698647603600431,
+      "loss": 2.6645,
+      "step": 4133
+    },
+    {
+      "epoch": 0.25430610236220474,
+      "grad_norm": 1.0944863557815552,
+      "learning_rate": 0.0001698509236323756,
+      "loss": 3.4317,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2543676181102362,
+      "grad_norm": 1.3057812452316284,
+      "learning_rate": 0.00016983708429267301,
+      "loss": 2.6861,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2544291338582677,
+      "grad_norm": 1.081725001335144,
+      "learning_rate": 0.00016982324234145294,
+      "loss": 2.9082,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2544906496062992,
+      "grad_norm": 1.0883471965789795,
+      "learning_rate": 0.00016980939777923296,
+      "loss": 3.0189,
+      "step": 4137
+    },
+    {
+      "epoch": 0.25455216535433073,
+      "grad_norm": 1.3699374198913574,
+      "learning_rate": 0.00016979555060653073,
+      "loss": 3.0605,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2546136811023622,
+      "grad_norm": 0.9379632472991943,
+      "learning_rate": 0.00016978170082386414,
+      "loss": 3.1624,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2546751968503937,
+      "grad_norm": 1.4727582931518555,
+      "learning_rate": 0.00016976784843175105,
+      "loss": 3.0379,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2547367125984252,
+      "grad_norm": 1.2544288635253906,
+      "learning_rate": 0.00016975399343070944,
+      "loss": 2.7525,
+      "step": 4141
+    },
+    {
+      "epoch": 0.2547982283464567,
+      "grad_norm": 1.3723382949829102,
+      "learning_rate": 0.00016974013582125746,
+      "loss": 2.7597,
+      "step": 4142
+    },
+    {
+      "epoch": 0.2548597440944882,
+      "grad_norm": 0.9803564548492432,
+      "learning_rate": 0.00016972627560391326,
+      "loss": 3.065,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2549212598425197,
+      "grad_norm": 1.0214924812316895,
+      "learning_rate": 0.00016971241277919513,
+      "loss": 3.0507,
+      "step": 4144
+    },
+    {
+      "epoch": 0.25498277559055116,
+      "grad_norm": 1.2652086019515991,
+      "learning_rate": 0.00016969854734762155,
+      "loss": 2.586,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2550442913385827,
+      "grad_norm": 1.3457527160644531,
+      "learning_rate": 0.00016968467930971094,
+      "loss": 2.8441,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2551058070866142,
+      "grad_norm": 1.0734384059906006,
+      "learning_rate": 0.00016967080866598185,
+      "loss": 2.9397,
+      "step": 4147
+    },
+    {
+      "epoch": 0.25516732283464566,
+      "grad_norm": 1.340882658958435,
+      "learning_rate": 0.00016965693541695303,
+      "loss": 3.3232,
+      "step": 4148
+    },
+    {
+      "epoch": 0.25522883858267714,
+      "grad_norm": 1.0902477502822876,
+      "learning_rate": 0.00016964305956314327,
+      "loss": 2.9829,
+      "step": 4149
+    },
+    {
+      "epoch": 0.2552903543307087,
+      "grad_norm": 0.9102745652198792,
+      "learning_rate": 0.00016962918110507142,
+      "loss": 2.7766,
+      "step": 4150
+    },
+    {
+      "epoch": 0.25535187007874016,
+      "grad_norm": 1.1998645067214966,
+      "learning_rate": 0.00016961530004325643,
+      "loss": 2.9886,
+      "step": 4151
+    },
+    {
+      "epoch": 0.25541338582677164,
+      "grad_norm": 1.0444422960281372,
+      "learning_rate": 0.0001696014163782174,
+      "loss": 2.7913,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2554749015748031,
+      "grad_norm": 1.3215972185134888,
+      "learning_rate": 0.00016958753011047354,
+      "loss": 2.8933,
+      "step": 4153
+    },
+    {
+      "epoch": 0.25553641732283466,
+      "grad_norm": 1.232641339302063,
+      "learning_rate": 0.00016957364124054407,
+      "loss": 2.2378,
+      "step": 4154
+    },
+    {
+      "epoch": 0.25559793307086615,
+      "grad_norm": 0.9893457889556885,
+      "learning_rate": 0.00016955974976894838,
+      "loss": 2.5272,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2556594488188976,
+      "grad_norm": 1.2060935497283936,
+      "learning_rate": 0.00016954585569620593,
+      "loss": 2.4436,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2557209645669291,
+      "grad_norm": 1.0435256958007812,
+      "learning_rate": 0.00016953195902283622,
+      "loss": 3.2328,
+      "step": 4157
+    },
+    {
+      "epoch": 0.25578248031496065,
+      "grad_norm": 0.9801449775695801,
+      "learning_rate": 0.000169518059749359,
+      "loss": 2.6317,
+      "step": 4158
+    },
+    {
+      "epoch": 0.25584399606299213,
+      "grad_norm": 1.268829345703125,
+      "learning_rate": 0.00016950415787629402,
+      "loss": 2.5228,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2559055118110236,
+      "grad_norm": 1.142693042755127,
+      "learning_rate": 0.00016949025340416105,
+      "loss": 2.6572,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2559670275590551,
+      "grad_norm": 1.3090786933898926,
+      "learning_rate": 0.00016947634633348013,
+      "loss": 2.4581,
+      "step": 4161
+    },
+    {
+      "epoch": 0.25602854330708663,
+      "grad_norm": 1.1883511543273926,
+      "learning_rate": 0.00016946243666477123,
+      "loss": 2.084,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2560900590551181,
+      "grad_norm": 1.190834641456604,
+      "learning_rate": 0.00016944852439855458,
+      "loss": 2.6535,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2561515748031496,
+      "grad_norm": 1.0638116598129272,
+      "learning_rate": 0.00016943460953535033,
+      "loss": 2.7923,
+      "step": 4164
+    },
+    {
+      "epoch": 0.2562130905511811,
+      "grad_norm": 1.1736892461776733,
+      "learning_rate": 0.00016942069207567887,
+      "loss": 3.1404,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2562746062992126,
+      "grad_norm": 1.1742072105407715,
+      "learning_rate": 0.00016940677202006063,
+      "loss": 2.8088,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2563361220472441,
+      "grad_norm": 1.0628341436386108,
+      "learning_rate": 0.00016939284936901615,
+      "loss": 2.6944,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2563976377952756,
+      "grad_norm": 1.1885006427764893,
+      "learning_rate": 0.00016937892412306605,
+      "loss": 2.784,
+      "step": 4168
+    },
+    {
+      "epoch": 0.25645915354330706,
+      "grad_norm": 1.102070927619934,
+      "learning_rate": 0.00016936499628273106,
+      "loss": 2.7155,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2565206692913386,
+      "grad_norm": 1.0509779453277588,
+      "learning_rate": 0.000169351065848532,
+      "loss": 2.9541,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2565821850393701,
+      "grad_norm": 1.0955532789230347,
+      "learning_rate": 0.00016933713282098976,
+      "loss": 3.5553,
+      "step": 4171
+    },
+    {
+      "epoch": 0.25664370078740156,
+      "grad_norm": 1.0051299333572388,
+      "learning_rate": 0.00016932319720062538,
+      "loss": 2.5855,
+      "step": 4172
+    },
+    {
+      "epoch": 0.25670521653543305,
+      "grad_norm": 1.4289250373840332,
+      "learning_rate": 0.00016930925898795999,
+      "loss": 2.699,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2567667322834646,
+      "grad_norm": 1.534692406654358,
+      "learning_rate": 0.00016929531818351483,
+      "loss": 2.5971,
+      "step": 4174
+    },
+    {
+      "epoch": 0.25682824803149606,
+      "grad_norm": 1.025661587715149,
+      "learning_rate": 0.00016928137478781116,
+      "loss": 2.9012,
+      "step": 4175
+    },
+    {
+      "epoch": 0.25688976377952755,
+      "grad_norm": 1.1760838031768799,
+      "learning_rate": 0.00016926742880137038,
+      "loss": 2.8726,
+      "step": 4176
+    },
+    {
+      "epoch": 0.25695127952755903,
+      "grad_norm": 1.0244145393371582,
+      "learning_rate": 0.000169253480224714,
+      "loss": 2.9053,
+      "step": 4177
+    },
+    {
+      "epoch": 0.25701279527559057,
+      "grad_norm": 1.207390308380127,
+      "learning_rate": 0.00016923952905836364,
+      "loss": 2.7299,
+      "step": 4178
+    },
+    {
+      "epoch": 0.25707431102362205,
+      "grad_norm": 1.08153235912323,
+      "learning_rate": 0.000169225575302841,
+      "loss": 2.9272,
+      "step": 4179
+    },
+    {
+      "epoch": 0.25713582677165353,
+      "grad_norm": 1.1758222579956055,
+      "learning_rate": 0.00016921161895866785,
+      "loss": 2.4416,
+      "step": 4180
+    },
+    {
+      "epoch": 0.257197342519685,
+      "grad_norm": 1.2819567918777466,
+      "learning_rate": 0.00016919766002636605,
+      "loss": 2.9011,
+      "step": 4181
+    },
+    {
+      "epoch": 0.25725885826771655,
+      "grad_norm": 1.5786445140838623,
+      "learning_rate": 0.00016918369850645767,
+      "loss": 2.8372,
+      "step": 4182
+    },
+    {
+      "epoch": 0.25732037401574803,
+      "grad_norm": 1.3524471521377563,
+      "learning_rate": 0.0001691697343994647,
+      "loss": 3.0666,
+      "step": 4183
+    },
+    {
+      "epoch": 0.2573818897637795,
+      "grad_norm": 1.0725582838058472,
+      "learning_rate": 0.0001691557677059094,
+      "loss": 2.7561,
+      "step": 4184
+    },
+    {
+      "epoch": 0.257443405511811,
+      "grad_norm": 1.1069633960723877,
+      "learning_rate": 0.000169141798426314,
+      "loss": 2.4945,
+      "step": 4185
+    },
+    {
+      "epoch": 0.25750492125984253,
+      "grad_norm": 1.1703448295593262,
+      "learning_rate": 0.0001691278265612009,
+      "loss": 2.1763,
+      "step": 4186
+    },
+    {
+      "epoch": 0.257566437007874,
+      "grad_norm": 1.1567227840423584,
+      "learning_rate": 0.00016911385211109252,
+      "loss": 2.8365,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2576279527559055,
+      "grad_norm": 1.201215386390686,
+      "learning_rate": 0.00016909987507651147,
+      "loss": 3.0719,
+      "step": 4188
+    },
+    {
+      "epoch": 0.257689468503937,
+      "grad_norm": 1.0969847440719604,
+      "learning_rate": 0.00016908589545798041,
+      "loss": 2.8523,
+      "step": 4189
+    },
+    {
+      "epoch": 0.2577509842519685,
+      "grad_norm": 1.2080692052841187,
+      "learning_rate": 0.00016907191325602212,
+      "loss": 2.4157,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2578125,
+      "grad_norm": 1.0881073474884033,
+      "learning_rate": 0.0001690579284711594,
+      "loss": 2.8645,
+      "step": 4191
+    },
+    {
+      "epoch": 0.2578740157480315,
+      "grad_norm": 1.159602165222168,
+      "learning_rate": 0.00016904394110391522,
+      "loss": 2.6913,
+      "step": 4192
+    },
+    {
+      "epoch": 0.257935531496063,
+      "grad_norm": 1.1168556213378906,
+      "learning_rate": 0.00016902995115481266,
+      "loss": 2.7854,
+      "step": 4193
+    },
+    {
+      "epoch": 0.2579970472440945,
+      "grad_norm": 1.1380579471588135,
+      "learning_rate": 0.00016901595862437485,
+      "loss": 2.423,
+      "step": 4194
+    },
+    {
+      "epoch": 0.258058562992126,
+      "grad_norm": 1.27704656124115,
+      "learning_rate": 0.00016900196351312503,
+      "loss": 3.0218,
+      "step": 4195
+    },
+    {
+      "epoch": 0.25812007874015747,
+      "grad_norm": 1.208048939704895,
+      "learning_rate": 0.00016898796582158653,
+      "loss": 2.6305,
+      "step": 4196
+    },
+    {
+      "epoch": 0.258181594488189,
+      "grad_norm": 1.1274224519729614,
+      "learning_rate": 0.0001689739655502828,
+      "loss": 1.5479,
+      "step": 4197
+    },
+    {
+      "epoch": 0.2582431102362205,
+      "grad_norm": 1.4853583574295044,
+      "learning_rate": 0.00016895996269973739,
+      "loss": 2.912,
+      "step": 4198
+    },
+    {
+      "epoch": 0.25830462598425197,
+      "grad_norm": 0.9844282269477844,
+      "learning_rate": 0.00016894595727047387,
+      "loss": 2.8621,
+      "step": 4199
+    },
+    {
+      "epoch": 0.25836614173228345,
+      "grad_norm": 1.0375475883483887,
+      "learning_rate": 0.000168931949263016,
+      "loss": 2.2904,
+      "step": 4200
+    },
+    {
+      "epoch": 0.258427657480315,
+      "grad_norm": 1.1193816661834717,
+      "learning_rate": 0.00016891793867788763,
+      "loss": 2.4221,
+      "step": 4201
+    },
+    {
+      "epoch": 0.25848917322834647,
+      "grad_norm": 1.268370509147644,
+      "learning_rate": 0.00016890392551561264,
+      "loss": 2.1428,
+      "step": 4202
+    },
+    {
+      "epoch": 0.25855068897637795,
+      "grad_norm": 1.2186672687530518,
+      "learning_rate": 0.00016888990977671507,
+      "loss": 2.7202,
+      "step": 4203
+    },
+    {
+      "epoch": 0.25861220472440943,
+      "grad_norm": 1.2408641576766968,
+      "learning_rate": 0.000168875891461719,
+      "loss": 2.9428,
+      "step": 4204
+    },
+    {
+      "epoch": 0.25867372047244097,
+      "grad_norm": 1.3038536310195923,
+      "learning_rate": 0.00016886187057114867,
+      "loss": 2.464,
+      "step": 4205
+    },
+    {
+      "epoch": 0.25873523622047245,
+      "grad_norm": 1.1026012897491455,
+      "learning_rate": 0.00016884784710552832,
+      "loss": 1.8477,
+      "step": 4206
+    },
+    {
+      "epoch": 0.25879675196850394,
+      "grad_norm": 1.4103397130966187,
+      "learning_rate": 0.00016883382106538245,
+      "loss": 2.9987,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2588582677165354,
+      "grad_norm": 1.0469348430633545,
+      "learning_rate": 0.00016881979245123545,
+      "loss": 3.1591,
+      "step": 4208
+    },
+    {
+      "epoch": 0.25891978346456695,
+      "grad_norm": 1.1734962463378906,
+      "learning_rate": 0.00016880576126361197,
+      "loss": 1.8181,
+      "step": 4209
+    },
+    {
+      "epoch": 0.25898129921259844,
+      "grad_norm": 1.6248393058776855,
+      "learning_rate": 0.0001687917275030367,
+      "loss": 3.3295,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2590428149606299,
+      "grad_norm": 1.1128031015396118,
+      "learning_rate": 0.00016877769117003445,
+      "loss": 2.5391,
+      "step": 4211
+    },
+    {
+      "epoch": 0.2591043307086614,
+      "grad_norm": 1.3814733028411865,
+      "learning_rate": 0.00016876365226513003,
+      "loss": 2.8154,
+      "step": 4212
+    },
+    {
+      "epoch": 0.25916584645669294,
+      "grad_norm": 1.262110948562622,
+      "learning_rate": 0.00016874961078884845,
+      "loss": 2.7804,
+      "step": 4213
+    },
+    {
+      "epoch": 0.2592273622047244,
+      "grad_norm": 1.5369398593902588,
+      "learning_rate": 0.0001687355667417148,
+      "loss": 2.5391,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2592888779527559,
+      "grad_norm": 1.0065503120422363,
+      "learning_rate": 0.00016872152012425423,
+      "loss": 2.5864,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2593503937007874,
+      "grad_norm": 1.245690941810608,
+      "learning_rate": 0.000168707470936992,
+      "loss": 2.697,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2594119094488189,
+      "grad_norm": 1.1505440473556519,
+      "learning_rate": 0.00016869341918045348,
+      "loss": 2.6874,
+      "step": 4217
+    },
+    {
+      "epoch": 0.2594734251968504,
+      "grad_norm": 1.515601634979248,
+      "learning_rate": 0.00016867936485516412,
+      "loss": 1.8135,
+      "step": 4218
+    },
+    {
+      "epoch": 0.2595349409448819,
+      "grad_norm": 1.2608680725097656,
+      "learning_rate": 0.00016866530796164953,
+      "loss": 2.8458,
+      "step": 4219
+    },
+    {
+      "epoch": 0.25959645669291337,
+      "grad_norm": 1.3908436298370361,
+      "learning_rate": 0.00016865124850043529,
+      "loss": 2.6237,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2596579724409449,
+      "grad_norm": 1.2010663747787476,
+      "learning_rate": 0.0001686371864720471,
+      "loss": 2.7704,
+      "step": 4221
+    },
+    {
+      "epoch": 0.2597194881889764,
+      "grad_norm": 1.0230482816696167,
+      "learning_rate": 0.00016862312187701094,
+      "loss": 2.5987,
+      "step": 4222
+    },
+    {
+      "epoch": 0.25978100393700787,
+      "grad_norm": 1.2878299951553345,
+      "learning_rate": 0.00016860905471585266,
+      "loss": 2.7278,
+      "step": 4223
+    },
+    {
+      "epoch": 0.25984251968503935,
+      "grad_norm": 1.0397971868515015,
+      "learning_rate": 0.00016859498498909828,
+      "loss": 3.175,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2599040354330709,
+      "grad_norm": 1.169396996498108,
+      "learning_rate": 0.00016858091269727397,
+      "loss": 2.0367,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2599655511811024,
+      "grad_norm": 1.3181240558624268,
+      "learning_rate": 0.00016856683784090596,
+      "loss": 2.3873,
+      "step": 4226
+    },
+    {
+      "epoch": 0.26002706692913385,
+      "grad_norm": 1.2954423427581787,
+      "learning_rate": 0.00016855276042052056,
+      "loss": 2.8128,
+      "step": 4227
+    },
+    {
+      "epoch": 0.26008858267716534,
+      "grad_norm": 1.1805468797683716,
+      "learning_rate": 0.00016853868043664415,
+      "loss": 2.3876,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2601500984251969,
+      "grad_norm": 1.2688809633255005,
+      "learning_rate": 0.00016852459788980331,
+      "loss": 2.6972,
+      "step": 4229
+    },
+    {
+      "epoch": 0.26021161417322836,
+      "grad_norm": 1.4530959129333496,
+      "learning_rate": 0.00016851051278052458,
+      "loss": 2.6095,
+      "step": 4230
+    },
+    {
+      "epoch": 0.26027312992125984,
+      "grad_norm": 1.108902931213379,
+      "learning_rate": 0.0001684964251093347,
+      "loss": 3.0759,
+      "step": 4231
+    },
+    {
+      "epoch": 0.2603346456692913,
+      "grad_norm": 0.9473598599433899,
+      "learning_rate": 0.00016848233487676048,
+      "loss": 2.8404,
+      "step": 4232
+    },
+    {
+      "epoch": 0.26039616141732286,
+      "grad_norm": 1.1496483087539673,
+      "learning_rate": 0.00016846824208332882,
+      "loss": 2.9499,
+      "step": 4233
+    },
+    {
+      "epoch": 0.26045767716535434,
+      "grad_norm": 1.1145013570785522,
+      "learning_rate": 0.00016845414672956666,
+      "loss": 2.6736,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2605191929133858,
+      "grad_norm": 0.999867856502533,
+      "learning_rate": 0.00016844004881600113,
+      "loss": 3.0143,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2605807086614173,
+      "grad_norm": 1.3063740730285645,
+      "learning_rate": 0.00016842594834315944,
+      "loss": 2.6365,
+      "step": 4236
+    },
+    {
+      "epoch": 0.26064222440944884,
+      "grad_norm": 1.183717131614685,
+      "learning_rate": 0.0001684118453115688,
+      "loss": 3.0665,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2607037401574803,
+      "grad_norm": 1.2420308589935303,
+      "learning_rate": 0.00016839773972175662,
+      "loss": 2.0174,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2607652559055118,
+      "grad_norm": 1.2182881832122803,
+      "learning_rate": 0.0001683836315742504,
+      "loss": 2.8208,
+      "step": 4239
+    },
+    {
+      "epoch": 0.2608267716535433,
+      "grad_norm": 0.9592204689979553,
+      "learning_rate": 0.00016836952086957764,
+      "loss": 3.0321,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2608882874015748,
+      "grad_norm": 1.0538135766983032,
+      "learning_rate": 0.00016835540760826607,
+      "loss": 3.0988,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2609498031496063,
+      "grad_norm": 1.1781244277954102,
+      "learning_rate": 0.00016834129179084338,
+      "loss": 2.9714,
+      "step": 4242
+    },
+    {
+      "epoch": 0.2610113188976378,
+      "grad_norm": 1.0579181909561157,
+      "learning_rate": 0.0001683271734178375,
+      "loss": 2.6347,
+      "step": 4243
+    },
+    {
+      "epoch": 0.26107283464566927,
+      "grad_norm": 0.9868262410163879,
+      "learning_rate": 0.00016831305248977632,
+      "loss": 1.5401,
+      "step": 4244
+    },
+    {
+      "epoch": 0.2611343503937008,
+      "grad_norm": 1.1695541143417358,
+      "learning_rate": 0.00016829892900718788,
+      "loss": 2.8166,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2611958661417323,
+      "grad_norm": 1.2437807321548462,
+      "learning_rate": 0.00016828480297060037,
+      "loss": 3.1538,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2612573818897638,
+      "grad_norm": 1.1678595542907715,
+      "learning_rate": 0.00016827067438054198,
+      "loss": 2.9103,
+      "step": 4247
+    },
+    {
+      "epoch": 0.26131889763779526,
+      "grad_norm": 0.9730347990989685,
+      "learning_rate": 0.00016825654323754106,
+      "loss": 2.9611,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2613804133858268,
+      "grad_norm": 1.2026481628417969,
+      "learning_rate": 0.00016824240954212605,
+      "loss": 2.9177,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2614419291338583,
+      "grad_norm": 1.2732776403427124,
+      "learning_rate": 0.0001682282732948254,
+      "loss": 2.2068,
+      "step": 4250
+    },
+    {
+      "epoch": 0.26150344488188976,
+      "grad_norm": 1.1544452905654907,
+      "learning_rate": 0.00016821413449616784,
+      "loss": 2.9825,
+      "step": 4251
+    },
+    {
+      "epoch": 0.26156496062992124,
+      "grad_norm": 1.1290838718414307,
+      "learning_rate": 0.000168199993146682,
+      "loss": 2.4718,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2616264763779528,
+      "grad_norm": 0.9685749411582947,
+      "learning_rate": 0.00016818584924689669,
+      "loss": 2.2344,
+      "step": 4253
+    },
+    {
+      "epoch": 0.26168799212598426,
+      "grad_norm": 1.2437050342559814,
+      "learning_rate": 0.00016817170279734086,
+      "loss": 2.5237,
+      "step": 4254
+    },
+    {
+      "epoch": 0.26174950787401574,
+      "grad_norm": 1.0967096090316772,
+      "learning_rate": 0.00016815755379854345,
+      "loss": 2.4158,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2618110236220472,
+      "grad_norm": 1.1440705060958862,
+      "learning_rate": 0.0001681434022510336,
+      "loss": 2.8422,
+      "step": 4256
+    },
+    {
+      "epoch": 0.26187253937007876,
+      "grad_norm": 0.9441343545913696,
+      "learning_rate": 0.00016812924815534053,
+      "loss": 2.8239,
+      "step": 4257
+    },
+    {
+      "epoch": 0.26193405511811024,
+      "grad_norm": 1.1715377569198608,
+      "learning_rate": 0.00016811509151199342,
+      "loss": 2.7592,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2619955708661417,
+      "grad_norm": 1.1698347330093384,
+      "learning_rate": 0.00016810093232152174,
+      "loss": 2.7459,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2620570866141732,
+      "grad_norm": 1.469275951385498,
+      "learning_rate": 0.00016808677058445493,
+      "loss": 2.8827,
+      "step": 4260
+    },
+    {
+      "epoch": 0.26211860236220474,
+      "grad_norm": 1.1307201385498047,
+      "learning_rate": 0.00016807260630132253,
+      "loss": 3.2763,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2621801181102362,
+      "grad_norm": 0.9475696086883545,
+      "learning_rate": 0.00016805843947265427,
+      "loss": 2.3013,
+      "step": 4262
+    },
+    {
+      "epoch": 0.2622416338582677,
+      "grad_norm": 1.4002058506011963,
+      "learning_rate": 0.00016804427009897986,
+      "loss": 3.2409,
+      "step": 4263
+    },
+    {
+      "epoch": 0.2623031496062992,
+      "grad_norm": 1.0544027090072632,
+      "learning_rate": 0.0001680300981808292,
+      "loss": 3.1587,
+      "step": 4264
+    },
+    {
+      "epoch": 0.26236466535433073,
+      "grad_norm": 1.1826729774475098,
+      "learning_rate": 0.00016801592371873218,
+      "loss": 2.36,
+      "step": 4265
+    },
+    {
+      "epoch": 0.2624261811023622,
+      "grad_norm": 1.293279767036438,
+      "learning_rate": 0.00016800174671321888,
+      "loss": 2.932,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2624876968503937,
+      "grad_norm": 0.9769501686096191,
+      "learning_rate": 0.00016798756716481945,
+      "loss": 2.7763,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2625492125984252,
+      "grad_norm": 1.0897001028060913,
+      "learning_rate": 0.00016797338507406413,
+      "loss": 3.0856,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2626107283464567,
+      "grad_norm": 1.1315609216690063,
+      "learning_rate": 0.00016795920044148326,
+      "loss": 3.2221,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2626722440944882,
+      "grad_norm": 1.3629577159881592,
+      "learning_rate": 0.00016794501326760722,
+      "loss": 3.1285,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2627337598425197,
+      "grad_norm": 1.4634761810302734,
+      "learning_rate": 0.00016793082355296652,
+      "loss": 2.6298,
+      "step": 4271
+    },
+    {
+      "epoch": 0.26279527559055116,
+      "grad_norm": 1.1661911010742188,
+      "learning_rate": 0.00016791663129809184,
+      "loss": 3.5633,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2628567913385827,
+      "grad_norm": 1.229202151298523,
+      "learning_rate": 0.00016790243650351388,
+      "loss": 2.3052,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2629183070866142,
+      "grad_norm": 1.0290642976760864,
+      "learning_rate": 0.0001678882391697634,
+      "loss": 3.1425,
+      "step": 4274
+    },
+    {
+      "epoch": 0.26297982283464566,
+      "grad_norm": 1.1298282146453857,
+      "learning_rate": 0.00016787403929737136,
+      "loss": 2.4351,
+      "step": 4275
+    },
+    {
+      "epoch": 0.26304133858267714,
+      "grad_norm": 1.1072921752929688,
+      "learning_rate": 0.00016785983688686872,
+      "loss": 2.7245,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2631028543307087,
+      "grad_norm": 1.2235804796218872,
+      "learning_rate": 0.00016784563193878655,
+      "loss": 2.5795,
+      "step": 4277
+    },
+    {
+      "epoch": 0.26316437007874016,
+      "grad_norm": 1.3396010398864746,
+      "learning_rate": 0.00016783142445365607,
+      "loss": 3.0338,
+      "step": 4278
+    },
+    {
+      "epoch": 0.26322588582677164,
+      "grad_norm": 1.0049115419387817,
+      "learning_rate": 0.00016781721443200856,
+      "loss": 3.199,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2632874015748031,
+      "grad_norm": 1.0840871334075928,
+      "learning_rate": 0.00016780300187437542,
+      "loss": 2.7325,
+      "step": 4280
+    },
+    {
+      "epoch": 0.26334891732283466,
+      "grad_norm": 1.271036982536316,
+      "learning_rate": 0.00016778878678128806,
+      "loss": 2.7851,
+      "step": 4281
+    },
+    {
+      "epoch": 0.26341043307086615,
+      "grad_norm": 1.3817417621612549,
+      "learning_rate": 0.00016777456915327807,
+      "loss": 2.6674,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2634719488188976,
+      "grad_norm": 1.1391136646270752,
+      "learning_rate": 0.00016776034899087714,
+      "loss": 3.031,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2635334645669291,
+      "grad_norm": 1.144614577293396,
+      "learning_rate": 0.000167746126294617,
+      "loss": 3.3151,
+      "step": 4284
+    },
+    {
+      "epoch": 0.26359498031496065,
+      "grad_norm": 1.2283201217651367,
+      "learning_rate": 0.00016773190106502946,
+      "loss": 2.6563,
+      "step": 4285
+    },
+    {
+      "epoch": 0.26365649606299213,
+      "grad_norm": 1.3877841234207153,
+      "learning_rate": 0.00016771767330264653,
+      "loss": 2.1989,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2637180118110236,
+      "grad_norm": 1.0959105491638184,
+      "learning_rate": 0.00016770344300800021,
+      "loss": 2.5562,
+      "step": 4287
+    },
+    {
+      "epoch": 0.2637795275590551,
+      "grad_norm": 1.2003686428070068,
+      "learning_rate": 0.00016768921018162268,
+      "loss": 3.1838,
+      "step": 4288
+    },
+    {
+      "epoch": 0.26384104330708663,
+      "grad_norm": 1.3038508892059326,
+      "learning_rate": 0.0001676749748240461,
+      "loss": 2.4693,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2639025590551181,
+      "grad_norm": 1.2474361658096313,
+      "learning_rate": 0.00016766073693580283,
+      "loss": 2.9957,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2639640748031496,
+      "grad_norm": 2.4690773487091064,
+      "learning_rate": 0.0001676464965174253,
+      "loss": 2.1884,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2640255905511811,
+      "grad_norm": 1.0141266584396362,
+      "learning_rate": 0.00016763225356944598,
+      "loss": 1.7334,
+      "step": 4292
+    },
+    {
+      "epoch": 0.2640871062992126,
+      "grad_norm": 1.1536662578582764,
+      "learning_rate": 0.00016761800809239757,
+      "loss": 2.5735,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2641486220472441,
+      "grad_norm": 0.9336086511611938,
+      "learning_rate": 0.00016760376008681264,
+      "loss": 2.9466,
+      "step": 4294
+    },
+    {
+      "epoch": 0.2642101377952756,
+      "grad_norm": 1.078698754310608,
+      "learning_rate": 0.00016758950955322408,
+      "loss": 3.0771,
+      "step": 4295
+    },
+    {
+      "epoch": 0.26427165354330706,
+      "grad_norm": 1.241478681564331,
+      "learning_rate": 0.00016757525649216474,
+      "loss": 3.0387,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2643331692913386,
+      "grad_norm": 0.8869643807411194,
+      "learning_rate": 0.00016756100090416763,
+      "loss": 2.9522,
+      "step": 4297
+    },
+    {
+      "epoch": 0.2643946850393701,
+      "grad_norm": 1.451230525970459,
+      "learning_rate": 0.0001675467427897658,
+      "loss": 2.3303,
+      "step": 4298
+    },
+    {
+      "epoch": 0.26445620078740156,
+      "grad_norm": 0.9746833443641663,
+      "learning_rate": 0.00016753248214949244,
+      "loss": 2.8661,
+      "step": 4299
+    },
+    {
+      "epoch": 0.26451771653543305,
+      "grad_norm": 1.2632886171340942,
+      "learning_rate": 0.00016751821898388082,
+      "loss": 2.5411,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2645792322834646,
+      "grad_norm": 1.2090383768081665,
+      "learning_rate": 0.00016750395329346432,
+      "loss": 2.9059,
+      "step": 4301
+    },
+    {
+      "epoch": 0.26464074803149606,
+      "grad_norm": 1.3085417747497559,
+      "learning_rate": 0.0001674896850787764,
+      "loss": 3.3719,
+      "step": 4302
+    },
+    {
+      "epoch": 0.26470226377952755,
+      "grad_norm": 1.1804871559143066,
+      "learning_rate": 0.00016747541434035051,
+      "loss": 3.1151,
+      "step": 4303
+    },
+    {
+      "epoch": 0.26476377952755903,
+      "grad_norm": 1.049797534942627,
+      "learning_rate": 0.00016746114107872044,
+      "loss": 2.4052,
+      "step": 4304
+    },
+    {
+      "epoch": 0.26482529527559057,
+      "grad_norm": 1.1012510061264038,
+      "learning_rate": 0.00016744686529441988,
+      "loss": 2.5089,
+      "step": 4305
+    },
+    {
+      "epoch": 0.26488681102362205,
+      "grad_norm": 1.0970401763916016,
+      "learning_rate": 0.0001674325869879826,
+      "loss": 2.7732,
+      "step": 4306
+    },
+    {
+      "epoch": 0.26494832677165353,
+      "grad_norm": 1.0483962297439575,
+      "learning_rate": 0.00016741830615994263,
+      "loss": 2.9166,
+      "step": 4307
+    },
+    {
+      "epoch": 0.265009842519685,
+      "grad_norm": 0.9129215478897095,
+      "learning_rate": 0.0001674040228108339,
+      "loss": 3.1278,
+      "step": 4308
+    },
+    {
+      "epoch": 0.26507135826771655,
+      "grad_norm": 1.2662101984024048,
+      "learning_rate": 0.0001673897369411906,
+      "loss": 2.9284,
+      "step": 4309
+    },
+    {
+      "epoch": 0.26513287401574803,
+      "grad_norm": 1.197811484336853,
+      "learning_rate": 0.0001673754485515469,
+      "loss": 3.045,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2651943897637795,
+      "grad_norm": 1.4913930892944336,
+      "learning_rate": 0.00016736115764243715,
+      "loss": 2.1947,
+      "step": 4311
+    },
+    {
+      "epoch": 0.265255905511811,
+      "grad_norm": 1.4967966079711914,
+      "learning_rate": 0.0001673468642143957,
+      "loss": 3.2884,
+      "step": 4312
+    },
+    {
+      "epoch": 0.26531742125984253,
+      "grad_norm": 1.0511850118637085,
+      "learning_rate": 0.00016733256826795704,
+      "loss": 2.9267,
+      "step": 4313
+    },
+    {
+      "epoch": 0.265378937007874,
+      "grad_norm": 1.04053795337677,
+      "learning_rate": 0.0001673182698036558,
+      "loss": 2.9214,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2654404527559055,
+      "grad_norm": 1.0982098579406738,
+      "learning_rate": 0.0001673039688220267,
+      "loss": 3.1689,
+      "step": 4315
+    },
+    {
+      "epoch": 0.265501968503937,
+      "grad_norm": 1.5090000629425049,
+      "learning_rate": 0.00016728966532360436,
+      "loss": 2.6719,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2655634842519685,
+      "grad_norm": 1.3309783935546875,
+      "learning_rate": 0.00016727535930892383,
+      "loss": 2.4914,
+      "step": 4317
+    },
+    {
+      "epoch": 0.265625,
+      "grad_norm": 1.2868901491165161,
+      "learning_rate": 0.00016726105077852,
+      "loss": 3.0166,
+      "step": 4318
+    },
+    {
+      "epoch": 0.2656865157480315,
+      "grad_norm": 1.2130858898162842,
+      "learning_rate": 0.0001672467397329279,
+      "loss": 3.068,
+      "step": 4319
+    },
+    {
+      "epoch": 0.265748031496063,
+      "grad_norm": 1.4112720489501953,
+      "learning_rate": 0.00016723242617268272,
+      "loss": 2.8793,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2658095472440945,
+      "grad_norm": 1.2182451486587524,
+      "learning_rate": 0.00016721811009831969,
+      "loss": 3.6496,
+      "step": 4321
+    },
+    {
+      "epoch": 0.265871062992126,
+      "grad_norm": 1.01311194896698,
+      "learning_rate": 0.00016720379151037416,
+      "loss": 3.0833,
+      "step": 4322
+    },
+    {
+      "epoch": 0.26593257874015747,
+      "grad_norm": 1.0822722911834717,
+      "learning_rate": 0.00016718947040938157,
+      "loss": 2.6409,
+      "step": 4323
+    },
+    {
+      "epoch": 0.265994094488189,
+      "grad_norm": 1.390677809715271,
+      "learning_rate": 0.0001671751467958775,
+      "loss": 2.9222,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2660556102362205,
+      "grad_norm": 1.1509617567062378,
+      "learning_rate": 0.00016716082067039743,
+      "loss": 2.9261,
+      "step": 4325
+    },
+    {
+      "epoch": 0.26611712598425197,
+      "grad_norm": 1.1809113025665283,
+      "learning_rate": 0.00016714649203347724,
+      "loss": 2.959,
+      "step": 4326
+    },
+    {
+      "epoch": 0.26617864173228345,
+      "grad_norm": 1.1221181154251099,
+      "learning_rate": 0.0001671321608856526,
+      "loss": 3.4771,
+      "step": 4327
+    },
+    {
+      "epoch": 0.266240157480315,
+      "grad_norm": 1.2904157638549805,
+      "learning_rate": 0.0001671178272274595,
+      "loss": 3.0477,
+      "step": 4328
+    },
+    {
+      "epoch": 0.26630167322834647,
+      "grad_norm": 1.2628248929977417,
+      "learning_rate": 0.00016710349105943396,
+      "loss": 2.7359,
+      "step": 4329
+    },
+    {
+      "epoch": 0.26636318897637795,
+      "grad_norm": 1.3969725370407104,
+      "learning_rate": 0.00016708915238211204,
+      "loss": 2.7909,
+      "step": 4330
+    },
+    {
+      "epoch": 0.26642470472440943,
+      "grad_norm": 1.0470685958862305,
+      "learning_rate": 0.00016707481119602988,
+      "loss": 2.7582,
+      "step": 4331
+    },
+    {
+      "epoch": 0.26648622047244097,
+      "grad_norm": 1.2803395986557007,
+      "learning_rate": 0.00016706046750172382,
+      "loss": 2.5376,
+      "step": 4332
+    },
+    {
+      "epoch": 0.26654773622047245,
+      "grad_norm": 1.0789825916290283,
+      "learning_rate": 0.0001670461212997302,
+      "loss": 2.6201,
+      "step": 4333
+    },
+    {
+      "epoch": 0.26660925196850394,
+      "grad_norm": 1.0347013473510742,
+      "learning_rate": 0.00016703177259058552,
+      "loss": 2.432,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2666707677165354,
+      "grad_norm": 1.2182813882827759,
+      "learning_rate": 0.0001670174213748263,
+      "loss": 2.274,
+      "step": 4335
+    },
+    {
+      "epoch": 0.26673228346456695,
+      "grad_norm": 1.7544316053390503,
+      "learning_rate": 0.00016700306765298924,
+      "loss": 2.3723,
+      "step": 4336
+    },
+    {
+      "epoch": 0.26679379921259844,
+      "grad_norm": 1.1529240608215332,
+      "learning_rate": 0.00016698871142561104,
+      "loss": 2.0731,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2668553149606299,
+      "grad_norm": 1.3767292499542236,
+      "learning_rate": 0.00016697435269322864,
+      "loss": 3.5007,
+      "step": 4338
+    },
+    {
+      "epoch": 0.2669168307086614,
+      "grad_norm": 1.219026803970337,
+      "learning_rate": 0.00016695999145637883,
+      "loss": 2.8877,
+      "step": 4339
+    },
+    {
+      "epoch": 0.26697834645669294,
+      "grad_norm": 1.1933655738830566,
+      "learning_rate": 0.00016694562771559876,
+      "loss": 2.9579,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2670398622047244,
+      "grad_norm": 1.0941102504730225,
+      "learning_rate": 0.0001669312614714255,
+      "loss": 2.6741,
+      "step": 4341
+    },
+    {
+      "epoch": 0.2671013779527559,
+      "grad_norm": 0.973625659942627,
+      "learning_rate": 0.0001669168927243963,
+      "loss": 3.1879,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2671628937007874,
+      "grad_norm": 1.0517559051513672,
+      "learning_rate": 0.00016690252147504844,
+      "loss": 3.6246,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2672244094488189,
+      "grad_norm": 1.114965558052063,
+      "learning_rate": 0.00016688814772391932,
+      "loss": 3.4565,
+      "step": 4344
+    },
+    {
+      "epoch": 0.2672859251968504,
+      "grad_norm": 1.0055075883865356,
+      "learning_rate": 0.00016687377147154648,
+      "loss": 2.956,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2673474409448819,
+      "grad_norm": 1.2515794038772583,
+      "learning_rate": 0.00016685939271846744,
+      "loss": 2.4219,
+      "step": 4346
+    },
+    {
+      "epoch": 0.26740895669291337,
+      "grad_norm": 1.1612004041671753,
+      "learning_rate": 0.00016684501146522,
+      "loss": 2.9927,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2674704724409449,
+      "grad_norm": 0.9721747040748596,
+      "learning_rate": 0.0001668306277123418,
+      "loss": 2.9751,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2675319881889764,
+      "grad_norm": 1.3007668256759644,
+      "learning_rate": 0.0001668162414603708,
+      "loss": 2.4699,
+      "step": 4349
+    },
+    {
+      "epoch": 0.26759350393700787,
+      "grad_norm": 1.056334137916565,
+      "learning_rate": 0.00016680185270984496,
+      "loss": 3.07,
+      "step": 4350
+    },
+    {
+      "epoch": 0.26765501968503935,
+      "grad_norm": 1.1383159160614014,
+      "learning_rate": 0.00016678746146130236,
+      "loss": 2.7071,
+      "step": 4351
+    },
+    {
+      "epoch": 0.2677165354330709,
+      "grad_norm": 0.9643104076385498,
+      "learning_rate": 0.00016677306771528105,
+      "loss": 2.9753,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2677780511811024,
+      "grad_norm": 1.3651759624481201,
+      "learning_rate": 0.0001667586714723194,
+      "loss": 2.8592,
+      "step": 4353
+    },
+    {
+      "epoch": 0.26783956692913385,
+      "grad_norm": 1.0184972286224365,
+      "learning_rate": 0.00016674427273295566,
+      "loss": 2.5898,
+      "step": 4354
+    },
+    {
+      "epoch": 0.26790108267716534,
+      "grad_norm": 1.5176093578338623,
+      "learning_rate": 0.0001667298714977283,
+      "loss": 2.6635,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2679625984251969,
+      "grad_norm": 2.1054625511169434,
+      "learning_rate": 0.00016671546776717583,
+      "loss": 2.3847,
+      "step": 4356
+    },
+    {
+      "epoch": 0.26802411417322836,
+      "grad_norm": 1.2859820127487183,
+      "learning_rate": 0.00016670106154183692,
+      "loss": 2.6519,
+      "step": 4357
+    },
+    {
+      "epoch": 0.26808562992125984,
+      "grad_norm": 1.0323145389556885,
+      "learning_rate": 0.00016668665282225023,
+      "loss": 3.3892,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2681471456692913,
+      "grad_norm": 1.0289911031723022,
+      "learning_rate": 0.00016667224160895456,
+      "loss": 2.814,
+      "step": 4359
+    },
+    {
+      "epoch": 0.26820866141732286,
+      "grad_norm": 1.2119157314300537,
+      "learning_rate": 0.00016665782790248885,
+      "loss": 3.048,
+      "step": 4360
+    },
+    {
+      "epoch": 0.26827017716535434,
+      "grad_norm": 1.4164531230926514,
+      "learning_rate": 0.00016664341170339206,
+      "loss": 2.71,
+      "step": 4361
+    },
+    {
+      "epoch": 0.2683316929133858,
+      "grad_norm": 0.9363687038421631,
+      "learning_rate": 0.00016662899301220327,
+      "loss": 2.272,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2683932086614173,
+      "grad_norm": 1.2547837495803833,
+      "learning_rate": 0.0001666145718294617,
+      "loss": 2.2085,
+      "step": 4363
+    },
+    {
+      "epoch": 0.26845472440944884,
+      "grad_norm": 1.1407623291015625,
+      "learning_rate": 0.0001666001481557066,
+      "loss": 2.3436,
+      "step": 4364
+    },
+    {
+      "epoch": 0.2685162401574803,
+      "grad_norm": 1.1112265586853027,
+      "learning_rate": 0.0001665857219914773,
+      "loss": 2.77,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2685777559055118,
+      "grad_norm": 1.094480276107788,
+      "learning_rate": 0.00016657129333731333,
+      "loss": 3.5103,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2686392716535433,
+      "grad_norm": 1.1037111282348633,
+      "learning_rate": 0.00016655686219375419,
+      "loss": 2.3685,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2687007874015748,
+      "grad_norm": 1.1919771432876587,
+      "learning_rate": 0.00016654242856133954,
+      "loss": 2.5578,
+      "step": 4368
+    },
+    {
+      "epoch": 0.2687623031496063,
+      "grad_norm": 1.0067487955093384,
+      "learning_rate": 0.00016652799244060908,
+      "loss": 3.1748,
+      "step": 4369
+    },
+    {
+      "epoch": 0.2688238188976378,
+      "grad_norm": 1.7239720821380615,
+      "learning_rate": 0.0001665135538321027,
+      "loss": 2.3719,
+      "step": 4370
+    },
+    {
+      "epoch": 0.26888533464566927,
+      "grad_norm": 1.7628542184829712,
+      "learning_rate": 0.0001664991127363603,
+      "loss": 2.6014,
+      "step": 4371
+    },
+    {
+      "epoch": 0.2689468503937008,
+      "grad_norm": 1.2281956672668457,
+      "learning_rate": 0.00016648466915392186,
+      "loss": 3.3253,
+      "step": 4372
+    },
+    {
+      "epoch": 0.2690083661417323,
+      "grad_norm": 1.0600703954696655,
+      "learning_rate": 0.00016647022308532755,
+      "loss": 3.038,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2690698818897638,
+      "grad_norm": 0.9593676924705505,
+      "learning_rate": 0.00016645577453111752,
+      "loss": 3.2476,
+      "step": 4374
+    },
+    {
+      "epoch": 0.26913139763779526,
+      "grad_norm": 1.3075501918792725,
+      "learning_rate": 0.00016644132349183215,
+      "loss": 2.8968,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2691929133858268,
+      "grad_norm": 1.0307234525680542,
+      "learning_rate": 0.0001664268699680117,
+      "loss": 2.7389,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2692544291338583,
+      "grad_norm": 0.9870452880859375,
+      "learning_rate": 0.00016641241396019675,
+      "loss": 3.0819,
+      "step": 4377
+    },
+    {
+      "epoch": 0.26931594488188976,
+      "grad_norm": 1.1197131872177124,
+      "learning_rate": 0.00016639795546892784,
+      "loss": 2.8782,
+      "step": 4378
+    },
+    {
+      "epoch": 0.26937746062992124,
+      "grad_norm": 1.0679792165756226,
+      "learning_rate": 0.00016638349449474564,
+      "loss": 2.7416,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2694389763779528,
+      "grad_norm": 1.2656769752502441,
+      "learning_rate": 0.00016636903103819093,
+      "loss": 3.0495,
+      "step": 4380
+    },
+    {
+      "epoch": 0.26950049212598426,
+      "grad_norm": 1.963358998298645,
+      "learning_rate": 0.00016635456509980453,
+      "loss": 3.5128,
+      "step": 4381
+    },
+    {
+      "epoch": 0.26956200787401574,
+      "grad_norm": 1.4801812171936035,
+      "learning_rate": 0.00016634009668012737,
+      "loss": 3.7267,
+      "step": 4382
+    },
+    {
+      "epoch": 0.2696235236220472,
+      "grad_norm": 1.2391738891601562,
+      "learning_rate": 0.00016632562577970054,
+      "loss": 2.7804,
+      "step": 4383
+    },
+    {
+      "epoch": 0.26968503937007876,
+      "grad_norm": 1.0176039934158325,
+      "learning_rate": 0.00016631115239906517,
+      "loss": 3.0435,
+      "step": 4384
+    },
+    {
+      "epoch": 0.26974655511811024,
+      "grad_norm": 1.0624140501022339,
+      "learning_rate": 0.00016629667653876243,
+      "loss": 3.2476,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2698080708661417,
+      "grad_norm": 1.5827122926712036,
+      "learning_rate": 0.00016628219819933367,
+      "loss": 2.89,
+      "step": 4386
+    },
+    {
+      "epoch": 0.2698695866141732,
+      "grad_norm": 1.267661690711975,
+      "learning_rate": 0.0001662677173813203,
+      "loss": 3.5029,
+      "step": 4387
+    },
+    {
+      "epoch": 0.26993110236220474,
+      "grad_norm": 1.3135602474212646,
+      "learning_rate": 0.00016625323408526384,
+      "loss": 2.6122,
+      "step": 4388
+    },
+    {
+      "epoch": 0.2699926181102362,
+      "grad_norm": 1.2642086744308472,
+      "learning_rate": 0.00016623874831170582,
+      "loss": 3.2586,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2700541338582677,
+      "grad_norm": 1.1053721904754639,
+      "learning_rate": 0.000166224260061188,
+      "loss": 2.3247,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2701156496062992,
+      "grad_norm": 1.1537970304489136,
+      "learning_rate": 0.00016620976933425215,
+      "loss": 2.6414,
+      "step": 4391
+    },
+    {
+      "epoch": 0.27017716535433073,
+      "grad_norm": 1.130031704902649,
+      "learning_rate": 0.0001661952761314401,
+      "loss": 2.5913,
+      "step": 4392
+    },
+    {
+      "epoch": 0.2702386811023622,
+      "grad_norm": 0.9636685848236084,
+      "learning_rate": 0.00016618078045329383,
+      "loss": 2.8091,
+      "step": 4393
+    },
+    {
+      "epoch": 0.2703001968503937,
+      "grad_norm": 1.24986732006073,
+      "learning_rate": 0.0001661662823003554,
+      "loss": 2.8126,
+      "step": 4394
+    },
+    {
+      "epoch": 0.2703617125984252,
+      "grad_norm": 0.9704762101173401,
+      "learning_rate": 0.000166151781673167,
+      "loss": 3.0355,
+      "step": 4395
+    },
+    {
+      "epoch": 0.2704232283464567,
+      "grad_norm": 1.1305038928985596,
+      "learning_rate": 0.00016613727857227079,
+      "loss": 2.7475,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2704847440944882,
+      "grad_norm": 1.039721131324768,
+      "learning_rate": 0.00016612277299820916,
+      "loss": 3.068,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2705462598425197,
+      "grad_norm": 1.0172123908996582,
+      "learning_rate": 0.00016610826495152455,
+      "loss": 3.0235,
+      "step": 4398
+    },
+    {
+      "epoch": 0.27060777559055116,
+      "grad_norm": 1.576098918914795,
+      "learning_rate": 0.00016609375443275948,
+      "loss": 2.5338,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2706692913385827,
+      "grad_norm": 0.9993608593940735,
+      "learning_rate": 0.00016607924144245652,
+      "loss": 3.0464,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2707308070866142,
+      "grad_norm": 1.0318354368209839,
+      "learning_rate": 0.0001660647259811584,
+      "loss": 3.4258,
+      "step": 4401
+    },
+    {
+      "epoch": 0.27079232283464566,
+      "grad_norm": 0.9997342824935913,
+      "learning_rate": 0.0001660502080494079,
+      "loss": 2.7907,
+      "step": 4402
+    },
+    {
+      "epoch": 0.27085383858267714,
+      "grad_norm": 1.4035478830337524,
+      "learning_rate": 0.00016603568764774797,
+      "loss": 2.6067,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2709153543307087,
+      "grad_norm": 1.2556425333023071,
+      "learning_rate": 0.0001660211647767215,
+      "loss": 4.5834,
+      "step": 4404
+    },
+    {
+      "epoch": 0.27097687007874016,
+      "grad_norm": 1.0163151025772095,
+      "learning_rate": 0.00016600663943687165,
+      "loss": 2.8649,
+      "step": 4405
+    },
+    {
+      "epoch": 0.27103838582677164,
+      "grad_norm": 0.9867238998413086,
+      "learning_rate": 0.00016599211162874152,
+      "loss": 3.0434,
+      "step": 4406
+    },
+    {
+      "epoch": 0.2710999015748031,
+      "grad_norm": 1.1090370416641235,
+      "learning_rate": 0.0001659775813528744,
+      "loss": 3.1042,
+      "step": 4407
+    },
+    {
+      "epoch": 0.27116141732283466,
+      "grad_norm": 1.202957034111023,
+      "learning_rate": 0.00016596304860981364,
+      "loss": 2.8044,
+      "step": 4408
+    },
+    {
+      "epoch": 0.27122293307086615,
+      "grad_norm": 1.19243586063385,
+      "learning_rate": 0.0001659485134001027,
+      "loss": 2.9444,
+      "step": 4409
+    },
+    {
+      "epoch": 0.2712844488188976,
+      "grad_norm": 1.229990005493164,
+      "learning_rate": 0.00016593397572428509,
+      "loss": 3.2949,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2713459645669291,
+      "grad_norm": 1.1516857147216797,
+      "learning_rate": 0.00016591943558290443,
+      "loss": 3.0932,
+      "step": 4411
+    },
+    {
+      "epoch": 0.27140748031496065,
+      "grad_norm": 1.0897278785705566,
+      "learning_rate": 0.00016590489297650446,
+      "loss": 3.62,
+      "step": 4412
+    },
+    {
+      "epoch": 0.27146899606299213,
+      "grad_norm": 1.3829576969146729,
+      "learning_rate": 0.00016589034790562897,
+      "loss": 2.134,
+      "step": 4413
+    },
+    {
+      "epoch": 0.2715305118110236,
+      "grad_norm": 1.2478686571121216,
+      "learning_rate": 0.00016587580037082193,
+      "loss": 3.234,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2715920275590551,
+      "grad_norm": 1.1638952493667603,
+      "learning_rate": 0.00016586125037262728,
+      "loss": 2.7239,
+      "step": 4415
+    },
+    {
+      "epoch": 0.27165354330708663,
+      "grad_norm": 1.2081092596054077,
+      "learning_rate": 0.0001658466979115891,
+      "loss": 2.4881,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2717150590551181,
+      "grad_norm": 1.3394947052001953,
+      "learning_rate": 0.0001658321429882516,
+      "loss": 2.5164,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2717765748031496,
+      "grad_norm": 1.3511383533477783,
+      "learning_rate": 0.00016581758560315902,
+      "loss": 3.4515,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2718380905511811,
+      "grad_norm": 1.1088223457336426,
+      "learning_rate": 0.00016580302575685578,
+      "loss": 2.6243,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2718996062992126,
+      "grad_norm": 1.18809974193573,
+      "learning_rate": 0.0001657884634498863,
+      "loss": 2.626,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2719611220472441,
+      "grad_norm": 1.339214563369751,
+      "learning_rate": 0.0001657738986827951,
+      "loss": 2.473,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2720226377952756,
+      "grad_norm": 1.1495537757873535,
+      "learning_rate": 0.00016575933145612686,
+      "loss": 3.0957,
+      "step": 4422
+    },
+    {
+      "epoch": 0.27208415354330706,
+      "grad_norm": 1.4753822088241577,
+      "learning_rate": 0.00016574476177042634,
+      "loss": 2.6314,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2721456692913386,
+      "grad_norm": 0.9901444911956787,
+      "learning_rate": 0.0001657301896262383,
+      "loss": 3.0543,
+      "step": 4424
+    },
+    {
+      "epoch": 0.2722071850393701,
+      "grad_norm": 1.0908039808273315,
+      "learning_rate": 0.00016571561502410771,
+      "loss": 2.9341,
+      "step": 4425
+    },
+    {
+      "epoch": 0.27226870078740156,
+      "grad_norm": 1.4802002906799316,
+      "learning_rate": 0.00016570103796457956,
+      "loss": 3.0314,
+      "step": 4426
+    },
+    {
+      "epoch": 0.27233021653543305,
+      "grad_norm": 1.219825029373169,
+      "learning_rate": 0.00016568645844819896,
+      "loss": 2.8791,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2723917322834646,
+      "grad_norm": 1.3126660585403442,
+      "learning_rate": 0.00016567187647551106,
+      "loss": 3.0014,
+      "step": 4428
+    },
+    {
+      "epoch": 0.27245324803149606,
+      "grad_norm": 1.1890817880630493,
+      "learning_rate": 0.0001656572920470612,
+      "loss": 2.5184,
+      "step": 4429
+    },
+    {
+      "epoch": 0.27251476377952755,
+      "grad_norm": 1.2670537233352661,
+      "learning_rate": 0.00016564270516339474,
+      "loss": 2.4825,
+      "step": 4430
+    },
+    {
+      "epoch": 0.27257627952755903,
+      "grad_norm": 1.3378400802612305,
+      "learning_rate": 0.0001656281158250571,
+      "loss": 2.351,
+      "step": 4431
+    },
+    {
+      "epoch": 0.27263779527559057,
+      "grad_norm": 1.2143644094467163,
+      "learning_rate": 0.0001656135240325939,
+      "loss": 3.2811,
+      "step": 4432
+    },
+    {
+      "epoch": 0.27269931102362205,
+      "grad_norm": 1.3588284254074097,
+      "learning_rate": 0.0001655989297865508,
+      "loss": 1.9924,
+      "step": 4433
+    },
+    {
+      "epoch": 0.27276082677165353,
+      "grad_norm": 1.497705340385437,
+      "learning_rate": 0.0001655843330874735,
+      "loss": 2.6679,
+      "step": 4434
+    },
+    {
+      "epoch": 0.272822342519685,
+      "grad_norm": 1.4287819862365723,
+      "learning_rate": 0.00016556973393590783,
+      "loss": 3.0193,
+      "step": 4435
+    },
+    {
+      "epoch": 0.27288385826771655,
+      "grad_norm": 1.3193975687026978,
+      "learning_rate": 0.00016555513233239974,
+      "loss": 2.9191,
+      "step": 4436
+    },
+    {
+      "epoch": 0.27294537401574803,
+      "grad_norm": 1.2715003490447998,
+      "learning_rate": 0.00016554052827749525,
+      "loss": 3.0399,
+      "step": 4437
+    },
+    {
+      "epoch": 0.2730068897637795,
+      "grad_norm": 1.385724425315857,
+      "learning_rate": 0.00016552592177174048,
+      "loss": 1.7499,
+      "step": 4438
+    },
+    {
+      "epoch": 0.273068405511811,
+      "grad_norm": 1.0676435232162476,
+      "learning_rate": 0.0001655113128156816,
+      "loss": 2.9519,
+      "step": 4439
+    },
+    {
+      "epoch": 0.27312992125984253,
+      "grad_norm": 1.1752756834030151,
+      "learning_rate": 0.00016549670140986495,
+      "loss": 2.8452,
+      "step": 4440
+    },
+    {
+      "epoch": 0.273191437007874,
+      "grad_norm": 1.0755422115325928,
+      "learning_rate": 0.0001654820875548369,
+      "loss": 2.655,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2732529527559055,
+      "grad_norm": 1.2339533567428589,
+      "learning_rate": 0.00016546747125114388,
+      "loss": 2.8496,
+      "step": 4442
+    },
+    {
+      "epoch": 0.273314468503937,
+      "grad_norm": 1.2926950454711914,
+      "learning_rate": 0.00016545285249933248,
+      "loss": 2.7654,
+      "step": 4443
+    },
+    {
+      "epoch": 0.2733759842519685,
+      "grad_norm": 1.195510983467102,
+      "learning_rate": 0.0001654382312999494,
+      "loss": 2.6252,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2734375,
+      "grad_norm": 1.1251013278961182,
+      "learning_rate": 0.00016542360765354135,
+      "loss": 2.7177,
+      "step": 4445
+    },
+    {
+      "epoch": 0.2734990157480315,
+      "grad_norm": 1.1626002788543701,
+      "learning_rate": 0.00016540898156065518,
+      "loss": 2.3911,
+      "step": 4446
+    },
+    {
+      "epoch": 0.273560531496063,
+      "grad_norm": 1.202102541923523,
+      "learning_rate": 0.00016539435302183783,
+      "loss": 2.5454,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2736220472440945,
+      "grad_norm": 1.527787685394287,
+      "learning_rate": 0.00016537972203763634,
+      "loss": 2.5319,
+      "step": 4448
+    },
+    {
+      "epoch": 0.273683562992126,
+      "grad_norm": 1.3849385976791382,
+      "learning_rate": 0.0001653650886085978,
+      "loss": 2.7823,
+      "step": 4449
+    },
+    {
+      "epoch": 0.27374507874015747,
+      "grad_norm": 1.2906588315963745,
+      "learning_rate": 0.00016535045273526945,
+      "loss": 2.4783,
+      "step": 4450
+    },
+    {
+      "epoch": 0.273806594488189,
+      "grad_norm": 1.1255592107772827,
+      "learning_rate": 0.00016533581441819853,
+      "loss": 3.5526,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2738681102362205,
+      "grad_norm": 1.0256321430206299,
+      "learning_rate": 0.00016532117365793253,
+      "loss": 2.8854,
+      "step": 4452
+    },
+    {
+      "epoch": 0.27392962598425197,
+      "grad_norm": 1.4777430295944214,
+      "learning_rate": 0.00016530653045501885,
+      "loss": 2.6567,
+      "step": 4453
+    },
+    {
+      "epoch": 0.27399114173228345,
+      "grad_norm": 1.0188243389129639,
+      "learning_rate": 0.00016529188481000508,
+      "loss": 2.9988,
+      "step": 4454
+    },
+    {
+      "epoch": 0.274052657480315,
+      "grad_norm": 1.338597059249878,
+      "learning_rate": 0.00016527723672343892,
+      "loss": 2.789,
+      "step": 4455
+    },
+    {
+      "epoch": 0.27411417322834647,
+      "grad_norm": 1.4032208919525146,
+      "learning_rate": 0.0001652625861958681,
+      "loss": 2.5,
+      "step": 4456
+    },
+    {
+      "epoch": 0.27417568897637795,
+      "grad_norm": 1.1649895906448364,
+      "learning_rate": 0.00016524793322784045,
+      "loss": 2.0308,
+      "step": 4457
+    },
+    {
+      "epoch": 0.27423720472440943,
+      "grad_norm": 1.2288763523101807,
+      "learning_rate": 0.00016523327781990395,
+      "loss": 3.2408,
+      "step": 4458
+    },
+    {
+      "epoch": 0.27429872047244097,
+      "grad_norm": 1.0959136486053467,
+      "learning_rate": 0.00016521861997260658,
+      "loss": 2.9078,
+      "step": 4459
+    },
+    {
+      "epoch": 0.27436023622047245,
+      "grad_norm": 1.0447324514389038,
+      "learning_rate": 0.0001652039596864965,
+      "loss": 3.0571,
+      "step": 4460
+    },
+    {
+      "epoch": 0.27442175196850394,
+      "grad_norm": 1.3763411045074463,
+      "learning_rate": 0.00016518929696212193,
+      "loss": 2.4179,
+      "step": 4461
+    },
+    {
+      "epoch": 0.2744832677165354,
+      "grad_norm": 1.1027061939239502,
+      "learning_rate": 0.00016517463180003113,
+      "loss": 2.5133,
+      "step": 4462
+    },
+    {
+      "epoch": 0.27454478346456695,
+      "grad_norm": 1.658150315284729,
+      "learning_rate": 0.00016515996420077256,
+      "loss": 2.8574,
+      "step": 4463
+    },
+    {
+      "epoch": 0.27460629921259844,
+      "grad_norm": 1.0638750791549683,
+      "learning_rate": 0.00016514529416489468,
+      "loss": 2.8053,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2746678149606299,
+      "grad_norm": 1.101629614830017,
+      "learning_rate": 0.00016513062169294602,
+      "loss": 3.2794,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2747293307086614,
+      "grad_norm": 1.1691670417785645,
+      "learning_rate": 0.00016511594678547532,
+      "loss": 2.9127,
+      "step": 4466
+    },
+    {
+      "epoch": 0.27479084645669294,
+      "grad_norm": 1.222704291343689,
+      "learning_rate": 0.00016510126944303127,
+      "loss": 1.926,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2748523622047244,
+      "grad_norm": 1.2800092697143555,
+      "learning_rate": 0.0001650865896661628,
+      "loss": 2.9975,
+      "step": 4468
+    },
+    {
+      "epoch": 0.2749138779527559,
+      "grad_norm": 1.044110655784607,
+      "learning_rate": 0.00016507190745541878,
+      "loss": 3.5124,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2749753937007874,
+      "grad_norm": 1.1445931196212769,
+      "learning_rate": 0.00016505722281134829,
+      "loss": 2.7124,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2750369094488189,
+      "grad_norm": 1.0354903936386108,
+      "learning_rate": 0.00016504253573450042,
+      "loss": 2.6591,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2750984251968504,
+      "grad_norm": 1.2743451595306396,
+      "learning_rate": 0.0001650278462254244,
+      "loss": 2.7941,
+      "step": 4472
+    },
+    {
+      "epoch": 0.2751599409448819,
+      "grad_norm": 1.153326153755188,
+      "learning_rate": 0.00016501315428466953,
+      "loss": 2.4768,
+      "step": 4473
+    },
+    {
+      "epoch": 0.27522145669291337,
+      "grad_norm": 1.4138609170913696,
+      "learning_rate": 0.00016499845991278523,
+      "loss": 2.5891,
+      "step": 4474
+    },
+    {
+      "epoch": 0.2752829724409449,
+      "grad_norm": 1.1395156383514404,
+      "learning_rate": 0.00016498376311032095,
+      "loss": 2.9593,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2753444881889764,
+      "grad_norm": 1.085139513015747,
+      "learning_rate": 0.00016496906387782634,
+      "loss": 3.44,
+      "step": 4476
+    },
+    {
+      "epoch": 0.27540600393700787,
+      "grad_norm": 1.1558607816696167,
+      "learning_rate": 0.000164954362215851,
+      "loss": 2.8723,
+      "step": 4477
+    },
+    {
+      "epoch": 0.27546751968503935,
+      "grad_norm": 1.2521228790283203,
+      "learning_rate": 0.00016493965812494467,
+      "loss": 2.7088,
+      "step": 4478
+    },
+    {
+      "epoch": 0.2755290354330709,
+      "grad_norm": 1.0921251773834229,
+      "learning_rate": 0.00016492495160565725,
+      "loss": 3.194,
+      "step": 4479
+    },
+    {
+      "epoch": 0.2755905511811024,
+      "grad_norm": 1.2646875381469727,
+      "learning_rate": 0.0001649102426585387,
+      "loss": 2.1945,
+      "step": 4480
+    },
+    {
+      "epoch": 0.27565206692913385,
+      "grad_norm": 1.3195098638534546,
+      "learning_rate": 0.00016489553128413902,
+      "loss": 3.5787,
+      "step": 4481
+    },
+    {
+      "epoch": 0.27571358267716534,
+      "grad_norm": 1.0435289144515991,
+      "learning_rate": 0.00016488081748300832,
+      "loss": 2.9228,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2757750984251969,
+      "grad_norm": 1.1351228952407837,
+      "learning_rate": 0.00016486610125569684,
+      "loss": 3.6468,
+      "step": 4483
+    },
+    {
+      "epoch": 0.27583661417322836,
+      "grad_norm": 0.9771934151649475,
+      "learning_rate": 0.00016485138260275486,
+      "loss": 2.9958,
+      "step": 4484
+    },
+    {
+      "epoch": 0.27589812992125984,
+      "grad_norm": 1.2808003425598145,
+      "learning_rate": 0.0001648366615247328,
+      "loss": 2.9261,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2759596456692913,
+      "grad_norm": 1.171127438545227,
+      "learning_rate": 0.00016482193802218114,
+      "loss": 2.5269,
+      "step": 4486
+    },
+    {
+      "epoch": 0.27602116141732286,
+      "grad_norm": 1.3161617517471313,
+      "learning_rate": 0.00016480721209565047,
+      "loss": 3.2274,
+      "step": 4487
+    },
+    {
+      "epoch": 0.27608267716535434,
+      "grad_norm": 1.146040916442871,
+      "learning_rate": 0.00016479248374569143,
+      "loss": 2.8736,
+      "step": 4488
+    },
+    {
+      "epoch": 0.2761441929133858,
+      "grad_norm": 1.4730421304702759,
+      "learning_rate": 0.0001647777529728548,
+      "loss": 2.6809,
+      "step": 4489
+    },
+    {
+      "epoch": 0.2762057086614173,
+      "grad_norm": 1.2807053327560425,
+      "learning_rate": 0.0001647630197776914,
+      "loss": 2.6754,
+      "step": 4490
+    },
+    {
+      "epoch": 0.27626722440944884,
+      "grad_norm": 1.0262514352798462,
+      "learning_rate": 0.0001647482841607522,
+      "loss": 2.456,
+      "step": 4491
+    },
+    {
+      "epoch": 0.2763287401574803,
+      "grad_norm": 1.075419545173645,
+      "learning_rate": 0.0001647335461225882,
+      "loss": 2.7862,
+      "step": 4492
+    },
+    {
+      "epoch": 0.2763902559055118,
+      "grad_norm": 1.0450094938278198,
+      "learning_rate": 0.00016471880566375056,
+      "loss": 3.1193,
+      "step": 4493
+    },
+    {
+      "epoch": 0.2764517716535433,
+      "grad_norm": 1.0800143480300903,
+      "learning_rate": 0.0001647040627847905,
+      "loss": 2.3655,
+      "step": 4494
+    },
+    {
+      "epoch": 0.2765132874015748,
+      "grad_norm": 1.0844542980194092,
+      "learning_rate": 0.00016468931748625924,
+      "loss": 3.2615,
+      "step": 4495
+    },
+    {
+      "epoch": 0.2765748031496063,
+      "grad_norm": 1.0831248760223389,
+      "learning_rate": 0.00016467456976870825,
+      "loss": 3.2333,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2766363188976378,
+      "grad_norm": 1.1633784770965576,
+      "learning_rate": 0.00016465981963268897,
+      "loss": 2.5751,
+      "step": 4497
+    },
+    {
+      "epoch": 0.27669783464566927,
+      "grad_norm": 1.1727222204208374,
+      "learning_rate": 0.000164645067078753,
+      "loss": 2.6393,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2767593503937008,
+      "grad_norm": 1.2071046829223633,
+      "learning_rate": 0.000164630312107452,
+      "loss": 2.9013,
+      "step": 4499
+    },
+    {
+      "epoch": 0.2768208661417323,
+      "grad_norm": 1.006325364112854,
+      "learning_rate": 0.00016461555471933767,
+      "loss": 2.6291,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2768823818897638,
+      "grad_norm": 0.9079983234405518,
+      "learning_rate": 0.00016460079491496194,
+      "loss": 2.6646,
+      "step": 4501
+    },
+    {
+      "epoch": 0.27694389763779526,
+      "grad_norm": 1.1979910135269165,
+      "learning_rate": 0.0001645860326948767,
+      "loss": 2.8501,
+      "step": 4502
+    },
+    {
+      "epoch": 0.2770054133858268,
+      "grad_norm": 1.342537760734558,
+      "learning_rate": 0.000164571268059634,
+      "loss": 3.021,
+      "step": 4503
+    },
+    {
+      "epoch": 0.2770669291338583,
+      "grad_norm": 1.039851188659668,
+      "learning_rate": 0.0001645565010097859,
+      "loss": 1.6222,
+      "step": 4504
+    },
+    {
+      "epoch": 0.27712844488188976,
+      "grad_norm": 1.4182703495025635,
+      "learning_rate": 0.00016454173154588464,
+      "loss": 2.143,
+      "step": 4505
+    },
+    {
+      "epoch": 0.27718996062992124,
+      "grad_norm": 1.1518765687942505,
+      "learning_rate": 0.00016452695966848256,
+      "loss": 2.9782,
+      "step": 4506
+    },
+    {
+      "epoch": 0.2772514763779528,
+      "grad_norm": 1.325704574584961,
+      "learning_rate": 0.00016451218537813197,
+      "loss": 2.6153,
+      "step": 4507
+    },
+    {
+      "epoch": 0.27731299212598426,
+      "grad_norm": 1.1088083982467651,
+      "learning_rate": 0.00016449740867538536,
+      "loss": 2.5065,
+      "step": 4508
+    },
+    {
+      "epoch": 0.27737450787401574,
+      "grad_norm": 1.0423314571380615,
+      "learning_rate": 0.0001644826295607953,
+      "loss": 2.7411,
+      "step": 4509
+    },
+    {
+      "epoch": 0.2774360236220472,
+      "grad_norm": 1.2622854709625244,
+      "learning_rate": 0.0001644678480349145,
+      "loss": 2.6502,
+      "step": 4510
+    },
+    {
+      "epoch": 0.27749753937007876,
+      "grad_norm": 0.9780692458152771,
+      "learning_rate": 0.00016445306409829566,
+      "loss": 3.2316,
+      "step": 4511
+    },
+    {
+      "epoch": 0.27755905511811024,
+      "grad_norm": 1.548874020576477,
+      "learning_rate": 0.00016443827775149158,
+      "loss": 2.5855,
+      "step": 4512
+    },
+    {
+      "epoch": 0.2776205708661417,
+      "grad_norm": 1.0133916139602661,
+      "learning_rate": 0.0001644234889950553,
+      "loss": 2.7544,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2776820866141732,
+      "grad_norm": 1.0868786573410034,
+      "learning_rate": 0.00016440869782953973,
+      "loss": 3.18,
+      "step": 4514
+    },
+    {
+      "epoch": 0.27774360236220474,
+      "grad_norm": 1.1225837469100952,
+      "learning_rate": 0.00016439390425549802,
+      "loss": 2.9284,
+      "step": 4515
+    },
+    {
+      "epoch": 0.2778051181102362,
+      "grad_norm": 1.0489013195037842,
+      "learning_rate": 0.00016437910827348334,
+      "loss": 2.6984,
+      "step": 4516
+    },
+    {
+      "epoch": 0.2778666338582677,
+      "grad_norm": 1.2729225158691406,
+      "learning_rate": 0.00016436430988404902,
+      "loss": 2.9792,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2779281496062992,
+      "grad_norm": 1.3959559202194214,
+      "learning_rate": 0.0001643495090877484,
+      "loss": 3.1693,
+      "step": 4518
+    },
+    {
+      "epoch": 0.27798966535433073,
+      "grad_norm": 0.9962127804756165,
+      "learning_rate": 0.000164334705885135,
+      "loss": 3.1211,
+      "step": 4519
+    },
+    {
+      "epoch": 0.2780511811023622,
+      "grad_norm": 1.0748634338378906,
+      "learning_rate": 0.0001643199002767623,
+      "loss": 2.085,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2781126968503937,
+      "grad_norm": 1.7125742435455322,
+      "learning_rate": 0.000164305092263184,
+      "loss": 1.8246,
+      "step": 4521
+    },
+    {
+      "epoch": 0.2781742125984252,
+      "grad_norm": 1.4049538373947144,
+      "learning_rate": 0.0001642902818449538,
+      "loss": 2.1135,
+      "step": 4522
+    },
+    {
+      "epoch": 0.2782357283464567,
+      "grad_norm": 1.230069875717163,
+      "learning_rate": 0.00016427546902262558,
+      "loss": 3.0872,
+      "step": 4523
+    },
+    {
+      "epoch": 0.2782972440944882,
+      "grad_norm": 1.1722221374511719,
+      "learning_rate": 0.00016426065379675323,
+      "loss": 2.4821,
+      "step": 4524
+    },
+    {
+      "epoch": 0.2783587598425197,
+      "grad_norm": 1.6638189554214478,
+      "learning_rate": 0.00016424583616789074,
+      "loss": 2.259,
+      "step": 4525
+    },
+    {
+      "epoch": 0.27842027559055116,
+      "grad_norm": 1.3233976364135742,
+      "learning_rate": 0.00016423101613659223,
+      "loss": 2.5269,
+      "step": 4526
+    },
+    {
+      "epoch": 0.2784817913385827,
+      "grad_norm": 1.5197041034698486,
+      "learning_rate": 0.0001642161937034119,
+      "loss": 2.2376,
+      "step": 4527
+    },
+    {
+      "epoch": 0.2785433070866142,
+      "grad_norm": 1.2002145051956177,
+      "learning_rate": 0.00016420136886890398,
+      "loss": 2.653,
+      "step": 4528
+    },
+    {
+      "epoch": 0.27860482283464566,
+      "grad_norm": 1.411747694015503,
+      "learning_rate": 0.00016418654163362286,
+      "loss": 2.7034,
+      "step": 4529
+    },
+    {
+      "epoch": 0.27866633858267714,
+      "grad_norm": 1.154043197631836,
+      "learning_rate": 0.00016417171199812301,
+      "loss": 2.4905,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2787278543307087,
+      "grad_norm": 1.0752205848693848,
+      "learning_rate": 0.00016415687996295894,
+      "loss": 2.9668,
+      "step": 4531
+    },
+    {
+      "epoch": 0.27878937007874016,
+      "grad_norm": 1.539167046546936,
+      "learning_rate": 0.00016414204552868533,
+      "loss": 2.0525,
+      "step": 4532
+    },
+    {
+      "epoch": 0.27885088582677164,
+      "grad_norm": 1.5718109607696533,
+      "learning_rate": 0.00016412720869585687,
+      "loss": 3.5094,
+      "step": 4533
+    },
+    {
+      "epoch": 0.2789124015748031,
+      "grad_norm": 1.356446385383606,
+      "learning_rate": 0.0001641123694650284,
+      "loss": 2.3523,
+      "step": 4534
+    },
+    {
+      "epoch": 0.27897391732283466,
+      "grad_norm": 1.1378636360168457,
+      "learning_rate": 0.0001640975278367548,
+      "loss": 2.7998,
+      "step": 4535
+    },
+    {
+      "epoch": 0.27903543307086615,
+      "grad_norm": 1.027816891670227,
+      "learning_rate": 0.00016408268381159108,
+      "loss": 3.5094,
+      "step": 4536
+    },
+    {
+      "epoch": 0.2790969488188976,
+      "grad_norm": 1.0989842414855957,
+      "learning_rate": 0.00016406783739009233,
+      "loss": 2.7049,
+      "step": 4537
+    },
+    {
+      "epoch": 0.2791584645669291,
+      "grad_norm": 1.1159745454788208,
+      "learning_rate": 0.0001640529885728137,
+      "loss": 2.8021,
+      "step": 4538
+    },
+    {
+      "epoch": 0.27921998031496065,
+      "grad_norm": 1.7120733261108398,
+      "learning_rate": 0.00016403813736031047,
+      "loss": 1.9252,
+      "step": 4539
+    },
+    {
+      "epoch": 0.27928149606299213,
+      "grad_norm": 1.04603111743927,
+      "learning_rate": 0.000164023283753138,
+      "loss": 1.7774,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2793430118110236,
+      "grad_norm": 1.139522910118103,
+      "learning_rate": 0.0001640084277518517,
+      "loss": 2.4915,
+      "step": 4541
+    },
+    {
+      "epoch": 0.2794045275590551,
+      "grad_norm": 1.1881873607635498,
+      "learning_rate": 0.00016399356935700715,
+      "loss": 2.7546,
+      "step": 4542
+    },
+    {
+      "epoch": 0.27946604330708663,
+      "grad_norm": 1.2111927270889282,
+      "learning_rate": 0.00016397870856915992,
+      "loss": 3.0754,
+      "step": 4543
+    },
+    {
+      "epoch": 0.2795275590551181,
+      "grad_norm": 1.1269322633743286,
+      "learning_rate": 0.00016396384538886575,
+      "loss": 2.9626,
+      "step": 4544
+    },
+    {
+      "epoch": 0.2795890748031496,
+      "grad_norm": 1.03859281539917,
+      "learning_rate": 0.00016394897981668045,
+      "loss": 2.8322,
+      "step": 4545
+    },
+    {
+      "epoch": 0.2796505905511811,
+      "grad_norm": 1.547787070274353,
+      "learning_rate": 0.00016393411185315987,
+      "loss": 2.262,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2797121062992126,
+      "grad_norm": 1.0490694046020508,
+      "learning_rate": 0.00016391924149886005,
+      "loss": 2.341,
+      "step": 4547
+    },
+    {
+      "epoch": 0.2797736220472441,
+      "grad_norm": 1.1219727993011475,
+      "learning_rate": 0.000163904368754337,
+      "loss": 2.2245,
+      "step": 4548
+    },
+    {
+      "epoch": 0.2798351377952756,
+      "grad_norm": 0.9666674733161926,
+      "learning_rate": 0.0001638894936201469,
+      "loss": 3.33,
+      "step": 4549
+    },
+    {
+      "epoch": 0.27989665354330706,
+      "grad_norm": 1.2354060411453247,
+      "learning_rate": 0.00016387461609684598,
+      "loss": 2.6879,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2799581692913386,
+      "grad_norm": 1.1452245712280273,
+      "learning_rate": 0.00016385973618499063,
+      "loss": 2.4078,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2800196850393701,
+      "grad_norm": 1.2627849578857422,
+      "learning_rate": 0.0001638448538851372,
+      "loss": 3.0456,
+      "step": 4552
+    },
+    {
+      "epoch": 0.28008120078740156,
+      "grad_norm": 1.2693103551864624,
+      "learning_rate": 0.00016382996919784225,
+      "loss": 2.5456,
+      "step": 4553
+    },
+    {
+      "epoch": 0.28014271653543305,
+      "grad_norm": 1.1190710067749023,
+      "learning_rate": 0.00016381508212366238,
+      "loss": 2.4344,
+      "step": 4554
+    },
+    {
+      "epoch": 0.2802042322834646,
+      "grad_norm": 1.0561959743499756,
+      "learning_rate": 0.0001638001926631543,
+      "loss": 3.3093,
+      "step": 4555
+    },
+    {
+      "epoch": 0.28026574803149606,
+      "grad_norm": 1.270707607269287,
+      "learning_rate": 0.00016378530081687478,
+      "loss": 3.1431,
+      "step": 4556
+    },
+    {
+      "epoch": 0.28032726377952755,
+      "grad_norm": 1.1912306547164917,
+      "learning_rate": 0.00016377040658538066,
+      "loss": 2.6842,
+      "step": 4557
+    },
+    {
+      "epoch": 0.28038877952755903,
+      "grad_norm": 1.3071626424789429,
+      "learning_rate": 0.00016375550996922892,
+      "loss": 2.8028,
+      "step": 4558
+    },
+    {
+      "epoch": 0.28045029527559057,
+      "grad_norm": 1.1712042093276978,
+      "learning_rate": 0.0001637406109689766,
+      "loss": 2.4101,
+      "step": 4559
+    },
+    {
+      "epoch": 0.28051181102362205,
+      "grad_norm": 1.0922691822052002,
+      "learning_rate": 0.00016372570958518093,
+      "loss": 3.0107,
+      "step": 4560
+    },
+    {
+      "epoch": 0.28057332677165353,
+      "grad_norm": 1.3452138900756836,
+      "learning_rate": 0.000163710805818399,
+      "loss": 3.0176,
+      "step": 4561
+    },
+    {
+      "epoch": 0.280634842519685,
+      "grad_norm": 1.3135766983032227,
+      "learning_rate": 0.00016369589966918824,
+      "loss": 2.5235,
+      "step": 4562
+    },
+    {
+      "epoch": 0.28069635826771655,
+      "grad_norm": 1.3672027587890625,
+      "learning_rate": 0.000163680991138106,
+      "loss": 2.2632,
+      "step": 4563
+    },
+    {
+      "epoch": 0.28075787401574803,
+      "grad_norm": 1.015019178390503,
+      "learning_rate": 0.00016366608022570976,
+      "loss": 2.629,
+      "step": 4564
+    },
+    {
+      "epoch": 0.2808193897637795,
+      "grad_norm": 1.1128273010253906,
+      "learning_rate": 0.00016365116693255714,
+      "loss": 3.3842,
+      "step": 4565
+    },
+    {
+      "epoch": 0.280880905511811,
+      "grad_norm": 1.322969675064087,
+      "learning_rate": 0.00016363625125920584,
+      "loss": 2.8292,
+      "step": 4566
+    },
+    {
+      "epoch": 0.28094242125984253,
+      "grad_norm": 1.1543601751327515,
+      "learning_rate": 0.00016362133320621354,
+      "loss": 3.0243,
+      "step": 4567
+    },
+    {
+      "epoch": 0.281003937007874,
+      "grad_norm": 1.0476384162902832,
+      "learning_rate": 0.00016360641277413817,
+      "loss": 2.5236,
+      "step": 4568
+    },
+    {
+      "epoch": 0.2810654527559055,
+      "grad_norm": 1.1128840446472168,
+      "learning_rate": 0.0001635914899635376,
+      "loss": 2.6385,
+      "step": 4569
+    },
+    {
+      "epoch": 0.281126968503937,
+      "grad_norm": 1.4447579383850098,
+      "learning_rate": 0.00016357656477496995,
+      "loss": 3.0849,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2811884842519685,
+      "grad_norm": 1.206144094467163,
+      "learning_rate": 0.00016356163720899328,
+      "loss": 2.4484,
+      "step": 4571
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 1.193233847618103,
+      "learning_rate": 0.00016354670726616578,
+      "loss": 2.1972,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2813115157480315,
+      "grad_norm": 1.1809132099151611,
+      "learning_rate": 0.00016353177494704581,
+      "loss": 2.711,
+      "step": 4573
+    },
+    {
+      "epoch": 0.281373031496063,
+      "grad_norm": 1.2280769348144531,
+      "learning_rate": 0.0001635168402521917,
+      "loss": 2.7763,
+      "step": 4574
+    },
+    {
+      "epoch": 0.2814345472440945,
+      "grad_norm": 1.4669212102890015,
+      "learning_rate": 0.00016350190318216193,
+      "loss": 2.269,
+      "step": 4575
+    },
+    {
+      "epoch": 0.281496062992126,
+      "grad_norm": 1.0834475755691528,
+      "learning_rate": 0.0001634869637375151,
+      "loss": 2.8497,
+      "step": 4576
+    },
+    {
+      "epoch": 0.28155757874015747,
+      "grad_norm": 1.0257607698440552,
+      "learning_rate": 0.00016347202191880985,
+      "loss": 2.8272,
+      "step": 4577
+    },
+    {
+      "epoch": 0.281619094488189,
+      "grad_norm": 1.4242370128631592,
+      "learning_rate": 0.0001634570777266049,
+      "loss": 3.3335,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2816806102362205,
+      "grad_norm": 1.177608609199524,
+      "learning_rate": 0.00016344213116145908,
+      "loss": 2.8864,
+      "step": 4579
+    },
+    {
+      "epoch": 0.28174212598425197,
+      "grad_norm": 1.9446598291397095,
+      "learning_rate": 0.00016342718222393132,
+      "loss": 2.6582,
+      "step": 4580
+    },
+    {
+      "epoch": 0.28180364173228345,
+      "grad_norm": 1.6401206254959106,
+      "learning_rate": 0.00016341223091458066,
+      "loss": 2.495,
+      "step": 4581
+    },
+    {
+      "epoch": 0.281865157480315,
+      "grad_norm": 1.2942825555801392,
+      "learning_rate": 0.00016339727723396608,
+      "loss": 2.1865,
+      "step": 4582
+    },
+    {
+      "epoch": 0.28192667322834647,
+      "grad_norm": 1.1726598739624023,
+      "learning_rate": 0.00016338232118264691,
+      "loss": 3.0893,
+      "step": 4583
+    },
+    {
+      "epoch": 0.28198818897637795,
+      "grad_norm": 1.2026127576828003,
+      "learning_rate": 0.00016336736276118235,
+      "loss": 2.2636,
+      "step": 4584
+    },
+    {
+      "epoch": 0.28204970472440943,
+      "grad_norm": 1.1527416706085205,
+      "learning_rate": 0.00016335240197013176,
+      "loss": 2.972,
+      "step": 4585
+    },
+    {
+      "epoch": 0.28211122047244097,
+      "grad_norm": 1.2873998880386353,
+      "learning_rate": 0.0001633374388100546,
+      "loss": 3.0162,
+      "step": 4586
+    },
+    {
+      "epoch": 0.28217273622047245,
+      "grad_norm": 0.9792348742485046,
+      "learning_rate": 0.0001633224732815104,
+      "loss": 2.966,
+      "step": 4587
+    },
+    {
+      "epoch": 0.28223425196850394,
+      "grad_norm": 1.1724910736083984,
+      "learning_rate": 0.0001633075053850588,
+      "loss": 2.5313,
+      "step": 4588
+    },
+    {
+      "epoch": 0.2822957677165354,
+      "grad_norm": 1.4435616731643677,
+      "learning_rate": 0.0001632925351212595,
+      "loss": 2.8292,
+      "step": 4589
+    },
+    {
+      "epoch": 0.28235728346456695,
+      "grad_norm": 1.6179641485214233,
+      "learning_rate": 0.00016327756249067234,
+      "loss": 2.3736,
+      "step": 4590
+    },
+    {
+      "epoch": 0.28241879921259844,
+      "grad_norm": 1.10062837600708,
+      "learning_rate": 0.00016326258749385718,
+      "loss": 2.9097,
+      "step": 4591
+    },
+    {
+      "epoch": 0.2824803149606299,
+      "grad_norm": 1.0056565999984741,
+      "learning_rate": 0.000163247610131374,
+      "loss": 3.2367,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2825418307086614,
+      "grad_norm": 1.291487455368042,
+      "learning_rate": 0.00016323263040378288,
+      "loss": 2.9828,
+      "step": 4593
+    },
+    {
+      "epoch": 0.28260334645669294,
+      "grad_norm": 1.1175187826156616,
+      "learning_rate": 0.00016321764831164396,
+      "loss": 3.2164,
+      "step": 4594
+    },
+    {
+      "epoch": 0.2826648622047244,
+      "grad_norm": 1.1013495922088623,
+      "learning_rate": 0.00016320266385551756,
+      "loss": 2.6171,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2827263779527559,
+      "grad_norm": 1.3635337352752686,
+      "learning_rate": 0.00016318767703596388,
+      "loss": 2.6397,
+      "step": 4596
+    },
+    {
+      "epoch": 0.2827878937007874,
+      "grad_norm": 1.1789071559906006,
+      "learning_rate": 0.0001631726878535435,
+      "loss": 2.8927,
+      "step": 4597
+    },
+    {
+      "epoch": 0.2828494094488189,
+      "grad_norm": 1.4106770753860474,
+      "learning_rate": 0.00016315769630881682,
+      "loss": 2.918,
+      "step": 4598
+    },
+    {
+      "epoch": 0.2829109251968504,
+      "grad_norm": 1.039588451385498,
+      "learning_rate": 0.00016314270240234445,
+      "loss": 3.0961,
+      "step": 4599
+    },
+    {
+      "epoch": 0.2829724409448819,
+      "grad_norm": 1.3607542514801025,
+      "learning_rate": 0.00016312770613468715,
+      "loss": 3.3129,
+      "step": 4600
+    },
+    {
+      "epoch": 0.28303395669291337,
+      "grad_norm": 1.1620734930038452,
+      "learning_rate": 0.00016311270750640563,
+      "loss": 2.4924,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2830954724409449,
+      "grad_norm": 1.1298425197601318,
+      "learning_rate": 0.00016309770651806079,
+      "loss": 3.2435,
+      "step": 4602
+    },
+    {
+      "epoch": 0.2831569881889764,
+      "grad_norm": 1.14798104763031,
+      "learning_rate": 0.00016308270317021354,
+      "loss": 3.1978,
+      "step": 4603
+    },
+    {
+      "epoch": 0.28321850393700787,
+      "grad_norm": 1.0764223337173462,
+      "learning_rate": 0.00016306769746342496,
+      "loss": 2.2716,
+      "step": 4604
+    },
+    {
+      "epoch": 0.28328001968503935,
+      "grad_norm": 1.0640770196914673,
+      "learning_rate": 0.00016305268939825618,
+      "loss": 2.5294,
+      "step": 4605
+    },
+    {
+      "epoch": 0.2833415354330709,
+      "grad_norm": 1.1964821815490723,
+      "learning_rate": 0.00016303767897526842,
+      "loss": 3.0617,
+      "step": 4606
+    },
+    {
+      "epoch": 0.2834030511811024,
+      "grad_norm": 1.2096500396728516,
+      "learning_rate": 0.00016302266619502292,
+      "loss": 2.283,
+      "step": 4607
+    },
+    {
+      "epoch": 0.28346456692913385,
+      "grad_norm": 1.1367918252944946,
+      "learning_rate": 0.0001630076510580812,
+      "loss": 2.6847,
+      "step": 4608
+    },
+    {
+      "epoch": 0.28352608267716534,
+      "grad_norm": 1.1303249597549438,
+      "learning_rate": 0.00016299263356500462,
+      "loss": 2.5583,
+      "step": 4609
+    },
+    {
+      "epoch": 0.2835875984251969,
+      "grad_norm": 1.2402222156524658,
+      "learning_rate": 0.00016297761371635483,
+      "loss": 2.2933,
+      "step": 4610
+    },
+    {
+      "epoch": 0.28364911417322836,
+      "grad_norm": 1.1847805976867676,
+      "learning_rate": 0.00016296259151269346,
+      "loss": 3.0955,
+      "step": 4611
+    },
+    {
+      "epoch": 0.28371062992125984,
+      "grad_norm": 1.2331749200820923,
+      "learning_rate": 0.00016294756695458222,
+      "loss": 2.512,
+      "step": 4612
+    },
+    {
+      "epoch": 0.2837721456692913,
+      "grad_norm": 1.1272627115249634,
+      "learning_rate": 0.000162932540042583,
+      "loss": 3.0267,
+      "step": 4613
+    },
+    {
+      "epoch": 0.28383366141732286,
+      "grad_norm": 1.2179044485092163,
+      "learning_rate": 0.00016291751077725773,
+      "loss": 2.722,
+      "step": 4614
+    },
+    {
+      "epoch": 0.28389517716535434,
+      "grad_norm": 1.0872946977615356,
+      "learning_rate": 0.00016290247915916838,
+      "loss": 2.7377,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2839566929133858,
+      "grad_norm": 1.247704267501831,
+      "learning_rate": 0.00016288744518887707,
+      "loss": 2.1537,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2840182086614173,
+      "grad_norm": 1.1106090545654297,
+      "learning_rate": 0.00016287240886694594,
+      "loss": 2.8288,
+      "step": 4617
+    },
+    {
+      "epoch": 0.28407972440944884,
+      "grad_norm": 1.1931582689285278,
+      "learning_rate": 0.00016285737019393733,
+      "loss": 2.5643,
+      "step": 4618
+    },
+    {
+      "epoch": 0.2841412401574803,
+      "grad_norm": 1.120021939277649,
+      "learning_rate": 0.00016284232917041357,
+      "loss": 2.4943,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2842027559055118,
+      "grad_norm": 1.2722420692443848,
+      "learning_rate": 0.00016282728579693713,
+      "loss": 2.4806,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2842642716535433,
+      "grad_norm": 1.0532110929489136,
+      "learning_rate": 0.00016281224007407054,
+      "loss": 3.0696,
+      "step": 4621
+    },
+    {
+      "epoch": 0.2843257874015748,
+      "grad_norm": 1.4109197854995728,
+      "learning_rate": 0.0001627971920023764,
+      "loss": 2.4117,
+      "step": 4622
+    },
+    {
+      "epoch": 0.2843873031496063,
+      "grad_norm": 1.916097640991211,
+      "learning_rate": 0.0001627821415824175,
+      "loss": 2.7785,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2844488188976378,
+      "grad_norm": 1.2830438613891602,
+      "learning_rate": 0.00016276708881475654,
+      "loss": 2.9282,
+      "step": 4624
+    },
+    {
+      "epoch": 0.28451033464566927,
+      "grad_norm": 0.9449663758277893,
+      "learning_rate": 0.00016275203369995646,
+      "loss": 3.1546,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2845718503937008,
+      "grad_norm": 1.4671975374221802,
+      "learning_rate": 0.00016273697623858023,
+      "loss": 2.2357,
+      "step": 4626
+    },
+    {
+      "epoch": 0.2846333661417323,
+      "grad_norm": 1.0886623859405518,
+      "learning_rate": 0.00016272191643119092,
+      "loss": 3.4477,
+      "step": 4627
+    },
+    {
+      "epoch": 0.2846948818897638,
+      "grad_norm": 1.087132215499878,
+      "learning_rate": 0.0001627068542783517,
+      "loss": 2.3603,
+      "step": 4628
+    },
+    {
+      "epoch": 0.28475639763779526,
+      "grad_norm": 1.7753138542175293,
+      "learning_rate": 0.0001626917897806258,
+      "loss": 2.0794,
+      "step": 4629
+    },
+    {
+      "epoch": 0.2848179133858268,
+      "grad_norm": 1.2319945096969604,
+      "learning_rate": 0.00016267672293857651,
+      "loss": 2.1138,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2848794291338583,
+      "grad_norm": 1.490250587463379,
+      "learning_rate": 0.0001626616537527673,
+      "loss": 2.7501,
+      "step": 4631
+    },
+    {
+      "epoch": 0.28494094488188976,
+      "grad_norm": 1.5039869546890259,
+      "learning_rate": 0.00016264658222376162,
+      "loss": 2.2682,
+      "step": 4632
+    },
+    {
+      "epoch": 0.28500246062992124,
+      "grad_norm": 1.2158503532409668,
+      "learning_rate": 0.00016263150835212314,
+      "loss": 2.7259,
+      "step": 4633
+    },
+    {
+      "epoch": 0.2850639763779528,
+      "grad_norm": 1.5854463577270508,
+      "learning_rate": 0.00016261643213841547,
+      "loss": 2.7086,
+      "step": 4634
+    },
+    {
+      "epoch": 0.28512549212598426,
+      "grad_norm": 1.1674830913543701,
+      "learning_rate": 0.0001626013535832024,
+      "loss": 2.9809,
+      "step": 4635
+    },
+    {
+      "epoch": 0.28518700787401574,
+      "grad_norm": 1.125846028327942,
+      "learning_rate": 0.00016258627268704777,
+      "loss": 3.3042,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2852485236220472,
+      "grad_norm": 0.9975308775901794,
+      "learning_rate": 0.00016257118945051552,
+      "loss": 2.9799,
+      "step": 4637
+    },
+    {
+      "epoch": 0.28531003937007876,
+      "grad_norm": 1.3956637382507324,
+      "learning_rate": 0.00016255610387416972,
+      "loss": 2.3276,
+      "step": 4638
+    },
+    {
+      "epoch": 0.28537155511811024,
+      "grad_norm": 1.1341049671173096,
+      "learning_rate": 0.00016254101595857443,
+      "loss": 3.2819,
+      "step": 4639
+    },
+    {
+      "epoch": 0.2854330708661417,
+      "grad_norm": 1.380044937133789,
+      "learning_rate": 0.00016252592570429388,
+      "loss": 3.1161,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2854945866141732,
+      "grad_norm": 1.285287618637085,
+      "learning_rate": 0.0001625108331118924,
+      "loss": 2.3044,
+      "step": 4641
+    },
+    {
+      "epoch": 0.28555610236220474,
+      "grad_norm": 1.094084620475769,
+      "learning_rate": 0.0001624957381819343,
+      "loss": 2.7824,
+      "step": 4642
+    },
+    {
+      "epoch": 0.2856176181102362,
+      "grad_norm": 1.209990382194519,
+      "learning_rate": 0.00016248064091498407,
+      "loss": 3.4525,
+      "step": 4643
+    },
+    {
+      "epoch": 0.2856791338582677,
+      "grad_norm": 1.0968564748764038,
+      "learning_rate": 0.0001624655413116063,
+      "loss": 3.1152,
+      "step": 4644
+    },
+    {
+      "epoch": 0.2857406496062992,
+      "grad_norm": 1.2321933507919312,
+      "learning_rate": 0.0001624504393723656,
+      "loss": 2.8979,
+      "step": 4645
+    },
+    {
+      "epoch": 0.28580216535433073,
+      "grad_norm": 1.3019940853118896,
+      "learning_rate": 0.00016243533509782668,
+      "loss": 2.5255,
+      "step": 4646
+    },
+    {
+      "epoch": 0.2858636811023622,
+      "grad_norm": 1.028449535369873,
+      "learning_rate": 0.0001624202284885544,
+      "loss": 3.0285,
+      "step": 4647
+    },
+    {
+      "epoch": 0.2859251968503937,
+      "grad_norm": 1.0399342775344849,
+      "learning_rate": 0.00016240511954511365,
+      "loss": 3.2138,
+      "step": 4648
+    },
+    {
+      "epoch": 0.2859867125984252,
+      "grad_norm": 1.9973114728927612,
+      "learning_rate": 0.00016239000826806936,
+      "loss": 2.4403,
+      "step": 4649
+    },
+    {
+      "epoch": 0.2860482283464567,
+      "grad_norm": 1.0806597471237183,
+      "learning_rate": 0.00016237489465798668,
+      "loss": 2.9748,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2861097440944882,
+      "grad_norm": 1.3599380254745483,
+      "learning_rate": 0.00016235977871543076,
+      "loss": 2.9253,
+      "step": 4651
+    },
+    {
+      "epoch": 0.2861712598425197,
+      "grad_norm": 1.350771427154541,
+      "learning_rate": 0.00016234466044096684,
+      "loss": 2.8231,
+      "step": 4652
+    },
+    {
+      "epoch": 0.28623277559055116,
+      "grad_norm": 1.117749571800232,
+      "learning_rate": 0.00016232953983516025,
+      "loss": 2.4457,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2862942913385827,
+      "grad_norm": 1.0279841423034668,
+      "learning_rate": 0.00016231441689857644,
+      "loss": 2.6744,
+      "step": 4654
+    },
+    {
+      "epoch": 0.2863558070866142,
+      "grad_norm": 1.0949769020080566,
+      "learning_rate": 0.00016229929163178092,
+      "loss": 2.7503,
+      "step": 4655
+    },
+    {
+      "epoch": 0.28641732283464566,
+      "grad_norm": 1.3003346920013428,
+      "learning_rate": 0.00016228416403533927,
+      "loss": 3.1031,
+      "step": 4656
+    },
+    {
+      "epoch": 0.28647883858267714,
+      "grad_norm": 1.4140925407409668,
+      "learning_rate": 0.0001622690341098172,
+      "loss": 2.075,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2865403543307087,
+      "grad_norm": 1.0773919820785522,
+      "learning_rate": 0.0001622539018557805,
+      "loss": 3.1721,
+      "step": 4658
+    },
+    {
+      "epoch": 0.28660187007874016,
+      "grad_norm": 1.142231822013855,
+      "learning_rate": 0.00016223876727379497,
+      "loss": 3.0057,
+      "step": 4659
+    },
+    {
+      "epoch": 0.28666338582677164,
+      "grad_norm": 1.143755555152893,
+      "learning_rate": 0.0001622236303644266,
+      "loss": 2.3121,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2867249015748031,
+      "grad_norm": 1.2686090469360352,
+      "learning_rate": 0.00016220849112824146,
+      "loss": 2.1877,
+      "step": 4661
+    },
+    {
+      "epoch": 0.28678641732283466,
+      "grad_norm": 1.4543949365615845,
+      "learning_rate": 0.0001621933495658056,
+      "loss": 2.0767,
+      "step": 4662
+    },
+    {
+      "epoch": 0.28684793307086615,
+      "grad_norm": 1.0781265497207642,
+      "learning_rate": 0.00016217820567768527,
+      "loss": 2.8554,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2869094488188976,
+      "grad_norm": 1.0645018815994263,
+      "learning_rate": 0.00016216305946444677,
+      "loss": 3.0458,
+      "step": 4664
+    },
+    {
+      "epoch": 0.2869709645669291,
+      "grad_norm": 1.1845537424087524,
+      "learning_rate": 0.00016214791092665647,
+      "loss": 2.8247,
+      "step": 4665
+    },
+    {
+      "epoch": 0.28703248031496065,
+      "grad_norm": 1.1565319299697876,
+      "learning_rate": 0.00016213276006488085,
+      "loss": 2.0706,
+      "step": 4666
+    },
+    {
+      "epoch": 0.28709399606299213,
+      "grad_norm": 1.3019042015075684,
+      "learning_rate": 0.00016211760687968646,
+      "loss": 2.5426,
+      "step": 4667
+    },
+    {
+      "epoch": 0.2871555118110236,
+      "grad_norm": 1.0797659158706665,
+      "learning_rate": 0.00016210245137163995,
+      "loss": 2.8736,
+      "step": 4668
+    },
+    {
+      "epoch": 0.2872170275590551,
+      "grad_norm": 1.0576016902923584,
+      "learning_rate": 0.00016208729354130806,
+      "loss": 2.7971,
+      "step": 4669
+    },
+    {
+      "epoch": 0.28727854330708663,
+      "grad_norm": 1.1042574644088745,
+      "learning_rate": 0.0001620721333892576,
+      "loss": 3.1608,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2873400590551181,
+      "grad_norm": 1.3180222511291504,
+      "learning_rate": 0.00016205697091605548,
+      "loss": 2.5966,
+      "step": 4671
+    },
+    {
+      "epoch": 0.2874015748031496,
+      "grad_norm": 1.2659834623336792,
+      "learning_rate": 0.00016204180612226866,
+      "loss": 2.0772,
+      "step": 4672
+    },
+    {
+      "epoch": 0.2874630905511811,
+      "grad_norm": 1.1729931831359863,
+      "learning_rate": 0.0001620266390084643,
+      "loss": 2.695,
+      "step": 4673
+    },
+    {
+      "epoch": 0.2875246062992126,
+      "grad_norm": 1.1730505228042603,
+      "learning_rate": 0.00016201146957520946,
+      "loss": 2.7529,
+      "step": 4674
+    },
+    {
+      "epoch": 0.2875861220472441,
+      "grad_norm": 1.1137505769729614,
+      "learning_rate": 0.00016199629782307143,
+      "loss": 3.1745,
+      "step": 4675
+    },
+    {
+      "epoch": 0.2876476377952756,
+      "grad_norm": 1.000640630722046,
+      "learning_rate": 0.00016198112375261762,
+      "loss": 3.0752,
+      "step": 4676
+    },
+    {
+      "epoch": 0.28770915354330706,
+      "grad_norm": 1.4522089958190918,
+      "learning_rate": 0.00016196594736441538,
+      "loss": 2.7203,
+      "step": 4677
+    },
+    {
+      "epoch": 0.2877706692913386,
+      "grad_norm": 0.8932562470436096,
+      "learning_rate": 0.0001619507686590322,
+      "loss": 3.1258,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2878321850393701,
+      "grad_norm": 1.0725454092025757,
+      "learning_rate": 0.00016193558763703575,
+      "loss": 2.5437,
+      "step": 4679
+    },
+    {
+      "epoch": 0.28789370078740156,
+      "grad_norm": 1.729404330253601,
+      "learning_rate": 0.0001619204042989937,
+      "loss": 3.0053,
+      "step": 4680
+    },
+    {
+      "epoch": 0.28795521653543305,
+      "grad_norm": 1.2627519369125366,
+      "learning_rate": 0.0001619052186454738,
+      "loss": 2.6547,
+      "step": 4681
+    },
+    {
+      "epoch": 0.2880167322834646,
+      "grad_norm": 1.1362733840942383,
+      "learning_rate": 0.0001618900306770439,
+      "loss": 2.2209,
+      "step": 4682
+    },
+    {
+      "epoch": 0.28807824803149606,
+      "grad_norm": 1.0540101528167725,
+      "learning_rate": 0.00016187484039427194,
+      "loss": 3.0951,
+      "step": 4683
+    },
+    {
+      "epoch": 0.28813976377952755,
+      "grad_norm": 1.3327182531356812,
+      "learning_rate": 0.00016185964779772603,
+      "loss": 2.3209,
+      "step": 4684
+    },
+    {
+      "epoch": 0.28820127952755903,
+      "grad_norm": 1.138779640197754,
+      "learning_rate": 0.00016184445288797418,
+      "loss": 3.0455,
+      "step": 4685
+    },
+    {
+      "epoch": 0.28826279527559057,
+      "grad_norm": 1.0782198905944824,
+      "learning_rate": 0.00016182925566558468,
+      "loss": 2.716,
+      "step": 4686
+    },
+    {
+      "epoch": 0.28832431102362205,
+      "grad_norm": 1.1749882698059082,
+      "learning_rate": 0.00016181405613112577,
+      "loss": 2.7789,
+      "step": 4687
+    },
+    {
+      "epoch": 0.28838582677165353,
+      "grad_norm": 1.312641978263855,
+      "learning_rate": 0.00016179885428516583,
+      "loss": 2.5217,
+      "step": 4688
+    },
+    {
+      "epoch": 0.288447342519685,
+      "grad_norm": 1.1090084314346313,
+      "learning_rate": 0.00016178365012827338,
+      "loss": 3.0321,
+      "step": 4689
+    },
+    {
+      "epoch": 0.28850885826771655,
+      "grad_norm": 1.4979618787765503,
+      "learning_rate": 0.0001617684436610169,
+      "loss": 2.7224,
+      "step": 4690
+    },
+    {
+      "epoch": 0.28857037401574803,
+      "grad_norm": 1.0562686920166016,
+      "learning_rate": 0.00016175323488396508,
+      "loss": 2.7644,
+      "step": 4691
+    },
+    {
+      "epoch": 0.2886318897637795,
+      "grad_norm": 1.1145896911621094,
+      "learning_rate": 0.00016173802379768662,
+      "loss": 2.6491,
+      "step": 4692
+    },
+    {
+      "epoch": 0.288693405511811,
+      "grad_norm": 1.1519232988357544,
+      "learning_rate": 0.0001617228104027503,
+      "loss": 2.6867,
+      "step": 4693
+    },
+    {
+      "epoch": 0.28875492125984253,
+      "grad_norm": 1.2490944862365723,
+      "learning_rate": 0.00016170759469972507,
+      "loss": 3.1507,
+      "step": 4694
+    },
+    {
+      "epoch": 0.288816437007874,
+      "grad_norm": 1.3353427648544312,
+      "learning_rate": 0.00016169237668917986,
+      "loss": 2.3449,
+      "step": 4695
+    },
+    {
+      "epoch": 0.2888779527559055,
+      "grad_norm": 1.3312934637069702,
+      "learning_rate": 0.00016167715637168382,
+      "loss": 3.1115,
+      "step": 4696
+    },
+    {
+      "epoch": 0.288939468503937,
+      "grad_norm": 1.1246024370193481,
+      "learning_rate": 0.00016166193374780603,
+      "loss": 2.8731,
+      "step": 4697
+    },
+    {
+      "epoch": 0.2890009842519685,
+      "grad_norm": 1.306317925453186,
+      "learning_rate": 0.00016164670881811574,
+      "loss": 3.2717,
+      "step": 4698
+    },
+    {
+      "epoch": 0.2890625,
+      "grad_norm": 1.2227855920791626,
+      "learning_rate": 0.00016163148158318232,
+      "loss": 2.9488,
+      "step": 4699
+    },
+    {
+      "epoch": 0.2891240157480315,
+      "grad_norm": 1.3367503881454468,
+      "learning_rate": 0.00016161625204357512,
+      "loss": 2.9613,
+      "step": 4700
+    },
+    {
+      "epoch": 0.289185531496063,
+      "grad_norm": 1.3013933897018433,
+      "learning_rate": 0.0001616010201998637,
+      "loss": 2.5116,
+      "step": 4701
+    },
+    {
+      "epoch": 0.2892470472440945,
+      "grad_norm": 1.0991991758346558,
+      "learning_rate": 0.00016158578605261761,
+      "loss": 2.4957,
+      "step": 4702
+    },
+    {
+      "epoch": 0.289308562992126,
+      "grad_norm": 1.2267694473266602,
+      "learning_rate": 0.0001615705496024065,
+      "loss": 2.6793,
+      "step": 4703
+    },
+    {
+      "epoch": 0.28937007874015747,
+      "grad_norm": 1.0323015451431274,
+      "learning_rate": 0.0001615553108498002,
+      "loss": 2.9171,
+      "step": 4704
+    },
+    {
+      "epoch": 0.289431594488189,
+      "grad_norm": 1.0315102338790894,
+      "learning_rate": 0.00016154006979536854,
+      "loss": 2.8994,
+      "step": 4705
+    },
+    {
+      "epoch": 0.2894931102362205,
+      "grad_norm": 1.048244833946228,
+      "learning_rate": 0.0001615248264396814,
+      "loss": 2.6477,
+      "step": 4706
+    },
+    {
+      "epoch": 0.28955462598425197,
+      "grad_norm": 1.1929936408996582,
+      "learning_rate": 0.00016150958078330882,
+      "loss": 3.3838,
+      "step": 4707
+    },
+    {
+      "epoch": 0.28961614173228345,
+      "grad_norm": 1.0030460357666016,
+      "learning_rate": 0.0001614943328268209,
+      "loss": 2.9808,
+      "step": 4708
+    },
+    {
+      "epoch": 0.289677657480315,
+      "grad_norm": 1.3420523405075073,
+      "learning_rate": 0.00016147908257078784,
+      "loss": 3.2273,
+      "step": 4709
+    },
+    {
+      "epoch": 0.28973917322834647,
+      "grad_norm": 1.009576439857483,
+      "learning_rate": 0.00016146383001577994,
+      "loss": 2.8232,
+      "step": 4710
+    },
+    {
+      "epoch": 0.28980068897637795,
+      "grad_norm": 1.4230022430419922,
+      "learning_rate": 0.00016144857516236748,
+      "loss": 2.6094,
+      "step": 4711
+    },
+    {
+      "epoch": 0.28986220472440943,
+      "grad_norm": 1.1399201154708862,
+      "learning_rate": 0.000161433318011121,
+      "loss": 2.797,
+      "step": 4712
+    },
+    {
+      "epoch": 0.28992372047244097,
+      "grad_norm": 0.9709253311157227,
+      "learning_rate": 0.00016141805856261096,
+      "loss": 2.6953,
+      "step": 4713
+    },
+    {
+      "epoch": 0.28998523622047245,
+      "grad_norm": 1.1418147087097168,
+      "learning_rate": 0.000161402796817408,
+      "loss": 2.9872,
+      "step": 4714
+    },
+    {
+      "epoch": 0.29004675196850394,
+      "grad_norm": 1.102578043937683,
+      "learning_rate": 0.00016138753277608286,
+      "loss": 2.9375,
+      "step": 4715
+    },
+    {
+      "epoch": 0.2901082677165354,
+      "grad_norm": 1.124630093574524,
+      "learning_rate": 0.00016137226643920627,
+      "loss": 2.7534,
+      "step": 4716
+    },
+    {
+      "epoch": 0.29016978346456695,
+      "grad_norm": 1.2306936979293823,
+      "learning_rate": 0.00016135699780734915,
+      "loss": 2.1115,
+      "step": 4717
+    },
+    {
+      "epoch": 0.29023129921259844,
+      "grad_norm": 0.9753066301345825,
+      "learning_rate": 0.00016134172688108244,
+      "loss": 2.7265,
+      "step": 4718
+    },
+    {
+      "epoch": 0.2902928149606299,
+      "grad_norm": 1.0537021160125732,
+      "learning_rate": 0.00016132645366097719,
+      "loss": 2.7996,
+      "step": 4719
+    },
+    {
+      "epoch": 0.2903543307086614,
+      "grad_norm": 1.0797885656356812,
+      "learning_rate": 0.00016131117814760458,
+      "loss": 2.5644,
+      "step": 4720
+    },
+    {
+      "epoch": 0.29041584645669294,
+      "grad_norm": 1.0860984325408936,
+      "learning_rate": 0.00016129590034153575,
+      "loss": 2.8915,
+      "step": 4721
+    },
+    {
+      "epoch": 0.2904773622047244,
+      "grad_norm": 1.2645087242126465,
+      "learning_rate": 0.00016128062024334204,
+      "loss": 2.7357,
+      "step": 4722
+    },
+    {
+      "epoch": 0.2905388779527559,
+      "grad_norm": 1.0989103317260742,
+      "learning_rate": 0.00016126533785359485,
+      "loss": 2.9714,
+      "step": 4723
+    },
+    {
+      "epoch": 0.2906003937007874,
+      "grad_norm": 1.1441895961761475,
+      "learning_rate": 0.00016125005317286565,
+      "loss": 3.0025,
+      "step": 4724
+    },
+    {
+      "epoch": 0.2906619094488189,
+      "grad_norm": 1.1458436250686646,
+      "learning_rate": 0.000161234766201726,
+      "loss": 3.2379,
+      "step": 4725
+    },
+    {
+      "epoch": 0.2907234251968504,
+      "grad_norm": 1.190189242362976,
+      "learning_rate": 0.00016121947694074757,
+      "loss": 2.0194,
+      "step": 4726
+    },
+    {
+      "epoch": 0.2907849409448819,
+      "grad_norm": 1.045846700668335,
+      "learning_rate": 0.00016120418539050205,
+      "loss": 2.7974,
+      "step": 4727
+    },
+    {
+      "epoch": 0.29084645669291337,
+      "grad_norm": 1.0596561431884766,
+      "learning_rate": 0.00016118889155156127,
+      "loss": 3.1409,
+      "step": 4728
+    },
+    {
+      "epoch": 0.2909079724409449,
+      "grad_norm": 1.2869417667388916,
+      "learning_rate": 0.00016117359542449717,
+      "loss": 2.7075,
+      "step": 4729
+    },
+    {
+      "epoch": 0.2909694881889764,
+      "grad_norm": 1.1219050884246826,
+      "learning_rate": 0.0001611582970098817,
+      "loss": 2.3325,
+      "step": 4730
+    },
+    {
+      "epoch": 0.29103100393700787,
+      "grad_norm": 1.2261369228363037,
+      "learning_rate": 0.00016114299630828695,
+      "loss": 2.6667,
+      "step": 4731
+    },
+    {
+      "epoch": 0.29109251968503935,
+      "grad_norm": 1.2787598371505737,
+      "learning_rate": 0.00016112769332028509,
+      "loss": 2.4422,
+      "step": 4732
+    },
+    {
+      "epoch": 0.2911540354330709,
+      "grad_norm": 1.1724467277526855,
+      "learning_rate": 0.00016111238804644834,
+      "loss": 2.4162,
+      "step": 4733
+    },
+    {
+      "epoch": 0.2912155511811024,
+      "grad_norm": 1.1549687385559082,
+      "learning_rate": 0.00016109708048734906,
+      "loss": 2.7817,
+      "step": 4734
+    },
+    {
+      "epoch": 0.29127706692913385,
+      "grad_norm": 1.1532858610153198,
+      "learning_rate": 0.00016108177064355967,
+      "loss": 2.7697,
+      "step": 4735
+    },
+    {
+      "epoch": 0.29133858267716534,
+      "grad_norm": 1.1412546634674072,
+      "learning_rate": 0.00016106645851565266,
+      "loss": 2.7651,
+      "step": 4736
+    },
+    {
+      "epoch": 0.2914000984251969,
+      "grad_norm": 1.304656982421875,
+      "learning_rate": 0.00016105114410420058,
+      "loss": 2.6116,
+      "step": 4737
+    },
+    {
+      "epoch": 0.29146161417322836,
+      "grad_norm": 1.6156342029571533,
+      "learning_rate": 0.00016103582740977616,
+      "loss": 2.292,
+      "step": 4738
+    },
+    {
+      "epoch": 0.29152312992125984,
+      "grad_norm": 0.9843847751617432,
+      "learning_rate": 0.00016102050843295215,
+      "loss": 3.1879,
+      "step": 4739
+    },
+    {
+      "epoch": 0.2915846456692913,
+      "grad_norm": 1.174578309059143,
+      "learning_rate": 0.00016100518717430138,
+      "loss": 3.2455,
+      "step": 4740
+    },
+    {
+      "epoch": 0.29164616141732286,
+      "grad_norm": 1.116148829460144,
+      "learning_rate": 0.0001609898636343968,
+      "loss": 2.6471,
+      "step": 4741
+    },
+    {
+      "epoch": 0.29170767716535434,
+      "grad_norm": 1.1496261358261108,
+      "learning_rate": 0.0001609745378138114,
+      "loss": 1.9436,
+      "step": 4742
+    },
+    {
+      "epoch": 0.2917691929133858,
+      "grad_norm": 1.0779645442962646,
+      "learning_rate": 0.00016095920971311825,
+      "loss": 2.8225,
+      "step": 4743
+    },
+    {
+      "epoch": 0.2918307086614173,
+      "grad_norm": 1.300087332725525,
+      "learning_rate": 0.0001609438793328906,
+      "loss": 2.5983,
+      "step": 4744
+    },
+    {
+      "epoch": 0.29189222440944884,
+      "grad_norm": 1.1327155828475952,
+      "learning_rate": 0.0001609285466737017,
+      "loss": 3.0851,
+      "step": 4745
+    },
+    {
+      "epoch": 0.2919537401574803,
+      "grad_norm": 1.0655517578125,
+      "learning_rate": 0.0001609132117361249,
+      "loss": 3.0672,
+      "step": 4746
+    },
+    {
+      "epoch": 0.2920152559055118,
+      "grad_norm": 1.049393653869629,
+      "learning_rate": 0.00016089787452073365,
+      "loss": 3.0633,
+      "step": 4747
+    },
+    {
+      "epoch": 0.2920767716535433,
+      "grad_norm": 1.1474688053131104,
+      "learning_rate": 0.0001608825350281015,
+      "loss": 2.4558,
+      "step": 4748
+    },
+    {
+      "epoch": 0.2921382874015748,
+      "grad_norm": 1.1824558973312378,
+      "learning_rate": 0.00016086719325880195,
+      "loss": 2.6746,
+      "step": 4749
+    },
+    {
+      "epoch": 0.2921998031496063,
+      "grad_norm": 1.8965190649032593,
+      "learning_rate": 0.00016085184921340884,
+      "loss": 2.5247,
+      "step": 4750
+    },
+    {
+      "epoch": 0.2922613188976378,
+      "grad_norm": 0.9413197636604309,
+      "learning_rate": 0.00016083650289249585,
+      "loss": 2.044,
+      "step": 4751
+    },
+    {
+      "epoch": 0.29232283464566927,
+      "grad_norm": 1.0574960708618164,
+      "learning_rate": 0.00016082115429663692,
+      "loss": 3.0787,
+      "step": 4752
+    },
+    {
+      "epoch": 0.2923843503937008,
+      "grad_norm": 1.1448652744293213,
+      "learning_rate": 0.00016080580342640595,
+      "loss": 2.738,
+      "step": 4753
+    },
+    {
+      "epoch": 0.2924458661417323,
+      "grad_norm": 1.0494568347930908,
+      "learning_rate": 0.000160790450282377,
+      "loss": 3.2105,
+      "step": 4754
+    },
+    {
+      "epoch": 0.2925073818897638,
+      "grad_norm": 1.2819693088531494,
+      "learning_rate": 0.00016077509486512418,
+      "loss": 2.6863,
+      "step": 4755
+    },
+    {
+      "epoch": 0.29256889763779526,
+      "grad_norm": 1.189011573791504,
+      "learning_rate": 0.0001607597371752217,
+      "loss": 2.6067,
+      "step": 4756
+    },
+    {
+      "epoch": 0.2926304133858268,
+      "grad_norm": 1.461963176727295,
+      "learning_rate": 0.00016074437721324384,
+      "loss": 2.6346,
+      "step": 4757
+    },
+    {
+      "epoch": 0.2926919291338583,
+      "grad_norm": 1.2153595685958862,
+      "learning_rate": 0.000160729014979765,
+      "loss": 2.4741,
+      "step": 4758
+    },
+    {
+      "epoch": 0.29275344488188976,
+      "grad_norm": 1.2260465621948242,
+      "learning_rate": 0.00016071365047535968,
+      "loss": 2.1056,
+      "step": 4759
+    },
+    {
+      "epoch": 0.29281496062992124,
+      "grad_norm": 0.9923728108406067,
+      "learning_rate": 0.00016069828370060237,
+      "loss": 2.6391,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2928764763779528,
+      "grad_norm": 1.1641820669174194,
+      "learning_rate": 0.00016068291465606767,
+      "loss": 3.0067,
+      "step": 4761
+    },
+    {
+      "epoch": 0.29293799212598426,
+      "grad_norm": 0.9043299555778503,
+      "learning_rate": 0.00016066754334233035,
+      "loss": 2.5144,
+      "step": 4762
+    },
+    {
+      "epoch": 0.29299950787401574,
+      "grad_norm": 1.9335565567016602,
+      "learning_rate": 0.00016065216975996523,
+      "loss": 2.6463,
+      "step": 4763
+    },
+    {
+      "epoch": 0.2930610236220472,
+      "grad_norm": 1.359946846961975,
+      "learning_rate": 0.00016063679390954712,
+      "loss": 2.0425,
+      "step": 4764
+    },
+    {
+      "epoch": 0.29312253937007876,
+      "grad_norm": 1.1720930337905884,
+      "learning_rate": 0.00016062141579165106,
+      "loss": 2.7887,
+      "step": 4765
+    },
+    {
+      "epoch": 0.29318405511811024,
+      "grad_norm": 1.0858263969421387,
+      "learning_rate": 0.0001606060354068521,
+      "loss": 2.6328,
+      "step": 4766
+    },
+    {
+      "epoch": 0.2932455708661417,
+      "grad_norm": 1.136880874633789,
+      "learning_rate": 0.00016059065275572535,
+      "loss": 2.8709,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2933070866141732,
+      "grad_norm": 0.9847744107246399,
+      "learning_rate": 0.00016057526783884603,
+      "loss": 2.7143,
+      "step": 4768
+    },
+    {
+      "epoch": 0.29336860236220474,
+      "grad_norm": 1.2673603296279907,
+      "learning_rate": 0.00016055988065678952,
+      "loss": 2.6165,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2934301181102362,
+      "grad_norm": 1.0774317979812622,
+      "learning_rate": 0.00016054449121013112,
+      "loss": 3.1619,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2934916338582677,
+      "grad_norm": 1.1627027988433838,
+      "learning_rate": 0.00016052909949944638,
+      "loss": 3.2281,
+      "step": 4771
+    },
+    {
+      "epoch": 0.2935531496062992,
+      "grad_norm": 0.961686372756958,
+      "learning_rate": 0.00016051370552531085,
+      "loss": 3.1325,
+      "step": 4772
+    },
+    {
+      "epoch": 0.29361466535433073,
+      "grad_norm": 1.1875594854354858,
+      "learning_rate": 0.00016049830928830015,
+      "loss": 2.8234,
+      "step": 4773
+    },
+    {
+      "epoch": 0.2936761811023622,
+      "grad_norm": 1.3072967529296875,
+      "learning_rate": 0.00016048291078899,
+      "loss": 2.465,
+      "step": 4774
+    },
+    {
+      "epoch": 0.2937376968503937,
+      "grad_norm": 1.1353789567947388,
+      "learning_rate": 0.00016046751002795628,
+      "loss": 3.0037,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2937992125984252,
+      "grad_norm": 0.9264965653419495,
+      "learning_rate": 0.00016045210700577487,
+      "loss": 2.9958,
+      "step": 4776
+    },
+    {
+      "epoch": 0.2938607283464567,
+      "grad_norm": 1.35205078125,
+      "learning_rate": 0.00016043670172302172,
+      "loss": 2.8761,
+      "step": 4777
+    },
+    {
+      "epoch": 0.2939222440944882,
+      "grad_norm": 1.062429428100586,
+      "learning_rate": 0.00016042129418027297,
+      "loss": 3.089,
+      "step": 4778
+    },
+    {
+      "epoch": 0.2939837598425197,
+      "grad_norm": 1.1638269424438477,
+      "learning_rate": 0.00016040588437810468,
+      "loss": 2.7863,
+      "step": 4779
+    },
+    {
+      "epoch": 0.29404527559055116,
+      "grad_norm": 1.3738300800323486,
+      "learning_rate": 0.00016039047231709318,
+      "loss": 2.3771,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2941067913385827,
+      "grad_norm": 1.0561339855194092,
+      "learning_rate": 0.00016037505799781476,
+      "loss": 3.0208,
+      "step": 4781
+    },
+    {
+      "epoch": 0.2941683070866142,
+      "grad_norm": 1.2762647867202759,
+      "learning_rate": 0.00016035964142084583,
+      "loss": 2.5474,
+      "step": 4782
+    },
+    {
+      "epoch": 0.29422982283464566,
+      "grad_norm": 1.1598780155181885,
+      "learning_rate": 0.00016034422258676289,
+      "loss": 2.452,
+      "step": 4783
+    },
+    {
+      "epoch": 0.29429133858267714,
+      "grad_norm": 1.1389797925949097,
+      "learning_rate": 0.0001603288014961425,
+      "loss": 2.2823,
+      "step": 4784
+    },
+    {
+      "epoch": 0.2943528543307087,
+      "grad_norm": 1.4961258172988892,
+      "learning_rate": 0.00016031337814956137,
+      "loss": 2.6135,
+      "step": 4785
+    },
+    {
+      "epoch": 0.29441437007874016,
+      "grad_norm": 1.4298272132873535,
+      "learning_rate": 0.00016029795254759617,
+      "loss": 1.2493,
+      "step": 4786
+    },
+    {
+      "epoch": 0.29447588582677164,
+      "grad_norm": 1.330047845840454,
+      "learning_rate": 0.00016028252469082378,
+      "loss": 2.7798,
+      "step": 4787
+    },
+    {
+      "epoch": 0.2945374015748031,
+      "grad_norm": 1.1190193891525269,
+      "learning_rate": 0.00016026709457982115,
+      "loss": 2.7411,
+      "step": 4788
+    },
+    {
+      "epoch": 0.29459891732283466,
+      "grad_norm": 1.1521906852722168,
+      "learning_rate": 0.0001602516622151652,
+      "loss": 2.6854,
+      "step": 4789
+    },
+    {
+      "epoch": 0.29466043307086615,
+      "grad_norm": 1.2481646537780762,
+      "learning_rate": 0.00016023622759743308,
+      "loss": 1.194,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2947219488188976,
+      "grad_norm": 1.361781120300293,
+      "learning_rate": 0.00016022079072720193,
+      "loss": 2.0404,
+      "step": 4791
+    },
+    {
+      "epoch": 0.2947834645669291,
+      "grad_norm": 1.1321898698806763,
+      "learning_rate": 0.000160205351605049,
+      "loss": 2.802,
+      "step": 4792
+    },
+    {
+      "epoch": 0.29484498031496065,
+      "grad_norm": 1.1061396598815918,
+      "learning_rate": 0.00016018991023155162,
+      "loss": 2.519,
+      "step": 4793
+    },
+    {
+      "epoch": 0.29490649606299213,
+      "grad_norm": 1.2198725938796997,
+      "learning_rate": 0.00016017446660728725,
+      "loss": 2.7551,
+      "step": 4794
+    },
+    {
+      "epoch": 0.2949680118110236,
+      "grad_norm": 1.3463919162750244,
+      "learning_rate": 0.0001601590207328334,
+      "loss": 3.1,
+      "step": 4795
+    },
+    {
+      "epoch": 0.2950295275590551,
+      "grad_norm": 1.1704790592193604,
+      "learning_rate": 0.00016014357260876757,
+      "loss": 2.8957,
+      "step": 4796
+    },
+    {
+      "epoch": 0.29509104330708663,
+      "grad_norm": 1.2451845407485962,
+      "learning_rate": 0.00016012812223566752,
+      "loss": 2.6597,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2951525590551181,
+      "grad_norm": 1.2524900436401367,
+      "learning_rate": 0.00016011266961411097,
+      "loss": 2.8709,
+      "step": 4798
+    },
+    {
+      "epoch": 0.2952140748031496,
+      "grad_norm": 1.0259252786636353,
+      "learning_rate": 0.0001600972147446758,
+      "loss": 2.4161,
+      "step": 4799
+    },
+    {
+      "epoch": 0.2952755905511811,
+      "grad_norm": 1.1970151662826538,
+      "learning_rate": 0.0001600817576279399,
+      "loss": 2.5862,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2953371062992126,
+      "grad_norm": 0.9495461583137512,
+      "learning_rate": 0.00016006629826448128,
+      "loss": 3.1948,
+      "step": 4801
+    },
+    {
+      "epoch": 0.2953986220472441,
+      "grad_norm": 1.0803916454315186,
+      "learning_rate": 0.00016005083665487804,
+      "loss": 2.7706,
+      "step": 4802
+    },
+    {
+      "epoch": 0.2954601377952756,
+      "grad_norm": 1.8056267499923706,
+      "learning_rate": 0.00016003537279970836,
+      "loss": 1.849,
+      "step": 4803
+    },
+    {
+      "epoch": 0.29552165354330706,
+      "grad_norm": 1.1690521240234375,
+      "learning_rate": 0.00016001990669955052,
+      "loss": 2.4229,
+      "step": 4804
+    },
+    {
+      "epoch": 0.2955831692913386,
+      "grad_norm": 1.0830731391906738,
+      "learning_rate": 0.00016000443835498284,
+      "loss": 1.914,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2956446850393701,
+      "grad_norm": 1.4319514036178589,
+      "learning_rate": 0.0001599889677665838,
+      "loss": 2.7561,
+      "step": 4806
+    },
+    {
+      "epoch": 0.29570620078740156,
+      "grad_norm": 1.3302457332611084,
+      "learning_rate": 0.00015997349493493182,
+      "loss": 2.9349,
+      "step": 4807
+    },
+    {
+      "epoch": 0.29576771653543305,
+      "grad_norm": 1.8057187795639038,
+      "learning_rate": 0.00015995801986060557,
+      "loss": 1.5229,
+      "step": 4808
+    },
+    {
+      "epoch": 0.2958292322834646,
+      "grad_norm": 0.9786076545715332,
+      "learning_rate": 0.0001599425425441837,
+      "loss": 3.3288,
+      "step": 4809
+    },
+    {
+      "epoch": 0.29589074803149606,
+      "grad_norm": 1.113142728805542,
+      "learning_rate": 0.000159927062986245,
+      "loss": 2.913,
+      "step": 4810
+    },
+    {
+      "epoch": 0.29595226377952755,
+      "grad_norm": 1.1066116094589233,
+      "learning_rate": 0.0001599115811873683,
+      "loss": 2.4115,
+      "step": 4811
+    },
+    {
+      "epoch": 0.29601377952755903,
+      "grad_norm": 1.5618215799331665,
+      "learning_rate": 0.00015989609714813256,
+      "loss": 2.6856,
+      "step": 4812
+    },
+    {
+      "epoch": 0.29607529527559057,
+      "grad_norm": 1.112648606300354,
+      "learning_rate": 0.00015988061086911673,
+      "loss": 2.8475,
+      "step": 4813
+    },
+    {
+      "epoch": 0.29613681102362205,
+      "grad_norm": 1.148156762123108,
+      "learning_rate": 0.0001598651223509,
+      "loss": 3.1612,
+      "step": 4814
+    },
+    {
+      "epoch": 0.29619832677165353,
+      "grad_norm": 1.1744130849838257,
+      "learning_rate": 0.0001598496315940615,
+      "loss": 2.74,
+      "step": 4815
+    },
+    {
+      "epoch": 0.296259842519685,
+      "grad_norm": 1.299706220626831,
+      "learning_rate": 0.0001598341385991805,
+      "loss": 3.0174,
+      "step": 4816
+    },
+    {
+      "epoch": 0.29632135826771655,
+      "grad_norm": 1.1013771295547485,
+      "learning_rate": 0.0001598186433668364,
+      "loss": 2.7062,
+      "step": 4817
+    },
+    {
+      "epoch": 0.29638287401574803,
+      "grad_norm": 1.3045375347137451,
+      "learning_rate": 0.00015980314589760853,
+      "loss": 2.3686,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2964443897637795,
+      "grad_norm": 1.1942012310028076,
+      "learning_rate": 0.0001597876461920765,
+      "loss": 3.1317,
+      "step": 4819
+    },
+    {
+      "epoch": 0.296505905511811,
+      "grad_norm": 0.9493500590324402,
+      "learning_rate": 0.00015977214425081992,
+      "loss": 3.2009,
+      "step": 4820
+    },
+    {
+      "epoch": 0.29656742125984253,
+      "grad_norm": 1.1484111547470093,
+      "learning_rate": 0.00015975664007441843,
+      "loss": 2.5279,
+      "step": 4821
+    },
+    {
+      "epoch": 0.296628937007874,
+      "grad_norm": 1.1988306045532227,
+      "learning_rate": 0.00015974113366345178,
+      "loss": 2.6955,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2966904527559055,
+      "grad_norm": 1.2097610235214233,
+      "learning_rate": 0.0001597256250184999,
+      "loss": 2.2937,
+      "step": 4823
+    },
+    {
+      "epoch": 0.296751968503937,
+      "grad_norm": 1.1901737451553345,
+      "learning_rate": 0.00015971011414014265,
+      "loss": 2.7386,
+      "step": 4824
+    },
+    {
+      "epoch": 0.2968134842519685,
+      "grad_norm": 1.206157922744751,
+      "learning_rate": 0.0001596946010289601,
+      "loss": 2.2775,
+      "step": 4825
+    },
+    {
+      "epoch": 0.296875,
+      "grad_norm": 1.215260624885559,
+      "learning_rate": 0.00015967908568553232,
+      "loss": 2.8824,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2969365157480315,
+      "grad_norm": 1.261181354522705,
+      "learning_rate": 0.00015966356811043954,
+      "loss": 2.6614,
+      "step": 4827
+    },
+    {
+      "epoch": 0.296998031496063,
+      "grad_norm": 1.0271024703979492,
+      "learning_rate": 0.00015964804830426198,
+      "loss": 2.4941,
+      "step": 4828
+    },
+    {
+      "epoch": 0.2970595472440945,
+      "grad_norm": 1.1587308645248413,
+      "learning_rate": 0.00015963252626758004,
+      "loss": 2.2462,
+      "step": 4829
+    },
+    {
+      "epoch": 0.297121062992126,
+      "grad_norm": 1.7452574968338013,
+      "learning_rate": 0.00015961700200097415,
+      "loss": 2.0479,
+      "step": 4830
+    },
+    {
+      "epoch": 0.29718257874015747,
+      "grad_norm": 1.2740708589553833,
+      "learning_rate": 0.0001596014755050248,
+      "loss": 2.4636,
+      "step": 4831
+    },
+    {
+      "epoch": 0.297244094488189,
+      "grad_norm": 1.3328737020492554,
+      "learning_rate": 0.0001595859467803126,
+      "loss": 2.7116,
+      "step": 4832
+    },
+    {
+      "epoch": 0.2973056102362205,
+      "grad_norm": 1.1771894693374634,
+      "learning_rate": 0.00015957041582741828,
+      "loss": 2.7654,
+      "step": 4833
+    },
+    {
+      "epoch": 0.29736712598425197,
+      "grad_norm": 1.1939533948898315,
+      "learning_rate": 0.00015955488264692255,
+      "loss": 2.2144,
+      "step": 4834
+    },
+    {
+      "epoch": 0.29742864173228345,
+      "grad_norm": 1.5889363288879395,
+      "learning_rate": 0.00015953934723940633,
+      "loss": 2.5764,
+      "step": 4835
+    },
+    {
+      "epoch": 0.297490157480315,
+      "grad_norm": 1.0495988130569458,
+      "learning_rate": 0.0001595238096054505,
+      "loss": 2.6551,
+      "step": 4836
+    },
+    {
+      "epoch": 0.29755167322834647,
+      "grad_norm": 1.2281873226165771,
+      "learning_rate": 0.00015950826974563614,
+      "loss": 2.6488,
+      "step": 4837
+    },
+    {
+      "epoch": 0.29761318897637795,
+      "grad_norm": 1.0579283237457275,
+      "learning_rate": 0.00015949272766054427,
+      "loss": 2.4045,
+      "step": 4838
+    },
+    {
+      "epoch": 0.29767470472440943,
+      "grad_norm": 1.0720999240875244,
+      "learning_rate": 0.00015947718335075615,
+      "loss": 2.8142,
+      "step": 4839
+    },
+    {
+      "epoch": 0.29773622047244097,
+      "grad_norm": 1.0044304132461548,
+      "learning_rate": 0.00015946163681685303,
+      "loss": 2.3738,
+      "step": 4840
+    },
+    {
+      "epoch": 0.29779773622047245,
+      "grad_norm": 1.313467025756836,
+      "learning_rate": 0.00015944608805941622,
+      "loss": 2.677,
+      "step": 4841
+    },
+    {
+      "epoch": 0.29785925196850394,
+      "grad_norm": 1.1899086236953735,
+      "learning_rate": 0.00015943053707902726,
+      "loss": 2.4287,
+      "step": 4842
+    },
+    {
+      "epoch": 0.2979207677165354,
+      "grad_norm": 1.7321226596832275,
+      "learning_rate": 0.00015941498387626757,
+      "loss": 3.009,
+      "step": 4843
+    },
+    {
+      "epoch": 0.29798228346456695,
+      "grad_norm": 1.0124009847640991,
+      "learning_rate": 0.00015939942845171883,
+      "loss": 3.0889,
+      "step": 4844
+    },
+    {
+      "epoch": 0.29804379921259844,
+      "grad_norm": 1.1032806634902954,
+      "learning_rate": 0.00015938387080596263,
+      "loss": 2.6792,
+      "step": 4845
+    },
+    {
+      "epoch": 0.2981053149606299,
+      "grad_norm": 1.2566081285476685,
+      "learning_rate": 0.00015936831093958081,
+      "loss": 2.4604,
+      "step": 4846
+    },
+    {
+      "epoch": 0.2981668307086614,
+      "grad_norm": 1.0227073431015015,
+      "learning_rate": 0.00015935274885315524,
+      "loss": 2.591,
+      "step": 4847
+    },
+    {
+      "epoch": 0.29822834645669294,
+      "grad_norm": 1.2103911638259888,
+      "learning_rate": 0.0001593371845472678,
+      "loss": 2.8048,
+      "step": 4848
+    },
+    {
+      "epoch": 0.2982898622047244,
+      "grad_norm": 1.1933239698410034,
+      "learning_rate": 0.00015932161802250053,
+      "loss": 2.6816,
+      "step": 4849
+    },
+    {
+      "epoch": 0.2983513779527559,
+      "grad_norm": 1.2193037271499634,
+      "learning_rate": 0.00015930604927943552,
+      "loss": 2.8474,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2984128937007874,
+      "grad_norm": 1.0995290279388428,
+      "learning_rate": 0.00015929047831865498,
+      "loss": 3.009,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2984744094488189,
+      "grad_norm": 1.1004085540771484,
+      "learning_rate": 0.00015927490514074116,
+      "loss": 3.3131,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2985359251968504,
+      "grad_norm": 1.2475188970565796,
+      "learning_rate": 0.00015925932974627643,
+      "loss": 2.8104,
+      "step": 4853
+    },
+    {
+      "epoch": 0.2985974409448819,
+      "grad_norm": 1.1066526174545288,
+      "learning_rate": 0.00015924375213584318,
+      "loss": 2.3882,
+      "step": 4854
+    },
+    {
+      "epoch": 0.29865895669291337,
+      "grad_norm": 1.107508659362793,
+      "learning_rate": 0.00015922817231002397,
+      "loss": 2.8593,
+      "step": 4855
+    },
+    {
+      "epoch": 0.2987204724409449,
+      "grad_norm": 1.278817892074585,
+      "learning_rate": 0.00015921259026940138,
+      "loss": 2.9576,
+      "step": 4856
+    },
+    {
+      "epoch": 0.2987819881889764,
+      "grad_norm": 1.2821513414382935,
+      "learning_rate": 0.0001591970060145581,
+      "loss": 3.5201,
+      "step": 4857
+    },
+    {
+      "epoch": 0.29884350393700787,
+      "grad_norm": 2.453383684158325,
+      "learning_rate": 0.00015918141954607688,
+      "loss": 2.2596,
+      "step": 4858
+    },
+    {
+      "epoch": 0.29890501968503935,
+      "grad_norm": 1.1788479089736938,
+      "learning_rate": 0.00015916583086454056,
+      "loss": 3.143,
+      "step": 4859
+    },
+    {
+      "epoch": 0.2989665354330709,
+      "grad_norm": 1.1688908338546753,
+      "learning_rate": 0.0001591502399705321,
+      "loss": 2.9908,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2990280511811024,
+      "grad_norm": 1.1092981100082397,
+      "learning_rate": 0.0001591346468646345,
+      "loss": 2.7064,
+      "step": 4861
+    },
+    {
+      "epoch": 0.29908956692913385,
+      "grad_norm": 1.1626051664352417,
+      "learning_rate": 0.00015911905154743084,
+      "loss": 2.8656,
+      "step": 4862
+    },
+    {
+      "epoch": 0.29915108267716534,
+      "grad_norm": 1.0698535442352295,
+      "learning_rate": 0.0001591034540195043,
+      "loss": 2.6558,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2992125984251969,
+      "grad_norm": 1.1347033977508545,
+      "learning_rate": 0.00015908785428143815,
+      "loss": 1.9224,
+      "step": 4864
+    },
+    {
+      "epoch": 0.29927411417322836,
+      "grad_norm": 1.867475152015686,
+      "learning_rate": 0.00015907225233381575,
+      "loss": 2.3776,
+      "step": 4865
+    },
+    {
+      "epoch": 0.29933562992125984,
+      "grad_norm": 1.2457207441329956,
+      "learning_rate": 0.0001590566481772205,
+      "loss": 2.6199,
+      "step": 4866
+    },
+    {
+      "epoch": 0.2993971456692913,
+      "grad_norm": 1.0371519327163696,
+      "learning_rate": 0.00015904104181223592,
+      "loss": 2.8045,
+      "step": 4867
+    },
+    {
+      "epoch": 0.29945866141732286,
+      "grad_norm": 1.1926041841506958,
+      "learning_rate": 0.0001590254332394456,
+      "loss": 2.6988,
+      "step": 4868
+    },
+    {
+      "epoch": 0.29952017716535434,
+      "grad_norm": 1.1906627416610718,
+      "learning_rate": 0.00015900982245943318,
+      "loss": 2.9875,
+      "step": 4869
+    },
+    {
+      "epoch": 0.2995816929133858,
+      "grad_norm": 1.0899044275283813,
+      "learning_rate": 0.0001589942094727825,
+      "loss": 2.4893,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2996432086614173,
+      "grad_norm": 1.4465601444244385,
+      "learning_rate": 0.00015897859428007732,
+      "loss": 2.5229,
+      "step": 4871
+    },
+    {
+      "epoch": 0.29970472440944884,
+      "grad_norm": 1.2007819414138794,
+      "learning_rate": 0.00015896297688190158,
+      "loss": 2.419,
+      "step": 4872
+    },
+    {
+      "epoch": 0.2997662401574803,
+      "grad_norm": 1.1753671169281006,
+      "learning_rate": 0.00015894735727883932,
+      "loss": 3.1114,
+      "step": 4873
+    },
+    {
+      "epoch": 0.2998277559055118,
+      "grad_norm": 1.1752164363861084,
+      "learning_rate": 0.00015893173547147456,
+      "loss": 2.6001,
+      "step": 4874
+    },
+    {
+      "epoch": 0.2998892716535433,
+      "grad_norm": 1.1844929456710815,
+      "learning_rate": 0.0001589161114603915,
+      "loss": 2.9935,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2999507874015748,
+      "grad_norm": 1.1879594326019287,
+      "learning_rate": 0.00015890048524617445,
+      "loss": 3.2049,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3000123031496063,
+      "grad_norm": 1.1767131090164185,
+      "learning_rate": 0.00015888485682940764,
+      "loss": 2.0942,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3000738188976378,
+      "grad_norm": 1.4607532024383545,
+      "learning_rate": 0.00015886922621067556,
+      "loss": 2.5526,
+      "step": 4878
+    },
+    {
+      "epoch": 0.30013533464566927,
+      "grad_norm": 2.6280839443206787,
+      "learning_rate": 0.00015885359339056268,
+      "loss": 1.8853,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3001968503937008,
+      "grad_norm": 1.6259530782699585,
+      "learning_rate": 0.00015883795836965357,
+      "loss": 3.8189,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3002583661417323,
+      "grad_norm": 1.2327033281326294,
+      "learning_rate": 0.00015882232114853296,
+      "loss": 2.8748,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3003198818897638,
+      "grad_norm": 2.110041856765747,
+      "learning_rate": 0.0001588066817277855,
+      "loss": 2.7716,
+      "step": 4882
+    },
+    {
+      "epoch": 0.30038139763779526,
+      "grad_norm": 1.164729356765747,
+      "learning_rate": 0.00015879104010799603,
+      "loss": 1.7172,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3004429133858268,
+      "grad_norm": 1.131719708442688,
+      "learning_rate": 0.00015877539628974952,
+      "loss": 2.8411,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3005044291338583,
+      "grad_norm": 1.0882172584533691,
+      "learning_rate": 0.00015875975027363094,
+      "loss": 3.1974,
+      "step": 4885
+    },
+    {
+      "epoch": 0.30056594488188976,
+      "grad_norm": 1.0828602313995361,
+      "learning_rate": 0.00015874410206022534,
+      "loss": 2.6994,
+      "step": 4886
+    },
+    {
+      "epoch": 0.30062746062992124,
+      "grad_norm": 1.9597864151000977,
+      "learning_rate": 0.0001587284516501179,
+      "loss": 3.451,
+      "step": 4887
+    },
+    {
+      "epoch": 0.3006889763779528,
+      "grad_norm": 0.9737358689308167,
+      "learning_rate": 0.00015871279904389383,
+      "loss": 3.1289,
+      "step": 4888
+    },
+    {
+      "epoch": 0.30075049212598426,
+      "grad_norm": 1.0725927352905273,
+      "learning_rate": 0.00015869714424213846,
+      "loss": 2.6943,
+      "step": 4889
+    },
+    {
+      "epoch": 0.30081200787401574,
+      "grad_norm": 1.0301021337509155,
+      "learning_rate": 0.00015868148724543723,
+      "loss": 2.6069,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3008735236220472,
+      "grad_norm": 1.2048838138580322,
+      "learning_rate": 0.00015866582805437558,
+      "loss": 2.6525,
+      "step": 4891
+    },
+    {
+      "epoch": 0.30093503937007876,
+      "grad_norm": 0.9619945287704468,
+      "learning_rate": 0.00015865016666953907,
+      "loss": 2.4772,
+      "step": 4892
+    },
+    {
+      "epoch": 0.30099655511811024,
+      "grad_norm": 1.194504976272583,
+      "learning_rate": 0.00015863450309151338,
+      "loss": 2.8235,
+      "step": 4893
+    },
+    {
+      "epoch": 0.3010580708661417,
+      "grad_norm": 1.366633653640747,
+      "learning_rate": 0.00015861883732088425,
+      "loss": 1.9301,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3011195866141732,
+      "grad_norm": 1.0207074880599976,
+      "learning_rate": 0.00015860316935823745,
+      "loss": 2.9589,
+      "step": 4895
+    },
+    {
+      "epoch": 0.30118110236220474,
+      "grad_norm": 1.144622564315796,
+      "learning_rate": 0.0001585874992041589,
+      "loss": 2.5622,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3012426181102362,
+      "grad_norm": 1.095579981803894,
+      "learning_rate": 0.00015857182685923458,
+      "loss": 3.0087,
+      "step": 4897
+    },
+    {
+      "epoch": 0.3013041338582677,
+      "grad_norm": 1.4540863037109375,
+      "learning_rate": 0.00015855615232405055,
+      "loss": 2.9627,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3013656496062992,
+      "grad_norm": 1.0110819339752197,
+      "learning_rate": 0.0001585404755991929,
+      "loss": 2.5849,
+      "step": 4899
+    },
+    {
+      "epoch": 0.30142716535433073,
+      "grad_norm": 1.063454031944275,
+      "learning_rate": 0.00015852479668524794,
+      "loss": 2.8331,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3014886811023622,
+      "grad_norm": 1.2303301095962524,
+      "learning_rate": 0.0001585091155828019,
+      "loss": 2.6966,
+      "step": 4901
+    },
+    {
+      "epoch": 0.3015501968503937,
+      "grad_norm": 1.0929139852523804,
+      "learning_rate": 0.0001584934322924412,
+      "loss": 2.6998,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3016117125984252,
+      "grad_norm": 1.2823587656021118,
+      "learning_rate": 0.00015847774681475229,
+      "loss": 2.8246,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3016732283464567,
+      "grad_norm": 1.099130630493164,
+      "learning_rate": 0.00015846205915032174,
+      "loss": 2.6582,
+      "step": 4904
+    },
+    {
+      "epoch": 0.3017347440944882,
+      "grad_norm": 0.9354981184005737,
+      "learning_rate": 0.00015844636929973618,
+      "loss": 1.8054,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3017962598425197,
+      "grad_norm": 1.10685396194458,
+      "learning_rate": 0.0001584306772635823,
+      "loss": 2.6691,
+      "step": 4906
+    },
+    {
+      "epoch": 0.30185777559055116,
+      "grad_norm": 1.1192599534988403,
+      "learning_rate": 0.00015841498304244693,
+      "loss": 2.7984,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3019192913385827,
+      "grad_norm": 1.0983449220657349,
+      "learning_rate": 0.00015839928663691693,
+      "loss": 3.4715,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3019808070866142,
+      "grad_norm": 1.267878770828247,
+      "learning_rate": 0.0001583835880475792,
+      "loss": 2.2569,
+      "step": 4909
+    },
+    {
+      "epoch": 0.30204232283464566,
+      "grad_norm": 1.9003444910049438,
+      "learning_rate": 0.00015836788727502087,
+      "loss": 2.263,
+      "step": 4910
+    },
+    {
+      "epoch": 0.30210383858267714,
+      "grad_norm": 1.167526364326477,
+      "learning_rate": 0.00015835218431982904,
+      "loss": 2.659,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3021653543307087,
+      "grad_norm": 1.3715718984603882,
+      "learning_rate": 0.00015833647918259085,
+      "loss": 2.176,
+      "step": 4912
+    },
+    {
+      "epoch": 0.30222687007874016,
+      "grad_norm": 1.1792789697647095,
+      "learning_rate": 0.00015832077186389367,
+      "loss": 2.5927,
+      "step": 4913
+    },
+    {
+      "epoch": 0.30228838582677164,
+      "grad_norm": 1.4575066566467285,
+      "learning_rate": 0.00015830506236432483,
+      "loss": 2.8282,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3023499015748031,
+      "grad_norm": 1.1031008958816528,
+      "learning_rate": 0.00015828935068447174,
+      "loss": 2.9883,
+      "step": 4915
+    },
+    {
+      "epoch": 0.30241141732283466,
+      "grad_norm": 1.3470453023910522,
+      "learning_rate": 0.000158273636824922,
+      "loss": 2.7243,
+      "step": 4916
+    },
+    {
+      "epoch": 0.30247293307086615,
+      "grad_norm": 1.3710490465164185,
+      "learning_rate": 0.0001582579207862632,
+      "loss": 2.5967,
+      "step": 4917
+    },
+    {
+      "epoch": 0.3025344488188976,
+      "grad_norm": 1.445923089981079,
+      "learning_rate": 0.000158242202569083,
+      "loss": 2.2759,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3025959645669291,
+      "grad_norm": 1.2783039808273315,
+      "learning_rate": 0.00015822648217396918,
+      "loss": 2.5118,
+      "step": 4919
+    },
+    {
+      "epoch": 0.30265748031496065,
+      "grad_norm": 0.96812504529953,
+      "learning_rate": 0.00015821075960150959,
+      "loss": 2.9461,
+      "step": 4920
+    },
+    {
+      "epoch": 0.30271899606299213,
+      "grad_norm": 1.2312179803848267,
+      "learning_rate": 0.00015819503485229222,
+      "loss": 2.6901,
+      "step": 4921
+    },
+    {
+      "epoch": 0.3027805118110236,
+      "grad_norm": 1.1178754568099976,
+      "learning_rate": 0.00015817930792690502,
+      "loss": 2.9925,
+      "step": 4922
+    },
+    {
+      "epoch": 0.3028420275590551,
+      "grad_norm": 1.053446650505066,
+      "learning_rate": 0.00015816357882593614,
+      "loss": 2.8614,
+      "step": 4923
+    },
+    {
+      "epoch": 0.30290354330708663,
+      "grad_norm": 0.9743262529373169,
+      "learning_rate": 0.00015814784754997373,
+      "loss": 2.7192,
+      "step": 4924
+    },
+    {
+      "epoch": 0.3029650590551181,
+      "grad_norm": 1.0512984991073608,
+      "learning_rate": 0.00015813211409960608,
+      "loss": 3.2518,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3030265748031496,
+      "grad_norm": 1.0666710138320923,
+      "learning_rate": 0.00015811637847542147,
+      "loss": 3.3294,
+      "step": 4926
+    },
+    {
+      "epoch": 0.3030880905511811,
+      "grad_norm": 1.2195631265640259,
+      "learning_rate": 0.00015810064067800842,
+      "loss": 2.2853,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3031496062992126,
+      "grad_norm": 1.3018420934677124,
+      "learning_rate": 0.00015808490070795536,
+      "loss": 1.9068,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3032111220472441,
+      "grad_norm": 1.1791508197784424,
+      "learning_rate": 0.00015806915856585093,
+      "loss": 2.3889,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3032726377952756,
+      "grad_norm": 1.1695011854171753,
+      "learning_rate": 0.00015805341425228372,
+      "loss": 3.382,
+      "step": 4930
+    },
+    {
+      "epoch": 0.30333415354330706,
+      "grad_norm": 1.1360746622085571,
+      "learning_rate": 0.0001580376677678426,
+      "loss": 3.297,
+      "step": 4931
+    },
+    {
+      "epoch": 0.3033956692913386,
+      "grad_norm": 1.284388780593872,
+      "learning_rate": 0.00015802191911311631,
+      "loss": 2.6343,
+      "step": 4932
+    },
+    {
+      "epoch": 0.3034571850393701,
+      "grad_norm": 0.9773507118225098,
+      "learning_rate": 0.00015800616828869375,
+      "loss": 3.0622,
+      "step": 4933
+    },
+    {
+      "epoch": 0.30351870078740156,
+      "grad_norm": 1.1578261852264404,
+      "learning_rate": 0.00015799041529516398,
+      "loss": 2.2675,
+      "step": 4934
+    },
+    {
+      "epoch": 0.30358021653543305,
+      "grad_norm": 1.2151103019714355,
+      "learning_rate": 0.00015797466013311602,
+      "loss": 2.4587,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3036417322834646,
+      "grad_norm": 1.145583152770996,
+      "learning_rate": 0.00015795890280313906,
+      "loss": 2.4722,
+      "step": 4936
+    },
+    {
+      "epoch": 0.30370324803149606,
+      "grad_norm": 1.1410161256790161,
+      "learning_rate": 0.00015794314330582233,
+      "loss": 2.3223,
+      "step": 4937
+    },
+    {
+      "epoch": 0.30376476377952755,
+      "grad_norm": 1.9711912870407104,
+      "learning_rate": 0.00015792738164175513,
+      "loss": 2.5291,
+      "step": 4938
+    },
+    {
+      "epoch": 0.30382627952755903,
+      "grad_norm": 0.9637991786003113,
+      "learning_rate": 0.00015791161781152687,
+      "loss": 2.5576,
+      "step": 4939
+    },
+    {
+      "epoch": 0.30388779527559057,
+      "grad_norm": 1.087793231010437,
+      "learning_rate": 0.00015789585181572702,
+      "loss": 2.7549,
+      "step": 4940
+    },
+    {
+      "epoch": 0.30394931102362205,
+      "grad_norm": 1.1802195310592651,
+      "learning_rate": 0.00015788008365494517,
+      "loss": 3.1309,
+      "step": 4941
+    },
+    {
+      "epoch": 0.30401082677165353,
+      "grad_norm": 1.6501368284225464,
+      "learning_rate": 0.00015786431332977092,
+      "loss": 2.5094,
+      "step": 4942
+    },
+    {
+      "epoch": 0.304072342519685,
+      "grad_norm": 1.4137097597122192,
+      "learning_rate": 0.000157848540840794,
+      "loss": 2.6409,
+      "step": 4943
+    },
+    {
+      "epoch": 0.30413385826771655,
+      "grad_norm": 1.223057508468628,
+      "learning_rate": 0.00015783276618860426,
+      "loss": 1.7302,
+      "step": 4944
+    },
+    {
+      "epoch": 0.30419537401574803,
+      "grad_norm": 0.9976323246955872,
+      "learning_rate": 0.00015781698937379155,
+      "loss": 2.8795,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3042568897637795,
+      "grad_norm": 1.0923974514007568,
+      "learning_rate": 0.00015780121039694583,
+      "loss": 2.337,
+      "step": 4946
+    },
+    {
+      "epoch": 0.304318405511811,
+      "grad_norm": 1.569874882698059,
+      "learning_rate": 0.00015778542925865712,
+      "loss": 2.9802,
+      "step": 4947
+    },
+    {
+      "epoch": 0.30437992125984253,
+      "grad_norm": 1.2471206188201904,
+      "learning_rate": 0.00015776964595951557,
+      "loss": 3.0251,
+      "step": 4948
+    },
+    {
+      "epoch": 0.304441437007874,
+      "grad_norm": 1.090998649597168,
+      "learning_rate": 0.00015775386050011144,
+      "loss": 2.8299,
+      "step": 4949
+    },
+    {
+      "epoch": 0.3045029527559055,
+      "grad_norm": 3.5250022411346436,
+      "learning_rate": 0.00015773807288103493,
+      "loss": 2.9418,
+      "step": 4950
+    },
+    {
+      "epoch": 0.304564468503937,
+      "grad_norm": 1.2576251029968262,
+      "learning_rate": 0.0001577222831028765,
+      "loss": 2.4823,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3046259842519685,
+      "grad_norm": 1.4356610774993896,
+      "learning_rate": 0.00015770649116622648,
+      "loss": 2.3831,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3046875,
+      "grad_norm": 1.0928410291671753,
+      "learning_rate": 0.00015769069707167552,
+      "loss": 2.5138,
+      "step": 4953
+    },
+    {
+      "epoch": 0.3047490157480315,
+      "grad_norm": 1.48811936378479,
+      "learning_rate": 0.00015767490081981418,
+      "loss": 2.2883,
+      "step": 4954
+    },
+    {
+      "epoch": 0.304810531496063,
+      "grad_norm": 1.278649091720581,
+      "learning_rate": 0.00015765910241123313,
+      "loss": 2.3182,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3048720472440945,
+      "grad_norm": 2.3534178733825684,
+      "learning_rate": 0.00015764330184652318,
+      "loss": 2.4166,
+      "step": 4956
+    },
+    {
+      "epoch": 0.304933562992126,
+      "grad_norm": 1.6755549907684326,
+      "learning_rate": 0.0001576274991262752,
+      "loss": 2.1275,
+      "step": 4957
+    },
+    {
+      "epoch": 0.30499507874015747,
+      "grad_norm": 1.2691560983657837,
+      "learning_rate": 0.00015761169425108004,
+      "loss": 2.7374,
+      "step": 4958
+    },
+    {
+      "epoch": 0.305056594488189,
+      "grad_norm": 1.212083339691162,
+      "learning_rate": 0.00015759588722152881,
+      "loss": 2.5095,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3051181102362205,
+      "grad_norm": 1.1684248447418213,
+      "learning_rate": 0.00015758007803821252,
+      "loss": 2.3215,
+      "step": 4960
+    },
+    {
+      "epoch": 0.30517962598425197,
+      "grad_norm": 1.2004276514053345,
+      "learning_rate": 0.0001575642667017224,
+      "loss": 3.0992,
+      "step": 4961
+    },
+    {
+      "epoch": 0.30524114173228345,
+      "grad_norm": 1.4189118146896362,
+      "learning_rate": 0.0001575484532126497,
+      "loss": 2.6097,
+      "step": 4962
+    },
+    {
+      "epoch": 0.305302657480315,
+      "grad_norm": 1.1605699062347412,
+      "learning_rate": 0.00015753263757158577,
+      "loss": 2.9488,
+      "step": 4963
+    },
+    {
+      "epoch": 0.30536417322834647,
+      "grad_norm": 1.1894681453704834,
+      "learning_rate": 0.00015751681977912196,
+      "loss": 3.2381,
+      "step": 4964
+    },
+    {
+      "epoch": 0.30542568897637795,
+      "grad_norm": 1.3443187475204468,
+      "learning_rate": 0.00015750099983584983,
+      "loss": 2.6636,
+      "step": 4965
+    },
+    {
+      "epoch": 0.30548720472440943,
+      "grad_norm": 1.246845006942749,
+      "learning_rate": 0.00015748517774236096,
+      "loss": 3.0721,
+      "step": 4966
+    },
+    {
+      "epoch": 0.30554872047244097,
+      "grad_norm": 1.0080130100250244,
+      "learning_rate": 0.000157469353499247,
+      "loss": 3.0471,
+      "step": 4967
+    },
+    {
+      "epoch": 0.30561023622047245,
+      "grad_norm": 1.1883370876312256,
+      "learning_rate": 0.0001574535271070996,
+      "loss": 2.9724,
+      "step": 4968
+    },
+    {
+      "epoch": 0.30567175196850394,
+      "grad_norm": 1.1136255264282227,
+      "learning_rate": 0.00015743769856651073,
+      "loss": 2.907,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3057332677165354,
+      "grad_norm": 1.4167506694793701,
+      "learning_rate": 0.0001574218678780722,
+      "loss": 2.7817,
+      "step": 4970
+    },
+    {
+      "epoch": 0.30579478346456695,
+      "grad_norm": 1.1148821115493774,
+      "learning_rate": 0.000157406035042376,
+      "loss": 2.4327,
+      "step": 4971
+    },
+    {
+      "epoch": 0.30585629921259844,
+      "grad_norm": 1.1239902973175049,
+      "learning_rate": 0.00015739020006001416,
+      "loss": 2.6749,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3059178149606299,
+      "grad_norm": 0.9773595929145813,
+      "learning_rate": 0.0001573743629315789,
+      "loss": 2.8234,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3059793307086614,
+      "grad_norm": 1.7513662576675415,
+      "learning_rate": 0.00015735852365766236,
+      "loss": 2.2961,
+      "step": 4974
+    },
+    {
+      "epoch": 0.30604084645669294,
+      "grad_norm": 1.2198741436004639,
+      "learning_rate": 0.00015734268223885692,
+      "loss": 2.6727,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3061023622047244,
+      "grad_norm": 1.0910673141479492,
+      "learning_rate": 0.00015732683867575488,
+      "loss": 2.9609,
+      "step": 4976
+    },
+    {
+      "epoch": 0.3061638779527559,
+      "grad_norm": 1.013731598854065,
+      "learning_rate": 0.00015731099296894873,
+      "loss": 2.9135,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3062253937007874,
+      "grad_norm": 1.122497320175171,
+      "learning_rate": 0.00015729514511903108,
+      "loss": 2.2983,
+      "step": 4978
+    },
+    {
+      "epoch": 0.3062869094488189,
+      "grad_norm": 1.4958105087280273,
+      "learning_rate": 0.0001572792951265944,
+      "loss": 1.4301,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3063484251968504,
+      "grad_norm": 1.095311164855957,
+      "learning_rate": 0.00015726344299223154,
+      "loss": 2.9066,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3064099409448819,
+      "grad_norm": 0.9803163409233093,
+      "learning_rate": 0.00015724758871653522,
+      "loss": 3.3278,
+      "step": 4981
+    },
+    {
+      "epoch": 0.30647145669291337,
+      "grad_norm": 1.2114437818527222,
+      "learning_rate": 0.0001572317323000983,
+      "loss": 2.711,
+      "step": 4982
+    },
+    {
+      "epoch": 0.3065329724409449,
+      "grad_norm": 1.1234915256500244,
+      "learning_rate": 0.0001572158737435137,
+      "loss": 2.8613,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3065944881889764,
+      "grad_norm": 1.1229381561279297,
+      "learning_rate": 0.0001572000130473745,
+      "loss": 2.469,
+      "step": 4984
+    },
+    {
+      "epoch": 0.30665600393700787,
+      "grad_norm": 1.217988133430481,
+      "learning_rate": 0.00015718415021227374,
+      "loss": 2.7776,
+      "step": 4985
+    },
+    {
+      "epoch": 0.30671751968503935,
+      "grad_norm": 1.203588843345642,
+      "learning_rate": 0.00015716828523880462,
+      "loss": 2.6794,
+      "step": 4986
+    },
+    {
+      "epoch": 0.3067790354330709,
+      "grad_norm": 1.2539359331130981,
+      "learning_rate": 0.00015715241812756044,
+      "loss": 2.4106,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3068405511811024,
+      "grad_norm": 1.1446422338485718,
+      "learning_rate": 0.0001571365488791345,
+      "loss": 2.1046,
+      "step": 4988
+    },
+    {
+      "epoch": 0.30690206692913385,
+      "grad_norm": 1.0805919170379639,
+      "learning_rate": 0.00015712067749412023,
+      "loss": 2.7846,
+      "step": 4989
+    },
+    {
+      "epoch": 0.30696358267716534,
+      "grad_norm": 1.1300969123840332,
+      "learning_rate": 0.00015710480397311113,
+      "loss": 2.5753,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3070250984251969,
+      "grad_norm": 1.2144203186035156,
+      "learning_rate": 0.00015708892831670082,
+      "loss": 2.5607,
+      "step": 4991
+    },
+    {
+      "epoch": 0.30708661417322836,
+      "grad_norm": 2.3810226917266846,
+      "learning_rate": 0.00015707305052548292,
+      "loss": 2.6663,
+      "step": 4992
+    },
+    {
+      "epoch": 0.30714812992125984,
+      "grad_norm": 1.583852767944336,
+      "learning_rate": 0.00015705717060005117,
+      "loss": 2.406,
+      "step": 4993
+    },
+    {
+      "epoch": 0.3072096456692913,
+      "grad_norm": 1.0219244956970215,
+      "learning_rate": 0.00015704128854099938,
+      "loss": 2.6639,
+      "step": 4994
+    },
+    {
+      "epoch": 0.30727116141732286,
+      "grad_norm": 1.257533073425293,
+      "learning_rate": 0.0001570254043489215,
+      "loss": 2.218,
+      "step": 4995
+    },
+    {
+      "epoch": 0.30733267716535434,
+      "grad_norm": 1.1776233911514282,
+      "learning_rate": 0.00015700951802441143,
+      "loss": 3.1125,
+      "step": 4996
+    },
+    {
+      "epoch": 0.3073941929133858,
+      "grad_norm": 1.4189305305480957,
+      "learning_rate": 0.0001569936295680633,
+      "loss": 2.7013,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3074557086614173,
+      "grad_norm": 1.184033751487732,
+      "learning_rate": 0.00015697773898047128,
+      "loss": 2.7092,
+      "step": 4998
+    },
+    {
+      "epoch": 0.30751722440944884,
+      "grad_norm": 1.1817634105682373,
+      "learning_rate": 0.00015696184626222949,
+      "loss": 2.9964,
+      "step": 4999
+    },
+    {
+      "epoch": 0.3075787401574803,
+      "grad_norm": 1.145363688468933,
+      "learning_rate": 0.00015694595141393228,
+      "loss": 2.2565,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3076402559055118,
+      "grad_norm": 1.0647703409194946,
+      "learning_rate": 0.00015693005443617402,
+      "loss": 2.3678,
+      "step": 5001
+    },
+    {
+      "epoch": 0.3077017716535433,
+      "grad_norm": 1.1380219459533691,
+      "learning_rate": 0.0001569141553295492,
+      "loss": 3.0301,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3077632874015748,
+      "grad_norm": 1.3118842840194702,
+      "learning_rate": 0.00015689825409465234,
+      "loss": 2.561,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3078248031496063,
+      "grad_norm": 1.1021641492843628,
+      "learning_rate": 0.00015688235073207802,
+      "loss": 2.5569,
+      "step": 5004
+    },
+    {
+      "epoch": 0.3078863188976378,
+      "grad_norm": 1.0430008172988892,
+      "learning_rate": 0.00015686644524242097,
+      "loss": 2.85,
+      "step": 5005
+    },
+    {
+      "epoch": 0.30794783464566927,
+      "grad_norm": 0.9734707474708557,
+      "learning_rate": 0.00015685053762627598,
+      "loss": 2.7396,
+      "step": 5006
+    },
+    {
+      "epoch": 0.3080093503937008,
+      "grad_norm": 1.103248953819275,
+      "learning_rate": 0.0001568346278842379,
+      "loss": 2.3959,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3080708661417323,
+      "grad_norm": 1.4353207349777222,
+      "learning_rate": 0.00015681871601690164,
+      "loss": 2.7698,
+      "step": 5008
+    },
+    {
+      "epoch": 0.3081323818897638,
+      "grad_norm": 1.0392906665802002,
+      "learning_rate": 0.0001568028020248622,
+      "loss": 3.5247,
+      "step": 5009
+    },
+    {
+      "epoch": 0.30819389763779526,
+      "grad_norm": 1.1516846418380737,
+      "learning_rate": 0.00015678688590871475,
+      "loss": 2.939,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3082554133858268,
+      "grad_norm": 1.0444000959396362,
+      "learning_rate": 0.0001567709676690544,
+      "loss": 2.7217,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3083169291338583,
+      "grad_norm": 0.987623393535614,
+      "learning_rate": 0.00015675504730647643,
+      "loss": 2.2654,
+      "step": 5012
+    },
+    {
+      "epoch": 0.30837844488188976,
+      "grad_norm": 1.1542468070983887,
+      "learning_rate": 0.00015673912482157615,
+      "loss": 2.4777,
+      "step": 5013
+    },
+    {
+      "epoch": 0.30843996062992124,
+      "grad_norm": 1.0824044942855835,
+      "learning_rate": 0.00015672320021494903,
+      "loss": 2.728,
+      "step": 5014
+    },
+    {
+      "epoch": 0.3085014763779528,
+      "grad_norm": 1.0599946975708008,
+      "learning_rate": 0.00015670727348719048,
+      "loss": 3.211,
+      "step": 5015
+    },
+    {
+      "epoch": 0.30856299212598426,
+      "grad_norm": 1.047646164894104,
+      "learning_rate": 0.00015669134463889612,
+      "loss": 3.3774,
+      "step": 5016
+    },
+    {
+      "epoch": 0.30862450787401574,
+      "grad_norm": 1.1178468465805054,
+      "learning_rate": 0.00015667541367066158,
+      "loss": 3.0545,
+      "step": 5017
+    },
+    {
+      "epoch": 0.3086860236220472,
+      "grad_norm": 1.83350670337677,
+      "learning_rate": 0.00015665948058308263,
+      "loss": 1.8551,
+      "step": 5018
+    },
+    {
+      "epoch": 0.30874753937007876,
+      "grad_norm": 1.24369215965271,
+      "learning_rate": 0.000156643545376755,
+      "loss": 2.5399,
+      "step": 5019
+    },
+    {
+      "epoch": 0.30880905511811024,
+      "grad_norm": 1.1607989072799683,
+      "learning_rate": 0.00015662760805227469,
+      "loss": 2.7008,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3088705708661417,
+      "grad_norm": 1.3206210136413574,
+      "learning_rate": 0.00015661166861023754,
+      "loss": 2.6652,
+      "step": 5021
+    },
+    {
+      "epoch": 0.3089320866141732,
+      "grad_norm": 1.1137993335723877,
+      "learning_rate": 0.0001565957270512397,
+      "loss": 2.836,
+      "step": 5022
+    },
+    {
+      "epoch": 0.30899360236220474,
+      "grad_norm": 1.1289048194885254,
+      "learning_rate": 0.00015657978337587722,
+      "loss": 2.2827,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3090551181102362,
+      "grad_norm": 1.1582903861999512,
+      "learning_rate": 0.00015656383758474638,
+      "loss": 2.2694,
+      "step": 5024
+    },
+    {
+      "epoch": 0.3091166338582677,
+      "grad_norm": 1.1440911293029785,
+      "learning_rate": 0.0001565478896784434,
+      "loss": 2.8036,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3091781496062992,
+      "grad_norm": 1.463137149810791,
+      "learning_rate": 0.00015653193965756469,
+      "loss": 3.0544,
+      "step": 5026
+    },
+    {
+      "epoch": 0.30923966535433073,
+      "grad_norm": 1.2239835262298584,
+      "learning_rate": 0.00015651598752270662,
+      "loss": 2.4924,
+      "step": 5027
+    },
+    {
+      "epoch": 0.3093011811023622,
+      "grad_norm": 1.001660704612732,
+      "learning_rate": 0.00015650003327446582,
+      "loss": 2.7216,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3093626968503937,
+      "grad_norm": 1.2627620697021484,
+      "learning_rate": 0.00015648407691343875,
+      "loss": 2.3954,
+      "step": 5029
+    },
+    {
+      "epoch": 0.3094242125984252,
+      "grad_norm": 1.2999376058578491,
+      "learning_rate": 0.00015646811844022222,
+      "loss": 3.1126,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3094857283464567,
+      "grad_norm": 1.1239020824432373,
+      "learning_rate": 0.00015645215785541294,
+      "loss": 2.3057,
+      "step": 5031
+    },
+    {
+      "epoch": 0.3095472440944882,
+      "grad_norm": 1.1862410306930542,
+      "learning_rate": 0.00015643619515960772,
+      "loss": 2.741,
+      "step": 5032
+    },
+    {
+      "epoch": 0.3096087598425197,
+      "grad_norm": 1.2169870138168335,
+      "learning_rate": 0.00015642023035340353,
+      "loss": 2.5968,
+      "step": 5033
+    },
+    {
+      "epoch": 0.30967027559055116,
+      "grad_norm": 1.1207643747329712,
+      "learning_rate": 0.0001564042634373973,
+      "loss": 3.0794,
+      "step": 5034
+    },
+    {
+      "epoch": 0.3097317913385827,
+      "grad_norm": 1.2218523025512695,
+      "learning_rate": 0.00015638829441218613,
+      "loss": 2.5161,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3097933070866142,
+      "grad_norm": 1.0304009914398193,
+      "learning_rate": 0.00015637232327836718,
+      "loss": 3.5939,
+      "step": 5036
+    },
+    {
+      "epoch": 0.30985482283464566,
+      "grad_norm": 1.1389825344085693,
+      "learning_rate": 0.0001563563500365377,
+      "loss": 2.8815,
+      "step": 5037
+    },
+    {
+      "epoch": 0.30991633858267714,
+      "grad_norm": 1.094187617301941,
+      "learning_rate": 0.00015634037468729498,
+      "loss": 2.5569,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3099778543307087,
+      "grad_norm": 1.0825653076171875,
+      "learning_rate": 0.0001563243972312364,
+      "loss": 2.3407,
+      "step": 5039
+    },
+    {
+      "epoch": 0.31003937007874016,
+      "grad_norm": 1.1862939596176147,
+      "learning_rate": 0.00015630841766895944,
+      "loss": 3.0729,
+      "step": 5040
+    },
+    {
+      "epoch": 0.31010088582677164,
+      "grad_norm": 1.145599603652954,
+      "learning_rate": 0.00015629243600106166,
+      "loss": 2.5047,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3101624015748031,
+      "grad_norm": 1.060102939605713,
+      "learning_rate": 0.00015627645222814066,
+      "loss": 2.6227,
+      "step": 5042
+    },
+    {
+      "epoch": 0.31022391732283466,
+      "grad_norm": 1.2564131021499634,
+      "learning_rate": 0.00015626046635079413,
+      "loss": 2.0169,
+      "step": 5043
+    },
+    {
+      "epoch": 0.31028543307086615,
+      "grad_norm": 1.5390654802322388,
+      "learning_rate": 0.0001562444783696199,
+      "loss": 2.7511,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3103469488188976,
+      "grad_norm": 1.1974594593048096,
+      "learning_rate": 0.00015622848828521576,
+      "loss": 3.1403,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3104084645669291,
+      "grad_norm": 1.3076635599136353,
+      "learning_rate": 0.00015621249609817976,
+      "loss": 2.3909,
+      "step": 5046
+    },
+    {
+      "epoch": 0.31046998031496065,
+      "grad_norm": 1.371691107749939,
+      "learning_rate": 0.00015619650180910983,
+      "loss": 2.1176,
+      "step": 5047
+    },
+    {
+      "epoch": 0.31053149606299213,
+      "grad_norm": 1.0407780408859253,
+      "learning_rate": 0.00015618050541860406,
+      "loss": 2.9792,
+      "step": 5048
+    },
+    {
+      "epoch": 0.3105930118110236,
+      "grad_norm": 1.1846849918365479,
+      "learning_rate": 0.00015616450692726073,
+      "loss": 3.2089,
+      "step": 5049
+    },
+    {
+      "epoch": 0.3106545275590551,
+      "grad_norm": 1.1221367120742798,
+      "learning_rate": 0.00015614850633567797,
+      "loss": 2.9485,
+      "step": 5050
+    },
+    {
+      "epoch": 0.31071604330708663,
+      "grad_norm": 1.0790314674377441,
+      "learning_rate": 0.00015613250364445417,
+      "loss": 2.7776,
+      "step": 5051
+    },
+    {
+      "epoch": 0.3107775590551181,
+      "grad_norm": 1.3141484260559082,
+      "learning_rate": 0.00015611649885418776,
+      "loss": 3.2728,
+      "step": 5052
+    },
+    {
+      "epoch": 0.3108390748031496,
+      "grad_norm": 1.1923140287399292,
+      "learning_rate": 0.0001561004919654772,
+      "loss": 2.318,
+      "step": 5053
+    },
+    {
+      "epoch": 0.3109005905511811,
+      "grad_norm": 1.136709451675415,
+      "learning_rate": 0.00015608448297892104,
+      "loss": 2.3121,
+      "step": 5054
+    },
+    {
+      "epoch": 0.3109621062992126,
+      "grad_norm": 1.1647640466690063,
+      "learning_rate": 0.00015606847189511796,
+      "loss": 2.6423,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3110236220472441,
+      "grad_norm": 1.4412775039672852,
+      "learning_rate": 0.0001560524587146667,
+      "loss": 2.7014,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3110851377952756,
+      "grad_norm": 1.2235043048858643,
+      "learning_rate": 0.00015603644343816604,
+      "loss": 2.3335,
+      "step": 5057
+    },
+    {
+      "epoch": 0.31114665354330706,
+      "grad_norm": 1.3097481727600098,
+      "learning_rate": 0.00015602042606621485,
+      "loss": 3.1649,
+      "step": 5058
+    },
+    {
+      "epoch": 0.3112081692913386,
+      "grad_norm": 1.3227657079696655,
+      "learning_rate": 0.00015600440659941208,
+      "loss": 2.9598,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3112696850393701,
+      "grad_norm": 1.220902919769287,
+      "learning_rate": 0.00015598838503835685,
+      "loss": 2.6392,
+      "step": 5060
+    },
+    {
+      "epoch": 0.31133120078740156,
+      "grad_norm": 1.0501551628112793,
+      "learning_rate": 0.00015597236138364818,
+      "loss": 2.8268,
+      "step": 5061
+    },
+    {
+      "epoch": 0.31139271653543305,
+      "grad_norm": 1.1263349056243896,
+      "learning_rate": 0.0001559563356358853,
+      "loss": 2.9938,
+      "step": 5062
+    },
+    {
+      "epoch": 0.3114542322834646,
+      "grad_norm": 1.0979342460632324,
+      "learning_rate": 0.0001559403077956675,
+      "loss": 3.1047,
+      "step": 5063
+    },
+    {
+      "epoch": 0.31151574803149606,
+      "grad_norm": 1.1408517360687256,
+      "learning_rate": 0.00015592427786359413,
+      "loss": 2.7366,
+      "step": 5064
+    },
+    {
+      "epoch": 0.31157726377952755,
+      "grad_norm": 1.3274396657943726,
+      "learning_rate": 0.00015590824584026463,
+      "loss": 2.0179,
+      "step": 5065
+    },
+    {
+      "epoch": 0.31163877952755903,
+      "grad_norm": 1.1601821184158325,
+      "learning_rate": 0.00015589221172627844,
+      "loss": 2.9996,
+      "step": 5066
+    },
+    {
+      "epoch": 0.31170029527559057,
+      "grad_norm": 1.404089093208313,
+      "learning_rate": 0.00015587617552223522,
+      "loss": 3.0257,
+      "step": 5067
+    },
+    {
+      "epoch": 0.31176181102362205,
+      "grad_norm": 1.07805597782135,
+      "learning_rate": 0.0001558601372287346,
+      "loss": 2.8665,
+      "step": 5068
+    },
+    {
+      "epoch": 0.31182332677165353,
+      "grad_norm": 1.4148898124694824,
+      "learning_rate": 0.00015584409684637636,
+      "loss": 2.6242,
+      "step": 5069
+    },
+    {
+      "epoch": 0.311884842519685,
+      "grad_norm": 1.050510287284851,
+      "learning_rate": 0.00015582805437576024,
+      "loss": 2.7846,
+      "step": 5070
+    },
+    {
+      "epoch": 0.31194635826771655,
+      "grad_norm": 1.7260642051696777,
+      "learning_rate": 0.00015581200981748623,
+      "loss": 2.9315,
+      "step": 5071
+    },
+    {
+      "epoch": 0.31200787401574803,
+      "grad_norm": 1.2342300415039062,
+      "learning_rate": 0.00015579596317215425,
+      "loss": 3.2656,
+      "step": 5072
+    },
+    {
+      "epoch": 0.3120693897637795,
+      "grad_norm": 1.1880096197128296,
+      "learning_rate": 0.0001557799144403644,
+      "loss": 2.8233,
+      "step": 5073
+    },
+    {
+      "epoch": 0.312130905511811,
+      "grad_norm": 1.1102819442749023,
+      "learning_rate": 0.00015576386362271678,
+      "loss": 2.9284,
+      "step": 5074
+    },
+    {
+      "epoch": 0.31219242125984253,
+      "grad_norm": 1.2920249700546265,
+      "learning_rate": 0.0001557478107198116,
+      "loss": 2.7078,
+      "step": 5075
+    },
+    {
+      "epoch": 0.312253937007874,
+      "grad_norm": 1.2453279495239258,
+      "learning_rate": 0.00015573175573224915,
+      "loss": 2.8142,
+      "step": 5076
+    },
+    {
+      "epoch": 0.3123154527559055,
+      "grad_norm": 1.2635228633880615,
+      "learning_rate": 0.0001557156986606298,
+      "loss": 2.8092,
+      "step": 5077
+    },
+    {
+      "epoch": 0.312376968503937,
+      "grad_norm": 1.0569782257080078,
+      "learning_rate": 0.000155699639505554,
+      "loss": 2.9217,
+      "step": 5078
+    },
+    {
+      "epoch": 0.3124384842519685,
+      "grad_norm": 1.2359199523925781,
+      "learning_rate": 0.0001556835782676223,
+      "loss": 3.0476,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 1.1401257514953613,
+      "learning_rate": 0.0001556675149474353,
+      "loss": 2.6683,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3125615157480315,
+      "grad_norm": 1.019160270690918,
+      "learning_rate": 0.00015565144954559357,
+      "loss": 2.8995,
+      "step": 5081
+    },
+    {
+      "epoch": 0.312623031496063,
+      "grad_norm": 1.2784544229507446,
+      "learning_rate": 0.000155635382062698,
+      "loss": 2.8359,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3126845472440945,
+      "grad_norm": 1.2276674509048462,
+      "learning_rate": 0.00015561931249934933,
+      "loss": 2.6039,
+      "step": 5083
+    },
+    {
+      "epoch": 0.312746062992126,
+      "grad_norm": 1.1401909589767456,
+      "learning_rate": 0.00015560324085614856,
+      "loss": 2.8365,
+      "step": 5084
+    },
+    {
+      "epoch": 0.31280757874015747,
+      "grad_norm": 1.1890449523925781,
+      "learning_rate": 0.0001555871671336966,
+      "loss": 2.5614,
+      "step": 5085
+    },
+    {
+      "epoch": 0.312869094488189,
+      "grad_norm": 1.1198787689208984,
+      "learning_rate": 0.00015557109133259458,
+      "loss": 2.0399,
+      "step": 5086
+    },
+    {
+      "epoch": 0.3129306102362205,
+      "grad_norm": 1.2756470441818237,
+      "learning_rate": 0.00015555501345344358,
+      "loss": 2.3112,
+      "step": 5087
+    },
+    {
+      "epoch": 0.31299212598425197,
+      "grad_norm": 1.0000739097595215,
+      "learning_rate": 0.00015553893349684491,
+      "loss": 2.1549,
+      "step": 5088
+    },
+    {
+      "epoch": 0.31305364173228345,
+      "grad_norm": 1.0322518348693848,
+      "learning_rate": 0.0001555228514633998,
+      "loss": 2.8699,
+      "step": 5089
+    },
+    {
+      "epoch": 0.313115157480315,
+      "grad_norm": 1.2883877754211426,
+      "learning_rate": 0.00015550676735370965,
+      "loss": 2.8784,
+      "step": 5090
+    },
+    {
+      "epoch": 0.31317667322834647,
+      "grad_norm": 1.354669213294983,
+      "learning_rate": 0.0001554906811683759,
+      "loss": 2.4169,
+      "step": 5091
+    },
+    {
+      "epoch": 0.31323818897637795,
+      "grad_norm": 1.212921380996704,
+      "learning_rate": 0.0001554745929080001,
+      "loss": 2.7589,
+      "step": 5092
+    },
+    {
+      "epoch": 0.31329970472440943,
+      "grad_norm": 1.4567354917526245,
+      "learning_rate": 0.0001554585025731839,
+      "loss": 2.1736,
+      "step": 5093
+    },
+    {
+      "epoch": 0.31336122047244097,
+      "grad_norm": 1.0887346267700195,
+      "learning_rate": 0.0001554424101645289,
+      "loss": 2.9931,
+      "step": 5094
+    },
+    {
+      "epoch": 0.31342273622047245,
+      "grad_norm": 1.4969717264175415,
+      "learning_rate": 0.00015542631568263694,
+      "loss": 2.6493,
+      "step": 5095
+    },
+    {
+      "epoch": 0.31348425196850394,
+      "grad_norm": 1.3519649505615234,
+      "learning_rate": 0.00015541021912810984,
+      "loss": 3.308,
+      "step": 5096
+    },
+    {
+      "epoch": 0.3135457677165354,
+      "grad_norm": 1.3046905994415283,
+      "learning_rate": 0.0001553941205015495,
+      "loss": 2.4982,
+      "step": 5097
+    },
+    {
+      "epoch": 0.31360728346456695,
+      "grad_norm": 1.3364003896713257,
+      "learning_rate": 0.00015537801980355798,
+      "loss": 3.1011,
+      "step": 5098
+    },
+    {
+      "epoch": 0.31366879921259844,
+      "grad_norm": 1.074690580368042,
+      "learning_rate": 0.00015536191703473726,
+      "loss": 2.8005,
+      "step": 5099
+    },
+    {
+      "epoch": 0.3137303149606299,
+      "grad_norm": 1.1206022500991821,
+      "learning_rate": 0.00015534581219568957,
+      "loss": 3.3627,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3137918307086614,
+      "grad_norm": 1.242507815361023,
+      "learning_rate": 0.00015532970528701716,
+      "loss": 3.0591,
+      "step": 5101
+    },
+    {
+      "epoch": 0.31385334645669294,
+      "grad_norm": 1.2259441614151,
+      "learning_rate": 0.00015531359630932228,
+      "loss": 2.6778,
+      "step": 5102
+    },
+    {
+      "epoch": 0.3139148622047244,
+      "grad_norm": 1.1966248750686646,
+      "learning_rate": 0.00015529748526320732,
+      "loss": 3.1805,
+      "step": 5103
+    },
+    {
+      "epoch": 0.3139763779527559,
+      "grad_norm": 1.1150904893875122,
+      "learning_rate": 0.00015528137214927473,
+      "loss": 2.8323,
+      "step": 5104
+    },
+    {
+      "epoch": 0.3140378937007874,
+      "grad_norm": 1.2046905755996704,
+      "learning_rate": 0.00015526525696812711,
+      "loss": 2.8125,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3140994094488189,
+      "grad_norm": 1.175865650177002,
+      "learning_rate": 0.00015524913972036708,
+      "loss": 2.4741,
+      "step": 5106
+    },
+    {
+      "epoch": 0.3141609251968504,
+      "grad_norm": 1.1325241327285767,
+      "learning_rate": 0.00015523302040659727,
+      "loss": 2.8021,
+      "step": 5107
+    },
+    {
+      "epoch": 0.3142224409448819,
+      "grad_norm": 1.1499979496002197,
+      "learning_rate": 0.00015521689902742047,
+      "loss": 2.8616,
+      "step": 5108
+    },
+    {
+      "epoch": 0.31428395669291337,
+      "grad_norm": 1.0621825456619263,
+      "learning_rate": 0.00015520077558343957,
+      "loss": 2.706,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3143454724409449,
+      "grad_norm": 1.1516525745391846,
+      "learning_rate": 0.00015518465007525746,
+      "loss": 3.08,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3144069881889764,
+      "grad_norm": 1.0844038724899292,
+      "learning_rate": 0.00015516852250347713,
+      "loss": 2.9461,
+      "step": 5111
+    },
+    {
+      "epoch": 0.31446850393700787,
+      "grad_norm": 1.4454748630523682,
+      "learning_rate": 0.00015515239286870172,
+      "loss": 2.246,
+      "step": 5112
+    },
+    {
+      "epoch": 0.31453001968503935,
+      "grad_norm": 1.0956681966781616,
+      "learning_rate": 0.00015513626117153435,
+      "loss": 2.3785,
+      "step": 5113
+    },
+    {
+      "epoch": 0.3145915354330709,
+      "grad_norm": 1.0651484727859497,
+      "learning_rate": 0.00015512012741257826,
+      "loss": 2.3798,
+      "step": 5114
+    },
+    {
+      "epoch": 0.3146530511811024,
+      "grad_norm": 1.0153863430023193,
+      "learning_rate": 0.00015510399159243676,
+      "loss": 2.4433,
+      "step": 5115
+    },
+    {
+      "epoch": 0.31471456692913385,
+      "grad_norm": 1.1630622148513794,
+      "learning_rate": 0.00015508785371171323,
+      "loss": 2.7015,
+      "step": 5116
+    },
+    {
+      "epoch": 0.31477608267716534,
+      "grad_norm": 1.0745515823364258,
+      "learning_rate": 0.00015507171377101116,
+      "loss": 3.107,
+      "step": 5117
+    },
+    {
+      "epoch": 0.3148375984251969,
+      "grad_norm": 1.3272719383239746,
+      "learning_rate": 0.0001550555717709341,
+      "loss": 3.4584,
+      "step": 5118
+    },
+    {
+      "epoch": 0.31489911417322836,
+      "grad_norm": 1.014405608177185,
+      "learning_rate": 0.00015503942771208563,
+      "loss": 2.9335,
+      "step": 5119
+    },
+    {
+      "epoch": 0.31496062992125984,
+      "grad_norm": 1.6725507974624634,
+      "learning_rate": 0.00015502328159506948,
+      "loss": 2.9332,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3150221456692913,
+      "grad_norm": 1.4744592905044556,
+      "learning_rate": 0.0001550071334204894,
+      "loss": 2.1802,
+      "step": 5121
+    },
+    {
+      "epoch": 0.31508366141732286,
+      "grad_norm": 1.1695510149002075,
+      "learning_rate": 0.0001549909831889493,
+      "loss": 2.7428,
+      "step": 5122
+    },
+    {
+      "epoch": 0.31514517716535434,
+      "grad_norm": 1.1932672262191772,
+      "learning_rate": 0.00015497483090105304,
+      "loss": 3.346,
+      "step": 5123
+    },
+    {
+      "epoch": 0.3152066929133858,
+      "grad_norm": 1.0129542350769043,
+      "learning_rate": 0.00015495867655740465,
+      "loss": 3.1668,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3152682086614173,
+      "grad_norm": 1.036478877067566,
+      "learning_rate": 0.00015494252015860822,
+      "loss": 2.6602,
+      "step": 5125
+    },
+    {
+      "epoch": 0.31532972440944884,
+      "grad_norm": 1.3557771444320679,
+      "learning_rate": 0.0001549263617052679,
+      "loss": 2.8435,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3153912401574803,
+      "grad_norm": 1.1496968269348145,
+      "learning_rate": 0.00015491020119798792,
+      "loss": 2.6379,
+      "step": 5127
+    },
+    {
+      "epoch": 0.3154527559055118,
+      "grad_norm": 1.0594522953033447,
+      "learning_rate": 0.00015489403863737264,
+      "loss": 2.8764,
+      "step": 5128
+    },
+    {
+      "epoch": 0.3155142716535433,
+      "grad_norm": 1.1536942720413208,
+      "learning_rate": 0.00015487787402402637,
+      "loss": 2.4829,
+      "step": 5129
+    },
+    {
+      "epoch": 0.3155757874015748,
+      "grad_norm": 1.0434961318969727,
+      "learning_rate": 0.00015486170735855366,
+      "loss": 2.7115,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3156373031496063,
+      "grad_norm": 1.1584323644638062,
+      "learning_rate": 0.00015484553864155898,
+      "loss": 2.5971,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3156988188976378,
+      "grad_norm": 0.996439516544342,
+      "learning_rate": 0.00015482936787364698,
+      "loss": 2.3309,
+      "step": 5132
+    },
+    {
+      "epoch": 0.31576033464566927,
+      "grad_norm": 1.1407891511917114,
+      "learning_rate": 0.00015481319505542237,
+      "loss": 2.8218,
+      "step": 5133
+    },
+    {
+      "epoch": 0.3158218503937008,
+      "grad_norm": 1.0651286840438843,
+      "learning_rate": 0.00015479702018748992,
+      "loss": 2.1685,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3158833661417323,
+      "grad_norm": 1.233332872390747,
+      "learning_rate": 0.00015478084327045445,
+      "loss": 2.4943,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3159448818897638,
+      "grad_norm": 1.2463935613632202,
+      "learning_rate": 0.00015476466430492094,
+      "loss": 2.6432,
+      "step": 5136
+    },
+    {
+      "epoch": 0.31600639763779526,
+      "grad_norm": 1.14547598361969,
+      "learning_rate": 0.00015474848329149435,
+      "loss": 2.6535,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3160679133858268,
+      "grad_norm": 1.2023497819900513,
+      "learning_rate": 0.00015473230023077975,
+      "loss": 2.4204,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3161294291338583,
+      "grad_norm": 1.1173592805862427,
+      "learning_rate": 0.00015471611512338234,
+      "loss": 3.0195,
+      "step": 5139
+    },
+    {
+      "epoch": 0.31619094488188976,
+      "grad_norm": 1.1788824796676636,
+      "learning_rate": 0.00015469992796990733,
+      "loss": 2.2387,
+      "step": 5140
+    },
+    {
+      "epoch": 0.31625246062992124,
+      "grad_norm": 1.1276782751083374,
+      "learning_rate": 0.00015468373877096,
+      "loss": 3.1789,
+      "step": 5141
+    },
+    {
+      "epoch": 0.3163139763779528,
+      "grad_norm": 1.1502629518508911,
+      "learning_rate": 0.00015466754752714582,
+      "loss": 2.1846,
+      "step": 5142
+    },
+    {
+      "epoch": 0.31637549212598426,
+      "grad_norm": 1.0899603366851807,
+      "learning_rate": 0.00015465135423907013,
+      "loss": 3.3993,
+      "step": 5143
+    },
+    {
+      "epoch": 0.31643700787401574,
+      "grad_norm": 1.4058409929275513,
+      "learning_rate": 0.00015463515890733857,
+      "loss": 2.1099,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3164985236220472,
+      "grad_norm": 1.1758254766464233,
+      "learning_rate": 0.00015461896153255673,
+      "loss": 2.9144,
+      "step": 5145
+    },
+    {
+      "epoch": 0.31656003937007876,
+      "grad_norm": 1.1036620140075684,
+      "learning_rate": 0.00015460276211533026,
+      "loss": 3.007,
+      "step": 5146
+    },
+    {
+      "epoch": 0.31662155511811024,
+      "grad_norm": 2.05633282661438,
+      "learning_rate": 0.000154586560656265,
+      "loss": 2.313,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3166830708661417,
+      "grad_norm": 1.0357699394226074,
+      "learning_rate": 0.00015457035715596673,
+      "loss": 3.477,
+      "step": 5148
+    },
+    {
+      "epoch": 0.3167445866141732,
+      "grad_norm": 1.306477665901184,
+      "learning_rate": 0.0001545541516150414,
+      "loss": 2.781,
+      "step": 5149
+    },
+    {
+      "epoch": 0.31680610236220474,
+      "grad_norm": 1.180249810218811,
+      "learning_rate": 0.000154537944034095,
+      "loss": 3.0342,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3168676181102362,
+      "grad_norm": 1.276918888092041,
+      "learning_rate": 0.0001545217344137336,
+      "loss": 3.3044,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3169291338582677,
+      "grad_norm": 1.59418785572052,
+      "learning_rate": 0.00015450552275456336,
+      "loss": 2.4448,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3169906496062992,
+      "grad_norm": 1.0821864604949951,
+      "learning_rate": 0.00015448930905719052,
+      "loss": 2.6449,
+      "step": 5153
+    },
+    {
+      "epoch": 0.31705216535433073,
+      "grad_norm": 1.053840160369873,
+      "learning_rate": 0.00015447309332222134,
+      "loss": 2.8567,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3171136811023622,
+      "grad_norm": 1.0531610250473022,
+      "learning_rate": 0.00015445687555026223,
+      "loss": 2.6233,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3171751968503937,
+      "grad_norm": 1.0734477043151855,
+      "learning_rate": 0.0001544406557419196,
+      "loss": 2.7251,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3172367125984252,
+      "grad_norm": 1.1375876665115356,
+      "learning_rate": 0.00015442443389780008,
+      "loss": 2.3755,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3172982283464567,
+      "grad_norm": 1.0814098119735718,
+      "learning_rate": 0.00015440821001851015,
+      "loss": 3.3607,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3173597440944882,
+      "grad_norm": 1.2837457656860352,
+      "learning_rate": 0.00015439198410465655,
+      "loss": 3.0095,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3174212598425197,
+      "grad_norm": 1.053125023841858,
+      "learning_rate": 0.0001543757561568461,
+      "loss": 2.8584,
+      "step": 5160
+    },
+    {
+      "epoch": 0.31748277559055116,
+      "grad_norm": 1.150827169418335,
+      "learning_rate": 0.00015435952617568553,
+      "loss": 1.7931,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3175442913385827,
+      "grad_norm": 1.2026658058166504,
+      "learning_rate": 0.0001543432941617818,
+      "loss": 2.6671,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3176058070866142,
+      "grad_norm": 1.2518446445465088,
+      "learning_rate": 0.00015432706011574195,
+      "loss": 2.8927,
+      "step": 5163
+    },
+    {
+      "epoch": 0.31766732283464566,
+      "grad_norm": 1.3294403553009033,
+      "learning_rate": 0.00015431082403817293,
+      "loss": 2.9957,
+      "step": 5164
+    },
+    {
+      "epoch": 0.31772883858267714,
+      "grad_norm": 1.1822839975357056,
+      "learning_rate": 0.00015429458592968197,
+      "loss": 2.3587,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3177903543307087,
+      "grad_norm": 1.5753675699234009,
+      "learning_rate": 0.00015427834579087623,
+      "loss": 2.5201,
+      "step": 5166
+    },
+    {
+      "epoch": 0.31785187007874016,
+      "grad_norm": 1.0831626653671265,
+      "learning_rate": 0.00015426210362236303,
+      "loss": 2.6416,
+      "step": 5167
+    },
+    {
+      "epoch": 0.31791338582677164,
+      "grad_norm": 1.131073236465454,
+      "learning_rate": 0.00015424585942474975,
+      "loss": 2.6272,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3179749015748031,
+      "grad_norm": 1.223253846168518,
+      "learning_rate": 0.00015422961319864375,
+      "loss": 2.5635,
+      "step": 5169
+    },
+    {
+      "epoch": 0.31803641732283466,
+      "grad_norm": 1.119757056236267,
+      "learning_rate": 0.00015421336494465267,
+      "loss": 2.698,
+      "step": 5170
+    },
+    {
+      "epoch": 0.31809793307086615,
+      "grad_norm": 1.1726499795913696,
+      "learning_rate": 0.000154197114663384,
+      "loss": 2.8842,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3181594488188976,
+      "grad_norm": 1.3384331464767456,
+      "learning_rate": 0.00015418086235544545,
+      "loss": 2.572,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3182209645669291,
+      "grad_norm": 1.0606106519699097,
+      "learning_rate": 0.0001541646080214448,
+      "loss": 2.6061,
+      "step": 5173
+    },
+    {
+      "epoch": 0.31828248031496065,
+      "grad_norm": 1.4063708782196045,
+      "learning_rate": 0.00015414835166198984,
+      "loss": 2.5987,
+      "step": 5174
+    },
+    {
+      "epoch": 0.31834399606299213,
+      "grad_norm": 1.2853461503982544,
+      "learning_rate": 0.00015413209327768846,
+      "loss": 2.4882,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3184055118110236,
+      "grad_norm": 1.2160193920135498,
+      "learning_rate": 0.00015411583286914865,
+      "loss": 2.9485,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3184670275590551,
+      "grad_norm": 1.083579659461975,
+      "learning_rate": 0.00015409957043697846,
+      "loss": 2.3843,
+      "step": 5177
+    },
+    {
+      "epoch": 0.31852854330708663,
+      "grad_norm": 1.6811126470565796,
+      "learning_rate": 0.00015408330598178595,
+      "loss": 4.2251,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3185900590551181,
+      "grad_norm": 1.2879703044891357,
+      "learning_rate": 0.0001540670395041794,
+      "loss": 2.5734,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3186515748031496,
+      "grad_norm": 1.0302507877349854,
+      "learning_rate": 0.0001540507710047671,
+      "loss": 2.8876,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3187130905511811,
+      "grad_norm": 1.072900652885437,
+      "learning_rate": 0.0001540345004841573,
+      "loss": 2.2951,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3187746062992126,
+      "grad_norm": 1.1929172277450562,
+      "learning_rate": 0.0001540182279429585,
+      "loss": 3.2698,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3188361220472441,
+      "grad_norm": 1.214979887008667,
+      "learning_rate": 0.0001540019533817792,
+      "loss": 3.2008,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3188976377952756,
+      "grad_norm": 1.1639785766601562,
+      "learning_rate": 0.00015398567680122795,
+      "loss": 2.6939,
+      "step": 5184
+    },
+    {
+      "epoch": 0.31895915354330706,
+      "grad_norm": 1.182238221168518,
+      "learning_rate": 0.00015396939820191342,
+      "loss": 2.4416,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3190206692913386,
+      "grad_norm": 1.2369346618652344,
+      "learning_rate": 0.00015395311758444437,
+      "loss": 3.094,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3190821850393701,
+      "grad_norm": 1.0608497858047485,
+      "learning_rate": 0.00015393683494942955,
+      "loss": 2.7008,
+      "step": 5187
+    },
+    {
+      "epoch": 0.31914370078740156,
+      "grad_norm": 1.1972993612289429,
+      "learning_rate": 0.00015392055029747788,
+      "loss": 2.1395,
+      "step": 5188
+    },
+    {
+      "epoch": 0.31920521653543305,
+      "grad_norm": 0.9763641953468323,
+      "learning_rate": 0.0001539042636291983,
+      "loss": 2.8263,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3192667322834646,
+      "grad_norm": 1.2631070613861084,
+      "learning_rate": 0.0001538879749451998,
+      "loss": 2.7265,
+      "step": 5190
+    },
+    {
+      "epoch": 0.31932824803149606,
+      "grad_norm": 0.9234073758125305,
+      "learning_rate": 0.00015387168424609154,
+      "loss": 2.9078,
+      "step": 5191
+    },
+    {
+      "epoch": 0.31938976377952755,
+      "grad_norm": 1.1408380270004272,
+      "learning_rate": 0.00015385539153248273,
+      "loss": 2.9932,
+      "step": 5192
+    },
+    {
+      "epoch": 0.31945127952755903,
+      "grad_norm": 1.0665607452392578,
+      "learning_rate": 0.00015383909680498253,
+      "loss": 3.3489,
+      "step": 5193
+    },
+    {
+      "epoch": 0.31951279527559057,
+      "grad_norm": 1.1575465202331543,
+      "learning_rate": 0.00015382280006420035,
+      "loss": 2.4184,
+      "step": 5194
+    },
+    {
+      "epoch": 0.31957431102362205,
+      "grad_norm": 1.3081769943237305,
+      "learning_rate": 0.00015380650131074556,
+      "loss": 2.8258,
+      "step": 5195
+    },
+    {
+      "epoch": 0.31963582677165353,
+      "grad_norm": 1.1539462804794312,
+      "learning_rate": 0.00015379020054522767,
+      "loss": 2.6779,
+      "step": 5196
+    },
+    {
+      "epoch": 0.319697342519685,
+      "grad_norm": 1.1435942649841309,
+      "learning_rate": 0.0001537738977682562,
+      "loss": 1.9498,
+      "step": 5197
+    },
+    {
+      "epoch": 0.31975885826771655,
+      "grad_norm": 1.350520372390747,
+      "learning_rate": 0.00015375759298044082,
+      "loss": 2.4243,
+      "step": 5198
+    },
+    {
+      "epoch": 0.31982037401574803,
+      "grad_norm": 1.663401484489441,
+      "learning_rate": 0.0001537412861823912,
+      "loss": 2.5964,
+      "step": 5199
+    },
+    {
+      "epoch": 0.3198818897637795,
+      "grad_norm": 1.2314056158065796,
+      "learning_rate": 0.00015372497737471718,
+      "loss": 2.0413,
+      "step": 5200
+    },
+    {
+      "epoch": 0.319943405511811,
+      "grad_norm": 1.400802493095398,
+      "learning_rate": 0.00015370866655802857,
+      "loss": 2.2563,
+      "step": 5201
+    },
+    {
+      "epoch": 0.32000492125984253,
+      "grad_norm": 1.3720860481262207,
+      "learning_rate": 0.00015369235373293534,
+      "loss": 1.7868,
+      "step": 5202
+    },
+    {
+      "epoch": 0.320066437007874,
+      "grad_norm": 1.2554841041564941,
+      "learning_rate": 0.00015367603890004743,
+      "loss": 3.1821,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3201279527559055,
+      "grad_norm": 1.3925338983535767,
+      "learning_rate": 0.00015365972205997502,
+      "loss": 3.2453,
+      "step": 5204
+    },
+    {
+      "epoch": 0.320189468503937,
+      "grad_norm": 1.187589168548584,
+      "learning_rate": 0.00015364340321332817,
+      "loss": 2.4053,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3202509842519685,
+      "grad_norm": 1.504937767982483,
+      "learning_rate": 0.0001536270823607172,
+      "loss": 2.3124,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3203125,
+      "grad_norm": 1.2485361099243164,
+      "learning_rate": 0.00015361075950275237,
+      "loss": 2.145,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3203740157480315,
+      "grad_norm": 1.0649597644805908,
+      "learning_rate": 0.00015359443464004407,
+      "loss": 2.9825,
+      "step": 5208
+    },
+    {
+      "epoch": 0.320435531496063,
+      "grad_norm": 1.1464604139328003,
+      "learning_rate": 0.0001535781077732028,
+      "loss": 3.0766,
+      "step": 5209
+    },
+    {
+      "epoch": 0.3204970472440945,
+      "grad_norm": 1.3525128364562988,
+      "learning_rate": 0.000153561778902839,
+      "loss": 2.7056,
+      "step": 5210
+    },
+    {
+      "epoch": 0.320558562992126,
+      "grad_norm": 1.1054641008377075,
+      "learning_rate": 0.00015354544802956342,
+      "loss": 2.9457,
+      "step": 5211
+    },
+    {
+      "epoch": 0.32062007874015747,
+      "grad_norm": 1.0405795574188232,
+      "learning_rate": 0.0001535291151539866,
+      "loss": 3.1587,
+      "step": 5212
+    },
+    {
+      "epoch": 0.320681594488189,
+      "grad_norm": 1.1991597414016724,
+      "learning_rate": 0.00015351278027671937,
+      "loss": 2.3659,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3207431102362205,
+      "grad_norm": 1.232959508895874,
+      "learning_rate": 0.00015349644339837255,
+      "loss": 3.5157,
+      "step": 5214
+    },
+    {
+      "epoch": 0.32080462598425197,
+      "grad_norm": 1.3933085203170776,
+      "learning_rate": 0.00015348010451955707,
+      "loss": 2.6755,
+      "step": 5215
+    },
+    {
+      "epoch": 0.32086614173228345,
+      "grad_norm": 1.248811960220337,
+      "learning_rate": 0.00015346376364088385,
+      "loss": 2.7057,
+      "step": 5216
+    },
+    {
+      "epoch": 0.320927657480315,
+      "grad_norm": 1.0312199592590332,
+      "learning_rate": 0.00015344742076296404,
+      "loss": 2.9386,
+      "step": 5217
+    },
+    {
+      "epoch": 0.32098917322834647,
+      "grad_norm": 1.1867161989212036,
+      "learning_rate": 0.0001534310758864087,
+      "loss": 2.7891,
+      "step": 5218
+    },
+    {
+      "epoch": 0.32105068897637795,
+      "grad_norm": 1.199964165687561,
+      "learning_rate": 0.00015341472901182909,
+      "loss": 3.032,
+      "step": 5219
+    },
+    {
+      "epoch": 0.32111220472440943,
+      "grad_norm": 1.3150421380996704,
+      "learning_rate": 0.00015339838013983642,
+      "loss": 2.4721,
+      "step": 5220
+    },
+    {
+      "epoch": 0.32117372047244097,
+      "grad_norm": 1.4395772218704224,
+      "learning_rate": 0.00015338202927104214,
+      "loss": 2.255,
+      "step": 5221
+    },
+    {
+      "epoch": 0.32123523622047245,
+      "grad_norm": 1.10757577419281,
+      "learning_rate": 0.00015336567640605755,
+      "loss": 2.4062,
+      "step": 5222
+    },
+    {
+      "epoch": 0.32129675196850394,
+      "grad_norm": 1.1340880393981934,
+      "learning_rate": 0.00015334932154549433,
+      "loss": 2.9254,
+      "step": 5223
+    },
+    {
+      "epoch": 0.3213582677165354,
+      "grad_norm": 1.1758360862731934,
+      "learning_rate": 0.00015333296468996393,
+      "loss": 2.8553,
+      "step": 5224
+    },
+    {
+      "epoch": 0.32141978346456695,
+      "grad_norm": 1.2455058097839355,
+      "learning_rate": 0.00015331660584007802,
+      "loss": 3.3667,
+      "step": 5225
+    },
+    {
+      "epoch": 0.32148129921259844,
+      "grad_norm": 1.1944704055786133,
+      "learning_rate": 0.00015330024499644838,
+      "loss": 3.1745,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3215428149606299,
+      "grad_norm": 1.260232925415039,
+      "learning_rate": 0.0001532838821596868,
+      "loss": 2.6143,
+      "step": 5227
+    },
+    {
+      "epoch": 0.3216043307086614,
+      "grad_norm": 1.054417610168457,
+      "learning_rate": 0.00015326751733040514,
+      "loss": 3.1765,
+      "step": 5228
+    },
+    {
+      "epoch": 0.32166584645669294,
+      "grad_norm": 1.090609073638916,
+      "learning_rate": 0.00015325115050921533,
+      "loss": 2.6675,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3217273622047244,
+      "grad_norm": 1.19331693649292,
+      "learning_rate": 0.00015323478169672946,
+      "loss": 2.8291,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3217888779527559,
+      "grad_norm": 1.0912829637527466,
+      "learning_rate": 0.00015321841089355957,
+      "loss": 2.9223,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3218503937007874,
+      "grad_norm": 1.1780345439910889,
+      "learning_rate": 0.0001532020381003179,
+      "loss": 3.1415,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3219119094488189,
+      "grad_norm": 1.0986944437026978,
+      "learning_rate": 0.00015318566331761665,
+      "loss": 2.9796,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3219734251968504,
+      "grad_norm": 1.3491836786270142,
+      "learning_rate": 0.00015316928654606813,
+      "loss": 2.5524,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3220349409448819,
+      "grad_norm": 1.237416386604309,
+      "learning_rate": 0.00015315290778628482,
+      "loss": 2.6672,
+      "step": 5235
+    },
+    {
+      "epoch": 0.32209645669291337,
+      "grad_norm": 1.271661400794983,
+      "learning_rate": 0.00015313652703887915,
+      "loss": 2.0179,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3221579724409449,
+      "grad_norm": 1.1317790746688843,
+      "learning_rate": 0.00015312014430446364,
+      "loss": 2.697,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3222194881889764,
+      "grad_norm": 1.164856195449829,
+      "learning_rate": 0.00015310375958365094,
+      "loss": 2.6417,
+      "step": 5238
+    },
+    {
+      "epoch": 0.32228100393700787,
+      "grad_norm": 1.4392765760421753,
+      "learning_rate": 0.0001530873728770538,
+      "loss": 2.4449,
+      "step": 5239
+    },
+    {
+      "epoch": 0.32234251968503935,
+      "grad_norm": 1.660623550415039,
+      "learning_rate": 0.00015307098418528486,
+      "loss": 2.753,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3224040354330709,
+      "grad_norm": 1.0715113878250122,
+      "learning_rate": 0.0001530545935089571,
+      "loss": 2.8595,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3224655511811024,
+      "grad_norm": 1.0867308378219604,
+      "learning_rate": 0.00015303820084868334,
+      "loss": 2.8795,
+      "step": 5242
+    },
+    {
+      "epoch": 0.32252706692913385,
+      "grad_norm": 1.1036630868911743,
+      "learning_rate": 0.00015302180620507664,
+      "loss": 2.7434,
+      "step": 5243
+    },
+    {
+      "epoch": 0.32258858267716534,
+      "grad_norm": 1.1693278551101685,
+      "learning_rate": 0.00015300540957875004,
+      "loss": 2.831,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3226500984251969,
+      "grad_norm": 1.2517393827438354,
+      "learning_rate": 0.0001529890109703167,
+      "loss": 2.4072,
+      "step": 5245
+    },
+    {
+      "epoch": 0.32271161417322836,
+      "grad_norm": 1.3229448795318604,
+      "learning_rate": 0.0001529726103803898,
+      "loss": 3.0092,
+      "step": 5246
+    },
+    {
+      "epoch": 0.32277312992125984,
+      "grad_norm": 1.3072962760925293,
+      "learning_rate": 0.0001529562078095827,
+      "loss": 2.2264,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3228346456692913,
+      "grad_norm": 1.1157249212265015,
+      "learning_rate": 0.0001529398032585087,
+      "loss": 2.5124,
+      "step": 5248
+    },
+    {
+      "epoch": 0.32289616141732286,
+      "grad_norm": 1.264760971069336,
+      "learning_rate": 0.00015292339672778125,
+      "loss": 2.9556,
+      "step": 5249
+    },
+    {
+      "epoch": 0.32295767716535434,
+      "grad_norm": 1.1906821727752686,
+      "learning_rate": 0.0001529069882180139,
+      "loss": 2.6385,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3230191929133858,
+      "grad_norm": 1.3362345695495605,
+      "learning_rate": 0.00015289057772982013,
+      "loss": 2.5581,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3230807086614173,
+      "grad_norm": 1.0462592840194702,
+      "learning_rate": 0.00015287416526381376,
+      "loss": 2.9339,
+      "step": 5252
+    },
+    {
+      "epoch": 0.32314222440944884,
+      "grad_norm": 1.2769001722335815,
+      "learning_rate": 0.00015285775082060844,
+      "loss": 2.1607,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3232037401574803,
+      "grad_norm": 1.3508915901184082,
+      "learning_rate": 0.0001528413344008179,
+      "loss": 2.958,
+      "step": 5254
+    },
+    {
+      "epoch": 0.3232652559055118,
+      "grad_norm": 1.144619107246399,
+      "learning_rate": 0.0001528249160050562,
+      "loss": 2.97,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3233267716535433,
+      "grad_norm": 1.3087143898010254,
+      "learning_rate": 0.00015280849563393715,
+      "loss": 3.3771,
+      "step": 5256
+    },
+    {
+      "epoch": 0.3233882874015748,
+      "grad_norm": 1.2242597341537476,
+      "learning_rate": 0.0001527920732880748,
+      "loss": 2.7315,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3234498031496063,
+      "grad_norm": 1.2692580223083496,
+      "learning_rate": 0.00015277564896808336,
+      "loss": 2.7366,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3235113188976378,
+      "grad_norm": 0.9895825386047363,
+      "learning_rate": 0.00015275922267457692,
+      "loss": 3.3919,
+      "step": 5259
+    },
+    {
+      "epoch": 0.32357283464566927,
+      "grad_norm": 1.111103892326355,
+      "learning_rate": 0.0001527427944081697,
+      "loss": 2.8675,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3236343503937008,
+      "grad_norm": 0.986619770526886,
+      "learning_rate": 0.0001527263641694761,
+      "loss": 3.0583,
+      "step": 5261
+    },
+    {
+      "epoch": 0.3236958661417323,
+      "grad_norm": 1.1608473062515259,
+      "learning_rate": 0.0001527099319591105,
+      "loss": 2.5602,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3237573818897638,
+      "grad_norm": 1.0787019729614258,
+      "learning_rate": 0.00015269349777768737,
+      "loss": 2.1887,
+      "step": 5263
+    },
+    {
+      "epoch": 0.32381889763779526,
+      "grad_norm": 1.094887137413025,
+      "learning_rate": 0.00015267706162582122,
+      "loss": 2.9657,
+      "step": 5264
+    },
+    {
+      "epoch": 0.3238804133858268,
+      "grad_norm": 1.107035517692566,
+      "learning_rate": 0.0001526606235041267,
+      "loss": 3.4392,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3239419291338583,
+      "grad_norm": 1.0666518211364746,
+      "learning_rate": 0.00015264418341321854,
+      "loss": 3.3727,
+      "step": 5266
+    },
+    {
+      "epoch": 0.32400344488188976,
+      "grad_norm": 1.268830418586731,
+      "learning_rate": 0.00015262774135371143,
+      "loss": 2.961,
+      "step": 5267
+    },
+    {
+      "epoch": 0.32406496062992124,
+      "grad_norm": 1.0983208417892456,
+      "learning_rate": 0.0001526112973262203,
+      "loss": 2.5042,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3241264763779528,
+      "grad_norm": 1.4266208410263062,
+      "learning_rate": 0.00015259485133136,
+      "loss": 2.5491,
+      "step": 5269
+    },
+    {
+      "epoch": 0.32418799212598426,
+      "grad_norm": 1.1242039203643799,
+      "learning_rate": 0.00015257840336974548,
+      "loss": 2.7528,
+      "step": 5270
+    },
+    {
+      "epoch": 0.32424950787401574,
+      "grad_norm": 1.1205854415893555,
+      "learning_rate": 0.00015256195344199192,
+      "loss": 2.3666,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3243110236220472,
+      "grad_norm": 1.1577759981155396,
+      "learning_rate": 0.0001525455015487144,
+      "loss": 2.6263,
+      "step": 5272
+    },
+    {
+      "epoch": 0.32437253937007876,
+      "grad_norm": 1.1390602588653564,
+      "learning_rate": 0.00015252904769052808,
+      "loss": 2.8338,
+      "step": 5273
+    },
+    {
+      "epoch": 0.32443405511811024,
+      "grad_norm": 1.1635687351226807,
+      "learning_rate": 0.00015251259186804833,
+      "loss": 2.8272,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3244955708661417,
+      "grad_norm": 1.2259422540664673,
+      "learning_rate": 0.00015249613408189046,
+      "loss": 3.1451,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3245570866141732,
+      "grad_norm": 1.6428111791610718,
+      "learning_rate": 0.00015247967433266988,
+      "loss": 3.5029,
+      "step": 5276
+    },
+    {
+      "epoch": 0.32461860236220474,
+      "grad_norm": 1.1985257863998413,
+      "learning_rate": 0.00015246321262100208,
+      "loss": 2.2243,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3246801181102362,
+      "grad_norm": 1.2266110181808472,
+      "learning_rate": 0.0001524467489475027,
+      "loss": 2.9105,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3247416338582677,
+      "grad_norm": 1.1297204494476318,
+      "learning_rate": 0.0001524302833127874,
+      "loss": 2.3919,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3248031496062992,
+      "grad_norm": 1.2393146753311157,
+      "learning_rate": 0.00015241381571747185,
+      "loss": 3.1126,
+      "step": 5280
+    },
+    {
+      "epoch": 0.32486466535433073,
+      "grad_norm": 1.214733600616455,
+      "learning_rate": 0.00015239734616217184,
+      "loss": 2.5971,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3249261811023622,
+      "grad_norm": 1.1953259706497192,
+      "learning_rate": 0.00015238087464750327,
+      "loss": 2.4298,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3249876968503937,
+      "grad_norm": 1.143499732017517,
+      "learning_rate": 0.00015236440117408204,
+      "loss": 2.8664,
+      "step": 5283
+    },
+    {
+      "epoch": 0.3250492125984252,
+      "grad_norm": 1.0649546384811401,
+      "learning_rate": 0.00015234792574252425,
+      "loss": 2.6315,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3251107283464567,
+      "grad_norm": 1.2813001871109009,
+      "learning_rate": 0.00015233144835344594,
+      "loss": 2.6265,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3251722440944882,
+      "grad_norm": 1.0568772554397583,
+      "learning_rate": 0.00015231496900746324,
+      "loss": 2.3375,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3252337598425197,
+      "grad_norm": 1.002536416053772,
+      "learning_rate": 0.0001522984877051924,
+      "loss": 3.0435,
+      "step": 5287
+    },
+    {
+      "epoch": 0.32529527559055116,
+      "grad_norm": 1.5388044118881226,
+      "learning_rate": 0.00015228200444724975,
+      "loss": 2.3014,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3253567913385827,
+      "grad_norm": 1.1350655555725098,
+      "learning_rate": 0.00015226551923425168,
+      "loss": 2.8219,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3254183070866142,
+      "grad_norm": 1.1343332529067993,
+      "learning_rate": 0.00015224903206681464,
+      "loss": 3.018,
+      "step": 5290
+    },
+    {
+      "epoch": 0.32547982283464566,
+      "grad_norm": 1.0402863025665283,
+      "learning_rate": 0.00015223254294555516,
+      "loss": 2.8208,
+      "step": 5291
+    },
+    {
+      "epoch": 0.32554133858267714,
+      "grad_norm": 1.0005759000778198,
+      "learning_rate": 0.00015221605187108977,
+      "loss": 2.5483,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3256028543307087,
+      "grad_norm": 1.076174259185791,
+      "learning_rate": 0.00015219955884403525,
+      "loss": 2.4217,
+      "step": 5293
+    },
+    {
+      "epoch": 0.32566437007874016,
+      "grad_norm": 1.1702107191085815,
+      "learning_rate": 0.00015218306386500832,
+      "loss": 2.5451,
+      "step": 5294
+    },
+    {
+      "epoch": 0.32572588582677164,
+      "grad_norm": 1.3694982528686523,
+      "learning_rate": 0.00015216656693462573,
+      "loss": 3.2456,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3257874015748031,
+      "grad_norm": 0.9946928024291992,
+      "learning_rate": 0.00015215006805350447,
+      "loss": 3.0354,
+      "step": 5296
+    },
+    {
+      "epoch": 0.32584891732283466,
+      "grad_norm": 1.6038711071014404,
+      "learning_rate": 0.00015213356722226144,
+      "loss": 2.6163,
+      "step": 5297
+    },
+    {
+      "epoch": 0.32591043307086615,
+      "grad_norm": 1.195353627204895,
+      "learning_rate": 0.0001521170644415137,
+      "loss": 2.7384,
+      "step": 5298
+    },
+    {
+      "epoch": 0.3259719488188976,
+      "grad_norm": 1.132960319519043,
+      "learning_rate": 0.00015210055971187836,
+      "loss": 2.9526,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3260334645669291,
+      "grad_norm": 0.8974447250366211,
+      "learning_rate": 0.00015208405303397262,
+      "loss": 2.8531,
+      "step": 5300
+    },
+    {
+      "epoch": 0.32609498031496065,
+      "grad_norm": 1.1465415954589844,
+      "learning_rate": 0.00015206754440841373,
+      "loss": 2.6303,
+      "step": 5301
+    },
+    {
+      "epoch": 0.32615649606299213,
+      "grad_norm": 1.1441839933395386,
+      "learning_rate": 0.000152051033835819,
+      "loss": 2.9095,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3262180118110236,
+      "grad_norm": 1.042211651802063,
+      "learning_rate": 0.00015203452131680586,
+      "loss": 3.263,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3262795275590551,
+      "grad_norm": 1.0661691427230835,
+      "learning_rate": 0.00015201800685199176,
+      "loss": 3.0052,
+      "step": 5304
+    },
+    {
+      "epoch": 0.32634104330708663,
+      "grad_norm": 1.1134999990463257,
+      "learning_rate": 0.00015200149044199428,
+      "loss": 2.9323,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3264025590551181,
+      "grad_norm": 1.1194888353347778,
+      "learning_rate": 0.00015198497208743103,
+      "loss": 2.774,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3264640748031496,
+      "grad_norm": 1.155953049659729,
+      "learning_rate": 0.0001519684517889197,
+      "loss": 2.696,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3265255905511811,
+      "grad_norm": 1.2100298404693604,
+      "learning_rate": 0.00015195192954707803,
+      "loss": 2.9274,
+      "step": 5308
+    },
+    {
+      "epoch": 0.3265871062992126,
+      "grad_norm": 1.2011116743087769,
+      "learning_rate": 0.00015193540536252392,
+      "loss": 2.7866,
+      "step": 5309
+    },
+    {
+      "epoch": 0.3266486220472441,
+      "grad_norm": 1.2744120359420776,
+      "learning_rate": 0.00015191887923587523,
+      "loss": 2.7286,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3267101377952756,
+      "grad_norm": 1.2265757322311401,
+      "learning_rate": 0.00015190235116774997,
+      "loss": 2.247,
+      "step": 5311
+    },
+    {
+      "epoch": 0.32677165354330706,
+      "grad_norm": 1.0270837545394897,
+      "learning_rate": 0.00015188582115876616,
+      "loss": 3.1353,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3268331692913386,
+      "grad_norm": 1.2417787313461304,
+      "learning_rate": 0.000151869289209542,
+      "loss": 3.0382,
+      "step": 5313
+    },
+    {
+      "epoch": 0.3268946850393701,
+      "grad_norm": 2.620262384414673,
+      "learning_rate": 0.00015185275532069565,
+      "loss": 2.6769,
+      "step": 5314
+    },
+    {
+      "epoch": 0.32695620078740156,
+      "grad_norm": 1.6736172437667847,
+      "learning_rate": 0.0001518362194928454,
+      "loss": 1.9884,
+      "step": 5315
+    },
+    {
+      "epoch": 0.32701771653543305,
+      "grad_norm": 1.5078926086425781,
+      "learning_rate": 0.00015181968172660958,
+      "loss": 2.8932,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3270792322834646,
+      "grad_norm": 1.049121379852295,
+      "learning_rate": 0.00015180314202260662,
+      "loss": 2.9594,
+      "step": 5317
+    },
+    {
+      "epoch": 0.32714074803149606,
+      "grad_norm": 1.2529737949371338,
+      "learning_rate": 0.00015178660038145497,
+      "loss": 2.7871,
+      "step": 5318
+    },
+    {
+      "epoch": 0.32720226377952755,
+      "grad_norm": 1.2858866453170776,
+      "learning_rate": 0.00015177005680377327,
+      "loss": 2.5354,
+      "step": 5319
+    },
+    {
+      "epoch": 0.32726377952755903,
+      "grad_norm": 1.2569963932037354,
+      "learning_rate": 0.00015175351129018014,
+      "loss": 2.6585,
+      "step": 5320
+    },
+    {
+      "epoch": 0.32732529527559057,
+      "grad_norm": 1.2429256439208984,
+      "learning_rate": 0.00015173696384129424,
+      "loss": 2.0304,
+      "step": 5321
+    },
+    {
+      "epoch": 0.32738681102362205,
+      "grad_norm": 1.1201138496398926,
+      "learning_rate": 0.00015172041445773443,
+      "loss": 2.253,
+      "step": 5322
+    },
+    {
+      "epoch": 0.32744832677165353,
+      "grad_norm": 1.3280043601989746,
+      "learning_rate": 0.00015170386314011949,
+      "loss": 2.3909,
+      "step": 5323
+    },
+    {
+      "epoch": 0.327509842519685,
+      "grad_norm": 1.189564824104309,
+      "learning_rate": 0.00015168730988906836,
+      "loss": 2.3045,
+      "step": 5324
+    },
+    {
+      "epoch": 0.32757135826771655,
+      "grad_norm": 0.9837059378623962,
+      "learning_rate": 0.0001516707547052001,
+      "loss": 2.873,
+      "step": 5325
+    },
+    {
+      "epoch": 0.32763287401574803,
+      "grad_norm": 1.3862055540084839,
+      "learning_rate": 0.00015165419758913368,
+      "loss": 2.7837,
+      "step": 5326
+    },
+    {
+      "epoch": 0.3276943897637795,
+      "grad_norm": 1.5095659494400024,
+      "learning_rate": 0.00015163763854148833,
+      "loss": 2.0641,
+      "step": 5327
+    },
+    {
+      "epoch": 0.327755905511811,
+      "grad_norm": 0.9454160332679749,
+      "learning_rate": 0.00015162107756288323,
+      "loss": 3.3649,
+      "step": 5328
+    },
+    {
+      "epoch": 0.32781742125984253,
+      "grad_norm": 1.2886433601379395,
+      "learning_rate": 0.00015160451465393772,
+      "loss": 2.2724,
+      "step": 5329
+    },
+    {
+      "epoch": 0.327878937007874,
+      "grad_norm": 1.4747880697250366,
+      "learning_rate": 0.00015158794981527107,
+      "loss": 2.9315,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3279404527559055,
+      "grad_norm": 1.025666356086731,
+      "learning_rate": 0.0001515713830475028,
+      "loss": 2.6563,
+      "step": 5331
+    },
+    {
+      "epoch": 0.328001968503937,
+      "grad_norm": 1.418283462524414,
+      "learning_rate": 0.00015155481435125232,
+      "loss": 2.6492,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3280634842519685,
+      "grad_norm": 1.3126718997955322,
+      "learning_rate": 0.00015153824372713933,
+      "loss": 3.035,
+      "step": 5333
+    },
+    {
+      "epoch": 0.328125,
+      "grad_norm": 1.104101538658142,
+      "learning_rate": 0.00015152167117578333,
+      "loss": 2.2701,
+      "step": 5334
+    },
+    {
+      "epoch": 0.3281865157480315,
+      "grad_norm": 1.2739133834838867,
+      "learning_rate": 0.00015150509669780416,
+      "loss": 2.7822,
+      "step": 5335
+    },
+    {
+      "epoch": 0.328248031496063,
+      "grad_norm": 1.0883976221084595,
+      "learning_rate": 0.0001514885202938216,
+      "loss": 2.6173,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3283095472440945,
+      "grad_norm": 1.1589553356170654,
+      "learning_rate": 0.0001514719419644554,
+      "loss": 2.4407,
+      "step": 5337
+    },
+    {
+      "epoch": 0.328371062992126,
+      "grad_norm": 1.2473783493041992,
+      "learning_rate": 0.0001514553617103257,
+      "loss": 2.0997,
+      "step": 5338
+    },
+    {
+      "epoch": 0.32843257874015747,
+      "grad_norm": 1.7242170572280884,
+      "learning_rate": 0.00015143877953205235,
+      "loss": 3.1545,
+      "step": 5339
+    },
+    {
+      "epoch": 0.328494094488189,
+      "grad_norm": 0.967196524143219,
+      "learning_rate": 0.00015142219543025544,
+      "loss": 2.8725,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3285556102362205,
+      "grad_norm": 1.0626559257507324,
+      "learning_rate": 0.0001514056094055552,
+      "loss": 2.8847,
+      "step": 5341
+    },
+    {
+      "epoch": 0.32861712598425197,
+      "grad_norm": 1.1302798986434937,
+      "learning_rate": 0.0001513890214585718,
+      "loss": 2.8424,
+      "step": 5342
+    },
+    {
+      "epoch": 0.32867864173228345,
+      "grad_norm": 1.101613998413086,
+      "learning_rate": 0.00015137243158992554,
+      "loss": 2.9312,
+      "step": 5343
+    },
+    {
+      "epoch": 0.328740157480315,
+      "grad_norm": 1.1434834003448486,
+      "learning_rate": 0.00015135583980023683,
+      "loss": 3.0379,
+      "step": 5344
+    },
+    {
+      "epoch": 0.32880167322834647,
+      "grad_norm": 1.1748497486114502,
+      "learning_rate": 0.00015133924609012604,
+      "loss": 2.9966,
+      "step": 5345
+    },
+    {
+      "epoch": 0.32886318897637795,
+      "grad_norm": 1.120949625968933,
+      "learning_rate": 0.00015132265046021375,
+      "loss": 3.262,
+      "step": 5346
+    },
+    {
+      "epoch": 0.32892470472440943,
+      "grad_norm": 1.2460200786590576,
+      "learning_rate": 0.0001513060529111205,
+      "loss": 2.5707,
+      "step": 5347
+    },
+    {
+      "epoch": 0.32898622047244097,
+      "grad_norm": 1.2404742240905762,
+      "learning_rate": 0.00015128945344346697,
+      "loss": 2.5686,
+      "step": 5348
+    },
+    {
+      "epoch": 0.32904773622047245,
+      "grad_norm": 1.1978825330734253,
+      "learning_rate": 0.0001512728520578739,
+      "loss": 1.9318,
+      "step": 5349
+    },
+    {
+      "epoch": 0.32910925196850394,
+      "grad_norm": 1.1507304906845093,
+      "learning_rate": 0.00015125624875496203,
+      "loss": 2.5276,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3291707677165354,
+      "grad_norm": 1.2954161167144775,
+      "learning_rate": 0.0001512396435353523,
+      "loss": 3.2282,
+      "step": 5351
+    },
+    {
+      "epoch": 0.32923228346456695,
+      "grad_norm": 1.1677956581115723,
+      "learning_rate": 0.0001512230363996656,
+      "loss": 2.5901,
+      "step": 5352
+    },
+    {
+      "epoch": 0.32929379921259844,
+      "grad_norm": 1.1810423135757446,
+      "learning_rate": 0.00015120642734852297,
+      "loss": 2.7687,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3293553149606299,
+      "grad_norm": 1.1539108753204346,
+      "learning_rate": 0.0001511898163825455,
+      "loss": 2.6577,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3294168307086614,
+      "grad_norm": 1.264840006828308,
+      "learning_rate": 0.00015117320350235434,
+      "loss": 2.6417,
+      "step": 5355
+    },
+    {
+      "epoch": 0.32947834645669294,
+      "grad_norm": 1.603467345237732,
+      "learning_rate": 0.0001511565887085707,
+      "loss": 3.3705,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3295398622047244,
+      "grad_norm": 1.4864848852157593,
+      "learning_rate": 0.0001511399720018159,
+      "loss": 1.6113,
+      "step": 5357
+    },
+    {
+      "epoch": 0.3296013779527559,
+      "grad_norm": 1.285179853439331,
+      "learning_rate": 0.00015112335338271135,
+      "loss": 3.3873,
+      "step": 5358
+    },
+    {
+      "epoch": 0.3296628937007874,
+      "grad_norm": 1.2156498432159424,
+      "learning_rate": 0.00015110673285187844,
+      "loss": 2.796,
+      "step": 5359
+    },
+    {
+      "epoch": 0.3297244094488189,
+      "grad_norm": 0.9492179155349731,
+      "learning_rate": 0.0001510901104099387,
+      "loss": 2.6512,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3297859251968504,
+      "grad_norm": 1.2655978202819824,
+      "learning_rate": 0.0001510734860575137,
+      "loss": 2.3839,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3298474409448819,
+      "grad_norm": 1.1758986711502075,
+      "learning_rate": 0.00015105685979522511,
+      "loss": 2.9532,
+      "step": 5362
+    },
+    {
+      "epoch": 0.32990895669291337,
+      "grad_norm": 1.0308341979980469,
+      "learning_rate": 0.0001510402316236947,
+      "loss": 2.4136,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3299704724409449,
+      "grad_norm": 1.4898951053619385,
+      "learning_rate": 0.0001510236015435442,
+      "loss": 2.9209,
+      "step": 5364
+    },
+    {
+      "epoch": 0.3300319881889764,
+      "grad_norm": 1.0468533039093018,
+      "learning_rate": 0.00015100696955539555,
+      "loss": 2.7928,
+      "step": 5365
+    },
+    {
+      "epoch": 0.33009350393700787,
+      "grad_norm": 0.9712848663330078,
+      "learning_rate": 0.00015099033565987063,
+      "loss": 2.6925,
+      "step": 5366
+    },
+    {
+      "epoch": 0.33015501968503935,
+      "grad_norm": 1.1470612287521362,
+      "learning_rate": 0.0001509736998575915,
+      "loss": 3.2203,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3302165354330709,
+      "grad_norm": 1.3374446630477905,
+      "learning_rate": 0.0001509570621491802,
+      "loss": 2.4199,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3302780511811024,
+      "grad_norm": 1.1936756372451782,
+      "learning_rate": 0.00015094042253525894,
+      "loss": 2.8351,
+      "step": 5369
+    },
+    {
+      "epoch": 0.33033956692913385,
+      "grad_norm": 1.1791021823883057,
+      "learning_rate": 0.00015092378101644996,
+      "loss": 2.5709,
+      "step": 5370
+    },
+    {
+      "epoch": 0.33040108267716534,
+      "grad_norm": 0.9776358008384705,
+      "learning_rate": 0.0001509071375933755,
+      "loss": 2.6427,
+      "step": 5371
+    },
+    {
+      "epoch": 0.3304625984251969,
+      "grad_norm": 1.0874519348144531,
+      "learning_rate": 0.00015089049226665798,
+      "loss": 3.2263,
+      "step": 5372
+    },
+    {
+      "epoch": 0.33052411417322836,
+      "grad_norm": 1.1934412717819214,
+      "learning_rate": 0.00015087384503691982,
+      "loss": 2.1662,
+      "step": 5373
+    },
+    {
+      "epoch": 0.33058562992125984,
+      "grad_norm": 1.2384800910949707,
+      "learning_rate": 0.0001508571959047835,
+      "loss": 2.2987,
+      "step": 5374
+    },
+    {
+      "epoch": 0.3306471456692913,
+      "grad_norm": 1.2896045446395874,
+      "learning_rate": 0.00015084054487087166,
+      "loss": 3.0988,
+      "step": 5375
+    },
+    {
+      "epoch": 0.33070866141732286,
+      "grad_norm": 1.140952706336975,
+      "learning_rate": 0.00015082389193580697,
+      "loss": 3.331,
+      "step": 5376
+    },
+    {
+      "epoch": 0.33077017716535434,
+      "grad_norm": 1.222532033920288,
+      "learning_rate": 0.00015080723710021208,
+      "loss": 2.5304,
+      "step": 5377
+    },
+    {
+      "epoch": 0.3308316929133858,
+      "grad_norm": 1.2123609781265259,
+      "learning_rate": 0.0001507905803647099,
+      "loss": 2.8902,
+      "step": 5378
+    },
+    {
+      "epoch": 0.3308932086614173,
+      "grad_norm": 1.1471251249313354,
+      "learning_rate": 0.00015077392172992314,
+      "loss": 2.4442,
+      "step": 5379
+    },
+    {
+      "epoch": 0.33095472440944884,
+      "grad_norm": 1.244316577911377,
+      "learning_rate": 0.00015075726119647486,
+      "loss": 2.1408,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3310162401574803,
+      "grad_norm": 1.318281888961792,
+      "learning_rate": 0.00015074059876498804,
+      "loss": 2.6854,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3310777559055118,
+      "grad_norm": 1.2111965417861938,
+      "learning_rate": 0.0001507239344360858,
+      "loss": 2.4844,
+      "step": 5382
+    },
+    {
+      "epoch": 0.3311392716535433,
+      "grad_norm": 1.1305720806121826,
+      "learning_rate": 0.0001507072682103912,
+      "loss": 2.7727,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3312007874015748,
+      "grad_norm": 1.1766597032546997,
+      "learning_rate": 0.00015069060008852757,
+      "loss": 3.1377,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3312623031496063,
+      "grad_norm": 1.3487646579742432,
+      "learning_rate": 0.00015067393007111812,
+      "loss": 2.4207,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3313238188976378,
+      "grad_norm": 1.1688032150268555,
+      "learning_rate": 0.00015065725815878626,
+      "loss": 2.591,
+      "step": 5386
+    },
+    {
+      "epoch": 0.33138533464566927,
+      "grad_norm": 1.0743569135665894,
+      "learning_rate": 0.00015064058435215546,
+      "loss": 2.663,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3314468503937008,
+      "grad_norm": 0.9893577694892883,
+      "learning_rate": 0.0001506239086518491,
+      "loss": 2.7532,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3315083661417323,
+      "grad_norm": 1.2350058555603027,
+      "learning_rate": 0.0001506072310584909,
+      "loss": 2.529,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3315698818897638,
+      "grad_norm": 1.4781651496887207,
+      "learning_rate": 0.00015059055157270445,
+      "loss": 2.7557,
+      "step": 5390
+    },
+    {
+      "epoch": 0.33163139763779526,
+      "grad_norm": 2.644047975540161,
+      "learning_rate": 0.0001505738701951135,
+      "loss": 2.8685,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3316929133858268,
+      "grad_norm": 1.2101587057113647,
+      "learning_rate": 0.00015055718692634176,
+      "loss": 2.1153,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3317544291338583,
+      "grad_norm": 1.6791183948516846,
+      "learning_rate": 0.00015054050176701314,
+      "loss": 2.2703,
+      "step": 5393
+    },
+    {
+      "epoch": 0.33181594488188976,
+      "grad_norm": 1.486531376838684,
+      "learning_rate": 0.00015052381471775163,
+      "loss": 2.4368,
+      "step": 5394
+    },
+    {
+      "epoch": 0.33187746062992124,
+      "grad_norm": 1.2635551691055298,
+      "learning_rate": 0.00015050712577918115,
+      "loss": 2.335,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3319389763779528,
+      "grad_norm": 1.2503938674926758,
+      "learning_rate": 0.00015049043495192577,
+      "loss": 2.2956,
+      "step": 5396
+    },
+    {
+      "epoch": 0.33200049212598426,
+      "grad_norm": 1.2946521043777466,
+      "learning_rate": 0.0001504737422366097,
+      "loss": 2.8094,
+      "step": 5397
+    },
+    {
+      "epoch": 0.33206200787401574,
+      "grad_norm": 1.2595237493515015,
+      "learning_rate": 0.00015045704763385712,
+      "loss": 2.9977,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3321235236220472,
+      "grad_norm": 1.2014905214309692,
+      "learning_rate": 0.0001504403511442923,
+      "loss": 2.9049,
+      "step": 5399
+    },
+    {
+      "epoch": 0.33218503937007876,
+      "grad_norm": 1.3412137031555176,
+      "learning_rate": 0.00015042365276853964,
+      "loss": 2.6282,
+      "step": 5400
+    },
+    {
+      "epoch": 0.33224655511811024,
+      "grad_norm": 1.5341516733169556,
+      "learning_rate": 0.0001504069525072235,
+      "loss": 2.7086,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3323080708661417,
+      "grad_norm": 1.0321565866470337,
+      "learning_rate": 0.00015039025036096842,
+      "loss": 3.2382,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3323695866141732,
+      "grad_norm": 1.0713040828704834,
+      "learning_rate": 0.00015037354633039898,
+      "loss": 3.0793,
+      "step": 5403
+    },
+    {
+      "epoch": 0.33243110236220474,
+      "grad_norm": 1.358391284942627,
+      "learning_rate": 0.00015035684041613976,
+      "loss": 2.5451,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3324926181102362,
+      "grad_norm": 1.1474759578704834,
+      "learning_rate": 0.00015034013261881553,
+      "loss": 2.7292,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3325541338582677,
+      "grad_norm": 1.2198325395584106,
+      "learning_rate": 0.00015032342293905105,
+      "loss": 2.67,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3326156496062992,
+      "grad_norm": 1.0146608352661133,
+      "learning_rate": 0.00015030671137747118,
+      "loss": 2.5184,
+      "step": 5407
+    },
+    {
+      "epoch": 0.33267716535433073,
+      "grad_norm": 1.3229371309280396,
+      "learning_rate": 0.0001502899979347008,
+      "loss": 2.8006,
+      "step": 5408
+    },
+    {
+      "epoch": 0.3327386811023622,
+      "grad_norm": 1.062180519104004,
+      "learning_rate": 0.0001502732826113649,
+      "loss": 2.9261,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3328001968503937,
+      "grad_norm": 1.1770308017730713,
+      "learning_rate": 0.00015025656540808857,
+      "loss": 2.8618,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3328617125984252,
+      "grad_norm": 1.2534539699554443,
+      "learning_rate": 0.00015023984632549694,
+      "loss": 2.4664,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3329232283464567,
+      "grad_norm": 1.0738805532455444,
+      "learning_rate": 0.00015022312536421524,
+      "loss": 2.0908,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3329847440944882,
+      "grad_norm": 1.2649497985839844,
+      "learning_rate": 0.00015020640252486867,
+      "loss": 2.3549,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3330462598425197,
+      "grad_norm": 1.1274800300598145,
+      "learning_rate": 0.0001501896778080826,
+      "loss": 2.9665,
+      "step": 5414
+    },
+    {
+      "epoch": 0.33310777559055116,
+      "grad_norm": 1.3844730854034424,
+      "learning_rate": 0.00015017295121448244,
+      "loss": 2.7858,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3331692913385827,
+      "grad_norm": 1.2255089282989502,
+      "learning_rate": 0.00015015622274469373,
+      "loss": 2.4694,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3332308070866142,
+      "grad_norm": 0.9738345742225647,
+      "learning_rate": 0.00015013949239934187,
+      "loss": 3.1317,
+      "step": 5417
+    },
+    {
+      "epoch": 0.33329232283464566,
+      "grad_norm": 1.097307562828064,
+      "learning_rate": 0.00015012276017905264,
+      "loss": 2.6519,
+      "step": 5418
+    },
+    {
+      "epoch": 0.33335383858267714,
+      "grad_norm": 1.1636780500411987,
+      "learning_rate": 0.0001501060260844517,
+      "loss": 3.2062,
+      "step": 5419
+    },
+    {
+      "epoch": 0.3334153543307087,
+      "grad_norm": 1.0731953382492065,
+      "learning_rate": 0.00015008929011616474,
+      "loss": 1.8386,
+      "step": 5420
+    },
+    {
+      "epoch": 0.33347687007874016,
+      "grad_norm": 2.107642889022827,
+      "learning_rate": 0.00015007255227481766,
+      "loss": 2.399,
+      "step": 5421
+    },
+    {
+      "epoch": 0.33353838582677164,
+      "grad_norm": 1.0547138452529907,
+      "learning_rate": 0.0001500558125610363,
+      "loss": 2.4204,
+      "step": 5422
+    },
+    {
+      "epoch": 0.3335999015748031,
+      "grad_norm": 0.9841210246086121,
+      "learning_rate": 0.00015003907097544666,
+      "loss": 2.0344,
+      "step": 5423
+    },
+    {
+      "epoch": 0.33366141732283466,
+      "grad_norm": 1.1117353439331055,
+      "learning_rate": 0.00015002232751867487,
+      "loss": 2.6479,
+      "step": 5424
+    },
+    {
+      "epoch": 0.33372293307086615,
+      "grad_norm": 1.3394081592559814,
+      "learning_rate": 0.0001500055821913469,
+      "loss": 2.73,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3337844488188976,
+      "grad_norm": 1.0384933948516846,
+      "learning_rate": 0.00014998883499408902,
+      "loss": 2.8317,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3338459645669291,
+      "grad_norm": 1.1462814807891846,
+      "learning_rate": 0.00014997208592752742,
+      "loss": 2.6224,
+      "step": 5427
+    },
+    {
+      "epoch": 0.33390748031496065,
+      "grad_norm": 1.1915855407714844,
+      "learning_rate": 0.0001499553349922885,
+      "loss": 3.223,
+      "step": 5428
+    },
+    {
+      "epoch": 0.33396899606299213,
+      "grad_norm": 1.0745042562484741,
+      "learning_rate": 0.0001499385821889986,
+      "loss": 3.0371,
+      "step": 5429
+    },
+    {
+      "epoch": 0.3340305118110236,
+      "grad_norm": 1.1002651453018188,
+      "learning_rate": 0.0001499218275182842,
+      "loss": 2.6787,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3340920275590551,
+      "grad_norm": 1.1791858673095703,
+      "learning_rate": 0.0001499050709807718,
+      "loss": 2.522,
+      "step": 5431
+    },
+    {
+      "epoch": 0.33415354330708663,
+      "grad_norm": 1.3051724433898926,
+      "learning_rate": 0.00014988831257708803,
+      "loss": 3.4843,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3342150590551181,
+      "grad_norm": 1.252703309059143,
+      "learning_rate": 0.00014987155230785956,
+      "loss": 3.2205,
+      "step": 5433
+    },
+    {
+      "epoch": 0.3342765748031496,
+      "grad_norm": 1.1797760725021362,
+      "learning_rate": 0.00014985479017371313,
+      "loss": 2.3989,
+      "step": 5434
+    },
+    {
+      "epoch": 0.3343380905511811,
+      "grad_norm": 1.2444872856140137,
+      "learning_rate": 0.00014983802617527555,
+      "loss": 2.8597,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3343996062992126,
+      "grad_norm": 1.2360728979110718,
+      "learning_rate": 0.00014982126031317367,
+      "loss": 2.9522,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3344611220472441,
+      "grad_norm": 1.08803391456604,
+      "learning_rate": 0.00014980449258803448,
+      "loss": 3.0855,
+      "step": 5437
+    },
+    {
+      "epoch": 0.3345226377952756,
+      "grad_norm": 1.1226625442504883,
+      "learning_rate": 0.00014978772300048497,
+      "loss": 1.9726,
+      "step": 5438
+    },
+    {
+      "epoch": 0.33458415354330706,
+      "grad_norm": 1.1167871952056885,
+      "learning_rate": 0.00014977095155115227,
+      "loss": 2.8439,
+      "step": 5439
+    },
+    {
+      "epoch": 0.3346456692913386,
+      "grad_norm": 1.2717304229736328,
+      "learning_rate": 0.00014975417824066348,
+      "loss": 2.4556,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3347071850393701,
+      "grad_norm": 1.0663129091262817,
+      "learning_rate": 0.0001497374030696459,
+      "loss": 1.861,
+      "step": 5441
+    },
+    {
+      "epoch": 0.33476870078740156,
+      "grad_norm": 1.0959603786468506,
+      "learning_rate": 0.00014972062603872677,
+      "loss": 2.8276,
+      "step": 5442
+    },
+    {
+      "epoch": 0.33483021653543305,
+      "grad_norm": 1.2632591724395752,
+      "learning_rate": 0.00014970384714853348,
+      "loss": 2.711,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3348917322834646,
+      "grad_norm": 1.2049498558044434,
+      "learning_rate": 0.00014968706639969346,
+      "loss": 3.1297,
+      "step": 5444
+    },
+    {
+      "epoch": 0.33495324803149606,
+      "grad_norm": 1.1399272680282593,
+      "learning_rate": 0.0001496702837928342,
+      "loss": 2.5219,
+      "step": 5445
+    },
+    {
+      "epoch": 0.33501476377952755,
+      "grad_norm": 0.9767244458198547,
+      "learning_rate": 0.00014965349932858334,
+      "loss": 2.5689,
+      "step": 5446
+    },
+    {
+      "epoch": 0.33507627952755903,
+      "grad_norm": 1.546247124671936,
+      "learning_rate": 0.00014963671300756843,
+      "loss": 2.3847,
+      "step": 5447
+    },
+    {
+      "epoch": 0.33513779527559057,
+      "grad_norm": 2.573253631591797,
+      "learning_rate": 0.00014961992483041728,
+      "loss": 0.7768,
+      "step": 5448
+    },
+    {
+      "epoch": 0.33519931102362205,
+      "grad_norm": 1.2903836965560913,
+      "learning_rate": 0.0001496031347977576,
+      "loss": 2.8683,
+      "step": 5449
+    },
+    {
+      "epoch": 0.33526082677165353,
+      "grad_norm": 1.1924371719360352,
+      "learning_rate": 0.00014958634291021726,
+      "loss": 3.1861,
+      "step": 5450
+    },
+    {
+      "epoch": 0.335322342519685,
+      "grad_norm": 1.1307212114334106,
+      "learning_rate": 0.00014956954916842423,
+      "loss": 2.3321,
+      "step": 5451
+    },
+    {
+      "epoch": 0.33538385826771655,
+      "grad_norm": 1.306231141090393,
+      "learning_rate": 0.00014955275357300644,
+      "loss": 3.0271,
+      "step": 5452
+    },
+    {
+      "epoch": 0.33544537401574803,
+      "grad_norm": 1.2216944694519043,
+      "learning_rate": 0.000149535956124592,
+      "loss": 2.0639,
+      "step": 5453
+    },
+    {
+      "epoch": 0.3355068897637795,
+      "grad_norm": 1.0456092357635498,
+      "learning_rate": 0.00014951915682380901,
+      "loss": 2.5767,
+      "step": 5454
+    },
+    {
+      "epoch": 0.335568405511811,
+      "grad_norm": 1.2703717947006226,
+      "learning_rate": 0.0001495023556712857,
+      "loss": 2.0837,
+      "step": 5455
+    },
+    {
+      "epoch": 0.33562992125984253,
+      "grad_norm": 1.1786175966262817,
+      "learning_rate": 0.0001494855526676503,
+      "loss": 2.4179,
+      "step": 5456
+    },
+    {
+      "epoch": 0.335691437007874,
+      "grad_norm": 1.2502678632736206,
+      "learning_rate": 0.00014946874781353116,
+      "loss": 2.5266,
+      "step": 5457
+    },
+    {
+      "epoch": 0.3357529527559055,
+      "grad_norm": 1.1646987199783325,
+      "learning_rate": 0.00014945194110955673,
+      "loss": 2.6906,
+      "step": 5458
+    },
+    {
+      "epoch": 0.335814468503937,
+      "grad_norm": 1.0794591903686523,
+      "learning_rate": 0.00014943513255635544,
+      "loss": 2.7915,
+      "step": 5459
+    },
+    {
+      "epoch": 0.3358759842519685,
+      "grad_norm": 1.2540720701217651,
+      "learning_rate": 0.00014941832215455584,
+      "loss": 2.9552,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3359375,
+      "grad_norm": 1.4533509016036987,
+      "learning_rate": 0.0001494015099047866,
+      "loss": 2.4856,
+      "step": 5461
+    },
+    {
+      "epoch": 0.3359990157480315,
+      "grad_norm": 1.126369595527649,
+      "learning_rate": 0.00014938469580767628,
+      "loss": 2.8998,
+      "step": 5462
+    },
+    {
+      "epoch": 0.336060531496063,
+      "grad_norm": 1.1490825414657593,
+      "learning_rate": 0.00014936787986385375,
+      "loss": 2.595,
+      "step": 5463
+    },
+    {
+      "epoch": 0.3361220472440945,
+      "grad_norm": 1.1185004711151123,
+      "learning_rate": 0.00014935106207394786,
+      "loss": 1.1721,
+      "step": 5464
+    },
+    {
+      "epoch": 0.336183562992126,
+      "grad_norm": 1.1813080310821533,
+      "learning_rate": 0.00014933424243858734,
+      "loss": 3.7544,
+      "step": 5465
+    },
+    {
+      "epoch": 0.33624507874015747,
+      "grad_norm": 1.0930202007293701,
+      "learning_rate": 0.00014931742095840132,
+      "loss": 2.8146,
+      "step": 5466
+    },
+    {
+      "epoch": 0.336306594488189,
+      "grad_norm": 1.1854456663131714,
+      "learning_rate": 0.0001493005976340187,
+      "loss": 2.7387,
+      "step": 5467
+    },
+    {
+      "epoch": 0.3363681102362205,
+      "grad_norm": 1.2658679485321045,
+      "learning_rate": 0.00014928377246606865,
+      "loss": 2.4942,
+      "step": 5468
+    },
+    {
+      "epoch": 0.33642962598425197,
+      "grad_norm": 1.3771445751190186,
+      "learning_rate": 0.00014926694545518035,
+      "loss": 2.4058,
+      "step": 5469
+    },
+    {
+      "epoch": 0.33649114173228345,
+      "grad_norm": 1.8860758543014526,
+      "learning_rate": 0.00014925011660198303,
+      "loss": 3.3379,
+      "step": 5470
+    },
+    {
+      "epoch": 0.336552657480315,
+      "grad_norm": 1.2572227716445923,
+      "learning_rate": 0.00014923328590710593,
+      "loss": 2.8537,
+      "step": 5471
+    },
+    {
+      "epoch": 0.33661417322834647,
+      "grad_norm": 1.1156630516052246,
+      "learning_rate": 0.0001492164533711785,
+      "loss": 1.804,
+      "step": 5472
+    },
+    {
+      "epoch": 0.33667568897637795,
+      "grad_norm": 1.3675307035446167,
+      "learning_rate": 0.00014919961899483012,
+      "loss": 3.261,
+      "step": 5473
+    },
+    {
+      "epoch": 0.33673720472440943,
+      "grad_norm": 1.2095377445220947,
+      "learning_rate": 0.00014918278277869035,
+      "loss": 2.9175,
+      "step": 5474
+    },
+    {
+      "epoch": 0.33679872047244097,
+      "grad_norm": 1.2480788230895996,
+      "learning_rate": 0.00014916594472338875,
+      "loss": 2.056,
+      "step": 5475
+    },
+    {
+      "epoch": 0.33686023622047245,
+      "grad_norm": 1.051906704902649,
+      "learning_rate": 0.00014914910482955498,
+      "loss": 2.8644,
+      "step": 5476
+    },
+    {
+      "epoch": 0.33692175196850394,
+      "grad_norm": 1.2276140451431274,
+      "learning_rate": 0.00014913226309781874,
+      "loss": 3.1211,
+      "step": 5477
+    },
+    {
+      "epoch": 0.3369832677165354,
+      "grad_norm": 1.174818515777588,
+      "learning_rate": 0.00014911541952880984,
+      "loss": 3.1067,
+      "step": 5478
+    },
+    {
+      "epoch": 0.33704478346456695,
+      "grad_norm": 1.2540795803070068,
+      "learning_rate": 0.00014909857412315813,
+      "loss": 2.7943,
+      "step": 5479
+    },
+    {
+      "epoch": 0.33710629921259844,
+      "grad_norm": 1.1826705932617188,
+      "learning_rate": 0.00014908172688149352,
+      "loss": 2.8975,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3371678149606299,
+      "grad_norm": 1.152567982673645,
+      "learning_rate": 0.00014906487780444601,
+      "loss": 1.8611,
+      "step": 5481
+    },
+    {
+      "epoch": 0.3372293307086614,
+      "grad_norm": 0.941408634185791,
+      "learning_rate": 0.00014904802689264568,
+      "loss": 2.4954,
+      "step": 5482
+    },
+    {
+      "epoch": 0.33729084645669294,
+      "grad_norm": 0.9211896061897278,
+      "learning_rate": 0.00014903117414672264,
+      "loss": 3.2115,
+      "step": 5483
+    },
+    {
+      "epoch": 0.3373523622047244,
+      "grad_norm": 1.030195713043213,
+      "learning_rate": 0.00014901431956730704,
+      "loss": 2.9298,
+      "step": 5484
+    },
+    {
+      "epoch": 0.3374138779527559,
+      "grad_norm": 1.101212739944458,
+      "learning_rate": 0.00014899746315502927,
+      "loss": 2.63,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3374753937007874,
+      "grad_norm": 1.103846788406372,
+      "learning_rate": 0.00014898060491051957,
+      "loss": 2.9967,
+      "step": 5486
+    },
+    {
+      "epoch": 0.3375369094488189,
+      "grad_norm": 1.136149525642395,
+      "learning_rate": 0.00014896374483440835,
+      "loss": 3.3563,
+      "step": 5487
+    },
+    {
+      "epoch": 0.3375984251968504,
+      "grad_norm": 1.2915152311325073,
+      "learning_rate": 0.00014894688292732612,
+      "loss": 2.7858,
+      "step": 5488
+    },
+    {
+      "epoch": 0.3376599409448819,
+      "grad_norm": 1.3610811233520508,
+      "learning_rate": 0.0001489300191899034,
+      "loss": 2.1608,
+      "step": 5489
+    },
+    {
+      "epoch": 0.33772145669291337,
+      "grad_norm": 1.159896969795227,
+      "learning_rate": 0.0001489131536227708,
+      "loss": 3.0675,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3377829724409449,
+      "grad_norm": 1.12271249294281,
+      "learning_rate": 0.00014889628622655903,
+      "loss": 2.5778,
+      "step": 5491
+    },
+    {
+      "epoch": 0.3378444881889764,
+      "grad_norm": 1.2394531965255737,
+      "learning_rate": 0.0001488794170018988,
+      "loss": 3.2474,
+      "step": 5492
+    },
+    {
+      "epoch": 0.33790600393700787,
+      "grad_norm": 1.2090963125228882,
+      "learning_rate": 0.00014886254594942088,
+      "loss": 2.7934,
+      "step": 5493
+    },
+    {
+      "epoch": 0.33796751968503935,
+      "grad_norm": 1.079709529876709,
+      "learning_rate": 0.00014884567306975628,
+      "loss": 2.5894,
+      "step": 5494
+    },
+    {
+      "epoch": 0.3380290354330709,
+      "grad_norm": 1.1220860481262207,
+      "learning_rate": 0.00014882879836353584,
+      "loss": 2.6748,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3380905511811024,
+      "grad_norm": 1.012208342552185,
+      "learning_rate": 0.00014881192183139062,
+      "loss": 2.7499,
+      "step": 5496
+    },
+    {
+      "epoch": 0.33815206692913385,
+      "grad_norm": 1.1212389469146729,
+      "learning_rate": 0.0001487950434739517,
+      "loss": 3.0724,
+      "step": 5497
+    },
+    {
+      "epoch": 0.33821358267716534,
+      "grad_norm": 3.260244131088257,
+      "learning_rate": 0.00014877816329185025,
+      "loss": 2.4724,
+      "step": 5498
+    },
+    {
+      "epoch": 0.3382750984251969,
+      "grad_norm": 1.069076657295227,
+      "learning_rate": 0.0001487612812857175,
+      "loss": 2.4125,
+      "step": 5499
+    },
+    {
+      "epoch": 0.33833661417322836,
+      "grad_norm": 1.3787832260131836,
+      "learning_rate": 0.00014874439745618472,
+      "loss": 2.0604,
+      "step": 5500
+    },
+    {
+      "epoch": 0.33839812992125984,
+      "grad_norm": 1.2042988538742065,
+      "learning_rate": 0.0001487275118038833,
+      "loss": 2.7352,
+      "step": 5501
+    },
+    {
+      "epoch": 0.3384596456692913,
+      "grad_norm": 1.0804394483566284,
+      "learning_rate": 0.00014871062432944462,
+      "loss": 2.3886,
+      "step": 5502
+    },
+    {
+      "epoch": 0.33852116141732286,
+      "grad_norm": 1.1313567161560059,
+      "learning_rate": 0.00014869373503350022,
+      "loss": 2.7795,
+      "step": 5503
+    },
+    {
+      "epoch": 0.33858267716535434,
+      "grad_norm": 1.0838290452957153,
+      "learning_rate": 0.00014867684391668164,
+      "loss": 2.4963,
+      "step": 5504
+    },
+    {
+      "epoch": 0.3386441929133858,
+      "grad_norm": 1.112964391708374,
+      "learning_rate": 0.00014865995097962059,
+      "loss": 2.4888,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3387057086614173,
+      "grad_norm": 1.071990966796875,
+      "learning_rate": 0.00014864305622294864,
+      "loss": 2.701,
+      "step": 5506
+    },
+    {
+      "epoch": 0.33876722440944884,
+      "grad_norm": 1.4821597337722778,
+      "learning_rate": 0.00014862615964729764,
+      "loss": 2.3623,
+      "step": 5507
+    },
+    {
+      "epoch": 0.3388287401574803,
+      "grad_norm": 1.1834548711776733,
+      "learning_rate": 0.00014860926125329944,
+      "loss": 2.2613,
+      "step": 5508
+    },
+    {
+      "epoch": 0.3388902559055118,
+      "grad_norm": 1.1466786861419678,
+      "learning_rate": 0.00014859236104158589,
+      "loss": 3.3763,
+      "step": 5509
+    },
+    {
+      "epoch": 0.3389517716535433,
+      "grad_norm": 1.1805636882781982,
+      "learning_rate": 0.000148575459012789,
+      "loss": 3.2712,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3390132874015748,
+      "grad_norm": 0.9815438389778137,
+      "learning_rate": 0.00014855855516754084,
+      "loss": 2.6282,
+      "step": 5511
+    },
+    {
+      "epoch": 0.3390748031496063,
+      "grad_norm": 1.2459264993667603,
+      "learning_rate": 0.00014854164950647345,
+      "loss": 2.34,
+      "step": 5512
+    },
+    {
+      "epoch": 0.3391363188976378,
+      "grad_norm": 1.2704801559448242,
+      "learning_rate": 0.00014852474203021907,
+      "loss": 2.9687,
+      "step": 5513
+    },
+    {
+      "epoch": 0.33919783464566927,
+      "grad_norm": 1.0497708320617676,
+      "learning_rate": 0.0001485078327394099,
+      "loss": 2.09,
+      "step": 5514
+    },
+    {
+      "epoch": 0.3392593503937008,
+      "grad_norm": 1.2186943292617798,
+      "learning_rate": 0.00014849092163467828,
+      "loss": 2.653,
+      "step": 5515
+    },
+    {
+      "epoch": 0.3393208661417323,
+      "grad_norm": 1.4061936140060425,
+      "learning_rate": 0.00014847400871665658,
+      "loss": 2.9079,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3393823818897638,
+      "grad_norm": 0.9999016523361206,
+      "learning_rate": 0.00014845709398597726,
+      "loss": 2.7424,
+      "step": 5517
+    },
+    {
+      "epoch": 0.33944389763779526,
+      "grad_norm": 1.0271272659301758,
+      "learning_rate": 0.00014844017744327284,
+      "loss": 2.7146,
+      "step": 5518
+    },
+    {
+      "epoch": 0.3395054133858268,
+      "grad_norm": 1.1185117959976196,
+      "learning_rate": 0.00014842325908917588,
+      "loss": 2.8615,
+      "step": 5519
+    },
+    {
+      "epoch": 0.3395669291338583,
+      "grad_norm": 1.3586937189102173,
+      "learning_rate": 0.00014840633892431905,
+      "loss": 2.1783,
+      "step": 5520
+    },
+    {
+      "epoch": 0.33962844488188976,
+      "grad_norm": 1.2131474018096924,
+      "learning_rate": 0.00014838941694933505,
+      "loss": 2.4064,
+      "step": 5521
+    },
+    {
+      "epoch": 0.33968996062992124,
+      "grad_norm": 1.188366174697876,
+      "learning_rate": 0.00014837249316485668,
+      "loss": 2.8206,
+      "step": 5522
+    },
+    {
+      "epoch": 0.3397514763779528,
+      "grad_norm": 1.0323402881622314,
+      "learning_rate": 0.00014835556757151683,
+      "loss": 3.1731,
+      "step": 5523
+    },
+    {
+      "epoch": 0.33981299212598426,
+      "grad_norm": 1.2320774793624878,
+      "learning_rate": 0.00014833864016994836,
+      "loss": 2.9841,
+      "step": 5524
+    },
+    {
+      "epoch": 0.33987450787401574,
+      "grad_norm": 1.0018118619918823,
+      "learning_rate": 0.0001483217109607843,
+      "loss": 2.8149,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3399360236220472,
+      "grad_norm": 1.2012104988098145,
+      "learning_rate": 0.0001483047799446577,
+      "loss": 2.7332,
+      "step": 5526
+    },
+    {
+      "epoch": 0.33999753937007876,
+      "grad_norm": 1.1041432619094849,
+      "learning_rate": 0.0001482878471222017,
+      "loss": 2.7707,
+      "step": 5527
+    },
+    {
+      "epoch": 0.34005905511811024,
+      "grad_norm": 1.137762427330017,
+      "learning_rate": 0.00014827091249404947,
+      "loss": 3.0038,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3401205708661417,
+      "grad_norm": 1.0103626251220703,
+      "learning_rate": 0.00014825397606083425,
+      "loss": 2.9524,
+      "step": 5529
+    },
+    {
+      "epoch": 0.3401820866141732,
+      "grad_norm": 1.3417940139770508,
+      "learning_rate": 0.00014823703782318942,
+      "loss": 2.5937,
+      "step": 5530
+    },
+    {
+      "epoch": 0.34024360236220474,
+      "grad_norm": 1.1307023763656616,
+      "learning_rate": 0.00014822009778174837,
+      "loss": 2.5731,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3403051181102362,
+      "grad_norm": 1.1418399810791016,
+      "learning_rate": 0.00014820315593714453,
+      "loss": 2.3494,
+      "step": 5532
+    },
+    {
+      "epoch": 0.3403666338582677,
+      "grad_norm": 1.1599332094192505,
+      "learning_rate": 0.00014818621229001146,
+      "loss": 2.4234,
+      "step": 5533
+    },
+    {
+      "epoch": 0.3404281496062992,
+      "grad_norm": 1.395423412322998,
+      "learning_rate": 0.00014816926684098274,
+      "loss": 3.1828,
+      "step": 5534
+    },
+    {
+      "epoch": 0.34048966535433073,
+      "grad_norm": 0.937284529209137,
+      "learning_rate": 0.000148152319590692,
+      "loss": 2.6694,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3405511811023622,
+      "grad_norm": 1.1536164283752441,
+      "learning_rate": 0.00014813537053977305,
+      "loss": 1.8866,
+      "step": 5536
+    },
+    {
+      "epoch": 0.3406126968503937,
+      "grad_norm": 1.213709831237793,
+      "learning_rate": 0.00014811841968885965,
+      "loss": 2.9355,
+      "step": 5537
+    },
+    {
+      "epoch": 0.3406742125984252,
+      "grad_norm": 1.1176024675369263,
+      "learning_rate": 0.00014810146703858566,
+      "loss": 2.7587,
+      "step": 5538
+    },
+    {
+      "epoch": 0.3407357283464567,
+      "grad_norm": 1.1363438367843628,
+      "learning_rate": 0.00014808451258958502,
+      "loss": 2.5242,
+      "step": 5539
+    },
+    {
+      "epoch": 0.3407972440944882,
+      "grad_norm": 1.3354774713516235,
+      "learning_rate": 0.00014806755634249173,
+      "loss": 2.6343,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3408587598425197,
+      "grad_norm": 1.0660209655761719,
+      "learning_rate": 0.0001480505982979399,
+      "loss": 3.0676,
+      "step": 5541
+    },
+    {
+      "epoch": 0.34092027559055116,
+      "grad_norm": 1.2456384897232056,
+      "learning_rate": 0.00014803363845656363,
+      "loss": 2.928,
+      "step": 5542
+    },
+    {
+      "epoch": 0.3409817913385827,
+      "grad_norm": 1.1733312606811523,
+      "learning_rate": 0.0001480166768189971,
+      "loss": 2.4196,
+      "step": 5543
+    },
+    {
+      "epoch": 0.3410433070866142,
+      "grad_norm": 1.0305663347244263,
+      "learning_rate": 0.0001479997133858746,
+      "loss": 2.6478,
+      "step": 5544
+    },
+    {
+      "epoch": 0.34110482283464566,
+      "grad_norm": 1.412433385848999,
+      "learning_rate": 0.00014798274815783045,
+      "loss": 2.7544,
+      "step": 5545
+    },
+    {
+      "epoch": 0.34116633858267714,
+      "grad_norm": 1.3832224607467651,
+      "learning_rate": 0.00014796578113549912,
+      "loss": 2.2153,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3412278543307087,
+      "grad_norm": 1.2565594911575317,
+      "learning_rate": 0.00014794881231951503,
+      "loss": 2.0315,
+      "step": 5547
+    },
+    {
+      "epoch": 0.34128937007874016,
+      "grad_norm": 1.1358835697174072,
+      "learning_rate": 0.00014793184171051272,
+      "loss": 2.7416,
+      "step": 5548
+    },
+    {
+      "epoch": 0.34135088582677164,
+      "grad_norm": 1.054184079170227,
+      "learning_rate": 0.00014791486930912685,
+      "loss": 2.369,
+      "step": 5549
+    },
+    {
+      "epoch": 0.3414124015748031,
+      "grad_norm": 1.0896164178848267,
+      "learning_rate": 0.00014789789511599199,
+      "loss": 2.9015,
+      "step": 5550
+    },
+    {
+      "epoch": 0.34147391732283466,
+      "grad_norm": 1.1173498630523682,
+      "learning_rate": 0.00014788091913174297,
+      "loss": 3.0277,
+      "step": 5551
+    },
+    {
+      "epoch": 0.34153543307086615,
+      "grad_norm": 1.9504423141479492,
+      "learning_rate": 0.00014786394135701456,
+      "loss": 2.3657,
+      "step": 5552
+    },
+    {
+      "epoch": 0.3415969488188976,
+      "grad_norm": 2.1835663318634033,
+      "learning_rate": 0.00014784696179244165,
+      "loss": 2.2744,
+      "step": 5553
+    },
+    {
+      "epoch": 0.3416584645669291,
+      "grad_norm": 1.0914654731750488,
+      "learning_rate": 0.00014782998043865912,
+      "loss": 2.4095,
+      "step": 5554
+    },
+    {
+      "epoch": 0.34171998031496065,
+      "grad_norm": 1.3150848150253296,
+      "learning_rate": 0.0001478129972963021,
+      "loss": 3.0445,
+      "step": 5555
+    },
+    {
+      "epoch": 0.34178149606299213,
+      "grad_norm": 1.0520861148834229,
+      "learning_rate": 0.00014779601236600555,
+      "loss": 2.585,
+      "step": 5556
+    },
+    {
+      "epoch": 0.3418430118110236,
+      "grad_norm": 1.060596227645874,
+      "learning_rate": 0.00014777902564840463,
+      "loss": 3.008,
+      "step": 5557
+    },
+    {
+      "epoch": 0.3419045275590551,
+      "grad_norm": 1.225386619567871,
+      "learning_rate": 0.00014776203714413462,
+      "loss": 2.4996,
+      "step": 5558
+    },
+    {
+      "epoch": 0.34196604330708663,
+      "grad_norm": 1.4280105829238892,
+      "learning_rate": 0.00014774504685383075,
+      "loss": 3.6258,
+      "step": 5559
+    },
+    {
+      "epoch": 0.3420275590551181,
+      "grad_norm": 1.0635112524032593,
+      "learning_rate": 0.00014772805477812834,
+      "loss": 2.6691,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3420890748031496,
+      "grad_norm": 1.270967721939087,
+      "learning_rate": 0.00014771106091766284,
+      "loss": 2.632,
+      "step": 5561
+    },
+    {
+      "epoch": 0.3421505905511811,
+      "grad_norm": 1.1207069158554077,
+      "learning_rate": 0.0001476940652730697,
+      "loss": 2.7196,
+      "step": 5562
+    },
+    {
+      "epoch": 0.3422121062992126,
+      "grad_norm": 1.2382382154464722,
+      "learning_rate": 0.00014767706784498446,
+      "loss": 2.6338,
+      "step": 5563
+    },
+    {
+      "epoch": 0.3422736220472441,
+      "grad_norm": 1.3431943655014038,
+      "learning_rate": 0.00014766006863404274,
+      "loss": 2.6618,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3423351377952756,
+      "grad_norm": 1.1696223020553589,
+      "learning_rate": 0.0001476430676408802,
+      "loss": 2.846,
+      "step": 5565
+    },
+    {
+      "epoch": 0.34239665354330706,
+      "grad_norm": 1.3321685791015625,
+      "learning_rate": 0.00014762606486613263,
+      "loss": 2.8591,
+      "step": 5566
+    },
+    {
+      "epoch": 0.3424581692913386,
+      "grad_norm": 1.4199285507202148,
+      "learning_rate": 0.00014760906031043578,
+      "loss": 2.7432,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3425196850393701,
+      "grad_norm": 1.1737245321273804,
+      "learning_rate": 0.00014759205397442556,
+      "loss": 2.8098,
+      "step": 5568
+    },
+    {
+      "epoch": 0.34258120078740156,
+      "grad_norm": 1.2047477960586548,
+      "learning_rate": 0.0001475750458587379,
+      "loss": 2.7348,
+      "step": 5569
+    },
+    {
+      "epoch": 0.34264271653543305,
+      "grad_norm": 1.2345035076141357,
+      "learning_rate": 0.0001475580359640088,
+      "loss": 2.5379,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3427042322834646,
+      "grad_norm": 1.0006518363952637,
+      "learning_rate": 0.00014754102429087435,
+      "loss": 2.2756,
+      "step": 5571
+    },
+    {
+      "epoch": 0.34276574803149606,
+      "grad_norm": 1.0978318452835083,
+      "learning_rate": 0.0001475240108399707,
+      "loss": 2.8136,
+      "step": 5572
+    },
+    {
+      "epoch": 0.34282726377952755,
+      "grad_norm": 1.1103577613830566,
+      "learning_rate": 0.000147506995611934,
+      "loss": 2.6644,
+      "step": 5573
+    },
+    {
+      "epoch": 0.34288877952755903,
+      "grad_norm": 1.1895262002944946,
+      "learning_rate": 0.00014748997860740062,
+      "loss": 2.5546,
+      "step": 5574
+    },
+    {
+      "epoch": 0.34295029527559057,
+      "grad_norm": 1.1306966543197632,
+      "learning_rate": 0.00014747295982700683,
+      "loss": 2.1031,
+      "step": 5575
+    },
+    {
+      "epoch": 0.34301181102362205,
+      "grad_norm": 1.1949869394302368,
+      "learning_rate": 0.0001474559392713891,
+      "loss": 2.9101,
+      "step": 5576
+    },
+    {
+      "epoch": 0.34307332677165353,
+      "grad_norm": 0.9476867914199829,
+      "learning_rate": 0.0001474389169411838,
+      "loss": 2.8135,
+      "step": 5577
+    },
+    {
+      "epoch": 0.343134842519685,
+      "grad_norm": 1.2111165523529053,
+      "learning_rate": 0.00014742189283702757,
+      "loss": 2.357,
+      "step": 5578
+    },
+    {
+      "epoch": 0.34319635826771655,
+      "grad_norm": 1.176069736480713,
+      "learning_rate": 0.00014740486695955697,
+      "loss": 3.2309,
+      "step": 5579
+    },
+    {
+      "epoch": 0.34325787401574803,
+      "grad_norm": 1.0057505369186401,
+      "learning_rate": 0.0001473878393094087,
+      "loss": 2.7051,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3433193897637795,
+      "grad_norm": 1.3429596424102783,
+      "learning_rate": 0.00014737080988721948,
+      "loss": 2.7975,
+      "step": 5581
+    },
+    {
+      "epoch": 0.343380905511811,
+      "grad_norm": 1.1150904893875122,
+      "learning_rate": 0.00014735377869362607,
+      "loss": 2.3992,
+      "step": 5582
+    },
+    {
+      "epoch": 0.34344242125984253,
+      "grad_norm": 1.3252249956130981,
+      "learning_rate": 0.00014733674572926545,
+      "loss": 2.5038,
+      "step": 5583
+    },
+    {
+      "epoch": 0.343503937007874,
+      "grad_norm": 1.2963114976882935,
+      "learning_rate": 0.00014731971099477445,
+      "loss": 2.8879,
+      "step": 5584
+    },
+    {
+      "epoch": 0.3435654527559055,
+      "grad_norm": 1.2344744205474854,
+      "learning_rate": 0.00014730267449079016,
+      "loss": 2.8373,
+      "step": 5585
+    },
+    {
+      "epoch": 0.343626968503937,
+      "grad_norm": 0.9122993350028992,
+      "learning_rate": 0.00014728563621794956,
+      "loss": 3.4007,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3436884842519685,
+      "grad_norm": 1.1985026597976685,
+      "learning_rate": 0.00014726859617688988,
+      "loss": 2.4827,
+      "step": 5587
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 1.0347760915756226,
+      "learning_rate": 0.00014725155436824826,
+      "loss": 2.6771,
+      "step": 5588
+    },
+    {
+      "epoch": 0.3438115157480315,
+      "grad_norm": 1.1410675048828125,
+      "learning_rate": 0.00014723451079266199,
+      "loss": 2.5012,
+      "step": 5589
+    },
+    {
+      "epoch": 0.343873031496063,
+      "grad_norm": 1.1605135202407837,
+      "learning_rate": 0.00014721746545076842,
+      "loss": 3.0681,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3439345472440945,
+      "grad_norm": 1.1621202230453491,
+      "learning_rate": 0.00014720041834320494,
+      "loss": 2.3812,
+      "step": 5591
+    },
+    {
+      "epoch": 0.343996062992126,
+      "grad_norm": 1.142979383468628,
+      "learning_rate": 0.00014718336947060898,
+      "loss": 2.7309,
+      "step": 5592
+    },
+    {
+      "epoch": 0.34405757874015747,
+      "grad_norm": 1.0919982194900513,
+      "learning_rate": 0.00014716631883361812,
+      "loss": 2.8598,
+      "step": 5593
+    },
+    {
+      "epoch": 0.344119094488189,
+      "grad_norm": 1.4260917901992798,
+      "learning_rate": 0.00014714926643286994,
+      "loss": 2.3942,
+      "step": 5594
+    },
+    {
+      "epoch": 0.3441806102362205,
+      "grad_norm": 1.098315954208374,
+      "learning_rate": 0.00014713221226900209,
+      "loss": 2.7102,
+      "step": 5595
+    },
+    {
+      "epoch": 0.34424212598425197,
+      "grad_norm": 0.9789575338363647,
+      "learning_rate": 0.00014711515634265233,
+      "loss": 3.4214,
+      "step": 5596
+    },
+    {
+      "epoch": 0.34430364173228345,
+      "grad_norm": 1.3516291379928589,
+      "learning_rate": 0.00014709809865445848,
+      "loss": 2.748,
+      "step": 5597
+    },
+    {
+      "epoch": 0.344365157480315,
+      "grad_norm": 1.2208577394485474,
+      "learning_rate": 0.00014708103920505833,
+      "loss": 2.7194,
+      "step": 5598
+    },
+    {
+      "epoch": 0.34442667322834647,
+      "grad_norm": 1.0689319372177124,
+      "learning_rate": 0.00014706397799508984,
+      "loss": 2.5389,
+      "step": 5599
+    },
+    {
+      "epoch": 0.34448818897637795,
+      "grad_norm": 1.2246854305267334,
+      "learning_rate": 0.00014704691502519102,
+      "loss": 2.9114,
+      "step": 5600
+    },
+    {
+      "epoch": 0.34454970472440943,
+      "grad_norm": 1.1809436082839966,
+      "learning_rate": 0.00014702985029599992,
+      "loss": 3.2169,
+      "step": 5601
+    },
+    {
+      "epoch": 0.34461122047244097,
+      "grad_norm": 1.1981674432754517,
+      "learning_rate": 0.00014701278380815463,
+      "loss": 2.8982,
+      "step": 5602
+    },
+    {
+      "epoch": 0.34467273622047245,
+      "grad_norm": 1.2102093696594238,
+      "learning_rate": 0.00014699571556229344,
+      "loss": 2.4443,
+      "step": 5603
+    },
+    {
+      "epoch": 0.34473425196850394,
+      "grad_norm": 1.0446252822875977,
+      "learning_rate": 0.0001469786455590545,
+      "loss": 2.4776,
+      "step": 5604
+    },
+    {
+      "epoch": 0.3447957677165354,
+      "grad_norm": 1.2244725227355957,
+      "learning_rate": 0.00014696157379907618,
+      "loss": 2.0446,
+      "step": 5605
+    },
+    {
+      "epoch": 0.34485728346456695,
+      "grad_norm": 1.1952320337295532,
+      "learning_rate": 0.0001469445002829969,
+      "loss": 2.8134,
+      "step": 5606
+    },
+    {
+      "epoch": 0.34491879921259844,
+      "grad_norm": 1.0914582014083862,
+      "learning_rate": 0.00014692742501145505,
+      "loss": 2.2139,
+      "step": 5607
+    },
+    {
+      "epoch": 0.3449803149606299,
+      "grad_norm": 1.098558783531189,
+      "learning_rate": 0.00014691034798508916,
+      "loss": 2.8153,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3450418307086614,
+      "grad_norm": 1.226540446281433,
+      "learning_rate": 0.00014689326920453786,
+      "loss": 3.003,
+      "step": 5609
+    },
+    {
+      "epoch": 0.34510334645669294,
+      "grad_norm": 1.0355476140975952,
+      "learning_rate": 0.00014687618867043978,
+      "loss": 1.7095,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3451648622047244,
+      "grad_norm": 1.0737543106079102,
+      "learning_rate": 0.0001468591063834336,
+      "loss": 2.8391,
+      "step": 5611
+    },
+    {
+      "epoch": 0.3452263779527559,
+      "grad_norm": 1.1828428506851196,
+      "learning_rate": 0.00014684202234415815,
+      "loss": 3.2662,
+      "step": 5612
+    },
+    {
+      "epoch": 0.3452878937007874,
+      "grad_norm": 1.1305407285690308,
+      "learning_rate": 0.00014682493655325228,
+      "loss": 1.8385,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3453494094488189,
+      "grad_norm": 1.2114399671554565,
+      "learning_rate": 0.0001468078490113549,
+      "loss": 2.361,
+      "step": 5614
+    },
+    {
+      "epoch": 0.3454109251968504,
+      "grad_norm": 1.2460321187973022,
+      "learning_rate": 0.00014679075971910495,
+      "loss": 3.0148,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3454724409448819,
+      "grad_norm": 0.981701135635376,
+      "learning_rate": 0.00014677366867714152,
+      "loss": 3.1694,
+      "step": 5616
+    },
+    {
+      "epoch": 0.34553395669291337,
+      "grad_norm": 1.1348145008087158,
+      "learning_rate": 0.00014675657588610364,
+      "loss": 2.4845,
+      "step": 5617
+    },
+    {
+      "epoch": 0.3455954724409449,
+      "grad_norm": 1.7058559656143188,
+      "learning_rate": 0.0001467394813466306,
+      "loss": 2.8863,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3456569881889764,
+      "grad_norm": 1.1592429876327515,
+      "learning_rate": 0.00014672238505936158,
+      "loss": 2.7994,
+      "step": 5619
+    },
+    {
+      "epoch": 0.34571850393700787,
+      "grad_norm": 1.2533406019210815,
+      "learning_rate": 0.00014670528702493587,
+      "loss": 2.2796,
+      "step": 5620
+    },
+    {
+      "epoch": 0.34578001968503935,
+      "grad_norm": 1.2520430088043213,
+      "learning_rate": 0.00014668818724399287,
+      "loss": 3.0207,
+      "step": 5621
+    },
+    {
+      "epoch": 0.3458415354330709,
+      "grad_norm": 1.2223916053771973,
+      "learning_rate": 0.00014667108571717199,
+      "loss": 3.1093,
+      "step": 5622
+    },
+    {
+      "epoch": 0.3459030511811024,
+      "grad_norm": 1.0729453563690186,
+      "learning_rate": 0.0001466539824451128,
+      "loss": 2.4297,
+      "step": 5623
+    },
+    {
+      "epoch": 0.34596456692913385,
+      "grad_norm": 1.0979942083358765,
+      "learning_rate": 0.00014663687742845475,
+      "loss": 2.2299,
+      "step": 5624
+    },
+    {
+      "epoch": 0.34602608267716534,
+      "grad_norm": 1.1446964740753174,
+      "learning_rate": 0.00014661977066783756,
+      "loss": 3.2453,
+      "step": 5625
+    },
+    {
+      "epoch": 0.3460875984251969,
+      "grad_norm": 1.100608468055725,
+      "learning_rate": 0.00014660266216390093,
+      "loss": 3.0728,
+      "step": 5626
+    },
+    {
+      "epoch": 0.34614911417322836,
+      "grad_norm": 1.0959219932556152,
+      "learning_rate": 0.00014658555191728455,
+      "loss": 2.6071,
+      "step": 5627
+    },
+    {
+      "epoch": 0.34621062992125984,
+      "grad_norm": 1.1335673332214355,
+      "learning_rate": 0.00014656843992862835,
+      "loss": 2.6664,
+      "step": 5628
+    },
+    {
+      "epoch": 0.3462721456692913,
+      "grad_norm": 1.0697386264801025,
+      "learning_rate": 0.00014655132619857217,
+      "loss": 2.797,
+      "step": 5629
+    },
+    {
+      "epoch": 0.34633366141732286,
+      "grad_norm": 1.23844575881958,
+      "learning_rate": 0.00014653421072775594,
+      "loss": 2.808,
+      "step": 5630
+    },
+    {
+      "epoch": 0.34639517716535434,
+      "grad_norm": 1.1883106231689453,
+      "learning_rate": 0.0001465170935168197,
+      "loss": 2.6342,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3464566929133858,
+      "grad_norm": 1.1294312477111816,
+      "learning_rate": 0.0001464999745664036,
+      "loss": 2.8796,
+      "step": 5632
+    },
+    {
+      "epoch": 0.3465182086614173,
+      "grad_norm": 1.2326726913452148,
+      "learning_rate": 0.0001464828538771477,
+      "loss": 2.9866,
+      "step": 5633
+    },
+    {
+      "epoch": 0.34657972440944884,
+      "grad_norm": 1.0879875421524048,
+      "learning_rate": 0.00014646573144969224,
+      "loss": 3.1238,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3466412401574803,
+      "grad_norm": 1.0808444023132324,
+      "learning_rate": 0.00014644860728467757,
+      "loss": 3.1795,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3467027559055118,
+      "grad_norm": 1.319830298423767,
+      "learning_rate": 0.00014643148138274398,
+      "loss": 1.9108,
+      "step": 5636
+    },
+    {
+      "epoch": 0.3467642716535433,
+      "grad_norm": 1.1608459949493408,
+      "learning_rate": 0.00014641435374453186,
+      "loss": 2.0049,
+      "step": 5637
+    },
+    {
+      "epoch": 0.3468257874015748,
+      "grad_norm": 1.2274948358535767,
+      "learning_rate": 0.00014639722437068177,
+      "loss": 2.8748,
+      "step": 5638
+    },
+    {
+      "epoch": 0.3468873031496063,
+      "grad_norm": 1.2649035453796387,
+      "learning_rate": 0.00014638009326183423,
+      "loss": 2.4499,
+      "step": 5639
+    },
+    {
+      "epoch": 0.3469488188976378,
+      "grad_norm": 1.1770399808883667,
+      "learning_rate": 0.00014636296041862975,
+      "loss": 2.6727,
+      "step": 5640
+    },
+    {
+      "epoch": 0.34701033464566927,
+      "grad_norm": 1.0218725204467773,
+      "learning_rate": 0.0001463458258417091,
+      "loss": 2.4832,
+      "step": 5641
+    },
+    {
+      "epoch": 0.3470718503937008,
+      "grad_norm": 1.3728197813034058,
+      "learning_rate": 0.00014632868953171304,
+      "loss": 2.6785,
+      "step": 5642
+    },
+    {
+      "epoch": 0.3471333661417323,
+      "grad_norm": 1.3338273763656616,
+      "learning_rate": 0.0001463115514892823,
+      "loss": 2.0622,
+      "step": 5643
+    },
+    {
+      "epoch": 0.3471948818897638,
+      "grad_norm": 1.3320562839508057,
+      "learning_rate": 0.00014629441171505776,
+      "loss": 2.5708,
+      "step": 5644
+    },
+    {
+      "epoch": 0.34725639763779526,
+      "grad_norm": 1.6091145277023315,
+      "learning_rate": 0.00014627727020968036,
+      "loss": 2.6763,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3473179133858268,
+      "grad_norm": 1.1792770624160767,
+      "learning_rate": 0.00014626012697379115,
+      "loss": 3.3468,
+      "step": 5646
+    },
+    {
+      "epoch": 0.3473794291338583,
+      "grad_norm": 1.1856772899627686,
+      "learning_rate": 0.00014624298200803109,
+      "loss": 2.7423,
+      "step": 5647
+    },
+    {
+      "epoch": 0.34744094488188976,
+      "grad_norm": 1.0906062126159668,
+      "learning_rate": 0.00014622583531304138,
+      "loss": 1.6815,
+      "step": 5648
+    },
+    {
+      "epoch": 0.34750246062992124,
+      "grad_norm": 1.2465893030166626,
+      "learning_rate": 0.00014620868688946322,
