diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24346 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998704047691045,
+  "eval_steps": 500,
+  "global_step": 34722,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008639682059700203,
+      "grad_norm": 11.771962803910544,
+      "learning_rate": 9.596928982725528e-08,
+      "loss": 0.6022,
+      "step": 10
+    },
+    {
+      "epoch": 0.0017279364119400407,
+      "grad_norm": 9.020646240501524,
+      "learning_rate": 1.9193857965451055e-07,
+      "loss": 0.5688,
+      "step": 20
+    },
+    {
+      "epoch": 0.0025919046179100607,
+      "grad_norm": 6.34090392679821,
+      "learning_rate": 2.8790786948176586e-07,
+      "loss": 0.4585,
+      "step": 30
+    },
+    {
+      "epoch": 0.0034558728238800814,
+      "grad_norm": 2.8058492274624522,
+      "learning_rate": 3.838771593090211e-07,
+      "loss": 0.3803,
+      "step": 40
+    },
+    {
+      "epoch": 0.004319841029850101,
+      "grad_norm": 2.6778181919123094,
+      "learning_rate": 4.798464491362765e-07,
+      "loss": 0.3372,
+      "step": 50
+    },
+    {
+      "epoch": 0.005183809235820121,
+      "grad_norm": 2.66961075763771,
+      "learning_rate": 5.758157389635317e-07,
+      "loss": 0.316,
+      "step": 60
+    },
+    {
+      "epoch": 0.0060477774417901425,
+      "grad_norm": 2.52693112946638,
+      "learning_rate": 6.717850287907871e-07,
+      "loss": 0.2944,
+      "step": 70
+    },
+    {
+      "epoch": 0.006911745647760163,
+      "grad_norm": 2.687818570630057,
+      "learning_rate": 7.677543186180422e-07,
+      "loss": 0.2864,
+      "step": 80
+    },
+    {
+      "epoch": 0.007775713853730183,
+      "grad_norm": 2.5730686332715154,
+      "learning_rate": 8.637236084452976e-07,
+      "loss": 0.2732,
+      "step": 90
+    },
+    {
+      "epoch": 0.008639682059700202,
+      "grad_norm": 2.6580593069171905,
+      "learning_rate": 9.59692898272553e-07,
+      "loss": 0.2706,
+      "step": 100
+    },
+    {
+      "epoch": 0.009503650265670223,
+      "grad_norm": 2.6483380675263417,
+      "learning_rate": 1.0556621880998082e-06,
+      "loss": 0.2658,
+      "step": 110
+    },
+    {
+      "epoch": 0.010367618471640243,
+      "grad_norm": 3.2034183330693926,
+      "learning_rate": 1.1516314779270634e-06,
+      "loss": 0.2574,
+      "step": 120
+    },
+    {
+      "epoch": 0.011231586677610265,
+      "grad_norm": 3.5452977484012838,
+      "learning_rate": 1.2476007677543187e-06,
+      "loss": 0.2564,
+      "step": 130
+    },
+    {
+      "epoch": 0.012095554883580285,
+      "grad_norm": 2.8513462750096714,
+      "learning_rate": 1.3435700575815741e-06,
+      "loss": 0.2488,
+      "step": 140
+    },
+    {
+      "epoch": 0.012959523089550305,
+      "grad_norm": 2.6333666283166988,
+      "learning_rate": 1.4395393474088292e-06,
+      "loss": 0.2462,
+      "step": 150
+    },
+    {
+      "epoch": 0.013823491295520325,
+      "grad_norm": 2.6748238297309967,
+      "learning_rate": 1.5355086372360844e-06,
+      "loss": 0.2462,
+      "step": 160
+    },
+    {
+      "epoch": 0.014687459501490346,
+      "grad_norm": 2.5303667333233997,
+      "learning_rate": 1.63147792706334e-06,
+      "loss": 0.2441,
+      "step": 170
+    },
+    {
+      "epoch": 0.015551427707460366,
+      "grad_norm": 2.9751307770273425,
+      "learning_rate": 1.7274472168905951e-06,
+      "loss": 0.2391,
+      "step": 180
+    },
+    {
+      "epoch": 0.016415395913430386,
+      "grad_norm": 2.758192405739691,
+      "learning_rate": 1.8234165067178506e-06,
+      "loss": 0.2352,
+      "step": 190
+    },
+    {
+      "epoch": 0.017279364119400405,
+      "grad_norm": 2.5969506757432677,
+      "learning_rate": 1.919385796545106e-06,
+      "loss": 0.234,
+      "step": 200
+    },
+    {
+      "epoch": 0.018143332325370427,
+      "grad_norm": 2.946805205411811,
+      "learning_rate": 2.015355086372361e-06,
+      "loss": 0.2368,
+      "step": 210
+    },
+    {
+      "epoch": 0.019007300531340445,
+      "grad_norm": 2.375602226438953,
+      "learning_rate": 2.1113243761996164e-06,
+      "loss": 0.2292,
+      "step": 220
+    },
+    {
+      "epoch": 0.019871268737310467,
+      "grad_norm": 2.479386341346247,
+      "learning_rate": 2.2072936660268714e-06,
+      "loss": 0.2282,
+      "step": 230
+    },
+    {
+      "epoch": 0.020735236943280486,
+      "grad_norm": 2.530227027626686,
+      "learning_rate": 2.303262955854127e-06,
+      "loss": 0.2283,
+      "step": 240
+    },
+    {
+      "epoch": 0.021599205149250508,
+      "grad_norm": 2.663697247032743,
+      "learning_rate": 2.3992322456813823e-06,
+      "loss": 0.2287,
+      "step": 250
+    },
+    {
+      "epoch": 0.02246317335522053,
+      "grad_norm": 2.4409362189300823,
+      "learning_rate": 2.4952015355086374e-06,
+      "loss": 0.2283,
+      "step": 260
+    },
+    {
+      "epoch": 0.023327141561190548,
+      "grad_norm": 2.4887477176122963,
+      "learning_rate": 2.5911708253358924e-06,
+      "loss": 0.2274,
+      "step": 270
+    },
+    {
+      "epoch": 0.02419110976716057,
+      "grad_norm": 2.169049552229543,
+      "learning_rate": 2.6871401151631483e-06,
+      "loss": 0.2259,
+      "step": 280
+    },
+    {
+      "epoch": 0.02505507797313059,
+      "grad_norm": 2.2321349960796644,
+      "learning_rate": 2.7831094049904033e-06,
+      "loss": 0.224,
+      "step": 290
+    },
+    {
+      "epoch": 0.02591904617910061,
+      "grad_norm": 2.123716642203702,
+      "learning_rate": 2.8790786948176584e-06,
+      "loss": 0.2261,
+      "step": 300
+    },
+    {
+      "epoch": 0.02678301438507063,
+      "grad_norm": 2.2304887712432158,
+      "learning_rate": 2.975047984644914e-06,
+      "loss": 0.2223,
+      "step": 310
+    },
+    {
+      "epoch": 0.02764698259104065,
+      "grad_norm": 2.3168128684841958,
+      "learning_rate": 3.071017274472169e-06,
+      "loss": 0.2258,
+      "step": 320
+    },
+    {
+      "epoch": 0.02851095079701067,
+      "grad_norm": 2.126683137374433,
+      "learning_rate": 3.1669865642994248e-06,
+      "loss": 0.2218,
+      "step": 330
+    },
+    {
+      "epoch": 0.02937491900298069,
+      "grad_norm": 2.167002578947182,
+      "learning_rate": 3.26295585412668e-06,
+      "loss": 0.2267,
+      "step": 340
+    },
+    {
+      "epoch": 0.03023888720895071,
+      "grad_norm": 2.3404259059137194,
+      "learning_rate": 3.358925143953935e-06,
+      "loss": 0.2154,
+      "step": 350
+    },
+    {
+      "epoch": 0.031102855414920732,
+      "grad_norm": 2.1511856683733623,
+      "learning_rate": 3.4548944337811903e-06,
+      "loss": 0.221,
+      "step": 360
+    },
+    {
+      "epoch": 0.031966823620890754,
+      "grad_norm": 2.158113277467384,
+      "learning_rate": 3.5508637236084453e-06,
+      "loss": 0.2235,
+      "step": 370
+    },
+    {
+      "epoch": 0.03283079182686077,
+      "grad_norm": 2.110044048252369,
+      "learning_rate": 3.6468330134357012e-06,
+      "loss": 0.2204,
+      "step": 380
+    },
+    {
+      "epoch": 0.03369476003283079,
+      "grad_norm": 2.0795923977620725,
+      "learning_rate": 3.7428023032629563e-06,
+      "loss": 0.2209,
+      "step": 390
+    },
+    {
+      "epoch": 0.03455872823880081,
+      "grad_norm": 1.9794767623756446,
+      "learning_rate": 3.838771593090212e-06,
+      "loss": 0.2186,
+      "step": 400
+    },
+    {
+      "epoch": 0.035422696444770835,
+      "grad_norm": 1.9015278291873687,
+      "learning_rate": 3.934740882917467e-06,
+      "loss": 0.217,
+      "step": 410
+    },
+    {
+      "epoch": 0.03628666465074085,
+      "grad_norm": 2.123413533371326,
+      "learning_rate": 4.030710172744722e-06,
+      "loss": 0.2176,
+      "step": 420
+    },
+    {
+      "epoch": 0.03715063285671087,
+      "grad_norm": 1.982609733635354,
+      "learning_rate": 4.126679462571978e-06,
+      "loss": 0.2161,
+      "step": 430
+    },
+    {
+      "epoch": 0.03801460106268089,
+      "grad_norm": 1.873506879383276,
+      "learning_rate": 4.222648752399233e-06,
+      "loss": 0.218,
+      "step": 440
+    },
+    {
+      "epoch": 0.038878569268650916,
+      "grad_norm": 1.9435107928689135,
+      "learning_rate": 4.318618042226488e-06,
+      "loss": 0.2177,
+      "step": 450
+    },
+    {
+      "epoch": 0.039742537474620934,
+      "grad_norm": 2.0252961794361086,
+      "learning_rate": 4.414587332053743e-06,
+      "loss": 0.2154,
+      "step": 460
+    },
+    {
+      "epoch": 0.04060650568059095,
+      "grad_norm": 1.700171956783739,
+      "learning_rate": 4.510556621880998e-06,
+      "loss": 0.2109,
+      "step": 470
+    },
+    {
+      "epoch": 0.04147047388656097,
+      "grad_norm": 2.053940448238235,
+      "learning_rate": 4.606525911708254e-06,
+      "loss": 0.2131,
+      "step": 480
+    },
+    {
+      "epoch": 0.042334442092531,
+      "grad_norm": 1.822531476696438,
+      "learning_rate": 4.702495201535509e-06,
+      "loss": 0.2124,
+      "step": 490
+    },
+    {
+      "epoch": 0.043198410298501015,
+      "grad_norm": 1.9175265658812415,
+      "learning_rate": 4.798464491362765e-06,
+      "loss": 0.2162,
+      "step": 500
+    },
+    {
+      "epoch": 0.044062378504471034,
+      "grad_norm": 1.766701105342602,
+      "learning_rate": 4.89443378119002e-06,
+      "loss": 0.2205,
+      "step": 510
+    },
+    {
+      "epoch": 0.04492634671044106,
+      "grad_norm": 1.835679557041508,
+      "learning_rate": 4.990403071017275e-06,
+      "loss": 0.2153,
+      "step": 520
+    },
+    {
+      "epoch": 0.04579031491641108,
+      "grad_norm": 1.7702803984984092,
+      "learning_rate": 5.086372360844531e-06,
+      "loss": 0.2133,
+      "step": 530
+    },
+    {
+      "epoch": 0.046654283122381096,
+      "grad_norm": 1.7541287271210348,
+      "learning_rate": 5.182341650671785e-06,
+      "loss": 0.2111,
+      "step": 540
+    },
+    {
+      "epoch": 0.047518251328351115,
+      "grad_norm": 1.679583479494552,
+      "learning_rate": 5.278310940499041e-06,
+      "loss": 0.218,
+      "step": 550
+    },
+    {
+      "epoch": 0.04838221953432114,
+      "grad_norm": 1.7372110816101225,
+      "learning_rate": 5.374280230326297e-06,
+      "loss": 0.2111,
+      "step": 560
+    },
+    {
+      "epoch": 0.04924618774029116,
+      "grad_norm": 1.718777295330965,
+      "learning_rate": 5.470249520153551e-06,
+      "loss": 0.2164,
+      "step": 570
+    },
+    {
+      "epoch": 0.05011015594626118,
+      "grad_norm": 1.6716688118747165,
+      "learning_rate": 5.566218809980807e-06,
+      "loss": 0.2186,
+      "step": 580
+    },
+    {
+      "epoch": 0.050974124152231196,
+      "grad_norm": 1.7059282422333841,
+      "learning_rate": 5.662188099808062e-06,
+      "loss": 0.2113,
+      "step": 590
+    },
+    {
+      "epoch": 0.05183809235820122,
+      "grad_norm": 1.7518290879187177,
+      "learning_rate": 5.758157389635317e-06,
+      "loss": 0.2148,
+      "step": 600
+    },
+    {
+      "epoch": 0.05270206056417124,
+      "grad_norm": 1.7399694944414077,
+      "learning_rate": 5.854126679462573e-06,
+      "loss": 0.2154,
+      "step": 610
+    },
+    {
+      "epoch": 0.05356602877014126,
+      "grad_norm": 1.6592091944436103,
+      "learning_rate": 5.950095969289828e-06,
+      "loss": 0.2154,
+      "step": 620
+    },
+    {
+      "epoch": 0.054429996976111276,
+      "grad_norm": 1.6428653305834138,
+      "learning_rate": 6.0460652591170836e-06,
+      "loss": 0.2122,
+      "step": 630
+    },
+    {
+      "epoch": 0.0552939651820813,
+      "grad_norm": 1.5199474469183176,
+      "learning_rate": 6.142034548944338e-06,
+      "loss": 0.2137,
+      "step": 640
+    },
+    {
+      "epoch": 0.05615793338805132,
+      "grad_norm": 1.495986683540699,
+      "learning_rate": 6.238003838771594e-06,
+      "loss": 0.207,
+      "step": 650
+    },
+    {
+      "epoch": 0.05702190159402134,
+      "grad_norm": 1.6130097470231044,
+      "learning_rate": 6.3339731285988495e-06,
+      "loss": 0.2094,
+      "step": 660
+    },
+    {
+      "epoch": 0.05788586979999136,
+      "grad_norm": 1.429539862570976,
+      "learning_rate": 6.429942418426104e-06,
+      "loss": 0.2144,
+      "step": 670
+    },
+    {
+      "epoch": 0.05874983800596138,
+      "grad_norm": 1.4829396307386704,
+      "learning_rate": 6.52591170825336e-06,
+      "loss": 0.2112,
+      "step": 680
+    },
+    {
+      "epoch": 0.0596138062119314,
+      "grad_norm": 1.5601103696909149,
+      "learning_rate": 6.621880998080615e-06,
+      "loss": 0.2146,
+      "step": 690
+    },
+    {
+      "epoch": 0.06047777441790142,
+      "grad_norm": 1.5121395416193373,
+      "learning_rate": 6.71785028790787e-06,
+      "loss": 0.2085,
+      "step": 700
+    },
+    {
+      "epoch": 0.06134174262387144,
+      "grad_norm": 1.5247176525676305,
+      "learning_rate": 6.8138195777351256e-06,
+      "loss": 0.2133,
+      "step": 710
+    },
+    {
+      "epoch": 0.062205710829841464,
+      "grad_norm": 1.5231955702558029,
+      "learning_rate": 6.909788867562381e-06,
+      "loss": 0.2105,
+      "step": 720
+    },
+    {
+      "epoch": 0.06306967903581148,
+      "grad_norm": 1.575431244960636,
+      "learning_rate": 7.005758157389636e-06,
+      "loss": 0.2112,
+      "step": 730
+    },
+    {
+      "epoch": 0.06393364724178151,
+      "grad_norm": 1.4052670709962776,
+      "learning_rate": 7.101727447216891e-06,
+      "loss": 0.214,
+      "step": 740
+    },
+    {
+      "epoch": 0.06479761544775152,
+      "grad_norm": 1.5981212939035538,
+      "learning_rate": 7.1976967370441466e-06,
+      "loss": 0.212,
+      "step": 750
+    },
+    {
+      "epoch": 0.06566158365372154,
+      "grad_norm": 1.3490598266529032,
+      "learning_rate": 7.2936660268714024e-06,
+      "loss": 0.2156,
+      "step": 760
+    },
+    {
+      "epoch": 0.06652555185969157,
+      "grad_norm": 1.469722333119062,
+      "learning_rate": 7.389635316698657e-06,
+      "loss": 0.2125,
+      "step": 770
+    },
+    {
+      "epoch": 0.06738952006566158,
+      "grad_norm": 1.4540654642423125,
+      "learning_rate": 7.4856046065259125e-06,
+      "loss": 0.2103,
+      "step": 780
+    },
+    {
+      "epoch": 0.06825348827163161,
+      "grad_norm": 1.5938247069436307,
+      "learning_rate": 7.581573896353167e-06,
+      "loss": 0.2074,
+      "step": 790
+    },
+    {
+      "epoch": 0.06911745647760162,
+      "grad_norm": 1.419450236768579,
+      "learning_rate": 7.677543186180423e-06,
+      "loss": 0.2156,
+      "step": 800
+    },
+    {
+      "epoch": 0.06998142468357164,
+      "grad_norm": 1.4817149124966364,
+      "learning_rate": 7.773512476007678e-06,
+      "loss": 0.2062,
+      "step": 810
+    },
+    {
+      "epoch": 0.07084539288954167,
+      "grad_norm": 1.3562338115519248,
+      "learning_rate": 7.869481765834934e-06,
+      "loss": 0.2145,
+      "step": 820
+    },
+    {
+      "epoch": 0.07170936109551168,
+      "grad_norm": 1.441182293209827,
+      "learning_rate": 7.965451055662189e-06,
+      "loss": 0.2096,
+      "step": 830
+    },
+    {
+      "epoch": 0.0725733293014817,
+      "grad_norm": 1.4090583980671276,
+      "learning_rate": 8.061420345489444e-06,
+      "loss": 0.2088,
+      "step": 840
+    },
+    {
+      "epoch": 0.07343729750745173,
+      "grad_norm": 1.3527688354129277,
+      "learning_rate": 8.157389635316699e-06,
+      "loss": 0.207,
+      "step": 850
+    },
+    {
+      "epoch": 0.07430126571342174,
+      "grad_norm": 1.3845049446769717,
+      "learning_rate": 8.253358925143955e-06,
+      "loss": 0.2087,
+      "step": 860
+    },
+    {
+      "epoch": 0.07516523391939177,
+      "grad_norm": 1.3620809094506507,
+      "learning_rate": 8.34932821497121e-06,
+      "loss": 0.2062,
+      "step": 870
+    },
+    {
+      "epoch": 0.07602920212536178,
+      "grad_norm": 1.2449980327729386,
+      "learning_rate": 8.445297504798465e-06,
+      "loss": 0.2286,
+      "step": 880
+    },
+    {
+      "epoch": 0.0768931703313318,
+      "grad_norm": 1.2459711984543538,
+      "learning_rate": 8.54126679462572e-06,
+      "loss": 0.2131,
+      "step": 890
+    },
+    {
+      "epoch": 0.07775713853730183,
+      "grad_norm": 1.435496434836962,
+      "learning_rate": 8.637236084452976e-06,
+      "loss": 0.2132,
+      "step": 900
+    },
+    {
+      "epoch": 0.07862110674327184,
+      "grad_norm": 1.2279448447576062,
+      "learning_rate": 8.73320537428023e-06,
+      "loss": 0.2058,
+      "step": 910
+    },
+    {
+      "epoch": 0.07948507494924187,
+      "grad_norm": 1.3411791343537367,
+      "learning_rate": 8.829174664107486e-06,
+      "loss": 0.2104,
+      "step": 920
+    },
+    {
+      "epoch": 0.0803490431552119,
+      "grad_norm": 1.3588577237929125,
+      "learning_rate": 8.925143953934742e-06,
+      "loss": 0.2086,
+      "step": 930
+    },
+    {
+      "epoch": 0.0812130113611819,
+      "grad_norm": 1.215089486089123,
+      "learning_rate": 9.021113243761996e-06,
+      "loss": 0.2116,
+      "step": 940
+    },
+    {
+      "epoch": 0.08207697956715193,
+      "grad_norm": 1.2690165783221579,
+      "learning_rate": 9.117082533589252e-06,
+      "loss": 0.2107,
+      "step": 950
+    },
+    {
+      "epoch": 0.08294094777312194,
+      "grad_norm": 1.1716070949493718,
+      "learning_rate": 9.213051823416507e-06,
+      "loss": 0.2123,
+      "step": 960
+    },
+    {
+      "epoch": 0.08380491597909197,
+      "grad_norm": 1.2018320494999504,
+      "learning_rate": 9.309021113243763e-06,
+      "loss": 0.2076,
+      "step": 970
+    },
+    {
+      "epoch": 0.084668884185062,
+      "grad_norm": 1.253590789081327,
+      "learning_rate": 9.404990403071018e-06,
+      "loss": 0.2053,
+      "step": 980
+    },
+    {
+      "epoch": 0.085532852391032,
+      "grad_norm": 1.3322390529767874,
+      "learning_rate": 9.500959692898273e-06,
+      "loss": 0.2109,
+      "step": 990
+    },
+    {
+      "epoch": 0.08639682059700203,
+      "grad_norm": 1.1375523323508605,
+      "learning_rate": 9.59692898272553e-06,
+      "loss": 0.2102,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08726078880297206,
+      "grad_norm": 1.426192453982837,
+      "learning_rate": 9.692898272552784e-06,
+      "loss": 0.2174,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08812475700894207,
+      "grad_norm": 1.2098268943786776,
+      "learning_rate": 9.78886756238004e-06,
+      "loss": 0.2101,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08898872521491209,
+      "grad_norm": 1.2055452181294146,
+      "learning_rate": 9.884836852207294e-06,
+      "loss": 0.2049,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08985269342088212,
+      "grad_norm": 1.1566160374036807,
+      "learning_rate": 9.98080614203455e-06,
+      "loss": 0.2098,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09071666162685213,
+      "grad_norm": 1.2194302882557917,
+      "learning_rate": 9.999998607883625e-06,
+      "loss": 0.2079,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09158062983282216,
+      "grad_norm": 1.0589518495138135,
+      "learning_rate": 9.999992952412176e-06,
+      "loss": 0.2098,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09244459803879217,
+      "grad_norm": 1.1650993014300575,
+      "learning_rate": 9.999982946583298e-06,
+      "loss": 0.209,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09330856624476219,
+      "grad_norm": 1.1322274380889616,
+      "learning_rate": 9.999968590405698e-06,
+      "loss": 0.2067,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09417253445073222,
+      "grad_norm": 1.199347187742619,
+      "learning_rate": 9.999949883891863e-06,
+      "loss": 0.2043,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09503650265670223,
+      "grad_norm": 1.2200569547783202,
+      "learning_rate": 9.999926827058076e-06,
+      "loss": 0.2069,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09590047086267225,
+      "grad_norm": 1.157827031372808,
+      "learning_rate": 9.999899419924391e-06,
+      "loss": 0.2167,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09676443906864228,
+      "grad_norm": 1.0754344065487103,
+      "learning_rate": 9.999867662514655e-06,
+      "loss": 0.2039,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09762840727461229,
+      "grad_norm": 1.1937972053687231,
+      "learning_rate": 9.999831554856503e-06,
+      "loss": 0.2103,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09849237548058232,
+      "grad_norm": 1.1174871907395723,
+      "learning_rate": 9.99979109698135e-06,
+      "loss": 0.203,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09935634368655233,
+      "grad_norm": 1.080160161013022,
+      "learning_rate": 9.999746288924396e-06,
+      "loss": 0.2023,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10022031189252235,
+      "grad_norm": 1.0920396314412057,
+      "learning_rate": 9.999697130724628e-06,
+      "loss": 0.205,
+      "step": 1160
+    },
+    {
+      "epoch": 0.10108428009849238,
+      "grad_norm": 1.173817518913107,
+      "learning_rate": 9.999643622424817e-06,
+      "loss": 0.2083,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10194824830446239,
+      "grad_norm": 1.0403379627522062,
+      "learning_rate": 9.99958576407152e-06,
+      "loss": 0.2004,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10281221651043242,
+      "grad_norm": 1.146331529719693,
+      "learning_rate": 9.999523555715077e-06,
+      "loss": 0.2089,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10367618471640244,
+      "grad_norm": 1.639211053109325,
+      "learning_rate": 9.999456997409614e-06,
+      "loss": 0.2056,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10454015292237245,
+      "grad_norm": 1.0282120502515748,
+      "learning_rate": 9.999386089213042e-06,
+      "loss": 0.2004,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10540412112834248,
+      "grad_norm": 1.1924096638792132,
+      "learning_rate": 9.999310831187056e-06,
+      "loss": 0.2038,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10626808933431249,
+      "grad_norm": 1.3235594482554829,
+      "learning_rate": 9.999231223397136e-06,
+      "loss": 0.2098,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10713205754028252,
+      "grad_norm": 1.0759458188187956,
+      "learning_rate": 9.999147265912545e-06,
+      "loss": 0.2045,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10799602574625254,
+      "grad_norm": 0.9954941962888572,
+      "learning_rate": 9.999058958806337e-06,
+      "loss": 0.2005,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10885999395222255,
+      "grad_norm": 1.1146268522277554,
+      "learning_rate": 9.998966302155337e-06,
+      "loss": 0.2018,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10972396215819258,
+      "grad_norm": 1.0822500532270867,
+      "learning_rate": 9.998869296040172e-06,
+      "loss": 0.2042,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1105879303641626,
+      "grad_norm": 1.0663634430114168,
+      "learning_rate": 9.99876794054524e-06,
+      "loss": 0.1999,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11145189857013262,
+      "grad_norm": 1.0487941810337988,
+      "learning_rate": 9.998662235758726e-06,
+      "loss": 0.2035,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11231586677610264,
+      "grad_norm": 1.0397680680828816,
+      "learning_rate": 9.998552181772608e-06,
+      "loss": 0.2008,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11317983498207267,
+      "grad_norm": 1.007838733865999,
+      "learning_rate": 9.998437778682632e-06,
+      "loss": 0.201,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11404380318804268,
+      "grad_norm": 0.9490225603657636,
+      "learning_rate": 9.998319026588341e-06,
+      "loss": 0.2013,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1149077713940127,
+      "grad_norm": 1.0866065144355808,
+      "learning_rate": 9.99819592559306e-06,
+      "loss": 0.2033,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11577173959998271,
+      "grad_norm": 1.0692981821008887,
+      "learning_rate": 9.998068475803893e-06,
+      "loss": 0.1964,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11663570780595274,
+      "grad_norm": 1.009125477008282,
+      "learning_rate": 9.99793667733173e-06,
+      "loss": 0.2033,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11749967601192277,
+      "grad_norm": 0.995413726812648,
+      "learning_rate": 9.997800530291249e-06,
+      "loss": 0.2022,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11836364421789278,
+      "grad_norm": 1.0153384644661287,
+      "learning_rate": 9.997660034800904e-06,
+      "loss": 0.2034,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1192276124238628,
+      "grad_norm": 1.0060721836065372,
+      "learning_rate": 9.99751519098294e-06,
+      "loss": 0.1981,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12009158062983283,
+      "grad_norm": 0.9395888677265961,
+      "learning_rate": 9.997365998963378e-06,
+      "loss": 0.2001,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12095554883580284,
+      "grad_norm": 0.9207516546879072,
+      "learning_rate": 9.997212458872026e-06,
+      "loss": 0.1958,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12181951704177287,
+      "grad_norm": 0.9562432904941996,
+      "learning_rate": 9.997054570842476e-06,
+      "loss": 0.1923,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12268348524774288,
+      "grad_norm": 0.9733048311881327,
+      "learning_rate": 9.996892335012106e-06,
+      "loss": 0.2021,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1235474534537129,
+      "grad_norm": 0.928328731096014,
+      "learning_rate": 9.996725751522066e-06,
+      "loss": 0.1962,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12441142165968293,
+      "grad_norm": 0.9312940268247356,
+      "learning_rate": 9.996554820517302e-06,
+      "loss": 0.1925,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12527538986565295,
+      "grad_norm": 0.9649248803993781,
+      "learning_rate": 9.996379542146532e-06,
+      "loss": 0.1953,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12613935807162296,
+      "grad_norm": 0.9689645983824585,
+      "learning_rate": 9.996199916562263e-06,
+      "loss": 0.1969,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12700332627759298,
+      "grad_norm": 1.0121475618102251,
+      "learning_rate": 9.99601594392078e-06,
+      "loss": 0.1944,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12786729448356302,
+      "grad_norm": 0.9444136452659068,
+      "learning_rate": 9.995827624382157e-06,
+      "loss": 0.1959,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12873126268953303,
+      "grad_norm": 1.0984999625010117,
+      "learning_rate": 9.995634958110243e-06,
+      "loss": 0.1979,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12959523089550304,
+      "grad_norm": 0.969715543160458,
+      "learning_rate": 9.995437945272671e-06,
+      "loss": 0.1948,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13045919910147308,
+      "grad_norm": 1.0241223672746604,
+      "learning_rate": 9.995236586040857e-06,
+      "loss": 0.1945,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1313231673074431,
+      "grad_norm": 0.964222667785021,
+      "learning_rate": 9.995030880589998e-06,
+      "loss": 0.1941,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1321871355134131,
+      "grad_norm": 0.9432330938575335,
+      "learning_rate": 9.994820829099074e-06,
+      "loss": 0.195,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13305110371938314,
+      "grad_norm": 0.9307652199560369,
+      "learning_rate": 9.994606431750842e-06,
+      "loss": 0.1854,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13391507192535315,
+      "grad_norm": 0.9310493339961098,
+      "learning_rate": 9.994387688731847e-06,
+      "loss": 0.1928,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13477904013132316,
+      "grad_norm": 0.9439479070573529,
+      "learning_rate": 9.994164600232412e-06,
+      "loss": 0.195,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13564300833729317,
+      "grad_norm": 0.9603028405137487,
+      "learning_rate": 9.993937166446635e-06,
+      "loss": 0.1898,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13650697654326321,
+      "grad_norm": 0.8970445894391301,
+      "learning_rate": 9.993705387572404e-06,
+      "loss": 0.193,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13737094474923323,
+      "grad_norm": 0.9137010906999963,
+      "learning_rate": 9.993469263811383e-06,
+      "loss": 0.1864,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13823491295520324,
+      "grad_norm": 0.8831586632426319,
+      "learning_rate": 9.993228795369017e-06,
+      "loss": 0.1872,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13909888116117328,
+      "grad_norm": 0.9134761502587991,
+      "learning_rate": 9.992983982454528e-06,
+      "loss": 0.1884,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1399628493671433,
+      "grad_norm": 0.9339429524570831,
+      "learning_rate": 9.992734825280926e-06,
+      "loss": 0.1917,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1408268175731133,
+      "grad_norm": 0.9113630951414461,
+      "learning_rate": 9.992481324064991e-06,
+      "loss": 0.1903,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14169078577908334,
+      "grad_norm": 0.9585246215583206,
+      "learning_rate": 9.99222347902729e-06,
+      "loss": 0.1902,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14255475398505335,
+      "grad_norm": 0.953890677401058,
+      "learning_rate": 9.991961290392166e-06,
+      "loss": 0.1875,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14341872219102336,
+      "grad_norm": 0.8503479328191382,
+      "learning_rate": 9.991694758387744e-06,
+      "loss": 0.1887,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1442826903969934,
+      "grad_norm": 0.9020194108982177,
+      "learning_rate": 9.991423883245925e-06,
+      "loss": 0.194,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1451466586029634,
+      "grad_norm": 0.9083223160164683,
+      "learning_rate": 9.99114866520239e-06,
+      "loss": 0.1938,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14601062680893342,
+      "grad_norm": 0.9090284777953003,
+      "learning_rate": 9.9908691044966e-06,
+      "loss": 0.189,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14687459501490346,
+      "grad_norm": 0.9353875805981315,
+      "learning_rate": 9.99058520137179e-06,
+      "loss": 0.1876,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14773856322087348,
+      "grad_norm": 0.9804296916341775,
+      "learning_rate": 9.990296956074979e-06,
+      "loss": 0.1882,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1486025314268435,
+      "grad_norm": 0.8978507801311811,
+      "learning_rate": 9.99000436885696e-06,
+      "loss": 0.1863,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1494664996328135,
+      "grad_norm": 0.9050475519755361,
+      "learning_rate": 9.989707439972306e-06,
+      "loss": 0.1833,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15033046783878354,
+      "grad_norm": 0.9315891329324208,
+      "learning_rate": 9.989406169679367e-06,
+      "loss": 0.1913,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15119443604475355,
+      "grad_norm": 0.922244731553486,
+      "learning_rate": 9.98910055824027e-06,
+      "loss": 0.1883,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15205840425072356,
+      "grad_norm": 0.8667210354836978,
+      "learning_rate": 9.988790605920917e-06,
+      "loss": 0.1852,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1529223724566936,
+      "grad_norm": 0.8814053572722416,
+      "learning_rate": 9.988476312990994e-06,
+      "loss": 0.1837,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1537863406626636,
+      "grad_norm": 0.8845096592429691,
+      "learning_rate": 9.988157679723953e-06,
+      "loss": 0.1827,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15465030886863362,
+      "grad_norm": 0.896642858191613,
+      "learning_rate": 9.98783470639703e-06,
+      "loss": 0.1853,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15551427707460366,
+      "grad_norm": 0.9515141760201947,
+      "learning_rate": 9.987507393291238e-06,
+      "loss": 0.186,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15637824528057367,
+      "grad_norm": 0.8838310226390965,
+      "learning_rate": 9.98717574069136e-06,
+      "loss": 0.1858,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15724221348654369,
+      "grad_norm": 1.0419827856105834,
+      "learning_rate": 9.98683974888596e-06,
+      "loss": 0.1857,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15810618169251373,
+      "grad_norm": 0.8554442292653766,
+      "learning_rate": 9.986499418167373e-06,
+      "loss": 0.1878,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15897014989848374,
+      "grad_norm": 0.9172557697246417,
+      "learning_rate": 9.986154748831715e-06,
+      "loss": 0.1892,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15983411810445375,
+      "grad_norm": 0.906975620762014,
+      "learning_rate": 9.98580574117887e-06,
+      "loss": 0.184,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1606980863104238,
+      "grad_norm": 0.8860481268543452,
+      "learning_rate": 9.9854523955125e-06,
+      "loss": 0.1891,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1615620545163938,
+      "grad_norm": 0.9766505643536513,
+      "learning_rate": 9.985094712140044e-06,
+      "loss": 0.1882,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1624260227223638,
+      "grad_norm": 0.8698571209777873,
+      "learning_rate": 9.98473269137271e-06,
+      "loss": 0.185,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16328999092833385,
+      "grad_norm": 0.8758942022132932,
+      "learning_rate": 9.984366333525483e-06,
+      "loss": 0.1875,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16415395913430386,
+      "grad_norm": 0.8078901518813743,
+      "learning_rate": 9.983995638917122e-06,
+      "loss": 0.186,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16501792734027387,
+      "grad_norm": 0.844074206643064,
+      "learning_rate": 9.98362060787016e-06,
+      "loss": 0.1829,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16588189554624388,
+      "grad_norm": 0.9462307510554506,
+      "learning_rate": 9.983241240710897e-06,
+      "loss": 0.1849,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16674586375221392,
+      "grad_norm": 0.8894512873356976,
+      "learning_rate": 9.982857537769412e-06,
+      "loss": 0.1804,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16760983195818394,
+      "grad_norm": 0.8862445385159019,
+      "learning_rate": 9.982469499379556e-06,
+      "loss": 0.1855,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16847380016415395,
+      "grad_norm": 0.8711563518645582,
+      "learning_rate": 9.982077125878948e-06,
+      "loss": 0.1814,
+      "step": 1950
+    },
+    {
+      "epoch": 0.169337768370124,
+      "grad_norm": 0.9050034309741795,
+      "learning_rate": 9.981680417608983e-06,
+      "loss": 0.1803,
+      "step": 1960
+    },
+    {
+      "epoch": 0.170201736576094,
+      "grad_norm": 0.9013350985231641,
+      "learning_rate": 9.981279374914826e-06,
+      "loss": 0.185,
+      "step": 1970
+    },
+    {
+      "epoch": 0.171065704782064,
+      "grad_norm": 0.8654792583323618,
+      "learning_rate": 9.980873998145413e-06,
+      "loss": 0.1825,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17192967298803405,
+      "grad_norm": 0.8348173364643074,
+      "learning_rate": 9.980464287653451e-06,
+      "loss": 0.1798,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17279364119400406,
+      "grad_norm": 0.8524350735085091,
+      "learning_rate": 9.980050243795418e-06,
+      "loss": 0.182,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17365760939997407,
+      "grad_norm": 0.8823380446120698,
+      "learning_rate": 9.979631866931562e-06,
+      "loss": 0.1838,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1745215776059441,
+      "grad_norm": 0.8990809252489962,
+      "learning_rate": 9.979209157425902e-06,
+      "loss": 0.1832,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17538554581191412,
+      "grad_norm": 0.8737258448461562,
+      "learning_rate": 9.978782115646226e-06,
+      "loss": 0.1833,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17624951401788413,
+      "grad_norm": 0.8539185597232698,
+      "learning_rate": 9.978350741964091e-06,
+      "loss": 0.1821,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17711348222385417,
+      "grad_norm": 0.8718315063650803,
+      "learning_rate": 9.977915036754822e-06,
+      "loss": 0.1838,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17797745042982419,
+      "grad_norm": 0.9045138495180984,
+      "learning_rate": 9.977475000397518e-06,
+      "loss": 0.1828,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1788414186357942,
+      "grad_norm": 0.8655192511724848,
+      "learning_rate": 9.97703063327504e-06,
+      "loss": 0.1842,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17970538684176424,
+      "grad_norm": 0.9058187600514871,
+      "learning_rate": 9.97658193577402e-06,
+      "loss": 0.1813,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18056935504773425,
+      "grad_norm": 0.8534047113131205,
+      "learning_rate": 9.976128908284857e-06,
+      "loss": 0.1823,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18143332325370426,
+      "grad_norm": 0.7845439254440908,
+      "learning_rate": 9.975671551201719e-06,
+      "loss": 0.1796,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18229729145967427,
+      "grad_norm": 0.8348503967840605,
+      "learning_rate": 9.97520986492254e-06,
+      "loss": 0.1803,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1831612596656443,
+      "grad_norm": 0.8434028762996415,
+      "learning_rate": 9.974743849849017e-06,
+      "loss": 0.1793,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18402522787161432,
+      "grad_norm": 0.8447485487459954,
+      "learning_rate": 9.974273506386623e-06,
+      "loss": 0.1735,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18488919607758433,
+      "grad_norm": 0.8030215035923475,
+      "learning_rate": 9.973798834944588e-06,
+      "loss": 0.1826,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18575316428355437,
+      "grad_norm": 0.8518647928321847,
+      "learning_rate": 9.97331983593591e-06,
+      "loss": 0.1812,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18661713248952438,
+      "grad_norm": 0.8494439395574299,
+      "learning_rate": 9.972836509777352e-06,
+      "loss": 0.1796,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1874811006954944,
+      "grad_norm": 0.806243507038616,
+      "learning_rate": 9.972348856889447e-06,
+      "loss": 0.1803,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18834506890146444,
+      "grad_norm": 0.7838675544587994,
+      "learning_rate": 9.971856877696483e-06,
+      "loss": 0.1808,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18920903710743445,
+      "grad_norm": 0.8679707131997498,
+      "learning_rate": 9.971360572626525e-06,
+      "loss": 0.1795,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19007300531340446,
+      "grad_norm": 0.8359795113698925,
+      "learning_rate": 9.970859942111387e-06,
+      "loss": 0.1829,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1909369735193745,
+      "grad_norm": 0.8141606983614745,
+      "learning_rate": 9.97035498658666e-06,
+      "loss": 0.1775,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1918009417253445,
+      "grad_norm": 0.8511617271228508,
+      "learning_rate": 9.969845706491686e-06,
+      "loss": 0.1819,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19266490993131452,
+      "grad_norm": 0.8181597729741305,
+      "learning_rate": 9.96933210226958e-06,
+      "loss": 0.1816,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19352887813728456,
+      "grad_norm": 0.8536969679710033,
+      "learning_rate": 9.968814174367214e-06,
+      "loss": 0.1795,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19439284634325457,
+      "grad_norm": 0.8697086511401918,
+      "learning_rate": 9.968291923235222e-06,
+      "loss": 0.1799,
+      "step": 2250
+    },
+    {
+      "epoch": 0.19525681454922458,
+      "grad_norm": 0.8874621853306252,
+      "learning_rate": 9.967765349328003e-06,
+      "loss": 0.1748,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19612078275519462,
+      "grad_norm": 0.828862563408282,
+      "learning_rate": 9.967234453103712e-06,
+      "loss": 0.1771,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19698475096116463,
+      "grad_norm": 0.7766510398391139,
+      "learning_rate": 9.966699235024266e-06,
+      "loss": 0.178,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19784871916713465,
+      "grad_norm": 0.854829690265069,
+      "learning_rate": 9.966159695555349e-06,
+      "loss": 0.177,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19871268737310466,
+      "grad_norm": 0.8222531787450125,
+      "learning_rate": 9.965615835166396e-06,
+      "loss": 0.1774,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1995766555790747,
+      "grad_norm": 0.8173464466969167,
+      "learning_rate": 9.965067654330604e-06,
+      "loss": 0.1749,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2004406237850447,
+      "grad_norm": 0.7951744192540607,
+      "learning_rate": 9.964515153524932e-06,
+      "loss": 0.1771,
+      "step": 2320
+    },
+    {
+      "epoch": 0.20130459199101472,
+      "grad_norm": 0.8413750210733838,
+      "learning_rate": 9.963958333230097e-06,
+      "loss": 0.1792,
+      "step": 2330
+    },
+    {
+      "epoch": 0.20216856019698476,
+      "grad_norm": 0.8451327183723681,
+      "learning_rate": 9.96339719393057e-06,
+      "loss": 0.1812,
+      "step": 2340
+    },
+    {
+      "epoch": 0.20303252840295477,
+      "grad_norm": 0.8346875351231713,
+      "learning_rate": 9.962831736114585e-06,
+      "loss": 0.1732,
+      "step": 2350
+    },
+    {
+      "epoch": 0.20389649660892478,
+      "grad_norm": 0.7717467067418516,
+      "learning_rate": 9.962261960274132e-06,
+      "loss": 0.1753,
+      "step": 2360
+    },
+    {
+      "epoch": 0.20476046481489482,
+      "grad_norm": 0.8407722193490899,
+      "learning_rate": 9.961687866904954e-06,
+      "loss": 0.1747,
+      "step": 2370
+    },
+    {
+      "epoch": 0.20562443302086483,
+      "grad_norm": 0.8444934823842906,
+      "learning_rate": 9.961109456506559e-06,
+      "loss": 0.1766,
+      "step": 2380
+    },
+    {
+      "epoch": 0.20648840122683484,
+      "grad_norm": 0.8247517545324368,
+      "learning_rate": 9.960526729582203e-06,
+      "loss": 0.1708,
+      "step": 2390
+    },
+    {
+      "epoch": 0.20735236943280488,
+      "grad_norm": 0.8276154600773534,
+      "learning_rate": 9.959939686638901e-06,
+      "loss": 0.1754,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2082163376387749,
+      "grad_norm": 0.7878727685300535,
+      "learning_rate": 9.959348328187424e-06,
+      "loss": 0.1769,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2090803058447449,
+      "grad_norm": 0.8280838186074931,
+      "learning_rate": 9.958752654742296e-06,
+      "loss": 0.1761,
+      "step": 2420
+    },
+    {
+      "epoch": 0.20994427405071495,
+      "grad_norm": 0.8279865260833438,
+      "learning_rate": 9.958152666821796e-06,
+      "loss": 0.1788,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21080824225668496,
+      "grad_norm": 0.8203639460882185,
+      "learning_rate": 9.957548364947959e-06,
+      "loss": 0.1784,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21167221046265497,
+      "grad_norm": 0.7849791611006334,
+      "learning_rate": 9.95693974964657e-06,
+      "loss": 0.1761,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21253617866862498,
+      "grad_norm": 0.7544264491850907,
+      "learning_rate": 9.956326821447168e-06,
+      "loss": 0.1735,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21340014687459502,
+      "grad_norm": 0.7935822340896542,
+      "learning_rate": 9.955709580883047e-06,
+      "loss": 0.175,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21426411508056503,
+      "grad_norm": 0.7652213678508909,
+      "learning_rate": 9.955088028491247e-06,
+      "loss": 0.1738,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21512808328653504,
+      "grad_norm": 0.8065981747060027,
+      "learning_rate": 9.954462164812568e-06,
+      "loss": 0.1753,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21599205149250508,
+      "grad_norm": 0.7773019508262726,
+      "learning_rate": 9.953831990391557e-06,
+      "loss": 0.1773,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2168560196984751,
+      "grad_norm": 0.8167560180005594,
+      "learning_rate": 9.95319750577651e-06,
+      "loss": 0.1766,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2177199879044451,
+      "grad_norm": 0.7745012569374494,
+      "learning_rate": 9.952558711519475e-06,
+      "loss": 0.1733,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21858395611041515,
+      "grad_norm": 0.7882505435416458,
+      "learning_rate": 9.951915608176247e-06,
+      "loss": 0.173,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21944792431638516,
+      "grad_norm": 0.7535315864659419,
+      "learning_rate": 9.951268196306379e-06,
+      "loss": 0.1724,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22031189252235517,
+      "grad_norm": 0.7903619770040948,
+      "learning_rate": 9.950616476473161e-06,
+      "loss": 0.1776,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2211758607283252,
+      "grad_norm": 0.827402074944646,
+      "learning_rate": 9.949960449243638e-06,
+      "loss": 0.1786,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22203982893429522,
+      "grad_norm": 0.7786893208413252,
+      "learning_rate": 9.9493001151886e-06,
+      "loss": 0.177,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22290379714026523,
+      "grad_norm": 0.7907751212309051,
+      "learning_rate": 9.94863547488259e-06,
+      "loss": 0.1761,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22376776534623527,
+      "grad_norm": 0.8621582903262195,
+      "learning_rate": 9.94796652890389e-06,
+      "loss": 0.1759,
+      "step": 2590
+    },
+    {
+      "epoch": 0.22463173355220528,
+      "grad_norm": 0.806757543369109,
+      "learning_rate": 9.947293277834531e-06,
+      "loss": 0.1754,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2254957017581753,
+      "grad_norm": 0.7527211655050535,
+      "learning_rate": 9.946615722260291e-06,
+      "loss": 0.1719,
+      "step": 2610
+    },
+    {
+      "epoch": 0.22635966996414533,
+      "grad_norm": 0.8447146520227704,
+      "learning_rate": 9.945933862770695e-06,
+      "loss": 0.1747,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22722363817011534,
+      "grad_norm": 0.7430272978878533,
+      "learning_rate": 9.94524769995901e-06,
+      "loss": 0.1723,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22808760637608536,
+      "grad_norm": 0.7758344166204706,
+      "learning_rate": 9.944557234422244e-06,
+      "loss": 0.173,
+      "step": 2640
+    },
+    {
+      "epoch": 0.22895157458205537,
+      "grad_norm": 0.7902705002047675,
+      "learning_rate": 9.943862466761154e-06,
+      "loss": 0.1776,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2298155427880254,
+      "grad_norm": 0.8419673410782607,
+      "learning_rate": 9.943163397580237e-06,
+      "loss": 0.1702,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23067951099399542,
+      "grad_norm": 0.8381561999082194,
+      "learning_rate": 9.94246002748774e-06,
+      "loss": 0.179,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23154347919996543,
+      "grad_norm": 0.7574228443477528,
+      "learning_rate": 9.94175235709564e-06,
+      "loss": 0.1731,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23240744740593547,
+      "grad_norm": 0.7911399295773064,
+      "learning_rate": 9.941040387019663e-06,
+      "loss": 0.1716,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23327141561190548,
+      "grad_norm": 0.7898506911596503,
+      "learning_rate": 9.940324117879276e-06,
+      "loss": 0.1731,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2341353838178755,
+      "grad_norm": 0.7981346247488054,
+      "learning_rate": 9.939603550297684e-06,
+      "loss": 0.1731,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23499935202384553,
+      "grad_norm": 0.7584900113832275,
+      "learning_rate": 9.938878684901834e-06,
+      "loss": 0.17,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23586332022981554,
+      "grad_norm": 0.803733635587881,
+      "learning_rate": 9.938149522322411e-06,
+      "loss": 0.1798,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23672728843578555,
+      "grad_norm": 0.8047802032077452,
+      "learning_rate": 9.937416063193841e-06,
+      "loss": 0.1754,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2375912566417556,
+      "grad_norm": 0.748847952157505,
+      "learning_rate": 9.936678308154283e-06,
+      "loss": 0.1721,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2384552248477256,
+      "grad_norm": 0.7198022855499399,
+      "learning_rate": 9.935936257845643e-06,
+      "loss": 0.1704,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23931919305369562,
+      "grad_norm": 1.0481929988558047,
+      "learning_rate": 9.935189912913555e-06,
+      "loss": 0.1702,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24018316125966566,
+      "grad_norm": 0.781807883224081,
+      "learning_rate": 9.934439274007392e-06,
+      "loss": 0.1752,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24104712946563567,
+      "grad_norm": 0.7830306240360945,
+      "learning_rate": 9.93368434178027e-06,
+      "loss": 0.1771,
+      "step": 2790
+    },
+    {
+      "epoch": 0.24191109767160568,
+      "grad_norm": 0.7546245521114135,
+      "learning_rate": 9.93292511688903e-06,
+      "loss": 0.1726,
+      "step": 2800
+    },
+    {
+      "epoch": 0.24277506587757572,
+      "grad_norm": 0.769112224734122,
+      "learning_rate": 9.932161599994253e-06,
+      "loss": 0.1732,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24363903408354573,
+      "grad_norm": 0.8248961320812401,
+      "learning_rate": 9.931393791760258e-06,
+      "loss": 0.1763,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24450300228951574,
+      "grad_norm": 0.7577418125024905,
+      "learning_rate": 9.930621692855089e-06,
+      "loss": 0.1715,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24536697049548575,
+      "grad_norm": 0.7300877318840987,
+      "learning_rate": 9.929845303950533e-06,
+      "loss": 0.1699,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2462309387014558,
+      "grad_norm": 0.7810888803907006,
+      "learning_rate": 9.929064625722103e-06,
+      "loss": 0.1727,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2470949069074258,
+      "grad_norm": 0.7601317930300447,
+      "learning_rate": 9.928279658849044e-06,
+      "loss": 0.1696,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24795887511339582,
+      "grad_norm": 0.7865052537791539,
+      "learning_rate": 9.927490404014335e-06,
+      "loss": 0.1749,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24882284331936586,
+      "grad_norm": 0.7668012354719749,
+      "learning_rate": 9.926696861904688e-06,
+      "loss": 0.1727,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24968681152533587,
+      "grad_norm": 0.7585293542729381,
+      "learning_rate": 9.925899033210537e-06,
+      "loss": 0.1687,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2505507797313059,
+      "grad_norm": 0.7357294583323162,
+      "learning_rate": 9.925096918626057e-06,
+      "loss": 0.1732,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2514147479372759,
+      "grad_norm": 0.7723496222824552,
+      "learning_rate": 9.924290518849143e-06,
+      "loss": 0.1667,
+      "step": 2910
+    },
+    {
+      "epoch": 0.25227871614324593,
+      "grad_norm": 0.7719742079374611,
+      "learning_rate": 9.92347983458142e-06,
+      "loss": 0.1672,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25314268434921594,
+      "grad_norm": 0.7707056771464336,
+      "learning_rate": 9.922664866528245e-06,
+      "loss": 0.1746,
+      "step": 2930
+    },
+    {
+      "epoch": 0.25400665255518595,
+      "grad_norm": 0.7292053975039586,
+      "learning_rate": 9.921845615398696e-06,
+      "loss": 0.1669,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25487062076115596,
+      "grad_norm": 0.7241856067663586,
+      "learning_rate": 9.921022081905584e-06,
+      "loss": 0.1663,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25573458896712603,
+      "grad_norm": 0.7466682681906929,
+      "learning_rate": 9.920194266765443e-06,
+      "loss": 0.172,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25659855717309604,
+      "grad_norm": 0.7453627408950962,
+      "learning_rate": 9.919362170698535e-06,
+      "loss": 0.1688,
+      "step": 2970
+    },
+    {
+      "epoch": 0.25746252537906605,
+      "grad_norm": 0.7355063590474924,
+      "learning_rate": 9.918525794428835e-06,
+      "loss": 0.1724,
+      "step": 2980
+    },
+    {
+      "epoch": 0.25832649358503607,
+      "grad_norm": 0.7790954819757584,
+      "learning_rate": 9.917685138684061e-06,
+      "loss": 0.1691,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2591904617910061,
+      "grad_norm": 0.765744357026,
+      "learning_rate": 9.91684020419564e-06,
+      "loss": 0.1706,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2600544299969761,
+      "grad_norm": 0.7430626764761369,
+      "learning_rate": 9.915990991698725e-06,
+      "loss": 0.1748,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26091839820294616,
+      "grad_norm": 0.7305081505894514,
+      "learning_rate": 9.915137501932196e-06,
+      "loss": 0.1694,
+      "step": 3020
+    },
+    {
+      "epoch": 0.26178236640891617,
+      "grad_norm": 0.7738500765507736,
+      "learning_rate": 9.91427973563865e-06,
+      "loss": 0.1667,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2626463346148862,
+      "grad_norm": 0.7902256019188909,
+      "learning_rate": 9.913417693564406e-06,
+      "loss": 0.1696,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2635103028208562,
+      "grad_norm": 0.7646679714717656,
+      "learning_rate": 9.912551376459502e-06,
+      "loss": 0.1657,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2643742710268262,
+      "grad_norm": 0.7451146026894049,
+      "learning_rate": 9.911680785077699e-06,
+      "loss": 0.1703,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2652382392327962,
+      "grad_norm": 0.7807148744435659,
+      "learning_rate": 9.910805920176472e-06,
+      "loss": 0.1706,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2661022074387663,
+      "grad_norm": 0.7263055380070994,
+      "learning_rate": 9.90992678251702e-06,
+      "loss": 0.1681,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2669661756447363,
+      "grad_norm": 0.7663865032837028,
+      "learning_rate": 9.909043372864256e-06,
+      "loss": 0.1667,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2678301438507063,
+      "grad_norm": 0.7186567166331693,
+      "learning_rate": 9.90815569198681e-06,
+      "loss": 0.1643,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2686941120566763,
+      "grad_norm": 0.7614040177140725,
+      "learning_rate": 9.90726374065703e-06,
+      "loss": 0.1712,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2695580802626463,
+      "grad_norm": 0.7593484797688069,
+      "learning_rate": 9.906367519650976e-06,
+      "loss": 0.1727,
+      "step": 3120
+    },
+    {
+      "epoch": 0.27042204846861634,
+      "grad_norm": 0.7784846405411727,
+      "learning_rate": 9.905467029748427e-06,
+      "loss": 0.1731,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27128601667458635,
+      "grad_norm": 0.6936575897699844,
+      "learning_rate": 9.904562271732877e-06,
+      "loss": 0.1709,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2721499848805564,
+      "grad_norm": 0.7455742600336882,
+      "learning_rate": 9.903653246391526e-06,
+      "loss": 0.1712,
+      "step": 3150
+    },
+    {
+      "epoch": 0.27301395308652643,
+      "grad_norm": 0.7639542012395595,
+      "learning_rate": 9.902739954515298e-06,
+      "loss": 0.1692,
+      "step": 3160
+    },
+    {
+      "epoch": 0.27387792129249644,
+      "grad_norm": 0.7654658583654708,
+      "learning_rate": 9.90182239689882e-06,
+      "loss": 0.1718,
+      "step": 3170
+    },
+    {
+      "epoch": 0.27474188949846645,
+      "grad_norm": 0.7254925144092764,
+      "learning_rate": 9.900900574340433e-06,
+      "loss": 0.1711,
+      "step": 3180
+    },
+    {
+      "epoch": 0.27560585770443646,
+      "grad_norm": 0.7784945029552791,
+      "learning_rate": 9.899974487642191e-06,
+      "loss": 0.1696,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2764698259104065,
+      "grad_norm": 0.7292362264828972,
+      "learning_rate": 9.899044137609857e-06,
+      "loss": 0.1682,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27733379411637654,
+      "grad_norm": 0.7488659895432636,
+      "learning_rate": 9.898109525052904e-06,
+      "loss": 0.1669,
+      "step": 3210
+    },
+    {
+      "epoch": 0.27819776232234655,
+      "grad_norm": 0.7356583051338673,
+      "learning_rate": 9.89717065078451e-06,
+      "loss": 0.1619,
+      "step": 3220
+    },
+    {
+      "epoch": 0.27906173052831657,
+      "grad_norm": 0.72285525580165,
+      "learning_rate": 9.896227515621567e-06,
+      "loss": 0.1661,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2799256987342866,
+      "grad_norm": 0.7339085499346532,
+      "learning_rate": 9.89528012038467e-06,
+      "loss": 0.1668,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2807896669402566,
+      "grad_norm": 0.7303332944076532,
+      "learning_rate": 9.89432846589812e-06,
+      "loss": 0.1686,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2816536351462266,
+      "grad_norm": 0.7801708995133046,
+      "learning_rate": 9.893372552989928e-06,
+      "loss": 0.167,
+      "step": 3260
+    },
+    {
+      "epoch": 0.28251760335219667,
+      "grad_norm": 0.7841519896635663,
+      "learning_rate": 9.892412382491808e-06,
+      "loss": 0.1687,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2833815715581667,
+      "grad_norm": 0.7379472562953742,
+      "learning_rate": 9.891447955239177e-06,
+      "loss": 0.1678,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2842455397641367,
+      "grad_norm": 0.7477316830102703,
+      "learning_rate": 9.890479272071156e-06,
+      "loss": 0.1682,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2851095079701067,
+      "grad_norm": 0.7499765853953004,
+      "learning_rate": 9.88950633383057e-06,
+      "loss": 0.173,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2859734761760767,
+      "grad_norm": 0.7617572357950421,
+      "learning_rate": 9.888529141363949e-06,
+      "loss": 0.1692,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2868374443820467,
+      "grad_norm": 0.7928336156705718,
+      "learning_rate": 9.887547695521518e-06,
+      "loss": 0.1655,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28770141258801674,
+      "grad_norm": 0.7303417343810876,
+      "learning_rate": 9.886561997157207e-06,
+      "loss": 0.163,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2885653807939868,
+      "grad_norm": 0.7364254164460936,
+      "learning_rate": 9.885572047128646e-06,
+      "loss": 0.1717,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2894293489999568,
+      "grad_norm": 0.7948365440268766,
+      "learning_rate": 9.884577846297163e-06,
+      "loss": 0.1718,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2902933172059268,
+      "grad_norm": 0.7305274500423871,
+      "learning_rate": 9.883579395527787e-06,
+      "loss": 0.1641,
+      "step": 3360
+    },
+    {
+      "epoch": 0.29115728541189684,
+      "grad_norm": 0.7134297636334582,
+      "learning_rate": 9.882576695689239e-06,
+      "loss": 0.1691,
+      "step": 3370
+    },
+    {
+      "epoch": 0.29202125361786685,
+      "grad_norm": 0.6925354450672917,
+      "learning_rate": 9.881569747653943e-06,
+      "loss": 0.1683,
+      "step": 3380
+    },
+    {
+      "epoch": 0.29288522182383686,
+      "grad_norm": 0.7386046590981896,
+      "learning_rate": 9.880558552298018e-06,
+      "loss": 0.1713,
+      "step": 3390
+    },
+    {
+      "epoch": 0.29374919002980693,
+      "grad_norm": 0.721795755790784,
+      "learning_rate": 9.879543110501276e-06,
+      "loss": 0.1667,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29461315823577694,
+      "grad_norm": 0.7281554532294857,
+      "learning_rate": 9.878523423147223e-06,
+      "loss": 0.1667,
+      "step": 3410
+    },
+    {
+      "epoch": 0.29547712644174695,
+      "grad_norm": 0.7232451772483279,
+      "learning_rate": 9.877499491123066e-06,
+      "loss": 0.1682,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29634109464771696,
+      "grad_norm": 0.712697831818986,
+      "learning_rate": 9.876471315319699e-06,
+      "loss": 0.1712,
+      "step": 3430
+    },
+    {
+      "epoch": 0.297205062853687,
+      "grad_norm": 0.7314848374059167,
+      "learning_rate": 9.875438896631706e-06,
+      "loss": 0.1669,
+      "step": 3440
+    },
+    {
+      "epoch": 0.298069031059657,
+      "grad_norm": 0.7223661801150361,
+      "learning_rate": 9.87440223595737e-06,
+      "loss": 0.1694,
+      "step": 3450
+    },
+    {
+      "epoch": 0.298932999265627,
+      "grad_norm": 0.7297043299090169,
+      "learning_rate": 9.873361334198661e-06,
+      "loss": 0.1603,
+      "step": 3460
+    },
+    {
+      "epoch": 0.29979696747159706,
+      "grad_norm": 0.7633676240410845,
+      "learning_rate": 9.872316192261238e-06,
+      "loss": 0.1657,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3006609356775671,
+      "grad_norm": 0.6987340528429398,
+      "learning_rate": 9.871266811054449e-06,
+      "loss": 0.1705,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3015249038835371,
+      "grad_norm": 0.7405716794127658,
+      "learning_rate": 9.870213191491335e-06,
+      "loss": 0.1653,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3023888720895071,
+      "grad_norm": 0.7479032250490073,
+      "learning_rate": 9.869155334488622e-06,
+      "loss": 0.1618,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3032528402954771,
+      "grad_norm": 0.7501069449128815,
+      "learning_rate": 9.86809324096672e-06,
+      "loss": 0.1619,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3041168085014471,
+      "grad_norm": 0.7702944945336244,
+      "learning_rate": 9.867026911849728e-06,
+      "loss": 0.1709,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3049807767074172,
+      "grad_norm": 0.7111948089944916,
+      "learning_rate": 9.865956348065431e-06,
+      "loss": 0.1695,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3058447449133872,
+      "grad_norm": 0.7294603405967993,
+      "learning_rate": 9.864881550545296e-06,
+      "loss": 0.1655,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3067087131193572,
+      "grad_norm": 0.7354788765353012,
+      "learning_rate": 9.863802520224474e-06,
+      "loss": 0.1651,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3075726813253272,
+      "grad_norm": 0.717913826033429,
+      "learning_rate": 9.862719258041804e-06,
+      "loss": 0.1623,
+      "step": 3560
+    },
+    {
+      "epoch": 0.30843664953129724,
+      "grad_norm": 0.7612142871404554,
+      "learning_rate": 9.8616317649398e-06,
+      "loss": 0.1717,
+      "step": 3570
+    },
+    {
+      "epoch": 0.30930061773726725,
+      "grad_norm": 0.7464556010648316,
+      "learning_rate": 9.860540041864662e-06,
+      "loss": 0.1642,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3101645859432373,
+      "grad_norm": 0.7331397054467885,
+      "learning_rate": 9.859444089766264e-06,
+      "loss": 0.1673,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3110285541492073,
+      "grad_norm": 0.7874566449044291,
+      "learning_rate": 9.85834390959817e-06,
+      "loss": 0.1658,
+      "step": 3600
+    },
+    {
+      "epoch": 0.31189252235517734,
+      "grad_norm": 0.7535513082033617,
+      "learning_rate": 9.85723950231761e-06,
+      "loss": 0.1675,
+      "step": 3610
+    },
+    {
+      "epoch": 0.31275649056114735,
+      "grad_norm": 0.7159751469299143,
+      "learning_rate": 9.856130868885505e-06,
+      "loss": 0.1679,
+      "step": 3620
+    },
+    {
+      "epoch": 0.31362045876711736,
+      "grad_norm": 0.6813273679617754,
+      "learning_rate": 9.855018010266443e-06,
+      "loss": 0.1664,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31448442697308737,
+      "grad_norm": 0.7173235203424494,
+      "learning_rate": 9.853900927428694e-06,
+      "loss": 0.163,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3153483951790574,
+      "grad_norm": 0.7041005574822984,
+      "learning_rate": 9.852779621344199e-06,
+      "loss": 0.1648,
+      "step": 3650
+    },
+    {
+      "epoch": 0.31621236338502745,
+      "grad_norm": 0.7814286351256942,
+      "learning_rate": 9.851654092988578e-06,
+      "loss": 0.1649,
+      "step": 3660
+    },
+    {
+      "epoch": 0.31707633159099746,
+      "grad_norm": 0.726258361940594,
+      "learning_rate": 9.850524343341121e-06,
+      "loss": 0.1662,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3179402997969675,
+      "grad_norm": 0.7291712487787225,
+      "learning_rate": 9.849390373384793e-06,
+      "loss": 0.1616,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3188042680029375,
+      "grad_norm": 0.7562894585114789,
+      "learning_rate": 9.84825218410623e-06,
+      "loss": 0.1646,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3196682362089075,
+      "grad_norm": 0.7025509334909738,
+      "learning_rate": 9.84710977649574e-06,
+      "loss": 0.1649,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3205322044148775,
+      "grad_norm": 0.7193689188065979,
+      "learning_rate": 9.845963151547302e-06,
+      "loss": 0.167,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3213961726208476,
+      "grad_norm": 0.7177232302086264,
+      "learning_rate": 9.84481231025856e-06,
+      "loss": 0.1648,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3222601408268176,
+      "grad_norm": 0.7181865122617623,
+      "learning_rate": 9.84365725363083e-06,
+      "loss": 0.166,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3231241090327876,
+      "grad_norm": 0.7522519602597432,
+      "learning_rate": 9.842497982669097e-06,
+      "loss": 0.1665,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3239880772387576,
+      "grad_norm": 0.7487144507984922,
+      "learning_rate": 9.84133449838201e-06,
+      "loss": 0.1601,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3248520454447276,
+      "grad_norm": 0.7042965860824565,
+      "learning_rate": 9.840166801781887e-06,
+      "loss": 0.1638,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32571601365069763,
+      "grad_norm": 0.7187756482975587,
+      "learning_rate": 9.838994893884705e-06,
+      "loss": 0.1695,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3265799818566677,
+      "grad_norm": 0.6995611012661251,
+      "learning_rate": 9.837818775710114e-06,
+      "loss": 0.1648,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3274439500626377,
+      "grad_norm": 0.7608691305524844,
+      "learning_rate": 9.836638448281417e-06,
+      "loss": 0.1654,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3283079182686077,
+      "grad_norm": 0.7289093633796809,
+      "learning_rate": 9.835453912625587e-06,
+      "loss": 0.1667,
+      "step": 3800
+    },
+    {
+      "epoch": 0.32917188647457774,
+      "grad_norm": 0.7629065996737572,
+      "learning_rate": 9.834265169773259e-06,
+      "loss": 0.1713,
+      "step": 3810
+    },
+    {
+      "epoch": 0.33003585468054775,
+      "grad_norm": 0.6727524986501837,
+      "learning_rate": 9.833072220758722e-06,
+      "loss": 0.1646,
+      "step": 3820
+    },
+    {
+      "epoch": 0.33089982288651776,
+      "grad_norm": 0.7491865346889776,
+      "learning_rate": 9.831875066619929e-06,
+      "loss": 0.1645,
+      "step": 3830
+    },
+    {
+      "epoch": 0.33176379109248777,
+      "grad_norm": 0.7645752796321156,
+      "learning_rate": 9.830673708398492e-06,
+      "loss": 0.1633,
+      "step": 3840
+    },
+    {
+      "epoch": 0.33262775929845784,
+      "grad_norm": 0.7036207961815555,
+      "learning_rate": 9.829468147139681e-06,
+      "loss": 0.1653,
+      "step": 3850
+    },
+    {
+      "epoch": 0.33349172750442785,
+      "grad_norm": 0.6558514059493622,
+      "learning_rate": 9.828258383892419e-06,
+      "loss": 0.1593,
+      "step": 3860
+    },
+    {
+      "epoch": 0.33435569571039786,
+      "grad_norm": 0.6523169696559058,
+      "learning_rate": 9.827044419709289e-06,
+      "loss": 0.1665,
+      "step": 3870
+    },
+    {
+      "epoch": 0.33521966391636787,
+      "grad_norm": 0.719938636914537,
+      "learning_rate": 9.825826255646532e-06,
+      "loss": 0.1654,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3360836321223379,
+      "grad_norm": 0.9142436257642391,
+      "learning_rate": 9.824603892764033e-06,
+      "loss": 0.1655,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3369476003283079,
+      "grad_norm": 0.7308837755716424,
+      "learning_rate": 9.82337733212534e-06,