+      "loss": 2.1266,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3475639763779528,
+      "grad_norm": 1.2261359691619873,
+      "learning_rate": 0.00014619153673793778,
+      "loss": 2.6624,
+      "step": 5650
+    },
+    {
+      "epoch": 0.34762549212598426,
+      "grad_norm": 0.9183244705200195,
+      "learning_rate": 0.00014617438485910647,
+      "loss": 2.6558,
+      "step": 5651
+    },
+    {
+      "epoch": 0.34768700787401574,
+      "grad_norm": 1.1117944717407227,
+      "learning_rate": 0.00014615723125361067,
+      "loss": 2.3557,
+      "step": 5652
+    },
+    {
+      "epoch": 0.3477485236220472,
+      "grad_norm": 1.0343488454818726,
+      "learning_rate": 0.00014614007592209175,
+      "loss": 3.1991,
+      "step": 5653
+    },
+    {
+      "epoch": 0.34781003937007876,
+      "grad_norm": 0.973900556564331,
+      "learning_rate": 0.0001461229188651913,
+      "loss": 2.8509,
+      "step": 5654
+    },
+    {
+      "epoch": 0.34787155511811024,
+      "grad_norm": 1.0809930562973022,
+      "learning_rate": 0.0001461057600835509,
+      "loss": 3.3746,
+      "step": 5655
+    },
+    {
+      "epoch": 0.3479330708661417,
+      "grad_norm": 1.2144532203674316,
+      "learning_rate": 0.00014608859957781212,
+      "loss": 2.1083,
+      "step": 5656
+    },
+    {
+      "epoch": 0.3479945866141732,
+      "grad_norm": 0.9983019232749939,
+      "learning_rate": 0.00014607143734861672,
+      "loss": 2.6826,
+      "step": 5657
+    },
+    {
+      "epoch": 0.34805610236220474,
+      "grad_norm": 1.1571950912475586,
+      "learning_rate": 0.0001460542733966065,
+      "loss": 2.6629,
+      "step": 5658
+    },
+    {
+      "epoch": 0.3481176181102362,
+      "grad_norm": 1.02420175075531,
+      "learning_rate": 0.00014603710772242322,
+      "loss": 2.8199,
+      "step": 5659
+    },
+    {
+      "epoch": 0.3481791338582677,
+      "grad_norm": 1.3770450353622437,
+      "learning_rate": 0.00014601994032670886,
+      "loss": 2.8841,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3482406496062992,
+      "grad_norm": 0.9440857172012329,
+      "learning_rate": 0.0001460027712101053,
+      "loss": 2.6542,
+      "step": 5661
+    },
+    {
+      "epoch": 0.34830216535433073,
+      "grad_norm": 0.9098318219184875,
+      "learning_rate": 0.00014598560037325462,
+      "loss": 2.4271,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3483636811023622,
+      "grad_norm": 1.2827099561691284,
+      "learning_rate": 0.00014596842781679892,
+      "loss": 2.1567,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3484251968503937,
+      "grad_norm": 1.2355599403381348,
+      "learning_rate": 0.00014595125354138037,
+      "loss": 3.1375,
+      "step": 5664
+    },
+    {
+      "epoch": 0.3484867125984252,
+      "grad_norm": 1.0608857870101929,
+      "learning_rate": 0.00014593407754764118,
+      "loss": 1.9595,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3485482283464567,
+      "grad_norm": 1.2017579078674316,
+      "learning_rate": 0.00014591689983622358,
+      "loss": 2.4611,
+      "step": 5666
+    },
+    {
+      "epoch": 0.3486097440944882,
+      "grad_norm": 1.1359628438949585,
+      "learning_rate": 0.00014589972040776999,
+      "loss": 2.945,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3486712598425197,
+      "grad_norm": 1.150930404663086,
+      "learning_rate": 0.0001458825392629228,
+      "loss": 1.7895,
+      "step": 5668
+    },
+    {
+      "epoch": 0.34873277559055116,
+      "grad_norm": 1.4479113817214966,
+      "learning_rate": 0.00014586535640232447,
+      "loss": 1.8402,
+      "step": 5669
+    },
+    {
+      "epoch": 0.3487942913385827,
+      "grad_norm": 1.2128782272338867,
+      "learning_rate": 0.00014584817182661757,
+      "loss": 2.6291,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3488558070866142,
+      "grad_norm": 1.0060373544692993,
+      "learning_rate": 0.0001458309855364447,
+      "loss": 2.7946,
+      "step": 5671
+    },
+    {
+      "epoch": 0.34891732283464566,
+      "grad_norm": 1.0398375988006592,
+      "learning_rate": 0.00014581379753244854,
+      "loss": 2.5065,
+      "step": 5672
+    },
+    {
+      "epoch": 0.34897883858267714,
+      "grad_norm": 1.5352152585983276,
+      "learning_rate": 0.0001457966078152718,
+      "loss": 2.5326,
+      "step": 5673
+    },
+    {
+      "epoch": 0.3490403543307087,
+      "grad_norm": 1.1795942783355713,
+      "learning_rate": 0.00014577941638555731,
+      "loss": 2.4938,
+      "step": 5674
+    },
+    {
+      "epoch": 0.34910187007874016,
+      "grad_norm": 1.1416047811508179,
+      "learning_rate": 0.0001457622232439479,
+      "loss": 2.4843,
+      "step": 5675
+    },
+    {
+      "epoch": 0.34916338582677164,
+      "grad_norm": 1.0240055322647095,
+      "learning_rate": 0.00014574502839108657,
+      "loss": 2.9168,
+      "step": 5676
+    },
+    {
+      "epoch": 0.3492249015748031,
+      "grad_norm": 1.197385549545288,
+      "learning_rate": 0.0001457278318276162,
+      "loss": 2.8684,
+      "step": 5677
+    },
+    {
+      "epoch": 0.34928641732283466,
+      "grad_norm": 1.2119923830032349,
+      "learning_rate": 0.00014571063355417994,
+      "loss": 3.1461,
+      "step": 5678
+    },
+    {
+      "epoch": 0.34934793307086615,
+      "grad_norm": 1.2897403240203857,
+      "learning_rate": 0.00014569343357142086,
+      "loss": 2.4273,
+      "step": 5679
+    },
+    {
+      "epoch": 0.3494094488188976,
+      "grad_norm": 1.1193853616714478,
+      "learning_rate": 0.0001456762318799822,
+      "loss": 2.6337,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3494709645669291,
+      "grad_norm": 1.212066888809204,
+      "learning_rate": 0.00014565902848050712,
+      "loss": 2.2013,
+      "step": 5681
+    },
+    {
+      "epoch": 0.34953248031496065,
+      "grad_norm": 1.1577565670013428,
+      "learning_rate": 0.00014564182337363897,
+      "loss": 2.9951,
+      "step": 5682
+    },
+    {
+      "epoch": 0.34959399606299213,
+      "grad_norm": 1.1848540306091309,
+      "learning_rate": 0.00014562461656002115,
+      "loss": 2.8646,
+      "step": 5683
+    },
+    {
+      "epoch": 0.3496555118110236,
+      "grad_norm": 1.1452038288116455,
+      "learning_rate": 0.0001456074080402971,
+      "loss": 3.4896,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3497170275590551,
+      "grad_norm": 1.3463855981826782,
+      "learning_rate": 0.0001455901978151103,
+      "loss": 2.7646,
+      "step": 5685
+    },
+    {
+      "epoch": 0.34977854330708663,
+      "grad_norm": 1.7771154642105103,
+      "learning_rate": 0.00014557298588510428,
+      "loss": 2.4017,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3498400590551181,
+      "grad_norm": 1.0504770278930664,
+      "learning_rate": 0.00014555577225092273,
+      "loss": 2.0251,
+      "step": 5687
+    },
+    {
+      "epoch": 0.3499015748031496,
+      "grad_norm": 1.2215855121612549,
+      "learning_rate": 0.00014553855691320933,
+      "loss": 2.8573,
+      "step": 5688
+    },
+    {
+      "epoch": 0.3499630905511811,
+      "grad_norm": 1.22599196434021,
+      "learning_rate": 0.00014552133987260783,
+      "loss": 2.268,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3500246062992126,
+      "grad_norm": 1.065226435661316,
+      "learning_rate": 0.00014550412112976206,
+      "loss": 2.8069,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3500861220472441,
+      "grad_norm": 1.224968433380127,
+      "learning_rate": 0.00014548690068531586,
+      "loss": 2.606,
+      "step": 5691
+    },
+    {
+      "epoch": 0.3501476377952756,
+      "grad_norm": 1.6194692850112915,
+      "learning_rate": 0.00014546967853991326,
+      "loss": 2.5526,
+      "step": 5692
+    },
+    {
+      "epoch": 0.35020915354330706,
+      "grad_norm": 1.2269930839538574,
+      "learning_rate": 0.00014545245469419822,
+      "loss": 2.6874,
+      "step": 5693
+    },
+    {
+      "epoch": 0.3502706692913386,
+      "grad_norm": 1.1159192323684692,
+      "learning_rate": 0.0001454352291488148,
+      "loss": 3.0038,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3503321850393701,
+      "grad_norm": 1.5146400928497314,
+      "learning_rate": 0.00014541800190440718,
+      "loss": 1.5139,
+      "step": 5695
+    },
+    {
+      "epoch": 0.35039370078740156,
+      "grad_norm": 1.040730595588684,
+      "learning_rate": 0.00014540077296161955,
+      "loss": 3.0693,
+      "step": 5696
+    },
+    {
+      "epoch": 0.35045521653543305,
+      "grad_norm": 1.100232720375061,
+      "learning_rate": 0.00014538354232109617,
+      "loss": 2.762,
+      "step": 5697
+    },
+    {
+      "epoch": 0.3505167322834646,
+      "grad_norm": 1.1635102033615112,
+      "learning_rate": 0.00014536630998348137,
+      "loss": 2.6213,
+      "step": 5698
+    },
+    {
+      "epoch": 0.35057824803149606,
+      "grad_norm": 1.1227120161056519,
+      "learning_rate": 0.00014534907594941954,
+      "loss": 2.6769,
+      "step": 5699
+    },
+    {
+      "epoch": 0.35063976377952755,
+      "grad_norm": 1.1472488641738892,
+      "learning_rate": 0.00014533184021955518,
+      "loss": 2.9009,
+      "step": 5700
+    },
+    {
+      "epoch": 0.35070127952755903,
+      "grad_norm": 1.029097557067871,
+      "learning_rate": 0.00014531460279453273,
+      "loss": 2.3507,
+      "step": 5701
+    },
+    {
+      "epoch": 0.35076279527559057,
+      "grad_norm": 1.2669126987457275,
+      "learning_rate": 0.00014529736367499684,
+      "loss": 2.7023,
+      "step": 5702
+    },
+    {
+      "epoch": 0.35082431102362205,
+      "grad_norm": 1.2648082971572876,
+      "learning_rate": 0.00014528012286159212,
+      "loss": 2.9905,
+      "step": 5703
+    },
+    {
+      "epoch": 0.35088582677165353,
+      "grad_norm": 1.253461241722107,
+      "learning_rate": 0.0001452628803549633,
+      "loss": 2.6219,
+      "step": 5704
+    },
+    {
+      "epoch": 0.350947342519685,
+      "grad_norm": 1.2121690511703491,
+      "learning_rate": 0.00014524563615575512,
+      "loss": 2.3728,
+      "step": 5705
+    },
+    {
+      "epoch": 0.35100885826771655,
+      "grad_norm": 1.1923835277557373,
+      "learning_rate": 0.0001452283902646125,
+      "loss": 2.1794,
+      "step": 5706
+    },
+    {
+      "epoch": 0.35107037401574803,
+      "grad_norm": 1.1955904960632324,
+      "learning_rate": 0.00014521114268218024,
+      "loss": 2.3443,
+      "step": 5707
+    },
+    {
+      "epoch": 0.3511318897637795,
+      "grad_norm": 1.0499769449234009,
+      "learning_rate": 0.00014519389340910337,
+      "loss": 2.9718,
+      "step": 5708
+    },
+    {
+      "epoch": 0.351193405511811,
+      "grad_norm": 1.2483720779418945,
+      "learning_rate": 0.00014517664244602689,
+      "loss": 2.8342,
+      "step": 5709
+    },
+    {
+      "epoch": 0.35125492125984253,
+      "grad_norm": 1.0230828523635864,
+      "learning_rate": 0.00014515938979359588,
+      "loss": 2.801,
+      "step": 5710
+    },
+    {
+      "epoch": 0.351316437007874,
+      "grad_norm": 1.1593916416168213,
+      "learning_rate": 0.00014514213545245553,
+      "loss": 2.8852,
+      "step": 5711
+    },
+    {
+      "epoch": 0.3513779527559055,
+      "grad_norm": 1.8764874935150146,
+      "learning_rate": 0.00014512487942325102,
+      "loss": 2.3402,
+      "step": 5712
+    },
+    {
+      "epoch": 0.351439468503937,
+      "grad_norm": 1.6455585956573486,
+      "learning_rate": 0.00014510762170662767,
+      "loss": 2.808,
+      "step": 5713
+    },
+    {
+      "epoch": 0.3515009842519685,
+      "grad_norm": 1.411838173866272,
+      "learning_rate": 0.00014509036230323075,
+      "loss": 2.9584,
+      "step": 5714
+    },
+    {
+      "epoch": 0.3515625,
+      "grad_norm": 1.053407073020935,
+      "learning_rate": 0.00014507310121370572,
+      "loss": 2.9738,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3516240157480315,
+      "grad_norm": 1.1628550291061401,
+      "learning_rate": 0.00014505583843869805,
+      "loss": 3.2222,
+      "step": 5716
+    },
+    {
+      "epoch": 0.351685531496063,
+      "grad_norm": 1.2567732334136963,
+      "learning_rate": 0.00014503857397885325,
+      "loss": 2.0262,
+      "step": 5717
+    },
+    {
+      "epoch": 0.3517470472440945,
+      "grad_norm": 1.2259002923965454,
+      "learning_rate": 0.00014502130783481694,
+      "loss": 2.3824,
+      "step": 5718
+    },
+    {
+      "epoch": 0.351808562992126,
+      "grad_norm": 1.2829350233078003,
+      "learning_rate": 0.00014500404000723479,
+      "loss": 2.2052,
+      "step": 5719
+    },
+    {
+      "epoch": 0.35187007874015747,
+      "grad_norm": 1.4014040231704712,
+      "learning_rate": 0.00014498677049675242,
+      "loss": 2.7539,
+      "step": 5720
+    },
+    {
+      "epoch": 0.351931594488189,
+      "grad_norm": 1.4025734663009644,
+      "learning_rate": 0.00014496949930401572,
+      "loss": 3.188,
+      "step": 5721
+    },
+    {
+      "epoch": 0.3519931102362205,
+      "grad_norm": 1.2160841226577759,
+      "learning_rate": 0.00014495222642967054,
+      "loss": 2.4371,
+      "step": 5722
+    },
+    {
+      "epoch": 0.35205462598425197,
+      "grad_norm": 1.3853944540023804,
+      "learning_rate": 0.0001449349518743627,
+      "loss": 2.8048,
+      "step": 5723
+    },
+    {
+      "epoch": 0.35211614173228345,
+      "grad_norm": 1.0029784440994263,
+      "learning_rate": 0.00014491767563873826,
+      "loss": 3.0287,
+      "step": 5724
+    },
+    {
+      "epoch": 0.352177657480315,
+      "grad_norm": 1.087759017944336,
+      "learning_rate": 0.0001449003977234432,
+      "loss": 2.8066,
+      "step": 5725
+    },
+    {
+      "epoch": 0.35223917322834647,
+      "grad_norm": 1.1711490154266357,
+      "learning_rate": 0.00014488311812912365,
+      "loss": 2.869,
+      "step": 5726
+    },
+    {
+      "epoch": 0.35230068897637795,
+      "grad_norm": 1.1180987358093262,
+      "learning_rate": 0.00014486583685642575,
+      "loss": 2.7001,
+      "step": 5727
+    },
+    {
+      "epoch": 0.35236220472440943,
+      "grad_norm": 1.086405873298645,
+      "learning_rate": 0.00014484855390599574,
+      "loss": 3.4129,
+      "step": 5728
+    },
+    {
+      "epoch": 0.35242372047244097,
+      "grad_norm": 1.0607857704162598,
+      "learning_rate": 0.0001448312692784799,
+      "loss": 2.1355,
+      "step": 5729
+    },
+    {
+      "epoch": 0.35248523622047245,
+      "grad_norm": 0.9999568462371826,
+      "learning_rate": 0.00014481398297452454,
+      "loss": 2.7715,
+      "step": 5730
+    },
+    {
+      "epoch": 0.35254675196850394,
+      "grad_norm": 1.8757684230804443,
+      "learning_rate": 0.00014479669499477615,
+      "loss": 1.4261,
+      "step": 5731
+    },
+    {
+      "epoch": 0.3526082677165354,
+      "grad_norm": 1.6431925296783447,
+      "learning_rate": 0.00014477940533988112,
+      "loss": 2.1853,
+      "step": 5732
+    },
+    {
+      "epoch": 0.35266978346456695,
+      "grad_norm": 1.2546237707138062,
+      "learning_rate": 0.00014476211401048605,
+      "loss": 2.9054,
+      "step": 5733
+    },
+    {
+      "epoch": 0.35273129921259844,
+      "grad_norm": 1.072466492652893,
+      "learning_rate": 0.00014474482100723752,
+      "loss": 2.8738,
+      "step": 5734
+    },
+    {
+      "epoch": 0.3527928149606299,
+      "grad_norm": 1.4189919233322144,
+      "learning_rate": 0.0001447275263307822,
+      "loss": 2.753,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3528543307086614,
+      "grad_norm": 1.0770916938781738,
+      "learning_rate": 0.00014471022998176675,
+      "loss": 3.0434,
+      "step": 5736
+    },
+    {
+      "epoch": 0.35291584645669294,
+      "grad_norm": 1.2783455848693848,
+      "learning_rate": 0.00014469293196083807,
+      "loss": 2.7243,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3529773622047244,
+      "grad_norm": 1.3390834331512451,
+      "learning_rate": 0.0001446756322686429,
+      "loss": 2.5843,
+      "step": 5738
+    },
+    {
+      "epoch": 0.3530388779527559,
+      "grad_norm": 2.364964008331299,
+      "learning_rate": 0.00014465833090582822,
+      "loss": 2.3063,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3531003937007874,
+      "grad_norm": 1.2726184129714966,
+      "learning_rate": 0.000144641027873041,
+      "loss": 2.6724,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3531619094488189,
+      "grad_norm": 1.0299506187438965,
+      "learning_rate": 0.00014462372317092825,
+      "loss": 2.8381,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3532234251968504,
+      "grad_norm": 1.118635892868042,
+      "learning_rate": 0.00014460641680013708,
+      "loss": 2.8838,
+      "step": 5742
+    },
+    {
+      "epoch": 0.3532849409448819,
+      "grad_norm": 0.9622350931167603,
+      "learning_rate": 0.00014458910876131466,
+      "loss": 2.9881,
+      "step": 5743
+    },
+    {
+      "epoch": 0.35334645669291337,
+      "grad_norm": 1.195245623588562,
+      "learning_rate": 0.0001445717990551082,
+      "loss": 2.6626,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3534079724409449,
+      "grad_norm": 1.198189616203308,
+      "learning_rate": 0.000144554487682165,
+      "loss": 2.0191,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3534694881889764,
+      "grad_norm": 1.1824406385421753,
+      "learning_rate": 0.00014453717464313242,
+      "loss": 2.491,
+      "step": 5746
+    },
+    {
+      "epoch": 0.35353100393700787,
+      "grad_norm": 1.1755725145339966,
+      "learning_rate": 0.00014451985993865782,
+      "loss": 3.0886,
+      "step": 5747
+    },
+    {
+      "epoch": 0.35359251968503935,
+      "grad_norm": 1.0812524557113647,
+      "learning_rate": 0.00014450254356938874,
+      "loss": 2.6617,
+      "step": 5748
+    },
+    {
+      "epoch": 0.3536540354330709,
+      "grad_norm": 1.0912212133407593,
+      "learning_rate": 0.00014448522553597265,
+      "loss": 2.1455,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3537155511811024,
+      "grad_norm": 1.3374228477478027,
+      "learning_rate": 0.0001444679058390572,
+      "loss": 2.5362,
+      "step": 5750
+    },
+    {
+      "epoch": 0.35377706692913385,
+      "grad_norm": 1.1447547674179077,
+      "learning_rate": 0.00014445058447929004,
+      "loss": 2.41,
+      "step": 5751
+    },
+    {
+      "epoch": 0.35383858267716534,
+      "grad_norm": 1.022451400756836,
+      "learning_rate": 0.0001444332614573189,
+      "loss": 2.7212,
+      "step": 5752
+    },
+    {
+      "epoch": 0.3539000984251969,
+      "grad_norm": 1.1377314329147339,
+      "learning_rate": 0.0001444159367737915,
+      "loss": 2.8607,
+      "step": 5753
+    },
+    {
+      "epoch": 0.35396161417322836,
+      "grad_norm": 1.2245326042175293,
+      "learning_rate": 0.00014439861042935575,
+      "loss": 2.9239,
+      "step": 5754
+    },
+    {
+      "epoch": 0.35402312992125984,
+      "grad_norm": 1.3145171403884888,
+      "learning_rate": 0.00014438128242465957,
+      "loss": 3.0507,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3540846456692913,
+      "grad_norm": 1.2522521018981934,
+      "learning_rate": 0.0001443639527603509,
+      "loss": 2.5388,
+      "step": 5756
+    },
+    {
+      "epoch": 0.35414616141732286,
+      "grad_norm": 1.2172229290008545,
+      "learning_rate": 0.00014434662143707774,
+      "loss": 3.0482,
+      "step": 5757
+    },
+    {
+      "epoch": 0.35420767716535434,
+      "grad_norm": 1.1669977903366089,
+      "learning_rate": 0.00014432928845548823,
+      "loss": 2.4006,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3542691929133858,
+      "grad_norm": 1.3646152019500732,
+      "learning_rate": 0.00014431195381623055,
+      "loss": 2.9879,
+      "step": 5759
+    },
+    {
+      "epoch": 0.3543307086614173,
+      "grad_norm": 1.009785532951355,
+      "learning_rate": 0.00014429461751995287,
+      "loss": 2.9162,
+      "step": 5760
+    },
+    {
+      "epoch": 0.35439222440944884,
+      "grad_norm": 1.1014037132263184,
+      "learning_rate": 0.0001442772795673035,
+      "loss": 2.4937,
+      "step": 5761
+    },
+    {
+      "epoch": 0.3544537401574803,
+      "grad_norm": 1.228633165359497,
+      "learning_rate": 0.00014425993995893075,
+      "loss": 2.5241,
+      "step": 5762
+    },
+    {
+      "epoch": 0.3545152559055118,
+      "grad_norm": 1.0965694189071655,
+      "learning_rate": 0.0001442425986954831,
+      "loss": 2.8867,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3545767716535433,
+      "grad_norm": 1.1242976188659668,
+      "learning_rate": 0.0001442252557776089,
+      "loss": 2.4513,
+      "step": 5764
+    },
+    {
+      "epoch": 0.3546382874015748,
+      "grad_norm": 1.2169724702835083,
+      "learning_rate": 0.00014420791120595675,
+      "loss": 2.1841,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3546998031496063,
+      "grad_norm": 0.930447518825531,
+      "learning_rate": 0.00014419056498117525,
+      "loss": 3.1169,
+      "step": 5766
+    },
+    {
+      "epoch": 0.3547613188976378,
+      "grad_norm": 1.2256653308868408,
+      "learning_rate": 0.00014417321710391302,
+      "loss": 3.5943,
+      "step": 5767
+    },
+    {
+      "epoch": 0.35482283464566927,
+      "grad_norm": 1.1452436447143555,
+      "learning_rate": 0.0001441558675748188,
+      "loss": 2.7132,
+      "step": 5768
+    },
+    {
+      "epoch": 0.3548843503937008,
+      "grad_norm": 1.1125651597976685,
+      "learning_rate": 0.0001441385163945413,
+      "loss": 2.7386,
+      "step": 5769
+    },
+    {
+      "epoch": 0.3549458661417323,
+      "grad_norm": 1.5557483434677124,
+      "learning_rate": 0.0001441211635637295,
+      "loss": 3.0688,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3550073818897638,
+      "grad_norm": 1.1778714656829834,
+      "learning_rate": 0.00014410380908303212,
+      "loss": 2.9986,
+      "step": 5771
+    },
+    {
+      "epoch": 0.35506889763779526,
+      "grad_norm": 1.0561258792877197,
+      "learning_rate": 0.00014408645295309826,
+      "loss": 3.0476,
+      "step": 5772
+    },
+    {
+      "epoch": 0.3551304133858268,
+      "grad_norm": 1.6193792819976807,
+      "learning_rate": 0.00014406909517457688,
+      "loss": 2.6566,
+      "step": 5773
+    },
+    {
+      "epoch": 0.3551919291338583,
+      "grad_norm": 1.025693416595459,
+      "learning_rate": 0.00014405173574811707,
+      "loss": 2.9415,
+      "step": 5774
+    },
+    {
+      "epoch": 0.35525344488188976,
+      "grad_norm": 1.411733865737915,
+      "learning_rate": 0.00014403437467436803,
+      "loss": 2.7058,
+      "step": 5775
+    },
+    {
+      "epoch": 0.35531496062992124,
+      "grad_norm": 1.0050512552261353,
+      "learning_rate": 0.00014401701195397888,
+      "loss": 3.1132,
+      "step": 5776
+    },
+    {
+      "epoch": 0.3553764763779528,
+      "grad_norm": 1.0643622875213623,
+      "learning_rate": 0.00014399964758759893,
+      "loss": 3.0132,
+      "step": 5777
+    },
+    {
+      "epoch": 0.35543799212598426,
+      "grad_norm": 1.5118457078933716,
+      "learning_rate": 0.0001439822815758775,
+      "loss": 1.876,
+      "step": 5778
+    },
+    {
+      "epoch": 0.35549950787401574,
+      "grad_norm": 1.122490406036377,
+      "learning_rate": 0.000143964913919464,
+      "loss": 2.8077,
+      "step": 5779
+    },
+    {
+      "epoch": 0.3555610236220472,
+      "grad_norm": 1.236480474472046,
+      "learning_rate": 0.00014394754461900786,
+      "loss": 2.8008,
+      "step": 5780
+    },
+    {
+      "epoch": 0.35562253937007876,
+      "grad_norm": 2.3513495922088623,
+      "learning_rate": 0.00014393017367515863,
+      "loss": 3.2542,
+      "step": 5781
+    },
+    {
+      "epoch": 0.35568405511811024,
+      "grad_norm": 1.146382451057434,
+      "learning_rate": 0.00014391280108856587,
+      "loss": 1.9381,
+      "step": 5782
+    },
+    {
+      "epoch": 0.3557455708661417,
+      "grad_norm": 1.1445420980453491,
+      "learning_rate": 0.0001438954268598792,
+      "loss": 2.4676,
+      "step": 5783
+    },
+    {
+      "epoch": 0.3558070866141732,
+      "grad_norm": 1.226932406425476,
+      "learning_rate": 0.00014387805098974835,
+      "loss": 3.3059,
+      "step": 5784
+    },
+    {
+      "epoch": 0.35586860236220474,
+      "grad_norm": 1.299184799194336,
+      "learning_rate": 0.00014386067347882302,
+      "loss": 3.0752,
+      "step": 5785
+    },
+    {
+      "epoch": 0.3559301181102362,
+      "grad_norm": 1.6067891120910645,
+      "learning_rate": 0.00014384329432775314,
+      "loss": 3.3711,
+      "step": 5786
+    },
+    {
+      "epoch": 0.3559916338582677,
+      "grad_norm": 1.046858549118042,
+      "learning_rate": 0.00014382591353718848,
+      "loss": 1.6836,
+      "step": 5787
+    },
+    {
+      "epoch": 0.3560531496062992,
+      "grad_norm": 1.3106520175933838,
+      "learning_rate": 0.00014380853110777906,
+      "loss": 2.2259,
+      "step": 5788
+    },
+    {
+      "epoch": 0.35611466535433073,
+      "grad_norm": 1.0951091051101685,
+      "learning_rate": 0.00014379114704017484,
+      "loss": 2.562,
+      "step": 5789
+    },
+    {
+      "epoch": 0.3561761811023622,
+      "grad_norm": 1.2188637256622314,
+      "learning_rate": 0.0001437737613350259,
+      "loss": 2.7062,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3562376968503937,
+      "grad_norm": 1.2561814785003662,
+      "learning_rate": 0.0001437563739929824,
+      "loss": 2.9564,
+      "step": 5791
+    },
+    {
+      "epoch": 0.3562992125984252,
+      "grad_norm": 1.2519866228103638,
+      "learning_rate": 0.0001437389850146945,
+      "loss": 2.5937,
+      "step": 5792
+    },
+    {
+      "epoch": 0.3563607283464567,
+      "grad_norm": 1.3290822505950928,
+      "learning_rate": 0.00014372159440081248,
+      "loss": 2.9712,
+      "step": 5793
+    },
+    {
+      "epoch": 0.3564222440944882,
+      "grad_norm": 1.2427573204040527,
+      "learning_rate": 0.0001437042021519866,
+      "loss": 2.3311,
+      "step": 5794
+    },
+    {
+      "epoch": 0.3564837598425197,
+      "grad_norm": 1.0927122831344604,
+      "learning_rate": 0.00014368680826886725,
+      "loss": 2.2031,
+      "step": 5795
+    },
+    {
+      "epoch": 0.35654527559055116,
+      "grad_norm": 1.098605751991272,
+      "learning_rate": 0.00014366941275210487,
+      "loss": 2.8206,
+      "step": 5796
+    },
+    {
+      "epoch": 0.3566067913385827,
+      "grad_norm": 1.297592043876648,
+      "learning_rate": 0.00014365201560234998,
+      "loss": 2.6697,
+      "step": 5797
+    },
+    {
+      "epoch": 0.3566683070866142,
+      "grad_norm": 1.1746894121170044,
+      "learning_rate": 0.00014363461682025312,
+      "loss": 2.6593,
+      "step": 5798
+    },
+    {
+      "epoch": 0.35672982283464566,
+      "grad_norm": 1.214983582496643,
+      "learning_rate": 0.0001436172164064649,
+      "loss": 2.7011,
+      "step": 5799
+    },
+    {
+      "epoch": 0.35679133858267714,
+      "grad_norm": 1.097553014755249,
+      "learning_rate": 0.00014359981436163602,
+      "loss": 1.6882,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3568528543307087,
+      "grad_norm": 0.997981071472168,
+      "learning_rate": 0.00014358241068641717,
+      "loss": 2.758,
+      "step": 5801
+    },
+    {
+      "epoch": 0.35691437007874016,
+      "grad_norm": 1.5933560132980347,
+      "learning_rate": 0.0001435650053814592,
+      "loss": 2.6892,
+      "step": 5802
+    },
+    {
+      "epoch": 0.35697588582677164,
+      "grad_norm": 1.3218094110488892,
+      "learning_rate": 0.00014354759844741294,
+      "loss": 2.7527,
+      "step": 5803
+    },
+    {
+      "epoch": 0.3570374015748031,
+      "grad_norm": 1.0995084047317505,
+      "learning_rate": 0.00014353018988492934,
+      "loss": 2.2924,
+      "step": 5804
+    },
+    {
+      "epoch": 0.35709891732283466,
+      "grad_norm": 1.1948720216751099,
+      "learning_rate": 0.00014351277969465936,
+      "loss": 2.7521,
+      "step": 5805
+    },
+    {
+      "epoch": 0.35716043307086615,
+      "grad_norm": 1.042844533920288,
+      "learning_rate": 0.00014349536787725407,
+      "loss": 2.9223,
+      "step": 5806
+    },
+    {
+      "epoch": 0.3572219488188976,
+      "grad_norm": 1.1261646747589111,
+      "learning_rate": 0.00014347795443336454,
+      "loss": 2.8659,
+      "step": 5807
+    },
+    {
+      "epoch": 0.3572834645669291,
+      "grad_norm": 1.077876091003418,
+      "learning_rate": 0.00014346053936364195,
+      "loss": 2.6069,
+      "step": 5808
+    },
+    {
+      "epoch": 0.35734498031496065,
+      "grad_norm": 1.0754034519195557,
+      "learning_rate": 0.00014344312266873754,
+      "loss": 2.7805,
+      "step": 5809
+    },
+    {
+      "epoch": 0.35740649606299213,
+      "grad_norm": 1.3852568864822388,
+      "learning_rate": 0.00014342570434930257,
+      "loss": 2.3518,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3574680118110236,
+      "grad_norm": 1.0706230401992798,
+      "learning_rate": 0.00014340828440598843,
+      "loss": 2.6259,
+      "step": 5811
+    },
+    {
+      "epoch": 0.3575295275590551,
+      "grad_norm": 1.630120873451233,
+      "learning_rate": 0.0001433908628394465,
+      "loss": 3.165,
+      "step": 5812
+    },
+    {
+      "epoch": 0.35759104330708663,
+      "grad_norm": 1.0981831550598145,
+      "learning_rate": 0.00014337343965032827,
+      "loss": 2.7485,
+      "step": 5813
+    },
+    {
+      "epoch": 0.3576525590551181,
+      "grad_norm": 1.1812095642089844,
+      "learning_rate": 0.0001433560148392852,
+      "loss": 2.2237,
+      "step": 5814
+    },
+    {
+      "epoch": 0.3577140748031496,
+      "grad_norm": 1.2231512069702148,
+      "learning_rate": 0.000143338588406969,
+      "loss": 2.7293,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3577755905511811,
+      "grad_norm": 1.0244133472442627,
+      "learning_rate": 0.0001433211603540312,
+      "loss": 1.8808,
+      "step": 5816
+    },
+    {
+      "epoch": 0.3578371062992126,
+      "grad_norm": 1.1639107465744019,
+      "learning_rate": 0.0001433037306811236,
+      "loss": 2.9767,
+      "step": 5817
+    },
+    {
+      "epoch": 0.3578986220472441,
+      "grad_norm": 0.9738418459892273,
+      "learning_rate": 0.00014328629938889796,
+      "loss": 2.3424,
+      "step": 5818
+    },
+    {
+      "epoch": 0.3579601377952756,
+      "grad_norm": 1.2450603246688843,
+      "learning_rate": 0.00014326886647800607,
+      "loss": 2.7776,
+      "step": 5819
+    },
+    {
+      "epoch": 0.35802165354330706,
+      "grad_norm": 1.2114429473876953,
+      "learning_rate": 0.00014325143194909983,
+      "loss": 2.8488,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3580831692913386,
+      "grad_norm": 1.0428086519241333,
+      "learning_rate": 0.00014323399580283124,
+      "loss": 2.7444,
+      "step": 5821
+    },
+    {
+      "epoch": 0.3581446850393701,
+      "grad_norm": 1.211296558380127,
+      "learning_rate": 0.00014321655803985225,
+      "loss": 2.7173,
+      "step": 5822
+    },
+    {
+      "epoch": 0.35820620078740156,
+      "grad_norm": 1.1859794855117798,
+      "learning_rate": 0.000143199118660815,
+      "loss": 2.679,
+      "step": 5823
+    },
+    {
+      "epoch": 0.35826771653543305,
+      "grad_norm": 1.2201050519943237,
+      "learning_rate": 0.0001431816776663716,
+      "loss": 2.3983,
+      "step": 5824
+    },
+    {
+      "epoch": 0.3583292322834646,
+      "grad_norm": 1.2235811948776245,
+      "learning_rate": 0.00014316423505717423,
+      "loss": 2.5185,
+      "step": 5825
+    },
+    {
+      "epoch": 0.35839074803149606,
+      "grad_norm": 1.159736156463623,
+      "learning_rate": 0.00014314679083387517,
+      "loss": 2.4313,
+      "step": 5826
+    },
+    {
+      "epoch": 0.35845226377952755,
+      "grad_norm": 1.1394689083099365,
+      "learning_rate": 0.00014312934499712674,
+      "loss": 2.8708,
+      "step": 5827
+    },
+    {
+      "epoch": 0.35851377952755903,
+      "grad_norm": 1.0801342725753784,
+      "learning_rate": 0.00014311189754758128,
+      "loss": 2.8695,
+      "step": 5828
+    },
+    {
+      "epoch": 0.35857529527559057,
+      "grad_norm": 1.0572946071624756,
+      "learning_rate": 0.00014309444848589124,
+      "loss": 3.3605,
+      "step": 5829
+    },
+    {
+      "epoch": 0.35863681102362205,
+      "grad_norm": 1.2364927530288696,
+      "learning_rate": 0.00014307699781270915,
+      "loss": 1.9985,
+      "step": 5830
+    },
+    {
+      "epoch": 0.35869832677165353,
+      "grad_norm": 1.2879528999328613,
+      "learning_rate": 0.00014305954552868753,
+      "loss": 2.6689,
+      "step": 5831
+    },
+    {
+      "epoch": 0.358759842519685,
+      "grad_norm": 0.9771284461021423,
+      "learning_rate": 0.00014304209163447902,
+      "loss": 2.3997,
+      "step": 5832
+    },
+    {
+      "epoch": 0.35882135826771655,
+      "grad_norm": 1.334873914718628,
+      "learning_rate": 0.0001430246361307363,
+      "loss": 2.2256,
+      "step": 5833
+    },
+    {
+      "epoch": 0.35888287401574803,
+      "grad_norm": 1.4086403846740723,
+      "learning_rate": 0.00014300717901811208,
+      "loss": 3.0957,
+      "step": 5834
+    },
+    {
+      "epoch": 0.3589443897637795,
+      "grad_norm": 1.1135802268981934,
+      "learning_rate": 0.0001429897202972592,
+      "loss": 2.7439,
+      "step": 5835
+    },
+    {
+      "epoch": 0.359005905511811,
+      "grad_norm": 1.1710203886032104,
+      "learning_rate": 0.00014297225996883053,
+      "loss": 2.412,
+      "step": 5836
+    },
+    {
+      "epoch": 0.35906742125984253,
+      "grad_norm": 1.0748392343521118,
+      "learning_rate": 0.0001429547980334789,
+      "loss": 2.8803,
+      "step": 5837
+    },
+    {
+      "epoch": 0.359128937007874,
+      "grad_norm": 1.0130051374435425,
+      "learning_rate": 0.00014293733449185735,
+      "loss": 2.9084,
+      "step": 5838
+    },
+    {
+      "epoch": 0.3591904527559055,
+      "grad_norm": 1.1416534185409546,
+      "learning_rate": 0.00014291986934461895,
+      "loss": 2.4617,
+      "step": 5839
+    },
+    {
+      "epoch": 0.359251968503937,
+      "grad_norm": 1.279102087020874,
+      "learning_rate": 0.00014290240259241674,
+      "loss": 2.9069,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3593134842519685,
+      "grad_norm": 1.0762577056884766,
+      "learning_rate": 0.0001428849342359039,
+      "loss": 2.6828,
+      "step": 5841
+    },
+    {
+      "epoch": 0.359375,
+      "grad_norm": 0.9565315246582031,
+      "learning_rate": 0.0001428674642757337,
+      "loss": 3.0648,
+      "step": 5842
+    },
+    {
+      "epoch": 0.3594365157480315,
+      "grad_norm": 1.0444016456604004,
+      "learning_rate": 0.00014284999271255933,
+      "loss": 3.3168,
+      "step": 5843
+    },
+    {
+      "epoch": 0.359498031496063,
+      "grad_norm": 1.3353841304779053,
+      "learning_rate": 0.00014283251954703413,
+      "loss": 2.9409,
+      "step": 5844
+    },
+    {
+      "epoch": 0.3595595472440945,
+      "grad_norm": 1.3569176197052002,
+      "learning_rate": 0.0001428150447798116,
+      "loss": 1.8036,
+      "step": 5845
+    },
+    {
+      "epoch": 0.359621062992126,
+      "grad_norm": 1.0666240453720093,
+      "learning_rate": 0.00014279756841154514,
+      "loss": 3.2303,
+      "step": 5846
+    },
+    {
+      "epoch": 0.35968257874015747,
+      "grad_norm": 1.3822942972183228,
+      "learning_rate": 0.00014278009044288822,
+      "loss": 3.1169,
+      "step": 5847
+    },
+    {
+      "epoch": 0.359744094488189,
+      "grad_norm": 1.0390353202819824,
+      "learning_rate": 0.0001427626108744945,
+      "loss": 3.0788,
+      "step": 5848
+    },
+    {
+      "epoch": 0.3598056102362205,
+      "grad_norm": 1.0188344717025757,
+      "learning_rate": 0.00014274512970701756,
+      "loss": 2.9807,
+      "step": 5849
+    },
+    {
+      "epoch": 0.35986712598425197,
+      "grad_norm": 1.4441202878952026,
+      "learning_rate": 0.00014272764694111116,
+      "loss": 2.2241,
+      "step": 5850
+    },
+    {
+      "epoch": 0.35992864173228345,
+      "grad_norm": 0.9970209002494812,
+      "learning_rate": 0.000142710162577429,
+      "loss": 1.8043,
+      "step": 5851
+    },
+    {
+      "epoch": 0.359990157480315,
+      "grad_norm": 1.1620795726776123,
+      "learning_rate": 0.0001426926766166249,
+      "loss": 1.9918,
+      "step": 5852
+    },
+    {
+      "epoch": 0.36005167322834647,
+      "grad_norm": 1.333421230316162,
+      "learning_rate": 0.0001426751890593528,
+      "loss": 2.716,
+      "step": 5853
+    },
+    {
+      "epoch": 0.36011318897637795,
+      "grad_norm": 1.0368493795394897,
+      "learning_rate": 0.00014265769990626654,
+      "loss": 2.9656,
+      "step": 5854
+    },
+    {
+      "epoch": 0.36017470472440943,
+      "grad_norm": 1.0542171001434326,
+      "learning_rate": 0.0001426402091580202,
+      "loss": 2.5838,
+      "step": 5855
+    },
+    {
+      "epoch": 0.36023622047244097,
+      "grad_norm": 1.0582154989242554,
+      "learning_rate": 0.00014262271681526783,
+      "loss": 2.8685,
+      "step": 5856
+    },
+    {
+      "epoch": 0.36029773622047245,
+      "grad_norm": 1.06461501121521,
+      "learning_rate": 0.00014260522287866346,
+      "loss": 2.6306,
+      "step": 5857
+    },
+    {
+      "epoch": 0.36035925196850394,
+      "grad_norm": 2.794530153274536,
+      "learning_rate": 0.00014258772734886138,
+      "loss": 2.7726,
+      "step": 5858
+    },
+    {
+      "epoch": 0.3604207677165354,
+      "grad_norm": 1.1736444234848022,
+      "learning_rate": 0.00014257023022651576,
+      "loss": 2.3321,
+      "step": 5859
+    },
+    {
+      "epoch": 0.36048228346456695,
+      "grad_norm": 1.1570266485214233,
+      "learning_rate": 0.00014255273151228087,
+      "loss": 3.1429,
+      "step": 5860
+    },
+    {
+      "epoch": 0.36054379921259844,
+      "grad_norm": 1.0097638368606567,
+      "learning_rate": 0.00014253523120681116,
+      "loss": 2.7003,
+      "step": 5861
+    },
+    {
+      "epoch": 0.3606053149606299,
+      "grad_norm": 1.339082956314087,
+      "learning_rate": 0.00014251772931076094,
+      "loss": 2.6221,
+      "step": 5862
+    },
+    {
+      "epoch": 0.3606668307086614,
+      "grad_norm": 1.0804494619369507,
+      "learning_rate": 0.00014250022582478478,
+      "loss": 1.6966,
+      "step": 5863
+    },
+    {
+      "epoch": 0.36072834645669294,
+      "grad_norm": 1.0970771312713623,
+      "learning_rate": 0.0001424827207495371,
+      "loss": 2.8759,
+      "step": 5864
+    },
+    {
+      "epoch": 0.3607898622047244,
+      "grad_norm": 0.9780701994895935,
+      "learning_rate": 0.0001424652140856726,
+      "loss": 2.0681,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3608513779527559,
+      "grad_norm": 1.0544068813323975,
+      "learning_rate": 0.00014244770583384588,
+      "loss": 2.8173,
+      "step": 5866
+    },
+    {
+      "epoch": 0.3609128937007874,
+      "grad_norm": 1.4332139492034912,
+      "learning_rate": 0.00014243019599471165,
+      "loss": 2.0992,
+      "step": 5867
+    },
+    {
+      "epoch": 0.3609744094488189,
+      "grad_norm": 1.159895658493042,
+      "learning_rate": 0.00014241268456892468,
+      "loss": 2.0989,
+      "step": 5868
+    },
+    {
+      "epoch": 0.3610359251968504,
+      "grad_norm": 1.4029850959777832,
+      "learning_rate": 0.0001423951715571398,
+      "loss": 1.9654,
+      "step": 5869
+    },
+    {
+      "epoch": 0.3610974409448819,
+      "grad_norm": 1.1144707202911377,
+      "learning_rate": 0.0001423776569600119,
+      "loss": 3.076,
+      "step": 5870
+    },
+    {
+      "epoch": 0.36115895669291337,
+      "grad_norm": 1.1498141288757324,
+      "learning_rate": 0.00014236014077819597,
+      "loss": 2.9924,
+      "step": 5871
+    },
+    {
+      "epoch": 0.3612204724409449,
+      "grad_norm": 1.1732629537582397,
+      "learning_rate": 0.00014234262301234698,
+      "loss": 1.6824,
+      "step": 5872
+    },
+    {
+      "epoch": 0.3612819881889764,
+      "grad_norm": 1.0188934803009033,
+      "learning_rate": 0.00014232510366311998,
+      "loss": 2.7147,
+      "step": 5873
+    },
+    {
+      "epoch": 0.36134350393700787,
+      "grad_norm": 1.2684439420700073,
+      "learning_rate": 0.0001423075827311701,
+      "loss": 2.9004,
+      "step": 5874
+    },
+    {
+      "epoch": 0.36140501968503935,
+      "grad_norm": 1.1494557857513428,
+      "learning_rate": 0.00014229006021715258,
+      "loss": 2.8925,
+      "step": 5875
+    },
+    {
+      "epoch": 0.3614665354330709,
+      "grad_norm": 1.1155261993408203,
+      "learning_rate": 0.00014227253612172257,
+      "loss": 2.6366,
+      "step": 5876
+    },
+    {
+      "epoch": 0.3615280511811024,
+      "grad_norm": 1.2758992910385132,
+      "learning_rate": 0.00014225501044553545,
+      "loss": 2.3735,
+      "step": 5877
+    },
+    {
+      "epoch": 0.36158956692913385,
+      "grad_norm": 1.2694777250289917,
+      "learning_rate": 0.0001422374831892466,
+      "loss": 2.4585,
+      "step": 5878
+    },
+    {
+      "epoch": 0.36165108267716534,
+      "grad_norm": 1.2701408863067627,
+      "learning_rate": 0.00014221995435351135,
+      "loss": 2.4274,
+      "step": 5879
+    },
+    {
+      "epoch": 0.3617125984251969,
+      "grad_norm": 1.0925604104995728,
+      "learning_rate": 0.00014220242393898525,
+      "loss": 2.7989,
+      "step": 5880
+    },
+    {
+      "epoch": 0.36177411417322836,
+      "grad_norm": 1.1124423742294312,
+      "learning_rate": 0.00014218489194632383,
+      "loss": 2.6233,
+      "step": 5881
+    },
+    {
+      "epoch": 0.36183562992125984,
+      "grad_norm": 1.0246856212615967,
+      "learning_rate": 0.00014216735837618263,
+      "loss": 3.0042,
+      "step": 5882
+    },
+    {
+      "epoch": 0.3618971456692913,
+      "grad_norm": 1.1834280490875244,
+      "learning_rate": 0.0001421498232292174,
+      "loss": 2.7528,
+      "step": 5883
+    },
+    {
+      "epoch": 0.36195866141732286,
+      "grad_norm": 1.2525266408920288,
+      "learning_rate": 0.00014213228650608383,
+      "loss": 2.7488,
+      "step": 5884
+    },
+    {
+      "epoch": 0.36202017716535434,
+      "grad_norm": 1.106852650642395,
+      "learning_rate": 0.0001421147482074377,
+      "loss": 2.632,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3620816929133858,
+      "grad_norm": 1.0097556114196777,
+      "learning_rate": 0.00014209720833393476,
+      "loss": 2.2642,
+      "step": 5886
+    },
+    {
+      "epoch": 0.3621432086614173,
+      "grad_norm": 1.5156587362289429,
+      "learning_rate": 0.000142079666886231,
+      "loss": 2.6001,
+      "step": 5887
+    },
+    {
+      "epoch": 0.36220472440944884,
+      "grad_norm": 1.187280297279358,
+      "learning_rate": 0.00014206212386498235,
+      "loss": 2.4852,
+      "step": 5888
+    },
+    {
+      "epoch": 0.3622662401574803,
+      "grad_norm": 1.1459951400756836,
+      "learning_rate": 0.00014204457927084478,
+      "loss": 2.3268,
+      "step": 5889
+    },
+    {
+      "epoch": 0.3623277559055118,
+      "grad_norm": 1.0864328145980835,
+      "learning_rate": 0.00014202703310447443,
+      "loss": 3.1231,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3623892716535433,
+      "grad_norm": 1.0455275774002075,
+      "learning_rate": 0.0001420094853665274,
+      "loss": 2.8953,
+      "step": 5891
+    },
+    {
+      "epoch": 0.3624507874015748,
+      "grad_norm": 1.5496771335601807,
+      "learning_rate": 0.00014199193605765983,
+      "loss": 2.2239,
+      "step": 5892
+    },
+    {
+      "epoch": 0.3625123031496063,
+      "grad_norm": 1.3016999959945679,
+      "learning_rate": 0.00014197438517852802,
+      "loss": 2.6107,
+      "step": 5893
+    },
+    {
+      "epoch": 0.3625738188976378,
+      "grad_norm": 1.3840243816375732,
+      "learning_rate": 0.00014195683272978825,
+      "loss": 2.2522,
+      "step": 5894
+    },
+    {
+      "epoch": 0.36263533464566927,
+      "grad_norm": 1.2727329730987549,
+      "learning_rate": 0.00014193927871209688,
+      "loss": 2.85,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3626968503937008,
+      "grad_norm": 1.6153764724731445,
+      "learning_rate": 0.00014192172312611038,
+      "loss": 2.4385,
+      "step": 5896
+    },
+    {
+      "epoch": 0.3627583661417323,
+      "grad_norm": 1.2129825353622437,
+      "learning_rate": 0.0001419041659724852,
+      "loss": 2.5675,
+      "step": 5897
+    },
+    {
+      "epoch": 0.3628198818897638,
+      "grad_norm": 1.135545253753662,
+      "learning_rate": 0.00014188660725187785,
+      "loss": 2.5476,
+      "step": 5898
+    },
+    {
+      "epoch": 0.36288139763779526,
+      "grad_norm": 0.9694842100143433,
+      "learning_rate": 0.00014186904696494496,
+      "loss": 2.7154,
+      "step": 5899
+    },
+    {
+      "epoch": 0.3629429133858268,
+      "grad_norm": 0.942356526851654,
+      "learning_rate": 0.00014185148511234322,
+      "loss": 3.1934,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3630044291338583,
+      "grad_norm": 1.155816912651062,
+      "learning_rate": 0.00014183392169472926,
+      "loss": 3.0004,
+      "step": 5901
+    },
+    {
+      "epoch": 0.36306594488188976,
+      "grad_norm": 1.2041157484054565,
+      "learning_rate": 0.0001418163567127599,
+      "loss": 2.5607,
+      "step": 5902
+    },
+    {
+      "epoch": 0.36312746062992124,
+      "grad_norm": 1.2720422744750977,
+      "learning_rate": 0.000141798790167092,
+      "loss": 2.4337,
+      "step": 5903
+    },
+    {
+      "epoch": 0.3631889763779528,
+      "grad_norm": 1.4063915014266968,
+      "learning_rate": 0.00014178122205838243,
+      "loss": 3.2736,
+      "step": 5904
+    },
+    {
+      "epoch": 0.36325049212598426,
+      "grad_norm": 1.3971705436706543,
+      "learning_rate": 0.00014176365238728812,
+      "loss": 1.8516,
+      "step": 5905
+    },
+    {
+      "epoch": 0.36331200787401574,
+      "grad_norm": 1.1394338607788086,
+      "learning_rate": 0.00014174608115446606,
+      "loss": 2.8719,
+      "step": 5906
+    },
+    {
+      "epoch": 0.3633735236220472,
+      "grad_norm": 1.3593043088912964,
+      "learning_rate": 0.0001417285083605734,
+      "loss": 2.5227,
+      "step": 5907
+    },
+    {
+      "epoch": 0.36343503937007876,
+      "grad_norm": 1.1026146411895752,
+      "learning_rate": 0.00014171093400626715,
+      "loss": 2.4206,
+      "step": 5908
+    },
+    {
+      "epoch": 0.36349655511811024,
+      "grad_norm": 1.2278774976730347,
+      "learning_rate": 0.00014169335809220462,
+      "loss": 2.8063,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3635580708661417,
+      "grad_norm": 1.3933086395263672,
+      "learning_rate": 0.000141675780619043,
+      "loss": 2.6782,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3636195866141732,
+      "grad_norm": 1.0827549695968628,
+      "learning_rate": 0.0001416582015874395,
+      "loss": 1.3603,
+      "step": 5911
+    },
+    {
+      "epoch": 0.36368110236220474,
+      "grad_norm": 1.1697765588760376,
+      "learning_rate": 0.0001416406209980516,
+      "loss": 2.8211,
+      "step": 5912
+    },
+    {
+      "epoch": 0.3637426181102362,
+      "grad_norm": 1.2548707723617554,
+      "learning_rate": 0.00014162303885153666,
+      "loss": 2.5391,
+      "step": 5913
+    },
+    {
+      "epoch": 0.3638041338582677,
+      "grad_norm": 1.1490145921707153,
+      "learning_rate": 0.00014160545514855215,
+      "loss": 3.0663,
+      "step": 5914
+    },
+    {
+      "epoch": 0.3638656496062992,
+      "grad_norm": 1.2477426528930664,
+      "learning_rate": 0.00014158786988975564,
+      "loss": 3.0192,
+      "step": 5915
+    },
+    {
+      "epoch": 0.36392716535433073,
+      "grad_norm": 2.259104013442993,
+      "learning_rate": 0.0001415702830758047,
+      "loss": 2.5954,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3639886811023622,
+      "grad_norm": 1.2731071710586548,
+      "learning_rate": 0.00014155269470735694,
+      "loss": 2.9515,
+      "step": 5917
+    },
+    {
+      "epoch": 0.3640501968503937,
+      "grad_norm": 1.08100426197052,
+      "learning_rate": 0.0001415351047850701,
+      "loss": 2.7069,
+      "step": 5918
+    },
+    {
+      "epoch": 0.3641117125984252,
+      "grad_norm": 1.1245132684707642,
+      "learning_rate": 0.00014151751330960196,
+      "loss": 2.4293,
+      "step": 5919
+    },
+    {
+      "epoch": 0.3641732283464567,
+      "grad_norm": 1.2853879928588867,
+      "learning_rate": 0.00014149992028161037,
+      "loss": 2.6944,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3642347440944882,
+      "grad_norm": 1.196007251739502,
+      "learning_rate": 0.0001414823257017531,
+      "loss": 2.713,
+      "step": 5921
+    },
+    {
+      "epoch": 0.3642962598425197,
+      "grad_norm": 1.0759581327438354,
+      "learning_rate": 0.0001414647295706882,
+      "loss": 2.8866,
+      "step": 5922
+    },
+    {
+      "epoch": 0.36435777559055116,
+      "grad_norm": 1.2459166049957275,
+      "learning_rate": 0.00014144713188907367,
+      "loss": 2.7631,
+      "step": 5923
+    },
+    {
+      "epoch": 0.3644192913385827,
+      "grad_norm": 1.0685508251190186,
+      "learning_rate": 0.00014142953265756743,
+      "loss": 2.6548,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3644808070866142,
+      "grad_norm": 1.1906906366348267,
+      "learning_rate": 0.00014141193187682774,
+      "loss": 2.5878,
+      "step": 5925
+    },
+    {
+      "epoch": 0.36454232283464566,
+      "grad_norm": 1.3111698627471924,
+      "learning_rate": 0.00014139432954751271,
+      "loss": 2.3309,
+      "step": 5926
+    },
+    {
+      "epoch": 0.36460383858267714,
+      "grad_norm": 1.165251612663269,
+      "learning_rate": 0.0001413767256702806,
+      "loss": 2.5037,
+      "step": 5927
+    },
+    {
+      "epoch": 0.3646653543307087,
+      "grad_norm": 1.0471251010894775,
+      "learning_rate": 0.00014135912024578964,
+      "loss": 2.8164,
+      "step": 5928
+    },
+    {
+      "epoch": 0.36472687007874016,
+      "grad_norm": 1.1354172229766846,
+      "learning_rate": 0.00014134151327469826,
+      "loss": 2.9538,
+      "step": 5929
+    },
+    {
+      "epoch": 0.36478838582677164,
+      "grad_norm": 1.3205845355987549,
+      "learning_rate": 0.00014132390475766476,
+      "loss": 2.6602,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3648499015748031,
+      "grad_norm": 0.9886876344680786,
+      "learning_rate": 0.00014130629469534764,
+      "loss": 2.674,
+      "step": 5931
+    },
+    {
+      "epoch": 0.36491141732283466,
+      "grad_norm": 1.059760570526123,
+      "learning_rate": 0.00014128868308840547,
+      "loss": 2.8664,
+      "step": 5932
+    },
+    {
+      "epoch": 0.36497293307086615,
+      "grad_norm": 1.1438698768615723,
+      "learning_rate": 0.00014127106993749674,
+      "loss": 2.842,
+      "step": 5933
+    },
+    {
+      "epoch": 0.3650344488188976,
+      "grad_norm": 1.4471333026885986,
+      "learning_rate": 0.00014125345524328015,
+      "loss": 2.5676,
+      "step": 5934
+    },
+    {
+      "epoch": 0.3650959645669291,
+      "grad_norm": 1.2641026973724365,
+      "learning_rate": 0.00014123583900641436,
+      "loss": 2.7982,
+      "step": 5935
+    },
+    {
+      "epoch": 0.36515748031496065,
+      "grad_norm": 1.2789355516433716,
+      "learning_rate": 0.00014121822122755817,
+      "loss": 2.9141,
+      "step": 5936
+    },
+    {
+      "epoch": 0.36521899606299213,
+      "grad_norm": 1.2175452709197998,
+      "learning_rate": 0.0001412006019073703,
+      "loss": 3.0002,
+      "step": 5937
+    },
+    {
+      "epoch": 0.3652805118110236,
+      "grad_norm": 1.2370004653930664,
+      "learning_rate": 0.00014118298104650966,
+      "loss": 1.9273,
+      "step": 5938
+    },
+    {
+      "epoch": 0.3653420275590551,
+      "grad_norm": 1.237525224685669,
+      "learning_rate": 0.0001411653586456352,
+      "loss": 2.2333,
+      "step": 5939
+    },
+    {
+      "epoch": 0.36540354330708663,
+      "grad_norm": 0.9971503019332886,
+      "learning_rate": 0.00014114773470540586,
+      "loss": 2.7406,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3654650590551181,
+      "grad_norm": 1.159501314163208,
+      "learning_rate": 0.00014113010922648067,
+      "loss": 2.7457,
+      "step": 5941
+    },
+    {
+      "epoch": 0.3655265748031496,
+      "grad_norm": 1.3211581707000732,
+      "learning_rate": 0.00014111248220951875,
+      "loss": 2.4702,
+      "step": 5942
+    },
+    {
+      "epoch": 0.3655880905511811,
+      "grad_norm": 1.2520309686660767,
+      "learning_rate": 0.00014109485365517926,
+      "loss": 2.6702,
+      "step": 5943
+    },
+    {
+      "epoch": 0.3656496062992126,
+      "grad_norm": 1.0072524547576904,
+      "learning_rate": 0.00014107722356412138,
+      "loss": 2.6172,
+      "step": 5944
+    },
+    {
+      "epoch": 0.3657111220472441,
+      "grad_norm": 1.3465410470962524,
+      "learning_rate": 0.0001410595919370044,
+      "loss": 2.8308,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3657726377952756,
+      "grad_norm": 1.021047830581665,
+      "learning_rate": 0.00014104195877448765,
+      "loss": 2.72,
+      "step": 5946
+    },
+    {
+      "epoch": 0.36583415354330706,
+      "grad_norm": 1.1119564771652222,
+      "learning_rate": 0.0001410243240772305,
+      "loss": 2.1493,
+      "step": 5947
+    },
+    {
+      "epoch": 0.3658956692913386,
+      "grad_norm": 1.1941128969192505,
+      "learning_rate": 0.00014100668784589238,
+      "loss": 3.6121,
+      "step": 5948
+    },
+    {
+      "epoch": 0.3659571850393701,
+      "grad_norm": 1.1287257671356201,
+      "learning_rate": 0.00014098905008113274,
+      "loss": 2.4209,
+      "step": 5949
+    },
+    {
+      "epoch": 0.36601870078740156,
+      "grad_norm": 1.0580445528030396,
+      "learning_rate": 0.00014097141078361124,
+      "loss": 3.2223,
+      "step": 5950
+    },
+    {
+      "epoch": 0.36608021653543305,
+      "grad_norm": 1.353436827659607,
+      "learning_rate": 0.00014095376995398743,
+      "loss": 2.7075,
+      "step": 5951
+    },
+    {
+      "epoch": 0.3661417322834646,
+      "grad_norm": 1.2519347667694092,
+      "learning_rate": 0.00014093612759292102,
+      "loss": 3.0163,
+      "step": 5952
+    },
+    {
+      "epoch": 0.36620324803149606,
+      "grad_norm": 1.0582761764526367,
+      "learning_rate": 0.00014091848370107165,
+      "loss": 2.6954,
+      "step": 5953
+    },
+    {
+      "epoch": 0.36626476377952755,
+      "grad_norm": 1.0018788576126099,
+      "learning_rate": 0.0001409008382790992,
+      "loss": 2.9129,
+      "step": 5954
+    },
+    {
+      "epoch": 0.36632627952755903,
+      "grad_norm": 1.1188297271728516,
+      "learning_rate": 0.00014088319132766344,
+      "loss": 2.7389,
+      "step": 5955
+    },
+    {
+      "epoch": 0.36638779527559057,
+      "grad_norm": 1.0715093612670898,
+      "learning_rate": 0.0001408655428474243,
+      "loss": 2.6597,
+      "step": 5956
+    },
+    {
+      "epoch": 0.36644931102362205,
+      "grad_norm": 1.267085075378418,
+      "learning_rate": 0.00014084789283904173,
+      "loss": 1.6784,
+      "step": 5957
+    },
+    {
+      "epoch": 0.36651082677165353,
+      "grad_norm": 1.0898419618606567,
+      "learning_rate": 0.00014083024130317576,
+      "loss": 1.8412,
+      "step": 5958
+    },
+    {
+      "epoch": 0.366572342519685,
+      "grad_norm": 1.2054049968719482,
+      "learning_rate": 0.0001408125882404864,
+      "loss": 2.4899,
+      "step": 5959
+    },
+    {
+      "epoch": 0.36663385826771655,
+      "grad_norm": 1.3260561227798462,
+      "learning_rate": 0.00014079493365163386,
+      "loss": 2.6179,
+      "step": 5960
+    },
+    {
+      "epoch": 0.36669537401574803,
+      "grad_norm": 1.3705940246582031,
+      "learning_rate": 0.00014077727753727829,
+      "loss": 2.5916,
+      "step": 5961
+    },
+    {
+      "epoch": 0.3667568897637795,
+      "grad_norm": 1.2106672525405884,
+      "learning_rate": 0.00014075961989807987,
+      "loss": 2.2216,
+      "step": 5962
+    },
+    {
+      "epoch": 0.366818405511811,
+      "grad_norm": 1.4452307224273682,
+      "learning_rate": 0.00014074196073469896,
+      "loss": 2.6995,
+      "step": 5963
+    },
+    {
+      "epoch": 0.36687992125984253,
+      "grad_norm": 1.4027457237243652,
+      "learning_rate": 0.00014072430004779594,
+      "loss": 3.2467,
+      "step": 5964
+    },
+    {
+      "epoch": 0.366941437007874,
+      "grad_norm": 1.733917474746704,
+      "learning_rate": 0.00014070663783803116,
+      "loss": 2.95,
+      "step": 5965
+    },
+    {
+      "epoch": 0.3670029527559055,
+      "grad_norm": 1.0921287536621094,
+      "learning_rate": 0.00014068897410606508,
+      "loss": 2.7101,
+      "step": 5966
+    },
+    {
+      "epoch": 0.367064468503937,
+      "grad_norm": 1.019783854484558,
+      "learning_rate": 0.0001406713088525583,
+      "loss": 2.8358,
+      "step": 5967
+    },
+    {
+      "epoch": 0.3671259842519685,
+      "grad_norm": 1.1198570728302002,
+      "learning_rate": 0.00014065364207817136,
+      "loss": 2.636,
+      "step": 5968
+    },
+    {
+      "epoch": 0.3671875,
+      "grad_norm": 1.3331079483032227,
+      "learning_rate": 0.00014063597378356486,
+      "loss": 2.4018,
+      "step": 5969
+    },
+    {
+      "epoch": 0.3672490157480315,
+      "grad_norm": 1.1102997064590454,
+      "learning_rate": 0.00014061830396939955,
+      "loss": 2.4707,
+      "step": 5970
+    },
+    {
+      "epoch": 0.367310531496063,
+      "grad_norm": 1.1189804077148438,
+      "learning_rate": 0.00014060063263633617,
+      "loss": 2.9256,
+      "step": 5971
+    },
+    {
+      "epoch": 0.3673720472440945,
+      "grad_norm": 1.1461769342422485,
+      "learning_rate": 0.00014058295978503554,
+      "loss": 2.7276,
+      "step": 5972
+    },
+    {
+      "epoch": 0.367433562992126,
+      "grad_norm": 1.0559407472610474,
+      "learning_rate": 0.00014056528541615847,
+      "loss": 2.3275,
+      "step": 5973
+    },
+    {
+      "epoch": 0.36749507874015747,
+      "grad_norm": 1.2454173564910889,
+      "learning_rate": 0.00014054760953036597,
+      "loss": 1.7245,
+      "step": 5974
+    },
+    {
+      "epoch": 0.367556594488189,
+      "grad_norm": 1.198763132095337,
+      "learning_rate": 0.00014052993212831894,
+      "loss": 3.0505,
+      "step": 5975
+    },
+    {
+      "epoch": 0.3676181102362205,
+      "grad_norm": 1.311190128326416,
+      "learning_rate": 0.00014051225321067844,
+      "loss": 3.2148,
+      "step": 5976
+    },
+    {
+      "epoch": 0.36767962598425197,
+      "grad_norm": 1.221556305885315,
+      "learning_rate": 0.0001404945727781056,
+      "loss": 2.4534,
+      "step": 5977
+    },
+    {
+      "epoch": 0.36774114173228345,
+      "grad_norm": 1.1493868827819824,
+      "learning_rate": 0.00014047689083126157,
+      "loss": 2.8196,
+      "step": 5978
+    },
+    {
+      "epoch": 0.367802657480315,
+      "grad_norm": 1.2901254892349243,
+      "learning_rate": 0.00014045920737080746,
+      "loss": 2.6287,
+      "step": 5979
+    },
+    {
+      "epoch": 0.36786417322834647,
+      "grad_norm": 1.1601101160049438,
+      "learning_rate": 0.00014044152239740465,
+      "loss": 2.1861,
+      "step": 5980
+    },
+    {
+      "epoch": 0.36792568897637795,
+      "grad_norm": 0.9924845695495605,
+      "learning_rate": 0.0001404238359117144,
+      "loss": 2.0936,
+      "step": 5981
+    },
+    {
+      "epoch": 0.36798720472440943,
+      "grad_norm": 1.334097146987915,
+      "learning_rate": 0.00014040614791439808,
+      "loss": 2.5699,
+      "step": 5982
+    },
+    {
+      "epoch": 0.36804872047244097,
+      "grad_norm": 1.164738416671753,
+      "learning_rate": 0.00014038845840611715,
+      "loss": 2.4961,
+      "step": 5983
+    },
+    {
+      "epoch": 0.36811023622047245,
+      "grad_norm": 1.076383113861084,
+      "learning_rate": 0.0001403707673875331,
+      "loss": 2.4244,
+      "step": 5984
+    },
+    {
+      "epoch": 0.36817175196850394,
+      "grad_norm": 1.2030115127563477,
+      "learning_rate": 0.00014035307485930746,
+      "loss": 3.1177,
+      "step": 5985
+    },
+    {
+      "epoch": 0.3682332677165354,
+      "grad_norm": 1.0214416980743408,
+      "learning_rate": 0.00014033538082210183,
+      "loss": 3.3569,
+      "step": 5986
+    },
+    {
+      "epoch": 0.36829478346456695,
+      "grad_norm": 0.9877448678016663,
+      "learning_rate": 0.00014031768527657788,
+      "loss": 2.9409,
+      "step": 5987
+    },
+    {
+      "epoch": 0.36835629921259844,
+      "grad_norm": 1.3242155313491821,
+      "learning_rate": 0.0001402999882233973,
+      "loss": 2.4339,
+      "step": 5988
+    },
+    {
+      "epoch": 0.3684178149606299,
+      "grad_norm": 1.1602596044540405,
+      "learning_rate": 0.00014028228966322189,
+      "loss": 3.0379,
+      "step": 5989
+    },
+    {
+      "epoch": 0.3684793307086614,
+      "grad_norm": 1.1425986289978027,
+      "learning_rate": 0.00014026458959671344,
+      "loss": 3.2266,
+      "step": 5990
+    },
+    {
+      "epoch": 0.36854084645669294,
+      "grad_norm": 1.0362073183059692,
+      "learning_rate": 0.00014024688802453388,
+      "loss": 2.6411,
+      "step": 5991
+    },
+    {
+      "epoch": 0.3686023622047244,
+      "grad_norm": 1.0232431888580322,
+      "learning_rate": 0.00014022918494734516,
+      "loss": 2.7655,
+      "step": 5992
+    },
+    {
+      "epoch": 0.3686638779527559,
+      "grad_norm": 1.203115701675415,
+      "learning_rate": 0.00014021148036580923,
+      "loss": 2.4964,
+      "step": 5993
+    },
+    {
+      "epoch": 0.3687253937007874,
+      "grad_norm": 1.1594287157058716,
+      "learning_rate": 0.00014019377428058812,
+      "loss": 2.9553,
+      "step": 5994
+    },
+    {
+      "epoch": 0.3687869094488189,
+      "grad_norm": 1.1111888885498047,
+      "learning_rate": 0.00014017606669234402,
+      "loss": 2.7162,
+      "step": 5995
+    },
+    {
+      "epoch": 0.3688484251968504,
+      "grad_norm": 1.2181708812713623,
+      "learning_rate": 0.00014015835760173906,
+      "loss": 3.354,
+      "step": 5996
+    },
+    {
+      "epoch": 0.3689099409448819,
+      "grad_norm": 1.3448917865753174,
+      "learning_rate": 0.00014014064700943545,
+      "loss": 2.5215,
+      "step": 5997
+    },
+    {
+      "epoch": 0.36897145669291337,
+      "grad_norm": 1.1280043125152588,
+      "learning_rate": 0.00014012293491609543,
+      "loss": 2.8198,
+      "step": 5998
+    },
+    {
+      "epoch": 0.3690329724409449,
+      "grad_norm": 1.1832728385925293,
+      "learning_rate": 0.00014010522132238142,
+      "loss": 2.6785,
+      "step": 5999
+    },
+    {
+      "epoch": 0.3690944881889764,
+      "grad_norm": 1.130179524421692,
+      "learning_rate": 0.00014008750622895576,
+      "loss": 2.6177,
+      "step": 6000
+    },
+    {
+      "epoch": 0.36915600393700787,
+      "grad_norm": 1.2036106586456299,
+      "learning_rate": 0.00014006978963648087,
+      "loss": 2.2098,
+      "step": 6001
+    },
+    {
+      "epoch": 0.36921751968503935,
+      "grad_norm": 1.2131121158599854,
+      "learning_rate": 0.00014005207154561933,
+      "loss": 2.8047,
+      "step": 6002
+    },
+    {
+      "epoch": 0.3692790354330709,
+      "grad_norm": 1.1777770519256592,
+      "learning_rate": 0.00014003435195703362,
+      "loss": 2.0805,
+      "step": 6003
+    },
+    {
+      "epoch": 0.3693405511811024,
+      "grad_norm": 1.2286607027053833,
+      "learning_rate": 0.00014001663087138638,
+      "loss": 2.7115,
+      "step": 6004
+    },
+    {
+      "epoch": 0.36940206692913385,
+      "grad_norm": 1.0615347623825073,
+      "learning_rate": 0.0001399989082893403,
+      "loss": 3.0492,
+      "step": 6005
+    },
+    {
+      "epoch": 0.36946358267716534,
+      "grad_norm": 1.3400251865386963,
+      "learning_rate": 0.0001399811842115581,
+      "loss": 2.4175,
+      "step": 6006
+    },
+    {
+      "epoch": 0.3695250984251969,
+      "grad_norm": 1.1637264490127563,
+      "learning_rate": 0.00013996345863870254,
+      "loss": 2.2295,
+      "step": 6007
+    },
+    {
+      "epoch": 0.36958661417322836,
+      "grad_norm": 2.1271605491638184,
+      "learning_rate": 0.00013994573157143645,
+      "loss": 2.699,
+      "step": 6008
+    },
+    {
+      "epoch": 0.36964812992125984,
+      "grad_norm": 1.0444923639297485,
+      "learning_rate": 0.00013992800301042277,
+      "loss": 1.7358,
+      "step": 6009
+    },
+    {
+      "epoch": 0.3697096456692913,
+      "grad_norm": 1.2003716230392456,
+      "learning_rate": 0.0001399102729563244,
+      "loss": 2.4702,
+      "step": 6010
+    },
+    {
+      "epoch": 0.36977116141732286,
+      "grad_norm": 1.066720724105835,
+      "learning_rate": 0.00013989254140980438,
+      "loss": 2.8061,
+      "step": 6011
+    },
+    {
+      "epoch": 0.36983267716535434,
+      "grad_norm": 1.129258394241333,
+      "learning_rate": 0.0001398748083715258,
+      "loss": 3.0757,
+      "step": 6012
+    },
+    {
+      "epoch": 0.3698941929133858,
+      "grad_norm": 1.4228111505508423,
+      "learning_rate": 0.00013985707384215167,
+      "loss": 2.3887,
+      "step": 6013
+    },
+    {
+      "epoch": 0.3699557086614173,
+      "grad_norm": 1.3455454111099243,
+      "learning_rate": 0.00013983933782234524,
+      "loss": 2.5025,
+      "step": 6014
+    },
+    {
+      "epoch": 0.37001722440944884,
+      "grad_norm": 1.2470759153366089,
+      "learning_rate": 0.00013982160031276975,
+      "loss": 2.6222,
+      "step": 6015
+    },
+    {
+      "epoch": 0.3700787401574803,
+      "grad_norm": 1.0869828462600708,
+      "learning_rate": 0.00013980386131408844,
+      "loss": 1.4779,
+      "step": 6016
+    },
+    {
+      "epoch": 0.3701402559055118,
+      "grad_norm": 1.008874535560608,
+      "learning_rate": 0.00013978612082696465,
+      "loss": 3.3044,
+      "step": 6017
+    },
+    {
+      "epoch": 0.3702017716535433,
+      "grad_norm": 1.156722903251648,
+      "learning_rate": 0.00013976837885206183,
+      "loss": 2.8565,
+      "step": 6018
+    },
+    {
+      "epoch": 0.3702632874015748,
+      "grad_norm": 1.5113179683685303,
+      "learning_rate": 0.00013975063539004338,
+      "loss": 2.5604,
+      "step": 6019
+    },
+    {
+      "epoch": 0.3703248031496063,
+      "grad_norm": 1.1064906120300293,
+      "learning_rate": 0.00013973289044157285,
+      "loss": 2.7055,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3703863188976378,
+      "grad_norm": 1.2013283967971802,
+      "learning_rate": 0.00013971514400731375,
+      "loss": 2.8203,
+      "step": 6021
+    },
+    {
+      "epoch": 0.37044783464566927,
+      "grad_norm": 1.311406135559082,
+      "learning_rate": 0.00013969739608792972,
+      "loss": 2.3498,
+      "step": 6022
+    },
+    {
+      "epoch": 0.3705093503937008,
+      "grad_norm": 1.1088025569915771,
+      "learning_rate": 0.00013967964668408443,
+      "loss": 2.8928,
+      "step": 6023
+    },
+    {
+      "epoch": 0.3705708661417323,
+      "grad_norm": 1.119327187538147,
+      "learning_rate": 0.00013966189579644163,
+      "loss": 2.6884,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3706323818897638,
+      "grad_norm": 1.1854051351547241,
+      "learning_rate": 0.0001396441434256651,
+      "loss": 2.9001,
+      "step": 6025
+    },
+    {
+      "epoch": 0.37069389763779526,
+      "grad_norm": 1.054888129234314,
+      "learning_rate": 0.00013962638957241863,
+      "loss": 2.7689,
+      "step": 6026
+    },
+    {
+      "epoch": 0.3707554133858268,
+      "grad_norm": 1.1727865934371948,
+      "learning_rate": 0.00013960863423736615,
+      "loss": 1.9127,
+      "step": 6027
+    },
+    {
+      "epoch": 0.3708169291338583,
+      "grad_norm": 1.1059250831604004,
+      "learning_rate": 0.00013959087742117165,
+      "loss": 2.8972,
+      "step": 6028
+    },
+    {
+      "epoch": 0.37087844488188976,
+      "grad_norm": 1.1120609045028687,
+      "learning_rate": 0.00013957311912449907,
+      "loss": 2.842,
+      "step": 6029
+    },
+    {
+      "epoch": 0.37093996062992124,
+      "grad_norm": 1.472446322441101,
+      "learning_rate": 0.0001395553593480125,
+      "loss": 3.1315,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3710014763779528,
+      "grad_norm": 0.9761093258857727,
+      "learning_rate": 0.0001395375980923761,
+      "loss": 2.8237,
+      "step": 6031
+    },
+    {
+      "epoch": 0.37106299212598426,
+      "grad_norm": 1.0463820695877075,
+      "learning_rate": 0.00013951983535825395,
+      "loss": 2.6912,
+      "step": 6032
+    },
+    {
+      "epoch": 0.37112450787401574,
+      "grad_norm": 1.315595030784607,
+      "learning_rate": 0.00013950207114631037,
+      "loss": 2.6079,
+      "step": 6033
+    },
+    {
+      "epoch": 0.3711860236220472,
+      "grad_norm": 1.205846905708313,
+      "learning_rate": 0.00013948430545720957,
+      "loss": 2.1876,
+      "step": 6034
+    },
+    {
+      "epoch": 0.37124753937007876,
+      "grad_norm": 1.1012307405471802,
+      "learning_rate": 0.0001394665382916159,
+      "loss": 2.3868,
+      "step": 6035
+    },
+    {
+      "epoch": 0.37130905511811024,
+      "grad_norm": 1.2094402313232422,
+      "learning_rate": 0.0001394487696501938,
+      "loss": 2.8191,
+      "step": 6036
+    },
+    {
+      "epoch": 0.3713705708661417,
+      "grad_norm": 1.2197163105010986,
+      "learning_rate": 0.00013943099953360767,
+      "loss": 2.7668,
+      "step": 6037
+    },
+    {
+      "epoch": 0.3714320866141732,
+      "grad_norm": 1.1748559474945068,
+      "learning_rate": 0.00013941322794252205,
+      "loss": 2.6843,
+      "step": 6038
+    },
+    {
+      "epoch": 0.37149360236220474,
+      "grad_norm": 1.1626619100570679,
+      "learning_rate": 0.00013939545487760144,
+      "loss": 3.0233,
+      "step": 6039
+    },
+    {
+      "epoch": 0.3715551181102362,
+      "grad_norm": 1.0607874393463135,
+      "learning_rate": 0.00013937768033951052,
+      "loss": 2.9796,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3716166338582677,
+      "grad_norm": 1.122658133506775,
+      "learning_rate": 0.00013935990432891395,
+      "loss": 2.4745,
+      "step": 6041
+    },
+    {
+      "epoch": 0.3716781496062992,
+      "grad_norm": 1.178472876548767,
+      "learning_rate": 0.00013934212684647642,
+      "loss": 2.25,
+      "step": 6042
+    },
+    {
+      "epoch": 0.37173966535433073,
+      "grad_norm": 1.1995198726654053,
+      "learning_rate": 0.0001393243478928627,
+      "loss": 2.3478,
+      "step": 6043
+    },
+    {
+      "epoch": 0.3718011811023622,
+      "grad_norm": 1.115504503250122,
+      "learning_rate": 0.0001393065674687377,
+      "loss": 3.2795,
+      "step": 6044
+    },
+    {
+      "epoch": 0.3718626968503937,
+      "grad_norm": 1.0471184253692627,
+      "learning_rate": 0.00013928878557476618,
+      "loss": 2.7803,
+      "step": 6045
+    },
+    {
+      "epoch": 0.3719242125984252,
+      "grad_norm": 1.1437947750091553,
+      "learning_rate": 0.00013927100221161322,
+      "loss": 2.5927,
+      "step": 6046
+    },
+    {
+      "epoch": 0.3719857283464567,
+      "grad_norm": 1.1769379377365112,
+      "learning_rate": 0.00013925321737994373,
+      "loss": 2.2625,
+      "step": 6047
+    },
+    {
+      "epoch": 0.3720472440944882,
+      "grad_norm": 1.5988656282424927,
+      "learning_rate": 0.00013923543108042283,
+      "loss": 2.4979,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3721087598425197,
+      "grad_norm": 1.4669188261032104,
+      "learning_rate": 0.00013921764331371555,
+      "loss": 2.0789,
+      "step": 6049
+    },
+    {
+      "epoch": 0.37217027559055116,
+      "grad_norm": 1.176474928855896,
+      "learning_rate": 0.0001391998540804871,
+      "loss": 2.1536,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3722317913385827,
+      "grad_norm": 1.137064814567566,
+      "learning_rate": 0.00013918206338140273,
+      "loss": 2.2896,
+      "step": 6051
+    },
+    {
+      "epoch": 0.3722933070866142,
+      "grad_norm": 1.0377345085144043,
+      "learning_rate": 0.00013916427121712764,
+      "loss": 3.2089,
+      "step": 6052
+    },
+    {
+      "epoch": 0.37235482283464566,
+      "grad_norm": 1.4989086389541626,
+      "learning_rate": 0.00013914647758832716,
+      "loss": 2.7028,
+      "step": 6053
+    },
+    {
+      "epoch": 0.37241633858267714,
+      "grad_norm": 0.9463605284690857,
+      "learning_rate": 0.00013912868249566678,
+      "loss": 2.9721,
+      "step": 6054
+    },
+    {
+      "epoch": 0.3724778543307087,
+      "grad_norm": 1.1945708990097046,
+      "learning_rate": 0.0001391108859398118,
+      "loss": 2.4389,
+      "step": 6055
+    },
+    {
+      "epoch": 0.37253937007874016,
+      "grad_norm": 1.429040789604187,
+      "learning_rate": 0.00013909308792142778,
+      "loss": 2.9564,
+      "step": 6056
+    },
+    {
+      "epoch": 0.37260088582677164,
+      "grad_norm": 1.1187944412231445,
+      "learning_rate": 0.00013907528844118032,
+      "loss": 2.4624,
+      "step": 6057
+    },
+    {
+      "epoch": 0.3726624015748031,
+      "grad_norm": 1.0699119567871094,
+      "learning_rate": 0.0001390574874997349,
+      "loss": 2.1178,
+      "step": 6058
+    },
+    {
+      "epoch": 0.37272391732283466,
+      "grad_norm": 1.3366745710372925,
+      "learning_rate": 0.00013903968509775726,
+      "loss": 2.5577,
+      "step": 6059
+    },
+    {
+      "epoch": 0.37278543307086615,
+      "grad_norm": 1.1140227317810059,
+      "learning_rate": 0.00013902188123591312,
+      "loss": 2.6699,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3728469488188976,
+      "grad_norm": 1.2734464406967163,
+      "learning_rate": 0.00013900407591486814,
+      "loss": 2.94,
+      "step": 6061
+    },
+    {
+      "epoch": 0.3729084645669291,
+      "grad_norm": 1.3211153745651245,
+      "learning_rate": 0.00013898626913528826,
+      "loss": 2.7012,
+      "step": 6062
+    },
+    {
+      "epoch": 0.37296998031496065,
+      "grad_norm": 1.0855523347854614,
+      "learning_rate": 0.00013896846089783932,
+      "loss": 2.6576,
+      "step": 6063
+    },
+    {
+      "epoch": 0.37303149606299213,
+      "grad_norm": 1.1625336408615112,
+      "learning_rate": 0.0001389506512031872,
+      "loss": 2.5043,
+      "step": 6064
+    },
+    {
+      "epoch": 0.3730930118110236,
+      "grad_norm": 1.2593998908996582,
+      "learning_rate": 0.00013893284005199794,
+      "loss": 2.621,
+      "step": 6065
+    },
+    {
+      "epoch": 0.3731545275590551,
+      "grad_norm": 1.0291095972061157,
+      "learning_rate": 0.00013891502744493755,
+      "loss": 2.3295,
+      "step": 6066
+    },
+    {
+      "epoch": 0.37321604330708663,
+      "grad_norm": 1.1604384183883667,
+      "learning_rate": 0.00013889721338267218,
+      "loss": 3.5285,
+      "step": 6067
+    },
+    {
+      "epoch": 0.3732775590551181,
+      "grad_norm": 1.1358870267868042,
+      "learning_rate": 0.00013887939786586787,
+      "loss": 3.577,
+      "step": 6068
+    },
+    {
+      "epoch": 0.3733390748031496,
+      "grad_norm": 1.7941579818725586,
+      "learning_rate": 0.0001388615808951909,
+      "loss": 2.9958,
+      "step": 6069
+    },
+    {
+      "epoch": 0.3734005905511811,
+      "grad_norm": 0.9642488956451416,
+      "learning_rate": 0.00013884376247130749,
+      "loss": 3.1242,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3734621062992126,
+      "grad_norm": 1.3350937366485596,
+      "learning_rate": 0.00013882594259488393,
+      "loss": 2.5797,
+      "step": 6071
+    },
+    {
+      "epoch": 0.3735236220472441,
+      "grad_norm": 1.4719228744506836,
+      "learning_rate": 0.00013880812126658665,
+      "loss": 1.9023,
+      "step": 6072
+    },
+    {
+      "epoch": 0.3735851377952756,
+      "grad_norm": 1.0268434286117554,
+      "learning_rate": 0.00013879029848708205,
+      "loss": 3.2002,
+      "step": 6073
+    },
+    {
+      "epoch": 0.37364665354330706,
+      "grad_norm": 1.0493897199630737,
+      "learning_rate": 0.00013877247425703656,
+      "loss": 2.5132,
+      "step": 6074
+    },
+    {
+      "epoch": 0.3737081692913386,
+      "grad_norm": 1.0271064043045044,
+      "learning_rate": 0.00013875464857711676,
+      "loss": 2.5378,
+      "step": 6075
+    },
+    {
+      "epoch": 0.3737696850393701,
+      "grad_norm": 1.381683349609375,
+      "learning_rate": 0.00013873682144798917,
+      "loss": 2.3806,
+      "step": 6076
+    },
+    {
+      "epoch": 0.37383120078740156,
+      "grad_norm": 1.206984281539917,
+      "learning_rate": 0.0001387189928703205,
+      "loss": 2.8804,
+      "step": 6077
+    },
+    {
+      "epoch": 0.37389271653543305,
+      "grad_norm": 1.1912376880645752,
+      "learning_rate": 0.00013870116284477738,
+      "loss": 2.5428,
+      "step": 6078
+    },
+    {
+      "epoch": 0.3739542322834646,
+      "grad_norm": 1.0777443647384644,
+      "learning_rate": 0.00013868333137202657,
+      "loss": 2.619,
+      "step": 6079
+    },
+    {
+      "epoch": 0.37401574803149606,
+      "grad_norm": 1.1592036485671997,
+      "learning_rate": 0.0001386654984527349,
+      "loss": 2.6095,
+      "step": 6080
+    },
+    {
+      "epoch": 0.37407726377952755,
+      "grad_norm": 1.2353136539459229,
+      "learning_rate": 0.00013864766408756913,
+      "loss": 2.8066,
+      "step": 6081
+    },
+    {
+      "epoch": 0.37413877952755903,
+      "grad_norm": 1.0832616090774536,
+      "learning_rate": 0.0001386298282771963,
+      "loss": 3.3146,
+      "step": 6082
+    },
+    {
+      "epoch": 0.37420029527559057,
+      "grad_norm": 1.0680965185165405,
+      "learning_rate": 0.00013861199102228324,
+      "loss": 2.7769,
+      "step": 6083
+    },
+    {
+      "epoch": 0.37426181102362205,
+      "grad_norm": 1.698930025100708,
+      "learning_rate": 0.00013859415232349705,
+      "loss": 2.8094,
+      "step": 6084
+    },
+    {
+      "epoch": 0.37432332677165353,
+      "grad_norm": 1.3260328769683838,
+      "learning_rate": 0.00013857631218150478,
+      "loss": 2.5473,
+      "step": 6085
+    },
+    {
+      "epoch": 0.374384842519685,
+      "grad_norm": 1.0456068515777588,
+      "learning_rate": 0.00013855847059697352,
+      "loss": 2.7746,
+      "step": 6086
+    },
+    {
+      "epoch": 0.37444635826771655,
+      "grad_norm": 1.2280977964401245,
+      "learning_rate": 0.00013854062757057047,
+      "loss": 2.4971,
+      "step": 6087
+    },
+    {
+      "epoch": 0.37450787401574803,
+      "grad_norm": 1.295070767402649,
+      "learning_rate": 0.00013852278310296285,
+      "loss": 2.2864,
+      "step": 6088
+    },
+    {
+      "epoch": 0.3745693897637795,
+      "grad_norm": 1.0330873727798462,
+      "learning_rate": 0.000138504937194818,
+      "loss": 2.3717,
+      "step": 6089
+    },
+    {
+      "epoch": 0.374630905511811,
+      "grad_norm": 2.5258429050445557,
+      "learning_rate": 0.00013848708984680314,
+      "loss": 2.7336,
+      "step": 6090
+    },
+    {
+      "epoch": 0.37469242125984253,
+      "grad_norm": 1.2947332859039307,
+      "learning_rate": 0.00013846924105958576,
+      "loss": 3.3824,
+      "step": 6091
+    },
+    {
+      "epoch": 0.374753937007874,
+      "grad_norm": 1.255558967590332,
+      "learning_rate": 0.0001384513908338333,
+      "loss": 2.7636,
+      "step": 6092
+    },
+    {
+      "epoch": 0.3748154527559055,
+      "grad_norm": 1.1667470932006836,
+      "learning_rate": 0.0001384335391702132,
+      "loss": 3.0538,
+      "step": 6093
+    },
+    {
+      "epoch": 0.374876968503937,
+      "grad_norm": 1.1965667009353638,
+      "learning_rate": 0.00013841568606939305,
+      "loss": 2.9856,
+      "step": 6094
+    },
+    {
+      "epoch": 0.3749384842519685,
+      "grad_norm": 1.1273250579833984,
+      "learning_rate": 0.0001383978315320405,
+      "loss": 2.7815,
+      "step": 6095
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 1.0379092693328857,
+      "learning_rate": 0.0001383799755588231,
+      "loss": 2.8547,
+      "step": 6096
+    },
+    {
+      "epoch": 0.3750615157480315,
+      "grad_norm": 1.0672259330749512,
+      "learning_rate": 0.00013836211815040865,
+      "loss": 3.321,
+      "step": 6097
+    },
+    {
+      "epoch": 0.375123031496063,
+      "grad_norm": 1.3441846370697021,
+      "learning_rate": 0.00013834425930746492,
+      "loss": 2.4207,
+      "step": 6098
+    },
+    {
+      "epoch": 0.3751845472440945,
+      "grad_norm": 1.1419907808303833,
+      "learning_rate": 0.00013832639903065962,
+      "loss": 2.9524,
+      "step": 6099
+    },
+    {
+      "epoch": 0.375246062992126,
+      "grad_norm": 1.0658490657806396,
+      "learning_rate": 0.00013830853732066078,
+      "loss": 2.9959,
+      "step": 6100
+    },
+    {
+      "epoch": 0.37530757874015747,
+      "grad_norm": 1.11767578125,
+      "learning_rate": 0.00013829067417813628,
+      "loss": 2.7422,
+      "step": 6101
+    },
+    {
+      "epoch": 0.375369094488189,
+      "grad_norm": 1.29624605178833,
+      "learning_rate": 0.00013827280960375404,
+      "loss": 2.9171,
+      "step": 6102
+    },
+    {
+      "epoch": 0.3754306102362205,
+      "grad_norm": 1.5357835292816162,
+      "learning_rate": 0.00013825494359818213,
+      "loss": 2.1155,
+      "step": 6103
+    },
+    {
+      "epoch": 0.37549212598425197,
+      "grad_norm": 1.4714573621749878,
+      "learning_rate": 0.00013823707616208866,
+      "loss": 2.438,
+      "step": 6104
+    },
+    {
+      "epoch": 0.37555364173228345,
+      "grad_norm": 1.4878414869308472,
+      "learning_rate": 0.00013821920729614179,
+      "loss": 1.8436,
+      "step": 6105
+    },
+    {
+      "epoch": 0.375615157480315,
+      "grad_norm": 1.1968672275543213,
+      "learning_rate": 0.0001382013370010096,
+      "loss": 2.556,
+      "step": 6106
+    },
+    {
+      "epoch": 0.37567667322834647,
+      "grad_norm": 1.3054059743881226,
+      "learning_rate": 0.0001381834652773605,
+      "loss": 3.4244,
+      "step": 6107
+    },
+    {
+      "epoch": 0.37573818897637795,
+      "grad_norm": 1.243179202079773,
+      "learning_rate": 0.00013816559212586268,
+      "loss": 2.7083,
+      "step": 6108
+    },
+    {
+      "epoch": 0.37579970472440943,
+      "grad_norm": 1.3788121938705444,
+      "learning_rate": 0.00013814771754718454,
+      "loss": 2.3166,
+      "step": 6109
+    },
+    {
+      "epoch": 0.37586122047244097,
+      "grad_norm": 1.4444854259490967,
+      "learning_rate": 0.00013812984154199447,
+      "loss": 2.7063,
+      "step": 6110
+    },
+    {
+      "epoch": 0.37592273622047245,
+      "grad_norm": 5.334558963775635,
+      "learning_rate": 0.00013811196411096097,
+      "loss": 2.8901,
+      "step": 6111
+    },
+    {
+      "epoch": 0.37598425196850394,
+      "grad_norm": 1.2012500762939453,
+      "learning_rate": 0.00013809408525475249,
+      "loss": 2.9617,
+      "step": 6112
+    },
+    {
+      "epoch": 0.3760457677165354,
+      "grad_norm": 1.3764123916625977,
+      "learning_rate": 0.00013807620497403769,
+      "loss": 2.8033,
+      "step": 6113
+    },
+    {
+      "epoch": 0.37610728346456695,
+      "grad_norm": 1.3140250444412231,
+      "learning_rate": 0.0001380583232694851,
+      "loss": 2.3354,
+      "step": 6114
+    },
+    {
+      "epoch": 0.37616879921259844,
+      "grad_norm": 1.256007194519043,
+      "learning_rate": 0.00013804044014176348,
+      "loss": 1.9113,
+      "step": 6115
+    },
+    {
+      "epoch": 0.3762303149606299,
+      "grad_norm": 1.2224398851394653,
+      "learning_rate": 0.00013802255559154146,
+      "loss": 2.6607,
+      "step": 6116
+    },
+    {
+      "epoch": 0.3762918307086614,
+      "grad_norm": 1.2764501571655273,
+      "learning_rate": 0.00013800466961948793,
+      "loss": 2.3425,
+      "step": 6117
+    },
+    {
+      "epoch": 0.37635334645669294,
+      "grad_norm": 1.04802405834198,
+      "learning_rate": 0.00013798678222627165,
+      "loss": 2.7731,
+      "step": 6118
+    },
+    {
+      "epoch": 0.3764148622047244,
+      "grad_norm": 1.5740153789520264,
+      "learning_rate": 0.00013796889341256155,
+      "loss": 3.5191,
+      "step": 6119
+    },
+    {
+      "epoch": 0.3764763779527559,
+      "grad_norm": 1.438682198524475,
+      "learning_rate": 0.00013795100317902655,
+      "loss": 2.5304,
+      "step": 6120
+    },
+    {
+      "epoch": 0.3765378937007874,
+      "grad_norm": 1.437948226928711,
+      "learning_rate": 0.0001379331115263357,
+      "loss": 3.4534,
+      "step": 6121
+    },
+    {
+      "epoch": 0.3765994094488189,
+      "grad_norm": 1.2981996536254883,
+      "learning_rate": 0.00013791521845515792,
+      "loss": 2.6794,
+      "step": 6122
+    },
+    {
+      "epoch": 0.3766609251968504,
+      "grad_norm": 1.1193989515304565,
+      "learning_rate": 0.0001378973239661624,
+      "loss": 2.6172,
+      "step": 6123
+    },
+    {
+      "epoch": 0.3767224409448819,
+      "grad_norm": 0.9117259383201599,
+      "learning_rate": 0.00013787942806001833,
+      "loss": 2.7242,
+      "step": 6124
+    },
+    {
+      "epoch": 0.37678395669291337,
+      "grad_norm": 1.2078338861465454,
+      "learning_rate": 0.00013786153073739484,
+      "loss": 2.0523,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3768454724409449,
+      "grad_norm": 1.173665165901184,
+      "learning_rate": 0.00013784363199896124,
+      "loss": 2.7493,
+      "step": 6126
+    },
+    {
+      "epoch": 0.3769069881889764,
+      "grad_norm": 1.069958209991455,
+      "learning_rate": 0.00013782573184538684,
+      "loss": 2.3606,
+      "step": 6127
+    },
+    {
+      "epoch": 0.37696850393700787,
+      "grad_norm": 1.0471746921539307,
+      "learning_rate": 0.00013780783027734096,
+      "loss": 2.2574,
+      "step": 6128
+    },
+    {
+      "epoch": 0.37703001968503935,
+      "grad_norm": 1.1033258438110352,
+      "learning_rate": 0.00013778992729549306,
+      "loss": 2.8243,
+      "step": 6129
+    },
+    {
+      "epoch": 0.3770915354330709,
+      "grad_norm": 1.5116955041885376,
+      "learning_rate": 0.0001377720229005126,
+      "loss": 2.4444,
+      "step": 6130
+    },
+    {
+      "epoch": 0.3771530511811024,
+      "grad_norm": 1.1921970844268799,
+      "learning_rate": 0.0001377541170930691,
+      "loss": 2.7981,
+      "step": 6131
+    },
+    {
+      "epoch": 0.37721456692913385,
+      "grad_norm": 1.1816980838775635,
+      "learning_rate": 0.00013773620987383216,
+      "loss": 3.156,
+      "step": 6132
+    },
+    {
+      "epoch": 0.37727608267716534,
+      "grad_norm": 1.2703490257263184,
+      "learning_rate": 0.00013771830124347138,
+      "loss": 2.7229,
+      "step": 6133
+    },
+    {
+      "epoch": 0.3773375984251969,
+      "grad_norm": 1.2482250928878784,
+      "learning_rate": 0.00013770039120265646,
+      "loss": 3.1275,
+      "step": 6134
+    },
+    {
+      "epoch": 0.37739911417322836,
+      "grad_norm": 1.0548374652862549,
+      "learning_rate": 0.00013768247975205715,
+      "loss": 3.0173,
+      "step": 6135
+    },
+    {
+      "epoch": 0.37746062992125984,
+      "grad_norm": 1.5765684843063354,
+      "learning_rate": 0.0001376645668923432,
+      "loss": 2.6623,
+      "step": 6136
+    },
+    {
+      "epoch": 0.3775221456692913,
+      "grad_norm": 1.1703650951385498,
+      "learning_rate": 0.0001376466526241845,
+      "loss": 3.5088,
+      "step": 6137
+    },
+    {
+      "epoch": 0.37758366141732286,
+      "grad_norm": 0.9809045791625977,
+      "learning_rate": 0.00013762873694825084,
+      "loss": 2.8027,
+      "step": 6138
+    },
+    {
+      "epoch": 0.37764517716535434,
+      "grad_norm": 1.309194564819336,
+      "learning_rate": 0.00013761081986521232,
+      "loss": 2.7047,
+      "step": 6139
+    },
+    {
+      "epoch": 0.3777066929133858,
+      "grad_norm": 1.02065908908844,
+      "learning_rate": 0.00013759290137573883,
+      "loss": 2.7762,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3777682086614173,
+      "grad_norm": 1.0215626955032349,
+      "learning_rate": 0.00013757498148050042,
+      "loss": 2.9074,
+      "step": 6141
+    },
+    {
+      "epoch": 0.37782972440944884,
+      "grad_norm": 1.159311056137085,
+      "learning_rate": 0.00013755706018016728,
+      "loss": 2.7036,
+      "step": 6142
+    },
+    {
+      "epoch": 0.3778912401574803,
+      "grad_norm": 1.6409995555877686,
+      "learning_rate": 0.0001375391374754095,
+      "loss": 2.041,
+      "step": 6143
+    },
+    {
+      "epoch": 0.3779527559055118,
+      "grad_norm": 1.307798147201538,
+      "learning_rate": 0.0001375212133668973,
+      "loss": 2.799,
+      "step": 6144
+    },
+    {
+      "epoch": 0.3780142716535433,
+      "grad_norm": 1.436171531677246,
+      "learning_rate": 0.00013750328785530093,
+      "loss": 2.2355,
+      "step": 6145
+    },
+    {
+      "epoch": 0.3780757874015748,
+      "grad_norm": 1.151341199874878,
+      "learning_rate": 0.00013748536094129073,
+      "loss": 2.7268,
+      "step": 6146
+    },
+    {
+      "epoch": 0.3781373031496063,
+      "grad_norm": 1.1790707111358643,
+      "learning_rate": 0.00013746743262553704,
+      "loss": 2.8553,
+      "step": 6147
+    },
+    {
+      "epoch": 0.3781988188976378,
+      "grad_norm": 1.2069833278656006,
+      "learning_rate": 0.00013744950290871033,
+      "loss": 1.6747,
+      "step": 6148
+    },
+    {
+      "epoch": 0.37826033464566927,
+      "grad_norm": 1.2560701370239258,
+      "learning_rate": 0.000137431571791481,
+      "loss": 2.9943,
+      "step": 6149
+    },
+    {
+      "epoch": 0.3783218503937008,
+      "grad_norm": 1.158545970916748,
+      "learning_rate": 0.00013741363927451965,
+      "loss": 2.7562,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3783833661417323,
+      "grad_norm": 1.2097970247268677,
+      "learning_rate": 0.00013739570535849677,
+      "loss": 2.6124,
+      "step": 6151
+    },
+    {
+      "epoch": 0.3784448818897638,
+      "grad_norm": 1.1537448167800903,
+      "learning_rate": 0.00013737777004408305,
+      "loss": 2.7392,
+      "step": 6152
+    },
+    {
+      "epoch": 0.37850639763779526,
+      "grad_norm": 1.3593229055404663,
+      "learning_rate": 0.00013735983333194917,
+      "loss": 2.3961,
+      "step": 6153
+    },
+    {
+      "epoch": 0.3785679133858268,
+      "grad_norm": 1.1193783283233643,
+      "learning_rate": 0.00013734189522276582,
+      "loss": 2.6198,
+      "step": 6154
+    },
+    {
+      "epoch": 0.3786294291338583,
+      "grad_norm": 1.2796448469161987,
+      "learning_rate": 0.00013732395571720382,
+      "loss": 1.959,
+      "step": 6155
+    },
+    {
+      "epoch": 0.37869094488188976,
+      "grad_norm": 1.1342086791992188,
+      "learning_rate": 0.00013730601481593403,
+      "loss": 2.7276,
+      "step": 6156
+    },
+    {
+      "epoch": 0.37875246062992124,
+      "grad_norm": 1.272155523300171,
+      "learning_rate": 0.00013728807251962727,
+      "loss": 2.7537,
+      "step": 6157
+    },
+    {
+      "epoch": 0.3788139763779528,
+      "grad_norm": 1.2854337692260742,
+      "learning_rate": 0.00013727012882895455,
+      "loss": 2.4087,
+      "step": 6158
+    },
+    {
+      "epoch": 0.37887549212598426,
+      "grad_norm": 1.2747315168380737,
+      "learning_rate": 0.00013725218374458683,
+      "loss": 2.2918,
+      "step": 6159
+    },
+    {
+      "epoch": 0.37893700787401574,
+      "grad_norm": 1.0817631483078003,
+      "learning_rate": 0.00013723423726719514,
+      "loss": 2.6668,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3789985236220472,
+      "grad_norm": 1.2318754196166992,
+      "learning_rate": 0.00013721628939745062,
+      "loss": 2.4347,
+      "step": 6161
+    },
+    {
+      "epoch": 0.37906003937007876,
+      "grad_norm": 1.2916200160980225,
+      "learning_rate": 0.0001371983401360244,
+      "loss": 2.5888,
+      "step": 6162
+    },
+    {
+      "epoch": 0.37912155511811024,
+      "grad_norm": 1.0191254615783691,
+      "learning_rate": 0.0001371803894835877,
+      "loss": 2.5389,
+      "step": 6163
+    },
+    {
+      "epoch": 0.3791830708661417,
+      "grad_norm": 1.1751911640167236,
+      "learning_rate": 0.0001371624374408117,
+      "loss": 3.0812,
+      "step": 6164
+    },
+    {
+      "epoch": 0.3792445866141732,
+      "grad_norm": 1.1753803491592407,
+      "learning_rate": 0.00013714448400836782,
+      "loss": 2.788,
+      "step": 6165
+    },
+    {
+      "epoch": 0.37930610236220474,
+      "grad_norm": 1.2773629426956177,
+      "learning_rate": 0.00013712652918692735,
+      "loss": 2.5154,
+      "step": 6166
+    },
+    {
+      "epoch": 0.3793676181102362,
+      "grad_norm": 1.1421626806259155,
+      "learning_rate": 0.00013710857297716172,
+      "loss": 3.2921,
+      "step": 6167
+    },
+    {
+      "epoch": 0.3794291338582677,
+      "grad_norm": 1.2475318908691406,
+      "learning_rate": 0.00013709061537974234,
+      "loss": 2.4234,
+      "step": 6168
+    },
+    {
+      "epoch": 0.3794906496062992,
+      "grad_norm": 1.1253471374511719,
+      "learning_rate": 0.0001370726563953408,
+      "loss": 2.2823,
+      "step": 6169
+    },
+    {
+      "epoch": 0.37955216535433073,
+      "grad_norm": 0.9177619218826294,
+      "learning_rate": 0.00013705469602462862,
+      "loss": 1.9591,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3796136811023622,
+      "grad_norm": 1.1155534982681274,
+      "learning_rate": 0.00013703673426827747,
+      "loss": 2.2371,
+      "step": 6171
+    },
+    {
+      "epoch": 0.3796751968503937,
+      "grad_norm": 1.0254722833633423,
+      "learning_rate": 0.00013701877112695893,
+      "loss": 2.9368,
+      "step": 6172
+    },
+    {
+      "epoch": 0.3797367125984252,
+      "grad_norm": 1.1882402896881104,
+      "learning_rate": 0.00013700080660134477,
+      "loss": 2.4445,
+      "step": 6173
+    },
+    {
+      "epoch": 0.3797982283464567,
+      "grad_norm": 1.157615065574646,
+      "learning_rate": 0.00013698284069210681,
+      "loss": 3.2435,
+      "step": 6174
+    },
+    {
+      "epoch": 0.3798597440944882,
+      "grad_norm": 1.1998556852340698,
+      "learning_rate": 0.00013696487339991681,
+      "loss": 2.2612,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3799212598425197,
+      "grad_norm": 1.1720874309539795,
+      "learning_rate": 0.00013694690472544665,
+      "loss": 2.854,
+      "step": 6176
+    },
+    {
+      "epoch": 0.37998277559055116,
+      "grad_norm": 1.753970742225647,
+      "learning_rate": 0.0001369289346693683,
+      "loss": 2.2565,
+      "step": 6177
+    },
+    {
+      "epoch": 0.3800442913385827,
+      "grad_norm": 1.7811822891235352,
+      "learning_rate": 0.0001369109632323537,
+      "loss": 2.0996,
+      "step": 6178
+    },
+    {
+      "epoch": 0.3801058070866142,
+      "grad_norm": 1.348998785018921,
+      "learning_rate": 0.00013689299041507488,
+      "loss": 2.3926,
+      "step": 6179
+    },
+    {
+      "epoch": 0.38016732283464566,
+      "grad_norm": 1.1093744039535522,
+      "learning_rate": 0.00013687501621820398,
+      "loss": 2.5999,
+      "step": 6180
+    },
+    {
+      "epoch": 0.38022883858267714,
+      "grad_norm": 1.1450116634368896,
+      "learning_rate": 0.00013685704064241307,
+      "loss": 1.788,
+      "step": 6181
+    },
+    {
+      "epoch": 0.3802903543307087,
+      "grad_norm": 1.3898329734802246,
+      "learning_rate": 0.00013683906368837435,
+      "loss": 2.8487,
+      "step": 6182
+    },
+    {
+      "epoch": 0.38035187007874016,
+      "grad_norm": 1.29795503616333,
+      "learning_rate": 0.00013682108535676007,
+      "loss": 2.2469,
+      "step": 6183
+    },
+    {
+      "epoch": 0.38041338582677164,
+      "grad_norm": 1.057680606842041,
+      "learning_rate": 0.00013680310564824255,
+      "loss": 2.3609,
+      "step": 6184
+    },
+    {
+      "epoch": 0.3804749015748031,
+      "grad_norm": 1.035338282585144,
+      "learning_rate": 0.0001367851245634941,
+      "loss": 2.6757,
+      "step": 6185
+    },
+    {
+      "epoch": 0.38053641732283466,
+      "grad_norm": 1.2595950365066528,
+      "learning_rate": 0.00013676714210318706,
+      "loss": 1.991,
+      "step": 6186
+    },
+    {
+      "epoch": 0.38059793307086615,
+      "grad_norm": 1.1283055543899536,
+      "learning_rate": 0.00013674915826799398,
+      "loss": 3.3447,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3806594488188976,
+      "grad_norm": 1.2197872400283813,
+      "learning_rate": 0.00013673117305858728,
+      "loss": 2.3156,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3807209645669291,
+      "grad_norm": 1.1528962850570679,
+      "learning_rate": 0.00013671318647563949,
+      "loss": 3.031,
+      "step": 6189
+    },
+    {
+      "epoch": 0.38078248031496065,
+      "grad_norm": 1.4378628730773926,
+      "learning_rate": 0.00013669519851982327,
+      "loss": 2.7475,
+      "step": 6190
+    },
+    {
+      "epoch": 0.38084399606299213,
+      "grad_norm": 1.2725030183792114,
+      "learning_rate": 0.00013667720919181126,
+      "loss": 2.8989,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3809055118110236,
+      "grad_norm": 1.1283795833587646,
+      "learning_rate": 0.00013665921849227613,
+      "loss": 2.9266,
+      "step": 6192
+    },
+    {
+      "epoch": 0.3809670275590551,
+      "grad_norm": 1.1469789743423462,
+      "learning_rate": 0.00013664122642189063,
+      "loss": 2.6159,
+      "step": 6193
+    },
+    {
+      "epoch": 0.38102854330708663,
+      "grad_norm": 1.6298500299453735,
+      "learning_rate": 0.00013662323298132759,
+      "loss": 3.2404,
+      "step": 6194
+    },
+    {
+      "epoch": 0.3810900590551181,
+      "grad_norm": 1.5568195581436157,
+      "learning_rate": 0.00013660523817125988,
+      "loss": 2.2737,
+      "step": 6195
+    },
+    {
+      "epoch": 0.3811515748031496,
+      "grad_norm": 0.9571136832237244,
+      "learning_rate": 0.00013658724199236034,
+      "loss": 2.8649,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3812130905511811,
+      "grad_norm": 0.9571226239204407,
+      "learning_rate": 0.00013656924444530199,
+      "loss": 2.9593,
+      "step": 6197
+    },
+    {
+      "epoch": 0.3812746062992126,
+      "grad_norm": 1.2323343753814697,
+      "learning_rate": 0.0001365512455307578,
+      "loss": 2.6008,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3813361220472441,
+      "grad_norm": 1.0545011758804321,
+      "learning_rate": 0.00013653324524940085,
+      "loss": 2.6242,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3813976377952756,
+      "grad_norm": 1.1564081907272339,
+      "learning_rate": 0.00013651524360190425,
+      "loss": 2.496,
+      "step": 6200
+    },
+    {
+      "epoch": 0.38145915354330706,
+      "grad_norm": 1.0912898778915405,
+      "learning_rate": 0.00013649724058894115,
+      "loss": 2.7421,
+      "step": 6201
+    },
+    {
+      "epoch": 0.3815206692913386,
+      "grad_norm": 1.3347585201263428,
+      "learning_rate": 0.00013647923621118476,
+      "loss": 2.4042,
+      "step": 6202
+    },
+    {
+      "epoch": 0.3815821850393701,
+      "grad_norm": 1.111948013305664,
+      "learning_rate": 0.00013646123046930834,
+      "loss": 2.1443,
+      "step": 6203
+    },
+    {
+      "epoch": 0.38164370078740156,
+      "grad_norm": 1.1867355108261108,
+      "learning_rate": 0.00013644322336398527,
+      "loss": 2.6239,
+      "step": 6204
+    },
+    {
+      "epoch": 0.38170521653543305,
+      "grad_norm": 1.0927470922470093,
+      "learning_rate": 0.00013642521489588879,
+      "loss": 3.1098,
+      "step": 6205
+    },
+    {
+      "epoch": 0.3817667322834646,
+      "grad_norm": 1.1535395383834839,
+      "learning_rate": 0.0001364072050656924,
+      "loss": 2.8536,
+      "step": 6206
+    },
+    {
+      "epoch": 0.38182824803149606,
+      "grad_norm": 1.06503427028656,
+      "learning_rate": 0.00013638919387406957,
+      "loss": 2.6713,
+      "step": 6207
+    },
+    {
+      "epoch": 0.38188976377952755,
+      "grad_norm": 1.2530103921890259,
+      "learning_rate": 0.0001363711813216938,
+      "loss": 2.2657,
+      "step": 6208
+    },
+    {
+      "epoch": 0.38195127952755903,
+      "grad_norm": 1.06989586353302,
+      "learning_rate": 0.00013635316740923865,
+      "loss": 2.4642,
+      "step": 6209
+    },
+    {
+      "epoch": 0.38201279527559057,
+      "grad_norm": 1.0189834833145142,
+      "learning_rate": 0.00013633515213737774,
+      "loss": 3.0187,
+      "step": 6210
+    },
+    {
+      "epoch": 0.38207431102362205,
+      "grad_norm": 1.229945421218872,
+      "learning_rate": 0.00013631713550678478,
+      "loss": 2.9001,
+      "step": 6211
+    },
+    {
+      "epoch": 0.38213582677165353,
+      "grad_norm": 1.2227282524108887,
+      "learning_rate": 0.00013629911751813345,
+      "loss": 2.1758,
+      "step": 6212
+    },
+    {
+      "epoch": 0.382197342519685,
+      "grad_norm": 1.4739022254943848,
+      "learning_rate": 0.00013628109817209755,
+      "loss": 2.3131,
+      "step": 6213
+    },
+    {
+      "epoch": 0.38225885826771655,
+      "grad_norm": 1.3487902879714966,
+      "learning_rate": 0.00013626307746935083,
+      "loss": 2.6125,
+      "step": 6214
+    },
+    {
+      "epoch": 0.38232037401574803,
+      "grad_norm": 1.1601017713546753,
+      "learning_rate": 0.00013624505541056726,
+      "loss": 3.1204,
+      "step": 6215
+    },
+    {
+      "epoch": 0.3823818897637795,
+      "grad_norm": 1.2807209491729736,
+      "learning_rate": 0.00013622703199642073,
+      "loss": 2.619,
+      "step": 6216
+    },
+    {
+      "epoch": 0.382443405511811,
+      "grad_norm": 1.105197548866272,
+      "learning_rate": 0.0001362090072275852,
+      "loss": 2.8416,
+      "step": 6217
+    },
+    {
+      "epoch": 0.38250492125984253,
+      "grad_norm": 1.1799781322479248,
+      "learning_rate": 0.0001361909811047347,
+      "loss": 2.7426,
+      "step": 6218
+    },
+    {
+      "epoch": 0.382566437007874,
+      "grad_norm": 1.2924667596817017,
+      "learning_rate": 0.00013617295362854334,
+      "loss": 2.7174,
+      "step": 6219
+    },
+    {
+      "epoch": 0.3826279527559055,
+      "grad_norm": 1.0743738412857056,
+      "learning_rate": 0.00013615492479968522,
+      "loss": 3.2278,
+      "step": 6220
+    },
+    {
+      "epoch": 0.382689468503937,
+      "grad_norm": 1.143014669418335,
+      "learning_rate": 0.00013613689461883447,
+      "loss": 2.4465,
+      "step": 6221
+    },
+    {
+      "epoch": 0.3827509842519685,
+      "grad_norm": 1.0862133502960205,
+      "learning_rate": 0.0001361188630866654,
+      "loss": 2.4409,
+      "step": 6222
+    },
+    {
+      "epoch": 0.3828125,
+      "grad_norm": 1.173578143119812,
+      "learning_rate": 0.00013610083020385228,
+      "loss": 2.5881,
+      "step": 6223
+    },
+    {
+      "epoch": 0.3828740157480315,
+      "grad_norm": 1.2310242652893066,
+      "learning_rate": 0.00013608279597106937,
+      "loss": 2.1278,
+      "step": 6224
+    },
+    {
+      "epoch": 0.382935531496063,
+      "grad_norm": 1.2232387065887451,
+      "learning_rate": 0.00013606476038899112,
+      "loss": 2.8153,
+      "step": 6225
+    },
+    {
+      "epoch": 0.3829970472440945,
+      "grad_norm": 1.1542248725891113,
+      "learning_rate": 0.00013604672345829194,
+      "loss": 3.1053,
+      "step": 6226
+    },
+    {
+      "epoch": 0.383058562992126,
+      "grad_norm": 1.1778006553649902,
+      "learning_rate": 0.00013602868517964626,
+      "loss": 2.702,
+      "step": 6227
+    },
+    {
+      "epoch": 0.38312007874015747,
+      "grad_norm": 1.236549735069275,
+      "learning_rate": 0.00013601064555372869,
+      "loss": 2.1237,
+      "step": 6228
+    },
+    {
+      "epoch": 0.383181594488189,
+      "grad_norm": 1.3501158952713013,
+      "learning_rate": 0.0001359926045812138,
+      "loss": 2.8772,
+      "step": 6229
+    },
+    {
+      "epoch": 0.3832431102362205,
+      "grad_norm": 1.2166824340820312,
+      "learning_rate": 0.00013597456226277616,
+      "loss": 2.7913,
+      "step": 6230
+    },
+    {
+      "epoch": 0.38330462598425197,
+      "grad_norm": 1.300389051437378,
+      "learning_rate": 0.0001359565185990905,
+      "loss": 2.3563,
+      "step": 6231
+    },
+    {
+      "epoch": 0.38336614173228345,
+      "grad_norm": 1.328697919845581,
+      "learning_rate": 0.00013593847359083153,
+      "loss": 2.7851,
+      "step": 6232
+    },
+    {
+      "epoch": 0.383427657480315,
+      "grad_norm": 1.1863569021224976,
+      "learning_rate": 0.00013592042723867409,
+      "loss": 2.3664,
+      "step": 6233
+    },
+    {
+      "epoch": 0.38348917322834647,
+      "grad_norm": 1.5099600553512573,
+      "learning_rate": 0.0001359023795432929,
+      "loss": 2.1804,
+      "step": 6234
+    },
+    {
+      "epoch": 0.38355068897637795,
+      "grad_norm": 1.1512418985366821,
+      "learning_rate": 0.00013588433050536297,
+      "loss": 3.0031,
+      "step": 6235
+    },
+    {
+      "epoch": 0.38361220472440943,
+      "grad_norm": 1.3219692707061768,
+      "learning_rate": 0.00013586628012555915,
+      "loss": 2.5039,
+      "step": 6236
+    },
+    {
+      "epoch": 0.38367372047244097,
+      "grad_norm": 1.2457973957061768,
+      "learning_rate": 0.00013584822840455646,
+      "loss": 2.7431,
+      "step": 6237
+    },
+    {
+      "epoch": 0.38373523622047245,
+      "grad_norm": 1.3940659761428833,
+      "learning_rate": 0.00013583017534302992,
+      "loss": 2.736,
+      "step": 6238
+    },
+    {
+      "epoch": 0.38379675196850394,
+      "grad_norm": 1.2981562614440918,
+      "learning_rate": 0.00013581212094165463,
+      "loss": 2.3712,
+      "step": 6239
+    },
+    {
+      "epoch": 0.3838582677165354,
+      "grad_norm": 1.4010294675827026,
+      "learning_rate": 0.0001357940652011057,
+      "loss": 2.5706,
+      "step": 6240
+    },
+    {
+      "epoch": 0.38391978346456695,
+      "grad_norm": 1.5789843797683716,
+      "learning_rate": 0.00013577600812205833,
+      "loss": 3.2681,
+      "step": 6241
+    },
+    {
+      "epoch": 0.38398129921259844,
+      "grad_norm": 1.1936103105545044,
+      "learning_rate": 0.00013575794970518775,
+      "loss": 2.4051,
+      "step": 6242
+    },
+    {
+      "epoch": 0.3840428149606299,
+      "grad_norm": 1.3350406885147095,
+      "learning_rate": 0.00013573988995116922,
+      "loss": 2.6919,
+      "step": 6243
+    },
+    {
+      "epoch": 0.3841043307086614,
+      "grad_norm": 1.3581995964050293,
+      "learning_rate": 0.00013572182886067811,
+      "loss": 2.6955,
+      "step": 6244
+    },
+    {
+      "epoch": 0.38416584645669294,
+      "grad_norm": 1.089999794960022,
+      "learning_rate": 0.00013570376643438982,
+      "loss": 2.5378,
+      "step": 6245
+    },
+    {
+      "epoch": 0.3842273622047244,
+      "grad_norm": 1.5015697479248047,
+      "learning_rate": 0.0001356857026729797,
+      "loss": 2.2976,
+      "step": 6246
+    },
+    {
+      "epoch": 0.3842888779527559,
+      "grad_norm": 1.0759966373443604,
+      "learning_rate": 0.00013566763757712333,
+      "loss": 2.8889,
+      "step": 6247
+    },
+    {
+      "epoch": 0.3843503937007874,
+      "grad_norm": 1.0936856269836426,
+      "learning_rate": 0.0001356495711474962,
+      "loss": 3.0764,
+      "step": 6248
+    },
+    {
+      "epoch": 0.3844119094488189,
+      "grad_norm": 1.306078553199768,
+      "learning_rate": 0.00013563150338477385,
+      "loss": 2.0962,
+      "step": 6249
+    },
+    {
+      "epoch": 0.3844734251968504,
+      "grad_norm": 1.3146740198135376,
+      "learning_rate": 0.000135613434289632,
+      "loss": 1.9985,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3845349409448819,
+      "grad_norm": 1.1629019975662231,
+      "learning_rate": 0.00013559536386274632,
+      "loss": 2.3276,
+      "step": 6251
+    },
+    {
+      "epoch": 0.38459645669291337,
+      "grad_norm": 1.145195722579956,
+      "learning_rate": 0.00013557729210479248,
+      "loss": 2.7259,
+      "step": 6252
+    },
+    {
+      "epoch": 0.3846579724409449,
+      "grad_norm": 1.3497421741485596,
+      "learning_rate": 0.0001355592190164463,
+      "loss": 1.9602,
+      "step": 6253
+    },
+    {
+      "epoch": 0.3847194881889764,
+      "grad_norm": 1.4891456365585327,
+      "learning_rate": 0.00013554114459838362,
+      "loss": 2.6215,
+      "step": 6254
+    },
+    {
+      "epoch": 0.38478100393700787,
+      "grad_norm": 1.2130391597747803,
+      "learning_rate": 0.00013552306885128033,
+      "loss": 3.12,
+      "step": 6255
+    },
+    {
+      "epoch": 0.38484251968503935,
+      "grad_norm": 1.0259039402008057,
+      "learning_rate": 0.0001355049917758123,
+      "loss": 2.5435,
+      "step": 6256
+    },
+    {
+      "epoch": 0.3849040354330709,
+      "grad_norm": 1.1373729705810547,
+      "learning_rate": 0.0001354869133726556,
+      "loss": 2.5271,
+      "step": 6257
+    },
+    {
+      "epoch": 0.3849655511811024,
+      "grad_norm": 1.1652559041976929,
+      "learning_rate": 0.0001354688336424862,
+      "loss": 2.309,
+      "step": 6258
+    },
+    {
+      "epoch": 0.38502706692913385,
+      "grad_norm": 1.136067271232605,
+      "learning_rate": 0.0001354507525859802,
+      "loss": 2.4825,
+      "step": 6259
+    },
+    {
+      "epoch": 0.38508858267716534,
+      "grad_norm": 1.1034777164459229,
+      "learning_rate": 0.00013543267020381375,
+      "loss": 2.9804,
+      "step": 6260
+    },
+    {
+      "epoch": 0.3851500984251969,
+      "grad_norm": 1.227372646331787,
+      "learning_rate": 0.00013541458649666305,
+      "loss": 2.8913,
+      "step": 6261
+    },
+    {
+      "epoch": 0.38521161417322836,
+      "grad_norm": 1.269643783569336,
+      "learning_rate": 0.0001353965014652042,
+      "loss": 3.1055,
+      "step": 6262
+    },
+    {
+      "epoch": 0.38527312992125984,
+      "grad_norm": 0.9603456854820251,
+      "learning_rate": 0.00013537841511011362,
+      "loss": 2.7396,
+      "step": 6263
+    },
+    {
+      "epoch": 0.3853346456692913,
+      "grad_norm": 1.2541306018829346,
+      "learning_rate": 0.0001353603274320676,
+      "loss": 2.318,
+      "step": 6264
+    },
+    {
+      "epoch": 0.38539616141732286,
+      "grad_norm": 1.2491638660430908,
+      "learning_rate": 0.00013534223843174252,
+      "loss": 2.7545,
+      "step": 6265
+    },
+    {
+      "epoch": 0.38545767716535434,
+      "grad_norm": 1.2328311204910278,
+      "learning_rate": 0.00013532414810981476,
+      "loss": 2.7755,
+      "step": 6266
+    },
+    {
+      "epoch": 0.3855191929133858,
+      "grad_norm": 1.244247317314148,
+      "learning_rate": 0.00013530605646696085,
+      "loss": 2.6535,
+      "step": 6267
+    },
+    {
+      "epoch": 0.3855807086614173,
+      "grad_norm": 1.3552314043045044,
+      "learning_rate": 0.0001352879635038573,
+      "loss": 2.3998,
+      "step": 6268
+    },
+    {
+      "epoch": 0.38564222440944884,
+      "grad_norm": 1.295143961906433,
+      "learning_rate": 0.0001352698692211807,
+      "loss": 2.2148,
+      "step": 6269
+    },
+    {
+      "epoch": 0.3857037401574803,
+      "grad_norm": 1.3830538988113403,
+      "learning_rate": 0.00013525177361960763,
+      "loss": 2.9429,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3857652559055118,
+      "grad_norm": 1.2406446933746338,
+      "learning_rate": 0.00013523367669981482,
+      "loss": 3.2898,
+      "step": 6271
+    },
+    {
+      "epoch": 0.3858267716535433,
+      "grad_norm": 1.2016392946243286,
+      "learning_rate": 0.00013521557846247895,
+      "loss": 2.5141,
+      "step": 6272
+    },
+    {
+      "epoch": 0.3858882874015748,
+      "grad_norm": 1.1626306772232056,
+      "learning_rate": 0.00013519747890827682,
+      "loss": 3.0015,
+      "step": 6273
+    },
+    {
+      "epoch": 0.3859498031496063,
+      "grad_norm": 0.9395928978919983,
+      "learning_rate": 0.00013517937803788526,
+      "loss": 3.012,
+      "step": 6274
+    },
+    {
+      "epoch": 0.3860113188976378,
+      "grad_norm": 1.0800964832305908,
+      "learning_rate": 0.00013516127585198108,
+      "loss": 2.0307,
+      "step": 6275
+    },
+    {
+      "epoch": 0.38607283464566927,
+      "grad_norm": 1.1150048971176147,
+      "learning_rate": 0.0001351431723512413,
+      "loss": 2.7302,
+      "step": 6276
+    },
+    {
+      "epoch": 0.3861343503937008,
+      "grad_norm": 1.1441500186920166,
+      "learning_rate": 0.00013512506753634286,
+      "loss": 2.4856,
+      "step": 6277
+    },
+    {
+      "epoch": 0.3861958661417323,
+      "grad_norm": 1.145790457725525,
+      "learning_rate": 0.0001351069614079627,
+      "loss": 1.7745,
+      "step": 6278
+    },
+    {
+      "epoch": 0.3862573818897638,
+      "grad_norm": 1.4161159992218018,
+      "learning_rate": 0.00013508885396677797,
+      "loss": 2.4994,
+      "step": 6279
+    },
+    {
+      "epoch": 0.38631889763779526,
+      "grad_norm": 1.0775854587554932,
+      "learning_rate": 0.00013507074521346577,
+      "loss": 2.4887,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3863804133858268,
+      "grad_norm": 1.4543919563293457,
+      "learning_rate": 0.00013505263514870322,
+      "loss": 2.7611,
+      "step": 6281
+    },
+    {
+      "epoch": 0.3864419291338583,
+      "grad_norm": 1.2676063776016235,
+      "learning_rate": 0.0001350345237731676,
+      "loss": 2.3684,
+      "step": 6282
+    },
+    {
+      "epoch": 0.38650344488188976,
+      "grad_norm": 1.0813559293746948,
+      "learning_rate": 0.00013501641108753617,
+      "loss": 2.8479,
+      "step": 6283
+    },
+    {
+      "epoch": 0.38656496062992124,
+      "grad_norm": 1.139850378036499,
+      "learning_rate": 0.00013499829709248616,
+      "loss": 2.204,
+      "step": 6284
+    },
+    {
+      "epoch": 0.3866264763779528,
+      "grad_norm": 1.2974610328674316,
+      "learning_rate": 0.00013498018178869506,
+      "loss": 2.3324,
+      "step": 6285
+    },
+    {
+      "epoch": 0.38668799212598426,
+      "grad_norm": 1.2752776145935059,
+      "learning_rate": 0.0001349620651768402,
+      "loss": 2.8343,
+      "step": 6286
+    },
+    {
+      "epoch": 0.38674950787401574,
+      "grad_norm": 1.5342998504638672,
+      "learning_rate": 0.00013494394725759905,
+      "loss": 2.6366,
+      "step": 6287
+    },
+    {
+      "epoch": 0.3868110236220472,
+      "grad_norm": 1.1882953643798828,
+      "learning_rate": 0.0001349258280316491,
+      "loss": 2.8569,
+      "step": 6288
+    },
+    {
+      "epoch": 0.38687253937007876,
+      "grad_norm": 1.1692194938659668,
+      "learning_rate": 0.00013490770749966796,
+      "loss": 2.4647,
+      "step": 6289
+    },
+    {
+      "epoch": 0.38693405511811024,
+      "grad_norm": 1.3821157217025757,
+      "learning_rate": 0.00013488958566233323,
+      "loss": 2.481,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3869955708661417,
+      "grad_norm": 1.2317073345184326,
+      "learning_rate": 0.0001348714625203225,
+      "loss": 2.4122,
+      "step": 6291
+    },
+    {
+      "epoch": 0.3870570866141732,
+      "grad_norm": 1.0679781436920166,
+      "learning_rate": 0.00013485333807431354,
+      "loss": 2.5358,
+      "step": 6292
+    },
+    {
+      "epoch": 0.38711860236220474,
+      "grad_norm": 1.3584920167922974,
+      "learning_rate": 0.0001348352123249841,
+      "loss": 2.9361,
+      "step": 6293
+    },
+    {
+      "epoch": 0.3871801181102362,
+      "grad_norm": 1.114866018295288,
+      "learning_rate": 0.00013481708527301193,
+      "loss": 3.362,
+      "step": 6294
+    },
+    {
+      "epoch": 0.3872416338582677,
+      "grad_norm": 1.2538416385650635,
+      "learning_rate": 0.00013479895691907495,
+      "loss": 2.33,
+      "step": 6295
+    },
+    {
+      "epoch": 0.3873031496062992,
+      "grad_norm": 1.600911259651184,
+      "learning_rate": 0.00013478082726385102,
+      "loss": 2.7761,
+      "step": 6296
+    },
+    {
+      "epoch": 0.38736466535433073,
+      "grad_norm": 1.1095097064971924,
+      "learning_rate": 0.00013476269630801808,
+      "loss": 3.022,
+      "step": 6297
+    },
+    {
+      "epoch": 0.3874261811023622,
+      "grad_norm": 1.0862045288085938,
+      "learning_rate": 0.00013474456405225416,
+      "loss": 3.1303,
+      "step": 6298
+    },
+    {
+      "epoch": 0.3874876968503937,
+      "grad_norm": 0.851887583732605,
+      "learning_rate": 0.00013472643049723727,
+      "loss": 1.3804,
+      "step": 6299
+    },
+    {
+      "epoch": 0.3875492125984252,
+      "grad_norm": 1.067521572113037,
+      "learning_rate": 0.00013470829564364554,
+      "loss": 2.5986,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3876107283464567,
+      "grad_norm": 1.0988916158676147,
+      "learning_rate": 0.00013469015949215703,
+      "loss": 2.714,
+      "step": 6301
+    },
+    {
+      "epoch": 0.3876722440944882,
+      "grad_norm": 1.2808791399002075,
+      "learning_rate": 0.00013467202204345006,
+      "loss": 2.6086,
+      "step": 6302
+    },
+    {
+      "epoch": 0.3877337598425197,
+      "grad_norm": 1.1220426559448242,
+      "learning_rate": 0.0001346538832982028,
+      "loss": 2.8855,
+      "step": 6303
+    },
+    {
+      "epoch": 0.38779527559055116,
+      "grad_norm": 1.1140552759170532,
+      "learning_rate": 0.0001346357432570935,
+      "loss": 2.5942,
+      "step": 6304
+    },
+    {
+      "epoch": 0.3878567913385827,
+      "grad_norm": 1.0192300081253052,
+      "learning_rate": 0.00013461760192080054,
+      "loss": 2.8458,
+      "step": 6305
+    },
+    {
+      "epoch": 0.3879183070866142,
+      "grad_norm": 1.2910425662994385,
+      "learning_rate": 0.00013459945929000233,
+      "loss": 2.9098,
+      "step": 6306
+    },
+    {
+      "epoch": 0.38797982283464566,
+      "grad_norm": 1.1545531749725342,
+      "learning_rate": 0.00013458131536537725,
+      "loss": 2.454,
+      "step": 6307
+    },
+    {
+      "epoch": 0.38804133858267714,
+      "grad_norm": 1.709187626838684,
+      "learning_rate": 0.00013456317014760383,
+      "loss": 2.5984,
+      "step": 6308
+    },
+    {
+      "epoch": 0.3881028543307087,
+      "grad_norm": 1.380447268486023,
+      "learning_rate": 0.00013454502363736054,
+      "loss": 2.6836,
+      "step": 6309
+    },
+    {
+      "epoch": 0.38816437007874016,
+      "grad_norm": 1.2324227094650269,
+      "learning_rate": 0.00013452687583532602,
+      "loss": 2.5363,
+      "step": 6310
+    },
+    {
+      "epoch": 0.38822588582677164,
+      "grad_norm": 1.2408297061920166,
+      "learning_rate": 0.00013450872674217887,
+      "loss": 2.1867,
+      "step": 6311
+    },
+    {
+      "epoch": 0.3882874015748031,
+      "grad_norm": 1.116060733795166,
+      "learning_rate": 0.00013449057635859775,
+      "loss": 2.0438,
+      "step": 6312
+    },
+    {
+      "epoch": 0.38834891732283466,
+      "grad_norm": 1.3586562871932983,
+      "learning_rate": 0.00013447242468526144,
+      "loss": 2.3818,
+      "step": 6313
+    },
+    {
+      "epoch": 0.38841043307086615,
+      "grad_norm": 1.075305461883545,
+      "learning_rate": 0.00013445427172284865,
+      "loss": 2.8343,
+      "step": 6314
+    },
+    {
+      "epoch": 0.3884719488188976,
+      "grad_norm": 1.2365964651107788,
+      "learning_rate": 0.0001344361174720382,
+      "loss": 2.6138,
+      "step": 6315
+    },
+    {
+      "epoch": 0.3885334645669291,
+      "grad_norm": 1.0897717475891113,
+      "learning_rate": 0.000134417961933509,
+      "loss": 2.5737,
+      "step": 6316
+    },
+    {
+      "epoch": 0.38859498031496065,
+      "grad_norm": 1.0422922372817993,
+      "learning_rate": 0.00013439980510793992,
+      "loss": 3.0037,
+      "step": 6317
+    },
+    {
+      "epoch": 0.38865649606299213,
+      "grad_norm": 1.1961582899093628,
+      "learning_rate": 0.00013438164699600998,
+      "loss": 2.5577,
+      "step": 6318
+    },
+    {
+      "epoch": 0.3887180118110236,
+      "grad_norm": 1.3072936534881592,
+      "learning_rate": 0.00013436348759839813,
+      "loss": 2.803,
+      "step": 6319
+    },
+    {
+      "epoch": 0.3887795275590551,
+      "grad_norm": 1.2356388568878174,
+      "learning_rate": 0.00013434532691578347,
+      "loss": 2.72,
+      "step": 6320
+    },
+    {
+      "epoch": 0.38884104330708663,
+      "grad_norm": 1.0316483974456787,
+      "learning_rate": 0.00013432716494884509,
+      "loss": 2.5605,
+      "step": 6321
+    },
+    {
+      "epoch": 0.3889025590551181,
+      "grad_norm": 1.1147468090057373,
+      "learning_rate": 0.00013430900169826218,
+      "loss": 2.5343,
+      "step": 6322
+    },
+    {
+      "epoch": 0.3889640748031496,
+      "grad_norm": 1.2991392612457275,
+      "learning_rate": 0.0001342908371647139,
+      "loss": 2.6626,
+      "step": 6323
+    },
+    {
+      "epoch": 0.3890255905511811,
+      "grad_norm": 1.174857258796692,
+      "learning_rate": 0.00013427267134887952,
+      "loss": 2.276,
+      "step": 6324
+    },
+    {
+      "epoch": 0.3890871062992126,
+      "grad_norm": 1.0855109691619873,
+      "learning_rate": 0.00013425450425143834,
+      "loss": 2.7732,
+      "step": 6325
+    },
+    {
+      "epoch": 0.3891486220472441,
+      "grad_norm": 1.1785802841186523,
+      "learning_rate": 0.0001342363358730697,
+      "loss": 2.8133,
+      "step": 6326
+    },
+    {
+      "epoch": 0.3892101377952756,
+      "grad_norm": 1.1648176908493042,
+      "learning_rate": 0.00013421816621445306,
+      "loss": 2.6141,
+      "step": 6327
+    },
+    {
+      "epoch": 0.38927165354330706,
+      "grad_norm": 1.2860115766525269,
+      "learning_rate": 0.00013419999527626775,
+      "loss": 1.9345,
+      "step": 6328
+    },
+    {
+      "epoch": 0.3893331692913386,
+      "grad_norm": 1.1447482109069824,
+      "learning_rate": 0.00013418182305919332,
+      "loss": 2.9367,
+      "step": 6329
+    },
+    {
+      "epoch": 0.3893946850393701,
+      "grad_norm": 1.1615244150161743,
+      "learning_rate": 0.00013416364956390936,
+      "loss": 2.5493,
+      "step": 6330
+    },
+    {
+      "epoch": 0.38945620078740156,
+      "grad_norm": 1.309278130531311,
+      "learning_rate": 0.00013414547479109537,
+      "loss": 2.4907,
+      "step": 6331
+    },
+    {
+      "epoch": 0.38951771653543305,
+      "grad_norm": 1.0784106254577637,
+      "learning_rate": 0.000134127298741431,
+      "loss": 2.5065,
+      "step": 6332
+    },
+    {
+      "epoch": 0.3895792322834646,
+      "grad_norm": 1.0308576822280884,
+      "learning_rate": 0.000134109121415596,
+      "loss": 2.8669,
+      "step": 6333
+    },
+    {
+      "epoch": 0.38964074803149606,
+      "grad_norm": 1.5188177824020386,
+      "learning_rate": 0.00013409094281427003,
+      "loss": 2.4885,
+      "step": 6334
+    },
+    {
+      "epoch": 0.38970226377952755,
+      "grad_norm": 1.117801547050476,
+      "learning_rate": 0.00013407276293813294,
+      "loss": 2.4098,
+      "step": 6335
+    },
+    {
+      "epoch": 0.38976377952755903,
+      "grad_norm": 1.3570841550827026,
+      "learning_rate": 0.00013405458178786445,
+      "loss": 2.8558,
+      "step": 6336
+    },
+    {
+      "epoch": 0.38982529527559057,
+      "grad_norm": 1.2044456005096436,
+      "learning_rate": 0.00013403639936414454,
+      "loss": 2.7626,
+      "step": 6337
+    },
+    {
+      "epoch": 0.38988681102362205,
+      "grad_norm": 1.2574272155761719,
+      "learning_rate": 0.00013401821566765308,
+      "loss": 2.4974,
+      "step": 6338
+    },
+    {
+      "epoch": 0.38994832677165353,
+      "grad_norm": 1.1001025438308716,
+      "learning_rate": 0.00013400003069907,
+      "loss": 1.8652,
+      "step": 6339
+    },
+    {
+      "epoch": 0.390009842519685,
+      "grad_norm": 1.1704692840576172,
+      "learning_rate": 0.00013398184445907542,
+      "loss": 2.7849,
+      "step": 6340
+    },
+    {
+      "epoch": 0.39007135826771655,
+      "grad_norm": 1.2496256828308105,
+      "learning_rate": 0.00013396365694834934,
+      "loss": 2.3222,
+      "step": 6341
+    },
+    {
+      "epoch": 0.39013287401574803,
+      "grad_norm": 1.270049810409546,
+      "learning_rate": 0.00013394546816757183,
+      "loss": 2.8886,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3901943897637795,
+      "grad_norm": 1.2966049909591675,
+      "learning_rate": 0.00013392727811742313,
+      "loss": 3.0599,
+      "step": 6343
+    },
+    {
+      "epoch": 0.390255905511811,
+      "grad_norm": 1.6204458475112915,
+      "learning_rate": 0.0001339090867985834,
+      "loss": 2.1748,
+      "step": 6344
+    },
+    {
+      "epoch": 0.39031742125984253,
+      "grad_norm": 1.0754252672195435,
+      "learning_rate": 0.0001338908942117329,
+      "loss": 2.0173,
+      "step": 6345
+    },
+    {
+      "epoch": 0.390378937007874,
+      "grad_norm": 1.0948486328125,
+      "learning_rate": 0.00013387270035755198,
+      "loss": 3.1655,
+      "step": 6346
+    },
+    {
+      "epoch": 0.3904404527559055,
+      "grad_norm": 1.4261516332626343,
+      "learning_rate": 0.0001338545052367209,
+      "loss": 2.2806,
+      "step": 6347
+    },
+    {
+      "epoch": 0.390501968503937,
+      "grad_norm": 1.4651297330856323,
+      "learning_rate": 0.00013383630884992014,
+      "loss": 2.8069,
+      "step": 6348
+    },
+    {
+      "epoch": 0.3905634842519685,
+      "grad_norm": 1.5324660539627075,
+      "learning_rate": 0.00013381811119783006,
+      "loss": 2.2549,
+      "step": 6349
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 1.2602962255477905,
+      "learning_rate": 0.0001337999122811312,
+      "loss": 2.1932,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3906865157480315,
+      "grad_norm": 1.124198079109192,
+      "learning_rate": 0.00013378171210050413,
+      "loss": 2.4957,
+      "step": 6351
+    },
+    {
+      "epoch": 0.390748031496063,
+      "grad_norm": 1.364725112915039,
+      "learning_rate": 0.0001337635106566294,
+      "loss": 2.7105,
+      "step": 6352
+    },
+    {
+      "epoch": 0.3908095472440945,
+      "grad_norm": 1.4484244585037231,
+      "learning_rate": 0.00013374530795018762,
+      "loss": 2.8969,
+      "step": 6353
+    },
+    {
+      "epoch": 0.390871062992126,
+      "grad_norm": 1.192599892616272,
+      "learning_rate": 0.00013372710398185952,
+      "loss": 2.5839,
+      "step": 6354
+    },
+    {
+      "epoch": 0.39093257874015747,
+      "grad_norm": 1.1410727500915527,
+      "learning_rate": 0.00013370889875232575,
+      "loss": 2.6653,
+      "step": 6355
+    },
+    {
+      "epoch": 0.390994094488189,
+      "grad_norm": 1.1915457248687744,
+      "learning_rate": 0.00013369069226226718,
+      "loss": 2.2302,
+      "step": 6356
+    },
+    {
+      "epoch": 0.3910556102362205,
+      "grad_norm": 1.1985487937927246,
+      "learning_rate": 0.00013367248451236457,
+      "loss": 2.6033,
+      "step": 6357
+    },
+    {
+      "epoch": 0.39111712598425197,
+      "grad_norm": 1.3343377113342285,
+      "learning_rate": 0.00013365427550329875,
+      "loss": 2.3314,
+      "step": 6358
+    },
+    {
+      "epoch": 0.39117864173228345,
+      "grad_norm": 1.2084442377090454,
+      "learning_rate": 0.00013363606523575077,
+      "loss": 3.3977,
+      "step": 6359
+    },
+    {
+      "epoch": 0.391240157480315,
+      "grad_norm": 1.2804722785949707,
+      "learning_rate": 0.00013361785371040146,
+      "loss": 2.8501,
+      "step": 6360
+    },
+    {
+      "epoch": 0.39130167322834647,
+      "grad_norm": 1.4358011484146118,
+      "learning_rate": 0.00013359964092793188,
+      "loss": 2.7921,
+      "step": 6361
+    },
+    {
+      "epoch": 0.39136318897637795,
+      "grad_norm": 1.3239156007766724,
+      "learning_rate": 0.00013358142688902312,
+      "loss": 2.589,
+      "step": 6362
+    },
+    {
+      "epoch": 0.39142470472440943,
+      "grad_norm": 1.265500545501709,
+      "learning_rate": 0.00013356321159435622,
+      "loss": 2.6918,
+      "step": 6363
+    },
+    {
+      "epoch": 0.39148622047244097,
+      "grad_norm": 1.213341236114502,
+      "learning_rate": 0.00013354499504461236,
+      "loss": 2.4508,
+      "step": 6364
+    },
+    {
+      "epoch": 0.39154773622047245,
+      "grad_norm": 1.1525547504425049,
+      "learning_rate": 0.00013352677724047274,
+      "loss": 2.983,
+      "step": 6365
+    },
+    {
+      "epoch": 0.39160925196850394,
+      "grad_norm": 1.5728507041931152,
+      "learning_rate": 0.0001335085581826186,
+      "loss": 2.3393,
+      "step": 6366
+    },
+    {
+      "epoch": 0.3916707677165354,
+      "grad_norm": 1.475266456604004,
+      "learning_rate": 0.00013349033787173124,
+      "loss": 2.1596,
+      "step": 6367
+    },
+    {
+      "epoch": 0.39173228346456695,
+      "grad_norm": 3.9797325134277344,
+      "learning_rate": 0.00013347211630849198,
+      "loss": 2.8732,
+      "step": 6368
+    },
+    {
+      "epoch": 0.39179379921259844,
+      "grad_norm": 1.19599187374115,
+      "learning_rate": 0.00013345389349358225,
+      "loss": 2.7259,
+      "step": 6369
+    },
+    {
+      "epoch": 0.3918553149606299,
+      "grad_norm": 1.1403005123138428,
+      "learning_rate": 0.00013343566942768342,
+      "loss": 2.8506,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3919168307086614,
+      "grad_norm": 1.1315650939941406,
+      "learning_rate": 0.000133417444111477,
+      "loss": 2.7835,
+      "step": 6371
+    },
+    {
+      "epoch": 0.39197834645669294,
+      "grad_norm": 1.0632174015045166,
+      "learning_rate": 0.00013339921754564448,
+      "loss": 2.6621,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3920398622047244,
+      "grad_norm": 1.1491283178329468,
+      "learning_rate": 0.0001333809897308675,
+      "loss": 2.414,
+      "step": 6373
+    },
+    {
+      "epoch": 0.3921013779527559,
+      "grad_norm": 1.1117137670516968,
+      "learning_rate": 0.00013336276066782762,
+      "loss": 3.0306,
+      "step": 6374
+    },
+    {
+      "epoch": 0.3921628937007874,
+      "grad_norm": 1.4499375820159912,
+      "learning_rate": 0.00013334453035720656,
+      "loss": 1.6181,
+      "step": 6375
+    },
+    {
+      "epoch": 0.3922244094488189,
+      "grad_norm": 1.0295602083206177,
+      "learning_rate": 0.000133326298799686,
+      "loss": 3.123,
+      "step": 6376
+    },
+    {
+      "epoch": 0.3922859251968504,
+      "grad_norm": 0.9873502850532532,
+      "learning_rate": 0.00013330806599594764,
+      "loss": 2.7394,
+      "step": 6377
+    },
+    {
+      "epoch": 0.3923474409448819,
+      "grad_norm": 1.0695582628250122,
+      "learning_rate": 0.0001332898319466734,
+      "loss": 2.5661,
+      "step": 6378
+    },
+    {
+      "epoch": 0.39240895669291337,
+      "grad_norm": 1.3478389978408813,
+      "learning_rate": 0.00013327159665254504,
+      "loss": 2.8956,
+      "step": 6379
+    },
+    {
+      "epoch": 0.3924704724409449,
+      "grad_norm": 1.2562843561172485,
+      "learning_rate": 0.00013325336011424453,
+      "loss": 3.0058,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3925319881889764,
+      "grad_norm": 1.074967384338379,
+      "learning_rate": 0.00013323512233245373,
+      "loss": 3.1097,
+      "step": 6381
+    },
+    {
+      "epoch": 0.39259350393700787,
+      "grad_norm": 1.335099697113037,
+      "learning_rate": 0.00013321688330785472,
+      "loss": 2.9214,
+      "step": 6382
+    },
+    {
+      "epoch": 0.39265501968503935,
+      "grad_norm": 1.2457351684570312,
+      "learning_rate": 0.00013319864304112947,
+      "loss": 2.4209,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3927165354330709,
+      "grad_norm": 1.182144045829773,
+      "learning_rate": 0.0001331804015329601,
+      "loss": 3.2536,
+      "step": 6384
+    },
+    {
+      "epoch": 0.3927780511811024,
+      "grad_norm": 1.168448805809021,
+      "learning_rate": 0.00013316215878402874,
+      "loss": 2.9065,
+      "step": 6385
+    },
+    {
+      "epoch": 0.39283956692913385,
+      "grad_norm": 1.1965535879135132,
+      "learning_rate": 0.00013314391479501752,
+      "loss": 3.1455,
+      "step": 6386
+    },
+    {
+      "epoch": 0.39290108267716534,
+      "grad_norm": 1.1517541408538818,
+      "learning_rate": 0.00013312566956660875,
+      "loss": 2.6292,
+      "step": 6387
+    },
+    {
+      "epoch": 0.3929625984251969,
+      "grad_norm": 0.9913952350616455,
+      "learning_rate": 0.00013310742309948466,
+      "loss": 3.092,
+      "step": 6388
+    },
+    {
+      "epoch": 0.39302411417322836,
+      "grad_norm": 1.1563293933868408,
+      "learning_rate": 0.00013308917539432753,
+      "loss": 2.2537,
+      "step": 6389
+    },
+    {
+      "epoch": 0.39308562992125984,
+      "grad_norm": 1.3809019327163696,
+      "learning_rate": 0.00013307092645181975,
+      "loss": 2.851,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3931471456692913,
+      "grad_norm": 1.2144056558609009,
+      "learning_rate": 0.00013305267627264372,
+      "loss": 2.5182,
+      "step": 6391
+    },
+    {
+      "epoch": 0.39320866141732286,
+      "grad_norm": 1.26202392578125,
+      "learning_rate": 0.00013303442485748193,
+      "loss": 3.1481,
+      "step": 6392
+    },
+    {
+      "epoch": 0.39327017716535434,
+      "grad_norm": 1.098955512046814,
+      "learning_rate": 0.00013301617220701684,
+      "loss": 3.2552,
+      "step": 6393
+    },
+    {
+      "epoch": 0.3933316929133858,
+      "grad_norm": 1.1577693223953247,
+      "learning_rate": 0.00013299791832193102,
+      "loss": 2.5377,
+      "step": 6394
+    },
+    {
+      "epoch": 0.3933932086614173,
+      "grad_norm": 1.0300003290176392,
+      "learning_rate": 0.00013297966320290704,
+      "loss": 3.2447,
+      "step": 6395
+    },
+    {
+      "epoch": 0.39345472440944884,
+      "grad_norm": 1.2467966079711914,
+      "learning_rate": 0.00013296140685062756,
+      "loss": 2.4848,
+      "step": 6396
+    },
+    {
+      "epoch": 0.3935162401574803,
+      "grad_norm": 1.1325417757034302,
+      "learning_rate": 0.00013294314926577528,
+      "loss": 1.9528,
+      "step": 6397
+    },
+    {
+      "epoch": 0.3935777559055118,
+      "grad_norm": 1.416006088256836,
+      "learning_rate": 0.00013292489044903293,
+      "loss": 2.1273,
+      "step": 6398
+    },
+    {
+      "epoch": 0.3936392716535433,
+      "grad_norm": 1.102643609046936,
+      "learning_rate": 0.0001329066304010832,
+      "loss": 2.4977,
+      "step": 6399
+    },
+    {
+      "epoch": 0.3937007874015748,
+      "grad_norm": 1.1197993755340576,
+      "learning_rate": 0.00013288836912260906,
+      "loss": 2.7661,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3937623031496063,
+      "grad_norm": 1.114104986190796,
+      "learning_rate": 0.00013287010661429326,
+      "loss": 3.0094,
+      "step": 6401
+    },
+    {
+      "epoch": 0.3938238188976378,
+      "grad_norm": 1.1608916521072388,
+      "learning_rate": 0.0001328518428768188,
+      "loss": 3.0083,
+      "step": 6402
+    },
+    {
+      "epoch": 0.39388533464566927,
+      "grad_norm": 1.1948562860488892,
+      "learning_rate": 0.00013283357791086855,
+      "loss": 2.4776,
+      "step": 6403
+    },
+    {
+      "epoch": 0.3939468503937008,
+      "grad_norm": 1.4539536237716675,
+      "learning_rate": 0.0001328153117171256,
+      "loss": 2.2011,
+      "step": 6404
+    },
+    {
+      "epoch": 0.3940083661417323,
+      "grad_norm": 1.1894135475158691,
+      "learning_rate": 0.00013279704429627299,
+      "loss": 2.4623,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3940698818897638,
+      "grad_norm": 1.0555897951126099,
+      "learning_rate": 0.00013277877564899377,
+      "loss": 2.6781,
+      "step": 6406
+    },
+    {
+      "epoch": 0.39413139763779526,
+      "grad_norm": 1.148749589920044,
+      "learning_rate": 0.00013276050577597115,
+      "loss": 3.1986,
+      "step": 6407
+    },
+    {
+      "epoch": 0.3941929133858268,
+      "grad_norm": 1.141707181930542,
+      "learning_rate": 0.0001327422346778883,
+      "loss": 2.3356,
+      "step": 6408
+    },
+    {
+      "epoch": 0.3942544291338583,
+      "grad_norm": 1.0810202360153198,
+      "learning_rate": 0.00013272396235542838,
+      "loss": 2.4722,
+      "step": 6409
+    },
+    {
+      "epoch": 0.39431594488188976,
+      "grad_norm": 1.1728975772857666,
+      "learning_rate": 0.0001327056888092748,
+      "loss": 2.2004,
+      "step": 6410
+    },
+    {
+      "epoch": 0.39437746062992124,
+      "grad_norm": 1.1826364994049072,
+      "learning_rate": 0.00013268741404011086,
+      "loss": 2.5934,
+      "step": 6411
+    },
+    {
+      "epoch": 0.3944389763779528,
+      "grad_norm": 1.2080717086791992,
+      "learning_rate": 0.00013266913804861984,
+      "loss": 3.2692,
+      "step": 6412
+    },
+    {
+      "epoch": 0.39450049212598426,
+      "grad_norm": 1.5751785039901733,
+      "learning_rate": 0.00013265086083548525,
+      "loss": 2.4947,
+      "step": 6413
+    },
+    {
+      "epoch": 0.39456200787401574,
+      "grad_norm": 1.1969319581985474,
+      "learning_rate": 0.00013263258240139053,
+      "loss": 2.6427,
+      "step": 6414
+    },
+    {
+      "epoch": 0.3946235236220472,
+      "grad_norm": 1.2061444520950317,
+      "learning_rate": 0.00013261430274701921,
+      "loss": 2.6979,
+      "step": 6415
+    },
+    {
+      "epoch": 0.39468503937007876,
+      "grad_norm": 1.2731539011001587,
+      "learning_rate": 0.00013259602187305482,
+      "loss": 2.5837,
+      "step": 6416
+    },
+    {
+      "epoch": 0.39474655511811024,
+      "grad_norm": 1.4690423011779785,
+      "learning_rate": 0.00013257773978018098,
+      "loss": 2.1731,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3948080708661417,
+      "grad_norm": 1.1020780801773071,
+      "learning_rate": 0.0001325594564690813,
+      "loss": 2.9015,
+      "step": 6418
+    },
+    {
+      "epoch": 0.3948695866141732,
+      "grad_norm": 1.1283117532730103,
+      "learning_rate": 0.00013254117194043952,
+      "loss": 2.4619,
+      "step": 6419
+    },
+    {
+      "epoch": 0.39493110236220474,
+      "grad_norm": 1.1527462005615234,
+      "learning_rate": 0.0001325228861949394,
+      "loss": 2.5234,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3949926181102362,
+      "grad_norm": 1.3457953929901123,
+      "learning_rate": 0.00013250459923326463,
+      "loss": 2.161,
+      "step": 6421
+    },
+    {
+      "epoch": 0.3950541338582677,
+      "grad_norm": 1.2901853322982788,
+      "learning_rate": 0.0001324863110560991,
+      "loss": 2.3689,
+      "step": 6422
+    },
+    {
+      "epoch": 0.3951156496062992,
+      "grad_norm": 1.2077940702438354,
+      "learning_rate": 0.00013246802166412672,
+      "loss": 3.0697,
+      "step": 6423
+    },
+    {
+      "epoch": 0.39517716535433073,
+      "grad_norm": 1.179761290550232,
+      "learning_rate": 0.0001324497310580314,
+      "loss": 2.6042,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3952386811023622,
+      "grad_norm": 1.012677788734436,
+      "learning_rate": 0.00013243143923849705,
+      "loss": 3.3456,
+      "step": 6425
+    },
+    {
+      "epoch": 0.3953001968503937,
+      "grad_norm": 0.9602512717247009,
+      "learning_rate": 0.0001324131462062077,
+      "loss": 3.3383,
+      "step": 6426
+    },
+    {
+      "epoch": 0.3953617125984252,
+      "grad_norm": 1.1417088508605957,
+      "learning_rate": 0.00013239485196184746,
+      "loss": 2.9884,
+      "step": 6427
+    },
+    {
+      "epoch": 0.3954232283464567,
+      "grad_norm": 1.0577692985534668,
+      "learning_rate": 0.00013237655650610034,
+      "loss": 3.2225,
+      "step": 6428
+    },
+    {
+      "epoch": 0.3954847440944882,
+      "grad_norm": 1.1149462461471558,
+      "learning_rate": 0.0001323582598396506,
+      "loss": 2.7121,
+      "step": 6429
+    },
+    {
+      "epoch": 0.3955462598425197,
+      "grad_norm": 1.6230721473693848,
+      "learning_rate": 0.00013233996196318235,
+      "loss": 2.1871,
+      "step": 6430
+    },
+    {
+      "epoch": 0.39560777559055116,
+      "grad_norm": 1.1378705501556396,
+      "learning_rate": 0.00013232166287737983,
+      "loss": 2.645,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3956692913385827,
+      "grad_norm": 1.0518540143966675,
+      "learning_rate": 0.0001323033625829274,
+      "loss": 2.6546,
+      "step": 6432
+    },
+    {
+      "epoch": 0.3957308070866142,
+      "grad_norm": 1.3892381191253662,
+      "learning_rate": 0.0001322850610805093,
+      "loss": 2.9644,
+      "step": 6433
+    },
+    {
+      "epoch": 0.39579232283464566,
+      "grad_norm": 1.0467280149459839,
+      "learning_rate": 0.00013226675837080997,
+      "loss": 2.8921,
+      "step": 6434
+    },
+    {
+      "epoch": 0.39585383858267714,
+      "grad_norm": 1.2269375324249268,
+      "learning_rate": 0.0001322484544545138,
+      "loss": 2.2651,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3959153543307087,
+      "grad_norm": 1.2481954097747803,
+      "learning_rate": 0.00013223014933230523,
+      "loss": 1.9135,
+      "step": 6436
+    },
+    {
+      "epoch": 0.39597687007874016,
+      "grad_norm": 1.060167670249939,
+      "learning_rate": 0.00013221184300486882,
+      "loss": 2.2057,
+      "step": 6437
+    },
+    {
+      "epoch": 0.39603838582677164,
+      "grad_norm": 1.3830407857894897,
+      "learning_rate": 0.00013219353547288908,
+      "loss": 2.2265,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3960999015748031,
+      "grad_norm": 1.1976724863052368,
+      "learning_rate": 0.00013217522673705063,
+      "loss": 2.6683,
+      "step": 6439
+    },
+    {
+      "epoch": 0.39616141732283466,
+      "grad_norm": 1.1082402467727661,
+      "learning_rate": 0.00013215691679803815,
+      "loss": 2.7737,
+      "step": 6440
+    },
+    {
+      "epoch": 0.39622293307086615,
+      "grad_norm": 0.9668769240379333,
+      "learning_rate": 0.00013213860565653625,
+      "loss": 3.2152,
+      "step": 6441
+    },
+    {
+      "epoch": 0.3962844488188976,
+      "grad_norm": 1.2569506168365479,
+      "learning_rate": 0.00013212029331322972,
+      "loss": 2.5205,
+      "step": 6442
+    },
+    {
+      "epoch": 0.3963459645669291,
+      "grad_norm": 1.1582003831863403,
+      "learning_rate": 0.00013210197976880333,
+      "loss": 2.5206,
+      "step": 6443
+    },
+    {
+      "epoch": 0.39640748031496065,
+      "grad_norm": 1.5741920471191406,
+      "learning_rate": 0.00013208366502394188,
+      "loss": 2.252,
+      "step": 6444
+    },
+    {
+      "epoch": 0.39646899606299213,
+      "grad_norm": 1.1761246919631958,
+      "learning_rate": 0.0001320653490793303,
+      "loss": 2.7724,
+      "step": 6445
+    },
+    {
+      "epoch": 0.3965305118110236,
+      "grad_norm": 1.127431035041809,
+      "learning_rate": 0.00013204703193565347,
+      "loss": 2.8515,
+      "step": 6446
+    },
+    {
+      "epoch": 0.3965920275590551,
+      "grad_norm": 1.5182491540908813,
+      "learning_rate": 0.00013202871359359629,
+      "loss": 2.5406,
+      "step": 6447
+    },
+    {
+      "epoch": 0.39665354330708663,
+      "grad_norm": 1.2019810676574707,
+      "learning_rate": 0.00013201039405384387,
+      "loss": 2.3189,
+      "step": 6448
+    },
+    {
+      "epoch": 0.3967150590551181,
+      "grad_norm": 1.0966938734054565,
+      "learning_rate": 0.0001319920733170812,
+      "loss": 2.6545,
+      "step": 6449
+    },
+    {
+      "epoch": 0.3967765748031496,
+      "grad_norm": 1.290155291557312,
+      "learning_rate": 0.0001319737513839934,
+      "loss": 2.3305,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3968380905511811,
+      "grad_norm": 1.4828225374221802,
+      "learning_rate": 0.0001319554282552655,
+      "loss": 2.4109,
+      "step": 6451
+    },
+    {
+      "epoch": 0.3968996062992126,
+      "grad_norm": 1.1077861785888672,
+      "learning_rate": 0.00013193710393158287,
+      "loss": 2.855,
+      "step": 6452
+    },
+    {
+      "epoch": 0.3969611220472441,
+      "grad_norm": 1.2893933057785034,
+      "learning_rate": 0.0001319187784136306,
+      "loss": 2.7039,
+      "step": 6453
+    },
+    {
+      "epoch": 0.3970226377952756,
+      "grad_norm": 1.240809679031372,
+      "learning_rate": 0.000131900451702094,
+      "loss": 2.7683,
+      "step": 6454
+    },
+    {
+      "epoch": 0.39708415354330706,
+      "grad_norm": 1.4413481950759888,
+      "learning_rate": 0.0001318821237976584,
+      "loss": 2.2008,
+      "step": 6455
+    },
+    {
+      "epoch": 0.3971456692913386,
+      "grad_norm": 1.0853023529052734,
+      "learning_rate": 0.00013186379470100915,
+      "loss": 3.3276,
+      "step": 6456
+    },
+    {
+      "epoch": 0.3972071850393701,
+      "grad_norm": 1.2685980796813965,
+      "learning_rate": 0.00013184546441283163,
+      "loss": 2.7829,
+      "step": 6457
+    },
+    {
+      "epoch": 0.39726870078740156,
+      "grad_norm": 1.1792718172073364,
+      "learning_rate": 0.00013182713293381133,
+      "loss": 2.5623,
+      "step": 6458
+    },
+    {
+      "epoch": 0.39733021653543305,
+      "grad_norm": 1.387479543685913,
+      "learning_rate": 0.00013180880026463374,
+      "loss": 2.6489,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3973917322834646,
+      "grad_norm": 1.418769359588623,
+      "learning_rate": 0.00013179046640598437,
+      "loss": 2.7415,
+      "step": 6460
+    },
+    {
+      "epoch": 0.39745324803149606,
+      "grad_norm": 1.0806790590286255,
+      "learning_rate": 0.00013177213135854887,
+      "loss": 3.0288,
+      "step": 6461
+    },
+    {
+      "epoch": 0.39751476377952755,
+      "grad_norm": 1.1316968202590942,
+      "learning_rate": 0.0001317537951230128,
+      "loss": 2.8091,
+      "step": 6462
+    },
+    {
+      "epoch": 0.39757627952755903,
+      "grad_norm": 1.156412959098816,
+      "learning_rate": 0.00013173545770006184,
+      "loss": 2.2898,
+      "step": 6463
+    },
+    {
+      "epoch": 0.39763779527559057,
+      "grad_norm": 1.1408354043960571,
+      "learning_rate": 0.00013171711909038173,
+      "loss": 2.1613,
+      "step": 6464
+    },
+    {
+      "epoch": 0.39769931102362205,
+      "grad_norm": 0.9243590235710144,
+      "learning_rate": 0.00013169877929465824,
+      "loss": 3.0611,
+      "step": 6465
+    },
+    {
+      "epoch": 0.39776082677165353,
+      "grad_norm": 1.4723342657089233,
+      "learning_rate": 0.00013168043831357715,
+      "loss": 2.4351,
+      "step": 6466
+    },
+    {
+      "epoch": 0.397822342519685,
+      "grad_norm": 1.2676960229873657,
+      "learning_rate": 0.00013166209614782432,
+      "loss": 2.8519,
+      "step": 6467
+    },
+    {
+      "epoch": 0.39788385826771655,
+      "grad_norm": 0.984671413898468,
+      "learning_rate": 0.00013164375279808564,
+      "loss": 3.1102,
+      "step": 6468
+    },
+    {
+      "epoch": 0.39794537401574803,
+      "grad_norm": 1.1413205862045288,
+      "learning_rate": 0.00013162540826504706,
+      "loss": 2.9538,
+      "step": 6469
+    },
+    {
+      "epoch": 0.3980068897637795,
+      "grad_norm": 1.3900903463363647,
+      "learning_rate": 0.00013160706254939457,
+      "loss": 2.4078,
+      "step": 6470
+    },
+    {
+      "epoch": 0.398068405511811,
+      "grad_norm": 1.206387996673584,
+      "learning_rate": 0.00013158871565181418,
+      "loss": 3.0732,
+      "step": 6471
+    },
+    {
+      "epoch": 0.39812992125984253,
+      "grad_norm": 1.1178852319717407,
+      "learning_rate": 0.000131570367572992,
+      "loss": 2.4688,
+      "step": 6472
+    },
+    {
+      "epoch": 0.398191437007874,
+      "grad_norm": 0.9662019610404968,
+      "learning_rate": 0.00013155201831361406,
+      "loss": 2.7711,
+      "step": 6473
+    },
+    {
+      "epoch": 0.3982529527559055,
+      "grad_norm": 1.478756308555603,
+      "learning_rate": 0.0001315336678743666,
+      "loss": 2.8635,
+      "step": 6474
+    },
+    {
+      "epoch": 0.398314468503937,
+      "grad_norm": 1.1803065538406372,
+      "learning_rate": 0.00013151531625593582,
+      "loss": 3.0459,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3983759842519685,
+      "grad_norm": 1.4492402076721191,
+      "learning_rate": 0.00013149696345900788,
+      "loss": 3.3378,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3984375,
+      "grad_norm": 2.775594711303711,
+      "learning_rate": 0.00013147860948426918,
+      "loss": 2.1252,
+      "step": 6477
+    },
+    {
+      "epoch": 0.3984990157480315,
+      "grad_norm": 1.5002890825271606,
+      "learning_rate": 0.00013146025433240604,
+      "loss": 2.4157,
+      "step": 6478
+    },
+    {
+      "epoch": 0.398560531496063,
+      "grad_norm": 1.159711241722107,
+      "learning_rate": 0.00013144189800410477,
+      "loss": 2.5513,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3986220472440945,
+      "grad_norm": 1.2590187788009644,
+      "learning_rate": 0.00013142354050005187,
+      "loss": 2.9428,
+      "step": 6480
+    },
+    {
+      "epoch": 0.398683562992126,
+      "grad_norm": 1.278133749961853,
+      "learning_rate": 0.00013140518182093376,
+      "loss": 2.6438,
+      "step": 6481
+    },
+    {
+      "epoch": 0.39874507874015747,
+      "grad_norm": 1.2202856540679932,
+      "learning_rate": 0.00013138682196743698,
+      "loss": 2.0449,
+      "step": 6482
+    },
+    {
+      "epoch": 0.398806594488189,
+      "grad_norm": 1.2411651611328125,
+      "learning_rate": 0.00013136846094024805,
+      "loss": 2.6459,
+      "step": 6483
+    },
+    {
+      "epoch": 0.3988681102362205,
+      "grad_norm": 1.3978685140609741,
+      "learning_rate": 0.00013135009874005364,
+      "loss": 2.353,
+      "step": 6484
+    },
+    {
+      "epoch": 0.39892962598425197,
+      "grad_norm": 1.0247361660003662,
+      "learning_rate": 0.00013133173536754032,
+      "loss": 2.6413,
+      "step": 6485
+    },
+    {
+      "epoch": 0.39899114173228345,
+      "grad_norm": 1.0229452848434448,
+      "learning_rate": 0.0001313133708233948,
+      "loss": 2.9032,
+      "step": 6486
+    },
+    {
+      "epoch": 0.399052657480315,
+      "grad_norm": 1.3141952753067017,
+      "learning_rate": 0.00013129500510830386,
+      "loss": 2.6962,
+      "step": 6487
+    },
+    {
+      "epoch": 0.39911417322834647,
+      "grad_norm": 1.0688655376434326,
+      "learning_rate": 0.00013127663822295422,
+      "loss": 2.6661,
+      "step": 6488
+    },
+    {
+      "epoch": 0.39917568897637795,
+      "grad_norm": 1.135474443435669,
+      "learning_rate": 0.00013125827016803268,
+      "loss": 1.9985,
+      "step": 6489
+    },
+    {
+      "epoch": 0.39923720472440943,
+      "grad_norm": 1.867540955543518,
+      "learning_rate": 0.0001312399009442262,
+      "loss": 2.1353,
+      "step": 6490
+    },
+    {
+      "epoch": 0.39929872047244097,
+      "grad_norm": 2.358492374420166,
+      "learning_rate": 0.00013122153055222156,
+      "loss": 2.247,
+      "step": 6491
+    },
+    {
+      "epoch": 0.39936023622047245,
+      "grad_norm": 1.251294493675232,
+      "learning_rate": 0.00013120315899270578,
+      "loss": 2.8559,
+      "step": 6492
+    },
+    {
+      "epoch": 0.39942175196850394,
+      "grad_norm": 1.1665611267089844,
+      "learning_rate": 0.00013118478626636588,
+      "loss": 2.7324,
+      "step": 6493
+    },
+    {
+      "epoch": 0.3994832677165354,
+      "grad_norm": 1.2105861902236938,
+      "learning_rate": 0.00013116641237388886,
+      "loss": 2.7267,
+      "step": 6494
+    },
+    {
+      "epoch": 0.39954478346456695,
+      "grad_norm": 1.2504451274871826,
+      "learning_rate": 0.00013114803731596177,
+      "loss": 2.592,
+      "step": 6495
+    },
+    {
+      "epoch": 0.39960629921259844,
+      "grad_norm": 1.0320755243301392,
+      "learning_rate": 0.0001311296610932718,
+      "loss": 2.6848,
+      "step": 6496
+    },
+    {
+      "epoch": 0.3996678149606299,
+      "grad_norm": 1.125938057899475,
+      "learning_rate": 0.0001311112837065061,
+      "loss": 2.5425,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3997293307086614,
+      "grad_norm": 1.8002840280532837,
+      "learning_rate": 0.00013109290515635186,
+      "loss": 3.5185,
+      "step": 6498
+    },
+    {
+      "epoch": 0.39979084645669294,
+      "grad_norm": 2.117408275604248,
+      "learning_rate": 0.00013107452544349634,
+      "loss": 2.0328,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3998523622047244,
+      "grad_norm": 1.341422200202942,
+      "learning_rate": 0.00013105614456862687,
+      "loss": 2.5148,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3999138779527559,
+      "grad_norm": 1.0982285737991333,
+      "learning_rate": 0.00013103776253243072,
+      "loss": 2.6488,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3999753937007874,
+      "grad_norm": 1.2407125234603882,
+      "learning_rate": 0.00013101937933559534,
+      "loss": 2.2321,
+      "step": 6502
+    },
+    {
+      "epoch": 0.4000369094488189,
+      "grad_norm": 1.0864880084991455,
+      "learning_rate": 0.00013100099497880815,
+      "loss": 2.9606,
+      "step": 6503
+    },
+    {
+      "epoch": 0.4000984251968504,
+      "grad_norm": 1.6242647171020508,
+      "learning_rate": 0.00013098260946275666,
+      "loss": 2.7567,
+      "step": 6504
+    },
+    {
+      "epoch": 0.4001599409448819,
+      "grad_norm": 1.759515643119812,
+      "learning_rate": 0.00013096422278812827,
+      "loss": 1.3752,
+      "step": 6505
+    },
+    {
+      "epoch": 0.40022145669291337,
+      "grad_norm": 1.130091667175293,
+      "learning_rate": 0.0001309458349556106,
+      "loss": 2.9247,
+      "step": 6506
+    },
+    {
+      "epoch": 0.4002829724409449,
+      "grad_norm": 1.3961849212646484,
+      "learning_rate": 0.0001309274459658913,
+      "loss": 2.4172,
+      "step": 6507
+    },
+    {
+      "epoch": 0.4003444881889764,
+      "grad_norm": 1.217262864112854,
+      "learning_rate": 0.00013090905581965797,
+      "loss": 2.8451,
+      "step": 6508
+    },
+    {
+      "epoch": 0.40040600393700787,
+      "grad_norm": 1.2774074077606201,
+      "learning_rate": 0.0001308906645175983,
+      "loss": 2.4947,
+      "step": 6509
+    },
+    {
+      "epoch": 0.40046751968503935,
+      "grad_norm": 1.1462191343307495,
+      "learning_rate": 0.00013087227206040004,
+      "loss": 2.6161,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4005290354330709,
+      "grad_norm": 1.4048763513565063,
+      "learning_rate": 0.00013085387844875094,
+      "loss": 2.11,
+      "step": 6511
+    },
+    {
+      "epoch": 0.4005905511811024,
+      "grad_norm": 1.0892643928527832,
+      "learning_rate": 0.00013083548368333884,
+      "loss": 1.8393,
+      "step": 6512
+    },
+    {
+      "epoch": 0.40065206692913385,
+      "grad_norm": 1.2532272338867188,
+      "learning_rate": 0.0001308170877648516,
+      "loss": 2.6723,
+      "step": 6513
+    },
+    {
+      "epoch": 0.40071358267716534,
+      "grad_norm": 1.06470787525177,
+      "learning_rate": 0.0001307986906939771,
+      "loss": 2.2954,
+      "step": 6514
+    },
+    {
+      "epoch": 0.4007750984251969,
+      "grad_norm": 1.0961564779281616,
+      "learning_rate": 0.00013078029247140332,
+      "loss": 2.1019,
+      "step": 6515
+    },
+    {
+      "epoch": 0.40083661417322836,
+      "grad_norm": 1.0018421411514282,
+      "learning_rate": 0.00013076189309781824,
+      "loss": 2.4582,
+      "step": 6516
+    },
+    {
+      "epoch": 0.40089812992125984,
+      "grad_norm": 1.5209519863128662,
+      "learning_rate": 0.0001307434925739099,
+      "loss": 2.1363,
+      "step": 6517
+    },
+    {
+      "epoch": 0.4009596456692913,
+      "grad_norm": 1.42621910572052,
+      "learning_rate": 0.00013072509090036633,
+      "loss": 2.4002,
+      "step": 6518
+    },
+    {
+      "epoch": 0.40102116141732286,
+      "grad_norm": 1.3731610774993896,
+      "learning_rate": 0.00013070668807787573,
+      "loss": 2.0885,
+      "step": 6519
+    },
+    {
+      "epoch": 0.40108267716535434,
+      "grad_norm": 0.9862780570983887,
+      "learning_rate": 0.00013068828410712618,
+      "loss": 2.6128,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4011441929133858,
+      "grad_norm": 1.1091578006744385,
+      "learning_rate": 0.00013066987898880595,
+      "loss": 1.9539,
+      "step": 6521
+    },
+    {
+      "epoch": 0.4012057086614173,
+      "grad_norm": 1.3388612270355225,
+      "learning_rate": 0.00013065147272360328,
+      "loss": 2.3256,
+      "step": 6522
+    },
+    {
+      "epoch": 0.40126722440944884,
+      "grad_norm": 1.2713919878005981,
+      "learning_rate": 0.00013063306531220642,
+      "loss": 2.4635,
+      "step": 6523
+    },
+    {
+      "epoch": 0.4013287401574803,
+      "grad_norm": 1.1900160312652588,
+      "learning_rate": 0.00013061465675530377,
+      "loss": 3.0874,
+      "step": 6524
+    },
+    {
+      "epoch": 0.4013902559055118,
+      "grad_norm": 1.0191236734390259,
+      "learning_rate": 0.0001305962470535836,
+      "loss": 2.817,
+      "step": 6525
+    },
+    {
+      "epoch": 0.4014517716535433,
+      "grad_norm": 1.3875048160552979,
+      "learning_rate": 0.00013057783620773448,
+      "loss": 2.5789,
+      "step": 6526
+    },
+    {
+      "epoch": 0.4015132874015748,
+      "grad_norm": 1.1270359754562378,
+      "learning_rate": 0.00013055942421844475,
+      "loss": 2.8107,
+      "step": 6527
+    },
+    {
+      "epoch": 0.4015748031496063,
+      "grad_norm": 1.2941638231277466,
+      "learning_rate": 0.00013054101108640295,
+      "loss": 2.276,
+      "step": 6528
+    },
+    {
+      "epoch": 0.4016363188976378,
+      "grad_norm": 1.2195637226104736,
+      "learning_rate": 0.00013052259681229766,
+      "loss": 1.6503,
+      "step": 6529
+    },
+    {
+      "epoch": 0.40169783464566927,
+      "grad_norm": 1.2477473020553589,
+      "learning_rate": 0.00013050418139681744,
+      "loss": 2.6131,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4017593503937008,
+      "grad_norm": 1.1942588090896606,
+      "learning_rate": 0.00013048576484065092,
+      "loss": 2.6242,
+      "step": 6531
+    },
+    {
+      "epoch": 0.4018208661417323,
+      "grad_norm": 1.3105266094207764,
+      "learning_rate": 0.00013046734714448682,
+      "loss": 2.5523,
+      "step": 6532
+    },
+    {
+      "epoch": 0.4018823818897638,
+      "grad_norm": 1.5419132709503174,
+      "learning_rate": 0.0001304489283090138,
+      "loss": 2.3581,
+      "step": 6533
+    },
+    {
+      "epoch": 0.40194389763779526,
+      "grad_norm": 1.2973966598510742,
+      "learning_rate": 0.00013043050833492066,
+      "loss": 2.5254,
+      "step": 6534
+    },
+    {
+      "epoch": 0.4020054133858268,
+      "grad_norm": 1.2171814441680908,
+      "learning_rate": 0.0001304120872228962,
+      "loss": 2.6582,
+      "step": 6535
+    },
+    {
+      "epoch": 0.4020669291338583,
+      "grad_norm": 0.9600286483764648,
+      "learning_rate": 0.00013039366497362924,
+      "loss": 3.2975,
+      "step": 6536
+    },
+    {
+      "epoch": 0.40212844488188976,
+      "grad_norm": 1.1188173294067383,
+      "learning_rate": 0.0001303752415878087,
+      "loss": 2.458,
+      "step": 6537
+    },
+    {
+      "epoch": 0.40218996062992124,
+      "grad_norm": 1.0977017879486084,
+      "learning_rate": 0.0001303568170661235,
+      "loss": 2.5845,
+      "step": 6538
+    },
+    {
+      "epoch": 0.4022514763779528,
+      "grad_norm": 1.2233117818832397,
+      "learning_rate": 0.00013033839140926264,
+      "loss": 2.8823,
+      "step": 6539
+    },
+    {
+      "epoch": 0.40231299212598426,
+      "grad_norm": 1.1194524765014648,
+      "learning_rate": 0.00013031996461791506,
+      "loss": 2.0674,
+      "step": 6540
+    },
+    {
+      "epoch": 0.40237450787401574,
+      "grad_norm": 1.3966907262802124,
+      "learning_rate": 0.0001303015366927699,
+      "loss": 2.1901,
+      "step": 6541
+    },
+    {
+      "epoch": 0.4024360236220472,
+      "grad_norm": 1.1549806594848633,
+      "learning_rate": 0.00013028310763451626,
+      "loss": 2.6131,
+      "step": 6542
+    },
+    {
+      "epoch": 0.40249753937007876,
+      "grad_norm": 1.219994306564331,
+      "learning_rate": 0.0001302646774438432,
+      "loss": 2.8881,
+      "step": 6543
+    },
+    {
+      "epoch": 0.40255905511811024,
+      "grad_norm": 1.219987154006958,
+      "learning_rate": 0.00013024624612144,
+      "loss": 1.7001,
+      "step": 6544
+    },
+    {
+      "epoch": 0.4026205708661417,
+      "grad_norm": 1.133165717124939,
+      "learning_rate": 0.00013022781366799587,
+      "loss": 3.1058,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4026820866141732,
+      "grad_norm": 1.0612021684646606,
+      "learning_rate": 0.00013020938008420003,
+      "loss": 2.9492,
+      "step": 6546
+    },
+    {
+      "epoch": 0.40274360236220474,
+      "grad_norm": 1.0946024656295776,
+      "learning_rate": 0.00013019094537074184,
+      "loss": 2.3882,
+      "step": 6547
+    },
+    {
+      "epoch": 0.4028051181102362,
+      "grad_norm": 1.3595072031021118,
+      "learning_rate": 0.00013017250952831067,
+      "loss": 2.6678,
+      "step": 6548
+    },
+    {
+      "epoch": 0.4028666338582677,
+      "grad_norm": 1.1296470165252686,
+      "learning_rate": 0.0001301540725575958,
+      "loss": 2.8742,
+      "step": 6549
+    },
+    {
+      "epoch": 0.4029281496062992,
+      "grad_norm": 1.10514497756958,
+      "learning_rate": 0.00013013563445928687,
+      "loss": 2.5133,
+      "step": 6550
+    },
+    {
+      "epoch": 0.40298966535433073,
+      "grad_norm": 1.42274010181427,
+      "learning_rate": 0.0001301171952340732,
+      "loss": 1.9633,
+      "step": 6551
+    },
+    {
+      "epoch": 0.4030511811023622,
+      "grad_norm": 1.0792776346206665,
+      "learning_rate": 0.0001300987548826444,
+      "loss": 2.9514,
+      "step": 6552
+    },
+    {
+      "epoch": 0.4031126968503937,
+      "grad_norm": 1.331674575805664,
+      "learning_rate": 0.00013008031340568994,
+      "loss": 2.4461,
+      "step": 6553
+    },
+    {
+      "epoch": 0.4031742125984252,
+      "grad_norm": 1.2898112535476685,
+      "learning_rate": 0.00013006187080389953,
+      "loss": 2.3525,
+      "step": 6554
+    },
+    {
+      "epoch": 0.4032357283464567,
+      "grad_norm": 0.9700909852981567,
+      "learning_rate": 0.00013004342707796277,
+      "loss": 3.164,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4032972440944882,
+      "grad_norm": 1.1516762971878052,
+      "learning_rate": 0.00013002498222856935,
+      "loss": 2.3069,
+      "step": 6556
+    },
+    {
+      "epoch": 0.4033587598425197,
+      "grad_norm": 0.9802262187004089,
+      "learning_rate": 0.00013000653625640907,
+      "loss": 2.9818,
+      "step": 6557
+    },
+    {
+      "epoch": 0.40342027559055116,
+      "grad_norm": 1.153828740119934,
+      "learning_rate": 0.0001299880891621716,
+      "loss": 2.5835,
+      "step": 6558
+    },
+    {
+      "epoch": 0.4034817913385827,
+      "grad_norm": 1.260665774345398,
+      "learning_rate": 0.00012996964094654686,
+      "loss": 3.0985,
+      "step": 6559
+    },
+    {
+      "epoch": 0.4035433070866142,
+      "grad_norm": 1.1178867816925049,
+      "learning_rate": 0.00012995119161022465,
+      "loss": 2.7901,
+      "step": 6560
+    },
+    {
+      "epoch": 0.40360482283464566,
+      "grad_norm": 1.3685038089752197,
+      "learning_rate": 0.00012993274115389488,
+      "loss": 2.4274,
+      "step": 6561
+    },
+    {
+      "epoch": 0.40366633858267714,
+      "grad_norm": 1.0914726257324219,
+      "learning_rate": 0.00012991428957824747,
+      "loss": 3.0933,
+      "step": 6562
+    },
+    {
+      "epoch": 0.4037278543307087,
+      "grad_norm": 1.0129648447036743,
+      "learning_rate": 0.00012989583688397246,
+      "loss": 2.312,
+      "step": 6563
+    },
+    {
+      "epoch": 0.40378937007874016,
+      "grad_norm": 1.2528806924819946,
+      "learning_rate": 0.00012987738307175986,
+      "loss": 2.5174,
+      "step": 6564
+    },
+    {
+      "epoch": 0.40385088582677164,
+      "grad_norm": 1.2032320499420166,
+      "learning_rate": 0.00012985892814229975,
+      "loss": 2.0718,
+      "step": 6565
+    },
+    {
+      "epoch": 0.4039124015748031,
+      "grad_norm": 1.316542625427246,
+      "learning_rate": 0.0001298404720962822,
+      "loss": 2.5793,
+      "step": 6566
+    },
+    {
+      "epoch": 0.40397391732283466,
+      "grad_norm": 1.2004907131195068,
+      "learning_rate": 0.00012982201493439742,
+      "loss": 2.5959,
+      "step": 6567
+    },
+    {
+      "epoch": 0.40403543307086615,
+      "grad_norm": 1.550864815711975,
+      "learning_rate": 0.00012980355665733555,
+      "loss": 2.764,
+      "step": 6568
+    },
+    {
+      "epoch": 0.4040969488188976,
+      "grad_norm": 1.4404295682907104,
+      "learning_rate": 0.00012978509726578683,
+      "loss": 3.0561,
+      "step": 6569
+    },
+    {
+      "epoch": 0.4041584645669291,
+      "grad_norm": 1.6805381774902344,
+      "learning_rate": 0.0001297666367604416,
+      "loss": 2.0067,
+      "step": 6570
+    },
+    {
+      "epoch": 0.40421998031496065,
+      "grad_norm": 1.0549691915512085,
+      "learning_rate": 0.00012974817514199008,
+      "loss": 2.2852,
+      "step": 6571
+    },
+    {
+      "epoch": 0.40428149606299213,
+      "grad_norm": 1.2243671417236328,
+      "learning_rate": 0.00012972971241112275,
+      "loss": 2.7804,
+      "step": 6572
+    },
+    {
+      "epoch": 0.4043430118110236,
+      "grad_norm": 1.0815359354019165,
+      "learning_rate": 0.00012971124856852995,
+      "loss": 3.1056,
+      "step": 6573
+    },
+    {
+      "epoch": 0.4044045275590551,
+      "grad_norm": 1.1805957555770874,
+      "learning_rate": 0.00012969278361490212,
+      "loss": 2.2696,
+      "step": 6574
+    },
+    {
+      "epoch": 0.40446604330708663,
+      "grad_norm": 1.253315806388855,
+      "learning_rate": 0.0001296743175509297,
+      "loss": 2.171,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4045275590551181,
+      "grad_norm": 1.1014631986618042,
+      "learning_rate": 0.00012965585037730336,
+      "loss": 2.2951,
+      "step": 6576
+    },
+    {
+      "epoch": 0.4045890748031496,
+      "grad_norm": 1.2247780561447144,
+      "learning_rate": 0.00012963738209471353,
+      "loss": 2.5611,
+      "step": 6577
+    },
+    {
+      "epoch": 0.4046505905511811,
+      "grad_norm": 1.0429021120071411,
+      "learning_rate": 0.00012961891270385086,
+      "loss": 2.5886,
+      "step": 6578
+    },
+    {
+      "epoch": 0.4047121062992126,
+      "grad_norm": 1.088425874710083,
+      "learning_rate": 0.00012960044220540603,
+      "loss": 2.6875,
+      "step": 6579
+    },
+    {
+      "epoch": 0.4047736220472441,
+      "grad_norm": 1.1614575386047363,
+      "learning_rate": 0.00012958197060006972,
+      "loss": 2.1603,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4048351377952756,
+      "grad_norm": 1.2770230770111084,
+      "learning_rate": 0.00012956349788853267,
+      "loss": 2.6695,
+      "step": 6581
+    },
+    {
+      "epoch": 0.40489665354330706,
+      "grad_norm": 1.1968960762023926,
+      "learning_rate": 0.00012954502407148564,
+      "loss": 2.3959,
+      "step": 6582
+    },
+    {
+      "epoch": 0.4049581692913386,
+      "grad_norm": 1.1799652576446533,
+      "learning_rate": 0.00012952654914961947,
+      "loss": 2.1177,
+      "step": 6583
+    },
+    {
+      "epoch": 0.4050196850393701,
+      "grad_norm": 1.0733405351638794,
+      "learning_rate": 0.000129508073123625,
+      "loss": 2.6041,
+      "step": 6584
+    },
+    {
+      "epoch": 0.40508120078740156,
+      "grad_norm": 1.5436042547225952,
+      "learning_rate": 0.00012948959599419312,
+      "loss": 2.3781,
+      "step": 6585
+    },
+    {
+      "epoch": 0.40514271653543305,
+      "grad_norm": 1.4213377237319946,
+      "learning_rate": 0.00012947111776201483,
+      "loss": 1.784,
+      "step": 6586
+    },
+    {
+      "epoch": 0.4052042322834646,
+      "grad_norm": 1.348888874053955,
+      "learning_rate": 0.00012945263842778103,
+      "loss": 2.9221,
+      "step": 6587
+    },
+    {
+      "epoch": 0.40526574803149606,
+      "grad_norm": 1.3313995599746704,
+      "learning_rate": 0.00012943415799218278,
+      "loss": 2.4742,
+      "step": 6588
+    },
+    {
+      "epoch": 0.40532726377952755,
+      "grad_norm": 1.5076372623443604,
+      "learning_rate": 0.0001294156764559112,
+      "loss": 2.7607,
+      "step": 6589
+    },
+    {
+      "epoch": 0.40538877952755903,
+      "grad_norm": 1.0561962127685547,
+      "learning_rate": 0.00012939719381965733,
+      "loss": 3.0593,
+      "step": 6590
+    },
+    {
+      "epoch": 0.40545029527559057,
+      "grad_norm": 2.450873374938965,
+      "learning_rate": 0.00012937871008411229,
+      "loss": 3.9186,
+      "step": 6591
+    },
+    {
+      "epoch": 0.40551181102362205,
+      "grad_norm": 1.6279890537261963,
+      "learning_rate": 0.00012936022524996735,
+      "loss": 1.4911,
+      "step": 6592
+    },
+    {
+      "epoch": 0.40557332677165353,
+      "grad_norm": 1.3286901712417603,
+      "learning_rate": 0.00012934173931791371,
+      "loss": 2.5799,
+      "step": 6593
+    },
+    {
+      "epoch": 0.405634842519685,
+      "grad_norm": 1.1558541059494019,
+      "learning_rate": 0.0001293232522886426,
+      "loss": 2.6443,
+      "step": 6594
+    },
+    {
+      "epoch": 0.40569635826771655,
+      "grad_norm": 1.0619460344314575,
+      "learning_rate": 0.0001293047641628454,
+      "loss": 2.9069,
+      "step": 6595
+    },
+    {
+      "epoch": 0.40575787401574803,
+      "grad_norm": 1.8375333547592163,
+      "learning_rate": 0.00012928627494121343,
+      "loss": 3.6908,
+      "step": 6596
+    },
+    {
+      "epoch": 0.4058193897637795,
+      "grad_norm": 1.8046550750732422,
+      "learning_rate": 0.00012926778462443803,
+      "loss": 2.5258,
+      "step": 6597
+    },
+    {
+      "epoch": 0.405880905511811,
+      "grad_norm": 1.2458773851394653,
+      "learning_rate": 0.00012924929321321072,
+      "loss": 3.2217,
+      "step": 6598
+    },
+    {
+      "epoch": 0.40594242125984253,
+      "grad_norm": 1.2661041021347046,
+      "learning_rate": 0.00012923080070822296,
+      "loss": 3.0661,
+      "step": 6599
+    },
+    {
+      "epoch": 0.406003937007874,
+      "grad_norm": 0.9913474917411804,
+      "learning_rate": 0.00012921230711016623,
+      "loss": 2.8317,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4060654527559055,
+      "grad_norm": 1.077295184135437,
+      "learning_rate": 0.0001291938124197321,
+      "loss": 2.6061,
+      "step": 6601
+    },
+    {
+      "epoch": 0.406126968503937,
+      "grad_norm": 1.126076340675354,
+      "learning_rate": 0.0001291753166376122,
+      "loss": 1.9479,
+      "step": 6602
+    },
+    {
+      "epoch": 0.4061884842519685,
+      "grad_norm": 1.2171235084533691,
+      "learning_rate": 0.00012915681976449813,
+      "loss": 2.6993,
+      "step": 6603
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 1.215165138244629,
+      "learning_rate": 0.00012913832180108157,
+      "loss": 2.491,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4063115157480315,
+      "grad_norm": 1.4639495611190796,
+      "learning_rate": 0.00012911982274805428,
+      "loss": 2.3891,
+      "step": 6605
+    },
+    {
+      "epoch": 0.406373031496063,
+      "grad_norm": 0.8992121815681458,
+      "learning_rate": 0.00012910132260610796,
+      "loss": 3.1227,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4064345472440945,
+      "grad_norm": 1.084008812904358,
+      "learning_rate": 0.00012908282137593448,
+      "loss": 2.8012,
+      "step": 6607
+    },
+    {
+      "epoch": 0.406496062992126,
+      "grad_norm": 1.1637858152389526,
+      "learning_rate": 0.00012906431905822566,
+      "loss": 2.2106,
+      "step": 6608
+    },
+    {
+      "epoch": 0.40655757874015747,
+      "grad_norm": 1.2345281839370728,
+      "learning_rate": 0.00012904581565367337,
+      "loss": 2.4323,
+      "step": 6609
+    },
+    {
+      "epoch": 0.406619094488189,
+      "grad_norm": 1.1088581085205078,
+      "learning_rate": 0.0001290273111629695,
+      "loss": 2.9948,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4066806102362205,
+      "grad_norm": 1.0943801403045654,
+      "learning_rate": 0.0001290088055868061,
+      "loss": 2.8768,
+      "step": 6611
+    },
+    {
+      "epoch": 0.40674212598425197,
+      "grad_norm": 1.1002767086029053,
+      "learning_rate": 0.0001289902989258751,
+      "loss": 3.3016,
+      "step": 6612
+    },
+    {
+      "epoch": 0.40680364173228345,
+      "grad_norm": 1.3504074811935425,
+      "learning_rate": 0.00012897179118086856,
+      "loss": 2.6107,
+      "step": 6613
+    },
+    {
+      "epoch": 0.406865157480315,
+      "grad_norm": 1.2977992296218872,
+      "learning_rate": 0.00012895328235247862,
+      "loss": 2.635,
+      "step": 6614
+    },
+    {
+      "epoch": 0.40692667322834647,
+      "grad_norm": 1.0522536039352417,
+      "learning_rate": 0.00012893477244139737,
+      "loss": 2.5855,
+      "step": 6615
+    },
+    {
+      "epoch": 0.40698818897637795,
+      "grad_norm": 1.2571325302124023,
+      "learning_rate": 0.00012891626144831693,
+      "loss": 1.974,
+      "step": 6616
+    },
+    {
+      "epoch": 0.40704970472440943,
+      "grad_norm": 1.2201308012008667,
+      "learning_rate": 0.00012889774937392958,
+      "loss": 2.1979,
+      "step": 6617
+    },
+    {
+      "epoch": 0.40711122047244097,
+      "grad_norm": 1.1529476642608643,
+      "learning_rate": 0.00012887923621892754,
+      "loss": 3.2229,
+      "step": 6618
+    },
+    {
+      "epoch": 0.40717273622047245,
+      "grad_norm": 1.2680972814559937,
+      "learning_rate": 0.0001288607219840031,
+      "loss": 3.1332,
+      "step": 6619
+    },
+    {
+      "epoch": 0.40723425196850394,
+      "grad_norm": 1.206460952758789,
+      "learning_rate": 0.00012884220666984858,
+      "loss": 3.0086,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4072957677165354,
+      "grad_norm": 1.7531989812850952,
+      "learning_rate": 0.00012882369027715637,
+      "loss": 3.9933,
+      "step": 6621
+    },
+    {
+      "epoch": 0.40735728346456695,
+      "grad_norm": 1.2447428703308105,
+      "learning_rate": 0.00012880517280661888,
+      "loss": 2.2771,
+      "step": 6622
+    },
+    {
+      "epoch": 0.40741879921259844,
+      "grad_norm": 1.3285930156707764,
+      "learning_rate": 0.00012878665425892854,
+      "loss": 2.4127,
+      "step": 6623
+    },
+    {
+      "epoch": 0.4074803149606299,
+      "grad_norm": 1.1181238889694214,
+      "learning_rate": 0.00012876813463477782,
+      "loss": 2.3403,
+      "step": 6624
+    },
+    {
+      "epoch": 0.4075418307086614,
+      "grad_norm": 1.2831430435180664,
+      "learning_rate": 0.00012874961393485933,
+      "loss": 2.4212,
+      "step": 6625
+    },
+    {
+      "epoch": 0.40760334645669294,
+      "grad_norm": 1.1888551712036133,
+      "learning_rate": 0.00012873109215986555,
+      "loss": 2.0268,
+      "step": 6626
+    },
+    {
+      "epoch": 0.4076648622047244,
+      "grad_norm": 1.3071162700653076,
+      "learning_rate": 0.00012871256931048917,
+      "loss": 1.9244,
+      "step": 6627
+    },
+    {
+      "epoch": 0.4077263779527559,
+      "grad_norm": 1.34785795211792,
+      "learning_rate": 0.00012869404538742277,
+      "loss": 2.4579,
+      "step": 6628
+    },
+    {
+      "epoch": 0.4077878937007874,
+      "grad_norm": 1.2152326107025146,
+      "learning_rate": 0.00012867552039135907,
+      "loss": 3.034,
+      "step": 6629
+    },
+    {
+      "epoch": 0.4078494094488189,
+      "grad_norm": 1.489168643951416,
+      "learning_rate": 0.00012865699432299086,
+      "loss": 2.2406,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4079109251968504,
+      "grad_norm": 1.0882667303085327,
+      "learning_rate": 0.0001286384671830108,
+      "loss": 2.4261,
+      "step": 6631
+    },
+    {
+      "epoch": 0.4079724409448819,
+      "grad_norm": 1.2059510946273804,
+      "learning_rate": 0.00012861993897211177,
+      "loss": 2.7258,
+      "step": 6632
+    },
+    {
+      "epoch": 0.40803395669291337,
+      "grad_norm": 1.3245511054992676,
+      "learning_rate": 0.00012860140969098665,
+      "loss": 2.9467,
+      "step": 6633
+    },
+    {
+      "epoch": 0.4080954724409449,
+      "grad_norm": 1.1523886919021606,
+      "learning_rate": 0.00012858287934032828,
+      "loss": 2.7265,
+      "step": 6634
+    },
+    {
+      "epoch": 0.4081569881889764,
+      "grad_norm": 1.1813641786575317,
+      "learning_rate": 0.00012856434792082958,
+      "loss": 2.8201,
+      "step": 6635
+    },
+    {
+      "epoch": 0.40821850393700787,
+      "grad_norm": 1.1105988025665283,
+      "learning_rate": 0.00012854581543318355,
+      "loss": 2.9132,
+      "step": 6636
+    },
+    {
+      "epoch": 0.40828001968503935,
+      "grad_norm": 1.1368381977081299,
+      "learning_rate": 0.0001285272818780832,
+      "loss": 2.7621,
+      "step": 6637
+    },
+    {
+      "epoch": 0.4083415354330709,
+      "grad_norm": 1.3018574714660645,
+      "learning_rate": 0.0001285087472562216,
+      "loss": 2.8555,
+      "step": 6638
+    },
+    {
+      "epoch": 0.4084030511811024,
+      "grad_norm": 1.2077096700668335,
+      "learning_rate": 0.0001284902115682918,
+      "loss": 3.1572,
+      "step": 6639
+    },
+    {
+      "epoch": 0.40846456692913385,
+      "grad_norm": 1.1184600591659546,
+      "learning_rate": 0.00012847167481498697,
+      "loss": 2.8499,
+      "step": 6640
+    },
+    {
+      "epoch": 0.40852608267716534,
+      "grad_norm": 1.190954327583313,
+      "learning_rate": 0.00012845313699700025,
+      "loss": 3.0887,
+      "step": 6641
+    },
+    {
+      "epoch": 0.4085875984251969,
+      "grad_norm": 1.3144118785858154,
+      "learning_rate": 0.00012843459811502487,
+      "loss": 2.4557,
+      "step": 6642
+    },
+    {
+      "epoch": 0.40864911417322836,
+      "grad_norm": 1.3445801734924316,
+      "learning_rate": 0.00012841605816975407,
+      "loss": 2.843,
+      "step": 6643
+    },
+    {
+      "epoch": 0.40871062992125984,
+      "grad_norm": 1.2726503610610962,
+      "learning_rate": 0.00012839751716188117,
+      "loss": 2.251,
+      "step": 6644
+    },
+    {
+      "epoch": 0.4087721456692913,
+      "grad_norm": 1.098514199256897,
+      "learning_rate": 0.0001283789750920995,
+      "loss": 2.9875,
+      "step": 6645
+    },
+    {
+      "epoch": 0.40883366141732286,
+      "grad_norm": 1.2307733297348022,
+      "learning_rate": 0.00012836043196110236,
+      "loss": 2.4026,
+      "step": 6646
+    },
+    {
+      "epoch": 0.40889517716535434,
+      "grad_norm": 1.206615686416626,
+      "learning_rate": 0.0001283418877695832,
+      "loss": 2.6437,
+      "step": 6647
+    },
+    {
+      "epoch": 0.4089566929133858,
+      "grad_norm": 1.0549659729003906,
+      "learning_rate": 0.0001283233425182355,
+      "loss": 2.8201,
+      "step": 6648
+    },
+    {
+      "epoch": 0.4090182086614173,
+      "grad_norm": 1.214282751083374,
+      "learning_rate": 0.00012830479620775273,
+      "loss": 3.4486,
+      "step": 6649
+    },
+    {
+      "epoch": 0.40907972440944884,
+      "grad_norm": 1.160422682762146,
+      "learning_rate": 0.0001282862488388284,
+      "loss": 2.8975,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4091412401574803,
+      "grad_norm": 1.0365824699401855,
+      "learning_rate": 0.0001282677004121561,
+      "loss": 2.2007,
+      "step": 6651
+    },
+    {
+      "epoch": 0.4092027559055118,
+      "grad_norm": 1.0889991521835327,
+      "learning_rate": 0.0001282491509284294,
+      "loss": 2.6256,
+      "step": 6652
+    },
+    {
+      "epoch": 0.4092642716535433,
+      "grad_norm": 1.3083139657974243,
+      "learning_rate": 0.000128230600388342,
+      "loss": 2.3858,
+      "step": 6653
+    },
+    {
+      "epoch": 0.4093257874015748,
+      "grad_norm": 1.16429603099823,
+      "learning_rate": 0.0001282120487925876,
+      "loss": 2.5565,
+      "step": 6654
+    },
+    {
+      "epoch": 0.4093873031496063,
+      "grad_norm": 1.320753812789917,
+      "learning_rate": 0.00012819349614185983,
+      "loss": 2.1906,
+      "step": 6655
+    },
+    {
+      "epoch": 0.4094488188976378,
+      "grad_norm": 1.1657683849334717,
+      "learning_rate": 0.00012817494243685254,
+      "loss": 2.9346,
+      "step": 6656
+    },
+    {
+      "epoch": 0.40951033464566927,
+      "grad_norm": 1.2465006113052368,
+      "learning_rate": 0.00012815638767825951,
+      "loss": 2.572,
+      "step": 6657
+    },
+    {
+      "epoch": 0.4095718503937008,
+      "grad_norm": 1.1282645463943481,
+      "learning_rate": 0.00012813783186677458,
+      "loss": 2.0881,
+      "step": 6658
+    },
+    {
+      "epoch": 0.4096333661417323,
+      "grad_norm": 1.1289640665054321,
+      "learning_rate": 0.00012811927500309164,
+      "loss": 2.5859,
+      "step": 6659
+    },
+    {
+      "epoch": 0.4096948818897638,
+      "grad_norm": 1.6361079216003418,
+      "learning_rate": 0.00012810071708790462,
+      "loss": 2.3866,
+      "step": 6660
+    },
+    {
+      "epoch": 0.40975639763779526,
+      "grad_norm": 1.2252602577209473,
+      "learning_rate": 0.00012808215812190748,
+      "loss": 3.0443,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4098179133858268,
+      "grad_norm": 1.1016401052474976,
+      "learning_rate": 0.0001280635981057942,
+      "loss": 2.1981,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4098794291338583,
+      "grad_norm": 1.157731294631958,
+      "learning_rate": 0.00012804503704025885,
+      "loss": 2.9258,
+      "step": 6663
+    },
+    {
+      "epoch": 0.40994094488188976,
+      "grad_norm": 1.12851083278656,
+      "learning_rate": 0.0001280264749259955,
+      "loss": 2.0845,
+      "step": 6664
+    },
+    {
+      "epoch": 0.41000246062992124,
+      "grad_norm": 1.0908666849136353,
+      "learning_rate": 0.00012800791176369823,
+      "loss": 2.5475,
+      "step": 6665
+    },
+    {
+      "epoch": 0.4100639763779528,
+      "grad_norm": 1.2514358758926392,
+      "learning_rate": 0.00012798934755406128,
+      "loss": 2.8665,
+      "step": 6666
+    },
+    {
+      "epoch": 0.41012549212598426,
+      "grad_norm": 1.228140950202942,
+      "learning_rate": 0.0001279707822977788,
+      "loss": 2.4558,
+      "step": 6667
+    },
+    {
+      "epoch": 0.41018700787401574,
+      "grad_norm": 1.4342073202133179,
+      "learning_rate": 0.00012795221599554504,
+      "loss": 2.4882,
+      "step": 6668
+    },
+    {
+      "epoch": 0.4102485236220472,
+      "grad_norm": 1.037083625793457,
+      "learning_rate": 0.00012793364864805424,
+      "loss": 3.0734,
+      "step": 6669
+    },
+    {
+      "epoch": 0.41031003937007876,
+      "grad_norm": 1.1267977952957153,
+      "learning_rate": 0.00012791508025600078,
+      "loss": 2.4592,
+      "step": 6670
+    },
+    {
+      "epoch": 0.41037155511811024,
+      "grad_norm": 0.9243610501289368,
+      "learning_rate": 0.00012789651082007893,
+      "loss": 2.7165,
+      "step": 6671
+    },
+    {
+      "epoch": 0.4104330708661417,
+      "grad_norm": 1.2092387676239014,
+      "learning_rate": 0.00012787794034098315,
+      "loss": 2.6924,
+      "step": 6672
+    },
+    {
+      "epoch": 0.4104945866141732,
+      "grad_norm": 1.0752989053726196,
+      "learning_rate": 0.00012785936881940785,
+      "loss": 2.8698,
+      "step": 6673
+    },
+    {
+      "epoch": 0.41055610236220474,
+      "grad_norm": 1.2019541263580322,
+      "learning_rate": 0.0001278407962560475,
+      "loss": 2.64,
+      "step": 6674
+    },
+    {
+      "epoch": 0.4106176181102362,
+      "grad_norm": 1.2149814367294312,
+      "learning_rate": 0.00012782222265159663,
+      "loss": 3.0635,
+      "step": 6675
+    },
+    {
+      "epoch": 0.4106791338582677,
+      "grad_norm": 0.9401673674583435,
+      "learning_rate": 0.00012780364800674978,
+      "loss": 2.7472,
+      "step": 6676
+    },
+    {
+      "epoch": 0.4107406496062992,
+      "grad_norm": 1.3106412887573242,
+      "learning_rate": 0.0001277850723222015,
+      "loss": 2.6623,
+      "step": 6677
+    },
+    {
+      "epoch": 0.41080216535433073,
+      "grad_norm": 1.502474308013916,
+      "learning_rate": 0.00012776649559864645,
+      "loss": 2.9547,
+      "step": 6678
+    },
+    {
+      "epoch": 0.4108636811023622,
+      "grad_norm": 1.302619218826294,
+      "learning_rate": 0.00012774791783677934,
+      "loss": 2.6668,
+      "step": 6679
+    },
+    {
+      "epoch": 0.4109251968503937,
+      "grad_norm": 1.3273581266403198,
+      "learning_rate": 0.0001277293390372948,
+      "loss": 2.5875,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4109867125984252,
+      "grad_norm": 1.0041760206222534,
+      "learning_rate": 0.0001277107592008876,
+      "loss": 3.0739,
+      "step": 6681
+    },
+    {
+      "epoch": 0.4110482283464567,
+      "grad_norm": 1.1806504726409912,
+      "learning_rate": 0.00012769217832825255,
+      "loss": 2.3494,
+      "step": 6682
+    },
+    {
+      "epoch": 0.4111097440944882,
+      "grad_norm": 1.2117207050323486,
+      "learning_rate": 0.0001276735964200844,
+      "loss": 2.762,
+      "step": 6683
+    },
+    {
+      "epoch": 0.4111712598425197,
+      "grad_norm": 1.1144672632217407,
+      "learning_rate": 0.0001276550134770781,
+      "loss": 3.1775,
+      "step": 6684
+    },
+    {
+      "epoch": 0.41123277559055116,
+      "grad_norm": 1.329048991203308,
+      "learning_rate": 0.00012763642949992852,
+      "loss": 2.9974,
+      "step": 6685
+    },
+    {
+      "epoch": 0.4112942913385827,
+      "grad_norm": 1.144780158996582,
+      "learning_rate": 0.00012761784448933052,
+      "loss": 2.8545,
+      "step": 6686
+    },
+    {
+      "epoch": 0.4113558070866142,
+      "grad_norm": 2.076564073562622,
+      "learning_rate": 0.00012759925844597919,
+      "loss": 2.4569,
+      "step": 6687
+    },
+    {
+      "epoch": 0.41141732283464566,
+      "grad_norm": 1.4109479188919067,
+      "learning_rate": 0.00012758067137056946,
+      "loss": 2.6644,
+      "step": 6688
+    },
+    {
+      "epoch": 0.41147883858267714,
+      "grad_norm": 1.2545678615570068,
+      "learning_rate": 0.00012756208326379642,
+      "loss": 2.3082,
+      "step": 6689
+    },
+    {
+      "epoch": 0.4115403543307087,
+      "grad_norm": 1.3586299419403076,
+      "learning_rate": 0.00012754349412635516,
+      "loss": 2.4216,
+      "step": 6690
+    },
+    {
+      "epoch": 0.41160187007874016,
+      "grad_norm": 1.0656200647354126,
+      "learning_rate": 0.00012752490395894078,
+      "loss": 2.8111,
+      "step": 6691
+    },
+    {
+      "epoch": 0.41166338582677164,
+      "grad_norm": 1.0217927694320679,
+      "learning_rate": 0.00012750631276224855,
+      "loss": 2.5985,
+      "step": 6692
+    },
+    {
+      "epoch": 0.4117249015748031,
+      "grad_norm": 1.097823977470398,
+      "learning_rate": 0.0001274877205369735,
+      "loss": 2.7543,
+      "step": 6693
+    },
+    {
+      "epoch": 0.41178641732283466,
+      "grad_norm": 1.2026325464248657,
+      "learning_rate": 0.00012746912728381105,
+      "loss": 2.67,
+      "step": 6694
+    },
+    {
+      "epoch": 0.41184793307086615,
+      "grad_norm": 1.1953365802764893,
+      "learning_rate": 0.0001274505330034564,
+      "loss": 2.1969,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4119094488188976,
+      "grad_norm": 1.1549417972564697,
+      "learning_rate": 0.00012743193769660486,
+      "loss": 3.0432,
+      "step": 6696
+    },
+    {
+      "epoch": 0.4119709645669291,
+      "grad_norm": 1.2192004919052124,
+      "learning_rate": 0.00012741334136395183,
+      "loss": 2.2871,
+      "step": 6697
+    },
+    {
+      "epoch": 0.41203248031496065,
+      "grad_norm": 1.2835253477096558,
+      "learning_rate": 0.00012739474400619272,
+      "loss": 2.5514,
+      "step": 6698
+    },
+    {
+      "epoch": 0.41209399606299213,
+      "grad_norm": 1.1784133911132812,
+      "learning_rate": 0.00012737614562402289,
+      "loss": 2.7369,
+      "step": 6699
+    },
+    {
+      "epoch": 0.4121555118110236,
+      "grad_norm": 1.294084906578064,
+      "learning_rate": 0.00012735754621813792,
+      "loss": 2.6827,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4122170275590551,
+      "grad_norm": 1.5118178129196167,
+      "learning_rate": 0.00012733894578923323,
+      "loss": 2.2294,
+      "step": 6701
+    },
+    {
+      "epoch": 0.41227854330708663,
+      "grad_norm": 1.0592111349105835,
+      "learning_rate": 0.00012732034433800445,
+      "loss": 2.7959,
+      "step": 6702
+    },
+    {
+      "epoch": 0.4123400590551181,
+      "grad_norm": 1.2080491781234741,
+      "learning_rate": 0.0001273017418651471,
+      "loss": 2.7858,
+      "step": 6703
+    },
+    {
+      "epoch": 0.4124015748031496,
+      "grad_norm": 1.3254318237304688,
+      "learning_rate": 0.00012728313837135686,
+      "loss": 3.0264,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4124630905511811,
+      "grad_norm": 1.2503247261047363,
+      "learning_rate": 0.00012726453385732936,
+      "loss": 3.0724,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4125246062992126,
+      "grad_norm": 1.3164807558059692,
+      "learning_rate": 0.00012724592832376034,
+      "loss": 2.9412,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4125861220472441,
+      "grad_norm": 1.1205253601074219,
+      "learning_rate": 0.00012722732177134554,
+      "loss": 2.7323,
+      "step": 6707
+    },
+    {
+      "epoch": 0.4126476377952756,
+      "grad_norm": 1.489079236984253,
+      "learning_rate": 0.00012720871420078073,
+      "loss": 2.5777,
+      "step": 6708
+    },
+    {
+      "epoch": 0.41270915354330706,
+      "grad_norm": 1.1784225702285767,
+      "learning_rate": 0.0001271901056127617,
+      "loss": 2.5423,
+      "step": 6709
+    },
+    {
+      "epoch": 0.4127706692913386,
+      "grad_norm": 1.1744532585144043,
+      "learning_rate": 0.00012717149600798434,
+      "loss": 3.3531,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4128321850393701,
+      "grad_norm": 0.9619330763816833,
+      "learning_rate": 0.00012715288538714457,
+      "loss": 3.1632,
+      "step": 6711
+    },
+    {
+      "epoch": 0.41289370078740156,
+      "grad_norm": 1.5881850719451904,
+      "learning_rate": 0.00012713427375093828,
+      "loss": 2.7199,
+      "step": 6712
+    },
+    {
+      "epoch": 0.41295521653543305,
+      "grad_norm": 2.113797664642334,
+      "learning_rate": 0.00012711566110006146,
+      "loss": 2.2658,
+      "step": 6713
+    },
+    {
+      "epoch": 0.4130167322834646,
+      "grad_norm": 1.2829395532608032,
+      "learning_rate": 0.00012709704743521012,
+      "loss": 3.0093,
+      "step": 6714
+    },
+    {
+      "epoch": 0.41307824803149606,
+      "grad_norm": 1.2305134534835815,
+      "learning_rate": 0.00012707843275708028,
+      "loss": 2.3613,
+      "step": 6715
+    },
+    {
+      "epoch": 0.41313976377952755,
+      "grad_norm": 1.04655921459198,
+      "learning_rate": 0.00012705981706636807,
+      "loss": 2.4573,
+      "step": 6716
+    },
+    {
+      "epoch": 0.41320127952755903,
+      "grad_norm": 1.2413405179977417,
+      "learning_rate": 0.0001270412003637696,
+      "loss": 3.1622,
+      "step": 6717
+    },
+    {
+      "epoch": 0.41326279527559057,
+      "grad_norm": 1.1774698495864868,
+      "learning_rate": 0.000127022582649981,
+      "loss": 2.562,
+      "step": 6718
+    },
+    {
+      "epoch": 0.41332431102362205,
+      "grad_norm": 1.170767903327942,
+      "learning_rate": 0.0001270039639256985,
+      "loss": 2.3892,
+      "step": 6719
+    },
+    {
+      "epoch": 0.41338582677165353,
+      "grad_norm": 1.0432016849517822,
+      "learning_rate": 0.00012698534419161833,
+      "loss": 2.7437,
+      "step": 6720
+    },
+    {
+      "epoch": 0.413447342519685,
+      "grad_norm": 1.1866225004196167,
+      "learning_rate": 0.00012696672344843677,
+      "loss": 2.3456,
+      "step": 6721
+    },
+    {
+      "epoch": 0.41350885826771655,
+      "grad_norm": 1.3061898946762085,
+      "learning_rate": 0.00012694810169685012,
+      "loss": 2.584,
+      "step": 6722
+    },
+    {
+      "epoch": 0.41357037401574803,
+      "grad_norm": 1.4312163591384888,
+      "learning_rate": 0.00012692947893755473,
+      "loss": 3.2241,
+      "step": 6723
+    },
+    {
+      "epoch": 0.4136318897637795,
+      "grad_norm": 1.0574748516082764,
+      "learning_rate": 0.00012691085517124696,
+      "loss": 2.6247,
+      "step": 6724
+    },
+    {
+      "epoch": 0.413693405511811,
+      "grad_norm": 1.046142816543579,
+      "learning_rate": 0.0001268922303986233,
+      "loss": 2.9039,
+      "step": 6725
+    },
+    {
+      "epoch": 0.41375492125984253,
+      "grad_norm": 1.2210922241210938,
+      "learning_rate": 0.0001268736046203802,
+      "loss": 2.8824,
+      "step": 6726
+    },
+    {
+      "epoch": 0.413816437007874,
+      "grad_norm": 1.300827145576477,
+      "learning_rate": 0.00012685497783721406,
+      "loss": 2.3899,
+      "step": 6727
+    },
+    {
+      "epoch": 0.4138779527559055,
+      "grad_norm": 1.140870451927185,
+      "learning_rate": 0.00012683635004982156,
+      "loss": 2.7789,
+      "step": 6728
+    },
+    {
+      "epoch": 0.413939468503937,
+      "grad_norm": 1.1740866899490356,
+      "learning_rate": 0.0001268177212588992,
+      "loss": 2.6938,
+      "step": 6729
+    },
+    {
+      "epoch": 0.4140009842519685,
+      "grad_norm": 1.4221967458724976,
+      "learning_rate": 0.00012679909146514362,
+      "loss": 2.5892,
+      "step": 6730
+    },
+    {
+      "epoch": 0.4140625,
+      "grad_norm": 1.2634954452514648,
+      "learning_rate": 0.0001267804606692514,
+      "loss": 2.4951,
+      "step": 6731
+    },
+    {
+      "epoch": 0.4141240157480315,
+      "grad_norm": 1.0183467864990234,
+      "learning_rate": 0.00012676182887191932,
+      "loss": 2.6952,
+      "step": 6732
+    },
+    {
+      "epoch": 0.414185531496063,
+      "grad_norm": 1.285434365272522,
+      "learning_rate": 0.00012674319607384408,
+      "loss": 2.1602,
+      "step": 6733
+    },
+    {
+      "epoch": 0.4142470472440945,
+      "grad_norm": 1.1299057006835938,
+      "learning_rate": 0.0001267245622757224,
+      "loss": 2.9596,
+      "step": 6734
+    },
+    {
+      "epoch": 0.414308562992126,
+      "grad_norm": 1.2350070476531982,
+      "learning_rate": 0.00012670592747825114,
+      "loss": 2.6641,
+      "step": 6735
+    },
+    {
+      "epoch": 0.41437007874015747,
+      "grad_norm": 1.1408334970474243,
+      "learning_rate": 0.00012668729168212713,
+      "loss": 2.3337,
+      "step": 6736
+    },
+    {
+      "epoch": 0.414431594488189,
+      "grad_norm": 1.1531099081039429,
+      "learning_rate": 0.00012666865488804717,
+      "loss": 1.9902,
+      "step": 6737
+    },
+    {
+      "epoch": 0.4144931102362205,
+      "grad_norm": 1.090077519416809,
+      "learning_rate": 0.00012665001709670826,
+      "loss": 2.7722,
+      "step": 6738
+    },
+    {
+      "epoch": 0.41455462598425197,
+      "grad_norm": 1.3731459379196167,
+      "learning_rate": 0.00012663137830880728,
+      "loss": 2.8407,
+      "step": 6739
+    },
+    {
+      "epoch": 0.41461614173228345,
+      "grad_norm": 1.4312993288040161,
+      "learning_rate": 0.0001266127385250413,
+      "loss": 3.1388,
+      "step": 6740
+    },
+    {
+      "epoch": 0.414677657480315,
+      "grad_norm": 1.2896028757095337,
+      "learning_rate": 0.00012659409774610724,
+      "loss": 2.2284,
+      "step": 6741
+    },
+    {
+      "epoch": 0.41473917322834647,
+      "grad_norm": 1.184605360031128,
+      "learning_rate": 0.00012657545597270225,
+      "loss": 1.8972,
+      "step": 6742
+    },
+    {
+      "epoch": 0.41480068897637795,
+      "grad_norm": 1.4814672470092773,
+      "learning_rate": 0.0001265568132055234,
+      "loss": 2.4564,
+      "step": 6743
+    },
+    {
+      "epoch": 0.41486220472440943,
+      "grad_norm": 1.0805320739746094,
+      "learning_rate": 0.0001265381694452678,
+      "loss": 1.6989,
+      "step": 6744
+    },
+    {
+      "epoch": 0.41492372047244097,
+      "grad_norm": 1.2953877449035645,
+      "learning_rate": 0.00012651952469263267,
+      "loss": 2.8596,
+      "step": 6745
+    },
+    {
+      "epoch": 0.41498523622047245,
+      "grad_norm": 1.2891323566436768,
+      "learning_rate": 0.00012650087894831518,
+      "loss": 2.2205,
+      "step": 6746
+    },
+    {
+      "epoch": 0.41504675196850394,
+      "grad_norm": 1.239094614982605,
+      "learning_rate": 0.0001264822322130126,
+      "loss": 2.9508,
+      "step": 6747
+    },
+    {
+      "epoch": 0.4151082677165354,
+      "grad_norm": 1.2005199193954468,
+      "learning_rate": 0.0001264635844874222,
+      "loss": 2.3134,
+      "step": 6748
+    },
+    {
+      "epoch": 0.41516978346456695,
+      "grad_norm": 1.3192403316497803,
+      "learning_rate": 0.00012644493577224127,
+      "loss": 2.8132,
+      "step": 6749
+    },
+    {
+      "epoch": 0.41523129921259844,
+      "grad_norm": 1.1186935901641846,
+      "learning_rate": 0.00012642628606816727,
+      "loss": 2.6321,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4152928149606299,
+      "grad_norm": 1.478867769241333,
+      "learning_rate": 0.00012640763537589747,
+      "loss": 2.8127,
+      "step": 6751
+    },
+    {
+      "epoch": 0.4153543307086614,
+      "grad_norm": 1.0495606660842896,
+      "learning_rate": 0.0001263889836961294,
+      "loss": 2.54,
+      "step": 6752
+    },
+    {
+      "epoch": 0.41541584645669294,
+      "grad_norm": 1.1298325061798096,
+      "learning_rate": 0.00012637033102956045,
+      "loss": 3.039,
+      "step": 6753
+    },
+    {
+      "epoch": 0.4154773622047244,
+      "grad_norm": 1.114201545715332,
+      "learning_rate": 0.00012635167737688817,
+      "loss": 2.5775,
+      "step": 6754
+    },
+    {
+      "epoch": 0.4155388779527559,
+      "grad_norm": 1.2703338861465454,
+      "learning_rate": 0.00012633302273881017,
+      "loss": 2.5128,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4156003937007874,
+      "grad_norm": 1.234359622001648,
+      "learning_rate": 0.00012631436711602387,
+      "loss": 3.0957,
+      "step": 6756
+    },
+    {
+      "epoch": 0.4156619094488189,
+      "grad_norm": 1.0051590204238892,
+      "learning_rate": 0.00012629571050922698,
+      "loss": 2.4537,
+      "step": 6757
+    },
+    {
+      "epoch": 0.4157234251968504,
+      "grad_norm": 1.2320126295089722,
+      "learning_rate": 0.0001262770529191172,
+      "loss": 2.7673,
+      "step": 6758
+    },
+    {
+      "epoch": 0.4157849409448819,
+      "grad_norm": 1.083877682685852,
+      "learning_rate": 0.00012625839434639213,
+      "loss": 2.5286,
+      "step": 6759
+    },
+    {
+      "epoch": 0.41584645669291337,
+      "grad_norm": 1.5409575700759888,
+      "learning_rate": 0.00012623973479174952,
+      "loss": 2.3591,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4159079724409449,
+      "grad_norm": 1.240633487701416,
+      "learning_rate": 0.00012622107425588718,
+      "loss": 3.132,
+      "step": 6761
+    },
+    {
+      "epoch": 0.4159694881889764,
+      "grad_norm": 1.0124597549438477,
+      "learning_rate": 0.00012620241273950285,
+      "loss": 2.6607,
+      "step": 6762
+    },
+    {
+      "epoch": 0.41603100393700787,
+      "grad_norm": 1.3015955686569214,
+      "learning_rate": 0.0001261837502432944,
+      "loss": 2.2763,
+      "step": 6763
+    },
+    {
+      "epoch": 0.41609251968503935,
+      "grad_norm": 1.1862910985946655,
+      "learning_rate": 0.00012616508676795967,
+      "loss": 2.0758,
+      "step": 6764
+    },
+    {
+      "epoch": 0.4161540354330709,
+      "grad_norm": 1.2923144102096558,
+      "learning_rate": 0.00012614642231419666,
+      "loss": 3.1218,
+      "step": 6765
+    },
+    {
+      "epoch": 0.4162155511811024,
+      "grad_norm": 1.3869863748550415,
+      "learning_rate": 0.00012612775688270318,
+      "loss": 2.8875,
+      "step": 6766
+    },
+    {
+      "epoch": 0.41627706692913385,
+      "grad_norm": 1.236663579940796,
+      "learning_rate": 0.0001261090904741773,
+      "loss": 2.2645,
+      "step": 6767
+    },
+    {
+      "epoch": 0.41633858267716534,
+      "grad_norm": 1.3731883764266968,
+      "learning_rate": 0.00012609042308931706,
+      "loss": 1.9335,
+      "step": 6768
+    },
+    {
+      "epoch": 0.4164000984251969,
+      "grad_norm": 1.5505197048187256,
+      "learning_rate": 0.00012607175472882044,
+      "loss": 2.0346,
+      "step": 6769
+    },
+    {
+      "epoch": 0.41646161417322836,
+      "grad_norm": 1.0244718790054321,
+      "learning_rate": 0.0001260530853933856,
+      "loss": 1.7814,
+      "step": 6770
+    },
+    {
+      "epoch": 0.41652312992125984,
+      "grad_norm": 1.3682419061660767,
+      "learning_rate": 0.00012603441508371067,
+      "loss": 3.493,
+      "step": 6771
+    },
+    {
+      "epoch": 0.4165846456692913,
+      "grad_norm": 1.3246978521347046,
+      "learning_rate": 0.00012601574380049374,
+      "loss": 2.318,
+      "step": 6772
+    },
+    {
+      "epoch": 0.41664616141732286,
+      "grad_norm": 1.243801474571228,
+      "learning_rate": 0.00012599707154443306,
+      "loss": 2.5353,
+      "step": 6773
+    },
+    {
+      "epoch": 0.41670767716535434,
+      "grad_norm": 1.1077210903167725,
+      "learning_rate": 0.00012597839831622688,
+      "loss": 3.0461,
+      "step": 6774
+    },
+    {
+      "epoch": 0.4167691929133858,
+      "grad_norm": 1.254346251487732,
+      "learning_rate": 0.00012595972411657348,
+      "loss": 2.6981,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4168307086614173,
+      "grad_norm": 1.0701122283935547,
+      "learning_rate": 0.0001259410489461711,
+      "loss": 2.8495,
+      "step": 6776
+    },
+    {
+      "epoch": 0.41689222440944884,
+      "grad_norm": 1.0210282802581787,
+      "learning_rate": 0.00012592237280571823,
+      "loss": 3.1392,
+      "step": 6777
+    },
+    {
+      "epoch": 0.4169537401574803,
+      "grad_norm": 1.2218111753463745,
+      "learning_rate": 0.00012590369569591309,
+      "loss": 1.9038,
+      "step": 6778
+    },
+    {
+      "epoch": 0.4170152559055118,
+      "grad_norm": 0.9258589148521423,
+      "learning_rate": 0.00012588501761745419,
+      "loss": 2.8178,
+      "step": 6779
+    },
+    {
+      "epoch": 0.4170767716535433,
+      "grad_norm": 1.2572994232177734,
+      "learning_rate": 0.00012586633857103997,
+      "loss": 2.7607,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4171382874015748,
+      "grad_norm": 1.4264432191848755,
+      "learning_rate": 0.00012584765855736896,
+      "loss": 2.4242,
+      "step": 6781
+    },
+    {
+      "epoch": 0.4171998031496063,
+      "grad_norm": 1.6254597902297974,
+      "learning_rate": 0.00012582897757713957,
+      "loss": 2.4152,
+      "step": 6782
+    },
+    {
+      "epoch": 0.4172613188976378,
+      "grad_norm": 1.1048884391784668,
+      "learning_rate": 0.0001258102956310505,
+      "loss": 2.9047,
+      "step": 6783
+    },
+    {
+      "epoch": 0.41732283464566927,
+      "grad_norm": 1.0339524745941162,
+      "learning_rate": 0.00012579161271980031,
+      "loss": 2.7643,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4173843503937008,
+      "grad_norm": 1.3044381141662598,
+      "learning_rate": 0.0001257729288440876,
+      "loss": 2.2941,
+      "step": 6785
+    },
+    {
+      "epoch": 0.4174458661417323,
+      "grad_norm": 1.1513782739639282,
+      "learning_rate": 0.00012575424400461106,
+      "loss": 2.7268,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4175073818897638,
+      "grad_norm": 1.2861382961273193,
+      "learning_rate": 0.00012573555820206945,
+      "loss": 2.3326,
+      "step": 6787
+    },
+    {
+      "epoch": 0.41756889763779526,
+      "grad_norm": 1.4622645378112793,
+      "learning_rate": 0.0001257168714371614,
+      "loss": 2.8893,
+      "step": 6788
+    },
+    {
+      "epoch": 0.4176304133858268,
+      "grad_norm": 1.1407907009124756,
+      "learning_rate": 0.00012569818371058577,
+      "loss": 2.969,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4176919291338583,
+      "grad_norm": 1.0292813777923584,
+      "learning_rate": 0.00012567949502304143,
+      "loss": 2.6088,
+      "step": 6790
+    },
+    {
+      "epoch": 0.41775344488188976,
+      "grad_norm": 1.5028873682022095,
+      "learning_rate": 0.0001256608053752271,
+      "loss": 2.5674,
+      "step": 6791
+    },
+    {
+      "epoch": 0.41781496062992124,
+      "grad_norm": 1.2340095043182373,
+      "learning_rate": 0.00012564211476784177,
+      "loss": 2.9806,
+      "step": 6792
+    },
+    {
+      "epoch": 0.4178764763779528,
+      "grad_norm": 1.2387328147888184,
+      "learning_rate": 0.00012562342320158434,
+      "loss": 2.8843,
+      "step": 6793
+    },
+    {
+      "epoch": 0.41793799212598426,
+      "grad_norm": 1.329601764678955,
+      "learning_rate": 0.00012560473067715374,
+      "loss": 2.2649,
+      "step": 6794
+    },
+    {
+      "epoch": 0.41799950787401574,
+      "grad_norm": 1.0052926540374756,
+      "learning_rate": 0.00012558603719524898,
+      "loss": 2.8157,
+      "step": 6795
+    },
+    {
+      "epoch": 0.4180610236220472,
+      "grad_norm": 1.156911015510559,
+      "learning_rate": 0.00012556734275656913,
+      "loss": 2.8031,
+      "step": 6796
+    },
+    {
+      "epoch": 0.41812253937007876,
+      "grad_norm": 1.0229289531707764,
+      "learning_rate": 0.00012554864736181323,
+      "loss": 2.8638,
+      "step": 6797
+    },
+    {
+      "epoch": 0.41818405511811024,
+      "grad_norm": 1.2157974243164062,
+      "learning_rate": 0.00012552995101168036,
+      "loss": 3.0301,
+      "step": 6798
+    },
+    {
+      "epoch": 0.4182455708661417,
+      "grad_norm": 1.0633349418640137,
+      "learning_rate": 0.00012551125370686967,
+      "loss": 2.7193,
+      "step": 6799
+    },
+    {
+      "epoch": 0.4183070866141732,
+      "grad_norm": 0.9671129584312439,
+      "learning_rate": 0.00012549255544808035,
+      "loss": 2.8332,
+      "step": 6800
+    },
+    {
+      "epoch": 0.41836860236220474,
+      "grad_norm": 1.2146166563034058,
+      "learning_rate": 0.00012547385623601161,
+      "loss": 2.8925,
+      "step": 6801
+    },
+    {
+      "epoch": 0.4184301181102362,
+      "grad_norm": 1.2499457597732544,
+      "learning_rate": 0.0001254551560713627,
+      "loss": 1.9363,
+      "step": 6802
+    },
+    {
+      "epoch": 0.4184916338582677,
+      "grad_norm": 1.202637791633606,
+      "learning_rate": 0.0001254364549548329,
+      "loss": 3.2319,
+      "step": 6803
+    },
+    {
+      "epoch": 0.4185531496062992,
+      "grad_norm": 1.0586092472076416,
+      "learning_rate": 0.00012541775288712155,
+      "loss": 2.1457,
+      "step": 6804
+    },
+    {
+      "epoch": 0.41861466535433073,
+      "grad_norm": 1.1105329990386963,
+      "learning_rate": 0.0001253990498689279,
+      "loss": 2.945,
+      "step": 6805
+    },
+    {
+      "epoch": 0.4186761811023622,
+      "grad_norm": 0.9774209260940552,
+      "learning_rate": 0.00012538034590095149,
+      "loss": 2.9138,
+      "step": 6806
+    },
+    {
+      "epoch": 0.4187376968503937,
+      "grad_norm": 1.083125352859497,
+      "learning_rate": 0.00012536164098389164,
+      "loss": 3.0652,
+      "step": 6807
+    },
+    {
+      "epoch": 0.4187992125984252,
+      "grad_norm": 1.2415733337402344,
+      "learning_rate": 0.0001253429351184478,
+      "loss": 2.4999,
+      "step": 6808
+    },
+    {
+      "epoch": 0.4188607283464567,
+      "grad_norm": 1.2355259656906128,
+      "learning_rate": 0.00012532422830531956,
+      "loss": 3.0526,
+      "step": 6809
+    },
+    {
+      "epoch": 0.4189222440944882,
+      "grad_norm": 1.549412488937378,
+      "learning_rate": 0.00012530552054520637,
+      "loss": 2.2537,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4189837598425197,
+      "grad_norm": 1.224049687385559,
+      "learning_rate": 0.00012528681183880783,
+      "loss": 2.4324,
+      "step": 6811
+    },
+    {
+      "epoch": 0.41904527559055116,
+      "grad_norm": 1.1016188859939575,
+      "learning_rate": 0.0001252681021868235,
+      "loss": 2.9823,
+      "step": 6812
+    },
+    {
+      "epoch": 0.4191067913385827,
+      "grad_norm": 1.1057404279708862,
+      "learning_rate": 0.00012524939158995308,
+      "loss": 3.0088,
+      "step": 6813
+    },
+    {
+      "epoch": 0.4191683070866142,
+      "grad_norm": 1.005022644996643,
+      "learning_rate": 0.00012523068004889618,
+      "loss": 2.9111,
+      "step": 6814
+    },
+    {
+      "epoch": 0.41922982283464566,
+      "grad_norm": 1.2268017530441284,
+      "learning_rate": 0.00012521196756435255,
+      "loss": 2.4699,
+      "step": 6815
+    },
+    {
+      "epoch": 0.41929133858267714,
+      "grad_norm": 1.2577327489852905,
+      "learning_rate": 0.00012519325413702195,
+      "loss": 2.5109,
+      "step": 6816
+    },
+    {
+      "epoch": 0.4193528543307087,
+      "grad_norm": 1.217483401298523,
+      "learning_rate": 0.0001251745397676041,
+      "loss": 2.8202,
+      "step": 6817
+    },
+    {
+      "epoch": 0.41941437007874016,
+      "grad_norm": 1.1106187105178833,
+      "learning_rate": 0.00012515582445679881,
+      "loss": 2.8197,
+      "step": 6818
+    },
+    {
+      "epoch": 0.41947588582677164,
+      "grad_norm": 1.3303033113479614,
+      "learning_rate": 0.00012513710820530598,
+      "loss": 3.1294,
+      "step": 6819
+    },
+    {
+      "epoch": 0.4195374015748031,
+      "grad_norm": 1.1282480955123901,
+      "learning_rate": 0.0001251183910138255,
+      "loss": 2.1772,
+      "step": 6820
+    },
+    {
+      "epoch": 0.41959891732283466,
+      "grad_norm": 1.1972078084945679,
+      "learning_rate": 0.0001250996728830572,
+      "loss": 2.3432,
+      "step": 6821
+    },
+    {
+      "epoch": 0.41966043307086615,
+      "grad_norm": 1.1761964559555054,
+      "learning_rate": 0.00012508095381370116,
+      "loss": 2.735,
+      "step": 6822
+    },
+    {
+      "epoch": 0.4197219488188976,
+      "grad_norm": 1.1288365125656128,
+      "learning_rate": 0.00012506223380645727,
+      "loss": 2.9398,
+      "step": 6823
+    },
+    {
+      "epoch": 0.4197834645669291,
+      "grad_norm": 1.0886878967285156,
+      "learning_rate": 0.00012504351286202558,
+      "loss": 2.4903,
+      "step": 6824
+    },
+    {
+      "epoch": 0.41984498031496065,
+      "grad_norm": 1.197951316833496,
+      "learning_rate": 0.00012502479098110616,
+      "loss": 2.4861,
+      "step": 6825
+    },
+    {
+      "epoch": 0.41990649606299213,
+      "grad_norm": 0.9787187576293945,
+      "learning_rate": 0.0001250060681643991,
+      "loss": 2.925,
+      "step": 6826
+    },
+    {
+      "epoch": 0.4199680118110236,
+      "grad_norm": 1.6437605619430542,
+      "learning_rate": 0.00012498734441260454,
+      "loss": 1.9542,
+      "step": 6827
+    },
+    {
+      "epoch": 0.4200295275590551,
+      "grad_norm": 1.850866675376892,
+      "learning_rate": 0.00012496861972642262,
+      "loss": 2.183,
+      "step": 6828
+    },
+    {
+      "epoch": 0.42009104330708663,
+      "grad_norm": 0.9835193157196045,
+      "learning_rate": 0.00012494989410655356,
+      "loss": 2.9284,
+      "step": 6829
+    },
+    {
+      "epoch": 0.4201525590551181,
+      "grad_norm": 2.7572524547576904,
+      "learning_rate": 0.0001249311675536976,
+      "loss": 2.3964,
+      "step": 6830
+    },
+    {
+      "epoch": 0.4202140748031496,
+      "grad_norm": 1.4726897478103638,
+      "learning_rate": 0.00012491244006855495,
+      "loss": 2.5754,
+      "step": 6831
+    },
+    {
+      "epoch": 0.4202755905511811,
+      "grad_norm": 1.2485735416412354,
+      "learning_rate": 0.000124893711651826,
+      "loss": 2.9919,
+      "step": 6832
+    },
+    {
+      "epoch": 0.4203371062992126,
+      "grad_norm": 1.2788827419281006,
+      "learning_rate": 0.000124874982304211,
+      "loss": 2.1822,
+      "step": 6833
+    },
+    {
+      "epoch": 0.4203986220472441,
+      "grad_norm": 1.0010789632797241,
+      "learning_rate": 0.00012485625202641041,
+      "loss": 3.0605,
+      "step": 6834
+    },
+    {
+      "epoch": 0.4204601377952756,
+      "grad_norm": 1.1891103982925415,
+      "learning_rate": 0.00012483752081912457,
+      "loss": 2.9962,
+      "step": 6835
+    },
+    {
+      "epoch": 0.42052165354330706,
+      "grad_norm": 1.216060757637024,
+      "learning_rate": 0.00012481878868305396,
+      "loss": 1.2999,
+      "step": 6836
+    },
+    {
+      "epoch": 0.4205831692913386,
+      "grad_norm": 1.1790834665298462,
+      "learning_rate": 0.00012480005561889908,
+      "loss": 2.4434,
+      "step": 6837
+    },
+    {
+      "epoch": 0.4206446850393701,
+      "grad_norm": 1.203153133392334,
+      "learning_rate": 0.00012478132162736035,
+      "loss": 2.8198,
+      "step": 6838
+    },
+    {
+      "epoch": 0.42070620078740156,
+      "grad_norm": 1.002672791481018,
+      "learning_rate": 0.00012476258670913844,
+      "loss": 3.0378,
+      "step": 6839
+    },
+    {
+      "epoch": 0.42076771653543305,
+      "grad_norm": 1.3367253541946411,
+      "learning_rate": 0.0001247438508649338,
+      "loss": 3.0727,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4208292322834646,
+      "grad_norm": 1.2348965406417847,
+      "learning_rate": 0.00012472511409544713,
+      "loss": 2.7329,
+      "step": 6841
+    },
+    {
+      "epoch": 0.42089074803149606,
+      "grad_norm": 1.2131905555725098,
+      "learning_rate": 0.0001247063764013791,
+      "loss": 2.8074,
+      "step": 6842
+    },
+    {
+      "epoch": 0.42095226377952755,
+      "grad_norm": 1.1310256719589233,
+      "learning_rate": 0.00012468763778343032,
+      "loss": 3.0939,
+      "step": 6843
+    },
+    {
+      "epoch": 0.42101377952755903,
+      "grad_norm": 1.4428311586380005,
+      "learning_rate": 0.00012466889824230156,
+      "loss": 2.583,
+      "step": 6844
+    },
+    {
+      "epoch": 0.42107529527559057,
+      "grad_norm": 1.2492027282714844,
+      "learning_rate": 0.0001246501577786936,
+      "loss": 2.174,
+      "step": 6845
+    },
+    {
+      "epoch": 0.42113681102362205,
+      "grad_norm": 1.1243464946746826,
+      "learning_rate": 0.00012463141639330712,
+      "loss": 2.5541,
+      "step": 6846
+    },
+    {
+      "epoch": 0.42119832677165353,
+      "grad_norm": 1.2969512939453125,
+      "learning_rate": 0.00012461267408684306,
+      "loss": 3.1736,
+      "step": 6847
+    },
+    {
+      "epoch": 0.421259842519685,
+      "grad_norm": 1.1958410739898682,
+      "learning_rate": 0.00012459393086000224,
+      "loss": 2.7123,
+      "step": 6848
+    },
+    {
+      "epoch": 0.42132135826771655,
+      "grad_norm": 1.315287470817566,
+      "learning_rate": 0.00012457518671348553,
+      "loss": 2.5246,
+      "step": 6849
+    },
+    {
+      "epoch": 0.42138287401574803,
+      "grad_norm": 1.1188747882843018,
+      "learning_rate": 0.00012455644164799384,
+      "loss": 2.5248,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4214443897637795,
+      "grad_norm": 1.2037899494171143,
+      "learning_rate": 0.0001245376956642282,
+      "loss": 2.6602,
+      "step": 6851
+    },
+    {
+      "epoch": 0.421505905511811,
+      "grad_norm": 2.6991820335388184,
+      "learning_rate": 0.0001245189487628896,
+      "loss": 1.7853,
+      "step": 6852
+    },
+    {
+      "epoch": 0.42156742125984253,
+      "grad_norm": 1.2053344249725342,
+      "learning_rate": 0.000124500200944679,
+      "loss": 2.3045,
+      "step": 6853
+    },
+    {
+      "epoch": 0.421628937007874,
+      "grad_norm": 1.2939118146896362,
+      "learning_rate": 0.0001244814522102975,
+      "loss": 2.3533,
+      "step": 6854
+    },
+    {
+      "epoch": 0.4216904527559055,
+      "grad_norm": 1.2586112022399902,
+      "learning_rate": 0.0001244627025604462,
+      "loss": 2.9755,
+      "step": 6855
+    },
+    {
+      "epoch": 0.421751968503937,
+      "grad_norm": 1.0985870361328125,
+      "learning_rate": 0.00012444395199582623,
+      "loss": 2.6938,
+      "step": 6856
+    },
+    {
+      "epoch": 0.4218134842519685,
+      "grad_norm": 1.1721208095550537,
+      "learning_rate": 0.00012442520051713878,
+      "loss": 2.1916,
+      "step": 6857
+    },
+    {
+      "epoch": 0.421875,
+      "grad_norm": 1.3470538854599,
+      "learning_rate": 0.00012440644812508503,
+      "loss": 2.8388,
+      "step": 6858
+    },
+    {
+      "epoch": 0.4219365157480315,
+      "grad_norm": 1.2786595821380615,
+      "learning_rate": 0.00012438769482036617,
+      "loss": 2.9907,
+      "step": 6859
+    },
+    {
+      "epoch": 0.421998031496063,
+      "grad_norm": 1.2025648355484009,
+      "learning_rate": 0.00012436894060368354,
+      "loss": 2.6068,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4220595472440945,
+      "grad_norm": 1.3339189291000366,
+      "learning_rate": 0.00012435018547573843,
+      "loss": 2.9071,
+      "step": 6861
+    },
+    {
+      "epoch": 0.422121062992126,
+      "grad_norm": 1.0840085744857788,
+      "learning_rate": 0.00012433142943723213,
+      "loss": 2.6353,
+      "step": 6862
+    },
+    {
+      "epoch": 0.42218257874015747,
+      "grad_norm": 1.4049098491668701,
+      "learning_rate": 0.00012431267248886606,
+      "loss": 2.8595,
+      "step": 6863
+    },
+    {
+      "epoch": 0.422244094488189,
+      "grad_norm": 1.0122581720352173,
+      "learning_rate": 0.00012429391463134162,
+      "loss": 2.6611,
+      "step": 6864
+    },
+    {
+      "epoch": 0.4223056102362205,
+      "grad_norm": 1.1310001611709595,
+      "learning_rate": 0.00012427515586536028,
+      "loss": 2.2713,
+      "step": 6865
+    },
+    {
+      "epoch": 0.42236712598425197,
+      "grad_norm": 1.0846912860870361,
+      "learning_rate": 0.00012425639619162342,
+      "loss": 2.5165,
+      "step": 6866
+    },
+    {
+      "epoch": 0.42242864173228345,
+      "grad_norm": 1.1606781482696533,
+      "learning_rate": 0.00012423763561083263,
+      "loss": 2.9196,
+      "step": 6867
+    },
+    {
+      "epoch": 0.422490157480315,
+      "grad_norm": 1.1683895587921143,
+      "learning_rate": 0.00012421887412368936,
+      "loss": 1.9502,
+      "step": 6868
+    },
+    {
+      "epoch": 0.42255167322834647,
+      "grad_norm": 1.5453615188598633,
+      "learning_rate": 0.0001242001117308953,
+      "loss": 2.2247,
+      "step": 6869
+    },
+    {
+      "epoch": 0.42261318897637795,
+      "grad_norm": 1.9233101606369019,
+      "learning_rate": 0.000124181348433152,
+      "loss": 2.5346,
+      "step": 6870
+    },
+    {
+      "epoch": 0.42267470472440943,
+      "grad_norm": 1.0947074890136719,
+      "learning_rate": 0.00012416258423116107,
+      "loss": 2.6609,
+      "step": 6871
+    },
+    {
+      "epoch": 0.42273622047244097,
+      "grad_norm": 1.2572311162948608,
+      "learning_rate": 0.00012414381912562424,
+      "loss": 2.0532,
+      "step": 6872
+    },
+    {
+      "epoch": 0.42279773622047245,
+      "grad_norm": 1.3745061159133911,
+      "learning_rate": 0.00012412505311724324,
+      "loss": 2.3732,
+      "step": 6873
+    },
+    {
+      "epoch": 0.42285925196850394,
+      "grad_norm": 1.3851933479309082,
+      "learning_rate": 0.00012410628620671975,
+      "loss": 3.2842,
+      "step": 6874
+    },
+    {
+      "epoch": 0.4229207677165354,
+      "grad_norm": 1.1068551540374756,
+      "learning_rate": 0.00012408751839475558,
+      "loss": 2.9506,
+      "step": 6875
+    },
+    {
+      "epoch": 0.42298228346456695,
+      "grad_norm": 1.2625681161880493,
+      "learning_rate": 0.00012406874968205257,
+      "loss": 2.7967,
+      "step": 6876
+    },
+    {
+      "epoch": 0.42304379921259844,
+      "grad_norm": 1.131888747215271,
+      "learning_rate": 0.00012404998006931248,
+      "loss": 2.8856,
+      "step": 6877
+    },
+    {
+      "epoch": 0.4231053149606299,
+      "grad_norm": 1.2009429931640625,
+      "learning_rate": 0.00012403120955723726,
+      "loss": 2.4685,
+      "step": 6878
+    },
+    {
+      "epoch": 0.4231668307086614,
+      "grad_norm": 1.1900200843811035,
+      "learning_rate": 0.0001240124381465288,
+      "loss": 2.2243,
+      "step": 6879
+    },
+    {
+      "epoch": 0.42322834645669294,
+      "grad_norm": 1.2239503860473633,
+      "learning_rate": 0.00012399366583788906,
+      "loss": 2.4883,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4232898622047244,
+      "grad_norm": 1.3763883113861084,
+      "learning_rate": 0.00012397489263202,
+      "loss": 2.575,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4233513779527559,
+      "grad_norm": 1.1037997007369995,
+      "learning_rate": 0.00012395611852962366,
+      "loss": 2.6985,
+      "step": 6882
+    },
+    {
+      "epoch": 0.4234128937007874,
+      "grad_norm": 1.4167383909225464,
+      "learning_rate": 0.00012393734353140207,
+      "loss": 2.2513,
+      "step": 6883
+    },
+    {
+      "epoch": 0.4234744094488189,
+      "grad_norm": 1.3297178745269775,
+      "learning_rate": 0.00012391856763805726,
+      "loss": 2.6157,
+      "step": 6884
+    },
+    {
+      "epoch": 0.4235359251968504,
+      "grad_norm": 1.4269884824752808,
+      "learning_rate": 0.00012389979085029146,
+      "loss": 2.9155,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4235974409448819,
+      "grad_norm": 1.0242115259170532,
+      "learning_rate": 0.00012388101316880673,
+      "loss": 2.0608,
+      "step": 6886
+    },
+    {
+      "epoch": 0.42365895669291337,
+      "grad_norm": 1.2060352563858032,
+      "learning_rate": 0.00012386223459430527,
+      "loss": 2.9442,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4237204724409449,
+      "grad_norm": 1.0271624326705933,
+      "learning_rate": 0.00012384345512748927,
+      "loss": 2.6872,
+      "step": 6888
+    },
+    {
+      "epoch": 0.4237819881889764,
+      "grad_norm": 1.081023931503296,
+      "learning_rate": 0.00012382467476906105,
+      "loss": 2.5569,
+      "step": 6889
+    },
+    {
+      "epoch": 0.42384350393700787,
+      "grad_norm": 1.125035047531128,
+      "learning_rate": 0.00012380589351972284,
+      "loss": 3.0067,
+      "step": 6890
+    },
+    {
+      "epoch": 0.42390501968503935,
+      "grad_norm": 1.1405842304229736,
+      "learning_rate": 0.0001237871113801769,
+      "loss": 2.4665,
+      "step": 6891
+    },
+    {
+      "epoch": 0.4239665354330709,
+      "grad_norm": 1.3058969974517822,
+      "learning_rate": 0.00012376832835112568,
+      "loss": 2.7748,
+      "step": 6892
+    },
+    {
+      "epoch": 0.4240280511811024,
+      "grad_norm": 1.106742024421692,
+      "learning_rate": 0.0001237495444332715,
+      "loss": 3.3617,
+      "step": 6893
+    },
+    {
+      "epoch": 0.42408956692913385,
+      "grad_norm": 1.2528691291809082,
+      "learning_rate": 0.0001237307596273168,
+      "loss": 2.6684,
+      "step": 6894
+    },
+    {
+      "epoch": 0.42415108267716534,
+      "grad_norm": 1.398490071296692,
+      "learning_rate": 0.000123711973933964,
+      "loss": 1.9979,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4242125984251969,
+      "grad_norm": 1.0098438262939453,
+      "learning_rate": 0.00012369318735391564,
+      "loss": 2.1242,
+      "step": 6896
+    },
+    {
+      "epoch": 0.42427411417322836,
+      "grad_norm": 1.013129472732544,
+      "learning_rate": 0.00012367439988787414,
+      "loss": 2.5198,
+      "step": 6897
+    },
+    {
+      "epoch": 0.42433562992125984,
+      "grad_norm": 1.2973878383636475,
+      "learning_rate": 0.00012365561153654213,
+      "loss": 2.4126,
+      "step": 6898
+    },
+    {
+      "epoch": 0.4243971456692913,
+      "grad_norm": 2.1047353744506836,
+      "learning_rate": 0.00012363682230062215,
+      "loss": 2.7511,
+      "step": 6899
+    },
+    {
+      "epoch": 0.42445866141732286,
+      "grad_norm": 1.5129436254501343,
+      "learning_rate": 0.0001236180321808168,
+      "loss": 2.3779,
+      "step": 6900
+    },
+    {
+      "epoch": 0.42452017716535434,
+      "grad_norm": 1.3512458801269531,
+      "learning_rate": 0.00012359924117782875,
+      "loss": 2.9503,
+      "step": 6901
+    },
+    {
+      "epoch": 0.4245816929133858,
+      "grad_norm": 1.2071129083633423,
+      "learning_rate": 0.0001235804492923607,
+      "loss": 2.513,
+      "step": 6902
+    },
+    {
+      "epoch": 0.4246432086614173,
+      "grad_norm": 1.1189714670181274,
+      "learning_rate": 0.0001235616565251153,
+      "loss": 1.7402,
+      "step": 6903
+    },
+    {
+      "epoch": 0.42470472440944884,
+      "grad_norm": 1.3587483167648315,
+      "learning_rate": 0.00012354286287679536,
+      "loss": 2.2877,
+      "step": 6904
+    },
+    {
+      "epoch": 0.4247662401574803,
+      "grad_norm": 1.1409732103347778,
+      "learning_rate": 0.0001235240683481036,
+      "loss": 2.7228,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4248277559055118,
+      "grad_norm": 1.2218971252441406,
+      "learning_rate": 0.00012350527293974285,
+      "loss": 2.8501,
+      "step": 6906
+    },
+    {
+      "epoch": 0.4248892716535433,
+      "grad_norm": 1.0022624731063843,
+      "learning_rate": 0.00012348647665241599,
+      "loss": 2.3176,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4249507874015748,
+      "grad_norm": 1.3137249946594238,
+      "learning_rate": 0.00012346767948682587,
+      "loss": 2.7854,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4250123031496063,
+      "grad_norm": 1.1475485563278198,
+      "learning_rate": 0.00012344888144367536,
+      "loss": 2.2751,
+      "step": 6909
+    },
+    {
+      "epoch": 0.4250738188976378,
+      "grad_norm": 1.1032676696777344,
+      "learning_rate": 0.00012343008252366745,
+      "loss": 2.6532,
+      "step": 6910
+    },
+    {
+      "epoch": 0.42513533464566927,
+      "grad_norm": 1.0176992416381836,
+      "learning_rate": 0.00012341128272750512,
+      "loss": 2.6967,
+      "step": 6911
+    },
+    {
+      "epoch": 0.4251968503937008,
+      "grad_norm": 1.0026066303253174,
+      "learning_rate": 0.00012339248205589138,
+      "loss": 3.3864,
+      "step": 6912
+    },
+    {
+      "epoch": 0.4252583661417323,
+      "grad_norm": 1.043777346611023,
+      "learning_rate": 0.0001233736805095292,
+      "loss": 2.8275,
+      "step": 6913
+    },
+    {
+      "epoch": 0.4253198818897638,
+      "grad_norm": 1.2229052782058716,
+      "learning_rate": 0.00012335487808912173,
+      "loss": 2.088,
+      "step": 6914
+    },
+    {
+      "epoch": 0.42538139763779526,
+      "grad_norm": 1.2192978858947754,
+      "learning_rate": 0.00012333607479537206,
+      "loss": 2.8531,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4254429133858268,
+      "grad_norm": 1.338576316833496,
+      "learning_rate": 0.0001233172706289833,
+      "loss": 2.4437,
+      "step": 6916
+    },
+    {
+      "epoch": 0.4255044291338583,
+      "grad_norm": 1.1928391456604004,
+      "learning_rate": 0.00012329846559065865,
+      "loss": 2.5708,
+      "step": 6917
+    },
+    {
+      "epoch": 0.42556594488188976,
+      "grad_norm": 1.1135594844818115,
+      "learning_rate": 0.0001232796596811013,
+      "loss": 3.4135,
+      "step": 6918
+    },
+    {
+      "epoch": 0.42562746062992124,
+      "grad_norm": 1.1263635158538818,
+      "learning_rate": 0.00012326085290101447,
+      "loss": 2.1758,
+      "step": 6919
+    },
+    {
+      "epoch": 0.4256889763779528,
+      "grad_norm": 1.0773768424987793,
+      "learning_rate": 0.00012324204525110147,
+      "loss": 2.5246,
+      "step": 6920
+    },
+    {
+      "epoch": 0.42575049212598426,
+      "grad_norm": 1.0745408535003662,
+      "learning_rate": 0.0001232232367320656,
+      "loss": 2.6817,
+      "step": 6921
+    },
+    {
+      "epoch": 0.42581200787401574,
+      "grad_norm": 1.287124514579773,
+      "learning_rate": 0.00012320442734461016,
+      "loss": 2.0967,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4258735236220472,
+      "grad_norm": 1.2152758836746216,
+      "learning_rate": 0.0001231856170894385,
+      "loss": 2.7537,
+      "step": 6923
+    },
+    {
+      "epoch": 0.42593503937007876,
+      "grad_norm": 1.3459244966506958,
+      "learning_rate": 0.00012316680596725406,
+      "loss": 2.7713,
+      "step": 6924
+    },
+    {
+      "epoch": 0.42599655511811024,
+      "grad_norm": 1.3404775857925415,
+      "learning_rate": 0.00012314799397876027,
+      "loss": 3.2382,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4260580708661417,
+      "grad_norm": 1.1203370094299316,
+      "learning_rate": 0.00012312918112466055,
+      "loss": 2.3928,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4261195866141732,
+      "grad_norm": 1.0845093727111816,
+      "learning_rate": 0.00012311036740565845,
+      "loss": 2.2467,
+      "step": 6927
+    },
+    {
+      "epoch": 0.42618110236220474,
+      "grad_norm": 1.4379754066467285,
+      "learning_rate": 0.00012309155282245747,
+      "loss": 2.2973,
+      "step": 6928
+    },
+    {
+      "epoch": 0.4262426181102362,
+      "grad_norm": 0.9805158972740173,
+      "learning_rate": 0.0001230727373757612,
+      "loss": 2.7944,
+      "step": 6929
+    },
+    {
+      "epoch": 0.4263041338582677,
+      "grad_norm": 1.3215094804763794,
+      "learning_rate": 0.0001230539210662732,
+      "loss": 2.2309,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4263656496062992,
+      "grad_norm": 1.0982195138931274,
+      "learning_rate": 0.0001230351038946971,
+      "loss": 3.2304,
+      "step": 6931
+    },
+    {
+      "epoch": 0.42642716535433073,
+      "grad_norm": 1.0652259588241577,
+      "learning_rate": 0.00012301628586173653,
+      "loss": 2.7923,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4264886811023622,
+      "grad_norm": 1.191305160522461,
+      "learning_rate": 0.00012299746696809526,
+      "loss": 2.4307,
+      "step": 6933
+    },
+    {
+      "epoch": 0.4265501968503937,
+      "grad_norm": 1.364743947982788,
+      "learning_rate": 0.00012297864721447693,
+      "loss": 2.7158,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4266117125984252,
+      "grad_norm": 1.350317358970642,
+      "learning_rate": 0.00012295982660158537,
+      "loss": 2.599,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4266732283464567,
+      "grad_norm": 1.2468663454055786,
+      "learning_rate": 0.00012294100513012425,
+      "loss": 3.0249,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4267347440944882,
+      "grad_norm": 1.1131064891815186,
+      "learning_rate": 0.00012292218280079753,
+      "loss": 2.9012,
+      "step": 6937
+    },
+    {
+      "epoch": 0.4267962598425197,
+      "grad_norm": 1.1289973258972168,
+      "learning_rate": 0.00012290335961430898,
+      "loss": 2.6093,
+      "step": 6938
+    },
+    {
+      "epoch": 0.42685777559055116,
+      "grad_norm": 1.0486177206039429,
+      "learning_rate": 0.00012288453557136246,
+      "loss": 2.9587,
+      "step": 6939
+    },
+    {
+      "epoch": 0.4269192913385827,
+      "grad_norm": 1.0173640251159668,
+      "learning_rate": 0.00012286571067266194,
+      "loss": 2.3625,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4269808070866142,
+      "grad_norm": 1.219332218170166,
+      "learning_rate": 0.00012284688491891134,
+      "loss": 2.5306,
+      "step": 6941
+    },
+    {
+      "epoch": 0.42704232283464566,
+      "grad_norm": 1.209686517715454,
+      "learning_rate": 0.00012282805831081463,
+      "loss": 2.4203,
+      "step": 6942
+    },
+    {
+      "epoch": 0.42710383858267714,
+      "grad_norm": 1.3207803964614868,
+      "learning_rate": 0.00012280923084907584,
+      "loss": 2.4745,
+      "step": 6943
+    },
+    {
+      "epoch": 0.4271653543307087,
+      "grad_norm": 1.5176093578338623,
+      "learning_rate": 0.00012279040253439902,
+      "loss": 2.7782,
+      "step": 6944
+    },
+    {
+      "epoch": 0.42722687007874016,
+      "grad_norm": 1.1669846773147583,
+      "learning_rate": 0.00012277157336748818,
+      "loss": 3.1851,
+      "step": 6945
+    },
+    {
+      "epoch": 0.42728838582677164,
+      "grad_norm": 1.2652240991592407,
+      "learning_rate": 0.00012275274334904755,
+      "loss": 2.3077,
+      "step": 6946
+    },
+    {
+      "epoch": 0.4273499015748031,
+      "grad_norm": 1.0378745794296265,
+      "learning_rate": 0.00012273391247978114,
+      "loss": 2.4449,
+      "step": 6947
+    },
+    {
+      "epoch": 0.42741141732283466,
+      "grad_norm": 1.1951764822006226,
+      "learning_rate": 0.00012271508076039316,
+      "loss": 2.4321,
+      "step": 6948
+    },
+    {
+      "epoch": 0.42747293307086615,
+      "grad_norm": 1.2423676252365112,
+      "learning_rate": 0.00012269624819158785,
+      "loss": 2.2492,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4275344488188976,
+      "grad_norm": 2.176313877105713,
+      "learning_rate": 0.0001226774147740694,
+      "loss": 1.9117,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4275959645669291,
+      "grad_norm": 1.349036693572998,
+      "learning_rate": 0.00012265858050854208,
+      "loss": 1.7979,
+      "step": 6951
+    },
+    {
+      "epoch": 0.42765748031496065,
+      "grad_norm": 1.4407473802566528,
+      "learning_rate": 0.00012263974539571023,
+      "loss": 3.3098,
+      "step": 6952
+    },
+    {
+      "epoch": 0.42771899606299213,
+      "grad_norm": 1.2230315208435059,
+      "learning_rate": 0.00012262090943627812,
+      "loss": 3.0547,
+      "step": 6953
+    },
+    {
+      "epoch": 0.4277805118110236,
+      "grad_norm": 1.1779378652572632,
+      "learning_rate": 0.00012260207263095017,
+      "loss": 2.5701,
+      "step": 6954
+    },
+    {
+      "epoch": 0.4278420275590551,
+      "grad_norm": 1.1627355813980103,
+      "learning_rate": 0.0001225832349804307,
+      "loss": 2.7567,
+      "step": 6955
+    },
+    {
+      "epoch": 0.42790354330708663,
+      "grad_norm": 1.4327890872955322,
+      "learning_rate": 0.00012256439648542416,
+      "loss": 2.4928,
+      "step": 6956
+    },
+    {
+      "epoch": 0.4279650590551181,
+      "grad_norm": 1.3839768171310425,
+      "learning_rate": 0.00012254555714663506,
+      "loss": 2.8999,
+      "step": 6957
+    },
+    {
+      "epoch": 0.4280265748031496,
+      "grad_norm": 1.184889793395996,
+      "learning_rate": 0.0001225267169647678,
+      "loss": 2.1658,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4280880905511811,
+      "grad_norm": 1.1563637256622314,
+      "learning_rate": 0.00012250787594052695,
+      "loss": 2.6354,
+      "step": 6959
+    },
+    {
+      "epoch": 0.4281496062992126,
+      "grad_norm": 1.2844438552856445,
+      "learning_rate": 0.00012248903407461708,
+      "loss": 2.3166,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4282111220472441,
+      "grad_norm": 1.2603497505187988,
+      "learning_rate": 0.00012247019136774268,
+      "loss": 3.1036,
+      "step": 6961
+    },
+    {
+      "epoch": 0.4282726377952756,
+      "grad_norm": 1.021361231803894,
+      "learning_rate": 0.00012245134782060845,
+      "loss": 2.4867,
+      "step": 6962
+    },
+    {
+      "epoch": 0.42833415354330706,
+      "grad_norm": 1.2378233671188354,
+      "learning_rate": 0.00012243250343391902,
+      "loss": 2.3733,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4283956692913386,
+      "grad_norm": 1.1183732748031616,
+      "learning_rate": 0.000122413658208379,
+      "loss": 3.0768,
+      "step": 6964
+    },
+    {
+      "epoch": 0.4284571850393701,
+      "grad_norm": 1.1489909887313843,
+      "learning_rate": 0.00012239481214469317,
+      "loss": 1.4906,
+      "step": 6965
+    },
+    {
+      "epoch": 0.42851870078740156,
+      "grad_norm": 1.3273217678070068,
+      "learning_rate": 0.00012237596524356626,
+      "loss": 2.5625,
+      "step": 6966
+    },
+    {
+      "epoch": 0.42858021653543305,
+      "grad_norm": 1.074753761291504,
+      "learning_rate": 0.000122357117505703,
+      "loss": 2.4144,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4286417322834646,
+      "grad_norm": 1.216833233833313,
+      "learning_rate": 0.0001223382689318082,
+      "loss": 2.3627,
+      "step": 6968
+    },
+    {
+      "epoch": 0.42870324803149606,
+      "grad_norm": 1.7644507884979248,
+      "learning_rate": 0.00012231941952258675,
+      "loss": 1.909,
+      "step": 6969
+    },
+    {
+      "epoch": 0.42876476377952755,
+      "grad_norm": 1.1173059940338135,
+      "learning_rate": 0.00012230056927874344,
+      "loss": 2.5845,
+      "step": 6970
+    },
+    {
+      "epoch": 0.42882627952755903,
+      "grad_norm": 1.130960464477539,
+      "learning_rate": 0.00012228171820098319,
+      "loss": 2.7563,
+      "step": 6971
+    },
+    {
+      "epoch": 0.42888779527559057,
+      "grad_norm": 0.992050051689148,
+      "learning_rate": 0.0001222628662900109,
+      "loss": 2.8582,
+      "step": 6972
+    },
+    {
+      "epoch": 0.42894931102362205,
+      "grad_norm": 1.146539330482483,
+      "learning_rate": 0.0001222440135465316,
+      "loss": 3.0628,
+      "step": 6973
+    },
+    {
+      "epoch": 0.42901082677165353,
+      "grad_norm": 1.0445929765701294,
+      "learning_rate": 0.0001222251599712502,
+      "loss": 1.7849,
+      "step": 6974
+    },
+    {
+      "epoch": 0.429072342519685,
+      "grad_norm": 1.262829065322876,
+      "learning_rate": 0.00012220630556487175,
+      "loss": 2.6161,
+      "step": 6975
+    },
+    {
+      "epoch": 0.42913385826771655,
+      "grad_norm": 1.2406837940216064,
+      "learning_rate": 0.00012218745032810136,
+      "loss": 2.3952,
+      "step": 6976
+    },
+    {
+      "epoch": 0.42919537401574803,
+      "grad_norm": 1.2055994272232056,
+      "learning_rate": 0.00012216859426164396,
+      "loss": 2.6128,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4292568897637795,
+      "grad_norm": 1.258163571357727,
+      "learning_rate": 0.0001221497373662048,
+      "loss": 2.5736,
+      "step": 6978
+    },
+    {
+      "epoch": 0.429318405511811,
+      "grad_norm": 1.0202784538269043,
+      "learning_rate": 0.00012213087964248895,
+      "loss": 2.9882,
+      "step": 6979
+    },
+    {
+      "epoch": 0.42937992125984253,
+      "grad_norm": 1.1213109493255615,
+      "learning_rate": 0.00012211202109120158,
+      "loss": 2.8783,
+      "step": 6980
+    },
+    {
+      "epoch": 0.429441437007874,
+      "grad_norm": 1.551295518875122,
+      "learning_rate": 0.00012209316171304797,
+      "loss": 2.0542,
+      "step": 6981
+    },
+    {
+      "epoch": 0.4295029527559055,
+      "grad_norm": 1.5103853940963745,
+      "learning_rate": 0.0001220743015087333,
+      "loss": 2.5946,
+      "step": 6982
+    },
+    {
+      "epoch": 0.429564468503937,
+      "grad_norm": 1.2213523387908936,
+      "learning_rate": 0.0001220554404789628,
+      "loss": 2.5596,
+      "step": 6983
+    },
+    {
+      "epoch": 0.4296259842519685,
+      "grad_norm": 1.494336485862732,
+      "learning_rate": 0.00012203657862444187,
+      "loss": 2.6984,
+      "step": 6984
+    },
+    {
+      "epoch": 0.4296875,
+      "grad_norm": 1.1665375232696533,
+      "learning_rate": 0.00012201771594587574,
+      "loss": 2.4217,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4297490157480315,
+      "grad_norm": 1.7508245706558228,
+      "learning_rate": 0.0001219988524439698,
+      "loss": 2.328,
+      "step": 6986
+    },
+    {
+      "epoch": 0.429810531496063,
+      "grad_norm": 1.1378891468048096,
+      "learning_rate": 0.00012197998811942948,
+      "loss": 2.6296,
+      "step": 6987
+    },
+    {
+      "epoch": 0.4298720472440945,
+      "grad_norm": 1.264721393585205,
+      "learning_rate": 0.00012196112297296016,
+      "loss": 2.128,
+      "step": 6988
+    },
+    {
+      "epoch": 0.429933562992126,
+      "grad_norm": 1.2267109155654907,
+      "learning_rate": 0.0001219422570052673,
+      "loss": 2.1298,
+      "step": 6989
+    },
+    {
+      "epoch": 0.42999507874015747,
+      "grad_norm": 1.2985330820083618,
+      "learning_rate": 0.00012192339021705637,
+      "loss": 2.2405,
+      "step": 6990
+    },
+    {
+      "epoch": 0.430056594488189,
+      "grad_norm": 1.2002582550048828,
+      "learning_rate": 0.00012190452260903289,
+      "loss": 2.5778,
+      "step": 6991
+    },
+    {
+      "epoch": 0.4301181102362205,
+      "grad_norm": 1.0283262729644775,
+      "learning_rate": 0.00012188565418190243,
+      "loss": 2.5588,
+      "step": 6992
+    },
+    {
+      "epoch": 0.43017962598425197,
+      "grad_norm": 1.1672319173812866,
+      "learning_rate": 0.00012186678493637051,
+      "loss": 2.618,
+      "step": 6993
+    },
+    {
+      "epoch": 0.43024114173228345,
+      "grad_norm": 1.1074950695037842,
+      "learning_rate": 0.00012184791487314279,
+      "loss": 2.6224,
+      "step": 6994
+    },
+    {
+      "epoch": 0.430302657480315,
+      "grad_norm": 1.114138126373291,
+      "learning_rate": 0.00012182904399292489,
+      "loss": 2.6247,
+      "step": 6995
+    },
+    {
+      "epoch": 0.43036417322834647,
+      "grad_norm": 1.0254430770874023,
+      "learning_rate": 0.00012181017229642244,
+      "loss": 2.8391,
+      "step": 6996
+    },
+    {
+      "epoch": 0.43042568897637795,
+      "grad_norm": 1.1879688501358032,
+      "learning_rate": 0.00012179129978434119,
+      "loss": 2.2514,
+      "step": 6997
+    },
+    {
+      "epoch": 0.43048720472440943,
+      "grad_norm": 1.2050609588623047,
+      "learning_rate": 0.00012177242645738682,
+      "loss": 2.2766,
+      "step": 6998
+    },
+    {
+      "epoch": 0.43054872047244097,
+      "grad_norm": 1.3303614854812622,
+      "learning_rate": 0.00012175355231626508,
+      "loss": 3.2614,
+      "step": 6999
+    },
+    {
+      "epoch": 0.43061023622047245,
+      "grad_norm": 0.9404337406158447,
+      "learning_rate": 0.00012173467736168182,
+      "loss": 3.0306,
+      "step": 7000
+    },
+    {
+      "epoch": 0.43067175196850394,
+      "grad_norm": 1.1103935241699219,
+      "learning_rate": 0.00012171580159434283,
+      "loss": 2.6362,
+      "step": 7001
+    },
+    {
+      "epoch": 0.4307332677165354,
+      "grad_norm": 1.355753779411316,
+      "learning_rate": 0.00012169692501495395,
+      "loss": 2.6966,
+      "step": 7002
+    },
+    {
+      "epoch": 0.43079478346456695,
+      "grad_norm": 1.1670459508895874,
+      "learning_rate": 0.00012167804762422104,
+      "loss": 2.3177,
+      "step": 7003
+    },
+    {
+      "epoch": 0.43085629921259844,
+      "grad_norm": 1.0329539775848389,
+      "learning_rate": 0.00012165916942285005,
+      "loss": 2.8563,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4309178149606299,
+      "grad_norm": 1.2866746187210083,
+      "learning_rate": 0.00012164029041154687,
+      "loss": 3.0104,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4309793307086614,
+      "grad_norm": 1.2891361713409424,
+      "learning_rate": 0.00012162141059101753,
+      "loss": 2.3023,
+      "step": 7006
+    },
+    {
+      "epoch": 0.43104084645669294,
+      "grad_norm": 1.1596152782440186,
+      "learning_rate": 0.00012160252996196797,
+      "loss": 3.211,
+      "step": 7007
+    },
+    {
+      "epoch": 0.4311023622047244,
+      "grad_norm": 1.0343667268753052,
+      "learning_rate": 0.00012158364852510427,
+      "loss": 3.033,
+      "step": 7008
+    },
+    {
+      "epoch": 0.4311638779527559,
+      "grad_norm": 1.3588271141052246,
+      "learning_rate": 0.00012156476628113245,
+      "loss": 2.9897,
+      "step": 7009
+    },
+    {
+      "epoch": 0.4312253937007874,
+      "grad_norm": 1.1370306015014648,
+      "learning_rate": 0.00012154588323075863,
+      "loss": 2.5707,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4312869094488189,
+      "grad_norm": 1.3938848972320557,
+      "learning_rate": 0.00012152699937468895,
+      "loss": 3.018,
+      "step": 7011
+    },
+    {
+      "epoch": 0.4313484251968504,
+      "grad_norm": 1.2585843801498413,
+      "learning_rate": 0.00012150811471362948,
+      "loss": 2.4839,
+      "step": 7012
+    },
+    {
+      "epoch": 0.4314099409448819,
+      "grad_norm": 1.21547532081604,
+      "learning_rate": 0.0001214892292482865,
+      "loss": 2.588,
+      "step": 7013
+    },
+    {
+      "epoch": 0.43147145669291337,
+      "grad_norm": 1.5154510736465454,
+      "learning_rate": 0.00012147034297936614,
+      "loss": 2.3198,
+      "step": 7014
+    },
+    {
+      "epoch": 0.4315329724409449,
+      "grad_norm": 0.9856167435646057,
+      "learning_rate": 0.00012145145590757467,
+      "loss": 2.3729,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4315944881889764,
+      "grad_norm": 1.116533875465393,
+      "learning_rate": 0.00012143256803361838,
+      "loss": 2.9238,
+      "step": 7016
+    },
+    {
+      "epoch": 0.43165600393700787,
+      "grad_norm": 1.169195532798767,
+      "learning_rate": 0.00012141367935820356,
+      "loss": 2.6837,
+      "step": 7017
+    },
+    {
+      "epoch": 0.43171751968503935,
+      "grad_norm": 1.1426366567611694,
+      "learning_rate": 0.00012139478988203652,
+      "loss": 2.7316,
+      "step": 7018
+    },
+    {
+      "epoch": 0.4317790354330709,
+      "grad_norm": 1.2899088859558105,
+      "learning_rate": 0.00012137589960582365,
+      "loss": 2.5265,
+      "step": 7019
+    },
+    {
+      "epoch": 0.4318405511811024,
+      "grad_norm": 1.1326189041137695,
+      "learning_rate": 0.00012135700853027135,
+      "loss": 3.0639,
+      "step": 7020
+    },
+    {
+      "epoch": 0.43190206692913385,
+      "grad_norm": 1.1627920866012573,
+      "learning_rate": 0.00012133811665608599,
+      "loss": 2.8512,
+      "step": 7021
+    },
+    {
+      "epoch": 0.43196358267716534,
+      "grad_norm": 0.9928020238876343,
+      "learning_rate": 0.00012131922398397406,
+      "loss": 3.3588,
+      "step": 7022
+    },
+    {
+      "epoch": 0.4320250984251969,
+      "grad_norm": 1.1421645879745483,
+      "learning_rate": 0.00012130033051464204,
+      "loss": 2.8508,
+      "step": 7023
+    },
+    {
+      "epoch": 0.43208661417322836,
+      "grad_norm": 1.4874497652053833,
+      "learning_rate": 0.00012128143624879644,
+      "loss": 2.2668,
+      "step": 7024
+    },
+    {
+      "epoch": 0.43214812992125984,
+      "grad_norm": 1.4555165767669678,
+      "learning_rate": 0.00012126254118714377,
+      "loss": 1.9052,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4322096456692913,
+      "grad_norm": 1.0918474197387695,
+      "learning_rate": 0.00012124364533039064,
+      "loss": 2.9274,
+      "step": 7026
+    },
+    {
+      "epoch": 0.43227116141732286,
+      "grad_norm": 1.0935596227645874,
+      "learning_rate": 0.00012122474867924363,
+      "loss": 2.8797,
+      "step": 7027
+    },
+    {
+      "epoch": 0.43233267716535434,
+      "grad_norm": 1.1776233911514282,
+      "learning_rate": 0.00012120585123440935,
+      "loss": 2.685,
+      "step": 7028
+    },
+    {
+      "epoch": 0.4323941929133858,
+      "grad_norm": 1.4966143369674683,
+      "learning_rate": 0.00012118695299659451,
+      "loss": 2.2044,
+      "step": 7029
+    },
+    {
+      "epoch": 0.4324557086614173,
+      "grad_norm": 1.0166382789611816,
+      "learning_rate": 0.00012116805396650577,
+      "loss": 3.2191,
+      "step": 7030
+    },
+    {
+      "epoch": 0.43251722440944884,
+      "grad_norm": 1.1562596559524536,
+      "learning_rate": 0.00012114915414484981,
+      "loss": 2.3611,
+      "step": 7031
+    },
+    {
+      "epoch": 0.4325787401574803,
+      "grad_norm": 1.3181148767471313,
+      "learning_rate": 0.00012113025353233345,
+      "loss": 2.7011,
+      "step": 7032
+    },
+    {
+      "epoch": 0.4326402559055118,
+      "grad_norm": 1.0377025604248047,
+      "learning_rate": 0.00012111135212966344,
+      "loss": 2.6747,
+      "step": 7033
+    },
+    {
+      "epoch": 0.4327017716535433,
+      "grad_norm": 1.2649480104446411,
+      "learning_rate": 0.00012109244993754654,
+      "loss": 2.1055,
+      "step": 7034
+    },
+    {
+      "epoch": 0.4327632874015748,
+      "grad_norm": 1.1698663234710693,
+      "learning_rate": 0.00012107354695668967,
+      "loss": 2.6211,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4328248031496063,
+      "grad_norm": 1.100456953048706,
+      "learning_rate": 0.00012105464318779965,
+      "loss": 2.8699,
+      "step": 7036
+    },
+    {
+      "epoch": 0.4328863188976378,
+      "grad_norm": 1.7050143480300903,
+      "learning_rate": 0.00012103573863158337,
+      "loss": 2.3573,
+      "step": 7037
+    },
+    {
+      "epoch": 0.43294783464566927,
+      "grad_norm": 1.259788155555725,
+      "learning_rate": 0.00012101683328874773,
+      "loss": 2.7441,
+      "step": 7038
+    },
+    {
+      "epoch": 0.4330093503937008,
+      "grad_norm": 0.9732722043991089,
+      "learning_rate": 0.00012099792715999976,
+      "loss": 3.2671,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4330708661417323,
+      "grad_norm": 1.1038271188735962,
+      "learning_rate": 0.0001209790202460464,
+      "loss": 1.5304,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4331323818897638,
+      "grad_norm": 1.2941348552703857,
+      "learning_rate": 0.00012096011254759462,
+      "loss": 2.5665,
+      "step": 7041
+    },
+    {
+      "epoch": 0.43319389763779526,
+      "grad_norm": 1.3657448291778564,
+      "learning_rate": 0.00012094120406535156,
+      "loss": 2.2306,
+      "step": 7042
+    },
+    {
+      "epoch": 0.4332554133858268,
+      "grad_norm": 1.0579688549041748,
+      "learning_rate": 0.00012092229480002424,
+      "loss": 3.1885,
+      "step": 7043
+    },
+    {
+      "epoch": 0.4333169291338583,
+      "grad_norm": 1.285692811012268,
+      "learning_rate": 0.00012090338475231973,
+      "loss": 2.5204,
+      "step": 7044
+    },
+    {
+      "epoch": 0.43337844488188976,
+      "grad_norm": 1.093909502029419,
+      "learning_rate": 0.0001208844739229452,
+      "loss": 2.9465,
+      "step": 7045
+    },
+    {
+      "epoch": 0.43343996062992124,
+      "grad_norm": 1.3970143795013428,
+      "learning_rate": 0.00012086556231260782,
+      "loss": 2.3047,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4335014763779528,
+      "grad_norm": 1.1407122611999512,
+      "learning_rate": 0.00012084664992201471,
+      "loss": 2.6771,
+      "step": 7047
+    },
+    {
+      "epoch": 0.43356299212598426,
+      "grad_norm": 1.0469684600830078,
+      "learning_rate": 0.00012082773675187317,
+      "loss": 2.896,
+      "step": 7048
+    },
+    {
+      "epoch": 0.43362450787401574,
+      "grad_norm": 1.6786179542541504,
+      "learning_rate": 0.00012080882280289045,
+      "loss": 3.3432,
+      "step": 7049
+    },
+    {
+      "epoch": 0.4336860236220472,
+      "grad_norm": 1.2435276508331299,
+      "learning_rate": 0.00012078990807577373,
+      "loss": 2.4454,
+      "step": 7050
+    },
+    {
+      "epoch": 0.43374753937007876,
+      "grad_norm": 1.1956802606582642,
+      "learning_rate": 0.00012077099257123042,
+      "loss": 2.2809,
+      "step": 7051
+    },
+    {
+      "epoch": 0.43380905511811024,
+      "grad_norm": 1.0416388511657715,
+      "learning_rate": 0.00012075207628996779,
+      "loss": 2.5015,
+      "step": 7052
+    },
+    {
+      "epoch": 0.4338705708661417,
+      "grad_norm": 1.1194076538085938,
+      "learning_rate": 0.00012073315923269319,
+      "loss": 2.4938,
+      "step": 7053
+    },
+    {
+      "epoch": 0.4339320866141732,
+      "grad_norm": 1.2051868438720703,
+      "learning_rate": 0.00012071424140011409,
+      "loss": 2.5274,
+      "step": 7054
+    },
+    {
+      "epoch": 0.43399360236220474,
+      "grad_norm": 1.0260369777679443,
+      "learning_rate": 0.00012069532279293787,
+      "loss": 2.665,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4340551181102362,
+      "grad_norm": 1.0112473964691162,
+      "learning_rate": 0.00012067640341187196,
+      "loss": 2.6483,
+      "step": 7056
+    },
+    {
+      "epoch": 0.4341166338582677,
+      "grad_norm": 1.263420581817627,
+      "learning_rate": 0.00012065748325762385,
+      "loss": 1.6791,
+      "step": 7057
+    },
+    {
+      "epoch": 0.4341781496062992,
+      "grad_norm": 1.4876956939697266,
+      "learning_rate": 0.00012063856233090107,
+      "loss": 2.446,
+      "step": 7058
+    },
+    {
+      "epoch": 0.43423966535433073,
+      "grad_norm": 1.2162725925445557,
+      "learning_rate": 0.00012061964063241115,
+      "loss": 2.395,
+      "step": 7059
+    },
+    {
+      "epoch": 0.4343011811023622,
+      "grad_norm": 1.025585412979126,
+      "learning_rate": 0.00012060071816286162,
+      "loss": 2.772,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4343626968503937,
+      "grad_norm": 1.145734190940857,
+      "learning_rate": 0.00012058179492296013,
+      "loss": 2.0034,
+      "step": 7061
+    },
+    {
+      "epoch": 0.4344242125984252,
+      "grad_norm": 1.2003819942474365,
+      "learning_rate": 0.00012056287091341428,
+      "loss": 2.7259,
+      "step": 7062
+    },
+    {
+      "epoch": 0.4344857283464567,
+      "grad_norm": 1.0789873600006104,
+      "learning_rate": 0.0001205439461349317,
+      "loss": 2.9956,
+      "step": 7063
+    },
+    {
+      "epoch": 0.4345472440944882,
+      "grad_norm": 0.9546241164207458,
+      "learning_rate": 0.00012052502058822008,
+      "loss": 2.9987,
+      "step": 7064
+    },
+    {
+      "epoch": 0.4346087598425197,
+      "grad_norm": 1.1777452230453491,
+      "learning_rate": 0.00012050609427398718,
+      "loss": 2.454,
+      "step": 7065
+    },
+    {
+      "epoch": 0.43467027559055116,
+      "grad_norm": 1.0669773817062378,
+      "learning_rate": 0.00012048716719294066,
+      "loss": 2.4194,
+      "step": 7066
+    },
+    {
+      "epoch": 0.4347317913385827,
+      "grad_norm": 1.2090517282485962,
+      "learning_rate": 0.00012046823934578835,
+      "loss": 2.5877,
+      "step": 7067
+    },
+    {
+      "epoch": 0.4347933070866142,
+      "grad_norm": 1.6355810165405273,
+      "learning_rate": 0.00012044931073323802,
+      "loss": 2.093,
+      "step": 7068
+    },
+    {
+      "epoch": 0.43485482283464566,
+      "grad_norm": 1.1965821981430054,
+      "learning_rate": 0.00012043038135599752,
+      "loss": 2.6092,
+      "step": 7069
+    },
+    {
+      "epoch": 0.43491633858267714,
+      "grad_norm": 1.3105109930038452,
+      "learning_rate": 0.00012041145121477462,
+      "loss": 2.3692,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4349778543307087,
+      "grad_norm": 1.0362977981567383,
+      "learning_rate": 0.00012039252031027731,
+      "loss": 3.1772,
+      "step": 7071
+    },
+    {
+      "epoch": 0.43503937007874016,
+      "grad_norm": 1.8564778566360474,
+      "learning_rate": 0.00012037358864321344,
+      "loss": 2.3171,
+      "step": 7072
+    },
+    {
+      "epoch": 0.43510088582677164,
+      "grad_norm": 1.228662371635437,
+      "learning_rate": 0.00012035465621429093,
+      "loss": 2.3308,
+      "step": 7073
+    },
+    {
+      "epoch": 0.4351624015748031,
+      "grad_norm": 1.080881953239441,
+      "learning_rate": 0.00012033572302421783,
+      "loss": 3.3849,
+      "step": 7074
+    },
+    {
+      "epoch": 0.43522391732283466,
+      "grad_norm": 1.2045519351959229,
+      "learning_rate": 0.00012031678907370207,
+      "loss": 2.5481,
+      "step": 7075
+    },
+    {
+      "epoch": 0.43528543307086615,
+      "grad_norm": 1.6561299562454224,
+      "learning_rate": 0.00012029785436345167,
+      "loss": 2.6816,
+      "step": 7076
+    },
+    {
+      "epoch": 0.4353469488188976,
+      "grad_norm": 1.2394176721572876,
+      "learning_rate": 0.0001202789188941747,
+      "loss": 3.2943,
+      "step": 7077
+    },
+    {
+      "epoch": 0.4354084645669291,
+      "grad_norm": 1.0766545534133911,
+      "learning_rate": 0.00012025998266657927,
+      "loss": 2.9846,
+      "step": 7078
+    },
+    {
+      "epoch": 0.43546998031496065,
+      "grad_norm": 1.286528468132019,
+      "learning_rate": 0.00012024104568137343,
+      "loss": 2.7241,
+      "step": 7079
+    },
+    {
+      "epoch": 0.43553149606299213,
+      "grad_norm": 1.3494936227798462,
+      "learning_rate": 0.00012022210793926536,
+      "loss": 3.1321,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4355930118110236,
+      "grad_norm": 1.23129403591156,
+      "learning_rate": 0.00012020316944096324,
+      "loss": 2.7559,
+      "step": 7081
+    },
+    {
+      "epoch": 0.4356545275590551,
+      "grad_norm": 1.1242988109588623,
+      "learning_rate": 0.00012018423018717522,
+      "loss": 2.6455,
+      "step": 7082
+    },
+    {
+      "epoch": 0.43571604330708663,
+      "grad_norm": 1.1114197969436646,
+      "learning_rate": 0.0001201652901786095,
+      "loss": 3.1626,
+      "step": 7083
+    },
+    {
+      "epoch": 0.4357775590551181,
+      "grad_norm": 1.311197280883789,
+      "learning_rate": 0.00012014634941597442,
+      "loss": 2.6442,
+      "step": 7084
+    },
+    {
+      "epoch": 0.4358390748031496,
+      "grad_norm": 1.141365647315979,
+      "learning_rate": 0.00012012740789997821,
+      "loss": 2.5279,
+      "step": 7085
+    },
+    {
+      "epoch": 0.4359005905511811,
+      "grad_norm": 1.1763490438461304,
+      "learning_rate": 0.00012010846563132918,
+      "loss": 3.1324,
+      "step": 7086
+    },
+    {
+      "epoch": 0.4359621062992126,
+      "grad_norm": 1.1047377586364746,
+      "learning_rate": 0.00012008952261073565,
+      "loss": 3.1894,
+      "step": 7087
+    },
+    {
+      "epoch": 0.4360236220472441,
+      "grad_norm": 1.1694533824920654,
+      "learning_rate": 0.000120070578838906,
+      "loss": 2.1923,
+      "step": 7088
+    },
+    {
+      "epoch": 0.4360851377952756,
+      "grad_norm": 1.1720759868621826,
+      "learning_rate": 0.00012005163431654862,
+      "loss": 2.4285,
+      "step": 7089
+    },
+    {
+      "epoch": 0.43614665354330706,
+      "grad_norm": 1.16497802734375,
+      "learning_rate": 0.00012003268904437193,
+      "loss": 2.425,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4362081692913386,
+      "grad_norm": 1.3612741231918335,
+      "learning_rate": 0.00012001374302308436,
+      "loss": 3.1504,
+      "step": 7091
+    },
+    {
+      "epoch": 0.4362696850393701,
+      "grad_norm": 1.497312068939209,
+      "learning_rate": 0.0001199947962533944,
+      "loss": 2.3727,
+      "step": 7092
+    },
+    {
+      "epoch": 0.43633120078740156,
+      "grad_norm": 1.2075265645980835,
+      "learning_rate": 0.00011997584873601058,
+      "loss": 2.6619,
+      "step": 7093
+    },
+    {
+      "epoch": 0.43639271653543305,
+      "grad_norm": 1.281172275543213,
+      "learning_rate": 0.00011995690047164139,
+      "loss": 2.88,
+      "step": 7094
+    },
+    {
+      "epoch": 0.4364542322834646,
+      "grad_norm": 1.0612024068832397,
+      "learning_rate": 0.00011993795146099537,
+      "loss": 2.5324,
+      "step": 7095
+    },
+    {
+      "epoch": 0.43651574803149606,
+      "grad_norm": 1.3100106716156006,
+      "learning_rate": 0.00011991900170478122,
+      "loss": 2.6338,
+      "step": 7096
+    },
+    {
+      "epoch": 0.43657726377952755,
+      "grad_norm": 1.0507242679595947,
+      "learning_rate": 0.00011990005120370743,
+      "loss": 3.1588,
+      "step": 7097
+    },
+    {
+      "epoch": 0.43663877952755903,
+      "grad_norm": 2.429821252822876,
+      "learning_rate": 0.00011988109995848267,
+      "loss": 2.4593,
+      "step": 7098
+    },
+    {
+      "epoch": 0.43670029527559057,
+      "grad_norm": 1.5973351001739502,
+      "learning_rate": 0.00011986214796981565,
+      "loss": 2.9513,
+      "step": 7099
+    },
+    {
+      "epoch": 0.43676181102362205,
+      "grad_norm": 1.138907551765442,
+      "learning_rate": 0.00011984319523841507,
+      "loss": 3.1553,
+      "step": 7100
+    },
+    {
+      "epoch": 0.43682332677165353,
+      "grad_norm": 1.2300089597702026,
+      "learning_rate": 0.0001198242417649896,
+      "loss": 2.9118,
+      "step": 7101
+    },
+    {
+      "epoch": 0.436884842519685,
+      "grad_norm": 1.264133334159851,
+      "learning_rate": 0.00011980528755024807,
+      "loss": 2.7968,
+      "step": 7102
+    },
+    {
+      "epoch": 0.43694635826771655,
+      "grad_norm": 1.2686898708343506,
+      "learning_rate": 0.00011978633259489922,
+      "loss": 2.9293,
+      "step": 7103
+    },
+    {
+      "epoch": 0.43700787401574803,
+      "grad_norm": 1.3213648796081543,
+      "learning_rate": 0.00011976737689965185,
+      "loss": 2.8714,
+      "step": 7104
+    },
+    {
+      "epoch": 0.4370693897637795,
+      "grad_norm": 1.3442591428756714,
+      "learning_rate": 0.00011974842046521479,
+      "loss": 3.028,
+      "step": 7105
+    },
+    {
+      "epoch": 0.437130905511811,
+      "grad_norm": 1.186560869216919,
+      "learning_rate": 0.00011972946329229697,
+      "loss": 2.8141,
+      "step": 7106
+    },
+    {
+      "epoch": 0.43719242125984253,
+      "grad_norm": 1.165114402770996,
+      "learning_rate": 0.00011971050538160721,
+      "loss": 3.3381,
+      "step": 7107
+    },
+    {
+      "epoch": 0.437253937007874,
+      "grad_norm": 1.518225908279419,
+      "learning_rate": 0.00011969154673385445,
+      "loss": 2.1052,
+      "step": 7108
+    },
+    {
+      "epoch": 0.4373154527559055,
+      "grad_norm": 1.3722882270812988,
+      "learning_rate": 0.00011967258734974767,
+      "loss": 2.4478,
+      "step": 7109
+    },
+    {
+      "epoch": 0.437376968503937,
+      "grad_norm": 0.9641464948654175,
+      "learning_rate": 0.00011965362722999581,
+      "loss": 2.5874,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4374384842519685,
+      "grad_norm": 1.1141730546951294,
+      "learning_rate": 0.00011963466637530785,
+      "loss": 2.1888,
+      "step": 7111
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 1.0608575344085693,
+      "learning_rate": 0.0001196157047863929,
+      "loss": 2.2662,
+      "step": 7112
+    },
+    {
+      "epoch": 0.4375615157480315,
+      "grad_norm": 1.2090116739273071,
+      "learning_rate": 0.00011959674246395995,
+      "loss": 2.2588,
+      "step": 7113
+    },
+    {
+      "epoch": 0.437623031496063,
+      "grad_norm": 1.3049064874649048,
+      "learning_rate": 0.0001195777794087181,
+      "loss": 2.6892,
+      "step": 7114
+    },
+    {
+      "epoch": 0.4376845472440945,
+      "grad_norm": 0.9880239367485046,
+      "learning_rate": 0.00011955881562137648,
+      "loss": 2.7545,
+      "step": 7115
+    },
+    {
+      "epoch": 0.437746062992126,
+      "grad_norm": 1.3604129552841187,
+      "learning_rate": 0.00011953985110264423,
+      "loss": 1.9825,
+      "step": 7116
+    },
+    {
+      "epoch": 0.43780757874015747,
+      "grad_norm": 1.2600555419921875,
+      "learning_rate": 0.00011952088585323049,
+      "loss": 2.6103,
+      "step": 7117
+    },
+    {
+      "epoch": 0.437869094488189,
+      "grad_norm": 1.573250412940979,
+      "learning_rate": 0.00011950191987384447,
+      "loss": 2.1017,
+      "step": 7118
+    },
+    {
+      "epoch": 0.4379306102362205,
+      "grad_norm": 1.4792803525924683,
+      "learning_rate": 0.0001194829531651954,
+      "loss": 2.7609,
+      "step": 7119
+    },
+    {
+      "epoch": 0.43799212598425197,
+      "grad_norm": 1.4353591203689575,
+      "learning_rate": 0.00011946398572799251,
+      "loss": 2.4947,
+      "step": 7120
+    },
+    {
+      "epoch": 0.43805364173228345,
+      "grad_norm": 1.2959630489349365,
+      "learning_rate": 0.00011944501756294508,
+      "loss": 2.733,
+      "step": 7121
+    },
+    {
+      "epoch": 0.438115157480315,
+      "grad_norm": 1.0670957565307617,
+      "learning_rate": 0.00011942604867076245,
+      "loss": 3.0026,
+      "step": 7122
+    },
+    {
+      "epoch": 0.43817667322834647,
+      "grad_norm": 1.1539281606674194,
+      "learning_rate": 0.00011940707905215392,
+      "loss": 2.7467,
+      "step": 7123
+    },
+    {
+      "epoch": 0.43823818897637795,
+      "grad_norm": 1.1544246673583984,
+      "learning_rate": 0.00011938810870782883,
+      "loss": 2.5118,
+      "step": 7124
+    },
+    {
+      "epoch": 0.43829970472440943,
+      "grad_norm": 1.123909592628479,
+      "learning_rate": 0.00011936913763849662,
+      "loss": 2.7307,
+      "step": 7125
+    },
+    {
+      "epoch": 0.43836122047244097,
+      "grad_norm": 1.1728512048721313,
+      "learning_rate": 0.00011935016584486665,
+      "loss": 2.385,
+      "step": 7126
+    },
+    {
+      "epoch": 0.43842273622047245,
+      "grad_norm": 0.9655293226242065,
+      "learning_rate": 0.00011933119332764835,
+      "loss": 2.7165,
+      "step": 7127
+    },
+    {
+      "epoch": 0.43848425196850394,
+      "grad_norm": 1.2217143774032593,
+      "learning_rate": 0.00011931222008755122,
+      "loss": 2.7318,
+      "step": 7128
+    },
+    {
+      "epoch": 0.4385457677165354,
+      "grad_norm": 1.3140840530395508,
+      "learning_rate": 0.00011929324612528478,
+      "loss": 2.2579,
+      "step": 7129
+    },
+    {
+      "epoch": 0.43860728346456695,
+      "grad_norm": 1.2232789993286133,
+      "learning_rate": 0.00011927427144155849,
+      "loss": 2.03,
+      "step": 7130
+    },
+    {
+      "epoch": 0.43866879921259844,
+      "grad_norm": 3.3285598754882812,
+      "learning_rate": 0.00011925529603708194,
+      "loss": 1.681,
+      "step": 7131
+    },
+    {
+      "epoch": 0.4387303149606299,
+      "grad_norm": 1.5776492357254028,
+      "learning_rate": 0.00011923631991256468,
+      "loss": 3.4079,
+      "step": 7132
+    },
+    {
+      "epoch": 0.4387918307086614,
+      "grad_norm": 1.2767186164855957,
+      "learning_rate": 0.0001192173430687163,
+      "loss": 1.9656,
+      "step": 7133
+    },
+    {
+      "epoch": 0.43885334645669294,
+      "grad_norm": 1.0749551057815552,
+      "learning_rate": 0.00011919836550624648,
+      "loss": 2.5289,
+      "step": 7134
+    },
+    {
+      "epoch": 0.4389148622047244,
+      "grad_norm": 1.1063836812973022,
+      "learning_rate": 0.00011917938722586484,
+      "loss": 2.7144,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4389763779527559,
+      "grad_norm": 1.0973597764968872,
+      "learning_rate": 0.00011916040822828103,
+      "loss": 2.0354,
+      "step": 7136
+    },
+    {
+      "epoch": 0.4390378937007874,
+      "grad_norm": 1.0838568210601807,
+      "learning_rate": 0.00011914142851420482,
+      "loss": 1.8081,
+      "step": 7137
+    },
+    {
+      "epoch": 0.4390994094488189,
+      "grad_norm": 1.077348232269287,
+      "learning_rate": 0.00011912244808434594,
+      "loss": 2.9547,
+      "step": 7138
+    },
+    {
+      "epoch": 0.4391609251968504,
+      "grad_norm": 1.3664768934249878,
+      "learning_rate": 0.00011910346693941412,
+      "loss": 2.5224,
+      "step": 7139
+    },
+    {
+      "epoch": 0.4392224409448819,
+      "grad_norm": 1.2688947916030884,
+      "learning_rate": 0.00011908448508011911,
+      "loss": 2.7177,
+      "step": 7140
+    },
+    {
+      "epoch": 0.43928395669291337,
+      "grad_norm": 1.2053114175796509,
+      "learning_rate": 0.00011906550250717083,
+      "loss": 2.6219,
+      "step": 7141
+    },
+    {
+      "epoch": 0.4393454724409449,
+      "grad_norm": 1.195398211479187,
+      "learning_rate": 0.00011904651922127907,
+      "loss": 2.3812,
+      "step": 7142
+    },
+    {
+      "epoch": 0.4394069881889764,
+      "grad_norm": 1.3708364963531494,
+      "learning_rate": 0.00011902753522315366,
+      "loss": 2.8756,
+      "step": 7143
+    },
+    {
+      "epoch": 0.43946850393700787,
+      "grad_norm": 1.2502259016036987,
+      "learning_rate": 0.00011900855051350457,
+      "loss": 2.262,
+      "step": 7144
+    },
+    {
+      "epoch": 0.43953001968503935,
+      "grad_norm": 1.1161712408065796,
+      "learning_rate": 0.00011898956509304166,
+      "loss": 2.8825,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4395915354330709,
+      "grad_norm": 1.1749244928359985,
+      "learning_rate": 0.00011897057896247492,
+      "loss": 2.0106,
+      "step": 7146
+    },
+    {
+      "epoch": 0.4396530511811024,
+      "grad_norm": 1.4153708219528198,
+      "learning_rate": 0.0001189515921225143,
+      "loss": 2.931,
+      "step": 7147
+    },
+    {
+      "epoch": 0.43971456692913385,
+      "grad_norm": 1.2032541036605835,
+      "learning_rate": 0.00011893260457386981,
+      "loss": 2.8399,
+      "step": 7148
+    },
+    {
+      "epoch": 0.43977608267716534,
+      "grad_norm": 1.1002966165542603,
+      "learning_rate": 0.00011891361631725148,
+      "loss": 2.7293,
+      "step": 7149
+    },
+    {
+      "epoch": 0.4398375984251969,
+      "grad_norm": 1.1804742813110352,
+      "learning_rate": 0.00011889462735336938,
+      "loss": 1.9825,
+      "step": 7150
+    },
+    {
+      "epoch": 0.43989911417322836,
+      "grad_norm": 1.2778346538543701,
+      "learning_rate": 0.00011887563768293357,
+      "loss": 2.5302,
+      "step": 7151
+    },
+    {
+      "epoch": 0.43996062992125984,
+      "grad_norm": 1.199497938156128,
+      "learning_rate": 0.00011885664730665418,
+      "loss": 2.629,
+      "step": 7152
+    },
+    {
+      "epoch": 0.4400221456692913,
+      "grad_norm": 1.2824193239212036,
+      "learning_rate": 0.00011883765622524128,
+      "loss": 2.1599,
+      "step": 7153
+    },
+    {
+      "epoch": 0.44008366141732286,
+      "grad_norm": 1.2248002290725708,
+      "learning_rate": 0.00011881866443940512,
+      "loss": 2.9058,
+      "step": 7154
+    },
+    {
+      "epoch": 0.44014517716535434,
+      "grad_norm": 1.4581111669540405,
+      "learning_rate": 0.00011879967194985584,
+      "loss": 2.9339,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4402066929133858,
+      "grad_norm": 1.2626214027404785,
+      "learning_rate": 0.00011878067875730363,
+      "loss": 2.6834,
+      "step": 7156
+    },
+    {
+      "epoch": 0.4402682086614173,
+      "grad_norm": 1.2205697298049927,
+      "learning_rate": 0.0001187616848624588,
+      "loss": 2.7472,
+      "step": 7157
+    },
+    {
+      "epoch": 0.44032972440944884,
+      "grad_norm": 1.2494611740112305,
+      "learning_rate": 0.00011874269026603156,
+      "loss": 2.6021,
+      "step": 7158
+    },
+    {
+      "epoch": 0.4403912401574803,
+      "grad_norm": 1.156605839729309,
+      "learning_rate": 0.00011872369496873222,
+      "loss": 3.0077,
+      "step": 7159
+    },
+    {
+      "epoch": 0.4404527559055118,
+      "grad_norm": 1.2024750709533691,
+      "learning_rate": 0.00011870469897127107,
+      "loss": 2.8646,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4405142716535433,
+      "grad_norm": 1.7748974561691284,
+      "learning_rate": 0.00011868570227435852,
+      "loss": 3.0755,
+      "step": 7161
+    },
+    {
+      "epoch": 0.4405757874015748,
+      "grad_norm": 1.2364670038223267,
+      "learning_rate": 0.00011866670487870486,
+      "loss": 2.9925,
+      "step": 7162
+    },
+    {
+      "epoch": 0.4406373031496063,
+      "grad_norm": 1.1104528903961182,
+      "learning_rate": 0.00011864770678502056,
+      "loss": 2.1242,
+      "step": 7163
+    },
+    {
+      "epoch": 0.4406988188976378,
+      "grad_norm": 1.2989481687545776,
+      "learning_rate": 0.00011862870799401597,
+      "loss": 2.6246,
+      "step": 7164
+    },
+    {
+      "epoch": 0.44076033464566927,
+      "grad_norm": 1.3994046449661255,
+      "learning_rate": 0.00011860970850640158,
+      "loss": 3.0776,
+      "step": 7165
+    },
+    {
+      "epoch": 0.4408218503937008,
+      "grad_norm": 1.3477647304534912,
+      "learning_rate": 0.0001185907083228879,
+      "loss": 2.2952,
+      "step": 7166
+    },
+    {
+      "epoch": 0.4408833661417323,
+      "grad_norm": 1.2764877080917358,
+      "learning_rate": 0.00011857170744418537,
+      "loss": 2.7436,
+      "step": 7167
+    },
+    {
+      "epoch": 0.4409448818897638,
+      "grad_norm": 1.0967106819152832,
+      "learning_rate": 0.00011855270587100454,
+      "loss": 2.7721,
+      "step": 7168
+    },
+    {
+      "epoch": 0.44100639763779526,
+      "grad_norm": 1.3345832824707031,
+      "learning_rate": 0.00011853370360405596,
+      "loss": 2.6874,
+      "step": 7169
+    },
+    {
+      "epoch": 0.4410679133858268,
+      "grad_norm": 1.1579394340515137,
+      "learning_rate": 0.00011851470064405021,
+      "loss": 2.9669,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4411294291338583,
+      "grad_norm": 1.3273961544036865,
+      "learning_rate": 0.00011849569699169789,
+      "loss": 2.0372,
+      "step": 7171
+    },
+    {
+      "epoch": 0.44119094488188976,
+      "grad_norm": 1.2626677751541138,
+      "learning_rate": 0.00011847669264770966,
+      "loss": 2.4239,
+      "step": 7172
+    },
+    {
+      "epoch": 0.44125246062992124,
+      "grad_norm": 1.2052749395370483,
+      "learning_rate": 0.00011845768761279616,
+      "loss": 2.6442,
+      "step": 7173
+    },
+    {
+      "epoch": 0.4413139763779528,
+      "grad_norm": 1.1929012537002563,
+      "learning_rate": 0.00011843868188766807,
+      "loss": 2.6005,
+      "step": 7174
+    },
+    {
+      "epoch": 0.44137549212598426,
+      "grad_norm": 1.2986267805099487,
+      "learning_rate": 0.00011841967547303604,
+      "loss": 1.999,
+      "step": 7175
+    },
+    {
+      "epoch": 0.44143700787401574,
+      "grad_norm": 1.059539556503296,
+      "learning_rate": 0.0001184006683696109,
+      "loss": 1.5428,
+      "step": 7176
+    },
+    {
+      "epoch": 0.4414985236220472,
+      "grad_norm": 1.2970342636108398,
+      "learning_rate": 0.00011838166057810338,
+      "loss": 2.4491,
+      "step": 7177
+    },
+    {
+      "epoch": 0.44156003937007876,
+      "grad_norm": 1.1799365282058716,
+      "learning_rate": 0.00011836265209922422,
+      "loss": 2.6401,
+      "step": 7178
+    },
+    {
+      "epoch": 0.44162155511811024,
+      "grad_norm": 1.105394959449768,
+      "learning_rate": 0.0001183436429336843,
+      "loss": 2.9881,
+      "step": 7179
+    },
+    {
+      "epoch": 0.4416830708661417,
+      "grad_norm": 1.5282782316207886,
+      "learning_rate": 0.00011832463308219439,
+      "loss": 2.5225,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4417445866141732,
+      "grad_norm": 1.3556627035140991,
+      "learning_rate": 0.00011830562254546537,
+      "loss": 1.5325,
+      "step": 7181
+    },
+    {
+      "epoch": 0.44180610236220474,
+      "grad_norm": 1.683369755744934,
+      "learning_rate": 0.00011828661132420817,
+      "loss": 2.6759,
+      "step": 7182
+    },
+    {
+      "epoch": 0.4418676181102362,
+      "grad_norm": 1.751595139503479,
+      "learning_rate": 0.0001182675994191337,
+      "loss": 2.3357,
+      "step": 7183
+    },
+    {
+      "epoch": 0.4419291338582677,
+      "grad_norm": 1.0622522830963135,
+      "learning_rate": 0.0001182485868309528,
+      "loss": 2.1792,
+      "step": 7184
+    },
+    {
+      "epoch": 0.4419906496062992,
+      "grad_norm": 1.067615032196045,
+      "learning_rate": 0.00011822957356037658,
+      "loss": 2.4159,
+      "step": 7185
+    },
+    {
+      "epoch": 0.44205216535433073,
+      "grad_norm": 1.2186185121536255,
+      "learning_rate": 0.00011821055960811596,
+      "loss": 2.5708,
+      "step": 7186
+    },
+    {
+      "epoch": 0.4421136811023622,
+      "grad_norm": 1.0816826820373535,
+      "learning_rate": 0.00011819154497488192,
+      "loss": 2.1755,
+      "step": 7187
+    },
+    {
+      "epoch": 0.4421751968503937,
+      "grad_norm": 1.1538405418395996,
+      "learning_rate": 0.00011817252966138554,
+      "loss": 2.4665,
+      "step": 7188
+    },
+    {
+      "epoch": 0.4422367125984252,
+      "grad_norm": 1.1488782167434692,
+      "learning_rate": 0.00011815351366833789,
+      "loss": 3.2295,
+      "step": 7189
+    },
+    {
+      "epoch": 0.4422982283464567,
+      "grad_norm": 1.1032652854919434,
+      "learning_rate": 0.00011813449699645008,
+      "loss": 2.6507,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4423597440944882,
+      "grad_norm": 0.9569475054740906,
+      "learning_rate": 0.00011811547964643314,
+      "loss": 2.7114,
+      "step": 7191
+    },
+    {
+      "epoch": 0.4424212598425197,
+      "grad_norm": 1.3242520093917847,
+      "learning_rate": 0.00011809646161899835,
+      "loss": 2.344,
+      "step": 7192
+    },
+    {
+      "epoch": 0.44248277559055116,
+      "grad_norm": 1.1049935817718506,
+      "learning_rate": 0.00011807744291485676,
+      "loss": 2.638,
+      "step": 7193
+    },
+    {
+      "epoch": 0.4425442913385827,
+      "grad_norm": 1.1701843738555908,
+      "learning_rate": 0.00011805842353471961,
+      "loss": 2.4909,
+      "step": 7194
+    },
+    {
+      "epoch": 0.4426058070866142,
+      "grad_norm": 1.251820683479309,
+      "learning_rate": 0.00011803940347929812,
+      "loss": 1.7874,
+      "step": 7195
+    },
+    {
+      "epoch": 0.44266732283464566,
+      "grad_norm": 1.3839404582977295,
+      "learning_rate": 0.00011802038274930356,
+      "loss": 2.7201,
+      "step": 7196
+    },
+    {
+      "epoch": 0.44272883858267714,
+      "grad_norm": 1.249441146850586,
+      "learning_rate": 0.00011800136134544713,
+      "loss": 1.5287,
+      "step": 7197
+    },
+    {
+      "epoch": 0.4427903543307087,
+      "grad_norm": 1.1504863500595093,
+      "learning_rate": 0.00011798233926844017,
+      "loss": 2.6881,
+      "step": 7198
+    },
+    {
+      "epoch": 0.44285187007874016,
+      "grad_norm": 1.3665584325790405,
+      "learning_rate": 0.00011796331651899402,
+      "loss": 2.6524,
+      "step": 7199
+    },
+    {
+      "epoch": 0.44291338582677164,
+      "grad_norm": 1.2188156843185425,
+      "learning_rate": 0.00011794429309781998,
+      "loss": 2.5968,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4429749015748031,
+      "grad_norm": 1.2852996587753296,
+      "learning_rate": 0.00011792526900562944,
+      "loss": 2.5254,
+      "step": 7201
+    },
+    {
+      "epoch": 0.44303641732283466,
+      "grad_norm": 1.3251903057098389,
+      "learning_rate": 0.0001179062442431338,
+      "loss": 3.4195,
+      "step": 7202
+    },
+    {
+      "epoch": 0.44309793307086615,
+      "grad_norm": 1.1013600826263428,
+      "learning_rate": 0.00011788721881104443,
+      "loss": 3.1426,
+      "step": 7203
+    },
+    {
+      "epoch": 0.4431594488188976,
+      "grad_norm": 1.2416629791259766,
+      "learning_rate": 0.00011786819271007288,
+      "loss": 2.5189,
+      "step": 7204
+    },
+    {
+      "epoch": 0.4432209645669291,
+      "grad_norm": 1.4106897115707397,
+      "learning_rate": 0.00011784916594093051,
+      "loss": 2.2508,
+      "step": 7205
+    },
+    {
+      "epoch": 0.44328248031496065,
+      "grad_norm": 1.0874766111373901,
+      "learning_rate": 0.00011783013850432891,
+      "loss": 2.236,
+      "step": 7206
+    },
+    {
+      "epoch": 0.44334399606299213,
+      "grad_norm": 1.075551152229309,
+      "learning_rate": 0.00011781111040097952,
+      "loss": 2.5979,
+      "step": 7207
+    },
+    {
+      "epoch": 0.4434055118110236,
+      "grad_norm": 1.4333488941192627,
+      "learning_rate": 0.00011779208163159393,
+      "loss": 2.0637,
+      "step": 7208
+    },
+    {
+      "epoch": 0.4434670275590551,
+      "grad_norm": 1.1240514516830444,
+      "learning_rate": 0.00011777305219688371,
+      "loss": 2.5879,
+      "step": 7209
+    },
+    {
+      "epoch": 0.44352854330708663,
+      "grad_norm": 1.144729733467102,
+      "learning_rate": 0.00011775402209756039,
+      "loss": 2.7828,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4435900590551181,
+      "grad_norm": 1.9028733968734741,
+      "learning_rate": 0.0001177349913343357,
+      "loss": 2.1174,
+      "step": 7211
+    },
+    {
+      "epoch": 0.4436515748031496,
+      "grad_norm": 1.3756535053253174,
+      "learning_rate": 0.00011771595990792121,
+      "loss": 2.5263,
+      "step": 7212
+    },
+    {
+      "epoch": 0.4437130905511811,
+      "grad_norm": 0.9380342960357666,
+      "learning_rate": 0.00011769692781902859,
+      "loss": 3.0029,
+      "step": 7213
+    },
+    {
+      "epoch": 0.4437746062992126,
+      "grad_norm": 1.3830829858779907,
+      "learning_rate": 0.00011767789506836957,
+      "loss": 3.1775,
+      "step": 7214
+    },
+    {
+      "epoch": 0.4438361220472441,
+      "grad_norm": 1.386940360069275,
+      "learning_rate": 0.00011765886165665585,
+      "loss": 2.956,
+      "step": 7215
+    },
+    {
+      "epoch": 0.4438976377952756,
+      "grad_norm": 1.1034284830093384,
+      "learning_rate": 0.00011763982758459915,
+      "loss": 1.7022,
+      "step": 7216
+    },
+    {
+      "epoch": 0.44395915354330706,
+      "grad_norm": 1.3818703889846802,
+      "learning_rate": 0.00011762079285291128,
+      "loss": 2.9934,
+      "step": 7217
+    },
+    {
+      "epoch": 0.4440206692913386,
+      "grad_norm": 1.2276676893234253,
+      "learning_rate": 0.00011760175746230402,
+      "loss": 3.1776,
+      "step": 7218
+    },
+    {
+      "epoch": 0.4440821850393701,
+      "grad_norm": 1.3717628717422485,
+      "learning_rate": 0.00011758272141348918,
+      "loss": 3.1436,
+      "step": 7219
+    },
+    {
+      "epoch": 0.44414370078740156,
+      "grad_norm": 1.1036455631256104,
+      "learning_rate": 0.00011756368470717858,
+      "loss": 1.8839,
+      "step": 7220
+    },
+    {
+      "epoch": 0.44420521653543305,
+      "grad_norm": 1.2763402462005615,
+      "learning_rate": 0.00011754464734408412,
+      "loss": 3.1934,
+      "step": 7221
+    },
+    {
+      "epoch": 0.4442667322834646,
+      "grad_norm": 1.4346411228179932,
+      "learning_rate": 0.0001175256093249177,
+      "loss": 2.8553,
+      "step": 7222
+    },
+    {
+      "epoch": 0.44432824803149606,
+      "grad_norm": 1.3370360136032104,
+      "learning_rate": 0.0001175065706503912,
+      "loss": 2.3439,
+      "step": 7223
+    },
+    {
+      "epoch": 0.44438976377952755,
+      "grad_norm": 1.229494333267212,
+      "learning_rate": 0.00011748753132121657,
+      "loss": 2.4663,
+      "step": 7224
+    },
+    {
+      "epoch": 0.44445127952755903,
+      "grad_norm": 1.1864898204803467,
+      "learning_rate": 0.0001174684913381058,
+      "loss": 2.594,
+      "step": 7225
+    },
+    {
+      "epoch": 0.44451279527559057,
+      "grad_norm": 1.028801441192627,
+      "learning_rate": 0.00011744945070177085,
+      "loss": 2.9575,
+      "step": 7226
+    },
+    {
+      "epoch": 0.44457431102362205,
+      "grad_norm": 1.2347455024719238,
+      "learning_rate": 0.00011743040941292373,
+      "loss": 2.7019,
+      "step": 7227
+    },
+    {
+      "epoch": 0.44463582677165353,
+      "grad_norm": 1.3209768533706665,
+      "learning_rate": 0.00011741136747227652,
+      "loss": 2.6128,
+      "step": 7228
+    },
+    {
+      "epoch": 0.444697342519685,
+      "grad_norm": 1.300422191619873,
+      "learning_rate": 0.00011739232488054124,
+      "loss": 2.5071,
+      "step": 7229
+    },
+    {
+      "epoch": 0.44475885826771655,
+      "grad_norm": 1.091811180114746,
+      "learning_rate": 0.00011737328163843,
+      "loss": 3.1318,
+      "step": 7230
+    },
+    {
+      "epoch": 0.44482037401574803,
+      "grad_norm": 1.823935866355896,
+      "learning_rate": 0.00011735423774665491,
+      "loss": 2.2408,
+      "step": 7231
+    },
+    {
+      "epoch": 0.4448818897637795,
+      "grad_norm": 1.2551757097244263,
+      "learning_rate": 0.00011733519320592808,
+      "loss": 2.5706,
+      "step": 7232
+    },
+    {
+      "epoch": 0.444943405511811,
+      "grad_norm": 1.2120661735534668,
+      "learning_rate": 0.00011731614801696171,
+      "loss": 3.0253,
+      "step": 7233
+    },
+    {
+      "epoch": 0.44500492125984253,
+      "grad_norm": 1.2380943298339844,
+      "learning_rate": 0.00011729710218046794,
+      "loss": 2.9325,
+      "step": 7234
+    },
+    {
+      "epoch": 0.445066437007874,
+      "grad_norm": 1.2461233139038086,
+      "learning_rate": 0.000117278055697159,
+      "loss": 2.7969,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4451279527559055,
+      "grad_norm": 1.253032922744751,
+      "learning_rate": 0.00011725900856774713,
+      "loss": 2.7593,
+      "step": 7236
+    },
+    {
+      "epoch": 0.445189468503937,
+      "grad_norm": 1.6806128025054932,
+      "learning_rate": 0.00011723996079294459,
+      "loss": 2.9805,
+      "step": 7237
+    },
+    {
+      "epoch": 0.4452509842519685,
+      "grad_norm": 1.1012886762619019,
+      "learning_rate": 0.00011722091237346362,
+      "loss": 2.1264,
+      "step": 7238
+    },
+    {
+      "epoch": 0.4453125,
+      "grad_norm": 1.1214865446090698,
+      "learning_rate": 0.00011720186331001656,
+      "loss": 3.1563,
+      "step": 7239
+    },
+    {
+      "epoch": 0.4453740157480315,
+      "grad_norm": 1.2745332717895508,
+      "learning_rate": 0.00011718281360331576,
+      "loss": 2.5355,
+      "step": 7240
+    },
+    {
+      "epoch": 0.445435531496063,
+      "grad_norm": 1.4137405157089233,
+      "learning_rate": 0.00011716376325407353,
+      "loss": 2.2728,
+      "step": 7241
+    },
+    {
+      "epoch": 0.4454970472440945,
+      "grad_norm": 1.0985984802246094,
+      "learning_rate": 0.00011714471226300223,
+      "loss": 2.917,
+      "step": 7242
+    },
+    {
+      "epoch": 0.445558562992126,
+      "grad_norm": 1.1094346046447754,
+      "learning_rate": 0.00011712566063081434,
+      "loss": 2.8014,
+      "step": 7243
+    },
+    {
+      "epoch": 0.44562007874015747,
+      "grad_norm": 1.0911563634872437,
+      "learning_rate": 0.00011710660835822222,
+      "loss": 2.5709,
+      "step": 7244
+    },
+    {
+      "epoch": 0.445681594488189,
+      "grad_norm": 1.2453900575637817,
+      "learning_rate": 0.00011708755544593832,
+      "loss": 3.0871,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4457431102362205,
+      "grad_norm": 1.7083230018615723,
+      "learning_rate": 0.00011706850189467515,
+      "loss": 2.5789,
+      "step": 7246
+    },
+    {
+      "epoch": 0.44580462598425197,
+      "grad_norm": 1.1251566410064697,
+      "learning_rate": 0.00011704944770514518,
+      "loss": 3.0354,
+      "step": 7247
+    },
+    {
+      "epoch": 0.44586614173228345,
+      "grad_norm": 1.4326071739196777,
+      "learning_rate": 0.00011703039287806094,
+      "loss": 2.15,
+      "step": 7248
+    },
+    {
+      "epoch": 0.445927657480315,
+      "grad_norm": 1.0890299081802368,
+      "learning_rate": 0.00011701133741413497,
+      "loss": 2.3728,
+      "step": 7249
+    },
+    {
+      "epoch": 0.44598917322834647,
+      "grad_norm": 1.4228633642196655,
+      "learning_rate": 0.00011699228131407984,
+      "loss": 2.8057,
+      "step": 7250
+    },
+    {
+      "epoch": 0.44605068897637795,
+      "grad_norm": 1.332116723060608,
+      "learning_rate": 0.00011697322457860813,
+      "loss": 2.4177,
+      "step": 7251
+    },
+    {
+      "epoch": 0.44611220472440943,
+      "grad_norm": 1.0967496633529663,
+      "learning_rate": 0.00011695416720843246,
+      "loss": 2.5435,
+      "step": 7252
+    },
+    {
+      "epoch": 0.44617372047244097,
+      "grad_norm": 1.191976547241211,
+      "learning_rate": 0.00011693510920426551,
+      "loss": 2.4837,
+      "step": 7253
+    },
+    {
+      "epoch": 0.44623523622047245,
+      "grad_norm": 1.2739129066467285,
+      "learning_rate": 0.00011691605056681989,
+      "loss": 2.7133,
+      "step": 7254
+    },
+    {
+      "epoch": 0.44629675196850394,
+      "grad_norm": 1.1976308822631836,
+      "learning_rate": 0.00011689699129680828,
+      "loss": 3.1896,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4463582677165354,
+      "grad_norm": 1.560451865196228,
+      "learning_rate": 0.00011687793139494346,
+      "loss": 2.7593,
+      "step": 7256
+    },
+    {
+      "epoch": 0.44641978346456695,
+      "grad_norm": 1.252142310142517,
+      "learning_rate": 0.0001168588708619381,
+      "loss": 2.4098,
+      "step": 7257
+    },
+    {
+      "epoch": 0.44648129921259844,
+      "grad_norm": 1.1385300159454346,
+      "learning_rate": 0.00011683980969850494,
+      "loss": 2.2223,
+      "step": 7258
+    },
+    {
+      "epoch": 0.4465428149606299,
+      "grad_norm": 1.197648048400879,
+      "learning_rate": 0.00011682074790535684,
+      "loss": 2.7455,
+      "step": 7259
+    },
+    {
+      "epoch": 0.4466043307086614,
+      "grad_norm": 1.2912260293960571,
+      "learning_rate": 0.00011680168548320658,
+      "loss": 2.3794,
+      "step": 7260
+    },
+    {
+      "epoch": 0.44666584645669294,
+      "grad_norm": 1.3540126085281372,
+      "learning_rate": 0.00011678262243276692,
+      "loss": 2.7022,
+      "step": 7261
+    },
+    {
+      "epoch": 0.4467273622047244,
+      "grad_norm": 1.8537861108779907,
+      "learning_rate": 0.00011676355875475081,
+      "loss": 1.901,
+      "step": 7262
+    },
+    {
+      "epoch": 0.4467888779527559,
+      "grad_norm": 1.5300440788269043,
+      "learning_rate": 0.00011674449444987105,
+      "loss": 2.5108,
+      "step": 7263
+    },
+    {
+      "epoch": 0.4468503937007874,
+      "grad_norm": 1.3302699327468872,
+      "learning_rate": 0.00011672542951884058,
+      "loss": 2.3085,
+      "step": 7264
+    },
+    {
+      "epoch": 0.4469119094488189,
+      "grad_norm": 1.2420654296875,
+      "learning_rate": 0.00011670636396237232,
+      "loss": 2.6926,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4469734251968504,
+      "grad_norm": 1.0990997552871704,
+      "learning_rate": 0.00011668729778117917,
+      "loss": 2.1637,
+      "step": 7266
+    },
+    {
+      "epoch": 0.4470349409448819,
+      "grad_norm": 1.2302645444869995,
+      "learning_rate": 0.00011666823097597418,
+      "loss": 2.3611,
+      "step": 7267
+    },
+    {
+      "epoch": 0.44709645669291337,
+      "grad_norm": 1.2194918394088745,
+      "learning_rate": 0.00011664916354747025,
+      "loss": 2.7379,
+      "step": 7268
+    },
+    {
+      "epoch": 0.4471579724409449,
+      "grad_norm": 1.2529313564300537,
+      "learning_rate": 0.00011663009549638047,
+      "loss": 2.5992,
+      "step": 7269
+    },
+    {
+      "epoch": 0.4472194881889764,
+      "grad_norm": 1.3029394149780273,
+      "learning_rate": 0.00011661102682341785,
+      "loss": 2.9766,
+      "step": 7270
+    },
+    {
+      "epoch": 0.44728100393700787,
+      "grad_norm": 1.3117903470993042,
+      "learning_rate": 0.00011659195752929546,
+      "loss": 2.4536,
+      "step": 7271
+    },
+    {
+      "epoch": 0.44734251968503935,
+      "grad_norm": 1.8248016834259033,
+      "learning_rate": 0.00011657288761472638,
+      "loss": 2.0768,
+      "step": 7272
+    },
+    {
+      "epoch": 0.4474040354330709,
+      "grad_norm": 1.173685908317566,
+      "learning_rate": 0.00011655381708042367,
+      "loss": 4.1345,
+      "step": 7273
+    },
+    {
+      "epoch": 0.4474655511811024,
+      "grad_norm": 1.2873947620391846,
+      "learning_rate": 0.00011653474592710054,
+      "loss": 2.5736,
+      "step": 7274
+    },
+    {
+      "epoch": 0.44752706692913385,
+      "grad_norm": 1.3928651809692383,
+      "learning_rate": 0.00011651567415547013,
+      "loss": 1.5622,
+      "step": 7275
+    },
+    {
+      "epoch": 0.44758858267716534,
+      "grad_norm": 1.3513374328613281,
+      "learning_rate": 0.00011649660176624558,
+      "loss": 2.2341,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4476500984251969,
+      "grad_norm": 1.4027477502822876,
+      "learning_rate": 0.00011647752876014011,
+      "loss": 2.2223,
+      "step": 7277
+    },
+    {
+      "epoch": 0.44771161417322836,
+      "grad_norm": 1.115465760231018,
+      "learning_rate": 0.00011645845513786694,
+      "loss": 2.7735,
+      "step": 7278
+    },
+    {
+      "epoch": 0.44777312992125984,
+      "grad_norm": 1.2389352321624756,
+      "learning_rate": 0.00011643938090013935,
+      "loss": 2.1265,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4478346456692913,
+      "grad_norm": 1.25967538356781,
+      "learning_rate": 0.00011642030604767054,
+      "loss": 1.9593,
+      "step": 7280
+    },
+    {
+      "epoch": 0.44789616141732286,
+      "grad_norm": 1.036478877067566,
+      "learning_rate": 0.00011640123058117386,
+      "loss": 2.747,
+      "step": 7281
+    },
+    {
+      "epoch": 0.44795767716535434,
+      "grad_norm": 1.4217177629470825,
+      "learning_rate": 0.00011638215450136266,
+      "loss": 1.9585,
+      "step": 7282
+    },
+    {
+      "epoch": 0.4480191929133858,
+      "grad_norm": 1.0223708152770996,
+      "learning_rate": 0.00011636307780895017,
+      "loss": 2.7922,
+      "step": 7283
+    },
+    {
+      "epoch": 0.4480807086614173,
+      "grad_norm": 1.5347734689712524,
+      "learning_rate": 0.00011634400050464984,
+      "loss": 2.8519,
+      "step": 7284
+    },
+    {
+      "epoch": 0.44814222440944884,
+      "grad_norm": 1.155022144317627,
+      "learning_rate": 0.00011632492258917503,
+      "loss": 3.0387,
+      "step": 7285
+    },
+    {
+      "epoch": 0.4482037401574803,
+      "grad_norm": 1.0992631912231445,
+      "learning_rate": 0.00011630584406323912,
+      "loss": 2.1917,
+      "step": 7286
+    },
+    {
+      "epoch": 0.4482652559055118,
+      "grad_norm": 1.2428480386734009,
+      "learning_rate": 0.00011628676492755559,
+      "loss": 2.7532,
+      "step": 7287
+    },
+    {
+      "epoch": 0.4483267716535433,
+      "grad_norm": 1.2663041353225708,
+      "learning_rate": 0.00011626768518283788,
+      "loss": 2.6375,
+      "step": 7288
+    },
+    {
+      "epoch": 0.4483882874015748,
+      "grad_norm": 1.4445244073867798,
+      "learning_rate": 0.00011624860482979945,
+      "loss": 3.0055,
+      "step": 7289
+    },
+    {
+      "epoch": 0.4484498031496063,
+      "grad_norm": 1.2867896556854248,
+      "learning_rate": 0.00011622952386915376,
+      "loss": 2.0193,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4485113188976378,
+      "grad_norm": 1.3427555561065674,
+      "learning_rate": 0.00011621044230161442,
+      "loss": 2.6105,
+      "step": 7291
+    },
+    {
+      "epoch": 0.44857283464566927,
+      "grad_norm": 1.2859402894973755,
+      "learning_rate": 0.00011619136012789494,
+      "loss": 2.8108,
+      "step": 7292
+    },
+    {
+      "epoch": 0.4486343503937008,
+      "grad_norm": 1.4225335121154785,
+      "learning_rate": 0.00011617227734870883,
+      "loss": 2.3492,
+      "step": 7293
+    },
+    {
+      "epoch": 0.4486958661417323,
+      "grad_norm": 1.5160778760910034,
+      "learning_rate": 0.00011615319396476975,
+      "loss": 2.6839,
+      "step": 7294
+    },
+    {
+      "epoch": 0.4487573818897638,
+      "grad_norm": 2.0981011390686035,
+      "learning_rate": 0.0001161341099767913,
+      "loss": 1.7038,
+      "step": 7295
+    },
+    {
+      "epoch": 0.44881889763779526,
+      "grad_norm": 1.2031580209732056,
+      "learning_rate": 0.00011611502538548708,
+      "loss": 2.1001,
+      "step": 7296
+    },
+    {
+      "epoch": 0.4488804133858268,
+      "grad_norm": 1.0975937843322754,
+      "learning_rate": 0.00011609594019157079,
+      "loss": 1.8795,
+      "step": 7297
+    },
+    {
+      "epoch": 0.4489419291338583,
+      "grad_norm": 1.4182782173156738,
+      "learning_rate": 0.00011607685439575608,
+      "loss": 2.4741,
+      "step": 7298
+    },
+    {
+      "epoch": 0.44900344488188976,
+      "grad_norm": 1.104740023612976,
+      "learning_rate": 0.00011605776799875668,
+      "loss": 2.6361,
+      "step": 7299
+    },
+    {
+      "epoch": 0.44906496062992124,
+      "grad_norm": 1.1961876153945923,
+      "learning_rate": 0.00011603868100128628,
+      "loss": 2.5943,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4491264763779528,
+      "grad_norm": 1.1602674722671509,
+      "learning_rate": 0.00011601959340405865,
+      "loss": 2.5276,
+      "step": 7301
+    },
+    {
+      "epoch": 0.44918799212598426,
+      "grad_norm": 1.2883363962173462,
+      "learning_rate": 0.00011600050520778755,
+      "loss": 2.2472,
+      "step": 7302
+    },
+    {
+      "epoch": 0.44924950787401574,
+      "grad_norm": 1.1798640489578247,
+      "learning_rate": 0.00011598141641318678,
+      "loss": 3.1927,
+      "step": 7303
+    },
+    {
+      "epoch": 0.4493110236220472,
+      "grad_norm": 1.1797027587890625,
+      "learning_rate": 0.00011596232702097016,
+      "loss": 2.1363,
+      "step": 7304
+    },
+    {
+      "epoch": 0.44937253937007876,
+      "grad_norm": 1.244814395904541,
+      "learning_rate": 0.0001159432370318515,
+      "loss": 2.2591,
+      "step": 7305
+    },
+    {
+      "epoch": 0.44943405511811024,
+      "grad_norm": 1.2478411197662354,
+      "learning_rate": 0.00011592414644654471,
+      "loss": 2.4774,
+      "step": 7306
+    },
+    {
+      "epoch": 0.4494955708661417,
+      "grad_norm": 1.2293131351470947,
+      "learning_rate": 0.0001159050552657636,
+      "loss": 2.2677,
+      "step": 7307
+    },
+    {
+      "epoch": 0.4495570866141732,
+      "grad_norm": 1.1017879247665405,
+      "learning_rate": 0.00011588596349022211,
+      "loss": 2.5649,
+      "step": 7308
+    },
+    {
+      "epoch": 0.44961860236220474,
+      "grad_norm": 1.2310853004455566,
+      "learning_rate": 0.0001158668711206342,
+      "loss": 2.7714,
+      "step": 7309
+    },
+    {
+      "epoch": 0.4496801181102362,
+      "grad_norm": 1.2154450416564941,
+      "learning_rate": 0.00011584777815771377,
+      "loss": 2.55,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4497416338582677,
+      "grad_norm": 2.1395814418792725,
+      "learning_rate": 0.00011582868460217483,
+      "loss": 2.4765,
+      "step": 7311
+    },
+    {
+      "epoch": 0.4498031496062992,
+      "grad_norm": 1.545474886894226,
+      "learning_rate": 0.0001158095904547313,
+      "loss": 2.5276,
+      "step": 7312
+    },
+    {
+      "epoch": 0.44986466535433073,
+      "grad_norm": 1.389259696006775,
+      "learning_rate": 0.00011579049571609728,
+      "loss": 1.8012,
+      "step": 7313
+    },
+    {
+      "epoch": 0.4499261811023622,
+      "grad_norm": 1.193975806236267,
+      "learning_rate": 0.00011577140038698678,
+      "loss": 2.5781,
+      "step": 7314
+    },
+    {
+      "epoch": 0.4499876968503937,
+      "grad_norm": 1.1628143787384033,
+      "learning_rate": 0.0001157523044681138,
+      "loss": 2.2457,
+      "step": 7315
+    },
+    {
+      "epoch": 0.4500492125984252,
+      "grad_norm": 1.2096275091171265,
+      "learning_rate": 0.00011573320796019251,
+      "loss": 2.177,
+      "step": 7316
+    },
+    {
+      "epoch": 0.4501107283464567,
+      "grad_norm": 1.1653021574020386,
+      "learning_rate": 0.00011571411086393698,
+      "loss": 1.8582,
+      "step": 7317
+    },
+    {
+      "epoch": 0.4501722440944882,
+      "grad_norm": 1.964378833770752,
+      "learning_rate": 0.00011569501318006129,
+      "loss": 2.628,
+      "step": 7318
+    },
+    {
+      "epoch": 0.4502337598425197,
+      "grad_norm": 1.386061191558838,
+      "learning_rate": 0.00011567591490927965,
+      "loss": 2.559,
+      "step": 7319
+    },
+    {
+      "epoch": 0.45029527559055116,
+      "grad_norm": 1.23321533203125,
+      "learning_rate": 0.0001156568160523062,
+      "loss": 2.8576,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4503567913385827,
+      "grad_norm": 1.0423307418823242,
+      "learning_rate": 0.00011563771660985512,
+      "loss": 2.5915,
+      "step": 7321
+    },
+    {
+      "epoch": 0.4504183070866142,
+      "grad_norm": 1.1803985834121704,
+      "learning_rate": 0.00011561861658264066,
+      "loss": 3.1455,
+      "step": 7322
+    },
+    {
+      "epoch": 0.45047982283464566,
+      "grad_norm": 1.1119779348373413,
+      "learning_rate": 0.00011559951597137701,
+      "loss": 2.4549,
+      "step": 7323
+    },
+    {
+      "epoch": 0.45054133858267714,
+      "grad_norm": 1.2498743534088135,
+      "learning_rate": 0.00011558041477677848,
+      "loss": 2.7025,
+      "step": 7324
+    },
+    {
+      "epoch": 0.4506028543307087,
+      "grad_norm": 1.1106958389282227,
+      "learning_rate": 0.00011556131299955928,
+      "loss": 2.5524,
+      "step": 7325
+    },
+    {
+      "epoch": 0.45066437007874016,
+      "grad_norm": 1.2016581296920776,
+      "learning_rate": 0.00011554221064043375,
+      "loss": 2.6537,
+      "step": 7326
+    },
+    {
+      "epoch": 0.45072588582677164,
+      "grad_norm": 1.227379322052002,
+      "learning_rate": 0.00011552310770011625,
+      "loss": 2.4144,
+      "step": 7327
+    },
+    {
+      "epoch": 0.4507874015748031,
+      "grad_norm": 1.1496564149856567,
+      "learning_rate": 0.00011550400417932103,
+      "loss": 2.9621,
+      "step": 7328
+    },
+    {
+      "epoch": 0.45084891732283466,
+      "grad_norm": 1.511233925819397,
+      "learning_rate": 0.00011548490007876253,
+      "loss": 1.9344,
+      "step": 7329
+    },
+    {
+      "epoch": 0.45091043307086615,
+      "grad_norm": 1.1234731674194336,
+      "learning_rate": 0.00011546579539915512,
+      "loss": 1.6903,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4509719488188976,
+      "grad_norm": 1.1009224653244019,
+      "learning_rate": 0.00011544669014121319,
+      "loss": 2.0221,
+      "step": 7331
+    },
+    {
+      "epoch": 0.4510334645669291,
+      "grad_norm": 1.731203317642212,
+      "learning_rate": 0.00011542758430565118,
+      "loss": 2.0417,
+      "step": 7332
+    },
+    {
+      "epoch": 0.45109498031496065,
+      "grad_norm": 1.2925058603286743,
+      "learning_rate": 0.00011540847789318357,
+      "loss": 2.6239,
+      "step": 7333
+    },
+    {
+      "epoch": 0.45115649606299213,
+      "grad_norm": 1.5959972143173218,
+      "learning_rate": 0.00011538937090452475,
+      "loss": 1.9801,
+      "step": 7334
+    },
+    {
+      "epoch": 0.4512180118110236,
+      "grad_norm": 1.1362104415893555,
+      "learning_rate": 0.00011537026334038932,
+      "loss": 2.9191,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4512795275590551,
+      "grad_norm": 1.1386749744415283,
+      "learning_rate": 0.00011535115520149173,
+      "loss": 2.6934,
+      "step": 7336
+    },
+    {
+      "epoch": 0.45134104330708663,
+      "grad_norm": 1.3225836753845215,
+      "learning_rate": 0.00011533204648854656,
+      "loss": 2.5196,
+      "step": 7337
+    },
+    {
+      "epoch": 0.4514025590551181,
+      "grad_norm": 1.2089521884918213,
+      "learning_rate": 0.0001153129372022683,
+      "loss": 2.3622,
+      "step": 7338
+    },
+    {
+      "epoch": 0.4514640748031496,
+      "grad_norm": 1.1841118335723877,
+      "learning_rate": 0.00011529382734337162,
+      "loss": 2.6317,
+      "step": 7339
+    },
+    {
+      "epoch": 0.4515255905511811,
+      "grad_norm": 1.1632344722747803,
+      "learning_rate": 0.00011527471691257105,
+      "loss": 2.1447,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4515871062992126,
+      "grad_norm": 1.510338544845581,
+      "learning_rate": 0.00011525560591058124,
+      "loss": 2.7587,
+      "step": 7341
+    },
+    {
+      "epoch": 0.4516486220472441,
+      "grad_norm": 1.1339621543884277,
+      "learning_rate": 0.00011523649433811685,
+      "loss": 2.8325,
+      "step": 7342
+    },
+    {
+      "epoch": 0.4517101377952756,
+      "grad_norm": 1.473028540611267,
+      "learning_rate": 0.00011521738219589257,
+      "loss": 2.5507,
+      "step": 7343
+    },
+    {
+      "epoch": 0.45177165354330706,
+      "grad_norm": 1.1745460033416748,
+      "learning_rate": 0.00011519826948462298,
+      "loss": 1.8565,
+      "step": 7344
+    },
+    {
+      "epoch": 0.4518331692913386,
+      "grad_norm": 1.1757361888885498,
+      "learning_rate": 0.0001151791562050229,
+      "loss": 3.0124,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4518946850393701,
+      "grad_norm": 1.3003017902374268,
+      "learning_rate": 0.00011516004235780702,
+      "loss": 2.5156,
+      "step": 7346
+    },
+    {
+      "epoch": 0.45195620078740156,
+      "grad_norm": 1.2706547975540161,
+      "learning_rate": 0.00011514092794369007,
+      "loss": 5.6786,
+      "step": 7347
+    },
+    {
+      "epoch": 0.45201771653543305,
+      "grad_norm": 1.4913169145584106,
+      "learning_rate": 0.00011512181296338686,
+      "loss": 2.764,
+      "step": 7348
+    },
+    {
+      "epoch": 0.4520792322834646,
+      "grad_norm": 1.1132372617721558,
+      "learning_rate": 0.00011510269741761217,
+      "loss": 3.2411,
+      "step": 7349
+    },
+    {
+      "epoch": 0.45214074803149606,
+      "grad_norm": 1.30504310131073,
+      "learning_rate": 0.00011508358130708077,
+      "loss": 2.303,
+      "step": 7350
+    },
+    {
+      "epoch": 0.45220226377952755,
+      "grad_norm": 1.2754011154174805,
+      "learning_rate": 0.00011506446463250758,
+      "loss": 2.9733,
+      "step": 7351
+    },
+    {
+      "epoch": 0.45226377952755903,
+      "grad_norm": 1.1175177097320557,
+      "learning_rate": 0.00011504534739460742,
+      "loss": 3.0436,
+      "step": 7352
+    },
+    {
+      "epoch": 0.45232529527559057,
+      "grad_norm": 1.2596765756607056,
+      "learning_rate": 0.00011502622959409513,
+      "loss": 2.9641,
+      "step": 7353
+    },
+    {
+      "epoch": 0.45238681102362205,
+      "grad_norm": 1.029317855834961,
+      "learning_rate": 0.00011500711123168566,
+      "loss": 2.5178,
+      "step": 7354
+    },
+    {
+      "epoch": 0.45244832677165353,
+      "grad_norm": 0.9787509441375732,
+      "learning_rate": 0.00011498799230809392,
+      "loss": 3.1227,
+      "step": 7355
+    },
+    {
+      "epoch": 0.452509842519685,
+      "grad_norm": 1.079369306564331,
+      "learning_rate": 0.00011496887282403484,
+      "loss": 2.2024,
+      "step": 7356
+    },
+    {
+      "epoch": 0.45257135826771655,
+      "grad_norm": 1.3101277351379395,
+      "learning_rate": 0.00011494975278022336,
+      "loss": 2.4608,
+      "step": 7357
+    },
+    {
+      "epoch": 0.45263287401574803,
+      "grad_norm": 1.087615966796875,
+      "learning_rate": 0.0001149306321773745,
+      "loss": 2.6378,
+      "step": 7358
+    },
+    {
+      "epoch": 0.4526943897637795,
+      "grad_norm": 1.4514349699020386,
+      "learning_rate": 0.00011491151101620326,
+      "loss": 3.1285,
+      "step": 7359
+    },
+    {
+      "epoch": 0.452755905511811,
+      "grad_norm": 1.2483412027359009,
+      "learning_rate": 0.00011489238929742466,
+      "loss": 2.942,
+      "step": 7360
+    },
+    {
+      "epoch": 0.45281742125984253,
+      "grad_norm": 1.3315091133117676,
+      "learning_rate": 0.00011487326702175374,
+      "loss": 2.1152,
+      "step": 7361
+    },
+    {
+      "epoch": 0.452878937007874,
+      "grad_norm": 1.1754701137542725,
+      "learning_rate": 0.00011485414418990556,
+      "loss": 3.016,
+      "step": 7362
+    },
+    {
+      "epoch": 0.4529404527559055,
+      "grad_norm": 1.1790379285812378,
+      "learning_rate": 0.00011483502080259521,
+      "loss": 2.5131,
+      "step": 7363
+    },
+    {
+      "epoch": 0.453001968503937,
+      "grad_norm": 1.1962275505065918,
+      "learning_rate": 0.00011481589686053783,
+      "loss": 2.8755,
+      "step": 7364
+    },
+    {
+      "epoch": 0.4530634842519685,
+      "grad_norm": 1.1536210775375366,
+      "learning_rate": 0.0001147967723644485,
+      "loss": 2.7498,
+      "step": 7365
+    },
+    {
+      "epoch": 0.453125,
+      "grad_norm": 1.1879163980484009,
+      "learning_rate": 0.00011477764731504239,
+      "loss": 3.0548,
+      "step": 7366
+    },
+    {
+      "epoch": 0.4531865157480315,
+      "grad_norm": 1.1382437944412231,
+      "learning_rate": 0.0001147585217130347,
+      "loss": 2.9789,
+      "step": 7367
+    },
+    {
+      "epoch": 0.453248031496063,
+      "grad_norm": 1.4250757694244385,
+      "learning_rate": 0.00011473939555914058,
+      "loss": 1.6124,
+      "step": 7368
+    },
+    {
+      "epoch": 0.4533095472440945,
+      "grad_norm": 1.1502742767333984,
+      "learning_rate": 0.00011472026885407525,
+      "loss": 2.7255,
+      "step": 7369
+    },
+    {
+      "epoch": 0.453371062992126,
+      "grad_norm": 1.0378069877624512,
+      "learning_rate": 0.00011470114159855394,
+      "loss": 2.812,
+      "step": 7370
+    },
+    {
+      "epoch": 0.45343257874015747,
+      "grad_norm": 1.1833856105804443,
+      "learning_rate": 0.00011468201379329193,
+      "loss": 3.2279,
+      "step": 7371
+    },
+    {
+      "epoch": 0.453494094488189,
+      "grad_norm": 1.151667833328247,
+      "learning_rate": 0.00011466288543900447,
+      "loss": 2.4242,
+      "step": 7372
+    },
+    {
+      "epoch": 0.4535556102362205,
+      "grad_norm": 1.2555532455444336,
+      "learning_rate": 0.00011464375653640684,
+      "loss": 1.7855,
+      "step": 7373
+    },
+    {
+      "epoch": 0.45361712598425197,
+      "grad_norm": 1.273081660270691,
+      "learning_rate": 0.0001146246270862144,
+      "loss": 2.4152,
+      "step": 7374
+    },
+    {
+      "epoch": 0.45367864173228345,
+      "grad_norm": 1.3050533533096313,
+      "learning_rate": 0.00011460549708914244,
+      "loss": 2.1362,
+      "step": 7375
+    },
+    {
+      "epoch": 0.453740157480315,
+      "grad_norm": 1.2728328704833984,
+      "learning_rate": 0.00011458636654590632,
+      "loss": 2.7394,
+      "step": 7376
+    },
+    {
+      "epoch": 0.45380167322834647,
+      "grad_norm": 1.2352139949798584,
+      "learning_rate": 0.00011456723545722143,
+      "loss": 2.8681,
+      "step": 7377
+    },
+    {
+      "epoch": 0.45386318897637795,
+      "grad_norm": 1.7464158535003662,
+      "learning_rate": 0.0001145481038238032,
+      "loss": 2.515,
+      "step": 7378
+    },
+    {
+      "epoch": 0.45392470472440943,
+      "grad_norm": 1.7305097579956055,
+      "learning_rate": 0.00011452897164636696,
+      "loss": 2.6091,
+      "step": 7379
+    },
+    {
+      "epoch": 0.45398622047244097,
+      "grad_norm": 1.1684355735778809,
+      "learning_rate": 0.00011450983892562824,
+      "loss": 2.4317,
+      "step": 7380
+    },
+    {
+      "epoch": 0.45404773622047245,
+      "grad_norm": 1.1020153760910034,
+      "learning_rate": 0.00011449070566230242,
+      "loss": 3.0998,
+      "step": 7381
+    },
+    {
+      "epoch": 0.45410925196850394,
+      "grad_norm": 2.2732198238372803,
+      "learning_rate": 0.00011447157185710501,
+      "loss": 2.7789,
+      "step": 7382
+    },
+    {
+      "epoch": 0.4541707677165354,
+      "grad_norm": 2.2341556549072266,
+      "learning_rate": 0.00011445243751075153,
+      "loss": 2.2133,
+      "step": 7383
+    },
+    {
+      "epoch": 0.45423228346456695,
+      "grad_norm": 1.290846347808838,
+      "learning_rate": 0.00011443330262395746,
+      "loss": 2.7065,
+      "step": 7384
+    },
+    {
+      "epoch": 0.45429379921259844,
+      "grad_norm": 1.3549619913101196,
+      "learning_rate": 0.00011441416719743835,
+      "loss": 2.5692,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4543553149606299,
+      "grad_norm": 1.3306411504745483,
+      "learning_rate": 0.00011439503123190978,
+      "loss": 2.2107,
+      "step": 7386
+    },
+    {
+      "epoch": 0.4544168307086614,
+      "grad_norm": 2.0418808460235596,
+      "learning_rate": 0.0001143758947280873,
+      "loss": 2.5019,
+      "step": 7387
+    },
+    {
+      "epoch": 0.45447834645669294,
+      "grad_norm": 1.1263066530227661,
+      "learning_rate": 0.0001143567576866865,
+      "loss": 2.4477,
+      "step": 7388
+    },
+    {
+      "epoch": 0.4545398622047244,
+      "grad_norm": 1.2538725137710571,
+      "learning_rate": 0.00011433762010842307,
+      "loss": 2.3897,
+      "step": 7389
+    },
+    {
+      "epoch": 0.4546013779527559,
+      "grad_norm": 1.703273892402649,
+      "learning_rate": 0.00011431848199401256,
+      "loss": 2.3574,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4546628937007874,
+      "grad_norm": 1.0554031133651733,
+      "learning_rate": 0.00011429934334417068,
+      "loss": 2.9208,
+      "step": 7391
+    },
+    {
+      "epoch": 0.4547244094488189,
+      "grad_norm": 1.0379326343536377,
+      "learning_rate": 0.00011428020415961307,
+      "loss": 3.0343,
+      "step": 7392
+    },
+    {
+      "epoch": 0.4547859251968504,
+      "grad_norm": 1.3141963481903076,
+      "learning_rate": 0.00011426106444105546,
+      "loss": 1.9997,
+      "step": 7393
+    },
+    {
+      "epoch": 0.4548474409448819,
+      "grad_norm": 1.0856488943099976,
+      "learning_rate": 0.0001142419241892136,
+      "loss": 3.2372,
+      "step": 7394
+    },
+    {
+      "epoch": 0.45490895669291337,
+      "grad_norm": 1.0863926410675049,
+      "learning_rate": 0.00011422278340480312,
+      "loss": 2.4207,
+      "step": 7395
+    },
+    {
+      "epoch": 0.4549704724409449,
+      "grad_norm": 1.1398016214370728,
+      "learning_rate": 0.00011420364208853991,
+      "loss": 2.5654,
+      "step": 7396
+    },
+    {
+      "epoch": 0.4550319881889764,
+      "grad_norm": 1.0828161239624023,
+      "learning_rate": 0.00011418450024113967,
+      "loss": 2.6037,
+      "step": 7397
+    },
+    {
+      "epoch": 0.45509350393700787,
+      "grad_norm": 1.1970421075820923,
+      "learning_rate": 0.00011416535786331822,
+      "loss": 2.9217,
+      "step": 7398
+    },
+    {
+      "epoch": 0.45515501968503935,
+      "grad_norm": 1.1582391262054443,
+      "learning_rate": 0.00011414621495579137,
+      "loss": 2.7537,
+      "step": 7399
+    },
+    {
+      "epoch": 0.4552165354330709,
+      "grad_norm": 1.137813687324524,
+      "learning_rate": 0.00011412707151927496,
+      "loss": 2.6815,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4552780511811024,
+      "grad_norm": 1.3191426992416382,
+      "learning_rate": 0.00011410792755448482,
+      "loss": 2.2367,
+      "step": 7401
+    },
+    {
+      "epoch": 0.45533956692913385,
+      "grad_norm": 1.0935916900634766,
+      "learning_rate": 0.00011408878306213691,
+      "loss": 2.6726,
+      "step": 7402
+    },
+    {
+      "epoch": 0.45540108267716534,
+      "grad_norm": 1.2322585582733154,
+      "learning_rate": 0.00011406963804294706,
+      "loss": 2.7186,
+      "step": 7403
+    },
+    {
+      "epoch": 0.4554625984251969,
+      "grad_norm": 1.043876051902771,
+      "learning_rate": 0.0001140504924976312,
+      "loss": 2.1142,
+      "step": 7404
+    },
+    {
+      "epoch": 0.45552411417322836,
+      "grad_norm": 1.2368282079696655,
+      "learning_rate": 0.00011403134642690522,
+      "loss": 2.5405,
+      "step": 7405
+    },
+    {
+      "epoch": 0.45558562992125984,
+      "grad_norm": 2.3215577602386475,
+      "learning_rate": 0.00011401219983148519,
+      "loss": 2.356,
+      "step": 7406
+    },
+    {
+      "epoch": 0.4556471456692913,
+      "grad_norm": 1.1463490724563599,
+      "learning_rate": 0.000113993052712087,
+      "loss": 2.9361,
+      "step": 7407
+    },
+    {
+      "epoch": 0.45570866141732286,
+      "grad_norm": 1.1289856433868408,
+      "learning_rate": 0.00011397390506942665,
+      "loss": 2.2132,
+      "step": 7408
+    },
+    {
+      "epoch": 0.45577017716535434,
+      "grad_norm": 1.4174175262451172,
+      "learning_rate": 0.0001139547569042202,
+      "loss": 2.3085,
+      "step": 7409
+    },
+    {
+      "epoch": 0.4558316929133858,
+      "grad_norm": 1.5956746339797974,
+      "learning_rate": 0.00011393560821718363,
+      "loss": 2.2563,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4558932086614173,
+      "grad_norm": 1.1217842102050781,
+      "learning_rate": 0.00011391645900903301,
+      "loss": 2.6645,
+      "step": 7411
+    },
+    {
+      "epoch": 0.45595472440944884,
+      "grad_norm": 1.1716842651367188,
+      "learning_rate": 0.00011389730928048445,
+      "loss": 2.5709,
+      "step": 7412
+    },
+    {
+      "epoch": 0.4560162401574803,
+      "grad_norm": 1.3322558403015137,
+      "learning_rate": 0.00011387815903225402,
+      "loss": 2.4588,
+      "step": 7413
+    },
+    {
+      "epoch": 0.4560777559055118,
+      "grad_norm": 1.1000797748565674,
+      "learning_rate": 0.0001138590082650578,
+      "loss": 2.3958,
+      "step": 7414
+    },
+    {
+      "epoch": 0.4561392716535433,
+      "grad_norm": 1.164061427116394,
+      "learning_rate": 0.00011383985697961197,
+      "loss": 2.5353,
+      "step": 7415
+    },
+    {
+      "epoch": 0.4562007874015748,
+      "grad_norm": 0.993810772895813,
+      "learning_rate": 0.00011382070517663268,
+      "loss": 2.7656,
+      "step": 7416
+    },
+    {
+      "epoch": 0.4562623031496063,
+      "grad_norm": 1.29926335811615,
+      "learning_rate": 0.00011380155285683607,
+      "loss": 2.3663,
+      "step": 7417
+    },
+    {
+      "epoch": 0.4563238188976378,
+      "grad_norm": 1.4724915027618408,
+      "learning_rate": 0.00011378240002093835,
+      "loss": 3.3359,
+      "step": 7418
+    },
+    {
+      "epoch": 0.45638533464566927,
+      "grad_norm": 1.0891472101211548,
+      "learning_rate": 0.0001137632466696557,
+      "loss": 2.9507,
+      "step": 7419
+    },
+    {
+      "epoch": 0.4564468503937008,
+      "grad_norm": 1.1248326301574707,
+      "learning_rate": 0.00011374409280370438,
+      "loss": 2.5213,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4565083661417323,
+      "grad_norm": 1.0272067785263062,
+      "learning_rate": 0.0001137249384238006,
+      "loss": 3.4131,
+      "step": 7421
+    },
+    {
+      "epoch": 0.4565698818897638,
+      "grad_norm": 1.3980693817138672,
+      "learning_rate": 0.0001137057835306607,
+      "loss": 2.2979,
+      "step": 7422
+    },
+    {
+      "epoch": 0.45663139763779526,
+      "grad_norm": 1.2884955406188965,
+      "learning_rate": 0.0001136866281250009,
+      "loss": 2.097,
+      "step": 7423
+    },
+    {
+      "epoch": 0.4566929133858268,
+      "grad_norm": 2.0077977180480957,
+      "learning_rate": 0.00011366747220753752,
+      "loss": 3.2247,
+      "step": 7424
+    },
+    {
+      "epoch": 0.4567544291338583,
+      "grad_norm": 1.353615403175354,
+      "learning_rate": 0.00011364831577898691,
+      "loss": 2.3711,
+      "step": 7425
+    },
+    {
+      "epoch": 0.45681594488188976,
+      "grad_norm": 1.2710208892822266,
+      "learning_rate": 0.0001136291588400654,
+      "loss": 2.7328,
+      "step": 7426
+    },
+    {
+      "epoch": 0.45687746062992124,
+      "grad_norm": 1.2182353734970093,
+      "learning_rate": 0.00011361000139148932,
+      "loss": 2.7602,
+      "step": 7427
+    },
+    {
+      "epoch": 0.4569389763779528,
+      "grad_norm": 1.6396665573120117,
+      "learning_rate": 0.0001135908434339751,
+      "loss": 2.12,
+      "step": 7428
+    },
+    {
+      "epoch": 0.45700049212598426,
+      "grad_norm": 1.102015733718872,
+      "learning_rate": 0.00011357168496823914,
+      "loss": 2.3088,
+      "step": 7429
+    },
+    {
+      "epoch": 0.45706200787401574,
+      "grad_norm": 1.4561246633529663,
+      "learning_rate": 0.00011355252599499779,
+      "loss": 2.4806,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4571235236220472,
+      "grad_norm": 1.0963581800460815,
+      "learning_rate": 0.00011353336651496756,
+      "loss": 2.9237,
+      "step": 7431
+    },
+    {
+      "epoch": 0.45718503937007876,
+      "grad_norm": 1.416367769241333,
+      "learning_rate": 0.0001135142065288649,
+      "loss": 2.1536,
+      "step": 7432
+    },
+    {
+      "epoch": 0.45724655511811024,
+      "grad_norm": 1.2952618598937988,
+      "learning_rate": 0.00011349504603740624,
+      "loss": 2.4657,
+      "step": 7433
+    },
+    {
+      "epoch": 0.4573080708661417,
+      "grad_norm": 0.9899798035621643,
+      "learning_rate": 0.00011347588504130815,
+      "loss": 2.8039,
+      "step": 7434
+    },
+    {
+      "epoch": 0.4573695866141732,
+      "grad_norm": 1.588598608970642,
+      "learning_rate": 0.00011345672354128708,
+      "loss": 2.9228,
+      "step": 7435
+    },
+    {
+      "epoch": 0.45743110236220474,
+      "grad_norm": 1.2251744270324707,
+      "learning_rate": 0.00011343756153805957,
+      "loss": 2.8363,
+      "step": 7436
+    },
+    {
+      "epoch": 0.4574926181102362,
+      "grad_norm": 0.9843764901161194,
+      "learning_rate": 0.00011341839903234222,
+      "loss": 3.1192,
+      "step": 7437
+    },
+    {
+      "epoch": 0.4575541338582677,
+      "grad_norm": 1.4905023574829102,
+      "learning_rate": 0.00011339923602485156,
+      "loss": 2.6487,
+      "step": 7438
+    },
+    {
+      "epoch": 0.4576156496062992,
+      "grad_norm": 1.1291911602020264,
+      "learning_rate": 0.0001133800725163042,
+      "loss": 2.832,
+      "step": 7439
+    },
+    {
+      "epoch": 0.45767716535433073,
+      "grad_norm": 1.1055399179458618,
+      "learning_rate": 0.00011336090850741672,
+      "loss": 1.9525,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4577386811023622,
+      "grad_norm": 1.0789868831634521,
+      "learning_rate": 0.00011334174399890576,
+      "loss": 2.5873,
+      "step": 7441
+    },
+    {
+      "epoch": 0.4578001968503937,
+      "grad_norm": 1.0202945470809937,
+      "learning_rate": 0.000113322578991488,
+      "loss": 2.7841,
+      "step": 7442
+    },
+    {
+      "epoch": 0.4578617125984252,
+      "grad_norm": 1.2449884414672852,
+      "learning_rate": 0.00011330341348588003,
+      "loss": 2.3168,
+      "step": 7443
+    },
+    {
+      "epoch": 0.4579232283464567,
+      "grad_norm": 1.320665717124939,
+      "learning_rate": 0.0001132842474827986,
+      "loss": 3.0309,
+      "step": 7444
+    },
+    {
+      "epoch": 0.4579847440944882,
+      "grad_norm": 1.1471648216247559,
+      "learning_rate": 0.0001132650809829604,
+      "loss": 2.6844,
+      "step": 7445
+    },
+    {
+      "epoch": 0.4580462598425197,
+      "grad_norm": 1.4947344064712524,
+      "learning_rate": 0.00011324591398708211,
+      "loss": 2.1379,
+      "step": 7446
+    },
+    {
+      "epoch": 0.45810777559055116,
+      "grad_norm": 1.1956945657730103,
+      "learning_rate": 0.00011322674649588052,
+      "loss": 2.6938,
+      "step": 7447
+    },
+    {
+      "epoch": 0.4581692913385827,
+      "grad_norm": 1.3076856136322021,
+      "learning_rate": 0.00011320757851007238,
+      "loss": 3.1313,
+      "step": 7448
+    },
+    {
+      "epoch": 0.4582308070866142,
+      "grad_norm": 1.1645845174789429,
+      "learning_rate": 0.00011318841003037441,
+      "loss": 2.7634,
+      "step": 7449
+    },
+    {
+      "epoch": 0.45829232283464566,
+      "grad_norm": 1.1717803478240967,
+      "learning_rate": 0.00011316924105750348,
+      "loss": 2.5476,
+      "step": 7450
+    },
+    {
+      "epoch": 0.45835383858267714,
+      "grad_norm": 1.1088520288467407,
+      "learning_rate": 0.00011315007159217636,
+      "loss": 2.9616,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4584153543307087,
+      "grad_norm": 1.1537271738052368,
+      "learning_rate": 0.00011313090163510991,
+      "loss": 2.481,
+      "step": 7452
+    },
+    {
+      "epoch": 0.45847687007874016,
+      "grad_norm": 1.491597294807434,
+      "learning_rate": 0.00011311173118702089,
+      "loss": 2.4096,
+      "step": 7453
+    },
+    {
+      "epoch": 0.45853838582677164,
+      "grad_norm": 1.1233630180358887,
+      "learning_rate": 0.00011309256024862633,
+      "loss": 3.3258,
+      "step": 7454
+    },
+    {
+      "epoch": 0.4585999015748031,
+      "grad_norm": 1.188530683517456,
+      "learning_rate": 0.00011307338882064295,
+      "loss": 2.303,
+      "step": 7455
+    },
+    {
+      "epoch": 0.45866141732283466,
+      "grad_norm": 1.1864863634109497,
+      "learning_rate": 0.00011305421690378777,
+      "loss": 2.6473,
+      "step": 7456
+    },
+    {
+      "epoch": 0.45872293307086615,
+      "grad_norm": 1.1912932395935059,
+      "learning_rate": 0.00011303504449877766,
+      "loss": 2.5271,
+      "step": 7457
+    },
+    {
+      "epoch": 0.4587844488188976,
+      "grad_norm": 1.2872440814971924,
+      "learning_rate": 0.00011301587160632954,
+      "loss": 2.6921,
+      "step": 7458
+    },
+    {
+      "epoch": 0.4588459645669291,
+      "grad_norm": 1.1583061218261719,
+      "learning_rate": 0.00011299669822716043,
+      "loss": 2.8903,
+      "step": 7459
+    },
+    {
+      "epoch": 0.45890748031496065,
+      "grad_norm": 1.187064528465271,
+      "learning_rate": 0.00011297752436198727,
+      "loss": 3.0802,
+      "step": 7460
+    },
+    {
+      "epoch": 0.45896899606299213,
+      "grad_norm": 1.1764544248580933,
+      "learning_rate": 0.00011295835001152708,
+      "loss": 2.0467,
+      "step": 7461
+    },
+    {
+      "epoch": 0.4590305118110236,
+      "grad_norm": 1.1773542165756226,
+      "learning_rate": 0.00011293917517649684,
+      "loss": 2.2544,
+      "step": 7462
+    },
+    {
+      "epoch": 0.4590920275590551,
+      "grad_norm": 1.199644684791565,
+      "learning_rate": 0.0001129199998576136,
+      "loss": 3.3159,
+      "step": 7463
+    },
+    {
+      "epoch": 0.45915354330708663,
+      "grad_norm": 1.1578497886657715,
+      "learning_rate": 0.00011290082405559441,
+      "loss": 2.2698,
+      "step": 7464
+    },
+    {
+      "epoch": 0.4592150590551181,
+      "grad_norm": 1.0924227237701416,
+      "learning_rate": 0.00011288164777115633,
+      "loss": 2.9672,
+      "step": 7465
+    },
+    {
+      "epoch": 0.4592765748031496,
+      "grad_norm": 1.087121605873108,
+      "learning_rate": 0.00011286247100501647,
+      "loss": 2.5173,
+      "step": 7466
+    },
+    {
+      "epoch": 0.4593380905511811,
+      "grad_norm": 1.1418917179107666,
+      "learning_rate": 0.00011284329375789192,
+      "loss": 2.777,
+      "step": 7467
+    },
+    {
+      "epoch": 0.4593996062992126,
+      "grad_norm": 2.184501886367798,
+      "learning_rate": 0.00011282411603049976,
+      "loss": 2.3694,
+      "step": 7468
+    },
+    {
+      "epoch": 0.4594611220472441,
+      "grad_norm": 1.7589701414108276,
+      "learning_rate": 0.00011280493782355721,
+      "loss": 2.5371,
+      "step": 7469
+    },
+    {
+      "epoch": 0.4595226377952756,
+      "grad_norm": 1.211014747619629,
+      "learning_rate": 0.00011278575913778138,
+      "loss": 2.2642,
+      "step": 7470
+    },
+    {
+      "epoch": 0.45958415354330706,
+      "grad_norm": 1.1309539079666138,
+      "learning_rate": 0.00011276657997388944,
+      "loss": 2.3792,
+      "step": 7471
+    },
+    {
+      "epoch": 0.4596456692913386,
+      "grad_norm": 1.1750813722610474,
+      "learning_rate": 0.00011274740033259861,
+      "loss": 2.0622,
+      "step": 7472
+    },
+    {
+      "epoch": 0.4597071850393701,
+      "grad_norm": 1.211849331855774,
+      "learning_rate": 0.0001127282202146261,
+      "loss": 2.6558,
+      "step": 7473
+    },
+    {
+      "epoch": 0.45976870078740156,
+      "grad_norm": 1.285157561302185,
+      "learning_rate": 0.00011270903962068913,
+      "loss": 2.2046,
+      "step": 7474
+    },
+    {
+      "epoch": 0.45983021653543305,
+      "grad_norm": 1.0701028108596802,
+      "learning_rate": 0.0001126898585515049,
+      "loss": 2.9989,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4598917322834646,
+      "grad_norm": 1.4782801866531372,
+      "learning_rate": 0.00011267067700779075,
+      "loss": 2.9252,
+      "step": 7476
+    },
+    {
+      "epoch": 0.45995324803149606,
+      "grad_norm": 1.243680715560913,
+      "learning_rate": 0.00011265149499026395,
+      "loss": 3.2809,
+      "step": 7477
+    },
+    {
+      "epoch": 0.46001476377952755,
+      "grad_norm": 2.015550136566162,
+      "learning_rate": 0.00011263231249964174,
+      "loss": 2.5724,
+      "step": 7478
+    },
+    {
+      "epoch": 0.46007627952755903,
+      "grad_norm": 1.133969783782959,
+      "learning_rate": 0.00011261312953664153,
+      "loss": 3.1682,
+      "step": 7479
+    },
+    {
+      "epoch": 0.46013779527559057,
+      "grad_norm": 1.3714519739151,
+      "learning_rate": 0.0001125939461019806,
+      "loss": 2.9396,
+      "step": 7480
+    },
+    {
+      "epoch": 0.46019931102362205,
+      "grad_norm": 1.0491626262664795,
+      "learning_rate": 0.00011257476219637628,
+      "loss": 2.2711,
+      "step": 7481
+    },
+    {
+      "epoch": 0.46026082677165353,
+      "grad_norm": 1.327006220817566,
+      "learning_rate": 0.000112555577820546,
+      "loss": 2.8028,
+      "step": 7482
+    },
+    {
+      "epoch": 0.460322342519685,
+      "grad_norm": 1.271369457244873,
+      "learning_rate": 0.00011253639297520713,
+      "loss": 2.6227,
+      "step": 7483
+    },
+    {
+      "epoch": 0.46038385826771655,
+      "grad_norm": 1.1536035537719727,
+      "learning_rate": 0.00011251720766107703,
+      "loss": 2.3323,
+      "step": 7484
+    },
+    {
+      "epoch": 0.46044537401574803,
+      "grad_norm": 1.125225305557251,
+      "learning_rate": 0.00011249802187887319,
+      "loss": 2.7334,
+      "step": 7485
+    },
+    {
+      "epoch": 0.4605068897637795,
+      "grad_norm": 1.4758952856063843,
+      "learning_rate": 0.00011247883562931302,
+      "loss": 2.6533,
+      "step": 7486
+    },
+    {
+      "epoch": 0.460568405511811,
+      "grad_norm": 1.1073410511016846,
+      "learning_rate": 0.00011245964891311397,
+      "loss": 2.645,
+      "step": 7487
+    },
+    {
+      "epoch": 0.46062992125984253,
+      "grad_norm": 1.3270540237426758,
+      "learning_rate": 0.00011244046173099354,
+      "loss": 2.7361,
+      "step": 7488
+    },
+    {
+      "epoch": 0.460691437007874,
+      "grad_norm": 1.1988890171051025,
+      "learning_rate": 0.0001124212740836692,
+      "loss": 3.1603,
+      "step": 7489
+    },
+    {
+      "epoch": 0.4607529527559055,
+      "grad_norm": 1.083404541015625,
+      "learning_rate": 0.00011240208597185848,
+      "loss": 2.9139,
+      "step": 7490
+    },
+    {
+      "epoch": 0.460814468503937,
+      "grad_norm": 1.1376014947891235,
+      "learning_rate": 0.00011238289739627892,
+      "loss": 2.6807,
+      "step": 7491
+    },
+    {
+      "epoch": 0.4608759842519685,
+      "grad_norm": 1.0313570499420166,
+      "learning_rate": 0.00011236370835764802,
+      "loss": 3.2745,
+      "step": 7492
+    },
+    {
+      "epoch": 0.4609375,
+      "grad_norm": 1.2576545476913452,
+      "learning_rate": 0.00011234451885668337,
+      "loss": 2.7145,
+      "step": 7493
+    },
+    {
+      "epoch": 0.4609990157480315,
+      "grad_norm": 1.2354590892791748,
+      "learning_rate": 0.00011232532889410256,
+      "loss": 2.3663,
+      "step": 7494
+    },
+    {
+      "epoch": 0.461060531496063,
+      "grad_norm": 1.0852904319763184,
+      "learning_rate": 0.00011230613847062317,
+      "loss": 3.0312,
+      "step": 7495
+    },
+    {
+      "epoch": 0.4611220472440945,
+      "grad_norm": 1.2628062963485718,
+      "learning_rate": 0.00011228694758696284,
+      "loss": 2.4435,
+      "step": 7496
+    },
+    {
+      "epoch": 0.461183562992126,
+      "grad_norm": 1.4178674221038818,
+      "learning_rate": 0.00011226775624383914,
+      "loss": 2.3724,
+      "step": 7497
+    },
+    {
+      "epoch": 0.46124507874015747,
+      "grad_norm": 1.2509007453918457,
+      "learning_rate": 0.00011224856444196981,
+      "loss": 2.6206,
+      "step": 7498
+    },
+    {
+      "epoch": 0.461306594488189,
+      "grad_norm": 1.134913682937622,
+      "learning_rate": 0.00011222937218207246,
+      "loss": 2.638,
+      "step": 7499
+    },
+    {
+      "epoch": 0.4613681102362205,
+      "grad_norm": 1.3161927461624146,
+      "learning_rate": 0.00011221017946486476,
+      "loss": 1.898,
+      "step": 7500
+    },
+    {
+      "epoch": 0.46142962598425197,
+      "grad_norm": 1.0478384494781494,
+      "learning_rate": 0.00011219098629106444,
+      "loss": 2.6078,
+      "step": 7501
+    },
+    {
+      "epoch": 0.46149114173228345,
+      "grad_norm": 1.3246824741363525,
+      "learning_rate": 0.00011217179266138924,
+      "loss": 2.6807,
+      "step": 7502
+    },
+    {
+      "epoch": 0.461552657480315,
+      "grad_norm": 1.457833170890808,
+      "learning_rate": 0.00011215259857655682,
+      "loss": 2.7717,
+      "step": 7503
+    },
+    {
+      "epoch": 0.46161417322834647,
+      "grad_norm": 1.1017242670059204,
+      "learning_rate": 0.00011213340403728501,
+      "loss": 2.9062,
+      "step": 7504
+    },
+    {
+      "epoch": 0.46167568897637795,
+      "grad_norm": 1.089836597442627,
+      "learning_rate": 0.00011211420904429155,
+      "loss": 2.7422,
+      "step": 7505
+    },
+    {
+      "epoch": 0.46173720472440943,
+      "grad_norm": 1.321114420890808,
+      "learning_rate": 0.00011209501359829422,
+      "loss": 2.8367,
+      "step": 7506
+    },
+    {
+      "epoch": 0.46179872047244097,
+      "grad_norm": 1.2452548742294312,
+      "learning_rate": 0.00011207581770001081,
+      "loss": 2.2712,
+      "step": 7507
+    },
+    {
+      "epoch": 0.46186023622047245,
+      "grad_norm": 1.4404551982879639,
+      "learning_rate": 0.00011205662135015915,
+      "loss": 2.8792,
+      "step": 7508
+    },
+    {
+      "epoch": 0.46192175196850394,
+      "grad_norm": 1.2659788131713867,
+      "learning_rate": 0.00011203742454945711,
+      "loss": 2.687,
+      "step": 7509
+    },
+    {
+      "epoch": 0.4619832677165354,
+      "grad_norm": 1.4567406177520752,
+      "learning_rate": 0.00011201822729862248,
+      "loss": 2.0913,
+      "step": 7510
+    },
+    {
+      "epoch": 0.46204478346456695,
+      "grad_norm": 1.2041363716125488,
+      "learning_rate": 0.00011199902959837319,
+      "loss": 2.5169,
+      "step": 7511
+    },
+    {
+      "epoch": 0.46210629921259844,
+      "grad_norm": 1.3648641109466553,
+      "learning_rate": 0.0001119798314494271,
+      "loss": 2.9307,
+      "step": 7512
+    },
+    {
+      "epoch": 0.4621678149606299,
+      "grad_norm": 1.1820363998413086,
+      "learning_rate": 0.00011196063285250208,
+      "loss": 2.3273,
+      "step": 7513
+    },
+    {
+      "epoch": 0.4622293307086614,
+      "grad_norm": 1.1343661546707153,
+      "learning_rate": 0.00011194143380831614,
+      "loss": 2.8638,
+      "step": 7514
+    },
+    {
+      "epoch": 0.46229084645669294,
+      "grad_norm": 1.1391226053237915,
+      "learning_rate": 0.00011192223431758715,
+      "loss": 2.9058,
+      "step": 7515
+    },
+    {
+      "epoch": 0.4623523622047244,
+      "grad_norm": 1.1530513763427734,
+      "learning_rate": 0.00011190303438103304,
+      "loss": 2.3355,
+      "step": 7516
+    },
+    {
+      "epoch": 0.4624138779527559,
+      "grad_norm": 1.107230305671692,
+      "learning_rate": 0.00011188383399937185,
+      "loss": 2.3581,
+      "step": 7517
+    },
+    {
+      "epoch": 0.4624753937007874,
+      "grad_norm": 1.3564728498458862,
+      "learning_rate": 0.00011186463317332152,
+      "loss": 2.5207,
+      "step": 7518
+    },
+    {
+      "epoch": 0.4625369094488189,
+      "grad_norm": 1.1170448064804077,
+      "learning_rate": 0.00011184543190360008,
+      "loss": 2.3036,
+      "step": 7519
+    },
+    {
+      "epoch": 0.4625984251968504,
+      "grad_norm": 1.284211277961731,
+      "learning_rate": 0.00011182623019092553,
+      "loss": 2.1578,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4626599409448819,
+      "grad_norm": 1.1809130907058716,
+      "learning_rate": 0.00011180702803601593,
+      "loss": 2.7293,
+      "step": 7521
+    },
+    {
+      "epoch": 0.46272145669291337,
+      "grad_norm": 1.0251449346542358,
+      "learning_rate": 0.0001117878254395893,
+      "loss": 2.9178,
+      "step": 7522
+    },
+    {
+      "epoch": 0.4627829724409449,
+      "grad_norm": 1.1819628477096558,
+      "learning_rate": 0.00011176862240236372,
+      "loss": 2.9494,
+      "step": 7523
+    },
+    {
+      "epoch": 0.4628444881889764,
+      "grad_norm": 1.2284741401672363,
+      "learning_rate": 0.00011174941892505733,
+      "loss": 2.842,
+      "step": 7524
+    },
+    {
+      "epoch": 0.46290600393700787,
+      "grad_norm": 1.3917728662490845,
+      "learning_rate": 0.00011173021500838816,
+      "loss": 2.7831,
+      "step": 7525
+    },
+    {
+      "epoch": 0.46296751968503935,
+      "grad_norm": 1.1448485851287842,
+      "learning_rate": 0.00011171101065307434,
+      "loss": 2.7571,
+      "step": 7526
+    },
+    {
+      "epoch": 0.4630290354330709,
+      "grad_norm": 1.0387544631958008,
+      "learning_rate": 0.00011169180585983408,
+      "loss": 2.8677,
+      "step": 7527
+    },
+    {
+      "epoch": 0.4630905511811024,
+      "grad_norm": 1.099817156791687,
+      "learning_rate": 0.00011167260062938544,
+      "loss": 2.8253,
+      "step": 7528
+    },
+    {
+      "epoch": 0.46315206692913385,
+      "grad_norm": 0.8956767916679382,
+      "learning_rate": 0.00011165339496244663,
+      "loss": 2.9606,
+      "step": 7529
+    },
+    {
+      "epoch": 0.46321358267716534,
+      "grad_norm": 1.6513981819152832,
+      "learning_rate": 0.00011163418885973582,
+      "loss": 1.9689,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4632750984251969,
+      "grad_norm": 1.1797908544540405,
+      "learning_rate": 0.00011161498232197124,
+      "loss": 2.9193,
+      "step": 7531
+    },
+    {
+      "epoch": 0.46333661417322836,
+      "grad_norm": 1.0555920600891113,
+      "learning_rate": 0.00011159577534987105,
+      "loss": 2.8958,
+      "step": 7532
+    },
+    {
+      "epoch": 0.46339812992125984,
+      "grad_norm": 1.0968912839889526,
+      "learning_rate": 0.00011157656794415354,
+      "loss": 2.2837,
+      "step": 7533
+    },
+    {
+      "epoch": 0.4634596456692913,
+      "grad_norm": 1.28616464138031,
+      "learning_rate": 0.00011155736010553695,
+      "loss": 2.5375,
+      "step": 7534
+    },
+    {
+      "epoch": 0.46352116141732286,
+      "grad_norm": 2.7004458904266357,
+      "learning_rate": 0.00011153815183473953,
+      "loss": 2.3276,
+      "step": 7535
+    },
+    {
+      "epoch": 0.46358267716535434,
+      "grad_norm": 1.7902352809906006,
+      "learning_rate": 0.00011151894313247957,
+      "loss": 1.0299,
+      "step": 7536
+    },
+    {
+      "epoch": 0.4636441929133858,
+      "grad_norm": 1.1578590869903564,
+      "learning_rate": 0.00011149973399947537,
+      "loss": 2.524,
+      "step": 7537
+    },
+    {
+      "epoch": 0.4637057086614173,
+      "grad_norm": 1.2919212579727173,
+      "learning_rate": 0.00011148052443644524,
+      "loss": 2.5198,
+      "step": 7538
+    },
+    {
+      "epoch": 0.46376722440944884,
+      "grad_norm": 1.2200170755386353,
+      "learning_rate": 0.00011146131444410752,
+      "loss": 2.4344,
+      "step": 7539
+    },
+    {
+      "epoch": 0.4638287401574803,
+      "grad_norm": 1.3752554655075073,
+      "learning_rate": 0.00011144210402318057,
+      "loss": 2.836,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4638902559055118,
+      "grad_norm": 1.2098129987716675,
+      "learning_rate": 0.0001114228931743827,
+      "loss": 2.3717,
+      "step": 7541
+    },
+    {
+      "epoch": 0.4639517716535433,
+      "grad_norm": 1.2419377565383911,
+      "learning_rate": 0.00011140368189843234,
+      "loss": 2.9954,
+      "step": 7542
+    },
+    {
+      "epoch": 0.4640132874015748,
+      "grad_norm": 1.1783561706542969,
+      "learning_rate": 0.00011138447019604786,
+      "loss": 2.4572,
+      "step": 7543
+    },
+    {
+      "epoch": 0.4640748031496063,
+      "grad_norm": 1.0008645057678223,
+      "learning_rate": 0.00011136525806794767,
+      "loss": 2.7371,
+      "step": 7544
+    },
+    {
+      "epoch": 0.4641363188976378,
+      "grad_norm": 1.5232173204421997,
+      "learning_rate": 0.00011134604551485019,
+      "loss": 2.3386,
+      "step": 7545
+    },
+    {
+      "epoch": 0.46419783464566927,
+      "grad_norm": 1.1098322868347168,
+      "learning_rate": 0.0001113268325374739,
+      "loss": 2.7325,
+      "step": 7546
+    },
+    {
+      "epoch": 0.4642593503937008,
+      "grad_norm": 1.4458621740341187,
+      "learning_rate": 0.00011130761913653723,
+      "loss": 2.4888,
+      "step": 7547
+    },
+    {
+      "epoch": 0.4643208661417323,
+      "grad_norm": 0.9378423094749451,
+      "learning_rate": 0.00011128840531275864,
+      "loss": 3.0821,
+      "step": 7548
+    },
+    {
+      "epoch": 0.4643823818897638,
+      "grad_norm": 1.3114824295043945,
+      "learning_rate": 0.00011126919106685665,
+      "loss": 1.5662,
+      "step": 7549
+    },
+    {
+      "epoch": 0.46444389763779526,
+      "grad_norm": 1.2314255237579346,
+      "learning_rate": 0.00011124997639954975,
+      "loss": 2.4152,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4645054133858268,
+      "grad_norm": 1.3071725368499756,
+      "learning_rate": 0.00011123076131155646,
+      "loss": 2.5453,
+      "step": 7551
+    },
+    {
+      "epoch": 0.4645669291338583,
+      "grad_norm": 1.2191252708435059,
+      "learning_rate": 0.00011121154580359534,
+      "loss": 2.5827,
+      "step": 7552
+    },
+    {
+      "epoch": 0.46462844488188976,
+      "grad_norm": 0.9909071326255798,
+      "learning_rate": 0.0001111923298763849,
+      "loss": 2.7269,
+      "step": 7553
+    },
+    {
+      "epoch": 0.46468996062992124,
+      "grad_norm": 2.447213888168335,
+      "learning_rate": 0.00011117311353064376,
+      "loss": 2.6651,
+      "step": 7554
+    },
+    {
+      "epoch": 0.4647514763779528,
+      "grad_norm": 1.7646270990371704,
+      "learning_rate": 0.00011115389676709045,
+      "loss": 2.388,
+      "step": 7555
+    },
+    {
+      "epoch": 0.46481299212598426,
+      "grad_norm": 1.3365836143493652,
+      "learning_rate": 0.00011113467958644361,
+      "loss": 2.355,
+      "step": 7556
+    },
+    {
+      "epoch": 0.46487450787401574,
+      "grad_norm": 1.59646737575531,
+      "learning_rate": 0.00011111546198942184,
+      "loss": 2.805,
+      "step": 7557
+    },
+    {
+      "epoch": 0.4649360236220472,
+      "grad_norm": 1.3909296989440918,
+      "learning_rate": 0.00011109624397674376,
+      "loss": 3.2943,
+      "step": 7558
+    },
+    {
+      "epoch": 0.46499753937007876,
+      "grad_norm": 1.3041176795959473,
+      "learning_rate": 0.00011107702554912805,
+      "loss": 2.7641,
+      "step": 7559
+    },
+    {
+      "epoch": 0.46505905511811024,
+      "grad_norm": 1.096982717514038,
+      "learning_rate": 0.00011105780670729335,
+      "loss": 2.081,
+      "step": 7560
+    },
+    {
+      "epoch": 0.4651205708661417,
+      "grad_norm": 1.1798291206359863,
+      "learning_rate": 0.00011103858745195832,
+      "loss": 2.7157,
+      "step": 7561
+    },
+    {
+      "epoch": 0.4651820866141732,
+      "grad_norm": 1.1666587591171265,
+      "learning_rate": 0.0001110193677838417,
+      "loss": 2.587,
+      "step": 7562
+    },
+    {
+      "epoch": 0.46524360236220474,
+      "grad_norm": 1.1927207708358765,
+      "learning_rate": 0.00011100014770366215,
+      "loss": 2.4255,
+      "step": 7563
+    },
+    {
+      "epoch": 0.4653051181102362,
+      "grad_norm": 1.2313164472579956,
+      "learning_rate": 0.00011098092721213845,
+      "loss": 2.051,
+      "step": 7564
+    },
+    {
+      "epoch": 0.4653666338582677,
+      "grad_norm": 1.2989484071731567,
+      "learning_rate": 0.00011096170630998925,
+      "loss": 3.0229,
+      "step": 7565
+    },
+    {
+      "epoch": 0.4654281496062992,
+      "grad_norm": 1.1795076131820679,
+      "learning_rate": 0.0001109424849979334,
+      "loss": 2.3802,
+      "step": 7566
+    },
+    {
+      "epoch": 0.46548966535433073,
+      "grad_norm": 1.1056756973266602,
+      "learning_rate": 0.0001109232632766896,
+      "loss": 2.3643,
+      "step": 7567
+    },
+    {
+      "epoch": 0.4655511811023622,
+      "grad_norm": 1.433199167251587,
+      "learning_rate": 0.00011090404114697668,
+      "loss": 3.2058,
+      "step": 7568
+    },
+    {
+      "epoch": 0.4656126968503937,
+      "grad_norm": 1.1854861974716187,
+      "learning_rate": 0.00011088481860951343,
+      "loss": 2.3543,
+      "step": 7569
+    },
+    {
+      "epoch": 0.4656742125984252,
+      "grad_norm": 1.0990711450576782,
+      "learning_rate": 0.00011086559566501862,
+      "loss": 1.9925,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4657357283464567,
+      "grad_norm": 1.4431357383728027,
+      "learning_rate": 0.00011084637231421116,
+      "loss": 2.0214,
+      "step": 7571
+    },
+    {
+      "epoch": 0.4657972440944882,
+      "grad_norm": 0.9754773378372192,
+      "learning_rate": 0.00011082714855780984,
+      "loss": 2.0102,
+      "step": 7572
+    },
+    {
+      "epoch": 0.4658587598425197,
+      "grad_norm": 1.1809860467910767,
+      "learning_rate": 0.00011080792439653353,
+      "loss": 2.7313,
+      "step": 7573
+    },
+    {
+      "epoch": 0.46592027559055116,
+      "grad_norm": 1.2801042795181274,
+      "learning_rate": 0.00011078869983110113,
+      "loss": 1.8353,
+      "step": 7574
+    },
+    {
+      "epoch": 0.4659817913385827,
+      "grad_norm": 1.6567684412002563,
+      "learning_rate": 0.00011076947486223152,
+      "loss": 2.1438,
+      "step": 7575
+    },
+    {
+      "epoch": 0.4660433070866142,
+      "grad_norm": 1.8561455011367798,
+      "learning_rate": 0.00011075024949064361,
+      "loss": 2.4554,
+      "step": 7576
+    },
+    {
+      "epoch": 0.46610482283464566,
+      "grad_norm": 1.0982288122177124,
+      "learning_rate": 0.00011073102371705628,
+      "loss": 2.824,
+      "step": 7577
+    },
+    {
+      "epoch": 0.46616633858267714,
+      "grad_norm": 1.2831101417541504,
+      "learning_rate": 0.00011071179754218852,
+      "loss": 2.762,
+      "step": 7578
+    },
+    {
+      "epoch": 0.4662278543307087,
+      "grad_norm": 1.128991961479187,
+      "learning_rate": 0.00011069257096675927,
+      "loss": 3.1164,
+      "step": 7579
+    },
+    {
+      "epoch": 0.46628937007874016,
+      "grad_norm": 1.2331570386886597,
+      "learning_rate": 0.00011067334399148748,
+      "loss": 2.6267,
+      "step": 7580
+    },
+    {
+      "epoch": 0.46635088582677164,
+      "grad_norm": 1.2459808588027954,
+      "learning_rate": 0.00011065411661709212,
+      "loss": 2.4801,
+      "step": 7581
+    },
+    {
+      "epoch": 0.4664124015748031,
+      "grad_norm": 1.1219555139541626,
+      "learning_rate": 0.00011063488884429225,
+      "loss": 2.7379,
+      "step": 7582
+    },
+    {
+      "epoch": 0.46647391732283466,
+      "grad_norm": 1.2033149003982544,
+      "learning_rate": 0.0001106156606738068,
+      "loss": 2.5072,
+      "step": 7583
+    },
+    {
+      "epoch": 0.46653543307086615,
+      "grad_norm": 1.2956069707870483,
+      "learning_rate": 0.00011059643210635486,
+      "loss": 2.8217,
+      "step": 7584
+    },
+    {
+      "epoch": 0.4665969488188976,
+      "grad_norm": 1.6253209114074707,
+      "learning_rate": 0.00011057720314265546,
+      "loss": 2.6344,
+      "step": 7585
+    },
+    {
+      "epoch": 0.4666584645669291,
+      "grad_norm": 1.1412335634231567,
+      "learning_rate": 0.00011055797378342761,
+      "loss": 1.7396,
+      "step": 7586
+    },
+    {
+      "epoch": 0.46671998031496065,
+      "grad_norm": 1.6939393281936646,
+      "learning_rate": 0.00011053874402939042,
+      "loss": 2.4518,
+      "step": 7587
+    },
+    {
+      "epoch": 0.46678149606299213,
+      "grad_norm": 1.005387544631958,
+      "learning_rate": 0.00011051951388126299,
+      "loss": 2.8164,
+      "step": 7588
+    },
+    {
+      "epoch": 0.4668430118110236,
+      "grad_norm": 1.0259987115859985,
+      "learning_rate": 0.00011050028333976439,
+      "loss": 3.076,
+      "step": 7589
+    },
+    {
+      "epoch": 0.4669045275590551,
+      "grad_norm": 1.21457040309906,
+      "learning_rate": 0.0001104810524056137,
+      "loss": 2.6431,
+      "step": 7590
+    },
+    {
+      "epoch": 0.46696604330708663,
+      "grad_norm": 1.1654926538467407,
+      "learning_rate": 0.00011046182107953016,
+      "loss": 2.8253,
+      "step": 7591
+    },
+    {
+      "epoch": 0.4670275590551181,
+      "grad_norm": 1.1187483072280884,
+      "learning_rate": 0.00011044258936223282,
+      "loss": 2.2508,
+      "step": 7592
+    },
+    {
+      "epoch": 0.4670890748031496,
+      "grad_norm": 1.183179497718811,
+      "learning_rate": 0.00011042335725444086,
+      "loss": 2.6989,
+      "step": 7593
+    },
+    {
+      "epoch": 0.4671505905511811,
+      "grad_norm": 1.2641783952713013,
+      "learning_rate": 0.00011040412475687347,
+      "loss": 2.5631,
+      "step": 7594
+    },
+    {
+      "epoch": 0.4672121062992126,
+      "grad_norm": 1.3808763027191162,
+      "learning_rate": 0.00011038489187024983,
+      "loss": 2.6681,
+      "step": 7595
+    },
+    {
+      "epoch": 0.4672736220472441,
+      "grad_norm": 1.2671910524368286,
+      "learning_rate": 0.00011036565859528914,
+      "loss": 2.2121,
+      "step": 7596
+    },
+    {
+      "epoch": 0.4673351377952756,
+      "grad_norm": 1.3744688034057617,
+      "learning_rate": 0.00011034642493271062,
+      "loss": 2.2386,
+      "step": 7597
+    },
+    {
+      "epoch": 0.46739665354330706,
+      "grad_norm": 1.2370597124099731,
+      "learning_rate": 0.00011032719088323348,
+      "loss": 3.0968,
+      "step": 7598
+    },
+    {
+      "epoch": 0.4674581692913386,
+      "grad_norm": 1.2883567810058594,
+      "learning_rate": 0.00011030795644757701,
+      "loss": 2.9752,
+      "step": 7599
+    },
+    {
+      "epoch": 0.4675196850393701,
+      "grad_norm": 1.4619052410125732,
+      "learning_rate": 0.00011028872162646045,
+      "loss": 2.375,
+      "step": 7600
+    },
+    {
+      "epoch": 0.46758120078740156,
+      "grad_norm": 1.3367952108383179,
+      "learning_rate": 0.00011026948642060308,
+      "loss": 2.6623,
+      "step": 7601
+    },
+    {
+      "epoch": 0.46764271653543305,
+      "grad_norm": 1.431070327758789,
+      "learning_rate": 0.00011025025083072418,
+      "loss": 2.1689,
+      "step": 7602
+    },
+    {
+      "epoch": 0.4677042322834646,
+      "grad_norm": 1.2704875469207764,
+      "learning_rate": 0.00011023101485754303,
+      "loss": 2.6432,
+      "step": 7603
+    },
+    {
+      "epoch": 0.46776574803149606,
+      "grad_norm": 1.43257474899292,
+      "learning_rate": 0.000110211778501779,
+      "loss": 2.9479,
+      "step": 7604
+    },
+    {
+      "epoch": 0.46782726377952755,
+      "grad_norm": 1.0917401313781738,
+      "learning_rate": 0.00011019254176415137,
+      "loss": 2.9419,
+      "step": 7605
+    },
+    {
+      "epoch": 0.46788877952755903,
+      "grad_norm": 1.1007083654403687,
+      "learning_rate": 0.00011017330464537954,
+      "loss": 1.9651,
+      "step": 7606
+    },
+    {
+      "epoch": 0.46795029527559057,
+      "grad_norm": 1.2726191282272339,
+      "learning_rate": 0.00011015406714618284,
+      "loss": 2.0534,
+      "step": 7607
+    },
+    {
+      "epoch": 0.46801181102362205,
+      "grad_norm": 1.1240215301513672,
+      "learning_rate": 0.00011013482926728063,
+      "loss": 2.6732,
+      "step": 7608
+    },
+    {
+      "epoch": 0.46807332677165353,
+      "grad_norm": 1.2308576107025146,
+      "learning_rate": 0.00011011559100939233,
+      "loss": 2.1825,
+      "step": 7609
+    },
+    {
+      "epoch": 0.468134842519685,
+      "grad_norm": 1.5886073112487793,
+      "learning_rate": 0.00011009635237323734,
+      "loss": 2.2996,
+      "step": 7610
+    },
+    {
+      "epoch": 0.46819635826771655,
+      "grad_norm": 1.5383607149124146,
+      "learning_rate": 0.00011007711335953508,
+      "loss": 3.4134,
+      "step": 7611
+    },
+    {
+      "epoch": 0.46825787401574803,
+      "grad_norm": 2.012673854827881,
+      "learning_rate": 0.00011005787396900493,
+      "loss": 2.202,
+      "step": 7612
+    },
+    {
+      "epoch": 0.4683193897637795,
+      "grad_norm": 1.0802404880523682,
+      "learning_rate": 0.00011003863420236642,
+      "loss": 2.717,
+      "step": 7613
+    },
+    {
+      "epoch": 0.468380905511811,
+      "grad_norm": 0.9925294518470764,
+      "learning_rate": 0.00011001939406033894,
+      "loss": 2.8307,
+      "step": 7614
+    },
+    {
+      "epoch": 0.46844242125984253,
+      "grad_norm": 1.172334909439087,
+      "learning_rate": 0.000110000153543642,
+      "loss": 2.4963,
+      "step": 7615
+    },
+    {
+      "epoch": 0.468503937007874,
+      "grad_norm": 1.362390398979187,
+      "learning_rate": 0.00010998091265299509,
+      "loss": 2.1387,
+      "step": 7616
+    },
+    {
+      "epoch": 0.4685654527559055,
+      "grad_norm": 1.343788981437683,
+      "learning_rate": 0.00010996167138911771,
+      "loss": 2.7207,
+      "step": 7617
+    },
+    {
+      "epoch": 0.468626968503937,
+      "grad_norm": 1.1983450651168823,
+      "learning_rate": 0.00010994242975272932,
+      "loss": 2.5859,
+      "step": 7618
+    },
+    {
+      "epoch": 0.4686884842519685,
+      "grad_norm": 1.1007428169250488,
+      "learning_rate": 0.00010992318774454953,
+      "loss": 2.657,
+      "step": 7619
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 1.1975181102752686,
+      "learning_rate": 0.00010990394536529787,
+      "loss": 2.8795,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4688115157480315,
+      "grad_norm": 1.0315169095993042,
+      "learning_rate": 0.00010988470261569384,
+      "loss": 3.0279,
+      "step": 7621
+    },
+    {
+      "epoch": 0.468873031496063,
+      "grad_norm": 1.2179715633392334,
+      "learning_rate": 0.00010986545949645708,
+      "loss": 2.6354,
+      "step": 7622
+    },
+    {
+      "epoch": 0.4689345472440945,
+      "grad_norm": 1.2402660846710205,
+      "learning_rate": 0.00010984621600830714,
+      "loss": 2.3576,
+      "step": 7623
+    },
+    {
+      "epoch": 0.468996062992126,
+      "grad_norm": 1.1161717176437378,
+      "learning_rate": 0.00010982697215196363,
+      "loss": 1.9187,
+      "step": 7624
+    },
+    {
+      "epoch": 0.46905757874015747,
+      "grad_norm": 1.4431785345077515,
+      "learning_rate": 0.00010980772792814613,
+      "loss": 2.2778,
+      "step": 7625
+    },
+    {
+      "epoch": 0.469119094488189,
+      "grad_norm": 1.4417706727981567,
+      "learning_rate": 0.00010978848333757432,
+      "loss": 2.5716,
+      "step": 7626
+    },
+    {
+      "epoch": 0.4691806102362205,
+      "grad_norm": 1.154289960861206,
+      "learning_rate": 0.00010976923838096782,
+      "loss": 1.9395,
+      "step": 7627
+    },
+    {
+      "epoch": 0.46924212598425197,
+      "grad_norm": 1.358993649482727,
+      "learning_rate": 0.00010974999305904626,
+      "loss": 2.5927,
+      "step": 7628
+    },
+    {
+      "epoch": 0.46930364173228345,
+      "grad_norm": 1.3140835762023926,
+      "learning_rate": 0.00010973074737252935,
+      "loss": 2.0553,
+      "step": 7629
+    },
+    {
+      "epoch": 0.469365157480315,
+      "grad_norm": 1.1131033897399902,
+      "learning_rate": 0.00010971150132213675,
+      "loss": 2.9199,
+      "step": 7630
+    },
+    {
+      "epoch": 0.46942667322834647,
+      "grad_norm": 1.6109658479690552,
+      "learning_rate": 0.00010969225490858813,
+      "loss": 2.8629,
+      "step": 7631
+    },
+    {
+      "epoch": 0.46948818897637795,
+      "grad_norm": 1.2623709440231323,
+      "learning_rate": 0.00010967300813260321,
+      "loss": 2.77,
+      "step": 7632
+    },
+    {
+      "epoch": 0.46954970472440943,
+      "grad_norm": 1.133603811264038,
+      "learning_rate": 0.00010965376099490177,
+      "loss": 2.906,
+      "step": 7633
+    },
+    {
+      "epoch": 0.46961122047244097,
+      "grad_norm": 1.2006651163101196,
+      "learning_rate": 0.00010963451349620344,
+      "loss": 2.472,
+      "step": 7634
+    },
+    {
+      "epoch": 0.46967273622047245,
+      "grad_norm": 1.2649246454238892,
+      "learning_rate": 0.00010961526563722808,
+      "loss": 2.0208,
+      "step": 7635
+    },
+    {
+      "epoch": 0.46973425196850394,
+      "grad_norm": 1.189030408859253,
+      "learning_rate": 0.00010959601741869539,
+      "loss": 2.6479,
+      "step": 7636
+    },
+    {
+      "epoch": 0.4697957677165354,
+      "grad_norm": 1.309166431427002,
+      "learning_rate": 0.00010957676884132517,
+      "loss": 2.4411,
+      "step": 7637
+    },
+    {
+      "epoch": 0.46985728346456695,
+      "grad_norm": 1.260404348373413,
+      "learning_rate": 0.00010955751990583718,
+      "loss": 2.2806,
+      "step": 7638
+    },
+    {
+      "epoch": 0.46991879921259844,
+      "grad_norm": 1.219727635383606,
+      "learning_rate": 0.00010953827061295123,
+      "loss": 2.6065,
+      "step": 7639
+    },
+    {
+      "epoch": 0.4699803149606299,
+      "grad_norm": 1.2770975828170776,
+      "learning_rate": 0.00010951902096338715,
+      "loss": 2.5774,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4700418307086614,
+      "grad_norm": 1.232928991317749,
+      "learning_rate": 0.00010949977095786479,
+      "loss": 2.0073,
+      "step": 7641
+    },
+    {
+      "epoch": 0.47010334645669294,
+      "grad_norm": 1.3008204698562622,
+      "learning_rate": 0.00010948052059710395,
+      "loss": 2.1513,
+      "step": 7642
+    },
+    {
+      "epoch": 0.4701648622047244,
+      "grad_norm": 1.3030552864074707,
+      "learning_rate": 0.0001094612698818245,
+      "loss": 2.161,
+      "step": 7643
+    },
+    {
+      "epoch": 0.4702263779527559,
+      "grad_norm": 1.2486708164215088,
+      "learning_rate": 0.0001094420188127463,
+      "loss": 2.4511,
+      "step": 7644
+    },
+    {
+      "epoch": 0.4702878937007874,
+      "grad_norm": 1.197933316230774,
+      "learning_rate": 0.00010942276739058928,
+      "loss": 2.6862,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4703494094488189,
+      "grad_norm": 1.32341730594635,
+      "learning_rate": 0.0001094035156160733,
+      "loss": 2.298,
+      "step": 7646
+    },
+    {
+      "epoch": 0.4704109251968504,
+      "grad_norm": 1.5598783493041992,
+      "learning_rate": 0.00010938426348991823,
+      "loss": 0.2959,
+      "step": 7647
+    },
+    {
+      "epoch": 0.4704724409448819,
+      "grad_norm": 1.2541531324386597,
+      "learning_rate": 0.00010936501101284406,
+      "loss": 2.1278,
+      "step": 7648
+    },
+    {
+      "epoch": 0.47053395669291337,
+      "grad_norm": 1.1206721067428589,
+      "learning_rate": 0.00010934575818557071,
+      "loss": 2.5506,
+      "step": 7649
+    },
+    {
+      "epoch": 0.4705954724409449,
+      "grad_norm": 1.275762915611267,
+      "learning_rate": 0.00010932650500881809,
+      "loss": 2.4367,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4706569881889764,
+      "grad_norm": 1.6805590391159058,
+      "learning_rate": 0.00010930725148330623,
+      "loss": 1.9211,
+      "step": 7651
+    },
+    {
+      "epoch": 0.47071850393700787,
+      "grad_norm": 1.301647424697876,
+      "learning_rate": 0.00010928799760975503,
+      "loss": 2.5303,
+      "step": 7652
+    },
+    {
+      "epoch": 0.47078001968503935,
+      "grad_norm": 1.2697278261184692,
+      "learning_rate": 0.0001092687433888845,
+      "loss": 3.0226,
+      "step": 7653
+    },
+    {
+      "epoch": 0.4708415354330709,
+      "grad_norm": 1.113952875137329,
+      "learning_rate": 0.00010924948882141469,
+      "loss": 2.5904,
+      "step": 7654
+    },
+    {
+      "epoch": 0.4709030511811024,
+      "grad_norm": 1.2876708507537842,
+      "learning_rate": 0.00010923023390806555,
+      "loss": 2.7133,
+      "step": 7655
+    },
+    {
+      "epoch": 0.47096456692913385,
+      "grad_norm": 1.0370928049087524,
+      "learning_rate": 0.00010921097864955716,
+      "loss": 2.5951,
+      "step": 7656
+    },
+    {
+      "epoch": 0.47102608267716534,
+      "grad_norm": 1.0678495168685913,
+      "learning_rate": 0.00010919172304660948,
+      "loss": 2.9118,
+      "step": 7657
+    },
+    {
+      "epoch": 0.4710875984251969,
+      "grad_norm": 1.341937780380249,
+      "learning_rate": 0.00010917246709994265,
+      "loss": 2.1618,
+      "step": 7658
+    },
+    {
+      "epoch": 0.47114911417322836,
+      "grad_norm": 1.0701593160629272,
+      "learning_rate": 0.0001091532108102767,
+      "loss": 2.5738,
+      "step": 7659
+    },
+    {
+      "epoch": 0.47121062992125984,
+      "grad_norm": 1.2086325883865356,
+      "learning_rate": 0.00010913395417833167,
+      "loss": 3.0803,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4712721456692913,
+      "grad_norm": 1.1487727165222168,
+      "learning_rate": 0.00010911469720482772,
+      "loss": 2.9427,
+      "step": 7661
+    },
+    {
+      "epoch": 0.47133366141732286,
+      "grad_norm": 1.281419277191162,
+      "learning_rate": 0.00010909543989048489,
+      "loss": 2.3995,
+      "step": 7662
+    },
+    {
+      "epoch": 0.47139517716535434,
+      "grad_norm": 1.290838360786438,
+      "learning_rate": 0.00010907618223602333,
+      "loss": 2.1623,
+      "step": 7663
+    },
+    {
+      "epoch": 0.4714566929133858,
+      "grad_norm": 1.0343679189682007,
+      "learning_rate": 0.00010905692424216321,
+      "loss": 1.9175,
+      "step": 7664
+    },
+    {
+      "epoch": 0.4715182086614173,
+      "grad_norm": 1.1979039907455444,
+      "learning_rate": 0.00010903766590962458,
+      "loss": 2.3026,
+      "step": 7665
+    },
+    {
+      "epoch": 0.47157972440944884,
+      "grad_norm": 1.0911281108856201,
+      "learning_rate": 0.00010901840723912764,
+      "loss": 2.481,
+      "step": 7666
+    },
+    {
+      "epoch": 0.4716412401574803,
+      "grad_norm": 1.2137645483016968,
+      "learning_rate": 0.00010899914823139259,
+      "loss": 2.1316,
+      "step": 7667
+    },
+    {
+      "epoch": 0.4717027559055118,
+      "grad_norm": 1.5783424377441406,
+      "learning_rate": 0.00010897988888713957,
+      "loss": 2.9379,
+      "step": 7668
+    },
+    {
+      "epoch": 0.4717642716535433,
+      "grad_norm": 1.2489967346191406,
+      "learning_rate": 0.00010896062920708876,
+      "loss": 3.1237,
+      "step": 7669
+    },
+    {
+      "epoch": 0.4718257874015748,
+      "grad_norm": 1.2888888120651245,
+      "learning_rate": 0.00010894136919196038,
+      "loss": 2.7124,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4718873031496063,
+      "grad_norm": 1.1065336465835571,
+      "learning_rate": 0.00010892210884247466,
+      "loss": 2.9666,
+      "step": 7671
+    },
+    {
+      "epoch": 0.4719488188976378,
+      "grad_norm": 1.1880484819412231,
+      "learning_rate": 0.00010890284815935183,
+      "loss": 2.1947,
+      "step": 7672
+    },
+    {
+      "epoch": 0.47201033464566927,
+      "grad_norm": 1.138203740119934,
+      "learning_rate": 0.0001088835871433121,
+      "loss": 2.104,
+      "step": 7673
+    },
+    {
+      "epoch": 0.4720718503937008,
+      "grad_norm": 1.2177066802978516,
+      "learning_rate": 0.00010886432579507579,
+      "loss": 2.3353,
+      "step": 7674
+    },
+    {
+      "epoch": 0.4721333661417323,
+      "grad_norm": 1.06807279586792,
+      "learning_rate": 0.00010884506411536308,
+      "loss": 2.3739,
+      "step": 7675
+    },
+    {
+      "epoch": 0.4721948818897638,
+      "grad_norm": 1.1452404260635376,
+      "learning_rate": 0.00010882580210489432,
+      "loss": 2.8543,
+      "step": 7676
+    },
+    {
+      "epoch": 0.47225639763779526,
+      "grad_norm": 1.2072272300720215,
+      "learning_rate": 0.00010880653976438977,
+      "loss": 2.2458,
+      "step": 7677
+    },
+    {
+      "epoch": 0.4723179133858268,
+      "grad_norm": 1.1615134477615356,
+      "learning_rate": 0.00010878727709456976,
+      "loss": 2.3375,
+      "step": 7678
+    },
+    {
+      "epoch": 0.4723794291338583,
+      "grad_norm": 1.0521788597106934,
+      "learning_rate": 0.00010876801409615454,
+      "loss": 2.5416,
+      "step": 7679
+    },
+    {
+      "epoch": 0.47244094488188976,
+      "grad_norm": 1.1450419425964355,
+      "learning_rate": 0.0001087487507698645,
+      "loss": 2.7379,
+      "step": 7680
+    },
+    {
+      "epoch": 0.47250246062992124,
+      "grad_norm": 1.23845636844635,
+      "learning_rate": 0.00010872948711641998,
+      "loss": 2.0611,
+      "step": 7681
+    },
+    {
+      "epoch": 0.4725639763779528,
+      "grad_norm": 1.3284415006637573,
+      "learning_rate": 0.00010871022313654129,
+      "loss": 1.9804,
+      "step": 7682
+    },
+    {
+      "epoch": 0.47262549212598426,
+      "grad_norm": 1.372694730758667,
+      "learning_rate": 0.00010869095883094885,
+      "loss": 1.9316,
+      "step": 7683
+    },
+    {
+      "epoch": 0.47268700787401574,
+      "grad_norm": 1.0894865989685059,
+      "learning_rate": 0.00010867169420036301,
+      "loss": 2.7848,
+      "step": 7684
+    },
+    {
+      "epoch": 0.4727485236220472,
+      "grad_norm": 1.2265185117721558,
+      "learning_rate": 0.00010865242924550413,
+      "loss": 2.508,
+      "step": 7685
+    },
+    {
+      "epoch": 0.47281003937007876,
+      "grad_norm": 1.4949102401733398,
+      "learning_rate": 0.00010863316396709269,
+      "loss": 2.324,
+      "step": 7686
+    },
+    {
+      "epoch": 0.47287155511811024,
+      "grad_norm": 1.3146452903747559,
+      "learning_rate": 0.00010861389836584904,
+      "loss": 1.9891,
+      "step": 7687
+    },
+    {
+      "epoch": 0.4729330708661417,
+      "grad_norm": 1.0558637380599976,
+      "learning_rate": 0.00010859463244249361,
+      "loss": 2.8929,
+      "step": 7688
+    },
+    {
+      "epoch": 0.4729945866141732,
+      "grad_norm": 1.252812147140503,
+      "learning_rate": 0.00010857536619774686,
+      "loss": 2.2231,
+      "step": 7689
+    },
+    {
+      "epoch": 0.47305610236220474,
+      "grad_norm": 1.0319552421569824,
+      "learning_rate": 0.00010855609963232925,
+      "loss": 2.2341,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4731176181102362,
+      "grad_norm": 1.1616640090942383,
+      "learning_rate": 0.0001085368327469612,
+      "loss": 2.3478,
+      "step": 7691
+    },
+    {
+      "epoch": 0.4731791338582677,
+      "grad_norm": 1.1990280151367188,
+      "learning_rate": 0.0001085175655423632,
+      "loss": 3.1619,
+      "step": 7692
+    },
+    {
+      "epoch": 0.4732406496062992,
+      "grad_norm": 1.5587260723114014,
+      "learning_rate": 0.00010849829801925577,
+      "loss": 2.4647,
+      "step": 7693
+    },
+    {
+      "epoch": 0.47330216535433073,
+      "grad_norm": 1.2629384994506836,
+      "learning_rate": 0.00010847903017835938,
+      "loss": 2.6463,
+      "step": 7694
+    },
+    {
+      "epoch": 0.4733636811023622,
+      "grad_norm": 1.2137272357940674,
+      "learning_rate": 0.00010845976202039451,
+      "loss": 2.5796,
+      "step": 7695
+    },
+    {
+      "epoch": 0.4734251968503937,
+      "grad_norm": 1.5153030157089233,
+      "learning_rate": 0.00010844049354608177,
+      "loss": 2.5749,
+      "step": 7696
+    },
+    {
+      "epoch": 0.4734867125984252,
+      "grad_norm": 1.1915168762207031,
+      "learning_rate": 0.0001084212247561416,
+      "loss": 2.6859,
+      "step": 7697
+    },
+    {
+      "epoch": 0.4735482283464567,
+      "grad_norm": 1.2707632780075073,
+      "learning_rate": 0.00010840195565129456,
+      "loss": 3.0267,
+      "step": 7698
+    },
+    {
+      "epoch": 0.4736097440944882,
+      "grad_norm": 1.1652952432632446,
+      "learning_rate": 0.00010838268623226126,
+      "loss": 2.9844,
+      "step": 7699
+    },
+    {
+      "epoch": 0.4736712598425197,
+      "grad_norm": 1.1826751232147217,
+      "learning_rate": 0.00010836341649976224,
+      "loss": 2.5927,
+      "step": 7700
+    },
+    {
+      "epoch": 0.47373277559055116,
+      "grad_norm": 1.4061943292617798,
+      "learning_rate": 0.00010834414645451807,
+      "loss": 2.4584,
+      "step": 7701
+    },
+    {
+      "epoch": 0.4737942913385827,
+      "grad_norm": 2.0851166248321533,
+      "learning_rate": 0.00010832487609724935,
+      "loss": 2.544,
+      "step": 7702
+    },
+    {
+      "epoch": 0.4738558070866142,
+      "grad_norm": 1.4661732912063599,
+      "learning_rate": 0.00010830560542867669,
+      "loss": 2.0434,
+      "step": 7703
+    },
+    {
+      "epoch": 0.47391732283464566,
+      "grad_norm": 1.2355715036392212,
+      "learning_rate": 0.00010828633444952072,
+      "loss": 2.6153,
+      "step": 7704
+    },
+    {
+      "epoch": 0.47397883858267714,
+      "grad_norm": 1.0726184844970703,
+      "learning_rate": 0.00010826706316050201,
+      "loss": 2.8131,
+      "step": 7705
+    },
+    {
+      "epoch": 0.4740403543307087,
+      "grad_norm": 1.404786467552185,
+      "learning_rate": 0.00010824779156234126,
+      "loss": 2.7954,
+      "step": 7706
+    },
+    {
+      "epoch": 0.47410187007874016,
+      "grad_norm": 1.0617549419403076,
+      "learning_rate": 0.00010822851965575909,
+      "loss": 2.4699,
+      "step": 7707
+    },
+    {
+      "epoch": 0.47416338582677164,
+      "grad_norm": 1.2709327936172485,
+      "learning_rate": 0.00010820924744147616,
+      "loss": 3.0608,
+      "step": 7708
+    },
+    {
+      "epoch": 0.4742249015748031,
+      "grad_norm": 1.8422532081604004,
+      "learning_rate": 0.00010818997492021314,
+      "loss": 2.1662,
+      "step": 7709
+    },
+    {
+      "epoch": 0.47428641732283466,
+      "grad_norm": 2.7369837760925293,
+      "learning_rate": 0.00010817070209269075,
+      "loss": 2.5573,
+      "step": 7710
+    },
+    {
+      "epoch": 0.47434793307086615,
+      "grad_norm": 1.2511111497879028,
+      "learning_rate": 0.00010815142895962969,
+      "loss": 2.559,
+      "step": 7711
+    },
+    {
+      "epoch": 0.4744094488188976,
+      "grad_norm": 1.155408263206482,
+      "learning_rate": 0.0001081321555217506,
+      "loss": 3.1843,
+      "step": 7712
+    },
+    {
+      "epoch": 0.4744709645669291,
+      "grad_norm": 1.2066247463226318,
+      "learning_rate": 0.00010811288177977425,
+      "loss": 2.471,
+      "step": 7713
+    },
+    {
+      "epoch": 0.47453248031496065,
+      "grad_norm": 1.404071569442749,
+      "learning_rate": 0.00010809360773442134,
+      "loss": 1.9486,
+      "step": 7714
+    },
+    {
+      "epoch": 0.47459399606299213,
+      "grad_norm": 1.1509944200515747,
+      "learning_rate": 0.00010807433338641266,
+      "loss": 2.4027,
+      "step": 7715
+    },
+    {
+      "epoch": 0.4746555118110236,
+      "grad_norm": 1.1945728063583374,
+      "learning_rate": 0.00010805505873646893,
+      "loss": 2.5116,
+      "step": 7716
+    },
+    {
+      "epoch": 0.4747170275590551,
+      "grad_norm": 1.190449833869934,
+      "learning_rate": 0.0001080357837853109,
+      "loss": 2.7133,
+      "step": 7717
+    },
+    {
+      "epoch": 0.47477854330708663,
+      "grad_norm": 1.0470480918884277,
+      "learning_rate": 0.00010801650853365939,
+      "loss": 2.7517,
+      "step": 7718
+    },
+    {
+      "epoch": 0.4748400590551181,
+      "grad_norm": 1.4977415800094604,
+      "learning_rate": 0.00010799723298223516,
+      "loss": 1.8844,
+      "step": 7719
+    },
+    {
+      "epoch": 0.4749015748031496,
+      "grad_norm": 1.1114091873168945,
+      "learning_rate": 0.00010797795713175898,
+      "loss": 2.4731,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4749630905511811,
+      "grad_norm": 1.0691838264465332,
+      "learning_rate": 0.00010795868098295172,
+      "loss": 2.9066,
+      "step": 7721
+    },
+    {
+      "epoch": 0.4750246062992126,
+      "grad_norm": 1.2894264459609985,
+      "learning_rate": 0.00010793940453653418,
+      "loss": 2.3421,
+      "step": 7722
+    },
+    {
+      "epoch": 0.4750861220472441,
+      "grad_norm": 1.3217569589614868,
+      "learning_rate": 0.00010792012779322713,
+      "loss": 2.4966,
+      "step": 7723
+    },
+    {
+      "epoch": 0.4751476377952756,
+      "grad_norm": 1.03250253200531,
+      "learning_rate": 0.00010790085075375151,
+      "loss": 3.0216,
+      "step": 7724
+    },
+    {
+      "epoch": 0.47520915354330706,
+      "grad_norm": 1.1231858730316162,
+      "learning_rate": 0.00010788157341882812,
+      "loss": 2.6474,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4752706692913386,
+      "grad_norm": 1.2640900611877441,
+      "learning_rate": 0.00010786229578917784,
+      "loss": 2.7678,
+      "step": 7726
+    },
+    {
+      "epoch": 0.4753321850393701,
+      "grad_norm": 1.1563228368759155,
+      "learning_rate": 0.00010784301786552151,
+      "loss": 2.8013,
+      "step": 7727
+    },
+    {
+      "epoch": 0.47539370078740156,
+      "grad_norm": 1.3315871953964233,
+      "learning_rate": 0.00010782373964858008,
+      "loss": 1.9903,
+      "step": 7728
+    },
+    {
+      "epoch": 0.47545521653543305,
+      "grad_norm": 1.0304712057113647,
+      "learning_rate": 0.00010780446113907438,
+      "loss": 2.6207,
+      "step": 7729
+    },
+    {
+      "epoch": 0.4755167322834646,
+      "grad_norm": 1.1773074865341187,
+      "learning_rate": 0.00010778518233772536,
+      "loss": 2.6894,
+      "step": 7730
+    },
+    {
+      "epoch": 0.47557824803149606,
+      "grad_norm": 1.0944392681121826,
+      "learning_rate": 0.00010776590324525395,
+      "loss": 2.5643,
+      "step": 7731
+    },
+    {
+      "epoch": 0.47563976377952755,
+      "grad_norm": 1.1008286476135254,
+      "learning_rate": 0.00010774662386238106,
+      "loss": 2.687,
+      "step": 7732
+    },
+    {
+      "epoch": 0.47570127952755903,
+      "grad_norm": 1.0779165029525757,
+      "learning_rate": 0.00010772734418982762,
+      "loss": 2.734,
+      "step": 7733
+    },
+    {
+      "epoch": 0.47576279527559057,
+      "grad_norm": 1.3196789026260376,
+      "learning_rate": 0.00010770806422831461,
+      "loss": 2.4729,
+      "step": 7734
+    },
+    {
+      "epoch": 0.47582431102362205,
+      "grad_norm": 1.1814632415771484,
+      "learning_rate": 0.00010768878397856297,
+      "loss": 2.7106,
+      "step": 7735
+    },
+    {
+      "epoch": 0.47588582677165353,
+      "grad_norm": 1.1796928644180298,
+      "learning_rate": 0.00010766950344129367,
+      "loss": 2.4892,
+      "step": 7736
+    },
+    {
+      "epoch": 0.475947342519685,
+      "grad_norm": 1.3555541038513184,
+      "learning_rate": 0.00010765022261722773,
+      "loss": 2.7007,
+      "step": 7737
+    },
+    {
+      "epoch": 0.47600885826771655,
+      "grad_norm": 1.2286978960037231,
+      "learning_rate": 0.00010763094150708612,
+      "loss": 2.1637,
+      "step": 7738
+    },
+    {
+      "epoch": 0.47607037401574803,
+      "grad_norm": 1.149582028388977,
+      "learning_rate": 0.00010761166011158986,
+      "loss": 2.4218,
+      "step": 7739
+    },
+    {
+      "epoch": 0.4761318897637795,
+      "grad_norm": 1.2856695652008057,
+      "learning_rate": 0.00010759237843145992,
+      "loss": 2.5818,
+      "step": 7740
+    },
+    {
+      "epoch": 0.476193405511811,
+      "grad_norm": 1.0762094259262085,
+      "learning_rate": 0.0001075730964674174,
+      "loss": 2.1573,
+      "step": 7741
+    },
+    {
+      "epoch": 0.47625492125984253,
+      "grad_norm": 1.0774520635604858,
+      "learning_rate": 0.00010755381422018329,
+      "loss": 2.2736,
+      "step": 7742
+    },
+    {
+      "epoch": 0.476316437007874,
+      "grad_norm": 1.2879606485366821,
+      "learning_rate": 0.00010753453169047862,
+      "loss": 2.6429,
+      "step": 7743
+    },
+    {
+      "epoch": 0.4763779527559055,
+      "grad_norm": 1.036560297012329,
+      "learning_rate": 0.00010751524887902452,
+      "loss": 1.9997,
+      "step": 7744
+    },
+    {
+      "epoch": 0.476439468503937,
+      "grad_norm": 1.0710084438323975,
+      "learning_rate": 0.000107495965786542,
+      "loss": 2.5302,
+      "step": 7745
+    },
+    {
+      "epoch": 0.4765009842519685,
+      "grad_norm": 1.1097811460494995,
+      "learning_rate": 0.00010747668241375215,
+      "loss": 2.9362,
+      "step": 7746
+    },
+    {
+      "epoch": 0.4765625,
+      "grad_norm": 1.3453803062438965,
+      "learning_rate": 0.00010745739876137608,
+      "loss": 2.803,
+      "step": 7747
+    },
+    {
+      "epoch": 0.4766240157480315,
+      "grad_norm": 1.0106500387191772,
+      "learning_rate": 0.0001074381148301349,
+      "loss": 2.688,
+      "step": 7748
+    },
+    {
+      "epoch": 0.476685531496063,
+      "grad_norm": 1.1535696983337402,
+      "learning_rate": 0.00010741883062074967,
+      "loss": 2.7452,
+      "step": 7749
+    },
+    {
+      "epoch": 0.4767470472440945,
+      "grad_norm": 1.1435014009475708,
+      "learning_rate": 0.00010739954613394157,
+      "loss": 3.1453,
+      "step": 7750
+    },
+    {
+      "epoch": 0.476808562992126,
+      "grad_norm": 1.1711208820343018,
+      "learning_rate": 0.00010738026137043166,
+      "loss": 2.0168,
+      "step": 7751
+    },
+    {
+      "epoch": 0.47687007874015747,
+      "grad_norm": 1.143204689025879,
+      "learning_rate": 0.00010736097633094116,
+      "loss": 3.167,
+      "step": 7752
+    },
+    {
+      "epoch": 0.476931594488189,
+      "grad_norm": 1.435957908630371,
+      "learning_rate": 0.00010734169101619118,
+      "loss": 2.7066,
+      "step": 7753
+    },
+    {
+      "epoch": 0.4769931102362205,
+      "grad_norm": 1.2399896383285522,
+      "learning_rate": 0.00010732240542690293,
+      "loss": 2.7077,
+      "step": 7754
+    },
+    {
+      "epoch": 0.47705462598425197,
+      "grad_norm": 1.3039630651474,
+      "learning_rate": 0.0001073031195637975,
+      "loss": 2.7414,
+      "step": 7755
+    },
+    {
+      "epoch": 0.47711614173228345,
+      "grad_norm": 1.5263702869415283,
+      "learning_rate": 0.00010728383342759614,
+      "loss": 2.0472,
+      "step": 7756
+    },
+    {
+      "epoch": 0.477177657480315,
+      "grad_norm": 1.0277162790298462,
+      "learning_rate": 0.00010726454701902003,
+      "loss": 2.4234,
+      "step": 7757
+    },
+    {
+      "epoch": 0.47723917322834647,
+      "grad_norm": 1.1086006164550781,
+      "learning_rate": 0.00010724526033879034,
+      "loss": 2.8529,
+      "step": 7758
+    },
+    {
+      "epoch": 0.47730068897637795,
+      "grad_norm": 1.1918141841888428,
+      "learning_rate": 0.00010722597338762837,
+      "loss": 2.7627,
+      "step": 7759
+    },
+    {
+      "epoch": 0.47736220472440943,
+      "grad_norm": 1.0762476921081543,
+      "learning_rate": 0.00010720668616625526,
+      "loss": 2.7753,
+      "step": 7760
+    },
+    {
+      "epoch": 0.47742372047244097,
+      "grad_norm": 1.207157850265503,
+      "learning_rate": 0.00010718739867539228,
+      "loss": 2.3392,
+      "step": 7761
+    },
+    {
+      "epoch": 0.47748523622047245,
+      "grad_norm": 1.3099734783172607,
+      "learning_rate": 0.00010716811091576064,
+      "loss": 2.9782,
+      "step": 7762
+    },
+    {
+      "epoch": 0.47754675196850394,
+      "grad_norm": 1.1279125213623047,
+      "learning_rate": 0.00010714882288808166,
+      "loss": 2.7324,
+      "step": 7763
+    },
+    {
+      "epoch": 0.4776082677165354,
+      "grad_norm": 1.1576510667800903,
+      "learning_rate": 0.00010712953459307655,
+      "loss": 2.4457,
+      "step": 7764
+    },
+    {
+      "epoch": 0.47766978346456695,
+      "grad_norm": 1.2729549407958984,
+      "learning_rate": 0.00010711024603146657,
+      "loss": 2.481,
+      "step": 7765
+    },
+    {
+      "epoch": 0.47773129921259844,
+      "grad_norm": 1.26987624168396,
+      "learning_rate": 0.00010709095720397309,
+      "loss": 2.1437,
+      "step": 7766
+    },
+    {
+      "epoch": 0.4777928149606299,
+      "grad_norm": 1.2099502086639404,
+      "learning_rate": 0.00010707166811131732,
+      "loss": 2.6638,
+      "step": 7767
+    },
+    {
+      "epoch": 0.4778543307086614,
+      "grad_norm": 1.2385361194610596,
+      "learning_rate": 0.00010705237875422058,
+      "loss": 1.8897,
+      "step": 7768
+    },
+    {
+      "epoch": 0.47791584645669294,
+      "grad_norm": 1.4185224771499634,
+      "learning_rate": 0.00010703308913340425,
+      "loss": 2.7462,
+      "step": 7769
+    },
+    {
+      "epoch": 0.4779773622047244,
+      "grad_norm": 1.0919464826583862,
+      "learning_rate": 0.00010701379924958957,
+      "loss": 2.7715,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4780388779527559,
+      "grad_norm": 1.2857152223587036,
+      "learning_rate": 0.00010699450910349791,
+      "loss": 2.6204,
+      "step": 7771
+    },
+    {
+      "epoch": 0.4781003937007874,
+      "grad_norm": 1.0661914348602295,
+      "learning_rate": 0.00010697521869585061,
+      "loss": 3.0345,
+      "step": 7772
+    },
+    {
+      "epoch": 0.4781619094488189,
+      "grad_norm": 1.349022388458252,
+      "learning_rate": 0.00010695592802736904,
+      "loss": 2.5863,
+      "step": 7773
+    },
+    {
+      "epoch": 0.4782234251968504,
+      "grad_norm": 1.13512122631073,
+      "learning_rate": 0.00010693663709877455,
+      "loss": 2.6408,
+      "step": 7774
+    },
+    {
+      "epoch": 0.4782849409448819,
+      "grad_norm": 1.1259815692901611,
+      "learning_rate": 0.00010691734591078848,
+      "loss": 2.7182,
+      "step": 7775
+    },
+    {
+      "epoch": 0.47834645669291337,
+      "grad_norm": 1.3265893459320068,
+      "learning_rate": 0.00010689805446413228,
+      "loss": 2.4741,
+      "step": 7776
+    },
+    {
+      "epoch": 0.4784079724409449,
+      "grad_norm": 1.130415439605713,
+      "learning_rate": 0.00010687876275952731,
+      "loss": 2.4132,
+      "step": 7777
+    },
+    {
+      "epoch": 0.4784694881889764,
+      "grad_norm": 1.333799123764038,
+      "learning_rate": 0.00010685947079769494,
+      "loss": 2.5455,
+      "step": 7778
+    },
+    {
+      "epoch": 0.47853100393700787,
+      "grad_norm": 1.2326173782348633,
+      "learning_rate": 0.00010684017857935662,
+      "loss": 3.1428,
+      "step": 7779
+    },
+    {
+      "epoch": 0.47859251968503935,
+      "grad_norm": 1.501041293144226,
+      "learning_rate": 0.00010682088610523378,
+      "loss": 2.4377,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4786540354330709,
+      "grad_norm": 1.2397665977478027,
+      "learning_rate": 0.00010680159337604781,
+      "loss": 3.008,
+      "step": 7781
+    },
+    {
+      "epoch": 0.4787155511811024,
+      "grad_norm": 1.2004473209381104,
+      "learning_rate": 0.00010678230039252019,
+      "loss": 2.9026,
+      "step": 7782
+    },
+    {
+      "epoch": 0.47877706692913385,
+      "grad_norm": 1.1404041051864624,
+      "learning_rate": 0.00010676300715537237,
+      "loss": 2.9682,
+      "step": 7783
+    },
+    {
+      "epoch": 0.47883858267716534,
+      "grad_norm": 1.2666747570037842,
+      "learning_rate": 0.00010674371366532579,
+      "loss": 2.5005,
+      "step": 7784
+    },
+    {
+      "epoch": 0.4789000984251969,
+      "grad_norm": 1.1277352571487427,
+      "learning_rate": 0.00010672441992310191,
+      "loss": 2.4258,
+      "step": 7785
+    },
+    {
+      "epoch": 0.47896161417322836,
+      "grad_norm": 1.2043049335479736,
+      "learning_rate": 0.00010670512592942223,
+      "loss": 3.2972,
+      "step": 7786
+    },
+    {
+      "epoch": 0.47902312992125984,
+      "grad_norm": 1.4163585901260376,
+      "learning_rate": 0.00010668583168500822,
+      "loss": 2.388,
+      "step": 7787
+    },
+    {
+      "epoch": 0.4790846456692913,
+      "grad_norm": 1.3338021039962769,
+      "learning_rate": 0.00010666653719058138,
+      "loss": 2.1071,
+      "step": 7788
+    },
+    {
+      "epoch": 0.47914616141732286,
+      "grad_norm": 1.4129936695098877,
+      "learning_rate": 0.00010664724244686324,
+      "loss": 2.2093,
+      "step": 7789
+    },
+    {
+      "epoch": 0.47920767716535434,
+      "grad_norm": 1.0139185190200806,
+      "learning_rate": 0.00010662794745457527,
+      "loss": 3.1339,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4792691929133858,
+      "grad_norm": 1.390896201133728,
+      "learning_rate": 0.00010660865221443905,
+      "loss": 2.5806,
+      "step": 7791
+    },
+    {
+      "epoch": 0.4793307086614173,
+      "grad_norm": 1.299834966659546,
+      "learning_rate": 0.00010658935672717607,
+      "loss": 2.5045,
+      "step": 7792
+    },
+    {
+      "epoch": 0.47939222440944884,
+      "grad_norm": 1.26209557056427,
+      "learning_rate": 0.00010657006099350792,
+      "loss": 2.5802,
+      "step": 7793
+    },
+    {
+      "epoch": 0.4794537401574803,
+      "grad_norm": 1.5077794790267944,
+      "learning_rate": 0.00010655076501415608,
+      "loss": 2.3455,
+      "step": 7794
+    },
+    {
+      "epoch": 0.4795152559055118,
+      "grad_norm": 1.2323601245880127,
+      "learning_rate": 0.00010653146878984219,
+      "loss": 2.2549,
+      "step": 7795
+    },
+    {
+      "epoch": 0.4795767716535433,
+      "grad_norm": 1.38791024684906,
+      "learning_rate": 0.00010651217232128779,
+      "loss": 2.3088,
+      "step": 7796
+    },
+    {
+      "epoch": 0.4796382874015748,
+      "grad_norm": 1.2347619533538818,
+      "learning_rate": 0.00010649287560921444,
+      "loss": 2.7779,
+      "step": 7797
+    },
+    {
+      "epoch": 0.4796998031496063,
+      "grad_norm": 1.3495279550552368,
+      "learning_rate": 0.00010647357865434376,
+      "loss": 2.7577,
+      "step": 7798
+    },
+    {
+      "epoch": 0.4797613188976378,
+      "grad_norm": 1.2226566076278687,
+      "learning_rate": 0.00010645428145739733,
+      "loss": 2.8934,
+      "step": 7799
+    },
+    {
+      "epoch": 0.47982283464566927,
+      "grad_norm": 1.1608039140701294,
+      "learning_rate": 0.00010643498401909675,
+      "loss": 2.8184,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4798843503937008,
+      "grad_norm": 1.1849979162216187,
+      "learning_rate": 0.00010641568634016367,
+      "loss": 2.6485,
+      "step": 7801
+    },
+    {
+      "epoch": 0.4799458661417323,
+      "grad_norm": 1.0929222106933594,
+      "learning_rate": 0.00010639638842131971,
+      "loss": 3.063,
+      "step": 7802
+    },
+    {
+      "epoch": 0.4800073818897638,
+      "grad_norm": 1.0996869802474976,
+      "learning_rate": 0.00010637709026328644,
+      "loss": 2.774,
+      "step": 7803
+    },
+    {
+      "epoch": 0.48006889763779526,
+      "grad_norm": 1.2681695222854614,
+      "learning_rate": 0.0001063577918667856,
+      "loss": 2.8833,
+      "step": 7804
+    },
+    {
+      "epoch": 0.4801304133858268,
+      "grad_norm": 1.2306712865829468,
+      "learning_rate": 0.0001063384932325388,
+      "loss": 2.8238,
+      "step": 7805
+    },
+    {
+      "epoch": 0.4801919291338583,
+      "grad_norm": 1.1357609033584595,
+      "learning_rate": 0.00010631919436126768,
+      "loss": 2.5615,
+      "step": 7806
+    },
+    {
+      "epoch": 0.48025344488188976,
+      "grad_norm": 1.347277283668518,
+      "learning_rate": 0.0001062998952536939,
+      "loss": 2.0788,
+      "step": 7807
+    },
+    {
+      "epoch": 0.48031496062992124,
+      "grad_norm": 1.5402146577835083,
+      "learning_rate": 0.00010628059591053921,
+      "loss": 2.3947,
+      "step": 7808
+    },
+    {
+      "epoch": 0.4803764763779528,
+      "grad_norm": 1.2318816184997559,
+      "learning_rate": 0.00010626129633252524,
+      "loss": 2.5927,
+      "step": 7809
+    },
+    {
+      "epoch": 0.48043799212598426,
+      "grad_norm": 1.0370197296142578,
+      "learning_rate": 0.00010624199652037368,
+      "loss": 2.6892,
+      "step": 7810
+    },
+    {
+      "epoch": 0.48049950787401574,
+      "grad_norm": 1.3109487295150757,
+      "learning_rate": 0.0001062226964748063,
+      "loss": 2.6086,
+      "step": 7811
+    },
+    {
+      "epoch": 0.4805610236220472,
+      "grad_norm": 1.3248285055160522,
+      "learning_rate": 0.00010620339619654474,
+      "loss": 2.246,
+      "step": 7812
+    },
+    {
+      "epoch": 0.48062253937007876,
+      "grad_norm": 1.3771083354949951,
+      "learning_rate": 0.00010618409568631073,
+      "loss": 3.103,
+      "step": 7813
+    },
+    {
+      "epoch": 0.48068405511811024,
+      "grad_norm": 1.237891674041748,
+      "learning_rate": 0.00010616479494482608,
+      "loss": 2.586,
+      "step": 7814
+    },
+    {
+      "epoch": 0.4807455708661417,
+      "grad_norm": 1.0313037633895874,
+      "learning_rate": 0.00010614549397281244,
+      "loss": 3.3108,
+      "step": 7815
+    },
+    {
+      "epoch": 0.4808070866141732,
+      "grad_norm": 1.1646825075149536,
+      "learning_rate": 0.00010612619277099159,
+      "loss": 2.2122,
+      "step": 7816
+    },
+    {
+      "epoch": 0.48086860236220474,
+      "grad_norm": 1.2296122312545776,
+      "learning_rate": 0.00010610689134008531,
+      "loss": 3.0839,
+      "step": 7817
+    },
+    {
+      "epoch": 0.4809301181102362,
+      "grad_norm": 0.9500906467437744,
+      "learning_rate": 0.00010608758968081536,
+      "loss": 2.5798,
+      "step": 7818
+    },
+    {
+      "epoch": 0.4809916338582677,
+      "grad_norm": 1.2065331935882568,
+      "learning_rate": 0.00010606828779390347,
+      "loss": 2.3291,
+      "step": 7819
+    },
+    {
+      "epoch": 0.4810531496062992,
+      "grad_norm": 1.2885408401489258,
+      "learning_rate": 0.00010604898568007147,
+      "loss": 2.1861,
+      "step": 7820
+    },
+    {
+      "epoch": 0.48111466535433073,
+      "grad_norm": 1.309748888015747,
+      "learning_rate": 0.00010602968334004115,
+      "loss": 2.7518,
+      "step": 7821
+    },
+    {
+      "epoch": 0.4811761811023622,
+      "grad_norm": 1.312630295753479,
+      "learning_rate": 0.00010601038077453427,
+      "loss": 2.4111,
+      "step": 7822
+    },
+    {
+      "epoch": 0.4812376968503937,
+      "grad_norm": 1.207488775253296,
+      "learning_rate": 0.00010599107798427265,
+      "loss": 2.7608,
+      "step": 7823
+    },
+    {
+      "epoch": 0.4812992125984252,
+      "grad_norm": 1.0699747800827026,
+      "learning_rate": 0.00010597177496997817,
+      "loss": 2.8212,
+      "step": 7824
+    },
+    {
+      "epoch": 0.4813607283464567,
+      "grad_norm": 1.260619878768921,
+      "learning_rate": 0.00010595247173237254,
+      "loss": 2.7426,
+      "step": 7825
+    },
+    {
+      "epoch": 0.4814222440944882,
+      "grad_norm": 1.1799503564834595,
+      "learning_rate": 0.00010593316827217772,
+      "loss": 2.8349,
+      "step": 7826
+    },
+    {
+      "epoch": 0.4814837598425197,
+      "grad_norm": 1.1671499013900757,
+      "learning_rate": 0.00010591386459011547,
+      "loss": 2.2009,
+      "step": 7827
+    },
+    {
+      "epoch": 0.48154527559055116,
+      "grad_norm": 1.1980592012405396,
+      "learning_rate": 0.00010589456068690767,
+      "loss": 2.5402,
+      "step": 7828
+    },
+    {
+      "epoch": 0.4816067913385827,
+      "grad_norm": 1.3584468364715576,
+      "learning_rate": 0.00010587525656327614,
+      "loss": 2.4827,
+      "step": 7829
+    },
+    {
+      "epoch": 0.4816683070866142,
+      "grad_norm": 1.2168675661087036,
+      "learning_rate": 0.0001058559522199428,
+      "loss": 2.3934,
+      "step": 7830
+    },
+    {
+      "epoch": 0.48172982283464566,
+      "grad_norm": 1.321364402770996,
+      "learning_rate": 0.0001058366476576295,
+      "loss": 2.2567,
+      "step": 7831
+    },
+    {
+      "epoch": 0.48179133858267714,
+      "grad_norm": 1.2849512100219727,
+      "learning_rate": 0.00010581734287705809,
+      "loss": 2.4467,
+      "step": 7832
+    },
+    {
+      "epoch": 0.4818528543307087,
+      "grad_norm": 1.2409931421279907,
+      "learning_rate": 0.00010579803787895052,
+      "loss": 2.4111,
+      "step": 7833
+    },
+    {
+      "epoch": 0.48191437007874016,
+      "grad_norm": 1.0939905643463135,
+      "learning_rate": 0.00010577873266402869,
+      "loss": 2.0383,
+      "step": 7834
+    },
+    {
+      "epoch": 0.48197588582677164,
+      "grad_norm": 1.2339200973510742,
+      "learning_rate": 0.00010575942723301444,
+      "loss": 2.6654,
+      "step": 7835
+    },
+    {
+      "epoch": 0.4820374015748031,
+      "grad_norm": 1.2574387788772583,
+      "learning_rate": 0.00010574012158662976,
+      "loss": 2.0879,
+      "step": 7836
+    },
+    {
+      "epoch": 0.48209891732283466,
+      "grad_norm": 1.0288536548614502,
+      "learning_rate": 0.00010572081572559654,
+      "loss": 2.7025,
+      "step": 7837
+    },
+    {
+      "epoch": 0.48216043307086615,
+      "grad_norm": 1.533151626586914,
+      "learning_rate": 0.00010570150965063667,
+      "loss": 2.471,
+      "step": 7838
+    },
+    {
+      "epoch": 0.4822219488188976,
+      "grad_norm": 1.054628610610962,
+      "learning_rate": 0.00010568220336247216,
+      "loss": 2.9545,
+      "step": 7839
+    },
+    {
+      "epoch": 0.4822834645669291,
+      "grad_norm": 1.520537257194519,
+      "learning_rate": 0.00010566289686182494,
+      "loss": 2.7831,
+      "step": 7840
+    },
+    {
+      "epoch": 0.48234498031496065,
+      "grad_norm": 1.1613606214523315,
+      "learning_rate": 0.00010564359014941694,
+      "loss": 2.6419,
+      "step": 7841
+    },
+    {
+      "epoch": 0.48240649606299213,
+      "grad_norm": 1.2349547147750854,
+      "learning_rate": 0.00010562428322597013,
+      "loss": 2.6834,
+      "step": 7842
+    },
+    {
+      "epoch": 0.4824680118110236,
+      "grad_norm": 1.2857916355133057,
+      "learning_rate": 0.00010560497609220649,
+      "loss": 2.6537,
+      "step": 7843
+    },
+    {
+      "epoch": 0.4825295275590551,
+      "grad_norm": 1.0249170064926147,
+      "learning_rate": 0.00010558566874884803,
+      "loss": 2.8272,
+      "step": 7844
+    },
+    {
+      "epoch": 0.48259104330708663,
+      "grad_norm": 1.3833844661712646,
+      "learning_rate": 0.00010556636119661667,
+      "loss": 2.6462,
+      "step": 7845
+    },
+    {
+      "epoch": 0.4826525590551181,
+      "grad_norm": 1.3254375457763672,
+      "learning_rate": 0.00010554705343623445,
+      "loss": 2.6709,
+      "step": 7846
+    },
+    {
+      "epoch": 0.4827140748031496,
+      "grad_norm": 1.2213642597198486,
+      "learning_rate": 0.0001055277454684234,
+      "loss": 2.1215,
+      "step": 7847
+    },
+    {
+      "epoch": 0.4827755905511811,
+      "grad_norm": 1.1348814964294434,
+      "learning_rate": 0.00010550843729390543,
+      "loss": 3.0419,
+      "step": 7848
+    },
+    {
+      "epoch": 0.4828371062992126,
+      "grad_norm": 1.2710570096969604,
+      "learning_rate": 0.00010548912891340266,
+      "loss": 2.5431,
+      "step": 7849
+    },
+    {
+      "epoch": 0.4828986220472441,
+      "grad_norm": 1.0805885791778564,
+      "learning_rate": 0.00010546982032763705,
+      "loss": 3.106,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4829601377952756,
+      "grad_norm": 1.3158317804336548,
+      "learning_rate": 0.00010545051153733067,
+      "loss": 2.2439,
+      "step": 7851
+    },
+    {
+      "epoch": 0.48302165354330706,
+      "grad_norm": 1.2625130414962769,
+      "learning_rate": 0.00010543120254320557,
+      "loss": 2.7296,
+      "step": 7852
+    },
+    {
+      "epoch": 0.4830831692913386,
+      "grad_norm": 1.2618921995162964,
+      "learning_rate": 0.00010541189334598379,
+      "loss": 2.067,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4831446850393701,
+      "grad_norm": 1.2230528593063354,
+      "learning_rate": 0.00010539258394638736,
+      "loss": 2.9685,
+      "step": 7854
+    },
+    {
+      "epoch": 0.48320620078740156,
+      "grad_norm": 1.3134373426437378,
+      "learning_rate": 0.00010537327434513833,
+      "loss": 2.5645,
+      "step": 7855
+    },
+    {
+      "epoch": 0.48326771653543305,
+      "grad_norm": 1.356021523475647,
+      "learning_rate": 0.00010535396454295883,
+      "loss": 2.761,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4833292322834646,
+      "grad_norm": 1.057382345199585,
+      "learning_rate": 0.00010533465454057092,
+      "loss": 2.9174,
+      "step": 7857
+    },
+    {
+      "epoch": 0.48339074803149606,
+      "grad_norm": 0.9945914149284363,
+      "learning_rate": 0.00010531534433869668,
+      "loss": 3.1995,
+      "step": 7858
+    },
+    {
+      "epoch": 0.48345226377952755,
+      "grad_norm": 1.3944025039672852,
+      "learning_rate": 0.00010529603393805818,
+      "loss": 2.9711,
+      "step": 7859
+    },
+    {
+      "epoch": 0.48351377952755903,
+      "grad_norm": 1.192490577697754,
+      "learning_rate": 0.00010527672333937753,
+      "loss": 2.0524,
+      "step": 7860
+    },
+    {
+      "epoch": 0.48357529527559057,
+      "grad_norm": 1.1862621307373047,
+      "learning_rate": 0.00010525741254337686,
+      "loss": 2.5585,
+      "step": 7861
+    },
+    {
+      "epoch": 0.48363681102362205,
+      "grad_norm": 1.0440675020217896,
+      "learning_rate": 0.00010523810155077831,
+      "loss": 3.0703,
+      "step": 7862
+    },
+    {
+      "epoch": 0.48369832677165353,
+      "grad_norm": 1.183127760887146,
+      "learning_rate": 0.00010521879036230396,
+      "loss": 2.5874,
+      "step": 7863
+    },
+    {
+      "epoch": 0.483759842519685,
+      "grad_norm": 1.4111837148666382,
+      "learning_rate": 0.0001051994789786759,
+      "loss": 2.648,
+      "step": 7864
+    },
+    {
+      "epoch": 0.48382135826771655,
+      "grad_norm": 1.1309943199157715,
+      "learning_rate": 0.00010518016740061638,
+      "loss": 3.188,
+      "step": 7865
+    },
+    {
+      "epoch": 0.48388287401574803,
+      "grad_norm": 1.3453580141067505,
+      "learning_rate": 0.00010516085562884745,
+      "loss": 2.4223,
+      "step": 7866
+    },
+    {
+      "epoch": 0.4839443897637795,
+      "grad_norm": 1.1905299425125122,
+      "learning_rate": 0.0001051415436640913,
+      "loss": 2.4032,
+      "step": 7867
+    },
+    {
+      "epoch": 0.484005905511811,
+      "grad_norm": 1.1905426979064941,
+      "learning_rate": 0.00010512223150707011,
+      "loss": 2.2857,
+      "step": 7868
+    },
+    {
+      "epoch": 0.48406742125984253,
+      "grad_norm": 1.0999646186828613,
+      "learning_rate": 0.000105102919158506,
+      "loss": 2.0898,
+      "step": 7869
+    },
+    {
+      "epoch": 0.484128937007874,
+      "grad_norm": 1.1502639055252075,
+      "learning_rate": 0.00010508360661912118,
+      "loss": 2.8045,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4841904527559055,
+      "grad_norm": 1.6024699211120605,
+      "learning_rate": 0.00010506429388963781,
+      "loss": 1.7062,
+      "step": 7871
+    },
+    {
+      "epoch": 0.484251968503937,
+      "grad_norm": 1.3829096555709839,
+      "learning_rate": 0.00010504498097077811,
+      "loss": 2.3337,
+      "step": 7872
+    },
+    {
+      "epoch": 0.4843134842519685,
+      "grad_norm": 1.5889042615890503,
+      "learning_rate": 0.00010502566786326422,
+      "loss": 2.3105,
+      "step": 7873
+    },
+    {
+      "epoch": 0.484375,
+      "grad_norm": 1.1662575006484985,
+      "learning_rate": 0.0001050063545678184,
+      "loss": 1.7254,
+      "step": 7874
+    },
+    {
+      "epoch": 0.4844365157480315,
+      "grad_norm": 1.3360068798065186,
+      "learning_rate": 0.00010498704108516284,
+      "loss": 2.5484,
+      "step": 7875
+    },
+    {
+      "epoch": 0.484498031496063,
+      "grad_norm": 1.2509642839431763,
+      "learning_rate": 0.00010496772741601974,
+      "loss": 2.4614,
+      "step": 7876
+    },
+    {
+      "epoch": 0.4845595472440945,
+      "grad_norm": 1.1272648572921753,
+      "learning_rate": 0.00010494841356111134,
+      "loss": 2.558,
+      "step": 7877
+    },
+    {
+      "epoch": 0.484621062992126,
+      "grad_norm": 1.283819317817688,
+      "learning_rate": 0.00010492909952115985,
+      "loss": 2.3546,
+      "step": 7878
+    },
+    {
+      "epoch": 0.48468257874015747,
+      "grad_norm": 1.1212811470031738,
+      "learning_rate": 0.00010490978529688756,
+      "loss": 2.6285,
+      "step": 7879
+    },
+    {
+      "epoch": 0.484744094488189,
+      "grad_norm": 1.1672048568725586,
+      "learning_rate": 0.00010489047088901663,
+      "loss": 2.7923,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4848056102362205,
+      "grad_norm": 1.3002959489822388,
+      "learning_rate": 0.00010487115629826938,
+      "loss": 1.9103,
+      "step": 7881
+    },
+    {
+      "epoch": 0.48486712598425197,
+      "grad_norm": 1.0481263399124146,
+      "learning_rate": 0.00010485184152536807,
+      "loss": 2.5313,
+      "step": 7882
+    },
+    {
+      "epoch": 0.48492864173228345,
+      "grad_norm": 1.2431764602661133,
+      "learning_rate": 0.0001048325265710349,
+      "loss": 2.6496,
+      "step": 7883
+    },
+    {
+      "epoch": 0.484990157480315,
+      "grad_norm": 1.2239537239074707,
+      "learning_rate": 0.00010481321143599221,
+      "loss": 2.5363,
+      "step": 7884
+    },
+    {
+      "epoch": 0.48505167322834647,
+      "grad_norm": 1.122721791267395,
+      "learning_rate": 0.00010479389612096224,
+      "loss": 3.2734,
+      "step": 7885
+    },
+    {
+      "epoch": 0.48511318897637795,
+      "grad_norm": 1.1417568922042847,
+      "learning_rate": 0.0001047745806266673,
+      "loss": 2.1864,
+      "step": 7886
+    },
+    {
+      "epoch": 0.48517470472440943,
+      "grad_norm": 1.3196957111358643,
+      "learning_rate": 0.00010475526495382965,
+      "loss": 2.756,
+      "step": 7887
+    },
+    {
+      "epoch": 0.48523622047244097,
+      "grad_norm": 1.257440447807312,
+      "learning_rate": 0.00010473594910317164,
+      "loss": 3.1846,
+      "step": 7888
+    },
+    {
+      "epoch": 0.48529773622047245,
+      "grad_norm": 1.2984037399291992,
+      "learning_rate": 0.0001047166330754155,
+      "loss": 2.3543,
+      "step": 7889
+    },
+    {
+      "epoch": 0.48535925196850394,
+      "grad_norm": 1.203837275505066,
+      "learning_rate": 0.0001046973168712836,
+      "loss": 3.0059,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4854207677165354,
+      "grad_norm": 1.3052457571029663,
+      "learning_rate": 0.00010467800049149826,
+      "loss": 2.7976,
+      "step": 7891
+    },
+    {
+      "epoch": 0.48548228346456695,
+      "grad_norm": 1.0131052732467651,
+      "learning_rate": 0.00010465868393678177,
+      "loss": 3.3918,
+      "step": 7892
+    },
+    {
+      "epoch": 0.48554379921259844,
+      "grad_norm": 1.1924010515213013,
+      "learning_rate": 0.00010463936720785646,
+      "loss": 2.7823,
+      "step": 7893
+    },
+    {
+      "epoch": 0.4856053149606299,
+      "grad_norm": 1.3123244047164917,
+      "learning_rate": 0.0001046200503054447,
+      "loss": 2.6468,
+      "step": 7894
+    },
+    {
+      "epoch": 0.4856668307086614,
+      "grad_norm": 0.908377468585968,
+      "learning_rate": 0.00010460073323026884,
+      "loss": 2.884,
+      "step": 7895
+    },
+    {
+      "epoch": 0.48572834645669294,
+      "grad_norm": 1.4268203973770142,
+      "learning_rate": 0.00010458141598305122,
+      "loss": 2.5744,
+      "step": 7896
+    },
+    {
+      "epoch": 0.4857898622047244,
+      "grad_norm": 1.2027627229690552,
+      "learning_rate": 0.00010456209856451418,
+      "loss": 2.7756,
+      "step": 7897
+    },
+    {
+      "epoch": 0.4858513779527559,
+      "grad_norm": 1.3398345708847046,
+      "learning_rate": 0.00010454278097538009,
+      "loss": 1.6274,
+      "step": 7898
+    },
+    {
+      "epoch": 0.4859128937007874,
+      "grad_norm": 1.4429328441619873,
+      "learning_rate": 0.00010452346321637128,
+      "loss": 2.9525,
+      "step": 7899
+    },
+    {
+      "epoch": 0.4859744094488189,
+      "grad_norm": 1.2730181217193604,
+      "learning_rate": 0.0001045041452882102,
+      "loss": 2.293,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4860359251968504,
+      "grad_norm": 1.3057634830474854,
+      "learning_rate": 0.00010448482719161922,
+      "loss": 2.4051,
+      "step": 7901
+    },
+    {
+      "epoch": 0.4860974409448819,
+      "grad_norm": 1.0964579582214355,
+      "learning_rate": 0.00010446550892732068,
+      "loss": 2.9011,
+      "step": 7902
+    },
+    {
+      "epoch": 0.48615895669291337,
+      "grad_norm": 1.1312830448150635,
+      "learning_rate": 0.00010444619049603703,
+      "loss": 2.8296,
+      "step": 7903
+    },
+    {
+      "epoch": 0.4862204724409449,
+      "grad_norm": 1.0698341131210327,
+      "learning_rate": 0.00010442687189849065,
+      "loss": 2.8115,
+      "step": 7904
+    },
+    {
+      "epoch": 0.4862819881889764,
+      "grad_norm": 1.2122777700424194,
+      "learning_rate": 0.00010440755313540391,
+      "loss": 1.8674,
+      "step": 7905
+    },
+    {
+      "epoch": 0.48634350393700787,
+      "grad_norm": 1.423335075378418,
+      "learning_rate": 0.0001043882342074993,
+      "loss": 2.2909,
+      "step": 7906
+    },
+    {
+      "epoch": 0.48640501968503935,
+      "grad_norm": 1.028366208076477,
+      "learning_rate": 0.00010436891511549918,
+      "loss": 1.8897,
+      "step": 7907
+    },
+    {
+      "epoch": 0.4864665354330709,
+      "grad_norm": 1.095047950744629,
+      "learning_rate": 0.00010434959586012598,
+      "loss": 2.2882,
+      "step": 7908
+    },
+    {
+      "epoch": 0.4865280511811024,
+      "grad_norm": 1.159894347190857,
+      "learning_rate": 0.00010433027644210219,
+      "loss": 2.8275,
+      "step": 7909
+    },
+    {
+      "epoch": 0.48658956692913385,
+      "grad_norm": 1.163830280303955,
+      "learning_rate": 0.0001043109568621502,
+      "loss": 2.3039,
+      "step": 7910
+    },
+    {
+      "epoch": 0.48665108267716534,
+      "grad_norm": 1.4256505966186523,
+      "learning_rate": 0.00010429163712099246,
+      "loss": 2.4747,
+      "step": 7911
+    },
+    {
+      "epoch": 0.4867125984251969,
+      "grad_norm": 1.2266684770584106,
+      "learning_rate": 0.00010427231721935138,
+      "loss": 1.9016,
+      "step": 7912
+    },
+    {
+      "epoch": 0.48677411417322836,
+      "grad_norm": 1.416464924812317,
+      "learning_rate": 0.00010425299715794949,
+      "loss": 1.9944,
+      "step": 7913
+    },
+    {
+      "epoch": 0.48683562992125984,
+      "grad_norm": 1.4612195491790771,
+      "learning_rate": 0.00010423367693750925,
+      "loss": 2.3085,
+      "step": 7914
+    },
+    {
+      "epoch": 0.4868971456692913,
+      "grad_norm": 1.2582570314407349,
+      "learning_rate": 0.00010421435655875305,
+      "loss": 2.5733,
+      "step": 7915
+    },
+    {
+      "epoch": 0.48695866141732286,
+      "grad_norm": 1.0579376220703125,
+      "learning_rate": 0.00010419503602240346,
+      "loss": 3.3609,
+      "step": 7916
+    },
+    {
+      "epoch": 0.48702017716535434,
+      "grad_norm": 1.3831584453582764,
+      "learning_rate": 0.00010417571532918289,
+      "loss": 2.5481,
+      "step": 7917
+    },
+    {
+      "epoch": 0.4870816929133858,
+      "grad_norm": 0.9880089163780212,
+      "learning_rate": 0.00010415639447981384,
+      "loss": 2.8064,
+      "step": 7918
+    },
+    {
+      "epoch": 0.4871432086614173,
+      "grad_norm": 1.3981678485870361,
+      "learning_rate": 0.00010413707347501882,
+      "loss": 1.7789,
+      "step": 7919
+    },
+    {
+      "epoch": 0.48720472440944884,
+      "grad_norm": 1.1640158891677856,
+      "learning_rate": 0.00010411775231552034,
+      "loss": 2.3099,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4872662401574803,
+      "grad_norm": 1.2891827821731567,
+      "learning_rate": 0.00010409843100204087,
+      "loss": 3.0013,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4873277559055118,
+      "grad_norm": 2.1194918155670166,
+      "learning_rate": 0.00010407910953530294,
+      "loss": 2.6593,
+      "step": 7922
+    },
+    {
+      "epoch": 0.4873892716535433,
+      "grad_norm": 0.8618022799491882,
+      "learning_rate": 0.00010405978791602907,
+      "loss": 1.758,
+      "step": 7923
+    },
+    {
+      "epoch": 0.4874507874015748,
+      "grad_norm": 1.0030373334884644,
+      "learning_rate": 0.00010404046614494174,
+      "loss": 2.7385,
+      "step": 7924
+    },
+    {
+      "epoch": 0.4875123031496063,
+      "grad_norm": 0.9397240281105042,
+      "learning_rate": 0.00010402114422276352,
+      "loss": 2.7006,
+      "step": 7925
+    },
+    {
+      "epoch": 0.4875738188976378,
+      "grad_norm": 1.2072759866714478,
+      "learning_rate": 0.00010400182215021695,
+      "loss": 2.6446,
+      "step": 7926
+    },
+    {
+      "epoch": 0.48763533464566927,
+      "grad_norm": 1.2449004650115967,
+      "learning_rate": 0.00010398249992802451,
+      "loss": 3.242,
+      "step": 7927
+    },
+    {
+      "epoch": 0.4876968503937008,
+      "grad_norm": 1.4644886255264282,
+      "learning_rate": 0.00010396317755690879,
+      "loss": 2.8172,
+      "step": 7928
+    },
+    {
+      "epoch": 0.4877583661417323,
+      "grad_norm": 1.265195608139038,
+      "learning_rate": 0.00010394385503759234,
+      "loss": 2.3172,
+      "step": 7929
+    },
+    {
+      "epoch": 0.4878198818897638,
+      "grad_norm": 1.0918716192245483,
+      "learning_rate": 0.00010392453237079772,
+      "loss": 2.5377,
+      "step": 7930
+    },
+    {
+      "epoch": 0.48788139763779526,
+      "grad_norm": 1.3099734783172607,
+      "learning_rate": 0.00010390520955724745,
+      "loss": 1.8148,
+      "step": 7931
+    },
+    {
+      "epoch": 0.4879429133858268,
+      "grad_norm": 1.0772491693496704,
+      "learning_rate": 0.00010388588659766411,
+      "loss": 2.9691,
+      "step": 7932
+    },
+    {
+      "epoch": 0.4880044291338583,
+      "grad_norm": 1.2083429098129272,
+      "learning_rate": 0.0001038665634927703,
+      "loss": 2.557,
+      "step": 7933
+    },
+    {
+      "epoch": 0.48806594488188976,
+      "grad_norm": 1.170527458190918,
+      "learning_rate": 0.00010384724024328856,
+      "loss": 1.8586,
+      "step": 7934
+    },
+    {
+      "epoch": 0.48812746062992124,
+      "grad_norm": 1.0114551782608032,
+      "learning_rate": 0.00010382791684994149,
+      "loss": 2.658,
+      "step": 7935
+    },
+    {
+      "epoch": 0.4881889763779528,
+      "grad_norm": 1.150352954864502,
+      "learning_rate": 0.00010380859331345169,
+      "loss": 2.236,
+      "step": 7936
+    },
+    {
+      "epoch": 0.48825049212598426,
+      "grad_norm": 1.2289948463439941,
+      "learning_rate": 0.00010378926963454172,
+      "loss": 2.4726,
+      "step": 7937
+    },
+    {
+      "epoch": 0.48831200787401574,
+      "grad_norm": 1.2644215822219849,
+      "learning_rate": 0.0001037699458139342,
+      "loss": 2.6844,
+      "step": 7938
+    },
+    {
+      "epoch": 0.4883735236220472,
+      "grad_norm": 1.3717159032821655,
+      "learning_rate": 0.00010375062185235173,
+      "loss": 2.8508,
+      "step": 7939
+    },
+    {
+      "epoch": 0.48843503937007876,
+      "grad_norm": 1.2760165929794312,
+      "learning_rate": 0.00010373129775051691,
+      "loss": 2.6247,
+      "step": 7940
+    },
+    {
+      "epoch": 0.48849655511811024,
+      "grad_norm": 1.476924180984497,
+      "learning_rate": 0.00010371197350915236,
+      "loss": 2.577,
+      "step": 7941
+    },
+    {
+      "epoch": 0.4885580708661417,
+      "grad_norm": 1.1164991855621338,
+      "learning_rate": 0.0001036926491289807,
+      "loss": 2.5783,
+      "step": 7942
+    },
+    {
+      "epoch": 0.4886195866141732,
+      "grad_norm": 1.3708242177963257,
+      "learning_rate": 0.0001036733246107246,
+      "loss": 2.0471,
+      "step": 7943
+    },
+    {
+      "epoch": 0.48868110236220474,
+      "grad_norm": 1.1813691854476929,
+      "learning_rate": 0.00010365399995510658,
+      "loss": 2.7973,
+      "step": 7944
+    },
+    {
+      "epoch": 0.4887426181102362,
+      "grad_norm": 1.251155972480774,
+      "learning_rate": 0.00010363467516284935,
+      "loss": 2.7601,
+      "step": 7945
+    },
+    {
+      "epoch": 0.4888041338582677,
+      "grad_norm": 1.19106125831604,
+      "learning_rate": 0.00010361535023467557,
+      "loss": 2.381,
+      "step": 7946
+    },
+    {
+      "epoch": 0.4888656496062992,
+      "grad_norm": 1.0438249111175537,
+      "learning_rate": 0.0001035960251713078,
+      "loss": 3.0623,
+      "step": 7947
+    },
+    {
+      "epoch": 0.48892716535433073,
+      "grad_norm": 1.0889898538589478,
+      "learning_rate": 0.00010357669997346877,
+      "loss": 2.9665,
+      "step": 7948
+    },
+    {
+      "epoch": 0.4889886811023622,
+      "grad_norm": 1.345669150352478,
+      "learning_rate": 0.00010355737464188111,
+      "loss": 2.2552,
+      "step": 7949
+    },
+    {
+      "epoch": 0.4890501968503937,
+      "grad_norm": 1.2294317483901978,
+      "learning_rate": 0.00010353804917726747,
+      "loss": 2.2883,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4891117125984252,
+      "grad_norm": 1.3435773849487305,
+      "learning_rate": 0.00010351872358035051,
+      "loss": 2.5949,
+      "step": 7951
+    },
+    {
+      "epoch": 0.4891732283464567,
+      "grad_norm": 1.025640606880188,
+      "learning_rate": 0.00010349939785185292,
+      "loss": 2.1165,
+      "step": 7952
+    },
+    {
+      "epoch": 0.4892347440944882,
+      "grad_norm": 1.1570559740066528,
+      "learning_rate": 0.00010348007199249734,
+      "loss": 2.76,
+      "step": 7953
+    },
+    {
+      "epoch": 0.4892962598425197,
+      "grad_norm": 1.3425816297531128,
+      "learning_rate": 0.00010346074600300651,
+      "loss": 2.8859,
+      "step": 7954
+    },
+    {
+      "epoch": 0.48935777559055116,
+      "grad_norm": 1.0541659593582153,
+      "learning_rate": 0.00010344141988410305,
+      "loss": 2.8481,
+      "step": 7955
+    },
+    {
+      "epoch": 0.4894192913385827,
+      "grad_norm": 1.3510029315948486,
+      "learning_rate": 0.00010342209363650969,
+      "loss": 2.7792,
+      "step": 7956
+    },
+    {
+      "epoch": 0.4894808070866142,
+      "grad_norm": 1.1471670866012573,
+      "learning_rate": 0.00010340276726094908,
+      "loss": 2.8064,
+      "step": 7957
+    },
+    {
+      "epoch": 0.48954232283464566,
+      "grad_norm": 0.9800888299942017,
+      "learning_rate": 0.000103383440758144,
+      "loss": 2.602,
+      "step": 7958
+    },
+    {
+      "epoch": 0.48960383858267714,
+      "grad_norm": 1.2650357484817505,
+      "learning_rate": 0.00010336411412881708,
+      "loss": 2.3029,
+      "step": 7959
+    },
+    {
+      "epoch": 0.4896653543307087,
+      "grad_norm": 1.3090823888778687,
+      "learning_rate": 0.00010334478737369102,
+      "loss": 2.4552,
+      "step": 7960
+    },
+    {
+      "epoch": 0.48972687007874016,
+      "grad_norm": 1.3167332410812378,
+      "learning_rate": 0.00010332546049348861,
+      "loss": 2.1281,
+      "step": 7961
+    },
+    {
+      "epoch": 0.48978838582677164,
+      "grad_norm": 1.2923017740249634,
+      "learning_rate": 0.0001033061334889325,
+      "loss": 2.3049,
+      "step": 7962
+    },
+    {
+      "epoch": 0.4898499015748031,
+      "grad_norm": 1.3940361738204956,
+      "learning_rate": 0.00010328680636074541,
+      "loss": 2.3733,
+      "step": 7963
+    },
+    {
+      "epoch": 0.48991141732283466,
+      "grad_norm": 1.2805423736572266,
+      "learning_rate": 0.00010326747910965013,
+      "loss": 2.1994,
+      "step": 7964
+    },
+    {
+      "epoch": 0.48997293307086615,
+      "grad_norm": 1.139833927154541,
+      "learning_rate": 0.00010324815173636934,
+      "loss": 2.5605,
+      "step": 7965
+    },
+    {
+      "epoch": 0.4900344488188976,
+      "grad_norm": 1.0853452682495117,
+      "learning_rate": 0.00010322882424162577,
+      "loss": 2.7859,
+      "step": 7966
+    },
+    {
+      "epoch": 0.4900959645669291,
+      "grad_norm": 1.1464111804962158,
+      "learning_rate": 0.00010320949662614218,
+      "loss": 2.8895,
+      "step": 7967
+    },
+    {
+      "epoch": 0.49015748031496065,
+      "grad_norm": 1.1818221807479858,
+      "learning_rate": 0.00010319016889064137,
+      "loss": 2.1947,
+      "step": 7968
+    },
+    {
+      "epoch": 0.49021899606299213,
+      "grad_norm": 0.9138051271438599,
+      "learning_rate": 0.00010317084103584598,
+      "loss": 2.6791,
+      "step": 7969
+    },
+    {
+      "epoch": 0.4902805118110236,
+      "grad_norm": 1.3087915182113647,
+      "learning_rate": 0.00010315151306247882,
+      "loss": 2.5465,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4903420275590551,
+      "grad_norm": 1.23324453830719,
+      "learning_rate": 0.00010313218497126265,
+      "loss": 2.4901,
+      "step": 7971
+    },
+    {
+      "epoch": 0.49040354330708663,
+      "grad_norm": 1.1738560199737549,
+      "learning_rate": 0.0001031128567629202,
+      "loss": 2.5094,
+      "step": 7972
+    },
+    {
+      "epoch": 0.4904650590551181,
+      "grad_norm": 1.226456880569458,
+      "learning_rate": 0.0001030935284381743,
+      "loss": 2.2681,
+      "step": 7973
+    },
+    {
+      "epoch": 0.4905265748031496,
+      "grad_norm": 1.4611310958862305,
+      "learning_rate": 0.00010307419999774767,
+      "loss": 2.1946,
+      "step": 7974
+    },
+    {
+      "epoch": 0.4905880905511811,
+      "grad_norm": 1.1263529062271118,
+      "learning_rate": 0.0001030548714423631,
+      "loss": 2.759,
+      "step": 7975
+    },
+    {
+      "epoch": 0.4906496062992126,
+      "grad_norm": 1.2053600549697876,
+      "learning_rate": 0.00010303554277274338,
+      "loss": 3.2066,
+      "step": 7976
+    },
+    {
+      "epoch": 0.4907111220472441,
+      "grad_norm": 1.218946099281311,
+      "learning_rate": 0.0001030162139896113,
+      "loss": 2.7414,
+      "step": 7977
+    },
+    {
+      "epoch": 0.4907726377952756,
+      "grad_norm": 1.099171757698059,
+      "learning_rate": 0.00010299688509368964,
+      "loss": 3.0225,
+      "step": 7978
+    },
+    {
+      "epoch": 0.49083415354330706,
+      "grad_norm": 1.2200366258621216,
+      "learning_rate": 0.00010297755608570114,
+      "loss": 2.4666,
+      "step": 7979
+    },
+    {
+      "epoch": 0.4908956692913386,
+      "grad_norm": 1.110369086265564,
+      "learning_rate": 0.0001029582269663687,
+      "loss": 2.3511,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4909571850393701,
+      "grad_norm": 1.1547049283981323,
+      "learning_rate": 0.00010293889773641506,
+      "loss": 2.6796,
+      "step": 7981
+    },
+    {
+      "epoch": 0.49101870078740156,
+      "grad_norm": 1.1721627712249756,
+      "learning_rate": 0.00010291956839656298,
+      "loss": 2.2721,
+      "step": 7982
+    },
+    {
+      "epoch": 0.49108021653543305,
+      "grad_norm": 1.137723684310913,
+      "learning_rate": 0.00010290023894753535,
+      "loss": 2.6317,
+      "step": 7983
+    },
+    {
+      "epoch": 0.4911417322834646,
+      "grad_norm": 1.1756068468093872,
+      "learning_rate": 0.00010288090939005499,
+      "loss": 2.7242,
+      "step": 7984
+    },
+    {
+      "epoch": 0.49120324803149606,
+      "grad_norm": 1.1580960750579834,
+      "learning_rate": 0.00010286157972484463,
+      "loss": 2.4613,
+      "step": 7985
+    },
+    {
+      "epoch": 0.49126476377952755,
+      "grad_norm": 1.1891722679138184,
+      "learning_rate": 0.00010284224995262719,
+      "loss": 2.2703,
+      "step": 7986
+    },
+    {
+      "epoch": 0.49132627952755903,
+      "grad_norm": 1.2213929891586304,
+      "learning_rate": 0.00010282292007412543,
+      "loss": 3.0383,
+      "step": 7987
+    },
+    {
+      "epoch": 0.49138779527559057,
+      "grad_norm": 1.0856600999832153,
+      "learning_rate": 0.00010280359009006219,
+      "loss": 2.6607,
+      "step": 7988
+    },
+    {
+      "epoch": 0.49144931102362205,
+      "grad_norm": 1.1710997819900513,
+      "learning_rate": 0.00010278426000116033,
+      "loss": 2.2095,
+      "step": 7989
+    },
+    {
+      "epoch": 0.49151082677165353,
+      "grad_norm": 1.4868483543395996,
+      "learning_rate": 0.00010276492980814268,
+      "loss": 2.6653,
+      "step": 7990
+    },
+    {
+      "epoch": 0.491572342519685,
+      "grad_norm": 1.037477970123291,
+      "learning_rate": 0.00010274559951173207,
+      "loss": 2.2176,
+      "step": 7991
+    },
+    {
+      "epoch": 0.49163385826771655,
+      "grad_norm": 1.5481314659118652,
+      "learning_rate": 0.0001027262691126513,
+      "loss": 2.6825,
+      "step": 7992
+    },
+    {
+      "epoch": 0.49169537401574803,
+      "grad_norm": 1.2877371311187744,
+      "learning_rate": 0.00010270693861162332,
+      "loss": 2.5172,
+      "step": 7993
+    },
+    {
+      "epoch": 0.4917568897637795,
+      "grad_norm": 1.3207194805145264,
+      "learning_rate": 0.00010268760800937093,
+      "loss": 2.167,
+      "step": 7994
+    },
+    {
+      "epoch": 0.491818405511811,
+      "grad_norm": 1.1924972534179688,
+      "learning_rate": 0.00010266827730661697,
+      "loss": 2.6477,
+      "step": 7995
+    },
+    {
+      "epoch": 0.49187992125984253,
+      "grad_norm": 1.1901606321334839,
+      "learning_rate": 0.00010264894650408432,
+      "loss": 1.9713,
+      "step": 7996
+    },
+    {
+      "epoch": 0.491941437007874,
+      "grad_norm": 1.1179689168930054,
+      "learning_rate": 0.00010262961560249584,
+      "loss": 2.7303,
+      "step": 7997
+    },
+    {
+      "epoch": 0.4920029527559055,
+      "grad_norm": 1.34584379196167,
+      "learning_rate": 0.0001026102846025744,
+      "loss": 2.8789,
+      "step": 7998
+    },
+    {
+      "epoch": 0.492064468503937,
+      "grad_norm": 1.2757887840270996,
+      "learning_rate": 0.00010259095350504286,
+      "loss": 2.5744,
+      "step": 7999
+    },
+    {
+      "epoch": 0.4921259842519685,
+      "grad_norm": 1.1413215398788452,
+      "learning_rate": 0.00010257162231062414,
+      "loss": 2.6961,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4921875,
+      "grad_norm": 1.3210691213607788,
+      "learning_rate": 0.00010255229102004105,
+      "loss": 1.9718,
+      "step": 8001
+    },
+    {
+      "epoch": 0.4922490157480315,
+      "grad_norm": 1.1422626972198486,
+      "learning_rate": 0.00010253295963401652,
+      "loss": 2.7728,
+      "step": 8002
+    },
+    {
+      "epoch": 0.492310531496063,
+      "grad_norm": 1.184571385383606,
+      "learning_rate": 0.00010251362815327343,
+      "loss": 2.9899,
+      "step": 8003
+    },
+    {
+      "epoch": 0.4923720472440945,
+      "grad_norm": 1.196838617324829,
+      "learning_rate": 0.00010249429657853469,
+      "loss": 2.6149,
+      "step": 8004
+    },
+    {
+      "epoch": 0.492433562992126,
+      "grad_norm": 1.0473861694335938,
+      "learning_rate": 0.00010247496491052313,
+      "loss": 3.1183,
+      "step": 8005
+    },
+    {
+      "epoch": 0.49249507874015747,
+      "grad_norm": 1.433659553527832,
+      "learning_rate": 0.0001024556331499617,
+      "loss": 2.9019,
+      "step": 8006
+    },
+    {
+      "epoch": 0.492556594488189,
+      "grad_norm": 1.2542264461517334,
+      "learning_rate": 0.00010243630129757324,
+      "loss": 3.0851,
+      "step": 8007
+    },
+    {
+      "epoch": 0.4926181102362205,
+      "grad_norm": 1.100046157836914,
+      "learning_rate": 0.00010241696935408076,
+      "loss": 2.7408,
+      "step": 8008
+    },
+    {
+      "epoch": 0.49267962598425197,
+      "grad_norm": 1.0076274871826172,
+      "learning_rate": 0.00010239763732020707,
+      "loss": 2.5726,
+      "step": 8009
+    },
+    {
+      "epoch": 0.49274114173228345,
+      "grad_norm": 1.3178813457489014,
+      "learning_rate": 0.0001023783051966751,
+      "loss": 2.8817,
+      "step": 8010
+    },
+    {
+      "epoch": 0.492802657480315,
+      "grad_norm": 1.109753131866455,
+      "learning_rate": 0.0001023589729842078,
+      "loss": 2.719,
+      "step": 8011
+    },
+    {
+      "epoch": 0.49286417322834647,
+      "grad_norm": 1.2878189086914062,
+      "learning_rate": 0.00010233964068352806,
+      "loss": 1.192,
+      "step": 8012
+    },
+    {
+      "epoch": 0.49292568897637795,
+      "grad_norm": 1.2201826572418213,
+      "learning_rate": 0.00010232030829535882,
+      "loss": 2.3492,
+      "step": 8013
+    },
+    {
+      "epoch": 0.49298720472440943,
+      "grad_norm": 1.3213222026824951,
+      "learning_rate": 0.00010230097582042295,
+      "loss": 2.3079,
+      "step": 8014
+    },
+    {
+      "epoch": 0.49304872047244097,
+      "grad_norm": 1.0434002876281738,
+      "learning_rate": 0.00010228164325944345,
+      "loss": 2.5101,
+      "step": 8015
+    },
+    {
+      "epoch": 0.49311023622047245,
+      "grad_norm": 1.7574670314788818,
+      "learning_rate": 0.0001022623106131432,
+      "loss": 2.3488,
+      "step": 8016
+    },
+    {
+      "epoch": 0.49317175196850394,
+      "grad_norm": 1.2045958042144775,
+      "learning_rate": 0.00010224297788224512,
+      "loss": 2.5341,
+      "step": 8017
+    },
+    {
+      "epoch": 0.4932332677165354,
+      "grad_norm": 1.1568257808685303,
+      "learning_rate": 0.00010222364506747224,
+      "loss": 2.4208,
+      "step": 8018
+    },
+    {
+      "epoch": 0.49329478346456695,
+      "grad_norm": 1.246476411819458,
+      "learning_rate": 0.0001022043121695474,
+      "loss": 2.8668,
+      "step": 8019
+    },
+    {
+      "epoch": 0.49335629921259844,
+      "grad_norm": 1.3676202297210693,
+      "learning_rate": 0.00010218497918919354,
+      "loss": 2.6675,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4934178149606299,
+      "grad_norm": 1.13251793384552,
+      "learning_rate": 0.00010216564612713368,
+      "loss": 3.0028,
+      "step": 8021
+    },
+    {
+      "epoch": 0.4934793307086614,
+      "grad_norm": 1.0290311574935913,
+      "learning_rate": 0.00010214631298409075,
+      "loss": 3.0253,
+      "step": 8022
+    },
+    {
+      "epoch": 0.49354084645669294,
+      "grad_norm": 1.1097441911697388,
+      "learning_rate": 0.00010212697976078764,
+      "loss": 2.447,
+      "step": 8023
+    },
+    {
+      "epoch": 0.4936023622047244,
+      "grad_norm": 1.1433817148208618,
+      "learning_rate": 0.00010210764645794737,
+      "loss": 2.5386,
+      "step": 8024
+    },
+    {
+      "epoch": 0.4936638779527559,
+      "grad_norm": 1.233014702796936,
+      "learning_rate": 0.00010208831307629288,
+      "loss": 2.9489,
+      "step": 8025
+    },
+    {
+      "epoch": 0.4937253937007874,
+      "grad_norm": 1.2549322843551636,
+      "learning_rate": 0.00010206897961654713,
+      "loss": 1.9906,
+      "step": 8026
+    },
+    {
+      "epoch": 0.4937869094488189,
+      "grad_norm": 1.2327656745910645,
+      "learning_rate": 0.00010204964607943306,
+      "loss": 2.933,
+      "step": 8027
+    },
+    {
+      "epoch": 0.4938484251968504,
+      "grad_norm": 1.345576524734497,
+      "learning_rate": 0.00010203031246567365,
+      "loss": 2.3047,
+      "step": 8028
+    },
+    {
+      "epoch": 0.4939099409448819,
+      "grad_norm": 1.2556227445602417,
+      "learning_rate": 0.0001020109787759919,
+      "loss": 3.1245,
+      "step": 8029
+    },
+    {
+      "epoch": 0.49397145669291337,
+      "grad_norm": 1.1539448499679565,
+      "learning_rate": 0.00010199164501111073,
+      "loss": 2.6894,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4940329724409449,
+      "grad_norm": 1.3034554719924927,
+      "learning_rate": 0.00010197231117175317,
+      "loss": 2.6764,
+      "step": 8031
+    },
+    {
+      "epoch": 0.4940944881889764,
+      "grad_norm": 1.2519049644470215,
+      "learning_rate": 0.00010195297725864216,
+      "loss": 2.4174,
+      "step": 8032
+    },
+    {
+      "epoch": 0.49415600393700787,
+      "grad_norm": 1.3792946338653564,
+      "learning_rate": 0.00010193364327250068,
+      "loss": 2.4119,
+      "step": 8033
+    },
+    {
+      "epoch": 0.49421751968503935,
+      "grad_norm": 1.2379392385482788,
+      "learning_rate": 0.00010191430921405174,
+      "loss": 2.6774,
+      "step": 8034
+    },
+    {
+      "epoch": 0.4942790354330709,
+      "grad_norm": 1.0859769582748413,
+      "learning_rate": 0.00010189497508401831,
+      "loss": 2.7038,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4943405511811024,
+      "grad_norm": 1.0914816856384277,
+      "learning_rate": 0.00010187564088312334,
+      "loss": 2.4735,
+      "step": 8036
+    },
+    {
+      "epoch": 0.49440206692913385,
+      "grad_norm": 1.4692819118499756,
+      "learning_rate": 0.00010185630661208989,
+      "loss": 2.7797,
+      "step": 8037
+    },
+    {
+      "epoch": 0.49446358267716534,
+      "grad_norm": 1.2143222093582153,
+      "learning_rate": 0.00010183697227164096,
+      "loss": 2.2054,
+      "step": 8038
+    },
+    {
+      "epoch": 0.4945250984251969,
+      "grad_norm": 1.0908098220825195,
+      "learning_rate": 0.00010181763786249946,
+      "loss": 1.3237,
+      "step": 8039
+    },
+    {
+      "epoch": 0.49458661417322836,
+      "grad_norm": 1.029659628868103,
+      "learning_rate": 0.00010179830338538844,
+      "loss": 2.8777,
+      "step": 8040
+    },
+    {
+      "epoch": 0.49464812992125984,
+      "grad_norm": 1.1105406284332275,
+      "learning_rate": 0.00010177896884103093,
+      "loss": 2.7996,
+      "step": 8041
+    },
+    {
+      "epoch": 0.4947096456692913,
+      "grad_norm": 1.286268949508667,
+      "learning_rate": 0.0001017596342301499,
+      "loss": 2.6397,
+      "step": 8042
+    },
+    {
+      "epoch": 0.49477116141732286,
+      "grad_norm": 1.282758355140686,
+      "learning_rate": 0.00010174029955346836,
+      "loss": 2.5483,
+      "step": 8043
+    },
+    {
+      "epoch": 0.49483267716535434,
+      "grad_norm": 1.2559682130813599,
+      "learning_rate": 0.0001017209648117093,
+      "loss": 2.9028,
+      "step": 8044
+    },
+    {
+      "epoch": 0.4948941929133858,
+      "grad_norm": 1.295530915260315,
+      "learning_rate": 0.00010170163000559575,
+      "loss": 2.5554,
+      "step": 8045
+    },
+    {
+      "epoch": 0.4949557086614173,
+      "grad_norm": 1.1703797578811646,
+      "learning_rate": 0.00010168229513585074,
+      "loss": 2.4397,
+      "step": 8046
+    },
+    {
+      "epoch": 0.49501722440944884,
+      "grad_norm": 1.225110650062561,
+      "learning_rate": 0.0001016629602031973,
+      "loss": 2.2932,
+      "step": 8047
+    },
+    {
+      "epoch": 0.4950787401574803,
+      "grad_norm": 1.1887551546096802,
+      "learning_rate": 0.0001016436252083584,
+      "loss": 2.6771,
+      "step": 8048
+    },
+    {
+      "epoch": 0.4951402559055118,
+      "grad_norm": 1.1910514831542969,
+      "learning_rate": 0.00010162429015205706,
+      "loss": 2.611,
+      "step": 8049
+    },
+    {
+      "epoch": 0.4952017716535433,
+      "grad_norm": 1.197386384010315,
+      "learning_rate": 0.00010160495503501635,
+      "loss": 2.7226,
+      "step": 8050
+    },
+    {
+      "epoch": 0.4952632874015748,
+      "grad_norm": 1.2510840892791748,
+      "learning_rate": 0.00010158561985795927,
+      "loss": 2.9115,
+      "step": 8051
+    },
+    {
+      "epoch": 0.4953248031496063,
+      "grad_norm": 1.3057364225387573,
+      "learning_rate": 0.00010156628462160882,
+      "loss": 2.5079,
+      "step": 8052
+    },
+    {
+      "epoch": 0.4953863188976378,
+      "grad_norm": 1.0038979053497314,
+      "learning_rate": 0.0001015469493266881,
+      "loss": 2.7093,
+      "step": 8053
+    },
+    {
+      "epoch": 0.49544783464566927,
+      "grad_norm": 2.0526084899902344,
+      "learning_rate": 0.00010152761397392008,
+      "loss": 1.3599,
+      "step": 8054
+    },
+    {
+      "epoch": 0.4955093503937008,
+      "grad_norm": 1.7808622121810913,
+      "learning_rate": 0.0001015082785640278,
+      "loss": 2.1613,
+      "step": 8055
+    },
+    {
+      "epoch": 0.4955708661417323,
+      "grad_norm": 1.2661195993423462,
+      "learning_rate": 0.00010148894309773434,
+      "loss": 2.4838,
+      "step": 8056
+    },
+    {
+      "epoch": 0.4956323818897638,
+      "grad_norm": 1.4272936582565308,
+      "learning_rate": 0.00010146960757576272,
+      "loss": 2.791,
+      "step": 8057
+    },
+    {
+      "epoch": 0.49569389763779526,
+      "grad_norm": 1.2137830257415771,
+      "learning_rate": 0.00010145027199883592,
+      "loss": 2.5872,
+      "step": 8058
+    },
+    {
+      "epoch": 0.4957554133858268,
+      "grad_norm": 1.3280816078186035,
+      "learning_rate": 0.00010143093636767709,
+      "loss": 2.3991,
+      "step": 8059
+    },
+    {
+      "epoch": 0.4958169291338583,
+      "grad_norm": 1.4151464700698853,
+      "learning_rate": 0.00010141160068300919,
+      "loss": 3.2209,
+      "step": 8060
+    },
+    {
+      "epoch": 0.49587844488188976,
+      "grad_norm": 1.3125890493392944,
+      "learning_rate": 0.00010139226494555532,
+      "loss": 2.2077,
+      "step": 8061
+    },
+    {
+      "epoch": 0.49593996062992124,
+      "grad_norm": 1.2735986709594727,
+      "learning_rate": 0.00010137292915603846,
+      "loss": 2.4783,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4960014763779528,
+      "grad_norm": 1.4047147035598755,
+      "learning_rate": 0.00010135359331518175,
+      "loss": 1.6548,
+      "step": 8063
+    },
+    {
+      "epoch": 0.49606299212598426,
+      "grad_norm": 1.3940439224243164,
+      "learning_rate": 0.0001013342574237082,
+      "loss": 2.5073,
+      "step": 8064
+    },
+    {
+      "epoch": 0.49612450787401574,
+      "grad_norm": 1.1963787078857422,
+      "learning_rate": 0.0001013149214823408,
+      "loss": 2.5131,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4961860236220472,
+      "grad_norm": 1.1749638319015503,
+      "learning_rate": 0.00010129558549180273,
+      "loss": 3.0031,
+      "step": 8066
+    },
+    {
+      "epoch": 0.49624753937007876,
+      "grad_norm": 1.1848801374435425,
+      "learning_rate": 0.00010127624945281697,
+      "loss": 2.8179,
+      "step": 8067
+    },
+    {
+      "epoch": 0.49630905511811024,
+      "grad_norm": 1.3229222297668457,
+      "learning_rate": 0.00010125691336610659,
+      "loss": 2.2487,
+      "step": 8068
+    },
+    {
+      "epoch": 0.4963705708661417,
+      "grad_norm": 1.2249375581741333,
+      "learning_rate": 0.00010123757723239466,
+      "loss": 2.5151,
+      "step": 8069
+    },
+    {
+      "epoch": 0.4964320866141732,
+      "grad_norm": 1.1547174453735352,
+      "learning_rate": 0.00010121824105240425,
+      "loss": 2.1027,
+      "step": 8070
+    },
+    {
+      "epoch": 0.49649360236220474,
+      "grad_norm": 1.5682426691055298,
+      "learning_rate": 0.0001011989048268584,
+      "loss": 2.8013,
+      "step": 8071
+    },
+    {
+      "epoch": 0.4965551181102362,
+      "grad_norm": 1.5286115407943726,
+      "learning_rate": 0.0001011795685564802,
+      "loss": 3.0014,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4966166338582677,
+      "grad_norm": 1.3152134418487549,
+      "learning_rate": 0.00010116023224199274,
+      "loss": 1.9073,
+      "step": 8073
+    },
+    {
+      "epoch": 0.4966781496062992,
+      "grad_norm": 1.0428591966629028,
+      "learning_rate": 0.00010114089588411906,
+      "loss": 2.3348,
+      "step": 8074
+    },
+    {
+      "epoch": 0.49673966535433073,
+      "grad_norm": 1.1752588748931885,
+      "learning_rate": 0.0001011215594835822,
+      "loss": 2.8242,
+      "step": 8075
+    },
+    {
+      "epoch": 0.4968011811023622,
+      "grad_norm": 1.5039944648742676,
+      "learning_rate": 0.0001011022230411053,
+      "loss": 1.5106,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4968626968503937,
+      "grad_norm": 1.13310706615448,
+      "learning_rate": 0.0001010828865574114,
+      "loss": 2.3659,
+      "step": 8077
+    },
+    {
+      "epoch": 0.4969242125984252,
+      "grad_norm": 1.2585477828979492,
+      "learning_rate": 0.00010106355003322359,
+      "loss": 2.8931,
+      "step": 8078
+    },
+    {
+      "epoch": 0.4969857283464567,
+      "grad_norm": 1.5000816583633423,
+      "learning_rate": 0.00010104421346926493,
+      "loss": 2.4892,
+      "step": 8079
+    },
+    {
+      "epoch": 0.4970472440944882,
+      "grad_norm": 1.1580995321273804,
+      "learning_rate": 0.0001010248768662585,
+      "loss": 2.9864,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4971087598425197,
+      "grad_norm": 1.44575035572052,
+      "learning_rate": 0.00010100554022492739,
+      "loss": 3.0695,
+      "step": 8081
+    },
+    {
+      "epoch": 0.49717027559055116,
+      "grad_norm": 1.275058627128601,
+      "learning_rate": 0.00010098620354599469,
+      "loss": 2.8842,
+      "step": 8082
+    },
+    {
+      "epoch": 0.4972317913385827,
+      "grad_norm": 1.3450970649719238,
+      "learning_rate": 0.00010096686683018349,
+      "loss": 2.7183,
+      "step": 8083
+    },
+    {
+      "epoch": 0.4972933070866142,
+      "grad_norm": 1.3037446737289429,
+      "learning_rate": 0.00010094753007821683,
+      "loss": 2.1158,
+      "step": 8084
+    },
+    {
+      "epoch": 0.49735482283464566,
+      "grad_norm": 1.3971941471099854,
+      "learning_rate": 0.00010092819329081785,
+      "loss": 2.0323,
+      "step": 8085
+    },
+    {
+      "epoch": 0.49741633858267714,
+      "grad_norm": 1.170823097229004,
+      "learning_rate": 0.00010090885646870962,
+      "loss": 2.8313,
+      "step": 8086
+    },
+    {
+      "epoch": 0.4974778543307087,
+      "grad_norm": 1.2828342914581299,
+      "learning_rate": 0.00010088951961261519,
+      "loss": 2.3837,
+      "step": 8087
+    },
+    {
+      "epoch": 0.49753937007874016,
+      "grad_norm": 1.033003807067871,
+      "learning_rate": 0.00010087018272325772,
+      "loss": 3.2079,
+      "step": 8088
+    },
+    {
+      "epoch": 0.49760088582677164,
+      "grad_norm": 1.1674925088882446,
+      "learning_rate": 0.00010085084580136028,
+      "loss": 2.127,
+      "step": 8089
+    },
+    {
+      "epoch": 0.4976624015748031,
+      "grad_norm": 1.1561458110809326,
+      "learning_rate": 0.00010083150884764591,
+      "loss": 2.8165,
+      "step": 8090
+    },
+    {
+      "epoch": 0.49772391732283466,
+      "grad_norm": 1.1977254152297974,
+      "learning_rate": 0.00010081217186283777,
+      "loss": 2.4528,
+      "step": 8091
+    },
+    {
+      "epoch": 0.49778543307086615,
+      "grad_norm": 1.2996398210525513,
+      "learning_rate": 0.00010079283484765896,
+      "loss": 2.6947,
+      "step": 8092
+    },
+    {
+      "epoch": 0.4978469488188976,
+      "grad_norm": 1.3598356246948242,
+      "learning_rate": 0.00010077349780283252,
+      "loss": 2.2397,
+      "step": 8093
+    },
+    {
+      "epoch": 0.4979084645669291,
+      "grad_norm": 1.0542891025543213,
+      "learning_rate": 0.00010075416072908155,
+      "loss": 2.9024,
+      "step": 8094
+    },
+    {
+      "epoch": 0.49796998031496065,
+      "grad_norm": 1.210607647895813,
+      "learning_rate": 0.0001007348236271292,
+      "loss": 2.6772,
+      "step": 8095
+    },
+    {
+      "epoch": 0.49803149606299213,
+      "grad_norm": 1.1633234024047852,
+      "learning_rate": 0.00010071548649769856,
+      "loss": 2.731,
+      "step": 8096
+    },
+    {
+      "epoch": 0.4980930118110236,
+      "grad_norm": 1.161020278930664,
+      "learning_rate": 0.00010069614934151268,
+      "loss": 2.4648,
+      "step": 8097
+    },
+    {
+      "epoch": 0.4981545275590551,
+      "grad_norm": 1.2638401985168457,
+      "learning_rate": 0.00010067681215929471,
+      "loss": 3.4223,
+      "step": 8098
+    },
+    {
+      "epoch": 0.49821604330708663,
+      "grad_norm": 1.0872392654418945,
+      "learning_rate": 0.00010065747495176774,
+      "loss": 2.4156,
+      "step": 8099
+    },
+    {
+      "epoch": 0.4982775590551181,
+      "grad_norm": 1.0621449947357178,
+      "learning_rate": 0.00010063813771965487,
+      "loss": 1.9639,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4983390748031496,
+      "grad_norm": 1.3117069005966187,
+      "learning_rate": 0.00010061880046367922,
+      "loss": 2.8071,
+      "step": 8101
+    },
+    {
+      "epoch": 0.4984005905511811,
+      "grad_norm": 1.2320237159729004,
+      "learning_rate": 0.00010059946318456388,
+      "loss": 2.4037,
+      "step": 8102
+    },
+    {
+      "epoch": 0.4984621062992126,
+      "grad_norm": 1.2433385848999023,
+      "learning_rate": 0.00010058012588303193,
+      "loss": 2.7318,
+      "step": 8103
+    },
+    {
+      "epoch": 0.4985236220472441,
+      "grad_norm": 1.3671197891235352,
+      "learning_rate": 0.00010056078855980655,
+      "loss": 3.238,
+      "step": 8104
+    },
+    {
+      "epoch": 0.4985851377952756,
+      "grad_norm": 1.4163215160369873,
+      "learning_rate": 0.00010054145121561079,
+      "loss": 2.4441,
+      "step": 8105
+    },
+    {
+      "epoch": 0.49864665354330706,
+      "grad_norm": 1.3558136224746704,
+      "learning_rate": 0.00010052211385116779,
+      "loss": 2.0056,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4987081692913386,
+      "grad_norm": 1.1587270498275757,
+      "learning_rate": 0.00010050277646720059,
+      "loss": 2.9386,
+      "step": 8107
+    },
+    {
+      "epoch": 0.4987696850393701,
+      "grad_norm": 1.157941460609436,
+      "learning_rate": 0.0001004834390644324,
+      "loss": 2.4685,
+      "step": 8108
+    },
+    {
+      "epoch": 0.49883120078740156,
+      "grad_norm": 1.191723108291626,
+      "learning_rate": 0.00010046410164358628,
+      "loss": 2.3262,
+      "step": 8109
+    },
+    {
+      "epoch": 0.49889271653543305,
+      "grad_norm": 1.0880285501480103,
+      "learning_rate": 0.00010044476420538534,
+      "loss": 2.7842,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4989542322834646,
+      "grad_norm": 1.2752193212509155,
+      "learning_rate": 0.00010042542675055269,
+      "loss": 2.8926,
+      "step": 8111
+    },
+    {
+      "epoch": 0.49901574803149606,
+      "grad_norm": 1.1879831552505493,
+      "learning_rate": 0.00010040608927981145,
+      "loss": 2.9553,
+      "step": 8112
+    },
+    {
+      "epoch": 0.49907726377952755,
+      "grad_norm": 1.2492519617080688,
+      "learning_rate": 0.00010038675179388475,
+      "loss": 1.964,
+      "step": 8113
+    },
+    {
+      "epoch": 0.49913877952755903,
+      "grad_norm": 1.0754824876785278,
+      "learning_rate": 0.00010036741429349568,
+      "loss": 2.5241,
+      "step": 8114
+    },
+    {
+      "epoch": 0.49920029527559057,
+      "grad_norm": 1.2809557914733887,
+      "learning_rate": 0.00010034807677936739,
+      "loss": 2.6503,
+      "step": 8115
+    },
+    {
+      "epoch": 0.49926181102362205,
+      "grad_norm": 1.1516820192337036,
+      "learning_rate": 0.00010032873925222295,
+      "loss": 1.7089,
+      "step": 8116
+    },
+    {
+      "epoch": 0.49932332677165353,
+      "grad_norm": 1.2596381902694702,
+      "learning_rate": 0.00010030940171278549,
+      "loss": 2.2103,
+      "step": 8117
+    },
+    {
+      "epoch": 0.499384842519685,
+      "grad_norm": 1.2199008464813232,
+      "learning_rate": 0.00010029006416177815,
+      "loss": 2.359,
+      "step": 8118
+    },
+    {
+      "epoch": 0.49944635826771655,
+      "grad_norm": 1.279309868812561,
+      "learning_rate": 0.00010027072659992399,
+      "loss": 2.4149,
+      "step": 8119
+    },
+    {
+      "epoch": 0.49950787401574803,
+      "grad_norm": 1.2775964736938477,
+      "learning_rate": 0.0001002513890279462,
+      "loss": 2.6423,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4995693897637795,
+      "grad_norm": 1.6271353960037231,
+      "learning_rate": 0.00010023205144656784,
+      "loss": 3.2335,
+      "step": 8121
+    },
+    {
+      "epoch": 0.499630905511811,
+      "grad_norm": 1.497721552848816,
+      "learning_rate": 0.00010021271385651205,
+      "loss": 1.8957,
+      "step": 8122
+    },
+    {
+      "epoch": 0.49969242125984253,
+      "grad_norm": 1.2965353727340698,
+      "learning_rate": 0.00010019337625850196,
+      "loss": 2.2103,
+      "step": 8123
+    },
+    {
+      "epoch": 0.499753937007874,
+      "grad_norm": 1.2277823686599731,
+      "learning_rate": 0.0001001740386532607,
+      "loss": 2.5947,
+      "step": 8124
+    },
+    {
+      "epoch": 0.4998154527559055,
+      "grad_norm": 1.2674821615219116,
+      "learning_rate": 0.0001001547010415113,
+      "loss": 2.9174,
+      "step": 8125
+    },
+    {
+      "epoch": 0.499876968503937,
+      "grad_norm": 1.2761168479919434,
+      "learning_rate": 0.000100135363423977,
+      "loss": 2.5984,
+      "step": 8126
+    },
+    {
+      "epoch": 0.4999384842519685,
+      "grad_norm": 1.1656324863433838,
+      "learning_rate": 0.00010011602580138084,
+      "loss": 2.9291,
+      "step": 8127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0356035232543945,
+      "learning_rate": 0.00010009668817444598,
+      "loss": 2.5448,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 2.5490832328796387,
+      "eval_runtime": 1035.7311,
+      "eval_samples_per_second": 3.702,
+      "eval_steps_per_second": 3.702,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5000615157480315,
+      "grad_norm": 1.151626706123352,
+      "learning_rate": 0.00010007735054389549,
+      "loss": 2.5131,
+      "step": 8129
+    },
+    {
+      "epoch": 0.500123031496063,
+      "grad_norm": 1.163151741027832,
+      "learning_rate": 0.00010005801291045254,
+      "loss": 1.7692,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5001845472440944,
+      "grad_norm": 0.9838603138923645,
+      "learning_rate": 0.00010003867527484023,
+      "loss": 2.8461,
+      "step": 8131
+    },
+    {
+      "epoch": 0.500246062992126,
+      "grad_norm": 1.3035002946853638,
+      "learning_rate": 0.00010001933763778168,
+      "loss": 1.8901,
+      "step": 8132
+    },
+    {
+      "epoch": 0.5003075787401575,
+      "grad_norm": 1.2065681219100952,
+      "learning_rate": 0.0001,
+      "loss": 2.0629,
+      "step": 8133
+    },
+    {
+      "epoch": 0.500369094488189,
+      "grad_norm": 1.0714071989059448,
+      "learning_rate": 9.998066236221835e-05,
+      "loss": 2.7526,
+      "step": 8134
+    },
+    {
+      "epoch": 0.5004306102362205,
+      "grad_norm": 1.0876179933547974,
+      "learning_rate": 9.996132472515978e-05,
+      "loss": 2.6774,
+      "step": 8135
+    },
+    {
+      "epoch": 0.500492125984252,
+      "grad_norm": 1.4426108598709106,
+      "learning_rate": 9.99419870895475e-05,
+      "loss": 2.381,
+      "step": 8136
+    },
+    {
+      "epoch": 0.5005536417322834,
+      "grad_norm": 1.2925910949707031,
+      "learning_rate": 9.992264945610453e-05,
+      "loss": 2.2366,
+      "step": 8137
+    },
+    {
+      "epoch": 0.5006151574803149,
+      "grad_norm": 1.153929591178894,
+      "learning_rate": 9.990331182555408e-05,
+      "loss": 2.5958,
+      "step": 8138
+    },
+    {
+      "epoch": 0.5006766732283464,
+      "grad_norm": 1.3726719617843628,
+      "learning_rate": 9.988397419861917e-05,
+      "loss": 2.9215,
+      "step": 8139
+    },
+    {
+      "epoch": 0.500738188976378,
+      "grad_norm": 1.1279505491256714,
+      "learning_rate": 9.986463657602305e-05,
+      "loss": 2.704,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5007997047244095,
+      "grad_norm": 1.215431809425354,
+      "learning_rate": 9.984529895848872e-05,
+      "loss": 2.7301,
+      "step": 8141
+    },
+    {
+      "epoch": 0.500861220472441,
+      "grad_norm": 1.425489902496338,
+      "learning_rate": 9.982596134673936e-05,
+      "loss": 2.5414,
+      "step": 8142
+    },
+    {
+      "epoch": 0.5009227362204725,
+      "grad_norm": 1.0466691255569458,
+      "learning_rate": 9.980662374149807e-05,
+      "loss": 2.7922,
+      "step": 8143
+    },
+    {
+      "epoch": 0.5009842519685039,
+      "grad_norm": 1.1872187852859497,
+      "learning_rate": 9.978728614348797e-05,
+      "loss": 2.4403,
+      "step": 8144
+    },
+    {
+      "epoch": 0.5010457677165354,
+      "grad_norm": 1.3240453004837036,
+      "learning_rate": 9.976794855343218e-05,
+      "loss": 2.4485,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5011072834645669,
+      "grad_norm": 1.1508365869522095,
+      "learning_rate": 9.974861097205384e-05,
+      "loss": 2.862,
+      "step": 8146
+    },
+    {
+      "epoch": 0.5011687992125984,
+      "grad_norm": 1.4839322566986084,
+      "learning_rate": 9.972927340007602e-05,
+      "loss": 2.2801,
+      "step": 8147
+    },
+    {
+      "epoch": 0.50123031496063,
+      "grad_norm": 1.4541816711425781,
+      "learning_rate": 9.97099358382219e-05,
+      "loss": 2.9963,
+      "step": 8148
+    },
+    {
+      "epoch": 0.5012918307086615,
+      "grad_norm": 1.1946548223495483,
+      "learning_rate": 9.969059828721453e-05,
+      "loss": 2.4183,
+      "step": 8149
+    },
+    {
+      "epoch": 0.5013533464566929,
+      "grad_norm": 1.214625597000122,
+      "learning_rate": 9.967126074777709e-05,
+      "loss": 2.4052,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5014148622047244,
+      "grad_norm": 1.2129005193710327,
+      "learning_rate": 9.965192322063263e-05,
+      "loss": 2.1566,
+      "step": 8151
+    },
+    {
+      "epoch": 0.5014763779527559,
+      "grad_norm": 1.1817898750305176,
+      "learning_rate": 9.963258570650433e-05,
+      "loss": 2.6206,
+      "step": 8152
+    },
+    {
+      "epoch": 0.5015378937007874,
+      "grad_norm": 1.1849191188812256,
+      "learning_rate": 9.961324820611526e-05,
+      "loss": 3.1425,
+      "step": 8153
+    },
+    {
+      "epoch": 0.5015994094488189,
+      "grad_norm": 1.2403525114059448,
+      "learning_rate": 9.959391072018856e-05,
+      "loss": 1.9956,
+      "step": 8154
+    },
+    {
+      "epoch": 0.5016609251968503,
+      "grad_norm": 1.1546263694763184,
+      "learning_rate": 9.95745732494473e-05,
+      "loss": 1.978,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5017224409448819,
+      "grad_norm": 1.2749762535095215,
+      "learning_rate": 9.95552357946147e-05,
+      "loss": 2.5367,
+      "step": 8156
+    },
+    {
+      "epoch": 0.5017839566929134,
+      "grad_norm": 1.4142885208129883,
+      "learning_rate": 9.953589835641374e-05,
+      "loss": 2.5807,
+      "step": 8157
+    },
+    {
+      "epoch": 0.5018454724409449,
+      "grad_norm": 1.2706938982009888,
+      "learning_rate": 9.951656093556762e-05,
+      "loss": 2.246,
+      "step": 8158
+    },
+    {
+      "epoch": 0.5019069881889764,
+      "grad_norm": 1.4304842948913574,
+      "learning_rate": 9.94972235327994e-05,
+      "loss": 2.7974,
+      "step": 8159
+    },
+    {
+      "epoch": 0.5019685039370079,
+      "grad_norm": 1.2382725477218628,
+      "learning_rate": 9.947788614883225e-05,
+      "loss": 2.3248,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5020300196850394,
+      "grad_norm": 1.1657508611679077,
+      "learning_rate": 9.945854878438921e-05,
+      "loss": 2.3165,
+      "step": 8161
+    },
+    {
+      "epoch": 0.5020915354330708,
+      "grad_norm": 1.0771254301071167,
+      "learning_rate": 9.943921144019347e-05,
+      "loss": 3.0908,
+      "step": 8162
+    },
+    {
+      "epoch": 0.5021530511811023,
+      "grad_norm": 1.187587857246399,
+      "learning_rate": 9.941987411696806e-05,
+      "loss": 1.9756,
+      "step": 8163
+    },
+    {
+      "epoch": 0.5022145669291339,
+      "grad_norm": 1.209520697593689,
+      "learning_rate": 9.940053681543614e-05,
+      "loss": 2.3186,
+      "step": 8164
+    },
+    {
+      "epoch": 0.5022760826771654,
+      "grad_norm": 1.1334996223449707,
+      "learning_rate": 9.938119953632078e-05,
+      "loss": 2.704,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5023375984251969,
+      "grad_norm": 1.312588095664978,
+      "learning_rate": 9.936186228034515e-05,
+      "loss": 1.8986,
+      "step": 8166
+    },
+    {
+      "epoch": 0.5023991141732284,
+      "grad_norm": 1.3123010396957397,
+      "learning_rate": 9.934252504823227e-05,
+      "loss": 2.7924,
+      "step": 8167
+    },
+    {
+      "epoch": 0.5024606299212598,
+      "grad_norm": 1.6456831693649292,
+      "learning_rate": 9.932318784070531e-05,
+      "loss": 2.5978,
+      "step": 8168
+    },
+    {
+      "epoch": 0.5025221456692913,
+      "grad_norm": 1.2465269565582275,
+      "learning_rate": 9.930385065848733e-05,
+      "loss": 2.7639,
+      "step": 8169
+    },
+    {
+      "epoch": 0.5025836614173228,
+      "grad_norm": 1.1093560457229614,
+      "learning_rate": 9.928451350230146e-05,
+      "loss": 2.4976,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5026451771653543,
+      "grad_norm": 1.182767629623413,
+      "learning_rate": 9.926517637287084e-05,
+      "loss": 3.1672,
+      "step": 8171
+    },
+    {
+      "epoch": 0.5027066929133859,
+      "grad_norm": 1.8756957054138184,
+      "learning_rate": 9.924583927091847e-05,
+      "loss": 2.1599,
+      "step": 8172
+    },
+    {
+      "epoch": 0.5027682086614174,
+      "grad_norm": 1.825042724609375,
+      "learning_rate": 9.922650219716753e-05,
+      "loss": 2.2735,
+      "step": 8173
+    },
+    {
+      "epoch": 0.5028297244094488,
+      "grad_norm": 1.1511675119400024,
+      "learning_rate": 9.920716515234108e-05,
+      "loss": 2.7775,
+      "step": 8174
+    },
+    {
+      "epoch": 0.5028912401574803,
+      "grad_norm": 1.0913325548171997,
+      "learning_rate": 9.918782813716226e-05,
+      "loss": 2.5916,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5029527559055118,
+      "grad_norm": 1.4564335346221924,
+      "learning_rate": 9.916849115235411e-05,
+      "loss": 2.736,
+      "step": 8176
+    },
+    {
+      "epoch": 0.5030142716535433,
+      "grad_norm": 1.2541983127593994,
+      "learning_rate": 9.914915419863977e-05,
+      "loss": 2.9391,
+      "step": 8177
+    },
+    {
+      "epoch": 0.5030757874015748,
+      "grad_norm": 1.1941059827804565,
+      "learning_rate": 9.91298172767423e-05,
+      "loss": 2.8088,
+      "step": 8178
+    },
+    {
+      "epoch": 0.5031373031496063,
+      "grad_norm": 1.3309944868087769,
+      "learning_rate": 9.911048038738482e-05,
+      "loss": 2.5874,
+      "step": 8179
+    },
+    {
+      "epoch": 0.5031988188976378,
+      "grad_norm": 1.3883872032165527,
+      "learning_rate": 9.909114353129041e-05,
+      "loss": 2.7013,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5032603346456693,
+      "grad_norm": 1.0932588577270508,
+      "learning_rate": 9.90718067091822e-05,
+      "loss": 3.0371,
+      "step": 8181
+    },
+    {
+      "epoch": 0.5033218503937008,
+      "grad_norm": 1.4260631799697876,
+      "learning_rate": 9.90524699217832e-05,
+      "loss": 2.2714,
+      "step": 8182
+    },
+    {
+      "epoch": 0.5033833661417323,
+      "grad_norm": 1.2842984199523926,
+      "learning_rate": 9.903313316981656e-05,
+      "loss": 2.5453,
+      "step": 8183
+    },
+    {
+      "epoch": 0.5034448818897638,
+      "grad_norm": 1.29935884475708,
+      "learning_rate": 9.901379645400532e-05,
+      "loss": 2.4399,
+      "step": 8184
+    },
+    {
+      "epoch": 0.5035063976377953,
+      "grad_norm": 1.1686079502105713,
+      "learning_rate": 9.899445977507265e-05,
+      "loss": 3.0432,
+      "step": 8185
+    },
+    {
+      "epoch": 0.5035679133858267,
+      "grad_norm": 1.3754950761795044,
+      "learning_rate": 9.89751231337415e-05,
+      "loss": 2.5592,
+      "step": 8186
+    },
+    {
+      "epoch": 0.5036294291338582,
+      "grad_norm": 1.6827775239944458,
+      "learning_rate": 9.895578653073512e-05,
+      "loss": 1.9851,
+      "step": 8187
+    },
+    {
+      "epoch": 0.5036909448818898,
+      "grad_norm": 1.4279011487960815,
+      "learning_rate": 9.893644996677643e-05,
+      "loss": 2.5675,
+      "step": 8188
+    },
+    {
+      "epoch": 0.5037524606299213,
+      "grad_norm": 1.278359293937683,
+      "learning_rate": 9.891711344258862e-05,
+      "loss": 2.5336,
+      "step": 8189
+    },
+    {
+      "epoch": 0.5038139763779528,
+      "grad_norm": 1.4357550144195557,
+      "learning_rate": 9.88977769588947e-05,
+      "loss": 2.812,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5038754921259843,
+      "grad_norm": 1.1608444452285767,
+      "learning_rate": 9.887844051641782e-05,
+      "loss": 2.0122,
+      "step": 8191
+    },
+    {
+      "epoch": 0.5039370078740157,
+      "grad_norm": 1.3115960359573364,
+      "learning_rate": 9.885910411588097e-05,
+      "loss": 2.2978,
+      "step": 8192
+    },
+    {
+      "epoch": 0.5039985236220472,
+      "grad_norm": 1.0956534147262573,
+      "learning_rate": 9.883976775800727e-05,
+      "loss": 2.9479,
+      "step": 8193
+    },
+    {
+      "epoch": 0.5040600393700787,
+      "grad_norm": 1.2396721839904785,
+      "learning_rate": 9.882043144351978e-05,
+      "loss": 2.3077,
+      "step": 8194
+    },
+    {
+      "epoch": 0.5041215551181102,
+      "grad_norm": 1.2556284666061401,
+      "learning_rate": 9.880109517314162e-05,
+      "loss": 2.0281,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5041830708661418,
+      "grad_norm": 1.1051130294799805,
+      "learning_rate": 9.878175894759574e-05,
+      "loss": 2.4245,
+      "step": 8196
+    },
+    {
+      "epoch": 0.5042445866141733,
+      "grad_norm": 1.0098998546600342,
+      "learning_rate": 9.876242276760536e-05,
+      "loss": 2.4394,
+      "step": 8197
+    },
+    {
+      "epoch": 0.5043061023622047,
+      "grad_norm": 1.2440876960754395,
+      "learning_rate": 9.874308663389342e-05,
+      "loss": 2.5832,
+      "step": 8198
+    },
+    {
+      "epoch": 0.5043676181102362,
+      "grad_norm": 1.3612111806869507,
+      "learning_rate": 9.872375054718305e-05,
+      "loss": 2.2529,
+      "step": 8199
+    },
+    {
+      "epoch": 0.5044291338582677,
+      "grad_norm": 1.273559331893921,
+      "learning_rate": 9.870441450819728e-05,
+      "loss": 2.4397,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5044906496062992,
+      "grad_norm": 1.1037129163742065,
+      "learning_rate": 9.86850785176592e-05,
+      "loss": 2.7724,
+      "step": 8201
+    },
+    {
+      "epoch": 0.5045521653543307,
+      "grad_norm": 1.168248176574707,
+      "learning_rate": 9.866574257629184e-05,
+      "loss": 2.8036,
+      "step": 8202
+    },
+    {
+      "epoch": 0.5046136811023622,
+      "grad_norm": 1.270625114440918,
+      "learning_rate": 9.864640668481827e-05,
+      "loss": 1.9176,
+      "step": 8203
+    },
+    {
+      "epoch": 0.5046751968503937,
+      "grad_norm": 1.3164377212524414,
+      "learning_rate": 9.862707084396153e-05,
+      "loss": 2.5821,
+      "step": 8204
+    },
+    {
+      "epoch": 0.5047367125984252,
+      "grad_norm": 1.3298513889312744,
+      "learning_rate": 9.86077350544447e-05,
+      "loss": 1.7541,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5047982283464567,
+      "grad_norm": 1.183562994003296,
+      "learning_rate": 9.85883993169908e-05,
+      "loss": 2.7212,
+      "step": 8206
+    },
+    {
+      "epoch": 0.5048597440944882,
+      "grad_norm": 1.123175024986267,
+      "learning_rate": 9.856906363232293e-05,
+      "loss": 2.0536,
+      "step": 8207
+    },
+    {
+      "epoch": 0.5049212598425197,
+      "grad_norm": 1.0674753189086914,
+      "learning_rate": 9.85497280011641e-05,
+      "loss": 2.5658,
+      "step": 8208
+    },
+    {
+      "epoch": 0.5049827755905512,
+      "grad_norm": 1.3060060739517212,
+      "learning_rate": 9.85303924242373e-05,
+      "loss": 3.1511,
+      "step": 8209
+    },
+    {
+      "epoch": 0.5050442913385826,
+      "grad_norm": 1.2013989686965942,
+      "learning_rate": 9.85110569022657e-05,
+      "loss": 2.815,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5051058070866141,
+      "grad_norm": 1.2356771230697632,
+      "learning_rate": 9.849172143597222e-05,
+      "loss": 2.527,
+      "step": 8211
+    },
+    {
+      "epoch": 0.5051673228346457,
+      "grad_norm": 1.0845333337783813,
+      "learning_rate": 9.847238602607996e-05,
+      "loss": 3.0083,
+      "step": 8212
+    },
+    {
+      "epoch": 0.5052288385826772,
+      "grad_norm": 1.1185427904129028,
+      "learning_rate": 9.845305067331193e-05,
+      "loss": 2.8733,
+      "step": 8213
+    },
+    {
+      "epoch": 0.5052903543307087,
+      "grad_norm": 1.0894584655761719,
+      "learning_rate": 9.843371537839121e-05,
+      "loss": 2.1396,
+      "step": 8214
+    },
+    {
+      "epoch": 0.5053518700787402,
+      "grad_norm": 1.2911232709884644,
+      "learning_rate": 9.841438014204075e-05,
+      "loss": 2.635,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5054133858267716,
+      "grad_norm": 1.3363804817199707,
+      "learning_rate": 9.839504496498369e-05,
+      "loss": 2.9784,
+      "step": 8216
+    },
+    {
+      "epoch": 0.5054749015748031,
+      "grad_norm": 1.0047510862350464,
+      "learning_rate": 9.837570984794296e-05,
+      "loss": 1.14,
+      "step": 8217
+    },
+    {
+      "epoch": 0.5055364173228346,
+      "grad_norm": 1.293960452079773,
+      "learning_rate": 9.835637479164164e-05,
+      "loss": 2.5504,
+      "step": 8218
+    },
+    {
+      "epoch": 0.5055979330708661,
+      "grad_norm": 1.1193047761917114,
+      "learning_rate": 9.833703979680274e-05,
+      "loss": 2.9558,
+      "step": 8219
+    },
+    {
+      "epoch": 0.5056594488188977,
+      "grad_norm": 1.4105498790740967,
+      "learning_rate": 9.83177048641493e-05,
+      "loss": 2.0412,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5057209645669292,
+      "grad_norm": 1.4895271062850952,
+      "learning_rate": 9.829836999440427e-05,
+      "loss": 2.395,
+      "step": 8221
+    },
+    {
+      "epoch": 0.5057824803149606,
+      "grad_norm": 1.4819875955581665,
+      "learning_rate": 9.827903518829073e-05,
+      "loss": 3.2384,
+      "step": 8222
+    },
+    {
+      "epoch": 0.5058439960629921,
+      "grad_norm": 1.0549601316452026,
+      "learning_rate": 9.825970044653168e-05,
+      "loss": 2.8894,
+      "step": 8223
+    },
+    {
+      "epoch": 0.5059055118110236,
+      "grad_norm": 1.3861445188522339,
+      "learning_rate": 9.824036576985013e-05,
+      "loss": 2.2243,
+      "step": 8224
+    },
+    {
+      "epoch": 0.5059670275590551,
+      "grad_norm": 1.5474681854248047,
+      "learning_rate": 9.822103115896908e-05,
+      "loss": 2.7923,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5060285433070866,
+      "grad_norm": 1.4457621574401855,
+      "learning_rate": 9.820169661461158e-05,
+      "loss": 2.2778,
+      "step": 8226
+    },
+    {
+      "epoch": 0.5060900590551181,
+      "grad_norm": 1.563105821609497,
+      "learning_rate": 9.818236213750056e-05,
+      "loss": 2.9542,
+      "step": 8227
+    },
+    {
+      "epoch": 0.5061515748031497,
+      "grad_norm": 1.5135773420333862,
+      "learning_rate": 9.816302772835908e-05,
+      "loss": 1.7652,
+      "step": 8228
+    },
+    {
+      "epoch": 0.5062130905511811,
+      "grad_norm": 1.3420755863189697,
+      "learning_rate": 9.814369338791009e-05,
+      "loss": 2.6548,
+      "step": 8229
+    },
+    {
+      "epoch": 0.5062746062992126,
+      "grad_norm": 1.3379322290420532,
+      "learning_rate": 9.812435911687667e-05,
+      "loss": 3.002,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5063361220472441,
+      "grad_norm": 1.1443718671798706,
+      "learning_rate": 9.81050249159817e-05,
+      "loss": 1.8304,
+      "step": 8231
+    },
+    {
+      "epoch": 0.5063976377952756,
+      "grad_norm": 1.1285450458526611,
+      "learning_rate": 9.808569078594829e-05,
+      "loss": 2.7694,
+      "step": 8232
+    },
+    {
+      "epoch": 0.5064591535433071,
+      "grad_norm": 1.1707381010055542,
+      "learning_rate": 9.806635672749933e-05,
+      "loss": 2.1892,
+      "step": 8233
+    },
+    {
+      "epoch": 0.5065206692913385,
+      "grad_norm": 1.1144012212753296,
+      "learning_rate": 9.804702274135786e-05,
+      "loss": 2.602,
+      "step": 8234
+    },
+    {
+      "epoch": 0.50658218503937,
+      "grad_norm": 1.1850855350494385,
+      "learning_rate": 9.802768882824683e-05,
+      "loss": 2.5967,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5066437007874016,
+      "grad_norm": 0.9958007335662842,
+      "learning_rate": 9.800835498888928e-05,
+      "loss": 3.0141,
+      "step": 8236
+    },
+    {
+      "epoch": 0.5067052165354331,
+      "grad_norm": 1.2061309814453125,
+      "learning_rate": 9.798902122400812e-05,
+      "loss": 2.7552,
+      "step": 8237
+    },
+    {
+      "epoch": 0.5067667322834646,
+      "grad_norm": 1.0050749778747559,
+      "learning_rate": 9.796968753432636e-05,
+      "loss": 2.8307,
+      "step": 8238
+    },
+    {
+      "epoch": 0.5068282480314961,
+      "grad_norm": 1.1546040773391724,
+      "learning_rate": 9.795035392056695e-05,
+      "loss": 2.6547,
+      "step": 8239
+    },
+    {
+      "epoch": 0.5068897637795275,
+      "grad_norm": 1.100258231163025,
+      "learning_rate": 9.79310203834529e-05,
+      "loss": 2.7101,
+      "step": 8240
+    },
+    {
+      "epoch": 0.506951279527559,
+      "grad_norm": 1.3437423706054688,
+      "learning_rate": 9.791168692370712e-05,
+      "loss": 2.0113,
+      "step": 8241
+    },
+    {
+      "epoch": 0.5070127952755905,
+      "grad_norm": 1.2487047910690308,
+      "learning_rate": 9.789235354205265e-05,
+      "loss": 2.3864,
+      "step": 8242
+    },
+    {
+      "epoch": 0.507074311023622,
+      "grad_norm": 1.1158820390701294,
+      "learning_rate": 9.787302023921235e-05,
+      "loss": 2.5119,
+      "step": 8243
+    },
+    {
+      "epoch": 0.5071358267716536,
+      "grad_norm": 2.3030762672424316,
+      "learning_rate": 9.785368701590928e-05,
+      "loss": 2.1374,
+      "step": 8244
+    },
+    {
+      "epoch": 0.5071973425196851,
+      "grad_norm": 1.1044385433197021,
+      "learning_rate": 9.783435387286635e-05,
+      "loss": 2.817,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5072588582677166,
+      "grad_norm": 1.1762449741363525,
+      "learning_rate": 9.781502081080647e-05,
+      "loss": 2.6981,
+      "step": 8246
+    },
+    {
+      "epoch": 0.507320374015748,
+      "grad_norm": 1.222424030303955,
+      "learning_rate": 9.779568783045265e-05,
+      "loss": 2.4885,
+      "step": 8247
+    },
+    {
+      "epoch": 0.5073818897637795,
+      "grad_norm": 1.1241434812545776,
+      "learning_rate": 9.777635493252779e-05,
+      "loss": 2.5205,
+      "step": 8248
+    },
+    {
+      "epoch": 0.507443405511811,
+      "grad_norm": 1.099979043006897,
+      "learning_rate": 9.775702211775489e-05,
+      "loss": 2.36,
+      "step": 8249
+    },
+    {
+      "epoch": 0.5075049212598425,
+      "grad_norm": 1.2585588693618774,
+      "learning_rate": 9.773768938685681e-05,
+      "loss": 2.3802,
+      "step": 8250
+    },
+    {
+      "epoch": 0.507566437007874,
+      "grad_norm": 1.1879340410232544,
+      "learning_rate": 9.77183567405566e-05,
+      "loss": 1.9232,
+      "step": 8251
+    },
+    {
+      "epoch": 0.5076279527559056,
+      "grad_norm": 1.1214536428451538,
+      "learning_rate": 9.769902417957707e-05,
+      "loss": 2.5404,
+      "step": 8252
+    },
+    {
+      "epoch": 0.507689468503937,
+      "grad_norm": 1.3413785696029663,
+      "learning_rate": 9.767969170464123e-05,
+      "loss": 2.7498,
+      "step": 8253
+    },
+    {
+      "epoch": 0.5077509842519685,
+      "grad_norm": 1.242747187614441,
+      "learning_rate": 9.766035931647195e-05,
+      "loss": 2.9952,
+      "step": 8254
+    },
+    {
+      "epoch": 0.5078125,
+      "grad_norm": 1.034842848777771,
+      "learning_rate": 9.764102701579224e-05,
+      "loss": 2.6782,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5078740157480315,
+      "grad_norm": 1.2424741983413696,
+      "learning_rate": 9.762169480332492e-05,
+      "loss": 2.6513,
+      "step": 8256
+    },
+    {
+      "epoch": 0.507935531496063,
+      "grad_norm": 1.1681605577468872,
+      "learning_rate": 9.760236267979297e-05,
+      "loss": 2.9224,
+      "step": 8257
+    },
+    {
+      "epoch": 0.5079970472440944,
+      "grad_norm": 1.0589063167572021,
+      "learning_rate": 9.758303064591926e-05,
+      "loss": 2.7005,
+      "step": 8258
+    },
+    {
+      "epoch": 0.508058562992126,
+      "grad_norm": 1.1972293853759766,
+      "learning_rate": 9.756369870242677e-05,
+      "loss": 2.8939,
+      "step": 8259
+    },
+    {
+      "epoch": 0.5081200787401575,
+      "grad_norm": 1.082395076751709,
+      "learning_rate": 9.754436685003833e-05,
+      "loss": 2.8107,
+      "step": 8260
+    },
+    {
+      "epoch": 0.508181594488189,
+      "grad_norm": 1.1019303798675537,
+      "learning_rate": 9.752503508947691e-05,
+      "loss": 2.9236,
+      "step": 8261
+    },
+    {
+      "epoch": 0.5082431102362205,
+      "grad_norm": 1.321231722831726,
+      "learning_rate": 9.750570342146534e-05,
+      "loss": 2.6732,
+      "step": 8262
+    },
+    {
+      "epoch": 0.508304625984252,
+      "grad_norm": 1.7180914878845215,
+      "learning_rate": 9.748637184672658e-05,
+      "loss": 2.558,
+      "step": 8263
+    },
+    {
+      "epoch": 0.5083661417322834,
+      "grad_norm": 1.2647619247436523,
+      "learning_rate": 9.746704036598347e-05,
+      "loss": 2.5049,
+      "step": 8264
+    },
+    {
+      "epoch": 0.5084276574803149,
+      "grad_norm": 1.1217907667160034,
+      "learning_rate": 9.744770897995897e-05,
+      "loss": 2.6489,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5084891732283464,
+      "grad_norm": 1.1405553817749023,
+      "learning_rate": 9.742837768937586e-05,
+      "loss": 2.7796,
+      "step": 8266
+    },
+    {
+      "epoch": 0.508550688976378,
+      "grad_norm": 1.4073686599731445,
+      "learning_rate": 9.740904649495715e-05,
+      "loss": 2.4181,
+      "step": 8267
+    },
+    {
+      "epoch": 0.5086122047244095,
+      "grad_norm": 1.2061423063278198,
+      "learning_rate": 9.738971539742561e-05,
+      "loss": 2.4634,
+      "step": 8268
+    },
+    {
+      "epoch": 0.508673720472441,
+      "grad_norm": 1.2314543724060059,
+      "learning_rate": 9.737038439750418e-05,
+      "loss": 2.3523,
+      "step": 8269
+    },
+    {
+      "epoch": 0.5087352362204725,
+      "grad_norm": 1.2566312551498413,
+      "learning_rate": 9.735105349591567e-05,
+      "loss": 2.9506,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5087967519685039,
+      "grad_norm": 1.2187609672546387,
+      "learning_rate": 9.733172269338306e-05,
+      "loss": 2.7641,
+      "step": 8271
+    },
+    {
+      "epoch": 0.5088582677165354,
+      "grad_norm": 1.0804346799850464,
+      "learning_rate": 9.731239199062909e-05,
+      "loss": 2.8846,
+      "step": 8272
+    },
+    {
+      "epoch": 0.5089197834645669,
+      "grad_norm": 1.3811393976211548,
+      "learning_rate": 9.72930613883767e-05,
+      "loss": 2.8608,
+      "step": 8273
+    },
+    {
+      "epoch": 0.5089812992125984,
+      "grad_norm": 1.264640212059021,
+      "learning_rate": 9.727373088734868e-05,
+      "loss": 2.8344,
+      "step": 8274
+    },
+    {
+      "epoch": 0.50904281496063,
+      "grad_norm": 1.7371588945388794,
+      "learning_rate": 9.725440048826796e-05,
+      "loss": 2.0939,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5091043307086615,
+      "grad_norm": 1.5706932544708252,
+      "learning_rate": 9.723507019185733e-05,
+      "loss": 2.3156,
+      "step": 8276
+    },
+    {
+      "epoch": 0.5091658464566929,
+      "grad_norm": 1.101751446723938,
+      "learning_rate": 9.721573999883968e-05,
+      "loss": 3.3836,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5092273622047244,
+      "grad_norm": 1.256461262702942,
+      "learning_rate": 9.719640990993782e-05,
+      "loss": 2.9956,
+      "step": 8278
+    },
+    {
+      "epoch": 0.5092888779527559,
+      "grad_norm": 1.3670318126678467,
+      "learning_rate": 9.717707992587459e-05,
+      "loss": 2.3688,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5093503937007874,
+      "grad_norm": 1.3191993236541748,
+      "learning_rate": 9.715775004737282e-05,
+      "loss": 2.3789,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5094119094488189,
+      "grad_norm": 1.2273883819580078,
+      "learning_rate": 9.713842027515538e-05,
+      "loss": 2.4261,
+      "step": 8281
+    },
+    {
+      "epoch": 0.5094734251968503,
+      "grad_norm": 1.4858580827713013,
+      "learning_rate": 9.711909060994506e-05,
+      "loss": 2.8606,
+      "step": 8282
+    },
+    {
+      "epoch": 0.5095349409448819,
+      "grad_norm": 1.1953883171081543,
+      "learning_rate": 9.709976105246466e-05,
+      "loss": 2.6939,
+      "step": 8283
+    },
+    {
+      "epoch": 0.5095964566929134,
+      "grad_norm": 1.1841386556625366,
+      "learning_rate": 9.708043160343704e-05,
+      "loss": 2.2907,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5096579724409449,
+      "grad_norm": 1.2318238019943237,
+      "learning_rate": 9.706110226358498e-05,
+      "loss": 2.6431,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5097194881889764,
+      "grad_norm": 1.2575410604476929,
+      "learning_rate": 9.704177303363135e-05,
+      "loss": 2.1619,
+      "step": 8286
+    },
+    {
+      "epoch": 0.5097810039370079,
+      "grad_norm": 1.1212905645370483,
+      "learning_rate": 9.702244391429888e-05,
+      "loss": 2.8778,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5098425196850394,
+      "grad_norm": 1.6685899496078491,
+      "learning_rate": 9.700311490631041e-05,
+      "loss": 1.9617,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5099040354330708,
+      "grad_norm": 1.4280076026916504,
+      "learning_rate": 9.698378601038872e-05,
+      "loss": 2.2429,
+      "step": 8289
+    },
+    {
+      "epoch": 0.5099655511811023,
+      "grad_norm": 1.322075605392456,
+      "learning_rate": 9.696445722725665e-05,
+      "loss": 2.8789,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5100270669291339,
+      "grad_norm": 1.91207754611969,
+      "learning_rate": 9.694512855763692e-05,
+      "loss": 2.7017,
+      "step": 8291
+    },
+    {
+      "epoch": 0.5100885826771654,
+      "grad_norm": 2.077484607696533,
+      "learning_rate": 9.692580000225236e-05,
+      "loss": 1.9991,
+      "step": 8292
+    },
+    {
+      "epoch": 0.5101500984251969,
+      "grad_norm": 1.2217861413955688,
+      "learning_rate": 9.690647156182572e-05,
+      "loss": 2.4345,
+      "step": 8293
+    },
+    {
+      "epoch": 0.5102116141732284,
+      "grad_norm": 1.031118631362915,
+      "learning_rate": 9.688714323707981e-05,
+      "loss": 2.9962,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5102731299212598,
+      "grad_norm": 1.1225663423538208,
+      "learning_rate": 9.686781502873738e-05,
+      "loss": 2.606,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5103346456692913,
+      "grad_norm": 1.0672177076339722,
+      "learning_rate": 9.684848693752123e-05,
+      "loss": 2.769,
+      "step": 8296
+    },
+    {
+      "epoch": 0.5103961614173228,
+      "grad_norm": 1.2578306198120117,
+      "learning_rate": 9.682915896415406e-05,
+      "loss": 2.6841,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5104576771653543,
+      "grad_norm": 1.4862686395645142,
+      "learning_rate": 9.680983110935868e-05,
+      "loss": 1.8898,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5105191929133859,
+      "grad_norm": 1.2153643369674683,
+      "learning_rate": 9.67905033738578e-05,
+      "loss": 3.1742,
+      "step": 8299
+    },
+    {
+      "epoch": 0.5105807086614174,
+      "grad_norm": 2.0164146423339844,
+      "learning_rate": 9.677117575837426e-05,
+      "loss": 2.3725,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5106422244094488,
+      "grad_norm": 1.3921352624893188,
+      "learning_rate": 9.675184826363067e-05,
+      "loss": 3.0913,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5107037401574803,
+      "grad_norm": 1.3516830205917358,
+      "learning_rate": 9.673252089034989e-05,
+      "loss": 2.3133,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5107652559055118,
+      "grad_norm": 1.0958331823349,
+      "learning_rate": 9.671319363925457e-05,
+      "loss": 2.6602,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5108267716535433,
+      "grad_norm": 1.2328940629959106,
+      "learning_rate": 9.669386651106752e-05,
+      "loss": 2.7942,
+      "step": 8304
+    },
+    {
+      "epoch": 0.5108882874015748,
+      "grad_norm": 1.441518783569336,
+      "learning_rate": 9.66745395065114e-05,
+      "loss": 2.356,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5109498031496063,
+      "grad_norm": 1.294377088546753,
+      "learning_rate": 9.665521262630899e-05,
+      "loss": 2.6406,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5110113188976378,
+      "grad_norm": 1.127592921257019,
+      "learning_rate": 9.663588587118294e-05,
+      "loss": 2.9736,
+      "step": 8307
+    },
+    {
+      "epoch": 0.5110728346456693,
+      "grad_norm": 1.2097939252853394,
+      "learning_rate": 9.661655924185603e-05,
+      "loss": 2.8633,
+      "step": 8308
+    },
+    {
+      "epoch": 0.5111343503937008,
+      "grad_norm": 1.1124361753463745,
+      "learning_rate": 9.65972327390509e-05,
+      "loss": 2.8511,
+      "step": 8309
+    },
+    {
+      "epoch": 0.5111958661417323,
+      "grad_norm": 1.1703095436096191,
+      "learning_rate": 9.657790636349032e-05,
+      "loss": 1.8177,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5112573818897638,
+      "grad_norm": 1.0999525785446167,
+      "learning_rate": 9.655858011589695e-05,
+      "loss": 2.8084,
+      "step": 8311
+    },
+    {
+      "epoch": 0.5113188976377953,
+      "grad_norm": 1.351492166519165,
+      "learning_rate": 9.653925399699351e-05,
+      "loss": 2.7862,
+      "step": 8312
+    },
+    {
+      "epoch": 0.5113804133858267,
+      "grad_norm": 1.2808340787887573,
+      "learning_rate": 9.651992800750266e-05,
+      "loss": 2.3833,
+      "step": 8313
+    },
+    {
+      "epoch": 0.5114419291338582,
+      "grad_norm": 1.3594101667404175,
+      "learning_rate": 9.65006021481471e-05,
+      "loss": 2.6631,
+      "step": 8314
+    },
+    {
+      "epoch": 0.5115034448818898,
+      "grad_norm": 1.3687288761138916,
+      "learning_rate": 9.648127641964949e-05,
+      "loss": 2.6728,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5115649606299213,
+      "grad_norm": 1.2315653562545776,
+      "learning_rate": 9.646195082273257e-05,
+      "loss": 2.842,
+      "step": 8316
+    },
+    {
+      "epoch": 0.5116264763779528,
+      "grad_norm": 1.308694839477539,
+      "learning_rate": 9.644262535811894e-05,
+      "loss": 2.7087,
+      "step": 8317
+    },
+    {
+      "epoch": 0.5116879921259843,
+      "grad_norm": 1.3265371322631836,
+      "learning_rate": 9.642330002653124e-05,
+      "loss": 2.284,
+      "step": 8318
+    },
+    {
+      "epoch": 0.5117495078740157,
+      "grad_norm": 2.0234012603759766,
+      "learning_rate": 9.640397482869225e-05,
+      "loss": 3.053,
+      "step": 8319
+    },
+    {
+      "epoch": 0.5118110236220472,
+      "grad_norm": 1.3492143154144287,
+      "learning_rate": 9.638464976532446e-05,
+      "loss": 2.138,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5118725393700787,
+      "grad_norm": 1.2306753396987915,
+      "learning_rate": 9.636532483715068e-05,
+      "loss": 2.0124,
+      "step": 8321
+    },
+    {
+      "epoch": 0.5119340551181102,
+      "grad_norm": 1.3044737577438354,
+      "learning_rate": 9.634600004489345e-05,
+      "loss": 2.8921,
+      "step": 8322
+    },
+    {
+      "epoch": 0.5119955708661418,
+      "grad_norm": 1.3941199779510498,
+      "learning_rate": 9.632667538927546e-05,
+      "loss": 2.8653,
+      "step": 8323
+    },
+    {
+      "epoch": 0.5120570866141733,
+      "grad_norm": 1.337144374847412,
+      "learning_rate": 9.63073508710193e-05,
+      "loss": 2.5311,
+      "step": 8324
+    },
+    {
+      "epoch": 0.5121186023622047,
+      "grad_norm": 1.1195112466812134,
+      "learning_rate": 9.628802649084767e-05,
+      "loss": 2.8029,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5121801181102362,
+      "grad_norm": 1.2608423233032227,
+      "learning_rate": 9.626870224948311e-05,
+      "loss": 2.729,
+      "step": 8326
+    },
+    {
+      "epoch": 0.5122416338582677,
+      "grad_norm": 1.1882933378219604,
+      "learning_rate": 9.62493781476483e-05,
+      "loss": 2.4503,
+      "step": 8327
+    },
+    {
+      "epoch": 0.5123031496062992,
+      "grad_norm": 1.2347484827041626,
+      "learning_rate": 9.623005418606582e-05,
+      "loss": 2.1584,
+      "step": 8328
+    },
+    {
+      "epoch": 0.5123646653543307,
+      "grad_norm": 1.1149687767028809,
+      "learning_rate": 9.621073036545831e-05,
+      "loss": 2.7443,
+      "step": 8329
+    },
+    {
+      "epoch": 0.5124261811023622,
+      "grad_norm": 1.2456820011138916,
+      "learning_rate": 9.619140668654833e-05,
+      "loss": 2.4788,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5124876968503937,
+      "grad_norm": 1.1861653327941895,
+      "learning_rate": 9.617208315005855e-05,
+      "loss": 2.8152,
+      "step": 8331
+    },
+    {
+      "epoch": 0.5125492125984252,
+      "grad_norm": 1.3436360359191895,
+      "learning_rate": 9.615275975671147e-05,
+      "loss": 3.0863,
+      "step": 8332
+    },
+    {
+      "epoch": 0.5126107283464567,
+      "grad_norm": 1.1888959407806396,
+      "learning_rate": 9.613343650722973e-05,
+      "loss": 3.1787,
+      "step": 8333
+    },
+    {
+      "epoch": 0.5126722440944882,
+      "grad_norm": 1.2109920978546143,
+      "learning_rate": 9.61141134023359e-05,
+      "loss": 2.8286,
+      "step": 8334
+    },
+    {
+      "epoch": 0.5127337598425197,
+      "grad_norm": 1.1453206539154053,
+      "learning_rate": 9.60947904427526e-05,
+      "loss": 3.0614,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5127952755905512,
+      "grad_norm": 1.4183378219604492,
+      "learning_rate": 9.607546762920229e-05,
+      "loss": 2.9301,
+      "step": 8336
+    },
+    {
+      "epoch": 0.5128567913385826,
+      "grad_norm": 1.0880351066589355,
+      "learning_rate": 9.605614496240768e-05,
+      "loss": 2.684,
+      "step": 8337
+    },
+    {
+      "epoch": 0.5129183070866141,
+      "grad_norm": 1.3476917743682861,
+      "learning_rate": 9.603682244309121e-05,
+      "loss": 2.9015,
+      "step": 8338
+    },
+    {
+      "epoch": 0.5129798228346457,
+      "grad_norm": 1.2906596660614014,
+      "learning_rate": 9.601750007197551e-05,
+      "loss": 2.6093,
+      "step": 8339
+    },
+    {
+      "epoch": 0.5130413385826772,
+      "grad_norm": 1.075652003288269,
+      "learning_rate": 9.599817784978306e-05,
+      "loss": 2.4942,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5131028543307087,
+      "grad_norm": 1.3694764375686646,
+      "learning_rate": 9.59788557772365e-05,
+      "loss": 2.3333,
+      "step": 8341
+    },
+    {
+      "epoch": 0.5131643700787402,
+      "grad_norm": 0.9970013499259949,
+      "learning_rate": 9.595953385505827e-05,
+      "loss": 2.4024,
+      "step": 8342
+    },
+    {
+      "epoch": 0.5132258858267716,
+      "grad_norm": 1.390303373336792,
+      "learning_rate": 9.594021208397097e-05,
+      "loss": 2.5837,
+      "step": 8343
+    },
+    {
+      "epoch": 0.5132874015748031,
+      "grad_norm": 1.2287286520004272,
+      "learning_rate": 9.592089046469707e-05,
+      "loss": 2.6251,
+      "step": 8344
+    },
+    {
+      "epoch": 0.5133489173228346,
+      "grad_norm": 1.1232932806015015,
+      "learning_rate": 9.590156899795916e-05,
+      "loss": 2.1708,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5134104330708661,
+      "grad_norm": 1.289667010307312,
+      "learning_rate": 9.588224768447966e-05,
+      "loss": 2.6276,
+      "step": 8346
+    },
+    {
+      "epoch": 0.5134719488188977,
+      "grad_norm": 1.1993868350982666,
+      "learning_rate": 9.586292652498119e-05,
+      "loss": 2.6062,
+      "step": 8347
+    },
+    {
+      "epoch": 0.5135334645669292,
+      "grad_norm": 1.1353377103805542,
+      "learning_rate": 9.584360552018617e-05,
+      "loss": 2.7378,
+      "step": 8348
+    },
+    {
+      "epoch": 0.5135949803149606,
+      "grad_norm": 1.41996431350708,
+      "learning_rate": 9.582428467081713e-05,
+      "loss": 2.2984,
+      "step": 8349
+    },
+    {
+      "epoch": 0.5136564960629921,
+      "grad_norm": 1.244768738746643,
+      "learning_rate": 9.580496397759655e-05,
+      "loss": 2.5005,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5137180118110236,
+      "grad_norm": 1.176455020904541,
+      "learning_rate": 9.578564344124696e-05,
+      "loss": 3.1444,
+      "step": 8351
+    },
+    {
+      "epoch": 0.5137795275590551,
+      "grad_norm": 1.2491726875305176,
+      "learning_rate": 9.576632306249079e-05,
+      "loss": 2.626,
+      "step": 8352
+    },
+    {
+      "epoch": 0.5138410433070866,
+      "grad_norm": 1.1721012592315674,
+      "learning_rate": 9.574700284205052e-05,
+      "loss": 3.1855,
+      "step": 8353
+    },
+    {
+      "epoch": 0.5139025590551181,
+      "grad_norm": 1.4534269571304321,
+      "learning_rate": 9.572768278064866e-05,
+      "loss": 2.4923,
+      "step": 8354
+    },
+    {
+      "epoch": 0.5139640748031497,
+      "grad_norm": 1.3617291450500488,
+      "learning_rate": 9.570836287900758e-05,
+      "loss": 2.7433,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5140255905511811,
+      "grad_norm": 1.2044216394424438,
+      "learning_rate": 9.568904313784986e-05,
+      "loss": 2.4935,
+      "step": 8356
+    },
+    {
+      "epoch": 0.5140871062992126,
+      "grad_norm": 1.1603797674179077,
+      "learning_rate": 9.566972355789782e-05,
+      "loss": 1.7764,
+      "step": 8357
+    },
+    {
+      "epoch": 0.5141486220472441,
+      "grad_norm": 1.1463041305541992,
+      "learning_rate": 9.565040413987403e-05,
+      "loss": 2.1221,
+      "step": 8358
+    },
+    {
+      "epoch": 0.5142101377952756,
+      "grad_norm": 1.6226743459701538,
+      "learning_rate": 9.563108488450083e-05,
+      "loss": 2.6505,
+      "step": 8359
+    },
+    {
+      "epoch": 0.5142716535433071,
+      "grad_norm": 1.2645647525787354,
+      "learning_rate": 9.561176579250075e-05,
+      "loss": 2.3891,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5143331692913385,
+      "grad_norm": 1.3753032684326172,
+      "learning_rate": 9.559244686459611e-05,
+      "loss": 2.1569,
+      "step": 8361
+    },
+    {
+      "epoch": 0.51439468503937,
+      "grad_norm": 1.1386163234710693,
+      "learning_rate": 9.55731281015094e-05,
+      "loss": 2.7079,
+      "step": 8362
+    },
+    {
+      "epoch": 0.5144562007874016,
+      "grad_norm": 1.2073734998703003,
+      "learning_rate": 9.5553809503963e-05,
+      "loss": 2.6479,
+      "step": 8363
+    },
+    {
+      "epoch": 0.5145177165354331,
+      "grad_norm": 2.165187358856201,
+      "learning_rate": 9.553449107267934e-05,
+      "loss": 2.2787,
+      "step": 8364
+    },
+    {
+      "epoch": 0.5145792322834646,
+      "grad_norm": 1.4761152267456055,
+      "learning_rate": 9.55151728083808e-05,
+      "loss": 1.9647,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5146407480314961,
+      "grad_norm": 1.4367363452911377,
+      "learning_rate": 9.549585471178983e-05,
+      "loss": 2.0361,
+      "step": 8366
+    },
+    {
+      "epoch": 0.5147022637795275,
+      "grad_norm": 1.601345181465149,
+      "learning_rate": 9.547653678362874e-05,
+      "loss": 1.9017,
+      "step": 8367
+    },
+    {
+      "epoch": 0.514763779527559,
+      "grad_norm": 1.1645532846450806,
+      "learning_rate": 9.545721902461996e-05,
+      "loss": 2.8126,
+      "step": 8368
+    },
+    {
+      "epoch": 0.5148252952755905,
+      "grad_norm": 1.2890074253082275,
+      "learning_rate": 9.543790143548585e-05,
+      "loss": 2.5191,
+      "step": 8369
+    },
+    {
+      "epoch": 0.514886811023622,
+      "grad_norm": 1.2488257884979248,
+      "learning_rate": 9.541858401694883e-05,
+      "loss": 3.0314,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5149483267716536,
+      "grad_norm": 1.3130491971969604,
+      "learning_rate": 9.539926676973118e-05,
+      "loss": 2.1884,
+      "step": 8371
+    },
+    {
+      "epoch": 0.5150098425196851,
+      "grad_norm": 1.3428913354873657,
+      "learning_rate": 9.53799496945553e-05,
+      "loss": 2.1661,
+      "step": 8372
+    },
+    {
+      "epoch": 0.5150713582677166,
+      "grad_norm": 1.1496449708938599,
+      "learning_rate": 9.536063279214354e-05,
+      "loss": 2.7763,
+      "step": 8373
+    },
+    {
+      "epoch": 0.515132874015748,
+      "grad_norm": 1.4637672901153564,
+      "learning_rate": 9.534131606321826e-05,
+      "loss": 2.7647,
+      "step": 8374
+    },
+    {
+      "epoch": 0.5151943897637795,
+      "grad_norm": 1.1782386302947998,
+      "learning_rate": 9.532199950850175e-05,
+      "loss": 2.1645,
+      "step": 8375
+    },
+    {
+      "epoch": 0.515255905511811,
+      "grad_norm": 1.1182315349578857,
+      "learning_rate": 9.530268312871641e-05,
+      "loss": 3.2692,
+      "step": 8376
+    },
+    {
+      "epoch": 0.5153174212598425,
+      "grad_norm": 1.1480664014816284,
+      "learning_rate": 9.528336692458451e-05,
+      "loss": 2.5605,
+      "step": 8377
+    },
+    {
+      "epoch": 0.515378937007874,
+      "grad_norm": 1.028917908668518,
+      "learning_rate": 9.526405089682839e-05,
+      "loss": 3.1313,
+      "step": 8378
+    },
+    {
+      "epoch": 0.5154404527559056,
+      "grad_norm": 1.4285340309143066,
+      "learning_rate": 9.524473504617035e-05,
+      "loss": 1.9159,
+      "step": 8379
+    },
+    {
+      "epoch": 0.515501968503937,
+      "grad_norm": 1.5165057182312012,
+      "learning_rate": 9.522541937333273e-05,
+      "loss": 2.8855,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5155634842519685,
+      "grad_norm": 1.2188700437545776,
+      "learning_rate": 9.520610387903775e-05,
+      "loss": 3.0389,
+      "step": 8381
+    },
+    {
+      "epoch": 0.515625,
+      "grad_norm": 1.4431930780410767,
+      "learning_rate": 9.518678856400781e-05,
+      "loss": 2.8342,
+      "step": 8382
+    },
+    {
+      "epoch": 0.5156865157480315,
+      "grad_norm": 1.1217416524887085,
+      "learning_rate": 9.51674734289651e-05,
+      "loss": 2.7395,
+      "step": 8383
+    },
+    {
+      "epoch": 0.515748031496063,
+      "grad_norm": 1.3188362121582031,
+      "learning_rate": 9.514815847463195e-05,
+      "loss": 2.6765,
+      "step": 8384
+    },
+    {
+      "epoch": 0.5158095472440944,
+      "grad_norm": 2.1636056900024414,
+      "learning_rate": 9.51288437017306e-05,
+      "loss": 2.7843,
+      "step": 8385
+    },
+    {
+      "epoch": 0.515871062992126,
+      "grad_norm": 1.448089838027954,
+      "learning_rate": 9.51095291109834e-05,
+      "loss": 3.088,
+      "step": 8386
+    },
+    {
+      "epoch": 0.5159325787401575,
+      "grad_norm": 1.3776205778121948,
+      "learning_rate": 9.509021470311247e-05,
+      "loss": 2.3374,
+      "step": 8387
+    },
+    {
+      "epoch": 0.515994094488189,
+      "grad_norm": 1.2051541805267334,
+      "learning_rate": 9.507090047884016e-05,
+      "loss": 2.9167,
+      "step": 8388
+    },
+    {
+      "epoch": 0.5160556102362205,
+      "grad_norm": 1.2855243682861328,
+      "learning_rate": 9.505158643888867e-05,
+      "loss": 2.5915,
+      "step": 8389
+    },
+    {
+      "epoch": 0.516117125984252,
+      "grad_norm": 1.4939861297607422,
+      "learning_rate": 9.503227258398027e-05,
+      "loss": 2.7322,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5161786417322834,
+      "grad_norm": 1.4462883472442627,
+      "learning_rate": 9.50129589148372e-05,
+      "loss": 2.6104,
+      "step": 8391
+    },
+    {
+      "epoch": 0.5162401574803149,
+      "grad_norm": 1.289072871208191,
+      "learning_rate": 9.499364543218162e-05,
+      "loss": 2.3665,
+      "step": 8392
+    },
+    {
+      "epoch": 0.5163016732283464,
+      "grad_norm": 1.2538114786148071,
+      "learning_rate": 9.49743321367358e-05,
+      "loss": 2.4174,
+      "step": 8393
+    },
+    {
+      "epoch": 0.516363188976378,
+      "grad_norm": 1.2645320892333984,
+      "learning_rate": 9.495501902922191e-05,
+      "loss": 2.2712,
+      "step": 8394
+    },
+    {
+      "epoch": 0.5164247047244095,
+      "grad_norm": 0.9368350505828857,
+      "learning_rate": 9.493570611036222e-05,
+      "loss": 2.9372,
+      "step": 8395
+    },
+    {
+      "epoch": 0.516486220472441,
+      "grad_norm": 1.1484442949295044,
+      "learning_rate": 9.491639338087884e-05,
+      "loss": 2.4583,
+      "step": 8396
+    },
+    {
+      "epoch": 0.5165477362204725,
+      "grad_norm": 1.184391736984253,
+      "learning_rate": 9.489708084149402e-05,
+      "loss": 2.21,
+      "step": 8397
+    },
+    {
+      "epoch": 0.5166092519685039,
+      "grad_norm": 1.2955951690673828,
+      "learning_rate": 9.48777684929299e-05,
+      "loss": 2.4881,
+      "step": 8398
+    },
+    {
+      "epoch": 0.5166707677165354,
+      "grad_norm": 1.240967035293579,
+      "learning_rate": 9.485845633590872e-05,
+      "loss": 2.7644,
+      "step": 8399
+    },
+    {
+      "epoch": 0.5167322834645669,
+      "grad_norm": 1.2152522802352905,
+      "learning_rate": 9.483914437115256e-05,
+      "loss": 2.4593,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5167937992125984,
+      "grad_norm": 1.5031617879867554,
+      "learning_rate": 9.481983259938367e-05,
+      "loss": 2.7893,
+      "step": 8401
+    },
+    {
+      "epoch": 0.51685531496063,
+      "grad_norm": 1.1953388452529907,
+      "learning_rate": 9.480052102132411e-05,
+      "loss": 2.763,
+      "step": 8402
+    },
+    {
+      "epoch": 0.5169168307086615,
+      "grad_norm": 1.275161623954773,
+      "learning_rate": 9.478120963769609e-05,
+      "loss": 2.8809,
+      "step": 8403
+    },
+    {
+      "epoch": 0.5169783464566929,
+      "grad_norm": 1.1862001419067383,
+      "learning_rate": 9.476189844922171e-05,
+      "loss": 2.4999,
+      "step": 8404
+    },
+    {
+      "epoch": 0.5170398622047244,
+      "grad_norm": 1.6839112043380737,
+      "learning_rate": 9.474258745662316e-05,
+      "loss": 2.2422,
+      "step": 8405
+    },
+    {
+      "epoch": 0.5171013779527559,
+      "grad_norm": 1.218889594078064,
+      "learning_rate": 9.472327666062249e-05,
+      "loss": 2.8073,
+      "step": 8406
+    },
+    {
+      "epoch": 0.5171628937007874,
+      "grad_norm": 1.034104824066162,
+      "learning_rate": 9.470396606194186e-05,
+      "loss": 3.0776,
+      "step": 8407
+    },
+    {
+      "epoch": 0.5172244094488189,
+      "grad_norm": 1.3821659088134766,
+      "learning_rate": 9.468465566130336e-05,
+      "loss": 1.9242,
+      "step": 8408
+    },
+    {
+      "epoch": 0.5172859251968503,
+      "grad_norm": 1.0870532989501953,
+      "learning_rate": 9.46653454594291e-05,
+      "loss": 2.7543,
+      "step": 8409
+    },
+    {
+      "epoch": 0.5173474409448819,
+      "grad_norm": 1.4062626361846924,
+      "learning_rate": 9.464603545704117e-05,
+      "loss": 2.8655,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5174089566929134,
+      "grad_norm": 1.3181178569793701,
+      "learning_rate": 9.462672565486169e-05,
+      "loss": 2.4832,
+      "step": 8411
+    },
+    {
+      "epoch": 0.5174704724409449,
+      "grad_norm": 1.1418591737747192,
+      "learning_rate": 9.460741605361267e-05,
+      "loss": 2.6392,
+      "step": 8412
+    },
+    {
+      "epoch": 0.5175319881889764,
+      "grad_norm": 1.3905919790267944,
+      "learning_rate": 9.458810665401625e-05,
+      "loss": 2.6156,
+      "step": 8413
+    },
+    {
+      "epoch": 0.5175935039370079,
+      "grad_norm": 1.201379418373108,
+      "learning_rate": 9.456879745679444e-05,
+      "loss": 2.3402,
+      "step": 8414
+    },
+    {
+      "epoch": 0.5176550196850394,
+      "grad_norm": 1.5362982749938965,
+      "learning_rate": 9.454948846266934e-05,
+      "loss": 3.0989,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5177165354330708,
+      "grad_norm": 1.1495211124420166,
+      "learning_rate": 9.453017967236294e-05,
+      "loss": 2.9025,
+      "step": 8416
+    },
+    {
+      "epoch": 0.5177780511811023,
+      "grad_norm": 1.2485389709472656,
+      "learning_rate": 9.451087108659737e-05,
+      "loss": 2.0843,
+      "step": 8417
+    },
+    {
+      "epoch": 0.5178395669291339,
+      "grad_norm": 1.3691059350967407,
+      "learning_rate": 9.449156270609457e-05,
+      "loss": 2.1294,
+      "step": 8418
+    },
+    {
+      "epoch": 0.5179010826771654,
+      "grad_norm": 1.1382609605789185,
+      "learning_rate": 9.447225453157664e-05,
+      "loss": 2.901,
+      "step": 8419
+    },
+    {
+      "epoch": 0.5179625984251969,
+      "grad_norm": 1.1466625928878784,
+      "learning_rate": 9.445294656376554e-05,
+      "loss": 2.1279,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5180241141732284,
+      "grad_norm": 1.0827083587646484,
+      "learning_rate": 9.443363880338335e-05,
+      "loss": 2.8205,
+      "step": 8421
+    },
+    {
+      "epoch": 0.5180856299212598,
+      "grad_norm": 1.2735350131988525,
+      "learning_rate": 9.441433125115199e-05,
+      "loss": 2.0662,
+      "step": 8422
+    },
+    {
+      "epoch": 0.5181471456692913,
+      "grad_norm": 1.1899627447128296,
+      "learning_rate": 9.439502390779352e-05,
+      "loss": 2.5143,
+      "step": 8423
+    },
+    {
+      "epoch": 0.5182086614173228,
+      "grad_norm": 1.2803102731704712,
+      "learning_rate": 9.437571677402986e-05,
+      "loss": 2.1104,
+      "step": 8424
+    },
+    {
+      "epoch": 0.5182701771653543,
+      "grad_norm": 1.191349744796753,
+      "learning_rate": 9.435640985058307e-05,
+      "loss": 2.6761,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5183316929133859,
+      "grad_norm": 1.4171900749206543,
+      "learning_rate": 9.433710313817506e-05,
+      "loss": 2.6375,
+      "step": 8426
+    },
+    {
+      "epoch": 0.5183932086614174,
+      "grad_norm": 1.1126469373703003,
+      "learning_rate": 9.431779663752786e-05,
+      "loss": 2.6738,
+      "step": 8427
+    },
+    {
+      "epoch": 0.5184547244094488,
+      "grad_norm": 1.368662714958191,
+      "learning_rate": 9.429849034936335e-05,
+      "loss": 2.7933,
+      "step": 8428
+    },
+    {
+      "epoch": 0.5185162401574803,
+      "grad_norm": 1.4597527980804443,
+      "learning_rate": 9.42791842744035e-05,
+      "loss": 2.3903,
+      "step": 8429
+    },
+    {
+      "epoch": 0.5185777559055118,
+      "grad_norm": 1.2142077684402466,
+      "learning_rate": 9.425987841337029e-05,
+      "loss": 2.1978,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5186392716535433,
+      "grad_norm": 1.4746044874191284,
+      "learning_rate": 9.424057276698558e-05,
+      "loss": 2.3142,
+      "step": 8431
+    },
+    {
+      "epoch": 0.5187007874015748,
+      "grad_norm": 1.0932053327560425,
+      "learning_rate": 9.422126733597136e-05,
+      "loss": 2.707,
+      "step": 8432
+    },
+    {
+      "epoch": 0.5187623031496063,
+      "grad_norm": 1.557570457458496,
+      "learning_rate": 9.420196212104948e-05,
+      "loss": 2.7661,
+      "step": 8433
+    },
+    {
+      "epoch": 0.5188238188976378,
+      "grad_norm": 1.3526796102523804,
+      "learning_rate": 9.418265712294192e-05,
+      "loss": 2.4931,
+      "step": 8434
+    },
+    {
+      "epoch": 0.5188853346456693,
+      "grad_norm": 1.5721501111984253,
+      "learning_rate": 9.416335234237053e-05,
+      "loss": 3.1682,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5189468503937008,
+      "grad_norm": 1.298606514930725,
+      "learning_rate": 9.414404778005724e-05,
+      "loss": 2.5437,
+      "step": 8436
+    },
+    {
+      "epoch": 0.5190083661417323,
+      "grad_norm": 1.6316375732421875,
+      "learning_rate": 9.412474343672388e-05,
+      "loss": 2.3957,
+      "step": 8437
+    },
+    {
+      "epoch": 0.5190698818897638,
+      "grad_norm": 1.544948697090149,
+      "learning_rate": 9.410543931309238e-05,
+      "loss": 2.1921,
+      "step": 8438
+    },
+    {
+      "epoch": 0.5191313976377953,
+      "grad_norm": 1.0365709066390991,
+      "learning_rate": 9.408613540988455e-05,
+      "loss": 2.0857,
+      "step": 8439
+    },
+    {
+      "epoch": 0.5191929133858267,
+      "grad_norm": 1.0804147720336914,
+      "learning_rate": 9.406683172782232e-05,
+      "loss": 2.8806,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5192544291338582,
+      "grad_norm": 0.9773882031440735,
+      "learning_rate": 9.404752826762747e-05,
+      "loss": 2.7181,
+      "step": 8441
+    },
+    {
+      "epoch": 0.5193159448818898,
+      "grad_norm": 1.147876262664795,
+      "learning_rate": 9.402822503002188e-05,
+      "loss": 2.6065,
+      "step": 8442
+    },
+    {
+      "epoch": 0.5193774606299213,
+      "grad_norm": 1.1819231510162354,
+      "learning_rate": 9.400892201572736e-05,
+      "loss": 2.8905,
+      "step": 8443
+    },
+    {
+      "epoch": 0.5194389763779528,
+      "grad_norm": 1.3232486248016357,
+      "learning_rate": 9.398961922546577e-05,
+      "loss": 3.207,
+      "step": 8444
+    },
+    {
+      "epoch": 0.5195004921259843,
+      "grad_norm": 1.2472833395004272,
+      "learning_rate": 9.397031665995887e-05,
+      "loss": 2.2115,
+      "step": 8445
+    },
+    {
+      "epoch": 0.5195620078740157,
+      "grad_norm": 1.1340419054031372,
+      "learning_rate": 9.395101431992857e-05,
+      "loss": 2.8528,
+      "step": 8446
+    },
+    {
+      "epoch": 0.5196235236220472,
+      "grad_norm": 1.177686333656311,
+      "learning_rate": 9.393171220609654e-05,
+      "loss": 2.9894,
+      "step": 8447
+    },
+    {
+      "epoch": 0.5196850393700787,
+      "grad_norm": 1.1574662923812866,
+      "learning_rate": 9.391241031918467e-05,
+      "loss": 2.4727,
+      "step": 8448
+    },
+    {
+      "epoch": 0.5197465551181102,
+      "grad_norm": 1.0406737327575684,
+      "learning_rate": 9.389310865991469e-05,
+      "loss": 2.6149,
+      "step": 8449
+    },
+    {
+      "epoch": 0.5198080708661418,
+      "grad_norm": 1.2490423917770386,
+      "learning_rate": 9.387380722900842e-05,
+      "loss": 2.7698,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5198695866141733,
+      "grad_norm": 1.2998334169387817,
+      "learning_rate": 9.385450602718755e-05,
+      "loss": 2.5481,
+      "step": 8451
+    },
+    {
+      "epoch": 0.5199311023622047,
+      "grad_norm": 1.0868903398513794,
+      "learning_rate": 9.383520505517394e-05,
+      "loss": 2.5161,
+      "step": 8452
+    },
+    {
+      "epoch": 0.5199926181102362,
+      "grad_norm": 1.7154780626296997,
+      "learning_rate": 9.381590431368926e-05,
+      "loss": 2.5827,
+      "step": 8453
+    },
+    {
+      "epoch": 0.5200541338582677,
+      "grad_norm": 1.2600760459899902,
+      "learning_rate": 9.379660380345528e-05,
+      "loss": 2.3657,
+      "step": 8454
+    },
+    {
+      "epoch": 0.5201156496062992,
+      "grad_norm": 1.160067081451416,
+      "learning_rate": 9.377730352519371e-05,
+      "loss": 2.8014,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5201771653543307,
+      "grad_norm": 1.1704394817352295,
+      "learning_rate": 9.375800347962633e-05,
+      "loss": 1.9324,
+      "step": 8456
+    },
+    {
+      "epoch": 0.5202386811023622,
+      "grad_norm": 1.5038212537765503,
+      "learning_rate": 9.373870366747478e-05,
+      "loss": 2.7734,
+      "step": 8457
+    },
+    {
+      "epoch": 0.5203001968503937,
+      "grad_norm": 1.085179328918457,
+      "learning_rate": 9.371940408946081e-05,
+      "loss": 2.9133,
+      "step": 8458
+    },
+    {
+      "epoch": 0.5203617125984252,
+      "grad_norm": 1.2596720457077026,
+      "learning_rate": 9.37001047463061e-05,
+      "loss": 2.0942,
+      "step": 8459
+    },
+    {
+      "epoch": 0.5204232283464567,
+      "grad_norm": 1.1428275108337402,
+      "learning_rate": 9.368080563873235e-05,
+      "loss": 2.7328,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5204847440944882,
+      "grad_norm": 1.1281404495239258,
+      "learning_rate": 9.366150676746122e-05,
+      "loss": 2.8503,
+      "step": 8461
+    },
+    {
+      "epoch": 0.5205462598425197,
+      "grad_norm": 1.3959786891937256,
+      "learning_rate": 9.364220813321442e-05,
+      "loss": 2.5245,
+      "step": 8462
+    },
+    {
+      "epoch": 0.5206077755905512,
+      "grad_norm": 1.5053291320800781,
+      "learning_rate": 9.362290973671355e-05,
+      "loss": 2.5125,
+      "step": 8463
+    },
+    {
+      "epoch": 0.5206692913385826,
+      "grad_norm": 1.281016230583191,
+      "learning_rate": 9.360361157868032e-05,
+      "loss": 2.4459,
+      "step": 8464
+    },
+    {
+      "epoch": 0.5207308070866141,
+      "grad_norm": 1.4141756296157837,
+      "learning_rate": 9.358431365983636e-05,
+      "loss": 2.2513,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5207923228346457,
+      "grad_norm": 1.2895288467407227,
+      "learning_rate": 9.356501598090328e-05,
+      "loss": 2.5639,
+      "step": 8466
+    },
+    {
+      "epoch": 0.5208538385826772,
+      "grad_norm": 1.1659737825393677,
+      "learning_rate": 9.354571854260271e-05,
+      "loss": 3.1204,
+      "step": 8467
+    },
+    {
+      "epoch": 0.5209153543307087,
+      "grad_norm": 1.2052810192108154,
+      "learning_rate": 9.352642134565627e-05,
+      "loss": 2.5131,
+      "step": 8468
+    },
+    {
+      "epoch": 0.5209768700787402,
+      "grad_norm": 1.1270692348480225,
+      "learning_rate": 9.35071243907856e-05,
+      "loss": 2.6937,
+      "step": 8469
+    },
+    {
+      "epoch": 0.5210383858267716,
+      "grad_norm": 1.3756937980651855,
+      "learning_rate": 9.348782767871222e-05,
+      "loss": 3.0885,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5210999015748031,
+      "grad_norm": 1.2651311159133911,
+      "learning_rate": 9.346853121015785e-05,
+      "loss": 3.0157,
+      "step": 8471
+    },
+    {
+      "epoch": 0.5211614173228346,
+      "grad_norm": 1.0190445184707642,
+      "learning_rate": 9.344923498584394e-05,
+      "loss": 2.5721,
+      "step": 8472
+    },
+    {
+      "epoch": 0.5212229330708661,
+      "grad_norm": 1.372620940208435,
+      "learning_rate": 9.342993900649213e-05,
+      "loss": 2.4428,
+      "step": 8473
+    },
+    {
+      "epoch": 0.5212844488188977,
+      "grad_norm": 1.6929939985275269,
+      "learning_rate": 9.341064327282394e-05,
+      "loss": 3.2695,
+      "step": 8474
+    },
+    {
+      "epoch": 0.5213459645669292,
+      "grad_norm": 1.130351185798645,
+      "learning_rate": 9.3391347785561e-05,
+      "loss": 2.0286,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5214074803149606,
+      "grad_norm": 1.1818442344665527,
+      "learning_rate": 9.337205254542476e-05,
+      "loss": 3.1754,
+      "step": 8476
+    },
+    {
+      "epoch": 0.5214689960629921,
+      "grad_norm": 1.1397937536239624,
+      "learning_rate": 9.335275755313682e-05,
+      "loss": 2.3249,
+      "step": 8477
+    },
+    {
+      "epoch": 0.5215305118110236,
+      "grad_norm": 1.3882256746292114,
+      "learning_rate": 9.333346280941865e-05,
+      "loss": 2.8363,
+      "step": 8478
+    },
+    {
+      "epoch": 0.5215920275590551,
+      "grad_norm": 1.5862228870391846,
+      "learning_rate": 9.331416831499182e-05,
+      "loss": 2.7082,
+      "step": 8479
+    },
+    {
+      "epoch": 0.5216535433070866,
+      "grad_norm": 1.3647496700286865,
+      "learning_rate": 9.32948740705778e-05,
+      "loss": 2.0551,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5217150590551181,
+      "grad_norm": 1.1723041534423828,
+      "learning_rate": 9.327558007689814e-05,
+      "loss": 2.558,
+      "step": 8481
+    },
+    {
+      "epoch": 0.5217765748031497,
+      "grad_norm": 1.488921046257019,
+      "learning_rate": 9.325628633467424e-05,
+      "loss": 2.3968,
+      "step": 8482
+    },
+    {
+      "epoch": 0.5218380905511811,
+      "grad_norm": 1.2988783121109009,
+      "learning_rate": 9.323699284462765e-05,
+      "loss": 2.5615,
+      "step": 8483
+    },
+    {
+      "epoch": 0.5218996062992126,
+      "grad_norm": 1.5239906311035156,
+      "learning_rate": 9.32176996074798e-05,
+      "loss": 2.4635,
+      "step": 8484
+    },
+    {
+      "epoch": 0.5219611220472441,
+      "grad_norm": 1.126315951347351,
+      "learning_rate": 9.31984066239522e-05,
+      "loss": 2.9043,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5220226377952756,
+      "grad_norm": 1.3617161512374878,
+      "learning_rate": 9.317911389476621e-05,
+      "loss": 2.6809,
+      "step": 8486
+    },
+    {
+      "epoch": 0.5220841535433071,
+      "grad_norm": 1.3588591814041138,
+      "learning_rate": 9.315982142064338e-05,
+      "loss": 2.4385,
+      "step": 8487
+    },
+    {
+      "epoch": 0.5221456692913385,
+      "grad_norm": 0.9860087633132935,
+      "learning_rate": 9.314052920230507e-05,
+      "loss": 2.9902,
+      "step": 8488
+    },
+    {
+      "epoch": 0.52220718503937,
+      "grad_norm": 1.217284083366394,
+      "learning_rate": 9.312123724047271e-05,
+      "loss": 2.8858,
+      "step": 8489
+    },
+    {
+      "epoch": 0.5222687007874016,
+      "grad_norm": 1.5354743003845215,
+      "learning_rate": 9.310194553586772e-05,
+      "loss": 2.3222,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5223302165354331,
+      "grad_norm": 1.431999921798706,
+      "learning_rate": 9.308265408921153e-05,
+      "loss": 2.9232,
+      "step": 8491
+    },
+    {
+      "epoch": 0.5223917322834646,
+      "grad_norm": 1.1941413879394531,
+      "learning_rate": 9.306336290122547e-05,
+      "loss": 2.1576,
+      "step": 8492
+    },
+    {
+      "epoch": 0.5224532480314961,
+      "grad_norm": 1.1846094131469727,
+      "learning_rate": 9.304407197263099e-05,
+      "loss": 2.8584,
+      "step": 8493
+    },
+    {
+      "epoch": 0.5225147637795275,
+      "grad_norm": 1.38862943649292,
+      "learning_rate": 9.302478130414938e-05,
+      "loss": 2.1742,
+      "step": 8494
+    },
+    {
+      "epoch": 0.522576279527559,
+      "grad_norm": 1.1079604625701904,
+      "learning_rate": 9.300549089650213e-05,
+      "loss": 2.5116,
+      "step": 8495
+    },
+    {
+      "epoch": 0.5226377952755905,
+      "grad_norm": 1.0407570600509644,
+      "learning_rate": 9.298620075041043e-05,
+      "loss": 2.8269,
+      "step": 8496
+    },
+    {
+      "epoch": 0.522699311023622,
+      "grad_norm": 1.3020771741867065,
+      "learning_rate": 9.296691086659578e-05,
+      "loss": 2.6954,
+      "step": 8497
+    },
+    {
+      "epoch": 0.5227608267716536,
+      "grad_norm": 1.2077611684799194,
+      "learning_rate": 9.29476212457794e-05,
+      "loss": 2.4762,
+      "step": 8498
+    },
+    {
+      "epoch": 0.5228223425196851,
+      "grad_norm": 1.3569668531417847,
+      "learning_rate": 9.292833188868269e-05,
+      "loss": 2.7311,
+      "step": 8499
+    },
+    {
+      "epoch": 0.5228838582677166,
+      "grad_norm": 1.3336117267608643,
+      "learning_rate": 9.290904279602692e-05,
+      "loss": 2.2211,
+      "step": 8500
+    },
+    {
+      "epoch": 0.522945374015748,
+      "grad_norm": 1.2464944124221802,
+      "learning_rate": 9.288975396853344e-05,
+      "loss": 2.4622,
+      "step": 8501
+    },
+    {
+      "epoch": 0.5230068897637795,
+      "grad_norm": 1.307917594909668,
+      "learning_rate": 9.28704654069235e-05,
+      "loss": 2.703,
+      "step": 8502
+    },
+    {
+      "epoch": 0.523068405511811,
+      "grad_norm": 1.211928367614746,
+      "learning_rate": 9.285117711191838e-05,
+      "loss": 1.8052,
+      "step": 8503
+    },
+    {
+      "epoch": 0.5231299212598425,
+      "grad_norm": 1.1911197900772095,
+      "learning_rate": 9.283188908423941e-05,
+      "loss": 2.5279,
+      "step": 8504
+    },
+    {
+      "epoch": 0.523191437007874,
+      "grad_norm": 1.1511297225952148,
+      "learning_rate": 9.281260132460775e-05,
+      "loss": 2.3037,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5232529527559056,
+      "grad_norm": 1.406070590019226,
+      "learning_rate": 9.279331383374479e-05,
+      "loss": 2.4466,
+      "step": 8506
+    },
+    {
+      "epoch": 0.523314468503937,
+      "grad_norm": 1.226250171661377,
+      "learning_rate": 9.277402661237165e-05,
+      "loss": 2.5223,
+      "step": 8507
+    },
+    {
+      "epoch": 0.5233759842519685,
+      "grad_norm": 1.1879825592041016,
+      "learning_rate": 9.275473966120967e-05,
+      "loss": 3.1536,
+      "step": 8508
+    },
+    {
+      "epoch": 0.5234375,
+      "grad_norm": 1.3035138845443726,
+      "learning_rate": 9.273545298097998e-05,
+      "loss": 1.8403,
+      "step": 8509
+    },
+    {
+      "epoch": 0.5234990157480315,
+      "grad_norm": 1.201723575592041,
+      "learning_rate": 9.271616657240391e-05,
+      "loss": 2.3943,
+      "step": 8510
+    },
+    {
+      "epoch": 0.523560531496063,
+      "grad_norm": 1.2490946054458618,
+      "learning_rate": 9.269688043620252e-05,
+      "loss": 2.0831,
+      "step": 8511
+    },
+    {
+      "epoch": 0.5236220472440944,
+      "grad_norm": 1.441981554031372,
+      "learning_rate": 9.267759457309713e-05,
+      "loss": 2.5312,
+      "step": 8512
+    },
+    {
+      "epoch": 0.523683562992126,
+      "grad_norm": 1.2779808044433594,
+      "learning_rate": 9.265830898380883e-05,
+      "loss": 2.1987,
+      "step": 8513
+    },
+    {
+      "epoch": 0.5237450787401575,
+      "grad_norm": 1.2633556127548218,
+      "learning_rate": 9.263902366905886e-05,
+      "loss": 2.8999,
+      "step": 8514
+    },
+    {
+      "epoch": 0.523806594488189,
+      "grad_norm": 1.2624938488006592,
+      "learning_rate": 9.261973862956835e-05,
+      "loss": 2.5933,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5238681102362205,
+      "grad_norm": 1.2054935693740845,
+      "learning_rate": 9.26004538660585e-05,
+      "loss": 2.427,
+      "step": 8516
+    },
+    {
+      "epoch": 0.523929625984252,
+      "grad_norm": 1.3381165266036987,
+      "learning_rate": 9.258116937925036e-05,
+      "loss": 2.333,
+      "step": 8517
+    },
+    {
+      "epoch": 0.5239911417322834,
+      "grad_norm": 1.7103657722473145,
+      "learning_rate": 9.256188516986515e-05,
+      "loss": 2.0686,
+      "step": 8518
+    },
+    {
+      "epoch": 0.5240526574803149,
+      "grad_norm": 1.1944801807403564,
+      "learning_rate": 9.254260123862393e-05,
+      "loss": 2.3128,
+      "step": 8519
+    },
+    {
+      "epoch": 0.5241141732283464,
+      "grad_norm": 1.1399720907211304,
+      "learning_rate": 9.252331758624788e-05,
+      "loss": 2.9729,
+      "step": 8520
+    },
+    {
+      "epoch": 0.524175688976378,
+      "grad_norm": 1.401511311531067,
+      "learning_rate": 9.250403421345803e-05,
+      "loss": 2.9034,
+      "step": 8521
+    },
+    {
+      "epoch": 0.5242372047244095,
+      "grad_norm": 1.2252548933029175,
+      "learning_rate": 9.248475112097551e-05,
+      "loss": 2.3465,
+      "step": 8522
+    },
+    {
+      "epoch": 0.524298720472441,
+      "grad_norm": 1.1924911737442017,
+      "learning_rate": 9.246546830952137e-05,
+      "loss": 3.1298,
+      "step": 8523
+    },
+    {
+      "epoch": 0.5243602362204725,
+      "grad_norm": 1.3554874658584595,
+      "learning_rate": 9.244618577981674e-05,
+      "loss": 2.0727,
+      "step": 8524
+    },
+    {
+      "epoch": 0.5244217519685039,
+      "grad_norm": 1.124437689781189,
+      "learning_rate": 9.242690353258261e-05,
+      "loss": 2.4585,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5244832677165354,
+      "grad_norm": 1.213649034500122,
+      "learning_rate": 9.24076215685401e-05,
+      "loss": 2.4393,
+      "step": 8526
+    },
+    {
+      "epoch": 0.5245447834645669,
+      "grad_norm": 1.3027397394180298,
+      "learning_rate": 9.238833988841015e-05,
+      "loss": 2.6671,
+      "step": 8527
+    },
+    {
+      "epoch": 0.5246062992125984,
+      "grad_norm": 1.1175583600997925,
+      "learning_rate": 9.236905849291389e-05,
+      "loss": 3.2433,
+      "step": 8528
+    },
+    {
+      "epoch": 0.52466781496063,
+      "grad_norm": 1.2004929780960083,
+      "learning_rate": 9.234977738277226e-05,
+      "loss": 2.9406,
+      "step": 8529
+    },
+    {
+      "epoch": 0.5247293307086615,
+      "grad_norm": 1.320556879043579,
+      "learning_rate": 9.233049655870635e-05,
+      "loss": 2.7506,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5247908464566929,
+      "grad_norm": 1.0276225805282593,
+      "learning_rate": 9.231121602143703e-05,
+      "loss": 3.2228,
+      "step": 8531
+    },
+    {
+      "epoch": 0.5248523622047244,
+      "grad_norm": 1.2492305040359497,
+      "learning_rate": 9.229193577168543e-05,
+      "loss": 2.1179,
+      "step": 8532
+    },
+    {
+      "epoch": 0.5249138779527559,
+      "grad_norm": 1.130123496055603,
+      "learning_rate": 9.227265581017238e-05,
+      "loss": 2.4824,
+      "step": 8533
+    },
+    {
+      "epoch": 0.5249753937007874,
+      "grad_norm": 1.217750072479248,
+      "learning_rate": 9.225337613761896e-05,
+      "loss": 3.0647,
+      "step": 8534
+    },
+    {
+      "epoch": 0.5250369094488189,
+      "grad_norm": 1.289331078529358,
+      "learning_rate": 9.223409675474605e-05,
+      "loss": 2.859,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5250984251968503,
+      "grad_norm": 1.4026092290878296,
+      "learning_rate": 9.221481766227466e-05,
+      "loss": 2.8118,
+      "step": 8536
+    },
+    {
+      "epoch": 0.5251599409448819,
+      "grad_norm": 1.4902969598770142,
+      "learning_rate": 9.219553886092563e-05,
+      "loss": 1.2866,
+      "step": 8537
+    },
+    {
+      "epoch": 0.5252214566929134,
+      "grad_norm": 1.1948285102844238,
+      "learning_rate": 9.217626035141995e-05,
+      "loss": 2.5936,
+      "step": 8538
+    },
+    {
+      "epoch": 0.5252829724409449,
+      "grad_norm": 1.186699628829956,
+      "learning_rate": 9.215698213447852e-05,
+      "loss": 2.5614,
+      "step": 8539
+    },
+    {
+      "epoch": 0.5253444881889764,
+      "grad_norm": 1.1271172761917114,
+      "learning_rate": 9.213770421082219e-05,
+      "loss": 2.8843,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5254060039370079,
+      "grad_norm": 1.2647290229797363,
+      "learning_rate": 9.211842658117192e-05,
+      "loss": 2.6139,
+      "step": 8541
+    },
+    {
+      "epoch": 0.5254675196850394,
+      "grad_norm": 1.0887911319732666,
+      "learning_rate": 9.209914924624851e-05,
+      "loss": 2.6745,
+      "step": 8542
+    },
+    {
+      "epoch": 0.5255290354330708,
+      "grad_norm": 1.2068647146224976,
+      "learning_rate": 9.20798722067729e-05,
+      "loss": 2.5876,
+      "step": 8543
+    },
+    {
+      "epoch": 0.5255905511811023,
+      "grad_norm": 1.5654188394546509,
+      "learning_rate": 9.206059546346584e-05,
+      "loss": 2.3946,
+      "step": 8544
+    },
+    {
+      "epoch": 0.5256520669291339,
+      "grad_norm": 1.0805264711380005,
+      "learning_rate": 9.204131901704833e-05,
+      "loss": 2.328,
+      "step": 8545
+    },
+    {
+      "epoch": 0.5257135826771654,
+      "grad_norm": 1.2874178886413574,
+      "learning_rate": 9.202204286824103e-05,
+      "loss": 2.8688,
+      "step": 8546
+    },
+    {
+      "epoch": 0.5257750984251969,
+      "grad_norm": 1.186677098274231,
+      "learning_rate": 9.200276701776489e-05,
+      "loss": 2.0794,
+      "step": 8547
+    },
+    {
+      "epoch": 0.5258366141732284,
+      "grad_norm": 1.5003809928894043,
+      "learning_rate": 9.198349146634064e-05,
+      "loss": 1.9677,
+      "step": 8548
+    },
+    {
+      "epoch": 0.5258981299212598,
+      "grad_norm": 1.3959518671035767,
+      "learning_rate": 9.196421621468911e-05,
+      "loss": 2.3441,
+      "step": 8549
+    },
+    {
+      "epoch": 0.5259596456692913,
+      "grad_norm": 1.2630321979522705,
+      "learning_rate": 9.194494126353108e-05,
+      "loss": 2.0574,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5260211614173228,
+      "grad_norm": 1.2830268144607544,
+      "learning_rate": 9.192566661358738e-05,
+      "loss": 2.0494,
+      "step": 8551
+    },
+    {
+      "epoch": 0.5260826771653543,
+      "grad_norm": 1.3451396226882935,
+      "learning_rate": 9.190639226557867e-05,
+      "loss": 2.0967,
+      "step": 8552
+    },
+    {
+      "epoch": 0.5261441929133859,
+      "grad_norm": 1.1340912580490112,
+      "learning_rate": 9.188711822022579e-05,
+      "loss": 2.4175,
+      "step": 8553
+    },
+    {
+      "epoch": 0.5262057086614174,
+      "grad_norm": 1.3679219484329224,
+      "learning_rate": 9.186784447824942e-05,
+      "loss": 2.6746,
+      "step": 8554
+    },
+    {
+      "epoch": 0.5262672244094488,
+      "grad_norm": 1.479862093925476,
+      "learning_rate": 9.184857104037036e-05,
+      "loss": 2.5995,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5263287401574803,
+      "grad_norm": 1.3289539813995361,
+      "learning_rate": 9.182929790730926e-05,
+      "loss": 2.8868,
+      "step": 8556
+    },
+    {
+      "epoch": 0.5263902559055118,
+      "grad_norm": 1.2152018547058105,
+      "learning_rate": 9.181002507978687e-05,
+      "loss": 2.7058,
+      "step": 8557
+    },
+    {
+      "epoch": 0.5264517716535433,
+      "grad_norm": 1.3641971349716187,
+      "learning_rate": 9.179075255852385e-05,
+      "loss": 2.1836,
+      "step": 8558
+    },
+    {
+      "epoch": 0.5265132874015748,
+      "grad_norm": 1.2815732955932617,
+      "learning_rate": 9.177148034424093e-05,
+      "loss": 3.0375,
+      "step": 8559
+    },
+    {
+      "epoch": 0.5265748031496063,
+      "grad_norm": 1.234792947769165,
+      "learning_rate": 9.175220843765875e-05,
+      "loss": 2.2089,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5266363188976378,
+      "grad_norm": 1.313652515411377,
+      "learning_rate": 9.173293683949802e-05,
+      "loss": 2.4722,
+      "step": 8561
+    },
+    {
+      "epoch": 0.5266978346456693,
+      "grad_norm": 1.3482919931411743,
+      "learning_rate": 9.17136655504793e-05,
+      "loss": 1.9801,
+      "step": 8562
+    },
+    {
+      "epoch": 0.5267593503937008,
+      "grad_norm": 1.4496039152145386,
+      "learning_rate": 9.169439457132332e-05,
+      "loss": 2.0577,
+      "step": 8563
+    },
+    {
+      "epoch": 0.5268208661417323,
+      "grad_norm": 1.1781237125396729,
+      "learning_rate": 9.167512390275064e-05,
+      "loss": 3.0943,
+      "step": 8564
+    },
+    {
+      "epoch": 0.5268823818897638,
+      "grad_norm": 1.2167519330978394,
+      "learning_rate": 9.165585354548195e-05,
+      "loss": 2.5149,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5269438976377953,
+      "grad_norm": 1.0746363401412964,
+      "learning_rate": 9.163658350023775e-05,
+      "loss": 2.254,
+      "step": 8566
+    },
+    {
+      "epoch": 0.5270054133858267,
+      "grad_norm": 1.133798360824585,
+      "learning_rate": 9.161731376773876e-05,
+      "loss": 2.947,
+      "step": 8567
+    },
+    {
+      "epoch": 0.5270669291338582,
+      "grad_norm": 1.1182947158813477,
+      "learning_rate": 9.159804434870543e-05,
+      "loss": 3.223,
+      "step": 8568
+    },
+    {
+      "epoch": 0.5271284448818898,
+      "grad_norm": 1.195249080657959,
+      "learning_rate": 9.157877524385843e-05,
+      "loss": 2.1223,
+      "step": 8569
+    },
+    {
+      "epoch": 0.5271899606299213,
+      "grad_norm": 1.1192834377288818,
+      "learning_rate": 9.155950645391824e-05,
+      "loss": 2.7875,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5272514763779528,
+      "grad_norm": 1.105224609375,
+      "learning_rate": 9.15402379796055e-05,
+      "loss": 2.3322,
+      "step": 8571
+    },
+    {
+      "epoch": 0.5273129921259843,
+      "grad_norm": 1.0616884231567383,
+      "learning_rate": 9.152096982164063e-05,
+      "loss": 2.9944,
+      "step": 8572
+    },
+    {
+      "epoch": 0.5273745078740157,
+      "grad_norm": 1.1799978017807007,
+      "learning_rate": 9.150170198074425e-05,
+      "loss": 2.6336,
+      "step": 8573
+    },
+    {
+      "epoch": 0.5274360236220472,
+      "grad_norm": 1.1640756130218506,
+      "learning_rate": 9.14824344576368e-05,
+      "loss": 2.7514,
+      "step": 8574
+    },
+    {
+      "epoch": 0.5274975393700787,
+      "grad_norm": 1.1542253494262695,
+      "learning_rate": 9.146316725303881e-05,
+      "loss": 2.7577,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5275590551181102,
+      "grad_norm": 1.1526012420654297,
+      "learning_rate": 9.14439003676708e-05,
+      "loss": 2.585,
+      "step": 8576
+    },
+    {
+      "epoch": 0.5276205708661418,
+      "grad_norm": 1.3873400688171387,
+      "learning_rate": 9.142463380225315e-05,
+      "loss": 2.5057,
+      "step": 8577
+    },
+    {
+      "epoch": 0.5276820866141733,
+      "grad_norm": 1.436115026473999,
+      "learning_rate": 9.140536755750641e-05,
+      "loss": 2.4327,
+      "step": 8578
+    },
+    {
+      "epoch": 0.5277436023622047,
+      "grad_norm": 1.343629002571106,
+      "learning_rate": 9.138610163415097e-05,
+      "loss": 2.547,
+      "step": 8579
+    },
+    {
+      "epoch": 0.5278051181102362,
+      "grad_norm": 1.3839319944381714,
+      "learning_rate": 9.136683603290735e-05,
+      "loss": 2.7001,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5278666338582677,
+      "grad_norm": 2.0926973819732666,
+      "learning_rate": 9.134757075449588e-05,
+      "loss": 2.2456,
+      "step": 8581
+    },
+    {
+      "epoch": 0.5279281496062992,
+      "grad_norm": 1.3647915124893188,
+      "learning_rate": 9.132830579963704e-05,
+      "loss": 3.074,
+      "step": 8582
+    },
+    {
+      "epoch": 0.5279896653543307,
+      "grad_norm": 1.223942756652832,
+      "learning_rate": 9.130904116905116e-05,
+      "loss": 2.0475,
+      "step": 8583
+    },
+    {
+      "epoch": 0.5280511811023622,
+      "grad_norm": 1.494646430015564,
+      "learning_rate": 9.128977686345873e-05,
+      "loss": 2.3863,
+      "step": 8584
+    },
+    {
+      "epoch": 0.5281126968503937,
+      "grad_norm": 1.1961517333984375,
+      "learning_rate": 9.127051288358003e-05,
+      "loss": 2.2621,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5281742125984252,
+      "grad_norm": 1.2081382274627686,
+      "learning_rate": 9.125124923013555e-05,
+      "loss": 2.8405,
+      "step": 8586
+    },
+    {
+      "epoch": 0.5282357283464567,
+      "grad_norm": 1.219925045967102,
+      "learning_rate": 9.12319859038455e-05,
+      "loss": 2.6901,
+      "step": 8587
+    },
+    {
+      "epoch": 0.5282972440944882,
+      "grad_norm": 1.4552139043807983,
+      "learning_rate": 9.121272290543031e-05,
+      "loss": 2.4783,
+      "step": 8588
+    },
+    {
+      "epoch": 0.5283587598425197,
+      "grad_norm": 1.2819626331329346,
+      "learning_rate": 9.119346023561025e-05,
+      "loss": 2.5762,
+      "step": 8589
+    },
+    {
+      "epoch": 0.5284202755905512,
+      "grad_norm": 1.0976682901382446,
+      "learning_rate": 9.117419789510572e-05,
+      "loss": 2.6208,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5284817913385826,
+      "grad_norm": 1.4499284029006958,
+      "learning_rate": 9.115493588463693e-05,
+      "loss": 2.4027,
+      "step": 8591
+    },
+    {
+      "epoch": 0.5285433070866141,
+      "grad_norm": 1.3378252983093262,
+      "learning_rate": 9.113567420492424e-05,
+      "loss": 2.9319,
+      "step": 8592
+    },
+    {
+      "epoch": 0.5286048228346457,
+      "grad_norm": 2.7534353733062744,
+      "learning_rate": 9.11164128566879e-05,
+      "loss": 1.8833,
+      "step": 8593
+    },
+    {
+      "epoch": 0.5286663385826772,
+      "grad_norm": 1.177855372428894,
+      "learning_rate": 9.109715184064818e-05,
+      "loss": 3.2369,
+      "step": 8594
+    },
+    {
+      "epoch": 0.5287278543307087,
+      "grad_norm": 1.0979416370391846,
+      "learning_rate": 9.107789115752533e-05,
+      "loss": 2.673,
+      "step": 8595
+    },
+    {
+      "epoch": 0.5287893700787402,
+      "grad_norm": 1.2452839612960815,
+      "learning_rate": 9.105863080803964e-05,
+      "loss": 2.1732,
+      "step": 8596
+    },
+    {
+      "epoch": 0.5288508858267716,
+      "grad_norm": 1.0209227800369263,
+      "learning_rate": 9.103937079291125e-05,
+      "loss": 2.7481,
+      "step": 8597
+    },
+    {
+      "epoch": 0.5289124015748031,
+      "grad_norm": 1.3166621923446655,
+      "learning_rate": 9.102011111286047e-05,
+      "loss": 2.291,
+      "step": 8598
+    },
+    {
+      "epoch": 0.5289739173228346,
+      "grad_norm": 1.3285654783248901,
+      "learning_rate": 9.10008517686074e-05,
+      "loss": 2.504,
+      "step": 8599
+    },
+    {
+      "epoch": 0.5290354330708661,
+      "grad_norm": 1.3434808254241943,
+      "learning_rate": 9.098159276087237e-05,
+      "loss": 2.2156,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5290969488188977,
+      "grad_norm": 1.289945363998413,
+      "learning_rate": 9.096233409037541e-05,
+      "loss": 2.004,
+      "step": 8601
+    },
+    {
+      "epoch": 0.5291584645669292,
+      "grad_norm": 1.4056504964828491,
+      "learning_rate": 9.094307575783683e-05,
+      "loss": 2.5349,
+      "step": 8602
+    },
+    {
+      "epoch": 0.5292199803149606,
+      "grad_norm": 1.3963468074798584,
+      "learning_rate": 9.092381776397666e-05,
+      "loss": 2.6443,
+      "step": 8603
+    },
+    {
+      "epoch": 0.5292814960629921,
+      "grad_norm": 1.0140279531478882,
+      "learning_rate": 9.090456010951512e-05,
+      "loss": 3.0773,
+      "step": 8604
+    },
+    {
+      "epoch": 0.5293430118110236,
+      "grad_norm": 1.143874168395996,
+      "learning_rate": 9.088530279517229e-05,
+      "loss": 2.2045,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5294045275590551,
+      "grad_norm": 1.3231228590011597,
+      "learning_rate": 9.086604582166835e-05,
+      "loss": 2.0783,
+      "step": 8606
+    },
+    {
+      "epoch": 0.5294660433070866,
+      "grad_norm": 1.7870367765426636,
+      "learning_rate": 9.084678918972334e-05,
+      "loss": 2.2336,
+      "step": 8607
+    },
+    {
+      "epoch": 0.5295275590551181,
+      "grad_norm": 1.2184032201766968,
+      "learning_rate": 9.082753290005738e-05,
+      "loss": 2.8822,
+      "step": 8608
+    },
+    {
+      "epoch": 0.5295890748031497,
+      "grad_norm": 1.3300234079360962,
+      "learning_rate": 9.080827695339052e-05,
+      "loss": 2.4717,
+      "step": 8609
+    },
+    {
+      "epoch": 0.5296505905511811,
+      "grad_norm": 1.0958620309829712,
+      "learning_rate": 9.078902135044287e-05,
+      "loss": 2.8807,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5297121062992126,
+      "grad_norm": 1.1329206228256226,
+      "learning_rate": 9.076976609193449e-05,
+      "loss": 2.6266,
+      "step": 8611
+    },
+    {
+      "epoch": 0.5297736220472441,
+      "grad_norm": 1.0770678520202637,
+      "learning_rate": 9.075051117858534e-05,
+      "loss": 2.5261,
+      "step": 8612
+    },
+    {
+      "epoch": 0.5298351377952756,
+      "grad_norm": 1.250490427017212,
+      "learning_rate": 9.073125661111552e-05,
+      "loss": 2.4202,
+      "step": 8613
+    },
+    {
+      "epoch": 0.5298966535433071,
+      "grad_norm": 1.0194065570831299,
+      "learning_rate": 9.071200239024499e-05,
+      "loss": 2.4511,
+      "step": 8614
+    },
+    {
+      "epoch": 0.5299581692913385,
+      "grad_norm": 1.2670960426330566,
+      "learning_rate": 9.069274851669382e-05,
+      "loss": 2.6488,
+      "step": 8615
+    },
+    {
+      "epoch": 0.53001968503937,
+      "grad_norm": 1.3194849491119385,
+      "learning_rate": 9.067349499118193e-05,
+      "loss": 3.0756,
+      "step": 8616
+    },
+    {
+      "epoch": 0.5300812007874016,
+      "grad_norm": 1.082755446434021,
+      "learning_rate": 9.065424181442934e-05,
+      "loss": 2.6274,
+      "step": 8617
+    },
+    {
+      "epoch": 0.5301427165354331,
+      "grad_norm": 1.5292015075683594,
+      "learning_rate": 9.063498898715595e-05,
+      "loss": 2.3079,
+      "step": 8618
+    },
+    {
+      "epoch": 0.5302042322834646,
+      "grad_norm": 1.1546927690505981,
+      "learning_rate": 9.061573651008182e-05,
+      "loss": 1.883,
+      "step": 8619
+    },
+    {
+      "epoch": 0.5302657480314961,
+      "grad_norm": 1.1553208827972412,
+      "learning_rate": 9.059648438392674e-05,
+      "loss": 2.8924,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5303272637795275,
+      "grad_norm": 1.526698112487793,
+      "learning_rate": 9.057723260941076e-05,
+      "loss": 1.9083,
+      "step": 8621
+    },
+    {
+      "epoch": 0.530388779527559,
+      "grad_norm": 1.2590194940567017,
+      "learning_rate": 9.055798118725371e-05,
+      "loss": 2.5943,
+      "step": 8622
+    },
+    {
+      "epoch": 0.5304502952755905,
+      "grad_norm": 1.1622639894485474,
+      "learning_rate": 9.053873011817555e-05,
+      "loss": 3.0964,
+      "step": 8623
+    },
+    {
+      "epoch": 0.530511811023622,
+      "grad_norm": 1.2679328918457031,
+      "learning_rate": 9.051947940289608e-05,
+      "loss": 2.6992,
+      "step": 8624
+    },
+    {
+      "epoch": 0.5305733267716536,
+      "grad_norm": 1.299763560295105,
+      "learning_rate": 9.050022904213526e-05,
+      "loss": 2.1941,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5306348425196851,
+      "grad_norm": 1.1989010572433472,
+      "learning_rate": 9.048097903661286e-05,
+      "loss": 2.3969,
+      "step": 8626
+    },
+    {
+      "epoch": 0.5306963582677166,
+      "grad_norm": 1.2524060010910034,
+      "learning_rate": 9.04617293870488e-05,
+      "loss": 2.6006,
+      "step": 8627
+    },
+    {
+      "epoch": 0.530757874015748,
+      "grad_norm": 1.14457368850708,
+      "learning_rate": 9.044248009416284e-05,
+      "loss": 2.9,
+      "step": 8628
+    },
+    {
+      "epoch": 0.5308193897637795,
+      "grad_norm": 1.189940094947815,
+      "learning_rate": 9.042323115867485e-05,
+      "loss": 3.1287,
+      "step": 8629
+    },
+    {
+      "epoch": 0.530880905511811,
+      "grad_norm": 1.376964807510376,
+      "learning_rate": 9.04039825813046e-05,
+      "loss": 2.23,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5309424212598425,
+      "grad_norm": 1.1464316844940186,
+      "learning_rate": 9.038473436277193e-05,
+      "loss": 2.9204,
+      "step": 8631
+    },
+    {
+      "epoch": 0.531003937007874,
+      "grad_norm": 1.1015790700912476,
+      "learning_rate": 9.036548650379655e-05,
+      "loss": 2.8338,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5310654527559056,
+      "grad_norm": 1.5035582780838013,
+      "learning_rate": 9.034623900509827e-05,
+      "loss": 2.5278,
+      "step": 8633
+    },
+    {
+      "epoch": 0.531126968503937,
+      "grad_norm": 1.0962461233139038,
+      "learning_rate": 9.032699186739678e-05,
+      "loss": 2.0576,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5311884842519685,
+      "grad_norm": 1.112338662147522,
+      "learning_rate": 9.030774509141191e-05,
+      "loss": 2.5556,
+      "step": 8635
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 1.2999188899993896,
+      "learning_rate": 9.028849867786326e-05,
+      "loss": 2.4837,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5313115157480315,
+      "grad_norm": 1.6167991161346436,
+      "learning_rate": 9.026925262747069e-05,
+      "loss": 2.3976,
+      "step": 8637
+    },
+    {
+      "epoch": 0.531373031496063,
+      "grad_norm": 1.0504248142242432,
+      "learning_rate": 9.025000694095374e-05,
+      "loss": 3.3787,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5314345472440944,
+      "grad_norm": 1.3823373317718506,
+      "learning_rate": 9.023076161903219e-05,
+      "loss": 2.3147,
+      "step": 8639
+    },
+    {
+      "epoch": 0.531496062992126,
+      "grad_norm": 1.2257128953933716,
+      "learning_rate": 9.021151666242566e-05,
+      "loss": 2.6343,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5315575787401575,
+      "grad_norm": 1.172044038772583,
+      "learning_rate": 9.019227207185389e-05,
+      "loss": 2.3993,
+      "step": 8641
+    },
+    {
+      "epoch": 0.531619094488189,
+      "grad_norm": 1.2722500562667847,
+      "learning_rate": 9.01730278480364e-05,
+      "loss": 2.9618,
+      "step": 8642
+    },
+    {
+      "epoch": 0.5316806102362205,
+      "grad_norm": 1.2153156995773315,
+      "learning_rate": 9.015378399169287e-05,
+      "loss": 2.4464,
+      "step": 8643
+    },
+    {
+      "epoch": 0.531742125984252,
+      "grad_norm": 1.2781696319580078,
+      "learning_rate": 9.013454050354293e-05,
+      "loss": 2.9123,
+      "step": 8644
+    },
+    {
+      "epoch": 0.5318036417322834,
+      "grad_norm": 1.4049205780029297,
+      "learning_rate": 9.011529738430619e-05,
+      "loss": 2.4736,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5318651574803149,
+      "grad_norm": 1.362953782081604,
+      "learning_rate": 9.009605463470213e-05,
+      "loss": 2.4493,
+      "step": 8646
+    },
+    {
+      "epoch": 0.5319266732283464,
+      "grad_norm": 1.2548307180404663,
+      "learning_rate": 9.007681225545048e-05,
+      "loss": 2.348,
+      "step": 8647
+    },
+    {
+      "epoch": 0.531988188976378,
+      "grad_norm": 1.3103933334350586,
+      "learning_rate": 9.00575702472707e-05,
+      "loss": 2.56,
+      "step": 8648
+    },
+    {
+      "epoch": 0.5320497047244095,
+      "grad_norm": 1.467149019241333,
+      "learning_rate": 9.003832861088233e-05,
+      "loss": 2.533,
+      "step": 8649
+    },
+    {
+      "epoch": 0.532111220472441,
+      "grad_norm": 1.7129554748535156,
+      "learning_rate": 9.001908734700496e-05,
+      "loss": 1.0949,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5321727362204725,
+      "grad_norm": 1.3182082176208496,
+      "learning_rate": 8.999984645635803e-05,
+      "loss": 2.9872,
+      "step": 8651
+    },
+    {
+      "epoch": 0.5322342519685039,
+      "grad_norm": 1.3403782844543457,
+      "learning_rate": 8.99806059396611e-05,
+      "loss": 2.3984,
+      "step": 8652
+    },
+    {
+      "epoch": 0.5322957677165354,
+      "grad_norm": 1.1112539768218994,
+      "learning_rate": 8.996136579763361e-05,
+      "loss": 2.1372,
+      "step": 8653
+    },
+    {
+      "epoch": 0.5323572834645669,
+      "grad_norm": 1.1487245559692383,
+      "learning_rate": 8.994212603099512e-05,
+      "loss": 2.569,
+      "step": 8654
+    },
+    {
+      "epoch": 0.5324187992125984,
+      "grad_norm": 1.4148730039596558,
+      "learning_rate": 8.992288664046496e-05,
+      "loss": 2.5154,
+      "step": 8655
+    },
+    {
+      "epoch": 0.53248031496063,
+      "grad_norm": 1.127703070640564,
+      "learning_rate": 8.990364762676269e-05,
+      "loss": 1.8923,
+      "step": 8656
+    },
+    {
+      "epoch": 0.5325418307086615,
+      "grad_norm": 1.604523777961731,
+      "learning_rate": 8.988440899060769e-05,
+      "loss": 2.5899,
+      "step": 8657
+    },
+    {
+      "epoch": 0.5326033464566929,
+      "grad_norm": 1.3536767959594727,
+      "learning_rate": 8.986517073271939e-05,
+      "loss": 2.5413,
+      "step": 8658
+    },
+    {
+      "epoch": 0.5326648622047244,
+      "grad_norm": 1.1236141920089722,
+      "learning_rate": 8.984593285381719e-05,
+      "loss": 2.3285,
+      "step": 8659
+    },
+    {
+      "epoch": 0.5327263779527559,
+      "grad_norm": 1.3442047834396362,
+      "learning_rate": 8.98266953546205e-05,
+      "loss": 2.3438,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5327878937007874,
+      "grad_norm": 1.0452406406402588,
+      "learning_rate": 8.980745823584866e-05,
+      "loss": 3.1289,
+      "step": 8661
+    },
+    {
+      "epoch": 0.5328494094488189,
+      "grad_norm": 1.4045372009277344,
+      "learning_rate": 8.978822149822104e-05,
+      "loss": 2.7164,
+      "step": 8662
+    },
+    {
+      "epoch": 0.5329109251968503,
+      "grad_norm": 1.236183524131775,
+      "learning_rate": 8.976898514245699e-05,
+      "loss": 2.7653,
+      "step": 8663
+    },
+    {
+      "epoch": 0.5329724409448819,
+      "grad_norm": 1.3095771074295044,
+      "learning_rate": 8.974974916927585e-05,
+      "loss": 2.4858,
+      "step": 8664
+    },
+    {
+      "epoch": 0.5330339566929134,
+      "grad_norm": 1.4881503582000732,
+      "learning_rate": 8.973051357939693e-05,
+      "loss": 2.7127,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5330954724409449,
+      "grad_norm": 1.3239408731460571,
+      "learning_rate": 8.971127837353957e-05,
+      "loss": 2.8861,
+      "step": 8666
+    },
+    {
+      "epoch": 0.5331569881889764,
+      "grad_norm": 1.3025262355804443,
+      "learning_rate": 8.969204355242299e-05,
+      "loss": 2.7921,
+      "step": 8667
+    },
+    {
+      "epoch": 0.5332185039370079,
+      "grad_norm": 1.1958730220794678,
+      "learning_rate": 8.967280911676653e-05,
+      "loss": 2.2355,
+      "step": 8668
+    },
+    {
+      "epoch": 0.5332800196850394,
+      "grad_norm": 1.2911648750305176,
+      "learning_rate": 8.965357506728939e-05,
+      "loss": 2.2876,
+      "step": 8669
+    },
+    {
+      "epoch": 0.5333415354330708,
+      "grad_norm": 1.1413195133209229,
+      "learning_rate": 8.96343414047109e-05,
+      "loss": 2.8555,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5334030511811023,
+      "grad_norm": 1.1050760746002197,
+      "learning_rate": 8.961510812975019e-05,
+      "loss": 2.9962,
+      "step": 8671
+    },
+    {
+      "epoch": 0.5334645669291339,
+      "grad_norm": 1.1859439611434937,
+      "learning_rate": 8.959587524312656e-05,
+      "loss": 2.6124,
+      "step": 8672
+    },
+    {
+      "epoch": 0.5335260826771654,
+      "grad_norm": 1.150641918182373,
+      "learning_rate": 8.957664274555915e-05,
+      "loss": 2.2478,
+      "step": 8673
+    },
+    {
+      "epoch": 0.5335875984251969,
+      "grad_norm": 1.363936185836792,
+      "learning_rate": 8.955741063776721e-05,
+      "loss": 2.7542,
+      "step": 8674
+    },
+    {
+      "epoch": 0.5336491141732284,
+      "grad_norm": 1.173126220703125,
+      "learning_rate": 8.953817892046985e-05,
+      "loss": 2.9101,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5337106299212598,
+      "grad_norm": 1.264100193977356,
+      "learning_rate": 8.95189475943863e-05,
+      "loss": 2.7445,
+      "step": 8676
+    },
+    {
+      "epoch": 0.5337721456692913,
+      "grad_norm": 1.2077105045318604,
+      "learning_rate": 8.949971666023565e-05,
+      "loss": 2.5321,
+      "step": 8677
+    },
+    {
+      "epoch": 0.5338336614173228,
+      "grad_norm": 1.5250957012176514,
+      "learning_rate": 8.948048611873704e-05,
+      "loss": 2.2564,
+      "step": 8678
+    },
+    {
+      "epoch": 0.5338951771653543,
+      "grad_norm": 1.4088741540908813,
+      "learning_rate": 8.946125597060958e-05,
+      "loss": 2.2293,
+      "step": 8679
+    },
+    {
+      "epoch": 0.5339566929133859,
+      "grad_norm": 1.1061025857925415,
+      "learning_rate": 8.944202621657243e-05,
+      "loss": 2.7751,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5340182086614174,
+      "grad_norm": 1.3386693000793457,