+      "loss": 0.1647,
+      "step": 3900
+    },
+    {
+      "epoch": 0.33781156853427796,
+      "grad_norm": 0.7170945250305703,
+      "learning_rate": 9.82214657479765e-06,
+      "loss": 0.165,
+      "step": 3910
+    },
+    {
+      "epoch": 0.338675536740248,
+      "grad_norm": 0.7101772423726191,
+      "learning_rate": 9.820911621851813e-06,
+      "loss": 0.1674,
+      "step": 3920
+    },
+    {
+      "epoch": 0.339539504946218,
+      "grad_norm": 0.7105871752653151,
+      "learning_rate": 9.819672474362324e-06,
+      "loss": 0.1661,
+      "step": 3930
+    },
+    {
+      "epoch": 0.340403473152188,
+      "grad_norm": 0.6928866893620583,
+      "learning_rate": 9.818429133407332e-06,
+      "loss": 0.1639,
+      "step": 3940
+    },
+    {
+      "epoch": 0.341267441358158,
+      "grad_norm": 0.7070762196788623,
+      "learning_rate": 9.817181600068636e-06,
+      "loss": 0.1598,
+      "step": 3950
+    },
+    {
+      "epoch": 0.342131409564128,
+      "grad_norm": 0.7131526299925895,
+      "learning_rate": 9.81592987543168e-06,
+      "loss": 0.1603,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3429953777700981,
+      "grad_norm": 0.6857882956396996,
+      "learning_rate": 9.814673960585556e-06,
+      "loss": 0.166,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3438593459760681,
+      "grad_norm": 0.7335700923086347,
+      "learning_rate": 9.813413856623002e-06,
+      "loss": 0.167,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3447233141820381,
+      "grad_norm": 0.6928897108991285,
+      "learning_rate": 9.812149564640397e-06,
+      "loss": 0.1612,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3455872823880081,
+      "grad_norm": 0.7266942679789092,
+      "learning_rate": 9.810881085737769e-06,
+      "loss": 0.1617,
+      "step": 4000
+    },
+    {
+      "epoch": 0.34645125059397813,
+      "grad_norm": 0.7271458299087374,
+      "learning_rate": 9.809608421018786e-06,
+      "loss": 0.1606,
+      "step": 4010
+    },
+    {
+      "epoch": 0.34731521879994814,
+      "grad_norm": 0.7087330966470788,
+      "learning_rate": 9.80833157159076e-06,
+      "loss": 0.1635,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34817918700591816,
+      "grad_norm": 0.6481443225031476,
+      "learning_rate": 9.807050538564644e-06,
+      "loss": 0.1577,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3490431552118882,
+      "grad_norm": 0.701382417957392,
+      "learning_rate": 9.805765323055025e-06,
+      "loss": 0.1603,
+      "step": 4040
+    },
+    {
+      "epoch": 0.34990712341785823,
+      "grad_norm": 0.6975531043581854,
+      "learning_rate": 9.804475926180139e-06,
+      "loss": 0.1658,
+      "step": 4050
+    },
+    {
+      "epoch": 0.35077109162382825,
+      "grad_norm": 0.738431014281204,
+      "learning_rate": 9.803182349061853e-06,
+      "loss": 0.1655,
+      "step": 4060
+    },
+    {
+      "epoch": 0.35163505982979826,
+      "grad_norm": 0.7081474595206806,
+      "learning_rate": 9.801884592825673e-06,
+      "loss": 0.1632,
+      "step": 4070
+    },
+    {
+      "epoch": 0.35249902803576827,
+      "grad_norm": 0.6865576639672578,
+      "learning_rate": 9.80058265860074e-06,
+      "loss": 0.1683,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3533629962417383,
+      "grad_norm": 0.7273806197401467,
+      "learning_rate": 9.799276547519836e-06,
+      "loss": 0.1663,
+      "step": 4090
+    },
+    {
+      "epoch": 0.35422696444770835,
+      "grad_norm": 0.7164869319313873,
+      "learning_rate": 9.797966260719369e-06,
+      "loss": 0.1632,
+      "step": 4100
+    },
+    {
+      "epoch": 0.35509093265367836,
+      "grad_norm": 0.6888892295698765,
+      "learning_rate": 9.796651799339383e-06,
+      "loss": 0.1642,
+      "step": 4110
+    },
+    {
+      "epoch": 0.35595490085964837,
+      "grad_norm": 0.6597106187788412,
+      "learning_rate": 9.795333164523557e-06,
+      "loss": 0.1662,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3568188690656184,
+      "grad_norm": 0.7121146624840744,
+      "learning_rate": 9.7940103574192e-06,
+      "loss": 0.1629,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3576828372715884,
+      "grad_norm": 0.7300742422291335,
+      "learning_rate": 9.792683379177249e-06,
+      "loss": 0.1631,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3585468054775584,
+      "grad_norm": 0.7310423779575642,
+      "learning_rate": 9.791352230952269e-06,
+      "loss": 0.1669,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3594107736835285,
+      "grad_norm": 0.7040579907657136,
+      "learning_rate": 9.790016913902458e-06,
+      "loss": 0.1657,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3602747418894985,
+      "grad_norm": 0.6613582124056474,
+      "learning_rate": 9.788677429189642e-06,
+      "loss": 0.165,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3611387100954685,
+      "grad_norm": 0.7137351700890041,
+      "learning_rate": 9.787333777979266e-06,
+      "loss": 0.1628,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3620026783014385,
+      "grad_norm": 0.6950454669747205,
+      "learning_rate": 9.785985961440405e-06,
+      "loss": 0.166,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3628666465074085,
+      "grad_norm": 0.690613286107089,
+      "learning_rate": 9.784633980745756e-06,
+      "loss": 0.1644,
+      "step": 4200
+    },
+    {
+      "epoch": 0.36373061471337853,
+      "grad_norm": 0.6770452196316469,
+      "learning_rate": 9.783277837071647e-06,
+      "loss": 0.1589,
+      "step": 4210
+    },
+    {
+      "epoch": 0.36459458291934854,
+      "grad_norm": 0.7351025780843091,
+      "learning_rate": 9.781917531598013e-06,
+      "loss": 0.1605,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3654585511253186,
+      "grad_norm": 0.7029824747231594,
+      "learning_rate": 9.780553065508424e-06,
+      "loss": 0.1589,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3663225193312886,
+      "grad_norm": 0.7065773579594148,
+      "learning_rate": 9.779184439990064e-06,
+      "loss": 0.1671,
+      "step": 4240
+    },
+    {
+      "epoch": 0.36718648753725863,
+      "grad_norm": 0.6659108075139156,
+      "learning_rate": 9.777811656233738e-06,
+      "loss": 0.1677,
+      "step": 4250
+    },
+    {
+      "epoch": 0.36805045574322864,
+      "grad_norm": 0.699215116558182,
+      "learning_rate": 9.776434715433863e-06,
+      "loss": 0.163,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36891442394919866,
+      "grad_norm": 0.706583481787705,
+      "learning_rate": 9.775053618788482e-06,
+      "loss": 0.1647,
+      "step": 4270
+    },
+    {
+      "epoch": 0.36977839215516867,
+      "grad_norm": 0.6874162890009886,
+      "learning_rate": 9.773668367499246e-06,
+      "loss": 0.1673,
+      "step": 4280
+    },
+    {
+      "epoch": 0.37064236036113873,
+      "grad_norm": 0.7546642378109868,
+      "learning_rate": 9.772278962771427e-06,
+      "loss": 0.164,
+      "step": 4290
+    },
+    {
+      "epoch": 0.37150632856710875,
+      "grad_norm": 0.6958279018093277,
+      "learning_rate": 9.770885405813907e-06,
+      "loss": 0.1639,
+      "step": 4300
+    },
+    {
+      "epoch": 0.37237029677307876,
+      "grad_norm": 0.6863783205097246,
+      "learning_rate": 9.769487697839184e-06,
+      "loss": 0.1606,
+      "step": 4310
+    },
+    {
+      "epoch": 0.37323426497904877,
+      "grad_norm": 0.7660038203574383,
+      "learning_rate": 9.768085840063363e-06,
+      "loss": 0.1658,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3740982331850188,
+      "grad_norm": 0.7310683895324374,
+      "learning_rate": 9.766679833706163e-06,
+      "loss": 0.165,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3749622013909888,
+      "grad_norm": 0.6682511713805321,
+      "learning_rate": 9.765269679990913e-06,
+      "loss": 0.1613,
+      "step": 4340
+    },
+    {
+      "epoch": 0.37582616959695886,
+      "grad_norm": 0.7295984036992738,
+      "learning_rate": 9.763855380144546e-06,
+      "loss": 0.1627,
+      "step": 4350
+    },
+    {
+      "epoch": 0.37669013780292887,
+      "grad_norm": 0.6661232017039217,
+      "learning_rate": 9.762436935397608e-06,
+      "loss": 0.1583,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3775541060088989,
+      "grad_norm": 0.6649991756043743,
+      "learning_rate": 9.76101434698425e-06,
+      "loss": 0.1618,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3784180742148689,
+      "grad_norm": 0.6923759634173425,
+      "learning_rate": 9.759587616142225e-06,
+      "loss": 0.163,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3792820424208389,
+      "grad_norm": 0.6871423621900502,
+      "learning_rate": 9.758156744112895e-06,
+      "loss": 0.1597,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3801460106268089,
+      "grad_norm": 0.7069305432340762,
+      "learning_rate": 9.75672173214122e-06,
+      "loss": 0.1674,
+      "step": 4400
+    },
+    {
+      "epoch": 0.38100997883277893,
+      "grad_norm": 0.7103059022932815,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.1606,
+      "step": 4410
+    },
+    {
+      "epoch": 0.381873947038749,
+      "grad_norm": 0.6899949710005489,
+      "learning_rate": 9.753839293368704e-06,
+      "loss": 0.1619,
+      "step": 4420
+    },
+    {
+      "epoch": 0.382737915244719,
+      "grad_norm": 0.6866404341566841,
+      "learning_rate": 9.752391869075791e-06,
+      "loss": 0.1621,
+      "step": 4430
+    },
+    {
+      "epoch": 0.383601883450689,
+      "grad_norm": 0.6747754230556255,
+      "learning_rate": 9.750940309856393e-06,
+      "loss": 0.1613,
+      "step": 4440
+    },
+    {
+      "epoch": 0.38446585165665903,
+      "grad_norm": 0.7249657876338548,
+      "learning_rate": 9.749484616973478e-06,
+      "loss": 0.1613,
+      "step": 4450
+    },
+    {
+      "epoch": 0.38532981986262904,
+      "grad_norm": 0.7105209288147724,
+      "learning_rate": 9.748024791693598e-06,
+      "loss": 0.1569,
+      "step": 4460
+    },
+    {
+      "epoch": 0.38619378806859905,
+      "grad_norm": 0.6574038542965889,
+      "learning_rate": 9.74656083528691e-06,
+      "loss": 0.1597,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3870577562745691,
+      "grad_norm": 0.6480703817246564,
+      "learning_rate": 9.745092749027163e-06,
+      "loss": 0.1603,
+      "step": 4480
+    },
+    {
+      "epoch": 0.38792172448053913,
+      "grad_norm": 0.6415644598250044,
+      "learning_rate": 9.743620534191698e-06,
+      "loss": 0.1588,
+      "step": 4490
+    },
+    {
+      "epoch": 0.38878569268650914,
+      "grad_norm": 0.6805308455140375,
+      "learning_rate": 9.74214419206145e-06,
+      "loss": 0.1629,
+      "step": 4500
+    },
+    {
+      "epoch": 0.38964966089247915,
+      "grad_norm": 0.7330716655075384,
+      "learning_rate": 9.74066372392094e-06,
+      "loss": 0.161,
+      "step": 4510
+    },
+    {
+      "epoch": 0.39051362909844917,
+      "grad_norm": 0.6818706767824103,
+      "learning_rate": 9.73917913105829e-06,
+      "loss": 0.1591,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3913775973044192,
+      "grad_norm": 0.7047990410076855,
+      "learning_rate": 9.737690414765198e-06,
+      "loss": 0.1641,
+      "step": 4530
+    },
+    {
+      "epoch": 0.39224156551038925,
+      "grad_norm": 0.6958885127192019,
+      "learning_rate": 9.736197576336957e-06,
+      "loss": 0.1598,
+      "step": 4540
+    },
+    {
+      "epoch": 0.39310553371635926,
+      "grad_norm": 0.7029461177460052,
+      "learning_rate": 9.734700617072444e-06,
+      "loss": 0.1561,
+      "step": 4550
+    },
+    {
+      "epoch": 0.39396950192232927,
+      "grad_norm": 0.6822164097545496,
+      "learning_rate": 9.733199538274124e-06,
+      "loss": 0.1576,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3948334701282993,
+      "grad_norm": 0.6805525304356631,
+      "learning_rate": 9.731694341248045e-06,
+      "loss": 0.1582,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3956974383342693,
+      "grad_norm": 0.7090666255186993,
+      "learning_rate": 9.730185027303837e-06,
+      "loss": 0.1625,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3965614065402393,
+      "grad_norm": 0.6775427538075873,
+      "learning_rate": 9.728671597754715e-06,
+      "loss": 0.1599,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3974253747462093,
+      "grad_norm": 0.7056790719470328,
+      "learning_rate": 9.727154053917469e-06,
+      "loss": 0.1628,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3982893429521794,
+      "grad_norm": 0.638674649902704,
+      "learning_rate": 9.725632397112474e-06,
+      "loss": 0.1629,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3991533111581494,
+      "grad_norm": 0.6550807852274794,
+      "learning_rate": 9.724106628663683e-06,
+      "loss": 0.16,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4000172793641194,
+      "grad_norm": 0.6460332536413842,
+      "learning_rate": 9.722576749898624e-06,
+      "loss": 0.1579,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4008812475700894,
+      "grad_norm": 0.6551551230297281,
+      "learning_rate": 9.721042762148405e-06,
+      "loss": 0.1622,
+      "step": 4640
+    },
+    {
+      "epoch": 0.40174521577605943,
+      "grad_norm": 0.6626205641580281,
+      "learning_rate": 9.719504666747704e-06,
+      "loss": 0.1559,
+      "step": 4650
+    },
+    {
+      "epoch": 0.40260918398202944,
+      "grad_norm": 0.7096427344280777,
+      "learning_rate": 9.717962465034778e-06,
+      "loss": 0.1658,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4034731521879995,
+      "grad_norm": 0.7129316785627647,
+      "learning_rate": 9.716416158351454e-06,
+      "loss": 0.1624,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4043371203939695,
+      "grad_norm": 0.6976638834788887,
+      "learning_rate": 9.714865748043129e-06,
+      "loss": 0.164,
+      "step": 4680
+    },
+    {
+      "epoch": 0.40520108859993953,
+      "grad_norm": 0.6658249350744622,
+      "learning_rate": 9.713311235458778e-06,
+      "loss": 0.1631,
+      "step": 4690
+    },
+    {
+      "epoch": 0.40606505680590954,
+      "grad_norm": 0.6840733659601906,
+      "learning_rate": 9.711752621950936e-06,
+      "loss": 0.1615,
+      "step": 4700
+    },
+    {
+      "epoch": 0.40692902501187955,
+      "grad_norm": 0.6840753154130336,
+      "learning_rate": 9.71018990887571e-06,
+      "loss": 0.1617,
+      "step": 4710
+    },
+    {
+      "epoch": 0.40779299321784956,
+      "grad_norm": 0.6630826297979107,
+      "learning_rate": 9.708623097592775e-06,
+      "loss": 0.1621,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4086569614238196,
+      "grad_norm": 0.7140002244423074,
+      "learning_rate": 9.70705218946537e-06,
+      "loss": 0.1622,
+      "step": 4730
+    },
+    {
+      "epoch": 0.40952092962978964,
+      "grad_norm": 0.6758179638610594,
+      "learning_rate": 9.705477185860302e-06,
+      "loss": 0.1594,
+      "step": 4740
+    },
+    {
+      "epoch": 0.41038489783575965,
+      "grad_norm": 0.6900535022098612,
+      "learning_rate": 9.703898088147935e-06,
+      "loss": 0.1632,
+      "step": 4750
+    },
+    {
+      "epoch": 0.41124886604172967,
+      "grad_norm": 0.6761698450465825,
+      "learning_rate": 9.702314897702203e-06,
+      "loss": 0.1614,
+      "step": 4760
+    },
+    {
+      "epoch": 0.4121128342476997,
+      "grad_norm": 0.6485182080321424,
+      "learning_rate": 9.700727615900591e-06,
+      "loss": 0.1604,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4129768024536697,
+      "grad_norm": 0.689619498277749,
+      "learning_rate": 9.699136244124155e-06,
+      "loss": 0.1613,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4138407706596397,
+      "grad_norm": 0.6848912848714906,
+      "learning_rate": 9.697540783757502e-06,
+      "loss": 0.1641,
+      "step": 4790
+    },
+    {
+      "epoch": 0.41470473886560977,
+      "grad_norm": 0.6931500377479651,
+      "learning_rate": 9.695941236188797e-06,
+      "loss": 0.1619,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4155687070715798,
+      "grad_norm": 0.6851757299026764,
+      "learning_rate": 9.694337602809765e-06,
+      "loss": 0.1566,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4164326752775498,
+      "grad_norm": 0.6624506770445044,
+      "learning_rate": 9.692729885015684e-06,
+      "loss": 0.1599,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4172966434835198,
+      "grad_norm": 0.6387799436823667,
+      "learning_rate": 9.691118084205382e-06,
+      "loss": 0.1637,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4181606116894898,
+      "grad_norm": 0.6918645541241718,
+      "learning_rate": 9.689502201781247e-06,
+      "loss": 0.1596,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4190245798954598,
+      "grad_norm": 0.6718823552411884,
+      "learning_rate": 9.68788223914921e-06,
+      "loss": 0.1599,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4198885481014299,
+      "grad_norm": 0.6412918843776836,
+      "learning_rate": 9.686258197718761e-06,
+      "loss": 0.1577,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4207525163073999,
+      "grad_norm": 0.674420383357471,
+      "learning_rate": 9.684630078902933e-06,
+      "loss": 0.1607,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4216164845133699,
+      "grad_norm": 0.6704220676917114,
+      "learning_rate": 9.682997884118303e-06,
+      "loss": 0.1656,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4224804527193399,
+      "grad_norm": 0.6878812416032285,
+      "learning_rate": 9.681361614785006e-06,
+      "loss": 0.1594,
+      "step": 4890
+    },
+    {
+      "epoch": 0.42334442092530994,
+      "grad_norm": 0.6553717926047634,
+      "learning_rate": 9.679721272326709e-06,
+      "loss": 0.1584,
+      "step": 4900
+    },
+    {
+      "epoch": 0.42420838913127995,
+      "grad_norm": 0.6774054340189088,
+      "learning_rate": 9.678076858170633e-06,
+      "loss": 0.1576,
+      "step": 4910
+    },
+    {
+      "epoch": 0.42507235733724996,
+      "grad_norm": 0.6738216331424101,
+      "learning_rate": 9.676428373747538e-06,
+      "loss": 0.1588,
+      "step": 4920
+    },
+    {
+      "epoch": 0.42593632554322003,
+      "grad_norm": 0.6741765178104295,
+      "learning_rate": 9.674775820491725e-06,
+      "loss": 0.1573,
+      "step": 4930
+    },
+    {
+      "epoch": 0.42680029374919004,
+      "grad_norm": 0.6760301728835072,
+      "learning_rate": 9.673119199841033e-06,
+      "loss": 0.1587,
+      "step": 4940
+    },
+    {
+      "epoch": 0.42766426195516005,
+      "grad_norm": 0.683102732085449,
+      "learning_rate": 9.671458513236845e-06,
+      "loss": 0.1599,
+      "step": 4950
+    },
+    {
+      "epoch": 0.42852823016113006,
+      "grad_norm": 0.6379446609802096,
+      "learning_rate": 9.669793762124079e-06,
+      "loss": 0.1565,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4293921983671001,
+      "grad_norm": 0.6798505709963193,
+      "learning_rate": 9.668124947951187e-06,
+      "loss": 0.1621,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4302561665730701,
+      "grad_norm": 0.6549161090260386,
+      "learning_rate": 9.666452072170163e-06,
+      "loss": 0.1582,
+      "step": 4980
+    },
+    {
+      "epoch": 0.43112013477904015,
+      "grad_norm": 0.6852988744956113,
+      "learning_rate": 9.664775136236528e-06,
+      "loss": 0.1566,
+      "step": 4990
+    },
+    {
+      "epoch": 0.43198410298501017,
+      "grad_norm": 0.6526290012979358,
+      "learning_rate": 9.663094141609337e-06,
+      "loss": 0.1587,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4328480711909802,
+      "grad_norm": 0.6668884578151805,
+      "learning_rate": 9.661409089751179e-06,
+      "loss": 0.1575,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4337120393969502,
+      "grad_norm": 0.6976587290426327,
+      "learning_rate": 9.659719982128172e-06,
+      "loss": 0.16,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4345760076029202,
+      "grad_norm": 0.7079200322711302,
+      "learning_rate": 9.65802682020996e-06,
+      "loss": 0.159,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4354399758088902,
+      "grad_norm": 0.6917877632076461,
+      "learning_rate": 9.656329605469724e-06,
+      "loss": 0.1547,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4363039440148603,
+      "grad_norm": 0.6797501891633531,
+      "learning_rate": 9.654628339384154e-06,
+      "loss": 0.161,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4371679122208303,
+      "grad_norm": 0.6635520032766078,
+      "learning_rate": 9.652923023433483e-06,
+      "loss": 0.1551,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4380318804268003,
+      "grad_norm": 0.6803906100574411,
+      "learning_rate": 9.651213659101456e-06,
+      "loss": 0.1606,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4388958486327703,
+      "grad_norm": 0.6700575525523713,
+      "learning_rate": 9.649500247875347e-06,
+      "loss": 0.1604,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4397598168387403,
+      "grad_norm": 0.6662538170197336,
+      "learning_rate": 9.647782791245945e-06,
+      "loss": 0.1568,
+      "step": 5090
+    },
+    {
+      "epoch": 0.44062378504471034,
+      "grad_norm": 0.6624543848438776,
+      "learning_rate": 9.646061290707566e-06,
+      "loss": 0.1574,
+      "step": 5100
+    },
+    {
+      "epoch": 0.44148775325068035,
+      "grad_norm": 0.7109126656654107,
+      "learning_rate": 9.644335747758037e-06,
+      "loss": 0.1591,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4423517214566504,
+      "grad_norm": 0.6473806141940632,
+      "learning_rate": 9.642606163898708e-06,
+      "loss": 0.1595,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4432156896626204,
+      "grad_norm": 0.7006474036535846,
+      "learning_rate": 9.640872540634443e-06,
+      "loss": 0.1557,
+      "step": 5130
+    },
+    {
+      "epoch": 0.44407965786859044,
+      "grad_norm": 0.641365283402573,
+      "learning_rate": 9.63913487947362e-06,
+      "loss": 0.1624,
+      "step": 5140
+    },
+    {
+      "epoch": 0.44494362607456045,
+      "grad_norm": 0.6959501153561491,
+      "learning_rate": 9.63739318192813e-06,
+      "loss": 0.1612,
+      "step": 5150
+    },
+    {
+      "epoch": 0.44580759428053046,
+      "grad_norm": 0.6750876307270886,
+      "learning_rate": 9.635647449513375e-06,
+      "loss": 0.1594,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4466715624865005,
+      "grad_norm": 0.6462815274785941,
+      "learning_rate": 9.633897683748271e-06,
+      "loss": 0.1563,
+      "step": 5170
+    },
+    {
+      "epoch": 0.44753553069247054,
+      "grad_norm": 0.6795189194974991,
+      "learning_rate": 9.632143886155242e-06,
+      "loss": 0.1569,
+      "step": 5180
+    },
+    {
+      "epoch": 0.44839949889844055,
+      "grad_norm": 0.6386005245305387,
+      "learning_rate": 9.630386058260219e-06,
+      "loss": 0.1595,
+      "step": 5190
+    },
+    {
+      "epoch": 0.44926346710441056,
+      "grad_norm": 0.6696051720370734,
+      "learning_rate": 9.628624201592637e-06,
+      "loss": 0.1559,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4501274353103806,
+      "grad_norm": 0.6525705870370956,
+      "learning_rate": 9.626858317685446e-06,
+      "loss": 0.1552,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4509914035163506,
+      "grad_norm": 0.6601726705245239,
+      "learning_rate": 9.625088408075088e-06,
+      "loss": 0.1547,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4518553717223206,
+      "grad_norm": 0.6949657010269852,
+      "learning_rate": 9.623314474301513e-06,
+      "loss": 0.1628,
+      "step": 5230
+    },
+    {
+      "epoch": 0.45271933992829066,
+      "grad_norm": 0.6272986645491211,
+      "learning_rate": 9.621536517908175e-06,
+      "loss": 0.1584,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4535833081342607,
+      "grad_norm": 0.6564336061317336,
+      "learning_rate": 9.619754540442023e-06,
+      "loss": 0.1561,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4544472763402307,
+      "grad_norm": 0.6339087400986222,
+      "learning_rate": 9.61796854345351e-06,
+      "loss": 0.1584,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4553112445462007,
+      "grad_norm": 0.655855968908132,
+      "learning_rate": 9.616178528496583e-06,
+      "loss": 0.1553,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4561752127521707,
+      "grad_norm": 0.6860757996153211,
+      "learning_rate": 9.61438449712868e-06,
+      "loss": 0.16,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4570391809581407,
+      "grad_norm": 0.6406762286870573,
+      "learning_rate": 9.612586450910744e-06,
+      "loss": 0.1536,
+      "step": 5290
+    },
+    {
+      "epoch": 0.45790314916411073,
+      "grad_norm": 0.6507725769700754,
+      "learning_rate": 9.610784391407204e-06,
+      "loss": 0.1602,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4587671173700808,
+      "grad_norm": 0.6513419928024363,
+      "learning_rate": 9.608978320185985e-06,
+      "loss": 0.154,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4596310855760508,
+      "grad_norm": 0.6557237617400382,
+      "learning_rate": 9.607168238818496e-06,
+      "loss": 0.1607,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4604950537820208,
+      "grad_norm": 0.6489510877793241,
+      "learning_rate": 9.605354148879643e-06,
+      "loss": 0.1543,
+      "step": 5330
+    },
+    {
+      "epoch": 0.46135902198799084,
+      "grad_norm": 0.6518534162138759,
+      "learning_rate": 9.603536051947815e-06,
+      "loss": 0.1537,
+      "step": 5340
+    },
+    {
+      "epoch": 0.46222299019396085,
+      "grad_norm": 0.6435402536124452,
+      "learning_rate": 9.601713949604887e-06,
+      "loss": 0.1602,
+      "step": 5350
+    },
+    {
+      "epoch": 0.46308695839993086,
+      "grad_norm": 0.656285821302999,
+      "learning_rate": 9.599887843436224e-06,
+      "loss": 0.1575,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4639509266059009,
+      "grad_norm": 0.6734174544832082,
+      "learning_rate": 9.598057735030668e-06,
+      "loss": 0.1615,
+      "step": 5370
+    },
+    {
+      "epoch": 0.46481489481187094,
+      "grad_norm": 0.6525119845384318,
+      "learning_rate": 9.59622362598055e-06,
+      "loss": 0.1555,
+      "step": 5380
+    },
+    {
+      "epoch": 0.46567886301784095,
+      "grad_norm": 0.6424183453875718,
+      "learning_rate": 9.594385517881673e-06,
+      "loss": 0.1584,
+      "step": 5390
+    },
+    {
+      "epoch": 0.46654283122381096,
+      "grad_norm": 0.6777539885626862,
+      "learning_rate": 9.592543412333329e-06,
+      "loss": 0.1623,
+      "step": 5400
+    },
+    {
+      "epoch": 0.467406799429781,
+      "grad_norm": 0.6259425254530941,
+      "learning_rate": 9.59069731093828e-06,
+      "loss": 0.1599,
+      "step": 5410
+    },
+    {
+      "epoch": 0.468270767635751,
+      "grad_norm": 0.6932018168621533,
+      "learning_rate": 9.588847215302772e-06,
+      "loss": 0.1592,
+      "step": 5420
+    },
+    {
+      "epoch": 0.46913473584172105,
+      "grad_norm": 0.6477717328424021,
+      "learning_rate": 9.586993127036522e-06,
+      "loss": 0.1588,
+      "step": 5430
+    },
+    {
+      "epoch": 0.46999870404769106,
+      "grad_norm": 0.6592836020610073,
+      "learning_rate": 9.58513504775272e-06,
+      "loss": 0.1635,
+      "step": 5440
+    },
+    {
+      "epoch": 0.4708626722536611,
+      "grad_norm": 0.6457972290946014,
+      "learning_rate": 9.583272979068032e-06,
+      "loss": 0.1573,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4717266404596311,
+      "grad_norm": 0.6707595353255882,
+      "learning_rate": 9.581406922602593e-06,
+      "loss": 0.1528,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4725906086656011,
+      "grad_norm": 0.654924509203685,
+      "learning_rate": 9.579536879980005e-06,
+      "loss": 0.154,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4734545768715711,
+      "grad_norm": 0.6498507426859098,
+      "learning_rate": 9.577662852827345e-06,
+      "loss": 0.1554,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4743185450775411,
+      "grad_norm": 0.6503674843306472,
+      "learning_rate": 9.575784842775152e-06,
+      "loss": 0.1578,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4751825132835112,
+      "grad_norm": 0.6685006182692143,
+      "learning_rate": 9.573902851457428e-06,
+      "loss": 0.1579,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4760464814894812,
+      "grad_norm": 0.647438727181266,
+      "learning_rate": 9.572016880511645e-06,
+      "loss": 0.1581,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4769104496954512,
+      "grad_norm": 0.6524775973331061,
+      "learning_rate": 9.570126931578734e-06,
+      "loss": 0.1558,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4777744179014212,
+      "grad_norm": 0.6328300944860435,
+      "learning_rate": 9.56823300630309e-06,
+      "loss": 0.1514,
+      "step": 5530
+    },
+    {
+      "epoch": 0.47863838610739123,
+      "grad_norm": 0.633087466179911,
+      "learning_rate": 9.566335106332563e-06,
+      "loss": 0.1595,
+      "step": 5540
+    },
+    {
+      "epoch": 0.47950235431336125,
+      "grad_norm": 0.6531923998474526,
+      "learning_rate": 9.564433233318466e-06,
+      "loss": 0.1535,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4803663225193313,
+      "grad_norm": 0.635543450438123,
+      "learning_rate": 9.562527388915565e-06,
+      "loss": 0.1557,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4812302907253013,
+      "grad_norm": 0.628235427511297,
+      "learning_rate": 9.560617574782085e-06,
+      "loss": 0.1601,
+      "step": 5570
+    },
+    {
+      "epoch": 0.48209425893127134,
+      "grad_norm": 0.6322211295045682,
+      "learning_rate": 9.558703792579702e-06,
+      "loss": 0.1598,
+      "step": 5580
+    },
+    {
+      "epoch": 0.48295822713724135,
+      "grad_norm": 0.6423726739039197,
+      "learning_rate": 9.556786043973547e-06,
+      "loss": 0.1557,
+      "step": 5590
+    },
+    {
+      "epoch": 0.48382219534321136,
+      "grad_norm": 0.6773207454316515,
+      "learning_rate": 9.554864330632198e-06,
+      "loss": 0.1585,
+      "step": 5600
+    },
+    {
+      "epoch": 0.48468616354918137,
+      "grad_norm": 0.6499962659091628,
+      "learning_rate": 9.55293865422769e-06,
+      "loss": 0.1583,
+      "step": 5610
+    },
+    {
+      "epoch": 0.48555013175515144,
+      "grad_norm": 0.6653000931931303,
+      "learning_rate": 9.551009016435495e-06,
+      "loss": 0.1604,
+      "step": 5620
+    },
+    {
+      "epoch": 0.48641409996112145,
+      "grad_norm": 0.6360345932914004,
+      "learning_rate": 9.549075418934543e-06,
+      "loss": 0.1521,
+      "step": 5630
+    },
+    {
+      "epoch": 0.48727806816709146,
+      "grad_norm": 0.5933880679418415,
+      "learning_rate": 9.547137863407204e-06,
+      "loss": 0.1577,
+      "step": 5640
+    },
+    {
+      "epoch": 0.48814203637306147,
+      "grad_norm": 0.629981336471149,
+      "learning_rate": 9.545196351539292e-06,
+      "loss": 0.1532,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4890060045790315,
+      "grad_norm": 0.6555386097808079,
+      "learning_rate": 9.543250885020061e-06,
+      "loss": 0.156,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4898699727850015,
+      "grad_norm": 0.6842451255068056,
+      "learning_rate": 9.54130146554221e-06,
+      "loss": 0.1587,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4907339409909715,
+      "grad_norm": 0.6713064235267874,
+      "learning_rate": 9.539348094801877e-06,
+      "loss": 0.1552,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4915979091969416,
+      "grad_norm": 0.6564079624166437,
+      "learning_rate": 9.537390774498637e-06,
+      "loss": 0.1537,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4924618774029116,
+      "grad_norm": 0.6286290451414437,
+      "learning_rate": 9.535429506335496e-06,
+      "loss": 0.1566,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4933258456088816,
+      "grad_norm": 0.6385916147206167,
+      "learning_rate": 9.533464292018906e-06,
+      "loss": 0.1553,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4941898138148516,
+      "grad_norm": 0.6518930348205345,
+      "learning_rate": 9.531495133258742e-06,
+      "loss": 0.1544,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4950537820208216,
+      "grad_norm": 0.629072344541979,
+      "learning_rate": 9.529522031768317e-06,
+      "loss": 0.1613,
+      "step": 5730
+    },
+    {
+      "epoch": 0.49591775022679163,
+      "grad_norm": 0.6572574558805094,
+      "learning_rate": 9.527544989264375e-06,
+      "loss": 0.1553,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4967817184327617,
+      "grad_norm": 0.631342912981415,
+      "learning_rate": 9.525564007467082e-06,
+      "loss": 0.1555,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4976456866387317,
+      "grad_norm": 0.610023752445402,
+      "learning_rate": 9.523579088100041e-06,
+      "loss": 0.1532,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4985096548447017,
+      "grad_norm": 0.6376457831648065,
+      "learning_rate": 9.521590232890272e-06,
+      "loss": 0.1577,
+      "step": 5770
+    },
+    {
+      "epoch": 0.49937362305067173,
+      "grad_norm": 0.6406158974539737,
+      "learning_rate": 9.519597443568227e-06,
+      "loss": 0.1552,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5002375912566418,
+      "grad_norm": 0.6608118104247712,
+      "learning_rate": 9.517600721867775e-06,
+      "loss": 0.1617,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5011015594626118,
+      "grad_norm": 0.6153155011611277,
+      "learning_rate": 9.51560006952621e-06,
+      "loss": 0.1566,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5019655276685818,
+      "grad_norm": 0.6539057272210915,
+      "learning_rate": 9.513595488284246e-06,
+      "loss": 0.1549,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5028294958745518,
+      "grad_norm": 0.6531761984313647,
+      "learning_rate": 9.511586979886013e-06,
+      "loss": 0.1552,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5036934640805218,
+      "grad_norm": 0.638972805350294,
+      "learning_rate": 9.509574546079061e-06,
+      "loss": 0.1536,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5045574322864919,
+      "grad_norm": 0.6454598366800842,
+      "learning_rate": 9.507558188614353e-06,
+      "loss": 0.1564,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5054214004924619,
+      "grad_norm": 0.6179767848023813,
+      "learning_rate": 9.505537909246266e-06,
+      "loss": 0.1615,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5062853686984319,
+      "grad_norm": 0.6574241098778968,
+      "learning_rate": 9.50351370973259e-06,
+      "loss": 0.1506,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5071493369044019,
+      "grad_norm": 0.6180722558466789,
+      "learning_rate": 9.501485591834525e-06,
+      "loss": 0.1522,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5080133051103719,
+      "grad_norm": 0.6538193809641687,
+      "learning_rate": 9.499453557316684e-06,
+      "loss": 0.155,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5088772733163419,
+      "grad_norm": 0.6838416890378471,
+      "learning_rate": 9.497417607947081e-06,
+      "loss": 0.1547,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5097412415223119,
+      "grad_norm": 0.634166321930832,
+      "learning_rate": 9.495377745497144e-06,
+      "loss": 0.1538,
+      "step": 5900
+    },
+    {
+      "epoch": 0.510605209728282,
+      "grad_norm": 0.6333864736473793,
+      "learning_rate": 9.493333971741698e-06,
+      "loss": 0.161,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5114691779342521,
+      "grad_norm": 0.6519770013554314,
+      "learning_rate": 9.491286288458978e-06,
+      "loss": 0.1507,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5123331461402221,
+      "grad_norm": 0.6532553915478285,
+      "learning_rate": 9.489234697430613e-06,
+      "loss": 0.1548,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5131971143461921,
+      "grad_norm": 0.6360296430234144,
+      "learning_rate": 9.48717920044164e-06,
+      "loss": 0.1523,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5140610825521621,
+      "grad_norm": 0.6562037627803998,
+      "learning_rate": 9.485119799280491e-06,
+      "loss": 0.1528,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5149250507581321,
+      "grad_norm": 0.6710431539588906,
+      "learning_rate": 9.483056495738994e-06,
+      "loss": 0.1525,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5157890189641021,
+      "grad_norm": 0.6421136526052125,
+      "learning_rate": 9.480989291612372e-06,
+      "loss": 0.1546,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5166529871700721,
+      "grad_norm": 0.6391902054075337,
+      "learning_rate": 9.478918188699243e-06,
+      "loss": 0.1542,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5175169553760421,
+      "grad_norm": 0.6556659472283152,
+      "learning_rate": 9.47684318880162e-06,
+      "loss": 0.158,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5183809235820122,
+      "grad_norm": 0.6546018179543279,
+      "learning_rate": 9.474764293724898e-06,
+      "loss": 0.1572,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5192448917879822,
+      "grad_norm": 0.6154480749406304,
+      "learning_rate": 9.472681505277872e-06,
+      "loss": 0.1543,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5201088599939522,
+      "grad_norm": 0.6403268547561236,
+      "learning_rate": 9.470594825272719e-06,
+      "loss": 0.151,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5209728281999222,
+      "grad_norm": 0.6236563458442675,
+      "learning_rate": 9.468504255525e-06,
+      "loss": 0.1542,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5218367964058923,
+      "grad_norm": 0.6877503233654196,
+      "learning_rate": 9.466409797853665e-06,
+      "loss": 0.1554,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5227007646118623,
+      "grad_norm": 0.643531814258169,
+      "learning_rate": 9.464311454081041e-06,
+      "loss": 0.1532,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5235647328178323,
+      "grad_norm": 0.6805286157046766,
+      "learning_rate": 9.462209226032843e-06,
+      "loss": 0.154,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5244287010238023,
+      "grad_norm": 0.61245472642869,
+      "learning_rate": 9.460103115538161e-06,
+      "loss": 0.1558,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5252926692297724,
+      "grad_norm": 0.6232098480791012,
+      "learning_rate": 9.457993124429466e-06,
+      "loss": 0.1546,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5261566374357424,
+      "grad_norm": 0.618449033777273,
+      "learning_rate": 9.4558792545426e-06,
+      "loss": 0.1557,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5270206056417124,
+      "grad_norm": 0.6397349484531721,
+      "learning_rate": 9.453761507716787e-06,
+      "loss": 0.1542,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5278845738476824,
+      "grad_norm": 0.6278611478363766,
+      "learning_rate": 9.451639885794621e-06,
+      "loss": 0.1574,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5287485420536524,
+      "grad_norm": 0.6235855930527535,
+      "learning_rate": 9.449514390622062e-06,
+      "loss": 0.1544,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5296125102596224,
+      "grad_norm": 0.6700763747048177,
+      "learning_rate": 9.447385024048454e-06,
+      "loss": 0.1521,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5304764784655924,
+      "grad_norm": 0.6090307268762309,
+      "learning_rate": 9.445251787926492e-06,
+      "loss": 0.1574,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5313404466715624,
+      "grad_norm": 0.6427231492601245,
+      "learning_rate": 9.443114684112251e-06,
+      "loss": 0.1492,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5322044148775326,
+      "grad_norm": 0.6555482109889144,
+      "learning_rate": 9.440973714465167e-06,
+      "loss": 0.1551,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5330683830835026,
+      "grad_norm": 0.6250258619357056,
+      "learning_rate": 9.438828880848039e-06,
+      "loss": 0.1524,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5339323512894726,
+      "grad_norm": 0.6288378897514569,
+      "learning_rate": 9.436680185127026e-06,
+      "loss": 0.1536,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5347963194954426,
+      "grad_norm": 0.6291255544505886,
+      "learning_rate": 9.434527629171653e-06,
+      "loss": 0.1584,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5356602877014126,
+      "grad_norm": 0.651577330866289,
+      "learning_rate": 9.432371214854797e-06,
+      "loss": 0.1522,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5365242559073826,
+      "grad_norm": 0.6177750730266219,
+      "learning_rate": 9.430210944052696e-06,
+      "loss": 0.1559,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5373882241133526,
+      "grad_norm": 0.6229158880730679,
+      "learning_rate": 9.428046818644942e-06,
+      "loss": 0.1574,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5382521923193226,
+      "grad_norm": 0.6628200588962239,
+      "learning_rate": 9.425878840514487e-06,
+      "loss": 0.1501,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5391161605252927,
+      "grad_norm": 0.6597060635008648,
+      "learning_rate": 9.423707011547622e-06,
+      "loss": 0.1474,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5399801287312627,
+      "grad_norm": 0.6543877462107659,
+      "learning_rate": 9.421531333634e-06,
+      "loss": 0.1562,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5408440969372327,
+      "grad_norm": 0.631809391193605,
+      "learning_rate": 9.419351808666618e-06,
+      "loss": 0.1532,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5417080651432027,
+      "grad_norm": 0.597981419686076,
+      "learning_rate": 9.417168438541821e-06,
+      "loss": 0.1501,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5425720333491727,
+      "grad_norm": 0.6143884751433082,
+      "learning_rate": 9.414981225159303e-06,
+      "loss": 0.1585,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5434360015551428,
+      "grad_norm": 0.6797038536136399,
+      "learning_rate": 9.412790170422094e-06,
+      "loss": 0.1538,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5442999697611128,
+      "grad_norm": 0.613381369055774,
+      "learning_rate": 9.410595276236574e-06,
+      "loss": 0.1505,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5451639379670828,
+      "grad_norm": 0.6353511536715715,
+      "learning_rate": 9.408396544512459e-06,
+      "loss": 0.1558,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5460279061730529,
+      "grad_norm": 0.6444380767578983,
+      "learning_rate": 9.406193977162806e-06,
+      "loss": 0.1492,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5468918743790229,
+      "grad_norm": 0.6205572841742492,
+      "learning_rate": 9.403987576104009e-06,
+      "loss": 0.1563,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5477558425849929,
+      "grad_norm": 0.5971746646391922,
+      "learning_rate": 9.401777343255796e-06,
+      "loss": 0.1528,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5486198107909629,
+      "grad_norm": 0.6233544276574221,
+      "learning_rate": 9.399563280541233e-06,
+      "loss": 0.1553,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5494837789969329,
+      "grad_norm": 0.6487060550718584,
+      "learning_rate": 9.397345389886715e-06,
+      "loss": 0.1523,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5503477472029029,
+      "grad_norm": 0.6296095891164247,
+      "learning_rate": 9.395123673221965e-06,
+      "loss": 0.1553,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5512117154088729,
+      "grad_norm": 0.7315998401798692,
+      "learning_rate": 9.39289813248004e-06,
+      "loss": 0.1567,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5520756836148429,
+      "grad_norm": 0.6450303515918221,
+      "learning_rate": 9.390668769597323e-06,
+      "loss": 0.1456,
+      "step": 6390
+    },
+    {
+      "epoch": 0.552939651820813,
+      "grad_norm": 0.6353752953847911,
+      "learning_rate": 9.38843558651352e-06,
+      "loss": 0.1591,
+      "step": 6400
+    },
+    {
+      "epoch": 0.553803620026783,
+      "grad_norm": 0.639900060325134,
+      "learning_rate": 9.386198585171666e-06,
+      "loss": 0.1582,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5546675882327531,
+      "grad_norm": 0.631243331570897,
+      "learning_rate": 9.38395776751811e-06,
+      "loss": 0.1535,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5555315564387231,
+      "grad_norm": 0.6120386993003224,
+      "learning_rate": 9.381713135502531e-06,
+      "loss": 0.1504,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5563955246446931,
+      "grad_norm": 0.6218134724182738,
+      "learning_rate": 9.379464691077919e-06,
+      "loss": 0.1523,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5572594928506631,
+      "grad_norm": 0.7523102118837981,
+      "learning_rate": 9.377212436200587e-06,
+      "loss": 0.1521,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5581234610566331,
+      "grad_norm": 0.6557050210004696,
+      "learning_rate": 9.374956372830158e-06,
+      "loss": 0.1561,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5589874292626031,
+      "grad_norm": 0.6201510619086964,
+      "learning_rate": 9.37269650292957e-06,
+      "loss": 0.155,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5598513974685732,
+      "grad_norm": 0.5811173881087803,
+      "learning_rate": 9.370432828465079e-06,
+      "loss": 0.1551,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5607153656745432,
+      "grad_norm": 0.6012143606073025,
+      "learning_rate": 9.36816535140624e-06,
+      "loss": 0.1543,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5615793338805132,
+      "grad_norm": 0.5825061635910226,
+      "learning_rate": 9.365894073725929e-06,
+      "loss": 0.1548,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5624433020864832,
+      "grad_norm": 0.6375174928498161,
+      "learning_rate": 9.363618997400319e-06,
+      "loss": 0.1542,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5633072702924532,
+      "grad_norm": 0.6218539072355832,
+      "learning_rate": 9.361340124408893e-06,
+      "loss": 0.1504,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5641712384984232,
+      "grad_norm": 0.6377893763114578,
+      "learning_rate": 9.359057456734437e-06,
+      "loss": 0.1528,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5650352067043933,
+      "grad_norm": 0.6059514862386471,
+      "learning_rate": 9.356770996363034e-06,
+      "loss": 0.1536,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5658991749103633,
+      "grad_norm": 0.6782562253690789,
+      "learning_rate": 9.354480745284075e-06,
+      "loss": 0.1522,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5667631431163334,
+      "grad_norm": 0.6530575495241211,
+      "learning_rate": 9.352186705490245e-06,
+      "loss": 0.1564,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5676271113223034,
+      "grad_norm": 0.6271306204208658,
+      "learning_rate": 9.349888878977525e-06,
+      "loss": 0.1579,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5684910795282734,
+      "grad_norm": 0.655379907159268,
+      "learning_rate": 9.347587267745188e-06,
+      "loss": 0.1587,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5693550477342434,
+      "grad_norm": 0.6207039064949125,
+      "learning_rate": 9.345281873795807e-06,
+      "loss": 0.1502,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5702190159402134,
+      "grad_norm": 0.6202852350840276,
+      "learning_rate": 9.342972699135238e-06,
+      "loss": 0.1482,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5710829841461834,
+      "grad_norm": 0.6170685026124321,
+      "learning_rate": 9.340659745772635e-06,
+      "loss": 0.1528,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5719469523521534,
+      "grad_norm": 0.6108278186924734,
+      "learning_rate": 9.338343015720434e-06,
+      "loss": 0.153,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5728109205581234,
+      "grad_norm": 0.6451428011838746,
+      "learning_rate": 9.33602251099436e-06,
+      "loss": 0.1483,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5736748887640934,
+      "grad_norm": 0.6672803417700074,
+      "learning_rate": 9.33369823361342e-06,
+      "loss": 0.1525,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5745388569700635,
+      "grad_norm": 0.638138381181463,
+      "learning_rate": 9.331370185599902e-06,
+      "loss": 0.1526,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5754028251760335,
+      "grad_norm": 0.6285291919757152,
+      "learning_rate": 9.32903836897938e-06,
+      "loss": 0.1579,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5762667933820036,
+      "grad_norm": 0.5894093953466101,
+      "learning_rate": 9.326702785780704e-06,
+      "loss": 0.1527,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5771307615879736,
+      "grad_norm": 0.6363863884972929,
+      "learning_rate": 9.324363438035998e-06,
+      "loss": 0.1522,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5779947297939436,
+      "grad_norm": 0.5954359232987835,
+      "learning_rate": 9.322020327780667e-06,
+      "loss": 0.147,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5788586979999136,
+      "grad_norm": 0.5924570670614575,
+      "learning_rate": 9.319673457053389e-06,
+      "loss": 0.1554,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5797226662058836,
+      "grad_norm": 0.6557729717170221,
+      "learning_rate": 9.31732282789611e-06,
+      "loss": 0.1548,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5805866344118537,
+      "grad_norm": 0.6252295859666467,
+      "learning_rate": 9.314968442354048e-06,
+      "loss": 0.1519,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5814506026178237,
+      "grad_norm": 0.6104984510258042,
+      "learning_rate": 9.312610302475691e-06,
+      "loss": 0.1545,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5823145708237937,
+      "grad_norm": 0.6084878036937621,
+      "learning_rate": 9.31024841031279e-06,
+      "loss": 0.1509,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5831785390297637,
+      "grad_norm": 0.615179111407142,
+      "learning_rate": 9.307882767920365e-06,
+      "loss": 0.1496,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5840425072357337,
+      "grad_norm": 0.6335012641621653,
+      "learning_rate": 9.305513377356696e-06,
+      "loss": 0.159,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5849064754417037,
+      "grad_norm": 0.6432893824981711,
+      "learning_rate": 9.303140240683326e-06,
+      "loss": 0.1558,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5857704436476737,
+      "grad_norm": 0.6288129691404086,
+      "learning_rate": 9.300763359965057e-06,
+      "loss": 0.1529,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5866344118536437,
+      "grad_norm": 0.5996719948442747,
+      "learning_rate": 9.298382737269944e-06,
+      "loss": 0.1561,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5874983800596139,
+      "grad_norm": 0.6318777413266485,
+      "learning_rate": 9.295998374669307e-06,
+      "loss": 0.1532,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5883623482655839,
+      "grad_norm": 0.6393146031927769,
+      "learning_rate": 9.29361027423771e-06,
+      "loss": 0.154,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5892263164715539,
+      "grad_norm": 0.6087589722272293,
+      "learning_rate": 9.291218438052978e-06,
+      "loss": 0.1449,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5900902846775239,
+      "grad_norm": 0.6053164128500995,
+      "learning_rate": 9.28882286819618e-06,
+      "loss": 0.1543,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5909542528834939,
+      "grad_norm": 0.6168383132645895,
+      "learning_rate": 9.286423566751638e-06,
+      "loss": 0.153,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5918182210894639,
+      "grad_norm": 0.6357370911729846,
+      "learning_rate": 9.284020535806917e-06,
+      "loss": 0.1493,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5926821892954339,
+      "grad_norm": 0.5856044394038176,
+      "learning_rate": 9.28161377745283e-06,
+      "loss": 0.1526,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5935461575014039,
+      "grad_norm": 0.595353891775203,
+      "learning_rate": 9.27920329378343e-06,
+      "loss": 0.1521,
+      "step": 6870
+    },
+    {
+      "epoch": 0.594410125707374,
+      "grad_norm": 0.5879333218876476,
+      "learning_rate": 9.276789086896015e-06,
+      "loss": 0.1517,
+      "step": 6880
+    },
+    {
+      "epoch": 0.595274093913344,
+      "grad_norm": 0.6437671621448833,
+      "learning_rate": 9.274371158891117e-06,
+      "loss": 0.1528,
+      "step": 6890
+    },
+    {
+      "epoch": 0.596138062119314,
+      "grad_norm": 0.6362829862629101,
+      "learning_rate": 9.271949511872514e-06,
+      "loss": 0.1574,
+      "step": 6900
+    },
+    {
+      "epoch": 0.597002030325284,
+      "grad_norm": 0.6324049229309815,
+      "learning_rate": 9.269524147947214e-06,
+      "loss": 0.1537,
+      "step": 6910
+    },
+    {
+      "epoch": 0.597865998531254,
+      "grad_norm": 0.6329882824337885,
+      "learning_rate": 9.267095069225456e-06,
+      "loss": 0.1519,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5987299667372241,
+      "grad_norm": 0.6005996025691578,
+      "learning_rate": 9.264662277820719e-06,
+      "loss": 0.1512,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5995939349431941,
+      "grad_norm": 0.6040443399864707,
+      "learning_rate": 9.262225775849707e-06,
+      "loss": 0.1525,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6004579031491641,
+      "grad_norm": 0.6159608848677605,
+      "learning_rate": 9.259785565432356e-06,
+      "loss": 0.1533,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6013218713551342,
+      "grad_norm": 0.6149382032923227,
+      "learning_rate": 9.257341648691822e-06,
+      "loss": 0.1509,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6021858395611042,
+      "grad_norm": 0.6290543016885738,
+      "learning_rate": 9.254894027754493e-06,
+      "loss": 0.1539,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6030498077670742,
+      "grad_norm": 0.5930637622387478,
+      "learning_rate": 9.25244270474998e-06,
+      "loss": 0.1543,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6039137759730442,
+      "grad_norm": 0.5657882487334296,
+      "learning_rate": 9.249987681811106e-06,
+      "loss": 0.1492,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6047777441790142,
+      "grad_norm": 0.6095682916220933,
+      "learning_rate": 9.247528961073925e-06,
+      "loss": 0.1516,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6056417123849842,
+      "grad_norm": 0.5992451055616469,
+      "learning_rate": 9.2450665446777e-06,
+      "loss": 0.1534,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6065056805909542,
+      "grad_norm": 0.5997401846439975,
+      "learning_rate": 9.242600434764912e-06,
+      "loss": 0.1521,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6073696487969242,
+      "grad_norm": 0.5933488089940263,
+      "learning_rate": 9.240130633481259e-06,
+      "loss": 0.1518,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6082336170028942,
+      "grad_norm": 0.6313843488784524,
+      "learning_rate": 9.237657142975643e-06,
+      "loss": 0.1538,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6090975852088644,
+      "grad_norm": 0.6413521714283614,
+      "learning_rate": 9.235179965400184e-06,
+      "loss": 0.152,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6099615534148344,
+      "grad_norm": 0.6184470237558511,
+      "learning_rate": 9.232699102910208e-06,
+      "loss": 0.1504,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6108255216208044,
+      "grad_norm": 0.5707911485975334,
+      "learning_rate": 9.230214557664241e-06,
+      "loss": 0.1504,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6116894898267744,
+      "grad_norm": 0.5777559907197692,
+      "learning_rate": 9.227726331824021e-06,
+      "loss": 0.1496,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6125534580327444,
+      "grad_norm": 0.5954819404017806,
+      "learning_rate": 9.225234427554485e-06,
+      "loss": 0.1509,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6134174262387144,
+      "grad_norm": 0.6192745295620272,
+      "learning_rate": 9.222738847023772e-06,
+      "loss": 0.151,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6142813944446844,
+      "grad_norm": 0.6038949399030311,
+      "learning_rate": 9.220239592403216e-06,
+      "loss": 0.1491,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6151453626506544,
+      "grad_norm": 0.5824949970844027,
+      "learning_rate": 9.217736665867352e-06,
+      "loss": 0.1473,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6160093308566245,
+      "grad_norm": 0.5954051916023153,
+      "learning_rate": 9.215230069593907e-06,
+      "loss": 0.1525,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6168732990625945,
+      "grad_norm": 0.6320195114436121,
+      "learning_rate": 9.212719805763806e-06,
+      "loss": 0.1541,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6177372672685645,
+      "grad_norm": 0.6929376486569512,
+      "learning_rate": 9.210205876561153e-06,
+      "loss": 0.1493,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6186012354745345,
+      "grad_norm": 0.6159597724690932,
+      "learning_rate": 9.207688284173257e-06,
+      "loss": 0.1489,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6194652036805045,
+      "grad_norm": 0.5943323896784803,
+      "learning_rate": 9.205167030790604e-06,
+      "loss": 0.1486,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6203291718864746,
+      "grad_norm": 0.6204073730643097,
+      "learning_rate": 9.202642118606866e-06,
+      "loss": 0.1503,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6211931400924446,
+      "grad_norm": 0.6241579805429692,
+      "learning_rate": 9.2001135498189e-06,
+      "loss": 0.1532,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6220571082984147,
+      "grad_norm": 0.6030366378179512,
+      "learning_rate": 9.19758132662675e-06,
+      "loss": 0.149,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6229210765043847,
+      "grad_norm": 0.6071124192692541,
+      "learning_rate": 9.195045451233627e-06,
+      "loss": 0.1541,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6237850447103547,
+      "grad_norm": 0.6158999535197959,
+      "learning_rate": 9.192505925845932e-06,
+      "loss": 0.1527,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6246490129163247,
+      "grad_norm": 0.5640793686230424,
+      "learning_rate": 9.189962752673234e-06,
+      "loss": 0.1465,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6255129811222947,
+      "grad_norm": 0.5975619759720371,
+      "learning_rate": 9.187415933928279e-06,
+      "loss": 0.1525,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6263769493282647,
+      "grad_norm": 0.5968403447579432,
+      "learning_rate": 9.184865471826988e-06,
+      "loss": 0.1506,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6272409175342347,
+      "grad_norm": 0.6042528506000148,
+      "learning_rate": 9.182311368588444e-06,
+      "loss": 0.1486,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6281048857402047,
+      "grad_norm": 0.606460473515699,
+      "learning_rate": 9.179753626434905e-06,
+      "loss": 0.1505,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6289688539461747,
+      "grad_norm": 0.5903817090048041,
+      "learning_rate": 9.17719224759179e-06,
+      "loss": 0.1518,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6298328221521448,
+      "grad_norm": 0.5952765730532862,
+      "learning_rate": 9.174627234287688e-06,
+      "loss": 0.1492,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6306967903581148,
+      "grad_norm": 0.622409000843169,
+      "learning_rate": 9.172058588754345e-06,
+      "loss": 0.1529,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6315607585640849,
+      "grad_norm": 0.57758523085189,
+      "learning_rate": 9.169486313226671e-06,
+      "loss": 0.1506,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6324247267700549,
+      "grad_norm": 0.6142820795880717,
+      "learning_rate": 9.166910409942731e-06,
+      "loss": 0.1486,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6332886949760249,
+      "grad_norm": 0.5969544846944455,
+      "learning_rate": 9.16433088114375e-06,
+      "loss": 0.152,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6341526631819949,
+      "grad_norm": 0.6077472401893858,
+      "learning_rate": 9.161747729074105e-06,
+      "loss": 0.1546,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6350166313879649,
+      "grad_norm": 0.6153979176395339,
+      "learning_rate": 9.159160955981326e-06,
+      "loss": 0.1519,
+      "step": 7350
+    },
+    {
+      "epoch": 0.635880599593935,
+      "grad_norm": 0.5856990458728484,
+      "learning_rate": 9.156570564116092e-06,
+      "loss": 0.1509,
+      "step": 7360
+    },
+    {
+      "epoch": 0.636744567799905,
+      "grad_norm": 0.5805924301676306,
+      "learning_rate": 9.153976555732233e-06,
+      "loss": 0.1441,
+      "step": 7370
+    },
+    {
+      "epoch": 0.637608536005875,
+      "grad_norm": 0.6405542270797587,
+      "learning_rate": 9.151378933086728e-06,
+      "loss": 0.1536,
+      "step": 7380
+    },
+    {
+      "epoch": 0.638472504211845,
+      "grad_norm": 0.6087345540466436,
+      "learning_rate": 9.148777698439695e-06,
+      "loss": 0.1532,
+      "step": 7390
+    },
+    {
+      "epoch": 0.639336472417815,
+      "grad_norm": 0.5881093647581587,
+      "learning_rate": 9.146172854054395e-06,
+      "loss": 0.1497,
+      "step": 7400
+    },
+    {
+      "epoch": 0.640200440623785,
+      "grad_norm": 0.653373547737377,
+      "learning_rate": 9.143564402197239e-06,
+      "loss": 0.1577,
+      "step": 7410
+    },
+    {
+      "epoch": 0.641064408829755,
+      "grad_norm": 0.5725873284917496,
+      "learning_rate": 9.140952345137762e-06,
+      "loss": 0.1482,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6419283770357251,
+      "grad_norm": 0.6062841375115021,
+      "learning_rate": 9.138336685148648e-06,
+      "loss": 0.1519,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6427923452416952,
+      "grad_norm": 0.6390131647406511,
+      "learning_rate": 9.13571742450571e-06,
+      "loss": 0.1472,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6436563134476652,
+      "grad_norm": 0.6325898539379184,
+      "learning_rate": 9.133094565487894e-06,
+      "loss": 0.153,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6445202816536352,
+      "grad_norm": 0.6244419302476996,
+      "learning_rate": 9.130468110377283e-06,
+      "loss": 0.1481,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6453842498596052,
+      "grad_norm": 0.5917216909133591,
+      "learning_rate": 9.127838061459077e-06,
+      "loss": 0.1481,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6462482180655752,
+      "grad_norm": 0.6155031693074737,
+      "learning_rate": 9.125204421021616e-06,
+      "loss": 0.1499,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6471121862715452,
+      "grad_norm": 0.602270438150178,
+      "learning_rate": 9.122567191356355e-06,
+      "loss": 0.1495,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6479761544775152,
+      "grad_norm": 0.6426993093044683,
+      "learning_rate": 9.119926374757876e-06,
+      "loss": 0.1527,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6488401226834852,
+      "grad_norm": 0.5912880315934217,
+      "learning_rate": 9.117281973523882e-06,
+      "loss": 0.1501,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6497040908894552,
+      "grad_norm": 0.6244924057208089,
+      "learning_rate": 9.114633989955194e-06,
+      "loss": 0.1519,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6505680590954253,
+      "grad_norm": 0.611934191875604,
+      "learning_rate": 9.111982426355753e-06,
+      "loss": 0.1469,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6514320273013953,
+      "grad_norm": 0.5864263801510056,
+      "learning_rate": 9.109327285032607e-06,
+      "loss": 0.149,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6522959955073653,
+      "grad_norm": 0.6010224495299872,
+      "learning_rate": 9.106668568295927e-06,
+      "loss": 0.1459,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6531599637133354,
+      "grad_norm": 0.6050030904357648,
+      "learning_rate": 9.104006278458986e-06,
+      "loss": 0.1458,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6540239319193054,
+      "grad_norm": 0.5676610160897111,
+      "learning_rate": 9.101340417838171e-06,
+      "loss": 0.1487,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6548879001252754,
+      "grad_norm": 0.6592513736662908,
+      "learning_rate": 9.098670988752975e-06,
+      "loss": 0.1479,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6557518683312454,
+      "grad_norm": 0.5927040039853829,
+      "learning_rate": 9.095997993525999e-06,
+      "loss": 0.1557,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6566158365372154,
+      "grad_norm": 0.6020603854415782,
+      "learning_rate": 9.093321434482935e-06,
+      "loss": 0.1571,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6574798047431855,
+      "grad_norm": 0.5872803841047225,
+      "learning_rate": 9.09064131395259e-06,
+      "loss": 0.1528,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6583437729491555,
+      "grad_norm": 0.5901679825406331,
+      "learning_rate": 9.087957634266862e-06,
+      "loss": 0.1502,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6592077411551255,
+      "grad_norm": 0.6311244536626748,
+      "learning_rate": 9.085270397760748e-06,
+      "loss": 0.1518,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6600717093610955,
+      "grad_norm": 0.5872882598881509,
+      "learning_rate": 9.082579606772339e-06,
+      "loss": 0.1528,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6609356775670655,
+      "grad_norm": 0.5882786648598662,
+      "learning_rate": 9.079885263642818e-06,
+      "loss": 0.1524,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6617996457730355,
+      "grad_norm": 0.5823254581363112,
+      "learning_rate": 9.077187370716461e-06,
+      "loss": 0.1505,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6626636139790055,
+      "grad_norm": 0.6065493971147171,
+      "learning_rate": 9.074485930340631e-06,
+      "loss": 0.1466,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6635275821849755,
+      "grad_norm": 0.6491173159325675,
+      "learning_rate": 9.071780944865775e-06,
+      "loss": 0.1472,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6643915503909457,
+      "grad_norm": 0.6409731127559477,
+      "learning_rate": 9.06907241664543e-06,
+      "loss": 0.1511,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6652555185969157,
+      "grad_norm": 0.581674663711164,
+      "learning_rate": 9.066360348036211e-06,
+      "loss": 0.1459,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6661194868028857,
+      "grad_norm": 0.6084235174422291,
+      "learning_rate": 9.063644741397814e-06,
+      "loss": 0.1534,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6669834550088557,
+      "grad_norm": 0.6142204723314244,
+      "learning_rate": 9.060925599093015e-06,
+      "loss": 0.1503,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6678474232148257,
+      "grad_norm": 0.6210363271358172,
+      "learning_rate": 9.058202923487669e-06,
+      "loss": 0.1501,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6687113914207957,
+      "grad_norm": 0.6328872498299684,
+      "learning_rate": 9.055476716950697e-06,
+      "loss": 0.15,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6695753596267657,
+      "grad_norm": 0.6076328150996245,
+      "learning_rate": 9.052746981854097e-06,
+      "loss": 0.1475,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6704393278327357,
+      "grad_norm": 0.5985368263170588,
+      "learning_rate": 9.050013720572941e-06,
+      "loss": 0.1479,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6713032960387058,
+      "grad_norm": 0.6388889436940819,
+      "learning_rate": 9.04727693548536e-06,
+      "loss": 0.1526,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6721672642446758,
+      "grad_norm": 0.603154034262408,
+      "learning_rate": 9.04453662897256e-06,
+      "loss": 0.1464,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6730312324506458,
+      "grad_norm": 0.5762078420887212,
+      "learning_rate": 9.041792803418808e-06,
+      "loss": 0.148,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6738952006566158,
+      "grad_norm": 0.6146161683014285,
+      "learning_rate": 9.039045461211426e-06,
+      "loss": 0.1521,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6747591688625859,
+      "grad_norm": 0.5739686729569367,
+      "learning_rate": 9.036294604740805e-06,
+      "loss": 0.1519,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6756231370685559,
+      "grad_norm": 0.5868335044090508,
+      "learning_rate": 9.03354023640039e-06,