+      "learning_rate": 8.942279685734455e-05,
+      "loss": 2.9614,
+      "step": 8681
+    },
+    {
+      "epoch": 0.5340797244094488,
+      "grad_norm": 1.2899184226989746,
+      "learning_rate": 8.940356789364515e-05,
+      "loss": 2.382,
+      "step": 8682
+    },
+    {
+      "epoch": 0.5341412401574803,
+      "grad_norm": 1.3455287218093872,
+      "learning_rate": 8.93843393261932e-05,
+      "loss": 2.824,
+      "step": 8683
+    },
+    {
+      "epoch": 0.5342027559055118,
+      "grad_norm": 1.157379388809204,
+      "learning_rate": 8.936511115570778e-05,
+      "loss": 2.3851,
+      "step": 8684
+    },
+    {
+      "epoch": 0.5342642716535433,
+      "grad_norm": 1.3755513429641724,
+      "learning_rate": 8.93458833829079e-05,
+      "loss": 2.7163,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5343257874015748,
+      "grad_norm": 1.1561988592147827,
+      "learning_rate": 8.932665600851256e-05,
+      "loss": 2.792,
+      "step": 8686
+    },
+    {
+      "epoch": 0.5343873031496063,
+      "grad_norm": 1.5234358310699463,
+      "learning_rate": 8.930742903324077e-05,
+      "loss": 2.1773,
+      "step": 8687
+    },
+    {
+      "epoch": 0.5344488188976378,
+      "grad_norm": 1.438562035560608,
+      "learning_rate": 8.92882024578115e-05,
+      "loss": 1.8933,
+      "step": 8688
+    },
+    {
+      "epoch": 0.5345103346456693,
+      "grad_norm": 1.5536041259765625,
+      "learning_rate": 8.926897628294376e-05,
+      "loss": 2.5116,
+      "step": 8689
+    },
+    {
+      "epoch": 0.5345718503937008,
+      "grad_norm": 1.31061589717865,
+      "learning_rate": 8.924975050935642e-05,
+      "loss": 2.3392,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5346333661417323,
+      "grad_norm": 1.1983445882797241,
+      "learning_rate": 8.923052513776851e-05,
+      "loss": 2.8569,
+      "step": 8691
+    },
+    {
+      "epoch": 0.5346948818897638,
+      "grad_norm": 1.947075605392456,
+      "learning_rate": 8.921130016889888e-05,
+      "loss": 1.2624,
+      "step": 8692
+    },
+    {
+      "epoch": 0.5347563976377953,
+      "grad_norm": 1.2658908367156982,
+      "learning_rate": 8.919207560346649e-05,
+      "loss": 2.9449,
+      "step": 8693
+    },
+    {
+      "epoch": 0.5348179133858267,
+      "grad_norm": 1.1357979774475098,
+      "learning_rate": 8.917285144219017e-05,
+      "loss": 1.6619,
+      "step": 8694
+    },
+    {
+      "epoch": 0.5348794291338582,
+      "grad_norm": 1.1922743320465088,
+      "learning_rate": 8.915362768578889e-05,
+      "loss": 2.6699,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5349409448818898,
+      "grad_norm": 2.0312256813049316,
+      "learning_rate": 8.91344043349814e-05,
+      "loss": 2.2378,
+      "step": 8696
+    },
+    {
+      "epoch": 0.5350024606299213,
+      "grad_norm": 1.1583735942840576,
+      "learning_rate": 8.911518139048662e-05,
+      "loss": 2.6654,
+      "step": 8697
+    },
+    {
+      "epoch": 0.5350639763779528,
+      "grad_norm": 1.0277013778686523,
+      "learning_rate": 8.909595885302335e-05,
+      "loss": 2.578,
+      "step": 8698
+    },
+    {
+      "epoch": 0.5351254921259843,
+      "grad_norm": 1.5449285507202148,
+      "learning_rate": 8.907673672331042e-05,
+      "loss": 2.8018,
+      "step": 8699
+    },
+    {
+      "epoch": 0.5351870078740157,
+      "grad_norm": 2.082502841949463,
+      "learning_rate": 8.905751500206661e-05,
+      "loss": 1.187,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5352485236220472,
+      "grad_norm": 1.0419214963912964,
+      "learning_rate": 8.903829369001077e-05,
+      "loss": 2.7932,
+      "step": 8701
+    },
+    {
+      "epoch": 0.5353100393700787,
+      "grad_norm": 1.23361337184906,
+      "learning_rate": 8.901907278786159e-05,
+      "loss": 2.6081,
+      "step": 8702
+    },
+    {
+      "epoch": 0.5353715551181102,
+      "grad_norm": 1.1025867462158203,
+      "learning_rate": 8.899985229633785e-05,
+      "loss": 2.0586,
+      "step": 8703
+    },
+    {
+      "epoch": 0.5354330708661418,
+      "grad_norm": 1.0770543813705444,
+      "learning_rate": 8.898063221615831e-05,
+      "loss": 2.1116,
+      "step": 8704
+    },
+    {
+      "epoch": 0.5354945866141733,
+      "grad_norm": 1.3573089838027954,
+      "learning_rate": 8.89614125480417e-05,
+      "loss": 2.4265,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5355561023622047,
+      "grad_norm": 1.522608757019043,
+      "learning_rate": 8.894219329270666e-05,
+      "loss": 1.9504,
+      "step": 8706
+    },
+    {
+      "epoch": 0.5356176181102362,
+      "grad_norm": 1.2594385147094727,
+      "learning_rate": 8.892297445087196e-05,
+      "loss": 2.8405,
+      "step": 8707
+    },
+    {
+      "epoch": 0.5356791338582677,
+      "grad_norm": 1.2395623922348022,
+      "learning_rate": 8.890375602325623e-05,
+      "loss": 2.3778,
+      "step": 8708
+    },
+    {
+      "epoch": 0.5357406496062992,
+      "grad_norm": 1.386103630065918,
+      "learning_rate": 8.888453801057818e-05,
+      "loss": 2.6054,
+      "step": 8709
+    },
+    {
+      "epoch": 0.5358021653543307,
+      "grad_norm": 1.4219967126846313,
+      "learning_rate": 8.886532041355639e-05,
+      "loss": 2.3281,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5358636811023622,
+      "grad_norm": 1.295594573020935,
+      "learning_rate": 8.884610323290958e-05,
+      "loss": 2.5941,
+      "step": 8711
+    },
+    {
+      "epoch": 0.5359251968503937,
+      "grad_norm": 1.4325065612792969,
+      "learning_rate": 8.882688646935626e-05,
+      "loss": 1.929,
+      "step": 8712
+    },
+    {
+      "epoch": 0.5359867125984252,
+      "grad_norm": 1.1643036603927612,
+      "learning_rate": 8.880767012361512e-05,
+      "loss": 2.2398,
+      "step": 8713
+    },
+    {
+      "epoch": 0.5360482283464567,
+      "grad_norm": 1.5428601503372192,
+      "learning_rate": 8.878845419640467e-05,
+      "loss": 2.6883,
+      "step": 8714
+    },
+    {
+      "epoch": 0.5361097440944882,
+      "grad_norm": 1.215552568435669,
+      "learning_rate": 8.876923868844357e-05,
+      "loss": 2.0691,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5361712598425197,
+      "grad_norm": 1.0925817489624023,
+      "learning_rate": 8.875002360045024e-05,
+      "loss": 3.0829,
+      "step": 8716
+    },
+    {
+      "epoch": 0.5362327755905512,
+      "grad_norm": 1.0572662353515625,
+      "learning_rate": 8.873080893314337e-05,
+      "loss": 2.2625,
+      "step": 8717
+    },
+    {
+      "epoch": 0.5362942913385826,
+      "grad_norm": 1.3075571060180664,
+      "learning_rate": 8.871159468724137e-05,
+      "loss": 2.5958,
+      "step": 8718
+    },
+    {
+      "epoch": 0.5363558070866141,
+      "grad_norm": 1.0799065828323364,
+      "learning_rate": 8.86923808634628e-05,
+      "loss": 2.6461,
+      "step": 8719
+    },
+    {
+      "epoch": 0.5364173228346457,
+      "grad_norm": 1.2986656427383423,
+      "learning_rate": 8.86731674625261e-05,
+      "loss": 2.4144,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5364788385826772,
+      "grad_norm": 1.1050500869750977,
+      "learning_rate": 8.865395448514983e-05,
+      "loss": 2.8143,
+      "step": 8721
+    },
+    {
+      "epoch": 0.5365403543307087,
+      "grad_norm": 1.1311110258102417,
+      "learning_rate": 8.863474193205238e-05,
+      "loss": 2.5064,
+      "step": 8722
+    },
+    {
+      "epoch": 0.5366018700787402,
+      "grad_norm": 1.3338242769241333,
+      "learning_rate": 8.861552980395218e-05,
+      "loss": 2.8352,
+      "step": 8723
+    },
+    {
+      "epoch": 0.5366633858267716,
+      "grad_norm": 1.1519842147827148,
+      "learning_rate": 8.859631810156773e-05,
+      "loss": 2.3552,
+      "step": 8724
+    },
+    {
+      "epoch": 0.5367249015748031,
+      "grad_norm": 1.3415734767913818,
+      "learning_rate": 8.857710682561732e-05,
+      "loss": 2.6689,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5367864173228346,
+      "grad_norm": 1.3348820209503174,
+      "learning_rate": 8.855789597681948e-05,
+      "loss": 2.8116,
+      "step": 8726
+    },
+    {
+      "epoch": 0.5368479330708661,
+      "grad_norm": 1.2648341655731201,
+      "learning_rate": 8.85386855558925e-05,
+      "loss": 2.5658,
+      "step": 8727
+    },
+    {
+      "epoch": 0.5369094488188977,
+      "grad_norm": 1.190193772315979,
+      "learning_rate": 8.851947556355477e-05,
+      "loss": 2.5695,
+      "step": 8728
+    },
+    {
+      "epoch": 0.5369709645669292,
+      "grad_norm": 1.1894975900650024,
+      "learning_rate": 8.850026600052464e-05,
+      "loss": 2.4681,
+      "step": 8729
+    },
+    {
+      "epoch": 0.5370324803149606,
+      "grad_norm": 1.315069556236267,
+      "learning_rate": 8.848105686752047e-05,
+      "loss": 2.6477,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5370939960629921,
+      "grad_norm": 1.3082104921340942,
+      "learning_rate": 8.84618481652605e-05,
+      "loss": 2.6003,
+      "step": 8731
+    },
+    {
+      "epoch": 0.5371555118110236,
+      "grad_norm": 1.1898934841156006,
+      "learning_rate": 8.844263989446309e-05,
+      "loss": 1.2392,
+      "step": 8732
+    },
+    {
+      "epoch": 0.5372170275590551,
+      "grad_norm": 1.191184401512146,
+      "learning_rate": 8.842343205584648e-05,
+      "loss": 2.8707,
+      "step": 8733
+    },
+    {
+      "epoch": 0.5372785433070866,
+      "grad_norm": 1.3112155199050903,
+      "learning_rate": 8.840422465012899e-05,
+      "loss": 2.3376,
+      "step": 8734
+    },
+    {
+      "epoch": 0.5373400590551181,
+      "grad_norm": 1.071143627166748,
+      "learning_rate": 8.83850176780288e-05,
+      "loss": 2.9566,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5374015748031497,
+      "grad_norm": 1.6955660581588745,
+      "learning_rate": 8.836581114026423e-05,
+      "loss": 2.5195,
+      "step": 8736
+    },
+    {
+      "epoch": 0.5374630905511811,
+      "grad_norm": 1.2053282260894775,
+      "learning_rate": 8.834660503755341e-05,
+      "loss": 2.5577,
+      "step": 8737
+    },
+    {
+      "epoch": 0.5375246062992126,
+      "grad_norm": 1.2308149337768555,
+      "learning_rate": 8.83273993706146e-05,
+      "loss": 2.5982,
+      "step": 8738
+    },
+    {
+      "epoch": 0.5375861220472441,
+      "grad_norm": 1.3036760091781616,
+      "learning_rate": 8.830819414016595e-05,
+      "loss": 1.9577,
+      "step": 8739
+    },
+    {
+      "epoch": 0.5376476377952756,
+      "grad_norm": 1.2440825700759888,
+      "learning_rate": 8.828898934692568e-05,
+      "loss": 2.7833,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5377091535433071,
+      "grad_norm": 1.1302919387817383,
+      "learning_rate": 8.826978499161187e-05,
+      "loss": 2.3033,
+      "step": 8741
+    },
+    {
+      "epoch": 0.5377706692913385,
+      "grad_norm": 1.4574803113937378,
+      "learning_rate": 8.825058107494271e-05,
+      "loss": 2.164,
+      "step": 8742
+    },
+    {
+      "epoch": 0.53783218503937,
+      "grad_norm": 1.1294766664505005,
+      "learning_rate": 8.823137759763627e-05,
+      "loss": 2.6951,
+      "step": 8743
+    },
+    {
+      "epoch": 0.5378937007874016,
+      "grad_norm": 1.2023780345916748,
+      "learning_rate": 8.821217456041071e-05,
+      "loss": 2.6122,
+      "step": 8744
+    },
+    {
+      "epoch": 0.5379552165354331,
+      "grad_norm": 1.6604779958724976,
+      "learning_rate": 8.819297196398407e-05,
+      "loss": 2.3682,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5380167322834646,
+      "grad_norm": 1.1492531299591064,
+      "learning_rate": 8.817376980907449e-05,
+      "loss": 2.8851,
+      "step": 8746
+    },
+    {
+      "epoch": 0.5380782480314961,
+      "grad_norm": 1.2418662309646606,
+      "learning_rate": 8.815456809639994e-05,
+      "loss": 2.4325,
+      "step": 8747
+    },
+    {
+      "epoch": 0.5381397637795275,
+      "grad_norm": 1.302384614944458,
+      "learning_rate": 8.813536682667848e-05,
+      "loss": 2.4184,
+      "step": 8748
+    },
+    {
+      "epoch": 0.538201279527559,
+      "grad_norm": 1.2689741849899292,
+      "learning_rate": 8.811616600062815e-05,
+      "loss": 1.7461,
+      "step": 8749
+    },
+    {
+      "epoch": 0.5382627952755905,
+      "grad_norm": 1.3056378364562988,
+      "learning_rate": 8.809696561896699e-05,
+      "loss": 3.0308,
+      "step": 8750
+    },
+    {
+      "epoch": 0.538324311023622,
+      "grad_norm": 1.204616904258728,
+      "learning_rate": 8.807776568241286e-05,
+      "loss": 2.7322,
+      "step": 8751
+    },
+    {
+      "epoch": 0.5383858267716536,
+      "grad_norm": 1.539953351020813,
+      "learning_rate": 8.805856619168389e-05,
+      "loss": 2.58,
+      "step": 8752
+    },
+    {
+      "epoch": 0.5384473425196851,
+      "grad_norm": 1.1654621362686157,
+      "learning_rate": 8.803936714749791e-05,
+      "loss": 2.6001,
+      "step": 8753
+    },
+    {
+      "epoch": 0.5385088582677166,
+      "grad_norm": 1.2360860109329224,
+      "learning_rate": 8.802016855057293e-05,
+      "loss": 3.3628,
+      "step": 8754
+    },
+    {
+      "epoch": 0.538570374015748,
+      "grad_norm": 1.1913121938705444,
+      "learning_rate": 8.800097040162681e-05,
+      "loss": 2.8874,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5386318897637795,
+      "grad_norm": 1.31353759765625,
+      "learning_rate": 8.798177270137754e-05,
+      "loss": 2.7431,
+      "step": 8756
+    },
+    {
+      "epoch": 0.538693405511811,
+      "grad_norm": 1.595366358757019,
+      "learning_rate": 8.796257545054291e-05,
+      "loss": 2.5362,
+      "step": 8757
+    },
+    {
+      "epoch": 0.5387549212598425,
+      "grad_norm": 1.151187777519226,
+      "learning_rate": 8.794337864984086e-05,
+      "loss": 2.1266,
+      "step": 8758
+    },
+    {
+      "epoch": 0.538816437007874,
+      "grad_norm": 1.1926968097686768,
+      "learning_rate": 8.792418229998924e-05,
+      "loss": 3.003,
+      "step": 8759
+    },
+    {
+      "epoch": 0.5388779527559056,
+      "grad_norm": 1.276667833328247,
+      "learning_rate": 8.79049864017058e-05,
+      "loss": 2.7718,
+      "step": 8760
+    },
+    {
+      "epoch": 0.538939468503937,
+      "grad_norm": 1.3400651216506958,
+      "learning_rate": 8.788579095570848e-05,
+      "loss": 2.4434,
+      "step": 8761
+    },
+    {
+      "epoch": 0.5390009842519685,
+      "grad_norm": 1.1322771310806274,
+      "learning_rate": 8.786659596271501e-05,
+      "loss": 2.8799,
+      "step": 8762
+    },
+    {
+      "epoch": 0.5390625,
+      "grad_norm": 1.3699135780334473,
+      "learning_rate": 8.784740142344319e-05,
+      "loss": 2.8124,
+      "step": 8763
+    },
+    {
+      "epoch": 0.5391240157480315,
+      "grad_norm": 1.4645859003067017,
+      "learning_rate": 8.782820733861079e-05,
+      "loss": 2.1125,
+      "step": 8764
+    },
+    {
+      "epoch": 0.539185531496063,
+      "grad_norm": 1.1866832971572876,
+      "learning_rate": 8.78090137089356e-05,
+      "loss": 2.6953,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5392470472440944,
+      "grad_norm": 1.2163447141647339,
+      "learning_rate": 8.778982053513528e-05,
+      "loss": 2.7586,
+      "step": 8766
+    },
+    {
+      "epoch": 0.539308562992126,
+      "grad_norm": 1.0986062288284302,
+      "learning_rate": 8.777062781792759e-05,
+      "loss": 2.6793,
+      "step": 8767
+    },
+    {
+      "epoch": 0.5393700787401575,
+      "grad_norm": 1.1431279182434082,
+      "learning_rate": 8.775143555803022e-05,
+      "loss": 2.5949,
+      "step": 8768
+    },
+    {
+      "epoch": 0.539431594488189,
+      "grad_norm": 1.1309306621551514,
+      "learning_rate": 8.773224375616089e-05,
+      "loss": 2.4925,
+      "step": 8769
+    },
+    {
+      "epoch": 0.5394931102362205,
+      "grad_norm": 1.3577401638031006,
+      "learning_rate": 8.771305241303719e-05,
+      "loss": 2.8101,
+      "step": 8770
+    },
+    {
+      "epoch": 0.539554625984252,
+      "grad_norm": 1.3671467304229736,
+      "learning_rate": 8.769386152937687e-05,
+      "loss": 2.2669,
+      "step": 8771
+    },
+    {
+      "epoch": 0.5396161417322834,
+      "grad_norm": 1.4068055152893066,
+      "learning_rate": 8.767467110589747e-05,
+      "loss": 2.2595,
+      "step": 8772
+    },
+    {
+      "epoch": 0.5396776574803149,
+      "grad_norm": 1.445812463760376,
+      "learning_rate": 8.765548114331665e-05,
+      "loss": 3.2761,
+      "step": 8773
+    },
+    {
+      "epoch": 0.5397391732283464,
+      "grad_norm": 1.2876851558685303,
+      "learning_rate": 8.763629164235199e-05,
+      "loss": 2.6321,
+      "step": 8774
+    },
+    {
+      "epoch": 0.539800688976378,
+      "grad_norm": 1.1312310695648193,
+      "learning_rate": 8.761710260372112e-05,
+      "loss": 2.3617,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5398622047244095,
+      "grad_norm": 1.396596908569336,
+      "learning_rate": 8.759791402814154e-05,
+      "loss": 2.5527,
+      "step": 8776
+    },
+    {
+      "epoch": 0.539923720472441,
+      "grad_norm": 1.3008530139923096,
+      "learning_rate": 8.757872591633081e-05,
+      "loss": 2.4015,
+      "step": 8777
+    },
+    {
+      "epoch": 0.5399852362204725,
+      "grad_norm": 1.4616023302078247,
+      "learning_rate": 8.755953826900647e-05,
+      "loss": 2.5041,
+      "step": 8778
+    },
+    {
+      "epoch": 0.5400467519685039,
+      "grad_norm": 1.3740216493606567,
+      "learning_rate": 8.754035108688604e-05,
+      "loss": 2.6614,
+      "step": 8779
+    },
+    {
+      "epoch": 0.5401082677165354,
+      "grad_norm": 1.2039421796798706,
+      "learning_rate": 8.752116437068699e-05,
+      "loss": 2.9102,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5401697834645669,
+      "grad_norm": 1.2440944910049438,
+      "learning_rate": 8.750197812112683e-05,
+      "loss": 1.7401,
+      "step": 8781
+    },
+    {
+      "epoch": 0.5402312992125984,
+      "grad_norm": 1.3347936868667603,
+      "learning_rate": 8.748279233892297e-05,
+      "loss": 2.6231,
+      "step": 8782
+    },
+    {
+      "epoch": 0.54029281496063,
+      "grad_norm": 1.1406652927398682,
+      "learning_rate": 8.74636070247929e-05,
+      "loss": 2.8069,
+      "step": 8783
+    },
+    {
+      "epoch": 0.5403543307086615,
+      "grad_norm": 1.8313581943511963,
+      "learning_rate": 8.7444422179454e-05,
+      "loss": 2.3731,
+      "step": 8784
+    },
+    {
+      "epoch": 0.5404158464566929,
+      "grad_norm": 0.9891850352287292,
+      "learning_rate": 8.742523780362374e-05,
+      "loss": 2.4648,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5404773622047244,
+      "grad_norm": 1.1830307245254517,
+      "learning_rate": 8.74060538980194e-05,
+      "loss": 2.0263,
+      "step": 8786
+    },
+    {
+      "epoch": 0.5405388779527559,
+      "grad_norm": 1.244865894317627,
+      "learning_rate": 8.738687046335849e-05,
+      "loss": 2.3886,
+      "step": 8787
+    },
+    {
+      "epoch": 0.5406003937007874,
+      "grad_norm": 1.0955219268798828,
+      "learning_rate": 8.736768750035824e-05,
+      "loss": 2.4725,
+      "step": 8788
+    },
+    {
+      "epoch": 0.5406619094488189,
+      "grad_norm": 1.0487056970596313,
+      "learning_rate": 8.734850500973607e-05,
+      "loss": 1.9832,
+      "step": 8789
+    },
+    {
+      "epoch": 0.5407234251968503,
+      "grad_norm": 1.284984827041626,
+      "learning_rate": 8.732932299220924e-05,
+      "loss": 2.5591,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5407849409448819,
+      "grad_norm": 1.1824369430541992,
+      "learning_rate": 8.731014144849512e-05,
+      "loss": 2.7972,
+      "step": 8791
+    },
+    {
+      "epoch": 0.5408464566929134,
+      "grad_norm": 1.1711713075637817,
+      "learning_rate": 8.729096037931091e-05,
+      "loss": 2.2159,
+      "step": 8792
+    },
+    {
+      "epoch": 0.5409079724409449,
+      "grad_norm": 1.2477095127105713,
+      "learning_rate": 8.727177978537393e-05,
+      "loss": 2.5586,
+      "step": 8793
+    },
+    {
+      "epoch": 0.5409694881889764,
+      "grad_norm": 1.3166677951812744,
+      "learning_rate": 8.725259966740139e-05,
+      "loss": 3.6604,
+      "step": 8794
+    },
+    {
+      "epoch": 0.5410310039370079,
+      "grad_norm": 1.0680614709854126,
+      "learning_rate": 8.72334200261106e-05,
+      "loss": 3.1785,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5410925196850394,
+      "grad_norm": 1.2459962368011475,
+      "learning_rate": 8.721424086221866e-05,
+      "loss": 2.403,
+      "step": 8796
+    },
+    {
+      "epoch": 0.5411540354330708,
+      "grad_norm": 1.3241482973098755,
+      "learning_rate": 8.719506217644281e-05,
+      "loss": 2.0946,
+      "step": 8797
+    },
+    {
+      "epoch": 0.5412155511811023,
+      "grad_norm": 1.3573532104492188,
+      "learning_rate": 8.717588396950025e-05,
+      "loss": 2.4795,
+      "step": 8798
+    },
+    {
+      "epoch": 0.5412770669291339,
+      "grad_norm": 1.4977253675460815,
+      "learning_rate": 8.71567062421081e-05,
+      "loss": 1.9989,
+      "step": 8799
+    },
+    {
+      "epoch": 0.5413385826771654,
+      "grad_norm": 1.2161363363265991,
+      "learning_rate": 8.713752899498358e-05,
+      "loss": 1.9574,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5414000984251969,
+      "grad_norm": 1.1256977319717407,
+      "learning_rate": 8.71183522288437e-05,
+      "loss": 2.6161,
+      "step": 8801
+    },
+    {
+      "epoch": 0.5414616141732284,
+      "grad_norm": 1.2889803647994995,
+      "learning_rate": 8.709917594440562e-05,
+      "loss": 1.982,
+      "step": 8802
+    },
+    {
+      "epoch": 0.5415231299212598,
+      "grad_norm": 1.0631427764892578,
+      "learning_rate": 8.708000014238641e-05,
+      "loss": 2.9272,
+      "step": 8803
+    },
+    {
+      "epoch": 0.5415846456692913,
+      "grad_norm": 1.1400431394577026,
+      "learning_rate": 8.706082482350322e-05,
+      "loss": 2.5712,
+      "step": 8804
+    },
+    {
+      "epoch": 0.5416461614173228,
+      "grad_norm": 1.0889811515808105,
+      "learning_rate": 8.704164998847294e-05,
+      "loss": 2.3363,
+      "step": 8805
+    },
+    {
+      "epoch": 0.5417076771653543,
+      "grad_norm": 1.1782684326171875,
+      "learning_rate": 8.702247563801276e-05,
+      "loss": 2.2689,
+      "step": 8806
+    },
+    {
+      "epoch": 0.5417691929133859,
+      "grad_norm": 1.3528907299041748,
+      "learning_rate": 8.700330177283958e-05,
+      "loss": 2.6836,
+      "step": 8807
+    },
+    {
+      "epoch": 0.5418307086614174,
+      "grad_norm": 1.5802597999572754,
+      "learning_rate": 8.698412839367047e-05,
+      "loss": 2.1928,
+      "step": 8808
+    },
+    {
+      "epoch": 0.5418922244094488,
+      "grad_norm": 1.1380723714828491,
+      "learning_rate": 8.696495550122236e-05,
+      "loss": 2.8446,
+      "step": 8809
+    },
+    {
+      "epoch": 0.5419537401574803,
+      "grad_norm": 1.0666948556900024,
+      "learning_rate": 8.694578309621228e-05,
+      "loss": 3.0364,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5420152559055118,
+      "grad_norm": 1.3439710140228271,
+      "learning_rate": 8.692661117935707e-05,
+      "loss": 3.287,
+      "step": 8811
+    },
+    {
+      "epoch": 0.5420767716535433,
+      "grad_norm": 1.115654706954956,
+      "learning_rate": 8.690743975137372e-05,
+      "loss": 2.4435,
+      "step": 8812
+    },
+    {
+      "epoch": 0.5421382874015748,
+      "grad_norm": 1.191228985786438,
+      "learning_rate": 8.688826881297909e-05,
+      "loss": 2.2531,
+      "step": 8813
+    },
+    {
+      "epoch": 0.5421998031496063,
+      "grad_norm": 1.1720094680786133,
+      "learning_rate": 8.686909836489012e-05,
+      "loss": 3.1772,
+      "step": 8814
+    },
+    {
+      "epoch": 0.5422613188976378,
+      "grad_norm": 1.3444079160690308,
+      "learning_rate": 8.684992840782364e-05,
+      "loss": 2.3018,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5423228346456693,
+      "grad_norm": 1.3302099704742432,
+      "learning_rate": 8.683075894249654e-05,
+      "loss": 2.2765,
+      "step": 8816
+    },
+    {
+      "epoch": 0.5423843503937008,
+      "grad_norm": 1.3410534858703613,
+      "learning_rate": 8.681158996962558e-05,
+      "loss": 1.828,
+      "step": 8817
+    },
+    {
+      "epoch": 0.5424458661417323,
+      "grad_norm": 1.319563388824463,
+      "learning_rate": 8.679242148992764e-05,
+      "loss": 2.0707,
+      "step": 8818
+    },
+    {
+      "epoch": 0.5425073818897638,
+      "grad_norm": 1.167117953300476,
+      "learning_rate": 8.677325350411946e-05,
+      "loss": 2.3534,
+      "step": 8819
+    },
+    {
+      "epoch": 0.5425688976377953,
+      "grad_norm": 1.3190840482711792,
+      "learning_rate": 8.67540860129179e-05,
+      "loss": 2.6919,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5426304133858267,
+      "grad_norm": 1.1833274364471436,
+      "learning_rate": 8.673491901703962e-05,
+      "loss": 2.9029,
+      "step": 8821
+    },
+    {
+      "epoch": 0.5426919291338582,
+      "grad_norm": 1.1940557956695557,
+      "learning_rate": 8.671575251720141e-05,
+      "loss": 2.205,
+      "step": 8822
+    },
+    {
+      "epoch": 0.5427534448818898,
+      "grad_norm": 1.2579244375228882,
+      "learning_rate": 8.669658651411997e-05,
+      "loss": 2.2826,
+      "step": 8823
+    },
+    {
+      "epoch": 0.5428149606299213,
+      "grad_norm": 1.2627367973327637,
+      "learning_rate": 8.667742100851203e-05,
+      "loss": 2.4572,
+      "step": 8824
+    },
+    {
+      "epoch": 0.5428764763779528,
+      "grad_norm": 1.113911509513855,
+      "learning_rate": 8.665825600109423e-05,
+      "loss": 2.4382,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5429379921259843,
+      "grad_norm": 1.924681544303894,
+      "learning_rate": 8.663909149258331e-05,
+      "loss": 2.2549,
+      "step": 8826
+    },
+    {
+      "epoch": 0.5429995078740157,
+      "grad_norm": 1.1472160816192627,
+      "learning_rate": 8.661992748369582e-05,
+      "loss": 3.1781,
+      "step": 8827
+    },
+    {
+      "epoch": 0.5430610236220472,
+      "grad_norm": 1.1975409984588623,
+      "learning_rate": 8.660076397514845e-05,
+      "loss": 2.5742,
+      "step": 8828
+    },
+    {
+      "epoch": 0.5431225393700787,
+      "grad_norm": 1.2177575826644897,
+      "learning_rate": 8.658160096765778e-05,
+      "loss": 2.331,
+      "step": 8829
+    },
+    {
+      "epoch": 0.5431840551181102,
+      "grad_norm": 1.0653860569000244,
+      "learning_rate": 8.656243846194044e-05,
+      "loss": 2.2039,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5432455708661418,
+      "grad_norm": 1.1382089853286743,
+      "learning_rate": 8.654327645871292e-05,
+      "loss": 2.6102,
+      "step": 8831
+    },
+    {
+      "epoch": 0.5433070866141733,
+      "grad_norm": 1.118179440498352,
+      "learning_rate": 8.652411495869189e-05,
+      "loss": 2.0475,
+      "step": 8832
+    },
+    {
+      "epoch": 0.5433686023622047,
+      "grad_norm": 1.3152010440826416,
+      "learning_rate": 8.650495396259377e-05,
+      "loss": 2.5908,
+      "step": 8833
+    },
+    {
+      "epoch": 0.5434301181102362,
+      "grad_norm": 1.2555575370788574,
+      "learning_rate": 8.648579347113512e-05,
+      "loss": 3.0015,
+      "step": 8834
+    },
+    {
+      "epoch": 0.5434916338582677,
+      "grad_norm": 1.1519676446914673,
+      "learning_rate": 8.646663348503248e-05,
+      "loss": 2.4258,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5435531496062992,
+      "grad_norm": 1.241322636604309,
+      "learning_rate": 8.644747400500224e-05,
+      "loss": 2.844,
+      "step": 8836
+    },
+    {
+      "epoch": 0.5436146653543307,
+      "grad_norm": 1.3633884191513062,
+      "learning_rate": 8.642831503176092e-05,
+      "loss": 2.9985,
+      "step": 8837
+    },
+    {
+      "epoch": 0.5436761811023622,
+      "grad_norm": 1.6306813955307007,
+      "learning_rate": 8.640915656602492e-05,
+      "loss": 2.421,
+      "step": 8838
+    },
+    {
+      "epoch": 0.5437376968503937,
+      "grad_norm": 1.2216682434082031,
+      "learning_rate": 8.638999860851072e-05,
+      "loss": 2.919,
+      "step": 8839
+    },
+    {
+      "epoch": 0.5437992125984252,
+      "grad_norm": 1.8269500732421875,
+      "learning_rate": 8.637084115993461e-05,
+      "loss": 2.3839,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5438607283464567,
+      "grad_norm": 1.6817045211791992,
+      "learning_rate": 8.635168422101311e-05,
+      "loss": 2.1631,
+      "step": 8841
+    },
+    {
+      "epoch": 0.5439222440944882,
+      "grad_norm": 1.14164400100708,
+      "learning_rate": 8.63325277924625e-05,
+      "loss": 3.2126,
+      "step": 8842
+    },
+    {
+      "epoch": 0.5439837598425197,
+      "grad_norm": 1.1335910558700562,
+      "learning_rate": 8.631337187499912e-05,
+      "loss": 2.5257,
+      "step": 8843
+    },
+    {
+      "epoch": 0.5440452755905512,
+      "grad_norm": 1.209763765335083,
+      "learning_rate": 8.629421646933932e-05,
+      "loss": 2.9599,
+      "step": 8844
+    },
+    {
+      "epoch": 0.5441067913385826,
+      "grad_norm": 1.152488350868225,
+      "learning_rate": 8.627506157619942e-05,
+      "loss": 1.6802,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5441683070866141,
+      "grad_norm": 1.160421371459961,
+      "learning_rate": 8.625590719629566e-05,
+      "loss": 2.8426,
+      "step": 8846
+    },
+    {
+      "epoch": 0.5442298228346457,
+      "grad_norm": 1.1597630977630615,
+      "learning_rate": 8.623675333034435e-05,
+      "loss": 2.5847,
+      "step": 8847
+    },
+    {
+      "epoch": 0.5442913385826772,
+      "grad_norm": 1.062598466873169,
+      "learning_rate": 8.621759997906168e-05,
+      "loss": 2.4286,
+      "step": 8848
+    },
+    {
+      "epoch": 0.5443528543307087,
+      "grad_norm": 1.049414038658142,
+      "learning_rate": 8.619844714316396e-05,
+      "loss": 2.6022,
+      "step": 8849
+    },
+    {
+      "epoch": 0.5444143700787402,
+      "grad_norm": 1.1322722434997559,
+      "learning_rate": 8.617929482336733e-05,
+      "loss": 3.2339,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5444758858267716,
+      "grad_norm": 1.2667335271835327,
+      "learning_rate": 8.616014302038804e-05,
+      "loss": 2.3283,
+      "step": 8851
+    },
+    {
+      "epoch": 0.5445374015748031,
+      "grad_norm": 1.098078966140747,
+      "learning_rate": 8.61409917349422e-05,
+      "loss": 2.2445,
+      "step": 8852
+    },
+    {
+      "epoch": 0.5445989173228346,
+      "grad_norm": 1.0120551586151123,
+      "learning_rate": 8.6121840967746e-05,
+      "loss": 2.7071,
+      "step": 8853
+    },
+    {
+      "epoch": 0.5446604330708661,
+      "grad_norm": 1.1079785823822021,
+      "learning_rate": 8.610269071951554e-05,
+      "loss": 2.4747,
+      "step": 8854
+    },
+    {
+      "epoch": 0.5447219488188977,
+      "grad_norm": 1.1735355854034424,
+      "learning_rate": 8.6083540990967e-05,
+      "loss": 2.4636,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5447834645669292,
+      "grad_norm": 1.4464056491851807,
+      "learning_rate": 8.606439178281639e-05,
+      "loss": 2.623,
+      "step": 8856
+    },
+    {
+      "epoch": 0.5448449803149606,
+      "grad_norm": 1.005497932434082,
+      "learning_rate": 8.604524309577984e-05,
+      "loss": 2.7395,
+      "step": 8857
+    },
+    {
+      "epoch": 0.5449064960629921,
+      "grad_norm": 1.2298945188522339,
+      "learning_rate": 8.602609493057335e-05,
+      "loss": 2.6381,
+      "step": 8858
+    },
+    {
+      "epoch": 0.5449680118110236,
+      "grad_norm": 1.3788515329360962,
+      "learning_rate": 8.600694728791302e-05,
+      "loss": 2.9211,
+      "step": 8859
+    },
+    {
+      "epoch": 0.5450295275590551,
+      "grad_norm": 1.2841507196426392,
+      "learning_rate": 8.598780016851481e-05,
+      "loss": 2.0844,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5450910433070866,
+      "grad_norm": 1.2354365587234497,
+      "learning_rate": 8.596865357309479e-05,
+      "loss": 2.5429,
+      "step": 8861
+    },
+    {
+      "epoch": 0.5451525590551181,
+      "grad_norm": 1.1815465688705444,
+      "learning_rate": 8.594950750236884e-05,
+      "loss": 2.67,
+      "step": 8862
+    },
+    {
+      "epoch": 0.5452140748031497,
+      "grad_norm": 1.134177565574646,
+      "learning_rate": 8.593036195705298e-05,
+      "loss": 1.7197,
+      "step": 8863
+    },
+    {
+      "epoch": 0.5452755905511811,
+      "grad_norm": 1.528662919998169,
+      "learning_rate": 8.59112169378631e-05,
+      "loss": 2.9006,
+      "step": 8864
+    },
+    {
+      "epoch": 0.5453371062992126,
+      "grad_norm": 1.3675415515899658,
+      "learning_rate": 8.589207244551519e-05,
+      "loss": 1.9531,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5453986220472441,
+      "grad_norm": 1.335365653038025,
+      "learning_rate": 8.587292848072505e-05,
+      "loss": 2.126,
+      "step": 8866
+    },
+    {
+      "epoch": 0.5454601377952756,
+      "grad_norm": 1.3274824619293213,
+      "learning_rate": 8.585378504420867e-05,
+      "loss": 2.8865,
+      "step": 8867
+    },
+    {
+      "epoch": 0.5455216535433071,
+      "grad_norm": 1.1278105974197388,
+      "learning_rate": 8.583464213668179e-05,
+      "loss": 2.0499,
+      "step": 8868
+    },
+    {
+      "epoch": 0.5455831692913385,
+      "grad_norm": 1.3276245594024658,
+      "learning_rate": 8.581549975886034e-05,
+      "loss": 2.486,
+      "step": 8869
+    },
+    {
+      "epoch": 0.54564468503937,
+      "grad_norm": 1.2530063390731812,
+      "learning_rate": 8.579635791146013e-05,
+      "loss": 3.3212,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5457062007874016,
+      "grad_norm": 1.1658209562301636,
+      "learning_rate": 8.577721659519689e-05,
+      "loss": 2.8086,
+      "step": 8871
+    },
+    {
+      "epoch": 0.5457677165354331,
+      "grad_norm": 1.244845986366272,
+      "learning_rate": 8.575807581078646e-05,
+      "loss": 2.5472,
+      "step": 8872
+    },
+    {
+      "epoch": 0.5458292322834646,
+      "grad_norm": 1.2484718561172485,
+      "learning_rate": 8.573893555894455e-05,
+      "loss": 1.8304,
+      "step": 8873
+    },
+    {
+      "epoch": 0.5458907480314961,
+      "grad_norm": 1.6217328310012817,
+      "learning_rate": 8.571979584038697e-05,
+      "loss": 1.8316,
+      "step": 8874
+    },
+    {
+      "epoch": 0.5459522637795275,
+      "grad_norm": 1.1605002880096436,
+      "learning_rate": 8.570065665582934e-05,
+      "loss": 3.0929,
+      "step": 8875
+    },
+    {
+      "epoch": 0.546013779527559,
+      "grad_norm": 1.266587495803833,
+      "learning_rate": 8.568151800598748e-05,
+      "loss": 2.6115,
+      "step": 8876
+    },
+    {
+      "epoch": 0.5460752952755905,
+      "grad_norm": 1.0485714673995972,
+      "learning_rate": 8.566237989157696e-05,
+      "loss": 2.6441,
+      "step": 8877
+    },
+    {
+      "epoch": 0.546136811023622,
+      "grad_norm": 1.2696110010147095,
+      "learning_rate": 8.56432423133135e-05,
+      "loss": 2.4683,
+      "step": 8878
+    },
+    {
+      "epoch": 0.5461983267716536,
+      "grad_norm": 1.0195101499557495,
+      "learning_rate": 8.562410527191271e-05,
+      "loss": 2.0071,
+      "step": 8879
+    },
+    {
+      "epoch": 0.5462598425196851,
+      "grad_norm": 1.2238309383392334,
+      "learning_rate": 8.560496876809026e-05,
+      "loss": 2.2759,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5463213582677166,
+      "grad_norm": 1.320476770401001,
+      "learning_rate": 8.558583280256167e-05,
+      "loss": 2.8653,
+      "step": 8881
+    },
+    {
+      "epoch": 0.546382874015748,
+      "grad_norm": 1.231910228729248,
+      "learning_rate": 8.556669737604257e-05,
+      "loss": 2.1107,
+      "step": 8882
+    },
+    {
+      "epoch": 0.5464443897637795,
+      "grad_norm": 1.4029037952423096,
+      "learning_rate": 8.554756248924848e-05,
+      "loss": 2.5879,
+      "step": 8883
+    },
+    {
+      "epoch": 0.546505905511811,
+      "grad_norm": 1.3876440525054932,
+      "learning_rate": 8.552842814289501e-05,
+      "loss": 2.9959,
+      "step": 8884
+    },
+    {
+      "epoch": 0.5465674212598425,
+      "grad_norm": 1.2476789951324463,
+      "learning_rate": 8.550929433769759e-05,
+      "loss": 2.405,
+      "step": 8885
+    },
+    {
+      "epoch": 0.546628937007874,
+      "grad_norm": 1.1562936305999756,
+      "learning_rate": 8.549016107437181e-05,
+      "loss": 2.3286,
+      "step": 8886
+    },
+    {
+      "epoch": 0.5466904527559056,
+      "grad_norm": 1.260839581489563,
+      "learning_rate": 8.547102835363305e-05,
+      "loss": 2.733,
+      "step": 8887
+    },
+    {
+      "epoch": 0.546751968503937,
+      "grad_norm": 1.1915110349655151,
+      "learning_rate": 8.545189617619684e-05,
+      "loss": 2.7281,
+      "step": 8888
+    },
+    {
+      "epoch": 0.5468134842519685,
+      "grad_norm": 1.2935409545898438,
+      "learning_rate": 8.543276454277856e-05,
+      "loss": 2.9694,
+      "step": 8889
+    },
+    {
+      "epoch": 0.546875,
+      "grad_norm": 1.262475609779358,
+      "learning_rate": 8.541363345409371e-05,
+      "loss": 3.0457,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5469365157480315,
+      "grad_norm": 1.2563663721084595,
+      "learning_rate": 8.539450291085759e-05,
+      "loss": 2.8769,
+      "step": 8891
+    },
+    {
+      "epoch": 0.546998031496063,
+      "grad_norm": 1.3513610363006592,
+      "learning_rate": 8.537537291378563e-05,
+      "loss": 2.5656,
+      "step": 8892
+    },
+    {
+      "epoch": 0.5470595472440944,
+      "grad_norm": 1.0964088439941406,
+      "learning_rate": 8.535624346359316e-05,
+      "loss": 2.6733,
+      "step": 8893
+    },
+    {
+      "epoch": 0.547121062992126,
+      "grad_norm": 1.100426197052002,
+      "learning_rate": 8.533711456099555e-05,
+      "loss": 2.5016,
+      "step": 8894
+    },
+    {
+      "epoch": 0.5471825787401575,
+      "grad_norm": 1.2670143842697144,
+      "learning_rate": 8.531798620670806e-05,
+      "loss": 2.5875,
+      "step": 8895
+    },
+    {
+      "epoch": 0.547244094488189,
+      "grad_norm": 1.2104041576385498,
+      "learning_rate": 8.529885840144607e-05,
+      "loss": 2.4344,
+      "step": 8896
+    },
+    {
+      "epoch": 0.5473056102362205,
+      "grad_norm": 1.3455275297164917,
+      "learning_rate": 8.527973114592476e-05,
+      "loss": 2.6397,
+      "step": 8897
+    },
+    {
+      "epoch": 0.547367125984252,
+      "grad_norm": 1.207993507385254,
+      "learning_rate": 8.526060444085943e-05,
+      "loss": 2.8344,
+      "step": 8898
+    },
+    {
+      "epoch": 0.5474286417322834,
+      "grad_norm": 1.2191689014434814,
+      "learning_rate": 8.524147828696531e-05,
+      "loss": 2.2603,
+      "step": 8899
+    },
+    {
+      "epoch": 0.5474901574803149,
+      "grad_norm": 1.2136601209640503,
+      "learning_rate": 8.522235268495763e-05,
+      "loss": 2.8025,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5475516732283464,
+      "grad_norm": 1.1354647874832153,
+      "learning_rate": 8.52032276355515e-05,
+      "loss": 1.7477,
+      "step": 8901
+    },
+    {
+      "epoch": 0.547613188976378,
+      "grad_norm": 1.1718404293060303,
+      "learning_rate": 8.51841031394622e-05,
+      "loss": 2.3827,
+      "step": 8902
+    },
+    {
+      "epoch": 0.5476747047244095,
+      "grad_norm": 1.1321096420288086,
+      "learning_rate": 8.516497919740479e-05,
+      "loss": 2.1999,
+      "step": 8903
+    },
+    {
+      "epoch": 0.547736220472441,
+      "grad_norm": 1.2326244115829468,
+      "learning_rate": 8.514585581009446e-05,
+      "loss": 2.7026,
+      "step": 8904
+    },
+    {
+      "epoch": 0.5477977362204725,
+      "grad_norm": 1.1679073572158813,
+      "learning_rate": 8.512673297824632e-05,
+      "loss": 3.1869,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5478592519685039,
+      "grad_norm": 1.3687595129013062,
+      "learning_rate": 8.510761070257538e-05,
+      "loss": 2.5087,
+      "step": 8906
+    },
+    {
+      "epoch": 0.5479207677165354,
+      "grad_norm": 1.0431780815124512,
+      "learning_rate": 8.508848898379678e-05,
+      "loss": 2.5412,
+      "step": 8907
+    },
+    {
+      "epoch": 0.5479822834645669,
+      "grad_norm": 1.2798442840576172,
+      "learning_rate": 8.506936782262552e-05,
+      "loss": 2.1717,
+      "step": 8908
+    },
+    {
+      "epoch": 0.5480437992125984,
+      "grad_norm": 1.076757550239563,
+      "learning_rate": 8.50502472197767e-05,
+      "loss": 2.5217,
+      "step": 8909
+    },
+    {
+      "epoch": 0.54810531496063,
+      "grad_norm": 1.3898628950119019,
+      "learning_rate": 8.50311271759652e-05,
+      "loss": 3.2662,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5481668307086615,
+      "grad_norm": 1.273162603378296,
+      "learning_rate": 8.501200769190613e-05,
+      "loss": 2.3234,
+      "step": 8911
+    },
+    {
+      "epoch": 0.5482283464566929,
+      "grad_norm": 8.415121078491211,
+      "learning_rate": 8.499288876831437e-05,
+      "loss": 2.7598,
+      "step": 8912
+    },
+    {
+      "epoch": 0.5482898622047244,
+      "grad_norm": 1.2618712186813354,
+      "learning_rate": 8.497377040590491e-05,
+      "loss": 2.3583,
+      "step": 8913
+    },
+    {
+      "epoch": 0.5483513779527559,
+      "grad_norm": 1.1365240812301636,
+      "learning_rate": 8.495465260539262e-05,
+      "loss": 2.9178,
+      "step": 8914
+    },
+    {
+      "epoch": 0.5484128937007874,
+      "grad_norm": 1.409030556678772,
+      "learning_rate": 8.493553536749246e-05,
+      "loss": 2.3188,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5484744094488189,
+      "grad_norm": 1.249393343925476,
+      "learning_rate": 8.491641869291924e-05,
+      "loss": 2.4551,
+      "step": 8916
+    },
+    {
+      "epoch": 0.5485359251968503,
+      "grad_norm": 1.1318520307540894,
+      "learning_rate": 8.489730258238788e-05,
+      "loss": 2.4964,
+      "step": 8917
+    },
+    {
+      "epoch": 0.5485974409448819,
+      "grad_norm": 1.1137759685516357,
+      "learning_rate": 8.487818703661316e-05,
+      "loss": 2.9299,
+      "step": 8918
+    },
+    {
+      "epoch": 0.5486589566929134,
+      "grad_norm": 1.7108560800552368,
+      "learning_rate": 8.485907205630997e-05,
+      "loss": 2.3074,
+      "step": 8919
+    },
+    {
+      "epoch": 0.5487204724409449,
+      "grad_norm": 1.0905139446258545,
+      "learning_rate": 8.483995764219301e-05,
+      "loss": 2.6654,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5487819881889764,
+      "grad_norm": 1.267396330833435,
+      "learning_rate": 8.482084379497713e-05,
+      "loss": 2.5443,
+      "step": 8921
+    },
+    {
+      "epoch": 0.5488435039370079,
+      "grad_norm": 1.2899881601333618,
+      "learning_rate": 8.480173051537703e-05,
+      "loss": 2.6449,
+      "step": 8922
+    },
+    {
+      "epoch": 0.5489050196850394,
+      "grad_norm": 1.0880955457687378,
+      "learning_rate": 8.478261780410748e-05,
+      "loss": 2.9376,
+      "step": 8923
+    },
+    {
+      "epoch": 0.5489665354330708,
+      "grad_norm": 1.505352258682251,
+      "learning_rate": 8.476350566188315e-05,
+      "loss": 2.3144,
+      "step": 8924
+    },
+    {
+      "epoch": 0.5490280511811023,
+      "grad_norm": 1.4679211378097534,
+      "learning_rate": 8.474439408941877e-05,
+      "loss": 2.931,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5490895669291339,
+      "grad_norm": 1.1554590463638306,
+      "learning_rate": 8.472528308742896e-05,
+      "loss": 2.4139,
+      "step": 8926
+    },
+    {
+      "epoch": 0.5491510826771654,
+      "grad_norm": 1.3823968172073364,
+      "learning_rate": 8.470617265662841e-05,
+      "loss": 2.1436,
+      "step": 8927
+    },
+    {
+      "epoch": 0.5492125984251969,
+      "grad_norm": 1.4095230102539062,
+      "learning_rate": 8.468706279773169e-05,
+      "loss": 2.6265,
+      "step": 8928
+    },
+    {
+      "epoch": 0.5492741141732284,
+      "grad_norm": 1.2220537662506104,
+      "learning_rate": 8.466795351145346e-05,
+      "loss": 3.1962,
+      "step": 8929
+    },
+    {
+      "epoch": 0.5493356299212598,
+      "grad_norm": 1.1394132375717163,
+      "learning_rate": 8.464884479850827e-05,
+      "loss": 2.5544,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5493971456692913,
+      "grad_norm": 1.2805854082107544,
+      "learning_rate": 8.46297366596107e-05,
+      "loss": 2.3305,
+      "step": 8931
+    },
+    {
+      "epoch": 0.5494586614173228,
+      "grad_norm": 1.1021754741668701,
+      "learning_rate": 8.461062909547525e-05,
+      "loss": 3.0336,
+      "step": 8932
+    },
+    {
+      "epoch": 0.5495201771653543,
+      "grad_norm": 1.311874270439148,
+      "learning_rate": 8.459152210681647e-05,
+      "loss": 2.3266,
+      "step": 8933
+    },
+    {
+      "epoch": 0.5495816929133859,
+      "grad_norm": 1.3347464799880981,
+      "learning_rate": 8.457241569434882e-05,
+      "loss": 2.6929,
+      "step": 8934
+    },
+    {
+      "epoch": 0.5496432086614174,
+      "grad_norm": 1.1930166482925415,
+      "learning_rate": 8.455330985878685e-05,
+      "loss": 2.7351,
+      "step": 8935
+    },
+    {
+      "epoch": 0.5497047244094488,
+      "grad_norm": 1.7373169660568237,
+      "learning_rate": 8.453420460084488e-05,
+      "loss": 1.9673,
+      "step": 8936
+    },
+    {
+      "epoch": 0.5497662401574803,
+      "grad_norm": 1.0834078788757324,
+      "learning_rate": 8.451509992123749e-05,
+      "loss": 2.597,
+      "step": 8937
+    },
+    {
+      "epoch": 0.5498277559055118,
+      "grad_norm": 1.3639235496520996,
+      "learning_rate": 8.449599582067898e-05,
+      "loss": 2.785,
+      "step": 8938
+    },
+    {
+      "epoch": 0.5498892716535433,
+      "grad_norm": 1.3027256727218628,
+      "learning_rate": 8.447689229988378e-05,
+      "loss": 2.2694,
+      "step": 8939
+    },
+    {
+      "epoch": 0.5499507874015748,
+      "grad_norm": 1.1252803802490234,
+      "learning_rate": 8.445778935956623e-05,
+      "loss": 2.7725,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5500123031496063,
+      "grad_norm": 1.6886695623397827,
+      "learning_rate": 8.443868700044074e-05,
+      "loss": 1.9403,
+      "step": 8941
+    },
+    {
+      "epoch": 0.5500738188976378,
+      "grad_norm": 1.4341826438903809,
+      "learning_rate": 8.441958522322156e-05,
+      "loss": 2.9156,
+      "step": 8942
+    },
+    {
+      "epoch": 0.5501353346456693,
+      "grad_norm": 1.0841460227966309,
+      "learning_rate": 8.4400484028623e-05,
+      "loss": 2.2835,
+      "step": 8943
+    },
+    {
+      "epoch": 0.5501968503937008,
+      "grad_norm": 1.2918897867202759,
+      "learning_rate": 8.438138341735939e-05,
+      "loss": 2.2255,
+      "step": 8944
+    },
+    {
+      "epoch": 0.5502583661417323,
+      "grad_norm": 1.1002864837646484,
+      "learning_rate": 8.436228339014492e-05,
+      "loss": 2.9938,
+      "step": 8945
+    },
+    {
+      "epoch": 0.5503198818897638,
+      "grad_norm": 1.157605767250061,
+      "learning_rate": 8.434318394769384e-05,
+      "loss": 2.1901,
+      "step": 8946
+    },
+    {
+      "epoch": 0.5503813976377953,
+      "grad_norm": 1.3631093502044678,
+      "learning_rate": 8.432408509072039e-05,
+      "loss": 2.0646,
+      "step": 8947
+    },
+    {
+      "epoch": 0.5504429133858267,
+      "grad_norm": 1.6524105072021484,
+      "learning_rate": 8.430498681993875e-05,
+      "loss": 2.0777,
+      "step": 8948
+    },
+    {
+      "epoch": 0.5505044291338582,
+      "grad_norm": 1.0770889520645142,
+      "learning_rate": 8.428588913606306e-05,
+      "loss": 2.7619,
+      "step": 8949
+    },
+    {
+      "epoch": 0.5505659448818898,
+      "grad_norm": 1.2170742750167847,
+      "learning_rate": 8.426679203980754e-05,
+      "loss": 1.8773,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5506274606299213,
+      "grad_norm": 1.1451493501663208,
+      "learning_rate": 8.424769553188623e-05,
+      "loss": 2.5094,
+      "step": 8951
+    },
+    {
+      "epoch": 0.5506889763779528,
+      "grad_norm": 1.1835969686508179,
+      "learning_rate": 8.422859961301329e-05,
+      "loss": 2.9864,
+      "step": 8952
+    },
+    {
+      "epoch": 0.5507504921259843,
+      "grad_norm": 1.1953946352005005,
+      "learning_rate": 8.420950428390274e-05,
+      "loss": 2.5024,
+      "step": 8953
+    },
+    {
+      "epoch": 0.5508120078740157,
+      "grad_norm": 1.2284317016601562,
+      "learning_rate": 8.419040954526873e-05,
+      "loss": 2.3366,
+      "step": 8954
+    },
+    {
+      "epoch": 0.5508735236220472,
+      "grad_norm": 1.3276218175888062,
+      "learning_rate": 8.41713153978252e-05,
+      "loss": 1.9188,
+      "step": 8955
+    },
+    {
+      "epoch": 0.5509350393700787,
+      "grad_norm": 2.132702112197876,
+      "learning_rate": 8.415222184228626e-05,
+      "loss": 2.4524,
+      "step": 8956
+    },
+    {
+      "epoch": 0.5509965551181102,
+      "grad_norm": 0.9937765002250671,
+      "learning_rate": 8.413312887936581e-05,
+      "loss": 2.953,
+      "step": 8957
+    },
+    {
+      "epoch": 0.5510580708661418,
+      "grad_norm": 1.3624837398529053,
+      "learning_rate": 8.41140365097779e-05,
+      "loss": 2.8069,
+      "step": 8958
+    },
+    {
+      "epoch": 0.5511195866141733,
+      "grad_norm": 1.4650756120681763,
+      "learning_rate": 8.409494473423641e-05,
+      "loss": 2.5594,
+      "step": 8959
+    },
+    {
+      "epoch": 0.5511811023622047,
+      "grad_norm": 1.2840869426727295,
+      "learning_rate": 8.407585355345534e-05,
+      "loss": 3.0372,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5512426181102362,
+      "grad_norm": 1.0557310581207275,
+      "learning_rate": 8.405676296814851e-05,
+      "loss": 2.8318,
+      "step": 8961
+    },
+    {
+      "epoch": 0.5513041338582677,
+      "grad_norm": 1.384508728981018,
+      "learning_rate": 8.403767297902987e-05,
+      "loss": 1.851,
+      "step": 8962
+    },
+    {
+      "epoch": 0.5513656496062992,
+      "grad_norm": 1.2186439037322998,
+      "learning_rate": 8.401858358681323e-05,
+      "loss": 3.19,
+      "step": 8963
+    },
+    {
+      "epoch": 0.5514271653543307,
+      "grad_norm": 1.1060711145401,
+      "learning_rate": 8.399949479221246e-05,
+      "loss": 2.0898,
+      "step": 8964
+    },
+    {
+      "epoch": 0.5514886811023622,
+      "grad_norm": 1.3388217687606812,
+      "learning_rate": 8.398040659594136e-05,
+      "loss": 2.3102,
+      "step": 8965
+    },
+    {
+      "epoch": 0.5515501968503937,
+      "grad_norm": 1.3828145265579224,
+      "learning_rate": 8.396131899871375e-05,
+      "loss": 2.8382,
+      "step": 8966
+    },
+    {
+      "epoch": 0.5516117125984252,
+      "grad_norm": 1.7992191314697266,
+      "learning_rate": 8.394223200124333e-05,
+      "loss": 2.5201,
+      "step": 8967
+    },
+    {
+      "epoch": 0.5516732283464567,
+      "grad_norm": 1.4443241357803345,
+      "learning_rate": 8.392314560424393e-05,
+      "loss": 2.4838,
+      "step": 8968
+    },
+    {
+      "epoch": 0.5517347440944882,
+      "grad_norm": 1.369299292564392,
+      "learning_rate": 8.390405980842921e-05,
+      "loss": 2.3035,
+      "step": 8969
+    },
+    {
+      "epoch": 0.5517962598425197,
+      "grad_norm": 1.2800227403640747,
+      "learning_rate": 8.388497461451294e-05,
+      "loss": 2.3217,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5518577755905512,
+      "grad_norm": 1.1831718683242798,
+      "learning_rate": 8.386589002320872e-05,
+      "loss": 3.4926,
+      "step": 8971
+    },
+    {
+      "epoch": 0.5519192913385826,
+      "grad_norm": 1.223617672920227,
+      "learning_rate": 8.384680603523026e-05,
+      "loss": 2.4527,
+      "step": 8972
+    },
+    {
+      "epoch": 0.5519808070866141,
+      "grad_norm": 1.3839843273162842,
+      "learning_rate": 8.382772265129118e-05,
+      "loss": 2.5806,
+      "step": 8973
+    },
+    {
+      "epoch": 0.5520423228346457,
+      "grad_norm": 1.4345366954803467,
+      "learning_rate": 8.38086398721051e-05,
+      "loss": 2.925,
+      "step": 8974
+    },
+    {
+      "epoch": 0.5521038385826772,
+      "grad_norm": 1.3443400859832764,
+      "learning_rate": 8.378955769838558e-05,
+      "loss": 2.5634,
+      "step": 8975
+    },
+    {
+      "epoch": 0.5521653543307087,
+      "grad_norm": 1.1740927696228027,
+      "learning_rate": 8.377047613084625e-05,
+      "loss": 2.622,
+      "step": 8976
+    },
+    {
+      "epoch": 0.5522268700787402,
+      "grad_norm": 1.559098720550537,
+      "learning_rate": 8.375139517020059e-05,
+      "loss": 2.3498,
+      "step": 8977
+    },
+    {
+      "epoch": 0.5522883858267716,
+      "grad_norm": 1.0443925857543945,
+      "learning_rate": 8.373231481716214e-05,
+      "loss": 2.8296,
+      "step": 8978
+    },
+    {
+      "epoch": 0.5523499015748031,
+      "grad_norm": 1.0732027292251587,
+      "learning_rate": 8.371323507244445e-05,
+      "loss": 2.6663,
+      "step": 8979
+    },
+    {
+      "epoch": 0.5524114173228346,
+      "grad_norm": 1.096166968345642,
+      "learning_rate": 8.36941559367609e-05,
+      "loss": 2.6877,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5524729330708661,
+      "grad_norm": 1.1726104021072388,
+      "learning_rate": 8.367507741082502e-05,
+      "loss": 2.6073,
+      "step": 8981
+    },
+    {
+      "epoch": 0.5525344488188977,
+      "grad_norm": 1.1891435384750366,
+      "learning_rate": 8.365599949535018e-05,
+      "loss": 2.6857,
+      "step": 8982
+    },
+    {
+      "epoch": 0.5525959645669292,
+      "grad_norm": 1.2831896543502808,
+      "learning_rate": 8.363692219104987e-05,
+      "loss": 2.6178,
+      "step": 8983
+    },
+    {
+      "epoch": 0.5526574803149606,
+      "grad_norm": 1.7385873794555664,
+      "learning_rate": 8.361784549863738e-05,
+      "loss": 2.0413,
+      "step": 8984
+    },
+    {
+      "epoch": 0.5527189960629921,
+      "grad_norm": 1.081113338470459,
+      "learning_rate": 8.359876941882616e-05,
+      "loss": 2.7985,
+      "step": 8985
+    },
+    {
+      "epoch": 0.5527805118110236,
+      "grad_norm": 1.2365363836288452,
+      "learning_rate": 8.357969395232947e-05,
+      "loss": 2.622,
+      "step": 8986
+    },
+    {
+      "epoch": 0.5528420275590551,
+      "grad_norm": 1.2759932279586792,
+      "learning_rate": 8.35606190998607e-05,
+      "loss": 2.5533,
+      "step": 8987
+    },
+    {
+      "epoch": 0.5529035433070866,
+      "grad_norm": 1.888225793838501,
+      "learning_rate": 8.354154486213308e-05,
+      "loss": 2.3405,
+      "step": 8988
+    },
+    {
+      "epoch": 0.5529650590551181,
+      "grad_norm": 1.243507742881775,
+      "learning_rate": 8.352247123985995e-05,
+      "loss": 2.3617,
+      "step": 8989
+    },
+    {
+      "epoch": 0.5530265748031497,
+      "grad_norm": 1.1883171796798706,
+      "learning_rate": 8.350339823375443e-05,
+      "loss": 2.1347,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5530880905511811,
+      "grad_norm": 1.1350346803665161,
+      "learning_rate": 8.348432584452992e-05,
+      "loss": 2.9448,
+      "step": 8991
+    },
+    {
+      "epoch": 0.5531496062992126,
+      "grad_norm": 1.0231155157089233,
+      "learning_rate": 8.346525407289947e-05,
+      "loss": 2.996,
+      "step": 8992
+    },
+    {
+      "epoch": 0.5532111220472441,
+      "grad_norm": 1.5021470785140991,
+      "learning_rate": 8.344618291957635e-05,
+      "loss": 2.3724,
+      "step": 8993
+    },
+    {
+      "epoch": 0.5532726377952756,
+      "grad_norm": 1.4093605279922485,
+      "learning_rate": 8.342711238527365e-05,
+      "loss": 2.5649,
+      "step": 8994
+    },
+    {
+      "epoch": 0.5533341535433071,
+      "grad_norm": 1.0429962873458862,
+      "learning_rate": 8.340804247070458e-05,
+      "loss": 2.5608,
+      "step": 8995
+    },
+    {
+      "epoch": 0.5533956692913385,
+      "grad_norm": 1.2350029945373535,
+      "learning_rate": 8.338897317658216e-05,
+      "loss": 2.5643,
+      "step": 8996
+    },
+    {
+      "epoch": 0.55345718503937,
+      "grad_norm": 1.3062829971313477,
+      "learning_rate": 8.336990450361954e-05,
+      "loss": 2.8084,
+      "step": 8997
+    },
+    {
+      "epoch": 0.5535187007874016,
+      "grad_norm": 1.55648672580719,
+      "learning_rate": 8.335083645252975e-05,
+      "loss": 2.2583,
+      "step": 8998
+    },
+    {
+      "epoch": 0.5535802165354331,
+      "grad_norm": 1.2165809869766235,
+      "learning_rate": 8.333176902402583e-05,
+      "loss": 2.153,
+      "step": 8999
+    },
+    {
+      "epoch": 0.5536417322834646,
+      "grad_norm": 1.2406800985336304,
+      "learning_rate": 8.331270221882082e-05,
+      "loss": 2.8323,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5537032480314961,
+      "grad_norm": 1.284838318824768,
+      "learning_rate": 8.329363603762772e-05,
+      "loss": 2.4804,
+      "step": 9001
+    },
+    {
+      "epoch": 0.5537647637795275,
+      "grad_norm": 1.147632122039795,
+      "learning_rate": 8.327457048115943e-05,
+      "loss": 2.5501,
+      "step": 9002
+    },
+    {
+      "epoch": 0.553826279527559,
+      "grad_norm": 1.2156237363815308,
+      "learning_rate": 8.325550555012897e-05,
+      "loss": 3.0274,
+      "step": 9003
+    },
+    {
+      "epoch": 0.5538877952755905,
+      "grad_norm": 1.329642415046692,
+      "learning_rate": 8.32364412452492e-05,
+      "loss": 2.6129,
+      "step": 9004
+    },
+    {
+      "epoch": 0.553949311023622,
+      "grad_norm": 1.2833995819091797,
+      "learning_rate": 8.32173775672331e-05,
+      "loss": 2.3412,
+      "step": 9005
+    },
+    {
+      "epoch": 0.5540108267716536,
+      "grad_norm": 1.3148090839385986,
+      "learning_rate": 8.319831451679345e-05,
+      "loss": 2.8072,
+      "step": 9006
+    },
+    {
+      "epoch": 0.5540723425196851,
+      "grad_norm": 1.2950634956359863,
+      "learning_rate": 8.317925209464317e-05,
+      "loss": 2.7521,
+      "step": 9007
+    },
+    {
+      "epoch": 0.5541338582677166,
+      "grad_norm": 1.058594822883606,
+      "learning_rate": 8.316019030149504e-05,
+      "loss": 1.917,
+      "step": 9008
+    },
+    {
+      "epoch": 0.554195374015748,
+      "grad_norm": 1.27119779586792,
+      "learning_rate": 8.314112913806193e-05,
+      "loss": 2.7467,
+      "step": 9009
+    },
+    {
+      "epoch": 0.5542568897637795,
+      "grad_norm": 1.2900911569595337,
+      "learning_rate": 8.312206860505655e-05,
+      "loss": 2.6097,
+      "step": 9010
+    },
+    {
+      "epoch": 0.554318405511811,
+      "grad_norm": 1.4236425161361694,
+      "learning_rate": 8.310300870319174e-05,
+      "loss": 2.3692,
+      "step": 9011
+    },
+    {
+      "epoch": 0.5543799212598425,
+      "grad_norm": 1.4934239387512207,
+      "learning_rate": 8.308394943318014e-05,
+      "loss": 1.949,
+      "step": 9012
+    },
+    {
+      "epoch": 0.554441437007874,
+      "grad_norm": 1.018650770187378,
+      "learning_rate": 8.306489079573452e-05,
+      "loss": 2.3747,
+      "step": 9013
+    },
+    {
+      "epoch": 0.5545029527559056,
+      "grad_norm": 1.169042706489563,
+      "learning_rate": 8.304583279156752e-05,
+      "loss": 1.9679,
+      "step": 9014
+    },
+    {
+      "epoch": 0.554564468503937,
+      "grad_norm": 1.340030312538147,
+      "learning_rate": 8.30267754213919e-05,
+      "loss": 2.0367,
+      "step": 9015
+    },
+    {
+      "epoch": 0.5546259842519685,
+      "grad_norm": 1.105520248413086,
+      "learning_rate": 8.30077186859202e-05,
+      "loss": 2.5576,
+      "step": 9016
+    },
+    {
+      "epoch": 0.5546875,
+      "grad_norm": 1.2825003862380981,
+      "learning_rate": 8.298866258586505e-05,
+      "loss": 2.7932,
+      "step": 9017
+    },
+    {
+      "epoch": 0.5547490157480315,
+      "grad_norm": 1.1106377840042114,
+      "learning_rate": 8.296960712193909e-05,
+      "loss": 2.3995,
+      "step": 9018
+    },
+    {
+      "epoch": 0.554810531496063,
+      "grad_norm": 1.2397065162658691,
+      "learning_rate": 8.295055229485483e-05,
+      "loss": 2.2574,
+      "step": 9019
+    },
+    {
+      "epoch": 0.5548720472440944,
+      "grad_norm": 1.2081729173660278,
+      "learning_rate": 8.293149810532487e-05,
+      "loss": 2.5459,
+      "step": 9020
+    },
+    {
+      "epoch": 0.554933562992126,
+      "grad_norm": 1.229720115661621,
+      "learning_rate": 8.29124445540617e-05,
+      "loss": 2.8052,
+      "step": 9021
+    },
+    {
+      "epoch": 0.5549950787401575,
+      "grad_norm": 1.0621192455291748,
+      "learning_rate": 8.289339164177781e-05,
+      "loss": 2.482,
+      "step": 9022
+    },
+    {
+      "epoch": 0.555056594488189,
+      "grad_norm": 1.2389838695526123,
+      "learning_rate": 8.287433936918569e-05,
+      "loss": 1.9128,
+      "step": 9023
+    },
+    {
+      "epoch": 0.5551181102362205,
+      "grad_norm": 1.1010658740997314,
+      "learning_rate": 8.28552877369978e-05,
+      "loss": 2.6431,
+      "step": 9024
+    },
+    {
+      "epoch": 0.555179625984252,
+      "grad_norm": 1.3278940916061401,
+      "learning_rate": 8.283623674592648e-05,
+      "loss": 2.8672,
+      "step": 9025
+    },
+    {
+      "epoch": 0.5552411417322834,
+      "grad_norm": 1.2146433591842651,
+      "learning_rate": 8.281718639668429e-05,
+      "loss": 2.7008,
+      "step": 9026
+    },
+    {
+      "epoch": 0.5553026574803149,
+      "grad_norm": 1.0125178098678589,
+      "learning_rate": 8.279813668998346e-05,
+      "loss": 2.9628,
+      "step": 9027
+    },
+    {
+      "epoch": 0.5553641732283464,
+      "grad_norm": 1.3827258348464966,
+      "learning_rate": 8.277908762653641e-05,
+      "loss": 3.0829,
+      "step": 9028
+    },
+    {
+      "epoch": 0.555425688976378,
+      "grad_norm": 0.990588366985321,
+      "learning_rate": 8.276003920705544e-05,
+      "loss": 3.1326,
+      "step": 9029
+    },
+    {
+      "epoch": 0.5554872047244095,
+      "grad_norm": 1.4268124103546143,
+      "learning_rate": 8.274099143225292e-05,
+      "loss": 3.2963,
+      "step": 9030
+    },
+    {
+      "epoch": 0.555548720472441,
+      "grad_norm": 1.230638027191162,
+      "learning_rate": 8.272194430284102e-05,
+      "loss": 2.7876,
+      "step": 9031
+    },
+    {
+      "epoch": 0.5556102362204725,
+      "grad_norm": 1.3605643510818481,
+      "learning_rate": 8.27028978195321e-05,
+      "loss": 2.543,
+      "step": 9032
+    },
+    {
+      "epoch": 0.5556717519685039,
+      "grad_norm": 1.3307706117630005,
+      "learning_rate": 8.268385198303831e-05,
+      "loss": 2.667,
+      "step": 9033
+    },
+    {
+      "epoch": 0.5557332677165354,
+      "grad_norm": 1.2672111988067627,
+      "learning_rate": 8.266480679407193e-05,
+      "loss": 2.1592,
+      "step": 9034
+    },
+    {
+      "epoch": 0.5557947834645669,
+      "grad_norm": 1.202646255493164,
+      "learning_rate": 8.26457622533451e-05,
+      "loss": 2.5863,
+      "step": 9035
+    },
+    {
+      "epoch": 0.5558562992125984,
+      "grad_norm": 1.1110224723815918,
+      "learning_rate": 8.262671836157002e-05,
+      "loss": 2.5346,
+      "step": 9036
+    },
+    {
+      "epoch": 0.55591781496063,
+      "grad_norm": 1.130860686302185,
+      "learning_rate": 8.260767511945877e-05,
+      "loss": 2.8072,
+      "step": 9037
+    },
+    {
+      "epoch": 0.5559793307086615,
+      "grad_norm": 1.3081088066101074,
+      "learning_rate": 8.258863252772349e-05,
+      "loss": 2.67,
+      "step": 9038
+    },
+    {
+      "epoch": 0.5560408464566929,
+      "grad_norm": 1.693129062652588,
+      "learning_rate": 8.256959058707626e-05,
+      "loss": 2.4196,
+      "step": 9039
+    },
+    {
+      "epoch": 0.5561023622047244,
+      "grad_norm": 1.6757149696350098,
+      "learning_rate": 8.255054929822919e-05,
+      "loss": 1.6243,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5561638779527559,
+      "grad_norm": 2.2931935787200928,
+      "learning_rate": 8.253150866189422e-05,
+      "loss": 2.6501,
+      "step": 9041
+    },
+    {
+      "epoch": 0.5562253937007874,
+      "grad_norm": 1.6642881631851196,
+      "learning_rate": 8.251246867878344e-05,
+      "loss": 2.4807,
+      "step": 9042
+    },
+    {
+      "epoch": 0.5562869094488189,
+      "grad_norm": 1.312231183052063,
+      "learning_rate": 8.249342934960882e-05,
+      "loss": 2.7788,
+      "step": 9043
+    },
+    {
+      "epoch": 0.5563484251968503,
+      "grad_norm": 1.2321979999542236,
+      "learning_rate": 8.247439067508231e-05,
+      "loss": 2.7967,
+      "step": 9044
+    },
+    {
+      "epoch": 0.5564099409448819,
+      "grad_norm": 1.1933305263519287,
+      "learning_rate": 8.245535265591586e-05,
+      "loss": 2.6717,
+      "step": 9045
+    },
+    {
+      "epoch": 0.5564714566929134,
+      "grad_norm": 1.340055227279663,
+      "learning_rate": 8.243631529282144e-05,
+      "loss": 2.1882,
+      "step": 9046
+    },
+    {
+      "epoch": 0.5565329724409449,
+      "grad_norm": 1.436790943145752,
+      "learning_rate": 8.241727858651084e-05,
+      "loss": 2.195,
+      "step": 9047
+    },
+    {
+      "epoch": 0.5565944881889764,
+      "grad_norm": 1.164202332496643,
+      "learning_rate": 8.2398242537696e-05,
+      "loss": 2.5697,
+      "step": 9048
+    },
+    {
+      "epoch": 0.5566560039370079,
+      "grad_norm": 1.2691435813903809,
+      "learning_rate": 8.237920714708871e-05,
+      "loss": 2.8349,
+      "step": 9049
+    },
+    {
+      "epoch": 0.5567175196850394,
+      "grad_norm": 1.1928706169128418,
+      "learning_rate": 8.236017241540088e-05,
+      "loss": 1.8971,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5567790354330708,
+      "grad_norm": 1.2121247053146362,
+      "learning_rate": 8.234113834334415e-05,
+      "loss": 2.728,
+      "step": 9051
+    },
+    {
+      "epoch": 0.5568405511811023,
+      "grad_norm": 1.1156668663024902,
+      "learning_rate": 8.232210493163044e-05,
+      "loss": 2.7961,
+      "step": 9052
+    },
+    {
+      "epoch": 0.5569020669291339,
+      "grad_norm": 1.2844558954238892,
+      "learning_rate": 8.230307218097143e-05,
+      "loss": 2.5321,
+      "step": 9053
+    },
+    {
+      "epoch": 0.5569635826771654,
+      "grad_norm": 1.1608729362487793,
+      "learning_rate": 8.228404009207881e-05,
+      "loss": 2.6949,
+      "step": 9054
+    },
+    {
+      "epoch": 0.5570250984251969,
+      "grad_norm": 1.1746633052825928,
+      "learning_rate": 8.226500866566435e-05,
+      "loss": 2.719,
+      "step": 9055
+    },
+    {
+      "epoch": 0.5570866141732284,
+      "grad_norm": 1.2119638919830322,
+      "learning_rate": 8.224597790243962e-05,
+      "loss": 2.5592,
+      "step": 9056
+    },
+    {
+      "epoch": 0.5571481299212598,
+      "grad_norm": 1.2113127708435059,
+      "learning_rate": 8.222694780311635e-05,
+      "loss": 2.1661,
+      "step": 9057
+    },
+    {
+      "epoch": 0.5572096456692913,
+      "grad_norm": 1.1043976545333862,
+      "learning_rate": 8.22079183684061e-05,
+      "loss": 2.3725,
+      "step": 9058
+    },
+    {
+      "epoch": 0.5572711614173228,
+      "grad_norm": 1.1766011714935303,
+      "learning_rate": 8.218888959902054e-05,
+      "loss": 2.6344,
+      "step": 9059
+    },
+    {
+      "epoch": 0.5573326771653543,
+      "grad_norm": 1.305519938468933,
+      "learning_rate": 8.216986149567113e-05,
+      "loss": 3.0672,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5573941929133859,
+      "grad_norm": 1.9721112251281738,
+      "learning_rate": 8.215083405906951e-05,
+      "loss": 2.7401,
+      "step": 9061
+    },
+    {
+      "epoch": 0.5574557086614174,
+      "grad_norm": 2.308136224746704,
+      "learning_rate": 8.213180728992716e-05,
+      "loss": 2.2009,
+      "step": 9062
+    },
+    {
+      "epoch": 0.5575172244094488,
+      "grad_norm": 1.2038949728012085,
+      "learning_rate": 8.211278118895558e-05,
+      "loss": 2.5643,
+      "step": 9063
+    },
+    {
+      "epoch": 0.5575787401574803,
+      "grad_norm": 1.3509740829467773,
+      "learning_rate": 8.209375575686623e-05,
+      "loss": 2.1229,
+      "step": 9064
+    },
+    {
+      "epoch": 0.5576402559055118,
+      "grad_norm": 1.3644425868988037,
+      "learning_rate": 8.207473099437061e-05,
+      "loss": 2.1926,
+      "step": 9065
+    },
+    {
+      "epoch": 0.5577017716535433,
+      "grad_norm": 1.148869276046753,
+      "learning_rate": 8.205570690218005e-05,
+      "loss": 2.6883,
+      "step": 9066
+    },
+    {
+      "epoch": 0.5577632874015748,
+      "grad_norm": 1.1595567464828491,
+      "learning_rate": 8.203668348100602e-05,
+      "loss": 2.6131,
+      "step": 9067
+    },
+    {
+      "epoch": 0.5578248031496063,
+      "grad_norm": 1.2226734161376953,
+      "learning_rate": 8.201766073155984e-05,
+      "loss": 2.4172,
+      "step": 9068
+    },
+    {
+      "epoch": 0.5578863188976378,
+      "grad_norm": 1.6399449110031128,
+      "learning_rate": 8.19986386545529e-05,
+      "loss": 2.4883,
+      "step": 9069
+    },
+    {
+      "epoch": 0.5579478346456693,
+      "grad_norm": 1.61493718624115,
+      "learning_rate": 8.197961725069646e-05,
+      "loss": 2.3362,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5580093503937008,
+      "grad_norm": 1.3458176851272583,
+      "learning_rate": 8.196059652070189e-05,
+      "loss": 1.8701,
+      "step": 9071
+    },
+    {
+      "epoch": 0.5580708661417323,
+      "grad_norm": 1.1627448797225952,
+      "learning_rate": 8.19415764652804e-05,
+      "loss": 2.7844,
+      "step": 9072
+    },
+    {
+      "epoch": 0.5581323818897638,
+      "grad_norm": 1.3090535402297974,
+      "learning_rate": 8.192255708514325e-05,
+      "loss": 2.1967,
+      "step": 9073
+    },
+    {
+      "epoch": 0.5581938976377953,
+      "grad_norm": 1.257612943649292,
+      "learning_rate": 8.190353838100166e-05,
+      "loss": 2.7652,
+      "step": 9074
+    },
+    {
+      "epoch": 0.5582554133858267,
+      "grad_norm": 1.2392849922180176,
+      "learning_rate": 8.188452035356688e-05,
+      "loss": 2.0805,
+      "step": 9075
+    },
+    {
+      "epoch": 0.5583169291338582,
+      "grad_norm": 1.4467984437942505,
+      "learning_rate": 8.186550300354996e-05,
+      "loss": 2.4196,
+      "step": 9076
+    },
+    {
+      "epoch": 0.5583784448818898,
+      "grad_norm": 1.0753750801086426,
+      "learning_rate": 8.184648633166213e-05,
+      "loss": 2.8142,
+      "step": 9077
+    },
+    {
+      "epoch": 0.5584399606299213,
+      "grad_norm": 1.2987184524536133,
+      "learning_rate": 8.182747033861447e-05,
+      "loss": 2.4727,
+      "step": 9078
+    },
+    {
+      "epoch": 0.5585014763779528,
+      "grad_norm": 1.431510329246521,
+      "learning_rate": 8.180845502511809e-05,
+      "loss": 2.853,
+      "step": 9079
+    },
+    {
+      "epoch": 0.5585629921259843,
+      "grad_norm": 1.1076995134353638,
+      "learning_rate": 8.178944039188405e-05,
+      "loss": 3.1117,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5586245078740157,
+      "grad_norm": 1.2093721628189087,
+      "learning_rate": 8.177042643962344e-05,
+      "loss": 3.0032,
+      "step": 9081
+    },
+    {
+      "epoch": 0.5586860236220472,
+      "grad_norm": 1.1717896461486816,
+      "learning_rate": 8.175141316904718e-05,
+      "loss": 2.1217,
+      "step": 9082
+    },
+    {
+      "epoch": 0.5587475393700787,
+      "grad_norm": 1.205816388130188,
+      "learning_rate": 8.173240058086633e-05,
+      "loss": 1.5781,
+      "step": 9083
+    },
+    {
+      "epoch": 0.5588090551181102,
+      "grad_norm": 1.4044934511184692,
+      "learning_rate": 8.171338867579181e-05,
+      "loss": 2.5881,
+      "step": 9084
+    },
+    {
+      "epoch": 0.5588705708661418,
+      "grad_norm": 1.202323317527771,
+      "learning_rate": 8.169437745453465e-05,
+      "loss": 2.635,
+      "step": 9085
+    },
+    {
+      "epoch": 0.5589320866141733,
+      "grad_norm": 1.4072846174240112,
+      "learning_rate": 8.167536691780561e-05,
+      "loss": 2.7628,
+      "step": 9086
+    },
+    {
+      "epoch": 0.5589936023622047,
+      "grad_norm": 1.4449740648269653,
+      "learning_rate": 8.165635706631574e-05,
+      "loss": 2.5065,
+      "step": 9087
+    },
+    {
+      "epoch": 0.5590551181102362,
+      "grad_norm": 1.1893812417984009,
+      "learning_rate": 8.163734790077578e-05,
+      "loss": 2.9229,
+      "step": 9088
+    },
+    {
+      "epoch": 0.5591166338582677,
+      "grad_norm": 1.1727690696716309,
+      "learning_rate": 8.161833942189665e-05,
+      "loss": 2.4703,
+      "step": 9089
+    },
+    {
+      "epoch": 0.5591781496062992,
+      "grad_norm": 1.0272319316864014,
+      "learning_rate": 8.159933163038913e-05,
+      "loss": 2.3147,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5592396653543307,
+      "grad_norm": 1.0309271812438965,
+      "learning_rate": 8.158032452696396e-05,
+      "loss": 2.4928,
+      "step": 9091
+    },
+    {
+      "epoch": 0.5593011811023622,
+      "grad_norm": 1.234958291053772,
+      "learning_rate": 8.156131811233198e-05,
+      "loss": 2.2842,
+      "step": 9092
+    },
+    {
+      "epoch": 0.5593626968503937,
+      "grad_norm": 1.1911085844039917,
+      "learning_rate": 8.154231238720386e-05,
+      "loss": 2.6443,
+      "step": 9093
+    },
+    {
+      "epoch": 0.5594242125984252,
+      "grad_norm": 1.4372330904006958,
+      "learning_rate": 8.152330735229037e-05,
+      "loss": 1.7767,
+      "step": 9094
+    },
+    {
+      "epoch": 0.5594857283464567,
+      "grad_norm": 1.1528939008712769,
+      "learning_rate": 8.150430300830212e-05,
+      "loss": 2.24,
+      "step": 9095
+    },
+    {
+      "epoch": 0.5595472440944882,
+      "grad_norm": 1.1776001453399658,
+      "learning_rate": 8.148529935594982e-05,
+      "loss": 2.9163,
+      "step": 9096
+    },
+    {
+      "epoch": 0.5596087598425197,
+      "grad_norm": 1.183300256729126,
+      "learning_rate": 8.146629639594407e-05,
+      "loss": 2.2146,
+      "step": 9097
+    },
+    {
+      "epoch": 0.5596702755905512,
+      "grad_norm": 1.3937350511550903,
+      "learning_rate": 8.14472941289955e-05,
+      "loss": 2.6076,
+      "step": 9098
+    },
+    {
+      "epoch": 0.5597317913385826,
+      "grad_norm": 1.1758222579956055,
+      "learning_rate": 8.142829255581465e-05,
+      "loss": 2.5644,
+      "step": 9099
+    },
+    {
+      "epoch": 0.5597933070866141,
+      "grad_norm": 1.3526443243026733,
+      "learning_rate": 8.140929167711216e-05,
+      "loss": 2.7231,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5598548228346457,
+      "grad_norm": 1.6000947952270508,
+      "learning_rate": 8.139029149359844e-05,
+      "loss": 1.7116,
+      "step": 9101
+    },
+    {
+      "epoch": 0.5599163385826772,
+      "grad_norm": 1.322876214981079,
+      "learning_rate": 8.137129200598407e-05,
+      "loss": 1.83,
+      "step": 9102
+    },
+    {
+      "epoch": 0.5599778543307087,
+      "grad_norm": 1.5365108251571655,
+      "learning_rate": 8.135229321497948e-05,
+      "loss": 2.2595,
+      "step": 9103
+    },
+    {
+      "epoch": 0.5600393700787402,
+      "grad_norm": 1.2642616033554077,
+      "learning_rate": 8.133329512129519e-05,
+      "loss": 2.8138,
+      "step": 9104
+    },
+    {
+      "epoch": 0.5601008858267716,
+      "grad_norm": 1.1399726867675781,
+      "learning_rate": 8.13142977256415e-05,
+      "loss": 2.8876,
+      "step": 9105
+    },
+    {
+      "epoch": 0.5601624015748031,
+      "grad_norm": 1.211965560913086,
+      "learning_rate": 8.129530102872895e-05,
+      "loss": 2.4365,
+      "step": 9106
+    },
+    {
+      "epoch": 0.5602239173228346,
+      "grad_norm": 1.2356669902801514,
+      "learning_rate": 8.127630503126781e-05,
+      "loss": 1.8686,
+      "step": 9107
+    },
+    {
+      "epoch": 0.5602854330708661,
+      "grad_norm": 1.1397418975830078,
+      "learning_rate": 8.125730973396847e-05,
+      "loss": 3.5115,
+      "step": 9108
+    },
+    {
+      "epoch": 0.5603469488188977,
+      "grad_norm": 1.2102426290512085,
+      "learning_rate": 8.123831513754121e-05,
+      "loss": 2.4475,
+      "step": 9109
+    },
+    {
+      "epoch": 0.5604084645669292,
+      "grad_norm": 1.3382971286773682,
+      "learning_rate": 8.12193212426964e-05,
+      "loss": 3.1611,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5604699803149606,
+      "grad_norm": 1.109318494796753,
+      "learning_rate": 8.120032805014419e-05,
+      "loss": 2.5729,
+      "step": 9111
+    },
+    {
+      "epoch": 0.5605314960629921,
+      "grad_norm": 1.2859606742858887,
+      "learning_rate": 8.11813355605949e-05,
+      "loss": 1.6988,
+      "step": 9112
+    },
+    {
+      "epoch": 0.5605930118110236,
+      "grad_norm": 1.3025386333465576,
+      "learning_rate": 8.116234377475871e-05,
+      "loss": 2.5253,
+      "step": 9113
+    },
+    {
+      "epoch": 0.5606545275590551,
+      "grad_norm": 1.2570918798446655,
+      "learning_rate": 8.114335269334585e-05,
+      "loss": 2.8536,
+      "step": 9114
+    },
+    {
+      "epoch": 0.5607160433070866,
+      "grad_norm": 1.0956817865371704,
+      "learning_rate": 8.112436231706642e-05,
+      "loss": 2.7102,
+      "step": 9115
+    },
+    {
+      "epoch": 0.5607775590551181,
+      "grad_norm": 1.266663908958435,
+      "learning_rate": 8.110537264663064e-05,
+      "loss": 2.6597,
+      "step": 9116
+    },
+    {
+      "epoch": 0.5608390748031497,
+      "grad_norm": 1.1453992128372192,
+      "learning_rate": 8.108638368274853e-05,
+      "loss": 2.9205,
+      "step": 9117
+    },
+    {
+      "epoch": 0.5609005905511811,
+      "grad_norm": 1.0274790525436401,
+      "learning_rate": 8.10673954261302e-05,
+      "loss": 2.5917,
+      "step": 9118
+    },
+    {
+      "epoch": 0.5609621062992126,
+      "grad_norm": 1.194158911705017,
+      "learning_rate": 8.104840787748569e-05,
+      "loss": 2.3913,
+      "step": 9119
+    },
+    {
+      "epoch": 0.5610236220472441,
+      "grad_norm": 1.213384985923767,
+      "learning_rate": 8.102942103752512e-05,
+      "loss": 2.4077,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5610851377952756,
+      "grad_norm": 1.1327756643295288,
+      "learning_rate": 8.101043490695836e-05,
+      "loss": 2.6814,
+      "step": 9121
+    },
+    {
+      "epoch": 0.5611466535433071,
+      "grad_norm": 1.0791884660720825,
+      "learning_rate": 8.099144948649547e-05,
+      "loss": 2.5092,
+      "step": 9122
+    },
+    {
+      "epoch": 0.5612081692913385,
+      "grad_norm": 1.5956511497497559,
+      "learning_rate": 8.097246477684633e-05,
+      "loss": 3.4896,
+      "step": 9123
+    },
+    {
+      "epoch": 0.56126968503937,
+      "grad_norm": 1.3603854179382324,
+      "learning_rate": 8.095348077872096e-05,
+      "loss": 2.2534,
+      "step": 9124
+    },
+    {
+      "epoch": 0.5613312007874016,
+      "grad_norm": 1.3530104160308838,
+      "learning_rate": 8.093449749282917e-05,
+      "loss": 1.9937,
+      "step": 9125
+    },
+    {
+      "epoch": 0.5613927165354331,
+      "grad_norm": 1.119555115699768,
+      "learning_rate": 8.09155149198809e-05,
+      "loss": 2.9469,
+      "step": 9126
+    },
+    {
+      "epoch": 0.5614542322834646,
+      "grad_norm": 1.32982337474823,
+      "learning_rate": 8.089653306058593e-05,
+      "loss": 2.7457,
+      "step": 9127
+    },
+    {
+      "epoch": 0.5615157480314961,
+      "grad_norm": 1.089260220527649,
+      "learning_rate": 8.08775519156541e-05,
+      "loss": 2.2942,
+      "step": 9128
+    },
+    {
+      "epoch": 0.5615772637795275,
+      "grad_norm": 1.3590487241744995,
+      "learning_rate": 8.085857148579521e-05,
+      "loss": 2.5083,
+      "step": 9129
+    },
+    {
+      "epoch": 0.561638779527559,
+      "grad_norm": 1.1758288145065308,
+      "learning_rate": 8.083959177171899e-05,
+      "loss": 2.5532,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5617002952755905,
+      "grad_norm": 1.5479586124420166,
+      "learning_rate": 8.08206127741352e-05,
+      "loss": 1.2914,
+      "step": 9131
+    },
+    {
+      "epoch": 0.561761811023622,
+      "grad_norm": 1.1076641082763672,
+      "learning_rate": 8.080163449375355e-05,
+      "loss": 2.5944,
+      "step": 9132
+    },
+    {
+      "epoch": 0.5618233267716536,
+      "grad_norm": 1.2012462615966797,
+      "learning_rate": 8.078265693128372e-05,
+      "loss": 2.2096,
+      "step": 9133
+    },
+    {
+      "epoch": 0.5618848425196851,
+      "grad_norm": 1.0987882614135742,
+      "learning_rate": 8.076368008743534e-05,
+      "loss": 2.9101,
+      "step": 9134
+    },
+    {
+      "epoch": 0.5619463582677166,
+      "grad_norm": 1.1895191669464111,
+      "learning_rate": 8.074470396291811e-05,
+      "loss": 2.0152,
+      "step": 9135
+    },
+    {
+      "epoch": 0.562007874015748,
+      "grad_norm": 1.7900383472442627,
+      "learning_rate": 8.072572855844153e-05,
+      "loss": 1.9469,
+      "step": 9136
+    },
+    {
+      "epoch": 0.5620693897637795,
+      "grad_norm": 1.6596347093582153,
+      "learning_rate": 8.070675387471525e-05,
+      "loss": 1.3668,
+      "step": 9137
+    },
+    {
+      "epoch": 0.562130905511811,
+      "grad_norm": 1.124116063117981,
+      "learning_rate": 8.068777991244879e-05,
+      "loss": 2.7573,
+      "step": 9138
+    },
+    {
+      "epoch": 0.5621924212598425,
+      "grad_norm": 1.251479148864746,
+      "learning_rate": 8.06688066723517e-05,
+      "loss": 2.6579,
+      "step": 9139
+    },
+    {
+      "epoch": 0.562253937007874,
+      "grad_norm": 1.3792665004730225,
+      "learning_rate": 8.064983415513337e-05,
+      "loss": 3.507,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5623154527559056,
+      "grad_norm": 1.1956738233566284,
+      "learning_rate": 8.063086236150343e-05,
+      "loss": 3.0758,
+      "step": 9141
+    },
+    {
+      "epoch": 0.562376968503937,
+      "grad_norm": 1.4517542123794556,
+      "learning_rate": 8.061189129217118e-05,
+      "loss": 2.1505,
+      "step": 9142
+    },
+    {
+      "epoch": 0.5624384842519685,
+      "grad_norm": 1.2727984189987183,
+      "learning_rate": 8.05929209478461e-05,
+      "loss": 2.6234,
+      "step": 9143
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 1.3219112157821655,
+      "learning_rate": 8.057395132923756e-05,
+      "loss": 2.6748,
+      "step": 9144
+    },
+    {
+      "epoch": 0.5625615157480315,
+      "grad_norm": 1.2739969491958618,
+      "learning_rate": 8.055498243705493e-05,
+      "loss": 2.2274,
+      "step": 9145
+    },
+    {
+      "epoch": 0.562623031496063,
+      "grad_norm": 1.2412452697753906,
+      "learning_rate": 8.05360142720075e-05,
+      "loss": 2.3993,
+      "step": 9146
+    },
+    {
+      "epoch": 0.5626845472440944,
+      "grad_norm": 1.2865512371063232,
+      "learning_rate": 8.051704683480463e-05,
+      "loss": 2.3962,
+      "step": 9147
+    },
+    {
+      "epoch": 0.562746062992126,
+      "grad_norm": 1.373644471168518,
+      "learning_rate": 8.049808012615553e-05,
+      "loss": 3.1876,
+      "step": 9148
+    },
+    {
+      "epoch": 0.5628075787401575,
+      "grad_norm": 1.10518479347229,
+      "learning_rate": 8.047911414676952e-05,
+      "loss": 2.9436,
+      "step": 9149
+    },
+    {
+      "epoch": 0.562869094488189,
+      "grad_norm": 1.403003454208374,
+      "learning_rate": 8.046014889735577e-05,
+      "loss": 2.5627,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5629306102362205,
+      "grad_norm": 1.4040237665176392,
+      "learning_rate": 8.044118437862353e-05,
+      "loss": 2.7652,
+      "step": 9151
+    },
+    {
+      "epoch": 0.562992125984252,
+      "grad_norm": 1.2873454093933105,
+      "learning_rate": 8.04222205912819e-05,
+      "loss": 2.1839,
+      "step": 9152
+    },
+    {
+      "epoch": 0.5630536417322834,
+      "grad_norm": 1.2363938093185425,
+      "learning_rate": 8.040325753604007e-05,
+      "loss": 2.4367,
+      "step": 9153
+    },
+    {
+      "epoch": 0.5631151574803149,
+      "grad_norm": 1.2406039237976074,
+      "learning_rate": 8.038429521360711e-05,
+      "loss": 2.6051,
+      "step": 9154
+    },
+    {
+      "epoch": 0.5631766732283464,
+      "grad_norm": 1.618685007095337,
+      "learning_rate": 8.036533362469216e-05,
+      "loss": 2.1161,
+      "step": 9155
+    },
+    {
+      "epoch": 0.563238188976378,
+      "grad_norm": 1.7355397939682007,
+      "learning_rate": 8.034637277000422e-05,
+      "loss": 2.0045,
+      "step": 9156
+    },
+    {
+      "epoch": 0.5632997047244095,
+      "grad_norm": 1.3122743368148804,
+      "learning_rate": 8.032741265025237e-05,
+      "loss": 1.8518,
+      "step": 9157
+    },
+    {
+      "epoch": 0.563361220472441,
+      "grad_norm": 1.3605140447616577,
+      "learning_rate": 8.030845326614555e-05,
+      "loss": 2.7069,
+      "step": 9158
+    },
+    {
+      "epoch": 0.5634227362204725,
+      "grad_norm": 1.7535420656204224,
+      "learning_rate": 8.028949461839281e-05,
+      "loss": 2.2622,
+      "step": 9159
+    },
+    {
+      "epoch": 0.5634842519685039,
+      "grad_norm": 1.080159306526184,
+      "learning_rate": 8.027053670770304e-05,
+      "loss": 2.4221,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5635457677165354,
+      "grad_norm": 1.1795457601547241,
+      "learning_rate": 8.025157953478522e-05,
+      "loss": 2.3999,
+      "step": 9161
+    },
+    {
+      "epoch": 0.5636072834645669,
+      "grad_norm": 1.2945882081985474,
+      "learning_rate": 8.023262310034816e-05,
+      "loss": 2.2714,
+      "step": 9162
+    },
+    {
+      "epoch": 0.5636687992125984,
+      "grad_norm": 1.2121529579162598,
+      "learning_rate": 8.02136674051008e-05,
+      "loss": 2.8061,
+      "step": 9163
+    },
+    {
+      "epoch": 0.56373031496063,
+      "grad_norm": 1.2960259914398193,
+      "learning_rate": 8.019471244975196e-05,
+      "loss": 2.4876,
+      "step": 9164
+    },
+    {
+      "epoch": 0.5637918307086615,
+      "grad_norm": 1.1135057210922241,
+      "learning_rate": 8.01757582350104e-05,
+      "loss": 2.8384,
+      "step": 9165
+    },
+    {
+      "epoch": 0.5638533464566929,
+      "grad_norm": 1.1140295267105103,
+      "learning_rate": 8.015680476158496e-05,
+      "loss": 2.4476,
+      "step": 9166
+    },
+    {
+      "epoch": 0.5639148622047244,
+      "grad_norm": 1.1740487813949585,
+      "learning_rate": 8.013785203018436e-05,
+      "loss": 2.8184,
+      "step": 9167
+    },
+    {
+      "epoch": 0.5639763779527559,
+      "grad_norm": 1.3001151084899902,
+      "learning_rate": 8.011890004151736e-05,
+      "loss": 2.5199,
+      "step": 9168
+    },
+    {
+      "epoch": 0.5640378937007874,
+      "grad_norm": 1.3785420656204224,
+      "learning_rate": 8.00999487962926e-05,
+      "loss": 2.6869,
+      "step": 9169
+    },
+    {
+      "epoch": 0.5640994094488189,
+      "grad_norm": 1.1850249767303467,
+      "learning_rate": 8.008099829521885e-05,
+      "loss": 2.3119,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5641609251968503,
+      "grad_norm": 1.4447604417800903,
+      "learning_rate": 8.006204853900464e-05,
+      "loss": 2.2108,
+      "step": 9171
+    },
+    {
+      "epoch": 0.5642224409448819,
+      "grad_norm": 1.1298967599868774,
+      "learning_rate": 8.004309952835865e-05,
+      "loss": 2.6576,
+      "step": 9172
+    },
+    {
+      "epoch": 0.5642839566929134,
+      "grad_norm": 1.2489978075027466,
+      "learning_rate": 8.002415126398944e-05,
+      "loss": 2.4106,
+      "step": 9173
+    },
+    {
+      "epoch": 0.5643454724409449,
+      "grad_norm": 1.2546497583389282,
+      "learning_rate": 8.000520374660564e-05,
+      "loss": 2.6458,
+      "step": 9174
+    },
+    {
+      "epoch": 0.5644069881889764,
+      "grad_norm": 1.193748950958252,
+      "learning_rate": 7.998625697691565e-05,
+      "loss": 3.0464,
+      "step": 9175
+    },
+    {
+      "epoch": 0.5644685039370079,
+      "grad_norm": 1.2276636362075806,
+      "learning_rate": 7.996731095562811e-05,
+      "loss": 3.3271,
+      "step": 9176
+    },
+    {
+      "epoch": 0.5645300196850394,
+      "grad_norm": 1.328851342201233,
+      "learning_rate": 7.99483656834514e-05,
+      "loss": 2.4769,
+      "step": 9177
+    },
+    {
+      "epoch": 0.5645915354330708,
+      "grad_norm": 1.2387118339538574,
+      "learning_rate": 7.992942116109404e-05,
+      "loss": 2.8669,
+      "step": 9178
+    },
+    {
+      "epoch": 0.5646530511811023,
+      "grad_norm": 1.2539589405059814,
+      "learning_rate": 7.991047738926436e-05,
+      "loss": 2.3381,
+      "step": 9179
+    },
+    {
+      "epoch": 0.5647145669291339,
+      "grad_norm": 1.3684818744659424,
+      "learning_rate": 7.989153436867087e-05,
+      "loss": 2.7196,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5647760826771654,
+      "grad_norm": 1.400826334953308,
+      "learning_rate": 7.98725921000218e-05,
+      "loss": 2.42,
+      "step": 9181
+    },
+    {
+      "epoch": 0.5648375984251969,
+      "grad_norm": 1.1567476987838745,
+      "learning_rate": 7.985365058402559e-05,
+      "loss": 2.7568,
+      "step": 9182
+    },
+    {
+      "epoch": 0.5648991141732284,
+      "grad_norm": 1.0899925231933594,
+      "learning_rate": 7.983470982139049e-05,
+      "loss": 2.2161,
+      "step": 9183
+    },
+    {
+      "epoch": 0.5649606299212598,
+      "grad_norm": 1.32108736038208,
+      "learning_rate": 7.981576981282481e-05,
+      "loss": 2.1475,
+      "step": 9184
+    },
+    {
+      "epoch": 0.5650221456692913,
+      "grad_norm": 1.1008050441741943,
+      "learning_rate": 7.979683055903677e-05,
+      "loss": 3.0107,
+      "step": 9185
+    },
+    {
+      "epoch": 0.5650836614173228,
+      "grad_norm": 1.340855598449707,
+      "learning_rate": 7.977789206073465e-05,
+      "loss": 2.1599,
+      "step": 9186
+    },
+    {
+      "epoch": 0.5651451771653543,
+      "grad_norm": 1.303739309310913,
+      "learning_rate": 7.975895431862657e-05,
+      "loss": 2.3252,
+      "step": 9187
+    },
+    {
+      "epoch": 0.5652066929133859,
+      "grad_norm": 1.1287062168121338,
+      "learning_rate": 7.974001733342074e-05,
+      "loss": 2.027,
+      "step": 9188
+    },
+    {
+      "epoch": 0.5652682086614174,
+      "grad_norm": 1.2099697589874268,
+      "learning_rate": 7.97210811058253e-05,
+      "loss": 2.7669,
+      "step": 9189
+    },
+    {
+      "epoch": 0.5653297244094488,
+      "grad_norm": 1.240808129310608,
+      "learning_rate": 7.970214563654836e-05,
+      "loss": 2.5962,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5653912401574803,
+      "grad_norm": 1.0957903861999512,
+      "learning_rate": 7.968321092629796e-05,
+      "loss": 2.4734,
+      "step": 9191
+    },
+    {
+      "epoch": 0.5654527559055118,
+      "grad_norm": 1.1646203994750977,
+      "learning_rate": 7.96642769757822e-05,
+      "loss": 2.3751,
+      "step": 9192
+    },
+    {
+      "epoch": 0.5655142716535433,
+      "grad_norm": 1.2147629261016846,
+      "learning_rate": 7.964534378570905e-05,
+      "loss": 2.3793,
+      "step": 9193
+    },
+    {
+      "epoch": 0.5655757874015748,
+      "grad_norm": 1.2839164733886719,
+      "learning_rate": 7.962641135678658e-05,
+      "loss": 2.3239,
+      "step": 9194
+    },
+    {
+      "epoch": 0.5656373031496063,
+      "grad_norm": 1.4352805614471436,
+      "learning_rate": 7.960747968972269e-05,
+      "loss": 2.1718,
+      "step": 9195
+    },
+    {
+      "epoch": 0.5656988188976378,
+      "grad_norm": 1.2264831066131592,
+      "learning_rate": 7.958854878522539e-05,
+      "loss": 3.2121,
+      "step": 9196
+    },
+    {
+      "epoch": 0.5657603346456693,
+      "grad_norm": 1.1765464544296265,
+      "learning_rate": 7.956961864400252e-05,
+      "loss": 2.0396,
+      "step": 9197
+    },
+    {
+      "epoch": 0.5658218503937008,
+      "grad_norm": 1.1629294157028198,
+      "learning_rate": 7.955068926676199e-05,
+      "loss": 2.4046,
+      "step": 9198
+    },
+    {
+      "epoch": 0.5658833661417323,
+      "grad_norm": 1.4533395767211914,
+      "learning_rate": 7.953176065421169e-05,
+      "loss": 2.6505,
+      "step": 9199
+    },
+    {
+      "epoch": 0.5659448818897638,
+      "grad_norm": 1.2621383666992188,
+      "learning_rate": 7.951283280705936e-05,
+      "loss": 2.476,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5660063976377953,
+      "grad_norm": 1.2761176824569702,
+      "learning_rate": 7.949390572601286e-05,
+      "loss": 3.0138,
+      "step": 9201
+    },
+    {
+      "epoch": 0.5660679133858267,
+      "grad_norm": 1.0156033039093018,
+      "learning_rate": 7.947497941177993e-05,
+      "loss": 2.779,
+      "step": 9202
+    },
+    {
+      "epoch": 0.5661294291338582,
+      "grad_norm": 1.214859962463379,
+      "learning_rate": 7.945605386506834e-05,
+      "loss": 2.8796,
+      "step": 9203
+    },
+    {
+      "epoch": 0.5661909448818898,
+      "grad_norm": 1.288453459739685,
+      "learning_rate": 7.943712908658575e-05,
+      "loss": 3.2114,
+      "step": 9204
+    },
+    {
+      "epoch": 0.5662524606299213,
+      "grad_norm": 1.2112808227539062,
+      "learning_rate": 7.941820507703991e-05,
+      "loss": 2.3132,
+      "step": 9205
+    },
+    {
+      "epoch": 0.5663139763779528,
+      "grad_norm": 1.2328121662139893,
+      "learning_rate": 7.93992818371384e-05,
+      "loss": 2.8385,
+      "step": 9206
+    },
+    {
+      "epoch": 0.5663754921259843,
+      "grad_norm": 1.3043404817581177,
+      "learning_rate": 7.93803593675889e-05,
+      "loss": 2.7284,
+      "step": 9207
+    },
+    {
+      "epoch": 0.5664370078740157,
+      "grad_norm": 1.4716458320617676,
+      "learning_rate": 7.936143766909895e-05,
+      "loss": 2.4726,
+      "step": 9208
+    },
+    {
+      "epoch": 0.5664985236220472,
+      "grad_norm": 1.2195005416870117,
+      "learning_rate": 7.934251674237618e-05,
+      "loss": 2.6175,
+      "step": 9209
+    },
+    {
+      "epoch": 0.5665600393700787,
+      "grad_norm": 1.5274282693862915,
+      "learning_rate": 7.932359658812805e-05,
+      "loss": 2.2337,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5666215551181102,
+      "grad_norm": 1.1941131353378296,
+      "learning_rate": 7.930467720706217e-05,
+      "loss": 1.3523,
+      "step": 9211
+    },
+    {
+      "epoch": 0.5666830708661418,
+      "grad_norm": 1.3573615550994873,
+      "learning_rate": 7.928575859988592e-05,
+      "loss": 2.2996,
+      "step": 9212
+    },
+    {
+      "epoch": 0.5667445866141733,
+      "grad_norm": 1.6931321620941162,
+      "learning_rate": 7.926684076730682e-05,
+      "loss": 2.3539,
+      "step": 9213
+    },
+    {
+      "epoch": 0.5668061023622047,
+      "grad_norm": 1.1456520557403564,
+      "learning_rate": 7.924792371003223e-05,
+      "loss": 2.2065,
+      "step": 9214
+    },
+    {
+      "epoch": 0.5668676181102362,
+      "grad_norm": 1.1350946426391602,
+      "learning_rate": 7.922900742876963e-05,
+      "loss": 2.8052,
+      "step": 9215
+    },
+    {
+      "epoch": 0.5669291338582677,
+      "grad_norm": 1.3472753763198853,
+      "learning_rate": 7.921009192422628e-05,
+      "loss": 2.5328,
+      "step": 9216
+    },
+    {
+      "epoch": 0.5669906496062992,
+      "grad_norm": 1.2361211776733398,
+      "learning_rate": 7.91911771971096e-05,
+      "loss": 3.0518,
+      "step": 9217
+    },
+    {
+      "epoch": 0.5670521653543307,
+      "grad_norm": 1.2100505828857422,
+      "learning_rate": 7.917226324812682e-05,
+      "loss": 2.6659,
+      "step": 9218
+    },
+    {
+      "epoch": 0.5671136811023622,
+      "grad_norm": 1.3139570951461792,
+      "learning_rate": 7.915335007798531e-05,
+      "loss": 2.4234,
+      "step": 9219
+    },
+    {
+      "epoch": 0.5671751968503937,
+      "grad_norm": 1.2872577905654907,
+      "learning_rate": 7.91344376873922e-05,
+      "loss": 2.1197,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5672367125984252,
+      "grad_norm": 1.315695881843567,
+      "learning_rate": 7.911552607705482e-05,
+      "loss": 2.3462,
+      "step": 9221
+    },
+    {
+      "epoch": 0.5672982283464567,
+      "grad_norm": 1.1550352573394775,
+      "learning_rate": 7.909661524768028e-05,
+      "loss": 2.3278,
+      "step": 9222
+    },
+    {
+      "epoch": 0.5673597440944882,
+      "grad_norm": 1.2324837446212769,
+      "learning_rate": 7.90777051999758e-05,
+      "loss": 2.3678,
+      "step": 9223
+    },
+    {
+      "epoch": 0.5674212598425197,
+      "grad_norm": 1.1903263330459595,
+      "learning_rate": 7.905879593464843e-05,
+      "loss": 2.8764,
+      "step": 9224
+    },
+    {
+      "epoch": 0.5674827755905512,
+      "grad_norm": 1.1871098279953003,
+      "learning_rate": 7.903988745240539e-05,
+      "loss": 1.9389,
+      "step": 9225
+    },
+    {
+      "epoch": 0.5675442913385826,
+      "grad_norm": 1.1394492387771606,
+      "learning_rate": 7.902097975395363e-05,
+      "loss": 2.7431,
+      "step": 9226
+    },
+    {
+      "epoch": 0.5676058070866141,
+      "grad_norm": 1.147477149963379,
+      "learning_rate": 7.900207284000027e-05,
+      "loss": 2.3722,
+      "step": 9227
+    },
+    {
+      "epoch": 0.5676673228346457,
+      "grad_norm": 1.3549307584762573,
+      "learning_rate": 7.898316671125226e-05,
+      "loss": 2.4954,
+      "step": 9228
+    },
+    {
+      "epoch": 0.5677288385826772,
+      "grad_norm": 1.3436921834945679,
+      "learning_rate": 7.896426136841665e-05,
+      "loss": 2.3096,
+      "step": 9229
+    },
+    {
+      "epoch": 0.5677903543307087,
+      "grad_norm": 1.3178197145462036,
+      "learning_rate": 7.894535681220036e-05,
+      "loss": 2.4336,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5678518700787402,
+      "grad_norm": 1.0593019723892212,
+      "learning_rate": 7.892645304331036e-05,
+      "loss": 1.9484,
+      "step": 9231
+    },
+    {
+      "epoch": 0.5679133858267716,
+      "grad_norm": 1.3977975845336914,
+      "learning_rate": 7.890755006245345e-05,
+      "loss": 3.0893,
+      "step": 9232
+    },
+    {
+      "epoch": 0.5679749015748031,
+      "grad_norm": 1.7110482454299927,
+      "learning_rate": 7.888864787033658e-05,
+      "loss": 3.1428,
+      "step": 9233
+    },
+    {
+      "epoch": 0.5680364173228346,
+      "grad_norm": 1.3258730173110962,
+      "learning_rate": 7.886974646766655e-05,
+      "loss": 2.4486,
+      "step": 9234
+    },
+    {
+      "epoch": 0.5680979330708661,
+      "grad_norm": 1.4200063943862915,
+      "learning_rate": 7.885084585515021e-05,
+      "loss": 2.0141,
+      "step": 9235
+    },
+    {
+      "epoch": 0.5681594488188977,
+      "grad_norm": 1.2310881614685059,
+      "learning_rate": 7.883194603349429e-05,
+      "loss": 2.3087,
+      "step": 9236
+    },
+    {
+      "epoch": 0.5682209645669292,
+      "grad_norm": 1.2604215145111084,
+      "learning_rate": 7.881304700340551e-05,
+      "loss": 2.9632,
+      "step": 9237
+    },
+    {
+      "epoch": 0.5682824803149606,
+      "grad_norm": 1.3975037336349487,
+      "learning_rate": 7.879414876559067e-05,
+      "loss": 2.5726,
+      "step": 9238
+    },
+    {
+      "epoch": 0.5683439960629921,
+      "grad_norm": 1.0382124185562134,
+      "learning_rate": 7.877525132075639e-05,
+      "loss": 2.6779,
+      "step": 9239
+    },
+    {
+      "epoch": 0.5684055118110236,
+      "grad_norm": 1.2095682621002197,
+      "learning_rate": 7.875635466960941e-05,
+      "loss": 2.5865,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5684670275590551,
+      "grad_norm": 1.2085011005401611,
+      "learning_rate": 7.873745881285625e-05,
+      "loss": 2.2053,
+      "step": 9241
+    },
+    {
+      "epoch": 0.5685285433070866,
+      "grad_norm": 1.213900089263916,
+      "learning_rate": 7.87185637512036e-05,
+      "loss": 2.5321,
+      "step": 9242
+    },
+    {
+      "epoch": 0.5685900590551181,
+      "grad_norm": 1.192629337310791,
+      "learning_rate": 7.869966948535798e-05,
+      "loss": 2.7393,
+      "step": 9243
+    },
+    {
+      "epoch": 0.5686515748031497,
+      "grad_norm": 1.6835094690322876,
+      "learning_rate": 7.868077601602599e-05,
+      "loss": 3.7731,
+      "step": 9244
+    },
+    {
+      "epoch": 0.5687130905511811,
+      "grad_norm": 1.215097427368164,
+      "learning_rate": 7.866188334391405e-05,
+      "loss": 2.6606,
+      "step": 9245
+    },
+    {
+      "epoch": 0.5687746062992126,
+      "grad_norm": 1.299522042274475,
+      "learning_rate": 7.86429914697287e-05,
+      "loss": 2.5891,
+      "step": 9246
+    },
+    {
+      "epoch": 0.5688361220472441,
+      "grad_norm": 1.2962617874145508,
+      "learning_rate": 7.862410039417636e-05,
+      "loss": 2.399,
+      "step": 9247
+    },
+    {
+      "epoch": 0.5688976377952756,
+      "grad_norm": 1.451585292816162,
+      "learning_rate": 7.86052101179635e-05,
+      "loss": 2.3002,
+      "step": 9248
+    },
+    {
+      "epoch": 0.5689591535433071,
+      "grad_norm": 1.2429418563842773,
+      "learning_rate": 7.858632064179645e-05,
+      "loss": 2.4169,
+      "step": 9249
+    },
+    {
+      "epoch": 0.5690206692913385,
+      "grad_norm": 1.3594270944595337,
+      "learning_rate": 7.856743196638166e-05,
+      "loss": 2.4236,
+      "step": 9250
+    },
+    {
+      "epoch": 0.56908218503937,
+      "grad_norm": 1.162048101425171,
+      "learning_rate": 7.854854409242536e-05,
+      "loss": 2.7935,
+      "step": 9251
+    },
+    {
+      "epoch": 0.5691437007874016,
+      "grad_norm": 1.1715792417526245,
+      "learning_rate": 7.85296570206339e-05,
+      "loss": 3.005,
+      "step": 9252
+    },
+    {
+      "epoch": 0.5692052165354331,
+      "grad_norm": 1.326958179473877,
+      "learning_rate": 7.851077075171353e-05,
+      "loss": 2.8664,
+      "step": 9253
+    },
+    {
+      "epoch": 0.5692667322834646,
+      "grad_norm": 1.301504135131836,
+      "learning_rate": 7.849188528637054e-05,
+      "loss": 2.2954,
+      "step": 9254
+    },
+    {
+      "epoch": 0.5693282480314961,
+      "grad_norm": 1.3988796472549438,
+      "learning_rate": 7.847300062531106e-05,
+      "loss": 2.7989,
+      "step": 9255
+    },
+    {
+      "epoch": 0.5693897637795275,
+      "grad_norm": 1.0259157419204712,
+      "learning_rate": 7.845411676924139e-05,
+      "loss": 1.2838,
+      "step": 9256
+    },
+    {
+      "epoch": 0.569451279527559,
+      "grad_norm": 1.1061937808990479,
+      "learning_rate": 7.843523371886755e-05,
+      "loss": 3.3946,
+      "step": 9257
+    },
+    {
+      "epoch": 0.5695127952755905,
+      "grad_norm": 1.0783976316452026,
+      "learning_rate": 7.841635147489574e-05,
+      "loss": 3.3835,
+      "step": 9258
+    },
+    {
+      "epoch": 0.569574311023622,
+      "grad_norm": 1.1084072589874268,
+      "learning_rate": 7.839747003803202e-05,
+      "loss": 2.6946,
+      "step": 9259
+    },
+    {
+      "epoch": 0.5696358267716536,
+      "grad_norm": 1.066986322402954,
+      "learning_rate": 7.83785894089825e-05,
+      "loss": 2.3866,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5696973425196851,
+      "grad_norm": 1.176194667816162,
+      "learning_rate": 7.835970958845314e-05,
+      "loss": 2.6566,
+      "step": 9261
+    },
+    {
+      "epoch": 0.5697588582677166,
+      "grad_norm": 1.3342446088790894,
+      "learning_rate": 7.834083057714998e-05,
+      "loss": 2.3455,
+      "step": 9262
+    },
+    {
+      "epoch": 0.569820374015748,
+      "grad_norm": 1.299846887588501,
+      "learning_rate": 7.832195237577896e-05,
+      "loss": 2.7758,
+      "step": 9263
+    },
+    {
+      "epoch": 0.5698818897637795,
+      "grad_norm": 1.1531678438186646,
+      "learning_rate": 7.830307498504607e-05,
+      "loss": 2.6776,
+      "step": 9264
+    },
+    {
+      "epoch": 0.569943405511811,
+      "grad_norm": 1.2765926122665405,
+      "learning_rate": 7.828419840565716e-05,
+      "loss": 2.383,
+      "step": 9265
+    },
+    {
+      "epoch": 0.5700049212598425,
+      "grad_norm": 1.2881745100021362,
+      "learning_rate": 7.826532263831819e-05,
+      "loss": 2.235,
+      "step": 9266
+    },
+    {
+      "epoch": 0.570066437007874,
+      "grad_norm": 1.2249394655227661,
+      "learning_rate": 7.82464476837349e-05,
+      "loss": 2.9174,
+      "step": 9267
+    },
+    {
+      "epoch": 0.5701279527559056,
+      "grad_norm": 1.1977345943450928,
+      "learning_rate": 7.82275735426132e-05,
+      "loss": 2.6152,
+      "step": 9268
+    },
+    {
+      "epoch": 0.570189468503937,
+      "grad_norm": 1.4781162738800049,
+      "learning_rate": 7.820870021565882e-05,
+      "loss": 2.4991,
+      "step": 9269
+    },
+    {
+      "epoch": 0.5702509842519685,
+      "grad_norm": 1.1107479333877563,
+      "learning_rate": 7.818982770357759e-05,
+      "loss": 2.9415,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5703125,
+      "grad_norm": 1.0871528387069702,
+      "learning_rate": 7.817095600707513e-05,
+      "loss": 2.9585,
+      "step": 9271
+    },
+    {
+      "epoch": 0.5703740157480315,
+      "grad_norm": 1.1770459413528442,
+      "learning_rate": 7.815208512685723e-05,
+      "loss": 2.9155,
+      "step": 9272
+    },
+    {
+      "epoch": 0.570435531496063,
+      "grad_norm": 1.2389638423919678,
+      "learning_rate": 7.813321506362951e-05,
+      "loss": 2.3535,
+      "step": 9273
+    },
+    {
+      "epoch": 0.5704970472440944,
+      "grad_norm": 1.1743546724319458,
+      "learning_rate": 7.811434581809759e-05,
+      "loss": 2.3768,
+      "step": 9274
+    },
+    {
+      "epoch": 0.570558562992126,
+      "grad_norm": 1.4517600536346436,
+      "learning_rate": 7.809547739096715e-05,
+      "loss": 2.4187,
+      "step": 9275
+    },
+    {
+      "epoch": 0.5706200787401575,
+      "grad_norm": 1.1864025592803955,
+      "learning_rate": 7.807660978294365e-05,
+      "loss": 3.1611,
+      "step": 9276
+    },
+    {
+      "epoch": 0.570681594488189,
+      "grad_norm": 1.1310100555419922,
+      "learning_rate": 7.805774299473274e-05,
+      "loss": 2.5468,
+      "step": 9277
+    },
+    {
+      "epoch": 0.5707431102362205,
+      "grad_norm": 1.2809008359909058,
+      "learning_rate": 7.803887702703986e-05,
+      "loss": 2.1027,
+      "step": 9278
+    },
+    {
+      "epoch": 0.570804625984252,
+      "grad_norm": 1.1592967510223389,
+      "learning_rate": 7.802001188057057e-05,
+      "loss": 2.4483,
+      "step": 9279
+    },
+    {
+      "epoch": 0.5708661417322834,
+      "grad_norm": 1.1143025159835815,
+      "learning_rate": 7.800114755603021e-05,
+      "loss": 2.717,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5709276574803149,
+      "grad_norm": 1.256090521812439,
+      "learning_rate": 7.79822840541243e-05,
+      "loss": 2.3159,
+      "step": 9281
+    },
+    {
+      "epoch": 0.5709891732283464,
+      "grad_norm": 1.1409530639648438,
+      "learning_rate": 7.796342137555817e-05,
+      "loss": 3.1681,
+      "step": 9282
+    },
+    {
+      "epoch": 0.571050688976378,
+      "grad_norm": 1.099615216255188,
+      "learning_rate": 7.79445595210372e-05,
+      "loss": 2.753,
+      "step": 9283
+    },
+    {
+      "epoch": 0.5711122047244095,
+      "grad_norm": 1.2897449731826782,
+      "learning_rate": 7.792569849126674e-05,
+      "loss": 2.6178,
+      "step": 9284
+    },
+    {
+      "epoch": 0.571173720472441,
+      "grad_norm": 1.1389482021331787,
+      "learning_rate": 7.790683828695208e-05,
+      "loss": 2.8359,
+      "step": 9285
+    },
+    {
+      "epoch": 0.5712352362204725,
+      "grad_norm": 1.3406850099563599,
+      "learning_rate": 7.788797890879843e-05,
+      "loss": 2.3994,
+      "step": 9286
+    },
+    {
+      "epoch": 0.5712967519685039,
+      "grad_norm": 1.0618406534194946,
+      "learning_rate": 7.78691203575111e-05,
+      "loss": 2.9031,
+      "step": 9287
+    },
+    {
+      "epoch": 0.5713582677165354,
+      "grad_norm": 1.3052003383636475,
+      "learning_rate": 7.785026263379524e-05,
+      "loss": 1.7135,
+      "step": 9288
+    },
+    {
+      "epoch": 0.5714197834645669,
+      "grad_norm": 1.0972894430160522,
+      "learning_rate": 7.783140573835609e-05,
+      "loss": 2.8625,
+      "step": 9289
+    },
+    {
+      "epoch": 0.5714812992125984,
+      "grad_norm": 1.2231553792953491,
+      "learning_rate": 7.781254967189869e-05,
+      "loss": 1.7819,
+      "step": 9290
+    },
+    {
+      "epoch": 0.57154281496063,
+      "grad_norm": 1.5628801584243774,
+      "learning_rate": 7.779369443512826e-05,
+      "loss": 2.2232,
+      "step": 9291
+    },
+    {
+      "epoch": 0.5716043307086615,
+      "grad_norm": 1.4547151327133179,
+      "learning_rate": 7.777484002874981e-05,
+      "loss": 2.2342,
+      "step": 9292
+    },
+    {
+      "epoch": 0.5716658464566929,
+      "grad_norm": 1.0455586910247803,
+      "learning_rate": 7.775598645346842e-05,
+      "loss": 2.6647,
+      "step": 9293
+    },
+    {
+      "epoch": 0.5717273622047244,
+      "grad_norm": 1.418337106704712,
+      "learning_rate": 7.773713370998908e-05,
+      "loss": 2.1554,
+      "step": 9294
+    },
+    {
+      "epoch": 0.5717888779527559,
+      "grad_norm": 1.2480595111846924,
+      "learning_rate": 7.771828179901685e-05,
+      "loss": 2.3633,
+      "step": 9295
+    },
+    {
+      "epoch": 0.5718503937007874,
+      "grad_norm": 1.1506893634796143,
+      "learning_rate": 7.769943072125658e-05,
+      "loss": 2.4473,
+      "step": 9296
+    },
+    {
+      "epoch": 0.5719119094488189,
+      "grad_norm": 1.3067868947982788,
+      "learning_rate": 7.768058047741328e-05,
+      "loss": 2.0705,
+      "step": 9297
+    },
+    {
+      "epoch": 0.5719734251968503,
+      "grad_norm": 1.2587029933929443,
+      "learning_rate": 7.766173106819178e-05,
+      "loss": 2.1508,
+      "step": 9298
+    },
+    {
+      "epoch": 0.5720349409448819,
+      "grad_norm": 1.3105967044830322,
+      "learning_rate": 7.764288249429701e-05,
+      "loss": 2.3555,
+      "step": 9299
+    },
+    {
+      "epoch": 0.5720964566929134,
+      "grad_norm": 1.2627415657043457,
+      "learning_rate": 7.762403475643373e-05,
+      "loss": 1.7639,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5721579724409449,
+      "grad_norm": 1.1247694492340088,
+      "learning_rate": 7.760518785530684e-05,
+      "loss": 2.6536,
+      "step": 9301
+    },
+    {
+      "epoch": 0.5722194881889764,
+      "grad_norm": 1.1833165884017944,
+      "learning_rate": 7.7586341791621e-05,
+      "loss": 2.7066,
+      "step": 9302
+    },
+    {
+      "epoch": 0.5722810039370079,
+      "grad_norm": 1.4916692972183228,
+      "learning_rate": 7.756749656608102e-05,
+      "loss": 1.4829,
+      "step": 9303
+    },
+    {
+      "epoch": 0.5723425196850394,
+      "grad_norm": 1.0690722465515137,
+      "learning_rate": 7.754865217939156e-05,
+      "loss": 1.9623,
+      "step": 9304
+    },
+    {
+      "epoch": 0.5724040354330708,
+      "grad_norm": 1.2861212491989136,
+      "learning_rate": 7.752980863225735e-05,
+      "loss": 1.8013,
+      "step": 9305
+    },
+    {
+      "epoch": 0.5724655511811023,
+      "grad_norm": 1.1186821460723877,
+      "learning_rate": 7.751096592538297e-05,
+      "loss": 2.0402,
+      "step": 9306
+    },
+    {
+      "epoch": 0.5725270669291339,
+      "grad_norm": 1.547999620437622,
+      "learning_rate": 7.749212405947306e-05,
+      "loss": 2.9698,
+      "step": 9307
+    },
+    {
+      "epoch": 0.5725885826771654,
+      "grad_norm": 1.2610899209976196,
+      "learning_rate": 7.747328303523221e-05,
+      "loss": 1.8604,
+      "step": 9308
+    },
+    {
+      "epoch": 0.5726500984251969,
+      "grad_norm": 1.5092477798461914,
+      "learning_rate": 7.745444285336496e-05,
+      "loss": 2.7154,
+      "step": 9309
+    },
+    {
+      "epoch": 0.5727116141732284,
+      "grad_norm": 1.1823476552963257,
+      "learning_rate": 7.743560351457586e-05,
+      "loss": 1.8109,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5727731299212598,
+      "grad_norm": 1.166075587272644,
+      "learning_rate": 7.741676501956933e-05,
+      "loss": 2.6241,
+      "step": 9311
+    },
+    {
+      "epoch": 0.5728346456692913,
+      "grad_norm": 1.227495789527893,
+      "learning_rate": 7.739792736904987e-05,
+      "loss": 2.5767,
+      "step": 9312
+    },
+    {
+      "epoch": 0.5728961614173228,
+      "grad_norm": 1.246163010597229,
+      "learning_rate": 7.737909056372188e-05,
+      "loss": 2.7713,
+      "step": 9313
+    },
+    {
+      "epoch": 0.5729576771653543,
+      "grad_norm": 1.565739393234253,
+      "learning_rate": 7.73602546042898e-05,
+      "loss": 2.2087,
+      "step": 9314
+    },
+    {
+      "epoch": 0.5730191929133859,
+      "grad_norm": 1.2883150577545166,
+      "learning_rate": 7.734141949145792e-05,
+      "loss": 2.1557,
+      "step": 9315
+    },
+    {
+      "epoch": 0.5730807086614174,
+      "grad_norm": 1.2486847639083862,
+      "learning_rate": 7.732258522593064e-05,
+      "loss": 1.9257,
+      "step": 9316
+    },
+    {
+      "epoch": 0.5731422244094488,
+      "grad_norm": 1.5033259391784668,
+      "learning_rate": 7.730375180841217e-05,
+      "loss": 1.9739,
+      "step": 9317
+    },
+    {
+      "epoch": 0.5732037401574803,
+      "grad_norm": 0.997412919998169,
+      "learning_rate": 7.728491923960685e-05,
+      "loss": 3.0926,
+      "step": 9318
+    },
+    {
+      "epoch": 0.5732652559055118,
+      "grad_norm": 1.19733726978302,
+      "learning_rate": 7.726608752021888e-05,
+      "loss": 2.7395,
+      "step": 9319
+    },
+    {
+      "epoch": 0.5733267716535433,
+      "grad_norm": 1.1179243326187134,
+      "learning_rate": 7.724725665095251e-05,
+      "loss": 2.658,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5733882874015748,
+      "grad_norm": 1.1265935897827148,
+      "learning_rate": 7.722842663251182e-05,
+      "loss": 2.5678,
+      "step": 9321
+    },
+    {
+      "epoch": 0.5734498031496063,
+      "grad_norm": 1.2549786567687988,
+      "learning_rate": 7.720959746560103e-05,
+      "loss": 2.5332,
+      "step": 9322
+    },
+    {
+      "epoch": 0.5735113188976378,
+      "grad_norm": 1.3540998697280884,
+      "learning_rate": 7.719076915092417e-05,
+      "loss": 2.3661,
+      "step": 9323
+    },
+    {
+      "epoch": 0.5735728346456693,
+      "grad_norm": 1.7507370710372925,
+      "learning_rate": 7.717194168918541e-05,
+      "loss": 2.4835,
+      "step": 9324
+    },
+    {
+      "epoch": 0.5736343503937008,
+      "grad_norm": 1.1641815900802612,
+      "learning_rate": 7.715311508108867e-05,
+      "loss": 2.7803,
+      "step": 9325
+    },
+    {
+      "epoch": 0.5736958661417323,
+      "grad_norm": 1.082871437072754,
+      "learning_rate": 7.71342893273381e-05,
+      "loss": 2.5273,
+      "step": 9326
+    },
+    {
+      "epoch": 0.5737573818897638,
+      "grad_norm": 1.3960578441619873,
+      "learning_rate": 7.711546442863755e-05,
+      "loss": 2.5766,
+      "step": 9327
+    },
+    {
+      "epoch": 0.5738188976377953,
+      "grad_norm": 1.3181875944137573,
+      "learning_rate": 7.709664038569107e-05,
+      "loss": 2.8572,
+      "step": 9328
+    },
+    {
+      "epoch": 0.5738804133858267,
+      "grad_norm": 1.1236287355422974,
+      "learning_rate": 7.707781719920248e-05,
+      "loss": 2.6418,
+      "step": 9329
+    },
+    {
+      "epoch": 0.5739419291338582,
+      "grad_norm": 1.163524866104126,
+      "learning_rate": 7.705899486987575e-05,
+      "loss": 2.2129,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5740034448818898,
+      "grad_norm": 1.259597897529602,
+      "learning_rate": 7.704017339841467e-05,
+      "loss": 2.1225,
+      "step": 9331
+    },
+    {
+      "epoch": 0.5740649606299213,
+      "grad_norm": 1.1443169116973877,
+      "learning_rate": 7.702135278552308e-05,
+      "loss": 2.7587,
+      "step": 9332
+    },
+    {
+      "epoch": 0.5741264763779528,
+      "grad_norm": 1.0731087923049927,
+      "learning_rate": 7.700253303190474e-05,
+      "loss": 2.7889,
+      "step": 9333
+    },
+    {
+      "epoch": 0.5741879921259843,
+      "grad_norm": 1.3807424306869507,
+      "learning_rate": 7.698371413826346e-05,
+      "loss": 2.8968,
+      "step": 9334
+    },
+    {
+      "epoch": 0.5742495078740157,
+      "grad_norm": 1.1772398948669434,
+      "learning_rate": 7.696489610530291e-05,
+      "loss": 2.9525,
+      "step": 9335
+    },
+    {
+      "epoch": 0.5743110236220472,
+      "grad_norm": 1.2594553232192993,
+      "learning_rate": 7.694607893372684e-05,
+      "loss": 1.7878,
+      "step": 9336
+    },
+    {
+      "epoch": 0.5743725393700787,
+      "grad_norm": 1.143923282623291,
+      "learning_rate": 7.692726262423882e-05,
+      "loss": 2.2797,
+      "step": 9337
+    },
+    {
+      "epoch": 0.5744340551181102,
+      "grad_norm": 1.1327874660491943,
+      "learning_rate": 7.690844717754253e-05,
+      "loss": 2.7342,
+      "step": 9338
+    },
+    {
+      "epoch": 0.5744955708661418,
+      "grad_norm": 1.300134301185608,
+      "learning_rate": 7.688963259434155e-05,
+      "loss": 2.6419,
+      "step": 9339
+    },
+    {
+      "epoch": 0.5745570866141733,
+      "grad_norm": 1.251158356666565,
+      "learning_rate": 7.687081887533947e-05,
+      "loss": 2.6341,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5746186023622047,
+      "grad_norm": 1.1814579963684082,
+      "learning_rate": 7.685200602123977e-05,
+      "loss": 2.0585,
+      "step": 9341
+    },
+    {
+      "epoch": 0.5746801181102362,
+      "grad_norm": 1.1902294158935547,
+      "learning_rate": 7.683319403274597e-05,
+      "loss": 2.4401,
+      "step": 9342
+    },
+    {
+      "epoch": 0.5747416338582677,
+      "grad_norm": 1.2025398015975952,
+      "learning_rate": 7.681438291056152e-05,
+      "loss": 2.443,
+      "step": 9343
+    },
+    {
+      "epoch": 0.5748031496062992,
+      "grad_norm": 1.7284257411956787,
+      "learning_rate": 7.679557265538988e-05,
+      "loss": 2.2067,
+      "step": 9344
+    },
+    {
+      "epoch": 0.5748646653543307,
+      "grad_norm": 1.239438772201538,
+      "learning_rate": 7.677676326793441e-05,
+      "loss": 2.3637,
+      "step": 9345
+    },
+    {
+      "epoch": 0.5749261811023622,
+      "grad_norm": 1.1454373598098755,
+      "learning_rate": 7.675795474889854e-05,
+      "loss": 2.468,
+      "step": 9346
+    },
+    {
+      "epoch": 0.5749876968503937,
+      "grad_norm": 1.2209702730178833,
+      "learning_rate": 7.673914709898555e-05,
+      "loss": 3.1205,
+      "step": 9347
+    },
+    {
+      "epoch": 0.5750492125984252,
+      "grad_norm": 1.3839753866195679,
+      "learning_rate": 7.672034031889872e-05,
+      "loss": 3.0548,
+      "step": 9348
+    },
+    {
+      "epoch": 0.5751107283464567,
+      "grad_norm": 1.3539707660675049,
+      "learning_rate": 7.67015344093414e-05,
+      "loss": 1.6862,
+      "step": 9349
+    },
+    {
+      "epoch": 0.5751722440944882,
+      "grad_norm": 1.1839995384216309,
+      "learning_rate": 7.668272937101673e-05,
+      "loss": 2.9972,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5752337598425197,
+      "grad_norm": 1.1765819787979126,
+      "learning_rate": 7.666392520462798e-05,
+      "loss": 2.3123,
+      "step": 9351
+    },
+    {
+      "epoch": 0.5752952755905512,
+      "grad_norm": 1.2683650255203247,
+      "learning_rate": 7.664512191087828e-05,
+      "loss": 2.3579,
+      "step": 9352
+    },
+    {
+      "epoch": 0.5753567913385826,
+      "grad_norm": 1.0331428050994873,
+      "learning_rate": 7.662631949047082e-05,
+      "loss": 2.9869,
+      "step": 9353
+    },
+    {
+      "epoch": 0.5754183070866141,
+      "grad_norm": 1.7075783014297485,
+      "learning_rate": 7.660751794410866e-05,
+      "loss": 3.2368,
+      "step": 9354
+    },
+    {
+      "epoch": 0.5754798228346457,
+      "grad_norm": 1.2772608995437622,
+      "learning_rate": 7.658871727249492e-05,
+      "loss": 2.3693,
+      "step": 9355
+    },
+    {
+      "epoch": 0.5755413385826772,
+      "grad_norm": 1.3124580383300781,
+      "learning_rate": 7.656991747633257e-05,
+      "loss": 2.3432,
+      "step": 9356
+    },
+    {
+      "epoch": 0.5756028543307087,
+      "grad_norm": 1.1719385385513306,
+      "learning_rate": 7.655111855632466e-05,
+      "loss": 3.0776,
+      "step": 9357
+    },
+    {
+      "epoch": 0.5756643700787402,
+      "grad_norm": 1.0214768648147583,
+      "learning_rate": 7.653232051317417e-05,
+      "loss": 3.2185,
+      "step": 9358
+    },
+    {
+      "epoch": 0.5757258858267716,
+      "grad_norm": 1.124907374382019,
+      "learning_rate": 7.651352334758406e-05,
+      "loss": 1.6564,
+      "step": 9359
+    },
+    {
+      "epoch": 0.5757874015748031,
+      "grad_norm": 1.333434820175171,
+      "learning_rate": 7.649472706025716e-05,
+      "loss": 2.11,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5758489173228346,
+      "grad_norm": 1.1822327375411987,
+      "learning_rate": 7.647593165189643e-05,
+      "loss": 2.8265,
+      "step": 9361
+    },
+    {
+      "epoch": 0.5759104330708661,
+      "grad_norm": 1.2103010416030884,
+      "learning_rate": 7.645713712320467e-05,
+      "loss": 2.5254,
+      "step": 9362
+    },
+    {
+      "epoch": 0.5759719488188977,
+      "grad_norm": 1.0550392866134644,
+      "learning_rate": 7.643834347488473e-05,
+      "loss": 2.1545,
+      "step": 9363
+    },
+    {
+      "epoch": 0.5760334645669292,
+      "grad_norm": 1.331407904624939,
+      "learning_rate": 7.641955070763932e-05,
+      "loss": 2.797,
+      "step": 9364
+    },
+    {
+      "epoch": 0.5760949803149606,
+      "grad_norm": 1.1479434967041016,
+      "learning_rate": 7.640075882217126e-05,
+      "loss": 2.7305,
+      "step": 9365
+    },
+    {
+      "epoch": 0.5761564960629921,
+      "grad_norm": 1.261290431022644,
+      "learning_rate": 7.63819678191832e-05,
+      "loss": 3.3686,
+      "step": 9366
+    },
+    {
+      "epoch": 0.5762180118110236,
+      "grad_norm": 1.1248596906661987,
+      "learning_rate": 7.636317769937789e-05,
+      "loss": 3.0713,
+      "step": 9367
+    },
+    {
+      "epoch": 0.5762795275590551,
+      "grad_norm": 1.1534780263900757,
+      "learning_rate": 7.634438846345787e-05,
+      "loss": 2.4419,
+      "step": 9368
+    },
+    {
+      "epoch": 0.5763410433070866,
+      "grad_norm": 1.1155916452407837,
+      "learning_rate": 7.632560011212588e-05,
+      "loss": 3.1738,
+      "step": 9369
+    },
+    {
+      "epoch": 0.5764025590551181,
+      "grad_norm": 1.4414855241775513,
+      "learning_rate": 7.630681264608437e-05,
+      "loss": 2.3832,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5764640748031497,
+      "grad_norm": 1.29059898853302,
+      "learning_rate": 7.6288026066036e-05,
+      "loss": 2.5198,
+      "step": 9371
+    },
+    {
+      "epoch": 0.5765255905511811,
+      "grad_norm": 1.281938076019287,
+      "learning_rate": 7.62692403726832e-05,
+      "loss": 2.97,
+      "step": 9372
+    },
+    {
+      "epoch": 0.5765871062992126,
+      "grad_norm": 1.307450294494629,
+      "learning_rate": 7.62504555667285e-05,
+      "loss": 2.746,
+      "step": 9373
+    },
+    {
+      "epoch": 0.5766486220472441,
+      "grad_norm": 1.130194067955017,
+      "learning_rate": 7.623167164887432e-05,
+      "loss": 2.5798,
+      "step": 9374
+    },
+    {
+      "epoch": 0.5767101377952756,
+      "grad_norm": 1.2476087808609009,
+      "learning_rate": 7.621288861982312e-05,
+      "loss": 2.3272,
+      "step": 9375
+    },
+    {
+      "epoch": 0.5767716535433071,
+      "grad_norm": 1.2416155338287354,
+      "learning_rate": 7.61941064802772e-05,
+      "loss": 3.1215,
+      "step": 9376
+    },
+    {
+      "epoch": 0.5768331692913385,
+      "grad_norm": 0.9849461913108826,
+      "learning_rate": 7.617532523093898e-05,
+      "loss": 3.1494,
+      "step": 9377
+    },
+    {
+      "epoch": 0.57689468503937,
+      "grad_norm": 1.22256600856781,
+      "learning_rate": 7.615654487251071e-05,
+      "loss": 2.7241,
+      "step": 9378
+    },
+    {
+      "epoch": 0.5769562007874016,
+      "grad_norm": 1.2150330543518066,
+      "learning_rate": 7.613776540569475e-05,
+      "loss": 2.2474,
+      "step": 9379
+    },
+    {
+      "epoch": 0.5770177165354331,
+      "grad_norm": 1.1532442569732666,
+      "learning_rate": 7.611898683119328e-05,
+      "loss": 2.6676,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5770792322834646,
+      "grad_norm": 1.5998376607894897,
+      "learning_rate": 7.610020914970857e-05,
+      "loss": 3.0108,
+      "step": 9381
+    },
+    {
+      "epoch": 0.5771407480314961,
+      "grad_norm": 1.8461917638778687,
+      "learning_rate": 7.608143236194272e-05,
+      "loss": 2.102,
+      "step": 9382
+    },
+    {
+      "epoch": 0.5772022637795275,
+      "grad_norm": 1.1816844940185547,
+      "learning_rate": 7.606265646859796e-05,
+      "loss": 2.4222,
+      "step": 9383
+    },
+    {
+      "epoch": 0.577263779527559,
+      "grad_norm": 1.0674670934677124,
+      "learning_rate": 7.604388147037639e-05,
+      "loss": 2.5805,
+      "step": 9384
+    },
+    {
+      "epoch": 0.5773252952755905,
+      "grad_norm": 1.2895221710205078,
+      "learning_rate": 7.602510736798002e-05,
+      "loss": 2.5747,
+      "step": 9385
+    },
+    {
+      "epoch": 0.577386811023622,
+      "grad_norm": 1.2674909830093384,
+      "learning_rate": 7.600633416211098e-05,
+      "loss": 2.4909,
+      "step": 9386
+    },
+    {
+      "epoch": 0.5774483267716536,
+      "grad_norm": 1.3358712196350098,
+      "learning_rate": 7.598756185347122e-05,
+      "loss": 2.4595,
+      "step": 9387
+    },
+    {
+      "epoch": 0.5775098425196851,
+      "grad_norm": 1.3965338468551636,
+      "learning_rate": 7.596879044276277e-05,
+      "loss": 2.4882,
+      "step": 9388
+    },
+    {
+      "epoch": 0.5775713582677166,
+      "grad_norm": 1.2473030090332031,
+      "learning_rate": 7.595001993068754e-05,
+      "loss": 2.0914,
+      "step": 9389
+    },
+    {
+      "epoch": 0.577632874015748,
+      "grad_norm": 1.5213251113891602,
+      "learning_rate": 7.59312503179475e-05,
+      "loss": 3.6832,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5776943897637795,
+      "grad_norm": 1.282532811164856,
+      "learning_rate": 7.591248160524444e-05,
+      "loss": 2.1024,
+      "step": 9391
+    },
+    {
+      "epoch": 0.577755905511811,
+      "grad_norm": 1.357256293296814,
+      "learning_rate": 7.589371379328027e-05,
+      "loss": 2.7612,
+      "step": 9392
+    },
+    {
+      "epoch": 0.5778174212598425,
+      "grad_norm": 1.2479603290557861,
+      "learning_rate": 7.587494688275679e-05,
+      "loss": 2.6364,
+      "step": 9393
+    },
+    {
+      "epoch": 0.577878937007874,
+      "grad_norm": 1.5714020729064941,
+      "learning_rate": 7.585618087437578e-05,
+      "loss": 2.2009,
+      "step": 9394
+    },
+    {
+      "epoch": 0.5779404527559056,
+      "grad_norm": 1.2428719997406006,
+      "learning_rate": 7.583741576883895e-05,
+      "loss": 1.9602,
+      "step": 9395
+    },
+    {
+      "epoch": 0.578001968503937,
+      "grad_norm": 1.2057615518569946,
+      "learning_rate": 7.581865156684805e-05,
+      "loss": 3.1639,
+      "step": 9396
+    },
+    {
+      "epoch": 0.5780634842519685,
+      "grad_norm": 1.3152683973312378,
+      "learning_rate": 7.579988826910473e-05,
+      "loss": 2.3,
+      "step": 9397
+    },
+    {
+      "epoch": 0.578125,
+      "grad_norm": 1.1861573457717896,
+      "learning_rate": 7.578112587631065e-05,
+      "loss": 2.2712,
+      "step": 9398
+    },
+    {
+      "epoch": 0.5781865157480315,
+      "grad_norm": 1.1275148391723633,
+      "learning_rate": 7.576236438916742e-05,
+      "loss": 2.7922,
+      "step": 9399
+    },
+    {
+      "epoch": 0.578248031496063,
+      "grad_norm": 1.4953693151474,
+      "learning_rate": 7.574360380837663e-05,
+      "loss": 2.4471,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5783095472440944,
+      "grad_norm": 1.2460459470748901,
+      "learning_rate": 7.572484413463976e-05,
+      "loss": 2.9866,
+      "step": 9401
+    },
+    {
+      "epoch": 0.578371062992126,
+      "grad_norm": 1.0710158348083496,
+      "learning_rate": 7.570608536865839e-05,
+      "loss": 3.1652,
+      "step": 9402
+    },
+    {
+      "epoch": 0.5784325787401575,
+      "grad_norm": 1.1415736675262451,
+      "learning_rate": 7.568732751113392e-05,
+      "loss": 2.1554,
+      "step": 9403
+    },
+    {
+      "epoch": 0.578494094488189,
+      "grad_norm": 1.306793212890625,
+      "learning_rate": 7.56685705627679e-05,
+      "loss": 2.7928,
+      "step": 9404
+    },
+    {
+      "epoch": 0.5785556102362205,
+      "grad_norm": 1.2712892293930054,
+      "learning_rate": 7.564981452426158e-05,
+      "loss": 2.5757,
+      "step": 9405
+    },
+    {
+      "epoch": 0.578617125984252,
+      "grad_norm": 1.0746692419052124,
+      "learning_rate": 7.563105939631649e-05,
+      "loss": 2.9677,
+      "step": 9406
+    },
+    {
+      "epoch": 0.5786786417322834,
+      "grad_norm": 1.0836094617843628,
+      "learning_rate": 7.561230517963384e-05,
+      "loss": 1.8323,
+      "step": 9407
+    },
+    {
+      "epoch": 0.5787401574803149,
+      "grad_norm": 1.423322319984436,
+      "learning_rate": 7.559355187491501e-05,
+      "loss": 3.2803,
+      "step": 9408
+    },
+    {
+      "epoch": 0.5788016732283464,
+      "grad_norm": 1.3719409704208374,
+      "learning_rate": 7.557479948286124e-05,
+      "loss": 2.2217,
+      "step": 9409
+    },
+    {
+      "epoch": 0.578863188976378,
+      "grad_norm": 1.2935377359390259,
+      "learning_rate": 7.555604800417379e-05,
+      "loss": 2.2811,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5789247047244095,
+      "grad_norm": 1.142168641090393,
+      "learning_rate": 7.553729743955381e-05,
+      "loss": 3.2503,
+      "step": 9411
+    },
+    {
+      "epoch": 0.578986220472441,
+      "grad_norm": 1.2838338613510132,
+      "learning_rate": 7.551854778970253e-05,
+      "loss": 2.8628,
+      "step": 9412
+    },
+    {
+      "epoch": 0.5790477362204725,
+      "grad_norm": 1.336432933807373,
+      "learning_rate": 7.549979905532102e-05,
+      "loss": 2.0052,
+      "step": 9413
+    },
+    {
+      "epoch": 0.5791092519685039,
+      "grad_norm": 1.1353296041488647,
+      "learning_rate": 7.548105123711042e-05,
+      "loss": 3.0713,
+      "step": 9414
+    },
+    {
+      "epoch": 0.5791707677165354,
+      "grad_norm": 1.3213386535644531,
+      "learning_rate": 7.546230433577177e-05,
+      "loss": 2.0359,
+      "step": 9415
+    },
+    {
+      "epoch": 0.5792322834645669,
+      "grad_norm": 1.1371912956237793,
+      "learning_rate": 7.544355835200617e-05,
+      "loss": 2.3327,
+      "step": 9416
+    },
+    {
+      "epoch": 0.5792937992125984,
+      "grad_norm": 1.3676575422286987,
+      "learning_rate": 7.54248132865145e-05,
+      "loss": 1.6847,
+      "step": 9417
+    },
+    {
+      "epoch": 0.57935531496063,
+      "grad_norm": 1.0769248008728027,
+      "learning_rate": 7.54060691399978e-05,
+      "loss": 2.4913,
+      "step": 9418
+    },
+    {
+      "epoch": 0.5794168307086615,
+      "grad_norm": 1.732871651649475,
+      "learning_rate": 7.538732591315694e-05,
+      "loss": 1.8495,
+      "step": 9419
+    },
+    {
+      "epoch": 0.5794783464566929,
+      "grad_norm": 1.1982194185256958,
+      "learning_rate": 7.53685836066929e-05,
+      "loss": 2.457,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5795398622047244,
+      "grad_norm": 1.2363871335983276,
+      "learning_rate": 7.534984222130646e-05,
+      "loss": 2.6165,
+      "step": 9421
+    },
+    {
+      "epoch": 0.5796013779527559,
+      "grad_norm": 1.3097918033599854,
+      "learning_rate": 7.533110175769846e-05,
+      "loss": 1.9245,
+      "step": 9422
+    },
+    {
+      "epoch": 0.5796628937007874,
+      "grad_norm": 1.152034878730774,
+      "learning_rate": 7.531236221656971e-05,
+      "loss": 2.7702,
+      "step": 9423
+    },
+    {
+      "epoch": 0.5797244094488189,
+      "grad_norm": 1.2753031253814697,
+      "learning_rate": 7.529362359862093e-05,
+      "loss": 2.5372,
+      "step": 9424
+    },
+    {
+      "epoch": 0.5797859251968503,
+      "grad_norm": 1.1776022911071777,
+      "learning_rate": 7.527488590455291e-05,
+      "loss": 3.5076,
+      "step": 9425
+    },
+    {
+      "epoch": 0.5798474409448819,
+      "grad_norm": 1.2680916786193848,
+      "learning_rate": 7.525614913506622e-05,
+      "loss": 2.285,
+      "step": 9426
+    },
+    {
+      "epoch": 0.5799089566929134,
+      "grad_norm": 1.3038790225982666,
+      "learning_rate": 7.523741329086163e-05,
+      "loss": 3.1444,
+      "step": 9427
+    },
+    {
+      "epoch": 0.5799704724409449,
+      "grad_norm": 1.0680415630340576,
+      "learning_rate": 7.521867837263966e-05,
+      "loss": 2.0183,
+      "step": 9428
+    },
+    {
+      "epoch": 0.5800319881889764,
+      "grad_norm": 1.1093069314956665,
+      "learning_rate": 7.519994438110099e-05,
+      "loss": 2.5941,
+      "step": 9429
+    },
+    {
+      "epoch": 0.5800935039370079,
+      "grad_norm": 1.1589809656143188,
+      "learning_rate": 7.518121131694605e-05,
+      "loss": 2.6579,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5801550196850394,
+      "grad_norm": 1.247496247291565,
+      "learning_rate": 7.516247918087545e-05,
+      "loss": 2.4742,
+      "step": 9431
+    },
+    {
+      "epoch": 0.5802165354330708,
+      "grad_norm": 1.0750221014022827,
+      "learning_rate": 7.514374797358961e-05,
+      "loss": 2.6259,
+      "step": 9432
+    },
+    {
+      "epoch": 0.5802780511811023,
+      "grad_norm": 1.1816842555999756,
+      "learning_rate": 7.512501769578902e-05,
+      "loss": 2.3439,
+      "step": 9433
+    },
+    {
+      "epoch": 0.5803395669291339,
+      "grad_norm": 1.3007900714874268,
+      "learning_rate": 7.510628834817401e-05,
+      "loss": 2.7623,
+      "step": 9434
+    },
+    {
+      "epoch": 0.5804010826771654,
+      "grad_norm": 1.2945692539215088,
+      "learning_rate": 7.508755993144508e-05,
+      "loss": 2.4217,
+      "step": 9435
+    },
+    {
+      "epoch": 0.5804625984251969,
+      "grad_norm": 1.3938874006271362,
+      "learning_rate": 7.506883244630244e-05,
+      "loss": 2.3794,
+      "step": 9436
+    },
+    {
+      "epoch": 0.5805241141732284,
+      "grad_norm": 1.387252926826477,
+      "learning_rate": 7.505010589344646e-05,
+      "loss": 3.4999,
+      "step": 9437
+    },
+    {
+      "epoch": 0.5805856299212598,
+      "grad_norm": 1.2773849964141846,
+      "learning_rate": 7.503138027357738e-05,
+      "loss": 2.0768,
+      "step": 9438
+    },
+    {
+      "epoch": 0.5806471456692913,
+      "grad_norm": 1.3712834119796753,
+      "learning_rate": 7.50126555873955e-05,
+      "loss": 2.2787,
+      "step": 9439
+    },
+    {
+      "epoch": 0.5807086614173228,
+      "grad_norm": 1.3196417093276978,
+      "learning_rate": 7.499393183560089e-05,
+      "loss": 2.6199,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5807701771653543,
+      "grad_norm": 1.6673611402511597,
+      "learning_rate": 7.497520901889386e-05,
+      "loss": 2.8314,
+      "step": 9441
+    },
+    {
+      "epoch": 0.5808316929133859,
+      "grad_norm": 2.131274700164795,
+      "learning_rate": 7.495648713797443e-05,
+      "loss": 1.9338,
+      "step": 9442
+    },
+    {
+      "epoch": 0.5808932086614174,
+      "grad_norm": 1.2376315593719482,
+      "learning_rate": 7.493776619354275e-05,
+      "loss": 2.4805,
+      "step": 9443
+    },
+    {
+      "epoch": 0.5809547244094488,
+      "grad_norm": 1.3028171062469482,
+      "learning_rate": 7.491904618629886e-05,
+      "loss": 2.7955,
+      "step": 9444
+    },
+    {
+      "epoch": 0.5810162401574803,
+      "grad_norm": 1.3285173177719116,
+      "learning_rate": 7.490032711694282e-05,
+      "loss": 2.5351,
+      "step": 9445
+    },
+    {
+      "epoch": 0.5810777559055118,
+      "grad_norm": 1.3121274709701538,
+      "learning_rate": 7.488160898617453e-05,
+      "loss": 2.0044,
+      "step": 9446
+    },
+    {
+      "epoch": 0.5811392716535433,
+      "grad_norm": 1.2963768243789673,
+      "learning_rate": 7.486289179469403e-05,
+      "loss": 2.843,
+      "step": 9447
+    },
+    {
+      "epoch": 0.5812007874015748,
+      "grad_norm": 1.328521490097046,
+      "learning_rate": 7.48441755432012e-05,
+      "loss": 2.6271,
+      "step": 9448
+    },
+    {
+      "epoch": 0.5812623031496063,
+      "grad_norm": 1.8903266191482544,
+      "learning_rate": 7.482546023239594e-05,
+      "loss": 1.7612,
+      "step": 9449
+    },
+    {
+      "epoch": 0.5813238188976378,
+      "grad_norm": 1.333668828010559,
+      "learning_rate": 7.480674586297807e-05,
+      "loss": 2.4109,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5813853346456693,
+      "grad_norm": 1.2022819519042969,
+      "learning_rate": 7.478803243564747e-05,
+      "loss": 2.1865,
+      "step": 9451
+    },
+    {
+      "epoch": 0.5814468503937008,
+      "grad_norm": 1.1251130104064941,
+      "learning_rate": 7.476931995110382e-05,
+      "loss": 2.4088,
+      "step": 9452
+    },
+    {
+      "epoch": 0.5815083661417323,
+      "grad_norm": 1.2902605533599854,
+      "learning_rate": 7.475060841004694e-05,
+      "loss": 2.8717,
+      "step": 9453
+    },
+    {
+      "epoch": 0.5815698818897638,
+      "grad_norm": 1.3226194381713867,
+      "learning_rate": 7.473189781317649e-05,
+      "loss": 2.7409,
+      "step": 9454
+    },
+    {
+      "epoch": 0.5816313976377953,
+      "grad_norm": 1.0819672346115112,
+      "learning_rate": 7.471318816119222e-05,
+      "loss": 2.7339,
+      "step": 9455
+    },
+    {
+      "epoch": 0.5816929133858267,
+      "grad_norm": 1.1876517534255981,
+      "learning_rate": 7.469447945479364e-05,
+      "loss": 2.3851,
+      "step": 9456
+    },
+    {
+      "epoch": 0.5817544291338582,
+      "grad_norm": 1.278283715248108,
+      "learning_rate": 7.467577169468045e-05,
+      "loss": 2.5152,
+      "step": 9457
+    },
+    {
+      "epoch": 0.5818159448818898,
+      "grad_norm": 1.067028284072876,
+      "learning_rate": 7.46570648815522e-05,
+      "loss": 2.8224,
+      "step": 9458
+    },
+    {
+      "epoch": 0.5818774606299213,
+      "grad_norm": 1.5735101699829102,
+      "learning_rate": 7.463835901610839e-05,
+      "loss": 2.4269,
+      "step": 9459
+    },
+    {
+      "epoch": 0.5819389763779528,
+      "grad_norm": 1.4029451608657837,
+      "learning_rate": 7.461965409904855e-05,
+      "loss": 1.9498,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5820004921259843,
+      "grad_norm": 1.179723858833313,
+      "learning_rate": 7.460095013107208e-05,
+      "loss": 3.0287,
+      "step": 9461
+    },
+    {
+      "epoch": 0.5820620078740157,
+      "grad_norm": 1.2248774766921997,
+      "learning_rate": 7.458224711287848e-05,
+      "loss": 2.9893,
+      "step": 9462
+    },
+    {
+      "epoch": 0.5821235236220472,
+      "grad_norm": 1.2670255899429321,
+      "learning_rate": 7.456354504516709e-05,
+      "loss": 2.8339,
+      "step": 9463
+    },
+    {
+      "epoch": 0.5821850393700787,
+      "grad_norm": 1.235329508781433,
+      "learning_rate": 7.454484392863732e-05,
+      "loss": 2.2683,
+      "step": 9464
+    },
+    {
+      "epoch": 0.5822465551181102,
+      "grad_norm": 1.1908749341964722,
+      "learning_rate": 7.452614376398841e-05,
+      "loss": 2.3737,
+      "step": 9465
+    },
+    {
+      "epoch": 0.5823080708661418,
+      "grad_norm": 1.50821053981781,
+      "learning_rate": 7.450744455191967e-05,
+      "loss": 2.7062,
+      "step": 9466
+    },
+    {
+      "epoch": 0.5823695866141733,
+      "grad_norm": 1.0632244348526,
+      "learning_rate": 7.448874629313036e-05,
+      "loss": 1.8627,
+      "step": 9467
+    },
+    {
+      "epoch": 0.5824311023622047,
+      "grad_norm": 1.3669689893722534,
+      "learning_rate": 7.447004898831969e-05,
+      "loss": 3.1486,
+      "step": 9468
+    },
+    {
+      "epoch": 0.5824926181102362,
+      "grad_norm": 1.1890381574630737,
+      "learning_rate": 7.445135263818681e-05,
+      "loss": 2.7083,
+      "step": 9469
+    },
+    {
+      "epoch": 0.5825541338582677,
+      "grad_norm": 1.2448678016662598,
+      "learning_rate": 7.443265724343091e-05,
+      "loss": 2.7476,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5826156496062992,
+      "grad_norm": 1.0791367292404175,
+      "learning_rate": 7.441396280475104e-05,
+      "loss": 2.6703,
+      "step": 9471
+    },
+    {
+      "epoch": 0.5826771653543307,
+      "grad_norm": 1.333409070968628,
+      "learning_rate": 7.43952693228463e-05,
+      "loss": 1.9184,
+      "step": 9472
+    },
+    {
+      "epoch": 0.5827386811023622,
+      "grad_norm": 1.4348021745681763,
+      "learning_rate": 7.437657679841569e-05,
+      "loss": 2.4417,
+      "step": 9473
+    },
+    {
+      "epoch": 0.5828001968503937,
+      "grad_norm": 1.0836724042892456,
+      "learning_rate": 7.435788523215828e-05,
+      "loss": 2.9073,
+      "step": 9474
+    },
+    {
+      "epoch": 0.5828617125984252,
+      "grad_norm": 1.241835117340088,
+      "learning_rate": 7.43391946247729e-05,
+      "loss": 2.1675,
+      "step": 9475
+    },
+    {
+      "epoch": 0.5829232283464567,
+      "grad_norm": 1.1651225090026855,
+      "learning_rate": 7.432050497695861e-05,
+      "loss": 2.7289,
+      "step": 9476
+    },
+    {
+      "epoch": 0.5829847440944882,
+      "grad_norm": 1.1361092329025269,
+      "learning_rate": 7.430181628941422e-05,
+      "loss": 3.0608,
+      "step": 9477
+    },
+    {
+      "epoch": 0.5830462598425197,
+      "grad_norm": 1.1721380949020386,
+      "learning_rate": 7.42831285628386e-05,
+      "loss": 2.8338,
+      "step": 9478
+    },
+    {
+      "epoch": 0.5831077755905512,
+      "grad_norm": 1.284976601600647,
+      "learning_rate": 7.426444179793057e-05,
+      "loss": 2.5324,
+      "step": 9479
+    },
+    {
+      "epoch": 0.5831692913385826,
+      "grad_norm": 1.199906349182129,
+      "learning_rate": 7.424575599538895e-05,
+      "loss": 2.6176,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5832308070866141,
+      "grad_norm": 1.4821046590805054,
+      "learning_rate": 7.422707115591241e-05,
+      "loss": 2.4452,
+      "step": 9481
+    },
+    {
+      "epoch": 0.5832923228346457,
+      "grad_norm": 1.2434676885604858,
+      "learning_rate": 7.42083872801997e-05,
+      "loss": 3.1298,
+      "step": 9482
+    },
+    {
+      "epoch": 0.5833538385826772,
+      "grad_norm": 8.83683967590332,
+      "learning_rate": 7.418970436894947e-05,
+      "loss": 2.0493,
+      "step": 9483
+    },
+    {
+      "epoch": 0.5834153543307087,
+      "grad_norm": 1.259078860282898,
+      "learning_rate": 7.417102242286042e-05,
+      "loss": 2.5141,
+      "step": 9484
+    },
+    {
+      "epoch": 0.5834768700787402,
+      "grad_norm": 1.6380629539489746,
+      "learning_rate": 7.415234144263106e-05,
+      "loss": 3.122,
+      "step": 9485
+    },
+    {
+      "epoch": 0.5835383858267716,
+      "grad_norm": 1.565039038658142,
+      "learning_rate": 7.413366142896003e-05,
+      "loss": 2.2962,
+      "step": 9486
+    },
+    {
+      "epoch": 0.5835999015748031,
+      "grad_norm": 1.27558434009552,
+      "learning_rate": 7.411498238254581e-05,
+      "loss": 2.9941,
+      "step": 9487
+    },
+    {
+      "epoch": 0.5836614173228346,
+      "grad_norm": 1.164050817489624,
+      "learning_rate": 7.409630430408692e-05,
+      "loss": 2.6597,
+      "step": 9488
+    },
+    {
+      "epoch": 0.5837229330708661,
+      "grad_norm": 1.4529438018798828,
+      "learning_rate": 7.40776271942818e-05,
+      "loss": 2.6229,
+      "step": 9489
+    },
+    {
+      "epoch": 0.5837844488188977,
+      "grad_norm": 1.8141453266143799,
+      "learning_rate": 7.40589510538289e-05,
+      "loss": 1.8766,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5838459645669292,
+      "grad_norm": 2.139641523361206,
+      "learning_rate": 7.404027588342655e-05,
+      "loss": 1.8936,
+      "step": 9491
+    },
+    {
+      "epoch": 0.5839074803149606,
+      "grad_norm": 1.3108205795288086,
+      "learning_rate": 7.402160168377313e-05,
+      "loss": 2.0452,
+      "step": 9492
+    },
+    {
+      "epoch": 0.5839689960629921,
+      "grad_norm": 1.2794313430786133,
+      "learning_rate": 7.400292845556697e-05,
+      "loss": 2.1318,
+      "step": 9493
+    },
+    {
+      "epoch": 0.5840305118110236,
+      "grad_norm": 1.2026152610778809,
+      "learning_rate": 7.398425619950629e-05,
+      "loss": 2.9334,
+      "step": 9494
+    },
+    {
+      "epoch": 0.5840920275590551,
+      "grad_norm": 1.1399273872375488,
+      "learning_rate": 7.39655849162894e-05,
+      "loss": 2.7342,
+      "step": 9495
+    },
+    {
+      "epoch": 0.5841535433070866,
+      "grad_norm": 1.1216129064559937,
+      "learning_rate": 7.394691460661442e-05,
+      "loss": 3.5655,
+      "step": 9496
+    },
+    {
+      "epoch": 0.5842150590551181,
+      "grad_norm": 1.357131838798523,
+      "learning_rate": 7.392824527117957e-05,
+      "loss": 2.203,
+      "step": 9497
+    },
+    {
+      "epoch": 0.5842765748031497,
+      "grad_norm": 1.6427232027053833,
+      "learning_rate": 7.390957691068296e-05,
+      "loss": 2.2476,
+      "step": 9498
+    },
+    {
+      "epoch": 0.5843380905511811,
+      "grad_norm": 1.2616726160049438,
+      "learning_rate": 7.389090952582272e-05,
+      "loss": 2.3561,
+      "step": 9499
+    },
+    {
+      "epoch": 0.5843996062992126,
+      "grad_norm": 1.0582351684570312,
+      "learning_rate": 7.387224311729684e-05,
+      "loss": 3.0872,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5844611220472441,
+      "grad_norm": 1.258435606956482,
+      "learning_rate": 7.385357768580341e-05,
+      "loss": 2.7857,
+      "step": 9501
+    },
+    {
+      "epoch": 0.5845226377952756,
+      "grad_norm": 1.3019524812698364,
+      "learning_rate": 7.383491323204035e-05,
+      "loss": 2.3283,
+      "step": 9502
+    },
+    {
+      "epoch": 0.5845841535433071,
+      "grad_norm": 1.2489678859710693,
+      "learning_rate": 7.381624975670565e-05,
+      "loss": 2.6399,
+      "step": 9503
+    },
+    {
+      "epoch": 0.5846456692913385,
+      "grad_norm": 1.1099514961242676,
+      "learning_rate": 7.379758726049717e-05,
+      "loss": 2.7177,
+      "step": 9504
+    },
+    {
+      "epoch": 0.58470718503937,
+      "grad_norm": 1.313863754272461,
+      "learning_rate": 7.377892574411288e-05,
+      "loss": 2.6103,
+      "step": 9505
+    },
+    {
+      "epoch": 0.5847687007874016,
+      "grad_norm": 1.3952645063400269,
+      "learning_rate": 7.37602652082505e-05,
+      "loss": 2.7123,
+      "step": 9506
+    },
+    {
+      "epoch": 0.5848302165354331,
+      "grad_norm": 1.2318283319473267,
+      "learning_rate": 7.374160565360791e-05,
+      "loss": 2.1977,
+      "step": 9507
+    },
+    {
+      "epoch": 0.5848917322834646,
+      "grad_norm": 1.2368923425674438,
+      "learning_rate": 7.372294708088283e-05,
+      "loss": 2.717,
+      "step": 9508
+    },
+    {
+      "epoch": 0.5849532480314961,
+      "grad_norm": 1.1926039457321167,
+      "learning_rate": 7.370428949077304e-05,
+      "loss": 2.5781,
+      "step": 9509
+    },
+    {
+      "epoch": 0.5850147637795275,
+      "grad_norm": 1.0960572957992554,
+      "learning_rate": 7.368563288397617e-05,
+      "loss": 2.7893,
+      "step": 9510
+    },
+    {
+      "epoch": 0.585076279527559,
+      "grad_norm": 1.3151315450668335,
+      "learning_rate": 7.36669772611899e-05,
+      "loss": 2.37,
+      "step": 9511
+    },
+    {
+      "epoch": 0.5851377952755905,
+      "grad_norm": 1.0522948503494263,
+      "learning_rate": 7.364832262311182e-05,
+      "loss": 3.1955,
+      "step": 9512
+    },
+    {
+      "epoch": 0.585199311023622,
+      "grad_norm": 1.1252800226211548,
+      "learning_rate": 7.362966897043957e-05,
+      "loss": 2.6706,
+      "step": 9513
+    },
+    {
+      "epoch": 0.5852608267716536,
+      "grad_norm": 1.26697838306427,
+      "learning_rate": 7.361101630387062e-05,
+      "loss": 2.4986,
+      "step": 9514
+    },
+    {
+      "epoch": 0.5853223425196851,
+      "grad_norm": 1.3215481042861938,
+      "learning_rate": 7.359236462410255e-05,
+      "loss": 2.3134,
+      "step": 9515
+    },
+    {
+      "epoch": 0.5853838582677166,
+      "grad_norm": 1.224061369895935,
+      "learning_rate": 7.357371393183277e-05,
+      "loss": 2.0923,
+      "step": 9516
+    },
+    {
+      "epoch": 0.585445374015748,
+      "grad_norm": 1.2115113735198975,
+      "learning_rate": 7.355506422775873e-05,
+      "loss": 2.8589,
+      "step": 9517
+    },
+    {
+      "epoch": 0.5855068897637795,
+      "grad_norm": 1.1043306589126587,
+      "learning_rate": 7.353641551257782e-05,
+      "loss": 3.0551,
+      "step": 9518
+    },
+    {
+      "epoch": 0.585568405511811,
+      "grad_norm": 1.2123963832855225,
+      "learning_rate": 7.351776778698745e-05,
+      "loss": 2.1328,
+      "step": 9519
+    },
+    {
+      "epoch": 0.5856299212598425,
+      "grad_norm": 1.0567623376846313,
+      "learning_rate": 7.349912105168481e-05,
+      "loss": 2.7912,
+      "step": 9520
+    },
+    {
+      "epoch": 0.585691437007874,
+      "grad_norm": 1.2979331016540527,
+      "learning_rate": 7.348047530736735e-05,
+      "loss": 2.8988,
+      "step": 9521
+    },
+    {
+      "epoch": 0.5857529527559056,
+      "grad_norm": 1.1528609991073608,
+      "learning_rate": 7.346183055473219e-05,
+      "loss": 2.9965,
+      "step": 9522
+    },
+    {
+      "epoch": 0.585814468503937,
+      "grad_norm": 1.3412275314331055,
+      "learning_rate": 7.344318679447661e-05,
+      "loss": 2.6964,
+      "step": 9523
+    },
+    {
+      "epoch": 0.5858759842519685,
+      "grad_norm": 1.1575555801391602,
+      "learning_rate": 7.342454402729775e-05,
+      "loss": 2.4179,
+      "step": 9524
+    },
+    {
+      "epoch": 0.5859375,
+      "grad_norm": 1.8113913536071777,
+      "learning_rate": 7.340590225389278e-05,
+      "loss": 2.9698,
+      "step": 9525
+    },
+    {
+      "epoch": 0.5859990157480315,
+      "grad_norm": 1.354473352432251,
+      "learning_rate": 7.338726147495873e-05,
+      "loss": 2.9346,
+      "step": 9526
+    },
+    {
+      "epoch": 0.586060531496063,
+      "grad_norm": 1.182441234588623,
+      "learning_rate": 7.336862169119274e-05,
+      "loss": 3.0774,
+      "step": 9527
+    },
+    {
+      "epoch": 0.5861220472440944,
+      "grad_norm": 1.298122763633728,
+      "learning_rate": 7.334998290329176e-05,
+      "loss": 1.9639,
+      "step": 9528
+    },
+    {
+      "epoch": 0.586183562992126,
+      "grad_norm": 1.143425464630127,
+      "learning_rate": 7.333134511195284e-05,
+      "loss": 2.4464,
+      "step": 9529
+    },
+    {
+      "epoch": 0.5862450787401575,
+      "grad_norm": 1.1556569337844849,
+      "learning_rate": 7.331270831787293e-05,
+      "loss": 2.8549,
+      "step": 9530
+    },
+    {
+      "epoch": 0.586306594488189,
+      "grad_norm": 1.2853516340255737,
+      "learning_rate": 7.329407252174887e-05,
+      "loss": 2.8422,
+      "step": 9531
+    },
+    {
+      "epoch": 0.5863681102362205,
+      "grad_norm": 1.3608694076538086,
+      "learning_rate": 7.32754377242776e-05,
+      "loss": 3.0478,
+      "step": 9532
+    },
+    {
+      "epoch": 0.586429625984252,
+      "grad_norm": 1.1551111936569214,
+      "learning_rate": 7.325680392615594e-05,
+      "loss": 3.0922,
+      "step": 9533
+    },
+    {
+      "epoch": 0.5864911417322834,
+      "grad_norm": 1.2141501903533936,
+      "learning_rate": 7.32381711280807e-05,
+      "loss": 2.3264,
+      "step": 9534
+    },
+    {
+      "epoch": 0.5865526574803149,
+      "grad_norm": 1.3426342010498047,
+      "learning_rate": 7.32195393307486e-05,
+      "loss": 2.403,
+      "step": 9535
+    },
+    {
+      "epoch": 0.5866141732283464,
+      "grad_norm": 1.2422668933868408,
+      "learning_rate": 7.320090853485644e-05,
+      "loss": 3.8036,
+      "step": 9536
+    },
+    {
+      "epoch": 0.586675688976378,
+      "grad_norm": 1.3138175010681152,
+      "learning_rate": 7.318227874110082e-05,
+      "loss": 2.3948,
+      "step": 9537
+    },
+    {
+      "epoch": 0.5867372047244095,
+      "grad_norm": 1.2485607862472534,
+      "learning_rate": 7.316364995017845e-05,
+      "loss": 2.3371,
+      "step": 9538
+    },
+    {
+      "epoch": 0.586798720472441,
+      "grad_norm": 1.2640172243118286,
+      "learning_rate": 7.314502216278593e-05,
+      "loss": 2.8464,
+      "step": 9539
+    },
+    {
+      "epoch": 0.5868602362204725,
+      "grad_norm": 1.2967250347137451,
+      "learning_rate": 7.312639537961988e-05,
+      "loss": 2.8639,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5869217519685039,
+      "grad_norm": 1.1572000980377197,
+      "learning_rate": 7.310776960137673e-05,
+      "loss": 2.5863,
+      "step": 9541
+    },
+    {
+      "epoch": 0.5869832677165354,
+      "grad_norm": 1.3907947540283203,
+      "learning_rate": 7.308914482875307e-05,
+      "loss": 2.6541,
+      "step": 9542
+    },
+    {
+      "epoch": 0.5870447834645669,
+      "grad_norm": 1.2465312480926514,
+      "learning_rate": 7.30705210624453e-05,
+      "loss": 2.3656,
+      "step": 9543
+    },
+    {
+      "epoch": 0.5871062992125984,
+      "grad_norm": 1.2768654823303223,
+      "learning_rate": 7.305189830314994e-05,
+      "loss": 2.1271,
+      "step": 9544
+    },
+    {
+      "epoch": 0.58716781496063,
+      "grad_norm": 1.3233426809310913,
+      "learning_rate": 7.303327655156327e-05,
+      "loss": 2.2507,
+      "step": 9545
+    },
+    {
+      "epoch": 0.5872293307086615,
+      "grad_norm": 1.358342170715332,
+      "learning_rate": 7.30146558083817e-05,
+      "loss": 2.2441,
+      "step": 9546
+    },
+    {
+      "epoch": 0.5872908464566929,
+      "grad_norm": 1.0867445468902588,
+      "learning_rate": 7.299603607430151e-05,
+      "loss": 3.2137,
+      "step": 9547
+    },
+    {
+      "epoch": 0.5873523622047244,
+      "grad_norm": 1.2551008462905884,
+      "learning_rate": 7.297741735001902e-05,
+      "loss": 2.6538,
+      "step": 9548
+    },
+    {
+      "epoch": 0.5874138779527559,
+      "grad_norm": 1.2032204866409302,
+      "learning_rate": 7.29587996362304e-05,
+      "loss": 2.5511,
+      "step": 9549
+    },
+    {
+      "epoch": 0.5874753937007874,
+      "grad_norm": 1.4030629396438599,
+      "learning_rate": 7.294018293363195e-05,
+      "loss": 1.9014,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5875369094488189,
+      "grad_norm": 1.0208221673965454,
+      "learning_rate": 7.292156724291972e-05,
+      "loss": 2.7766,
+      "step": 9551
+    },
+    {
+      "epoch": 0.5875984251968503,
+      "grad_norm": 1.2399163246154785,
+      "learning_rate": 7.29029525647899e-05,
+      "loss": 2.3948,
+      "step": 9552
+    },
+    {
+      "epoch": 0.5876599409448819,
+      "grad_norm": 1.2214555740356445,
+      "learning_rate": 7.288433889993854e-05,
+      "loss": 2.9189,
+      "step": 9553
+    },
+    {
+      "epoch": 0.5877214566929134,
+      "grad_norm": 1.4506546258926392,
+      "learning_rate": 7.286572624906174e-05,
+      "loss": 2.1699,
+      "step": 9554
+    },
+    {
+      "epoch": 0.5877829724409449,
+      "grad_norm": 1.1982344388961792,
+      "learning_rate": 7.284711461285542e-05,
+      "loss": 2.2841,
+      "step": 9555
+    },
+    {
+      "epoch": 0.5878444881889764,
+      "grad_norm": 1.0833228826522827,
+      "learning_rate": 7.282850399201566e-05,
+      "loss": 2.7456,
+      "step": 9556
+    },
+    {
+      "epoch": 0.5879060039370079,
+      "grad_norm": 1.0922080278396606,
+      "learning_rate": 7.28098943872383e-05,
+      "loss": 2.5978,
+      "step": 9557
+    },
+    {
+      "epoch": 0.5879675196850394,
+      "grad_norm": 1.157679557800293,
+      "learning_rate": 7.27912857992193e-05,
+      "loss": 2.4749,
+      "step": 9558
+    },
+    {
+      "epoch": 0.5880290354330708,
+      "grad_norm": 1.2058453559875488,
+      "learning_rate": 7.277267822865446e-05,
+      "loss": 2.0552,
+      "step": 9559
+    },
+    {
+      "epoch": 0.5880905511811023,
+      "grad_norm": 1.0337985754013062,
+      "learning_rate": 7.275407167623968e-05,
+      "loss": 2.4378,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5881520669291339,
+      "grad_norm": 1.2371912002563477,
+      "learning_rate": 7.273546614267063e-05,
+      "loss": 1.8899,
+      "step": 9561
+    },
+    {
+      "epoch": 0.5882135826771654,
+      "grad_norm": 1.1419559717178345,
+      "learning_rate": 7.271686162864318e-05,
+      "loss": 2.6472,
+      "step": 9562
+    },
+    {
+      "epoch": 0.5882750984251969,
+      "grad_norm": 1.2071956396102905,
+      "learning_rate": 7.269825813485292e-05,
+      "loss": 2.195,
+      "step": 9563
+    },
+    {
+      "epoch": 0.5883366141732284,
+      "grad_norm": 1.2703624963760376,
+      "learning_rate": 7.267965566199558e-05,
+      "loss": 2.8851,
+      "step": 9564
+    },
+    {
+      "epoch": 0.5883981299212598,
+      "grad_norm": 1.8477718830108643,
+      "learning_rate": 7.266105421076676e-05,
+      "loss": 1.9904,
+      "step": 9565
+    },
+    {
+      "epoch": 0.5884596456692913,
+      "grad_norm": 1.6508145332336426,
+      "learning_rate": 7.264245378186212e-05,
+      "loss": 2.6468,
+      "step": 9566
+    },
+    {
+      "epoch": 0.5885211614173228,
+      "grad_norm": 1.2575228214263916,
+      "learning_rate": 7.262385437597712e-05,
+      "loss": 2.3937,
+      "step": 9567
+    },
+    {
+      "epoch": 0.5885826771653543,
+      "grad_norm": 1.350451111793518,
+      "learning_rate": 7.26052559938073e-05,
+      "loss": 2.6089,
+      "step": 9568
+    },
+    {
+      "epoch": 0.5886441929133859,
+      "grad_norm": 1.6133369207382202,
+      "learning_rate": 7.258665863604819e-05,
+      "loss": 2.4711,
+      "step": 9569
+    },
+    {
+      "epoch": 0.5887057086614174,
+      "grad_norm": 1.1698945760726929,
+      "learning_rate": 7.256806230339515e-05,
+      "loss": 2.4009,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5887672244094488,
+      "grad_norm": 1.289487361907959,
+      "learning_rate": 7.254946699654363e-05,
+      "loss": 2.3479,
+      "step": 9571
+    },
+    {
+      "epoch": 0.5888287401574803,
+      "grad_norm": 1.1837997436523438,
+      "learning_rate": 7.253087271618897e-05,
+      "loss": 2.7427,
+      "step": 9572
+    },
+    {
+      "epoch": 0.5888902559055118,
+      "grad_norm": 1.4038089513778687,
+      "learning_rate": 7.25122794630265e-05,
+      "loss": 2.5598,
+      "step": 9573
+    },
+    {
+      "epoch": 0.5889517716535433,
+      "grad_norm": 1.11372971534729,
+      "learning_rate": 7.24936872377515e-05,
+      "loss": 3.2793,
+      "step": 9574
+    },
+    {
+      "epoch": 0.5890132874015748,
+      "grad_norm": 1.264888048171997,
+      "learning_rate": 7.247509604105925e-05,
+      "loss": 2.6101,
+      "step": 9575
+    },
+    {
+      "epoch": 0.5890748031496063,
+      "grad_norm": 1.12351393699646,
+      "learning_rate": 7.245650587364488e-05,
+      "loss": 2.0584,
+      "step": 9576
+    },
+    {
+      "epoch": 0.5891363188976378,
+      "grad_norm": 1.3896863460540771,
+      "learning_rate": 7.243791673620362e-05,
+      "loss": 2.8602,
+      "step": 9577
+    },
+    {
+      "epoch": 0.5891978346456693,
+      "grad_norm": 1.5057560205459595,
+      "learning_rate": 7.241932862943057e-05,
+      "loss": 2.9783,
+      "step": 9578
+    },
+    {
+      "epoch": 0.5892593503937008,
+      "grad_norm": 1.2729685306549072,
+      "learning_rate": 7.240074155402088e-05,
+      "loss": 3.0315,
+      "step": 9579
+    },
+    {
+      "epoch": 0.5893208661417323,
+      "grad_norm": 1.1346815824508667,
+      "learning_rate": 7.238215551066951e-05,
+      "loss": 2.7229,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5893823818897638,
+      "grad_norm": 1.1457980871200562,
+      "learning_rate": 7.236357050007154e-05,
+      "loss": 2.8525,
+      "step": 9581
+    },
+    {
+      "epoch": 0.5894438976377953,
+      "grad_norm": 1.2143340110778809,
+      "learning_rate": 7.234498652292192e-05,
+      "loss": 2.5292,
+      "step": 9582
+    },
+    {
+      "epoch": 0.5895054133858267,
+      "grad_norm": 1.1487284898757935,
+      "learning_rate": 7.23264035799156e-05,
+      "loss": 2.4008,
+      "step": 9583
+    },
+    {
+      "epoch": 0.5895669291338582,
+      "grad_norm": 1.2658342123031616,
+      "learning_rate": 7.230782167174747e-05,
+      "loss": 2.3409,
+      "step": 9584
+    },
+    {
+      "epoch": 0.5896284448818898,
+      "grad_norm": 1.1222198009490967,
+      "learning_rate": 7.228924079911242e-05,
+      "loss": 3.1253,
+      "step": 9585
+    },
+    {
+      "epoch": 0.5896899606299213,
+      "grad_norm": 1.1487822532653809,
+      "learning_rate": 7.227066096270522e-05,
+      "loss": 2.0661,
+      "step": 9586
+    },
+    {
+      "epoch": 0.5897514763779528,
+      "grad_norm": 1.0486475229263306,
+      "learning_rate": 7.225208216322068e-05,
+      "loss": 2.8696,
+      "step": 9587
+    },
+    {
+      "epoch": 0.5898129921259843,
+      "grad_norm": 1.0871034860610962,
+      "learning_rate": 7.223350440135353e-05,
+      "loss": 2.5216,
+      "step": 9588
+    },
+    {
+      "epoch": 0.5898745078740157,
+      "grad_norm": 1.4998164176940918,
+      "learning_rate": 7.221492767779852e-05,
+      "loss": 2.3445,
+      "step": 9589
+    },
+    {
+      "epoch": 0.5899360236220472,
+      "grad_norm": 1.168900966644287,
+      "learning_rate": 7.219635199325023e-05,
+      "loss": 3.1308,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5899975393700787,
+      "grad_norm": 2.1708385944366455,
+      "learning_rate": 7.21777773484034e-05,
+      "loss": 1.9767,
+      "step": 9591
+    },
+    {
+      "epoch": 0.5900590551181102,
+      "grad_norm": 1.3558375835418701,
+      "learning_rate": 7.215920374395251e-05,
+      "loss": 2.908,
+      "step": 9592
+    },
+    {
+      "epoch": 0.5901205708661418,
+      "grad_norm": 1.3803366422653198,
+      "learning_rate": 7.214063118059216e-05,
+      "loss": 2.5975,
+      "step": 9593
+    },
+    {
+      "epoch": 0.5901820866141733,
+      "grad_norm": 1.6738849878311157,
+      "learning_rate": 7.212205965901686e-05,
+      "loss": 2.6399,
+      "step": 9594
+    },
+    {
+      "epoch": 0.5902436023622047,
+      "grad_norm": 1.07210373878479,
+      "learning_rate": 7.210348917992111e-05,
+      "loss": 2.2755,
+      "step": 9595
+    },
+    {
+      "epoch": 0.5903051181102362,
+      "grad_norm": 1.2866789102554321,
+      "learning_rate": 7.208491974399927e-05,
+      "loss": 2.9913,
+      "step": 9596
+    },
+    {
+      "epoch": 0.5903666338582677,
+      "grad_norm": 1.8100920915603638,
+      "learning_rate": 7.206635135194579e-05,
+      "loss": 1.1165,
+      "step": 9597
+    },
+    {
+      "epoch": 0.5904281496062992,
+      "grad_norm": 1.0568140745162964,
+      "learning_rate": 7.204778400445498e-05,
+      "loss": 2.8263,
+      "step": 9598
+    },
+    {
+      "epoch": 0.5904896653543307,
+      "grad_norm": 1.1272863149642944,
+      "learning_rate": 7.202921770222121e-05,
+      "loss": 2.4461,
+      "step": 9599
+    },
+    {
+      "epoch": 0.5905511811023622,
+      "grad_norm": 1.2776812314987183,
+      "learning_rate": 7.20106524459387e-05,
+      "loss": 3.1814,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5906126968503937,
+      "grad_norm": 1.1858736276626587,
+      "learning_rate": 7.199208823630177e-05,
+      "loss": 2.241,
+      "step": 9601
+    },
+    {
+      "epoch": 0.5906742125984252,
+      "grad_norm": 1.5619585514068604,
+      "learning_rate": 7.197352507400452e-05,
+      "loss": 2.6011,
+      "step": 9602
+    },
+    {
+      "epoch": 0.5907357283464567,
+      "grad_norm": 1.3898437023162842,
+      "learning_rate": 7.195496295974117e-05,
+      "loss": 3.1819,
+      "step": 9603
+    },
+    {
+      "epoch": 0.5907972440944882,
+      "grad_norm": 1.1402697563171387,
+      "learning_rate": 7.193640189420585e-05,
+      "loss": 2.2959,
+      "step": 9604
+    },
+    {
+      "epoch": 0.5908587598425197,
+      "grad_norm": 1.2436654567718506,
+      "learning_rate": 7.191784187809256e-05,
+      "loss": 2.7142,
+      "step": 9605
+    },
+    {
+      "epoch": 0.5909202755905512,
+      "grad_norm": 1.0439344644546509,
+      "learning_rate": 7.189928291209541e-05,
+      "loss": 2.6005,
+      "step": 9606
+    },
+    {
+      "epoch": 0.5909817913385826,
+      "grad_norm": 1.1607829332351685,
+      "learning_rate": 7.188072499690837e-05,
+      "loss": 2.8736,
+      "step": 9607
+    },
+    {
+      "epoch": 0.5910433070866141,
+      "grad_norm": 1.3906867504119873,
+      "learning_rate": 7.186216813322547e-05,
+      "loss": 2.7675,
+      "step": 9608
+    },
+    {
+      "epoch": 0.5911048228346457,
+      "grad_norm": 1.3931610584259033,
+      "learning_rate": 7.184361232174051e-05,
+      "loss": 2.74,
+      "step": 9609
+    },
+    {
+      "epoch": 0.5911663385826772,
+      "grad_norm": 1.1953396797180176,
+      "learning_rate": 7.18250575631475e-05,
+      "loss": 2.8302,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5912278543307087,
+      "grad_norm": 1.3276770114898682,
+      "learning_rate": 7.18065038581402e-05,
+      "loss": 3.0935,
+      "step": 9611
+    },
+    {
+      "epoch": 0.5912893700787402,
+      "grad_norm": 1.5057029724121094,
+      "learning_rate": 7.178795120741246e-05,
+      "loss": 2.2074,
+      "step": 9612
+    },
+    {
+      "epoch": 0.5913508858267716,
+      "grad_norm": 1.0908234119415283,
+      "learning_rate": 7.176939961165801e-05,
+      "loss": 2.197,
+      "step": 9613
+    },
+    {
+      "epoch": 0.5914124015748031,
+      "grad_norm": 3.8892836570739746,
+      "learning_rate": 7.175084907157063e-05,
+      "loss": 3.7508,
+      "step": 9614
+    },
+    {
+      "epoch": 0.5914739173228346,
+      "grad_norm": 1.4228222370147705,
+      "learning_rate": 7.173229958784394e-05,
+      "loss": 3.5476,
+      "step": 9615
+    },
+    {
+      "epoch": 0.5915354330708661,
+      "grad_norm": 1.2084765434265137,
+      "learning_rate": 7.171375116117164e-05,
+      "loss": 3.051,
+      "step": 9616
+    },
+    {
+      "epoch": 0.5915969488188977,
+      "grad_norm": 1.881251573562622,
+      "learning_rate": 7.16952037922473e-05,
+      "loss": 2.2475,
+      "step": 9617
+    },
+    {
+      "epoch": 0.5916584645669292,
+      "grad_norm": 1.1060031652450562,
+      "learning_rate": 7.167665748176452e-05,
+      "loss": 2.8654,
+      "step": 9618
+    },
+    {
+      "epoch": 0.5917199803149606,
+      "grad_norm": 1.2688839435577393,
+      "learning_rate": 7.165811223041681e-05,
+      "loss": 2.4894,
+      "step": 9619
+    },
+    {
+      "epoch": 0.5917814960629921,
+      "grad_norm": 1.2343467473983765,
+      "learning_rate": 7.163956803889768e-05,
+      "loss": 1.8402,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5918430118110236,
+      "grad_norm": 1.286982536315918,
+      "learning_rate": 7.162102490790054e-05,
+      "loss": 2.6834,
+      "step": 9621
+    },
+    {
+      "epoch": 0.5919045275590551,
+      "grad_norm": 1.194109559059143,
+      "learning_rate": 7.160248283811884e-05,
+      "loss": 2.5117,
+      "step": 9622
+    },
+    {
+      "epoch": 0.5919660433070866,
+      "grad_norm": 1.270883321762085,
+      "learning_rate": 7.158394183024591e-05,
+      "loss": 2.5112,
+      "step": 9623
+    },
+    {
+      "epoch": 0.5920275590551181,
+      "grad_norm": 1.2839806079864502,
+      "learning_rate": 7.156540188497515e-05,
+      "loss": 2.7704,
+      "step": 9624
+    },
+    {
+      "epoch": 0.5920890748031497,
+      "grad_norm": 1.3871183395385742,
+      "learning_rate": 7.154686300299973e-05,
+      "loss": 2.3837,
+      "step": 9625
+    },
+    {
+      "epoch": 0.5921505905511811,
+      "grad_norm": 1.4812896251678467,
+      "learning_rate": 7.152832518501304e-05,
+      "loss": 2.768,
+      "step": 9626
+    },
+    {
+      "epoch": 0.5922121062992126,
+      "grad_norm": 1.163131594657898,
+      "learning_rate": 7.15097884317082e-05,
+      "loss": 2.5745,
+      "step": 9627
+    },
+    {
+      "epoch": 0.5922736220472441,
+      "grad_norm": 1.117163062095642,
+      "learning_rate": 7.149125274377842e-05,
+      "loss": 2.8962,
+      "step": 9628
+    },
+    {
+      "epoch": 0.5923351377952756,
+      "grad_norm": 1.3535606861114502,
+      "learning_rate": 7.147271812191679e-05,
+      "loss": 2.3537,
+      "step": 9629
+    },
+    {
+      "epoch": 0.5923966535433071,
+      "grad_norm": 1.318419098854065,
+      "learning_rate": 7.145418456681647e-05,
+      "loss": 2.4946,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5924581692913385,
+      "grad_norm": 1.037214994430542,
+      "learning_rate": 7.143565207917044e-05,
+      "loss": 3.2397,
+      "step": 9631
+    },
+    {
+      "epoch": 0.59251968503937,
+      "grad_norm": 0.9954841732978821,
+      "learning_rate": 7.141712065967177e-05,
+      "loss": 2.3008,
+      "step": 9632
+    },
+    {
+      "epoch": 0.5925812007874016,
+      "grad_norm": 1.0396883487701416,
+      "learning_rate": 7.139859030901336e-05,
+      "loss": 2.7228,
+      "step": 9633
+    },
+    {
+      "epoch": 0.5926427165354331,
+      "grad_norm": 1.2117633819580078,
+      "learning_rate": 7.138006102788824e-05,
+      "loss": 1.9148,
+      "step": 9634
+    },
+    {
+      "epoch": 0.5927042322834646,
+      "grad_norm": 1.1652575731277466,
+      "learning_rate": 7.136153281698919e-05,
+      "loss": 2.1017,
+      "step": 9635
+    },
+    {
+      "epoch": 0.5927657480314961,
+      "grad_norm": 1.5712169408798218,
+      "learning_rate": 7.134300567700918e-05,
+      "loss": 2.307,
+      "step": 9636
+    },
+    {
+      "epoch": 0.5928272637795275,
+      "grad_norm": 1.2511379718780518,
+      "learning_rate": 7.132447960864093e-05,
+      "loss": 2.3134,
+      "step": 9637
+    },
+    {
+      "epoch": 0.592888779527559,
+      "grad_norm": 1.332441806793213,
+      "learning_rate": 7.130595461257724e-05,
+      "loss": 2.4579,
+      "step": 9638
+    },
+    {
+      "epoch": 0.5929502952755905,
+      "grad_norm": 1.4579700231552124,
+      "learning_rate": 7.128743068951085e-05,
+      "loss": 2.8909,
+      "step": 9639
+    },
+    {
+      "epoch": 0.593011811023622,
+      "grad_norm": 1.1809604167938232,
+      "learning_rate": 7.126890784013447e-05,
+      "loss": 2.5445,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5930733267716536,
+      "grad_norm": 1.180201530456543,
+      "learning_rate": 7.125038606514072e-05,
+      "loss": 2.6102,
+      "step": 9641
+    },
+    {
+      "epoch": 0.5931348425196851,
+      "grad_norm": 1.4831254482269287,
+      "learning_rate": 7.123186536522218e-05,
+      "loss": 2.2319,
+      "step": 9642
+    },
+    {
+      "epoch": 0.5931963582677166,
+      "grad_norm": 1.2154369354248047,
+      "learning_rate": 7.121334574107153e-05,
+      "loss": 2.5281,
+      "step": 9643
+    },
+    {
+      "epoch": 0.593257874015748,
+      "grad_norm": 1.144901156425476,
+      "learning_rate": 7.119482719338114e-05,
+      "loss": 2.3119,
+      "step": 9644
+    },
+    {
+      "epoch": 0.5933193897637795,
+      "grad_norm": 1.1862964630126953,
+      "learning_rate": 7.117630972284366e-05,
+      "loss": 2.433,
+      "step": 9645
+    },
+    {
+      "epoch": 0.593380905511811,
+      "grad_norm": 1.2937769889831543,
+      "learning_rate": 7.115779333015145e-05,
+      "loss": 2.3215,
+      "step": 9646
+    },
+    {
+      "epoch": 0.5934424212598425,
+      "grad_norm": 1.2588516473770142,
+      "learning_rate": 7.113927801599694e-05,
+      "loss": 1.8771,
+      "step": 9647
+    },
+    {
+      "epoch": 0.593503937007874,
+      "grad_norm": 1.3341014385223389,
+      "learning_rate": 7.112076378107248e-05,
+      "loss": 2.8501,
+      "step": 9648
+    },
+    {
+      "epoch": 0.5935654527559056,
+      "grad_norm": 1.504839539527893,
+      "learning_rate": 7.110225062607046e-05,
+      "loss": 1.9887,
+      "step": 9649
+    },
+    {
+      "epoch": 0.593626968503937,
+      "grad_norm": 1.1849561929702759,
+      "learning_rate": 7.10837385516831e-05,
+      "loss": 2.1455,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5936884842519685,
+      "grad_norm": 1.3252397775650024,
+      "learning_rate": 7.10652275586027e-05,
+      "loss": 2.5352,
+      "step": 9651
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 1.3112633228302002,
+      "learning_rate": 7.10467176475214e-05,
+      "loss": 3.1582,
+      "step": 9652
+    },
+    {
+      "epoch": 0.5938115157480315,
+      "grad_norm": 1.2508816719055176,
+      "learning_rate": 7.102820881913146e-05,
+      "loss": 2.7599,
+      "step": 9653
+    },
+    {
+      "epoch": 0.593873031496063,
+      "grad_norm": 1.375533938407898,
+      "learning_rate": 7.100970107412493e-05,
+      "loss": 2.371,
+      "step": 9654
+    },
+    {
+      "epoch": 0.5939345472440944,
+      "grad_norm": 1.2248200178146362,
+      "learning_rate": 7.099119441319396e-05,
+      "loss": 1.8463,
+      "step": 9655
+    },
+    {
+      "epoch": 0.593996062992126,
+      "grad_norm": 1.1945736408233643,
+      "learning_rate": 7.097268883703052e-05,
+      "loss": 2.4573,
+      "step": 9656
+    },
+    {
+      "epoch": 0.5940575787401575,
+      "grad_norm": 1.2012856006622314,
+      "learning_rate": 7.095418434632668e-05,
+      "loss": 2.5878,
+      "step": 9657
+    },
+    {
+      "epoch": 0.594119094488189,
+      "grad_norm": 1.2135837078094482,
+      "learning_rate": 7.093568094177436e-05,
+      "loss": 2.0996,
+      "step": 9658
+    },
+    {
+      "epoch": 0.5941806102362205,
+      "grad_norm": 1.2312592267990112,
+      "learning_rate": 7.091717862406556e-05,
+      "loss": 2.7136,
+      "step": 9659
+    },
+    {
+      "epoch": 0.594242125984252,
+      "grad_norm": 1.2631800174713135,
+      "learning_rate": 7.089867739389205e-05,
+      "loss": 2.1784,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5943036417322834,
+      "grad_norm": 1.2338900566101074,
+      "learning_rate": 7.088017725194576e-05,
+      "loss": 2.9363,
+      "step": 9661
+    },
+    {
+      "epoch": 0.5943651574803149,
+      "grad_norm": 1.317040205001831,
+      "learning_rate": 7.086167819891842e-05,
+      "loss": 2.3651,
+      "step": 9662
+    },
+    {
+      "epoch": 0.5944266732283464,
+      "grad_norm": 1.3236840963363647,
+      "learning_rate": 7.08431802355019e-05,
+      "loss": 2.6636,
+      "step": 9663
+    },
+    {
+      "epoch": 0.594488188976378,
+      "grad_norm": 1.2847830057144165,
+      "learning_rate": 7.08246833623878e-05,
+      "loss": 2.9288,
+      "step": 9664
+    },
+    {
+      "epoch": 0.5945497047244095,
+      "grad_norm": 1.530174970626831,
+      "learning_rate": 7.080618758026791e-05,
+      "loss": 2.0714,
+      "step": 9665
+    },
+    {
+      "epoch": 0.594611220472441,
+      "grad_norm": 1.3241755962371826,
+      "learning_rate": 7.078769288983378e-05,
+      "loss": 2.3025,
+      "step": 9666
+    },
+    {
+      "epoch": 0.5946727362204725,
+      "grad_norm": 1.5671757459640503,
+      "learning_rate": 7.076919929177705e-05,
+      "loss": 2.2452,
+      "step": 9667
+    },
+    {
+      "epoch": 0.5947342519685039,
+      "grad_norm": 1.4040197134017944,
+      "learning_rate": 7.075070678678928e-05,
+      "loss": 2.1811,
+      "step": 9668
+    },
+    {
+      "epoch": 0.5947957677165354,
+      "grad_norm": 1.349133849143982,
+      "learning_rate": 7.073221537556199e-05,
+      "loss": 2.4419,
+      "step": 9669
+    },
+    {
+      "epoch": 0.5948572834645669,
+      "grad_norm": 1.1672110557556152,
+      "learning_rate": 7.07137250587866e-05,
+      "loss": 3.2323,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5949187992125984,
+      "grad_norm": 1.266252040863037,
+      "learning_rate": 7.069523583715463e-05,
+      "loss": 2.6759,
+      "step": 9671
+    },
+    {
+      "epoch": 0.59498031496063,
+      "grad_norm": 1.1715503931045532,
+      "learning_rate": 7.06767477113574e-05,
+      "loss": 2.8236,
+      "step": 9672
+    },
+    {
+      "epoch": 0.5950418307086615,
+      "grad_norm": 1.2521014213562012,
+      "learning_rate": 7.065826068208632e-05,
+      "loss": 2.7889,
+      "step": 9673
+    },
+    {
+      "epoch": 0.5951033464566929,
+      "grad_norm": 1.2800747156143188,
+      "learning_rate": 7.063977475003265e-05,
+      "loss": 2.1983,
+      "step": 9674
+    },
+    {
+      "epoch": 0.5951648622047244,
+      "grad_norm": 1.1723014116287231,
+      "learning_rate": 7.062128991588774e-05,
+      "loss": 2.8603,
+      "step": 9675
+    },
+    {
+      "epoch": 0.5952263779527559,
+      "grad_norm": 1.124780535697937,
+      "learning_rate": 7.060280618034272e-05,
+      "loss": 2.8667,
+      "step": 9676
+    },
+    {
+      "epoch": 0.5952878937007874,
+      "grad_norm": 1.3518683910369873,
+      "learning_rate": 7.058432354408883e-05,
+      "loss": 2.3847,
+      "step": 9677
+    },
+    {
+      "epoch": 0.5953494094488189,
+      "grad_norm": 1.1710835695266724,
+      "learning_rate": 7.056584200781726e-05,
+      "loss": 2.6398,
+      "step": 9678
+    },
+    {
+      "epoch": 0.5954109251968503,
+      "grad_norm": 1.4091523885726929,
+      "learning_rate": 7.054736157221898e-05,
+      "loss": 2.2346,
+      "step": 9679
+    },
+    {
+      "epoch": 0.5954724409448819,
+      "grad_norm": 1.2959779500961304,
+      "learning_rate": 7.052888223798522e-05,
+      "loss": 1.6816,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5955339566929134,
+      "grad_norm": 1.241396427154541,
+      "learning_rate": 7.051040400580689e-05,
+      "loss": 2.7893,
+      "step": 9681
+    },
+    {
+      "epoch": 0.5955954724409449,
+      "grad_norm": 1.4071707725524902,
+      "learning_rate": 7.049192687637504e-05,
+      "loss": 2.5366,
+      "step": 9682
+    },
+    {
+      "epoch": 0.5956569881889764,
+      "grad_norm": 1.1618720293045044,
+      "learning_rate": 7.047345085038055e-05,
+      "loss": 2.8515,
+      "step": 9683
+    },
+    {
+      "epoch": 0.5957185039370079,
+      "grad_norm": 1.4001322984695435,
+      "learning_rate": 7.04549759285144e-05,
+      "loss": 2.2501,
+      "step": 9684
+    },
+    {
+      "epoch": 0.5957800196850394,
+      "grad_norm": 1.392026424407959,
+      "learning_rate": 7.043650211146735e-05,
+      "loss": 2.2065,
+      "step": 9685
+    },
+    {
+      "epoch": 0.5958415354330708,
+      "grad_norm": 1.0913108587265015,
+      "learning_rate": 7.04180293999303e-05,
+      "loss": 2.6148,
+      "step": 9686
+    },
+    {
+      "epoch": 0.5959030511811023,
+      "grad_norm": 1.4895914793014526,
+      "learning_rate": 7.039955779459399e-05,
+      "loss": 2.4539,
+      "step": 9687
+    },
+    {
+      "epoch": 0.5959645669291339,
+      "grad_norm": 1.3191239833831787,
+      "learning_rate": 7.038108729614917e-05,
+      "loss": 2.194,
+      "step": 9688
+    },
+    {
+      "epoch": 0.5960260826771654,
+      "grad_norm": 1.3489962816238403,
+      "learning_rate": 7.03626179052865e-05,
+      "loss": 2.1516,
+      "step": 9689
+    },
+    {
+      "epoch": 0.5960875984251969,
+      "grad_norm": 1.7849889993667603,
+      "learning_rate": 7.034414962269671e-05,
+      "loss": 2.1009,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5961491141732284,
+      "grad_norm": 1.3596599102020264,
+      "learning_rate": 7.03256824490703e-05,
+      "loss": 2.2928,
+      "step": 9691
+    },
+    {
+      "epoch": 0.5962106299212598,
+      "grad_norm": 1.2029955387115479,
+      "learning_rate": 7.030721638509794e-05,
+      "loss": 3.0379,
+      "step": 9692
+    },
+    {
+      "epoch": 0.5962721456692913,
+      "grad_norm": 1.348219633102417,
+      "learning_rate": 7.028875143147008e-05,
+      "loss": 2.6593,
+      "step": 9693
+    },
+    {
+      "epoch": 0.5963336614173228,
+      "grad_norm": 1.32565438747406,
+      "learning_rate": 7.027028758887728e-05,
+      "loss": 3.1386,
+      "step": 9694
+    },
+    {
+      "epoch": 0.5963951771653543,
+      "grad_norm": 1.1406625509262085,
+      "learning_rate": 7.025182485800991e-05,
+      "loss": 2.7297,
+      "step": 9695
+    },
+    {
+      "epoch": 0.5964566929133859,
+      "grad_norm": 1.456684947013855,
+      "learning_rate": 7.023336323955844e-05,
+      "loss": 2.3214,
+      "step": 9696
+    },
+    {
+      "epoch": 0.5965182086614174,
+      "grad_norm": 1.2167088985443115,
+      "learning_rate": 7.021490273421319e-05,
+      "loss": 2.6119,
+      "step": 9697
+    },
+    {
+      "epoch": 0.5965797244094488,
+      "grad_norm": 1.4250439405441284,
+      "learning_rate": 7.019644334266449e-05,
+      "loss": 2.8944,
+      "step": 9698
+    },
+    {
+      "epoch": 0.5966412401574803,
+      "grad_norm": 1.4159214496612549,
+      "learning_rate": 7.01779850656026e-05,
+      "loss": 2.1251,
+      "step": 9699
+    },
+    {
+      "epoch": 0.5967027559055118,
+      "grad_norm": 1.0244404077529907,
+      "learning_rate": 7.015952790371783e-05,
+      "loss": 2.9601,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5967642716535433,
+      "grad_norm": 1.246267318725586,
+      "learning_rate": 7.014107185770027e-05,
+      "loss": 2.4104,
+      "step": 9701
+    },
+    {
+      "epoch": 0.5968257874015748,
+      "grad_norm": 1.197293996810913,
+      "learning_rate": 7.012261692824015e-05,
+      "loss": 2.9444,
+      "step": 9702
+    },
+    {
+      "epoch": 0.5968873031496063,
+      "grad_norm": 1.1242711544036865,
+      "learning_rate": 7.010416311602753e-05,
+      "loss": 2.9023,
+      "step": 9703
+    },
+    {
+      "epoch": 0.5969488188976378,
+      "grad_norm": 1.2174556255340576,
+      "learning_rate": 7.008571042175256e-05,
+      "loss": 2.6615,
+      "step": 9704
+    },
+    {
+      "epoch": 0.5970103346456693,
+      "grad_norm": 1.1696577072143555,
+      "learning_rate": 7.006725884610513e-05,
+      "loss": 2.2725,
+      "step": 9705
+    },
+    {
+      "epoch": 0.5970718503937008,
+      "grad_norm": 1.373318076133728,
+      "learning_rate": 7.004880838977538e-05,
+      "loss": 2.2351,
+      "step": 9706
+    },
+    {
+      "epoch": 0.5971333661417323,
+      "grad_norm": 1.3037840127944946,
+      "learning_rate": 7.003035905345317e-05,
+      "loss": 2.9409,
+      "step": 9707
+    },
+    {
+      "epoch": 0.5971948818897638,
+      "grad_norm": 1.2341737747192383,
+      "learning_rate": 7.001191083782841e-05,
+      "loss": 2.3851,
+      "step": 9708
+    },
+    {
+      "epoch": 0.5972563976377953,
+      "grad_norm": 1.2113019227981567,
+      "learning_rate": 6.999346374359095e-05,
+      "loss": 2.5246,
+      "step": 9709
+    },
+    {
+      "epoch": 0.5973179133858267,
+      "grad_norm": 1.7731746435165405,
+      "learning_rate": 6.997501777143066e-05,
+      "loss": 3.5125,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5973794291338582,
+      "grad_norm": 1.6535590887069702,
+      "learning_rate": 6.995657292203725e-05,
+      "loss": 2.4176,
+      "step": 9711
+    },
+    {
+      "epoch": 0.5974409448818898,
+      "grad_norm": 1.2697978019714355,
+      "learning_rate": 6.99381291961005e-05,
+      "loss": 2.2578,
+      "step": 9712
+    },
+    {
+      "epoch": 0.5975024606299213,
+      "grad_norm": 1.2351146936416626,
+      "learning_rate": 6.991968659431006e-05,
+      "loss": 2.4339,
+      "step": 9713
+    },
+    {
+      "epoch": 0.5975639763779528,
+      "grad_norm": 1.2556118965148926,
+      "learning_rate": 6.990124511735565e-05,
+      "loss": 2.6468,
+      "step": 9714
+    },
+    {
+      "epoch": 0.5976254921259843,
+      "grad_norm": 1.3576292991638184,
+      "learning_rate": 6.988280476592685e-05,
+      "loss": 2.273,
+      "step": 9715
+    },
+    {
+      "epoch": 0.5976870078740157,
+      "grad_norm": 1.4696630239486694,
+      "learning_rate": 6.986436554071316e-05,
+      "loss": 2.5006,
+      "step": 9716
+    },
+    {
+      "epoch": 0.5977485236220472,
+      "grad_norm": 1.2378016710281372,
+      "learning_rate": 6.984592744240419e-05,
+      "loss": 2.6147,
+      "step": 9717
+    },
+    {
+      "epoch": 0.5978100393700787,
+      "grad_norm": 1.4497205018997192,
+      "learning_rate": 6.982749047168937e-05,
+      "loss": 2.6433,
+      "step": 9718
+    },
+    {
+      "epoch": 0.5978715551181102,
+      "grad_norm": 1.1344549655914307,
+      "learning_rate": 6.98090546292582e-05,
+      "loss": 2.6857,
+      "step": 9719
+    },
+    {
+      "epoch": 0.5979330708661418,
+      "grad_norm": 1.2649824619293213,
+      "learning_rate": 6.97906199158e-05,
+      "loss": 2.0831,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5979945866141733,
+      "grad_norm": 1.1775246858596802,
+      "learning_rate": 6.977218633200417e-05,
+      "loss": 2.42,
+      "step": 9721
+    },
+    {
+      "epoch": 0.5980561023622047,
+      "grad_norm": 1.2324252128601074,
+      "learning_rate": 6.975375387856001e-05,
+      "loss": 2.7767,
+      "step": 9722
+    },
+    {
+      "epoch": 0.5981176181102362,
+      "grad_norm": 1.2812089920043945,
+      "learning_rate": 6.973532255615682e-05,
+      "loss": 2.6512,
+      "step": 9723
+    },
+    {
+      "epoch": 0.5981791338582677,
+      "grad_norm": 1.2760658264160156,
+      "learning_rate": 6.971689236548377e-05,
+      "loss": 2.5336,
+      "step": 9724
+    },
+    {
+      "epoch": 0.5982406496062992,
+      "grad_norm": 1.1751998662948608,
+      "learning_rate": 6.969846330723013e-05,
+      "loss": 2.8693,
+      "step": 9725
+    },
+    {
+      "epoch": 0.5983021653543307,
+      "grad_norm": 1.1719560623168945,
+      "learning_rate": 6.968003538208497e-05,
+      "loss": 2.4467,
+      "step": 9726
+    },
+    {
+      "epoch": 0.5983636811023622,
+      "grad_norm": 1.5010043382644653,
+      "learning_rate": 6.966160859073741e-05,
+      "loss": 2.6809,
+      "step": 9727
+    },
+    {
+      "epoch": 0.5984251968503937,
+      "grad_norm": 1.0644341707229614,
+      "learning_rate": 6.964318293387651e-05,
+      "loss": 2.389,
+      "step": 9728
+    },
+    {
+      "epoch": 0.5984867125984252,
+      "grad_norm": 1.3973894119262695,
+      "learning_rate": 6.962475841219133e-05,
+      "loss": 2.601,
+      "step": 9729
+    },
+    {
+      "epoch": 0.5985482283464567,
+      "grad_norm": 1.2713767290115356,
+      "learning_rate": 6.960633502637079e-05,
+      "loss": 2.066,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5986097440944882,
+      "grad_norm": 1.242937684059143,
+      "learning_rate": 6.958791277710384e-05,
+      "loss": 2.5706,
+      "step": 9731
+    },
+    {
+      "epoch": 0.5986712598425197,
+      "grad_norm": 1.4932661056518555,
+      "learning_rate": 6.956949166507936e-05,
+      "loss": 2.1152,
+      "step": 9732
+    },
+    {
+      "epoch": 0.5987327755905512,
+      "grad_norm": 2.218651533126831,
+      "learning_rate": 6.955107169098622e-05,
+      "loss": 1.9707,
+      "step": 9733
+    },
+    {
+      "epoch": 0.5987942913385826,
+      "grad_norm": 1.4475960731506348,
+      "learning_rate": 6.953265285551319e-05,
+      "loss": 2.5062,
+      "step": 9734
+    },
+    {
+      "epoch": 0.5988558070866141,
+      "grad_norm": 1.1530733108520508,
+      "learning_rate": 6.951423515934909e-05,
+      "loss": 2.7671,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5989173228346457,
+      "grad_norm": 1.3029019832611084,
+      "learning_rate": 6.949581860318257e-05,
+      "loss": 2.1651,
+      "step": 9736
+    },
+    {
+      "epoch": 0.5989788385826772,
+      "grad_norm": 1.3225599527359009,
+      "learning_rate": 6.947740318770237e-05,
+      "loss": 2.2486,
+      "step": 9737
+    },
+    {
+      "epoch": 0.5990403543307087,
+      "grad_norm": 1.3841599225997925,
+      "learning_rate": 6.945898891359705e-05,
+      "loss": 2.4055,
+      "step": 9738
+    },
+    {
+      "epoch": 0.5991018700787402,
+      "grad_norm": 1.4796360731124878,
+      "learning_rate": 6.94405757815553e-05,
+      "loss": 2.5718,
+      "step": 9739
+    },
+    {
+      "epoch": 0.5991633858267716,
+      "grad_norm": 1.185036540031433,
+      "learning_rate": 6.942216379226553e-05,
+      "loss": 2.5143,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5992249015748031,
+      "grad_norm": 1.2543270587921143,
+      "learning_rate": 6.940375294641638e-05,
+      "loss": 2.5062,
+      "step": 9741
+    },
+    {
+      "epoch": 0.5992864173228346,
+      "grad_norm": 1.509778380393982,
+      "learning_rate": 6.938534324469626e-05,
+      "loss": 2.9807,
+      "step": 9742
+    },
+    {
+      "epoch": 0.5993479330708661,
+      "grad_norm": 1.3855265378952026,
+      "learning_rate": 6.936693468779358e-05,
+      "loss": 2.5772,
+      "step": 9743
+    },
+    {
+      "epoch": 0.5994094488188977,
+      "grad_norm": 1.210081934928894,
+      "learning_rate": 6.934852727639673e-05,
+      "loss": 2.6252,
+      "step": 9744
+    },
+    {
+      "epoch": 0.5994709645669292,
+      "grad_norm": 1.3241639137268066,
+      "learning_rate": 6.933012101119407e-05,
+      "loss": 2.6661,
+      "step": 9745
+    },
+    {
+      "epoch": 0.5995324803149606,
+      "grad_norm": 1.24818754196167,
+      "learning_rate": 6.931171589287382e-05,
+      "loss": 2.5726,
+      "step": 9746
+    },
+    {
+      "epoch": 0.5995939960629921,
+      "grad_norm": 1.3061069250106812,
+      "learning_rate": 6.92933119221243e-05,
+      "loss": 2.1433,
+      "step": 9747
+    },
+    {
+      "epoch": 0.5996555118110236,
+      "grad_norm": 1.6160396337509155,
+      "learning_rate": 6.927490909963367e-05,
+      "loss": 2.4867,
+      "step": 9748
+    },
+    {
+      "epoch": 0.5997170275590551,
+      "grad_norm": 1.1909364461898804,
+      "learning_rate": 6.925650742609013e-05,
+      "loss": 2.471,
+      "step": 9749
+    },
+    {
+      "epoch": 0.5997785433070866,
+      "grad_norm": 1.1884629726409912,
+      "learning_rate": 6.923810690218176e-05,
+      "loss": 3.069,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5998400590551181,
+      "grad_norm": 1.469449758529663,
+      "learning_rate": 6.92197075285967e-05,
+      "loss": 3.0396,
+      "step": 9751
+    },
+    {
+      "epoch": 0.5999015748031497,
+      "grad_norm": 1.4044455289840698,
+      "learning_rate": 6.920130930602292e-05,
+      "loss": 2.3506,
+      "step": 9752
+    },
+    {
+      "epoch": 0.5999630905511811,
+      "grad_norm": 1.3117482662200928,
+      "learning_rate": 6.918291223514841e-05,
+      "loss": 2.4551,
+      "step": 9753
+    },
+    {
+      "epoch": 0.6000246062992126,
+      "grad_norm": 1.318999171257019,
+      "learning_rate": 6.91645163166612e-05,
+      "loss": 2.1631,
+      "step": 9754
+    },
+    {
+      "epoch": 0.6000861220472441,
+      "grad_norm": 1.1680430173873901,
+      "learning_rate": 6.914612155124909e-05,
+      "loss": 2.0121,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6001476377952756,
+      "grad_norm": 1.593333125114441,
+      "learning_rate": 6.91277279396e-05,
+      "loss": 2.8822,
+      "step": 9756
+    },
+    {
+      "epoch": 0.6002091535433071,
+      "grad_norm": 1.2421700954437256,
+      "learning_rate": 6.910933548240171e-05,
+      "loss": 2.3849,
+      "step": 9757
+    },
+    {
+      "epoch": 0.6002706692913385,
+      "grad_norm": 1.3078434467315674,
+      "learning_rate": 6.909094418034208e-05,
+      "loss": 2.1616,
+      "step": 9758
+    },
+    {
+      "epoch": 0.60033218503937,
+      "grad_norm": 1.1922647953033447,
+      "learning_rate": 6.90725540341087e-05,
+      "loss": 1.9463,
+      "step": 9759
+    },
+    {
+      "epoch": 0.6003937007874016,
+      "grad_norm": 1.1400395631790161,
+      "learning_rate": 6.905416504438942e-05,
+      "loss": 1.8715,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6004552165354331,
+      "grad_norm": 1.3099052906036377,
+      "learning_rate": 6.903577721187177e-05,
+      "loss": 1.9974,
+      "step": 9761
+    },
+    {
+      "epoch": 0.6005167322834646,
+      "grad_norm": 1.349771499633789,
+      "learning_rate": 6.90173905372434e-05,
+      "loss": 2.4739,
+      "step": 9762
+    },
+    {
+      "epoch": 0.6005782480314961,
+      "grad_norm": 1.485307216644287,
+      "learning_rate": 6.899900502119186e-05,
+      "loss": 3.1013,
+      "step": 9763
+    },
+    {
+      "epoch": 0.6006397637795275,
+      "grad_norm": 1.188683032989502,
+      "learning_rate": 6.89806206644047e-05,
+      "loss": 2.3478,
+      "step": 9764
+    },
+    {
+      "epoch": 0.600701279527559,
+      "grad_norm": 1.3666459321975708,
+      "learning_rate": 6.89622374675693e-05,
+      "loss": 2.2448,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6007627952755905,
+      "grad_norm": 1.0517417192459106,
+      "learning_rate": 6.894385543137318e-05,
+      "loss": 1.6877,
+      "step": 9766
+    },
+    {
+      "epoch": 0.600824311023622,
+      "grad_norm": 1.3021575212478638,
+      "learning_rate": 6.892547455650367e-05,
+      "loss": 2.047,
+      "step": 9767
+    },
+    {
+      "epoch": 0.6008858267716536,
+      "grad_norm": 1.3485373258590698,
+      "learning_rate": 6.890709484364816e-05,
+      "loss": 2.8679,
+      "step": 9768
+    },
+    {
+      "epoch": 0.6009473425196851,
+      "grad_norm": 1.2261005640029907,
+      "learning_rate": 6.88887162934939e-05,
+      "loss": 2.1654,
+      "step": 9769
+    },
+    {
+      "epoch": 0.6010088582677166,
+      "grad_norm": 1.2353347539901733,
+      "learning_rate": 6.887033890672821e-05,
+      "loss": 2.5595,
+      "step": 9770
+    },
+    {
+      "epoch": 0.601070374015748,
+      "grad_norm": 1.6333057880401611,
+      "learning_rate": 6.885196268403823e-05,
+      "loss": 3.5585,
+      "step": 9771
+    },
+    {
+      "epoch": 0.6011318897637795,
+      "grad_norm": 1.3541569709777832,
+      "learning_rate": 6.883358762611116e-05,
+      "loss": 1.8298,
+      "step": 9772
+    },
+    {
+      "epoch": 0.601193405511811,
+      "grad_norm": 1.206099510192871,
+      "learning_rate": 6.881521373363413e-05,
+      "loss": 2.9482,
+      "step": 9773
+    },
+    {
+      "epoch": 0.6012549212598425,
+      "grad_norm": 1.389768362045288,
+      "learning_rate": 6.879684100729424e-05,
+      "loss": 2.5601,
+      "step": 9774
+    },
+    {
+      "epoch": 0.601316437007874,
+      "grad_norm": 1.239404320716858,
+      "learning_rate": 6.877846944777843e-05,
+      "loss": 2.1426,
+      "step": 9775
+    },
+    {
+      "epoch": 0.6013779527559056,
+      "grad_norm": 1.572130560874939,
+      "learning_rate": 6.876009905577385e-05,
+      "loss": 2.8947,
+      "step": 9776
+    },
+    {
+      "epoch": 0.601439468503937,
+      "grad_norm": 1.151388168334961,
+      "learning_rate": 6.874172983196732e-05,
+      "loss": 2.1233,
+      "step": 9777
+    },
+    {
+      "epoch": 0.6015009842519685,
+      "grad_norm": 1.328539252281189,
+      "learning_rate": 6.872336177704582e-05,
+      "loss": 2.9636,
+      "step": 9778
+    },
+    {
+      "epoch": 0.6015625,
+      "grad_norm": 1.4218841791152954,
+      "learning_rate": 6.870499489169615e-05,
+      "loss": 2.3554,
+      "step": 9779
+    },
+    {
+      "epoch": 0.6016240157480315,
+      "grad_norm": 1.100711464881897,
+      "learning_rate": 6.86866291766052e-05,
+      "loss": 2.535,
+      "step": 9780
+    },
+    {
+      "epoch": 0.601685531496063,
+      "grad_norm": 1.0382953882217407,
+      "learning_rate": 6.866826463245969e-05,
+      "loss": 2.8138,
+      "step": 9781
+    },
+    {
+      "epoch": 0.6017470472440944,
+      "grad_norm": 1.1608577966690063,
+      "learning_rate": 6.864990125994639e-05,
+      "loss": 2.9802,
+      "step": 9782
+    },
+    {
+      "epoch": 0.601808562992126,
+      "grad_norm": 1.2833484411239624,
+      "learning_rate": 6.863153905975194e-05,
+      "loss": 2.2022,
+      "step": 9783
+    },
+    {
+      "epoch": 0.6018700787401575,
+      "grad_norm": 1.6915305852890015,
+      "learning_rate": 6.861317803256306e-05,
+      "loss": 2.637,
+      "step": 9784
+    },
+    {
+      "epoch": 0.601931594488189,
+      "grad_norm": 1.5136306285858154,
+      "learning_rate": 6.859481817906623e-05,
+      "loss": 2.4295,
+      "step": 9785
+    },
+    {
+      "epoch": 0.6019931102362205,
+      "grad_norm": 1.3044019937515259,
+      "learning_rate": 6.857645949994817e-05,
+      "loss": 2.3577,
+      "step": 9786
+    },
+    {
+      "epoch": 0.602054625984252,
+      "grad_norm": 1.1345988512039185,
+      "learning_rate": 6.855810199589523e-05,
+      "loss": 2.6126,
+      "step": 9787
+    },
+    {
+      "epoch": 0.6021161417322834,
+      "grad_norm": 1.140833854675293,
+      "learning_rate": 6.853974566759398e-05,
+      "loss": 2.342,
+      "step": 9788
+    },
+    {
+      "epoch": 0.6021776574803149,
+      "grad_norm": 1.5404249429702759,
+      "learning_rate": 6.852139051573085e-05,
+      "loss": 2.465,
+      "step": 9789
+    },
+    {
+      "epoch": 0.6022391732283464,
+      "grad_norm": 1.158915400505066,
+      "learning_rate": 6.850303654099213e-05,
+      "loss": 2.4003,
+      "step": 9790
+    },
+    {
+      "epoch": 0.602300688976378,
+      "grad_norm": 1.2202861309051514,
+      "learning_rate": 6.848468374406425e-05,
+      "loss": 2.8531,
+      "step": 9791
+    },
+    {
+      "epoch": 0.6023622047244095,
+      "grad_norm": 1.2401710748672485,
+      "learning_rate": 6.846633212563342e-05,
+      "loss": 2.5155,
+      "step": 9792
+    },
+    {
+      "epoch": 0.602423720472441,
+      "grad_norm": 1.2648224830627441,
+      "learning_rate": 6.8447981686386e-05,
+      "loss": 2.6477,
+      "step": 9793
+    },
+    {
+      "epoch": 0.6024852362204725,
+      "grad_norm": 1.1879395246505737,
+      "learning_rate": 6.842963242700804e-05,
+      "loss": 2.2459,
+      "step": 9794
+    },
+    {
+      "epoch": 0.6025467519685039,
+      "grad_norm": 1.2818528413772583,
+      "learning_rate": 6.841128434818585e-05,
+      "loss": 2.9418,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6026082677165354,
+      "grad_norm": 1.4008591175079346,
+      "learning_rate": 6.839293745060544e-05,
+      "loss": 2.4008,
+      "step": 9796
+    },
+    {
+      "epoch": 0.6026697834645669,
+      "grad_norm": 1.2720427513122559,
+      "learning_rate": 6.837459173495295e-05,
+      "loss": 3.3257,
+      "step": 9797
+    },
+    {
+      "epoch": 0.6027312992125984,
+      "grad_norm": 1.274795413017273,
+      "learning_rate": 6.835624720191437e-05,
+      "loss": 2.5172,
+      "step": 9798
+    },
+    {
+      "epoch": 0.60279281496063,
+      "grad_norm": 1.2063535451889038,
+      "learning_rate": 6.833790385217573e-05,
+      "loss": 2.5407,
+      "step": 9799
+    },
+    {
+      "epoch": 0.6028543307086615,
+      "grad_norm": 3.3712992668151855,
+      "learning_rate": 6.831956168642289e-05,
+      "loss": 3.0663,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6029158464566929,
+      "grad_norm": 0.9995375275611877,
+      "learning_rate": 6.83012207053418e-05,
+      "loss": 1.7239,
+      "step": 9801
+    },
+    {
+      "epoch": 0.6029773622047244,
+      "grad_norm": 1.5556999444961548,
+      "learning_rate": 6.828288090961829e-05,
+      "loss": 1.9563,
+      "step": 9802
+    },
+    {
+      "epoch": 0.6030388779527559,
+      "grad_norm": 1.331146001815796,
+      "learning_rate": 6.826454229993819e-05,
+      "loss": 2.8649,
+      "step": 9803
+    },
+    {
+      "epoch": 0.6031003937007874,
+      "grad_norm": 1.3083586692810059,
+      "learning_rate": 6.824620487698721e-05,
+      "loss": 2.8557,
+      "step": 9804
+    },
+    {
+      "epoch": 0.6031619094488189,
+      "grad_norm": 1.3747366666793823,
+      "learning_rate": 6.822786864145116e-05,
+      "loss": 2.1743,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6032234251968503,
+      "grad_norm": 1.4892027378082275,
+      "learning_rate": 6.820953359401562e-05,
+      "loss": 2.5145,
+      "step": 9806
+    },
+    {
+      "epoch": 0.6032849409448819,
+      "grad_norm": 1.160078763961792,
+      "learning_rate": 6.819119973536628e-05,
+      "loss": 2.4621,
+      "step": 9807
+    },
+    {
+      "epoch": 0.6033464566929134,
+      "grad_norm": 1.1065906286239624,
+      "learning_rate": 6.817286706618866e-05,
+      "loss": 1.8857,
+      "step": 9808
+    },
+    {
+      "epoch": 0.6034079724409449,
+      "grad_norm": 1.0710889101028442,
+      "learning_rate": 6.81545355871684e-05,
+      "loss": 2.2346,
+      "step": 9809
+    },
+    {
+      "epoch": 0.6034694881889764,
+      "grad_norm": 1.190867304801941,
+      "learning_rate": 6.813620529899087e-05,
+      "loss": 2.7516,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6035310039370079,
+      "grad_norm": 1.247504472732544,
+      "learning_rate": 6.811787620234162e-05,
+      "loss": 2.4462,
+      "step": 9811
+    },
+    {
+      "epoch": 0.6035925196850394,
+      "grad_norm": 1.648169994354248,
+      "learning_rate": 6.8099548297906e-05,
+      "loss": 2.0171,
+      "step": 9812
+    },
+    {
+      "epoch": 0.6036540354330708,
+      "grad_norm": 1.1716581583023071,
+      "learning_rate": 6.80812215863694e-05,
+      "loss": 2.7897,
+      "step": 9813
+    },
+    {
+      "epoch": 0.6037155511811023,
+      "grad_norm": 1.2515931129455566,
+      "learning_rate": 6.806289606841713e-05,
+      "loss": 2.3783,
+      "step": 9814
+    },
+    {
+      "epoch": 0.6037770669291339,
+      "grad_norm": 1.3095088005065918,
+      "learning_rate": 6.804457174473448e-05,
+      "loss": 3.14,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6038385826771654,
+      "grad_norm": 1.202420711517334,
+      "learning_rate": 6.802624861600664e-05,
+      "loss": 2.2885,
+      "step": 9816
+    },
+    {
+      "epoch": 0.6039000984251969,
+      "grad_norm": 1.3738082647323608,
+      "learning_rate": 6.800792668291881e-05,
+      "loss": 1.6196,
+      "step": 9817
+    },
+    {
+      "epoch": 0.6039616141732284,
+      "grad_norm": 1.4599957466125488,
+      "learning_rate": 6.798960594615611e-05,
+      "loss": 2.6161,
+      "step": 9818
+    },
+    {
+      "epoch": 0.6040231299212598,
+      "grad_norm": 1.1938358545303345,
+      "learning_rate": 6.797128640640371e-05,
+      "loss": 2.4976,
+      "step": 9819
+    },
+    {
+      "epoch": 0.6040846456692913,
+      "grad_norm": 1.099785566329956,
+      "learning_rate": 6.795296806434653e-05,
+      "loss": 2.3429,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6041461614173228,
+      "grad_norm": 1.1921950578689575,
+      "learning_rate": 6.79346509206697e-05,
+      "loss": 2.5,
+      "step": 9821
+    },
+    {
+      "epoch": 0.6042076771653543,
+      "grad_norm": 1.3081262111663818,
+      "learning_rate": 6.79163349760581e-05,
+      "loss": 2.2204,
+      "step": 9822
+    },
+    {
+      "epoch": 0.6042691929133859,
+      "grad_norm": 1.0453537702560425,
+      "learning_rate": 6.789802023119667e-05,
+      "loss": 2.785,
+      "step": 9823
+    },
+    {
+      "epoch": 0.6043307086614174,
+      "grad_norm": 1.2076890468597412,
+      "learning_rate": 6.787970668677033e-05,
+      "loss": 2.7768,
+      "step": 9824
+    },
+    {
+      "epoch": 0.6043922244094488,
+      "grad_norm": 1.2312395572662354,
+      "learning_rate": 6.786139434346377e-05,
+      "loss": 2.0599,
+      "step": 9825
+    },
+    {
+      "epoch": 0.6044537401574803,
+      "grad_norm": 1.4213868379592896,
+      "learning_rate": 6.784308320196192e-05,
+      "loss": 2.2071,
+      "step": 9826
+    },
+    {
+      "epoch": 0.6045152559055118,
+      "grad_norm": 1.4287081956863403,
+      "learning_rate": 6.782477326294938e-05,
+      "loss": 2.9921,
+      "step": 9827
+    },
+    {
+      "epoch": 0.6045767716535433,
+      "grad_norm": 1.5035371780395508,
+      "learning_rate": 6.780646452711097e-05,
+      "loss": 2.0811,
+      "step": 9828
+    },
+    {
+      "epoch": 0.6046382874015748,
+      "grad_norm": 1.6684101819992065,
+      "learning_rate": 6.778815699513119e-05,
+      "loss": 2.5903,
+      "step": 9829
+    },
+    {
+      "epoch": 0.6046998031496063,
+      "grad_norm": 1.2748204469680786,
+      "learning_rate": 6.776985066769481e-05,
+      "loss": 2.4001,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6047613188976378,
+      "grad_norm": 1.0269943475723267,
+      "learning_rate": 6.775154554548624e-05,
+      "loss": 2.4392,
+      "step": 9831
+    },
+    {
+      "epoch": 0.6048228346456693,
+      "grad_norm": 1.159308910369873,
+      "learning_rate": 6.773324162919006e-05,
+      "loss": 2.5888,
+      "step": 9832
+    },
+    {
+      "epoch": 0.6048843503937008,
+      "grad_norm": 1.2944666147232056,
+      "learning_rate": 6.77149389194907e-05,
+      "loss": 2.3568,
+      "step": 9833
+    },
+    {
+      "epoch": 0.6049458661417323,
+      "grad_norm": 1.14823579788208,
+      "learning_rate": 6.769663741707264e-05,
+      "loss": 2.4926,
+      "step": 9834
+    },
+    {
+      "epoch": 0.6050073818897638,
+      "grad_norm": 1.1430994272232056,
+      "learning_rate": 6.767833712262017e-05,
+      "loss": 3.1435,
+      "step": 9835
+    },
+    {
+      "epoch": 0.6050688976377953,
+      "grad_norm": 1.2137141227722168,
+      "learning_rate": 6.766003803681769e-05,
+      "loss": 2.8037,
+      "step": 9836
+    },
+    {
+      "epoch": 0.6051304133858267,
+      "grad_norm": 1.2866195440292358,
+      "learning_rate": 6.764174016034943e-05,
+      "loss": 2.6365,
+      "step": 9837
+    },
+    {
+      "epoch": 0.6051919291338582,
+      "grad_norm": 1.1594384908676147,
+      "learning_rate": 6.762344349389966e-05,
+      "loss": 2.5398,
+      "step": 9838
+    },
+    {
+      "epoch": 0.6052534448818898,
+      "grad_norm": 1.3699404001235962,
+      "learning_rate": 6.760514803815257e-05,
+      "loss": 2.0791,
+      "step": 9839
+    },
+    {
+      "epoch": 0.6053149606299213,
+      "grad_norm": 1.1321613788604736,
+      "learning_rate": 6.758685379379233e-05,
+      "loss": 2.5185,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6053764763779528,
+      "grad_norm": 1.2013226747512817,
+      "learning_rate": 6.756856076150297e-05,
+      "loss": 2.4164,
+      "step": 9841
+    },
+    {
+      "epoch": 0.6054379921259843,
+      "grad_norm": 1.0886785984039307,
+      "learning_rate": 6.755026894196863e-05,
+      "loss": 2.7956,
+      "step": 9842
+    },
+    {
+      "epoch": 0.6054995078740157,
+      "grad_norm": 1.1373172998428345,
+      "learning_rate": 6.753197833587327e-05,
+      "loss": 2.8163,
+      "step": 9843
+    },
+    {
+      "epoch": 0.6055610236220472,
+      "grad_norm": 1.2237169742584229,
+      "learning_rate": 6.751368894390091e-05,
+      "loss": 2.4284,
+      "step": 9844
+    },
+    {
+      "epoch": 0.6056225393700787,
+      "grad_norm": 1.3746575117111206,
+      "learning_rate": 6.74954007667354e-05,
+      "loss": 2.4786,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6056840551181102,
+      "grad_norm": 1.4543513059616089,
+      "learning_rate": 6.747711380506065e-05,
+      "loss": 2.494,
+      "step": 9846
+    },
+    {
+      "epoch": 0.6057455708661418,
+      "grad_norm": 1.272995114326477,
+      "learning_rate": 6.745882805956047e-05,
+      "loss": 2.2964,
+      "step": 9847
+    },
+    {
+      "epoch": 0.6058070866141733,
+      "grad_norm": 1.3691567182540894,
+      "learning_rate": 6.744054353091871e-05,
+      "loss": 2.7541,
+      "step": 9848
+    },
+    {
+      "epoch": 0.6058686023622047,
+      "grad_norm": 1.2515000104904175,
+      "learning_rate": 6.742226021981903e-05,
+      "loss": 2.2724,
+      "step": 9849
+    },
+    {
+      "epoch": 0.6059301181102362,
+      "grad_norm": 1.3069610595703125,
+      "learning_rate": 6.740397812694521e-05,
+      "loss": 2.0085,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6059916338582677,
+      "grad_norm": 1.2609158754348755,
+      "learning_rate": 6.738569725298081e-05,
+      "loss": 2.4462,
+      "step": 9851
+    },
+    {
+      "epoch": 0.6060531496062992,
+      "grad_norm": 1.2488808631896973,
+      "learning_rate": 6.736741759860947e-05,
+      "loss": 2.6337,
+      "step": 9852
+    },
+    {
+      "epoch": 0.6061146653543307,
+      "grad_norm": 1.3000465631484985,
+      "learning_rate": 6.734913916451475e-05,
+      "loss": 2.3547,
+      "step": 9853
+    },
+    {
+      "epoch": 0.6061761811023622,
+      "grad_norm": 1.2373818159103394,
+      "learning_rate": 6.733086195138019e-05,
+      "loss": 2.3199,
+      "step": 9854
+    },
+    {
+      "epoch": 0.6062376968503937,
+      "grad_norm": 1.2090634107589722,
+      "learning_rate": 6.731258595988915e-05,
+      "loss": 2.4964,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6062992125984252,
+      "grad_norm": 1.317165732383728,
+      "learning_rate": 6.72943111907252e-05,
+      "loss": 3.3069,
+      "step": 9856
+    },
+    {
+      "epoch": 0.6063607283464567,
+      "grad_norm": 1.2187089920043945,
+      "learning_rate": 6.72760376445716e-05,
+      "loss": 2.3504,
+      "step": 9857
+    },
+    {
+      "epoch": 0.6064222440944882,
+      "grad_norm": 1.2370244264602661,
+      "learning_rate": 6.725776532211174e-05,
+      "loss": 2.0984,
+      "step": 9858
+    },
+    {
+      "epoch": 0.6064837598425197,
+      "grad_norm": 1.24349844455719,
+      "learning_rate": 6.723949422402886e-05,
+      "loss": 2.0836,
+      "step": 9859
+    },
+    {
+      "epoch": 0.6065452755905512,
+      "grad_norm": 1.279971957206726,
+      "learning_rate": 6.722122435100625e-05,
+      "loss": 2.3919,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6066067913385826,
+      "grad_norm": 1.00607430934906,
+      "learning_rate": 6.720295570372705e-05,
+      "loss": 3.0803,
+      "step": 9861
+    },
+    {
+      "epoch": 0.6066683070866141,
+      "grad_norm": 1.3849806785583496,
+      "learning_rate": 6.718468828287442e-05,
+      "loss": 1.9739,
+      "step": 9862
+    },
+    {
+      "epoch": 0.6067298228346457,
+      "grad_norm": 1.2006856203079224,
+      "learning_rate": 6.71664220891315e-05,
+      "loss": 2.3185,
+      "step": 9863
+    },
+    {
+      "epoch": 0.6067913385826772,
+      "grad_norm": 1.2134038209915161,
+      "learning_rate": 6.714815712318123e-05,
+      "loss": 3.1472,
+      "step": 9864
+    },
+    {
+      "epoch": 0.6068528543307087,
+      "grad_norm": 1.2650344371795654,
+      "learning_rate": 6.712989338570679e-05,
+      "loss": 2.4316,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6069143700787402,
+      "grad_norm": 1.6455477476119995,
+      "learning_rate": 6.711163087739098e-05,
+      "loss": 2.0094,
+      "step": 9866
+    },
+    {
+      "epoch": 0.6069758858267716,
+      "grad_norm": 1.32550048828125,
+      "learning_rate": 6.709336959891681e-05,
+      "loss": 2.4815,
+      "step": 9867
+    },
+    {
+      "epoch": 0.6070374015748031,
+      "grad_norm": 1.2488782405853271,
+      "learning_rate": 6.707510955096712e-05,
+      "loss": 2.4812,
+      "step": 9868
+    },
+    {
+      "epoch": 0.6070989173228346,
+      "grad_norm": 1.255958080291748,
+      "learning_rate": 6.705685073422477e-05,
+      "loss": 2.7191,
+      "step": 9869
+    },
+    {
+      "epoch": 0.6071604330708661,
+      "grad_norm": 1.7345296144485474,
+      "learning_rate": 6.703859314937246e-05,
+      "loss": 2.2044,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6072219488188977,
+      "grad_norm": 1.1317100524902344,
+      "learning_rate": 6.702033679709299e-05,
+      "loss": 2.5546,
+      "step": 9871
+    },
+    {
+      "epoch": 0.6072834645669292,
+      "grad_norm": 1.3361108303070068,
+      "learning_rate": 6.7002081678069e-05,
+      "loss": 2.4484,
+      "step": 9872
+    },
+    {
+      "epoch": 0.6073449803149606,
+      "grad_norm": 1.1897008419036865,
+      "learning_rate": 6.698382779298319e-05,
+      "loss": 2.3438,
+      "step": 9873
+    },
+    {
+      "epoch": 0.6074064960629921,
+      "grad_norm": 1.1920042037963867,
+      "learning_rate": 6.696557514251809e-05,
+      "loss": 2.3815,
+      "step": 9874
+    },
+    {
+      "epoch": 0.6074680118110236,
+      "grad_norm": 1.0794498920440674,
+      "learning_rate": 6.69473237273563e-05,
+      "loss": 2.7649,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6075295275590551,
+      "grad_norm": 1.316332221031189,
+      "learning_rate": 6.692907354818028e-05,
+      "loss": 2.1612,
+      "step": 9876
+    },
+    {
+      "epoch": 0.6075910433070866,
+      "grad_norm": 1.4497408866882324,
+      "learning_rate": 6.69108246056725e-05,
+      "loss": 2.0134,
+      "step": 9877
+    },
+    {
+      "epoch": 0.6076525590551181,
+      "grad_norm": 1.1618759632110596,
+      "learning_rate": 6.689257690051536e-05,
+      "loss": 3.1027,
+      "step": 9878
+    },
+    {
+      "epoch": 0.6077140748031497,
+      "grad_norm": 1.6131585836410522,
+      "learning_rate": 6.687433043339129e-05,
+      "loss": 2.208,
+      "step": 9879
+    },
+    {
+      "epoch": 0.6077755905511811,
+      "grad_norm": 1.1727057695388794,
+      "learning_rate": 6.685608520498247e-05,
+      "loss": 3.1713,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6078371062992126,
+      "grad_norm": 1.2874152660369873,
+      "learning_rate": 6.68378412159713e-05,
+      "loss": 2.5,
+      "step": 9881
+    },
+    {
+      "epoch": 0.6078986220472441,
+      "grad_norm": 1.3059587478637695,
+      "learning_rate": 6.681959846703991e-05,
+      "loss": 2.4024,
+      "step": 9882
+    },
+    {
+      "epoch": 0.6079601377952756,
+      "grad_norm": 1.1582242250442505,
+      "learning_rate": 6.680135695887056e-05,
+      "loss": 2.7884,
+      "step": 9883
+    },
+    {
+      "epoch": 0.6080216535433071,
+      "grad_norm": 1.4074013233184814,
+      "learning_rate": 6.67831166921453e-05,
+      "loss": 2.8935,
+      "step": 9884
+    },
+    {
+      "epoch": 0.6080831692913385,
+      "grad_norm": 1.2798842191696167,
+      "learning_rate": 6.676487766754629e-05,
+      "loss": 2.8304,
+      "step": 9885
+    },
+    {
+      "epoch": 0.60814468503937,
+      "grad_norm": 1.1342332363128662,
+      "learning_rate": 6.674663988575551e-05,
+      "loss": 2.1095,
+      "step": 9886
+    },
+    {
+      "epoch": 0.6082062007874016,
+      "grad_norm": 1.2877187728881836,
+      "learning_rate": 6.672840334745497e-05,
+      "loss": 2.3674,
+      "step": 9887
+    },
+    {
+      "epoch": 0.6082677165354331,
+      "grad_norm": 1.250684142112732,
+      "learning_rate": 6.671016805332662e-05,
+      "loss": 2.3781,
+      "step": 9888
+    },
+    {
+      "epoch": 0.6083292322834646,
+      "grad_norm": 1.2997536659240723,
+      "learning_rate": 6.669193400405238e-05,
+      "loss": 2.1897,
+      "step": 9889
+    },
+    {
+      "epoch": 0.6083907480314961,
+      "grad_norm": 1.1337984800338745,
+      "learning_rate": 6.667370120031402e-05,
+      "loss": 2.8638,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6084522637795275,
+      "grad_norm": 1.2194536924362183,
+      "learning_rate": 6.665546964279346e-05,
+      "loss": 2.358,
+      "step": 9891
+    },
+    {
+      "epoch": 0.608513779527559,
+      "grad_norm": 1.2106194496154785,
+      "learning_rate": 6.663723933217236e-05,
+      "loss": 2.6535,
+      "step": 9892
+    },
+    {
+      "epoch": 0.6085752952755905,
+      "grad_norm": 1.2954257726669312,
+      "learning_rate": 6.66190102691325e-05,
+      "loss": 1.8418,
+      "step": 9893
+    },
+    {
+      "epoch": 0.608636811023622,
+      "grad_norm": 1.210898756980896,
+      "learning_rate": 6.66007824543555e-05,
+      "loss": 2.7113,
+      "step": 9894
+    },
+    {
+      "epoch": 0.6086983267716536,
+      "grad_norm": 1.175424575805664,
+      "learning_rate": 6.658255588852305e-05,
+      "loss": 2.6473,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6087598425196851,
+      "grad_norm": 1.2861524820327759,
+      "learning_rate": 6.65643305723166e-05,
+      "loss": 1.8501,
+      "step": 9896
+    },
+    {
+      "epoch": 0.6088213582677166,
+      "grad_norm": 1.404852271080017,
+      "learning_rate": 6.654610650641778e-05,
+      "loss": 2.0002,
+      "step": 9897
+    },
+    {
+      "epoch": 0.608882874015748,
+      "grad_norm": 1.2286819219589233,
+      "learning_rate": 6.652788369150805e-05,
+      "loss": 2.6356,
+      "step": 9898
+    },
+    {
+      "epoch": 0.6089443897637795,
+      "grad_norm": 1.8991138935089111,
+      "learning_rate": 6.650966212826877e-05,
+      "loss": 2.9315,
+      "step": 9899
+    },
+    {
+      "epoch": 0.609005905511811,
+      "grad_norm": 2.013005495071411,
+      "learning_rate": 6.649144181738142e-05,
+      "loss": 2.0118,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6090674212598425,
+      "grad_norm": 1.3486236333847046,
+      "learning_rate": 6.647322275952728e-05,
+      "loss": 2.5908,
+      "step": 9901
+    },
+    {
+      "epoch": 0.609128937007874,
+      "grad_norm": 1.4934053421020508,
+      "learning_rate": 6.645500495538768e-05,
+      "loss": 1.9111,
+      "step": 9902
+    },
+    {
+      "epoch": 0.6091904527559056,
+      "grad_norm": 1.2645920515060425,
+      "learning_rate": 6.64367884056438e-05,
+      "loss": 2.4529,
+      "step": 9903
+    },
+    {
+      "epoch": 0.609251968503937,
+      "grad_norm": 1.2851815223693848,
+      "learning_rate": 6.641857311097694e-05,
+      "loss": 2.3,
+      "step": 9904
+    },
+    {
+      "epoch": 0.6093134842519685,
+      "grad_norm": 1.1074784994125366,
+      "learning_rate": 6.640035907206814e-05,
+      "loss": 2.5097,
+      "step": 9905
+    },
+    {
+      "epoch": 0.609375,
+      "grad_norm": 1.218320608139038,
+      "learning_rate": 6.638214628959857e-05,
+      "loss": 2.7382,
+      "step": 9906
+    },
+    {
+      "epoch": 0.6094365157480315,
+      "grad_norm": 1.2223756313323975,
+      "learning_rate": 6.636393476424927e-05,
+      "loss": 2.1791,
+      "step": 9907
+    },
+    {
+      "epoch": 0.609498031496063,
+      "grad_norm": 1.296067476272583,
+      "learning_rate": 6.634572449670127e-05,
+      "loss": 2.605,
+      "step": 9908
+    },
+    {
+      "epoch": 0.6095595472440944,
+      "grad_norm": 1.2814337015151978,
+      "learning_rate": 6.632751548763546e-05,
+      "loss": 2.6637,
+      "step": 9909
+    },
+    {
+      "epoch": 0.609621062992126,
+      "grad_norm": 1.2572617530822754,
+      "learning_rate": 6.630930773773287e-05,
+      "loss": 2.9575,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6096825787401575,
+      "grad_norm": 1.1589611768722534,
+      "learning_rate": 6.629110124767426e-05,
+      "loss": 2.3994,
+      "step": 9911
+    },
+    {
+      "epoch": 0.609744094488189,
+      "grad_norm": 1.253210425376892,
+      "learning_rate": 6.627289601814053e-05,
+      "loss": 2.0886,
+      "step": 9912
+    },
+    {
+      "epoch": 0.6098056102362205,
+      "grad_norm": 1.4568487405776978,
+      "learning_rate": 6.625469204981239e-05,
+      "loss": 2.4445,
+      "step": 9913
+    },
+    {
+      "epoch": 0.609867125984252,
+      "grad_norm": 1.3033747673034668,
+      "learning_rate": 6.623648934337063e-05,
+      "loss": 3.4234,
+      "step": 9914
+    },
+    {
+      "epoch": 0.6099286417322834,
+      "grad_norm": 1.3172202110290527,
+      "learning_rate": 6.621828789949586e-05,
+      "loss": 2.5081,
+      "step": 9915
+    },
+    {
+      "epoch": 0.6099901574803149,
+      "grad_norm": 1.7578140497207642,
+      "learning_rate": 6.620008771886879e-05,
+      "loss": 1.7361,
+      "step": 9916
+    },
+    {
+      "epoch": 0.6100516732283464,
+      "grad_norm": 1.1477214097976685,
+      "learning_rate": 6.618188880216993e-05,
+      "loss": 2.0823,
+      "step": 9917
+    },
+    {
+      "epoch": 0.610113188976378,
+      "grad_norm": 1.4548250436782837,
+      "learning_rate": 6.616369115007989e-05,
+      "loss": 2.7614,
+      "step": 9918
+    },
+    {
+      "epoch": 0.6101747047244095,
+      "grad_norm": 1.1552423238754272,
+      "learning_rate": 6.614549476327909e-05,
+      "loss": 1.539,
+      "step": 9919
+    },
+    {
+      "epoch": 0.610236220472441,
+      "grad_norm": 1.2290503978729248,
+      "learning_rate": 6.612729964244806e-05,
+      "loss": 2.7602,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6102977362204725,
+      "grad_norm": 1.2544962167739868,
+      "learning_rate": 6.610910578826708e-05,
+      "loss": 2.4125,
+      "step": 9921
+    },
+    {
+      "epoch": 0.6103592519685039,
+      "grad_norm": 1.1365222930908203,
+      "learning_rate": 6.60909132014166e-05,
+      "loss": 2.5054,
+      "step": 9922
+    },
+    {
+      "epoch": 0.6104207677165354,
+      "grad_norm": 1.158895492553711,
+      "learning_rate": 6.607272188257687e-05,
+      "loss": 2.7694,
+      "step": 9923
+    },
+    {
+      "epoch": 0.6104822834645669,
+      "grad_norm": 1.305262804031372,
+      "learning_rate": 6.60545318324282e-05,
+      "loss": 2.2157,
+      "step": 9924
+    },
+    {
+      "epoch": 0.6105437992125984,
+      "grad_norm": 1.3269683122634888,
+      "learning_rate": 6.603634305165067e-05,
+      "loss": 2.6283,
+      "step": 9925
+    },
+    {
+      "epoch": 0.61060531496063,
+      "grad_norm": 1.2068876028060913,
+      "learning_rate": 6.60181555409246e-05,
+      "loss": 2.3154,
+      "step": 9926
+    },
+    {
+      "epoch": 0.6106668307086615,
+      "grad_norm": 1.3783241510391235,
+      "learning_rate": 6.599996930092997e-05,
+      "loss": 2.0196,
+      "step": 9927
+    },
+    {
+      "epoch": 0.6107283464566929,
+      "grad_norm": 1.1323922872543335,
+      "learning_rate": 6.598178433234694e-05,
+      "loss": 3.1019,
+      "step": 9928
+    },
+    {
+      "epoch": 0.6107898622047244,
+      "grad_norm": 1.0198924541473389,
+      "learning_rate": 6.596360063585546e-05,
+      "loss": 3.4072,
+      "step": 9929
+    },
+    {
+      "epoch": 0.6108513779527559,
+      "grad_norm": 1.140397310256958,
+      "learning_rate": 6.594541821213556e-05,
+      "loss": 2.4506,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6109128937007874,
+      "grad_norm": 1.187270164489746,
+      "learning_rate": 6.592723706186709e-05,
+      "loss": 2.514,
+      "step": 9931
+    },
+    {
+      "epoch": 0.6109744094488189,
+      "grad_norm": 1.1521977186203003,
+      "learning_rate": 6.590905718572997e-05,
+      "loss": 2.9674,
+      "step": 9932
+    },
+    {
+      "epoch": 0.6110359251968503,
+      "grad_norm": 1.3418937921524048,
+      "learning_rate": 6.5890878584404e-05,
+      "loss": 2.4367,
+      "step": 9933
+    },
+    {
+      "epoch": 0.6110974409448819,
+      "grad_norm": 1.1673429012298584,
+      "learning_rate": 6.587270125856902e-05,
+      "loss": 2.3442,
+      "step": 9934
+    },
+    {
+      "epoch": 0.6111589566929134,
+      "grad_norm": 1.2690478563308716,
+      "learning_rate": 6.585452520890468e-05,
+      "loss": 2.963,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6112204724409449,
+      "grad_norm": 1.401421070098877,
+      "learning_rate": 6.58363504360907e-05,
+      "loss": 3.1228,
+      "step": 9936
+    },
+    {
+      "epoch": 0.6112819881889764,
+      "grad_norm": 1.2472777366638184,
+      "learning_rate": 6.581817694080669e-05,
+      "loss": 3.0542,
+      "step": 9937
+    },
+    {
+      "epoch": 0.6113435039370079,
+      "grad_norm": 1.1277310848236084,
+      "learning_rate": 6.580000472373227e-05,
+      "loss": 2.559,
+      "step": 9938
+    },
+    {
+      "epoch": 0.6114050196850394,
+      "grad_norm": 1.3178750276565552,
+      "learning_rate": 6.5781833785547e-05,
+      "loss": 2.5867,
+      "step": 9939
+    },
+    {
+      "epoch": 0.6114665354330708,
+      "grad_norm": 1.4806267023086548,
+      "learning_rate": 6.576366412693032e-05,
+      "loss": 2.0764,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6115280511811023,
+      "grad_norm": 1.2648476362228394,
+      "learning_rate": 6.57454957485617e-05,
+      "loss": 2.7435,
+      "step": 9941
+    },
+    {
+      "epoch": 0.6115895669291339,
+      "grad_norm": 1.2418053150177002,
+      "learning_rate": 6.57273286511205e-05,
+      "loss": 2.7222,
+      "step": 9942
+    },
+    {
+      "epoch": 0.6116510826771654,
+      "grad_norm": 1.0629422664642334,
+      "learning_rate": 6.570916283528615e-05,
+      "loss": 2.4852,
+      "step": 9943
+    },
+    {
+      "epoch": 0.6117125984251969,
+      "grad_norm": 1.2950987815856934,
+      "learning_rate": 6.569099830173784e-05,
+      "loss": 3.3434,
+      "step": 9944
+    },
+    {
+      "epoch": 0.6117741141732284,
+      "grad_norm": 1.6810799837112427,
+      "learning_rate": 6.567283505115494e-05,
+      "loss": 2.5752,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6118356299212598,
+      "grad_norm": 1.3515088558197021,
+      "learning_rate": 6.565467308421656e-05,
+      "loss": 2.2207,
+      "step": 9946
+    },
+    {
+      "epoch": 0.6118971456692913,
+      "grad_norm": 1.201960802078247,
+      "learning_rate": 6.56365124016019e-05,
+      "loss": 1.7079,
+      "step": 9947
+    },
+    {
+      "epoch": 0.6119586614173228,
+      "grad_norm": 1.2055708169937134,
+      "learning_rate": 6.561835300399005e-05,
+      "loss": 2.4095,
+      "step": 9948
+    },
+    {
+      "epoch": 0.6120201771653543,
+      "grad_norm": 1.3365743160247803,
+      "learning_rate": 6.560019489206011e-05,
+      "loss": 2.6103,
+      "step": 9949
+    },
+    {
+      "epoch": 0.6120816929133859,
+      "grad_norm": 0.9585121273994446,
+      "learning_rate": 6.558203806649104e-05,
+      "loss": 2.917,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6121432086614174,
+      "grad_norm": 1.3141615390777588,
+      "learning_rate": 6.556388252796182e-05,
+      "loss": 2.6614,
+      "step": 9951
+    },
+    {
+      "epoch": 0.6122047244094488,
+      "grad_norm": 1.2913241386413574,
+      "learning_rate": 6.554572827715138e-05,
+      "loss": 1.8855,
+      "step": 9952
+    },
+    {
+      "epoch": 0.6122662401574803,
+      "grad_norm": 1.0286751985549927,
+      "learning_rate": 6.552757531473858e-05,
+      "loss": 1.6959,
+      "step": 9953
+    },
+    {
+      "epoch": 0.6123277559055118,
+      "grad_norm": 1.6366677284240723,
+      "learning_rate": 6.550942364140223e-05,
+      "loss": 1.9125,
+      "step": 9954
+    },
+    {
+      "epoch": 0.6123892716535433,
+      "grad_norm": 1.3765082359313965,
+      "learning_rate": 6.549127325782116e-05,
+      "loss": 1.9519,
+      "step": 9955
+    },
+    {
+      "epoch": 0.6124507874015748,
+      "grad_norm": 1.268033742904663,
+      "learning_rate": 6.547312416467399e-05,
+      "loss": 2.5659,
+      "step": 9956
+    },
+    {
+      "epoch": 0.6125123031496063,
+      "grad_norm": 1.1905689239501953,
+      "learning_rate": 6.545497636263946e-05,
+      "loss": 2.6337,
+      "step": 9957
+    },
+    {
+      "epoch": 0.6125738188976378,
+      "grad_norm": 1.2558722496032715,
+      "learning_rate": 6.543682985239619e-05,
+      "loss": 2.0806,
+      "step": 9958
+    },
+    {
+      "epoch": 0.6126353346456693,
+      "grad_norm": 1.1983689069747925,
+      "learning_rate": 6.541868463462277e-05,
+      "loss": 2.8654,
+      "step": 9959
+    },
+    {
+      "epoch": 0.6126968503937008,
+      "grad_norm": 1.1447417736053467,
+      "learning_rate": 6.54005407099977e-05,
+      "loss": 2.8489,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6127583661417323,
+      "grad_norm": 1.2398898601531982,
+      "learning_rate": 6.538239807919948e-05,
+      "loss": 2.6072,
+      "step": 9961
+    },
+    {
+      "epoch": 0.6128198818897638,
+      "grad_norm": 1.3114826679229736,
+      "learning_rate": 6.536425674290651e-05,
+      "loss": 1.5905,
+      "step": 9962
+    },
+    {
+      "epoch": 0.6128813976377953,
+      "grad_norm": 1.4007800817489624,
+      "learning_rate": 6.534611670179724e-05,
+      "loss": 2.4417,
+      "step": 9963
+    },
+    {
+      "epoch": 0.6129429133858267,
+      "grad_norm": 1.1505769491195679,
+      "learning_rate": 6.532797795654995e-05,
+      "loss": 2.6702,
+      "step": 9964
+    },
+    {
+      "epoch": 0.6130044291338582,
+      "grad_norm": 1.2526730298995972,
+      "learning_rate": 6.530984050784298e-05,
+      "loss": 2.5346,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6130659448818898,
+      "grad_norm": 1.4007632732391357,
+      "learning_rate": 6.52917043563545e-05,
+      "loss": 2.5696,
+      "step": 9966
+    },
+    {
+      "epoch": 0.6131274606299213,
+      "grad_norm": 1.2367206811904907,
+      "learning_rate": 6.527356950276275e-05,
+      "loss": 2.1768,
+      "step": 9967
+    },
+    {
+      "epoch": 0.6131889763779528,
+      "grad_norm": 1.1425297260284424,
+      "learning_rate": 6.525543594774586e-05,
+      "loss": 2.2477,
+      "step": 9968
+    },
+    {
+      "epoch": 0.6132504921259843,
+      "grad_norm": 1.4670515060424805,
+      "learning_rate": 6.523730369198195e-05,
+      "loss": 2.556,
+      "step": 9969
+    },
+    {
+      "epoch": 0.6133120078740157,
+      "grad_norm": 1.3531266450881958,
+      "learning_rate": 6.521917273614897e-05,
+      "loss": 2.3117,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6133735236220472,
+      "grad_norm": 1.2766920328140259,
+      "learning_rate": 6.520104308092507e-05,
+      "loss": 2.1203,
+      "step": 9971
+    },
+    {
+      "epoch": 0.6134350393700787,
+      "grad_norm": 1.135575532913208,
+      "learning_rate": 6.518291472698807e-05,
+      "loss": 2.9463,
+      "step": 9972
+    },
+    {
+      "epoch": 0.6134965551181102,
+      "grad_norm": 1.2094404697418213,
+      "learning_rate": 6.516478767501593e-05,
+      "loss": 2.3618,
+      "step": 9973
+    },
+    {
+      "epoch": 0.6135580708661418,
+      "grad_norm": 1.2912994623184204,
+      "learning_rate": 6.514666192568649e-05,
+      "loss": 2.2121,
+      "step": 9974
+    },
+    {
+      "epoch": 0.6136195866141733,
+      "grad_norm": 1.2794997692108154,
+      "learning_rate": 6.512853747967752e-05,
+      "loss": 2.2857,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6136811023622047,
+      "grad_norm": 1.4437661170959473,
+      "learning_rate": 6.511041433766683e-05,
+      "loss": 1.7745,
+      "step": 9976
+    },
+    {
+      "epoch": 0.6137426181102362,
+      "grad_norm": 2.6816375255584717,
+      "learning_rate": 6.509229250033204e-05,
+      "loss": 2.1755,
+      "step": 9977
+    },
+    {
+      "epoch": 0.6138041338582677,
+      "grad_norm": 1.5832775831222534,
+      "learning_rate": 6.507417196835093e-05,
+      "loss": 2.4872,
+      "step": 9978
+    },
+    {
+      "epoch": 0.6138656496062992,
+      "grad_norm": 1.3269991874694824,
+      "learning_rate": 6.505605274240096e-05,
+      "loss": 2.751,
+      "step": 9979
+    },
+    {
+      "epoch": 0.6139271653543307,
+      "grad_norm": 1.1965776681900024,
+      "learning_rate": 6.503793482315985e-05,
+      "loss": 1.9833,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6139886811023622,
+      "grad_norm": 1.1406358480453491,
+      "learning_rate": 6.501981821130497e-05,
+      "loss": 2.0865,
+      "step": 9981
+    },
+    {
+      "epoch": 0.6140501968503937,
+      "grad_norm": 1.130051612854004,
+      "learning_rate": 6.500170290751383e-05,
+      "loss": 2.0815,
+      "step": 9982
+    },
+    {
+      "epoch": 0.6141117125984252,
+      "grad_norm": 1.5882502794265747,
+      "learning_rate": 6.498358891246385e-05,
+      "loss": 2.4462,
+      "step": 9983
+    },
+    {
+      "epoch": 0.6141732283464567,
+      "grad_norm": 1.3367186784744263,
+      "learning_rate": 6.496547622683244e-05,
+      "loss": 2.4553,
+      "step": 9984
+    },
+    {
+      "epoch": 0.6142347440944882,
+      "grad_norm": 1.3846070766448975,
+      "learning_rate": 6.49473648512968e-05,
+      "loss": 2.7128,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6142962598425197,
+      "grad_norm": 1.5313129425048828,
+      "learning_rate": 6.492925478653429e-05,
+      "loss": 2.5011,
+      "step": 9986
+    },
+    {
+      "epoch": 0.6143577755905512,
+      "grad_norm": 1.1601166725158691,
+      "learning_rate": 6.491114603322206e-05,
+      "loss": 2.9627,
+      "step": 9987
+    },
+    {
+      "epoch": 0.6144192913385826,
+      "grad_norm": 1.1552687883377075,
+      "learning_rate": 6.489303859203732e-05,
+      "loss": 2.3291,
+      "step": 9988
+    },
+    {
+      "epoch": 0.6144808070866141,
+      "grad_norm": 1.2493581771850586,
+      "learning_rate": 6.487493246365717e-05,
+      "loss": 2.7818,
+      "step": 9989
+    },
+    {
+      "epoch": 0.6145423228346457,
+      "grad_norm": 1.003635287284851,
+      "learning_rate": 6.485682764875871e-05,
+      "loss": 2.1517,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6146038385826772,
+      "grad_norm": 1.468835473060608,
+      "learning_rate": 6.48387241480189e-05,
+      "loss": 2.2885,
+      "step": 9991
+    },
+    {
+      "epoch": 0.6146653543307087,
+      "grad_norm": 1.4445658922195435,
+      "learning_rate": 6.482062196211477e-05,
+      "loss": 2.6572,
+      "step": 9992
+    },
+    {
+      "epoch": 0.6147268700787402,
+      "grad_norm": 2.472860097885132,
+      "learning_rate": 6.480252109172318e-05,
+      "loss": 2.9957,
+      "step": 9993
+    },
+    {
+      "epoch": 0.6147883858267716,
+      "grad_norm": 1.0906403064727783,
+      "learning_rate": 6.478442153752108e-05,
+      "loss": 1.5956,
+      "step": 9994
+    },
+    {
+      "epoch": 0.6148499015748031,
+      "grad_norm": 1.1635187864303589,
+      "learning_rate": 6.476632330018522e-05,
+      "loss": 2.4317,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6149114173228346,
+      "grad_norm": 1.2242778539657593,
+      "learning_rate": 6.47482263803924e-05,
+      "loss": 2.2348,
+      "step": 9996
+    },
+    {
+      "epoch": 0.6149729330708661,
+      "grad_norm": 1.1635382175445557,
+      "learning_rate": 6.473013077881933e-05,
+      "loss": 2.5587,
+      "step": 9997
+    },
+    {
+      "epoch": 0.6150344488188977,
+      "grad_norm": 1.1727509498596191,
+      "learning_rate": 6.471203649614272e-05,
+      "loss": 2.5755,
+      "step": 9998
+    },
+    {
+      "epoch": 0.6150959645669292,
+      "grad_norm": 1.3178962469100952,
+      "learning_rate": 6.469394353303915e-05,
+      "loss": 2.4219,
+      "step": 9999
+    },
+    {
+      "epoch": 0.6151574803149606,
+      "grad_norm": 1.6078696250915527,
+      "learning_rate": 6.467585189018527e-05,
+      "loss": 2.1316,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6152189960629921,
+      "grad_norm": 1.4673467874526978,
+      "learning_rate": 6.465776156825751e-05,
+      "loss": 2.1503,
+      "step": 10001
+    },
+    {
+      "epoch": 0.6152805118110236,
+      "grad_norm": 1.0599462985992432,
+      "learning_rate": 6.463967256793242e-05,
+      "loss": 2.6423,
+      "step": 10002
+    },
+    {
+      "epoch": 0.6153420275590551,
+      "grad_norm": 1.0258418321609497,
+      "learning_rate": 6.462158488988636e-05,
+      "loss": 2.8419,
+      "step": 10003
+    },
+    {
+      "epoch": 0.6154035433070866,
+      "grad_norm": 1.2376627922058105,
+      "learning_rate": 6.460349853479582e-05,
+      "loss": 2.3234,
+      "step": 10004
+    },
+    {
+      "epoch": 0.6154650590551181,
+      "grad_norm": 1.4908173084259033,
+      "learning_rate": 6.458541350333698e-05,
+      "loss": 3.0512,
+      "step": 10005
+    },
+    {
+      "epoch": 0.6155265748031497,
+      "grad_norm": 1.435834527015686,
+      "learning_rate": 6.456732979618626e-05,
+      "loss": 1.8099,
+      "step": 10006
+    },
+    {
+      "epoch": 0.6155880905511811,
+      "grad_norm": 2.8630428314208984,
+      "learning_rate": 6.454924741401978e-05,
+      "loss": 3.3542,
+      "step": 10007
+    },
+    {
+      "epoch": 0.6156496062992126,
+      "grad_norm": 1.1984026432037354,
+      "learning_rate": 6.45311663575138e-05,
+      "loss": 3.5731,
+      "step": 10008
+    },
+    {
+      "epoch": 0.6157111220472441,
+      "grad_norm": 1.2768272161483765,
+      "learning_rate": 6.451308662734443e-05,
+      "loss": 2.5504,
+      "step": 10009
+    },
+    {
+      "epoch": 0.6157726377952756,
+      "grad_norm": 1.3978346586227417,
+      "learning_rate": 6.449500822418771e-05,
+      "loss": 2.7397,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6158341535433071,
+      "grad_norm": 1.1426795721054077,
+      "learning_rate": 6.447693114871972e-05,
+      "loss": 2.5183,
+      "step": 10011
+    },
+    {
+      "epoch": 0.6158956692913385,
+      "grad_norm": 1.0217516422271729,
+      "learning_rate": 6.44588554016164e-05,
+      "loss": 2.6334,
+      "step": 10012
+    },
+    {
+      "epoch": 0.61595718503937,
+      "grad_norm": 1.1943399906158447,
+      "learning_rate": 6.444078098355375e-05,
+      "loss": 2.2052,
+      "step": 10013
+    },
+    {
+      "epoch": 0.6160187007874016,
+      "grad_norm": 1.4352926015853882,
+      "learning_rate": 6.442270789520754e-05,
+      "loss": 2.0499,
+      "step": 10014
+    },
+    {
+      "epoch": 0.6160802165354331,
+      "grad_norm": 1.367427945137024,
+      "learning_rate": 6.440463613725374e-05,
+      "loss": 2.6974,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6161417322834646,
+      "grad_norm": 1.5337677001953125,
+      "learning_rate": 6.438656571036801e-05,
+      "loss": 1.8093,
+      "step": 10016
+    },
+    {
+      "epoch": 0.6162032480314961,
+      "grad_norm": 1.0856999158859253,
+      "learning_rate": 6.436849661522616e-05,
+      "loss": 3.1556,
+      "step": 10017
+    },
+    {
+      "epoch": 0.6162647637795275,
+      "grad_norm": 1.111114501953125,
+      "learning_rate": 6.435042885250383e-05,
+      "loss": 2.3869,
+      "step": 10018
+    },
+    {
+      "epoch": 0.616326279527559,
+      "grad_norm": 1.2178068161010742,
+      "learning_rate": 6.433236242287671e-05,
+      "loss": 2.89,
+      "step": 10019
+    },
+    {
+      "epoch": 0.6163877952755905,
+      "grad_norm": 1.192901849746704,
+      "learning_rate": 6.431429732702033e-05,
+      "loss": 2.22,
+      "step": 10020
+    },
+    {
+      "epoch": 0.616449311023622,
+      "grad_norm": 1.346298336982727,
+      "learning_rate": 6.429623356561023e-05,
+      "loss": 2.1953,
+      "step": 10021
+    },
+    {
+      "epoch": 0.6165108267716536,
+      "grad_norm": 1.6151511669158936,
+      "learning_rate": 6.42781711393219e-05,
+      "loss": 2.2082,
+      "step": 10022
+    },
+    {
+      "epoch": 0.6165723425196851,
+      "grad_norm": 1.5129948854446411,
+      "learning_rate": 6.426011004883079e-05,
+      "loss": 2.3638,
+      "step": 10023
+    },
+    {
+      "epoch": 0.6166338582677166,
+      "grad_norm": 1.1338258981704712,
+      "learning_rate": 6.424205029481227e-05,
+      "loss": 2.8681,
+      "step": 10024
+    },
+    {
+      "epoch": 0.616695374015748,
+      "grad_norm": 1.3467066287994385,
+      "learning_rate": 6.42239918779417e-05,
+      "loss": 2.3743,
+      "step": 10025
+    },
+    {
+      "epoch": 0.6167568897637795,
+      "grad_norm": 1.0533477067947388,
+      "learning_rate": 6.420593479889432e-05,
+      "loss": 2.5505,
+      "step": 10026
+    },
+    {
+      "epoch": 0.616818405511811,
+      "grad_norm": 1.3526017665863037,
+      "learning_rate": 6.418787905834539e-05,
+      "loss": 1.7107,
+      "step": 10027
+    },
+    {
+      "epoch": 0.6168799212598425,
+      "grad_norm": 1.1077406406402588,
+      "learning_rate": 6.416982465697007e-05,
+      "loss": 2.8591,
+      "step": 10028
+    },
+    {
+      "epoch": 0.616941437007874,
+      "grad_norm": 1.2821557521820068,
+      "learning_rate": 6.415177159544356e-05,
+      "loss": 2.6681,
+      "step": 10029
+    },
+    {
+      "epoch": 0.6170029527559056,
+      "grad_norm": 1.2309085130691528,
+      "learning_rate": 6.413371987444085e-05,
+      "loss": 2.1786,
+      "step": 10030
+    },
+    {
+      "epoch": 0.617064468503937,
+      "grad_norm": 1.6199880838394165,
+      "learning_rate": 6.411566949463705e-05,
+      "loss": 1.9468,
+      "step": 10031
+    },
+    {
+      "epoch": 0.6171259842519685,
+      "grad_norm": 1.500494122505188,
+      "learning_rate": 6.409762045670708e-05,
+      "loss": 2.7167,
+      "step": 10032
+    },
+    {
+      "epoch": 0.6171875,
+      "grad_norm": 1.1370658874511719,
+      "learning_rate": 6.407957276132595e-05,
+      "loss": 2.7296,
+      "step": 10033
+    },
+    {
+      "epoch": 0.6172490157480315,
+      "grad_norm": 1.2411946058273315,
+      "learning_rate": 6.406152640916848e-05,
+      "loss": 2.3601,
+      "step": 10034
+    },
+    {
+      "epoch": 0.617310531496063,
+      "grad_norm": 1.499547004699707,
+      "learning_rate": 6.404348140090954e-05,
+      "loss": 2.3312,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6173720472440944,
+      "grad_norm": 1.3036158084869385,
+      "learning_rate": 6.402543773722387e-05,
+      "loss": 2.6413,
+      "step": 10036
+    },
+    {
+      "epoch": 0.617433562992126,
+      "grad_norm": 1.377973198890686,
+      "learning_rate": 6.400739541878624e-05,
+      "loss": 2.3004,
+      "step": 10037
+    },
+    {
+      "epoch": 0.6174950787401575,
+      "grad_norm": 1.3145301342010498,
+      "learning_rate": 6.398935444627131e-05,
+      "loss": 1.9756,
+      "step": 10038
+    },
+    {
+      "epoch": 0.617556594488189,
+      "grad_norm": 1.2318575382232666,
+      "learning_rate": 6.397131482035376e-05,
+      "loss": 2.8086,
+      "step": 10039
+    },
+    {
+      "epoch": 0.6176181102362205,
+      "grad_norm": 1.3628387451171875,
+      "learning_rate": 6.395327654170809e-05,
+      "loss": 2.335,
+      "step": 10040
+    },
+    {
+      "epoch": 0.617679625984252,
+      "grad_norm": 1.1208727359771729,
+      "learning_rate": 6.393523961100891e-05,
+      "loss": 2.658,
+      "step": 10041
+    },
+    {
+      "epoch": 0.6177411417322834,
+      "grad_norm": 1.2011529207229614,
+      "learning_rate": 6.391720402893065e-05,
+      "loss": 2.7714,
+      "step": 10042
+    },
+    {
+      "epoch": 0.6178026574803149,
+      "grad_norm": 1.4932416677474976,
+      "learning_rate": 6.389916979614776e-05,
+      "loss": 2.8353,
+      "step": 10043
+    },
+    {
+      "epoch": 0.6178641732283464,
+      "grad_norm": 1.3902095556259155,
+      "learning_rate": 6.38811369133346e-05,
+      "loss": 2.9183,
+      "step": 10044
+    },
+    {
+      "epoch": 0.617925688976378,
+      "grad_norm": 1.226865530014038,
+      "learning_rate": 6.386310538116554e-05,
+      "loss": 2.3419,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6179872047244095,
+      "grad_norm": 1.158708095550537,
+      "learning_rate": 6.384507520031483e-05,
+      "loss": 2.774,
+      "step": 10046
+    },
+    {
+      "epoch": 0.618048720472441,
+      "grad_norm": 1.185225248336792,
+      "learning_rate": 6.38270463714567e-05,
+      "loss": 2.4296,
+      "step": 10047
+    },
+    {
+      "epoch": 0.6181102362204725,
+      "grad_norm": 1.222567081451416,
+      "learning_rate": 6.380901889526534e-05,
+      "loss": 2.3837,
+      "step": 10048
+    },
+    {
+      "epoch": 0.6181717519685039,
+      "grad_norm": 1.1994961500167847,
+      "learning_rate": 6.37909927724148e-05,
+      "loss": 3.1437,
+      "step": 10049
+    },
+    {
+      "epoch": 0.6182332677165354,
+      "grad_norm": 1.1933703422546387,
+      "learning_rate": 6.37729680035793e-05,
+      "loss": 2.6135,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6182947834645669,
+      "grad_norm": 1.2592458724975586,
+      "learning_rate": 6.375494458943277e-05,
+      "loss": 2.559,
+      "step": 10051
+    },
+    {
+      "epoch": 0.6183562992125984,
+      "grad_norm": 1.3141868114471436,
+      "learning_rate": 6.373692253064919e-05,
+      "loss": 2.5128,
+      "step": 10052
+    },
+    {
+      "epoch": 0.61841781496063,
+      "grad_norm": 1.302083969116211,
+      "learning_rate": 6.37189018279025e-05,
+      "loss": 2.4997,
+      "step": 10053
+    },
+    {
+      "epoch": 0.6184793307086615,
+      "grad_norm": 1.3207964897155762,
+      "learning_rate": 6.370088248186661e-05,
+      "loss": 2.8673,
+      "step": 10054
+    },
+    {
+      "epoch": 0.6185408464566929,
+      "grad_norm": 1.402989387512207,
+      "learning_rate": 6.368286449321526e-05,
+      "loss": 2.7289,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6186023622047244,
+      "grad_norm": 1.1700905561447144,
+      "learning_rate": 6.366484786262228e-05,
+      "loss": 2.5756,
+      "step": 10056
+    },
+    {
+      "epoch": 0.6186638779527559,
+      "grad_norm": 1.4201911687850952,
+      "learning_rate": 6.364683259076136e-05,
+      "loss": 2.1808,
+      "step": 10057
+    },
+    {
+      "epoch": 0.6187253937007874,
+      "grad_norm": 1.4372082948684692,
+      "learning_rate": 6.362881867830625e-05,
+      "loss": 2.4462,
+      "step": 10058
+    },
+    {
+      "epoch": 0.6187869094488189,
+      "grad_norm": 1.3192102909088135,
+      "learning_rate": 6.361080612593045e-05,
+      "loss": 2.4766,
+      "step": 10059
+    },
+    {
+      "epoch": 0.6188484251968503,
+      "grad_norm": 1.0897926092147827,
+      "learning_rate": 6.359279493430763e-05,
+      "loss": 3.2522,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6189099409448819,
+      "grad_norm": 1.4265435934066772,
+      "learning_rate": 6.357478510411124e-05,
+      "loss": 1.9142,
+      "step": 10061
+    },
+    {
+      "epoch": 0.6189714566929134,
+      "grad_norm": 1.1115761995315552,
+      "learning_rate": 6.355677663601478e-05,
+      "loss": 2.8257,
+      "step": 10062
+    },
+    {
+      "epoch": 0.6190329724409449,
+      "grad_norm": 1.3028048276901245,
+      "learning_rate": 6.353876953069164e-05,
+      "loss": 2.1055,
+      "step": 10063
+    },
+    {
+      "epoch": 0.6190944881889764,
+      "grad_norm": 1.085206151008606,
+      "learning_rate": 6.352076378881526e-05,
+      "loss": 2.1712,
+      "step": 10064
+    },
+    {
+      "epoch": 0.6191560039370079,
+      "grad_norm": 1.1967097520828247,
+      "learning_rate": 6.350275941105886e-05,
+      "loss": 2.3922,
+      "step": 10065
+    },
+    {
+      "epoch": 0.6192175196850394,
+      "grad_norm": 1.7705422639846802,
+      "learning_rate": 6.348475639809577e-05,
+      "loss": 2.5146,
+      "step": 10066
+    },
+    {
+      "epoch": 0.6192790354330708,
+      "grad_norm": 1.6953725814819336,
+      "learning_rate": 6.346675475059914e-05,
+      "loss": 2.3483,
+      "step": 10067
+    },
+    {
+      "epoch": 0.6193405511811023,
+      "grad_norm": 1.5565510988235474,
+      "learning_rate": 6.34487544692422e-05,
+      "loss": 2.7917,
+      "step": 10068
+    },
+    {
+      "epoch": 0.6194020669291339,
+      "grad_norm": 1.3432745933532715,
+      "learning_rate": 6.3430755554698e-05,
+      "loss": 1.8481,
+      "step": 10069
+    },
+    {
+      "epoch": 0.6194635826771654,
+      "grad_norm": 1.1567538976669312,
+      "learning_rate": 6.341275800763967e-05,
+      "loss": 2.7181,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6195250984251969,
+      "grad_norm": 1.2993537187576294,
+      "learning_rate": 6.339476182874013e-05,
+      "loss": 2.2262,
+      "step": 10071
+    },
+    {
+      "epoch": 0.6195866141732284,
+      "grad_norm": 1.1579804420471191,
+      "learning_rate": 6.337676701867241e-05,
+      "loss": 2.1768,
+      "step": 10072
+    },
+    {
+      "epoch": 0.6196481299212598,
+      "grad_norm": 1.937639594078064,
+      "learning_rate": 6.335877357810936e-05,
+      "loss": 2.2075,
+      "step": 10073
+    },
+    {
+      "epoch": 0.6197096456692913,
+      "grad_norm": 1.1906492710113525,
+      "learning_rate": 6.334078150772392e-05,
+      "loss": 1.9022,
+      "step": 10074
+    },
+    {
+      "epoch": 0.6197711614173228,
+      "grad_norm": 1.097245693206787,
+      "learning_rate": 6.332279080818875e-05,
+      "loss": 2.8971,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6198326771653543,
+      "grad_norm": 1.2933012247085571,
+      "learning_rate": 6.330480148017674e-05,
+      "loss": 2.4479,
+      "step": 10076
+    },
+    {
+      "epoch": 0.6198941929133859,
+      "grad_norm": 1.5928488969802856,
+      "learning_rate": 6.328681352436052e-05,
+      "loss": 2.3167,
+      "step": 10077
+    },
+    {
+      "epoch": 0.6199557086614174,
+      "grad_norm": 1.4319413900375366,
+      "learning_rate": 6.326882694141276e-05,
+      "loss": 2.3211,
+      "step": 10078
+    },
+    {
+      "epoch": 0.6200172244094488,
+      "grad_norm": 1.1178529262542725,
+      "learning_rate": 6.325084173200604e-05,
+      "loss": 2.0707,
+      "step": 10079
+    },
+    {
+      "epoch": 0.6200787401574803,
+      "grad_norm": 1.2462244033813477,
+      "learning_rate": 6.323285789681296e-05,
+      "loss": 2.4926,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6201402559055118,
+      "grad_norm": 1.0790820121765137,
+      "learning_rate": 6.321487543650593e-05,
+      "loss": 2.7833,
+      "step": 10081
+    },
+    {
+      "epoch": 0.6202017716535433,
+      "grad_norm": 1.418547511100769,
+      "learning_rate": 6.319689435175746e-05,
+      "loss": 2.4098,
+      "step": 10082
+    },
+    {
+      "epoch": 0.6202632874015748,
+      "grad_norm": 1.2943347692489624,
+      "learning_rate": 6.317891464323995e-05,
+      "loss": 2.4024,
+      "step": 10083
+    },
+    {
+      "epoch": 0.6203248031496063,
+      "grad_norm": 1.0284454822540283,
+      "learning_rate": 6.316093631162567e-05,
+      "loss": 2.9419,
+      "step": 10084
+    },
+    {
+      "epoch": 0.6203863188976378,
+      "grad_norm": 1.1266297101974487,
+      "learning_rate": 6.314295935758697e-05,
+      "loss": 2.9281,
+      "step": 10085
+    },
+    {
+      "epoch": 0.6204478346456693,
+      "grad_norm": 1.6275001764297485,
+      "learning_rate": 6.312498378179606e-05,
+      "loss": 2.5633,
+      "step": 10086
+    },
+    {
+      "epoch": 0.6205093503937008,
+      "grad_norm": 1.1478339433670044,
+      "learning_rate": 6.310700958492514e-05,
+      "loss": 2.5928,
+      "step": 10087
+    },
+    {
+      "epoch": 0.6205708661417323,
+      "grad_norm": 1.1452122926712036,
+      "learning_rate": 6.308903676764632e-05,
+      "loss": 2.144,
+      "step": 10088
+    },
+    {
+      "epoch": 0.6206323818897638,
+      "grad_norm": 1.3413805961608887,
+      "learning_rate": 6.307106533063175e-05,
+      "loss": 2.8429,
+      "step": 10089
+    },
+    {
+      "epoch": 0.6206938976377953,
+      "grad_norm": 1.2152105569839478,
+      "learning_rate": 6.305309527455338e-05,
+      "loss": 2.7019,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6207554133858267,
+      "grad_norm": 1.2290918827056885,
+      "learning_rate": 6.303512660008324e-05,
+      "loss": 2.2294,
+      "step": 10091
+    },
+    {
+      "epoch": 0.6208169291338582,
+      "grad_norm": 1.2166019678115845,
+      "learning_rate": 6.301715930789321e-05,
+      "loss": 2.9022,
+      "step": 10092
+    },
+    {
+      "epoch": 0.6208784448818898,
+      "grad_norm": 1.2473539113998413,
+      "learning_rate": 6.299919339865526e-05,
+      "loss": 2.5436,
+      "step": 10093
+    },
+    {
+      "epoch": 0.6209399606299213,
+      "grad_norm": 1.3066595792770386,
+      "learning_rate": 6.298122887304108e-05,
+      "loss": 2.4947,
+      "step": 10094
+    },
+    {
+      "epoch": 0.6210014763779528,
+      "grad_norm": 1.311134696006775,
+      "learning_rate": 6.29632657317226e-05,
+      "loss": 2.4676,
+      "step": 10095
+    },
+    {
+      "epoch": 0.6210629921259843,
+      "grad_norm": 1.243613839149475,
+      "learning_rate": 6.294530397537139e-05,
+      "loss": 3.0632,
+      "step": 10096
+    },
+    {
+      "epoch": 0.6211245078740157,
+      "grad_norm": 1.12485671043396,
+      "learning_rate": 6.292734360465923e-05,
+      "loss": 1.9561,
+      "step": 10097
+    },
+    {
+      "epoch": 0.6211860236220472,
+      "grad_norm": 1.2850180864334106,
+      "learning_rate": 6.290938462025766e-05,
+      "loss": 2.2571,
+      "step": 10098
+    },
+    {
+      "epoch": 0.6212475393700787,
+      "grad_norm": 1.2819117307662964,
+      "learning_rate": 6.289142702283835e-05,
+      "loss": 2.7698,
+      "step": 10099
+    },
+    {
+      "epoch": 0.6213090551181102,
+      "grad_norm": 1.2213656902313232,
+      "learning_rate": 6.287347081307267e-05,
+      "loss": 2.3563,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6213705708661418,
+      "grad_norm": 1.4798129796981812,
+      "learning_rate": 6.285551599163221e-05,
+      "loss": 2.277,
+      "step": 10101
+    },
+    {
+      "epoch": 0.6214320866141733,
+      "grad_norm": 1.5503888130187988,
+      "learning_rate": 6.283756255918829e-05,
+      "loss": 1.5057,
+      "step": 10102
+    },
+    {
+      "epoch": 0.6214936023622047,
+      "grad_norm": 1.3612425327301025,
+      "learning_rate": 6.281961051641234e-05,
+      "loss": 2.374,
+      "step": 10103
+    },
+    {
+      "epoch": 0.6215551181102362,
+      "grad_norm": 2.305363416671753,
+      "learning_rate": 6.280165986397559e-05,
+      "loss": 1.8085,
+      "step": 10104
+    },
+    {
+      "epoch": 0.6216166338582677,
+      "grad_norm": 1.3427214622497559,
+      "learning_rate": 6.278371060254939e-05,
+      "loss": 3.4096,
+      "step": 10105
+    },
+    {
+      "epoch": 0.6216781496062992,
+      "grad_norm": 1.2374521493911743,
+      "learning_rate": 6.276576273280485e-05,
+      "loss": 3.3801,
+      "step": 10106
+    },
+    {
+      "epoch": 0.6217396653543307,
+      "grad_norm": 1.249342918395996,
+      "learning_rate": 6.27478162554132e-05,
+      "loss": 2.3176,
+      "step": 10107
+    },
+    {
+      "epoch": 0.6218011811023622,
+      "grad_norm": 1.1801801919937134,
+      "learning_rate": 6.272987117104545e-05,
+      "loss": 1.9743,
+      "step": 10108
+    },
+    {
+      "epoch": 0.6218626968503937,
+      "grad_norm": 1.3358277082443237,
+      "learning_rate": 6.271192748037275e-05,
+      "loss": 2.3235,
+      "step": 10109
+    },
+    {
+      "epoch": 0.6219242125984252,
+      "grad_norm": 1.2548657655715942,
+      "learning_rate": 6.269398518406599e-05,
+      "loss": 2.614,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6219857283464567,
+      "grad_norm": 1.0892750024795532,
+      "learning_rate": 6.267604428279619e-05,
+      "loss": 2.5528,
+      "step": 10111
+    },
+    {
+      "epoch": 0.6220472440944882,
+      "grad_norm": 1.1972535848617554,
+      "learning_rate": 6.265810477723417e-05,
+      "loss": 2.068,
+      "step": 10112
+    },
+    {
+      "epoch": 0.6221087598425197,
+      "grad_norm": 1.2442843914031982,
+      "learning_rate": 6.264016666805084e-05,
+      "loss": 2.4085,
+      "step": 10113
+    },
+    {
+      "epoch": 0.6221702755905512,
+      "grad_norm": 1.1631054878234863,
+      "learning_rate": 6.262222995591694e-05,
+      "loss": 2.57,
+      "step": 10114
+    },
+    {
+      "epoch": 0.6222317913385826,
+      "grad_norm": 1.442786693572998,
+      "learning_rate": 6.260429464150325e-05,
+      "loss": 2.2125,
+      "step": 10115
+    },
+    {
+      "epoch": 0.6222933070866141,
+      "grad_norm": 1.3744157552719116,
+      "learning_rate": 6.258636072548037e-05,
+      "loss": 2.8222,
+      "step": 10116
+    },
+    {
+      "epoch": 0.6223548228346457,
+      "grad_norm": 1.1757303476333618,
+      "learning_rate": 6.2568428208519e-05,
+      "loss": 2.4475,
+      "step": 10117
+    },
+    {
+      "epoch": 0.6224163385826772,
+      "grad_norm": 1.3509361743927002,
+      "learning_rate": 6.255049709128971e-05,
+      "loss": 2.8164,
+      "step": 10118
+    },
+    {
+      "epoch": 0.6224778543307087,
+      "grad_norm": 1.0567899942398071,
+      "learning_rate": 6.253256737446295e-05,
+      "loss": 2.5298,
+      "step": 10119
+    },
+    {
+      "epoch": 0.6225393700787402,
+      "grad_norm": 1.310828685760498,
+      "learning_rate": 6.251463905870929e-05,
+      "loss": 2.3584,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6226008858267716,
+      "grad_norm": 1.0736228227615356,
+      "learning_rate": 6.249671214469908e-05,
+      "loss": 2.7919,
+      "step": 10121
+    },
+    {
+      "epoch": 0.6226624015748031,
+      "grad_norm": 1.2445387840270996,
+      "learning_rate": 6.247878663310274e-05,
+      "loss": 2.4302,
+      "step": 10122
+    },
+    {
+      "epoch": 0.6227239173228346,
+      "grad_norm": 1.8830958604812622,
+      "learning_rate": 6.246086252459051e-05,
+      "loss": 1.893,
+      "step": 10123
+    },
+    {
+      "epoch": 0.6227854330708661,
+      "grad_norm": 1.2238677740097046,
+      "learning_rate": 6.244293981983275e-05,
+      "loss": 2.3166,
+      "step": 10124
+    },
+    {
+      "epoch": 0.6228469488188977,
+      "grad_norm": 1.003596305847168,
+      "learning_rate": 6.242501851949957e-05,
+      "loss": 2.5859,
+      "step": 10125
+    },
+    {
+      "epoch": 0.6229084645669292,
+      "grad_norm": 1.3009426593780518,
+      "learning_rate": 6.240709862426122e-05,
+      "loss": 2.3511,
+      "step": 10126
+    },
+    {
+      "epoch": 0.6229699803149606,
+      "grad_norm": 1.217091679573059,
+      "learning_rate": 6.238918013478772e-05,
+      "loss": 2.8183,
+      "step": 10127
+    },
+    {
+      "epoch": 0.6230314960629921,
+      "grad_norm": 1.4321101903915405,
+      "learning_rate": 6.237126305174918e-05,
+      "loss": 2.7488,
+      "step": 10128
+    },
+    {
+      "epoch": 0.6230930118110236,
+      "grad_norm": 1.6164016723632812,
+      "learning_rate": 6.235334737581555e-05,
+      "loss": 2.2657,
+      "step": 10129
+    },
+    {
+      "epoch": 0.6231545275590551,
+      "grad_norm": 1.207777500152588,
+      "learning_rate": 6.233543310765685e-05,
+      "loss": 2.8305,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6232160433070866,
+      "grad_norm": 1.220777153968811,
+      "learning_rate": 6.231752024794289e-05,
+      "loss": 2.3884,
+      "step": 10131
+    },
+    {
+      "epoch": 0.6232775590551181,
+      "grad_norm": 1.1685165166854858,
+      "learning_rate": 6.229960879734356e-05,
+      "loss": 2.3146,
+      "step": 10132
+    },
+    {
+      "epoch": 0.6233390748031497,
+      "grad_norm": 1.308658242225647,
+      "learning_rate": 6.228169875652863e-05,
+      "loss": 2.8546,
+      "step": 10133
+    },
+    {
+      "epoch": 0.6234005905511811,
+      "grad_norm": 1.260244607925415,
+      "learning_rate": 6.226379012616787e-05,
+      "loss": 2.9341,
+      "step": 10134
+    },
+    {
+      "epoch": 0.6234621062992126,
+      "grad_norm": 1.4029947519302368,
+      "learning_rate": 6.224588290693091e-05,
+      "loss": 2.1899,
+      "step": 10135
+    },
+    {
+      "epoch": 0.6235236220472441,
+      "grad_norm": 1.2456897497177124,
+      "learning_rate": 6.222797709948743e-05,
+      "loss": 2.1087,
+      "step": 10136
+    },
+    {
+      "epoch": 0.6235851377952756,
+      "grad_norm": 1.3476418256759644,
+      "learning_rate": 6.221007270450695e-05,
+      "loss": 2.2591,
+      "step": 10137
+    },
+    {
+      "epoch": 0.6236466535433071,
+      "grad_norm": 1.23239004611969,
+      "learning_rate": 6.219216972265907e-05,
+      "loss": 2.7253,
+      "step": 10138
+    },
+    {
+      "epoch": 0.6237081692913385,
+      "grad_norm": 1.3404265642166138,
+      "learning_rate": 6.217426815461318e-05,
+      "loss": 2.173,
+      "step": 10139
+    },
+    {
+      "epoch": 0.62376968503937,
+      "grad_norm": 1.4532690048217773,
+      "learning_rate": 6.215636800103879e-05,
+      "loss": 2.5224,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6238312007874016,
+      "grad_norm": 1.3185724020004272,
+      "learning_rate": 6.213846926260515e-05,
+      "loss": 1.9225,
+      "step": 10141
+    },
+    {
+      "epoch": 0.6238927165354331,
+      "grad_norm": 1.1841901540756226,
+      "learning_rate": 6.212057193998168e-05,
+      "loss": 1.9475,
+      "step": 10142
+    },
+    {
+      "epoch": 0.6239542322834646,
+      "grad_norm": 1.244384765625,
+      "learning_rate": 6.210267603383759e-05,
+      "loss": 2.3302,
+      "step": 10143
+    },
+    {
+      "epoch": 0.6240157480314961,
+      "grad_norm": 1.3590806722640991,
+      "learning_rate": 6.208478154484212e-05,
+      "loss": 2.9686,
+      "step": 10144
+    },
+    {
+      "epoch": 0.6240772637795275,
+      "grad_norm": 1.25499427318573,
+      "learning_rate": 6.206688847366435e-05,
+      "loss": 2.8475,
+      "step": 10145
+    },
+    {
+      "epoch": 0.624138779527559,
+      "grad_norm": 1.2252644300460815,
+      "learning_rate": 6.204899682097347e-05,
+      "loss": 2.5287,
+      "step": 10146
+    },
+    {
+      "epoch": 0.6242002952755905,
+      "grad_norm": 1.3025792837142944,
+      "learning_rate": 6.203110658743845e-05,
+      "loss": 2.4578,
+      "step": 10147
+    },
+    {
+      "epoch": 0.624261811023622,
+      "grad_norm": 1.2601529359817505,
+      "learning_rate": 6.201321777372835e-05,
+      "loss": 2.1232,
+      "step": 10148
+    },
+    {
+      "epoch": 0.6243233267716536,
+      "grad_norm": 1.3341368436813354,
+      "learning_rate": 6.199533038051206e-05,
+      "loss": 2.7184,
+      "step": 10149
+    },
+    {
+      "epoch": 0.6243848425196851,
+      "grad_norm": 2.2852389812469482,
+      "learning_rate": 6.197744440845855e-05,
+      "loss": 2.0007,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6244463582677166,
+      "grad_norm": 1.4149699211120605,
+      "learning_rate": 6.195955985823655e-05,
+      "loss": 3.5628,
+      "step": 10151
+    },
+    {
+      "epoch": 0.624507874015748,
+      "grad_norm": 1.281107783317566,
+      "learning_rate": 6.194167673051492e-05,
+      "loss": 2.6774,
+      "step": 10152
+    },
+    {
+      "epoch": 0.6245693897637795,
+      "grad_norm": 1.3444494009017944,
+      "learning_rate": 6.192379502596232e-05,
+      "loss": 2.79,
+      "step": 10153
+    },
+    {
+      "epoch": 0.624630905511811,
+      "grad_norm": 1.3526204824447632,
+      "learning_rate": 6.190591474524752e-05,
+      "loss": 2.7461,
+      "step": 10154
+    },
+    {
+      "epoch": 0.6246924212598425,
+      "grad_norm": 1.4579297304153442,
+      "learning_rate": 6.188803588903907e-05,
+      "loss": 2.6898,
+      "step": 10155
+    },
+    {
+      "epoch": 0.624753937007874,
+      "grad_norm": 1.1271418333053589,
+      "learning_rate": 6.187015845800555e-05,
+      "loss": 2.6736,
+      "step": 10156
+    },
+    {
+      "epoch": 0.6248154527559056,
+      "grad_norm": 1.0747624635696411,
+      "learning_rate": 6.18522824528155e-05,
+      "loss": 2.0629,
+      "step": 10157
+    },
+    {
+      "epoch": 0.624876968503937,
+      "grad_norm": 1.3723021745681763,
+      "learning_rate": 6.183440787413734e-05,
+      "loss": 2.1446,
+      "step": 10158
+    },
+    {
+      "epoch": 0.6249384842519685,
+      "grad_norm": 1.1472758054733276,
+      "learning_rate": 6.181653472263956e-05,
+      "loss": 2.8575,
+      "step": 10159
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 1.3141944408416748,
+      "learning_rate": 6.17986629989904e-05,
+      "loss": 2.6771,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6250615157480315,
+      "grad_norm": 1.4066745042800903,
+      "learning_rate": 6.178079270385828e-05,
+      "loss": 3.3003,
+      "step": 10161
+    },
+    {
+      "epoch": 0.625123031496063,
+      "grad_norm": 1.4480763673782349,
+      "learning_rate": 6.176292383791136e-05,
+      "loss": 2.4467,
+      "step": 10162
+    },
+    {
+      "epoch": 0.6251845472440944,
+      "grad_norm": 1.1592953205108643,
+      "learning_rate": 6.174505640181791e-05,
+      "loss": 2.6876,
+      "step": 10163
+    },
+    {
+      "epoch": 0.625246062992126,
+      "grad_norm": 1.289048194885254,
+      "learning_rate": 6.172719039624598e-05,
+      "loss": 2.218,
+      "step": 10164
+    },
+    {
+      "epoch": 0.6253075787401575,
+      "grad_norm": 2.2846672534942627,
+      "learning_rate": 6.170932582186377e-05,
+      "loss": 1.9576,
+      "step": 10165
+    },
+    {
+      "epoch": 0.625369094488189,
+      "grad_norm": 1.4060072898864746,
+      "learning_rate": 6.169146267933923e-05,
+      "loss": 2.5511,
+      "step": 10166
+    },
+    {
+      "epoch": 0.6254306102362205,
+      "grad_norm": 1.2106257677078247,
+      "learning_rate": 6.167360096934039e-05,
+      "loss": 3.1052,
+      "step": 10167
+    },
+    {
+      "epoch": 0.625492125984252,
+      "grad_norm": 1.2178758382797241,
+      "learning_rate": 6.165574069253513e-05,
+      "loss": 2.7933,
+      "step": 10168
+    },
+    {
+      "epoch": 0.6255536417322834,
+      "grad_norm": 1.3241297006607056,
+      "learning_rate": 6.163788184959139e-05,
+      "loss": 2.5665,
+      "step": 10169
+    },
+    {
+      "epoch": 0.6256151574803149,
+      "grad_norm": 1.1468652486801147,
+      "learning_rate": 6.162002444117694e-05,
+      "loss": 2.4387,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6256766732283464,
+      "grad_norm": 1.1631616353988647,
+      "learning_rate": 6.160216846795956e-05,
+      "loss": 2.6992,
+      "step": 10171
+    },
+    {
+      "epoch": 0.625738188976378,
+      "grad_norm": 1.122131586074829,
+      "learning_rate": 6.158431393060695e-05,
+      "loss": 2.4965,
+      "step": 10172
+    },
+    {
+      "epoch": 0.6257997047244095,
+      "grad_norm": 1.193292498588562,
+      "learning_rate": 6.15664608297868e-05,
+      "loss": 2.0971,
+      "step": 10173
+    },
+    {
+      "epoch": 0.625861220472441,
+      "grad_norm": 1.2983843088150024,
+      "learning_rate": 6.154860916616672e-05,
+      "loss": 2.8474,
+      "step": 10174
+    },
+    {
+      "epoch": 0.6259227362204725,
+      "grad_norm": 1.158073902130127,
+      "learning_rate": 6.153075894041426e-05,
+      "loss": 2.4626,
+      "step": 10175
+    },
+    {
+      "epoch": 0.6259842519685039,
+      "grad_norm": 1.2244696617126465,
+      "learning_rate": 6.151291015319687e-05,
+      "loss": 2.5891,
+      "step": 10176
+    },
+    {
+      "epoch": 0.6260457677165354,
+      "grad_norm": 1.2899560928344727,
+      "learning_rate": 6.149506280518204e-05,
+      "loss": 2.2274,
+      "step": 10177
+    },
+    {
+      "epoch": 0.6261072834645669,
+      "grad_norm": 1.3660080432891846,
+      "learning_rate": 6.147721689703714e-05,
+      "loss": 2.8642,
+      "step": 10178
+    },
+    {
+      "epoch": 0.6261687992125984,
+      "grad_norm": 1.5656416416168213,
+      "learning_rate": 6.145937242942956e-05,
+      "loss": 2.4694,
+      "step": 10179
+    },
+    {
+      "epoch": 0.62623031496063,
+      "grad_norm": 1.2641816139221191,
+      "learning_rate": 6.14415294030265e-05,
+      "loss": 2.2185,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6262918307086615,
+      "grad_norm": 1.3015506267547607,
+      "learning_rate": 6.142368781849526e-05,
+      "loss": 2.5361,
+      "step": 10181
+    },
+    {
+      "epoch": 0.6263533464566929,
+      "grad_norm": 1.2133718729019165,
+      "learning_rate": 6.140584767650296e-05,
+      "loss": 3.0053,
+      "step": 10182
+    },
+    {
+      "epoch": 0.6264148622047244,
+      "grad_norm": 1.3628414869308472,
+      "learning_rate": 6.138800897771677e-05,
+      "loss": 2.3437,
+      "step": 10183
+    },
+    {
+      "epoch": 0.6264763779527559,
+      "grad_norm": 1.1454676389694214,
+      "learning_rate": 6.137017172280373e-05,
+      "loss": 2.6077,
+      "step": 10184
+    },
+    {
+      "epoch": 0.6265378937007874,
+      "grad_norm": 1.3676772117614746,
+      "learning_rate": 6.135233591243088e-05,
+      "loss": 2.2069,
+      "step": 10185
+    },
+    {
+      "epoch": 0.6265994094488189,
+      "grad_norm": 1.1011106967926025,
+      "learning_rate": 6.133450154726514e-05,
+      "loss": 3.1453,
+      "step": 10186
+    },
+    {
+      "epoch": 0.6266609251968503,
+      "grad_norm": 1.9241080284118652,
+      "learning_rate": 6.131666862797346e-05,
+      "loss": 3.4232,
+      "step": 10187
+    },
+    {
+      "epoch": 0.6267224409448819,
+      "grad_norm": 1.2994272708892822,
+      "learning_rate": 6.129883715522263e-05,
+      "loss": 2.3664,
+      "step": 10188
+    },
+    {
+      "epoch": 0.6267839566929134,
+      "grad_norm": 1.1599706411361694,
+      "learning_rate": 6.128100712967955e-05,
+      "loss": 2.4738,
+      "step": 10189
+    },
+    {
+      "epoch": 0.6268454724409449,
+      "grad_norm": 1.2325810194015503,
+      "learning_rate": 6.126317855201081e-05,
+      "loss": 2.7739,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6269069881889764,
+      "grad_norm": 1.3120455741882324,
+      "learning_rate": 6.124535142288328e-05,
+      "loss": 2.5228,
+      "step": 10191
+    },
+    {
+      "epoch": 0.6269685039370079,
+      "grad_norm": 1.1382724046707153,
+      "learning_rate": 6.122752574296346e-05,
+      "loss": 2.4078,
+      "step": 10192
+    },
+    {
+      "epoch": 0.6270300196850394,
+      "grad_norm": 1.3179192543029785,
+      "learning_rate": 6.120970151291798e-05,
+      "loss": 2.5259,
+      "step": 10193
+    },
+    {
+      "epoch": 0.6270915354330708,
+      "grad_norm": 1.1968183517456055,
+      "learning_rate": 6.119187873341339e-05,
+      "loss": 2.5189,
+      "step": 10194
+    },
+    {
+      "epoch": 0.6271530511811023,
+      "grad_norm": 1.187404990196228,
+      "learning_rate": 6.11740574051161e-05,
+      "loss": 2.9735,
+      "step": 10195
+    },
+    {
+      "epoch": 0.6272145669291339,
+      "grad_norm": 1.4889875650405884,
+      "learning_rate": 6.115623752869258e-05,
+      "loss": 1.2535,
+      "step": 10196
+    },
+    {
+      "epoch": 0.6272760826771654,
+      "grad_norm": 1.1879204511642456,
+      "learning_rate": 6.113841910480915e-05,
+      "loss": 2.4349,
+      "step": 10197
+    },
+    {
+      "epoch": 0.6273375984251969,
+      "grad_norm": 1.1211365461349487,
+      "learning_rate": 6.11206021341322e-05,
+      "loss": 2.4725,
+      "step": 10198
+    },
+    {
+      "epoch": 0.6273991141732284,
+      "grad_norm": 1.2725521326065063,
+      "learning_rate": 6.110278661732787e-05,
+      "loss": 2.2356,
+      "step": 10199
+    },
+    {
+      "epoch": 0.6274606299212598,
+      "grad_norm": 1.4264994859695435,
+      "learning_rate": 6.108497255506247e-05,
+      "loss": 2.4428,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6275221456692913,
+      "grad_norm": 1.4899061918258667,
+      "learning_rate": 6.106715994800207e-05,
+      "loss": 2.5351,
+      "step": 10201
+    },
+    {
+      "epoch": 0.6275836614173228,
+      "grad_norm": 1.0473905801773071,
+      "learning_rate": 6.104934879681281e-05,
+      "loss": 2.8321,
+      "step": 10202
+    },
+    {
+      "epoch": 0.6276451771653543,
+      "grad_norm": 1.298926591873169,
+      "learning_rate": 6.1031539102160705e-05,
+      "loss": 2.5632,
+      "step": 10203
+    },
+    {
+      "epoch": 0.6277066929133859,
+      "grad_norm": 1.2749563455581665,
+      "learning_rate": 6.101373086471177e-05,
+      "loss": 2.1845,
+      "step": 10204
+    },
+    {
+      "epoch": 0.6277682086614174,
+      "grad_norm": 1.2950206995010376,
+      "learning_rate": 6.099592408513186e-05,
+      "loss": 2.5532,
+      "step": 10205
+    },
+    {
+      "epoch": 0.6278297244094488,
+      "grad_norm": 1.328344464302063,
+      "learning_rate": 6.097811876408694e-05,
+      "loss": 2.9033,
+      "step": 10206
+    },
+    {
+      "epoch": 0.6278912401574803,
+      "grad_norm": 1.3610998392105103,
+      "learning_rate": 6.096031490224274e-05,
+      "loss": 2.7806,
+      "step": 10207
+    },
+    {
+      "epoch": 0.6279527559055118,
+      "grad_norm": 1.2104934453964233,
+      "learning_rate": 6.094251250026513e-05,
+      "loss": 2.7558,
+      "step": 10208
+    },
+    {
+      "epoch": 0.6280142716535433,
+      "grad_norm": 1.181870937347412,
+      "learning_rate": 6.0924711558819694e-05,
+      "loss": 2.8655,
+      "step": 10209
+    },
+    {
+      "epoch": 0.6280757874015748,
+      "grad_norm": 1.429458498954773,
+      "learning_rate": 6.090691207857222e-05,
+      "loss": 2.6002,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6281373031496063,
+      "grad_norm": 1.2025614976882935,
+      "learning_rate": 6.08891140601882e-05,
+      "loss": 2.7802,
+      "step": 10211
+    },
+    {
+      "epoch": 0.6281988188976378,
+      "grad_norm": 1.1285308599472046,
+      "learning_rate": 6.087131750433326e-05,
+      "loss": 2.1994,
+      "step": 10212
+    },
+    {
+      "epoch": 0.6282603346456693,
+      "grad_norm": 1.332010269165039,
+      "learning_rate": 6.0853522411672815e-05,
+      "loss": 2.5753,
+      "step": 10213
+    },
+    {
+      "epoch": 0.6283218503937008,
+      "grad_norm": 1.114488959312439,
+      "learning_rate": 6.0835728782872405e-05,
+      "loss": 2.8084,
+      "step": 10214
+    },
+    {
+      "epoch": 0.6283833661417323,
+      "grad_norm": 1.405869960784912,
+      "learning_rate": 6.081793661859731e-05,
+      "loss": 2.8119,
+      "step": 10215
+    },
+    {
+      "epoch": 0.6284448818897638,
+      "grad_norm": 1.181693434715271,
+      "learning_rate": 6.080014591951291e-05,
+      "loss": 2.4295,
+      "step": 10216
+    },
+    {
+      "epoch": 0.6285063976377953,
+      "grad_norm": 1.0984737873077393,
+      "learning_rate": 6.0782356686284446e-05,
+      "loss": 2.7529,
+      "step": 10217
+    },
+    {
+      "epoch": 0.6285679133858267,
+      "grad_norm": 1.2532286643981934,
+      "learning_rate": 6.076456891957719e-05,
+      "loss": 2.485,
+      "step": 10218
+    },
+    {
+      "epoch": 0.6286294291338582,
+      "grad_norm": 1.072227120399475,
+      "learning_rate": 6.0746782620056256e-05,
+      "loss": 2.7494,
+      "step": 10219
+    },
+    {
+      "epoch": 0.6286909448818898,
+      "grad_norm": 1.2854969501495361,
+      "learning_rate": 6.07289977883868e-05,
+      "loss": 2.6703,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6287524606299213,
+      "grad_norm": 1.1064329147338867,
+      "learning_rate": 6.071121442523381e-05,
+      "loss": 1.8869,
+      "step": 10221
+    },
+    {
+      "epoch": 0.6288139763779528,
+      "grad_norm": 1.4013066291809082,
+      "learning_rate": 6.069343253126234e-05,
+      "loss": 2.6983,
+      "step": 10222
+    },
+    {
+      "epoch": 0.6288754921259843,
+      "grad_norm": 1.4874142408370972,
+      "learning_rate": 6.067565210713728e-05,
+      "loss": 2.3124,
+      "step": 10223
+    },
+    {
+      "epoch": 0.6289370078740157,
+      "grad_norm": 1.2347391843795776,
+      "learning_rate": 6.065787315352362e-05,
+      "loss": 2.9404,
+      "step": 10224
+    },
+    {
+      "epoch": 0.6289985236220472,
+      "grad_norm": 1.1643022298812866,
+      "learning_rate": 6.064009567108604e-05,
+      "loss": 2.083,
+      "step": 10225
+    },
+    {
+      "epoch": 0.6290600393700787,
+      "grad_norm": 1.2757688760757446,
+      "learning_rate": 6.062231966048948e-05,
+      "loss": 2.0693,
+      "step": 10226
+    },
+    {
+      "epoch": 0.6291215551181102,
+      "grad_norm": 1.3381280899047852,
+      "learning_rate": 6.0604545122398546e-05,
+      "loss": 2.7734,
+      "step": 10227
+    },
+    {
+      "epoch": 0.6291830708661418,
+      "grad_norm": 1.1808091402053833,
+      "learning_rate": 6.058677205747797e-05,
+      "loss": 2.9015,
+      "step": 10228
+    },
+    {
+      "epoch": 0.6292445866141733,
+      "grad_norm": 1.3489655256271362,
+      "learning_rate": 6.056900046639237e-05,
+      "loss": 2.1407,
+      "step": 10229
+    },
+    {
+      "epoch": 0.6293061023622047,
+      "grad_norm": 1.2871475219726562,
+      "learning_rate": 6.055123034980623e-05,
+      "loss": 2.2043,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6293676181102362,
+      "grad_norm": 1.230087399482727,
+      "learning_rate": 6.053346170838412e-05,
+      "loss": 2.5549,
+      "step": 10231
+    },
+    {
+      "epoch": 0.6294291338582677,
+      "grad_norm": 1.4666649103164673,
+      "learning_rate": 6.051569454279046e-05,
+      "loss": 2.1608,
+      "step": 10232
+    },
+    {
+      "epoch": 0.6294906496062992,
+      "grad_norm": 2.0833566188812256,
+      "learning_rate": 6.049792885368969e-05,
+      "loss": 2.9383,
+      "step": 10233
+    },
+    {
+      "epoch": 0.6295521653543307,
+      "grad_norm": 0.9998828172683716,
+      "learning_rate": 6.0480164641746075e-05,
+      "loss": 2.6978,
+      "step": 10234
+    },
+    {
+      "epoch": 0.6296136811023622,
+      "grad_norm": 1.3620706796646118,
+      "learning_rate": 6.0462401907623947e-05,
+      "loss": 2.4862,
+      "step": 10235
+    },
+    {
+      "epoch": 0.6296751968503937,
+      "grad_norm": 1.0233407020568848,
+      "learning_rate": 6.0444640651987494e-05,
+      "loss": 2.7126,
+      "step": 10236
+    },
+    {
+      "epoch": 0.6297367125984252,
+      "grad_norm": 1.0879510641098022,
+      "learning_rate": 6.0426880875500946e-05,
+      "loss": 2.4925,
+      "step": 10237
+    },
+    {
+      "epoch": 0.6297982283464567,
+      "grad_norm": 1.2622991800308228,
+      "learning_rate": 6.040912257882837e-05,
+      "loss": 2.6215,
+      "step": 10238
+    },
+    {
+      "epoch": 0.6298597440944882,
+      "grad_norm": 1.6257309913635254,
+      "learning_rate": 6.0391365762633876e-05,
+      "loss": 2.6098,
+      "step": 10239
+    },
+    {
+      "epoch": 0.6299212598425197,
+      "grad_norm": 1.2026573419570923,
+      "learning_rate": 6.0373610427581407e-05,
+      "loss": 2.0734,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6299827755905512,
+      "grad_norm": 1.1668815612792969,
+      "learning_rate": 6.0355856574334954e-05,
+      "loss": 2.0799,
+      "step": 10241
+    },
+    {
+      "epoch": 0.6300442913385826,
+      "grad_norm": 1.0896989107131958,
+      "learning_rate": 6.033810420355839e-05,
+      "loss": 1.9723,
+      "step": 10242
+    },
+    {
+      "epoch": 0.6301058070866141,
+      "grad_norm": 1.2751038074493408,
+      "learning_rate": 6.032035331591561e-05,
+      "loss": 2.5133,
+      "step": 10243
+    },
+    {
+      "epoch": 0.6301673228346457,
+      "grad_norm": 1.0980479717254639,
+      "learning_rate": 6.03026039120703e-05,
+      "loss": 2.7715,
+      "step": 10244
+    },
+    {
+      "epoch": 0.6302288385826772,
+      "grad_norm": 1.2263882160186768,
+      "learning_rate": 6.028485599268629e-05,
+      "loss": 2.2104,
+      "step": 10245
+    },
+    {
+      "epoch": 0.6302903543307087,
+      "grad_norm": 1.3599485158920288,
+      "learning_rate": 6.026710955842718e-05,
+      "loss": 2.8206,
+      "step": 10246
+    },
+    {
+      "epoch": 0.6303518700787402,
+      "grad_norm": 1.2272576093673706,
+      "learning_rate": 6.0249364609956626e-05,
+      "loss": 3.0584,
+      "step": 10247
+    },
+    {
+      "epoch": 0.6304133858267716,
+      "grad_norm": 1.0615452527999878,
+      "learning_rate": 6.0231621147938165e-05,
+      "loss": 3.0438,
+      "step": 10248
+    },
+    {
+      "epoch": 0.6304749015748031,
+      "grad_norm": 1.1737464666366577,
+      "learning_rate": 6.021387917303536e-05,
+      "loss": 2.5469,
+      "step": 10249
+    },
+    {
+      "epoch": 0.6305364173228346,
+      "grad_norm": 1.2753177881240845,
+      "learning_rate": 6.019613868591158e-05,
+      "loss": 2.2276,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6305979330708661,
+      "grad_norm": 1.6283141374588013,
+      "learning_rate": 6.017839968723029e-05,
+      "loss": 2.3862,
+      "step": 10251
+    },
+    {
+      "epoch": 0.6306594488188977,
+      "grad_norm": 1.2563461065292358,
+      "learning_rate": 6.016066217765477e-05,
+      "loss": 2.5013,
+      "step": 10252
+    },
+    {
+      "epoch": 0.6307209645669292,
+      "grad_norm": 1.2668206691741943,
+      "learning_rate": 6.014292615784836e-05,
+      "loss": 2.359,
+      "step": 10253
+    },
+    {
+      "epoch": 0.6307824803149606,
+      "grad_norm": 1.0796971321105957,
+      "learning_rate": 6.012519162847423e-05,
+      "loss": 2.129,
+      "step": 10254
+    },
+    {
+      "epoch": 0.6308439960629921,
+      "grad_norm": 1.4074152708053589,
+      "learning_rate": 6.010745859019564e-05,
+      "loss": 3.6063,
+      "step": 10255
+    },
+    {
+      "epoch": 0.6309055118110236,
+      "grad_norm": 1.3130453824996948,
+      "learning_rate": 6.008972704367559e-05,
+      "loss": 1.5237,
+      "step": 10256
+    },
+    {
+      "epoch": 0.6309670275590551,
+      "grad_norm": 1.059128999710083,
+      "learning_rate": 6.007199698957725e-05,
+      "loss": 2.7724,
+      "step": 10257
+    },
+    {
+      "epoch": 0.6310285433070866,
+      "grad_norm": 1.2578731775283813,
+      "learning_rate": 6.005426842856354e-05,
+      "loss": 2.7258,
+      "step": 10258
+    },
+    {
+      "epoch": 0.6310900590551181,
+      "grad_norm": 1.5343773365020752,
+      "learning_rate": 6.0036541361297505e-05,
+      "loss": 2.7799,
+      "step": 10259
+    },
+    {
+      "epoch": 0.6311515748031497,
+      "grad_norm": 1.7473505735397339,
+      "learning_rate": 6.0018815788441926e-05,
+      "loss": 1.5177,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6312130905511811,
+      "grad_norm": 1.0876039266586304,
+      "learning_rate": 6.000109171065971e-05,
+      "loss": 3.0033,
+      "step": 10261
+    },
+    {
+      "epoch": 0.6312746062992126,
+      "grad_norm": 1.530023455619812,
+      "learning_rate": 5.998336912861362e-05,
+      "loss": 1.9594,
+      "step": 10262
+    },
+    {
+      "epoch": 0.6313361220472441,
+      "grad_norm": 1.448938250541687,
+      "learning_rate": 5.9965648042966395e-05,
+      "loss": 2.745,
+      "step": 10263
+    },
+    {
+      "epoch": 0.6313976377952756,
+      "grad_norm": 1.1306142807006836,
+      "learning_rate": 5.994792845438067e-05,
+      "loss": 2.5552,
+      "step": 10264
+    },
+    {
+      "epoch": 0.6314591535433071,
+      "grad_norm": 1.541654348373413,
+      "learning_rate": 5.9930210363519136e-05,
+      "loss": 2.5373,
+      "step": 10265
+    },
+    {
+      "epoch": 0.6315206692913385,
+      "grad_norm": 1.0811362266540527,
+      "learning_rate": 5.991249377104429e-05,
+      "loss": 2.8377,
+      "step": 10266
+    },
+    {
+      "epoch": 0.63158218503937,
+      "grad_norm": 1.1200770139694214,
+      "learning_rate": 5.989477867761859e-05,
+      "loss": 2.2964,
+      "step": 10267
+    },
+    {
+      "epoch": 0.6316437007874016,
+      "grad_norm": 1.0180352926254272,
+      "learning_rate": 5.9877065083904604e-05,
+      "loss": 2.9685,
+      "step": 10268
+    },
+    {
+      "epoch": 0.6317052165354331,
+      "grad_norm": 1.1554234027862549,
+      "learning_rate": 5.98593529905646e-05,
+      "loss": 2.4324,
+      "step": 10269
+    },
+    {
+      "epoch": 0.6317667322834646,
+      "grad_norm": 1.4335999488830566,
+      "learning_rate": 5.9841642398260975e-05,
+      "loss": 2.8386,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6318282480314961,
+      "grad_norm": 1.2605304718017578,
+      "learning_rate": 5.9823933307655985e-05,
+      "loss": 3.1491,
+      "step": 10271
+    },
+    {
+      "epoch": 0.6318897637795275,
+      "grad_norm": 1.2434344291687012,
+      "learning_rate": 5.980622571941189e-05,
+      "loss": 1.6335,
+      "step": 10272
+    },
+    {
+      "epoch": 0.631951279527559,
+      "grad_norm": 1.185629963874817,
+      "learning_rate": 5.97885196341908e-05,
+      "loss": 2.9325,
+      "step": 10273
+    },
+    {
+      "epoch": 0.6320127952755905,
+      "grad_norm": 1.2775521278381348,
+      "learning_rate": 5.9770815052654897e-05,
+      "loss": 2.4387,
+      "step": 10274
+    },
+    {
+      "epoch": 0.632074311023622,
+      "grad_norm": 1.3654481172561646,
+      "learning_rate": 5.975311197546613e-05,
+      "loss": 2.986,
+      "step": 10275
+    },
+    {
+      "epoch": 0.6321358267716536,
+      "grad_norm": 1.165680170059204,
+      "learning_rate": 5.973541040328659e-05,
+      "loss": 2.4384,
+      "step": 10276
+    },
+    {
+      "epoch": 0.6321973425196851,
+      "grad_norm": 1.4631868600845337,
+      "learning_rate": 5.9717710336778145e-05,
+      "loss": 2.3532,
+      "step": 10277
+    },
+    {
+      "epoch": 0.6322588582677166,
+      "grad_norm": 1.348726749420166,
+      "learning_rate": 5.970001177660276e-05,
+      "loss": 2.3424,
+      "step": 10278
+    },
+    {
+      "epoch": 0.632320374015748,
+      "grad_norm": 2.151646137237549,
+      "learning_rate": 5.968231472342215e-05,
+      "loss": 2.3778,
+      "step": 10279
+    },
+    {
+      "epoch": 0.6323818897637795,
+      "grad_norm": 1.7059216499328613,
+      "learning_rate": 5.966461917789822e-05,
+      "loss": 2.1209,
+      "step": 10280
+    },
+    {
+      "epoch": 0.632443405511811,
+      "grad_norm": 1.1370270252227783,
+      "learning_rate": 5.964692514069257e-05,
+      "loss": 2.965,
+      "step": 10281
+    },
+    {
+      "epoch": 0.6325049212598425,
+      "grad_norm": 1.2065664529800415,
+      "learning_rate": 5.962923261246692e-05,
+      "loss": 3.0768,
+      "step": 10282
+    },
+    {
+      "epoch": 0.632566437007874,
+      "grad_norm": 1.3268829584121704,
+      "learning_rate": 5.961154159388285e-05,
+      "loss": 2.5429,
+      "step": 10283
+    },
+    {
+      "epoch": 0.6326279527559056,
+      "grad_norm": 1.0364468097686768,
+      "learning_rate": 5.959385208560195e-05,
+      "loss": 2.7072,
+      "step": 10284
+    },
+    {
+      "epoch": 0.632689468503937,
+      "grad_norm": 1.200774908065796,
+      "learning_rate": 5.957616408828562e-05,
+      "loss": 1.8656,
+      "step": 10285
+    },
+    {
+      "epoch": 0.6327509842519685,
+      "grad_norm": 1.150553822517395,
+      "learning_rate": 5.955847760259537e-05,
+      "loss": 2.7571,
+      "step": 10286
+    },
+    {
+      "epoch": 0.6328125,
+      "grad_norm": 1.312976360321045,
+      "learning_rate": 5.9540792629192535e-05,
+      "loss": 2.8346,
+      "step": 10287
+    },
+    {
+      "epoch": 0.6328740157480315,
+      "grad_norm": 1.2798765897750854,
+      "learning_rate": 5.952310916873847e-05,
+      "loss": 2.4275,
+      "step": 10288
+    },
+    {
+      "epoch": 0.632935531496063,
+      "grad_norm": 1.1253478527069092,
+      "learning_rate": 5.9505427221894384e-05,
+      "loss": 2.33,
+      "step": 10289
+    },
+    {
+      "epoch": 0.6329970472440944,
+      "grad_norm": 1.2770934104919434,
+      "learning_rate": 5.9487746789321566e-05,
+      "loss": 2.4826,
+      "step": 10290
+    },
+    {
+      "epoch": 0.633058562992126,
+      "grad_norm": 1.080826997756958,
+      "learning_rate": 5.947006787168108e-05,
+      "loss": 2.6981,
+      "step": 10291
+    },
+    {
+      "epoch": 0.6331200787401575,
+      "grad_norm": 1.2276320457458496,
+      "learning_rate": 5.945239046963407e-05,
+      "loss": 2.8207,
+      "step": 10292
+    },
+    {
+      "epoch": 0.633181594488189,
+      "grad_norm": 1.3139642477035522,
+      "learning_rate": 5.943471458384152e-05,
+      "loss": 1.6957,
+      "step": 10293
+    },
+    {
+      "epoch": 0.6332431102362205,
+      "grad_norm": 1.520217776298523,
+      "learning_rate": 5.94170402149645e-05,
+      "loss": 1.9376,
+      "step": 10294
+    },
+    {
+      "epoch": 0.633304625984252,
+      "grad_norm": 1.1750743389129639,
+      "learning_rate": 5.939936736366385e-05,
+      "loss": 3.1552,
+      "step": 10295
+    },
+    {
+      "epoch": 0.6333661417322834,
+      "grad_norm": 1.0848885774612427,
+      "learning_rate": 5.9381696030600465e-05,
+      "loss": 2.652,
+      "step": 10296
+    },
+    {
+      "epoch": 0.6334276574803149,
+      "grad_norm": 1.359116554260254,
+      "learning_rate": 5.9364026216435154e-05,
+      "loss": 2.6986,
+      "step": 10297
+    },
+    {
+      "epoch": 0.6334891732283464,
+      "grad_norm": 1.535845398902893,
+      "learning_rate": 5.9346357921828674e-05,
+      "loss": 2.7012,
+      "step": 10298
+    },
+    {
+      "epoch": 0.633550688976378,
+      "grad_norm": 1.3191213607788086,
+      "learning_rate": 5.9328691147441706e-05,
+      "loss": 2.5776,
+      "step": 10299
+    },
+    {
+      "epoch": 0.6336122047244095,
+      "grad_norm": 1.373891830444336,
+      "learning_rate": 5.9311025893934934e-05,
+      "loss": 1.7389,
+      "step": 10300
+    },
+    {
+      "epoch": 0.633673720472441,
+      "grad_norm": 1.2382034063339233,
+      "learning_rate": 5.929336216196887e-05,
+      "loss": 2.2505,
+      "step": 10301
+    },
+    {
+      "epoch": 0.6337352362204725,
+      "grad_norm": 1.0997861623764038,
+      "learning_rate": 5.927569995220409e-05,
+      "loss": 2.7288,
+      "step": 10302
+    },
+    {
+      "epoch": 0.6337967519685039,
+      "grad_norm": 1.2652037143707275,
+      "learning_rate": 5.925803926530107e-05,
+      "loss": 2.8437,
+      "step": 10303
+    },
+    {
+      "epoch": 0.6338582677165354,
+      "grad_norm": 1.2900429964065552,
+      "learning_rate": 5.924038010192017e-05,
+      "loss": 2.5992,
+      "step": 10304
+    },
+    {
+      "epoch": 0.6339197834645669,
+      "grad_norm": 1.4565637111663818,
+      "learning_rate": 5.922272246272177e-05,
+      "loss": 1.7199,
+      "step": 10305
+    },
+    {
+      "epoch": 0.6339812992125984,
+      "grad_norm": 1.2098041772842407,
+      "learning_rate": 5.920506634836616e-05,
+      "loss": 2.7492,
+      "step": 10306
+    },
+    {
+      "epoch": 0.63404281496063,
+      "grad_norm": 1.1226601600646973,
+      "learning_rate": 5.918741175951361e-05,
+      "loss": 2.6787,
+      "step": 10307
+    },
+    {
+      "epoch": 0.6341043307086615,
+      "grad_norm": 1.2799534797668457,
+      "learning_rate": 5.916975869682426e-05,
+      "loss": 1.9189,
+      "step": 10308
+    },
+    {
+      "epoch": 0.6341658464566929,
+      "grad_norm": 1.1415210962295532,
+      "learning_rate": 5.91521071609583e-05,
+      "loss": 2.666,
+      "step": 10309
+    },
+    {
+      "epoch": 0.6342273622047244,
+      "grad_norm": 1.1998608112335205,
+      "learning_rate": 5.913445715257573e-05,
+      "loss": 2.486,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6342888779527559,
+      "grad_norm": 1.1692533493041992,
+      "learning_rate": 5.91168086723366e-05,
+      "loss": 2.8295,
+      "step": 10311
+    },
+    {
+      "epoch": 0.6343503937007874,
+      "grad_norm": 1.1678292751312256,
+      "learning_rate": 5.909916172090083e-05,
+      "loss": 2.3887,
+      "step": 10312
+    },
+    {
+      "epoch": 0.6344119094488189,
+      "grad_norm": 1.1897130012512207,
+      "learning_rate": 5.908151629892839e-05,
+      "loss": 2.5936,
+      "step": 10313
+    },
+    {
+      "epoch": 0.6344734251968503,
+      "grad_norm": 1.4545423984527588,
+      "learning_rate": 5.9063872407079e-05,
+      "loss": 2.6088,
+      "step": 10314
+    },
+    {
+      "epoch": 0.6345349409448819,
+      "grad_norm": 1.213422179222107,
+      "learning_rate": 5.90462300460126e-05,
+      "loss": 2.547,
+      "step": 10315
+    },
+    {
+      "epoch": 0.6345964566929134,
+      "grad_norm": 1.2611663341522217,
+      "learning_rate": 5.9028589216388765e-05,
+      "loss": 3.1675,
+      "step": 10316
+    },
+    {
+      "epoch": 0.6346579724409449,
+      "grad_norm": 1.3046282529830933,
+      "learning_rate": 5.9010949918867265e-05,
+      "loss": 2.338,
+      "step": 10317
+    },
+    {
+      "epoch": 0.6347194881889764,
+      "grad_norm": 1.2680141925811768,
+      "learning_rate": 5.899331215410766e-05,
+      "loss": 2.1925,
+      "step": 10318
+    },
+    {
+      "epoch": 0.6347810039370079,
+      "grad_norm": 1.1840208768844604,
+      "learning_rate": 5.8975675922769555e-05,
+      "loss": 2.5032,
+      "step": 10319
+    },
+    {
+      "epoch": 0.6348425196850394,
+      "grad_norm": 1.1632028818130493,
+      "learning_rate": 5.895804122551237e-05,
+      "loss": 2.3797,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6349040354330708,
+      "grad_norm": 1.2038706541061401,
+      "learning_rate": 5.894040806299561e-05,
+      "loss": 2.4064,
+      "step": 10321
+    },
+    {
+      "epoch": 0.6349655511811023,
+      "grad_norm": 1.284420132637024,
+      "learning_rate": 5.892277643587861e-05,
+      "loss": 2.5232,
+      "step": 10322
+    },
+    {
+      "epoch": 0.6350270669291339,
+      "grad_norm": 1.233149766921997,
+      "learning_rate": 5.890514634482074e-05,
+      "loss": 2.0195,
+      "step": 10323
+    },
+    {
+      "epoch": 0.6350885826771654,
+      "grad_norm": 1.1353529691696167,
+      "learning_rate": 5.888751779048123e-05,
+      "loss": 2.8231,
+      "step": 10324
+    },
+    {
+      "epoch": 0.6351500984251969,
+      "grad_norm": 1.1946697235107422,
+      "learning_rate": 5.886989077351934e-05,
+      "loss": 2.3913,
+      "step": 10325
+    },
+    {
+      "epoch": 0.6352116141732284,
+      "grad_norm": 1.2025212049484253,
+      "learning_rate": 5.885226529459415e-05,
+      "loss": 3.06,
+      "step": 10326
+    },
+    {
+      "epoch": 0.6352731299212598,
+      "grad_norm": 1.4875926971435547,
+      "learning_rate": 5.8834641354364825e-05,
+      "loss": 2.0856,
+      "step": 10327
+    },
+    {
+      "epoch": 0.6353346456692913,
+      "grad_norm": 1.328292727470398,
+      "learning_rate": 5.881701895349033e-05,
+      "loss": 2.6931,
+      "step": 10328
+    },
+    {
+      "epoch": 0.6353961614173228,
+      "grad_norm": 1.5664781332015991,
+      "learning_rate": 5.8799398092629734e-05,
+      "loss": 1.8321,
+      "step": 10329
+    },
+    {
+      "epoch": 0.6354576771653543,
+      "grad_norm": 1.2252286672592163,
+      "learning_rate": 5.8781778772441864e-05,
+      "loss": 2.3026,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6355191929133859,
+      "grad_norm": 1.1536482572555542,
+      "learning_rate": 5.8764160993585645e-05,
+      "loss": 2.4361,
+      "step": 10331
+    },
+    {
+      "epoch": 0.6355807086614174,
+      "grad_norm": 1.3792572021484375,
+      "learning_rate": 5.874654475671986e-05,
+      "loss": 2.0817,
+      "step": 10332
+    },
+    {
+      "epoch": 0.6356422244094488,
+      "grad_norm": 1.3076045513153076,
+      "learning_rate": 5.872893006250326e-05,
+      "loss": 2.1412,
+      "step": 10333
+    },
+    {
+      "epoch": 0.6357037401574803,
+      "grad_norm": 1.2874819040298462,
+      "learning_rate": 5.871131691159455e-05,
+      "loss": 2.7093,
+      "step": 10334
+    },
+    {
+      "epoch": 0.6357652559055118,
+      "grad_norm": 1.7464412450790405,
+      "learning_rate": 5.869370530465239e-05,
+      "loss": 2.2888,
+      "step": 10335
+    },
+    {
+      "epoch": 0.6358267716535433,
+      "grad_norm": 1.5957410335540771,
+      "learning_rate": 5.867609524233526e-05,
+      "loss": 2.1197,
+      "step": 10336
+    },
+    {
+      "epoch": 0.6358882874015748,
+      "grad_norm": 1.1142007112503052,
+      "learning_rate": 5.8658486725301786e-05,
+      "loss": 3.0225,
+      "step": 10337
+    },
+    {
+      "epoch": 0.6359498031496063,
+      "grad_norm": 1.2098757028579712,
+      "learning_rate": 5.864087975421035e-05,
+      "loss": 2.9069,
+      "step": 10338
+    },
+    {
+      "epoch": 0.6360113188976378,
+      "grad_norm": 1.296210765838623,
+      "learning_rate": 5.862327432971943e-05,
+      "loss": 2.769,
+      "step": 10339
+    },
+    {
+      "epoch": 0.6360728346456693,
+      "grad_norm": 1.1862573623657227,
+      "learning_rate": 5.86056704524873e-05,
+      "loss": 2.5428,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6361343503937008,
+      "grad_norm": 1.1046514511108398,
+      "learning_rate": 5.858806812317226e-05,
+      "loss": 2.866,
+      "step": 10341
+    },
+    {
+      "epoch": 0.6361958661417323,
+      "grad_norm": 1.2893248796463013,
+      "learning_rate": 5.8570467342432565e-05,
+      "loss": 2.7065,
+      "step": 10342
+    },
+    {
+      "epoch": 0.6362573818897638,
+      "grad_norm": 1.7922064065933228,
+      "learning_rate": 5.8552868110926377e-05,
+      "loss": 2.292,
+      "step": 10343
+    },
+    {
+      "epoch": 0.6363188976377953,
+      "grad_norm": 1.2797832489013672,
+      "learning_rate": 5.853527042931182e-05,
+      "loss": 1.819,
+      "step": 10344
+    },
+    {
+      "epoch": 0.6363804133858267,
+      "grad_norm": 1.1985318660736084,
+      "learning_rate": 5.851767429824691e-05,
+      "loss": 2.2906,
+      "step": 10345
+    },
+    {
+      "epoch": 0.6364419291338582,
+      "grad_norm": 1.418782114982605,
+      "learning_rate": 5.850007971838968e-05,
+      "loss": 2.7077,
+      "step": 10346
+    },
+    {
+      "epoch": 0.6365034448818898,
+      "grad_norm": 1.4858863353729248,
+      "learning_rate": 5.848248669039804e-05,
+      "loss": 2.7798,
+      "step": 10347
+    },
+    {
+      "epoch": 0.6365649606299213,
+      "grad_norm": 1.0528167486190796,
+      "learning_rate": 5.846489521492994e-05,
+      "loss": 2.6302,
+      "step": 10348
+    },
+    {
+      "epoch": 0.6366264763779528,
+      "grad_norm": 1.721489429473877,
+      "learning_rate": 5.8447305292643086e-05,
+      "loss": 2.2301,
+      "step": 10349
+    },
+    {
+      "epoch": 0.6366879921259843,
+      "grad_norm": 1.5252255201339722,
+      "learning_rate": 5.842971692419536e-05,
+      "loss": 2.6929,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6367495078740157,
+      "grad_norm": 1.144049048423767,
+      "learning_rate": 5.84121301102444e-05,
+      "loss": 2.6182,
+      "step": 10351
+    },
+    {
+      "epoch": 0.6368110236220472,
+      "grad_norm": 1.397394061088562,
+      "learning_rate": 5.839454485144787e-05,
+      "loss": 2.4565,
+      "step": 10352
+    },
+    {
+      "epoch": 0.6368725393700787,
+      "grad_norm": 1.0754096508026123,
+      "learning_rate": 5.837696114846336e-05,
+      "loss": 1.5913,
+      "step": 10353
+    },
+    {
+      "epoch": 0.6369340551181102,
+      "grad_norm": 1.1689589023590088,
+      "learning_rate": 5.8359379001948445e-05,
+      "loss": 2.5848,
+      "step": 10354
+    },
+    {
+      "epoch": 0.6369955708661418,
+      "grad_norm": 1.495650291442871,
+      "learning_rate": 5.834179841256051e-05,
+      "loss": 2.1707,
+      "step": 10355
+    },
+    {
+      "epoch": 0.6370570866141733,
+      "grad_norm": 0.9965783953666687,
+      "learning_rate": 5.832421938095708e-05,
+      "loss": 2.8287,
+      "step": 10356
+    },
+    {
+      "epoch": 0.6371186023622047,
+      "grad_norm": 1.2521198987960815,
+      "learning_rate": 5.83066419077954e-05,
+      "loss": 2.7478,
+      "step": 10357
+    },
+    {
+      "epoch": 0.6371801181102362,
+      "grad_norm": 1.432503581047058,
+      "learning_rate": 5.828906599373283e-05,
+      "loss": 2.6181,
+      "step": 10358
+    },
+    {
+      "epoch": 0.6372416338582677,
+      "grad_norm": 1.4436856508255005,
+      "learning_rate": 5.82714916394266e-05,
+      "loss": 2.4888,
+      "step": 10359
+    },
+    {
+      "epoch": 0.6373031496062992,
+      "grad_norm": 1.3020265102386475,
+      "learning_rate": 5.825391884553395e-05,
+      "loss": 2.7769,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6373646653543307,
+      "grad_norm": 1.1274093389511108,
+      "learning_rate": 5.8236347612711906e-05,
+      "loss": 2.3195,
+      "step": 10361
+    },
+    {
+      "epoch": 0.6374261811023622,
+      "grad_norm": 1.323288083076477,
+      "learning_rate": 5.8218777941617616e-05,
+      "loss": 2.5855,
+      "step": 10362
+    },
+    {
+      "epoch": 0.6374876968503937,
+      "grad_norm": 1.402686595916748,
+      "learning_rate": 5.820120983290802e-05,
+      "loss": 2.4767,
+      "step": 10363
+    },
+    {
+      "epoch": 0.6375492125984252,
+      "grad_norm": 1.152733325958252,
+      "learning_rate": 5.8183643287240095e-05,
+      "loss": 3.1203,
+      "step": 10364
+    },
+    {
+      "epoch": 0.6376107283464567,
+      "grad_norm": 1.263121247291565,
+      "learning_rate": 5.816607830527074e-05,
+      "loss": 2.5978,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6376722440944882,
+      "grad_norm": 1.1382938623428345,
+      "learning_rate": 5.814851488765682e-05,
+      "loss": 2.4737,
+      "step": 10366
+    },
+    {
+      "epoch": 0.6377337598425197,
+      "grad_norm": 1.183079481124878,
+      "learning_rate": 5.8130953035055036e-05,
+      "loss": 2.7844,
+      "step": 10367
+    },
+    {
+      "epoch": 0.6377952755905512,
+      "grad_norm": 1.1693123579025269,
+      "learning_rate": 5.8113392748122176e-05,
+      "loss": 2.7124,
+      "step": 10368
+    },
+    {
+      "epoch": 0.6378567913385826,
+      "grad_norm": 1.2350260019302368,
+      "learning_rate": 5.809583402751483e-05,
+      "loss": 2.5799,
+      "step": 10369
+    },
+    {
+      "epoch": 0.6379183070866141,
+      "grad_norm": 2.16494083404541,
+      "learning_rate": 5.807827687388963e-05,
+      "loss": 2.3051,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6379798228346457,
+      "grad_norm": 1.929491639137268,
+      "learning_rate": 5.80607212879031e-05,
+      "loss": 2.224,
+      "step": 10371
+    },
+    {
+      "epoch": 0.6380413385826772,
+      "grad_norm": 1.3259142637252808,
+      "learning_rate": 5.8043167270211774e-05,
+      "loss": 2.1452,
+      "step": 10372
+    },
+    {
+      "epoch": 0.6381028543307087,
+      "grad_norm": 1.396801471710205,
+      "learning_rate": 5.8025614821471994e-05,
+      "loss": 2.8623,
+      "step": 10373
+    },
+    {
+      "epoch": 0.6381643700787402,
+      "grad_norm": 1.2585667371749878,
+      "learning_rate": 5.800806394234022e-05,
+      "loss": 2.3837,
+      "step": 10374
+    },
+    {
+      "epoch": 0.6382258858267716,
+      "grad_norm": 1.3107242584228516,
+      "learning_rate": 5.799051463347265e-05,
+      "loss": 2.3707,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6382874015748031,
+      "grad_norm": 1.190915822982788,
+      "learning_rate": 5.797296689552558e-05,
+      "loss": 3.0845,
+      "step": 10376
+    },
+    {
+      "epoch": 0.6383489173228346,
+      "grad_norm": 1.0429978370666504,
+      "learning_rate": 5.7955420729155226e-05,
+      "loss": 2.6292,
+      "step": 10377
+    },
+    {
+      "epoch": 0.6384104330708661,
+      "grad_norm": 1.3205758333206177,
+      "learning_rate": 5.793787613501766e-05,
+      "loss": 2.3775,
+      "step": 10378
+    },
+    {
+      "epoch": 0.6384719488188977,
+      "grad_norm": 1.1345819234848022,
+      "learning_rate": 5.792033311376903e-05,
+      "loss": 2.7833,
+      "step": 10379
+    },
+    {
+      "epoch": 0.6385334645669292,
+      "grad_norm": 1.3800116777420044,
+      "learning_rate": 5.790279166606527e-05,
+      "loss": 2.418,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6385949803149606,
+      "grad_norm": 1.3646222352981567,
+      "learning_rate": 5.788525179256238e-05,
+      "loss": 2.1838,
+      "step": 10381
+    },
+    {
+      "epoch": 0.6386564960629921,
+      "grad_norm": 1.497115969657898,
+      "learning_rate": 5.78677134939162e-05,
+      "loss": 2.328,
+      "step": 10382
+    },
+    {
+      "epoch": 0.6387180118110236,
+      "grad_norm": 1.2678014039993286,
+      "learning_rate": 5.785017677078259e-05,
+      "loss": 2.0584,
+      "step": 10383
+    },
+    {
+      "epoch": 0.6387795275590551,
+      "grad_norm": 1.1423008441925049,
+      "learning_rate": 5.783264162381736e-05,
+      "loss": 2.5751,
+      "step": 10384
+    },
+    {
+      "epoch": 0.6388410433070866,
+      "grad_norm": 1.4952630996704102,
+      "learning_rate": 5.7815108053676224e-05,
+      "loss": 2.3973,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6389025590551181,
+      "grad_norm": 1.374259114265442,
+      "learning_rate": 5.779757606101478e-05,
+      "loss": 2.8806,
+      "step": 10386
+    },
+    {
+      "epoch": 0.6389640748031497,
+      "grad_norm": 1.3719562292099,
+      "learning_rate": 5.7780045646488704e-05,
+      "loss": 1.9355,
+      "step": 10387
+    },
+    {
+      "epoch": 0.6390255905511811,
+      "grad_norm": 1.5889806747436523,
+      "learning_rate": 5.7762516810753454e-05,
+      "loss": 2.7872,
+      "step": 10388
+    },
+    {
+      "epoch": 0.6390871062992126,
+      "grad_norm": 1.402217149734497,
+      "learning_rate": 5.7744989554464555e-05,
+      "loss": 2.3584,
+      "step": 10389
+    },
+    {
+      "epoch": 0.6391486220472441,
+      "grad_norm": 1.135439157485962,
+      "learning_rate": 5.772746387827743e-05,
+      "loss": 2.7127,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6392101377952756,
+      "grad_norm": 1.2233389616012573,
+      "learning_rate": 5.770993978284747e-05,
+      "loss": 2.6129,
+      "step": 10391
+    },
+    {
+      "epoch": 0.6392716535433071,
+      "grad_norm": 1.226732611656189,
+      "learning_rate": 5.769241726882991e-05,
+      "loss": 2.2444,
+      "step": 10392
+    },
+    {
+      "epoch": 0.6393331692913385,
+      "grad_norm": 1.1987651586532593,
+      "learning_rate": 5.767489633688007e-05,
+      "loss": 2.4766,
+      "step": 10393
+    },
+    {
+      "epoch": 0.63939468503937,
+      "grad_norm": 1.274125099182129,
+      "learning_rate": 5.7657376987653057e-05,
+      "loss": 2.4956,
+      "step": 10394
+    },
+    {
+      "epoch": 0.6394562007874016,
+      "grad_norm": 1.145537257194519,
+      "learning_rate": 5.763985922180404e-05,
+      "loss": 2.6143,
+      "step": 10395
+    },
+    {
+      "epoch": 0.6395177165354331,
+      "grad_norm": 1.2906227111816406,
+      "learning_rate": 5.762234303998807e-05,
+      "loss": 2.2637,
+      "step": 10396
+    },
+    {
+      "epoch": 0.6395792322834646,
+      "grad_norm": 1.274940013885498,
+      "learning_rate": 5.760482844286022e-05,
+      "loss": 3.1862,
+      "step": 10397
+    },
+    {
+      "epoch": 0.6396407480314961,
+      "grad_norm": 1.2594871520996094,
+      "learning_rate": 5.7587315431075326e-05,
+      "loss": 2.8336,
+      "step": 10398
+    },
+    {
+      "epoch": 0.6397022637795275,
+      "grad_norm": 1.210656762123108,
+      "learning_rate": 5.7569804005288396e-05,
+      "loss": 2.5841,
+      "step": 10399
+    },
+    {
+      "epoch": 0.639763779527559,
+      "grad_norm": 1.3315685987472534,
+      "learning_rate": 5.7552294166154144e-05,
+      "loss": 2.3201,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6398252952755905,
+      "grad_norm": 1.134869933128357,
+      "learning_rate": 5.753478591432744e-05,
+      "loss": 2.6011,
+      "step": 10401
+    },
+    {
+      "epoch": 0.639886811023622,
+      "grad_norm": 1.167527198791504,
+      "learning_rate": 5.751727925046286e-05,
+      "loss": 2.2547,
+      "step": 10402
+    },
+    {
+      "epoch": 0.6399483267716536,
+      "grad_norm": 1.2946345806121826,
+      "learning_rate": 5.749977417521525e-05,
+      "loss": 2.1884,
+      "step": 10403
+    },
+    {
+      "epoch": 0.6400098425196851,
+      "grad_norm": 1.1870077848434448,
+      "learning_rate": 5.748227068923904e-05,
+      "loss": 2.6232,
+      "step": 10404
+    },
+    {
+      "epoch": 0.6400713582677166,
+      "grad_norm": 1.1773215532302856,
+      "learning_rate": 5.746476879318887e-05,
+      "loss": 2.7588,
+      "step": 10405
+    },
+    {
+      "epoch": 0.640132874015748,
+      "grad_norm": 1.235770344734192,
+      "learning_rate": 5.744726848771912e-05,
+      "loss": 2.2785,
+      "step": 10406
+    },
+    {
+      "epoch": 0.6401943897637795,
+      "grad_norm": 1.4210387468338013,
+      "learning_rate": 5.742976977348429e-05,
+      "loss": 2.2386,
+      "step": 10407
+    },
+    {
+      "epoch": 0.640255905511811,
+      "grad_norm": 1.3912811279296875,
+      "learning_rate": 5.7412272651138644e-05,
+      "loss": 2.1832,
+      "step": 10408
+    },
+    {
+      "epoch": 0.6403174212598425,
+      "grad_norm": 1.3009040355682373,
+      "learning_rate": 5.739477712133654e-05,
+      "loss": 2.2882,
+      "step": 10409
+    },
+    {
+      "epoch": 0.640378937007874,
+      "grad_norm": 1.374758005142212,
+      "learning_rate": 5.737728318473219e-05,
+      "loss": 2.7295,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6404404527559056,
+      "grad_norm": 1.2679381370544434,
+      "learning_rate": 5.735979084197981e-05,
+      "loss": 2.3287,
+      "step": 10411
+    },
+    {
+      "epoch": 0.640501968503937,
+      "grad_norm": 1.2650173902511597,
+      "learning_rate": 5.7342300093733494e-05,
+      "loss": 2.963,
+      "step": 10412
+    },
+    {
+      "epoch": 0.6405634842519685,
+      "grad_norm": 1.1514476537704468,
+      "learning_rate": 5.732481094064724e-05,
+      "loss": 2.7025,
+      "step": 10413
+    },
+    {
+      "epoch": 0.640625,
+      "grad_norm": 1.2881736755371094,
+      "learning_rate": 5.73073233833751e-05,
+      "loss": 2.6563,
+      "step": 10414
+    },
+    {
+      "epoch": 0.6406865157480315,
+      "grad_norm": 1.1032053232192993,
+      "learning_rate": 5.728983742257101e-05,
+      "loss": 2.8734,
+      "step": 10415
+    },
+    {
+      "epoch": 0.640748031496063,
+      "grad_norm": 1.3732157945632935,
+      "learning_rate": 5.727235305888887e-05,
+      "loss": 2.7271,
+      "step": 10416
+    },
+    {
+      "epoch": 0.6408095472440944,
+      "grad_norm": 1.1567269563674927,
+      "learning_rate": 5.7254870292982446e-05,
+      "loss": 2.2353,
+      "step": 10417
+    },
+    {
+      "epoch": 0.640871062992126,
+      "grad_norm": 1.402845859527588,
+      "learning_rate": 5.723738912550555e-05,
+      "loss": 2.614,
+      "step": 10418
+    },
+    {
+      "epoch": 0.6409325787401575,
+      "grad_norm": 1.348281979560852,
+      "learning_rate": 5.7219909557111806e-05,
+      "loss": 2.3555,
+      "step": 10419
+    },
+    {
+      "epoch": 0.640994094488189,
+      "grad_norm": 1.4317915439605713,
+      "learning_rate": 5.720243158845494e-05,
+      "loss": 2.2728,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6410556102362205,
+      "grad_norm": 1.3130459785461426,
+      "learning_rate": 5.718495522018841e-05,
+      "loss": 2.4798,
+      "step": 10421
+    },
+    {
+      "epoch": 0.641117125984252,
+      "grad_norm": 1.2788034677505493,
+      "learning_rate": 5.716748045296588e-05,
+      "loss": 2.4596,
+      "step": 10422
+    },
+    {
+      "epoch": 0.6411786417322834,
+      "grad_norm": 1.0403231382369995,
+      "learning_rate": 5.715000728744071e-05,
+      "loss": 2.7668,
+      "step": 10423
+    },
+    {
+      "epoch": 0.6412401574803149,
+      "grad_norm": 1.1000057458877563,
+      "learning_rate": 5.7132535724266375e-05,
+      "loss": 2.3478,
+      "step": 10424
+    },
+    {
+      "epoch": 0.6413016732283464,
+      "grad_norm": 1.1211646795272827,
+      "learning_rate": 5.711506576409611e-05,
+      "loss": 2.6973,
+      "step": 10425
+    },
+    {
+      "epoch": 0.641363188976378,
+      "grad_norm": 1.1895620822906494,
+      "learning_rate": 5.7097597407583315e-05,
+      "loss": 2.3635,
+      "step": 10426
+    },
+    {
+      "epoch": 0.6414247047244095,
+      "grad_norm": 1.3355445861816406,
+      "learning_rate": 5.7080130655381095e-05,
+      "loss": 2.1768,
+      "step": 10427
+    },
+    {
+      "epoch": 0.641486220472441,
+      "grad_norm": 1.4502538442611694,
+      "learning_rate": 5.706266550814266e-05,
+      "loss": 2.6726,
+      "step": 10428
+    },
+    {
+      "epoch": 0.6415477362204725,
+      "grad_norm": 2.08309268951416,
+      "learning_rate": 5.704520196652111e-05,
+      "loss": 1.9439,
+      "step": 10429
+    },
+    {
+      "epoch": 0.6416092519685039,
+      "grad_norm": 1.5753313302993774,
+      "learning_rate": 5.702774003116953e-05,
+      "loss": 1.9676,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6416707677165354,
+      "grad_norm": 1.3515396118164062,
+      "learning_rate": 5.70102797027408e-05,
+      "loss": 2.8611,
+      "step": 10431
+    },
+    {
+      "epoch": 0.6417322834645669,
+      "grad_norm": 1.1936767101287842,
+      "learning_rate": 5.699282098188794e-05,
+      "loss": 1.9763,
+      "step": 10432
+    },
+    {
+      "epoch": 0.6417937992125984,
+      "grad_norm": 1.1445626020431519,
+      "learning_rate": 5.697536386926372e-05,
+      "loss": 3.0263,
+      "step": 10433
+    },
+    {
+      "epoch": 0.64185531496063,
+      "grad_norm": 1.1304891109466553,
+      "learning_rate": 5.695790836552098e-05,
+      "loss": 2.3371,
+      "step": 10434
+    },
+    {
+      "epoch": 0.6419168307086615,
+      "grad_norm": 1.1627142429351807,
+      "learning_rate": 5.6940454471312465e-05,
+      "loss": 2.9617,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6419783464566929,
+      "grad_norm": 1.1733192205429077,
+      "learning_rate": 5.692300218729087e-05,
+      "loss": 2.8492,
+      "step": 10436
+    },
+    {
+      "epoch": 0.6420398622047244,
+      "grad_norm": 1.7732006311416626,
+      "learning_rate": 5.690555151410877e-05,
+      "loss": 2.1686,
+      "step": 10437
+    },
+    {
+      "epoch": 0.6421013779527559,
+      "grad_norm": 1.098564624786377,
+      "learning_rate": 5.6888102452418756e-05,
+      "loss": 2.8149,
+      "step": 10438
+    },
+    {
+      "epoch": 0.6421628937007874,
+      "grad_norm": 1.5232152938842773,
+      "learning_rate": 5.68706550028733e-05,
+      "loss": 2.667,
+      "step": 10439
+    },
+    {
+      "epoch": 0.6422244094488189,
+      "grad_norm": 1.351861834526062,
+      "learning_rate": 5.685320916612482e-05,
+      "loss": 2.3048,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6422859251968503,
+      "grad_norm": 1.3866534233093262,
+      "learning_rate": 5.6835764942825743e-05,
+      "loss": 2.2399,
+      "step": 10441
+    },
+    {
+      "epoch": 0.6423474409448819,
+      "grad_norm": 1.3162959814071655,
+      "learning_rate": 5.681832233362841e-05,
+      "loss": 1.9372,
+      "step": 10442
+    },
+    {
+      "epoch": 0.6424089566929134,
+      "grad_norm": 1.2619242668151855,
+      "learning_rate": 5.6800881339184994e-05,
+      "loss": 2.9464,
+      "step": 10443
+    },
+    {
+      "epoch": 0.6424704724409449,
+      "grad_norm": 1.3033154010772705,
+      "learning_rate": 5.678344196014777e-05,
+      "loss": 2.3487,
+      "step": 10444
+    },
+    {
+      "epoch": 0.6425319881889764,
+      "grad_norm": 1.5090821981430054,
+      "learning_rate": 5.6766004197168775e-05,
+      "loss": 2.5964,
+      "step": 10445
+    },
+    {
+      "epoch": 0.6425935039370079,
+      "grad_norm": 1.367497444152832,
+      "learning_rate": 5.674856805090021e-05,
+      "loss": 1.9335,
+      "step": 10446
+    },
+    {
+      "epoch": 0.6426550196850394,
+      "grad_norm": 1.226319432258606,
+      "learning_rate": 5.673113352199393e-05,
+      "loss": 1.9237,
+      "step": 10447
+    },
+    {
+      "epoch": 0.6427165354330708,
+      "grad_norm": 1.2113677263259888,
+      "learning_rate": 5.671370061110206e-05,
+      "loss": 2.1012,
+      "step": 10448
+    },
+    {
+      "epoch": 0.6427780511811023,
+      "grad_norm": 1.3447067737579346,
+      "learning_rate": 5.6696269318876425e-05,
+      "loss": 2.6383,
+      "step": 10449
+    },
+    {
+      "epoch": 0.6428395669291339,
+      "grad_norm": 2.2250919342041016,
+      "learning_rate": 5.66788396459688e-05,
+      "loss": 2.2345,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6429010826771654,
+      "grad_norm": 1.6088268756866455,
+      "learning_rate": 5.6661411593031066e-05,
+      "loss": 2.0953,
+      "step": 10451
+    },
+    {
+      "epoch": 0.6429625984251969,
+      "grad_norm": 1.7350788116455078,
+      "learning_rate": 5.6643985160714806e-05,
+      "loss": 1.9935,
+      "step": 10452
+    },
+    {
+      "epoch": 0.6430241141732284,
+      "grad_norm": 1.2609155178070068,
+      "learning_rate": 5.662656034967176e-05,
+      "loss": 2.9216,
+      "step": 10453
+    },
+    {
+      "epoch": 0.6430856299212598,
+      "grad_norm": 1.1011873483657837,
+      "learning_rate": 5.6609137160553504e-05,
+      "loss": 2.7698,
+      "step": 10454
+    },
+    {
+      "epoch": 0.6431471456692913,
+      "grad_norm": 1.3566416501998901,
+      "learning_rate": 5.65917155940116e-05,
+      "loss": 2.4843,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6432086614173228,
+      "grad_norm": 1.2615042924880981,
+      "learning_rate": 5.657429565069743e-05,
+      "loss": 2.5406,
+      "step": 10456
+    },
+    {
+      "epoch": 0.6432701771653543,
+      "grad_norm": 1.1843277215957642,
+      "learning_rate": 5.6556877331262515e-05,
+      "loss": 2.4617,
+      "step": 10457
+    },
+    {
+      "epoch": 0.6433316929133859,
+      "grad_norm": 1.7209250926971436,
+      "learning_rate": 5.6539460636358086e-05,
+      "loss": 2.1748,
+      "step": 10458
+    },
+    {
+      "epoch": 0.6433932086614174,
+      "grad_norm": 1.1684218645095825,
+      "learning_rate": 5.652204556663549e-05,
+      "loss": 2.9972,
+      "step": 10459
+    },
+    {
+      "epoch": 0.6434547244094488,
+      "grad_norm": 1.2308473587036133,
+      "learning_rate": 5.650463212274595e-05,
+      "loss": 2.6921,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6435162401574803,
+      "grad_norm": 1.28851318359375,
+      "learning_rate": 5.648722030534067e-05,
+      "loss": 3.0521,
+      "step": 10461
+    },
+    {
+      "epoch": 0.6435777559055118,
+      "grad_norm": 1.533997654914856,
+      "learning_rate": 5.6469810115070686e-05,
+      "loss": 2.519,
+      "step": 10462
+    },
+    {
+      "epoch": 0.6436392716535433,
+      "grad_norm": 1.2082085609436035,
+      "learning_rate": 5.645240155258711e-05,
+      "loss": 2.6555,
+      "step": 10463
+    },
+    {
+      "epoch": 0.6437007874015748,
+      "grad_norm": 1.4465044736862183,
+      "learning_rate": 5.643499461854084e-05,
+      "loss": 1.994,
+      "step": 10464
+    },
+    {
+      "epoch": 0.6437623031496063,
+      "grad_norm": 1.3134442567825317,
+      "learning_rate": 5.6417589313582856e-05,
+      "loss": 2.714,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6438238188976378,
+      "grad_norm": 1.2468665838241577,
+      "learning_rate": 5.6400185638364004e-05,
+      "loss": 2.4291,
+      "step": 10466
+    },
+    {
+      "epoch": 0.6438853346456693,
+      "grad_norm": 1.1722811460494995,
+      "learning_rate": 5.638278359353513e-05,
+      "loss": 2.8774,
+      "step": 10467
+    },
+    {
+      "epoch": 0.6439468503937008,
+      "grad_norm": 1.2970179319381714,
+      "learning_rate": 5.636538317974689e-05,
+      "loss": 2.4903,
+      "step": 10468
+    },
+    {
+      "epoch": 0.6440083661417323,
+      "grad_norm": 1.188423752784729,
+      "learning_rate": 5.6347984397650045e-05,
+      "loss": 2.6186,
+      "step": 10469
+    },
+    {
+      "epoch": 0.6440698818897638,
+      "grad_norm": 1.3921535015106201,
+      "learning_rate": 5.6330587247895126e-05,
+      "loss": 2.4363,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6441313976377953,
+      "grad_norm": 1.3403078317642212,
+      "learning_rate": 5.6313191731132796e-05,
+      "loss": 2.095,
+      "step": 10471
+    },
+    {
+      "epoch": 0.6441929133858267,
+      "grad_norm": 1.166049838066101,
+      "learning_rate": 5.629579784801344e-05,
+      "loss": 2.7976,
+      "step": 10472
+    },
+    {
+      "epoch": 0.6442544291338582,
+      "grad_norm": 1.7939597368240356,
+      "learning_rate": 5.6278405599187556e-05,
+      "loss": 2.624,
+      "step": 10473
+    },
+    {
+      "epoch": 0.6443159448818898,
+      "grad_norm": 1.588094711303711,
+      "learning_rate": 5.626101498530549e-05,
+      "loss": 2.9046,
+      "step": 10474
+    },
+    {
+      "epoch": 0.6443774606299213,
+      "grad_norm": 1.206839680671692,
+      "learning_rate": 5.62436260070176e-05,
+      "loss": 1.9485,
+      "step": 10475
+    },
+    {
+      "epoch": 0.6444389763779528,
+      "grad_norm": 1.1354613304138184,
+      "learning_rate": 5.622623866497408e-05,
+      "loss": 2.0656,
+      "step": 10476
+    },
+    {
+      "epoch": 0.6445004921259843,
+      "grad_norm": 1.2618441581726074,
+      "learning_rate": 5.620885295982519e-05,
+      "loss": 2.8171,
+      "step": 10477
+    },
+    {
+      "epoch": 0.6445620078740157,
+      "grad_norm": 1.240996241569519,
+      "learning_rate": 5.619146889222097e-05,
+      "loss": 2.3504,
+      "step": 10478
+    },
+    {
+      "epoch": 0.6446235236220472,
+      "grad_norm": 1.0451699495315552,
+      "learning_rate": 5.617408646281153e-05,
+      "loss": 2.4933,
+      "step": 10479
+    },
+    {
+      "epoch": 0.6446850393700787,
+      "grad_norm": 1.027446985244751,
+      "learning_rate": 5.615670567224687e-05,
+      "loss": 2.9271,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6447465551181102,
+      "grad_norm": 1.4417206048965454,
+      "learning_rate": 5.613932652117697e-05,
+      "loss": 2.1634,
+      "step": 10481
+    },
+    {
+      "epoch": 0.6448080708661418,
+      "grad_norm": 1.1361106634140015,
+      "learning_rate": 5.612194901025166e-05,
+      "loss": 2.7764,
+      "step": 10482
+    },
+    {
+      "epoch": 0.6448695866141733,
+      "grad_norm": 1.2581249475479126,
+      "learning_rate": 5.610457314012083e-05,
+      "loss": 2.5223,
+      "step": 10483
+    },
+    {
+      "epoch": 0.6449311023622047,
+      "grad_norm": 1.3377870321273804,
+      "learning_rate": 5.6087198911434145e-05,
+      "loss": 2.3987,
+      "step": 10484
+    },
+    {
+      "epoch": 0.6449926181102362,
+      "grad_norm": 1.2266559600830078,
+      "learning_rate": 5.606982632484137e-05,
+      "loss": 2.1871,
+      "step": 10485
+    },
+    {
+      "epoch": 0.6450541338582677,
+      "grad_norm": 1.1468557119369507,
+      "learning_rate": 5.605245538099215e-05,
+      "loss": 2.2018,
+      "step": 10486
+    },
+    {
+      "epoch": 0.6451156496062992,
+      "grad_norm": 1.3594081401824951,
+      "learning_rate": 5.603508608053601e-05,
+      "loss": 2.2857,
+      "step": 10487
+    },
+    {
+      "epoch": 0.6451771653543307,
+      "grad_norm": 1.2175798416137695,
+      "learning_rate": 5.6017718424122536e-05,
+      "loss": 2.7332,
+      "step": 10488
+    },
+    {
+      "epoch": 0.6452386811023622,
+      "grad_norm": 1.2714027166366577,
+      "learning_rate": 5.600035241240111e-05,
+      "loss": 2.7111,
+      "step": 10489
+    },
+    {
+      "epoch": 0.6453001968503937,
+      "grad_norm": 1.1958401203155518,
+      "learning_rate": 5.598298804602118e-05,
+      "loss": 3.059,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6453617125984252,
+      "grad_norm": 1.414358139038086,
+      "learning_rate": 5.5965625325631984e-05,
+      "loss": 2.5471,
+      "step": 10491
+    },
+    {
+      "epoch": 0.6454232283464567,
+      "grad_norm": 1.0951354503631592,
+      "learning_rate": 5.594826425188293e-05,
+      "loss": 2.2785,
+      "step": 10492
+    },
+    {
+      "epoch": 0.6454847440944882,
+      "grad_norm": 1.1466975212097168,
+      "learning_rate": 5.593090482542311e-05,
+      "loss": 2.6287,
+      "step": 10493
+    },
+    {
+      "epoch": 0.6455462598425197,
+      "grad_norm": 1.1120296716690063,
+      "learning_rate": 5.5913547046901774e-05,
+      "loss": 2.4382,
+      "step": 10494
+    },
+    {
+      "epoch": 0.6456077755905512,
+      "grad_norm": 1.0422160625457764,
+      "learning_rate": 5.5896190916967885e-05,
+      "loss": 2.2776,
+      "step": 10495
+    },
+    {
+      "epoch": 0.6456692913385826,
+      "grad_norm": 1.1623096466064453,
+      "learning_rate": 5.587883643627058e-05,
+      "loss": 2.2063,
+      "step": 10496
+    },
+    {
+      "epoch": 0.6457308070866141,
+      "grad_norm": 1.3006945848464966,
+      "learning_rate": 5.586148360545871e-05,
+      "loss": 2.6295,
+      "step": 10497
+    },
+    {
+      "epoch": 0.6457923228346457,
+      "grad_norm": 1.406202793121338,
+      "learning_rate": 5.5844132425181236e-05,
+      "loss": 2.6067,
+      "step": 10498
+    },
+    {
+      "epoch": 0.6458538385826772,
+      "grad_norm": 1.3154250383377075,
+      "learning_rate": 5.582678289608701e-05,
+      "loss": 2.2938,
+      "step": 10499
+    },
+    {
+      "epoch": 0.6459153543307087,
+      "grad_norm": 1.3916015625,
+      "learning_rate": 5.5809435018824796e-05,
+      "loss": 2.4301,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6459768700787402,
+      "grad_norm": 1.300726056098938,
+      "learning_rate": 5.5792088794043275e-05,
+      "loss": 2.4955,
+      "step": 10501
+    },
+    {
+      "epoch": 0.6460383858267716,
+      "grad_norm": 1.25248384475708,
+      "learning_rate": 5.577474422239115e-05,
+      "loss": 2.5286,
+      "step": 10502
+    },
+    {
+      "epoch": 0.6460999015748031,
+      "grad_norm": 1.1986738443374634,
+      "learning_rate": 5.575740130451696e-05,
+      "loss": 2.153,
+      "step": 10503
+    },
+    {
+      "epoch": 0.6461614173228346,
+      "grad_norm": 1.286765456199646,
+      "learning_rate": 5.5740060041069255e-05,
+      "loss": 2.9469,
+      "step": 10504
+    },
+    {
+      "epoch": 0.6462229330708661,
+      "grad_norm": 1.3007996082305908,
+      "learning_rate": 5.5722720432696504e-05,
+      "loss": 2.6147,
+      "step": 10505
+    },
+    {
+      "epoch": 0.6462844488188977,
+      "grad_norm": 1.3663699626922607,
+      "learning_rate": 5.5705382480047155e-05,
+      "loss": 2.2405,
+      "step": 10506
+    },
+    {
+      "epoch": 0.6463459645669292,
+      "grad_norm": 1.3433891534805298,
+      "learning_rate": 5.568804618376946e-05,
+      "loss": 3.16,
+      "step": 10507
+    },
+    {
+      "epoch": 0.6464074803149606,
+      "grad_norm": 1.398112177848816,
+      "learning_rate": 5.5670711544511775e-05,
+      "loss": 3.0242,
+      "step": 10508
+    },
+    {
+      "epoch": 0.6464689960629921,
+      "grad_norm": 1.784881591796875,
+      "learning_rate": 5.565337856292228e-05,
+      "loss": 2.4541,
+      "step": 10509
+    },
+    {
+      "epoch": 0.6465305118110236,
+      "grad_norm": 1.7204060554504395,
+      "learning_rate": 5.5636047239649123e-05,
+      "loss": 2.028,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6465920275590551,
+      "grad_norm": 1.2349950075149536,
+      "learning_rate": 5.561871757534043e-05,
+      "loss": 3.1313,
+      "step": 10511
+    },
+    {
+      "epoch": 0.6466535433070866,
+      "grad_norm": 1.247835397720337,
+      "learning_rate": 5.560138957064425e-05,
+      "loss": 2.2458,
+      "step": 10512
+    },
+    {
+      "epoch": 0.6467150590551181,
+      "grad_norm": 1.5593767166137695,
+      "learning_rate": 5.5584063226208504e-05,
+      "loss": 2.0542,
+      "step": 10513
+    },
+    {
+      "epoch": 0.6467765748031497,
+      "grad_norm": 1.1927181482315063,
+      "learning_rate": 5.556673854268115e-05,
+      "loss": 2.3742,
+      "step": 10514
+    },
+    {
+      "epoch": 0.6468380905511811,
+      "grad_norm": 1.4074063301086426,
+      "learning_rate": 5.554941552070997e-05,
+      "loss": 2.7591,
+      "step": 10515
+    },
+    {
+      "epoch": 0.6468996062992126,
+      "grad_norm": 1.0685439109802246,
+      "learning_rate": 5.5532094160942826e-05,
+      "loss": 2.9804,
+      "step": 10516
+    },
+    {
+      "epoch": 0.6469611220472441,
+      "grad_norm": 1.2638992071151733,
+      "learning_rate": 5.551477446402733e-05,
+      "loss": 1.9976,
+      "step": 10517
+    },
+    {
+      "epoch": 0.6470226377952756,
+      "grad_norm": 1.279167890548706,
+      "learning_rate": 5.5497456430611284e-05,
+      "loss": 2.7701,
+      "step": 10518
+    },
+    {
+      "epoch": 0.6470841535433071,
+      "grad_norm": 1.7008371353149414,
+      "learning_rate": 5.548014006134217e-05,
+      "loss": 2.08,
+      "step": 10519
+    },
+    {
+      "epoch": 0.6471456692913385,
+      "grad_norm": 2.0301215648651123,
+      "learning_rate": 5.546282535686762e-05,
+      "loss": 2.1627,
+      "step": 10520
+    },
+    {
+      "epoch": 0.64720718503937,
+      "grad_norm": 1.6089473962783813,
+      "learning_rate": 5.5445512317835015e-05,
+      "loss": 2.0122,
+      "step": 10521
+    },
+    {
+      "epoch": 0.6472687007874016,
+      "grad_norm": 1.3980172872543335,
+      "learning_rate": 5.542820094489183e-05,
+      "loss": 2.3449,
+      "step": 10522
+    },
+    {
+      "epoch": 0.6473302165354331,
+      "grad_norm": 1.2032711505889893,
+      "learning_rate": 5.541089123868537e-05,
+      "loss": 2.5229,
+      "step": 10523
+    },
+    {
+      "epoch": 0.6473917322834646,
+      "grad_norm": 1.4879852533340454,
+      "learning_rate": 5.539358319986292e-05,
+      "loss": 2.4923,
+      "step": 10524
+    },
+    {
+      "epoch": 0.6474532480314961,
+      "grad_norm": 1.0909044742584229,
+      "learning_rate": 5.5376276829071784e-05,
+      "loss": 2.7892,
+      "step": 10525
+    },
+    {
+      "epoch": 0.6475147637795275,
+      "grad_norm": 1.2012158632278442,
+      "learning_rate": 5.535897212695902e-05,
+      "loss": 1.6542,
+      "step": 10526
+    },
+    {
+      "epoch": 0.647576279527559,
+      "grad_norm": 1.1467952728271484,
+      "learning_rate": 5.534166909417182e-05,
+      "loss": 2.35,
+      "step": 10527
+    },
+    {
+      "epoch": 0.6476377952755905,
+      "grad_norm": 1.4226492643356323,
+      "learning_rate": 5.532436773135713e-05,
+      "loss": 2.3493,
+      "step": 10528
+    },
+    {
+      "epoch": 0.647699311023622,
+      "grad_norm": 1.3275327682495117,
+      "learning_rate": 5.530706803916197e-05,
+      "loss": 1.6204,
+      "step": 10529
+    },
+    {
+      "epoch": 0.6477608267716536,
+      "grad_norm": 1.266038179397583,
+      "learning_rate": 5.528977001823324e-05,
+      "loss": 2.0053,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6478223425196851,
+      "grad_norm": 1.2530661821365356,
+      "learning_rate": 5.5272473669217864e-05,
+      "loss": 2.4659,
+      "step": 10531
+    },
+    {
+      "epoch": 0.6478838582677166,
+      "grad_norm": 1.6407445669174194,
+      "learning_rate": 5.5255178992762505e-05,
+      "loss": 2.3854,
+      "step": 10532
+    },
+    {
+      "epoch": 0.647945374015748,
+      "grad_norm": 1.3260033130645752,
+      "learning_rate": 5.523788598951398e-05,
+      "loss": 3.0095,
+      "step": 10533
+    },
+    {
+      "epoch": 0.6480068897637795,
+      "grad_norm": 1.360180139541626,
+      "learning_rate": 5.522059466011889e-05,
+      "loss": 2.8986,
+      "step": 10534
+    },
+    {
+      "epoch": 0.648068405511811,
+      "grad_norm": 1.1564717292785645,
+      "learning_rate": 5.520330500522387e-05,
+      "loss": 2.5347,
+      "step": 10535
+    },
+    {
+      "epoch": 0.6481299212598425,
+      "grad_norm": 1.2071079015731812,
+      "learning_rate": 5.518601702547545e-05,
+      "loss": 2.3936,
+      "step": 10536
+    },
+    {
+      "epoch": 0.648191437007874,
+      "grad_norm": 1.328425645828247,
+      "learning_rate": 5.5168730721520135e-05,
+      "loss": 1.5425,
+      "step": 10537
+    },
+    {
+      "epoch": 0.6482529527559056,
+      "grad_norm": 1.293694019317627,
+      "learning_rate": 5.5151446094004266e-05,
+      "loss": 2.4054,
+      "step": 10538
+    },
+    {
+      "epoch": 0.648314468503937,
+      "grad_norm": 1.300125241279602,
+      "learning_rate": 5.513416314357427e-05,
+      "loss": 2.8825,
+      "step": 10539
+    },
+    {
+      "epoch": 0.6483759842519685,
+      "grad_norm": 1.508357286453247,
+      "learning_rate": 5.511688187087636e-05,
+      "loss": 2.3425,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6484375,
+      "grad_norm": 1.324403166770935,
+      "learning_rate": 5.509960227655683e-05,
+      "loss": 2.4755,
+      "step": 10541
+    },
+    {
+      "epoch": 0.6484990157480315,
+      "grad_norm": 1.4822428226470947,
+      "learning_rate": 5.508232436126176e-05,
+      "loss": 2.4186,
+      "step": 10542
+    },
+    {
+      "epoch": 0.648560531496063,
+      "grad_norm": 1.7318201065063477,
+      "learning_rate": 5.506504812563729e-05,
+      "loss": 1.994,
+      "step": 10543
+    },
+    {
+      "epoch": 0.6486220472440944,
+      "grad_norm": 1.2803921699523926,
+      "learning_rate": 5.504777357032947e-05,
+      "loss": 2.5802,
+      "step": 10544
+    },
+    {
+      "epoch": 0.648683562992126,
+      "grad_norm": 1.2772639989852905,
+      "learning_rate": 5.503050069598428e-05,
+      "loss": 2.5687,
+      "step": 10545
+    },
+    {
+      "epoch": 0.6487450787401575,
+      "grad_norm": 1.423445463180542,
+      "learning_rate": 5.5013229503247566e-05,
+      "loss": 2.3379,
+      "step": 10546
+    },
+    {
+      "epoch": 0.648806594488189,
+      "grad_norm": 1.389050006866455,
+      "learning_rate": 5.499595999276528e-05,
+      "loss": 2.3992,
+      "step": 10547
+    },
+    {
+      "epoch": 0.6488681102362205,
+      "grad_norm": 1.26420259475708,
+      "learning_rate": 5.4978692165183074e-05,
+      "loss": 2.5493,
+      "step": 10548
+    },
+    {
+      "epoch": 0.648929625984252,
+      "grad_norm": 1.6106375455856323,
+      "learning_rate": 5.496142602114674e-05,
+      "loss": 2.3013,
+      "step": 10549
+    },
+    {
+      "epoch": 0.6489911417322834,
+      "grad_norm": 1.5344908237457275,
+      "learning_rate": 5.494416156130194e-05,
+      "loss": 2.3791,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6490526574803149,
+      "grad_norm": 1.18315851688385,
+      "learning_rate": 5.49268987862943e-05,
+      "loss": 2.3769,
+      "step": 10551
+    },
+    {
+      "epoch": 0.6491141732283464,
+      "grad_norm": 1.078532099723816,
+      "learning_rate": 5.490963769676927e-05,
+      "loss": 2.4782,
+      "step": 10552
+    },
+    {
+      "epoch": 0.649175688976378,
+      "grad_norm": 1.2425532341003418,
+      "learning_rate": 5.489237829337239e-05,
+      "loss": 1.98,
+      "step": 10553
+    },
+    {
+      "epoch": 0.6492372047244095,
+      "grad_norm": 1.5262258052825928,
+      "learning_rate": 5.4875120576749e-05,
+      "loss": 2.876,
+      "step": 10554
+    },
+    {
+      "epoch": 0.649298720472441,
+      "grad_norm": 1.6382309198379517,
+      "learning_rate": 5.485786454754449e-05,
+      "loss": 2.4798,
+      "step": 10555
+    },
+    {
+      "epoch": 0.6493602362204725,
+      "grad_norm": 1.203305721282959,
+      "learning_rate": 5.4840610206404106e-05,
+      "loss": 2.2569,
+      "step": 10556
+    },
+    {
+      "epoch": 0.6494217519685039,
+      "grad_norm": 1.071831464767456,
+      "learning_rate": 5.4823357553973144e-05,
+      "loss": 2.8233,
+      "step": 10557
+    },
+    {
+      "epoch": 0.6494832677165354,
+      "grad_norm": 1.2939215898513794,
+      "learning_rate": 5.480610659089664e-05,
+      "loss": 2.2317,
+      "step": 10558
+    },
+    {
+      "epoch": 0.6495447834645669,
+      "grad_norm": 1.1742585897445679,
+      "learning_rate": 5.478885731781979e-05,
+      "loss": 3.0992,
+      "step": 10559
+    },
+    {
+      "epoch": 0.6496062992125984,
+      "grad_norm": 1.163936972618103,
+      "learning_rate": 5.4771609735387575e-05,
+      "loss": 2.5359,
+      "step": 10560
+    },
+    {
+      "epoch": 0.64966781496063,
+      "grad_norm": 1.199110746383667,
+      "learning_rate": 5.4754363844244905e-05,
+      "loss": 2.769,
+      "step": 10561
+    },
+    {
+      "epoch": 0.6497293307086615,
+      "grad_norm": 1.197044014930725,
+      "learning_rate": 5.473711964503674e-05,
+      "loss": 2.4576,
+      "step": 10562
+    },
+    {
+      "epoch": 0.6497908464566929,
+      "grad_norm": 1.1779528856277466,
+      "learning_rate": 5.4719877138407896e-05,
+      "loss": 2.656,
+      "step": 10563
+    },
+    {
+      "epoch": 0.6498523622047244,
+      "grad_norm": 1.2033590078353882,
+      "learning_rate": 5.470263632500321e-05,
+      "loss": 2.6962,
+      "step": 10564
+    },
+    {
+      "epoch": 0.6499138779527559,
+      "grad_norm": 1.2091118097305298,
+      "learning_rate": 5.4685397205467305e-05,
+      "loss": 2.6324,
+      "step": 10565
+    },
+    {
+      "epoch": 0.6499753937007874,
+      "grad_norm": 1.3461220264434814,
+      "learning_rate": 5.466815978044488e-05,
+      "loss": 2.2583,
+      "step": 10566
+    },
+    {
+      "epoch": 0.6500369094488189,
+      "grad_norm": 1.2122828960418701,
+      "learning_rate": 5.465092405058048e-05,
+      "loss": 1.756,
+      "step": 10567
+    },
+    {
+      "epoch": 0.6500984251968503,
+      "grad_norm": 1.106953501701355,
+      "learning_rate": 5.4633690016518644e-05,
+      "loss": 2.4467,
+      "step": 10568
+    },
+    {
+      "epoch": 0.6501599409448819,
+      "grad_norm": 1.7400914430618286,
+      "learning_rate": 5.461645767890384e-05,
+      "loss": 2.1561,
+      "step": 10569
+    },
+    {
+      "epoch": 0.6502214566929134,
+      "grad_norm": 1.3059147596359253,
+      "learning_rate": 5.459922703838048e-05,
+      "loss": 2.4896,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6502829724409449,
+      "grad_norm": 1.1103229522705078,
+      "learning_rate": 5.458199809559283e-05,
+      "loss": 2.4553,
+      "step": 10571
+    },
+    {
+      "epoch": 0.6503444881889764,
+      "grad_norm": 1.5678291320800781,
+      "learning_rate": 5.456477085118523e-05,
+      "loss": 2.2528,
+      "step": 10572
+    },
+    {
+      "epoch": 0.6504060039370079,
+      "grad_norm": 1.2629272937774658,
+      "learning_rate": 5.454754530580182e-05,
+      "loss": 2.6203,
+      "step": 10573
+    },
+    {
+      "epoch": 0.6504675196850394,
+      "grad_norm": 1.0992947816848755,
+      "learning_rate": 5.453032146008676e-05,
+      "loss": 3.2546,
+      "step": 10574
+    },
+    {
+      "epoch": 0.6505290354330708,
+      "grad_norm": 1.302968144416809,
+      "learning_rate": 5.451309931468413e-05,
+      "loss": 2.3445,
+      "step": 10575
+    },
+    {
+      "epoch": 0.6505905511811023,
+      "grad_norm": 1.267256498336792,
+      "learning_rate": 5.4495878870237985e-05,
+      "loss": 2.2914,
+      "step": 10576
+    },
+    {
+      "epoch": 0.6506520669291339,
+      "grad_norm": 1.4068706035614014,
+      "learning_rate": 5.447866012739219e-05,
+      "loss": 2.154,
+      "step": 10577
+    },
+    {
+      "epoch": 0.6507135826771654,
+      "grad_norm": 1.529467225074768,
+      "learning_rate": 5.446144308679071e-05,
+      "loss": 2.247,
+      "step": 10578
+    },
+    {
+      "epoch": 0.6507750984251969,
+      "grad_norm": 1.2866579294204712,
+      "learning_rate": 5.444422774907729e-05,
+      "loss": 2.1947,
+      "step": 10579
+    },
+    {
+      "epoch": 0.6508366141732284,
+      "grad_norm": 1.2729465961456299,
+      "learning_rate": 5.442701411489573e-05,
+      "loss": 2.9731,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6508981299212598,
+      "grad_norm": 1.034032940864563,
+      "learning_rate": 5.4409802184889716e-05,
+      "loss": 2.9002,
+      "step": 10581
+    },
+    {
+      "epoch": 0.6509596456692913,
+      "grad_norm": 1.150286316871643,
+      "learning_rate": 5.439259195970292e-05,
+      "loss": 2.5285,
+      "step": 10582
+    },
+    {
+      "epoch": 0.6510211614173228,
+      "grad_norm": 1.2458866834640503,
+      "learning_rate": 5.4375383439978845e-05,
+      "loss": 2.8135,
+      "step": 10583
+    },
+    {
+      "epoch": 0.6510826771653543,
+      "grad_norm": 1.9218538999557495,
+      "learning_rate": 5.435817662636105e-05,
+      "loss": 2.6517,
+      "step": 10584
+    },
+    {
+      "epoch": 0.6511441929133859,
+      "grad_norm": 0.3047711253166199,
+      "learning_rate": 5.43409715194929e-05,
+      "loss": 0.0269,
+      "step": 10585
+    },
+    {
+      "epoch": 0.6512057086614174,
+      "grad_norm": 2.262274742126465,
+      "learning_rate": 5.432376812001785e-05,
+      "loss": 1.6751,
+      "step": 10586
+    },
+    {
+      "epoch": 0.6512672244094488,
+      "grad_norm": 1.326090693473816,
+      "learning_rate": 5.430656642857914e-05,
+      "loss": 3.6155,
+      "step": 10587
+    },
+    {
+      "epoch": 0.6513287401574803,
+      "grad_norm": 1.3823928833007812,
+      "learning_rate": 5.428936644582005e-05,
+      "loss": 2.9213,
+      "step": 10588
+    },
+    {
+      "epoch": 0.6513902559055118,
+      "grad_norm": 1.313279628753662,
+      "learning_rate": 5.4272168172383786e-05,
+      "loss": 2.4212,
+      "step": 10589
+    },
+    {
+      "epoch": 0.6514517716535433,
+      "grad_norm": 1.286077618598938,
+      "learning_rate": 5.425497160891346e-05,
+      "loss": 2.4423,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6515132874015748,
+      "grad_norm": 1.301202654838562,
+      "learning_rate": 5.423777675605208e-05,
+      "loss": 2.5855,
+      "step": 10591
+    },
+    {
+      "epoch": 0.6515748031496063,
+      "grad_norm": 1.2140475511550903,
+      "learning_rate": 5.422058361444272e-05,
+      "loss": 1.7276,
+      "step": 10592
+    },
+    {
+      "epoch": 0.6516363188976378,
+      "grad_norm": 1.2546757459640503,
+      "learning_rate": 5.420339218472822e-05,
+      "loss": 2.6971,
+      "step": 10593
+    },
+    {
+      "epoch": 0.6516978346456693,
+      "grad_norm": 1.310766339302063,
+      "learning_rate": 5.4186202467551486e-05,
+      "loss": 2.4788,
+      "step": 10594
+    },
+    {
+      "epoch": 0.6517593503937008,
+      "grad_norm": 1.2278507947921753,
+      "learning_rate": 5.41690144635553e-05,
+      "loss": 2.6543,
+      "step": 10595
+    },
+    {
+      "epoch": 0.6518208661417323,
+      "grad_norm": 1.4046400785446167,
+      "learning_rate": 5.415182817338247e-05,
+      "loss": 2.2904,
+      "step": 10596
+    },
+    {
+      "epoch": 0.6518823818897638,
+      "grad_norm": 1.1292082071304321,
+      "learning_rate": 5.4134643597675595e-05,
+      "loss": 2.5854,
+      "step": 10597
+    },
+    {
+      "epoch": 0.6519438976377953,
+      "grad_norm": 1.3355419635772705,
+      "learning_rate": 5.4117460737077265e-05,
+      "loss": 2.6233,
+      "step": 10598
+    },
+    {
+      "epoch": 0.6520054133858267,
+      "grad_norm": 1.3183377981185913,
+      "learning_rate": 5.410027959223005e-05,
+      "loss": 1.9218,
+      "step": 10599
+    },
+    {
+      "epoch": 0.6520669291338582,
+      "grad_norm": 1.2502188682556152,
+      "learning_rate": 5.408310016377643e-05,
+      "loss": 1.8925,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6521284448818898,
+      "grad_norm": 1.2384216785430908,
+      "learning_rate": 5.4065922452358874e-05,
+      "loss": 2.4862,
+      "step": 10601
+    },
+    {
+      "epoch": 0.6521899606299213,
+      "grad_norm": 1.3901035785675049,
+      "learning_rate": 5.4048746458619635e-05,
+      "loss": 2.4444,
+      "step": 10602
+    },
+    {
+      "epoch": 0.6522514763779528,
+      "grad_norm": 1.1518229246139526,
+      "learning_rate": 5.403157218320109e-05,
+      "loss": 2.6698,
+      "step": 10603
+    },
+    {
+      "epoch": 0.6523129921259843,
+      "grad_norm": 1.1991026401519775,
+      "learning_rate": 5.401439962674539e-05,
+      "loss": 2.2137,
+      "step": 10604
+    },
+    {
+      "epoch": 0.6523745078740157,
+      "grad_norm": 1.4119088649749756,
+      "learning_rate": 5.399722878989475e-05,
+      "loss": 2.2349,
+      "step": 10605
+    },
+    {
+      "epoch": 0.6524360236220472,
+      "grad_norm": 1.2300037145614624,
+      "learning_rate": 5.398005967329116e-05,
+      "loss": 2.3393,
+      "step": 10606
+    },
+    {
+      "epoch": 0.6524975393700787,
+      "grad_norm": 1.247649073600769,
+      "learning_rate": 5.3962892277576806e-05,
+      "loss": 2.6664,
+      "step": 10607
+    },
+    {
+      "epoch": 0.6525590551181102,
+      "grad_norm": 1.2306113243103027,
+      "learning_rate": 5.394572660339352e-05,
+      "loss": 2.2044,
+      "step": 10608
+    },
+    {
+      "epoch": 0.6526205708661418,
+      "grad_norm": 1.0454108715057373,
+      "learning_rate": 5.3928562651383305e-05,
+      "loss": 2.3741,
+      "step": 10609
+    },
+    {
+      "epoch": 0.6526820866141733,
+      "grad_norm": 1.095218539237976,
+      "learning_rate": 5.39114004221879e-05,
+      "loss": 2.3907,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6527436023622047,
+      "grad_norm": 1.3889614343643188,
+      "learning_rate": 5.3894239916449164e-05,
+      "loss": 1.8448,
+      "step": 10611
+    },
+    {
+      "epoch": 0.6528051181102362,
+      "grad_norm": 1.5182931423187256,
+      "learning_rate": 5.3877081134808716e-05,
+      "loss": 2.8537,
+      "step": 10612
+    },
+    {
+      "epoch": 0.6528666338582677,
+      "grad_norm": 1.1572792530059814,
+      "learning_rate": 5.3859924077908256e-05,
+      "loss": 2.9487,
+      "step": 10613
+    },
+    {
+      "epoch": 0.6529281496062992,
+      "grad_norm": 1.0570787191390991,
+      "learning_rate": 5.384276874638934e-05,
+      "loss": 1.7573,
+      "step": 10614
+    },
+    {
+      "epoch": 0.6529896653543307,
+      "grad_norm": 1.484793782234192,
+      "learning_rate": 5.382561514089354e-05,
+      "loss": 2.3499,
+      "step": 10615
+    },
+    {
+      "epoch": 0.6530511811023622,
+      "grad_norm": 1.1552637815475464,
+      "learning_rate": 5.380846326206221e-05,
+      "loss": 1.875,
+      "step": 10616
+    },
+    {
+      "epoch": 0.6531126968503937,
+      "grad_norm": 1.23251473903656,
+      "learning_rate": 5.3791313110536834e-05,
+      "loss": 2.2319,
+      "step": 10617
+    },
+    {
+      "epoch": 0.6531742125984252,
+      "grad_norm": 1.2432875633239746,
+      "learning_rate": 5.377416468695864e-05,
+      "loss": 2.817,
+      "step": 10618
+    },
+    {
+      "epoch": 0.6532357283464567,
+      "grad_norm": 1.1615185737609863,
+      "learning_rate": 5.375701799196892e-05,
+      "loss": 2.7477,
+      "step": 10619
+    },
+    {
+      "epoch": 0.6532972440944882,
+      "grad_norm": 1.3300831317901611,
+      "learning_rate": 5.373987302620886e-05,
+      "loss": 2.678,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6533587598425197,
+      "grad_norm": 1.3375121355056763,
+      "learning_rate": 5.3722729790319645e-05,
+      "loss": 2.0528,
+      "step": 10621
+    },
+    {
+      "epoch": 0.6534202755905512,
+      "grad_norm": 1.081632375717163,
+      "learning_rate": 5.370558828494225e-05,
+      "loss": 2.959,
+      "step": 10622
+    },
+    {
+      "epoch": 0.6534817913385826,
+      "grad_norm": 1.1320751905441284,
+      "learning_rate": 5.368844851071775e-05,
+      "loss": 2.6166,
+      "step": 10623
+    },
+    {
+      "epoch": 0.6535433070866141,
+      "grad_norm": 1.1104224920272827,
+      "learning_rate": 5.367131046828699e-05,
+      "loss": 2.6814,
+      "step": 10624
+    },
+    {
+      "epoch": 0.6536048228346457,
+      "grad_norm": 1.3165977001190186,
+      "learning_rate": 5.365417415829088e-05,
+      "loss": 2.4738,
+      "step": 10625
+    },
+    {
+      "epoch": 0.6536663385826772,
+      "grad_norm": 1.1792192459106445,
+      "learning_rate": 5.3637039581370226e-05,
+      "loss": 2.0492,
+      "step": 10626
+    },
+    {
+      "epoch": 0.6537278543307087,
+      "grad_norm": 1.1577813625335693,
+      "learning_rate": 5.3619906738165816e-05,
+      "loss": 1.9697,
+      "step": 10627
+    },
+    {
+      "epoch": 0.6537893700787402,
+      "grad_norm": 1.3493719100952148,
+      "learning_rate": 5.360277562931821e-05,
+      "loss": 2.7058,
+      "step": 10628
+    },
+    {
+      "epoch": 0.6538508858267716,
+      "grad_norm": 1.193872332572937,
+      "learning_rate": 5.3585646255468135e-05,
+      "loss": 2.4382,
+      "step": 10629
+    },
+    {
+      "epoch": 0.6539124015748031,
+      "grad_norm": 1.338686466217041,
+      "learning_rate": 5.356851861725604e-05,
+      "loss": 2.2181,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6539739173228346,
+      "grad_norm": 0.9732441306114197,
+      "learning_rate": 5.355139271532247e-05,
+      "loss": 1.8448,
+      "step": 10631
+    },
+    {
+      "epoch": 0.6540354330708661,
+      "grad_norm": 1.21633780002594,
+      "learning_rate": 5.353426855030773e-05,
+      "loss": 2.3192,
+      "step": 10632
+    },
+    {
+      "epoch": 0.6540969488188977,
+      "grad_norm": 1.039315104484558,
+      "learning_rate": 5.3517146122852326e-05,
+      "loss": 3.0477,
+      "step": 10633
+    },
+    {
+      "epoch": 0.6541584645669292,
+      "grad_norm": 1.5558083057403564,
+      "learning_rate": 5.350002543359646e-05,
+      "loss": 2.4631,
+      "step": 10634
+    },
+    {
+      "epoch": 0.6542199803149606,
+      "grad_norm": 1.308925986289978,
+      "learning_rate": 5.348290648318031e-05,
+      "loss": 2.3399,
+      "step": 10635
+    },
+    {
+      "epoch": 0.6542814960629921,
+      "grad_norm": 1.1069941520690918,
+      "learning_rate": 5.3465789272244116e-05,
+      "loss": 3.3031,
+      "step": 10636
+    },
+    {
+      "epoch": 0.6543430118110236,
+      "grad_norm": 1.2356458902359009,
+      "learning_rate": 5.344867380142788e-05,
+      "loss": 2.4194,
+      "step": 10637
+    },
+    {
+      "epoch": 0.6544045275590551,
+      "grad_norm": 1.351383924484253,
+      "learning_rate": 5.343156007137167e-05,
+      "loss": 2.8508,
+      "step": 10638
+    },
+    {
+      "epoch": 0.6544660433070866,
+      "grad_norm": 1.1779519319534302,
+      "learning_rate": 5.341444808271543e-05,
+      "loss": 2.7506,
+      "step": 10639
+    },
+    {
+      "epoch": 0.6545275590551181,
+      "grad_norm": 1.0210636854171753,
+      "learning_rate": 5.3397337836099106e-05,
+      "loss": 2.4687,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6545890748031497,
+      "grad_norm": 1.382954716682434,
+      "learning_rate": 5.3380229332162446e-05,
+      "loss": 2.7014,
+      "step": 10641
+    },
+    {
+      "epoch": 0.6546505905511811,
+      "grad_norm": 1.113101601600647,
+      "learning_rate": 5.33631225715453e-05,
+      "loss": 2.9806,
+      "step": 10642
+    },
+    {
+      "epoch": 0.6547121062992126,
+      "grad_norm": 1.2273107767105103,
+      "learning_rate": 5.334601755488726e-05,
+      "loss": 2.6606,
+      "step": 10643
+    },
+    {
+      "epoch": 0.6547736220472441,
+      "grad_norm": 1.3000305891036987,
+      "learning_rate": 5.332891428282802e-05,
+      "loss": 1.3136,
+      "step": 10644
+    },
+    {
+      "epoch": 0.6548351377952756,
+      "grad_norm": 1.2516337633132935,
+      "learning_rate": 5.331181275600714e-05,
+      "loss": 2.7703,
+      "step": 10645
+    },
+    {
+      "epoch": 0.6548966535433071,
+      "grad_norm": 1.3572369813919067,
+      "learning_rate": 5.329471297506416e-05,
+      "loss": 2.2554,
+      "step": 10646
+    },
+    {
+      "epoch": 0.6549581692913385,
+      "grad_norm": 1.3297009468078613,
+      "learning_rate": 5.3277614940638454e-05,
+      "loss": 2.1001,
+      "step": 10647
+    },
+    {
+      "epoch": 0.65501968503937,
+      "grad_norm": 1.5870095491409302,
+      "learning_rate": 5.326051865336944e-05,
+      "loss": 2.3181,
+      "step": 10648
+    },
+    {
+      "epoch": 0.6550812007874016,
+      "grad_norm": 1.3694627285003662,
+      "learning_rate": 5.3243424113896365e-05,
+      "loss": 2.1999,
+      "step": 10649
+    },
+    {
+      "epoch": 0.6551427165354331,
+      "grad_norm": 1.2796106338500977,
+      "learning_rate": 5.3226331322858514e-05,
+      "loss": 2.6295,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6552042322834646,
+      "grad_norm": 1.3538422584533691,
+      "learning_rate": 5.3209240280895046e-05,
+      "loss": 3.1607,
+      "step": 10651
+    },
+    {
+      "epoch": 0.6552657480314961,
+      "grad_norm": 1.0823967456817627,
+      "learning_rate": 5.319215098864513e-05,
+      "loss": 2.3792,
+      "step": 10652
+    },
+    {
+      "epoch": 0.6553272637795275,
+      "grad_norm": 1.5794247388839722,
+      "learning_rate": 5.317506344674771e-05,
+      "loss": 1.8417,
+      "step": 10653
+    },
+    {
+      "epoch": 0.655388779527559,
+      "grad_norm": 1.2769852876663208,
+      "learning_rate": 5.315797765584185e-05,
+      "loss": 2.5336,
+      "step": 10654
+    },
+    {
+      "epoch": 0.6554502952755905,
+      "grad_norm": 1.9083820581436157,
+      "learning_rate": 5.31408936165664e-05,
+      "loss": 2.2595,
+      "step": 10655
+    },
+    {
+      "epoch": 0.655511811023622,
+      "grad_norm": 1.7309584617614746,
+      "learning_rate": 5.312381132956027e-05,
+      "loss": 2.0493,
+      "step": 10656
+    },
+    {
+      "epoch": 0.6555733267716536,
+      "grad_norm": 1.0591477155685425,
+      "learning_rate": 5.3106730795462166e-05,
+      "loss": 3.178,
+      "step": 10657
+    },
+    {
+      "epoch": 0.6556348425196851,
+      "grad_norm": 1.6291264295578003,
+      "learning_rate": 5.308965201491085e-05,
+      "loss": 2.5384,
+      "step": 10658
+    },
+    {
+      "epoch": 0.6556963582677166,
+      "grad_norm": 1.1161224842071533,
+      "learning_rate": 5.3072574988544966e-05,
+      "loss": 3.0381,
+      "step": 10659
+    },
+    {
+      "epoch": 0.655757874015748,
+      "grad_norm": 1.3584715127944946,
+      "learning_rate": 5.305549971700314e-05,
+      "loss": 2.2213,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6558193897637795,
+      "grad_norm": 1.490796685218811,
+      "learning_rate": 5.303842620092382e-05,
+      "loss": 1.7401,
+      "step": 10661
+    },
+    {
+      "epoch": 0.655880905511811,
+      "grad_norm": 1.3031976222991943,
+      "learning_rate": 5.302135444094554e-05,
+      "loss": 2.2281,
+      "step": 10662
+    },
+    {
+      "epoch": 0.6559424212598425,
+      "grad_norm": 1.3590469360351562,
+      "learning_rate": 5.3004284437706586e-05,
+      "loss": 1.8536,
+      "step": 10663
+    },
+    {
+      "epoch": 0.656003937007874,
+      "grad_norm": 1.408909559249878,
+      "learning_rate": 5.298721619184534e-05,
+      "loss": 2.7479,
+      "step": 10664
+    },
+    {
+      "epoch": 0.6560654527559056,
+      "grad_norm": 1.5089309215545654,
+      "learning_rate": 5.2970149704000074e-05,
+      "loss": 2.1648,
+      "step": 10665
+    },
+    {
+      "epoch": 0.656126968503937,
+      "grad_norm": 1.2181538343429565,
+      "learning_rate": 5.2953084974808995e-05,
+      "loss": 3.1049,
+      "step": 10666
+    },
+    {
+      "epoch": 0.6561884842519685,
+      "grad_norm": 1.289602518081665,
+      "learning_rate": 5.293602200491017e-05,
+      "loss": 2.9102,
+      "step": 10667
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 1.2312513589859009,
+      "learning_rate": 5.291896079494172e-05,
+      "loss": 2.1973,
+      "step": 10668
+    },
+    {
+      "epoch": 0.6563115157480315,
+      "grad_norm": 1.224658489227295,
+      "learning_rate": 5.290190134554156e-05,
+      "loss": 1.6609,
+      "step": 10669
+    },
+    {
+      "epoch": 0.656373031496063,
+      "grad_norm": 1.2067112922668457,
+      "learning_rate": 5.288484365734765e-05,
+      "loss": 2.6543,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6564345472440944,
+      "grad_norm": 1.1949599981307983,
+      "learning_rate": 5.286778773099792e-05,
+      "loss": 2.4209,
+      "step": 10671
+    },
+    {
+      "epoch": 0.656496062992126,
+      "grad_norm": 1.646979808807373,
+      "learning_rate": 5.2850733567130085e-05,
+      "loss": 1.9101,
+      "step": 10672
+    },
+    {
+      "epoch": 0.6565575787401575,
+      "grad_norm": 1.4489370584487915,
+      "learning_rate": 5.283368116638192e-05,
+      "loss": 2.5679,
+      "step": 10673
+    },
+    {
+      "epoch": 0.656619094488189,
+      "grad_norm": 1.296639323234558,
+      "learning_rate": 5.281663052939105e-05,
+      "loss": 2.3878,
+      "step": 10674
+    },
+    {
+      "epoch": 0.6566806102362205,
+      "grad_norm": 1.515486478805542,
+      "learning_rate": 5.279958165679513e-05,
+      "loss": 1.7637,
+      "step": 10675
+    },
+    {
+      "epoch": 0.656742125984252,
+      "grad_norm": 1.1737570762634277,
+      "learning_rate": 5.278253454923158e-05,
+      "loss": 2.6752,
+      "step": 10676
+    },
+    {
+      "epoch": 0.6568036417322834,
+      "grad_norm": 1.1428142786026,
+      "learning_rate": 5.276548920733802e-05,
+      "loss": 2.5928,
+      "step": 10677
+    },
+    {
+      "epoch": 0.6568651574803149,
+      "grad_norm": 1.4634525775909424,
+      "learning_rate": 5.274844563175174e-05,
+      "loss": 2.6279,
+      "step": 10678
+    },
+    {
+      "epoch": 0.6569266732283464,
+      "grad_norm": 2.25390887260437,
+      "learning_rate": 5.273140382311015e-05,
+      "loss": 1.9073,
+      "step": 10679
+    },
+    {
+      "epoch": 0.656988188976378,
+      "grad_norm": 1.0792981386184692,
+      "learning_rate": 5.2714363782050447e-05,
+      "loss": 2.9142,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6570497047244095,
+      "grad_norm": 1.250333547592163,
+      "learning_rate": 5.2697325509209905e-05,
+      "loss": 2.2051,
+      "step": 10681
+    },
+    {
+      "epoch": 0.657111220472441,
+      "grad_norm": 1.4579371213912964,
+      "learning_rate": 5.268028900522558e-05,
+      "loss": 2.3158,
+      "step": 10682
+    },
+    {
+      "epoch": 0.6571727362204725,
+      "grad_norm": 1.1378724575042725,
+      "learning_rate": 5.266325427073459e-05,
+      "loss": 2.6913,
+      "step": 10683
+    },
+    {
+      "epoch": 0.6572342519685039,
+      "grad_norm": 1.227869987487793,
+      "learning_rate": 5.264622130637392e-05,
+      "loss": 2.6468,
+      "step": 10684
+    },
+    {
+      "epoch": 0.6572957677165354,
+      "grad_norm": 1.213568091392517,
+      "learning_rate": 5.262919011278057e-05,
+      "loss": 2.0565,
+      "step": 10685
+    },
+    {
+      "epoch": 0.6573572834645669,
+      "grad_norm": 1.3986119031906128,
+      "learning_rate": 5.2612160690591317e-05,
+      "loss": 2.3502,
+      "step": 10686
+    },
+    {
+      "epoch": 0.6574187992125984,
+      "grad_norm": 1.2438442707061768,
+      "learning_rate": 5.2595133040443056e-05,
+      "loss": 3.2201,
+      "step": 10687
+    },
+    {
+      "epoch": 0.65748031496063,
+      "grad_norm": 1.8161461353302002,
+      "learning_rate": 5.257810716297244e-05,
+      "loss": 1.6265,
+      "step": 10688
+    },
+    {
+      "epoch": 0.6575418307086615,
+      "grad_norm": 1.149717926979065,
+      "learning_rate": 5.256108305881619e-05,
+      "loss": 1.5148,
+      "step": 10689
+    },
+    {
+      "epoch": 0.6576033464566929,
+      "grad_norm": 1.1838204860687256,
+      "learning_rate": 5.254406072861091e-05,
+      "loss": 2.333,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6576648622047244,
+      "grad_norm": 1.182685136795044,
+      "learning_rate": 5.252704017299317e-05,
+      "loss": 2.7848,
+      "step": 10691
+    },
+    {
+      "epoch": 0.6577263779527559,
+      "grad_norm": 1.168862223625183,
+      "learning_rate": 5.251002139259937e-05,
+      "loss": 2.1906,
+      "step": 10692
+    },
+    {
+      "epoch": 0.6577878937007874,
+      "grad_norm": 1.288252830505371,
+      "learning_rate": 5.2493004388066015e-05,
+      "loss": 2.9877,
+      "step": 10693
+    },
+    {
+      "epoch": 0.6578494094488189,
+      "grad_norm": 1.3099123239517212,
+      "learning_rate": 5.2475989160029325e-05,
+      "loss": 2.3542,
+      "step": 10694
+    },
+    {
+      "epoch": 0.6579109251968503,
+      "grad_norm": 1.2529489994049072,
+      "learning_rate": 5.2458975709125655e-05,
+      "loss": 1.8984,
+      "step": 10695
+    },
+    {
+      "epoch": 0.6579724409448819,
+      "grad_norm": 1.5020006895065308,
+      "learning_rate": 5.24419640359912e-05,
+      "loss": 2.0827,
+      "step": 10696
+    },
+    {
+      "epoch": 0.6580339566929134,
+      "grad_norm": 1.2476330995559692,
+      "learning_rate": 5.242495414126212e-05,
+      "loss": 2.559,
+      "step": 10697
+    },
+    {
+      "epoch": 0.6580954724409449,
+      "grad_norm": 1.2071291208267212,
+      "learning_rate": 5.240794602557444e-05,
+      "loss": 2.8327,
+      "step": 10698
+    },
+    {
+      "epoch": 0.6581569881889764,
+      "grad_norm": 1.1584715843200684,
+      "learning_rate": 5.2390939689564244e-05,
+      "loss": 2.2079,
+      "step": 10699
+    },
+    {
+      "epoch": 0.6582185039370079,
+      "grad_norm": 1.1098252534866333,
+      "learning_rate": 5.237393513386739e-05,
+      "loss": 2.3519,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6582800196850394,
+      "grad_norm": 1.3832684755325317,
+      "learning_rate": 5.2356932359119826e-05,
+      "loss": 2.7401,
+      "step": 10701
+    },
+    {
+      "epoch": 0.6583415354330708,
+      "grad_norm": 1.0987809896469116,
+      "learning_rate": 5.233993136595724e-05,
+      "loss": 2.5873,
+      "step": 10702
+    },
+    {
+      "epoch": 0.6584030511811023,
+      "grad_norm": 1.4232206344604492,
+      "learning_rate": 5.232293215501555e-05,
+      "loss": 2.8013,
+      "step": 10703
+    },
+    {
+      "epoch": 0.6584645669291339,
+      "grad_norm": 1.078839898109436,
+      "learning_rate": 5.2305934726930306e-05,
+      "loss": 2.4302,
+      "step": 10704
+    },
+    {
+      "epoch": 0.6585260826771654,
+      "grad_norm": 1.128528356552124,
+      "learning_rate": 5.2288939082337186e-05,
+      "loss": 2.7558,
+      "step": 10705
+    },
+    {
+      "epoch": 0.6585875984251969,
+      "grad_norm": 1.2938066720962524,
+      "learning_rate": 5.22719452218717e-05,
+      "loss": 1.696,
+      "step": 10706
+    },
+    {
+      "epoch": 0.6586491141732284,
+      "grad_norm": 1.4400185346603394,
+      "learning_rate": 5.225495314616928e-05,
+      "loss": 2.2873,
+      "step": 10707
+    },
+    {
+      "epoch": 0.6587106299212598,
+      "grad_norm": 1.3730417490005493,
+      "learning_rate": 5.2237962855865386e-05,
+      "loss": 2.3477,
+      "step": 10708
+    },
+    {
+      "epoch": 0.6587721456692913,
+      "grad_norm": 1.4268970489501953,
+      "learning_rate": 5.2220974351595356e-05,
+      "loss": 1.7037,
+      "step": 10709
+    },
+    {
+      "epoch": 0.6588336614173228,
+      "grad_norm": 1.308838963508606,
+      "learning_rate": 5.22039876339945e-05,
+      "loss": 2.5651,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6588951771653543,
+      "grad_norm": 1.2436983585357666,
+      "learning_rate": 5.218700270369794e-05,
+      "loss": 2.4572,
+      "step": 10711
+    },
+    {
+      "epoch": 0.6589566929133859,
+      "grad_norm": 1.1445012092590332,
+      "learning_rate": 5.217001956134091e-05,
+      "loss": 2.6092,
+      "step": 10712
+    },
+    {
+      "epoch": 0.6590182086614174,
+      "grad_norm": 1.1246190071105957,
+      "learning_rate": 5.2153038207558415e-05,
+      "loss": 2.9147,
+      "step": 10713
+    },
+    {
+      "epoch": 0.6590797244094488,
+      "grad_norm": 1.1137704849243164,
+      "learning_rate": 5.213605864298548e-05,
+      "loss": 2.8441,
+      "step": 10714
+    },
+    {
+      "epoch": 0.6591412401574803,
+      "grad_norm": 1.293726921081543,
+      "learning_rate": 5.2119080868257044e-05,
+      "loss": 2.361,
+      "step": 10715
+    },
+    {
+      "epoch": 0.6592027559055118,
+      "grad_norm": 1.2179419994354248,
+      "learning_rate": 5.210210488400804e-05,
+      "loss": 2.6016,
+      "step": 10716
+    },
+    {
+      "epoch": 0.6592642716535433,
+      "grad_norm": 1.2452737092971802,
+      "learning_rate": 5.208513069087318e-05,
+      "loss": 2.7047,
+      "step": 10717
+    },
+    {
+      "epoch": 0.6593257874015748,
+      "grad_norm": 1.293905258178711,
+      "learning_rate": 5.20681582894873e-05,
+      "loss": 2.5087,
+      "step": 10718
+    },
+    {
+      "epoch": 0.6593873031496063,
+      "grad_norm": 1.2083128690719604,
+      "learning_rate": 5.205118768048499e-05,
+      "loss": 2.2596,
+      "step": 10719
+    },
+    {
+      "epoch": 0.6594488188976378,
+      "grad_norm": 1.3131811618804932,
+      "learning_rate": 5.203421886450093e-05,
+      "loss": 2.6031,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6595103346456693,
+      "grad_norm": 1.2405710220336914,
+      "learning_rate": 5.201725184216952e-05,
+      "loss": 2.4198,
+      "step": 10721
+    },
+    {
+      "epoch": 0.6595718503937008,
+      "grad_norm": 1.5388637781143188,
+      "learning_rate": 5.2000286614125435e-05,
+      "loss": 1.6394,
+      "step": 10722
+    },
+    {
+      "epoch": 0.6596333661417323,
+      "grad_norm": 1.3226144313812256,
+      "learning_rate": 5.198332318100293e-05,
+      "loss": 2.5452,
+      "step": 10723
+    },
+    {
+      "epoch": 0.6596948818897638,
+      "grad_norm": 1.3175575733184814,
+      "learning_rate": 5.1966361543436435e-05,
+      "loss": 2.4229,
+      "step": 10724
+    },
+    {
+      "epoch": 0.6597563976377953,
+      "grad_norm": 1.3757050037384033,
+      "learning_rate": 5.194940170206013e-05,
+      "loss": 2.5956,
+      "step": 10725
+    },
+    {
+      "epoch": 0.6598179133858267,
+      "grad_norm": 1.3764300346374512,
+      "learning_rate": 5.193244365750829e-05,
+      "loss": 2.1745,
+      "step": 10726
+    },
+    {
+      "epoch": 0.6598794291338582,
+      "grad_norm": 1.2200627326965332,
+      "learning_rate": 5.191548741041501e-05,
+      "loss": 2.4958,
+      "step": 10727
+    },
+    {
+      "epoch": 0.6599409448818898,
+      "grad_norm": 1.6334590911865234,
+      "learning_rate": 5.189853296141436e-05,
+      "loss": 2.7032,
+      "step": 10728
+    },
+    {
+      "epoch": 0.6600024606299213,
+      "grad_norm": 1.1826916933059692,
+      "learning_rate": 5.1881580311140355e-05,
+      "loss": 2.8649,
+      "step": 10729
+    },
+    {
+      "epoch": 0.6600639763779528,
+      "grad_norm": 1.2553688287734985,
+      "learning_rate": 5.1864629460226974e-05,
+      "loss": 2.6618,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6601254921259843,
+      "grad_norm": 1.284104824066162,
+      "learning_rate": 5.184768040930801e-05,
+      "loss": 2.5707,
+      "step": 10731
+    },
+    {
+      "epoch": 0.6601870078740157,
+      "grad_norm": 1.3201652765274048,
+      "learning_rate": 5.183073315901732e-05,
+      "loss": 2.7264,
+      "step": 10732
+    },
+    {
+      "epoch": 0.6602485236220472,
+      "grad_norm": 1.2165430784225464,
+      "learning_rate": 5.181378770998857e-05,
+      "loss": 2.0186,
+      "step": 10733
+    },
+    {
+      "epoch": 0.6603100393700787,
+      "grad_norm": 1.3396203517913818,
+      "learning_rate": 5.1796844062855475e-05,
+      "loss": 1.9543,
+      "step": 10734
+    },
+    {
+      "epoch": 0.6603715551181102,
+      "grad_norm": 1.3140437602996826,
+      "learning_rate": 5.177990221825162e-05,
+      "loss": 2.5284,
+      "step": 10735
+    },
+    {
+      "epoch": 0.6604330708661418,
+      "grad_norm": 1.452309250831604,
+      "learning_rate": 5.176296217681058e-05,
+      "loss": 2.2244,
+      "step": 10736
+    },
+    {
+      "epoch": 0.6604945866141733,
+      "grad_norm": 1.1423903703689575,
+      "learning_rate": 5.174602393916574e-05,
+      "loss": 2.6786,
+      "step": 10737
+    },
+    {
+      "epoch": 0.6605561023622047,
+      "grad_norm": 1.289901852607727,
+      "learning_rate": 5.172908750595057e-05,
+      "loss": 2.6243,
+      "step": 10738
+    },
+    {
+      "epoch": 0.6606176181102362,
+      "grad_norm": 1.3368136882781982,
+      "learning_rate": 5.1712152877798314e-05,
+      "loss": 2.1773,
+      "step": 10739
+    },
+    {
+      "epoch": 0.6606791338582677,
+      "grad_norm": 1.210960030555725,
+      "learning_rate": 5.1695220055342294e-05,
+      "loss": 2.8329,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6607406496062992,
+      "grad_norm": 1.3882012367248535,
+      "learning_rate": 5.167828903921568e-05,
+      "loss": 1.8417,
+      "step": 10741
+    },
+    {
+      "epoch": 0.6608021653543307,
+      "grad_norm": 1.4131888151168823,
+      "learning_rate": 5.1661359830051645e-05,
+      "loss": 2.3266,
+      "step": 10742
+    },
+    {
+      "epoch": 0.6608636811023622,
+      "grad_norm": 1.224347710609436,
+      "learning_rate": 5.164443242848323e-05,
+      "loss": 2.6197,
+      "step": 10743
+    },
+    {
+      "epoch": 0.6609251968503937,
+      "grad_norm": 1.3466224670410156,
+      "learning_rate": 5.1627506835143345e-05,
+      "loss": 2.4003,
+      "step": 10744
+    },
+    {
+      "epoch": 0.6609867125984252,
+      "grad_norm": 1.3424056768417358,
+      "learning_rate": 5.161058305066501e-05,
+      "loss": 2.5432,
+      "step": 10745
+    },
+    {
+      "epoch": 0.6610482283464567,
+      "grad_norm": 1.396362543106079,
+      "learning_rate": 5.1593661075681e-05,
+      "loss": 2.8615,
+      "step": 10746
+    },
+    {
+      "epoch": 0.6611097440944882,
+      "grad_norm": 1.1575514078140259,
+      "learning_rate": 5.1576740910824164e-05,
+      "loss": 2.685,
+      "step": 10747
+    },
+    {
+      "epoch": 0.6611712598425197,
+      "grad_norm": 1.3703327178955078,
+      "learning_rate": 5.155982255672718e-05,
+      "loss": 2.315,
+      "step": 10748
+    },
+    {
+      "epoch": 0.6612327755905512,
+      "grad_norm": 1.4593313932418823,
+      "learning_rate": 5.154290601402277e-05,
+      "loss": 2.7415,
+      "step": 10749
+    },
+    {
+      "epoch": 0.6612942913385826,
+      "grad_norm": 1.2439932823181152,
+      "learning_rate": 5.1525991283343434e-05,
+      "loss": 2.141,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6613558070866141,
+      "grad_norm": 1.3340901136398315,
+      "learning_rate": 5.150907836532177e-05,
+      "loss": 2.5459,
+      "step": 10751
+    },
+    {
+      "epoch": 0.6614173228346457,
+      "grad_norm": 1.2621408700942993,
+      "learning_rate": 5.149216726059014e-05,
+      "loss": 1.7769,
+      "step": 10752
+    },
+    {
+      "epoch": 0.6614788385826772,
+      "grad_norm": 1.2665208578109741,
+      "learning_rate": 5.1475257969780965e-05,
+      "loss": 2.6397,
+      "step": 10753
+    },
+    {
+      "epoch": 0.6615403543307087,
+      "grad_norm": 1.1335926055908203,
+      "learning_rate": 5.145835049352655e-05,
+      "loss": 2.7172,
+      "step": 10754
+    },
+    {
+      "epoch": 0.6616018700787402,
+      "grad_norm": 1.2990705966949463,
+      "learning_rate": 5.144144483245921e-05,
+      "loss": 2.2704,
+      "step": 10755
+    },
+    {
+      "epoch": 0.6616633858267716,
+      "grad_norm": 2.6633405685424805,
+      "learning_rate": 5.142454098721101e-05,
+      "loss": 2.2725,
+      "step": 10756
+    },
+    {
+      "epoch": 0.6617249015748031,
+      "grad_norm": 0.9321319460868835,
+      "learning_rate": 5.1407638958414164e-05,
+      "loss": 0.7263,
+      "step": 10757
+    },
+    {
+      "epoch": 0.6617864173228346,
+      "grad_norm": 1.379913568496704,
+      "learning_rate": 5.139073874670061e-05,
+      "loss": 2.4932,
+      "step": 10758
+    },
+    {
+      "epoch": 0.6618479330708661,
+      "grad_norm": 1.2416187524795532,
+      "learning_rate": 5.137384035270237e-05,
+      "loss": 2.3673,
+      "step": 10759
+    },
+    {
+      "epoch": 0.6619094488188977,
+      "grad_norm": 1.3388009071350098,
+      "learning_rate": 5.135694377705137e-05,
+      "loss": 2.5981,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6619709645669292,
+      "grad_norm": 1.350550651550293,
+      "learning_rate": 5.1340049020379464e-05,
+      "loss": 2.3937,
+      "step": 10761
+    },
+    {
+      "epoch": 0.6620324803149606,
+      "grad_norm": 1.5334696769714355,
+      "learning_rate": 5.1323156083318345e-05,
+      "loss": 2.1755,
+      "step": 10762
+    },
+    {
+      "epoch": 0.6620939960629921,
+      "grad_norm": 1.1031583547592163,
+      "learning_rate": 5.130626496649982e-05,
+      "loss": 2.4117,
+      "step": 10763
+    },
+    {
+      "epoch": 0.6621555118110236,
+      "grad_norm": 1.4201834201812744,
+      "learning_rate": 5.12893756705554e-05,
+      "loss": 2.9552,
+      "step": 10764
+    },
+    {
+      "epoch": 0.6622170275590551,
+      "grad_norm": 1.0882256031036377,
+      "learning_rate": 5.127248819611672e-05,
+      "loss": 2.9142,
+      "step": 10765
+    },
+    {
+      "epoch": 0.6622785433070866,
+      "grad_norm": 1.4735019207000732,
+      "learning_rate": 5.125560254381526e-05,
+      "loss": 2.4531,
+      "step": 10766
+    },
+    {
+      "epoch": 0.6623400590551181,
+      "grad_norm": 1.1840914487838745,
+      "learning_rate": 5.1238718714282505e-05,
+      "loss": 2.4473,
+      "step": 10767
+    },
+    {
+      "epoch": 0.6624015748031497,
+      "grad_norm": 1.3189023733139038,
+      "learning_rate": 5.122183670814973e-05,
+      "loss": 2.4075,
+      "step": 10768
+    },
+    {
+      "epoch": 0.6624630905511811,
+      "grad_norm": 1.2070316076278687,
+      "learning_rate": 5.120495652604831e-05,
+      "loss": 2.6462,
+      "step": 10769
+    },
+    {
+      "epoch": 0.6625246062992126,
+      "grad_norm": 1.1294593811035156,
+      "learning_rate": 5.118807816860939e-05,
+      "loss": 2.9605,
+      "step": 10770
+    },
+    {
+      "epoch": 0.6625861220472441,
+      "grad_norm": 1.194267988204956,
+      "learning_rate": 5.117120163646421e-05,
+      "loss": 2.0464,
+      "step": 10771
+    },
+    {
+      "epoch": 0.6626476377952756,
+      "grad_norm": 1.1311472654342651,
+      "learning_rate": 5.115432693024376e-05,
+      "loss": 3.0052,
+      "step": 10772
+    },
+    {
+      "epoch": 0.6627091535433071,
+      "grad_norm": 1.3056628704071045,
+      "learning_rate": 5.1137454050579106e-05,
+      "loss": 2.3017,
+      "step": 10773
+    },
+    {
+      "epoch": 0.6627706692913385,
+      "grad_norm": 1.3296836614608765,
+      "learning_rate": 5.112058299810121e-05,
+      "loss": 2.7242,
+      "step": 10774
+    },
+    {
+      "epoch": 0.66283218503937,
+      "grad_norm": 1.415744423866272,
+      "learning_rate": 5.110371377344099e-05,
+      "loss": 2.2476,
+      "step": 10775
+    },
+    {
+      "epoch": 0.6628937007874016,
+      "grad_norm": 1.2013813257217407,
+      "learning_rate": 5.108684637722919e-05,
+      "loss": 3.1045,
+      "step": 10776
+    },
+    {
+      "epoch": 0.6629552165354331,
+      "grad_norm": 1.2173823118209839,
+      "learning_rate": 5.106998081009662e-05,
+      "loss": 2.406,
+      "step": 10777
+    },
+    {
+      "epoch": 0.6630167322834646,
+      "grad_norm": 1.771910548210144,
+      "learning_rate": 5.105311707267388e-05,
+      "loss": 2.5943,
+      "step": 10778
+    },
+    {
+      "epoch": 0.6630782480314961,
+      "grad_norm": 1.140499234199524,
+      "learning_rate": 5.103625516559165e-05,
+      "loss": 2.676,
+      "step": 10779
+    },
+    {
+      "epoch": 0.6631397637795275,
+      "grad_norm": 1.2815961837768555,
+      "learning_rate": 5.1019395089480474e-05,
+      "loss": 3.0533,
+      "step": 10780
+    },
+    {
+      "epoch": 0.663201279527559,
+      "grad_norm": 1.222311019897461,
+      "learning_rate": 5.100253684497076e-05,
+      "loss": 2.7003,
+      "step": 10781
+    },
+    {
+      "epoch": 0.6632627952755905,
+      "grad_norm": 1.0759825706481934,
+      "learning_rate": 5.0985680432692985e-05,
+      "loss": 2.7822,
+      "step": 10782
+    },
+    {
+      "epoch": 0.663324311023622,
+      "grad_norm": 1.5367465019226074,
+      "learning_rate": 5.096882585327741e-05,
+      "loss": 1.6695,
+      "step": 10783
+    },
+    {
+      "epoch": 0.6633858267716536,
+      "grad_norm": 1.088484525680542,
+      "learning_rate": 5.0951973107354354e-05,
+      "loss": 2.4486,
+      "step": 10784
+    },
+    {
+      "epoch": 0.6634473425196851,
+      "grad_norm": 1.3981140851974487,
+      "learning_rate": 5.0935122195553995e-05,
+      "loss": 2.1251,
+      "step": 10785
+    },
+    {
+      "epoch": 0.6635088582677166,
+      "grad_norm": 1.6364890336990356,
+      "learning_rate": 5.091827311850652e-05,
+      "loss": 3.3384,
+      "step": 10786
+    },
+    {
+      "epoch": 0.663570374015748,
+      "grad_norm": 1.3947302103042603,
+      "learning_rate": 5.090142587684189e-05,
+      "loss": 2.2067,
+      "step": 10787
+    },
+    {
+      "epoch": 0.6636318897637795,
+      "grad_norm": 1.2971687316894531,
+      "learning_rate": 5.08845804711902e-05,
+      "loss": 2.4875,
+      "step": 10788
+    },
+    {
+      "epoch": 0.663693405511811,
+      "grad_norm": 1.266240119934082,
+      "learning_rate": 5.086773690218128e-05,
+      "loss": 2.6008,
+      "step": 10789
+    },
+    {
+      "epoch": 0.6637549212598425,
+      "grad_norm": 1.2368900775909424,
+      "learning_rate": 5.0850895170445077e-05,
+      "loss": 2.6926,
+      "step": 10790
+    },
+    {
+      "epoch": 0.663816437007874,
+      "grad_norm": 1.2704402208328247,
+      "learning_rate": 5.0834055276611245e-05,
+      "loss": 2.7869,
+      "step": 10791
+    },
+    {
+      "epoch": 0.6638779527559056,
+      "grad_norm": 1.4094371795654297,
+      "learning_rate": 5.081721722130968e-05,
+      "loss": 2.2496,
+      "step": 10792
+    },
+    {
+      "epoch": 0.663939468503937,
+      "grad_norm": 1.2761175632476807,
+      "learning_rate": 5.080038100516989e-05,
+      "loss": 2.8463,
+      "step": 10793
+    },
+    {
+      "epoch": 0.6640009842519685,
+      "grad_norm": 1.238707184791565,
+      "learning_rate": 5.078354662882156e-05,
+      "loss": 2.9144,
+      "step": 10794
+    },
+    {
+      "epoch": 0.6640625,
+      "grad_norm": 1.1358689069747925,
+      "learning_rate": 5.07667140928941e-05,
+      "loss": 2.6235,
+      "step": 10795
+    },
+    {
+      "epoch": 0.6641240157480315,
+      "grad_norm": 1.5177242755889893,
+      "learning_rate": 5.074988339801703e-05,
+      "loss": 2.167,
+      "step": 10796
+    },
+    {
+      "epoch": 0.664185531496063,
+      "grad_norm": 1.197731852531433,
+      "learning_rate": 5.073305454481967e-05,
+      "loss": 2.3399,
+      "step": 10797
+    },
+    {
+      "epoch": 0.6642470472440944,
+      "grad_norm": 1.228052020072937,
+      "learning_rate": 5.071622753393134e-05,
+      "loss": 2.4394,
+      "step": 10798
+    },
+    {
+      "epoch": 0.664308562992126,
+      "grad_norm": 1.1605525016784668,
+      "learning_rate": 5.069940236598128e-05,
+      "loss": 2.6886,
+      "step": 10799
+    },
+    {
+      "epoch": 0.6643700787401575,
+      "grad_norm": 1.2467238903045654,
+      "learning_rate": 5.068257904159872e-05,
+      "loss": 2.1938,
+      "step": 10800
+    },
+    {
+      "epoch": 0.664431594488189,
+      "grad_norm": 1.3888403177261353,
+      "learning_rate": 5.066575756141265e-05,
+      "loss": 2.3801,
+      "step": 10801
+    },
+    {
+      "epoch": 0.6644931102362205,
+      "grad_norm": 1.0377135276794434,
+      "learning_rate": 5.064893792605221e-05,
+      "loss": 2.2763,
+      "step": 10802
+    },
+    {
+      "epoch": 0.664554625984252,
+      "grad_norm": 1.3867692947387695,
+      "learning_rate": 5.063212013614624e-05,
+      "loss": 2.3325,
+      "step": 10803
+    },
+    {
+      "epoch": 0.6646161417322834,
+      "grad_norm": 1.2828871011734009,
+      "learning_rate": 5.06153041923237e-05,
+      "loss": 2.5366,
+      "step": 10804
+    },
+    {
+      "epoch": 0.6646776574803149,
+      "grad_norm": 1.3477833271026611,
+      "learning_rate": 5.059849009521342e-05,
+      "loss": 1.9625,
+      "step": 10805
+    },
+    {
+      "epoch": 0.6647391732283464,
+      "grad_norm": 1.4552456140518188,
+      "learning_rate": 5.0581677845444165e-05,
+      "loss": 1.9051,
+      "step": 10806
+    },
+    {
+      "epoch": 0.664800688976378,
+      "grad_norm": 1.4559563398361206,
+      "learning_rate": 5.056486744364457e-05,
+      "loss": 1.9557,
+      "step": 10807
+    },
+    {
+      "epoch": 0.6648622047244095,
+      "grad_norm": 1.5508204698562622,
+      "learning_rate": 5.0548058890443294e-05,
+      "loss": 2.5319,
+      "step": 10808
+    },
+    {
+      "epoch": 0.664923720472441,
+      "grad_norm": 1.6508272886276245,
+      "learning_rate": 5.053125218646884e-05,
+      "loss": 1.9947,
+      "step": 10809
+    },
+    {
+      "epoch": 0.6649852362204725,
+      "grad_norm": 1.2731437683105469,
+      "learning_rate": 5.05144473323497e-05,
+      "loss": 2.2524,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6650467519685039,
+      "grad_norm": 1.2343039512634277,
+      "learning_rate": 5.0497644328714286e-05,
+      "loss": 2.7942,
+      "step": 10811
+    },
+    {
+      "epoch": 0.6651082677165354,
+      "grad_norm": 1.2786234617233276,
+      "learning_rate": 5.0480843176190995e-05,
+      "loss": 2.8436,
+      "step": 10812
+    },
+    {
+      "epoch": 0.6651697834645669,
+      "grad_norm": 1.2154825925827026,
+      "learning_rate": 5.0464043875407996e-05,
+      "loss": 2.3755,
+      "step": 10813
+    },
+    {
+      "epoch": 0.6652312992125984,
+      "grad_norm": 1.2900404930114746,
+      "learning_rate": 5.044724642699359e-05,
+      "loss": 2.4635,
+      "step": 10814
+    },
+    {
+      "epoch": 0.66529281496063,
+      "grad_norm": 1.2262550592422485,
+      "learning_rate": 5.043045083157579e-05,
+      "loss": 2.6287,
+      "step": 10815
+    },
+    {
+      "epoch": 0.6653543307086615,
+      "grad_norm": 1.2447093725204468,
+      "learning_rate": 5.0413657089782786e-05,
+      "loss": 2.6995,
+      "step": 10816
+    },
+    {
+      "epoch": 0.6654158464566929,
+      "grad_norm": 1.3233191967010498,
+      "learning_rate": 5.039686520224245e-05,
+      "loss": 2.1041,
+      "step": 10817
+    },
+    {
+      "epoch": 0.6654773622047244,
+      "grad_norm": 1.511148452758789,
+      "learning_rate": 5.0380075169582766e-05,
+      "loss": 2.1735,
+      "step": 10818
+    },
+    {
+      "epoch": 0.6655388779527559,
+      "grad_norm": 1.3799885511398315,
+      "learning_rate": 5.0363286992431615e-05,
+      "loss": 2.4075,
+      "step": 10819
+    },
+    {
+      "epoch": 0.6656003937007874,
+      "grad_norm": 1.1338056325912476,
+      "learning_rate": 5.034650067141671e-05,
+      "loss": 2.243,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6656619094488189,
+      "grad_norm": 1.748327612876892,
+      "learning_rate": 5.0329716207165845e-05,
+      "loss": 3.3164,
+      "step": 10821
+    },
+    {
+      "epoch": 0.6657234251968503,
+      "grad_norm": 1.2357250452041626,
+      "learning_rate": 5.031293360030659e-05,
+      "loss": 2.547,
+      "step": 10822
+    },
+    {
+      "epoch": 0.6657849409448819,
+      "grad_norm": 1.254472255706787,
+      "learning_rate": 5.029615285146656e-05,
+      "loss": 1.9187,
+      "step": 10823
+    },
+    {
+      "epoch": 0.6658464566929134,
+      "grad_norm": 1.1062663793563843,
+      "learning_rate": 5.027937396127325e-05,
+      "loss": 2.5615,
+      "step": 10824
+    },
+    {
+      "epoch": 0.6659079724409449,
+      "grad_norm": 1.218879222869873,
+      "learning_rate": 5.026259693035413e-05,
+      "loss": 2.6294,
+      "step": 10825
+    },
+    {
+      "epoch": 0.6659694881889764,
+      "grad_norm": 1.1417057514190674,
+      "learning_rate": 5.024582175933652e-05,
+      "loss": 2.1029,
+      "step": 10826
+    },
+    {
+      "epoch": 0.6660310039370079,
+      "grad_norm": 1.146204948425293,
+      "learning_rate": 5.0229048448847774e-05,
+      "loss": 1.7807,
+      "step": 10827
+    },
+    {
+      "epoch": 0.6660925196850394,
+      "grad_norm": 1.1512508392333984,
+      "learning_rate": 5.0212276999515054e-05,
+      "loss": 2.0285,
+      "step": 10828
+    },
+    {
+      "epoch": 0.6661540354330708,
+      "grad_norm": 1.2468593120574951,
+      "learning_rate": 5.019550741196555e-05,
+      "loss": 2.5802,
+      "step": 10829
+    },
+    {
+      "epoch": 0.6662155511811023,
+      "grad_norm": 1.45399808883667,
+      "learning_rate": 5.0178739686826336e-05,
+      "loss": 2.5027,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6662770669291339,
+      "grad_norm": 1.1503413915634155,
+      "learning_rate": 5.0161973824724496e-05,
+      "loss": 2.1703,
+      "step": 10831
+    },
+    {
+      "epoch": 0.6663385826771654,
+      "grad_norm": 1.1485544443130493,
+      "learning_rate": 5.014520982628689e-05,
+      "loss": 2.3154,
+      "step": 10832
+    },
+    {
+      "epoch": 0.6664000984251969,
+      "grad_norm": 1.5176888704299927,
+      "learning_rate": 5.0128447692140465e-05,
+      "loss": 2.6813,
+      "step": 10833
+    },
+    {
+      "epoch": 0.6664616141732284,
+      "grad_norm": 1.492994785308838,
+      "learning_rate": 5.011168742291199e-05,
+      "loss": 2.3623,
+      "step": 10834
+    },
+    {
+      "epoch": 0.6665231299212598,
+      "grad_norm": 1.1636732816696167,
+      "learning_rate": 5.009492901922825e-05,
+      "loss": 1.488,
+      "step": 10835
+    },
+    {
+      "epoch": 0.6665846456692913,
+      "grad_norm": 1.302748441696167,
+      "learning_rate": 5.0078172481715804e-05,
+      "loss": 1.9147,
+      "step": 10836
+    },
+    {
+      "epoch": 0.6666461614173228,
+      "grad_norm": 1.4329144954681396,
+      "learning_rate": 5.006141781100142e-05,
+      "loss": 2.3866,
+      "step": 10837
+    },
+    {
+      "epoch": 0.6667076771653543,
+      "grad_norm": 1.2661057710647583,
+      "learning_rate": 5.00446650077115e-05,
+      "loss": 2.1908,
+      "step": 10838
+    },
+    {
+      "epoch": 0.6667691929133859,
+      "grad_norm": 1.3834327459335327,
+      "learning_rate": 5.002791407247258e-05,
+      "loss": 2.4648,
+      "step": 10839
+    },
+    {
+      "epoch": 0.6668307086614174,
+      "grad_norm": 1.246362328529358,
+      "learning_rate": 5.0011165005911e-05,
+      "loss": 2.7607,
+      "step": 10840
+    },
+    {
+      "epoch": 0.6668922244094488,
+      "grad_norm": 1.133460283279419,
+      "learning_rate": 4.9994417808653136e-05,
+      "loss": 1.4993,
+      "step": 10841
+    },
+    {
+      "epoch": 0.6669537401574803,
+      "grad_norm": 1.3861373662948608,
+      "learning_rate": 4.997767248132517e-05,
+      "loss": 6.6135,
+      "step": 10842
+    },
+    {
+      "epoch": 0.6670152559055118,
+      "grad_norm": 1.329242467880249,
+      "learning_rate": 4.9960929024553314e-05,
+      "loss": 2.4699,
+      "step": 10843
+    },
+    {
+      "epoch": 0.6670767716535433,
+      "grad_norm": 1.3653202056884766,
+      "learning_rate": 4.99441874389637e-05,
+      "loss": 2.6119,
+      "step": 10844
+    },
+    {
+      "epoch": 0.6671382874015748,
+      "grad_norm": 1.0571746826171875,
+      "learning_rate": 4.992744772518239e-05,
+      "loss": 3.0523,
+      "step": 10845
+    },
+    {
+      "epoch": 0.6671998031496063,
+      "grad_norm": 1.1760951280593872,
+      "learning_rate": 4.991070988383527e-05,
+      "loss": 3.0363,
+      "step": 10846
+    },
+    {
+      "epoch": 0.6672613188976378,
+      "grad_norm": 1.323391079902649,
+      "learning_rate": 4.989397391554835e-05,
+      "loss": 2.7837,
+      "step": 10847
+    },
+    {
+      "epoch": 0.6673228346456693,
+      "grad_norm": 1.2477766275405884,
+      "learning_rate": 4.9877239820947366e-05,
+      "loss": 2.7579,
+      "step": 10848
+    },
+    {
+      "epoch": 0.6673843503937008,
+      "grad_norm": 1.2509872913360596,
+      "learning_rate": 4.9860507600658114e-05,
+      "loss": 2.9731,
+      "step": 10849
+    },
+    {
+      "epoch": 0.6674458661417323,
+      "grad_norm": 1.2697372436523438,
+      "learning_rate": 4.9843777255306304e-05,
+      "loss": 2.0578,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6675073818897638,
+      "grad_norm": 1.2553149461746216,
+      "learning_rate": 4.982704878551757e-05,
+      "loss": 2.5003,
+      "step": 10851
+    },
+    {
+      "epoch": 0.6675688976377953,
+      "grad_norm": 1.5868961811065674,
+      "learning_rate": 4.9810322191917405e-05,
+      "loss": 2.5584,
+      "step": 10852
+    },
+    {
+      "epoch": 0.6676304133858267,
+      "grad_norm": 1.2272861003875732,
+      "learning_rate": 4.979359747513137e-05,
+      "loss": 1.4797,
+      "step": 10853
+    },
+    {
+      "epoch": 0.6676919291338582,
+      "grad_norm": 1.1665765047073364,
+      "learning_rate": 4.977687463578479e-05,
+      "loss": 3.298,
+      "step": 10854
+    },
+    {
+      "epoch": 0.6677534448818898,
+      "grad_norm": 1.2773758172988892,
+      "learning_rate": 4.976015367450304e-05,
+      "loss": 2.9622,
+      "step": 10855
+    },
+    {
+      "epoch": 0.6678149606299213,
+      "grad_norm": 1.2494651079177856,
+      "learning_rate": 4.974343459191144e-05,
+      "loss": 1.6921,
+      "step": 10856
+    },
+    {
+      "epoch": 0.6678764763779528,
+      "grad_norm": 1.0940953493118286,
+      "learning_rate": 4.9726717388635114e-05,
+      "loss": 1.7943,
+      "step": 10857
+    },
+    {
+      "epoch": 0.6679379921259843,
+      "grad_norm": 1.259549617767334,
+      "learning_rate": 4.9710002065299256e-05,
+      "loss": 2.7934,
+      "step": 10858
+    },
+    {
+      "epoch": 0.6679995078740157,
+      "grad_norm": 1.3434315919876099,
+      "learning_rate": 4.9693288622528865e-05,
+      "loss": 2.6262,
+      "step": 10859
+    },
+    {
+      "epoch": 0.6680610236220472,
+      "grad_norm": 1.0822266340255737,
+      "learning_rate": 4.967657706094899e-05,
+      "loss": 2.8189,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6681225393700787,
+      "grad_norm": 1.2629457712173462,
+      "learning_rate": 4.9659867381184496e-05,
+      "loss": 1.9171,
+      "step": 10861
+    },
+    {
+      "epoch": 0.6681840551181102,
+      "grad_norm": 1.2892048358917236,
+      "learning_rate": 4.964315958386025e-05,
+      "loss": 3.0865,
+      "step": 10862
+    },
+    {
+      "epoch": 0.6682455708661418,
+      "grad_norm": 1.3551520109176636,
+      "learning_rate": 4.962645366960105e-05,
+      "loss": 2.4233,
+      "step": 10863
+    },
+    {
+      "epoch": 0.6683070866141733,
+      "grad_norm": 1.3005518913269043,
+      "learning_rate": 4.960974963903161e-05,
+      "loss": 2.4608,
+      "step": 10864
+    },
+    {
+      "epoch": 0.6683686023622047,
+      "grad_norm": 1.4073541164398193,
+      "learning_rate": 4.959304749277652e-05,
+      "loss": 2.6988,
+      "step": 10865
+    },
+    {
+      "epoch": 0.6684301181102362,
+      "grad_norm": 1.1141064167022705,
+      "learning_rate": 4.957634723146043e-05,
+      "loss": 2.0008,
+      "step": 10866
+    },
+    {
+      "epoch": 0.6684916338582677,
+      "grad_norm": 1.3466217517852783,
+      "learning_rate": 4.955964885570773e-05,
+      "loss": 2.4532,
+      "step": 10867
+    },
+    {
+      "epoch": 0.6685531496062992,
+      "grad_norm": 1.1573903560638428,
+      "learning_rate": 4.9542952366142906e-05,
+      "loss": 2.576,
+      "step": 10868
+    },
+    {
+      "epoch": 0.6686146653543307,
+      "grad_norm": 1.476111650466919,
+      "learning_rate": 4.95262577633903e-05,
+      "loss": 2.4944,
+      "step": 10869
+    },
+    {
+      "epoch": 0.6686761811023622,
+      "grad_norm": 1.2561159133911133,
+      "learning_rate": 4.9509565048074245e-05,
+      "loss": 2.29,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6687376968503937,
+      "grad_norm": 1.2556567192077637,
+      "learning_rate": 4.949287422081888e-05,
+      "loss": 2.3874,
+      "step": 10871
+    },
+    {
+      "epoch": 0.6687992125984252,
+      "grad_norm": 1.3146686553955078,
+      "learning_rate": 4.9476185282248425e-05,
+      "loss": 2.6918,
+      "step": 10872
+    },
+    {
+      "epoch": 0.6688607283464567,
+      "grad_norm": 1.1355254650115967,
+      "learning_rate": 4.9459498232986865e-05,
+      "loss": 2.617,
+      "step": 10873
+    },
+    {
+      "epoch": 0.6689222440944882,
+      "grad_norm": 1.3619017601013184,
+      "learning_rate": 4.944281307365826e-05,
+      "loss": 2.2163,
+      "step": 10874
+    },
+    {
+      "epoch": 0.6689837598425197,
+      "grad_norm": 1.3649113178253174,
+      "learning_rate": 4.942612980488651e-05,
+      "loss": 3.0197,
+      "step": 10875
+    },
+    {
+      "epoch": 0.6690452755905512,
+      "grad_norm": 1.130345344543457,
+      "learning_rate": 4.9409448427295555e-05,
+      "loss": 2.5077,
+      "step": 10876
+    },
+    {
+      "epoch": 0.6691067913385826,
+      "grad_norm": 1.3990577459335327,
+      "learning_rate": 4.939276894150908e-05,
+      "loss": 2.2543,
+      "step": 10877
+    },
+    {
+      "epoch": 0.6691683070866141,
+      "grad_norm": 1.1082613468170166,
+      "learning_rate": 4.93760913481509e-05,
+      "loss": 2.9544,
+      "step": 10878
+    },
+    {
+      "epoch": 0.6692298228346457,
+      "grad_norm": 1.3741199970245361,
+      "learning_rate": 4.935941564784457e-05,
+      "loss": 2.7199,
+      "step": 10879
+    },
+    {
+      "epoch": 0.6692913385826772,
+      "grad_norm": 1.019669532775879,
+      "learning_rate": 4.934274184121373e-05,
+      "loss": 2.3258,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6693528543307087,
+      "grad_norm": 1.1261855363845825,
+      "learning_rate": 4.932606992888187e-05,
+      "loss": 2.6146,
+      "step": 10881
+    },
+    {
+      "epoch": 0.6694143700787402,
+      "grad_norm": 1.2338635921478271,
+      "learning_rate": 4.930939991147245e-05,
+      "loss": 2.461,
+      "step": 10882
+    },
+    {
+      "epoch": 0.6694758858267716,
+      "grad_norm": 1.2385798692703247,
+      "learning_rate": 4.929273178960879e-05,
+      "loss": 2.4334,
+      "step": 10883
+    },
+    {
+      "epoch": 0.6695374015748031,
+      "grad_norm": 1.2630261182785034,
+      "learning_rate": 4.927606556391424e-05,
+      "loss": 2.3011,
+      "step": 10884
+    },
+    {
+      "epoch": 0.6695989173228346,
+      "grad_norm": 1.2654186487197876,
+      "learning_rate": 4.925940123501197e-05,
+      "loss": 2.6279,
+      "step": 10885
+    },
+    {
+      "epoch": 0.6696604330708661,
+      "grad_norm": 1.0032626390457153,
+      "learning_rate": 4.924273880352518e-05,
+      "loss": 2.7011,
+      "step": 10886
+    },
+    {
+      "epoch": 0.6697219488188977,
+      "grad_norm": 1.245406150817871,
+      "learning_rate": 4.9226078270076895e-05,
+      "loss": 2.5151,
+      "step": 10887
+    },
+    {
+      "epoch": 0.6697834645669292,
+      "grad_norm": 1.3606789112091064,
+      "learning_rate": 4.9209419635290156e-05,
+      "loss": 2.7812,
+      "step": 10888
+    },
+    {
+      "epoch": 0.6698449803149606,
+      "grad_norm": 1.223998785018921,
+      "learning_rate": 4.919276289978791e-05,
+      "loss": 2.33,
+      "step": 10889
+    },
+    {
+      "epoch": 0.6699064960629921,
+      "grad_norm": 1.166269063949585,
+      "learning_rate": 4.917610806419306e-05,
+      "loss": 1.4744,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6699680118110236,
+      "grad_norm": 1.3318418264389038,
+      "learning_rate": 4.9159455129128377e-05,
+      "loss": 2.2424,
+      "step": 10891
+    },
+    {
+      "epoch": 0.6700295275590551,
+      "grad_norm": 1.2199673652648926,
+      "learning_rate": 4.914280409521652e-05,
+      "loss": 2.3419,
+      "step": 10892
+    },
+    {
+      "epoch": 0.6700910433070866,
+      "grad_norm": 1.2611676454544067,
+      "learning_rate": 4.912615496308022e-05,
+      "loss": 2.5163,
+      "step": 10893
+    },
+    {
+      "epoch": 0.6701525590551181,
+      "grad_norm": 1.2865095138549805,
+      "learning_rate": 4.910950773334203e-05,
+      "loss": 2.5941,
+      "step": 10894
+    },
+    {
+      "epoch": 0.6702140748031497,
+      "grad_norm": 1.5025373697280884,
+      "learning_rate": 4.909286240662453e-05,
+      "loss": 2.0627,
+      "step": 10895
+    },
+    {
+      "epoch": 0.6702755905511811,
+      "grad_norm": 1.6943250894546509,
+      "learning_rate": 4.907621898355006e-05,
+      "loss": 2.2019,
+      "step": 10896
+    },
+    {
+      "epoch": 0.6703371062992126,
+      "grad_norm": 1.429404616355896,
+      "learning_rate": 4.905957746474108e-05,
+      "loss": 2.3607,
+      "step": 10897
+    },
+    {
+      "epoch": 0.6703986220472441,
+      "grad_norm": 1.6115286350250244,
+      "learning_rate": 4.9042937850819814e-05,
+      "loss": 2.499,
+      "step": 10898
+    },
+    {
+      "epoch": 0.6704601377952756,
+      "grad_norm": 1.4400733709335327,
+      "learning_rate": 4.9026300142408534e-05,
+      "loss": 2.7394,
+      "step": 10899
+    },
+    {
+      "epoch": 0.6705216535433071,
+      "grad_norm": 1.1263729333877563,
+      "learning_rate": 4.900966434012939e-05,
+      "loss": 2.8963,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6705831692913385,
+      "grad_norm": 1.4247043132781982,
+      "learning_rate": 4.89930304446045e-05,
+      "loss": 2.619,
+      "step": 10901
+    },
+    {
+      "epoch": 0.67064468503937,
+      "grad_norm": 1.0747654438018799,
+      "learning_rate": 4.897639845645582e-05,
+      "loss": 2.5351,
+      "step": 10902
+    },
+    {
+      "epoch": 0.6707062007874016,
+      "grad_norm": 1.381488561630249,
+      "learning_rate": 4.895976837630536e-05,
+      "loss": 1.9419,
+      "step": 10903
+    },
+    {
+      "epoch": 0.6707677165354331,
+      "grad_norm": 1.6519640684127808,
+      "learning_rate": 4.8943140204774916e-05,
+      "loss": 2.7324,
+      "step": 10904
+    },
+    {
+      "epoch": 0.6708292322834646,
+      "grad_norm": 1.3441612720489502,
+      "learning_rate": 4.892651394248636e-05,
+      "loss": 2.66,
+      "step": 10905
+    },
+    {
+      "epoch": 0.6708907480314961,
+      "grad_norm": 1.258495807647705,
+      "learning_rate": 4.890988959006132e-05,
+      "loss": 2.2388,
+      "step": 10906
+    },
+    {
+      "epoch": 0.6709522637795275,
+      "grad_norm": 1.0355350971221924,
+      "learning_rate": 4.889326714812159e-05,
+      "loss": 2.5965,
+      "step": 10907
+    },
+    {
+      "epoch": 0.671013779527559,
+      "grad_norm": 1.2877267599105835,
+      "learning_rate": 4.8876646617288655e-05,
+      "loss": 2.0723,
+      "step": 10908
+    },
+    {
+      "epoch": 0.6710752952755905,
+      "grad_norm": 1.2249616384506226,
+      "learning_rate": 4.886002799818411e-05,
+      "loss": 2.9869,
+      "step": 10909
+    },
+    {
+      "epoch": 0.671136811023622,
+      "grad_norm": 1.2455617189407349,
+      "learning_rate": 4.8843411291429306e-05,
+      "loss": 2.3255,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6711983267716536,
+      "grad_norm": 1.3002265691757202,
+      "learning_rate": 4.882679649764571e-05,
+      "loss": 2.9649,
+      "step": 10911
+    },
+    {
+      "epoch": 0.6712598425196851,
+      "grad_norm": 1.8223140239715576,
+      "learning_rate": 4.881018361745453e-05,
+      "loss": 2.7978,
+      "step": 10912
+    },
+    {
+      "epoch": 0.6713213582677166,
+      "grad_norm": 1.636309027671814,
+      "learning_rate": 4.879357265147704e-05,
+      "loss": 1.8612,
+      "step": 10913
+    },
+    {
+      "epoch": 0.671382874015748,
+      "grad_norm": 1.3642929792404175,
+      "learning_rate": 4.87769636003344e-05,
+      "loss": 2.0863,
+      "step": 10914
+    },
+    {
+      "epoch": 0.6714443897637795,
+      "grad_norm": 2.0885679721832275,
+      "learning_rate": 4.876035646464773e-05,
+      "loss": 2.1119,
+      "step": 10915
+    },
+    {
+      "epoch": 0.671505905511811,
+      "grad_norm": 1.5309391021728516,
+      "learning_rate": 4.874375124503797e-05,
+      "loss": 2.427,
+      "step": 10916
+    },
+    {
+      "epoch": 0.6715674212598425,
+      "grad_norm": 1.4427118301391602,
+      "learning_rate": 4.872714794212615e-05,
+      "loss": 1.4551,
+      "step": 10917
+    },
+    {
+      "epoch": 0.671628937007874,
+      "grad_norm": 1.093066692352295,
+      "learning_rate": 4.8710546556533043e-05,
+      "loss": 3.0455,
+      "step": 10918
+    },
+    {
+      "epoch": 0.6716904527559056,
+      "grad_norm": 1.145269751548767,
+      "learning_rate": 4.86939470888795e-05,
+      "loss": 2.1398,
+      "step": 10919
+    },
+    {
+      "epoch": 0.671751968503937,
+      "grad_norm": 1.0239776372909546,
+      "learning_rate": 4.867734953978624e-05,
+      "loss": 3.0266,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6718134842519685,
+      "grad_norm": 1.139328122138977,
+      "learning_rate": 4.866075390987397e-05,
+      "loss": 2.3005,
+      "step": 10921
+    },
+    {
+      "epoch": 0.671875,
+      "grad_norm": 1.2338848114013672,
+      "learning_rate": 4.864416019976319e-05,
+      "loss": 2.8176,
+      "step": 10922
+    },
+    {
+      "epoch": 0.6719365157480315,
+      "grad_norm": 2.0420241355895996,
+      "learning_rate": 4.862756841007449e-05,
+      "loss": 2.9019,
+      "step": 10923
+    },
+    {
+      "epoch": 0.671998031496063,
+      "grad_norm": 1.0556044578552246,
+      "learning_rate": 4.861097854142822e-05,
+      "loss": 1.6668,
+      "step": 10924
+    },
+    {
+      "epoch": 0.6720595472440944,
+      "grad_norm": 1.2915326356887817,
+      "learning_rate": 4.8594390594444805e-05,
+      "loss": 2.3829,
+      "step": 10925
+    },
+    {
+      "epoch": 0.672121062992126,
+      "grad_norm": 1.0253710746765137,
+      "learning_rate": 4.857780456974455e-05,
+      "loss": 1.8619,
+      "step": 10926
+    },
+    {
+      "epoch": 0.6721825787401575,
+      "grad_norm": 1.2738628387451172,
+      "learning_rate": 4.856122046794769e-05,
+      "loss": 2.4298,
+      "step": 10927
+    },
+    {
+      "epoch": 0.672244094488189,
+      "grad_norm": 1.2042149305343628,
+      "learning_rate": 4.854463828967435e-05,
+      "loss": 2.8463,
+      "step": 10928
+    },
+    {
+      "epoch": 0.6723056102362205,
+      "grad_norm": 1.2877291440963745,
+      "learning_rate": 4.852805803554459e-05,
+      "loss": 2.3488,
+      "step": 10929
+    },
+    {
+      "epoch": 0.672367125984252,
+      "grad_norm": 1.3814904689788818,
+      "learning_rate": 4.851147970617848e-05,
+      "loss": 2.2847,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6724286417322834,
+      "grad_norm": 1.3166254758834839,
+      "learning_rate": 4.8494903302195874e-05,
+      "loss": 2.3722,
+      "step": 10931
+    },
+    {
+      "epoch": 0.6724901574803149,
+      "grad_norm": 1.4740488529205322,
+      "learning_rate": 4.847832882421669e-05,
+      "loss": 1.9944,
+      "step": 10932
+    },
+    {
+      "epoch": 0.6725516732283464,
+      "grad_norm": 1.3193703889846802,
+      "learning_rate": 4.846175627286072e-05,
+      "loss": 3.1071,
+      "step": 10933
+    },
+    {
+      "epoch": 0.672613188976378,
+      "grad_norm": 1.1793124675750732,
+      "learning_rate": 4.8445185648747705e-05,
+      "loss": 2.482,
+      "step": 10934
+    },
+    {
+      "epoch": 0.6726747047244095,
+      "grad_norm": 1.2118158340454102,
+      "learning_rate": 4.842861695249724e-05,
+      "loss": 2.4321,
+      "step": 10935
+    },
+    {
+      "epoch": 0.672736220472441,
+      "grad_norm": 1.184425950050354,
+      "learning_rate": 4.8412050184728975e-05,
+      "loss": 2.5292,
+      "step": 10936
+    },
+    {
+      "epoch": 0.6727977362204725,
+      "grad_norm": 1.0693926811218262,
+      "learning_rate": 4.839548534606232e-05,
+      "loss": 2.6619,
+      "step": 10937
+    },
+    {
+      "epoch": 0.6728592519685039,
+      "grad_norm": 1.2010961771011353,
+      "learning_rate": 4.837892243711677e-05,
+      "loss": 2.1544,
+      "step": 10938
+    },
+    {
+      "epoch": 0.6729207677165354,
+      "grad_norm": 1.2651519775390625,
+      "learning_rate": 4.836236145851166e-05,
+      "loss": 2.4173,
+      "step": 10939
+    },
+    {
+      "epoch": 0.6729822834645669,
+      "grad_norm": 1.2388523817062378,
+      "learning_rate": 4.834580241086635e-05,
+      "loss": 2.587,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6730437992125984,
+      "grad_norm": 1.1940997838974,
+      "learning_rate": 4.832924529479994e-05,
+      "loss": 2.2148,
+      "step": 10941
+    },
+    {
+      "epoch": 0.67310531496063,
+      "grad_norm": 1.264302372932434,
+      "learning_rate": 4.831269011093168e-05,
+      "loss": 2.3126,
+      "step": 10942
+    },
+    {
+      "epoch": 0.6731668307086615,
+      "grad_norm": 1.1838268041610718,
+      "learning_rate": 4.829613685988056e-05,
+      "loss": 2.581,
+      "step": 10943
+    },
+    {
+      "epoch": 0.6732283464566929,
+      "grad_norm": 1.353926658630371,
+      "learning_rate": 4.8279585542265604e-05,
+      "loss": 2.3379,
+      "step": 10944
+    },
+    {
+      "epoch": 0.6732898622047244,
+      "grad_norm": 1.2540717124938965,
+      "learning_rate": 4.8263036158705745e-05,
+      "loss": 2.591,
+      "step": 10945
+    },
+    {
+      "epoch": 0.6733513779527559,
+      "grad_norm": 1.4070345163345337,
+      "learning_rate": 4.8246488709819885e-05,
+      "loss": 2.3318,
+      "step": 10946
+    },
+    {
+      "epoch": 0.6734128937007874,
+      "grad_norm": 1.2134063243865967,
+      "learning_rate": 4.8229943196226724e-05,
+      "loss": 2.9763,
+      "step": 10947
+    },
+    {
+      "epoch": 0.6734744094488189,
+      "grad_norm": 1.2097704410552979,
+      "learning_rate": 4.821339961854505e-05,
+      "loss": 2.1304,
+      "step": 10948
+    },
+    {
+      "epoch": 0.6735359251968503,
+      "grad_norm": 1.3813742399215698,
+      "learning_rate": 4.819685797739342e-05,
+      "loss": 2.5171,
+      "step": 10949
+    },
+    {
+      "epoch": 0.6735974409448819,
+      "grad_norm": 1.6204333305358887,
+      "learning_rate": 4.8180318273390434e-05,
+      "loss": 2.322,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6736589566929134,
+      "grad_norm": 1.4478901624679565,
+      "learning_rate": 4.81637805071546e-05,
+      "loss": 2.1804,
+      "step": 10951
+    },
+    {
+      "epoch": 0.6737204724409449,
+      "grad_norm": 1.1422576904296875,
+      "learning_rate": 4.814724467930436e-05,
+      "loss": 2.1053,
+      "step": 10952
+    },
+    {
+      "epoch": 0.6737819881889764,
+      "grad_norm": 1.3441320657730103,
+      "learning_rate": 4.813071079045799e-05,
+      "loss": 2.4759,
+      "step": 10953
+    },
+    {
+      "epoch": 0.6738435039370079,
+      "grad_norm": 1.2971433401107788,
+      "learning_rate": 4.811417884123385e-05,
+      "loss": 2.033,
+      "step": 10954
+    },
+    {
+      "epoch": 0.6739050196850394,
+      "grad_norm": 1.1884788274765015,
+      "learning_rate": 4.809764883225005e-05,
+      "loss": 2.8421,
+      "step": 10955
+    },
+    {
+      "epoch": 0.6739665354330708,
+      "grad_norm": 1.2180469036102295,
+      "learning_rate": 4.808112076412481e-05,
+      "loss": 2.8087,
+      "step": 10956
+    },
+    {
+      "epoch": 0.6740280511811023,
+      "grad_norm": 1.2243521213531494,
+      "learning_rate": 4.806459463747611e-05,
+      "loss": 3.2128,
+      "step": 10957
+    },
+    {
+      "epoch": 0.6740895669291339,
+      "grad_norm": 1.389818549156189,
+      "learning_rate": 4.804807045292198e-05,
+      "loss": 2.5867,
+      "step": 10958
+    },
+    {
+      "epoch": 0.6741510826771654,
+      "grad_norm": 1.1736928224563599,
+      "learning_rate": 4.803154821108031e-05,
+      "loss": 2.1753,
+      "step": 10959
+    },
+    {
+      "epoch": 0.6742125984251969,
+      "grad_norm": 1.316419005393982,
+      "learning_rate": 4.8015027912569e-05,
+      "loss": 2.3913,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6742741141732284,
+      "grad_norm": 1.435109257698059,
+      "learning_rate": 4.7998509558005725e-05,
+      "loss": 2.8781,
+      "step": 10961
+    },
+    {
+      "epoch": 0.6743356299212598,
+      "grad_norm": 1.1918389797210693,
+      "learning_rate": 4.7981993148008264e-05,
+      "loss": 2.6555,
+      "step": 10962
+    },
+    {
+      "epoch": 0.6743971456692913,
+      "grad_norm": 1.3831456899642944,
+      "learning_rate": 4.7965478683194156e-05,
+      "loss": 2.1089,
+      "step": 10963
+    },
+    {
+      "epoch": 0.6744586614173228,
+      "grad_norm": 1.5962190628051758,
+      "learning_rate": 4.7948966164181e-05,
+      "loss": 2.3705,
+      "step": 10964
+    },
+    {
+      "epoch": 0.6745201771653543,
+      "grad_norm": 1.1753572225570679,
+      "learning_rate": 4.79324555915863e-05,
+      "loss": 2.8514,
+      "step": 10965
+    },
+    {
+      "epoch": 0.6745816929133859,
+      "grad_norm": 1.3819910287857056,
+      "learning_rate": 4.7915946966027394e-05,
+      "loss": 2.8995,
+      "step": 10966
+    },
+    {
+      "epoch": 0.6746432086614174,
+      "grad_norm": 1.3701659440994263,
+      "learning_rate": 4.7899440288121666e-05,
+      "loss": 2.3624,
+      "step": 10967
+    },
+    {
+      "epoch": 0.6747047244094488,
+      "grad_norm": 1.300148367881775,
+      "learning_rate": 4.788293555848632e-05,
+      "loss": 2.3264,
+      "step": 10968
+    },
+    {
+      "epoch": 0.6747662401574803,
+      "grad_norm": 1.241076946258545,
+      "learning_rate": 4.786643277773858e-05,
+      "loss": 1.5794,
+      "step": 10969
+    },
+    {
+      "epoch": 0.6748277559055118,
+      "grad_norm": 1.1558411121368408,
+      "learning_rate": 4.784993194649554e-05,
+      "loss": 2.2963,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6748892716535433,
+      "grad_norm": 1.2742496728897095,
+      "learning_rate": 4.7833433065374286e-05,
+      "loss": 2.4914,
+      "step": 10971
+    },
+    {
+      "epoch": 0.6749507874015748,
+      "grad_norm": 1.4196809530258179,
+      "learning_rate": 4.7816936134991707e-05,
+      "loss": 2.5894,
+      "step": 10972
+    },
+    {
+      "epoch": 0.6750123031496063,
+      "grad_norm": 1.178125023841858,
+      "learning_rate": 4.780044115596478e-05,
+      "loss": 1.7453,
+      "step": 10973
+    },
+    {
+      "epoch": 0.6750738188976378,
+      "grad_norm": 1.2733104228973389,
+      "learning_rate": 4.7783948128910235e-05,
+      "loss": 2.658,
+      "step": 10974
+    },
+    {
+      "epoch": 0.6751353346456693,
+      "grad_norm": 1.56722092628479,
+      "learning_rate": 4.776745705444491e-05,
+      "loss": 2.3331,
+      "step": 10975
+    },
+    {
+      "epoch": 0.6751968503937008,
+      "grad_norm": 1.6588324308395386,
+      "learning_rate": 4.775096793318537e-05,
+      "loss": 2.3139,
+      "step": 10976
+    },
+    {
+      "epoch": 0.6752583661417323,
+      "grad_norm": 1.183461308479309,
+      "learning_rate": 4.773448076574833e-05,
+      "loss": 2.7569,
+      "step": 10977
+    },
+    {
+      "epoch": 0.6753198818897638,
+      "grad_norm": 1.4360005855560303,
+      "learning_rate": 4.771799555275025e-05,
+      "loss": 3.1241,
+      "step": 10978
+    },
+    {
+      "epoch": 0.6753813976377953,
+      "grad_norm": 1.4643524885177612,
+      "learning_rate": 4.770151229480764e-05,
+      "loss": 2.3599,
+      "step": 10979
+    },
+    {
+      "epoch": 0.6754429133858267,
+      "grad_norm": 1.3818893432617188,
+      "learning_rate": 4.768503099253679e-05,
+      "loss": 2.7514,
+      "step": 10980
+    },
+    {
+      "epoch": 0.6755044291338582,
+      "grad_norm": 1.2348276376724243,
+      "learning_rate": 4.7668551646554125e-05,
+      "loss": 1.6035,
+      "step": 10981
+    },
+    {
+      "epoch": 0.6755659448818898,
+      "grad_norm": 1.4189397096633911,
+      "learning_rate": 4.7652074257475776e-05,
+      "loss": 2.3651,
+      "step": 10982
+    },
+    {
+      "epoch": 0.6756274606299213,
+      "grad_norm": 1.262601613998413,
+      "learning_rate": 4.763559882591795e-05,
+      "loss": 2.5894,
+      "step": 10983
+    },
+    {
+      "epoch": 0.6756889763779528,
+      "grad_norm": 1.134673833847046,
+      "learning_rate": 4.7619125352496726e-05,
+      "loss": 2.6222,
+      "step": 10984
+    },
+    {
+      "epoch": 0.6757504921259843,
+      "grad_norm": 1.1099512577056885,
+      "learning_rate": 4.760265383782818e-05,
+      "loss": 2.7409,
+      "step": 10985
+    },
+    {
+      "epoch": 0.6758120078740157,
+      "grad_norm": 1.1378703117370605,
+      "learning_rate": 4.758618428252817e-05,
+      "loss": 2.7207,
+      "step": 10986
+    },
+    {
+      "epoch": 0.6758735236220472,
+      "grad_norm": 1.1778202056884766,
+      "learning_rate": 4.7569716687212626e-05,
+      "loss": 2.6012,
+      "step": 10987
+    },
+    {
+      "epoch": 0.6759350393700787,
+      "grad_norm": 1.5281822681427002,
+      "learning_rate": 4.755325105249728e-05,
+      "loss": 2.3076,
+      "step": 10988
+    },
+    {
+      "epoch": 0.6759965551181102,
+      "grad_norm": 1.121415376663208,
+      "learning_rate": 4.7536787378997906e-05,
+      "loss": 3.0467,
+      "step": 10989
+    },
+    {
+      "epoch": 0.6760580708661418,
+      "grad_norm": 1.3928142786026,
+      "learning_rate": 4.752032566733013e-05,
+      "loss": 2.8396,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6761195866141733,
+      "grad_norm": 1.1481980085372925,
+      "learning_rate": 4.750386591810958e-05,
+      "loss": 3.0393,
+      "step": 10991
+    },
+    {
+      "epoch": 0.6761811023622047,
+      "grad_norm": 1.3785027265548706,
+      "learning_rate": 4.7487408131951684e-05,
+      "loss": 2.2319,
+      "step": 10992
+    },
+    {
+      "epoch": 0.6762426181102362,
+      "grad_norm": 1.2361165285110474,
+      "learning_rate": 4.747095230947194e-05,
+      "loss": 2.7021,
+      "step": 10993
+    },
+    {
+      "epoch": 0.6763041338582677,
+      "grad_norm": 1.2883888483047485,
+      "learning_rate": 4.7454498451285635e-05,
+      "loss": 1.9765,
+      "step": 10994
+    },
+    {
+      "epoch": 0.6763656496062992,
+      "grad_norm": 1.3085488080978394,
+      "learning_rate": 4.743804655800809e-05,
+      "loss": 2.0191,
+      "step": 10995
+    },
+    {
+      "epoch": 0.6764271653543307,
+      "grad_norm": 1.4938784837722778,
+      "learning_rate": 4.7421596630254505e-05,
+      "loss": 2.4889,
+      "step": 10996
+    },
+    {
+      "epoch": 0.6764886811023622,
+      "grad_norm": 1.1948869228363037,
+      "learning_rate": 4.740514866864005e-05,
+      "loss": 2.6266,
+      "step": 10997
+    },
+    {
+      "epoch": 0.6765501968503937,
+      "grad_norm": 1.3295094966888428,
+      "learning_rate": 4.738870267377974e-05,
+      "loss": 1.9062,
+      "step": 10998
+    },
+    {
+      "epoch": 0.6766117125984252,
+      "grad_norm": 1.1380668878555298,
+      "learning_rate": 4.7372258646288604e-05,
+      "loss": 3.0486,
+      "step": 10999
+    },
+    {
+      "epoch": 0.6766732283464567,
+      "grad_norm": 1.3474626541137695,
+      "learning_rate": 4.7355816586781534e-05,
+      "loss": 1.9487,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6767347440944882,
+      "grad_norm": 1.2492839097976685,
+      "learning_rate": 4.7339376495873345e-05,
+      "loss": 2.5681,
+      "step": 11001
+    },
+    {
+      "epoch": 0.6767962598425197,
+      "grad_norm": 1.1594313383102417,
+      "learning_rate": 4.7322938374178815e-05,
+      "loss": 2.364,
+      "step": 11002
+    },
+    {
+      "epoch": 0.6768577755905512,
+      "grad_norm": 1.4735267162322998,
+      "learning_rate": 4.7306502222312664e-05,
+      "loss": 1.8698,
+      "step": 11003
+    },
+    {
+      "epoch": 0.6769192913385826,
+      "grad_norm": 0.9942310452461243,
+      "learning_rate": 4.729006804088955e-05,
+      "loss": 2.9157,
+      "step": 11004
+    },
+    {
+      "epoch": 0.6769808070866141,
+      "grad_norm": 1.2157108783721924,
+      "learning_rate": 4.727363583052392e-05,
+      "loss": 1.7041,
+      "step": 11005
+    },
+    {
+      "epoch": 0.6770423228346457,
+      "grad_norm": 1.4522472620010376,
+      "learning_rate": 4.725720559183034e-05,
+      "loss": 2.3748,
+      "step": 11006
+    },
+    {
+      "epoch": 0.6771038385826772,
+      "grad_norm": 1.217360019683838,
+      "learning_rate": 4.7240777325423133e-05,
+      "loss": 2.9529,
+      "step": 11007
+    },
+    {
+      "epoch": 0.6771653543307087,
+      "grad_norm": 1.0702288150787354,
+      "learning_rate": 4.7224351031916656e-05,
+      "loss": 1.9274,
+      "step": 11008
+    },
+    {
+      "epoch": 0.6772268700787402,
+      "grad_norm": 1.1996551752090454,
+      "learning_rate": 4.720792671192518e-05,
+      "loss": 2.0308,
+      "step": 11009
+    },
+    {
+      "epoch": 0.6772883858267716,
+      "grad_norm": 1.153746247291565,
+      "learning_rate": 4.719150436606289e-05,
+      "loss": 2.4475,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6773499015748031,
+      "grad_norm": 1.355107307434082,
+      "learning_rate": 4.7175083994943835e-05,
+      "loss": 2.1959,
+      "step": 11011
+    },
+    {
+      "epoch": 0.6774114173228346,
+      "grad_norm": 1.5471177101135254,
+      "learning_rate": 4.7158665599182126e-05,
+      "loss": 2.531,
+      "step": 11012
+    },
+    {
+      "epoch": 0.6774729330708661,
+      "grad_norm": 1.3952678442001343,
+      "learning_rate": 4.714224917939163e-05,
+      "loss": 2.4313,
+      "step": 11013
+    },
+    {
+      "epoch": 0.6775344488188977,
+      "grad_norm": 1.5182172060012817,
+      "learning_rate": 4.712583473618627e-05,
+      "loss": 2.4388,
+      "step": 11014
+    },
+    {
+      "epoch": 0.6775959645669292,
+      "grad_norm": 1.196738839149475,
+      "learning_rate": 4.710942227017985e-05,
+      "loss": 2.6591,
+      "step": 11015
+    },
+    {
+      "epoch": 0.6776574803149606,
+      "grad_norm": 1.3594398498535156,
+      "learning_rate": 4.709301178198615e-05,
+      "loss": 2.3883,
+      "step": 11016
+    },
+    {
+      "epoch": 0.6777189960629921,
+      "grad_norm": 1.2308427095413208,
+      "learning_rate": 4.707660327221877e-05,
+      "loss": 1.7827,
+      "step": 11017
+    },
+    {
+      "epoch": 0.6777805118110236,
+      "grad_norm": 19.371315002441406,
+      "learning_rate": 4.706019674149134e-05,
+      "loss": 2.925,
+      "step": 11018
+    },
+    {
+      "epoch": 0.6778420275590551,
+      "grad_norm": 1.3357356786727905,
+      "learning_rate": 4.704379219041731e-05,
+      "loss": 2.9334,
+      "step": 11019
+    },
+    {
+      "epoch": 0.6779035433070866,
+      "grad_norm": 1.2318631410598755,
+      "learning_rate": 4.702738961961022e-05,
+      "loss": 2.554,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6779650590551181,
+      "grad_norm": 1.5908502340316772,
+      "learning_rate": 4.7010989029683285e-05,
+      "loss": 2.1845,
+      "step": 11021
+    },
+    {
+      "epoch": 0.6780265748031497,
+      "grad_norm": 1.2538155317306519,
+      "learning_rate": 4.699459042124996e-05,
+      "loss": 2.6238,
+      "step": 11022
+    },
+    {
+      "epoch": 0.6780880905511811,
+      "grad_norm": 1.1305484771728516,
+      "learning_rate": 4.6978193794923356e-05,
+      "loss": 2.7131,
+      "step": 11023
+    },
+    {
+      "epoch": 0.6781496062992126,
+      "grad_norm": 1.2794581651687622,
+      "learning_rate": 4.6961799151316675e-05,
+      "loss": 2.5533,
+      "step": 11024
+    },
+    {
+      "epoch": 0.6782111220472441,
+      "grad_norm": 1.2934045791625977,
+      "learning_rate": 4.694540649104292e-05,
+      "loss": 2.3595,
+      "step": 11025
+    },
+    {
+      "epoch": 0.6782726377952756,
+      "grad_norm": 1.51053786277771,
+      "learning_rate": 4.692901581471516e-05,
+      "loss": 2.4094,
+      "step": 11026
+    },
+    {
+      "epoch": 0.6783341535433071,
+      "grad_norm": 1.3664504289627075,
+      "learning_rate": 4.691262712294624e-05,
+      "loss": 2.3117,
+      "step": 11027
+    },
+    {
+      "epoch": 0.6783956692913385,
+      "grad_norm": 1.105607271194458,
+      "learning_rate": 4.6896240416349046e-05,
+      "loss": 2.7812,
+      "step": 11028
+    },
+    {
+      "epoch": 0.67845718503937,
+      "grad_norm": 1.2424882650375366,
+      "learning_rate": 4.687985569553634e-05,
+      "loss": 2.2609,
+      "step": 11029
+    },
+    {
+      "epoch": 0.6785187007874016,
+      "grad_norm": 0.9974716901779175,
+      "learning_rate": 4.686347296112087e-05,
+      "loss": 1.3402,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6785802165354331,
+      "grad_norm": 1.1944910287857056,
+      "learning_rate": 4.684709221371517e-05,
+      "loss": 2.2172,
+      "step": 11031
+    },
+    {
+      "epoch": 0.6786417322834646,
+      "grad_norm": 1.275079369544983,
+      "learning_rate": 4.6830713453931876e-05,
+      "loss": 2.117,
+      "step": 11032
+    },
+    {
+      "epoch": 0.6787032480314961,
+      "grad_norm": 1.3143846988677979,
+      "learning_rate": 4.681433668238338e-05,
+      "loss": 2.9062,
+      "step": 11033
+    },
+    {
+      "epoch": 0.6787647637795275,
+      "grad_norm": 1.3747814893722534,
+      "learning_rate": 4.679796189968212e-05,
+      "loss": 2.2516,
+      "step": 11034
+    },
+    {
+      "epoch": 0.678826279527559,
+      "grad_norm": 1.4378585815429688,
+      "learning_rate": 4.678158910644041e-05,
+      "loss": 2.2726,
+      "step": 11035
+    },
+    {
+      "epoch": 0.6788877952755905,
+      "grad_norm": 1.082740306854248,
+      "learning_rate": 4.6765218303270565e-05,
+      "loss": 2.8099,
+      "step": 11036
+    },
+    {
+      "epoch": 0.678949311023622,
+      "grad_norm": 1.293700098991394,
+      "learning_rate": 4.674884949078471e-05,
+      "loss": 2.0416,
+      "step": 11037
+    },
+    {
+      "epoch": 0.6790108267716536,
+      "grad_norm": 1.2946163415908813,
+      "learning_rate": 4.673248266959491e-05,
+      "loss": 2.5644,
+      "step": 11038
+    },
+    {
+      "epoch": 0.6790723425196851,
+      "grad_norm": 1.5091261863708496,
+      "learning_rate": 4.671611784031322e-05,
+      "loss": 2.6287,
+      "step": 11039
+    },
+    {
+      "epoch": 0.6791338582677166,
+      "grad_norm": 1.1938222646713257,
+      "learning_rate": 4.6699755003551616e-05,
+      "loss": 2.1237,
+      "step": 11040
+    },
+    {
+      "epoch": 0.679195374015748,
+      "grad_norm": 1.2113499641418457,
+      "learning_rate": 4.6683394159922e-05,
+      "loss": 2.4684,
+      "step": 11041
+    },
+    {
+      "epoch": 0.6792568897637795,
+      "grad_norm": 1.5067740678787231,
+      "learning_rate": 4.66670353100361e-05,
+      "loss": 1.9862,
+      "step": 11042
+    },
+    {
+      "epoch": 0.679318405511811,
+      "grad_norm": 1.1450941562652588,
+      "learning_rate": 4.6650678454505727e-05,
+      "loss": 2.835,
+      "step": 11043
+    },
+    {
+      "epoch": 0.6793799212598425,
+      "grad_norm": 1.4744349718093872,
+      "learning_rate": 4.6634323593942455e-05,
+      "loss": 2.3593,
+      "step": 11044
+    },
+    {
+      "epoch": 0.679441437007874,
+      "grad_norm": 1.2466731071472168,
+      "learning_rate": 4.661797072895794e-05,
+      "loss": 2.2111,
+      "step": 11045
+    },
+    {
+      "epoch": 0.6795029527559056,
+      "grad_norm": 1.2698062658309937,
+      "learning_rate": 4.6601619860163615e-05,
+      "loss": 2.4286,
+      "step": 11046
+    },
+    {
+      "epoch": 0.679564468503937,
+      "grad_norm": 1.1055563688278198,
+      "learning_rate": 4.658527098817096e-05,
+      "loss": 2.4679,
+      "step": 11047
+    },
+    {
+      "epoch": 0.6796259842519685,
+      "grad_norm": 1.2048237323760986,
+      "learning_rate": 4.6568924113591306e-05,
+      "loss": 2.662,
+      "step": 11048
+    },
+    {
+      "epoch": 0.6796875,
+      "grad_norm": 1.167694330215454,
+      "learning_rate": 4.6552579237036e-05,
+      "loss": 2.4466,
+      "step": 11049
+    },
+    {
+      "epoch": 0.6797490157480315,
+      "grad_norm": 1.3255537748336792,
+      "learning_rate": 4.653623635911616e-05,
+      "loss": 1.8312,
+      "step": 11050
+    },
+    {
+      "epoch": 0.679810531496063,
+      "grad_norm": 1.283203363418579,
+      "learning_rate": 4.6519895480443e-05,
+      "loss": 2.515,
+      "step": 11051
+    },
+    {
+      "epoch": 0.6798720472440944,
+      "grad_norm": 1.1581982374191284,
+      "learning_rate": 4.6503556601627485e-05,
+      "loss": 2.7305,
+      "step": 11052
+    },
+    {
+      "epoch": 0.679933562992126,
+      "grad_norm": 1.253989338874817,
+      "learning_rate": 4.6487219723280664e-05,
+      "loss": 1.951,
+      "step": 11053
+    },
+    {
+      "epoch": 0.6799950787401575,
+      "grad_norm": 1.2387480735778809,
+      "learning_rate": 4.6470884846013417e-05,
+      "loss": 2.7075,
+      "step": 11054
+    },
+    {
+      "epoch": 0.680056594488189,
+      "grad_norm": 1.3724225759506226,
+      "learning_rate": 4.6454551970436635e-05,
+      "loss": 2.6011,
+      "step": 11055
+    },
+    {
+      "epoch": 0.6801181102362205,
+      "grad_norm": 1.177573800086975,
+      "learning_rate": 4.6438221097161e-05,
+      "loss": 2.8829,
+      "step": 11056
+    },
+    {
+      "epoch": 0.680179625984252,
+      "grad_norm": 1.1713110208511353,
+      "learning_rate": 4.642189222679725e-05,
+      "loss": 2.8043,
+      "step": 11057
+    },
+    {
+      "epoch": 0.6802411417322834,
+      "grad_norm": 1.3698190450668335,
+      "learning_rate": 4.640556535995594e-05,
+      "loss": 2.3107,
+      "step": 11058
+    },
+    {
+      "epoch": 0.6803026574803149,
+      "grad_norm": 1.5799258947372437,
+      "learning_rate": 4.6389240497247645e-05,
+      "loss": 2.4955,
+      "step": 11059
+    },
+    {
+      "epoch": 0.6803641732283464,
+      "grad_norm": 1.12698233127594,
+      "learning_rate": 4.63729176392828e-05,
+      "loss": 2.1341,
+      "step": 11060
+    },
+    {
+      "epoch": 0.680425688976378,
+      "grad_norm": 1.0933713912963867,
+      "learning_rate": 4.6356596786671846e-05,
+      "loss": 2.8224,
+      "step": 11061
+    },
+    {
+      "epoch": 0.6804872047244095,
+      "grad_norm": 1.1930580139160156,
+      "learning_rate": 4.6340277940025e-05,
+      "loss": 2.4388,
+      "step": 11062
+    },
+    {
+      "epoch": 0.680548720472441,
+      "grad_norm": 1.1630215644836426,
+      "learning_rate": 4.63239610999526e-05,
+      "loss": 2.1063,
+      "step": 11063
+    },
+    {
+      "epoch": 0.6806102362204725,
+      "grad_norm": 1.1992642879486084,
+      "learning_rate": 4.63076462670647e-05,
+      "loss": 3.0188,
+      "step": 11064
+    },
+    {
+      "epoch": 0.6806717519685039,
+      "grad_norm": 1.1598749160766602,
+      "learning_rate": 4.629133344197143e-05,
+      "loss": 2.9517,
+      "step": 11065
+    },
+    {
+      "epoch": 0.6807332677165354,
+      "grad_norm": 1.1524288654327393,
+      "learning_rate": 4.627502262528282e-05,
+      "loss": 2.109,
+      "step": 11066
+    },
+    {
+      "epoch": 0.6807947834645669,
+      "grad_norm": 1.3702690601348877,
+      "learning_rate": 4.62587138176088e-05,
+      "loss": 3.2586,
+      "step": 11067
+    },
+    {
+      "epoch": 0.6808562992125984,
+      "grad_norm": 1.2222012281417847,
+      "learning_rate": 4.6242407019559186e-05,
+      "loss": 2.6198,
+      "step": 11068
+    },
+    {
+      "epoch": 0.68091781496063,
+      "grad_norm": 1.2179198265075684,
+      "learning_rate": 4.622610223174383e-05,
+      "loss": 2.2872,
+      "step": 11069
+    },
+    {
+      "epoch": 0.6809793307086615,
+      "grad_norm": 1.4528218507766724,
+      "learning_rate": 4.620979945477235e-05,
+      "loss": 2.1149,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6810408464566929,
+      "grad_norm": 1.1031570434570312,
+      "learning_rate": 4.619349868925448e-05,
+      "loss": 2.3036,
+      "step": 11071
+    },
+    {
+      "epoch": 0.6811023622047244,
+      "grad_norm": 1.4848666191101074,
+      "learning_rate": 4.6177199935799675e-05,
+      "loss": 2.3963,
+      "step": 11072
+    },
+    {
+      "epoch": 0.6811638779527559,
+      "grad_norm": 1.3045158386230469,
+      "learning_rate": 4.616090319501748e-05,
+      "loss": 2.4997,
+      "step": 11073
+    },
+    {
+      "epoch": 0.6812253937007874,
+      "grad_norm": 1.1826212406158447,
+      "learning_rate": 4.614460846751732e-05,
+      "loss": 3.2398,
+      "step": 11074
+    },
+    {
+      "epoch": 0.6812869094488189,
+      "grad_norm": 1.4865082502365112,
+      "learning_rate": 4.6128315753908456e-05,
+      "loss": 1.8132,
+      "step": 11075
+    },
+    {
+      "epoch": 0.6813484251968503,
+      "grad_norm": 1.3005702495574951,
+      "learning_rate": 4.6112025054800237e-05,
+      "loss": 2.0942,
+      "step": 11076
+    },
+    {
+      "epoch": 0.6814099409448819,
+      "grad_norm": 1.6940679550170898,
+      "learning_rate": 4.609573637080174e-05,
+      "loss": 2.6641,
+      "step": 11077
+    },
+    {
+      "epoch": 0.6814714566929134,
+      "grad_norm": 1.8289302587509155,
+      "learning_rate": 4.607944970252214e-05,
+      "loss": 1.8033,
+      "step": 11078
+    },
+    {
+      "epoch": 0.6815329724409449,
+      "grad_norm": 2.845411777496338,
+      "learning_rate": 4.606316505057044e-05,
+      "loss": 2.1823,
+      "step": 11079
+    },
+    {
+      "epoch": 0.6815944881889764,
+      "grad_norm": 1.2701581716537476,
+      "learning_rate": 4.604688241555566e-05,
+      "loss": 3.117,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6816560039370079,
+      "grad_norm": 1.380090594291687,
+      "learning_rate": 4.603060179808657e-05,
+      "loss": 2.9545,
+      "step": 11081
+    },
+    {
+      "epoch": 0.6817175196850394,
+      "grad_norm": 1.3848391771316528,
+      "learning_rate": 4.601432319877208e-05,
+      "loss": 2.1548,
+      "step": 11082
+    },
+    {
+      "epoch": 0.6817790354330708,
+      "grad_norm": 1.1974236965179443,
+      "learning_rate": 4.5998046618220835e-05,
+      "loss": 2.4321,
+      "step": 11083
+    },
+    {
+      "epoch": 0.6818405511811023,
+      "grad_norm": 1.2485660314559937,
+      "learning_rate": 4.5981772057041514e-05,
+      "loss": 3.0784,
+      "step": 11084
+    },
+    {
+      "epoch": 0.6819020669291339,
+      "grad_norm": 1.2622580528259277,
+      "learning_rate": 4.5965499515842715e-05,
+      "loss": 3.2727,
+      "step": 11085
+    },
+    {
+      "epoch": 0.6819635826771654,
+      "grad_norm": 1.352099061012268,
+      "learning_rate": 4.594922899523295e-05,
+      "loss": 2.7875,
+      "step": 11086
+    },
+    {
+      "epoch": 0.6820250984251969,
+      "grad_norm": 1.109038233757019,
+      "learning_rate": 4.593296049582061e-05,
+      "loss": 2.3897,
+      "step": 11087
+    },
+    {
+      "epoch": 0.6820866141732284,
+      "grad_norm": 1.15196692943573,
+      "learning_rate": 4.591669401821408e-05,
+      "loss": 2.5199,
+      "step": 11088
+    },
+    {
+      "epoch": 0.6821481299212598,
+      "grad_norm": 1.3234212398529053,
+      "learning_rate": 4.5900429563021585e-05,
+      "loss": 2.2981,
+      "step": 11089
+    },
+    {
+      "epoch": 0.6822096456692913,
+      "grad_norm": 1.15436589717865,
+      "learning_rate": 4.58841671308514e-05,
+      "loss": 1.884,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6822711614173228,
+      "grad_norm": 1.2366632223129272,
+      "learning_rate": 4.5867906722311525e-05,
+      "loss": 2.3352,
+      "step": 11091
+    },
+    {
+      "epoch": 0.6823326771653543,
+      "grad_norm": 1.2324519157409668,
+      "learning_rate": 4.585164833801017e-05,
+      "loss": 1.9565,
+      "step": 11092
+    },
+    {
+      "epoch": 0.6823941929133859,
+      "grad_norm": 1.3285222053527832,
+      "learning_rate": 4.583539197855519e-05,
+      "loss": 2.8396,
+      "step": 11093
+    },
+    {
+      "epoch": 0.6824557086614174,
+      "grad_norm": 1.4986624717712402,
+      "learning_rate": 4.581913764455456e-05,
+      "loss": 2.2448,
+      "step": 11094
+    },
+    {
+      "epoch": 0.6825172244094488,
+      "grad_norm": 1.2683942317962646,
+      "learning_rate": 4.580288533661601e-05,
+      "loss": 2.4229,
+      "step": 11095
+    },
+    {
+      "epoch": 0.6825787401574803,
+      "grad_norm": 1.3523683547973633,
+      "learning_rate": 4.5786635055347385e-05,
+      "loss": 2.1446,
+      "step": 11096
+    },
+    {
+      "epoch": 0.6826402559055118,
+      "grad_norm": 1.196533441543579,
+      "learning_rate": 4.5770386801356255e-05,
+      "loss": 2.5917,
+      "step": 11097
+    },
+    {
+      "epoch": 0.6827017716535433,
+      "grad_norm": 1.332192301750183,
+      "learning_rate": 4.575414057525028e-05,
+      "loss": 2.4173,
+      "step": 11098
+    },
+    {
+      "epoch": 0.6827632874015748,
+      "grad_norm": 1.1829917430877686,
+      "learning_rate": 4.573789637763697e-05,
+      "loss": 2.481,
+      "step": 11099
+    },
+    {
+      "epoch": 0.6828248031496063,
+      "grad_norm": 1.4137818813323975,
+      "learning_rate": 4.572165420912379e-05,
+      "loss": 2.5215,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6828863188976378,
+      "grad_norm": 1.3990817070007324,
+      "learning_rate": 4.570541407031804e-05,
+      "loss": 2.4423,
+      "step": 11101
+    },
+    {
+      "epoch": 0.6829478346456693,
+      "grad_norm": 1.252293586730957,
+      "learning_rate": 4.568917596182709e-05,
+      "loss": 2.5112,
+      "step": 11102
+    },
+    {
+      "epoch": 0.6830093503937008,
+      "grad_norm": 1.346498966217041,
+      "learning_rate": 4.567293988425807e-05,
+      "loss": 2.4671,
+      "step": 11103
+    },
+    {
+      "epoch": 0.6830708661417323,
+      "grad_norm": 1.3447003364562988,
+      "learning_rate": 4.565670583821817e-05,
+      "loss": 2.4836,
+      "step": 11104
+    },
+    {
+      "epoch": 0.6831323818897638,
+      "grad_norm": 1.1447691917419434,
+      "learning_rate": 4.5640473824314443e-05,
+      "loss": 2.9592,
+      "step": 11105
+    },
+    {
+      "epoch": 0.6831938976377953,
+      "grad_norm": 1.2404060363769531,
+      "learning_rate": 4.5624243843153916e-05,
+      "loss": 2.4373,
+      "step": 11106
+    },
+    {
+      "epoch": 0.6832554133858267,
+      "grad_norm": 1.4039735794067383,
+      "learning_rate": 4.5608015895343426e-05,
+      "loss": 2.0961,
+      "step": 11107
+    },
+    {
+      "epoch": 0.6833169291338582,
+      "grad_norm": 1.4295811653137207,
+      "learning_rate": 4.5591789981489876e-05,
+      "loss": 2.8806,
+      "step": 11108
+    },
+    {
+      "epoch": 0.6833784448818898,
+      "grad_norm": 1.1650720834732056,
+      "learning_rate": 4.557556610219995e-05,
+      "loss": 2.1438,
+      "step": 11109
+    },
+    {
+      "epoch": 0.6834399606299213,
+      "grad_norm": 1.4595915079116821,
+      "learning_rate": 4.5559344258080385e-05,
+      "loss": 2.4614,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6835014763779528,
+      "grad_norm": 1.2536201477050781,
+      "learning_rate": 4.554312444973781e-05,
+      "loss": 2.4054,
+      "step": 11111
+    },
+    {
+      "epoch": 0.6835629921259843,
+      "grad_norm": 1.3394126892089844,
+      "learning_rate": 4.552690667777868e-05,
+      "loss": 1.9163,
+      "step": 11112
+    },
+    {
+      "epoch": 0.6836245078740157,
+      "grad_norm": 1.4614338874816895,
+      "learning_rate": 4.551069094280953e-05,
+      "loss": 2.2399,
+      "step": 11113
+    },
+    {
+      "epoch": 0.6836860236220472,
+      "grad_norm": 1.3099151849746704,
+      "learning_rate": 4.549447724543665e-05,
+      "loss": 2.7628,
+      "step": 11114
+    },
+    {
+      "epoch": 0.6837475393700787,
+      "grad_norm": 1.1502212285995483,
+      "learning_rate": 4.547826558626645e-05,
+      "loss": 1.294,
+      "step": 11115
+    },
+    {
+      "epoch": 0.6838090551181102,
+      "grad_norm": 1.5598812103271484,
+      "learning_rate": 4.546205596590504e-05,
+      "loss": 2.2112,
+      "step": 11116
+    },
+    {
+      "epoch": 0.6838705708661418,
+      "grad_norm": 1.3119310140609741,
+      "learning_rate": 4.544584838495862e-05,
+      "loss": 1.8538,
+      "step": 11117
+    },
+    {
+      "epoch": 0.6839320866141733,
+      "grad_norm": 1.217003583908081,
+      "learning_rate": 4.542964284403328e-05,
+      "loss": 2.1556,
+      "step": 11118
+    },
+    {
+      "epoch": 0.6839936023622047,
+      "grad_norm": 1.4028228521347046,
+      "learning_rate": 4.5413439343735045e-05,
+      "loss": 1.8772,
+      "step": 11119
+    },
+    {
+      "epoch": 0.6840551181102362,
+      "grad_norm": 1.174930453300476,
+      "learning_rate": 4.539723788466974e-05,
+      "loss": 2.8025,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6841166338582677,
+      "grad_norm": 1.14895761013031,
+      "learning_rate": 4.538103846744331e-05,
+      "loss": 2.3764,
+      "step": 11121
+    },
+    {
+      "epoch": 0.6841781496062992,
+      "grad_norm": 1.2605655193328857,
+      "learning_rate": 4.5364841092661456e-05,
+      "loss": 3.0155,
+      "step": 11122
+    },
+    {
+      "epoch": 0.6842396653543307,
+      "grad_norm": 1.3300578594207764,
+      "learning_rate": 4.534864576092988e-05,
+      "loss": 2.5308,
+      "step": 11123
+    },
+    {
+      "epoch": 0.6843011811023622,
+      "grad_norm": 1.2820647954940796,
+      "learning_rate": 4.53324524728542e-05,
+      "loss": 1.4143,
+      "step": 11124
+    },
+    {
+      "epoch": 0.6843626968503937,
+      "grad_norm": 1.3928333520889282,
+      "learning_rate": 4.531626122904001e-05,
+      "loss": 2.0811,
+      "step": 11125
+    },
+    {
+      "epoch": 0.6844242125984252,
+      "grad_norm": 1.2873207330703735,
+      "learning_rate": 4.530007203009269e-05,
+      "loss": 2.2745,
+      "step": 11126
+    },
+    {
+      "epoch": 0.6844857283464567,
+      "grad_norm": 1.4645237922668457,
+      "learning_rate": 4.52838848766177e-05,
+      "loss": 2.0691,
+      "step": 11127
+    },
+    {
+      "epoch": 0.6845472440944882,
+      "grad_norm": 1.1395366191864014,
+      "learning_rate": 4.526769976922026e-05,
+      "loss": 2.6263,
+      "step": 11128
+    },
+    {
+      "epoch": 0.6846087598425197,
+      "grad_norm": 1.1960272789001465,
+      "learning_rate": 4.525151670850567e-05,
+      "loss": 2.6405,
+      "step": 11129
+    },
+    {
+      "epoch": 0.6846702755905512,
+      "grad_norm": 1.3328568935394287,
+      "learning_rate": 4.523533569507905e-05,
+      "loss": 2.9263,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6847317913385826,
+      "grad_norm": 1.2884347438812256,
+      "learning_rate": 4.5219156729545556e-05,
+      "loss": 2.5568,
+      "step": 11131
+    },
+    {
+      "epoch": 0.6847933070866141,
+      "grad_norm": 1.2649637460708618,
+      "learning_rate": 4.520297981251008e-05,
+      "loss": 2.7696,
+      "step": 11132
+    },
+    {
+      "epoch": 0.6848548228346457,
+      "grad_norm": 1.3475596904754639,
+      "learning_rate": 4.5186804944577644e-05,
+      "loss": 2.1466,
+      "step": 11133
+    },
+    {
+      "epoch": 0.6849163385826772,
+      "grad_norm": 1.3511691093444824,
+      "learning_rate": 4.517063212635303e-05,
+      "loss": 2.3812,
+      "step": 11134
+    },
+    {
+      "epoch": 0.6849778543307087,
+      "grad_norm": 1.2180882692337036,
+      "learning_rate": 4.5154461358441056e-05,
+      "loss": 2.7168,
+      "step": 11135
+    },
+    {
+      "epoch": 0.6850393700787402,
+      "grad_norm": 1.158154845237732,
+      "learning_rate": 4.513829264144635e-05,
+      "loss": 2.5026,
+      "step": 11136
+    },
+    {
+      "epoch": 0.6851008858267716,
+      "grad_norm": 1.1797356605529785,
+      "learning_rate": 4.512212597597363e-05,
+      "loss": 2.5531,
+      "step": 11137
+    },
+    {
+      "epoch": 0.6851624015748031,
+      "grad_norm": 1.2036038637161255,
+      "learning_rate": 4.5105961362627366e-05,
+      "loss": 2.4398,
+      "step": 11138
+    },
+    {
+      "epoch": 0.6852239173228346,
+      "grad_norm": 1.0745625495910645,
+      "learning_rate": 4.508979880201209e-05,
+      "loss": 3.0124,
+      "step": 11139
+    },
+    {
+      "epoch": 0.6852854330708661,
+      "grad_norm": 1.2286382913589478,
+      "learning_rate": 4.507363829473211e-05,
+      "loss": 2.6318,
+      "step": 11140
+    },
+    {
+      "epoch": 0.6853469488188977,
+      "grad_norm": 1.5769041776657104,
+      "learning_rate": 4.505747984139181e-05,
+      "loss": 2.2831,
+      "step": 11141
+    },
+    {
+      "epoch": 0.6854084645669292,
+      "grad_norm": 1.1688791513442993,
+      "learning_rate": 4.504132344259537e-05,
+      "loss": 2.8109,
+      "step": 11142
+    },
+    {
+      "epoch": 0.6854699803149606,
+      "grad_norm": 1.7312004566192627,
+      "learning_rate": 4.502516909894697e-05,
+      "loss": 2.1346,
+      "step": 11143
+    },
+    {
+      "epoch": 0.6855314960629921,
+      "grad_norm": 1.1531943082809448,
+      "learning_rate": 4.500901681105071e-05,
+      "loss": 2.4591,
+      "step": 11144
+    },
+    {
+      "epoch": 0.6855930118110236,
+      "grad_norm": 1.5502070188522339,
+      "learning_rate": 4.499286657951061e-05,
+      "loss": 2.3319,
+      "step": 11145
+    },
+    {
+      "epoch": 0.6856545275590551,
+      "grad_norm": 1.1939208507537842,
+      "learning_rate": 4.4976718404930526e-05,
+      "loss": 2.1204,
+      "step": 11146
+    },
+    {
+      "epoch": 0.6857160433070866,
+      "grad_norm": 1.433640956878662,
+      "learning_rate": 4.49605722879144e-05,
+      "loss": 2.2351,
+      "step": 11147
+    },
+    {
+      "epoch": 0.6857775590551181,
+      "grad_norm": 1.281654953956604,
+      "learning_rate": 4.4944428229065925e-05,
+      "loss": 2.4989,
+      "step": 11148
+    },
+    {
+      "epoch": 0.6858390748031497,
+      "grad_norm": 1.1996405124664307,
+      "learning_rate": 4.4928286228988835e-05,
+      "loss": 2.3556,
+      "step": 11149
+    },
+    {
+      "epoch": 0.6859005905511811,
+      "grad_norm": 1.217638373374939,
+      "learning_rate": 4.4912146288286795e-05,
+      "loss": 2.4241,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6859621062992126,
+      "grad_norm": 1.395323634147644,
+      "learning_rate": 4.489600840756326e-05,
+      "loss": 2.0875,
+      "step": 11151
+    },
+    {
+      "epoch": 0.6860236220472441,
+      "grad_norm": 1.3502075672149658,
+      "learning_rate": 4.487987258742179e-05,
+      "loss": 2.0748,
+      "step": 11152
+    },
+    {
+      "epoch": 0.6860851377952756,
+      "grad_norm": 1.2015812397003174,
+      "learning_rate": 4.4863738828465686e-05,
+      "loss": 2.1351,
+      "step": 11153
+    },
+    {
+      "epoch": 0.6861466535433071,
+      "grad_norm": 1.6101709604263306,
+      "learning_rate": 4.48476071312983e-05,
+      "loss": 1.9835,
+      "step": 11154
+    },
+    {
+      "epoch": 0.6862081692913385,
+      "grad_norm": 1.1352142095565796,
+      "learning_rate": 4.483147749652287e-05,
+      "loss": 3.0199,
+      "step": 11155
+    },
+    {
+      "epoch": 0.68626968503937,
+      "grad_norm": 1.1461927890777588,
+      "learning_rate": 4.481534992474258e-05,
+      "loss": 1.9949,
+      "step": 11156
+    },
+    {
+      "epoch": 0.6863312007874016,
+      "grad_norm": 1.2451581954956055,
+      "learning_rate": 4.479922441656046e-05,
+      "loss": 2.3869,
+      "step": 11157
+    },
+    {
+      "epoch": 0.6863927165354331,
+      "grad_norm": 1.2874330282211304,
+      "learning_rate": 4.478310097257957e-05,
+      "loss": 2.5974,
+      "step": 11158
+    },
+    {
+      "epoch": 0.6864542322834646,
+      "grad_norm": 1.2548911571502686,
+      "learning_rate": 4.476697959340277e-05,
+      "loss": 2.7028,
+      "step": 11159
+    },
+    {
+      "epoch": 0.6865157480314961,
+      "grad_norm": 1.1209766864776611,
+      "learning_rate": 4.475086027963299e-05,
+      "loss": 2.8027,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6865772637795275,
+      "grad_norm": 1.1497410535812378,
+      "learning_rate": 4.47347430318729e-05,
+      "loss": 2.0451,
+      "step": 11161
+    },
+    {
+      "epoch": 0.686638779527559,
+      "grad_norm": 1.2751883268356323,
+      "learning_rate": 4.471862785072528e-05,
+      "loss": 2.6146,
+      "step": 11162
+    },
+    {
+      "epoch": 0.6867002952755905,
+      "grad_norm": 1.1978905200958252,
+      "learning_rate": 4.47025147367927e-05,
+      "loss": 2.6053,
+      "step": 11163
+    },
+    {
+      "epoch": 0.686761811023622,
+      "grad_norm": 1.32436203956604,
+      "learning_rate": 4.468640369067777e-05,
+      "loss": 2.0798,
+      "step": 11164
+    },
+    {
+      "epoch": 0.6868233267716536,
+      "grad_norm": 1.2411268949508667,
+      "learning_rate": 4.467029471298286e-05,
+      "loss": 3.1255,
+      "step": 11165
+    },
+    {
+      "epoch": 0.6868848425196851,
+      "grad_norm": 1.3170537948608398,
+      "learning_rate": 4.465418780431044e-05,
+      "loss": 2.4766,
+      "step": 11166
+    },
+    {
+      "epoch": 0.6869463582677166,
+      "grad_norm": 1.3640908002853394,
+      "learning_rate": 4.463808296526275e-05,
+      "loss": 2.89,
+      "step": 11167
+    },
+    {
+      "epoch": 0.687007874015748,
+      "grad_norm": 1.0832957029342651,
+      "learning_rate": 4.462198019644205e-05,
+      "loss": 2.747,
+      "step": 11168
+    },
+    {
+      "epoch": 0.6870693897637795,
+      "grad_norm": 1.3680444955825806,
+      "learning_rate": 4.4605879498450485e-05,
+      "loss": 2.1118,
+      "step": 11169
+    },
+    {
+      "epoch": 0.687130905511811,
+      "grad_norm": 1.2572617530822754,
+      "learning_rate": 4.458978087189018e-05,
+      "loss": 1.9092,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6871924212598425,
+      "grad_norm": 1.2069846391677856,
+      "learning_rate": 4.4573684317363065e-05,
+      "loss": 3.0303,
+      "step": 11171
+    },
+    {
+      "epoch": 0.687253937007874,
+      "grad_norm": 1.8827955722808838,
+      "learning_rate": 4.455758983547112e-05,
+      "loss": 2.2382,
+      "step": 11172
+    },
+    {
+      "epoch": 0.6873154527559056,
+      "grad_norm": 1.3835970163345337,
+      "learning_rate": 4.454149742681613e-05,
+      "loss": 2.4376,
+      "step": 11173
+    },
+    {
+      "epoch": 0.687376968503937,
+      "grad_norm": 1.2673808336257935,
+      "learning_rate": 4.452540709199989e-05,
+      "loss": 1.9244,
+      "step": 11174
+    },
+    {
+      "epoch": 0.6874384842519685,
+      "grad_norm": 1.134384036064148,
+      "learning_rate": 4.450931883162408e-05,
+      "loss": 2.7361,
+      "step": 11175
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 1.2948858737945557,
+      "learning_rate": 4.449323264629037e-05,
+      "loss": 1.397,
+      "step": 11176
+    },
+    {
+      "epoch": 0.6875615157480315,
+      "grad_norm": 1.3720020055770874,
+      "learning_rate": 4.4477148536600196e-05,
+      "loss": 2.735,
+      "step": 11177
+    },
+    {
+      "epoch": 0.687623031496063,
+      "grad_norm": 1.1966447830200195,
+      "learning_rate": 4.4461066503155115e-05,
+      "loss": 2.8206,
+      "step": 11178
+    },
+    {
+      "epoch": 0.6876845472440944,
+      "grad_norm": 2.1086270809173584,
+      "learning_rate": 4.444498654655641e-05,
+      "loss": 2.1591,
+      "step": 11179
+    },
+    {
+      "epoch": 0.687746062992126,
+      "grad_norm": 1.404229760169983,
+      "learning_rate": 4.442890866740542e-05,
+      "loss": 2.3778,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6878075787401575,
+      "grad_norm": 1.4363739490509033,
+      "learning_rate": 4.441283286630338e-05,
+      "loss": 2.2527,
+      "step": 11181
+    },
+    {
+      "epoch": 0.687869094488189,
+      "grad_norm": 1.2843161821365356,
+      "learning_rate": 4.4396759143851466e-05,
+      "loss": 2.4453,
+      "step": 11182
+    },
+    {
+      "epoch": 0.6879306102362205,
+      "grad_norm": 1.1517702341079712,
+      "learning_rate": 4.438068750065065e-05,
+      "loss": 2.6777,
+      "step": 11183
+    },
+    {
+      "epoch": 0.687992125984252,
+      "grad_norm": 1.1420795917510986,
+      "learning_rate": 4.4364617937302047e-05,
+      "loss": 2.5259,
+      "step": 11184
+    },
+    {
+      "epoch": 0.6880536417322834,
+      "grad_norm": 1.1812490224838257,
+      "learning_rate": 4.434855045440648e-05,
+      "loss": 2.3535,
+      "step": 11185
+    },
+    {
+      "epoch": 0.6881151574803149,
+      "grad_norm": 1.2152148485183716,
+      "learning_rate": 4.433248505256478e-05,
+      "loss": 2.7407,
+      "step": 11186
+    },
+    {
+      "epoch": 0.6881766732283464,
+      "grad_norm": 1.1670631170272827,
+      "learning_rate": 4.431642173237772e-05,
+      "loss": 2.277,
+      "step": 11187
+    },
+    {
+      "epoch": 0.688238188976378,
+      "grad_norm": 1.3489243984222412,
+      "learning_rate": 4.4300360494445994e-05,
+      "loss": 2.5913,
+      "step": 11188
+    },
+    {
+      "epoch": 0.6882997047244095,
+      "grad_norm": 1.2122858762741089,
+      "learning_rate": 4.428430133937022e-05,
+      "loss": 1.8903,
+      "step": 11189
+    },
+    {
+      "epoch": 0.688361220472441,
+      "grad_norm": 1.2869611978530884,
+      "learning_rate": 4.426824426775088e-05,
+      "loss": 2.2313,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6884227362204725,
+      "grad_norm": 1.3683571815490723,
+      "learning_rate": 4.425218928018846e-05,
+      "loss": 2.1571,
+      "step": 11191
+    },
+    {
+      "epoch": 0.6884842519685039,
+      "grad_norm": 1.3616451025009155,
+      "learning_rate": 4.423613637728327e-05,
+      "loss": 2.6306,
+      "step": 11192
+    },
+    {
+      "epoch": 0.6885457677165354,
+      "grad_norm": 1.172034740447998,
+      "learning_rate": 4.422008555963563e-05,
+      "loss": 2.3254,
+      "step": 11193
+    },
+    {
+      "epoch": 0.6886072834645669,
+      "grad_norm": 1.3781684637069702,
+      "learning_rate": 4.420403682784575e-05,
+      "loss": 1.442,
+      "step": 11194
+    },
+    {
+      "epoch": 0.6886687992125984,
+      "grad_norm": 1.1122219562530518,
+      "learning_rate": 4.4187990182513805e-05,
+      "loss": 2.373,
+      "step": 11195
+    },
+    {
+      "epoch": 0.68873031496063,
+      "grad_norm": 1.687192678451538,
+      "learning_rate": 4.4171945624239775e-05,
+      "loss": 2.3533,
+      "step": 11196
+    },
+    {
+      "epoch": 0.6887918307086615,
+      "grad_norm": 2.7163963317871094,
+      "learning_rate": 4.41559031536237e-05,
+      "loss": 3.1445,
+      "step": 11197
+    },
+    {
+      "epoch": 0.6888533464566929,
+      "grad_norm": 1.2237862348556519,
+      "learning_rate": 4.413986277126543e-05,
+      "loss": 1.9397,
+      "step": 11198
+    },
+    {
+      "epoch": 0.6889148622047244,
+      "grad_norm": 1.2723268270492554,
+      "learning_rate": 4.41238244777648e-05,
+      "loss": 2.6938,
+      "step": 11199
+    },
+    {
+      "epoch": 0.6889763779527559,
+      "grad_norm": 1.4682092666625977,
+      "learning_rate": 4.410778827372156e-05,
+      "loss": 2.7472,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6890378937007874,
+      "grad_norm": 1.4850456714630127,
+      "learning_rate": 4.409175415973541e-05,
+      "loss": 2.861,
+      "step": 11201
+    },
+    {
+      "epoch": 0.6890994094488189,
+      "grad_norm": 1.5862480401992798,
+      "learning_rate": 4.407572213640587e-05,
+      "loss": 2.314,
+      "step": 11202
+    },
+    {
+      "epoch": 0.6891609251968503,
+      "grad_norm": 1.2887494564056396,
+      "learning_rate": 4.4059692204332515e-05,
+      "loss": 2.9634,
+      "step": 11203
+    },
+    {
+      "epoch": 0.6892224409448819,
+      "grad_norm": 1.2522789239883423,
+      "learning_rate": 4.40436643641147e-05,
+      "loss": 2.2315,
+      "step": 11204
+    },
+    {
+      "epoch": 0.6892839566929134,
+      "grad_norm": 1.0713680982589722,
+      "learning_rate": 4.402763861635186e-05,
+      "loss": 2.9744,
+      "step": 11205
+    },
+    {
+      "epoch": 0.6893454724409449,
+      "grad_norm": 1.3885293006896973,
+      "learning_rate": 4.4011614961643155e-05,
+      "loss": 2.9498,
+      "step": 11206
+    },
+    {
+      "epoch": 0.6894069881889764,
+      "grad_norm": 1.0817904472351074,
+      "learning_rate": 4.399559340058791e-05,
+      "loss": 3.0236,
+      "step": 11207
+    },
+    {
+      "epoch": 0.6894685039370079,
+      "grad_norm": 1.233761191368103,
+      "learning_rate": 4.3979573933785155e-05,
+      "loss": 2.7347,
+      "step": 11208
+    },
+    {
+      "epoch": 0.6895300196850394,
+      "grad_norm": 1.167093276977539,
+      "learning_rate": 4.3963556561834e-05,
+      "loss": 2.2962,
+      "step": 11209
+    },
+    {
+      "epoch": 0.6895915354330708,
+      "grad_norm": 1.2021968364715576,
+      "learning_rate": 4.3947541285333305e-05,
+      "loss": 1.8508,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6896530511811023,
+      "grad_norm": 1.4468834400177002,
+      "learning_rate": 4.393152810488206e-05,
+      "loss": 1.9206,
+      "step": 11211
+    },
+    {
+      "epoch": 0.6897145669291339,
+      "grad_norm": 1.286860704421997,
+      "learning_rate": 4.391551702107898e-05,
+      "loss": 1.9647,
+      "step": 11212
+    },
+    {
+      "epoch": 0.6897760826771654,
+      "grad_norm": 1.3548715114593506,
+      "learning_rate": 4.389950803452282e-05,
+      "loss": 2.2771,
+      "step": 11213
+    },
+    {
+      "epoch": 0.6898375984251969,
+      "grad_norm": 1.2413135766983032,
+      "learning_rate": 4.3883501145812246e-05,
+      "loss": 2.1204,
+      "step": 11214
+    },
+    {
+      "epoch": 0.6898991141732284,
+      "grad_norm": 1.330621600151062,
+      "learning_rate": 4.386749635554584e-05,
+      "loss": 2.2273,
+      "step": 11215
+    },
+    {
+      "epoch": 0.6899606299212598,
+      "grad_norm": 1.0392115116119385,
+      "learning_rate": 4.385149366432204e-05,
+      "loss": 2.8314,
+      "step": 11216
+    },
+    {
+      "epoch": 0.6900221456692913,
+      "grad_norm": 1.2225290536880493,
+      "learning_rate": 4.383549307273932e-05,
+      "loss": 1.9247,
+      "step": 11217
+    },
+    {
+      "epoch": 0.6900836614173228,
+      "grad_norm": 1.3462010622024536,
+      "learning_rate": 4.3819494581395934e-05,
+      "loss": 2.1863,
+      "step": 11218
+    },
+    {
+      "epoch": 0.6901451771653543,
+      "grad_norm": 1.2545464038848877,
+      "learning_rate": 4.380349819089018e-05,
+      "loss": 2.8272,
+      "step": 11219
+    },
+    {
+      "epoch": 0.6902066929133859,
+      "grad_norm": 1.3328280448913574,
+      "learning_rate": 4.378750390182024e-05,
+      "loss": 2.3485,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6902682086614174,
+      "grad_norm": 1.2099566459655762,
+      "learning_rate": 4.377151171478424e-05,
+      "loss": 2.5146,
+      "step": 11221
+    },
+    {
+      "epoch": 0.6903297244094488,
+      "grad_norm": 1.3009681701660156,
+      "learning_rate": 4.3755521630380157e-05,
+      "loss": 2.5029,
+      "step": 11222
+    },
+    {
+      "epoch": 0.6903912401574803,
+      "grad_norm": 1.195154070854187,
+      "learning_rate": 4.37395336492059e-05,
+      "loss": 2.7177,
+      "step": 11223
+    },
+    {
+      "epoch": 0.6904527559055118,
+      "grad_norm": 1.4023717641830444,
+      "learning_rate": 4.3723547771859374e-05,
+      "loss": 2.477,
+      "step": 11224
+    },
+    {
+      "epoch": 0.6905142716535433,
+      "grad_norm": 1.2023096084594727,
+      "learning_rate": 4.370756399893835e-05,
+      "loss": 2.527,
+      "step": 11225
+    },
+    {
+      "epoch": 0.6905757874015748,
+      "grad_norm": 1.4967448711395264,
+      "learning_rate": 4.369158233104058e-05,
+      "loss": 2.4052,
+      "step": 11226
+    },
+    {
+      "epoch": 0.6906373031496063,
+      "grad_norm": 1.278667688369751,
+      "learning_rate": 4.367560276876361e-05,
+      "loss": 1.8693,
+      "step": 11227
+    },
+    {
+      "epoch": 0.6906988188976378,
+      "grad_norm": 1.0510917901992798,
+      "learning_rate": 4.3659625312705055e-05,
+      "loss": 2.4868,
+      "step": 11228
+    },
+    {
+      "epoch": 0.6907603346456693,
+      "grad_norm": 1.4220894575119019,
+      "learning_rate": 4.364364996346232e-05,
+      "loss": 1.9747,
+      "step": 11229
+    },
+    {
+      "epoch": 0.6908218503937008,
+      "grad_norm": 1.212723731994629,
+      "learning_rate": 4.362767672163285e-05,
+      "loss": 2.6693,
+      "step": 11230
+    },
+    {
+      "epoch": 0.6908833661417323,
+      "grad_norm": 1.46207857131958,
+      "learning_rate": 4.36117055878139e-05,
+      "loss": 2.6914,
+      "step": 11231
+    },
+    {
+      "epoch": 0.6909448818897638,
+      "grad_norm": 1.1487475633621216,
+      "learning_rate": 4.359573656260274e-05,
+      "loss": 2.4971,
+      "step": 11232
+    },
+    {
+      "epoch": 0.6910063976377953,
+      "grad_norm": 1.6013535261154175,
+      "learning_rate": 4.35797696465965e-05,
+      "loss": 2.3341,
+      "step": 11233
+    },
+    {
+      "epoch": 0.6910679133858267,
+      "grad_norm": 1.1072216033935547,
+      "learning_rate": 4.3563804840392306e-05,
+      "loss": 2.0061,
+      "step": 11234
+    },
+    {
+      "epoch": 0.6911294291338582,
+      "grad_norm": 1.3103364706039429,
+      "learning_rate": 4.354784214458708e-05,
+      "loss": 3.0928,
+      "step": 11235
+    },
+    {
+      "epoch": 0.6911909448818898,
+      "grad_norm": 1.4101709127426147,
+      "learning_rate": 4.353188155977781e-05,
+      "loss": 2.4098,
+      "step": 11236
+    },
+    {
+      "epoch": 0.6912524606299213,
+      "grad_norm": 1.2414617538452148,
+      "learning_rate": 4.351592308656126e-05,
+      "loss": 2.8197,
+      "step": 11237
+    },
+    {
+      "epoch": 0.6913139763779528,
+      "grad_norm": 1.235758900642395,
+      "learning_rate": 4.3499966725534223e-05,
+      "loss": 2.288,
+      "step": 11238
+    },
+    {
+      "epoch": 0.6913754921259843,
+      "grad_norm": 1.384408712387085,
+      "learning_rate": 4.348401247729338e-05,
+      "loss": 2.2785,
+      "step": 11239
+    },
+    {
+      "epoch": 0.6914370078740157,
+      "grad_norm": 1.4128786325454712,
+      "learning_rate": 4.346806034243536e-05,
+      "loss": 2.7476,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6914985236220472,
+      "grad_norm": 1.2287607192993164,
+      "learning_rate": 4.345211032155661e-05,
+      "loss": 2.606,
+      "step": 11241
+    },
+    {
+      "epoch": 0.6915600393700787,
+      "grad_norm": 1.4173192977905273,
+      "learning_rate": 4.343616241525366e-05,
+      "loss": 1.8898,
+      "step": 11242
+    },
+    {
+      "epoch": 0.6916215551181102,
+      "grad_norm": 1.1756000518798828,
+      "learning_rate": 4.342021662412279e-05,
+      "loss": 2.5972,
+      "step": 11243
+    },
+    {
+      "epoch": 0.6916830708661418,
+      "grad_norm": 1.1357073783874512,
+      "learning_rate": 4.3404272948760314e-05,
+      "loss": 3.0749,
+      "step": 11244
+    },
+    {
+      "epoch": 0.6917445866141733,
+      "grad_norm": 1.2155001163482666,
+      "learning_rate": 4.338833138976245e-05,
+      "loss": 1.8265,
+      "step": 11245
+    },
+    {
+      "epoch": 0.6918061023622047,
+      "grad_norm": 1.3489793539047241,
+      "learning_rate": 4.3372391947725344e-05,
+      "loss": 2.0003,
+      "step": 11246
+    },
+    {
+      "epoch": 0.6918676181102362,
+      "grad_norm": 1.1678704023361206,
+      "learning_rate": 4.335645462324499e-05,
+      "loss": 2.6462,
+      "step": 11247
+    },
+    {
+      "epoch": 0.6919291338582677,
+      "grad_norm": 1.3216971158981323,
+      "learning_rate": 4.334051941691741e-05,
+      "loss": 2.0741,
+      "step": 11248
+    },
+    {
+      "epoch": 0.6919906496062992,
+      "grad_norm": 1.444808840751648,
+      "learning_rate": 4.332458632933842e-05,
+      "loss": 2.4728,
+      "step": 11249
+    },
+    {
+      "epoch": 0.6920521653543307,
+      "grad_norm": 1.5208234786987305,
+      "learning_rate": 4.330865536110388e-05,
+      "loss": 2.2947,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6921136811023622,
+      "grad_norm": 1.3833328485488892,
+      "learning_rate": 4.329272651280951e-05,
+      "loss": 2.209,
+      "step": 11251
+    },
+    {
+      "epoch": 0.6921751968503937,
+      "grad_norm": 1.1708029508590698,
+      "learning_rate": 4.327679978505099e-05,
+      "loss": 3.0215,
+      "step": 11252
+    },
+    {
+      "epoch": 0.6922367125984252,
+      "grad_norm": 1.3440003395080566,
+      "learning_rate": 4.3260875178423834e-05,
+      "loss": 2.749,
+      "step": 11253
+    },
+    {
+      "epoch": 0.6922982283464567,
+      "grad_norm": 1.2830920219421387,
+      "learning_rate": 4.3244952693523586e-05,
+      "loss": 3.2256,
+      "step": 11254
+    },
+    {
+      "epoch": 0.6923597440944882,
+      "grad_norm": 2.4820048809051514,
+      "learning_rate": 4.322903233094561e-05,
+      "loss": 2.8968,
+      "step": 11255
+    },
+    {
+      "epoch": 0.6924212598425197,
+      "grad_norm": 1.9391471147537231,
+      "learning_rate": 4.3213114091285276e-05,
+      "loss": 1.6636,
+      "step": 11256
+    },
+    {
+      "epoch": 0.6924827755905512,
+      "grad_norm": 1.303881287574768,
+      "learning_rate": 4.31971979751378e-05,
+      "loss": 2.2848,
+      "step": 11257
+    },
+    {
+      "epoch": 0.6925442913385826,
+      "grad_norm": 1.134962558746338,
+      "learning_rate": 4.318128398309839e-05,
+      "loss": 2.6389,
+      "step": 11258
+    },
+    {
+      "epoch": 0.6926058070866141,
+      "grad_norm": 1.281459927558899,
+      "learning_rate": 4.3165372115762145e-05,
+      "loss": 2.1883,
+      "step": 11259
+    },
+    {
+      "epoch": 0.6926673228346457,
+      "grad_norm": 1.1458675861358643,
+      "learning_rate": 4.3149462373724035e-05,
+      "loss": 2.1082,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6927288385826772,
+      "grad_norm": 1.2152944803237915,
+      "learning_rate": 4.313355475757906e-05,
+      "loss": 2.7852,
+      "step": 11261
+    },
+    {
+      "epoch": 0.6927903543307087,
+      "grad_norm": 1.1291477680206299,
+      "learning_rate": 4.311764926792201e-05,
+      "loss": 2.899,
+      "step": 11262
+    },
+    {
+      "epoch": 0.6928518700787402,
+      "grad_norm": 1.0770987272262573,
+      "learning_rate": 4.310174590534769e-05,
+      "loss": 2.267,
+      "step": 11263
+    },
+    {
+      "epoch": 0.6929133858267716,
+      "grad_norm": 1.0586423873901367,
+      "learning_rate": 4.30858446704508e-05,
+      "loss": 2.7654,
+      "step": 11264
+    },
+    {
+      "epoch": 0.6929749015748031,
+      "grad_norm": 1.2603076696395874,
+      "learning_rate": 4.3069945563826e-05,
+      "loss": 2.5789,
+      "step": 11265
+    },
+    {
+      "epoch": 0.6930364173228346,
+      "grad_norm": 1.4907711744308472,
+      "learning_rate": 4.305404858606773e-05,
+      "loss": 2.5028,
+      "step": 11266
+    },
+    {
+      "epoch": 0.6930979330708661,
+      "grad_norm": 1.4425654411315918,
+      "learning_rate": 4.303815373777056e-05,
+      "loss": 2.1801,
+      "step": 11267
+    },
+    {
+      "epoch": 0.6931594488188977,
+      "grad_norm": 1.5709338188171387,
+      "learning_rate": 4.302226101952877e-05,
+      "loss": 2.4568,
+      "step": 11268
+    },
+    {
+      "epoch": 0.6932209645669292,
+      "grad_norm": 1.2621530294418335,
+      "learning_rate": 4.300637043193669e-05,
+      "loss": 2.212,
+      "step": 11269
+    },
+    {
+      "epoch": 0.6932824803149606,
+      "grad_norm": 1.741391658782959,
+      "learning_rate": 4.2990481975588556e-05,
+      "loss": 2.2924,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6933439960629921,
+      "grad_norm": 1.1366198062896729,
+      "learning_rate": 4.2974595651078556e-05,
+      "loss": 2.4777,
+      "step": 11271
+    },
+    {
+      "epoch": 0.6934055118110236,
+      "grad_norm": 2.137523651123047,
+      "learning_rate": 4.295871145900063e-05,
+      "loss": 2.0066,
+      "step": 11272
+    },
+    {
+      "epoch": 0.6934670275590551,
+      "grad_norm": 1.1961666345596313,
+      "learning_rate": 4.2942829399948884e-05,
+      "loss": 2.5586,
+      "step": 11273
+    },
+    {
+      "epoch": 0.6935285433070866,
+      "grad_norm": 1.4898905754089355,
+      "learning_rate": 4.292694947451712e-05,
+      "loss": 2.0759,
+      "step": 11274
+    },
+    {
+      "epoch": 0.6935900590551181,
+      "grad_norm": 1.3798511028289795,
+      "learning_rate": 4.291107168329923e-05,
+      "loss": 2.8649,
+      "step": 11275
+    },
+    {
+      "epoch": 0.6936515748031497,
+      "grad_norm": 1.2069584131240845,
+      "learning_rate": 4.289519602688885e-05,
+      "loss": 2.7142,
+      "step": 11276
+    },
+    {
+      "epoch": 0.6937130905511811,
+      "grad_norm": 1.1763979196548462,
+      "learning_rate": 4.2879322505879784e-05,
+      "loss": 2.6848,
+      "step": 11277
+    },
+    {
+      "epoch": 0.6937746062992126,
+      "grad_norm": 1.2458277940750122,
+      "learning_rate": 4.28634511208655e-05,
+      "loss": 2.1969,
+      "step": 11278
+    },
+    {
+      "epoch": 0.6938361220472441,
+      "grad_norm": 1.2525115013122559,
+      "learning_rate": 4.284758187243959e-05,
+      "loss": 2.4207,
+      "step": 11279
+    },
+    {
+      "epoch": 0.6938976377952756,
+      "grad_norm": 1.2972807884216309,
+      "learning_rate": 4.2831714761195384e-05,
+      "loss": 3.1035,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6939591535433071,
+      "grad_norm": 1.2260913848876953,
+      "learning_rate": 4.2815849787726314e-05,
+      "loss": 2.5925,
+      "step": 11281
+    },
+    {
+      "epoch": 0.6940206692913385,
+      "grad_norm": 1.291373610496521,
+      "learning_rate": 4.279998695262555e-05,
+      "loss": 2.294,
+      "step": 11282
+    },
+    {
+      "epoch": 0.69408218503937,
+      "grad_norm": 1.6216243505477905,
+      "learning_rate": 4.278412625648631e-05,
+      "loss": 2.5094,
+      "step": 11283
+    },
+    {
+      "epoch": 0.6941437007874016,
+      "grad_norm": 1.2946150302886963,
+      "learning_rate": 4.276826769990172e-05,
+      "loss": 2.6568,
+      "step": 11284
+    },
+    {
+      "epoch": 0.6942052165354331,
+      "grad_norm": 1.18653404712677,
+      "learning_rate": 4.275241128346481e-05,
+      "loss": 2.4724,
+      "step": 11285
+    },
+    {
+      "epoch": 0.6942667322834646,
+      "grad_norm": 1.1980500221252441,
+      "learning_rate": 4.273655700776846e-05,
+      "loss": 2.083,
+      "step": 11286
+    },
+    {
+      "epoch": 0.6943282480314961,
+      "grad_norm": 1.4068262577056885,
+      "learning_rate": 4.272070487340561e-05,
+      "loss": 2.7943,
+      "step": 11287
+    },
+    {
+      "epoch": 0.6943897637795275,
+      "grad_norm": 1.3685897588729858,
+      "learning_rate": 4.2704854880968956e-05,
+      "loss": 2.9661,
+      "step": 11288
+    },
+    {
+      "epoch": 0.694451279527559,
+      "grad_norm": 1.2533668279647827,
+      "learning_rate": 4.2689007031051255e-05,
+      "loss": 2.5979,
+      "step": 11289
+    },
+    {
+      "epoch": 0.6945127952755905,
+      "grad_norm": 1.5037707090377808,
+      "learning_rate": 4.267316132424512e-05,
+      "loss": 2.2459,
+      "step": 11290
+    },
+    {
+      "epoch": 0.694574311023622,
+      "grad_norm": 1.43562912940979,
+      "learning_rate": 4.2657317761143104e-05,
+      "loss": 2.42,
+      "step": 11291
+    },
+    {
+      "epoch": 0.6946358267716536,
+      "grad_norm": 1.267535924911499,
+      "learning_rate": 4.2641476342337625e-05,
+      "loss": 2.3267,
+      "step": 11292
+    },
+    {
+      "epoch": 0.6946973425196851,
+      "grad_norm": 1.1653761863708496,
+      "learning_rate": 4.2625637068421135e-05,
+      "loss": 2.5529,
+      "step": 11293
+    },
+    {
+      "epoch": 0.6947588582677166,
+      "grad_norm": 1.3965281248092651,
+      "learning_rate": 4.260979993998587e-05,
+      "loss": 2.847,
+      "step": 11294
+    },
+    {
+      "epoch": 0.694820374015748,
+      "grad_norm": 1.1059683561325073,
+      "learning_rate": 4.259396495762402e-05,
+      "loss": 2.901,
+      "step": 11295
+    },
+    {
+      "epoch": 0.6948818897637795,
+      "grad_norm": 1.2100886106491089,
+      "learning_rate": 4.257813212192784e-05,
+      "loss": 2.8088,
+      "step": 11296
+    },
+    {
+      "epoch": 0.694943405511811,
+      "grad_norm": 1.1184515953063965,
+      "learning_rate": 4.256230143348929e-05,
+      "loss": 2.7662,
+      "step": 11297
+    },
+    {
+      "epoch": 0.6950049212598425,
+      "grad_norm": 1.2728956937789917,
+      "learning_rate": 4.2546472892900425e-05,
+      "loss": 2.777,
+      "step": 11298
+    },
+    {
+      "epoch": 0.695066437007874,
+      "grad_norm": 1.080295205116272,
+      "learning_rate": 4.253064650075306e-05,
+      "loss": 2.5544,
+      "step": 11299
+    },
+    {
+      "epoch": 0.6951279527559056,
+      "grad_norm": 1.3041399717330933,
+      "learning_rate": 4.25148222576391e-05,
+      "loss": 2.3113,
+      "step": 11300
+    },
+    {
+      "epoch": 0.695189468503937,
+      "grad_norm": 1.2580946683883667,
+      "learning_rate": 4.2499000164150195e-05,
+      "loss": 2.672,
+      "step": 11301
+    },
+    {
+      "epoch": 0.6952509842519685,
+      "grad_norm": 1.4405885934829712,
+      "learning_rate": 4.248318022087806e-05,
+      "loss": 2.1042,
+      "step": 11302
+    },
+    {
+      "epoch": 0.6953125,
+      "grad_norm": 1.6247988939285278,
+      "learning_rate": 4.246736242841426e-05,
+      "loss": 2.0112,
+      "step": 11303
+    },
+    {
+      "epoch": 0.6953740157480315,
+      "grad_norm": 1.1664005517959595,
+      "learning_rate": 4.2451546787350326e-05,
+      "loss": 3.0367,
+      "step": 11304
+    },
+    {
+      "epoch": 0.695435531496063,
+      "grad_norm": 1.2842973470687866,
+      "learning_rate": 4.243573329827761e-05,
+      "loss": 2.3401,
+      "step": 11305
+    },
+    {
+      "epoch": 0.6954970472440944,
+      "grad_norm": 1.6202166080474854,
+      "learning_rate": 4.241992196178752e-05,
+      "loss": 2.526,
+      "step": 11306
+    },
+    {
+      "epoch": 0.695558562992126,
+      "grad_norm": 1.2473046779632568,
+      "learning_rate": 4.2404112778471236e-05,
+      "loss": 1.9717,
+      "step": 11307
+    },
+    {
+      "epoch": 0.6956200787401575,
+      "grad_norm": 1.2551443576812744,
+      "learning_rate": 4.238830574891998e-05,
+      "loss": 2.4217,
+      "step": 11308
+    },
+    {
+      "epoch": 0.695681594488189,
+      "grad_norm": 1.281956672668457,
+      "learning_rate": 4.237250087372482e-05,
+      "loss": 3.0905,
+      "step": 11309
+    },
+    {
+      "epoch": 0.6957431102362205,
+      "grad_norm": 1.287843108177185,
+      "learning_rate": 4.2356698153476836e-05,
+      "loss": 2.5865,
+      "step": 11310
+    },
+    {
+      "epoch": 0.695804625984252,
+      "grad_norm": 1.3267438411712646,
+      "learning_rate": 4.234089758876687e-05,
+      "loss": 2.3473,
+      "step": 11311
+    },
+    {
+      "epoch": 0.6958661417322834,
+      "grad_norm": 1.428458571434021,
+      "learning_rate": 4.232509918018586e-05,
+      "loss": 2.5675,
+      "step": 11312
+    },
+    {
+      "epoch": 0.6959276574803149,
+      "grad_norm": 1.2809300422668457,
+      "learning_rate": 4.230930292832449e-05,
+      "loss": 2.2738,
+      "step": 11313
+    },
+    {
+      "epoch": 0.6959891732283464,
+      "grad_norm": 1.4004228115081787,
+      "learning_rate": 4.2293508833773506e-05,
+      "loss": 2.5118,
+      "step": 11314
+    },
+    {
+      "epoch": 0.696050688976378,
+      "grad_norm": 1.3359259366989136,
+      "learning_rate": 4.227771689712352e-05,
+      "loss": 2.102,
+      "step": 11315
+    },
+    {
+      "epoch": 0.6961122047244095,
+      "grad_norm": 1.1879509687423706,
+      "learning_rate": 4.226192711896507e-05,
+      "loss": 2.4804,
+      "step": 11316
+    },
+    {
+      "epoch": 0.696173720472441,
+      "grad_norm": 1.268984079360962,
+      "learning_rate": 4.224613949988857e-05,
+      "loss": 2.0874,
+      "step": 11317
+    },
+    {
+      "epoch": 0.6962352362204725,
+      "grad_norm": 1.1617166996002197,
+      "learning_rate": 4.2230354040484443e-05,
+      "loss": 2.5276,
+      "step": 11318
+    },
+    {
+      "epoch": 0.6962967519685039,
+      "grad_norm": 1.2116093635559082,
+      "learning_rate": 4.22145707413429e-05,
+      "loss": 2.6212,
+      "step": 11319
+    },
+    {
+      "epoch": 0.6963582677165354,
+      "grad_norm": 1.034482717514038,
+      "learning_rate": 4.219878960305423e-05,
+      "loss": 2.6064,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6964197834645669,
+      "grad_norm": 1.2718614339828491,
+      "learning_rate": 4.218301062620844e-05,
+      "loss": 2.4462,
+      "step": 11321
+    },
+    {
+      "epoch": 0.6964812992125984,
+      "grad_norm": 1.4299131631851196,
+      "learning_rate": 4.2167233811395746e-05,
+      "loss": 2.1502,
+      "step": 11322
+    },
+    {
+      "epoch": 0.69654281496063,
+      "grad_norm": 1.2111974954605103,
+      "learning_rate": 4.2151459159205974e-05,
+      "loss": 2.5932,
+      "step": 11323
+    },
+    {
+      "epoch": 0.6966043307086615,
+      "grad_norm": 1.3308709859848022,
+      "learning_rate": 4.213568667022909e-05,
+      "loss": 3.0284,
+      "step": 11324
+    },
+    {
+      "epoch": 0.6966658464566929,
+      "grad_norm": 1.1488608121871948,
+      "learning_rate": 4.211991634505484e-05,
+      "loss": 2.424,
+      "step": 11325
+    },
+    {
+      "epoch": 0.6967273622047244,
+      "grad_norm": 1.1220159530639648,
+      "learning_rate": 4.2104148184272994e-05,
+      "loss": 2.0525,
+      "step": 11326
+    },
+    {
+      "epoch": 0.6967888779527559,
+      "grad_norm": 1.1743003129959106,
+      "learning_rate": 4.208838218847315e-05,
+      "loss": 2.4778,
+      "step": 11327
+    },
+    {
+      "epoch": 0.6968503937007874,
+      "grad_norm": 1.1173937320709229,
+      "learning_rate": 4.207261835824488e-05,
+      "loss": 3.0891,
+      "step": 11328
+    },
+    {
+      "epoch": 0.6969119094488189,
+      "grad_norm": 1.4518802165985107,
+      "learning_rate": 4.205685669417767e-05,
+      "loss": 2.4141,
+      "step": 11329
+    },
+    {
+      "epoch": 0.6969734251968503,
+      "grad_norm": 1.2542383670806885,
+      "learning_rate": 4.204109719686096e-05,
+      "loss": 2.5012,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6970349409448819,
+      "grad_norm": 1.3453904390335083,
+      "learning_rate": 4.202533986688402e-05,
+      "loss": 2.5236,
+      "step": 11331
+    },
+    {
+      "epoch": 0.6970964566929134,
+      "grad_norm": 1.7769324779510498,
+      "learning_rate": 4.200958470483606e-05,
+      "loss": 1.7593,
+      "step": 11332
+    },
+    {
+      "epoch": 0.6971579724409449,
+      "grad_norm": 1.182513952255249,
+      "learning_rate": 4.199383171130627e-05,
+      "loss": 2.2832,
+      "step": 11333
+    },
+    {
+      "epoch": 0.6972194881889764,
+      "grad_norm": 1.0525118112564087,
+      "learning_rate": 4.197808088688372e-05,
+      "loss": 2.9694,
+      "step": 11334
+    },
+    {
+      "epoch": 0.6972810039370079,
+      "grad_norm": 1.2044512033462524,
+      "learning_rate": 4.1962332232157444e-05,
+      "loss": 2.5647,
+      "step": 11335
+    },
+    {
+      "epoch": 0.6973425196850394,
+      "grad_norm": 1.3873416185379028,
+      "learning_rate": 4.194658574771627e-05,
+      "loss": 2.5895,
+      "step": 11336
+    },
+    {
+      "epoch": 0.6974040354330708,
+      "grad_norm": 1.4256078004837036,
+      "learning_rate": 4.193084143414912e-05,
+      "loss": 2.3315,
+      "step": 11337
+    },
+    {
+      "epoch": 0.6974655511811023,
+      "grad_norm": 1.1519702672958374,
+      "learning_rate": 4.1915099292044667e-05,
+      "loss": 2.2888,
+      "step": 11338
+    },
+    {
+      "epoch": 0.6975270669291339,
+      "grad_norm": 1.3596478700637817,
+      "learning_rate": 4.189935932199161e-05,
+      "loss": 2.4211,
+      "step": 11339
+    },
+    {
+      "epoch": 0.6975885826771654,
+      "grad_norm": 1.2986646890640259,
+      "learning_rate": 4.188362152457853e-05,
+      "loss": 2.8833,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6976500984251969,
+      "grad_norm": 1.3820533752441406,
+      "learning_rate": 4.1867885900393975e-05,
+      "loss": 2.4884,
+      "step": 11341
+    },
+    {
+      "epoch": 0.6977116141732284,
+      "grad_norm": 1.1238341331481934,
+      "learning_rate": 4.1852152450026295e-05,
+      "loss": 1.7272,
+      "step": 11342
+    },
+    {
+      "epoch": 0.6977731299212598,
+      "grad_norm": 1.4369009733200073,
+      "learning_rate": 4.183642117406391e-05,
+      "loss": 2.348,
+      "step": 11343
+    },
+    {
+      "epoch": 0.6978346456692913,
+      "grad_norm": 1.219205379486084,
+      "learning_rate": 4.1820692073095016e-05,
+      "loss": 2.7121,
+      "step": 11344
+    },
+    {
+      "epoch": 0.6978961614173228,
+      "grad_norm": 1.5061286687850952,
+      "learning_rate": 4.180496514770785e-05,
+      "loss": 2.2125,
+      "step": 11345
+    },
+    {
+      "epoch": 0.6979576771653543,
+      "grad_norm": 1.3512591123580933,
+      "learning_rate": 4.178924039849045e-05,
+      "loss": 3.0644,
+      "step": 11346
+    },
+    {
+      "epoch": 0.6980191929133859,
+      "grad_norm": 1.6097140312194824,
+      "learning_rate": 4.177351782603086e-05,
+      "loss": 2.5984,
+      "step": 11347
+    },
+    {
+      "epoch": 0.6980807086614174,
+      "grad_norm": 1.3334541320800781,
+      "learning_rate": 4.1757797430917035e-05,
+      "loss": 1.6931,
+      "step": 11348
+    },
+    {
+      "epoch": 0.6981422244094488,
+      "grad_norm": 1.2941229343414307,
+      "learning_rate": 4.174207921373685e-05,
+      "loss": 2.5406,
+      "step": 11349
+    },
+    {
+      "epoch": 0.6982037401574803,
+      "grad_norm": 1.2301042079925537,
+      "learning_rate": 4.172636317507801e-05,
+      "loss": 2.6227,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6982652559055118,
+      "grad_norm": 1.125053882598877,
+      "learning_rate": 4.1710649315528285e-05,
+      "loss": 2.9163,
+      "step": 11351
+    },
+    {
+      "epoch": 0.6983267716535433,
+      "grad_norm": 1.1280810832977295,
+      "learning_rate": 4.16949376356752e-05,
+      "loss": 2.463,
+      "step": 11352
+    },
+    {
+      "epoch": 0.6983882874015748,
+      "grad_norm": 1.3273347616195679,
+      "learning_rate": 4.167922813610633e-05,
+      "loss": 2.2463,
+      "step": 11353
+    },
+    {
+      "epoch": 0.6984498031496063,
+      "grad_norm": 1.1092815399169922,
+      "learning_rate": 4.166352081740913e-05,
+      "loss": 2.8036,
+      "step": 11354
+    },
+    {
+      "epoch": 0.6985113188976378,
+      "grad_norm": 1.4338841438293457,
+      "learning_rate": 4.1647815680170986e-05,
+      "loss": 2.0129,
+      "step": 11355
+    },
+    {
+      "epoch": 0.6985728346456693,
+      "grad_norm": 1.1723943948745728,
+      "learning_rate": 4.1632112724979135e-05,
+      "loss": 2.3507,
+      "step": 11356
+    },
+    {
+      "epoch": 0.6986343503937008,
+      "grad_norm": 1.340692162513733,
+      "learning_rate": 4.161641195242082e-05,
+      "loss": 2.7112,
+      "step": 11357
+    },
+    {
+      "epoch": 0.6986958661417323,
+      "grad_norm": 1.3262096643447876,
+      "learning_rate": 4.160071336308311e-05,
+      "loss": 2.311,
+      "step": 11358
+    },
+    {
+      "epoch": 0.6987573818897638,
+      "grad_norm": 1.0990618467330933,
+      "learning_rate": 4.158501695755308e-05,
+      "loss": 2.7852,
+      "step": 11359
+    },
+    {
+      "epoch": 0.6988188976377953,
+      "grad_norm": 1.2954801321029663,
+      "learning_rate": 4.1569322736417684e-05,
+      "loss": 2.0731,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6988804133858267,
+      "grad_norm": 1.3475290536880493,
+      "learning_rate": 4.155363070026383e-05,
+      "loss": 2.4178,
+      "step": 11361
+    },
+    {
+      "epoch": 0.6989419291338582,
+      "grad_norm": 1.20216703414917,
+      "learning_rate": 4.153794084967825e-05,
+      "loss": 2.5934,
+      "step": 11362
+    },
+    {
+      "epoch": 0.6990034448818898,
+      "grad_norm": 1.1665531396865845,
+      "learning_rate": 4.152225318524773e-05,
+      "loss": 2.7497,
+      "step": 11363
+    },
+    {
+      "epoch": 0.6990649606299213,
+      "grad_norm": 1.2056910991668701,
+      "learning_rate": 4.150656770755882e-05,
+      "loss": 2.7518,
+      "step": 11364
+    },
+    {
+      "epoch": 0.6991264763779528,
+      "grad_norm": 1.3660712242126465,
+      "learning_rate": 4.1490884417198105e-05,
+      "loss": 2.3414,
+      "step": 11365
+    },
+    {
+      "epoch": 0.6991879921259843,
+      "grad_norm": 1.1794114112854004,
+      "learning_rate": 4.1475203314752057e-05,
+      "loss": 2.4433,
+      "step": 11366
+    },
+    {
+      "epoch": 0.6992495078740157,
+      "grad_norm": 1.2566263675689697,
+      "learning_rate": 4.1459524400807104e-05,
+      "loss": 2.5646,
+      "step": 11367
+    },
+    {
+      "epoch": 0.6993110236220472,
+      "grad_norm": 1.3603278398513794,
+      "learning_rate": 4.1443847675949497e-05,
+      "loss": 2.1058,
+      "step": 11368
+    },
+    {
+      "epoch": 0.6993725393700787,
+      "grad_norm": 1.3335293531417847,
+      "learning_rate": 4.142817314076545e-05,
+      "loss": 2.2742,
+      "step": 11369
+    },
+    {
+      "epoch": 0.6994340551181102,
+      "grad_norm": 1.2813295125961304,
+      "learning_rate": 4.1412500795841146e-05,
+      "loss": 2.6837,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6994955708661418,
+      "grad_norm": 1.556107521057129,
+      "learning_rate": 4.139683064176259e-05,
+      "loss": 2.1957,
+      "step": 11371
+    },
+    {
+      "epoch": 0.6995570866141733,
+      "grad_norm": 1.1190979480743408,
+      "learning_rate": 4.138116267911578e-05,
+      "loss": 3.0877,
+      "step": 11372
+    },
+    {
+      "epoch": 0.6996186023622047,
+      "grad_norm": 1.2065739631652832,
+      "learning_rate": 4.136549690848662e-05,
+      "loss": 2.3135,
+      "step": 11373
+    },
+    {
+      "epoch": 0.6996801181102362,
+      "grad_norm": 1.3256093263626099,
+      "learning_rate": 4.1349833330460965e-05,
+      "loss": 2.5544,
+      "step": 11374
+    },
+    {
+      "epoch": 0.6997416338582677,
+      "grad_norm": 1.4891879558563232,
+      "learning_rate": 4.1334171945624454e-05,
+      "loss": 2.2266,
+      "step": 11375
+    },
+    {
+      "epoch": 0.6998031496062992,
+      "grad_norm": 1.105311632156372,
+      "learning_rate": 4.131851275456282e-05,
+      "loss": 2.7877,
+      "step": 11376
+    },
+    {
+      "epoch": 0.6998646653543307,
+      "grad_norm": 1.7288824319839478,
+      "learning_rate": 4.130285575786156e-05,
+      "loss": 2.0237,
+      "step": 11377
+    },
+    {
+      "epoch": 0.6999261811023622,
+      "grad_norm": 1.241281509399414,
+      "learning_rate": 4.12872009561062e-05,
+      "loss": 2.8276,
+      "step": 11378
+    },
+    {
+      "epoch": 0.6999876968503937,
+      "grad_norm": 1.6037333011627197,
+      "learning_rate": 4.1271548349882115e-05,
+      "loss": 2.196,
+      "step": 11379
+    },
+    {
+      "epoch": 0.7000492125984252,
+      "grad_norm": 1.198914885520935,
+      "learning_rate": 4.125589793977469e-05,
+      "loss": 2.5491,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7001107283464567,
+      "grad_norm": 1.5560712814331055,
+      "learning_rate": 4.124024972636907e-05,
+      "loss": 2.3501,
+      "step": 11381
+    },
+    {
+      "epoch": 0.7001722440944882,
+      "grad_norm": 1.45331609249115,
+      "learning_rate": 4.1224603710250496e-05,
+      "loss": 2.3128,
+      "step": 11382
+    },
+    {
+      "epoch": 0.7002337598425197,
+      "grad_norm": 1.4529458284378052,
+      "learning_rate": 4.120895989200397e-05,
+      "loss": 2.7736,
+      "step": 11383
+    },
+    {
+      "epoch": 0.7002952755905512,
+      "grad_norm": 1.371948480606079,
+      "learning_rate": 4.119331827221452e-05,
+      "loss": 2.3866,
+      "step": 11384
+    },
+    {
+      "epoch": 0.7003567913385826,
+      "grad_norm": 1.3687083721160889,
+      "learning_rate": 4.1177678851467046e-05,
+      "loss": 2.4403,
+      "step": 11385
+    },
+    {
+      "epoch": 0.7004183070866141,
+      "grad_norm": 1.55045747756958,
+      "learning_rate": 4.1162041630346424e-05,
+      "loss": 2.8669,
+      "step": 11386
+    },
+    {
+      "epoch": 0.7004798228346457,
+      "grad_norm": 1.878119945526123,
+      "learning_rate": 4.114640660943732e-05,
+      "loss": 2.0248,
+      "step": 11387
+    },
+    {
+      "epoch": 0.7005413385826772,
+      "grad_norm": 1.1364487409591675,
+      "learning_rate": 4.113077378932445e-05,
+      "loss": 3.0828,
+      "step": 11388
+    },
+    {
+      "epoch": 0.7006028543307087,
+      "grad_norm": 1.4168285131454468,
+      "learning_rate": 4.111514317059236e-05,
+      "loss": 1.4931,
+      "step": 11389
+    },
+    {
+      "epoch": 0.7006643700787402,
+      "grad_norm": 1.4241546392440796,
+      "learning_rate": 4.10995147538256e-05,
+      "loss": 2.1102,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7007258858267716,
+      "grad_norm": 1.2659838199615479,
+      "learning_rate": 4.108388853960846e-05,
+      "loss": 2.7532,
+      "step": 11391
+    },
+    {
+      "epoch": 0.7007874015748031,
+      "grad_norm": 1.3436626195907593,
+      "learning_rate": 4.106826452852545e-05,
+      "loss": 2.2682,
+      "step": 11392
+    },
+    {
+      "epoch": 0.7008489173228346,
+      "grad_norm": 1.3055418729782104,
+      "learning_rate": 4.1052642721160696e-05,
+      "loss": 2.6248,
+      "step": 11393
+    },
+    {
+      "epoch": 0.7009104330708661,
+      "grad_norm": 1.278188705444336,
+      "learning_rate": 4.103702311809844e-05,
+      "loss": 2.3359,
+      "step": 11394
+    },
+    {
+      "epoch": 0.7009719488188977,
+      "grad_norm": 1.3455841541290283,
+      "learning_rate": 4.102140571992269e-05,
+      "loss": 2.3769,
+      "step": 11395
+    },
+    {
+      "epoch": 0.7010334645669292,
+      "grad_norm": 1.2654314041137695,
+      "learning_rate": 4.100579052721754e-05,
+      "loss": 2.822,
+      "step": 11396
+    },
+    {
+      "epoch": 0.7010949803149606,
+      "grad_norm": 1.2344489097595215,
+      "learning_rate": 4.099017754056682e-05,
+      "loss": 2.1705,
+      "step": 11397
+    },
+    {
+      "epoch": 0.7011564960629921,
+      "grad_norm": 1.3126988410949707,
+      "learning_rate": 4.097456676055441e-05,
+      "loss": 2.8369,
+      "step": 11398
+    },
+    {
+      "epoch": 0.7012180118110236,
+      "grad_norm": 1.6614022254943848,
+      "learning_rate": 4.095895818776409e-05,
+      "loss": 2.5388,
+      "step": 11399
+    },
+    {
+      "epoch": 0.7012795275590551,
+      "grad_norm": 1.1389780044555664,
+      "learning_rate": 4.094335182277952e-05,
+      "loss": 2.2518,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7013410433070866,
+      "grad_norm": 1.3145006895065308,
+      "learning_rate": 4.092774766618426e-05,
+      "loss": 3.1604,
+      "step": 11401
+    },
+    {
+      "epoch": 0.7014025590551181,
+      "grad_norm": 1.2309099435806274,
+      "learning_rate": 4.0912145718561875e-05,
+      "loss": 2.8744,
+      "step": 11402
+    },
+    {
+      "epoch": 0.7014640748031497,
+      "grad_norm": 1.1506516933441162,
+      "learning_rate": 4.0896545980495714e-05,
+      "loss": 2.7672,
+      "step": 11403
+    },
+    {
+      "epoch": 0.7015255905511811,
+      "grad_norm": 1.2609691619873047,
+      "learning_rate": 4.088094845256918e-05,
+      "loss": 2.7793,
+      "step": 11404
+    },
+    {
+      "epoch": 0.7015871062992126,
+      "grad_norm": 1.2344261407852173,
+      "learning_rate": 4.0865353135365536e-05,
+      "loss": 2.8316,
+      "step": 11405
+    },
+    {
+      "epoch": 0.7016486220472441,
+      "grad_norm": 1.326507806777954,
+      "learning_rate": 4.084976002946792e-05,
+      "loss": 2.7164,
+      "step": 11406
+    },
+    {
+      "epoch": 0.7017101377952756,
+      "grad_norm": 1.3567970991134644,
+      "learning_rate": 4.083416913545948e-05,
+      "loss": 2.448,
+      "step": 11407
+    },
+    {
+      "epoch": 0.7017716535433071,
+      "grad_norm": 1.6553720235824585,
+      "learning_rate": 4.081858045392315e-05,
+      "loss": 2.2488,
+      "step": 11408
+    },
+    {
+      "epoch": 0.7018331692913385,
+      "grad_norm": 1.2343477010726929,
+      "learning_rate": 4.080299398544195e-05,
+      "loss": 1.2398,
+      "step": 11409
+    },
+    {
+      "epoch": 0.70189468503937,
+      "grad_norm": 1.1423189640045166,
+      "learning_rate": 4.078740973059861e-05,
+      "loss": 1.8958,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7019562007874016,
+      "grad_norm": 1.6313785314559937,
+      "learning_rate": 4.077182768997604e-05,
+      "loss": 2.6414,
+      "step": 11411
+    },
+    {
+      "epoch": 0.7020177165354331,
+      "grad_norm": 1.3787307739257812,
+      "learning_rate": 4.075624786415681e-05,
+      "loss": 2.3021,
+      "step": 11412
+    },
+    {
+      "epoch": 0.7020792322834646,
+      "grad_norm": 1.4699547290802002,
+      "learning_rate": 4.07406702537236e-05,
+      "loss": 2.9635,
+      "step": 11413
+    },
+    {
+      "epoch": 0.7021407480314961,
+      "grad_norm": 1.1505178213119507,
+      "learning_rate": 4.072509485925885e-05,
+      "loss": 2.7939,
+      "step": 11414
+    },
+    {
+      "epoch": 0.7022022637795275,
+      "grad_norm": 1.1895263195037842,
+      "learning_rate": 4.0709521681345054e-05,
+      "loss": 2.0935,
+      "step": 11415
+    },
+    {
+      "epoch": 0.702263779527559,
+      "grad_norm": 1.2155646085739136,
+      "learning_rate": 4.06939507205645e-05,
+      "loss": 2.5607,
+      "step": 11416
+    },
+    {
+      "epoch": 0.7023252952755905,
+      "grad_norm": 1.1435967683792114,
+      "learning_rate": 4.067838197749949e-05,
+      "loss": 2.5521,
+      "step": 11417
+    },
+    {
+      "epoch": 0.702386811023622,
+      "grad_norm": 1.2211400270462036,
+      "learning_rate": 4.066281545273222e-05,
+      "loss": 2.4322,
+      "step": 11418
+    },
+    {
+      "epoch": 0.7024483267716536,
+      "grad_norm": 1.299149751663208,
+      "learning_rate": 4.06472511468448e-05,
+      "loss": 2.4128,
+      "step": 11419
+    },
+    {
+      "epoch": 0.7025098425196851,
+      "grad_norm": 1.5158054828643799,
+      "learning_rate": 4.063168906041919e-05,
+      "loss": 2.1959,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7025713582677166,
+      "grad_norm": 1.2828751802444458,
+      "learning_rate": 4.06161291940374e-05,
+      "loss": 2.7831,
+      "step": 11421
+    },
+    {
+      "epoch": 0.702632874015748,
+      "grad_norm": 1.2753183841705322,
+      "learning_rate": 4.060057154828122e-05,
+      "loss": 2.1375,
+      "step": 11422
+    },
+    {
+      "epoch": 0.7026943897637795,
+      "grad_norm": 1.5692718029022217,
+      "learning_rate": 4.0585016123732434e-05,
+      "loss": 2.2446,
+      "step": 11423
+    },
+    {
+      "epoch": 0.702755905511811,
+      "grad_norm": 1.3382245302200317,
+      "learning_rate": 4.056946292097273e-05,
+      "loss": 2.2406,
+      "step": 11424
+    },
+    {
+      "epoch": 0.7028174212598425,
+      "grad_norm": 1.2420227527618408,
+      "learning_rate": 4.055391194058377e-05,
+      "loss": 2.9222,
+      "step": 11425
+    },
+    {
+      "epoch": 0.702878937007874,
+      "grad_norm": 1.3774936199188232,
+      "learning_rate": 4.0538363183146985e-05,
+      "loss": 2.2459,
+      "step": 11426
+    },
+    {
+      "epoch": 0.7029404527559056,
+      "grad_norm": 1.1329669952392578,
+      "learning_rate": 4.052281664924388e-05,
+      "loss": 2.9131,
+      "step": 11427
+    },
+    {
+      "epoch": 0.703001968503937,
+      "grad_norm": 1.4601072072982788,
+      "learning_rate": 4.0507272339455746e-05,
+      "loss": 2.3294,
+      "step": 11428
+    },
+    {
+      "epoch": 0.7030634842519685,
+      "grad_norm": 1.3724474906921387,
+      "learning_rate": 4.049173025436388e-05,
+      "loss": 2.415,
+      "step": 11429
+    },
+    {
+      "epoch": 0.703125,
+      "grad_norm": 1.9650874137878418,
+      "learning_rate": 4.047619039454948e-05,
+      "loss": 2.1964,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7031865157480315,
+      "grad_norm": 1.353798508644104,
+      "learning_rate": 4.046065276059369e-05,
+      "loss": 2.0337,
+      "step": 11431
+    },
+    {
+      "epoch": 0.703248031496063,
+      "grad_norm": 1.2807143926620483,
+      "learning_rate": 4.0445117353077434e-05,
+      "loss": 2.8537,
+      "step": 11432
+    },
+    {
+      "epoch": 0.7033095472440944,
+      "grad_norm": 1.302891492843628,
+      "learning_rate": 4.042958417258175e-05,
+      "loss": 2.8474,
+      "step": 11433
+    },
+    {
+      "epoch": 0.703371062992126,
+      "grad_norm": 1.3808226585388184,
+      "learning_rate": 4.0414053219687395e-05,
+      "loss": 2.3576,
+      "step": 11434
+    },
+    {
+      "epoch": 0.7034325787401575,
+      "grad_norm": 1.130129337310791,
+      "learning_rate": 4.0398524494975245e-05,
+      "loss": 2.5071,
+      "step": 11435
+    },
+    {
+      "epoch": 0.703494094488189,
+      "grad_norm": 1.393938660621643,
+      "learning_rate": 4.038299799902584e-05,
+      "loss": 2.2434,
+      "step": 11436
+    },
+    {
+      "epoch": 0.7035556102362205,
+      "grad_norm": 1.1713905334472656,
+      "learning_rate": 4.036747373241996e-05,
+      "loss": 2.9271,
+      "step": 11437
+    },
+    {
+      "epoch": 0.703617125984252,
+      "grad_norm": 1.3187675476074219,
+      "learning_rate": 4.0351951695738e-05,
+      "loss": 2.4692,
+      "step": 11438
+    },
+    {
+      "epoch": 0.7036786417322834,
+      "grad_norm": 1.3271307945251465,
+      "learning_rate": 4.033643188956048e-05,
+      "loss": 2.629,
+      "step": 11439
+    },
+    {
+      "epoch": 0.7037401574803149,
+      "grad_norm": 1.1393064260482788,
+      "learning_rate": 4.0320914314467684e-05,
+      "loss": 2.5882,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7038016732283464,
+      "grad_norm": 1.423404335975647,
+      "learning_rate": 4.030539897103994e-05,
+      "loss": 1.7973,
+      "step": 11441
+    },
+    {
+      "epoch": 0.703863188976378,
+      "grad_norm": 1.145493507385254,
+      "learning_rate": 4.0289885859857366e-05,
+      "loss": 2.4988,
+      "step": 11442
+    },
+    {
+      "epoch": 0.7039247047244095,
+      "grad_norm": 1.1628642082214355,
+      "learning_rate": 4.0274374981500116e-05,
+      "loss": 2.526,
+      "step": 11443
+    },
+    {
+      "epoch": 0.703986220472441,
+      "grad_norm": 1.394178867340088,
+      "learning_rate": 4.025886633654825e-05,
+      "loss": 1.9826,
+      "step": 11444
+    },
+    {
+      "epoch": 0.7040477362204725,
+      "grad_norm": 1.9058444499969482,
+      "learning_rate": 4.02433599255816e-05,
+      "loss": 1.8918,
+      "step": 11445
+    },
+    {
+      "epoch": 0.7041092519685039,
+      "grad_norm": 1.2976551055908203,
+      "learning_rate": 4.0227855749180135e-05,
+      "loss": 2.4059,
+      "step": 11446
+    },
+    {
+      "epoch": 0.7041707677165354,
+      "grad_norm": 1.1373690366744995,
+      "learning_rate": 4.021235380792351e-05,
+      "loss": 2.7878,
+      "step": 11447
+    },
+    {
+      "epoch": 0.7042322834645669,
+      "grad_norm": 1.1708335876464844,
+      "learning_rate": 4.019685410239148e-05,
+      "loss": 1.9883,
+      "step": 11448
+    },
+    {
+      "epoch": 0.7042937992125984,
+      "grad_norm": 1.2949050664901733,
+      "learning_rate": 4.018135663316363e-05,
+      "loss": 1.5941,
+      "step": 11449
+    },
+    {
+      "epoch": 0.70435531496063,
+      "grad_norm": 1.4170546531677246,
+      "learning_rate": 4.0165861400819515e-05,
+      "loss": 2.179,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7044168307086615,
+      "grad_norm": 1.3050153255462646,
+      "learning_rate": 4.015036840593851e-05,
+      "loss": 2.4807,
+      "step": 11451
+    },
+    {
+      "epoch": 0.7044783464566929,
+      "grad_norm": 1.1319853067398071,
+      "learning_rate": 4.0134877649100034e-05,
+      "loss": 2.4474,
+      "step": 11452
+    },
+    {
+      "epoch": 0.7045398622047244,
+      "grad_norm": 1.2002419233322144,
+      "learning_rate": 4.011938913088328e-05,
+      "loss": 2.5646,
+      "step": 11453
+    },
+    {
+      "epoch": 0.7046013779527559,
+      "grad_norm": 1.2314529418945312,
+      "learning_rate": 4.010390285186747e-05,
+      "loss": 2.7794,
+      "step": 11454
+    },
+    {
+      "epoch": 0.7046628937007874,
+      "grad_norm": 1.4415767192840576,
+      "learning_rate": 4.0088418812631703e-05,
+      "loss": 2.9459,
+      "step": 11455
+    },
+    {
+      "epoch": 0.7047244094488189,
+      "grad_norm": 1.1256098747253418,
+      "learning_rate": 4.007293701375503e-05,
+      "loss": 2.8995,
+      "step": 11456
+    },
+    {
+      "epoch": 0.7047859251968503,
+      "grad_norm": 1.199942946434021,
+      "learning_rate": 4.00574574558163e-05,
+      "loss": 2.7676,
+      "step": 11457
+    },
+    {
+      "epoch": 0.7048474409448819,
+      "grad_norm": 1.5720136165618896,
+      "learning_rate": 4.0041980139394477e-05,
+      "loss": 2.2878,
+      "step": 11458
+    },
+    {
+      "epoch": 0.7049089566929134,
+      "grad_norm": 1.3074284791946411,
+      "learning_rate": 4.002650506506821e-05,
+      "loss": 2.5381,
+      "step": 11459
+    },
+    {
+      "epoch": 0.7049704724409449,
+      "grad_norm": 1.3835163116455078,
+      "learning_rate": 4.001103223341627e-05,
+      "loss": 2.7158,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7050319881889764,
+      "grad_norm": 1.1115061044692993,
+      "learning_rate": 3.9995561645017175e-05,
+      "loss": 2.9416,
+      "step": 11461
+    },
+    {
+      "epoch": 0.7050935039370079,
+      "grad_norm": 1.1659679412841797,
+      "learning_rate": 3.9980093300449486e-05,
+      "loss": 2.6452,
+      "step": 11462
+    },
+    {
+      "epoch": 0.7051550196850394,
+      "grad_norm": 1.2101013660430908,
+      "learning_rate": 3.996462720029163e-05,
+      "loss": 2.5789,
+      "step": 11463
+    },
+    {
+      "epoch": 0.7052165354330708,
+      "grad_norm": 0.9922416806221008,
+      "learning_rate": 3.994916334512199e-05,
+      "loss": 1.3861,
+      "step": 11464
+    },
+    {
+      "epoch": 0.7052780511811023,
+      "grad_norm": 1.1939550638198853,
+      "learning_rate": 3.993370173551874e-05,
+      "loss": 1.5875,
+      "step": 11465
+    },
+    {
+      "epoch": 0.7053395669291339,
+      "grad_norm": 1.3633400201797485,
+      "learning_rate": 3.9918242372060145e-05,
+      "loss": 2.6146,
+      "step": 11466
+    },
+    {
+      "epoch": 0.7054010826771654,
+      "grad_norm": 1.1926758289337158,
+      "learning_rate": 3.9902785255324224e-05,
+      "loss": 2.4347,
+      "step": 11467
+    },
+    {
+      "epoch": 0.7054625984251969,
+      "grad_norm": 1.1585379838943481,
+      "learning_rate": 3.988733038588903e-05,
+      "loss": 2.9858,
+      "step": 11468
+    },
+    {
+      "epoch": 0.7055241141732284,
+      "grad_norm": 1.5856844186782837,
+      "learning_rate": 3.9871877764332476e-05,
+      "loss": 2.6914,
+      "step": 11469
+    },
+    {
+      "epoch": 0.7055856299212598,
+      "grad_norm": 1.3630695343017578,
+      "learning_rate": 3.985642739123245e-05,
+      "loss": 2.8419,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7056471456692913,
+      "grad_norm": 1.3207831382751465,
+      "learning_rate": 3.9840979267166636e-05,
+      "loss": 2.5923,
+      "step": 11471
+    },
+    {
+      "epoch": 0.7057086614173228,
+      "grad_norm": 1.288610577583313,
+      "learning_rate": 3.982553339271277e-05,
+      "loss": 2.5203,
+      "step": 11472
+    },
+    {
+      "epoch": 0.7057701771653543,
+      "grad_norm": 1.318735957145691,
+      "learning_rate": 3.981008976844838e-05,
+      "loss": 2.5132,
+      "step": 11473
+    },
+    {
+      "epoch": 0.7058316929133859,
+      "grad_norm": 1.2245995998382568,
+      "learning_rate": 3.9794648394951004e-05,
+      "loss": 2.8984,
+      "step": 11474
+    },
+    {
+      "epoch": 0.7058932086614174,
+      "grad_norm": 1.3665931224822998,
+      "learning_rate": 3.9779209272798065e-05,
+      "loss": 2.5278,
+      "step": 11475
+    },
+    {
+      "epoch": 0.7059547244094488,
+      "grad_norm": 1.256838083267212,
+      "learning_rate": 3.9763772402566934e-05,
+      "loss": 2.2492,
+      "step": 11476
+    },
+    {
+      "epoch": 0.7060162401574803,
+      "grad_norm": 1.2207403182983398,
+      "learning_rate": 3.97483377848348e-05,
+      "loss": 2.497,
+      "step": 11477
+    },
+    {
+      "epoch": 0.7060777559055118,
+      "grad_norm": 1.4652100801467896,
+      "learning_rate": 3.973290542017889e-05,
+      "loss": 2.6263,
+      "step": 11478
+    },
+    {
+      "epoch": 0.7061392716535433,
+      "grad_norm": 1.6983269453048706,
+      "learning_rate": 3.9717475309176254e-05,
+      "loss": 2.0968,
+      "step": 11479
+    },
+    {
+      "epoch": 0.7062007874015748,
+      "grad_norm": 1.603281021118164,
+      "learning_rate": 3.970204745240384e-05,
+      "loss": 3.8601,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7062623031496063,
+      "grad_norm": 1.1717936992645264,
+      "learning_rate": 3.9686621850438675e-05,
+      "loss": 3.0707,
+      "step": 11481
+    },
+    {
+      "epoch": 0.7063238188976378,
+      "grad_norm": 1.493999719619751,
+      "learning_rate": 3.967119850385751e-05,
+      "loss": 2.3128,
+      "step": 11482
+    },
+    {
+      "epoch": 0.7063853346456693,
+      "grad_norm": 1.2631914615631104,
+      "learning_rate": 3.965577741323715e-05,
+      "loss": 1.2283,
+      "step": 11483
+    },
+    {
+      "epoch": 0.7064468503937008,
+      "grad_norm": 1.2738841772079468,
+      "learning_rate": 3.964035857915419e-05,
+      "loss": 1.9321,
+      "step": 11484
+    },
+    {
+      "epoch": 0.7065083661417323,
+      "grad_norm": 1.2241605520248413,
+      "learning_rate": 3.9624942002185286e-05,
+      "loss": 1.9433,
+      "step": 11485
+    },
+    {
+      "epoch": 0.7065698818897638,
+      "grad_norm": 1.2088782787322998,
+      "learning_rate": 3.960952768290686e-05,
+      "loss": 2.6256,
+      "step": 11486
+    },
+    {
+      "epoch": 0.7066313976377953,
+      "grad_norm": 1.4334502220153809,
+      "learning_rate": 3.959411562189533e-05,
+      "loss": 2.5711,
+      "step": 11487
+    },
+    {
+      "epoch": 0.7066929133858267,
+      "grad_norm": 1.3019689321517944,
+      "learning_rate": 3.957870581972706e-05,
+      "loss": 2.0114,
+      "step": 11488
+    },
+    {
+      "epoch": 0.7067544291338582,
+      "grad_norm": 1.3110214471817017,
+      "learning_rate": 3.956329827697831e-05,
+      "loss": 2.3823,
+      "step": 11489
+    },
+    {
+      "epoch": 0.7068159448818898,
+      "grad_norm": 1.2873128652572632,
+      "learning_rate": 3.9547892994225155e-05,
+      "loss": 2.1114,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7068774606299213,
+      "grad_norm": 1.3636795282363892,
+      "learning_rate": 3.953248997204376e-05,
+      "loss": 2.3281,
+      "step": 11491
+    },
+    {
+      "epoch": 0.7069389763779528,
+      "grad_norm": 1.5408352613449097,
+      "learning_rate": 3.951708921101001e-05,
+      "loss": 2.9132,
+      "step": 11492
+    },
+    {
+      "epoch": 0.7070004921259843,
+      "grad_norm": 1.1864523887634277,
+      "learning_rate": 3.950169071169988e-05,
+      "loss": 2.2854,
+      "step": 11493
+    },
+    {
+      "epoch": 0.7070620078740157,
+      "grad_norm": 1.383840560913086,
+      "learning_rate": 3.948629447468917e-05,
+      "loss": 2.1585,
+      "step": 11494
+    },
+    {
+      "epoch": 0.7071235236220472,
+      "grad_norm": 1.7038685083389282,
+      "learning_rate": 3.9470900500553634e-05,
+      "loss": 2.0102,
+      "step": 11495
+    },
+    {
+      "epoch": 0.7071850393700787,
+      "grad_norm": 1.1167140007019043,
+      "learning_rate": 3.945550878986888e-05,
+      "loss": 2.3424,
+      "step": 11496
+    },
+    {
+      "epoch": 0.7072465551181102,
+      "grad_norm": 1.2755109071731567,
+      "learning_rate": 3.9440119343210514e-05,
+      "loss": 2.8302,
+      "step": 11497
+    },
+    {
+      "epoch": 0.7073080708661418,
+      "grad_norm": 1.151110053062439,
+      "learning_rate": 3.9424732161153964e-05,
+      "loss": 2.9005,
+      "step": 11498
+    },
+    {
+      "epoch": 0.7073695866141733,
+      "grad_norm": 1.0670644044876099,
+      "learning_rate": 3.940934724427465e-05,
+      "loss": 1.7617,
+      "step": 11499
+    },
+    {
+      "epoch": 0.7074311023622047,
+      "grad_norm": 1.4836344718933105,
+      "learning_rate": 3.939396459314789e-05,
+      "loss": 2.2841,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7074926181102362,
+      "grad_norm": 1.2821238040924072,
+      "learning_rate": 3.937858420834895e-05,
+      "loss": 3.0206,
+      "step": 11501
+    },
+    {
+      "epoch": 0.7075541338582677,
+      "grad_norm": 1.2671197652816772,
+      "learning_rate": 3.936320609045287e-05,
+      "loss": 2.1462,
+      "step": 11502
+    },
+    {
+      "epoch": 0.7076156496062992,
+      "grad_norm": 1.4277632236480713,
+      "learning_rate": 3.934783024003481e-05,
+      "loss": 2.6121,
+      "step": 11503
+    },
+    {
+      "epoch": 0.7076771653543307,
+      "grad_norm": 1.3140496015548706,
+      "learning_rate": 3.933245665766966e-05,
+      "loss": 2.5759,
+      "step": 11504
+    },
+    {
+      "epoch": 0.7077386811023622,
+      "grad_norm": 1.1051963567733765,
+      "learning_rate": 3.931708534393237e-05,
+      "loss": 2.7385,
+      "step": 11505
+    },
+    {
+      "epoch": 0.7078001968503937,
+      "grad_norm": 1.1892433166503906,
+      "learning_rate": 3.930171629939764e-05,
+      "loss": 2.1284,
+      "step": 11506
+    },
+    {
+      "epoch": 0.7078617125984252,
+      "grad_norm": 1.3047523498535156,
+      "learning_rate": 3.928634952464033e-05,
+      "loss": 2.5696,
+      "step": 11507
+    },
+    {
+      "epoch": 0.7079232283464567,
+      "grad_norm": 1.3893890380859375,
+      "learning_rate": 3.927098502023496e-05,
+      "loss": 2.1641,
+      "step": 11508
+    },
+    {
+      "epoch": 0.7079847440944882,
+      "grad_norm": 1.3062366247177124,
+      "learning_rate": 3.925562278675616e-05,
+      "loss": 2.1742,
+      "step": 11509
+    },
+    {
+      "epoch": 0.7080462598425197,
+      "grad_norm": 1.3311398029327393,
+      "learning_rate": 3.924026282477831e-05,
+      "loss": 3.1032,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7081077755905512,
+      "grad_norm": 1.2769169807434082,
+      "learning_rate": 3.922490513487585e-05,
+      "loss": 2.5476,
+      "step": 11511
+    },
+    {
+      "epoch": 0.7081692913385826,
+      "grad_norm": 1.7183542251586914,
+      "learning_rate": 3.920954971762303e-05,
+      "loss": 2.4097,
+      "step": 11512
+    },
+    {
+      "epoch": 0.7082308070866141,
+      "grad_norm": 1.3309288024902344,
+      "learning_rate": 3.919419657359406e-05,
+      "loss": 2.8665,
+      "step": 11513
+    },
+    {
+      "epoch": 0.7082923228346457,
+      "grad_norm": 1.1842262744903564,
+      "learning_rate": 3.9178845703363076e-05,
+      "loss": 2.6679,
+      "step": 11514
+    },
+    {
+      "epoch": 0.7083538385826772,
+      "grad_norm": 1.2171181440353394,
+      "learning_rate": 3.916349710750415e-05,
+      "loss": 2.5158,
+      "step": 11515
+    },
+    {
+      "epoch": 0.7084153543307087,
+      "grad_norm": 1.3396886587142944,
+      "learning_rate": 3.9148150786591206e-05,
+      "loss": 3.0515,
+      "step": 11516
+    },
+    {
+      "epoch": 0.7084768700787402,
+      "grad_norm": 1.3525973558425903,
+      "learning_rate": 3.9132806741198067e-05,
+      "loss": 2.7038,
+      "step": 11517
+    },
+    {
+      "epoch": 0.7085383858267716,
+      "grad_norm": 1.4169843196868896,
+      "learning_rate": 3.911746497189854e-05,
+      "loss": 2.6689,
+      "step": 11518
+    },
+    {
+      "epoch": 0.7085999015748031,
+      "grad_norm": 1.097171425819397,
+      "learning_rate": 3.910212547926635e-05,
+      "loss": 2.8922,
+      "step": 11519
+    },
+    {
+      "epoch": 0.7086614173228346,
+      "grad_norm": 1.0633293390274048,
+      "learning_rate": 3.908678826387512e-05,
+      "loss": 2.254,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7087229330708661,
+      "grad_norm": 1.1689337491989136,
+      "learning_rate": 3.90714533262983e-05,
+      "loss": 2.9579,
+      "step": 11521
+    },
+    {
+      "epoch": 0.7087844488188977,
+      "grad_norm": 1.160155177116394,
+      "learning_rate": 3.905612066710943e-05,
+      "loss": 2.2808,
+      "step": 11522
+    },
+    {
+      "epoch": 0.7088459645669292,
+      "grad_norm": 1.4438343048095703,
+      "learning_rate": 3.9040790286881765e-05,
+      "loss": 2.1759,
+      "step": 11523
+    },
+    {
+      "epoch": 0.7089074803149606,
+      "grad_norm": 1.9965481758117676,
+      "learning_rate": 3.902546218618863e-05,
+      "loss": 1.8883,
+      "step": 11524
+    },
+    {
+      "epoch": 0.7089689960629921,
+      "grad_norm": 1.1832844018936157,
+      "learning_rate": 3.901013636560321e-05,
+      "loss": 2.3318,
+      "step": 11525
+    },
+    {
+      "epoch": 0.7090305118110236,
+      "grad_norm": 1.375765323638916,
+      "learning_rate": 3.8994812825698634e-05,
+      "loss": 2.5183,
+      "step": 11526
+    },
+    {
+      "epoch": 0.7090920275590551,
+      "grad_norm": 1.4517887830734253,
+      "learning_rate": 3.897949156704786e-05,
+      "loss": 2.0059,
+      "step": 11527
+    },
+    {
+      "epoch": 0.7091535433070866,
+      "grad_norm": 1.1776238679885864,
+      "learning_rate": 3.896417259022386e-05,
+      "loss": 2.5973,
+      "step": 11528
+    },
+    {
+      "epoch": 0.7092150590551181,
+      "grad_norm": 1.3567326068878174,
+      "learning_rate": 3.894885589579943e-05,
+      "loss": 2.3547,
+      "step": 11529
+    },
+    {
+      "epoch": 0.7092765748031497,
+      "grad_norm": 1.1598007678985596,
+      "learning_rate": 3.89335414843474e-05,
+      "loss": 2.8615,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7093380905511811,
+      "grad_norm": 1.360379695892334,
+      "learning_rate": 3.8918229356440364e-05,
+      "loss": 2.1362,
+      "step": 11531
+    },
+    {
+      "epoch": 0.7093996062992126,
+      "grad_norm": 1.3018360137939453,
+      "learning_rate": 3.8902919512650945e-05,
+      "loss": 2.6865,
+      "step": 11532
+    },
+    {
+      "epoch": 0.7094611220472441,
+      "grad_norm": 1.2760051488876343,
+      "learning_rate": 3.888761195355165e-05,
+      "loss": 2.3621,
+      "step": 11533
+    },
+    {
+      "epoch": 0.7095226377952756,
+      "grad_norm": 1.5302636623382568,
+      "learning_rate": 3.8872306679714945e-05,
+      "loss": 2.4472,
+      "step": 11534
+    },
+    {
+      "epoch": 0.7095841535433071,
+      "grad_norm": 1.1558959484100342,
+      "learning_rate": 3.885700369171306e-05,
+      "loss": 1.66,
+      "step": 11535
+    },
+    {
+      "epoch": 0.7096456692913385,
+      "grad_norm": 1.4265578985214233,
+      "learning_rate": 3.884170299011834e-05,
+      "loss": 2.3636,
+      "step": 11536
+    },
+    {
+      "epoch": 0.70970718503937,
+      "grad_norm": 1.2831860780715942,
+      "learning_rate": 3.882640457550285e-05,
+      "loss": 2.0883,
+      "step": 11537
+    },
+    {
+      "epoch": 0.7097687007874016,
+      "grad_norm": 1.1214302778244019,
+      "learning_rate": 3.8811108448438724e-05,
+      "loss": 2.4804,
+      "step": 11538
+    },
+    {
+      "epoch": 0.7098302165354331,
+      "grad_norm": 1.1466583013534546,
+      "learning_rate": 3.879581460949795e-05,
+      "loss": 2.4017,
+      "step": 11539
+    },
+    {
+      "epoch": 0.7098917322834646,
+      "grad_norm": 1.1779085397720337,
+      "learning_rate": 3.878052305925246e-05,
+      "loss": 1.7668,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7099532480314961,
+      "grad_norm": 1.1091883182525635,
+      "learning_rate": 3.8765233798274e-05,
+      "loss": 2.5927,
+      "step": 11541
+    },
+    {
+      "epoch": 0.7100147637795275,
+      "grad_norm": 1.5773255825042725,
+      "learning_rate": 3.8749946827134374e-05,
+      "loss": 2.3092,
+      "step": 11542
+    },
+    {
+      "epoch": 0.710076279527559,
+      "grad_norm": 1.1589010953903198,
+      "learning_rate": 3.873466214640516e-05,
+      "loss": 2.8095,
+      "step": 11543
+    },
+    {
+      "epoch": 0.7101377952755905,
+      "grad_norm": 1.2126344442367554,
+      "learning_rate": 3.871937975665797e-05,
+      "loss": 3.3995,
+      "step": 11544
+    },
+    {
+      "epoch": 0.710199311023622,
+      "grad_norm": 1.322676181793213,
+      "learning_rate": 3.8704099658464255e-05,
+      "loss": 2.6524,
+      "step": 11545
+    },
+    {
+      "epoch": 0.7102608267716536,
+      "grad_norm": 1.2105664014816284,
+      "learning_rate": 3.868882185239545e-05,
+      "loss": 3.0512,
+      "step": 11546
+    },
+    {
+      "epoch": 0.7103223425196851,
+      "grad_norm": 1.2760450839996338,
+      "learning_rate": 3.8673546339022803e-05,
+      "loss": 1.81,
+      "step": 11547
+    },
+    {
+      "epoch": 0.7103838582677166,
+      "grad_norm": 1.7503397464752197,
+      "learning_rate": 3.8658273118917586e-05,
+      "loss": 2.115,
+      "step": 11548
+    },
+    {
+      "epoch": 0.710445374015748,
+      "grad_norm": 1.1482179164886475,
+      "learning_rate": 3.864300219265087e-05,
+      "loss": 1.9237,
+      "step": 11549
+    },
+    {
+      "epoch": 0.7105068897637795,
+      "grad_norm": 1.2381162643432617,
+      "learning_rate": 3.8627733560793734e-05,
+      "loss": 2.548,
+      "step": 11550
+    },
+    {
+      "epoch": 0.710568405511811,
+      "grad_norm": 1.2354072332382202,
+      "learning_rate": 3.861246722391715e-05,
+      "loss": 2.7429,
+      "step": 11551
+    },
+    {
+      "epoch": 0.7106299212598425,
+      "grad_norm": 1.3888118267059326,
+      "learning_rate": 3.859720318259201e-05,
+      "loss": 2.1865,
+      "step": 11552
+    },
+    {
+      "epoch": 0.710691437007874,
+      "grad_norm": 1.0690685510635376,
+      "learning_rate": 3.858194143738908e-05,
+      "loss": 2.6002,
+      "step": 11553
+    },
+    {
+      "epoch": 0.7107529527559056,
+      "grad_norm": 1.436797857284546,
+      "learning_rate": 3.8566681988879035e-05,
+      "loss": 2.4693,
+      "step": 11554
+    },
+    {
+      "epoch": 0.710814468503937,
+      "grad_norm": 1.119325041770935,
+      "learning_rate": 3.855142483763255e-05,
+      "loss": 2.8789,
+      "step": 11555
+    },
+    {
+      "epoch": 0.7108759842519685,
+      "grad_norm": 1.1515002250671387,
+      "learning_rate": 3.8536169984220105e-05,
+      "loss": 2.5862,
+      "step": 11556
+    },
+    {
+      "epoch": 0.7109375,
+      "grad_norm": 1.1689637899398804,
+      "learning_rate": 3.8520917429212166e-05,
+      "loss": 2.6475,
+      "step": 11557
+    },
+    {
+      "epoch": 0.7109990157480315,
+      "grad_norm": 1.4010792970657349,
+      "learning_rate": 3.8505667173179094e-05,
+      "loss": 3.2613,
+      "step": 11558
+    },
+    {
+      "epoch": 0.711060531496063,
+      "grad_norm": 1.201626181602478,
+      "learning_rate": 3.8490419216691206e-05,
+      "loss": 2.7622,
+      "step": 11559
+    },
+    {
+      "epoch": 0.7111220472440944,
+      "grad_norm": 1.0983119010925293,
+      "learning_rate": 3.847517356031861e-05,
+      "loss": 2.8848,
+      "step": 11560
+    },
+    {
+      "epoch": 0.711183562992126,
+      "grad_norm": 1.2421927452087402,
+      "learning_rate": 3.84599302046315e-05,
+      "loss": 2.4295,
+      "step": 11561
+    },
+    {
+      "epoch": 0.7112450787401575,
+      "grad_norm": 1.3935506343841553,
+      "learning_rate": 3.84446891501998e-05,
+      "loss": 2.4925,
+      "step": 11562
+    },
+    {
+      "epoch": 0.711306594488189,
+      "grad_norm": 1.4336262941360474,
+      "learning_rate": 3.842945039759348e-05,
+      "loss": 2.0038,
+      "step": 11563
+    },
+    {
+      "epoch": 0.7113681102362205,
+      "grad_norm": 1.2119282484054565,
+      "learning_rate": 3.84142139473824e-05,
+      "loss": 2.8096,
+      "step": 11564
+    },
+    {
+      "epoch": 0.711429625984252,
+      "grad_norm": 1.2533329725265503,
+      "learning_rate": 3.8398979800136337e-05,
+      "loss": 2.3214,
+      "step": 11565
+    },
+    {
+      "epoch": 0.7114911417322834,
+      "grad_norm": 1.1096575260162354,
+      "learning_rate": 3.83837479564249e-05,
+      "loss": 2.4464,
+      "step": 11566
+    },
+    {
+      "epoch": 0.7115526574803149,
+      "grad_norm": 1.2184727191925049,
+      "learning_rate": 3.836851841681774e-05,
+      "loss": 2.7973,
+      "step": 11567
+    },
+    {
+      "epoch": 0.7116141732283464,
+      "grad_norm": 1.541491985321045,
+      "learning_rate": 3.8353291181884286e-05,
+      "loss": 2.6258,
+      "step": 11568
+    },
+    {
+      "epoch": 0.711675688976378,
+      "grad_norm": 1.3863837718963623,
+      "learning_rate": 3.8338066252193996e-05,
+      "loss": 2.3195,
+      "step": 11569
+    },
+    {
+      "epoch": 0.7117372047244095,
+      "grad_norm": 1.2436977624893188,
+      "learning_rate": 3.832284362831619e-05,
+      "loss": 2.5337,
+      "step": 11570
+    },
+    {
+      "epoch": 0.711798720472441,
+      "grad_norm": 1.1483503580093384,
+      "learning_rate": 3.830762331082014e-05,
+      "loss": 2.404,
+      "step": 11571
+    },
+    {
+      "epoch": 0.7118602362204725,
+      "grad_norm": 1.438163161277771,
+      "learning_rate": 3.8292405300274935e-05,
+      "loss": 2.6898,
+      "step": 11572
+    },
+    {
+      "epoch": 0.7119217519685039,
+      "grad_norm": 1.1250652074813843,
+      "learning_rate": 3.827718959724973e-05,
+      "loss": 2.7461,
+      "step": 11573
+    },
+    {
+      "epoch": 0.7119832677165354,
+      "grad_norm": 1.2833493947982788,
+      "learning_rate": 3.826197620231341e-05,
+      "loss": 2.778,
+      "step": 11574
+    },
+    {
+      "epoch": 0.7120447834645669,
+      "grad_norm": 1.1940213441848755,
+      "learning_rate": 3.824676511603497e-05,
+      "loss": 2.1743,
+      "step": 11575
+    },
+    {
+      "epoch": 0.7121062992125984,
+      "grad_norm": 1.4337599277496338,
+      "learning_rate": 3.823155633898312e-05,
+      "loss": 2.6779,
+      "step": 11576
+    },
+    {
+      "epoch": 0.71216781496063,
+      "grad_norm": 1.5166285037994385,
+      "learning_rate": 3.821634987172663e-05,
+      "loss": 2.7616,
+      "step": 11577
+    },
+    {
+      "epoch": 0.7122293307086615,
+      "grad_norm": 1.281347632408142,
+      "learning_rate": 3.820114571483415e-05,
+      "loss": 2.4382,
+      "step": 11578
+    },
+    {
+      "epoch": 0.7122908464566929,
+      "grad_norm": 1.2893019914627075,
+      "learning_rate": 3.8185943868874255e-05,
+      "loss": 2.0838,
+      "step": 11579
+    },
+    {
+      "epoch": 0.7123523622047244,
+      "grad_norm": 1.4669790267944336,
+      "learning_rate": 3.817074433441534e-05,
+      "loss": 2.7203,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7124138779527559,
+      "grad_norm": 1.2268128395080566,
+      "learning_rate": 3.8155547112025846e-05,
+      "loss": 2.627,
+      "step": 11581
+    },
+    {
+      "epoch": 0.7124753937007874,
+      "grad_norm": 1.2651742696762085,
+      "learning_rate": 3.814035220227401e-05,
+      "loss": 2.5466,
+      "step": 11582
+    },
+    {
+      "epoch": 0.7125369094488189,
+      "grad_norm": 1.2133398056030273,
+      "learning_rate": 3.812515960572805e-05,
+      "loss": 2.4801,
+      "step": 11583
+    },
+    {
+      "epoch": 0.7125984251968503,
+      "grad_norm": 1.1724555492401123,
+      "learning_rate": 3.810996932295611e-05,
+      "loss": 2.6582,
+      "step": 11584
+    },
+    {
+      "epoch": 0.7126599409448819,
+      "grad_norm": 1.3960292339324951,
+      "learning_rate": 3.809478135452623e-05,
+      "loss": 2.63,
+      "step": 11585
+    },
+    {
+      "epoch": 0.7127214566929134,
+      "grad_norm": 1.2330083847045898,
+      "learning_rate": 3.807959570100631e-05,
+      "loss": 2.3676,
+      "step": 11586
+    },
+    {
+      "epoch": 0.7127829724409449,
+      "grad_norm": 1.316048502922058,
+      "learning_rate": 3.8064412362964264e-05,
+      "loss": 2.5935,
+      "step": 11587
+    },
+    {
+      "epoch": 0.7128444881889764,
+      "grad_norm": 1.6337345838546753,
+      "learning_rate": 3.804923134096781e-05,
+      "loss": 2.676,
+      "step": 11588
+    },
+    {
+      "epoch": 0.7129060039370079,
+      "grad_norm": 1.0411579608917236,
+      "learning_rate": 3.803405263558464e-05,
+      "loss": 2.7608,
+      "step": 11589
+    },
+    {
+      "epoch": 0.7129675196850394,
+      "grad_norm": 1.2435951232910156,
+      "learning_rate": 3.801887624738242e-05,
+      "loss": 2.7763,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7130290354330708,
+      "grad_norm": 1.203134298324585,
+      "learning_rate": 3.800370217692857e-05,
+      "loss": 2.5107,
+      "step": 11591
+    },
+    {
+      "epoch": 0.7130905511811023,
+      "grad_norm": 1.149448275566101,
+      "learning_rate": 3.798853042479058e-05,
+      "loss": 2.7525,
+      "step": 11592
+    },
+    {
+      "epoch": 0.7131520669291339,
+      "grad_norm": 1.2259821891784668,
+      "learning_rate": 3.797336099153575e-05,
+      "loss": 2.3428,
+      "step": 11593
+    },
+    {
+      "epoch": 0.7132135826771654,
+      "grad_norm": 1.3198069334030151,
+      "learning_rate": 3.795819387773137e-05,
+      "loss": 2.0629,
+      "step": 11594
+    },
+    {
+      "epoch": 0.7132750984251969,
+      "grad_norm": 1.1208547353744507,
+      "learning_rate": 3.794302908394453e-05,
+      "loss": 2.8075,
+      "step": 11595
+    },
+    {
+      "epoch": 0.7133366141732284,
+      "grad_norm": 1.2945438623428345,
+      "learning_rate": 3.792786661074243e-05,
+      "loss": 2.7955,
+      "step": 11596
+    },
+    {
+      "epoch": 0.7133981299212598,
+      "grad_norm": 1.5281891822814941,
+      "learning_rate": 3.7912706458691946e-05,
+      "loss": 2.5312,
+      "step": 11597
+    },
+    {
+      "epoch": 0.7134596456692913,
+      "grad_norm": 1.4349414110183716,
+      "learning_rate": 3.789754862836008e-05,
+      "loss": 2.2326,
+      "step": 11598
+    },
+    {
+      "epoch": 0.7135211614173228,
+      "grad_norm": 1.1802762746810913,
+      "learning_rate": 3.7882393120313565e-05,
+      "loss": 3.1196,
+      "step": 11599
+    },
+    {
+      "epoch": 0.7135826771653543,
+      "grad_norm": 1.2450141906738281,
+      "learning_rate": 3.7867239935119206e-05,
+      "loss": 2.3413,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7136441929133859,
+      "grad_norm": 1.2813291549682617,
+      "learning_rate": 3.7852089073343566e-05,
+      "loss": 2.7316,
+      "step": 11601
+    },
+    {
+      "epoch": 0.7137057086614174,
+      "grad_norm": 1.0916022062301636,
+      "learning_rate": 3.783694053555326e-05,
+      "loss": 2.3812,
+      "step": 11602
+    },
+    {
+      "epoch": 0.7137672244094488,
+      "grad_norm": 1.285207748413086,
+      "learning_rate": 3.782179432231474e-05,
+      "loss": 2.0867,
+      "step": 11603
+    },
+    {
+      "epoch": 0.7138287401574803,
+      "grad_norm": 1.4066834449768066,
+      "learning_rate": 3.780665043419444e-05,
+      "loss": 1.9546,
+      "step": 11604
+    },
+    {
+      "epoch": 0.7138902559055118,
+      "grad_norm": 1.1577272415161133,
+      "learning_rate": 3.779150887175857e-05,
+      "loss": 2.1927,
+      "step": 11605
+    },
+    {
+      "epoch": 0.7139517716535433,
+      "grad_norm": 1.4455896615982056,
+      "learning_rate": 3.777636963557343e-05,
+      "loss": 2.4581,
+      "step": 11606
+    },
+    {
+      "epoch": 0.7140132874015748,
+      "grad_norm": 1.2743539810180664,
+      "learning_rate": 3.7761232726205054e-05,
+      "loss": 2.7952,
+      "step": 11607
+    },
+    {
+      "epoch": 0.7140748031496063,
+      "grad_norm": 1.312334656715393,
+      "learning_rate": 3.7746098144219525e-05,
+      "loss": 2.4898,
+      "step": 11608
+    },
+    {
+      "epoch": 0.7141363188976378,
+      "grad_norm": 1.281887412071228,
+      "learning_rate": 3.773096589018279e-05,
+      "loss": 3.2678,
+      "step": 11609
+    },
+    {
+      "epoch": 0.7141978346456693,
+      "grad_norm": 1.2567912340164185,
+      "learning_rate": 3.771583596466074e-05,
+      "loss": 2.7683,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7142593503937008,
+      "grad_norm": 1.1119362115859985,
+      "learning_rate": 3.7700708368219085e-05,
+      "loss": 2.633,
+      "step": 11611
+    },
+    {
+      "epoch": 0.7143208661417323,
+      "grad_norm": 1.1444907188415527,
+      "learning_rate": 3.768558310142358e-05,
+      "loss": 2.5086,
+      "step": 11612
+    },
+    {
+      "epoch": 0.7143823818897638,
+      "grad_norm": 1.2926464080810547,
+      "learning_rate": 3.767046016483976e-05,
+      "loss": 2.8966,
+      "step": 11613
+    },
+    {
+      "epoch": 0.7144438976377953,
+      "grad_norm": 1.3072245121002197,
+      "learning_rate": 3.7655339559033165e-05,
+      "loss": 1.8058,
+      "step": 11614
+    },
+    {
+      "epoch": 0.7145054133858267,
+      "grad_norm": 1.1779320240020752,
+      "learning_rate": 3.764022128456923e-05,
+      "loss": 2.5607,
+      "step": 11615
+    },
+    {
+      "epoch": 0.7145669291338582,
+      "grad_norm": 1.2803990840911865,
+      "learning_rate": 3.762510534201333e-05,
+      "loss": 3.012,
+      "step": 11616
+    },
+    {
+      "epoch": 0.7146284448818898,
+      "grad_norm": 1.2755900621414185,
+      "learning_rate": 3.760999173193065e-05,
+      "loss": 2.1636,
+      "step": 11617
+    },
+    {
+      "epoch": 0.7146899606299213,
+      "grad_norm": 1.2274677753448486,
+      "learning_rate": 3.759488045488641e-05,
+      "loss": 2.2793,
+      "step": 11618
+    },
+    {
+      "epoch": 0.7147514763779528,
+      "grad_norm": 1.2291501760482788,
+      "learning_rate": 3.757977151144562e-05,
+      "loss": 2.6142,
+      "step": 11619
+    },
+    {
+      "epoch": 0.7148129921259843,
+      "grad_norm": 1.614369511604309,
+      "learning_rate": 3.756466490217335e-05,
+      "loss": 2.0265,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7148745078740157,
+      "grad_norm": 1.8595967292785645,
+      "learning_rate": 3.754956062763439e-05,
+      "loss": 1.9278,
+      "step": 11621
+    },
+    {
+      "epoch": 0.7149360236220472,
+      "grad_norm": 1.2571266889572144,
+      "learning_rate": 3.75344586883937e-05,
+      "loss": 2.5008,
+      "step": 11622
+    },
+    {
+      "epoch": 0.7149975393700787,
+      "grad_norm": 1.4159725904464722,
+      "learning_rate": 3.751935908501592e-05,
+      "loss": 1.9916,
+      "step": 11623
+    },
+    {
+      "epoch": 0.7150590551181102,
+      "grad_norm": 1.1552990674972534,
+      "learning_rate": 3.750426181806572e-05,
+      "loss": 2.6834,
+      "step": 11624
+    },
+    {
+      "epoch": 0.7151205708661418,
+      "grad_norm": 1.3313719034194946,
+      "learning_rate": 3.748916688810765e-05,
+      "loss": 2.6906,
+      "step": 11625
+    },
+    {
+      "epoch": 0.7151820866141733,
+      "grad_norm": 1.384554147720337,
+      "learning_rate": 3.7474074295706144e-05,
+      "loss": 2.2166,
+      "step": 11626
+    },
+    {
+      "epoch": 0.7152436023622047,
+      "grad_norm": 1.2205703258514404,
+      "learning_rate": 3.7458984041425594e-05,
+      "loss": 2.6176,
+      "step": 11627
+    },
+    {
+      "epoch": 0.7153051181102362,
+      "grad_norm": 1.4862256050109863,
+      "learning_rate": 3.74438961258303e-05,
+      "loss": 2.377,
+      "step": 11628
+    },
+    {
+      "epoch": 0.7153666338582677,
+      "grad_norm": 1.1929333209991455,
+      "learning_rate": 3.7428810549484514e-05,
+      "loss": 2.4071,
+      "step": 11629
+    },
+    {
+      "epoch": 0.7154281496062992,
+      "grad_norm": 1.2207903861999512,
+      "learning_rate": 3.7413727312952265e-05,
+      "loss": 2.2447,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7154896653543307,
+      "grad_norm": 1.2480076551437378,
+      "learning_rate": 3.7398646416797656e-05,
+      "loss": 2.2562,
+      "step": 11631
+    },
+    {
+      "epoch": 0.7155511811023622,
+      "grad_norm": 1.4144788980484009,
+      "learning_rate": 3.738356786158457e-05,
+      "loss": 2.8838,
+      "step": 11632
+    },
+    {
+      "epoch": 0.7156126968503937,
+      "grad_norm": 1.5117623805999756,
+      "learning_rate": 3.736849164787688e-05,
+      "loss": 2.6135,
+      "step": 11633
+    },
+    {
+      "epoch": 0.7156742125984252,
+      "grad_norm": 1.1877944469451904,
+      "learning_rate": 3.735341777623836e-05,
+      "loss": 2.0768,
+      "step": 11634
+    },
+    {
+      "epoch": 0.7157357283464567,
+      "grad_norm": 1.3870924711227417,
+      "learning_rate": 3.7338346247232736e-05,
+      "loss": 2.7529,
+      "step": 11635
+    },
+    {
+      "epoch": 0.7157972440944882,
+      "grad_norm": 1.4035099744796753,
+      "learning_rate": 3.73232770614235e-05,
+      "loss": 2.3157,
+      "step": 11636
+    },
+    {
+      "epoch": 0.7158587598425197,
+      "grad_norm": 1.2267934083938599,
+      "learning_rate": 3.730821021937425e-05,
+      "loss": 2.8566,
+      "step": 11637
+    },
+    {
+      "epoch": 0.7159202755905512,
+      "grad_norm": 1.3071563243865967,
+      "learning_rate": 3.729314572164833e-05,
+      "loss": 2.5211,
+      "step": 11638
+    },
+    {
+      "epoch": 0.7159817913385826,
+      "grad_norm": 1.438123106956482,
+      "learning_rate": 3.7278083568809084e-05,
+      "loss": 2.5465,
+      "step": 11639
+    },
+    {
+      "epoch": 0.7160433070866141,
+      "grad_norm": 1.1488988399505615,
+      "learning_rate": 3.726302376141977e-05,
+      "loss": 2.2796,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7161048228346457,
+      "grad_norm": 1.4376230239868164,
+      "learning_rate": 3.7247966300043566e-05,
+      "loss": 2.3658,
+      "step": 11641
+    },
+    {
+      "epoch": 0.7161663385826772,
+      "grad_norm": 1.3787140846252441,
+      "learning_rate": 3.723291118524348e-05,
+      "loss": 2.107,
+      "step": 11642
+    },
+    {
+      "epoch": 0.7162278543307087,
+      "grad_norm": 1.350182056427002,
+      "learning_rate": 3.7217858417582554e-05,
+      "loss": 2.5733,
+      "step": 11643
+    },
+    {
+      "epoch": 0.7162893700787402,
+      "grad_norm": 1.144217848777771,
+      "learning_rate": 3.72028079976236e-05,
+      "loss": 2.6992,
+      "step": 11644
+    },
+    {
+      "epoch": 0.7163508858267716,
+      "grad_norm": 1.2938629388809204,
+      "learning_rate": 3.718775992592949e-05,
+      "loss": 1.7127,
+      "step": 11645
+    },
+    {
+      "epoch": 0.7164124015748031,
+      "grad_norm": 1.2201993465423584,
+      "learning_rate": 3.7172714203062884e-05,
+      "loss": 1.8277,
+      "step": 11646
+    },
+    {
+      "epoch": 0.7164739173228346,
+      "grad_norm": 1.3499064445495605,
+      "learning_rate": 3.715767082958643e-05,
+      "loss": 2.1764,
+      "step": 11647
+    },
+    {
+      "epoch": 0.7165354330708661,
+      "grad_norm": 1.445281744003296,
+      "learning_rate": 3.714262980606267e-05,
+      "loss": 2.1053,
+      "step": 11648
+    },
+    {
+      "epoch": 0.7165969488188977,
+      "grad_norm": 1.6332470178604126,
+      "learning_rate": 3.712759113305407e-05,
+      "loss": 2.4357,
+      "step": 11649
+    },
+    {
+      "epoch": 0.7166584645669292,
+      "grad_norm": 1.4761842489242554,
+      "learning_rate": 3.711255481112297e-05,
+      "loss": 2.3586,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7167199803149606,
+      "grad_norm": 1.2986505031585693,
+      "learning_rate": 3.7097520840831666e-05,
+      "loss": 2.7446,
+      "step": 11651
+    },
+    {
+      "epoch": 0.7167814960629921,
+      "grad_norm": 1.3048765659332275,
+      "learning_rate": 3.708248922274229e-05,
+      "loss": 2.1914,
+      "step": 11652
+    },
+    {
+      "epoch": 0.7168430118110236,
+      "grad_norm": 1.2400672435760498,
+      "learning_rate": 3.706745995741699e-05,
+      "loss": 3.021,
+      "step": 11653
+    },
+    {
+      "epoch": 0.7169045275590551,
+      "grad_norm": 1.4402581453323364,
+      "learning_rate": 3.7052433045417766e-05,
+      "loss": 2.7073,
+      "step": 11654
+    },
+    {
+      "epoch": 0.7169660433070866,
+      "grad_norm": 1.6982648372650146,
+      "learning_rate": 3.7037408487306566e-05,
+      "loss": 2.0464,
+      "step": 11655
+    },
+    {
+      "epoch": 0.7170275590551181,
+      "grad_norm": 1.2637518644332886,
+      "learning_rate": 3.702238628364517e-05,
+      "loss": 1.6606,
+      "step": 11656
+    },
+    {
+      "epoch": 0.7170890748031497,
+      "grad_norm": 1.3622041940689087,
+      "learning_rate": 3.70073664349954e-05,
+      "loss": 2.1433,
+      "step": 11657
+    },
+    {
+      "epoch": 0.7171505905511811,
+      "grad_norm": 1.4817968606948853,
+      "learning_rate": 3.699234894191882e-05,
+      "loss": 2.043,
+      "step": 11658
+    },
+    {
+      "epoch": 0.7172121062992126,
+      "grad_norm": 1.0837593078613281,
+      "learning_rate": 3.6977333804977056e-05,
+      "loss": 2.8579,
+      "step": 11659
+    },
+    {
+      "epoch": 0.7172736220472441,
+      "grad_norm": 1.4242103099822998,
+      "learning_rate": 3.696232102473158e-05,
+      "loss": 2.2207,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7173351377952756,
+      "grad_norm": 1.3375438451766968,
+      "learning_rate": 3.6947310601743825e-05,
+      "loss": 2.1642,
+      "step": 11661
+    },
+    {
+      "epoch": 0.7173966535433071,
+      "grad_norm": 1.355574131011963,
+      "learning_rate": 3.693230253657507e-05,
+      "loss": 2.3103,
+      "step": 11662
+    },
+    {
+      "epoch": 0.7174581692913385,
+      "grad_norm": 1.2483994960784912,
+      "learning_rate": 3.691729682978647e-05,
+      "loss": 2.553,
+      "step": 11663
+    },
+    {
+      "epoch": 0.71751968503937,
+      "grad_norm": 1.2945903539657593,
+      "learning_rate": 3.6902293481939265e-05,
+      "loss": 2.3053,
+      "step": 11664
+    },
+    {
+      "epoch": 0.7175812007874016,
+      "grad_norm": 1.2607200145721436,
+      "learning_rate": 3.688729249359436e-05,
+      "loss": 2.5603,
+      "step": 11665
+    },
+    {
+      "epoch": 0.7176427165354331,
+      "grad_norm": 1.039909839630127,
+      "learning_rate": 3.687229386531287e-05,
+      "loss": 2.6712,
+      "step": 11666
+    },
+    {
+      "epoch": 0.7177042322834646,
+      "grad_norm": 1.3438429832458496,
+      "learning_rate": 3.685729759765554e-05,
+      "loss": 2.5613,
+      "step": 11667
+    },
+    {
+      "epoch": 0.7177657480314961,
+      "grad_norm": 1.3311586380004883,
+      "learning_rate": 3.684230369118322e-05,
+      "loss": 2.0361,
+      "step": 11668
+    },
+    {
+      "epoch": 0.7178272637795275,
+      "grad_norm": 1.6782035827636719,
+      "learning_rate": 3.682731214645653e-05,
+      "loss": 2.7052,
+      "step": 11669
+    },
+    {
+      "epoch": 0.717888779527559,
+      "grad_norm": 1.2912354469299316,
+      "learning_rate": 3.6812322964036136e-05,
+      "loss": 1.8602,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7179502952755905,
+      "grad_norm": 1.2957042455673218,
+      "learning_rate": 3.679733614448249e-05,
+      "loss": 2.6179,
+      "step": 11671
+    },
+    {
+      "epoch": 0.718011811023622,
+      "grad_norm": 1.436549425125122,
+      "learning_rate": 3.6782351688356044e-05,
+      "loss": 1.5245,
+      "step": 11672
+    },
+    {
+      "epoch": 0.7180733267716536,
+      "grad_norm": 1.3319896459579468,
+      "learning_rate": 3.676736959621714e-05,
+      "loss": 2.8668,
+      "step": 11673
+    },
+    {
+      "epoch": 0.7181348425196851,
+      "grad_norm": 1.4609068632125854,
+      "learning_rate": 3.6752389868626034e-05,
+      "loss": 2.1551,
+      "step": 11674
+    },
+    {
+      "epoch": 0.7181963582677166,
+      "grad_norm": 1.248576045036316,
+      "learning_rate": 3.6737412506142855e-05,
+      "loss": 2.6479,
+      "step": 11675
+    },
+    {
+      "epoch": 0.718257874015748,
+      "grad_norm": 1.332768201828003,
+      "learning_rate": 3.67224375093277e-05,
+      "loss": 2.1675,
+      "step": 11676
+    },
+    {
+      "epoch": 0.7183193897637795,
+      "grad_norm": 1.4903850555419922,
+      "learning_rate": 3.6707464878740516e-05,
+      "loss": 1.2792,
+      "step": 11677
+    },
+    {
+      "epoch": 0.718380905511811,
+      "grad_norm": 1.1136527061462402,
+      "learning_rate": 3.669249461494122e-05,
+      "loss": 2.3457,
+      "step": 11678
+    },
+    {
+      "epoch": 0.7184424212598425,
+      "grad_norm": 1.4492982625961304,
+      "learning_rate": 3.6677526718489606e-05,
+      "loss": 1.8649,
+      "step": 11679
+    },
+    {
+      "epoch": 0.718503937007874,
+      "grad_norm": 1.2982844114303589,
+      "learning_rate": 3.6662561189945434e-05,
+      "loss": 2.6223,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7185654527559056,
+      "grad_norm": 1.3859153985977173,
+      "learning_rate": 3.664759802986826e-05,
+      "loss": 2.281,
+      "step": 11681
+    },
+    {
+      "epoch": 0.718626968503937,
+      "grad_norm": 1.1992931365966797,
+      "learning_rate": 3.6632637238817694e-05,
+      "loss": 2.8204,
+      "step": 11682
+    },
+    {
+      "epoch": 0.7186884842519685,
+      "grad_norm": 1.2387803792953491,
+      "learning_rate": 3.661767881735311e-05,
+      "loss": 2.4159,
+      "step": 11683
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 1.1628577709197998,
+      "learning_rate": 3.6602722766033905e-05,
+      "loss": 2.5498,
+      "step": 11684
+    },
+    {
+      "epoch": 0.7188115157480315,
+      "grad_norm": 1.4340437650680542,
+      "learning_rate": 3.6587769085419353e-05,
+      "loss": 2.0748,
+      "step": 11685
+    },
+    {
+      "epoch": 0.718873031496063,
+      "grad_norm": 1.8489739894866943,
+      "learning_rate": 3.657281777606869e-05,
+      "loss": 2.6033,
+      "step": 11686
+    },
+    {
+      "epoch": 0.7189345472440944,
+      "grad_norm": 1.4951868057250977,
+      "learning_rate": 3.6557868838540924e-05,
+      "loss": 2.1699,
+      "step": 11687
+    },
+    {
+      "epoch": 0.718996062992126,
+      "grad_norm": 1.2771003246307373,
+      "learning_rate": 3.654292227339513e-05,
+      "loss": 2.2721,
+      "step": 11688
+    },
+    {
+      "epoch": 0.7190575787401575,
+      "grad_norm": 1.3423595428466797,
+      "learning_rate": 3.652797808119016e-05,
+      "loss": 2.0995,
+      "step": 11689
+    },
+    {
+      "epoch": 0.719119094488189,
+      "grad_norm": 1.3850088119506836,
+      "learning_rate": 3.651303626248491e-05,
+      "loss": 2.479,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7191806102362205,
+      "grad_norm": 1.259966492652893,
+      "learning_rate": 3.649809681783804e-05,
+      "loss": 2.6628,
+      "step": 11691
+    },
+    {
+      "epoch": 0.719242125984252,
+      "grad_norm": 1.0979254245758057,
+      "learning_rate": 3.648315974780831e-05,
+      "loss": 3.0561,
+      "step": 11692
+    },
+    {
+      "epoch": 0.7193036417322834,
+      "grad_norm": 2.2217612266540527,
+      "learning_rate": 3.6468225052954195e-05,
+      "loss": 1.227,
+      "step": 11693
+    },
+    {
+      "epoch": 0.7193651574803149,
+      "grad_norm": 1.2583540678024292,
+      "learning_rate": 3.6453292733834235e-05,
+      "loss": 3.0873,
+      "step": 11694
+    },
+    {
+      "epoch": 0.7194266732283464,
+      "grad_norm": 1.2549370527267456,
+      "learning_rate": 3.643836279100674e-05,
+      "loss": 2.9005,
+      "step": 11695
+    },
+    {
+      "epoch": 0.719488188976378,
+      "grad_norm": 1.3164347410202026,
+      "learning_rate": 3.642343522503008e-05,
+      "loss": 2.3265,
+      "step": 11696
+    },
+    {
+      "epoch": 0.7195497047244095,
+      "grad_norm": 1.1510121822357178,
+      "learning_rate": 3.6408510036462406e-05,
+      "loss": 3.0492,
+      "step": 11697
+    },
+    {
+      "epoch": 0.719611220472441,
+      "grad_norm": 1.202017903327942,
+      "learning_rate": 3.639358722586186e-05,
+      "loss": 2.6352,
+      "step": 11698
+    },
+    {
+      "epoch": 0.7196727362204725,
+      "grad_norm": 1.6390314102172852,
+      "learning_rate": 3.6378666793786496e-05,
+      "loss": 2.7849,
+      "step": 11699
+    },
+    {
+      "epoch": 0.7197342519685039,
+      "grad_norm": 1.2407679557800293,
+      "learning_rate": 3.63637487407942e-05,
+      "loss": 2.4973,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7197957677165354,
+      "grad_norm": 1.4559763669967651,
+      "learning_rate": 3.63488330674429e-05,
+      "loss": 2.2087,
+      "step": 11701
+    },
+    {
+      "epoch": 0.7198572834645669,
+      "grad_norm": 1.1650235652923584,
+      "learning_rate": 3.633391977429027e-05,
+      "loss": 2.3571,
+      "step": 11702
+    },
+    {
+      "epoch": 0.7199187992125984,
+      "grad_norm": 1.4068934917449951,
+      "learning_rate": 3.631900886189403e-05,
+      "loss": 2.7363,
+      "step": 11703
+    },
+    {
+      "epoch": 0.71998031496063,
+      "grad_norm": 1.3051484823226929,
+      "learning_rate": 3.630410033081178e-05,
+      "loss": 2.7853,
+      "step": 11704
+    },
+    {
+      "epoch": 0.7200418307086615,
+      "grad_norm": 1.3786773681640625,
+      "learning_rate": 3.628919418160101e-05,
+      "loss": 2.2465,
+      "step": 11705
+    },
+    {
+      "epoch": 0.7201033464566929,
+      "grad_norm": 1.5577940940856934,
+      "learning_rate": 3.627429041481909e-05,
+      "loss": 2.9318,
+      "step": 11706
+    },
+    {
+      "epoch": 0.7201648622047244,
+      "grad_norm": 1.2584038972854614,
+      "learning_rate": 3.6259389031023405e-05,
+      "loss": 2.5644,
+      "step": 11707
+    },
+    {
+      "epoch": 0.7202263779527559,
+      "grad_norm": 1.303151249885559,
+      "learning_rate": 3.624449003077111e-05,
+      "loss": 2.7301,
+      "step": 11708
+    },
+    {
+      "epoch": 0.7202878937007874,
+      "grad_norm": 1.350799322128296,
+      "learning_rate": 3.62295934146194e-05,
+      "loss": 2.6824,
+      "step": 11709
+    },
+    {
+      "epoch": 0.7203494094488189,
+      "grad_norm": 1.1669620275497437,
+      "learning_rate": 3.6214699183125245e-05,
+      "loss": 2.4993,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7204109251968503,
+      "grad_norm": 1.474509596824646,
+      "learning_rate": 3.619980733684573e-05,
+      "loss": 2.4335,
+      "step": 11711
+    },
+    {
+      "epoch": 0.7204724409448819,
+      "grad_norm": 1.2688404321670532,
+      "learning_rate": 3.6184917876337624e-05,
+      "loss": 2.8818,
+      "step": 11712
+    },
+    {
+      "epoch": 0.7205339566929134,
+      "grad_norm": 1.2064685821533203,
+      "learning_rate": 3.6170030802157775e-05,
+      "loss": 2.6296,
+      "step": 11713
+    },
+    {
+      "epoch": 0.7205954724409449,
+      "grad_norm": 1.2382420301437378,
+      "learning_rate": 3.6155146114862826e-05,
+      "loss": 2.5163,
+      "step": 11714
+    },
+    {
+      "epoch": 0.7206569881889764,
+      "grad_norm": 1.2627203464508057,
+      "learning_rate": 3.6140263815009426e-05,
+      "loss": 2.7422,
+      "step": 11715
+    },
+    {
+      "epoch": 0.7207185039370079,
+      "grad_norm": 1.2176969051361084,
+      "learning_rate": 3.612538390315404e-05,
+      "loss": 2.1713,
+      "step": 11716
+    },
+    {
+      "epoch": 0.7207800196850394,
+      "grad_norm": 1.4880679845809937,
+      "learning_rate": 3.611050637985313e-05,
+      "loss": 2.5804,
+      "step": 11717
+    },
+    {
+      "epoch": 0.7208415354330708,
+      "grad_norm": 1.0030490159988403,
+      "learning_rate": 3.6095631245663006e-05,
+      "loss": 1.7758,
+      "step": 11718
+    },
+    {
+      "epoch": 0.7209030511811023,
+      "grad_norm": 1.0668634176254272,
+      "learning_rate": 3.6080758501139976e-05,
+      "loss": 3.1167,
+      "step": 11719
+    },
+    {
+      "epoch": 0.7209645669291339,
+      "grad_norm": 1.3600850105285645,
+      "learning_rate": 3.6065888146840124e-05,
+      "loss": 2.0308,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7210260826771654,