+      "loss": 0.1497,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6764871052745259,
+      "grad_norm": 0.6111017627175641,
+      "learning_rate": 9.030782358586684e-06,
+      "loss": 0.1528,
+      "step": 7830
+    },
+    {
+      "epoch": 0.677351073480496,
+      "grad_norm": 0.6252200831513823,
+      "learning_rate": 9.028020973699237e-06,
+      "loss": 0.1524,
+      "step": 7840
+    },
+    {
+      "epoch": 0.678215041686466,
+      "grad_norm": 0.6345447586951666,
+      "learning_rate": 9.025256084140656e-06,
+      "loss": 0.1502,
+      "step": 7850
+    },
+    {
+      "epoch": 0.679079009892436,
+      "grad_norm": 0.6300359296600807,
+      "learning_rate": 9.022487692316599e-06,
+      "loss": 0.1495,
+      "step": 7860
+    },
+    {
+      "epoch": 0.679942978098406,
+      "grad_norm": 0.5918604092279216,
+      "learning_rate": 9.019715800635764e-06,
+      "loss": 0.1482,
+      "step": 7870
+    },
+    {
+      "epoch": 0.680806946304376,
+      "grad_norm": 0.552573528938169,
+      "learning_rate": 9.0169404115099e-06,
+      "loss": 0.1503,
+      "step": 7880
+    },
+    {
+      "epoch": 0.681670914510346,
+      "grad_norm": 0.6159373617370572,
+      "learning_rate": 9.014161527353798e-06,
+      "loss": 0.1482,
+      "step": 7890
+    },
+    {
+      "epoch": 0.682534882716316,
+      "grad_norm": 0.5782099461611947,
+      "learning_rate": 9.01137915058529e-06,
+      "loss": 0.155,
+      "step": 7900
+    },
+    {
+      "epoch": 0.683398850922286,
+      "grad_norm": 0.5935877354981174,
+      "learning_rate": 9.008593283625242e-06,
+      "loss": 0.149,
+      "step": 7910
+    },
+    {
+      "epoch": 0.684262819128256,
+      "grad_norm": 0.5559859018366894,
+      "learning_rate": 9.005803928897563e-06,
+      "loss": 0.1466,
+      "step": 7920
+    },
+    {
+      "epoch": 0.685126787334226,
+      "grad_norm": 0.6204828417543897,
+      "learning_rate": 9.003011088829197e-06,
+      "loss": 0.1509,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6859907555401962,
+      "grad_norm": 0.5691521373817178,
+      "learning_rate": 9.000214765850115e-06,
+      "loss": 0.1468,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6868547237461662,
+      "grad_norm": 0.5852804581194763,
+      "learning_rate": 8.997414962393323e-06,
+      "loss": 0.1463,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6877186919521362,
+      "grad_norm": 0.6049762063656603,
+      "learning_rate": 8.994611680894853e-06,
+      "loss": 0.1471,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6885826601581062,
+      "grad_norm": 0.5777602090413144,
+      "learning_rate": 8.991804923793762e-06,
+      "loss": 0.1494,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6894466283640762,
+      "grad_norm": 0.6044551263155569,
+      "learning_rate": 8.988994693532136e-06,
+      "loss": 0.1469,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6903105965700462,
+      "grad_norm": 0.5783047865363863,
+      "learning_rate": 8.986180992555077e-06,
+      "loss": 0.148,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6911745647760162,
+      "grad_norm": 0.6233075599008563,
+      "learning_rate": 8.983363823310712e-06,
+      "loss": 0.1525,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6920385329819863,
+      "grad_norm": 0.6006254139480321,
+      "learning_rate": 8.98054318825018e-06,
+      "loss": 0.1509,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6929025011879563,
+      "grad_norm": 0.5769597479586498,
+      "learning_rate": 8.977719089827638e-06,
+      "loss": 0.1428,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6937664693939263,
+      "grad_norm": 0.5885334854420028,
+      "learning_rate": 8.974891530500263e-06,
+      "loss": 0.1461,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6946304375998963,
+      "grad_norm": 0.58922621960749,
+      "learning_rate": 8.97206051272823e-06,
+      "loss": 0.1468,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6954944058058663,
+      "grad_norm": 0.5900619322694628,
+      "learning_rate": 8.969226038974737e-06,
+      "loss": 0.1489,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6963583740118363,
+      "grad_norm": 0.5721375599740359,
+      "learning_rate": 8.966388111705977e-06,
+      "loss": 0.1445,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6972223422178064,
+      "grad_norm": 0.6060535974988339,
+      "learning_rate": 8.963546733391155e-06,
+      "loss": 0.1513,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6980863104237764,
+      "grad_norm": 0.6121984155980555,
+      "learning_rate": 8.960701906502476e-06,
+      "loss": 0.1502,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6989502786297465,
+      "grad_norm": 0.5883564250085476,
+      "learning_rate": 8.957853633515148e-06,
+      "loss": 0.1481,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6998142468357165,
+      "grad_norm": 0.5995778157693623,
+      "learning_rate": 8.955001916907376e-06,
+      "loss": 0.1516,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7006782150416865,
+      "grad_norm": 0.6060847874402603,
+      "learning_rate": 8.952146759160356e-06,
+      "loss": 0.1488,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7015421832476565,
+      "grad_norm": 0.5781934980960773,
+      "learning_rate": 8.949288162758287e-06,
+      "loss": 0.1492,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7024061514536265,
+      "grad_norm": 0.6051536158517672,
+      "learning_rate": 8.946426130188357e-06,
+      "loss": 0.1503,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7032701196595965,
+      "grad_norm": 0.6139354449588857,
+      "learning_rate": 8.943560663940739e-06,
+      "loss": 0.1495,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7041340878655665,
+      "grad_norm": 0.6126263867946146,
+      "learning_rate": 8.940691766508597e-06,
+      "loss": 0.1476,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7049980560715365,
+      "grad_norm": 0.6302258584565575,
+      "learning_rate": 8.937819440388086e-06,
+      "loss": 0.1473,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7058620242775066,
+      "grad_norm": 0.5923426197318572,
+      "learning_rate": 8.93494368807833e-06,
+      "loss": 0.1448,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7067259924834766,
+      "grad_norm": 0.5819670560556821,
+      "learning_rate": 8.93206451208145e-06,
+      "loss": 0.1471,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7075899606894466,
+      "grad_norm": 0.5704325384697982,
+      "learning_rate": 8.929181914902532e-06,
+      "loss": 0.1484,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7084539288954167,
+      "grad_norm": 0.5854395681335803,
+      "learning_rate": 8.926295899049651e-06,
+      "loss": 0.1467,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7093178971013867,
+      "grad_norm": 0.5956571082587886,
+      "learning_rate": 8.923406467033846e-06,
+      "loss": 0.1502,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7101818653073567,
+      "grad_norm": 0.6013675361000554,
+      "learning_rate": 8.920513621369138e-06,
+      "loss": 0.1421,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7110458335133267,
+      "grad_norm": 0.6169527609070847,
+      "learning_rate": 8.917617364572509e-06,
+      "loss": 0.1457,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7119098017192967,
+      "grad_norm": 0.615664017102032,
+      "learning_rate": 8.914717699163913e-06,
+      "loss": 0.1528,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7127737699252668,
+      "grad_norm": 0.5929835165146116,
+      "learning_rate": 8.91181462766627e-06,
+      "loss": 0.1528,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7136377381312368,
+      "grad_norm": 0.5769838984731995,
+      "learning_rate": 8.908908152605468e-06,
+      "loss": 0.1467,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7145017063372068,
+      "grad_norm": 0.5979484694169591,
+      "learning_rate": 8.905998276510347e-06,
+      "loss": 0.1495,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7153656745431768,
+      "grad_norm": 0.6250913684486913,
+      "learning_rate": 8.903085001912707e-06,
+      "loss": 0.1503,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7162296427491468,
+      "grad_norm": 0.5743702279111842,
+      "learning_rate": 8.90016833134732e-06,
+      "loss": 0.1521,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7170936109551168,
+      "grad_norm": 0.5800789077997158,
+      "learning_rate": 8.89724826735189e-06,
+      "loss": 0.1488,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7179575791610868,
+      "grad_norm": 0.6123653987611494,
+      "learning_rate": 8.894324812467092e-06,
+      "loss": 0.1411,
+      "step": 8310
+    },
+    {
+      "epoch": 0.718821547367057,
+      "grad_norm": 0.6413369452138997,
+      "learning_rate": 8.891397969236541e-06,
+      "loss": 0.1465,
+      "step": 8320
+    },
+    {
+      "epoch": 0.719685515573027,
+      "grad_norm": 0.6314994629599093,
+      "learning_rate": 8.888467740206805e-06,
+      "loss": 0.1482,
+      "step": 8330
+    },
+    {
+      "epoch": 0.720549483778997,
+      "grad_norm": 0.5891434300309477,
+      "learning_rate": 8.885534127927397e-06,
+      "loss": 0.1478,
+      "step": 8340
+    },
+    {
+      "epoch": 0.721413451984967,
+      "grad_norm": 0.5858090399630496,
+      "learning_rate": 8.882597134950772e-06,
+      "loss": 0.1472,
+      "step": 8350
+    },
+    {
+      "epoch": 0.722277420190937,
+      "grad_norm": 0.5638968847011544,
+      "learning_rate": 8.879656763832327e-06,
+      "loss": 0.1425,
+      "step": 8360
+    },
+    {
+      "epoch": 0.723141388396907,
+      "grad_norm": 0.6067904865153183,
+      "learning_rate": 8.876713017130398e-06,
+      "loss": 0.1459,
+      "step": 8370
+    },
+    {
+      "epoch": 0.724005356602877,
+      "grad_norm": 0.5851959853995922,
+      "learning_rate": 8.87376589740626e-06,
+      "loss": 0.1423,
+      "step": 8380
+    },
+    {
+      "epoch": 0.724869324808847,
+      "grad_norm": 0.5670882649673757,
+      "learning_rate": 8.870815407224121e-06,
+      "loss": 0.1422,
+      "step": 8390
+    },
+    {
+      "epoch": 0.725733293014817,
+      "grad_norm": 0.5792761245461205,
+      "learning_rate": 8.867861549151123e-06,
+      "loss": 0.1511,
+      "step": 8400
+    },
+    {
+      "epoch": 0.726597261220787,
+      "grad_norm": 0.6384499720411182,
+      "learning_rate": 8.864904325757336e-06,
+      "loss": 0.1458,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7274612294267571,
+      "grad_norm": 0.6239567584341648,
+      "learning_rate": 8.861943739615761e-06,
+      "loss": 0.1481,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7283251976327271,
+      "grad_norm": 0.6526576277949593,
+      "learning_rate": 8.85897979330232e-06,
+      "loss": 0.1457,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7291891658386971,
+      "grad_norm": 0.6192375294210014,
+      "learning_rate": 8.856012489395865e-06,
+      "loss": 0.1515,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7300531340446672,
+      "grad_norm": 0.5829504676468671,
+      "learning_rate": 8.853041830478165e-06,
+      "loss": 0.1443,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7309171022506372,
+      "grad_norm": 0.6240893602044929,
+      "learning_rate": 8.85006781913391e-06,
+      "loss": 0.1468,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7317810704566072,
+      "grad_norm": 0.6402177545734322,
+      "learning_rate": 8.847090457950704e-06,
+      "loss": 0.1504,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7326450386625772,
+      "grad_norm": 0.6354668520522381,
+      "learning_rate": 8.84410974951907e-06,
+      "loss": 0.1463,
+      "step": 8480
+    },
+    {
+      "epoch": 0.7335090068685473,
+      "grad_norm": 0.5727464227472098,
+      "learning_rate": 8.841125696432438e-06,
+      "loss": 0.1498,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7343729750745173,
+      "grad_norm": 0.6018868846257793,
+      "learning_rate": 8.838138301287156e-06,
+      "loss": 0.1467,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7352369432804873,
+      "grad_norm": 0.5644741322670542,
+      "learning_rate": 8.835147566682472e-06,
+      "loss": 0.1463,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7361009114864573,
+      "grad_norm": 0.5673974911271266,
+      "learning_rate": 8.832153495220543e-06,
+      "loss": 0.1475,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7369648796924273,
+      "grad_norm": 0.5964122079459192,
+      "learning_rate": 8.829156089506427e-06,
+      "loss": 0.1503,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7378288478983973,
+      "grad_norm": 0.5715091500499475,
+      "learning_rate": 8.826155352148089e-06,
+      "loss": 0.143,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7386928161043673,
+      "grad_norm": 0.6035211545064121,
+      "learning_rate": 8.823151285756383e-06,
+      "loss": 0.1494,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7395567843103373,
+      "grad_norm": 0.5289010319710973,
+      "learning_rate": 8.820143892945072e-06,
+      "loss": 0.1465,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7404207525163073,
+      "grad_norm": 0.6314524880705613,
+      "learning_rate": 8.817133176330802e-06,
+      "loss": 0.1486,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7412847207222775,
+      "grad_norm": 0.6019438096859134,
+      "learning_rate": 8.814119138533117e-06,
+      "loss": 0.149,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7421486889282475,
+      "grad_norm": 0.5685350683665455,
+      "learning_rate": 8.811101782174447e-06,
+      "loss": 0.1482,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7430126571342175,
+      "grad_norm": 0.5672312522177472,
+      "learning_rate": 8.808081109880113e-06,
+      "loss": 0.149,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7438766253401875,
+      "grad_norm": 0.6033392735761053,
+      "learning_rate": 8.80505712427832e-06,
+      "loss": 0.1478,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7447405935461575,
+      "grad_norm": 0.5852922151350154,
+      "learning_rate": 8.802029828000157e-06,
+      "loss": 0.1461,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7456045617521275,
+      "grad_norm": 0.5895046658401459,
+      "learning_rate": 8.798999223679586e-06,
+      "loss": 0.1462,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7464685299580975,
+      "grad_norm": 0.5961991481966116,
+      "learning_rate": 8.79596531395346e-06,
+      "loss": 0.1462,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7473324981640675,
+      "grad_norm": 0.5457241905915233,
+      "learning_rate": 8.792928101461493e-06,
+      "loss": 0.1482,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7481964663700376,
+      "grad_norm": 0.5709887805971291,
+      "learning_rate": 8.789887588846288e-06,
+      "loss": 0.1464,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7490604345760076,
+      "grad_norm": 0.6179251161053763,
+      "learning_rate": 8.786843778753311e-06,
+      "loss": 0.1512,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7499244027819776,
+      "grad_norm": 0.5417188053967987,
+      "learning_rate": 8.783796673830896e-06,
+      "loss": 0.1473,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7507883709879476,
+      "grad_norm": 0.565076720183294,
+      "learning_rate": 8.780746276730246e-06,
+      "loss": 0.145,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7516523391939177,
+      "grad_norm": 0.5855244277950163,
+      "learning_rate": 8.777692590105429e-06,
+      "loss": 0.1462,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7525163073998877,
+      "grad_norm": 0.5497821035086375,
+      "learning_rate": 8.774635616613373e-06,
+      "loss": 0.1442,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7533802756058577,
+      "grad_norm": 0.6039032637599129,
+      "learning_rate": 8.771575358913871e-06,
+      "loss": 0.1462,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7542442438118278,
+      "grad_norm": 0.5750476580302071,
+      "learning_rate": 8.768511819669566e-06,
+      "loss": 0.1453,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7551082120177978,
+      "grad_norm": 0.5568443722713766,
+      "learning_rate": 8.765445001545961e-06,
+      "loss": 0.1503,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7559721802237678,
+      "grad_norm": 0.6349949989616961,
+      "learning_rate": 8.76237490721141e-06,
+      "loss": 0.1458,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7568361484297378,
+      "grad_norm": 0.5900255153823822,
+      "learning_rate": 8.75930153933712e-06,
+      "loss": 0.1488,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7577001166357078,
+      "grad_norm": 0.5726499989030004,
+      "learning_rate": 8.756224900597144e-06,
+      "loss": 0.1463,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7585640848416778,
+      "grad_norm": 0.5794097752200494,
+      "learning_rate": 8.75314499366838e-06,
+      "loss": 0.1476,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7594280530476478,
+      "grad_norm": 0.5592726753611104,
+      "learning_rate": 8.750061821230573e-06,
+      "loss": 0.1456,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7602920212536178,
+      "grad_norm": 0.5642859448954136,
+      "learning_rate": 8.746975385966305e-06,
+      "loss": 0.1444,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7611559894595878,
+      "grad_norm": 0.602949892871976,
+      "learning_rate": 8.743885690561002e-06,
+      "loss": 0.1476,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7620199576655579,
+      "grad_norm": 0.578749309560873,
+      "learning_rate": 8.740792737702921e-06,
+      "loss": 0.1464,
+      "step": 8820
+    },
+    {
+      "epoch": 0.762883925871528,
+      "grad_norm": 0.5827033788877268,
+      "learning_rate": 8.737696530083158e-06,
+      "loss": 0.1444,
+      "step": 8830
+    },
+    {
+      "epoch": 0.763747894077498,
+      "grad_norm": 0.5737302603011082,
+      "learning_rate": 8.734597070395635e-06,
+      "loss": 0.1476,
+      "step": 8840
+    },
+    {
+      "epoch": 0.764611862283468,
+      "grad_norm": 0.567276002499956,
+      "learning_rate": 8.731494361337111e-06,
+      "loss": 0.147,
+      "step": 8850
+    },
+    {
+      "epoch": 0.765475830489438,
+      "grad_norm": 0.5785412025187946,
+      "learning_rate": 8.72838840560717e-06,
+      "loss": 0.1466,
+      "step": 8860
+    },
+    {
+      "epoch": 0.766339798695408,
+      "grad_norm": 0.6021940515191849,
+      "learning_rate": 8.725279205908214e-06,
+      "loss": 0.1424,
+      "step": 8870
+    },
+    {
+      "epoch": 0.767203766901378,
+      "grad_norm": 0.5678686047684466,
+      "learning_rate": 8.722166764945476e-06,
+      "loss": 0.1416,
+      "step": 8880
+    },
+    {
+      "epoch": 0.768067735107348,
+      "grad_norm": 0.5983675801989564,
+      "learning_rate": 8.719051085427007e-06,
+      "loss": 0.1445,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7689317033133181,
+      "grad_norm": 0.5508988572074687,
+      "learning_rate": 8.71593217006367e-06,
+      "loss": 0.1466,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7697956715192881,
+      "grad_norm": 0.5884231126943428,
+      "learning_rate": 8.712810021569153e-06,
+      "loss": 0.1509,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7706596397252581,
+      "grad_norm": 0.5702397352103349,
+      "learning_rate": 8.709684642659952e-06,
+      "loss": 0.15,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7715236079312281,
+      "grad_norm": 0.5646379676204086,
+      "learning_rate": 8.706556036055372e-06,
+      "loss": 0.1448,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7723875761371981,
+      "grad_norm": 0.5908474945066845,
+      "learning_rate": 8.703424204477527e-06,
+      "loss": 0.1484,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7732515443431681,
+      "grad_norm": 0.5582469316100711,
+      "learning_rate": 8.700289150651342e-06,
+      "loss": 0.1452,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7741155125491382,
+      "grad_norm": 0.5758318223072699,
+      "learning_rate": 8.69715087730454e-06,
+      "loss": 0.1439,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7749794807551083,
+      "grad_norm": 0.5851791361562063,
+      "learning_rate": 8.694009387167643e-06,
+      "loss": 0.1461,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7758434489610783,
+      "grad_norm": 0.55539409751672,
+      "learning_rate": 8.690864682973983e-06,
+      "loss": 0.1429,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7767074171670483,
+      "grad_norm": 0.5972513387417647,
+      "learning_rate": 8.687716767459677e-06,
+      "loss": 0.1438,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7775713853730183,
+      "grad_norm": 0.5449750895753459,
+      "learning_rate": 8.68456564336364e-06,
+      "loss": 0.1423,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7784353535789883,
+      "grad_norm": 0.5629247273553248,
+      "learning_rate": 8.681411313427584e-06,
+      "loss": 0.1406,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7792993217849583,
+      "grad_norm": 0.5815650093028251,
+      "learning_rate": 8.678253780395997e-06,
+      "loss": 0.1498,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7801632899909283,
+      "grad_norm": 0.6061694848721607,
+      "learning_rate": 8.67509304701617e-06,
+      "loss": 0.147,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7810272581968983,
+      "grad_norm": 0.6055037881598447,
+      "learning_rate": 8.671929116038167e-06,
+      "loss": 0.1468,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7818912264028683,
+      "grad_norm": 0.5850345947610444,
+      "learning_rate": 8.66876199021484e-06,
+      "loss": 0.1484,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7827551946088384,
+      "grad_norm": 0.5542978640162831,
+      "learning_rate": 8.665591672301816e-06,
+      "loss": 0.1426,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7836191628148084,
+      "grad_norm": 0.5928333489110242,
+      "learning_rate": 8.662418165057507e-06,
+      "loss": 0.1429,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7844831310207785,
+      "grad_norm": 0.5648483381701337,
+      "learning_rate": 8.659241471243088e-06,
+      "loss": 0.1445,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7853470992267485,
+      "grad_norm": 0.6297067607299209,
+      "learning_rate": 8.656061593622521e-06,
+      "loss": 0.1486,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7862110674327185,
+      "grad_norm": 0.5783598666839724,
+      "learning_rate": 8.652878534962523e-06,
+      "loss": 0.1449,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7870750356386885,
+      "grad_norm": 0.5808145834168469,
+      "learning_rate": 8.649692298032594e-06,
+      "loss": 0.1452,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7879390038446585,
+      "grad_norm": 0.6257064540825436,
+      "learning_rate": 8.646502885604988e-06,
+      "loss": 0.1439,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7888029720506285,
+      "grad_norm": 0.5958553918867221,
+      "learning_rate": 8.643310300454726e-06,
+      "loss": 0.1498,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7896669402565986,
+      "grad_norm": 0.5718667220315514,
+      "learning_rate": 8.640114545359589e-06,
+      "loss": 0.1423,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7905309084625686,
+      "grad_norm": 0.5828008232292384,
+      "learning_rate": 8.636915623100116e-06,
+      "loss": 0.1427,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7913948766685386,
+      "grad_norm": 0.6078309638455678,
+      "learning_rate": 8.633713536459603e-06,
+      "loss": 0.1493,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7922588448745086,
+      "grad_norm": 0.5805499311435128,
+      "learning_rate": 8.6305082882241e-06,
+      "loss": 0.1452,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7931228130804786,
+      "grad_norm": 0.5523180415389615,
+      "learning_rate": 8.627299881182402e-06,
+      "loss": 0.1484,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7939867812864486,
+      "grad_norm": 0.5731511610127191,
+      "learning_rate": 8.624088318126057e-06,
+      "loss": 0.1472,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7948507494924186,
+      "grad_norm": 0.5711686241765782,
+      "learning_rate": 8.620873601849362e-06,
+      "loss": 0.1441,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7957147176983888,
+      "grad_norm": 0.5559112922512367,
+      "learning_rate": 8.617655735149354e-06,
+      "loss": 0.15,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7965786859043588,
+      "grad_norm": 0.5942008849380511,
+      "learning_rate": 8.614434720825805e-06,
+      "loss": 0.1453,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7974426541103288,
+      "grad_norm": 0.5826039170890981,
+      "learning_rate": 8.611210561681236e-06,
+      "loss": 0.147,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7983066223162988,
+      "grad_norm": 0.5809325892616363,
+      "learning_rate": 8.6079832605209e-06,
+      "loss": 0.1439,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7991705905222688,
+      "grad_norm": 0.551222305517412,
+      "learning_rate": 8.604752820152782e-06,
+      "loss": 0.1451,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8000345587282388,
+      "grad_norm": 0.569466216351549,
+      "learning_rate": 8.601519243387602e-06,
+      "loss": 0.1408,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8008985269342088,
+      "grad_norm": 0.5632939536324367,
+      "learning_rate": 8.598282533038804e-06,
+      "loss": 0.1414,
+      "step": 9270
+    },
+    {
+      "epoch": 0.8017624951401788,
+      "grad_norm": 0.5945112711943039,
+      "learning_rate": 8.595042691922564e-06,
+      "loss": 0.149,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8026264633461488,
+      "grad_norm": 0.6257070347054505,
+      "learning_rate": 8.591799722857779e-06,
+      "loss": 0.1468,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8034904315521189,
+      "grad_norm": 0.5502878627500125,
+      "learning_rate": 8.588553628666067e-06,
+      "loss": 0.1459,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8043543997580889,
+      "grad_norm": 0.5542111041310896,
+      "learning_rate": 8.585304412171767e-06,
+      "loss": 0.143,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8052183679640589,
+      "grad_norm": 0.5326126029738462,
+      "learning_rate": 8.582052076201932e-06,
+      "loss": 0.1423,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8060823361700289,
+      "grad_norm": 0.5885736349948085,
+      "learning_rate": 8.578796623586332e-06,
+      "loss": 0.1447,
+      "step": 9330
+    },
+    {
+      "epoch": 0.806946304375999,
+      "grad_norm": 0.584316686720796,
+      "learning_rate": 8.575538057157448e-06,
+      "loss": 0.1448,
+      "step": 9340
+    },
+    {
+      "epoch": 0.807810272581969,
+      "grad_norm": 0.5664241111477385,
+      "learning_rate": 8.572276379750468e-06,
+      "loss": 0.145,
+      "step": 9350
+    },
+    {
+      "epoch": 0.808674240787939,
+      "grad_norm": 0.5993574255434805,
+      "learning_rate": 8.569011594203291e-06,
+      "loss": 0.1458,
+      "step": 9360
+    },
+    {
+      "epoch": 0.809538208993909,
+      "grad_norm": 0.5888740832156935,
+      "learning_rate": 8.565743703356514e-06,
+      "loss": 0.1422,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8104021771998791,
+      "grad_norm": 0.5360589804242816,
+      "learning_rate": 8.562472710053444e-06,
+      "loss": 0.1405,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8112661454058491,
+      "grad_norm": 0.576297123132208,
+      "learning_rate": 8.559198617140081e-06,
+      "loss": 0.1421,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8121301136118191,
+      "grad_norm": 0.589692619006291,
+      "learning_rate": 8.555921427465124e-06,
+      "loss": 0.1462,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8129940818177891,
+      "grad_norm": 0.586639880883335,
+      "learning_rate": 8.552641143879965e-06,
+      "loss": 0.1418,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8138580500237591,
+      "grad_norm": 0.5728715638136743,
+      "learning_rate": 8.549357769238689e-06,
+      "loss": 0.1482,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8147220182297291,
+      "grad_norm": 0.5577477890301851,
+      "learning_rate": 8.546071306398075e-06,
+      "loss": 0.142,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8155859864356991,
+      "grad_norm": 0.5773828818872149,
+      "learning_rate": 8.54278175821758e-06,
+      "loss": 0.1462,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8164499546416691,
+      "grad_norm": 0.6148625238176575,
+      "learning_rate": 8.539489127559352e-06,
+      "loss": 0.1439,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8173139228476392,
+      "grad_norm": 0.5809787205443636,
+      "learning_rate": 8.53619341728822e-06,
+      "loss": 0.1446,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8181778910536093,
+      "grad_norm": 0.5673805345122123,
+      "learning_rate": 8.532894630271691e-06,
+      "loss": 0.1464,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8190418592595793,
+      "grad_norm": 0.5714831001212896,
+      "learning_rate": 8.529592769379947e-06,
+      "loss": 0.1467,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8199058274655493,
+      "grad_norm": 0.8142466454645951,
+      "learning_rate": 8.52628783748585e-06,
+      "loss": 0.146,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8207697956715193,
+      "grad_norm": 0.5736789389590432,
+      "learning_rate": 8.52297983746493e-06,
+      "loss": 0.1467,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8216337638774893,
+      "grad_norm": 0.5758982922754664,
+      "learning_rate": 8.519668772195389e-06,
+      "loss": 0.1437,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8224977320834593,
+      "grad_norm": 0.5451983109559408,
+      "learning_rate": 8.516354644558091e-06,
+      "loss": 0.1428,
+      "step": 9520
+    },
+    {
+      "epoch": 0.8233617002894293,
+      "grad_norm": 0.5781843496988237,
+      "learning_rate": 8.513037457436571e-06,
+      "loss": 0.1421,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8242256684953994,
+      "grad_norm": 0.5660860358702046,
+      "learning_rate": 8.509717213717023e-06,
+      "loss": 0.1466,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8250896367013694,
+      "grad_norm": 0.5875923545017205,
+      "learning_rate": 8.5063939162883e-06,
+      "loss": 0.1471,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8259536049073394,
+      "grad_norm": 0.5793638698834187,
+      "learning_rate": 8.503067568041912e-06,
+      "loss": 0.1435,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8268175731133094,
+      "grad_norm": 0.6082894408541321,
+      "learning_rate": 8.499738171872024e-06,
+      "loss": 0.1476,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8276815413192794,
+      "grad_norm": 0.5410052725196663,
+      "learning_rate": 8.496405730675453e-06,
+      "loss": 0.1413,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8285455095252495,
+      "grad_norm": 0.5668481687999716,
+      "learning_rate": 8.493070247351665e-06,
+      "loss": 0.1428,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8294094777312195,
+      "grad_norm": 0.5396188203926512,
+      "learning_rate": 8.48973172480277e-06,
+      "loss": 0.1452,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8302734459371895,
+      "grad_norm": 0.5944775835927977,
+      "learning_rate": 8.48639016593353e-06,
+      "loss": 0.1416,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8311374141431596,
+      "grad_norm": 0.5707292481605472,
+      "learning_rate": 8.483045573651345e-06,
+      "loss": 0.1443,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8320013823491296,
+      "grad_norm": 0.5596948995294987,
+      "learning_rate": 8.479697950866249e-06,
+      "loss": 0.1414,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8328653505550996,
+      "grad_norm": 0.5789700110781804,
+      "learning_rate": 8.476347300490919e-06,
+      "loss": 0.1428,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8337293187610696,
+      "grad_norm": 0.5816059812515278,
+      "learning_rate": 8.472993625440666e-06,
+      "loss": 0.1503,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8345932869670396,
+      "grad_norm": 0.5583523771021892,
+      "learning_rate": 8.469636928633426e-06,
+      "loss": 0.1464,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8354572551730096,
+      "grad_norm": 0.5783974918171014,
+      "learning_rate": 8.466277212989778e-06,
+      "loss": 0.1421,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8363212233789796,
+      "grad_norm": 0.5637539984292849,
+      "learning_rate": 8.462914481432912e-06,
+      "loss": 0.1407,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8371851915849496,
+      "grad_norm": 0.5469313058157097,
+      "learning_rate": 8.459548736888651e-06,
+      "loss": 0.1441,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8380491597909197,
+      "grad_norm": 0.5781885951173833,
+      "learning_rate": 8.456179982285437e-06,
+      "loss": 0.1423,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8389131279968897,
+      "grad_norm": 0.5929979133896947,
+      "learning_rate": 8.452808220554332e-06,
+      "loss": 0.1419,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8397770962028598,
+      "grad_norm": 0.5688459632744697,
+      "learning_rate": 8.449433454629015e-06,
+      "loss": 0.1462,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8406410644088298,
+      "grad_norm": 0.5603455231856802,
+      "learning_rate": 8.446055687445774e-06,
+      "loss": 0.1448,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8415050326147998,
+      "grad_norm": 0.5859270782944502,
+      "learning_rate": 8.442674921943516e-06,
+      "loss": 0.142,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8423690008207698,
+      "grad_norm": 0.5798335827513441,
+      "learning_rate": 8.439291161063751e-06,
+      "loss": 0.141,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8432329690267398,
+      "grad_norm": 0.5565217504939,
+      "learning_rate": 8.4359044077506e-06,
+      "loss": 0.1428,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8440969372327098,
+      "grad_norm": 0.6040073775552014,
+      "learning_rate": 8.43251466495078e-06,
+      "loss": 0.1396,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8449609054386799,
+      "grad_norm": 0.5687283030783332,
+      "learning_rate": 8.429121935613614e-06,
+      "loss": 0.1396,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8458248736446499,
+      "grad_norm": 0.5425553417122125,
+      "learning_rate": 8.425726222691027e-06,
+      "loss": 0.1434,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8466888418506199,
+      "grad_norm": 0.5518060633249242,
+      "learning_rate": 8.422327529137534e-06,
+      "loss": 0.1459,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8475528100565899,
+      "grad_norm": 0.5433804487253319,
+      "learning_rate": 8.418925857910245e-06,
+      "loss": 0.1408,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8484167782625599,
+      "grad_norm": 0.5442315827085944,
+      "learning_rate": 8.415521211968862e-06,
+      "loss": 0.1477,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8492807464685299,
+      "grad_norm": 0.5521307095439536,
+      "learning_rate": 8.412113594275676e-06,
+      "loss": 0.1432,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8501447146744999,
+      "grad_norm": 0.5785067887210904,
+      "learning_rate": 8.408703007795559e-06,
+      "loss": 0.1442,
+      "step": 9840
+    },
+    {
+      "epoch": 0.85100868288047,
+      "grad_norm": 0.5888646780531525,
+      "learning_rate": 8.405289455495971e-06,
+      "loss": 0.1432,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8518726510864401,
+      "grad_norm": 0.5869356385591026,
+      "learning_rate": 8.401872940346952e-06,
+      "loss": 0.1435,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8527366192924101,
+      "grad_norm": 0.5647084533037493,
+      "learning_rate": 8.398453465321115e-06,
+      "loss": 0.1413,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8536005874983801,
+      "grad_norm": 0.5805929245861594,
+      "learning_rate": 8.395031033393655e-06,
+      "loss": 0.1471,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8544645557043501,
+      "grad_norm": 0.5818692090965515,
+      "learning_rate": 8.391605647542336e-06,
+      "loss": 0.1415,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8553285239103201,
+      "grad_norm": 0.5623579419758806,
+      "learning_rate": 8.388177310747494e-06,
+      "loss": 0.1445,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8561924921162901,
+      "grad_norm": 0.5842301010806707,
+      "learning_rate": 8.384746025992026e-06,
+      "loss": 0.1388,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8570564603222601,
+      "grad_norm": 0.552275823798508,
+      "learning_rate": 8.381311796261407e-06,
+      "loss": 0.1454,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8579204285282301,
+      "grad_norm": 0.5814597997265838,
+      "learning_rate": 8.37787462454366e-06,
+      "loss": 0.1426,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8587843967342002,
+      "grad_norm": 0.5391815636893464,
+      "learning_rate": 8.374434513829377e-06,
+      "loss": 0.1425,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8596483649401702,
+      "grad_norm": 0.5909735682466266,
+      "learning_rate": 8.370991467111705e-06,
+      "loss": 0.1457,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8605123331461402,
+      "grad_norm": 0.5908399739792674,
+      "learning_rate": 8.367545487386345e-06,
+      "loss": 0.1423,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8613763013521103,
+      "grad_norm": 0.5702025368686541,
+      "learning_rate": 8.36409657765155e-06,
+      "loss": 0.1407,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8622402695580803,
+      "grad_norm": 0.616393750494224,
+      "learning_rate": 8.36064474090812e-06,
+      "loss": 0.1454,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8631042377640503,
+      "grad_norm": 0.5793933000509822,
+      "learning_rate": 8.357189980159405e-06,
+      "loss": 0.142,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8639682059700203,
+      "grad_norm": 0.6120735127654011,
+      "learning_rate": 8.353732298411298e-06,
+      "loss": 0.143,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8648321741759903,
+      "grad_norm": 0.5672142424827209,
+      "learning_rate": 8.350271698672236e-06,
+      "loss": 0.1465,
+      "step": 10010
+    },
+    {
+      "epoch": 0.8656961423819604,
+      "grad_norm": 0.5902729900228314,
+      "learning_rate": 8.34680818395319e-06,
+      "loss": 0.1435,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8665601105879304,
+      "grad_norm": 0.5691844420715534,
+      "learning_rate": 8.343341757267671e-06,
+      "loss": 0.1439,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8674240787939004,
+      "grad_norm": 0.5525245356489779,
+      "learning_rate": 8.339872421631718e-06,
+      "loss": 0.1421,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8682880469998704,
+      "grad_norm": 0.5633902551254112,
+      "learning_rate": 8.336400180063913e-06,
+      "loss": 0.1405,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8691520152058404,
+      "grad_norm": 0.5522452533333609,
+      "learning_rate": 8.332925035585351e-06,
+      "loss": 0.1406,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8700159834118104,
+      "grad_norm": 0.5683843284307902,
+      "learning_rate": 8.329446991219665e-06,
+      "loss": 0.1414,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8708799516177804,
+      "grad_norm": 0.8877940351300201,
+      "learning_rate": 8.325966049993004e-06,
+      "loss": 0.1443,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8717439198237504,
+      "grad_norm": 0.6050744894035649,
+      "learning_rate": 8.322482214934044e-06,
+      "loss": 0.1441,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8726078880297206,
+      "grad_norm": 0.5650088110025322,
+      "learning_rate": 8.318995489073968e-06,
+      "loss": 0.1416,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8734718562356906,
+      "grad_norm": 0.5571218501034207,
+      "learning_rate": 8.315505875446489e-06,
+      "loss": 0.1452,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8743358244416606,
+      "grad_norm": 0.531945157846496,
+      "learning_rate": 8.31201337708782e-06,
+      "loss": 0.1413,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8751997926476306,
+      "grad_norm": 0.5793899013617272,
+      "learning_rate": 8.308517997036687e-06,
+      "loss": 0.1419,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8760637608536006,
+      "grad_norm": 0.5522987674972991,
+      "learning_rate": 8.305019738334328e-06,
+      "loss": 0.1429,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8769277290595706,
+      "grad_norm": 0.599533080514142,
+      "learning_rate": 8.301518604024481e-06,
+      "loss": 0.1483,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8777916972655406,
+      "grad_norm": 0.5820611165673861,
+      "learning_rate": 8.298014597153387e-06,
+      "loss": 0.1449,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8786556654715106,
+      "grad_norm": 0.552105008952654,
+      "learning_rate": 8.294507720769789e-06,
+      "loss": 0.1422,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8795196336774807,
+      "grad_norm": 0.589431986283032,
+      "learning_rate": 8.290997977924922e-06,
+      "loss": 0.142,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8803836018834507,
+      "grad_norm": 0.5912403064747072,
+      "learning_rate": 8.287485371672518e-06,
+      "loss": 0.1455,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8812475700894207,
+      "grad_norm": 0.5835849936868334,
+      "learning_rate": 8.283969905068803e-06,
+      "loss": 0.141,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8821115382953907,
+      "grad_norm": 0.5560030512889691,
+      "learning_rate": 8.280451581172484e-06,
+      "loss": 0.1424,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8829755065013607,
+      "grad_norm": 0.5646674155359306,
+      "learning_rate": 8.276930403044759e-06,
+      "loss": 0.1457,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8838394747073308,
+      "grad_norm": 0.5737466726861257,
+      "learning_rate": 8.273406373749313e-06,
+      "loss": 0.1441,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8847034429133008,
+      "grad_norm": 0.5523276034253872,
+      "learning_rate": 8.269879496352304e-06,
+      "loss": 0.1405,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8855674111192708,
+      "grad_norm": 0.5534688383692349,
+      "learning_rate": 8.266349773922372e-06,
+      "loss": 0.137,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8864313793252409,
+      "grad_norm": 0.601059059729315,
+      "learning_rate": 8.262817209530636e-06,
+      "loss": 0.1398,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8872953475312109,
+      "grad_norm": 0.548502422051078,
+      "learning_rate": 8.259281806250678e-06,
+      "loss": 0.1411,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8881593157371809,
+      "grad_norm": 0.5333366385953059,
+      "learning_rate": 8.255743567158561e-06,
+      "loss": 0.142,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8890232839431509,
+      "grad_norm": 0.5154244022617628,
+      "learning_rate": 8.252202495332808e-06,
+      "loss": 0.1385,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8898872521491209,
+      "grad_norm": 0.5613558968123625,
+      "learning_rate": 8.248658593854408e-06,
+      "loss": 0.138,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8907512203550909,
+      "grad_norm": 0.5449671665400901,
+      "learning_rate": 8.245111865806816e-06,
+      "loss": 0.1427,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8916151885610609,
+      "grad_norm": 0.6077993286632034,
+      "learning_rate": 8.24156231427594e-06,
+      "loss": 0.1397,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8924791567670309,
+      "grad_norm": 0.5669974577554227,
+      "learning_rate": 8.23800994235015e-06,
+      "loss": 0.1409,
+      "step": 10330
+    },
+    {
+      "epoch": 0.893343124973001,
+      "grad_norm": 0.553947870504879,
+      "learning_rate": 8.234454753120268e-06,
+      "loss": 0.1427,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8942070931789711,
+      "grad_norm": 0.5480355609367276,
+      "learning_rate": 8.230896749679566e-06,
+      "loss": 0.1445,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8950710613849411,
+      "grad_norm": 0.567683345822826,
+      "learning_rate": 8.227335935123766e-06,
+      "loss": 0.1417,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8959350295909111,
+      "grad_norm": 0.5923492791517332,
+      "learning_rate": 8.223772312551035e-06,
+      "loss": 0.1452,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8967989977968811,
+      "grad_norm": 0.559847173910797,
+      "learning_rate": 8.220205885061986e-06,
+      "loss": 0.142,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8976629660028511,
+      "grad_norm": 0.52678466638974,
+      "learning_rate": 8.216636655759666e-06,
+      "loss": 0.1388,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8985269342088211,
+      "grad_norm": 0.5679754532984375,
+      "learning_rate": 8.213064627749567e-06,
+      "loss": 0.1431,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8993909024147911,
+      "grad_norm": 0.535769991144621,
+      "learning_rate": 8.209489804139614e-06,
+      "loss": 0.142,
+      "step": 10410
+    },
+    {
+      "epoch": 0.9002548706207611,
+      "grad_norm": 0.5849730580217068,
+      "learning_rate": 8.205912188040164e-06,
+      "loss": 0.1458,
+      "step": 10420
+    },
+    {
+      "epoch": 0.9011188388267312,
+      "grad_norm": 0.5895526919570224,
+      "learning_rate": 8.202331782564e-06,
+      "loss": 0.1412,
+      "step": 10430
+    },
+    {
+      "epoch": 0.9019828070327012,
+      "grad_norm": 0.5626343062982044,
+      "learning_rate": 8.198748590826336e-06,
+      "loss": 0.1417,
+      "step": 10440
+    },
+    {
+      "epoch": 0.9028467752386712,
+      "grad_norm": 0.5514554215350627,
+      "learning_rate": 8.195162615944809e-06,
+      "loss": 0.1388,
+      "step": 10450
+    },
+    {
+      "epoch": 0.9037107434446412,
+      "grad_norm": 0.596576561211133,
+      "learning_rate": 8.191573861039481e-06,
+      "loss": 0.138,
+      "step": 10460
+    },
+    {
+      "epoch": 0.9045747116506112,
+      "grad_norm": 0.5608018302461737,
+      "learning_rate": 8.187982329232826e-06,
+      "loss": 0.1406,
+      "step": 10470
+    },
+    {
+      "epoch": 0.9054386798565813,
+      "grad_norm": 0.5922211818250301,
+      "learning_rate": 8.18438802364974e-06,
+      "loss": 0.1407,
+      "step": 10480
+    },
+    {
+      "epoch": 0.9063026480625513,
+      "grad_norm": 0.5626611609158868,
+      "learning_rate": 8.18079094741753e-06,
+      "loss": 0.1403,
+      "step": 10490
+    },
+    {
+      "epoch": 0.9071666162685214,
+      "grad_norm": 0.55739369913719,
+      "learning_rate": 8.177191103665912e-06,
+      "loss": 0.1428,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9080305844744914,
+      "grad_norm": 0.5589210208749367,
+      "learning_rate": 8.173588495527013e-06,
+      "loss": 0.1444,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9088945526804614,
+      "grad_norm": 0.5948997818661523,
+      "learning_rate": 8.169983126135366e-06,
+      "loss": 0.1406,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9097585208864314,
+      "grad_norm": 0.5637575295315885,
+      "learning_rate": 8.166374998627903e-06,
+      "loss": 0.1419,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9106224890924014,
+      "grad_norm": 0.5520322935532269,
+      "learning_rate": 8.162764116143956e-06,
+      "loss": 0.1428,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9114864572983714,
+      "grad_norm": 0.5476065819654563,
+      "learning_rate": 8.159150481825256e-06,
+      "loss": 0.1445,
+      "step": 10550
+    },
+    {
+      "epoch": 0.9123504255043414,
+      "grad_norm": 0.5607845386057297,
+      "learning_rate": 8.155534098815929e-06,
+      "loss": 0.1409,
+      "step": 10560
+    },
+    {
+      "epoch": 0.9132143937103114,
+      "grad_norm": 0.5895145175268459,
+      "learning_rate": 8.15191497026249e-06,
+      "loss": 0.1392,
+      "step": 10570
+    },
+    {
+      "epoch": 0.9140783619162814,
+      "grad_norm": 0.5894062695205331,
+      "learning_rate": 8.148293099313843e-06,
+      "loss": 0.1353,
+      "step": 10580
+    },
+    {
+      "epoch": 0.9149423301222515,
+      "grad_norm": 0.5743939777288788,
+      "learning_rate": 8.14466848912128e-06,
+      "loss": 0.1382,
+      "step": 10590
+    },
+    {
+      "epoch": 0.9158062983282215,
+      "grad_norm": 0.5738967722486941,
+      "learning_rate": 8.141041142838475e-06,
+      "loss": 0.1437,
+      "step": 10600
+    },
+    {
+      "epoch": 0.9166702665341916,
+      "grad_norm": 0.5508385159709146,
+      "learning_rate": 8.137411063621488e-06,
+      "loss": 0.1395,
+      "step": 10610
+    },
+    {
+      "epoch": 0.9175342347401616,
+      "grad_norm": 0.5549344070840297,
+      "learning_rate": 8.133778254628744e-06,
+      "loss": 0.1391,
+      "step": 10620
+    },
+    {
+      "epoch": 0.9183982029461316,
+      "grad_norm": 0.5557337487994121,
+      "learning_rate": 8.130142719021055e-06,
+      "loss": 0.1428,
+      "step": 10630
+    },
+    {
+      "epoch": 0.9192621711521016,
+      "grad_norm": 0.5335590128180864,
+      "learning_rate": 8.126504459961601e-06,
+      "loss": 0.1435,
+      "step": 10640
+    },
+    {
+      "epoch": 0.9201261393580716,
+      "grad_norm": 0.5618082432643633,
+      "learning_rate": 8.122863480615932e-06,
+      "loss": 0.1434,
+      "step": 10650
+    },
+    {
+      "epoch": 0.9209901075640416,
+      "grad_norm": 0.5483727536324223,
+      "learning_rate": 8.119219784151964e-06,
+      "loss": 0.1407,
+      "step": 10660
+    },
+    {
+      "epoch": 0.9218540757700117,
+      "grad_norm": 0.5225716204079307,
+      "learning_rate": 8.11557337373998e-06,
+      "loss": 0.1426,
+      "step": 10670
+    },
+    {
+      "epoch": 0.9227180439759817,
+      "grad_norm": 0.5802251936874245,
+      "learning_rate": 8.11192425255262e-06,
+      "loss": 0.1414,
+      "step": 10680
+    },
+    {
+      "epoch": 0.9235820121819517,
+      "grad_norm": 0.549636716436263,
+      "learning_rate": 8.108272423764883e-06,
+      "loss": 0.1439,
+      "step": 10690
+    },
+    {
+      "epoch": 0.9244459803879217,
+      "grad_norm": 0.5654550374218119,
+      "learning_rate": 8.104617890554129e-06,
+      "loss": 0.1427,
+      "step": 10700
+    },
+    {
+      "epoch": 0.9253099485938917,
+      "grad_norm": 0.5673693824798013,
+      "learning_rate": 8.100960656100069e-06,
+      "loss": 0.1429,
+      "step": 10710
+    },
+    {
+      "epoch": 0.9261739167998617,
+      "grad_norm": 0.5132529576114598,
+      "learning_rate": 8.097300723584757e-06,
+      "loss": 0.1377,
+      "step": 10720
+    },
+    {
+      "epoch": 0.9270378850058317,
+      "grad_norm": 0.5334736844206937,
+      "learning_rate": 8.093638096192606e-06,
+      "loss": 0.1405,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9279018532118019,
+      "grad_norm": 0.5510539851521975,
+      "learning_rate": 8.089972777110366e-06,
+      "loss": 0.1383,
+      "step": 10740
+    },
+    {
+      "epoch": 0.9287658214177719,
+      "grad_norm": 0.543077012923459,
+      "learning_rate": 8.08630476952713e-06,
+      "loss": 0.1408,
+      "step": 10750
+    },
+    {
+      "epoch": 0.9296297896237419,
+      "grad_norm": 0.5386576487008168,
+      "learning_rate": 8.082634076634334e-06,
+      "loss": 0.142,
+      "step": 10760
+    },
+    {
+      "epoch": 0.9304937578297119,
+      "grad_norm": 0.5565406495408177,
+      "learning_rate": 8.078960701625746e-06,
+      "loss": 0.1373,
+      "step": 10770
+    },
+    {
+      "epoch": 0.9313577260356819,
+      "grad_norm": 0.5948123321042771,
+      "learning_rate": 8.07528464769747e-06,
+      "loss": 0.1382,
+      "step": 10780
+    },
+    {
+      "epoch": 0.9322216942416519,
+      "grad_norm": 0.5446315002732925,
+      "learning_rate": 8.071605918047938e-06,
+      "loss": 0.1379,
+      "step": 10790
+    },
+    {
+      "epoch": 0.9330856624476219,
+      "grad_norm": 0.5539597975139731,
+      "learning_rate": 8.067924515877914e-06,
+      "loss": 0.138,
+      "step": 10800
+    },
+    {
+      "epoch": 0.9339496306535919,
+      "grad_norm": 0.5552871380668192,
+      "learning_rate": 8.064240444390487e-06,
+      "loss": 0.1402,
+      "step": 10810
+    },
+    {
+      "epoch": 0.934813598859562,
+      "grad_norm": 0.5673123265243116,
+      "learning_rate": 8.060553706791066e-06,
+      "loss": 0.1431,
+      "step": 10820
+    },
+    {
+      "epoch": 0.935677567065532,
+      "grad_norm": 0.5909182056894116,
+      "learning_rate": 8.05686430628738e-06,
+      "loss": 0.1429,
+      "step": 10830
+    },
+    {
+      "epoch": 0.936541535271502,
+      "grad_norm": 0.5504559786897197,
+      "learning_rate": 8.053172246089476e-06,
+      "loss": 0.1454,
+      "step": 10840
+    },
+    {
+      "epoch": 0.937405503477472,
+      "grad_norm": 0.5806631353235165,
+      "learning_rate": 8.049477529409712e-06,
+      "loss": 0.1472,
+      "step": 10850
+    },
+    {
+      "epoch": 0.9382694716834421,
+      "grad_norm": 0.5709906117492967,
+      "learning_rate": 8.045780159462769e-06,
+      "loss": 0.1392,
+      "step": 10860
+    },
+    {
+      "epoch": 0.9391334398894121,
+      "grad_norm": 0.5940409325929809,
+      "learning_rate": 8.042080139465617e-06,
+      "loss": 0.1371,
+      "step": 10870
+    },
+    {
+      "epoch": 0.9399974080953821,
+      "grad_norm": 0.5644194425180563,
+      "learning_rate": 8.03837747263755e-06,
+      "loss": 0.1401,
+      "step": 10880
+    },
+    {
+      "epoch": 0.9408613763013521,
+      "grad_norm": 0.5513413048942039,
+      "learning_rate": 8.034672162200153e-06,
+      "loss": 0.1433,
+      "step": 10890
+    },
+    {
+      "epoch": 0.9417253445073221,
+      "grad_norm": 0.5451574269752696,
+      "learning_rate": 8.030964211377317e-06,
+      "loss": 0.1409,
+      "step": 10900
+    },
+    {
+      "epoch": 0.9425893127132922,
+      "grad_norm": 0.5651146180037745,
+      "learning_rate": 8.027253623395231e-06,
+      "loss": 0.1382,
+      "step": 10910
+    },
+    {
+      "epoch": 0.9434532809192622,
+      "grad_norm": 0.5398822960393462,
+      "learning_rate": 8.023540401482373e-06,
+      "loss": 0.1431,
+      "step": 10920
+    },
+    {
+      "epoch": 0.9443172491252322,
+      "grad_norm": 0.5437530423433509,
+      "learning_rate": 8.019824548869518e-06,
+      "loss": 0.1425,
+      "step": 10930
+    },
+    {
+      "epoch": 0.9451812173312022,
+      "grad_norm": 0.5546789213874745,
+      "learning_rate": 8.016106068789727e-06,
+      "loss": 0.1373,
+      "step": 10940
+    },
+    {
+      "epoch": 0.9460451855371722,
+      "grad_norm": 0.5471020932209029,
+      "learning_rate": 8.012384964478346e-06,
+      "loss": 0.138,
+      "step": 10950
+    },
+    {
+      "epoch": 0.9469091537431422,
+      "grad_norm": 0.5697798890296741,
+      "learning_rate": 8.008661239173012e-06,
+      "loss": 0.1407,
+      "step": 10960
+    },
+    {
+      "epoch": 0.9477731219491122,
+      "grad_norm": 0.5575286625262722,
+      "learning_rate": 8.004934896113633e-06,
+      "loss": 0.1397,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9486370901550822,
+      "grad_norm": 0.5265828755856463,
+      "learning_rate": 8.001205938542398e-06,
+      "loss": 0.1424,
+      "step": 10980
+    },
+    {
+      "epoch": 0.9495010583610524,
+      "grad_norm": 0.551199244101437,
+      "learning_rate": 7.997474369703772e-06,
+      "loss": 0.1419,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9503650265670224,
+      "grad_norm": 0.5545419227985952,
+      "learning_rate": 7.993740192844493e-06,
+      "loss": 0.1408,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9512289947729924,
+      "grad_norm": 0.5783342185948568,
+      "learning_rate": 7.990003411213562e-06,
+      "loss": 0.1387,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9520929629789624,
+      "grad_norm": 0.5520321411170842,
+      "learning_rate": 7.986264028062256e-06,
+      "loss": 0.1414,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9529569311849324,
+      "grad_norm": 0.5376112537460771,
+      "learning_rate": 7.982522046644106e-06,
+      "loss": 0.1407,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9538208993909024,
+      "grad_norm": 0.5973209607733642,
+      "learning_rate": 7.97877747021491e-06,
+      "loss": 0.1381,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9546848675968724,
+      "grad_norm": 0.5722729711850041,
+      "learning_rate": 7.975030302032722e-06,
+      "loss": 0.143,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9555488358028424,
+      "grad_norm": 0.5505304492880758,
+      "learning_rate": 7.971280545357851e-06,
+      "loss": 0.1429,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9564128040088125,
+      "grad_norm": 0.5430702520493753,
+      "learning_rate": 7.967528203452856e-06,
+      "loss": 0.1377,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9572767722147825,
+      "grad_norm": 0.5379484526659242,
+      "learning_rate": 7.963773279582548e-06,
+      "loss": 0.1425,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9581407404207525,
+      "grad_norm": 0.59660987229937,
+      "learning_rate": 7.960015777013984e-06,
+      "loss": 0.1404,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9590047086267225,
+      "grad_norm": 0.5421623266199536,
+      "learning_rate": 7.956255699016466e-06,
+      "loss": 0.1424,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9598686768326925,
+      "grad_norm": 0.5322756578697195,
+      "learning_rate": 7.952493048861534e-06,
+      "loss": 0.1396,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9607326450386626,
+      "grad_norm": 0.6187852901428197,
+      "learning_rate": 7.948727829822967e-06,
+      "loss": 0.141,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9615966132446326,
+      "grad_norm": 0.6065087922615199,
+      "learning_rate": 7.94496004517678e-06,
+      "loss": 0.1411,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9624605814506026,
+      "grad_norm": 0.5659172316900728,
+      "learning_rate": 7.941189698201218e-06,
+      "loss": 0.1396,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9633245496565727,
+      "grad_norm": 0.5317942345273444,
+      "learning_rate": 7.937416792176758e-06,
+      "loss": 0.141,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9641885178625427,
+      "grad_norm": 0.554066543556405,
+      "learning_rate": 7.933641330386104e-06,
+      "loss": 0.1397,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9650524860685127,
+      "grad_norm": 0.5254678736887252,
+      "learning_rate": 7.929863316114179e-06,
+      "loss": 0.1388,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9659164542744827,
+      "grad_norm": 0.5302754296193825,
+      "learning_rate": 7.926082752648135e-06,
+      "loss": 0.1444,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9667804224804527,
+      "grad_norm": 0.5698851699270783,
+      "learning_rate": 7.922299643277331e-06,
+      "loss": 0.1409,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9676443906864227,
+      "grad_norm": 0.5464315412902958,
+      "learning_rate": 7.918513991293352e-06,
+      "loss": 0.1414,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9685083588923927,
+      "grad_norm": 0.5599688882235971,
+      "learning_rate": 7.91472579998999e-06,
+      "loss": 0.1402,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9693723270983627,
+      "grad_norm": 0.5753649910885886,
+      "learning_rate": 7.910935072663245e-06,
+      "loss": 0.1409,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9702362953043328,
+      "grad_norm": 0.5646681396820311,
+      "learning_rate": 7.907141812611325e-06,
+      "loss": 0.141,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9711002635103029,
+      "grad_norm": 0.5455395876393425,
+      "learning_rate": 7.903346023134645e-06,
+      "loss": 0.14,
+      "step": 11240
+    },
+    {
+      "epoch": 0.9719642317162729,
+      "grad_norm": 0.5643721404005844,
+      "learning_rate": 7.899547707535816e-06,
+      "loss": 0.1405,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9728281999222429,
+      "grad_norm": 0.5734263716016804,
+      "learning_rate": 7.895746869119647e-06,
+      "loss": 0.1386,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9736921681282129,
+      "grad_norm": 0.5754019383795476,
+      "learning_rate": 7.891943511193148e-06,
+      "loss": 0.1412,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9745561363341829,
+      "grad_norm": 0.5357745666067263,
+      "learning_rate": 7.888137637065514e-06,
+      "loss": 0.1407,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9754201045401529,
+      "grad_norm": 0.5463072339025498,
+      "learning_rate": 7.884329250048134e-06,
+      "loss": 0.1369,
+      "step": 11290
+    },
+    {
+      "epoch": 0.9762840727461229,
+      "grad_norm": 0.563329757491908,
+      "learning_rate": 7.880518353454576e-06,
+      "loss": 0.1366,
+      "step": 11300
+    },
+    {
+      "epoch": 0.977148040952093,
+      "grad_norm": 0.5750546590943822,
+      "learning_rate": 7.876704950600607e-06,
+      "loss": 0.1379,
+      "step": 11310
+    },
+    {
+      "epoch": 0.978012009158063,
+      "grad_norm": 0.5285988056427497,
+      "learning_rate": 7.872889044804155e-06,
+      "loss": 0.1399,
+      "step": 11320
+    },
+    {
+      "epoch": 0.978875977364033,
+      "grad_norm": 0.532439311490372,
+      "learning_rate": 7.869070639385343e-06,
+      "loss": 0.1422,
+      "step": 11330
+    },
+    {
+      "epoch": 0.979739945570003,
+      "grad_norm": 0.5398829784687519,
+      "learning_rate": 7.865249737666458e-06,
+      "loss": 0.1375,
+      "step": 11340
+    },
+    {
+      "epoch": 0.980603913775973,
+      "grad_norm": 0.5315198727011683,
+      "learning_rate": 7.861426342971962e-06,
+      "loss": 0.1333,
+      "step": 11350
+    },
+    {
+      "epoch": 0.981467881981943,
+      "grad_norm": 0.5484806587096321,
+      "learning_rate": 7.857600458628485e-06,
+      "loss": 0.1383,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9823318501879131,
+      "grad_norm": 0.5632626864322672,
+      "learning_rate": 7.85377208796483e-06,
+      "loss": 0.1408,
+      "step": 11370
+    },
+    {
+      "epoch": 0.9831958183938831,
+      "grad_norm": 0.5756731418315327,
+      "learning_rate": 7.849941234311952e-06,
+      "loss": 0.1423,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9840597865998532,
+      "grad_norm": 0.569984058319945,
+      "learning_rate": 7.846107901002976e-06,
+      "loss": 0.1418,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9849237548058232,
+      "grad_norm": 0.5394728565154906,
+      "learning_rate": 7.84227209137318e-06,
+      "loss": 0.1397,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9857877230117932,
+      "grad_norm": 0.5294475608511082,
+      "learning_rate": 7.838433808759994e-06,
+      "loss": 0.1417,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9866516912177632,
+      "grad_norm": 0.5788220534347303,
+      "learning_rate": 7.834593056503009e-06,
+      "loss": 0.1403,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9875156594237332,
+      "grad_norm": 0.5495378815031298,
+      "learning_rate": 7.830749837943952e-06,
+      "loss": 0.1438,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9883796276297032,
+      "grad_norm": 0.5602649448652921,
+      "learning_rate": 7.826904156426706e-06,
+      "loss": 0.1349,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9892435958356732,
+      "grad_norm": 0.510550749078522,
+      "learning_rate": 7.823056015297294e-06,
+      "loss": 0.1389,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9901075640416432,
+      "grad_norm": 0.5618787773731678,
+      "learning_rate": 7.819205417903879e-06,
+      "loss": 0.1391,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9909715322476133,
+      "grad_norm": 0.5415142910705364,
+      "learning_rate": 7.81535236759676e-06,
+      "loss": 0.1383,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9918355004535833,
+      "grad_norm": 0.5378807964604567,
+      "learning_rate": 7.81149686772837e-06,
+      "loss": 0.1381,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9926994686595533,
+      "grad_norm": 0.5086407896763732,
+      "learning_rate": 7.807638921653275e-06,
+      "loss": 0.1396,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9935634368655234,
+      "grad_norm": 0.5763401451203161,
+      "learning_rate": 7.803778532728168e-06,
+      "loss": 0.1385,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9944274050714934,
+      "grad_norm": 0.5283421801533311,
+      "learning_rate": 7.799915704311869e-06,
+      "loss": 0.1392,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9952913732774634,
+      "grad_norm": 0.5407041178210695,
+      "learning_rate": 7.79605043976532e-06,
+      "loss": 0.1378,
+      "step": 11520
+    },
+    {
+      "epoch": 0.9961553414834334,
+      "grad_norm": 0.5444185941688278,
+      "learning_rate": 7.792182742451579e-06,
+      "loss": 0.1389,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9970193096894034,
+      "grad_norm": 0.5155116418138258,
+      "learning_rate": 7.78831261573583e-06,
+      "loss": 0.1362,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9978832778953735,
+      "grad_norm": 0.5558368170116909,
+      "learning_rate": 7.784440062985357e-06,
+      "loss": 0.1402,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9987472461013435,
+      "grad_norm": 0.5370499131190705,
+      "learning_rate": 7.78056508756957e-06,
+      "loss": 0.1378,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9996112143073135,
+      "grad_norm": 0.5314120316897529,
+      "learning_rate": 7.776687692859972e-06,
+      "loss": 0.1361,
+      "step": 11570
+    },
+    {
+      "epoch": 1.0004751825132836,
+      "grad_norm": 0.6484295050544892,
+      "learning_rate": 7.772807882230184e-06,
+      "loss": 0.1253,
+      "step": 11580
+    },
+    {
+      "epoch": 1.0013391507192535,
+      "grad_norm": 0.5288373554942792,
+      "learning_rate": 7.768925659055922e-06,
+      "loss": 0.1179,
+      "step": 11590
+    },
+    {
+      "epoch": 1.0022031189252236,
+      "grad_norm": 0.5226558697799325,
+      "learning_rate": 7.765041026715e-06,
+      "loss": 0.1162,
+      "step": 11600
+    },
+    {
+      "epoch": 1.0030670871311935,
+      "grad_norm": 0.5559986175018743,
+      "learning_rate": 7.761153988587336e-06,
+      "loss": 0.1136,
+      "step": 11610
+    },
+    {
+      "epoch": 1.0039310553371636,
+      "grad_norm": 0.5738333391751079,
+      "learning_rate": 7.757264548054931e-06,
+      "loss": 0.1129,
+      "step": 11620
+    },
+    {
+      "epoch": 1.0047950235431335,
+      "grad_norm": 0.5874896405994214,
+      "learning_rate": 7.75337270850188e-06,
+      "loss": 0.1155,
+      "step": 11630
+    },
+    {
+      "epoch": 1.0056589917491037,
+      "grad_norm": 0.57700834782927,
+      "learning_rate": 7.749478473314371e-06,
+      "loss": 0.114,
+      "step": 11640
+    },
+    {
+      "epoch": 1.0065229599550736,
+      "grad_norm": 0.5253879750024364,
+      "learning_rate": 7.745581845880668e-06,
+      "loss": 0.1147,
+      "step": 11650
+    },
+    {
+      "epoch": 1.0073869281610437,
+      "grad_norm": 0.578799776263045,
+      "learning_rate": 7.741682829591124e-06,
+      "loss": 0.1162,
+      "step": 11660
+    },
+    {
+      "epoch": 1.0082508963670136,
+      "grad_norm": 0.5520505063573035,
+      "learning_rate": 7.737781427838163e-06,
+      "loss": 0.1111,
+      "step": 11670
+    },
+    {
+      "epoch": 1.0091148645729837,
+      "grad_norm": 0.5445981382109637,
+      "learning_rate": 7.733877644016288e-06,
+      "loss": 0.1146,
+      "step": 11680
+    },
+    {
+      "epoch": 1.0099788327789538,
+      "grad_norm": 0.51984767612585,
+      "learning_rate": 7.729971481522079e-06,
+      "loss": 0.115,
+      "step": 11690
+    },
+    {
+      "epoch": 1.0108428009849237,
+      "grad_norm": 0.579207833886054,
+      "learning_rate": 7.726062943754177e-06,
+      "loss": 0.1146,
+      "step": 11700
+    },
+    {
+      "epoch": 1.0117067691908939,
+      "grad_norm": 0.6003541965292548,
+      "learning_rate": 7.722152034113299e-06,
+      "loss": 0.114,
+      "step": 11710
+    },
+    {
+      "epoch": 1.0125707373968638,
+      "grad_norm": 0.5344757173700486,
+      "learning_rate": 7.718238756002214e-06,
+      "loss": 0.114,
+      "step": 11720
+    },
+    {
+      "epoch": 1.0134347056028339,
+      "grad_norm": 0.5717675830192779,
+      "learning_rate": 7.714323112825764e-06,
+      "loss": 0.1142,
+      "step": 11730
+    },
+    {
+      "epoch": 1.0142986738088038,
+      "grad_norm": 0.5589395318387775,
+      "learning_rate": 7.710405107990841e-06,
+      "loss": 0.1124,
+      "step": 11740
+    },
+    {
+      "epoch": 1.015162642014774,
+      "grad_norm": 0.5191208223948365,
+      "learning_rate": 7.706484744906394e-06,
+      "loss": 0.1132,
+      "step": 11750
+    },
+    {
+      "epoch": 1.0160266102207438,
+      "grad_norm": 0.5739974899486876,
+      "learning_rate": 7.702562026983425e-06,
+      "loss": 0.113,
+      "step": 11760
+    },
+    {
+      "epoch": 1.016890578426714,
+      "grad_norm": 0.5538985992217317,
+      "learning_rate": 7.698636957634984e-06,
+      "loss": 0.1125,
+      "step": 11770
+    },
+    {
+      "epoch": 1.0177545466326838,
+      "grad_norm": 0.5556788352650281,
+      "learning_rate": 7.694709540276165e-06,
+      "loss": 0.1086,
+      "step": 11780
+    },
+    {
+      "epoch": 1.018618514838654,
+      "grad_norm": 0.557737388023279,
+      "learning_rate": 7.690779778324106e-06,
+      "loss": 0.1139,
+      "step": 11790
+    },
+    {
+      "epoch": 1.0194824830446239,
+      "grad_norm": 0.5533298134289029,
+      "learning_rate": 7.68684767519799e-06,
+      "loss": 0.1166,
+      "step": 11800
+    },
+    {
+      "epoch": 1.020346451250594,
+      "grad_norm": 0.5575744213533839,
+      "learning_rate": 7.68291323431903e-06,
+      "loss": 0.1123,
+      "step": 11810
+    },
+    {
+      "epoch": 1.021210419456564,
+      "grad_norm": 0.5576624574436335,
+      "learning_rate": 7.678976459110475e-06,
+      "loss": 0.1147,
+      "step": 11820
+    },
+    {
+      "epoch": 1.022074387662534,
+      "grad_norm": 0.5582794921471507,
+      "learning_rate": 7.675037352997606e-06,
+      "loss": 0.1128,
+      "step": 11830
+    },
+    {
+      "epoch": 1.0229383558685041,
+      "grad_norm": 0.5765353203515869,
+      "learning_rate": 7.67109591940773e-06,
+      "loss": 0.1135,
+      "step": 11840
+    },
+    {
+      "epoch": 1.023802324074474,
+      "grad_norm": 0.5594470994345416,
+      "learning_rate": 7.667152161770185e-06,
+      "loss": 0.1132,
+      "step": 11850
+    },
+    {
+      "epoch": 1.0246662922804441,
+      "grad_norm": 0.5843630847176365,
+      "learning_rate": 7.663206083516323e-06,
+      "loss": 0.1148,
+      "step": 11860
+    },
+    {
+      "epoch": 1.025530260486414,
+      "grad_norm": 0.5413541740109529,
+      "learning_rate": 7.659257688079524e-06,
+      "loss": 0.1113,
+      "step": 11870
+    },
+    {
+      "epoch": 1.0263942286923842,
+      "grad_norm": 0.5822217976514313,
+      "learning_rate": 7.655306978895173e-06,
+      "loss": 0.116,
+      "step": 11880
+    },
+    {
+      "epoch": 1.027258196898354,
+      "grad_norm": 0.5979006652450949,
+      "learning_rate": 7.651353959400678e-06,
+      "loss": 0.1135,
+      "step": 11890
+    },
+    {
+      "epoch": 1.0281221651043242,
+      "grad_norm": 0.5665894234503989,
+      "learning_rate": 7.647398633035452e-06,
+      "loss": 0.1144,
+      "step": 11900
+    },
+    {
+      "epoch": 1.028986133310294,
+      "grad_norm": 0.5375518216139221,
+      "learning_rate": 7.643441003240918e-06,
+      "loss": 0.1177,
+      "step": 11910
+    },
+    {
+      "epoch": 1.0298501015162642,
+      "grad_norm": 0.5546406845601831,
+      "learning_rate": 7.639481073460501e-06,
+      "loss": 0.1154,
+      "step": 11920
+    },
+    {
+      "epoch": 1.0307140697222341,
+      "grad_norm": 0.5670480535796274,
+      "learning_rate": 7.635518847139627e-06,
+      "loss": 0.112,
+      "step": 11930
+    },
+    {
+      "epoch": 1.0315780379282042,
+      "grad_norm": 0.5887507475464324,
+      "learning_rate": 7.631554327725722e-06,
+      "loss": 0.1095,
+      "step": 11940
+    },
+    {
+      "epoch": 1.0324420061341744,
+      "grad_norm": 0.5896886428008532,
+      "learning_rate": 7.627587518668205e-06,
+      "loss": 0.1166,
+      "step": 11950
+    },
+    {
+      "epoch": 1.0333059743401443,
+      "grad_norm": 0.5690679992282057,
+      "learning_rate": 7.623618423418488e-06,
+      "loss": 0.1135,
+      "step": 11960
+    },
+    {
+      "epoch": 1.0341699425461144,
+      "grad_norm": 0.5763884873413326,
+      "learning_rate": 7.619647045429975e-06,
+      "loss": 0.1145,
+      "step": 11970
+    },
+    {
+      "epoch": 1.0350339107520843,
+      "grad_norm": 0.5547331293683502,
+      "learning_rate": 7.615673388158052e-06,
+      "loss": 0.1125,
+      "step": 11980
+    },
+    {
+      "epoch": 1.0358978789580544,
+      "grad_norm": 0.5809540964348433,
+      "learning_rate": 7.6116974550600895e-06,
+      "loss": 0.1147,
+      "step": 11990
+    },
+    {
+      "epoch": 1.0367618471640243,
+      "grad_norm": 0.5617878261436202,
+      "learning_rate": 7.60771924959544e-06,
+      "loss": 0.1152,
+      "step": 12000
+    },
+    {
+      "epoch": 1.0376258153699944,
+      "grad_norm": 0.5691052166008522,
+      "learning_rate": 7.603738775225429e-06,
+      "loss": 0.1118,
+      "step": 12010
+    },
+    {
+      "epoch": 1.0384897835759643,
+      "grad_norm": 0.5927470323848189,
+      "learning_rate": 7.599756035413359e-06,
+      "loss": 0.1142,
+      "step": 12020
+    },
+    {
+      "epoch": 1.0393537517819345,
+      "grad_norm": 0.5348052360596652,
+      "learning_rate": 7.595771033624507e-06,
+      "loss": 0.1138,
+      "step": 12030
+    },
+    {
+      "epoch": 1.0402177199879044,
+      "grad_norm": 0.5942881290559787,
+      "learning_rate": 7.5917837733261104e-06,
+      "loss": 0.1136,
+      "step": 12040
+    },
+    {
+      "epoch": 1.0410816881938745,
+      "grad_norm": 0.5870254311070779,
+      "learning_rate": 7.5877942579873755e-06,
+      "loss": 0.1111,
+      "step": 12050
+    },
+    {
+      "epoch": 1.0419456563998444,
+      "grad_norm": 0.5560655753288578,
+      "learning_rate": 7.583802491079473e-06,
+      "loss": 0.1166,
+      "step": 12060
+    },
+    {
+      "epoch": 1.0428096246058145,
+      "grad_norm": 0.5812311683484401,
+      "learning_rate": 7.579808476075529e-06,
+      "loss": 0.1134,
+      "step": 12070
+    },
+    {
+      "epoch": 1.0436735928117846,
+      "grad_norm": 0.587965901597796,
+      "learning_rate": 7.575812216450626e-06,
+      "loss": 0.1131,
+      "step": 12080
+    },
+    {
+      "epoch": 1.0445375610177545,
+      "grad_norm": 0.5807654543387688,
+      "learning_rate": 7.571813715681804e-06,
+      "loss": 0.1107,
+      "step": 12090
+    },
+    {
+      "epoch": 1.0454015292237246,
+      "grad_norm": 0.5819264052078913,
+      "learning_rate": 7.567812977248046e-06,
+      "loss": 0.1123,
+      "step": 12100
+    },
+    {
+      "epoch": 1.0462654974296945,
+      "grad_norm": 0.5915930314280453,
+      "learning_rate": 7.5638100046302855e-06,
+      "loss": 0.1131,
+      "step": 12110
+    },
+    {
+      "epoch": 1.0471294656356647,
+      "grad_norm": 0.5258533900711518,
+      "learning_rate": 7.5598048013114015e-06,
+      "loss": 0.1136,
+      "step": 12120
+    },
+    {
+      "epoch": 1.0479934338416346,
+      "grad_norm": 0.5836672177555081,
+      "learning_rate": 7.555797370776212e-06,
+      "loss": 0.1181,
+      "step": 12130
+    },
+    {
+      "epoch": 1.0488574020476047,
+      "grad_norm": 0.563741150677117,
+      "learning_rate": 7.551787716511472e-06,
+      "loss": 0.1143,
+      "step": 12140
+    },
+    {
+      "epoch": 1.0497213702535746,
+      "grad_norm": 0.5745656695349525,
+      "learning_rate": 7.547775842005871e-06,
+      "loss": 0.114,
+      "step": 12150
+    },
+    {
+      "epoch": 1.0505853384595447,
+      "grad_norm": 0.5502725240340415,
+      "learning_rate": 7.543761750750034e-06,
+      "loss": 0.114,
+      "step": 12160
+    },
+    {
+      "epoch": 1.0514493066655146,
+      "grad_norm": 0.5573865072308392,
+      "learning_rate": 7.53974544623651e-06,
+      "loss": 0.1123,
+      "step": 12170
+    },
+    {
+      "epoch": 1.0523132748714847,
+      "grad_norm": 0.5424078163988334,
+      "learning_rate": 7.535726931959781e-06,
+      "loss": 0.1125,
+      "step": 12180
+    },
+    {
+      "epoch": 1.0531772430774549,
+      "grad_norm": 0.6021571647004282,
+      "learning_rate": 7.531706211416239e-06,
+      "loss": 0.1138,
+      "step": 12190
+    },
+    {
+      "epoch": 1.0540412112834248,
+      "grad_norm": 0.5664122542057618,
+      "learning_rate": 7.527683288104208e-06,
+      "loss": 0.1115,
+      "step": 12200
+    },
+    {
+      "epoch": 1.0549051794893949,
+      "grad_norm": 0.5716818115749401,
+      "learning_rate": 7.523658165523924e-06,
+      "loss": 0.1114,
+      "step": 12210
+    },
+    {
+      "epoch": 1.0557691476953648,
+      "grad_norm": 0.6063170462255332,
+      "learning_rate": 7.5196308471775345e-06,
+      "loss": 0.1112,
+      "step": 12220
+    },
+    {
+      "epoch": 1.056633115901335,
+      "grad_norm": 1.2632340648066684,
+      "learning_rate": 7.5156013365691005e-06,
+      "loss": 0.1161,
+      "step": 12230
+    },
+    {
+      "epoch": 1.0574970841073048,
+      "grad_norm": 0.5509074426756472,
+      "learning_rate": 7.51156963720459e-06,
+      "loss": 0.1151,
+      "step": 12240
+    },
+    {
+      "epoch": 1.058361052313275,
+      "grad_norm": 0.5466414245187583,
+      "learning_rate": 7.5075357525918725e-06,
+      "loss": 0.1128,
+      "step": 12250
+    },
+    {
+      "epoch": 1.0592250205192448,
+      "grad_norm": 0.596160426575199,
+      "learning_rate": 7.503499686240719e-06,
+      "loss": 0.1114,
+      "step": 12260
+    },
+    {
+      "epoch": 1.060088988725215,
+      "grad_norm": 0.5642196283298498,
+      "learning_rate": 7.499461441662807e-06,
+      "loss": 0.1136,
+      "step": 12270
+    },
+    {
+      "epoch": 1.0609529569311849,
+      "grad_norm": 0.5849920421619905,
+      "learning_rate": 7.495421022371697e-06,
+      "loss": 0.1134,
+      "step": 12280
+    },
+    {
+      "epoch": 1.061816925137155,
+      "grad_norm": 0.5787275769912021,
+      "learning_rate": 7.491378431882851e-06,
+      "loss": 0.1149,
+      "step": 12290
+    },
+    {
+      "epoch": 1.0626808933431249,
+      "grad_norm": 0.5483503453537147,
+      "learning_rate": 7.487333673713615e-06,
+      "loss": 0.1167,
+      "step": 12300
+    },
+    {
+      "epoch": 1.063544861549095,
+      "grad_norm": 0.5359573243330554,
+      "learning_rate": 7.483286751383224e-06,
+      "loss": 0.1146,
+      "step": 12310
+    },
+    {
+      "epoch": 1.064408829755065,
+      "grad_norm": 0.5875705499528636,
+      "learning_rate": 7.479237668412793e-06,
+      "loss": 0.112,
+      "step": 12320
+    },
+    {
+      "epoch": 1.065272797961035,
+      "grad_norm": 0.5561394006831806,
+      "learning_rate": 7.47518642832532e-06,
+      "loss": 0.1139,
+      "step": 12330
+    },
+    {
+      "epoch": 1.0661367661670051,
+      "grad_norm": 0.5633666250557524,
+      "learning_rate": 7.47113303464568e-06,
+      "loss": 0.1125,
+      "step": 12340
+    },
+    {
+      "epoch": 1.067000734372975,
+      "grad_norm": 0.5623851970677978,
+      "learning_rate": 7.4670774909006174e-06,
+      "loss": 0.1181,
+      "step": 12350
+    },
+    {
+      "epoch": 1.0678647025789452,
+      "grad_norm": 0.5587675735584775,
+      "learning_rate": 7.463019800618749e-06,
+      "loss": 0.1133,
+      "step": 12360
+    },
+    {
+      "epoch": 1.068728670784915,
+      "grad_norm": 0.584981094727008,
+      "learning_rate": 7.458959967330565e-06,
+      "loss": 0.1135,
+      "step": 12370
+    },
+    {
+      "epoch": 1.0695926389908852,
+      "grad_norm": 0.6284088915860303,
+      "learning_rate": 7.4548979945684105e-06,
+      "loss": 0.1184,
+      "step": 12380
+    },
+    {
+      "epoch": 1.070456607196855,
+      "grad_norm": 0.5931271494335132,
+      "learning_rate": 7.450833885866502e-06,
+      "loss": 0.1125,
+      "step": 12390
+    },
+    {
+      "epoch": 1.0713205754028252,
+      "grad_norm": 0.5550532429183336,
+      "learning_rate": 7.446767644760906e-06,
+      "loss": 0.1127,
+      "step": 12400
+    },
+    {
+      "epoch": 1.0721845436087951,
+      "grad_norm": 0.5711627384957843,
+      "learning_rate": 7.4426992747895495e-06,
+      "loss": 0.1144,
+      "step": 12410
+    },
+    {
+      "epoch": 1.0730485118147652,
+      "grad_norm": 0.5497441584250377,
+      "learning_rate": 7.43862877949221e-06,
+      "loss": 0.1152,
+      "step": 12420
+    },
+    {
+      "epoch": 1.0739124800207351,
+      "grad_norm": 0.5849304669143405,
+      "learning_rate": 7.434556162410514e-06,
+      "loss": 0.1133,
+      "step": 12430
+    },
+    {
+      "epoch": 1.0747764482267053,
+      "grad_norm": 0.5584597859335196,
+      "learning_rate": 7.430481427087935e-06,
+      "loss": 0.114,
+      "step": 12440
+    },
+    {
+      "epoch": 1.0756404164326754,
+      "grad_norm": 0.5509002638468832,
+      "learning_rate": 7.426404577069789e-06,
+      "loss": 0.1137,
+      "step": 12450
+    },
+    {
+      "epoch": 1.0765043846386453,
+      "grad_norm": 0.5694890975843275,
+      "learning_rate": 7.42232561590323e-06,
+      "loss": 0.1143,
+      "step": 12460
+    },
+    {
+      "epoch": 1.0773683528446154,
+      "grad_norm": 0.6020473634715721,
+      "learning_rate": 7.418244547137254e-06,
+      "loss": 0.1139,
+      "step": 12470
+    },
+    {
+      "epoch": 1.0782323210505853,
+      "grad_norm": 0.5867109716162718,
+      "learning_rate": 7.414161374322686e-06,
+      "loss": 0.1159,
+      "step": 12480
+    },
+    {
+      "epoch": 1.0790962892565554,
+      "grad_norm": 0.5644144690792395,
+      "learning_rate": 7.410076101012184e-06,
+      "loss": 0.1146,
+      "step": 12490
+    },
+    {
+      "epoch": 1.0799602574625253,
+      "grad_norm": 0.6177197577941457,
+      "learning_rate": 7.405988730760231e-06,
+      "loss": 0.116,
+      "step": 12500
+    },
+    {
+      "epoch": 1.0808242256684955,
+      "grad_norm": 0.5853046102952433,
+      "learning_rate": 7.401899267123137e-06,
+      "loss": 0.1145,
+      "step": 12510
+    },
+    {
+      "epoch": 1.0816881938744654,
+      "grad_norm": 0.5697893771470309,
+      "learning_rate": 7.397807713659034e-06,
+      "loss": 0.1162,
+      "step": 12520
+    },
+    {
+      "epoch": 1.0825521620804355,
+      "grad_norm": 0.573783818853617,
+      "learning_rate": 7.39371407392787e-06,
+      "loss": 0.1159,
+      "step": 12530
+    },
+    {
+      "epoch": 1.0834161302864054,
+      "grad_norm": 0.5563259533540806,
+      "learning_rate": 7.38961835149141e-06,
+      "loss": 0.1156,
+      "step": 12540
+    },
+    {
+      "epoch": 1.0842800984923755,
+      "grad_norm": 0.5776456652808264,
+      "learning_rate": 7.38552054991323e-06,
+      "loss": 0.1168,
+      "step": 12550
+    },
+    {
+      "epoch": 1.0851440666983454,
+      "grad_norm": 0.5605064163789284,
+      "learning_rate": 7.381420672758714e-06,
+      "loss": 0.1129,
+      "step": 12560
+    },
+    {
+      "epoch": 1.0860080349043155,
+      "grad_norm": 0.5724005520433846,
+      "learning_rate": 7.377318723595055e-06,
+      "loss": 0.1141,
+      "step": 12570
+    },
+    {
+      "epoch": 1.0868720031102856,
+      "grad_norm": 0.5763764037044753,
+      "learning_rate": 7.373214705991245e-06,
+      "loss": 0.1143,
+      "step": 12580
+    },
+    {
+      "epoch": 1.0877359713162555,
+      "grad_norm": 0.5517897969244039,
+      "learning_rate": 7.36910862351808e-06,
+      "loss": 0.1128,
+      "step": 12590
+    },
+    {
+      "epoch": 1.0885999395222257,
+      "grad_norm": 0.5459275054821126,
+      "learning_rate": 7.36500047974815e-06,
+      "loss": 0.1148,
+      "step": 12600
+    },
+    {
+      "epoch": 1.0894639077281956,
+      "grad_norm": 0.5503423920751671,
+      "learning_rate": 7.360890278255838e-06,
+      "loss": 0.1112,
+      "step": 12610
+    },
+    {
+      "epoch": 1.0903278759341657,
+      "grad_norm": 0.5640084967541655,
+      "learning_rate": 7.356778022617318e-06,
+      "loss": 0.1112,
+      "step": 12620
+    },
+    {
+      "epoch": 1.0911918441401356,
+      "grad_norm": 0.5875745843307169,
+      "learning_rate": 7.352663716410553e-06,
+      "loss": 0.118,
+      "step": 12630
+    },
+    {
+      "epoch": 1.0920558123461057,
+      "grad_norm": 0.5323955570867941,
+      "learning_rate": 7.3485473632152844e-06,
+      "loss": 0.1147,
+      "step": 12640
+    },
+    {
+      "epoch": 1.0929197805520756,
+      "grad_norm": 0.5483966268609873,
+      "learning_rate": 7.344428966613044e-06,
+      "loss": 0.114,
+      "step": 12650
+    },
+    {
+      "epoch": 1.0937837487580457,
+      "grad_norm": 0.549406276310017,
+      "learning_rate": 7.340308530187134e-06,
+      "loss": 0.1108,
+      "step": 12660
+    },
+    {
+      "epoch": 1.0946477169640156,
+      "grad_norm": 0.5433161641383483,
+      "learning_rate": 7.336186057522633e-06,
+      "loss": 0.1159,
+      "step": 12670
+    },
+    {
+      "epoch": 1.0955116851699858,
+      "grad_norm": 0.5842351306204783,
+      "learning_rate": 7.332061552206393e-06,
+      "loss": 0.114,
+      "step": 12680
+    },
+    {
+      "epoch": 1.0963756533759557,
+      "grad_norm": 0.5616172693897599,
+      "learning_rate": 7.327935017827034e-06,
+      "loss": 0.1114,
+      "step": 12690
+    },
+    {
+      "epoch": 1.0972396215819258,
+      "grad_norm": 0.6017902269821186,
+      "learning_rate": 7.323806457974939e-06,
+      "loss": 0.115,
+      "step": 12700
+    },
+    {
+      "epoch": 1.098103589787896,
+      "grad_norm": 0.5724606965291857,
+      "learning_rate": 7.319675876242256e-06,
+      "loss": 0.1141,
+      "step": 12710
+    },
+    {
+      "epoch": 1.0989675579938658,
+      "grad_norm": 0.5895805939601668,
+      "learning_rate": 7.315543276222894e-06,
+      "loss": 0.1133,
+      "step": 12720
+    },
+    {
+      "epoch": 1.099831526199836,
+      "grad_norm": 0.5827595105149493,
+      "learning_rate": 7.3114086615125125e-06,
+      "loss": 0.115,
+      "step": 12730
+    },
+    {
+      "epoch": 1.1006954944058058,
+      "grad_norm": 0.5300031140016738,
+      "learning_rate": 7.3072720357085284e-06,
+      "loss": 0.1141,
+      "step": 12740
+    },
+    {
+      "epoch": 1.101559462611776,
+      "grad_norm": 0.5659944671796406,
+      "learning_rate": 7.303133402410104e-06,
+      "loss": 0.1103,
+      "step": 12750
+    },
+    {
+      "epoch": 1.1024234308177459,
+      "grad_norm": 0.5575335822538651,
+      "learning_rate": 7.298992765218156e-06,
+      "loss": 0.1189,
+      "step": 12760
+    },
+    {
+      "epoch": 1.103287399023716,
+      "grad_norm": 0.5816088261759306,
+      "learning_rate": 7.294850127735336e-06,
+      "loss": 0.1125,
+      "step": 12770
+    },
+    {
+      "epoch": 1.1041513672296859,
+      "grad_norm": 0.5377912164914728,
+      "learning_rate": 7.2907054935660415e-06,
+      "loss": 0.1142,
+      "step": 12780
+    },
+    {
+      "epoch": 1.105015335435656,
+      "grad_norm": 0.5171845913840188,
+      "learning_rate": 7.286558866316405e-06,
+      "loss": 0.1144,
+      "step": 12790
+    },
+    {
+      "epoch": 1.105879303641626,
+      "grad_norm": 0.5625513547764542,
+      "learning_rate": 7.282410249594294e-06,
+      "loss": 0.1139,
+      "step": 12800
+    },
+    {
+      "epoch": 1.106743271847596,
+      "grad_norm": 0.548817914014772,
+      "learning_rate": 7.278259647009308e-06,
+      "loss": 0.1125,
+      "step": 12810
+    },
+    {
+      "epoch": 1.1076072400535661,
+      "grad_norm": 0.5619622892577025,
+      "learning_rate": 7.27410706217277e-06,
+      "loss": 0.1123,
+      "step": 12820
+    },
+    {
+      "epoch": 1.108471208259536,
+      "grad_norm": 0.5958312479893616,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.1151,
+      "step": 12830
+    },
+    {
+      "epoch": 1.1093351764655062,
+      "grad_norm": 0.6065189288813805,
+      "learning_rate": 7.265795960198971e-06,
+      "loss": 0.1145,
+      "step": 12840
+    },
+    {
+      "epoch": 1.110199144671476,
+      "grad_norm": 0.574744744019306,
+      "learning_rate": 7.261637450292972e-06,
+      "loss": 0.1123,
+      "step": 12850
+    },
+    {
+      "epoch": 1.1110631128774462,
+      "grad_norm": 0.5523033030399436,
+      "learning_rate": 7.257476972597941e-06,
+      "loss": 0.1169,
+      "step": 12860
+    },
+    {
+      "epoch": 1.111927081083416,
+      "grad_norm": 0.5615343679114696,
+      "learning_rate": 7.2533145307337995e-06,
+      "loss": 0.1116,
+      "step": 12870
+    },
+    {
+      "epoch": 1.1127910492893862,
+      "grad_norm": 0.5892480180243681,
+      "learning_rate": 7.249150128322171e-06,
+      "loss": 0.1167,
+      "step": 12880
+    },
+    {
+      "epoch": 1.1136550174953561,
+      "grad_norm": 0.5835398687472092,
+      "learning_rate": 7.244983768986391e-06,
+      "loss": 0.1143,
+      "step": 12890
+    },
+    {
+      "epoch": 1.1145189857013262,
+      "grad_norm": 0.5420813706717189,
+      "learning_rate": 7.240815456351493e-06,
+      "loss": 0.1142,
+      "step": 12900
+    },
+    {
+      "epoch": 1.1153829539072961,
+      "grad_norm": 0.5560693442080733,
+      "learning_rate": 7.236645194044215e-06,
+      "loss": 0.1124,
+      "step": 12910
+    },
+    {
+      "epoch": 1.1162469221132663,
+      "grad_norm": 0.582775487561968,
+      "learning_rate": 7.232472985692985e-06,
+      "loss": 0.1158,
+      "step": 12920
+    },
+    {
+      "epoch": 1.1171108903192362,
+      "grad_norm": 0.5813772164210899,
+      "learning_rate": 7.228298834927932e-06,
+      "loss": 0.1151,
+      "step": 12930
+    },
+    {
+      "epoch": 1.1179748585252063,
+      "grad_norm": 0.551749722566045,
+      "learning_rate": 7.224122745380866e-06,
+      "loss": 0.1144,
+      "step": 12940
+    },
+    {
+      "epoch": 1.1188388267311762,
+      "grad_norm": 0.5987437110852261,
+      "learning_rate": 7.2199447206852926e-06,
+      "loss": 0.1131,
+      "step": 12950
+    },
+    {
+      "epoch": 1.1197027949371463,
+      "grad_norm": 0.5823318038186402,
+      "learning_rate": 7.215764764476392e-06,
+      "loss": 0.1144,
+      "step": 12960
+    },
+    {
+      "epoch": 1.1205667631431164,
+      "grad_norm": 0.562823286330617,
+      "learning_rate": 7.211582880391036e-06,
+      "loss": 0.1138,
+      "step": 12970
+    },
+    {
+      "epoch": 1.1214307313490863,
+      "grad_norm": 0.5689016713168833,
+      "learning_rate": 7.2073990720677625e-06,
+      "loss": 0.1082,
+      "step": 12980
+    },
+    {
+      "epoch": 1.1222946995550565,
+      "grad_norm": 0.5669520425360047,
+      "learning_rate": 7.203213343146793e-06,
+      "loss": 0.1123,
+      "step": 12990
+    },
+    {
+      "epoch": 1.1231586677610264,
+      "grad_norm": 0.5909865997664461,
+      "learning_rate": 7.199025697270014e-06,
+      "loss": 0.1168,
+      "step": 13000
+    },
+    {
+      "epoch": 1.1240226359669965,
+      "grad_norm": 0.6073490868130594,
+      "learning_rate": 7.194836138080983e-06,
+      "loss": 0.1164,
+      "step": 13010
+    },
+    {
+      "epoch": 1.1248866041729664,
+      "grad_norm": 0.5738517459776369,
+      "learning_rate": 7.19064466922492e-06,
+      "loss": 0.1115,
+      "step": 13020
+    },
+    {
+      "epoch": 1.1257505723789365,
+      "grad_norm": 0.5341322056648838,
+      "learning_rate": 7.186451294348708e-06,
+      "loss": 0.1127,
+      "step": 13030
+    },
+    {
+      "epoch": 1.1266145405849064,
+      "grad_norm": 0.5659073456165932,
+      "learning_rate": 7.182256017100888e-06,
+      "loss": 0.1132,
+      "step": 13040
+    },
+    {
+      "epoch": 1.1274785087908765,
+      "grad_norm": 0.5733078496207322,
+      "learning_rate": 7.178058841131658e-06,
+      "loss": 0.1152,
+      "step": 13050
+    },
+    {
+      "epoch": 1.1283424769968464,
+      "grad_norm": 0.5917250921625605,
+      "learning_rate": 7.173859770092863e-06,
+      "loss": 0.1137,
+      "step": 13060
+    },
+    {
+      "epoch": 1.1292064452028165,
+      "grad_norm": 0.5724064155200794,
+      "learning_rate": 7.1696588076380025e-06,
+      "loss": 0.1124,
+      "step": 13070
+    },
+    {
+      "epoch": 1.1300704134087867,
+      "grad_norm": 0.5839270882164534,
+      "learning_rate": 7.165455957422219e-06,
+      "loss": 0.1173,
+      "step": 13080
+    },
+    {
+      "epoch": 1.1309343816147566,
+      "grad_norm": 0.5818668132052538,
+      "learning_rate": 7.161251223102297e-06,
+      "loss": 0.1116,
+      "step": 13090
+    },
+    {
+      "epoch": 1.1317983498207267,
+      "grad_norm": 0.562653394134604,
+      "learning_rate": 7.15704460833666e-06,
+      "loss": 0.1169,
+      "step": 13100
+    },
+    {
+      "epoch": 1.1326623180266966,
+      "grad_norm": 0.5369258651613542,
+      "learning_rate": 7.152836116785372e-06,
+      "loss": 0.1144,
+      "step": 13110
+    },
+    {
+      "epoch": 1.1335262862326667,
+      "grad_norm": 0.5609598754088422,
+      "learning_rate": 7.148625752110125e-06,
+      "loss": 0.1128,
+      "step": 13120
+    },
+    {
+      "epoch": 1.1343902544386366,
+      "grad_norm": 0.5307583270088041,
+      "learning_rate": 7.1444135179742424e-06,
+      "loss": 0.1129,
+      "step": 13130
+    },
+    {
+      "epoch": 1.1352542226446067,
+      "grad_norm": 0.5794753705639407,
+      "learning_rate": 7.140199418042674e-06,
+      "loss": 0.116,
+      "step": 13140
+    },
+    {
+      "epoch": 1.1361181908505766,
+      "grad_norm": 0.5765150526409482,
+      "learning_rate": 7.135983455981993e-06,
+      "loss": 0.1161,
+      "step": 13150
+    },
+    {
+      "epoch": 1.1369821590565468,
+      "grad_norm": 0.588585379901811,
+      "learning_rate": 7.131765635460394e-06,
+      "loss": 0.1156,
+      "step": 13160
+    },
+    {
+      "epoch": 1.1378461272625167,
+      "grad_norm": 0.5268321880508435,
+      "learning_rate": 7.127545960147685e-06,
+      "loss": 0.1105,
+      "step": 13170
+    },
+    {
+      "epoch": 1.1387100954684868,
+      "grad_norm": 0.5409759474589731,
+      "learning_rate": 7.123324433715293e-06,
+      "loss": 0.1139,
+      "step": 13180
+    },
+    {
+      "epoch": 1.1395740636744567,
+      "grad_norm": 0.548753740483243,
+      "learning_rate": 7.119101059836252e-06,
+      "loss": 0.1115,
+      "step": 13190
+    },
+    {
+      "epoch": 1.1404380318804268,
+      "grad_norm": 0.5473787441617439,
+      "learning_rate": 7.114875842185205e-06,
+      "loss": 0.1132,
+      "step": 13200
+    },
+    {
+      "epoch": 1.1413020000863967,
+      "grad_norm": 0.5521716011944642,
+      "learning_rate": 7.110648784438396e-06,
+      "loss": 0.1132,
+      "step": 13210
+    },
+    {
+      "epoch": 1.1421659682923668,
+      "grad_norm": 0.5452012339396002,
+      "learning_rate": 7.1064198902736766e-06,
+      "loss": 0.1136,
+      "step": 13220
+    },
+    {
+      "epoch": 1.143029936498337,
+      "grad_norm": 0.573456706312834,
+      "learning_rate": 7.10218916337049e-06,
+      "loss": 0.1117,
+      "step": 13230
+    },
+    {
+      "epoch": 1.1438939047043069,
+      "grad_norm": 0.5486166950146599,
+      "learning_rate": 7.097956607409876e-06,
+      "loss": 0.1138,
+      "step": 13240
+    },
+    {
+      "epoch": 1.144757872910277,
+      "grad_norm": 0.5526828504425823,
+      "learning_rate": 7.093722226074467e-06,
+      "loss": 0.1099,
+      "step": 13250
+    },
+    {
+      "epoch": 1.1456218411162469,
+      "grad_norm": 0.5772698801339367,
+      "learning_rate": 7.089486023048482e-06,
+      "loss": 0.1102,
+      "step": 13260
+    },
+    {
+      "epoch": 1.146485809322217,
+      "grad_norm": 0.6049052486348215,
+      "learning_rate": 7.0852480020177265e-06,
+      "loss": 0.1124,
+      "step": 13270
+    },
+    {
+      "epoch": 1.147349777528187,
+      "grad_norm": 0.5755929622844455,
+      "learning_rate": 7.081008166669585e-06,
+      "loss": 0.1113,
+      "step": 13280
+    },
+    {
+      "epoch": 1.148213745734157,
+      "grad_norm": 0.5789717682270602,
+      "learning_rate": 7.076766520693024e-06,
+      "loss": 0.113,
+      "step": 13290
+    },
+    {
+      "epoch": 1.149077713940127,
+      "grad_norm": 0.5590626048251914,
+      "learning_rate": 7.072523067778583e-06,
+      "loss": 0.1154,
+      "step": 13300
+    },
+    {
+      "epoch": 1.149941682146097,
+      "grad_norm": 0.5562146192899278,
+      "learning_rate": 7.068277811618376e-06,
+      "loss": 0.1147,
+      "step": 13310
+    },
+    {
+      "epoch": 1.150805650352067,
+      "grad_norm": 0.5391519600284189,
+      "learning_rate": 7.064030755906084e-06,
+      "loss": 0.1132,
+      "step": 13320
+    },
+    {
+      "epoch": 1.151669618558037,
+      "grad_norm": 0.5689682085308547,
+      "learning_rate": 7.059781904336953e-06,
+      "loss": 0.1128,
+      "step": 13330
+    },
+    {
+      "epoch": 1.1525335867640072,
+      "grad_norm": 0.5685437828904983,
+      "learning_rate": 7.055531260607795e-06,
+      "loss": 0.1151,
+      "step": 13340
+    },
+    {
+      "epoch": 1.153397554969977,
+      "grad_norm": 0.5947489843605617,
+      "learning_rate": 7.051278828416979e-06,
+      "loss": 0.112,
+      "step": 13350
+    },
+    {
+      "epoch": 1.1542615231759472,
+      "grad_norm": 0.5463410502004773,
+      "learning_rate": 7.047024611464428e-06,
+      "loss": 0.1138,
+      "step": 13360
+    },
+    {
+      "epoch": 1.1551254913819171,
+      "grad_norm": 0.5374793646965029,
+      "learning_rate": 7.042768613451623e-06,
+      "loss": 0.1132,
+      "step": 13370
+    },
+    {
+      "epoch": 1.1559894595878872,
+      "grad_norm": 0.5966329054253161,
+      "learning_rate": 7.038510838081588e-06,
+      "loss": 0.1125,
+      "step": 13380
+    },
+    {
+      "epoch": 1.1568534277938571,
+      "grad_norm": 0.5531816864338223,
+      "learning_rate": 7.0342512890589e-06,
+      "loss": 0.1124,
+      "step": 13390
+    },
+    {
+      "epoch": 1.1577173959998273,
+      "grad_norm": 0.6014056601060235,
+      "learning_rate": 7.029989970089675e-06,
+      "loss": 0.1121,
+      "step": 13400
+    },
+    {
+      "epoch": 1.1585813642057972,
+      "grad_norm": 0.5984871684047993,
+      "learning_rate": 7.025726884881572e-06,
+      "loss": 0.1135,
+      "step": 13410
+    },
+    {
+      "epoch": 1.1594453324117673,
+      "grad_norm": 0.5583075318226408,
+      "learning_rate": 7.021462037143783e-06,
+      "loss": 0.1119,
+      "step": 13420
+    },
+    {
+      "epoch": 1.1603093006177372,
+      "grad_norm": 0.5502086604536702,
+      "learning_rate": 7.017195430587037e-06,
+      "loss": 0.1161,
+      "step": 13430
+    },
+    {
+      "epoch": 1.1611732688237073,
+      "grad_norm": 0.5918176410941008,
+      "learning_rate": 7.012927068923592e-06,
+      "loss": 0.1145,
+      "step": 13440
+    },
+    {
+      "epoch": 1.1620372370296774,
+      "grad_norm": 0.5545319587622841,
+      "learning_rate": 7.008656955867232e-06,
+      "loss": 0.1123,
+      "step": 13450
+    },
+    {
+      "epoch": 1.1629012052356473,
+      "grad_norm": 0.5673106420372879,
+      "learning_rate": 7.004385095133268e-06,
+      "loss": 0.1152,
+      "step": 13460
+    },
+    {
+      "epoch": 1.1637651734416172,
+      "grad_norm": 0.5840240557788225,
+      "learning_rate": 7.000111490438527e-06,
+      "loss": 0.1107,
+      "step": 13470
+    },
+    {
+      "epoch": 1.1646291416475874,
+      "grad_norm": 0.543595956222494,
+      "learning_rate": 6.9958361455013556e-06,
+      "loss": 0.1145,
+      "step": 13480
+    },
+    {
+      "epoch": 1.1654931098535575,
+      "grad_norm": 0.5490360713143707,
+      "learning_rate": 6.991559064041618e-06,
+      "loss": 0.1151,
+      "step": 13490
+    },
+    {
+      "epoch": 1.1663570780595274,
+      "grad_norm": 0.5905952151929723,
+      "learning_rate": 6.987280249780682e-06,
+      "loss": 0.1123,
+      "step": 13500
+    },
+    {
+      "epoch": 1.1672210462654975,
+      "grad_norm": 0.5708684897331763,
+      "learning_rate": 6.9829997064414315e-06,
+      "loss": 0.1117,
+      "step": 13510
+    },
+    {
+      "epoch": 1.1680850144714674,
+      "grad_norm": 0.5707805140043376,
+      "learning_rate": 6.9787174377482454e-06,
+      "loss": 0.1129,
+      "step": 13520
+    },
+    {
+      "epoch": 1.1689489826774375,
+      "grad_norm": 0.5734976835058482,
+      "learning_rate": 6.9744334474270134e-06,
+      "loss": 0.1127,
+      "step": 13530
+    },
+    {
+      "epoch": 1.1698129508834074,
+      "grad_norm": 0.5754813896572345,
+      "learning_rate": 6.970147739205115e-06,
+      "loss": 0.1111,
+      "step": 13540
+    },
+    {
+      "epoch": 1.1706769190893775,
+      "grad_norm": 0.5805142687488006,
+      "learning_rate": 6.965860316811432e-06,
+      "loss": 0.1096,
+      "step": 13550
+    },
+    {
+      "epoch": 1.1715408872953474,
+      "grad_norm": 0.5677452272337598,
+      "learning_rate": 6.96157118397633e-06,
+      "loss": 0.1114,
+      "step": 13560
+    },
+    {
+      "epoch": 1.1724048555013176,
+      "grad_norm": 0.5478940239346025,
+      "learning_rate": 6.957280344431669e-06,
+      "loss": 0.1139,
+      "step": 13570
+    },
+    {
+      "epoch": 1.1732688237072875,
+      "grad_norm": 0.5705600986096185,
+      "learning_rate": 6.9529878019107886e-06,
+      "loss": 0.1141,
+      "step": 13580
+    },
+    {
+      "epoch": 1.1741327919132576,
+      "grad_norm": 0.5671172827221018,
+      "learning_rate": 6.948693560148515e-06,
+      "loss": 0.1153,
+      "step": 13590
+    },
+    {
+      "epoch": 1.1749967601192277,
+      "grad_norm": 0.5887655397357853,
+      "learning_rate": 6.944397622881151e-06,
+      "loss": 0.1166,
+      "step": 13600
+    },
+    {
+      "epoch": 1.1758607283251976,
+      "grad_norm": 0.5712255557841519,
+      "learning_rate": 6.940099993846472e-06,
+      "loss": 0.1116,
+      "step": 13610
+    },
+    {
+      "epoch": 1.1767246965311677,
+      "grad_norm": 0.5388407173904846,
+      "learning_rate": 6.935800676783731e-06,
+      "loss": 0.113,
+      "step": 13620
+    },
+    {
+      "epoch": 1.1775886647371376,
+      "grad_norm": 0.5524086717947396,
+      "learning_rate": 6.931499675433644e-06,
+      "loss": 0.1106,
+      "step": 13630
+    },
+    {
+      "epoch": 1.1784526329431078,
+      "grad_norm": 0.5900683311612895,
+      "learning_rate": 6.927196993538396e-06,
+      "loss": 0.1143,
+      "step": 13640
+    },
+    {
+      "epoch": 1.1793166011490777,
+      "grad_norm": 0.5571666223561705,
+      "learning_rate": 6.922892634841632e-06,
+      "loss": 0.1105,
+      "step": 13650
+    },
+    {
+      "epoch": 1.1801805693550478,
+      "grad_norm": 0.5449665981514987,
+      "learning_rate": 6.91858660308846e-06,
+      "loss": 0.1141,
+      "step": 13660
+    },
+    {
+      "epoch": 1.1810445375610177,
+      "grad_norm": 0.5464355283512453,
+      "learning_rate": 6.914278902025436e-06,
+      "loss": 0.1123,
+      "step": 13670
+    },
+    {
+      "epoch": 1.1819085057669878,
+      "grad_norm": 0.5519986042239243,
+      "learning_rate": 6.909969535400578e-06,
+      "loss": 0.1125,
+      "step": 13680
+    },
+    {
+      "epoch": 1.1827724739729577,
+      "grad_norm": 0.5614519421544537,
+      "learning_rate": 6.905658506963344e-06,
+      "loss": 0.1141,
+      "step": 13690
+    },
+    {
+      "epoch": 1.1836364421789278,
+      "grad_norm": 0.5704123213334359,
+      "learning_rate": 6.901345820464647e-06,
+      "loss": 0.1149,
+      "step": 13700
+    },
+    {
+      "epoch": 1.184500410384898,
+      "grad_norm": 0.5776361497185494,
+      "learning_rate": 6.897031479656833e-06,
+      "loss": 0.1123,
+      "step": 13710
+    },
+    {
+      "epoch": 1.1853643785908679,
+      "grad_norm": 0.5459654368909052,
+      "learning_rate": 6.8927154882936955e-06,
+      "loss": 0.1129,
+      "step": 13720
+    },
+    {
+      "epoch": 1.1862283467968378,
+      "grad_norm": 0.5748613451436307,
+      "learning_rate": 6.888397850130458e-06,
+      "loss": 0.115,
+      "step": 13730
+    },
+    {
+      "epoch": 1.1870923150028079,
+      "grad_norm": 0.5431693832897374,
+      "learning_rate": 6.884078568923782e-06,
+      "loss": 0.1123,
+      "step": 13740
+    },
+    {
+      "epoch": 1.187956283208778,
+      "grad_norm": 0.5623779523522215,
+      "learning_rate": 6.879757648431755e-06,
+      "loss": 0.1127,
+      "step": 13750
+    },
+    {
+      "epoch": 1.188820251414748,
+      "grad_norm": 0.592553622911199,
+      "learning_rate": 6.875435092413894e-06,
+      "loss": 0.1153,
+      "step": 13760
+    },
+    {
+      "epoch": 1.189684219620718,
+      "grad_norm": 0.5595958783240502,
+      "learning_rate": 6.871110904631136e-06,
+      "loss": 0.1138,
+      "step": 13770
+    },
+    {
+      "epoch": 1.190548187826688,
+      "grad_norm": 0.5633116470772369,
+      "learning_rate": 6.866785088845838e-06,
+      "loss": 0.1143,
+      "step": 13780
+    },
+    {
+      "epoch": 1.191412156032658,
+      "grad_norm": 0.5909659506315208,
+      "learning_rate": 6.862457648821773e-06,
+      "loss": 0.1134,
+      "step": 13790
+    },
+    {
+      "epoch": 1.192276124238628,
+      "grad_norm": 0.5739285382431173,
+      "learning_rate": 6.858128588324131e-06,
+      "loss": 0.1133,
+      "step": 13800
+    },
+    {
+      "epoch": 1.193140092444598,
+      "grad_norm": 0.5721923307570116,
+      "learning_rate": 6.853797911119509e-06,
+      "loss": 0.1137,
+      "step": 13810
+    },
+    {
+      "epoch": 1.194004060650568,
+      "grad_norm": 0.5613644285392295,
+      "learning_rate": 6.84946562097591e-06,
+      "loss": 0.1122,
+      "step": 13820
+    },
+    {
+      "epoch": 1.194868028856538,
+      "grad_norm": 0.5542391369025897,
+      "learning_rate": 6.845131721662741e-06,
+      "loss": 0.114,
+      "step": 13830
+    },
+    {
+      "epoch": 1.195731997062508,
+      "grad_norm": 0.5437931235251273,
+      "learning_rate": 6.840796216950813e-06,
+      "loss": 0.1104,
+      "step": 13840
+    },
+    {
+      "epoch": 1.1965959652684781,
+      "grad_norm": 0.5727878132218367,
+      "learning_rate": 6.836459110612326e-06,
+      "loss": 0.1161,
+      "step": 13850
+    },
+    {
+      "epoch": 1.1974599334744482,
+      "grad_norm": 0.5552817579336806,
+      "learning_rate": 6.83212040642088e-06,
+      "loss": 0.1136,
+      "step": 13860
+    },
+    {
+      "epoch": 1.1983239016804181,
+      "grad_norm": 0.559250789005052,
+      "learning_rate": 6.827780108151464e-06,
+      "loss": 0.1125,
+      "step": 13870
+    },
+    {
+      "epoch": 1.1991878698863883,
+      "grad_norm": 0.5937634902485177,
+      "learning_rate": 6.823438219580452e-06,
+      "loss": 0.1083,
+      "step": 13880
+    },
+    {
+      "epoch": 1.2000518380923582,
+      "grad_norm": 0.5552214918377005,
+      "learning_rate": 6.819094744485601e-06,
+      "loss": 0.1129,
+      "step": 13890
+    },
+    {
+      "epoch": 1.2009158062983283,
+      "grad_norm": 0.5655419429967947,
+      "learning_rate": 6.8147496866460525e-06,
+      "loss": 0.1101,
+      "step": 13900
+    },
+    {
+      "epoch": 1.2017797745042982,
+      "grad_norm": 0.5839619215629936,
+      "learning_rate": 6.810403049842324e-06,
+      "loss": 0.1085,
+      "step": 13910
+    },
+    {
+      "epoch": 1.2026437427102683,
+      "grad_norm": 0.5409388759672646,
+      "learning_rate": 6.806054837856301e-06,
+      "loss": 0.1138,
+      "step": 13920
+    },
+    {
+      "epoch": 1.2035077109162382,
+      "grad_norm": 0.57144561714602,
+      "learning_rate": 6.801705054471248e-06,
+      "loss": 0.1145,
+      "step": 13930
+    },
+    {
+      "epoch": 1.2043716791222083,
+      "grad_norm": 0.5424545215147586,
+      "learning_rate": 6.79735370347179e-06,
+      "loss": 0.1139,
+      "step": 13940
+    },
+    {
+      "epoch": 1.2052356473281782,
+      "grad_norm": 0.5618561017996767,
+      "learning_rate": 6.793000788643923e-06,
+      "loss": 0.1116,
+      "step": 13950
+    },
+    {
+      "epoch": 1.2060996155341484,
+      "grad_norm": 0.5676546554249075,
+      "learning_rate": 6.788646313774996e-06,
+      "loss": 0.1133,
+      "step": 13960
+    },
+    {
+      "epoch": 1.2069635837401185,
+      "grad_norm": 0.6010536091884849,
+      "learning_rate": 6.784290282653719e-06,
+      "loss": 0.1125,
+      "step": 13970
+    },
+    {
+      "epoch": 1.2078275519460884,
+      "grad_norm": 0.6034476266033024,
+      "learning_rate": 6.779932699070157e-06,
+      "loss": 0.109,
+      "step": 13980
+    },
+    {
+      "epoch": 1.2086915201520585,
+      "grad_norm": 0.5223870934855904,
+      "learning_rate": 6.775573566815725e-06,
+      "loss": 0.1104,
+      "step": 13990
+    },
+    {
+      "epoch": 1.2095554883580284,
+      "grad_norm": 0.5894416024611014,
+      "learning_rate": 6.771212889683182e-06,
+      "loss": 0.1118,
+      "step": 14000
+    },
+    {
+      "epoch": 1.2104194565639985,
+      "grad_norm": 0.5336729589337562,
+      "learning_rate": 6.766850671466637e-06,
+      "loss": 0.1133,
+      "step": 14010
+    },
+    {
+      "epoch": 1.2112834247699684,
+      "grad_norm": 0.6024966427336127,
+      "learning_rate": 6.762486915961536e-06,
+      "loss": 0.1095,
+      "step": 14020
+    },
+    {
+      "epoch": 1.2121473929759385,
+      "grad_norm": 0.5413840123335933,
+      "learning_rate": 6.758121626964665e-06,
+      "loss": 0.1132,
+      "step": 14030
+    },
+    {
+      "epoch": 1.2130113611819084,
+      "grad_norm": 0.582683552400572,
+      "learning_rate": 6.753754808274139e-06,
+      "loss": 0.1142,
+      "step": 14040
+    },
+    {
+      "epoch": 1.2138753293878786,
+      "grad_norm": 0.5593554231560006,
+      "learning_rate": 6.749386463689413e-06,
+      "loss": 0.111,
+      "step": 14050
+    },
+    {
+      "epoch": 1.2147392975938485,
+      "grad_norm": 0.5534453036111682,
+      "learning_rate": 6.74501659701126e-06,
+      "loss": 0.112,
+      "step": 14060
+    },
+    {
+      "epoch": 1.2156032657998186,
+      "grad_norm": 0.5373879399272095,
+      "learning_rate": 6.7406452120417846e-06,
+      "loss": 0.1104,
+      "step": 14070
+    },
+    {
+      "epoch": 1.2164672340057887,
+      "grad_norm": 0.5605673893423937,
+      "learning_rate": 6.736272312584408e-06,
+      "loss": 0.1146,
+      "step": 14080
+    },
+    {
+      "epoch": 1.2173312022117586,
+      "grad_norm": 0.5282153572951253,
+      "learning_rate": 6.7318979024438725e-06,
+      "loss": 0.1108,
+      "step": 14090
+    },
+    {
+      "epoch": 1.2181951704177285,
+      "grad_norm": 0.5738009147157338,
+      "learning_rate": 6.7275219854262295e-06,
+      "loss": 0.1101,
+      "step": 14100
+    },
+    {
+      "epoch": 1.2190591386236986,
+      "grad_norm": 0.5537976410855348,
+      "learning_rate": 6.7231445653388494e-06,
+      "loss": 0.1117,
+      "step": 14110
+    },
+    {
+      "epoch": 1.2199231068296688,
+      "grad_norm": 0.536439976576386,
+      "learning_rate": 6.718765645990402e-06,
+      "loss": 0.1126,
+      "step": 14120
+    },
+    {
+      "epoch": 1.2207870750356387,
+      "grad_norm": 0.5614207764845589,
+      "learning_rate": 6.7143852311908695e-06,
+      "loss": 0.11,
+      "step": 14130
+    },
+    {
+      "epoch": 1.2216510432416088,
+      "grad_norm": 0.5799281943109638,
+      "learning_rate": 6.7100033247515265e-06,
+      "loss": 0.1118,
+      "step": 14140
+    },
+    {
+      "epoch": 1.2225150114475787,
+      "grad_norm": 0.5334465896125076,
+      "learning_rate": 6.705619930484954e-06,
+      "loss": 0.1137,
+      "step": 14150
+    },
+    {
+      "epoch": 1.2233789796535488,
+      "grad_norm": 0.5418264694625892,
+      "learning_rate": 6.701235052205023e-06,
+      "loss": 0.1118,
+      "step": 14160
+    },
+    {
+      "epoch": 1.2242429478595187,
+      "grad_norm": 0.5716467405913869,
+      "learning_rate": 6.696848693726896e-06,
+      "loss": 0.1137,
+      "step": 14170
+    },
+    {
+      "epoch": 1.2251069160654888,
+      "grad_norm": 0.5600841783319294,
+      "learning_rate": 6.692460858867025e-06,
+      "loss": 0.1117,
+      "step": 14180
+    },
+    {
+      "epoch": 1.2259708842714587,
+      "grad_norm": 0.5430963957009048,
+      "learning_rate": 6.6880715514431424e-06,
+      "loss": 0.11,
+      "step": 14190
+    },
+    {
+      "epoch": 1.2268348524774288,
+      "grad_norm": 0.610965208613663,
+      "learning_rate": 6.683680775274267e-06,
+      "loss": 0.1161,
+      "step": 14200
+    },
+    {
+      "epoch": 1.2276988206833988,
+      "grad_norm": 0.5428600711535735,
+      "learning_rate": 6.679288534180692e-06,
+      "loss": 0.114,
+      "step": 14210
+    },
+    {
+      "epoch": 1.2285627888893689,
+      "grad_norm": 0.5257402970258083,
+      "learning_rate": 6.67489483198399e-06,
+      "loss": 0.1127,
+      "step": 14220
+    },
+    {
+      "epoch": 1.229426757095339,
+      "grad_norm": 0.5695987153021643,
+      "learning_rate": 6.670499672506996e-06,
+      "loss": 0.109,
+      "step": 14230
+    },
+    {
+      "epoch": 1.230290725301309,
+      "grad_norm": 0.5421509104065498,
+      "learning_rate": 6.666103059573823e-06,
+      "loss": 0.1144,
+      "step": 14240
+    },
+    {
+      "epoch": 1.231154693507279,
+      "grad_norm": 0.5700089455833822,
+      "learning_rate": 6.661704997009841e-06,
+      "loss": 0.1096,
+      "step": 14250
+    },
+    {
+      "epoch": 1.232018661713249,
+      "grad_norm": 0.5770791849329917,
+      "learning_rate": 6.657305488641687e-06,
+      "loss": 0.1124,
+      "step": 14260
+    },
+    {
+      "epoch": 1.232882629919219,
+      "grad_norm": 0.5585275622314702,
+      "learning_rate": 6.65290453829725e-06,
+      "loss": 0.1153,
+      "step": 14270
+    },
+    {
+      "epoch": 1.233746598125189,
+      "grad_norm": 0.5494171663494277,
+      "learning_rate": 6.648502149805679e-06,
+      "loss": 0.1166,
+      "step": 14280
+    },
+    {
+      "epoch": 1.234610566331159,
+      "grad_norm": 0.5765287928053029,
+      "learning_rate": 6.6440983269973725e-06,
+      "loss": 0.1122,
+      "step": 14290
+    },
+    {
+      "epoch": 1.235474534537129,
+      "grad_norm": 0.5763938405118092,
+      "learning_rate": 6.639693073703974e-06,
+      "loss": 0.1076,
+      "step": 14300
+    },
+    {
+      "epoch": 1.236338502743099,
+      "grad_norm": 0.5494980151341401,
+      "learning_rate": 6.635286393758376e-06,
+      "loss": 0.1143,
+      "step": 14310
+    },
+    {
+      "epoch": 1.237202470949069,
+      "grad_norm": 0.5680501144954855,
+      "learning_rate": 6.63087829099471e-06,
+      "loss": 0.1137,
+      "step": 14320
+    },
+    {
+      "epoch": 1.2380664391550391,
+      "grad_norm": 0.5452528726216044,
+      "learning_rate": 6.6264687692483455e-06,
+      "loss": 0.111,
+      "step": 14330
+    },
+    {
+      "epoch": 1.2389304073610092,
+      "grad_norm": 0.575161065532565,
+      "learning_rate": 6.6220578323558885e-06,
+      "loss": 0.1155,
+      "step": 14340
+    },
+    {
+      "epoch": 1.2397943755669791,
+      "grad_norm": 0.5825666496800802,
+      "learning_rate": 6.617645484155173e-06,
+      "loss": 0.1138,
+      "step": 14350
+    },
+    {
+      "epoch": 1.240658343772949,
+      "grad_norm": 0.5573343725422848,
+      "learning_rate": 6.6132317284852656e-06,
+      "loss": 0.1115,
+      "step": 14360
+    },
+    {
+      "epoch": 1.2415223119789192,
+      "grad_norm": 0.5976405807104025,
+      "learning_rate": 6.60881656918645e-06,
+      "loss": 0.1119,
+      "step": 14370
+    },
+    {
+      "epoch": 1.2423862801848893,
+      "grad_norm": 0.5773790829133977,
+      "learning_rate": 6.60440001010024e-06,
+      "loss": 0.1109,
+      "step": 14380
+    },
+    {
+      "epoch": 1.2432502483908592,
+      "grad_norm": 0.5596129375427722,
+      "learning_rate": 6.599982055069363e-06,
+      "loss": 0.1133,
+      "step": 14390
+    },
+    {
+      "epoch": 1.2441142165968293,
+      "grad_norm": 0.5986154505530558,
+      "learning_rate": 6.595562707937759e-06,
+      "loss": 0.1112,
+      "step": 14400
+    },
+    {
+      "epoch": 1.2449781848027992,
+      "grad_norm": 0.5850475888093254,
+      "learning_rate": 6.591141972550581e-06,
+      "loss": 0.1122,
+      "step": 14410
+    },
+    {
+      "epoch": 1.2458421530087693,
+      "grad_norm": 0.5438809542880516,
+      "learning_rate": 6.5867198527541935e-06,
+      "loss": 0.1124,
+      "step": 14420
+    },
+    {
+      "epoch": 1.2467061212147392,
+      "grad_norm": 0.5854222234007489,
+      "learning_rate": 6.582296352396158e-06,
+      "loss": 0.1145,
+      "step": 14430
+    },
+    {
+      "epoch": 1.2475700894207093,
+      "grad_norm": 0.5647614923048787,
+      "learning_rate": 6.577871475325245e-06,
+      "loss": 0.1131,
+      "step": 14440
+    },
+    {
+      "epoch": 1.2484340576266792,
+      "grad_norm": 0.5446916876914634,
+      "learning_rate": 6.573445225391417e-06,
+      "loss": 0.1128,
+      "step": 14450
+    },
+    {
+      "epoch": 1.2492980258326494,
+      "grad_norm": 0.556275506719302,
+      "learning_rate": 6.569017606445836e-06,
+      "loss": 0.1104,
+      "step": 14460
+    },
+    {
+      "epoch": 1.2501619940386193,
+      "grad_norm": 0.5439613445971405,
+      "learning_rate": 6.564588622340848e-06,
+      "loss": 0.1085,
+      "step": 14470
+    },
+    {
+      "epoch": 1.2510259622445894,
+      "grad_norm": 0.5628308747399765,
+      "learning_rate": 6.560158276929996e-06,
+      "loss": 0.119,
+      "step": 14480
+    },
+    {
+      "epoch": 1.2518899304505595,
+      "grad_norm": 0.561991898239134,
+      "learning_rate": 6.555726574068e-06,
+      "loss": 0.1146,
+      "step": 14490
+    },
+    {
+      "epoch": 1.2527538986565294,
+      "grad_norm": 0.5751250413019706,
+      "learning_rate": 6.551293517610764e-06,
+      "loss": 0.1085,
+      "step": 14500
+    },
+    {
+      "epoch": 1.2536178668624995,
+      "grad_norm": 0.5640021015599359,
+      "learning_rate": 6.546859111415371e-06,
+      "loss": 0.1102,
+      "step": 14510
+    },
+    {
+      "epoch": 1.2544818350684694,
+      "grad_norm": 0.5635763265614476,
+      "learning_rate": 6.5424233593400735e-06,
+      "loss": 0.1145,
+      "step": 14520
+    },
+    {
+      "epoch": 1.2553458032744396,
+      "grad_norm": 0.591286004727335,
+      "learning_rate": 6.537986265244302e-06,
+      "loss": 0.111,
+      "step": 14530
+    },
+    {
+      "epoch": 1.2562097714804095,
+      "grad_norm": 0.5522099040559655,
+      "learning_rate": 6.533547832988647e-06,
+      "loss": 0.1111,
+      "step": 14540
+    },
+    {
+      "epoch": 1.2570737396863796,
+      "grad_norm": 0.5470523497296542,
+      "learning_rate": 6.529108066434872e-06,
+      "loss": 0.1114,
+      "step": 14550
+    },
+    {
+      "epoch": 1.2579377078923495,
+      "grad_norm": 0.568333276620229,
+      "learning_rate": 6.524666969445892e-06,
+      "loss": 0.1119,
+      "step": 14560
+    },
+    {
+      "epoch": 1.2588016760983196,
+      "grad_norm": 0.5850616288956318,
+      "learning_rate": 6.520224545885789e-06,
+      "loss": 0.1096,
+      "step": 14570
+    },
+    {
+      "epoch": 1.2596656443042895,
+      "grad_norm": 0.5773097418715036,
+      "learning_rate": 6.515780799619791e-06,
+      "loss": 0.1135,
+      "step": 14580
+    },
+    {
+      "epoch": 1.2605296125102596,
+      "grad_norm": 0.5388132754358779,
+      "learning_rate": 6.511335734514283e-06,
+      "loss": 0.1119,
+      "step": 14590
+    },
+    {
+      "epoch": 1.2613935807162298,
+      "grad_norm": 0.5577823715228171,
+      "learning_rate": 6.506889354436792e-06,
+      "loss": 0.1176,
+      "step": 14600
+    },
+    {
+      "epoch": 1.2622575489221997,
+      "grad_norm": 0.5339935975649758,
+      "learning_rate": 6.5024416632559935e-06,
+      "loss": 0.107,
+      "step": 14610
+    },
+    {
+      "epoch": 1.2631215171281696,
+      "grad_norm": 0.5732228640887302,
+      "learning_rate": 6.4979926648417025e-06,
+      "loss": 0.1124,
+      "step": 14620
+    },
+    {
+      "epoch": 1.2639854853341397,
+      "grad_norm": 0.5903353302881912,
+      "learning_rate": 6.493542363064871e-06,
+      "loss": 0.1145,
+      "step": 14630
+    },
+    {
+      "epoch": 1.2648494535401098,
+      "grad_norm": 0.5470991075480361,
+      "learning_rate": 6.489090761797583e-06,
+      "loss": 0.108,
+      "step": 14640
+    },
+    {
+      "epoch": 1.2657134217460797,
+      "grad_norm": 0.5625592769987781,
+      "learning_rate": 6.484637864913059e-06,
+      "loss": 0.1114,
+      "step": 14650
+    },
+    {
+      "epoch": 1.2665773899520498,
+      "grad_norm": 0.5689791373198402,
+      "learning_rate": 6.4801836762856406e-06,
+      "loss": 0.1133,
+      "step": 14660
+    },
+    {
+      "epoch": 1.2674413581580197,
+      "grad_norm": 0.5444276374757011,
+      "learning_rate": 6.475728199790796e-06,
+      "loss": 0.1094,
+      "step": 14670
+    },
+    {
+      "epoch": 1.2683053263639898,
+      "grad_norm": 0.5667653808009618,
+      "learning_rate": 6.471271439305115e-06,
+      "loss": 0.112,
+      "step": 14680
+    },
+    {
+      "epoch": 1.2691692945699597,
+      "grad_norm": 0.5344168071211037,
+      "learning_rate": 6.466813398706302e-06,
+      "loss": 0.1128,
+      "step": 14690
+    },
+    {
+      "epoch": 1.2700332627759299,
+      "grad_norm": 0.5510709964262109,
+      "learning_rate": 6.462354081873177e-06,
+      "loss": 0.1101,
+      "step": 14700
+    },
+    {
+      "epoch": 1.2708972309819,
+      "grad_norm": 0.5653351061563336,
+      "learning_rate": 6.457893492685671e-06,
+      "loss": 0.1126,
+      "step": 14710
+    },
+    {
+      "epoch": 1.27176119918787,
+      "grad_norm": 0.5565782210839264,
+      "learning_rate": 6.453431635024817e-06,
+      "loss": 0.111,
+      "step": 14720
+    },
+    {
+      "epoch": 1.2726251673938398,
+      "grad_norm": 0.5547676333122443,
+      "learning_rate": 6.4489685127727606e-06,
+      "loss": 0.1086,
+      "step": 14730
+    },
+    {
+      "epoch": 1.27348913559981,
+      "grad_norm": 0.544001901072926,
+      "learning_rate": 6.44450412981274e-06,
+      "loss": 0.1111,
+      "step": 14740
+    },
+    {
+      "epoch": 1.27435310380578,
+      "grad_norm": 0.5267941742084936,
+      "learning_rate": 6.440038490029092e-06,
+      "loss": 0.112,
+      "step": 14750
+    },
+    {
+      "epoch": 1.27521707201175,
+      "grad_norm": 0.5453459720243373,
+      "learning_rate": 6.435571597307251e-06,
+      "loss": 0.1119,
+      "step": 14760
+    },
+    {
+      "epoch": 1.27608104021772,
+      "grad_norm": 0.5647611308613276,
+      "learning_rate": 6.431103455533735e-06,
+      "loss": 0.1155,
+      "step": 14770
+    },
+    {
+      "epoch": 1.27694500842369,
+      "grad_norm": 0.5702592798618671,
+      "learning_rate": 6.426634068596154e-06,
+      "loss": 0.1077,
+      "step": 14780
+    },
+    {
+      "epoch": 1.27780897662966,
+      "grad_norm": 0.6036801878221308,
+      "learning_rate": 6.422163440383202e-06,
+      "loss": 0.1168,
+      "step": 14790
+    },
+    {
+      "epoch": 1.27867294483563,
+      "grad_norm": 0.5962193986965203,
+      "learning_rate": 6.417691574784647e-06,
+      "loss": 0.1155,
+      "step": 14800
+    },
+    {
+      "epoch": 1.2795369130416,
+      "grad_norm": 0.593769084774705,
+      "learning_rate": 6.413218475691337e-06,
+      "loss": 0.1136,
+      "step": 14810
+    },
+    {
+      "epoch": 1.28040088124757,
+      "grad_norm": 0.5378674069612178,
+      "learning_rate": 6.408744146995197e-06,
+      "loss": 0.1127,
+      "step": 14820
+    },
+    {
+      "epoch": 1.2812648494535401,
+      "grad_norm": 0.552753211413115,
+      "learning_rate": 6.404268592589214e-06,
+      "loss": 0.1095,
+      "step": 14830
+    },
+    {
+      "epoch": 1.28212881765951,
+      "grad_norm": 0.5624083853062832,
+      "learning_rate": 6.39979181636745e-06,
+      "loss": 0.1084,
+      "step": 14840
+    },
+    {
+      "epoch": 1.2829927858654802,
+      "grad_norm": 0.5601341611146992,
+      "learning_rate": 6.395313822225022e-06,
+      "loss": 0.1125,
+      "step": 14850
+    },
+    {
+      "epoch": 1.2838567540714503,
+      "grad_norm": 0.5172364101610524,
+      "learning_rate": 6.390834614058114e-06,
+      "loss": 0.1114,
+      "step": 14860
+    },
+    {
+      "epoch": 1.2847207222774202,
+      "grad_norm": 0.5229962139350934,
+      "learning_rate": 6.38635419576396e-06,
+      "loss": 0.1132,
+      "step": 14870
+    },
+    {
+      "epoch": 1.28558469048339,
+      "grad_norm": 0.5756669235276612,
+      "learning_rate": 6.381872571240852e-06,
+      "loss": 0.1147,
+      "step": 14880
+    },
+    {
+      "epoch": 1.2864486586893602,
+      "grad_norm": 0.5563681551307071,
+      "learning_rate": 6.377389744388127e-06,
+      "loss": 0.1095,
+      "step": 14890
+    },
+    {
+      "epoch": 1.2873126268953303,
+      "grad_norm": 0.5854403852531498,
+      "learning_rate": 6.372905719106172e-06,
+      "loss": 0.1128,
+      "step": 14900
+    },
+    {
+      "epoch": 1.2881765951013002,
+      "grad_norm": 0.6036186605098479,
+      "learning_rate": 6.368420499296413e-06,
+      "loss": 0.1126,
+      "step": 14910
+    },
+    {
+      "epoch": 1.2890405633072703,
+      "grad_norm": 0.593748655103996,
+      "learning_rate": 6.36393408886132e-06,
+      "loss": 0.1148,
+      "step": 14920
+    },
+    {
+      "epoch": 1.2899045315132402,
+      "grad_norm": 0.5552533246969583,
+      "learning_rate": 6.359446491704394e-06,
+      "loss": 0.11,
+      "step": 14930
+    },
+    {
+      "epoch": 1.2907684997192104,
+      "grad_norm": 0.5879665603388455,
+      "learning_rate": 6.3549577117301735e-06,
+      "loss": 0.1121,
+      "step": 14940
+    },
+    {
+      "epoch": 1.2916324679251803,
+      "grad_norm": 0.553609513620935,
+      "learning_rate": 6.35046775284422e-06,
+      "loss": 0.1108,
+      "step": 14950
+    },
+    {
+      "epoch": 1.2924964361311504,
+      "grad_norm": 0.5804539813996858,
+      "learning_rate": 6.345976618953127e-06,
+      "loss": 0.1125,
+      "step": 14960
+    },
+    {
+      "epoch": 1.2933604043371205,
+      "grad_norm": 0.5404050350479398,
+      "learning_rate": 6.341484313964506e-06,
+      "loss": 0.108,
+      "step": 14970
+    },
+    {
+      "epoch": 1.2942243725430904,
+      "grad_norm": 0.5717453327587577,
+      "learning_rate": 6.33699084178699e-06,
+      "loss": 0.1132,
+      "step": 14980
+    },
+    {
+      "epoch": 1.2950883407490603,
+      "grad_norm": 0.5571547661988168,
+      "learning_rate": 6.332496206330227e-06,
+      "loss": 0.1118,
+      "step": 14990
+    },
+    {
+      "epoch": 1.2959523089550304,
+      "grad_norm": 0.5894118107294782,
+      "learning_rate": 6.328000411504876e-06,
+      "loss": 0.1119,
+      "step": 15000
+    },
+    {
+      "epoch": 1.2968162771610006,
+      "grad_norm": 0.6271319736229325,
+      "learning_rate": 6.323503461222602e-06,
+      "loss": 0.1093,
+      "step": 15010
+    },
+    {
+      "epoch": 1.2976802453669705,
+      "grad_norm": 0.5714163744986458,
+      "learning_rate": 6.319005359396084e-06,
+      "loss": 0.1163,
+      "step": 15020
+    },
+    {
+      "epoch": 1.2985442135729406,
+      "grad_norm": 0.570485713384322,
+      "learning_rate": 6.314506109938996e-06,
+      "loss": 0.1133,
+      "step": 15030
+    },
+    {
+      "epoch": 1.2994081817789105,
+      "grad_norm": 0.6031381548468219,
+      "learning_rate": 6.31000571676601e-06,
+      "loss": 0.1148,
+      "step": 15040
+    },
+    {
+      "epoch": 1.3002721499848806,
+      "grad_norm": 0.5782702057393322,
+      "learning_rate": 6.305504183792799e-06,
+      "loss": 0.1112,
+      "step": 15050
+    },
+    {
+      "epoch": 1.3011361181908505,
+      "grad_norm": 0.5618969513353681,
+      "learning_rate": 6.30100151493602e-06,
+      "loss": 0.1132,
+      "step": 15060
+    },
+    {
+      "epoch": 1.3020000863968206,
+      "grad_norm": 0.5627948160701411,
+      "learning_rate": 6.296497714113325e-06,
+      "loss": 0.1113,
+      "step": 15070
+    },
+    {
+      "epoch": 1.3028640546027908,
+      "grad_norm": 0.5534126614895483,
+      "learning_rate": 6.291992785243348e-06,
+      "loss": 0.112,
+      "step": 15080
+    },
+    {
+      "epoch": 1.3037280228087607,
+      "grad_norm": 0.5912501317649836,
+      "learning_rate": 6.287486732245705e-06,
+      "loss": 0.112,
+      "step": 15090
+    },
+    {
+      "epoch": 1.3045919910147306,
+      "grad_norm": 0.5600010807396356,
+      "learning_rate": 6.282979559040988e-06,
+      "loss": 0.1141,
+      "step": 15100
+    },
+    {
+      "epoch": 1.3054559592207007,
+      "grad_norm": 0.5935733410748419,
+      "learning_rate": 6.278471269550766e-06,
+      "loss": 0.1139,
+      "step": 15110
+    },
+    {
+      "epoch": 1.3063199274266708,
+      "grad_norm": 0.5600831456724733,
+      "learning_rate": 6.2739618676975785e-06,
+      "loss": 0.1116,
+      "step": 15120
+    },
+    {
+      "epoch": 1.3071838956326407,
+      "grad_norm": 0.5599466998893358,
+      "learning_rate": 6.2694513574049354e-06,
+      "loss": 0.1102,
+      "step": 15130
+    },
+    {
+      "epoch": 1.3080478638386106,
+      "grad_norm": 0.5559964600507826,
+      "learning_rate": 6.264939742597304e-06,
+      "loss": 0.1102,
+      "step": 15140
+    },
+    {
+      "epoch": 1.3089118320445807,
+      "grad_norm": 0.5786344058859312,
+      "learning_rate": 6.2604270272001235e-06,
+      "loss": 0.1131,
+      "step": 15150
+    },
+    {
+      "epoch": 1.3097758002505508,
+      "grad_norm": 0.5657515488388618,
+      "learning_rate": 6.255913215139778e-06,
+      "loss": 0.113,
+      "step": 15160
+    },
+    {
+      "epoch": 1.3106397684565207,
+      "grad_norm": 0.5863407258722373,
+      "learning_rate": 6.251398310343617e-06,
+      "loss": 0.1118,
+      "step": 15170
+    },
+    {
+      "epoch": 1.3115037366624909,
+      "grad_norm": 0.5442709891236638,
+      "learning_rate": 6.246882316739932e-06,
+      "loss": 0.1129,
+      "step": 15180
+    },
+    {
+      "epoch": 1.3123677048684608,
+      "grad_norm": 0.6002637189048005,
+      "learning_rate": 6.242365238257969e-06,
+      "loss": 0.1137,
+      "step": 15190
+    },
+    {
+      "epoch": 1.313231673074431,
+      "grad_norm": 0.5842893880761071,
+      "learning_rate": 6.237847078827914e-06,
+      "loss": 0.1126,
+      "step": 15200
+    },
+    {
+      "epoch": 1.3140956412804008,
+      "grad_norm": 0.5830066962552539,
+      "learning_rate": 6.233327842380894e-06,
+      "loss": 0.1161,
+      "step": 15210
+    },
+    {
+      "epoch": 1.314959609486371,
+      "grad_norm": 0.5373321121823259,
+      "learning_rate": 6.228807532848973e-06,
+      "loss": 0.111,
+      "step": 15220
+    },
+    {
+      "epoch": 1.315823577692341,
+      "grad_norm": 0.6108127472256695,
+      "learning_rate": 6.224286154165148e-06,
+      "loss": 0.1101,
+      "step": 15230
+    },
+    {
+      "epoch": 1.316687545898311,
+      "grad_norm": 0.5696760440842511,
+      "learning_rate": 6.219763710263349e-06,
+      "loss": 0.1081,
+      "step": 15240
+    },
+    {
+      "epoch": 1.3175515141042808,
+      "grad_norm": 0.5489297133762001,
+      "learning_rate": 6.21524020507843e-06,
+      "loss": 0.1118,
+      "step": 15250
+    },
+    {
+      "epoch": 1.318415482310251,
+      "grad_norm": 0.5613132623439822,
+      "learning_rate": 6.21071564254617e-06,
+      "loss": 0.1143,
+      "step": 15260
+    },
+    {
+      "epoch": 1.319279450516221,
+      "grad_norm": 0.5526160450209798,
+      "learning_rate": 6.206190026603267e-06,
+      "loss": 0.1126,
+      "step": 15270
+    },
+    {
+      "epoch": 1.320143418722191,
+      "grad_norm": 0.5769836694482383,
+      "learning_rate": 6.201663361187336e-06,
+      "loss": 0.1087,
+      "step": 15280
+    },
+    {
+      "epoch": 1.321007386928161,
+      "grad_norm": 0.55598238886065,
+      "learning_rate": 6.1971356502369065e-06,
+      "loss": 0.1131,
+      "step": 15290
+    },
+    {
+      "epoch": 1.321871355134131,
+      "grad_norm": 0.5735365686216224,
+      "learning_rate": 6.192606897691414e-06,
+      "loss": 0.1126,
+      "step": 15300
+    },
+    {
+      "epoch": 1.3227353233401011,
+      "grad_norm": 0.5783112362966369,
+      "learning_rate": 6.188077107491204e-06,
+      "loss": 0.1112,
+      "step": 15310
+    },
+    {
+      "epoch": 1.323599291546071,
+      "grad_norm": 0.5683067930605199,
+      "learning_rate": 6.183546283577523e-06,
+      "loss": 0.1121,
+      "step": 15320
+    },
+    {
+      "epoch": 1.3244632597520412,
+      "grad_norm": 0.5464109155323802,
+      "learning_rate": 6.1790144298925146e-06,
+      "loss": 0.1137,
+      "step": 15330
+    },
+    {
+      "epoch": 1.3253272279580113,
+      "grad_norm": 0.5585746473787109,
+      "learning_rate": 6.174481550379225e-06,
+      "loss": 0.1134,
+      "step": 15340
+    },
+    {
+      "epoch": 1.3261911961639812,
+      "grad_norm": 0.5545762428074108,
+      "learning_rate": 6.1699476489815876e-06,
+      "loss": 0.1149,
+      "step": 15350
+    },
+    {
+      "epoch": 1.327055164369951,
+      "grad_norm": 0.5748347832493429,
+      "learning_rate": 6.165412729644426e-06,
+      "loss": 0.109,
+      "step": 15360
+    },
+    {
+      "epoch": 1.3279191325759212,
+      "grad_norm": 0.5842776295530736,
+      "learning_rate": 6.1608767963134495e-06,
+      "loss": 0.1072,
+      "step": 15370
+    },
+    {
+      "epoch": 1.3287831007818913,
+      "grad_norm": 0.5575191047466767,
+      "learning_rate": 6.156339852935251e-06,
+      "loss": 0.1117,
+      "step": 15380
+    },
+    {
+      "epoch": 1.3296470689878612,
+      "grad_norm": 0.5450731649894345,
+      "learning_rate": 6.1518019034572995e-06,
+      "loss": 0.1144,
+      "step": 15390
+    },
+    {
+      "epoch": 1.3305110371938313,
+      "grad_norm": 0.5624842087427264,
+      "learning_rate": 6.147262951827943e-06,
+      "loss": 0.1107,
+      "step": 15400
+    },
+    {
+      "epoch": 1.3313750053998012,
+      "grad_norm": 0.566230353504596,
+      "learning_rate": 6.142723001996398e-06,
+      "loss": 0.1078,
+      "step": 15410
+    },
+    {
+      "epoch": 1.3322389736057714,
+      "grad_norm": 0.5596105650561354,
+      "learning_rate": 6.138182057912751e-06,
+      "loss": 0.1114,
+      "step": 15420
+    },
+    {
+      "epoch": 1.3331029418117413,
+      "grad_norm": 0.5631168310453766,
+      "learning_rate": 6.133640123527953e-06,
+      "loss": 0.1094,
+      "step": 15430
+    },
+    {
+      "epoch": 1.3339669100177114,
+      "grad_norm": 0.5757668723185331,
+      "learning_rate": 6.12909720279382e-06,
+      "loss": 0.1133,
+      "step": 15440
+    },
+    {
+      "epoch": 1.3348308782236813,
+      "grad_norm": 0.5656227206741703,
+      "learning_rate": 6.12455329966302e-06,
+      "loss": 0.1102,
+      "step": 15450
+    },
+    {
+      "epoch": 1.3356948464296514,
+      "grad_norm": 0.5675203492355643,
+      "learning_rate": 6.120008418089082e-06,
+      "loss": 0.1108,
+      "step": 15460
+    },
+    {
+      "epoch": 1.3365588146356213,
+      "grad_norm": 0.5609540526824188,
+      "learning_rate": 6.11546256202638e-06,
+      "loss": 0.1083,
+      "step": 15470
+    },
+    {
+      "epoch": 1.3374227828415914,
+      "grad_norm": 0.5453368748675613,
+      "learning_rate": 6.110915735430142e-06,
+      "loss": 0.1068,
+      "step": 15480
+    },
+    {
+      "epoch": 1.3382867510475616,
+      "grad_norm": 0.5783208678504869,
+      "learning_rate": 6.106367942256437e-06,
+      "loss": 0.114,
+      "step": 15490
+    },
+    {
+      "epoch": 1.3391507192535315,
+      "grad_norm": 0.5772617722487127,
+      "learning_rate": 6.1018191864621764e-06,
+      "loss": 0.1103,
+      "step": 15500
+    },
+    {
+      "epoch": 1.3400146874595014,
+      "grad_norm": 0.5726131201792536,
+      "learning_rate": 6.097269472005107e-06,
+      "loss": 0.1108,
+      "step": 15510
+    },
+    {
+      "epoch": 1.3408786556654715,
+      "grad_norm": 0.5397368294651602,
+      "learning_rate": 6.092718802843814e-06,
+      "loss": 0.1106,
+      "step": 15520
+    },
+    {
+      "epoch": 1.3417426238714416,
+      "grad_norm": 0.5591733492489677,
+      "learning_rate": 6.088167182937706e-06,
+      "loss": 0.1129,
+      "step": 15530
+    },
+    {
+      "epoch": 1.3426065920774115,
+      "grad_norm": 0.5561490349875766,
+      "learning_rate": 6.083614616247028e-06,
+      "loss": 0.112,
+      "step": 15540
+    },
+    {
+      "epoch": 1.3434705602833816,
+      "grad_norm": 0.5819373721015771,
+      "learning_rate": 6.0790611067328395e-06,
+      "loss": 0.1072,
+      "step": 15550
+    },
+    {
+      "epoch": 1.3443345284893515,
+      "grad_norm": 0.5359474980622055,
+      "learning_rate": 6.074506658357029e-06,
+      "loss": 0.1109,
+      "step": 15560
+    },
+    {
+      "epoch": 1.3451984966953217,
+      "grad_norm": 0.5720230417354536,
+      "learning_rate": 6.069951275082295e-06,
+      "loss": 0.1078,
+      "step": 15570
+    },
+    {
+      "epoch": 1.3460624649012916,
+      "grad_norm": 0.5432165606654724,
+      "learning_rate": 6.065394960872154e-06,
+      "loss": 0.1078,
+      "step": 15580
+    },
+    {
+      "epoch": 1.3469264331072617,
+      "grad_norm": 0.5956848004967984,
+      "learning_rate": 6.06083771969093e-06,
+      "loss": 0.1087,
+      "step": 15590
+    },
+    {
+      "epoch": 1.3477904013132318,
+      "grad_norm": 0.5831628365743788,
+      "learning_rate": 6.056279555503752e-06,
+      "loss": 0.1075,
+      "step": 15600
+    },
+    {
+      "epoch": 1.3486543695192017,
+      "grad_norm": 0.5360439964712416,
+      "learning_rate": 6.0517204722765585e-06,
+      "loss": 0.1083,
+      "step": 15610
+    },
+    {
+      "epoch": 1.3495183377251716,
+      "grad_norm": 0.5608239905295521,
+      "learning_rate": 6.047160473976081e-06,
+      "loss": 0.1145,
+      "step": 15620
+    },
+    {
+      "epoch": 1.3503823059311417,
+      "grad_norm": 0.5681136455973719,