+      "grad_norm": 1.164121389389038,
+      "learning_rate": 3.605102018331957e-05,
+      "loss": 2.0879,
+      "step": 11721
+    },
+    {
+      "epoch": 0.7210875984251969,
+      "grad_norm": 1.4049639701843262,
+      "learning_rate": 3.603615461113425e-05,
+      "loss": 2.8016,
+      "step": 11722
+    },
+    {
+      "epoch": 0.7211491141732284,
+      "grad_norm": 1.3573439121246338,
+      "learning_rate": 3.602129143084008e-05,
+      "loss": 2.2287,
+      "step": 11723
+    },
+    {
+      "epoch": 0.7212106299212598,
+      "grad_norm": 1.303357720375061,
+      "learning_rate": 3.600643064299285e-05,
+      "loss": 2.4804,
+      "step": 11724
+    },
+    {
+      "epoch": 0.7212721456692913,
+      "grad_norm": 1.3328369855880737,
+      "learning_rate": 3.599157224814831e-05,
+      "loss": 2.1626,
+      "step": 11725
+    },
+    {
+      "epoch": 0.7213336614173228,
+      "grad_norm": 1.3646676540374756,
+      "learning_rate": 3.597671624686201e-05,
+      "loss": 2.4551,
+      "step": 11726
+    },
+    {
+      "epoch": 0.7213951771653543,
+      "grad_norm": 1.1923068761825562,
+      "learning_rate": 3.596186263968955e-05,
+      "loss": 2.5338,
+      "step": 11727
+    },
+    {
+      "epoch": 0.7214566929133859,
+      "grad_norm": 1.176573395729065,
+      "learning_rate": 3.5947011427186305e-05,
+      "loss": 2.8934,
+      "step": 11728
+    },
+    {
+      "epoch": 0.7215182086614174,
+      "grad_norm": 1.2065784931182861,
+      "learning_rate": 3.593216260990767e-05,
+      "loss": 2.7295,
+      "step": 11729
+    },
+    {
+      "epoch": 0.7215797244094488,
+      "grad_norm": 1.3020740747451782,
+      "learning_rate": 3.591731618840891e-05,
+      "loss": 2.4431,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7216412401574803,
+      "grad_norm": 1.4897356033325195,
+      "learning_rate": 3.590247216324522e-05,
+      "loss": 1.824,
+      "step": 11731
+    },
+    {
+      "epoch": 0.7217027559055118,
+      "grad_norm": 1.0682860612869263,
+      "learning_rate": 3.5887630534971604e-05,
+      "loss": 2.4282,
+      "step": 11732
+    },
+    {
+      "epoch": 0.7217642716535433,
+      "grad_norm": 1.4875530004501343,
+      "learning_rate": 3.5872791304143146e-05,
+      "loss": 2.1533,
+      "step": 11733
+    },
+    {
+      "epoch": 0.7218257874015748,
+      "grad_norm": 1.5022841691970825,
+      "learning_rate": 3.585795447131468e-05,
+      "loss": 2.6661,
+      "step": 11734
+    },
+    {
+      "epoch": 0.7218873031496063,
+      "grad_norm": 1.071591854095459,
+      "learning_rate": 3.584312003704109e-05,
+      "loss": 2.9861,
+      "step": 11735
+    },
+    {
+      "epoch": 0.7219488188976378,
+      "grad_norm": 1.3651520013809204,
+      "learning_rate": 3.582828800187702e-05,
+      "loss": 2.1562,
+      "step": 11736
+    },
+    {
+      "epoch": 0.7220103346456693,
+      "grad_norm": 1.3871190547943115,
+      "learning_rate": 3.581345836637715e-05,
+      "loss": 2.9587,
+      "step": 11737
+    },
+    {
+      "epoch": 0.7220718503937008,
+      "grad_norm": 1.3736779689788818,
+      "learning_rate": 3.579863113109606e-05,
+      "loss": 2.4926,
+      "step": 11738
+    },
+    {
+      "epoch": 0.7221333661417323,
+      "grad_norm": 1.2952133417129517,
+      "learning_rate": 3.578380629658813e-05,
+      "loss": 2.3625,
+      "step": 11739
+    },
+    {
+      "epoch": 0.7221948818897638,
+      "grad_norm": 1.3800122737884521,
+      "learning_rate": 3.5768983863407805e-05,
+      "loss": 2.5113,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7222563976377953,
+      "grad_norm": 1.0975048542022705,
+      "learning_rate": 3.575416383210928e-05,
+      "loss": 2.2664,
+      "step": 11741
+    },
+    {
+      "epoch": 0.7223179133858267,
+      "grad_norm": 1.3393220901489258,
+      "learning_rate": 3.573934620324678e-05,
+      "loss": 2.8994,
+      "step": 11742
+    },
+    {
+      "epoch": 0.7223794291338582,
+      "grad_norm": 1.131794810295105,
+      "learning_rate": 3.572453097737441e-05,
+      "loss": 2.9739,
+      "step": 11743
+    },
+    {
+      "epoch": 0.7224409448818898,
+      "grad_norm": 1.1704496145248413,
+      "learning_rate": 3.5709718155046214e-05,
+      "loss": 2.8199,
+      "step": 11744
+    },
+    {
+      "epoch": 0.7225024606299213,
+      "grad_norm": 1.4019544124603271,
+      "learning_rate": 3.569490773681603e-05,
+      "loss": 1.9546,
+      "step": 11745
+    },
+    {
+      "epoch": 0.7225639763779528,
+      "grad_norm": 1.5987391471862793,
+      "learning_rate": 3.568009972323775e-05,
+      "loss": 2.5761,
+      "step": 11746
+    },
+    {
+      "epoch": 0.7226254921259843,
+      "grad_norm": 1.1356744766235352,
+      "learning_rate": 3.566529411486504e-05,
+      "loss": 2.4301,
+      "step": 11747
+    },
+    {
+      "epoch": 0.7226870078740157,
+      "grad_norm": 1.420599102973938,
+      "learning_rate": 3.565049091225161e-05,
+      "loss": 2.3575,
+      "step": 11748
+    },
+    {
+      "epoch": 0.7227485236220472,
+      "grad_norm": 1.0645538568496704,
+      "learning_rate": 3.563569011595099e-05,
+      "loss": 2.8165,
+      "step": 11749
+    },
+    {
+      "epoch": 0.7228100393700787,
+      "grad_norm": 1.4148012399673462,
+      "learning_rate": 3.562089172651668e-05,
+      "loss": 2.071,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7228715551181102,
+      "grad_norm": 1.4034701585769653,
+      "learning_rate": 3.560609574450201e-05,
+      "loss": 2.0038,
+      "step": 11751
+    },
+    {
+      "epoch": 0.7229330708661418,
+      "grad_norm": 1.4347132444381714,
+      "learning_rate": 3.559130217046032e-05,
+      "loss": 2.4236,
+      "step": 11752
+    },
+    {
+      "epoch": 0.7229945866141733,
+      "grad_norm": 1.2813942432403564,
+      "learning_rate": 3.557651100494473e-05,
+      "loss": 2.8146,
+      "step": 11753
+    },
+    {
+      "epoch": 0.7230561023622047,
+      "grad_norm": 1.147791862487793,
+      "learning_rate": 3.55617222485084e-05,
+      "loss": 2.2532,
+      "step": 11754
+    },
+    {
+      "epoch": 0.7231176181102362,
+      "grad_norm": 1.20417058467865,
+      "learning_rate": 3.554693590170435e-05,
+      "loss": 2.1302,
+      "step": 11755
+    },
+    {
+      "epoch": 0.7231791338582677,
+      "grad_norm": 1.3217746019363403,
+      "learning_rate": 3.553215196508551e-05,
+      "loss": 3.1224,
+      "step": 11756
+    },
+    {
+      "epoch": 0.7232406496062992,
+      "grad_norm": 1.302856206893921,
+      "learning_rate": 3.551737043920469e-05,
+      "loss": 2.6701,
+      "step": 11757
+    },
+    {
+      "epoch": 0.7233021653543307,
+      "grad_norm": 1.3399449586868286,
+      "learning_rate": 3.5502591324614674e-05,
+      "loss": 2.1124,
+      "step": 11758
+    },
+    {
+      "epoch": 0.7233636811023622,
+      "grad_norm": 1.3461081981658936,
+      "learning_rate": 3.5487814621868064e-05,
+      "loss": 2.4803,
+      "step": 11759
+    },
+    {
+      "epoch": 0.7234251968503937,
+      "grad_norm": 1.2890937328338623,
+      "learning_rate": 3.5473040331517495e-05,
+      "loss": 2.4399,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7234867125984252,
+      "grad_norm": 1.243652582168579,
+      "learning_rate": 3.5458268454115374e-05,
+      "loss": 2.1319,
+      "step": 11761
+    },
+    {
+      "epoch": 0.7235482283464567,
+      "grad_norm": 1.6673349142074585,
+      "learning_rate": 3.544349899021411e-05,
+      "loss": 2.2436,
+      "step": 11762
+    },
+    {
+      "epoch": 0.7236097440944882,
+      "grad_norm": 1.3866935968399048,
+      "learning_rate": 3.542873194036601e-05,
+      "loss": 2.5829,
+      "step": 11763
+    },
+    {
+      "epoch": 0.7236712598425197,
+      "grad_norm": 1.301857590675354,
+      "learning_rate": 3.541396730512331e-05,
+      "loss": 2.3196,
+      "step": 11764
+    },
+    {
+      "epoch": 0.7237327755905512,
+      "grad_norm": 1.290812611579895,
+      "learning_rate": 3.5399205085038054e-05,
+      "loss": 1.979,
+      "step": 11765
+    },
+    {
+      "epoch": 0.7237942913385826,
+      "grad_norm": 1.6388111114501953,
+      "learning_rate": 3.538444528066234e-05,
+      "loss": 2.4459,
+      "step": 11766
+    },
+    {
+      "epoch": 0.7238558070866141,
+      "grad_norm": 1.1340117454528809,
+      "learning_rate": 3.536968789254803e-05,
+      "loss": 2.7865,
+      "step": 11767
+    },
+    {
+      "epoch": 0.7239173228346457,
+      "grad_norm": 1.460411548614502,
+      "learning_rate": 3.535493292124701e-05,
+      "loss": 3.3436,
+      "step": 11768
+    },
+    {
+      "epoch": 0.7239788385826772,
+      "grad_norm": 1.2965377569198608,
+      "learning_rate": 3.534018036731103e-05,
+      "loss": 2.5107,
+      "step": 11769
+    },
+    {
+      "epoch": 0.7240403543307087,
+      "grad_norm": 1.0903172492980957,
+      "learning_rate": 3.532543023129178e-05,
+      "loss": 2.6456,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7241018700787402,
+      "grad_norm": 1.454087495803833,
+      "learning_rate": 3.531068251374077e-05,
+      "loss": 2.3631,
+      "step": 11771
+    },
+    {
+      "epoch": 0.7241633858267716,
+      "grad_norm": 1.1947343349456787,
+      "learning_rate": 3.5295937215209554e-05,
+      "loss": 2.4605,
+      "step": 11772
+    },
+    {
+      "epoch": 0.7242249015748031,
+      "grad_norm": 1.3679181337356567,
+      "learning_rate": 3.528119433624945e-05,
+      "loss": 2.9036,
+      "step": 11773
+    },
+    {
+      "epoch": 0.7242864173228346,
+      "grad_norm": 1.389110803604126,
+      "learning_rate": 3.526645387741179e-05,
+      "loss": 2.266,
+      "step": 11774
+    },
+    {
+      "epoch": 0.7243479330708661,
+      "grad_norm": 1.1721547842025757,
+      "learning_rate": 3.5251715839247835e-05,
+      "loss": 2.7864,
+      "step": 11775
+    },
+    {
+      "epoch": 0.7244094488188977,
+      "grad_norm": 1.2596698999404907,
+      "learning_rate": 3.5236980222308615e-05,
+      "loss": 2.6616,
+      "step": 11776
+    },
+    {
+      "epoch": 0.7244709645669292,
+      "grad_norm": 1.1546409130096436,
+      "learning_rate": 3.522224702714525e-05,
+      "loss": 2.8381,
+      "step": 11777
+    },
+    {
+      "epoch": 0.7245324803149606,
+      "grad_norm": 1.1545964479446411,
+      "learning_rate": 3.520751625430861e-05,
+      "loss": 1.415,
+      "step": 11778
+    },
+    {
+      "epoch": 0.7245939960629921,
+      "grad_norm": 1.1218464374542236,
+      "learning_rate": 3.519278790434958e-05,
+      "loss": 2.7519,
+      "step": 11779
+    },
+    {
+      "epoch": 0.7246555118110236,
+      "grad_norm": 1.3647416830062866,
+      "learning_rate": 3.5178061977818854e-05,
+      "loss": 2.1736,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7247170275590551,
+      "grad_norm": 1.3437831401824951,
+      "learning_rate": 3.516333847526722e-05,
+      "loss": 2.3158,
+      "step": 11781
+    },
+    {
+      "epoch": 0.7247785433070866,
+      "grad_norm": 1.3177611827850342,
+      "learning_rate": 3.5148617397245155e-05,
+      "loss": 2.5851,
+      "step": 11782
+    },
+    {
+      "epoch": 0.7248400590551181,
+      "grad_norm": 1.5285824537277222,
+      "learning_rate": 3.5133898744303206e-05,
+      "loss": 3.4147,
+      "step": 11783
+    },
+    {
+      "epoch": 0.7249015748031497,
+      "grad_norm": 1.172063946723938,
+      "learning_rate": 3.5119182516991714e-05,
+      "loss": 2.4541,
+      "step": 11784
+    },
+    {
+      "epoch": 0.7249630905511811,
+      "grad_norm": 1.1563133001327515,
+      "learning_rate": 3.510446871586104e-05,
+      "loss": 2.8849,
+      "step": 11785
+    },
+    {
+      "epoch": 0.7250246062992126,
+      "grad_norm": 1.2307254076004028,
+      "learning_rate": 3.508975734146133e-05,
+      "loss": 2.551,
+      "step": 11786
+    },
+    {
+      "epoch": 0.7250861220472441,
+      "grad_norm": 1.1735150814056396,
+      "learning_rate": 3.507504839434276e-05,
+      "loss": 1.7965,
+      "step": 11787
+    },
+    {
+      "epoch": 0.7251476377952756,
+      "grad_norm": 1.3848451375961304,
+      "learning_rate": 3.506034187505534e-05,
+      "loss": 2.6572,
+      "step": 11788
+    },
+    {
+      "epoch": 0.7252091535433071,
+      "grad_norm": 1.2366960048675537,
+      "learning_rate": 3.504563778414904e-05,
+      "loss": 2.735,
+      "step": 11789
+    },
+    {
+      "epoch": 0.7252706692913385,
+      "grad_norm": 1.2331539392471313,
+      "learning_rate": 3.503093612217368e-05,
+      "loss": 2.7312,
+      "step": 11790
+    },
+    {
+      "epoch": 0.72533218503937,
+      "grad_norm": 1.3618950843811035,
+      "learning_rate": 3.5016236889679054e-05,
+      "loss": 2.9882,
+      "step": 11791
+    },
+    {
+      "epoch": 0.7253937007874016,
+      "grad_norm": 1.410597562789917,
+      "learning_rate": 3.500154008721478e-05,
+      "loss": 2.5097,
+      "step": 11792
+    },
+    {
+      "epoch": 0.7254552165354331,
+      "grad_norm": 1.2044261693954468,
+      "learning_rate": 3.498684571533046e-05,
+      "loss": 2.3489,
+      "step": 11793
+    },
+    {
+      "epoch": 0.7255167322834646,
+      "grad_norm": 1.2570551633834839,
+      "learning_rate": 3.49721537745756e-05,
+      "loss": 1.7755,
+      "step": 11794
+    },
+    {
+      "epoch": 0.7255782480314961,
+      "grad_norm": 1.329640507698059,
+      "learning_rate": 3.4957464265499604e-05,
+      "loss": 2.5945,
+      "step": 11795
+    },
+    {
+      "epoch": 0.7256397637795275,
+      "grad_norm": 1.245543122291565,
+      "learning_rate": 3.494277718865173e-05,
+      "loss": 2.8413,
+      "step": 11796
+    },
+    {
+      "epoch": 0.725701279527559,
+      "grad_norm": 1.360540747642517,
+      "learning_rate": 3.4928092544581246e-05,
+      "loss": 2.716,
+      "step": 11797
+    },
+    {
+      "epoch": 0.7257627952755905,
+      "grad_norm": 1.2113115787506104,
+      "learning_rate": 3.4913410333837224e-05,
+      "loss": 2.1535,
+      "step": 11798
+    },
+    {
+      "epoch": 0.725824311023622,
+      "grad_norm": 1.0818395614624023,
+      "learning_rate": 3.489873055696873e-05,
+      "loss": 2.289,
+      "step": 11799
+    },
+    {
+      "epoch": 0.7258858267716536,
+      "grad_norm": 1.2313308715820312,
+      "learning_rate": 3.488405321452468e-05,
+      "loss": 1.5995,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7259473425196851,
+      "grad_norm": 1.3132903575897217,
+      "learning_rate": 3.486937830705399e-05,
+      "loss": 2.3183,
+      "step": 11801
+    },
+    {
+      "epoch": 0.7260088582677166,
+      "grad_norm": 1.139052152633667,
+      "learning_rate": 3.485470583510534e-05,
+      "loss": 2.8071,
+      "step": 11802
+    },
+    {
+      "epoch": 0.726070374015748,
+      "grad_norm": 1.2952848672866821,
+      "learning_rate": 3.484003579922745e-05,
+      "loss": 2.2842,
+      "step": 11803
+    },
+    {
+      "epoch": 0.7261318897637795,
+      "grad_norm": 1.4194741249084473,
+      "learning_rate": 3.4825368199968866e-05,
+      "loss": 2.6518,
+      "step": 11804
+    },
+    {
+      "epoch": 0.726193405511811,
+      "grad_norm": 1.0732437372207642,
+      "learning_rate": 3.481070303787811e-05,
+      "loss": 1.8174,
+      "step": 11805
+    },
+    {
+      "epoch": 0.7262549212598425,
+      "grad_norm": 1.410650610923767,
+      "learning_rate": 3.479604031350349e-05,
+      "loss": 2.3586,
+      "step": 11806
+    },
+    {
+      "epoch": 0.726316437007874,
+      "grad_norm": 1.5178674459457397,
+      "learning_rate": 3.478138002739344e-05,
+      "loss": 3.0153,
+      "step": 11807
+    },
+    {
+      "epoch": 0.7263779527559056,
+      "grad_norm": 1.3632397651672363,
+      "learning_rate": 3.476672218009608e-05,
+      "loss": 2.1028,
+      "step": 11808
+    },
+    {
+      "epoch": 0.726439468503937,
+      "grad_norm": 1.3082128763198853,
+      "learning_rate": 3.475206677215959e-05,
+      "loss": 2.3928,
+      "step": 11809
+    },
+    {
+      "epoch": 0.7265009842519685,
+      "grad_norm": 1.6077502965927124,
+      "learning_rate": 3.473741380413197e-05,
+      "loss": 2.7237,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7265625,
+      "grad_norm": 1.1554794311523438,
+      "learning_rate": 3.472276327656112e-05,
+      "loss": 2.3714,
+      "step": 11811
+    },
+    {
+      "epoch": 0.7266240157480315,
+      "grad_norm": 1.274156928062439,
+      "learning_rate": 3.4708115189994926e-05,
+      "loss": 1.9932,
+      "step": 11812
+    },
+    {
+      "epoch": 0.726685531496063,
+      "grad_norm": 1.3066976070404053,
+      "learning_rate": 3.469346954498116e-05,
+      "loss": 2.4591,
+      "step": 11813
+    },
+    {
+      "epoch": 0.7267470472440944,
+      "grad_norm": 1.3176190853118896,
+      "learning_rate": 3.467882634206751e-05,
+      "loss": 2.0492,
+      "step": 11814
+    },
+    {
+      "epoch": 0.726808562992126,
+      "grad_norm": 1.3694310188293457,
+      "learning_rate": 3.4664185581801464e-05,
+      "loss": 2.6682,
+      "step": 11815
+    },
+    {
+      "epoch": 0.7268700787401575,
+      "grad_norm": 1.4658230543136597,
+      "learning_rate": 3.4649547264730596e-05,
+      "loss": 2.4086,
+      "step": 11816
+    },
+    {
+      "epoch": 0.726931594488189,
+      "grad_norm": 1.3221163749694824,
+      "learning_rate": 3.463491139140222e-05,
+      "loss": 2.6446,
+      "step": 11817
+    },
+    {
+      "epoch": 0.7269931102362205,
+      "grad_norm": 1.230334997177124,
+      "learning_rate": 3.462027796236368e-05,
+      "loss": 2.7616,
+      "step": 11818
+    },
+    {
+      "epoch": 0.727054625984252,
+      "grad_norm": 1.3783255815505981,
+      "learning_rate": 3.4605646978162164e-05,
+      "loss": 2.5968,
+      "step": 11819
+    },
+    {
+      "epoch": 0.7271161417322834,
+      "grad_norm": 1.2874590158462524,
+      "learning_rate": 3.459101843934485e-05,
+      "loss": 2.4343,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7271776574803149,
+      "grad_norm": 1.1671257019042969,
+      "learning_rate": 3.4576392346458676e-05,
+      "loss": 1.49,
+      "step": 11821
+    },
+    {
+      "epoch": 0.7272391732283464,
+      "grad_norm": 1.5604835748672485,
+      "learning_rate": 3.4561768700050644e-05,
+      "loss": 2.4387,
+      "step": 11822
+    },
+    {
+      "epoch": 0.727300688976378,
+      "grad_norm": 1.388006329536438,
+      "learning_rate": 3.454714750066754e-05,
+      "loss": 2.2798,
+      "step": 11823
+    },
+    {
+      "epoch": 0.7273622047244095,
+      "grad_norm": 1.322603464126587,
+      "learning_rate": 3.4532528748856176e-05,
+      "loss": 2.4681,
+      "step": 11824
+    },
+    {
+      "epoch": 0.727423720472441,
+      "grad_norm": 1.0927265882492065,
+      "learning_rate": 3.4517912445163115e-05,
+      "loss": 2.571,
+      "step": 11825
+    },
+    {
+      "epoch": 0.7274852362204725,
+      "grad_norm": 1.116620659828186,
+      "learning_rate": 3.450329859013506e-05,
+      "loss": 2.5466,
+      "step": 11826
+    },
+    {
+      "epoch": 0.7275467519685039,
+      "grad_norm": 1.3404793739318848,
+      "learning_rate": 3.448868718431838e-05,
+      "loss": 2.5123,
+      "step": 11827
+    },
+    {
+      "epoch": 0.7276082677165354,
+      "grad_norm": 1.2363288402557373,
+      "learning_rate": 3.4474078228259546e-05,
+      "loss": 2.5492,
+      "step": 11828
+    },
+    {
+      "epoch": 0.7276697834645669,
+      "grad_norm": 1.3532307147979736,
+      "learning_rate": 3.445947172250476e-05,
+      "loss": 2.4043,
+      "step": 11829
+    },
+    {
+      "epoch": 0.7277312992125984,
+      "grad_norm": 1.1407816410064697,
+      "learning_rate": 3.444486766760029e-05,
+      "loss": 2.9656,
+      "step": 11830
+    },
+    {
+      "epoch": 0.72779281496063,
+      "grad_norm": 1.5835058689117432,
+      "learning_rate": 3.44302660640922e-05,
+      "loss": 3.012,
+      "step": 11831
+    },
+    {
+      "epoch": 0.7278543307086615,
+      "grad_norm": 1.1210376024246216,
+      "learning_rate": 3.4415666912526534e-05,
+      "loss": 1.4638,
+      "step": 11832
+    },
+    {
+      "epoch": 0.7279158464566929,
+      "grad_norm": 1.2735034227371216,
+      "learning_rate": 3.4401070213449216e-05,
+      "loss": 2.2918,
+      "step": 11833
+    },
+    {
+      "epoch": 0.7279773622047244,
+      "grad_norm": 1.8206309080123901,
+      "learning_rate": 3.438647596740612e-05,
+      "loss": 1.9468,
+      "step": 11834
+    },
+    {
+      "epoch": 0.7280388779527559,
+      "grad_norm": 1.3515393733978271,
+      "learning_rate": 3.4371884174942903e-05,
+      "loss": 2.4751,
+      "step": 11835
+    },
+    {
+      "epoch": 0.7281003937007874,
+      "grad_norm": 1.1877667903900146,
+      "learning_rate": 3.4357294836605315e-05,
+      "loss": 3.3536,
+      "step": 11836
+    },
+    {
+      "epoch": 0.7281619094488189,
+      "grad_norm": 1.0401036739349365,
+      "learning_rate": 3.434270795293882e-05,
+      "loss": 2.7489,
+      "step": 11837
+    },
+    {
+      "epoch": 0.7282234251968503,
+      "grad_norm": 1.2418360710144043,
+      "learning_rate": 3.4328123524488944e-05,
+      "loss": 2.6613,
+      "step": 11838
+    },
+    {
+      "epoch": 0.7282849409448819,
+      "grad_norm": 1.2684520483016968,
+      "learning_rate": 3.431354155180105e-05,
+      "loss": 2.5836,
+      "step": 11839
+    },
+    {
+      "epoch": 0.7283464566929134,
+      "grad_norm": 1.4100667238235474,
+      "learning_rate": 3.429896203542046e-05,
+      "loss": 2.2364,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7284079724409449,
+      "grad_norm": 1.5065404176712036,
+      "learning_rate": 3.428438497589229e-05,
+      "loss": 2.0008,
+      "step": 11841
+    },
+    {
+      "epoch": 0.7284694881889764,
+      "grad_norm": 1.4909019470214844,
+      "learning_rate": 3.426981037376171e-05,
+      "loss": 2.6665,
+      "step": 11842
+    },
+    {
+      "epoch": 0.7285310039370079,
+      "grad_norm": 1.2278257608413696,
+      "learning_rate": 3.4255238229573685e-05,
+      "loss": 2.22,
+      "step": 11843
+    },
+    {
+      "epoch": 0.7285925196850394,
+      "grad_norm": 1.0035121440887451,
+      "learning_rate": 3.424066854387313e-05,
+      "loss": 2.5175,
+      "step": 11844
+    },
+    {
+      "epoch": 0.7286540354330708,
+      "grad_norm": 1.5547271966934204,
+      "learning_rate": 3.42261013172049e-05,
+      "loss": 1.7764,
+      "step": 11845
+    },
+    {
+      "epoch": 0.7287155511811023,
+      "grad_norm": 1.4335345029830933,
+      "learning_rate": 3.421153655011374e-05,
+      "loss": 0.3111,
+      "step": 11846
+    },
+    {
+      "epoch": 0.7287770669291339,
+      "grad_norm": 1.1689411401748657,
+      "learning_rate": 3.4196974243144264e-05,
+      "loss": 2.4131,
+      "step": 11847
+    },
+    {
+      "epoch": 0.7288385826771654,
+      "grad_norm": 1.3407446146011353,
+      "learning_rate": 3.4182414396841e-05,
+      "loss": 2.2147,
+      "step": 11848
+    },
+    {
+      "epoch": 0.7289000984251969,
+      "grad_norm": 1.225178837776184,
+      "learning_rate": 3.416785701174846e-05,
+      "loss": 2.1641,
+      "step": 11849
+    },
+    {
+      "epoch": 0.7289616141732284,
+      "grad_norm": 1.2177708148956299,
+      "learning_rate": 3.4153302088410944e-05,
+      "loss": 2.2214,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7290231299212598,
+      "grad_norm": 1.2921661138534546,
+      "learning_rate": 3.413874962737276e-05,
+      "loss": 2.1215,
+      "step": 11851
+    },
+    {
+      "epoch": 0.7290846456692913,
+      "grad_norm": 1.3320841789245605,
+      "learning_rate": 3.412419962917808e-05,
+      "loss": 2.2775,
+      "step": 11852
+    },
+    {
+      "epoch": 0.7291461614173228,
+      "grad_norm": 1.238175630569458,
+      "learning_rate": 3.4109652094371036e-05,
+      "loss": 2.3993,
+      "step": 11853
+    },
+    {
+      "epoch": 0.7292076771653543,
+      "grad_norm": 1.3642925024032593,
+      "learning_rate": 3.409510702349555e-05,
+      "loss": 2.3822,
+      "step": 11854
+    },
+    {
+      "epoch": 0.7292691929133859,
+      "grad_norm": 1.267783522605896,
+      "learning_rate": 3.408056441709562e-05,
+      "loss": 2.2234,
+      "step": 11855
+    },
+    {
+      "epoch": 0.7293307086614174,
+      "grad_norm": 1.2738475799560547,
+      "learning_rate": 3.4066024275714945e-05,
+      "loss": 2.5019,
+      "step": 11856
+    },
+    {
+      "epoch": 0.7293922244094488,
+      "grad_norm": 1.2559155225753784,
+      "learning_rate": 3.405148659989732e-05,
+      "loss": 2.3085,
+      "step": 11857
+    },
+    {
+      "epoch": 0.7294537401574803,
+      "grad_norm": 1.3619484901428223,
+      "learning_rate": 3.4036951390186356e-05,
+      "loss": 2.107,
+      "step": 11858
+    },
+    {
+      "epoch": 0.7295152559055118,
+      "grad_norm": 1.1234254837036133,
+      "learning_rate": 3.402241864712562e-05,
+      "loss": 2.3308,
+      "step": 11859
+    },
+    {
+      "epoch": 0.7295767716535433,
+      "grad_norm": 1.375422477722168,
+      "learning_rate": 3.40078883712585e-05,
+      "loss": 2.2364,
+      "step": 11860
+    },
+    {
+      "epoch": 0.7296382874015748,
+      "grad_norm": 1.0716291666030884,
+      "learning_rate": 3.39933605631284e-05,
+      "loss": 2.9414,
+      "step": 11861
+    },
+    {
+      "epoch": 0.7296998031496063,
+      "grad_norm": 1.3678841590881348,
+      "learning_rate": 3.397883522327852e-05,
+      "loss": 2.3351,
+      "step": 11862
+    },
+    {
+      "epoch": 0.7297613188976378,
+      "grad_norm": 1.3514962196350098,
+      "learning_rate": 3.396431235225205e-05,
+      "loss": 2.0304,
+      "step": 11863
+    },
+    {
+      "epoch": 0.7298228346456693,
+      "grad_norm": 1.3312469720840454,
+      "learning_rate": 3.394979195059208e-05,
+      "loss": 2.3856,
+      "step": 11864
+    },
+    {
+      "epoch": 0.7298843503937008,
+      "grad_norm": 1.2129018306732178,
+      "learning_rate": 3.393527401884161e-05,
+      "loss": 2.153,
+      "step": 11865
+    },
+    {
+      "epoch": 0.7299458661417323,
+      "grad_norm": 1.7026900053024292,
+      "learning_rate": 3.392075855754349e-05,
+      "loss": 2.1947,
+      "step": 11866
+    },
+    {
+      "epoch": 0.7300073818897638,
+      "grad_norm": 1.1991586685180664,
+      "learning_rate": 3.3906245567240543e-05,
+      "loss": 2.7067,
+      "step": 11867
+    },
+    {
+      "epoch": 0.7300688976377953,
+      "grad_norm": 1.1206928491592407,
+      "learning_rate": 3.389173504847545e-05,
+      "loss": 2.8174,
+      "step": 11868
+    },
+    {
+      "epoch": 0.7301304133858267,
+      "grad_norm": 1.3459813594818115,
+      "learning_rate": 3.3877227001790826e-05,
+      "loss": 2.7311,
+      "step": 11869
+    },
+    {
+      "epoch": 0.7301919291338582,
+      "grad_norm": 1.307350993156433,
+      "learning_rate": 3.38627214277292e-05,
+      "loss": 1.9344,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7302534448818898,
+      "grad_norm": 1.3744747638702393,
+      "learning_rate": 3.384821832683304e-05,
+      "loss": 2.2258,
+      "step": 11871
+    },
+    {
+      "epoch": 0.7303149606299213,
+      "grad_norm": 1.2625923156738281,
+      "learning_rate": 3.383371769964461e-05,
+      "loss": 2.2736,
+      "step": 11872
+    },
+    {
+      "epoch": 0.7303764763779528,
+      "grad_norm": 2.9754385948181152,
+      "learning_rate": 3.38192195467062e-05,
+      "loss": 2.273,
+      "step": 11873
+    },
+    {
+      "epoch": 0.7304379921259843,
+      "grad_norm": 1.3496522903442383,
+      "learning_rate": 3.380472386855993e-05,
+      "loss": 3.3115,
+      "step": 11874
+    },
+    {
+      "epoch": 0.7304995078740157,
+      "grad_norm": 1.1897042989730835,
+      "learning_rate": 3.37902306657479e-05,
+      "loss": 2.6234,
+      "step": 11875
+    },
+    {
+      "epoch": 0.7305610236220472,
+      "grad_norm": 1.2712377309799194,
+      "learning_rate": 3.377573993881201e-05,
+      "loss": 2.2047,
+      "step": 11876
+    },
+    {
+      "epoch": 0.7306225393700787,
+      "grad_norm": 1.2454420328140259,
+      "learning_rate": 3.376125168829418e-05,
+      "loss": 2.3449,
+      "step": 11877
+    },
+    {
+      "epoch": 0.7306840551181102,
+      "grad_norm": 1.3462640047073364,
+      "learning_rate": 3.374676591473618e-05,
+      "loss": 2.3527,
+      "step": 11878
+    },
+    {
+      "epoch": 0.7307455708661418,
+      "grad_norm": 1.177037000656128,
+      "learning_rate": 3.373228261867972e-05,
+      "loss": 1.9686,
+      "step": 11879
+    },
+    {
+      "epoch": 0.7308070866141733,
+      "grad_norm": 1.6600385904312134,
+      "learning_rate": 3.371780180066634e-05,
+      "loss": 2.0058,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7308686023622047,
+      "grad_norm": 1.1421992778778076,
+      "learning_rate": 3.370332346123761e-05,
+      "loss": 1.2936,
+      "step": 11881
+    },
+    {
+      "epoch": 0.7309301181102362,
+      "grad_norm": 1.251875877380371,
+      "learning_rate": 3.368884760093487e-05,
+      "loss": 2.4142,
+      "step": 11882
+    },
+    {
+      "epoch": 0.7309916338582677,
+      "grad_norm": 1.2569682598114014,
+      "learning_rate": 3.3674374220299466e-05,
+      "loss": 2.5073,
+      "step": 11883
+    },
+    {
+      "epoch": 0.7310531496062992,
+      "grad_norm": 1.119256615638733,
+      "learning_rate": 3.3659903319872656e-05,
+      "loss": 3.1785,
+      "step": 11884
+    },
+    {
+      "epoch": 0.7311146653543307,
+      "grad_norm": 1.3712078332901,
+      "learning_rate": 3.364543490019551e-05,
+      "loss": 2.3437,
+      "step": 11885
+    },
+    {
+      "epoch": 0.7311761811023622,
+      "grad_norm": 1.2211884260177612,
+      "learning_rate": 3.363096896180912e-05,
+      "loss": 2.6844,
+      "step": 11886
+    },
+    {
+      "epoch": 0.7312376968503937,
+      "grad_norm": 1.2017546892166138,
+      "learning_rate": 3.361650550525438e-05,
+      "loss": 2.0836,
+      "step": 11887
+    },
+    {
+      "epoch": 0.7312992125984252,
+      "grad_norm": 1.3599283695220947,
+      "learning_rate": 3.360204453107218e-05,
+      "loss": 2.1162,
+      "step": 11888
+    },
+    {
+      "epoch": 0.7313607283464567,
+      "grad_norm": 1.0491085052490234,
+      "learning_rate": 3.3587586039803256e-05,
+      "loss": 2.8452,
+      "step": 11889
+    },
+    {
+      "epoch": 0.7314222440944882,
+      "grad_norm": 1.1752591133117676,
+      "learning_rate": 3.357313003198832e-05,
+      "loss": 2.6444,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7314837598425197,
+      "grad_norm": 1.4511895179748535,
+      "learning_rate": 3.3558676508167885e-05,
+      "loss": 2.0213,
+      "step": 11891
+    },
+    {
+      "epoch": 0.7315452755905512,
+      "grad_norm": 1.401616096496582,
+      "learning_rate": 3.354422546888249e-05,
+      "loss": 2.4651,
+      "step": 11892
+    },
+    {
+      "epoch": 0.7316067913385826,
+      "grad_norm": 1.4942244291305542,
+      "learning_rate": 3.352977691467247e-05,
+      "loss": 2.2161,
+      "step": 11893
+    },
+    {
+      "epoch": 0.7316683070866141,
+      "grad_norm": 1.221798300743103,
+      "learning_rate": 3.3515330846078175e-05,
+      "loss": 2.5853,
+      "step": 11894
+    },
+    {
+      "epoch": 0.7317298228346457,
+      "grad_norm": 1.1444816589355469,
+      "learning_rate": 3.350088726363971e-05,
+      "loss": 2.3971,
+      "step": 11895
+    },
+    {
+      "epoch": 0.7317913385826772,
+      "grad_norm": 1.154004454612732,
+      "learning_rate": 3.348644616789732e-05,
+      "loss": 2.9546,
+      "step": 11896
+    },
+    {
+      "epoch": 0.7318528543307087,
+      "grad_norm": 1.202808141708374,
+      "learning_rate": 3.3472007559390926e-05,
+      "loss": 1.8189,
+      "step": 11897
+    },
+    {
+      "epoch": 0.7319143700787402,
+      "grad_norm": 1.1081815958023071,
+      "learning_rate": 3.3457571438660505e-05,
+      "loss": 2.6444,
+      "step": 11898
+    },
+    {
+      "epoch": 0.7319758858267716,
+      "grad_norm": 1.9659062623977661,
+      "learning_rate": 3.344313780624583e-05,
+      "loss": 1.8914,
+      "step": 11899
+    },
+    {
+      "epoch": 0.7320374015748031,
+      "grad_norm": 1.4177637100219727,
+      "learning_rate": 3.3428706662686706e-05,
+      "loss": 2.6633,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7320989173228346,
+      "grad_norm": 1.4269458055496216,
+      "learning_rate": 3.3414278008522706e-05,
+      "loss": 1.6755,
+      "step": 11901
+    },
+    {
+      "epoch": 0.7321604330708661,
+      "grad_norm": 1.2679647207260132,
+      "learning_rate": 3.33998518442934e-05,
+      "loss": 2.9949,
+      "step": 11902
+    },
+    {
+      "epoch": 0.7322219488188977,
+      "grad_norm": 1.4194835424423218,
+      "learning_rate": 3.338542817053829e-05,
+      "loss": 2.4966,
+      "step": 11903
+    },
+    {
+      "epoch": 0.7322834645669292,
+      "grad_norm": 1.1309475898742676,
+      "learning_rate": 3.337100698779674e-05,
+      "loss": 2.6615,
+      "step": 11904
+    },
+    {
+      "epoch": 0.7323449803149606,
+      "grad_norm": 1.413487434387207,
+      "learning_rate": 3.335658829660795e-05,
+      "loss": 1.871,
+      "step": 11905
+    },
+    {
+      "epoch": 0.7324064960629921,
+      "grad_norm": 1.214284062385559,
+      "learning_rate": 3.334217209751118e-05,
+      "loss": 2.5944,
+      "step": 11906
+    },
+    {
+      "epoch": 0.7324680118110236,
+      "grad_norm": 1.1562731266021729,
+      "learning_rate": 3.332775839104545e-05,
+      "loss": 1.8436,
+      "step": 11907
+    },
+    {
+      "epoch": 0.7325295275590551,
+      "grad_norm": 1.2491925954818726,
+      "learning_rate": 3.331334717774979e-05,
+      "loss": 2.4533,
+      "step": 11908
+    },
+    {
+      "epoch": 0.7325910433070866,
+      "grad_norm": 1.1226376295089722,
+      "learning_rate": 3.3298938458163073e-05,
+      "loss": 2.9537,
+      "step": 11909
+    },
+    {
+      "epoch": 0.7326525590551181,
+      "grad_norm": 1.3607547283172607,
+      "learning_rate": 3.328453223282416e-05,
+      "loss": 2.5231,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7327140748031497,
+      "grad_norm": 1.1941211223602295,
+      "learning_rate": 3.32701285022717e-05,
+      "loss": 2.4691,
+      "step": 11911
+    },
+    {
+      "epoch": 0.7327755905511811,
+      "grad_norm": 1.160727620124817,
+      "learning_rate": 3.325572726704437e-05,
+      "loss": 2.6415,
+      "step": 11912
+    },
+    {
+      "epoch": 0.7328371062992126,
+      "grad_norm": 1.2110838890075684,
+      "learning_rate": 3.324132852768063e-05,
+      "loss": 2.7126,
+      "step": 11913
+    },
+    {
+      "epoch": 0.7328986220472441,
+      "grad_norm": 1.1305994987487793,
+      "learning_rate": 3.322693228471895e-05,
+      "loss": 2.2162,
+      "step": 11914
+    },
+    {
+      "epoch": 0.7329601377952756,
+      "grad_norm": 1.4227879047393799,
+      "learning_rate": 3.321253853869766e-05,
+      "loss": 2.8941,
+      "step": 11915
+    },
+    {
+      "epoch": 0.7330216535433071,
+      "grad_norm": 1.3113723993301392,
+      "learning_rate": 3.3198147290155044e-05,
+      "loss": 2.1096,
+      "step": 11916
+    },
+    {
+      "epoch": 0.7330831692913385,
+      "grad_norm": 1.398732304573059,
+      "learning_rate": 3.3183758539629184e-05,
+      "loss": 2.646,
+      "step": 11917
+    },
+    {
+      "epoch": 0.73314468503937,
+      "grad_norm": 1.2225292921066284,
+      "learning_rate": 3.316937228765822e-05,
+      "loss": 1.8828,
+      "step": 11918
+    },
+    {
+      "epoch": 0.7332062007874016,
+      "grad_norm": 1.2312171459197998,
+      "learning_rate": 3.3154988534780076e-05,
+      "loss": 2.0617,
+      "step": 11919
+    },
+    {
+      "epoch": 0.7332677165354331,
+      "grad_norm": 1.3349649906158447,
+      "learning_rate": 3.3140607281532574e-05,
+      "loss": 1.5851,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7333292322834646,
+      "grad_norm": 1.1890990734100342,
+      "learning_rate": 3.312622852845356e-05,
+      "loss": 3.0197,
+      "step": 11921
+    },
+    {
+      "epoch": 0.7333907480314961,
+      "grad_norm": 1.2687405347824097,
+      "learning_rate": 3.311185227608069e-05,
+      "loss": 2.6235,
+      "step": 11922
+    },
+    {
+      "epoch": 0.7334522637795275,
+      "grad_norm": 1.2408859729766846,
+      "learning_rate": 3.309747852495161e-05,
+      "loss": 3.1195,
+      "step": 11923
+    },
+    {
+      "epoch": 0.733513779527559,
+      "grad_norm": 2.5663342475891113,
+      "learning_rate": 3.308310727560373e-05,
+      "loss": 2.3001,
+      "step": 11924
+    },
+    {
+      "epoch": 0.7335752952755905,
+      "grad_norm": 1.7501299381256104,
+      "learning_rate": 3.306873852857454e-05,
+      "loss": 1.9655,
+      "step": 11925
+    },
+    {
+      "epoch": 0.733636811023622,
+      "grad_norm": 1.3660250902175903,
+      "learning_rate": 3.305437228440127e-05,
+      "loss": 2.5931,
+      "step": 11926
+    },
+    {
+      "epoch": 0.7336983267716536,
+      "grad_norm": 1.2977477312088013,
+      "learning_rate": 3.304000854362118e-05,
+      "loss": 2.549,
+      "step": 11927
+    },
+    {
+      "epoch": 0.7337598425196851,
+      "grad_norm": 1.3749216794967651,
+      "learning_rate": 3.302564730677139e-05,
+      "loss": 2.3293,
+      "step": 11928
+    },
+    {
+      "epoch": 0.7338213582677166,
+      "grad_norm": 1.1386386156082153,
+      "learning_rate": 3.3011288574388966e-05,
+      "loss": 2.6358,
+      "step": 11929
+    },
+    {
+      "epoch": 0.733882874015748,
+      "grad_norm": 1.1909444332122803,
+      "learning_rate": 3.299693234701078e-05,
+      "loss": 2.8636,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7339443897637795,
+      "grad_norm": 1.1767821311950684,
+      "learning_rate": 3.298257862517373e-05,
+      "loss": 2.4361,
+      "step": 11931
+    },
+    {
+      "epoch": 0.734005905511811,
+      "grad_norm": 1.464532732963562,
+      "learning_rate": 3.296822740941452e-05,
+      "loss": 2.13,
+      "step": 11932
+    },
+    {
+      "epoch": 0.7340674212598425,
+      "grad_norm": 1.6738370656967163,
+      "learning_rate": 3.2953878700269815e-05,
+      "loss": 1.3183,
+      "step": 11933
+    },
+    {
+      "epoch": 0.734128937007874,
+      "grad_norm": 1.7223058938980103,
+      "learning_rate": 3.293953249827619e-05,
+      "loss": 2.1369,
+      "step": 11934
+    },
+    {
+      "epoch": 0.7341904527559056,
+      "grad_norm": 1.3046752214431763,
+      "learning_rate": 3.292518880397014e-05,
+      "loss": 2.3721,
+      "step": 11935
+    },
+    {
+      "epoch": 0.734251968503937,
+      "grad_norm": 1.2947990894317627,
+      "learning_rate": 3.291084761788798e-05,
+      "loss": 2.4211,
+      "step": 11936
+    },
+    {
+      "epoch": 0.7343134842519685,
+      "grad_norm": 1.2917044162750244,
+      "learning_rate": 3.289650894056605e-05,
+      "loss": 1.9546,
+      "step": 11937
+    },
+    {
+      "epoch": 0.734375,
+      "grad_norm": 1.312216877937317,
+      "learning_rate": 3.2882172772540484e-05,
+      "loss": 2.2524,
+      "step": 11938
+    },
+    {
+      "epoch": 0.7344365157480315,
+      "grad_norm": 1.437898874282837,
+      "learning_rate": 3.2867839114347385e-05,
+      "loss": 2.5169,
+      "step": 11939
+    },
+    {
+      "epoch": 0.734498031496063,
+      "grad_norm": 1.6197056770324707,
+      "learning_rate": 3.2853507966522766e-05,
+      "loss": 1.5912,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7345595472440944,
+      "grad_norm": 1.067899227142334,
+      "learning_rate": 3.283917932960257e-05,
+      "loss": 2.55,
+      "step": 11941
+    },
+    {
+      "epoch": 0.734621062992126,
+      "grad_norm": 1.8140100240707397,
+      "learning_rate": 3.282485320412253e-05,
+      "loss": 1.9888,
+      "step": 11942
+    },
+    {
+      "epoch": 0.7346825787401575,
+      "grad_norm": 1.3311716318130493,
+      "learning_rate": 3.2810529590618435e-05,
+      "loss": 2.9949,
+      "step": 11943
+    },
+    {
+      "epoch": 0.734744094488189,
+      "grad_norm": 1.3622338771820068,
+      "learning_rate": 3.2796208489625835e-05,
+      "loss": 2.4859,
+      "step": 11944
+    },
+    {
+      "epoch": 0.7348056102362205,
+      "grad_norm": 1.2859363555908203,
+      "learning_rate": 3.2781889901680344e-05,
+      "loss": 2.5923,
+      "step": 11945
+    },
+    {
+      "epoch": 0.734867125984252,
+      "grad_norm": 1.2313098907470703,
+      "learning_rate": 3.276757382731731e-05,
+      "loss": 2.5803,
+      "step": 11946
+    },
+    {
+      "epoch": 0.7349286417322834,
+      "grad_norm": 1.0303536653518677,
+      "learning_rate": 3.275326026707212e-05,
+      "loss": 2.5796,
+      "step": 11947
+    },
+    {
+      "epoch": 0.7349901574803149,
+      "grad_norm": 1.3519495725631714,
+      "learning_rate": 3.2738949221480007e-05,
+      "loss": 2.1145,
+      "step": 11948
+    },
+    {
+      "epoch": 0.7350516732283464,
+      "grad_norm": 1.644269347190857,
+      "learning_rate": 3.272464069107618e-05,
+      "loss": 2.3666,
+      "step": 11949
+    },
+    {
+      "epoch": 0.735113188976378,
+      "grad_norm": 1.1893062591552734,
+      "learning_rate": 3.271033467639563e-05,
+      "loss": 2.2021,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7351747047244095,
+      "grad_norm": 1.374372124671936,
+      "learning_rate": 3.2696031177973364e-05,
+      "loss": 2.1123,
+      "step": 11951
+    },
+    {
+      "epoch": 0.735236220472441,
+      "grad_norm": 1.6465411186218262,
+      "learning_rate": 3.268173019634421e-05,
+      "loss": 2.3494,
+      "step": 11952
+    },
+    {
+      "epoch": 0.7352977362204725,
+      "grad_norm": 1.3777672052383423,
+      "learning_rate": 3.266743173204295e-05,
+      "loss": 2.7793,
+      "step": 11953
+    },
+    {
+      "epoch": 0.7353592519685039,
+      "grad_norm": 1.181043028831482,
+      "learning_rate": 3.2653135785604305e-05,
+      "loss": 2.3117,
+      "step": 11954
+    },
+    {
+      "epoch": 0.7354207677165354,
+      "grad_norm": 1.493939995765686,
+      "learning_rate": 3.263884235756288e-05,
+      "loss": 2.5164,
+      "step": 11955
+    },
+    {
+      "epoch": 0.7354822834645669,
+      "grad_norm": 1.4274671077728271,
+      "learning_rate": 3.262455144845312e-05,
+      "loss": 3.3146,
+      "step": 11956
+    },
+    {
+      "epoch": 0.7355437992125984,
+      "grad_norm": 1.1522557735443115,
+      "learning_rate": 3.2610263058809424e-05,
+      "loss": 2.5782,
+      "step": 11957
+    },
+    {
+      "epoch": 0.73560531496063,
+      "grad_norm": 1.01256263256073,
+      "learning_rate": 3.259597718916611e-05,
+      "loss": 2.1916,
+      "step": 11958
+    },
+    {
+      "epoch": 0.7356668307086615,
+      "grad_norm": 1.4355789422988892,
+      "learning_rate": 3.258169384005738e-05,
+      "loss": 2.406,
+      "step": 11959
+    },
+    {
+      "epoch": 0.7357283464566929,
+      "grad_norm": 1.330152988433838,
+      "learning_rate": 3.256741301201741e-05,
+      "loss": 2.2709,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7357898622047244,
+      "grad_norm": 1.3647537231445312,
+      "learning_rate": 3.2553134705580144e-05,
+      "loss": 2.3165,
+      "step": 11961
+    },
+    {
+      "epoch": 0.7358513779527559,
+      "grad_norm": 1.148589849472046,
+      "learning_rate": 3.253885892127958e-05,
+      "loss": 2.607,
+      "step": 11962
+    },
+    {
+      "epoch": 0.7359128937007874,
+      "grad_norm": 1.271702527999878,
+      "learning_rate": 3.252458565964949e-05,
+      "loss": 3.2351,
+      "step": 11963
+    },
+    {
+      "epoch": 0.7359744094488189,
+      "grad_norm": 1.3790823221206665,
+      "learning_rate": 3.251031492122367e-05,
+      "loss": 2.0265,
+      "step": 11964
+    },
+    {
+      "epoch": 0.7360359251968503,
+      "grad_norm": 1.4471570253372192,
+      "learning_rate": 3.2496046706535675e-05,
+      "loss": 2.5793,
+      "step": 11965
+    },
+    {
+      "epoch": 0.7360974409448819,
+      "grad_norm": 1.3770700693130493,
+      "learning_rate": 3.2481781016119185e-05,
+      "loss": 2.3924,
+      "step": 11966
+    },
+    {
+      "epoch": 0.7361589566929134,
+      "grad_norm": 1.3881632089614868,
+      "learning_rate": 3.2467517850507565e-05,
+      "loss": 3.0308,
+      "step": 11967
+    },
+    {
+      "epoch": 0.7362204724409449,
+      "grad_norm": 1.2460942268371582,
+      "learning_rate": 3.2453257210234234e-05,
+      "loss": 1.9554,
+      "step": 11968
+    },
+    {
+      "epoch": 0.7362819881889764,
+      "grad_norm": 1.1106305122375488,
+      "learning_rate": 3.24389990958324e-05,
+      "loss": 2.5566,
+      "step": 11969
+    },
+    {
+      "epoch": 0.7363435039370079,
+      "grad_norm": 1.7026475667953491,
+      "learning_rate": 3.242474350783531e-05,
+      "loss": 2.3929,
+      "step": 11970
+    },
+    {
+      "epoch": 0.7364050196850394,
+      "grad_norm": 1.1165045499801636,
+      "learning_rate": 3.2410490446775954e-05,
+      "loss": 2.6792,
+      "step": 11971
+    },
+    {
+      "epoch": 0.7364665354330708,
+      "grad_norm": 1.1326860189437866,
+      "learning_rate": 3.239623991318738e-05,
+      "loss": 3.0802,
+      "step": 11972
+    },
+    {
+      "epoch": 0.7365280511811023,
+      "grad_norm": 1.4783470630645752,
+      "learning_rate": 3.238199190760246e-05,
+      "loss": 2.5427,
+      "step": 11973
+    },
+    {
+      "epoch": 0.7365895669291339,
+      "grad_norm": 1.5552659034729004,
+      "learning_rate": 3.236774643055403e-05,
+      "loss": 2.5186,
+      "step": 11974
+    },
+    {
+      "epoch": 0.7366510826771654,
+      "grad_norm": 1.0885107517242432,
+      "learning_rate": 3.235350348257471e-05,
+      "loss": 2.4992,
+      "step": 11975
+    },
+    {
+      "epoch": 0.7367125984251969,
+      "grad_norm": 1.2793525457382202,
+      "learning_rate": 3.23392630641972e-05,
+      "loss": 2.0114,
+      "step": 11976
+    },
+    {
+      "epoch": 0.7367741141732284,
+      "grad_norm": 1.514946699142456,
+      "learning_rate": 3.2325025175953925e-05,
+      "loss": 2.9527,
+      "step": 11977
+    },
+    {
+      "epoch": 0.7368356299212598,
+      "grad_norm": 1.3508167266845703,
+      "learning_rate": 3.231078981837734e-05,
+      "loss": 3.375,
+      "step": 11978
+    },
+    {
+      "epoch": 0.7368971456692913,
+      "grad_norm": 1.6033669710159302,
+      "learning_rate": 3.229655699199978e-05,
+      "loss": 1.813,
+      "step": 11979
+    },
+    {
+      "epoch": 0.7369586614173228,
+      "grad_norm": 1.261840581893921,
+      "learning_rate": 3.2282326697353496e-05,
+      "loss": 2.7351,
+      "step": 11980
+    },
+    {
+      "epoch": 0.7370201771653543,
+      "grad_norm": 1.367807388305664,
+      "learning_rate": 3.226809893497055e-05,
+      "loss": 2.6689,
+      "step": 11981
+    },
+    {
+      "epoch": 0.7370816929133859,
+      "grad_norm": 1.216030478477478,
+      "learning_rate": 3.225387370538305e-05,
+      "loss": 2.674,
+      "step": 11982
+    },
+    {
+      "epoch": 0.7371432086614174,
+      "grad_norm": 1.4151339530944824,
+      "learning_rate": 3.2239651009122885e-05,
+      "loss": 2.5815,
+      "step": 11983
+    },
+    {
+      "epoch": 0.7372047244094488,
+      "grad_norm": 1.1124818325042725,
+      "learning_rate": 3.2225430846721924e-05,
+      "loss": 2.6045,
+      "step": 11984
+    },
+    {
+      "epoch": 0.7372662401574803,
+      "grad_norm": 1.3834439516067505,
+      "learning_rate": 3.221121321871193e-05,
+      "loss": 2.2414,
+      "step": 11985
+    },
+    {
+      "epoch": 0.7373277559055118,
+      "grad_norm": 1.3170793056488037,
+      "learning_rate": 3.21969981256246e-05,
+      "loss": 1.8829,
+      "step": 11986
+    },
+    {
+      "epoch": 0.7373892716535433,
+      "grad_norm": 1.2752559185028076,
+      "learning_rate": 3.2182785567991426e-05,
+      "loss": 2.3074,
+      "step": 11987
+    },
+    {
+      "epoch": 0.7374507874015748,
+      "grad_norm": 1.215622067451477,
+      "learning_rate": 3.2168575546343936e-05,
+      "loss": 2.4131,
+      "step": 11988
+    },
+    {
+      "epoch": 0.7375123031496063,
+      "grad_norm": 1.3014047145843506,
+      "learning_rate": 3.2154368061213456e-05,
+      "loss": 2.8032,
+      "step": 11989
+    },
+    {
+      "epoch": 0.7375738188976378,
+      "grad_norm": 1.5451740026474,
+      "learning_rate": 3.2140163113131326e-05,
+      "loss": 2.518,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7376353346456693,
+      "grad_norm": 1.2590343952178955,
+      "learning_rate": 3.212596070262863e-05,
+      "loss": 3.3792,
+      "step": 11991
+    },
+    {
+      "epoch": 0.7376968503937008,
+      "grad_norm": 1.2206352949142456,
+      "learning_rate": 3.2111760830236606e-05,
+      "loss": 2.3231,
+      "step": 11992
+    },
+    {
+      "epoch": 0.7377583661417323,
+      "grad_norm": 1.2387523651123047,
+      "learning_rate": 3.209756349648616e-05,
+      "loss": 2.6178,
+      "step": 11993
+    },
+    {
+      "epoch": 0.7378198818897638,
+      "grad_norm": 1.163020372390747,
+      "learning_rate": 3.208336870190818e-05,
+      "loss": 2.424,
+      "step": 11994
+    },
+    {
+      "epoch": 0.7378813976377953,
+      "grad_norm": 1.137170433998108,
+      "learning_rate": 3.206917644703352e-05,
+      "loss": 2.1975,
+      "step": 11995
+    },
+    {
+      "epoch": 0.7379429133858267,
+      "grad_norm": 1.34241783618927,
+      "learning_rate": 3.205498673239283e-05,
+      "loss": 2.3861,
+      "step": 11996
+    },
+    {
+      "epoch": 0.7380044291338582,
+      "grad_norm": 1.4110126495361328,
+      "learning_rate": 3.204079955851678e-05,
+      "loss": 2.793,
+      "step": 11997
+    },
+    {
+      "epoch": 0.7380659448818898,
+      "grad_norm": 1.2855457067489624,
+      "learning_rate": 3.202661492593587e-05,
+      "loss": 2.6347,
+      "step": 11998
+    },
+    {
+      "epoch": 0.7381274606299213,
+      "grad_norm": 1.653258204460144,
+      "learning_rate": 3.2012432835180564e-05,
+      "loss": 2.1382,
+      "step": 11999
+    },
+    {
+      "epoch": 0.7381889763779528,
+      "grad_norm": 1.2025396823883057,
+      "learning_rate": 3.1998253286781124e-05,
+      "loss": 2.5214,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7382504921259843,
+      "grad_norm": 1.3120582103729248,
+      "learning_rate": 3.198407628126786e-05,
+      "loss": 2.3505,
+      "step": 12001
+    },
+    {
+      "epoch": 0.7383120078740157,
+      "grad_norm": 1.274434208869934,
+      "learning_rate": 3.1969901819170845e-05,
+      "loss": 2.5004,
+      "step": 12002
+    },
+    {
+      "epoch": 0.7383735236220472,
+      "grad_norm": 1.2838214635849,
+      "learning_rate": 3.1955729901020146e-05,
+      "loss": 2.0967,
+      "step": 12003
+    },
+    {
+      "epoch": 0.7384350393700787,
+      "grad_norm": 2.1306629180908203,
+      "learning_rate": 3.1941560527345735e-05,
+      "loss": 2.6812,
+      "step": 12004
+    },
+    {
+      "epoch": 0.7384965551181102,
+      "grad_norm": 1.233364224433899,
+      "learning_rate": 3.192739369867749e-05,
+      "loss": 0.3707,
+      "step": 12005
+    },
+    {
+      "epoch": 0.7385580708661418,
+      "grad_norm": 1.5203951597213745,
+      "learning_rate": 3.19132294155451e-05,
+      "loss": 2.8078,
+      "step": 12006
+    },
+    {
+      "epoch": 0.7386195866141733,
+      "grad_norm": 1.420128345489502,
+      "learning_rate": 3.18990676784783e-05,
+      "loss": 1.9802,
+      "step": 12007
+    },
+    {
+      "epoch": 0.7386811023622047,
+      "grad_norm": 1.4092605113983154,
+      "learning_rate": 3.188490848800659e-05,
+      "loss": 2.6896,
+      "step": 12008
+    },
+    {
+      "epoch": 0.7387426181102362,
+      "grad_norm": 1.3199554681777954,
+      "learning_rate": 3.187075184465952e-05,
+      "loss": 3.0581,
+      "step": 12009
+    },
+    {
+      "epoch": 0.7388041338582677,
+      "grad_norm": 1.4309242963790894,
+      "learning_rate": 3.185659774896637e-05,
+      "loss": 2.4541,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7388656496062992,
+      "grad_norm": 1.4412243366241455,
+      "learning_rate": 3.1842446201456545e-05,
+      "loss": 2.5411,
+      "step": 12011
+    },
+    {
+      "epoch": 0.7389271653543307,
+      "grad_norm": 1.3360707759857178,
+      "learning_rate": 3.182829720265914e-05,
+      "loss": 1.9841,
+      "step": 12012
+    },
+    {
+      "epoch": 0.7389886811023622,
+      "grad_norm": 1.4353846311569214,
+      "learning_rate": 3.1814150753103323e-05,
+      "loss": 2.2473,
+      "step": 12013
+    },
+    {
+      "epoch": 0.7390501968503937,
+      "grad_norm": 1.1643097400665283,
+      "learning_rate": 3.1800006853318023e-05,
+      "loss": 2.8066,
+      "step": 12014
+    },
+    {
+      "epoch": 0.7391117125984252,
+      "grad_norm": 1.1517003774642944,
+      "learning_rate": 3.17858655038322e-05,
+      "loss": 2.8753,
+      "step": 12015
+    },
+    {
+      "epoch": 0.7391732283464567,
+      "grad_norm": 1.1340795755386353,
+      "learning_rate": 3.177172670517461e-05,
+      "loss": 2.9791,
+      "step": 12016
+    },
+    {
+      "epoch": 0.7392347440944882,
+      "grad_norm": 1.2147998809814453,
+      "learning_rate": 3.1757590457873975e-05,
+      "loss": 2.408,
+      "step": 12017
+    },
+    {
+      "epoch": 0.7392962598425197,
+      "grad_norm": 1.47923743724823,
+      "learning_rate": 3.1743456762458935e-05,
+      "loss": 2.7444,
+      "step": 12018
+    },
+    {
+      "epoch": 0.7393577755905512,
+      "grad_norm": 1.1364686489105225,
+      "learning_rate": 3.172932561945804e-05,
+      "loss": 2.7782,
+      "step": 12019
+    },
+    {
+      "epoch": 0.7394192913385826,
+      "grad_norm": 1.1875269412994385,
+      "learning_rate": 3.171519702939965e-05,
+      "loss": 2.4423,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7394808070866141,
+      "grad_norm": 1.2356469631195068,
+      "learning_rate": 3.170107099281214e-05,
+      "loss": 2.4433,
+      "step": 12021
+    },
+    {
+      "epoch": 0.7395423228346457,
+      "grad_norm": 1.1915510892868042,
+      "learning_rate": 3.168694751022371e-05,
+      "loss": 2.5221,
+      "step": 12022
+    },
+    {
+      "epoch": 0.7396038385826772,
+      "grad_norm": 1.3115324974060059,
+      "learning_rate": 3.1672826582162516e-05,
+      "loss": 1.8168,
+      "step": 12023
+    },
+    {
+      "epoch": 0.7396653543307087,
+      "grad_norm": 1.474535346031189,
+      "learning_rate": 3.165870820915661e-05,
+      "loss": 2.3369,
+      "step": 12024
+    },
+    {
+      "epoch": 0.7397268700787402,
+      "grad_norm": 1.1048924922943115,
+      "learning_rate": 3.1644592391733964e-05,
+      "loss": 2.6074,
+      "step": 12025
+    },
+    {
+      "epoch": 0.7397883858267716,
+      "grad_norm": 1.3457329273223877,
+      "learning_rate": 3.163047913042236e-05,
+      "loss": 2.2167,
+      "step": 12026
+    },
+    {
+      "epoch": 0.7398499015748031,
+      "grad_norm": 1.404428243637085,
+      "learning_rate": 3.161636842574964e-05,
+      "loss": 2.1962,
+      "step": 12027
+    },
+    {
+      "epoch": 0.7399114173228346,
+      "grad_norm": 1.1719242334365845,
+      "learning_rate": 3.160226027824339e-05,
+      "loss": 2.9608,
+      "step": 12028
+    },
+    {
+      "epoch": 0.7399729330708661,
+      "grad_norm": 1.147246241569519,
+      "learning_rate": 3.1588154688431216e-05,
+      "loss": 2.219,
+      "step": 12029
+    },
+    {
+      "epoch": 0.7400344488188977,
+      "grad_norm": 1.4012198448181152,
+      "learning_rate": 3.1574051656840595e-05,
+      "loss": 1.9034,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7400959645669292,
+      "grad_norm": 1.330908179283142,
+      "learning_rate": 3.155995118399888e-05,
+      "loss": 2.2808,
+      "step": 12031
+    },
+    {
+      "epoch": 0.7401574803149606,
+      "grad_norm": 1.1598882675170898,
+      "learning_rate": 3.154585327043338e-05,
+      "loss": 2.7234,
+      "step": 12032
+    },
+    {
+      "epoch": 0.7402189960629921,
+      "grad_norm": 1.234269142150879,
+      "learning_rate": 3.153175791667122e-05,
+      "loss": 2.7427,
+      "step": 12033
+    },
+    {
+      "epoch": 0.7402805118110236,
+      "grad_norm": 1.3319603204727173,
+      "learning_rate": 3.151766512323956e-05,
+      "loss": 2.7153,
+      "step": 12034
+    },
+    {
+      "epoch": 0.7403420275590551,
+      "grad_norm": 1.1731311082839966,
+      "learning_rate": 3.150357489066532e-05,
+      "loss": 2.4677,
+      "step": 12035
+    },
+    {
+      "epoch": 0.7404035433070866,
+      "grad_norm": 1.332895278930664,
+      "learning_rate": 3.148948721947545e-05,
+      "loss": 3.0218,
+      "step": 12036
+    },
+    {
+      "epoch": 0.7404650590551181,
+      "grad_norm": 1.232611894607544,
+      "learning_rate": 3.1475402110196715e-05,
+      "loss": 1.8421,
+      "step": 12037
+    },
+    {
+      "epoch": 0.7405265748031497,
+      "grad_norm": 1.3988744020462036,
+      "learning_rate": 3.1461319563355876e-05,
+      "loss": 2.4823,
+      "step": 12038
+    },
+    {
+      "epoch": 0.7405880905511811,
+      "grad_norm": 1.1435611248016357,
+      "learning_rate": 3.144723957947947e-05,
+      "loss": 2.6299,
+      "step": 12039
+    },
+    {
+      "epoch": 0.7406496062992126,
+      "grad_norm": 1.2949559688568115,
+      "learning_rate": 3.143316215909408e-05,
+      "loss": 1.9825,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7407111220472441,
+      "grad_norm": 1.3249008655548096,
+      "learning_rate": 3.141908730272605e-05,
+      "loss": 2.2368,
+      "step": 12041
+    },
+    {
+      "epoch": 0.7407726377952756,
+      "grad_norm": 1.1867393255233765,
+      "learning_rate": 3.140501501090173e-05,
+      "loss": 2.4066,
+      "step": 12042
+    },
+    {
+      "epoch": 0.7408341535433071,
+      "grad_norm": 2.445821523666382,
+      "learning_rate": 3.1390945284147354e-05,
+      "loss": 2.4259,
+      "step": 12043
+    },
+    {
+      "epoch": 0.7408956692913385,
+      "grad_norm": 1.2987943887710571,
+      "learning_rate": 3.13768781229891e-05,
+      "loss": 1.984,
+      "step": 12044
+    },
+    {
+      "epoch": 0.74095718503937,
+      "grad_norm": 1.4713129997253418,
+      "learning_rate": 3.136281352795289e-05,
+      "loss": 2.7295,
+      "step": 12045
+    },
+    {
+      "epoch": 0.7410187007874016,
+      "grad_norm": 1.2950202226638794,
+      "learning_rate": 3.134875149956478e-05,
+      "loss": 2.6624,
+      "step": 12046
+    },
+    {
+      "epoch": 0.7410802165354331,
+      "grad_norm": 1.1923394203186035,
+      "learning_rate": 3.133469203835051e-05,
+      "loss": 2.1991,
+      "step": 12047
+    },
+    {
+      "epoch": 0.7411417322834646,
+      "grad_norm": 1.4190202951431274,
+      "learning_rate": 3.132063514483587e-05,
+      "loss": 2.6523,
+      "step": 12048
+    },
+    {
+      "epoch": 0.7412032480314961,
+      "grad_norm": 1.2241793870925903,
+      "learning_rate": 3.130658081954651e-05,
+      "loss": 2.6638,
+      "step": 12049
+    },
+    {
+      "epoch": 0.7412647637795275,
+      "grad_norm": 1.1679461002349854,
+      "learning_rate": 3.1292529063008015e-05,
+      "loss": 2.553,
+      "step": 12050
+    },
+    {
+      "epoch": 0.741326279527559,
+      "grad_norm": 1.2998133897781372,
+      "learning_rate": 3.127847987574578e-05,
+      "loss": 2.4934,
+      "step": 12051
+    },
+    {
+      "epoch": 0.7413877952755905,
+      "grad_norm": 1.1690130233764648,
+      "learning_rate": 3.126443325828523e-05,
+      "loss": 2.5601,
+      "step": 12052
+    },
+    {
+      "epoch": 0.741449311023622,
+      "grad_norm": 1.560801386833191,
+      "learning_rate": 3.1250389211151556e-05,
+      "loss": 2.087,
+      "step": 12053
+    },
+    {
+      "epoch": 0.7415108267716536,
+      "grad_norm": 1.2630293369293213,
+      "learning_rate": 3.1236347734869976e-05,
+      "loss": 1.9877,
+      "step": 12054
+    },
+    {
+      "epoch": 0.7415723425196851,
+      "grad_norm": 1.292263388633728,
+      "learning_rate": 3.122230882996554e-05,
+      "loss": 2.0871,
+      "step": 12055
+    },
+    {
+      "epoch": 0.7416338582677166,
+      "grad_norm": 1.2207120656967163,
+      "learning_rate": 3.1208272496963284e-05,
+      "loss": 2.1808,
+      "step": 12056
+    },
+    {
+      "epoch": 0.741695374015748,
+      "grad_norm": 1.380871057510376,
+      "learning_rate": 3.1194238736388006e-05,
+      "loss": 2.5437,
+      "step": 12057
+    },
+    {
+      "epoch": 0.7417568897637795,
+      "grad_norm": 1.186340570449829,
+      "learning_rate": 3.118020754876456e-05,
+      "loss": 2.8821,
+      "step": 12058
+    },
+    {
+      "epoch": 0.741818405511811,
+      "grad_norm": 1.1960570812225342,
+      "learning_rate": 3.116617893461758e-05,
+      "loss": 2.8341,
+      "step": 12059
+    },
+    {
+      "epoch": 0.7418799212598425,
+      "grad_norm": 1.6343187093734741,
+      "learning_rate": 3.11521528944717e-05,
+      "loss": 1.6248,
+      "step": 12060
+    },
+    {
+      "epoch": 0.741941437007874,
+      "grad_norm": 1.4457907676696777,
+      "learning_rate": 3.113812942885137e-05,
+      "loss": 2.6396,
+      "step": 12061
+    },
+    {
+      "epoch": 0.7420029527559056,
+      "grad_norm": 1.020368218421936,
+      "learning_rate": 3.1124108538281006e-05,
+      "loss": 2.7855,
+      "step": 12062
+    },
+    {
+      "epoch": 0.742064468503937,
+      "grad_norm": 1.2890223264694214,
+      "learning_rate": 3.1110090223284924e-05,
+      "loss": 2.4933,
+      "step": 12063
+    },
+    {
+      "epoch": 0.7421259842519685,
+      "grad_norm": 1.332316517829895,
+      "learning_rate": 3.109607448438736e-05,
+      "loss": 2.2302,
+      "step": 12064
+    },
+    {
+      "epoch": 0.7421875,
+      "grad_norm": 1.3884661197662354,
+      "learning_rate": 3.1082061322112364e-05,
+      "loss": 2.9091,
+      "step": 12065
+    },
+    {
+      "epoch": 0.7422490157480315,
+      "grad_norm": 1.162190556526184,
+      "learning_rate": 3.106805073698401e-05,
+      "loss": 2.9835,
+      "step": 12066
+    },
+    {
+      "epoch": 0.742310531496063,
+      "grad_norm": 1.4122551679611206,
+      "learning_rate": 3.1054042729526146e-05,
+      "loss": 2.5521,
+      "step": 12067
+    },
+    {
+      "epoch": 0.7423720472440944,
+      "grad_norm": 1.3018922805786133,
+      "learning_rate": 3.104003730026264e-05,
+      "loss": 2.2967,
+      "step": 12068
+    },
+    {
+      "epoch": 0.742433562992126,
+      "grad_norm": 2.1664538383483887,
+      "learning_rate": 3.1026034449717226e-05,
+      "loss": 1.8714,
+      "step": 12069
+    },
+    {
+      "epoch": 0.7424950787401575,
+      "grad_norm": 1.63238525390625,
+      "learning_rate": 3.1012034178413484e-05,
+      "loss": 1.9559,
+      "step": 12070
+    },
+    {
+      "epoch": 0.742556594488189,
+      "grad_norm": 1.4016382694244385,
+      "learning_rate": 3.0998036486875015e-05,
+      "loss": 2.4658,
+      "step": 12071
+    },
+    {
+      "epoch": 0.7426181102362205,
+      "grad_norm": 1.3074501752853394,
+      "learning_rate": 3.098404137562518e-05,
+      "loss": 2.0604,
+      "step": 12072
+    },
+    {
+      "epoch": 0.742679625984252,
+      "grad_norm": 1.1733782291412354,
+      "learning_rate": 3.097004884518736e-05,
+      "loss": 2.855,
+      "step": 12073
+    },
+    {
+      "epoch": 0.7427411417322834,
+      "grad_norm": 1.374462366104126,
+      "learning_rate": 3.0956058896084786e-05,
+      "loss": 1.7533,
+      "step": 12074
+    },
+    {
+      "epoch": 0.7428026574803149,
+      "grad_norm": 1.3904799222946167,
+      "learning_rate": 3.094207152884065e-05,
+      "loss": 1.7062,
+      "step": 12075
+    },
+    {
+      "epoch": 0.7428641732283464,
+      "grad_norm": 1.144176959991455,
+      "learning_rate": 3.092808674397791e-05,
+      "loss": 2.9139,
+      "step": 12076
+    },
+    {
+      "epoch": 0.742925688976378,
+      "grad_norm": 1.361730933189392,
+      "learning_rate": 3.0914104542019616e-05,
+      "loss": 2.3188,
+      "step": 12077
+    },
+    {
+      "epoch": 0.7429872047244095,
+      "grad_norm": 1.2844016551971436,
+      "learning_rate": 3.090012492348855e-05,
+      "loss": 2.5989,
+      "step": 12078
+    },
+    {
+      "epoch": 0.743048720472441,
+      "grad_norm": 1.3131269216537476,
+      "learning_rate": 3.088614788890752e-05,
+      "loss": 2.2145,
+      "step": 12079
+    },
+    {
+      "epoch": 0.7431102362204725,
+      "grad_norm": 1.2367439270019531,
+      "learning_rate": 3.0872173438799115e-05,
+      "loss": 2.5062,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7431717519685039,
+      "grad_norm": 1.2449818849563599,
+      "learning_rate": 3.085820157368602e-05,
+      "loss": 2.5383,
+      "step": 12081
+    },
+    {
+      "epoch": 0.7432332677165354,
+      "grad_norm": 1.3998771905899048,
+      "learning_rate": 3.0844232294090604e-05,
+      "loss": 2.0039,
+      "step": 12082
+    },
+    {
+      "epoch": 0.7432947834645669,
+      "grad_norm": 1.3212970495224,
+      "learning_rate": 3.0830265600535314e-05,
+      "loss": 2.7381,
+      "step": 12083
+    },
+    {
+      "epoch": 0.7433562992125984,
+      "grad_norm": 1.2298836708068848,
+      "learning_rate": 3.0816301493542355e-05,
+      "loss": 2.8504,
+      "step": 12084
+    },
+    {
+      "epoch": 0.74341781496063,
+      "grad_norm": 1.3363972902297974,
+      "learning_rate": 3.0802339973633976e-05,
+      "loss": 2.4948,
+      "step": 12085
+    },
+    {
+      "epoch": 0.7434793307086615,
+      "grad_norm": 1.2720152139663696,
+      "learning_rate": 3.078838104133218e-05,
+      "loss": 2.26,
+      "step": 12086
+    },
+    {
+      "epoch": 0.7435408464566929,
+      "grad_norm": 1.1738191843032837,
+      "learning_rate": 3.077442469715901e-05,
+      "loss": 2.3382,
+      "step": 12087
+    },
+    {
+      "epoch": 0.7436023622047244,
+      "grad_norm": 1.2439279556274414,
+      "learning_rate": 3.0760470941636345e-05,
+      "loss": 3.0878,
+      "step": 12088
+    },
+    {
+      "epoch": 0.7436638779527559,
+      "grad_norm": 1.2042397260665894,
+      "learning_rate": 3.074651977528601e-05,
+      "loss": 2.7174,
+      "step": 12089
+    },
+    {
+      "epoch": 0.7437253937007874,
+      "grad_norm": 1.2152615785598755,
+      "learning_rate": 3.073257119862963e-05,
+      "loss": 2.4591,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7437869094488189,
+      "grad_norm": 1.3277435302734375,
+      "learning_rate": 3.071862521218888e-05,
+      "loss": 2.4443,
+      "step": 12091
+    },
+    {
+      "epoch": 0.7438484251968503,
+      "grad_norm": 1.3157694339752197,
+      "learning_rate": 3.0704681816485184e-05,
+      "loss": 2.9378,
+      "step": 12092
+    },
+    {
+      "epoch": 0.7439099409448819,
+      "grad_norm": 1.1577956676483154,
+      "learning_rate": 3.069074101203999e-05,
+      "loss": 2.0311,
+      "step": 12093
+    },
+    {
+      "epoch": 0.7439714566929134,
+      "grad_norm": 1.0327965021133423,
+      "learning_rate": 3.067680279937461e-05,
+      "loss": 2.0554,
+      "step": 12094
+    },
+    {
+      "epoch": 0.7440329724409449,
+      "grad_norm": 1.4998440742492676,
+      "learning_rate": 3.0662867179010266e-05,
+      "loss": 2.5783,
+      "step": 12095
+    },
+    {
+      "epoch": 0.7440944881889764,
+      "grad_norm": 1.1651467084884644,
+      "learning_rate": 3.0648934151468034e-05,
+      "loss": 2.5864,
+      "step": 12096
+    },
+    {
+      "epoch": 0.7441560039370079,
+      "grad_norm": 1.2508670091629028,
+      "learning_rate": 3.0635003717268986e-05,
+      "loss": 2.1229,
+      "step": 12097
+    },
+    {
+      "epoch": 0.7442175196850394,
+      "grad_norm": 1.2969597578048706,
+      "learning_rate": 3.062107587693397e-05,
+      "loss": 2.336,
+      "step": 12098
+    },
+    {
+      "epoch": 0.7442790354330708,
+      "grad_norm": 1.194684386253357,
+      "learning_rate": 3.060715063098385e-05,
+      "loss": 2.1663,
+      "step": 12099
+    },
+    {
+      "epoch": 0.7443405511811023,
+      "grad_norm": 1.330588936805725,
+      "learning_rate": 3.059322797993935e-05,
+      "loss": 2.5866,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7444020669291339,
+      "grad_norm": 1.2779790163040161,
+      "learning_rate": 3.057930792432113e-05,
+      "loss": 2.2023,
+      "step": 12101
+    },
+    {
+      "epoch": 0.7444635826771654,
+      "grad_norm": 1.2358951568603516,
+      "learning_rate": 3.056539046464967e-05,
+      "loss": 2.3473,
+      "step": 12102
+    },
+    {
+      "epoch": 0.7445250984251969,
+      "grad_norm": 1.2052713632583618,
+      "learning_rate": 3.055147560144545e-05,
+      "loss": 2.3489,
+      "step": 12103
+    },
+    {
+      "epoch": 0.7445866141732284,
+      "grad_norm": 1.8419594764709473,
+      "learning_rate": 3.05375633352288e-05,
+      "loss": 1.9864,
+      "step": 12104
+    },
+    {
+      "epoch": 0.7446481299212598,
+      "grad_norm": 1.3473279476165771,
+      "learning_rate": 3.052365366651991e-05,
+      "loss": 2.4587,
+      "step": 12105
+    },
+    {
+      "epoch": 0.7447096456692913,
+      "grad_norm": 1.291125774383545,
+      "learning_rate": 3.0509746595838963e-05,
+      "loss": 2.5936,
+      "step": 12106
+    },
+    {
+      "epoch": 0.7447711614173228,
+      "grad_norm": 1.3877842426300049,
+      "learning_rate": 3.049584212370601e-05,
+      "loss": 3.5382,
+      "step": 12107
+    },
+    {
+      "epoch": 0.7448326771653543,
+      "grad_norm": 1.0595825910568237,
+      "learning_rate": 3.0481940250641016e-05,
+      "loss": 2.8543,
+      "step": 12108
+    },
+    {
+      "epoch": 0.7448941929133859,
+      "grad_norm": 1.3978735208511353,
+      "learning_rate": 3.046804097716379e-05,
+      "loss": 2.0116,
+      "step": 12109
+    },
+    {
+      "epoch": 0.7449557086614174,
+      "grad_norm": 1.400151252746582,
+      "learning_rate": 3.045414430379414e-05,
+      "loss": 2.9086,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7450172244094488,
+      "grad_norm": 1.1348668336868286,
+      "learning_rate": 3.044025023105166e-05,
+      "loss": 2.7526,
+      "step": 12111
+    },
+    {
+      "epoch": 0.7450787401574803,
+      "grad_norm": 1.1841189861297607,
+      "learning_rate": 3.0426358759455943e-05,
+      "loss": 2.5152,
+      "step": 12112
+    },
+    {
+      "epoch": 0.7451402559055118,
+      "grad_norm": 1.2008134126663208,
+      "learning_rate": 3.041246988952646e-05,
+      "loss": 2.8371,
+      "step": 12113
+    },
+    {
+      "epoch": 0.7452017716535433,
+      "grad_norm": 1.2083262205123901,
+      "learning_rate": 3.0398583621782605e-05,
+      "loss": 2.8492,
+      "step": 12114
+    },
+    {
+      "epoch": 0.7452632874015748,
+      "grad_norm": 1.662874460220337,
+      "learning_rate": 3.038469995674359e-05,
+      "loss": 2.3798,
+      "step": 12115
+    },
+    {
+      "epoch": 0.7453248031496063,
+      "grad_norm": 2.13716197013855,
+      "learning_rate": 3.0370818894928632e-05,
+      "loss": 0.7793,
+      "step": 12116
+    },
+    {
+      "epoch": 0.7453863188976378,
+      "grad_norm": 1.2519826889038086,
+      "learning_rate": 3.035694043685675e-05,
+      "loss": 2.8724,
+      "step": 12117
+    },
+    {
+      "epoch": 0.7454478346456693,
+      "grad_norm": 1.5728609561920166,
+      "learning_rate": 3.0343064583046966e-05,
+      "loss": 2.8918,
+      "step": 12118
+    },
+    {
+      "epoch": 0.7455093503937008,
+      "grad_norm": 1.3232778310775757,
+      "learning_rate": 3.032919133401815e-05,
+      "loss": 2.4075,
+      "step": 12119
+    },
+    {
+      "epoch": 0.7455708661417323,
+      "grad_norm": 1.4790136814117432,
+      "learning_rate": 3.0315320690289106e-05,
+      "loss": 2.5002,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7456323818897638,
+      "grad_norm": 1.3567655086517334,
+      "learning_rate": 3.030145265237846e-05,
+      "loss": 2.4578,
+      "step": 12121
+    },
+    {
+      "epoch": 0.7456938976377953,
+      "grad_norm": 1.2847551107406616,
+      "learning_rate": 3.028758722080488e-05,
+      "loss": 2.427,
+      "step": 12122
+    },
+    {
+      "epoch": 0.7457554133858267,
+      "grad_norm": 1.2162450551986694,
+      "learning_rate": 3.0273724396086765e-05,
+      "loss": 3.1107,
+      "step": 12123
+    },
+    {
+      "epoch": 0.7458169291338582,
+      "grad_norm": 1.4985822439193726,
+      "learning_rate": 3.0259864178742593e-05,
+      "loss": 3.2377,
+      "step": 12124
+    },
+    {
+      "epoch": 0.7458784448818898,
+      "grad_norm": 1.567840337753296,
+      "learning_rate": 3.0246006569290554e-05,
+      "loss": 2.255,
+      "step": 12125
+    },
+    {
+      "epoch": 0.7459399606299213,
+      "grad_norm": 0.9719260334968567,
+      "learning_rate": 3.0232151568248977e-05,
+      "loss": 2.8772,
+      "step": 12126
+    },
+    {
+      "epoch": 0.7460014763779528,
+      "grad_norm": 1.0704666376113892,
+      "learning_rate": 3.0218299176135865e-05,
+      "loss": 2.0536,
+      "step": 12127
+    },
+    {
+      "epoch": 0.7460629921259843,
+      "grad_norm": 1.3368271589279175,
+      "learning_rate": 3.020444939346928e-05,
+      "loss": 2.0352,
+      "step": 12128
+    },
+    {
+      "epoch": 0.7461245078740157,
+      "grad_norm": 1.4131321907043457,
+      "learning_rate": 3.019060222076706e-05,
+      "loss": 2.1726,
+      "step": 12129
+    },
+    {
+      "epoch": 0.7461860236220472,
+      "grad_norm": 1.0800281763076782,
+      "learning_rate": 3.0176757658547084e-05,
+      "loss": 2.5066,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7462475393700787,
+      "grad_norm": 1.4624783992767334,
+      "learning_rate": 3.0162915707327e-05,
+      "loss": 2.4695,
+      "step": 12131
+    },
+    {
+      "epoch": 0.7463090551181102,
+      "grad_norm": 1.3322432041168213,
+      "learning_rate": 3.014907636762444e-05,
+      "loss": 2.7888,
+      "step": 12132
+    },
+    {
+      "epoch": 0.7463705708661418,
+      "grad_norm": 1.367374300956726,
+      "learning_rate": 3.0135239639956925e-05,
+      "loss": 2.2621,
+      "step": 12133
+    },
+    {
+      "epoch": 0.7464320866141733,
+      "grad_norm": 1.4627711772918701,
+      "learning_rate": 3.0121405524841905e-05,
+      "loss": 2.0863,
+      "step": 12134
+    },
+    {
+      "epoch": 0.7464936023622047,
+      "grad_norm": 1.1717560291290283,
+      "learning_rate": 3.0107574022796626e-05,
+      "loss": 2.7982,
+      "step": 12135
+    },
+    {
+      "epoch": 0.7465551181102362,
+      "grad_norm": 1.2802815437316895,
+      "learning_rate": 3.009374513433839e-05,
+      "loss": 2.5118,
+      "step": 12136
+    },
+    {
+      "epoch": 0.7466166338582677,
+      "grad_norm": 1.5940032005310059,
+      "learning_rate": 3.0079918859984236e-05,
+      "loss": 2.4934,
+      "step": 12137
+    },
+    {
+      "epoch": 0.7466781496062992,
+      "grad_norm": 1.176928997039795,
+      "learning_rate": 3.0066095200251243e-05,
+      "loss": 2.1233,
+      "step": 12138
+    },
+    {
+      "epoch": 0.7467396653543307,
+      "grad_norm": 1.5546274185180664,
+      "learning_rate": 3.0052274155656313e-05,
+      "loss": 2.3501,
+      "step": 12139
+    },
+    {
+      "epoch": 0.7468011811023622,
+      "grad_norm": 1.290088415145874,
+      "learning_rate": 3.003845572671633e-05,
+      "loss": 2.4666,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7468626968503937,
+      "grad_norm": 1.422597885131836,
+      "learning_rate": 3.0024639913947995e-05,
+      "loss": 2.551,
+      "step": 12141
+    },
+    {
+      "epoch": 0.7469242125984252,
+      "grad_norm": 1.257097840309143,
+      "learning_rate": 3.001082671786789e-05,
+      "loss": 3.0366,
+      "step": 12142
+    },
+    {
+      "epoch": 0.7469857283464567,
+      "grad_norm": 1.1140332221984863,
+      "learning_rate": 2.9997016138992596e-05,
+      "loss": 2.1405,
+      "step": 12143
+    },
+    {
+      "epoch": 0.7470472440944882,
+      "grad_norm": 1.452077031135559,
+      "learning_rate": 2.998320817783855e-05,
+      "loss": 2.2206,
+      "step": 12144
+    },
+    {
+      "epoch": 0.7471087598425197,
+      "grad_norm": 1.1249759197235107,
+      "learning_rate": 2.9969402834922135e-05,
+      "loss": 2.6277,
+      "step": 12145
+    },
+    {
+      "epoch": 0.7471702755905512,
+      "grad_norm": 1.3058544397354126,
+      "learning_rate": 2.9955600110759517e-05,
+      "loss": 2.3367,
+      "step": 12146
+    },
+    {
+      "epoch": 0.7472317913385826,
+      "grad_norm": 1.261332392692566,
+      "learning_rate": 2.9941800005866904e-05,
+      "loss": 2.1461,
+      "step": 12147
+    },
+    {
+      "epoch": 0.7472933070866141,
+      "grad_norm": 1.0800762176513672,
+      "learning_rate": 2.992800252076029e-05,
+      "loss": 2.5115,
+      "step": 12148
+    },
+    {
+      "epoch": 0.7473548228346457,
+      "grad_norm": 1.4616767168045044,
+      "learning_rate": 2.9914207655955672e-05,
+      "loss": 2.3276,
+      "step": 12149
+    },
+    {
+      "epoch": 0.7474163385826772,
+      "grad_norm": 1.4262237548828125,
+      "learning_rate": 2.9900415411968864e-05,
+      "loss": 2.4325,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7474778543307087,
+      "grad_norm": 1.076545000076294,
+      "learning_rate": 2.9886625789315615e-05,
+      "loss": 2.0884,
+      "step": 12151
+    },
+    {
+      "epoch": 0.7475393700787402,
+      "grad_norm": 1.8959076404571533,
+      "learning_rate": 2.9872838788511614e-05,
+      "loss": 3.2388,
+      "step": 12152
+    },
+    {
+      "epoch": 0.7476008858267716,
+      "grad_norm": 1.527564287185669,
+      "learning_rate": 2.985905441007242e-05,
+      "loss": 2.309,
+      "step": 12153
+    },
+    {
+      "epoch": 0.7476624015748031,
+      "grad_norm": 1.239639163017273,
+      "learning_rate": 2.9845272654513455e-05,
+      "loss": 2.3654,
+      "step": 12154
+    },
+    {
+      "epoch": 0.7477239173228346,
+      "grad_norm": 1.3235551118850708,
+      "learning_rate": 2.983149352235012e-05,
+      "loss": 1.7318,
+      "step": 12155
+    },
+    {
+      "epoch": 0.7477854330708661,
+      "grad_norm": 1.1277711391448975,
+      "learning_rate": 2.9817717014097634e-05,
+      "loss": 2.0332,
+      "step": 12156
+    },
+    {
+      "epoch": 0.7478469488188977,
+      "grad_norm": 1.2229974269866943,
+      "learning_rate": 2.9803943130271182e-05,
+      "loss": 2.3324,
+      "step": 12157
+    },
+    {
+      "epoch": 0.7479084645669292,
+      "grad_norm": 1.444467306137085,
+      "learning_rate": 2.979017187138583e-05,
+      "loss": 2.5432,
+      "step": 12158
+    },
+    {
+      "epoch": 0.7479699803149606,
+      "grad_norm": 1.3530080318450928,
+      "learning_rate": 2.9776403237956586e-05,
+      "loss": 3.1181,
+      "step": 12159
+    },
+    {
+      "epoch": 0.7480314960629921,
+      "grad_norm": 1.095489263534546,
+      "learning_rate": 2.9762637230498248e-05,
+      "loss": 2.3387,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7480930118110236,
+      "grad_norm": 1.170441746711731,
+      "learning_rate": 2.974887384952565e-05,
+      "loss": 2.4595,
+      "step": 12161
+    },
+    {
+      "epoch": 0.7481545275590551,
+      "grad_norm": 1.0662544965744019,
+      "learning_rate": 2.9735113095553412e-05,
+      "loss": 2.4819,
+      "step": 12162
+    },
+    {
+      "epoch": 0.7482160433070866,
+      "grad_norm": 1.512778401374817,
+      "learning_rate": 2.972135496909614e-05,
+      "loss": 2.7184,
+      "step": 12163
+    },
+    {
+      "epoch": 0.7482775590551181,
+      "grad_norm": 1.160326600074768,
+      "learning_rate": 2.97075994706683e-05,
+      "loss": 2.7068,
+      "step": 12164
+    },
+    {
+      "epoch": 0.7483390748031497,
+      "grad_norm": 1.7742239236831665,
+      "learning_rate": 2.9693846600784324e-05,
+      "loss": 1.7768,
+      "step": 12165
+    },
+    {
+      "epoch": 0.7484005905511811,
+      "grad_norm": 1.2121423482894897,
+      "learning_rate": 2.9680096359958397e-05,
+      "loss": 1.7836,
+      "step": 12166
+    },
+    {
+      "epoch": 0.7484621062992126,
+      "grad_norm": 1.2967588901519775,
+      "learning_rate": 2.9666348748704798e-05,
+      "loss": 2.9687,
+      "step": 12167
+    },
+    {
+      "epoch": 0.7485236220472441,
+      "grad_norm": 1.3774651288986206,
+      "learning_rate": 2.9652603767537536e-05,
+      "loss": 2.6335,
+      "step": 12168
+    },
+    {
+      "epoch": 0.7485851377952756,
+      "grad_norm": 1.4340474605560303,
+      "learning_rate": 2.9638861416970608e-05,
+      "loss": 2.3121,
+      "step": 12169
+    },
+    {
+      "epoch": 0.7486466535433071,
+      "grad_norm": 1.2312376499176025,
+      "learning_rate": 2.9625121697517933e-05,
+      "loss": 2.409,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7487081692913385,
+      "grad_norm": 1.2365868091583252,
+      "learning_rate": 2.9611384609693304e-05,
+      "loss": 2.4084,
+      "step": 12171
+    },
+    {
+      "epoch": 0.74876968503937,
+      "grad_norm": 1.2462691068649292,
+      "learning_rate": 2.959765015401037e-05,
+      "loss": 1.8872,
+      "step": 12172
+    },
+    {
+      "epoch": 0.7488312007874016,
+      "grad_norm": 1.20871102809906,
+      "learning_rate": 2.9583918330982785e-05,
+      "loss": 3.0169,
+      "step": 12173
+    },
+    {
+      "epoch": 0.7488927165354331,
+      "grad_norm": 1.2128801345825195,
+      "learning_rate": 2.957018914112396e-05,
+      "loss": 2.185,
+      "step": 12174
+    },
+    {
+      "epoch": 0.7489542322834646,
+      "grad_norm": 1.4017871618270874,
+      "learning_rate": 2.9556462584947374e-05,
+      "loss": 2.5577,
+      "step": 12175
+    },
+    {
+      "epoch": 0.7490157480314961,
+      "grad_norm": 1.2775802612304688,
+      "learning_rate": 2.9542738662966242e-05,
+      "loss": 2.3149,
+      "step": 12176
+    },
+    {
+      "epoch": 0.7490772637795275,
+      "grad_norm": 2.1313135623931885,
+      "learning_rate": 2.952901737569381e-05,
+      "loss": 2.0786,
+      "step": 12177
+    },
+    {
+      "epoch": 0.749138779527559,
+      "grad_norm": 1.218780279159546,
+      "learning_rate": 2.9515298723643204e-05,
+      "loss": 3.0645,
+      "step": 12178
+    },
+    {
+      "epoch": 0.7492002952755905,
+      "grad_norm": 1.2972476482391357,
+      "learning_rate": 2.9501582707327358e-05,
+      "loss": 2.6427,
+      "step": 12179
+    },
+    {
+      "epoch": 0.749261811023622,
+      "grad_norm": 1.6338187456130981,
+      "learning_rate": 2.948786932725923e-05,
+      "loss": 2.0585,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7493233267716536,
+      "grad_norm": 1.4515211582183838,
+      "learning_rate": 2.9474158583951583e-05,
+      "loss": 2.6601,
+      "step": 12181
+    },
+    {
+      "epoch": 0.7493848425196851,
+      "grad_norm": 1.1990731954574585,
+      "learning_rate": 2.9460450477917144e-05,
+      "loss": 2.3463,
+      "step": 12182
+    },
+    {
+      "epoch": 0.7494463582677166,
+      "grad_norm": 1.524301528930664,
+      "learning_rate": 2.944674500966852e-05,
+      "loss": 2.3446,
+      "step": 12183
+    },
+    {
+      "epoch": 0.749507874015748,
+      "grad_norm": 1.2029885053634644,
+      "learning_rate": 2.943304217971824e-05,
+      "loss": 1.8697,
+      "step": 12184
+    },
+    {
+      "epoch": 0.7495693897637795,
+      "grad_norm": 1.3747395277023315,
+      "learning_rate": 2.9419341988578665e-05,
+      "loss": 2.1811,
+      "step": 12185
+    },
+    {
+      "epoch": 0.749630905511811,
+      "grad_norm": 1.0560526847839355,
+      "learning_rate": 2.940564443676216e-05,
+      "loss": 1.2646,
+      "step": 12186
+    },
+    {
+      "epoch": 0.7496924212598425,
+      "grad_norm": 1.4818862676620483,
+      "learning_rate": 2.939194952478087e-05,
+      "loss": 2.015,
+      "step": 12187
+    },
+    {
+      "epoch": 0.749753937007874,
+      "grad_norm": 1.5494214296340942,
+      "learning_rate": 2.9378257253146955e-05,
+      "loss": 1.9706,
+      "step": 12188
+    },
+    {
+      "epoch": 0.7498154527559056,
+      "grad_norm": 1.4632768630981445,
+      "learning_rate": 2.936456762237242e-05,
+      "loss": 2.3239,
+      "step": 12189
+    },
+    {
+      "epoch": 0.749876968503937,
+      "grad_norm": 1.3418123722076416,
+      "learning_rate": 2.9350880632969215e-05,
+      "loss": 2.4849,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7499384842519685,
+      "grad_norm": 1.1313115358352661,
+      "learning_rate": 2.9337196285449087e-05,
+      "loss": 2.9896,
+      "step": 12191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3468115329742432,
+      "learning_rate": 2.932351458032382e-05,
+      "loss": 2.6703,
+      "step": 12192
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 2.44388484954834,
+      "eval_runtime": 1035.8452,
+      "eval_samples_per_second": 3.701,
+      "eval_steps_per_second": 3.701,
+      "step": 12192
+    },
+    {
+      "epoch": 0.7500615157480315,
+      "grad_norm": 1.2086018323898315,
+      "learning_rate": 2.930983551810498e-05,
+      "loss": 2.8774,
+      "step": 12193
+    },
+    {
+      "epoch": 0.750123031496063,
+      "grad_norm": 1.1370933055877686,
+      "learning_rate": 2.9296159099304144e-05,
+      "loss": 2.5447,
+      "step": 12194
+    },
+    {
+      "epoch": 0.7501845472440944,
+      "grad_norm": 1.2050659656524658,
+      "learning_rate": 2.928248532443264e-05,
+      "loss": 2.903,
+      "step": 12195
+    },
+    {
+      "epoch": 0.750246062992126,
+      "grad_norm": 1.22750985622406,
+      "learning_rate": 2.9268814194001914e-05,
+      "loss": 2.3759,
+      "step": 12196
+    },
+    {
+      "epoch": 0.7503075787401575,
+      "grad_norm": 1.2183829545974731,
+      "learning_rate": 2.925514570852309e-05,
+      "loss": 2.5434,
+      "step": 12197
+    },
+    {
+      "epoch": 0.750369094488189,
+      "grad_norm": 1.2962971925735474,
+      "learning_rate": 2.924147986850737e-05,
+      "loss": 2.8124,
+      "step": 12198
+    },
+    {
+      "epoch": 0.7504306102362205,
+      "grad_norm": 1.3642160892486572,
+      "learning_rate": 2.9227816674465714e-05,
+      "loss": 3.0115,
+      "step": 12199
+    },
+    {
+      "epoch": 0.750492125984252,
+      "grad_norm": 1.1127068996429443,
+      "learning_rate": 2.921415612690911e-05,
+      "loss": 1.7771,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7505536417322834,
+      "grad_norm": 1.30765962600708,
+      "learning_rate": 2.9200498226348326e-05,
+      "loss": 2.2354,
+      "step": 12201
+    },
+    {
+      "epoch": 0.7506151574803149,
+      "grad_norm": 1.3515886068344116,
+      "learning_rate": 2.9186842973294116e-05,
+      "loss": 2.2345,
+      "step": 12202
+    },
+    {
+      "epoch": 0.7506766732283464,
+      "grad_norm": 1.2290550470352173,
+      "learning_rate": 2.9173190368257108e-05,
+      "loss": 4.0623,
+      "step": 12203
+    },
+    {
+      "epoch": 0.750738188976378,
+      "grad_norm": 1.2699711322784424,
+      "learning_rate": 2.9159540411747876e-05,
+      "loss": 1.9906,
+      "step": 12204
+    },
+    {
+      "epoch": 0.7507997047244095,
+      "grad_norm": 1.2224503755569458,
+      "learning_rate": 2.914589310427678e-05,
+      "loss": 2.4613,
+      "step": 12205
+    },
+    {
+      "epoch": 0.750861220472441,
+      "grad_norm": 1.5232954025268555,
+      "learning_rate": 2.9132248446354227e-05,
+      "loss": 1.8715,
+      "step": 12206
+    },
+    {
+      "epoch": 0.7509227362204725,
+      "grad_norm": 1.0865464210510254,
+      "learning_rate": 2.911860643849037e-05,
+      "loss": 3.2263,
+      "step": 12207
+    },
+    {
+      "epoch": 0.7509842519685039,
+      "grad_norm": 1.056138277053833,
+      "learning_rate": 2.91049670811954e-05,
+      "loss": 2.4246,
+      "step": 12208
+    },
+    {
+      "epoch": 0.7510457677165354,
+      "grad_norm": 1.4132282733917236,
+      "learning_rate": 2.909133037497933e-05,
+      "loss": 2.4697,
+      "step": 12209
+    },
+    {
+      "epoch": 0.7511072834645669,
+      "grad_norm": 1.0624934434890747,
+      "learning_rate": 2.9077696320352143e-05,
+      "loss": 2.2875,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7511687992125984,
+      "grad_norm": 1.4179502725601196,
+      "learning_rate": 2.9064064917823607e-05,
+      "loss": 2.5318,
+      "step": 12211
+    },
+    {
+      "epoch": 0.75123031496063,
+      "grad_norm": 1.2336153984069824,
+      "learning_rate": 2.9050436167903517e-05,
+      "loss": 2.7973,
+      "step": 12212
+    },
+    {
+      "epoch": 0.7512918307086615,
+      "grad_norm": 1.504095196723938,
+      "learning_rate": 2.9036810071101474e-05,
+      "loss": 2.1953,
+      "step": 12213
+    },
+    {
+      "epoch": 0.7513533464566929,
+      "grad_norm": 1.2243423461914062,
+      "learning_rate": 2.9023186627927023e-05,
+      "loss": 2.4797,
+      "step": 12214
+    },
+    {
+      "epoch": 0.7514148622047244,
+      "grad_norm": 1.161818504333496,
+      "learning_rate": 2.9009565838889642e-05,
+      "loss": 2.4778,
+      "step": 12215
+    },
+    {
+      "epoch": 0.7514763779527559,
+      "grad_norm": 1.2591023445129395,
+      "learning_rate": 2.8995947704498617e-05,
+      "loss": 2.4347,
+      "step": 12216
+    },
+    {
+      "epoch": 0.7515378937007874,
+      "grad_norm": 1.3104937076568604,
+      "learning_rate": 2.8982332225263255e-05,
+      "loss": 2.9441,
+      "step": 12217
+    },
+    {
+      "epoch": 0.7515994094488189,
+      "grad_norm": 1.5252248048782349,
+      "learning_rate": 2.8968719401692624e-05,
+      "loss": 2.6371,
+      "step": 12218
+    },
+    {
+      "epoch": 0.7516609251968503,
+      "grad_norm": 1.2877498865127563,
+      "learning_rate": 2.895510923429584e-05,
+      "loss": 2.6531,
+      "step": 12219
+    },
+    {
+      "epoch": 0.7517224409448819,
+      "grad_norm": 1.2315967082977295,
+      "learning_rate": 2.8941501723581788e-05,
+      "loss": 2.3482,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7517839566929134,
+      "grad_norm": 1.3631584644317627,
+      "learning_rate": 2.8927896870059325e-05,
+      "loss": 2.6943,
+      "step": 12221
+    },
+    {
+      "epoch": 0.7518454724409449,
+      "grad_norm": 1.2653639316558838,
+      "learning_rate": 2.891429467423722e-05,
+      "loss": 2.1538,
+      "step": 12222
+    },
+    {
+      "epoch": 0.7519069881889764,
+      "grad_norm": 1.430284857749939,
+      "learning_rate": 2.8900695136624133e-05,
+      "loss": 2.8605,
+      "step": 12223
+    },
+    {
+      "epoch": 0.7519685039370079,
+      "grad_norm": 1.323126196861267,
+      "learning_rate": 2.8887098257728573e-05,
+      "loss": 1.6131,
+      "step": 12224
+    },
+    {
+      "epoch": 0.7520300196850394,
+      "grad_norm": 1.0929495096206665,
+      "learning_rate": 2.8873504038059018e-05,
+      "loss": 1.3898,
+      "step": 12225
+    },
+    {
+      "epoch": 0.7520915354330708,
+      "grad_norm": 1.571635127067566,
+      "learning_rate": 2.8859912478123774e-05,
+      "loss": 2.2231,
+      "step": 12226
+    },
+    {
+      "epoch": 0.7521530511811023,
+      "grad_norm": 1.5210857391357422,
+      "learning_rate": 2.884632357843111e-05,
+      "loss": 1.7403,
+      "step": 12227
+    },
+    {
+      "epoch": 0.7522145669291339,
+      "grad_norm": 1.2766331434249878,
+      "learning_rate": 2.883273733948918e-05,
+      "loss": 2.7562,
+      "step": 12228
+    },
+    {
+      "epoch": 0.7522760826771654,
+      "grad_norm": 1.5239782333374023,
+      "learning_rate": 2.8819153761806073e-05,
+      "loss": 1.7924,
+      "step": 12229
+    },
+    {
+      "epoch": 0.7523375984251969,
+      "grad_norm": 1.283625841140747,
+      "learning_rate": 2.880557284588966e-05,
+      "loss": 2.4372,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7523991141732284,
+      "grad_norm": 1.5842127799987793,
+      "learning_rate": 2.8791994592247875e-05,
+      "loss": 2.3437,
+      "step": 12231
+    },
+    {
+      "epoch": 0.7524606299212598,
+      "grad_norm": 1.4838825464248657,
+      "learning_rate": 2.877841900138839e-05,
+      "loss": 1.3749,
+      "step": 12232
+    },
+    {
+      "epoch": 0.7525221456692913,
+      "grad_norm": 1.2549479007720947,
+      "learning_rate": 2.8764846073818886e-05,
+      "loss": 2.4446,
+      "step": 12233
+    },
+    {
+      "epoch": 0.7525836614173228,
+      "grad_norm": 1.2133760452270508,
+      "learning_rate": 2.8751275810046917e-05,
+      "loss": 2.2472,
+      "step": 12234
+    },
+    {
+      "epoch": 0.7526451771653543,
+      "grad_norm": 1.319292664527893,
+      "learning_rate": 2.8737708210579983e-05,
+      "loss": 2.5679,
+      "step": 12235
+    },
+    {
+      "epoch": 0.7527066929133859,
+      "grad_norm": 1.291015386581421,
+      "learning_rate": 2.8724143275925353e-05,
+      "loss": 2.1099,
+      "step": 12236
+    },
+    {
+      "epoch": 0.7527682086614174,
+      "grad_norm": 1.273210048675537,
+      "learning_rate": 2.8710581006590354e-05,
+      "loss": 2.9161,
+      "step": 12237
+    },
+    {
+      "epoch": 0.7528297244094488,
+      "grad_norm": 1.3474407196044922,
+      "learning_rate": 2.869702140308207e-05,
+      "loss": 2.3962,
+      "step": 12238
+    },
+    {
+      "epoch": 0.7528912401574803,
+      "grad_norm": 1.1449286937713623,
+      "learning_rate": 2.8683464465907595e-05,
+      "loss": 2.7067,
+      "step": 12239
+    },
+    {
+      "epoch": 0.7529527559055118,
+      "grad_norm": 1.4007174968719482,
+      "learning_rate": 2.866991019557387e-05,
+      "loss": 2.4998,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7530142716535433,
+      "grad_norm": 1.3513872623443604,
+      "learning_rate": 2.8656358592587795e-05,
+      "loss": 2.8492,
+      "step": 12241
+    },
+    {
+      "epoch": 0.7530757874015748,
+      "grad_norm": 1.2817132472991943,
+      "learning_rate": 2.8642809657456048e-05,
+      "loss": 1.9208,
+      "step": 12242
+    },
+    {
+      "epoch": 0.7531373031496063,
+      "grad_norm": 1.1714670658111572,
+      "learning_rate": 2.8629263390685357e-05,
+      "loss": 1.3671,
+      "step": 12243
+    },
+    {
+      "epoch": 0.7531988188976378,
+      "grad_norm": 1.317557692527771,
+      "learning_rate": 2.8615719792782205e-05,
+      "loss": 2.3663,
+      "step": 12244
+    },
+    {
+      "epoch": 0.7532603346456693,
+      "grad_norm": 1.2830867767333984,
+      "learning_rate": 2.8602178864253115e-05,
+      "loss": 2.3945,
+      "step": 12245
+    },
+    {
+      "epoch": 0.7533218503937008,
+      "grad_norm": 1.8097264766693115,
+      "learning_rate": 2.858864060560439e-05,
+      "loss": 2.6714,
+      "step": 12246
+    },
+    {
+      "epoch": 0.7533833661417323,
+      "grad_norm": 1.3132768869400024,
+      "learning_rate": 2.8575105017342295e-05,
+      "loss": 2.1814,
+      "step": 12247
+    },
+    {
+      "epoch": 0.7534448818897638,
+      "grad_norm": 1.2021795511245728,
+      "learning_rate": 2.856157209997299e-05,
+      "loss": 2.3811,
+      "step": 12248
+    },
+    {
+      "epoch": 0.7535063976377953,
+      "grad_norm": 1.1827187538146973,
+      "learning_rate": 2.8548041854002582e-05,
+      "loss": 2.4373,
+      "step": 12249
+    },
+    {
+      "epoch": 0.7535679133858267,
+      "grad_norm": 1.3228251934051514,
+      "learning_rate": 2.8534514279936973e-05,
+      "loss": 2.5918,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7536294291338582,
+      "grad_norm": 1.37088942527771,
+      "learning_rate": 2.8520989378282004e-05,
+      "loss": 1.9389,
+      "step": 12251
+    },
+    {
+      "epoch": 0.7536909448818898,
+      "grad_norm": 1.071250081062317,
+      "learning_rate": 2.8507467149543444e-05,
+      "loss": 2.5363,
+      "step": 12252
+    },
+    {
+      "epoch": 0.7537524606299213,
+      "grad_norm": 1.0889266729354858,
+      "learning_rate": 2.8493947594226966e-05,
+      "loss": 2.654,
+      "step": 12253
+    },
+    {
+      "epoch": 0.7538139763779528,
+      "grad_norm": 1.2061353921890259,
+      "learning_rate": 2.8480430712838146e-05,
+      "loss": 2.5007,
+      "step": 12254
+    },
+    {
+      "epoch": 0.7538754921259843,
+      "grad_norm": 1.4711155891418457,
+      "learning_rate": 2.846691650588239e-05,
+      "loss": 2.8393,
+      "step": 12255
+    },
+    {
+      "epoch": 0.7539370078740157,
+      "grad_norm": 1.2944018840789795,
+      "learning_rate": 2.8453404973865105e-05,
+      "loss": 2.6424,
+      "step": 12256
+    },
+    {
+      "epoch": 0.7539985236220472,
+      "grad_norm": 1.0470002889633179,
+      "learning_rate": 2.8439896117291486e-05,
+      "loss": 2.0081,
+      "step": 12257
+    },
+    {
+      "epoch": 0.7540600393700787,
+      "grad_norm": 1.3451687097549438,
+      "learning_rate": 2.842638993666672e-05,
+      "loss": 1.8886,
+      "step": 12258
+    },
+    {
+      "epoch": 0.7541215551181102,
+      "grad_norm": 1.628450632095337,
+      "learning_rate": 2.8412886432495865e-05,
+      "loss": 1.6407,
+      "step": 12259
+    },
+    {
+      "epoch": 0.7541830708661418,
+      "grad_norm": 1.3331528902053833,
+      "learning_rate": 2.8399385605283913e-05,
+      "loss": 3.2254,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7542445866141733,
+      "grad_norm": 1.144888997077942,
+      "learning_rate": 2.8385887455535654e-05,
+      "loss": 2.5544,
+      "step": 12261
+    },
+    {
+      "epoch": 0.7543061023622047,
+      "grad_norm": 1.0681467056274414,
+      "learning_rate": 2.8372391983755898e-05,
+      "loss": 2.9058,
+      "step": 12262
+    },
+    {
+      "epoch": 0.7543676181102362,
+      "grad_norm": 1.3220428228378296,
+      "learning_rate": 2.835889919044924e-05,
+      "loss": 2.119,
+      "step": 12263
+    },
+    {
+      "epoch": 0.7544291338582677,
+      "grad_norm": 1.527512788772583,
+      "learning_rate": 2.834540907612031e-05,
+      "loss": 2.2351,
+      "step": 12264
+    },
+    {
+      "epoch": 0.7544906496062992,
+      "grad_norm": 1.3116260766983032,
+      "learning_rate": 2.8331921641273462e-05,
+      "loss": 2.7852,
+      "step": 12265
+    },
+    {
+      "epoch": 0.7545521653543307,
+      "grad_norm": 1.2802958488464355,
+      "learning_rate": 2.8318436886413167e-05,
+      "loss": 2.4411,
+      "step": 12266
+    },
+    {
+      "epoch": 0.7546136811023622,
+      "grad_norm": 1.2344988584518433,
+      "learning_rate": 2.8304954812043604e-05,
+      "loss": 1.4716,
+      "step": 12267
+    },
+    {
+      "epoch": 0.7546751968503937,
+      "grad_norm": 2.165882110595703,
+      "learning_rate": 2.8291475418668965e-05,
+      "loss": 1.7741,
+      "step": 12268
+    },
+    {
+      "epoch": 0.7547367125984252,
+      "grad_norm": 1.3645144701004028,
+      "learning_rate": 2.8277998706793274e-05,
+      "loss": 2.4523,
+      "step": 12269
+    },
+    {
+      "epoch": 0.7547982283464567,
+      "grad_norm": 1.4316508769989014,
+      "learning_rate": 2.8264524676920524e-05,
+      "loss": 1.7743,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7548597440944882,
+      "grad_norm": 1.4915177822113037,
+      "learning_rate": 2.8251053329554498e-05,
+      "loss": 2.4265,
+      "step": 12271
+    },
+    {
+      "epoch": 0.7549212598425197,
+      "grad_norm": 1.290751338005066,
+      "learning_rate": 2.8237584665199014e-05,
+      "loss": 2.013,
+      "step": 12272
+    },
+    {
+      "epoch": 0.7549827755905512,
+      "grad_norm": 1.2434791326522827,
+      "learning_rate": 2.8224118684357694e-05,
+      "loss": 2.6668,
+      "step": 12273
+    },
+    {
+      "epoch": 0.7550442913385826,
+      "grad_norm": 1.1239006519317627,
+      "learning_rate": 2.8210655387534136e-05,
+      "loss": 1.8477,
+      "step": 12274
+    },
+    {
+      "epoch": 0.7551058070866141,
+      "grad_norm": 1.5972089767456055,
+      "learning_rate": 2.8197194775231728e-05,
+      "loss": 2.4227,
+      "step": 12275
+    },
+    {
+      "epoch": 0.7551673228346457,
+      "grad_norm": 1.281661868095398,
+      "learning_rate": 2.818373684795388e-05,
+      "loss": 2.1115,
+      "step": 12276
+    },
+    {
+      "epoch": 0.7552288385826772,
+      "grad_norm": 1.1991761922836304,
+      "learning_rate": 2.8170281606203786e-05,
+      "loss": 2.1133,
+      "step": 12277
+    },
+    {
+      "epoch": 0.7552903543307087,
+      "grad_norm": 1.3464531898498535,
+      "learning_rate": 2.8156829050484635e-05,
+      "loss": 2.2046,
+      "step": 12278
+    },
+    {
+      "epoch": 0.7553518700787402,
+      "grad_norm": 1.2643153667449951,
+      "learning_rate": 2.8143379181299455e-05,
+      "loss": 2.2635,
+      "step": 12279
+    },
+    {
+      "epoch": 0.7554133858267716,
+      "grad_norm": 1.3029268980026245,
+      "learning_rate": 2.812993199915125e-05,
+      "loss": 2.3326,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7554749015748031,
+      "grad_norm": 1.3490033149719238,
+      "learning_rate": 2.81164875045428e-05,
+      "loss": 1.6573,
+      "step": 12281
+    },
+    {
+      "epoch": 0.7555364173228346,
+      "grad_norm": 1.5597668886184692,
+      "learning_rate": 2.8103045697976905e-05,
+      "loss": 2.4127,
+      "step": 12282
+    },
+    {
+      "epoch": 0.7555979330708661,
+      "grad_norm": 1.5566529035568237,
+      "learning_rate": 2.8089606579956174e-05,
+      "loss": 2.0837,
+      "step": 12283
+    },
+    {
+      "epoch": 0.7556594488188977,
+      "grad_norm": 1.2290185689926147,
+      "learning_rate": 2.807617015098316e-05,
+      "loss": 2.187,
+      "step": 12284
+    },
+    {
+      "epoch": 0.7557209645669292,
+      "grad_norm": 1.4737097024917603,
+      "learning_rate": 2.8062736411560333e-05,
+      "loss": 2.3052,
+      "step": 12285
+    },
+    {
+      "epoch": 0.7557824803149606,
+      "grad_norm": 1.2593914270401,
+      "learning_rate": 2.8049305362190058e-05,
+      "loss": 2.0657,
+      "step": 12286
+    },
+    {
+      "epoch": 0.7558439960629921,
+      "grad_norm": 1.490357518196106,
+      "learning_rate": 2.8035877003374556e-05,
+      "loss": 1.7219,
+      "step": 12287
+    },
+    {
+      "epoch": 0.7559055118110236,
+      "grad_norm": 1.2197480201721191,
+      "learning_rate": 2.802245133561594e-05,
+      "loss": 2.8176,
+      "step": 12288
+    },
+    {
+      "epoch": 0.7559670275590551,
+      "grad_norm": 1.8359957933425903,
+      "learning_rate": 2.8009028359416323e-05,
+      "loss": 2.1226,
+      "step": 12289
+    },
+    {
+      "epoch": 0.7560285433070866,
+      "grad_norm": 1.4316469430923462,
+      "learning_rate": 2.799560807527757e-05,
+      "loss": 1.5429,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7560900590551181,
+      "grad_norm": 1.2742751836776733,
+      "learning_rate": 2.7982190483701577e-05,
+      "loss": 2.8141,
+      "step": 12291
+    },
+    {
+      "epoch": 0.7561515748031497,
+      "grad_norm": 1.1839746236801147,
+      "learning_rate": 2.7968775585190076e-05,
+      "loss": 2.1706,
+      "step": 12292
+    },
+    {
+      "epoch": 0.7562130905511811,
+      "grad_norm": 1.573228120803833,
+      "learning_rate": 2.7955363380244737e-05,
+      "loss": 2.0488,
+      "step": 12293
+    },
+    {
+      "epoch": 0.7562746062992126,
+      "grad_norm": 1.2714544534683228,
+      "learning_rate": 2.7941953869367054e-05,
+      "loss": 1.651,
+      "step": 12294
+    },
+    {
+      "epoch": 0.7563361220472441,
+      "grad_norm": 1.2749834060668945,
+      "learning_rate": 2.7928547053058507e-05,
+      "loss": 2.029,
+      "step": 12295
+    },
+    {
+      "epoch": 0.7563976377952756,
+      "grad_norm": 1.454814076423645,
+      "learning_rate": 2.7915142931820405e-05,
+      "loss": 2.3466,
+      "step": 12296
+    },
+    {
+      "epoch": 0.7564591535433071,
+      "grad_norm": 1.202738642692566,
+      "learning_rate": 2.790174150615399e-05,
+      "loss": 2.1889,
+      "step": 12297
+    },
+    {
+      "epoch": 0.7565206692913385,
+      "grad_norm": 1.3822451829910278,
+      "learning_rate": 2.7888342776560406e-05,
+      "loss": 2.4758,
+      "step": 12298
+    },
+    {
+      "epoch": 0.75658218503937,
+      "grad_norm": 2.818451166152954,
+      "learning_rate": 2.7874946743540743e-05,
+      "loss": 3.559,
+      "step": 12299
+    },
+    {
+      "epoch": 0.7566437007874016,
+      "grad_norm": 1.4375497102737427,
+      "learning_rate": 2.786155340759585e-05,
+      "loss": 3.728,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7567052165354331,
+      "grad_norm": 1.224476933479309,
+      "learning_rate": 2.7848162769226637e-05,
+      "loss": 2.7687,
+      "step": 12301
+    },
+    {
+      "epoch": 0.7567667322834646,
+      "grad_norm": 1.2960422039031982,
+      "learning_rate": 2.7834774828933773e-05,
+      "loss": 1.9452,
+      "step": 12302
+    },
+    {
+      "epoch": 0.7568282480314961,
+      "grad_norm": 1.482041597366333,
+      "learning_rate": 2.7821389587217938e-05,
+      "loss": 3.682,
+      "step": 12303
+    },
+    {
+      "epoch": 0.7568897637795275,
+      "grad_norm": 1.3603119850158691,
+      "learning_rate": 2.7808007044579643e-05,
+      "loss": 1.7984,
+      "step": 12304
+    },
+    {
+      "epoch": 0.756951279527559,
+      "grad_norm": 1.3533368110656738,
+      "learning_rate": 2.7794627201519364e-05,
+      "loss": 2.6607,
+      "step": 12305
+    },
+    {
+      "epoch": 0.7570127952755905,
+      "grad_norm": 1.3508394956588745,
+      "learning_rate": 2.7781250058537367e-05,
+      "loss": 1.99,
+      "step": 12306
+    },
+    {
+      "epoch": 0.757074311023622,
+      "grad_norm": 1.1464353799819946,
+      "learning_rate": 2.776787561613394e-05,
+      "loss": 2.2557,
+      "step": 12307
+    },
+    {
+      "epoch": 0.7571358267716536,
+      "grad_norm": 1.1903971433639526,
+      "learning_rate": 2.7754503874809168e-05,
+      "loss": 2.5378,
+      "step": 12308
+    },
+    {
+      "epoch": 0.7571973425196851,
+      "grad_norm": 1.2996307611465454,
+      "learning_rate": 2.7741134835063132e-05,
+      "loss": 1.7441,
+      "step": 12309
+    },
+    {
+      "epoch": 0.7572588582677166,
+      "grad_norm": 1.293467402458191,
+      "learning_rate": 2.7727768497395658e-05,
+      "loss": 2.2545,
+      "step": 12310
+    },
+    {
+      "epoch": 0.757320374015748,
+      "grad_norm": 1.3009763956069946,
+      "learning_rate": 2.7714404862306707e-05,
+      "loss": 2.7247,
+      "step": 12311
+    },
+    {
+      "epoch": 0.7573818897637795,
+      "grad_norm": 1.22845458984375,
+      "learning_rate": 2.77010439302959e-05,
+      "loss": 2.9028,
+      "step": 12312
+    },
+    {
+      "epoch": 0.757443405511811,
+      "grad_norm": 1.2821581363677979,
+      "learning_rate": 2.7687685701862943e-05,
+      "loss": 2.3607,
+      "step": 12313
+    },
+    {
+      "epoch": 0.7575049212598425,
+      "grad_norm": 1.3725416660308838,
+      "learning_rate": 2.767433017750727e-05,
+      "loss": 2.6523,
+      "step": 12314
+    },
+    {
+      "epoch": 0.757566437007874,
+      "grad_norm": 1.9038777351379395,
+      "learning_rate": 2.766097735772839e-05,
+      "loss": 2.4199,
+      "step": 12315
+    },
+    {
+      "epoch": 0.7576279527559056,
+      "grad_norm": 1.2626076936721802,
+      "learning_rate": 2.7647627243025553e-05,
+      "loss": 1.9942,
+      "step": 12316
+    },
+    {
+      "epoch": 0.757689468503937,
+      "grad_norm": 1.1935184001922607,
+      "learning_rate": 2.7634279833898004e-05,
+      "loss": 2.3686,
+      "step": 12317
+    },
+    {
+      "epoch": 0.7577509842519685,
+      "grad_norm": 1.1291757822036743,
+      "learning_rate": 2.7620935130844873e-05,
+      "loss": 2.108,
+      "step": 12318
+    },
+    {
+      "epoch": 0.7578125,
+      "grad_norm": 1.1853879690170288,
+      "learning_rate": 2.7607593134365196e-05,
+      "loss": 2.2345,
+      "step": 12319
+    },
+    {
+      "epoch": 0.7578740157480315,
+      "grad_norm": 1.3480526208877563,
+      "learning_rate": 2.759425384495784e-05,
+      "loss": 2.7643,
+      "step": 12320
+    },
+    {
+      "epoch": 0.757935531496063,
+      "grad_norm": 1.381865382194519,
+      "learning_rate": 2.7580917263121676e-05,
+      "loss": 2.4718,
+      "step": 12321
+    },
+    {
+      "epoch": 0.7579970472440944,
+      "grad_norm": 1.3516179323196411,
+      "learning_rate": 2.7567583389355344e-05,
+      "loss": 3.0538,
+      "step": 12322
+    },
+    {
+      "epoch": 0.758058562992126,
+      "grad_norm": 1.1055347919464111,
+      "learning_rate": 2.7554252224157505e-05,
+      "loss": 2.4468,
+      "step": 12323
+    },
+    {
+      "epoch": 0.7581200787401575,
+      "grad_norm": 1.4135123491287231,
+      "learning_rate": 2.7540923768026707e-05,
+      "loss": 2.5522,
+      "step": 12324
+    },
+    {
+      "epoch": 0.758181594488189,
+      "grad_norm": 1.1282674074172974,
+      "learning_rate": 2.7527598021461276e-05,
+      "loss": 3.048,
+      "step": 12325
+    },
+    {
+      "epoch": 0.7582431102362205,
+      "grad_norm": 1.6315019130706787,
+      "learning_rate": 2.751427498495959e-05,
+      "loss": 2.5012,
+      "step": 12326
+    },
+    {
+      "epoch": 0.758304625984252,
+      "grad_norm": 1.188609004020691,
+      "learning_rate": 2.7500954659019806e-05,
+      "loss": 2.1008,
+      "step": 12327
+    },
+    {
+      "epoch": 0.7583661417322834,
+      "grad_norm": 1.2013826370239258,
+      "learning_rate": 2.7487637044140048e-05,
+      "loss": 2.2891,
+      "step": 12328
+    },
+    {
+      "epoch": 0.7584276574803149,
+      "grad_norm": 1.7251348495483398,
+      "learning_rate": 2.7474322140818322e-05,
+      "loss": 1.8135,
+      "step": 12329
+    },
+    {
+      "epoch": 0.7584891732283464,
+      "grad_norm": 1.343966007232666,
+      "learning_rate": 2.746100994955256e-05,
+      "loss": 2.3179,
+      "step": 12330
+    },
+    {
+      "epoch": 0.758550688976378,
+      "grad_norm": 1.1787996292114258,
+      "learning_rate": 2.744770047084051e-05,
+      "loss": 2.8267,
+      "step": 12331
+    },
+    {
+      "epoch": 0.7586122047244095,
+      "grad_norm": 1.7608596086502075,
+      "learning_rate": 2.743439370517994e-05,
+      "loss": 1.7147,
+      "step": 12332
+    },
+    {
+      "epoch": 0.758673720472441,
+      "grad_norm": 1.6152199506759644,
+      "learning_rate": 2.7421089653068366e-05,
+      "loss": 2.0108,
+      "step": 12333
+    },
+    {
+      "epoch": 0.7587352362204725,
+      "grad_norm": 1.3478434085845947,
+      "learning_rate": 2.7407788315003356e-05,
+      "loss": 2.4126,
+      "step": 12334
+    },
+    {
+      "epoch": 0.7587967519685039,
+      "grad_norm": 1.2028359174728394,
+      "learning_rate": 2.739448969148226e-05,
+      "loss": 2.9294,
+      "step": 12335
+    },
+    {
+      "epoch": 0.7588582677165354,
+      "grad_norm": 1.3527559041976929,
+      "learning_rate": 2.7381193783002392e-05,
+      "loss": 2.123,
+      "step": 12336
+    },
+    {
+      "epoch": 0.7589197834645669,
+      "grad_norm": 1.056169033050537,
+      "learning_rate": 2.7367900590060925e-05,
+      "loss": 2.9709,
+      "step": 12337
+    },
+    {
+      "epoch": 0.7589812992125984,
+      "grad_norm": 1.4998979568481445,
+      "learning_rate": 2.735461011315501e-05,
+      "loss": 2.3309,
+      "step": 12338
+    },
+    {
+      "epoch": 0.75904281496063,
+      "grad_norm": 1.2076013088226318,
+      "learning_rate": 2.734132235278156e-05,
+      "loss": 2.7938,
+      "step": 12339
+    },
+    {
+      "epoch": 0.7591043307086615,
+      "grad_norm": 1.237858772277832,
+      "learning_rate": 2.732803730943754e-05,
+      "loss": 2.4237,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7591658464566929,
+      "grad_norm": 1.2293809652328491,
+      "learning_rate": 2.731475498361965e-05,
+      "loss": 2.623,
+      "step": 12341
+    },
+    {
+      "epoch": 0.7592273622047244,
+      "grad_norm": 1.4025835990905762,
+      "learning_rate": 2.7301475375824627e-05,
+      "loss": 2.5766,
+      "step": 12342
+    },
+    {
+      "epoch": 0.7592888779527559,
+      "grad_norm": 1.2150996923446655,
+      "learning_rate": 2.7288198486549042e-05,
+      "loss": 2.5074,
+      "step": 12343
+    },
+    {
+      "epoch": 0.7593503937007874,
+      "grad_norm": 1.0606135129928589,
+      "learning_rate": 2.727492431628942e-05,
+      "loss": 2.8588,
+      "step": 12344
+    },
+    {
+      "epoch": 0.7594119094488189,
+      "grad_norm": 1.30427885055542,
+      "learning_rate": 2.7261652865542063e-05,
+      "loss": 3.086,
+      "step": 12345
+    },
+    {
+      "epoch": 0.7594734251968503,
+      "grad_norm": 1.2183337211608887,
+      "learning_rate": 2.724838413480333e-05,
+      "loss": 2.5502,
+      "step": 12346
+    },
+    {
+      "epoch": 0.7595349409448819,
+      "grad_norm": 1.321625828742981,
+      "learning_rate": 2.7235118124569316e-05,
+      "loss": 2.5128,
+      "step": 12347
+    },
+    {
+      "epoch": 0.7595964566929134,
+      "grad_norm": 1.1546738147735596,
+      "learning_rate": 2.7221854835336136e-05,
+      "loss": 2.6663,
+      "step": 12348
+    },
+    {
+      "epoch": 0.7596579724409449,
+      "grad_norm": 1.3055211305618286,
+      "learning_rate": 2.7208594267599774e-05,
+      "loss": 2.2685,
+      "step": 12349
+    },
+    {
+      "epoch": 0.7597194881889764,
+      "grad_norm": 1.4275857210159302,
+      "learning_rate": 2.719533642185611e-05,
+      "loss": 2.2697,
+      "step": 12350
+    },
+    {
+      "epoch": 0.7597810039370079,
+      "grad_norm": 1.3287726640701294,
+      "learning_rate": 2.718208129860087e-05,
+      "loss": 2.9816,
+      "step": 12351
+    },
+    {
+      "epoch": 0.7598425196850394,
+      "grad_norm": 1.1468677520751953,
+      "learning_rate": 2.7168828898329767e-05,
+      "loss": 1.9523,
+      "step": 12352
+    },
+    {
+      "epoch": 0.7599040354330708,
+      "grad_norm": 1.4502012729644775,
+      "learning_rate": 2.715557922153833e-05,
+      "loss": 2.6342,
+      "step": 12353
+    },
+    {
+      "epoch": 0.7599655511811023,
+      "grad_norm": 1.3809700012207031,
+      "learning_rate": 2.7142332268722026e-05,
+      "loss": 2.4673,
+      "step": 12354
+    },
+    {
+      "epoch": 0.7600270669291339,
+      "grad_norm": 1.3664281368255615,
+      "learning_rate": 2.7129088040376226e-05,
+      "loss": 2.0033,
+      "step": 12355
+    },
+    {
+      "epoch": 0.7600885826771654,
+      "grad_norm": 1.2426568269729614,
+      "learning_rate": 2.7115846536996237e-05,
+      "loss": 2.8146,
+      "step": 12356
+    },
+    {
+      "epoch": 0.7601500984251969,
+      "grad_norm": 1.1429537534713745,
+      "learning_rate": 2.7102607759077137e-05,
+      "loss": 2.4526,
+      "step": 12357
+    },
+    {
+      "epoch": 0.7602116141732284,
+      "grad_norm": 1.5048365592956543,
+      "learning_rate": 2.7089371707114054e-05,
+      "loss": 2.8474,
+      "step": 12358
+    },
+    {
+      "epoch": 0.7602731299212598,
+      "grad_norm": 1.1577996015548706,
+      "learning_rate": 2.707613838160187e-05,
+      "loss": 2.4207,
+      "step": 12359
+    },
+    {
+      "epoch": 0.7603346456692913,
+      "grad_norm": 1.4565132856369019,
+      "learning_rate": 2.706290778303552e-05,
+      "loss": 1.6517,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7603961614173228,
+      "grad_norm": 1.3377671241760254,
+      "learning_rate": 2.7049679911909666e-05,
+      "loss": 2.5918,
+      "step": 12361
+    },
+    {
+      "epoch": 0.7604576771653543,
+      "grad_norm": 1.1865622997283936,
+      "learning_rate": 2.703645476871901e-05,
+      "loss": 2.6497,
+      "step": 12362
+    },
+    {
+      "epoch": 0.7605191929133859,
+      "grad_norm": 1.4317630529403687,
+      "learning_rate": 2.7023232353958118e-05,
+      "loss": 2.6408,
+      "step": 12363
+    },
+    {
+      "epoch": 0.7605807086614174,
+      "grad_norm": 1.2287341356277466,
+      "learning_rate": 2.7010012668121378e-05,
+      "loss": 1.6899,
+      "step": 12364
+    },
+    {
+      "epoch": 0.7606422244094488,
+      "grad_norm": 1.3731741905212402,
+      "learning_rate": 2.6996795711703183e-05,
+      "loss": 2.2984,
+      "step": 12365
+    },
+    {
+      "epoch": 0.7607037401574803,
+      "grad_norm": 1.8333574533462524,
+      "learning_rate": 2.698358148519773e-05,
+      "loss": 2.1626,
+      "step": 12366
+    },
+    {
+      "epoch": 0.7607652559055118,
+      "grad_norm": 1.3829450607299805,
+      "learning_rate": 2.6970369989099163e-05,
+      "loss": 1.9312,
+      "step": 12367
+    },
+    {
+      "epoch": 0.7608267716535433,
+      "grad_norm": 1.2114617824554443,
+      "learning_rate": 2.6957161223901552e-05,
+      "loss": 2.433,
+      "step": 12368
+    },
+    {
+      "epoch": 0.7608882874015748,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 2.6943955190098834e-05,
+      "loss": 2.7377,
+      "step": 12369
+    },
+    {
+      "epoch": 0.7609498031496063,
+      "grad_norm": 1.1774392127990723,
+      "learning_rate": 2.6930751888184802e-05,
+      "loss": 2.4818,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7610113188976378,
+      "grad_norm": 1.087149739265442,
+      "learning_rate": 2.6917551318653223e-05,
+      "loss": 2.3322,
+      "step": 12371
+    },
+    {
+      "epoch": 0.7610728346456693,
+      "grad_norm": 1.4572222232818604,
+      "learning_rate": 2.690435348199768e-05,
+      "loss": 2.63,
+      "step": 12372
+    },
+    {
+      "epoch": 0.7611343503937008,
+      "grad_norm": 1.068434715270996,
+      "learning_rate": 2.6891158378711723e-05,
+      "loss": 2.6965,
+      "step": 12373
+    },
+    {
+      "epoch": 0.7611958661417323,
+      "grad_norm": 1.221816062927246,
+      "learning_rate": 2.687796600928877e-05,
+      "loss": 2.289,
+      "step": 12374
+    },
+    {
+      "epoch": 0.7612573818897638,
+      "grad_norm": 1.38926100730896,
+      "learning_rate": 2.686477637422219e-05,
+      "loss": 2.7296,
+      "step": 12375
+    },
+    {
+      "epoch": 0.7613188976377953,
+      "grad_norm": 1.3371025323867798,
+      "learning_rate": 2.6851589474005125e-05,
+      "loss": 2.1527,
+      "step": 12376
+    },
+    {
+      "epoch": 0.7613804133858267,
+      "grad_norm": 1.5292549133300781,
+      "learning_rate": 2.6838405309130753e-05,
+      "loss": 2.614,
+      "step": 12377
+    },
+    {
+      "epoch": 0.7614419291338582,
+      "grad_norm": 1.1400392055511475,
+      "learning_rate": 2.682522388009204e-05,
+      "loss": 2.249,
+      "step": 12378
+    },
+    {
+      "epoch": 0.7615034448818898,
+      "grad_norm": 1.1934058666229248,
+      "learning_rate": 2.681204518738195e-05,
+      "loss": 2.997,
+      "step": 12379
+    },
+    {
+      "epoch": 0.7615649606299213,
+      "grad_norm": 1.3338383436203003,
+      "learning_rate": 2.6798869231493197e-05,
+      "loss": 2.7504,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7616264763779528,
+      "grad_norm": 1.1836891174316406,
+      "learning_rate": 2.6785696012918614e-05,
+      "loss": 2.3006,
+      "step": 12381
+    },
+    {
+      "epoch": 0.7616879921259843,
+      "grad_norm": 1.3242024183273315,
+      "learning_rate": 2.6772525532150716e-05,
+      "loss": 2.6732,
+      "step": 12382
+    },
+    {
+      "epoch": 0.7617495078740157,
+      "grad_norm": 1.2735884189605713,
+      "learning_rate": 2.675935778968207e-05,
+      "loss": 3.0716,
+      "step": 12383
+    },
+    {
+      "epoch": 0.7618110236220472,
+      "grad_norm": 1.1315122842788696,
+      "learning_rate": 2.674619278600501e-05,
+      "loss": 2.8367,
+      "step": 12384
+    },
+    {
+      "epoch": 0.7618725393700787,
+      "grad_norm": 1.8667430877685547,
+      "learning_rate": 2.673303052161189e-05,
+      "loss": 2.2629,
+      "step": 12385
+    },
+    {
+      "epoch": 0.7619340551181102,
+      "grad_norm": 1.2265647649765015,
+      "learning_rate": 2.6719870996994868e-05,
+      "loss": 2.4461,
+      "step": 12386
+    },
+    {
+      "epoch": 0.7619955708661418,
+      "grad_norm": 1.0732736587524414,
+      "learning_rate": 2.6706714212646034e-05,
+      "loss": 1.97,
+      "step": 12387
+    },
+    {
+      "epoch": 0.7620570866141733,
+      "grad_norm": 1.3521064519882202,
+      "learning_rate": 2.6693560169057397e-05,
+      "loss": 2.6995,
+      "step": 12388
+    },
+    {
+      "epoch": 0.7621186023622047,
+      "grad_norm": 1.1499258279800415,
+      "learning_rate": 2.6680408866720864e-05,
+      "loss": 3.1615,
+      "step": 12389
+    },
+    {
+      "epoch": 0.7621801181102362,
+      "grad_norm": 1.4854274988174438,
+      "learning_rate": 2.6667260306128182e-05,
+      "loss": 2.1491,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7622416338582677,
+      "grad_norm": 2.092366933822632,
+      "learning_rate": 2.6654114487771076e-05,
+      "loss": 2.5506,
+      "step": 12391
+    },
+    {
+      "epoch": 0.7623031496062992,
+      "grad_norm": 1.4760669469833374,
+      "learning_rate": 2.664097141214107e-05,
+      "loss": 2.6191,
+      "step": 12392
+    },
+    {
+      "epoch": 0.7623646653543307,
+      "grad_norm": 1.2919038534164429,
+      "learning_rate": 2.662783107972966e-05,
+      "loss": 2.7469,
+      "step": 12393
+    },
+    {
+      "epoch": 0.7624261811023622,
+      "grad_norm": 1.3613179922103882,
+      "learning_rate": 2.661469349102824e-05,
+      "loss": 2.1488,
+      "step": 12394
+    },
+    {
+      "epoch": 0.7624876968503937,
+      "grad_norm": 1.1055041551589966,
+      "learning_rate": 2.6601558646528114e-05,
+      "loss": 3.0539,
+      "step": 12395
+    },
+    {
+      "epoch": 0.7625492125984252,
+      "grad_norm": 1.8468188047409058,
+      "learning_rate": 2.658842654672038e-05,
+      "loss": 2.4462,
+      "step": 12396
+    },
+    {
+      "epoch": 0.7626107283464567,
+      "grad_norm": 1.2859699726104736,
+      "learning_rate": 2.657529719209617e-05,
+      "loss": 2.5021,
+      "step": 12397
+    },
+    {
+      "epoch": 0.7626722440944882,
+      "grad_norm": 1.2098698616027832,
+      "learning_rate": 2.6562170583146418e-05,
+      "loss": 1.9865,
+      "step": 12398
+    },
+    {
+      "epoch": 0.7627337598425197,
+      "grad_norm": 1.3592673540115356,
+      "learning_rate": 2.6549046720361913e-05,
+      "loss": 2.2839,
+      "step": 12399
+    },
+    {
+      "epoch": 0.7627952755905512,
+      "grad_norm": 1.4842720031738281,
+      "learning_rate": 2.6535925604233558e-05,
+      "loss": 2.8151,
+      "step": 12400
+    },
+    {
+      "epoch": 0.7628567913385826,
+      "grad_norm": 1.2576227188110352,
+      "learning_rate": 2.652280723525191e-05,
+      "loss": 2.5461,
+      "step": 12401
+    },
+    {
+      "epoch": 0.7629183070866141,
+      "grad_norm": 1.3165488243103027,
+      "learning_rate": 2.6509691613907583e-05,
+      "loss": 2.507,
+      "step": 12402
+    },
+    {
+      "epoch": 0.7629798228346457,
+      "grad_norm": 1.259328842163086,
+      "learning_rate": 2.649657874069096e-05,
+      "loss": 2.1091,
+      "step": 12403
+    },
+    {
+      "epoch": 0.7630413385826772,
+      "grad_norm": 1.3064770698547363,
+      "learning_rate": 2.6483468616092454e-05,
+      "loss": 2.5945,
+      "step": 12404
+    },
+    {
+      "epoch": 0.7631028543307087,
+      "grad_norm": 1.2694432735443115,
+      "learning_rate": 2.647036124060226e-05,
+      "loss": 2.4528,
+      "step": 12405
+    },
+    {
+      "epoch": 0.7631643700787402,
+      "grad_norm": 1.3805938959121704,
+      "learning_rate": 2.6457256614710547e-05,
+      "loss": 2.1035,
+      "step": 12406
+    },
+    {
+      "epoch": 0.7632258858267716,
+      "grad_norm": 1.1920113563537598,
+      "learning_rate": 2.6444154738907336e-05,
+      "loss": 2.4875,
+      "step": 12407
+    },
+    {
+      "epoch": 0.7632874015748031,
+      "grad_norm": 1.0717958211898804,
+      "learning_rate": 2.643105561368262e-05,
+      "loss": 2.2339,
+      "step": 12408
+    },
+    {
+      "epoch": 0.7633489173228346,
+      "grad_norm": 1.1770024299621582,
+      "learning_rate": 2.6417959239526157e-05,
+      "loss": 2.9132,
+      "step": 12409
+    },
+    {
+      "epoch": 0.7634104330708661,
+      "grad_norm": 1.525758147239685,
+      "learning_rate": 2.6404865616927743e-05,
+      "loss": 2.4223,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7634719488188977,
+      "grad_norm": 1.1526836156845093,
+      "learning_rate": 2.639177474637695e-05,
+      "loss": 2.8516,
+      "step": 12411
+    },
+    {
+      "epoch": 0.7635334645669292,
+      "grad_norm": 1.185744285583496,
+      "learning_rate": 2.6378686628363335e-05,
+      "loss": 2.7757,
+      "step": 12412
+    },
+    {
+      "epoch": 0.7635949803149606,
+      "grad_norm": 1.5308468341827393,
+      "learning_rate": 2.6365601263376306e-05,
+      "loss": 2.3339,
+      "step": 12413
+    },
+    {
+      "epoch": 0.7636564960629921,
+      "grad_norm": 1.426340103149414,
+      "learning_rate": 2.6352518651905235e-05,
+      "loss": 2.1205,
+      "step": 12414
+    },
+    {
+      "epoch": 0.7637180118110236,
+      "grad_norm": 1.1326920986175537,
+      "learning_rate": 2.6339438794439263e-05,
+      "loss": 2.8413,
+      "step": 12415
+    },
+    {
+      "epoch": 0.7637795275590551,
+      "grad_norm": 1.3931468725204468,
+      "learning_rate": 2.632636169146756e-05,
+      "loss": 2.8749,
+      "step": 12416
+    },
+    {
+      "epoch": 0.7638410433070866,
+      "grad_norm": 1.1263303756713867,
+      "learning_rate": 2.6313287343479086e-05,
+      "loss": 2.7375,
+      "step": 12417
+    },
+    {
+      "epoch": 0.7639025590551181,
+      "grad_norm": 1.3875828981399536,
+      "learning_rate": 2.6300215750962786e-05,
+      "loss": 2.0723,
+      "step": 12418
+    },
+    {
+      "epoch": 0.7639640748031497,
+      "grad_norm": 1.5641953945159912,
+      "learning_rate": 2.6287146914407446e-05,
+      "loss": 1.8948,
+      "step": 12419
+    },
+    {
+      "epoch": 0.7640255905511811,
+      "grad_norm": 1.660504937171936,
+      "learning_rate": 2.6274080834301817e-05,
+      "loss": 2.4641,
+      "step": 12420
+    },
+    {
+      "epoch": 0.7640871062992126,
+      "grad_norm": 1.2455116510391235,
+      "learning_rate": 2.6261017511134424e-05,
+      "loss": 1.2519,
+      "step": 12421
+    },
+    {
+      "epoch": 0.7641486220472441,
+      "grad_norm": 1.4041827917099,
+      "learning_rate": 2.6247956945393825e-05,
+      "loss": 2.3108,
+      "step": 12422
+    },
+    {
+      "epoch": 0.7642101377952756,
+      "grad_norm": 1.3994035720825195,
+      "learning_rate": 2.6234899137568348e-05,
+      "loss": 2.2653,
+      "step": 12423
+    },
+    {
+      "epoch": 0.7642716535433071,
+      "grad_norm": 1.4595671892166138,
+      "learning_rate": 2.6221844088146352e-05,
+      "loss": 1.9738,
+      "step": 12424
+    },
+    {
+      "epoch": 0.7643331692913385,
+      "grad_norm": 2.220447063446045,
+      "learning_rate": 2.6208791797615928e-05,
+      "loss": 2.0453,
+      "step": 12425
+    },
+    {
+      "epoch": 0.76439468503937,
+      "grad_norm": 1.2412874698638916,
+      "learning_rate": 2.6195742266465285e-05,
+      "loss": 2.6821,
+      "step": 12426
+    },
+    {
+      "epoch": 0.7644562007874016,
+      "grad_norm": 1.2325760126113892,
+      "learning_rate": 2.6182695495182307e-05,
+      "loss": 2.2511,
+      "step": 12427
+    },
+    {
+      "epoch": 0.7645177165354331,
+      "grad_norm": 1.2212493419647217,
+      "learning_rate": 2.616965148425493e-05,
+      "loss": 2.6295,
+      "step": 12428
+    },
+    {
+      "epoch": 0.7645792322834646,
+      "grad_norm": 1.3464983701705933,
+      "learning_rate": 2.6156610234170864e-05,
+      "loss": 1.8749,
+      "step": 12429
+    },
+    {
+      "epoch": 0.7646407480314961,
+      "grad_norm": 1.3751558065414429,
+      "learning_rate": 2.614357174541784e-05,
+      "loss": 2.6135,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7647022637795275,
+      "grad_norm": 1.6329851150512695,
+      "learning_rate": 2.6130536018483375e-05,
+      "loss": 2.0561,
+      "step": 12431
+    },
+    {
+      "epoch": 0.764763779527559,
+      "grad_norm": 1.2278341054916382,
+      "learning_rate": 2.611750305385495e-05,
+      "loss": 2.3626,
+      "step": 12432
+    },
+    {
+      "epoch": 0.7648252952755905,
+      "grad_norm": 1.3102818727493286,
+      "learning_rate": 2.6104472852019946e-05,
+      "loss": 2.6246,
+      "step": 12433
+    },
+    {
+      "epoch": 0.764886811023622,
+      "grad_norm": 1.4872817993164062,
+      "learning_rate": 2.609144541346562e-05,
+      "loss": 2.7629,
+      "step": 12434
+    },
+    {
+      "epoch": 0.7649483267716536,
+      "grad_norm": 1.4072515964508057,
+      "learning_rate": 2.607842073867911e-05,
+      "loss": 2.5667,
+      "step": 12435
+    },
+    {
+      "epoch": 0.7650098425196851,
+      "grad_norm": 1.2878233194351196,
+      "learning_rate": 2.6065398828147437e-05,
+      "loss": 2.6747,
+      "step": 12436
+    },
+    {
+      "epoch": 0.7650713582677166,
+      "grad_norm": 1.480086326599121,
+      "learning_rate": 2.6052379682357586e-05,
+      "loss": 2.5026,
+      "step": 12437
+    },
+    {
+      "epoch": 0.765132874015748,
+      "grad_norm": 1.2093867063522339,
+      "learning_rate": 2.603936330179638e-05,
+      "loss": 2.238,
+      "step": 12438
+    },
+    {
+      "epoch": 0.7651943897637795,
+      "grad_norm": 1.418361783027649,
+      "learning_rate": 2.6026349686950614e-05,
+      "loss": 2.1571,
+      "step": 12439
+    },
+    {
+      "epoch": 0.765255905511811,
+      "grad_norm": 1.214308261871338,
+      "learning_rate": 2.6013338838306832e-05,
+      "loss": 2.3385,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7653174212598425,
+      "grad_norm": 1.0453122854232788,
+      "learning_rate": 2.6000330756351664e-05,
+      "loss": 2.6533,
+      "step": 12441
+    },
+    {
+      "epoch": 0.765378937007874,
+      "grad_norm": 1.2994251251220703,
+      "learning_rate": 2.598732544157145e-05,
+      "loss": 2.6328,
+      "step": 12442
+    },
+    {
+      "epoch": 0.7654404527559056,
+      "grad_norm": 1.3245925903320312,
+      "learning_rate": 2.5974322894452576e-05,
+      "loss": 2.0839,
+      "step": 12443
+    },
+    {
+      "epoch": 0.765501968503937,
+      "grad_norm": 1.2184478044509888,
+      "learning_rate": 2.5961323115481227e-05,
+      "loss": 2.4423,
+      "step": 12444
+    },
+    {
+      "epoch": 0.7655634842519685,
+      "grad_norm": 1.2239923477172852,
+      "learning_rate": 2.5948326105143583e-05,
+      "loss": 2.5852,
+      "step": 12445
+    },
+    {
+      "epoch": 0.765625,
+      "grad_norm": 1.2295602560043335,
+      "learning_rate": 2.5935331863925583e-05,
+      "loss": 1.8003,
+      "step": 12446
+    },
+    {
+      "epoch": 0.7656865157480315,
+      "grad_norm": 1.7520620822906494,
+      "learning_rate": 2.5922340392313217e-05,
+      "loss": 2.1098,
+      "step": 12447
+    },
+    {
+      "epoch": 0.765748031496063,
+      "grad_norm": 1.323722243309021,
+      "learning_rate": 2.5909351690792217e-05,
+      "loss": 2.4877,
+      "step": 12448
+    },
+    {
+      "epoch": 0.7658095472440944,
+      "grad_norm": 1.2463053464889526,
+      "learning_rate": 2.589636575984835e-05,
+      "loss": 1.8711,
+      "step": 12449
+    },
+    {
+      "epoch": 0.765871062992126,
+      "grad_norm": 1.230785846710205,
+      "learning_rate": 2.5883382599967153e-05,
+      "loss": 2.5934,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7659325787401575,
+      "grad_norm": 1.3742331266403198,
+      "learning_rate": 2.587040221163417e-05,
+      "loss": 2.1761,
+      "step": 12451
+    },
+    {
+      "epoch": 0.765994094488189,
+      "grad_norm": 1.0644816160202026,
+      "learning_rate": 2.585742459533479e-05,
+      "loss": 2.7616,
+      "step": 12452
+    },
+    {
+      "epoch": 0.7660556102362205,
+      "grad_norm": 1.7579659223556519,
+      "learning_rate": 2.5844449751554323e-05,
+      "loss": 2.0266,
+      "step": 12453
+    },
+    {
+      "epoch": 0.766117125984252,
+      "grad_norm": 1.110132098197937,
+      "learning_rate": 2.5831477680777905e-05,
+      "loss": 2.2923,
+      "step": 12454
+    },
+    {
+      "epoch": 0.7661786417322834,
+      "grad_norm": 1.1454174518585205,
+      "learning_rate": 2.581850838349068e-05,
+      "loss": 3.1287,
+      "step": 12455
+    },
+    {
+      "epoch": 0.7662401574803149,
+      "grad_norm": 1.2692084312438965,
+      "learning_rate": 2.580554186017756e-05,
+      "loss": 1.5769,
+      "step": 12456
+    },
+    {
+      "epoch": 0.7663016732283464,
+      "grad_norm": 1.472168207168579,
+      "learning_rate": 2.579257811132345e-05,
+      "loss": 2.2422,
+      "step": 12457
+    },
+    {
+      "epoch": 0.766363188976378,
+      "grad_norm": 1.2828521728515625,
+      "learning_rate": 2.5779617137413138e-05,
+      "loss": 2.6922,
+      "step": 12458
+    },
+    {
+      "epoch": 0.7664247047244095,
+      "grad_norm": 2.0303597450256348,
+      "learning_rate": 2.576665893893131e-05,
+      "loss": 2.7667,
+      "step": 12459
+    },
+    {
+      "epoch": 0.766486220472441,
+      "grad_norm": 1.3631747961044312,
+      "learning_rate": 2.575370351636246e-05,
+      "loss": 1.9738,
+      "step": 12460
+    },
+    {
+      "epoch": 0.7665477362204725,
+      "grad_norm": 1.6201263666152954,
+      "learning_rate": 2.5740750870191132e-05,
+      "loss": 2.1998,
+      "step": 12461
+    },
+    {
+      "epoch": 0.7666092519685039,
+      "grad_norm": 1.224955439567566,
+      "learning_rate": 2.5727801000901596e-05,
+      "loss": 2.8739,
+      "step": 12462
+    },
+    {
+      "epoch": 0.7666707677165354,
+      "grad_norm": 1.445063591003418,
+      "learning_rate": 2.5714853908978166e-05,
+      "loss": 2.3721,
+      "step": 12463
+    },
+    {
+      "epoch": 0.7667322834645669,
+      "grad_norm": 1.1583741903305054,
+      "learning_rate": 2.570190959490496e-05,
+      "loss": 2.5061,
+      "step": 12464
+    },
+    {
+      "epoch": 0.7667937992125984,
+      "grad_norm": 1.1933739185333252,
+      "learning_rate": 2.5688968059166075e-05,
+      "loss": 2.7343,
+      "step": 12465
+    },
+    {
+      "epoch": 0.76685531496063,
+      "grad_norm": 1.1673918962478638,
+      "learning_rate": 2.5676029302245386e-05,
+      "loss": 2.4885,
+      "step": 12466
+    },
+    {
+      "epoch": 0.7669168307086615,
+      "grad_norm": 1.3467117547988892,
+      "learning_rate": 2.5663093324626786e-05,
+      "loss": 1.7138,
+      "step": 12467
+    },
+    {
+      "epoch": 0.7669783464566929,
+      "grad_norm": 1.3110458850860596,
+      "learning_rate": 2.5650160126793955e-05,
+      "loss": 2.5788,
+      "step": 12468
+    },
+    {
+      "epoch": 0.7670398622047244,
+      "grad_norm": 1.350419044494629,
+      "learning_rate": 2.5637229709230548e-05,
+      "loss": 2.0888,
+      "step": 12469
+    },
+    {
+      "epoch": 0.7671013779527559,
+      "grad_norm": 1.2892991304397583,
+      "learning_rate": 2.562430207242009e-05,
+      "loss": 2.4845,
+      "step": 12470
+    },
+    {
+      "epoch": 0.7671628937007874,
+      "grad_norm": 1.240056037902832,
+      "learning_rate": 2.5611377216846032e-05,
+      "loss": 2.9101,
+      "step": 12471
+    },
+    {
+      "epoch": 0.7672244094488189,
+      "grad_norm": 1.2861526012420654,
+      "learning_rate": 2.5598455142991673e-05,
+      "loss": 1.7915,
+      "step": 12472
+    },
+    {
+      "epoch": 0.7672859251968503,
+      "grad_norm": 1.2508224248886108,
+      "learning_rate": 2.558553585134018e-05,
+      "loss": 1.8771,
+      "step": 12473
+    },
+    {
+      "epoch": 0.7673474409448819,
+      "grad_norm": 1.3159549236297607,
+      "learning_rate": 2.5572619342374727e-05,
+      "loss": 2.3411,
+      "step": 12474
+    },
+    {
+      "epoch": 0.7674089566929134,
+      "grad_norm": 1.2046922445297241,
+      "learning_rate": 2.5559705616578266e-05,
+      "loss": 2.5022,
+      "step": 12475
+    },
+    {
+      "epoch": 0.7674704724409449,
+      "grad_norm": 1.2706228494644165,
+      "learning_rate": 2.5546794674433726e-05,
+      "loss": 2.1569,
+      "step": 12476
+    },
+    {
+      "epoch": 0.7675319881889764,
+      "grad_norm": 2.0438952445983887,
+      "learning_rate": 2.553388651642391e-05,
+      "loss": 2.8306,
+      "step": 12477
+    },
+    {
+      "epoch": 0.7675935039370079,
+      "grad_norm": 1.491504192352295,
+      "learning_rate": 2.5520981143031518e-05,
+      "loss": 3.6252,
+      "step": 12478
+    },
+    {
+      "epoch": 0.7676550196850394,
+      "grad_norm": 1.2224547863006592,
+      "learning_rate": 2.55080785547391e-05,
+      "loss": 2.2505,
+      "step": 12479
+    },
+    {
+      "epoch": 0.7677165354330708,
+      "grad_norm": 1.2243478298187256,
+      "learning_rate": 2.5495178752029203e-05,
+      "loss": 1.8689,
+      "step": 12480
+    },
+    {
+      "epoch": 0.7677780511811023,
+      "grad_norm": 1.143164873123169,
+      "learning_rate": 2.5482281735384138e-05,
+      "loss": 2.3354,
+      "step": 12481
+    },
+    {
+      "epoch": 0.7678395669291339,
+      "grad_norm": 1.314723014831543,
+      "learning_rate": 2.5469387505286214e-05,
+      "loss": 2.5264,
+      "step": 12482
+    },
+    {
+      "epoch": 0.7679010826771654,
+      "grad_norm": 1.1125574111938477,
+      "learning_rate": 2.5456496062217594e-05,
+      "loss": 2.8574,
+      "step": 12483
+    },
+    {
+      "epoch": 0.7679625984251969,
+      "grad_norm": 1.1409457921981812,
+      "learning_rate": 2.5443607406660397e-05,
+      "loss": 2.891,
+      "step": 12484
+    },
+    {
+      "epoch": 0.7680241141732284,
+      "grad_norm": 1.3121589422225952,
+      "learning_rate": 2.5430721539096502e-05,
+      "loss": 2.272,
+      "step": 12485
+    },
+    {
+      "epoch": 0.7680856299212598,
+      "grad_norm": 1.4759961366653442,
+      "learning_rate": 2.541783846000785e-05,
+      "loss": 2.7297,
+      "step": 12486
+    },
+    {
+      "epoch": 0.7681471456692913,
+      "grad_norm": 1.3961162567138672,
+      "learning_rate": 2.5404958169876127e-05,
+      "loss": 2.31,
+      "step": 12487
+    },
+    {
+      "epoch": 0.7682086614173228,
+      "grad_norm": 1.3953959941864014,
+      "learning_rate": 2.5392080669183017e-05,
+      "loss": 2.2144,
+      "step": 12488
+    },
+    {
+      "epoch": 0.7682701771653543,
+      "grad_norm": 1.120513916015625,
+      "learning_rate": 2.5379205958410068e-05,
+      "loss": 2.6213,
+      "step": 12489
+    },
+    {
+      "epoch": 0.7683316929133859,
+      "grad_norm": 1.315743327140808,
+      "learning_rate": 2.5366334038038743e-05,
+      "loss": 2.6079,
+      "step": 12490
+    },
+    {
+      "epoch": 0.7683932086614174,
+      "grad_norm": 1.3847860097885132,
+      "learning_rate": 2.535346490855033e-05,
+      "loss": 2.4223,
+      "step": 12491
+    },
+    {
+      "epoch": 0.7684547244094488,
+      "grad_norm": 1.158681869506836,
+      "learning_rate": 2.534059857042611e-05,
+      "loss": 1.8824,
+      "step": 12492
+    },
+    {
+      "epoch": 0.7685162401574803,
+      "grad_norm": 1.667690396308899,
+      "learning_rate": 2.532773502414717e-05,
+      "loss": 2.3579,
+      "step": 12493
+    },
+    {
+      "epoch": 0.7685777559055118,
+      "grad_norm": 1.3437180519104004,
+      "learning_rate": 2.531487427019459e-05,
+      "loss": 2.6346,
+      "step": 12494
+    },
+    {
+      "epoch": 0.7686392716535433,
+      "grad_norm": 1.2326830625534058,
+      "learning_rate": 2.5302016309049183e-05,
+      "loss": 2.5906,
+      "step": 12495
+    },
+    {
+      "epoch": 0.7687007874015748,
+      "grad_norm": 1.4271893501281738,
+      "learning_rate": 2.5289161141191907e-05,
+      "loss": 2.6707,
+      "step": 12496
+    },
+    {
+      "epoch": 0.7687623031496063,
+      "grad_norm": 1.2029739618301392,
+      "learning_rate": 2.5276308767103373e-05,
+      "loss": 2.6995,
+      "step": 12497
+    },
+    {
+      "epoch": 0.7688238188976378,
+      "grad_norm": 1.3487439155578613,
+      "learning_rate": 2.526345918726426e-05,
+      "loss": 2.3168,
+      "step": 12498
+    },
+    {
+      "epoch": 0.7688853346456693,
+      "grad_norm": 2.018555164337158,
+      "learning_rate": 2.5250612402154995e-05,
+      "loss": 2.1266,
+      "step": 12499
+    },
+    {
+      "epoch": 0.7689468503937008,
+      "grad_norm": 1.1207687854766846,
+      "learning_rate": 2.5237768412256035e-05,
+      "loss": 2.883,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7690083661417323,
+      "grad_norm": 1.3576616048812866,
+      "learning_rate": 2.5224927218047623e-05,
+      "loss": 2.5113,
+      "step": 12501
+    },
+    {
+      "epoch": 0.7690698818897638,
+      "grad_norm": 1.330940842628479,
+      "learning_rate": 2.5212088820009983e-05,
+      "loss": 2.9818,
+      "step": 12502
+    },
+    {
+      "epoch": 0.7691313976377953,
+      "grad_norm": 1.210128903388977,
+      "learning_rate": 2.5199253218623187e-05,
+      "loss": 2.4922,
+      "step": 12503
+    },
+    {
+      "epoch": 0.7691929133858267,
+      "grad_norm": 1.840056300163269,
+      "learning_rate": 2.5186420414367252e-05,
+      "loss": 1.4937,
+      "step": 12504
+    },
+    {
+      "epoch": 0.7692544291338582,
+      "grad_norm": 2.2785120010375977,
+      "learning_rate": 2.517359040772199e-05,
+      "loss": 2.567,
+      "step": 12505
+    },
+    {
+      "epoch": 0.7693159448818898,
+      "grad_norm": 1.2045214176177979,
+      "learning_rate": 2.5160763199167238e-05,
+      "loss": 2.6119,
+      "step": 12506
+    },
+    {
+      "epoch": 0.7693774606299213,
+      "grad_norm": 1.0994946956634521,
+      "learning_rate": 2.514793878918259e-05,
+      "loss": 2.3135,
+      "step": 12507
+    },
+    {
+      "epoch": 0.7694389763779528,
+      "grad_norm": 1.1828445196151733,
+      "learning_rate": 2.5135117178247647e-05,
+      "loss": 2.7503,
+      "step": 12508
+    },
+    {
+      "epoch": 0.7695004921259843,
+      "grad_norm": 1.9715324640274048,
+      "learning_rate": 2.5122298366841902e-05,
+      "loss": 2.4643,
+      "step": 12509
+    },
+    {
+      "epoch": 0.7695620078740157,
+      "grad_norm": 1.265136957168579,
+      "learning_rate": 2.5109482355444625e-05,
+      "loss": 2.263,
+      "step": 12510
+    },
+    {
+      "epoch": 0.7696235236220472,
+      "grad_norm": 1.2767716646194458,
+      "learning_rate": 2.5096669144535146e-05,
+      "loss": 2.3856,
+      "step": 12511
+    },
+    {
+      "epoch": 0.7696850393700787,
+      "grad_norm": 1.296063780784607,
+      "learning_rate": 2.5083858734592536e-05,
+      "loss": 2.6308,
+      "step": 12512
+    },
+    {
+      "epoch": 0.7697465551181102,
+      "grad_norm": 1.160260796546936,
+      "learning_rate": 2.5071051126095868e-05,
+      "loss": 1.7085,
+      "step": 12513
+    },
+    {
+      "epoch": 0.7698080708661418,
+      "grad_norm": 1.268680214881897,
+      "learning_rate": 2.505824631952407e-05,
+      "loss": 2.2482,
+      "step": 12514
+    },
+    {
+      "epoch": 0.7698695866141733,
+      "grad_norm": 1.1597925424575806,
+      "learning_rate": 2.5045444315355993e-05,
+      "loss": 2.0579,
+      "step": 12515
+    },
+    {
+      "epoch": 0.7699311023622047,
+      "grad_norm": 1.268763542175293,
+      "learning_rate": 2.503264511407033e-05,
+      "loss": 2.5736,
+      "step": 12516
+    },
+    {
+      "epoch": 0.7699926181102362,
+      "grad_norm": 1.1385772228240967,
+      "learning_rate": 2.501984871614572e-05,
+      "loss": 2.535,
+      "step": 12517
+    },
+    {
+      "epoch": 0.7700541338582677,
+      "grad_norm": 1.1676610708236694,
+      "learning_rate": 2.5007055122060653e-05,
+      "loss": 2.5638,
+      "step": 12518
+    },
+    {
+      "epoch": 0.7701156496062992,
+      "grad_norm": 1.3950194120407104,
+      "learning_rate": 2.4994264332293584e-05,
+      "loss": 2.851,
+      "step": 12519
+    },
+    {
+      "epoch": 0.7701771653543307,
+      "grad_norm": 1.2881479263305664,
+      "learning_rate": 2.4981476347322742e-05,
+      "loss": 2.2012,
+      "step": 12520
+    },
+    {
+      "epoch": 0.7702386811023622,
+      "grad_norm": 1.264055609703064,
+      "learning_rate": 2.496869116762638e-05,
+      "loss": 2.6031,
+      "step": 12521
+    },
+    {
+      "epoch": 0.7703001968503937,
+      "grad_norm": 1.3180564641952515,
+      "learning_rate": 2.4955908793682582e-05,
+      "loss": 2.2863,
+      "step": 12522
+    },
+    {
+      "epoch": 0.7703617125984252,
+      "grad_norm": 1.105276346206665,
+      "learning_rate": 2.4943129225969354e-05,
+      "loss": 2.3472,
+      "step": 12523
+    },
+    {
+      "epoch": 0.7704232283464567,
+      "grad_norm": 1.32320237159729,
+      "learning_rate": 2.4930352464964547e-05,
+      "loss": 2.5216,
+      "step": 12524
+    },
+    {
+      "epoch": 0.7704847440944882,
+      "grad_norm": 1.3477294445037842,
+      "learning_rate": 2.491757851114599e-05,
+      "loss": 2.3385,
+      "step": 12525
+    },
+    {
+      "epoch": 0.7705462598425197,
+      "grad_norm": 1.2841030359268188,
+      "learning_rate": 2.490480736499129e-05,
+      "loss": 1.7084,
+      "step": 12526
+    },
+    {
+      "epoch": 0.7706077755905512,
+      "grad_norm": 1.3242231607437134,
+      "learning_rate": 2.4892039026978052e-05,
+      "loss": 2.4531,
+      "step": 12527
+    },
+    {
+      "epoch": 0.7706692913385826,
+      "grad_norm": 1.315011978149414,
+      "learning_rate": 2.4879273497583745e-05,
+      "loss": 2.1753,
+      "step": 12528
+    },
+    {
+      "epoch": 0.7707308070866141,
+      "grad_norm": 1.33985435962677,
+      "learning_rate": 2.4866510777285758e-05,
+      "loss": 3.2112,
+      "step": 12529
+    },
+    {
+      "epoch": 0.7707923228346457,
+      "grad_norm": 1.3704442977905273,
+      "learning_rate": 2.4853750866561277e-05,
+      "loss": 2.7839,
+      "step": 12530
+    },
+    {
+      "epoch": 0.7708538385826772,
+      "grad_norm": 1.268569827079773,
+      "learning_rate": 2.4840993765887512e-05,
+      "loss": 2.1646,
+      "step": 12531
+    },
+    {
+      "epoch": 0.7709153543307087,
+      "grad_norm": 1.4061428308486938,
+      "learning_rate": 2.4828239475741456e-05,
+      "loss": 3.2307,
+      "step": 12532
+    },
+    {
+      "epoch": 0.7709768700787402,
+      "grad_norm": 1.3367695808410645,
+      "learning_rate": 2.4815487996600074e-05,
+      "loss": 2.4976,
+      "step": 12533
+    },
+    {
+      "epoch": 0.7710383858267716,
+      "grad_norm": 1.0294123888015747,
+      "learning_rate": 2.4802739328940196e-05,
+      "loss": 2.8935,
+      "step": 12534
+    },
+    {
+      "epoch": 0.7710999015748031,
+      "grad_norm": 1.1790599822998047,
+      "learning_rate": 2.4789993473238594e-05,
+      "loss": 2.621,
+      "step": 12535
+    },
+    {
+      "epoch": 0.7711614173228346,
+      "grad_norm": 1.5232110023498535,
+      "learning_rate": 2.477725042997182e-05,
+      "loss": 1.6688,
+      "step": 12536
+    },
+    {
+      "epoch": 0.7712229330708661,
+      "grad_norm": 1.7374019622802734,
+      "learning_rate": 2.4764510199616454e-05,
+      "loss": 2.0447,
+      "step": 12537
+    },
+    {
+      "epoch": 0.7712844488188977,
+      "grad_norm": 1.3912395238876343,
+      "learning_rate": 2.4751772782648864e-05,
+      "loss": 1.7894,
+      "step": 12538
+    },
+    {
+      "epoch": 0.7713459645669292,
+      "grad_norm": 1.5154178142547607,
+      "learning_rate": 2.473903817954536e-05,
+      "loss": 2.5811,
+      "step": 12539
+    },
+    {
+      "epoch": 0.7714074803149606,
+      "grad_norm": 1.343747615814209,
+      "learning_rate": 2.472630639078217e-05,
+      "loss": 2.6122,
+      "step": 12540
+    },
+    {
+      "epoch": 0.7714689960629921,
+      "grad_norm": 1.534749150276184,
+      "learning_rate": 2.471357741683541e-05,
+      "loss": 2.245,
+      "step": 12541
+    },
+    {
+      "epoch": 0.7715305118110236,
+      "grad_norm": 1.1004406213760376,
+      "learning_rate": 2.4700851258181025e-05,
+      "loss": 1.184,
+      "step": 12542
+    },
+    {
+      "epoch": 0.7715920275590551,
+      "grad_norm": 1.3607484102249146,
+      "learning_rate": 2.4688127915294945e-05,
+      "loss": 2.5094,
+      "step": 12543
+    },
+    {
+      "epoch": 0.7716535433070866,
+      "grad_norm": 1.2048970460891724,
+      "learning_rate": 2.4675407388652926e-05,
+      "loss": 2.6396,
+      "step": 12544
+    },
+    {
+      "epoch": 0.7717150590551181,
+      "grad_norm": 1.867926836013794,
+      "learning_rate": 2.466268967873062e-05,
+      "loss": 2.0174,
+      "step": 12545
+    },
+    {
+      "epoch": 0.7717765748031497,
+      "grad_norm": 1.1136280298233032,
+      "learning_rate": 2.464997478600364e-05,
+      "loss": 2.6671,
+      "step": 12546
+    },
+    {
+      "epoch": 0.7718380905511811,
+      "grad_norm": 1.4027094841003418,
+      "learning_rate": 2.463726271094743e-05,
+      "loss": 2.2181,
+      "step": 12547
+    },
+    {
+      "epoch": 0.7718996062992126,
+      "grad_norm": 1.1538459062576294,
+      "learning_rate": 2.462455345403739e-05,
+      "loss": 2.853,
+      "step": 12548
+    },
+    {
+      "epoch": 0.7719611220472441,
+      "grad_norm": 1.2513765096664429,
+      "learning_rate": 2.4611847015748723e-05,
+      "loss": 2.4712,
+      "step": 12549
+    },
+    {
+      "epoch": 0.7720226377952756,
+      "grad_norm": 1.1755805015563965,
+      "learning_rate": 2.4599143396556622e-05,
+      "loss": 2.8251,
+      "step": 12550
+    },
+    {
+      "epoch": 0.7720841535433071,
+      "grad_norm": 1.3297518491744995,
+      "learning_rate": 2.458644259693609e-05,
+      "loss": 2.1352,
+      "step": 12551
+    },
+    {
+      "epoch": 0.7721456692913385,
+      "grad_norm": 1.4885890483856201,
+      "learning_rate": 2.4573744617362083e-05,
+      "loss": 2.3928,
+      "step": 12552
+    },
+    {
+      "epoch": 0.77220718503937,
+      "grad_norm": 1.3989919424057007,
+      "learning_rate": 2.4561049458309438e-05,
+      "loss": 2.6481,
+      "step": 12553
+    },
+    {
+      "epoch": 0.7722687007874016,
+      "grad_norm": 1.3157896995544434,
+      "learning_rate": 2.4548357120252918e-05,
+      "loss": 2.2706,
+      "step": 12554
+    },
+    {
+      "epoch": 0.7723302165354331,
+      "grad_norm": 1.1753439903259277,
+      "learning_rate": 2.453566760366708e-05,
+      "loss": 1.6178,
+      "step": 12555
+    },
+    {
+      "epoch": 0.7723917322834646,
+      "grad_norm": 1.2404921054840088,
+      "learning_rate": 2.4522980909026503e-05,
+      "loss": 2.4092,
+      "step": 12556
+    },
+    {
+      "epoch": 0.7724532480314961,
+      "grad_norm": 1.6572799682617188,
+      "learning_rate": 2.451029703680554e-05,
+      "loss": 1.7755,
+      "step": 12557
+    },
+    {
+      "epoch": 0.7725147637795275,
+      "grad_norm": 1.4854962825775146,
+      "learning_rate": 2.4497615987478516e-05,
+      "loss": 1.9376,
+      "step": 12558
+    },
+    {
+      "epoch": 0.772576279527559,
+      "grad_norm": 1.6977163553237915,
+      "learning_rate": 2.4484937761519656e-05,
+      "loss": 2.6713,
+      "step": 12559
+    },
+    {
+      "epoch": 0.7726377952755905,
+      "grad_norm": 1.364522099494934,
+      "learning_rate": 2.4472262359403065e-05,
+      "loss": 2.471,
+      "step": 12560
+    },
+    {
+      "epoch": 0.772699311023622,
+      "grad_norm": 1.3086196184158325,
+      "learning_rate": 2.4459589781602676e-05,
+      "loss": 2.2885,
+      "step": 12561
+    },
+    {
+      "epoch": 0.7727608267716536,
+      "grad_norm": 1.3950082063674927,
+      "learning_rate": 2.4446920028592433e-05,
+      "loss": 2.1589,
+      "step": 12562
+    },
+    {
+      "epoch": 0.7728223425196851,
+      "grad_norm": 1.2944353818893433,
+      "learning_rate": 2.443425310084606e-05,
+      "loss": 2.2234,
+      "step": 12563
+    },
+    {
+      "epoch": 0.7728838582677166,
+      "grad_norm": 1.357410192489624,
+      "learning_rate": 2.4421588998837274e-05,
+      "loss": 2.1761,
+      "step": 12564
+    },
+    {
+      "epoch": 0.772945374015748,
+      "grad_norm": 1.2813769578933716,
+      "learning_rate": 2.4408927723039575e-05,
+      "loss": 2.8664,
+      "step": 12565
+    },
+    {
+      "epoch": 0.7730068897637795,
+      "grad_norm": 1.40409255027771,
+      "learning_rate": 2.4396269273926532e-05,
+      "loss": 2.4793,
+      "step": 12566
+    },
+    {
+      "epoch": 0.773068405511811,
+      "grad_norm": 1.1196950674057007,
+      "learning_rate": 2.4383613651971404e-05,
+      "loss": 2.8931,
+      "step": 12567
+    },
+    {
+      "epoch": 0.7731299212598425,
+      "grad_norm": 1.562888264656067,
+      "learning_rate": 2.4370960857647508e-05,
+      "loss": 2.0616,
+      "step": 12568
+    },
+    {
+      "epoch": 0.773191437007874,
+      "grad_norm": 1.6832231283187866,
+      "learning_rate": 2.4358310891427927e-05,
+      "loss": 1.449,
+      "step": 12569
+    },
+    {
+      "epoch": 0.7732529527559056,
+      "grad_norm": 1.334938883781433,
+      "learning_rate": 2.434566375378575e-05,
+      "loss": 2.3646,
+      "step": 12570
+    },
+    {
+      "epoch": 0.773314468503937,
+      "grad_norm": 1.1576743125915527,
+      "learning_rate": 2.433301944519387e-05,
+      "loss": 2.4653,
+      "step": 12571
+    },
+    {
+      "epoch": 0.7733759842519685,
+      "grad_norm": 1.4233694076538086,
+      "learning_rate": 2.4320377966125128e-05,
+      "loss": 2.161,
+      "step": 12572
+    },
+    {
+      "epoch": 0.7734375,
+      "grad_norm": 1.316839337348938,
+      "learning_rate": 2.430773931705226e-05,
+      "loss": 2.5127,
+      "step": 12573
+    },
+    {
+      "epoch": 0.7734990157480315,
+      "grad_norm": 1.2649221420288086,
+      "learning_rate": 2.429510349844788e-05,
+      "loss": 2.42,
+      "step": 12574
+    },
+    {
+      "epoch": 0.773560531496063,
+      "grad_norm": 1.119444489479065,
+      "learning_rate": 2.428247051078447e-05,
+      "loss": 2.9033,
+      "step": 12575
+    },
+    {
+      "epoch": 0.7736220472440944,
+      "grad_norm": 1.3452377319335938,
+      "learning_rate": 2.426984035453448e-05,
+      "loss": 2.0223,
+      "step": 12576
+    },
+    {
+      "epoch": 0.773683562992126,
+      "grad_norm": 1.4136850833892822,
+      "learning_rate": 2.4257213030170146e-05,
+      "loss": 1.5129,
+      "step": 12577
+    },
+    {
+      "epoch": 0.7737450787401575,
+      "grad_norm": 1.227026343345642,
+      "learning_rate": 2.4244588538163693e-05,
+      "loss": 2.6485,
+      "step": 12578
+    },
+    {
+      "epoch": 0.773806594488189,
+      "grad_norm": 1.6707650423049927,
+      "learning_rate": 2.42319668789872e-05,
+      "loss": 2.6978,
+      "step": 12579
+    },
+    {
+      "epoch": 0.7738681102362205,
+      "grad_norm": 2.2144885063171387,
+      "learning_rate": 2.4219348053112677e-05,
+      "loss": 1.8816,
+      "step": 12580
+    },
+    {
+      "epoch": 0.773929625984252,
+      "grad_norm": 1.17075777053833,
+      "learning_rate": 2.4206732061011984e-05,
+      "loss": 2.6104,
+      "step": 12581
+    },
+    {
+      "epoch": 0.7739911417322834,
+      "grad_norm": 1.275229811668396,
+      "learning_rate": 2.4194118903156847e-05,
+      "loss": 1.5727,
+      "step": 12582
+    },
+    {
+      "epoch": 0.7740526574803149,
+      "grad_norm": 1.2303013801574707,
+      "learning_rate": 2.4181508580018984e-05,
+      "loss": 2.521,
+      "step": 12583
+    },
+    {
+      "epoch": 0.7741141732283464,
+      "grad_norm": 1.0589929819107056,
+      "learning_rate": 2.4168901092069863e-05,
+      "loss": 2.77,
+      "step": 12584
+    },
+    {
+      "epoch": 0.774175688976378,
+      "grad_norm": 1.3988171815872192,
+      "learning_rate": 2.4156296439781058e-05,
+      "loss": 2.2589,
+      "step": 12585
+    },
+    {
+      "epoch": 0.7742372047244095,
+      "grad_norm": 1.3318636417388916,
+      "learning_rate": 2.4143694623623814e-05,
+      "loss": 2.5763,
+      "step": 12586
+    },
+    {
+      "epoch": 0.774298720472441,
+      "grad_norm": 1.1848145723342896,
+      "learning_rate": 2.4131095644069446e-05,
+      "loss": 2.3498,
+      "step": 12587
+    },
+    {
+      "epoch": 0.7743602362204725,
+      "grad_norm": 1.2361299991607666,
+      "learning_rate": 2.4118499501588998e-05,
+      "loss": 2.3755,
+      "step": 12588
+    },
+    {
+      "epoch": 0.7744217519685039,
+      "grad_norm": 1.344258189201355,
+      "learning_rate": 2.410590619665358e-05,
+      "loss": 2.9881,
+      "step": 12589
+    },
+    {
+      "epoch": 0.7744832677165354,
+      "grad_norm": 1.6243754625320435,
+      "learning_rate": 2.4093315729734056e-05,
+      "loss": 2.1392,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7745447834645669,
+      "grad_norm": 1.361473798751831,
+      "learning_rate": 2.4080728101301242e-05,
+      "loss": 2.8313,
+      "step": 12591
+    },
+    {
+      "epoch": 0.7746062992125984,
+      "grad_norm": 1.419929027557373,
+      "learning_rate": 2.4068143311825864e-05,
+      "loss": 2.4709,
+      "step": 12592
+    },
+    {
+      "epoch": 0.77466781496063,
+      "grad_norm": 1.2229639291763306,
+      "learning_rate": 2.4055561361778543e-05,
+      "loss": 2.3458,
+      "step": 12593
+    },
+    {
+      "epoch": 0.7747293307086615,
+      "grad_norm": 1.5566822290420532,
+      "learning_rate": 2.4042982251629718e-05,
+      "loss": 2.4505,
+      "step": 12594
+    },
+    {
+      "epoch": 0.7747908464566929,
+      "grad_norm": 1.0611220598220825,
+      "learning_rate": 2.4030405981849845e-05,
+      "loss": 2.8751,
+      "step": 12595
+    },
+    {
+      "epoch": 0.7748523622047244,
+      "grad_norm": 1.2490371465682983,
+      "learning_rate": 2.401783255290915e-05,
+      "loss": 3.0444,
+      "step": 12596
+    },
+    {
+      "epoch": 0.7749138779527559,
+      "grad_norm": 1.2611677646636963,
+      "learning_rate": 2.4005261965277814e-05,
+      "loss": 2.5391,
+      "step": 12597
+    },
+    {
+      "epoch": 0.7749753937007874,
+      "grad_norm": 1.513735294342041,
+      "learning_rate": 2.3992694219425926e-05,
+      "loss": 2.2197,
+      "step": 12598
+    },
+    {
+      "epoch": 0.7750369094488189,
+      "grad_norm": 1.2781203985214233,
+      "learning_rate": 2.3980129315823475e-05,
+      "loss": 2.8731,
+      "step": 12599
+    },
+    {
+      "epoch": 0.7750984251968503,
+      "grad_norm": 1.3171786069869995,
+      "learning_rate": 2.3967567254940258e-05,
+      "loss": 2.2818,
+      "step": 12600
+    },
+    {
+      "epoch": 0.7751599409448819,
+      "grad_norm": 1.3200405836105347,
+      "learning_rate": 2.3955008037246097e-05,
+      "loss": 2.4687,
+      "step": 12601
+    },
+    {
+      "epoch": 0.7752214566929134,
+      "grad_norm": 1.39409339427948,
+      "learning_rate": 2.3942451663210564e-05,
+      "loss": 2.5878,
+      "step": 12602
+    },
+    {
+      "epoch": 0.7752829724409449,
+      "grad_norm": 1.5101338624954224,
+      "learning_rate": 2.392989813330322e-05,
+      "loss": 2.9786,
+      "step": 12603
+    },
+    {
+      "epoch": 0.7753444881889764,
+      "grad_norm": 1.6587945222854614,
+      "learning_rate": 2.391734744799351e-05,
+      "loss": 1.1174,
+      "step": 12604
+    },
+    {
+      "epoch": 0.7754060039370079,
+      "grad_norm": 1.2777998447418213,
+      "learning_rate": 2.3904799607750794e-05,
+      "loss": 2.6269,
+      "step": 12605
+    },
+    {
+      "epoch": 0.7754675196850394,
+      "grad_norm": 1.3118892908096313,
+      "learning_rate": 2.3892254613044218e-05,
+      "loss": 2.1942,
+      "step": 12606
+    },
+    {
+      "epoch": 0.7755290354330708,
+      "grad_norm": 1.2248913049697876,
+      "learning_rate": 2.3879712464342964e-05,
+      "loss": 2.2181,
+      "step": 12607
+    },
+    {
+      "epoch": 0.7755905511811023,
+      "grad_norm": 1.423987865447998,
+      "learning_rate": 2.3867173162115974e-05,
+      "loss": 2.0649,
+      "step": 12608
+    },
+    {
+      "epoch": 0.7756520669291339,
+      "grad_norm": 1.3389211893081665,
+      "learning_rate": 2.3854636706832212e-05,
+      "loss": 2.3726,
+      "step": 12609
+    },
+    {
+      "epoch": 0.7757135826771654,
+      "grad_norm": 1.5788817405700684,
+      "learning_rate": 2.384210309896038e-05,
+      "loss": 2.1254,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7757750984251969,
+      "grad_norm": 1.2401155233383179,
+      "learning_rate": 2.3829572338969287e-05,
+      "loss": 2.676,
+      "step": 12611
+    },
+    {
+      "epoch": 0.7758366141732284,
+      "grad_norm": 1.304577112197876,
+      "learning_rate": 2.3817044427327417e-05,
+      "loss": 2.6482,
+      "step": 12612
+    },
+    {
+      "epoch": 0.7758981299212598,
+      "grad_norm": 1.3702248334884644,
+      "learning_rate": 2.3804519364503296e-05,
+      "loss": 2.136,
+      "step": 12613
+    },
+    {
+      "epoch": 0.7759596456692913,
+      "grad_norm": 1.4225828647613525,
+      "learning_rate": 2.3791997150965263e-05,
+      "loss": 2.6643,
+      "step": 12614
+    },
+    {
+      "epoch": 0.7760211614173228,
+      "grad_norm": 1.4749395847320557,
+      "learning_rate": 2.3779477787181605e-05,
+      "loss": 2.49,
+      "step": 12615
+    },
+    {
+      "epoch": 0.7760826771653543,
+      "grad_norm": 1.5789858102798462,
+      "learning_rate": 2.3766961273620436e-05,
+      "loss": 2.7041,
+      "step": 12616
+    },
+    {
+      "epoch": 0.7761441929133859,
+      "grad_norm": 1.4443390369415283,
+      "learning_rate": 2.375444761074982e-05,
+      "loss": 2.137,
+      "step": 12617
+    },
+    {
+      "epoch": 0.7762057086614174,
+      "grad_norm": 1.1430456638336182,
+      "learning_rate": 2.3741936799037745e-05,
+      "loss": 2.6306,
+      "step": 12618
+    },
+    {
+      "epoch": 0.7762672244094488,
+      "grad_norm": 1.116387963294983,
+      "learning_rate": 2.3729428838951982e-05,
+      "loss": 2.3057,
+      "step": 12619
+    },
+    {
+      "epoch": 0.7763287401574803,
+      "grad_norm": 1.2001343965530396,
+      "learning_rate": 2.3716923730960305e-05,
+      "loss": 2.3335,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7763902559055118,
+      "grad_norm": 1.2112313508987427,
+      "learning_rate": 2.370442147553028e-05,
+      "loss": 2.7414,
+      "step": 12621
+    },
+    {
+      "epoch": 0.7764517716535433,
+      "grad_norm": 1.3347088098526,
+      "learning_rate": 2.369192207312947e-05,
+      "loss": 2.7698,
+      "step": 12622
+    },
+    {
+      "epoch": 0.7765132874015748,
+      "grad_norm": 1.2197576761245728,
+      "learning_rate": 2.3679425524225263e-05,
+      "loss": 2.763,
+      "step": 12623
+    },
+    {
+      "epoch": 0.7765748031496063,
+      "grad_norm": 1.1881568431854248,
+      "learning_rate": 2.3666931829284988e-05,
+      "loss": 2.6202,
+      "step": 12624
+    },
+    {
+      "epoch": 0.7766363188976378,
+      "grad_norm": 1.3025109767913818,
+      "learning_rate": 2.3654440988775794e-05,
+      "loss": 2.367,
+      "step": 12625
+    },
+    {
+      "epoch": 0.7766978346456693,
+      "grad_norm": 1.2061775922775269,
+      "learning_rate": 2.364195300316483e-05,
+      "loss": 2.2283,
+      "step": 12626
+    },
+    {
+      "epoch": 0.7767593503937008,
+      "grad_norm": 1.2375426292419434,
+      "learning_rate": 2.3629467872918997e-05,
+      "loss": 2.866,
+      "step": 12627
+    },
+    {
+      "epoch": 0.7768208661417323,
+      "grad_norm": 1.1760656833648682,
+      "learning_rate": 2.3616985598505227e-05,
+      "loss": 2.3826,
+      "step": 12628
+    },
+    {
+      "epoch": 0.7768823818897638,
+      "grad_norm": 1.1272270679473877,
+      "learning_rate": 2.3604506180390263e-05,
+      "loss": 2.4527,
+      "step": 12629
+    },
+    {
+      "epoch": 0.7769438976377953,
+      "grad_norm": 1.4077277183532715,
+      "learning_rate": 2.3592029619040802e-05,
+      "loss": 2.629,
+      "step": 12630
+    },
+    {
+      "epoch": 0.7770054133858267,
+      "grad_norm": 1.2250478267669678,
+      "learning_rate": 2.3579555914923347e-05,
+      "loss": 2.8346,
+      "step": 12631
+    },
+    {
+      "epoch": 0.7770669291338582,
+      "grad_norm": 1.2388429641723633,
+      "learning_rate": 2.3567085068504392e-05,
+      "loss": 1.684,
+      "step": 12632
+    },
+    {
+      "epoch": 0.7771284448818898,
+      "grad_norm": 1.305193305015564,
+      "learning_rate": 2.355461708025023e-05,
+      "loss": 2.5413,
+      "step": 12633
+    },
+    {
+      "epoch": 0.7771899606299213,
+      "grad_norm": 1.338988184928894,
+      "learning_rate": 2.3542151950627132e-05,
+      "loss": 2.7746,
+      "step": 12634
+    },
+    {
+      "epoch": 0.7772514763779528,
+      "grad_norm": 1.2770826816558838,
+      "learning_rate": 2.3529689680101198e-05,
+      "loss": 2.0583,
+      "step": 12635
+    },
+    {
+      "epoch": 0.7773129921259843,
+      "grad_norm": 1.1756641864776611,
+      "learning_rate": 2.3517230269138455e-05,
+      "loss": 2.8048,
+      "step": 12636
+    },
+    {
+      "epoch": 0.7773745078740157,
+      "grad_norm": 1.3228302001953125,
+      "learning_rate": 2.350477371820482e-05,
+      "loss": 2.0098,
+      "step": 12637
+    },
+    {
+      "epoch": 0.7774360236220472,
+      "grad_norm": 1.252831220626831,
+      "learning_rate": 2.3492320027766136e-05,
+      "loss": 2.6059,
+      "step": 12638
+    },
+    {
+      "epoch": 0.7774975393700787,
+      "grad_norm": 1.533451795578003,
+      "learning_rate": 2.3479869198288028e-05,
+      "loss": 2.3633,
+      "step": 12639
+    },
+    {
+      "epoch": 0.7775590551181102,
+      "grad_norm": 1.1937273740768433,
+      "learning_rate": 2.346742123023615e-05,
+      "loss": 2.9175,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7776205708661418,
+      "grad_norm": 1.3478251695632935,
+      "learning_rate": 2.3454976124075933e-05,
+      "loss": 2.9546,
+      "step": 12641
+    },
+    {
+      "epoch": 0.7776820866141733,
+      "grad_norm": 1.0751696825027466,
+      "learning_rate": 2.344253388027279e-05,
+      "loss": 2.5892,
+      "step": 12642
+    },
+    {
+      "epoch": 0.7777436023622047,
+      "grad_norm": 1.384627103805542,
+      "learning_rate": 2.3430094499291978e-05,
+      "loss": 3.2762,
+      "step": 12643
+    },
+    {
+      "epoch": 0.7778051181102362,
+      "grad_norm": 1.4812572002410889,
+      "learning_rate": 2.3417657981598697e-05,
+      "loss": 2.4549,
+      "step": 12644
+    },
+    {
+      "epoch": 0.7778666338582677,
+      "grad_norm": 1.4356160163879395,
+      "learning_rate": 2.3405224327657938e-05,
+      "loss": 3.0578,
+      "step": 12645
+    },
+    {
+      "epoch": 0.7779281496062992,
+      "grad_norm": 1.1808711290359497,
+      "learning_rate": 2.339279353793472e-05,
+      "loss": 2.8527,
+      "step": 12646
+    },
+    {
+      "epoch": 0.7779896653543307,
+      "grad_norm": 1.195343017578125,
+      "learning_rate": 2.3380365612893818e-05,
+      "loss": 2.7909,
+      "step": 12647
+    },
+    {
+      "epoch": 0.7780511811023622,
+      "grad_norm": 1.2965295314788818,
+      "learning_rate": 2.3367940553e-05,
+      "loss": 2.621,
+      "step": 12648
+    },
+    {
+      "epoch": 0.7781126968503937,
+      "grad_norm": 1.4322726726531982,
+      "learning_rate": 2.3355518358717887e-05,
+      "loss": 2.5306,
+      "step": 12649
+    },
+    {
+      "epoch": 0.7781742125984252,
+      "grad_norm": 1.2416120767593384,
+      "learning_rate": 2.334309903051204e-05,
+      "loss": 2.6703,
+      "step": 12650
+    },
+    {
+      "epoch": 0.7782357283464567,
+      "grad_norm": 0.9230075478553772,
+      "learning_rate": 2.3330682568846816e-05,
+      "loss": 1.8861,
+      "step": 12651
+    },
+    {
+      "epoch": 0.7782972440944882,
+      "grad_norm": 1.399463415145874,
+      "learning_rate": 2.3318268974186564e-05,
+      "loss": 2.4239,
+      "step": 12652
+    },
+    {
+      "epoch": 0.7783587598425197,
+      "grad_norm": 1.2923696041107178,
+      "learning_rate": 2.330585824699544e-05,
+      "loss": 2.6884,
+      "step": 12653
+    },
+    {
+      "epoch": 0.7784202755905512,
+      "grad_norm": 1.8387752771377563,
+      "learning_rate": 2.3293450387737558e-05,
+      "loss": 2.2161,
+      "step": 12654
+    },
+    {
+      "epoch": 0.7784817913385826,
+      "grad_norm": 1.4334815740585327,
+      "learning_rate": 2.328104539687693e-05,
+      "loss": 2.2008,
+      "step": 12655
+    },
+    {
+      "epoch": 0.7785433070866141,
+      "grad_norm": 1.4217314720153809,
+      "learning_rate": 2.3268643274877388e-05,
+      "loss": 2.3644,
+      "step": 12656
+    },
+    {
+      "epoch": 0.7786048228346457,
+      "grad_norm": 1.1175901889801025,
+      "learning_rate": 2.3256244022202746e-05,
+      "loss": 2.0438,
+      "step": 12657
+    },
+    {
+      "epoch": 0.7786663385826772,
+      "grad_norm": 1.491843819618225,
+      "learning_rate": 2.3243847639316607e-05,
+      "loss": 2.4313,
+      "step": 12658
+    },
+    {
+      "epoch": 0.7787278543307087,
+      "grad_norm": 1.4700809717178345,
+      "learning_rate": 2.3231454126682593e-05,
+      "loss": 2.3469,
+      "step": 12659
+    },
+    {
+      "epoch": 0.7787893700787402,
+      "grad_norm": 1.0915924310684204,
+      "learning_rate": 2.3219063484764103e-05,
+      "loss": 2.7791,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7788508858267716,
+      "grad_norm": 1.2550650835037231,
+      "learning_rate": 2.3206675714024483e-05,
+      "loss": 2.2856,
+      "step": 12661
+    },
+    {
+      "epoch": 0.7789124015748031,
+      "grad_norm": 1.3316518068313599,
+      "learning_rate": 2.319429081492699e-05,
+      "loss": 2.8609,
+      "step": 12662
+    },
+    {
+      "epoch": 0.7789739173228346,
+      "grad_norm": 1.0814074277877808,
+      "learning_rate": 2.3181908787934758e-05,
+      "loss": 2.9725,
+      "step": 12663
+    },
+    {
+      "epoch": 0.7790354330708661,
+      "grad_norm": 1.1905925273895264,
+      "learning_rate": 2.3169529633510766e-05,
+      "loss": 2.5883,
+      "step": 12664
+    },
+    {
+      "epoch": 0.7790969488188977,
+      "grad_norm": 1.2646161317825317,
+      "learning_rate": 2.3157153352117977e-05,
+      "loss": 1.6198,
+      "step": 12665
+    },
+    {
+      "epoch": 0.7791584645669292,
+      "grad_norm": 1.1933616399765015,
+      "learning_rate": 2.314477994421914e-05,
+      "loss": 2.9773,
+      "step": 12666
+    },
+    {
+      "epoch": 0.7792199803149606,
+      "grad_norm": 1.2367888689041138,
+      "learning_rate": 2.3132409410276966e-05,
+      "loss": 2.6199,
+      "step": 12667
+    },
+    {
+      "epoch": 0.7792814960629921,
+      "grad_norm": 1.3217194080352783,
+      "learning_rate": 2.312004175075406e-05,
+      "loss": 2.2708,
+      "step": 12668
+    },
+    {
+      "epoch": 0.7793430118110236,
+      "grad_norm": 1.2766157388687134,
+      "learning_rate": 2.3107676966112935e-05,
+      "loss": 2.1438,
+      "step": 12669
+    },
+    {
+      "epoch": 0.7794045275590551,
+      "grad_norm": 1.452368140220642,
+      "learning_rate": 2.309531505681589e-05,
+      "loss": 2.9892,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7794660433070866,
+      "grad_norm": 1.1740721464157104,
+      "learning_rate": 2.3082956023325263e-05,
+      "loss": 1.9202,
+      "step": 12671
+    },
+    {
+      "epoch": 0.7795275590551181,
+      "grad_norm": 1.584816336631775,
+      "learning_rate": 2.3070599866103148e-05,
+      "loss": 2.9933,
+      "step": 12672
+    },
+    {
+      "epoch": 0.7795890748031497,
+      "grad_norm": 1.273175597190857,
+      "learning_rate": 2.3058246585611632e-05,
+      "loss": 3.2586,
+      "step": 12673
+    },
+    {
+      "epoch": 0.7796505905511811,
+      "grad_norm": 1.55315363407135,
+      "learning_rate": 2.3045896182312642e-05,
+      "loss": 2.3966,
+      "step": 12674
+    },
+    {
+      "epoch": 0.7797121062992126,
+      "grad_norm": 1.171857237815857,
+      "learning_rate": 2.303354865666807e-05,
+      "loss": 2.5632,
+      "step": 12675
+    },
+    {
+      "epoch": 0.7797736220472441,
+      "grad_norm": 1.2977532148361206,
+      "learning_rate": 2.3021204009139563e-05,
+      "loss": 2.3089,
+      "step": 12676
+    },
+    {
+      "epoch": 0.7798351377952756,
+      "grad_norm": 1.3328734636306763,
+      "learning_rate": 2.300886224018881e-05,
+      "loss": 2.4039,
+      "step": 12677
+    },
+    {
+      "epoch": 0.7798966535433071,
+      "grad_norm": 1.4718269109725952,
+      "learning_rate": 2.2996523350277277e-05,
+      "loss": 2.7059,
+      "step": 12678
+    },
+    {
+      "epoch": 0.7799581692913385,
+      "grad_norm": 1.512586236000061,
+      "learning_rate": 2.2984187339866404e-05,
+      "loss": 2.256,
+      "step": 12679
+    },
+    {
+      "epoch": 0.78001968503937,
+      "grad_norm": 1.4767831563949585,
+      "learning_rate": 2.297185420941742e-05,
+      "loss": 2.8548,
+      "step": 12680
+    },
+    {
+      "epoch": 0.7800812007874016,
+      "grad_norm": 1.077396035194397,
+      "learning_rate": 2.2959523959391626e-05,
+      "loss": 2.7613,
+      "step": 12681
+    },
+    {
+      "epoch": 0.7801427165354331,
+      "grad_norm": 1.3728997707366943,
+      "learning_rate": 2.294719659025002e-05,
+      "loss": 2.683,
+      "step": 12682
+    },
+    {
+      "epoch": 0.7802042322834646,
+      "grad_norm": 1.3152856826782227,
+      "learning_rate": 2.2934872102453632e-05,
+      "loss": 2.6787,
+      "step": 12683
+    },
+    {
+      "epoch": 0.7802657480314961,
+      "grad_norm": 1.316115140914917,
+      "learning_rate": 2.292255049646327e-05,
+      "loss": 2.7332,
+      "step": 12684
+    },
+    {
+      "epoch": 0.7803272637795275,
+      "grad_norm": 1.4300413131713867,
+      "learning_rate": 2.2910231772739743e-05,
+      "loss": 2.7496,
+      "step": 12685
+    },
+    {
+      "epoch": 0.780388779527559,
+      "grad_norm": 1.3615635633468628,
+      "learning_rate": 2.2897915931743663e-05,
+      "loss": 1.9489,
+      "step": 12686
+    },
+    {
+      "epoch": 0.7804502952755905,
+      "grad_norm": 1.346184253692627,
+      "learning_rate": 2.2885602973935595e-05,
+      "loss": 2.3887,
+      "step": 12687
+    },
+    {
+      "epoch": 0.780511811023622,
+      "grad_norm": 1.4321609735488892,
+      "learning_rate": 2.2873292899775976e-05,
+      "loss": 2.3401,
+      "step": 12688
+    },
+    {
+      "epoch": 0.7805733267716536,
+      "grad_norm": 1.4070417881011963,
+      "learning_rate": 2.286098570972516e-05,
+      "loss": 2.8855,
+      "step": 12689
+    },
+    {
+      "epoch": 0.7806348425196851,
+      "grad_norm": 1.1585370302200317,
+      "learning_rate": 2.284868140424331e-05,
+      "loss": 2.7701,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7806963582677166,
+      "grad_norm": 1.1841137409210205,
+      "learning_rate": 2.283637998379059e-05,
+      "loss": 2.4238,
+      "step": 12691
+    },
+    {
+      "epoch": 0.780757874015748,
+      "grad_norm": 1.1318453550338745,
+      "learning_rate": 2.2824081448826962e-05,
+      "loss": 2.9256,
+      "step": 12692
+    },
+    {
+      "epoch": 0.7808193897637795,
+      "grad_norm": 3.2751522064208984,
+      "learning_rate": 2.2811785799812334e-05,
+      "loss": 2.3226,
+      "step": 12693
+    },
+    {
+      "epoch": 0.780880905511811,
+      "grad_norm": 1.6884108781814575,
+      "learning_rate": 2.2799493037206544e-05,
+      "loss": 2.7887,
+      "step": 12694
+    },
+    {
+      "epoch": 0.7809424212598425,
+      "grad_norm": 1.3511431217193604,
+      "learning_rate": 2.2787203161469195e-05,
+      "loss": 3.3797,
+      "step": 12695
+    },
+    {
+      "epoch": 0.781003937007874,
+      "grad_norm": 1.5997170209884644,
+      "learning_rate": 2.2774916173059936e-05,
+      "loss": 2.467,
+      "step": 12696
+    },
+    {
+      "epoch": 0.7810654527559056,
+      "grad_norm": 1.4545574188232422,
+      "learning_rate": 2.276263207243815e-05,
+      "loss": 1.7537,
+      "step": 12697
+    },
+    {
+      "epoch": 0.781126968503937,
+      "grad_norm": 1.2247586250305176,
+      "learning_rate": 2.2750350860063273e-05,
+      "loss": 2.95,
+      "step": 12698
+    },
+    {
+      "epoch": 0.7811884842519685,
+      "grad_norm": 1.1950867176055908,
+      "learning_rate": 2.273807253639447e-05,
+      "loss": 2.0868,
+      "step": 12699
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 1.5014487504959106,
+      "learning_rate": 2.272579710189098e-05,
+      "loss": 2.5428,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7813115157480315,
+      "grad_norm": 1.6025679111480713,
+      "learning_rate": 2.2713524557011757e-05,
+      "loss": 2.8163,
+      "step": 12701
+    },
+    {
+      "epoch": 0.781373031496063,
+      "grad_norm": 1.2113839387893677,
+      "learning_rate": 2.2701254902215775e-05,
+      "loss": 2.2746,
+      "step": 12702
+    },
+    {
+      "epoch": 0.7814345472440944,
+      "grad_norm": 2.320495367050171,
+      "learning_rate": 2.268898813796181e-05,
+      "loss": 2.0644,
+      "step": 12703
+    },
+    {
+      "epoch": 0.781496062992126,
+      "grad_norm": 1.431241750717163,
+      "learning_rate": 2.2676724264708616e-05,
+      "loss": 2.1444,
+      "step": 12704
+    },
+    {
+      "epoch": 0.7815575787401575,
+      "grad_norm": 1.3318101167678833,
+      "learning_rate": 2.266446328291474e-05,
+      "loss": 2.9754,
+      "step": 12705
+    },
+    {
+      "epoch": 0.781619094488189,
+      "grad_norm": 1.3765008449554443,
+      "learning_rate": 2.265220519303871e-05,
+      "loss": 2.4136,
+      "step": 12706
+    },
+    {
+      "epoch": 0.7816806102362205,
+      "grad_norm": 1.4289578199386597,
+      "learning_rate": 2.263994999553889e-05,
+      "loss": 2.171,
+      "step": 12707
+    },
+    {
+      "epoch": 0.781742125984252,
+      "grad_norm": 1.3000191450119019,
+      "learning_rate": 2.2627697690873606e-05,
+      "loss": 2.5967,
+      "step": 12708
+    },
+    {
+      "epoch": 0.7818036417322834,
+      "grad_norm": 1.1050853729248047,
+      "learning_rate": 2.2615448279500962e-05,
+      "loss": 2.5192,
+      "step": 12709
+    },
+    {
+      "epoch": 0.7818651574803149,
+      "grad_norm": 1.3721330165863037,
+      "learning_rate": 2.260320176187907e-05,
+      "loss": 1.915,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7819266732283464,
+      "grad_norm": 1.3283506631851196,
+      "learning_rate": 2.2590958138465824e-05,
+      "loss": 2.1,
+      "step": 12711
+    },
+    {
+      "epoch": 0.781988188976378,
+      "grad_norm": 1.4251114130020142,
+      "learning_rate": 2.2578717409719097e-05,
+      "loss": 2.5662,
+      "step": 12712
+    },
+    {
+      "epoch": 0.7820497047244095,
+      "grad_norm": 1.246773362159729,
+      "learning_rate": 2.2566479576096632e-05,
+      "loss": 1.9586,
+      "step": 12713
+    },
+    {
+      "epoch": 0.782111220472441,
+      "grad_norm": 1.262018084526062,
+      "learning_rate": 2.2554244638056077e-05,
+      "loss": 2.0455,
+      "step": 12714
+    },
+    {
+      "epoch": 0.7821727362204725,
+      "grad_norm": 1.2271478176116943,
+      "learning_rate": 2.2542012596054884e-05,
+      "loss": 2.1158,
+      "step": 12715
+    },
+    {
+      "epoch": 0.7822342519685039,
+      "grad_norm": 1.3130266666412354,
+      "learning_rate": 2.2529783450550546e-05,
+      "loss": 2.8128,
+      "step": 12716
+    },
+    {
+      "epoch": 0.7822957677165354,
+      "grad_norm": 1.9315967559814453,
+      "learning_rate": 2.2517557202000284e-05,
+      "loss": 2.0668,
+      "step": 12717
+    },
+    {
+      "epoch": 0.7823572834645669,
+      "grad_norm": 1.3976874351501465,
+      "learning_rate": 2.2505333850861333e-05,
+      "loss": 2.3709,
+      "step": 12718
+    },
+    {
+      "epoch": 0.7824187992125984,
+      "grad_norm": 1.1874810457229614,
+      "learning_rate": 2.249311339759077e-05,
+      "loss": 2.5513,
+      "step": 12719
+    },
+    {
+      "epoch": 0.78248031496063,
+      "grad_norm": 1.184054970741272,
+      "learning_rate": 2.248089584264561e-05,
+      "loss": 2.8601,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7825418307086615,
+      "grad_norm": 1.4041991233825684,
+      "learning_rate": 2.2468681186482654e-05,
+      "loss": 2.412,
+      "step": 12721
+    },
+    {
+      "epoch": 0.7826033464566929,
+      "grad_norm": 1.8719768524169922,
+      "learning_rate": 2.2456469429558725e-05,
+      "loss": 1.7855,
+      "step": 12722
+    },
+    {
+      "epoch": 0.7826648622047244,
+      "grad_norm": 1.258997917175293,
+      "learning_rate": 2.244426057233041e-05,
+      "loss": 3.0674,
+      "step": 12723
+    },
+    {
+      "epoch": 0.7827263779527559,
+      "grad_norm": 1.2964032888412476,
+      "learning_rate": 2.2432054615254327e-05,
+      "loss": 2.4757,
+      "step": 12724
+    },
+    {
+      "epoch": 0.7827878937007874,
+      "grad_norm": 1.1949800252914429,
+      "learning_rate": 2.241985155878681e-05,
+      "loss": 2.8283,
+      "step": 12725
+    },
+    {
+      "epoch": 0.7828494094488189,
+      "grad_norm": 1.313338279724121,
+      "learning_rate": 2.24076514033843e-05,
+      "loss": 2.434,
+      "step": 12726
+    },
+    {
+      "epoch": 0.7829109251968503,
+      "grad_norm": 1.4273321628570557,
+      "learning_rate": 2.2395454149502938e-05,
+      "loss": 2.3853,
+      "step": 12727
+    },
+    {
+      "epoch": 0.7829724409448819,
+      "grad_norm": 1.3070664405822754,
+      "learning_rate": 2.2383259797598875e-05,
+      "loss": 2.5566,
+      "step": 12728
+    },
+    {
+      "epoch": 0.7830339566929134,
+      "grad_norm": 1.295442819595337,
+      "learning_rate": 2.2371068348128098e-05,
+      "loss": 2.3926,
+      "step": 12729
+    },
+    {
+      "epoch": 0.7830954724409449,
+      "grad_norm": 1.0626894235610962,
+      "learning_rate": 2.235887980154646e-05,
+      "loss": 2.4944,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7831569881889764,
+      "grad_norm": 1.2439215183258057,
+      "learning_rate": 2.2346694158309788e-05,
+      "loss": 2.8438,
+      "step": 12731
+    },
+    {
+      "epoch": 0.7832185039370079,
+      "grad_norm": 1.169219970703125,
+      "learning_rate": 2.2334511418873738e-05,
+      "loss": 2.0559,
+      "step": 12732
+    },
+    {
+      "epoch": 0.7832800196850394,
+      "grad_norm": 1.8684487342834473,
+      "learning_rate": 2.2322331583693924e-05,
+      "loss": 2.198,
+      "step": 12733
+    },
+    {
+      "epoch": 0.7833415354330708,
+      "grad_norm": 2.150590181350708,
+      "learning_rate": 2.2310154653225733e-05,
+      "loss": 1.8264,
+      "step": 12734
+    },
+    {
+      "epoch": 0.7834030511811023,
+      "grad_norm": 1.136293649673462,
+      "learning_rate": 2.2297980627924576e-05,
+      "loss": 2.8856,
+      "step": 12735
+    },
+    {
+      "epoch": 0.7834645669291339,
+      "grad_norm": 1.491180658340454,
+      "learning_rate": 2.2285809508245637e-05,
+      "loss": 2.215,
+      "step": 12736
+    },
+    {
+      "epoch": 0.7835260826771654,
+      "grad_norm": 1.2195098400115967,
+      "learning_rate": 2.2273641294644075e-05,
+      "loss": 3.0288,
+      "step": 12737
+    },
+    {
+      "epoch": 0.7835875984251969,
+      "grad_norm": 1.3956141471862793,
+      "learning_rate": 2.226147598757492e-05,
+      "loss": 2.5101,
+      "step": 12738
+    },
+    {
+      "epoch": 0.7836491141732284,
+      "grad_norm": 1.2553534507751465,
+      "learning_rate": 2.2249313587493104e-05,
+      "loss": 2.7902,
+      "step": 12739
+    },
+    {
+      "epoch": 0.7837106299212598,
+      "grad_norm": 1.1313070058822632,
+      "learning_rate": 2.2237154094853384e-05,
+      "loss": 2.4264,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7837721456692913,
+      "grad_norm": 1.3982542753219604,
+      "learning_rate": 2.2224997510110512e-05,
+      "loss": 2.21,
+      "step": 12741
+    },
+    {
+      "epoch": 0.7838336614173228,
+      "grad_norm": 1.3233275413513184,
+      "learning_rate": 2.2212843833719022e-05,
+      "loss": 1.5142,
+      "step": 12742
+    },
+    {
+      "epoch": 0.7838951771653543,
+      "grad_norm": 1.2544862031936646,
+      "learning_rate": 2.2200693066133415e-05,
+      "loss": 2.3346,
+      "step": 12743
+    },
+    {
+      "epoch": 0.7839566929133859,
+      "grad_norm": 1.2789925336837769,
+      "learning_rate": 2.2188545207808064e-05,
+      "loss": 2.3822,
+      "step": 12744
+    },
+    {
+      "epoch": 0.7840182086614174,
+      "grad_norm": 1.057080864906311,
+      "learning_rate": 2.2176400259197272e-05,
+      "loss": 3.056,
+      "step": 12745
+    },
+    {
+      "epoch": 0.7840797244094488,
+      "grad_norm": 1.2054427862167358,
+      "learning_rate": 2.2164258220755118e-05,
+      "loss": 3.0179,
+      "step": 12746
+    },
+    {
+      "epoch": 0.7841412401574803,
+      "grad_norm": 1.1744166612625122,
+      "learning_rate": 2.215211909293571e-05,
+      "loss": 2.8799,
+      "step": 12747
+    },
+    {
+      "epoch": 0.7842027559055118,
+      "grad_norm": 1.3034284114837646,
+      "learning_rate": 2.2139982876192933e-05,
+      "loss": 2.9144,
+      "step": 12748
+    },
+    {
+      "epoch": 0.7842642716535433,
+      "grad_norm": 1.404818058013916,
+      "learning_rate": 2.2127849570980664e-05,
+      "loss": 1.6748,
+      "step": 12749
+    },
+    {
+      "epoch": 0.7843257874015748,
+      "grad_norm": 1.3183417320251465,
+      "learning_rate": 2.211571917775256e-05,
+      "loss": 2.3764,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7843873031496063,
+      "grad_norm": 1.3093316555023193,
+      "learning_rate": 2.210359169696227e-05,
+      "loss": 2.6374,
+      "step": 12751
+    },
+    {
+      "epoch": 0.7844488188976378,
+      "grad_norm": 1.3519234657287598,
+      "learning_rate": 2.209146712906328e-05,
+      "loss": 2.3225,
+      "step": 12752
+    },
+    {
+      "epoch": 0.7845103346456693,
+      "grad_norm": 1.339794635772705,
+      "learning_rate": 2.2079345474509026e-05,
+      "loss": 2.0493,
+      "step": 12753
+    },
+    {
+      "epoch": 0.7845718503937008,
+      "grad_norm": 1.3772575855255127,
+      "learning_rate": 2.206722673375273e-05,
+      "loss": 2.2464,
+      "step": 12754
+    },
+    {
+      "epoch": 0.7846333661417323,
+      "grad_norm": 1.1763056516647339,
+      "learning_rate": 2.205511090724761e-05,
+      "loss": 2.6527,
+      "step": 12755
+    },
+    {
+      "epoch": 0.7846948818897638,
+      "grad_norm": 1.6303354501724243,
+      "learning_rate": 2.204299799544668e-05,
+      "loss": 2.4281,
+      "step": 12756
+    },
+    {
+      "epoch": 0.7847563976377953,
+      "grad_norm": 1.2989119291305542,
+      "learning_rate": 2.203088799880293e-05,
+      "loss": 2.2099,
+      "step": 12757
+    },
+    {
+      "epoch": 0.7848179133858267,
+      "grad_norm": 1.1517181396484375,
+      "learning_rate": 2.20187809177692e-05,
+      "loss": 2.7334,
+      "step": 12758
+    },
+    {
+      "epoch": 0.7848794291338582,
+      "grad_norm": 1.2421023845672607,
+      "learning_rate": 2.200667675279825e-05,
+      "loss": 3.343,
+      "step": 12759
+    },
+    {
+      "epoch": 0.7849409448818898,
+      "grad_norm": 1.2110435962677002,
+      "learning_rate": 2.1994575504342664e-05,
+      "loss": 2.639,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7850024606299213,
+      "grad_norm": 1.35634183883667,
+      "learning_rate": 2.198247717285501e-05,
+      "loss": 2.9017,
+      "step": 12761
+    },
+    {
+      "epoch": 0.7850639763779528,
+      "grad_norm": 1.3180335760116577,
+      "learning_rate": 2.197038175878765e-05,
+      "loss": 1.9541,
+      "step": 12762
+    },
+    {
+      "epoch": 0.7851254921259843,
+      "grad_norm": 1.549756407737732,
+      "learning_rate": 2.19582892625929e-05,
+      "loss": 2.599,
+      "step": 12763
+    },
+    {
+      "epoch": 0.7851870078740157,
+      "grad_norm": 1.1524138450622559,
+      "learning_rate": 2.1946199684722967e-05,
+      "loss": 2.188,
+      "step": 12764
+    },
+    {
+      "epoch": 0.7852485236220472,
+      "grad_norm": 1.4090490341186523,
+      "learning_rate": 2.1934113025629943e-05,
+      "loss": 2.2402,
+      "step": 12765
+    },
+    {
+      "epoch": 0.7853100393700787,
+      "grad_norm": 1.2141872644424438,
+      "learning_rate": 2.19220292857658e-05,
+      "loss": 1.97,
+      "step": 12766
+    },
+    {
+      "epoch": 0.7853715551181102,
+      "grad_norm": 1.1773533821105957,
+      "learning_rate": 2.1909948465582342e-05,
+      "loss": 2.6845,
+      "step": 12767
+    },
+    {
+      "epoch": 0.7854330708661418,
+      "grad_norm": 1.4730557203292847,
+      "learning_rate": 2.1897870565531407e-05,
+      "loss": 2.568,
+      "step": 12768
+    },
+    {
+      "epoch": 0.7854945866141733,
+      "grad_norm": 1.645254135131836,
+      "learning_rate": 2.188579558606455e-05,
+      "loss": 2.3927,
+      "step": 12769
+    },
+    {
+      "epoch": 0.7855561023622047,
+      "grad_norm": 1.3156942129135132,
+      "learning_rate": 2.1873723527633404e-05,
+      "loss": 2.5127,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7856176181102362,
+      "grad_norm": 1.2736717462539673,
+      "learning_rate": 2.1861654390689334e-05,
+      "loss": 3.1228,
+      "step": 12771
+    },
+    {
+      "epoch": 0.7856791338582677,
+      "grad_norm": 1.2354490756988525,
+      "learning_rate": 2.1849588175683712e-05,
+      "loss": 2.0722,
+      "step": 12772
+    },
+    {
+      "epoch": 0.7857406496062992,
+      "grad_norm": 1.2948546409606934,
+      "learning_rate": 2.1837524883067685e-05,
+      "loss": 2.7698,
+      "step": 12773
+    },
+    {
+      "epoch": 0.7858021653543307,
+      "grad_norm": 1.334715723991394,
+      "learning_rate": 2.1825464513292403e-05,
+      "loss": 3.0061,
+      "step": 12774
+    },
+    {
+      "epoch": 0.7858636811023622,
+      "grad_norm": 1.195112943649292,
+      "learning_rate": 2.1813407066808798e-05,
+      "loss": 2.3373,
+      "step": 12775
+    },
+    {
+      "epoch": 0.7859251968503937,
+      "grad_norm": 1.4758480787277222,
+      "learning_rate": 2.1801352544067812e-05,
+      "loss": 3.3413,
+      "step": 12776
+    },
+    {
+      "epoch": 0.7859867125984252,
+      "grad_norm": 1.3459298610687256,
+      "learning_rate": 2.1789300945520173e-05,
+      "loss": 2.2029,
+      "step": 12777
+    },
+    {
+      "epoch": 0.7860482283464567,
+      "grad_norm": 1.1737968921661377,
+      "learning_rate": 2.17772522716166e-05,
+      "loss": 1.917,
+      "step": 12778
+    },
+    {
+      "epoch": 0.7861097440944882,
+      "grad_norm": 1.4249787330627441,
+      "learning_rate": 2.176520652280758e-05,
+      "loss": 2.309,
+      "step": 12779
+    },
+    {
+      "epoch": 0.7861712598425197,
+      "grad_norm": 1.2008780241012573,
+      "learning_rate": 2.1753163699543622e-05,
+      "loss": 3.0976,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7862327755905512,
+      "grad_norm": 1.238271951675415,
+      "learning_rate": 2.1741123802274988e-05,
+      "loss": 2.4956,
+      "step": 12781
+    },
+    {
+      "epoch": 0.7862942913385826,
+      "grad_norm": 1.261231780052185,
+      "learning_rate": 2.172908683145195e-05,
+      "loss": 2.5126,
+      "step": 12782
+    },
+    {
+      "epoch": 0.7863558070866141,
+      "grad_norm": 1.3411625623703003,
+      "learning_rate": 2.1717052787524605e-05,
+      "loss": 2.2225,
+      "step": 12783
+    },
+    {
+      "epoch": 0.7864173228346457,
+      "grad_norm": 1.3011184930801392,
+      "learning_rate": 2.1705021670942994e-05,
+      "loss": 2.0796,
+      "step": 12784
+    },
+    {
+      "epoch": 0.7864788385826772,
+      "grad_norm": 1.228279948234558,
+      "learning_rate": 2.1692993482156966e-05,
+      "loss": 2.6569,
+      "step": 12785
+    },
+    {
+      "epoch": 0.7865403543307087,
+      "grad_norm": 1.3273593187332153,
+      "learning_rate": 2.168096822161636e-05,
+      "loss": 2.1469,
+      "step": 12786
+    },
+    {
+      "epoch": 0.7866018700787402,
+      "grad_norm": 1.1115565299987793,
+      "learning_rate": 2.1668945889770786e-05,
+      "loss": 2.8188,
+      "step": 12787
+    },
+    {
+      "epoch": 0.7866633858267716,
+      "grad_norm": 1.3492381572723389,
+      "learning_rate": 2.165692648706986e-05,
+      "loss": 2.2787,
+      "step": 12788
+    },
+    {
+      "epoch": 0.7867249015748031,
+      "grad_norm": 1.2288682460784912,
+      "learning_rate": 2.1644910013963027e-05,
+      "loss": 2.3777,
+      "step": 12789
+    },
+    {
+      "epoch": 0.7867864173228346,
+      "grad_norm": 1.1987385749816895,
+      "learning_rate": 2.1632896470899666e-05,
+      "loss": 2.2448,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7868479330708661,
+      "grad_norm": 1.4388203620910645,
+      "learning_rate": 2.1620885858328966e-05,
+      "loss": 1.9636,
+      "step": 12791
+    },
+    {
+      "epoch": 0.7869094488188977,
+      "grad_norm": 1.276955485343933,
+      "learning_rate": 2.1608878176700098e-05,
+      "loss": 1.9154,
+      "step": 12792
+    },
+    {
+      "epoch": 0.7869709645669292,
+      "grad_norm": 1.4449435472488403,
+      "learning_rate": 2.159687342646205e-05,
+      "loss": 2.0987,
+      "step": 12793
+    },
+    {
+      "epoch": 0.7870324803149606,
+      "grad_norm": 1.1365503072738647,
+      "learning_rate": 2.158487160806377e-05,
+      "loss": 2.1003,
+      "step": 12794
+    },
+    {
+      "epoch": 0.7870939960629921,
+      "grad_norm": 1.4787429571151733,
+      "learning_rate": 2.1572872721953985e-05,
+      "loss": 2.9411,
+      "step": 12795
+    },
+    {
+      "epoch": 0.7871555118110236,
+      "grad_norm": 1.263980746269226,
+      "learning_rate": 2.1560876768581494e-05,
+      "loss": 2.5481,
+      "step": 12796
+    },
+    {
+      "epoch": 0.7872170275590551,
+      "grad_norm": 1.4181249141693115,
+      "learning_rate": 2.1548883748394788e-05,
+      "loss": 2.1494,
+      "step": 12797
+    },
+    {
+      "epoch": 0.7872785433070866,
+      "grad_norm": 1.4303257465362549,
+      "learning_rate": 2.1536893661842405e-05,
+      "loss": 2.5806,
+      "step": 12798
+    },
+    {
+      "epoch": 0.7873400590551181,
+      "grad_norm": 1.13947331905365,
+      "learning_rate": 2.1524906509372656e-05,
+      "loss": 2.2582,
+      "step": 12799
+    },
+    {
+      "epoch": 0.7874015748031497,
+      "grad_norm": 1.196307897567749,
+      "learning_rate": 2.1512922291433834e-05,
+      "loss": 2.6044,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7874630905511811,
+      "grad_norm": 1.130100131034851,
+      "learning_rate": 2.150094100847404e-05,
+      "loss": 3.0652,
+      "step": 12801
+    },
+    {
+      "epoch": 0.7875246062992126,
+      "grad_norm": 1.3777663707733154,
+      "learning_rate": 2.1488962660941337e-05,
+      "loss": 2.8263,
+      "step": 12802
+    },
+    {
+      "epoch": 0.7875861220472441,
+      "grad_norm": 1.2435983419418335,
+      "learning_rate": 2.1476987249283652e-05,
+      "loss": 2.1898,
+      "step": 12803
+    },
+    {
+      "epoch": 0.7876476377952756,
+      "grad_norm": 1.326291799545288,
+      "learning_rate": 2.1465014773948766e-05,
+      "loss": 1.778,
+      "step": 12804
+    },
+    {
+      "epoch": 0.7877091535433071,
+      "grad_norm": 1.4497143030166626,
+      "learning_rate": 2.1453045235384438e-05,
+      "loss": 2.4241,
+      "step": 12805
+    },
+    {
+      "epoch": 0.7877706692913385,
+      "grad_norm": 1.7673338651657104,
+      "learning_rate": 2.144107863403819e-05,
+      "loss": 2.2077,
+      "step": 12806
+    },
+    {
+      "epoch": 0.78783218503937,
+      "grad_norm": 1.2234450578689575,
+      "learning_rate": 2.1429114970357556e-05,
+      "loss": 1.8158,
+      "step": 12807
+    },
+    {
+      "epoch": 0.7878937007874016,
+      "grad_norm": 2.768216848373413,
+      "learning_rate": 2.1417154244789896e-05,
+      "loss": 1.8363,
+      "step": 12808
+    },
+    {
+      "epoch": 0.7879552165354331,
+      "grad_norm": 1.433071255683899,
+      "learning_rate": 2.1405196457782496e-05,
+      "loss": 1.9537,
+      "step": 12809
+    },
+    {
+      "epoch": 0.7880167322834646,
+      "grad_norm": 1.397233009338379,
+      "learning_rate": 2.1393241609782477e-05,
+      "loss": 2.2157,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7880782480314961,
+      "grad_norm": 1.1358915567398071,
+      "learning_rate": 2.1381289701236918e-05,
+      "loss": 2.4626,
+      "step": 12811
+    },
+    {
+      "epoch": 0.7881397637795275,
+      "grad_norm": 1.3158377408981323,
+      "learning_rate": 2.1369340732592702e-05,
+      "loss": 2.7902,
+      "step": 12812
+    },
+    {
+      "epoch": 0.788201279527559,
+      "grad_norm": 1.1765570640563965,
+      "learning_rate": 2.1357394704296695e-05,
+      "loss": 2.6009,
+      "step": 12813
+    },
+    {
+      "epoch": 0.7882627952755905,
+      "grad_norm": 1.3383753299713135,
+      "learning_rate": 2.13454516167956e-05,
+      "loss": 2.6027,
+      "step": 12814
+    },
+    {
+      "epoch": 0.788324311023622,
+      "grad_norm": 1.4320870637893677,
+      "learning_rate": 2.133351147053605e-05,
+      "loss": 2.6155,
+      "step": 12815
+    },
+    {
+      "epoch": 0.7883858267716536,
+      "grad_norm": 1.3207398653030396,
+      "learning_rate": 2.1321574265964495e-05,
+      "loss": 2.4021,
+      "step": 12816
+    },
+    {
+      "epoch": 0.7884473425196851,
+      "grad_norm": 1.5097616910934448,
+      "learning_rate": 2.1309640003527363e-05,
+      "loss": 2.5452,
+      "step": 12817
+    },
+    {
+      "epoch": 0.7885088582677166,
+      "grad_norm": 1.2109522819519043,
+      "learning_rate": 2.129770868367087e-05,
+      "loss": 2.3018,
+      "step": 12818
+    },
+    {
+      "epoch": 0.788570374015748,
+      "grad_norm": 1.3923404216766357,
+      "learning_rate": 2.1285780306841254e-05,
+      "loss": 2.6434,
+      "step": 12819
+    },
+    {
+      "epoch": 0.7886318897637795,
+      "grad_norm": 1.5719817876815796,
+      "learning_rate": 2.12738548734845e-05,
+      "loss": 2.3587,
+      "step": 12820
+    },
+    {
+      "epoch": 0.788693405511811,
+      "grad_norm": 1.3424049615859985,
+      "learning_rate": 2.126193238404659e-05,
+      "loss": 2.5135,
+      "step": 12821
+    },
+    {
+      "epoch": 0.7887549212598425,
+      "grad_norm": 1.2188950777053833,
+      "learning_rate": 2.125001283897334e-05,
+      "loss": 2.0272,
+      "step": 12822
+    },
+    {
+      "epoch": 0.788816437007874,
+      "grad_norm": 1.3092567920684814,
+      "learning_rate": 2.1238096238710525e-05,
+      "loss": 2.6271,
+      "step": 12823
+    },
+    {
+      "epoch": 0.7888779527559056,
+      "grad_norm": 1.2862963676452637,
+      "learning_rate": 2.1226182583703703e-05,
+      "loss": 1.9629,
+      "step": 12824
+    },
+    {
+      "epoch": 0.788939468503937,
+      "grad_norm": 1.1642861366271973,
+      "learning_rate": 2.1214271874398405e-05,
+      "loss": 2.8558,
+      "step": 12825
+    },
+    {
+      "epoch": 0.7890009842519685,
+      "grad_norm": 1.3631834983825684,
+      "learning_rate": 2.1202364111240004e-05,
+      "loss": 2.4411,
+      "step": 12826
+    },
+    {
+      "epoch": 0.7890625,
+      "grad_norm": 1.2883447408676147,
+      "learning_rate": 2.119045929467379e-05,
+      "loss": 2.5882,
+      "step": 12827
+    },
+    {
+      "epoch": 0.7891240157480315,
+      "grad_norm": 1.2234328985214233,
+      "learning_rate": 2.1178557425144952e-05,
+      "loss": 2.5854,
+      "step": 12828
+    },
+    {
+      "epoch": 0.789185531496063,
+      "grad_norm": 1.3339874744415283,
+      "learning_rate": 2.116665850309858e-05,
+      "loss": 3.1319,
+      "step": 12829
+    },
+    {
+      "epoch": 0.7892470472440944,
+      "grad_norm": 1.1288529634475708,
+      "learning_rate": 2.1154762528979555e-05,
+      "loss": 2.7527,
+      "step": 12830
+    },
+    {
+      "epoch": 0.789308562992126,
+      "grad_norm": 1.1802557706832886,
+      "learning_rate": 2.1142869503232786e-05,
+      "loss": 1.6182,
+      "step": 12831
+    },
+    {
+      "epoch": 0.7893700787401575,
+      "grad_norm": 1.45371413230896,
+      "learning_rate": 2.1130979426302945e-05,
+      "loss": 2.8613,
+      "step": 12832
+    },
+    {
+      "epoch": 0.789431594488189,
+      "grad_norm": 1.2978949546813965,
+      "learning_rate": 2.11190922986347e-05,
+      "loss": 1.8409,
+      "step": 12833
+    },
+    {
+      "epoch": 0.7894931102362205,
+      "grad_norm": 1.4058138132095337,
+      "learning_rate": 2.110720812067254e-05,
+      "loss": 2.7634,
+      "step": 12834
+    },
+    {
+      "epoch": 0.789554625984252,
+      "grad_norm": 1.2292518615722656,
+      "learning_rate": 2.109532689286091e-05,
+      "loss": 2.2735,
+      "step": 12835
+    },
+    {
+      "epoch": 0.7896161417322834,
+      "grad_norm": 1.2662049531936646,
+      "learning_rate": 2.108344861564404e-05,
+      "loss": 1.8692,
+      "step": 12836
+    },
+    {
+      "epoch": 0.7896776574803149,
+      "grad_norm": 1.4744348526000977,
+      "learning_rate": 2.1071573289466174e-05,
+      "loss": 2.2107,
+      "step": 12837
+    },
+    {
+      "epoch": 0.7897391732283464,
+      "grad_norm": 1.1897774934768677,
+      "learning_rate": 2.1059700914771352e-05,
+      "loss": 2.1845,
+      "step": 12838
+    },
+    {
+      "epoch": 0.789800688976378,
+      "grad_norm": 1.4833554029464722,
+      "learning_rate": 2.10478314920035e-05,
+      "loss": 2.3679,
+      "step": 12839
+    },
+    {
+      "epoch": 0.7898622047244095,
+      "grad_norm": 1.132999300956726,
+      "learning_rate": 2.1035965021606497e-05,
+      "loss": 2.7493,
+      "step": 12840
+    },
+    {
+      "epoch": 0.789923720472441,
+      "grad_norm": 1.4470469951629639,
+      "learning_rate": 2.1024101504024096e-05,
+      "loss": 2.453,
+      "step": 12841
+    },
+    {
+      "epoch": 0.7899852362204725,
+      "grad_norm": 1.2097375392913818,
+      "learning_rate": 2.1012240939699935e-05,
+      "loss": 2.4828,
+      "step": 12842
+    },
+    {
+      "epoch": 0.7900467519685039,
+      "grad_norm": 1.7974684238433838,
+      "learning_rate": 2.10003833290775e-05,
+      "loss": 2.4356,
+      "step": 12843
+    },
+    {
+      "epoch": 0.7901082677165354,
+      "grad_norm": 1.2770586013793945,
+      "learning_rate": 2.0988528672600228e-05,
+      "loss": 1.9935,
+      "step": 12844
+    },
+    {
+      "epoch": 0.7901697834645669,
+      "grad_norm": 1.0289283990859985,
+      "learning_rate": 2.0976676970711385e-05,
+      "loss": 2.1757,
+      "step": 12845
+    },
+    {
+      "epoch": 0.7902312992125984,
+      "grad_norm": 1.2066290378570557,
+      "learning_rate": 2.0964828223854172e-05,
+      "loss": 2.8751,
+      "step": 12846
+    },
+    {
+      "epoch": 0.79029281496063,
+      "grad_norm": 1.1635410785675049,
+      "learning_rate": 2.095298243247168e-05,
+      "loss": 2.8226,
+      "step": 12847
+    },
+    {
+      "epoch": 0.7903543307086615,
+      "grad_norm": 1.4196237325668335,
+      "learning_rate": 2.0941139597006885e-05,
+      "loss": 2.5188,
+      "step": 12848
+    },
+    {
+      "epoch": 0.7904158464566929,
+      "grad_norm": 1.3061391115188599,
+      "learning_rate": 2.0929299717902595e-05,
+      "loss": 2.0133,
+      "step": 12849
+    },
+    {
+      "epoch": 0.7904773622047244,
+      "grad_norm": 1.1969066858291626,
+      "learning_rate": 2.091746279560163e-05,
+      "loss": 2.4708,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7905388779527559,
+      "grad_norm": 1.3146122694015503,
+      "learning_rate": 2.090562883054653e-05,
+      "loss": 2.3032,
+      "step": 12851
+    },
+    {
+      "epoch": 0.7906003937007874,
+      "grad_norm": 1.2730257511138916,
+      "learning_rate": 2.0893797823179884e-05,
+      "loss": 2.4098,
+      "step": 12852
+    },
+    {
+      "epoch": 0.7906619094488189,
+      "grad_norm": 1.2403643131256104,
+      "learning_rate": 2.0881969773944086e-05,
+      "loss": 1.9251,
+      "step": 12853
+    },
+    {
+      "epoch": 0.7907234251968503,
+      "grad_norm": 1.2096692323684692,
+      "learning_rate": 2.0870144683281477e-05,
+      "loss": 2.807,
+      "step": 12854
+    },
+    {
+      "epoch": 0.7907849409448819,
+      "grad_norm": 1.4897541999816895,
+      "learning_rate": 2.085832255163418e-05,
+      "loss": 2.5424,
+      "step": 12855
+    },
+    {
+      "epoch": 0.7908464566929134,
+      "grad_norm": 1.2959318161010742,
+      "learning_rate": 2.084650337944435e-05,
+      "loss": 2.7279,
+      "step": 12856
+    },
+    {
+      "epoch": 0.7909079724409449,
+      "grad_norm": 1.4430500268936157,
+      "learning_rate": 2.083468716715389e-05,
+      "loss": 2.7768,
+      "step": 12857
+    },
+    {
+      "epoch": 0.7909694881889764,
+      "grad_norm": 1.17197585105896,
+      "learning_rate": 2.082287391520469e-05,
+      "loss": 2.6556,
+      "step": 12858
+    },
+    {
+      "epoch": 0.7910310039370079,
+      "grad_norm": 1.2804100513458252,
+      "learning_rate": 2.0811063624038507e-05,
+      "loss": 1.6019,
+      "step": 12859
+    },
+    {
+      "epoch": 0.7910925196850394,
+      "grad_norm": 1.1927742958068848,
+      "learning_rate": 2.079925629409699e-05,
+      "loss": 3.1301,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7911540354330708,
+      "grad_norm": 1.391116976737976,
+      "learning_rate": 2.0787451925821644e-05,
+      "loss": 3.0243,
+      "step": 12861
+    },
+    {
+      "epoch": 0.7912155511811023,
+      "grad_norm": 1.1633292436599731,
+      "learning_rate": 2.0775650519653912e-05,
+      "loss": 2.7798,
+      "step": 12862
+    },
+    {
+      "epoch": 0.7912770669291339,
+      "grad_norm": 1.5094703435897827,
+      "learning_rate": 2.076385207603505e-05,
+      "loss": 1.9409,
+      "step": 12863
+    },
+    {
+      "epoch": 0.7913385826771654,
+      "grad_norm": 1.2898657321929932,
+      "learning_rate": 2.075205659540632e-05,
+      "loss": 2.7132,
+      "step": 12864
+    },
+    {
+      "epoch": 0.7914000984251969,
+      "grad_norm": 1.658064842224121,
+      "learning_rate": 2.0740264078208748e-05,
+      "loss": 1.4928,
+      "step": 12865
+    },
+    {
+      "epoch": 0.7914616141732284,
+      "grad_norm": 1.3127723932266235,
+      "learning_rate": 2.0728474524883334e-05,
+      "loss": 2.612,
+      "step": 12866
+    },
+    {
+      "epoch": 0.7915231299212598,
+      "grad_norm": 1.056848168373108,
+      "learning_rate": 2.0716687935870947e-05,
+      "loss": 2.3546,
+      "step": 12867
+    },
+    {
+      "epoch": 0.7915846456692913,
+      "grad_norm": 1.2657554149627686,
+      "learning_rate": 2.0704904311612362e-05,
+      "loss": 2.0285,
+      "step": 12868
+    },
+    {
+      "epoch": 0.7916461614173228,
+      "grad_norm": 1.1668055057525635,
+      "learning_rate": 2.069312365254816e-05,
+      "loss": 2.6943,
+      "step": 12869
+    },
+    {
+      "epoch": 0.7917076771653543,
+      "grad_norm": 1.350904941558838,
+      "learning_rate": 2.0681345959118948e-05,
+      "loss": 2.5161,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7917691929133859,
+      "grad_norm": 1.2717827558517456,
+      "learning_rate": 2.066957123176506e-05,
+      "loss": 2.7367,
+      "step": 12871
+    },
+    {
+      "epoch": 0.7918307086614174,
+      "grad_norm": 1.2398159503936768,
+      "learning_rate": 2.0657799470926864e-05,
+      "loss": 2.6587,
+      "step": 12872
+    },
+    {
+      "epoch": 0.7918922244094488,
+      "grad_norm": 1.2930747270584106,
+      "learning_rate": 2.0646030677044525e-05,
+      "loss": 2.429,
+      "step": 12873
+    },
+    {
+      "epoch": 0.7919537401574803,
+      "grad_norm": 1.298759937286377,
+      "learning_rate": 2.0634264850558195e-05,
+      "loss": 2.1772,
+      "step": 12874
+    },
+    {
+      "epoch": 0.7920152559055118,
+      "grad_norm": 1.2826111316680908,
+      "learning_rate": 2.0622501991907793e-05,
+      "loss": 2.5396,
+      "step": 12875
+    },
+    {
+      "epoch": 0.7920767716535433,
+      "grad_norm": 1.6014238595962524,
+      "learning_rate": 2.0610742101533176e-05,
+      "loss": 2.4062,
+      "step": 12876
+    },
+    {
+      "epoch": 0.7921382874015748,
+      "grad_norm": 1.1682312488555908,
+      "learning_rate": 2.0598985179874107e-05,
+      "loss": 2.5345,
+      "step": 12877
+    },
+    {
+      "epoch": 0.7921998031496063,
+      "grad_norm": 1.4236277341842651,
+      "learning_rate": 2.0587231227370253e-05,
+      "loss": 2.2096,
+      "step": 12878
+    },
+    {
+      "epoch": 0.7922613188976378,
+      "grad_norm": 1.5456010103225708,
+      "learning_rate": 2.0575480244461144e-05,
+      "loss": 2.6451,
+      "step": 12879
+    },
+    {
+      "epoch": 0.7923228346456693,
+      "grad_norm": 1.3569893836975098,
+      "learning_rate": 2.0563732231586175e-05,
+      "loss": 2.067,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7923843503937008,
+      "grad_norm": 1.486645221710205,
+      "learning_rate": 2.0551987189184697e-05,
+      "loss": 2.1495,
+      "step": 12881
+    },
+    {
+      "epoch": 0.7924458661417323,
+      "grad_norm": 1.3094428777694702,
+      "learning_rate": 2.0540245117695855e-05,
+      "loss": 2.3109,
+      "step": 12882
+    },
+    {
+      "epoch": 0.7925073818897638,
+      "grad_norm": 1.6613448858261108,
+      "learning_rate": 2.05285060175588e-05,
+      "loss": 2.4018,
+      "step": 12883
+    },
+    {
+      "epoch": 0.7925688976377953,
+      "grad_norm": 1.1298714876174927,
+      "learning_rate": 2.0516769889212418e-05,
+      "loss": 2.6518,
+      "step": 12884
+    },
+    {
+      "epoch": 0.7926304133858267,
+      "grad_norm": 1.759159803390503,
+      "learning_rate": 2.0505036733095672e-05,
+      "loss": 1.9482,
+      "step": 12885
+    },
+    {
+      "epoch": 0.7926919291338582,
+      "grad_norm": 1.3185054063796997,
+      "learning_rate": 2.049330654964726e-05,
+      "loss": 2.6953,
+      "step": 12886
+    },
+    {
+      "epoch": 0.7927534448818898,
+      "grad_norm": 1.1706594228744507,
+      "learning_rate": 2.0481579339305877e-05,
+      "loss": 2.9357,
+      "step": 12887
+    },
+    {
+      "epoch": 0.7928149606299213,
+      "grad_norm": 1.2884975671768188,
+      "learning_rate": 2.0469855102509983e-05,
+      "loss": 2.7387,
+      "step": 12888
+    },
+    {
+      "epoch": 0.7928764763779528,
+      "grad_norm": 1.3240770101547241,
+      "learning_rate": 2.045813383969807e-05,
+      "loss": 2.664,
+      "step": 12889
+    },
+    {
+      "epoch": 0.7929379921259843,
+      "grad_norm": 1.4691420793533325,
+      "learning_rate": 2.0446415551308374e-05,
+      "loss": 2.0684,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7929995078740157,
+      "grad_norm": 1.159980058670044,
+      "learning_rate": 2.0434700237779136e-05,
+      "loss": 2.8504,
+      "step": 12891
+    },
+    {
+      "epoch": 0.7930610236220472,
+      "grad_norm": 1.2611578702926636,
+      "learning_rate": 2.042298789954845e-05,
+      "loss": 1.9439,
+      "step": 12892
+    },
+    {
+      "epoch": 0.7931225393700787,
+      "grad_norm": 1.8536195755004883,
+      "learning_rate": 2.04112785370543e-05,
+      "loss": 2.9499,
+      "step": 12893
+    },
+    {
+      "epoch": 0.7931840551181102,
+      "grad_norm": 1.4931670427322388,
+      "learning_rate": 2.0399572150734514e-05,
+      "loss": 2.4613,
+      "step": 12894
+    },
+    {
+      "epoch": 0.7932455708661418,
+      "grad_norm": 1.3644415140151978,
+      "learning_rate": 2.0387868741026893e-05,
+      "loss": 2.6977,
+      "step": 12895
+    },
+    {
+      "epoch": 0.7933070866141733,
+      "grad_norm": 1.1392385959625244,
+      "learning_rate": 2.0376168308369025e-05,
+      "loss": 2.573,
+      "step": 12896
+    },
+    {
+      "epoch": 0.7933686023622047,
+      "grad_norm": 1.3584532737731934,
+      "learning_rate": 2.036447085319846e-05,
+      "loss": 2.3798,
+      "step": 12897
+    },
+    {
+      "epoch": 0.7934301181102362,
+      "grad_norm": 1.2739866971969604,
+      "learning_rate": 2.035277637595263e-05,
+      "loss": 2.5835,
+      "step": 12898
+    },
+    {
+      "epoch": 0.7934916338582677,
+      "grad_norm": 1.592454195022583,
+      "learning_rate": 2.034108487706887e-05,
+      "loss": 1.8534,
+      "step": 12899
+    },
+    {
+      "epoch": 0.7935531496062992,
+      "grad_norm": 1.1251877546310425,
+      "learning_rate": 2.032939635698431e-05,
+      "loss": 2.9246,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7936146653543307,
+      "grad_norm": 1.2500253915786743,
+      "learning_rate": 2.0317710816136104e-05,
+      "loss": 2.5286,
+      "step": 12901
+    },
+    {
+      "epoch": 0.7936761811023622,
+      "grad_norm": 1.147540807723999,
+      "learning_rate": 2.0306028254961164e-05,
+      "loss": 2.5675,
+      "step": 12902
+    },
+    {
+      "epoch": 0.7937376968503937,
+      "grad_norm": 1.3221890926361084,
+      "learning_rate": 2.0294348673896378e-05,
+      "loss": 2.4509,
+      "step": 12903
+    },
+    {
+      "epoch": 0.7937992125984252,
+      "grad_norm": 1.4798563718795776,
+      "learning_rate": 2.0282672073378506e-05,
+      "loss": 3.3014,
+      "step": 12904
+    },
+    {
+      "epoch": 0.7938607283464567,
+      "grad_norm": 1.2798662185668945,
+      "learning_rate": 2.0270998453844204e-05,
+      "loss": 2.5132,
+      "step": 12905
+    },
+    {
+      "epoch": 0.7939222440944882,
+      "grad_norm": 1.54253351688385,
+      "learning_rate": 2.0259327815729955e-05,
+      "loss": 2.1235,
+      "step": 12906
+    },
+    {
+      "epoch": 0.7939837598425197,
+      "grad_norm": 1.256797194480896,
+      "learning_rate": 2.0247660159472224e-05,
+      "loss": 2.6573,
+      "step": 12907
+    },
+    {
+      "epoch": 0.7940452755905512,
+      "grad_norm": 1.1589962244033813,
+      "learning_rate": 2.0235995485507275e-05,
+      "loss": 2.9202,
+      "step": 12908
+    },
+    {
+      "epoch": 0.7941067913385826,
+      "grad_norm": 1.2364766597747803,
+      "learning_rate": 2.0224333794271332e-05,
+      "loss": 2.6991,
+      "step": 12909
+    },
+    {
+      "epoch": 0.7941683070866141,
+      "grad_norm": 1.1539188623428345,
+      "learning_rate": 2.0212675086200417e-05,
+      "loss": 2.2259,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7942298228346457,
+      "grad_norm": 1.2489629983901978,
+      "learning_rate": 2.02010193617306e-05,
+      "loss": 1.9379,
+      "step": 12911
+    },
+    {
+      "epoch": 0.7942913385826772,
+      "grad_norm": 1.2902107238769531,
+      "learning_rate": 2.0189366621297688e-05,
+      "loss": 2.8463,
+      "step": 12912
+    },
+    {
+      "epoch": 0.7943528543307087,
+      "grad_norm": 1.3836935758590698,
+      "learning_rate": 2.0177716865337402e-05,
+      "loss": 2.4852,
+      "step": 12913
+    },
+    {
+      "epoch": 0.7944143700787402,
+      "grad_norm": 1.38995361328125,
+      "learning_rate": 2.0166070094285427e-05,
+      "loss": 2.3599,
+      "step": 12914
+    },
+    {
+      "epoch": 0.7944758858267716,
+      "grad_norm": 1.2400559186935425,
+      "learning_rate": 2.0154426308577236e-05,
+      "loss": 2.5092,
+      "step": 12915
+    },
+    {
+      "epoch": 0.7945374015748031,
+      "grad_norm": 1.5271861553192139,
+      "learning_rate": 2.0142785508648278e-05,
+      "loss": 2.4194,
+      "step": 12916
+    },
+    {
+      "epoch": 0.7945989173228346,
+      "grad_norm": 1.5985357761383057,
+      "learning_rate": 2.013114769493384e-05,
+      "loss": 1.187,
+      "step": 12917
+    },
+    {
+      "epoch": 0.7946604330708661,
+      "grad_norm": 1.2544834613800049,
+      "learning_rate": 2.011951286786915e-05,
+      "loss": 1.8313,
+      "step": 12918
+    },
+    {
+      "epoch": 0.7947219488188977,
+      "grad_norm": 1.3763432502746582,
+      "learning_rate": 2.010788102788922e-05,
+      "loss": 2.609,
+      "step": 12919
+    },
+    {
+      "epoch": 0.7947834645669292,
+      "grad_norm": 1.3689898252487183,
+      "learning_rate": 2.0096252175429076e-05,
+      "loss": 1.6076,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7948449803149606,
+      "grad_norm": 1.2006311416625977,
+      "learning_rate": 2.0084626310923517e-05,
+      "loss": 2.5021,
+      "step": 12921
+    },
+    {
+      "epoch": 0.7949064960629921,
+      "grad_norm": 1.489780068397522,
+      "learning_rate": 2.0073003434807315e-05,
+      "loss": 2.4449,
+      "step": 12922
+    },
+    {
+      "epoch": 0.7949680118110236,
+      "grad_norm": 1.3168599605560303,
+      "learning_rate": 2.00613835475151e-05,
+      "loss": 2.6185,
+      "step": 12923
+    },
+    {
+      "epoch": 0.7950295275590551,
+      "grad_norm": 1.199528455734253,
+      "learning_rate": 2.0049766649481417e-05,
+      "loss": 2.141,
+      "step": 12924
+    },
+    {
+      "epoch": 0.7950910433070866,
+      "grad_norm": 1.3019713163375854,
+      "learning_rate": 2.0038152741140625e-05,
+      "loss": 2.9906,
+      "step": 12925
+    },
+    {
+      "epoch": 0.7951525590551181,
+      "grad_norm": 1.2956538200378418,
+      "learning_rate": 2.0026541822927058e-05,
+      "loss": 1.951,
+      "step": 12926
+    },
+    {
+      "epoch": 0.7952140748031497,
+      "grad_norm": 1.3159948587417603,
+      "learning_rate": 2.001493389527487e-05,
+      "loss": 2.4464,
+      "step": 12927
+    },
+    {
+      "epoch": 0.7952755905511811,
+      "grad_norm": 1.191163420677185,
+      "learning_rate": 2.000332895861814e-05,
+      "loss": 2.7004,
+      "step": 12928
+    },
+    {
+      "epoch": 0.7953371062992126,
+      "grad_norm": 1.4366928339004517,
+      "learning_rate": 1.9991727013390826e-05,
+      "loss": 1.4323,
+      "step": 12929
+    },
+    {
+      "epoch": 0.7953986220472441,
+      "grad_norm": 1.1493600606918335,
+      "learning_rate": 1.9980128060026815e-05,
+      "loss": 1.9706,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7954601377952756,
+      "grad_norm": 1.306700348854065,
+      "learning_rate": 1.996853209895978e-05,
+      "loss": 2.6023,
+      "step": 12931
+    },
+    {
+      "epoch": 0.7955216535433071,
+      "grad_norm": 1.395777702331543,
+      "learning_rate": 1.995693913062341e-05,
+      "loss": 2.8172,
+      "step": 12932
+    },
+    {
+      "epoch": 0.7955831692913385,
+      "grad_norm": 1.293657898902893,
+      "learning_rate": 1.9945349155451153e-05,
+      "loss": 2.2218,
+      "step": 12933
+    },
+    {
+      "epoch": 0.79564468503937,
+      "grad_norm": 1.2819981575012207,
+      "learning_rate": 1.993376217387647e-05,
+      "loss": 2.8104,
+      "step": 12934
+    },
+    {
+      "epoch": 0.7957062007874016,
+      "grad_norm": 1.1192998886108398,
+      "learning_rate": 1.9922178186332595e-05,
+      "loss": 2.631,
+      "step": 12935
+    },
+    {
+      "epoch": 0.7957677165354331,
+      "grad_norm": 1.2924690246582031,
+      "learning_rate": 1.9910597193252722e-05,
+      "loss": 2.5397,
+      "step": 12936
+    },
+    {
+      "epoch": 0.7958292322834646,
+      "grad_norm": 1.2423112392425537,
+      "learning_rate": 1.9899019195069935e-05,
+      "loss": 2.5777,
+      "step": 12937
+    },
+    {
+      "epoch": 0.7958907480314961,
+      "grad_norm": 1.302066445350647,
+      "learning_rate": 1.9887444192217186e-05,
+      "loss": 2.7725,
+      "step": 12938
+    },
+    {
+      "epoch": 0.7959522637795275,
+      "grad_norm": 1.31121027469635,
+      "learning_rate": 1.9875872185127288e-05,
+      "loss": 1.3784,
+      "step": 12939
+    },
+    {
+      "epoch": 0.796013779527559,
+      "grad_norm": 1.2079071998596191,
+      "learning_rate": 1.986430317423301e-05,
+      "loss": 2.4685,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7960752952755905,
+      "grad_norm": 1.2274036407470703,
+      "learning_rate": 1.9852737159966927e-05,
+      "loss": 2.169,
+      "step": 12941
+    },
+    {
+      "epoch": 0.796136811023622,
+      "grad_norm": 1.3228216171264648,
+      "learning_rate": 1.9841174142761553e-05,
+      "loss": 2.1825,
+      "step": 12942
+    },
+    {
+      "epoch": 0.7961983267716536,
+      "grad_norm": 1.343519687652588,
+      "learning_rate": 1.9829614123049278e-05,
+      "loss": 2.1341,
+      "step": 12943
+    },
+    {
+      "epoch": 0.7962598425196851,
+      "grad_norm": 1.233752965927124,
+      "learning_rate": 1.981805710126243e-05,
+      "loss": 2.7576,
+      "step": 12944
+    },
+    {
+      "epoch": 0.7963213582677166,
+      "grad_norm": 1.3076859712600708,
+      "learning_rate": 1.9806503077833115e-05,
+      "loss": 2.5518,
+      "step": 12945
+    },
+    {
+      "epoch": 0.796382874015748,
+      "grad_norm": 1.369969129562378,
+      "learning_rate": 1.979495205319344e-05,
+      "loss": 2.2196,
+      "step": 12946
+    },
+    {
+      "epoch": 0.7964443897637795,
+      "grad_norm": 2.1460812091827393,
+      "learning_rate": 1.9783404027775286e-05,
+      "loss": 2.8845,
+      "step": 12947
+    },
+    {
+      "epoch": 0.796505905511811,
+      "grad_norm": 1.025295615196228,
+      "learning_rate": 1.977185900201054e-05,
+      "loss": 2.7627,
+      "step": 12948
+    },
+    {
+      "epoch": 0.7965674212598425,
+      "grad_norm": 1.188286304473877,
+      "learning_rate": 1.9760316976330916e-05,
+      "loss": 2.5287,
+      "step": 12949
+    },
+    {
+      "epoch": 0.796628937007874,
+      "grad_norm": 1.3593220710754395,
+      "learning_rate": 1.9748777951167984e-05,
+      "loss": 2.2758,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7966904527559056,
+      "grad_norm": 1.1738111972808838,
+      "learning_rate": 1.9737241926953288e-05,
+      "loss": 2.734,
+      "step": 12951
+    },
+    {
+      "epoch": 0.796751968503937,
+      "grad_norm": 1.169301986694336,
+      "learning_rate": 1.9725708904118155e-05,
+      "loss": 2.9094,
+      "step": 12952
+    },
+    {
+      "epoch": 0.7968134842519685,
+      "grad_norm": 1.1480480432510376,
+      "learning_rate": 1.9714178883093926e-05,
+      "loss": 2.3048,
+      "step": 12953
+    },
+    {
+      "epoch": 0.796875,
+      "grad_norm": 1.8889679908752441,
+      "learning_rate": 1.970265186431166e-05,
+      "loss": 2.4073,
+      "step": 12954
+    },
+    {
+      "epoch": 0.7969365157480315,
+      "grad_norm": 1.0547865629196167,
+      "learning_rate": 1.9691127848202516e-05,
+      "loss": 1.6448,
+      "step": 12955
+    },
+    {
+      "epoch": 0.796998031496063,
+      "grad_norm": 1.2742869853973389,
+      "learning_rate": 1.9679606835197363e-05,
+      "loss": 3.1694,
+      "step": 12956
+    },
+    {
+      "epoch": 0.7970595472440944,
+      "grad_norm": 1.5699018239974976,
+      "learning_rate": 1.966808882572705e-05,
+      "loss": 1.6297,
+      "step": 12957
+    },
+    {
+      "epoch": 0.797121062992126,
+      "grad_norm": 1.169142484664917,
+      "learning_rate": 1.9656573820222247e-05,
+      "loss": 2.4003,
+      "step": 12958
+    },
+    {
+      "epoch": 0.7971825787401575,
+      "grad_norm": 1.3209651708602905,
+      "learning_rate": 1.9645061819113595e-05,
+      "loss": 2.4336,
+      "step": 12959
+    },
+    {
+      "epoch": 0.797244094488189,
+      "grad_norm": 2.4292209148406982,
+      "learning_rate": 1.9633552822831545e-05,
+      "loss": 2.5091,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7973056102362205,
+      "grad_norm": 1.1789586544036865,
+      "learning_rate": 1.9622046831806485e-05,
+      "loss": 2.5998,
+      "step": 12961
+    },
+    {
+      "epoch": 0.797367125984252,
+      "grad_norm": 1.5018647909164429,
+      "learning_rate": 1.9610543846468666e-05,
+      "loss": 2.8939,
+      "step": 12962
+    },
+    {
+      "epoch": 0.7974286417322834,
+      "grad_norm": 1.3072960376739502,
+      "learning_rate": 1.9599043867248267e-05,
+      "loss": 2.8441,
+      "step": 12963
+    },
+    {
+      "epoch": 0.7974901574803149,
+      "grad_norm": 1.1780195236206055,
+      "learning_rate": 1.958754689457528e-05,
+      "loss": 1.9851,
+      "step": 12964
+    },
+    {
+      "epoch": 0.7975516732283464,
+      "grad_norm": 1.2903399467468262,
+      "learning_rate": 1.9576052928879673e-05,
+      "loss": 1.9488,
+      "step": 12965
+    },
+    {
+      "epoch": 0.797613188976378,
+      "grad_norm": 1.2609857320785522,
+      "learning_rate": 1.95645619705912e-05,
+      "loss": 1.9723,
+      "step": 12966
+    },
+    {
+      "epoch": 0.7976747047244095,
+      "grad_norm": 1.3976479768753052,
+      "learning_rate": 1.9553074020139594e-05,
+      "loss": 2.869,
+      "step": 12967
+    },
+    {
+      "epoch": 0.797736220472441,
+      "grad_norm": 1.2475148439407349,
+      "learning_rate": 1.9541589077954437e-05,
+      "loss": 3.0299,
+      "step": 12968
+    },
+    {
+      "epoch": 0.7977977362204725,
+      "grad_norm": 1.089210867881775,
+      "learning_rate": 1.9530107144465213e-05,
+      "loss": 1.7744,
+      "step": 12969
+    },
+    {
+      "epoch": 0.7978592519685039,
+      "grad_norm": 1.481336236000061,
+      "learning_rate": 1.9518628220101254e-05,
+      "loss": 1.6968,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7979207677165354,
+      "grad_norm": 1.399004578590393,
+      "learning_rate": 1.9507152305291854e-05,
+      "loss": 1.9607,
+      "step": 12971
+    },
+    {
+      "epoch": 0.7979822834645669,
+      "grad_norm": 2.5625739097595215,
+      "learning_rate": 1.949567940046608e-05,
+      "loss": 3.0159,
+      "step": 12972
+    },
+    {
+      "epoch": 0.7980437992125984,
+      "grad_norm": 1.9204922914505005,
+      "learning_rate": 1.9484209506052997e-05,
+      "loss": 2.9984,
+      "step": 12973
+    },
+    {
+      "epoch": 0.79810531496063,
+      "grad_norm": 1.1177102327346802,
+      "learning_rate": 1.947274262248151e-05,
+      "loss": 2.7919,
+      "step": 12974
+    },
+    {
+      "epoch": 0.7981668307086615,
+      "grad_norm": 1.2037595510482788,
+      "learning_rate": 1.9461278750180445e-05,
+      "loss": 2.707,
+      "step": 12975
+    },
+    {
+      "epoch": 0.7982283464566929,
+      "grad_norm": 1.3776748180389404,
+      "learning_rate": 1.9449817889578424e-05,
+      "loss": 2.6701,
+      "step": 12976
+    },
+    {
+      "epoch": 0.7982898622047244,
+      "grad_norm": 1.4851738214492798,
+      "learning_rate": 1.9438360041104087e-05,
+      "loss": 3.3487,
+      "step": 12977
+    },
+    {
+      "epoch": 0.7983513779527559,
+      "grad_norm": 1.3111202716827393,
+      "learning_rate": 1.9426905205185843e-05,
+      "loss": 2.0228,
+      "step": 12978
+    },
+    {
+      "epoch": 0.7984128937007874,
+      "grad_norm": 1.3463236093521118,
+      "learning_rate": 1.941545338225207e-05,
+      "loss": 3.4442,
+      "step": 12979
+    },
+    {
+      "epoch": 0.7984744094488189,
+      "grad_norm": 1.1275500059127808,
+      "learning_rate": 1.940400457273094e-05,
+      "loss": 2.4793,
+      "step": 12980
+    },
+    {
+      "epoch": 0.7985359251968503,
+      "grad_norm": 1.1899031400680542,
+      "learning_rate": 1.9392558777050675e-05,
+      "loss": 3.2215,
+      "step": 12981
+    },
+    {
+      "epoch": 0.7985974409448819,
+      "grad_norm": 1.4464659690856934,
+      "learning_rate": 1.93811159956392e-05,
+      "loss": 2.9149,
+      "step": 12982
+    },
+    {
+      "epoch": 0.7986589566929134,
+      "grad_norm": 1.244110345840454,
+      "learning_rate": 1.936967622892447e-05,
+      "loss": 2.6427,
+      "step": 12983
+    },
+    {
+      "epoch": 0.7987204724409449,
+      "grad_norm": 1.275148630142212,
+      "learning_rate": 1.9358239477334227e-05,
+      "loss": 2.6088,
+      "step": 12984
+    },
+    {
+      "epoch": 0.7987819881889764,
+      "grad_norm": 1.1777790784835815,
+      "learning_rate": 1.934680574129618e-05,
+      "loss": 2.5987,
+      "step": 12985
+    },
+    {
+      "epoch": 0.7988435039370079,
+      "grad_norm": 1.4025583267211914,
+      "learning_rate": 1.9335375021237833e-05,
+      "loss": 2.4189,
+      "step": 12986
+    },
+    {
+      "epoch": 0.7989050196850394,
+      "grad_norm": 1.2714653015136719,
+      "learning_rate": 1.9323947317586667e-05,
+      "loss": 3.0428,
+      "step": 12987
+    },
+    {
+      "epoch": 0.7989665354330708,
+      "grad_norm": 1.1836544275283813,
+      "learning_rate": 1.9312522630770035e-05,
+      "loss": 2.0945,
+      "step": 12988
+    },
+    {
+      "epoch": 0.7990280511811023,
+      "grad_norm": 1.1503077745437622,
+      "learning_rate": 1.9301100961215103e-05,
+      "loss": 2.6223,
+      "step": 12989
+    },
+    {
+      "epoch": 0.7990895669291339,
+      "grad_norm": 1.2984336614608765,
+      "learning_rate": 1.9289682309349023e-05,
+      "loss": 3.1746,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7991510826771654,
+      "grad_norm": 1.2327347993850708,
+      "learning_rate": 1.9278266675598756e-05,
+      "loss": 2.4402,
+      "step": 12991
+    },
+    {
+      "epoch": 0.7992125984251969,
+      "grad_norm": 1.39115571975708,
+      "learning_rate": 1.9266854060391194e-05,
+      "loss": 2.6308,
+      "step": 12992
+    },
+    {
+      "epoch": 0.7992741141732284,
+      "grad_norm": 1.3198124170303345,
+      "learning_rate": 1.9255444464153115e-05,
+      "loss": 2.6094,
+      "step": 12993
+    },
+    {
+      "epoch": 0.7993356299212598,
+      "grad_norm": 1.1374561786651611,
+      "learning_rate": 1.9244037887311195e-05,
+      "loss": 2.5494,
+      "step": 12994
+    },
+    {
+      "epoch": 0.7993971456692913,
+      "grad_norm": 1.5067616701126099,
+      "learning_rate": 1.923263433029192e-05,
+      "loss": 2.9218,
+      "step": 12995
+    },
+    {
+      "epoch": 0.7994586614173228,
+      "grad_norm": 1.1429238319396973,
+      "learning_rate": 1.9221233793521775e-05,
+      "loss": 2.7613,
+      "step": 12996
+    },
+    {
+      "epoch": 0.7995201771653543,
+      "grad_norm": 1.222848653793335,
+      "learning_rate": 1.920983627742703e-05,
+      "loss": 1.949,
+      "step": 12997
+    },
+    {
+      "epoch": 0.7995816929133859,
+      "grad_norm": 1.2194252014160156,
+      "learning_rate": 1.919844178243393e-05,
+      "loss": 2.598,
+      "step": 12998
+    },
+    {
+      "epoch": 0.7996432086614174,
+      "grad_norm": 1.2153878211975098,
+      "learning_rate": 1.9187050308968503e-05,
+      "loss": 2.3137,
+      "step": 12999
+    },
+    {
+      "epoch": 0.7997047244094488,
+      "grad_norm": 1.366070032119751,
+      "learning_rate": 1.9175661857456806e-05,
+      "loss": 2.3681,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7997662401574803,
+      "grad_norm": 1.267747402191162,
+      "learning_rate": 1.9164276428324656e-05,
+      "loss": 2.4882,
+      "step": 13001
+    },
+    {
+      "epoch": 0.7998277559055118,
+      "grad_norm": 1.2874491214752197,
+      "learning_rate": 1.9152894021997836e-05,
+      "loss": 2.4763,
+      "step": 13002
+    },
+    {
+      "epoch": 0.7998892716535433,
+      "grad_norm": 1.401138424873352,
+      "learning_rate": 1.9141514638901937e-05,
+      "loss": 1.2927,
+      "step": 13003
+    },
+    {
+      "epoch": 0.7999507874015748,
+      "grad_norm": 1.1102139949798584,
+      "learning_rate": 1.9130138279462528e-05,
+      "loss": 2.4736,
+      "step": 13004
+    },
+    {
+      "epoch": 0.8000123031496063,
+      "grad_norm": 1.2983354330062866,
+      "learning_rate": 1.9118764944104995e-05,
+      "loss": 1.673,
+      "step": 13005
+    },
+    {
+      "epoch": 0.8000738188976378,
+      "grad_norm": 1.1588900089263916,
+      "learning_rate": 1.9107394633254638e-05,
+      "loss": 2.8094,
+      "step": 13006
+    },
+    {
+      "epoch": 0.8001353346456693,
+      "grad_norm": 1.3462607860565186,
+      "learning_rate": 1.9096027347336653e-05,
+      "loss": 2.5854,
+      "step": 13007
+    },
+    {
+      "epoch": 0.8001968503937008,
+      "grad_norm": 1.2202513217926025,
+      "learning_rate": 1.9084663086776133e-05,
+      "loss": 2.1271,
+      "step": 13008
+    },
+    {
+      "epoch": 0.8002583661417323,
+      "grad_norm": 1.2773528099060059,
+      "learning_rate": 1.9073301851998006e-05,
+      "loss": 2.4571,
+      "step": 13009
+    },
+    {
+      "epoch": 0.8003198818897638,
+      "grad_norm": 1.1206612586975098,
+      "learning_rate": 1.9061943643427137e-05,
+      "loss": 2.3443,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8003813976377953,
+      "grad_norm": 1.354202389717102,
+      "learning_rate": 1.905058846148824e-05,
+      "loss": 3.1174,
+      "step": 13011
+    },
+    {
+      "epoch": 0.8004429133858267,
+      "grad_norm": 1.4680148363113403,
+      "learning_rate": 1.9039236306605944e-05,
+      "loss": 1.607,
+      "step": 13012
+    },
+    {
+      "epoch": 0.8005044291338582,
+      "grad_norm": 1.8763668537139893,
+      "learning_rate": 1.902788717920476e-05,
+      "loss": 3.7951,
+      "step": 13013
+    },
+    {
+      "epoch": 0.8005659448818898,
+      "grad_norm": 1.134484052658081,
+      "learning_rate": 1.9016541079709116e-05,
+      "loss": 2.7158,
+      "step": 13014
+    },
+    {
+      "epoch": 0.8006274606299213,
+      "grad_norm": 1.3149088621139526,
+      "learning_rate": 1.9005198008543222e-05,
+      "loss": 2.3348,
+      "step": 13015
+    },
+    {
+      "epoch": 0.8006889763779528,
+      "grad_norm": 1.6340004205703735,
+      "learning_rate": 1.8993857966131323e-05,
+      "loss": 2.1532,
+      "step": 13016
+    },
+    {
+      "epoch": 0.8007504921259843,
+      "grad_norm": 1.8721481561660767,
+      "learning_rate": 1.8982520952897396e-05,
+      "loss": 2.1048,
+      "step": 13017
+    },
+    {
+      "epoch": 0.8008120078740157,
+      "grad_norm": 1.1178109645843506,
+      "learning_rate": 1.897118696926542e-05,
+      "loss": 2.729,
+      "step": 13018
+    },
+    {
+      "epoch": 0.8008735236220472,
+      "grad_norm": 1.4530807733535767,
+      "learning_rate": 1.8959856015659228e-05,
+      "loss": 1.6791,
+      "step": 13019
+    },
+    {
+      "epoch": 0.8009350393700787,
+      "grad_norm": 1.450715184211731,
+      "learning_rate": 1.894852809250256e-05,
+      "loss": 1.9677,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8009965551181102,
+      "grad_norm": 1.2351480722427368,
+      "learning_rate": 1.8937203200218945e-05,
+      "loss": 2.1877,
+      "step": 13021
+    },
+    {
+      "epoch": 0.8010580708661418,
+      "grad_norm": 1.3603171110153198,
+      "learning_rate": 1.8925881339231944e-05,
+      "loss": 2.1335,
+      "step": 13022
+    },
+    {
+      "epoch": 0.8011195866141733,
+      "grad_norm": 1.2572474479675293,
+      "learning_rate": 1.8914562509964906e-05,
+      "loss": 2.5696,
+      "step": 13023
+    },
+    {
+      "epoch": 0.8011811023622047,
+      "grad_norm": 1.1161917448043823,
+      "learning_rate": 1.890324671284105e-05,
+      "loss": 2.8837,
+      "step": 13024
+    },
+    {
+      "epoch": 0.8012426181102362,
+      "grad_norm": 1.3316766023635864,
+      "learning_rate": 1.8891933948283557e-05,
+      "loss": 2.4938,
+      "step": 13025
+    },
+    {
+      "epoch": 0.8013041338582677,
+      "grad_norm": 1.3478590250015259,
+      "learning_rate": 1.888062421671547e-05,
+      "loss": 1.8353,
+      "step": 13026
+    },
+    {
+      "epoch": 0.8013656496062992,
+      "grad_norm": 1.3830442428588867,
+      "learning_rate": 1.8869317518559736e-05,
+      "loss": 1.7678,
+      "step": 13027
+    },
+    {
+      "epoch": 0.8014271653543307,
+      "grad_norm": 1.433241605758667,
+      "learning_rate": 1.8858013854239087e-05,
+      "loss": 2.4221,
+      "step": 13028
+    },
+    {
+      "epoch": 0.8014886811023622,
+      "grad_norm": 1.1218252182006836,
+      "learning_rate": 1.8846713224176295e-05,
+      "loss": 2.95,
+      "step": 13029
+    },
+    {
+      "epoch": 0.8015501968503937,
+      "grad_norm": 1.2383790016174316,
+      "learning_rate": 1.8835415628793885e-05,
+      "loss": 2.2918,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8016117125984252,
+      "grad_norm": 1.391884446144104,
+      "learning_rate": 1.8824121068514332e-05,
+      "loss": 2.6042,
+      "step": 13031
+    },
+    {
+      "epoch": 0.8016732283464567,
+      "grad_norm": 1.1908718347549438,
+      "learning_rate": 1.8812829543760014e-05,
+      "loss": 2.5663,
+      "step": 13032
+    },
+    {
+      "epoch": 0.8017347440944882,
+      "grad_norm": 1.3192514181137085,
+      "learning_rate": 1.8801541054953175e-05,
+      "loss": 2.6979,
+      "step": 13033
+    },
+    {
+      "epoch": 0.8017962598425197,
+      "grad_norm": 1.5275129079818726,
+      "learning_rate": 1.8790255602515903e-05,
+      "loss": 1.8904,
+      "step": 13034
+    },
+    {
+      "epoch": 0.8018577755905512,
+      "grad_norm": 1.2716100215911865,
+      "learning_rate": 1.8778973186870265e-05,
+      "loss": 2.7025,
+      "step": 13035
+    },
+    {
+      "epoch": 0.8019192913385826,
+      "grad_norm": 1.2974964380264282,
+      "learning_rate": 1.876769380843809e-05,
+      "loss": 2.1817,
+      "step": 13036
+    },
+    {
+      "epoch": 0.8019808070866141,
+      "grad_norm": 1.2766268253326416,
+      "learning_rate": 1.8756417467641217e-05,
+      "loss": 2.3988,
+      "step": 13037
+    },
+    {
+      "epoch": 0.8020423228346457,
+      "grad_norm": 1.437147855758667,
+      "learning_rate": 1.8745144164901297e-05,
+      "loss": 2.4745,
+      "step": 13038
+    },
+    {
+      "epoch": 0.8021038385826772,
+      "grad_norm": 1.2746093273162842,
+      "learning_rate": 1.873387390063992e-05,
+      "loss": 2.8156,
+      "step": 13039
+    },
+    {
+      "epoch": 0.8021653543307087,
+      "grad_norm": 1.3084590435028076,
+      "learning_rate": 1.8722606675278487e-05,
+      "loss": 2.3111,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8022268700787402,
+      "grad_norm": 1.2123706340789795,
+      "learning_rate": 1.871134248923837e-05,
+      "loss": 2.4435,
+      "step": 13041
+    },
+    {
+      "epoch": 0.8022883858267716,
+      "grad_norm": 1.2440812587738037,
+      "learning_rate": 1.8700081342940745e-05,
+      "loss": 2.3132,
+      "step": 13042
+    },
+    {
+      "epoch": 0.8023499015748031,
+      "grad_norm": 1.0367642641067505,
+      "learning_rate": 1.8688823236806728e-05,
+      "loss": 2.6091,
+      "step": 13043
+    },
+    {
+      "epoch": 0.8024114173228346,
+      "grad_norm": 1.1757265329360962,
+      "learning_rate": 1.867756817125732e-05,
+      "loss": 2.3076,
+      "step": 13044
+    },
+    {
+      "epoch": 0.8024729330708661,
+      "grad_norm": 1.3323231935501099,
+      "learning_rate": 1.8666316146713426e-05,
+      "loss": 2.2758,
+      "step": 13045
+    },
+    {
+      "epoch": 0.8025344488188977,
+      "grad_norm": 1.3831466436386108,
+      "learning_rate": 1.8655067163595764e-05,
+      "loss": 2.3548,
+      "step": 13046
+    },
+    {
+      "epoch": 0.8025959645669292,
+      "grad_norm": 1.2948150634765625,
+      "learning_rate": 1.864382122232501e-05,
+      "loss": 2.554,
+      "step": 13047
+    },
+    {
+      "epoch": 0.8026574803149606,
+      "grad_norm": 1.3629885911941528,
+      "learning_rate": 1.863257832332166e-05,
+      "loss": 2.2642,
+      "step": 13048
+    },
+    {
+      "epoch": 0.8027189960629921,
+      "grad_norm": 1.4320225715637207,
+      "learning_rate": 1.86213384670062e-05,
+      "loss": 2.1273,
+      "step": 13049
+    },
+    {
+      "epoch": 0.8027805118110236,
+      "grad_norm": 1.5128700733184814,
+      "learning_rate": 1.8610101653798863e-05,
+      "loss": 2.1864,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8028420275590551,
+      "grad_norm": 1.380567193031311,
+      "learning_rate": 1.8598867884119897e-05,
+      "loss": 2.7117,
+      "step": 13051
+    },
+    {
+      "epoch": 0.8029035433070866,
+      "grad_norm": 1.2538764476776123,
+      "learning_rate": 1.858763715838935e-05,
+      "loss": 3.232,
+      "step": 13052
+    },
+    {
+      "epoch": 0.8029650590551181,
+      "grad_norm": 1.3277387619018555,
+      "learning_rate": 1.8576409477027247e-05,
+      "loss": 2.6483,
+      "step": 13053
+    },
+    {
+      "epoch": 0.8030265748031497,
+      "grad_norm": 1.5336436033248901,
+      "learning_rate": 1.8565184840453364e-05,
+      "loss": 3.1905,
+      "step": 13054
+    },
+    {
+      "epoch": 0.8030880905511811,
+      "grad_norm": 1.1462100744247437,
+      "learning_rate": 1.85539632490875e-05,
+      "loss": 0.5819,
+      "step": 13055
+    },
+    {
+      "epoch": 0.8031496062992126,
+      "grad_norm": 1.2398706674575806,
+      "learning_rate": 1.8542744703349235e-05,
+      "loss": 1.5354,
+      "step": 13056
+    },
+    {
+      "epoch": 0.8032111220472441,
+      "grad_norm": 1.5372872352600098,
+      "learning_rate": 1.8531529203658094e-05,
+      "loss": 2.2792,
+      "step": 13057
+    },
+    {
+      "epoch": 0.8032726377952756,
+      "grad_norm": 1.1738109588623047,
+      "learning_rate": 1.8520316750433485e-05,
+      "loss": 3.2654,
+      "step": 13058
+    },
+    {
+      "epoch": 0.8033341535433071,
+      "grad_norm": 1.4800021648406982,
+      "learning_rate": 1.8509107344094712e-05,
+      "loss": 2.2079,
+      "step": 13059
+    },
+    {
+      "epoch": 0.8033956692913385,
+      "grad_norm": 1.1456631422042847,
+      "learning_rate": 1.849790098506091e-05,
+      "loss": 2.3364,
+      "step": 13060
+    },
+    {
+      "epoch": 0.80345718503937,
+      "grad_norm": 1.482414960861206,
+      "learning_rate": 1.848669767375113e-05,
+      "loss": 2.2414,
+      "step": 13061
+    },
+    {
+      "epoch": 0.8035187007874016,
+      "grad_norm": 1.0402238368988037,
+      "learning_rate": 1.8475497410584318e-05,
+      "loss": 0.9077,
+      "step": 13062
+    },
+    {
+      "epoch": 0.8035802165354331,
+      "grad_norm": 1.0236271619796753,
+      "learning_rate": 1.8464300195979313e-05,
+      "loss": 1.9087,
+      "step": 13063
+    },
+    {
+      "epoch": 0.8036417322834646,
+      "grad_norm": 1.920827865600586,
+      "learning_rate": 1.845310603035485e-05,
+      "loss": 2.1856,
+      "step": 13064
+    },
+    {
+      "epoch": 0.8037032480314961,
+      "grad_norm": 1.538880705833435,
+      "learning_rate": 1.8441914914129475e-05,
+      "loss": 2.0884,
+      "step": 13065
+    },
+    {
+      "epoch": 0.8037647637795275,
+      "grad_norm": 1.538716435432434,
+      "learning_rate": 1.8430726847721735e-05,
+      "loss": 2.6668,
+      "step": 13066
+    },
+    {
+      "epoch": 0.803826279527559,
+      "grad_norm": 1.2531718015670776,
+      "learning_rate": 1.8419541831549935e-05,
+      "loss": 2.3964,
+      "step": 13067
+    },
+    {
+      "epoch": 0.8038877952755905,
+      "grad_norm": 1.3188838958740234,
+      "learning_rate": 1.8408359866032397e-05,
+      "loss": 2.3527,
+      "step": 13068
+    },
+    {
+      "epoch": 0.803949311023622,
+      "grad_norm": 1.3382505178451538,
+      "learning_rate": 1.839718095158718e-05,
+      "loss": 1.6263,
+      "step": 13069
+    },
+    {
+      "epoch": 0.8040108267716536,
+      "grad_norm": 1.265297770500183,
+      "learning_rate": 1.838600508863242e-05,
+      "loss": 2.7778,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8040723425196851,
+      "grad_norm": 1.2603132724761963,
+      "learning_rate": 1.837483227758594e-05,
+      "loss": 1.9444,
+      "step": 13071
+    },
+    {
+      "epoch": 0.8041338582677166,
+      "grad_norm": 1.2178065776824951,
+      "learning_rate": 1.8363662518865608e-05,
+      "loss": 2.4867,
+      "step": 13072
+    },
+    {
+      "epoch": 0.804195374015748,
+      "grad_norm": 1.3083672523498535,
+      "learning_rate": 1.835249581288907e-05,
+      "loss": 2.4591,
+      "step": 13073
+    },
+    {
+      "epoch": 0.8042568897637795,
+      "grad_norm": 1.2694755792617798,
+      "learning_rate": 1.834133216007392e-05,
+      "loss": 3.0424,
+      "step": 13074
+    },
+    {
+      "epoch": 0.804318405511811,
+      "grad_norm": 1.209943413734436,
+      "learning_rate": 1.8330171560837584e-05,
+      "loss": 3.2308,
+      "step": 13075
+    },
+    {
+      "epoch": 0.8043799212598425,
+      "grad_norm": 1.2132383584976196,
+      "learning_rate": 1.8319014015597424e-05,
+      "loss": 2.7553,
+      "step": 13076
+    },
+    {
+      "epoch": 0.804441437007874,
+      "grad_norm": 1.2193735837936401,
+      "learning_rate": 1.8307859524770686e-05,
+      "loss": 1.9781,
+      "step": 13077
+    },
+    {
+      "epoch": 0.8045029527559056,
+      "grad_norm": 1.1989127397537231,
+      "learning_rate": 1.8296708088774493e-05,
+      "loss": 2.0371,
+      "step": 13078
+    },
+    {
+      "epoch": 0.804564468503937,
+      "grad_norm": 1.2943744659423828,
+      "learning_rate": 1.8285559708025802e-05,
+      "loss": 2.5059,
+      "step": 13079
+    },
+    {
+      "epoch": 0.8046259842519685,
+      "grad_norm": 1.490322232246399,
+      "learning_rate": 1.8274414382941553e-05,
+      "loss": 2.5329,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8046875,
+      "grad_norm": 1.4031542539596558,
+      "learning_rate": 1.8263272113938456e-05,
+      "loss": 2.2003,
+      "step": 13081
+    },
+    {
+      "epoch": 0.8047490157480315,
+      "grad_norm": 1.3648266792297363,
+      "learning_rate": 1.8252132901433217e-05,
+      "loss": 2.1881,
+      "step": 13082
+    },
+    {
+      "epoch": 0.804810531496063,
+      "grad_norm": 2.0331625938415527,
+      "learning_rate": 1.8240996745842366e-05,
+      "loss": 2.5945,
+      "step": 13083
+    },
+    {
+      "epoch": 0.8048720472440944,
+      "grad_norm": 1.6492342948913574,
+      "learning_rate": 1.822986364758236e-05,
+      "loss": 2.3703,
+      "step": 13084
+    },
+    {
+      "epoch": 0.804933562992126,
+      "grad_norm": 1.0498759746551514,
+      "learning_rate": 1.821873360706946e-05,
+      "loss": 3.1693,
+      "step": 13085
+    },
+    {
+      "epoch": 0.8049950787401575,
+      "grad_norm": 1.458282470703125,
+      "learning_rate": 1.820760662471993e-05,
+      "loss": 2.1577,
+      "step": 13086
+    },
+    {
+      "epoch": 0.805056594488189,
+      "grad_norm": 1.3395437002182007,
+      "learning_rate": 1.81964827009498e-05,
+      "loss": 1.7683,
+      "step": 13087
+    },
+    {
+      "epoch": 0.8051181102362205,
+      "grad_norm": 1.2321865558624268,
+      "learning_rate": 1.8185361836175063e-05,
+      "loss": 2.9681,
+      "step": 13088
+    },
+    {
+      "epoch": 0.805179625984252,
+      "grad_norm": 1.3533505201339722,
+      "learning_rate": 1.8174244030811592e-05,
+      "loss": 2.2943,
+      "step": 13089
+    },
+    {
+      "epoch": 0.8052411417322834,
+      "grad_norm": 1.2034692764282227,
+      "learning_rate": 1.816312928527514e-05,
+      "loss": 2.1203,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8053026574803149,
+      "grad_norm": 1.2289488315582275,
+      "learning_rate": 1.8152017599981295e-05,
+      "loss": 2.8092,
+      "step": 13091
+    },
+    {
+      "epoch": 0.8053641732283464,
+      "grad_norm": 1.3160037994384766,
+      "learning_rate": 1.8140908975345618e-05,
+      "loss": 2.4223,
+      "step": 13092
+    },
+    {
+      "epoch": 0.805425688976378,
+      "grad_norm": 1.3923983573913574,
+      "learning_rate": 1.8129803411783463e-05,
+      "loss": 2.5504,
+      "step": 13093
+    },
+    {
+      "epoch": 0.8054872047244095,
+      "grad_norm": 1.3113600015640259,
+      "learning_rate": 1.8118700909710163e-05,
+      "loss": 2.7903,
+      "step": 13094
+    },
+    {
+      "epoch": 0.805548720472441,
+      "grad_norm": 1.3774816989898682,
+      "learning_rate": 1.8107601469540824e-05,
+      "loss": 2.1687,
+      "step": 13095
+    },
+    {
+      "epoch": 0.8056102362204725,
+      "grad_norm": 1.425734519958496,
+      "learning_rate": 1.8096505091690587e-05,
+      "loss": 2.0293,
+      "step": 13096
+    },
+    {
+      "epoch": 0.8056717519685039,
+      "grad_norm": 1.3330949544906616,
+      "learning_rate": 1.8085411776574356e-05,
+      "loss": 2.5791,
+      "step": 13097
+    },
+    {
+      "epoch": 0.8057332677165354,
+      "grad_norm": 1.413309931755066,
+      "learning_rate": 1.807432152460693e-05,
+      "loss": 2.9301,
+      "step": 13098
+    },
+    {
+      "epoch": 0.8057947834645669,
+      "grad_norm": 1.4059644937515259,
+      "learning_rate": 1.806323433620307e-05,
+      "loss": 2.3683,
+      "step": 13099
+    },
+    {
+      "epoch": 0.8058562992125984,
+      "grad_norm": 1.5310533046722412,
+      "learning_rate": 1.805215021177733e-05,
+      "loss": 2.1039,
+      "step": 13100
+    },
+    {
+      "epoch": 0.80591781496063,
+      "grad_norm": 1.287520408630371,
+      "learning_rate": 1.804106915174423e-05,
+      "loss": 2.3117,
+      "step": 13101
+    },
+    {
+      "epoch": 0.8059793307086615,
+      "grad_norm": 1.3392493724822998,
+      "learning_rate": 1.802999115651811e-05,
+      "loss": 2.8616,
+      "step": 13102
+    },
+    {
+      "epoch": 0.8060408464566929,
+      "grad_norm": 1.4228535890579224,
+      "learning_rate": 1.801891622651328e-05,
+      "loss": 3.1741,
+      "step": 13103
+    },
+    {
+      "epoch": 0.8061023622047244,
+      "grad_norm": 1.19877290725708,
+      "learning_rate": 1.800784436214382e-05,
+      "loss": 2.6669,
+      "step": 13104
+    },
+    {
+      "epoch": 0.8061638779527559,
+      "grad_norm": 1.306725025177002,
+      "learning_rate": 1.799677556382381e-05,
+      "loss": 2.6186,
+      "step": 13105
+    },
+    {
+      "epoch": 0.8062253937007874,
+      "grad_norm": 1.2066916227340698,
+      "learning_rate": 1.798570983196709e-05,
+      "loss": 2.6717,
+      "step": 13106
+    },
+    {
+      "epoch": 0.8062869094488189,
+      "grad_norm": 1.3491255044937134,
+      "learning_rate": 1.797464716698751e-05,
+      "loss": 2.4631,
+      "step": 13107
+    },
+    {
+      "epoch": 0.8063484251968503,
+      "grad_norm": 1.8758172988891602,
+      "learning_rate": 1.7963587569298735e-05,
+      "loss": 2.1525,
+      "step": 13108
+    },
+    {
+      "epoch": 0.8064099409448819,
+      "grad_norm": 1.378912329673767,
+      "learning_rate": 1.795253103931437e-05,
+      "loss": 2.8272,
+      "step": 13109
+    },
+    {
+      "epoch": 0.8064714566929134,
+      "grad_norm": 1.2320888042449951,
+      "learning_rate": 1.7941477577447796e-05,
+      "loss": 2.3432,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8065329724409449,
+      "grad_norm": 2.491453170776367,
+      "learning_rate": 1.793042718411243e-05,
+      "loss": 1.8121,
+      "step": 13111
+    },
+    {
+      "epoch": 0.8065944881889764,
+      "grad_norm": 1.0194262266159058,
+      "learning_rate": 1.791937985972142e-05,
+      "loss": 2.9477,
+      "step": 13112
+    },
+    {
+      "epoch": 0.8066560039370079,
+      "grad_norm": 1.3233487606048584,
+      "learning_rate": 1.7908335604687932e-05,
+      "loss": 2.8231,
+      "step": 13113
+    },
+    {
+      "epoch": 0.8067175196850394,
+      "grad_norm": 1.3141087293624878,
+      "learning_rate": 1.7897294419424894e-05,
+      "loss": 2.5464,
+      "step": 13114
+    },
+    {
+      "epoch": 0.8067790354330708,
+      "grad_norm": 1.113013505935669,
+      "learning_rate": 1.7886256304345272e-05,
+      "loss": 2.1887,
+      "step": 13115
+    },
+    {
+      "epoch": 0.8068405511811023,
+      "grad_norm": 1.2935870885849,
+      "learning_rate": 1.7875221259861762e-05,
+      "loss": 2.7012,
+      "step": 13116
+    },
+    {
+      "epoch": 0.8069020669291339,
+      "grad_norm": 1.2602733373641968,
+      "learning_rate": 1.7864189286387068e-05,
+      "loss": 2.3184,
+      "step": 13117
+    },
+    {
+      "epoch": 0.8069635826771654,
+      "grad_norm": 1.180620551109314,
+      "learning_rate": 1.7853160384333666e-05,
+      "loss": 2.659,
+      "step": 13118
+    },
+    {
+      "epoch": 0.8070250984251969,
+      "grad_norm": 1.1716477870941162,
+      "learning_rate": 1.7842134554114022e-05,
+      "loss": 2.1764,
+      "step": 13119
+    },
+    {
+      "epoch": 0.8070866141732284,
+      "grad_norm": 1.301313877105713,
+      "learning_rate": 1.78311117961404e-05,
+      "loss": 2.3346,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8071481299212598,
+      "grad_norm": 1.2676937580108643,
+      "learning_rate": 1.7820092110825005e-05,
+      "loss": 2.3183,
+      "step": 13121
+    },
+    {
+      "epoch": 0.8072096456692913,
+      "grad_norm": 1.2977949380874634,
+      "learning_rate": 1.780907549857992e-05,
+      "loss": 2.4564,
+      "step": 13122
+    },
+    {
+      "epoch": 0.8072711614173228,
+      "grad_norm": 1.2946932315826416,
+      "learning_rate": 1.7798061959817124e-05,
+      "loss": 2.7517,
+      "step": 13123
+    },
+    {
+      "epoch": 0.8073326771653543,
+      "grad_norm": 1.2124667167663574,
+      "learning_rate": 1.7787051494948426e-05,
+      "loss": 1.9501,
+      "step": 13124
+    },
+    {
+      "epoch": 0.8073941929133859,
+      "grad_norm": 1.3479036092758179,
+      "learning_rate": 1.777604410438559e-05,
+      "loss": 2.3136,
+      "step": 13125
+    },
+    {
+      "epoch": 0.8074557086614174,
+      "grad_norm": 1.4973690509796143,
+      "learning_rate": 1.7765039788540183e-05,
+      "loss": 2.1946,
+      "step": 13126
+    },
+    {
+      "epoch": 0.8075172244094488,
+      "grad_norm": 1.290776014328003,
+      "learning_rate": 1.7754038547823736e-05,
+      "loss": 2.5327,
+      "step": 13127
+    },
+    {
+      "epoch": 0.8075787401574803,
+      "grad_norm": 1.3507400751113892,
+      "learning_rate": 1.7743040382647624e-05,
+      "loss": 2.3347,
+      "step": 13128
+    },
+    {
+      "epoch": 0.8076402559055118,
+      "grad_norm": 1.7093273401260376,
+      "learning_rate": 1.7732045293423162e-05,
+      "loss": 2.5755,
+      "step": 13129
+    },
+    {
+      "epoch": 0.8077017716535433,
+      "grad_norm": 1.16239333152771,
+      "learning_rate": 1.772105328056143e-05,
+      "loss": 1.9837,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8077632874015748,
+      "grad_norm": 1.2607611417770386,
+      "learning_rate": 1.771006434447353e-05,
+      "loss": 2.1943,
+      "step": 13131
+    },
+    {
+      "epoch": 0.8078248031496063,
+      "grad_norm": 1.3359800577163696,
+      "learning_rate": 1.7699078485570352e-05,
+      "loss": 2.2656,
+      "step": 13132
+    },
+    {
+      "epoch": 0.8078863188976378,
+      "grad_norm": 1.4509446620941162,
+      "learning_rate": 1.768809570426271e-05,
+      "loss": 1.8462,
+      "step": 13133
+    },
+    {
+      "epoch": 0.8079478346456693,
+      "grad_norm": 1.2293123006820679,
+      "learning_rate": 1.7677116000961326e-05,
+      "loss": 2.2645,
+      "step": 13134
+    },
+    {
+      "epoch": 0.8080093503937008,
+      "grad_norm": 1.425833821296692,
+      "learning_rate": 1.766613937607673e-05,
+      "loss": 1.7145,
+      "step": 13135
+    },
+    {
+      "epoch": 0.8080708661417323,
+      "grad_norm": 1.1558380126953125,
+      "learning_rate": 1.7655165830019448e-05,
+      "loss": 2.5536,
+      "step": 13136
+    },
+    {
+      "epoch": 0.8081323818897638,
+      "grad_norm": 1.2877697944641113,
+      "learning_rate": 1.7644195363199768e-05,
+      "loss": 2.5404,
+      "step": 13137
+    },
+    {
+      "epoch": 0.8081938976377953,
+      "grad_norm": 1.1913703680038452,
+      "learning_rate": 1.7633227976027977e-05,
+      "loss": 2.327,
+      "step": 13138
+    },
+    {
+      "epoch": 0.8082554133858267,
+      "grad_norm": 1.2726811170578003,
+      "learning_rate": 1.762226366891414e-05,
+      "loss": 2.7427,
+      "step": 13139
+    },
+    {
+      "epoch": 0.8083169291338582,
+      "grad_norm": 1.2294036149978638,
+      "learning_rate": 1.7611302442268307e-05,
+      "loss": 2.3399,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8083784448818898,
+      "grad_norm": 1.5712177753448486,
+      "learning_rate": 1.7600344296500338e-05,
+      "loss": 2.2954,
+      "step": 13141
+    },
+    {
+      "epoch": 0.8084399606299213,
+      "grad_norm": 1.5613617897033691,
+      "learning_rate": 1.758938923202005e-05,
+      "loss": 2.4645,
+      "step": 13142
+    },
+    {
+      "epoch": 0.8085014763779528,
+      "grad_norm": 2.223153829574585,
+      "learning_rate": 1.7578437249237035e-05,
+      "loss": 2.9222,
+      "step": 13143
+    },
+    {
+      "epoch": 0.8085629921259843,
+      "grad_norm": 1.2042828798294067,
+      "learning_rate": 1.7567488348560913e-05,
+      "loss": 2.5938,
+      "step": 13144
+    },
+    {
+      "epoch": 0.8086245078740157,
+      "grad_norm": 2.094348430633545,
+      "learning_rate": 1.7556542530401032e-05,
+      "loss": 2.0135,
+      "step": 13145
+    },
+    {
+      "epoch": 0.8086860236220472,
+      "grad_norm": 1.2293734550476074,
+      "learning_rate": 1.7545599795166746e-05,
+      "loss": 2.5396,
+      "step": 13146
+    },
+    {
+      "epoch": 0.8087475393700787,
+      "grad_norm": 1.4111498594284058,
+      "learning_rate": 1.7534660143267255e-05,
+      "loss": 2.1071,
+      "step": 13147
+    },
+    {
+      "epoch": 0.8088090551181102,
+      "grad_norm": 1.378665804862976,
+      "learning_rate": 1.752372357511165e-05,
+      "loss": 1.6138,
+      "step": 13148
+    },
+    {
+      "epoch": 0.8088705708661418,
+      "grad_norm": 1.1471022367477417,
+      "learning_rate": 1.7512790091108865e-05,
+      "loss": 2.5351,
+      "step": 13149
+    },
+    {
+      "epoch": 0.8089320866141733,
+      "grad_norm": 1.4410924911499023,
+      "learning_rate": 1.75018596916678e-05,
+      "loss": 2.2661,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8089936023622047,
+      "grad_norm": 1.1685380935668945,
+      "learning_rate": 1.7490932377197124e-05,
+      "loss": 2.7587,
+      "step": 13151
+    },
+    {
+      "epoch": 0.8090551181102362,
+      "grad_norm": 1.3004990816116333,
+      "learning_rate": 1.7480008148105497e-05,
+      "loss": 2.1294,
+      "step": 13152
+    },
+    {
+      "epoch": 0.8091166338582677,
+      "grad_norm": 0.9942352771759033,
+      "learning_rate": 1.746908700480142e-05,
+      "loss": 2.1587,
+      "step": 13153
+    },
+    {
+      "epoch": 0.8091781496062992,
+      "grad_norm": 1.1937875747680664,
+      "learning_rate": 1.745816894769332e-05,
+      "loss": 2.604,
+      "step": 13154
+    },
+    {
+      "epoch": 0.8092396653543307,
+      "grad_norm": 1.288239598274231,
+      "learning_rate": 1.7447253977189405e-05,
+      "loss": 2.3545,
+      "step": 13155
+    },
+    {
+      "epoch": 0.8093011811023622,
+      "grad_norm": 1.3179763555526733,
+      "learning_rate": 1.7436342093697876e-05,
+      "loss": 2.4276,
+      "step": 13156
+    },
+    {
+      "epoch": 0.8093626968503937,
+      "grad_norm": 1.4244749546051025,
+      "learning_rate": 1.742543329762676e-05,
+      "loss": 2.7454,
+      "step": 13157
+    },
+    {
+      "epoch": 0.8094242125984252,
+      "grad_norm": 1.2034142017364502,
+      "learning_rate": 1.7414527589383978e-05,
+      "loss": 2.6816,
+      "step": 13158
+    },
+    {
+      "epoch": 0.8094857283464567,
+      "grad_norm": 1.3212426900863647,
+      "learning_rate": 1.7403624969377352e-05,
+      "loss": 2.7718,
+      "step": 13159
+    },
+    {
+      "epoch": 0.8095472440944882,
+      "grad_norm": 1.2495567798614502,
+      "learning_rate": 1.7392725438014624e-05,
+      "loss": 2.6228,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8096087598425197,
+      "grad_norm": 1.2297592163085938,
+      "learning_rate": 1.7381828995703296e-05,
+      "loss": 2.4645,
+      "step": 13161
+    },
+    {
+      "epoch": 0.8096702755905512,
+      "grad_norm": 1.2431696653366089,
+      "learning_rate": 1.7370935642850894e-05,
+      "loss": 2.8216,
+      "step": 13162
+    },
+    {
+      "epoch": 0.8097317913385826,
+      "grad_norm": 1.2064034938812256,
+      "learning_rate": 1.7360045379864732e-05,
+      "loss": 2.1615,
+      "step": 13163
+    },
+    {
+      "epoch": 0.8097933070866141,
+      "grad_norm": 1.376287579536438,
+      "learning_rate": 1.7349158207152073e-05,
+      "loss": 2.5721,
+      "step": 13164
+    },
+    {
+      "epoch": 0.8098548228346457,
+      "grad_norm": 1.523431658744812,
+      "learning_rate": 1.7338274125120012e-05,
+      "loss": 2.5168,
+      "step": 13165
+    },
+    {
+      "epoch": 0.8099163385826772,
+      "grad_norm": 1.5549918413162231,
+      "learning_rate": 1.7327393134175552e-05,
+      "loss": 2.7721,
+      "step": 13166
+    },
+    {
+      "epoch": 0.8099778543307087,
+      "grad_norm": 1.5907013416290283,
+      "learning_rate": 1.7316515234725607e-05,
+      "loss": 3.1483,
+      "step": 13167
+    },
+    {
+      "epoch": 0.8100393700787402,
+      "grad_norm": 1.3275463581085205,
+      "learning_rate": 1.730564042717695e-05,
+      "loss": 2.2391,
+      "step": 13168
+    },
+    {
+      "epoch": 0.8101008858267716,
+      "grad_norm": 1.1064081192016602,
+      "learning_rate": 1.7294768711936225e-05,
+      "loss": 2.0888,
+      "step": 13169
+    },
+    {
+      "epoch": 0.8101624015748031,
+      "grad_norm": 1.4038712978363037,
+      "learning_rate": 1.7283900089409954e-05,
+      "loss": 2.4882,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8102239173228346,
+      "grad_norm": 1.274312138557434,
+      "learning_rate": 1.7273034560004576e-05,
+      "loss": 2.4893,
+      "step": 13171
+    },
+    {
+      "epoch": 0.8102854330708661,
+      "grad_norm": 1.2652628421783447,
+      "learning_rate": 1.726217212412641e-05,
+      "loss": 2.7054,
+      "step": 13172
+    },
+    {
+      "epoch": 0.8103469488188977,
+      "grad_norm": 1.3022034168243408,
+      "learning_rate": 1.7251312782181674e-05,
+      "loss": 2.552,
+      "step": 13173
+    },
+    {
+      "epoch": 0.8104084645669292,
+      "grad_norm": 1.546471357345581,
+      "learning_rate": 1.72404565345764e-05,
+      "loss": 2.9506,
+      "step": 13174
+    },
+    {
+      "epoch": 0.8104699803149606,
+      "grad_norm": 1.3230705261230469,
+      "learning_rate": 1.7229603381716587e-05,
+      "loss": 2.3405,
+      "step": 13175
+    },
+    {
+      "epoch": 0.8105314960629921,
+      "grad_norm": 1.2838871479034424,
+      "learning_rate": 1.7218753324008053e-05,
+      "loss": 3.0855,
+      "step": 13176
+    },
+    {
+      "epoch": 0.8105930118110236,
+      "grad_norm": 1.3608779907226562,
+      "learning_rate": 1.720790636185653e-05,
+      "loss": 2.467,
+      "step": 13177
+    },
+    {
+      "epoch": 0.8106545275590551,
+      "grad_norm": 1.486692190170288,
+      "learning_rate": 1.719706249566766e-05,
+      "loss": 2.5132,
+      "step": 13178
+    },
+    {
+      "epoch": 0.8107160433070866,
+      "grad_norm": 1.214760422706604,
+      "learning_rate": 1.7186221725846953e-05,
+      "loss": 2.7338,
+      "step": 13179
+    },
+    {
+      "epoch": 0.8107775590551181,
+      "grad_norm": 1.399381399154663,
+      "learning_rate": 1.7175384052799736e-05,
+      "loss": 2.2561,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8108390748031497,
+      "grad_norm": 1.3154181241989136,
+      "learning_rate": 1.7164549476931346e-05,
+      "loss": 2.2619,
+      "step": 13181
+    },
+    {
+      "epoch": 0.8109005905511811,
+      "grad_norm": 1.3063381910324097,
+      "learning_rate": 1.715371799864688e-05,
+      "loss": 2.2023,
+      "step": 13182
+    },
+    {
+      "epoch": 0.8109621062992126,
+      "grad_norm": 1.353845238685608,
+      "learning_rate": 1.714288961835142e-05,
+      "loss": 2.5599,
+      "step": 13183
+    },
+    {
+      "epoch": 0.8110236220472441,
+      "grad_norm": 1.3603301048278809,
+      "learning_rate": 1.7132064336449814e-05,
+      "loss": 2.2951,
+      "step": 13184
+    },
+    {
+      "epoch": 0.8110851377952756,
+      "grad_norm": 1.386696219444275,
+      "learning_rate": 1.7121242153346973e-05,
+      "loss": 2.9278,
+      "step": 13185
+    },
+    {
+      "epoch": 0.8111466535433071,
+      "grad_norm": 1.2524367570877075,
+      "learning_rate": 1.71104230694475e-05,
+      "loss": 2.4741,
+      "step": 13186
+    },
+    {
+      "epoch": 0.8112081692913385,
+      "grad_norm": 1.3182202577590942,
+      "learning_rate": 1.709960708515602e-05,
+      "loss": 2.5382,
+      "step": 13187
+    },
+    {
+      "epoch": 0.81126968503937,
+      "grad_norm": 1.0464056730270386,
+      "learning_rate": 1.7088794200876945e-05,
+      "loss": 3.2399,
+      "step": 13188
+    },
+    {
+      "epoch": 0.8113312007874016,
+      "grad_norm": 1.282415509223938,
+      "learning_rate": 1.707798441701468e-05,
+      "loss": 2.2149,
+      "step": 13189
+    },
+    {
+      "epoch": 0.8113927165354331,
+      "grad_norm": 1.3730645179748535,
+      "learning_rate": 1.7067177733973373e-05,
+      "loss": 2.0757,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8114542322834646,
+      "grad_norm": 1.3553229570388794,
+      "learning_rate": 1.7056374152157174e-05,
+      "loss": 2.5048,
+      "step": 13191
+    },
+    {
+      "epoch": 0.8115157480314961,
+      "grad_norm": 1.3787275552749634,
+      "learning_rate": 1.7045573671970083e-05,
+      "loss": 2.6106,
+      "step": 13192
+    },
+    {
+      "epoch": 0.8115772637795275,
+      "grad_norm": 1.5488100051879883,
+      "learning_rate": 1.703477629381599e-05,
+      "loss": 2.2608,
+      "step": 13193
+    },
+    {
+      "epoch": 0.811638779527559,
+      "grad_norm": 1.4894200563430786,
+      "learning_rate": 1.7023982018098616e-05,
+      "loss": 1.7885,
+      "step": 13194
+    },
+    {
+      "epoch": 0.8117002952755905,
+      "grad_norm": 1.257846713066101,
+      "learning_rate": 1.701319084522165e-05,
+      "loss": 2.4055,
+      "step": 13195
+    },
+    {
+      "epoch": 0.811761811023622,
+      "grad_norm": 1.3135520219802856,
+      "learning_rate": 1.7002402775588565e-05,
+      "loss": 2.0133,
+      "step": 13196
+    },
+    {
+      "epoch": 0.8118233267716536,
+      "grad_norm": 1.115867018699646,
+      "learning_rate": 1.6991617809602824e-05,
+      "loss": 2.2858,
+      "step": 13197
+    },
+    {
+      "epoch": 0.8118848425196851,
+      "grad_norm": 1.2142263650894165,
+      "learning_rate": 1.6980835947667706e-05,
+      "loss": 2.5343,
+      "step": 13198
+    },
+    {
+      "epoch": 0.8119463582677166,
+      "grad_norm": 1.2992466688156128,
+      "learning_rate": 1.6970057190186416e-05,
+      "loss": 2.4945,
+      "step": 13199
+    },
+    {
+      "epoch": 0.812007874015748,
+      "grad_norm": 1.2698917388916016,
+      "learning_rate": 1.6959281537561976e-05,
+      "loss": 2.1083,
+      "step": 13200
+    },
+    {
+      "epoch": 0.8120693897637795,
+      "grad_norm": 1.413948655128479,
+      "learning_rate": 1.6948508990197386e-05,
+      "loss": 2.1321,
+      "step": 13201
+    },
+    {
+      "epoch": 0.812130905511811,
+      "grad_norm": 1.2880628108978271,
+      "learning_rate": 1.6937739548495425e-05,
+      "loss": 2.2441,
+      "step": 13202
+    },
+    {
+      "epoch": 0.8121924212598425,
+      "grad_norm": 1.1256190538406372,
+      "learning_rate": 1.6926973212858833e-05,
+      "loss": 2.6969,
+      "step": 13203
+    },
+    {
+      "epoch": 0.812253937007874,
+      "grad_norm": 1.2085752487182617,
+      "learning_rate": 1.6916209983690225e-05,
+      "loss": 2.6558,
+      "step": 13204
+    },
+    {
+      "epoch": 0.8123154527559056,
+      "grad_norm": 1.5164936780929565,
+      "learning_rate": 1.6905449861392087e-05,
+      "loss": 2.2623,
+      "step": 13205
+    },
+    {
+      "epoch": 0.812376968503937,
+      "grad_norm": 1.2904844284057617,
+      "learning_rate": 1.6894692846366787e-05,
+      "loss": 2.7274,
+      "step": 13206
+    },
+    {
+      "epoch": 0.8124384842519685,
+      "grad_norm": 1.3104965686798096,
+      "learning_rate": 1.688393893901654e-05,
+      "loss": 2.8852,
+      "step": 13207
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 3.61515736579895,
+      "learning_rate": 1.6873188139743533e-05,
+      "loss": 2.7234,
+      "step": 13208
+    },
+    {
+      "epoch": 0.8125615157480315,
+      "grad_norm": 1.852840542793274,
+      "learning_rate": 1.6862440448949735e-05,
+      "loss": 3.1968,
+      "step": 13209
+    },
+    {
+      "epoch": 0.812623031496063,
+      "grad_norm": 1.3329403400421143,
+      "learning_rate": 1.6851695867037077e-05,
+      "loss": 2.2697,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8126845472440944,
+      "grad_norm": 2.834207057952881,
+      "learning_rate": 1.684095439440735e-05,
+      "loss": 2.9106,
+      "step": 13211
+    },
+    {
+      "epoch": 0.812746062992126,
+      "grad_norm": 1.7754141092300415,
+      "learning_rate": 1.6830216031462243e-05,
+      "loss": 3.6819,
+      "step": 13212
+    },
+    {
+      "epoch": 0.8128075787401575,
+      "grad_norm": 1.2771193981170654,
+      "learning_rate": 1.681948077860326e-05,
+      "loss": 2.9425,
+      "step": 13213
+    },
+    {
+      "epoch": 0.812869094488189,
+      "grad_norm": 1.2081372737884521,
+      "learning_rate": 1.6808748636231897e-05,
+      "loss": 2.6563,
+      "step": 13214
+    },
+    {
+      "epoch": 0.8129306102362205,
+      "grad_norm": 1.5887606143951416,
+      "learning_rate": 1.6798019604749428e-05,
+      "loss": 1.8258,
+      "step": 13215
+    },
+    {
+      "epoch": 0.812992125984252,
+      "grad_norm": 1.5622949600219727,
+      "learning_rate": 1.678729368455707e-05,
+      "loss": 1.9213,
+      "step": 13216
+    },
+    {
+      "epoch": 0.8130536417322834,
+      "grad_norm": 1.3849626779556274,
+      "learning_rate": 1.677657087605592e-05,
+      "loss": 1.9392,
+      "step": 13217
+    },
+    {
+      "epoch": 0.8131151574803149,
+      "grad_norm": 1.6281894445419312,
+      "learning_rate": 1.6765851179646985e-05,
+      "loss": 2.8208,
+      "step": 13218
+    },
+    {
+      "epoch": 0.8131766732283464,
+      "grad_norm": 2.7532529830932617,
+      "learning_rate": 1.675513459573106e-05,
+      "loss": 2.204,
+      "step": 13219
+    },
+    {
+      "epoch": 0.813238188976378,
+      "grad_norm": 1.4430147409439087,
+      "learning_rate": 1.674442112470894e-05,
+      "loss": 1.9188,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8132997047244095,
+      "grad_norm": 1.1667847633361816,
+      "learning_rate": 1.6733710766981205e-05,
+      "loss": 2.6078,
+      "step": 13221
+    },
+    {
+      "epoch": 0.813361220472441,
+      "grad_norm": 1.6660677194595337,
+      "learning_rate": 1.6723003522948378e-05,
+      "loss": 2.6563,
+      "step": 13222
+    },
+    {
+      "epoch": 0.8134227362204725,
+      "grad_norm": 1.6529884338378906,
+      "learning_rate": 1.6712299393010856e-05,
+      "loss": 2.3829,
+      "step": 13223
+    },
+    {
+      "epoch": 0.8134842519685039,
+      "grad_norm": 1.399314522743225,
+      "learning_rate": 1.670159837756894e-05,
+      "loss": 2.8837,
+      "step": 13224
+    },
+    {
+      "epoch": 0.8135457677165354,
+      "grad_norm": 1.2912274599075317,
+      "learning_rate": 1.669090047702273e-05,
+      "loss": 2.2195,
+      "step": 13225
+    },
+    {
+      "epoch": 0.8136072834645669,
+      "grad_norm": 1.3038244247436523,
+      "learning_rate": 1.668020569177232e-05,
+      "loss": 2.2339,
+      "step": 13226
+    },
+    {
+      "epoch": 0.8136687992125984,
+      "grad_norm": 1.273941993713379,
+      "learning_rate": 1.666951402221759e-05,
+      "loss": 2.6898,
+      "step": 13227
+    },
+    {
+      "epoch": 0.81373031496063,
+      "grad_norm": 1.4212791919708252,
+      "learning_rate": 1.6658825468758376e-05,
+      "loss": 1.9952,
+      "step": 13228
+    },
+    {
+      "epoch": 0.8137918307086615,
+      "grad_norm": 1.1913577318191528,
+      "learning_rate": 1.6648140031794358e-05,
+      "loss": 2.5759,
+      "step": 13229
+    },
+    {
+      "epoch": 0.8138533464566929,
+      "grad_norm": 1.3243151903152466,
+      "learning_rate": 1.6637457711725157e-05,
+      "loss": 2.4479,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8139148622047244,
+      "grad_norm": 1.0919286012649536,
+      "learning_rate": 1.662677850895016e-05,
+      "loss": 2.8159,
+      "step": 13231
+    },
+    {
+      "epoch": 0.8139763779527559,
+      "grad_norm": 1.2811585664749146,