+      "learning_rate": 6.04259956456985e-06,
+      "loss": 0.1091,
+      "step": 15630
+    },
+    {
+      "epoch": 1.3512462741371118,
+      "grad_norm": 0.5358581606501858,
+      "learning_rate": 6.038037748026187e-06,
+      "loss": 0.1095,
+      "step": 15640
+    },
+    {
+      "epoch": 1.3521102423430817,
+      "grad_norm": 0.5761867356055901,
+      "learning_rate": 6.0334750283142056e-06,
+      "loss": 0.1113,
+      "step": 15650
+    },
+    {
+      "epoch": 1.3529742105490519,
+      "grad_norm": 0.5717376371598626,
+      "learning_rate": 6.028911409403801e-06,
+      "loss": 0.1122,
+      "step": 15660
+    },
+    {
+      "epoch": 1.3538381787550218,
+      "grad_norm": 0.5571214583689651,
+      "learning_rate": 6.024346895265659e-06,
+      "loss": 0.1127,
+      "step": 15670
+    },
+    {
+      "epoch": 1.354702146960992,
+      "grad_norm": 0.5875745499147877,
+      "learning_rate": 6.019781489871235e-06,
+      "loss": 0.1083,
+      "step": 15680
+    },
+    {
+      "epoch": 1.3555661151669618,
+      "grad_norm": 0.5546175269594011,
+      "learning_rate": 6.015215197192763e-06,
+      "loss": 0.1079,
+      "step": 15690
+    },
+    {
+      "epoch": 1.356430083372932,
+      "grad_norm": 0.5450808281935141,
+      "learning_rate": 6.010648021203252e-06,
+      "loss": 0.1112,
+      "step": 15700
+    },
+    {
+      "epoch": 1.3572940515789018,
+      "grad_norm": 0.549235755180578,
+      "learning_rate": 6.006079965876476e-06,
+      "loss": 0.1085,
+      "step": 15710
+    },
+    {
+      "epoch": 1.358158019784872,
+      "grad_norm": 0.5623635978497271,
+      "learning_rate": 6.001511035186975e-06,
+      "loss": 0.1078,
+      "step": 15720
+    },
+    {
+      "epoch": 1.3590219879908418,
+      "grad_norm": 0.5484310920570102,
+      "learning_rate": 5.996941233110052e-06,
+      "loss": 0.1092,
+      "step": 15730
+    },
+    {
+      "epoch": 1.359885956196812,
+      "grad_norm": 0.5154476200776394,
+      "learning_rate": 5.992370563621766e-06,
+      "loss": 0.1085,
+      "step": 15740
+    },
+    {
+      "epoch": 1.360749924402782,
+      "grad_norm": 0.5553387563055833,
+      "learning_rate": 5.987799030698932e-06,
+      "loss": 0.112,
+      "step": 15750
+    },
+    {
+      "epoch": 1.361613892608752,
+      "grad_norm": 0.5580242137209885,
+      "learning_rate": 5.9832266383191154e-06,
+      "loss": 0.1118,
+      "step": 15760
+    },
+    {
+      "epoch": 1.3624778608147219,
+      "grad_norm": 0.5283878918575788,
+      "learning_rate": 5.978653390460632e-06,
+      "loss": 0.1087,
+      "step": 15770
+    },
+    {
+      "epoch": 1.363341829020692,
+      "grad_norm": 0.5785508559189512,
+      "learning_rate": 5.974079291102538e-06,
+      "loss": 0.1108,
+      "step": 15780
+    },
+    {
+      "epoch": 1.3642057972266621,
+      "grad_norm": 0.55241795550289,
+      "learning_rate": 5.969504344224635e-06,
+      "loss": 0.1118,
+      "step": 15790
+    },
+    {
+      "epoch": 1.365069765432632,
+      "grad_norm": 0.5351889601182599,
+      "learning_rate": 5.964928553807455e-06,
+      "loss": 0.1095,
+      "step": 15800
+    },
+    {
+      "epoch": 1.3659337336386022,
+      "grad_norm": 0.5964305804829771,
+      "learning_rate": 5.960351923832274e-06,
+      "loss": 0.1085,
+      "step": 15810
+    },
+    {
+      "epoch": 1.366797701844572,
+      "grad_norm": 0.5659343518769593,
+      "learning_rate": 5.955774458281088e-06,
+      "loss": 0.1119,
+      "step": 15820
+    },
+    {
+      "epoch": 1.3676616700505422,
+      "grad_norm": 0.5387151801430248,
+      "learning_rate": 5.951196161136629e-06,
+      "loss": 0.1104,
+      "step": 15830
+    },
+    {
+      "epoch": 1.368525638256512,
+      "grad_norm": 1.5548648788419452,
+      "learning_rate": 5.946617036382344e-06,
+      "loss": 0.1102,
+      "step": 15840
+    },
+    {
+      "epoch": 1.3693896064624822,
+      "grad_norm": 0.5558801260199372,
+      "learning_rate": 5.942037088002408e-06,
+      "loss": 0.111,
+      "step": 15850
+    },
+    {
+      "epoch": 1.3702535746684523,
+      "grad_norm": 0.575072919526527,
+      "learning_rate": 5.937456319981706e-06,
+      "loss": 0.1137,
+      "step": 15860
+    },
+    {
+      "epoch": 1.3711175428744222,
+      "grad_norm": 0.5530395900840259,
+      "learning_rate": 5.932874736305843e-06,
+      "loss": 0.1101,
+      "step": 15870
+    },
+    {
+      "epoch": 1.3719815110803921,
+      "grad_norm": 0.5983896943740621,
+      "learning_rate": 5.9282923409611285e-06,
+      "loss": 0.1075,
+      "step": 15880
+    },
+    {
+      "epoch": 1.3728454792863622,
+      "grad_norm": 0.5768004665520459,
+      "learning_rate": 5.9237091379345765e-06,
+      "loss": 0.1107,
+      "step": 15890
+    },
+    {
+      "epoch": 1.3737094474923324,
+      "grad_norm": 0.569097515223848,
+      "learning_rate": 5.919125131213912e-06,
+      "loss": 0.1089,
+      "step": 15900
+    },
+    {
+      "epoch": 1.3745734156983023,
+      "grad_norm": 0.5312624276461287,
+      "learning_rate": 5.91454032478755e-06,
+      "loss": 0.11,
+      "step": 15910
+    },
+    {
+      "epoch": 1.3754373839042724,
+      "grad_norm": 0.5328951454517763,
+      "learning_rate": 5.909954722644608e-06,
+      "loss": 0.1088,
+      "step": 15920
+    },
+    {
+      "epoch": 1.3763013521102423,
+      "grad_norm": 0.5722786730159535,
+      "learning_rate": 5.905368328774893e-06,
+      "loss": 0.1071,
+      "step": 15930
+    },
+    {
+      "epoch": 1.3771653203162124,
+      "grad_norm": 0.582982357808919,
+      "learning_rate": 5.9007811471689e-06,
+      "loss": 0.1107,
+      "step": 15940
+    },
+    {
+      "epoch": 1.3780292885221823,
+      "grad_norm": 0.5409670522081398,
+      "learning_rate": 5.896193181817811e-06,
+      "loss": 0.1123,
+      "step": 15950
+    },
+    {
+      "epoch": 1.3788932567281524,
+      "grad_norm": 0.5525488935001114,
+      "learning_rate": 5.891604436713491e-06,
+      "loss": 0.1101,
+      "step": 15960
+    },
+    {
+      "epoch": 1.3797572249341226,
+      "grad_norm": 0.5840786377822249,
+      "learning_rate": 5.887014915848478e-06,
+      "loss": 0.1097,
+      "step": 15970
+    },
+    {
+      "epoch": 1.3806211931400925,
+      "grad_norm": 0.5384412553004606,
+      "learning_rate": 5.882424623215993e-06,
+      "loss": 0.1088,
+      "step": 15980
+    },
+    {
+      "epoch": 1.3814851613460624,
+      "grad_norm": 0.5833292422162293,
+      "learning_rate": 5.877833562809922e-06,
+      "loss": 0.1123,
+      "step": 15990
+    },
+    {
+      "epoch": 1.3823491295520325,
+      "grad_norm": 0.5517932567301533,
+      "learning_rate": 5.873241738624824e-06,
+      "loss": 0.1057,
+      "step": 16000
+    },
+    {
+      "epoch": 1.3832130977580026,
+      "grad_norm": 0.5756476697861506,
+      "learning_rate": 5.8686491546559154e-06,
+      "loss": 0.1086,
+      "step": 16010
+    },
+    {
+      "epoch": 1.3840770659639725,
+      "grad_norm": 0.5785581813009524,
+      "learning_rate": 5.8640558148990835e-06,
+      "loss": 0.1115,
+      "step": 16020
+    },
+    {
+      "epoch": 1.3849410341699424,
+      "grad_norm": 0.5943858429441137,
+      "learning_rate": 5.8594617233508645e-06,
+      "loss": 0.1092,
+      "step": 16030
+    },
+    {
+      "epoch": 1.3858050023759125,
+      "grad_norm": 0.5745575971627046,
+      "learning_rate": 5.854866884008452e-06,
+      "loss": 0.1102,
+      "step": 16040
+    },
+    {
+      "epoch": 1.3866689705818827,
+      "grad_norm": 0.5348702159362239,
+      "learning_rate": 5.850271300869691e-06,
+      "loss": 0.1096,
+      "step": 16050
+    },
+    {
+      "epoch": 1.3875329387878526,
+      "grad_norm": 0.5694557334714356,
+      "learning_rate": 5.845674977933074e-06,
+      "loss": 0.1085,
+      "step": 16060
+    },
+    {
+      "epoch": 1.3883969069938227,
+      "grad_norm": 0.5693489028372023,
+      "learning_rate": 5.841077919197734e-06,
+      "loss": 0.1086,
+      "step": 16070
+    },
+    {
+      "epoch": 1.3892608751997926,
+      "grad_norm": 0.5593707766156559,
+      "learning_rate": 5.8364801286634486e-06,
+      "loss": 0.1095,
+      "step": 16080
+    },
+    {
+      "epoch": 1.3901248434057627,
+      "grad_norm": 0.6124636333798967,
+      "learning_rate": 5.831881610330628e-06,
+      "loss": 0.1166,
+      "step": 16090
+    },
+    {
+      "epoch": 1.3909888116117326,
+      "grad_norm": 0.5846656666878277,
+      "learning_rate": 5.8272823682003196e-06,
+      "loss": 0.1111,
+      "step": 16100
+    },
+    {
+      "epoch": 1.3918527798177027,
+      "grad_norm": 0.5720609742683443,
+      "learning_rate": 5.822682406274197e-06,
+      "loss": 0.1097,
+      "step": 16110
+    },
+    {
+      "epoch": 1.3927167480236728,
+      "grad_norm": 0.5513132732611473,
+      "learning_rate": 5.818081728554562e-06,
+      "loss": 0.1106,
+      "step": 16120
+    },
+    {
+      "epoch": 1.3935807162296427,
+      "grad_norm": 0.5149649775459123,
+      "learning_rate": 5.81348033904434e-06,
+      "loss": 0.1114,
+      "step": 16130
+    },
+    {
+      "epoch": 1.3944446844356126,
+      "grad_norm": 0.5869210395463293,
+      "learning_rate": 5.808878241747071e-06,
+      "loss": 0.1049,
+      "step": 16140
+    },
+    {
+      "epoch": 1.3953086526415828,
+      "grad_norm": 0.5625464154152859,
+      "learning_rate": 5.804275440666918e-06,
+      "loss": 0.1097,
+      "step": 16150
+    },
+    {
+      "epoch": 1.396172620847553,
+      "grad_norm": 0.5813811467929083,
+      "learning_rate": 5.7996719398086485e-06,
+      "loss": 0.1119,
+      "step": 16160
+    },
+    {
+      "epoch": 1.3970365890535228,
+      "grad_norm": 0.5725158526835552,
+      "learning_rate": 5.795067743177648e-06,
+      "loss": 0.1099,
+      "step": 16170
+    },
+    {
+      "epoch": 1.397900557259493,
+      "grad_norm": 0.5720353741750168,
+      "learning_rate": 5.790462854779898e-06,
+      "loss": 0.1083,
+      "step": 16180
+    },
+    {
+      "epoch": 1.3987645254654628,
+      "grad_norm": 0.5757257456663112,
+      "learning_rate": 5.785857278621989e-06,
+      "loss": 0.1074,
+      "step": 16190
+    },
+    {
+      "epoch": 1.399628493671433,
+      "grad_norm": 0.5751688492722273,
+      "learning_rate": 5.7812510187111045e-06,
+      "loss": 0.111,
+      "step": 16200
+    },
+    {
+      "epoch": 1.4004924618774028,
+      "grad_norm": 0.5308149515056841,
+      "learning_rate": 5.776644079055029e-06,
+      "loss": 0.1086,
+      "step": 16210
+    },
+    {
+      "epoch": 1.401356430083373,
+      "grad_norm": 0.5457555860479729,
+      "learning_rate": 5.772036463662133e-06,
+      "loss": 0.1095,
+      "step": 16220
+    },
+    {
+      "epoch": 1.402220398289343,
+      "grad_norm": 0.5525027323997239,
+      "learning_rate": 5.7674281765413775e-06,
+      "loss": 0.1098,
+      "step": 16230
+    },
+    {
+      "epoch": 1.403084366495313,
+      "grad_norm": 0.5552174838299228,
+      "learning_rate": 5.7628192217023075e-06,
+      "loss": 0.109,
+      "step": 16240
+    },
+    {
+      "epoch": 1.4039483347012829,
+      "grad_norm": 0.6052659498732853,
+      "learning_rate": 5.758209603155047e-06,
+      "loss": 0.1072,
+      "step": 16250
+    },
+    {
+      "epoch": 1.404812302907253,
+      "grad_norm": 0.5614718168621989,
+      "learning_rate": 5.7535993249103e-06,
+      "loss": 0.1106,
+      "step": 16260
+    },
+    {
+      "epoch": 1.4056762711132231,
+      "grad_norm": 0.5536105288949624,
+      "learning_rate": 5.748988390979346e-06,
+      "loss": 0.11,
+      "step": 16270
+    },
+    {
+      "epoch": 1.406540239319193,
+      "grad_norm": 0.5611334021841203,
+      "learning_rate": 5.74437680537403e-06,
+      "loss": 0.1101,
+      "step": 16280
+    },
+    {
+      "epoch": 1.4074042075251632,
+      "grad_norm": 0.5573869486007782,
+      "learning_rate": 5.73976457210677e-06,
+      "loss": 0.1105,
+      "step": 16290
+    },
+    {
+      "epoch": 1.408268175731133,
+      "grad_norm": 0.5516200215230308,
+      "learning_rate": 5.73515169519054e-06,
+      "loss": 0.1082,
+      "step": 16300
+    },
+    {
+      "epoch": 1.4091321439371032,
+      "grad_norm": 0.5526404729155173,
+      "learning_rate": 5.730538178638881e-06,
+      "loss": 0.1101,
+      "step": 16310
+    },
+    {
+      "epoch": 1.409996112143073,
+      "grad_norm": 0.5719197743944662,
+      "learning_rate": 5.725924026465888e-06,
+      "loss": 0.1112,
+      "step": 16320
+    },
+    {
+      "epoch": 1.4108600803490432,
+      "grad_norm": 0.5694236895947357,
+      "learning_rate": 5.721309242686209e-06,
+      "loss": 0.1093,
+      "step": 16330
+    },
+    {
+      "epoch": 1.411724048555013,
+      "grad_norm": 0.6141799055072968,
+      "learning_rate": 5.716693831315041e-06,
+      "loss": 0.1126,
+      "step": 16340
+    },
+    {
+      "epoch": 1.4125880167609832,
+      "grad_norm": 0.5781600972939118,
+      "learning_rate": 5.7120777963681276e-06,
+      "loss": 0.1133,
+      "step": 16350
+    },
+    {
+      "epoch": 1.4134519849669531,
+      "grad_norm": 0.5489573576838745,
+      "learning_rate": 5.707461141861753e-06,
+      "loss": 0.1085,
+      "step": 16360
+    },
+    {
+      "epoch": 1.4143159531729232,
+      "grad_norm": 0.5511980848938117,
+      "learning_rate": 5.702843871812745e-06,
+      "loss": 0.1093,
+      "step": 16370
+    },
+    {
+      "epoch": 1.4151799213788934,
+      "grad_norm": 0.5694496019360336,
+      "learning_rate": 5.698225990238463e-06,
+      "loss": 0.1069,
+      "step": 16380
+    },
+    {
+      "epoch": 1.4160438895848633,
+      "grad_norm": 0.587017896333731,
+      "learning_rate": 5.6936075011568e-06,
+      "loss": 0.1109,
+      "step": 16390
+    },
+    {
+      "epoch": 1.4169078577908332,
+      "grad_norm": 0.568063496878268,
+      "learning_rate": 5.688988408586177e-06,
+      "loss": 0.1093,
+      "step": 16400
+    },
+    {
+      "epoch": 1.4177718259968033,
+      "grad_norm": 0.5679995822925583,
+      "learning_rate": 5.684368716545541e-06,
+      "loss": 0.106,
+      "step": 16410
+    },
+    {
+      "epoch": 1.4186357942027734,
+      "grad_norm": 0.5734866434303718,
+      "learning_rate": 5.6797484290543594e-06,
+      "loss": 0.1119,
+      "step": 16420
+    },
+    {
+      "epoch": 1.4194997624087433,
+      "grad_norm": 0.5431629848080799,
+      "learning_rate": 5.6751275501326185e-06,
+      "loss": 0.11,
+      "step": 16430
+    },
+    {
+      "epoch": 1.4203637306147134,
+      "grad_norm": 0.5720416624760498,
+      "learning_rate": 5.67050608380082e-06,
+      "loss": 0.1087,
+      "step": 16440
+    },
+    {
+      "epoch": 1.4212276988206833,
+      "grad_norm": 0.580302420152164,
+      "learning_rate": 5.665884034079974e-06,
+      "loss": 0.1087,
+      "step": 16450
+    },
+    {
+      "epoch": 1.4220916670266535,
+      "grad_norm": 0.5840176383870129,
+      "learning_rate": 5.6612614049916e-06,
+      "loss": 0.1068,
+      "step": 16460
+    },
+    {
+      "epoch": 1.4229556352326234,
+      "grad_norm": 0.5619479511353946,
+      "learning_rate": 5.656638200557723e-06,
+      "loss": 0.1064,
+      "step": 16470
+    },
+    {
+      "epoch": 1.4238196034385935,
+      "grad_norm": 0.563149143445443,
+      "learning_rate": 5.652014424800865e-06,
+      "loss": 0.112,
+      "step": 16480
+    },
+    {
+      "epoch": 1.4246835716445636,
+      "grad_norm": 0.554179718385933,
+      "learning_rate": 5.647390081744047e-06,
+      "loss": 0.1118,
+      "step": 16490
+    },
+    {
+      "epoch": 1.4255475398505335,
+      "grad_norm": 0.5790448425729009,
+      "learning_rate": 5.6427651754107856e-06,
+      "loss": 0.113,
+      "step": 16500
+    },
+    {
+      "epoch": 1.4264115080565034,
+      "grad_norm": 0.5413167122533629,
+      "learning_rate": 5.638139709825085e-06,
+      "loss": 0.1105,
+      "step": 16510
+    },
+    {
+      "epoch": 1.4272754762624735,
+      "grad_norm": 0.5403591817971234,
+      "learning_rate": 5.633513689011436e-06,
+      "loss": 0.1094,
+      "step": 16520
+    },
+    {
+      "epoch": 1.4281394444684437,
+      "grad_norm": 0.5719167784121851,
+      "learning_rate": 5.628887116994812e-06,
+      "loss": 0.1069,
+      "step": 16530
+    },
+    {
+      "epoch": 1.4290034126744136,
+      "grad_norm": 0.5571643381657813,
+      "learning_rate": 5.624259997800671e-06,
+      "loss": 0.1082,
+      "step": 16540
+    },
+    {
+      "epoch": 1.4298673808803837,
+      "grad_norm": 0.5539391800128624,
+      "learning_rate": 5.61963233545494e-06,
+      "loss": 0.1076,
+      "step": 16550
+    },
+    {
+      "epoch": 1.4307313490863536,
+      "grad_norm": 0.5737886682031937,
+      "learning_rate": 5.615004133984022e-06,
+      "loss": 0.1087,
+      "step": 16560
+    },
+    {
+      "epoch": 1.4315953172923237,
+      "grad_norm": 0.5466801931023841,
+      "learning_rate": 5.610375397414788e-06,
+      "loss": 0.1101,
+      "step": 16570
+    },
+    {
+      "epoch": 1.4324592854982936,
+      "grad_norm": 0.5797886336713219,
+      "learning_rate": 5.605746129774577e-06,
+      "loss": 0.1114,
+      "step": 16580
+    },
+    {
+      "epoch": 1.4333232537042637,
+      "grad_norm": 0.5529740818600277,
+      "learning_rate": 5.601116335091189e-06,
+      "loss": 0.1074,
+      "step": 16590
+    },
+    {
+      "epoch": 1.4341872219102336,
+      "grad_norm": 0.5715134987232192,
+      "learning_rate": 5.59648601739288e-06,
+      "loss": 0.1084,
+      "step": 16600
+    },
+    {
+      "epoch": 1.4350511901162037,
+      "grad_norm": 0.5565474264744146,
+      "learning_rate": 5.591855180708365e-06,
+      "loss": 0.1082,
+      "step": 16610
+    },
+    {
+      "epoch": 1.4359151583221736,
+      "grad_norm": 0.5771393921573549,
+      "learning_rate": 5.587223829066807e-06,
+      "loss": 0.1132,
+      "step": 16620
+    },
+    {
+      "epoch": 1.4367791265281438,
+      "grad_norm": 0.5354337704232355,
+      "learning_rate": 5.582591966497818e-06,
+      "loss": 0.1077,
+      "step": 16630
+    },
+    {
+      "epoch": 1.437643094734114,
+      "grad_norm": 0.5704092054091621,
+      "learning_rate": 5.577959597031459e-06,
+      "loss": 0.1084,
+      "step": 16640
+    },
+    {
+      "epoch": 1.4385070629400838,
+      "grad_norm": 0.547835523974159,
+      "learning_rate": 5.573326724698225e-06,
+      "loss": 0.1069,
+      "step": 16650
+    },
+    {
+      "epoch": 1.4393710311460537,
+      "grad_norm": 0.5550047278667364,
+      "learning_rate": 5.568693353529053e-06,
+      "loss": 0.1092,
+      "step": 16660
+    },
+    {
+      "epoch": 1.4402349993520238,
+      "grad_norm": 0.5598159535674883,
+      "learning_rate": 5.564059487555311e-06,
+      "loss": 0.11,
+      "step": 16670
+    },
+    {
+      "epoch": 1.441098967557994,
+      "grad_norm": 0.5500359408388827,
+      "learning_rate": 5.559425130808802e-06,
+      "loss": 0.1093,
+      "step": 16680
+    },
+    {
+      "epoch": 1.4419629357639638,
+      "grad_norm": 0.5748371580710966,
+      "learning_rate": 5.554790287321752e-06,
+      "loss": 0.1095,
+      "step": 16690
+    },
+    {
+      "epoch": 1.442826903969934,
+      "grad_norm": 0.5589845183126114,
+      "learning_rate": 5.550154961126812e-06,
+      "loss": 0.11,
+      "step": 16700
+    },
+    {
+      "epoch": 1.4436908721759039,
+      "grad_norm": 0.5596557149578248,
+      "learning_rate": 5.5455191562570535e-06,
+      "loss": 0.1084,
+      "step": 16710
+    },
+    {
+      "epoch": 1.444554840381874,
+      "grad_norm": 0.555957437652107,
+      "learning_rate": 5.540882876745962e-06,
+      "loss": 0.1099,
+      "step": 16720
+    },
+    {
+      "epoch": 1.4454188085878439,
+      "grad_norm": 0.5697811632279,
+      "learning_rate": 5.53624612662744e-06,
+      "loss": 0.11,
+      "step": 16730
+    },
+    {
+      "epoch": 1.446282776793814,
+      "grad_norm": 0.540863041698352,
+      "learning_rate": 5.531608909935795e-06,
+      "loss": 0.1078,
+      "step": 16740
+    },
+    {
+      "epoch": 1.4471467449997841,
+      "grad_norm": 0.5439712322944413,
+      "learning_rate": 5.526971230705744e-06,
+      "loss": 0.1092,
+      "step": 16750
+    },
+    {
+      "epoch": 1.448010713205754,
+      "grad_norm": 0.5539653073070816,
+      "learning_rate": 5.522333092972406e-06,
+      "loss": 0.1067,
+      "step": 16760
+    },
+    {
+      "epoch": 1.448874681411724,
+      "grad_norm": 0.5757732852795685,
+      "learning_rate": 5.517694500771298e-06,
+      "loss": 0.1077,
+      "step": 16770
+    },
+    {
+      "epoch": 1.449738649617694,
+      "grad_norm": 0.5523898348981371,
+      "learning_rate": 5.513055458138329e-06,
+      "loss": 0.1098,
+      "step": 16780
+    },
+    {
+      "epoch": 1.4506026178236642,
+      "grad_norm": 0.53039159499405,
+      "learning_rate": 5.508415969109808e-06,
+      "loss": 0.1091,
+      "step": 16790
+    },
+    {
+      "epoch": 1.451466586029634,
+      "grad_norm": 0.575717777450302,
+      "learning_rate": 5.503776037722425e-06,
+      "loss": 0.1115,
+      "step": 16800
+    },
+    {
+      "epoch": 1.4523305542356042,
+      "grad_norm": 0.5295410373537287,
+      "learning_rate": 5.4991356680132576e-06,
+      "loss": 0.1106,
+      "step": 16810
+    },
+    {
+      "epoch": 1.453194522441574,
+      "grad_norm": 0.5637895554867309,
+      "learning_rate": 5.494494864019767e-06,
+      "loss": 0.1106,
+      "step": 16820
+    },
+    {
+      "epoch": 1.4540584906475442,
+      "grad_norm": 0.5647698345641413,
+      "learning_rate": 5.489853629779789e-06,
+      "loss": 0.1113,
+      "step": 16830
+    },
+    {
+      "epoch": 1.4549224588535141,
+      "grad_norm": 0.5480780997759342,
+      "learning_rate": 5.485211969331536e-06,
+      "loss": 0.1104,
+      "step": 16840
+    },
+    {
+      "epoch": 1.4557864270594842,
+      "grad_norm": 0.5962694168716901,
+      "learning_rate": 5.480569886713586e-06,
+      "loss": 0.1083,
+      "step": 16850
+    },
+    {
+      "epoch": 1.4566503952654544,
+      "grad_norm": 0.5468143792266299,
+      "learning_rate": 5.475927385964892e-06,
+      "loss": 0.105,
+      "step": 16860
+    },
+    {
+      "epoch": 1.4575143634714243,
+      "grad_norm": 0.5627954667617832,
+      "learning_rate": 5.4712844711247685e-06,
+      "loss": 0.1097,
+      "step": 16870
+    },
+    {
+      "epoch": 1.4583783316773942,
+      "grad_norm": 0.5498247339749494,
+      "learning_rate": 5.4666411462328835e-06,
+      "loss": 0.1073,
+      "step": 16880
+    },
+    {
+      "epoch": 1.4592422998833643,
+      "grad_norm": 0.5593645586338002,
+      "learning_rate": 5.461997415329274e-06,
+      "loss": 0.1099,
+      "step": 16890
+    },
+    {
+      "epoch": 1.4601062680893344,
+      "grad_norm": 0.5707187551887513,
+      "learning_rate": 5.457353282454318e-06,
+      "loss": 0.1121,
+      "step": 16900
+    },
+    {
+      "epoch": 1.4609702362953043,
+      "grad_norm": 0.553904006667131,
+      "learning_rate": 5.452708751648753e-06,
+      "loss": 0.1072,
+      "step": 16910
+    },
+    {
+      "epoch": 1.4618342045012742,
+      "grad_norm": 0.529572172260642,
+      "learning_rate": 5.448063826953654e-06,
+      "loss": 0.1091,
+      "step": 16920
+    },
+    {
+      "epoch": 1.4626981727072443,
+      "grad_norm": 0.5472381570541359,
+      "learning_rate": 5.443418512410449e-06,
+      "loss": 0.1072,
+      "step": 16930
+    },
+    {
+      "epoch": 1.4635621409132145,
+      "grad_norm": 0.5704255859503703,
+      "learning_rate": 5.438772812060895e-06,
+      "loss": 0.1099,
+      "step": 16940
+    },
+    {
+      "epoch": 1.4644261091191844,
+      "grad_norm": 0.5661130823236894,
+      "learning_rate": 5.434126729947091e-06,
+      "loss": 0.1067,
+      "step": 16950
+    },
+    {
+      "epoch": 1.4652900773251545,
+      "grad_norm": 0.5824585865876896,
+      "learning_rate": 5.4294802701114645e-06,
+      "loss": 0.1075,
+      "step": 16960
+    },
+    {
+      "epoch": 1.4661540455311244,
+      "grad_norm": 0.5712779312011059,
+      "learning_rate": 5.424833436596774e-06,
+      "loss": 0.1106,
+      "step": 16970
+    },
+    {
+      "epoch": 1.4670180137370945,
+      "grad_norm": 0.556641188854497,
+      "learning_rate": 5.420186233446104e-06,
+      "loss": 0.1087,
+      "step": 16980
+    },
+    {
+      "epoch": 1.4678819819430644,
+      "grad_norm": 0.5747662022205271,
+      "learning_rate": 5.415538664702858e-06,
+      "loss": 0.1073,
+      "step": 16990
+    },
+    {
+      "epoch": 1.4687459501490345,
+      "grad_norm": 0.5379635353303245,
+      "learning_rate": 5.410890734410761e-06,
+      "loss": 0.1053,
+      "step": 17000
+    },
+    {
+      "epoch": 1.4696099183550047,
+      "grad_norm": 0.6016166805731681,
+      "learning_rate": 5.406242446613845e-06,
+      "loss": 0.1103,
+      "step": 17010
+    },
+    {
+      "epoch": 1.4704738865609746,
+      "grad_norm": 0.5651647377002924,
+      "learning_rate": 5.401593805356464e-06,
+      "loss": 0.1066,
+      "step": 17020
+    },
+    {
+      "epoch": 1.4713378547669445,
+      "grad_norm": 0.53397079179655,
+      "learning_rate": 5.39694481468327e-06,
+      "loss": 0.1108,
+      "step": 17030
+    },
+    {
+      "epoch": 1.4722018229729146,
+      "grad_norm": 0.6010434080095977,
+      "learning_rate": 5.392295478639226e-06,
+      "loss": 0.1098,
+      "step": 17040
+    },
+    {
+      "epoch": 1.4730657911788847,
+      "grad_norm": 0.5546090791053712,
+      "learning_rate": 5.38764580126959e-06,
+      "loss": 0.1114,
+      "step": 17050
+    },
+    {
+      "epoch": 1.4739297593848546,
+      "grad_norm": 0.5273120526879177,
+      "learning_rate": 5.382995786619921e-06,
+      "loss": 0.1088,
+      "step": 17060
+    },
+    {
+      "epoch": 1.4747937275908247,
+      "grad_norm": 0.5567574055473559,
+      "learning_rate": 5.378345438736068e-06,
+      "loss": 0.1085,
+      "step": 17070
+    },
+    {
+      "epoch": 1.4756576957967946,
+      "grad_norm": 0.5714232269488697,
+      "learning_rate": 5.373694761664174e-06,
+      "loss": 0.1074,
+      "step": 17080
+    },
+    {
+      "epoch": 1.4765216640027647,
+      "grad_norm": 0.5534188993303146,
+      "learning_rate": 5.369043759450664e-06,
+      "loss": 0.1112,
+      "step": 17090
+    },
+    {
+      "epoch": 1.4773856322087346,
+      "grad_norm": 0.5533405667659507,
+      "learning_rate": 5.364392436142251e-06,
+      "loss": 0.1084,
+      "step": 17100
+    },
+    {
+      "epoch": 1.4782496004147048,
+      "grad_norm": 0.5715199286886745,
+      "learning_rate": 5.359740795785921e-06,
+      "loss": 0.1088,
+      "step": 17110
+    },
+    {
+      "epoch": 1.479113568620675,
+      "grad_norm": 0.57254667157946,
+      "learning_rate": 5.3550888424289415e-06,
+      "loss": 0.1056,
+      "step": 17120
+    },
+    {
+      "epoch": 1.4799775368266448,
+      "grad_norm": 0.5199024237592625,
+      "learning_rate": 5.3504365801188495e-06,
+      "loss": 0.1052,
+      "step": 17130
+    },
+    {
+      "epoch": 1.4808415050326147,
+      "grad_norm": 0.5813582218704558,
+      "learning_rate": 5.3457840129034535e-06,
+      "loss": 0.1047,
+      "step": 17140
+    },
+    {
+      "epoch": 1.4817054732385848,
+      "grad_norm": 0.5390351359241751,
+      "learning_rate": 5.341131144830822e-06,
+      "loss": 0.1068,
+      "step": 17150
+    },
+    {
+      "epoch": 1.482569441444555,
+      "grad_norm": 0.5829892827363654,
+      "learning_rate": 5.336477979949291e-06,
+      "loss": 0.1063,
+      "step": 17160
+    },
+    {
+      "epoch": 1.4834334096505248,
+      "grad_norm": 0.5613668918351362,
+      "learning_rate": 5.33182452230745e-06,
+      "loss": 0.1093,
+      "step": 17170
+    },
+    {
+      "epoch": 1.484297377856495,
+      "grad_norm": 0.5668262800654991,
+      "learning_rate": 5.327170775954149e-06,
+      "loss": 0.1067,
+      "step": 17180
+    },
+    {
+      "epoch": 1.4851613460624649,
+      "grad_norm": 0.563123194809761,
+      "learning_rate": 5.322516744938482e-06,
+      "loss": 0.1068,
+      "step": 17190
+    },
+    {
+      "epoch": 1.486025314268435,
+      "grad_norm": 0.5702164031662597,
+      "learning_rate": 5.317862433309797e-06,
+      "loss": 0.1123,
+      "step": 17200
+    },
+    {
+      "epoch": 1.4868892824744049,
+      "grad_norm": 0.5573428134086994,
+      "learning_rate": 5.3132078451176815e-06,
+      "loss": 0.1092,
+      "step": 17210
+    },
+    {
+      "epoch": 1.487753250680375,
+      "grad_norm": 0.5423066018915117,
+      "learning_rate": 5.308552984411968e-06,
+      "loss": 0.1065,
+      "step": 17220
+    },
+    {
+      "epoch": 1.488617218886345,
+      "grad_norm": 0.5895533833702804,
+      "learning_rate": 5.30389785524272e-06,
+      "loss": 0.1121,
+      "step": 17230
+    },
+    {
+      "epoch": 1.489481187092315,
+      "grad_norm": 0.5609587275234417,
+      "learning_rate": 5.299242461660243e-06,
+      "loss": 0.1065,
+      "step": 17240
+    },
+    {
+      "epoch": 1.490345155298285,
+      "grad_norm": 0.575671022001389,
+      "learning_rate": 5.294586807715066e-06,
+      "loss": 0.1087,
+      "step": 17250
+    },
+    {
+      "epoch": 1.491209123504255,
+      "grad_norm": 0.5345702719825773,
+      "learning_rate": 5.289930897457946e-06,
+      "loss": 0.1077,
+      "step": 17260
+    },
+    {
+      "epoch": 1.4920730917102252,
+      "grad_norm": 0.5371496978514604,
+      "learning_rate": 5.285274734939864e-06,
+      "loss": 0.1103,
+      "step": 17270
+    },
+    {
+      "epoch": 1.492937059916195,
+      "grad_norm": 0.5593334573221135,
+      "learning_rate": 5.280618324212018e-06,
+      "loss": 0.1074,
+      "step": 17280
+    },
+    {
+      "epoch": 1.493801028122165,
+      "grad_norm": 0.5295295244997914,
+      "learning_rate": 5.275961669325828e-06,
+      "loss": 0.1063,
+      "step": 17290
+    },
+    {
+      "epoch": 1.494664996328135,
+      "grad_norm": 0.5383691997834555,
+      "learning_rate": 5.271304774332917e-06,
+      "loss": 0.1089,
+      "step": 17300
+    },
+    {
+      "epoch": 1.4955289645341052,
+      "grad_norm": 0.5378075832729008,
+      "learning_rate": 5.266647643285126e-06,
+      "loss": 0.1071,
+      "step": 17310
+    },
+    {
+      "epoch": 1.4963929327400751,
+      "grad_norm": 0.5864711882360478,
+      "learning_rate": 5.261990280234498e-06,
+      "loss": 0.1082,
+      "step": 17320
+    },
+    {
+      "epoch": 1.4972569009460452,
+      "grad_norm": 0.5833725727517236,
+      "learning_rate": 5.257332689233275e-06,
+      "loss": 0.109,
+      "step": 17330
+    },
+    {
+      "epoch": 1.4981208691520151,
+      "grad_norm": 0.617406542429736,
+      "learning_rate": 5.252674874333902e-06,
+      "loss": 0.1103,
+      "step": 17340
+    },
+    {
+      "epoch": 1.4989848373579853,
+      "grad_norm": 0.5393750705119263,
+      "learning_rate": 5.248016839589015e-06,
+      "loss": 0.1065,
+      "step": 17350
+    },
+    {
+      "epoch": 1.4998488055639552,
+      "grad_norm": 0.5525154054728131,
+      "learning_rate": 5.243358589051445e-06,
+      "loss": 0.1086,
+      "step": 17360
+    },
+    {
+      "epoch": 1.5007127737699253,
+      "grad_norm": 0.5663281890236571,
+      "learning_rate": 5.238700126774208e-06,
+      "loss": 0.1106,
+      "step": 17370
+    },
+    {
+      "epoch": 1.5015767419758954,
+      "grad_norm": 0.5639045716521068,
+      "learning_rate": 5.234041456810501e-06,
+      "loss": 0.1099,
+      "step": 17380
+    },
+    {
+      "epoch": 1.5024407101818653,
+      "grad_norm": 0.5501571786531163,
+      "learning_rate": 5.229382583213713e-06,
+      "loss": 0.1084,
+      "step": 17390
+    },
+    {
+      "epoch": 1.5033046783878352,
+      "grad_norm": 0.5555533506524184,
+      "learning_rate": 5.224723510037395e-06,
+      "loss": 0.1096,
+      "step": 17400
+    },
+    {
+      "epoch": 1.5041686465938053,
+      "grad_norm": 0.5565015152184756,
+      "learning_rate": 5.220064241335286e-06,
+      "loss": 0.1115,
+      "step": 17410
+    },
+    {
+      "epoch": 1.5050326147997755,
+      "grad_norm": 0.5594405883971818,
+      "learning_rate": 5.215404781161283e-06,
+      "loss": 0.1086,
+      "step": 17420
+    },
+    {
+      "epoch": 1.5058965830057454,
+      "grad_norm": 0.560011821779313,
+      "learning_rate": 5.2107451335694595e-06,
+      "loss": 0.1075,
+      "step": 17430
+    },
+    {
+      "epoch": 1.5067605512117153,
+      "grad_norm": 0.543119102461241,
+      "learning_rate": 5.206085302614045e-06,
+      "loss": 0.1071,
+      "step": 17440
+    },
+    {
+      "epoch": 1.5076245194176854,
+      "grad_norm": 0.5357688437385529,
+      "learning_rate": 5.201425292349434e-06,
+      "loss": 0.1114,
+      "step": 17450
+    },
+    {
+      "epoch": 1.5084884876236555,
+      "grad_norm": 0.5504332290569045,
+      "learning_rate": 5.196765106830171e-06,
+      "loss": 0.1096,
+      "step": 17460
+    },
+    {
+      "epoch": 1.5093524558296254,
+      "grad_norm": 0.567852538652432,
+      "learning_rate": 5.192104750110959e-06,
+      "loss": 0.1074,
+      "step": 17470
+    },
+    {
+      "epoch": 1.5102164240355955,
+      "grad_norm": 0.5322729979849891,
+      "learning_rate": 5.187444226246645e-06,
+      "loss": 0.1099,
+      "step": 17480
+    },
+    {
+      "epoch": 1.5110803922415657,
+      "grad_norm": 0.5756617595017037,
+      "learning_rate": 5.182783539292226e-06,
+      "loss": 0.1086,
+      "step": 17490
+    },
+    {
+      "epoch": 1.5119443604475356,
+      "grad_norm": 0.5678975606126713,
+      "learning_rate": 5.178122693302835e-06,
+      "loss": 0.1057,
+      "step": 17500
+    },
+    {
+      "epoch": 1.5128083286535055,
+      "grad_norm": 0.5461058418335905,
+      "learning_rate": 5.173461692333751e-06,
+      "loss": 0.1068,
+      "step": 17510
+    },
+    {
+      "epoch": 1.5136722968594756,
+      "grad_norm": 0.571376612212014,
+      "learning_rate": 5.168800540440381e-06,
+      "loss": 0.1075,
+      "step": 17520
+    },
+    {
+      "epoch": 1.5145362650654457,
+      "grad_norm": 0.5497908782485466,
+      "learning_rate": 5.164139241678267e-06,
+      "loss": 0.1072,
+      "step": 17530
+    },
+    {
+      "epoch": 1.5154002332714156,
+      "grad_norm": 0.5914438759722356,
+      "learning_rate": 5.159477800103079e-06,
+      "loss": 0.1085,
+      "step": 17540
+    },
+    {
+      "epoch": 1.5162642014773855,
+      "grad_norm": 0.5316510326275006,
+      "learning_rate": 5.1548162197706075e-06,
+      "loss": 0.1075,
+      "step": 17550
+    },
+    {
+      "epoch": 1.5171281696833556,
+      "grad_norm": 0.5621679449680816,
+      "learning_rate": 5.150154504736768e-06,
+      "loss": 0.1071,
+      "step": 17560
+    },
+    {
+      "epoch": 1.5179921378893257,
+      "grad_norm": 0.5372659772513624,
+      "learning_rate": 5.145492659057593e-06,
+      "loss": 0.1061,
+      "step": 17570
+    },
+    {
+      "epoch": 1.5188561060952956,
+      "grad_norm": 0.5710203008791778,
+      "learning_rate": 5.140830686789224e-06,
+      "loss": 0.1069,
+      "step": 17580
+    },
+    {
+      "epoch": 1.5197200743012658,
+      "grad_norm": 0.5481229869659677,
+      "learning_rate": 5.1361685919879175e-06,
+      "loss": 0.1087,
+      "step": 17590
+    },
+    {
+      "epoch": 1.520584042507236,
+      "grad_norm": 0.5589983135098491,
+      "learning_rate": 5.131506378710035e-06,
+      "loss": 0.1109,
+      "step": 17600
+    },
+    {
+      "epoch": 1.5214480107132058,
+      "grad_norm": 0.5635199037596682,
+      "learning_rate": 5.126844051012041e-06,
+      "loss": 0.1103,
+      "step": 17610
+    },
+    {
+      "epoch": 1.5223119789191757,
+      "grad_norm": 0.5500151319574998,
+      "learning_rate": 5.122181612950499e-06,
+      "loss": 0.108,
+      "step": 17620
+    },
+    {
+      "epoch": 1.5231759471251458,
+      "grad_norm": 0.5371844113385824,
+      "learning_rate": 5.11751906858207e-06,
+      "loss": 0.1072,
+      "step": 17630
+    },
+    {
+      "epoch": 1.524039915331116,
+      "grad_norm": 0.5485761580061022,
+      "learning_rate": 5.112856421963507e-06,
+      "loss": 0.1072,
+      "step": 17640
+    },
+    {
+      "epoch": 1.5249038835370858,
+      "grad_norm": 0.5700920471032753,
+      "learning_rate": 5.108193677151648e-06,
+      "loss": 0.1085,
+      "step": 17650
+    },
+    {
+      "epoch": 1.5257678517430557,
+      "grad_norm": 0.5514566079923485,
+      "learning_rate": 5.103530838203427e-06,
+      "loss": 0.1071,
+      "step": 17660
+    },
+    {
+      "epoch": 1.5266318199490259,
+      "grad_norm": 0.5712587946627912,
+      "learning_rate": 5.0988679091758465e-06,
+      "loss": 0.1057,
+      "step": 17670
+    },
+    {
+      "epoch": 1.527495788154996,
+      "grad_norm": 0.523278063242322,
+      "learning_rate": 5.094204894125997e-06,
+      "loss": 0.1055,
+      "step": 17680
+    },
+    {
+      "epoch": 1.5283597563609659,
+      "grad_norm": 0.547975825208195,
+      "learning_rate": 5.0895417971110384e-06,
+      "loss": 0.1069,
+      "step": 17690
+    },
+    {
+      "epoch": 1.5292237245669358,
+      "grad_norm": 0.5649900655239077,
+      "learning_rate": 5.0848786221882065e-06,
+      "loss": 0.1071,
+      "step": 17700
+    },
+    {
+      "epoch": 1.530087692772906,
+      "grad_norm": 0.5235846263354882,
+      "learning_rate": 5.0802153734148e-06,
+      "loss": 0.1086,
+      "step": 17710
+    },
+    {
+      "epoch": 1.530951660978876,
+      "grad_norm": 0.558486058207212,
+      "learning_rate": 5.075552054848188e-06,
+      "loss": 0.108,
+      "step": 17720
+    },
+    {
+      "epoch": 1.531815629184846,
+      "grad_norm": 0.5591344019040175,
+      "learning_rate": 5.070888670545794e-06,
+      "loss": 0.1078,
+      "step": 17730
+    },
+    {
+      "epoch": 1.532679597390816,
+      "grad_norm": 0.5860722333653754,
+      "learning_rate": 5.066225224565102e-06,
+      "loss": 0.1085,
+      "step": 17740
+    },
+    {
+      "epoch": 1.5335435655967862,
+      "grad_norm": 0.5963012918305679,
+      "learning_rate": 5.061561720963649e-06,
+      "loss": 0.1098,
+      "step": 17750
+    },
+    {
+      "epoch": 1.534407533802756,
+      "grad_norm": 0.5604735839188759,
+      "learning_rate": 5.056898163799023e-06,
+      "loss": 0.1081,
+      "step": 17760
+    },
+    {
+      "epoch": 1.535271502008726,
+      "grad_norm": 0.5521582524965631,
+      "learning_rate": 5.052234557128859e-06,
+      "loss": 0.1056,
+      "step": 17770
+    },
+    {
+      "epoch": 1.536135470214696,
+      "grad_norm": 0.5863548246000648,
+      "learning_rate": 5.04757090501083e-06,
+      "loss": 0.1073,
+      "step": 17780
+    },
+    {
+      "epoch": 1.5369994384206662,
+      "grad_norm": 0.5505531729226412,
+      "learning_rate": 5.042907211502654e-06,
+      "loss": 0.1059,
+      "step": 17790
+    },
+    {
+      "epoch": 1.5378634066266361,
+      "grad_norm": 0.5527561782797406,
+      "learning_rate": 5.038243480662086e-06,
+      "loss": 0.107,
+      "step": 17800
+    },
+    {
+      "epoch": 1.538727374832606,
+      "grad_norm": 0.5403035943659682,
+      "learning_rate": 5.033579716546908e-06,
+      "loss": 0.1066,
+      "step": 17810
+    },
+    {
+      "epoch": 1.5395913430385761,
+      "grad_norm": 0.6183154805975839,
+      "learning_rate": 5.028915923214935e-06,
+      "loss": 0.1063,
+      "step": 17820
+    },
+    {
+      "epoch": 1.5404553112445463,
+      "grad_norm": 0.5501718905367079,
+      "learning_rate": 5.0242521047240076e-06,
+      "loss": 0.1036,
+      "step": 17830
+    },
+    {
+      "epoch": 1.5413192794505162,
+      "grad_norm": 0.5332692223400205,
+      "learning_rate": 5.019588265131984e-06,
+      "loss": 0.1061,
+      "step": 17840
+    },
+    {
+      "epoch": 1.5421832476564863,
+      "grad_norm": 0.5613621146761566,
+      "learning_rate": 5.014924408496746e-06,
+      "loss": 0.1068,
+      "step": 17850
+    },
+    {
+      "epoch": 1.5430472158624564,
+      "grad_norm": 0.5581157336544168,
+      "learning_rate": 5.010260538876187e-06,
+      "loss": 0.1075,
+      "step": 17860
+    },
+    {
+      "epoch": 1.5439111840684263,
+      "grad_norm": 0.559150234766518,
+      "learning_rate": 5.0055966603282136e-06,
+      "loss": 0.1046,
+      "step": 17870
+    },
+    {
+      "epoch": 1.5447751522743962,
+      "grad_norm": 0.5523464636226372,
+      "learning_rate": 5.000932776910739e-06,
+      "loss": 0.1081,
+      "step": 17880
+    },
+    {
+      "epoch": 1.5456391204803663,
+      "grad_norm": 0.5874851979850498,
+      "learning_rate": 4.996268892681681e-06,
+      "loss": 0.1059,
+      "step": 17890
+    },
+    {
+      "epoch": 1.5465030886863365,
+      "grad_norm": 0.5599016847752923,
+      "learning_rate": 4.991605011698957e-06,
+      "loss": 0.105,
+      "step": 17900
+    },
+    {
+      "epoch": 1.5473670568923064,
+      "grad_norm": 0.5536885812459234,
+      "learning_rate": 4.9869411380204825e-06,
+      "loss": 0.1114,
+      "step": 17910
+    },
+    {
+      "epoch": 1.5482310250982763,
+      "grad_norm": 0.5850370040739463,
+      "learning_rate": 4.982277275704169e-06,
+      "loss": 0.108,
+      "step": 17920
+    },
+    {
+      "epoch": 1.5490949933042464,
+      "grad_norm": 0.5476002826516605,
+      "learning_rate": 4.977613428807913e-06,
+      "loss": 0.1114,
+      "step": 17930
+    },
+    {
+      "epoch": 1.5499589615102165,
+      "grad_norm": 0.5458827789684254,
+      "learning_rate": 4.972949601389601e-06,
+      "loss": 0.1055,
+      "step": 17940
+    },
+    {
+      "epoch": 1.5508229297161864,
+      "grad_norm": 0.5413640872702045,
+      "learning_rate": 4.968285797507104e-06,
+      "loss": 0.1113,
+      "step": 17950
+    },
+    {
+      "epoch": 1.5516868979221563,
+      "grad_norm": 0.5752827421319834,
+      "learning_rate": 4.963622021218271e-06,
+      "loss": 0.1087,
+      "step": 17960
+    },
+    {
+      "epoch": 1.5525508661281266,
+      "grad_norm": 0.5860674587550517,
+      "learning_rate": 4.958958276580922e-06,
+      "loss": 0.1097,
+      "step": 17970
+    },
+    {
+      "epoch": 1.5534148343340966,
+      "grad_norm": 0.5444099654254021,
+      "learning_rate": 4.954294567652858e-06,
+      "loss": 0.1082,
+      "step": 17980
+    },
+    {
+      "epoch": 1.5542788025400665,
+      "grad_norm": 0.5578751187043065,
+      "learning_rate": 4.949630898491842e-06,
+      "loss": 0.1085,
+      "step": 17990
+    },
+    {
+      "epoch": 1.5551427707460366,
+      "grad_norm": 0.5607469002061576,
+      "learning_rate": 4.9449672731556095e-06,
+      "loss": 0.1103,
+      "step": 18000
+    },
+    {
+      "epoch": 1.5560067389520067,
+      "grad_norm": 0.597520663614332,
+      "learning_rate": 4.940303695701848e-06,
+      "loss": 0.1085,
+      "step": 18010
+    },
+    {
+      "epoch": 1.5568707071579766,
+      "grad_norm": 0.5635599902531142,
+      "learning_rate": 4.935640170188212e-06,
+      "loss": 0.1044,
+      "step": 18020
+    },
+    {
+      "epoch": 1.5577346753639465,
+      "grad_norm": 0.5477678866631618,
+      "learning_rate": 4.930976700672308e-06,
+      "loss": 0.1067,
+      "step": 18030
+    },
+    {
+      "epoch": 1.5585986435699166,
+      "grad_norm": 0.5633674903097964,
+      "learning_rate": 4.926313291211695e-06,
+      "loss": 0.1074,
+      "step": 18040
+    },
+    {
+      "epoch": 1.5594626117758867,
+      "grad_norm": 0.5457039922820861,
+      "learning_rate": 4.921649945863875e-06,
+      "loss": 0.1072,
+      "step": 18050
+    },
+    {
+      "epoch": 1.5603265799818566,
+      "grad_norm": 0.5529102573041352,
+      "learning_rate": 4.916986668686299e-06,
+      "loss": 0.1028,
+      "step": 18060
+    },
+    {
+      "epoch": 1.5611905481878265,
+      "grad_norm": 0.5526616247690039,
+      "learning_rate": 4.912323463736358e-06,
+      "loss": 0.1067,
+      "step": 18070
+    },
+    {
+      "epoch": 1.5620545163937967,
+      "grad_norm": 0.5906101164405569,
+      "learning_rate": 4.907660335071379e-06,
+      "loss": 0.107,
+      "step": 18080
+    },
+    {
+      "epoch": 1.5629184845997668,
+      "grad_norm": 0.5802445703142322,
+      "learning_rate": 4.902997286748623e-06,
+      "loss": 0.1073,
+      "step": 18090
+    },
+    {
+      "epoch": 1.5637824528057367,
+      "grad_norm": 0.5497703406075652,
+      "learning_rate": 4.898334322825279e-06,
+      "loss": 0.1045,
+      "step": 18100
+    },
+    {
+      "epoch": 1.5646464210117068,
+      "grad_norm": 0.5638475765108427,
+      "learning_rate": 4.893671447358469e-06,
+      "loss": 0.1069,
+      "step": 18110
+    },
+    {
+      "epoch": 1.565510389217677,
+      "grad_norm": 0.5474191633068053,
+      "learning_rate": 4.889008664405229e-06,
+      "loss": 0.1067,
+      "step": 18120
+    },
+    {
+      "epoch": 1.5663743574236468,
+      "grad_norm": 0.5726552269200086,
+      "learning_rate": 4.8843459780225214e-06,
+      "loss": 0.105,
+      "step": 18130
+    },
+    {
+      "epoch": 1.5672383256296167,
+      "grad_norm": 0.57494985885623,
+      "learning_rate": 4.879683392267223e-06,
+      "loss": 0.108,
+      "step": 18140
+    },
+    {
+      "epoch": 1.5681022938355869,
+      "grad_norm": 0.5758913701642095,
+      "learning_rate": 4.875020911196123e-06,
+      "loss": 0.1071,
+      "step": 18150
+    },
+    {
+      "epoch": 1.568966262041557,
+      "grad_norm": 0.5750957605006866,
+      "learning_rate": 4.870358538865916e-06,
+      "loss": 0.1085,
+      "step": 18160
+    },
+    {
+      "epoch": 1.5698302302475269,
+      "grad_norm": 0.5600079931570073,
+      "learning_rate": 4.865696279333207e-06,
+      "loss": 0.107,
+      "step": 18170
+    },
+    {
+      "epoch": 1.5706941984534968,
+      "grad_norm": 0.5617150407874216,
+      "learning_rate": 4.861034136654501e-06,
+      "loss": 0.1086,
+      "step": 18180
+    },
+    {
+      "epoch": 1.571558166659467,
+      "grad_norm": 0.5759016246754355,
+      "learning_rate": 4.8563721148862015e-06,
+      "loss": 0.1034,
+      "step": 18190
+    },
+    {
+      "epoch": 1.572422134865437,
+      "grad_norm": 0.5381126057860631,
+      "learning_rate": 4.851710218084605e-06,
+      "loss": 0.1057,
+      "step": 18200
+    },
+    {
+      "epoch": 1.573286103071407,
+      "grad_norm": 0.577124912521626,
+      "learning_rate": 4.8470484503059e-06,
+      "loss": 0.1053,
+      "step": 18210
+    },
+    {
+      "epoch": 1.5741500712773768,
+      "grad_norm": 0.5509976895411456,
+      "learning_rate": 4.8423868156061644e-06,
+      "loss": 0.1045,
+      "step": 18220
+    },
+    {
+      "epoch": 1.5750140394833472,
+      "grad_norm": 0.5820369404762893,
+      "learning_rate": 4.837725318041362e-06,
+      "loss": 0.1089,
+      "step": 18230
+    },
+    {
+      "epoch": 1.575878007689317,
+      "grad_norm": 0.55949728961922,
+      "learning_rate": 4.833063961667331e-06,
+      "loss": 0.1073,
+      "step": 18240
+    },
+    {
+      "epoch": 1.576741975895287,
+      "grad_norm": 0.5710029043616441,
+      "learning_rate": 4.828402750539791e-06,
+      "loss": 0.1062,
+      "step": 18250
+    },
+    {
+      "epoch": 1.577605944101257,
+      "grad_norm": 0.5382770988006425,
+      "learning_rate": 4.823741688714335e-06,
+      "loss": 0.1075,
+      "step": 18260
+    },
+    {
+      "epoch": 1.5784699123072272,
+      "grad_norm": 0.5658950513457499,
+      "learning_rate": 4.8190807802464285e-06,
+      "loss": 0.1083,
+      "step": 18270
+    },
+    {
+      "epoch": 1.5793338805131971,
+      "grad_norm": 0.5364382299197196,
+      "learning_rate": 4.814420029191395e-06,
+      "loss": 0.106,
+      "step": 18280
+    },
+    {
+      "epoch": 1.580197848719167,
+      "grad_norm": 0.5524618471037588,
+      "learning_rate": 4.8097594396044326e-06,
+      "loss": 0.106,
+      "step": 18290
+    },
+    {
+      "epoch": 1.5810618169251371,
+      "grad_norm": 0.5628924598931844,
+      "learning_rate": 4.8050990155405894e-06,
+      "loss": 0.1069,
+      "step": 18300
+    },
+    {
+      "epoch": 1.5819257851311073,
+      "grad_norm": 0.5420207388695794,
+      "learning_rate": 4.800438761054775e-06,
+      "loss": 0.1069,
+      "step": 18310
+    },
+    {
+      "epoch": 1.5827897533370772,
+      "grad_norm": 0.5910637255202297,
+      "learning_rate": 4.7957786802017485e-06,
+      "loss": 0.105,
+      "step": 18320
+    },
+    {
+      "epoch": 1.583653721543047,
+      "grad_norm": 0.5691128173632788,
+      "learning_rate": 4.791118777036119e-06,
+      "loss": 0.1067,
+      "step": 18330
+    },
+    {
+      "epoch": 1.5845176897490172,
+      "grad_norm": 0.5519021523237372,
+      "learning_rate": 4.786459055612341e-06,
+      "loss": 0.1077,
+      "step": 18340
+    },
+    {
+      "epoch": 1.5853816579549873,
+      "grad_norm": 0.5528173834315456,
+      "learning_rate": 4.781799519984715e-06,
+      "loss": 0.1051,
+      "step": 18350
+    },
+    {
+      "epoch": 1.5862456261609572,
+      "grad_norm": 0.5221856325986058,
+      "learning_rate": 4.7771401742073696e-06,
+      "loss": 0.1053,
+      "step": 18360
+    },
+    {
+      "epoch": 1.5871095943669273,
+      "grad_norm": 0.5619510960078813,
+      "learning_rate": 4.772481022334278e-06,
+      "loss": 0.1062,
+      "step": 18370
+    },
+    {
+      "epoch": 1.5879735625728975,
+      "grad_norm": 0.5186024822105305,
+      "learning_rate": 4.76782206841924e-06,
+      "loss": 0.1062,
+      "step": 18380
+    },
+    {
+      "epoch": 1.5888375307788674,
+      "grad_norm": 0.5525545768358348,
+      "learning_rate": 4.763163316515886e-06,
+      "loss": 0.1069,
+      "step": 18390
+    },
+    {
+      "epoch": 1.5897014989848373,
+      "grad_norm": 0.5551666634710379,
+      "learning_rate": 4.7585047706776674e-06,
+      "loss": 0.1064,
+      "step": 18400
+    },
+    {
+      "epoch": 1.5905654671908074,
+      "grad_norm": 0.5738819938747015,
+      "learning_rate": 4.753846434957856e-06,
+      "loss": 0.1084,
+      "step": 18410
+    },
+    {
+      "epoch": 1.5914294353967775,
+      "grad_norm": 0.5471225790659157,
+      "learning_rate": 4.749188313409548e-06,
+      "loss": 0.104,
+      "step": 18420
+    },
+    {
+      "epoch": 1.5922934036027474,
+      "grad_norm": 0.5619829238273804,
+      "learning_rate": 4.744530410085641e-06,
+      "loss": 0.1055,
+      "step": 18430
+    },
+    {
+      "epoch": 1.5931573718087173,
+      "grad_norm": 0.5637950090825805,
+      "learning_rate": 4.739872729038853e-06,
+      "loss": 0.1086,
+      "step": 18440
+    },
+    {
+      "epoch": 1.5940213400146874,
+      "grad_norm": 0.5903713369226199,
+      "learning_rate": 4.735215274321706e-06,
+      "loss": 0.1063,
+      "step": 18450
+    },
+    {
+      "epoch": 1.5948853082206575,
+      "grad_norm": 0.5685279528763301,
+      "learning_rate": 4.7305580499865256e-06,
+      "loss": 0.1077,
+      "step": 18460
+    },
+    {
+      "epoch": 1.5957492764266274,
+      "grad_norm": 0.5384106682889425,
+      "learning_rate": 4.725901060085431e-06,
+      "loss": 0.1061,
+      "step": 18470
+    },
+    {
+      "epoch": 1.5966132446325976,
+      "grad_norm": 0.5777818816201741,
+      "learning_rate": 4.721244308670344e-06,
+      "loss": 0.1084,
+      "step": 18480
+    },
+    {
+      "epoch": 1.5974772128385677,
+      "grad_norm": 0.5523159068046266,
+      "learning_rate": 4.71658779979298e-06,
+      "loss": 0.1057,
+      "step": 18490
+    },
+    {
+      "epoch": 1.5983411810445376,
+      "grad_norm": 0.5948813436945792,
+      "learning_rate": 4.711931537504836e-06,
+      "loss": 0.1085,
+      "step": 18500
+    },
+    {
+      "epoch": 1.5992051492505075,
+      "grad_norm": 0.5407671177901976,
+      "learning_rate": 4.7072755258572014e-06,
+      "loss": 0.1044,
+      "step": 18510
+    },
+    {
+      "epoch": 1.6000691174564776,
+      "grad_norm": 0.5731745458031607,
+      "learning_rate": 4.7026197689011425e-06,
+      "loss": 0.1058,
+      "step": 18520
+    },
+    {
+      "epoch": 1.6009330856624477,
+      "grad_norm": 0.5522669562395803,
+      "learning_rate": 4.697964270687507e-06,
+      "loss": 0.1051,
+      "step": 18530
+    },
+    {
+      "epoch": 1.6017970538684176,
+      "grad_norm": 0.5441669651971076,
+      "learning_rate": 4.693309035266921e-06,
+      "loss": 0.1058,
+      "step": 18540
+    },
+    {
+      "epoch": 1.6026610220743875,
+      "grad_norm": 0.5659568723921515,
+      "learning_rate": 4.6886540666897725e-06,
+      "loss": 0.1097,
+      "step": 18550
+    },
+    {
+      "epoch": 1.6035249902803577,
+      "grad_norm": 0.568976782651086,
+      "learning_rate": 4.683999369006223e-06,
+      "loss": 0.1057,
+      "step": 18560
+    },
+    {
+      "epoch": 1.6043889584863278,
+      "grad_norm": 0.5368332076086545,
+      "learning_rate": 4.679344946266202e-06,
+      "loss": 0.1071,
+      "step": 18570
+    },
+    {
+      "epoch": 1.6052529266922977,
+      "grad_norm": 0.5472796671016149,
+      "learning_rate": 4.674690802519394e-06,
+      "loss": 0.1059,
+      "step": 18580
+    },
+    {
+      "epoch": 1.6061168948982676,
+      "grad_norm": 0.5871536707812725,
+      "learning_rate": 4.67003694181524e-06,
+      "loss": 0.1065,
+      "step": 18590
+    },
+    {
+      "epoch": 1.6069808631042377,
+      "grad_norm": 0.5637887574564212,
+      "learning_rate": 4.665383368202939e-06,
+      "loss": 0.1044,
+      "step": 18600
+    },
+    {
+      "epoch": 1.6078448313102078,
+      "grad_norm": 0.5616396771556457,
+      "learning_rate": 4.660730085731438e-06,
+      "loss": 0.1054,
+      "step": 18610
+    },
+    {
+      "epoch": 1.6087087995161777,
+      "grad_norm": 0.5556472856134798,
+      "learning_rate": 4.656077098449434e-06,
+      "loss": 0.1032,
+      "step": 18620
+    },
+    {
+      "epoch": 1.6095727677221479,
+      "grad_norm": 0.5683986382033807,
+      "learning_rate": 4.65142441040536e-06,
+      "loss": 0.1072,
+      "step": 18630
+    },
+    {
+      "epoch": 1.610436735928118,
+      "grad_norm": 0.583395037447458,
+      "learning_rate": 4.646772025647394e-06,
+      "loss": 0.1051,
+      "step": 18640
+    },
+    {
+      "epoch": 1.6113007041340879,
+      "grad_norm": 0.5816202934421579,
+      "learning_rate": 4.642119948223449e-06,
+      "loss": 0.1072,
+      "step": 18650
+    },
+    {
+      "epoch": 1.6121646723400578,
+      "grad_norm": 0.5657989657195068,
+      "learning_rate": 4.6374681821811745e-06,
+      "loss": 0.1073,
+      "step": 18660
+    },
+    {
+      "epoch": 1.613028640546028,
+      "grad_norm": 0.5791187780948374,
+      "learning_rate": 4.6328167315679396e-06,
+      "loss": 0.1034,
+      "step": 18670
+    },
+    {
+      "epoch": 1.613892608751998,
+      "grad_norm": 0.5715678977702422,
+      "learning_rate": 4.6281656004308464e-06,
+      "loss": 0.1017,
+      "step": 18680
+    },
+    {
+      "epoch": 1.614756576957968,
+      "grad_norm": 0.5734621932193357,
+      "learning_rate": 4.62351479281672e-06,
+      "loss": 0.1064,
+      "step": 18690
+    },
+    {
+      "epoch": 1.6156205451639378,
+      "grad_norm": 0.5530026596429509,
+      "learning_rate": 4.618864312772098e-06,
+      "loss": 0.1079,
+      "step": 18700
+    },
+    {
+      "epoch": 1.616484513369908,
+      "grad_norm": 0.5601065046640386,
+      "learning_rate": 4.614214164343237e-06,
+      "loss": 0.1058,
+      "step": 18710
+    },
+    {
+      "epoch": 1.617348481575878,
+      "grad_norm": 0.5760129273781665,
+      "learning_rate": 4.609564351576103e-06,
+      "loss": 0.1059,
+      "step": 18720
+    },
+    {
+      "epoch": 1.618212449781848,
+      "grad_norm": 0.5558664942203434,
+      "learning_rate": 4.604914878516376e-06,
+      "loss": 0.1055,
+      "step": 18730
+    },
+    {
+      "epoch": 1.619076417987818,
+      "grad_norm": 0.5550058318089843,
+      "learning_rate": 4.60026574920943e-06,
+      "loss": 0.107,
+      "step": 18740
+    },
+    {
+      "epoch": 1.6199403861937882,
+      "grad_norm": 0.5546104276887956,
+      "learning_rate": 4.595616967700346e-06,
+      "loss": 0.1058,
+      "step": 18750
+    },
+    {
+      "epoch": 1.6208043543997581,
+      "grad_norm": 0.563362489721722,
+      "learning_rate": 4.590968538033906e-06,
+      "loss": 0.1028,
+      "step": 18760
+    },
+    {
+      "epoch": 1.621668322605728,
+      "grad_norm": 0.5664210910466141,
+      "learning_rate": 4.586320464254579e-06,
+      "loss": 0.1062,
+      "step": 18770
+    },
+    {
+      "epoch": 1.6225322908116981,
+      "grad_norm": 0.5604114011064872,
+      "learning_rate": 4.581672750406527e-06,
+      "loss": 0.1047,
+      "step": 18780
+    },
+    {
+      "epoch": 1.6233962590176683,
+      "grad_norm": 0.5739443373151654,
+      "learning_rate": 4.5770254005336005e-06,
+      "loss": 0.1076,
+      "step": 18790
+    },
+    {
+      "epoch": 1.6242602272236382,
+      "grad_norm": 0.5438322646921915,
+      "learning_rate": 4.5723784186793305e-06,
+      "loss": 0.1015,
+      "step": 18800
+    },
+    {
+      "epoch": 1.625124195429608,
+      "grad_norm": 0.5713437054250342,
+      "learning_rate": 4.56773180888693e-06,
+      "loss": 0.1049,
+      "step": 18810
+    },
+    {
+      "epoch": 1.6259881636355782,
+      "grad_norm": 0.5203624733377038,
+      "learning_rate": 4.563085575199288e-06,
+      "loss": 0.1033,
+      "step": 18820
+    },
+    {
+      "epoch": 1.6268521318415483,
+      "grad_norm": 0.5345371509745694,
+      "learning_rate": 4.558439721658962e-06,
+      "loss": 0.1029,
+      "step": 18830
+    },
+    {
+      "epoch": 1.6277161000475182,
+      "grad_norm": 0.5512000186661318,
+      "learning_rate": 4.5537942523081856e-06,
+      "loss": 0.1069,
+      "step": 18840
+    },
+    {
+      "epoch": 1.6285800682534881,
+      "grad_norm": 0.5732026077183064,
+      "learning_rate": 4.549149171188856e-06,
+      "loss": 0.1055,
+      "step": 18850
+    },
+    {
+      "epoch": 1.6294440364594585,
+      "grad_norm": 0.53897272246241,
+      "learning_rate": 4.5445044823425285e-06,
+      "loss": 0.1032,
+      "step": 18860
+    },
+    {
+      "epoch": 1.6303080046654284,
+      "grad_norm": 0.5385728364611817,
+      "learning_rate": 4.5398601898104215e-06,
+      "loss": 0.1045,
+      "step": 18870
+    },
+    {
+      "epoch": 1.6311719728713983,
+      "grad_norm": 0.5739375455349807,
+      "learning_rate": 4.535216297633407e-06,
+      "loss": 0.1042,
+      "step": 18880
+    },
+    {
+      "epoch": 1.6320359410773684,
+      "grad_norm": 0.5851177785810917,
+      "learning_rate": 4.53057280985201e-06,
+      "loss": 0.1107,
+      "step": 18890
+    },
+    {
+      "epoch": 1.6328999092833385,
+      "grad_norm": 0.5626896150925973,
+      "learning_rate": 4.5259297305064006e-06,
+      "loss": 0.1067,
+      "step": 18900
+    },
+    {
+      "epoch": 1.6337638774893084,
+      "grad_norm": 0.5579093379958637,
+      "learning_rate": 4.521287063636397e-06,
+      "loss": 0.1035,
+      "step": 18910
+    },
+    {
+      "epoch": 1.6346278456952783,
+      "grad_norm": 0.5463165649754007,
+      "learning_rate": 4.516644813281455e-06,
+      "loss": 0.1072,
+      "step": 18920
+    },
+    {
+      "epoch": 1.6354918139012484,
+      "grad_norm": 0.5299271950173727,
+      "learning_rate": 4.512002983480674e-06,
+      "loss": 0.1045,
+      "step": 18930
+    },
+    {
+      "epoch": 1.6363557821072185,
+      "grad_norm": 0.5295341692293466,
+      "learning_rate": 4.507361578272779e-06,
+      "loss": 0.1058,
+      "step": 18940
+    },
+    {
+      "epoch": 1.6372197503131884,
+      "grad_norm": 0.5659514898847705,
+      "learning_rate": 4.50272060169613e-06,
+      "loss": 0.1056,
+      "step": 18950
+    },
+    {
+      "epoch": 1.6380837185191583,
+      "grad_norm": 0.5355035407457457,
+      "learning_rate": 4.49808005778872e-06,
+      "loss": 0.1058,
+      "step": 18960
+    },
+    {
+      "epoch": 1.6389476867251285,
+      "grad_norm": 0.5813365434148485,
+      "learning_rate": 4.493439950588152e-06,
+      "loss": 0.1048,
+      "step": 18970
+    },
+    {
+      "epoch": 1.6398116549310986,
+      "grad_norm": 0.5465368834634429,
+      "learning_rate": 4.48880028413166e-06,
+      "loss": 0.1094,
+      "step": 18980
+    },
+    {
+      "epoch": 1.6406756231370685,
+      "grad_norm": 0.5594639004899338,
+      "learning_rate": 4.484161062456093e-06,
+      "loss": 0.1077,
+      "step": 18990
+    },
+    {
+      "epoch": 1.6415395913430386,
+      "grad_norm": 0.5619680445715887,
+      "learning_rate": 4.479522289597909e-06,
+      "loss": 0.1036,
+      "step": 19000
+    },
+    {
+      "epoch": 1.6424035595490087,
+      "grad_norm": 0.5458294496106755,
+      "learning_rate": 4.474883969593179e-06,
+      "loss": 0.1042,
+      "step": 19010
+    },
+    {
+      "epoch": 1.6432675277549786,
+      "grad_norm": 0.5718266897624316,
+      "learning_rate": 4.470246106477575e-06,
+      "loss": 0.1033,
+      "step": 19020
+    },
+    {
+      "epoch": 1.6441314959609485,
+      "grad_norm": 0.5726300062386809,
+      "learning_rate": 4.46560870428638e-06,
+      "loss": 0.108,
+      "step": 19030
+    },
+    {
+      "epoch": 1.6449954641669187,
+      "grad_norm": 0.5593823250439539,
+      "learning_rate": 4.460971767054469e-06,
+      "loss": 0.1054,
+      "step": 19040
+    },
+    {
+      "epoch": 1.6458594323728888,
+      "grad_norm": 0.5771387893259979,
+      "learning_rate": 4.456335298816314e-06,
+      "loss": 0.1038,
+      "step": 19050
+    },
+    {
+      "epoch": 1.6467234005788587,
+      "grad_norm": 0.5206178813990037,
+      "learning_rate": 4.45169930360598e-06,
+      "loss": 0.1047,
+      "step": 19060
+    },
+    {
+      "epoch": 1.6475873687848286,
+      "grad_norm": 0.5546056481802352,
+      "learning_rate": 4.4470637854571195e-06,
+      "loss": 0.1053,
+      "step": 19070
+    },
+    {
+      "epoch": 1.6484513369907987,
+      "grad_norm": 0.563121998454811,
+      "learning_rate": 4.442428748402974e-06,
+      "loss": 0.1025,
+      "step": 19080
+    },
+    {
+      "epoch": 1.6493153051967688,
+      "grad_norm": 0.5740974281010763,
+      "learning_rate": 4.437794196476357e-06,
+      "loss": 0.1077,
+      "step": 19090
+    },
+    {
+      "epoch": 1.6501792734027387,
+      "grad_norm": 0.5798995188610356,
+      "learning_rate": 4.433160133709668e-06,
+      "loss": 0.1077,
+      "step": 19100
+    },
+    {
+      "epoch": 1.6510432416087086,
+      "grad_norm": 0.5399967033915721,
+      "learning_rate": 4.428526564134879e-06,
+      "loss": 0.1051,
+      "step": 19110
+    },
+    {
+      "epoch": 1.651907209814679,
+      "grad_norm": 0.6047608211253362,
+      "learning_rate": 4.423893491783535e-06,
+      "loss": 0.1051,
+      "step": 19120
+    },
+    {
+      "epoch": 1.6527711780206489,
+      "grad_norm": 0.5567507402078828,
+      "learning_rate": 4.4192609206867395e-06,
+      "loss": 0.1044,
+      "step": 19130
+    },
+    {
+      "epoch": 1.6536351462266188,
+      "grad_norm": 0.5509176536860414,
+      "learning_rate": 4.414628854875171e-06,
+      "loss": 0.1082,
+      "step": 19140
+    },
+    {
+      "epoch": 1.654499114432589,
+      "grad_norm": 0.5807724228673988,
+      "learning_rate": 4.409997298379062e-06,
+      "loss": 0.1021,
+      "step": 19150
+    },
+    {
+      "epoch": 1.655363082638559,
+      "grad_norm": 0.5588511431747389,
+      "learning_rate": 4.405366255228206e-06,
+      "loss": 0.1057,
+      "step": 19160
+    },
+    {
+      "epoch": 1.656227050844529,
+      "grad_norm": 0.5760437138864196,
+      "learning_rate": 4.400735729451943e-06,
+      "loss": 0.0994,
+      "step": 19170
+    },
+    {
+      "epoch": 1.6570910190504988,
+      "grad_norm": 0.5862621716618907,
+      "learning_rate": 4.396105725079169e-06,
+      "loss": 0.1041,
+      "step": 19180
+    },
+    {
+      "epoch": 1.657954987256469,
+      "grad_norm": 0.5555879959070747,
+      "learning_rate": 4.391476246138326e-06,
+      "loss": 0.1052,
+      "step": 19190
+    },
+    {
+      "epoch": 1.658818955462439,
+      "grad_norm": 0.5952847226961522,
+      "learning_rate": 4.386847296657396e-06,
+      "loss": 0.1081,
+      "step": 19200
+    },
+    {
+      "epoch": 1.659682923668409,
+      "grad_norm": 0.5348380660765963,
+      "learning_rate": 4.382218880663902e-06,
+      "loss": 0.107,
+      "step": 19210
+    },
+    {
+      "epoch": 1.6605468918743789,
+      "grad_norm": 0.5151912291244257,
+      "learning_rate": 4.3775910021849e-06,
+      "loss": 0.1031,
+      "step": 19220
+    },
+    {
+      "epoch": 1.661410860080349,
+      "grad_norm": 0.6003358328007403,
+      "learning_rate": 4.372963665246986e-06,
+      "loss": 0.1064,
+      "step": 19230
+    },
+    {
+      "epoch": 1.6622748282863191,
+      "grad_norm": 0.5544041061975947,
+      "learning_rate": 4.368336873876273e-06,
+      "loss": 0.1014,
+      "step": 19240
+    },
+    {
+      "epoch": 1.663138796492289,
+      "grad_norm": 0.54913783519371,
+      "learning_rate": 4.36371063209841e-06,
+      "loss": 0.1036,
+      "step": 19250
+    },
+    {
+      "epoch": 1.6640027646982591,
+      "grad_norm": 0.5583738275839267,
+      "learning_rate": 4.359084943938564e-06,
+      "loss": 0.1064,
+      "step": 19260
+    },
+    {
+      "epoch": 1.6648667329042293,
+      "grad_norm": 0.5738850263871381,
+      "learning_rate": 4.35445981342142e-06,
+      "loss": 0.1049,
+      "step": 19270
+    },
+    {
+      "epoch": 1.6657307011101992,
+      "grad_norm": 0.5524237103340026,
+      "learning_rate": 4.349835244571175e-06,
+      "loss": 0.1042,
+      "step": 19280
+    },
+    {
+      "epoch": 1.666594669316169,
+      "grad_norm": 0.5706676298342727,
+      "learning_rate": 4.345211241411543e-06,
+      "loss": 0.1093,
+      "step": 19290
+    },
+    {
+      "epoch": 1.6674586375221392,
+      "grad_norm": 0.5848544135945213,
+      "learning_rate": 4.340587807965743e-06,
+      "loss": 0.1072,
+      "step": 19300
+    },
+    {
+      "epoch": 1.6683226057281093,
+      "grad_norm": 0.6287334419221372,
+      "learning_rate": 4.335964948256497e-06,
+      "loss": 0.1074,
+      "step": 19310
+    },
+    {
+      "epoch": 1.6691865739340792,
+      "grad_norm": 0.5777601253584108,
+      "learning_rate": 4.331342666306029e-06,
+      "loss": 0.1049,
+      "step": 19320
+    },
+    {
+      "epoch": 1.670050542140049,
+      "grad_norm": 0.5518672383982035,
+      "learning_rate": 4.32672096613606e-06,
+      "loss": 0.1076,
+      "step": 19330
+    },
+    {
+      "epoch": 1.6709145103460192,
+      "grad_norm": 0.5457043407209156,
+      "learning_rate": 4.322099851767807e-06,
+      "loss": 0.1063,
+      "step": 19340
+    },
+    {
+      "epoch": 1.6717784785519894,
+      "grad_norm": 0.5424461831045879,
+      "learning_rate": 4.317479327221976e-06,
+      "loss": 0.1063,
+      "step": 19350
+    },
+    {
+      "epoch": 1.6726424467579593,
+      "grad_norm": 0.5805324592882917,
+      "learning_rate": 4.3128593965187555e-06,
+      "loss": 0.1064,
+      "step": 19360
+    },
+    {
+      "epoch": 1.6735064149639294,
+      "grad_norm": 0.6130304203689181,
+      "learning_rate": 4.3082400636778236e-06,
+      "loss": 0.1042,
+      "step": 19370
+    },
+    {
+      "epoch": 1.6743703831698995,
+      "grad_norm": 0.5810864835602517,
+      "learning_rate": 4.303621332718336e-06,
+      "loss": 0.1037,
+      "step": 19380
+    },
+    {
+      "epoch": 1.6752343513758694,
+      "grad_norm": 0.5318618871960004,
+      "learning_rate": 4.299003207658926e-06,
+      "loss": 0.1095,
+      "step": 19390
+    },
+    {
+      "epoch": 1.6760983195818393,
+      "grad_norm": 0.5575880416342754,
+      "learning_rate": 4.294385692517696e-06,
+      "loss": 0.1069,
+      "step": 19400
+    },
+    {
+      "epoch": 1.6769622877878094,
+      "grad_norm": 0.5760120387957547,
+      "learning_rate": 4.289768791312219e-06,
+      "loss": 0.1031,
+      "step": 19410
+    },
+    {
+      "epoch": 1.6778262559937795,
+      "grad_norm": 0.6238448718107859,
+      "learning_rate": 4.2851525080595356e-06,
+      "loss": 0.1034,
+      "step": 19420
+    },
+    {
+      "epoch": 1.6786902241997494,
+      "grad_norm": 0.5617405307505385,
+      "learning_rate": 4.280536846776151e-06,
+      "loss": 0.1054,
+      "step": 19430
+    },
+    {
+      "epoch": 1.6795541924057193,
+      "grad_norm": 0.5540885463967568,
+      "learning_rate": 4.275921811478021e-06,
+      "loss": 0.1081,
+      "step": 19440
+    },
+    {
+      "epoch": 1.6804181606116895,
+      "grad_norm": 0.5559250887793661,
+      "learning_rate": 4.271307406180565e-06,
+      "loss": 0.1091,
+      "step": 19450
+    },
+    {
+      "epoch": 1.6812821288176596,
+      "grad_norm": 0.5462162389581664,
+      "learning_rate": 4.266693634898648e-06,
+      "loss": 0.1045,
+      "step": 19460
+    },
+    {
+      "epoch": 1.6821460970236295,
+      "grad_norm": 0.5768486265339877,
+      "learning_rate": 4.262080501646594e-06,
+      "loss": 0.1079,
+      "step": 19470
+    },
+    {
+      "epoch": 1.6830100652295994,
+      "grad_norm": 0.5679698629432213,
+      "learning_rate": 4.257468010438156e-06,
+      "loss": 0.1049,
+      "step": 19480
+    },
+    {
+      "epoch": 1.6838740334355695,
+      "grad_norm": 0.560233750108615,
+      "learning_rate": 4.25285616528654e-06,
+      "loss": 0.1053,
+      "step": 19490
+    },
+    {
+      "epoch": 1.6847380016415396,
+      "grad_norm": 0.5195629545137693,
+      "learning_rate": 4.248244970204388e-06,
+      "loss": 0.1025,
+      "step": 19500
+    },
+    {
+      "epoch": 1.6856019698475095,
+      "grad_norm": 0.5613215865113323,
+      "learning_rate": 4.243634429203774e-06,
+      "loss": 0.1029,
+      "step": 19510
+    },
+    {
+      "epoch": 1.6864659380534797,
+      "grad_norm": 0.5652931780024508,
+      "learning_rate": 4.2390245462962035e-06,
+      "loss": 0.1048,
+      "step": 19520
+    },
+    {
+      "epoch": 1.6873299062594498,
+      "grad_norm": 0.5816856644432576,
+      "learning_rate": 4.234415325492608e-06,
+      "loss": 0.1057,
+      "step": 19530
+    },
+    {
+      "epoch": 1.6881938744654197,
+      "grad_norm": 0.6059617188417905,
+      "learning_rate": 4.229806770803349e-06,
+      "loss": 0.1027,
+      "step": 19540
+    },
+    {
+      "epoch": 1.6890578426713896,
+      "grad_norm": 0.5445694254477236,
+      "learning_rate": 4.225198886238201e-06,
+      "loss": 0.1041,
+      "step": 19550
+    },
+    {
+      "epoch": 1.6899218108773597,
+      "grad_norm": 0.5581999508448194,
+      "learning_rate": 4.220591675806359e-06,
+      "loss": 0.1068,
+      "step": 19560
+    },
+    {
+      "epoch": 1.6907857790833298,
+      "grad_norm": 0.5546346046115186,
+      "learning_rate": 4.215985143516431e-06,
+      "loss": 0.1044,
+      "step": 19570
+    },
+    {
+      "epoch": 1.6916497472892997,
+      "grad_norm": 0.5439755044085209,
+      "learning_rate": 4.211379293376438e-06,
+      "loss": 0.1033,
+      "step": 19580
+    },
+    {
+      "epoch": 1.6925137154952696,
+      "grad_norm": 0.5656359038198634,
+      "learning_rate": 4.2067741293938e-06,
+      "loss": 0.1048,
+      "step": 19590
+    },
+    {
+      "epoch": 1.6933776837012398,
+      "grad_norm": 0.5590831186640056,
+      "learning_rate": 4.202169655575347e-06,
+      "loss": 0.1011,
+      "step": 19600
+    },
+    {
+      "epoch": 1.6942416519072099,
+      "grad_norm": 0.5394716638115005,
+      "learning_rate": 4.197565875927306e-06,
+      "loss": 0.1027,
+      "step": 19610
+    },
+    {
+      "epoch": 1.6951056201131798,
+      "grad_norm": 0.5634450989871348,
+      "learning_rate": 4.1929627944552995e-06,
+      "loss": 0.1058,
+      "step": 19620
+    },
+    {
+      "epoch": 1.69596958831915,
+      "grad_norm": 0.5515417589498646,
+      "learning_rate": 4.188360415164344e-06,
+      "loss": 0.1039,
+      "step": 19630
+    },
+    {
+      "epoch": 1.69683355652512,
+      "grad_norm": 0.5421289498514134,
+      "learning_rate": 4.183758742058842e-06,
+      "loss": 0.1052,
+      "step": 19640
+    },
+    {
+      "epoch": 1.69769752473109,
+      "grad_norm": 0.5844255281676082,
+      "learning_rate": 4.179157779142585e-06,
+      "loss": 0.1006,
+      "step": 19650
+    },
+    {
+      "epoch": 1.6985614929370598,
+      "grad_norm": 0.5736360543034765,
+      "learning_rate": 4.174557530418748e-06,
+      "loss": 0.1057,
+      "step": 19660
+    },
+    {
+      "epoch": 1.69942546114303,
+      "grad_norm": 0.5552882278271722,
+      "learning_rate": 4.169957999889877e-06,
+      "loss": 0.1028,
+      "step": 19670
+    },
+    {
+      "epoch": 1.700289429349,
+      "grad_norm": 0.544532249195041,
+      "learning_rate": 4.165359191557901e-06,
+      "loss": 0.103,
+      "step": 19680
+    },
+    {
+      "epoch": 1.70115339755497,
+      "grad_norm": 0.5246566731943877,
+      "learning_rate": 4.160761109424115e-06,
+      "loss": 0.1018,
+      "step": 19690
+    },
+    {
+      "epoch": 1.7020173657609399,
+      "grad_norm": 0.5417849366104356,
+      "learning_rate": 4.15616375748919e-06,
+      "loss": 0.1039,
+      "step": 19700
+    },
+    {
+      "epoch": 1.70288133396691,
+      "grad_norm": 0.5741977557459841,
+      "learning_rate": 4.151567139753152e-06,
+      "loss": 0.1062,
+      "step": 19710
+    },
+    {
+      "epoch": 1.7037453021728801,
+      "grad_norm": 0.5637437127823944,
+      "learning_rate": 4.1469712602153935e-06,
+      "loss": 0.1064,
+      "step": 19720
+    },
+    {
+      "epoch": 1.70460927037885,
+      "grad_norm": 0.5947331247456926,
+      "learning_rate": 4.142376122874664e-06,
+      "loss": 0.1059,
+      "step": 19730
+    },
+    {
+      "epoch": 1.70547323858482,
+      "grad_norm": 0.5566283619723043,
+      "learning_rate": 4.137781731729069e-06,
+      "loss": 0.1015,
+      "step": 19740
+    },
+    {
+      "epoch": 1.7063372067907903,
+      "grad_norm": 0.5582794032645616,
+      "learning_rate": 4.1331880907760595e-06,
+      "loss": 0.1089,
+      "step": 19750
+    },
+    {
+      "epoch": 1.7072011749967602,
+      "grad_norm": 0.5836144846557693,
+      "learning_rate": 4.1285952040124375e-06,
+      "loss": 0.1022,
+      "step": 19760
+    },
+    {
+      "epoch": 1.70806514320273,
+      "grad_norm": 0.5676899148191863,
+      "learning_rate": 4.124003075434351e-06,
+      "loss": 0.1094,
+      "step": 19770
+    },
+    {
+      "epoch": 1.7089291114087002,
+      "grad_norm": 0.5207672775373055,
+      "learning_rate": 4.119411709037286e-06,
+      "loss": 0.1022,
+      "step": 19780
+    },
+    {
+      "epoch": 1.7097930796146703,
+      "grad_norm": 0.5410470442889012,
+      "learning_rate": 4.114821108816063e-06,
+      "loss": 0.1012,
+      "step": 19790
+    },
+    {
+      "epoch": 1.7106570478206402,
+      "grad_norm": 0.5620634468728611,
+      "learning_rate": 4.110231278764837e-06,
+      "loss": 0.1015,
+      "step": 19800
+    },
+    {
+      "epoch": 1.71152101602661,
+      "grad_norm": 0.5625517889943247,
+      "learning_rate": 4.1056422228770986e-06,
+      "loss": 0.1042,
+      "step": 19810
+    },
+    {
+      "epoch": 1.7123849842325802,
+      "grad_norm": 0.5551478643367547,
+      "learning_rate": 4.101053945145655e-06,
+      "loss": 0.1054,
+      "step": 19820
+    },
+    {
+      "epoch": 1.7132489524385504,
+      "grad_norm": 0.5961782039221952,
+      "learning_rate": 4.0964664495626436e-06,
+      "loss": 0.1068,
+      "step": 19830
+    },
+    {
+      "epoch": 1.7141129206445203,
+      "grad_norm": 0.577866925877064,
+      "learning_rate": 4.091879740119518e-06,
+      "loss": 0.1051,
+      "step": 19840
+    },
+    {
+      "epoch": 1.7149768888504902,
+      "grad_norm": 0.5938704854801597,
+      "learning_rate": 4.087293820807054e-06,
+      "loss": 0.1056,
+      "step": 19850
+    },
+    {
+      "epoch": 1.7158408570564603,
+      "grad_norm": 0.5403360078049187,
+      "learning_rate": 4.082708695615326e-06,
+      "loss": 0.103,
+      "step": 19860
+    },
+    {
+      "epoch": 1.7167048252624304,
+      "grad_norm": 0.5435756955040987,
+      "learning_rate": 4.078124368533733e-06,
+      "loss": 0.104,
+      "step": 19870
+    },
+    {
+      "epoch": 1.7175687934684003,
+      "grad_norm": 0.5541908421361005,
+      "learning_rate": 4.07354084355097e-06,
+      "loss": 0.1004,
+      "step": 19880
+    },
+    {
+      "epoch": 1.7184327616743704,
+      "grad_norm": 0.5886329233256653,
+      "learning_rate": 4.06895812465504e-06,
+      "loss": 0.1068,
+      "step": 19890
+    },
+    {
+      "epoch": 1.7192967298803405,
+      "grad_norm": 0.5707722202962463,
+      "learning_rate": 4.064376215833238e-06,
+      "loss": 0.105,
+      "step": 19900
+    },
+    {
+      "epoch": 1.7201606980863104,
+      "grad_norm": 0.5674191409460011,
+      "learning_rate": 4.05979512107216e-06,
+      "loss": 0.103,
+      "step": 19910
+    },
+    {
+      "epoch": 1.7210246662922803,
+      "grad_norm": 0.5519385356483035,
+      "learning_rate": 4.055214844357692e-06,
+      "loss": 0.1028,
+      "step": 19920
+    },
+    {
+      "epoch": 1.7218886344982505,
+      "grad_norm": 0.5548644777530287,
+      "learning_rate": 4.050635389675006e-06,
+      "loss": 0.1019,
+      "step": 19930
+    },
+    {
+      "epoch": 1.7227526027042206,
+      "grad_norm": 0.5631549051283312,
+      "learning_rate": 4.046056761008561e-06,
+      "loss": 0.1054,
+      "step": 19940
+    },
+    {
+      "epoch": 1.7236165709101905,
+      "grad_norm": 0.5606486954370578,
+      "learning_rate": 4.041478962342098e-06,
+      "loss": 0.1024,
+      "step": 19950
+    },
+    {
+      "epoch": 1.7244805391161604,
+      "grad_norm": 0.5667956495178411,
+      "learning_rate": 4.036901997658632e-06,
+      "loss": 0.1031,
+      "step": 19960
+    },
+    {
+      "epoch": 1.7253445073221305,
+      "grad_norm": 0.5455738969443668,
+      "learning_rate": 4.03232587094046e-06,
+      "loss": 0.1016,
+      "step": 19970
+    },
+    {
+      "epoch": 1.7262084755281006,
+      "grad_norm": 0.5615537609092984,
+      "learning_rate": 4.0277505861691405e-06,
+      "loss": 0.1049,
+      "step": 19980
+    },
+    {
+      "epoch": 1.7270724437340705,
+      "grad_norm": 0.6077668734633745,
+      "learning_rate": 4.023176147325505e-06,
+      "loss": 0.1054,
+      "step": 19990
+    },
+    {
+      "epoch": 1.7279364119400407,
+      "grad_norm": 0.5353436552303115,
+      "learning_rate": 4.018602558389648e-06,
+      "loss": 0.1048,
+      "step": 20000
+    },
+    {
+      "epoch": 1.7288003801460108,
+      "grad_norm": 0.547027306615421,
+      "learning_rate": 4.014029823340928e-06,
+      "loss": 0.1022,
+      "step": 20010
+    },
+    {
+      "epoch": 1.7296643483519807,
+      "grad_norm": 0.5953265053533482,
+      "learning_rate": 4.0094579461579505e-06,
+      "loss": 0.1055,
+      "step": 20020
+    },
+    {
+      "epoch": 1.7305283165579506,
+      "grad_norm": 0.5527840216631419,
+      "learning_rate": 4.0048869308185856e-06,
+      "loss": 0.104,
+      "step": 20030
+    },
+    {
+      "epoch": 1.7313922847639207,
+      "grad_norm": 0.5631809649129795,
+      "learning_rate": 4.0003167812999465e-06,
+      "loss": 0.1031,
+      "step": 20040
+    },
+    {
+      "epoch": 1.7322562529698908,
+      "grad_norm": 0.5420978278693646,
+      "learning_rate": 3.9957475015784e-06,
+      "loss": 0.1024,
+      "step": 20050
+    },
+    {
+      "epoch": 1.7331202211758607,
+      "grad_norm": 0.5495208887748646,
+      "learning_rate": 3.9911790956295455e-06,
+      "loss": 0.099,
+      "step": 20060
+    },
+    {
+      "epoch": 1.7339841893818306,
+      "grad_norm": 0.6077615563313703,
+      "learning_rate": 3.986611567428231e-06,
+      "loss": 0.1039,
+      "step": 20070
+    },
+    {
+      "epoch": 1.7348481575878008,
+      "grad_norm": 0.5575610185779836,
+      "learning_rate": 3.982044920948542e-06,
+      "loss": 0.1041,
+      "step": 20080
+    },
+    {
+      "epoch": 1.7357121257937709,
+      "grad_norm": 0.5712717988692463,
+      "learning_rate": 3.977479160163786e-06,
+      "loss": 0.1036,
+      "step": 20090
+    },
+    {
+      "epoch": 1.7365760939997408,
+      "grad_norm": 0.5868440064927699,
+      "learning_rate": 3.972914289046512e-06,
+      "loss": 0.1051,
+      "step": 20100
+    },
+    {
+      "epoch": 1.7374400622057107,
+      "grad_norm": 0.5571265568035261,
+      "learning_rate": 3.968350311568487e-06,
+      "loss": 0.1059,
+      "step": 20110
+    },
+    {
+      "epoch": 1.7383040304116808,
+      "grad_norm": 0.5571341134249902,
+      "learning_rate": 3.963787231700707e-06,
+      "loss": 0.1009,
+      "step": 20120
+    },
+    {
+      "epoch": 1.739167998617651,
+      "grad_norm": 0.5442273293247737,
+      "learning_rate": 3.959225053413379e-06,
+      "loss": 0.1056,
+      "step": 20130
+    },
+    {
+      "epoch": 1.7400319668236208,
+      "grad_norm": 0.5392709739565602,
+      "learning_rate": 3.954663780675932e-06,
+      "loss": 0.1044,
+      "step": 20140
+    },
+    {
+      "epoch": 1.740895935029591,
+      "grad_norm": 0.5714865715662484,
+      "learning_rate": 3.950103417457004e-06,
+      "loss": 0.1003,
+      "step": 20150
+    },
+    {
+      "epoch": 1.741759903235561,
+      "grad_norm": 0.5611686800757524,
+      "learning_rate": 3.945543967724444e-06,
+      "loss": 0.1062,
+      "step": 20160
+    },
+    {
+      "epoch": 1.742623871441531,
+      "grad_norm": 0.562876377608983,
+      "learning_rate": 3.940985435445303e-06,
+      "loss": 0.1041,
+      "step": 20170
+    },
+    {
+      "epoch": 1.7434878396475009,
+      "grad_norm": 0.5365688293514245,
+      "learning_rate": 3.936427824585836e-06,
+      "loss": 0.1044,
+      "step": 20180
+    },
+    {
+      "epoch": 1.744351807853471,
+      "grad_norm": 0.5784734187598308,
+      "learning_rate": 3.931871139111497e-06,
+      "loss": 0.1012,
+      "step": 20190
+    },
+    {
+      "epoch": 1.7452157760594411,
+      "grad_norm": 0.56924590426923,
+      "learning_rate": 3.927315382986935e-06,
+      "loss": 0.1048,
+      "step": 20200
+    },
+    {
+      "epoch": 1.746079744265411,
+      "grad_norm": 0.5868943193747296,
+      "learning_rate": 3.922760560175984e-06,
+      "loss": 0.103,
+      "step": 20210
+    },
+    {
+      "epoch": 1.746943712471381,
+      "grad_norm": 0.5684990052889753,
+      "learning_rate": 3.918206674641674e-06,
+      "loss": 0.1054,
+      "step": 20220
+    },
+    {
+      "epoch": 1.747807680677351,
+      "grad_norm": 0.5984985926149132,
+      "learning_rate": 3.913653730346219e-06,
+      "loss": 0.1005,
+      "step": 20230
+    },
+    {
+      "epoch": 1.7486716488833212,
+      "grad_norm": 0.5719741375137701,
+      "learning_rate": 3.909101731251008e-06,
+      "loss": 0.1034,
+      "step": 20240
+    },
+    {
+      "epoch": 1.749535617089291,
+      "grad_norm": 0.5445905911014346,
+      "learning_rate": 3.904550681316613e-06,
+      "loss": 0.1004,
+      "step": 20250
+    },
+    {
+      "epoch": 1.7503995852952612,
+      "grad_norm": 0.5535256139758667,
+      "learning_rate": 3.900000584502777e-06,
+      "loss": 0.0999,
+      "step": 20260
+    },
+    {
+      "epoch": 1.7512635535012313,
+      "grad_norm": 0.5764065311727894,
+      "learning_rate": 3.8954514447684154e-06,
+      "loss": 0.1007,
+      "step": 20270
+    },
+    {
+      "epoch": 1.7521275217072012,
+      "grad_norm": 0.5458153301564745,
+      "learning_rate": 3.890903266071614e-06,
+      "loss": 0.1033,
+      "step": 20280
+    },
+    {
+      "epoch": 1.752991489913171,
+      "grad_norm": 0.562777505638518,
+      "learning_rate": 3.886356052369613e-06,
+      "loss": 0.1022,
+      "step": 20290
+    },
+    {
+      "epoch": 1.7538554581191412,
+      "grad_norm": 0.5552263607405525,
+      "learning_rate": 3.881809807618822e-06,
+      "loss": 0.1014,
+      "step": 20300
+    },
+    {
+      "epoch": 1.7547194263251114,
+      "grad_norm": 0.5809887010445461,
+      "learning_rate": 3.8772645357748055e-06,
+      "loss": 0.1025,
+      "step": 20310
+    },
+    {
+      "epoch": 1.7555833945310813,
+      "grad_norm": 0.5851596272145743,
+      "learning_rate": 3.872720240792281e-06,
+      "loss": 0.1038,
+      "step": 20320
+    },
+    {
+      "epoch": 1.7564473627370512,
+      "grad_norm": 0.5919873050029621,
+      "learning_rate": 3.868176926625111e-06,
+      "loss": 0.1023,
+      "step": 20330
+    },
+    {
+      "epoch": 1.7573113309430213,
+      "grad_norm": 0.5763379694868309,
+      "learning_rate": 3.863634597226314e-06,
+      "loss": 0.1012,
+      "step": 20340
+    },
+    {
+      "epoch": 1.7581752991489914,
+      "grad_norm": 0.5913953702067776,
+      "learning_rate": 3.859093256548044e-06,
+      "loss": 0.1058,
+      "step": 20350
+    },
+    {
+      "epoch": 1.7590392673549613,
+      "grad_norm": 0.5320930353822502,
+      "learning_rate": 3.854552908541601e-06,
+      "loss": 0.101,
+      "step": 20360
+    },
+    {
+      "epoch": 1.7599032355609312,
+      "grad_norm": 0.5513593841607388,
+      "learning_rate": 3.850013557157413e-06,
+      "loss": 0.1017,
+      "step": 20370
+    },
+    {
+      "epoch": 1.7607672037669013,
+      "grad_norm": 0.5272766571509331,
+      "learning_rate": 3.845475206345048e-06,
+      "loss": 0.105,
+      "step": 20380
+    },
+    {
+      "epoch": 1.7616311719728714,
+      "grad_norm": 0.561909459049623,
+      "learning_rate": 3.840937860053204e-06,
+      "loss": 0.1056,
+      "step": 20390
+    },
+    {
+      "epoch": 1.7624951401788413,
+      "grad_norm": 0.5623218739821189,
+      "learning_rate": 3.836401522229698e-06,
+      "loss": 0.1018,
+      "step": 20400
+    },
+    {
+      "epoch": 1.7633591083848115,
+      "grad_norm": 0.5713647997114392,
+      "learning_rate": 3.831866196821476e-06,
+      "loss": 0.1009,
+      "step": 20410
+    },
+    {
+      "epoch": 1.7642230765907816,
+      "grad_norm": 0.5537859362481026,
+      "learning_rate": 3.827331887774599e-06,
+      "loss": 0.1021,
+      "step": 20420
+    },
+    {
+      "epoch": 1.7650870447967515,
+      "grad_norm": 0.5606600806201089,
+      "learning_rate": 3.8227985990342495e-06,
+      "loss": 0.0982,
+      "step": 20430
+    },
+    {
+      "epoch": 1.7659510130027214,
+      "grad_norm": 0.5701330136609685,
+      "learning_rate": 3.818266334544714e-06,
+      "loss": 0.1042,
+      "step": 20440
+    },
+    {
+      "epoch": 1.7668149812086915,
+      "grad_norm": 0.5731712287364021,
+      "learning_rate": 3.813735098249395e-06,
+      "loss": 0.1039,
+      "step": 20450
+    },
+    {
+      "epoch": 1.7676789494146616,
+      "grad_norm": 0.5673002836546218,
+      "learning_rate": 3.8092048940907944e-06,
+      "loss": 0.1043,
+      "step": 20460
+    },
+    {
+      "epoch": 1.7685429176206315,
+      "grad_norm": 0.5670893355009318,
+      "learning_rate": 3.8046757260105244e-06,
+      "loss": 0.1024,
+      "step": 20470
+    },
+    {
+      "epoch": 1.7694068858266014,
+      "grad_norm": 0.5944192806216618,
+      "learning_rate": 3.800147597949285e-06,
+      "loss": 0.1017,
+      "step": 20480
+    },
+    {
+      "epoch": 1.7702708540325716,
+      "grad_norm": 0.5670087860330885,
+      "learning_rate": 3.7956205138468795e-06,
+      "loss": 0.1015,
+      "step": 20490
+    },
+    {
+      "epoch": 1.7711348222385417,
+      "grad_norm": 0.5817270080000759,
+      "learning_rate": 3.7910944776422e-06,
+      "loss": 0.1039,
+      "step": 20500
+    },
+    {
+      "epoch": 1.7719987904445116,
+      "grad_norm": 0.5724007713380568,
+      "learning_rate": 3.7865694932732296e-06,
+      "loss": 0.1051,
+      "step": 20510
+    },
+    {
+      "epoch": 1.7728627586504817,
+      "grad_norm": 0.5473492286689364,
+      "learning_rate": 3.7820455646770284e-06,
+      "loss": 0.1024,
+      "step": 20520
+    },
+    {
+      "epoch": 1.7737267268564518,
+      "grad_norm": 0.5485247247016158,
+      "learning_rate": 3.7775226957897465e-06,
+      "loss": 0.1019,
+      "step": 20530
+    },
+    {
+      "epoch": 1.7745906950624217,
+      "grad_norm": 0.5643486978024851,
+      "learning_rate": 3.773000890546609e-06,
+      "loss": 0.104,
+      "step": 20540
+    },
+    {
+      "epoch": 1.7754546632683916,
+      "grad_norm": 0.5796798917661853,
+      "learning_rate": 3.7684801528819155e-06,
+      "loss": 0.1047,
+      "step": 20550
+    },
+    {
+      "epoch": 1.7763186314743618,
+      "grad_norm": 0.545404357770531,
+      "learning_rate": 3.763960486729035e-06,
+      "loss": 0.1056,
+      "step": 20560
+    },
+    {
+      "epoch": 1.7771825996803319,
+      "grad_norm": 0.5357278759624914,
+      "learning_rate": 3.759441896020405e-06,
+      "loss": 0.1036,
+      "step": 20570
+    },
+    {
+      "epoch": 1.7780465678863018,
+      "grad_norm": 0.5918980827105917,
+      "learning_rate": 3.7549243846875288e-06,
+      "loss": 0.1016,
+      "step": 20580
+    },
+    {
+      "epoch": 1.7789105360922717,
+      "grad_norm": 0.557293641693346,
+      "learning_rate": 3.750407956660973e-06,
+      "loss": 0.1025,
+      "step": 20590
+    },
+    {
+      "epoch": 1.7797745042982418,
+      "grad_norm": 0.5442810225575131,
+      "learning_rate": 3.745892615870353e-06,
+      "loss": 0.0984,
+      "step": 20600
+    },
+    {
+      "epoch": 1.780638472504212,
+      "grad_norm": 0.5844983519187161,
+      "learning_rate": 3.741378366244346e-06,
+      "loss": 0.1044,
+      "step": 20610
+    },
+    {
+      "epoch": 1.7815024407101818,
+      "grad_norm": 0.6092327070617354,
+      "learning_rate": 3.736865211710678e-06,
+      "loss": 0.1032,
+      "step": 20620
+    },
+    {
+      "epoch": 1.7823664089161517,
+      "grad_norm": 0.5845750451198817,
+      "learning_rate": 3.732353156196123e-06,
+      "loss": 0.106,
+      "step": 20630
+    },
+    {
+      "epoch": 1.783230377122122,
+      "grad_norm": 0.5142430347585016,
+      "learning_rate": 3.727842203626494e-06,
+      "loss": 0.0974,
+      "step": 20640
+    },
+    {
+      "epoch": 1.784094345328092,
+      "grad_norm": 0.5477628186400496,
+      "learning_rate": 3.723332357926649e-06,
+      "loss": 0.1003,
+      "step": 20650
+    },
+    {
+      "epoch": 1.7849583135340619,
+      "grad_norm": 0.5303665814483901,
+      "learning_rate": 3.7188236230204826e-06,
+      "loss": 0.1077,
+      "step": 20660
+    },
+    {
+      "epoch": 1.785822281740032,
+      "grad_norm": 0.5762679439825226,
+      "learning_rate": 3.7143160028309215e-06,
+      "loss": 0.1047,
+      "step": 20670
+    },
+    {
+      "epoch": 1.7866862499460021,
+      "grad_norm": 0.5530133329338524,
+      "learning_rate": 3.7098095012799216e-06,
+      "loss": 0.0995,
+      "step": 20680
+    },
+    {
+      "epoch": 1.787550218151972,
+      "grad_norm": 0.5482584661223466,
+      "learning_rate": 3.7053041222884688e-06,
+      "loss": 0.1026,
+      "step": 20690
+    },
+    {
+      "epoch": 1.788414186357942,
+      "grad_norm": 0.580595778741035,
+      "learning_rate": 3.7007998697765713e-06,
+      "loss": 0.106,
+      "step": 20700
+    },
+    {
+      "epoch": 1.789278154563912,
+      "grad_norm": 0.576869662947042,
+      "learning_rate": 3.696296747663253e-06,
+      "loss": 0.102,
+      "step": 20710
+    },
+    {
+      "epoch": 1.7901421227698822,
+      "grad_norm": 0.5597875975808141,
+      "learning_rate": 3.6917947598665593e-06,
+      "loss": 0.1014,
+      "step": 20720
+    },
+    {
+      "epoch": 1.791006090975852,
+      "grad_norm": 0.558351611350602,
+      "learning_rate": 3.6872939103035465e-06,
+      "loss": 0.1006,
+      "step": 20730
+    },
+    {
+      "epoch": 1.791870059181822,
+      "grad_norm": 0.5606337066934608,
+      "learning_rate": 3.682794202890284e-06,
+      "loss": 0.105,
+      "step": 20740
+    },
+    {
+      "epoch": 1.792734027387792,
+      "grad_norm": 0.5690262394802477,
+      "learning_rate": 3.6782956415418404e-06,
+      "loss": 0.1015,
+      "step": 20750
+    },
+    {
+      "epoch": 1.7935979955937622,
+      "grad_norm": 0.5554666935458995,
+      "learning_rate": 3.673798230172293e-06,
+      "loss": 0.1012,
+      "step": 20760
+    },
+    {
+      "epoch": 1.794461963799732,
+      "grad_norm": 0.5826019998816877,
+      "learning_rate": 3.6693019726947154e-06,
+      "loss": 0.1007,
+      "step": 20770
+    },
+    {
+      "epoch": 1.7953259320057022,
+      "grad_norm": 0.552502735816188,
+      "learning_rate": 3.6648068730211816e-06,
+      "loss": 0.1005,
+      "step": 20780
+    },
+    {
+      "epoch": 1.7961899002116724,
+      "grad_norm": 0.5604779925921183,
+      "learning_rate": 3.660312935062752e-06,
+      "loss": 0.1015,
+      "step": 20790
+    },
+    {
+      "epoch": 1.7970538684176423,
+      "grad_norm": 0.553123810366442,
+      "learning_rate": 3.65582016272948e-06,
+      "loss": 0.1009,
+      "step": 20800
+    },
+    {
+      "epoch": 1.7979178366236122,
+      "grad_norm": 0.574040004269223,
+      "learning_rate": 3.651328559930404e-06,
+      "loss": 0.1024,
+      "step": 20810
+    },
+    {
+      "epoch": 1.7987818048295823,
+      "grad_norm": 0.5789047190106877,
+      "learning_rate": 3.6468381305735485e-06,
+      "loss": 0.1001,
+      "step": 20820
+    },
+    {
+      "epoch": 1.7996457730355524,
+      "grad_norm": 0.5743348463122552,
+      "learning_rate": 3.6423488785659085e-06,
+      "loss": 0.1063,
+      "step": 20830
+    },
+    {
+      "epoch": 1.8005097412415223,
+      "grad_norm": 0.568363700602923,
+      "learning_rate": 3.637860807813462e-06,
+      "loss": 0.1022,
+      "step": 20840
+    },
+    {
+      "epoch": 1.8013737094474922,
+      "grad_norm": 0.5632244458817115,
+      "learning_rate": 3.6333739222211576e-06,
+      "loss": 0.1005,
+      "step": 20850
+    },
+    {
+      "epoch": 1.8022376776534623,
+      "grad_norm": 0.5535634272232598,
+      "learning_rate": 3.628888225692912e-06,
+      "loss": 0.1038,
+      "step": 20860
+    },
+    {
+      "epoch": 1.8031016458594324,
+      "grad_norm": 0.5617936512836789,
+      "learning_rate": 3.6244037221316066e-06,
+      "loss": 0.1015,
+      "step": 20870
+    },
+    {
+      "epoch": 1.8039656140654023,
+      "grad_norm": 0.5676594590712748,
+      "learning_rate": 3.619920415439084e-06,
+      "loss": 0.0987,
+      "step": 20880
+    },
+    {
+      "epoch": 1.8048295822713725,
+      "grad_norm": 0.5548626950647346,
+      "learning_rate": 3.615438309516148e-06,
+      "loss": 0.1008,
+      "step": 20890
+    },
+    {
+      "epoch": 1.8056935504773426,
+      "grad_norm": 0.5472900095239701,
+      "learning_rate": 3.610957408262561e-06,
+      "loss": 0.103,
+      "step": 20900
+    },
+    {
+      "epoch": 1.8065575186833125,
+      "grad_norm": 0.556208859315823,
+      "learning_rate": 3.606477715577026e-06,
+      "loss": 0.1029,
+      "step": 20910
+    },
+    {
+      "epoch": 1.8074214868892824,
+      "grad_norm": 0.5843164924577277,
+      "learning_rate": 3.6019992353572047e-06,
+      "loss": 0.1014,
+      "step": 20920
+    },
+    {
+      "epoch": 1.8082854550952525,
+      "grad_norm": 0.5261056588336221,
+      "learning_rate": 3.5975219714997025e-06,
+      "loss": 0.1008,
+      "step": 20930
+    },
+    {
+      "epoch": 1.8091494233012226,
+      "grad_norm": 0.5291736647089412,
+      "learning_rate": 3.5930459279000606e-06,
+      "loss": 0.1035,
+      "step": 20940
+    },
+    {
+      "epoch": 1.8100133915071925,
+      "grad_norm": 0.5657988364870209,
+      "learning_rate": 3.588571108452764e-06,
+      "loss": 0.1034,
+      "step": 20950
+    },
+    {
+      "epoch": 1.8108773597131624,
+      "grad_norm": 0.5825009374567696,
+      "learning_rate": 3.5840975170512314e-06,
+      "loss": 0.1002,
+      "step": 20960
+    },
+    {
+      "epoch": 1.8117413279191326,
+      "grad_norm": 0.5580209417771652,
+      "learning_rate": 3.579625157587814e-06,
+      "loss": 0.101,
+      "step": 20970
+    },
+    {
+      "epoch": 1.8126052961251027,
+      "grad_norm": 0.523370115805435,
+      "learning_rate": 3.575154033953787e-06,
+      "loss": 0.1032,
+      "step": 20980
+    },
+    {
+      "epoch": 1.8134692643310726,
+      "grad_norm": 0.5803826900106622,
+      "learning_rate": 3.570684150039353e-06,
+      "loss": 0.1014,
+      "step": 20990
+    },
+    {
+      "epoch": 1.8143332325370425,
+      "grad_norm": 0.53633215338986,
+      "learning_rate": 3.5662155097336378e-06,
+      "loss": 0.1021,
+      "step": 21000
+    },
+    {
+      "epoch": 1.8151972007430126,
+      "grad_norm": 0.5457685347468603,
+      "learning_rate": 3.5617481169246845e-06,
+      "loss": 0.1062,
+      "step": 21010
+    },
+    {
+      "epoch": 1.8160611689489827,
+      "grad_norm": 0.5436165338718727,
+      "learning_rate": 3.557281975499446e-06,
+      "loss": 0.1,
+      "step": 21020
+    },
+    {
+      "epoch": 1.8169251371549526,
+      "grad_norm": 0.5219347875259863,
+      "learning_rate": 3.5528170893437918e-06,
+      "loss": 0.1022,
+      "step": 21030
+    },
+    {
+      "epoch": 1.8177891053609228,
+      "grad_norm": 0.5459117725002699,
+      "learning_rate": 3.5483534623424988e-06,
+      "loss": 0.1026,
+      "step": 21040
+    },
+    {
+      "epoch": 1.8186530735668929,
+      "grad_norm": 0.5344167085133646,
+      "learning_rate": 3.5438910983792465e-06,
+      "loss": 0.0988,
+      "step": 21050
+    },
+    {
+      "epoch": 1.8195170417728628,
+      "grad_norm": 0.533995913655794,
+      "learning_rate": 3.539430001336614e-06,
+      "loss": 0.1022,
+      "step": 21060
+    },
+    {
+      "epoch": 1.8203810099788327,
+      "grad_norm": 0.5395042423170748,
+      "learning_rate": 3.534970175096083e-06,
+      "loss": 0.1031,
+      "step": 21070
+    },
+    {
+      "epoch": 1.8212449781848028,
+      "grad_norm": 0.5508954228284384,
+      "learning_rate": 3.5305116235380233e-06,
+      "loss": 0.1036,
+      "step": 21080
+    },
+    {
+      "epoch": 1.822108946390773,
+      "grad_norm": 0.5459122344449423,
+      "learning_rate": 3.526054350541704e-06,
+      "loss": 0.1028,
+      "step": 21090
+    },
+    {
+      "epoch": 1.8229729145967428,
+      "grad_norm": 0.573507303035393,
+      "learning_rate": 3.521598359985271e-06,
+      "loss": 0.1016,
+      "step": 21100
+    },
+    {
+      "epoch": 1.8238368828027127,
+      "grad_norm": 0.5437055323857766,
+      "learning_rate": 3.5171436557457628e-06,
+      "loss": 0.0984,
+      "step": 21110
+    },
+    {
+      "epoch": 1.8247008510086828,
+      "grad_norm": 0.5670371375647348,
+      "learning_rate": 3.512690241699096e-06,
+      "loss": 0.1008,
+      "step": 21120
+    },
+    {
+      "epoch": 1.825564819214653,
+      "grad_norm": 0.5628275899816054,
+      "learning_rate": 3.5082381217200668e-06,
+      "loss": 0.0983,
+      "step": 21130
+    },
+    {
+      "epoch": 1.8264287874206229,
+      "grad_norm": 0.57987227106917,
+      "learning_rate": 3.5037872996823384e-06,
+      "loss": 0.1009,
+      "step": 21140
+    },
+    {
+      "epoch": 1.827292755626593,
+      "grad_norm": 0.5597332618992171,
+      "learning_rate": 3.4993377794584532e-06,
+      "loss": 0.1042,
+      "step": 21150
+    },
+    {
+      "epoch": 1.8281567238325631,
+      "grad_norm": 0.5527884910029922,
+      "learning_rate": 3.4948895649198156e-06,
+      "loss": 0.1002,
+      "step": 21160
+    },
+    {
+      "epoch": 1.829020692038533,
+      "grad_norm": 0.5399126337821516,
+      "learning_rate": 3.4904426599366985e-06,
+      "loss": 0.0998,
+      "step": 21170
+    },
+    {
+      "epoch": 1.829884660244503,
+      "grad_norm": 0.5533812170631007,
+      "learning_rate": 3.4859970683782283e-06,
+      "loss": 0.1029,
+      "step": 21180
+    },
+    {
+      "epoch": 1.830748628450473,
+      "grad_norm": 0.570090142425241,
+      "learning_rate": 3.481552794112395e-06,
+      "loss": 0.1002,
+      "step": 21190
+    },
+    {
+      "epoch": 1.8316125966564432,
+      "grad_norm": 0.576237314844463,
+      "learning_rate": 3.4771098410060433e-06,
+      "loss": 0.101,
+      "step": 21200
+    },
+    {
+      "epoch": 1.832476564862413,
+      "grad_norm": 0.5785261238185759,
+      "learning_rate": 3.47266821292486e-06,
+      "loss": 0.1026,
+      "step": 21210
+    },
+    {
+      "epoch": 1.833340533068383,
+      "grad_norm": 0.5391411473596741,
+      "learning_rate": 3.4682279137333874e-06,
+      "loss": 0.1009,
+      "step": 21220
+    },
+    {
+      "epoch": 1.834204501274353,
+      "grad_norm": 0.5717021882218313,
+      "learning_rate": 3.463788947295008e-06,
+      "loss": 0.1017,
+      "step": 21230
+    },
+    {
+      "epoch": 1.8350684694803232,
+      "grad_norm": 0.5641109625964653,
+      "learning_rate": 3.459351317471948e-06,
+      "loss": 0.1015,
+      "step": 21240
+    },
+    {
+      "epoch": 1.835932437686293,
+      "grad_norm": 0.5841912468112753,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.1009,
+      "step": 21250
+    },
+    {
+      "epoch": 1.836796405892263,
+      "grad_norm": 0.5775667698271872,
+      "learning_rate": 3.4504800831148523e-06,
+      "loss": 0.1015,
+      "step": 21260
+    },
+    {
+      "epoch": 1.8376603740982334,
+      "grad_norm": 0.5511909835861091,
+      "learning_rate": 3.446046486299437e-06,
+      "loss": 0.1001,
+      "step": 21270
+    },
+    {
+      "epoch": 1.8385243423042033,
+      "grad_norm": 0.5344861997274931,
+      "learning_rate": 3.4416142415365718e-06,
+      "loss": 0.1046,
+      "step": 21280
+    },
+    {
+      "epoch": 1.8393883105101732,
+      "grad_norm": 0.5756081848954616,
+      "learning_rate": 3.437183352682628e-06,
+      "loss": 0.1033,
+      "step": 21290
+    },
+    {
+      "epoch": 1.8402522787161433,
+      "grad_norm": 0.5671991668813597,
+      "learning_rate": 3.432753823592804e-06,
+      "loss": 0.1046,
+      "step": 21300
+    },
+    {
+      "epoch": 1.8411162469221134,
+      "grad_norm": 0.5782180776536016,
+      "learning_rate": 3.42832565812111e-06,
+      "loss": 0.0986,
+      "step": 21310
+    },
+    {
+      "epoch": 1.8419802151280833,
+      "grad_norm": 0.5327885416091026,
+      "learning_rate": 3.4238988601203766e-06,
+      "loss": 0.103,
+      "step": 21320
+    },
+    {
+      "epoch": 1.8428441833340532,
+      "grad_norm": 0.5301622627094087,
+      "learning_rate": 3.4194734334422343e-06,
+      "loss": 0.1022,
+      "step": 21330
+    },
+    {
+      "epoch": 1.8437081515400233,
+      "grad_norm": 0.542883154178525,
+      "learning_rate": 3.4150493819371282e-06,
+      "loss": 0.1029,
+      "step": 21340
+    },
+    {
+      "epoch": 1.8445721197459934,
+      "grad_norm": 0.5485117027957886,
+      "learning_rate": 3.4106267094543068e-06,
+      "loss": 0.1033,
+      "step": 21350
+    },
+    {
+      "epoch": 1.8454360879519633,
+      "grad_norm": 0.547398116269584,
+      "learning_rate": 3.4062054198418143e-06,
+      "loss": 0.1027,
+      "step": 21360
+    },
+    {
+      "epoch": 1.8463000561579332,
+      "grad_norm": 0.5467991450188698,
+      "learning_rate": 3.401785516946495e-06,
+      "loss": 0.1043,
+      "step": 21370
+    },
+    {
+      "epoch": 1.8471640243639034,
+      "grad_norm": 0.5856305997104947,
+      "learning_rate": 3.397367004613985e-06,
+      "loss": 0.1013,
+      "step": 21380
+    },
+    {
+      "epoch": 1.8480279925698735,
+      "grad_norm": 0.5382540415719735,
+      "learning_rate": 3.3929498866887124e-06,
+      "loss": 0.1011,
+      "step": 21390
+    },
+    {
+      "epoch": 1.8488919607758434,
+      "grad_norm": 0.5519626832639604,
+      "learning_rate": 3.3885341670138915e-06,
+      "loss": 0.098,
+      "step": 21400
+    },
+    {
+      "epoch": 1.8497559289818135,
+      "grad_norm": 0.5320256820466911,
+      "learning_rate": 3.384119849431517e-06,
+      "loss": 0.105,
+      "step": 21410
+    },
+    {
+      "epoch": 1.8506198971877836,
+      "grad_norm": 0.568130486279502,
+      "learning_rate": 3.3797069377823676e-06,
+      "loss": 0.1002,
+      "step": 21420
+    },
+    {
+      "epoch": 1.8514838653937535,
+      "grad_norm": 0.5478982214510582,
+      "learning_rate": 3.3752954359059976e-06,
+      "loss": 0.1039,
+      "step": 21430
+    },
+    {
+      "epoch": 1.8523478335997234,
+      "grad_norm": 0.544232492931912,
+      "learning_rate": 3.3708853476407365e-06,
+      "loss": 0.1004,
+      "step": 21440
+    },
+    {
+      "epoch": 1.8532118018056936,
+      "grad_norm": 0.5585103563773959,
+      "learning_rate": 3.366476676823677e-06,
+      "loss": 0.1019,
+      "step": 21450
+    },
+    {
+      "epoch": 1.8540757700116637,
+      "grad_norm": 0.5455231576521419,
+      "learning_rate": 3.3620694272906874e-06,
+      "loss": 0.0977,
+      "step": 21460
+    },
+    {
+      "epoch": 1.8549397382176336,
+      "grad_norm": 0.5379648194855917,
+      "learning_rate": 3.357663602876392e-06,
+      "loss": 0.0979,
+      "step": 21470
+    },
+    {
+      "epoch": 1.8558037064236035,
+      "grad_norm": 0.5726162354617428,
+      "learning_rate": 3.3532592074141823e-06,
+      "loss": 0.0994,
+      "step": 21480
+    },
+    {
+      "epoch": 1.8566676746295736,
+      "grad_norm": 0.5827018315751672,
+      "learning_rate": 3.3488562447361978e-06,
+      "loss": 0.1,
+      "step": 21490
+    },
+    {
+      "epoch": 1.8575316428355437,
+      "grad_norm": 0.5635050703224438,
+      "learning_rate": 3.344454718673339e-06,
+      "loss": 0.1029,
+      "step": 21500
+    },
+    {
+      "epoch": 1.8583956110415136,
+      "grad_norm": 0.5565811080368595,
+      "learning_rate": 3.3400546330552554e-06,
+      "loss": 0.1026,
+      "step": 21510
+    },
+    {
+      "epoch": 1.8592595792474835,
+      "grad_norm": 0.564486228128983,
+      "learning_rate": 3.3356559917103377e-06,
+      "loss": 0.1019,
+      "step": 21520
+    },
+    {
+      "epoch": 1.8601235474534539,
+      "grad_norm": 0.5605995453561945,
+      "learning_rate": 3.3312587984657246e-06,
+      "loss": 0.1006,
+      "step": 21530
+    },
+    {
+      "epoch": 1.8609875156594238,
+      "grad_norm": 0.6100355923575964,
+      "learning_rate": 3.326863057147295e-06,
+      "loss": 0.1039,
+      "step": 21540
+    },
+    {
+      "epoch": 1.8618514838653937,
+      "grad_norm": 0.5762711693846725,
+      "learning_rate": 3.3224687715796656e-06,
+      "loss": 0.1014,
+      "step": 21550
+    },
+    {
+      "epoch": 1.8627154520713638,
+      "grad_norm": 0.5366489167927376,
+      "learning_rate": 3.3180759455861798e-06,
+      "loss": 0.1003,
+      "step": 21560
+    },
+    {
+      "epoch": 1.863579420277334,
+      "grad_norm": 0.5662365857970131,
+      "learning_rate": 3.3136845829889185e-06,
+      "loss": 0.1027,
+      "step": 21570
+    },
+    {
+      "epoch": 1.8644433884833038,
+      "grad_norm": 0.5459962285652393,
+      "learning_rate": 3.309294687608685e-06,
+      "loss": 0.1005,
+      "step": 21580
+    },
+    {
+      "epoch": 1.8653073566892737,
+      "grad_norm": 0.573806154613363,
+      "learning_rate": 3.3049062632650098e-06,
+      "loss": 0.1011,
+      "step": 21590
+    },
+    {
+      "epoch": 1.8661713248952438,
+      "grad_norm": 0.5266529570129646,
+      "learning_rate": 3.3005193137761376e-06,
+      "loss": 0.1023,
+      "step": 21600
+    },
+    {
+      "epoch": 1.867035293101214,
+      "grad_norm": 0.5777135015426488,
+      "learning_rate": 3.2961338429590364e-06,
+      "loss": 0.1022,
+      "step": 21610
+    },
+    {
+      "epoch": 1.8678992613071839,
+      "grad_norm": 0.5710616793282451,
+      "learning_rate": 3.2917498546293823e-06,
+      "loss": 0.1024,
+      "step": 21620
+    },
+    {
+      "epoch": 1.8687632295131538,
+      "grad_norm": 0.5774894136897458,
+      "learning_rate": 3.287367352601568e-06,
+      "loss": 0.1003,
+      "step": 21630
+    },
+    {
+      "epoch": 1.869627197719124,
+      "grad_norm": 0.5741261431440448,
+      "learning_rate": 3.282986340688684e-06,
+      "loss": 0.1028,
+      "step": 21640
+    },
+    {
+      "epoch": 1.870491165925094,
+      "grad_norm": 0.5883458021179613,
+      "learning_rate": 3.278606822702532e-06,
+      "loss": 0.1008,
+      "step": 21650
+    },
+    {
+      "epoch": 1.871355134131064,
+      "grad_norm": 0.574927450088088,
+      "learning_rate": 3.274228802453611e-06,
+      "loss": 0.1003,
+      "step": 21660
+    },
+    {
+      "epoch": 1.872219102337034,
+      "grad_norm": 0.5477269695224006,
+      "learning_rate": 3.269852283751117e-06,
+      "loss": 0.1023,
+      "step": 21670
+    },
+    {
+      "epoch": 1.8730830705430042,
+      "grad_norm": 0.5438420121822475,
+      "learning_rate": 3.2654772704029385e-06,
+      "loss": 0.1005,
+      "step": 21680
+    },
+    {
+      "epoch": 1.873947038748974,
+      "grad_norm": 0.5773522247749471,
+      "learning_rate": 3.2611037662156553e-06,
+      "loss": 0.101,
+      "step": 21690
+    },
+    {
+      "epoch": 1.874811006954944,
+      "grad_norm": 0.5270650909460688,
+      "learning_rate": 3.256731774994534e-06,
+      "loss": 0.1009,
+      "step": 21700
+    },
+    {
+      "epoch": 1.875674975160914,
+      "grad_norm": 0.5697439029107149,
+      "learning_rate": 3.2523613005435273e-06,
+      "loss": 0.1018,
+      "step": 21710
+    },
+    {
+      "epoch": 1.8765389433668842,
+      "grad_norm": 0.5889980436086851,
+      "learning_rate": 3.247992346665262e-06,
+      "loss": 0.103,
+      "step": 21720
+    },
+    {
+      "epoch": 1.877402911572854,
+      "grad_norm": 0.5855966040283817,
+      "learning_rate": 3.243624917161048e-06,
+      "loss": 0.1004,
+      "step": 21730
+    },
+    {
+      "epoch": 1.878266879778824,
+      "grad_norm": 0.5645627450678173,
+      "learning_rate": 3.239259015830865e-06,
+      "loss": 0.1003,
+      "step": 21740
+    },
+    {
+      "epoch": 1.8791308479847941,
+      "grad_norm": 0.5607021768032766,
+      "learning_rate": 3.234894646473368e-06,
+      "loss": 0.1045,
+      "step": 21750
+    },
+    {
+      "epoch": 1.8799948161907643,
+      "grad_norm": 0.6018622097238112,
+      "learning_rate": 3.230531812885872e-06,
+      "loss": 0.1023,
+      "step": 21760
+    },
+    {
+      "epoch": 1.8808587843967342,
+      "grad_norm": 0.5716356881800182,
+      "learning_rate": 3.2261705188643612e-06,
+      "loss": 0.1022,
+      "step": 21770
+    },
+    {
+      "epoch": 1.8817227526027043,
+      "grad_norm": 0.5520205261963792,
+      "learning_rate": 3.2218107682034775e-06,
+      "loss": 0.1013,
+      "step": 21780
+    },
+    {
+      "epoch": 1.8825867208086744,
+      "grad_norm": 0.5495910812136405,
+      "learning_rate": 3.217452564696522e-06,
+      "loss": 0.1014,
+      "step": 21790
+    },
+    {
+      "epoch": 1.8834506890146443,
+      "grad_norm": 0.5624942994289897,
+      "learning_rate": 3.2130959121354475e-06,
+      "loss": 0.1001,
+      "step": 21800
+    },
+    {
+      "epoch": 1.8843146572206142,
+      "grad_norm": 0.6136256619750384,
+      "learning_rate": 3.208740814310859e-06,
+      "loss": 0.1002,
+      "step": 21810
+    },
+    {
+      "epoch": 1.8851786254265843,
+      "grad_norm": 0.5520828631133291,
+      "learning_rate": 3.20438727501201e-06,
+      "loss": 0.1051,
+      "step": 21820
+    },
+    {
+      "epoch": 1.8860425936325544,
+      "grad_norm": 0.6030386575456828,
+      "learning_rate": 3.2000352980267936e-06,
+      "loss": 0.1001,
+      "step": 21830
+    },
+    {
+      "epoch": 1.8869065618385243,
+      "grad_norm": 0.5712910795331343,
+      "learning_rate": 3.195684887141747e-06,
+      "loss": 0.0997,
+      "step": 21840
+    },
+    {
+      "epoch": 1.8877705300444942,
+      "grad_norm": 0.5486849114411699,
+      "learning_rate": 3.191336046142044e-06,
+      "loss": 0.101,
+      "step": 21850
+    },
+    {
+      "epoch": 1.8886344982504644,
+      "grad_norm": 0.6141548602098744,
+      "learning_rate": 3.1869887788114943e-06,
+      "loss": 0.1029,
+      "step": 21860
+    },
+    {
+      "epoch": 1.8894984664564345,
+      "grad_norm": 0.5775840450930885,
+      "learning_rate": 3.1826430889325344e-06,
+      "loss": 0.1029,
+      "step": 21870
+    },
+    {
+      "epoch": 1.8903624346624044,
+      "grad_norm": 0.5540741540255073,
+      "learning_rate": 3.1782989802862313e-06,
+      "loss": 0.0997,
+      "step": 21880
+    },
+    {
+      "epoch": 1.8912264028683743,
+      "grad_norm": 0.5645731722914674,
+      "learning_rate": 3.173956456652274e-06,
+      "loss": 0.1025,
+      "step": 21890
+    },
+    {
+      "epoch": 1.8920903710743444,
+      "grad_norm": 0.5934655084949406,
+      "learning_rate": 3.169615521808978e-06,
+      "loss": 0.0969,
+      "step": 21900
+    },
+    {
+      "epoch": 1.8929543392803145,
+      "grad_norm": 0.5541470186648709,
+      "learning_rate": 3.165276179533267e-06,
+      "loss": 0.1033,
+      "step": 21910
+    },
+    {
+      "epoch": 1.8938183074862844,
+      "grad_norm": 0.5738429556296889,
+      "learning_rate": 3.1609384336006866e-06,
+      "loss": 0.1016,
+      "step": 21920
+    },
+    {
+      "epoch": 1.8946822756922546,
+      "grad_norm": 0.546682642116901,
+      "learning_rate": 3.156602287785391e-06,
+      "loss": 0.1013,
+      "step": 21930
+    },
+    {
+      "epoch": 1.8955462438982247,
+      "grad_norm": 0.5727258889036533,
+      "learning_rate": 3.1522677458601447e-06,
+      "loss": 0.1018,
+      "step": 21940
+    },
+    {
+      "epoch": 1.8964102121041946,
+      "grad_norm": 0.5750188686261928,
+      "learning_rate": 3.1479348115963105e-06,
+      "loss": 0.0991,
+      "step": 21950
+    },
+    {
+      "epoch": 1.8972741803101645,
+      "grad_norm": 0.5675659013009478,
+      "learning_rate": 3.143603488763858e-06,
+      "loss": 0.1002,
+      "step": 21960
+    },
+    {
+      "epoch": 1.8981381485161346,
+      "grad_norm": 0.5432009469281879,
+      "learning_rate": 3.1392737811313535e-06,
+      "loss": 0.1008,
+      "step": 21970
+    },
+    {
+      "epoch": 1.8990021167221047,
+      "grad_norm": 0.5323856224850143,
+      "learning_rate": 3.134945692465957e-06,
+      "loss": 0.1009,
+      "step": 21980
+    },
+    {
+      "epoch": 1.8998660849280746,
+      "grad_norm": 0.5519361955646229,
+      "learning_rate": 3.13061922653342e-06,
+      "loss": 0.1026,
+      "step": 21990
+    },
+    {
+      "epoch": 1.9007300531340445,
+      "grad_norm": 0.5620259952273624,
+      "learning_rate": 3.1262943870980823e-06,
+      "loss": 0.0992,
+      "step": 22000
+    },
+    {
+      "epoch": 1.9015940213400147,
+      "grad_norm": 0.5686870753402948,
+      "learning_rate": 3.121971177922869e-06,
+      "loss": 0.1001,
+      "step": 22010
+    },
+    {
+      "epoch": 1.9024579895459848,
+      "grad_norm": 0.6088252851839825,
+      "learning_rate": 3.1176496027692886e-06,
+      "loss": 0.1027,
+      "step": 22020
+    },
+    {
+      "epoch": 1.9033219577519547,
+      "grad_norm": 0.5844603975272236,
+      "learning_rate": 3.113329665397422e-06,
+      "loss": 0.1015,
+      "step": 22030
+    },
+    {
+      "epoch": 1.9041859259579248,
+      "grad_norm": 0.5606232894350783,
+      "learning_rate": 3.1090113695659297e-06,
+      "loss": 0.1022,
+      "step": 22040
+    },
+    {
+      "epoch": 1.905049894163895,
+      "grad_norm": 0.5659034726528784,
+      "learning_rate": 3.1046947190320463e-06,
+      "loss": 0.1013,
+      "step": 22050
+    },
+    {
+      "epoch": 1.9059138623698648,
+      "grad_norm": 0.5340403917026766,
+      "learning_rate": 3.100379717551567e-06,
+      "loss": 0.0991,
+      "step": 22060
+    },
+    {
+      "epoch": 1.9067778305758347,
+      "grad_norm": 0.5618283207865042,
+      "learning_rate": 3.0960663688788596e-06,
+      "loss": 0.0996,
+      "step": 22070
+    },
+    {
+      "epoch": 1.9076417987818048,
+      "grad_norm": 0.5531777403622569,
+      "learning_rate": 3.0917546767668504e-06,
+      "loss": 0.103,
+      "step": 22080
+    },
+    {
+      "epoch": 1.908505766987775,
+      "grad_norm": 0.5497027735815795,
+      "learning_rate": 3.0874446449670258e-06,
+      "loss": 0.0972,
+      "step": 22090
+    },
+    {
+      "epoch": 1.9093697351937449,
+      "grad_norm": 0.5697514613470495,
+      "learning_rate": 3.083136277229426e-06,
+      "loss": 0.1011,
+      "step": 22100
+    },
+    {
+      "epoch": 1.9102337033997148,
+      "grad_norm": 0.5476742899641973,
+      "learning_rate": 3.0788295773026437e-06,
+      "loss": 0.1014,
+      "step": 22110
+    },
+    {
+      "epoch": 1.9110976716056849,
+      "grad_norm": 0.5677127957502706,
+      "learning_rate": 3.0745245489338217e-06,
+      "loss": 0.1023,
+      "step": 22120
+    },
+    {
+      "epoch": 1.911961639811655,
+      "grad_norm": 0.5983383473334517,
+      "learning_rate": 3.0702211958686503e-06,
+      "loss": 0.1004,
+      "step": 22130
+    },
+    {
+      "epoch": 1.912825608017625,
+      "grad_norm": 0.5395467401934758,
+      "learning_rate": 3.065919521851354e-06,
+      "loss": 0.1011,
+      "step": 22140
+    },
+    {
+      "epoch": 1.9136895762235948,
+      "grad_norm": 0.5372402138956013,
+      "learning_rate": 3.0616195306247054e-06,
+      "loss": 0.0996,
+      "step": 22150
+    },
+    {
+      "epoch": 1.9145535444295652,
+      "grad_norm": 0.5403566249572875,
+      "learning_rate": 3.057321225930009e-06,
+      "loss": 0.0999,
+      "step": 22160
+    },
+    {
+      "epoch": 1.915417512635535,
+      "grad_norm": 0.5544226990298059,
+      "learning_rate": 3.0530246115071036e-06,
+      "loss": 0.1,
+      "step": 22170
+    },
+    {
+      "epoch": 1.916281480841505,
+      "grad_norm": 0.5826205818540283,
+      "learning_rate": 3.0487296910943543e-06,
+      "loss": 0.0996,
+      "step": 22180
+    },
+    {
+      "epoch": 1.917145449047475,
+      "grad_norm": 0.5743500225747563,
+      "learning_rate": 3.044436468428655e-06,
+      "loss": 0.103,
+      "step": 22190
+    },
+    {
+      "epoch": 1.9180094172534452,
+      "grad_norm": 0.5633148475444488,
+      "learning_rate": 3.040144947245422e-06,
+      "loss": 0.1002,
+      "step": 22200
+    },
+    {
+      "epoch": 1.918873385459415,
+      "grad_norm": 0.5352323628595675,
+      "learning_rate": 3.0358551312785912e-06,
+      "loss": 0.0993,
+      "step": 22210
+    },
+    {
+      "epoch": 1.919737353665385,
+      "grad_norm": 0.5599302140701811,
+      "learning_rate": 3.0315670242606123e-06,
+      "loss": 0.1013,
+      "step": 22220
+    },
+    {
+      "epoch": 1.9206013218713551,
+      "grad_norm": 0.5660720954378141,
+      "learning_rate": 3.02728062992245e-06,
+      "loss": 0.1014,
+      "step": 22230
+    },
+    {
+      "epoch": 1.9214652900773252,
+      "grad_norm": 0.5380748855737479,
+      "learning_rate": 3.0229959519935814e-06,
+      "loss": 0.101,
+      "step": 22240
+    },
+    {
+      "epoch": 1.9223292582832952,
+      "grad_norm": 0.5884805058572077,
+      "learning_rate": 3.018712994201989e-06,
+      "loss": 0.102,
+      "step": 22250
+    },
+    {
+      "epoch": 1.923193226489265,
+      "grad_norm": 0.5874042489492537,
+      "learning_rate": 3.0144317602741535e-06,
+      "loss": 0.1006,
+      "step": 22260
+    },
+    {
+      "epoch": 1.9240571946952352,
+      "grad_norm": 0.5327072747468091,
+      "learning_rate": 3.010152253935061e-06,
+      "loss": 0.0965,
+      "step": 22270
+    },
+    {
+      "epoch": 1.9249211629012053,
+      "grad_norm": 0.5672748667596161,
+      "learning_rate": 3.005874478908195e-06,
+      "loss": 0.0992,
+      "step": 22280
+    },
+    {
+      "epoch": 1.9257851311071752,
+      "grad_norm": 0.5779026537091281,
+      "learning_rate": 3.00159843891553e-06,
+      "loss": 0.1008,
+      "step": 22290
+    },
+    {
+      "epoch": 1.9266490993131453,
+      "grad_norm": 0.5775941103917845,
+      "learning_rate": 2.9973241376775314e-06,
+      "loss": 0.102,
+      "step": 22300
+    },
+    {
+      "epoch": 1.9275130675191154,
+      "grad_norm": 0.5614889405016127,
+      "learning_rate": 2.9930515789131507e-06,
+      "loss": 0.1024,
+      "step": 22310
+    },
+    {
+      "epoch": 1.9283770357250853,
+      "grad_norm": 0.5207831073526223,
+      "learning_rate": 2.9887807663398283e-06,
+      "loss": 0.0995,
+      "step": 22320
+    },
+    {
+      "epoch": 1.9292410039310552,
+      "grad_norm": 0.5518036833957318,
+      "learning_rate": 2.9845117036734773e-06,
+      "loss": 0.0981,
+      "step": 22330
+    },
+    {
+      "epoch": 1.9301049721370254,
+      "grad_norm": 0.5848049624329538,
+      "learning_rate": 2.9802443946284944e-06,
+      "loss": 0.1027,
+      "step": 22340
+    },
+    {
+      "epoch": 1.9309689403429955,
+      "grad_norm": 0.5595521149024182,
+      "learning_rate": 2.9759788429177493e-06,
+      "loss": 0.0969,
+      "step": 22350
+    },
+    {
+      "epoch": 1.9318329085489654,
+      "grad_norm": 0.5774209309801077,
+      "learning_rate": 2.9717150522525838e-06,
+      "loss": 0.098,
+      "step": 22360
+    },
+    {
+      "epoch": 1.9326968767549353,
+      "grad_norm": 0.582103892789509,
+      "learning_rate": 2.967453026342802e-06,
+      "loss": 0.1037,
+      "step": 22370
+    },
+    {
+      "epoch": 1.9335608449609054,
+      "grad_norm": 0.596134016903453,
+      "learning_rate": 2.9631927688966783e-06,
+      "loss": 0.1022,
+      "step": 22380
+    },
+    {
+      "epoch": 1.9344248131668755,
+      "grad_norm": 0.5787978494309993,
+      "learning_rate": 2.9589342836209473e-06,
+      "loss": 0.1014,
+      "step": 22390
+    },
+    {
+      "epoch": 1.9352887813728454,
+      "grad_norm": 0.5598636799101697,
+      "learning_rate": 2.9546775742208e-06,
+      "loss": 0.0982,
+      "step": 22400
+    },
+    {
+      "epoch": 1.9361527495788153,
+      "grad_norm": 0.5457627399040831,
+      "learning_rate": 2.9504226443998827e-06,
+      "loss": 0.1003,
+      "step": 22410
+    },
+    {
+      "epoch": 1.9370167177847857,
+      "grad_norm": 0.5273927080875033,
+      "learning_rate": 2.946169497860294e-06,
+      "loss": 0.0977,
+      "step": 22420
+    },
+    {
+      "epoch": 1.9378806859907556,
+      "grad_norm": 0.5515535907865661,
+      "learning_rate": 2.94191813830258e-06,
+      "loss": 0.1016,
+      "step": 22430
+    },
+    {
+      "epoch": 1.9387446541967255,
+      "grad_norm": 0.5858401856027835,
+      "learning_rate": 2.9376685694257358e-06,
+      "loss": 0.098,
+      "step": 22440
+    },
+    {
+      "epoch": 1.9396086224026956,
+      "grad_norm": 0.5326821139702289,
+      "learning_rate": 2.933420794927191e-06,
+      "loss": 0.0999,
+      "step": 22450
+    },
+    {
+      "epoch": 1.9404725906086657,
+      "grad_norm": 0.6165635449377432,
+      "learning_rate": 2.9291748185028197e-06,
+      "loss": 0.099,
+      "step": 22460
+    },
+    {
+      "epoch": 1.9413365588146356,
+      "grad_norm": 0.5420456009991401,
+      "learning_rate": 2.9249306438469305e-06,
+      "loss": 0.0992,
+      "step": 22470
+    },
+    {
+      "epoch": 1.9422005270206055,
+      "grad_norm": 0.57820086837121,
+      "learning_rate": 2.9206882746522665e-06,
+      "loss": 0.1002,
+      "step": 22480
+    },
+    {
+      "epoch": 1.9430644952265756,
+      "grad_norm": 0.5818903495695285,
+      "learning_rate": 2.9164477146099923e-06,
+      "loss": 0.0996,
+      "step": 22490
+    },
+    {
+      "epoch": 1.9439284634325458,
+      "grad_norm": 0.5443267800752883,
+      "learning_rate": 2.9122089674097076e-06,
+      "loss": 0.1014,
+      "step": 22500
+    },
+    {
+      "epoch": 1.9447924316385157,
+      "grad_norm": 0.5500539383572447,
+      "learning_rate": 2.907972036739427e-06,
+      "loss": 0.1022,
+      "step": 22510
+    },
+    {
+      "epoch": 1.9456563998444856,
+      "grad_norm": 0.6347343648821124,
+      "learning_rate": 2.9037369262855947e-06,
+      "loss": 0.1,
+      "step": 22520
+    },
+    {
+      "epoch": 1.9465203680504557,
+      "grad_norm": 0.5514477780439476,
+      "learning_rate": 2.8995036397330563e-06,
+      "loss": 0.0992,
+      "step": 22530
+    },
+    {
+      "epoch": 1.9473843362564258,
+      "grad_norm": 0.586202809665758,
+      "learning_rate": 2.8952721807650854e-06,
+      "loss": 0.105,
+      "step": 22540
+    },
+    {
+      "epoch": 1.9482483044623957,
+      "grad_norm": 0.5545892197299369,
+      "learning_rate": 2.8910425530633556e-06,
+      "loss": 0.0984,
+      "step": 22550
+    },
+    {
+      "epoch": 1.9491122726683658,
+      "grad_norm": 0.5730214745435419,
+      "learning_rate": 2.886814760307952e-06,
+      "loss": 0.0993,
+      "step": 22560
+    },
+    {
+      "epoch": 1.949976240874336,
+      "grad_norm": 0.5458153748028479,
+      "learning_rate": 2.8825888061773623e-06,
+      "loss": 0.0988,
+      "step": 22570
+    },
+    {
+      "epoch": 1.9508402090803059,
+      "grad_norm": 0.5921035939497012,
+      "learning_rate": 2.87836469434847e-06,
+      "loss": 0.0946,
+      "step": 22580
+    },
+    {
+      "epoch": 1.9517041772862758,
+      "grad_norm": 0.5560783693469633,
+      "learning_rate": 2.8741424284965645e-06,
+      "loss": 0.1013,
+      "step": 22590
+    },
+    {
+      "epoch": 1.9525681454922459,
+      "grad_norm": 0.549086003919069,
+      "learning_rate": 2.869922012295322e-06,
+      "loss": 0.1005,
+      "step": 22600
+    },
+    {
+      "epoch": 1.953432113698216,
+      "grad_norm": 0.5793251391076313,
+      "learning_rate": 2.865703449416812e-06,
+      "loss": 0.1035,
+      "step": 22610
+    },
+    {
+      "epoch": 1.954296081904186,
+      "grad_norm": 0.5570124709906242,
+      "learning_rate": 2.8614867435314877e-06,
+      "loss": 0.0954,
+      "step": 22620
+    },
+    {
+      "epoch": 1.9551600501101558,
+      "grad_norm": 0.5329637013463601,
+      "learning_rate": 2.8572718983081972e-06,
+      "loss": 0.0993,
+      "step": 22630
+    },
+    {
+      "epoch": 1.956024018316126,
+      "grad_norm": 0.5332785700544586,
+      "learning_rate": 2.853058917414153e-06,
+      "loss": 0.0956,
+      "step": 22640
+    },
+    {
+      "epoch": 1.956887986522096,
+      "grad_norm": 0.570819195946033,
+      "learning_rate": 2.848847804514962e-06,
+      "loss": 0.1007,
+      "step": 22650
+    },
+    {
+      "epoch": 1.957751954728066,
+      "grad_norm": 0.5495513030265846,
+      "learning_rate": 2.844638563274596e-06,
+      "loss": 0.1002,
+      "step": 22660
+    },
+    {
+      "epoch": 1.958615922934036,
+      "grad_norm": 0.5975360173842512,
+      "learning_rate": 2.840431197355401e-06,
+      "loss": 0.1002,
+      "step": 22670
+    },
+    {
+      "epoch": 1.9594798911400062,
+      "grad_norm": 0.560563297872084,
+      "learning_rate": 2.836225710418091e-06,
+      "loss": 0.0988,
+      "step": 22680
+    },
+    {
+      "epoch": 1.960343859345976,
+      "grad_norm": 0.5717780583220105,
+      "learning_rate": 2.8320221061217446e-06,
+      "loss": 0.1013,
+      "step": 22690
+    },
+    {
+      "epoch": 1.961207827551946,
+      "grad_norm": 0.5525346469001077,
+      "learning_rate": 2.827820388123806e-06,
+      "loss": 0.1004,
+      "step": 22700
+    },
+    {
+      "epoch": 1.9620717957579161,
+      "grad_norm": 0.5827285043087085,
+      "learning_rate": 2.823620560080074e-06,
+      "loss": 0.1011,
+      "step": 22710
+    },
+    {
+      "epoch": 1.9629357639638862,
+      "grad_norm": 0.5752788801952603,
+      "learning_rate": 2.8194226256447033e-06,
+      "loss": 0.099,
+      "step": 22720
+    },
+    {
+      "epoch": 1.9637997321698561,
+      "grad_norm": 0.557491580653571,
+      "learning_rate": 2.815226588470201e-06,
+      "loss": 0.1018,
+      "step": 22730
+    },
+    {
+      "epoch": 1.964663700375826,
+      "grad_norm": 0.5916884024884534,
+      "learning_rate": 2.811032452207429e-06,
+      "loss": 0.0974,
+      "step": 22740
+    },
+    {
+      "epoch": 1.9655276685817962,
+      "grad_norm": 0.5386096553878962,
+      "learning_rate": 2.8068402205055878e-06,
+      "loss": 0.0989,
+      "step": 22750
+    },
+    {
+      "epoch": 1.9663916367877663,
+      "grad_norm": 0.5710691676756725,
+      "learning_rate": 2.8026498970122245e-06,
+      "loss": 0.0955,
+      "step": 22760
+    },
+    {
+      "epoch": 1.9672556049937362,
+      "grad_norm": 0.5556909406918205,
+      "learning_rate": 2.798461485373225e-06,
+      "loss": 0.1005,
+      "step": 22770
+    },
+    {
+      "epoch": 1.968119573199706,
+      "grad_norm": 0.557444256716542,
+      "learning_rate": 2.79427498923281e-06,
+      "loss": 0.0969,
+      "step": 22780
+    },
+    {
+      "epoch": 1.9689835414056762,
+      "grad_norm": 0.5905330627179569,
+      "learning_rate": 2.7900904122335415e-06,
+      "loss": 0.0987,
+      "step": 22790
+    },
+    {
+      "epoch": 1.9698475096116463,
+      "grad_norm": 0.5826168610406894,
+      "learning_rate": 2.7859077580162976e-06,
+      "loss": 0.0997,
+      "step": 22800
+    },
+    {
+      "epoch": 1.9707114778176162,
+      "grad_norm": 0.5600933255055477,
+      "learning_rate": 2.781727030220298e-06,
+      "loss": 0.0986,
+      "step": 22810
+    },
+    {
+      "epoch": 1.9715754460235864,
+      "grad_norm": 0.54806352388108,
+      "learning_rate": 2.777548232483075e-06,
+      "loss": 0.0976,
+      "step": 22820
+    },
+    {
+      "epoch": 1.9724394142295565,
+      "grad_norm": 0.5551221906142857,
+      "learning_rate": 2.773371368440494e-06,
+      "loss": 0.0983,
+      "step": 22830
+    },
+    {
+      "epoch": 1.9733033824355264,
+      "grad_norm": 0.5723114218083482,
+      "learning_rate": 2.7691964417267214e-06,
+      "loss": 0.0996,
+      "step": 22840
+    },
+    {
+      "epoch": 1.9741673506414963,
+      "grad_norm": 0.6044292557590281,
+      "learning_rate": 2.765023455974253e-06,
+      "loss": 0.0982,
+      "step": 22850
+    },
+    {
+      "epoch": 1.9750313188474664,
+      "grad_norm": 0.5746034008271442,
+      "learning_rate": 2.7608524148138878e-06,
+      "loss": 0.0988,
+      "step": 22860
+    },
+    {
+      "epoch": 1.9758952870534365,
+      "grad_norm": 0.6032867504998831,
+      "learning_rate": 2.7566833218747353e-06,
+      "loss": 0.0999,
+      "step": 22870
+    },
+    {
+      "epoch": 1.9767592552594064,
+      "grad_norm": 0.5802365470670267,
+      "learning_rate": 2.7525161807842095e-06,
+      "loss": 0.1009,
+      "step": 22880
+    },
+    {
+      "epoch": 1.9776232234653763,
+      "grad_norm": 0.5781767957801193,
+      "learning_rate": 2.748350995168023e-06,
+      "loss": 0.0985,
+      "step": 22890
+    },
+    {
+      "epoch": 1.9784871916713465,
+      "grad_norm": 0.5519359594377506,
+      "learning_rate": 2.744187768650198e-06,
+      "loss": 0.0989,
+      "step": 22900
+    },
+    {
+      "epoch": 1.9793511598773166,
+      "grad_norm": 0.573944538698511,
+      "learning_rate": 2.7400265048530355e-06,
+      "loss": 0.0983,
+      "step": 22910
+    },
+    {
+      "epoch": 1.9802151280832865,
+      "grad_norm": 0.5723397613793625,
+      "learning_rate": 2.7358672073971433e-06,
+      "loss": 0.098,
+      "step": 22920
+    },
+    {
+      "epoch": 1.9810790962892566,
+      "grad_norm": 0.5510971107060265,
+      "learning_rate": 2.7317098799014086e-06,
+      "loss": 0.1021,
+      "step": 22930
+    },
+    {
+      "epoch": 1.9819430644952267,
+      "grad_norm": 0.5351223179335077,
+      "learning_rate": 2.727554525983015e-06,
+      "loss": 0.0958,
+      "step": 22940
+    },
+    {
+      "epoch": 1.9828070327011966,
+      "grad_norm": 0.5911713108321606,
+      "learning_rate": 2.723401149257414e-06,
+      "loss": 0.0987,
+      "step": 22950
+    },
+    {
+      "epoch": 1.9836710009071665,
+      "grad_norm": 0.555620541087196,
+      "learning_rate": 2.7192497533383517e-06,
+      "loss": 0.0974,
+      "step": 22960
+    },
+    {
+      "epoch": 1.9845349691131366,
+      "grad_norm": 0.531489300609613,
+      "learning_rate": 2.7151003418378426e-06,
+      "loss": 0.0998,
+      "step": 22970
+    },
+    {
+      "epoch": 1.9853989373191068,
+      "grad_norm": 0.56067493893783,
+      "learning_rate": 2.710952918366176e-06,
+      "loss": 0.0986,
+      "step": 22980
+    },
+    {
+      "epoch": 1.9862629055250767,
+      "grad_norm": 0.5115950990313571,
+      "learning_rate": 2.706807486531912e-06,
+      "loss": 0.0993,
+      "step": 22990
+    },
+    {
+      "epoch": 1.9871268737310466,
+      "grad_norm": 0.56036178938333,
+      "learning_rate": 2.7026640499418764e-06,
+      "loss": 0.0959,
+      "step": 23000
+    },
+    {
+      "epoch": 1.9879908419370167,
+      "grad_norm": 0.5820731208668521,
+      "learning_rate": 2.698522612201163e-06,
+      "loss": 0.1015,
+      "step": 23010
+    },
+    {
+      "epoch": 1.9888548101429868,
+      "grad_norm": 0.5561062988471792,
+      "learning_rate": 2.6943831769131235e-06,
+      "loss": 0.0998,
+      "step": 23020
+    },
+    {
+      "epoch": 1.9897187783489567,
+      "grad_norm": 0.550576715080664,
+      "learning_rate": 2.690245747679366e-06,
+      "loss": 0.0995,
+      "step": 23030
+    },
+    {
+      "epoch": 1.9905827465549266,
+      "grad_norm": 0.5607772814673408,
+      "learning_rate": 2.686110328099753e-06,
+      "loss": 0.0981,
+      "step": 23040
+    },
+    {
+      "epoch": 1.991446714760897,
+      "grad_norm": 0.5649321440620626,
+      "learning_rate": 2.681976921772405e-06,
+      "loss": 0.0948,
+      "step": 23050
+    },
+    {
+      "epoch": 1.9923106829668669,
+      "grad_norm": 0.5547901042849294,
+      "learning_rate": 2.6778455322936837e-06,
+      "loss": 0.1017,
+      "step": 23060
+    },
+    {
+      "epoch": 1.9931746511728368,
+      "grad_norm": 0.5433895881280426,
+      "learning_rate": 2.6737161632581983e-06,
+      "loss": 0.0972,
+      "step": 23070
+    },
+    {
+      "epoch": 1.9940386193788069,
+      "grad_norm": 0.5879816117729628,
+      "learning_rate": 2.6695888182588005e-06,
+      "loss": 0.0983,
+      "step": 23080
+    },
+    {
+      "epoch": 1.994902587584777,
+      "grad_norm": 0.5639706915617735,
+      "learning_rate": 2.665463500886577e-06,
+      "loss": 0.0976,
+      "step": 23090
+    },
+    {
+      "epoch": 1.995766555790747,
+      "grad_norm": 0.5597079533515157,
+      "learning_rate": 2.661340214730862e-06,
+      "loss": 0.0992,
+      "step": 23100
+    },
+    {
+      "epoch": 1.9966305239967168,
+      "grad_norm": 0.5975735164456436,
+      "learning_rate": 2.6572189633792052e-06,
+      "loss": 0.0966,
+      "step": 23110
+    },
+    {
+      "epoch": 1.997494492202687,
+      "grad_norm": 0.5431942288707596,
+      "learning_rate": 2.6530997504174007e-06,
+      "loss": 0.099,
+      "step": 23120
+    },
+    {
+      "epoch": 1.998358460408657,
+      "grad_norm": 0.5610255705894158,
+      "learning_rate": 2.64898257942946e-06,
+      "loss": 0.1023,
+      "step": 23130
+    },
+    {
+      "epoch": 1.999222428614627,
+      "grad_norm": 0.5565024056381163,
+      "learning_rate": 2.6448674539976264e-06,
+      "loss": 0.0952,
+      "step": 23140
+    },
+    {
+      "epoch": 2.000086396820597,
+      "grad_norm": 0.5762272252445255,
+      "learning_rate": 2.6407543777023514e-06,
+      "loss": 0.096,
+      "step": 23150
+    },
+    {
+      "epoch": 2.000950365026567,
+      "grad_norm": 0.5941202253662053,
+      "learning_rate": 2.6366433541223156e-06,
+      "loss": 0.0673,
+      "step": 23160
+    },
+    {
+      "epoch": 2.001814333232537,
+      "grad_norm": 0.6016764334805269,
+      "learning_rate": 2.6325343868344064e-06,
+      "loss": 0.0668,
+      "step": 23170
+    },
+    {
+      "epoch": 2.002678301438507,
+      "grad_norm": 0.5768547489229648,
+      "learning_rate": 2.6284274794137243e-06,
+      "loss": 0.0668,
+      "step": 23180
+    },
+    {
+      "epoch": 2.003542269644477,
+      "grad_norm": 0.5653152032508831,
+      "learning_rate": 2.624322635433577e-06,
+      "loss": 0.0636,
+      "step": 23190
+    },
+    {
+      "epoch": 2.0044062378504472,
+      "grad_norm": 0.6356562558227943,
+      "learning_rate": 2.620219858465477e-06,
+      "loss": 0.0653,
+      "step": 23200
+    },
+    {
+      "epoch": 2.005270206056417,
+      "grad_norm": 0.6010359673980143,
+      "learning_rate": 2.616119152079142e-06,
+      "loss": 0.0668,
+      "step": 23210
+    },
+    {
+      "epoch": 2.006134174262387,
+      "grad_norm": 0.5909469980255928,
+      "learning_rate": 2.6120205198424786e-06,
+      "loss": 0.0637,
+      "step": 23220
+    },
+    {
+      "epoch": 2.006998142468357,
+      "grad_norm": 0.5797202598631477,
+      "learning_rate": 2.6079239653215994e-06,
+      "loss": 0.0665,
+      "step": 23230
+    },
+    {
+      "epoch": 2.0078621106743273,
+      "grad_norm": 0.5595299434577446,
+      "learning_rate": 2.603829492080802e-06,
+      "loss": 0.0648,
+      "step": 23240
+    },
+    {
+      "epoch": 2.008726078880297,
+      "grad_norm": 0.5868298303035886,
+      "learning_rate": 2.5997371036825814e-06,
+      "loss": 0.0649,
+      "step": 23250
+    },
+    {
+      "epoch": 2.009590047086267,
+      "grad_norm": 0.6136273885472115,
+      "learning_rate": 2.5956468036876046e-06,
+      "loss": 0.0658,
+      "step": 23260
+    },
+    {
+      "epoch": 2.0104540152922374,
+      "grad_norm": 0.6199336639686098,
+      "learning_rate": 2.591558595654737e-06,
+      "loss": 0.0661,
+      "step": 23270
+    },
+    {
+      "epoch": 2.0113179834982073,
+      "grad_norm": 0.5663941452077751,
+      "learning_rate": 2.587472483141015e-06,
+      "loss": 0.0635,
+      "step": 23280
+    },
+    {
+      "epoch": 2.0121819517041772,
+      "grad_norm": 0.5937353796830521,
+      "learning_rate": 2.5833884697016513e-06,
+      "loss": 0.0652,
+      "step": 23290
+    },
+    {
+      "epoch": 2.013045919910147,
+      "grad_norm": 0.6077597290157493,
+      "learning_rate": 2.5793065588900367e-06,
+      "loss": 0.0635,
+      "step": 23300
+    },
+    {
+      "epoch": 2.0139098881161175,
+      "grad_norm": 0.608614512624362,
+      "learning_rate": 2.575226754257728e-06,
+      "loss": 0.0648,
+      "step": 23310
+    },
+    {
+      "epoch": 2.0147738563220874,
+      "grad_norm": 0.6411985030217807,
+      "learning_rate": 2.571149059354456e-06,
+      "loss": 0.065,
+      "step": 23320
+    },
+    {
+      "epoch": 2.0156378245280573,
+      "grad_norm": 0.5980106388366775,
+      "learning_rate": 2.5670734777281093e-06,
+      "loss": 0.0645,
+      "step": 23330
+    },
+    {
+      "epoch": 2.016501792734027,
+      "grad_norm": 0.6363825389790632,
+      "learning_rate": 2.56300001292474e-06,
+      "loss": 0.0625,
+      "step": 23340
+    },
+    {
+      "epoch": 2.0173657609399975,
+      "grad_norm": 0.6223195747565223,
+      "learning_rate": 2.5589286684885584e-06,
+      "loss": 0.0647,
+      "step": 23350
+    },
+    {
+      "epoch": 2.0182297291459674,
+      "grad_norm": 0.6163821451502053,
+      "learning_rate": 2.5548594479619315e-06,
+      "loss": 0.0653,
+      "step": 23360
+    },
+    {
+      "epoch": 2.0190936973519373,
+      "grad_norm": 0.5917851425986639,
+      "learning_rate": 2.5507923548853775e-06,
+      "loss": 0.0661,
+      "step": 23370
+    },
+    {
+      "epoch": 2.0199576655579077,
+      "grad_norm": 0.6484282253432566,
+      "learning_rate": 2.546727392797561e-06,
+      "loss": 0.0613,
+      "step": 23380
+    },
+    {
+      "epoch": 2.0208216337638776,
+      "grad_norm": 0.6039945969146345,
+      "learning_rate": 2.5426645652352968e-06,
+      "loss": 0.0639,
+      "step": 23390