+      "learning_rate": 1.6616102423868772e-05,
+      "loss": 2.9313,
+      "step": 13232
+    },
+    {
+      "epoch": 0.8140378937007874,
+      "grad_norm": 1.1574584245681763,
+      "learning_rate": 1.6605429456880174e-05,
+      "loss": 2.4271,
+      "step": 13233
+    },
+    {
+      "epoch": 0.8140994094488189,
+      "grad_norm": 1.4370276927947998,
+      "learning_rate": 1.6594759608383515e-05,
+      "loss": 2.3727,
+      "step": 13234
+    },
+    {
+      "epoch": 0.8141609251968503,
+      "grad_norm": 1.4103235006332397,
+      "learning_rate": 1.6584092878777734e-05,
+      "loss": 2.302,
+      "step": 13235
+    },
+    {
+      "epoch": 0.8142224409448819,
+      "grad_norm": 1.2546942234039307,
+      "learning_rate": 1.657342926846175e-05,
+      "loss": 2.842,
+      "step": 13236
+    },
+    {
+      "epoch": 0.8142839566929134,
+      "grad_norm": 1.211719036102295,
+      "learning_rate": 1.6562768777834302e-05,
+      "loss": 2.4873,
+      "step": 13237
+    },
+    {
+      "epoch": 0.8143454724409449,
+      "grad_norm": 1.2038111686706543,
+      "learning_rate": 1.655211140729408e-05,
+      "loss": 2.3375,
+      "step": 13238
+    },
+    {
+      "epoch": 0.8144069881889764,
+      "grad_norm": 1.223832607269287,
+      "learning_rate": 1.6541457157239537e-05,
+      "loss": 2.832,
+      "step": 13239
+    },
+    {
+      "epoch": 0.8144685039370079,
+      "grad_norm": 1.623427152633667,
+      "learning_rate": 1.6530806028069145e-05,
+      "loss": 2.5493,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8145300196850394,
+      "grad_norm": 1.2690458297729492,
+      "learning_rate": 1.6520158020181153e-05,
+      "loss": 2.3254,
+      "step": 13241
+    },
+    {
+      "epoch": 0.8145915354330708,
+      "grad_norm": 1.130517601966858,
+      "learning_rate": 1.6509513133973752e-05,
+      "loss": 2.0486,
+      "step": 13242
+    },
+    {
+      "epoch": 0.8146530511811023,
+      "grad_norm": 1.210444450378418,
+      "learning_rate": 1.6498871369845027e-05,
+      "loss": 2.6136,
+      "step": 13243
+    },
+    {
+      "epoch": 0.8147145669291339,
+      "grad_norm": 1.311969518661499,
+      "learning_rate": 1.6488232728192866e-05,
+      "loss": 2.5613,
+      "step": 13244
+    },
+    {
+      "epoch": 0.8147760826771654,
+      "grad_norm": 1.1468111276626587,
+      "learning_rate": 1.6477597209415162e-05,
+      "loss": 2.6207,
+      "step": 13245
+    },
+    {
+      "epoch": 0.8148375984251969,
+      "grad_norm": 1.55246901512146,
+      "learning_rate": 1.646696481390956e-05,
+      "loss": 2.4662,
+      "step": 13246
+    },
+    {
+      "epoch": 0.8148991141732284,
+      "grad_norm": 1.383973479270935,
+      "learning_rate": 1.6456335542073686e-05,
+      "loss": 2.6456,
+      "step": 13247
+    },
+    {
+      "epoch": 0.8149606299212598,
+      "grad_norm": 1.353857398033142,
+      "learning_rate": 1.6445709394304996e-05,
+      "loss": 1.9191,
+      "step": 13248
+    },
+    {
+      "epoch": 0.8150221456692913,
+      "grad_norm": 1.598220944404602,
+      "learning_rate": 1.6435086371000908e-05,
+      "loss": 2.6416,
+      "step": 13249
+    },
+    {
+      "epoch": 0.8150836614173228,
+      "grad_norm": 1.1329282522201538,
+      "learning_rate": 1.6424466472558575e-05,
+      "loss": 1.8208,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8151451771653543,
+      "grad_norm": 1.3860739469528198,
+      "learning_rate": 1.6413849699375205e-05,
+      "loss": 2.5323,
+      "step": 13251
+    },
+    {
+      "epoch": 0.8152066929133859,
+      "grad_norm": 1.770241141319275,
+      "learning_rate": 1.6403236051847738e-05,
+      "loss": 2.0693,
+      "step": 13252
+    },
+    {
+      "epoch": 0.8152682086614174,
+      "grad_norm": 1.5646443367004395,
+      "learning_rate": 1.6392625530373117e-05,
+      "loss": 1.7709,
+      "step": 13253
+    },
+    {
+      "epoch": 0.8153297244094488,
+      "grad_norm": 1.440061092376709,
+      "learning_rate": 1.6382018135348044e-05,
+      "loss": 2.6924,
+      "step": 13254
+    },
+    {
+      "epoch": 0.8153912401574803,
+      "grad_norm": 1.2531615495681763,
+      "learning_rate": 1.637141386716928e-05,
+      "loss": 2.9187,
+      "step": 13255
+    },
+    {
+      "epoch": 0.8154527559055118,
+      "grad_norm": 1.3854233026504517,
+      "learning_rate": 1.6360812726233286e-05,
+      "loss": 2.208,
+      "step": 13256
+    },
+    {
+      "epoch": 0.8155142716535433,
+      "grad_norm": 1.4510592222213745,
+      "learning_rate": 1.635021471293653e-05,
+      "loss": 2.5739,
+      "step": 13257
+    },
+    {
+      "epoch": 0.8155757874015748,
+      "grad_norm": 1.2897865772247314,
+      "learning_rate": 1.633961982767528e-05,
+      "loss": 2.8526,
+      "step": 13258
+    },
+    {
+      "epoch": 0.8156373031496063,
+      "grad_norm": 1.3138848543167114,
+      "learning_rate": 1.6329028070845763e-05,
+      "loss": 2.2859,
+      "step": 13259
+    },
+    {
+      "epoch": 0.8156988188976378,
+      "grad_norm": 1.4424421787261963,
+      "learning_rate": 1.6318439442844023e-05,
+      "loss": 2.8722,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8157603346456693,
+      "grad_norm": 1.5419111251831055,
+      "learning_rate": 1.630785394406601e-05,
+      "loss": 3.4464,
+      "step": 13261
+    },
+    {
+      "epoch": 0.8158218503937008,
+      "grad_norm": 1.141876220703125,
+      "learning_rate": 1.6297271574907592e-05,
+      "loss": 2.8993,
+      "step": 13262
+    },
+    {
+      "epoch": 0.8158833661417323,
+      "grad_norm": 1.1926287412643433,
+      "learning_rate": 1.6286692335764496e-05,
+      "loss": 3.0456,
+      "step": 13263
+    },
+    {
+      "epoch": 0.8159448818897638,
+      "grad_norm": 1.9502602815628052,
+      "learning_rate": 1.627611622703229e-05,
+      "loss": 1.7039,
+      "step": 13264
+    },
+    {
+      "epoch": 0.8160063976377953,
+      "grad_norm": 1.2224844694137573,
+      "learning_rate": 1.6265543249106497e-05,
+      "loss": 1.6361,
+      "step": 13265
+    },
+    {
+      "epoch": 0.8160679133858267,
+      "grad_norm": 1.3126051425933838,
+      "learning_rate": 1.6254973402382455e-05,
+      "loss": 3.0458,
+      "step": 13266
+    },
+    {
+      "epoch": 0.8161294291338582,
+      "grad_norm": 1.2915457487106323,
+      "learning_rate": 1.624440668725542e-05,
+      "loss": 1.5513,
+      "step": 13267
+    },
+    {
+      "epoch": 0.8161909448818898,
+      "grad_norm": 1.2216092348098755,
+      "learning_rate": 1.623384310412055e-05,
+      "loss": 2.2495,
+      "step": 13268
+    },
+    {
+      "epoch": 0.8162524606299213,
+      "grad_norm": 1.2340238094329834,
+      "learning_rate": 1.6223282653372886e-05,
+      "loss": 2.429,
+      "step": 13269
+    },
+    {
+      "epoch": 0.8163139763779528,
+      "grad_norm": 1.247847080230713,
+      "learning_rate": 1.6212725335407263e-05,
+      "loss": 3.0943,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8163754921259843,
+      "grad_norm": 1.139695644378662,
+      "learning_rate": 1.6202171150618527e-05,
+      "loss": 1.8641,
+      "step": 13271
+    },
+    {
+      "epoch": 0.8164370078740157,
+      "grad_norm": 1.0678439140319824,
+      "learning_rate": 1.6191620099401296e-05,
+      "loss": 3.1716,
+      "step": 13272
+    },
+    {
+      "epoch": 0.8164985236220472,
+      "grad_norm": 1.2906105518341064,
+      "learning_rate": 1.6181072182150146e-05,
+      "loss": 2.9943,
+      "step": 13273
+    },
+    {
+      "epoch": 0.8165600393700787,
+      "grad_norm": 1.1840271949768066,
+      "learning_rate": 1.6170527399259518e-05,
+      "loss": 3.1513,
+      "step": 13274
+    },
+    {
+      "epoch": 0.8166215551181102,
+      "grad_norm": 1.2665059566497803,
+      "learning_rate": 1.615998575112373e-05,
+      "loss": 2.6871,
+      "step": 13275
+    },
+    {
+      "epoch": 0.8166830708661418,
+      "grad_norm": 1.4069886207580566,
+      "learning_rate": 1.614944723813696e-05,
+      "loss": 2.3472,
+      "step": 13276
+    },
+    {
+      "epoch": 0.8167445866141733,
+      "grad_norm": 1.1506041288375854,
+      "learning_rate": 1.6138911860693318e-05,
+      "loss": 2.4996,
+      "step": 13277
+    },
+    {
+      "epoch": 0.8168061023622047,
+      "grad_norm": 1.333082914352417,
+      "learning_rate": 1.6128379619186718e-05,
+      "loss": 2.7022,
+      "step": 13278
+    },
+    {
+      "epoch": 0.8168676181102362,
+      "grad_norm": 1.3929152488708496,
+      "learning_rate": 1.611785051401107e-05,
+      "loss": 2.0831,
+      "step": 13279
+    },
+    {
+      "epoch": 0.8169291338582677,
+      "grad_norm": 1.370152235031128,
+      "learning_rate": 1.6107324545560053e-05,
+      "loss": 2.4269,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8169906496062992,
+      "grad_norm": 1.2704473733901978,
+      "learning_rate": 1.6096801714227293e-05,
+      "loss": 2.7793,
+      "step": 13281
+    },
+    {
+      "epoch": 0.8170521653543307,
+      "grad_norm": 1.2233288288116455,
+      "learning_rate": 1.6086282020406318e-05,
+      "loss": 2.7449,
+      "step": 13282
+    },
+    {
+      "epoch": 0.8171136811023622,
+      "grad_norm": 1.2074921131134033,
+      "learning_rate": 1.6075765464490456e-05,
+      "loss": 2.2542,
+      "step": 13283
+    },
+    {
+      "epoch": 0.8171751968503937,
+      "grad_norm": 1.2233645915985107,
+      "learning_rate": 1.6065252046873013e-05,
+      "loss": 2.7592,
+      "step": 13284
+    },
+    {
+      "epoch": 0.8172367125984252,
+      "grad_norm": 1.5743070840835571,
+      "learning_rate": 1.605474176794709e-05,
+      "loss": 2.2142,
+      "step": 13285
+    },
+    {
+      "epoch": 0.8172982283464567,
+      "grad_norm": 1.3237063884735107,
+      "learning_rate": 1.6044234628105726e-05,
+      "loss": 2.523,
+      "step": 13286
+    },
+    {
+      "epoch": 0.8173597440944882,
+      "grad_norm": 1.2012920379638672,
+      "learning_rate": 1.603373062774185e-05,
+      "loss": 1.8403,
+      "step": 13287
+    },
+    {
+      "epoch": 0.8174212598425197,
+      "grad_norm": 1.3842172622680664,
+      "learning_rate": 1.602322976724825e-05,
+      "loss": 2.2714,
+      "step": 13288
+    },
+    {
+      "epoch": 0.8174827755905512,
+      "grad_norm": 1.1872310638427734,
+      "learning_rate": 1.6012732047017577e-05,
+      "loss": 2.5739,
+      "step": 13289
+    },
+    {
+      "epoch": 0.8175442913385826,
+      "grad_norm": 1.3606383800506592,
+      "learning_rate": 1.6002237467442415e-05,
+      "loss": 2.8432,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8176058070866141,
+      "grad_norm": 1.286643147468567,
+      "learning_rate": 1.5991746028915168e-05,
+      "loss": 2.9947,
+      "step": 13291
+    },
+    {
+      "epoch": 0.8176673228346457,
+      "grad_norm": 1.4329686164855957,
+      "learning_rate": 1.598125773182817e-05,
+      "loss": 2.4658,
+      "step": 13292
+    },
+    {
+      "epoch": 0.8177288385826772,
+      "grad_norm": 1.3069977760314941,
+      "learning_rate": 1.5970772576573646e-05,
+      "loss": 2.4354,
+      "step": 13293
+    },
+    {
+      "epoch": 0.8177903543307087,
+      "grad_norm": 1.4102084636688232,
+      "learning_rate": 1.5960290563543677e-05,
+      "loss": 2.0851,
+      "step": 13294
+    },
+    {
+      "epoch": 0.8178518700787402,
+      "grad_norm": 1.4084643125534058,
+      "learning_rate": 1.5949811693130203e-05,
+      "loss": 2.3149,
+      "step": 13295
+    },
+    {
+      "epoch": 0.8179133858267716,
+      "grad_norm": 1.0934959650039673,
+      "learning_rate": 1.593933596572512e-05,
+      "loss": 2.7075,
+      "step": 13296
+    },
+    {
+      "epoch": 0.8179749015748031,
+      "grad_norm": 1.3397984504699707,
+      "learning_rate": 1.5928863381720115e-05,
+      "loss": 2.7553,
+      "step": 13297
+    },
+    {
+      "epoch": 0.8180364173228346,
+      "grad_norm": 1.1977938413619995,
+      "learning_rate": 1.591839394150685e-05,
+      "loss": 2.5712,
+      "step": 13298
+    },
+    {
+      "epoch": 0.8180979330708661,
+      "grad_norm": 1.9777425527572632,
+      "learning_rate": 1.590792764547675e-05,
+      "loss": 2.4157,
+      "step": 13299
+    },
+    {
+      "epoch": 0.8181594488188977,
+      "grad_norm": 1.3933818340301514,
+      "learning_rate": 1.5897464494021287e-05,
+      "loss": 2.675,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8182209645669292,
+      "grad_norm": 1.8641341924667358,
+      "learning_rate": 1.5887004487531666e-05,
+      "loss": 1.6412,
+      "step": 13301
+    },
+    {
+      "epoch": 0.8182824803149606,
+      "grad_norm": 1.2244417667388916,
+      "learning_rate": 1.5876547626399074e-05,
+      "loss": 2.5089,
+      "step": 13302
+    },
+    {
+      "epoch": 0.8183439960629921,
+      "grad_norm": 1.2733970880508423,
+      "learning_rate": 1.586609391101449e-05,
+      "loss": 3.0015,
+      "step": 13303
+    },
+    {
+      "epoch": 0.8184055118110236,
+      "grad_norm": 1.435435175895691,
+      "learning_rate": 1.585564334176888e-05,
+      "loss": 2.3071,
+      "step": 13304
+    },
+    {
+      "epoch": 0.8184670275590551,
+      "grad_norm": 1.6054364442825317,
+      "learning_rate": 1.584519591905298e-05,
+      "loss": 1.9804,
+      "step": 13305
+    },
+    {
+      "epoch": 0.8185285433070866,
+      "grad_norm": 1.3752797842025757,
+      "learning_rate": 1.5834751643257493e-05,
+      "loss": 2.3375,
+      "step": 13306
+    },
+    {
+      "epoch": 0.8185900590551181,
+      "grad_norm": 1.1813825368881226,
+      "learning_rate": 1.582431051477298e-05,
+      "loss": 2.5843,
+      "step": 13307
+    },
+    {
+      "epoch": 0.8186515748031497,
+      "grad_norm": 1.3917057514190674,
+      "learning_rate": 1.5813872533989893e-05,
+      "loss": 2.1266,
+      "step": 13308
+    },
+    {
+      "epoch": 0.8187130905511811,
+      "grad_norm": 1.3416591882705688,
+      "learning_rate": 1.580343770129852e-05,
+      "loss": 2.0946,
+      "step": 13309
+    },
+    {
+      "epoch": 0.8187746062992126,
+      "grad_norm": 1.3026589155197144,
+      "learning_rate": 1.5793006017089096e-05,
+      "loss": 2.3273,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8188361220472441,
+      "grad_norm": 1.3370742797851562,
+      "learning_rate": 1.578257748175168e-05,
+      "loss": 2.8353,
+      "step": 13311
+    },
+    {
+      "epoch": 0.8188976377952756,
+      "grad_norm": 1.3669099807739258,
+      "learning_rate": 1.577215209567626e-05,
+      "loss": 2.4579,
+      "step": 13312
+    },
+    {
+      "epoch": 0.8189591535433071,
+      "grad_norm": 1.230405569076538,
+      "learning_rate": 1.5761729859252683e-05,
+      "loss": 2.2508,
+      "step": 13313
+    },
+    {
+      "epoch": 0.8190206692913385,
+      "grad_norm": 1.1335731744766235,
+      "learning_rate": 1.575131077287071e-05,
+      "loss": 2.58,
+      "step": 13314
+    },
+    {
+      "epoch": 0.81908218503937,
+      "grad_norm": 1.2621557712554932,
+      "learning_rate": 1.5740894836919906e-05,
+      "loss": 3.0216,
+      "step": 13315
+    },
+    {
+      "epoch": 0.8191437007874016,
+      "grad_norm": 1.2785500288009644,
+      "learning_rate": 1.573048205178982e-05,
+      "loss": 2.6457,
+      "step": 13316
+    },
+    {
+      "epoch": 0.8192052165354331,
+      "grad_norm": 1.1986680030822754,
+      "learning_rate": 1.5720072417869768e-05,
+      "loss": 1.5395,
+      "step": 13317
+    },
+    {
+      "epoch": 0.8192667322834646,
+      "grad_norm": 1.376085638999939,
+      "learning_rate": 1.5709665935549056e-05,
+      "loss": 2.8677,
+      "step": 13318
+    },
+    {
+      "epoch": 0.8193282480314961,
+      "grad_norm": 1.3148901462554932,
+      "learning_rate": 1.569926260521686e-05,
+      "loss": 2.4573,
+      "step": 13319
+    },
+    {
+      "epoch": 0.8193897637795275,
+      "grad_norm": 1.1751477718353271,
+      "learning_rate": 1.568886242726213e-05,
+      "loss": 2.3401,
+      "step": 13320
+    },
+    {
+      "epoch": 0.819451279527559,
+      "grad_norm": 1.3517918586730957,
+      "learning_rate": 1.5678465402073838e-05,
+      "loss": 2.6529,
+      "step": 13321
+    },
+    {
+      "epoch": 0.8195127952755905,
+      "grad_norm": 1.2348692417144775,
+      "learning_rate": 1.566807153004073e-05,
+      "loss": 2.0304,
+      "step": 13322
+    },
+    {
+      "epoch": 0.819574311023622,
+      "grad_norm": 1.3764070272445679,
+      "learning_rate": 1.5657680811551523e-05,
+      "loss": 2.7895,
+      "step": 13323
+    },
+    {
+      "epoch": 0.8196358267716536,
+      "grad_norm": 1.2620036602020264,
+      "learning_rate": 1.5647293246994722e-05,
+      "loss": 3.2361,
+      "step": 13324
+    },
+    {
+      "epoch": 0.8196973425196851,
+      "grad_norm": 1.3418689966201782,
+      "learning_rate": 1.563690883675879e-05,
+      "loss": 2.7602,
+      "step": 13325
+    },
+    {
+      "epoch": 0.8197588582677166,
+      "grad_norm": 2.102783441543579,
+      "learning_rate": 1.5626527581232054e-05,
+      "loss": 3.5965,
+      "step": 13326
+    },
+    {
+      "epoch": 0.819820374015748,
+      "grad_norm": 1.3679285049438477,
+      "learning_rate": 1.5616149480802723e-05,
+      "loss": 2.6194,
+      "step": 13327
+    },
+    {
+      "epoch": 0.8198818897637795,
+      "grad_norm": 1.288310170173645,
+      "learning_rate": 1.560577453585884e-05,
+      "loss": 2.6044,
+      "step": 13328
+    },
+    {
+      "epoch": 0.819943405511811,
+      "grad_norm": 1.260560393333435,
+      "learning_rate": 1.5595402746788423e-05,
+      "loss": 2.2857,
+      "step": 13329
+    },
+    {
+      "epoch": 0.8200049212598425,
+      "grad_norm": 1.4170923233032227,
+      "learning_rate": 1.5585034113979258e-05,
+      "loss": 2.5134,
+      "step": 13330
+    },
+    {
+      "epoch": 0.820066437007874,
+      "grad_norm": 1.5757063627243042,
+      "learning_rate": 1.5574668637819114e-05,
+      "loss": 2.247,
+      "step": 13331
+    },
+    {
+      "epoch": 0.8201279527559056,
+      "grad_norm": 1.202911376953125,
+      "learning_rate": 1.5564306318695586e-05,
+      "loss": 2.7673,
+      "step": 13332
+    },
+    {
+      "epoch": 0.820189468503937,
+      "grad_norm": 1.2541674375534058,
+      "learning_rate": 1.5553947156996207e-05,
+      "loss": 2.134,
+      "step": 13333
+    },
+    {
+      "epoch": 0.8202509842519685,
+      "grad_norm": 1.1653385162353516,
+      "learning_rate": 1.5543591153108293e-05,
+      "loss": 2.7698,
+      "step": 13334
+    },
+    {
+      "epoch": 0.8203125,
+      "grad_norm": 1.0731141567230225,
+      "learning_rate": 1.5533238307419163e-05,
+      "loss": 2.2136,
+      "step": 13335
+    },
+    {
+      "epoch": 0.8203740157480315,
+      "grad_norm": 1.1717336177825928,
+      "learning_rate": 1.552288862031588e-05,
+      "loss": 2.5474,
+      "step": 13336
+    },
+    {
+      "epoch": 0.820435531496063,
+      "grad_norm": 1.277491569519043,
+      "learning_rate": 1.5512542092185512e-05,
+      "loss": 2.0533,
+      "step": 13337
+    },
+    {
+      "epoch": 0.8204970472440944,
+      "grad_norm": 1.3656032085418701,
+      "learning_rate": 1.5502198723414963e-05,
+      "loss": 2.7512,
+      "step": 13338
+    },
+    {
+      "epoch": 0.820558562992126,
+      "grad_norm": 1.1786260604858398,
+      "learning_rate": 1.549185851439102e-05,
+      "loss": 2.5806,
+      "step": 13339
+    },
+    {
+      "epoch": 0.8206200787401575,
+      "grad_norm": 1.2115302085876465,
+      "learning_rate": 1.5481521465500327e-05,
+      "loss": 1.7958,
+      "step": 13340
+    },
+    {
+      "epoch": 0.820681594488189,
+      "grad_norm": 1.1955537796020508,
+      "learning_rate": 1.5471187577129452e-05,
+      "loss": 2.5389,
+      "step": 13341
+    },
+    {
+      "epoch": 0.8207431102362205,
+      "grad_norm": 1.2955206632614136,
+      "learning_rate": 1.5460856849664807e-05,
+      "loss": 2.8724,
+      "step": 13342
+    },
+    {
+      "epoch": 0.820804625984252,
+      "grad_norm": 1.3311328887939453,
+      "learning_rate": 1.5450529283492698e-05,
+      "loss": 2.2209,
+      "step": 13343
+    },
+    {
+      "epoch": 0.8208661417322834,
+      "grad_norm": 1.4799860715866089,
+      "learning_rate": 1.5440204878999343e-05,
+      "loss": 2.071,
+      "step": 13344
+    },
+    {
+      "epoch": 0.8209276574803149,
+      "grad_norm": 1.2998546361923218,
+      "learning_rate": 1.542988363657083e-05,
+      "loss": 2.7947,
+      "step": 13345
+    },
+    {
+      "epoch": 0.8209891732283464,
+      "grad_norm": 1.2487809658050537,
+      "learning_rate": 1.5419565556593063e-05,
+      "loss": 1.9816,
+      "step": 13346
+    },
+    {
+      "epoch": 0.821050688976378,
+      "grad_norm": 1.2446560859680176,
+      "learning_rate": 1.5409250639451943e-05,
+      "loss": 2.9502,
+      "step": 13347
+    },
+    {
+      "epoch": 0.8211122047244095,
+      "grad_norm": 1.307973861694336,
+      "learning_rate": 1.5398938885533132e-05,
+      "loss": 1.7596,
+      "step": 13348
+    },
+    {
+      "epoch": 0.821173720472441,
+      "grad_norm": 1.3423563241958618,
+      "learning_rate": 1.5388630295222284e-05,
+      "loss": 2.6908,
+      "step": 13349
+    },
+    {
+      "epoch": 0.8212352362204725,
+      "grad_norm": 1.2519017457962036,
+      "learning_rate": 1.5378324868904826e-05,
+      "loss": 2.7336,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8212967519685039,
+      "grad_norm": 1.1938420534133911,
+      "learning_rate": 1.536802260696616e-05,
+      "loss": 2.8009,
+      "step": 13351
+    },
+    {
+      "epoch": 0.8213582677165354,
+      "grad_norm": 1.1959526538848877,
+      "learning_rate": 1.5357723509791523e-05,
+      "loss": 2.0399,
+      "step": 13352
+    },
+    {
+      "epoch": 0.8214197834645669,
+      "grad_norm": 1.176472783088684,
+      "learning_rate": 1.534742757776608e-05,
+      "loss": 2.4856,
+      "step": 13353
+    },
+    {
+      "epoch": 0.8214812992125984,
+      "grad_norm": 1.2808300256729126,
+      "learning_rate": 1.53371348112748e-05,
+      "loss": 2.3265,
+      "step": 13354
+    },
+    {
+      "epoch": 0.82154281496063,
+      "grad_norm": 1.2825826406478882,
+      "learning_rate": 1.5326845210702566e-05,
+      "loss": 2.0346,
+      "step": 13355
+    },
+    {
+      "epoch": 0.8216043307086615,
+      "grad_norm": 1.3003032207489014,
+      "learning_rate": 1.5316558776434174e-05,
+      "loss": 2.2296,
+      "step": 13356
+    },
+    {
+      "epoch": 0.8216658464566929,
+      "grad_norm": 1.851033091545105,
+      "learning_rate": 1.530627550885427e-05,
+      "loss": 2.4716,
+      "step": 13357
+    },
+    {
+      "epoch": 0.8217273622047244,
+      "grad_norm": 2.6680195331573486,
+      "learning_rate": 1.529599540834743e-05,
+      "loss": 1.7595,
+      "step": 13358
+    },
+    {
+      "epoch": 0.8217888779527559,
+      "grad_norm": 1.289062738418579,
+      "learning_rate": 1.5285718475298004e-05,
+      "loss": 2.2268,
+      "step": 13359
+    },
+    {
+      "epoch": 0.8218503937007874,
+      "grad_norm": 1.4848994016647339,
+      "learning_rate": 1.527544471009036e-05,
+      "loss": 2.8535,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8219119094488189,
+      "grad_norm": 1.3916161060333252,
+      "learning_rate": 1.5265174113108615e-05,
+      "loss": 2.2379,
+      "step": 13361
+    },
+    {
+      "epoch": 0.8219734251968503,
+      "grad_norm": 1.2544633150100708,
+      "learning_rate": 1.5254906684736869e-05,
+      "loss": 2.0243,
+      "step": 13362
+    },
+    {
+      "epoch": 0.8220349409448819,
+      "grad_norm": 1.30370032787323,
+      "learning_rate": 1.5244642425359068e-05,
+      "loss": 2.0995,
+      "step": 13363
+    },
+    {
+      "epoch": 0.8220964566929134,
+      "grad_norm": 1.145849585533142,
+      "learning_rate": 1.5234381335359039e-05,
+      "loss": 2.2129,
+      "step": 13364
+    },
+    {
+      "epoch": 0.8221579724409449,
+      "grad_norm": 1.606382131576538,
+      "learning_rate": 1.5224123415120473e-05,
+      "loss": 2.3491,
+      "step": 13365
+    },
+    {
+      "epoch": 0.8222194881889764,
+      "grad_norm": 1.3311532735824585,
+      "learning_rate": 1.5213868665026988e-05,
+      "loss": 2.3739,
+      "step": 13366
+    },
+    {
+      "epoch": 0.8222810039370079,
+      "grad_norm": 1.3524125814437866,
+      "learning_rate": 1.5203617085462007e-05,
+      "loss": 2.5045,
+      "step": 13367
+    },
+    {
+      "epoch": 0.8223425196850394,
+      "grad_norm": 1.1455767154693604,
+      "learning_rate": 1.5193368676808927e-05,
+      "loss": 2.2666,
+      "step": 13368
+    },
+    {
+      "epoch": 0.8224040354330708,
+      "grad_norm": 1.1298049688339233,
+      "learning_rate": 1.518312343945093e-05,
+      "loss": 2.4648,
+      "step": 13369
+    },
+    {
+      "epoch": 0.8224655511811023,
+      "grad_norm": 1.4011069536209106,
+      "learning_rate": 1.5172881373771208e-05,
+      "loss": 2.8346,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8225270669291339,
+      "grad_norm": 1.56666898727417,
+      "learning_rate": 1.5162642480152689e-05,
+      "loss": 2.1494,
+      "step": 13371
+    },
+    {
+      "epoch": 0.8225885826771654,
+      "grad_norm": 1.3970715999603271,
+      "learning_rate": 1.5152406758978288e-05,
+      "loss": 2.8273,
+      "step": 13372
+    },
+    {
+      "epoch": 0.8226500984251969,
+      "grad_norm": 1.2909027338027954,
+      "learning_rate": 1.5142174210630744e-05,
+      "loss": 2.8118,
+      "step": 13373
+    },
+    {
+      "epoch": 0.8227116141732284,
+      "grad_norm": 1.361882209777832,
+      "learning_rate": 1.5131944835492717e-05,
+      "loss": 2.2243,
+      "step": 13374
+    },
+    {
+      "epoch": 0.8227731299212598,
+      "grad_norm": 2.251225233078003,
+      "learning_rate": 1.5121718633946702e-05,
+      "loss": 2.0159,
+      "step": 13375
+    },
+    {
+      "epoch": 0.8228346456692913,
+      "grad_norm": 1.2760025262832642,
+      "learning_rate": 1.5111495606375104e-05,
+      "loss": 2.5015,
+      "step": 13376
+    },
+    {
+      "epoch": 0.8228961614173228,
+      "grad_norm": 1.412812352180481,
+      "learning_rate": 1.5101275753160227e-05,
+      "loss": 3.164,
+      "step": 13377
+    },
+    {
+      "epoch": 0.8229576771653543,
+      "grad_norm": 1.2941250801086426,
+      "learning_rate": 1.5091059074684255e-05,
+      "loss": 2.4459,
+      "step": 13378
+    },
+    {
+      "epoch": 0.8230191929133859,
+      "grad_norm": 1.1791201829910278,
+      "learning_rate": 1.5080845571329172e-05,
+      "loss": 3.2668,
+      "step": 13379
+    },
+    {
+      "epoch": 0.8230807086614174,
+      "grad_norm": 1.3215348720550537,
+      "learning_rate": 1.5070635243476983e-05,
+      "loss": 2.2818,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8231422244094488,
+      "grad_norm": 1.1759097576141357,
+      "learning_rate": 1.5060428091509415e-05,
+      "loss": 2.2013,
+      "step": 13381
+    },
+    {
+      "epoch": 0.8232037401574803,
+      "grad_norm": 1.2831801176071167,
+      "learning_rate": 1.5050224115808209e-05,
+      "loss": 1.9531,
+      "step": 13382
+    },
+    {
+      "epoch": 0.8232652559055118,
+      "grad_norm": 1.2402325868606567,
+      "learning_rate": 1.5040023316754936e-05,
+      "loss": 2.7337,
+      "step": 13383
+    },
+    {
+      "epoch": 0.8233267716535433,
+      "grad_norm": 1.400732398033142,
+      "learning_rate": 1.502982569473106e-05,
+      "loss": 2.7394,
+      "step": 13384
+    },
+    {
+      "epoch": 0.8233882874015748,
+      "grad_norm": 1.4231297969818115,
+      "learning_rate": 1.5019631250117861e-05,
+      "loss": 2.2261,
+      "step": 13385
+    },
+    {
+      "epoch": 0.8234498031496063,
+      "grad_norm": 1.1960914134979248,
+      "learning_rate": 1.5009439983296626e-05,
+      "loss": 2.4747,
+      "step": 13386
+    },
+    {
+      "epoch": 0.8235113188976378,
+      "grad_norm": 1.2601139545440674,
+      "learning_rate": 1.4999251894648393e-05,
+      "loss": 2.4151,
+      "step": 13387
+    },
+    {
+      "epoch": 0.8235728346456693,
+      "grad_norm": 1.2874860763549805,
+      "learning_rate": 1.4989066984554156e-05,
+      "loss": 2.9209,
+      "step": 13388
+    },
+    {
+      "epoch": 0.8236343503937008,
+      "grad_norm": 1.332614779472351,
+      "learning_rate": 1.497888525339478e-05,
+      "loss": 2.6659,
+      "step": 13389
+    },
+    {
+      "epoch": 0.8236958661417323,
+      "grad_norm": 1.3907560110092163,
+      "learning_rate": 1.4968706701551028e-05,
+      "loss": 1.7161,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8237573818897638,
+      "grad_norm": 1.2892963886260986,
+      "learning_rate": 1.4958531329403502e-05,
+      "loss": 1.6958,
+      "step": 13391
+    },
+    {
+      "epoch": 0.8238188976377953,
+      "grad_norm": 1.4631807804107666,
+      "learning_rate": 1.4948359137332667e-05,
+      "loss": 2.226,
+      "step": 13392
+    },
+    {
+      "epoch": 0.8238804133858267,
+      "grad_norm": 1.2229204177856445,
+      "learning_rate": 1.4938190125718976e-05,
+      "loss": 2.541,
+      "step": 13393
+    },
+    {
+      "epoch": 0.8239419291338582,
+      "grad_norm": 1.2793203592300415,
+      "learning_rate": 1.4928024294942623e-05,
+      "loss": 2.9526,
+      "step": 13394
+    },
+    {
+      "epoch": 0.8240034448818898,
+      "grad_norm": 1.289116382598877,
+      "learning_rate": 1.4917861645383791e-05,
+      "loss": 1.6876,
+      "step": 13395
+    },
+    {
+      "epoch": 0.8240649606299213,
+      "grad_norm": 1.75065016746521,
+      "learning_rate": 1.49077021774225e-05,
+      "loss": 2.2813,
+      "step": 13396
+    },
+    {
+      "epoch": 0.8241264763779528,
+      "grad_norm": 1.4745858907699585,
+      "learning_rate": 1.4897545891438691e-05,
+      "loss": 2.2151,
+      "step": 13397
+    },
+    {
+      "epoch": 0.8241879921259843,
+      "grad_norm": 1.602856993675232,
+      "learning_rate": 1.4887392787812094e-05,
+      "loss": 2.6489,
+      "step": 13398
+    },
+    {
+      "epoch": 0.8242495078740157,
+      "grad_norm": 1.3704450130462646,
+      "learning_rate": 1.4877242866922425e-05,
+      "loss": 2.6683,
+      "step": 13399
+    },
+    {
+      "epoch": 0.8243110236220472,
+      "grad_norm": 1.595384120941162,
+      "learning_rate": 1.4867096129149193e-05,
+      "loss": 2.5106,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8243725393700787,
+      "grad_norm": 1.1539013385772705,
+      "learning_rate": 1.4856952574871851e-05,
+      "loss": 2.537,
+      "step": 13401
+    },
+    {
+      "epoch": 0.8244340551181102,
+      "grad_norm": 1.4620544910430908,
+      "learning_rate": 1.4846812204469718e-05,
+      "loss": 2.5549,
+      "step": 13402
+    },
+    {
+      "epoch": 0.8244955708661418,
+      "grad_norm": 1.3157626390457153,
+      "learning_rate": 1.4836675018322e-05,
+      "loss": 2.4209,
+      "step": 13403
+    },
+    {
+      "epoch": 0.8245570866141733,
+      "grad_norm": 1.29737389087677,
+      "learning_rate": 1.4826541016807727e-05,
+      "loss": 2.3206,
+      "step": 13404
+    },
+    {
+      "epoch": 0.8246186023622047,
+      "grad_norm": 1.216288447380066,
+      "learning_rate": 1.4816410200305908e-05,
+      "loss": 2.1895,
+      "step": 13405
+    },
+    {
+      "epoch": 0.8246801181102362,
+      "grad_norm": 1.3200807571411133,
+      "learning_rate": 1.4806282569195329e-05,
+      "loss": 2.3399,
+      "step": 13406
+    },
+    {
+      "epoch": 0.8247416338582677,
+      "grad_norm": 1.2356030941009521,
+      "learning_rate": 1.479615812385472e-05,
+      "loss": 2.3865,
+      "step": 13407
+    },
+    {
+      "epoch": 0.8248031496062992,
+      "grad_norm": 1.578279972076416,
+      "learning_rate": 1.47860368646627e-05,
+      "loss": 2.0491,
+      "step": 13408
+    },
+    {
+      "epoch": 0.8248646653543307,
+      "grad_norm": 2.116122245788574,
+      "learning_rate": 1.4775918791997745e-05,
+      "loss": 2.2208,
+      "step": 13409
+    },
+    {
+      "epoch": 0.8249261811023622,
+      "grad_norm": 1.1377906799316406,
+      "learning_rate": 1.4765803906238185e-05,
+      "loss": 2.5488,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8249876968503937,
+      "grad_norm": 1.176858901977539,
+      "learning_rate": 1.4755692207762306e-05,
+      "loss": 2.1065,
+      "step": 13411
+    },
+    {
+      "epoch": 0.8250492125984252,
+      "grad_norm": 1.514740228652954,
+      "learning_rate": 1.4745583696948184e-05,
+      "loss": 2.1275,
+      "step": 13412
+    },
+    {
+      "epoch": 0.8251107283464567,
+      "grad_norm": 1.305877923965454,
+      "learning_rate": 1.4735478374173861e-05,
+      "loss": 2.1058,
+      "step": 13413
+    },
+    {
+      "epoch": 0.8251722440944882,
+      "grad_norm": 1.3873614072799683,
+      "learning_rate": 1.4725376239817146e-05,
+      "loss": 2.6808,
+      "step": 13414
+    },
+    {
+      "epoch": 0.8252337598425197,
+      "grad_norm": 1.1359659433364868,
+      "learning_rate": 1.4715277294255914e-05,
+      "loss": 1.9242,
+      "step": 13415
+    },
+    {
+      "epoch": 0.8252952755905512,
+      "grad_norm": 1.5833005905151367,
+      "learning_rate": 1.4705181537867718e-05,
+      "loss": 2.4857,
+      "step": 13416
+    },
+    {
+      "epoch": 0.8253567913385826,
+      "grad_norm": 1.340254783630371,
+      "learning_rate": 1.4695088971030136e-05,
+      "loss": 2.2845,
+      "step": 13417
+    },
+    {
+      "epoch": 0.8254183070866141,
+      "grad_norm": 1.1349413394927979,
+      "learning_rate": 1.468499959412054e-05,
+      "loss": 2.0249,
+      "step": 13418
+    },
+    {
+      "epoch": 0.8254798228346457,
+      "grad_norm": 1.2487751245498657,
+      "learning_rate": 1.4674913407516244e-05,
+      "loss": 2.3158,
+      "step": 13419
+    },
+    {
+      "epoch": 0.8255413385826772,
+      "grad_norm": 1.851384162902832,
+      "learning_rate": 1.4664830411594377e-05,
+      "loss": 1.8227,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8256028543307087,
+      "grad_norm": 1.2711199522018433,
+      "learning_rate": 1.4654750606732015e-05,
+      "loss": 2.1081,
+      "step": 13421
+    },
+    {
+      "epoch": 0.8256643700787402,
+      "grad_norm": 1.2693164348602295,
+      "learning_rate": 1.4644673993306091e-05,
+      "loss": 2.7193,
+      "step": 13422
+    },
+    {
+      "epoch": 0.8257258858267716,
+      "grad_norm": 1.3096375465393066,
+      "learning_rate": 1.4634600571693413e-05,
+      "loss": 2.7289,
+      "step": 13423
+    },
+    {
+      "epoch": 0.8257874015748031,
+      "grad_norm": 1.640021562576294,
+      "learning_rate": 1.4624530342270648e-05,
+      "loss": 2.7177,
+      "step": 13424
+    },
+    {
+      "epoch": 0.8258489173228346,
+      "grad_norm": 1.146618127822876,
+      "learning_rate": 1.4614463305414405e-05,
+      "loss": 2.6492,
+      "step": 13425
+    },
+    {
+      "epoch": 0.8259104330708661,
+      "grad_norm": 1.2580705881118774,
+      "learning_rate": 1.4604399461501084e-05,
+      "loss": 3.141,
+      "step": 13426
+    },
+    {
+      "epoch": 0.8259719488188977,
+      "grad_norm": 1.250205397605896,
+      "learning_rate": 1.4594338810907049e-05,
+      "loss": 2.6803,
+      "step": 13427
+    },
+    {
+      "epoch": 0.8260334645669292,
+      "grad_norm": 1.2586356401443481,
+      "learning_rate": 1.4584281354008521e-05,
+      "loss": 2.0856,
+      "step": 13428
+    },
+    {
+      "epoch": 0.8260949803149606,
+      "grad_norm": 1.3580535650253296,
+      "learning_rate": 1.457422709118157e-05,
+      "loss": 2.2191,
+      "step": 13429
+    },
+    {
+      "epoch": 0.8261564960629921,
+      "grad_norm": 1.2086689472198486,
+      "learning_rate": 1.456417602280219e-05,
+      "loss": 2.1051,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8262180118110236,
+      "grad_norm": 1.3138248920440674,
+      "learning_rate": 1.4554128149246204e-05,
+      "loss": 2.3359,
+      "step": 13431
+    },
+    {
+      "epoch": 0.8262795275590551,
+      "grad_norm": 1.4127150774002075,
+      "learning_rate": 1.454408347088937e-05,
+      "loss": 2.6739,
+      "step": 13432
+    },
+    {
+      "epoch": 0.8263410433070866,
+      "grad_norm": 1.208866834640503,
+      "learning_rate": 1.4534041988107295e-05,
+      "loss": 2.6312,
+      "step": 13433
+    },
+    {
+      "epoch": 0.8264025590551181,
+      "grad_norm": 1.2832660675048828,
+      "learning_rate": 1.45240037012755e-05,
+      "loss": 2.0447,
+      "step": 13434
+    },
+    {
+      "epoch": 0.8264640748031497,
+      "grad_norm": 1.3717795610427856,
+      "learning_rate": 1.4513968610769323e-05,
+      "loss": 2.4922,
+      "step": 13435
+    },
+    {
+      "epoch": 0.8265255905511811,
+      "grad_norm": 1.2197751998901367,
+      "learning_rate": 1.4503936716964062e-05,
+      "loss": 2.0574,
+      "step": 13436
+    },
+    {
+      "epoch": 0.8265871062992126,
+      "grad_norm": 1.2487410306930542,
+      "learning_rate": 1.4493908020234793e-05,
+      "loss": 2.6314,
+      "step": 13437
+    },
+    {
+      "epoch": 0.8266486220472441,
+      "grad_norm": 1.3177988529205322,
+      "learning_rate": 1.4483882520956604e-05,
+      "loss": 2.6959,
+      "step": 13438
+    },
+    {
+      "epoch": 0.8267101377952756,
+      "grad_norm": 1.3181246519088745,
+      "learning_rate": 1.447386021950432e-05,
+      "loss": 2.2173,
+      "step": 13439
+    },
+    {
+      "epoch": 0.8267716535433071,
+      "grad_norm": 1.0826162099838257,
+      "learning_rate": 1.4463841116252774e-05,
+      "loss": 2.1826,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8268331692913385,
+      "grad_norm": 1.2722645998001099,
+      "learning_rate": 1.44538252115766e-05,
+      "loss": 3.1042,
+      "step": 13441
+    },
+    {
+      "epoch": 0.82689468503937,
+      "grad_norm": 1.41724693775177,
+      "learning_rate": 1.4443812505850363e-05,
+      "loss": 1.7753,
+      "step": 13442
+    },
+    {
+      "epoch": 0.8269562007874016,
+      "grad_norm": 1.4945446252822876,
+      "learning_rate": 1.4433802999448443e-05,
+      "loss": 2.5729,
+      "step": 13443
+    },
+    {
+      "epoch": 0.8270177165354331,
+      "grad_norm": 1.3562010526657104,
+      "learning_rate": 1.4423796692745194e-05,
+      "loss": 2.6448,
+      "step": 13444
+    },
+    {
+      "epoch": 0.8270792322834646,
+      "grad_norm": 1.8992455005645752,
+      "learning_rate": 1.4413793586114733e-05,
+      "loss": 1.8846,
+      "step": 13445
+    },
+    {
+      "epoch": 0.8271407480314961,
+      "grad_norm": 1.2319872379302979,
+      "learning_rate": 1.4403793679931143e-05,
+      "loss": 2.537,
+      "step": 13446
+    },
+    {
+      "epoch": 0.8272022637795275,
+      "grad_norm": 1.383763074874878,
+      "learning_rate": 1.4393796974568374e-05,
+      "loss": 2.7724,
+      "step": 13447
+    },
+    {
+      "epoch": 0.827263779527559,
+      "grad_norm": 1.616618037223816,
+      "learning_rate": 1.438380347040027e-05,
+      "loss": 2.5843,
+      "step": 13448
+    },
+    {
+      "epoch": 0.8273252952755905,
+      "grad_norm": 1.4952280521392822,
+      "learning_rate": 1.4373813167800488e-05,
+      "loss": 2.1945,
+      "step": 13449
+    },
+    {
+      "epoch": 0.827386811023622,
+      "grad_norm": 1.3261864185333252,
+      "learning_rate": 1.4363826067142649e-05,
+      "loss": 2.7073,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8274483267716536,
+      "grad_norm": 1.1845760345458984,
+      "learning_rate": 1.435384216880018e-05,
+      "loss": 2.93,
+      "step": 13451
+    },
+    {
+      "epoch": 0.8275098425196851,
+      "grad_norm": 1.3976995944976807,
+      "learning_rate": 1.4343861473146424e-05,
+      "loss": 2.0125,
+      "step": 13452
+    },
+    {
+      "epoch": 0.8275713582677166,
+      "grad_norm": 1.181937336921692,
+      "learning_rate": 1.4333883980554619e-05,
+      "loss": 3.0793,
+      "step": 13453
+    },
+    {
+      "epoch": 0.827632874015748,
+      "grad_norm": 1.751163363456726,
+      "learning_rate": 1.4323909691397897e-05,
+      "loss": 2.11,
+      "step": 13454
+    },
+    {
+      "epoch": 0.8276943897637795,
+      "grad_norm": 1.2348593473434448,
+      "learning_rate": 1.4313938606049181e-05,
+      "loss": 2.2917,
+      "step": 13455
+    },
+    {
+      "epoch": 0.827755905511811,
+      "grad_norm": 1.401218056678772,
+      "learning_rate": 1.4303970724881388e-05,
+      "loss": 2.8771,
+      "step": 13456
+    },
+    {
+      "epoch": 0.8278174212598425,
+      "grad_norm": 1.8048532009124756,
+      "learning_rate": 1.4294006048267217e-05,
+      "loss": 1.4967,
+      "step": 13457
+    },
+    {
+      "epoch": 0.827878937007874,
+      "grad_norm": 1.1902509927749634,
+      "learning_rate": 1.4284044576579303e-05,
+      "loss": 1.7998,
+      "step": 13458
+    },
+    {
+      "epoch": 0.8279404527559056,
+      "grad_norm": 1.4024416208267212,
+      "learning_rate": 1.427408631019016e-05,
+      "loss": 2.3577,
+      "step": 13459
+    },
+    {
+      "epoch": 0.828001968503937,
+      "grad_norm": 1.5433158874511719,
+      "learning_rate": 1.426413124947219e-05,
+      "loss": 2.7255,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8280634842519685,
+      "grad_norm": 1.2900489568710327,
+      "learning_rate": 1.4254179394797607e-05,
+      "loss": 1.856,
+      "step": 13461
+    },
+    {
+      "epoch": 0.828125,
+      "grad_norm": 1.394742727279663,
+      "learning_rate": 1.424423074653861e-05,
+      "loss": 2.0899,
+      "step": 13462
+    },
+    {
+      "epoch": 0.8281865157480315,
+      "grad_norm": 1.300533652305603,
+      "learning_rate": 1.4234285305067196e-05,
+      "loss": 2.2524,
+      "step": 13463
+    },
+    {
+      "epoch": 0.828248031496063,
+      "grad_norm": 1.1617172956466675,
+      "learning_rate": 1.4224343070755242e-05,
+      "loss": 2.5453,
+      "step": 13464
+    },
+    {
+      "epoch": 0.8283095472440944,
+      "grad_norm": 1.3372902870178223,
+      "learning_rate": 1.4214404043974573e-05,
+      "loss": 2.0935,
+      "step": 13465
+    },
+    {
+      "epoch": 0.828371062992126,
+      "grad_norm": 1.181625247001648,
+      "learning_rate": 1.4204468225096824e-05,
+      "loss": 2.6392,
+      "step": 13466
+    },
+    {
+      "epoch": 0.8284325787401575,
+      "grad_norm": 1.4105547666549683,
+      "learning_rate": 1.4194535614493575e-05,
+      "loss": 2.5537,
+      "step": 13467
+    },
+    {
+      "epoch": 0.828494094488189,
+      "grad_norm": 1.37892746925354,
+      "learning_rate": 1.4184606212536222e-05,
+      "loss": 2.2563,
+      "step": 13468
+    },
+    {
+      "epoch": 0.8285556102362205,
+      "grad_norm": 1.2905385494232178,
+      "learning_rate": 1.4174680019596087e-05,
+      "loss": 2.4002,
+      "step": 13469
+    },
+    {
+      "epoch": 0.828617125984252,
+      "grad_norm": 1.3023617267608643,
+      "learning_rate": 1.416475703604433e-05,
+      "loss": 2.695,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8286786417322834,
+      "grad_norm": 1.2161227464675903,
+      "learning_rate": 1.4154837262252029e-05,
+      "loss": 2.7119,
+      "step": 13471
+    },
+    {
+      "epoch": 0.8287401574803149,
+      "grad_norm": 1.233875036239624,
+      "learning_rate": 1.4144920698590126e-05,
+      "loss": 2.5763,
+      "step": 13472
+    },
+    {
+      "epoch": 0.8288016732283464,
+      "grad_norm": 1.5652263164520264,
+      "learning_rate": 1.4135007345429464e-05,
+      "loss": 1.6546,
+      "step": 13473
+    },
+    {
+      "epoch": 0.828863188976378,
+      "grad_norm": 1.273348331451416,
+      "learning_rate": 1.4125097203140713e-05,
+      "loss": 2.3848,
+      "step": 13474
+    },
+    {
+      "epoch": 0.8289247047244095,
+      "grad_norm": 1.4840478897094727,
+      "learning_rate": 1.4115190272094502e-05,
+      "loss": 2.5868,
+      "step": 13475
+    },
+    {
+      "epoch": 0.828986220472441,
+      "grad_norm": 1.2372716665267944,
+      "learning_rate": 1.4105286552661246e-05,
+      "loss": 2.3486,
+      "step": 13476
+    },
+    {
+      "epoch": 0.8290477362204725,
+      "grad_norm": 1.2446871995925903,
+      "learning_rate": 1.40953860452113e-05,
+      "loss": 2.4053,
+      "step": 13477
+    },
+    {
+      "epoch": 0.8291092519685039,
+      "grad_norm": 1.9095473289489746,
+      "learning_rate": 1.4085488750114905e-05,
+      "loss": 2.4048,
+      "step": 13478
+    },
+    {
+      "epoch": 0.8291707677165354,
+      "grad_norm": 1.357962727546692,
+      "learning_rate": 1.4075594667742186e-05,
+      "loss": 2.6444,
+      "step": 13479
+    },
+    {
+      "epoch": 0.8292322834645669,
+      "grad_norm": 1.2555745840072632,
+      "learning_rate": 1.406570379846307e-05,
+      "loss": 2.4771,
+      "step": 13480
+    },
+    {
+      "epoch": 0.8292937992125984,
+      "grad_norm": 1.3750083446502686,
+      "learning_rate": 1.405581614264747e-05,
+      "loss": 2.9658,
+      "step": 13481
+    },
+    {
+      "epoch": 0.82935531496063,
+      "grad_norm": 1.2771682739257812,
+      "learning_rate": 1.4045931700665094e-05,
+      "loss": 2.417,
+      "step": 13482
+    },
+    {
+      "epoch": 0.8294168307086615,
+      "grad_norm": 1.4624372720718384,
+      "learning_rate": 1.4036050472885598e-05,
+      "loss": 2.5453,
+      "step": 13483
+    },
+    {
+      "epoch": 0.8294783464566929,
+      "grad_norm": 1.2762340307235718,
+      "learning_rate": 1.4026172459678422e-05,
+      "loss": 2.2518,
+      "step": 13484
+    },
+    {
+      "epoch": 0.8295398622047244,
+      "grad_norm": 1.3824785947799683,
+      "learning_rate": 1.4016297661413036e-05,
+      "loss": 2.222,
+      "step": 13485
+    },
+    {
+      "epoch": 0.8296013779527559,
+      "grad_norm": 1.1869293451309204,
+      "learning_rate": 1.4006426078458634e-05,
+      "loss": 2.3718,
+      "step": 13486
+    },
+    {
+      "epoch": 0.8296628937007874,
+      "grad_norm": 1.218076467514038,
+      "learning_rate": 1.399655771118441e-05,
+      "loss": 1.6794,
+      "step": 13487
+    },
+    {
+      "epoch": 0.8297244094488189,
+      "grad_norm": 1.0566259622573853,
+      "learning_rate": 1.3986692559959335e-05,
+      "loss": 2.4098,
+      "step": 13488
+    },
+    {
+      "epoch": 0.8297859251968503,
+      "grad_norm": 1.291303038597107,
+      "learning_rate": 1.397683062515236e-05,
+      "loss": 3.4405,
+      "step": 13489
+    },
+    {
+      "epoch": 0.8298474409448819,
+      "grad_norm": 1.2819740772247314,
+      "learning_rate": 1.396697190713221e-05,
+      "loss": 2.0346,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8299089566929134,
+      "grad_norm": 1.3591059446334839,
+      "learning_rate": 1.3957116406267589e-05,
+      "loss": 2.5825,
+      "step": 13491
+    },
+    {
+      "epoch": 0.8299704724409449,
+      "grad_norm": 1.1924370527267456,
+      "learning_rate": 1.3947264122927018e-05,
+      "loss": 3.1996,
+      "step": 13492
+    },
+    {
+      "epoch": 0.8300319881889764,
+      "grad_norm": 1.2846355438232422,
+      "learning_rate": 1.3937415057478954e-05,
+      "loss": 2.9128,
+      "step": 13493
+    },
+    {
+      "epoch": 0.8300935039370079,
+      "grad_norm": 1.3276803493499756,
+      "learning_rate": 1.3927569210291635e-05,
+      "loss": 2.1286,
+      "step": 13494
+    },
+    {
+      "epoch": 0.8301550196850394,
+      "grad_norm": 1.258808970451355,
+      "learning_rate": 1.3917726581733314e-05,
+      "loss": 2.6538,
+      "step": 13495
+    },
+    {
+      "epoch": 0.8302165354330708,
+      "grad_norm": 1.574002981185913,
+      "learning_rate": 1.3907887172171975e-05,
+      "loss": 2.3021,
+      "step": 13496
+    },
+    {
+      "epoch": 0.8302780511811023,
+      "grad_norm": 1.1833086013793945,
+      "learning_rate": 1.3898050981975597e-05,
+      "loss": 2.7815,
+      "step": 13497
+    },
+    {
+      "epoch": 0.8303395669291339,
+      "grad_norm": 1.263008952140808,
+      "learning_rate": 1.3888218011512e-05,
+      "loss": 2.2309,
+      "step": 13498
+    },
+    {
+      "epoch": 0.8304010826771654,
+      "grad_norm": 1.55586576461792,
+      "learning_rate": 1.3878388261148889e-05,
+      "loss": 2.3424,
+      "step": 13499
+    },
+    {
+      "epoch": 0.8304625984251969,
+      "grad_norm": 1.0271414518356323,
+      "learning_rate": 1.3868561731253838e-05,
+      "loss": 2.4085,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8305241141732284,
+      "grad_norm": 1.4371401071548462,
+      "learning_rate": 1.3858738422194273e-05,
+      "loss": 2.1892,
+      "step": 13501
+    },
+    {
+      "epoch": 0.8305856299212598,
+      "grad_norm": 1.3282567262649536,
+      "learning_rate": 1.3848918334337556e-05,
+      "loss": 2.2268,
+      "step": 13502
+    },
+    {
+      "epoch": 0.8306471456692913,
+      "grad_norm": 1.21855890750885,
+      "learning_rate": 1.3839101468050897e-05,
+      "loss": 2.3687,
+      "step": 13503
+    },
+    {
+      "epoch": 0.8307086614173228,
+      "grad_norm": 1.4760675430297852,
+      "learning_rate": 1.382928782370142e-05,
+      "loss": 2.5532,
+      "step": 13504
+    },
+    {
+      "epoch": 0.8307701771653543,
+      "grad_norm": 1.178659439086914,
+      "learning_rate": 1.3819477401656055e-05,
+      "loss": 2.691,
+      "step": 13505
+    },
+    {
+      "epoch": 0.8308316929133859,
+      "grad_norm": 1.156262755393982,
+      "learning_rate": 1.3809670202281699e-05,
+      "loss": 2.2022,
+      "step": 13506
+    },
+    {
+      "epoch": 0.8308932086614174,
+      "grad_norm": 2.0985732078552246,
+      "learning_rate": 1.3799866225945046e-05,
+      "loss": 2.6794,
+      "step": 13507
+    },
+    {
+      "epoch": 0.8309547244094488,
+      "grad_norm": 1.2600396871566772,
+      "learning_rate": 1.3790065473012759e-05,
+      "loss": 1.7584,
+      "step": 13508
+    },
+    {
+      "epoch": 0.8310162401574803,
+      "grad_norm": 1.4835541248321533,
+      "learning_rate": 1.3780267943851277e-05,
+      "loss": 2.0223,
+      "step": 13509
+    },
+    {
+      "epoch": 0.8310777559055118,
+      "grad_norm": 1.125472903251648,
+      "learning_rate": 1.3770473638826998e-05,
+      "loss": 2.4793,
+      "step": 13510
+    },
+    {
+      "epoch": 0.8311392716535433,
+      "grad_norm": 1.3009390830993652,
+      "learning_rate": 1.3760682558306181e-05,
+      "loss": 2.5768,
+      "step": 13511
+    },
+    {
+      "epoch": 0.8312007874015748,
+      "grad_norm": 1.4079819917678833,
+      "learning_rate": 1.3750894702654971e-05,
+      "loss": 2.7238,
+      "step": 13512
+    },
+    {
+      "epoch": 0.8312623031496063,
+      "grad_norm": 1.2897136211395264,
+      "learning_rate": 1.3741110072239339e-05,
+      "loss": 2.3471,
+      "step": 13513
+    },
+    {
+      "epoch": 0.8313238188976378,
+      "grad_norm": 1.36943519115448,
+      "learning_rate": 1.3731328667425213e-05,
+      "loss": 2.2013,
+      "step": 13514
+    },
+    {
+      "epoch": 0.8313853346456693,
+      "grad_norm": 1.1729415655136108,
+      "learning_rate": 1.3721550488578339e-05,
+      "loss": 2.7949,
+      "step": 13515
+    },
+    {
+      "epoch": 0.8314468503937008,
+      "grad_norm": 1.1480236053466797,
+      "learning_rate": 1.3711775536064363e-05,
+      "loss": 2.3988,
+      "step": 13516
+    },
+    {
+      "epoch": 0.8315083661417323,
+      "grad_norm": 1.4465248584747314,
+      "learning_rate": 1.370200381024883e-05,
+      "loss": 2.3974,
+      "step": 13517
+    },
+    {
+      "epoch": 0.8315698818897638,
+      "grad_norm": 1.2057955265045166,
+      "learning_rate": 1.3692235311497171e-05,
+      "loss": 2.4137,
+      "step": 13518
+    },
+    {
+      "epoch": 0.8316313976377953,
+      "grad_norm": 1.2424349784851074,
+      "learning_rate": 1.3682470040174621e-05,
+      "loss": 1.7755,
+      "step": 13519
+    },
+    {
+      "epoch": 0.8316929133858267,
+      "grad_norm": 1.2211989164352417,
+      "learning_rate": 1.3672707996646394e-05,
+      "loss": 2.7897,
+      "step": 13520
+    },
+    {
+      "epoch": 0.8317544291338582,
+      "grad_norm": 1.1864173412322998,
+      "learning_rate": 1.3662949181277485e-05,
+      "loss": 2.5169,
+      "step": 13521
+    },
+    {
+      "epoch": 0.8318159448818898,
+      "grad_norm": 1.3299440145492554,
+      "learning_rate": 1.3653193594432856e-05,
+      "loss": 2.8882,
+      "step": 13522
+    },
+    {
+      "epoch": 0.8318774606299213,
+      "grad_norm": 1.3242558240890503,
+      "learning_rate": 1.3643441236477295e-05,
+      "loss": 3.0996,
+      "step": 13523
+    },
+    {
+      "epoch": 0.8319389763779528,
+      "grad_norm": 1.2453540563583374,
+      "learning_rate": 1.3633692107775531e-05,
+      "loss": 2.8602,
+      "step": 13524
+    },
+    {
+      "epoch": 0.8320004921259843,
+      "grad_norm": 1.3467925786972046,
+      "learning_rate": 1.3623946208692051e-05,
+      "loss": 1.9426,
+      "step": 13525
+    },
+    {
+      "epoch": 0.8320620078740157,
+      "grad_norm": 1.3340381383895874,
+      "learning_rate": 1.3614203539591374e-05,
+      "loss": 2.2932,
+      "step": 13526
+    },
+    {
+      "epoch": 0.8321235236220472,
+      "grad_norm": 1.264677882194519,
+      "learning_rate": 1.3604464100837754e-05,
+      "loss": 1.6843,
+      "step": 13527
+    },
+    {
+      "epoch": 0.8321850393700787,
+      "grad_norm": 1.5906867980957031,
+      "learning_rate": 1.3594727892795423e-05,
+      "loss": 2.1791,
+      "step": 13528
+    },
+    {
+      "epoch": 0.8322465551181102,
+      "grad_norm": 1.1642792224884033,
+      "learning_rate": 1.3584994915828463e-05,
+      "loss": 2.8518,
+      "step": 13529
+    },
+    {
+      "epoch": 0.8323080708661418,
+      "grad_norm": 1.33290433883667,
+      "learning_rate": 1.3575265170300844e-05,
+      "loss": 2.2191,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8323695866141733,
+      "grad_norm": 2.048370599746704,
+      "learning_rate": 1.356553865657637e-05,
+      "loss": 4.3363,
+      "step": 13531
+    },
+    {
+      "epoch": 0.8324311023622047,
+      "grad_norm": 1.4916508197784424,
+      "learning_rate": 1.355581537501881e-05,
+      "loss": 2.8856,
+      "step": 13532
+    },
+    {
+      "epoch": 0.8324926181102362,
+      "grad_norm": 1.3993210792541504,
+      "learning_rate": 1.3546095325991704e-05,
+      "loss": 2.4486,
+      "step": 13533
+    },
+    {
+      "epoch": 0.8325541338582677,
+      "grad_norm": 1.3771326541900635,
+      "learning_rate": 1.353637850985856e-05,
+      "loss": 2.5945,
+      "step": 13534
+    },
+    {
+      "epoch": 0.8326156496062992,
+      "grad_norm": 1.4008641242980957,
+      "learning_rate": 1.3526664926982724e-05,
+      "loss": 2.5207,
+      "step": 13535
+    },
+    {
+      "epoch": 0.8326771653543307,
+      "grad_norm": 1.304033637046814,
+      "learning_rate": 1.3516954577727414e-05,
+      "loss": 2.4335,
+      "step": 13536
+    },
+    {
+      "epoch": 0.8327386811023622,
+      "grad_norm": 1.5123192071914673,
+      "learning_rate": 1.3507247462455797e-05,
+      "loss": 2.7516,
+      "step": 13537
+    },
+    {
+      "epoch": 0.8328001968503937,
+      "grad_norm": 1.2659631967544556,
+      "learning_rate": 1.3497543581530792e-05,
+      "loss": 2.6913,
+      "step": 13538
+    },
+    {
+      "epoch": 0.8328617125984252,
+      "grad_norm": 1.30533766746521,
+      "learning_rate": 1.3487842935315342e-05,
+      "loss": 2.5126,
+      "step": 13539
+    },
+    {
+      "epoch": 0.8329232283464567,
+      "grad_norm": 1.2237850427627563,
+      "learning_rate": 1.3478145524172125e-05,
+      "loss": 2.0071,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8329847440944882,
+      "grad_norm": 1.4037691354751587,
+      "learning_rate": 1.3468451348463817e-05,
+      "loss": 2.684,
+      "step": 13541
+    },
+    {
+      "epoch": 0.8330462598425197,
+      "grad_norm": 1.3330198526382446,
+      "learning_rate": 1.3458760408552918e-05,
+      "loss": 2.8082,
+      "step": 13542
+    },
+    {
+      "epoch": 0.8331077755905512,
+      "grad_norm": 1.4718306064605713,
+      "learning_rate": 1.3449072704801824e-05,
+      "loss": 2.0498,
+      "step": 13543
+    },
+    {
+      "epoch": 0.8331692913385826,
+      "grad_norm": 1.232833981513977,
+      "learning_rate": 1.343938823757277e-05,
+      "loss": 2.6049,
+      "step": 13544
+    },
+    {
+      "epoch": 0.8332308070866141,
+      "grad_norm": 1.283528447151184,
+      "learning_rate": 1.3429707007227943e-05,
+      "loss": 2.2144,
+      "step": 13545
+    },
+    {
+      "epoch": 0.8332923228346457,
+      "grad_norm": 1.1374297142028809,
+      "learning_rate": 1.3420029014129332e-05,
+      "loss": 2.877,
+      "step": 13546
+    },
+    {
+      "epoch": 0.8333538385826772,
+      "grad_norm": 1.3956019878387451,
+      "learning_rate": 1.3410354258638846e-05,
+      "loss": 2.6634,
+      "step": 13547
+    },
+    {
+      "epoch": 0.8334153543307087,
+      "grad_norm": 1.3988920450210571,
+      "learning_rate": 1.3400682741118265e-05,
+      "loss": 2.2579,
+      "step": 13548
+    },
+    {
+      "epoch": 0.8334768700787402,
+      "grad_norm": 1.2732181549072266,
+      "learning_rate": 1.3391014461929297e-05,
+      "loss": 2.9693,
+      "step": 13549
+    },
+    {
+      "epoch": 0.8335383858267716,
+      "grad_norm": 1.4336764812469482,
+      "learning_rate": 1.3381349421433409e-05,
+      "loss": 2.2807,
+      "step": 13550
+    },
+    {
+      "epoch": 0.8335999015748031,
+      "grad_norm": 1.202507495880127,
+      "learning_rate": 1.337168761999208e-05,
+      "loss": 2.1553,
+      "step": 13551
+    },
+    {
+      "epoch": 0.8336614173228346,
+      "grad_norm": 1.1301060914993286,
+      "learning_rate": 1.3362029057966552e-05,
+      "loss": 2.4327,
+      "step": 13552
+    },
+    {
+      "epoch": 0.8337229330708661,
+      "grad_norm": 1.5090234279632568,
+      "learning_rate": 1.335237373571806e-05,
+      "loss": 1.989,
+      "step": 13553
+    },
+    {
+      "epoch": 0.8337844488188977,
+      "grad_norm": 1.371552586555481,
+      "learning_rate": 1.3342721653607593e-05,
+      "loss": 2.4325,
+      "step": 13554
+    },
+    {
+      "epoch": 0.8338459645669292,
+      "grad_norm": 1.627500057220459,
+      "learning_rate": 1.333307281199615e-05,
+      "loss": 2.0287,
+      "step": 13555
+    },
+    {
+      "epoch": 0.8339074803149606,
+      "grad_norm": 1.2254263162612915,
+      "learning_rate": 1.332342721124451e-05,
+      "loss": 1.7508,
+      "step": 13556
+    },
+    {
+      "epoch": 0.8339689960629921,
+      "grad_norm": 1.3046092987060547,
+      "learning_rate": 1.3313784851713384e-05,
+      "loss": 2.612,
+      "step": 13557
+    },
+    {
+      "epoch": 0.8340305118110236,
+      "grad_norm": 1.1988481283187866,
+      "learning_rate": 1.3304145733763307e-05,
+      "loss": 2.7787,
+      "step": 13558
+    },
+    {
+      "epoch": 0.8340920275590551,
+      "grad_norm": 1.3754034042358398,
+      "learning_rate": 1.3294509857754767e-05,
+      "loss": 2.3729,
+      "step": 13559
+    },
+    {
+      "epoch": 0.8341535433070866,
+      "grad_norm": 1.2305526733398438,
+      "learning_rate": 1.3284877224048064e-05,
+      "loss": 1.9325,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8342150590551181,
+      "grad_norm": 1.351078987121582,
+      "learning_rate": 1.3275247833003413e-05,
+      "loss": 2.3623,
+      "step": 13561
+    },
+    {
+      "epoch": 0.8342765748031497,
+      "grad_norm": 2.0601260662078857,
+      "learning_rate": 1.32656216849809e-05,
+      "loss": 2.0822,
+      "step": 13562
+    },
+    {
+      "epoch": 0.8343380905511811,
+      "grad_norm": 1.1850743293762207,
+      "learning_rate": 1.3255998780340506e-05,
+      "loss": 1.9806,
+      "step": 13563
+    },
+    {
+      "epoch": 0.8343996062992126,
+      "grad_norm": 1.344904899597168,
+      "learning_rate": 1.3246379119442054e-05,
+      "loss": 2.6679,
+      "step": 13564
+    },
+    {
+      "epoch": 0.8344611220472441,
+      "grad_norm": 1.2401747703552246,
+      "learning_rate": 1.3236762702645278e-05,
+      "loss": 2.6084,
+      "step": 13565
+    },
+    {
+      "epoch": 0.8345226377952756,
+      "grad_norm": 1.2583757638931274,
+      "learning_rate": 1.3227149530309757e-05,
+      "loss": 2.0724,
+      "step": 13566
+    },
+    {
+      "epoch": 0.8345841535433071,
+      "grad_norm": 1.0712159872055054,
+      "learning_rate": 1.3217539602794993e-05,
+      "loss": 2.6939,
+      "step": 13567
+    },
+    {
+      "epoch": 0.8346456692913385,
+      "grad_norm": 1.2889022827148438,
+      "learning_rate": 1.320793292046032e-05,
+      "loss": 2.2829,
+      "step": 13568
+    },
+    {
+      "epoch": 0.83470718503937,
+      "grad_norm": 1.3243582248687744,
+      "learning_rate": 1.3198329483665028e-05,
+      "loss": 2.4005,
+      "step": 13569
+    },
+    {
+      "epoch": 0.8347687007874016,
+      "grad_norm": 1.1312612295150757,
+      "learning_rate": 1.3188729292768153e-05,
+      "loss": 2.781,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8348302165354331,
+      "grad_norm": 1.869491457939148,
+      "learning_rate": 1.3179132348128764e-05,
+      "loss": 3.5675,
+      "step": 13571
+    },
+    {
+      "epoch": 0.8348917322834646,
+      "grad_norm": 2.1644952297210693,
+      "learning_rate": 1.3169538650105684e-05,
+      "loss": 1.8476,
+      "step": 13572
+    },
+    {
+      "epoch": 0.8349532480314961,
+      "grad_norm": 1.7244873046875,
+      "learning_rate": 1.315994819905767e-05,
+      "loss": 2.6177,
+      "step": 13573
+    },
+    {
+      "epoch": 0.8350147637795275,
+      "grad_norm": 1.3625125885009766,
+      "learning_rate": 1.315036099534338e-05,
+      "loss": 2.1781,
+      "step": 13574
+    },
+    {
+      "epoch": 0.835076279527559,
+      "grad_norm": 1.4001919031143188,
+      "learning_rate": 1.3140777039321272e-05,
+      "loss": 2.3614,
+      "step": 13575
+    },
+    {
+      "epoch": 0.8351377952755905,
+      "grad_norm": 1.2943285703659058,
+      "learning_rate": 1.3131196331349804e-05,
+      "loss": 2.806,
+      "step": 13576
+    },
+    {
+      "epoch": 0.835199311023622,
+      "grad_norm": 1.5758682489395142,
+      "learning_rate": 1.3121618871787167e-05,
+      "loss": 2.9269,
+      "step": 13577
+    },
+    {
+      "epoch": 0.8352608267716536,
+      "grad_norm": 1.3780648708343506,
+      "learning_rate": 1.3112044660991552e-05,
+      "loss": 2.3427,
+      "step": 13578
+    },
+    {
+      "epoch": 0.8353223425196851,
+      "grad_norm": 1.2078310251235962,
+      "learning_rate": 1.310247369932095e-05,
+      "loss": 2.8494,
+      "step": 13579
+    },
+    {
+      "epoch": 0.8353838582677166,
+      "grad_norm": 1.1595937013626099,
+      "learning_rate": 1.3092905987133286e-05,
+      "loss": 2.9214,
+      "step": 13580
+    },
+    {
+      "epoch": 0.835445374015748,
+      "grad_norm": 1.412774920463562,
+      "learning_rate": 1.308334152478632e-05,
+      "loss": 2.7087,
+      "step": 13581
+    },
+    {
+      "epoch": 0.8355068897637795,
+      "grad_norm": 1.7293038368225098,
+      "learning_rate": 1.307378031263775e-05,
+      "loss": 2.0795,
+      "step": 13582
+    },
+    {
+      "epoch": 0.835568405511811,
+      "grad_norm": 1.3709344863891602,
+      "learning_rate": 1.3064222351045052e-05,
+      "loss": 2.6821,
+      "step": 13583
+    },
+    {
+      "epoch": 0.8356299212598425,
+      "grad_norm": 1.5760574340820312,
+      "learning_rate": 1.3054667640365702e-05,
+      "loss": 1.7929,
+      "step": 13584
+    },
+    {
+      "epoch": 0.835691437007874,
+      "grad_norm": 1.368003487586975,
+      "learning_rate": 1.304511618095693e-05,
+      "loss": 2.3667,
+      "step": 13585
+    },
+    {
+      "epoch": 0.8357529527559056,
+      "grad_norm": 1.4018962383270264,
+      "learning_rate": 1.3035567973175944e-05,
+      "loss": 2.1535,
+      "step": 13586
+    },
+    {
+      "epoch": 0.835814468503937,
+      "grad_norm": 1.1832720041275024,
+      "learning_rate": 1.3026023017379796e-05,
+      "loss": 2.0326,
+      "step": 13587
+    },
+    {
+      "epoch": 0.8358759842519685,
+      "grad_norm": 1.4939030408859253,
+      "learning_rate": 1.3016481313925421e-05,
+      "loss": 2.5962,
+      "step": 13588
+    },
+    {
+      "epoch": 0.8359375,
+      "grad_norm": 1.2471052408218384,
+      "learning_rate": 1.3006942863169591e-05,
+      "loss": 2.9111,
+      "step": 13589
+    },
+    {
+      "epoch": 0.8359990157480315,
+      "grad_norm": 1.2167617082595825,
+      "learning_rate": 1.299740766546903e-05,
+      "loss": 2.3462,
+      "step": 13590
+    },
+    {
+      "epoch": 0.836060531496063,
+      "grad_norm": 1.2807562351226807,
+      "learning_rate": 1.2987875721180265e-05,
+      "loss": 2.1923,
+      "step": 13591
+    },
+    {
+      "epoch": 0.8361220472440944,
+      "grad_norm": 1.296579122543335,
+      "learning_rate": 1.2978347030659754e-05,
+      "loss": 2.5867,
+      "step": 13592
+    },
+    {
+      "epoch": 0.836183562992126,
+      "grad_norm": 1.3863022327423096,
+      "learning_rate": 1.2968821594263824e-05,
+      "loss": 2.197,
+      "step": 13593
+    },
+    {
+      "epoch": 0.8362450787401575,
+      "grad_norm": 1.1739342212677002,
+      "learning_rate": 1.295929941234868e-05,
+      "loss": 2.0566,
+      "step": 13594
+    },
+    {
+      "epoch": 0.836306594488189,
+      "grad_norm": 1.3653764724731445,
+      "learning_rate": 1.2949780485270369e-05,
+      "loss": 2.5457,
+      "step": 13595
+    },
+    {
+      "epoch": 0.8363681102362205,
+      "grad_norm": 1.2128218412399292,
+      "learning_rate": 1.2940264813384884e-05,
+      "loss": 2.0335,
+      "step": 13596
+    },
+    {
+      "epoch": 0.836429625984252,
+      "grad_norm": 1.3160734176635742,
+      "learning_rate": 1.2930752397048007e-05,
+      "loss": 2.3088,
+      "step": 13597
+    },
+    {
+      "epoch": 0.8364911417322834,
+      "grad_norm": 1.3113471269607544,
+      "learning_rate": 1.292124323661551e-05,
+      "loss": 2.4324,
+      "step": 13598
+    },
+    {
+      "epoch": 0.8365526574803149,
+      "grad_norm": 1.4006191492080688,
+      "learning_rate": 1.2911737332442896e-05,
+      "loss": 2.0817,
+      "step": 13599
+    },
+    {
+      "epoch": 0.8366141732283464,
+      "grad_norm": 1.7055399417877197,
+      "learning_rate": 1.290223468488575e-05,
+      "loss": 1.0918,
+      "step": 13600
+    },
+    {
+      "epoch": 0.836675688976378,
+      "grad_norm": 1.2941954135894775,
+      "learning_rate": 1.289273529429932e-05,
+      "loss": 2.7951,
+      "step": 13601
+    },
+    {
+      "epoch": 0.8367372047244095,
+      "grad_norm": 1.2248085737228394,
+      "learning_rate": 1.2883239161038896e-05,
+      "loss": 2.0911,
+      "step": 13602
+    },
+    {
+      "epoch": 0.836798720472441,
+      "grad_norm": 1.2399998903274536,
+      "learning_rate": 1.2873746285459532e-05,
+      "loss": 2.4971,
+      "step": 13603
+    },
+    {
+      "epoch": 0.8368602362204725,
+      "grad_norm": 1.257677674293518,
+      "learning_rate": 1.2864256667916241e-05,
+      "loss": 2.5597,
+      "step": 13604
+    },
+    {
+      "epoch": 0.8369217519685039,
+      "grad_norm": 1.5782458782196045,
+      "learning_rate": 1.2854770308763842e-05,
+      "loss": 2.4564,
+      "step": 13605
+    },
+    {
+      "epoch": 0.8369832677165354,
+      "grad_norm": 1.1403424739837646,
+      "learning_rate": 1.2845287208357115e-05,
+      "loss": 2.6913,
+      "step": 13606
+    },
+    {
+      "epoch": 0.8370447834645669,
+      "grad_norm": 1.5229240655899048,
+      "learning_rate": 1.2835807367050646e-05,
+      "loss": 2.7011,
+      "step": 13607
+    },
+    {
+      "epoch": 0.8371062992125984,
+      "grad_norm": 1.7133482694625854,
+      "learning_rate": 1.282633078519897e-05,
+      "loss": 2.2325,
+      "step": 13608
+    },
+    {
+      "epoch": 0.83716781496063,
+      "grad_norm": 1.2745038270950317,
+      "learning_rate": 1.2816857463156396e-05,
+      "loss": 3.2719,
+      "step": 13609
+    },
+    {
+      "epoch": 0.8372293307086615,
+      "grad_norm": 1.4832465648651123,
+      "learning_rate": 1.2807387401277237e-05,
+      "loss": 2.1075,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8372908464566929,
+      "grad_norm": 1.5067470073699951,
+      "learning_rate": 1.2797920599915569e-05,
+      "loss": 2.7724,
+      "step": 13611
+    },
+    {
+      "epoch": 0.8373523622047244,
+      "grad_norm": 1.2706706523895264,
+      "learning_rate": 1.2788457059425408e-05,
+      "loss": 2.3362,
+      "step": 13612
+    },
+    {
+      "epoch": 0.8374138779527559,
+      "grad_norm": 1.253536343574524,
+      "learning_rate": 1.2778996780160678e-05,
+      "loss": 2.002,
+      "step": 13613
+    },
+    {
+      "epoch": 0.8374753937007874,
+      "grad_norm": 1.623046875,
+      "learning_rate": 1.2769539762475092e-05,
+      "loss": 1.7583,
+      "step": 13614
+    },
+    {
+      "epoch": 0.8375369094488189,
+      "grad_norm": 1.1740460395812988,
+      "learning_rate": 1.2760086006722328e-05,
+      "loss": 2.7435,
+      "step": 13615
+    },
+    {
+      "epoch": 0.8375984251968503,
+      "grad_norm": 1.611578106880188,
+      "learning_rate": 1.2750635513255859e-05,
+      "loss": 2.1841,
+      "step": 13616
+    },
+    {
+      "epoch": 0.8376599409448819,
+      "grad_norm": 1.2438437938690186,
+      "learning_rate": 1.2741188282429106e-05,
+      "loss": 2.9892,
+      "step": 13617
+    },
+    {
+      "epoch": 0.8377214566929134,
+      "grad_norm": 1.2118769884109497,
+      "learning_rate": 1.273174431459535e-05,
+      "loss": 2.3729,
+      "step": 13618
+    },
+    {
+      "epoch": 0.8377829724409449,
+      "grad_norm": 1.2090352773666382,
+      "learning_rate": 1.2722303610107756e-05,
+      "loss": 2.4983,
+      "step": 13619
+    },
+    {
+      "epoch": 0.8378444881889764,
+      "grad_norm": 1.2116634845733643,
+      "learning_rate": 1.2712866169319305e-05,
+      "loss": 2.401,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8379060039370079,
+      "grad_norm": 1.2096996307373047,
+      "learning_rate": 1.2703431992582948e-05,
+      "loss": 2.4245,
+      "step": 13621
+    },
+    {
+      "epoch": 0.8379675196850394,
+      "grad_norm": 1.2659159898757935,
+      "learning_rate": 1.2694001080251449e-05,
+      "loss": 2.1571,
+      "step": 13622
+    },
+    {
+      "epoch": 0.8380290354330708,
+      "grad_norm": 1.4008598327636719,
+      "learning_rate": 1.2684573432677494e-05,
+      "loss": 2.4757,
+      "step": 13623
+    },
+    {
+      "epoch": 0.8380905511811023,
+      "grad_norm": 1.171976923942566,
+      "learning_rate": 1.2675149050213586e-05,
+      "loss": 2.795,
+      "step": 13624
+    },
+    {
+      "epoch": 0.8381520669291339,
+      "grad_norm": 1.3007534742355347,
+      "learning_rate": 1.2665727933212168e-05,
+      "loss": 2.8178,
+      "step": 13625
+    },
+    {
+      "epoch": 0.8382135826771654,
+      "grad_norm": 1.2835655212402344,
+      "learning_rate": 1.2656310082025536e-05,
+      "loss": 2.5095,
+      "step": 13626
+    },
+    {
+      "epoch": 0.8382750984251969,
+      "grad_norm": 1.2666163444519043,
+      "learning_rate": 1.2646895497005884e-05,
+      "loss": 2.2561,
+      "step": 13627
+    },
+    {
+      "epoch": 0.8383366141732284,
+      "grad_norm": 1.2583529949188232,
+      "learning_rate": 1.2637484178505221e-05,
+      "loss": 2.275,
+      "step": 13628
+    },
+    {
+      "epoch": 0.8383981299212598,
+      "grad_norm": 1.2335114479064941,
+      "learning_rate": 1.2628076126875532e-05,
+      "loss": 2.0196,
+      "step": 13629
+    },
+    {
+      "epoch": 0.8384596456692913,
+      "grad_norm": 1.219804048538208,
+      "learning_rate": 1.2618671342468579e-05,
+      "loss": 2.1938,
+      "step": 13630
+    },
+    {
+      "epoch": 0.8385211614173228,
+      "grad_norm": 1.5214662551879883,
+      "learning_rate": 1.2609269825636061e-05,
+      "loss": 2.2254,
+      "step": 13631
+    },
+    {
+      "epoch": 0.8385826771653543,
+      "grad_norm": 2.306699752807617,
+      "learning_rate": 1.259987157672955e-05,
+      "loss": 3.5153,
+      "step": 13632
+    },
+    {
+      "epoch": 0.8386441929133859,
+      "grad_norm": 1.2488040924072266,
+      "learning_rate": 1.2590476596100497e-05,
+      "loss": 2.7434,
+      "step": 13633
+    },
+    {
+      "epoch": 0.8387057086614174,
+      "grad_norm": 1.3333855867385864,
+      "learning_rate": 1.2581084884100192e-05,
+      "loss": 2.2889,
+      "step": 13634
+    },
+    {
+      "epoch": 0.8387672244094488,
+      "grad_norm": 1.193900227546692,
+      "learning_rate": 1.2571696441079882e-05,
+      "loss": 2.3137,
+      "step": 13635
+    },
+    {
+      "epoch": 0.8388287401574803,
+      "grad_norm": 1.1202424764633179,
+      "learning_rate": 1.256231126739058e-05,
+      "loss": 2.2406,
+      "step": 13636
+    },
+    {
+      "epoch": 0.8388902559055118,
+      "grad_norm": 1.4108457565307617,
+      "learning_rate": 1.2552929363383281e-05,
+      "loss": 2.5282,
+      "step": 13637
+    },
+    {
+      "epoch": 0.8389517716535433,
+      "grad_norm": 1.3150359392166138,
+      "learning_rate": 1.2543550729408793e-05,
+      "loss": 2.6501,
+      "step": 13638
+    },
+    {
+      "epoch": 0.8390132874015748,
+      "grad_norm": 1.2150161266326904,
+      "learning_rate": 1.253417536581787e-05,
+      "loss": 2.414,
+      "step": 13639
+    },
+    {
+      "epoch": 0.8390748031496063,
+      "grad_norm": 1.4027377367019653,
+      "learning_rate": 1.2524803272961038e-05,
+      "loss": 2.5283,
+      "step": 13640
+    },
+    {
+      "epoch": 0.8391363188976378,
+      "grad_norm": 1.0625300407409668,
+      "learning_rate": 1.2515434451188811e-05,
+      "loss": 2.8354,
+      "step": 13641
+    },
+    {
+      "epoch": 0.8391978346456693,
+      "grad_norm": 1.192327618598938,
+      "learning_rate": 1.2506068900851486e-05,
+      "loss": 3.1863,
+      "step": 13642
+    },
+    {
+      "epoch": 0.8392593503937008,
+      "grad_norm": 1.3221269845962524,
+      "learning_rate": 1.2496706622299304e-05,
+      "loss": 2.2988,
+      "step": 13643
+    },
+    {
+      "epoch": 0.8393208661417323,
+      "grad_norm": 1.3040111064910889,
+      "learning_rate": 1.2487347615882372e-05,
+      "loss": 2.2362,
+      "step": 13644
+    },
+    {
+      "epoch": 0.8393823818897638,
+      "grad_norm": 1.4271565675735474,
+      "learning_rate": 1.2477991881950668e-05,
+      "loss": 3.1506,
+      "step": 13645
+    },
+    {
+      "epoch": 0.8394438976377953,
+      "grad_norm": 1.4169963598251343,
+      "learning_rate": 1.2468639420854011e-05,
+      "loss": 3.1101,
+      "step": 13646
+    },
+    {
+      "epoch": 0.8395054133858267,
+      "grad_norm": 1.181119441986084,
+      "learning_rate": 1.2459290232942167e-05,
+      "loss": 2.5236,
+      "step": 13647
+    },
+    {
+      "epoch": 0.8395669291338582,
+      "grad_norm": 1.1437426805496216,
+      "learning_rate": 1.2449944318564732e-05,
+      "loss": 2.675,
+      "step": 13648
+    },
+    {
+      "epoch": 0.8396284448818898,
+      "grad_norm": 1.599632740020752,
+      "learning_rate": 1.2440601678071174e-05,
+      "loss": 2.1791,
+      "step": 13649
+    },
+    {
+      "epoch": 0.8396899606299213,
+      "grad_norm": 1.274694800376892,
+      "learning_rate": 1.2431262311810865e-05,
+      "loss": 2.5552,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8397514763779528,
+      "grad_norm": 1.3688350915908813,
+      "learning_rate": 1.242192622013304e-05,
+      "loss": 2.6007,
+      "step": 13651
+    },
+    {
+      "epoch": 0.8398129921259843,
+      "grad_norm": 1.4171006679534912,
+      "learning_rate": 1.2412593403386852e-05,
+      "loss": 2.1708,
+      "step": 13652
+    },
+    {
+      "epoch": 0.8398745078740157,
+      "grad_norm": 1.1318281888961792,
+      "learning_rate": 1.2403263861921255e-05,
+      "loss": 2.3373,
+      "step": 13653
+    },
+    {
+      "epoch": 0.8399360236220472,
+      "grad_norm": 1.3334121704101562,
+      "learning_rate": 1.2393937596085148e-05,
+      "loss": 2.2395,
+      "step": 13654
+    },
+    {
+      "epoch": 0.8399975393700787,
+      "grad_norm": 1.2424814701080322,
+      "learning_rate": 1.2384614606227252e-05,
+      "loss": 2.526,
+      "step": 13655
+    },
+    {
+      "epoch": 0.8400590551181102,
+      "grad_norm": 1.4639027118682861,
+      "learning_rate": 1.2375294892696221e-05,
+      "loss": 2.5097,
+      "step": 13656
+    },
+    {
+      "epoch": 0.8401205708661418,
+      "grad_norm": 1.1336926221847534,
+      "learning_rate": 1.2365978455840543e-05,
+      "loss": 2.2823,
+      "step": 13657
+    },
+    {
+      "epoch": 0.8401820866141733,
+      "grad_norm": 1.3283576965332031,
+      "learning_rate": 1.2356665296008629e-05,
+      "loss": 1.8328,
+      "step": 13658
+    },
+    {
+      "epoch": 0.8402436023622047,
+      "grad_norm": 1.9180575609207153,
+      "learning_rate": 1.2347355413548712e-05,
+      "loss": 2.3502,
+      "step": 13659
+    },
+    {
+      "epoch": 0.8403051181102362,
+      "grad_norm": 1.3671773672103882,
+      "learning_rate": 1.2338048808808945e-05,
+      "loss": 2.1493,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8403666338582677,
+      "grad_norm": 1.5986398458480835,
+      "learning_rate": 1.2328745482137328e-05,
+      "loss": 2.4561,
+      "step": 13661
+    },
+    {
+      "epoch": 0.8404281496062992,
+      "grad_norm": 1.1187318563461304,
+      "learning_rate": 1.2319445433881749e-05,
+      "loss": 2.6089,
+      "step": 13662
+    },
+    {
+      "epoch": 0.8404896653543307,
+      "grad_norm": 1.5529307126998901,
+      "learning_rate": 1.2310148664389998e-05,
+      "loss": 2.4699,
+      "step": 13663
+    },
+    {
+      "epoch": 0.8405511811023622,
+      "grad_norm": 1.4062273502349854,
+      "learning_rate": 1.2300855174009729e-05,
+      "loss": 2.4293,
+      "step": 13664
+    },
+    {
+      "epoch": 0.8406126968503937,
+      "grad_norm": 1.220240831375122,
+      "learning_rate": 1.2291564963088442e-05,
+      "loss": 2.9951,
+      "step": 13665
+    },
+    {
+      "epoch": 0.8406742125984252,
+      "grad_norm": 1.220771074295044,
+      "learning_rate": 1.228227803197356e-05,
+      "loss": 2.6386,
+      "step": 13666
+    },
+    {
+      "epoch": 0.8407357283464567,
+      "grad_norm": 1.3844984769821167,
+      "learning_rate": 1.2272994381012338e-05,
+      "loss": 2.6573,
+      "step": 13667
+    },
+    {
+      "epoch": 0.8407972440944882,
+      "grad_norm": 1.1545155048370361,
+      "learning_rate": 1.2263714010551975e-05,
+      "loss": 2.0362,
+      "step": 13668
+    },
+    {
+      "epoch": 0.8408587598425197,
+      "grad_norm": 1.5356110334396362,
+      "learning_rate": 1.2254436920939415e-05,
+      "loss": 2.7361,
+      "step": 13669
+    },
+    {
+      "epoch": 0.8409202755905512,
+      "grad_norm": 1.2055127620697021,
+      "learning_rate": 1.2245163112521696e-05,
+      "loss": 2.804,
+      "step": 13670
+    },
+    {
+      "epoch": 0.8409817913385826,
+      "grad_norm": 1.3078187704086304,
+      "learning_rate": 1.2235892585645502e-05,
+      "loss": 2.3982,
+      "step": 13671
+    },
+    {
+      "epoch": 0.8410433070866141,
+      "grad_norm": 1.372135043144226,
+      "learning_rate": 1.2226625340657572e-05,
+      "loss": 2.8969,
+      "step": 13672
+    },
+    {
+      "epoch": 0.8411048228346457,
+      "grad_norm": 1.3359997272491455,
+      "learning_rate": 1.221736137790439e-05,
+      "loss": 2.0635,
+      "step": 13673
+    },
+    {
+      "epoch": 0.8411663385826772,
+      "grad_norm": 1.4095755815505981,
+      "learning_rate": 1.2208100697732427e-05,
+      "loss": 2.6264,
+      "step": 13674
+    },
+    {
+      "epoch": 0.8412278543307087,
+      "grad_norm": 1.2855359315872192,
+      "learning_rate": 1.2198843300487938e-05,
+      "loss": 2.5193,
+      "step": 13675
+    },
+    {
+      "epoch": 0.8412893700787402,
+      "grad_norm": 1.1503627300262451,
+      "learning_rate": 1.2189589186517113e-05,
+      "loss": 2.9144,
+      "step": 13676
+    },
+    {
+      "epoch": 0.8413508858267716,
+      "grad_norm": 1.275216817855835,
+      "learning_rate": 1.218033835616601e-05,
+      "loss": 2.7984,
+      "step": 13677
+    },
+    {
+      "epoch": 0.8414124015748031,
+      "grad_norm": 1.2851852178573608,
+      "learning_rate": 1.2171090809780572e-05,
+      "loss": 2.2125,
+      "step": 13678
+    },
+    {
+      "epoch": 0.8414739173228346,
+      "grad_norm": 1.1762380599975586,
+      "learning_rate": 1.216184654770658e-05,
+      "loss": 2.0905,
+      "step": 13679
+    },
+    {
+      "epoch": 0.8415354330708661,
+      "grad_norm": 1.29166841506958,
+      "learning_rate": 1.2152605570289743e-05,
+      "loss": 2.4709,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8415969488188977,
+      "grad_norm": 1.285739779472351,
+      "learning_rate": 1.2143367877875588e-05,
+      "loss": 2.3638,
+      "step": 13681
+    },
+    {
+      "epoch": 0.8416584645669292,
+      "grad_norm": 1.463472604751587,
+      "learning_rate": 1.2134133470809572e-05,
+      "loss": 2.2857,
+      "step": 13682
+    },
+    {
+      "epoch": 0.8417199803149606,
+      "grad_norm": 1.2882285118103027,
+      "learning_rate": 1.2124902349437007e-05,
+      "loss": 2.8008,
+      "step": 13683
+    },
+    {
+      "epoch": 0.8417814960629921,
+      "grad_norm": 1.101500391960144,
+      "learning_rate": 1.2115674514103114e-05,
+      "loss": 2.6215,
+      "step": 13684
+    },
+    {
+      "epoch": 0.8418430118110236,
+      "grad_norm": 1.0550726652145386,
+      "learning_rate": 1.2106449965152932e-05,
+      "loss": 2.7931,
+      "step": 13685
+    },
+    {
+      "epoch": 0.8419045275590551,
+      "grad_norm": 1.3480501174926758,
+      "learning_rate": 1.2097228702931396e-05,
+      "loss": 2.5838,
+      "step": 13686
+    },
+    {
+      "epoch": 0.8419660433070866,
+      "grad_norm": 1.2139631509780884,
+      "learning_rate": 1.2088010727783372e-05,
+      "loss": 2.2365,
+      "step": 13687
+    },
+    {
+      "epoch": 0.8420275590551181,
+      "grad_norm": 1.2261632680892944,
+      "learning_rate": 1.2078796040053485e-05,
+      "loss": 2.504,
+      "step": 13688
+    },
+    {
+      "epoch": 0.8420890748031497,
+      "grad_norm": 1.133685827255249,
+      "learning_rate": 1.2069584640086418e-05,
+      "loss": 1.9387,
+      "step": 13689
+    },
+    {
+      "epoch": 0.8421505905511811,
+      "grad_norm": 1.2950674295425415,
+      "learning_rate": 1.2060376528226548e-05,
+      "loss": 2.8495,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8422121062992126,
+      "grad_norm": 1.310829520225525,
+      "learning_rate": 1.2051171704818243e-05,
+      "loss": 3.2889,
+      "step": 13691
+    },
+    {
+      "epoch": 0.8422736220472441,
+      "grad_norm": 1.2586102485656738,
+      "learning_rate": 1.2041970170205686e-05,
+      "loss": 2.5678,
+      "step": 13692
+    },
+    {
+      "epoch": 0.8423351377952756,
+      "grad_norm": 1.2848570346832275,
+      "learning_rate": 1.2032771924732999e-05,
+      "loss": 2.1968,
+      "step": 13693
+    },
+    {
+      "epoch": 0.8423966535433071,
+      "grad_norm": 1.2932555675506592,
+      "learning_rate": 1.20235769687441e-05,
+      "loss": 2.6343,
+      "step": 13694
+    },
+    {
+      "epoch": 0.8424581692913385,
+      "grad_norm": 1.2156277894973755,
+      "learning_rate": 1.2014385302582853e-05,
+      "loss": 2.7647,
+      "step": 13695
+    },
+    {
+      "epoch": 0.84251968503937,
+      "grad_norm": 1.4103583097457886,
+      "learning_rate": 1.2005196926592977e-05,
+      "loss": 2.8839,
+      "step": 13696
+    },
+    {
+      "epoch": 0.8425812007874016,
+      "grad_norm": 1.196150302886963,
+      "learning_rate": 1.1996011841118082e-05,
+      "loss": 2.2951,
+      "step": 13697
+    },
+    {
+      "epoch": 0.8426427165354331,
+      "grad_norm": 1.850738763809204,
+      "learning_rate": 1.1986830046501596e-05,
+      "loss": 1.4756,
+      "step": 13698
+    },
+    {
+      "epoch": 0.8427042322834646,
+      "grad_norm": 1.2263292074203491,
+      "learning_rate": 1.197765154308692e-05,
+      "loss": 2.9075,
+      "step": 13699
+    },
+    {
+      "epoch": 0.8427657480314961,
+      "grad_norm": 1.2355576753616333,
+      "learning_rate": 1.1968476331217226e-05,
+      "loss": 2.4531,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8428272637795275,
+      "grad_norm": 1.2170320749282837,
+      "learning_rate": 1.195930441123564e-05,
+      "loss": 2.515,
+      "step": 13701
+    },
+    {
+      "epoch": 0.842888779527559,
+      "grad_norm": 1.273209571838379,
+      "learning_rate": 1.1950135783485139e-05,
+      "loss": 1.9819,
+      "step": 13702
+    },
+    {
+      "epoch": 0.8429502952755905,
+      "grad_norm": 1.2122076749801636,
+      "learning_rate": 1.1940970448308597e-05,
+      "loss": 2.5828,
+      "step": 13703
+    },
+    {
+      "epoch": 0.843011811023622,
+      "grad_norm": 1.2691715955734253,
+      "learning_rate": 1.1931808406048716e-05,
+      "loss": 1.9301,
+      "step": 13704
+    },
+    {
+      "epoch": 0.8430733267716536,
+      "grad_norm": 1.200900673866272,
+      "learning_rate": 1.1922649657048135e-05,
+      "loss": 2.5265,
+      "step": 13705
+    },
+    {
+      "epoch": 0.8431348425196851,
+      "grad_norm": 1.4065887928009033,
+      "learning_rate": 1.1913494201649311e-05,
+      "loss": 1.9757,
+      "step": 13706
+    },
+    {
+      "epoch": 0.8431963582677166,
+      "grad_norm": 1.221243143081665,
+      "learning_rate": 1.1904342040194606e-05,
+      "loss": 2.0726,
+      "step": 13707
+    },
+    {
+      "epoch": 0.843257874015748,
+      "grad_norm": 1.4975045919418335,
+      "learning_rate": 1.1895193173026287e-05,
+      "loss": 2.2026,
+      "step": 13708
+    },
+    {
+      "epoch": 0.8433193897637795,
+      "grad_norm": 1.4645968675613403,
+      "learning_rate": 1.1886047600486472e-05,
+      "loss": 2.2661,
+      "step": 13709
+    },
+    {
+      "epoch": 0.843380905511811,
+      "grad_norm": 1.4314088821411133,
+      "learning_rate": 1.187690532291712e-05,
+      "loss": 2.1367,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8434424212598425,
+      "grad_norm": 1.3296453952789307,
+      "learning_rate": 1.1867766340660147e-05,
+      "loss": 2.2087,
+      "step": 13711
+    },
+    {
+      "epoch": 0.843503937007874,
+      "grad_norm": 1.3155999183654785,
+      "learning_rate": 1.1858630654057245e-05,
+      "loss": 2.3021,
+      "step": 13712
+    },
+    {
+      "epoch": 0.8435654527559056,
+      "grad_norm": 1.2513282299041748,
+      "learning_rate": 1.1849498263450088e-05,
+      "loss": 2.4841,
+      "step": 13713
+    },
+    {
+      "epoch": 0.843626968503937,
+      "grad_norm": 1.2732762098312378,
+      "learning_rate": 1.1840369169180121e-05,
+      "loss": 2.087,
+      "step": 13714
+    },
+    {
+      "epoch": 0.8436884842519685,
+      "grad_norm": 1.0967062711715698,
+      "learning_rate": 1.1831243371588796e-05,
+      "loss": 2.3059,
+      "step": 13715
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 1.286159634590149,
+      "learning_rate": 1.1822120871017295e-05,
+      "loss": 2.3461,
+      "step": 13716
+    },
+    {
+      "epoch": 0.8438115157480315,
+      "grad_norm": 1.283081293106079,
+      "learning_rate": 1.1813001667806811e-05,
+      "loss": 2.5372,
+      "step": 13717
+    },
+    {
+      "epoch": 0.843873031496063,
+      "grad_norm": 1.4488012790679932,
+      "learning_rate": 1.1803885762298305e-05,
+      "loss": 2.5893,
+      "step": 13718
+    },
+    {
+      "epoch": 0.8439345472440944,
+      "grad_norm": 1.1712696552276611,
+      "learning_rate": 1.1794773154832683e-05,
+      "loss": 2.4008,
+      "step": 13719
+    },
+    {
+      "epoch": 0.843996062992126,
+      "grad_norm": 1.22873854637146,
+      "learning_rate": 1.1785663845750683e-05,
+      "loss": 2.5857,
+      "step": 13720
+    },
+    {
+      "epoch": 0.8440575787401575,
+      "grad_norm": 1.3353785276412964,
+      "learning_rate": 1.1776557835392965e-05,
+      "loss": 1.7833,
+      "step": 13721
+    },
+    {
+      "epoch": 0.844119094488189,
+      "grad_norm": 1.4903745651245117,
+      "learning_rate": 1.1767455124100058e-05,
+      "loss": 3.0025,
+      "step": 13722
+    },
+    {
+      "epoch": 0.8441806102362205,
+      "grad_norm": 1.3567651510238647,
+      "learning_rate": 1.17583557122123e-05,
+      "loss": 3.4372,
+      "step": 13723
+    },
+    {
+      "epoch": 0.844242125984252,
+      "grad_norm": 1.1737229824066162,
+      "learning_rate": 1.1749259600070018e-05,
+      "loss": 3.3526,
+      "step": 13724
+    },
+    {
+      "epoch": 0.8443036417322834,
+      "grad_norm": 1.2479194402694702,
+      "learning_rate": 1.1740166788013308e-05,
+      "loss": 2.1965,
+      "step": 13725
+    },
+    {
+      "epoch": 0.8443651574803149,
+      "grad_norm": 1.4185010194778442,
+      "learning_rate": 1.1731077276382196e-05,
+      "loss": 2.6415,
+      "step": 13726
+    },
+    {
+      "epoch": 0.8444266732283464,
+      "grad_norm": 1.4740978479385376,
+      "learning_rate": 1.1721991065516602e-05,
+      "loss": 2.9843,
+      "step": 13727
+    },
+    {
+      "epoch": 0.844488188976378,
+      "grad_norm": 1.0915299654006958,
+      "learning_rate": 1.171290815575632e-05,
+      "loss": 3.1297,
+      "step": 13728
+    },
+    {
+      "epoch": 0.8445497047244095,
+      "grad_norm": 1.135255217552185,
+      "learning_rate": 1.1703828547440942e-05,
+      "loss": 2.4903,
+      "step": 13729
+    },
+    {
+      "epoch": 0.844611220472441,
+      "grad_norm": 1.2454313039779663,
+      "learning_rate": 1.1694752240910045e-05,
+      "loss": 2.3965,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8446727362204725,
+      "grad_norm": 1.2866120338439941,
+      "learning_rate": 1.168567923650299e-05,
+      "loss": 2.7689,
+      "step": 13731
+    },
+    {
+      "epoch": 0.8447342519685039,
+      "grad_norm": 1.4883472919464111,
+      "learning_rate": 1.1676609534559068e-05,
+      "loss": 2.8191,
+      "step": 13732
+    },
+    {
+      "epoch": 0.8447957677165354,
+      "grad_norm": 1.3294907808303833,
+      "learning_rate": 1.1667543135417458e-05,
+      "loss": 1.925,
+      "step": 13733
+    },
+    {
+      "epoch": 0.8448572834645669,
+      "grad_norm": 1.233880877494812,
+      "learning_rate": 1.1658480039417197e-05,
+      "loss": 1.9845,
+      "step": 13734
+    },
+    {
+      "epoch": 0.8449187992125984,
+      "grad_norm": 1.1034820079803467,
+      "learning_rate": 1.1649420246897147e-05,
+      "loss": 1.8297,
+      "step": 13735
+    },
+    {
+      "epoch": 0.84498031496063,
+      "grad_norm": 1.3591946363449097,
+      "learning_rate": 1.1640363758196149e-05,
+      "loss": 2.8625,
+      "step": 13736
+    },
+    {
+      "epoch": 0.8450418307086615,
+      "grad_norm": 1.3164713382720947,
+      "learning_rate": 1.163131057365282e-05,
+      "loss": 2.3689,
+      "step": 13737
+    },
+    {
+      "epoch": 0.8451033464566929,
+      "grad_norm": 1.1923396587371826,
+      "learning_rate": 1.1622260693605747e-05,
+      "loss": 2.3801,
+      "step": 13738
+    },
+    {
+      "epoch": 0.8451648622047244,
+      "grad_norm": 1.3118754625320435,
+      "learning_rate": 1.1613214118393289e-05,
+      "loss": 1.8286,
+      "step": 13739
+    },
+    {
+      "epoch": 0.8452263779527559,
+      "grad_norm": 1.3853856325149536,
+      "learning_rate": 1.1604170848353767e-05,
+      "loss": 2.0973,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8452878937007874,
+      "grad_norm": 1.2646265029907227,
+      "learning_rate": 1.1595130883825345e-05,
+      "loss": 2.6449,
+      "step": 13741
+    },
+    {
+      "epoch": 0.8453494094488189,
+      "grad_norm": 1.3541043996810913,
+      "learning_rate": 1.1586094225146083e-05,
+      "loss": 2.8288,
+      "step": 13742
+    },
+    {
+      "epoch": 0.8454109251968503,
+      "grad_norm": 1.365952730178833,
+      "learning_rate": 1.1577060872653878e-05,
+      "loss": 2.7656,
+      "step": 13743
+    },
+    {
+      "epoch": 0.8454724409448819,
+      "grad_norm": 1.177779197692871,
+      "learning_rate": 1.1568030826686549e-05,
+      "loss": 2.4264,
+      "step": 13744
+    },
+    {
+      "epoch": 0.8455339566929134,
+      "grad_norm": 1.103421926498413,
+      "learning_rate": 1.1559004087581738e-05,
+      "loss": 2.398,
+      "step": 13745
+    },
+    {
+      "epoch": 0.8455954724409449,
+      "grad_norm": 1.2030235528945923,
+      "learning_rate": 1.1549980655677017e-05,
+      "loss": 2.1862,
+      "step": 13746
+    },
+    {
+      "epoch": 0.8456569881889764,
+      "grad_norm": 1.313463807106018,
+      "learning_rate": 1.1540960531309797e-05,
+      "loss": 2.2923,
+      "step": 13747
+    },
+    {
+      "epoch": 0.8457185039370079,
+      "grad_norm": 1.2810721397399902,
+      "learning_rate": 1.1531943714817428e-05,
+      "loss": 2.367,
+      "step": 13748
+    },
+    {
+      "epoch": 0.8457800196850394,
+      "grad_norm": 1.3858108520507812,
+      "learning_rate": 1.1522930206537019e-05,
+      "loss": 2.8288,
+      "step": 13749
+    },
+    {
+      "epoch": 0.8458415354330708,
+      "grad_norm": 1.3079357147216797,
+      "learning_rate": 1.151392000680569e-05,
+      "loss": 2.9217,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8459030511811023,
+      "grad_norm": 1.0928544998168945,
+      "learning_rate": 1.1504913115960314e-05,
+      "loss": 2.2358,
+      "step": 13751
+    },
+    {
+      "epoch": 0.8459645669291339,
+      "grad_norm": 1.2487144470214844,
+      "learning_rate": 1.1495909534337734e-05,
+      "loss": 2.2875,
+      "step": 13752
+    },
+    {
+      "epoch": 0.8460260826771654,
+      "grad_norm": 1.3527944087982178,
+      "learning_rate": 1.1486909262274614e-05,
+      "loss": 2.2407,
+      "step": 13753
+    },
+    {
+      "epoch": 0.8460875984251969,
+      "grad_norm": 1.3500410318374634,
+      "learning_rate": 1.1477912300107551e-05,
+      "loss": 2.3321,
+      "step": 13754
+    },
+    {
+      "epoch": 0.8461491141732284,
+      "grad_norm": 1.3522080183029175,
+      "learning_rate": 1.1468918648172932e-05,
+      "loss": 2.5886,
+      "step": 13755
+    },
+    {
+      "epoch": 0.8462106299212598,
+      "grad_norm": 1.360816240310669,
+      "learning_rate": 1.1459928306807121e-05,
+      "loss": 2.6313,
+      "step": 13756
+    },
+    {
+      "epoch": 0.8462721456692913,
+      "grad_norm": 1.3301945924758911,
+      "learning_rate": 1.1450941276346272e-05,
+      "loss": 2.3229,
+      "step": 13757
+    },
+    {
+      "epoch": 0.8463336614173228,
+      "grad_norm": 1.1984654664993286,
+      "learning_rate": 1.1441957557126415e-05,
+      "loss": 2.3431,
+      "step": 13758
+    },
+    {
+      "epoch": 0.8463951771653543,
+      "grad_norm": 1.5962812900543213,
+      "learning_rate": 1.143297714948357e-05,
+      "loss": 2.2836,
+      "step": 13759
+    },
+    {
+      "epoch": 0.8464566929133859,
+      "grad_norm": 1.2980073690414429,
+      "learning_rate": 1.1424000053753503e-05,
+      "loss": 2.3379,
+      "step": 13760
+    },
+    {
+      "epoch": 0.8465182086614174,
+      "grad_norm": 1.2164725065231323,
+      "learning_rate": 1.1415026270271933e-05,
+      "loss": 3.1252,
+      "step": 13761
+    },
+    {
+      "epoch": 0.8465797244094488,
+      "grad_norm": 1.4226937294006348,
+      "learning_rate": 1.1406055799374404e-05,
+      "loss": 1.9951,
+      "step": 13762
+    },
+    {
+      "epoch": 0.8466412401574803,
+      "grad_norm": 1.3796135187149048,
+      "learning_rate": 1.1397088641396392e-05,
+      "loss": 2.5522,
+      "step": 13763
+    },
+    {
+      "epoch": 0.8467027559055118,
+      "grad_norm": 1.192012906074524,
+      "learning_rate": 1.1388124796673184e-05,
+      "loss": 2.0314,
+      "step": 13764
+    },
+    {
+      "epoch": 0.8467642716535433,
+      "grad_norm": 1.2250401973724365,
+      "learning_rate": 1.1379164265539998e-05,
+      "loss": 2.7146,
+      "step": 13765
+    },
+    {
+      "epoch": 0.8468257874015748,
+      "grad_norm": 1.569508671760559,
+      "learning_rate": 1.1370207048331894e-05,
+      "loss": 2.4442,
+      "step": 13766
+    },
+    {
+      "epoch": 0.8468873031496063,
+      "grad_norm": 1.5268635749816895,
+      "learning_rate": 1.1361253145383865e-05,
+      "loss": 1.7029,
+      "step": 13767
+    },
+    {
+      "epoch": 0.8469488188976378,
+      "grad_norm": 1.3357620239257812,
+      "learning_rate": 1.135230255703068e-05,
+      "loss": 2.5108,
+      "step": 13768
+    },
+    {
+      "epoch": 0.8470103346456693,
+      "grad_norm": 1.4824695587158203,
+      "learning_rate": 1.1343355283607082e-05,
+      "loss": 2.1587,
+      "step": 13769
+    },
+    {
+      "epoch": 0.8470718503937008,
+      "grad_norm": 1.3418958187103271,
+      "learning_rate": 1.1334411325447613e-05,
+      "loss": 2.5667,
+      "step": 13770
+    },
+    {
+      "epoch": 0.8471333661417323,
+      "grad_norm": 1.101348876953125,
+      "learning_rate": 1.1325470682886752e-05,
+      "loss": 2.5505,
+      "step": 13771
+    },
+    {
+      "epoch": 0.8471948818897638,
+      "grad_norm": 1.3826435804367065,
+      "learning_rate": 1.131653335625883e-05,
+      "loss": 2.3666,
+      "step": 13772
+    },
+    {
+      "epoch": 0.8472563976377953,
+      "grad_norm": 1.6115244626998901,
+      "learning_rate": 1.1307599345898056e-05,
+      "loss": 0.9832,
+      "step": 13773
+    },
+    {
+      "epoch": 0.8473179133858267,
+      "grad_norm": 1.341300129890442,
+      "learning_rate": 1.12986686521385e-05,
+      "loss": 2.0041,
+      "step": 13774
+    },
+    {
+      "epoch": 0.8473794291338582,
+      "grad_norm": 1.1289455890655518,
+      "learning_rate": 1.1289741275314126e-05,
+      "loss": 1.7074,
+      "step": 13775
+    },
+    {
+      "epoch": 0.8474409448818898,
+      "grad_norm": 1.2736248970031738,
+      "learning_rate": 1.1280817215758755e-05,
+      "loss": 2.2554,
+      "step": 13776
+    },
+    {
+      "epoch": 0.8475024606299213,
+      "grad_norm": 1.2944586277008057,
+      "learning_rate": 1.127189647380611e-05,
+      "loss": 2.1365,
+      "step": 13777
+    },
+    {
+      "epoch": 0.8475639763779528,
+      "grad_norm": 1.3818540573120117,
+      "learning_rate": 1.1262979049789779e-05,
+      "loss": 2.8171,
+      "step": 13778
+    },
+    {
+      "epoch": 0.8476254921259843,
+      "grad_norm": 1.1405590772628784,
+      "learning_rate": 1.125406494404323e-05,
+      "loss": 2.2144,
+      "step": 13779
+    },
+    {
+      "epoch": 0.8476870078740157,
+      "grad_norm": 1.4534968137741089,
+      "learning_rate": 1.1245154156899773e-05,
+      "loss": 2.1474,
+      "step": 13780
+    },
+    {
+      "epoch": 0.8477485236220472,
+      "grad_norm": 1.4314963817596436,
+      "learning_rate": 1.1236246688692664e-05,
+      "loss": 1.9152,
+      "step": 13781
+    },
+    {
+      "epoch": 0.8478100393700787,
+      "grad_norm": 1.2401763200759888,
+      "learning_rate": 1.1227342539754949e-05,
+      "loss": 2.4979,
+      "step": 13782
+    },
+    {
+      "epoch": 0.8478715551181102,
+      "grad_norm": 1.279443621635437,
+      "learning_rate": 1.1218441710419636e-05,
+      "loss": 2.2776,
+      "step": 13783
+    },
+    {
+      "epoch": 0.8479330708661418,
+      "grad_norm": 1.4419350624084473,
+      "learning_rate": 1.1209544201019496e-05,
+      "loss": 2.712,
+      "step": 13784
+    },
+    {
+      "epoch": 0.8479945866141733,
+      "grad_norm": 1.2495033740997314,
+      "learning_rate": 1.1200650011887338e-05,
+      "loss": 2.6741,
+      "step": 13785
+    },
+    {
+      "epoch": 0.8480561023622047,
+      "grad_norm": 1.2560948133468628,
+      "learning_rate": 1.1191759143355696e-05,
+      "loss": 2.5899,
+      "step": 13786
+    },
+    {
+      "epoch": 0.8481176181102362,
+      "grad_norm": 1.4384446144104004,
+      "learning_rate": 1.118287159575706e-05,
+      "loss": 2.3014,
+      "step": 13787
+    },
+    {
+      "epoch": 0.8481791338582677,
+      "grad_norm": 1.45972740650177,
+      "learning_rate": 1.1173987369423755e-05,
+      "loss": 2.3084,
+      "step": 13788
+    },
+    {
+      "epoch": 0.8482406496062992,
+      "grad_norm": 1.2630457878112793,
+      "learning_rate": 1.1165106464688035e-05,
+      "loss": 2.1069,
+      "step": 13789
+    },
+    {
+      "epoch": 0.8483021653543307,
+      "grad_norm": 1.212841272354126,
+      "learning_rate": 1.1156228881881958e-05,
+      "loss": 2.2832,
+      "step": 13790
+    },
+    {
+      "epoch": 0.8483636811023622,
+      "grad_norm": 1.3289573192596436,
+      "learning_rate": 1.1147354621337514e-05,
+      "loss": 2.1416,
+      "step": 13791
+    },
+    {
+      "epoch": 0.8484251968503937,
+      "grad_norm": 1.2460647821426392,
+      "learning_rate": 1.1138483683386558e-05,
+      "loss": 2.5006,
+      "step": 13792
+    },
+    {
+      "epoch": 0.8484867125984252,
+      "grad_norm": 1.1480956077575684,
+      "learning_rate": 1.1129616068360815e-05,
+      "loss": 2.201,
+      "step": 13793
+    },
+    {
+      "epoch": 0.8485482283464567,
+      "grad_norm": 1.5536139011383057,
+      "learning_rate": 1.1120751776591886e-05,
+      "loss": 2.179,
+      "step": 13794
+    },
+    {
+      "epoch": 0.8486097440944882,
+      "grad_norm": 1.1986815929412842,
+      "learning_rate": 1.1111890808411207e-05,
+      "loss": 2.7594,
+      "step": 13795
+    },
+    {
+      "epoch": 0.8486712598425197,
+      "grad_norm": 1.245856523513794,
+      "learning_rate": 1.1103033164150168e-05,
+      "loss": 2.7476,
+      "step": 13796
+    },
+    {
+      "epoch": 0.8487327755905512,
+      "grad_norm": 1.316697597503662,
+      "learning_rate": 1.109417884413998e-05,
+      "loss": 2.3743,
+      "step": 13797
+    },
+    {
+      "epoch": 0.8487942913385826,
+      "grad_norm": 1.2320882081985474,
+      "learning_rate": 1.108532784871177e-05,
+      "loss": 2.18,
+      "step": 13798
+    },
+    {
+      "epoch": 0.8488558070866141,
+      "grad_norm": 1.1701856851577759,
+      "learning_rate": 1.1076480178196469e-05,
+      "loss": 1.8925,
+      "step": 13799
+    },
+    {
+      "epoch": 0.8489173228346457,
+      "grad_norm": 1.5237599611282349,
+      "learning_rate": 1.1067635832924995e-05,
+      "loss": 2.1411,
+      "step": 13800
+    },
+    {
+      "epoch": 0.8489788385826772,
+      "grad_norm": 1.140046238899231,
+      "learning_rate": 1.1058794813228013e-05,
+      "loss": 2.8941,
+      "step": 13801
+    },
+    {
+      "epoch": 0.8490403543307087,
+      "grad_norm": 1.4699050188064575,
+      "learning_rate": 1.104995711943615e-05,
+      "loss": 2.8122,
+      "step": 13802
+    },
+    {
+      "epoch": 0.8491018700787402,
+      "grad_norm": 1.2289642095565796,
+      "learning_rate": 1.1041122751879895e-05,
+      "loss": 2.2095,
+      "step": 13803
+    },
+    {
+      "epoch": 0.8491633858267716,
+      "grad_norm": 1.3334200382232666,
+      "learning_rate": 1.103229171088962e-05,
+      "loss": 1.9389,
+      "step": 13804
+    },
+    {
+      "epoch": 0.8492249015748031,
+      "grad_norm": 1.4216147661209106,
+      "learning_rate": 1.1023463996795524e-05,
+      "loss": 2.9944,
+      "step": 13805
+    },
+    {
+      "epoch": 0.8492864173228346,
+      "grad_norm": 1.3043752908706665,
+      "learning_rate": 1.1014639609927735e-05,
+      "loss": 2.5714,
+      "step": 13806
+    },
+    {
+      "epoch": 0.8493479330708661,
+      "grad_norm": 1.9065130949020386,
+      "learning_rate": 1.100581855061622e-05,
+      "loss": 2.0897,
+      "step": 13807
+    },
+    {
+      "epoch": 0.8494094488188977,
+      "grad_norm": 1.3134068250656128,
+      "learning_rate": 1.099700081919086e-05,
+      "loss": 2.5948,
+      "step": 13808
+    },
+    {
+      "epoch": 0.8494709645669292,
+      "grad_norm": 1.1680930852890015,
+      "learning_rate": 1.0988186415981361e-05,
+      "loss": 2.12,
+      "step": 13809
+    },
+    {
+      "epoch": 0.8495324803149606,
+      "grad_norm": 1.4130477905273438,
+      "learning_rate": 1.0979375341317344e-05,
+      "loss": 2.5358,
+      "step": 13810
+    },
+    {
+      "epoch": 0.8495939960629921,
+      "grad_norm": 1.3562417030334473,
+      "learning_rate": 1.0970567595528292e-05,
+      "loss": 1.8803,
+      "step": 13811
+    },
+    {
+      "epoch": 0.8496555118110236,
+      "grad_norm": 1.3325459957122803,
+      "learning_rate": 1.0961763178943595e-05,
+      "loss": 2.2996,
+      "step": 13812
+    },
+    {
+      "epoch": 0.8497170275590551,
+      "grad_norm": 1.2994492053985596,
+      "learning_rate": 1.0952962091892449e-05,
+      "loss": 2.1923,
+      "step": 13813
+    },
+    {
+      "epoch": 0.8497785433070866,
+      "grad_norm": 2.0027172565460205,
+      "learning_rate": 1.0944164334703987e-05,
+      "loss": 1.7421,
+      "step": 13814
+    },
+    {
+      "epoch": 0.8498400590551181,
+      "grad_norm": 1.3543521165847778,
+      "learning_rate": 1.0935369907707183e-05,
+      "loss": 2.9135,
+      "step": 13815
+    },
+    {
+      "epoch": 0.8499015748031497,
+      "grad_norm": 1.3042141199111938,
+      "learning_rate": 1.0926578811230904e-05,
+      "loss": 2.9326,
+      "step": 13816
+    },
+    {
+      "epoch": 0.8499630905511811,
+      "grad_norm": 1.2459628582000732,
+      "learning_rate": 1.0917791045603887e-05,
+      "loss": 2.7595,
+      "step": 13817
+    },
+    {
+      "epoch": 0.8500246062992126,
+      "grad_norm": 1.1306394338607788,
+      "learning_rate": 1.0909006611154771e-05,
+      "loss": 2.266,
+      "step": 13818
+    },
+    {
+      "epoch": 0.8500861220472441,
+      "grad_norm": 1.255751609802246,
+      "learning_rate": 1.0900225508211992e-05,
+      "loss": 2.8841,
+      "step": 13819
+    },
+    {
+      "epoch": 0.8501476377952756,
+      "grad_norm": 1.4316967725753784,
+      "learning_rate": 1.0891447737103977e-05,
+      "loss": 2.4893,
+      "step": 13820
+    },
+    {
+      "epoch": 0.8502091535433071,
+      "grad_norm": 1.2047688961029053,
+      "learning_rate": 1.0882673298158919e-05,
+      "loss": 3.1239,
+      "step": 13821
+    },
+    {
+      "epoch": 0.8502706692913385,
+      "grad_norm": 1.3686553239822388,
+      "learning_rate": 1.0873902191704932e-05,
+      "loss": 2.4068,
+      "step": 13822
+    },
+    {
+      "epoch": 0.85033218503937,
+      "grad_norm": 1.7859638929367065,
+      "learning_rate": 1.0865134418070033e-05,
+      "loss": 2.8667,
+      "step": 13823
+    },
+    {
+      "epoch": 0.8503937007874016,
+      "grad_norm": 1.8979356288909912,
+      "learning_rate": 1.0856369977582092e-05,
+      "loss": 2.3022,
+      "step": 13824
+    },
+    {
+      "epoch": 0.8504552165354331,
+      "grad_norm": 1.128695011138916,
+      "learning_rate": 1.0847608870568814e-05,
+      "loss": 2.5898,
+      "step": 13825
+    },
+    {
+      "epoch": 0.8505167322834646,
+      "grad_norm": 1.2987042665481567,
+      "learning_rate": 1.0838851097357872e-05,
+      "loss": 2.3004,
+      "step": 13826
+    },
+    {
+      "epoch": 0.8505782480314961,
+      "grad_norm": 1.2391985654830933,
+      "learning_rate": 1.0830096658276689e-05,
+      "loss": 1.7804,
+      "step": 13827
+    },
+    {
+      "epoch": 0.8506397637795275,
+      "grad_norm": 1.205972671508789,
+      "learning_rate": 1.0821345553652661e-05,
+      "loss": 2.5418,
+      "step": 13828
+    },
+    {
+      "epoch": 0.850701279527559,
+      "grad_norm": 1.2481526136398315,
+      "learning_rate": 1.081259778381305e-05,
+      "loss": 2.1519,
+      "step": 13829
+    },
+    {
+      "epoch": 0.8507627952755905,
+      "grad_norm": 1.4377919435501099,
+      "learning_rate": 1.080385334908497e-05,
+      "loss": 1.655,
+      "step": 13830
+    },
+    {
+      "epoch": 0.850824311023622,
+      "grad_norm": 1.1296794414520264,
+      "learning_rate": 1.0795112249795402e-05,
+      "loss": 1.9603,
+      "step": 13831
+    },
+    {
+      "epoch": 0.8508858267716536,
+      "grad_norm": 1.271087646484375,
+      "learning_rate": 1.07863744862712e-05,
+      "loss": 2.6496,
+      "step": 13832
+    },
+    {
+      "epoch": 0.8509473425196851,
+      "grad_norm": 1.1932538747787476,
+      "learning_rate": 1.0777640058839145e-05,
+      "loss": 2.3875,
+      "step": 13833
+    },
+    {
+      "epoch": 0.8510088582677166,
+      "grad_norm": 1.4675498008728027,
+      "learning_rate": 1.0768908967825808e-05,
+      "loss": 2.0793,
+      "step": 13834
+    },
+    {
+      "epoch": 0.851070374015748,
+      "grad_norm": 1.4625883102416992,
+      "learning_rate": 1.0760181213557718e-05,
+      "loss": 2.3143,
+      "step": 13835
+    },
+    {
+      "epoch": 0.8511318897637795,
+      "grad_norm": 1.8298494815826416,
+      "learning_rate": 1.0751456796361226e-05,
+      "loss": 1.8321,
+      "step": 13836
+    },
+    {
+      "epoch": 0.851193405511811,
+      "grad_norm": 1.471952199935913,
+      "learning_rate": 1.0742735716562613e-05,
+      "loss": 2.3994,
+      "step": 13837
+    },
+    {
+      "epoch": 0.8512549212598425,
+      "grad_norm": 1.3895916938781738,
+      "learning_rate": 1.0734017974487954e-05,
+      "loss": 2.4276,
+      "step": 13838
+    },
+    {
+      "epoch": 0.851316437007874,
+      "grad_norm": 1.3208204507827759,
+      "learning_rate": 1.0725303570463275e-05,
+      "loss": 2.7411,
+      "step": 13839
+    },
+    {
+      "epoch": 0.8513779527559056,
+      "grad_norm": 1.2634611129760742,
+      "learning_rate": 1.0716592504814416e-05,
+      "loss": 2.4884,
+      "step": 13840
+    },
+    {
+      "epoch": 0.851439468503937,
+      "grad_norm": 1.264572024345398,
+      "learning_rate": 1.070788477786714e-05,
+      "loss": 2.5305,
+      "step": 13841
+    },
+    {
+      "epoch": 0.8515009842519685,
+      "grad_norm": 1.247751235961914,
+      "learning_rate": 1.0699180389947072e-05,
+      "loss": 2.6808,
+      "step": 13842
+    },
+    {
+      "epoch": 0.8515625,
+      "grad_norm": 1.2448049783706665,
+      "learning_rate": 1.069047934137971e-05,
+      "loss": 2.6472,
+      "step": 13843
+    },
+    {
+      "epoch": 0.8516240157480315,
+      "grad_norm": 1.368669033050537,
+      "learning_rate": 1.0681781632490406e-05,
+      "loss": 2.4658,
+      "step": 13844
+    },
+    {
+      "epoch": 0.851685531496063,
+      "grad_norm": 1.3527156114578247,
+      "learning_rate": 1.0673087263604441e-05,
+      "loss": 2.3541,
+      "step": 13845
+    },
+    {
+      "epoch": 0.8517470472440944,
+      "grad_norm": 1.5523855686187744,
+      "learning_rate": 1.066439623504688e-05,
+      "loss": 2.4836,
+      "step": 13846
+    },
+    {
+      "epoch": 0.851808562992126,
+      "grad_norm": 1.34539794921875,
+      "learning_rate": 1.0655708547142762e-05,
+      "loss": 2.258,
+      "step": 13847
+    },
+    {
+      "epoch": 0.8518700787401575,
+      "grad_norm": 1.405846118927002,
+      "learning_rate": 1.0647024200216937e-05,
+      "loss": 2.4388,
+      "step": 13848
+    },
+    {
+      "epoch": 0.851931594488189,
+      "grad_norm": 1.6408737897872925,
+      "learning_rate": 1.0638343194594192e-05,
+      "loss": 1.6421,
+      "step": 13849
+    },
+    {
+      "epoch": 0.8519931102362205,
+      "grad_norm": 1.3058422803878784,
+      "learning_rate": 1.0629665530599087e-05,
+      "loss": 2.6707,
+      "step": 13850
+    },
+    {
+      "epoch": 0.852054625984252,
+      "grad_norm": 1.1245269775390625,
+      "learning_rate": 1.0620991208556175e-05,
+      "loss": 2.6791,
+      "step": 13851
+    },
+    {
+      "epoch": 0.8521161417322834,
+      "grad_norm": 1.2310056686401367,
+      "learning_rate": 1.0612320228789774e-05,
+      "loss": 2.7816,
+      "step": 13852
+    },
+    {
+      "epoch": 0.8521776574803149,
+      "grad_norm": 1.2487196922302246,
+      "learning_rate": 1.060365259162418e-05,
+      "loss": 2.6616,
+      "step": 13853
+    },
+    {
+      "epoch": 0.8522391732283464,
+      "grad_norm": 1.3431615829467773,
+      "learning_rate": 1.059498829738348e-05,
+      "loss": 2.6552,
+      "step": 13854
+    },
+    {
+      "epoch": 0.852300688976378,
+      "grad_norm": 1.5759514570236206,
+      "learning_rate": 1.0586327346391667e-05,
+      "loss": 2.4878,
+      "step": 13855
+    },
+    {
+      "epoch": 0.8523622047244095,
+      "grad_norm": 1.2880116701126099,
+      "learning_rate": 1.057766973897264e-05,
+      "loss": 2.3123,
+      "step": 13856
+    },
+    {
+      "epoch": 0.852423720472441,
+      "grad_norm": 1.4064669609069824,
+      "learning_rate": 1.056901547545014e-05,
+      "loss": 2.2426,
+      "step": 13857
+    },
+    {
+      "epoch": 0.8524852362204725,
+      "grad_norm": 1.4035742282867432,
+      "learning_rate": 1.0560364556147772e-05,
+      "loss": 2.4552,
+      "step": 13858
+    },
+    {
+      "epoch": 0.8525467519685039,
+      "grad_norm": 1.2103263139724731,
+      "learning_rate": 1.0551716981389049e-05,
+      "loss": 2.0122,
+      "step": 13859
+    },
+    {
+      "epoch": 0.8526082677165354,
+      "grad_norm": 1.358180046081543,
+      "learning_rate": 1.054307275149733e-05,
+      "loss": 2.4096,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8526697834645669,
+      "grad_norm": 2.044321060180664,
+      "learning_rate": 1.0534431866795846e-05,
+      "loss": 2.5865,
+      "step": 13861
+    },
+    {
+      "epoch": 0.8527312992125984,
+      "grad_norm": 1.4150965213775635,
+      "learning_rate": 1.0525794327607752e-05,
+      "loss": 2.3558,
+      "step": 13862
+    },
+    {
+      "epoch": 0.85279281496063,
+      "grad_norm": 1.2725275754928589,
+      "learning_rate": 1.0517160134256032e-05,
+      "loss": 2.5857,
+      "step": 13863
+    },
+    {
+      "epoch": 0.8528543307086615,
+      "grad_norm": 1.4701579809188843,
+      "learning_rate": 1.050852928706355e-05,
+      "loss": 2.9768,
+      "step": 13864
+    },
+    {
+      "epoch": 0.8529158464566929,
+      "grad_norm": 1.253490924835205,
+      "learning_rate": 1.049990178635306e-05,
+      "loss": 2.3518,
+      "step": 13865
+    },
+    {
+      "epoch": 0.8529773622047244,
+      "grad_norm": 1.2281146049499512,
+      "learning_rate": 1.0491277632447172e-05,
+      "loss": 2.3957,
+      "step": 13866
+    },
+    {
+      "epoch": 0.8530388779527559,
+      "grad_norm": 1.1100058555603027,
+      "learning_rate": 1.048265682566838e-05,
+      "loss": 2.8626,
+      "step": 13867
+    },
+    {
+      "epoch": 0.8531003937007874,
+      "grad_norm": 1.3342180252075195,
+      "learning_rate": 1.0474039366339073e-05,
+      "loss": 2.2422,
+      "step": 13868
+    },
+    {
+      "epoch": 0.8531619094488189,
+      "grad_norm": 1.3251824378967285,
+      "learning_rate": 1.046542525478147e-05,
+      "loss": 2.1548,
+      "step": 13869
+    },
+    {
+      "epoch": 0.8532234251968503,
+      "grad_norm": 1.353447675704956,
+      "learning_rate": 1.0456814491317723e-05,
+      "loss": 2.2326,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8532849409448819,
+      "grad_norm": 1.3146060705184937,
+      "learning_rate": 1.044820707626979e-05,
+      "loss": 2.3906,
+      "step": 13871
+    },
+    {
+      "epoch": 0.8533464566929134,
+      "grad_norm": 1.3376247882843018,
+      "learning_rate": 1.0439603009959575e-05,
+      "loss": 2.668,
+      "step": 13872
+    },
+    {
+      "epoch": 0.8534079724409449,
+      "grad_norm": 1.1654306650161743,
+      "learning_rate": 1.0431002292708758e-05,
+      "loss": 2.8179,
+      "step": 13873
+    },
+    {
+      "epoch": 0.8534694881889764,
+      "grad_norm": 1.4011186361312866,
+      "learning_rate": 1.0422404924839057e-05,
+      "loss": 2.3231,
+      "step": 13874
+    },
+    {
+      "epoch": 0.8535310039370079,
+      "grad_norm": 2.47766375541687,
+      "learning_rate": 1.0413810906671884e-05,
+      "loss": 1.4357,
+      "step": 13875
+    },
+    {
+      "epoch": 0.8535925196850394,
+      "grad_norm": 1.4660532474517822,
+      "learning_rate": 1.0405220238528657e-05,
+      "loss": 2.9652,
+      "step": 13876
+    },
+    {
+      "epoch": 0.8536540354330708,
+      "grad_norm": 1.4981998205184937,
+      "learning_rate": 1.0396632920730588e-05,
+      "loss": 2.5167,
+      "step": 13877
+    },
+    {
+      "epoch": 0.8537155511811023,
+      "grad_norm": 1.1988065242767334,
+      "learning_rate": 1.0388048953598817e-05,
+      "loss": 2.4841,
+      "step": 13878
+    },
+    {
+      "epoch": 0.8537770669291339,
+      "grad_norm": 1.2002959251403809,
+      "learning_rate": 1.0379468337454313e-05,
+      "loss": 2.4675,
+      "step": 13879
+    },
+    {
+      "epoch": 0.8538385826771654,
+      "grad_norm": 1.3115462064743042,
+      "learning_rate": 1.037089107261795e-05,
+      "loss": 2.5768,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8539000984251969,
+      "grad_norm": 1.2006709575653076,
+      "learning_rate": 1.0362317159410473e-05,
+      "loss": 3.0301,
+      "step": 13881
+    },
+    {
+      "epoch": 0.8539616141732284,
+      "grad_norm": 1.1721590757369995,
+      "learning_rate": 1.0353746598152526e-05,
+      "loss": 2.3249,
+      "step": 13882
+    },
+    {
+      "epoch": 0.8540231299212598,
+      "grad_norm": 1.1839641332626343,
+      "learning_rate": 1.0345179389164561e-05,
+      "loss": 1.6463,
+      "step": 13883
+    },
+    {
+      "epoch": 0.8540846456692913,
+      "grad_norm": 1.328313946723938,
+      "learning_rate": 1.0336615532766969e-05,
+      "loss": 2.7287,
+      "step": 13884
+    },
+    {
+      "epoch": 0.8541461614173228,
+      "grad_norm": 1.1122115850448608,
+      "learning_rate": 1.032805502927996e-05,
+      "loss": 2.8104,
+      "step": 13885
+    },
+    {
+      "epoch": 0.8542076771653543,
+      "grad_norm": 1.2297695875167847,
+      "learning_rate": 1.0319497879023666e-05,
+      "loss": 2.3602,
+      "step": 13886
+    },
+    {
+      "epoch": 0.8542691929133859,
+      "grad_norm": 1.2426873445510864,
+      "learning_rate": 1.0310944082318086e-05,
+      "loss": 1.6959,
+      "step": 13887
+    },
+    {
+      "epoch": 0.8543307086614174,
+      "grad_norm": 1.4008464813232422,
+      "learning_rate": 1.0302393639483088e-05,
+      "loss": 2.2136,
+      "step": 13888
+    },
+    {
+      "epoch": 0.8543922244094488,
+      "grad_norm": 1.343506932258606,
+      "learning_rate": 1.0293846550838382e-05,
+      "loss": 2.8087,
+      "step": 13889
+    },
+    {
+      "epoch": 0.8544537401574803,
+      "grad_norm": 1.20205557346344,
+      "learning_rate": 1.0285302816703623e-05,
+      "loss": 2.938,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8545152559055118,
+      "grad_norm": 1.3004475831985474,
+      "learning_rate": 1.0276762437398257e-05,
+      "loss": 2.1246,
+      "step": 13891
+    },
+    {
+      "epoch": 0.8545767716535433,
+      "grad_norm": 1.3312363624572754,
+      "learning_rate": 1.0268225413241672e-05,
+      "loss": 2.2372,
+      "step": 13892
+    },
+    {
+      "epoch": 0.8546382874015748,
+      "grad_norm": 1.14280104637146,
+      "learning_rate": 1.025969174455309e-05,
+      "loss": 2.5677,
+      "step": 13893
+    },
+    {
+      "epoch": 0.8546998031496063,
+      "grad_norm": 1.1982172727584839,
+      "learning_rate": 1.0251161431651645e-05,
+      "loss": 2.0206,
+      "step": 13894
+    },
+    {
+      "epoch": 0.8547613188976378,
+      "grad_norm": 1.324776530265808,
+      "learning_rate": 1.0242634474856305e-05,
+      "loss": 2.01,
+      "step": 13895
+    },
+    {
+      "epoch": 0.8548228346456693,
+      "grad_norm": 1.1179285049438477,
+      "learning_rate": 1.0234110874485947e-05,
+      "loss": 2.3062,
+      "step": 13896
+    },
+    {
+      "epoch": 0.8548843503937008,
+      "grad_norm": 1.2538461685180664,
+      "learning_rate": 1.0225590630859272e-05,
+      "loss": 2.148,
+      "step": 13897
+    },
+    {
+      "epoch": 0.8549458661417323,
+      "grad_norm": 1.3949424028396606,
+      "learning_rate": 1.0217073744294925e-05,
+      "loss": 2.4264,
+      "step": 13898
+    },
+    {
+      "epoch": 0.8550073818897638,
+      "grad_norm": 1.3263459205627441,
+      "learning_rate": 1.0208560215111352e-05,
+      "loss": 2.5141,
+      "step": 13899
+    },
+    {
+      "epoch": 0.8550688976377953,
+      "grad_norm": 1.3121215105056763,
+      "learning_rate": 1.0200050043626963e-05,
+      "loss": 2.4723,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8551304133858267,
+      "grad_norm": 1.34665846824646,
+      "learning_rate": 1.019154323015995e-05,
+      "loss": 2.7266,
+      "step": 13901
+    },
+    {
+      "epoch": 0.8551919291338582,
+      "grad_norm": 1.1040046215057373,
+      "learning_rate": 1.0183039775028457e-05,
+      "loss": 2.6212,
+      "step": 13902
+    },
+    {
+      "epoch": 0.8552534448818898,
+      "grad_norm": 1.2612262964248657,
+      "learning_rate": 1.017453967855042e-05,
+      "loss": 2.1825,
+      "step": 13903
+    },
+    {
+      "epoch": 0.8553149606299213,
+      "grad_norm": 1.2827825546264648,
+      "learning_rate": 1.016604294104374e-05,
+      "loss": 2.7995,
+      "step": 13904
+    },
+    {
+      "epoch": 0.8553764763779528,
+      "grad_norm": 1.4743224382400513,
+      "learning_rate": 1.0157549562826107e-05,
+      "loss": 1.9554,
+      "step": 13905
+    },
+    {
+      "epoch": 0.8554379921259843,
+      "grad_norm": 1.1002705097198486,
+      "learning_rate": 1.0149059544215145e-05,
+      "loss": 2.7336,
+      "step": 13906
+    },
+    {
+      "epoch": 0.8554995078740157,
+      "grad_norm": 1.6048365831375122,
+      "learning_rate": 1.0140572885528342e-05,
+      "loss": 3.2818,
+      "step": 13907
+    },
+    {
+      "epoch": 0.8555610236220472,
+      "grad_norm": 1.3405523300170898,
+      "learning_rate": 1.0132089587083038e-05,
+      "loss": 2.1503,
+      "step": 13908
+    },
+    {
+      "epoch": 0.8556225393700787,
+      "grad_norm": 1.1941040754318237,
+      "learning_rate": 1.0123609649196487e-05,
+      "loss": 2.884,
+      "step": 13909
+    },
+    {
+      "epoch": 0.8556840551181102,
+      "grad_norm": 1.36247980594635,
+      "learning_rate": 1.0115133072185745e-05,
+      "loss": 2.408,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8557455708661418,
+      "grad_norm": 1.17240309715271,
+      "learning_rate": 1.0106659856367806e-05,
+      "loss": 2.3354,
+      "step": 13911
+    },
+    {
+      "epoch": 0.8558070866141733,
+      "grad_norm": 1.0114010572433472,
+      "learning_rate": 1.0098190002059537e-05,
+      "loss": 2.9453,
+      "step": 13912
+    },
+    {
+      "epoch": 0.8558686023622047,
+      "grad_norm": 1.3542909622192383,
+      "learning_rate": 1.0089723509577676e-05,
+      "loss": 2.4231,
+      "step": 13913
+    },
+    {
+      "epoch": 0.8559301181102362,
+      "grad_norm": 1.098615288734436,
+      "learning_rate": 1.0081260379238789e-05,
+      "loss": 2.6802,
+      "step": 13914
+    },
+    {
+      "epoch": 0.8559916338582677,
+      "grad_norm": 1.2057031393051147,
+      "learning_rate": 1.0072800611359378e-05,
+      "loss": 2.4823,
+      "step": 13915
+    },
+    {
+      "epoch": 0.8560531496062992,
+      "grad_norm": 1.2575291395187378,
+      "learning_rate": 1.006434420625575e-05,
+      "loss": 2.0601,
+      "step": 13916
+    },
+    {
+      "epoch": 0.8561146653543307,
+      "grad_norm": 1.3085041046142578,
+      "learning_rate": 1.0055891164244158e-05,
+      "loss": 2.2982,
+      "step": 13917
+    },
+    {
+      "epoch": 0.8561761811023622,
+      "grad_norm": 1.1648962497711182,
+      "learning_rate": 1.0047441485640696e-05,
+      "loss": 2.5674,
+      "step": 13918
+    },
+    {
+      "epoch": 0.8562376968503937,
+      "grad_norm": 1.296623706817627,
+      "learning_rate": 1.0038995170761356e-05,
+      "loss": 2.5036,
+      "step": 13919
+    },
+    {
+      "epoch": 0.8562992125984252,
+      "grad_norm": 1.4672554731369019,
+      "learning_rate": 1.0030552219921929e-05,
+      "loss": 2.4256,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8563607283464567,
+      "grad_norm": 1.3729712963104248,
+      "learning_rate": 1.0022112633438197e-05,
+      "loss": 2.2113,
+      "step": 13921
+    },
+    {
+      "epoch": 0.8564222440944882,
+      "grad_norm": 1.2296574115753174,
+      "learning_rate": 1.0013676411625694e-05,
+      "loss": 2.6955,
+      "step": 13922
+    },
+    {
+      "epoch": 0.8564837598425197,
+      "grad_norm": 1.2238317728042603,
+      "learning_rate": 1.0005243554799948e-05,
+      "loss": 2.9709,
+      "step": 13923
+    },
+    {
+      "epoch": 0.8565452755905512,
+      "grad_norm": 1.318827509880066,
+      "learning_rate": 9.996814063276238e-06,
+      "loss": 2.5271,
+      "step": 13924
+    },
+    {
+      "epoch": 0.8566067913385826,
+      "grad_norm": 1.3629155158996582,
+      "learning_rate": 9.988387937369814e-06,
+      "loss": 2.6424,
+      "step": 13925
+    },
+    {
+      "epoch": 0.8566683070866141,
+      "grad_norm": 1.2358825206756592,
+      "learning_rate": 9.979965177395755e-06,
+      "loss": 2.0617,
+      "step": 13926
+    },
+    {
+      "epoch": 0.8567298228346457,
+      "grad_norm": 1.2557175159454346,
+      "learning_rate": 9.971545783669068e-06,
+      "loss": 2.3237,
+      "step": 13927
+    },
+    {
+      "epoch": 0.8567913385826772,
+      "grad_norm": 1.3172041177749634,
+      "learning_rate": 9.96312975650452e-06,
+      "loss": 2.0213,
+      "step": 13928
+    },
+    {
+      "epoch": 0.8568528543307087,
+      "grad_norm": 1.4669172763824463,
+      "learning_rate": 9.954717096216893e-06,
+      "loss": 2.4482,
+      "step": 13929
+    },
+    {
+      "epoch": 0.8569143700787402,
+      "grad_norm": 1.1009221076965332,
+      "learning_rate": 9.946307803120714e-06,
+      "loss": 2.5571,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8569758858267716,
+      "grad_norm": 1.2708520889282227,
+      "learning_rate": 9.937901877530465e-06,
+      "loss": 2.3068,
+      "step": 13931
+    },
+    {
+      "epoch": 0.8570374015748031,
+      "grad_norm": 1.2283766269683838,
+      "learning_rate": 9.929499319760493e-06,
+      "loss": 2.3598,
+      "step": 13932
+    },
+    {
+      "epoch": 0.8570989173228346,
+      "grad_norm": 1.4754126071929932,
+      "learning_rate": 9.921100130125017e-06,
+      "loss": 1.5623,
+      "step": 13933
+    },
+    {
+      "epoch": 0.8571604330708661,
+      "grad_norm": 1.396822452545166,
+      "learning_rate": 9.912704308938091e-06,
+      "loss": 2.7477,
+      "step": 13934
+    },
+    {
+      "epoch": 0.8572219488188977,
+      "grad_norm": 1.1162456274032593,
+      "learning_rate": 9.904311856513693e-06,
+      "loss": 2.2636,
+      "step": 13935
+    },
+    {
+      "epoch": 0.8572834645669292,
+      "grad_norm": 1.5473335981369019,
+      "learning_rate": 9.895922773165623e-06,
+      "loss": 2.3299,
+      "step": 13936
+    },
+    {
+      "epoch": 0.8573449803149606,
+      "grad_norm": 1.4095804691314697,
+      "learning_rate": 9.88753705920762e-06,
+      "loss": 3.165,
+      "step": 13937
+    },
+    {
+      "epoch": 0.8574064960629921,
+      "grad_norm": 1.6549206972122192,
+      "learning_rate": 9.879154714953232e-06,
+      "loss": 2.2536,
+      "step": 13938
+    },
+    {
+      "epoch": 0.8574680118110236,
+      "grad_norm": 1.2792991399765015,
+      "learning_rate": 9.870775740715966e-06,
+      "loss": 3.0541,
+      "step": 13939
+    },
+    {
+      "epoch": 0.8575295275590551,
+      "grad_norm": 1.390647530555725,
+      "learning_rate": 9.86240013680908e-06,
+      "loss": 2.6834,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8575910433070866,
+      "grad_norm": 1.2216488122940063,
+      "learning_rate": 9.854027903545837e-06,
+      "loss": 2.9968,
+      "step": 13941
+    },
+    {
+      "epoch": 0.8576525590551181,
+      "grad_norm": 1.112931489944458,
+      "learning_rate": 9.845659041239275e-06,
+      "loss": 2.0257,
+      "step": 13942
+    },
+    {
+      "epoch": 0.8577140748031497,
+      "grad_norm": 1.45236337184906,
+      "learning_rate": 9.83729355020232e-06,
+      "loss": 2.232,
+      "step": 13943
+    },
+    {
+      "epoch": 0.8577755905511811,
+      "grad_norm": 1.4286140203475952,
+      "learning_rate": 9.828931430747856e-06,
+      "loss": 2.2674,
+      "step": 13944
+    },
+    {
+      "epoch": 0.8578371062992126,
+      "grad_norm": 1.163808822631836,
+      "learning_rate": 9.820572683188534e-06,
+      "loss": 2.5538,
+      "step": 13945
+    },
+    {
+      "epoch": 0.8578986220472441,
+      "grad_norm": 1.3510642051696777,
+      "learning_rate": 9.812217307836958e-06,
+      "loss": 2.0628,
+      "step": 13946
+    },
+    {
+      "epoch": 0.8579601377952756,
+      "grad_norm": 1.2939059734344482,
+      "learning_rate": 9.803865305005545e-06,
+      "loss": 2.918,
+      "step": 13947
+    },
+    {
+      "epoch": 0.8580216535433071,
+      "grad_norm": 1.3396854400634766,
+      "learning_rate": 9.795516675006622e-06,
+      "loss": 2.7998,
+      "step": 13948
+    },
+    {
+      "epoch": 0.8580831692913385,
+      "grad_norm": 1.3699856996536255,
+      "learning_rate": 9.787171418152374e-06,
+      "loss": 2.2199,
+      "step": 13949
+    },
+    {
+      "epoch": 0.85814468503937,
+      "grad_norm": 1.2852540016174316,
+      "learning_rate": 9.778829534754874e-06,
+      "loss": 2.5041,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8582062007874016,
+      "grad_norm": 1.3732192516326904,
+      "learning_rate": 9.770491025126061e-06,
+      "loss": 2.7373,
+      "step": 13951
+    },
+    {
+      "epoch": 0.8582677165354331,
+      "grad_norm": 1.360052227973938,
+      "learning_rate": 9.762155889577762e-06,
+      "loss": 2.4937,
+      "step": 13952
+    },
+    {
+      "epoch": 0.8583292322834646,
+      "grad_norm": 1.7133138179779053,
+      "learning_rate": 9.75382412842164e-06,
+      "loss": 2.2986,
+      "step": 13953
+    },
+    {
+      "epoch": 0.8583907480314961,
+      "grad_norm": 1.1932847499847412,
+      "learning_rate": 9.745495741969291e-06,
+      "loss": 2.9085,
+      "step": 13954
+    },
+    {
+      "epoch": 0.8584522637795275,
+      "grad_norm": 1.2273894548416138,
+      "learning_rate": 9.737170730532109e-06,
+      "loss": 2.7713,
+      "step": 13955
+    },
+    {
+      "epoch": 0.858513779527559,
+      "grad_norm": 1.328170895576477,
+      "learning_rate": 9.728849094421422e-06,
+      "loss": 2.3048,
+      "step": 13956
+    },
+    {
+      "epoch": 0.8585752952755905,
+      "grad_norm": 1.3576860427856445,
+      "learning_rate": 9.720530833948404e-06,
+      "loss": 1.9811,
+      "step": 13957
+    },
+    {
+      "epoch": 0.858636811023622,
+      "grad_norm": 1.3222525119781494,
+      "learning_rate": 9.712215949424153e-06,
+      "loss": 2.5582,
+      "step": 13958
+    },
+    {
+      "epoch": 0.8586983267716536,
+      "grad_norm": 1.3740819692611694,
+      "learning_rate": 9.70390444115954e-06,
+      "loss": 2.283,
+      "step": 13959
+    },
+    {
+      "epoch": 0.8587598425196851,
+      "grad_norm": 1.1071152687072754,
+      "learning_rate": 9.695596309465415e-06,
+      "loss": 2.3327,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8588213582677166,
+      "grad_norm": 1.2647974491119385,
+      "learning_rate": 9.687291554652422e-06,
+      "loss": 2.2667,
+      "step": 13961
+    },
+    {
+      "epoch": 0.858882874015748,
+      "grad_norm": 1.5621942281723022,
+      "learning_rate": 9.678990177031122e-06,
+      "loss": 2.7463,
+      "step": 13962
+    },
+    {
+      "epoch": 0.8589443897637795,
+      "grad_norm": 1.2865906953811646,
+      "learning_rate": 9.670692176911955e-06,
+      "loss": 1.7237,
+      "step": 13963
+    },
+    {
+      "epoch": 0.859005905511811,
+      "grad_norm": 1.27509343624115,
+      "learning_rate": 9.662397554605218e-06,
+      "loss": 2.0648,
+      "step": 13964
+    },
+    {
+      "epoch": 0.8590674212598425,
+      "grad_norm": 1.1291416883468628,
+      "learning_rate": 9.654106310421062e-06,
+      "loss": 2.3843,
+      "step": 13965
+    },
+    {
+      "epoch": 0.859128937007874,
+      "grad_norm": 1.3033523559570312,
+      "learning_rate": 9.645818444669564e-06,
+      "loss": 3.0677,
+      "step": 13966
+    },
+    {
+      "epoch": 0.8591904527559056,
+      "grad_norm": 1.1324145793914795,
+      "learning_rate": 9.637533957660616e-06,
+      "loss": 2.6259,
+      "step": 13967
+    },
+    {
+      "epoch": 0.859251968503937,
+      "grad_norm": 1.3530502319335938,
+      "learning_rate": 9.62925284970404e-06,
+      "loss": 1.6421,
+      "step": 13968
+    },
+    {
+      "epoch": 0.8593134842519685,
+      "grad_norm": 1.4172512292861938,
+      "learning_rate": 9.620975121109443e-06,
+      "loss": 2.4892,
+      "step": 13969
+    },
+    {
+      "epoch": 0.859375,
+      "grad_norm": 1.797371745109558,
+      "learning_rate": 9.612700772186445e-06,
+      "loss": 1.9546,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8594365157480315,
+      "grad_norm": 1.0898594856262207,
+      "learning_rate": 9.604429803244409e-06,
+      "loss": 2.8264,
+      "step": 13971
+    },
+    {
+      "epoch": 0.859498031496063,
+      "grad_norm": 1.3102433681488037,
+      "learning_rate": 9.596162214592664e-06,
+      "loss": 2.3057,
+      "step": 13972
+    },
+    {
+      "epoch": 0.8595595472440944,
+      "grad_norm": 1.1767890453338623,
+      "learning_rate": 9.587898006540319e-06,
+      "loss": 2.9744,
+      "step": 13973
+    },
+    {
+      "epoch": 0.859621062992126,
+      "grad_norm": 1.3254629373550415,
+      "learning_rate": 9.57963717939646e-06,
+      "loss": 2.811,
+      "step": 13974
+    },
+    {
+      "epoch": 0.8596825787401575,
+      "grad_norm": 1.7924271821975708,
+      "learning_rate": 9.571379733469955e-06,
+      "loss": 2.1786,
+      "step": 13975
+    },
+    {
+      "epoch": 0.859744094488189,
+      "grad_norm": 1.3303780555725098,
+      "learning_rate": 9.563125669069606e-06,
+      "loss": 2.8779,
+      "step": 13976
+    },
+    {
+      "epoch": 0.8598056102362205,
+      "grad_norm": 1.0737122297286987,
+      "learning_rate": 9.554874986504059e-06,
+      "loss": 2.6044,
+      "step": 13977
+    },
+    {
+      "epoch": 0.859867125984252,
+      "grad_norm": 1.2619965076446533,
+      "learning_rate": 9.546627686081878e-06,
+      "loss": 2.8935,
+      "step": 13978
+    },
+    {
+      "epoch": 0.8599286417322834,
+      "grad_norm": 1.164231300354004,
+      "learning_rate": 9.538383768111437e-06,
+      "loss": 2.9231,
+      "step": 13979
+    },
+    {
+      "epoch": 0.8599901574803149,
+      "grad_norm": 1.4015257358551025,
+      "learning_rate": 9.530143232901e-06,
+      "loss": 2.6795,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8600516732283464,
+      "grad_norm": 1.3529123067855835,
+      "learning_rate": 9.521906080758724e-06,
+      "loss": 2.303,
+      "step": 13981
+    },
+    {
+      "epoch": 0.860113188976378,
+      "grad_norm": 1.2805054187774658,
+      "learning_rate": 9.51367231199265e-06,
+      "loss": 2.3363,
+      "step": 13982
+    },
+    {
+      "epoch": 0.8601747047244095,
+      "grad_norm": 1.316666841506958,
+      "learning_rate": 9.505441926910674e-06,
+      "loss": 2.5728,
+      "step": 13983
+    },
+    {
+      "epoch": 0.860236220472441,
+      "grad_norm": 1.2471046447753906,
+      "learning_rate": 9.497214925820551e-06,
+      "loss": 2.7773,
+      "step": 13984
+    },
+    {
+      "epoch": 0.8602977362204725,
+      "grad_norm": 1.1836612224578857,
+      "learning_rate": 9.48899130902995e-06,
+      "loss": 2.6299,
+      "step": 13985
+    },
+    {
+      "epoch": 0.8603592519685039,
+      "grad_norm": 1.2841635942459106,
+      "learning_rate": 9.48077107684634e-06,
+      "loss": 2.4967,
+      "step": 13986
+    },
+    {
+      "epoch": 0.8604207677165354,
+      "grad_norm": 1.217665672302246,
+      "learning_rate": 9.472554229577168e-06,
+      "loss": 2.5932,
+      "step": 13987
+    },
+    {
+      "epoch": 0.8604822834645669,
+      "grad_norm": 1.577118992805481,
+      "learning_rate": 9.464340767529645e-06,
+      "loss": 2.7495,
+      "step": 13988
+    },
+    {
+      "epoch": 0.8605437992125984,
+      "grad_norm": 1.3386741876602173,
+      "learning_rate": 9.456130691010967e-06,
+      "loss": 2.0362,
+      "step": 13989
+    },
+    {
+      "epoch": 0.86060531496063,
+      "grad_norm": 1.1416951417922974,
+      "learning_rate": 9.44792400032809e-06,
+      "loss": 2.3267,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8606668307086615,
+      "grad_norm": 1.2471740245819092,
+      "learning_rate": 9.439720695787945e-06,
+      "loss": 1.9596,
+      "step": 13991
+    },
+    {
+      "epoch": 0.8607283464566929,
+      "grad_norm": 1.2003973722457886,
+      "learning_rate": 9.431520777697255e-06,
+      "loss": 2.2609,
+      "step": 13992
+    },
+    {
+      "epoch": 0.8607898622047244,
+      "grad_norm": 1.273317575454712,
+      "learning_rate": 9.423324246362674e-06,
+      "loss": 2.4102,
+      "step": 13993
+    },
+    {
+      "epoch": 0.8608513779527559,
+      "grad_norm": 2.288073778152466,
+      "learning_rate": 9.415131102090691e-06,
+      "loss": 1.9301,
+      "step": 13994
+    },
+    {
+      "epoch": 0.8609128937007874,
+      "grad_norm": 1.249241590499878,
+      "learning_rate": 9.406941345187681e-06,
+      "loss": 1.4867,
+      "step": 13995
+    },
+    {
+      "epoch": 0.8609744094488189,
+      "grad_norm": 1.3160735368728638,
+      "learning_rate": 9.3987549759599e-06,
+      "loss": 2.668,
+      "step": 13996
+    },
+    {
+      "epoch": 0.8610359251968503,
+      "grad_norm": 1.2620989084243774,
+      "learning_rate": 9.390571994713504e-06,
+      "loss": 2.2325,
+      "step": 13997
+    },
+    {
+      "epoch": 0.8610974409448819,
+      "grad_norm": 1.224686622619629,
+      "learning_rate": 9.382392401754437e-06,
+      "loss": 2.8753,
+      "step": 13998
+    },
+    {
+      "epoch": 0.8611589566929134,
+      "grad_norm": 1.973161220550537,
+      "learning_rate": 9.374216197388619e-06,
+      "loss": 2.2988,
+      "step": 13999
+    },
+    {
+      "epoch": 0.8612204724409449,
+      "grad_norm": 1.1927942037582397,
+      "learning_rate": 9.366043381921763e-06,
+      "loss": 2.6039,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8612819881889764,
+      "grad_norm": 1.3392022848129272,
+      "learning_rate": 9.35787395565948e-06,
+      "loss": 1.6933,
+      "step": 14001
+    },
+    {
+      "epoch": 0.8613435039370079,
+      "grad_norm": 1.2357146739959717,
+      "learning_rate": 9.34970791890728e-06,
+      "loss": 2.0801,
+      "step": 14002
+    },
+    {
+      "epoch": 0.8614050196850394,
+      "grad_norm": 1.2877540588378906,
+      "learning_rate": 9.341545271970553e-06,
+      "loss": 2.1746,
+      "step": 14003
+    },
+    {
+      "epoch": 0.8614665354330708,
+      "grad_norm": 1.1682462692260742,
+      "learning_rate": 9.333386015154478e-06,
+      "loss": 2.4668,
+      "step": 14004
+    },
+    {
+      "epoch": 0.8615280511811023,
+      "grad_norm": 1.1500688791275024,
+      "learning_rate": 9.325230148764219e-06,
+      "loss": 2.5399,
+      "step": 14005
+    },
+    {
+      "epoch": 0.8615895669291339,
+      "grad_norm": 2.0737738609313965,
+      "learning_rate": 9.317077673104713e-06,
+      "loss": 1.7089,
+      "step": 14006
+    },
+    {
+      "epoch": 0.8616510826771654,
+      "grad_norm": 1.2435898780822754,
+      "learning_rate": 9.308928588480848e-06,
+      "loss": 2.1174,
+      "step": 14007
+    },
+    {
+      "epoch": 0.8617125984251969,
+      "grad_norm": 1.6855230331420898,
+      "learning_rate": 9.300782895197336e-06,
+      "loss": 2.1371,
+      "step": 14008
+    },
+    {
+      "epoch": 0.8617741141732284,
+      "grad_norm": 1.4401233196258545,
+      "learning_rate": 9.29264059355881e-06,
+      "loss": 2.7741,
+      "step": 14009
+    },
+    {
+      "epoch": 0.8618356299212598,
+      "grad_norm": 1.3158395290374756,
+      "learning_rate": 9.284501683869717e-06,
+      "loss": 2.7021,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8618971456692913,
+      "grad_norm": 1.4105370044708252,
+      "learning_rate": 9.276366166434436e-06,
+      "loss": 2.6224,
+      "step": 14011
+    },
+    {
+      "epoch": 0.8619586614173228,
+      "grad_norm": 1.2420649528503418,
+      "learning_rate": 9.268234041557144e-06,
+      "loss": 2.7767,
+      "step": 14012
+    },
+    {
+      "epoch": 0.8620201771653543,
+      "grad_norm": 1.2428544759750366,
+      "learning_rate": 9.260105309541989e-06,
+      "loss": 2.4979,
+      "step": 14013
+    },
+    {
+      "epoch": 0.8620816929133859,
+      "grad_norm": 1.226555585861206,
+      "learning_rate": 9.251979970692903e-06,
+      "loss": 2.4149,
+      "step": 14014
+    },
+    {
+      "epoch": 0.8621432086614174,
+      "grad_norm": 1.36845064163208,
+      "learning_rate": 9.243858025313734e-06,
+      "loss": 2.3686,
+      "step": 14015
+    },
+    {
+      "epoch": 0.8622047244094488,
+      "grad_norm": 1.2694603204727173,
+      "learning_rate": 9.235739473708228e-06,
+      "loss": 1.8899,
+      "step": 14016
+    },
+    {
+      "epoch": 0.8622662401574803,
+      "grad_norm": 1.1000690460205078,
+      "learning_rate": 9.227624316179929e-06,
+      "loss": 1.9672,
+      "step": 14017
+    },
+    {
+      "epoch": 0.8623277559055118,
+      "grad_norm": 1.2241405248641968,
+      "learning_rate": 9.21951255303234e-06,
+      "loss": 2.2666,
+      "step": 14018
+    },
+    {
+      "epoch": 0.8623892716535433,
+      "grad_norm": 1.2288271188735962,
+      "learning_rate": 9.21140418456875e-06,
+      "loss": 2.281,
+      "step": 14019
+    },
+    {
+      "epoch": 0.8624507874015748,
+      "grad_norm": 1.2780499458312988,
+      "learning_rate": 9.203299211092397e-06,
+      "loss": 2.6546,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8625123031496063,
+      "grad_norm": 1.2496006488800049,
+      "learning_rate": 9.19519763290636e-06,
+      "loss": 2.2015,
+      "step": 14021
+    },
+    {
+      "epoch": 0.8625738188976378,
+      "grad_norm": 1.2654917240142822,
+      "learning_rate": 9.187099450313607e-06,
+      "loss": 2.6784,
+      "step": 14022
+    },
+    {
+      "epoch": 0.8626353346456693,
+      "grad_norm": 1.1232088804244995,
+      "learning_rate": 9.17900466361693e-06,
+      "loss": 1.9827,
+      "step": 14023
+    },
+    {
+      "epoch": 0.8626968503937008,
+      "grad_norm": 1.2875484228134155,
+      "learning_rate": 9.170913273119063e-06,
+      "loss": 2.8327,
+      "step": 14024
+    },
+    {
+      "epoch": 0.8627583661417323,
+      "grad_norm": 1.62112557888031,
+      "learning_rate": 9.162825279122545e-06,
+      "loss": 2.3025,
+      "step": 14025
+    },
+    {
+      "epoch": 0.8628198818897638,
+      "grad_norm": 1.8459941148757935,
+      "learning_rate": 9.154740681929841e-06,
+      "loss": 1.7986,
+      "step": 14026
+    },
+    {
+      "epoch": 0.8628813976377953,
+      "grad_norm": 1.2569661140441895,
+      "learning_rate": 9.146659481843267e-06,
+      "loss": 2.7207,
+      "step": 14027
+    },
+    {
+      "epoch": 0.8629429133858267,
+      "grad_norm": 1.551573395729065,
+      "learning_rate": 9.138581679165036e-06,
+      "loss": 2.1154,
+      "step": 14028
+    },
+    {
+      "epoch": 0.8630044291338582,
+      "grad_norm": 1.1754206418991089,
+      "learning_rate": 9.130507274197175e-06,
+      "loss": 2.8286,
+      "step": 14029
+    },
+    {
+      "epoch": 0.8630659448818898,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.122436267241652e-06,
+      "loss": 2.3799,
+      "step": 14030
+    },
+    {
+      "epoch": 0.8631274606299213,
+      "grad_norm": 1.3329954147338867,
+      "learning_rate": 9.11436865860026e-06,
+      "loss": 2.7611,
+      "step": 14031
+    },
+    {
+      "epoch": 0.8631889763779528,
+      "grad_norm": 1.1809698343276978,
+      "learning_rate": 9.106304448574677e-06,
+      "loss": 2.5078,
+      "step": 14032
+    },
+    {
+      "epoch": 0.8632504921259843,
+      "grad_norm": 1.2812434434890747,
+      "learning_rate": 9.098243637466474e-06,
+      "loss": 2.9189,
+      "step": 14033
+    },
+    {
+      "epoch": 0.8633120078740157,
+      "grad_norm": 1.190760612487793,
+      "learning_rate": 9.090186225577091e-06,
+      "loss": 2.1504,
+      "step": 14034
+    },
+    {
+      "epoch": 0.8633735236220472,
+      "grad_norm": 1.297427773475647,
+      "learning_rate": 9.082132213207807e-06,
+      "loss": 2.6582,
+      "step": 14035
+    },
+    {
+      "epoch": 0.8634350393700787,
+      "grad_norm": 1.4968986511230469,
+      "learning_rate": 9.074081600659811e-06,
+      "loss": 2.8201,
+      "step": 14036
+    },
+    {
+      "epoch": 0.8634965551181102,
+      "grad_norm": 1.3134163618087769,
+      "learning_rate": 9.066034388234145e-06,
+      "loss": 2.6923,
+      "step": 14037
+    },
+    {
+      "epoch": 0.8635580708661418,
+      "grad_norm": 1.2083148956298828,
+      "learning_rate": 9.057990576231735e-06,
+      "loss": 2.464,
+      "step": 14038
+    },
+    {
+      "epoch": 0.8636195866141733,
+      "grad_norm": 1.3729135990142822,
+      "learning_rate": 9.049950164953357e-06,
+      "loss": 2.5804,
+      "step": 14039
+    },
+    {
+      "epoch": 0.8636811023622047,
+      "grad_norm": 1.3434351682662964,
+      "learning_rate": 9.0419131546997e-06,
+      "loss": 2.5367,
+      "step": 14040
+    },
+    {
+      "epoch": 0.8637426181102362,
+      "grad_norm": 1.4821816682815552,
+      "learning_rate": 9.033879545771284e-06,
+      "loss": 2.38,
+      "step": 14041
+    },
+    {
+      "epoch": 0.8638041338582677,
+      "grad_norm": 1.2711180448532104,
+      "learning_rate": 9.02584933846855e-06,
+      "loss": 2.5579,
+      "step": 14042
+    },
+    {
+      "epoch": 0.8638656496062992,
+      "grad_norm": 1.3138251304626465,
+      "learning_rate": 9.017822533091758e-06,
+      "loss": 2.5188,
+      "step": 14043
+    },
+    {
+      "epoch": 0.8639271653543307,
+      "grad_norm": 1.3596148490905762,
+      "learning_rate": 9.009799129941077e-06,
+      "loss": 2.9986,
+      "step": 14044
+    },
+    {
+      "epoch": 0.8639886811023622,
+      "grad_norm": 1.2614679336547852,
+      "learning_rate": 9.001779129316523e-06,
+      "loss": 2.4569,
+      "step": 14045
+    },
+    {
+      "epoch": 0.8640501968503937,
+      "grad_norm": 1.1457555294036865,
+      "learning_rate": 8.993762531518012e-06,
+      "loss": 2.5227,
+      "step": 14046
+    },
+    {
+      "epoch": 0.8641117125984252,
+      "grad_norm": 1.1355466842651367,
+      "learning_rate": 8.985749336845306e-06,
+      "loss": 2.3998,
+      "step": 14047
+    },
+    {
+      "epoch": 0.8641732283464567,
+      "grad_norm": 1.4730881452560425,
+      "learning_rate": 8.977739545598096e-06,
+      "loss": 2.5717,
+      "step": 14048
+    },
+    {
+      "epoch": 0.8642347440944882,
+      "grad_norm": 1.3865768909454346,
+      "learning_rate": 8.969733158075855e-06,
+      "loss": 2.2984,
+      "step": 14049
+    },
+    {
+      "epoch": 0.8642962598425197,
+      "grad_norm": 1.2393232583999634,
+      "learning_rate": 8.961730174577998e-06,
+      "loss": 2.6241,
+      "step": 14050
+    },
+    {
+      "epoch": 0.8643577755905512,
+      "grad_norm": 1.3583917617797852,
+      "learning_rate": 8.953730595403786e-06,
+      "loss": 2.358,
+      "step": 14051
+    },
+    {
+      "epoch": 0.8644192913385826,
+      "grad_norm": 1.9939013719558716,
+      "learning_rate": 8.945734420852358e-06,
+      "loss": 2.4522,
+      "step": 14052
+    },
+    {
+      "epoch": 0.8644808070866141,
+      "grad_norm": 1.4536031484603882,
+      "learning_rate": 8.937741651222741e-06,
+      "loss": 2.6153,
+      "step": 14053
+    },
+    {
+      "epoch": 0.8645423228346457,
+      "grad_norm": 1.3550282716751099,
+      "learning_rate": 8.929752286813808e-06,
+      "loss": 2.3724,
+      "step": 14054
+    },
+    {
+      "epoch": 0.8646038385826772,
+      "grad_norm": 1.2844212055206299,
+      "learning_rate": 8.92176632792432e-06,
+      "loss": 2.549,
+      "step": 14055
+    },
+    {
+      "epoch": 0.8646653543307087,
+      "grad_norm": 1.3215166330337524,
+      "learning_rate": 8.91378377485289e-06,
+      "loss": 2.0132,
+      "step": 14056
+    },
+    {
+      "epoch": 0.8647268700787402,
+      "grad_norm": 1.272778868675232,
+      "learning_rate": 8.905804627898063e-06,
+      "loss": 2.5375,
+      "step": 14057
+    },
+    {
+      "epoch": 0.8647883858267716,
+      "grad_norm": 1.5495765209197998,
+      "learning_rate": 8.897828887358139e-06,
+      "loss": 2.5078,
+      "step": 14058
+    },
+    {
+      "epoch": 0.8648499015748031,
+      "grad_norm": 1.368237018585205,
+      "learning_rate": 8.889856553531461e-06,
+      "loss": 2.6052,
+      "step": 14059
+    },
+    {
+      "epoch": 0.8649114173228346,
+      "grad_norm": 1.2437336444854736,
+      "learning_rate": 8.88188762671608e-06,
+      "loss": 2.6436,
+      "step": 14060
+    },
+    {
+      "epoch": 0.8649729330708661,
+      "grad_norm": 1.3383865356445312,
+      "learning_rate": 8.873922107210043e-06,
+      "loss": 2.7597,
+      "step": 14061
+    },
+    {
+      "epoch": 0.8650344488188977,
+      "grad_norm": 1.3399720191955566,
+      "learning_rate": 8.865959995311158e-06,
+      "loss": 2.2658,
+      "step": 14062
+    },
+    {
+      "epoch": 0.8650959645669292,
+      "grad_norm": 1.4372003078460693,
+      "learning_rate": 8.858001291317209e-06,
+      "loss": 2.1163,
+      "step": 14063
+    },
+    {
+      "epoch": 0.8651574803149606,
+      "grad_norm": 1.5520179271697998,
+      "learning_rate": 8.850045995525768e-06,
+      "loss": 2.582,
+      "step": 14064
+    },
+    {
+      "epoch": 0.8652189960629921,
+      "grad_norm": 1.3409626483917236,
+      "learning_rate": 8.842094108234344e-06,
+      "loss": 2.1645,
+      "step": 14065
+    },
+    {
+      "epoch": 0.8652805118110236,
+      "grad_norm": 1.4509973526000977,
+      "learning_rate": 8.834145629740298e-06,
+      "loss": 2.2941,
+      "step": 14066
+    },
+    {
+      "epoch": 0.8653420275590551,
+      "grad_norm": 1.1905908584594727,
+      "learning_rate": 8.826200560340859e-06,
+      "loss": 2.8545,
+      "step": 14067
+    },
+    {
+      "epoch": 0.8654035433070866,
+      "grad_norm": 1.2669042348861694,
+      "learning_rate": 8.818258900333109e-06,
+      "loss": 2.9668,
+      "step": 14068
+    },
+    {
+      "epoch": 0.8654650590551181,
+      "grad_norm": 1.5343475341796875,
+      "learning_rate": 8.810320650014048e-06,
+      "loss": 2.4664,
+      "step": 14069
+    },
+    {
+      "epoch": 0.8655265748031497,
+      "grad_norm": 1.214093565940857,
+      "learning_rate": 8.802385809680491e-06,
+      "loss": 2.5508,
+      "step": 14070
+    },
+    {
+      "epoch": 0.8655880905511811,
+      "grad_norm": 1.2915093898773193,
+      "learning_rate": 8.794454379629168e-06,
+      "loss": 2.9254,
+      "step": 14071
+    },
+    {
+      "epoch": 0.8656496062992126,
+      "grad_norm": 1.3428475856781006,
+      "learning_rate": 8.786526360156677e-06,
+      "loss": 2.4527,
+      "step": 14072
+    },
+    {
+      "epoch": 0.8657111220472441,
+      "grad_norm": 1.2964283227920532,
+      "learning_rate": 8.7786017515595e-06,
+      "loss": 2.6109,
+      "step": 14073
+    },
+    {
+      "epoch": 0.8657726377952756,
+      "grad_norm": 1.3835539817810059,
+      "learning_rate": 8.770680554133937e-06,
+      "loss": 2.2184,
+      "step": 14074
+    },
+    {
+      "epoch": 0.8658341535433071,
+      "grad_norm": 1.2582675218582153,
+      "learning_rate": 8.762762768176224e-06,
+      "loss": 2.3818,
+      "step": 14075
+    },
+    {
+      "epoch": 0.8658956692913385,
+      "grad_norm": 1.3718618154525757,
+      "learning_rate": 8.754848393982418e-06,
+      "loss": 2.5431,
+      "step": 14076
+    },
+    {
+      "epoch": 0.86595718503937,
+      "grad_norm": 1.1487138271331787,
+      "learning_rate": 8.746937431848478e-06,
+      "loss": 2.9677,
+      "step": 14077
+    },
+    {
+      "epoch": 0.8660187007874016,
+      "grad_norm": 1.2716060876846313,
+      "learning_rate": 8.739029882070248e-06,
+      "loss": 2.0453,
+      "step": 14078
+    },
+    {
+      "epoch": 0.8660802165354331,
+      "grad_norm": 1.3146408796310425,
+      "learning_rate": 8.731125744943424e-06,
+      "loss": 1.9487,
+      "step": 14079
+    },
+    {
+      "epoch": 0.8661417322834646,
+      "grad_norm": 1.1532090902328491,
+      "learning_rate": 8.723225020763559e-06,
+      "loss": 2.7587,
+      "step": 14080
+    },
+    {
+      "epoch": 0.8662032480314961,
+      "grad_norm": 1.2910743951797485,
+      "learning_rate": 8.715327709826115e-06,
+      "loss": 2.719,
+      "step": 14081
+    },
+    {
+      "epoch": 0.8662647637795275,
+      "grad_norm": 1.3898738622665405,
+      "learning_rate": 8.70743381242638e-06,
+      "loss": 3.1396,
+      "step": 14082
+    },
+    {
+      "epoch": 0.866326279527559,
+      "grad_norm": 1.1866575479507446,
+      "learning_rate": 8.699543328859583e-06,
+      "loss": 2.6768,
+      "step": 14083
+    },
+    {
+      "epoch": 0.8663877952755905,
+      "grad_norm": 1.315994381904602,
+      "learning_rate": 8.691656259420711e-06,
+      "loss": 2.1189,
+      "step": 14084
+    },
+    {
+      "epoch": 0.866449311023622,
+      "grad_norm": 1.2520545721054077,
+      "learning_rate": 8.683772604404794e-06,
+      "loss": 2.6913,
+      "step": 14085
+    },
+    {
+      "epoch": 0.8665108267716536,
+      "grad_norm": 1.3977640867233276,
+      "learning_rate": 8.675892364106564e-06,
+      "loss": 2.135,
+      "step": 14086
+    },
+    {
+      "epoch": 0.8665723425196851,
+      "grad_norm": 1.4590426683425903,
+      "learning_rate": 8.668015538820751e-06,
+      "loss": 2.809,
+      "step": 14087
+    },
+    {
+      "epoch": 0.8666338582677166,
+      "grad_norm": 1.583214521408081,
+      "learning_rate": 8.660142128841863e-06,
+      "loss": 2.4726,
+      "step": 14088
+    },
+    {
+      "epoch": 0.866695374015748,
+      "grad_norm": 1.1586024761199951,
+      "learning_rate": 8.652272134464322e-06,
+      "loss": 2.2903,
+      "step": 14089
+    },
+    {
+      "epoch": 0.8667568897637795,
+      "grad_norm": 1.2090239524841309,
+      "learning_rate": 8.644405555982438e-06,
+      "loss": 3.0484,
+      "step": 14090
+    },
+    {
+      "epoch": 0.866818405511811,
+      "grad_norm": 1.3761941194534302,
+      "learning_rate": 8.63654239369036e-06,
+      "loss": 2.6384,
+      "step": 14091
+    },
+    {
+      "epoch": 0.8668799212598425,
+      "grad_norm": 1.2169564962387085,
+      "learning_rate": 8.628682647882169e-06,
+      "loss": 2.9799,
+      "step": 14092
+    },
+    {
+      "epoch": 0.866941437007874,
+      "grad_norm": 1.483056902885437,
+      "learning_rate": 8.620826318851728e-06,
+      "loss": 2.7358,
+      "step": 14093
+    },
+    {
+      "epoch": 0.8670029527559056,
+      "grad_norm": 1.211889624595642,
+      "learning_rate": 8.612973406892854e-06,
+      "loss": 2.1197,
+      "step": 14094
+    },
+    {
+      "epoch": 0.867064468503937,
+      "grad_norm": 1.323366641998291,
+      "learning_rate": 8.605123912299173e-06,
+      "loss": 2.8276,
+      "step": 14095
+    },
+    {
+      "epoch": 0.8671259842519685,
+      "grad_norm": 1.5127116441726685,
+      "learning_rate": 8.597277835364225e-06,
+      "loss": 1.9615,
+      "step": 14096
+    },
+    {
+      "epoch": 0.8671875,
+      "grad_norm": 1.3138967752456665,
+      "learning_rate": 8.589435176381411e-06,
+      "loss": 2.4053,
+      "step": 14097
+    },
+    {
+      "epoch": 0.8672490157480315,
+      "grad_norm": 1.3913681507110596,
+      "learning_rate": 8.581595935644016e-06,
+      "loss": 2.6322,
+      "step": 14098
+    },
+    {
+      "epoch": 0.867310531496063,
+      "grad_norm": 1.3626700639724731,
+      "learning_rate": 8.57376011344515e-06,
+      "loss": 2.6835,
+      "step": 14099
+    },
+    {
+      "epoch": 0.8673720472440944,
+      "grad_norm": 1.2137432098388672,
+      "learning_rate": 8.56592771007787e-06,
+      "loss": 3.1406,
+      "step": 14100
+    },
+    {
+      "epoch": 0.867433562992126,
+      "grad_norm": 1.4024875164031982,
+      "learning_rate": 8.55809872583503e-06,
+      "loss": 2.184,
+      "step": 14101
+    },
+    {
+      "epoch": 0.8674950787401575,
+      "grad_norm": 1.2782297134399414,
+      "learning_rate": 8.550273161009414e-06,
+      "loss": 2.4902,
+      "step": 14102
+    },
+    {
+      "epoch": 0.867556594488189,
+      "grad_norm": 1.3942896127700806,
+      "learning_rate": 8.542451015893605e-06,
+      "loss": 2.1689,
+      "step": 14103
+    },
+    {
+      "epoch": 0.8676181102362205,
+      "grad_norm": 1.2766207456588745,
+      "learning_rate": 8.534632290780197e-06,
+      "loss": 2.23,
+      "step": 14104
+    },
+    {
+      "epoch": 0.867679625984252,
+      "grad_norm": 1.2604835033416748,
+      "learning_rate": 8.526816985961484e-06,
+      "loss": 1.9838,
+      "step": 14105
+    },
+    {
+      "epoch": 0.8677411417322834,
+      "grad_norm": 1.3604243993759155,
+      "learning_rate": 8.519005101729771e-06,
+      "loss": 2.4873,
+      "step": 14106
+    },
+    {
+      "epoch": 0.8678026574803149,
+      "grad_norm": 1.34578275680542,
+      "learning_rate": 8.51119663837714e-06,
+      "loss": 1.6862,
+      "step": 14107
+    },
+    {
+      "epoch": 0.8678641732283464,
+      "grad_norm": 1.3458623886108398,
+      "learning_rate": 8.50339159619562e-06,
+      "loss": 2.2938,
+      "step": 14108
+    },
+    {
+      "epoch": 0.867925688976378,
+      "grad_norm": 1.4173238277435303,
+      "learning_rate": 8.495589975477025e-06,
+      "loss": 2.3537,
+      "step": 14109
+    },
+    {
+      "epoch": 0.8679872047244095,
+      "grad_norm": 1.4008333683013916,
+      "learning_rate": 8.487791776513132e-06,
+      "loss": 2.0599,
+      "step": 14110
+    },
+    {
+      "epoch": 0.868048720472441,
+      "grad_norm": 1.7199045419692993,
+      "learning_rate": 8.47999699959554e-06,
+      "loss": 2.4645,
+      "step": 14111
+    },
+    {
+      "epoch": 0.8681102362204725,
+      "grad_norm": 1.6829296350479126,
+      "learning_rate": 8.47220564501574e-06,
+      "loss": 1.9412,
+      "step": 14112
+    },
+    {
+      "epoch": 0.8681717519685039,
+      "grad_norm": 1.2998974323272705,
+      "learning_rate": 8.464417713065064e-06,
+      "loss": 2.7287,
+      "step": 14113
+    },
+    {
+      "epoch": 0.8682332677165354,
+      "grad_norm": 1.0725066661834717,
+      "learning_rate": 8.45663320403477e-06,
+      "loss": 2.8998,
+      "step": 14114
+    },
+    {
+      "epoch": 0.8682947834645669,
+      "grad_norm": 1.2946985960006714,
+      "learning_rate": 8.448852118215921e-06,
+      "loss": 2.8822,
+      "step": 14115
+    },
+    {
+      "epoch": 0.8683562992125984,
+      "grad_norm": 1.878522515296936,
+      "learning_rate": 8.441074455899489e-06,
+      "loss": 2.342,
+      "step": 14116
+    },
+    {
+      "epoch": 0.86841781496063,
+      "grad_norm": 1.0662479400634766,
+      "learning_rate": 8.433300217376328e-06,
+      "loss": 2.9693,
+      "step": 14117
+    },
+    {
+      "epoch": 0.8684793307086615,
+      "grad_norm": 1.4255255460739136,
+      "learning_rate": 8.425529402937171e-06,
+      "loss": 2.2506,
+      "step": 14118
+    },
+    {
+      "epoch": 0.8685408464566929,
+      "grad_norm": 1.1871894598007202,
+      "learning_rate": 8.417762012872555e-06,
+      "loss": 2.622,
+      "step": 14119
+    },
+    {
+      "epoch": 0.8686023622047244,
+      "grad_norm": 1.284087061882019,
+      "learning_rate": 8.409998047472988e-06,
+      "loss": 1.9616,
+      "step": 14120
+    },
+    {
+      "epoch": 0.8686638779527559,
+      "grad_norm": 1.3172767162322998,
+      "learning_rate": 8.402237507028754e-06,
+      "loss": 2.3947,
+      "step": 14121
+    },
+    {
+      "epoch": 0.8687253937007874,
+      "grad_norm": 1.2848008871078491,
+      "learning_rate": 8.394480391830073e-06,
+      "loss": 2.2792,
+      "step": 14122
+    },
+    {
+      "epoch": 0.8687869094488189,
+      "grad_norm": 1.2367521524429321,
+      "learning_rate": 8.386726702167025e-06,
+      "loss": 1.8382,
+      "step": 14123
+    },
+    {
+      "epoch": 0.8688484251968503,
+      "grad_norm": 1.3185133934020996,
+      "learning_rate": 8.378976438329567e-06,
+      "loss": 1.9318,
+      "step": 14124
+    },
+    {
+      "epoch": 0.8689099409448819,
+      "grad_norm": 1.5728802680969238,
+      "learning_rate": 8.371229600607488e-06,
+      "loss": 1.8948,
+      "step": 14125
+    },
+    {
+      "epoch": 0.8689714566929134,
+      "grad_norm": 1.1163465976715088,
+      "learning_rate": 8.36348618929047e-06,
+      "loss": 2.385,
+      "step": 14126
+    },
+    {
+      "epoch": 0.8690329724409449,
+      "grad_norm": 1.2998450994491577,
+      "learning_rate": 8.355746204668113e-06,
+      "loss": 3.2319,
+      "step": 14127
+    },
+    {
+      "epoch": 0.8690944881889764,
+      "grad_norm": 1.1091629266738892,
+      "learning_rate": 8.348009647029808e-06,
+      "loss": 1.0375,
+      "step": 14128
+    },
+    {
+      "epoch": 0.8691560039370079,
+      "grad_norm": 1.4331127405166626,
+      "learning_rate": 8.34027651666487e-06,
+      "loss": 2.0892,
+      "step": 14129
+    },
+    {
+      "epoch": 0.8692175196850394,
+      "grad_norm": 1.2238578796386719,
+      "learning_rate": 8.332546813862486e-06,
+      "loss": 1.9639,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8692790354330708,
+      "grad_norm": 1.356292963027954,
+      "learning_rate": 8.324820538911716e-06,
+      "loss": 2.485,
+      "step": 14131
+    },
+    {
+      "epoch": 0.8693405511811023,
+      "grad_norm": 1.3753610849380493,
+      "learning_rate": 8.31709769210145e-06,
+      "loss": 2.5487,
+      "step": 14132
+    },
+    {
+      "epoch": 0.8694020669291339,
+      "grad_norm": 1.4242981672286987,
+      "learning_rate": 8.309378273720503e-06,
+      "loss": 2.406,
+      "step": 14133
+    },
+    {
+      "epoch": 0.8694635826771654,
+      "grad_norm": 1.3640331029891968,
+      "learning_rate": 8.30166228405751e-06,
+      "loss": 2.7875,
+      "step": 14134
+    },
+    {
+      "epoch": 0.8695250984251969,
+      "grad_norm": 1.3524518013000488,
+      "learning_rate": 8.293949723401029e-06,
+      "loss": 2.0248,
+      "step": 14135
+    },
+    {
+      "epoch": 0.8695866141732284,
+      "grad_norm": 1.3500173091888428,
+      "learning_rate": 8.286240592039462e-06,
+      "loss": 2.4374,
+      "step": 14136
+    },
+    {
+      "epoch": 0.8696481299212598,
+      "grad_norm": 1.2449162006378174,
+      "learning_rate": 8.278534890261091e-06,
+      "loss": 2.7673,
+      "step": 14137
+    },
+    {
+      "epoch": 0.8697096456692913,
+      "grad_norm": 1.210629940032959,
+      "learning_rate": 8.270832618354063e-06,
+      "loss": 2.5949,
+      "step": 14138
+    },
+    {
+      "epoch": 0.8697711614173228,
+      "grad_norm": 1.2721108198165894,
+      "learning_rate": 8.263133776606402e-06,
+      "loss": 2.4258,
+      "step": 14139
+    },
+    {
+      "epoch": 0.8698326771653543,
+      "grad_norm": 1.4230594635009766,
+      "learning_rate": 8.255438365305989e-06,
+      "loss": 1.9967,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8698941929133859,
+      "grad_norm": 1.382490634918213,
+      "learning_rate": 8.247746384740606e-06,
+      "loss": 3.1081,
+      "step": 14141
+    },
+    {
+      "epoch": 0.8699557086614174,
+      "grad_norm": 1.3590558767318726,
+      "learning_rate": 8.240057835197867e-06,
+      "loss": 2.3684,
+      "step": 14142
+    },
+    {
+      "epoch": 0.8700172244094488,
+      "grad_norm": 1.3809738159179688,
+      "learning_rate": 8.23237271696533e-06,
+      "loss": 2.2847,
+      "step": 14143
+    },
+    {
+      "epoch": 0.8700787401574803,
+      "grad_norm": 1.170311689376831,
+      "learning_rate": 8.224691030330323e-06,
+      "loss": 2.7258,
+      "step": 14144
+    },
+    {
+      "epoch": 0.8701402559055118,
+      "grad_norm": 1.0964503288269043,
+      "learning_rate": 8.217012775580123e-06,
+      "loss": 3.1623,
+      "step": 14145
+    },
+    {
+      "epoch": 0.8702017716535433,
+      "grad_norm": 1.3058570623397827,
+      "learning_rate": 8.20933795300185e-06,
+      "loss": 2.8764,
+      "step": 14146
+    },
+    {
+      "epoch": 0.8702632874015748,
+      "grad_norm": 1.2361488342285156,
+      "learning_rate": 8.201666562882482e-06,
+      "loss": 2.7466,
+      "step": 14147
+    },
+    {
+      "epoch": 0.8703248031496063,
+      "grad_norm": 1.5874764919281006,
+      "learning_rate": 8.193998605508913e-06,
+      "loss": 2.3412,
+      "step": 14148
+    },
+    {
+      "epoch": 0.8703863188976378,
+      "grad_norm": 1.1206063032150269,
+      "learning_rate": 8.18633408116789e-06,
+      "loss": 2.956,
+      "step": 14149
+    },
+    {
+      "epoch": 0.8704478346456693,
+      "grad_norm": 1.3066354990005493,
+      "learning_rate": 8.178672990145986e-06,
+      "loss": 2.3183,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8705093503937008,
+      "grad_norm": 1.9235862493515015,
+      "learning_rate": 8.171015332729715e-06,
+      "loss": 2.3993,
+      "step": 14151
+    },
+    {
+      "epoch": 0.8705708661417323,
+      "grad_norm": 1.2593624591827393,
+      "learning_rate": 8.163361109205404e-06,
+      "loss": 1.8556,
+      "step": 14152
+    },
+    {
+      "epoch": 0.8706323818897638,
+      "grad_norm": 1.2920002937316895,
+      "learning_rate": 8.15571031985931e-06,
+      "loss": 2.5898,
+      "step": 14153
+    },
+    {
+      "epoch": 0.8706938976377953,
+      "grad_norm": 1.3575658798217773,
+      "learning_rate": 8.148062964977498e-06,
+      "loss": 2.7695,
+      "step": 14154
+    },
+    {
+      "epoch": 0.8707554133858267,
+      "grad_norm": 1.1614975929260254,
+      "learning_rate": 8.140419044845949e-06,
+      "loss": 2.6905,
+      "step": 14155
+    },
+    {
+      "epoch": 0.8708169291338582,
+      "grad_norm": 1.134259581565857,
+      "learning_rate": 8.132778559750498e-06,
+      "loss": 2.5973,
+      "step": 14156
+    },
+    {
+      "epoch": 0.8708784448818898,
+      "grad_norm": 1.3407979011535645,
+      "learning_rate": 8.125141509976886e-06,
+      "loss": 2.9895,
+      "step": 14157
+    },
+    {
+      "epoch": 0.8709399606299213,
+      "grad_norm": 1.3887531757354736,
+      "learning_rate": 8.117507895810662e-06,
+      "loss": 2.2034,
+      "step": 14158
+    },
+    {
+      "epoch": 0.8710014763779528,
+      "grad_norm": 1.506487488746643,
+      "learning_rate": 8.109877717537306e-06,
+      "loss": 2.4118,
+      "step": 14159
+    },
+    {
+      "epoch": 0.8710629921259843,
+      "grad_norm": 1.3634254932403564,
+      "learning_rate": 8.102250975442106e-06,
+      "loss": 3.1055,
+      "step": 14160
+    },
+    {
+      "epoch": 0.8711245078740157,
+      "grad_norm": 1.3989475965499878,
+      "learning_rate": 8.094627669810284e-06,
+      "loss": 1.9512,
+      "step": 14161
+    },
+    {
+      "epoch": 0.8711860236220472,
+      "grad_norm": 1.220679759979248,
+      "learning_rate": 8.087007800926938e-06,
+      "loss": 2.5601,
+      "step": 14162
+    },
+    {
+      "epoch": 0.8712475393700787,
+      "grad_norm": 1.3246560096740723,
+      "learning_rate": 8.07939136907695e-06,
+      "loss": 2.0956,
+      "step": 14163
+    },
+    {
+      "epoch": 0.8713090551181102,
+      "grad_norm": 1.3454142808914185,
+      "learning_rate": 8.07177837454518e-06,
+      "loss": 2.0538,
+      "step": 14164
+    },
+    {
+      "epoch": 0.8713705708661418,
+      "grad_norm": 1.2951537370681763,
+      "learning_rate": 8.06416881761628e-06,
+      "loss": 2.6878,
+      "step": 14165
+    },
+    {
+      "epoch": 0.8714320866141733,
+      "grad_norm": 1.3106507062911987,
+      "learning_rate": 8.05656269857482e-06,
+      "loss": 2.3139,
+      "step": 14166
+    },
+    {
+      "epoch": 0.8714936023622047,
+      "grad_norm": 1.2016959190368652,
+      "learning_rate": 8.04896001770522e-06,
+      "loss": 2.6324,
+      "step": 14167
+    },
+    {
+      "epoch": 0.8715551181102362,
+      "grad_norm": 1.219949722290039,
+      "learning_rate": 8.041360775291796e-06,
+      "loss": 2.6859,
+      "step": 14168
+    },
+    {
+      "epoch": 0.8716166338582677,
+      "grad_norm": 1.4508742094039917,
+      "learning_rate": 8.033764971618696e-06,
+      "loss": 2.6046,
+      "step": 14169
+    },
+    {
+      "epoch": 0.8716781496062992,
+      "grad_norm": 1.3843413591384888,
+      "learning_rate": 8.026172606969972e-06,
+      "loss": 2.7247,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8717396653543307,
+      "grad_norm": 1.3420552015304565,
+      "learning_rate": 8.018583681629522e-06,
+      "loss": 2.6294,
+      "step": 14171
+    },
+    {
+      "epoch": 0.8718011811023622,
+      "grad_norm": 1.1746011972427368,
+      "learning_rate": 8.01099819588116e-06,
+      "loss": 2.533,
+      "step": 14172
+    },
+    {
+      "epoch": 0.8718626968503937,
+      "grad_norm": 1.3281667232513428,
+      "learning_rate": 8.00341615000848e-06,
+      "loss": 2.1772,
+      "step": 14173
+    },
+    {
+      "epoch": 0.8719242125984252,
+      "grad_norm": 1.264695644378662,
+      "learning_rate": 7.995837544295093e-06,
+      "loss": 2.3538,
+      "step": 14174
+    },
+    {
+      "epoch": 0.8719857283464567,
+      "grad_norm": 1.2244898080825806,
+      "learning_rate": 7.988262379024325e-06,
+      "loss": 2.7529,
+      "step": 14175
+    },
+    {
+      "epoch": 0.8720472440944882,
+      "grad_norm": 1.3193626403808594,
+      "learning_rate": 7.980690654479506e-06,
+      "loss": 2.059,
+      "step": 14176
+    },
+    {
+      "epoch": 0.8721087598425197,
+      "grad_norm": 1.1449438333511353,
+      "learning_rate": 7.973122370943709e-06,
+      "loss": 2.583,
+      "step": 14177
+    },
+    {
+      "epoch": 0.8721702755905512,
+      "grad_norm": 1.3681868314743042,
+      "learning_rate": 7.965557528700008e-06,
+      "loss": 2.1982,
+      "step": 14178
+    },
+    {
+      "epoch": 0.8722317913385826,
+      "grad_norm": 1.1340041160583496,
+      "learning_rate": 7.95799612803123e-06,
+      "loss": 2.5075,
+      "step": 14179
+    },
+    {
+      "epoch": 0.8722933070866141,
+      "grad_norm": 1.3354839086532593,
+      "learning_rate": 7.950438169220154e-06,
+      "loss": 2.0678,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8723548228346457,
+      "grad_norm": 1.312503457069397,
+      "learning_rate": 7.942883652549415e-06,
+      "loss": 2.1482,
+      "step": 14181
+    },
+    {
+      "epoch": 0.8724163385826772,
+      "grad_norm": 1.3014541864395142,
+      "learning_rate": 7.935332578301524e-06,
+      "loss": 2.1299,
+      "step": 14182
+    },
+    {
+      "epoch": 0.8724778543307087,
+      "grad_norm": 1.3158801794052124,
+      "learning_rate": 7.92778494675881e-06,
+      "loss": 2.0157,
+      "step": 14183
+    },
+    {
+      "epoch": 0.8725393700787402,
+      "grad_norm": 1.2984039783477783,
+      "learning_rate": 7.920240758203545e-06,
+      "loss": 2.3525,
+      "step": 14184
+    },
+    {
+      "epoch": 0.8726008858267716,
+      "grad_norm": 1.3641633987426758,
+      "learning_rate": 7.912700012917806e-06,
+      "loss": 2.0724,
+      "step": 14185
+    },
+    {
+      "epoch": 0.8726624015748031,
+      "grad_norm": 1.3439209461212158,
+      "learning_rate": 7.905162711183589e-06,
+      "loss": 3.0159,
+      "step": 14186
+    },
+    {
+      "epoch": 0.8727239173228346,
+      "grad_norm": 1.399796724319458,
+      "learning_rate": 7.897628853282756e-06,
+      "loss": 2.1318,
+      "step": 14187
+    },
+    {
+      "epoch": 0.8727854330708661,
+      "grad_norm": 1.3454399108886719,
+      "learning_rate": 7.890098439497052e-06,
+      "loss": 2.3211,
+      "step": 14188
+    },
+    {
+      "epoch": 0.8728469488188977,
+      "grad_norm": 1.7120541334152222,
+      "learning_rate": 7.882571470108013e-06,
+      "loss": 2.0348,
+      "step": 14189
+    },
+    {
+      "epoch": 0.8729084645669292,
+      "grad_norm": 1.1379919052124023,
+      "learning_rate": 7.875047945397174e-06,
+      "loss": 1.7767,
+      "step": 14190
+    },
+    {
+      "epoch": 0.8729699803149606,
+      "grad_norm": 1.167424201965332,
+      "learning_rate": 7.867527865645818e-06,
+      "loss": 2.4536,
+      "step": 14191
+    },
+    {
+      "epoch": 0.8730314960629921,
+      "grad_norm": 1.5523369312286377,
+      "learning_rate": 7.860011231135167e-06,
+      "loss": 2.0236,
+      "step": 14192
+    },
+    {
+      "epoch": 0.8730930118110236,
+      "grad_norm": 1.2261406183242798,
+      "learning_rate": 7.852498042146317e-06,
+      "loss": 2.5856,
+      "step": 14193
+    },
+    {
+      "epoch": 0.8731545275590551,
+      "grad_norm": 1.3705096244812012,
+      "learning_rate": 7.84498829896022e-06,
+      "loss": 2.4709,
+      "step": 14194
+    },
+    {
+      "epoch": 0.8732160433070866,
+      "grad_norm": 1.3576514720916748,
+      "learning_rate": 7.837482001857687e-06,
+      "loss": 2.2726,
+      "step": 14195
+    },
+    {
+      "epoch": 0.8732775590551181,
+      "grad_norm": 1.5740668773651123,
+      "learning_rate": 7.829979151119426e-06,
+      "loss": 2.9261,
+      "step": 14196
+    },
+    {
+      "epoch": 0.8733390748031497,
+      "grad_norm": 1.0360807180404663,
+      "learning_rate": 7.822479747025969e-06,
+      "loss": 2.864,
+      "step": 14197
+    },
+    {
+      "epoch": 0.8734005905511811,
+      "grad_norm": 1.2101691961288452,
+      "learning_rate": 7.814983789857799e-06,
+      "loss": 2.4828,
+      "step": 14198
+    },
+    {
+      "epoch": 0.8734621062992126,
+      "grad_norm": 1.239941120147705,
+      "learning_rate": 7.807491279895175e-06,
+      "loss": 2.5027,
+      "step": 14199
+    },
+    {
+      "epoch": 0.8735236220472441,
+      "grad_norm": 1.561737298965454,
+      "learning_rate": 7.80000221741829e-06,
+      "loss": 2.925,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8735851377952756,
+      "grad_norm": 1.2969794273376465,
+      "learning_rate": 7.792516602707235e-06,
+      "loss": 2.1638,
+      "step": 14201
+    },
+    {
+      "epoch": 0.8736466535433071,
+      "grad_norm": 1.6309168338775635,
+      "learning_rate": 7.78503443604186e-06,
+      "loss": 2.6187,
+      "step": 14202
+    },
+    {
+      "epoch": 0.8737081692913385,
+      "grad_norm": 1.7116584777832031,
+      "learning_rate": 7.777555717702023e-06,
+      "loss": 1.2473,
+      "step": 14203
+    },
+    {
+      "epoch": 0.87376968503937,
+      "grad_norm": 1.2490482330322266,
+      "learning_rate": 7.770080447967332e-06,
+      "loss": 2.0527,
+      "step": 14204
+    },
+    {
+      "epoch": 0.8738312007874016,
+      "grad_norm": 1.3418537378311157,
+      "learning_rate": 7.762608627117352e-06,
+      "loss": 2.7593,
+      "step": 14205
+    },
+    {
+      "epoch": 0.8738927165354331,
+      "grad_norm": 1.620682954788208,
+      "learning_rate": 7.75514025543147e-06,
+      "loss": 2.5137,
+      "step": 14206
+    },
+    {
+      "epoch": 0.8739542322834646,
+      "grad_norm": 1.222316026687622,
+      "learning_rate": 7.747675333188997e-06,
+      "loss": 2.8799,
+      "step": 14207
+    },
+    {
+      "epoch": 0.8740157480314961,
+      "grad_norm": 1.7461000680923462,
+      "learning_rate": 7.740213860669032e-06,
+      "loss": 2.3534,
+      "step": 14208
+    },
+    {
+      "epoch": 0.8740772637795275,
+      "grad_norm": 1.3065959215164185,
+      "learning_rate": 7.73275583815063e-06,
+      "loss": 1.8548,
+      "step": 14209
+    },
+    {
+      "epoch": 0.874138779527559,
+      "grad_norm": 1.559866189956665,
+      "learning_rate": 7.725301265912643e-06,
+      "loss": 2.4872,
+      "step": 14210
+    },
+    {
+      "epoch": 0.8742002952755905,
+      "grad_norm": 1.4208554029464722,
+      "learning_rate": 7.717850144233852e-06,
+      "loss": 2.4384,
+      "step": 14211
+    },
+    {
+      "epoch": 0.874261811023622,
+      "grad_norm": 1.2533073425292969,
+      "learning_rate": 7.710402473392898e-06,
+      "loss": 3.2322,
+      "step": 14212
+    },
+    {
+      "epoch": 0.8743233267716536,
+      "grad_norm": 1.2931339740753174,
+      "learning_rate": 7.702958253668269e-06,
+      "loss": 2.8197,
+      "step": 14213
+    },
+    {
+      "epoch": 0.8743848425196851,
+      "grad_norm": 1.317873239517212,
+      "learning_rate": 7.695517485338333e-06,
+      "loss": 2.3312,
+      "step": 14214
+    },
+    {
+      "epoch": 0.8744463582677166,
+      "grad_norm": 1.377102017402649,
+      "learning_rate": 7.688080168681355e-06,
+      "loss": 2.4823,
+      "step": 14215
+    },
+    {
+      "epoch": 0.874507874015748,
+      "grad_norm": 1.2397310733795166,
+      "learning_rate": 7.680646303975413e-06,
+      "loss": 1.6405,
+      "step": 14216
+    },
+    {
+      "epoch": 0.8745693897637795,
+      "grad_norm": 1.2771815061569214,
+      "learning_rate": 7.67321589149852e-06,
+      "loss": 2.6226,
+      "step": 14217
+    },
+    {
+      "epoch": 0.874630905511811,
+      "grad_norm": 1.3502699136734009,
+      "learning_rate": 7.665788931528516e-06,
+      "loss": 2.6026,
+      "step": 14218
+    },
+    {
+      "epoch": 0.8746924212598425,
+      "grad_norm": 1.5712958574295044,
+      "learning_rate": 7.658365424343161e-06,
+      "loss": 2.3639,
+      "step": 14219
+    },
+    {
+      "epoch": 0.874753937007874,
+      "grad_norm": 1.3669606447219849,
+      "learning_rate": 7.650945370220008e-06,
+      "loss": 2.0367,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8748154527559056,
+      "grad_norm": 1.5274395942687988,
+      "learning_rate": 7.64352876943656e-06,
+      "loss": 2.4016,
+      "step": 14221
+    },
+    {
+      "epoch": 0.874876968503937,
+      "grad_norm": 1.4089378118515015,
+      "learning_rate": 7.636115622270135e-06,
+      "loss": 2.2848,
+      "step": 14222
+    },
+    {
+      "epoch": 0.8749384842519685,
+      "grad_norm": 1.1322544813156128,
+      "learning_rate": 7.628705928997959e-06,
+      "loss": 2.281,
+      "step": 14223
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 1.2044309377670288,
+      "learning_rate": 7.621299689897088e-06,
+      "loss": 2.4352,
+      "step": 14224
+    },
+    {
+      "epoch": 0.8750615157480315,
+      "grad_norm": 1.2480745315551758,
+      "learning_rate": 7.613896905244489e-06,
+      "loss": 2.6491,
+      "step": 14225
+    },
+    {
+      "epoch": 0.875123031496063,
+      "grad_norm": 1.2100026607513428,
+      "learning_rate": 7.6064975753169844e-06,
+      "loss": 2.5526,
+      "step": 14226
+    },
+    {
+      "epoch": 0.8751845472440944,
+      "grad_norm": 1.5959985256195068,
+      "learning_rate": 7.5991017003912975e-06,
+      "loss": 2.1446,
+      "step": 14227
+    },
+    {
+      "epoch": 0.875246062992126,
+      "grad_norm": 1.2399877309799194,
+      "learning_rate": 7.591709280743942e-06,
+      "loss": 2.1716,
+      "step": 14228
+    },
+    {
+      "epoch": 0.8753075787401575,
+      "grad_norm": 1.2864036560058594,
+      "learning_rate": 7.5843203166513946e-06,
+      "loss": 2.1931,
+      "step": 14229
+    },
+    {
+      "epoch": 0.875369094488189,
+      "grad_norm": 1.4055044651031494,
+      "learning_rate": 7.5769348083899235e-06,
+      "loss": 2.45,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8754306102362205,
+      "grad_norm": 1.3332494497299194,
+      "learning_rate": 7.5695527562357314e-06,
+      "loss": 2.6583,
+      "step": 14231
+    },
+    {
+      "epoch": 0.875492125984252,
+      "grad_norm": 1.2349852323532104,
+      "learning_rate": 7.562174160464852e-06,
+      "loss": 2.5901,
+      "step": 14232
+    },
+    {
+      "epoch": 0.8755536417322834,
+      "grad_norm": 1.3386850357055664,
+      "learning_rate": 7.554799021353231e-06,
+      "loss": 3.223,
+      "step": 14233
+    },
+    {
+      "epoch": 0.8756151574803149,
+      "grad_norm": 1.4092293977737427,
+      "learning_rate": 7.547427339176627e-06,
+      "loss": 2.9085,
+      "step": 14234
+    },
+    {
+      "epoch": 0.8756766732283464,
+      "grad_norm": 1.2027376890182495,
+      "learning_rate": 7.540059114210729e-06,
+      "loss": 2.2422,
+      "step": 14235
+    },
+    {
+      "epoch": 0.875738188976378,
+      "grad_norm": 1.4700349569320679,
+      "learning_rate": 7.532694346731029e-06,
+      "loss": 2.0722,
+      "step": 14236
+    },
+    {
+      "epoch": 0.8757997047244095,
+      "grad_norm": 1.253238558769226,
+      "learning_rate": 7.5253330370129385e-06,
+      "loss": 2.5654,
+      "step": 14237
+    },
+    {
+      "epoch": 0.875861220472441,
+      "grad_norm": 1.4013458490371704,
+      "learning_rate": 7.517975185331772e-06,
+      "loss": 2.5661,
+      "step": 14238
+    },
+    {
+      "epoch": 0.8759227362204725,
+      "grad_norm": 1.3631877899169922,
+      "learning_rate": 7.510620791962608e-06,
+      "loss": 2.3447,
+      "step": 14239
+    },
+    {
+      "epoch": 0.8759842519685039,
+      "grad_norm": 1.2340792417526245,
+      "learning_rate": 7.503269857180517e-06,
+      "loss": 2.3016,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8760457677165354,
+      "grad_norm": 1.3014453649520874,
+      "learning_rate": 7.495922381260345e-06,
+      "loss": 2.6168,
+      "step": 14241
+    },
+    {
+      "epoch": 0.8761072834645669,
+      "grad_norm": 1.2402410507202148,
+      "learning_rate": 7.488578364476872e-06,
+      "loss": 2.4676,
+      "step": 14242
+    },
+    {
+      "epoch": 0.8761687992125984,
+      "grad_norm": 1.583211064338684,
+      "learning_rate": 7.481237807104669e-06,
+      "loss": 2.1178,
+      "step": 14243
+    },
+    {
+      "epoch": 0.87623031496063,
+      "grad_norm": 1.211759328842163,
+      "learning_rate": 7.473900709418314e-06,
+      "loss": 2.6452,
+      "step": 14244
+    },
+    {
+      "epoch": 0.8762918307086615,
+      "grad_norm": 1.6390793323516846,
+      "learning_rate": 7.466567071692121e-06,
+      "loss": 3.1908,
+      "step": 14245
+    },
+    {
+      "epoch": 0.8763533464566929,
+      "grad_norm": 1.4891331195831299,
+      "learning_rate": 7.45923689420035e-06,
+      "loss": 2.6123,
+      "step": 14246
+    },
+    {
+      "epoch": 0.8764148622047244,
+      "grad_norm": 1.796600103378296,
+      "learning_rate": 7.451910177217081e-06,
+      "loss": 1.7834,
+      "step": 14247
+    },
+    {
+      "epoch": 0.8764763779527559,
+      "grad_norm": 1.2292022705078125,
+      "learning_rate": 7.4445869210163276e-06,
+      "loss": 2.1617,
+      "step": 14248
+    },
+    {
+      "epoch": 0.8765378937007874,
+      "grad_norm": 1.260959506034851,
+      "learning_rate": 7.437267125871894e-06,
+      "loss": 2.7883,
+      "step": 14249
+    },
+    {
+      "epoch": 0.8765994094488189,
+      "grad_norm": 1.3159595727920532,
+      "learning_rate": 7.429950792057538e-06,
+      "loss": 2.3512,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8766609251968503,
+      "grad_norm": 1.4605381488800049,
+      "learning_rate": 7.422637919846831e-06,
+      "loss": 3.1474,
+      "step": 14251
+    },
+    {
+      "epoch": 0.8767224409448819,
+      "grad_norm": 1.1965643167495728,
+      "learning_rate": 7.415328509513264e-06,
+      "loss": 2.5114,
+      "step": 14252
+    },
+    {
+      "epoch": 0.8767839566929134,
+      "grad_norm": 1.2324635982513428,
+      "learning_rate": 7.408022561330119e-06,
+      "loss": 2.4657,
+      "step": 14253
+    },
+    {
+      "epoch": 0.8768454724409449,
+      "grad_norm": 1.327559471130371,
+      "learning_rate": 7.400720075570644e-06,
+      "loss": 2.4157,
+      "step": 14254
+    },
+    {
+      "epoch": 0.8769069881889764,
+      "grad_norm": 1.5077461004257202,
+      "learning_rate": 7.393421052507876e-06,
+      "loss": 1.8669,
+      "step": 14255
+    },
+    {
+      "epoch": 0.8769685039370079,
+      "grad_norm": 1.3091078996658325,
+      "learning_rate": 7.386125492414764e-06,
+      "loss": 3.02,
+      "step": 14256
+    },
+    {
+      "epoch": 0.8770300196850394,
+      "grad_norm": 1.0459542274475098,
+      "learning_rate": 7.378833395564133e-06,
+      "loss": 3.1947,
+      "step": 14257
+    },
+    {
+      "epoch": 0.8770915354330708,
+      "grad_norm": 1.4276344776153564,
+      "learning_rate": 7.371544762228688e-06,
+      "loss": 1.8699,
+      "step": 14258
+    },
+    {
+      "epoch": 0.8771530511811023,
+      "grad_norm": 1.3299336433410645,
+      "learning_rate": 7.364259592680933e-06,
+      "loss": 2.5936,
+      "step": 14259
+    },
+    {
+      "epoch": 0.8772145669291339,
+      "grad_norm": 1.2765815258026123,
+      "learning_rate": 7.3569778871933395e-06,
+      "loss": 1.9279,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8772760826771654,
+      "grad_norm": 1.1550086736679077,
+      "learning_rate": 7.349699646038166e-06,
+      "loss": 2.0674,
+      "step": 14261
+    },
+    {
+      "epoch": 0.8773375984251969,
+      "grad_norm": 1.7996257543563843,
+      "learning_rate": 7.342424869487585e-06,
+      "loss": 2.4282,
+      "step": 14262
+    },
+    {
+      "epoch": 0.8773991141732284,
+      "grad_norm": 1.271180510520935,
+      "learning_rate": 7.335153557813645e-06,
+      "loss": 1.9654,
+      "step": 14263
+    },
+    {
+      "epoch": 0.8774606299212598,
+      "grad_norm": 1.2545807361602783,
+      "learning_rate": 7.327885711288273e-06,
+      "loss": 2.4046,
+      "step": 14264
+    },
+    {
+      "epoch": 0.8775221456692913,
+      "grad_norm": 1.423548936843872,
+      "learning_rate": 7.320621330183197e-06,
+      "loss": 2.2046,
+      "step": 14265
+    },
+    {
+      "epoch": 0.8775836614173228,
+      "grad_norm": 1.2393211126327515,
+      "learning_rate": 7.3133604147701096e-06,
+      "loss": 2.5957,
+      "step": 14266
+    },
+    {
+      "epoch": 0.8776451771653543,
+      "grad_norm": 1.2487374544143677,
+      "learning_rate": 7.306102965320482e-06,
+      "loss": 2.1441,
+      "step": 14267
+    },
+    {
+      "epoch": 0.8777066929133859,
+      "grad_norm": 1.4358090162277222,
+      "learning_rate": 7.298848982105755e-06,
+      "loss": 2.3838,
+      "step": 14268
+    },
+    {
+      "epoch": 0.8777682086614174,
+      "grad_norm": 1.1676946878433228,
+      "learning_rate": 7.29159846539712e-06,
+      "loss": 2.6056,
+      "step": 14269
+    },
+    {
+      "epoch": 0.8778297244094488,
+      "grad_norm": 1.4866762161254883,
+      "learning_rate": 7.284351415465784e-06,
+      "loss": 2.2508,
+      "step": 14270
+    },
+    {
+      "epoch": 0.8778912401574803,
+      "grad_norm": 1.1814361810684204,
+      "learning_rate": 7.277107832582686e-06,
+      "loss": 2.5688,
+      "step": 14271
+    },
+    {
+      "epoch": 0.8779527559055118,
+      "grad_norm": 1.4398648738861084,
+      "learning_rate": 7.269867717018742e-06,
+      "loss": 2.2404,
+      "step": 14272
+    },
+    {
+      "epoch": 0.8780142716535433,
+      "grad_norm": 1.2629632949829102,
+      "learning_rate": 7.262631069044667e-06,
+      "loss": 2.5227,
+      "step": 14273
+    },
+    {
+      "epoch": 0.8780757874015748,
+      "grad_norm": 1.4366885423660278,
+      "learning_rate": 7.2553978889310594e-06,
+      "loss": 2.1639,
+      "step": 14274
+    },
+    {
+      "epoch": 0.8781373031496063,
+      "grad_norm": 1.2542423009872437,
+      "learning_rate": 7.2481681769484e-06,
+      "loss": 2.7626,
+      "step": 14275
+    },
+    {
+      "epoch": 0.8781988188976378,
+      "grad_norm": 1.1697551012039185,
+      "learning_rate": 7.2409419333670625e-06,
+      "loss": 2.6432,
+      "step": 14276
+    },
+    {
+      "epoch": 0.8782603346456693,
+      "grad_norm": 1.3339182138442993,
+      "learning_rate": 7.233719158457275e-06,
+      "loss": 2.3131,
+      "step": 14277
+    },
+    {
+      "epoch": 0.8783218503937008,
+      "grad_norm": 1.1672865152359009,
+      "learning_rate": 7.226499852489088e-06,
+      "loss": 2.6875,
+      "step": 14278
+    },
+    {
+      "epoch": 0.8783833661417323,
+      "grad_norm": 1.267358660697937,
+      "learning_rate": 7.219284015732508e-06,
+      "loss": 2.7208,
+      "step": 14279
+    },
+    {
+      "epoch": 0.8784448818897638,
+      "grad_norm": 1.3159652948379517,
+      "learning_rate": 7.2120716484573305e-06,
+      "loss": 2.4627,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8785063976377953,
+      "grad_norm": 1.3787634372711182,
+      "learning_rate": 7.204862750933272e-06,
+      "loss": 2.3673,
+      "step": 14281
+    },
+    {
+      "epoch": 0.8785679133858267,
+      "grad_norm": 1.2633166313171387,
+      "learning_rate": 7.197657323429907e-06,
+      "loss": 2.3104,
+      "step": 14282
+    },
+    {
+      "epoch": 0.8786294291338582,
+      "grad_norm": 1.422722339630127,
+      "learning_rate": 7.190455366216697e-06,
+      "loss": 2.5984,
+      "step": 14283
+    },
+    {
+      "epoch": 0.8786909448818898,
+      "grad_norm": 1.181548833847046,
+      "learning_rate": 7.183256879562916e-06,
+      "loss": 2.8811,
+      "step": 14284
+    },
+    {
+      "epoch": 0.8787524606299213,
+      "grad_norm": 1.3228480815887451,
+      "learning_rate": 7.176061863737782e-06,
+      "loss": 2.9214,
+      "step": 14285
+    },
+    {
+      "epoch": 0.8788139763779528,
+      "grad_norm": 1.4322422742843628,
+      "learning_rate": 7.168870319010323e-06,
+      "loss": 2.7109,
+      "step": 14286
+    },
+    {
+      "epoch": 0.8788754921259843,
+      "grad_norm": 1.474165439605713,
+      "learning_rate": 7.161682245649492e-06,
+      "loss": 2.372,
+      "step": 14287
+    },
+    {
+      "epoch": 0.8789370078740157,
+      "grad_norm": 1.3044118881225586,
+      "learning_rate": 7.15449764392403e-06,
+      "loss": 1.8207,
+      "step": 14288
+    },
+    {
+      "epoch": 0.8789985236220472,
+      "grad_norm": 1.193328857421875,
+      "learning_rate": 7.147316514102664e-06,
+      "loss": 2.4578,
+      "step": 14289
+    },
+    {
+      "epoch": 0.8790600393700787,
+      "grad_norm": 1.2279224395751953,
+      "learning_rate": 7.140138856453882e-06,
+      "loss": 2.3871,
+      "step": 14290
+    },
+    {
+      "epoch": 0.8791215551181102,
+      "grad_norm": 1.3665519952774048,
+      "learning_rate": 7.132964671246123e-06,
+      "loss": 2.6925,
+      "step": 14291
+    },
+    {
+      "epoch": 0.8791830708661418,
+      "grad_norm": 1.0687668323516846,
+      "learning_rate": 7.1257939587476285e-06,
+      "loss": 2.3174,
+      "step": 14292
+    },
+    {
+      "epoch": 0.8792445866141733,
+      "grad_norm": 1.8727937936782837,
+      "learning_rate": 7.118626719226585e-06,
+      "loss": 2.3005,
+      "step": 14293
+    },
+    {
+      "epoch": 0.8793061023622047,
+      "grad_norm": 1.3458565473556519,
+      "learning_rate": 7.111462952950954e-06,
+      "loss": 2.4518,
+      "step": 14294
+    },
+    {
+      "epoch": 0.8793676181102362,
+      "grad_norm": 1.447424292564392,
+      "learning_rate": 7.104302660188655e-06,
+      "loss": 2.5878,
+      "step": 14295
+    },
+    {
+      "epoch": 0.8794291338582677,
+      "grad_norm": 1.448594570159912,
+      "learning_rate": 7.097145841207431e-06,
+      "loss": 2.4354,
+      "step": 14296
+    },
+    {
+      "epoch": 0.8794906496062992,
+      "grad_norm": 1.2685097455978394,
+      "learning_rate": 7.089992496274933e-06,
+      "loss": 2.2757,
+      "step": 14297
+    },
+    {
+      "epoch": 0.8795521653543307,
+      "grad_norm": 1.4819082021713257,
+      "learning_rate": 7.082842625658615e-06,
+      "loss": 2.8147,
+      "step": 14298
+    },
+    {
+      "epoch": 0.8796136811023622,
+      "grad_norm": 1.4581961631774902,
+      "learning_rate": 7.075696229625872e-06,
+      "loss": 2.0154,
+      "step": 14299
+    },
+    {
+      "epoch": 0.8796751968503937,
+      "grad_norm": 1.362847924232483,
+      "learning_rate": 7.068553308443926e-06,
+      "loss": 2.5046,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8797367125984252,
+      "grad_norm": 1.463780403137207,
+      "learning_rate": 7.0614138623798845e-06,
+      "loss": 2.5866,
+      "step": 14301
+    },
+    {
+      "epoch": 0.8797982283464567,
+      "grad_norm": 1.2736371755599976,
+      "learning_rate": 7.054277891700722e-06,
+      "loss": 2.4763,
+      "step": 14302
+    },
+    {
+      "epoch": 0.8798597440944882,
+      "grad_norm": 1.1836585998535156,
+      "learning_rate": 7.0471453966732935e-06,
+      "loss": 2.1462,
+      "step": 14303
+    },
+    {
+      "epoch": 0.8799212598425197,
+      "grad_norm": 1.2194602489471436,
+      "learning_rate": 7.040016377564307e-06,
+      "loss": 1.3121,
+      "step": 14304
+    },
+    {
+      "epoch": 0.8799827755905512,
+      "grad_norm": 1.419682502746582,
+      "learning_rate": 7.03289083464036e-06,
+      "loss": 2.1998,
+      "step": 14305
+    },
+    {
+      "epoch": 0.8800442913385826,
+      "grad_norm": 1.2857944965362549,
+      "learning_rate": 7.025768768167873e-06,
+      "loss": 2.5766,
+      "step": 14306
+    },
+    {
+      "epoch": 0.8801058070866141,
+      "grad_norm": 1.4315351247787476,
+      "learning_rate": 7.018650178413211e-06,
+      "loss": 2.2961,
+      "step": 14307
+    },
+    {
+      "epoch": 0.8801673228346457,
+      "grad_norm": 1.3617703914642334,
+      "learning_rate": 7.01153506564256e-06,
+      "loss": 2.2249,
+      "step": 14308
+    },
+    {
+      "epoch": 0.8802288385826772,
+      "grad_norm": 2.255680799484253,
+      "learning_rate": 7.004423430121965e-06,
+      "loss": 2.4285,
+      "step": 14309
+    },
+    {
+      "epoch": 0.8802903543307087,
+      "grad_norm": 1.355715274810791,
+      "learning_rate": 6.997315272117389e-06,
+      "loss": 1.6052,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8803518700787402,
+      "grad_norm": 1.2718427181243896,
+      "learning_rate": 6.990210591894608e-06,
+      "loss": 2.1986,
+      "step": 14311
+    },
+    {
+      "epoch": 0.8804133858267716,
+      "grad_norm": 1.09809410572052,
+      "learning_rate": 6.983109389719334e-06,
+      "loss": 3.2407,
+      "step": 14312
+    },
+    {
+      "epoch": 0.8804749015748031,
+      "grad_norm": 1.2555861473083496,
+      "learning_rate": 6.976011665857085e-06,
+      "loss": 2.2763,
+      "step": 14313
+    },
+    {
+      "epoch": 0.8805364173228346,
+      "grad_norm": 1.5596665143966675,
+      "learning_rate": 6.968917420573273e-06,
+      "loss": 2.5528,
+      "step": 14314
+    },
+    {
+      "epoch": 0.8805979330708661,
+      "grad_norm": 1.227049708366394,
+      "learning_rate": 6.961826654133196e-06,
+      "loss": 2.8003,
+      "step": 14315
+    },
+    {
+      "epoch": 0.8806594488188977,
+      "grad_norm": 1.1754910945892334,
+      "learning_rate": 6.954739366802021e-06,
+      "loss": 2.4511,
+      "step": 14316
+    },
+    {
+      "epoch": 0.8807209645669292,
+      "grad_norm": 1.2621737718582153,
+      "learning_rate": 6.9476555588447455e-06,
+      "loss": 2.593,
+      "step": 14317
+    },
+    {
+      "epoch": 0.8807824803149606,
+      "grad_norm": 1.4819873571395874,
+      "learning_rate": 6.9405752305262915e-06,
+      "loss": 1.8998,
+      "step": 14318
+    },
+    {
+      "epoch": 0.8808439960629921,
+      "grad_norm": 1.3452268838882446,
+      "learning_rate": 6.933498382111392e-06,
+      "loss": 2.4897,
+      "step": 14319
+    },
+    {
+      "epoch": 0.8809055118110236,
+      "grad_norm": 1.3651405572891235,
+      "learning_rate": 6.9264250138647125e-06,
+      "loss": 2.2243,
+      "step": 14320
+    },
+    {
+      "epoch": 0.8809670275590551,
+      "grad_norm": 1.0549097061157227,
+      "learning_rate": 6.919355126050731e-06,
+      "loss": 2.3514,
+      "step": 14321
+    },
+    {
+      "epoch": 0.8810285433070866,
+      "grad_norm": 1.5706559419631958,
+      "learning_rate": 6.912288718933857e-06,
+      "loss": 2.4582,
+      "step": 14322
+    },
+    {
+      "epoch": 0.8810900590551181,
+      "grad_norm": 1.2202489376068115,
+      "learning_rate": 6.905225792778303e-06,
+      "loss": 2.5065,
+      "step": 14323
+    },
+    {
+      "epoch": 0.8811515748031497,
+      "grad_norm": 1.1324734687805176,
+      "learning_rate": 6.8981663478482005e-06,
+      "loss": 3.0407,
+      "step": 14324
+    },
+    {
+      "epoch": 0.8812130905511811,
+      "grad_norm": 1.6679904460906982,
+      "learning_rate": 6.891110384407517e-06,
+      "loss": 2.3781,
+      "step": 14325
+    },
+    {
+      "epoch": 0.8812746062992126,
+      "grad_norm": 1.4288064241409302,
+      "learning_rate": 6.884057902720109e-06,
+      "loss": 2.1974,
+      "step": 14326
+    },
+    {
+      "epoch": 0.8813361220472441,
+      "grad_norm": 1.2382127046585083,
+      "learning_rate": 6.87700890304972e-06,
+      "loss": 2.0946,
+      "step": 14327
+    },
+    {
+      "epoch": 0.8813976377952756,
+      "grad_norm": 1.2880120277404785,
+      "learning_rate": 6.8699633856599385e-06,
+      "loss": 2.9407,
+      "step": 14328
+    },
+    {
+      "epoch": 0.8814591535433071,
+      "grad_norm": 1.1440913677215576,
+      "learning_rate": 6.8629213508142e-06,
+      "loss": 2.6521,
+      "step": 14329
+    },
+    {
+      "epoch": 0.8815206692913385,
+      "grad_norm": 1.2691240310668945,
+      "learning_rate": 6.85588279877587e-06,
+      "loss": 2.4695,
+      "step": 14330
+    },
+    {
+      "epoch": 0.88158218503937,
+      "grad_norm": 1.3318490982055664,
+      "learning_rate": 6.848847729808128e-06,
+      "loss": 2.599,
+      "step": 14331
+    },
+    {
+      "epoch": 0.8816437007874016,
+      "grad_norm": 1.2354272603988647,
+      "learning_rate": 6.841816144174051e-06,
+      "loss": 2.7655,
+      "step": 14332
+    },
+    {
+      "epoch": 0.8817052165354331,
+      "grad_norm": 1.3392033576965332,
+      "learning_rate": 6.834788042136586e-06,
+      "loss": 1.7424,
+      "step": 14333
+    },
+    {
+      "epoch": 0.8817667322834646,
+      "grad_norm": 1.2901972532272339,
+      "learning_rate": 6.827763423958566e-06,
+      "loss": 2.2428,
+      "step": 14334
+    },
+    {
+      "epoch": 0.8818282480314961,
+      "grad_norm": 1.3491712808609009,
+      "learning_rate": 6.8207422899026374e-06,
+      "loss": 2.5099,
+      "step": 14335
+    },
+    {
+      "epoch": 0.8818897637795275,
+      "grad_norm": 1.6372740268707275,
+      "learning_rate": 6.813724640231378e-06,
+      "loss": 1.8178,
+      "step": 14336
+    },
+    {
+      "epoch": 0.881951279527559,
+      "grad_norm": 1.2194948196411133,
+      "learning_rate": 6.806710475207179e-06,
+      "loss": 2.6425,
+      "step": 14337
+    },
+    {
+      "epoch": 0.8820127952755905,
+      "grad_norm": 1.4317655563354492,
+      "learning_rate": 6.799699795092363e-06,
+      "loss": 2.6102,
+      "step": 14338
+    },
+    {
+      "epoch": 0.882074311023622,
+      "grad_norm": 2.039426803588867,
+      "learning_rate": 6.792692600149053e-06,
+      "loss": 2.4028,
+      "step": 14339
+    },
+    {
+      "epoch": 0.8821358267716536,
+      "grad_norm": 1.223251461982727,
+      "learning_rate": 6.785688890639308e-06,
+      "loss": 2.7842,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8821973425196851,
+      "grad_norm": 1.199070930480957,
+      "learning_rate": 6.778688666825028e-06,
+      "loss": 2.4402,
+      "step": 14341
+    },
+    {
+      "epoch": 0.8822588582677166,
+      "grad_norm": 1.8702844381332397,
+      "learning_rate": 6.771691928967983e-06,
+      "loss": 2.8234,
+      "step": 14342
+    },
+    {
+      "epoch": 0.882320374015748,
+      "grad_norm": 1.2319902181625366,
+      "learning_rate": 6.764698677329795e-06,
+      "loss": 2.8538,
+      "step": 14343
+    },
+    {
+      "epoch": 0.8823818897637795,
+      "grad_norm": 1.2050966024398804,
+      "learning_rate": 6.757708912171989e-06,
+      "loss": 2.2919,
+      "step": 14344
+    },
+    {
+      "epoch": 0.882443405511811,
+      "grad_norm": 1.5184810161590576,
+      "learning_rate": 6.750722633755923e-06,
+      "loss": 2.5509,
+      "step": 14345
+    },
+    {
+      "epoch": 0.8825049212598425,
+      "grad_norm": 1.1830291748046875,
+      "learning_rate": 6.743739842342867e-06,
+      "loss": 3.0199,
+      "step": 14346
+    },
+    {
+      "epoch": 0.882566437007874,
+      "grad_norm": 1.1832629442214966,
+      "learning_rate": 6.736760538193943e-06,
+      "loss": 2.189,
+      "step": 14347
+    },
+    {
+      "epoch": 0.8826279527559056,
+      "grad_norm": 1.5826785564422607,
+      "learning_rate": 6.729784721570098e-06,
+      "loss": 3.0086,
+      "step": 14348
+    },
+    {
+      "epoch": 0.882689468503937,
+      "grad_norm": 1.8101909160614014,
+      "learning_rate": 6.722812392732248e-06,
+      "loss": 2.4152,
+      "step": 14349
+    },
+    {
+      "epoch": 0.8827509842519685,
+      "grad_norm": 1.1685924530029297,
+      "learning_rate": 6.715843551941059e-06,
+      "loss": 2.3889,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8828125,
+      "grad_norm": 1.547708511352539,
+      "learning_rate": 6.708878199457158e-06,
+      "loss": 2.5972,
+      "step": 14351
+    },
+    {
+      "epoch": 0.8828740157480315,
+      "grad_norm": 1.2326407432556152,
+      "learning_rate": 6.701916335541003e-06,
+      "loss": 2.9657,
+      "step": 14352
+    },
+    {
+      "epoch": 0.882935531496063,
+      "grad_norm": 1.1623655557632446,
+      "learning_rate": 6.694957960452952e-06,
+      "loss": 3.0,
+      "step": 14353
+    },
+    {
+      "epoch": 0.8829970472440944,
+      "grad_norm": 1.5128180980682373,
+      "learning_rate": 6.6880030744531754e-06,
+      "loss": 2.5015,
+      "step": 14354
+    },
+    {
+      "epoch": 0.883058562992126,
+      "grad_norm": 1.1304162740707397,
+      "learning_rate": 6.681051677801775e-06,
+      "loss": 2.6497,
+      "step": 14355
+    },
+    {
+      "epoch": 0.8831200787401575,
+      "grad_norm": 1.3304351568222046,
+      "learning_rate": 6.674103770758666e-06,
+      "loss": 2.2904,
+      "step": 14356
+    },
+    {
+      "epoch": 0.883181594488189,
+      "grad_norm": 1.275112509727478,
+      "learning_rate": 6.667159353583696e-06,
+      "loss": 1.941,
+      "step": 14357
+    },
+    {
+      "epoch": 0.8832431102362205,
+      "grad_norm": 1.331048846244812,
+      "learning_rate": 6.660218426536491e-06,
+      "loss": 2.33,
+      "step": 14358
+    },
+    {
+      "epoch": 0.883304625984252,
+      "grad_norm": 1.6630622148513794,
+      "learning_rate": 6.6532809898766756e-06,
+      "loss": 2.5021,
+      "step": 14359
+    },
+    {
+      "epoch": 0.8833661417322834,
+      "grad_norm": 1.2661693096160889,
+      "learning_rate": 6.646347043863621e-06,
+      "loss": 2.5768,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8834276574803149,
+      "grad_norm": 1.203163743019104,
+      "learning_rate": 6.639416588756642e-06,
+      "loss": 2.6538,
+      "step": 14361
+    },
+    {
+      "epoch": 0.8834891732283464,
+      "grad_norm": 1.3806143999099731,
+      "learning_rate": 6.6324896248148746e-06,
+      "loss": 2.5428,
+      "step": 14362
+    },
+    {
+      "epoch": 0.883550688976378,
+      "grad_norm": 1.3749902248382568,
+      "learning_rate": 6.62556615229738e-06,
+      "loss": 2.5519,
+      "step": 14363
+    },
+    {
+      "epoch": 0.8836122047244095,
+      "grad_norm": 1.5285485982894897,
+      "learning_rate": 6.6186461714630275e-06,
+      "loss": 2.4852,
+      "step": 14364
+    },
+    {
+      "epoch": 0.883673720472441,
+      "grad_norm": 1.7190754413604736,
+      "learning_rate": 6.6117296825706e-06,
+      "loss": 2.7076,
+      "step": 14365
+    },
+    {
+      "epoch": 0.8837352362204725,
+      "grad_norm": 1.4488574266433716,
+      "learning_rate": 6.604816685878734e-06,
+      "loss": 2.1052,
+      "step": 14366
+    },
+    {
+      "epoch": 0.8837967519685039,
+      "grad_norm": 1.441824197769165,
+      "learning_rate": 6.597907181645957e-06,
+      "loss": 2.2998,
+      "step": 14367
+    },
+    {
+      "epoch": 0.8838582677165354,
+      "grad_norm": 1.1734803915023804,
+      "learning_rate": 6.5910011701306064e-06,
+      "loss": 2.7159,
+      "step": 14368
+    },
+    {
+      "epoch": 0.8839197834645669,
+      "grad_norm": 1.2108895778656006,
+      "learning_rate": 6.584098651590975e-06,
+      "loss": 2.6145,
+      "step": 14369
+    },
+    {
+      "epoch": 0.8839812992125984,
+      "grad_norm": 1.3722084760665894,
+      "learning_rate": 6.5771996262851244e-06,
+      "loss": 2.4919,
+      "step": 14370
+    },
+    {
+      "epoch": 0.88404281496063,
+      "grad_norm": 1.349402904510498,
+      "learning_rate": 6.570304094471069e-06,
+      "loss": 1.8204,
+      "step": 14371
+    },
+    {
+      "epoch": 0.8841043307086615,
+      "grad_norm": 1.341759204864502,
+      "learning_rate": 6.56341205640667e-06,
+      "loss": 2.5669,
+      "step": 14372
+    },
+    {
+      "epoch": 0.8841658464566929,
+      "grad_norm": 1.4156715869903564,
+      "learning_rate": 6.556523512349655e-06,
+      "loss": 2.6453,
+      "step": 14373
+    },
+    {
+      "epoch": 0.8842273622047244,
+      "grad_norm": 1.5092707872390747,
+      "learning_rate": 6.549638462557595e-06,
+      "loss": 2.5618,
+      "step": 14374
+    },
+    {
+      "epoch": 0.8842888779527559,
+      "grad_norm": 1.4039231538772583,
+      "learning_rate": 6.5427569072879725e-06,
+      "loss": 1.8253,
+      "step": 14375
+    },
+    {
+      "epoch": 0.8843503937007874,
+      "grad_norm": 1.3395329713821411,
+      "learning_rate": 6.535878846798093e-06,
+      "loss": 2.1823,
+      "step": 14376
+    },
+    {
+      "epoch": 0.8844119094488189,
+      "grad_norm": 1.2661007642745972,
+      "learning_rate": 6.529004281345175e-06,
+      "loss": 3.0053,
+      "step": 14377
+    },
+    {
+      "epoch": 0.8844734251968503,
+      "grad_norm": 1.3010724782943726,
+      "learning_rate": 6.522133211186288e-06,
+      "loss": 2.0759,
+      "step": 14378
+    },
+    {
+      "epoch": 0.8845349409448819,
+      "grad_norm": 1.2456632852554321,
+      "learning_rate": 6.5152656365783935e-06,
+      "loss": 2.535,
+      "step": 14379
+    },
+    {
+      "epoch": 0.8845964566929134,
+      "grad_norm": 1.344762921333313,
+      "learning_rate": 6.5084015577782656e-06,
+      "loss": 2.3432,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8846579724409449,
+      "grad_norm": 1.2149444818496704,
+      "learning_rate": 6.501540975042597e-06,
+      "loss": 2.7853,
+      "step": 14381
+    },
+    {
+      "epoch": 0.8847194881889764,
+      "grad_norm": 1.27865469455719,
+      "learning_rate": 6.49468388862795e-06,
+      "loss": 2.0226,
+      "step": 14382
+    },
+    {
+      "epoch": 0.8847810039370079,
+      "grad_norm": 1.4306972026824951,
+      "learning_rate": 6.487830298790698e-06,
+      "loss": 1.9357,
+      "step": 14383
+    },
+    {
+      "epoch": 0.8848425196850394,
+      "grad_norm": 1.0283515453338623,
+      "learning_rate": 6.480980205787157e-06,
+      "loss": 2.8178,
+      "step": 14384
+    },
+    {
+      "epoch": 0.8849040354330708,
+      "grad_norm": 1.4829102754592896,
+      "learning_rate": 6.47413360987349e-06,
+      "loss": 2.6865,
+      "step": 14385
+    },
+    {
+      "epoch": 0.8849655511811023,
+      "grad_norm": 1.2583624124526978,
+      "learning_rate": 6.467290511305713e-06,
+      "loss": 2.2324,
+      "step": 14386
+    },
+    {
+      "epoch": 0.8850270669291339,
+      "grad_norm": 1.3076798915863037,
+      "learning_rate": 6.4604509103397104e-06,
+      "loss": 2.6436,
+      "step": 14387
+    },
+    {
+      "epoch": 0.8850885826771654,
+      "grad_norm": 1.2008742094039917,
+      "learning_rate": 6.4536148072312675e-06,
+      "loss": 2.8248,
+      "step": 14388
+    },
+    {
+      "epoch": 0.8851500984251969,
+      "grad_norm": 1.3841599225997925,
+      "learning_rate": 6.4467822022359904e-06,
+      "loss": 2.5491,
+      "step": 14389
+    },
+    {
+      "epoch": 0.8852116141732284,
+      "grad_norm": 1.4213135242462158,
+      "learning_rate": 6.439953095609386e-06,
+      "loss": 2.6755,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8852731299212598,
+      "grad_norm": 1.5618133544921875,
+      "learning_rate": 6.433127487606827e-06,
+      "loss": 2.2902,
+      "step": 14391
+    },
+    {
+      "epoch": 0.8853346456692913,
+      "grad_norm": 1.261317491531372,
+      "learning_rate": 6.426305378483577e-06,
+      "loss": 2.7797,
+      "step": 14392
+    },
+    {
+      "epoch": 0.8853961614173228,
+      "grad_norm": 2.180544137954712,
+      "learning_rate": 6.419486768494709e-06,
+      "loss": 2.0911,
+      "step": 14393
+    },
+    {
+      "epoch": 0.8854576771653543,
+      "grad_norm": 1.3141839504241943,
+      "learning_rate": 6.4126716578952414e-06,
+      "loss": 2.4566,
+      "step": 14394
+    },
+    {
+      "epoch": 0.8855191929133859,
+      "grad_norm": 1.2465360164642334,
+      "learning_rate": 6.405860046939971e-06,
+      "loss": 2.4594,
+      "step": 14395
+    },
+    {
+      "epoch": 0.8855807086614174,
+      "grad_norm": 1.3154994249343872,
+      "learning_rate": 6.3990519358836376e-06,
+      "loss": 2.1203,
+      "step": 14396
+    },
+    {
+      "epoch": 0.8856422244094488,
+      "grad_norm": 1.2480666637420654,
+      "learning_rate": 6.3922473249808266e-06,
+      "loss": 2.7559,
+      "step": 14397
+    },
+    {
+      "epoch": 0.8857037401574803,
+      "grad_norm": 1.2520732879638672,
+      "learning_rate": 6.385446214486024e-06,
+      "loss": 2.7123,
+      "step": 14398
+    },
+    {
+      "epoch": 0.8857652559055118,
+      "grad_norm": 1.27925443649292,
+      "learning_rate": 6.378648604653492e-06,
+      "loss": 2.2627,
+      "step": 14399
+    },
+    {
+      "epoch": 0.8858267716535433,
+      "grad_norm": 1.4419338703155518,
+      "learning_rate": 6.371854495737483e-06,
+      "loss": 2.5025,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8858882874015748,
+      "grad_norm": 1.2551053762435913,
+      "learning_rate": 6.365063887992006e-06,
+      "loss": 2.5842,
+      "step": 14401
+    },
+    {
+      "epoch": 0.8859498031496063,
+      "grad_norm": 1.2079284191131592,
+      "learning_rate": 6.358276781671047e-06,
+      "loss": 2.8827,
+      "step": 14402
+    },
+    {
+      "epoch": 0.8860113188976378,
+      "grad_norm": 1.2867610454559326,
+      "learning_rate": 6.3514931770283335e-06,
+      "loss": 2.3998,
+      "step": 14403
+    },
+    {
+      "epoch": 0.8860728346456693,
+      "grad_norm": 1.112509846687317,
+      "learning_rate": 6.34471307431761e-06,
+      "loss": 2.7458,
+      "step": 14404
+    },
+    {
+      "epoch": 0.8861343503937008,
+      "grad_norm": 1.3085359334945679,
+      "learning_rate": 6.337936473792361e-06,
+      "loss": 2.9998,
+      "step": 14405
+    },
+    {
+      "epoch": 0.8861958661417323,
+      "grad_norm": 1.384928822517395,
+      "learning_rate": 6.3311633757060395e-06,
+      "loss": 2.7273,
+      "step": 14406
+    },
+    {
+      "epoch": 0.8862573818897638,
+      "grad_norm": 1.3721575736999512,
+      "learning_rate": 6.3243937803118655e-06,
+      "loss": 2.8529,
+      "step": 14407
+    },
+    {
+      "epoch": 0.8863188976377953,
+      "grad_norm": 1.0989842414855957,
+      "learning_rate": 6.317627687863048e-06,
+      "loss": 2.494,
+      "step": 14408
+    },
+    {
+      "epoch": 0.8863804133858267,
+      "grad_norm": 1.340013861656189,
+      "learning_rate": 6.31086509861254e-06,
+      "loss": 2.1261,
+      "step": 14409
+    },
+    {
+      "epoch": 0.8864419291338582,
+      "grad_norm": 1.4868924617767334,
+      "learning_rate": 6.30410601281326e-06,
+      "loss": 2.2397,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8865034448818898,
+      "grad_norm": 1.2514797449111938,
+      "learning_rate": 6.297350430717941e-06,
+      "loss": 2.4965,
+      "step": 14411
+    },
+    {
+      "epoch": 0.8865649606299213,
+      "grad_norm": 1.762535810470581,
+      "learning_rate": 6.290598352579235e-06,
+      "loss": 2.6794,
+      "step": 14412
+    },
+    {
+      "epoch": 0.8866264763779528,
+      "grad_norm": 1.4000494480133057,
+      "learning_rate": 6.283849778649609e-06,
+      "loss": 3.1242,
+      "step": 14413
+    },
+    {
+      "epoch": 0.8866879921259843,
+      "grad_norm": 1.2120771408081055,
+      "learning_rate": 6.277104709181425e-06,
+      "loss": 3.2919,
+      "step": 14414
+    },
+    {
+      "epoch": 0.8867495078740157,
+      "grad_norm": 1.380929946899414,
+      "learning_rate": 6.270363144426905e-06,
+      "loss": 2.4219,
+      "step": 14415
+    },
+    {
+      "epoch": 0.8868110236220472,
+      "grad_norm": 1.5676155090332031,
+      "learning_rate": 6.263625084638158e-06,
+      "loss": 2.6088,
+      "step": 14416
+    },
+    {
+      "epoch": 0.8868725393700787,
+      "grad_norm": 1.3585619926452637,
+      "learning_rate": 6.256890530067139e-06,
+      "loss": 2.3714,
+      "step": 14417
+    },
+    {
+      "epoch": 0.8869340551181102,
+      "grad_norm": 1.2827287912368774,
+      "learning_rate": 6.250159480965701e-06,
+      "loss": 3.1442,
+      "step": 14418
+    },
+    {
+      "epoch": 0.8869955708661418,
+      "grad_norm": 1.1738455295562744,
+      "learning_rate": 6.243431937585531e-06,
+      "loss": 2.6572,
+      "step": 14419
+    },
+    {
+      "epoch": 0.8870570866141733,
+      "grad_norm": 1.3140966892242432,
+      "learning_rate": 6.236707900178196e-06,
+      "loss": 2.2451,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8871186023622047,
+      "grad_norm": 1.4515926837921143,
+      "learning_rate": 6.229987368995149e-06,
+      "loss": 2.818,
+      "step": 14421
+    },
+    {
+      "epoch": 0.8871801181102362,
+      "grad_norm": 1.4564028978347778,
+      "learning_rate": 6.223270344287702e-06,
+      "loss": 2.496,
+      "step": 14422
+    },
+    {
+      "epoch": 0.8872416338582677,
+      "grad_norm": 1.4078413248062134,
+      "learning_rate": 6.216556826307041e-06,
+      "loss": 2.6445,
+      "step": 14423
+    },
+    {
+      "epoch": 0.8873031496062992,
+      "grad_norm": 2.3313748836517334,
+      "learning_rate": 6.209846815304188e-06,
+      "loss": 2.4722,
+      "step": 14424
+    },
+    {
+      "epoch": 0.8873646653543307,
+      "grad_norm": 1.1446809768676758,
+      "learning_rate": 6.203140311530087e-06,
+      "loss": 2.8298,
+      "step": 14425
+    },
+    {
+      "epoch": 0.8874261811023622,
+      "grad_norm": 1.3410069942474365,
+      "learning_rate": 6.1964373152355036e-06,
+      "loss": 2.3442,
+      "step": 14426
+    },
+    {
+      "epoch": 0.8874876968503937,
+      "grad_norm": 1.4469988346099854,
+      "learning_rate": 6.189737826671116e-06,
+      "loss": 3.1293,
+      "step": 14427
+    },
+    {
+      "epoch": 0.8875492125984252,
+      "grad_norm": 1.1984758377075195,
+      "learning_rate": 6.1830418460874116e-06,
+      "loss": 2.3769,
+      "step": 14428
+    },
+    {
+      "epoch": 0.8876107283464567,
+      "grad_norm": 1.262671709060669,
+      "learning_rate": 6.176349373734813e-06,
+      "loss": 2.199,
+      "step": 14429
+    },
+    {
+      "epoch": 0.8876722440944882,
+      "grad_norm": 1.279427409172058,
+      "learning_rate": 6.169660409863576e-06,
+      "loss": 2.0868,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8877337598425197,
+      "grad_norm": 1.242490530014038,
+      "learning_rate": 6.1629749547238325e-06,
+      "loss": 2.0744,
+      "step": 14431
+    },
+    {
+      "epoch": 0.8877952755905512,
+      "grad_norm": 1.3096176385879517,
+      "learning_rate": 6.156293008565561e-06,
+      "loss": 2.282,
+      "step": 14432
+    },
+    {
+      "epoch": 0.8878567913385826,
+      "grad_norm": 1.4127908945083618,
+      "learning_rate": 6.149614571638662e-06,
+      "loss": 2.743,
+      "step": 14433
+    },
+    {
+      "epoch": 0.8879183070866141,
+      "grad_norm": 1.3221334218978882,
+      "learning_rate": 6.142939644192847e-06,
+      "loss": 2.5368,
+      "step": 14434
+    },
+    {
+      "epoch": 0.8879798228346457,
+      "grad_norm": 1.2131885290145874,
+      "learning_rate": 6.136268226477715e-06,
+      "loss": 2.6871,
+      "step": 14435
+    },
+    {
+      "epoch": 0.8880413385826772,
+      "grad_norm": 1.8299199342727661,
+      "learning_rate": 6.129600318742757e-06,
+      "loss": 3.073,
+      "step": 14436
+    },
+    {
+      "epoch": 0.8881028543307087,
+      "grad_norm": 1.2436434030532837,
+      "learning_rate": 6.122935921237327e-06,
+      "loss": 2.3358,
+      "step": 14437
+    },
+    {
+      "epoch": 0.8881643700787402,
+      "grad_norm": 1.2901175022125244,
+      "learning_rate": 6.1162750342106055e-06,
+      "loss": 2.1728,
+      "step": 14438
+    },
+    {
+      "epoch": 0.8882258858267716,
+      "grad_norm": 1.612849235534668,
+      "learning_rate": 6.109617657911704e-06,
+      "loss": 2.1944,
+      "step": 14439
+    },
+    {
+      "epoch": 0.8882874015748031,
+      "grad_norm": 1.3614344596862793,
+      "learning_rate": 6.102963792589544e-06,
+      "loss": 2.2594,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8883489173228346,
+      "grad_norm": 1.3904043436050415,
+      "learning_rate": 6.096313438492951e-06,
+      "loss": 2.2456,
+      "step": 14441
+    },
+    {
+      "epoch": 0.8884104330708661,
+      "grad_norm": 1.0774749517440796,
+      "learning_rate": 6.089666595870613e-06,
+      "loss": 2.7754,
+      "step": 14442
+    },
+    {
+      "epoch": 0.8884719488188977,
+      "grad_norm": 1.1611465215682983,
+      "learning_rate": 6.083023264971099e-06,
+      "loss": 1.3509,
+      "step": 14443
+    },
+    {
+      "epoch": 0.8885334645669292,
+      "grad_norm": 1.637982726097107,
+      "learning_rate": 6.07638344604281e-06,
+      "loss": 1.3077,
+      "step": 14444
+    },
+    {
+      "epoch": 0.8885949803149606,
+      "grad_norm": 1.2563163042068481,
+      "learning_rate": 6.0697471393340585e-06,
+      "loss": 2.2595,
+      "step": 14445
+    },
+    {
+      "epoch": 0.8886564960629921,
+      "grad_norm": 1.2826125621795654,
+      "learning_rate": 6.063114345092979e-06,
+      "loss": 3.164,
+      "step": 14446
+    },
+    {
+      "epoch": 0.8887180118110236,
+      "grad_norm": 1.359391689300537,
+      "learning_rate": 6.056485063567619e-06,
+      "loss": 2.3292,
+      "step": 14447
+    },
+    {
+      "epoch": 0.8887795275590551,
+      "grad_norm": 1.1270469427108765,
+      "learning_rate": 6.049859295005866e-06,
+      "loss": 2.4779,
+      "step": 14448
+    },
+    {
+      "epoch": 0.8888410433070866,
+      "grad_norm": 1.328676462173462,
+      "learning_rate": 6.043237039655525e-06,
+      "loss": 2.5433,
+      "step": 14449
+    },
+    {
+      "epoch": 0.8889025590551181,
+      "grad_norm": 1.3413052558898926,
+      "learning_rate": 6.036618297764174e-06,
+      "loss": 1.7352,
+      "step": 14450
+    },
+    {
+      "epoch": 0.8889640748031497,
+      "grad_norm": 1.119203805923462,
+      "learning_rate": 6.030003069579359e-06,
+      "loss": 1.6469,
+      "step": 14451
+    },
+    {
+      "epoch": 0.8890255905511811,
+      "grad_norm": 1.376248836517334,
+      "learning_rate": 6.0233913553484286e-06,
+      "loss": 2.3228,
+      "step": 14452
+    },
+    {
+      "epoch": 0.8890871062992126,
+      "grad_norm": 1.198083519935608,
+      "learning_rate": 6.01678315531865e-06,
+      "loss": 2.2096,
+      "step": 14453
+    },
+    {
+      "epoch": 0.8891486220472441,
+      "grad_norm": 1.2403004169464111,
+      "learning_rate": 6.010178469737104e-06,
+      "loss": 2.3337,
+      "step": 14454
+    },
+    {
+      "epoch": 0.8892101377952756,
+      "grad_norm": 1.2248462438583374,
+      "learning_rate": 6.003577298850771e-06,
+      "loss": 2.3738,
+      "step": 14455
+    },
+    {
+      "epoch": 0.8892716535433071,
+      "grad_norm": 1.3084514141082764,
+      "learning_rate": 5.99697964290653e-06,
+      "loss": 2.7075,
+      "step": 14456
+    },
+    {
+      "epoch": 0.8893331692913385,
+      "grad_norm": 1.1439790725708008,
+      "learning_rate": 5.990385502151064e-06,
+      "loss": 2.4696,
+      "step": 14457
+    },
+    {
+      "epoch": 0.88939468503937,
+      "grad_norm": 1.2253504991531372,
+      "learning_rate": 5.983794876830973e-06,
+      "loss": 2.0875,
+      "step": 14458
+    },
+    {
+      "epoch": 0.8894562007874016,
+      "grad_norm": 1.4060022830963135,
+      "learning_rate": 5.977207767192705e-06,
+      "loss": 2.2741,
+      "step": 14459
+    },
+    {
+      "epoch": 0.8895177165354331,
+      "grad_norm": 1.445488452911377,
+      "learning_rate": 5.970624173482564e-06,
+      "loss": 2.6464,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8895792322834646,
+      "grad_norm": 1.2058241367340088,
+      "learning_rate": 5.964044095946764e-06,
+      "loss": 2.1613,
+      "step": 14461
+    },
+    {
+      "epoch": 0.8896407480314961,
+      "grad_norm": 1.2410424947738647,
+      "learning_rate": 5.957467534831373e-06,
+      "loss": 2.7111,
+      "step": 14462
+    },
+    {
+      "epoch": 0.8897022637795275,
+      "grad_norm": 1.2024637460708618,
+      "learning_rate": 5.9508944903822865e-06,
+      "loss": 2.3448,
+      "step": 14463
+    },
+    {
+      "epoch": 0.889763779527559,
+      "grad_norm": 1.7883799076080322,
+      "learning_rate": 5.944324962845327e-06,
+      "loss": 2.0614,
+      "step": 14464
+    },
+    {
+      "epoch": 0.8898252952755905,
+      "grad_norm": 1.2505123615264893,
+      "learning_rate": 5.937758952466133e-06,
+      "loss": 2.3049,
+      "step": 14465
+    },
+    {
+      "epoch": 0.889886811023622,
+      "grad_norm": 1.2728313207626343,
+      "learning_rate": 5.931196459490241e-06,
+      "loss": 2.2713,
+      "step": 14466
+    },
+    {
+      "epoch": 0.8899483267716536,
+      "grad_norm": 1.231302261352539,
+      "learning_rate": 5.924637484163065e-06,
+      "loss": 1.5406,
+      "step": 14467
+    },
+    {
+      "epoch": 0.8900098425196851,
+      "grad_norm": 1.1399306058883667,
+      "learning_rate": 5.918082026729876e-06,
+      "loss": 2.7773,
+      "step": 14468
+    },
+    {
+      "epoch": 0.8900713582677166,
+      "grad_norm": 1.412088394165039,
+      "learning_rate": 5.9115300874358015e-06,
+      "loss": 2.1758,
+      "step": 14469
+    },
+    {
+      "epoch": 0.890132874015748,
+      "grad_norm": 1.4227807521820068,
+      "learning_rate": 5.904981666525855e-06,
+      "loss": 2.9486,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8901943897637795,
+      "grad_norm": 1.21261727809906,
+      "learning_rate": 5.898436764244897e-06,
+      "loss": 1.8286,
+      "step": 14471
+    },
+    {
+      "epoch": 0.890255905511811,
+      "grad_norm": 1.5869179964065552,
+      "learning_rate": 5.891895380837698e-06,
+      "loss": 2.151,
+      "step": 14472
+    },
+    {
+      "epoch": 0.8903174212598425,
+      "grad_norm": 1.482186198234558,
+      "learning_rate": 5.885357516548806e-06,
+      "loss": 2.513,
+      "step": 14473
+    },
+    {
+      "epoch": 0.890378937007874,
+      "grad_norm": 1.4285998344421387,
+      "learning_rate": 5.8788231716227825e-06,
+      "loss": 2.0907,
+      "step": 14474
+    },
+    {
+      "epoch": 0.8904404527559056,
+      "grad_norm": 1.7347558736801147,
+      "learning_rate": 5.872292346303931e-06,
+      "loss": 2.1786,
+      "step": 14475
+    },
+    {
+      "epoch": 0.890501968503937,
+      "grad_norm": 1.172324538230896,
+      "learning_rate": 5.86576504083648e-06,
+      "loss": 2.2347,
+      "step": 14476
+    },
+    {
+      "epoch": 0.8905634842519685,
+      "grad_norm": 1.3206031322479248,
+      "learning_rate": 5.8592412554644985e-06,
+      "loss": 2.4812,
+      "step": 14477
+    },
+    {
+      "epoch": 0.890625,
+      "grad_norm": 1.522623062133789,
+      "learning_rate": 5.85272099043197e-06,
+      "loss": 2.6107,
+      "step": 14478
+    },
+    {
+      "epoch": 0.8906865157480315,
+      "grad_norm": 1.2759383916854858,
+      "learning_rate": 5.8462042459826786e-06,
+      "loss": 2.0872,
+      "step": 14479
+    },
+    {
+      "epoch": 0.890748031496063,
+      "grad_norm": 1.4432759284973145,
+      "learning_rate": 5.839691022360327e-06,
+      "loss": 2.3167,
+      "step": 14480
+    },
+    {
+      "epoch": 0.8908095472440944,
+      "grad_norm": 1.2708547115325928,
+      "learning_rate": 5.8331813198084896e-06,
+      "loss": 1.9321,
+      "step": 14481
+    },
+    {
+      "epoch": 0.890871062992126,
+      "grad_norm": 1.3248392343521118,
+      "learning_rate": 5.826675138570581e-06,
+      "loss": 2.3059,
+      "step": 14482
+    },
+    {
+      "epoch": 0.8909325787401575,
+      "grad_norm": 1.967054843902588,
+      "learning_rate": 5.820172478889896e-06,
+      "loss": 2.3589,
+      "step": 14483
+    },
+    {
+      "epoch": 0.890994094488189,
+      "grad_norm": 1.217212438583374,
+      "learning_rate": 5.813673341009618e-06,
+      "loss": 2.4282,
+      "step": 14484
+    },
+    {
+      "epoch": 0.8910556102362205,
+      "grad_norm": 1.2188317775726318,
+      "learning_rate": 5.80717772517273e-06,
+      "loss": 1.9919,
+      "step": 14485
+    },
+    {
+      "epoch": 0.891117125984252,
+      "grad_norm": 1.2751448154449463,
+      "learning_rate": 5.800685631622171e-06,
+      "loss": 2.1736,
+      "step": 14486
+    },
+    {
+      "epoch": 0.8911786417322834,
+      "grad_norm": 1.3072350025177002,
+      "learning_rate": 5.794197060600692e-06,
+      "loss": 2.2338,
+      "step": 14487
+    },
+    {
+      "epoch": 0.8912401574803149,
+      "grad_norm": 1.417542576789856,
+      "learning_rate": 5.787712012350965e-06,
+      "loss": 2.4001,
+      "step": 14488
+    },
+    {
+      "epoch": 0.8913016732283464,
+      "grad_norm": 1.4527872800827026,
+      "learning_rate": 5.781230487115441e-06,
+      "loss": 1.9582,
+      "step": 14489
+    },
+    {
+      "epoch": 0.891363188976378,
+      "grad_norm": 1.216389536857605,
+      "learning_rate": 5.774752485136536e-06,
+      "loss": 2.6924,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8914247047244095,
+      "grad_norm": 1.370866298675537,
+      "learning_rate": 5.768278006656458e-06,
+      "loss": 2.1547,
+      "step": 14491
+    },
+    {
+      "epoch": 0.891486220472441,
+      "grad_norm": 1.4205741882324219,
+      "learning_rate": 5.761807051917334e-06,
+      "loss": 3.0339,
+      "step": 14492
+    },
+    {
+      "epoch": 0.8915477362204725,
+      "grad_norm": 1.769304633140564,
+      "learning_rate": 5.75533962116116e-06,
+      "loss": 2.5601,
+      "step": 14493
+    },
+    {
+      "epoch": 0.8916092519685039,
+      "grad_norm": 1.2722500562667847,
+      "learning_rate": 5.748875714629731e-06,
+      "loss": 2.4279,
+      "step": 14494
+    },
+    {
+      "epoch": 0.8916707677165354,
+      "grad_norm": 1.5763546228408813,
+      "learning_rate": 5.7424153325648215e-06,
+      "loss": 3.2236,
+      "step": 14495
+    },
+    {
+      "epoch": 0.8917322834645669,
+      "grad_norm": 1.3652963638305664,
+      "learning_rate": 5.7359584752079585e-06,
+      "loss": 2.226,
+      "step": 14496
+    },
+    {
+      "epoch": 0.8917937992125984,
+      "grad_norm": 1.434954047203064,
+      "learning_rate": 5.7295051428006394e-06,
+      "loss": 1.2564,
+      "step": 14497
+    },
+    {
+      "epoch": 0.89185531496063,
+      "grad_norm": 1.3128414154052734,
+      "learning_rate": 5.723055335584148e-06,
+      "loss": 2.3476,
+      "step": 14498
+    },
+    {
+      "epoch": 0.8919168307086615,
+      "grad_norm": 1.379725456237793,
+      "learning_rate": 5.716609053799682e-06,
+      "loss": 2.4317,
+      "step": 14499
+    },
+    {
+      "epoch": 0.8919783464566929,
+      "grad_norm": 1.3706448078155518,
+      "learning_rate": 5.7101662976882905e-06,
+      "loss": 2.3409,
+      "step": 14500
+    },
+    {
+      "epoch": 0.8920398622047244,
+      "grad_norm": 1.3684409856796265,
+      "learning_rate": 5.703727067490927e-06,
+      "loss": 2.6323,
+      "step": 14501
+    },
+    {
+      "epoch": 0.8921013779527559,
+      "grad_norm": 1.4222509860992432,
+      "learning_rate": 5.697291363448354e-06,
+      "loss": 2.2151,
+      "step": 14502
+    },
+    {
+      "epoch": 0.8921628937007874,
+      "grad_norm": 1.1070902347564697,
+      "learning_rate": 5.690859185801245e-06,
+      "loss": 2.7961,
+      "step": 14503
+    },
+    {
+      "epoch": 0.8922244094488189,
+      "grad_norm": 1.4439001083374023,
+      "learning_rate": 5.684430534790097e-06,
+      "loss": 2.7001,
+      "step": 14504
+    },
+    {
+      "epoch": 0.8922859251968503,
+      "grad_norm": 1.198531985282898,
+      "learning_rate": 5.6780054106553405e-06,
+      "loss": 2.7693,
+      "step": 14505
+    },
+    {
+      "epoch": 0.8923474409448819,
+      "grad_norm": 1.2593134641647339,
+      "learning_rate": 5.671583813637227e-06,
+      "loss": 2.0522,
+      "step": 14506
+    },
+    {
+      "epoch": 0.8924089566929134,
+      "grad_norm": 1.2180583477020264,
+      "learning_rate": 5.6651657439758975e-06,
+      "loss": 2.7159,
+      "step": 14507
+    },
+    {
+      "epoch": 0.8924704724409449,
+      "grad_norm": 1.4328283071517944,
+      "learning_rate": 5.658751201911338e-06,
+      "loss": 2.1025,
+      "step": 14508
+    },
+    {
+      "epoch": 0.8925319881889764,
+      "grad_norm": 1.4781503677368164,
+      "learning_rate": 5.6523401876834246e-06,
+      "loss": 3.6739,
+      "step": 14509
+    },
+    {
+      "epoch": 0.8925935039370079,
+      "grad_norm": 3.0380613803863525,
+      "learning_rate": 5.645932701531886e-06,
+      "loss": 2.475,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8926550196850394,
+      "grad_norm": 1.4009969234466553,
+      "learning_rate": 5.639528743696332e-06,
+      "loss": 0.765,
+      "step": 14511
+    },
+    {
+      "epoch": 0.8927165354330708,
+      "grad_norm": 1.3252174854278564,
+      "learning_rate": 5.633128314416225e-06,
+      "loss": 2.0233,
+      "step": 14512
+    },
+    {
+      "epoch": 0.8927780511811023,
+      "grad_norm": 1.0751221179962158,
+      "learning_rate": 5.626731413930941e-06,
+      "loss": 2.5875,
+      "step": 14513
+    },
+    {
+      "epoch": 0.8928395669291339,
+      "grad_norm": 1.416378140449524,
+      "learning_rate": 5.620338042479645e-06,
+      "loss": 2.2592,
+      "step": 14514
+    },
+    {
+      "epoch": 0.8929010826771654,
+      "grad_norm": 1.1805192232131958,
+      "learning_rate": 5.613948200301445e-06,
+      "loss": 2.5961,
+      "step": 14515
+    },
+    {
+      "epoch": 0.8929625984251969,
+      "grad_norm": 1.2386568784713745,
+      "learning_rate": 5.607561887635249e-06,
+      "loss": 1.9698,
+      "step": 14516
+    },
+    {
+      "epoch": 0.8930241141732284,
+      "grad_norm": 1.554761290550232,
+      "learning_rate": 5.6011791047199e-06,
+      "loss": 2.4715,
+      "step": 14517
+    },
+    {
+      "epoch": 0.8930856299212598,
+      "grad_norm": 1.3523590564727783,
+      "learning_rate": 5.5947998517940635e-06,
+      "loss": 1.971,
+      "step": 14518
+    },
+    {
+      "epoch": 0.8931471456692913,
+      "grad_norm": 1.4379605054855347,
+      "learning_rate": 5.588424129096315e-06,
+      "loss": 2.6708,
+      "step": 14519
+    },
+    {
+      "epoch": 0.8932086614173228,
+      "grad_norm": 1.4059735536575317,
+      "learning_rate": 5.5820519368650405e-06,
+      "loss": 2.0546,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8932701771653543,
+      "grad_norm": 1.3857216835021973,
+      "learning_rate": 5.575683275338539e-06,
+      "loss": 2.3145,
+      "step": 14521
+    },
+    {
+      "epoch": 0.8933316929133859,
+      "grad_norm": 1.1637784242630005,
+      "learning_rate": 5.569318144754954e-06,
+      "loss": 2.1673,
+      "step": 14522
+    },
+    {
+      "epoch": 0.8933932086614174,
+      "grad_norm": 1.1909840106964111,
+      "learning_rate": 5.562956545352316e-06,
+      "loss": 2.8234,
+      "step": 14523
+    },
+    {
+      "epoch": 0.8934547244094488,
+      "grad_norm": 1.2633169889450073,
+      "learning_rate": 5.556598477368502e-06,
+      "loss": 1.6586,
+      "step": 14524
+    },
+    {
+      "epoch": 0.8935162401574803,
+      "grad_norm": 1.4145236015319824,
+      "learning_rate": 5.5502439410412665e-06,
+      "loss": 2.613,
+      "step": 14525
+    },
+    {
+      "epoch": 0.8935777559055118,
+      "grad_norm": 1.2425386905670166,
+      "learning_rate": 5.543892936608242e-06,
+      "loss": 2.9624,
+      "step": 14526
+    },
+    {
+      "epoch": 0.8936392716535433,
+      "grad_norm": 1.1693185567855835,
+      "learning_rate": 5.5375454643069365e-06,
+      "loss": 2.7335,
+      "step": 14527
+    },
+    {
+      "epoch": 0.8937007874015748,
+      "grad_norm": 1.9256967306137085,
+      "learning_rate": 5.531201524374674e-06,
+      "loss": 2.0708,
+      "step": 14528
+    },
+    {
+      "epoch": 0.8937623031496063,
+      "grad_norm": 1.090990662574768,
+      "learning_rate": 5.5248611170487295e-06,
+      "loss": 2.7846,
+      "step": 14529
+    },
+    {
+      "epoch": 0.8938238188976378,
+      "grad_norm": 1.3784393072128296,
+      "learning_rate": 5.518524242566148e-06,
+      "loss": 2.7507,
+      "step": 14530
+    },
+    {
+      "epoch": 0.8938853346456693,
+      "grad_norm": 1.2880878448486328,
+      "learning_rate": 5.512190901163916e-06,
+      "loss": 2.7416,
+      "step": 14531
+    },
+    {
+      "epoch": 0.8939468503937008,
+      "grad_norm": 1.3620144128799438,
+      "learning_rate": 5.505861093078879e-06,
+      "loss": 2.1122,
+      "step": 14532
+    },
+    {
+      "epoch": 0.8940083661417323,
+      "grad_norm": 1.2291381359100342,
+      "learning_rate": 5.499534818547714e-06,
+      "loss": 2.697,
+      "step": 14533
+    },
+    {
+      "epoch": 0.8940698818897638,
+      "grad_norm": 1.5854872465133667,
+      "learning_rate": 5.493212077807009e-06,
+      "loss": 1.9853,
+      "step": 14534
+    },
+    {
+      "epoch": 0.8941313976377953,
+      "grad_norm": 1.2109755277633667,
+      "learning_rate": 5.4868928710931745e-06,
+      "loss": 2.4754,
+      "step": 14535
+    },
+    {
+      "epoch": 0.8941929133858267,
+      "grad_norm": 1.4700685739517212,
+      "learning_rate": 5.480577198642522e-06,
+      "loss": 2.5086,
+      "step": 14536
+    },
+    {
+      "epoch": 0.8942544291338582,
+      "grad_norm": 1.6791919469833374,
+      "learning_rate": 5.474265060691242e-06,
+      "loss": 1.9828,
+      "step": 14537
+    },
+    {
+      "epoch": 0.8943159448818898,
+      "grad_norm": 1.2971140146255493,
+      "learning_rate": 5.467956457475365e-06,
+      "loss": 2.3966,
+      "step": 14538
+    },
+    {
+      "epoch": 0.8943774606299213,
+      "grad_norm": 1.3147655725479126,
+      "learning_rate": 5.461651389230771e-06,
+      "loss": 2.6452,
+      "step": 14539
+    },
+    {
+      "epoch": 0.8944389763779528,
+      "grad_norm": 1.1735681295394897,
+      "learning_rate": 5.455349856193271e-06,
+      "loss": 2.256,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8945004921259843,
+      "grad_norm": 1.2259509563446045,
+      "learning_rate": 5.449051858598486e-06,
+      "loss": 2.1855,
+      "step": 14541
+    },
+    {
+      "epoch": 0.8945620078740157,
+      "grad_norm": 1.271925926208496,
+      "learning_rate": 5.44275739668193e-06,
+      "loss": 2.8859,
+      "step": 14542
+    },
+    {
+      "epoch": 0.8946235236220472,
+      "grad_norm": 1.2929446697235107,
+      "learning_rate": 5.436466470678969e-06,
+      "loss": 2.5245,
+      "step": 14543
+    },
+    {
+      "epoch": 0.8946850393700787,
+      "grad_norm": 1.4886528253555298,
+      "learning_rate": 5.430179080824882e-06,
+      "loss": 2.3443,
+      "step": 14544
+    },
+    {
+      "epoch": 0.8947465551181102,
+      "grad_norm": 1.3172470331192017,
+      "learning_rate": 5.423895227354747e-06,
+      "loss": 2.5944,
+      "step": 14545
+    },
+    {
+      "epoch": 0.8948080708661418,
+      "grad_norm": 1.2743608951568604,
+      "learning_rate": 5.417614910503577e-06,
+      "loss": 2.6492,
+      "step": 14546
+    },
+    {
+      "epoch": 0.8948695866141733,
+      "grad_norm": 1.200703501701355,
+      "learning_rate": 5.411338130506183e-06,
+      "loss": 2.7522,
+      "step": 14547
+    },
+    {
+      "epoch": 0.8949311023622047,
+      "grad_norm": 1.1908267736434937,
+      "learning_rate": 5.405064887597322e-06,
+      "loss": 2.9831,
+      "step": 14548
+    },
+    {
+      "epoch": 0.8949926181102362,
+      "grad_norm": 1.3428568840026855,
+      "learning_rate": 5.39879518201154e-06,
+      "loss": 2.244,
+      "step": 14549
+    },
+    {
+      "epoch": 0.8950541338582677,
+      "grad_norm": 1.3493317365646362,
+      "learning_rate": 5.392529013983316e-06,
+      "loss": 2.4496,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8951156496062992,
+      "grad_norm": 1.43540620803833,
+      "learning_rate": 5.386266383746952e-06,
+      "loss": 2.6199,
+      "step": 14551
+    },
+    {
+      "epoch": 0.8951771653543307,
+      "grad_norm": 1.2794917821884155,
+      "learning_rate": 5.380007291536659e-06,
+      "loss": 1.6952,
+      "step": 14552
+    },
+    {
+      "epoch": 0.8952386811023622,
+      "grad_norm": 1.3037686347961426,
+      "learning_rate": 5.3737517375864634e-06,
+      "loss": 2.703,
+      "step": 14553
+    },
+    {
+      "epoch": 0.8953001968503937,
+      "grad_norm": 1.2225921154022217,
+      "learning_rate": 5.367499722130321e-06,
+      "loss": 2.1397,
+      "step": 14554
+    },
+    {
+      "epoch": 0.8953617125984252,
+      "grad_norm": 1.279569387435913,
+      "learning_rate": 5.361251245401988e-06,
+      "loss": 2.5719,
+      "step": 14555
+    },
+    {
+      "epoch": 0.8954232283464567,
+      "grad_norm": 1.3561780452728271,
+      "learning_rate": 5.3550063076351465e-06,
+      "loss": 2.3356,
+      "step": 14556
+    },
+    {
+      "epoch": 0.8954847440944882,
+      "grad_norm": 2.262644052505493,
+      "learning_rate": 5.348764909063297e-06,
+      "loss": 2.4803,
+      "step": 14557
+    },
+    {
+      "epoch": 0.8955462598425197,
+      "grad_norm": 1.2370535135269165,
+      "learning_rate": 5.342527049919876e-06,
+      "loss": 3.2019,
+      "step": 14558
+    },
+    {
+      "epoch": 0.8956077755905512,
+      "grad_norm": 1.1814799308776855,
+      "learning_rate": 5.336292730438108e-06,
+      "loss": 2.4692,
+      "step": 14559
+    },
+    {
+      "epoch": 0.8956692913385826,
+      "grad_norm": 1.2794253826141357,
+      "learning_rate": 5.33006195085114e-06,
+      "loss": 1.7537,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8957308070866141,
+      "grad_norm": 1.180512547492981,
+      "learning_rate": 5.323834711391951e-06,
+      "loss": 2.5075,
+      "step": 14561
+    },
+    {
+      "epoch": 0.8957923228346457,
+      "grad_norm": 1.3085519075393677,
+      "learning_rate": 5.317611012293411e-06,
+      "loss": 2.5339,
+      "step": 14562
+    },
+    {
+      "epoch": 0.8958538385826772,
+      "grad_norm": 1.3337653875350952,
+      "learning_rate": 5.311390853788256e-06,
+      "loss": 2.8611,
+      "step": 14563
+    },
+    {
+      "epoch": 0.8959153543307087,
+      "grad_norm": 1.3432122468948364,
+      "learning_rate": 5.305174236109101e-06,
+      "loss": 2.5184,
+      "step": 14564
+    },
+    {
+      "epoch": 0.8959768700787402,
+      "grad_norm": 1.2620221376419067,
+      "learning_rate": 5.29896115948838e-06,
+      "loss": 1.9629,
+      "step": 14565
+    },
+    {
+      "epoch": 0.8960383858267716,
+      "grad_norm": 1.2500380277633667,
+      "learning_rate": 5.292751624158454e-06,
+      "loss": 2.1391,
+      "step": 14566
+    },
+    {
+      "epoch": 0.8960999015748031,
+      "grad_norm": 1.243711233139038,
+      "learning_rate": 5.2865456303515244e-06,
+      "loss": 2.7089,
+      "step": 14567
+    },
+    {
+      "epoch": 0.8961614173228346,
+      "grad_norm": 1.9905685186386108,
+      "learning_rate": 5.280343178299629e-06,
+      "loss": 1.7381,
+      "step": 14568
+    },
+    {
+      "epoch": 0.8962229330708661,
+      "grad_norm": 1.125632882118225,
+      "learning_rate": 5.2741442682347375e-06,
+      "loss": 3.1455,
+      "step": 14569
+    },
+    {
+      "epoch": 0.8962844488188977,
+      "grad_norm": 1.6993154287338257,
+      "learning_rate": 5.2679489003886305e-06,
+      "loss": 2.2469,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8963459645669292,
+      "grad_norm": 1.0438892841339111,
+      "learning_rate": 5.261757074993023e-06,
+      "loss": 2.9155,
+      "step": 14571
+    },
+    {
+      "epoch": 0.8964074803149606,
+      "grad_norm": 1.2492576837539673,
+      "learning_rate": 5.255568792279409e-06,
+      "loss": 2.6133,
+      "step": 14572
+    },
+    {
+      "epoch": 0.8964689960629921,
+      "grad_norm": 1.306219458580017,
+      "learning_rate": 5.249384052479223e-06,
+      "loss": 2.6095,
+      "step": 14573
+    },
+    {
+      "epoch": 0.8965305118110236,
+      "grad_norm": 1.0759294033050537,
+      "learning_rate": 5.243202855823715e-06,
+      "loss": 3.1108,
+      "step": 14574
+    },
+    {
+      "epoch": 0.8965920275590551,
+      "grad_norm": 1.4685763120651245,
+      "learning_rate": 5.237025202544044e-06,
+      "loss": 2.1585,
+      "step": 14575
+    },
+    {
+      "epoch": 0.8966535433070866,
+      "grad_norm": 1.222621202468872,
+      "learning_rate": 5.230851092871214e-06,
+      "loss": 2.7263,
+      "step": 14576
+    },
+    {
+      "epoch": 0.8967150590551181,
+      "grad_norm": 1.442571997642517,
+      "learning_rate": 5.224680527036119e-06,
+      "loss": 1.7529,
+      "step": 14577
+    },
+    {
+      "epoch": 0.8967765748031497,
+      "grad_norm": 1.2135932445526123,
+      "learning_rate": 5.218513505269485e-06,
+      "loss": 2.5951,
+      "step": 14578
+    },
+    {
+      "epoch": 0.8968380905511811,
+      "grad_norm": 1.334593653678894,
+      "learning_rate": 5.212350027801938e-06,
+      "loss": 2.085,
+      "step": 14579
+    },
+    {
+      "epoch": 0.8968996062992126,
+      "grad_norm": 1.2296526432037354,
+      "learning_rate": 5.206190094863927e-06,
+      "loss": 3.0911,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8969611220472441,
+      "grad_norm": 1.3543670177459717,
+      "learning_rate": 5.200033706685836e-06,
+      "loss": 2.2483,
+      "step": 14581
+    },
+    {
+      "epoch": 0.8970226377952756,
+      "grad_norm": 1.1895270347595215,
+      "learning_rate": 5.193880863497857e-06,
+      "loss": 3.0504,
+      "step": 14582
+    },
+    {
+      "epoch": 0.8970841535433071,
+      "grad_norm": 1.26246976852417,
+      "learning_rate": 5.1877315655300965e-06,
+      "loss": 2.3808,
+      "step": 14583
+    },
+    {
+      "epoch": 0.8971456692913385,
+      "grad_norm": 1.2573225498199463,
+      "learning_rate": 5.181585813012479e-06,
+      "loss": 2.7195,
+      "step": 14584
+    },
+    {
+      "epoch": 0.89720718503937,
+      "grad_norm": 1.5006000995635986,
+      "learning_rate": 5.175443606174835e-06,
+      "loss": 2.4202,
+      "step": 14585
+    },
+    {
+      "epoch": 0.8972687007874016,
+      "grad_norm": 1.2723604440689087,
+      "learning_rate": 5.169304945246833e-06,
+      "loss": 2.3204,
+      "step": 14586
+    },
+    {
+      "epoch": 0.8973302165354331,
+      "grad_norm": 1.3605889081954956,
+      "learning_rate": 5.163169830458059e-06,
+      "loss": 2.6737,
+      "step": 14587
+    },
+    {
+      "epoch": 0.8973917322834646,
+      "grad_norm": 1.3197062015533447,
+      "learning_rate": 5.157038262037872e-06,
+      "loss": 2.8765,
+      "step": 14588
+    },
+    {
+      "epoch": 0.8974532480314961,
+      "grad_norm": 2.1672348976135254,
+      "learning_rate": 5.150910240215623e-06,
+      "loss": 2.0816,
+      "step": 14589
+    },
+    {
+      "epoch": 0.8975147637795275,
+      "grad_norm": 1.335844874382019,
+      "learning_rate": 5.144785765220428e-06,
+      "loss": 2.7293,
+      "step": 14590
+    },
+    {
+      "epoch": 0.897576279527559,
+      "grad_norm": 1.4181535243988037,
+      "learning_rate": 5.138664837281327e-06,
+      "loss": 2.5904,
+      "step": 14591
+    },
+    {
+      "epoch": 0.8976377952755905,
+      "grad_norm": 1.326682209968567,
+      "learning_rate": 5.13254745662719e-06,
+      "loss": 2.3179,
+      "step": 14592
+    },
+    {
+      "epoch": 0.897699311023622,
+      "grad_norm": 1.4147453308105469,
+      "learning_rate": 5.126433623486793e-06,
+      "loss": 2.4742,
+      "step": 14593
+    },
+    {
+      "epoch": 0.8977608267716536,
+      "grad_norm": 1.1410118341445923,
+      "learning_rate": 5.120323338088739e-06,
+      "loss": 2.6957,
+      "step": 14594
+    },
+    {
+      "epoch": 0.8978223425196851,
+      "grad_norm": 1.0861049890518188,
+      "learning_rate": 5.114216600661525e-06,
+      "loss": 2.0881,
+      "step": 14595
+    },
+    {
+      "epoch": 0.8978838582677166,
+      "grad_norm": 1.2766735553741455,
+      "learning_rate": 5.108113411433513e-06,
+      "loss": 2.593,
+      "step": 14596
+    },
+    {
+      "epoch": 0.897945374015748,
+      "grad_norm": 1.1955246925354004,
+      "learning_rate": 5.1020137706329405e-06,
+      "loss": 2.6413,
+      "step": 14597
+    },
+    {
+      "epoch": 0.8980068897637795,
+      "grad_norm": 1.2102389335632324,
+      "learning_rate": 5.095917678487871e-06,
+      "loss": 2.3651,
+      "step": 14598
+    },
+    {
+      "epoch": 0.898068405511811,
+      "grad_norm": 1.3162448406219482,
+      "learning_rate": 5.0898251352263e-06,
+      "loss": 2.5867,
+      "step": 14599
+    },
+    {
+      "epoch": 0.8981299212598425,
+      "grad_norm": 1.1766011714935303,
+      "learning_rate": 5.083736141076023e-06,
+      "loss": 1.6492,
+      "step": 14600
+    },
+    {
+      "epoch": 0.898191437007874,
+      "grad_norm": 1.9870439767837524,
+      "learning_rate": 5.0776506962647465e-06,
+      "loss": 2.2629,
+      "step": 14601
+    },
+    {
+      "epoch": 0.8982529527559056,
+      "grad_norm": 1.3329181671142578,
+      "learning_rate": 5.071568801020044e-06,
+      "loss": 2.5796,
+      "step": 14602
+    },
+    {
+      "epoch": 0.898314468503937,
+      "grad_norm": 1.3233801126480103,
+      "learning_rate": 5.065490455569322e-06,
+      "loss": 1.802,
+      "step": 14603
+    },
+    {
+      "epoch": 0.8983759842519685,
+      "grad_norm": 1.4256162643432617,
+      "learning_rate": 5.059415660139888e-06,
+      "loss": 2.9132,
+      "step": 14604
+    },
+    {
+      "epoch": 0.8984375,
+      "grad_norm": 1.2794150114059448,
+      "learning_rate": 5.053344414958905e-06,
+      "loss": 2.4349,
+      "step": 14605
+    },
+    {
+      "epoch": 0.8984990157480315,
+      "grad_norm": 1.3244179487228394,
+      "learning_rate": 5.047276720253402e-06,
+      "loss": 2.6508,
+      "step": 14606
+    },
+    {
+      "epoch": 0.898560531496063,
+      "grad_norm": 1.371914267539978,
+      "learning_rate": 5.041212576250276e-06,
+      "loss": 2.2163,
+      "step": 14607
+    },
+    {
+      "epoch": 0.8986220472440944,
+      "grad_norm": 1.2046005725860596,
+      "learning_rate": 5.035151983176311e-06,
+      "loss": 2.3073,
+      "step": 14608
+    },
+    {
+      "epoch": 0.898683562992126,
+      "grad_norm": 1.247689127922058,
+      "learning_rate": 5.029094941258105e-06,
+      "loss": 2.4827,
+      "step": 14609
+    },
+    {
+      "epoch": 0.8987450787401575,
+      "grad_norm": 1.1581108570098877,
+      "learning_rate": 5.023041450722199e-06,
+      "loss": 2.1083,
+      "step": 14610
+    },
+    {
+      "epoch": 0.898806594488189,
+      "grad_norm": 1.343044400215149,
+      "learning_rate": 5.0169915117949106e-06,
+      "loss": 2.449,
+      "step": 14611
+    },
+    {
+      "epoch": 0.8988681102362205,
+      "grad_norm": 1.4282152652740479,
+      "learning_rate": 5.010945124702526e-06,
+      "loss": 2.4501,
+      "step": 14612
+    },
+    {
+      "epoch": 0.898929625984252,
+      "grad_norm": 1.2875409126281738,
+      "learning_rate": 5.0049022896710986e-06,
+      "loss": 2.5413,
+      "step": 14613
+    },
+    {
+      "epoch": 0.8989911417322834,
+      "grad_norm": 1.161148190498352,
+      "learning_rate": 4.998863006926624e-06,
+      "loss": 2.5012,
+      "step": 14614
+    },
+    {
+      "epoch": 0.8990526574803149,
+      "grad_norm": 1.3472241163253784,
+      "learning_rate": 4.992827276694922e-06,
+      "loss": 1.9504,
+      "step": 14615
+    },
+    {
+      "epoch": 0.8991141732283464,
+      "grad_norm": 1.1727232933044434,
+      "learning_rate": 4.986795099201725e-06,
+      "loss": 2.9137,
+      "step": 14616
+    },
+    {
+      "epoch": 0.899175688976378,
+      "grad_norm": 1.443488597869873,
+      "learning_rate": 4.980766474672571e-06,
+      "loss": 2.3365,
+      "step": 14617
+    },
+    {
+      "epoch": 0.8992372047244095,
+      "grad_norm": 1.2698702812194824,
+      "learning_rate": 4.9747414033329275e-06,
+      "loss": 2.1997,
+      "step": 14618
+    },
+    {
+      "epoch": 0.899298720472441,
+      "grad_norm": 1.2918559312820435,
+      "learning_rate": 4.968719885408058e-06,
+      "loss": 1.5582,
+      "step": 14619
+    },
+    {
+      "epoch": 0.8993602362204725,
+      "grad_norm": 1.173209547996521,
+      "learning_rate": 4.962701921123158e-06,
+      "loss": 1.7364,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8994217519685039,
+      "grad_norm": 1.322935700416565,
+      "learning_rate": 4.956687510703262e-06,
+      "loss": 2.6134,
+      "step": 14621
+    },
+    {
+      "epoch": 0.8994832677165354,
+      "grad_norm": 1.2027537822723389,
+      "learning_rate": 4.9506766543732985e-06,
+      "loss": 3.1078,
+      "step": 14622
+    },
+    {
+      "epoch": 0.8995447834645669,
+      "grad_norm": 1.3308210372924805,
+      "learning_rate": 4.944669352357989e-06,
+      "loss": 2.5078,
+      "step": 14623
+    },
+    {
+      "epoch": 0.8996062992125984,
+      "grad_norm": 1.4796481132507324,
+      "learning_rate": 4.9386656048820315e-06,
+      "loss": 1.5757,
+      "step": 14624
+    },
+    {
+      "epoch": 0.89966781496063,
+      "grad_norm": 1.2510219812393188,
+      "learning_rate": 4.93266541216989e-06,
+      "loss": 2.1582,
+      "step": 14625
+    },
+    {
+      "epoch": 0.8997293307086615,
+      "grad_norm": 1.3587743043899536,
+      "learning_rate": 4.926668774445942e-06,
+      "loss": 2.4965,
+      "step": 14626
+    },
+    {
+      "epoch": 0.8997908464566929,
+      "grad_norm": 1.547088623046875,
+      "learning_rate": 4.92067569193444e-06,
+      "loss": 2.7749,
+      "step": 14627
+    },
+    {
+      "epoch": 0.8998523622047244,
+      "grad_norm": 1.837960124015808,
+      "learning_rate": 4.9146861648595164e-06,
+      "loss": 2.488,
+      "step": 14628
+    },
+    {
+      "epoch": 0.8999138779527559,
+      "grad_norm": 1.2870874404907227,
+      "learning_rate": 4.908700193445092e-06,
+      "loss": 2.2103,
+      "step": 14629
+    },
+    {
+      "epoch": 0.8999753937007874,
+      "grad_norm": 1.296996831893921,
+      "learning_rate": 4.902717777915056e-06,
+      "loss": 2.299,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9000369094488189,
+      "grad_norm": 1.3951354026794434,
+      "learning_rate": 4.8967389184930824e-06,
+      "loss": 2.2735,
+      "step": 14631
+    },
+    {
+      "epoch": 0.9000984251968503,
+      "grad_norm": 1.299291968345642,
+      "learning_rate": 4.8907636154027715e-06,
+      "loss": 2.4892,
+      "step": 14632
+    },
+    {
+      "epoch": 0.9001599409448819,
+      "grad_norm": 1.2442082166671753,
+      "learning_rate": 4.884791868867545e-06,
+      "loss": 2.6576,
+      "step": 14633
+    },
+    {
+      "epoch": 0.9002214566929134,
+      "grad_norm": 1.278187870979309,
+      "learning_rate": 4.878823679110744e-06,
+      "loss": 2.2003,
+      "step": 14634
+    },
+    {
+      "epoch": 0.9002829724409449,
+      "grad_norm": 1.3023046255111694,
+      "learning_rate": 4.872859046355516e-06,
+      "loss": 2.1046,
+      "step": 14635
+    },
+    {
+      "epoch": 0.9003444881889764,
+      "grad_norm": 1.2825868129730225,
+      "learning_rate": 4.8668979708249355e-06,
+      "loss": 2.308,
+      "step": 14636
+    },
+    {
+      "epoch": 0.9004060039370079,
+      "grad_norm": 1.2572031021118164,
+      "learning_rate": 4.860940452741869e-06,
+      "loss": 2.5066,
+      "step": 14637
+    },
+    {
+      "epoch": 0.9004675196850394,
+      "grad_norm": 1.2117568254470825,
+      "learning_rate": 4.854986492329139e-06,
+      "loss": 1.8226,
+      "step": 14638
+    },
+    {
+      "epoch": 0.9005290354330708,
+      "grad_norm": 1.2444490194320679,
+      "learning_rate": 4.849036089809356e-06,
+      "loss": 2.2799,
+      "step": 14639
+    },
+    {
+      "epoch": 0.9005905511811023,
+      "grad_norm": 1.2125370502471924,
+      "learning_rate": 4.843089245405052e-06,
+      "loss": 2.8665,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9006520669291339,
+      "grad_norm": 1.4146745204925537,
+      "learning_rate": 4.837145959338607e-06,
+      "loss": 2.6448,
+      "step": 14641
+    },
+    {
+      "epoch": 0.9007135826771654,
+      "grad_norm": 1.4155547618865967,
+      "learning_rate": 4.831206231832253e-06,
+      "loss": 2.0613,
+      "step": 14642
+    },
+    {
+      "epoch": 0.9007750984251969,
+      "grad_norm": 1.0809540748596191,
+      "learning_rate": 4.825270063108112e-06,
+      "loss": 2.5991,
+      "step": 14643
+    },
+    {
+      "epoch": 0.9008366141732284,
+      "grad_norm": 1.3494318723678589,
+      "learning_rate": 4.819337453388162e-06,
+      "loss": 2.3688,
+      "step": 14644
+    },
+    {
+      "epoch": 0.9008981299212598,
+      "grad_norm": 1.4018962383270264,
+      "learning_rate": 4.813408402894237e-06,
+      "loss": 2.669,
+      "step": 14645
+    },
+    {
+      "epoch": 0.9009596456692913,
+      "grad_norm": 2.2655398845672607,
+      "learning_rate": 4.807482911848071e-06,
+      "loss": 1.7554,
+      "step": 14646
+    },
+    {
+      "epoch": 0.9010211614173228,
+      "grad_norm": 1.443344235420227,
+      "learning_rate": 4.801560980471254e-06,
+      "loss": 2.5225,
+      "step": 14647
+    },
+    {
+      "epoch": 0.9010826771653543,
+      "grad_norm": 1.3248958587646484,
+      "learning_rate": 4.795642608985196e-06,
+      "loss": 2.4253,
+      "step": 14648
+    },
+    {
+      "epoch": 0.9011441929133859,
+      "grad_norm": 1.5098063945770264,
+      "learning_rate": 4.7897277976112435e-06,
+      "loss": 2.6817,
+      "step": 14649
+    },
+    {
+      "epoch": 0.9012057086614174,
+      "grad_norm": 1.2444242238998413,
+      "learning_rate": 4.783816546570552e-06,
+      "loss": 2.2335,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9012672244094488,
+      "grad_norm": 1.5128728151321411,
+      "learning_rate": 4.7779088560841785e-06,
+      "loss": 2.7233,
+      "step": 14651
+    },
+    {
+      "epoch": 0.9013287401574803,
+      "grad_norm": 1.308783769607544,
+      "learning_rate": 4.7720047263730475e-06,
+      "loss": 2.9193,
+      "step": 14652
+    },
+    {
+      "epoch": 0.9013902559055118,
+      "grad_norm": 1.1101206541061401,
+      "learning_rate": 4.766104157657936e-06,
+      "loss": 2.0472,
+      "step": 14653
+    },
+    {
+      "epoch": 0.9014517716535433,
+      "grad_norm": 1.2506060600280762,
+      "learning_rate": 4.76020715015949e-06,
+      "loss": 1.8153,
+      "step": 14654
+    },
+    {
+      "epoch": 0.9015132874015748,
+      "grad_norm": 1.1887929439544678,
+      "learning_rate": 4.754313704098235e-06,
+      "loss": 1.6733,
+      "step": 14655
+    },
+    {
+      "epoch": 0.9015748031496063,
+      "grad_norm": 1.2233394384384155,
+      "learning_rate": 4.748423819694525e-06,
+      "loss": 2.3981,
+      "step": 14656
+    },
+    {
+      "epoch": 0.9016363188976378,
+      "grad_norm": 1.1174113750457764,
+      "learning_rate": 4.7425374971686424e-06,
+      "loss": 1.5072,
+      "step": 14657
+    },
+    {
+      "epoch": 0.9016978346456693,
+      "grad_norm": 1.3344355821609497,
+      "learning_rate": 4.736654736740653e-06,
+      "loss": 2.5688,
+      "step": 14658
+    },
+    {
+      "epoch": 0.9017593503937008,
+      "grad_norm": 1.3110787868499756,
+      "learning_rate": 4.7307755386306165e-06,
+      "loss": 2.3747,
+      "step": 14659
+    },
+    {
+      "epoch": 0.9018208661417323,
+      "grad_norm": 1.3375577926635742,
+      "learning_rate": 4.724899903058311e-06,
+      "loss": 2.8106,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9018823818897638,
+      "grad_norm": 1.6613600254058838,
+      "learning_rate": 4.7190278302434944e-06,
+      "loss": 2.901,
+      "step": 14661
+    },
+    {
+      "epoch": 0.9019438976377953,
+      "grad_norm": 1.1624813079833984,
+      "learning_rate": 4.713159320405735e-06,
+      "loss": 1.9754,
+      "step": 14662
+    },
+    {
+      "epoch": 0.9020054133858267,
+      "grad_norm": 1.2456127405166626,
+      "learning_rate": 4.707294373764482e-06,
+      "loss": 2.0627,
+      "step": 14663
+    },
+    {
+      "epoch": 0.9020669291338582,
+      "grad_norm": 1.2370291948318481,
+      "learning_rate": 4.701432990539045e-06,
+      "loss": 3.07,
+      "step": 14664
+    },
+    {
+      "epoch": 0.9021284448818898,
+      "grad_norm": 1.4557350873947144,
+      "learning_rate": 4.695575170948618e-06,
+      "loss": 2.5893,
+      "step": 14665
+    },
+    {
+      "epoch": 0.9021899606299213,
+      "grad_norm": 1.3272873163223267,
+      "learning_rate": 4.689720915212248e-06,
+      "loss": 1.9978,
+      "step": 14666
+    },
+    {
+      "epoch": 0.9022514763779528,
+      "grad_norm": 1.319372534751892,
+      "learning_rate": 4.683870223548858e-06,
+      "loss": 2.3811,
+      "step": 14667
+    },
+    {
+      "epoch": 0.9023129921259843,
+      "grad_norm": 1.366273045539856,
+      "learning_rate": 4.6780230961772195e-06,
+      "loss": 2.0557,
+      "step": 14668
+    },
+    {
+      "epoch": 0.9023745078740157,
+      "grad_norm": 1.308659553527832,
+      "learning_rate": 4.672179533316001e-06,
+      "loss": 2.7045,
+      "step": 14669
+    },
+    {
+      "epoch": 0.9024360236220472,
+      "grad_norm": 1.3252365589141846,
+      "learning_rate": 4.666339535183684e-06,
+      "loss": 1.9952,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9024975393700787,
+      "grad_norm": 1.2375667095184326,
+      "learning_rate": 4.66050310199867e-06,
+      "loss": 2.745,
+      "step": 14671
+    },
+    {
+      "epoch": 0.9025590551181102,
+      "grad_norm": 1.4387691020965576,
+      "learning_rate": 4.654670233979219e-06,
+      "loss": 2.6645,
+      "step": 14672
+    },
+    {
+      "epoch": 0.9026205708661418,
+      "grad_norm": 1.7043403387069702,
+      "learning_rate": 4.648840931343456e-06,
+      "loss": 2.9522,
+      "step": 14673
+    },
+    {
+      "epoch": 0.9026820866141733,
+      "grad_norm": 1.435488224029541,
+      "learning_rate": 4.643015194309342e-06,
+      "loss": 2.6371,
+      "step": 14674
+    },
+    {
+      "epoch": 0.9027436023622047,
+      "grad_norm": 1.147553563117981,
+      "learning_rate": 4.637193023094732e-06,
+      "loss": 2.5495,
+      "step": 14675
+    },
+    {
+      "epoch": 0.9028051181102362,
+      "grad_norm": 1.3058502674102783,
+      "learning_rate": 4.631374417917355e-06,
+      "loss": 2.79,
+      "step": 14676
+    },
+    {
+      "epoch": 0.9028666338582677,
+      "grad_norm": 1.2444502115249634,
+      "learning_rate": 4.625559378994759e-06,
+      "loss": 2.3808,
+      "step": 14677
+    },
+    {
+      "epoch": 0.9029281496062992,
+      "grad_norm": 1.120280385017395,
+      "learning_rate": 4.619747906544447e-06,
+      "loss": 2.438,
+      "step": 14678
+    },
+    {
+      "epoch": 0.9029896653543307,
+      "grad_norm": 1.2747539281845093,
+      "learning_rate": 4.6139400007836895e-06,
+      "loss": 2.7353,
+      "step": 14679
+    },
+    {
+      "epoch": 0.9030511811023622,
+      "grad_norm": 1.2087355852127075,
+      "learning_rate": 4.608135661929702e-06,
+      "loss": 2.7154,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9031126968503937,
+      "grad_norm": 1.126596212387085,
+      "learning_rate": 4.602334890199511e-06,
+      "loss": 2.5753,
+      "step": 14681
+    },
+    {
+      "epoch": 0.9031742125984252,
+      "grad_norm": 1.3403067588806152,
+      "learning_rate": 4.596537685810054e-06,
+      "loss": 2.73,
+      "step": 14682
+    },
+    {
+      "epoch": 0.9032357283464567,
+      "grad_norm": 1.2626731395721436,
+      "learning_rate": 4.590744048978102e-06,
+      "loss": 1.8041,
+      "step": 14683
+    },
+    {
+      "epoch": 0.9032972440944882,
+      "grad_norm": 1.2197802066802979,
+      "learning_rate": 4.584953979920292e-06,
+      "loss": 2.3163,
+      "step": 14684
+    },
+    {
+      "epoch": 0.9033587598425197,
+      "grad_norm": 1.217928409576416,
+      "learning_rate": 4.579167478853153e-06,
+      "loss": 2.7194,
+      "step": 14685
+    },
+    {
+      "epoch": 0.9034202755905512,
+      "grad_norm": 1.1178916692733765,
+      "learning_rate": 4.573384545993087e-06,
+      "loss": 2.3843,
+      "step": 14686
+    },
+    {
+      "epoch": 0.9034817913385826,
+      "grad_norm": 1.4043710231781006,
+      "learning_rate": 4.567605181556323e-06,
+      "loss": 2.341,
+      "step": 14687
+    },
+    {
+      "epoch": 0.9035433070866141,
+      "grad_norm": 1.3827368021011353,
+      "learning_rate": 4.561829385758976e-06,
+      "loss": 2.4919,
+      "step": 14688
+    },
+    {
+      "epoch": 0.9036048228346457,
+      "grad_norm": 1.188826322555542,
+      "learning_rate": 4.556057158817029e-06,
+      "loss": 2.6152,
+      "step": 14689
+    },
+    {
+      "epoch": 0.9036663385826772,
+      "grad_norm": 1.340112328529358,
+      "learning_rate": 4.550288500946331e-06,
+      "loss": 2.2445,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9037278543307087,
+      "grad_norm": 1.226772665977478,
+      "learning_rate": 4.5445234123626005e-06,
+      "loss": 2.576,
+      "step": 14691
+    },
+    {
+      "epoch": 0.9037893700787402,
+      "grad_norm": 1.0882834196090698,
+      "learning_rate": 4.5387618932814295e-06,
+      "loss": 2.0935,
+      "step": 14692
+    },
+    {
+      "epoch": 0.9038508858267716,
+      "grad_norm": 1.2286558151245117,
+      "learning_rate": 4.533003943918257e-06,
+      "loss": 2.8123,
+      "step": 14693
+    },
+    {
+      "epoch": 0.9039124015748031,
+      "grad_norm": 1.3732227087020874,
+      "learning_rate": 4.5272495644884e-06,
+      "loss": 2.2589,
+      "step": 14694
+    },
+    {
+      "epoch": 0.9039739173228346,
+      "grad_norm": 1.5498008728027344,
+      "learning_rate": 4.521498755207021e-06,
+      "loss": 2.5218,
+      "step": 14695
+    },
+    {
+      "epoch": 0.9040354330708661,
+      "grad_norm": 2.2214269638061523,
+      "learning_rate": 4.5157515162892015e-06,
+      "loss": 2.4164,
+      "step": 14696
+    },
+    {
+      "epoch": 0.9040969488188977,
+      "grad_norm": 1.517791986465454,
+      "learning_rate": 4.5100078479498265e-06,
+      "loss": 2.3854,
+      "step": 14697
+    },
+    {
+      "epoch": 0.9041584645669292,
+      "grad_norm": 1.3612926006317139,
+      "learning_rate": 4.504267750403712e-06,
+      "loss": 1.6259,
+      "step": 14698
+    },
+    {
+      "epoch": 0.9042199803149606,
+      "grad_norm": 1.3070085048675537,
+      "learning_rate": 4.498531223865466e-06,
+      "loss": 2.411,
+      "step": 14699
+    },
+    {
+      "epoch": 0.9042814960629921,
+      "grad_norm": 1.246006965637207,
+      "learning_rate": 4.492798268549636e-06,
+      "loss": 2.0099,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9043430118110236,
+      "grad_norm": 1.406135082244873,
+      "learning_rate": 4.487068884670587e-06,
+      "loss": 2.6563,
+      "step": 14701
+    },
+    {
+      "epoch": 0.9044045275590551,
+      "grad_norm": 1.1935691833496094,
+      "learning_rate": 4.481343072442567e-06,
+      "loss": 1.9544,
+      "step": 14702
+    },
+    {
+      "epoch": 0.9044660433070866,
+      "grad_norm": 1.2314890623092651,
+      "learning_rate": 4.475620832079664e-06,
+      "loss": 2.5129,
+      "step": 14703
+    },
+    {
+      "epoch": 0.9045275590551181,
+      "grad_norm": 1.3213697671890259,
+      "learning_rate": 4.469902163795914e-06,
+      "loss": 2.5743,
+      "step": 14704
+    },
+    {
+      "epoch": 0.9045890748031497,
+      "grad_norm": 1.3776129484176636,
+      "learning_rate": 4.464187067805114e-06,
+      "loss": 2.1589,
+      "step": 14705
+    },
+    {
+      "epoch": 0.9046505905511811,
+      "grad_norm": 1.3762623071670532,
+      "learning_rate": 4.458475544321017e-06,
+      "loss": 2.5906,
+      "step": 14706
+    },
+    {
+      "epoch": 0.9047121062992126,
+      "grad_norm": 1.5461126565933228,
+      "learning_rate": 4.452767593557161e-06,
+      "loss": 2.3084,
+      "step": 14707
+    },
+    {
+      "epoch": 0.9047736220472441,
+      "grad_norm": 1.4099920988082886,
+      "learning_rate": 4.4470632157270205e-06,
+      "loss": 3.4983,
+      "step": 14708
+    },
+    {
+      "epoch": 0.9048351377952756,
+      "grad_norm": 1.480168342590332,
+      "learning_rate": 4.441362411043892e-06,
+      "loss": 1.5916,
+      "step": 14709
+    },
+    {
+      "epoch": 0.9048966535433071,
+      "grad_norm": 1.4026674032211304,
+      "learning_rate": 4.43566517972096e-06,
+      "loss": 2.3569,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9049581692913385,
+      "grad_norm": 1.2408127784729004,
+      "learning_rate": 4.429971521971266e-06,
+      "loss": 2.4606,
+      "step": 14711
+    },
+    {
+      "epoch": 0.90501968503937,
+      "grad_norm": 1.3480879068374634,
+      "learning_rate": 4.424281438007738e-06,
+      "loss": 2.4336,
+      "step": 14712
+    },
+    {
+      "epoch": 0.9050812007874016,
+      "grad_norm": 1.6334198713302612,
+      "learning_rate": 4.418594928043141e-06,
+      "loss": 2.9147,
+      "step": 14713
+    },
+    {
+      "epoch": 0.9051427165354331,
+      "grad_norm": 1.2199394702911377,
+      "learning_rate": 4.412911992290103e-06,
+      "loss": 2.5978,
+      "step": 14714
+    },
+    {
+      "epoch": 0.9052042322834646,
+      "grad_norm": 1.2959684133529663,
+      "learning_rate": 4.407232630961144e-06,
+      "loss": 1.9758,
+      "step": 14715
+    },
+    {
+      "epoch": 0.9052657480314961,
+      "grad_norm": 1.5756319761276245,
+      "learning_rate": 4.401556844268651e-06,
+      "loss": 2.4021,
+      "step": 14716
+    },
+    {
+      "epoch": 0.9053272637795275,
+      "grad_norm": 1.2450453042984009,
+      "learning_rate": 4.395884632424862e-06,
+      "loss": 2.1191,
+      "step": 14717
+    },
+    {
+      "epoch": 0.905388779527559,
+      "grad_norm": 1.2476919889450073,
+      "learning_rate": 4.390215995641889e-06,
+      "loss": 2.0851,
+      "step": 14718
+    },
+    {
+      "epoch": 0.9054502952755905,
+      "grad_norm": 1.6831368207931519,
+      "learning_rate": 4.384550934131704e-06,
+      "loss": 1.4568,
+      "step": 14719
+    },
+    {
+      "epoch": 0.905511811023622,
+      "grad_norm": 1.5568535327911377,
+      "learning_rate": 4.378889448106138e-06,
+      "loss": 1.5635,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9055733267716536,
+      "grad_norm": 1.286919355392456,
+      "learning_rate": 4.373231537776912e-06,
+      "loss": 2.4534,
+      "step": 14721
+    },
+    {
+      "epoch": 0.9056348425196851,
+      "grad_norm": 1.239933967590332,
+      "learning_rate": 4.3675772033555886e-06,
+      "loss": 2.6904,
+      "step": 14722
+    },
+    {
+      "epoch": 0.9056963582677166,
+      "grad_norm": 1.1641831398010254,
+      "learning_rate": 4.361926445053632e-06,
+      "loss": 2.652,
+      "step": 14723
+    },
+    {
+      "epoch": 0.905757874015748,
+      "grad_norm": 1.358272910118103,
+      "learning_rate": 4.35627926308233e-06,
+      "loss": 2.1858,
+      "step": 14724
+    },
+    {
+      "epoch": 0.9058193897637795,
+      "grad_norm": 1.3560689687728882,
+      "learning_rate": 4.350635657652869e-06,
+      "loss": 2.2197,
+      "step": 14725
+    },
+    {
+      "epoch": 0.905880905511811,
+      "grad_norm": 3.030529260635376,
+      "learning_rate": 4.344995628976256e-06,
+      "loss": 2.2117,
+      "step": 14726
+    },
+    {
+      "epoch": 0.9059424212598425,
+      "grad_norm": 1.7532249689102173,
+      "learning_rate": 4.339359177263447e-06,
+      "loss": 1.8429,
+      "step": 14727
+    },
+    {
+      "epoch": 0.906003937007874,
+      "grad_norm": 1.313029408454895,
+      "learning_rate": 4.333726302725161e-06,
+      "loss": 2.1241,
+      "step": 14728
+    },
+    {
+      "epoch": 0.9060654527559056,
+      "grad_norm": 1.2485074996948242,
+      "learning_rate": 4.3280970055720625e-06,
+      "loss": 2.6371,
+      "step": 14729
+    },
+    {
+      "epoch": 0.906126968503937,
+      "grad_norm": 1.371032476425171,
+      "learning_rate": 4.322471286014662e-06,
+      "loss": 2.7138,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9061884842519685,
+      "grad_norm": 1.2976468801498413,
+      "learning_rate": 4.316849144263335e-06,
+      "loss": 1.8593,
+      "step": 14731
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 1.2128686904907227,
+      "learning_rate": 4.311230580528292e-06,
+      "loss": 1.0335,
+      "step": 14732
+    },
+    {
+      "epoch": 0.9063115157480315,
+      "grad_norm": 1.2638025283813477,
+      "learning_rate": 4.305615595019663e-06,
+      "loss": 2.3647,
+      "step": 14733
+    },
+    {
+      "epoch": 0.906373031496063,
+      "grad_norm": 1.158164620399475,
+      "learning_rate": 4.300004187947382e-06,
+      "loss": 2.4636,
+      "step": 14734
+    },
+    {
+      "epoch": 0.9064345472440944,
+      "grad_norm": 1.301378607749939,
+      "learning_rate": 4.294396359521324e-06,
+      "loss": 1.88,
+      "step": 14735
+    },
+    {
+      "epoch": 0.906496062992126,
+      "grad_norm": 1.3223358392715454,
+      "learning_rate": 4.2887921099511566e-06,
+      "loss": 2.5056,
+      "step": 14736
+    },
+    {
+      "epoch": 0.9065575787401575,
+      "grad_norm": 1.1693177223205566,
+      "learning_rate": 4.283191439446488e-06,
+      "loss": 2.8004,
+      "step": 14737
+    },
+    {
+      "epoch": 0.906619094488189,
+      "grad_norm": 1.1522172689437866,
+      "learning_rate": 4.277594348216707e-06,
+      "loss": 1.237,
+      "step": 14738
+    },
+    {
+      "epoch": 0.9066806102362205,
+      "grad_norm": 1.41617751121521,
+      "learning_rate": 4.272000836471157e-06,
+      "loss": 2.3499,
+      "step": 14739
+    },
+    {
+      "epoch": 0.906742125984252,
+      "grad_norm": 1.1640950441360474,
+      "learning_rate": 4.26641090441896e-06,
+      "loss": 2.3635,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9068036417322834,
+      "grad_norm": 1.2562663555145264,
+      "learning_rate": 4.260824552269171e-06,
+      "loss": 2.1918,
+      "step": 14741
+    },
+    {
+      "epoch": 0.9068651574803149,
+      "grad_norm": 1.3253443241119385,
+      "learning_rate": 4.25524178023069e-06,
+      "loss": 2.5055,
+      "step": 14742
+    },
+    {
+      "epoch": 0.9069266732283464,
+      "grad_norm": 1.342405080795288,
+      "learning_rate": 4.2496625885122935e-06,
+      "loss": 2.8715,
+      "step": 14743
+    },
+    {
+      "epoch": 0.906988188976378,
+      "grad_norm": 1.1478320360183716,
+      "learning_rate": 4.244086977322581e-06,
+      "loss": 2.7461,
+      "step": 14744
+    },
+    {
+      "epoch": 0.9070497047244095,
+      "grad_norm": 1.3068281412124634,
+      "learning_rate": 4.238514946870087e-06,
+      "loss": 2.2879,
+      "step": 14745
+    },
+    {
+      "epoch": 0.907111220472441,