+    },
+    {
+      "epoch": 2.0216856019698475,
+      "grad_norm": 0.6437986864420733,
+      "learning_rate": 2.5386038757335364e-06,
+      "loss": 0.0646,
+      "step": 23400
+    },
+    {
+      "epoch": 2.0225495701758174,
+      "grad_norm": 0.6132933888487462,
+      "learning_rate": 2.5345453278253818e-06,
+      "loss": 0.0617,
+      "step": 23410
+    },
+    {
+      "epoch": 2.0234135383817877,
+      "grad_norm": 0.6140098088819063,
+      "learning_rate": 2.530488925042056e-06,
+      "loss": 0.0637,
+      "step": 23420
+    },
+    {
+      "epoch": 2.0242775065877576,
+      "grad_norm": 0.6116175505543225,
+      "learning_rate": 2.5264346709129315e-06,
+      "loss": 0.0656,
+      "step": 23430
+    },
+    {
+      "epoch": 2.0251414747937275,
+      "grad_norm": 0.6161269270983905,
+      "learning_rate": 2.5223825689655024e-06,
+      "loss": 0.0633,
+      "step": 23440
+    },
+    {
+      "epoch": 2.0260054429996974,
+      "grad_norm": 0.6097595430473298,
+      "learning_rate": 2.5183326227253915e-06,
+      "loss": 0.064,
+      "step": 23450
+    },
+    {
+      "epoch": 2.0268694112056678,
+      "grad_norm": 0.6039768047331386,
+      "learning_rate": 2.514284835716343e-06,
+      "loss": 0.0651,
+      "step": 23460
+    },
+    {
+      "epoch": 2.0277333794116377,
+      "grad_norm": 0.5964483205003952,
+      "learning_rate": 2.5102392114602326e-06,
+      "loss": 0.0626,
+      "step": 23470
+    },
+    {
+      "epoch": 2.0285973476176076,
+      "grad_norm": 0.6335610652801382,
+      "learning_rate": 2.506195753477043e-06,
+      "loss": 0.0648,
+      "step": 23480
+    },
+    {
+      "epoch": 2.0294613158235775,
+      "grad_norm": 0.6032288219480831,
+      "learning_rate": 2.502154465284879e-06,
+      "loss": 0.0647,
+      "step": 23490
+    },
+    {
+      "epoch": 2.030325284029548,
+      "grad_norm": 0.6081416747954673,
+      "learning_rate": 2.4981153503999536e-06,
+      "loss": 0.0636,
+      "step": 23500
+    },
+    {
+      "epoch": 2.0311892522355177,
+      "grad_norm": 0.6190433175441461,
+      "learning_rate": 2.494078412336589e-06,
+      "loss": 0.0626,
+      "step": 23510
+    },
+    {
+      "epoch": 2.0320532204414876,
+      "grad_norm": 0.6352529516782883,
+      "learning_rate": 2.4900436546072206e-06,
+      "loss": 0.0657,
+      "step": 23520
+    },
+    {
+      "epoch": 2.032917188647458,
+      "grad_norm": 0.60396342566828,
+      "learning_rate": 2.486011080722374e-06,
+      "loss": 0.0638,
+      "step": 23530
+    },
+    {
+      "epoch": 2.033781156853428,
+      "grad_norm": 0.6321000602851499,
+      "learning_rate": 2.4819806941906855e-06,
+      "loss": 0.0632,
+      "step": 23540
+    },
+    {
+      "epoch": 2.0346451250593978,
+      "grad_norm": 0.6408999563026007,
+      "learning_rate": 2.477952498518883e-06,
+      "loss": 0.0672,
+      "step": 23550
+    },
+    {
+      "epoch": 2.0355090932653677,
+      "grad_norm": 0.6099633041788447,
+      "learning_rate": 2.473926497211794e-06,
+      "loss": 0.063,
+      "step": 23560
+    },
+    {
+      "epoch": 2.036373061471338,
+      "grad_norm": 0.6080948793986606,
+      "learning_rate": 2.469902693772325e-06,
+      "loss": 0.0611,
+      "step": 23570
+    },
+    {
+      "epoch": 2.037237029677308,
+      "grad_norm": 0.6341190621707892,
+      "learning_rate": 2.4658810917014823e-06,
+      "loss": 0.0637,
+      "step": 23580
+    },
+    {
+      "epoch": 2.038100997883278,
+      "grad_norm": 0.6540975857228662,
+      "learning_rate": 2.4618616944983525e-06,
+      "loss": 0.0644,
+      "step": 23590
+    },
+    {
+      "epoch": 2.0389649660892477,
+      "grad_norm": 0.6366781118625203,
+      "learning_rate": 2.457844505660102e-06,
+      "loss": 0.0648,
+      "step": 23600
+    },
+    {
+      "epoch": 2.039828934295218,
+      "grad_norm": 0.6317871739227648,
+      "learning_rate": 2.453829528681977e-06,
+      "loss": 0.0639,
+      "step": 23610
+    },
+    {
+      "epoch": 2.040692902501188,
+      "grad_norm": 0.625094462950415,
+      "learning_rate": 2.4498167670572977e-06,
+      "loss": 0.0661,
+      "step": 23620
+    },
+    {
+      "epoch": 2.041556870707158,
+      "grad_norm": 0.6075843535052505,
+      "learning_rate": 2.4458062242774627e-06,
+      "loss": 0.0669,
+      "step": 23630
+    },
+    {
+      "epoch": 2.042420838913128,
+      "grad_norm": 0.6231492470009503,
+      "learning_rate": 2.441797903831934e-06,
+      "loss": 0.0645,
+      "step": 23640
+    },
+    {
+      "epoch": 2.043284807119098,
+      "grad_norm": 0.6013322781301382,
+      "learning_rate": 2.437791809208241e-06,
+      "loss": 0.0653,
+      "step": 23650
+    },
+    {
+      "epoch": 2.044148775325068,
+      "grad_norm": 0.6332588139476704,
+      "learning_rate": 2.433787943891976e-06,
+      "loss": 0.0663,
+      "step": 23660
+    },
+    {
+      "epoch": 2.045012743531038,
+      "grad_norm": 0.5952863252399502,
+      "learning_rate": 2.429786311366796e-06,
+      "loss": 0.065,
+      "step": 23670
+    },
+    {
+      "epoch": 2.0458767117370082,
+      "grad_norm": 0.6412555707079038,
+      "learning_rate": 2.4257869151144128e-06,
+      "loss": 0.064,
+      "step": 23680
+    },
+    {
+      "epoch": 2.046740679942978,
+      "grad_norm": 0.6266314277669499,
+      "learning_rate": 2.421789758614589e-06,
+      "loss": 0.0616,
+      "step": 23690
+    },
+    {
+      "epoch": 2.047604648148948,
+      "grad_norm": 0.6024010395703102,
+      "learning_rate": 2.417794845345142e-06,
+      "loss": 0.0627,
+      "step": 23700
+    },
+    {
+      "epoch": 2.048468616354918,
+      "grad_norm": 0.6333664605446039,
+      "learning_rate": 2.4138021787819365e-06,
+      "loss": 0.063,
+      "step": 23710
+    },
+    {
+      "epoch": 2.0493325845608883,
+      "grad_norm": 0.640445584802518,
+      "learning_rate": 2.409811762398887e-06,
+      "loss": 0.0625,
+      "step": 23720
+    },
+    {
+      "epoch": 2.050196552766858,
+      "grad_norm": 0.6389702484627462,
+      "learning_rate": 2.405823599667939e-06,
+      "loss": 0.0629,
+      "step": 23730
+    },
+    {
+      "epoch": 2.051060520972828,
+      "grad_norm": 0.6205228444403665,
+      "learning_rate": 2.40183769405909e-06,
+      "loss": 0.0629,
+      "step": 23740
+    },
+    {
+      "epoch": 2.0519244891787984,
+      "grad_norm": 0.6340486050763664,
+      "learning_rate": 2.3978540490403645e-06,
+      "loss": 0.062,
+      "step": 23750
+    },
+    {
+      "epoch": 2.0527884573847683,
+      "grad_norm": 0.6367859468897559,
+      "learning_rate": 2.393872668077824e-06,
+      "loss": 0.0622,
+      "step": 23760
+    },
+    {
+      "epoch": 2.0536524255907382,
+      "grad_norm": 0.5904483099597296,
+      "learning_rate": 2.3898935546355584e-06,
+      "loss": 0.0643,
+      "step": 23770
+    },
+    {
+      "epoch": 2.054516393796708,
+      "grad_norm": 0.6290343818423563,
+      "learning_rate": 2.385916712175688e-06,
+      "loss": 0.0635,
+      "step": 23780
+    },
+    {
+      "epoch": 2.0553803620026785,
+      "grad_norm": 0.6312722588263152,
+      "learning_rate": 2.3819421441583545e-06,
+      "loss": 0.0609,
+      "step": 23790
+    },
+    {
+      "epoch": 2.0562443302086484,
+      "grad_norm": 0.634502209125537,
+      "learning_rate": 2.3779698540417202e-06,
+      "loss": 0.0624,
+      "step": 23800
+    },
+    {
+      "epoch": 2.0571082984146183,
+      "grad_norm": 0.6736741707092652,
+      "learning_rate": 2.373999845281966e-06,
+      "loss": 0.0628,
+      "step": 23810
+    },
+    {
+      "epoch": 2.057972266620588,
+      "grad_norm": 0.6568078983053083,
+      "learning_rate": 2.3700321213332873e-06,
+      "loss": 0.0664,
+      "step": 23820
+    },
+    {
+      "epoch": 2.0588362348265585,
+      "grad_norm": 0.6025816101246598,
+      "learning_rate": 2.3660666856478975e-06,
+      "loss": 0.064,
+      "step": 23830
+    },
+    {
+      "epoch": 2.0597002030325284,
+      "grad_norm": 0.6673143783562789,
+      "learning_rate": 2.3621035416760062e-06,
+      "loss": 0.0639,
+      "step": 23840
+    },
+    {
+      "epoch": 2.0605641712384983,
+      "grad_norm": 0.5777579786946239,
+      "learning_rate": 2.3581426928658436e-06,
+      "loss": 0.0637,
+      "step": 23850
+    },
+    {
+      "epoch": 2.0614281394444682,
+      "grad_norm": 0.5945588080678396,
+      "learning_rate": 2.354184142663631e-06,
+      "loss": 0.0631,
+      "step": 23860
+    },
+    {
+      "epoch": 2.0622921076504386,
+      "grad_norm": 0.6272905472891154,
+      "learning_rate": 2.3502278945136007e-06,
+      "loss": 0.0663,
+      "step": 23870
+    },
+    {
+      "epoch": 2.0631560758564085,
+      "grad_norm": 0.5814381114745243,
+      "learning_rate": 2.34627395185797e-06,
+      "loss": 0.0657,
+      "step": 23880
+    },
+    {
+      "epoch": 2.0640200440623784,
+      "grad_norm": 0.6194731552781829,
+      "learning_rate": 2.3423223181369613e-06,
+      "loss": 0.0658,
+      "step": 23890
+    },
+    {
+      "epoch": 2.0648840122683487,
+      "grad_norm": 0.6423636477498306,
+      "learning_rate": 2.3383729967887814e-06,
+      "loss": 0.0651,
+      "step": 23900
+    },
+    {
+      "epoch": 2.0657479804743186,
+      "grad_norm": 0.6494081770950135,
+      "learning_rate": 2.334425991249627e-06,
+      "loss": 0.0648,
+      "step": 23910
+    },
+    {
+      "epoch": 2.0666119486802885,
+      "grad_norm": 0.6163722203310215,
+      "learning_rate": 2.3304813049536802e-06,
+      "loss": 0.0624,
+      "step": 23920
+    },
+    {
+      "epoch": 2.0674759168862584,
+      "grad_norm": 0.6247311437412909,
+      "learning_rate": 2.3265389413331023e-06,
+      "loss": 0.0632,
+      "step": 23930
+    },
+    {
+      "epoch": 2.0683398850922288,
+      "grad_norm": 0.6531618747835833,
+      "learning_rate": 2.3225989038180415e-06,
+      "loss": 0.0634,
+      "step": 23940
+    },
+    {
+      "epoch": 2.0692038532981987,
+      "grad_norm": 0.6648124763866536,
+      "learning_rate": 2.3186611958366135e-06,
+      "loss": 0.0625,
+      "step": 23950
+    },
+    {
+      "epoch": 2.0700678215041686,
+      "grad_norm": 0.615739462809398,
+      "learning_rate": 2.314725820814911e-06,
+      "loss": 0.0641,
+      "step": 23960
+    },
+    {
+      "epoch": 2.0709317897101385,
+      "grad_norm": 0.6131030682970773,
+      "learning_rate": 2.3107927821769954e-06,
+      "loss": 0.0638,
+      "step": 23970
+    },
+    {
+      "epoch": 2.071795757916109,
+      "grad_norm": 0.6084449651691136,
+      "learning_rate": 2.306862083344899e-06,
+      "loss": 0.0641,
+      "step": 23980
+    },
+    {
+      "epoch": 2.0726597261220787,
+      "grad_norm": 0.6477878147729919,
+      "learning_rate": 2.3029337277386136e-06,
+      "loss": 0.0635,
+      "step": 23990
+    },
+    {
+      "epoch": 2.0735236943280486,
+      "grad_norm": 0.6254935837523002,
+      "learning_rate": 2.299007718776096e-06,
+      "loss": 0.0634,
+      "step": 24000
+    },
+    {
+      "epoch": 2.074387662534019,
+      "grad_norm": 0.634446437880187,
+      "learning_rate": 2.2950840598732572e-06,
+      "loss": 0.0618,
+      "step": 24010
+    },
+    {
+      "epoch": 2.075251630739989,
+      "grad_norm": 0.627483584074575,
+      "learning_rate": 2.2911627544439668e-06,
+      "loss": 0.0631,
+      "step": 24020
+    },
+    {
+      "epoch": 2.0761155989459588,
+      "grad_norm": 0.6316220677935596,
+      "learning_rate": 2.287243805900046e-06,
+      "loss": 0.0646,
+      "step": 24030
+    },
+    {
+      "epoch": 2.0769795671519287,
+      "grad_norm": 0.638069305465654,
+      "learning_rate": 2.2833272176512626e-06,
+      "loss": 0.0627,
+      "step": 24040
+    },
+    {
+      "epoch": 2.077843535357899,
+      "grad_norm": 0.6679409341830631,
+      "learning_rate": 2.2794129931053368e-06,
+      "loss": 0.0614,
+      "step": 24050
+    },
+    {
+      "epoch": 2.078707503563869,
+      "grad_norm": 0.6324265674564412,
+      "learning_rate": 2.275501135667927e-06,
+      "loss": 0.0622,
+      "step": 24060
+    },
+    {
+      "epoch": 2.079571471769839,
+      "grad_norm": 0.6643281873207468,
+      "learning_rate": 2.2715916487426338e-06,
+      "loss": 0.0615,
+      "step": 24070
+    },
+    {
+      "epoch": 2.0804354399758087,
+      "grad_norm": 0.647967935021621,
+      "learning_rate": 2.2676845357309922e-06,
+      "loss": 0.0631,
+      "step": 24080
+    },
+    {
+      "epoch": 2.081299408181779,
+      "grad_norm": 0.6292387792234486,
+      "learning_rate": 2.26377980003248e-06,
+      "loss": 0.0619,
+      "step": 24090
+    },
+    {
+      "epoch": 2.082163376387749,
+      "grad_norm": 0.5903159353895882,
+      "learning_rate": 2.2598774450444976e-06,
+      "loss": 0.0641,
+      "step": 24100
+    },
+    {
+      "epoch": 2.083027344593719,
+      "grad_norm": 0.664983677188501,
+      "learning_rate": 2.255977474162379e-06,
+      "loss": 0.0641,
+      "step": 24110
+    },
+    {
+      "epoch": 2.0838913127996888,
+      "grad_norm": 0.593897600532623,
+      "learning_rate": 2.252079890779382e-06,
+      "loss": 0.0635,
+      "step": 24120
+    },
+    {
+      "epoch": 2.084755281005659,
+      "grad_norm": 0.6117589894042372,
+      "learning_rate": 2.2481846982866843e-06,
+      "loss": 0.061,
+      "step": 24130
+    },
+    {
+      "epoch": 2.085619249211629,
+      "grad_norm": 0.6177291800970711,
+      "learning_rate": 2.2442919000733947e-06,
+      "loss": 0.0626,
+      "step": 24140
+    },
+    {
+      "epoch": 2.086483217417599,
+      "grad_norm": 0.6117096845706067,
+      "learning_rate": 2.240401499526522e-06,
+      "loss": 0.0635,
+      "step": 24150
+    },
+    {
+      "epoch": 2.0873471856235692,
+      "grad_norm": 0.6329557650601069,
+      "learning_rate": 2.2365135000310018e-06,
+      "loss": 0.0605,
+      "step": 24160
+    },
+    {
+      "epoch": 2.088211153829539,
+      "grad_norm": 0.5882562424900543,
+      "learning_rate": 2.232627904969677e-06,
+      "loss": 0.0622,
+      "step": 24170
+    },
+    {
+      "epoch": 2.089075122035509,
+      "grad_norm": 0.6579717668444609,
+      "learning_rate": 2.2287447177232957e-06,
+      "loss": 0.0622,
+      "step": 24180
+    },
+    {
+      "epoch": 2.089939090241479,
+      "grad_norm": 0.6186116937347731,
+      "learning_rate": 2.224863941670513e-06,
+      "loss": 0.0629,
+      "step": 24190
+    },
+    {
+      "epoch": 2.0908030584474493,
+      "grad_norm": 0.6228935783095523,
+      "learning_rate": 2.2209855801878886e-06,
+      "loss": 0.0613,
+      "step": 24200
+    },
+    {
+      "epoch": 2.091667026653419,
+      "grad_norm": 0.6404729334159126,
+      "learning_rate": 2.2171096366498785e-06,
+      "loss": 0.0624,
+      "step": 24210
+    },
+    {
+      "epoch": 2.092530994859389,
+      "grad_norm": 0.6022716371431075,
+      "learning_rate": 2.213236114428834e-06,
+      "loss": 0.0625,
+      "step": 24220
+    },
+    {
+      "epoch": 2.093394963065359,
+      "grad_norm": 0.6422295848132793,
+      "learning_rate": 2.2093650168950022e-06,
+      "loss": 0.0611,
+      "step": 24230
+    },
+    {
+      "epoch": 2.0942589312713293,
+      "grad_norm": 0.6411305363281238,
+      "learning_rate": 2.2054963474165165e-06,
+      "loss": 0.0631,
+      "step": 24240
+    },
+    {
+      "epoch": 2.0951228994772992,
+      "grad_norm": 0.6223287089938836,
+      "learning_rate": 2.2016301093594055e-06,
+      "loss": 0.0631,
+      "step": 24250
+    },
+    {
+      "epoch": 2.095986867683269,
+      "grad_norm": 0.6423015109263637,
+      "learning_rate": 2.1977663060875746e-06,
+      "loss": 0.0633,
+      "step": 24260
+    },
+    {
+      "epoch": 2.0968508358892395,
+      "grad_norm": 0.623734559704998,
+      "learning_rate": 2.1939049409628143e-06,
+      "loss": 0.0635,
+      "step": 24270
+    },
+    {
+      "epoch": 2.0977148040952094,
+      "grad_norm": 0.655630654823313,
+      "learning_rate": 2.1900460173447925e-06,
+      "loss": 0.062,
+      "step": 24280
+    },
+    {
+      "epoch": 2.0985787723011793,
+      "grad_norm": 0.684247340874467,
+      "learning_rate": 2.1861895385910535e-06,
+      "loss": 0.0638,
+      "step": 24290
+    },
+    {
+      "epoch": 2.099442740507149,
+      "grad_norm": 0.6820724809192963,
+      "learning_rate": 2.1823355080570145e-06,
+      "loss": 0.0621,
+      "step": 24300
+    },
+    {
+      "epoch": 2.1003067087131195,
+      "grad_norm": 0.6534096381485632,
+      "learning_rate": 2.178483929095961e-06,
+      "loss": 0.0636,
+      "step": 24310
+    },
+    {
+      "epoch": 2.1011706769190894,
+      "grad_norm": 0.6510147646848921,
+      "learning_rate": 2.1746348050590486e-06,
+      "loss": 0.0614,
+      "step": 24320
+    },
+    {
+      "epoch": 2.1020346451250593,
+      "grad_norm": 0.6041154097070152,
+      "learning_rate": 2.170788139295295e-06,
+      "loss": 0.0634,
+      "step": 24330
+    },
+    {
+      "epoch": 2.1028986133310292,
+      "grad_norm": 0.6120032247248519,
+      "learning_rate": 2.166943935151578e-06,
+      "loss": 0.0624,
+      "step": 24340
+    },
+    {
+      "epoch": 2.1037625815369996,
+      "grad_norm": 0.6462962519025485,
+      "learning_rate": 2.163102195972634e-06,
+      "loss": 0.0668,
+      "step": 24350
+    },
+    {
+      "epoch": 2.1046265497429695,
+      "grad_norm": 0.7056923665766052,
+      "learning_rate": 2.159262925101058e-06,
+      "loss": 0.0636,
+      "step": 24360
+    },
+    {
+      "epoch": 2.1054905179489394,
+      "grad_norm": 0.643509016704401,
+      "learning_rate": 2.1554261258772936e-06,
+      "loss": 0.0661,
+      "step": 24370
+    },
+    {
+      "epoch": 2.1063544861549097,
+      "grad_norm": 0.6549146285728764,
+      "learning_rate": 2.151591801639635e-06,
+      "loss": 0.0619,
+      "step": 24380
+    },
+    {
+      "epoch": 2.1072184543608796,
+      "grad_norm": 0.6581508720395497,
+      "learning_rate": 2.147759955724223e-06,
+      "loss": 0.0618,
+      "step": 24390
+    },
+    {
+      "epoch": 2.1080824225668495,
+      "grad_norm": 0.6517376056398865,
+      "learning_rate": 2.1439305914650398e-06,
+      "loss": 0.0612,
+      "step": 24400
+    },
+    {
+      "epoch": 2.1089463907728194,
+      "grad_norm": 0.6663625422110055,
+      "learning_rate": 2.1401037121939176e-06,
+      "loss": 0.0629,
+      "step": 24410
+    },
+    {
+      "epoch": 2.1098103589787898,
+      "grad_norm": 0.6199637112899523,
+      "learning_rate": 2.1362793212405112e-06,
+      "loss": 0.0606,
+      "step": 24420
+    },
+    {
+      "epoch": 2.1106743271847597,
+      "grad_norm": 0.6068817093511195,
+      "learning_rate": 2.1324574219323244e-06,
+      "loss": 0.0633,
+      "step": 24430
+    },
+    {
+      "epoch": 2.1115382953907296,
+      "grad_norm": 0.6391262425524984,
+      "learning_rate": 2.128638017594685e-06,
+      "loss": 0.0634,
+      "step": 24440
+    },
+    {
+      "epoch": 2.1124022635966995,
+      "grad_norm": 0.6545074639548908,
+      "learning_rate": 2.124821111550756e-06,
+      "loss": 0.0635,
+      "step": 24450
+    },
+    {
+      "epoch": 2.11326623180267,
+      "grad_norm": 0.6283092609786927,
+      "learning_rate": 2.1210067071215174e-06,
+      "loss": 0.0607,
+      "step": 24460
+    },
+    {
+      "epoch": 2.1141302000086397,
+      "grad_norm": 0.6265855386605454,
+      "learning_rate": 2.117194807625784e-06,
+      "loss": 0.0606,
+      "step": 24470
+    },
+    {
+      "epoch": 2.1149941682146096,
+      "grad_norm": 0.668275663546584,
+      "learning_rate": 2.113385416380182e-06,
+      "loss": 0.0641,
+      "step": 24480
+    },
+    {
+      "epoch": 2.1158581364205795,
+      "grad_norm": 0.6468286051725382,
+      "learning_rate": 2.1095785366991602e-06,
+      "loss": 0.0612,
+      "step": 24490
+    },
+    {
+      "epoch": 2.11672210462655,
+      "grad_norm": 0.6338607054600975,
+      "learning_rate": 2.1057741718949803e-06,
+      "loss": 0.0625,
+      "step": 24500
+    },
+    {
+      "epoch": 2.1175860728325198,
+      "grad_norm": 0.6284795065484166,
+      "learning_rate": 2.1019723252777152e-06,
+      "loss": 0.0604,
+      "step": 24510
+    },
+    {
+      "epoch": 2.1184500410384897,
+      "grad_norm": 0.6653757272633931,
+      "learning_rate": 2.09817300015525e-06,
+      "loss": 0.0639,
+      "step": 24520
+    },
+    {
+      "epoch": 2.11931400924446,
+      "grad_norm": 0.6132171778029453,
+      "learning_rate": 2.094376199833274e-06,
+      "loss": 0.0595,
+      "step": 24530
+    },
+    {
+      "epoch": 2.12017797745043,
+      "grad_norm": 0.6337994827967848,
+      "learning_rate": 2.0905819276152777e-06,
+      "loss": 0.0624,
+      "step": 24540
+    },
+    {
+      "epoch": 2.1210419456564,
+      "grad_norm": 0.6200981039293022,
+      "learning_rate": 2.086790186802554e-06,
+      "loss": 0.0627,
+      "step": 24550
+    },
+    {
+      "epoch": 2.1219059138623697,
+      "grad_norm": 0.607923399933756,
+      "learning_rate": 2.0830009806941943e-06,
+      "loss": 0.0633,
+      "step": 24560
+    },
+    {
+      "epoch": 2.12276988206834,
+      "grad_norm": 0.6327477772847702,
+      "learning_rate": 2.079214312587083e-06,
+      "loss": 0.0628,
+      "step": 24570
+    },
+    {
+      "epoch": 2.12363385027431,
+      "grad_norm": 0.6290055999012083,
+      "learning_rate": 2.0754301857758968e-06,
+      "loss": 0.0612,
+      "step": 24580
+    },
+    {
+      "epoch": 2.12449781848028,
+      "grad_norm": 0.6093290964229778,
+      "learning_rate": 2.0716486035531007e-06,
+      "loss": 0.0608,
+      "step": 24590
+    },
+    {
+      "epoch": 2.1253617866862498,
+      "grad_norm": 0.6269656092540906,
+      "learning_rate": 2.0678695692089467e-06,
+      "loss": 0.0602,
+      "step": 24600
+    },
+    {
+      "epoch": 2.12622575489222,
+      "grad_norm": 0.6486656468551327,
+      "learning_rate": 2.064093086031469e-06,
+      "loss": 0.0598,
+      "step": 24610
+    },
+    {
+      "epoch": 2.12708972309819,
+      "grad_norm": 0.631027633175365,
+      "learning_rate": 2.0603191573064814e-06,
+      "loss": 0.0631,
+      "step": 24620
+    },
+    {
+      "epoch": 2.12795369130416,
+      "grad_norm": 0.6583211890310452,
+      "learning_rate": 2.0565477863175785e-06,
+      "loss": 0.063,
+      "step": 24630
+    },
+    {
+      "epoch": 2.12881765951013,
+      "grad_norm": 0.6045310854163645,
+      "learning_rate": 2.052778976346127e-06,
+      "loss": 0.0634,
+      "step": 24640
+    },
+    {
+      "epoch": 2.1296816277161,
+      "grad_norm": 0.6640903285647783,
+      "learning_rate": 2.0490127306712656e-06,
+      "loss": 0.064,
+      "step": 24650
+    },
+    {
+      "epoch": 2.13054559592207,
+      "grad_norm": 0.6438233531716371,
+      "learning_rate": 2.0452490525698986e-06,
+      "loss": 0.0646,
+      "step": 24660
+    },
+    {
+      "epoch": 2.13140956412804,
+      "grad_norm": 0.617131494530124,
+      "learning_rate": 2.041487945316705e-06,
+      "loss": 0.0629,
+      "step": 24670
+    },
+    {
+      "epoch": 2.1322735323340103,
+      "grad_norm": 0.6065166880037826,
+      "learning_rate": 2.037729412184118e-06,
+      "loss": 0.0639,
+      "step": 24680
+    },
+    {
+      "epoch": 2.13313750053998,
+      "grad_norm": 0.6478988234190833,
+      "learning_rate": 2.0339734564423365e-06,
+      "loss": 0.0632,
+      "step": 24690
+    },
+    {
+      "epoch": 2.13400146874595,
+      "grad_norm": 0.6298992819336114,
+      "learning_rate": 2.030220081359314e-06,
+      "loss": 0.061,
+      "step": 24700
+    },
+    {
+      "epoch": 2.13486543695192,
+      "grad_norm": 0.650993251243365,
+      "learning_rate": 2.0264692902007576e-06,
+      "loss": 0.06,
+      "step": 24710
+    },
+    {
+      "epoch": 2.1357294051578903,
+      "grad_norm": 0.6637271378066566,
+      "learning_rate": 2.022721086230135e-06,
+      "loss": 0.0618,
+      "step": 24720
+    },
+    {
+      "epoch": 2.1365933733638602,
+      "grad_norm": 0.6671955218278872,
+      "learning_rate": 2.018975472708648e-06,
+      "loss": 0.0624,
+      "step": 24730
+    },
+    {
+      "epoch": 2.13745734156983,
+      "grad_norm": 0.6433645148089315,
+      "learning_rate": 2.015232452895258e-06,
+      "loss": 0.0612,
+      "step": 24740
+    },
+    {
+      "epoch": 2.1383213097758,
+      "grad_norm": 0.6355581302816811,
+      "learning_rate": 2.0114920300466624e-06,
+      "loss": 0.0628,
+      "step": 24750
+    },
+    {
+      "epoch": 2.1391852779817704,
+      "grad_norm": 0.6174139399477326,
+      "learning_rate": 2.0077542074173044e-06,
+      "loss": 0.0632,
+      "step": 24760
+    },
+    {
+      "epoch": 2.1400492461877403,
+      "grad_norm": 0.667036387076024,
+      "learning_rate": 2.0040189882593556e-06,
+      "loss": 0.0626,
+      "step": 24770
+    },
+    {
+      "epoch": 2.14091321439371,
+      "grad_norm": 0.6256015668288021,
+      "learning_rate": 2.0002863758227332e-06,
+      "loss": 0.0638,
+      "step": 24780
+    },
+    {
+      "epoch": 2.1417771825996805,
+      "grad_norm": 0.6350675918812858,
+      "learning_rate": 1.9965563733550797e-06,
+      "loss": 0.0615,
+      "step": 24790
+    },
+    {
+      "epoch": 2.1426411508056504,
+      "grad_norm": 0.6560277084131226,
+      "learning_rate": 1.992828984101768e-06,
+      "loss": 0.0607,
+      "step": 24800
+    },
+    {
+      "epoch": 2.1435051190116203,
+      "grad_norm": 0.622302271239541,
+      "learning_rate": 1.989104211305898e-06,
+      "loss": 0.0607,
+      "step": 24810
+    },
+    {
+      "epoch": 2.1443690872175902,
+      "grad_norm": 0.6220514071659232,
+      "learning_rate": 1.985382058208292e-06,
+      "loss": 0.0643,
+      "step": 24820
+    },
+    {
+      "epoch": 2.1452330554235606,
+      "grad_norm": 0.6285288781254312,
+      "learning_rate": 1.9816625280474966e-06,
+      "loss": 0.0595,
+      "step": 24830
+    },
+    {
+      "epoch": 2.1460970236295305,
+      "grad_norm": 0.5995374132946757,
+      "learning_rate": 1.9779456240597718e-06,
+      "loss": 0.0628,
+      "step": 24840
+    },
+    {
+      "epoch": 2.1469609918355004,
+      "grad_norm": 0.6358530584409989,
+      "learning_rate": 1.9742313494790943e-06,
+      "loss": 0.0637,
+      "step": 24850
+    },
+    {
+      "epoch": 2.1478249600414703,
+      "grad_norm": 0.6407208020987373,
+      "learning_rate": 1.970519707537151e-06,
+      "loss": 0.0616,
+      "step": 24860
+    },
+    {
+      "epoch": 2.1486889282474406,
+      "grad_norm": 0.6356132639497752,
+      "learning_rate": 1.9668107014633458e-06,
+      "loss": 0.0626,
+      "step": 24870
+    },
+    {
+      "epoch": 2.1495528964534105,
+      "grad_norm": 0.654976860526239,
+      "learning_rate": 1.963104334484777e-06,
+      "loss": 0.0613,
+      "step": 24880
+    },
+    {
+      "epoch": 2.1504168646593804,
+      "grad_norm": 0.6227712775313601,
+      "learning_rate": 1.9594006098262584e-06,
+      "loss": 0.0625,
+      "step": 24890
+    },
+    {
+      "epoch": 2.1512808328653508,
+      "grad_norm": 0.661425967242727,
+      "learning_rate": 1.955699530710298e-06,
+      "loss": 0.066,
+      "step": 24900
+    },
+    {
+      "epoch": 2.1521448010713207,
+      "grad_norm": 0.6501765614101469,
+      "learning_rate": 1.952001100357104e-06,
+      "loss": 0.0617,
+      "step": 24910
+    },
+    {
+      "epoch": 2.1530087692772906,
+      "grad_norm": 0.6565869828321924,
+      "learning_rate": 1.9483053219845786e-06,
+      "loss": 0.0609,
+      "step": 24920
+    },
+    {
+      "epoch": 2.1538727374832605,
+      "grad_norm": 0.6170406429516456,
+      "learning_rate": 1.9446121988083176e-06,
+      "loss": 0.0618,
+      "step": 24930
+    },
+    {
+      "epoch": 2.154736705689231,
+      "grad_norm": 0.6332346071442021,
+      "learning_rate": 1.9409217340416094e-06,
+      "loss": 0.0659,
+      "step": 24940
+    },
+    {
+      "epoch": 2.1556006738952007,
+      "grad_norm": 0.6310607589991424,
+      "learning_rate": 1.9372339308954243e-06,
+      "loss": 0.0617,
+      "step": 24950
+    },
+    {
+      "epoch": 2.1564646421011706,
+      "grad_norm": 0.631341824342489,
+      "learning_rate": 1.9335487925784203e-06,
+      "loss": 0.0619,
+      "step": 24960
+    },
+    {
+      "epoch": 2.1573286103071405,
+      "grad_norm": 0.6055231058000567,
+      "learning_rate": 1.9298663222969333e-06,
+      "loss": 0.0611,
+      "step": 24970
+    },
+    {
+      "epoch": 2.158192578513111,
+      "grad_norm": 0.6519031072151518,
+      "learning_rate": 1.926186523254984e-06,
+      "loss": 0.0623,
+      "step": 24980
+    },
+    {
+      "epoch": 2.1590565467190808,
+      "grad_norm": 0.6532029733553798,
+      "learning_rate": 1.9225093986542633e-06,
+      "loss": 0.0628,
+      "step": 24990
+    },
+    {
+      "epoch": 2.1599205149250507,
+      "grad_norm": 0.6371257726134606,
+      "learning_rate": 1.9188349516941363e-06,
+      "loss": 0.0617,
+      "step": 25000
+    },
+    {
+      "epoch": 2.160784483131021,
+      "grad_norm": 0.6517130228940046,
+      "learning_rate": 1.91516318557164e-06,
+      "loss": 0.062,
+      "step": 25010
+    },
+    {
+      "epoch": 2.161648451336991,
+      "grad_norm": 0.6314055320125311,
+      "learning_rate": 1.911494103481476e-06,
+      "loss": 0.0624,
+      "step": 25020
+    },
+    {
+      "epoch": 2.162512419542961,
+      "grad_norm": 0.644728408364859,
+      "learning_rate": 1.907827708616018e-06,
+      "loss": 0.0611,
+      "step": 25030
+    },
+    {
+      "epoch": 2.1633763877489307,
+      "grad_norm": 0.6783305329471674,
+      "learning_rate": 1.904164004165288e-06,
+      "loss": 0.063,
+      "step": 25040
+    },
+    {
+      "epoch": 2.164240355954901,
+      "grad_norm": 0.6226296409212154,
+      "learning_rate": 1.9005029933169815e-06,
+      "loss": 0.0613,
+      "step": 25050
+    },
+    {
+      "epoch": 2.165104324160871,
+      "grad_norm": 0.6627104553870933,
+      "learning_rate": 1.8968446792564405e-06,
+      "loss": 0.0611,
+      "step": 25060
+    },
+    {
+      "epoch": 2.165968292366841,
+      "grad_norm": 0.6672474621528279,
+      "learning_rate": 1.893189065166669e-06,
+      "loss": 0.0611,
+      "step": 25070
+    },
+    {
+      "epoch": 2.1668322605728108,
+      "grad_norm": 0.6591423831765176,
+      "learning_rate": 1.8895361542283103e-06,
+      "loss": 0.0618,
+      "step": 25080
+    },
+    {
+      "epoch": 2.167696228778781,
+      "grad_norm": 0.6253784506850278,
+      "learning_rate": 1.8858859496196685e-06,
+      "loss": 0.0596,
+      "step": 25090
+    },
+    {
+      "epoch": 2.168560196984751,
+      "grad_norm": 0.6112277593734251,
+      "learning_rate": 1.8822384545166845e-06,
+      "loss": 0.0614,
+      "step": 25100
+    },
+    {
+      "epoch": 2.169424165190721,
+      "grad_norm": 0.6230633387448746,
+      "learning_rate": 1.8785936720929437e-06,
+      "loss": 0.0621,
+      "step": 25110
+    },
+    {
+      "epoch": 2.170288133396691,
+      "grad_norm": 0.6411573873586922,
+      "learning_rate": 1.874951605519673e-06,
+      "loss": 0.0615,
+      "step": 25120
+    },
+    {
+      "epoch": 2.171152101602661,
+      "grad_norm": 0.6496078251100311,
+      "learning_rate": 1.8713122579657333e-06,
+      "loss": 0.0624,
+      "step": 25130
+    },
+    {
+      "epoch": 2.172016069808631,
+      "grad_norm": 0.6297655237072545,
+      "learning_rate": 1.8676756325976265e-06,
+      "loss": 0.0625,
+      "step": 25140
+    },
+    {
+      "epoch": 2.172880038014601,
+      "grad_norm": 0.6170767552073277,
+      "learning_rate": 1.8640417325794735e-06,
+      "loss": 0.0604,
+      "step": 25150
+    },
+    {
+      "epoch": 2.1737440062205713,
+      "grad_norm": 0.6272151430305527,
+      "learning_rate": 1.8604105610730378e-06,
+      "loss": 0.0613,
+      "step": 25160
+    },
+    {
+      "epoch": 2.174607974426541,
+      "grad_norm": 0.6675092753122248,
+      "learning_rate": 1.8567821212376986e-06,
+      "loss": 0.0639,
+      "step": 25170
+    },
+    {
+      "epoch": 2.175471942632511,
+      "grad_norm": 0.6224564596460632,
+      "learning_rate": 1.8531564162304677e-06,
+      "loss": 0.0627,
+      "step": 25180
+    },
+    {
+      "epoch": 2.176335910838481,
+      "grad_norm": 0.6068390661175612,
+      "learning_rate": 1.8495334492059653e-06,
+      "loss": 0.0621,
+      "step": 25190
+    },
+    {
+      "epoch": 2.1771998790444513,
+      "grad_norm": 0.6523381112434008,
+      "learning_rate": 1.8459132233164407e-06,
+      "loss": 0.0622,
+      "step": 25200
+    },
+    {
+      "epoch": 2.1780638472504212,
+      "grad_norm": 0.6416349928882249,
+      "learning_rate": 1.8422957417117531e-06,
+      "loss": 0.0631,
+      "step": 25210
+    },
+    {
+      "epoch": 2.178927815456391,
+      "grad_norm": 0.6530937293777992,
+      "learning_rate": 1.8386810075393735e-06,
+      "loss": 0.061,
+      "step": 25220
+    },
+    {
+      "epoch": 2.179791783662361,
+      "grad_norm": 0.6660230625168588,
+      "learning_rate": 1.8350690239443841e-06,
+      "loss": 0.0602,
+      "step": 25230
+    },
+    {
+      "epoch": 2.1806557518683314,
+      "grad_norm": 0.6125379807881822,
+      "learning_rate": 1.8314597940694711e-06,
+      "loss": 0.0613,
+      "step": 25240
+    },
+    {
+      "epoch": 2.1815197200743013,
+      "grad_norm": 0.6556372780382884,
+      "learning_rate": 1.8278533210549304e-06,
+      "loss": 0.0648,
+      "step": 25250
+    },
+    {
+      "epoch": 2.182383688280271,
+      "grad_norm": 0.6875735101398713,
+      "learning_rate": 1.8242496080386546e-06,
+      "loss": 0.0624,
+      "step": 25260
+    },
+    {
+      "epoch": 2.183247656486241,
+      "grad_norm": 0.6386277390230637,
+      "learning_rate": 1.8206486581561356e-06,
+      "loss": 0.0639,
+      "step": 25270
+    },
+    {
+      "epoch": 2.1841116246922114,
+      "grad_norm": 0.6383759095811737,
+      "learning_rate": 1.8170504745404598e-06,
+      "loss": 0.0635,
+      "step": 25280
+    },
+    {
+      "epoch": 2.1849755928981813,
+      "grad_norm": 0.6291029502412889,
+      "learning_rate": 1.8134550603223123e-06,
+      "loss": 0.0615,
+      "step": 25290
+    },
+    {
+      "epoch": 2.1858395611041512,
+      "grad_norm": 0.6597481436350051,
+      "learning_rate": 1.8098624186299628e-06,
+      "loss": 0.0637,
+      "step": 25300
+    },
+    {
+      "epoch": 2.1867035293101216,
+      "grad_norm": 0.6060866620443429,
+      "learning_rate": 1.8062725525892716e-06,
+      "loss": 0.0593,
+      "step": 25310
+    },
+    {
+      "epoch": 2.1875674975160915,
+      "grad_norm": 0.6594341783898489,
+      "learning_rate": 1.802685465323682e-06,
+      "loss": 0.062,
+      "step": 25320
+    },
+    {
+      "epoch": 2.1884314657220614,
+      "grad_norm": 0.6967919531607281,
+      "learning_rate": 1.7991011599542202e-06,
+      "loss": 0.062,
+      "step": 25330
+    },
+    {
+      "epoch": 2.1892954339280313,
+      "grad_norm": 0.6355968136614126,
+      "learning_rate": 1.7955196395994967e-06,
+      "loss": 0.0619,
+      "step": 25340
+    },
+    {
+      "epoch": 2.1901594021340016,
+      "grad_norm": 0.6493195599762654,
+      "learning_rate": 1.7919409073756883e-06,
+      "loss": 0.0602,
+      "step": 25350
+    },
+    {
+      "epoch": 2.1910233703399715,
+      "grad_norm": 0.6687212211442863,
+      "learning_rate": 1.7883649663965574e-06,
+      "loss": 0.0619,
+      "step": 25360
+    },
+    {
+      "epoch": 2.1918873385459414,
+      "grad_norm": 0.6783993190646149,
+      "learning_rate": 1.7847918197734293e-06,
+      "loss": 0.0627,
+      "step": 25370
+    },
+    {
+      "epoch": 2.1927513067519113,
+      "grad_norm": 0.6823882867138366,
+      "learning_rate": 1.781221470615206e-06,
+      "loss": 0.0609,
+      "step": 25380
+    },
+    {
+      "epoch": 2.1936152749578817,
+      "grad_norm": 0.6046134509758881,
+      "learning_rate": 1.7776539220283446e-06,
+      "loss": 0.0629,
+      "step": 25390
+    },
+    {
+      "epoch": 2.1944792431638516,
+      "grad_norm": 0.6632724484704178,
+      "learning_rate": 1.774089177116876e-06,
+      "loss": 0.0628,
+      "step": 25400
+    },
+    {
+      "epoch": 2.1953432113698215,
+      "grad_norm": 0.6093237196847797,
+      "learning_rate": 1.7705272389823869e-06,
+      "loss": 0.0622,
+      "step": 25410
+    },
+    {
+      "epoch": 2.196207179575792,
+      "grad_norm": 0.6369052508210911,
+      "learning_rate": 1.7669681107240223e-06,
+      "loss": 0.0607,
+      "step": 25420
+    },
+    {
+      "epoch": 2.1970711477817617,
+      "grad_norm": 0.6407569635936984,
+      "learning_rate": 1.763411795438482e-06,
+      "loss": 0.0611,
+      "step": 25430
+    },
+    {
+      "epoch": 2.1979351159877316,
+      "grad_norm": 0.6563720363166178,
+      "learning_rate": 1.7598582962200172e-06,
+      "loss": 0.0599,
+      "step": 25440
+    },
+    {
+      "epoch": 2.1987990841937015,
+      "grad_norm": 0.673275092712552,
+      "learning_rate": 1.7563076161604364e-06,
+      "loss": 0.0598,
+      "step": 25450
+    },
+    {
+      "epoch": 2.199663052399672,
+      "grad_norm": 0.6380876243254755,
+      "learning_rate": 1.7527597583490825e-06,
+      "loss": 0.0606,
+      "step": 25460
+    },
+    {
+      "epoch": 2.2005270206056418,
+      "grad_norm": 0.6450355340439092,
+      "learning_rate": 1.7492147258728538e-06,
+      "loss": 0.0621,
+      "step": 25470
+    },
+    {
+      "epoch": 2.2013909888116117,
+      "grad_norm": 0.6226050474103794,
+      "learning_rate": 1.745672521816184e-06,
+      "loss": 0.061,
+      "step": 25480
+    },
+    {
+      "epoch": 2.2022549570175816,
+      "grad_norm": 0.6409167634399674,
+      "learning_rate": 1.7421331492610533e-06,
+      "loss": 0.0614,
+      "step": 25490
+    },
+    {
+      "epoch": 2.203118925223552,
+      "grad_norm": 0.6730822678941681,
+      "learning_rate": 1.7385966112869657e-06,
+      "loss": 0.0591,
+      "step": 25500
+    },
+    {
+      "epoch": 2.203982893429522,
+      "grad_norm": 0.6970778786367295,
+      "learning_rate": 1.7350629109709715e-06,
+      "loss": 0.0607,
+      "step": 25510
+    },
+    {
+      "epoch": 2.2048468616354917,
+      "grad_norm": 0.6125821916693323,
+      "learning_rate": 1.731532051387646e-06,
+      "loss": 0.0596,
+      "step": 25520
+    },
+    {
+      "epoch": 2.205710829841462,
+      "grad_norm": 0.6656767212696019,
+      "learning_rate": 1.7280040356090933e-06,
+      "loss": 0.0607,
+      "step": 25530
+    },
+    {
+      "epoch": 2.206574798047432,
+      "grad_norm": 0.6411004681356632,
+      "learning_rate": 1.724478866704944e-06,
+      "loss": 0.062,
+      "step": 25540
+    },
+    {
+      "epoch": 2.207438766253402,
+      "grad_norm": 0.6764489258462001,
+      "learning_rate": 1.720956547742349e-06,
+      "loss": 0.0612,
+      "step": 25550
+    },
+    {
+      "epoch": 2.2083027344593718,
+      "grad_norm": 0.6128942114008559,
+      "learning_rate": 1.7174370817859854e-06,
+      "loss": 0.0608,
+      "step": 25560
+    },
+    {
+      "epoch": 2.209166702665342,
+      "grad_norm": 0.6339329827203459,
+      "learning_rate": 1.7139204718980434e-06,
+      "loss": 0.0625,
+      "step": 25570
+    },
+    {
+      "epoch": 2.210030670871312,
+      "grad_norm": 0.6452354260110682,
+      "learning_rate": 1.710406721138229e-06,
+      "loss": 0.0614,
+      "step": 25580
+    },
+    {
+      "epoch": 2.210894639077282,
+      "grad_norm": 0.6166881292995755,
+      "learning_rate": 1.7068958325637591e-06,
+      "loss": 0.0616,
+      "step": 25590
+    },
+    {
+      "epoch": 2.211758607283252,
+      "grad_norm": 0.6476715295266554,
+      "learning_rate": 1.7033878092293655e-06,
+      "loss": 0.061,
+      "step": 25600
+    },
+    {
+      "epoch": 2.212622575489222,
+      "grad_norm": 0.6517855464246195,
+      "learning_rate": 1.699882654187282e-06,
+      "loss": 0.0627,
+      "step": 25610
+    },
+    {
+      "epoch": 2.213486543695192,
+      "grad_norm": 0.6340260543769496,
+      "learning_rate": 1.6963803704872478e-06,
+      "loss": 0.0611,
+      "step": 25620
+    },
+    {
+      "epoch": 2.214350511901162,
+      "grad_norm": 0.6298347310498605,
+      "learning_rate": 1.6928809611765051e-06,
+      "loss": 0.0616,
+      "step": 25630
+    },
+    {
+      "epoch": 2.2152144801071323,
+      "grad_norm": 0.6243243135913467,
+      "learning_rate": 1.6893844292997923e-06,
+      "loss": 0.0615,
+      "step": 25640
+    },
+    {
+      "epoch": 2.216078448313102,
+      "grad_norm": 0.6340777870224308,
+      "learning_rate": 1.6858907778993516e-06,
+      "loss": 0.0609,
+      "step": 25650
+    },
+    {
+      "epoch": 2.216942416519072,
+      "grad_norm": 0.6083314908430885,
+      "learning_rate": 1.6824000100149067e-06,
+      "loss": 0.0596,
+      "step": 25660
+    },
+    {
+      "epoch": 2.217806384725042,
+      "grad_norm": 0.6726539489645692,
+      "learning_rate": 1.678912128683685e-06,
+      "loss": 0.0621,
+      "step": 25670
+    },
+    {
+      "epoch": 2.2186703529310123,
+      "grad_norm": 0.6194556913540952,
+      "learning_rate": 1.6754271369403934e-06,
+      "loss": 0.0613,
+      "step": 25680
+    },
+    {
+      "epoch": 2.2195343211369822,
+      "grad_norm": 0.6207525296451107,
+      "learning_rate": 1.671945037817233e-06,
+      "loss": 0.0616,
+      "step": 25690
+    },
+    {
+      "epoch": 2.220398289342952,
+      "grad_norm": 0.681932107424091,
+      "learning_rate": 1.6684658343438769e-06,
+      "loss": 0.0621,
+      "step": 25700
+    },
+    {
+      "epoch": 2.221262257548922,
+      "grad_norm": 0.6855599096619783,
+      "learning_rate": 1.6649895295474895e-06,
+      "loss": 0.0619,
+      "step": 25710
+    },
+    {
+      "epoch": 2.2221262257548924,
+      "grad_norm": 0.6129639845472115,
+      "learning_rate": 1.6615161264527075e-06,
+      "loss": 0.0607,
+      "step": 25720
+    },
+    {
+      "epoch": 2.2229901939608623,
+      "grad_norm": 0.6626689168785096,
+      "learning_rate": 1.6580456280816442e-06,
+      "loss": 0.0599,
+      "step": 25730
+    },
+    {
+      "epoch": 2.223854162166832,
+      "grad_norm": 0.6226827855598428,
+      "learning_rate": 1.654578037453885e-06,
+      "loss": 0.0602,
+      "step": 25740
+    },
+    {
+      "epoch": 2.224718130372802,
+      "grad_norm": 0.6147897546663305,
+      "learning_rate": 1.6511133575864846e-06,
+      "loss": 0.0595,
+      "step": 25750
+    },
+    {
+      "epoch": 2.2255820985787724,
+      "grad_norm": 0.6263674040275825,
+      "learning_rate": 1.6476515914939718e-06,
+      "loss": 0.0642,
+      "step": 25760
+    },
+    {
+      "epoch": 2.2264460667847423,
+      "grad_norm": 0.6413262252107033,
+      "learning_rate": 1.6441927421883274e-06,
+      "loss": 0.0617,
+      "step": 25770
+    },
+    {
+      "epoch": 2.2273100349907122,
+      "grad_norm": 0.6571817695841334,
+      "learning_rate": 1.6407368126790074e-06,
+      "loss": 0.0614,
+      "step": 25780
+    },
+    {
+      "epoch": 2.228174003196682,
+      "grad_norm": 0.639650797211082,
+      "learning_rate": 1.637283805972918e-06,
+      "loss": 0.0625,
+      "step": 25790
+    },
+    {
+      "epoch": 2.2290379714026525,
+      "grad_norm": 0.6241557566612796,
+      "learning_rate": 1.6338337250744313e-06,
+      "loss": 0.0594,
+      "step": 25800
+    },
+    {
+      "epoch": 2.2299019396086224,
+      "grad_norm": 0.6436803684620909,
+      "learning_rate": 1.6303865729853618e-06,
+      "loss": 0.0591,
+      "step": 25810
+    },
+    {
+      "epoch": 2.2307659078145923,
+      "grad_norm": 0.6593942452366395,
+      "learning_rate": 1.6269423527049876e-06,
+      "loss": 0.0605,
+      "step": 25820
+    },
+    {
+      "epoch": 2.2316298760205626,
+      "grad_norm": 0.7094732125345027,
+      "learning_rate": 1.623501067230029e-06,
+      "loss": 0.0596,
+      "step": 25830
+    },
+    {
+      "epoch": 2.2324938442265325,
+      "grad_norm": 0.639451168612842,
+      "learning_rate": 1.6200627195546547e-06,
+      "loss": 0.0587,
+      "step": 25840
+    },
+    {
+      "epoch": 2.2333578124325024,
+      "grad_norm": 0.666192956241704,
+      "learning_rate": 1.6166273126704773e-06,
+      "loss": 0.0592,
+      "step": 25850
+    },
+    {
+      "epoch": 2.2342217806384723,
+      "grad_norm": 0.651541166383784,
+      "learning_rate": 1.6131948495665483e-06,
+      "loss": 0.0627,
+      "step": 25860
+    },
+    {
+      "epoch": 2.2350857488444427,
+      "grad_norm": 0.6671884346428812,
+      "learning_rate": 1.609765333229364e-06,
+      "loss": 0.0592,
+      "step": 25870
+    },
+    {
+      "epoch": 2.2359497170504126,
+      "grad_norm": 0.6312027856638661,
+      "learning_rate": 1.6063387666428514e-06,
+      "loss": 0.0616,
+      "step": 25880
+    },
+    {
+      "epoch": 2.2368136852563825,
+      "grad_norm": 0.6351034279727996,
+      "learning_rate": 1.6029151527883724e-06,
+      "loss": 0.0619,
+      "step": 25890
+    },
+    {
+      "epoch": 2.2376776534623524,
+      "grad_norm": 0.650998398374031,
+      "learning_rate": 1.5994944946447187e-06,
+      "loss": 0.0596,
+      "step": 25900
+    },
+    {
+      "epoch": 2.2385416216683227,
+      "grad_norm": 0.64185933391637,
+      "learning_rate": 1.5960767951881146e-06,
+      "loss": 0.0613,
+      "step": 25910
+    },
+    {
+      "epoch": 2.2394055898742926,
+      "grad_norm": 0.6821387202412239,
+      "learning_rate": 1.5926620573922048e-06,
+      "loss": 0.0605,
+      "step": 25920
+    },
+    {
+      "epoch": 2.2402695580802625,
+      "grad_norm": 0.6584422635160079,
+      "learning_rate": 1.5892502842280605e-06,
+      "loss": 0.0631,
+      "step": 25930
+    },
+    {
+      "epoch": 2.241133526286233,
+      "grad_norm": 0.6719097501541714,
+      "learning_rate": 1.5858414786641723e-06,
+      "loss": 0.0598,
+      "step": 25940
+    },
+    {
+      "epoch": 2.2419974944922028,
+      "grad_norm": 0.6524586802098732,
+      "learning_rate": 1.5824356436664467e-06,
+      "loss": 0.0604,
+      "step": 25950
+    },
+    {
+      "epoch": 2.2428614626981727,
+      "grad_norm": 0.6516842713970796,
+      "learning_rate": 1.5790327821982137e-06,
+      "loss": 0.0625,
+      "step": 25960
+    },
+    {
+      "epoch": 2.2437254309041426,
+      "grad_norm": 0.6683175188643405,
+      "learning_rate": 1.5756328972202023e-06,
+      "loss": 0.0591,
+      "step": 25970
+    },
+    {
+      "epoch": 2.244589399110113,
+      "grad_norm": 0.625411302041235,
+      "learning_rate": 1.5722359916905656e-06,
+      "loss": 0.0593,
+      "step": 25980
+    },
+    {
+      "epoch": 2.245453367316083,
+      "grad_norm": 0.6293011621106597,
+      "learning_rate": 1.5688420685648565e-06,
+      "loss": 0.0622,
+      "step": 25990
+    },
+    {
+      "epoch": 2.2463173355220527,
+      "grad_norm": 0.6852578163384612,
+      "learning_rate": 1.5654511307960346e-06,
+      "loss": 0.0616,
+      "step": 26000
+    },
+    {
+      "epoch": 2.2471813037280226,
+      "grad_norm": 0.6510391632446941,
+      "learning_rate": 1.5620631813344611e-06,
+      "loss": 0.0643,
+      "step": 26010
+    },
+    {
+      "epoch": 2.248045271933993,
+      "grad_norm": 0.6375311713032594,
+      "learning_rate": 1.5586782231279012e-06,
+      "loss": 0.0628,
+      "step": 26020
+    },
+    {
+      "epoch": 2.248909240139963,
+      "grad_norm": 0.6571909140263892,
+      "learning_rate": 1.5552962591215137e-06,
+      "loss": 0.0603,
+      "step": 26030
+    },
+    {
+      "epoch": 2.2497732083459328,
+      "grad_norm": 0.6435249357904795,
+      "learning_rate": 1.5519172922578529e-06,
+      "loss": 0.0629,
+      "step": 26040
+    },
+    {
+      "epoch": 2.250637176551903,
+      "grad_norm": 0.6516646258238982,
+      "learning_rate": 1.5485413254768655e-06,
+      "loss": 0.0583,
+      "step": 26050
+    },
+    {
+      "epoch": 2.251501144757873,
+      "grad_norm": 0.6565838894870303,
+      "learning_rate": 1.5451683617158864e-06,
+      "loss": 0.0611,
+      "step": 26060
+    },
+    {
+      "epoch": 2.252365112963843,
+      "grad_norm": 0.6896405697177518,
+      "learning_rate": 1.541798403909644e-06,
+      "loss": 0.0601,
+      "step": 26070
+    },
+    {
+      "epoch": 2.253229081169813,
+      "grad_norm": 0.6312058690706274,
+      "learning_rate": 1.5384314549902407e-06,
+      "loss": 0.0598,
+      "step": 26080
+    },
+    {
+      "epoch": 2.254093049375783,
+      "grad_norm": 0.689835159961576,
+      "learning_rate": 1.5350675178871717e-06,
+      "loss": 0.0618,
+      "step": 26090
+    },
+    {
+      "epoch": 2.254957017581753,
+      "grad_norm": 0.6158946907902951,
+      "learning_rate": 1.531706595527303e-06,
+      "loss": 0.0604,
+      "step": 26100
+    },
+    {
+      "epoch": 2.255820985787723,
+      "grad_norm": 0.6099539265032519,
+      "learning_rate": 1.5283486908348872e-06,
+      "loss": 0.0589,
+      "step": 26110
+    },
+    {
+      "epoch": 2.256684953993693,
+      "grad_norm": 0.6450195523068343,
+      "learning_rate": 1.5249938067315379e-06,
+      "loss": 0.0613,
+      "step": 26120
+    },
+    {
+      "epoch": 2.257548922199663,
+      "grad_norm": 0.6487167626847626,
+      "learning_rate": 1.5216419461362542e-06,
+      "loss": 0.0591,
+      "step": 26130
+    },
+    {
+      "epoch": 2.258412890405633,
+      "grad_norm": 0.6753222278690582,
+      "learning_rate": 1.5182931119653965e-06,
+      "loss": 0.0595,
+      "step": 26140
+    },
+    {
+      "epoch": 2.259276858611603,
+      "grad_norm": 0.7071305222388669,
+      "learning_rate": 1.5149473071326941e-06,
+      "loss": 0.0609,
+      "step": 26150
+    },
+    {
+      "epoch": 2.2601408268175733,
+      "grad_norm": 0.664863282231861,
+      "learning_rate": 1.5116045345492403e-06,
+      "loss": 0.06,
+      "step": 26160
+    },
+    {
+      "epoch": 2.2610047950235432,
+      "grad_norm": 0.6528376283630767,
+      "learning_rate": 1.5082647971234886e-06,
+      "loss": 0.0635,
+      "step": 26170
+    },
+    {
+      "epoch": 2.261868763229513,
+      "grad_norm": 0.6492573255834764,
+      "learning_rate": 1.5049280977612575e-06,
+      "loss": 0.0627,
+      "step": 26180
+    },
+    {
+      "epoch": 2.262732731435483,
+      "grad_norm": 0.6627021537499391,
+      "learning_rate": 1.501594439365715e-06,
+      "loss": 0.0598,
+      "step": 26190
+    },
+    {
+      "epoch": 2.2635966996414534,
+      "grad_norm": 0.6806417650959726,
+      "learning_rate": 1.4982638248373871e-06,
+      "loss": 0.0637,
+      "step": 26200
+    },
+    {
+      "epoch": 2.2644606678474233,
+      "grad_norm": 0.62424115966496,
+      "learning_rate": 1.4949362570741493e-06,
+      "loss": 0.0592,
+      "step": 26210
+    },
+    {
+      "epoch": 2.265324636053393,
+      "grad_norm": 0.6688986783593902,
+      "learning_rate": 1.4916117389712303e-06,
+      "loss": 0.0603,
+      "step": 26220
+    },
+    {
+      "epoch": 2.266188604259363,
+      "grad_norm": 0.7018681679297502,
+      "learning_rate": 1.4882902734212013e-06,
+      "loss": 0.0637,
+      "step": 26230
+    },
+    {
+      "epoch": 2.2670525724653334,
+      "grad_norm": 0.5925240661361009,
+      "learning_rate": 1.484971863313978e-06,
+      "loss": 0.061,
+      "step": 26240
+    },
+    {
+      "epoch": 2.2679165406713033,
+      "grad_norm": 0.6414504235333198,
+      "learning_rate": 1.4816565115368199e-06,
+      "loss": 0.064,
+      "step": 26250
+    },
+    {
+      "epoch": 2.2687805088772732,
+      "grad_norm": 0.656382879329087,
+      "learning_rate": 1.4783442209743225e-06,
+      "loss": 0.0597,
+      "step": 26260
+    },
+    {
+      "epoch": 2.2696444770832436,
+      "grad_norm": 0.6691090047775298,
+      "learning_rate": 1.4750349945084209e-06,
+      "loss": 0.0605,
+      "step": 26270
+    },
+    {
+      "epoch": 2.2705084452892135,
+      "grad_norm": 0.657652046855535,
+      "learning_rate": 1.4717288350183805e-06,
+      "loss": 0.0599,
+      "step": 26280
+    },
+    {
+      "epoch": 2.2713724134951834,
+      "grad_norm": 0.6423009118728417,
+      "learning_rate": 1.4684257453808032e-06,
+      "loss": 0.0593,
+      "step": 26290
+    },
+    {
+      "epoch": 2.2722363817011533,
+      "grad_norm": 0.6485166849916811,
+      "learning_rate": 1.4651257284696164e-06,
+      "loss": 0.0595,
+      "step": 26300
+    },
+    {
+      "epoch": 2.273100349907123,
+      "grad_norm": 0.6207446565216592,
+      "learning_rate": 1.4618287871560737e-06,
+      "loss": 0.0601,
+      "step": 26310
+    },
+    {
+      "epoch": 2.2739643181130935,
+      "grad_norm": 0.6633360470389493,
+      "learning_rate": 1.4585349243087538e-06,
+      "loss": 0.061,
+      "step": 26320
+    },
+    {
+      "epoch": 2.2748282863190634,
+      "grad_norm": 0.6355126266677902,
+      "learning_rate": 1.4552441427935582e-06,
+      "loss": 0.0585,
+      "step": 26330
+    },
+    {
+      "epoch": 2.2756922545250333,
+      "grad_norm": 0.6310809829654427,
+      "learning_rate": 1.4519564454737063e-06,
+      "loss": 0.06,
+      "step": 26340
+    },
+    {
+      "epoch": 2.2765562227310037,
+      "grad_norm": 0.67846250533861,
+      "learning_rate": 1.4486718352097323e-06,
+      "loss": 0.0604,
+      "step": 26350
+    },
+    {
+      "epoch": 2.2774201909369736,
+      "grad_norm": 0.664061169733949,
+      "learning_rate": 1.4453903148594866e-06,
+      "loss": 0.0632,
+      "step": 26360
+    },
+    {
+      "epoch": 2.2782841591429435,
+      "grad_norm": 0.6697299183841422,
+      "learning_rate": 1.4421118872781291e-06,
+      "loss": 0.0605,
+      "step": 26370
+    },
+    {
+      "epoch": 2.2791481273489134,
+      "grad_norm": 0.6498208864924574,
+      "learning_rate": 1.438836555318135e-06,
+      "loss": 0.0639,
+      "step": 26380
+    },
+    {
+      "epoch": 2.2800120955548837,
+      "grad_norm": 0.64671706463022,
+      "learning_rate": 1.4355643218292742e-06,
+      "loss": 0.0615,
+      "step": 26390
+    },
+    {
+      "epoch": 2.2808760637608536,
+      "grad_norm": 0.7394098399800825,
+      "learning_rate": 1.4322951896586334e-06,
+      "loss": 0.0606,
+      "step": 26400
+    },
+    {
+      "epoch": 2.2817400319668235,
+      "grad_norm": 0.642376609395183,
+      "learning_rate": 1.4290291616505918e-06,
+      "loss": 0.0605,
+      "step": 26410
+    },
+    {
+      "epoch": 2.2826040001727934,
+      "grad_norm": 0.6566571466559815,
+      "learning_rate": 1.4257662406468353e-06,
+      "loss": 0.0573,
+      "step": 26420
+    },
+    {
+      "epoch": 2.2834679683787638,
+      "grad_norm": 0.6520952949887607,
+      "learning_rate": 1.4225064294863372e-06,
+      "loss": 0.0606,
+      "step": 26430
+    },
+    {
+      "epoch": 2.2843319365847337,
+      "grad_norm": 0.6838279782209414,
+      "learning_rate": 1.4192497310053748e-06,
+      "loss": 0.0623,
+      "step": 26440
+    },
+    {
+      "epoch": 2.2851959047907036,
+      "grad_norm": 0.6446957305304464,
+      "learning_rate": 1.41599614803751e-06,
+      "loss": 0.0599,
+      "step": 26450
+    },
+    {
+      "epoch": 2.286059872996674,
+      "grad_norm": 0.6410457377021014,
+      "learning_rate": 1.4127456834135978e-06,
+      "loss": 0.0625,
+      "step": 26460
+    },
+    {
+      "epoch": 2.286923841202644,
+      "grad_norm": 0.6186720566777877,
+      "learning_rate": 1.4094983399617784e-06,
+      "loss": 0.0614,
+      "step": 26470
+    },
+    {
+      "epoch": 2.2877878094086137,
+      "grad_norm": 0.6939594322670045,
+      "learning_rate": 1.4062541205074742e-06,
+      "loss": 0.0606,
+      "step": 26480
+    },
+    {
+      "epoch": 2.2886517776145836,
+      "grad_norm": 0.6742929364984191,
+      "learning_rate": 1.4030130278733967e-06,
+      "loss": 0.058,
+      "step": 26490
+    },
+    {
+      "epoch": 2.289515745820554,
+      "grad_norm": 0.6389179635631196,
+      "learning_rate": 1.3997750648795295e-06,
+      "loss": 0.0609,
+      "step": 26500
+    },
+    {
+      "epoch": 2.290379714026524,
+      "grad_norm": 0.6664027390477426,
+      "learning_rate": 1.3965402343431362e-06,
+      "loss": 0.0593,
+      "step": 26510
+    },
+    {
+      "epoch": 2.2912436822324938,
+      "grad_norm": 0.6522631321278886,
+      "learning_rate": 1.3933085390787531e-06,
+      "loss": 0.0611,
+      "step": 26520
+    },
+    {
+      "epoch": 2.2921076504384637,
+      "grad_norm": 0.6604097749419889,
+      "learning_rate": 1.3900799818981947e-06,
+      "loss": 0.06,
+      "step": 26530
+    },
+    {
+      "epoch": 2.292971618644434,
+      "grad_norm": 0.6406399475479331,
+      "learning_rate": 1.3868545656105342e-06,
+      "loss": 0.0586,
+      "step": 26540
+    },
+    {
+      "epoch": 2.293835586850404,
+      "grad_norm": 0.59954172486234,
+      "learning_rate": 1.3836322930221225e-06,
+      "loss": 0.061,
+      "step": 26550
+    },
+    {
+      "epoch": 2.294699555056374,
+      "grad_norm": 0.6664300784047551,
+      "learning_rate": 1.3804131669365705e-06,
+      "loss": 0.0595,
+      "step": 26560
+    },
+    {
+      "epoch": 2.295563523262344,
+      "grad_norm": 0.6188454289593853,
+      "learning_rate": 1.3771971901547515e-06,
+      "loss": 0.06,
+      "step": 26570
+    },
+    {
+      "epoch": 2.296427491468314,
+      "grad_norm": 0.6424177988440397,
+      "learning_rate": 1.3739843654747986e-06,
+      "loss": 0.0606,
+      "step": 26580
+    },
+    {
+      "epoch": 2.297291459674284,
+      "grad_norm": 0.6688160131458641,
+      "learning_rate": 1.370774695692102e-06,
+      "loss": 0.0606,
+      "step": 26590
+    },
+    {
+      "epoch": 2.298155427880254,
+      "grad_norm": 0.6683429083926847,
+      "learning_rate": 1.3675681835993109e-06,
+      "loss": 0.0613,
+      "step": 26600
+    },
+    {
+      "epoch": 2.299019396086224,
+      "grad_norm": 0.7006858852613675,
+      "learning_rate": 1.3643648319863222e-06,
+      "loss": 0.0588,
+      "step": 26610
+    },
+    {
+      "epoch": 2.299883364292194,
+      "grad_norm": 0.6778287862856338,
+      "learning_rate": 1.3611646436402849e-06,
+      "loss": 0.0626,
+      "step": 26620
+    },
+    {
+      "epoch": 2.300747332498164,
+      "grad_norm": 0.6624181219982882,
+      "learning_rate": 1.3579676213455934e-06,
+      "loss": 0.0618,
+      "step": 26630
+    },
+    {
+      "epoch": 2.301611300704134,
+      "grad_norm": 0.7008422542870012,
+      "learning_rate": 1.3547737678838934e-06,
+      "loss": 0.0607,
+      "step": 26640
+    },
+    {
+      "epoch": 2.3024752689101042,
+      "grad_norm": 0.6624919987207697,
+      "learning_rate": 1.3515830860340678e-06,
+      "loss": 0.0606,
+      "step": 26650
+    },
+    {
+      "epoch": 2.303339237116074,
+      "grad_norm": 0.6147145335368871,
+      "learning_rate": 1.3483955785722418e-06,
+      "loss": 0.0622,
+      "step": 26660
+    },
+    {
+      "epoch": 2.304203205322044,
+      "grad_norm": 0.6281804015316972,
+      "learning_rate": 1.3452112482717788e-06,
+      "loss": 0.0572,
+      "step": 26670
+    },
+    {
+      "epoch": 2.3050671735280144,
+      "grad_norm": 0.6451228796733127,
+      "learning_rate": 1.3420300979032758e-06,
+      "loss": 0.0606,
+      "step": 26680
+    },
+    {
+      "epoch": 2.3059311417339843,
+      "grad_norm": 0.6169719002668468,
+      "learning_rate": 1.3388521302345703e-06,
+      "loss": 0.0607,
+      "step": 26690
+    },
+    {
+      "epoch": 2.306795109939954,
+      "grad_norm": 0.685603442557472,
+      "learning_rate": 1.3356773480307178e-06,
+      "loss": 0.0618,
+      "step": 26700
+    },
+    {
+      "epoch": 2.307659078145924,
+      "grad_norm": 0.6443542909589884,
+      "learning_rate": 1.332505754054016e-06,
+      "loss": 0.0596,
+      "step": 26710
+    },
+    {
+      "epoch": 2.3085230463518944,
+      "grad_norm": 0.6542177335627326,
+      "learning_rate": 1.3293373510639772e-06,
+      "loss": 0.0599,
+      "step": 26720
+    },
+    {
+      "epoch": 2.3093870145578643,
+      "grad_norm": 0.618466824135546,
+      "learning_rate": 1.326172141817349e-06,
+      "loss": 0.058,
+      "step": 26730
+    },
+    {
+      "epoch": 2.3102509827638342,
+      "grad_norm": 0.6754024486903456,
+      "learning_rate": 1.3230101290680859e-06,
+      "loss": 0.0608,
+      "step": 26740
+    },
+    {
+      "epoch": 2.311114950969804,
+      "grad_norm": 0.6259336996138705,
+      "learning_rate": 1.3198513155673742e-06,
+      "loss": 0.0602,
+      "step": 26750
+    },
+    {
+      "epoch": 2.3119789191757745,
+      "grad_norm": 0.6952082803278199,
+      "learning_rate": 1.3166957040636102e-06,
+      "loss": 0.0611,
+      "step": 26760
+    },
+    {
+      "epoch": 2.3128428873817444,
+      "grad_norm": 0.679152130889011,
+      "learning_rate": 1.3135432973024044e-06,
+      "loss": 0.0572,
+      "step": 26770
+    },
+    {
+      "epoch": 2.3137068555877143,
+      "grad_norm": 0.6706211214233345,
+      "learning_rate": 1.3103940980265805e-06,
+      "loss": 0.0604,
+      "step": 26780
+    },
+    {
+      "epoch": 2.3145708237936846,
+      "grad_norm": 0.617309011922249,
+      "learning_rate": 1.3072481089761697e-06,
+      "loss": 0.0575,
+      "step": 26790
+    },
+    {
+      "epoch": 2.3154347919996545,
+      "grad_norm": 0.6615416949022743,
+      "learning_rate": 1.3041053328884128e-06,
+      "loss": 0.0616,
+      "step": 26800
+    },
+    {
+      "epoch": 2.3162987602056244,
+      "grad_norm": 0.6747162247811561,
+      "learning_rate": 1.3009657724977537e-06,
+      "loss": 0.0619,
+      "step": 26810
+    },
+    {
+      "epoch": 2.3171627284115943,
+      "grad_norm": 0.6664271188966158,
+      "learning_rate": 1.2978294305358374e-06,
+      "loss": 0.0615,
+      "step": 26820
+    },
+    {
+      "epoch": 2.3180266966175647,
+      "grad_norm": 0.65748922479625,
+      "learning_rate": 1.294696309731508e-06,
+      "loss": 0.0623,
+      "step": 26830
+    },
+    {
+      "epoch": 2.3188906648235346,
+      "grad_norm": 0.6685930838215557,
+      "learning_rate": 1.2915664128108123e-06,
+      "loss": 0.0604,
+      "step": 26840
+    },
+    {
+      "epoch": 2.3197546330295045,
+      "grad_norm": 0.6127967842512623,
+      "learning_rate": 1.288439742496984e-06,
+      "loss": 0.0599,
+      "step": 26850
+    },
+    {
+      "epoch": 2.3206186012354744,
+      "grad_norm": 0.6643011680928712,
+      "learning_rate": 1.2853163015104563e-06,
+      "loss": 0.063,
+      "step": 26860
+    },
+    {
+      "epoch": 2.3214825694414447,
+      "grad_norm": 0.6399180186475122,
+      "learning_rate": 1.2821960925688493e-06,
+      "loss": 0.0601,
+      "step": 26870
+    },
+    {
+      "epoch": 2.3223465376474146,
+      "grad_norm": 0.6309313709379819,
+      "learning_rate": 1.2790791183869717e-06,
+      "loss": 0.0597,
+      "step": 26880
+    },
+    {
+      "epoch": 2.3232105058533845,
+      "grad_norm": 0.6292393124967992,
+      "learning_rate": 1.2759653816768175e-06,
+      "loss": 0.0613,
+      "step": 26890
+    },
+    {
+      "epoch": 2.324074474059355,
+      "grad_norm": 0.6272135357664029,
+      "learning_rate": 1.2728548851475625e-06,
+      "loss": 0.0608,
+      "step": 26900
+    },
+    {
+      "epoch": 2.3249384422653248,
+      "grad_norm": 0.6534930026134436,
+      "learning_rate": 1.2697476315055673e-06,
+      "loss": 0.0603,
+      "step": 26910
+    },
+    {
+      "epoch": 2.3258024104712947,
+      "grad_norm": 0.6627039375093233,
+      "learning_rate": 1.2666436234543684e-06,
+      "loss": 0.0602,
+      "step": 26920
+    },
+    {
+      "epoch": 2.3266663786772646,
+      "grad_norm": 0.6998958974545966,
+      "learning_rate": 1.2635428636946773e-06,
+      "loss": 0.0587,
+      "step": 26930
+    },
+    {
+      "epoch": 2.3275303468832345,
+      "grad_norm": 0.664922356390914,
+      "learning_rate": 1.2604453549243796e-06,
+      "loss": 0.0607,
+      "step": 26940
+    },
+    {
+      "epoch": 2.328394315089205,
+      "grad_norm": 0.6545529848823667,
+      "learning_rate": 1.2573510998385357e-06,
+      "loss": 0.0606,
+      "step": 26950
+    },
+    {
+      "epoch": 2.3292582832951747,
+      "grad_norm": 0.6572234115129042,
+      "learning_rate": 1.2542601011293714e-06,
+      "loss": 0.0614,
+      "step": 26960
+    },
+    {
+      "epoch": 2.3301222515011446,
+      "grad_norm": 0.623619780479768,
+      "learning_rate": 1.2511723614862803e-06,
+      "loss": 0.0586,
+      "step": 26970
+    },
+    {
+      "epoch": 2.330986219707115,
+      "grad_norm": 0.6760745961536608,
+      "learning_rate": 1.2480878835958199e-06,
+      "loss": 0.0603,
+      "step": 26980
+    },
+    {
+      "epoch": 2.331850187913085,
+      "grad_norm": 0.6446461921885446,
+      "learning_rate": 1.245006670141709e-06,
+      "loss": 0.0606,
+      "step": 26990
+    },
+    {
+      "epoch": 2.3327141561190547,
+      "grad_norm": 0.6720018979198659,
+      "learning_rate": 1.2419287238048327e-06,
+      "loss": 0.0615,
+      "step": 27000
+    },
+    {
+      "epoch": 2.3335781243250246,
+      "grad_norm": 0.6209878169214561,
+      "learning_rate": 1.238854047263221e-06,
+      "loss": 0.0624,
+      "step": 27010
+    },
+    {
+      "epoch": 2.334442092530995,
+      "grad_norm": 0.6223483037469352,
+      "learning_rate": 1.2357826431920706e-06,
+      "loss": 0.059,
+      "step": 27020
+    },
+    {
+      "epoch": 2.335306060736965,
+      "grad_norm": 0.6514212276691695,
+      "learning_rate": 1.2327145142637237e-06,
+      "loss": 0.0605,
+      "step": 27030
+    },
+    {
+      "epoch": 2.336170028942935,
+      "grad_norm": 0.6010763844851291,
+      "learning_rate": 1.2296496631476807e-06,
+      "loss": 0.0606,
+      "step": 27040
+    },
+    {
+      "epoch": 2.3370339971489047,
+      "grad_norm": 0.6531152999704667,
+      "learning_rate": 1.2265880925105777e-06,
+      "loss": 0.0591,
+      "step": 27050
+    },
+    {
+      "epoch": 2.337897965354875,
+      "grad_norm": 0.6706702248713575,
+      "learning_rate": 1.2235298050162093e-06,
+      "loss": 0.0592,
+      "step": 27060
+    },
+    {
+      "epoch": 2.338761933560845,
+      "grad_norm": 0.6501204240351136,
+      "learning_rate": 1.2204748033255054e-06,
+      "loss": 0.0601,
+      "step": 27070
+    },
+    {
+      "epoch": 2.339625901766815,
+      "grad_norm": 0.6360109585574683,
+      "learning_rate": 1.2174230900965407e-06,
+      "loss": 0.0622,
+      "step": 27080
+    },
+    {
+      "epoch": 2.340489869972785,
+      "grad_norm": 0.7292331220131177,
+      "learning_rate": 1.214374667984527e-06,
+      "loss": 0.0602,
+      "step": 27090
+    },
+    {
+      "epoch": 2.341353838178755,
+      "grad_norm": 0.64431736347878,
+      "learning_rate": 1.2113295396418124e-06,
+      "loss": 0.06,
+      "step": 27100
+    },
+    {
+      "epoch": 2.342217806384725,
+      "grad_norm": 0.6693597385480407,
+      "learning_rate": 1.2082877077178834e-06,
+      "loss": 0.0604,
+      "step": 27110
+    },
+    {
+      "epoch": 2.343081774590695,
+      "grad_norm": 0.70203505820828,
+      "learning_rate": 1.2052491748593492e-06,
+      "loss": 0.0584,
+      "step": 27120
+    },
+    {
+      "epoch": 2.3439457427966652,
+      "grad_norm": 0.6529968481749552,
+      "learning_rate": 1.2022139437099601e-06,
+      "loss": 0.0582,
+      "step": 27130
+    },
+    {
+      "epoch": 2.344809711002635,
+      "grad_norm": 0.6292470117300617,
+      "learning_rate": 1.1991820169105827e-06,
+      "loss": 0.0591,
+      "step": 27140
+    },
+    {
+      "epoch": 2.345673679208605,
+      "grad_norm": 0.6937281317516001,
+      "learning_rate": 1.19615339709922e-06,
+      "loss": 0.0611,
+      "step": 27150
+    },
+    {
+      "epoch": 2.346537647414575,
+      "grad_norm": 0.667882030426861,
+      "learning_rate": 1.193128086910985e-06,
+      "loss": 0.0572,
+      "step": 27160
+    },
+    {
+      "epoch": 2.3474016156205453,
+      "grad_norm": 0.6415713468157893,
+      "learning_rate": 1.1901060889781224e-06,
+      "loss": 0.0596,
+      "step": 27170
+    },
+    {
+      "epoch": 2.348265583826515,
+      "grad_norm": 0.6686307087579261,
+      "learning_rate": 1.1870874059299875e-06,
+      "loss": 0.0594,
+      "step": 27180
+    },
+    {
+      "epoch": 2.349129552032485,
+      "grad_norm": 0.6509117568868412,
+      "learning_rate": 1.1840720403930555e-06,
+      "loss": 0.0594,
+      "step": 27190
+    },
+    {
+      "epoch": 2.3499935202384554,
+      "grad_norm": 0.6287633557030508,
+      "learning_rate": 1.1810599949909124e-06,
+      "loss": 0.0589,
+      "step": 27200
+    },
+    {
+      "epoch": 2.3508574884444253,
+      "grad_norm": 0.6378929565344178,
+      "learning_rate": 1.1780512723442556e-06,
+      "loss": 0.0607,
+      "step": 27210
+    },
+    {
+      "epoch": 2.3517214566503952,
+      "grad_norm": 0.6682225214396356,
+      "learning_rate": 1.1750458750708959e-06,
+      "loss": 0.0609,
+      "step": 27220
+    },
+    {
+      "epoch": 2.352585424856365,
+      "grad_norm": 0.664213964785015,
+      "learning_rate": 1.1720438057857458e-06,
+      "loss": 0.0589,
+      "step": 27230
+    },
+    {
+      "epoch": 2.3534493930623355,
+      "grad_norm": 0.7004254892208797,
+      "learning_rate": 1.169045067100824e-06,
+      "loss": 0.0593,
+      "step": 27240
+    },
+    {
+      "epoch": 2.3543133612683054,
+      "grad_norm": 0.6778013453063583,
+      "learning_rate": 1.1660496616252498e-06,
+      "loss": 0.0593,
+      "step": 27250
+    },
+    {
+      "epoch": 2.3551773294742753,
+      "grad_norm": 0.6625578892854007,
+      "learning_rate": 1.1630575919652459e-06,
+      "loss": 0.0592,
+      "step": 27260
+    },
+    {
+      "epoch": 2.356041297680245,
+      "grad_norm": 0.6889032338164695,
+      "learning_rate": 1.16006886072413e-06,