+      "grad_norm": 1.390403151512146,
+      "learning_rate": 4.2329464973631325e-06,
+      "loss": 2.4021,
+      "step": 14746
+    },
+    {
+      "epoch": 0.9071727362204725,
+      "grad_norm": 1.1168655157089233,
+      "learning_rate": 4.227381629009974e-06,
+      "loss": 1.9592,
+      "step": 14747
+    },
+    {
+      "epoch": 0.9072342519685039,
+      "grad_norm": 1.4085651636123657,
+      "learning_rate": 4.22182034201869e-06,
+      "loss": 2.5513,
+      "step": 14748
+    },
+    {
+      "epoch": 0.9072957677165354,
+      "grad_norm": 1.3563090562820435,
+      "learning_rate": 4.216262636597268e-06,
+      "loss": 2.4522,
+      "step": 14749
+    },
+    {
+      "epoch": 0.9073572834645669,
+      "grad_norm": 1.2262643575668335,
+      "learning_rate": 4.210708512953521e-06,
+      "loss": 1.4809,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9074187992125984,
+      "grad_norm": 1.5197542905807495,
+      "learning_rate": 4.205157971295126e-06,
+      "loss": 2.0424,
+      "step": 14751
+    },
+    {
+      "epoch": 0.90748031496063,
+      "grad_norm": 1.2077562808990479,
+      "learning_rate": 4.199611011829674e-06,
+      "loss": 1.7101,
+      "step": 14752
+    },
+    {
+      "epoch": 0.9075418307086615,
+      "grad_norm": 1.3539457321166992,
+      "learning_rate": 4.1940676347645534e-06,
+      "loss": 2.7852,
+      "step": 14753
+    },
+    {
+      "epoch": 0.9076033464566929,
+      "grad_norm": 1.2342476844787598,
+      "learning_rate": 4.188527840307077e-06,
+      "loss": 2.0997,
+      "step": 14754
+    },
+    {
+      "epoch": 0.9076648622047244,
+      "grad_norm": 1.3068161010742188,
+      "learning_rate": 4.18299162866439e-06,
+      "loss": 2.4122,
+      "step": 14755
+    },
+    {
+      "epoch": 0.9077263779527559,
+      "grad_norm": 1.2914049625396729,
+      "learning_rate": 4.177459000043549e-06,
+      "loss": 2.6404,
+      "step": 14756
+    },
+    {
+      "epoch": 0.9077878937007874,
+      "grad_norm": 1.7760944366455078,
+      "learning_rate": 4.1719299546514105e-06,
+      "loss": 2.277,
+      "step": 14757
+    },
+    {
+      "epoch": 0.9078494094488189,
+      "grad_norm": 1.351822018623352,
+      "learning_rate": 4.166404492694742e-06,
+      "loss": 1.6652,
+      "step": 14758
+    },
+    {
+      "epoch": 0.9079109251968503,
+      "grad_norm": 1.2728838920593262,
+      "learning_rate": 4.160882614380157e-06,
+      "loss": 3.0264,
+      "step": 14759
+    },
+    {
+      "epoch": 0.9079724409448819,
+      "grad_norm": 1.18834388256073,
+      "learning_rate": 4.155364319914157e-06,
+      "loss": 2.6672,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9080339566929134,
+      "grad_norm": 1.111777901649475,
+      "learning_rate": 4.149849609503076e-06,
+      "loss": 2.6414,
+      "step": 14761
+    },
+    {
+      "epoch": 0.9080954724409449,
+      "grad_norm": 1.5799317359924316,
+      "learning_rate": 4.1443384833531604e-06,
+      "loss": 2.4703,
+      "step": 14762
+    },
+    {
+      "epoch": 0.9081569881889764,
+      "grad_norm": 1.6122400760650635,
+      "learning_rate": 4.1388309416704575e-06,
+      "loss": 2.2414,
+      "step": 14763
+    },
+    {
+      "epoch": 0.9082185039370079,
+      "grad_norm": 1.5058997869491577,
+      "learning_rate": 4.133326984660968e-06,
+      "loss": 2.1205,
+      "step": 14764
+    },
+    {
+      "epoch": 0.9082800196850394,
+      "grad_norm": 1.3980737924575806,
+      "learning_rate": 4.127826612530461e-06,
+      "loss": 2.4464,
+      "step": 14765
+    },
+    {
+      "epoch": 0.9083415354330708,
+      "grad_norm": 1.1982488632202148,
+      "learning_rate": 4.122329825484639e-06,
+      "loss": 2.6626,
+      "step": 14766
+    },
+    {
+      "epoch": 0.9084030511811023,
+      "grad_norm": 1.2310413122177124,
+      "learning_rate": 4.116836623729059e-06,
+      "loss": 2.3188,
+      "step": 14767
+    },
+    {
+      "epoch": 0.9084645669291339,
+      "grad_norm": 1.2555168867111206,
+      "learning_rate": 4.111347007469146e-06,
+      "loss": 2.9984,
+      "step": 14768
+    },
+    {
+      "epoch": 0.9085260826771654,
+      "grad_norm": 1.6299006938934326,
+      "learning_rate": 4.105860976910148e-06,
+      "loss": 2.115,
+      "step": 14769
+    },
+    {
+      "epoch": 0.9085875984251969,
+      "grad_norm": 1.2633634805679321,
+      "learning_rate": 4.100378532257232e-06,
+      "loss": 2.1909,
+      "step": 14770
+    },
+    {
+      "epoch": 0.9086491141732284,
+      "grad_norm": 1.1600860357284546,
+      "learning_rate": 4.094899673715402e-06,
+      "loss": 2.3547,
+      "step": 14771
+    },
+    {
+      "epoch": 0.9087106299212598,
+      "grad_norm": 1.3090146780014038,
+      "learning_rate": 4.089424401489561e-06,
+      "loss": 1.7545,
+      "step": 14772
+    },
+    {
+      "epoch": 0.9087721456692913,
+      "grad_norm": 1.274096965789795,
+      "learning_rate": 4.083952715784412e-06,
+      "loss": 1.6651,
+      "step": 14773
+    },
+    {
+      "epoch": 0.9088336614173228,
+      "grad_norm": 1.5732364654541016,
+      "learning_rate": 4.078484616804601e-06,
+      "loss": 2.8508,
+      "step": 14774
+    },
+    {
+      "epoch": 0.9088951771653543,
+      "grad_norm": 1.3416237831115723,
+      "learning_rate": 4.073020104754588e-06,
+      "loss": 2.8928,
+      "step": 14775
+    },
+    {
+      "epoch": 0.9089566929133859,
+      "grad_norm": 1.3467551469802856,
+      "learning_rate": 4.067559179838731e-06,
+      "loss": 2.2999,
+      "step": 14776
+    },
+    {
+      "epoch": 0.9090182086614174,
+      "grad_norm": 1.2781902551651,
+      "learning_rate": 4.062101842261212e-06,
+      "loss": 2.1971,
+      "step": 14777
+    },
+    {
+      "epoch": 0.9090797244094488,
+      "grad_norm": 1.4517076015472412,
+      "learning_rate": 4.056648092226145e-06,
+      "loss": 2.6448,
+      "step": 14778
+    },
+    {
+      "epoch": 0.9091412401574803,
+      "grad_norm": 1.402185320854187,
+      "learning_rate": 4.051197929937422e-06,
+      "loss": 2.362,
+      "step": 14779
+    },
+    {
+      "epoch": 0.9092027559055118,
+      "grad_norm": 1.3886609077453613,
+      "learning_rate": 4.04575135559887e-06,
+      "loss": 2.4396,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9092642716535433,
+      "grad_norm": 1.4630340337753296,
+      "learning_rate": 4.040308369414159e-06,
+      "loss": 2.6012,
+      "step": 14781
+    },
+    {
+      "epoch": 0.9093257874015748,
+      "grad_norm": 1.5027821063995361,
+      "learning_rate": 4.034868971586847e-06,
+      "loss": 1.7655,
+      "step": 14782
+    },
+    {
+      "epoch": 0.9093873031496063,
+      "grad_norm": 1.3326245546340942,
+      "learning_rate": 4.029433162320306e-06,
+      "loss": 2.8847,
+      "step": 14783
+    },
+    {
+      "epoch": 0.9094488188976378,
+      "grad_norm": 1.323238730430603,
+      "learning_rate": 4.02400094181783e-06,
+      "loss": 2.4169,
+      "step": 14784
+    },
+    {
+      "epoch": 0.9095103346456693,
+      "grad_norm": 1.3950142860412598,
+      "learning_rate": 4.018572310282531e-06,
+      "loss": 2.3388,
+      "step": 14785
+    },
+    {
+      "epoch": 0.9095718503937008,
+      "grad_norm": 1.4484930038452148,
+      "learning_rate": 4.0131472679174165e-06,
+      "loss": 2.5052,
+      "step": 14786
+    },
+    {
+      "epoch": 0.9096333661417323,
+      "grad_norm": 1.2543833255767822,
+      "learning_rate": 4.007725814925367e-06,
+      "loss": 2.8287,
+      "step": 14787
+    },
+    {
+      "epoch": 0.9096948818897638,
+      "grad_norm": 1.1526458263397217,
+      "learning_rate": 4.0023079515090985e-06,
+      "loss": 2.1572,
+      "step": 14788
+    },
+    {
+      "epoch": 0.9097563976377953,
+      "grad_norm": 1.2830592393875122,
+      "learning_rate": 3.996893677871227e-06,
+      "loss": 2.6977,
+      "step": 14789
+    },
+    {
+      "epoch": 0.9098179133858267,
+      "grad_norm": 1.4895191192626953,
+      "learning_rate": 3.991482994214191e-06,
+      "loss": 2.8828,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9098794291338582,
+      "grad_norm": 1.8590654134750366,
+      "learning_rate": 3.986075900740327e-06,
+      "loss": 1.6962,
+      "step": 14791
+    },
+    {
+      "epoch": 0.9099409448818898,
+      "grad_norm": 1.1768838167190552,
+      "learning_rate": 3.980672397651841e-06,
+      "loss": 2.8651,
+      "step": 14792
+    },
+    {
+      "epoch": 0.9100024606299213,
+      "grad_norm": 1.3102003335952759,
+      "learning_rate": 3.975272485150805e-06,
+      "loss": 2.321,
+      "step": 14793
+    },
+    {
+      "epoch": 0.9100639763779528,
+      "grad_norm": 1.3173953294754028,
+      "learning_rate": 3.969876163439113e-06,
+      "loss": 2.1865,
+      "step": 14794
+    },
+    {
+      "epoch": 0.9101254921259843,
+      "grad_norm": 1.2017900943756104,
+      "learning_rate": 3.9644834327185795e-06,
+      "loss": 2.6251,
+      "step": 14795
+    },
+    {
+      "epoch": 0.9101870078740157,
+      "grad_norm": 1.351407766342163,
+      "learning_rate": 3.959094293190857e-06,
+      "loss": 2.3679,
+      "step": 14796
+    },
+    {
+      "epoch": 0.9102485236220472,
+      "grad_norm": 1.3511159420013428,
+      "learning_rate": 3.953708745057472e-06,
+      "loss": 2.741,
+      "step": 14797
+    },
+    {
+      "epoch": 0.9103100393700787,
+      "grad_norm": 1.210368275642395,
+      "learning_rate": 3.948326788519807e-06,
+      "loss": 2.3571,
+      "step": 14798
+    },
+    {
+      "epoch": 0.9103715551181102,
+      "grad_norm": 1.365233302116394,
+      "learning_rate": 3.942948423779114e-06,
+      "loss": 2.7709,
+      "step": 14799
+    },
+    {
+      "epoch": 0.9104330708661418,
+      "grad_norm": 1.2501029968261719,
+      "learning_rate": 3.9375736510365194e-06,
+      "loss": 2.1442,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9104945866141733,
+      "grad_norm": 1.3530961275100708,
+      "learning_rate": 3.932202470493029e-06,
+      "loss": 1.9746,
+      "step": 14801
+    },
+    {
+      "epoch": 0.9105561023622047,
+      "grad_norm": 1.158180594444275,
+      "learning_rate": 3.926834882349473e-06,
+      "loss": 2.7951,
+      "step": 14802
+    },
+    {
+      "epoch": 0.9106176181102362,
+      "grad_norm": 1.128093957901001,
+      "learning_rate": 3.921470886806566e-06,
+      "loss": 2.6459,
+      "step": 14803
+    },
+    {
+      "epoch": 0.9106791338582677,
+      "grad_norm": 1.0321640968322754,
+      "learning_rate": 3.916110484064906e-06,
+      "loss": 2.7324,
+      "step": 14804
+    },
+    {
+      "epoch": 0.9107406496062992,
+      "grad_norm": 1.3670142889022827,
+      "learning_rate": 3.91075367432493e-06,
+      "loss": 2.1312,
+      "step": 14805
+    },
+    {
+      "epoch": 0.9108021653543307,
+      "grad_norm": 1.2759501934051514,
+      "learning_rate": 3.9054004577869565e-06,
+      "loss": 2.5279,
+      "step": 14806
+    },
+    {
+      "epoch": 0.9108636811023622,
+      "grad_norm": 1.2857043743133545,
+      "learning_rate": 3.90005083465117e-06,
+      "loss": 1.9165,
+      "step": 14807
+    },
+    {
+      "epoch": 0.9109251968503937,
+      "grad_norm": 1.4286123514175415,
+      "learning_rate": 3.894704805117621e-06,
+      "loss": 2.547,
+      "step": 14808
+    },
+    {
+      "epoch": 0.9109867125984252,
+      "grad_norm": 1.3192908763885498,
+      "learning_rate": 3.889362369386218e-06,
+      "loss": 1.8413,
+      "step": 14809
+    },
+    {
+      "epoch": 0.9110482283464567,
+      "grad_norm": 1.1883831024169922,
+      "learning_rate": 3.8840235276567215e-06,
+      "loss": 2.0251,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9111097440944882,
+      "grad_norm": 1.2771527767181396,
+      "learning_rate": 3.878688280128784e-06,
+      "loss": 2.7708,
+      "step": 14811
+    },
+    {
+      "epoch": 0.9111712598425197,
+      "grad_norm": 1.2664098739624023,
+      "learning_rate": 3.873356627001923e-06,
+      "loss": 1.8648,
+      "step": 14812
+    },
+    {
+      "epoch": 0.9112327755905512,
+      "grad_norm": 1.4332420825958252,
+      "learning_rate": 3.868028568475513e-06,
+      "loss": 2.0665,
+      "step": 14813
+    },
+    {
+      "epoch": 0.9112942913385826,
+      "grad_norm": 1.9316673278808594,
+      "learning_rate": 3.862704104748782e-06,
+      "loss": 2.0747,
+      "step": 14814
+    },
+    {
+      "epoch": 0.9113558070866141,
+      "grad_norm": 1.1833841800689697,
+      "learning_rate": 3.857383236020851e-06,
+      "loss": 1.9287,
+      "step": 14815
+    },
+    {
+      "epoch": 0.9114173228346457,
+      "grad_norm": 1.210323452949524,
+      "learning_rate": 3.85206596249067e-06,
+      "loss": 2.3068,
+      "step": 14816
+    },
+    {
+      "epoch": 0.9114788385826772,
+      "grad_norm": 2.279601573944092,
+      "learning_rate": 3.846752284357092e-06,
+      "loss": 1.5044,
+      "step": 14817
+    },
+    {
+      "epoch": 0.9115403543307087,
+      "grad_norm": 1.5289119482040405,
+      "learning_rate": 3.841442201818813e-06,
+      "loss": 0.9397,
+      "step": 14818
+    },
+    {
+      "epoch": 0.9116018700787402,
+      "grad_norm": 1.3447675704956055,
+      "learning_rate": 3.836135715074407e-06,
+      "loss": 2.4393,
+      "step": 14819
+    },
+    {
+      "epoch": 0.9116633858267716,
+      "grad_norm": 1.0974675416946411,
+      "learning_rate": 3.830832824322284e-06,
+      "loss": 1.2468,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9117249015748031,
+      "grad_norm": 4.336716175079346,
+      "learning_rate": 3.825533529760783e-06,
+      "loss": 1.5602,
+      "step": 14821
+    },
+    {
+      "epoch": 0.9117864173228346,
+      "grad_norm": 1.2768415212631226,
+      "learning_rate": 3.820237831588025e-06,
+      "loss": 1.9566,
+      "step": 14822
+    },
+    {
+      "epoch": 0.9118479330708661,
+      "grad_norm": 1.4816330671310425,
+      "learning_rate": 3.814945730002073e-06,
+      "loss": 2.6417,
+      "step": 14823
+    },
+    {
+      "epoch": 0.9119094488188977,
+      "grad_norm": 1.261085033416748,
+      "learning_rate": 3.809657225200791e-06,
+      "loss": 2.536,
+      "step": 14824
+    },
+    {
+      "epoch": 0.9119709645669292,
+      "grad_norm": 1.413359522819519,
+      "learning_rate": 3.8043723173819655e-06,
+      "loss": 2.9612,
+      "step": 14825
+    },
+    {
+      "epoch": 0.9120324803149606,
+      "grad_norm": 1.7891135215759277,
+      "learning_rate": 3.7990910067432163e-06,
+      "loss": 2.3878,
+      "step": 14826
+    },
+    {
+      "epoch": 0.9120939960629921,
+      "grad_norm": 1.3202098608016968,
+      "learning_rate": 3.7938132934820293e-06,
+      "loss": 2.0676,
+      "step": 14827
+    },
+    {
+      "epoch": 0.9121555118110236,
+      "grad_norm": 1.0210189819335938,
+      "learning_rate": 3.788539177795769e-06,
+      "loss": 1.2848,
+      "step": 14828
+    },
+    {
+      "epoch": 0.9122170275590551,
+      "grad_norm": 1.1684471368789673,
+      "learning_rate": 3.7832686598816446e-06,
+      "loss": 2.5827,
+      "step": 14829
+    },
+    {
+      "epoch": 0.9122785433070866,
+      "grad_norm": 1.4220963716506958,
+      "learning_rate": 3.7780017399367652e-06,
+      "loss": 2.293,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9123400590551181,
+      "grad_norm": 1.244244933128357,
+      "learning_rate": 3.7727384181580616e-06,
+      "loss": 2.4351,
+      "step": 14831
+    },
+    {
+      "epoch": 0.9124015748031497,
+      "grad_norm": 1.2462438344955444,
+      "learning_rate": 3.767478694742366e-06,
+      "loss": 2.3403,
+      "step": 14832
+    },
+    {
+      "epoch": 0.9124630905511811,
+      "grad_norm": 1.1582615375518799,
+      "learning_rate": 3.7622225698863643e-06,
+      "loss": 2.8542,
+      "step": 14833
+    },
+    {
+      "epoch": 0.9125246062992126,
+      "grad_norm": 1.15077805519104,
+      "learning_rate": 3.7569700437866006e-06,
+      "loss": 2.9031,
+      "step": 14834
+    },
+    {
+      "epoch": 0.9125861220472441,
+      "grad_norm": 1.1822367906570435,
+      "learning_rate": 3.751721116639484e-06,
+      "loss": 2.3655,
+      "step": 14835
+    },
+    {
+      "epoch": 0.9126476377952756,
+      "grad_norm": 1.3646268844604492,
+      "learning_rate": 3.7464757886413127e-06,
+      "loss": 2.5552,
+      "step": 14836
+    },
+    {
+      "epoch": 0.9127091535433071,
+      "grad_norm": 1.3656249046325684,
+      "learning_rate": 3.7412340599882078e-06,
+      "loss": 2.559,
+      "step": 14837
+    },
+    {
+      "epoch": 0.9127706692913385,
+      "grad_norm": 1.430585265159607,
+      "learning_rate": 3.7359959308762125e-06,
+      "loss": 2.0307,
+      "step": 14838
+    },
+    {
+      "epoch": 0.91283218503937,
+      "grad_norm": 1.1706851720809937,
+      "learning_rate": 3.7307614015011818e-06,
+      "loss": 2.4382,
+      "step": 14839
+    },
+    {
+      "epoch": 0.9128937007874016,
+      "grad_norm": 1.3610016107559204,
+      "learning_rate": 3.725530472058869e-06,
+      "loss": 2.3442,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9129552165354331,
+      "grad_norm": 1.3129727840423584,
+      "learning_rate": 3.720303142744863e-06,
+      "loss": 2.1835,
+      "step": 14841
+    },
+    {
+      "epoch": 0.9130167322834646,
+      "grad_norm": 1.2437856197357178,
+      "learning_rate": 3.7150794137546517e-06,
+      "loss": 2.6341,
+      "step": 14842
+    },
+    {
+      "epoch": 0.9130782480314961,
+      "grad_norm": 1.4965144395828247,
+      "learning_rate": 3.709859285283557e-06,
+      "loss": 2.403,
+      "step": 14843
+    },
+    {
+      "epoch": 0.9131397637795275,
+      "grad_norm": 1.218025803565979,
+      "learning_rate": 3.7046427575268215e-06,
+      "loss": 2.2068,
+      "step": 14844
+    },
+    {
+      "epoch": 0.913201279527559,
+      "grad_norm": 1.220780372619629,
+      "learning_rate": 3.6994298306794793e-06,
+      "loss": 2.3837,
+      "step": 14845
+    },
+    {
+      "epoch": 0.9132627952755905,
+      "grad_norm": 1.4215481281280518,
+      "learning_rate": 3.694220504936485e-06,
+      "loss": 2.4911,
+      "step": 14846
+    },
+    {
+      "epoch": 0.913324311023622,
+      "grad_norm": 1.3588038682937622,
+      "learning_rate": 3.6890147804926056e-06,
+      "loss": 2.5662,
+      "step": 14847
+    },
+    {
+      "epoch": 0.9133858267716536,
+      "grad_norm": 1.3849800825119019,
+      "learning_rate": 3.683812657542551e-06,
+      "loss": 2.5238,
+      "step": 14848
+    },
+    {
+      "epoch": 0.9134473425196851,
+      "grad_norm": 1.4534624814987183,
+      "learning_rate": 3.6786141362808226e-06,
+      "loss": 1.9979,
+      "step": 14849
+    },
+    {
+      "epoch": 0.9135088582677166,
+      "grad_norm": 1.282839059829712,
+      "learning_rate": 3.6734192169018077e-06,
+      "loss": 2.4238,
+      "step": 14850
+    },
+    {
+      "epoch": 0.913570374015748,
+      "grad_norm": 1.5266846418380737,
+      "learning_rate": 3.6682278995997967e-06,
+      "loss": 2.2879,
+      "step": 14851
+    },
+    {
+      "epoch": 0.9136318897637795,
+      "grad_norm": 1.3013076782226562,
+      "learning_rate": 3.6630401845689e-06,
+      "loss": 3.0869,
+      "step": 14852
+    },
+    {
+      "epoch": 0.913693405511811,
+      "grad_norm": 1.3238890171051025,
+      "learning_rate": 3.6578560720031074e-06,
+      "loss": 2.5994,
+      "step": 14853
+    },
+    {
+      "epoch": 0.9137549212598425,
+      "grad_norm": 1.3743644952774048,
+      "learning_rate": 3.652675562096286e-06,
+      "loss": 1.8851,
+      "step": 14854
+    },
+    {
+      "epoch": 0.913816437007874,
+      "grad_norm": 1.2397602796554565,
+      "learning_rate": 3.6474986550421476e-06,
+      "loss": 2.8703,
+      "step": 14855
+    },
+    {
+      "epoch": 0.9138779527559056,
+      "grad_norm": 1.3326243162155151,
+      "learning_rate": 3.6423253510342815e-06,
+      "loss": 2.3221,
+      "step": 14856
+    },
+    {
+      "epoch": 0.913939468503937,
+      "grad_norm": 1.5964508056640625,
+      "learning_rate": 3.637155650266133e-06,
+      "loss": 2.5141,
+      "step": 14857
+    },
+    {
+      "epoch": 0.9140009842519685,
+      "grad_norm": 1.2330305576324463,
+      "learning_rate": 3.6319895529310475e-06,
+      "loss": 1.8832,
+      "step": 14858
+    },
+    {
+      "epoch": 0.9140625,
+      "grad_norm": 1.2258371114730835,
+      "learning_rate": 3.626827059222182e-06,
+      "loss": 2.6197,
+      "step": 14859
+    },
+    {
+      "epoch": 0.9141240157480315,
+      "grad_norm": 1.2549926042556763,
+      "learning_rate": 3.6216681693325926e-06,
+      "loss": 2.3425,
+      "step": 14860
+    },
+    {
+      "epoch": 0.914185531496063,
+      "grad_norm": 1.3400713205337524,
+      "learning_rate": 3.616512883455203e-06,
+      "loss": 2.5369,
+      "step": 14861
+    },
+    {
+      "epoch": 0.9142470472440944,
+      "grad_norm": 1.3550245761871338,
+      "learning_rate": 3.6113612017827593e-06,
+      "loss": 2.1074,
+      "step": 14862
+    },
+    {
+      "epoch": 0.914308562992126,
+      "grad_norm": 1.279319405555725,
+      "learning_rate": 3.6062131245079513e-06,
+      "loss": 2.1567,
+      "step": 14863
+    },
+    {
+      "epoch": 0.9143700787401575,
+      "grad_norm": 1.2893837690353394,
+      "learning_rate": 3.6010686518232474e-06,
+      "loss": 2.4223,
+      "step": 14864
+    },
+    {
+      "epoch": 0.914431594488189,
+      "grad_norm": 1.5926761627197266,
+      "learning_rate": 3.5959277839210605e-06,
+      "loss": 2.2731,
+      "step": 14865
+    },
+    {
+      "epoch": 0.9144931102362205,
+      "grad_norm": 1.1471095085144043,
+      "learning_rate": 3.5907905209936034e-06,
+      "loss": 2.5855,
+      "step": 14866
+    },
+    {
+      "epoch": 0.914554625984252,
+      "grad_norm": 1.3749594688415527,
+      "learning_rate": 3.585656863232989e-06,
+      "loss": 2.3838,
+      "step": 14867
+    },
+    {
+      "epoch": 0.9146161417322834,
+      "grad_norm": 1.3469007015228271,
+      "learning_rate": 3.580526810831175e-06,
+      "loss": 2.3181,
+      "step": 14868
+    },
+    {
+      "epoch": 0.9146776574803149,
+      "grad_norm": 1.452276349067688,
+      "learning_rate": 3.5754003639800195e-06,
+      "loss": 2.2051,
+      "step": 14869
+    },
+    {
+      "epoch": 0.9147391732283464,
+      "grad_norm": 1.3750723600387573,
+      "learning_rate": 3.5702775228712017e-06,
+      "loss": 1.5854,
+      "step": 14870
+    },
+    {
+      "epoch": 0.914800688976378,
+      "grad_norm": 1.360840916633606,
+      "learning_rate": 3.565158287696313e-06,
+      "loss": 2.283,
+      "step": 14871
+    },
+    {
+      "epoch": 0.9148622047244095,
+      "grad_norm": 0.9484022855758667,
+      "learning_rate": 3.560042658646756e-06,
+      "loss": 1.3523,
+      "step": 14872
+    },
+    {
+      "epoch": 0.914923720472441,
+      "grad_norm": 1.0661417245864868,
+      "learning_rate": 3.5549306359138557e-06,
+      "loss": 1.9543,
+      "step": 14873
+    },
+    {
+      "epoch": 0.9149852362204725,
+      "grad_norm": 1.6220300197601318,
+      "learning_rate": 3.5498222196887366e-06,
+      "loss": 2.01,
+      "step": 14874
+    },
+    {
+      "epoch": 0.9150467519685039,
+      "grad_norm": 1.3706786632537842,
+      "learning_rate": 3.544717410162457e-06,
+      "loss": 2.4491,
+      "step": 14875
+    },
+    {
+      "epoch": 0.9151082677165354,
+      "grad_norm": 1.4976423978805542,
+      "learning_rate": 3.5396162075258863e-06,
+      "loss": 1.7556,
+      "step": 14876
+    },
+    {
+      "epoch": 0.9151697834645669,
+      "grad_norm": 1.272991418838501,
+      "learning_rate": 3.5345186119698172e-06,
+      "loss": 2.437,
+      "step": 14877
+    },
+    {
+      "epoch": 0.9152312992125984,
+      "grad_norm": 1.3229100704193115,
+      "learning_rate": 3.529424623684818e-06,
+      "loss": 2.2905,
+      "step": 14878
+    },
+    {
+      "epoch": 0.91529281496063,
+      "grad_norm": 1.178420901298523,
+      "learning_rate": 3.5243342428614268e-06,
+      "loss": 2.1352,
+      "step": 14879
+    },
+    {
+      "epoch": 0.9153543307086615,
+      "grad_norm": 1.3431545495986938,
+      "learning_rate": 3.519247469689957e-06,
+      "loss": 2.5856,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9154158464566929,
+      "grad_norm": 1.3094093799591064,
+      "learning_rate": 3.5141643043606343e-06,
+      "loss": 2.2899,
+      "step": 14881
+    },
+    {
+      "epoch": 0.9154773622047244,
+      "grad_norm": 1.1875661611557007,
+      "learning_rate": 3.5090847470635514e-06,
+      "loss": 2.902,
+      "step": 14882
+    },
+    {
+      "epoch": 0.9155388779527559,
+      "grad_norm": 1.3382484912872314,
+      "learning_rate": 3.504008797988656e-06,
+      "loss": 2.7229,
+      "step": 14883
+    },
+    {
+      "epoch": 0.9156003937007874,
+      "grad_norm": 1.5001813173294067,
+      "learning_rate": 3.498936457325752e-06,
+      "loss": 2.8168,
+      "step": 14884
+    },
+    {
+      "epoch": 0.9156619094488189,
+      "grad_norm": 1.330907940864563,
+      "learning_rate": 3.4938677252645325e-06,
+      "loss": 2.5846,
+      "step": 14885
+    },
+    {
+      "epoch": 0.9157234251968503,
+      "grad_norm": 1.5708725452423096,
+      "learning_rate": 3.488802601994512e-06,
+      "loss": 3.1195,
+      "step": 14886
+    },
+    {
+      "epoch": 0.9157849409448819,
+      "grad_norm": 1.4944894313812256,
+      "learning_rate": 3.483741087705128e-06,
+      "loss": 2.5579,
+      "step": 14887
+    },
+    {
+      "epoch": 0.9158464566929134,
+      "grad_norm": 1.1453076601028442,
+      "learning_rate": 3.478683182585607e-06,
+      "loss": 2.4121,
+      "step": 14888
+    },
+    {
+      "epoch": 0.9159079724409449,
+      "grad_norm": 1.461733102798462,
+      "learning_rate": 3.4736288868251423e-06,
+      "loss": 2.1139,
+      "step": 14889
+    },
+    {
+      "epoch": 0.9159694881889764,
+      "grad_norm": 1.2411220073699951,
+      "learning_rate": 3.4685782006127045e-06,
+      "loss": 2.9623,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9160310039370079,
+      "grad_norm": 1.316239833831787,
+      "learning_rate": 3.463531124137176e-06,
+      "loss": 2.2261,
+      "step": 14891
+    },
+    {
+      "epoch": 0.9160925196850394,
+      "grad_norm": 1.2944642305374146,
+      "learning_rate": 3.4584876575872725e-06,
+      "loss": 2.8933,
+      "step": 14892
+    },
+    {
+      "epoch": 0.9161540354330708,
+      "grad_norm": 1.205958604812622,
+      "learning_rate": 3.4534478011516103e-06,
+      "loss": 2.4002,
+      "step": 14893
+    },
+    {
+      "epoch": 0.9162155511811023,
+      "grad_norm": 1.586063027381897,
+      "learning_rate": 3.448411555018638e-06,
+      "loss": 2.0429,
+      "step": 14894
+    },
+    {
+      "epoch": 0.9162770669291339,
+      "grad_norm": 1.7181259393692017,
+      "learning_rate": 3.443378919376672e-06,
+      "loss": 2.3382,
+      "step": 14895
+    },
+    {
+      "epoch": 0.9163385826771654,
+      "grad_norm": 1.6940549612045288,
+      "learning_rate": 3.438349894413928e-06,
+      "loss": 2.8626,
+      "step": 14896
+    },
+    {
+      "epoch": 0.9164000984251969,
+      "grad_norm": 2.2654452323913574,
+      "learning_rate": 3.433324480318467e-06,
+      "loss": 2.1526,
+      "step": 14897
+    },
+    {
+      "epoch": 0.9164616141732284,
+      "grad_norm": 1.1904348134994507,
+      "learning_rate": 3.4283026772782056e-06,
+      "loss": 2.2273,
+      "step": 14898
+    },
+    {
+      "epoch": 0.9165231299212598,
+      "grad_norm": 1.226873517036438,
+      "learning_rate": 3.4232844854809045e-06,
+      "loss": 2.3885,
+      "step": 14899
+    },
+    {
+      "epoch": 0.9165846456692913,
+      "grad_norm": 1.3174716234207153,
+      "learning_rate": 3.4182699051142465e-06,
+      "loss": 2.4802,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9166461614173228,
+      "grad_norm": 1.159097671508789,
+      "learning_rate": 3.4132589363657265e-06,
+      "loss": 2.6998,
+      "step": 14901
+    },
+    {
+      "epoch": 0.9167076771653543,
+      "grad_norm": 1.1498242616653442,
+      "learning_rate": 3.4082515794227608e-06,
+      "loss": 2.649,
+      "step": 14902
+    },
+    {
+      "epoch": 0.9167691929133859,
+      "grad_norm": 1.3734033107757568,
+      "learning_rate": 3.4032478344725673e-06,
+      "loss": 2.4523,
+      "step": 14903
+    },
+    {
+      "epoch": 0.9168307086614174,
+      "grad_norm": 1.0398086309432983,
+      "learning_rate": 3.398247701702273e-06,
+      "loss": 2.7038,
+      "step": 14904
+    },
+    {
+      "epoch": 0.9168922244094488,
+      "grad_norm": 1.33258056640625,
+      "learning_rate": 3.3932511812988286e-06,
+      "loss": 2.6925,
+      "step": 14905
+    },
+    {
+      "epoch": 0.9169537401574803,
+      "grad_norm": 1.2219699621200562,
+      "learning_rate": 3.3882582734491074e-06,
+      "loss": 2.7372,
+      "step": 14906
+    },
+    {
+      "epoch": 0.9170152559055118,
+      "grad_norm": 1.503219723701477,
+      "learning_rate": 3.383268978339804e-06,
+      "loss": 2.5183,
+      "step": 14907
+    },
+    {
+      "epoch": 0.9170767716535433,
+      "grad_norm": 1.3425105810165405,
+      "learning_rate": 3.3782832961575027e-06,
+      "loss": 2.6728,
+      "step": 14908
+    },
+    {
+      "epoch": 0.9171382874015748,
+      "grad_norm": 1.1429001092910767,
+      "learning_rate": 3.37330122708861e-06,
+      "loss": 2.7708,
+      "step": 14909
+    },
+    {
+      "epoch": 0.9171998031496063,
+      "grad_norm": 1.11369788646698,
+      "learning_rate": 3.3683227713194654e-06,
+      "loss": 1.808,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9172613188976378,
+      "grad_norm": 1.142833948135376,
+      "learning_rate": 3.363347929036198e-06,
+      "loss": 2.8716,
+      "step": 14911
+    },
+    {
+      "epoch": 0.9173228346456693,
+      "grad_norm": 1.1605851650238037,
+      "learning_rate": 3.358376700424881e-06,
+      "loss": 2.7459,
+      "step": 14912
+    },
+    {
+      "epoch": 0.9173843503937008,
+      "grad_norm": 1.1224573850631714,
+      "learning_rate": 3.353409085671366e-06,
+      "loss": 2.8307,
+      "step": 14913
+    },
+    {
+      "epoch": 0.9174458661417323,
+      "grad_norm": 1.3408676385879517,
+      "learning_rate": 3.348445084961438e-06,
+      "loss": 2.5254,
+      "step": 14914
+    },
+    {
+      "epoch": 0.9175073818897638,
+      "grad_norm": 1.448216199874878,
+      "learning_rate": 3.3434846984807145e-06,
+      "loss": 2.8199,
+      "step": 14915
+    },
+    {
+      "epoch": 0.9175688976377953,
+      "grad_norm": 1.3381648063659668,
+      "learning_rate": 3.338527926414703e-06,
+      "loss": 2.8852,
+      "step": 14916
+    },
+    {
+      "epoch": 0.9176304133858267,
+      "grad_norm": 1.2103426456451416,
+      "learning_rate": 3.333574768948744e-06,
+      "loss": 2.544,
+      "step": 14917
+    },
+    {
+      "epoch": 0.9176919291338582,
+      "grad_norm": 1.473357081413269,
+      "learning_rate": 3.3286252262680673e-06,
+      "loss": 2.8561,
+      "step": 14918
+    },
+    {
+      "epoch": 0.9177534448818898,
+      "grad_norm": 1.3973251581192017,
+      "learning_rate": 3.3236792985577468e-06,
+      "loss": 2.3656,
+      "step": 14919
+    },
+    {
+      "epoch": 0.9178149606299213,
+      "grad_norm": 1.314427375793457,
+      "learning_rate": 3.3187369860027353e-06,
+      "loss": 2.1166,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9178764763779528,
+      "grad_norm": 1.127894639968872,
+      "learning_rate": 3.3137982887878504e-06,
+      "loss": 2.7149,
+      "step": 14921
+    },
+    {
+      "epoch": 0.9179379921259843,
+      "grad_norm": 1.5698391199111938,
+      "learning_rate": 3.3088632070977788e-06,
+      "loss": 2.5152,
+      "step": 14922
+    },
+    {
+      "epoch": 0.9179995078740157,
+      "grad_norm": 1.2401435375213623,
+      "learning_rate": 3.3039317411170613e-06,
+      "loss": 1.9204,
+      "step": 14923
+    },
+    {
+      "epoch": 0.9180610236220472,
+      "grad_norm": 1.5835529565811157,
+      "learning_rate": 3.2990038910301056e-06,
+      "loss": 2.6545,
+      "step": 14924
+    },
+    {
+      "epoch": 0.9181225393700787,
+      "grad_norm": 1.243470549583435,
+      "learning_rate": 3.2940796570211764e-06,
+      "loss": 3.0385,
+      "step": 14925
+    },
+    {
+      "epoch": 0.9181840551181102,
+      "grad_norm": 1.3221266269683838,
+      "learning_rate": 3.289159039274414e-06,
+      "loss": 2.2768,
+      "step": 14926
+    },
+    {
+      "epoch": 0.9182455708661418,
+      "grad_norm": 1.2065258026123047,
+      "learning_rate": 3.284242037973839e-06,
+      "loss": 2.5271,
+      "step": 14927
+    },
+    {
+      "epoch": 0.9183070866141733,
+      "grad_norm": 1.209539532661438,
+      "learning_rate": 3.279328653303315e-06,
+      "loss": 2.2147,
+      "step": 14928
+    },
+    {
+      "epoch": 0.9183686023622047,
+      "grad_norm": 1.0978971719741821,
+      "learning_rate": 3.274418885446562e-06,
+      "loss": 2.74,
+      "step": 14929
+    },
+    {
+      "epoch": 0.9184301181102362,
+      "grad_norm": 1.1435068845748901,
+      "learning_rate": 3.269512734587199e-06,
+      "loss": 2.9082,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9184916338582677,
+      "grad_norm": 1.2869057655334473,
+      "learning_rate": 3.2646102009086688e-06,
+      "loss": 2.172,
+      "step": 14931
+    },
+    {
+      "epoch": 0.9185531496062992,
+      "grad_norm": 1.2679330110549927,
+      "learning_rate": 3.259711284594291e-06,
+      "loss": 2.5457,
+      "step": 14932
+    },
+    {
+      "epoch": 0.9186146653543307,
+      "grad_norm": 1.5072489976882935,
+      "learning_rate": 3.2548159858272975e-06,
+      "loss": 2.2185,
+      "step": 14933
+    },
+    {
+      "epoch": 0.9186761811023622,
+      "grad_norm": 1.3625460863113403,
+      "learning_rate": 3.249924304790697e-06,
+      "loss": 2.1164,
+      "step": 14934
+    },
+    {
+      "epoch": 0.9187376968503937,
+      "grad_norm": 1.3004515171051025,
+      "learning_rate": 3.245036241667454e-06,
+      "loss": 2.3023,
+      "step": 14935
+    },
+    {
+      "epoch": 0.9187992125984252,
+      "grad_norm": 2.825190305709839,
+      "learning_rate": 3.2401517966403226e-06,
+      "loss": 2.4618,
+      "step": 14936
+    },
+    {
+      "epoch": 0.9188607283464567,
+      "grad_norm": 1.8050260543823242,
+      "learning_rate": 3.2352709698919793e-06,
+      "loss": 2.2059,
+      "step": 14937
+    },
+    {
+      "epoch": 0.9189222440944882,
+      "grad_norm": 1.2965973615646362,
+      "learning_rate": 3.230393761604911e-06,
+      "loss": 2.3126,
+      "step": 14938
+    },
+    {
+      "epoch": 0.9189837598425197,
+      "grad_norm": 1.2105101346969604,
+      "learning_rate": 3.2255201719615157e-06,
+      "loss": 2.659,
+      "step": 14939
+    },
+    {
+      "epoch": 0.9190452755905512,
+      "grad_norm": 1.137555718421936,
+      "learning_rate": 3.2206502011440375e-06,
+      "loss": 2.6609,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9191067913385826,
+      "grad_norm": 1.4134135246276855,
+      "learning_rate": 3.215783849334597e-06,
+      "loss": 2.5792,
+      "step": 14941
+    },
+    {
+      "epoch": 0.9191683070866141,
+      "grad_norm": 1.1993614435195923,
+      "learning_rate": 3.2109211167151486e-06,
+      "loss": 2.4959,
+      "step": 14942
+    },
+    {
+      "epoch": 0.9192298228346457,
+      "grad_norm": 1.281904935836792,
+      "learning_rate": 3.206062003467547e-06,
+      "loss": 2.1666,
+      "step": 14943
+    },
+    {
+      "epoch": 0.9192913385826772,
+      "grad_norm": 1.2631081342697144,
+      "learning_rate": 3.2012065097734687e-06,
+      "loss": 2.3977,
+      "step": 14944
+    },
+    {
+      "epoch": 0.9193528543307087,
+      "grad_norm": 1.4289640188217163,
+      "learning_rate": 3.1963546358145135e-06,
+      "loss": 2.6279,
+      "step": 14945
+    },
+    {
+      "epoch": 0.9194143700787402,
+      "grad_norm": 1.376168966293335,
+      "learning_rate": 3.1915063817721023e-06,
+      "loss": 2.1334,
+      "step": 14946
+    },
+    {
+      "epoch": 0.9194758858267716,
+      "grad_norm": 1.389813780784607,
+      "learning_rate": 3.1866617478275463e-06,
+      "loss": 1.9341,
+      "step": 14947
+    },
+    {
+      "epoch": 0.9195374015748031,
+      "grad_norm": 1.3098649978637695,
+      "learning_rate": 3.1818207341619775e-06,
+      "loss": 2.0469,
+      "step": 14948
+    },
+    {
+      "epoch": 0.9195989173228346,
+      "grad_norm": 1.2163838148117065,
+      "learning_rate": 3.176983340956452e-06,
+      "loss": 1.8747,
+      "step": 14949
+    },
+    {
+      "epoch": 0.9196604330708661,
+      "grad_norm": 1.1591484546661377,
+      "learning_rate": 3.1721495683918467e-06,
+      "loss": 2.684,
+      "step": 14950
+    },
+    {
+      "epoch": 0.9197219488188977,
+      "grad_norm": 1.2738840579986572,
+      "learning_rate": 3.1673194166489063e-06,
+      "loss": 2.6815,
+      "step": 14951
+    },
+    {
+      "epoch": 0.9197834645669292,
+      "grad_norm": 1.3847320079803467,
+      "learning_rate": 3.162492885908275e-06,
+      "loss": 2.902,
+      "step": 14952
+    },
+    {
+      "epoch": 0.9198449803149606,
+      "grad_norm": 1.5000911951065063,
+      "learning_rate": 3.1576699763504415e-06,
+      "loss": 2.5389,
+      "step": 14953
+    },
+    {
+      "epoch": 0.9199064960629921,
+      "grad_norm": 1.4120088815689087,
+      "learning_rate": 3.1528506881557285e-06,
+      "loss": 2.4314,
+      "step": 14954
+    },
+    {
+      "epoch": 0.9199680118110236,
+      "grad_norm": 1.3306649923324585,
+      "learning_rate": 3.1480350215043807e-06,
+      "loss": 2.0182,
+      "step": 14955
+    },
+    {
+      "epoch": 0.9200295275590551,
+      "grad_norm": 3.423346996307373,
+      "learning_rate": 3.1432229765764433e-06,
+      "loss": 1.4128,
+      "step": 14956
+    },
+    {
+      "epoch": 0.9200910433070866,
+      "grad_norm": 1.0798507928848267,
+      "learning_rate": 3.1384145535518937e-06,
+      "loss": 2.5337,
+      "step": 14957
+    },
+    {
+      "epoch": 0.9201525590551181,
+      "grad_norm": 1.2227872610092163,
+      "learning_rate": 3.1336097526105e-06,
+      "loss": 3.0841,
+      "step": 14958
+    },
+    {
+      "epoch": 0.9202140748031497,
+      "grad_norm": 1.2526723146438599,
+      "learning_rate": 3.128808573931985e-06,
+      "loss": 1.8601,
+      "step": 14959
+    },
+    {
+      "epoch": 0.9202755905511811,
+      "grad_norm": 1.18284010887146,
+      "learning_rate": 3.124011017695849e-06,
+      "loss": 2.6096,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9203371062992126,
+      "grad_norm": 1.2156131267547607,
+      "learning_rate": 3.119217084081527e-06,
+      "loss": 2.7487,
+      "step": 14961
+    },
+    {
+      "epoch": 0.9203986220472441,
+      "grad_norm": 1.3333009481430054,
+      "learning_rate": 3.114426773268242e-06,
+      "loss": 2.1544,
+      "step": 14962
+    },
+    {
+      "epoch": 0.9204601377952756,
+      "grad_norm": 1.5726362466812134,
+      "learning_rate": 3.1096400854351503e-06,
+      "loss": 2.3536,
+      "step": 14963
+    },
+    {
+      "epoch": 0.9205216535433071,
+      "grad_norm": 1.1999744176864624,
+      "learning_rate": 3.1048570207612426e-06,
+      "loss": 2.7796,
+      "step": 14964
+    },
+    {
+      "epoch": 0.9205831692913385,
+      "grad_norm": 1.2202249765396118,
+      "learning_rate": 3.1000775794253866e-06,
+      "loss": 3.0988,
+      "step": 14965
+    },
+    {
+      "epoch": 0.92064468503937,
+      "grad_norm": 1.3737030029296875,
+      "learning_rate": 3.095301761606284e-06,
+      "loss": 2.3429,
+      "step": 14966
+    },
+    {
+      "epoch": 0.9207062007874016,
+      "grad_norm": 1.2969704866409302,
+      "learning_rate": 3.0905295674825696e-06,
+      "loss": 2.8182,
+      "step": 14967
+    },
+    {
+      "epoch": 0.9207677165354331,
+      "grad_norm": 1.3713687658309937,
+      "learning_rate": 3.0857609972326453e-06,
+      "loss": 2.2477,
+      "step": 14968
+    },
+    {
+      "epoch": 0.9208292322834646,
+      "grad_norm": 1.2285069227218628,
+      "learning_rate": 3.0809960510348567e-06,
+      "loss": 1.5708,
+      "step": 14969
+    },
+    {
+      "epoch": 0.9208907480314961,
+      "grad_norm": 2.282942295074463,
+      "learning_rate": 3.0762347290673844e-06,
+      "loss": 1.619,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9209522637795275,
+      "grad_norm": 1.0067511796951294,
+      "learning_rate": 3.071477031508263e-06,
+      "loss": 2.6626,
+      "step": 14971
+    },
+    {
+      "epoch": 0.921013779527559,
+      "grad_norm": 1.2261677980422974,
+      "learning_rate": 3.0667229585354174e-06,
+      "loss": 2.4816,
+      "step": 14972
+    },
+    {
+      "epoch": 0.9210752952755905,
+      "grad_norm": 1.3047438859939575,
+      "learning_rate": 3.061972510326616e-06,
+      "loss": 2.1793,
+      "step": 14973
+    },
+    {
+      "epoch": 0.921136811023622,
+      "grad_norm": 1.4523539543151855,
+      "learning_rate": 3.057225687059506e-06,
+      "loss": 2.1863,
+      "step": 14974
+    },
+    {
+      "epoch": 0.9211983267716536,
+      "grad_norm": 1.5482515096664429,
+      "learning_rate": 3.05248248891159e-06,
+      "loss": 2.2377,
+      "step": 14975
+    },
+    {
+      "epoch": 0.9212598425196851,
+      "grad_norm": 1.2791473865509033,
+      "learning_rate": 3.0477429160602366e-06,
+      "loss": 2.4612,
+      "step": 14976
+    },
+    {
+      "epoch": 0.9213213582677166,
+      "grad_norm": 1.287775993347168,
+      "learning_rate": 3.043006968682649e-06,
+      "loss": 2.3121,
+      "step": 14977
+    },
+    {
+      "epoch": 0.921382874015748,
+      "grad_norm": 1.1732624769210815,
+      "learning_rate": 3.0382746469559854e-06,
+      "loss": 2.7964,
+      "step": 14978
+    },
+    {
+      "epoch": 0.9214443897637795,
+      "grad_norm": 1.2135237455368042,
+      "learning_rate": 3.0335459510571597e-06,
+      "loss": 2.444,
+      "step": 14979
+    },
+    {
+      "epoch": 0.921505905511811,
+      "grad_norm": 1.2233564853668213,
+      "learning_rate": 3.0288208811630305e-06,
+      "loss": 1.678,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9215674212598425,
+      "grad_norm": 1.2243993282318115,
+      "learning_rate": 3.0240994374502674e-06,
+      "loss": 2.1016,
+      "step": 14981
+    },
+    {
+      "epoch": 0.921628937007874,
+      "grad_norm": 1.2018227577209473,
+      "learning_rate": 3.0193816200954294e-06,
+      "loss": 2.6564,
+      "step": 14982
+    },
+    {
+      "epoch": 0.9216904527559056,
+      "grad_norm": 3.1180307865142822,
+      "learning_rate": 3.0146674292749423e-06,
+      "loss": 3.8543,
+      "step": 14983
+    },
+    {
+      "epoch": 0.921751968503937,
+      "grad_norm": 1.3626888990402222,
+      "learning_rate": 3.009956865165087e-06,
+      "loss": 3.1799,
+      "step": 14984
+    },
+    {
+      "epoch": 0.9218134842519685,
+      "grad_norm": 1.1972389221191406,
+      "learning_rate": 3.005249927942011e-06,
+      "loss": 2.8437,
+      "step": 14985
+    },
+    {
+      "epoch": 0.921875,
+      "grad_norm": 1.0872361660003662,
+      "learning_rate": 3.000546617781741e-06,
+      "loss": 1.8535,
+      "step": 14986
+    },
+    {
+      "epoch": 0.9219365157480315,
+      "grad_norm": 1.2778534889221191,
+      "learning_rate": 2.995846934860136e-06,
+      "loss": 2.2029,
+      "step": 14987
+    },
+    {
+      "epoch": 0.921998031496063,
+      "grad_norm": 1.2036457061767578,
+      "learning_rate": 2.9911508793529443e-06,
+      "loss": 2.3411,
+      "step": 14988
+    },
+    {
+      "epoch": 0.9220595472440944,
+      "grad_norm": 1.6229556798934937,
+      "learning_rate": 2.986458451435781e-06,
+      "loss": 2.3918,
+      "step": 14989
+    },
+    {
+      "epoch": 0.922121062992126,
+      "grad_norm": 1.3630669116973877,
+      "learning_rate": 2.981769651284094e-06,
+      "loss": 2.3414,
+      "step": 14990
+    },
+    {
+      "epoch": 0.9221825787401575,
+      "grad_norm": 1.2053226232528687,
+      "learning_rate": 2.9770844790732446e-06,
+      "loss": 2.4086,
+      "step": 14991
+    },
+    {
+      "epoch": 0.922244094488189,
+      "grad_norm": 1.3762578964233398,
+      "learning_rate": 2.9724029349784242e-06,
+      "loss": 2.8633,
+      "step": 14992
+    },
+    {
+      "epoch": 0.9223056102362205,
+      "grad_norm": 1.0957151651382446,
+      "learning_rate": 2.967725019174683e-06,
+      "loss": 2.5328,
+      "step": 14993
+    },
+    {
+      "epoch": 0.922367125984252,
+      "grad_norm": 1.2161085605621338,
+      "learning_rate": 2.9630507318369695e-06,
+      "loss": 2.4133,
+      "step": 14994
+    },
+    {
+      "epoch": 0.9224286417322834,
+      "grad_norm": 1.2673746347427368,
+      "learning_rate": 2.9583800731400657e-06,
+      "loss": 2.2225,
+      "step": 14995
+    },
+    {
+      "epoch": 0.9224901574803149,
+      "grad_norm": 1.4829912185668945,
+      "learning_rate": 2.9537130432586104e-06,
+      "loss": 2.7283,
+      "step": 14996
+    },
+    {
+      "epoch": 0.9225516732283464,
+      "grad_norm": 1.3728923797607422,
+      "learning_rate": 2.9490496423671523e-06,
+      "loss": 2.2563,
+      "step": 14997
+    },
+    {
+      "epoch": 0.922613188976378,
+      "grad_norm": 1.1942042112350464,
+      "learning_rate": 2.944389870640085e-06,
+      "loss": 2.5473,
+      "step": 14998
+    },
+    {
+      "epoch": 0.9226747047244095,
+      "grad_norm": 1.3309508562088013,
+      "learning_rate": 2.9397337282516145e-06,
+      "loss": 2.8466,
+      "step": 14999
+    },
+    {
+      "epoch": 0.922736220472441,
+      "grad_norm": 1.2300019264221191,
+      "learning_rate": 2.935081215375901e-06,
+      "loss": 2.3194,
+      "step": 15000
+    },
+    {
+      "epoch": 0.9227977362204725,
+      "grad_norm": 1.2234553098678589,
+      "learning_rate": 2.9304323321868833e-06,
+      "loss": 2.4898,
+      "step": 15001
+    },
+    {
+      "epoch": 0.9228592519685039,
+      "grad_norm": 1.2454698085784912,
+      "learning_rate": 2.925787078858433e-06,
+      "loss": 2.2168,
+      "step": 15002
+    },
+    {
+      "epoch": 0.9229207677165354,
+      "grad_norm": 1.3029184341430664,
+      "learning_rate": 2.9211454555642337e-06,
+      "loss": 2.4249,
+      "step": 15003
+    },
+    {
+      "epoch": 0.9229822834645669,
+      "grad_norm": 1.455735683441162,
+      "learning_rate": 2.91650746247788e-06,
+      "loss": 2.5922,
+      "step": 15004
+    },
+    {
+      "epoch": 0.9230437992125984,
+      "grad_norm": 1.2576669454574585,
+      "learning_rate": 2.9118730997727883e-06,
+      "loss": 2.7151,
+      "step": 15005
+    },
+    {
+      "epoch": 0.92310531496063,
+      "grad_norm": 1.2355424165725708,
+      "learning_rate": 2.907242367622276e-06,
+      "loss": 2.734,
+      "step": 15006
+    },
+    {
+      "epoch": 0.9231668307086615,
+      "grad_norm": 1.4352236986160278,
+      "learning_rate": 2.9026152661994823e-06,
+      "loss": 2.2102,
+      "step": 15007
+    },
+    {
+      "epoch": 0.9232283464566929,
+      "grad_norm": 1.226032018661499,
+      "learning_rate": 2.897991795677446e-06,
+      "loss": 2.607,
+      "step": 15008
+    },
+    {
+      "epoch": 0.9232898622047244,
+      "grad_norm": 1.3209402561187744,
+      "learning_rate": 2.8933719562290627e-06,
+      "loss": 2.4265,
+      "step": 15009
+    },
+    {
+      "epoch": 0.9233513779527559,
+      "grad_norm": 1.486373782157898,
+      "learning_rate": 2.8887557480270942e-06,
+      "loss": 2.9095,
+      "step": 15010
+    },
+    {
+      "epoch": 0.9234128937007874,
+      "grad_norm": 1.5048071146011353,
+      "learning_rate": 2.8841431712441467e-06,
+      "loss": 2.8233,
+      "step": 15011
+    },
+    {
+      "epoch": 0.9234744094488189,
+      "grad_norm": 1.2480140924453735,
+      "learning_rate": 2.879534226052716e-06,
+      "loss": 2.8506,
+      "step": 15012
+    },
+    {
+      "epoch": 0.9235359251968503,
+      "grad_norm": 1.204193115234375,
+      "learning_rate": 2.8749289126251523e-06,
+      "loss": 1.7521,
+      "step": 15013
+    },
+    {
+      "epoch": 0.9235974409448819,
+      "grad_norm": 9.499128341674805,
+      "learning_rate": 2.8703272311336517e-06,
+      "loss": 2.0013,
+      "step": 15014
+    },
+    {
+      "epoch": 0.9236589566929134,
+      "grad_norm": 1.355277180671692,
+      "learning_rate": 2.8657291817502986e-06,
+      "loss": 2.2719,
+      "step": 15015
+    },
+    {
+      "epoch": 0.9237204724409449,
+      "grad_norm": 1.1917270421981812,
+      "learning_rate": 2.861134764647044e-06,
+      "loss": 2.3382,
+      "step": 15016
+    },
+    {
+      "epoch": 0.9237819881889764,
+      "grad_norm": 1.2488659620285034,
+      "learning_rate": 2.856543979995696e-06,
+      "loss": 2.4451,
+      "step": 15017
+    },
+    {
+      "epoch": 0.9238435039370079,
+      "grad_norm": 1.2410706281661987,
+      "learning_rate": 2.8519568279679056e-06,
+      "loss": 2.4511,
+      "step": 15018
+    },
+    {
+      "epoch": 0.9239050196850394,
+      "grad_norm": 1.0903749465942383,
+      "learning_rate": 2.847373308735235e-06,
+      "loss": 2.7995,
+      "step": 15019
+    },
+    {
+      "epoch": 0.9239665354330708,
+      "grad_norm": 1.2177677154541016,
+      "learning_rate": 2.8427934224690477e-06,
+      "loss": 2.4451,
+      "step": 15020
+    },
+    {
+      "epoch": 0.9240280511811023,
+      "grad_norm": 1.3279011249542236,
+      "learning_rate": 2.838217169340629e-06,
+      "loss": 2.6676,
+      "step": 15021
+    },
+    {
+      "epoch": 0.9240895669291339,
+      "grad_norm": 1.267417073249817,
+      "learning_rate": 2.833644549521086e-06,
+      "loss": 2.3469,
+      "step": 15022
+    },
+    {
+      "epoch": 0.9241510826771654,
+      "grad_norm": 1.4050756692886353,
+      "learning_rate": 2.829075563181449e-06,
+      "loss": 2.5516,
+      "step": 15023
+    },
+    {
+      "epoch": 0.9242125984251969,
+      "grad_norm": 1.4913681745529175,
+      "learning_rate": 2.8245102104925257e-06,
+      "loss": 2.0501,
+      "step": 15024
+    },
+    {
+      "epoch": 0.9242741141732284,
+      "grad_norm": 1.3268245458602905,
+      "learning_rate": 2.8199484916250685e-06,
+      "loss": 2.1986,
+      "step": 15025
+    },
+    {
+      "epoch": 0.9243356299212598,
+      "grad_norm": 1.325339913368225,
+      "learning_rate": 2.8153904067496296e-06,
+      "loss": 2.8165,
+      "step": 15026
+    },
+    {
+      "epoch": 0.9243971456692913,
+      "grad_norm": 1.4203113317489624,
+      "learning_rate": 2.810835956036695e-06,
+      "loss": 2.436,
+      "step": 15027
+    },
+    {
+      "epoch": 0.9244586614173228,
+      "grad_norm": 1.0710705518722534,
+      "learning_rate": 2.806285139656539e-06,
+      "loss": 2.3282,
+      "step": 15028
+    },
+    {
+      "epoch": 0.9245201771653543,
+      "grad_norm": 1.6387141942977905,
+      "learning_rate": 2.801737957779349e-06,
+      "loss": 2.5081,
+      "step": 15029
+    },
+    {
+      "epoch": 0.9245816929133859,
+      "grad_norm": 1.2929949760437012,
+      "learning_rate": 2.797194410575166e-06,
+      "loss": 1.9003,
+      "step": 15030
+    },
+    {
+      "epoch": 0.9246432086614174,
+      "grad_norm": 1.350862979888916,
+      "learning_rate": 2.792654498213909e-06,
+      "loss": 2.6794,
+      "step": 15031
+    },
+    {
+      "epoch": 0.9247047244094488,
+      "grad_norm": 1.4683421850204468,
+      "learning_rate": 2.7881182208653213e-06,
+      "loss": 2.1288,
+      "step": 15032
+    },
+    {
+      "epoch": 0.9247662401574803,
+      "grad_norm": 1.440258502960205,
+      "learning_rate": 2.783585578699044e-06,
+      "loss": 2.3229,
+      "step": 15033
+    },
+    {
+      "epoch": 0.9248277559055118,
+      "grad_norm": 1.3295676708221436,
+      "learning_rate": 2.779056571884575e-06,
+      "loss": 2.7157,
+      "step": 15034
+    },
+    {
+      "epoch": 0.9248892716535433,
+      "grad_norm": 1.5362801551818848,
+      "learning_rate": 2.7745312005912683e-06,
+      "loss": 2.4155,
+      "step": 15035
+    },
+    {
+      "epoch": 0.9249507874015748,
+      "grad_norm": 1.4193776845932007,
+      "learning_rate": 2.7700094649883436e-06,
+      "loss": 2.9138,
+      "step": 15036
+    },
+    {
+      "epoch": 0.9250123031496063,
+      "grad_norm": 1.352905511856079,
+      "learning_rate": 2.7654913652449098e-06,
+      "loss": 2.2871,
+      "step": 15037
+    },
+    {
+      "epoch": 0.9250738188976378,
+      "grad_norm": 1.4431235790252686,
+      "learning_rate": 2.760976901529899e-06,
+      "loss": 2.3855,
+      "step": 15038
+    },
+    {
+      "epoch": 0.9251353346456693,
+      "grad_norm": 1.2315689325332642,
+      "learning_rate": 2.756466074012143e-06,
+      "loss": 2.5751,
+      "step": 15039
+    },
+    {
+      "epoch": 0.9251968503937008,
+      "grad_norm": 1.8122719526290894,
+      "learning_rate": 2.7519588828602947e-06,
+      "loss": 2.3272,
+      "step": 15040
+    },
+    {
+      "epoch": 0.9252583661417323,
+      "grad_norm": 1.5133764743804932,
+      "learning_rate": 2.7474553282429093e-06,
+      "loss": 2.7143,
+      "step": 15041
+    },
+    {
+      "epoch": 0.9253198818897638,
+      "grad_norm": 1.1567318439483643,
+      "learning_rate": 2.742955410328407e-06,
+      "loss": 1.7993,
+      "step": 15042
+    },
+    {
+      "epoch": 0.9253813976377953,
+      "grad_norm": 1.2947725057601929,
+      "learning_rate": 2.7384591292850647e-06,
+      "loss": 2.6192,
+      "step": 15043
+    },
+    {
+      "epoch": 0.9254429133858267,
+      "grad_norm": 1.30339777469635,
+      "learning_rate": 2.733966485281003e-06,
+      "loss": 2.5891,
+      "step": 15044
+    },
+    {
+      "epoch": 0.9255044291338582,
+      "grad_norm": 1.155674695968628,
+      "learning_rate": 2.7294774784842213e-06,
+      "loss": 2.6554,
+      "step": 15045
+    },
+    {
+      "epoch": 0.9255659448818898,
+      "grad_norm": 1.287615418434143,
+      "learning_rate": 2.7249921090625854e-06,
+      "loss": 1.867,
+      "step": 15046
+    },
+    {
+      "epoch": 0.9256274606299213,
+      "grad_norm": 1.091744303703308,
+      "learning_rate": 2.7205103771838048e-06,
+      "loss": 2.5324,
+      "step": 15047
+    },
+    {
+      "epoch": 0.9256889763779528,
+      "grad_norm": 1.3998385667800903,
+      "learning_rate": 2.716032283015524e-06,
+      "loss": 2.196,
+      "step": 15048
+    },
+    {
+      "epoch": 0.9257504921259843,
+      "grad_norm": 1.7612462043762207,
+      "learning_rate": 2.7115578267251416e-06,
+      "loss": 2.2463,
+      "step": 15049
+    },
+    {
+      "epoch": 0.9258120078740157,
+      "grad_norm": 1.4149997234344482,
+      "learning_rate": 2.7070870084800138e-06,
+      "loss": 2.2393,
+      "step": 15050
+    },
+    {
+      "epoch": 0.9258735236220472,
+      "grad_norm": 1.3980536460876465,
+      "learning_rate": 2.702619828447306e-06,
+      "loss": 1.9331,
+      "step": 15051
+    },
+    {
+      "epoch": 0.9259350393700787,
+      "grad_norm": 1.2768127918243408,
+      "learning_rate": 2.6981562867940734e-06,
+      "loss": 2.5353,
+      "step": 15052
+    },
+    {
+      "epoch": 0.9259965551181102,
+      "grad_norm": 1.2102843523025513,
+      "learning_rate": 2.6936963836872277e-06,
+      "loss": 2.5467,
+      "step": 15053
+    },
+    {
+      "epoch": 0.9260580708661418,
+      "grad_norm": 1.3853262662887573,
+      "learning_rate": 2.6892401192935346e-06,
+      "loss": 2.0451,
+      "step": 15054
+    },
+    {
+      "epoch": 0.9261195866141733,
+      "grad_norm": 1.2100189924240112,
+      "learning_rate": 2.684787493779639e-06,
+      "loss": 2.2699,
+      "step": 15055
+    },
+    {
+      "epoch": 0.9261811023622047,
+      "grad_norm": 1.239369511604309,
+      "learning_rate": 2.6803385073120635e-06,
+      "loss": 1.9995,
+      "step": 15056
+    },
+    {
+      "epoch": 0.9262426181102362,
+      "grad_norm": 1.425161361694336,
+      "learning_rate": 2.6758931600571525e-06,
+      "loss": 2.7088,
+      "step": 15057
+    },
+    {
+      "epoch": 0.9263041338582677,
+      "grad_norm": 1.3268083333969116,
+      "learning_rate": 2.6714514521811395e-06,
+      "loss": 1.9415,
+      "step": 15058
+    },
+    {
+      "epoch": 0.9263656496062992,
+      "grad_norm": 1.3725427389144897,
+      "learning_rate": 2.6670133838501255e-06,
+      "loss": 1.8143,
+      "step": 15059
+    },
+    {
+      "epoch": 0.9264271653543307,
+      "grad_norm": 1.3138651847839355,
+      "learning_rate": 2.662578955230055e-06,
+      "loss": 3.0489,
+      "step": 15060
+    },
+    {
+      "epoch": 0.9264886811023622,
+      "grad_norm": 1.2466282844543457,
+      "learning_rate": 2.6581481664867734e-06,
+      "loss": 1.899,
+      "step": 15061
+    },
+    {
+      "epoch": 0.9265501968503937,
+      "grad_norm": 1.389285922050476,
+      "learning_rate": 2.6537210177859593e-06,
+      "loss": 2.9405,
+      "step": 15062
+    },
+    {
+      "epoch": 0.9266117125984252,
+      "grad_norm": 1.241398811340332,
+      "learning_rate": 2.649297509293158e-06,
+      "loss": 1.9232,
+      "step": 15063
+    },
+    {
+      "epoch": 0.9266732283464567,
+      "grad_norm": 1.1887680292129517,
+      "learning_rate": 2.644877641173804e-06,
+      "loss": 2.3501,
+      "step": 15064
+    },
+    {
+      "epoch": 0.9267347440944882,
+      "grad_norm": 1.3712414503097534,
+      "learning_rate": 2.6404614135931315e-06,
+      "loss": 1.9852,
+      "step": 15065
+    },
+    {
+      "epoch": 0.9267962598425197,
+      "grad_norm": 1.3945090770721436,
+      "learning_rate": 2.63604882671632e-06,
+      "loss": 2.6026,
+      "step": 15066
+    },
+    {
+      "epoch": 0.9268577755905512,
+      "grad_norm": 1.326849341392517,
+      "learning_rate": 2.6316398807083696e-06,
+      "loss": 1.6433,
+      "step": 15067
+    },
+    {
+      "epoch": 0.9269192913385826,
+      "grad_norm": 1.5246357917785645,
+      "learning_rate": 2.62723457573415e-06,
+      "loss": 3.0158,
+      "step": 15068
+    },
+    {
+      "epoch": 0.9269808070866141,
+      "grad_norm": 1.038488507270813,
+      "learning_rate": 2.622832911958384e-06,
+      "loss": 3.4344,
+      "step": 15069
+    },
+    {
+      "epoch": 0.9270423228346457,
+      "grad_norm": 1.3156594038009644,
+      "learning_rate": 2.6184348895456955e-06,
+      "loss": 2.2507,
+      "step": 15070
+    },
+    {
+      "epoch": 0.9271038385826772,
+      "grad_norm": 1.6917285919189453,
+      "learning_rate": 2.614040508660509e-06,
+      "loss": 2.7148,
+      "step": 15071
+    },
+    {
+      "epoch": 0.9271653543307087,
+      "grad_norm": 1.2517491579055786,
+      "learning_rate": 2.609649769467182e-06,
+      "loss": 3.0524,
+      "step": 15072
+    },
+    {
+      "epoch": 0.9272268700787402,
+      "grad_norm": 2.031463861465454,
+      "learning_rate": 2.6052626721298603e-06,
+      "loss": 3.059,
+      "step": 15073
+    },
+    {
+      "epoch": 0.9272883858267716,
+      "grad_norm": 1.763569712638855,
+      "learning_rate": 2.6008792168126573e-06,
+      "loss": 0.8877,
+      "step": 15074
+    },
+    {
+      "epoch": 0.9273499015748031,
+      "grad_norm": 1.169087290763855,
+      "learning_rate": 2.596499403679453e-06,
+      "loss": 2.3082,
+      "step": 15075
+    },
+    {
+      "epoch": 0.9274114173228346,
+      "grad_norm": 1.1180739402770996,
+      "learning_rate": 2.592123232894028e-06,
+      "loss": 2.7037,
+      "step": 15076
+    },
+    {
+      "epoch": 0.9274729330708661,
+      "grad_norm": 1.2888476848602295,
+      "learning_rate": 2.58775070462004e-06,
+      "loss": 2.4258,
+      "step": 15077
+    },
+    {
+      "epoch": 0.9275344488188977,
+      "grad_norm": 1.3701986074447632,
+      "learning_rate": 2.5833818190209915e-06,
+      "loss": 2.7227,
+      "step": 15078
+    },
+    {
+      "epoch": 0.9275959645669292,
+      "grad_norm": 1.3849083185195923,
+      "learning_rate": 2.5790165762602405e-06,
+      "loss": 2.2985,
+      "step": 15079
+    },
+    {
+      "epoch": 0.9276574803149606,
+      "grad_norm": 1.1566966772079468,
+      "learning_rate": 2.5746549765010342e-06,
+      "loss": 2.716,
+      "step": 15080
+    },
+    {
+      "epoch": 0.9277189960629921,
+      "grad_norm": 1.197548508644104,
+      "learning_rate": 2.570297019906498e-06,
+      "loss": 2.5247,
+      "step": 15081
+    },
+    {
+      "epoch": 0.9277805118110236,
+      "grad_norm": 1.2896808385849,
+      "learning_rate": 2.565942706639546e-06,
+      "loss": 2.3978,
+      "step": 15082
+    },
+    {
+      "epoch": 0.9278420275590551,
+      "grad_norm": 1.3294532299041748,
+      "learning_rate": 2.5615920368630364e-06,
+      "loss": 2.2851,
+      "step": 15083
+    },
+    {
+      "epoch": 0.9279035433070866,
+      "grad_norm": 1.195075273513794,
+      "learning_rate": 2.5572450107396506e-06,
+      "loss": 2.2472,
+      "step": 15084
+    },
+    {
+      "epoch": 0.9279650590551181,
+      "grad_norm": 1.2546125650405884,
+      "learning_rate": 2.5529016284319475e-06,
+      "loss": 2.1051,
+      "step": 15085
+    },
+    {
+      "epoch": 0.9280265748031497,
+      "grad_norm": 1.2792394161224365,
+      "learning_rate": 2.548561890102341e-06,
+      "loss": 3.0691,
+      "step": 15086
+    },
+    {
+      "epoch": 0.9280880905511811,
+      "grad_norm": 1.3017951250076294,
+      "learning_rate": 2.544225795913124e-06,
+      "loss": 2.2088,
+      "step": 15087
+    },
+    {
+      "epoch": 0.9281496062992126,
+      "grad_norm": 1.2333441972732544,
+      "learning_rate": 2.539893346026434e-06,
+      "loss": 2.093,
+      "step": 15088
+    },
+    {
+      "epoch": 0.9282111220472441,
+      "grad_norm": 1.3465170860290527,
+      "learning_rate": 2.5355645406042848e-06,
+      "loss": 2.862,
+      "step": 15089
+    },
+    {
+      "epoch": 0.9282726377952756,
+      "grad_norm": 1.396145224571228,
+      "learning_rate": 2.5312393798085254e-06,
+      "loss": 2.3715,
+      "step": 15090
+    },
+    {
+      "epoch": 0.9283341535433071,
+      "grad_norm": 1.5074092149734497,
+      "learning_rate": 2.526917863800926e-06,
+      "loss": 2.4121,
+      "step": 15091
+    },
+    {
+      "epoch": 0.9283956692913385,
+      "grad_norm": 1.3675276041030884,
+      "learning_rate": 2.5225999927430687e-06,
+      "loss": 2.5652,
+      "step": 15092
+    },
+    {
+      "epoch": 0.92845718503937,
+      "grad_norm": 1.2964110374450684,
+      "learning_rate": 2.5182857667964243e-06,
+      "loss": 1.9538,
+      "step": 15093
+    },
+    {
+      "epoch": 0.9285187007874016,
+      "grad_norm": 1.2249747514724731,
+      "learning_rate": 2.513975186122319e-06,
+      "loss": 2.2173,
+      "step": 15094
+    },
+    {
+      "epoch": 0.9285802165354331,
+      "grad_norm": 1.0796375274658203,
+      "learning_rate": 2.509668250881947e-06,
+      "loss": 2.9261,
+      "step": 15095
+    },
+    {
+      "epoch": 0.9286417322834646,
+      "grad_norm": 1.5826860666275024,
+      "learning_rate": 2.505364961236367e-06,
+      "loss": 2.5443,
+      "step": 15096
+    },
+    {
+      "epoch": 0.9287032480314961,
+      "grad_norm": 1.124995231628418,
+      "learning_rate": 2.5010653173464847e-06,
+      "loss": 2.7916,
+      "step": 15097
+    },
+    {
+      "epoch": 0.9287647637795275,
+      "grad_norm": 1.412154197692871,
+      "learning_rate": 2.4967693193730936e-06,
+      "loss": 2.6496,
+      "step": 15098
+    },
+    {
+      "epoch": 0.928826279527559,
+      "grad_norm": 1.3778599500656128,
+      "learning_rate": 2.492476967476831e-06,
+      "loss": 2.2037,
+      "step": 15099
+    },
+    {
+      "epoch": 0.9288877952755905,
+      "grad_norm": 1.3193507194519043,
+      "learning_rate": 2.488188261818214e-06,
+      "loss": 2.2125,
+      "step": 15100
+    },
+    {
+      "epoch": 0.928949311023622,
+      "grad_norm": 1.2084764242172241,
+      "learning_rate": 2.4839032025576247e-06,
+      "loss": 2.2645,
+      "step": 15101
+    },
+    {
+      "epoch": 0.9290108267716536,
+      "grad_norm": 1.1589490175247192,
+      "learning_rate": 2.4796217898552798e-06,
+      "loss": 2.8265,
+      "step": 15102
+    },
+    {
+      "epoch": 0.9290723425196851,
+      "grad_norm": 1.4786821603775024,
+      "learning_rate": 2.4753440238713068e-06,
+      "loss": 1.9877,
+      "step": 15103
+    },
+    {
+      "epoch": 0.9291338582677166,
+      "grad_norm": 1.229475975036621,
+      "learning_rate": 2.4710699047656327e-06,
+      "loss": 2.6711,
+      "step": 15104
+    },
+    {
+      "epoch": 0.929195374015748,
+      "grad_norm": 1.2645939588546753,
+      "learning_rate": 2.4667994326981194e-06,
+      "loss": 2.1389,
+      "step": 15105
+    },
+    {
+      "epoch": 0.9292568897637795,
+      "grad_norm": 1.2738370895385742,
+      "learning_rate": 2.4625326078284384e-06,
+      "loss": 2.3296,
+      "step": 15106
+    },
+    {
+      "epoch": 0.929318405511811,
+      "grad_norm": 1.6479301452636719,
+      "learning_rate": 2.4582694303161734e-06,
+      "loss": 2.3468,
+      "step": 15107
+    },
+    {
+      "epoch": 0.9293799212598425,
+      "grad_norm": 1.3750627040863037,
+      "learning_rate": 2.4540099003207083e-06,
+      "loss": 2.5497,
+      "step": 15108
+    },
+    {
+      "epoch": 0.929441437007874,
+      "grad_norm": 1.226776123046875,
+      "learning_rate": 2.44975401800136e-06,
+      "loss": 2.1629,
+      "step": 15109
+    },
+    {
+      "epoch": 0.9295029527559056,
+      "grad_norm": 1.377002477645874,
+      "learning_rate": 2.4455017835172347e-06,
+      "loss": 1.6773,
+      "step": 15110
+    },
+    {
+      "epoch": 0.929564468503937,
+      "grad_norm": 1.2629683017730713,
+      "learning_rate": 2.441253197027371e-06,
+      "loss": 2.1938,
+      "step": 15111
+    },
+    {
+      "epoch": 0.9296259842519685,
+      "grad_norm": 1.3945860862731934,
+      "learning_rate": 2.437008258690632e-06,
+      "loss": 2.2822,
+      "step": 15112
+    },
+    {
+      "epoch": 0.9296875,
+      "grad_norm": 1.442958950996399,
+      "learning_rate": 2.4327669686657673e-06,
+      "loss": 2.7845,
+      "step": 15113
+    },
+    {
+      "epoch": 0.9297490157480315,
+      "grad_norm": 1.3601179122924805,
+      "learning_rate": 2.428529327111362e-06,
+      "loss": 2.651,
+      "step": 15114
+    },
+    {
+      "epoch": 0.929810531496063,
+      "grad_norm": 1.3101822137832642,
+      "learning_rate": 2.4242953341858996e-06,
+      "loss": 1.8765,
+      "step": 15115
+    },
+    {
+      "epoch": 0.9298720472440944,
+      "grad_norm": 1.7629625797271729,
+      "learning_rate": 2.4200649900476767e-06,
+      "loss": 2.245,
+      "step": 15116
+    },
+    {
+      "epoch": 0.929933562992126,
+      "grad_norm": 1.6794872283935547,
+      "learning_rate": 2.415838294854911e-06,
+      "loss": 2.6233,
+      "step": 15117
+    },
+    {
+      "epoch": 0.9299950787401575,
+      "grad_norm": 1.1789405345916748,
+      "learning_rate": 2.4116152487656417e-06,
+      "loss": 2.6042,
+      "step": 15118
+    },
+    {
+      "epoch": 0.930056594488189,
+      "grad_norm": 1.4006072282791138,
+      "learning_rate": 2.407395851937799e-06,
+      "loss": 2.0979,
+      "step": 15119
+    },
+    {
+      "epoch": 0.9301181102362205,
+      "grad_norm": 1.2016994953155518,
+      "learning_rate": 2.403180104529168e-06,
+      "loss": 2.3288,
+      "step": 15120
+    },
+    {
+      "epoch": 0.930179625984252,
+      "grad_norm": 1.2973265647888184,
+      "learning_rate": 2.3989680066973775e-06,
+      "loss": 2.3736,
+      "step": 15121
+    },
+    {
+      "epoch": 0.9302411417322834,
+      "grad_norm": 1.2329484224319458,
+      "learning_rate": 2.394759558599946e-06,
+      "loss": 2.5351,
+      "step": 15122
+    },
+    {
+      "epoch": 0.9303026574803149,
+      "grad_norm": 1.2375224828720093,
+      "learning_rate": 2.390554760394248e-06,
+      "loss": 2.8606,
+      "step": 15123
+    },
+    {
+      "epoch": 0.9303641732283464,
+      "grad_norm": 1.3490478992462158,
+      "learning_rate": 2.386353612237513e-06,
+      "loss": 2.2966,
+      "step": 15124
+    },
+    {
+      "epoch": 0.930425688976378,
+      "grad_norm": 1.3881314992904663,
+      "learning_rate": 2.3821561142868376e-06,
+      "loss": 2.0356,
+      "step": 15125
+    },
+    {
+      "epoch": 0.9304872047244095,
+      "grad_norm": 1.3954355716705322,
+      "learning_rate": 2.3779622666992075e-06,
+      "loss": 2.4883,
+      "step": 15126
+    },
+    {
+      "epoch": 0.930548720472441,
+      "grad_norm": 1.2523610591888428,
+      "learning_rate": 2.3737720696314193e-06,
+      "loss": 2.5589,
+      "step": 15127
+    },
+    {
+      "epoch": 0.9306102362204725,
+      "grad_norm": 1.2566587924957275,
+      "learning_rate": 2.369585523240192e-06,
+      "loss": 2.5216,
+      "step": 15128
+    },
+    {
+      "epoch": 0.9306717519685039,
+      "grad_norm": 1.1380382776260376,
+      "learning_rate": 2.365402627682056e-06,
+      "loss": 2.3974,
+      "step": 15129
+    },
+    {
+      "epoch": 0.9307332677165354,
+      "grad_norm": 1.2759718894958496,
+      "learning_rate": 2.361223383113431e-06,
+      "loss": 1.7406,
+      "step": 15130
+    },
+    {
+      "epoch": 0.9307947834645669,
+      "grad_norm": 1.4399021863937378,
+      "learning_rate": 2.357047789690603e-06,
+      "loss": 2.6188,
+      "step": 15131
+    },
+    {
+      "epoch": 0.9308562992125984,
+      "grad_norm": 1.4500389099121094,
+      "learning_rate": 2.3528758475697242e-06,
+      "loss": 2.1165,
+      "step": 15132
+    },
+    {
+      "epoch": 0.93091781496063,
+      "grad_norm": 1.459618330001831,
+      "learning_rate": 2.3487075569067817e-06,
+      "loss": 2.211,
+      "step": 15133
+    },
+    {
+      "epoch": 0.9309793307086615,
+      "grad_norm": 1.3123799562454224,
+      "learning_rate": 2.344542917857673e-06,
+      "loss": 2.4713,
+      "step": 15134
+    },
+    {
+      "epoch": 0.9310408464566929,
+      "grad_norm": 1.4291914701461792,
+      "learning_rate": 2.340381930578106e-06,
+      "loss": 2.6538,
+      "step": 15135
+    },
+    {
+      "epoch": 0.9311023622047244,
+      "grad_norm": 1.296897053718567,
+      "learning_rate": 2.3362245952236905e-06,
+      "loss": 2.9399,
+      "step": 15136
+    },
+    {
+      "epoch": 0.9311638779527559,
+      "grad_norm": 1.1761277914047241,
+      "learning_rate": 2.332070911949891e-06,
+      "loss": 2.6533,
+      "step": 15137
+    },
+    {
+      "epoch": 0.9312253937007874,
+      "grad_norm": 1.1693366765975952,
+      "learning_rate": 2.3279208809120378e-06,
+      "loss": 1.8292,
+      "step": 15138
+    },
+    {
+      "epoch": 0.9312869094488189,
+      "grad_norm": 1.316658854484558,
+      "learning_rate": 2.3237745022652858e-06,
+      "loss": 2.4471,
+      "step": 15139
+    },
+    {
+      "epoch": 0.9313484251968503,
+      "grad_norm": 1.3381258249282837,
+      "learning_rate": 2.3196317761647325e-06,
+      "loss": 1.737,
+      "step": 15140
+    },
+    {
+      "epoch": 0.9314099409448819,
+      "grad_norm": 1.1654354333877563,
+      "learning_rate": 2.315492702765254e-06,
+      "loss": 2.4431,
+      "step": 15141
+    },
+    {
+      "epoch": 0.9314714566929134,
+      "grad_norm": 1.1188464164733887,
+      "learning_rate": 2.31135728222166e-06,
+      "loss": 2.458,
+      "step": 15142
+    },
+    {
+      "epoch": 0.9315329724409449,
+      "grad_norm": 1.2135519981384277,
+      "learning_rate": 2.30722551468856e-06,
+      "loss": 2.5417,
+      "step": 15143
+    },
+    {
+      "epoch": 0.9315944881889764,
+      "grad_norm": 1.3898452520370483,
+      "learning_rate": 2.3030974003204863e-06,
+      "loss": 2.767,
+      "step": 15144
+    },
+    {
+      "epoch": 0.9316560039370079,
+      "grad_norm": 1.2748713493347168,
+      "learning_rate": 2.298972939271793e-06,
+      "loss": 2.4818,
+      "step": 15145
+    },
+    {
+      "epoch": 0.9317175196850394,
+      "grad_norm": 1.2781662940979004,
+      "learning_rate": 2.294852131696723e-06,
+      "loss": 2.4678,
+      "step": 15146
+    },
+    {
+      "epoch": 0.9317790354330708,
+      "grad_norm": 1.1862280368804932,
+      "learning_rate": 2.2907349777493536e-06,
+      "loss": 2.8653,
+      "step": 15147
+    },
+    {
+      "epoch": 0.9318405511811023,
+      "grad_norm": 1.1781927347183228,
+      "learning_rate": 2.2866214775836613e-06,
+      "loss": 2.9618,
+      "step": 15148
+    },
+    {
+      "epoch": 0.9319020669291339,
+      "grad_norm": 1.3334780931472778,
+      "learning_rate": 2.282511631353457e-06,
+      "loss": 1.4339,
+      "step": 15149
+    },
+    {
+      "epoch": 0.9319635826771654,
+      "grad_norm": 1.2404488325119019,
+      "learning_rate": 2.278405439212428e-06,
+      "loss": 2.566,
+      "step": 15150
+    },
+    {
+      "epoch": 0.9320250984251969,
+      "grad_norm": 1.135316014289856,
+      "learning_rate": 2.274302901314129e-06,
+      "loss": 2.1524,
+      "step": 15151
+    },
+    {
+      "epoch": 0.9320866141732284,
+      "grad_norm": 1.356313705444336,
+      "learning_rate": 2.270204017811972e-06,
+      "loss": 2.0247,
+      "step": 15152
+    },
+    {
+      "epoch": 0.9321481299212598,
+      "grad_norm": 1.2034223079681396,
+      "learning_rate": 2.2661087888592225e-06,
+      "loss": 2.2368,
+      "step": 15153
+    },
+    {
+      "epoch": 0.9322096456692913,
+      "grad_norm": 1.3805840015411377,
+      "learning_rate": 2.262017214609036e-06,
+      "loss": 2.5914,
+      "step": 15154
+    },
+    {
+      "epoch": 0.9322711614173228,
+      "grad_norm": 1.3851525783538818,
+      "learning_rate": 2.257929295214389e-06,
+      "loss": 2.1349,
+      "step": 15155
+    },
+    {
+      "epoch": 0.9323326771653543,
+      "grad_norm": 1.2181543111801147,
+      "learning_rate": 2.253845030828161e-06,
+      "loss": 2.3202,
+      "step": 15156
+    },
+    {
+      "epoch": 0.9323941929133859,
+      "grad_norm": 1.1068302392959595,
+      "learning_rate": 2.2497644216030956e-06,
+      "loss": 1.8328,
+      "step": 15157
+    },
+    {
+      "epoch": 0.9324557086614174,
+      "grad_norm": 1.2718347311019897,
+      "learning_rate": 2.2456874676917594e-06,
+      "loss": 2.8384,
+      "step": 15158
+    },
+    {
+      "epoch": 0.9325172244094488,
+      "grad_norm": 1.2891833782196045,
+      "learning_rate": 2.241614169246631e-06,
+      "loss": 2.1838,
+      "step": 15159
+    },
+    {
+      "epoch": 0.9325787401574803,
+      "grad_norm": 1.4161888360977173,
+      "learning_rate": 2.237544526419999e-06,
+      "loss": 2.6524,
+      "step": 15160
+    },
+    {
+      "epoch": 0.9326402559055118,
+      "grad_norm": 1.2744642496109009,
+      "learning_rate": 2.2334785393640866e-06,
+      "loss": 2.1569,
+      "step": 15161
+    },
+    {
+      "epoch": 0.9327017716535433,
+      "grad_norm": 1.3285348415374756,
+      "learning_rate": 2.2294162082308835e-06,
+      "loss": 2.6057,
+      "step": 15162
+    },
+    {
+      "epoch": 0.9327632874015748,
+      "grad_norm": 1.4146273136138916,
+      "learning_rate": 2.2253575331723563e-06,
+      "loss": 1.7246,
+      "step": 15163
+    },
+    {
+      "epoch": 0.9328248031496063,
+      "grad_norm": 1.2681734561920166,
+      "learning_rate": 2.22130251434024e-06,
+      "loss": 2.1463,
+      "step": 15164
+    },
+    {
+      "epoch": 0.9328863188976378,
+      "grad_norm": 1.5601460933685303,
+      "learning_rate": 2.2172511518861903e-06,
+      "loss": 2.0451,
+      "step": 15165
+    },
+    {
+      "epoch": 0.9329478346456693,
+      "grad_norm": 1.3360729217529297,
+      "learning_rate": 2.213203445961687e-06,
+      "loss": 2.6818,
+      "step": 15166
+    },
+    {
+      "epoch": 0.9330093503937008,
+      "grad_norm": 1.2029186487197876,
+      "learning_rate": 2.209159396718108e-06,
+      "loss": 1.9988,
+      "step": 15167
+    },
+    {
+      "epoch": 0.9330708661417323,
+      "grad_norm": 1.2018578052520752,
+      "learning_rate": 2.205119004306666e-06,
+      "loss": 2.4243,
+      "step": 15168
+    },
+    {
+      "epoch": 0.9331323818897638,
+      "grad_norm": 1.3787946701049805,
+      "learning_rate": 2.2010822688784516e-06,
+      "loss": 2.5732,
+      "step": 15169
+    },
+    {
+      "epoch": 0.9331938976377953,
+      "grad_norm": 1.06744384765625,
+      "learning_rate": 2.1970491905844216e-06,
+      "loss": 2.9089,
+      "step": 15170
+    },
+    {
+      "epoch": 0.9332554133858267,
+      "grad_norm": 1.4380135536193848,
+      "learning_rate": 2.1930197695753883e-06,
+      "loss": 2.3238,
+      "step": 15171
+    },
+    {
+      "epoch": 0.9333169291338582,
+      "grad_norm": 1.3796650171279907,
+      "learning_rate": 2.188994006002032e-06,
+      "loss": 2.108,
+      "step": 15172
+    },
+    {
+      "epoch": 0.9333784448818898,
+      "grad_norm": 1.2532763481140137,
+      "learning_rate": 2.184971900014887e-06,
+      "loss": 2.8044,
+      "step": 15173
+    },
+    {
+      "epoch": 0.9334399606299213,
+      "grad_norm": 1.239436388015747,
+      "learning_rate": 2.180953451764356e-06,
+      "loss": 2.5722,
+      "step": 15174
+    },
+    {
+      "epoch": 0.9335014763779528,
+      "grad_norm": 3.1487934589385986,
+      "learning_rate": 2.176938661400718e-06,
+      "loss": 2.1355,
+      "step": 15175
+    },
+    {
+      "epoch": 0.9335629921259843,
+      "grad_norm": 1.85431969165802,
+      "learning_rate": 2.1729275290740982e-06,
+      "loss": 2.1127,
+      "step": 15176
+    },
+    {
+      "epoch": 0.9336245078740157,
+      "grad_norm": 1.1325210332870483,
+      "learning_rate": 2.1689200549344868e-06,
+      "loss": 2.9749,
+      "step": 15177
+    },
+    {
+      "epoch": 0.9336860236220472,
+      "grad_norm": 1.1475552320480347,
+      "learning_rate": 2.1649162391317424e-06,
+      "loss": 1.8635,
+      "step": 15178
+    },
+    {
+      "epoch": 0.9337475393700787,
+      "grad_norm": 1.3188238143920898,
+      "learning_rate": 2.1609160818156005e-06,
+      "loss": 2.8948,
+      "step": 15179
+    },
+    {
+      "epoch": 0.9338090551181102,
+      "grad_norm": 1.2616816759109497,
+      "learning_rate": 2.1569195831356193e-06,
+      "loss": 1.9926,
+      "step": 15180
+    },
+    {
+      "epoch": 0.9338705708661418,
+      "grad_norm": 1.274449110031128,
+      "learning_rate": 2.152926743241257e-06,
+      "loss": 1.6851,
+      "step": 15181
+    },
+    {
+      "epoch": 0.9339320866141733,
+      "grad_norm": 1.311814308166504,
+      "learning_rate": 2.1489375622818273e-06,
+      "loss": 2.0144,
+      "step": 15182
+    },
+    {
+      "epoch": 0.9339936023622047,
+      "grad_norm": 1.2078078985214233,
+      "learning_rate": 2.1449520404065003e-06,
+      "loss": 2.8224,
+      "step": 15183
+    },
+    {
+      "epoch": 0.9340551181102362,
+      "grad_norm": 1.2363253831863403,
+      "learning_rate": 2.140970177764312e-06,
+      "loss": 2.0596,
+      "step": 15184
+    },
+    {
+      "epoch": 0.9341166338582677,
+      "grad_norm": 1.2213716506958008,
+      "learning_rate": 2.1369919745041654e-06,
+      "loss": 2.4592,
+      "step": 15185
+    },
+    {
+      "epoch": 0.9341781496062992,
+      "grad_norm": 1.228011965751648,
+      "learning_rate": 2.1330174307748196e-06,
+      "loss": 2.2228,
+      "step": 15186
+    },
+    {
+      "epoch": 0.9342396653543307,
+      "grad_norm": 1.2556381225585938,
+      "learning_rate": 2.1290465467249e-06,
+      "loss": 2.2745,
+      "step": 15187
+    },
+    {
+      "epoch": 0.9343011811023622,
+      "grad_norm": 1.221698522567749,
+      "learning_rate": 2.1250793225028876e-06,
+      "loss": 2.4035,
+      "step": 15188
+    },
+    {
+      "epoch": 0.9343626968503937,
+      "grad_norm": 1.3926589488983154,
+      "learning_rate": 2.121115758257153e-06,
+      "loss": 2.2845,
+      "step": 15189
+    },
+    {
+      "epoch": 0.9344242125984252,
+      "grad_norm": 1.2245324850082397,
+      "learning_rate": 2.1171558541358992e-06,
+      "loss": 2.6893,
+      "step": 15190
+    },
+    {
+      "epoch": 0.9344857283464567,
+      "grad_norm": 1.1169379949569702,
+      "learning_rate": 2.1131996102872086e-06,
+      "loss": 2.4553,
+      "step": 15191
+    },
+    {
+      "epoch": 0.9345472440944882,
+      "grad_norm": 1.4057230949401855,
+      "learning_rate": 2.1092470268590292e-06,
+      "loss": 1.9927,
+      "step": 15192
+    },
+    {
+      "epoch": 0.9346087598425197,
+      "grad_norm": 1.258409023284912,
+      "learning_rate": 2.1052981039991427e-06,
+      "loss": 2.3062,
+      "step": 15193
+    },
+    {
+      "epoch": 0.9346702755905512,
+      "grad_norm": 1.1806799173355103,
+      "learning_rate": 2.1013528418552308e-06,
+      "loss": 2.4702,
+      "step": 15194
+    },
+    {
+      "epoch": 0.9347317913385826,
+      "grad_norm": 1.2019002437591553,
+      "learning_rate": 2.0974112405748204e-06,
+      "loss": 2.6909,
+      "step": 15195
+    },
+    {
+      "epoch": 0.9347933070866141,
+      "grad_norm": 1.2757489681243896,
+      "learning_rate": 2.0934733003053262e-06,
+      "loss": 1.6435,
+      "step": 15196
+    },
+    {
+      "epoch": 0.9348548228346457,
+      "grad_norm": 1.1609543561935425,
+      "learning_rate": 2.0895390211939757e-06,
+      "loss": 2.1463,
+      "step": 15197
+    },
+    {
+      "epoch": 0.9349163385826772,
+      "grad_norm": 1.4767603874206543,
+      "learning_rate": 2.0856084033879175e-06,
+      "loss": 2.7262,
+      "step": 15198
+    },
+    {
+      "epoch": 0.9349778543307087,
+      "grad_norm": 1.3187074661254883,
+      "learning_rate": 2.0816814470341008e-06,
+      "loss": 2.2773,
+      "step": 15199
+    },
+    {
+      "epoch": 0.9350393700787402,
+      "grad_norm": 1.2752724885940552,
+      "learning_rate": 2.0777581522793964e-06,
+      "loss": 2.8298,
+      "step": 15200
+    },
+    {
+      "epoch": 0.9351008858267716,
+      "grad_norm": 1.2746418714523315,
+      "learning_rate": 2.07383851927051e-06,
+      "loss": 2.5514,
+      "step": 15201
+    },
+    {
+      "epoch": 0.9351624015748031,
+      "grad_norm": 1.2463234663009644,
+      "learning_rate": 2.069922548154013e-06,
+      "loss": 2.2114,
+      "step": 15202
+    },
+    {
+      "epoch": 0.9352239173228346,
+      "grad_norm": 1.1816285848617554,
+      "learning_rate": 2.0660102390763437e-06,
+      "loss": 2.583,
+      "step": 15203
+    },
+    {
+      "epoch": 0.9352854330708661,
+      "grad_norm": 1.461143136024475,
+      "learning_rate": 2.062101592183796e-06,
+      "loss": 2.0877,
+      "step": 15204
+    },
+    {
+      "epoch": 0.9353469488188977,
+      "grad_norm": 1.1961520910263062,
+      "learning_rate": 2.0581966076225202e-06,
+      "loss": 2.5389,
+      "step": 15205
+    },
+    {
+      "epoch": 0.9354084645669292,
+      "grad_norm": 1.3907393217086792,
+      "learning_rate": 2.0542952855385543e-06,
+      "loss": 2.813,
+      "step": 15206
+    },
+    {
+      "epoch": 0.9354699803149606,
+      "grad_norm": 1.2961784601211548,
+      "learning_rate": 2.0503976260777936e-06,
+      "loss": 2.6177,
+      "step": 15207
+    },
+    {
+      "epoch": 0.9355314960629921,
+      "grad_norm": 1.3918938636779785,
+      "learning_rate": 2.046503629385976e-06,
+      "loss": 2.5105,
+      "step": 15208
+    },
+    {
+      "epoch": 0.9355930118110236,
+      "grad_norm": 1.8295176029205322,
+      "learning_rate": 2.0426132956087196e-06,
+      "loss": 2.2984,
+      "step": 15209
+    },
+    {
+      "epoch": 0.9356545275590551,
+      "grad_norm": 1.212807059288025,
+      "learning_rate": 2.038726624891507e-06,
+      "loss": 1.6401,
+      "step": 15210
+    },
+    {
+      "epoch": 0.9357160433070866,
+      "grad_norm": 1.5004680156707764,
+      "learning_rate": 2.034843617379656e-06,
+      "loss": 1.7951,
+      "step": 15211
+    },
+    {
+      "epoch": 0.9357775590551181,
+      "grad_norm": 1.3549164533615112,
+      "learning_rate": 2.0309642732184053e-06,
+      "loss": 2.4403,
+      "step": 15212
+    },
+    {
+      "epoch": 0.9358390748031497,
+      "grad_norm": 1.2684719562530518,
+      "learning_rate": 2.027088592552784e-06,
+      "loss": 2.324,
+      "step": 15213
+    },
+    {
+      "epoch": 0.9359005905511811,
+      "grad_norm": 1.2798713445663452,
+      "learning_rate": 2.0232165755277422e-06,
+      "loss": 2.4616,
+      "step": 15214
+    },
+    {
+      "epoch": 0.9359621062992126,
+      "grad_norm": 2.3612747192382812,
+      "learning_rate": 2.0193482222880646e-06,
+      "loss": 2.6804,
+      "step": 15215
+    },
+    {
+      "epoch": 0.9360236220472441,
+      "grad_norm": 1.8359897136688232,
+      "learning_rate": 2.015483532978413e-06,
+      "loss": 2.0388,
+      "step": 15216
+    },
+    {
+      "epoch": 0.9360851377952756,
+      "grad_norm": 1.1741111278533936,
+      "learning_rate": 2.011622507743294e-06,
+      "loss": 2.2494,
+      "step": 15217
+    },
+    {
+      "epoch": 0.9361466535433071,
+      "grad_norm": 1.2592540979385376,
+      "learning_rate": 2.007765146727103e-06,
+      "loss": 2.9751,
+      "step": 15218
+    },
+    {
+      "epoch": 0.9362081692913385,
+      "grad_norm": 1.094698429107666,
+      "learning_rate": 2.0039114500740695e-06,
+      "loss": 2.8834,
+      "step": 15219
+    },
+    {
+      "epoch": 0.93626968503937,
+      "grad_norm": 1.504066824913025,
+      "learning_rate": 2.0000614179283115e-06,
+      "loss": 2.2876,
+      "step": 15220
+    },
+    {
+      "epoch": 0.9363312007874016,
+      "grad_norm": 1.1998733282089233,
+      "learning_rate": 1.9962150504337806e-06,
+      "loss": 2.8516,
+      "step": 15221
+    },
+    {
+      "epoch": 0.9363927165354331,
+      "grad_norm": 1.5641897916793823,
+      "learning_rate": 1.9923723477343504e-06,
+      "loss": 2.2033,
+      "step": 15222
+    },
+    {
+      "epoch": 0.9364542322834646,
+      "grad_norm": 1.617488980293274,
+      "learning_rate": 1.9885333099736615e-06,
+      "loss": 2.8204,
+      "step": 15223
+    },
+    {
+      "epoch": 0.9365157480314961,
+      "grad_norm": 1.195371389389038,
+      "learning_rate": 1.9846979372953213e-06,
+      "loss": 3.0388,
+      "step": 15224
+    },
+    {
+      "epoch": 0.9365772637795275,
+      "grad_norm": 1.4491513967514038,
+      "learning_rate": 1.980866229842715e-06,
+      "loss": 2.4526,
+      "step": 15225
+    },
+    {
+      "epoch": 0.936638779527559,
+      "grad_norm": 1.106795072555542,
+      "learning_rate": 1.9770381877591393e-06,
+      "loss": 2.4808,
+      "step": 15226
+    },
+    {
+      "epoch": 0.9367002952755905,
+      "grad_norm": 1.2925002574920654,
+      "learning_rate": 1.9732138111877574e-06,
+      "loss": 2.3403,
+      "step": 15227
+    },
+    {
+      "epoch": 0.936761811023622,
+      "grad_norm": 1.4945420026779175,
+      "learning_rate": 1.969393100271566e-06,
+      "loss": 1.9861,
+      "step": 15228
+    },
+    {
+      "epoch": 0.9368233267716536,
+      "grad_norm": 1.1004513502120972,
+      "learning_rate": 1.9655760551534285e-06,
+      "loss": 2.8359,
+      "step": 15229
+    },
+    {
+      "epoch": 0.9368848425196851,
+      "grad_norm": 1.2998160123825073,
+      "learning_rate": 1.9617626759760977e-06,
+      "loss": 2.5337,
+      "step": 15230
+    },
+    {
+      "epoch": 0.9369463582677166,
+      "grad_norm": 1.3569294214248657,
+      "learning_rate": 1.9579529628821704e-06,
+      "loss": 2.4915,
+      "step": 15231
+    },
+    {
+      "epoch": 0.937007874015748,
+      "grad_norm": 1.3343439102172852,
+      "learning_rate": 1.954146916014088e-06,
+      "loss": 2.6982,
+      "step": 15232
+    },
+    {
+      "epoch": 0.9370693897637795,
+      "grad_norm": 1.376447319984436,
+      "learning_rate": 1.9503445355142147e-06,
+      "loss": 2.7172,
+      "step": 15233
+    },
+    {
+      "epoch": 0.937130905511811,
+      "grad_norm": 1.1458114385604858,
+      "learning_rate": 1.9465458215247035e-06,
+      "loss": 2.635,
+      "step": 15234
+    },
+    {
+      "epoch": 0.9371924212598425,
+      "grad_norm": 1.317236065864563,
+      "learning_rate": 1.9427507741876184e-06,
+      "loss": 2.816,
+      "step": 15235
+    },
+    {
+      "epoch": 0.937253937007874,
+      "grad_norm": 1.3905723094940186,
+      "learning_rate": 1.938959393644879e-06,
+      "loss": 2.1789,
+      "step": 15236
+    },
+    {
+      "epoch": 0.9373154527559056,
+      "grad_norm": 1.2227129936218262,
+      "learning_rate": 1.93517168003825e-06,
+      "loss": 2.6391,
+      "step": 15237
+    },
+    {
+      "epoch": 0.937376968503937,
+      "grad_norm": 1.3057440519332886,
+      "learning_rate": 1.931387633509374e-06,
+      "loss": 2.1938,
+      "step": 15238
+    },
+    {
+      "epoch": 0.9374384842519685,
+      "grad_norm": 1.2935945987701416,
+      "learning_rate": 1.927607254199748e-06,
+      "loss": 2.4959,
+      "step": 15239
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 1.3087893724441528,
+      "learning_rate": 1.9238305422507486e-06,
+      "loss": 2.7248,
+      "step": 15240
+    },
+    {
+      "epoch": 0.9375615157480315,
+      "grad_norm": 1.1436400413513184,
+      "learning_rate": 1.9200574978036068e-06,
+      "loss": 2.6151,
+      "step": 15241
+    },
+    {
+      "epoch": 0.937623031496063,
+      "grad_norm": 1.2677878141403198,
+      "learning_rate": 1.916288120999399e-06,
+      "loss": 1.9035,
+      "step": 15242
+    },
+    {
+      "epoch": 0.9376845472440944,
+      "grad_norm": 1.7726011276245117,
+      "learning_rate": 1.9125224119791006e-06,
+      "loss": 2.2034,
+      "step": 15243
+    },
+    {
+      "epoch": 0.937746062992126,
+      "grad_norm": 1.8024632930755615,
+      "learning_rate": 1.9087603708835e-06,
+      "loss": 1.8147,
+      "step": 15244
+    },
+    {
+      "epoch": 0.9378075787401575,
+      "grad_norm": 1.2977474927902222,
+      "learning_rate": 1.905001997853284e-06,
+      "loss": 3.4075,
+      "step": 15245
+    },
+    {
+      "epoch": 0.937869094488189,
+      "grad_norm": 1.3116579055786133,
+      "learning_rate": 1.901247293029007e-06,
+      "loss": 2.3049,
+      "step": 15246
+    },
+    {
+      "epoch": 0.9379306102362205,
+      "grad_norm": 1.208544135093689,
+      "learning_rate": 1.8974962565510678e-06,
+      "loss": 2.6746,
+      "step": 15247
+    },
+    {
+      "epoch": 0.937992125984252,
+      "grad_norm": 1.325448989868164,
+      "learning_rate": 1.8937488885597321e-06,
+      "loss": 3.3007,
+      "step": 15248
+    },
+    {
+      "epoch": 0.9380536417322834,
+      "grad_norm": 1.1738418340682983,
+      "learning_rate": 1.890005189195143e-06,
+      "loss": 2.7899,
+      "step": 15249
+    },
+    {
+      "epoch": 0.9381151574803149,
+      "grad_norm": 1.1732416152954102,
+      "learning_rate": 1.8862651585972667e-06,
+      "loss": 2.391,
+      "step": 15250
+    },
+    {
+      "epoch": 0.9381766732283464,
+      "grad_norm": 1.2499340772628784,
+      "learning_rate": 1.8825287969059801e-06,
+      "loss": 2.5477,
+      "step": 15251
+    },
+    {
+      "epoch": 0.938238188976378,
+      "grad_norm": 1.4404222965240479,
+      "learning_rate": 1.8787961042610047e-06,
+      "loss": 2.3654,
+      "step": 15252
+    },
+    {
+      "epoch": 0.9382997047244095,
+      "grad_norm": 1.2408455610275269,
+      "learning_rate": 1.8750670808019177e-06,
+      "loss": 2.3995,
+      "step": 15253
+    },
+    {
+      "epoch": 0.938361220472441,
+      "grad_norm": 1.4040318727493286,
+      "learning_rate": 1.8713417266681632e-06,
+      "loss": 2.2698,
+      "step": 15254
+    },
+    {
+      "epoch": 0.9384227362204725,
+      "grad_norm": 1.7109489440917969,
+      "learning_rate": 1.867620041999041e-06,
+      "loss": 2.694,
+      "step": 15255
+    },
+    {
+      "epoch": 0.9384842519685039,
+      "grad_norm": 1.2882298231124878,
+      "learning_rate": 1.8639020269337283e-06,
+      "loss": 2.571,
+      "step": 15256
+    },
+    {
+      "epoch": 0.9385457677165354,
+      "grad_norm": 1.4602793455123901,
+      "learning_rate": 1.8601876816112695e-06,
+      "loss": 2.9495,
+      "step": 15257
+    },
+    {
+      "epoch": 0.9386072834645669,
+      "grad_norm": 1.33034348487854,
+      "learning_rate": 1.8564770061705316e-06,
+      "loss": 2.1403,
+      "step": 15258
+    },
+    {
+      "epoch": 0.9386687992125984,
+      "grad_norm": 1.2954976558685303,
+      "learning_rate": 1.8527700007503036e-06,
+      "loss": 2.6962,
+      "step": 15259
+    },
+    {
+      "epoch": 0.93873031496063,
+      "grad_norm": 1.2929960489273071,
+      "learning_rate": 1.8490666654891853e-06,
+      "loss": 2.8636,
+      "step": 15260
+    },
+    {
+      "epoch": 0.9387918307086615,
+      "grad_norm": 1.4473017454147339,
+      "learning_rate": 1.8453670005256774e-06,
+      "loss": 1.9359,
+      "step": 15261
+    },
+    {
+      "epoch": 0.9388533464566929,
+      "grad_norm": 1.2862704992294312,
+      "learning_rate": 1.8416710059981024e-06,
+      "loss": 2.8639,
+      "step": 15262
+    },
+    {
+      "epoch": 0.9389148622047244,
+      "grad_norm": 1.4852159023284912,
+      "learning_rate": 1.8379786820447053e-06,
+      "loss": 2.3288,
+      "step": 15263
+    },
+    {
+      "epoch": 0.9389763779527559,
+      "grad_norm": 1.5293407440185547,
+      "learning_rate": 1.8342900288035314e-06,
+      "loss": 2.1486,
+      "step": 15264
+    },
+    {
+      "epoch": 0.9390378937007874,
+      "grad_norm": 1.2516520023345947,
+      "learning_rate": 1.8306050464125146e-06,
+      "loss": 2.1922,
+      "step": 15265
+    },
+    {
+      "epoch": 0.9390994094488189,
+      "grad_norm": 1.5216255187988281,
+      "learning_rate": 1.8269237350094781e-06,
+      "loss": 2.6076,
+      "step": 15266
+    },
+    {
+      "epoch": 0.9391609251968503,
+      "grad_norm": 1.2666856050491333,
+      "learning_rate": 1.8232460947320562e-06,
+      "loss": 2.2438,
+      "step": 15267
+    },
+    {
+      "epoch": 0.9392224409448819,
+      "grad_norm": 1.5203399658203125,
+      "learning_rate": 1.8195721257177832e-06,
+      "loss": 2.4143,
+      "step": 15268
+    },
+    {
+      "epoch": 0.9392839566929134,
+      "grad_norm": 1.1878981590270996,
+      "learning_rate": 1.8159018281040384e-06,
+      "loss": 2.661,
+      "step": 15269
+    },
+    {
+      "epoch": 0.9393454724409449,
+      "grad_norm": 1.2035224437713623,
+      "learning_rate": 1.8122352020280787e-06,
+      "loss": 2.6388,
+      "step": 15270
+    },
+    {
+      "epoch": 0.9394069881889764,
+      "grad_norm": 1.3554424047470093,
+      "learning_rate": 1.808572247627016e-06,
+      "loss": 1.8218,
+      "step": 15271
+    },
+    {
+      "epoch": 0.9394685039370079,
+      "grad_norm": 1.1823772192001343,
+      "learning_rate": 1.8049129650378194e-06,
+      "loss": 2.9796,
+      "step": 15272
+    },
+    {
+      "epoch": 0.9395300196850394,
+      "grad_norm": 1.4193603992462158,
+      "learning_rate": 1.8012573543973232e-06,
+      "loss": 2.4893,
+      "step": 15273
+    },
+    {
+      "epoch": 0.9395915354330708,
+      "grad_norm": 1.4013700485229492,
+      "learning_rate": 1.7976054158422407e-06,
+      "loss": 1.6542,
+      "step": 15274
+    },
+    {
+      "epoch": 0.9396530511811023,
+      "grad_norm": 1.4430749416351318,
+      "learning_rate": 1.7939571495091178e-06,
+      "loss": 2.234,
+      "step": 15275
+    },
+    {
+      "epoch": 0.9397145669291339,
+      "grad_norm": 1.3722418546676636,
+      "learning_rate": 1.7903125555343903e-06,
+      "loss": 1.724,
+      "step": 15276
+    },
+    {
+      "epoch": 0.9397760826771654,
+      "grad_norm": 1.259476900100708,
+      "learning_rate": 1.7866716340543266e-06,
+      "loss": 2.4005,
+      "step": 15277
+    },
+    {
+      "epoch": 0.9398375984251969,
+      "grad_norm": 1.3432716131210327,
+      "learning_rate": 1.7830343852050957e-06,
+      "loss": 2.1848,
+      "step": 15278
+    },
+    {
+      "epoch": 0.9398991141732284,
+      "grad_norm": 1.4525094032287598,
+      "learning_rate": 1.779400809122711e-06,
+      "loss": 2.2962,
+      "step": 15279
+    },
+    {
+      "epoch": 0.9399606299212598,
+      "grad_norm": 1.249059796333313,
+      "learning_rate": 1.7757709059430526e-06,
+      "loss": 2.9817,
+      "step": 15280
+    },
+    {
+      "epoch": 0.9400221456692913,
+      "grad_norm": 1.6554709672927856,
+      "learning_rate": 1.7721446758018345e-06,
+      "loss": 2.2243,
+      "step": 15281
+    },
+    {
+      "epoch": 0.9400836614173228,
+      "grad_norm": 1.3425332307815552,
+      "learning_rate": 1.7685221188346812e-06,
+      "loss": 2.8414,
+      "step": 15282
+    },
+    {
+      "epoch": 0.9401451771653543,
+      "grad_norm": 1.3891528844833374,
+      "learning_rate": 1.76490323517704e-06,
+      "loss": 2.5929,
+      "step": 15283
+    },
+    {
+      "epoch": 0.9402066929133859,
+      "grad_norm": 1.144274115562439,
+      "learning_rate": 1.7612880249642472e-06,
+      "loss": 2.6195,
+      "step": 15284
+    },
+    {
+      "epoch": 0.9402682086614174,
+      "grad_norm": 1.4236217737197876,
+      "learning_rate": 1.7576764883314833e-06,
+      "loss": 2.843,
+      "step": 15285
+    },
+    {
+      "epoch": 0.9403297244094488,
+      "grad_norm": 2.9258339405059814,
+      "learning_rate": 1.7540686254138072e-06,
+      "loss": 2.2195,
+      "step": 15286
+    },
+    {
+      "epoch": 0.9403912401574803,
+      "grad_norm": 1.3927582502365112,
+      "learning_rate": 1.7504644363461332e-06,
+      "loss": 1.9173,
+      "step": 15287
+    },
+    {
+      "epoch": 0.9404527559055118,
+      "grad_norm": 1.351745843887329,
+      "learning_rate": 1.746863921263242e-06,
+      "loss": 1.5506,
+      "step": 15288
+    },
+    {
+      "epoch": 0.9405142716535433,
+      "grad_norm": 1.302361011505127,
+      "learning_rate": 1.7432670802997486e-06,
+      "loss": 2.4631,
+      "step": 15289
+    },
+    {
+      "epoch": 0.9405757874015748,
+      "grad_norm": 1.3299295902252197,
+      "learning_rate": 1.7396739135901784e-06,
+      "loss": 2.5345,
+      "step": 15290
+    },
+    {
+      "epoch": 0.9406373031496063,
+      "grad_norm": 1.1715402603149414,
+      "learning_rate": 1.736084421268891e-06,
+      "loss": 2.2266,
+      "step": 15291
+    },
+    {
+      "epoch": 0.9406988188976378,
+      "grad_norm": 1.2495388984680176,
+      "learning_rate": 1.7324986034701118e-06,
+      "loss": 2.8119,
+      "step": 15292
+    },
+    {
+      "epoch": 0.9407603346456693,
+      "grad_norm": 1.4012154340744019,
+      "learning_rate": 1.728916460327934e-06,
+      "loss": 2.0746,
+      "step": 15293
+    },
+    {
+      "epoch": 0.9408218503937008,
+      "grad_norm": 1.5128456354141235,
+      "learning_rate": 1.7253379919763057e-06,
+      "loss": 1.9452,
+      "step": 15294
+    },
+    {
+      "epoch": 0.9408833661417323,
+      "grad_norm": 1.2013189792633057,
+      "learning_rate": 1.7217631985490312e-06,
+      "loss": 2.9009,
+      "step": 15295
+    },
+    {
+      "epoch": 0.9409448818897638,
+      "grad_norm": 1.179030418395996,
+      "learning_rate": 1.7181920801798035e-06,
+      "loss": 3.003,
+      "step": 15296
+    },
+    {
+      "epoch": 0.9410063976377953,
+      "grad_norm": 1.3699395656585693,
+      "learning_rate": 1.7146246370021601e-06,
+      "loss": 2.5382,
+      "step": 15297
+    },
+    {
+      "epoch": 0.9410679133858267,
+      "grad_norm": 1.399265170097351,
+      "learning_rate": 1.7110608691495056e-06,
+      "loss": 2.1975,
+      "step": 15298
+    },
+    {
+      "epoch": 0.9411294291338582,
+      "grad_norm": 1.5220972299575806,
+      "learning_rate": 1.7075007767550888e-06,
+      "loss": 2.0624,
+      "step": 15299
+    },
+    {
+      "epoch": 0.9411909448818898,
+      "grad_norm": 1.4624673128128052,
+      "learning_rate": 1.7039443599520589e-06,
+      "loss": 1.9734,
+      "step": 15300
+    },
+    {
+      "epoch": 0.9412524606299213,
+      "grad_norm": 1.3858426809310913,
+      "learning_rate": 1.700391618873398e-06,
+      "loss": 2.4051,
+      "step": 15301
+    },
+    {
+      "epoch": 0.9413139763779528,
+      "grad_norm": 2.3654584884643555,
+      "learning_rate": 1.696842553651956e-06,
+      "loss": 2.2775,
+      "step": 15302
+    },
+    {
+      "epoch": 0.9413754921259843,
+      "grad_norm": 1.5312387943267822,
+      "learning_rate": 1.6932971644204375e-06,
+      "loss": 3.5087,
+      "step": 15303
+    },
+    {
+      "epoch": 0.9414370078740157,
+      "grad_norm": 1.3032336235046387,
+      "learning_rate": 1.6897554513114366e-06,
+      "loss": 1.9835,
+      "step": 15304
+    },
+    {
+      "epoch": 0.9414985236220472,
+      "grad_norm": 1.1545963287353516,
+      "learning_rate": 1.6862174144573918e-06,
+      "loss": 2.7199,
+      "step": 15305
+    },
+    {
+      "epoch": 0.9415600393700787,
+      "grad_norm": 1.2893446683883667,
+      "learning_rate": 1.682683053990608e-06,
+      "loss": 2.3369,
+      "step": 15306
+    },
+    {
+      "epoch": 0.9416215551181102,
+      "grad_norm": 1.420468807220459,
+      "learning_rate": 1.6791523700432465e-06,
+      "loss": 2.1388,
+      "step": 15307
+    },
+    {
+      "epoch": 0.9416830708661418,
+      "grad_norm": 1.2672510147094727,
+      "learning_rate": 1.675625362747324e-06,
+      "loss": 2.1218,
+      "step": 15308
+    },
+    {
+      "epoch": 0.9417445866141733,
+      "grad_norm": 1.0958353281021118,
+      "learning_rate": 1.6721020322347347e-06,
+      "loss": 1.8163,
+      "step": 15309
+    },
+    {
+      "epoch": 0.9418061023622047,
+      "grad_norm": 1.302221417427063,
+      "learning_rate": 1.6685823786372513e-06,
+      "loss": 2.7945,
+      "step": 15310
+    },
+    {
+      "epoch": 0.9418676181102362,
+      "grad_norm": 1.311853051185608,
+      "learning_rate": 1.6650664020864791e-06,
+      "loss": 2.3946,
+      "step": 15311
+    },
+    {
+      "epoch": 0.9419291338582677,
+      "grad_norm": 1.350628137588501,
+      "learning_rate": 1.6615541027138803e-06,
+      "loss": 1.8675,
+      "step": 15312
+    },
+    {
+      "epoch": 0.9419906496062992,
+      "grad_norm": 1.1917823553085327,
+      "learning_rate": 1.6580454806508161e-06,
+      "loss": 2.2028,
+      "step": 15313
+    },
+    {
+      "epoch": 0.9420521653543307,
+      "grad_norm": 1.3030128479003906,
+      "learning_rate": 1.6545405360284816e-06,
+      "loss": 2.1501,
+      "step": 15314
+    },
+    {
+      "epoch": 0.9421136811023622,
+      "grad_norm": 1.1726900339126587,
+      "learning_rate": 1.6510392689779386e-06,
+      "loss": 2.4165,
+      "step": 15315
+    },
+    {
+      "epoch": 0.9421751968503937,
+      "grad_norm": 1.5347737073898315,
+      "learning_rate": 1.6475416796301156e-06,
+      "loss": 2.1821,
+      "step": 15316
+    },
+    {
+      "epoch": 0.9422367125984252,
+      "grad_norm": 1.3170112371444702,
+      "learning_rate": 1.6440477681158084e-06,
+      "loss": 2.2815,
+      "step": 15317
+    },
+    {
+      "epoch": 0.9422982283464567,
+      "grad_norm": 1.1884169578552246,
+      "learning_rate": 1.6405575345656676e-06,
+      "loss": 2.2303,
+      "step": 15318
+    },
+    {
+      "epoch": 0.9423597440944882,
+      "grad_norm": 1.2656147480010986,
+      "learning_rate": 1.6370709791102113e-06,
+      "loss": 2.1244,
+      "step": 15319
+    },
+    {
+      "epoch": 0.9424212598425197,
+      "grad_norm": 1.2801895141601562,
+      "learning_rate": 1.6335881018798128e-06,
+      "loss": 2.2685,
+      "step": 15320
+    },
+    {
+      "epoch": 0.9424827755905512,
+      "grad_norm": 1.3092862367630005,
+      "learning_rate": 1.6301089030047011e-06,
+      "loss": 2.402,
+      "step": 15321
+    },
+    {
+      "epoch": 0.9425442913385826,
+      "grad_norm": 1.5814423561096191,
+      "learning_rate": 1.6266333826149948e-06,
+      "loss": 2.4722,
+      "step": 15322
+    },
+    {
+      "epoch": 0.9426058070866141,
+      "grad_norm": 1.527178168296814,
+      "learning_rate": 1.6231615408406675e-06,
+      "loss": 2.3587,
+      "step": 15323
+    },
+    {
+      "epoch": 0.9426673228346457,
+      "grad_norm": 1.1196706295013428,
+      "learning_rate": 1.6196933778115263e-06,
+      "loss": 2.073,
+      "step": 15324
+    },
+    {
+      "epoch": 0.9427288385826772,
+      "grad_norm": 1.329309105873108,
+      "learning_rate": 1.616228893657279e-06,
+      "loss": 2.4283,
+      "step": 15325
+    },
+    {
+      "epoch": 0.9427903543307087,
+      "grad_norm": 1.3445128202438354,
+      "learning_rate": 1.6127680885074548e-06,
+      "loss": 2.7376,
+      "step": 15326
+    },
+    {
+      "epoch": 0.9428518700787402,
+      "grad_norm": 1.484100580215454,
+      "learning_rate": 1.6093109624914838e-06,
+      "loss": 2.2996,
+      "step": 15327
+    },
+    {
+      "epoch": 0.9429133858267716,
+      "grad_norm": 1.3234502077102661,
+      "learning_rate": 1.6058575157386401e-06,
+      "loss": 2.5454,
+      "step": 15328
+    },
+    {
+      "epoch": 0.9429749015748031,
+      "grad_norm": 1.4183136224746704,
+      "learning_rate": 1.6024077483780654e-06,
+      "loss": 2.537,
+      "step": 15329
+    },
+    {
+      "epoch": 0.9430364173228346,
+      "grad_norm": 1.1828879117965698,
+      "learning_rate": 1.5989616605387669e-06,
+      "loss": 2.3687,
+      "step": 15330
+    },
+    {
+      "epoch": 0.9430979330708661,
+      "grad_norm": 1.1767504215240479,
+      "learning_rate": 1.5955192523495976e-06,
+      "loss": 2.567,
+      "step": 15331
+    },
+    {
+      "epoch": 0.9431594488188977,
+      "grad_norm": 1.1984174251556396,
+      "learning_rate": 1.5920805239392877e-06,
+      "loss": 1.8883,
+      "step": 15332
+    },
+    {
+      "epoch": 0.9432209645669292,
+      "grad_norm": 1.2656266689300537,
+      "learning_rate": 1.5886454754364454e-06,
+      "loss": 2.3989,
+      "step": 15333
+    },
+    {
+      "epoch": 0.9432824803149606,
+      "grad_norm": 1.2032526731491089,
+      "learning_rate": 1.5852141069694904e-06,
+      "loss": 2.6129,
+      "step": 15334
+    },
+    {
+      "epoch": 0.9433439960629921,
+      "grad_norm": 1.244342565536499,
+      "learning_rate": 1.5817864186667531e-06,
+      "loss": 2.4863,
+      "step": 15335
+    },
+    {
+      "epoch": 0.9434055118110236,
+      "grad_norm": 1.3120938539505005,
+      "learning_rate": 1.5783624106563976e-06,
+      "loss": 2.2736,
+      "step": 15336
+    },
+    {
+      "epoch": 0.9434670275590551,
+      "grad_norm": 1.3700250387191772,
+      "learning_rate": 1.5749420830664995e-06,
+      "loss": 1.9922,
+      "step": 15337
+    },
+    {
+      "epoch": 0.9435285433070866,
+      "grad_norm": 1.232769250869751,
+      "learning_rate": 1.5715254360249232e-06,
+      "loss": 2.3427,
+      "step": 15338
+    },
+    {
+      "epoch": 0.9435900590551181,
+      "grad_norm": 1.3280576467514038,
+      "learning_rate": 1.568112469659433e-06,
+      "loss": 2.6843,
+      "step": 15339
+    },
+    {
+      "epoch": 0.9436515748031497,
+      "grad_norm": 1.13644540309906,
+      "learning_rate": 1.564703184097671e-06,
+      "loss": 2.4467,
+      "step": 15340
+    },
+    {
+      "epoch": 0.9437130905511811,
+      "grad_norm": 1.2990481853485107,
+      "learning_rate": 1.5612975794671247e-06,
+      "loss": 2.4552,
+      "step": 15341
+    },
+    {
+      "epoch": 0.9437746062992126,
+      "grad_norm": 1.1781578063964844,
+      "learning_rate": 1.5578956558951363e-06,
+      "loss": 1.6425,
+      "step": 15342
+    },
+    {
+      "epoch": 0.9438361220472441,
+      "grad_norm": 1.2962654829025269,
+      "learning_rate": 1.5544974135089263e-06,
+      "loss": 2.2528,
+      "step": 15343
+    },
+    {
+      "epoch": 0.9438976377952756,
+      "grad_norm": 1.601889729499817,
+      "learning_rate": 1.5511028524355598e-06,
+      "loss": 2.6141,
+      "step": 15344
+    },
+    {
+      "epoch": 0.9439591535433071,
+      "grad_norm": 1.1765259504318237,
+      "learning_rate": 1.5477119728019795e-06,
+      "loss": 2.4499,
+      "step": 15345
+    },
+    {
+      "epoch": 0.9440206692913385,
+      "grad_norm": 1.3150969743728638,
+      "learning_rate": 1.5443247747349953e-06,
+      "loss": 2.8938,
+      "step": 15346
+    },
+    {
+      "epoch": 0.94408218503937,
+      "grad_norm": 1.2426766157150269,
+      "learning_rate": 1.5409412583612503e-06,
+      "loss": 2.7466,
+      "step": 15347
+    },
+    {
+      "epoch": 0.9441437007874016,
+      "grad_norm": 1.3664613962173462,
+      "learning_rate": 1.5375614238072877e-06,
+      "loss": 2.4701,
+      "step": 15348
+    },
+    {
+      "epoch": 0.9442052165354331,
+      "grad_norm": 1.1110583543777466,
+      "learning_rate": 1.5341852711994842e-06,
+      "loss": 1.5967,
+      "step": 15349
+    },
+    {
+      "epoch": 0.9442667322834646,
+      "grad_norm": 1.4912399053573608,
+      "learning_rate": 1.5308128006640942e-06,
+      "loss": 2.7773,
+      "step": 15350
+    },
+    {
+      "epoch": 0.9443282480314961,
+      "grad_norm": 1.2439725399017334,
+      "learning_rate": 1.527444012327217e-06,
+      "loss": 2.1993,
+      "step": 15351
+    },
+    {
+      "epoch": 0.9443897637795275,
+      "grad_norm": 1.3677555322647095,
+      "learning_rate": 1.5240789063148407e-06,
+      "loss": 1.922,
+      "step": 15352
+    },
+    {
+      "epoch": 0.944451279527559,
+      "grad_norm": 1.1284335851669312,
+      "learning_rate": 1.5207174827527982e-06,
+      "loss": 3.2102,
+      "step": 15353
+    },
+    {
+      "epoch": 0.9445127952755905,
+      "grad_norm": 1.3043136596679688,
+      "learning_rate": 1.5173597417667773e-06,
+      "loss": 2.3116,
+      "step": 15354
+    },
+    {
+      "epoch": 0.944574311023622,
+      "grad_norm": 1.2207112312316895,
+      "learning_rate": 1.5140056834823556e-06,
+      "loss": 2.215,
+      "step": 15355
+    },
+    {
+      "epoch": 0.9446358267716536,
+      "grad_norm": 1.5419986248016357,
+      "learning_rate": 1.5106553080249552e-06,
+      "loss": 2.8495,
+      "step": 15356
+    },
+    {
+      "epoch": 0.9446973425196851,
+      "grad_norm": 1.9795472621917725,
+      "learning_rate": 1.5073086155198423e-06,
+      "loss": 2.3514,
+      "step": 15357
+    },
+    {
+      "epoch": 0.9447588582677166,
+      "grad_norm": 1.7472119331359863,
+      "learning_rate": 1.5039656060921835e-06,
+      "loss": 1.1269,
+      "step": 15358
+    },
+    {
+      "epoch": 0.944820374015748,
+      "grad_norm": 1.3216514587402344,
+      "learning_rate": 1.500626279866979e-06,
+      "loss": 2.1295,
+      "step": 15359
+    },
+    {
+      "epoch": 0.9448818897637795,
+      "grad_norm": 1.2717502117156982,
+      "learning_rate": 1.4972906369691064e-06,
+      "loss": 2.75,
+      "step": 15360
+    },
+    {
+      "epoch": 0.944943405511811,
+      "grad_norm": 1.2228740453720093,
+      "learning_rate": 1.4939586775232884e-06,
+      "loss": 2.5368,
+      "step": 15361
+    },
+    {
+      "epoch": 0.9450049212598425,
+      "grad_norm": 1.2576797008514404,
+      "learning_rate": 1.4906304016541473e-06,
+      "loss": 2.6799,
+      "step": 15362
+    },
+    {
+      "epoch": 0.945066437007874,
+      "grad_norm": 1.269008755683899,
+      "learning_rate": 1.487305809486117e-06,
+      "loss": 1.738,
+      "step": 15363
+    },
+    {
+      "epoch": 0.9451279527559056,
+      "grad_norm": 1.197523593902588,
+      "learning_rate": 1.4839849011435313e-06,
+      "loss": 2.6758,
+      "step": 15364
+    },
+    {
+      "epoch": 0.945189468503937,
+      "grad_norm": 1.4199039936065674,
+      "learning_rate": 1.480667676750569e-06,
+      "loss": 2.4282,
+      "step": 15365
+    },
+    {
+      "epoch": 0.9452509842519685,
+      "grad_norm": 1.5774574279785156,
+      "learning_rate": 1.4773541364312748e-06,
+      "loss": 2.1567,
+      "step": 15366
+    },
+    {
+      "epoch": 0.9453125,
+      "grad_norm": 1.3314203023910522,
+      "learning_rate": 1.47404428030955e-06,
+      "loss": 2.2005,
+      "step": 15367
+    },
+    {
+      "epoch": 0.9453740157480315,
+      "grad_norm": 1.2304906845092773,
+      "learning_rate": 1.4707381085091842e-06,
+      "loss": 2.5855,
+      "step": 15368
+    },
+    {
+      "epoch": 0.945435531496063,
+      "grad_norm": 1.526516079902649,
+      "learning_rate": 1.467435621153801e-06,
+      "loss": 2.5989,
+      "step": 15369
+    },
+    {
+      "epoch": 0.9454970472440944,
+      "grad_norm": 1.4304156303405762,
+      "learning_rate": 1.4641368183668902e-06,
+      "loss": 2.5767,
+      "step": 15370
+    },
+    {
+      "epoch": 0.945558562992126,
+      "grad_norm": 1.3033299446105957,
+      "learning_rate": 1.460841700271809e-06,
+      "loss": 2.6291,
+      "step": 15371
+    },
+    {
+      "epoch": 0.9456200787401575,
+      "grad_norm": 1.7187018394470215,
+      "learning_rate": 1.4575502669917808e-06,
+      "loss": 2.6027,
+      "step": 15372
+    },
+    {
+      "epoch": 0.945681594488189,
+      "grad_norm": 1.1661909818649292,
+      "learning_rate": 1.4542625186498737e-06,
+      "loss": 2.8875,
+      "step": 15373
+    },
+    {
+      "epoch": 0.9457431102362205,
+      "grad_norm": 1.4358869791030884,
+      "learning_rate": 1.4509784553690563e-06,
+      "loss": 2.0616,
+      "step": 15374
+    },
+    {
+      "epoch": 0.945804625984252,
+      "grad_norm": 1.2102667093276978,
+      "learning_rate": 1.4476980772721305e-06,
+      "loss": 1.9826,
+      "step": 15375
+    },
+    {
+      "epoch": 0.9458661417322834,
+      "grad_norm": 1.1151635646820068,
+      "learning_rate": 1.444421384481731e-06,
+      "loss": 2.645,
+      "step": 15376
+    },
+    {
+      "epoch": 0.9459276574803149,
+      "grad_norm": 1.1918047666549683,
+      "learning_rate": 1.441148377120427e-06,
+      "loss": 3.0606,
+      "step": 15377
+    },
+    {
+      "epoch": 0.9459891732283464,
+      "grad_norm": 1.3524401187896729,
+      "learning_rate": 1.437879055310587e-06,
+      "loss": 2.2271,
+      "step": 15378
+    },
+    {
+      "epoch": 0.946050688976378,
+      "grad_norm": 1.5792558193206787,
+      "learning_rate": 1.4346134191744687e-06,
+      "loss": 2.154,
+      "step": 15379
+    },
+    {
+      "epoch": 0.9461122047244095,
+      "grad_norm": 1.3975802659988403,
+      "learning_rate": 1.4313514688341967e-06,
+      "loss": 1.5438,
+      "step": 15380
+    },
+    {
+      "epoch": 0.946173720472441,
+      "grad_norm": 1.8646305799484253,
+      "learning_rate": 1.4280932044117512e-06,
+      "loss": 2.0,
+      "step": 15381
+    },
+    {
+      "epoch": 0.9462352362204725,
+      "grad_norm": 1.4153906106948853,
+      "learning_rate": 1.424838626028957e-06,
+      "loss": 2.6497,
+      "step": 15382
+    },
+    {
+      "epoch": 0.9462967519685039,
+      "grad_norm": 1.0514554977416992,
+      "learning_rate": 1.421587733807539e-06,
+      "loss": 3.1155,
+      "step": 15383
+    },
+    {
+      "epoch": 0.9463582677165354,
+      "grad_norm": 1.548849105834961,
+      "learning_rate": 1.4183405278690444e-06,
+      "loss": 3.2417,
+      "step": 15384
+    },
+    {
+      "epoch": 0.9464197834645669,
+      "grad_norm": 1.1193547248840332,
+      "learning_rate": 1.415097008334898e-06,
+      "loss": 2.1969,
+      "step": 15385
+    },
+    {
+      "epoch": 0.9464812992125984,
+      "grad_norm": 1.5561708211898804,
+      "learning_rate": 1.4118571753264143e-06,
+      "loss": 1.9743,
+      "step": 15386
+    },
+    {
+      "epoch": 0.94654281496063,
+      "grad_norm": 1.2089277505874634,
+      "learning_rate": 1.4086210289647294e-06,
+      "loss": 2.4584,
+      "step": 15387
+    },
+    {
+      "epoch": 0.9466043307086615,
+      "grad_norm": 1.3267555236816406,
+      "learning_rate": 1.4053885693708468e-06,
+      "loss": 2.7664,
+      "step": 15388
+    },
+    {
+      "epoch": 0.9466658464566929,
+      "grad_norm": 1.1610093116760254,
+      "learning_rate": 1.4021597966656696e-06,
+      "loss": 2.5269,
+      "step": 15389
+    },
+    {
+      "epoch": 0.9467273622047244,
+      "grad_norm": 1.438423991203308,
+      "learning_rate": 1.3989347109699124e-06,
+      "loss": 2.8799,
+      "step": 15390
+    },
+    {
+      "epoch": 0.9467888779527559,
+      "grad_norm": 1.406345248222351,
+      "learning_rate": 1.3957133124041788e-06,
+      "loss": 2.6765,
+      "step": 15391
+    },
+    {
+      "epoch": 0.9468503937007874,
+      "grad_norm": 1.2790188789367676,
+      "learning_rate": 1.392495601088928e-06,
+      "loss": 2.1758,
+      "step": 15392
+    },
+    {
+      "epoch": 0.9469119094488189,
+      "grad_norm": 1.305721640586853,
+      "learning_rate": 1.3892815771445078e-06,
+      "loss": 1.9686,
+      "step": 15393
+    },
+    {
+      "epoch": 0.9469734251968503,
+      "grad_norm": 1.198533296585083,
+      "learning_rate": 1.3860712406910669e-06,
+      "loss": 2.4997,
+      "step": 15394
+    },
+    {
+      "epoch": 0.9470349409448819,
+      "grad_norm": 1.326365351676941,
+      "learning_rate": 1.3828645918486981e-06,
+      "loss": 1.964,
+      "step": 15395
+    },
+    {
+      "epoch": 0.9470964566929134,
+      "grad_norm": 1.3397300243377686,
+      "learning_rate": 1.3796616307372723e-06,
+      "loss": 1.9798,
+      "step": 15396
+    },
+    {
+      "epoch": 0.9471579724409449,
+      "grad_norm": 1.391588568687439,
+      "learning_rate": 1.3764623574765933e-06,
+      "loss": 2.0589,
+      "step": 15397
+    },
+    {
+      "epoch": 0.9472194881889764,
+      "grad_norm": 1.4299522638320923,
+      "learning_rate": 1.373266772186277e-06,
+      "loss": 2.3392,
+      "step": 15398
+    },
+    {
+      "epoch": 0.9472810039370079,
+      "grad_norm": 1.19617760181427,
+      "learning_rate": 1.370074874985816e-06,
+      "loss": 1.5124,
+      "step": 15399
+    },
+    {
+      "epoch": 0.9473425196850394,
+      "grad_norm": 1.2611356973648071,
+      "learning_rate": 1.3668866659945823e-06,
+      "loss": 2.3511,
+      "step": 15400
+    },
+    {
+      "epoch": 0.9474040354330708,
+      "grad_norm": 1.2695822715759277,
+      "learning_rate": 1.363702145331802e-06,
+      "loss": 2.7213,
+      "step": 15401
+    },
+    {
+      "epoch": 0.9474655511811023,
+      "grad_norm": 1.4542473554611206,
+      "learning_rate": 1.3605213131165473e-06,
+      "loss": 2.2923,
+      "step": 15402
+    },
+    {
+      "epoch": 0.9475270669291339,
+      "grad_norm": 1.4335880279541016,
+      "learning_rate": 1.357344169467778e-06,
+      "loss": 2.1972,
+      "step": 15403
+    },
+    {
+      "epoch": 0.9475885826771654,
+      "grad_norm": 1.3172589540481567,
+      "learning_rate": 1.354170714504266e-06,
+      "loss": 2.5797,
+      "step": 15404
+    },
+    {
+      "epoch": 0.9476500984251969,
+      "grad_norm": 1.2687771320343018,
+      "learning_rate": 1.351000948344716e-06,
+      "loss": 2.3981,
+      "step": 15405
+    },
+    {
+      "epoch": 0.9477116141732284,
+      "grad_norm": 1.1029621362686157,
+      "learning_rate": 1.347834871107656e-06,
+      "loss": 2.4753,
+      "step": 15406
+    },
+    {
+      "epoch": 0.9477731299212598,
+      "grad_norm": 1.135325312614441,
+      "learning_rate": 1.3446724829114688e-06,
+      "loss": 1.7765,
+      "step": 15407
+    },
+    {
+      "epoch": 0.9478346456692913,
+      "grad_norm": 1.3786964416503906,
+      "learning_rate": 1.341513783874404e-06,
+      "loss": 2.5035,
+      "step": 15408
+    },
+    {
+      "epoch": 0.9478961614173228,
+      "grad_norm": 1.493129014968872,
+      "learning_rate": 1.338358774114601e-06,
+      "loss": 2.5804,
+      "step": 15409
+    },
+    {
+      "epoch": 0.9479576771653543,
+      "grad_norm": 1.4692604541778564,
+      "learning_rate": 1.3352074537500203e-06,
+      "loss": 2.4731,
+      "step": 15410
+    },
+    {
+      "epoch": 0.9480191929133859,
+      "grad_norm": 1.1698211431503296,
+      "learning_rate": 1.3320598228985125e-06,
+      "loss": 2.496,
+      "step": 15411
+    },
+    {
+      "epoch": 0.9480807086614174,
+      "grad_norm": 1.3588851690292358,
+      "learning_rate": 1.3289158816777835e-06,
+      "loss": 2.5644,
+      "step": 15412
+    },
+    {
+      "epoch": 0.9481422244094488,
+      "grad_norm": 1.208864450454712,
+      "learning_rate": 1.3257756302053947e-06,
+      "loss": 2.188,
+      "step": 15413
+    },
+    {
+      "epoch": 0.9482037401574803,
+      "grad_norm": 1.1858694553375244,
+      "learning_rate": 1.3226390685987745e-06,
+      "loss": 2.7041,
+      "step": 15414
+    },
+    {
+      "epoch": 0.9482652559055118,
+      "grad_norm": 1.2472031116485596,
+      "learning_rate": 1.3195061969752065e-06,
+      "loss": 2.4491,
+      "step": 15415
+    },
+    {
+      "epoch": 0.9483267716535433,
+      "grad_norm": 1.1734763383865356,
+      "learning_rate": 1.316377015451864e-06,
+      "loss": 2.1777,
+      "step": 15416
+    },
+    {
+      "epoch": 0.9483882874015748,
+      "grad_norm": 1.3058788776397705,
+      "learning_rate": 1.313251524145731e-06,
+      "loss": 2.2712,
+      "step": 15417
+    },
+    {
+      "epoch": 0.9484498031496063,
+      "grad_norm": 1.2125071287155151,
+      "learning_rate": 1.310129723173703e-06,
+      "loss": 2.558,
+      "step": 15418
+    },
+    {
+      "epoch": 0.9485113188976378,
+      "grad_norm": 1.1081565618515015,
+      "learning_rate": 1.307011612652509e-06,
+      "loss": 2.3482,
+      "step": 15419
+    },
+    {
+      "epoch": 0.9485728346456693,
+      "grad_norm": 1.2128994464874268,
+      "learning_rate": 1.3038971926987665e-06,
+      "loss": 3.0332,
+      "step": 15420
+    },
+    {
+      "epoch": 0.9486343503937008,
+      "grad_norm": 1.0887916088104248,
+      "learning_rate": 1.3007864634289047e-06,
+      "loss": 3.1804,
+      "step": 15421
+    },
+    {
+      "epoch": 0.9486958661417323,
+      "grad_norm": 1.525044560432434,
+      "learning_rate": 1.297679424959286e-06,
+      "loss": 1.5854,
+      "step": 15422
+    },
+    {
+      "epoch": 0.9487573818897638,
+      "grad_norm": 1.7074503898620605,
+      "learning_rate": 1.2945760774060733e-06,
+      "loss": 2.306,
+      "step": 15423
+    },
+    {
+      "epoch": 0.9488188976377953,
+      "grad_norm": 1.4394917488098145,
+      "learning_rate": 1.291476420885307e-06,
+      "loss": 2.1172,
+      "step": 15424
+    },
+    {
+      "epoch": 0.9488804133858267,
+      "grad_norm": 1.4571417570114136,
+      "learning_rate": 1.2883804555129165e-06,
+      "loss": 2.1536,
+      "step": 15425
+    },
+    {
+      "epoch": 0.9489419291338582,
+      "grad_norm": 1.6450220346450806,
+      "learning_rate": 1.2852881814046757e-06,
+      "loss": 2.3423,
+      "step": 15426
+    },
+    {
+      "epoch": 0.9490034448818898,
+      "grad_norm": 1.2844294309616089,
+      "learning_rate": 1.2821995986761925e-06,
+      "loss": 2.8018,
+      "step": 15427
+    },
+    {
+      "epoch": 0.9490649606299213,
+      "grad_norm": 1.3342326879501343,
+      "learning_rate": 1.2791147074429966e-06,
+      "loss": 2.5839,
+      "step": 15428
+    },
+    {
+      "epoch": 0.9491264763779528,
+      "grad_norm": 1.2171103954315186,
+      "learning_rate": 1.2760335078204177e-06,
+      "loss": 2.0986,
+      "step": 15429
+    },
+    {
+      "epoch": 0.9491879921259843,
+      "grad_norm": 1.285799264907837,
+      "learning_rate": 1.2729559999236751e-06,
+      "loss": 2.6438,
+      "step": 15430
+    },
+    {
+      "epoch": 0.9492495078740157,
+      "grad_norm": 1.401767611503601,
+      "learning_rate": 1.2698821838678765e-06,
+      "loss": 2.0122,
+      "step": 15431
+    },
+    {
+      "epoch": 0.9493110236220472,
+      "grad_norm": 1.280747413635254,
+      "learning_rate": 1.2668120597679411e-06,
+      "loss": 2.3855,
+      "step": 15432
+    },
+    {
+      "epoch": 0.9493725393700787,
+      "grad_norm": 1.2766493558883667,
+      "learning_rate": 1.2637456277386884e-06,
+      "loss": 2.6075,
+      "step": 15433
+    },
+    {
+      "epoch": 0.9494340551181102,
+      "grad_norm": 1.3193137645721436,
+      "learning_rate": 1.2606828878947818e-06,
+      "loss": 2.0509,
+      "step": 15434
+    },
+    {
+      "epoch": 0.9494955708661418,
+      "grad_norm": 1.3155157566070557,
+      "learning_rate": 1.2576238403507522e-06,
+      "loss": 2.9572,
+      "step": 15435
+    },
+    {
+      "epoch": 0.9495570866141733,
+      "grad_norm": 1.298224925994873,
+      "learning_rate": 1.254568485220975e-06,
+      "loss": 2.1307,
+      "step": 15436
+    },
+    {
+      "epoch": 0.9496186023622047,
+      "grad_norm": 1.5549665689468384,
+      "learning_rate": 1.251516822619725e-06,
+      "loss": 3.0595,
+      "step": 15437
+    },
+    {
+      "epoch": 0.9496801181102362,
+      "grad_norm": 1.3109790086746216,
+      "learning_rate": 1.248468852661111e-06,
+      "loss": 2.8622,
+      "step": 15438
+    },
+    {
+      "epoch": 0.9497416338582677,
+      "grad_norm": 1.2766871452331543,
+      "learning_rate": 1.2454245754591087e-06,
+      "loss": 2.3637,
+      "step": 15439
+    },
+    {
+      "epoch": 0.9498031496062992,
+      "grad_norm": 1.7490711212158203,
+      "learning_rate": 1.2423839911275604e-06,
+      "loss": 2.5317,
+      "step": 15440
+    },
+    {
+      "epoch": 0.9498646653543307,
+      "grad_norm": 1.2979559898376465,
+      "learning_rate": 1.2393470997801527e-06,
+      "loss": 1.9116,
+      "step": 15441
+    },
+    {
+      "epoch": 0.9499261811023622,
+      "grad_norm": 1.215429425239563,
+      "learning_rate": 1.2363139015304726e-06,
+      "loss": 2.714,
+      "step": 15442
+    },
+    {
+      "epoch": 0.9499876968503937,
+      "grad_norm": 1.079417109489441,
+      "learning_rate": 1.2332843964919182e-06,
+      "loss": 2.7399,
+      "step": 15443
+    },
+    {
+      "epoch": 0.9500492125984252,
+      "grad_norm": 1.3546937704086304,
+      "learning_rate": 1.230258584777788e-06,
+      "loss": 2.3396,
+      "step": 15444
+    },
+    {
+      "epoch": 0.9501107283464567,
+      "grad_norm": 1.226890206336975,
+      "learning_rate": 1.227236466501236e-06,
+      "loss": 2.2433,
+      "step": 15445
+    },
+    {
+      "epoch": 0.9501722440944882,
+      "grad_norm": 1.2885030508041382,
+      "learning_rate": 1.2242180417752712e-06,
+      "loss": 2.5944,
+      "step": 15446
+    },
+    {
+      "epoch": 0.9502337598425197,
+      "grad_norm": 1.3554818630218506,
+      "learning_rate": 1.2212033107127596e-06,
+      "loss": 2.7603,
+      "step": 15447
+    },
+    {
+      "epoch": 0.9502952755905512,
+      "grad_norm": 1.2917685508728027,
+      "learning_rate": 1.2181922734264439e-06,
+      "loss": 2.0797,
+      "step": 15448
+    },
+    {
+      "epoch": 0.9503567913385826,
+      "grad_norm": 1.2206690311431885,
+      "learning_rate": 1.215184930028912e-06,
+      "loss": 2.091,
+      "step": 15449
+    },
+    {
+      "epoch": 0.9504183070866141,
+      "grad_norm": 1.2341517210006714,
+      "learning_rate": 1.2121812806326183e-06,
+      "loss": 2.3333,
+      "step": 15450
+    },
+    {
+      "epoch": 0.9504798228346457,
+      "grad_norm": 1.4679365158081055,
+      "learning_rate": 1.2091813253498952e-06,
+      "loss": 2.9057,
+      "step": 15451
+    },
+    {
+      "epoch": 0.9505413385826772,
+      "grad_norm": 1.4749324321746826,
+      "learning_rate": 1.2061850642929195e-06,
+      "loss": 0.9705,
+      "step": 15452
+    },
+    {
+      "epoch": 0.9506028543307087,
+      "grad_norm": 1.1920175552368164,
+      "learning_rate": 1.2031924975737352e-06,
+      "loss": 1.9744,
+      "step": 15453
+    },
+    {
+      "epoch": 0.9506643700787402,
+      "grad_norm": 1.4823651313781738,
+      "learning_rate": 1.2002036253042304e-06,
+      "loss": 2.6343,
+      "step": 15454
+    },
+    {
+      "epoch": 0.9507258858267716,
+      "grad_norm": 1.4012479782104492,
+      "learning_rate": 1.1972184475961933e-06,
+      "loss": 2.1872,
+      "step": 15455
+    },
+    {
+      "epoch": 0.9507874015748031,
+      "grad_norm": 1.5947407484054565,
+      "learning_rate": 1.1942369645612461e-06,
+      "loss": 2.4025,
+      "step": 15456
+    },
+    {
+      "epoch": 0.9508489173228346,
+      "grad_norm": 1.3378145694732666,
+      "learning_rate": 1.1912591763108883e-06,
+      "loss": 2.1873,
+      "step": 15457
+    },
+    {
+      "epoch": 0.9509104330708661,
+      "grad_norm": 1.4158122539520264,
+      "learning_rate": 1.1882850829564641e-06,
+      "loss": 2.0159,
+      "step": 15458
+    },
+    {
+      "epoch": 0.9509719488188977,
+      "grad_norm": 1.2198761701583862,
+      "learning_rate": 1.1853146846091844e-06,
+      "loss": 2.3269,
+      "step": 15459
+    },
+    {
+      "epoch": 0.9510334645669292,
+      "grad_norm": 1.2923243045806885,
+      "learning_rate": 1.1823479813801274e-06,
+      "loss": 2.1919,
+      "step": 15460
+    },
+    {
+      "epoch": 0.9510949803149606,
+      "grad_norm": 1.3461817502975464,
+      "learning_rate": 1.1793849733802487e-06,
+      "loss": 2.611,
+      "step": 15461
+    },
+    {
+      "epoch": 0.9511564960629921,
+      "grad_norm": 1.2448787689208984,
+      "learning_rate": 1.1764256607203039e-06,
+      "loss": 1.89,
+      "step": 15462
+    },
+    {
+      "epoch": 0.9512180118110236,
+      "grad_norm": 1.5573008060455322,
+      "learning_rate": 1.1734700435110047e-06,
+      "loss": 1.8951,
+      "step": 15463
+    },
+    {
+      "epoch": 0.9512795275590551,
+      "grad_norm": 1.2429157495498657,
+      "learning_rate": 1.1705181218628513e-06,
+      "loss": 2.5426,
+      "step": 15464
+    },
+    {
+      "epoch": 0.9513410433070866,
+      "grad_norm": 1.273783564567566,
+      "learning_rate": 1.1675698958862336e-06,
+      "loss": 2.5591,
+      "step": 15465
+    },
+    {
+      "epoch": 0.9514025590551181,
+      "grad_norm": 1.21027672290802,
+      "learning_rate": 1.1646253656913964e-06,
+      "loss": 2.3814,
+      "step": 15466
+    },
+    {
+      "epoch": 0.9514640748031497,
+      "grad_norm": 1.3805503845214844,
+      "learning_rate": 1.1616845313884518e-06,
+      "loss": 2.1855,
+      "step": 15467
+    },
+    {
+      "epoch": 0.9515255905511811,
+      "grad_norm": 1.5273360013961792,
+      "learning_rate": 1.158747393087356e-06,
+      "loss": 2.3904,
+      "step": 15468
+    },
+    {
+      "epoch": 0.9515871062992126,
+      "grad_norm": 1.0985063314437866,
+      "learning_rate": 1.1558139508979659e-06,
+      "loss": 2.7073,
+      "step": 15469
+    },
+    {
+      "epoch": 0.9516486220472441,
+      "grad_norm": 1.300294041633606,
+      "learning_rate": 1.1528842049299603e-06,
+      "loss": 2.4065,
+      "step": 15470
+    },
+    {
+      "epoch": 0.9517101377952756,
+      "grad_norm": 1.876949429512024,
+      "learning_rate": 1.149958155292896e-06,
+      "loss": 2.2826,
+      "step": 15471
+    },
+    {
+      "epoch": 0.9517716535433071,
+      "grad_norm": 1.397376537322998,
+      "learning_rate": 1.147035802096197e-06,
+      "loss": 1.875,
+      "step": 15472
+    },
+    {
+      "epoch": 0.9518331692913385,
+      "grad_norm": 1.246342420578003,
+      "learning_rate": 1.1441171454491418e-06,
+      "loss": 2.9542,
+      "step": 15473
+    },
+    {
+      "epoch": 0.95189468503937,
+      "grad_norm": 1.1984071731567383,
+      "learning_rate": 1.1412021854608657e-06,
+      "loss": 2.6461,
+      "step": 15474
+    },
+    {
+      "epoch": 0.9519562007874016,
+      "grad_norm": 1.3544998168945312,
+      "learning_rate": 1.1382909222403704e-06,
+      "loss": 2.6576,
+      "step": 15475
+    },
+    {
+      "epoch": 0.9520177165354331,
+      "grad_norm": 1.314465880393982,
+      "learning_rate": 1.1353833558965355e-06,
+      "loss": 2.1483,
+      "step": 15476
+    },
+    {
+      "epoch": 0.9520792322834646,
+      "grad_norm": 1.21009361743927,
+      "learning_rate": 1.132479486538074e-06,
+      "loss": 2.4342,
+      "step": 15477
+    },
+    {
+      "epoch": 0.9521407480314961,
+      "grad_norm": 1.7414733171463013,
+      "learning_rate": 1.1295793142735767e-06,
+      "loss": 2.5048,
+      "step": 15478
+    },
+    {
+      "epoch": 0.9522022637795275,
+      "grad_norm": 1.7509785890579224,
+      "learning_rate": 1.1266828392115124e-06,
+      "loss": 2.8886,
+      "step": 15479
+    },
+    {
+      "epoch": 0.952263779527559,
+      "grad_norm": 1.2881535291671753,
+      "learning_rate": 1.1237900614601614e-06,
+      "loss": 2.7659,
+      "step": 15480
+    },
+    {
+      "epoch": 0.9523252952755905,
+      "grad_norm": 1.2985039949417114,
+      "learning_rate": 1.120900981127715e-06,
+      "loss": 2.8422,
+      "step": 15481
+    },
+    {
+      "epoch": 0.952386811023622,
+      "grad_norm": 1.2713862657546997,
+      "learning_rate": 1.1180155983222085e-06,
+      "loss": 2.4176,
+      "step": 15482
+    },
+    {
+      "epoch": 0.9524483267716536,
+      "grad_norm": 1.2674838304519653,
+      "learning_rate": 1.1151339131515448e-06,
+      "loss": 3.1389,
+      "step": 15483
+    },
+    {
+      "epoch": 0.9525098425196851,
+      "grad_norm": 1.2830209732055664,
+      "learning_rate": 1.112255925723471e-06,
+      "loss": 2.3751,
+      "step": 15484
+    },
+    {
+      "epoch": 0.9525713582677166,
+      "grad_norm": 1.1965936422348022,
+      "learning_rate": 1.1093816361456122e-06,
+      "loss": 2.1142,
+      "step": 15485
+    },
+    {
+      "epoch": 0.952632874015748,
+      "grad_norm": 1.336840033531189,
+      "learning_rate": 1.106511044525449e-06,
+      "loss": 2.7873,
+      "step": 15486
+    },
+    {
+      "epoch": 0.9526943897637795,
+      "grad_norm": 1.2177313566207886,
+      "learning_rate": 1.1036441509703287e-06,
+      "loss": 2.2613,
+      "step": 15487
+    },
+    {
+      "epoch": 0.952755905511811,
+      "grad_norm": 1.516982078552246,
+      "learning_rate": 1.1007809555874548e-06,
+      "loss": 1.8582,
+      "step": 15488
+    },
+    {
+      "epoch": 0.9528174212598425,
+      "grad_norm": 1.7143723964691162,
+      "learning_rate": 1.097921458483897e-06,
+      "loss": 2.4678,
+      "step": 15489
+    },
+    {
+      "epoch": 0.952878937007874,
+      "grad_norm": 1.7892767190933228,
+      "learning_rate": 1.0950656597665809e-06,
+      "loss": 1.8904,
+      "step": 15490
+    },
+    {
+      "epoch": 0.9529404527559056,
+      "grad_norm": 1.279440999031067,
+      "learning_rate": 1.09221355954231e-06,
+      "loss": 2.1287,
+      "step": 15491
+    },
+    {
+      "epoch": 0.953001968503937,
+      "grad_norm": 1.2300564050674438,
+      "learning_rate": 1.089365157917721e-06,
+      "loss": 3.0652,
+      "step": 15492
+    },
+    {
+      "epoch": 0.9530634842519685,
+      "grad_norm": 1.2872345447540283,
+      "learning_rate": 1.086520454999329e-06,
+      "loss": 2.254,
+      "step": 15493
+    },
+    {
+      "epoch": 0.953125,
+      "grad_norm": 1.3514409065246582,
+      "learning_rate": 1.0836794508935155e-06,
+      "loss": 2.3537,
+      "step": 15494
+    },
+    {
+      "epoch": 0.9531865157480315,
+      "grad_norm": 1.1729588508605957,
+      "learning_rate": 1.0808421457065176e-06,
+      "loss": 2.7866,
+      "step": 15495
+    },
+    {
+      "epoch": 0.953248031496063,
+      "grad_norm": 1.2098771333694458,
+      "learning_rate": 1.0780085395444506e-06,
+      "loss": 2.3766,
+      "step": 15496
+    },
+    {
+      "epoch": 0.9533095472440944,
+      "grad_norm": 1.4084092378616333,
+      "learning_rate": 1.0751786325132406e-06,
+      "loss": 2.5313,
+      "step": 15497
+    },
+    {
+      "epoch": 0.953371062992126,
+      "grad_norm": 1.283894658088684,
+      "learning_rate": 1.0723524247187478e-06,
+      "loss": 2.6551,
+      "step": 15498
+    },
+    {
+      "epoch": 0.9534325787401575,
+      "grad_norm": 1.5116373300552368,
+      "learning_rate": 1.0695299162666205e-06,
+      "loss": 2.789,
+      "step": 15499
+    },
+    {
+      "epoch": 0.953494094488189,
+      "grad_norm": 1.3317044973373413,
+      "learning_rate": 1.0667111072624303e-06,
+      "loss": 2.7676,
+      "step": 15500
+    },
+    {
+      "epoch": 0.9535556102362205,
+      "grad_norm": 1.322313904762268,
+      "learning_rate": 1.0638959978115815e-06,
+      "loss": 2.6456,
+      "step": 15501
+    },
+    {
+      "epoch": 0.953617125984252,
+      "grad_norm": 1.8986787796020508,
+      "learning_rate": 1.0610845880193343e-06,
+      "loss": 1.258,
+      "step": 15502
+    },
+    {
+      "epoch": 0.9536786417322834,
+      "grad_norm": 1.4150943756103516,
+      "learning_rate": 1.058276877990827e-06,
+      "loss": 2.2969,
+      "step": 15503
+    },
+    {
+      "epoch": 0.9537401574803149,
+      "grad_norm": 1.319623351097107,
+      "learning_rate": 1.0554728678310533e-06,
+      "loss": 2.3021,
+      "step": 15504
+    },
+    {
+      "epoch": 0.9538016732283464,
+      "grad_norm": 1.6745070219039917,
+      "learning_rate": 1.0526725576448627e-06,
+      "loss": 2.2905,
+      "step": 15505
+    },
+    {
+      "epoch": 0.953863188976378,
+      "grad_norm": 1.1708883047103882,
+      "learning_rate": 1.0498759475369712e-06,
+      "loss": 2.1983,
+      "step": 15506
+    },
+    {
+      "epoch": 0.9539247047244095,
+      "grad_norm": 1.635132074356079,
+      "learning_rate": 1.0470830376119511e-06,
+      "loss": 2.3869,
+      "step": 15507
+    },
+    {
+      "epoch": 0.953986220472441,
+      "grad_norm": 1.1674400568008423,
+      "learning_rate": 1.044293827974263e-06,
+      "loss": 2.7243,
+      "step": 15508
+    },
+    {
+      "epoch": 0.9540477362204725,
+      "grad_norm": 1.2961081266403198,
+      "learning_rate": 1.0415083187281904e-06,
+      "loss": 2.1026,
+      "step": 15509
+    },
+    {
+      "epoch": 0.9541092519685039,
+      "grad_norm": 1.4253239631652832,
+      "learning_rate": 1.0387265099778943e-06,
+      "loss": 1.907,
+      "step": 15510
+    },
+    {
+      "epoch": 0.9541707677165354,
+      "grad_norm": 2.3681204319000244,
+      "learning_rate": 1.0359484018274024e-06,
+      "loss": 2.0645,
+      "step": 15511
+    },
+    {
+      "epoch": 0.9542322834645669,
+      "grad_norm": 1.3078935146331787,
+      "learning_rate": 1.0331739943806096e-06,
+      "loss": 2.4675,
+      "step": 15512
+    },
+    {
+      "epoch": 0.9542937992125984,
+      "grad_norm": 1.214093565940857,
+      "learning_rate": 1.030403287741244e-06,
+      "loss": 2.3338,
+      "step": 15513
+    },
+    {
+      "epoch": 0.95435531496063,
+      "grad_norm": 1.4095491170883179,
+      "learning_rate": 1.0276362820129338e-06,
+      "loss": 2.3477,
+      "step": 15514
+    },
+    {
+      "epoch": 0.9544168307086615,
+      "grad_norm": 1.135481595993042,
+      "learning_rate": 1.0248729772991405e-06,
+      "loss": 2.3234,
+      "step": 15515
+    },
+    {
+      "epoch": 0.9544783464566929,
+      "grad_norm": 1.2709184885025024,
+      "learning_rate": 1.022113373703204e-06,
+      "loss": 2.6297,
+      "step": 15516
+    },
+    {
+      "epoch": 0.9545398622047244,
+      "grad_norm": 1.1795029640197754,
+      "learning_rate": 1.0193574713282972e-06,
+      "loss": 2.7479,
+      "step": 15517
+    },
+    {
+      "epoch": 0.9546013779527559,
+      "grad_norm": 1.3014748096466064,
+      "learning_rate": 1.0166052702775043e-06,
+      "loss": 2.7399,
+      "step": 15518
+    },
+    {
+      "epoch": 0.9546628937007874,
+      "grad_norm": 1.3901933431625366,
+      "learning_rate": 1.0138567706537094e-06,
+      "loss": 2.297,
+      "step": 15519
+    },
+    {
+      "epoch": 0.9547244094488189,
+      "grad_norm": 1.485610008239746,
+      "learning_rate": 1.0111119725597195e-06,
+      "loss": 1.8946,
+      "step": 15520
+    },
+    {
+      "epoch": 0.9547859251968503,
+      "grad_norm": 1.3014438152313232,
+      "learning_rate": 1.0083708760981747e-06,
+      "loss": 2.3883,
+      "step": 15521
+    },
+    {
+      "epoch": 0.9548474409448819,
+      "grad_norm": 1.4037678241729736,
+      "learning_rate": 1.0056334813715484e-06,
+      "loss": 1.9268,
+      "step": 15522
+    },
+    {
+      "epoch": 0.9549089566929134,
+      "grad_norm": 1.2700433731079102,
+      "learning_rate": 1.0028997884822366e-06,
+      "loss": 2.9392,
+      "step": 15523
+    },
+    {
+      "epoch": 0.9549704724409449,
+      "grad_norm": 1.2891130447387695,
+      "learning_rate": 1.0001697975324464e-06,
+      "loss": 2.6143,
+      "step": 15524
+    },
+    {
+      "epoch": 0.9550319881889764,
+      "grad_norm": 1.2822294235229492,
+      "learning_rate": 9.97443508624263e-07,
+      "loss": 2.125,
+      "step": 15525
+    },
+    {
+      "epoch": 0.9550935039370079,
+      "grad_norm": 1.284371018409729,
+      "learning_rate": 9.947209218596376e-07,
+      "loss": 2.3312,
+      "step": 15526
+    },
+    {
+      "epoch": 0.9551550196850394,
+      "grad_norm": 1.2534099817276,
+      "learning_rate": 9.920020373403894e-07,
+      "loss": 2.7218,
+      "step": 15527
+    },
+    {
+      "epoch": 0.9552165354330708,
+      "grad_norm": 1.2660061120986938,
+      "learning_rate": 9.8928685516817e-07,
+      "loss": 2.3838,
+      "step": 15528
+    },
+    {
+      "epoch": 0.9552780511811023,
+      "grad_norm": 1.3203731775283813,
+      "learning_rate": 9.865753754445427e-07,
+      "loss": 2.3085,
+      "step": 15529
+    },
+    {
+      "epoch": 0.9553395669291339,
+      "grad_norm": 1.296766757965088,
+      "learning_rate": 9.8386759827086e-07,
+      "loss": 2.4622,
+      "step": 15530
+    },
+    {
+      "epoch": 0.9554010826771654,
+      "grad_norm": 1.3852046728134155,
+      "learning_rate": 9.811635237484186e-07,
+      "loss": 2.0702,
+      "step": 15531
+    },
+    {
+      "epoch": 0.9554625984251969,
+      "grad_norm": 1.2552815675735474,
+      "learning_rate": 9.784631519783039e-07,
+      "loss": 2.9487,
+      "step": 15532
+    },
+    {
+      "epoch": 0.9555241141732284,
+      "grad_norm": 1.3964896202087402,
+      "learning_rate": 9.757664830615131e-07,
+      "loss": 2.3929,
+      "step": 15533
+    },
+    {
+      "epoch": 0.9555856299212598,
+      "grad_norm": 1.3921113014221191,
+      "learning_rate": 9.73073517098877e-07,
+      "loss": 2.4203,
+      "step": 15534
+    },
+    {
+      "epoch": 0.9556471456692913,
+      "grad_norm": 1.3683496713638306,
+      "learning_rate": 9.703842541911145e-07,
+      "loss": 2.4657,
+      "step": 15535
+    },
+    {
+      "epoch": 0.9557086614173228,
+      "grad_norm": 1.2647162675857544,
+      "learning_rate": 9.676986944387567e-07,
+      "loss": 2.6629,
+      "step": 15536
+    },
+    {
+      "epoch": 0.9557701771653543,
+      "grad_norm": 1.2612446546554565,
+      "learning_rate": 9.650168379422675e-07,
+      "loss": 2.2434,
+      "step": 15537
+    },
+    {
+      "epoch": 0.9558316929133859,
+      "grad_norm": 1.607458233833313,
+      "learning_rate": 9.62338684801911e-07,
+      "loss": 2.7242,
+      "step": 15538
+    },
+    {
+      "epoch": 0.9558932086614174,
+      "grad_norm": 1.6731505393981934,
+      "learning_rate": 9.596642351178298e-07,
+      "loss": 1.1475,
+      "step": 15539
+    },
+    {
+      "epoch": 0.9559547244094488,
+      "grad_norm": 1.2311077117919922,
+      "learning_rate": 9.569934889900434e-07,
+      "loss": 2.7382,
+      "step": 15540
+    },
+    {
+      "epoch": 0.9560162401574803,
+      "grad_norm": 1.4108431339263916,
+      "learning_rate": 9.543264465184277e-07,
+      "loss": 2.3729,
+      "step": 15541
+    },
+    {
+      "epoch": 0.9560777559055118,
+      "grad_norm": 1.4876067638397217,
+      "learning_rate": 9.516631078027028e-07,
+      "loss": 2.9347,
+      "step": 15542
+    },
+    {
+      "epoch": 0.9561392716535433,
+      "grad_norm": 1.1523449420928955,
+      "learning_rate": 9.490034729424779e-07,
+      "loss": 2.542,
+      "step": 15543
+    },
+    {
+      "epoch": 0.9562007874015748,
+      "grad_norm": 1.2194024324417114,
+      "learning_rate": 9.463475420371959e-07,
+      "loss": 2.5184,
+      "step": 15544
+    },
+    {
+      "epoch": 0.9562623031496063,
+      "grad_norm": 1.199638843536377,
+      "learning_rate": 9.436953151861661e-07,
+      "loss": 2.3546,
+      "step": 15545
+    },
+    {
+      "epoch": 0.9563238188976378,
+      "grad_norm": 1.2757607698440552,
+      "learning_rate": 9.410467924885868e-07,
+      "loss": 2.3617,
+      "step": 15546
+    },
+    {
+      "epoch": 0.9563853346456693,
+      "grad_norm": 1.2878806591033936,
+      "learning_rate": 9.384019740435012e-07,
+      "loss": 2.3521,
+      "step": 15547
+    },
+    {
+      "epoch": 0.9564468503937008,
+      "grad_norm": 1.4266314506530762,
+      "learning_rate": 9.357608599497858e-07,
+      "loss": 2.4162,
+      "step": 15548
+    },
+    {
+      "epoch": 0.9565083661417323,
+      "grad_norm": 1.2723851203918457,
+      "learning_rate": 9.33123450306217e-07,
+      "loss": 2.5116,
+      "step": 15549
+    },
+    {
+      "epoch": 0.9565698818897638,
+      "grad_norm": 1.384329915046692,
+      "learning_rate": 9.304897452114269e-07,
+      "loss": 2.5923,
+      "step": 15550
+    },
+    {
+      "epoch": 0.9566313976377953,
+      "grad_norm": 1.2373137474060059,
+      "learning_rate": 9.278597447638814e-07,
+      "loss": 2.7368,
+      "step": 15551
+    },
+    {
+      "epoch": 0.9566929133858267,
+      "grad_norm": 1.3513286113739014,
+      "learning_rate": 9.252334490619463e-07,
+      "loss": 2.1108,
+      "step": 15552
+    },
+    {
+      "epoch": 0.9567544291338582,
+      "grad_norm": 1.2935590744018555,
+      "learning_rate": 9.226108582038318e-07,
+      "loss": 2.5932,
+      "step": 15553
+    },
+    {
+      "epoch": 0.9568159448818898,
+      "grad_norm": 1.3380405902862549,
+      "learning_rate": 9.199919722875928e-07,
+      "loss": 2.6572,
+      "step": 15554
+    },
+    {
+      "epoch": 0.9568774606299213,
+      "grad_norm": 2.143334150314331,
+      "learning_rate": 9.173767914111731e-07,
+      "loss": 2.0628,
+      "step": 15555
+    },
+    {
+      "epoch": 0.9569389763779528,
+      "grad_norm": 1.6413837671279907,
+      "learning_rate": 9.147653156723613e-07,
+      "loss": 2.8306,
+      "step": 15556
+    },
+    {
+      "epoch": 0.9570004921259843,
+      "grad_norm": 1.6470887660980225,
+      "learning_rate": 9.121575451688125e-07,
+      "loss": 2.3789,
+      "step": 15557
+    },
+    {
+      "epoch": 0.9570620078740157,
+      "grad_norm": 1.2643756866455078,
+      "learning_rate": 9.095534799980376e-07,
+      "loss": 2.8437,
+      "step": 15558
+    },
+    {
+      "epoch": 0.9571235236220472,
+      "grad_norm": 1.3260769844055176,
+      "learning_rate": 9.069531202574144e-07,
+      "loss": 2.9814,
+      "step": 15559
+    },
+    {
+      "epoch": 0.9571850393700787,
+      "grad_norm": 1.2577069997787476,
+      "learning_rate": 9.043564660441983e-07,
+      "loss": 1.8499,
+      "step": 15560
+    },
+    {
+      "epoch": 0.9572465551181102,
+      "grad_norm": 1.4543627500534058,
+      "learning_rate": 9.017635174554673e-07,
+      "loss": 2.2651,
+      "step": 15561
+    },
+    {
+      "epoch": 0.9573080708661418,
+      "grad_norm": 1.3284239768981934,
+      "learning_rate": 8.991742745881992e-07,
+      "loss": 2.7151,
+      "step": 15562
+    },
+    {
+      "epoch": 0.9573695866141733,
+      "grad_norm": 1.170554757118225,
+      "learning_rate": 8.965887375392168e-07,
+      "loss": 1.8702,
+      "step": 15563
+    },
+    {
+      "epoch": 0.9574311023622047,
+      "grad_norm": 1.1663365364074707,
+      "learning_rate": 8.94006906405187e-07,
+      "loss": 2.4665,
+      "step": 15564
+    },
+    {
+      "epoch": 0.9574926181102362,
+      "grad_norm": 1.6475023031234741,
+      "learning_rate": 8.914287812826772e-07,
+      "loss": 1.8803,
+      "step": 15565
+    },
+    {
+      "epoch": 0.9575541338582677,
+      "grad_norm": 1.275475263595581,
+      "learning_rate": 8.888543622680878e-07,
+      "loss": 2.579,
+      "step": 15566
+    },
+    {
+      "epoch": 0.9576156496062992,
+      "grad_norm": 1.1459749937057495,
+      "learning_rate": 8.862836494576865e-07,
+      "loss": 2.1453,
+      "step": 15567
+    },
+    {
+      "epoch": 0.9576771653543307,
+      "grad_norm": 1.1838659048080444,
+      "learning_rate": 8.837166429476073e-07,
+      "loss": 3.1104,
+      "step": 15568
+    },
+    {
+      "epoch": 0.9577386811023622,
+      "grad_norm": 1.2422442436218262,
+      "learning_rate": 8.811533428338293e-07,
+      "loss": 2.5568,
+      "step": 15569
+    },
+    {
+      "epoch": 0.9578001968503937,
+      "grad_norm": 1.3349069356918335,
+      "learning_rate": 8.785937492122087e-07,
+      "loss": 2.1549,
+      "step": 15570
+    },
+    {
+      "epoch": 0.9578617125984252,
+      "grad_norm": 1.3229939937591553,
+      "learning_rate": 8.760378621784804e-07,
+      "loss": 2.4934,
+      "step": 15571
+    },
+    {
+      "epoch": 0.9579232283464567,
+      "grad_norm": 1.3982917070388794,
+      "learning_rate": 8.734856818282011e-07,
+      "loss": 1.9233,
+      "step": 15572
+    },
+    {
+      "epoch": 0.9579847440944882,
+      "grad_norm": 0.7013957500457764,
+      "learning_rate": 8.709372082568056e-07,
+      "loss": 0.3184,
+      "step": 15573
+    },
+    {
+      "epoch": 0.9580462598425197,
+      "grad_norm": 1.1449782848358154,
+      "learning_rate": 8.683924415596067e-07,
+      "loss": 2.3359,
+      "step": 15574
+    },
+    {
+      "epoch": 0.9581077755905512,
+      "grad_norm": 1.337902545928955,
+      "learning_rate": 8.658513818317615e-07,
+      "loss": 2.1348,
+      "step": 15575
+    },
+    {
+      "epoch": 0.9581692913385826,
+      "grad_norm": 1.483574628829956,
+      "learning_rate": 8.633140291682828e-07,
+      "loss": 2.32,
+      "step": 15576
+    },
+    {
+      "epoch": 0.9582308070866141,
+      "grad_norm": 1.2396143674850464,
+      "learning_rate": 8.607803836640504e-07,
+      "loss": 2.1044,
+      "step": 15577
+    },
+    {
+      "epoch": 0.9582923228346457,
+      "grad_norm": 1.2322349548339844,
+      "learning_rate": 8.582504454138218e-07,
+      "loss": 3.0118,
+      "step": 15578
+    },
+    {
+      "epoch": 0.9583538385826772,
+      "grad_norm": 1.2137715816497803,
+      "learning_rate": 8.557242145121991e-07,
+      "loss": 2.4545,
+      "step": 15579
+    },
+    {
+      "epoch": 0.9584153543307087,
+      "grad_norm": 1.1854603290557861,
+      "learning_rate": 8.532016910536511e-07,
+      "loss": 2.0502,
+      "step": 15580
+    },
+    {
+      "epoch": 0.9584768700787402,
+      "grad_norm": 1.4001988172531128,
+      "learning_rate": 8.506828751324913e-07,
+      "loss": 2.4111,
+      "step": 15581
+    },
+    {
+      "epoch": 0.9585383858267716,
+      "grad_norm": 1.2837283611297607,
+      "learning_rate": 8.481677668429334e-07,
+      "loss": 2.1056,
+      "step": 15582
+    },
+    {
+      "epoch": 0.9585999015748031,
+      "grad_norm": 1.3740195035934448,
+      "learning_rate": 8.456563662790129e-07,
+      "loss": 1.4621,
+      "step": 15583
+    },
+    {
+      "epoch": 0.9586614173228346,
+      "grad_norm": 1.2298839092254639,
+      "learning_rate": 8.431486735346328e-07,
+      "loss": 2.6669,
+      "step": 15584
+    },
+    {
+      "epoch": 0.9587229330708661,
+      "grad_norm": 1.1437709331512451,
+      "learning_rate": 8.406446887035846e-07,
+      "loss": 2.7532,
+      "step": 15585
+    },
+    {
+      "epoch": 0.9587844488188977,
+      "grad_norm": 1.4593154191970825,
+      "learning_rate": 8.381444118795045e-07,
+      "loss": 2.5307,
+      "step": 15586
+    },
+    {
+      "epoch": 0.9588459645669292,
+      "grad_norm": 1.219886302947998,
+      "learning_rate": 8.356478431558734e-07,
+      "loss": 2.7508,
+      "step": 15587
+    },
+    {
+      "epoch": 0.9589074803149606,
+      "grad_norm": 1.4509294033050537,
+      "learning_rate": 8.33154982626061e-07,
+      "loss": 2.3196,
+      "step": 15588
+    },
+    {
+      "epoch": 0.9589689960629921,
+      "grad_norm": 1.3187768459320068,
+      "learning_rate": 8.306658303832815e-07,
+      "loss": 2.5781,
+      "step": 15589
+    },
+    {
+      "epoch": 0.9590305118110236,
+      "grad_norm": 1.225074052810669,
+      "learning_rate": 8.281803865206162e-07,
+      "loss": 1.1577,
+      "step": 15590
+    },
+    {
+      "epoch": 0.9590920275590551,
+      "grad_norm": 1.234833836555481,
+      "learning_rate": 8.256986511310017e-07,
+      "loss": 2.3606,
+      "step": 15591
+    },
+    {
+      "epoch": 0.9591535433070866,
+      "grad_norm": 1.2626068592071533,
+      "learning_rate": 8.232206243072527e-07,
+      "loss": 3.2291,
+      "step": 15592
+    },
+    {
+      "epoch": 0.9592150590551181,
+      "grad_norm": 1.323028802871704,
+      "learning_rate": 8.207463061420285e-07,
+      "loss": 1.8824,
+      "step": 15593
+    },
+    {
+      "epoch": 0.9592765748031497,
+      "grad_norm": 1.3446701765060425,
+      "learning_rate": 8.182756967278549e-07,
+      "loss": 3.0223,
+      "step": 15594
+    },
+    {
+      "epoch": 0.9593380905511811,
+      "grad_norm": 1.5662436485290527,
+      "learning_rate": 8.158087961571137e-07,
+      "loss": 1.4564,
+      "step": 15595
+    },
+    {
+      "epoch": 0.9593996062992126,
+      "grad_norm": 1.1178752183914185,
+      "learning_rate": 8.133456045220534e-07,
+      "loss": 1.3354,
+      "step": 15596
+    },
+    {
+      "epoch": 0.9594611220472441,
+      "grad_norm": 1.2154089212417603,
+      "learning_rate": 8.108861219148e-07,
+      "loss": 2.6835,
+      "step": 15597
+    },
+    {
+      "epoch": 0.9595226377952756,
+      "grad_norm": 1.5087462663650513,
+      "learning_rate": 8.084303484273025e-07,
+      "loss": 1.7705,
+      "step": 15598
+    },
+    {
+      "epoch": 0.9595841535433071,
+      "grad_norm": 1.4349182844161987,
+      "learning_rate": 8.059782841514097e-07,
+      "loss": 2.2648,
+      "step": 15599
+    },
+    {
+      "epoch": 0.9596456692913385,
+      "grad_norm": 1.4489890336990356,
+      "learning_rate": 8.035299291788034e-07,
+      "loss": 2.2894,
+      "step": 15600
+    },
+    {
+      "epoch": 0.95970718503937,
+      "grad_norm": 1.3545798063278198,
+      "learning_rate": 8.01085283601044e-07,
+      "loss": 2.3849,
+      "step": 15601
+    },
+    {
+      "epoch": 0.9597687007874016,
+      "grad_norm": 1.2737141847610474,
+      "learning_rate": 7.986443475095473e-07,
+      "loss": 2.5355,
+      "step": 15602
+    },
+    {
+      "epoch": 0.9598302165354331,
+      "grad_norm": 1.2782533168792725,
+      "learning_rate": 7.962071209955846e-07,
+      "loss": 1.8376,
+      "step": 15603
+    },
+    {
+      "epoch": 0.9598917322834646,
+      "grad_norm": 1.1650242805480957,
+      "learning_rate": 7.937736041502941e-07,
+      "loss": 2.602,
+      "step": 15604
+    },
+    {
+      "epoch": 0.9599532480314961,
+      "grad_norm": 1.4927664995193481,
+      "learning_rate": 7.91343797064703e-07,
+      "loss": 2.749,
+      "step": 15605
+    },
+    {
+      "epoch": 0.9600147637795275,
+      "grad_norm": 1.318527102470398,
+      "learning_rate": 7.889176998296388e-07,
+      "loss": 2.1829,
+      "step": 15606
+    },
+    {
+      "epoch": 0.960076279527559,
+      "grad_norm": 1.4417206048965454,
+      "learning_rate": 7.864953125358399e-07,
+      "loss": 2.0962,
+      "step": 15607
+    },
+    {
+      "epoch": 0.9601377952755905,
+      "grad_norm": 1.6041160821914673,
+      "learning_rate": 7.840766352738782e-07,
+      "loss": 2.992,
+      "step": 15608
+    },
+    {
+      "epoch": 0.960199311023622,
+      "grad_norm": 1.4392280578613281,
+      "learning_rate": 7.816616681342148e-07,
+      "loss": 2.2949,
+      "step": 15609
+    },
+    {
+      "epoch": 0.9602608267716536,
+      "grad_norm": 1.4053643941879272,
+      "learning_rate": 7.792504112071441e-07,
+      "loss": 2.4271,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9603223425196851,
+      "grad_norm": 2.1080384254455566,
+      "learning_rate": 7.768428645828385e-07,
+      "loss": 2.7628,
+      "step": 15611
+    },
+    {
+      "epoch": 0.9603838582677166,
+      "grad_norm": 1.1814566850662231,
+      "learning_rate": 7.744390283513258e-07,
+      "loss": 2.7431,
+      "step": 15612
+    },
+    {
+      "epoch": 0.960445374015748,
+      "grad_norm": 1.537744402885437,
+      "learning_rate": 7.720389026025121e-07,
+      "loss": 2.5221,
+      "step": 15613
+    },
+    {
+      "epoch": 0.9605068897637795,
+      "grad_norm": 2.0069777965545654,
+      "learning_rate": 7.696424874261143e-07,
+      "loss": 2.6927,
+      "step": 15614
+    },
+    {
+      "epoch": 0.960568405511811,
+      "grad_norm": 1.7385438680648804,
+      "learning_rate": 7.67249782911772e-07,
+      "loss": 3.5506,
+      "step": 15615
+    },
+    {
+      "epoch": 0.9606299212598425,
+      "grad_norm": 1.2386784553527832,
+      "learning_rate": 7.648607891489356e-07,
+      "loss": 2.0787,
+      "step": 15616
+    },
+    {
+      "epoch": 0.960691437007874,
+      "grad_norm": 1.376741647720337,
+      "learning_rate": 7.624755062269783e-07,
+      "loss": 2.3255,
+      "step": 15617
+    },
+    {
+      "epoch": 0.9607529527559056,
+      "grad_norm": 1.2846322059631348,
+      "learning_rate": 7.600939342350622e-07,
+      "loss": 2.3387,
+      "step": 15618
+    },
+    {
+      "epoch": 0.960814468503937,
+      "grad_norm": 1.3151763677597046,
+      "learning_rate": 7.577160732622602e-07,
+      "loss": 2.0845,
+      "step": 15619
+    },
+    {
+      "epoch": 0.9608759842519685,
+      "grad_norm": 1.3139876127243042,
+      "learning_rate": 7.55341923397479e-07,
+      "loss": 1.3237,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9609375,
+      "grad_norm": 1.5021556615829468,
+      "learning_rate": 7.529714847295144e-07,
+      "loss": 2.2174,
+      "step": 15621
+    },
+    {
+      "epoch": 0.9609990157480315,
+      "grad_norm": 1.2681010961532593,
+      "learning_rate": 7.506047573469955e-07,
+      "loss": 1.9876,
+      "step": 15622
+    },
+    {
+      "epoch": 0.961060531496063,
+      "grad_norm": 1.6095095872879028,
+      "learning_rate": 7.48241741338429e-07,
+      "loss": 2.4728,
+      "step": 15623
+    },
+    {
+      "epoch": 0.9611220472440944,
+      "grad_norm": 1.2201817035675049,
+      "learning_rate": 7.45882436792178e-07,
+      "loss": 2.352,
+      "step": 15624
+    },
+    {
+      "epoch": 0.961183562992126,
+      "grad_norm": 1.3690054416656494,
+      "learning_rate": 7.435268437964716e-07,
+      "loss": 1.7824,
+      "step": 15625
+    },
+    {
+      "epoch": 0.9612450787401575,
+      "grad_norm": 1.3794076442718506,
+      "learning_rate": 7.411749624393949e-07,
+      "loss": 2.2845,
+      "step": 15626
+    },
+    {
+      "epoch": 0.961306594488189,
+      "grad_norm": 1.140856385231018,
+      "learning_rate": 7.388267928088887e-07,
+      "loss": 3.3053,
+      "step": 15627
+    },
+    {
+      "epoch": 0.9613681102362205,
+      "grad_norm": 1.279187560081482,
+      "learning_rate": 7.364823349927608e-07,
+      "loss": 2.973,
+      "step": 15628
+    },
+    {
+      "epoch": 0.961429625984252,
+      "grad_norm": 1.4426383972167969,
+      "learning_rate": 7.341415890786851e-07,
+      "loss": 2.772,
+      "step": 15629
+    },
+    {
+      "epoch": 0.9614911417322834,
+      "grad_norm": 1.1984729766845703,
+      "learning_rate": 7.318045551541919e-07,
+      "loss": 2.3179,
+      "step": 15630
+    },
+    {
+      "epoch": 0.9615526574803149,
+      "grad_norm": 1.1640381813049316,
+      "learning_rate": 7.294712333066778e-07,
+      "loss": 2.4266,
+      "step": 15631
+    },
+    {
+      "epoch": 0.9616141732283464,
+      "grad_norm": 1.5057977437973022,
+      "learning_rate": 7.271416236233952e-07,
+      "loss": 2.3414,
+      "step": 15632
+    },
+    {
+      "epoch": 0.961675688976378,
+      "grad_norm": 1.345380425453186,
+      "learning_rate": 7.248157261914523e-07,
+      "loss": 2.5725,
+      "step": 15633
+    },
+    {
+      "epoch": 0.9617372047244095,
+      "grad_norm": 1.3230210542678833,
+      "learning_rate": 7.22493541097824e-07,
+      "loss": 2.061,
+      "step": 15634
+    },
+    {
+      "epoch": 0.961798720472441,
+      "grad_norm": 1.238608479499817,
+      "learning_rate": 7.201750684293518e-07,
+      "loss": 2.2228,
+      "step": 15635
+    },
+    {
+      "epoch": 0.9618602362204725,
+      "grad_norm": 1.4114161729812622,
+      "learning_rate": 7.178603082727331e-07,
+      "loss": 2.588,
+      "step": 15636
+    },
+    {
+      "epoch": 0.9619217519685039,
+      "grad_norm": 1.3225206136703491,
+      "learning_rate": 7.155492607145209e-07,
+      "loss": 2.2606,
+      "step": 15637
+    },
+    {
+      "epoch": 0.9619832677165354,
+      "grad_norm": 1.487419843673706,
+      "learning_rate": 7.132419258411571e-07,
+      "loss": 2.833,
+      "step": 15638
+    },
+    {
+      "epoch": 0.9620447834645669,
+      "grad_norm": 1.1770795583724976,
+      "learning_rate": 7.109383037388951e-07,
+      "loss": 2.9634,
+      "step": 15639
+    },
+    {
+      "epoch": 0.9621062992125984,
+      "grad_norm": 1.2905209064483643,
+      "learning_rate": 7.086383944938879e-07,
+      "loss": 2.3592,
+      "step": 15640
+    },
+    {
+      "epoch": 0.96216781496063,
+      "grad_norm": 1.2743598222732544,
+      "learning_rate": 7.063421981921448e-07,
+      "loss": 1.8676,
+      "step": 15641
+    },
+    {
+      "epoch": 0.9622293307086615,
+      "grad_norm": 1.3640108108520508,
+      "learning_rate": 7.040497149195302e-07,
+      "loss": 2.6531,
+      "step": 15642
+    },
+    {
+      "epoch": 0.9622908464566929,
+      "grad_norm": 1.2635902166366577,
+      "learning_rate": 7.017609447617534e-07,
+      "loss": 2.1623,
+      "step": 15643
+    },
+    {
+      "epoch": 0.9623523622047244,
+      "grad_norm": 1.2792080640792847,
+      "learning_rate": 6.994758878044349e-07,
+      "loss": 2.7625,
+      "step": 15644
+    },
+    {
+      "epoch": 0.9624138779527559,
+      "grad_norm": 1.558062195777893,
+      "learning_rate": 6.971945441329842e-07,
+      "loss": 2.2389,
+      "step": 15645
+    },
+    {
+      "epoch": 0.9624753937007874,
+      "grad_norm": 1.2055001258850098,
+      "learning_rate": 6.949169138327438e-07,
+      "loss": 2.6381,
+      "step": 15646
+    },
+    {
+      "epoch": 0.9625369094488189,
+      "grad_norm": 1.2350903749465942,
+      "learning_rate": 6.92642996988857e-07,
+      "loss": 2.5247,
+      "step": 15647
+    },
+    {
+      "epoch": 0.9625984251968503,
+      "grad_norm": 1.2773486375808716,
+      "learning_rate": 6.903727936863891e-07,
+      "loss": 2.531,
+      "step": 15648
+    },
+    {
+      "epoch": 0.9626599409448819,
+      "grad_norm": 1.524214744567871,
+      "learning_rate": 6.881063040101943e-07,
+      "loss": 2.2739,
+      "step": 15649
+    },
+    {
+      "epoch": 0.9627214566929134,
+      "grad_norm": 1.1394048929214478,
+      "learning_rate": 6.858435280450603e-07,
+      "loss": 3.0072,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9627829724409449,
+      "grad_norm": 1.2380037307739258,
+      "learning_rate": 6.835844658755752e-07,
+      "loss": 2.8801,
+      "step": 15651
+    },
+    {
+      "epoch": 0.9628444881889764,
+      "grad_norm": 2.5848703384399414,
+      "learning_rate": 6.81329117586238e-07,
+      "loss": 1.5874,
+      "step": 15652
+    },
+    {
+      "epoch": 0.9629060039370079,
+      "grad_norm": 1.127270221710205,
+      "learning_rate": 6.7907748326137e-07,
+      "loss": 1.8928,
+      "step": 15653
+    },
+    {
+      "epoch": 0.9629675196850394,
+      "grad_norm": 1.3350677490234375,
+      "learning_rate": 6.768295629851707e-07,
+      "loss": 2.8241,
+      "step": 15654
+    },
+    {
+      "epoch": 0.9630290354330708,
+      "grad_norm": 1.359440565109253,
+      "learning_rate": 6.74585356841706e-07,
+      "loss": 2.2809,
+      "step": 15655
+    },
+    {
+      "epoch": 0.9630905511811023,
+      "grad_norm": 1.271693229675293,
+      "learning_rate": 6.723448649149089e-07,
+      "loss": 2.13,
+      "step": 15656
+    },
+    {
+      "epoch": 0.9631520669291339,
+      "grad_norm": 1.279301404953003,
+      "learning_rate": 6.701080872885346e-07,
+      "loss": 2.4623,
+      "step": 15657
+    },
+    {
+      "epoch": 0.9632135826771654,
+      "grad_norm": 1.5274990797042847,
+      "learning_rate": 6.678750240462384e-07,
+      "loss": 1.9425,
+      "step": 15658
+    },
+    {
+      "epoch": 0.9632750984251969,
+      "grad_norm": 1.1446696519851685,
+      "learning_rate": 6.656456752715313e-07,
+      "loss": 1.7426,
+      "step": 15659
+    },
+    {
+      "epoch": 0.9633366141732284,
+      "grad_norm": 1.1155427694320679,
+      "learning_rate": 6.634200410477687e-07,
+      "loss": 2.6929,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9633981299212598,
+      "grad_norm": 1.5808273553848267,
+      "learning_rate": 6.61198121458173e-07,
+      "loss": 2.5495,
+      "step": 15661
+    },
+    {
+      "epoch": 0.9634596456692913,
+      "grad_norm": 1.3090065717697144,
+      "learning_rate": 6.589799165858557e-07,
+      "loss": 2.6369,
+      "step": 15662
+    },
+    {
+      "epoch": 0.9635211614173228,
+      "grad_norm": 1.172630786895752,
+      "learning_rate": 6.567654265137391e-07,
+      "loss": 2.6853,
+      "step": 15663
+    },
+    {
+      "epoch": 0.9635826771653543,
+      "grad_norm": 1.5915862321853638,
+      "learning_rate": 6.545546513246459e-07,
+      "loss": 2.3617,
+      "step": 15664
+    },
+    {
+      "epoch": 0.9636441929133859,
+      "grad_norm": 1.2860027551651,
+      "learning_rate": 6.523475911012433e-07,
+      "loss": 2.0572,
+      "step": 15665
+    },
+    {
+      "epoch": 0.9637057086614174,
+      "grad_norm": 1.387419581413269,
+      "learning_rate": 6.501442459260654e-07,
+      "loss": 2.2409,
+      "step": 15666
+    },
+    {
+      "epoch": 0.9637672244094488,
+      "grad_norm": 1.3963629007339478,
+      "learning_rate": 6.479446158814906e-07,
+      "loss": 2.2902,
+      "step": 15667
+    },
+    {
+      "epoch": 0.9638287401574803,
+      "grad_norm": 1.3884119987487793,
+      "learning_rate": 6.457487010497976e-07,
+      "loss": 2.2404,
+      "step": 15668
+    },
+    {
+      "epoch": 0.9638902559055118,
+      "grad_norm": 1.1980581283569336,
+      "learning_rate": 6.435565015130873e-07,
+      "loss": 2.5959,
+      "step": 15669
+    },
+    {
+      "epoch": 0.9639517716535433,
+      "grad_norm": 1.2948195934295654,
+      "learning_rate": 6.413680173533388e-07,
+      "loss": 3.198,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9640132874015748,
+      "grad_norm": 1.312346339225769,
+      "learning_rate": 6.391832486523863e-07,
+      "loss": 2.3529,
+      "step": 15671
+    },
+    {
+      "epoch": 0.9640748031496063,
+      "grad_norm": 1.4699727296829224,
+      "learning_rate": 6.370021954919203e-07,
+      "loss": 2.5575,
+      "step": 15672
+    },
+    {
+      "epoch": 0.9641363188976378,
+      "grad_norm": 1.503238320350647,
+      "learning_rate": 6.348248579535088e-07,
+      "loss": 2.2402,
+      "step": 15673
+    },
+    {
+      "epoch": 0.9641978346456693,
+      "grad_norm": 1.3571182489395142,
+      "learning_rate": 6.326512361185754e-07,
+      "loss": 2.9087,
+      "step": 15674
+    },
+    {
+      "epoch": 0.9642593503937008,
+      "grad_norm": 1.2743169069290161,
+      "learning_rate": 6.304813300684109e-07,
+      "loss": 2.7425,
+      "step": 15675
+    },
+    {
+      "epoch": 0.9643208661417323,
+      "grad_norm": 1.2106361389160156,
+      "learning_rate": 6.283151398841281e-07,
+      "loss": 2.9711,
+      "step": 15676
+    },
+    {
+      "epoch": 0.9643823818897638,
+      "grad_norm": 1.4526644945144653,
+      "learning_rate": 6.26152665646762e-07,
+      "loss": 2.988,
+      "step": 15677
+    },
+    {
+      "epoch": 0.9644438976377953,
+      "grad_norm": 1.3470962047576904,
+      "learning_rate": 6.239939074371481e-07,
+      "loss": 2.9781,
+      "step": 15678
+    },
+    {
+      "epoch": 0.9645054133858267,
+      "grad_norm": 1.4137965440750122,
+      "learning_rate": 6.21838865336033e-07,
+      "loss": 2.6907,
+      "step": 15679
+    },
+    {
+      "epoch": 0.9645669291338582,
+      "grad_norm": 1.431391954421997,
+      "learning_rate": 6.196875394239965e-07,
+      "loss": 2.4222,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9646284448818898,
+      "grad_norm": 1.191960096359253,
+      "learning_rate": 6.175399297814854e-07,
+      "loss": 2.2892,
+      "step": 15681
+    },
+    {
+      "epoch": 0.9646899606299213,
+      "grad_norm": 1.3796143531799316,
+      "learning_rate": 6.153960364888134e-07,
+      "loss": 2.0681,
+      "step": 15682
+    },
+    {
+      "epoch": 0.9647514763779528,
+      "grad_norm": 1.5141513347625732,
+      "learning_rate": 6.132558596261384e-07,
+      "loss": 2.5305,
+      "step": 15683
+    },
+    {
+      "epoch": 0.9648129921259843,
+      "grad_norm": 1.156854271888733,
+      "learning_rate": 6.111193992735076e-07,
+      "loss": 1.5899,
+      "step": 15684
+    },
+    {
+      "epoch": 0.9648745078740157,
+      "grad_norm": 1.2190375328063965,
+      "learning_rate": 6.089866555107904e-07,
+      "loss": 2.4831,
+      "step": 15685
+    },
+    {
+      "epoch": 0.9649360236220472,
+      "grad_norm": 1.2105406522750854,
+      "learning_rate": 6.068576284177563e-07,
+      "loss": 2.8443,
+      "step": 15686
+    },
+    {
+      "epoch": 0.9649975393700787,
+      "grad_norm": 1.4682172536849976,
+      "learning_rate": 6.047323180740305e-07,
+      "loss": 2.4424,
+      "step": 15687
+    },
+    {
+      "epoch": 0.9650590551181102,
+      "grad_norm": 1.1875646114349365,
+      "learning_rate": 6.026107245590607e-07,
+      "loss": 2.972,
+      "step": 15688
+    },
+    {
+      "epoch": 0.9651205708661418,
+      "grad_norm": 1.2409573793411255,
+      "learning_rate": 6.004928479521943e-07,
+      "loss": 2.9373,
+      "step": 15689
+    },
+    {
+      "epoch": 0.9651820866141733,
+      "grad_norm": 1.250030279159546,
+      "learning_rate": 5.983786883326348e-07,
+      "loss": 2.6767,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9652436023622047,
+      "grad_norm": 1.2314567565917969,
+      "learning_rate": 5.9626824577943e-07,
+      "loss": 2.8936,
+      "step": 15691
+    },
+    {
+      "epoch": 0.9653051181102362,
+      "grad_norm": 1.5724626779556274,
+      "learning_rate": 5.941615203714945e-07,
+      "loss": 1.9926,
+      "step": 15692
+    },
+    {
+      "epoch": 0.9653666338582677,
+      "grad_norm": 1.2996635437011719,
+      "learning_rate": 5.920585121876321e-07,
+      "loss": 2.181,
+      "step": 15693
+    },
+    {
+      "epoch": 0.9654281496062992,
+      "grad_norm": 1.4124492406845093,
+      "learning_rate": 5.899592213064575e-07,
+      "loss": 3.4194,
+      "step": 15694
+    },
+    {
+      "epoch": 0.9654896653543307,
+      "grad_norm": 1.3270601034164429,
+      "learning_rate": 5.878636478064859e-07,
+      "loss": 3.01,
+      "step": 15695
+    },
+    {
+      "epoch": 0.9655511811023622,
+      "grad_norm": 1.3321137428283691,
+      "learning_rate": 5.857717917660654e-07,
+      "loss": 2.331,
+      "step": 15696
+    },
+    {
+      "epoch": 0.9656126968503937,
+      "grad_norm": 1.1634430885314941,
+      "learning_rate": 5.836836532634448e-07,
+      "loss": 1.9742,
+      "step": 15697
+    },
+    {
+      "epoch": 0.9656742125984252,
+      "grad_norm": 1.161369800567627,
+      "learning_rate": 5.81599232376695e-07,
+      "loss": 2.4063,
+      "step": 15698
+    },
+    {
+      "epoch": 0.9657357283464567,
+      "grad_norm": 1.2901676893234253,
+      "learning_rate": 5.795185291837535e-07,
+      "loss": 2.5177,
+      "step": 15699
+    },
+    {
+      "epoch": 0.9657972440944882,
+      "grad_norm": 1.3641135692596436,
+      "learning_rate": 5.774415437624358e-07,
+      "loss": 2.8969,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9658587598425197,
+      "grad_norm": 1.277353048324585,
+      "learning_rate": 5.753682761904134e-07,
+      "loss": 2.6935,
+      "step": 15701
+    },
+    {
+      "epoch": 0.9659202755905512,
+      "grad_norm": 1.5561186075210571,
+      "learning_rate": 5.732987265452017e-07,
+      "loss": 2.0555,
+      "step": 15702
+    },
+    {
+      "epoch": 0.9659817913385826,
+      "grad_norm": 1.3260079622268677,
+      "learning_rate": 5.712328949042167e-07,
+      "loss": 2.1819,
+      "step": 15703
+    },
+    {
+      "epoch": 0.9660433070866141,
+      "grad_norm": 5.488504409790039,
+      "learning_rate": 5.691707813446745e-07,
+      "loss": 2.7608,
+      "step": 15704
+    },
+    {
+      "epoch": 0.9661048228346457,
+      "grad_norm": 1.352052092552185,
+      "learning_rate": 5.671123859437133e-07,
+      "loss": 2.3233,
+      "step": 15705
+    },
+    {
+      "epoch": 0.9661663385826772,
+      "grad_norm": 1.6265449523925781,
+      "learning_rate": 5.650577087782938e-07,
+      "loss": 2.615,
+      "step": 15706
+    },
+    {
+      "epoch": 0.9662278543307087,
+      "grad_norm": 1.5247191190719604,
+      "learning_rate": 5.630067499252434e-07,
+      "loss": 1.8737,
+      "step": 15707
+    },
+    {
+      "epoch": 0.9662893700787402,
+      "grad_norm": 1.2846484184265137,
+      "learning_rate": 5.609595094612786e-07,
+      "loss": 2.0447,
+      "step": 15708
+    },
+    {
+      "epoch": 0.9663508858267716,
+      "grad_norm": 1.096246600151062,
+      "learning_rate": 5.58915987462938e-07,
+      "loss": 1.9696,
+      "step": 15709
+    },
+    {
+      "epoch": 0.9664124015748031,
+      "grad_norm": 1.3244662284851074,
+      "learning_rate": 5.568761840066272e-07,
+      "loss": 2.7966,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9664739173228346,
+      "grad_norm": 1.1828210353851318,
+      "learning_rate": 5.548400991686519e-07,
+      "loss": 2.5721,
+      "step": 15711
+    },
+    {
+      "epoch": 0.9665354330708661,
+      "grad_norm": 1.281810998916626,
+      "learning_rate": 5.528077330251292e-07,
+      "loss": 2.2965,
+      "step": 15712
+    },
+    {
+      "epoch": 0.9665969488188977,
+      "grad_norm": 2.880530595779419,
+      "learning_rate": 5.507790856520645e-07,
+      "loss": 2.7606,
+      "step": 15713
+    },
+    {
+      "epoch": 0.9666584645669292,
+      "grad_norm": 1.2327064275741577,
+      "learning_rate": 5.487541571253197e-07,
+      "loss": 3.2352,
+      "step": 15714
+    },
+    {
+      "epoch": 0.9667199803149606,
+      "grad_norm": 1.1719205379486084,
+      "learning_rate": 5.467329475206117e-07,
+      "loss": 2.4248,
+      "step": 15715
+    },
+    {
+      "epoch": 0.9667814960629921,
+      "grad_norm": 1.3731155395507812,
+      "learning_rate": 5.447154569135359e-07,
+      "loss": 2.149,
+      "step": 15716
+    },
+    {
+      "epoch": 0.9668430118110236,
+      "grad_norm": 1.2671878337860107,
+      "learning_rate": 5.427016853795097e-07,
+      "loss": 2.4598,
+      "step": 15717
+    },
+    {
+      "epoch": 0.9669045275590551,
+      "grad_norm": 1.8511446714401245,
+      "learning_rate": 5.406916329938505e-07,
+      "loss": 2.3457,
+      "step": 15718
+    },
+    {
+      "epoch": 0.9669660433070866,
+      "grad_norm": 1.242748737335205,
+      "learning_rate": 5.386852998317316e-07,
+      "loss": 1.9765,
+      "step": 15719
+    },
+    {
+      "epoch": 0.9670275590551181,
+      "grad_norm": 1.3449674844741821,
+      "learning_rate": 5.366826859681706e-07,
+      "loss": 1.906,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9670890748031497,
+      "grad_norm": 1.1767064332962036,
+      "learning_rate": 5.346837914780523e-07,
+      "loss": 2.6128,
+      "step": 15721
+    },
+    {
+      "epoch": 0.9671505905511811,
+      "grad_norm": 1.4295063018798828,
+      "learning_rate": 5.326886164361277e-07,
+      "loss": 2.7473,
+      "step": 15722
+    },
+    {
+      "epoch": 0.9672121062992126,
+      "grad_norm": 1.4322341680526733,
+      "learning_rate": 5.306971609170041e-07,
+      "loss": 2.1012,
+      "step": 15723
+    },
+    {
+      "epoch": 0.9672736220472441,
+      "grad_norm": 1.2690966129302979,
+      "learning_rate": 5.287094249951441e-07,
+      "loss": 2.5482,
+      "step": 15724
+    },
+    {
+      "epoch": 0.9673351377952756,
+      "grad_norm": 2.710144519805908,
+      "learning_rate": 5.267254087448881e-07,
+      "loss": 2.849,
+      "step": 15725
+    },
+    {
+      "epoch": 0.9673966535433071,
+      "grad_norm": 1.7360481023788452,
+      "learning_rate": 5.247451122404323e-07,
+      "loss": 2.109,
+      "step": 15726
+    },
+    {
+      "epoch": 0.9674581692913385,
+      "grad_norm": 1.2432267665863037,
+      "learning_rate": 5.227685355558176e-07,
+      "loss": 1.8234,
+      "step": 15727
+    },
+    {
+      "epoch": 0.96751968503937,
+      "grad_norm": 1.3344964981079102,
+      "learning_rate": 5.207956787649515e-07,
+      "loss": 1.7037,
+      "step": 15728
+    },
+    {
+      "epoch": 0.9675812007874016,
+      "grad_norm": 1.4470767974853516,
+      "learning_rate": 5.188265419416194e-07,
+      "loss": 2.0027,
+      "step": 15729
+    },
+    {
+      "epoch": 0.9676427165354331,
+      "grad_norm": 1.237662434577942,
+      "learning_rate": 5.168611251594513e-07,
+      "loss": 2.8439,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9677042322834646,
+      "grad_norm": 1.2874306440353394,
+      "learning_rate": 5.148994284919439e-07,
+      "loss": 2.2792,
+      "step": 15731
+    },
+    {
+      "epoch": 0.9677657480314961,
+      "grad_norm": 1.3013176918029785,
+      "learning_rate": 5.129414520124609e-07,
+      "loss": 2.592,
+      "step": 15732
+    },
+    {
+      "epoch": 0.9678272637795275,
+      "grad_norm": 1.264587163925171,
+      "learning_rate": 5.109871957941991e-07,
+      "loss": 2.1931,
+      "step": 15733
+    },
+    {
+      "epoch": 0.967888779527559,
+      "grad_norm": 1.2288742065429688,
+      "learning_rate": 5.090366599102669e-07,
+      "loss": 2.5994,
+      "step": 15734
+    },
+    {
+      "epoch": 0.9679502952755905,
+      "grad_norm": 1.304774522781372,
+      "learning_rate": 5.070898444335725e-07,
+      "loss": 2.6478,
+      "step": 15735
+    },
+    {
+      "epoch": 0.968011811023622,
+      "grad_norm": 1.6302281618118286,
+      "learning_rate": 5.051467494369355e-07,
+      "loss": 2.4167,
+      "step": 15736
+    },
+    {
+      "epoch": 0.9680733267716536,
+      "grad_norm": 1.690407156944275,
+      "learning_rate": 5.032073749930089e-07,
+      "loss": 1.9666,
+      "step": 15737
+    },
+    {
+      "epoch": 0.9681348425196851,
+      "grad_norm": 1.3365520238876343,
+      "learning_rate": 5.012717211743234e-07,
+      "loss": 2.471,
+      "step": 15738
+    },
+    {
+      "epoch": 0.9681963582677166,
+      "grad_norm": 1.4256707429885864,
+      "learning_rate": 4.993397880532547e-07,
+      "loss": 2.6046,
+      "step": 15739
+    },
+    {
+      "epoch": 0.968257874015748,
+      "grad_norm": 1.1148436069488525,
+      "learning_rate": 4.974115757020448e-07,
+      "loss": 2.7848,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9683193897637795,
+      "grad_norm": 1.4302046298980713,
+      "learning_rate": 4.954870841927917e-07,
+      "loss": 2.4139,
+      "step": 15741
+    },
+    {
+      "epoch": 0.968380905511811,
+      "grad_norm": 1.238240122795105,
+      "learning_rate": 4.935663135974822e-07,
+      "loss": 2.4925,
+      "step": 15742
+    },
+    {
+      "epoch": 0.9684424212598425,
+      "grad_norm": 1.330694556236267,
+      "learning_rate": 4.916492639879255e-07,
+      "loss": 0.9629,
+      "step": 15743
+    },
+    {
+      "epoch": 0.968503937007874,
+      "grad_norm": 1.2735792398452759,
+      "learning_rate": 4.897359354358089e-07,
+      "loss": 2.626,
+      "step": 15744
+    },
+    {
+      "epoch": 0.9685654527559056,
+      "grad_norm": 1.4602596759796143,
+      "learning_rate": 4.87826328012686e-07,
+      "loss": 2.6249,
+      "step": 15745
+    },
+    {
+      "epoch": 0.968626968503937,
+      "grad_norm": 1.487383484840393,
+      "learning_rate": 4.859204417899665e-07,
+      "loss": 2.5862,
+      "step": 15746
+    },
+    {
+      "epoch": 0.9686884842519685,
+      "grad_norm": 1.0613491535186768,
+      "learning_rate": 4.840182768389157e-07,
+      "loss": 2.4512,
+      "step": 15747
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 12.466569900512695,
+      "learning_rate": 4.821198332306653e-07,
+      "loss": 1.7956,
+      "step": 15748
+    },
+    {
+      "epoch": 0.9688115157480315,
+      "grad_norm": 1.379395842552185,
+      "learning_rate": 4.802251110362033e-07,
+      "loss": 2.3091,
+      "step": 15749
+    },
+    {
+      "epoch": 0.968873031496063,
+      "grad_norm": 1.1642518043518066,
+      "learning_rate": 4.783341103263838e-07,
+      "loss": 2.7364,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9689345472440944,
+      "grad_norm": 1.2093671560287476,
+      "learning_rate": 4.7644683117191725e-07,
+      "loss": 2.7969,
+      "step": 15751
+    },
+    {
+      "epoch": 0.968996062992126,
+      "grad_norm": 1.4263415336608887,
+      "learning_rate": 4.7456327364339135e-07,
+      "loss": 2.5751,
+      "step": 15752
+    },
+    {
+      "epoch": 0.9690575787401575,
+      "grad_norm": 1.2373050451278687,
+      "learning_rate": 4.726834378112277e-07,
+      "loss": 2.5051,
+      "step": 15753
+    },
+    {
+      "epoch": 0.969119094488189,
+      "grad_norm": 1.188022255897522,
+      "learning_rate": 4.7080732374571446e-07,
+      "loss": 2.6022,
+      "step": 15754
+    },
+    {
+      "epoch": 0.9691806102362205,
+      "grad_norm": 1.3111598491668701,
+      "learning_rate": 4.6893493151702886e-07,
+      "loss": 2.6213,
+      "step": 15755
+    },
+    {
+      "epoch": 0.969242125984252,
+      "grad_norm": 1.550711750984192,
+      "learning_rate": 4.6706626119517037e-07,
+      "loss": 2.93,
+      "step": 15756
+    },
+    {
+      "epoch": 0.9693036417322834,
+      "grad_norm": 1.1857855319976807,
+      "learning_rate": 4.6520131285001657e-07,
+      "loss": 1.8692,
+      "step": 15757
+    },
+    {
+      "epoch": 0.9693651574803149,
+      "grad_norm": 1.2785775661468506,
+      "learning_rate": 4.6334008655132264e-07,
+      "loss": 2.2142,
+      "step": 15758
+    },
+    {
+      "epoch": 0.9694266732283464,
+      "grad_norm": 1.1482257843017578,
+      "learning_rate": 4.6148258236867747e-07,
+      "loss": 2.358,
+      "step": 15759
+    },
+    {
+      "epoch": 0.969488188976378,
+      "grad_norm": 1.1911166906356812,
+      "learning_rate": 4.5962880037153653e-07,
+      "loss": 1.3046,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9695497047244095,
+      "grad_norm": 1.4464876651763916,
+      "learning_rate": 4.577787406292333e-07,
+      "loss": 2.9079,
+      "step": 15761
+    },
+    {
+      "epoch": 0.969611220472441,
+      "grad_norm": 1.464797854423523,
+      "learning_rate": 4.5593240321093467e-07,
+      "loss": 2.451,
+      "step": 15762
+    },
+    {
+      "epoch": 0.9696727362204725,
+      "grad_norm": 1.3844650983810425,
+      "learning_rate": 4.540897881856965e-07,
+      "loss": 2.3453,
+      "step": 15763
+    },
+    {
+      "epoch": 0.9697342519685039,
+      "grad_norm": 1.1274930238723755,
+      "learning_rate": 4.522508956224081e-07,
+      "loss": 2.8981,
+      "step": 15764
+    },
+    {
+      "epoch": 0.9697957677165354,
+      "grad_norm": 2.071282386779785,
+      "learning_rate": 4.504157255898589e-07,
+      "loss": 2.5745,
+      "step": 15765
+    },
+    {
+      "epoch": 0.9698572834645669,
+      "grad_norm": 1.6221636533737183,
+      "learning_rate": 4.485842781566385e-07,
+      "loss": 3.0721,
+      "step": 15766
+    },
+    {
+      "epoch": 0.9699187992125984,
+      "grad_norm": 1.2843923568725586,
+      "learning_rate": 4.467565533912699e-07,
+      "loss": 1.9281,
+      "step": 15767
+    },
+    {
+      "epoch": 0.96998031496063,
+      "grad_norm": 1.29819917678833,
+      "learning_rate": 4.4493255136207614e-07,
+      "loss": 2.3189,
+      "step": 15768
+    },
+    {
+      "epoch": 0.9700418307086615,
+      "grad_norm": 1.4912307262420654,
+      "learning_rate": 4.4311227213725823e-07,
+      "loss": 2.496,
+      "step": 15769
+    },
+    {
+      "epoch": 0.9701033464566929,
+      "grad_norm": 1.531643271446228,
+      "learning_rate": 4.4129571578490624e-07,
+      "loss": 2.9915,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9701648622047244,
+      "grad_norm": 2.086723804473877,
+      "learning_rate": 4.3948288237294356e-07,
+      "loss": 1.7969,
+      "step": 15771
+    },
+    {
+      "epoch": 0.9702263779527559,
+      "grad_norm": 1.3748822212219238,
+      "learning_rate": 4.376737719691604e-07,
+      "loss": 2.1083,
+      "step": 15772
+    },
+    {
+      "epoch": 0.9702878937007874,
+      "grad_norm": 1.3617067337036133,
+      "learning_rate": 4.358683846411915e-07,
+      "loss": 2.2721,
+      "step": 15773
+    },
+    {
+      "epoch": 0.9703494094488189,
+      "grad_norm": 1.3305469751358032,
+      "learning_rate": 4.340667204565718e-07,
+      "loss": 2.3915,
+      "step": 15774
+    },
+    {
+      "epoch": 0.9704109251968503,
+      "grad_norm": 1.3735666275024414,
+      "learning_rate": 4.3226877948264743e-07,
+      "loss": 2.4312,
+      "step": 15775
+    },
+    {
+      "epoch": 0.9704724409448819,
+      "grad_norm": 1.3628467321395874,
+      "learning_rate": 4.3047456178667565e-07,
+      "loss": 2.8027,
+      "step": 15776
+    },
+    {
+      "epoch": 0.9705339566929134,
+      "grad_norm": 1.1885515451431274,
+      "learning_rate": 4.286840674357473e-07,
+      "loss": 2.1789,
+      "step": 15777
+    },
+    {
+      "epoch": 0.9705954724409449,
+      "grad_norm": 1.37073814868927,
+      "learning_rate": 4.2689729649679766e-07,
+      "loss": 1.9115,
+      "step": 15778
+    },
+    {
+      "epoch": 0.9706569881889764,
+      "grad_norm": 1.5532366037368774,
+      "learning_rate": 4.251142490366622e-07,
+      "loss": 1.7381,
+      "step": 15779
+    },
+    {
+      "epoch": 0.9707185039370079,
+      "grad_norm": 1.3696569204330444,
+      "learning_rate": 4.233349251219987e-07,
+      "loss": 2.4376,
+      "step": 15780
+    },
+    {
+      "epoch": 0.9707800196850394,
+      "grad_norm": 1.2605400085449219,
+      "learning_rate": 4.21559324819365e-07,
+      "loss": 2.2753,
+      "step": 15781
+    },
+    {
+      "epoch": 0.9708415354330708,
+      "grad_norm": 1.29276704788208,
+      "learning_rate": 4.197874481951414e-07,
+      "loss": 2.4712,
+      "step": 15782
+    },
+    {
+      "epoch": 0.9709030511811023,
+      "grad_norm": 1.2072042226791382,
+      "learning_rate": 4.1801929531559703e-07,
+      "loss": 1.9012,
+      "step": 15783
+    },
+    {
+      "epoch": 0.9709645669291339,
+      "grad_norm": 1.3276649713516235,
+      "learning_rate": 4.1625486624684574e-07,
+      "loss": 2.4023,
+      "step": 15784
+    },
+    {
+      "epoch": 0.9710260826771654,
+      "grad_norm": 1.4134495258331299,
+      "learning_rate": 4.1449416105486805e-07,
+      "loss": 2.3308,
+      "step": 15785
+    },
+    {
+      "epoch": 0.9710875984251969,
+      "grad_norm": 1.1426512002944946,
+      "learning_rate": 4.127371798055002e-07,
+      "loss": 2.4167,
+      "step": 15786
+    },
+    {
+      "epoch": 0.9711491141732284,
+      "grad_norm": 1.2632418870925903,
+      "learning_rate": 4.109839225644563e-07,
+      "loss": 2.6614,
+      "step": 15787
+    },
+    {
+      "epoch": 0.9712106299212598,
+      "grad_norm": 1.3939414024353027,
+      "learning_rate": 4.0923438939728387e-07,
+      "loss": 2.8745,
+      "step": 15788
+    },
+    {
+      "epoch": 0.9712721456692913,
+      "grad_norm": 1.359014868736267,
+      "learning_rate": 4.074885803694084e-07,
+      "loss": 2.6995,
+      "step": 15789
+    },
+    {
+      "epoch": 0.9713336614173228,
+      "grad_norm": 1.3496167659759521,
+      "learning_rate": 4.057464955461221e-07,
+      "loss": 2.9765,
+      "step": 15790
+    },
+    {
+      "epoch": 0.9713951771653543,
+      "grad_norm": 1.2840732336044312,
+      "learning_rate": 4.040081349925617e-07,
+      "loss": 2.1654,
+      "step": 15791
+    },
+    {
+      "epoch": 0.9714566929133859,
+      "grad_norm": 1.2447259426116943,
+      "learning_rate": 4.02273498773742e-07,
+      "loss": 1.7575,
+      "step": 15792
+    },
+    {
+      "epoch": 0.9715182086614174,
+      "grad_norm": 1.4341068267822266,
+      "learning_rate": 4.0054258695451097e-07,
+      "loss": 2.2441,
+      "step": 15793
+    },
+    {
+      "epoch": 0.9715797244094488,
+      "grad_norm": 1.2828034162521362,
+      "learning_rate": 3.988153995996169e-07,
+      "loss": 2.5076,
+      "step": 15794
+    },
+    {
+      "epoch": 0.9716412401574803,
+      "grad_norm": 1.2866408824920654,
+      "learning_rate": 3.970919367736192e-07,
+      "loss": 2.8089,
+      "step": 15795
+    },
+    {
+      "epoch": 0.9717027559055118,
+      "grad_norm": 1.126671552658081,
+      "learning_rate": 3.9537219854099974e-07,
+      "loss": 1.3953,
+      "step": 15796
+    },
+    {
+      "epoch": 0.9717642716535433,
+      "grad_norm": 1.3030691146850586,
+      "learning_rate": 3.936561849660403e-07,
+      "loss": 2.2619,
+      "step": 15797
+    },
+    {
+      "epoch": 0.9718257874015748,
+      "grad_norm": 1.2543025016784668,
+      "learning_rate": 3.9194389611292294e-07,
+      "loss": 2.4769,
+      "step": 15798
+    },
+    {
+      "epoch": 0.9718873031496063,
+      "grad_norm": 1.3201347589492798,
+      "learning_rate": 3.902353320456742e-07,
+      "loss": 3.0737,
+      "step": 15799
+    },
+    {
+      "epoch": 0.9719488188976378,
+      "grad_norm": 1.242813229560852,
+      "learning_rate": 3.8853049282817634e-07,
+      "loss": 2.2713,
+      "step": 15800
+    },
+    {
+      "epoch": 0.9720103346456693,
+      "grad_norm": 1.576049566268921,
+      "learning_rate": 3.8682937852418946e-07,
+      "loss": 2.9368,
+      "step": 15801
+    },
+    {
+      "epoch": 0.9720718503937008,
+      "grad_norm": 1.286734700202942,
+      "learning_rate": 3.851319891973404e-07,
+      "loss": 1.2955,
+      "step": 15802
+    },
+    {
+      "epoch": 0.9721333661417323,
+      "grad_norm": 1.465498924255371,
+      "learning_rate": 3.8343832491107844e-07,
+      "loss": 2.4538,
+      "step": 15803
+    },
+    {
+      "epoch": 0.9721948818897638,
+      "grad_norm": 1.3682917356491089,
+      "learning_rate": 3.8174838572874184e-07,
+      "loss": 2.6151,
+      "step": 15804
+    },
+    {
+      "epoch": 0.9722563976377953,
+      "grad_norm": 1.184009075164795,
+      "learning_rate": 3.8006217171353554e-07,
+      "loss": 1.9901,
+      "step": 15805
+    },
+    {
+      "epoch": 0.9723179133858267,
+      "grad_norm": 1.3515723943710327,
+      "learning_rate": 3.78379682928498e-07,
+      "loss": 2.3574,
+      "step": 15806
+    },
+    {
+      "epoch": 0.9723794291338582,
+      "grad_norm": 1.270646333694458,
+      "learning_rate": 3.767009194365678e-07,
+      "loss": 2.3498,
+      "step": 15807
+    },
+    {
+      "epoch": 0.9724409448818898,
+      "grad_norm": 1.2812057733535767,
+      "learning_rate": 3.750258813005059e-07,
+      "loss": 2.688,
+      "step": 15808
+    },
+    {
+      "epoch": 0.9725024606299213,
+      "grad_norm": 1.3432518243789673,
+      "learning_rate": 3.733545685829398e-07,
+      "loss": 2.8517,
+      "step": 15809
+    },
+    {
+      "epoch": 0.9725639763779528,
+      "grad_norm": 1.3773226737976074,
+      "learning_rate": 3.716869813463975e-07,
+      "loss": 2.3767,
+      "step": 15810
+    },
+    {
+      "epoch": 0.9726254921259843,
+      "grad_norm": 1.1533784866333008,
+      "learning_rate": 3.700231196532067e-07,
+      "loss": 2.7073,
+      "step": 15811
+    },
+    {
+      "epoch": 0.9726870078740157,
+      "grad_norm": 1.3323701620101929,
+      "learning_rate": 3.683629835656066e-07,
+      "loss": 2.8513,
+      "step": 15812
+    },
+    {
+      "epoch": 0.9727485236220472,
+      "grad_norm": 1.3052810430526733,
+      "learning_rate": 3.667065731456698e-07,
+      "loss": 2.3045,
+      "step": 15813
+    },
+    {
+      "epoch": 0.9728100393700787,
+      "grad_norm": 1.3455829620361328,
+      "learning_rate": 3.650538884553245e-07,
+      "loss": 2.0967,
+      "step": 15814
+    },
+    {
+      "epoch": 0.9728715551181102,
+      "grad_norm": 1.2006338834762573,
+      "learning_rate": 3.6340492955639903e-07,
+      "loss": 2.4608,
+      "step": 15815
+    },
+    {
+      "epoch": 0.9729330708661418,
+      "grad_norm": 1.4298021793365479,
+      "learning_rate": 3.61759696510533e-07,
+      "loss": 1.8174,
+      "step": 15816
+    },
+    {
+      "epoch": 0.9729945866141733,
+      "grad_norm": 1.7693294286727905,
+      "learning_rate": 3.601181893792549e-07,
+      "loss": 1.9655,
+      "step": 15817
+    },
+    {
+      "epoch": 0.9730561023622047,
+      "grad_norm": 1.3947510719299316,
+      "learning_rate": 3.58480408223949e-07,
+      "loss": 2.9283,
+      "step": 15818
+    },
+    {
+      "epoch": 0.9731176181102362,
+      "grad_norm": 1.275413155555725,
+      "learning_rate": 3.5684635310586633e-07,
+      "loss": 2.5987,
+      "step": 15819
+    },
+    {
+      "epoch": 0.9731791338582677,
+      "grad_norm": 1.4320645332336426,
+      "learning_rate": 3.552160240860913e-07,
+      "loss": 2.4373,
+      "step": 15820
+    },
+    {
+      "epoch": 0.9732406496062992,
+      "grad_norm": 1.2888295650482178,
+      "learning_rate": 3.5358942122561966e-07,
+      "loss": 2.096,
+      "step": 15821
+    },
+    {
+      "epoch": 0.9733021653543307,
+      "grad_norm": 1.5256470441818237,
+      "learning_rate": 3.5196654458524715e-07,
+      "loss": 2.407,
+      "step": 15822
+    },
+    {
+      "epoch": 0.9733636811023622,
+      "grad_norm": 1.6735856533050537,
+      "learning_rate": 3.5034739422568074e-07,
+      "loss": 2.3796,
+      "step": 15823
+    },
+    {
+      "epoch": 0.9734251968503937,
+      "grad_norm": 1.3600356578826904,
+      "learning_rate": 3.4873197020746095e-07,
+      "loss": 2.5343,
+      "step": 15824
+    },
+    {
+      "epoch": 0.9734867125984252,
+      "grad_norm": 1.3545360565185547,
+      "learning_rate": 3.471202725909839e-07,
+      "loss": 2.2755,
+      "step": 15825
+    },
+    {
+      "epoch": 0.9735482283464567,
+      "grad_norm": 2.1818480491638184,
+      "learning_rate": 3.455123014365347e-07,
+      "loss": 2.0238,
+      "step": 15826
+    },
+    {
+      "epoch": 0.9736097440944882,
+      "grad_norm": 1.6492441892623901,
+      "learning_rate": 3.4390805680424296e-07,
+      "loss": 2.1228,
+      "step": 15827
+    },
+    {
+      "epoch": 0.9736712598425197,
+      "grad_norm": 1.1041011810302734,
+      "learning_rate": 3.4230753875408306e-07,
+      "loss": 2.2773,
+      "step": 15828
+    },
+    {
+      "epoch": 0.9737327755905512,
+      "grad_norm": 1.6212236881256104,
+      "learning_rate": 3.407107473459292e-07,
+      "loss": 2.5171,
+      "step": 15829
+    },
+    {
+      "epoch": 0.9737942913385826,
+      "grad_norm": 1.1018040180206299,
+      "learning_rate": 3.391176826394671e-07,
+      "loss": 2.6732,
+      "step": 15830
+    },
+    {
+      "epoch": 0.9738558070866141,
+      "grad_norm": 1.2644505500793457,
+      "learning_rate": 3.375283446942823e-07,
+      "loss": 2.622,
+      "step": 15831
+    },
+    {
+      "epoch": 0.9739173228346457,
+      "grad_norm": 1.4967069625854492,
+      "learning_rate": 3.35942733569794e-07,
+      "loss": 1.9601,
+      "step": 15832
+    },
+    {
+      "epoch": 0.9739788385826772,
+      "grad_norm": 1.538697600364685,
+      "learning_rate": 3.343608493253103e-07,
+      "loss": 3.2723,
+      "step": 15833
+    },
+    {
+      "epoch": 0.9740403543307087,
+      "grad_norm": 1.157845139503479,
+      "learning_rate": 3.327826920199839e-07,
+      "loss": 2.665,
+      "step": 15834
+    },
+    {
+      "epoch": 0.9741018700787402,
+      "grad_norm": 1.5932539701461792,
+      "learning_rate": 3.312082617128232e-07,
+      "loss": 2.1221,
+      "step": 15835
+    },
+    {
+      "epoch": 0.9741633858267716,
+      "grad_norm": 1.1353089809417725,
+      "learning_rate": 3.2963755846271426e-07,
+      "loss": 1.7399,
+      "step": 15836
+    },
+    {
+      "epoch": 0.9742249015748031,
+      "grad_norm": 1.5373746156692505,
+      "learning_rate": 3.2807058232837693e-07,
+      "loss": 1.7421,
+      "step": 15837
+    },
+    {
+      "epoch": 0.9742864173228346,
+      "grad_norm": 1.3493603467941284,
+      "learning_rate": 3.2650733336840877e-07,
+      "loss": 2.9283,
+      "step": 15838
+    },
+    {
+      "epoch": 0.9743479330708661,
+      "grad_norm": 1.3362170457839966,
+      "learning_rate": 3.249478116412741e-07,
+      "loss": 2.4734,
+      "step": 15839
+    },
+    {
+      "epoch": 0.9744094488188977,
+      "grad_norm": 1.1665818691253662,
+      "learning_rate": 3.2339201720528177e-07,
+      "loss": 2.2701,
+      "step": 15840
+    },
+    {
+      "epoch": 0.9744709645669292,
+      "grad_norm": 1.3074346780776978,
+      "learning_rate": 3.218399501186298e-07,
+      "loss": 2.6449,
+      "step": 15841
+    },
+    {
+      "epoch": 0.9745324803149606,
+      "grad_norm": 1.2741414308547974,
+      "learning_rate": 3.202916104393383e-07,
+      "loss": 2.7027,
+      "step": 15842
+    },
+    {
+      "epoch": 0.9745939960629921,
+      "grad_norm": 1.239997386932373,
+      "learning_rate": 3.187469982253055e-07,
+      "loss": 2.0118,
+      "step": 15843
+    },
+    {
+      "epoch": 0.9746555118110236,
+      "grad_norm": 1.2969248294830322,
+      "learning_rate": 3.172061135343074e-07,
+      "loss": 2.8303,
+      "step": 15844
+    },
+    {
+      "epoch": 0.9747170275590551,
+      "grad_norm": 1.4356331825256348,
+      "learning_rate": 3.1566895642394236e-07,
+      "loss": 2.2353,
+      "step": 15845
+    },
+    {
+      "epoch": 0.9747785433070866,
+      "grad_norm": 1.1817688941955566,
+      "learning_rate": 3.141355269517088e-07,
+      "loss": 1.9185,
+      "step": 15846
+    },
+    {
+      "epoch": 0.9748400590551181,
+      "grad_norm": 1.3410853147506714,
+      "learning_rate": 3.126058251749497e-07,
+      "loss": 3.0871,
+      "step": 15847
+    },
+    {
+      "epoch": 0.9749015748031497,
+      "grad_norm": 1.2362356185913086,
+      "learning_rate": 3.110798511508528e-07,
+      "loss": 2.8952,
+      "step": 15848
+    },
+    {
+      "epoch": 0.9749630905511811,
+      "grad_norm": 1.2280429601669312,
+      "learning_rate": 3.0955760493648346e-07,
+      "loss": 2.3965,
+      "step": 15849
+    },
+    {
+      "epoch": 0.9750246062992126,
+      "grad_norm": 1.2037022113800049,
+      "learning_rate": 3.08039086588785e-07,
+      "loss": 2.4907,
+      "step": 15850
+    },
+    {
+      "epoch": 0.9750861220472441,
+      "grad_norm": 1.2332780361175537,
+      "learning_rate": 3.0652429616451205e-07,
+      "loss": 2.3254,
+      "step": 15851
+    },
+    {
+      "epoch": 0.9751476377952756,
+      "grad_norm": 1.1763156652450562,
+      "learning_rate": 3.050132337203304e-07,
+      "loss": 2.5389,
+      "step": 15852
+    },
+    {
+      "epoch": 0.9752091535433071,
+      "grad_norm": 1.375347375869751,
+      "learning_rate": 3.0350589931273933e-07,
+      "loss": 2.3266,
+      "step": 15853
+    },
+    {
+      "epoch": 0.9752706692913385,
+      "grad_norm": 1.4314281940460205,
+      "learning_rate": 3.0200229299811587e-07,
+      "loss": 2.4752,
+      "step": 15854
+    },
+    {
+      "epoch": 0.97533218503937,
+      "grad_norm": 1.4876974821090698,
+      "learning_rate": 3.0050241483265964e-07,
+      "loss": 2.3756,
+      "step": 15855
+    },
+    {
+      "epoch": 0.9753937007874016,
+      "grad_norm": 1.3537023067474365,
+      "learning_rate": 2.990062648724812e-07,
+      "loss": 2.0891,
+      "step": 15856
+    },
+    {
+      "epoch": 0.9754552165354331,
+      "grad_norm": 1.3611165285110474,
+      "learning_rate": 2.9751384317351363e-07,
+      "loss": 2.3077,
+      "step": 15857
+    },
+    {
+      "epoch": 0.9755167322834646,
+      "grad_norm": 1.401503324508667,
+      "learning_rate": 2.9602514979157893e-07,
+      "loss": 2.4064,
+      "step": 15858
+    },
+    {
+      "epoch": 0.9755782480314961,
+      "grad_norm": 1.467042088508606,
+      "learning_rate": 2.9454018478233257e-07,
+      "loss": 2.5116,
+      "step": 15859
+    },
+    {
+      "epoch": 0.9756397637795275,
+      "grad_norm": 1.4409202337265015,
+      "learning_rate": 2.930589482013191e-07,
+      "loss": 3.0566,
+      "step": 15860
+    },
+    {
+      "epoch": 0.975701279527559,
+      "grad_norm": 1.4353641271591187,
+      "learning_rate": 2.915814401039163e-07,
+      "loss": 2.8857,
+      "step": 15861
+    },
+    {
+      "epoch": 0.9757627952755905,
+      "grad_norm": 1.1870088577270508,
+      "learning_rate": 2.901076605453801e-07,
+      "loss": 1.6666,
+      "step": 15862
+    },
+    {
+      "epoch": 0.975824311023622,
+      "grad_norm": 1.9847238063812256,
+      "learning_rate": 2.886376095808108e-07,
+      "loss": 2.3518,
+      "step": 15863
+    },
+    {
+      "epoch": 0.9758858267716536,
+      "grad_norm": 1.3779948949813843,
+      "learning_rate": 2.871712872651977e-07,
+      "loss": 2.7434,
+      "step": 15864
+    },
+    {
+      "epoch": 0.9759473425196851,
+      "grad_norm": 1.282639980316162,
+      "learning_rate": 2.8570869365336374e-07,
+      "loss": 2.3331,
+      "step": 15865
+    },
+    {
+      "epoch": 0.9760088582677166,
+      "grad_norm": 1.3895840644836426,
+      "learning_rate": 2.842498287999984e-07,
+      "loss": 2.4863,
+      "step": 15866
+    },
+    {
+      "epoch": 0.976070374015748,
+      "grad_norm": 1.2272924184799194,
+      "learning_rate": 2.827946927596581e-07,
+      "loss": 2.9668,
+      "step": 15867
+    },
+    {
+      "epoch": 0.9761318897637795,
+      "grad_norm": 1.4481498003005981,
+      "learning_rate": 2.8134328558676594e-07,
+      "loss": 2.2325,
+      "step": 15868
+    },
+    {
+      "epoch": 0.976193405511811,
+      "grad_norm": 1.4165005683898926,
+      "learning_rate": 2.7989560733557853e-07,
+      "loss": 2.2569,
+      "step": 15869
+    },
+    {
+      "epoch": 0.9762549212598425,
+      "grad_norm": 1.4401319026947021,
+      "learning_rate": 2.784516580602525e-07,
+      "loss": 2.488,
+      "step": 15870
+    },
+    {
+      "epoch": 0.976316437007874,
+      "grad_norm": 1.4553871154785156,
+      "learning_rate": 2.77011437814767e-07,
+      "loss": 2.3672,
+      "step": 15871
+    },
+    {
+      "epoch": 0.9763779527559056,
+      "grad_norm": 1.340387225151062,
+      "learning_rate": 2.755749466529789e-07,
+      "loss": 2.3537,
+      "step": 15872
+    },
+    {
+      "epoch": 0.976439468503937,
+      "grad_norm": 1.1635823249816895,
+      "learning_rate": 2.7414218462861187e-07,
+      "loss": 2.3216,
+      "step": 15873
+    },
+    {
+      "epoch": 0.9765009842519685,
+      "grad_norm": 1.3356642723083496,
+      "learning_rate": 2.7271315179524525e-07,
+      "loss": 2.6131,
+      "step": 15874
+    },
+    {
+      "epoch": 0.9765625,
+      "grad_norm": 1.3838838338851929,
+      "learning_rate": 2.7128784820630305e-07,
+      "loss": 2.4839,
+      "step": 15875
+    },
+    {
+      "epoch": 0.9766240157480315,
+      "grad_norm": 1.5713809728622437,
+      "learning_rate": 2.6986627391509813e-07,
+      "loss": 2.3361,
+      "step": 15876
+    },
+    {
+      "epoch": 0.976685531496063,
+      "grad_norm": 1.3006007671356201,
+      "learning_rate": 2.6844842897477686e-07,
+      "loss": 1.8498,
+      "step": 15877
+    },
+    {
+      "epoch": 0.9767470472440944,
+      "grad_norm": 1.2507625818252563,
+      "learning_rate": 2.670343134383746e-07,
+      "loss": 2.5892,
+      "step": 15878
+    },
+    {
+      "epoch": 0.976808562992126,
+      "grad_norm": 1.1625349521636963,
+      "learning_rate": 2.656239273587602e-07,
+      "loss": 2.3866,
+      "step": 15879
+    },
+    {
+      "epoch": 0.9768700787401575,
+      "grad_norm": 1.1821539402008057,
+      "learning_rate": 2.642172707886803e-07,
+      "loss": 2.7876,
+      "step": 15880
+    },
+    {
+      "epoch": 0.976931594488189,
+      "grad_norm": 1.145930290222168,
+      "learning_rate": 2.6281434378072624e-07,
+      "loss": 2.9147,
+      "step": 15881
+    },
+    {
+      "epoch": 0.9769931102362205,
+      "grad_norm": 1.1941183805465698,
+      "learning_rate": 2.614151463873671e-07,
+      "loss": 2.4407,
+      "step": 15882
+    },
+    {
+      "epoch": 0.977054625984252,
+      "grad_norm": 1.5458219051361084,
+      "learning_rate": 2.600196786609277e-07,
+      "loss": 2.1748,
+      "step": 15883
+    },
+    {
+      "epoch": 0.9771161417322834,
+      "grad_norm": 1.3398237228393555,
+      "learning_rate": 2.5862794065358854e-07,
+      "loss": 2.542,
+      "step": 15884
+    },
+    {
+      "epoch": 0.9771776574803149,
+      "grad_norm": 1.1226738691329956,
+      "learning_rate": 2.5723993241738577e-07,
+      "loss": 2.7162,
+      "step": 15885
+    },
+    {
+      "epoch": 0.9772391732283464,
+      "grad_norm": 1.2778303623199463,
+      "learning_rate": 2.558556540042334e-07,
+      "loss": 2.6784,
+      "step": 15886
+    },
+    {
+      "epoch": 0.977300688976378,
+      "grad_norm": 1.5260999202728271,
+      "learning_rate": 2.5447510546587894e-07,
+      "loss": 2.4666,
+      "step": 15887
+    },
+    {
+      "epoch": 0.9773622047244095,
+      "grad_norm": 1.275529384613037,
+      "learning_rate": 2.5309828685397e-07,
+      "loss": 2.5457,
+      "step": 15888
+    },
+    {
+      "epoch": 0.977423720472441,
+      "grad_norm": 1.3936446905136108,
+      "learning_rate": 2.517251982199875e-07,
+      "loss": 3.0454,
+      "step": 15889
+    },
+    {
+      "epoch": 0.9774852362204725,
+      "grad_norm": 1.3073153495788574,
+      "learning_rate": 2.503558396152572e-07,
+      "loss": 2.0715,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9775467519685039,
+      "grad_norm": 1.465147614479065,
+      "learning_rate": 2.489902110910047e-07,
+      "loss": 2.0755,
+      "step": 15891
+    },
+    {
+      "epoch": 0.9776082677165354,
+      "grad_norm": 1.2706907987594604,
+      "learning_rate": 2.476283126982892e-07,
+      "loss": 2.2687,
+      "step": 15892
+    },
+    {
+      "epoch": 0.9776697834645669,
+      "grad_norm": 1.083569049835205,
+      "learning_rate": 2.4627014448804777e-07,
+      "loss": 2.7236,
+      "step": 15893
+    },
+    {
+      "epoch": 0.9777312992125984,
+      "grad_norm": 1.4399019479751587,
+      "learning_rate": 2.449157065110508e-07,
+      "loss": 2.1199,
+      "step": 15894
+    },
+    {
+      "epoch": 0.97779281496063,
+      "grad_norm": 1.350998044013977,
+      "learning_rate": 2.435649988179689e-07,
+      "loss": 2.1498,
+      "step": 15895
+    },
+    {
+      "epoch": 0.9778543307086615,
+      "grad_norm": 1.3650400638580322,
+      "learning_rate": 2.4221802145928395e-07,
+      "loss": 2.6779,
+      "step": 15896
+    },
+    {
+      "epoch": 0.9779158464566929,
+      "grad_norm": 1.1939518451690674,
+      "learning_rate": 2.4087477448538897e-07,
+      "loss": 2.0388,
+      "step": 15897
+    },
+    {
+      "epoch": 0.9779773622047244,
+      "grad_norm": 1.403867244720459,
+      "learning_rate": 2.3953525794651044e-07,
+      "loss": 2.4406,
+      "step": 15898
+    },
+    {
+      "epoch": 0.9780388779527559,
+      "grad_norm": 1.3058817386627197,
+      "learning_rate": 2.381994718927305e-07,
+      "loss": 2.9261,
+      "step": 15899
+    },
+    {
+      "epoch": 0.9781003937007874,
+      "grad_norm": 1.2546288967132568,
+      "learning_rate": 2.368674163739981e-07,
+      "loss": 2.3953,
+      "step": 15900
+    },
+    {
+      "epoch": 0.9781619094488189,
+      "grad_norm": 1.2046606540679932,
+      "learning_rate": 2.3553909144012898e-07,
+      "loss": 1.9605,
+      "step": 15901
+    },
+    {
+      "epoch": 0.9782234251968503,
+      "grad_norm": 1.3906537294387817,
+      "learning_rate": 2.3421449714080555e-07,
+      "loss": 2.6343,
+      "step": 15902
+    },
+    {
+      "epoch": 0.9782849409448819,
+      "grad_norm": 1.4849977493286133,
+      "learning_rate": 2.3289363352554383e-07,
+      "loss": 2.3655,
+      "step": 15903
+    },
+    {
+      "epoch": 0.9783464566929134,
+      "grad_norm": 1.4336485862731934,
+      "learning_rate": 2.315765006437376e-07,
+      "loss": 2.5171,
+      "step": 15904
+    },
+    {
+      "epoch": 0.9784079724409449,
+      "grad_norm": 1.2071067094802856,
+      "learning_rate": 2.3026309854465856e-07,
+      "loss": 2.6291,
+      "step": 15905
+    },
+    {
+      "epoch": 0.9784694881889764,
+      "grad_norm": 1.1896963119506836,
+      "learning_rate": 2.2895342727740076e-07,
+      "loss": 2.9066,
+      "step": 15906
+    },
+    {
+      "epoch": 0.9785310039370079,
+      "grad_norm": 1.1849664449691772,
+      "learning_rate": 2.2764748689094728e-07,
+      "loss": 2.0586,
+      "step": 15907
+    },
+    {
+      "epoch": 0.9785925196850394,
+      "grad_norm": 1.3824048042297363,
+      "learning_rate": 2.2634527743413682e-07,
+      "loss": 2.6484,
+      "step": 15908
+    },
+    {
+      "epoch": 0.9786540354330708,
+      "grad_norm": 1.374638319015503,
+      "learning_rate": 2.250467989556526e-07,
+      "loss": 3.8704,
+      "step": 15909
+    },
+    {
+      "epoch": 0.9787155511811023,
+      "grad_norm": 1.3884894847869873,
+      "learning_rate": 2.2375205150405586e-07,
+      "loss": 2.5771,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9787770669291339,
+      "grad_norm": 1.257631540298462,
+      "learning_rate": 2.224610351277745e-07,
+      "loss": 2.5341,
+      "step": 15911
+    },
+    {
+      "epoch": 0.9788385826771654,
+      "grad_norm": 1.305263876914978,
+      "learning_rate": 2.2117374987506988e-07,
+      "loss": 2.3545,
+      "step": 15912
+    },
+    {
+      "epoch": 0.9789000984251969,
+      "grad_norm": 1.3030636310577393,
+      "learning_rate": 2.1989019579409242e-07,
+      "loss": 2.1693,
+      "step": 15913
+    },
+    {
+      "epoch": 0.9789616141732284,
+      "grad_norm": 1.397101879119873,
+      "learning_rate": 2.1861037293281483e-07,
+      "loss": 2.5412,
+      "step": 15914
+    },
+    {
+      "epoch": 0.9790231299212598,
+      "grad_norm": 2.4480772018432617,
+      "learning_rate": 2.1733428133913213e-07,
+      "loss": 1.5865,
+      "step": 15915
+    },
+    {
+      "epoch": 0.9790846456692913,
+      "grad_norm": 1.3145854473114014,
+      "learning_rate": 2.1606192106072842e-07,
+      "loss": 2.3521,
+      "step": 15916
+    },
+    {
+      "epoch": 0.9791461614173228,
+      "grad_norm": 1.191148281097412,
+      "learning_rate": 2.1479329214521004e-07,
+      "loss": 2.6992,
+      "step": 15917
+    },
+    {
+      "epoch": 0.9792076771653543,
+      "grad_norm": 1.6723952293395996,
+      "learning_rate": 2.1352839463999462e-07,
+      "loss": 2.826,
+      "step": 15918
+    },
+    {
+      "epoch": 0.9792691929133859,
+      "grad_norm": 1.3244352340698242,
+      "learning_rate": 2.12267228592411e-07,
+      "loss": 1.419,
+      "step": 15919
+    },
+    {
+      "epoch": 0.9793307086614174,
+      "grad_norm": 1.2205731868743896,
+      "learning_rate": 2.110097940495881e-07,
+      "loss": 1.934,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9793922244094488,
+      "grad_norm": 1.3408907651901245,
+      "learning_rate": 2.09756091058555e-07,
+      "loss": 2.2972,
+      "step": 15921
+    },
+    {
+      "epoch": 0.9794537401574803,
+      "grad_norm": 1.3876562118530273,
+      "learning_rate": 2.085061196662075e-07,
+      "loss": 2.497,
+      "step": 15922
+    },
+    {
+      "epoch": 0.9795152559055118,
+      "grad_norm": 1.2368251085281372,
+      "learning_rate": 2.0725987991927488e-07,
+      "loss": 2.4641,
+      "step": 15923
+    },
+    {
+      "epoch": 0.9795767716535433,
+      "grad_norm": 1.225169062614441,
+      "learning_rate": 2.060173718643643e-07,
+      "loss": 2.5164,
+      "step": 15924
+    },
+    {
+      "epoch": 0.9796382874015748,
+      "grad_norm": 1.2678732872009277,
+      "learning_rate": 2.0477859554793865e-07,
+      "loss": 2.6069,
+      "step": 15925
+    },
+    {
+      "epoch": 0.9796998031496063,
+      "grad_norm": 1.511687159538269,
+      "learning_rate": 2.035435510163164e-07,
+      "loss": 2.482,
+      "step": 15926
+    },
+    {
+      "epoch": 0.9797613188976378,
+      "grad_norm": 1.332966685295105,
+      "learning_rate": 2.023122383156939e-07,
+      "loss": 2.8428,
+      "step": 15927
+    },
+    {
+      "epoch": 0.9798228346456693,
+      "grad_norm": 1.3544002771377563,
+      "learning_rate": 2.0108465749208994e-07,
+      "loss": 2.3408,
+      "step": 15928
+    },
+    {
+      "epoch": 0.9798843503937008,
+      "grad_norm": 1.3964519500732422,
+      "learning_rate": 1.998608085914344e-07,
+      "loss": 2.221,
+      "step": 15929
+    },
+    {
+      "epoch": 0.9799458661417323,
+      "grad_norm": 1.3117096424102783,
+      "learning_rate": 1.9864069165949072e-07,
+      "loss": 2.8856,
+      "step": 15930
+    },
+    {
+      "epoch": 0.9800073818897638,
+      "grad_norm": 1.13992178440094,
+      "learning_rate": 1.9742430674186685e-07,
+      "loss": 2.8856,
+      "step": 15931
+    },
+    {
+      "epoch": 0.9800688976377953,
+      "grad_norm": 1.3081655502319336,
+      "learning_rate": 1.962116538840708e-07,
+      "loss": 2.2132,
+      "step": 15932
+    },
+    {
+      "epoch": 0.9801304133858267,
+      "grad_norm": 1.1626695394515991,
+      "learning_rate": 1.9500273313143302e-07,
+      "loss": 2.7299,
+      "step": 15933
+    },
+    {
+      "epoch": 0.9801919291338582,
+      "grad_norm": 1.3857423067092896,
+      "learning_rate": 1.9379754452916177e-07,
+      "loss": 2.413,
+      "step": 15934
+    },
+    {
+      "epoch": 0.9802534448818898,
+      "grad_norm": 1.1748541593551636,
+      "learning_rate": 1.9259608812233209e-07,
+      "loss": 2.8933,
+      "step": 15935
+    },
+    {
+      "epoch": 0.9803149606299213,
+      "grad_norm": 1.3325790166854858,
+      "learning_rate": 1.9139836395586363e-07,
+      "loss": 2.4716,
+      "step": 15936
+    },
+    {
+      "epoch": 0.9803764763779528,
+      "grad_norm": 1.298359751701355,
+      "learning_rate": 1.902043720745428e-07,
+      "loss": 2.1817,
+      "step": 15937
+    },
+    {
+      "epoch": 0.9804379921259843,
+      "grad_norm": 1.2691904306411743,
+      "learning_rate": 1.890141125230338e-07,
+      "loss": 2.6458,
+      "step": 15938
+    },
+    {
+      "epoch": 0.9804995078740157,
+      "grad_norm": 1.5235674381256104,
+      "learning_rate": 1.8782758534583446e-07,
+      "loss": 2.307,
+      "step": 15939
+    },
+    {
+      "epoch": 0.9805610236220472,
+      "grad_norm": 1.406773328781128,
+      "learning_rate": 1.8664479058730921e-07,
+      "loss": 2.3122,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9806225393700787,
+      "grad_norm": 1.45215904712677,
+      "learning_rate": 1.854657282917005e-07,
+      "loss": 2.1902,
+      "step": 15941
+    },
+    {
+      "epoch": 0.9806840551181102,
+      "grad_norm": 1.432749629020691,
+      "learning_rate": 1.8429039850309525e-07,
+      "loss": 2.2034,
+      "step": 15942
+    },
+    {
+      "epoch": 0.9807455708661418,
+      "grad_norm": 1.3459323644638062,
+      "learning_rate": 1.8311880126543612e-07,
+      "loss": 2.726,
+      "step": 15943
+    },
+    {
+      "epoch": 0.9808070866141733,
+      "grad_norm": 1.4736461639404297,
+      "learning_rate": 1.8195093662254359e-07,
+      "loss": 2.2371,
+      "step": 15944
+    },
+    {
+      "epoch": 0.9808686023622047,
+      "grad_norm": 1.0819653272628784,
+      "learning_rate": 1.8078680461807162e-07,
+      "loss": 2.404,
+      "step": 15945
+    },
+    {
+      "epoch": 0.9809301181102362,
+      "grad_norm": 1.6717368364334106,
+      "learning_rate": 1.796264052955854e-07,
+      "loss": 3.1195,
+      "step": 15946
+    },
+    {
+      "epoch": 0.9809916338582677,
+      "grad_norm": 1.5173876285552979,
+      "learning_rate": 1.7846973869843908e-07,
+      "loss": 2.1021,
+      "step": 15947
+    },
+    {
+      "epoch": 0.9810531496062992,
+      "grad_norm": 1.482761263847351,
+      "learning_rate": 1.773168048699203e-07,
+      "loss": 2.3628,
+      "step": 15948
+    },
+    {
+      "epoch": 0.9811146653543307,
+      "grad_norm": 1.6092276573181152,
+      "learning_rate": 1.7616760385311682e-07,
+      "loss": 1.7161,
+      "step": 15949
+    },
+    {
+      "epoch": 0.9811761811023622,
+      "grad_norm": 1.0579639673233032,
+      "learning_rate": 1.7502213569101644e-07,
+      "loss": 2.4103,
+      "step": 15950
+    },
+    {
+      "epoch": 0.9812376968503937,
+      "grad_norm": 1.3098341226577759,
+      "learning_rate": 1.738804004264516e-07,
+      "loss": 1.9625,
+      "step": 15951
+    },
+    {
+      "epoch": 0.9812992125984252,
+      "grad_norm": 1.2257238626480103,
+      "learning_rate": 1.7274239810211034e-07,
+      "loss": 1.8124,
+      "step": 15952
+    },
+    {
+      "epoch": 0.9813607283464567,
+      "grad_norm": 1.2383122444152832,
+      "learning_rate": 1.7160812876054755e-07,
+      "loss": 2.6774,
+      "step": 15953
+    },
+    {
+      "epoch": 0.9814222440944882,
+      "grad_norm": 1.3962665796279907,
+      "learning_rate": 1.7047759244418482e-07,
+      "loss": 1.6153,
+      "step": 15954
+    },
+    {
+      "epoch": 0.9814837598425197,
+      "grad_norm": 1.1472549438476562,
+      "learning_rate": 1.6935078919529945e-07,
+      "loss": 2.3307,
+      "step": 15955
+    },
+    {
+      "epoch": 0.9815452755905512,
+      "grad_norm": 1.4073847532272339,
+      "learning_rate": 1.682277190560244e-07,
+      "loss": 2.868,
+      "step": 15956
+    },
+    {
+      "epoch": 0.9816067913385826,
+      "grad_norm": 1.2981034517288208,
+      "learning_rate": 1.6710838206834833e-07,
+      "loss": 2.4604,
+      "step": 15957
+    },
+    {
+      "epoch": 0.9816683070866141,
+      "grad_norm": 1.0974900722503662,
+      "learning_rate": 1.6599277827413772e-07,
+      "loss": 3.1434,
+      "step": 15958
+    },
+    {
+      "epoch": 0.9817298228346457,
+      "grad_norm": 1.4839599132537842,
+      "learning_rate": 1.6488090771511478e-07,
+      "loss": 3.3106,
+      "step": 15959
+    },
+    {
+      "epoch": 0.9817913385826772,
+      "grad_norm": 1.172499418258667,
+      "learning_rate": 1.637727704328351e-07,
+      "loss": 2.2373,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9818528543307087,
+      "grad_norm": 1.2118163108825684,
+      "learning_rate": 1.626683664687656e-07,
+      "loss": 2.8315,
+      "step": 15961
+    },
+    {
+      "epoch": 0.9819143700787402,
+      "grad_norm": 1.364288091659546,
+      "learning_rate": 1.615676958641843e-07,
+      "loss": 2.4819,
+      "step": 15962
+    },
+    {
+      "epoch": 0.9819758858267716,
+      "grad_norm": 1.191705346107483,
+      "learning_rate": 1.604707586602583e-07,
+      "loss": 2.3912,
+      "step": 15963
+    },
+    {
+      "epoch": 0.9820374015748031,
+      "grad_norm": 1.247281551361084,
+      "learning_rate": 1.593775548980103e-07,
+      "loss": 2.7388,
+      "step": 15964
+    },
+    {
+      "epoch": 0.9820989173228346,
+      "grad_norm": 1.255872368812561,
+      "learning_rate": 1.582880846183077e-07,
+      "loss": 3.1458,
+      "step": 15965
+    },
+    {
+      "epoch": 0.9821604330708661,
+      "grad_norm": 1.4052519798278809,
+      "learning_rate": 1.572023478618956e-07,
+      "loss": 2.7743,
+      "step": 15966
+    },
+    {
+      "epoch": 0.9822219488188977,
+      "grad_norm": 1.968884825706482,
+      "learning_rate": 1.5612034466938597e-07,
+      "loss": 2.4145,
+      "step": 15967
+    },
+    {
+      "epoch": 0.9822834645669292,
+      "grad_norm": 1.2916620969772339,
+      "learning_rate": 1.5504207508122427e-07,
+      "loss": 1.7935,
+      "step": 15968
+    },
+    {
+      "epoch": 0.9823449803149606,
+      "grad_norm": 1.3969812393188477,
+      "learning_rate": 1.5396753913774486e-07,
+      "loss": 2.229,
+      "step": 15969
+    },
+    {
+      "epoch": 0.9824064960629921,
+      "grad_norm": 1.5504415035247803,
+      "learning_rate": 1.5289673687911566e-07,
+      "loss": 2.1218,
+      "step": 15970
+    },
+    {
+      "epoch": 0.9824680118110236,
+      "grad_norm": 1.2651127576828003,
+      "learning_rate": 1.5182966834539348e-07,
+      "loss": 2.7432,
+      "step": 15971
+    },
+    {
+      "epoch": 0.9825295275590551,
+      "grad_norm": 1.2472562789916992,
+      "learning_rate": 1.5076633357646863e-07,
+      "loss": 2.8086,
+      "step": 15972
+    },
+    {
+      "epoch": 0.9825910433070866,
+      "grad_norm": 1.3126990795135498,
+      "learning_rate": 1.4970673261210932e-07,
+      "loss": 2.0063,
+      "step": 15973
+    },
+    {
+      "epoch": 0.9826525590551181,
+      "grad_norm": 1.4068855047225952,
+      "learning_rate": 1.486508654919283e-07,
+      "loss": 2.6056,
+      "step": 15974
+    },
+    {
+      "epoch": 0.9827140748031497,
+      "grad_norm": 1.1377321481704712,
+      "learning_rate": 1.4759873225542732e-07,
+      "loss": 2.5352,
+      "step": 15975
+    },
+    {
+      "epoch": 0.9827755905511811,
+      "grad_norm": 1.3437339067459106,
+      "learning_rate": 1.4655033294194155e-07,
+      "loss": 2.505,
+      "step": 15976
+    },
+    {
+      "epoch": 0.9828371062992126,
+      "grad_norm": 1.5044078826904297,
+      "learning_rate": 1.4550566759068408e-07,
+      "loss": 2.5901,
+      "step": 15977
+    },
+    {
+      "epoch": 0.9828986220472441,
+      "grad_norm": 1.4546416997909546,
+      "learning_rate": 1.4446473624069035e-07,
+      "loss": 2.6899,
+      "step": 15978
+    },
+    {
+      "epoch": 0.9829601377952756,
+      "grad_norm": 1.4116569757461548,
+      "learning_rate": 1.4342753893091808e-07,
+      "loss": 1.9534,
+      "step": 15979
+    },
+    {
+      "epoch": 0.9830216535433071,
+      "grad_norm": 1.3173134326934814,
+      "learning_rate": 1.423940757001363e-07,
+      "loss": 2.5056,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9830831692913385,
+      "grad_norm": 1.3962268829345703,
+      "learning_rate": 1.4136434658700292e-07,
+      "loss": 2.5961,
+      "step": 15981
+    },
+    {
+      "epoch": 0.98314468503937,
+      "grad_norm": 1.253759503364563,
+      "learning_rate": 1.4033835163000942e-07,
+      "loss": 2.1302,
+      "step": 15982
+    },
+    {
+      "epoch": 0.9832062007874016,
+      "grad_norm": 1.2525475025177002,
+      "learning_rate": 1.3931609086753616e-07,
+      "loss": 2.8553,
+      "step": 15983
+    },
+    {
+      "epoch": 0.9832677165354331,
+      "grad_norm": 1.375497579574585,
+      "learning_rate": 1.3829756433778596e-07,
+      "loss": 2.3115,
+      "step": 15984
+    },
+    {
+      "epoch": 0.9833292322834646,
+      "grad_norm": 1.2153010368347168,
+      "learning_rate": 1.3728277207887274e-07,
+      "loss": 2.0597,
+      "step": 15985
+    },
+    {
+      "epoch": 0.9833907480314961,
+      "grad_norm": 1.579459547996521,
+      "learning_rate": 1.3627171412874396e-07,
+      "loss": 2.4549,
+      "step": 15986
+    },
+    {
+      "epoch": 0.9834522637795275,
+      "grad_norm": 1.2026474475860596,
+      "learning_rate": 1.3526439052518046e-07,
+      "loss": 2.4086,
+      "step": 15987
+    },
+    {
+      "epoch": 0.983513779527559,
+      "grad_norm": 1.2915359735488892,
+      "learning_rate": 1.3426080130587438e-07,
+      "loss": 2.7552,
+      "step": 15988
+    },
+    {
+      "epoch": 0.9835752952755905,
+      "grad_norm": 1.2145395278930664,
+      "learning_rate": 1.332609465083512e-07,
+      "loss": 2.6987,
+      "step": 15989
+    },
+    {
+      "epoch": 0.983636811023622,
+      "grad_norm": 1.2455953359603882,
+      "learning_rate": 1.3226482616999213e-07,
+      "loss": 2.2447,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9836983267716536,
+      "grad_norm": 1.3810456991195679,
+      "learning_rate": 1.312724403280563e-07,
+      "loss": 2.1623,
+      "step": 15991
+    },
+    {
+      "epoch": 0.9837598425196851,
+      "grad_norm": 1.339171051979065,
+      "learning_rate": 1.3028378901964733e-07,
+      "loss": 2.3625,
+      "step": 15992
+    },
+    {
+      "epoch": 0.9838213582677166,
+      "grad_norm": 1.4556241035461426,
+      "learning_rate": 1.2929887228172455e-07,
+      "loss": 2.6964,
+      "step": 15993
+    },
+    {
+      "epoch": 0.983882874015748,
+      "grad_norm": 1.2316755056381226,
+      "learning_rate": 1.2831769015114735e-07,
+      "loss": 2.5972,
+      "step": 15994
+    },
+    {
+      "epoch": 0.9839443897637795,
+      "grad_norm": 1.0659993886947632,
+      "learning_rate": 1.2734024266457533e-07,
+      "loss": 2.2434,
+      "step": 15995
+    },
+    {
+      "epoch": 0.984005905511811,
+      "grad_norm": 1.1520153284072876,
+      "learning_rate": 1.2636652985857922e-07,
+      "loss": 1.5313,
+      "step": 15996
+    },
+    {
+      "epoch": 0.9840674212598425,
+      "grad_norm": 1.284934401512146,
+      "learning_rate": 1.2539655176956321e-07,
+      "loss": 2.1832,
+      "step": 15997
+    },
+    {
+      "epoch": 0.984128937007874,
+      "grad_norm": 1.2974530458450317,
+      "learning_rate": 1.2443030843379833e-07,
+      "loss": 2.0515,
+      "step": 15998
+    },
+    {
+      "epoch": 0.9841904527559056,
+      "grad_norm": 1.3396075963974,
+      "learning_rate": 1.2346779988741118e-07,
+      "loss": 2.5285,
+      "step": 15999
+    },
+    {
+      "epoch": 0.984251968503937,
+      "grad_norm": 1.3877894878387451,
+      "learning_rate": 1.2250902616641745e-07,
+      "loss": 2.2622,
+      "step": 16000
+    },
+    {
+      "epoch": 0.9843134842519685,
+      "grad_norm": 1.2569154500961304,
+      "learning_rate": 1.2155398730664402e-07,
+      "loss": 2.8968,
+      "step": 16001
+    },
+    {
+      "epoch": 0.984375,
+      "grad_norm": 1.399823784828186,
+      "learning_rate": 1.2060268334381785e-07,
+      "loss": 2.4177,
+      "step": 16002
+    },
+    {
+      "epoch": 0.9844365157480315,
+      "grad_norm": 1.2548209428787231,
+      "learning_rate": 1.196551143134994e-07,
+      "loss": 2.5451,
+      "step": 16003
+    },
+    {
+      "epoch": 0.984498031496063,
+      "grad_norm": 1.3889284133911133,
+      "learning_rate": 1.1871128025113809e-07,
+      "loss": 2.6071,
+      "step": 16004
+    },
+    {
+      "epoch": 0.9845595472440944,
+      "grad_norm": 1.2751808166503906,
+      "learning_rate": 1.177711811920279e-07,
+      "loss": 2.4584,
+      "step": 16005
+    },
+    {
+      "epoch": 0.984621062992126,
+      "grad_norm": 1.3154929876327515,
+      "learning_rate": 1.168348171713074e-07,
+      "loss": 2.2394,
+      "step": 16006
+    },
+    {
+      "epoch": 0.9846825787401575,
+      "grad_norm": 1.0917940139770508,
+      "learning_rate": 1.1590218822400412e-07,
+      "loss": 3.0815,
+      "step": 16007
+    },
+    {
+      "epoch": 0.984744094488189,
+      "grad_norm": 1.2893391847610474,
+      "learning_rate": 1.1497329438499016e-07,
+      "loss": 2.5829,
+      "step": 16008
+    },
+    {
+      "epoch": 0.9848056102362205,
+      "grad_norm": 1.5119787454605103,
+      "learning_rate": 1.1404813568900441e-07,
+      "loss": 2.3208,
+      "step": 16009
+    },
+    {
+      "epoch": 0.984867125984252,
+      "grad_norm": 1.1267410516738892,
+      "learning_rate": 1.131267121706303e-07,
+      "loss": 2.9807,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9849286417322834,
+      "grad_norm": 1.2979251146316528,
+      "learning_rate": 1.1220902386434029e-07,
+      "loss": 2.5369,
+      "step": 16011
+    },
+    {
+      "epoch": 0.9849901574803149,
+      "grad_norm": 1.4502509832382202,
+      "learning_rate": 1.1129507080444023e-07,
+      "loss": 3.035,
+      "step": 16012
+    },
+    {
+      "epoch": 0.9850516732283464,
+      "grad_norm": 1.3267252445220947,
+      "learning_rate": 1.1038485302511392e-07,
+      "loss": 2.119,
+      "step": 16013
+    },
+    {
+      "epoch": 0.985113188976378,
+      "grad_norm": 1.179134726524353,
+      "learning_rate": 1.0947837056038968e-07,
+      "loss": 3.1038,
+      "step": 16014
+    },
+    {
+      "epoch": 0.9851747047244095,
+      "grad_norm": 1.2201045751571655,
+      "learning_rate": 1.0857562344417372e-07,
+      "loss": 2.2937,
+      "step": 16015
+    },
+    {
+      "epoch": 0.985236220472441,
+      "grad_norm": 1.5298445224761963,
+      "learning_rate": 1.0767661171021681e-07,
+      "loss": 2.2217,
+      "step": 16016
+    },
+    {
+      "epoch": 0.9852977362204725,
+      "grad_norm": 1.3166524171829224,
+      "learning_rate": 1.0678133539213653e-07,
+      "loss": 2.342,
+      "step": 16017
+    },
+    {
+      "epoch": 0.9853592519685039,
+      "grad_norm": 1.4243286848068237,
+      "learning_rate": 1.0588979452341719e-07,
+      "loss": 2.4259,
+      "step": 16018
+    },
+    {
+      "epoch": 0.9854207677165354,
+      "grad_norm": 1.3094757795333862,
+      "learning_rate": 1.0500198913739878e-07,
+      "loss": 2.5639,
+      "step": 16019
+    },
+    {
+      "epoch": 0.9854822834645669,
+      "grad_norm": 1.2670255899429321,
+      "learning_rate": 1.0411791926727699e-07,
+      "loss": 2.1349,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9855437992125984,
+      "grad_norm": 1.2946134805679321,
+      "learning_rate": 1.0323758494610314e-07,
+      "loss": 2.8823,
+      "step": 16021
+    },
+    {
+      "epoch": 0.98560531496063,
+      "grad_norm": 1.2590707540512085,
+      "learning_rate": 1.0236098620680646e-07,
+      "loss": 2.5834,
+      "step": 16022
+    },
+    {
+      "epoch": 0.9856668307086615,
+      "grad_norm": 1.225516676902771,
+      "learning_rate": 1.0148812308216071e-07,
+      "loss": 2.5146,
+      "step": 16023
+    },
+    {
+      "epoch": 0.9857283464566929,
+      "grad_norm": 1.490442156791687,
+      "learning_rate": 1.0061899560481758e-07,
+      "loss": 2.2146,
+      "step": 16024
+    },
+    {
+      "epoch": 0.9857898622047244,
+      "grad_norm": 1.3249846696853638,
+      "learning_rate": 9.975360380726218e-08,
+      "loss": 2.1775,
+      "step": 16025
+    },
+    {
+      "epoch": 0.9858513779527559,
+      "grad_norm": 1.1188991069793701,
+      "learning_rate": 9.889194772186861e-08,
+      "loss": 2.3977,
+      "step": 16026
+    },
+    {
+      "epoch": 0.9859128937007874,
+      "grad_norm": 1.2425357103347778,
+      "learning_rate": 9.803402738085555e-08,
+      "loss": 2.3081,
+      "step": 16027
+    },
+    {
+      "epoch": 0.9859744094488189,
+      "grad_norm": 1.4566948413848877,
+      "learning_rate": 9.717984281629733e-08,
+      "loss": 2.3842,
+      "step": 16028
+    },
+    {
+      "epoch": 0.9860359251968503,
+      "grad_norm": 1.5744017362594604,
+      "learning_rate": 9.632939406013508e-08,
+      "loss": 2.3431,
+      "step": 16029
+    },
+    {
+      "epoch": 0.9860974409448819,
+      "grad_norm": 1.6948823928833008,
+      "learning_rate": 9.548268114417668e-08,
+      "loss": 2.6354,
+      "step": 16030
+    },
+    {
+      "epoch": 0.9861589566929134,
+      "grad_norm": 1.3400027751922607,
+      "learning_rate": 9.463970410008571e-08,
+      "loss": 1.7408,
+      "step": 16031
+    },
+    {
+      "epoch": 0.9862204724409449,
+      "grad_norm": 1.2109706401824951,
+      "learning_rate": 9.380046295938139e-08,
+      "loss": 2.2136,
+      "step": 16032
+    },
+    {
+      "epoch": 0.9862819881889764,
+      "grad_norm": 1.137123465538025,
+      "learning_rate": 9.29649577534497e-08,
+      "loss": 2.8654,
+      "step": 16033
+    },
+    {
+      "epoch": 0.9863435039370079,
+      "grad_norm": 1.1907649040222168,
+      "learning_rate": 9.213318851353236e-08,
+      "loss": 2.6868,
+      "step": 16034
+    },
+    {
+      "epoch": 0.9864050196850394,
+      "grad_norm": 1.3360350131988525,
+      "learning_rate": 9.130515527072669e-08,
+      "loss": 2.1621,
+      "step": 16035
+    },
+    {
+      "epoch": 0.9864665354330708,
+      "grad_norm": 1.3369245529174805,
+      "learning_rate": 9.048085805600793e-08,
+      "loss": 2.5548,
+      "step": 16036
+    },
+    {
+      "epoch": 0.9865280511811023,
+      "grad_norm": 1.534206748008728,
+      "learning_rate": 8.966029690018474e-08,
+      "loss": 1.6238,
+      "step": 16037
+    },
+    {
+      "epoch": 0.9865895669291339,
+      "grad_norm": 1.5774458646774292,
+      "learning_rate": 8.884347183396591e-08,
+      "loss": 1.8665,
+      "step": 16038
+    },
+    {
+      "epoch": 0.9866510826771654,
+      "grad_norm": 1.21379554271698,
+      "learning_rate": 8.803038288787146e-08,
+      "loss": 2.5592,
+      "step": 16039
+    },
+    {
+      "epoch": 0.9867125984251969,
+      "grad_norm": 1.2384052276611328,
+      "learning_rate": 8.722103009232152e-08,
+      "loss": 2.3066,
+      "step": 16040
+    },
+    {
+      "epoch": 0.9867741141732284,
+      "grad_norm": 1.3460220098495483,
+      "learning_rate": 8.641541347756966e-08,
+      "loss": 2.3691,
+      "step": 16041
+    },
+    {
+      "epoch": 0.9868356299212598,
+      "grad_norm": 1.2880373001098633,
+      "learning_rate": 8.561353307374731e-08,
+      "loss": 2.3721,
+      "step": 16042
+    },
+    {
+      "epoch": 0.9868971456692913,
+      "grad_norm": 1.3671026229858398,
+      "learning_rate": 8.481538891084162e-08,
+      "loss": 2.3201,
+      "step": 16043
+    },
+    {
+      "epoch": 0.9869586614173228,
+      "grad_norm": 1.2889901399612427,
+      "learning_rate": 8.402098101870648e-08,
+      "loss": 2.3709,
+      "step": 16044
+    },
+    {
+      "epoch": 0.9870201771653543,
+      "grad_norm": 1.5192277431488037,
+      "learning_rate": 8.323030942702925e-08,
+      "loss": 3.3849,
+      "step": 16045
+    },
+    {
+      "epoch": 0.9870816929133859,
+      "grad_norm": 1.3334119319915771,
+      "learning_rate": 8.244337416539738e-08,
+      "loss": 2.9939,
+      "step": 16046
+    },
+    {
+      "epoch": 0.9871432086614174,
+      "grad_norm": 1.4060015678405762,
+      "learning_rate": 8.166017526322067e-08,
+      "loss": 2.1959,
+      "step": 16047
+    },
+    {
+      "epoch": 0.9872047244094488,
+      "grad_norm": 1.193689227104187,
+      "learning_rate": 8.088071274979791e-08,
+      "loss": 1.7845,
+      "step": 16048
+    },
+    {
+      "epoch": 0.9872662401574803,
+      "grad_norm": 1.2801579236984253,
+      "learning_rate": 8.010498665427246e-08,
+      "loss": 2.1863,
+      "step": 16049
+    },
+    {
+      "epoch": 0.9873277559055118,
+      "grad_norm": 1.127291202545166,
+      "learning_rate": 7.933299700564333e-08,
+      "loss": 3.1105,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9873892716535433,
+      "grad_norm": 1.288087248802185,
+      "learning_rate": 7.856474383279856e-08,
+      "loss": 2.4586,
+      "step": 16051
+    },
+    {
+      "epoch": 0.9874507874015748,
+      "grad_norm": 1.390275239944458,
+      "learning_rate": 7.780022716444845e-08,
+      "loss": 2.8827,
+      "step": 16052
+    },
+    {
+      "epoch": 0.9875123031496063,
+      "grad_norm": 1.2848610877990723,
+      "learning_rate": 7.703944702919241e-08,
+      "loss": 2.0975,
+      "step": 16053
+    },
+    {
+      "epoch": 0.9875738188976378,
+      "grad_norm": 1.233802080154419,
+      "learning_rate": 7.628240345546322e-08,
+      "loss": 2.9106,
+      "step": 16054
+    },
+    {
+      "epoch": 0.9876353346456693,
+      "grad_norm": 1.557951807975769,
+      "learning_rate": 7.552909647159379e-08,
+      "loss": 2.0913,
+      "step": 16055
+    },
+    {
+      "epoch": 0.9876968503937008,
+      "grad_norm": 1.3015267848968506,
+      "learning_rate": 7.477952610573934e-08,
+      "loss": 2.4682,
+      "step": 16056
+    },
+    {
+      "epoch": 0.9877583661417323,
+      "grad_norm": 1.1892119646072388,
+      "learning_rate": 7.403369238593305e-08,
+      "loss": 2.7316,
+      "step": 16057
+    },
+    {
+      "epoch": 0.9878198818897638,
+      "grad_norm": 1.4115209579467773,
+      "learning_rate": 7.32915953400637e-08,
+      "loss": 2.7424,
+      "step": 16058
+    },
+    {
+      "epoch": 0.9878813976377953,
+      "grad_norm": 2.5708963871002197,
+      "learning_rate": 7.255323499588684e-08,
+      "loss": 1.7809,
+      "step": 16059
+    },
+    {
+      "epoch": 0.9879429133858267,
+      "grad_norm": 1.2964695692062378,
+      "learning_rate": 7.181861138100265e-08,
+      "loss": 2.1612,
+      "step": 16060
+    },
+    {
+      "epoch": 0.9880044291338582,
+      "grad_norm": 1.2159690856933594,
+      "learning_rate": 7.108772452288915e-08,
+      "loss": 2.8399,
+      "step": 16061
+    },
+    {
+      "epoch": 0.9880659448818898,
+      "grad_norm": 1.4144372940063477,
+      "learning_rate": 7.03605744488689e-08,
+      "loss": 2.6321,
+      "step": 16062
+    },
+    {
+      "epoch": 0.9881274606299213,
+      "grad_norm": 1.5908268690109253,
+      "learning_rate": 6.963716118615348e-08,
+      "loss": 2.5881,
+      "step": 16063
+    },
+    {
+      "epoch": 0.9881889763779528,
+      "grad_norm": 1.3019975423812866,
+      "learning_rate": 6.891748476177683e-08,
+      "loss": 2.4961,
+      "step": 16064
+    },
+    {
+      "epoch": 0.9882504921259843,
+      "grad_norm": 1.2386705875396729,
+      "learning_rate": 6.820154520265076e-08,
+      "loss": 2.7417,
+      "step": 16065
+    },
+    {
+      "epoch": 0.9883120078740157,
+      "grad_norm": 1.4610657691955566,
+      "learning_rate": 6.748934253555384e-08,
+      "loss": 2.0964,
+      "step": 16066
+    },
+    {
+      "epoch": 0.9883735236220472,
+      "grad_norm": 1.4129503965377808,
+      "learning_rate": 6.67808767871203e-08,
+      "loss": 2.1991,
+      "step": 16067
+    },
+    {
+      "epoch": 0.9884350393700787,
+      "grad_norm": 1.179695725440979,
+      "learning_rate": 6.6076147983829e-08,
+      "loss": 2.5806,
+      "step": 16068
+    },
+    {
+      "epoch": 0.9884965551181102,
+      "grad_norm": 1.4267154932022095,
+      "learning_rate": 6.53751561520477e-08,
+      "loss": 2.0853,
+      "step": 16069
+    },
+    {
+      "epoch": 0.9885580708661418,
+      "grad_norm": 1.2209826707839966,
+      "learning_rate": 6.467790131798878e-08,
+      "loss": 2.1956,
+      "step": 16070
+    },
+    {
+      "epoch": 0.9886195866141733,
+      "grad_norm": 1.5524290800094604,
+      "learning_rate": 6.39843835077203e-08,
+      "loss": 1.8949,
+      "step": 16071
+    },
+    {
+      "epoch": 0.9886811023622047,
+      "grad_norm": 1.215566873550415,
+      "learning_rate": 6.329460274717702e-08,
+      "loss": 3.1553,
+      "step": 16072
+    },
+    {
+      "epoch": 0.9887426181102362,
+      "grad_norm": 1.2435275316238403,
+      "learning_rate": 6.260855906214946e-08,
+      "loss": 2.5148,
+      "step": 16073
+    },
+    {
+      "epoch": 0.9888041338582677,
+      "grad_norm": 1.3007451295852661,
+      "learning_rate": 6.192625247830596e-08,
+      "loss": 1.98,
+      "step": 16074
+    },
+    {
+      "epoch": 0.9888656496062992,
+      "grad_norm": 1.2617827653884888,
+      "learning_rate": 6.124768302113725e-08,
+      "loss": 2.2429,
+      "step": 16075
+    },
+    {
+      "epoch": 0.9889271653543307,
+      "grad_norm": 1.2708863019943237,
+      "learning_rate": 6.057285071604524e-08,
+      "loss": 2.0602,
+      "step": 16076
+    },
+    {
+      "epoch": 0.9889886811023622,
+      "grad_norm": 1.4186418056488037,
+      "learning_rate": 5.990175558824307e-08,
+      "loss": 2.4272,
+      "step": 16077
+    },
+    {
+      "epoch": 0.9890501968503937,
+      "grad_norm": 1.3406007289886475,
+      "learning_rate": 5.92343976628329e-08,
+      "loss": 2.3763,
+      "step": 16078
+    },
+    {
+      "epoch": 0.9891117125984252,
+      "grad_norm": 1.3506231307983398,
+      "learning_rate": 5.8570776964772536e-08,
+      "loss": 2.4947,
+      "step": 16079
+    },
+    {
+      "epoch": 0.9891732283464567,
+      "grad_norm": 1.2809171676635742,
+      "learning_rate": 5.7910893518875467e-08,
+      "loss": 2.2025,
+      "step": 16080
+    },
+    {
+      "epoch": 0.9892347440944882,
+      "grad_norm": 1.3396687507629395,
+      "learning_rate": 5.725474734982195e-08,
+      "loss": 2.4032,
+      "step": 16081
+    },
+    {
+      "epoch": 0.9892962598425197,
+      "grad_norm": 1.4260282516479492,
+      "learning_rate": 5.6602338482136806e-08,
+      "loss": 2.3881,
+      "step": 16082
+    },
+    {
+      "epoch": 0.9893577755905512,
+      "grad_norm": 1.5706126689910889,
+      "learning_rate": 5.5953666940233854e-08,
+      "loss": 2.5073,
+      "step": 16083
+    },
+    {
+      "epoch": 0.9894192913385826,
+      "grad_norm": 1.191962718963623,
+      "learning_rate": 5.530873274836035e-08,
+      "loss": 2.6832,
+      "step": 16084
+    },
+    {
+      "epoch": 0.9894808070866141,
+      "grad_norm": 1.3756036758422852,
+      "learning_rate": 5.4667535930630345e-08,
+      "loss": 2.4535,
+      "step": 16085
+    },
+    {
+      "epoch": 0.9895423228346457,
+      "grad_norm": 1.4520944356918335,
+      "learning_rate": 5.403007651102465e-08,
+      "loss": 2.5369,
+      "step": 16086
+    },
+    {
+      "epoch": 0.9896038385826772,
+      "grad_norm": 1.3959380388259888,
+      "learning_rate": 5.3396354513379765e-08,
+      "loss": 2.424,
+      "step": 16087
+    },
+    {
+      "epoch": 0.9896653543307087,
+      "grad_norm": 1.2744417190551758,
+      "learning_rate": 5.276636996138784e-08,
+      "loss": 2.1669,
+      "step": 16088
+    },
+    {
+      "epoch": 0.9897268700787402,
+      "grad_norm": 1.3946237564086914,
+      "learning_rate": 5.214012287861891e-08,
+      "loss": 2.2291,
+      "step": 16089
+    },
+    {
+      "epoch": 0.9897883858267716,
+      "grad_norm": 1.1781197786331177,
+      "learning_rate": 5.151761328848759e-08,
+      "loss": 2.721,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9898499015748031,
+      "grad_norm": 1.274348497390747,
+      "learning_rate": 5.0898841214264134e-08,
+      "loss": 1.8409,
+      "step": 16091
+    },
+    {
+      "epoch": 0.9899114173228346,
+      "grad_norm": 1.6762739419937134,
+      "learning_rate": 5.0283806679096714e-08,
+      "loss": 2.5009,
+      "step": 16092
+    },
+    {
+      "epoch": 0.9899729330708661,
+      "grad_norm": 1.2039622068405151,
+      "learning_rate": 4.967250970597803e-08,
+      "loss": 2.5567,
+      "step": 16093
+    },
+    {
+      "epoch": 0.9900344488188977,
+      "grad_norm": 1.2219923734664917,
+      "learning_rate": 4.906495031776759e-08,
+      "loss": 1.8239,
+      "step": 16094
+    },
+    {
+      "epoch": 0.9900959645669292,
+      "grad_norm": 1.3085881471633911,
+      "learning_rate": 4.846112853718055e-08,
+      "loss": 1.5162,
+      "step": 16095
+    },
+    {
+      "epoch": 0.9901574803149606,
+      "grad_norm": 1.2277593612670898,
+      "learning_rate": 4.786104438680994e-08,
+      "loss": 3.2804,
+      "step": 16096
+    },
+    {
+      "epoch": 0.9902189960629921,
+      "grad_norm": 1.2673791646957397,
+      "learning_rate": 4.726469788909338e-08,
+      "loss": 2.3187,
+      "step": 16097
+    },
+    {
+      "epoch": 0.9902805118110236,
+      "grad_norm": 1.1675165891647339,
+      "learning_rate": 4.667208906631304e-08,
+      "loss": 2.5876,
+      "step": 16098
+    },
+    {
+      "epoch": 0.9903420275590551,
+      "grad_norm": 1.1706905364990234,
+      "learning_rate": 4.608321794065118e-08,
+      "loss": 2.4598,
+      "step": 16099
+    },
+    {
+      "epoch": 0.9904035433070866,
+      "grad_norm": 1.4808545112609863,
+      "learning_rate": 4.5498084534112416e-08,
+      "loss": 2.5707,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9904650590551181,
+      "grad_norm": 1.3952466249465942,
+      "learning_rate": 4.491668886859035e-08,
+      "loss": 2.5183,
+      "step": 16101
+    },
+    {
+      "epoch": 0.9905265748031497,
+      "grad_norm": 1.214261770248413,
+      "learning_rate": 4.433903096581205e-08,
+      "loss": 2.3013,
+      "step": 16102
+    },
+    {
+      "epoch": 0.9905880905511811,
+      "grad_norm": 1.3719124794006348,
+      "learning_rate": 4.376511084739354e-08,
+      "loss": 2.3739,
+      "step": 16103
+    },
+    {
+      "epoch": 0.9906496062992126,
+      "grad_norm": 1.2257877588272095,
+      "learning_rate": 4.319492853478435e-08,
+      "loss": 2.4899,
+      "step": 16104
+    },
+    {
+      "epoch": 0.9907111220472441,
+      "grad_norm": 1.2540814876556396,
+      "learning_rate": 4.2628484049311854e-08,
+      "loss": 2.1685,
+      "step": 16105
+    },
+    {
+      "epoch": 0.9907726377952756,
+      "grad_norm": 1.4710447788238525,
+      "learning_rate": 4.2065777412148014e-08,
+      "loss": 2.7428,
+      "step": 16106
+    },
+    {
+      "epoch": 0.9908341535433071,
+      "grad_norm": 1.372375249862671,
+      "learning_rate": 4.1506808644353746e-08,
+      "loss": 2.9355,
+      "step": 16107
+    },
+    {
+      "epoch": 0.9908956692913385,
+      "grad_norm": 1.855323314666748,
+      "learning_rate": 4.095157776681235e-08,
+      "loss": 3.3641,
+      "step": 16108
+    },
+    {
+      "epoch": 0.99095718503937,
+      "grad_norm": 1.307522177696228,
+      "learning_rate": 4.040008480030721e-08,
+      "loss": 2.7549,
+      "step": 16109
+    },
+    {
+      "epoch": 0.9910187007874016,
+      "grad_norm": 1.1822417974472046,
+      "learning_rate": 3.985232976544406e-08,
+      "loss": 2.8402,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9910802165354331,
+      "grad_norm": 1.3645930290222168,
+      "learning_rate": 3.9308312682706515e-08,
+      "loss": 2.9684,
+      "step": 16111
+    },
+    {
+      "epoch": 0.9911417322834646,
+      "grad_norm": 1.4104423522949219,
+      "learning_rate": 3.8768033572444964e-08,
+      "loss": 2.4163,
+      "step": 16112
+    },
+    {
+      "epoch": 0.9912032480314961,
+      "grad_norm": 1.1773960590362549,
+      "learning_rate": 3.8231492454854355e-08,
+      "loss": 1.8506,
+      "step": 16113
+    },
+    {
+      "epoch": 0.9912647637795275,
+      "grad_norm": 1.3912580013275146,
+      "learning_rate": 3.769868935000753e-08,
+      "loss": 2.4865,
+      "step": 16114
+    },
+    {
+      "epoch": 0.991326279527559,
+      "grad_norm": 1.3015347719192505,
+      "learning_rate": 3.7169624277832995e-08,
+      "loss": 2.7151,
+      "step": 16115
+    },
+    {
+      "epoch": 0.9913877952755905,
+      "grad_norm": 1.179495096206665,
+      "learning_rate": 3.6644297258092705e-08,
+      "loss": 2.606,
+      "step": 16116
+    },
+    {
+      "epoch": 0.991449311023622,
+      "grad_norm": 1.3827400207519531,
+      "learning_rate": 3.612270831044873e-08,
+      "loss": 2.6788,
+      "step": 16117
+    },
+    {
+      "epoch": 0.9915108267716536,
+      "grad_norm": 1.1779468059539795,
+      "learning_rate": 3.560485745440767e-08,
+      "loss": 2.1559,
+      "step": 16118
+    },
+    {
+      "epoch": 0.9915723425196851,
+      "grad_norm": 1.289240837097168,
+      "learning_rate": 3.509074470932072e-08,
+      "loss": 2.7661,
+      "step": 16119
+    },
+    {
+      "epoch": 0.9916338582677166,
+      "grad_norm": 1.3787519931793213,
+      "learning_rate": 3.458037009442805e-08,
+      "loss": 2.2919,
+      "step": 16120
+    },
+    {
+      "epoch": 0.991695374015748,
+      "grad_norm": 1.351762294769287,
+      "learning_rate": 3.407373362880328e-08,
+      "loss": 2.2883,
+      "step": 16121
+    },
+    {
+      "epoch": 0.9917568897637795,
+      "grad_norm": 1.3389158248901367,
+      "learning_rate": 3.357083533139793e-08,
+      "loss": 1.8806,
+      "step": 16122
+    },
+    {
+      "epoch": 0.991818405511811,
+      "grad_norm": 1.1809654235839844,
+      "learning_rate": 3.307167522100807e-08,
+      "loss": 2.8622,
+      "step": 16123
+    },
+    {
+      "epoch": 0.9918799212598425,
+      "grad_norm": 1.2415204048156738,
+      "learning_rate": 3.257625331630765e-08,
+      "loss": 2.4372,
+      "step": 16124
+    },
+    {
+      "epoch": 0.991941437007874,
+      "grad_norm": 1.3878473043441772,
+      "learning_rate": 3.208456963582629e-08,
+      "loss": 2.3651,
+      "step": 16125
+    },
+    {
+      "epoch": 0.9920029527559056,
+      "grad_norm": 1.1555670499801636,
+      "learning_rate": 3.1596624197949285e-08,
+      "loss": 2.3091,
+      "step": 16126
+    },
+    {
+      "epoch": 0.992064468503937,
+      "grad_norm": 1.3836883306503296,
+      "learning_rate": 3.11124170209176e-08,
+      "loss": 2.2683,
+      "step": 16127
+    },
+    {
+      "epoch": 0.9921259842519685,
+      "grad_norm": 1.2736722230911255,
+      "learning_rate": 3.063194812282788e-08,
+      "loss": 1.8699,
+      "step": 16128
+    },
+    {
+      "epoch": 0.9921875,
+      "grad_norm": 1.2867659330368042,
+      "learning_rate": 3.0155217521676826e-08,
+      "loss": 2.3699,
+      "step": 16129
+    },
+    {
+      "epoch": 0.9922490157480315,
+      "grad_norm": 1.235126256942749,
+      "learning_rate": 2.968222523526132e-08,
+      "loss": 2.9219,
+      "step": 16130
+    },
+    {
+      "epoch": 0.992310531496063,
+      "grad_norm": 1.250683307647705,
+      "learning_rate": 2.9212971281278313e-08,
+      "loss": 2.5925,
+      "step": 16131
+    },
+    {
+      "epoch": 0.9923720472440944,
+      "grad_norm": 1.7160577774047852,
+      "learning_rate": 2.8747455677291534e-08,
+      "loss": 1.9651,
+      "step": 16132
+    },
+    {
+      "epoch": 0.992433562992126,
+      "grad_norm": 1.5094026327133179,
+      "learning_rate": 2.828567844068708e-08,
+      "loss": 2.3161,
+      "step": 16133
+    },
+    {
+      "epoch": 0.9924950787401575,
+      "grad_norm": 1.2316983938217163,
+      "learning_rate": 2.7827639588740018e-08,
+      "loss": 2.3002,
+      "step": 16134
+    },
+    {
+      "epoch": 0.992556594488189,
+      "grad_norm": 1.4474190473556519,
+      "learning_rate": 2.737333913858109e-08,
+      "loss": 2.5084,
+      "step": 16135
+    },
+    {
+      "epoch": 0.9926181102362205,
+      "grad_norm": 1.2941186428070068,
+      "learning_rate": 2.6922777107185605e-08,
+      "loss": 2.3728,
+      "step": 16136
+    },
+    {
+      "epoch": 0.992679625984252,
+      "grad_norm": 1.1684719324111938,
+      "learning_rate": 2.647595351142895e-08,
+      "loss": 2.8692,
+      "step": 16137
+    },
+    {
+      "epoch": 0.9927411417322834,
+      "grad_norm": 1.2455724477767944,
+      "learning_rate": 2.6032868367997788e-08,
+      "loss": 2.3148,
+      "step": 16138
+    },
+    {
+      "epoch": 0.9928026574803149,
+      "grad_norm": 1.6542049646377563,
+      "learning_rate": 2.5593521693467737e-08,
+      "loss": 2.2023,
+      "step": 16139
+    },
+    {
+      "epoch": 0.9928641732283464,
+      "grad_norm": 1.3546415567398071,
+      "learning_rate": 2.5157913504259e-08,
+      "loss": 2.8442,
+      "step": 16140
+    },
+    {
+      "epoch": 0.992925688976378,
+      "grad_norm": 1.2493747472763062,
+      "learning_rate": 2.4726043816680755e-08,
+      "loss": 2.4843,
+      "step": 16141
+    },
+    {
+      "epoch": 0.9929872047244095,
+      "grad_norm": 1.6142423152923584,
+      "learning_rate": 2.429791264686454e-08,
+      "loss": 1.8975,
+      "step": 16142
+    },
+    {
+      "epoch": 0.993048720472441,
+      "grad_norm": 1.164808988571167,
+      "learning_rate": 2.3873520010830873e-08,
+      "loss": 2.4825,
+      "step": 16143
+    },
+    {
+      "epoch": 0.9931102362204725,
+      "grad_norm": 1.1936159133911133,
+      "learning_rate": 2.3452865924444845e-08,
+      "loss": 2.7341,
+      "step": 16144
+    },
+    {
+      "epoch": 0.9931717519685039,
+      "grad_norm": 1.442173957824707,
+      "learning_rate": 2.303595040343831e-08,
+      "loss": 1.885,
+      "step": 16145
+    },
+    {
+      "epoch": 0.9932332677165354,
+      "grad_norm": 1.4827704429626465,
+      "learning_rate": 2.2622773463398805e-08,
+      "loss": 2.5285,
+      "step": 16146
+    },
+    {
+      "epoch": 0.9932947834645669,
+      "grad_norm": 1.5850555896759033,
+      "learning_rate": 2.2213335119780632e-08,
+      "loss": 2.614,
+      "step": 16147
+    },
+    {
+      "epoch": 0.9933562992125984,
+      "grad_norm": 1.3055200576782227,
+      "learning_rate": 2.1807635387882663e-08,
+      "loss": 1.9057,
+      "step": 16148
+    },
+    {
+      "epoch": 0.99341781496063,
+      "grad_norm": 1.1458539962768555,
+      "learning_rate": 2.140567428289275e-08,
+      "loss": 1.9522,
+      "step": 16149
+    },
+    {
+      "epoch": 0.9934793307086615,
+      "grad_norm": 1.2515246868133545,
+      "learning_rate": 2.1007451819832214e-08,
+      "loss": 2.4292,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9935408464566929,
+      "grad_norm": 1.261828064918518,
+      "learning_rate": 2.0612968013600243e-08,
+      "loss": 2.5662,
+      "step": 16151
+    },
+    {
+      "epoch": 0.9936023622047244,
+      "grad_norm": 1.209226369857788,
+      "learning_rate": 2.02222228789406e-08,
+      "loss": 2.2577,
+      "step": 16152
+    },
+    {
+      "epoch": 0.9936638779527559,
+      "grad_norm": 1.2197949886322021,
+      "learning_rate": 1.9835216430463823e-08,
+      "loss": 3.0379,
+      "step": 16153
+    },
+    {
+      "epoch": 0.9937253937007874,
+      "grad_norm": 1.1909258365631104,
+      "learning_rate": 1.9451948682647213e-08,
+      "loss": 2.486,
+      "step": 16154
+    },
+    {
+      "epoch": 0.9937869094488189,
+      "grad_norm": 1.441320776939392,
+      "learning_rate": 1.9072419649823757e-08,
+      "loss": 2.2896,
+      "step": 16155
+    },
+    {
+      "epoch": 0.9938484251968503,
+      "grad_norm": 1.308146357536316,
+      "learning_rate": 1.8696629346182104e-08,
+      "loss": 2.8939,
+      "step": 16156
+    },
+    {
+      "epoch": 0.9939099409448819,
+      "grad_norm": 1.1508235931396484,
+      "learning_rate": 1.8324577785777673e-08,
+      "loss": 2.821,
+      "step": 16157
+    },
+    {
+      "epoch": 0.9939714566929134,
+      "grad_norm": 1.5300345420837402,
+      "learning_rate": 1.795626498251046e-08,
+      "loss": 2.1814,
+      "step": 16158
+    },
+    {
+      "epoch": 0.9940329724409449,
+      "grad_norm": 1.2314157485961914,
+      "learning_rate": 1.7591690950169436e-08,
+      "loss": 1.9143,
+      "step": 16159
+    },
+    {
+      "epoch": 0.9940944881889764,
+      "grad_norm": 1.1897647380828857,
+      "learning_rate": 1.723085570238814e-08,
+      "loss": 1.7678,
+      "step": 16160
+    },
+    {
+      "epoch": 0.9941560039370079,
+      "grad_norm": 1.3295621871948242,
+      "learning_rate": 1.687375925264467e-08,
+      "loss": 1.9412,
+      "step": 16161
+    },
+    {
+      "epoch": 0.9942175196850394,
+      "grad_norm": 1.3653603792190552,
+      "learning_rate": 1.6520401614306126e-08,
+      "loss": 2.4934,
+      "step": 16162
+    },
+    {
+      "epoch": 0.9942790354330708,
+      "grad_norm": 1.2734991312026978,
+      "learning_rate": 1.617078280058415e-08,
+      "loss": 2.5606,
+      "step": 16163
+    },
+    {
+      "epoch": 0.9943405511811023,
+      "grad_norm": 1.563109040260315,
+      "learning_rate": 1.5824902824546072e-08,
+      "loss": 2.0279,
+      "step": 16164
+    },
+    {
+      "epoch": 0.9944020669291339,
+      "grad_norm": 1.0039583444595337,
+      "learning_rate": 1.5482761699125993e-08,
+      "loss": 2.448,
+      "step": 16165
+    },
+    {
+      "epoch": 0.9944635826771654,
+      "grad_norm": 1.2287806272506714,
+      "learning_rate": 1.5144359437124778e-08,
+      "loss": 2.4984,
+      "step": 16166
+    },
+    {
+      "epoch": 0.9945250984251969,
+      "grad_norm": 1.5290900468826294,
+      "learning_rate": 1.4809696051198973e-08,
+      "loss": 2.0429,
+      "step": 16167
+    },
+    {
+      "epoch": 0.9945866141732284,
+      "grad_norm": 1.5401517152786255,
+      "learning_rate": 1.4478771553849691e-08,
+      "loss": 2.1211,
+      "step": 16168
+    },
+    {
+      "epoch": 0.9946481299212598,
+      "grad_norm": 1.3851276636123657,
+      "learning_rate": 1.4151585957455914e-08,
+      "loss": 2.0488,
+      "step": 16169
+    },
+    {
+      "epoch": 0.9947096456692913,
+      "grad_norm": 1.2631639242172241,
+      "learning_rate": 1.3828139274263408e-08,
+      "loss": 2.7858,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9947711614173228,
+      "grad_norm": 1.1703933477401733,
+      "learning_rate": 1.3508431516362496e-08,
+      "loss": 2.34,
+      "step": 16171
+    },
+    {
+      "epoch": 0.9948326771653543,
+      "grad_norm": 1.2917424440383911,
+      "learning_rate": 1.3192462695710283e-08,
+      "loss": 2.3181,
+      "step": 16172
+    },
+    {
+      "epoch": 0.9948941929133859,
+      "grad_norm": 1.2497583627700806,
+      "learning_rate": 1.288023282410844e-08,
+      "loss": 2.5386,
+      "step": 16173
+    },
+    {
+      "epoch": 0.9949557086614174,
+      "grad_norm": 1.1946818828582764,
+      "learning_rate": 1.2571741913247614e-08,
+      "loss": 2.1753,
+      "step": 16174
+    },
+    {
+      "epoch": 0.9950172244094488,
+      "grad_norm": 1.2874400615692139,
+      "learning_rate": 1.2266989974651922e-08,
+      "loss": 2.1093,
+      "step": 16175
+    },
+    {
+      "epoch": 0.9950787401574803,
+      "grad_norm": 1.2245092391967773,
+      "learning_rate": 1.1965977019734454e-08,
+      "loss": 2.6949,
+      "step": 16176
+    },
+    {
+      "epoch": 0.9951402559055118,
+      "grad_norm": 1.0957703590393066,
+      "learning_rate": 1.1668703059730667e-08,
+      "loss": 2.8198,
+      "step": 16177
+    },
+    {
+      "epoch": 0.9952017716535433,
+      "grad_norm": 1.7793428897857666,
+      "learning_rate": 1.1375168105776101e-08,
+      "loss": 2.649,
+      "step": 16178
+    },
+    {
+      "epoch": 0.9952632874015748,
+      "grad_norm": 1.3644343614578247,
+      "learning_rate": 1.1085372168828656e-08,
+      "loss": 2.1677,
+      "step": 16179
+    },
+    {
+      "epoch": 0.9953248031496063,
+      "grad_norm": 1.33750319480896,
+      "learning_rate": 1.0799315259746312e-08,
+      "loss": 2.5238,
+      "step": 16180
+    },
+    {
+      "epoch": 0.9953863188976378,
+      "grad_norm": 1.3272625207901,
+      "learning_rate": 1.0516997389209416e-08,
+      "loss": 2.5607,
+      "step": 16181
+    },
+    {
+      "epoch": 0.9954478346456693,
+      "grad_norm": 1.2352735996246338,
+      "learning_rate": 1.0238418567776186e-08,
+      "loss": 2.3914,
+      "step": 16182
+    },
+    {
+      "epoch": 0.9955093503937008,
+      "grad_norm": 1.4413154125213623,
+      "learning_rate": 9.963578805871621e-09,
+      "loss": 2.8642,
+      "step": 16183
+    },
+    {
+      "epoch": 0.9955708661417323,
+      "grad_norm": 1.1157711744308472,
+      "learning_rate": 9.69247811376528e-09,
+      "loss": 2.3471,
+      "step": 16184
+    },
+    {
+      "epoch": 0.9956323818897638,
+      "grad_norm": 1.2960158586502075,
+      "learning_rate": 9.425116501604602e-09,
+      "loss": 2.7801,
+      "step": 16185
+    },
+    {
+      "epoch": 0.9956938976377953,
+      "grad_norm": 1.1165038347244263,
+      "learning_rate": 9.161493979381597e-09,
+      "loss": 2.2281,
+      "step": 16186
+    },
+    {
+      "epoch": 0.9957554133858267,
+      "grad_norm": 1.705283761024475,
+      "learning_rate": 8.901610556955042e-09,
+      "loss": 2.0827,
+      "step": 16187
+    },
+    {
+      "epoch": 0.9958169291338582,
+      "grad_norm": 1.2364954948425293,
+      "learning_rate": 8.64546624403939e-09,
+      "loss": 2.3371,
+      "step": 16188
+    },
+    {
+      "epoch": 0.9958784448818898,
+      "grad_norm": 1.1396315097808838,
+      "learning_rate": 8.393061050215868e-09,
+      "loss": 2.1704,
+      "step": 16189
+    },
+    {
+      "epoch": 0.9959399606299213,
+      "grad_norm": 1.3961118459701538,
+      "learning_rate": 8.144394984921366e-09,
+      "loss": 2.1539,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9960014763779528,
+      "grad_norm": 1.303037405014038,
+      "learning_rate": 7.89946805747066e-09,
+      "loss": 2.43,
+      "step": 16191
+    },
+    {
+      "epoch": 0.9960629921259843,
+      "grad_norm": 1.2882696390151978,
+      "learning_rate": 7.658280276989782e-09,
+      "loss": 2.9118,
+      "step": 16192
+    },
+    {
+      "epoch": 0.9961245078740157,
+      "grad_norm": 1.3498719930648804,
+      "learning_rate": 7.420831652527049e-09,
+      "loss": 2.0685,
+      "step": 16193
+    },
+    {
+      "epoch": 0.9961860236220472,
+      "grad_norm": 1.4371861219406128,
+      "learning_rate": 7.187122192953144e-09,
+      "loss": 1.704,
+      "step": 16194
+    },
+    {
+      "epoch": 0.9962475393700787,
+      "grad_norm": 1.2344551086425781,
+      "learning_rate": 6.957151906994419e-09,
+      "loss": 3.0795,
+      "step": 16195
+    },
+    {
+      "epoch": 0.9963090551181102,
+      "grad_norm": 1.3989298343658447,
+      "learning_rate": 6.730920803266205e-09,
+      "loss": 2.3119,
+      "step": 16196
+    },
+    {
+      "epoch": 0.9963705708661418,
+      "grad_norm": 1.0916481018066406,
+      "learning_rate": 6.508428890228402e-09,
+      "loss": 2.6452,
+      "step": 16197
+    },
+    {
+      "epoch": 0.9964320866141733,
+      "grad_norm": 1.369829773902893,
+      "learning_rate": 6.28967617619658e-09,
+      "loss": 2.3569,
+      "step": 16198
+    },
+    {
+      "epoch": 0.9964936023622047,
+      "grad_norm": 1.3699060678482056,
+      "learning_rate": 6.074662669353081e-09,
+      "loss": 1.9711,
+      "step": 16199
+    },
+    {
+      "epoch": 0.9965551181102362,
+      "grad_norm": 1.224433422088623,
+      "learning_rate": 5.863388377724821e-09,
+      "loss": 2.9823,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9966166338582677,
+      "grad_norm": 1.0552810430526733,
+      "learning_rate": 5.655853309227688e-09,
+      "loss": 2.3971,
+      "step": 16201
+    },
+    {
+      "epoch": 0.9966781496062992,
+      "grad_norm": 1.2044363021850586,
+      "learning_rate": 5.4520574716221405e-09,
+      "loss": 2.6341,
+      "step": 16202
+    },
+    {
+      "epoch": 0.9967396653543307,
+      "grad_norm": 1.2307614088058472,
+      "learning_rate": 5.25200087252431e-09,
+      "loss": 2.082,
+      "step": 16203
+    },
+    {
+      "epoch": 0.9968011811023622,
+      "grad_norm": 1.1975641250610352,
+      "learning_rate": 5.055683519417098e-09,
+      "loss": 2.8433,
+      "step": 16204
+    },
+    {
+      "epoch": 0.9968626968503937,
+      "grad_norm": 1.225172996520996,
+      "learning_rate": 4.86310541963908e-09,
+      "loss": 1.9877,
+      "step": 16205
+    },
+    {
+      "epoch": 0.9969242125984252,
+      "grad_norm": 1.329525113105774,
+      "learning_rate": 4.674266580384501e-09,
+      "loss": 2.0639,
+      "step": 16206
+    },
+    {
+      "epoch": 0.9969857283464567,
+      "grad_norm": 1.291792631149292,
+      "learning_rate": 4.489167008725481e-09,
+      "loss": 2.4884,
+      "step": 16207
+    },
+    {
+      "epoch": 0.9970472440944882,
+      "grad_norm": 1.3201563358306885,
+      "learning_rate": 4.307806711589812e-09,
+      "loss": 2.7049,
+      "step": 16208
+    },
+    {
+      "epoch": 0.9971087598425197,
+      "grad_norm": 1.2290339469909668,
+      "learning_rate": 4.130185695749855e-09,
+      "loss": 1.7876,
+      "step": 16209
+    },
+    {
+      "epoch": 0.9971702755905512,
+      "grad_norm": 1.3505645990371704,
+      "learning_rate": 3.9563039678447435e-09,
+      "loss": 1.6638,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9972317913385826,
+      "grad_norm": 1.3249911069869995,
+      "learning_rate": 3.786161534380383e-09,
+      "loss": 2.9362,
+      "step": 16211
+    },
+    {
+      "epoch": 0.9972933070866141,
+      "grad_norm": 1.1298621892929077,
+      "learning_rate": 3.619758401718354e-09,
+      "loss": 2.6523,
+      "step": 16212
+    },
+    {
+      "epoch": 0.9973548228346457,
+      "grad_norm": 1.641198992729187,
+      "learning_rate": 3.4570945760870054e-09,
+      "loss": 2.4935,
+      "step": 16213
+    },
+    {
+      "epoch": 0.9974163385826772,
+      "grad_norm": 1.2929364442825317,
+      "learning_rate": 3.2981700635592583e-09,
+      "loss": 1.6051,
+      "step": 16214
+    },
+    {
+      "epoch": 0.9974778543307087,
+      "grad_norm": 1.2387019395828247,
+      "learning_rate": 3.142984870085908e-09,
+      "loss": 2.0274,
+      "step": 16215
+    },
+    {
+      "epoch": 0.9975393700787402,
+      "grad_norm": 1.4370115995407104,
+      "learning_rate": 2.991539001462318e-09,
+      "loss": 2.4308,
+      "step": 16216
+    },
+    {
+      "epoch": 0.9976008858267716,
+      "grad_norm": 1.1813925504684448,
+      "learning_rate": 2.8438324633617285e-09,
+      "loss": 2.3044,
+      "step": 16217
+    },
+    {
+      "epoch": 0.9976624015748031,
+      "grad_norm": 1.27799391746521,
+      "learning_rate": 2.699865261301948e-09,
+      "loss": 2.2151,
+      "step": 16218
+    },
+    {
+      "epoch": 0.9977239173228346,
+      "grad_norm": 1.7172576189041138,
+      "learning_rate": 2.5596374006675584e-09,
+      "loss": 2.1447,
+      "step": 16219
+    },
+    {
+      "epoch": 0.9977854330708661,
+      "grad_norm": 1.539899468421936,
+      "learning_rate": 2.4231488866988116e-09,
+      "loss": 1.9249,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9978469488188977,
+      "grad_norm": 1.618799090385437,
+      "learning_rate": 2.2903997245027344e-09,
+      "loss": 2.1121,
+      "step": 16221
+    },
+    {
+      "epoch": 0.9979084645669292,
+      "grad_norm": 1.2632156610488892,
+      "learning_rate": 2.161389919042023e-09,
+      "loss": 2.1664,
+      "step": 16222
+    },
+    {
+      "epoch": 0.9979699803149606,
+      "grad_norm": 1.7379083633422852,
+      "learning_rate": 2.0361194751461475e-09,
+      "loss": 1.7165,
+      "step": 16223
+    },
+    {
+      "epoch": 0.9980314960629921,
+      "grad_norm": 1.6356664896011353,
+      "learning_rate": 1.9145883974891474e-09,
+      "loss": 3.0794,
+      "step": 16224
+    },
+    {
+      "epoch": 0.9980930118110236,
+      "grad_norm": 1.501373291015625,
+      "learning_rate": 1.7967966906340395e-09,
+      "loss": 2.2427,
+      "step": 16225
+    },
+    {
+      "epoch": 0.9981545275590551,
+      "grad_norm": 1.2663825750350952,
+      "learning_rate": 1.682744358966204e-09,
+      "loss": 2.3045,
+      "step": 16226
+    },
+    {
+      "epoch": 0.9982160433070866,
+      "grad_norm": 1.3478509187698364,
+      "learning_rate": 1.5724314067488977e-09,
+      "loss": 2.0898,
+      "step": 16227
+    },
+    {
+      "epoch": 0.9982775590551181,
+      "grad_norm": 2.245968818664551,
+      "learning_rate": 1.4658578381232523e-09,
+      "loss": 2.8536,
+      "step": 16228
+    },
+    {
+      "epoch": 0.9983390748031497,
+      "grad_norm": 1.4235966205596924,
+      "learning_rate": 1.3630236570749689e-09,
+      "loss": 3.2873,
+      "step": 16229
+    },
+    {
+      "epoch": 0.9984005905511811,
+      "grad_norm": 1.1721590757369995,
+      "learning_rate": 1.2639288674343165e-09,
+      "loss": 2.2078,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9984621062992126,
+      "grad_norm": 1.3136934041976929,
+      "learning_rate": 1.1685734729205423e-09,
+      "loss": 2.0562,
+      "step": 16231
+    },
+    {
+      "epoch": 0.9985236220472441,
+      "grad_norm": 1.3088420629501343,
+      "learning_rate": 1.07695747708636e-09,
+      "loss": 2.3287,
+      "step": 16232
+    },
+    {
+      "epoch": 0.9985851377952756,
+      "grad_norm": 1.2811261415481567,
+      "learning_rate": 9.890808833623588e-10,
+      "loss": 2.1424,
+      "step": 16233
+    },
+    {
+      "epoch": 0.9986466535433071,
+      "grad_norm": 1.2544751167297363,
+      "learning_rate": 9.049436950459012e-10,
+      "loss": 2.4255,
+      "step": 16234
+    },
+    {
+      "epoch": 0.9987081692913385,
+      "grad_norm": 1.353908896446228,
+      "learning_rate": 8.245459152678159e-10,
+      "loss": 2.003,
+      "step": 16235
+    },
+    {
+      "epoch": 0.99876968503937,
+      "grad_norm": 1.3439821004867554,
+      "learning_rate": 7.478875470479097e-10,
+      "loss": 2.2545,
+      "step": 16236
+    },
+    {
+      "epoch": 0.9988312007874016,
+      "grad_norm": 2.229583740234375,
+      "learning_rate": 6.749685932394556e-10,
+      "loss": 2.0754,
+      "step": 16237
+    },
+    {
+      "epoch": 0.9988927165354331,
+      "grad_norm": 1.352307915687561,
+      "learning_rate": 6.057890565736024e-10,
+      "loss": 2.1556,
+      "step": 16238
+    },
+    {
+      "epoch": 0.9989542322834646,
+      "grad_norm": 1.5719561576843262,
+      "learning_rate": 5.40348939648272e-10,
+      "loss": 2.7886,
+      "step": 16239
+    },
+    {
+      "epoch": 0.9990157480314961,
+      "grad_norm": 1.576890230178833,
+      "learning_rate": 4.786482448948526e-10,
+      "loss": 2.0422,
+      "step": 16240
+    },
+    {
+      "epoch": 0.9990772637795275,
+      "grad_norm": 1.1568195819854736,
+      "learning_rate": 4.206869746226083e-10,
+      "loss": 2.3374,
+      "step": 16241
+    },
+    {
+      "epoch": 0.999138779527559,
+      "grad_norm": 1.9935656785964966,
+      "learning_rate": 3.664651310186784e-10,
+      "loss": 2.4335,
+      "step": 16242
+    },
+    {
+      "epoch": 0.9992002952755905,
+      "grad_norm": 1.9384757280349731,
+      "learning_rate": 3.1598271608146433e-10,
+      "loss": 2.6868,
+      "step": 16243
+    },
+    {
+      "epoch": 0.999261811023622,
+      "grad_norm": 1.4490323066711426,
+      "learning_rate": 2.692397317205497e-10,
+      "loss": 2.4637,
+      "step": 16244
+    },
+    {
+      "epoch": 0.9993233267716536,
+      "grad_norm": 1.384597659111023,
+      "learning_rate": 2.262361796789847e-10,
+      "loss": 1.6906,
+      "step": 16245
+    },
+    {
+      "epoch": 0.9993848425196851,
+      "grad_norm": 1.5742255449295044,
+      "learning_rate": 1.869720615554904e-10,
+      "loss": 2.6218,
+      "step": 16246
+    },
+    {
+      "epoch": 0.9994463582677166,
+      "grad_norm": 1.349867582321167,
+      "learning_rate": 1.514473788266635e-10,
+      "loss": 2.6824,
+      "step": 16247
+    },
+    {
+      "epoch": 0.999507874015748,
+      "grad_norm": 1.2955880165100098,
+      "learning_rate": 1.1966213281366933e-10,
+      "loss": 2.5114,
+      "step": 16248
+    },
+    {
+      "epoch": 0.9995693897637795,
+      "grad_norm": 1.3865091800689697,
+      "learning_rate": 9.161632472665105e-11,
+      "loss": 2.7845,
+      "step": 16249
+    },
+    {
+      "epoch": 0.999630905511811,
+      "grad_norm": 2.378154754638672,
+      "learning_rate": 6.73099555870138e-11,
+      "loss": 2.2386,
+      "step": 16250
+    },
+    {
+      "epoch": 0.9996924212598425,
+      "grad_norm": 1.2805712223052979,
+      "learning_rate": 4.674302631624272e-11,
+      "loss": 2.129,
+      "step": 16251
+    },
+    {
+      "epoch": 0.999753937007874,
+      "grad_norm": 1.2453176975250244,
+      "learning_rate": 2.991553768039168e-11,
+      "loss": 2.8667,
+      "step": 16252
+    },
+    {
+      "epoch": 0.9998154527559056,
+      "grad_norm": 1.3301904201507568,
+      "learning_rate": 1.6827490312287807e-11,
+      "loss": 2.1615,
+      "step": 16253
+    },
+    {
+      "epoch": 0.999876968503937,
+      "grad_norm": 1.5151896476745605,
+      "learning_rate": 7.478884700429234e-12,
+      "loss": 2.0887,
+      "step": 16254
+    },
+    {
+      "epoch": 0.9999384842519685,
+      "grad_norm": 1.2759456634521484,
+      "learning_rate": 1.8697211889850963e-12,
+      "loss": 2.5007,
+      "step": 16255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3316564559936523,
+      "learning_rate": 0.0,
+      "loss": 2.5003,
+      "step": 16256
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.4201905727386475,
+      "eval_runtime": 1035.8225,
+      "eval_samples_per_second": 3.701,
+      "eval_steps_per_second": 3.701,
+      "step": 16256
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 16256,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 5.869851444605092e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}