+      "loss": 0.0598,
+      "step": 27270
+    },
+    {
+      "epoch": 2.3569052658862155,
+      "grad_norm": 0.656085217318829,
+      "learning_rate": 1.1570834705023148e-06,
+      "loss": 0.058,
+      "step": 27280
+    },
+    {
+      "epoch": 2.3577692340921854,
+      "grad_norm": 0.6322543933666873,
+      "learning_rate": 1.1541014238973076e-06,
+      "loss": 0.0594,
+      "step": 27290
+    },
+    {
+      "epoch": 2.3586332022981553,
+      "grad_norm": 0.6621617238381241,
+      "learning_rate": 1.1511227235037036e-06,
+      "loss": 0.0603,
+      "step": 27300
+    },
+    {
+      "epoch": 2.3594971705041257,
+      "grad_norm": 0.6675369069452474,
+      "learning_rate": 1.1481473719131935e-06,
+      "loss": 0.0591,
+      "step": 27310
+    },
+    {
+      "epoch": 2.3603611387100956,
+      "grad_norm": 0.6403904619439585,
+      "learning_rate": 1.1451753717145436e-06,
+      "loss": 0.057,
+      "step": 27320
+    },
+    {
+      "epoch": 2.3612251069160655,
+      "grad_norm": 0.6974195948916855,
+      "learning_rate": 1.1422067254936148e-06,
+      "loss": 0.0564,
+      "step": 27330
+    },
+    {
+      "epoch": 2.3620890751220354,
+      "grad_norm": 0.6844415275124478,
+      "learning_rate": 1.139241435833342e-06,
+      "loss": 0.0605,
+      "step": 27340
+    },
+    {
+      "epoch": 2.3629530433280057,
+      "grad_norm": 0.6308979148922937,
+      "learning_rate": 1.1362795053137477e-06,
+      "loss": 0.0591,
+      "step": 27350
+    },
+    {
+      "epoch": 2.3638170115339756,
+      "grad_norm": 0.6868448909356765,
+      "learning_rate": 1.1333209365119198e-06,
+      "loss": 0.0606,
+      "step": 27360
+    },
+    {
+      "epoch": 2.3646809797399455,
+      "grad_norm": 0.6814825334027916,
+      "learning_rate": 1.130365732002034e-06,
+      "loss": 0.0597,
+      "step": 27370
+    },
+    {
+      "epoch": 2.3655449479459154,
+      "grad_norm": 0.6505325443513633,
+      "learning_rate": 1.1274138943553303e-06,
+      "loss": 0.0601,
+      "step": 27380
+    },
+    {
+      "epoch": 2.3664089161518858,
+      "grad_norm": 0.6516230369514702,
+      "learning_rate": 1.1244654261401233e-06,
+      "loss": 0.0573,
+      "step": 27390
+    },
+    {
+      "epoch": 2.3672728843578557,
+      "grad_norm": 0.6351111112315094,
+      "learning_rate": 1.1215203299217941e-06,
+      "loss": 0.0591,
+      "step": 27400
+    },
+    {
+      "epoch": 2.3681368525638256,
+      "grad_norm": 0.6323350777816571,
+      "learning_rate": 1.1185786082627897e-06,
+      "loss": 0.0582,
+      "step": 27410
+    },
+    {
+      "epoch": 2.369000820769796,
+      "grad_norm": 0.669953614513429,
+      "learning_rate": 1.1156402637226254e-06,
+      "loss": 0.0601,
+      "step": 27420
+    },
+    {
+      "epoch": 2.369864788975766,
+      "grad_norm": 0.6477445841820543,
+      "learning_rate": 1.112705298857869e-06,
+      "loss": 0.0608,
+      "step": 27430
+    },
+    {
+      "epoch": 2.3707287571817357,
+      "grad_norm": 0.6971705503297287,
+      "learning_rate": 1.1097737162221584e-06,
+      "loss": 0.0585,
+      "step": 27440
+    },
+    {
+      "epoch": 2.3715927253877056,
+      "grad_norm": 0.6874010419235855,
+      "learning_rate": 1.1068455183661796e-06,
+      "loss": 0.0611,
+      "step": 27450
+    },
+    {
+      "epoch": 2.3724566935936755,
+      "grad_norm": 0.6634814836649913,
+      "learning_rate": 1.103920707837683e-06,
+      "loss": 0.0573,
+      "step": 27460
+    },
+    {
+      "epoch": 2.373320661799646,
+      "grad_norm": 0.6668055692462801,
+      "learning_rate": 1.10099928718146e-06,
+      "loss": 0.0606,
+      "step": 27470
+    },
+    {
+      "epoch": 2.3741846300056157,
+      "grad_norm": 0.659507511140097,
+      "learning_rate": 1.0980812589393637e-06,
+      "loss": 0.0584,
+      "step": 27480
+    },
+    {
+      "epoch": 2.3750485982115856,
+      "grad_norm": 0.7058506378776868,
+      "learning_rate": 1.095166625650289e-06,
+      "loss": 0.0605,
+      "step": 27490
+    },
+    {
+      "epoch": 2.375912566417556,
+      "grad_norm": 0.6616478710874778,
+      "learning_rate": 1.0922553898501799e-06,
+      "loss": 0.0575,
+      "step": 27500
+    },
+    {
+      "epoch": 2.376776534623526,
+      "grad_norm": 0.6416220397649288,
+      "learning_rate": 1.0893475540720215e-06,
+      "loss": 0.0574,
+      "step": 27510
+    },
+    {
+      "epoch": 2.377640502829496,
+      "grad_norm": 0.6851881227890764,
+      "learning_rate": 1.086443120845842e-06,
+      "loss": 0.0603,
+      "step": 27520
+    },
+    {
+      "epoch": 2.378504471035466,
+      "grad_norm": 0.643914935400431,
+      "learning_rate": 1.0835420926987123e-06,
+      "loss": 0.0594,
+      "step": 27530
+    },
+    {
+      "epoch": 2.379368439241436,
+      "grad_norm": 0.7065001618716296,
+      "learning_rate": 1.0806444721547367e-06,
+      "loss": 0.0598,
+      "step": 27540
+    },
+    {
+      "epoch": 2.380232407447406,
+      "grad_norm": 0.6531887811707487,
+      "learning_rate": 1.077750261735055e-06,
+      "loss": 0.0632,
+      "step": 27550
+    },
+    {
+      "epoch": 2.381096375653376,
+      "grad_norm": 0.6782902569583262,
+      "learning_rate": 1.0748594639578391e-06,
+      "loss": 0.0597,
+      "step": 27560
+    },
+    {
+      "epoch": 2.3819603438593457,
+      "grad_norm": 0.6590868743045272,
+      "learning_rate": 1.0719720813382972e-06,
+      "loss": 0.0596,
+      "step": 27570
+    },
+    {
+      "epoch": 2.382824312065316,
+      "grad_norm": 0.6852303874619127,
+      "learning_rate": 1.0690881163886602e-06,
+      "loss": 0.0586,
+      "step": 27580
+    },
+    {
+      "epoch": 2.383688280271286,
+      "grad_norm": 0.6691472842481307,
+      "learning_rate": 1.066207571618187e-06,
+      "loss": 0.0586,
+      "step": 27590
+    },
+    {
+      "epoch": 2.384552248477256,
+      "grad_norm": 0.6758493976482691,
+      "learning_rate": 1.0633304495331614e-06,
+      "loss": 0.0587,
+      "step": 27600
+    },
+    {
+      "epoch": 2.3854162166832262,
+      "grad_norm": 0.6360569860336877,
+      "learning_rate": 1.0604567526368875e-06,
+      "loss": 0.0598,
+      "step": 27610
+    },
+    {
+      "epoch": 2.386280184889196,
+      "grad_norm": 0.6688496501564594,
+      "learning_rate": 1.057586483429694e-06,
+      "loss": 0.0602,
+      "step": 27620
+    },
+    {
+      "epoch": 2.387144153095166,
+      "grad_norm": 0.6318508440444366,
+      "learning_rate": 1.054719644408919e-06,
+      "loss": 0.0599,
+      "step": 27630
+    },
+    {
+      "epoch": 2.388008121301136,
+      "grad_norm": 0.6679409504648013,
+      "learning_rate": 1.051856238068925e-06,
+      "loss": 0.0589,
+      "step": 27640
+    },
+    {
+      "epoch": 2.3888720895071063,
+      "grad_norm": 0.6845032317565701,
+      "learning_rate": 1.0489962669010817e-06,
+      "loss": 0.0584,
+      "step": 27650
+    },
+    {
+      "epoch": 2.389736057713076,
+      "grad_norm": 0.6218207202922958,
+      "learning_rate": 1.0461397333937728e-06,
+      "loss": 0.0586,
+      "step": 27660
+    },
+    {
+      "epoch": 2.390600025919046,
+      "grad_norm": 0.6504975420124806,
+      "learning_rate": 1.0432866400323883e-06,
+      "loss": 0.0593,
+      "step": 27670
+    },
+    {
+      "epoch": 2.391463994125016,
+      "grad_norm": 0.6627505995283275,
+      "learning_rate": 1.0404369892993299e-06,
+      "loss": 0.0601,
+      "step": 27680
+    },
+    {
+      "epoch": 2.3923279623309863,
+      "grad_norm": 0.6866333978458881,
+      "learning_rate": 1.037590783673999e-06,
+      "loss": 0.0603,
+      "step": 27690
+    },
+    {
+      "epoch": 2.3931919305369562,
+      "grad_norm": 0.6274084673491703,
+      "learning_rate": 1.0347480256328025e-06,
+      "loss": 0.0597,
+      "step": 27700
+    },
+    {
+      "epoch": 2.394055898742926,
+      "grad_norm": 0.6381370089501482,
+      "learning_rate": 1.0319087176491455e-06,
+      "loss": 0.0577,
+      "step": 27710
+    },
+    {
+      "epoch": 2.3949198669488965,
+      "grad_norm": 0.6704993469854148,
+      "learning_rate": 1.0290728621934315e-06,
+      "loss": 0.0599,
+      "step": 27720
+    },
+    {
+      "epoch": 2.3957838351548664,
+      "grad_norm": 0.6716724508184789,
+      "learning_rate": 1.0262404617330652e-06,
+      "loss": 0.0592,
+      "step": 27730
+    },
+    {
+      "epoch": 2.3966478033608363,
+      "grad_norm": 0.6583842494843922,
+      "learning_rate": 1.023411518732435e-06,
+      "loss": 0.058,
+      "step": 27740
+    },
+    {
+      "epoch": 2.397511771566806,
+      "grad_norm": 0.655274656221077,
+      "learning_rate": 1.0205860356529318e-06,
+      "loss": 0.059,
+      "step": 27750
+    },
+    {
+      "epoch": 2.3983757397727765,
+      "grad_norm": 0.6339739798600978,
+      "learning_rate": 1.0177640149529277e-06,
+      "loss": 0.0609,
+      "step": 27760
+    },
+    {
+      "epoch": 2.3992397079787464,
+      "grad_norm": 0.6884819006309391,
+      "learning_rate": 1.014945459087791e-06,
+      "loss": 0.0601,
+      "step": 27770
+    },
+    {
+      "epoch": 2.4001036761847163,
+      "grad_norm": 0.6504434154351232,
+      "learning_rate": 1.0121303705098646e-06,
+      "loss": 0.0583,
+      "step": 27780
+    },
+    {
+      "epoch": 2.400967644390686,
+      "grad_norm": 0.6478456711603433,
+      "learning_rate": 1.0093187516684832e-06,
+      "loss": 0.0624,
+      "step": 27790
+    },
+    {
+      "epoch": 2.4018316125966566,
+      "grad_norm": 0.6253213623206553,
+      "learning_rate": 1.0065106050099599e-06,
+      "loss": 0.0587,
+      "step": 27800
+    },
+    {
+      "epoch": 2.4026955808026265,
+      "grad_norm": 0.6790278170027676,
+      "learning_rate": 1.003705932977585e-06,
+      "loss": 0.0596,
+      "step": 27810
+    },
+    {
+      "epoch": 2.4035595490085964,
+      "grad_norm": 0.6798115066515837,
+      "learning_rate": 1.0009047380116283e-06,
+      "loss": 0.0607,
+      "step": 27820
+    },
+    {
+      "epoch": 2.4044235172145667,
+      "grad_norm": 0.6604675110327878,
+      "learning_rate": 9.98107022549331e-07,
+      "loss": 0.0604,
+      "step": 27830
+    },
+    {
+      "epoch": 2.4052874854205366,
+      "grad_norm": 0.6790462682499643,
+      "learning_rate": 9.95312789024912e-07,
+      "loss": 0.0568,
+      "step": 27840
+    },
+    {
+      "epoch": 2.4061514536265065,
+      "grad_norm": 0.6128078717477331,
+      "learning_rate": 9.925220398695562e-07,
+      "loss": 0.0596,
+      "step": 27850
+    },
+    {
+      "epoch": 2.4070154218324764,
+      "grad_norm": 0.6753776573614281,
+      "learning_rate": 9.897347775114185e-07,
+      "loss": 0.0608,
+      "step": 27860
+    },
+    {
+      "epoch": 2.4078793900384468,
+      "grad_norm": 0.645529040989447,
+      "learning_rate": 9.86951004375618e-07,
+      "loss": 0.0586,
+      "step": 27870
+    },
+    {
+      "epoch": 2.4087433582444167,
+      "grad_norm": 0.6658258287934132,
+      "learning_rate": 9.841707228842428e-07,
+      "loss": 0.0596,
+      "step": 27880
+    },
+    {
+      "epoch": 2.4096073264503866,
+      "grad_norm": 0.6567521212851223,
+      "learning_rate": 9.81393935456339e-07,
+      "loss": 0.0576,
+      "step": 27890
+    },
+    {
+      "epoch": 2.4104712946563565,
+      "grad_norm": 0.660656906400927,
+      "learning_rate": 9.786206445079127e-07,
+      "loss": 0.058,
+      "step": 27900
+    },
+    {
+      "epoch": 2.411335262862327,
+      "grad_norm": 0.6380899683509319,
+      "learning_rate": 9.75850852451929e-07,
+      "loss": 0.0589,
+      "step": 27910
+    },
+    {
+      "epoch": 2.4121992310682967,
+      "grad_norm": 0.6517954005215502,
+      "learning_rate": 9.73084561698307e-07,
+      "loss": 0.0599,
+      "step": 27920
+    },
+    {
+      "epoch": 2.4130631992742666,
+      "grad_norm": 0.6807352997523919,
+      "learning_rate": 9.703217746539256e-07,
+      "loss": 0.0604,
+      "step": 27930
+    },
+    {
+      "epoch": 2.413927167480237,
+      "grad_norm": 0.633359951134548,
+      "learning_rate": 9.675624937226037e-07,
+      "loss": 0.0582,
+      "step": 27940
+    },
+    {
+      "epoch": 2.414791135686207,
+      "grad_norm": 0.661588990732132,
+      "learning_rate": 9.648067213051216e-07,
+      "loss": 0.062,
+      "step": 27950
+    },
+    {
+      "epoch": 2.4156551038921767,
+      "grad_norm": 0.6643855642016916,
+      "learning_rate": 9.620544597991992e-07,
+      "loss": 0.059,
+      "step": 27960
+    },
+    {
+      "epoch": 2.4165190720981466,
+      "grad_norm": 0.6772126749791726,
+      "learning_rate": 9.593057115995053e-07,
+      "loss": 0.06,
+      "step": 27970
+    },
+    {
+      "epoch": 2.417383040304117,
+      "grad_norm": 0.6194181138318597,
+      "learning_rate": 9.565604790976485e-07,
+      "loss": 0.0578,
+      "step": 27980
+    },
+    {
+      "epoch": 2.418247008510087,
+      "grad_norm": 0.6764386780365282,
+      "learning_rate": 9.53818764682184e-07,
+      "loss": 0.058,
+      "step": 27990
+    },
+    {
+      "epoch": 2.419110976716057,
+      "grad_norm": 0.6927664968419396,
+      "learning_rate": 9.510805707386006e-07,
+      "loss": 0.0599,
+      "step": 28000
+    },
+    {
+      "epoch": 2.4199749449220267,
+      "grad_norm": 0.674259380191846,
+      "learning_rate": 9.483458996493267e-07,
+      "loss": 0.0599,
+      "step": 28010
+    },
+    {
+      "epoch": 2.420838913127997,
+      "grad_norm": 0.6678938458437413,
+      "learning_rate": 9.456147537937249e-07,
+      "loss": 0.0607,
+      "step": 28020
+    },
+    {
+      "epoch": 2.421702881333967,
+      "grad_norm": 0.6678011366556437,
+      "learning_rate": 9.42887135548089e-07,
+      "loss": 0.0579,
+      "step": 28030
+    },
+    {
+      "epoch": 2.422566849539937,
+      "grad_norm": 0.6879787431287953,
+      "learning_rate": 9.401630472856499e-07,
+      "loss": 0.0601,
+      "step": 28040
+    },
+    {
+      "epoch": 2.423430817745907,
+      "grad_norm": 0.6705502097546254,
+      "learning_rate": 9.374424913765567e-07,
+      "loss": 0.0579,
+      "step": 28050
+    },
+    {
+      "epoch": 2.424294785951877,
+      "grad_norm": 0.6599368510358516,
+      "learning_rate": 9.347254701878943e-07,
+      "loss": 0.06,
+      "step": 28060
+    },
+    {
+      "epoch": 2.425158754157847,
+      "grad_norm": 0.6826991683827245,
+      "learning_rate": 9.320119860836674e-07,
+      "loss": 0.0614,
+      "step": 28070
+    },
+    {
+      "epoch": 2.426022722363817,
+      "grad_norm": 0.6762089597321239,
+      "learning_rate": 9.293020414248072e-07,
+      "loss": 0.0594,
+      "step": 28080
+    },
+    {
+      "epoch": 2.426886690569787,
+      "grad_norm": 0.6460291522786121,
+      "learning_rate": 9.265956385691583e-07,
+      "loss": 0.0579,
+      "step": 28090
+    },
+    {
+      "epoch": 2.427750658775757,
+      "grad_norm": 0.6506399156335623,
+      "learning_rate": 9.238927798714908e-07,
+      "loss": 0.0581,
+      "step": 28100
+    },
+    {
+      "epoch": 2.428614626981727,
+      "grad_norm": 0.6684113157689684,
+      "learning_rate": 9.211934676834882e-07,
+      "loss": 0.0594,
+      "step": 28110
+    },
+    {
+      "epoch": 2.429478595187697,
+      "grad_norm": 0.6356040848491912,
+      "learning_rate": 9.184977043537474e-07,
+      "loss": 0.0573,
+      "step": 28120
+    },
+    {
+      "epoch": 2.4303425633936673,
+      "grad_norm": 0.6847493066510186,
+      "learning_rate": 9.158054922277787e-07,
+      "loss": 0.0605,
+      "step": 28130
+    },
+    {
+      "epoch": 2.431206531599637,
+      "grad_norm": 0.7050255385387857,
+      "learning_rate": 9.131168336480018e-07,
+      "loss": 0.059,
+      "step": 28140
+    },
+    {
+      "epoch": 2.432070499805607,
+      "grad_norm": 0.6719753806637919,
+      "learning_rate": 9.104317309537469e-07,
+      "loss": 0.0585,
+      "step": 28150
+    },
+    {
+      "epoch": 2.4329344680115774,
+      "grad_norm": 0.6482007708428015,
+      "learning_rate": 9.077501864812476e-07,
+      "loss": 0.0606,
+      "step": 28160
+    },
+    {
+      "epoch": 2.4337984362175473,
+      "grad_norm": 0.6634719349909597,
+      "learning_rate": 9.050722025636427e-07,
+      "loss": 0.0557,
+      "step": 28170
+    },
+    {
+      "epoch": 2.4346624044235172,
+      "grad_norm": 0.6551058466757594,
+      "learning_rate": 9.023977815309714e-07,
+      "loss": 0.0594,
+      "step": 28180
+    },
+    {
+      "epoch": 2.435526372629487,
+      "grad_norm": 0.6744881656835849,
+      "learning_rate": 8.997269257101776e-07,
+      "loss": 0.0607,
+      "step": 28190
+    },
+    {
+      "epoch": 2.436390340835457,
+      "grad_norm": 0.6918028486996344,
+      "learning_rate": 8.970596374250984e-07,
+      "loss": 0.0606,
+      "step": 28200
+    },
+    {
+      "epoch": 2.4372543090414274,
+      "grad_norm": 0.6315554234710237,
+      "learning_rate": 8.94395918996468e-07,
+      "loss": 0.0573,
+      "step": 28210
+    },
+    {
+      "epoch": 2.4381182772473973,
+      "grad_norm": 0.6532510959054035,
+      "learning_rate": 8.917357727419157e-07,
+      "loss": 0.0581,
+      "step": 28220
+    },
+    {
+      "epoch": 2.438982245453367,
+      "grad_norm": 0.6657451241588958,
+      "learning_rate": 8.890792009759624e-07,
+      "loss": 0.0607,
+      "step": 28230
+    },
+    {
+      "epoch": 2.4398462136593375,
+      "grad_norm": 0.6798649150950511,
+      "learning_rate": 8.864262060100182e-07,
+      "loss": 0.0582,
+      "step": 28240
+    },
+    {
+      "epoch": 2.4407101818653074,
+      "grad_norm": 0.6867048576047796,
+      "learning_rate": 8.837767901523808e-07,
+      "loss": 0.0588,
+      "step": 28250
+    },
+    {
+      "epoch": 2.4415741500712773,
+      "grad_norm": 0.6368106790875463,
+      "learning_rate": 8.81130955708237e-07,
+      "loss": 0.0578,
+      "step": 28260
+    },
+    {
+      "epoch": 2.442438118277247,
+      "grad_norm": 0.6514293184898988,
+      "learning_rate": 8.784887049796537e-07,
+      "loss": 0.0553,
+      "step": 28270
+    },
+    {
+      "epoch": 2.4433020864832176,
+      "grad_norm": 0.6861528684747734,
+      "learning_rate": 8.758500402655811e-07,
+      "loss": 0.0583,
+      "step": 28280
+    },
+    {
+      "epoch": 2.4441660546891875,
+      "grad_norm": 0.6408952542356624,
+      "learning_rate": 8.732149638618481e-07,
+      "loss": 0.0587,
+      "step": 28290
+    },
+    {
+      "epoch": 2.4450300228951574,
+      "grad_norm": 0.6583448327327999,
+      "learning_rate": 8.70583478061166e-07,
+      "loss": 0.0599,
+      "step": 28300
+    },
+    {
+      "epoch": 2.4458939911011273,
+      "grad_norm": 0.6841081059724552,
+      "learning_rate": 8.679555851531168e-07,
+      "loss": 0.058,
+      "step": 28310
+    },
+    {
+      "epoch": 2.4467579593070976,
+      "grad_norm": 0.6287400345247842,
+      "learning_rate": 8.653312874241587e-07,
+      "loss": 0.0594,
+      "step": 28320
+    },
+    {
+      "epoch": 2.4476219275130675,
+      "grad_norm": 0.6443569285000627,
+      "learning_rate": 8.627105871576214e-07,
+      "loss": 0.0575,
+      "step": 28330
+    },
+    {
+      "epoch": 2.4484858957190374,
+      "grad_norm": 0.6537062566918245,
+      "learning_rate": 8.600934866337035e-07,
+      "loss": 0.0595,
+      "step": 28340
+    },
+    {
+      "epoch": 2.4493498639250078,
+      "grad_norm": 0.6728442936950617,
+      "learning_rate": 8.57479988129476e-07,
+      "loss": 0.0608,
+      "step": 28350
+    },
+    {
+      "epoch": 2.4502138321309777,
+      "grad_norm": 0.6392937359718057,
+      "learning_rate": 8.548700939188686e-07,
+      "loss": 0.0572,
+      "step": 28360
+    },
+    {
+      "epoch": 2.4510778003369476,
+      "grad_norm": 0.6616002855883878,
+      "learning_rate": 8.522638062726823e-07,
+      "loss": 0.0582,
+      "step": 28370
+    },
+    {
+      "epoch": 2.4519417685429175,
+      "grad_norm": 0.6671488566913149,
+      "learning_rate": 8.496611274585759e-07,
+      "loss": 0.0582,
+      "step": 28380
+    },
+    {
+      "epoch": 2.452805736748888,
+      "grad_norm": 0.6564631235016541,
+      "learning_rate": 8.470620597410689e-07,
+      "loss": 0.061,
+      "step": 28390
+    },
+    {
+      "epoch": 2.4536697049548577,
+      "grad_norm": 0.6637031889827743,
+      "learning_rate": 8.444666053815375e-07,
+      "loss": 0.0622,
+      "step": 28400
+    },
+    {
+      "epoch": 2.4545336731608276,
+      "grad_norm": 0.7176647827953652,
+      "learning_rate": 8.418747666382188e-07,
+      "loss": 0.059,
+      "step": 28410
+    },
+    {
+      "epoch": 2.4553976413667975,
+      "grad_norm": 0.6896783584801283,
+      "learning_rate": 8.392865457662002e-07,
+      "loss": 0.0604,
+      "step": 28420
+    },
+    {
+      "epoch": 2.456261609572768,
+      "grad_norm": 0.6828733344280327,
+      "learning_rate": 8.367019450174208e-07,
+      "loss": 0.0603,
+      "step": 28430
+    },
+    {
+      "epoch": 2.4571255777787377,
+      "grad_norm": 0.6659408708812785,
+      "learning_rate": 8.341209666406724e-07,
+      "loss": 0.0593,
+      "step": 28440
+    },
+    {
+      "epoch": 2.4579895459847076,
+      "grad_norm": 0.6766561239252696,
+      "learning_rate": 8.315436128815918e-07,
+      "loss": 0.059,
+      "step": 28450
+    },
+    {
+      "epoch": 2.458853514190678,
+      "grad_norm": 0.6530791811491967,
+      "learning_rate": 8.289698859826667e-07,
+      "loss": 0.0576,
+      "step": 28460
+    },
+    {
+      "epoch": 2.459717482396648,
+      "grad_norm": 0.6605880314782795,
+      "learning_rate": 8.263997881832258e-07,
+      "loss": 0.0575,
+      "step": 28470
+    },
+    {
+      "epoch": 2.460581450602618,
+      "grad_norm": 0.6717909641695353,
+      "learning_rate": 8.238333217194411e-07,
+      "loss": 0.0578,
+      "step": 28480
+    },
+    {
+      "epoch": 2.4614454188085877,
+      "grad_norm": 0.6542064505610335,
+      "learning_rate": 8.212704888243245e-07,
+      "loss": 0.0576,
+      "step": 28490
+    },
+    {
+      "epoch": 2.462309387014558,
+      "grad_norm": 0.7133803141730897,
+      "learning_rate": 8.187112917277279e-07,
+      "loss": 0.0575,
+      "step": 28500
+    },
+    {
+      "epoch": 2.463173355220528,
+      "grad_norm": 0.6987238128077268,
+      "learning_rate": 8.161557326563374e-07,
+      "loss": 0.0601,
+      "step": 28510
+    },
+    {
+      "epoch": 2.464037323426498,
+      "grad_norm": 0.658559293565418,
+      "learning_rate": 8.136038138336754e-07,
+      "loss": 0.0573,
+      "step": 28520
+    },
+    {
+      "epoch": 2.4649012916324677,
+      "grad_norm": 0.6867165061226779,
+      "learning_rate": 8.110555374800988e-07,
+      "loss": 0.0612,
+      "step": 28530
+    },
+    {
+      "epoch": 2.465765259838438,
+      "grad_norm": 0.6616775930013254,
+      "learning_rate": 8.085109058127916e-07,
+      "loss": 0.0589,
+      "step": 28540
+    },
+    {
+      "epoch": 2.466629228044408,
+      "grad_norm": 0.6748442619397084,
+      "learning_rate": 8.059699210457695e-07,
+      "loss": 0.0602,
+      "step": 28550
+    },
+    {
+      "epoch": 2.467493196250378,
+      "grad_norm": 0.6784319465044278,
+      "learning_rate": 8.034325853898716e-07,
+      "loss": 0.0601,
+      "step": 28560
+    },
+    {
+      "epoch": 2.4683571644563482,
+      "grad_norm": 0.6696709196103713,
+      "learning_rate": 8.008989010527674e-07,
+      "loss": 0.0583,
+      "step": 28570
+    },
+    {
+      "epoch": 2.469221132662318,
+      "grad_norm": 0.6752973689391882,
+      "learning_rate": 7.983688702389447e-07,
+      "loss": 0.0593,
+      "step": 28580
+    },
+    {
+      "epoch": 2.470085100868288,
+      "grad_norm": 0.6121634874033778,
+      "learning_rate": 7.958424951497157e-07,
+      "loss": 0.0586,
+      "step": 28590
+    },
+    {
+      "epoch": 2.470949069074258,
+      "grad_norm": 0.6580181912566206,
+      "learning_rate": 7.933197779832091e-07,
+      "loss": 0.0568,
+      "step": 28600
+    },
+    {
+      "epoch": 2.4718130372802283,
+      "grad_norm": 0.6552538023898506,
+      "learning_rate": 7.908007209343716e-07,
+      "loss": 0.0578,
+      "step": 28610
+    },
+    {
+      "epoch": 2.472677005486198,
+      "grad_norm": 0.6669667599185005,
+      "learning_rate": 7.882853261949692e-07,
+      "loss": 0.0593,
+      "step": 28620
+    },
+    {
+      "epoch": 2.473540973692168,
+      "grad_norm": 0.6340279306362967,
+      "learning_rate": 7.857735959535739e-07,
+      "loss": 0.055,
+      "step": 28630
+    },
+    {
+      "epoch": 2.474404941898138,
+      "grad_norm": 0.6530282367866861,
+      "learning_rate": 7.832655323955773e-07,
+      "loss": 0.0596,
+      "step": 28640
+    },
+    {
+      "epoch": 2.4752689101041083,
+      "grad_norm": 0.66428227074611,
+      "learning_rate": 7.807611377031738e-07,
+      "loss": 0.0576,
+      "step": 28650
+    },
+    {
+      "epoch": 2.4761328783100782,
+      "grad_norm": 0.5941794758457453,
+      "learning_rate": 7.782604140553734e-07,
+      "loss": 0.0583,
+      "step": 28660
+    },
+    {
+      "epoch": 2.476996846516048,
+      "grad_norm": 0.6814535079374869,
+      "learning_rate": 7.757633636279826e-07,
+      "loss": 0.0605,
+      "step": 28670
+    },
+    {
+      "epoch": 2.4778608147220185,
+      "grad_norm": 0.6802629053649004,
+      "learning_rate": 7.732699885936201e-07,
+      "loss": 0.0576,
+      "step": 28680
+    },
+    {
+      "epoch": 2.4787247829279884,
+      "grad_norm": 0.6726871693359314,
+      "learning_rate": 7.707802911217027e-07,
+      "loss": 0.0591,
+      "step": 28690
+    },
+    {
+      "epoch": 2.4795887511339583,
+      "grad_norm": 0.6870407955530046,
+      "learning_rate": 7.682942733784476e-07,
+      "loss": 0.0587,
+      "step": 28700
+    },
+    {
+      "epoch": 2.480452719339928,
+      "grad_norm": 0.6289813857003265,
+      "learning_rate": 7.658119375268714e-07,
+      "loss": 0.0576,
+      "step": 28710
+    },
+    {
+      "epoch": 2.481316687545898,
+      "grad_norm": 0.6398280282296972,
+      "learning_rate": 7.633332857267856e-07,
+      "loss": 0.0582,
+      "step": 28720
+    },
+    {
+      "epoch": 2.4821806557518684,
+      "grad_norm": 0.643099284809275,
+      "learning_rate": 7.608583201348002e-07,
+      "loss": 0.0594,
+      "step": 28730
+    },
+    {
+      "epoch": 2.4830446239578383,
+      "grad_norm": 0.6469507902323824,
+      "learning_rate": 7.583870429043134e-07,
+      "loss": 0.0557,
+      "step": 28740
+    },
+    {
+      "epoch": 2.483908592163808,
+      "grad_norm": 0.664417794823054,
+      "learning_rate": 7.55919456185516e-07,
+      "loss": 0.0563,
+      "step": 28750
+    },
+    {
+      "epoch": 2.4847725603697786,
+      "grad_norm": 0.654464666999886,
+      "learning_rate": 7.534555621253875e-07,
+      "loss": 0.0568,
+      "step": 28760
+    },
+    {
+      "epoch": 2.4856365285757485,
+      "grad_norm": 0.6736856787614522,
+      "learning_rate": 7.509953628676963e-07,
+      "loss": 0.0591,
+      "step": 28770
+    },
+    {
+      "epoch": 2.4865004967817184,
+      "grad_norm": 0.6689201937782084,
+      "learning_rate": 7.485388605529942e-07,
+      "loss": 0.0605,
+      "step": 28780
+    },
+    {
+      "epoch": 2.4873644649876883,
+      "grad_norm": 0.671068501395764,
+      "learning_rate": 7.460860573186168e-07,
+      "loss": 0.0591,
+      "step": 28790
+    },
+    {
+      "epoch": 2.4882284331936586,
+      "grad_norm": 0.6849956789413019,
+      "learning_rate": 7.43636955298681e-07,
+      "loss": 0.0605,
+      "step": 28800
+    },
+    {
+      "epoch": 2.4890924013996285,
+      "grad_norm": 0.6174444958621187,
+      "learning_rate": 7.411915566240835e-07,
+      "loss": 0.0572,
+      "step": 28810
+    },
+    {
+      "epoch": 2.4899563696055984,
+      "grad_norm": 0.6613532292211507,
+      "learning_rate": 7.387498634224988e-07,
+      "loss": 0.0612,
+      "step": 28820
+    },
+    {
+      "epoch": 2.4908203378115683,
+      "grad_norm": 0.6281340387625589,
+      "learning_rate": 7.36311877818377e-07,
+      "loss": 0.057,
+      "step": 28830
+    },
+    {
+      "epoch": 2.4916843060175387,
+      "grad_norm": 0.6472955772427801,
+      "learning_rate": 7.338776019329452e-07,
+      "loss": 0.0595,
+      "step": 28840
+    },
+    {
+      "epoch": 2.4925482742235086,
+      "grad_norm": 0.6465898107352807,
+      "learning_rate": 7.314470378841987e-07,
+      "loss": 0.0579,
+      "step": 28850
+    },
+    {
+      "epoch": 2.4934122424294785,
+      "grad_norm": 0.6872080656870608,
+      "learning_rate": 7.290201877869052e-07,
+      "loss": 0.0577,
+      "step": 28860
+    },
+    {
+      "epoch": 2.494276210635449,
+      "grad_norm": 0.6546298045170468,
+      "learning_rate": 7.26597053752599e-07,
+      "loss": 0.0593,
+      "step": 28870
+    },
+    {
+      "epoch": 2.4951401788414187,
+      "grad_norm": 0.6501124516401879,
+      "learning_rate": 7.241776378895865e-07,
+      "loss": 0.0567,
+      "step": 28880
+    },
+    {
+      "epoch": 2.4960041470473886,
+      "grad_norm": 0.6489435699910041,
+      "learning_rate": 7.217619423029332e-07,
+      "loss": 0.0588,
+      "step": 28890
+    },
+    {
+      "epoch": 2.4968681152533585,
+      "grad_norm": 0.6535221870925897,
+      "learning_rate": 7.193499690944706e-07,
+      "loss": 0.0592,
+      "step": 28900
+    },
+    {
+      "epoch": 2.497732083459329,
+      "grad_norm": 0.6771730869265828,
+      "learning_rate": 7.169417203627898e-07,
+      "loss": 0.0579,
+      "step": 28910
+    },
+    {
+      "epoch": 2.4985960516652987,
+      "grad_norm": 0.6359104906466674,
+      "learning_rate": 7.145371982032423e-07,
+      "loss": 0.0575,
+      "step": 28920
+    },
+    {
+      "epoch": 2.4994600198712686,
+      "grad_norm": 0.6722845799249196,
+      "learning_rate": 7.121364047079405e-07,
+      "loss": 0.059,
+      "step": 28930
+    },
+    {
+      "epoch": 2.5003239880772385,
+      "grad_norm": 0.6749302305814222,
+      "learning_rate": 7.097393419657439e-07,
+      "loss": 0.0584,
+      "step": 28940
+    },
+    {
+      "epoch": 2.501187956283209,
+      "grad_norm": 0.6397819866220034,
+      "learning_rate": 7.07346012062275e-07,
+      "loss": 0.0596,
+      "step": 28950
+    },
+    {
+      "epoch": 2.502051924489179,
+      "grad_norm": 0.649234976789809,
+      "learning_rate": 7.049564170799034e-07,
+      "loss": 0.0588,
+      "step": 28960
+    },
+    {
+      "epoch": 2.5029158926951487,
+      "grad_norm": 0.6194457878662328,
+      "learning_rate": 7.025705590977528e-07,
+      "loss": 0.0586,
+      "step": 28970
+    },
+    {
+      "epoch": 2.503779860901119,
+      "grad_norm": 0.6254725233188696,
+      "learning_rate": 7.001884401916898e-07,
+      "loss": 0.0584,
+      "step": 28980
+    },
+    {
+      "epoch": 2.504643829107089,
+      "grad_norm": 0.6533137706948257,
+      "learning_rate": 6.978100624343332e-07,
+      "loss": 0.0568,
+      "step": 28990
+    },
+    {
+      "epoch": 2.505507797313059,
+      "grad_norm": 0.6726580638707064,
+      "learning_rate": 6.954354278950443e-07,
+      "loss": 0.0582,
+      "step": 29000
+    },
+    {
+      "epoch": 2.5063717655190287,
+      "grad_norm": 0.7037083198887285,
+      "learning_rate": 6.930645386399277e-07,
+      "loss": 0.0614,
+      "step": 29010
+    },
+    {
+      "epoch": 2.507235733724999,
+      "grad_norm": 0.7033536212322138,
+      "learning_rate": 6.906973967318287e-07,
+      "loss": 0.0608,
+      "step": 29020
+    },
+    {
+      "epoch": 2.508099701930969,
+      "grad_norm": 0.6409072157382765,
+      "learning_rate": 6.883340042303333e-07,
+      "loss": 0.0574,
+      "step": 29030
+    },
+    {
+      "epoch": 2.508963670136939,
+      "grad_norm": 0.6445925601552681,
+      "learning_rate": 6.859743631917653e-07,
+      "loss": 0.0583,
+      "step": 29040
+    },
+    {
+      "epoch": 2.509827638342909,
+      "grad_norm": 0.6992038269097037,
+      "learning_rate": 6.836184756691838e-07,
+      "loss": 0.0597,
+      "step": 29050
+    },
+    {
+      "epoch": 2.510691606548879,
+      "grad_norm": 0.6687242847138208,
+      "learning_rate": 6.812663437123823e-07,
+      "loss": 0.058,
+      "step": 29060
+    },
+    {
+      "epoch": 2.511555574754849,
+      "grad_norm": 0.6608479078655375,
+      "learning_rate": 6.789179693678855e-07,
+      "loss": 0.059,
+      "step": 29070
+    },
+    {
+      "epoch": 2.512419542960819,
+      "grad_norm": 0.6961967420294427,
+      "learning_rate": 6.765733546789527e-07,
+      "loss": 0.0596,
+      "step": 29080
+    },
+    {
+      "epoch": 2.5132835111667893,
+      "grad_norm": 0.6380443357234777,
+      "learning_rate": 6.742325016855655e-07,
+      "loss": 0.0579,
+      "step": 29090
+    },
+    {
+      "epoch": 2.514147479372759,
+      "grad_norm": 0.6437761706113116,
+      "learning_rate": 6.718954124244386e-07,
+      "loss": 0.0576,
+      "step": 29100
+    },
+    {
+      "epoch": 2.515011447578729,
+      "grad_norm": 0.7021053853712458,
+      "learning_rate": 6.695620889290095e-07,
+      "loss": 0.058,
+      "step": 29110
+    },
+    {
+      "epoch": 2.515875415784699,
+      "grad_norm": 0.6825429370809767,
+      "learning_rate": 6.672325332294383e-07,
+      "loss": 0.0571,
+      "step": 29120
+    },
+    {
+      "epoch": 2.516739383990669,
+      "grad_norm": 0.679356433703693,
+      "learning_rate": 6.649067473526083e-07,
+      "loss": 0.0574,
+      "step": 29130
+    },
+    {
+      "epoch": 2.517603352196639,
+      "grad_norm": 0.658900017381107,
+      "learning_rate": 6.625847333221213e-07,
+      "loss": 0.0604,
+      "step": 29140
+    },
+    {
+      "epoch": 2.518467320402609,
+      "grad_norm": 0.6650829530549738,
+      "learning_rate": 6.602664931583008e-07,
+      "loss": 0.0577,
+      "step": 29150
+    },
+    {
+      "epoch": 2.519331288608579,
+      "grad_norm": 0.7048816158879708,
+      "learning_rate": 6.579520288781826e-07,
+      "loss": 0.059,
+      "step": 29160
+    },
+    {
+      "epoch": 2.5201952568145494,
+      "grad_norm": 0.6645875550709922,
+      "learning_rate": 6.556413424955188e-07,
+      "loss": 0.0594,
+      "step": 29170
+    },
+    {
+      "epoch": 2.5210592250205193,
+      "grad_norm": 0.6709488583655743,
+      "learning_rate": 6.533344360207744e-07,
+      "loss": 0.0603,
+      "step": 29180
+    },
+    {
+      "epoch": 2.521923193226489,
+      "grad_norm": 0.6671497082111582,
+      "learning_rate": 6.510313114611272e-07,
+      "loss": 0.0593,
+      "step": 29190
+    },
+    {
+      "epoch": 2.5227871614324595,
+      "grad_norm": 0.6281321494275784,
+      "learning_rate": 6.487319708204625e-07,
+      "loss": 0.0571,
+      "step": 29200
+    },
+    {
+      "epoch": 2.5236511296384294,
+      "grad_norm": 0.6805803298969804,
+      "learning_rate": 6.464364160993736e-07,
+      "loss": 0.0593,
+      "step": 29210
+    },
+    {
+      "epoch": 2.5245150978443993,
+      "grad_norm": 0.6850029853982884,
+      "learning_rate": 6.441446492951597e-07,
+      "loss": 0.06,
+      "step": 29220
+    },
+    {
+      "epoch": 2.525379066050369,
+      "grad_norm": 0.6418791460040815,
+      "learning_rate": 6.418566724018232e-07,
+      "loss": 0.0552,
+      "step": 29230
+    },
+    {
+      "epoch": 2.526243034256339,
+      "grad_norm": 0.6482480515608605,
+      "learning_rate": 6.39572487410075e-07,
+      "loss": 0.058,
+      "step": 29240
+    },
+    {
+      "epoch": 2.5271070024623095,
+      "grad_norm": 0.6396174623886076,
+      "learning_rate": 6.372920963073165e-07,
+      "loss": 0.0599,
+      "step": 29250
+    },
+    {
+      "epoch": 2.5279709706682794,
+      "grad_norm": 0.6809497391776808,
+      "learning_rate": 6.350155010776576e-07,
+      "loss": 0.0608,
+      "step": 29260
+    },
+    {
+      "epoch": 2.5288349388742493,
+      "grad_norm": 0.6779821207469485,
+      "learning_rate": 6.32742703701899e-07,
+      "loss": 0.0572,
+      "step": 29270
+    },
+    {
+      "epoch": 2.5296989070802196,
+      "grad_norm": 0.6653538492392551,
+      "learning_rate": 6.304737061575438e-07,
+      "loss": 0.0583,
+      "step": 29280
+    },
+    {
+      "epoch": 2.5305628752861895,
+      "grad_norm": 0.6416808234550777,
+      "learning_rate": 6.282085104187796e-07,
+      "loss": 0.0614,
+      "step": 29290
+    },
+    {
+      "epoch": 2.5314268434921594,
+      "grad_norm": 0.6550871279640734,
+      "learning_rate": 6.259471184564952e-07,
+      "loss": 0.0574,
+      "step": 29300
+    },
+    {
+      "epoch": 2.5322908116981298,
+      "grad_norm": 0.6707751927794339,
+      "learning_rate": 6.236895322382653e-07,
+      "loss": 0.0583,
+      "step": 29310
+    },
+    {
+      "epoch": 2.5331547799040997,
+      "grad_norm": 0.6483868684196846,
+      "learning_rate": 6.214357537283527e-07,
+      "loss": 0.0568,
+      "step": 29320
+    },
+    {
+      "epoch": 2.5340187481100696,
+      "grad_norm": 0.6548519923942192,
+      "learning_rate": 6.191857848877097e-07,
+      "loss": 0.0569,
+      "step": 29330
+    },
+    {
+      "epoch": 2.5348827163160395,
+      "grad_norm": 0.684998353287605,
+      "learning_rate": 6.16939627673972e-07,
+      "loss": 0.0592,
+      "step": 29340
+    },
+    {
+      "epoch": 2.5357466845220094,
+      "grad_norm": 0.6359323968029438,
+      "learning_rate": 6.146972840414623e-07,
+      "loss": 0.0598,
+      "step": 29350
+    },
+    {
+      "epoch": 2.5366106527279797,
+      "grad_norm": 0.6175361169902326,
+      "learning_rate": 6.124587559411782e-07,
+      "loss": 0.0557,
+      "step": 29360
+    },
+    {
+      "epoch": 2.5374746209339496,
+      "grad_norm": 0.6772844589228746,
+      "learning_rate": 6.102240453208052e-07,
+      "loss": 0.0605,
+      "step": 29370
+    },
+    {
+      "epoch": 2.5383385891399195,
+      "grad_norm": 0.6512332430117873,
+      "learning_rate": 6.07993154124702e-07,
+      "loss": 0.0605,
+      "step": 29380
+    },
+    {
+      "epoch": 2.53920255734589,
+      "grad_norm": 0.6989309110098257,
+      "learning_rate": 6.057660842939095e-07,
+      "loss": 0.0592,
+      "step": 29390
+    },
+    {
+      "epoch": 2.5400665255518597,
+      "grad_norm": 0.6773524899173674,
+      "learning_rate": 6.035428377661362e-07,
+      "loss": 0.0589,
+      "step": 29400
+    },
+    {
+      "epoch": 2.5409304937578296,
+      "grad_norm": 0.668998479726973,
+      "learning_rate": 6.013234164757709e-07,
+      "loss": 0.0591,
+      "step": 29410
+    },
+    {
+      "epoch": 2.5417944619638,
+      "grad_norm": 0.6703732268508873,
+      "learning_rate": 5.99107822353871e-07,
+      "loss": 0.057,
+      "step": 29420
+    },
+    {
+      "epoch": 2.54265843016977,
+      "grad_norm": 0.6635245821998534,
+      "learning_rate": 5.968960573281645e-07,
+      "loss": 0.0595,
+      "step": 29430
+    },
+    {
+      "epoch": 2.54352239837574,
+      "grad_norm": 0.6525266456918762,
+      "learning_rate": 5.946881233230473e-07,
+      "loss": 0.0564,
+      "step": 29440
+    },
+    {
+      "epoch": 2.5443863665817097,
+      "grad_norm": 0.687436696980093,
+      "learning_rate": 5.924840222595818e-07,
+      "loss": 0.0583,
+      "step": 29450
+    },
+    {
+      "epoch": 2.5452503347876796,
+      "grad_norm": 0.6639549603653401,
+      "learning_rate": 5.902837560554981e-07,
+      "loss": 0.062,
+      "step": 29460
+    },
+    {
+      "epoch": 2.54611430299365,
+      "grad_norm": 0.664926820131592,
+      "learning_rate": 5.880873266251869e-07,
+      "loss": 0.0595,
+      "step": 29470
+    },
+    {
+      "epoch": 2.54697827119962,
+      "grad_norm": 0.6871249087091793,
+      "learning_rate": 5.858947358797018e-07,
+      "loss": 0.0596,
+      "step": 29480
+    },
+    {
+      "epoch": 2.5478422394055897,
+      "grad_norm": 0.667792891284496,
+      "learning_rate": 5.837059857267546e-07,
+      "loss": 0.058,
+      "step": 29490
+    },
+    {
+      "epoch": 2.54870620761156,
+      "grad_norm": 0.6350186657516332,
+      "learning_rate": 5.815210780707192e-07,
+      "loss": 0.0599,
+      "step": 29500
+    },
+    {
+      "epoch": 2.54957017581753,
+      "grad_norm": 0.6961370444377433,
+      "learning_rate": 5.793400148126233e-07,
+      "loss": 0.0583,
+      "step": 29510
+    },
+    {
+      "epoch": 2.5504341440235,
+      "grad_norm": 0.6484114304314101,
+      "learning_rate": 5.77162797850151e-07,
+      "loss": 0.0579,
+      "step": 29520
+    },
+    {
+      "epoch": 2.5512981122294702,
+      "grad_norm": 0.6489574041467437,
+      "learning_rate": 5.749894290776381e-07,
+      "loss": 0.0581,
+      "step": 29530
+    },
+    {
+      "epoch": 2.55216208043544,
+      "grad_norm": 0.6704512119761233,
+      "learning_rate": 5.728199103860738e-07,
+      "loss": 0.0582,
+      "step": 29540
+    },
+    {
+      "epoch": 2.55302604864141,
+      "grad_norm": 0.6880307479956702,
+      "learning_rate": 5.70654243663099e-07,
+      "loss": 0.0568,
+      "step": 29550
+    },
+    {
+      "epoch": 2.55389001684738,
+      "grad_norm": 0.6705087029929916,
+      "learning_rate": 5.684924307929984e-07,
+      "loss": 0.0571,
+      "step": 29560
+    },
+    {
+      "epoch": 2.55475398505335,
+      "grad_norm": 0.6728745082079638,
+      "learning_rate": 5.663344736567083e-07,
+      "loss": 0.0558,
+      "step": 29570
+    },
+    {
+      "epoch": 2.55561795325932,
+      "grad_norm": 0.6830199354155625,
+      "learning_rate": 5.641803741318069e-07,
+      "loss": 0.0598,
+      "step": 29580
+    },
+    {
+      "epoch": 2.55648192146529,
+      "grad_norm": 0.6522985920938468,
+      "learning_rate": 5.620301340925199e-07,
+      "loss": 0.0551,
+      "step": 29590
+    },
+    {
+      "epoch": 2.55734588967126,
+      "grad_norm": 0.6888286792099857,
+      "learning_rate": 5.598837554097092e-07,
+      "loss": 0.0569,
+      "step": 29600
+    },
+    {
+      "epoch": 2.5582098578772303,
+      "grad_norm": 0.6438224709357185,
+      "learning_rate": 5.577412399508831e-07,
+      "loss": 0.0581,
+      "step": 29610
+    },
+    {
+      "epoch": 2.5590738260832,
+      "grad_norm": 0.7114816339042405,
+      "learning_rate": 5.556025895801847e-07,
+      "loss": 0.0572,
+      "step": 29620
+    },
+    {
+      "epoch": 2.55993779428917,
+      "grad_norm": 0.6307847616008638,
+      "learning_rate": 5.534678061583953e-07,
+      "loss": 0.0565,
+      "step": 29630
+    },
+    {
+      "epoch": 2.56080176249514,
+      "grad_norm": 0.665655599351334,
+      "learning_rate": 5.513368915429318e-07,
+      "loss": 0.0574,
+      "step": 29640
+    },
+    {
+      "epoch": 2.56166573070111,
+      "grad_norm": 0.6160826997111912,
+      "learning_rate": 5.492098475878432e-07,
+      "loss": 0.0559,
+      "step": 29650
+    },
+    {
+      "epoch": 2.5625296989070803,
+      "grad_norm": 0.6430126390669085,
+      "learning_rate": 5.470866761438165e-07,
+      "loss": 0.0571,
+      "step": 29660
+    },
+    {
+      "epoch": 2.56339366711305,
+      "grad_norm": 0.6612250849594098,
+      "learning_rate": 5.449673790581611e-07,
+      "loss": 0.0578,
+      "step": 29670
+    },
+    {
+      "epoch": 2.56425763531902,
+      "grad_norm": 0.6560418863856894,
+      "learning_rate": 5.428519581748215e-07,
+      "loss": 0.0563,
+      "step": 29680
+    },
+    {
+      "epoch": 2.5651216035249904,
+      "grad_norm": 0.6643504836285753,
+      "learning_rate": 5.40740415334367e-07,
+      "loss": 0.0564,
+      "step": 29690
+    },
+    {
+      "epoch": 2.5659855717309603,
+      "grad_norm": 0.6771315503647855,
+      "learning_rate": 5.386327523739954e-07,
+      "loss": 0.0576,
+      "step": 29700
+    },
+    {
+      "epoch": 2.56684953993693,
+      "grad_norm": 0.6363414935150137,
+      "learning_rate": 5.365289711275235e-07,
+      "loss": 0.0576,
+      "step": 29710
+    },
+    {
+      "epoch": 2.5677135081429006,
+      "grad_norm": 0.6531883331299517,
+      "learning_rate": 5.34429073425396e-07,
+      "loss": 0.0591,
+      "step": 29720
+    },
+    {
+      "epoch": 2.5685774763488705,
+      "grad_norm": 0.720546804487095,
+      "learning_rate": 5.323330610946769e-07,
+      "loss": 0.0568,
+      "step": 29730
+    },
+    {
+      "epoch": 2.5694414445548404,
+      "grad_norm": 0.6329942697006272,
+      "learning_rate": 5.302409359590483e-07,
+      "loss": 0.0587,
+      "step": 29740
+    },
+    {
+      "epoch": 2.5703054127608103,
+      "grad_norm": 0.6687880309557287,
+      "learning_rate": 5.281526998388115e-07,
+      "loss": 0.0602,
+      "step": 29750
+    },
+    {
+      "epoch": 2.57116938096678,
+      "grad_norm": 0.6776722689320893,
+      "learning_rate": 5.260683545508827e-07,
+      "loss": 0.0619,
+      "step": 29760
+    },
+    {
+      "epoch": 2.5720333491727505,
+      "grad_norm": 0.6712745107615856,
+      "learning_rate": 5.239879019087957e-07,
+      "loss": 0.0601,
+      "step": 29770
+    },
+    {
+      "epoch": 2.5728973173787204,
+      "grad_norm": 0.6886401833521799,
+      "learning_rate": 5.219113437226946e-07,
+      "loss": 0.0591,
+      "step": 29780
+    },
+    {
+      "epoch": 2.5737612855846903,
+      "grad_norm": 0.6579663523334925,
+      "learning_rate": 5.198386817993367e-07,
+      "loss": 0.0585,
+      "step": 29790
+    },
+    {
+      "epoch": 2.5746252537906607,
+      "grad_norm": 0.6409753322445424,
+      "learning_rate": 5.17769917942087e-07,
+      "loss": 0.0573,
+      "step": 29800
+    },
+    {
+      "epoch": 2.5754892219966306,
+      "grad_norm": 0.6611922702209809,
+      "learning_rate": 5.157050539509228e-07,
+      "loss": 0.059,
+      "step": 29810
+    },
+    {
+      "epoch": 2.5763531902026005,
+      "grad_norm": 0.6499977380134889,
+      "learning_rate": 5.136440916224245e-07,
+      "loss": 0.0572,
+      "step": 29820
+    },
+    {
+      "epoch": 2.577217158408571,
+      "grad_norm": 0.6872934986169129,
+      "learning_rate": 5.11587032749779e-07,
+      "loss": 0.056,
+      "step": 29830
+    },
+    {
+      "epoch": 2.5780811266145407,
+      "grad_norm": 0.6732910099436433,
+      "learning_rate": 5.095338791227783e-07,
+      "loss": 0.0585,
+      "step": 29840
+    },
+    {
+      "epoch": 2.5789450948205106,
+      "grad_norm": 0.6708757861303539,
+      "learning_rate": 5.074846325278127e-07,
+      "loss": 0.0587,
+      "step": 29850
+    },
+    {
+      "epoch": 2.5798090630264805,
+      "grad_norm": 0.6783891979386204,
+      "learning_rate": 5.054392947478798e-07,
+      "loss": 0.056,
+      "step": 29860
+    },
+    {
+      "epoch": 2.5806730312324504,
+      "grad_norm": 0.6931804578103389,
+      "learning_rate": 5.033978675625679e-07,
+      "loss": 0.0572,
+      "step": 29870
+    },
+    {
+      "epoch": 2.5815369994384207,
+      "grad_norm": 0.6481825220806289,
+      "learning_rate": 5.013603527480704e-07,
+      "loss": 0.0581,
+      "step": 29880
+    },
+    {
+      "epoch": 2.5824009676443906,
+      "grad_norm": 0.6817496187695342,
+      "learning_rate": 4.993267520771705e-07,
+      "loss": 0.0566,
+      "step": 29890
+    },
+    {
+      "epoch": 2.5832649358503605,
+      "grad_norm": 0.6841775081681513,
+      "learning_rate": 4.972970673192529e-07,
+      "loss": 0.0588,
+      "step": 29900
+    },
+    {
+      "epoch": 2.584128904056331,
+      "grad_norm": 0.6475734993713166,
+      "learning_rate": 4.952713002402859e-07,
+      "loss": 0.0588,
+      "step": 29910
+    },
+    {
+      "epoch": 2.584992872262301,
+      "grad_norm": 0.6662802149829226,
+      "learning_rate": 4.93249452602837e-07,
+      "loss": 0.0579,
+      "step": 29920
+    },
+    {
+      "epoch": 2.5858568404682707,
+      "grad_norm": 0.6365663998049235,
+      "learning_rate": 4.912315261660611e-07,
+      "loss": 0.058,
+      "step": 29930
+    },
+    {
+      "epoch": 2.586720808674241,
+      "grad_norm": 0.6486599808954889,
+      "learning_rate": 4.892175226856994e-07,
+      "loss": 0.0584,
+      "step": 29940
+    },
+    {
+      "epoch": 2.587584776880211,
+      "grad_norm": 0.6658531724786169,
+      "learning_rate": 4.872074439140817e-07,
+      "loss": 0.0576,
+      "step": 29950
+    },
+    {
+      "epoch": 2.588448745086181,
+      "grad_norm": 0.7116085843786459,
+      "learning_rate": 4.852012916001225e-07,
+      "loss": 0.0594,
+      "step": 29960
+    },
+    {
+      "epoch": 2.5893127132921507,
+      "grad_norm": 0.6671752461840659,
+      "learning_rate": 4.831990674893222e-07,
+      "loss": 0.0573,
+      "step": 29970
+    },
+    {
+      "epoch": 2.5901766814981206,
+      "grad_norm": 0.6356609240356549,
+      "learning_rate": 4.812007733237583e-07,
+      "loss": 0.0582,
+      "step": 29980
+    },
+    {
+      "epoch": 2.591040649704091,
+      "grad_norm": 0.6244169993393772,
+      "learning_rate": 4.792064108420941e-07,
+      "loss": 0.0557,
+      "step": 29990
+    },
+    {
+      "epoch": 2.591904617910061,
+      "grad_norm": 0.6540666502258284,
+      "learning_rate": 4.772159817795685e-07,
+      "loss": 0.0582,
+      "step": 30000
+    },
+    {
+      "epoch": 2.592768586116031,
+      "grad_norm": 0.6358928816445137,
+      "learning_rate": 4.752294878680025e-07,
+      "loss": 0.0576,
+      "step": 30010
+    },
+    {
+      "epoch": 2.593632554322001,
+      "grad_norm": 0.6572956423353731,
+      "learning_rate": 4.7324693083578563e-07,
+      "loss": 0.0584,
+      "step": 30020
+    },
+    {
+      "epoch": 2.594496522527971,
+      "grad_norm": 0.6604128875290302,
+      "learning_rate": 4.7126831240789097e-07,
+      "loss": 0.059,
+      "step": 30030
+    },
+    {
+      "epoch": 2.595360490733941,
+      "grad_norm": 0.656347890741661,
+      "learning_rate": 4.692936343058579e-07,
+      "loss": 0.0584,
+      "step": 30040
+    },
+    {
+      "epoch": 2.5962244589399113,
+      "grad_norm": 0.6991884383243728,
+      "learning_rate": 4.673228982478012e-07,
+      "loss": 0.0579,
+      "step": 30050
+    },
+    {
+      "epoch": 2.597088427145881,
+      "grad_norm": 0.6979127438966883,
+      "learning_rate": 4.653561059484035e-07,
+      "loss": 0.0582,
+      "step": 30060
+    },
+    {
+      "epoch": 2.597952395351851,
+      "grad_norm": 0.674144107741082,
+      "learning_rate": 4.633932591189172e-07,
+      "loss": 0.0581,
+      "step": 30070
+    },
+    {
+      "epoch": 2.598816363557821,
+      "grad_norm": 0.6871709263612947,
+      "learning_rate": 4.6143435946716276e-07,
+      "loss": 0.0614,
+      "step": 30080
+    },
+    {
+      "epoch": 2.599680331763791,
+      "grad_norm": 0.6587668438432055,
+      "learning_rate": 4.594794086975252e-07,
+      "loss": 0.0581,
+      "step": 30090
+    },
+    {
+      "epoch": 2.600544299969761,
+      "grad_norm": 0.6601269223981766,
+      "learning_rate": 4.575284085109527e-07,
+      "loss": 0.0566,
+      "step": 30100
+    },
+    {
+      "epoch": 2.601408268175731,
+      "grad_norm": 0.6767525995745622,
+      "learning_rate": 4.555813606049575e-07,
+      "loss": 0.0554,
+      "step": 30110
+    },
+    {
+      "epoch": 2.602272236381701,
+      "grad_norm": 0.6629565179971297,
+      "learning_rate": 4.5363826667361443e-07,
+      "loss": 0.0564,
+      "step": 30120
+    },
+    {
+      "epoch": 2.6031362045876714,
+      "grad_norm": 0.7046146642092987,
+      "learning_rate": 4.5169912840755505e-07,
+      "loss": 0.0588,
+      "step": 30130
+    },
+    {
+      "epoch": 2.6040001727936413,
+      "grad_norm": 0.6772719079025742,
+      "learning_rate": 4.4976394749397076e-07,
+      "loss": 0.0565,
+      "step": 30140
+    },
+    {
+      "epoch": 2.604864140999611,
+      "grad_norm": 0.6742492402219806,
+      "learning_rate": 4.478327256166104e-07,
+      "loss": 0.0551,
+      "step": 30150
+    },
+    {
+      "epoch": 2.6057281092055815,
+      "grad_norm": 0.670700818891851,
+      "learning_rate": 4.459054644557759e-07,
+      "loss": 0.0572,
+      "step": 30160
+    },
+    {
+      "epoch": 2.6065920774115514,
+      "grad_norm": 0.6811115544062041,
+      "learning_rate": 4.43982165688327e-07,
+      "loss": 0.0583,
+      "step": 30170
+    },
+    {
+      "epoch": 2.6074560456175213,
+      "grad_norm": 0.6771983529090304,
+      "learning_rate": 4.4206283098767067e-07,
+      "loss": 0.0569,
+      "step": 30180
+    },
+    {
+      "epoch": 2.608320013823491,
+      "grad_norm": 0.7017740502312722,
+      "learning_rate": 4.4014746202377e-07,
+      "loss": 0.0585,
+      "step": 30190
+    },
+    {
+      "epoch": 2.609183982029461,
+      "grad_norm": 0.6755203458518678,
+      "learning_rate": 4.3823606046313415e-07,
+      "loss": 0.0567,
+      "step": 30200
+    },
+    {
+      "epoch": 2.6100479502354315,
+      "grad_norm": 0.6772379176657675,
+      "learning_rate": 4.363286279688217e-07,
+      "loss": 0.0582,
+      "step": 30210
+    },
+    {
+      "epoch": 2.6109119184414014,
+      "grad_norm": 0.6695431410155791,
+      "learning_rate": 4.3442516620043674e-07,
+      "loss": 0.0597,
+      "step": 30220
+    },
+    {
+      "epoch": 2.6117758866473713,
+      "grad_norm": 0.6567571285238,
+      "learning_rate": 4.325256768141312e-07,
+      "loss": 0.0574,
+      "step": 30230
+    },
+    {
+      "epoch": 2.6126398548533416,
+      "grad_norm": 0.6599745824932721,
+      "learning_rate": 4.306301614625979e-07,
+      "loss": 0.057,
+      "step": 30240
+    },
+    {
+      "epoch": 2.6135038230593115,
+      "grad_norm": 0.6378082167289888,
+      "learning_rate": 4.287386217950734e-07,
+      "loss": 0.0569,
+      "step": 30250
+    },
+    {
+      "epoch": 2.6143677912652814,
+      "grad_norm": 0.6809435093701421,
+      "learning_rate": 4.268510594573344e-07,
+      "loss": 0.0572,
+      "step": 30260
+    },
+    {
+      "epoch": 2.6152317594712513,
+      "grad_norm": 0.6741354796855408,
+      "learning_rate": 4.249674760916961e-07,
+      "loss": 0.057,
+      "step": 30270
+    },
+    {
+      "epoch": 2.616095727677221,
+      "grad_norm": 0.6663299971055163,
+      "learning_rate": 4.2308787333701697e-07,
+      "loss": 0.0573,
+      "step": 30280
+    },
+    {
+      "epoch": 2.6169596958831916,
+      "grad_norm": 0.6923628741436045,
+      "learning_rate": 4.2121225282868273e-07,
+      "loss": 0.0575,
+      "step": 30290
+    },
+    {
+      "epoch": 2.6178236640891615,
+      "grad_norm": 0.6422042356059123,
+      "learning_rate": 4.193406161986241e-07,
+      "loss": 0.0567,
+      "step": 30300
+    },
+    {
+      "epoch": 2.6186876322951314,
+      "grad_norm": 0.6971541286553281,
+      "learning_rate": 4.174729650752979e-07,
+      "loss": 0.059,
+      "step": 30310
+    },
+    {
+      "epoch": 2.6195516005011017,
+      "grad_norm": 0.668983792614538,
+      "learning_rate": 4.1560930108369925e-07,
+      "loss": 0.0575,
+      "step": 30320
+    },
+    {
+      "epoch": 2.6204155687070716,
+      "grad_norm": 0.6783496859385131,
+      "learning_rate": 4.1374962584534886e-07,
+      "loss": 0.0582,
+      "step": 30330
+    },
+    {
+      "epoch": 2.6212795369130415,
+      "grad_norm": 0.6566315043249467,
+      "learning_rate": 4.1189394097830073e-07,
+      "loss": 0.0587,
+      "step": 30340
+    },
+    {
+      "epoch": 2.622143505119012,
+      "grad_norm": 0.6665224926516164,
+      "learning_rate": 4.1004224809713497e-07,
+      "loss": 0.0569,
+      "step": 30350
+    },
+    {
+      "epoch": 2.6230074733249817,
+      "grad_norm": 0.6610369842339688,
+      "learning_rate": 4.081945488129602e-07,
+      "loss": 0.0566,
+      "step": 30360
+    },
+    {
+      "epoch": 2.6238714415309516,
+      "grad_norm": 0.6659586041314874,
+      "learning_rate": 4.06350844733408e-07,
+      "loss": 0.0584,
+      "step": 30370
+    },
+    {
+      "epoch": 2.6247354097369215,
+      "grad_norm": 0.6998362957468676,
+      "learning_rate": 4.0451113746263426e-07,
+      "loss": 0.0572,
+      "step": 30380
+    },
+    {
+      "epoch": 2.6255993779428914,
+      "grad_norm": 0.6629821943664789,
+      "learning_rate": 4.0267542860132017e-07,
+      "loss": 0.0604,
+      "step": 30390
+    },
+    {
+      "epoch": 2.626463346148862,
+      "grad_norm": 0.6851188724926308,
+      "learning_rate": 4.008437197466647e-07,
+      "loss": 0.0576,
+      "step": 30400
+    },
+    {
+      "epoch": 2.6273273143548317,
+      "grad_norm": 0.6358552887813863,
+      "learning_rate": 3.990160124923875e-07,
+      "loss": 0.0591,
+      "step": 30410
+    },
+    {
+      "epoch": 2.6281912825608016,
+      "grad_norm": 0.6822769123702047,
+      "learning_rate": 3.9719230842872714e-07,
+      "loss": 0.058,
+      "step": 30420
+    },
+    {
+      "epoch": 2.629055250766772,
+      "grad_norm": 0.6760968470919131,
+      "learning_rate": 3.9537260914243924e-07,
+      "loss": 0.058,
+      "step": 30430
+    },
+    {
+      "epoch": 2.629919218972742,
+      "grad_norm": 0.679594017303257,
+      "learning_rate": 3.9355691621679403e-07,
+      "loss": 0.059,
+      "step": 30440
+    },
+    {
+      "epoch": 2.6307831871787117,
+      "grad_norm": 0.6778234438718367,
+      "learning_rate": 3.9174523123157617e-07,
+      "loss": 0.0579,
+      "step": 30450
+    },
+    {
+      "epoch": 2.631647155384682,
+      "grad_norm": 0.7006233738319552,
+      "learning_rate": 3.8993755576308413e-07,
+      "loss": 0.0608,
+      "step": 30460
+    },
+    {
+      "epoch": 2.632511123590652,
+      "grad_norm": 0.6485879407092817,
+      "learning_rate": 3.8813389138412595e-07,
+      "loss": 0.0551,
+      "step": 30470
+    },
+    {
+      "epoch": 2.633375091796622,
+      "grad_norm": 0.688314346867616,
+      "learning_rate": 3.863342396640213e-07,
+      "loss": 0.0578,
+      "step": 30480
+    },
+    {
+      "epoch": 2.634239060002592,
+      "grad_norm": 0.6914084623518867,
+      "learning_rate": 3.845386021685971e-07,
+      "loss": 0.0587,
+      "step": 30490
+    },
+    {
+      "epoch": 2.6351030282085617,
+      "grad_norm": 0.686323558943809,
+      "learning_rate": 3.827469804601908e-07,
+      "loss": 0.0569,
+      "step": 30500
+    },
+    {
+      "epoch": 2.635966996414532,
+      "grad_norm": 0.6703731117375387,
+      "learning_rate": 3.8095937609764157e-07,
+      "loss": 0.0581,
+      "step": 30510
+    },
+    {
+      "epoch": 2.636830964620502,
+      "grad_norm": 0.6690328063750695,
+      "learning_rate": 3.791757906362958e-07,
+      "loss": 0.0572,
+      "step": 30520
+    },
+    {
+      "epoch": 2.637694932826472,
+      "grad_norm": 0.6455426316238811,
+      "learning_rate": 3.7739622562800224e-07,
+      "loss": 0.0565,
+      "step": 30530
+    },
+    {
+      "epoch": 2.638558901032442,
+      "grad_norm": 0.6888587274112533,
+      "learning_rate": 3.7562068262111286e-07,
+      "loss": 0.0581,
+      "step": 30540
+    },
+    {
+      "epoch": 2.639422869238412,
+      "grad_norm": 0.6825385826605149,
+      "learning_rate": 3.738491631604779e-07,
+      "loss": 0.0598,
+      "step": 30550
+    },
+    {
+      "epoch": 2.640286837444382,
+      "grad_norm": 0.690570583937555,
+      "learning_rate": 3.72081668787449e-07,
+      "loss": 0.0562,
+      "step": 30560
+    },
+    {
+      "epoch": 2.6411508056503523,
+      "grad_norm": 0.6766329158244576,
+      "learning_rate": 3.703182010398748e-07,
+      "loss": 0.058,
+      "step": 30570
+    },
+    {
+      "epoch": 2.642014773856322,
+      "grad_norm": 0.67715272342871,
+      "learning_rate": 3.6855876145209914e-07,
+      "loss": 0.0575,
+      "step": 30580
+    },
+    {
+      "epoch": 2.642878742062292,
+      "grad_norm": 0.6789455137223994,
+      "learning_rate": 3.668033515549646e-07,
+      "loss": 0.0574,
+      "step": 30590
+    },
+    {
+      "epoch": 2.643742710268262,
+      "grad_norm": 0.6383316891492063,
+      "learning_rate": 3.6505197287580285e-07,
+      "loss": 0.0582,
+      "step": 30600
+    },
+    {
+      "epoch": 2.644606678474232,
+      "grad_norm": 0.6629664084374244,
+      "learning_rate": 3.633046269384427e-07,
+      "loss": 0.0585,
+      "step": 30610
+    },
+    {
+      "epoch": 2.6454706466802023,
+      "grad_norm": 0.6913305154208267,
+      "learning_rate": 3.615613152632008e-07,
+      "loss": 0.0585,
+      "step": 30620
+    },
+    {
+      "epoch": 2.646334614886172,
+      "grad_norm": 0.6718079255061178,
+      "learning_rate": 3.5982203936688776e-07,
+      "loss": 0.0572,
+      "step": 30630
+    },
+    {
+      "epoch": 2.647198583092142,
+      "grad_norm": 0.6686806554991838,
+      "learning_rate": 3.5808680076279577e-07,
+      "loss": 0.0553,
+      "step": 30640
+    },
+    {
+      "epoch": 2.6480625512981124,
+      "grad_norm": 0.6736141930150831,
+      "learning_rate": 3.563556009607122e-07,
+      "loss": 0.0593,
+      "step": 30650
+    },
+    {
+      "epoch": 2.6489265195040823,
+      "grad_norm": 0.6240399939775275,
+      "learning_rate": 3.546284414669055e-07,
+      "loss": 0.0564,
+      "step": 30660
+    },
+    {
+      "epoch": 2.649790487710052,
+      "grad_norm": 0.6560981438220517,
+      "learning_rate": 3.5290532378412956e-07,
+      "loss": 0.0565,
+      "step": 30670
+    },
+    {
+      "epoch": 2.6506544559160226,
+      "grad_norm": 0.6680071817104607,
+      "learning_rate": 3.5118624941162303e-07,
+      "loss": 0.0566,
+      "step": 30680
+    },
+    {
+      "epoch": 2.6515184241219925,
+      "grad_norm": 0.6625073324317056,
+      "learning_rate": 3.4947121984510393e-07,
+      "loss": 0.0581,
+      "step": 30690
+    },
+    {
+      "epoch": 2.6523823923279624,
+      "grad_norm": 0.6807614072790078,
+      "learning_rate": 3.477602365767746e-07,
+      "loss": 0.0559,
+      "step": 30700
+    },
+    {
+      "epoch": 2.6532463605339323,
+      "grad_norm": 0.6908495010112696,
+      "learning_rate": 3.460533010953138e-07,
+      "loss": 0.0585,
+      "step": 30710
+    },
+    {
+      "epoch": 2.654110328739902,
+      "grad_norm": 0.6892337877997854,
+      "learning_rate": 3.4435041488588063e-07,
+      "loss": 0.0581,
+      "step": 30720
+    },
+    {
+      "epoch": 2.6549742969458725,
+      "grad_norm": 0.6433255896995709,
+      "learning_rate": 3.4265157943010774e-07,
+      "loss": 0.0557,
+      "step": 30730
+    },
+    {
+      "epoch": 2.6558382651518424,
+      "grad_norm": 0.6486141164989712,
+      "learning_rate": 3.4095679620610834e-07,
+      "loss": 0.0585,
+      "step": 30740
+    },
+    {
+      "epoch": 2.6567022333578123,
+      "grad_norm": 0.6826727156921892,
+      "learning_rate": 3.3926606668846395e-07,
+      "loss": 0.058,
+      "step": 30750
+    },
+    {
+      "epoch": 2.6575662015637826,
+      "grad_norm": 0.658775821207378,
+      "learning_rate": 3.375793923482351e-07,
+      "loss": 0.0569,
+      "step": 30760
+    },
+    {
+      "epoch": 2.6584301697697525,
+      "grad_norm": 0.7123926912059468,
+      "learning_rate": 3.3589677465294957e-07,
+      "loss": 0.0544,
+      "step": 30770
+    },
+    {
+      "epoch": 2.6592941379757224,
+      "grad_norm": 0.693975415540836,
+      "learning_rate": 3.3421821506660736e-07,
+      "loss": 0.0586,
+      "step": 30780
+    },
+    {
+      "epoch": 2.6601581061816924,
+      "grad_norm": 0.66409621627837,
+      "learning_rate": 3.3254371504967744e-07,
+      "loss": 0.057,
+      "step": 30790
+    },
+    {
+      "epoch": 2.6610220743876627,
+      "grad_norm": 0.6408974196362415,
+      "learning_rate": 3.3087327605909603e-07,
+      "loss": 0.058,
+      "step": 30800
+    },
+    {
+      "epoch": 2.6618860425936326,
+      "grad_norm": 0.6548935799283724,
+      "learning_rate": 3.292068995482672e-07,
+      "loss": 0.0583,
+      "step": 30810
+    },
+    {
+      "epoch": 2.6627500107996025,
+      "grad_norm": 0.6114378558059209,
+      "learning_rate": 3.2754458696705957e-07,
+      "loss": 0.0583,
+      "step": 30820
+    },
+    {
+      "epoch": 2.6636139790055724,
+      "grad_norm": 0.6532677500781641,
+      "learning_rate": 3.2588633976180616e-07,
+      "loss": 0.0565,
+      "step": 30830
+    },
+    {
+      "epoch": 2.6644779472115427,
+      "grad_norm": 0.6621994992284654,
+      "learning_rate": 3.242321593753017e-07,
+      "loss": 0.0578,
+      "step": 30840
+    },
+    {
+      "epoch": 2.6653419154175126,
+      "grad_norm": 0.6659908441376852,
+      "learning_rate": 3.225820472468044e-07,
+      "loss": 0.0548,
+      "step": 30850
+    },
+    {
+      "epoch": 2.6662058836234825,
+      "grad_norm": 0.6884289390924854,
+      "learning_rate": 3.2093600481203135e-07,
+      "loss": 0.0566,
+      "step": 30860
+    },
+    {
+      "epoch": 2.667069851829453,
+      "grad_norm": 0.662687602293797,
+      "learning_rate": 3.1929403350315913e-07,
+      "loss": 0.0578,
+      "step": 30870
+    },
+    {
+      "epoch": 2.667933820035423,
+      "grad_norm": 0.6549345644376964,
+      "learning_rate": 3.176561347488227e-07,
+      "loss": 0.0571,
+      "step": 30880
+    },
+    {
+      "epoch": 2.6687977882413927,
+      "grad_norm": 0.6848731424249099,
+      "learning_rate": 3.1602230997411166e-07,
+      "loss": 0.0591,
+      "step": 30890
+    },
+    {
+      "epoch": 2.6696617564473626,
+      "grad_norm": 0.6238420529746252,
+      "learning_rate": 3.1439256060057486e-07,
+      "loss": 0.0594,
+      "step": 30900
+    },
+    {
+      "epoch": 2.6705257246533325,
+      "grad_norm": 0.6593033839201444,
+      "learning_rate": 3.1276688804620983e-07,
+      "loss": 0.058,
+      "step": 30910
+    },
+    {
+      "epoch": 2.671389692859303,
+      "grad_norm": 0.6324222514200223,
+      "learning_rate": 3.111452937254722e-07,
+      "loss": 0.0555,
+      "step": 30920
+    },
+    {
+      "epoch": 2.6722536610652727,
+      "grad_norm": 0.6712429491334161,
+      "learning_rate": 3.095277790492646e-07,
+      "loss": 0.059,
+      "step": 30930
+    },
+    {
+      "epoch": 2.6731176292712426,
+      "grad_norm": 0.6485249395528295,
+      "learning_rate": 3.0791434542494615e-07,
+      "loss": 0.057,
+      "step": 30940
+    },
+    {
+      "epoch": 2.673981597477213,
+      "grad_norm": 0.6944594330088194,
+      "learning_rate": 3.063049942563173e-07,
+      "loss": 0.0562,
+      "step": 30950
+    },
+    {
+      "epoch": 2.674845565683183,
+      "grad_norm": 0.6280925503430318,
+      "learning_rate": 3.04699726943633e-07,
+      "loss": 0.056,
+      "step": 30960
+    },
+    {
+      "epoch": 2.675709533889153,
+      "grad_norm": 0.7066752206732005,
+      "learning_rate": 3.0309854488359213e-07,
+      "loss": 0.0578,
+      "step": 30970
+    },
+    {
+      "epoch": 2.676573502095123,
+      "grad_norm": 0.6903537649768017,
+      "learning_rate": 3.015014494693391e-07,
+      "loss": 0.0554,
+      "step": 30980
+    },
+    {
+      "epoch": 2.677437470301093,
+      "grad_norm": 0.6860432662236696,
+      "learning_rate": 2.9990844209046355e-07,
+      "loss": 0.057,
+      "step": 30990
+    },
+    {
+      "epoch": 2.678301438507063,
+      "grad_norm": 0.6777716875811373,
+      "learning_rate": 2.983195241329967e-07,
+      "loss": 0.0569,
+      "step": 31000
+    },
+    {
+      "epoch": 2.679165406713033,
+      "grad_norm": 0.6624799407879325,
+      "learning_rate": 2.967346969794138e-07,
+      "loss": 0.0558,
+      "step": 31010
+    },
+    {
+      "epoch": 2.6800293749190027,
+      "grad_norm": 0.6752161383911821,
+      "learning_rate": 2.9515396200862964e-07,
+      "loss": 0.0567,
+      "step": 31020
+    },
+    {
+      "epoch": 2.680893343124973,
+      "grad_norm": 0.6767628777851237,
+      "learning_rate": 2.935773205959985e-07,
+      "loss": 0.0583,
+      "step": 31030
+    },
+    {
+      "epoch": 2.681757311330943,
+      "grad_norm": 0.6467792303059987,
+      "learning_rate": 2.9200477411331197e-07,
+      "loss": 0.0576,
+      "step": 31040
+    },
+    {
+      "epoch": 2.682621279536913,
+      "grad_norm": 0.6338846148774004,
+      "learning_rate": 2.904363239288022e-07,
+      "loss": 0.0557,
+      "step": 31050
+    },
+    {
+      "epoch": 2.683485247742883,
+      "grad_norm": 0.6403686317193569,
+      "learning_rate": 2.8887197140713144e-07,
+      "loss": 0.0581,
+      "step": 31060
+    },
+    {
+      "epoch": 2.684349215948853,
+      "grad_norm": 0.6463691594607338,
+      "learning_rate": 2.873117179094037e-07,
+      "loss": 0.0608,
+      "step": 31070
+    },
+    {
+      "epoch": 2.685213184154823,
+      "grad_norm": 0.6908061227463191,
+      "learning_rate": 2.857555647931509e-07,
+      "loss": 0.0581,
+      "step": 31080
+    },
+    {
+      "epoch": 2.6860771523607934,
+      "grad_norm": 0.6817739973491456,
+      "learning_rate": 2.8420351341234e-07,
+      "loss": 0.0579,
+      "step": 31090
+    },
+    {
+      "epoch": 2.6869411205667633,
+      "grad_norm": 0.7034592398074475,
+      "learning_rate": 2.8265556511736846e-07,
+      "loss": 0.0578,
+      "step": 31100
+    },
+    {
+      "epoch": 2.687805088772733,
+      "grad_norm": 0.7360862346410634,
+      "learning_rate": 2.811117212550629e-07,
+      "loss": 0.0572,
+      "step": 31110
+    },
+    {
+      "epoch": 2.688669056978703,
+      "grad_norm": 0.6841861146827901,
+      "learning_rate": 2.7957198316868164e-07,
+      "loss": 0.0592,
+      "step": 31120
+    },
+    {
+      "epoch": 2.689533025184673,
+      "grad_norm": 0.6698510034953965,
+      "learning_rate": 2.7803635219790736e-07,
+      "loss": 0.0585,
+      "step": 31130
+    },
+    {
+      "epoch": 2.6903969933906433,
+      "grad_norm": 0.6243962845040457,
+      "learning_rate": 2.765048296788514e-07,
+      "loss": 0.0563,
+      "step": 31140
+    },
+    {
+      "epoch": 2.691260961596613,
+      "grad_norm": 0.6725898416366625,
+      "learning_rate": 2.749774169440489e-07,
+      "loss": 0.0578,
+      "step": 31150
+    },
+    {
+      "epoch": 2.692124929802583,
+      "grad_norm": 0.6513848193495176,
+      "learning_rate": 2.7345411532246126e-07,
+      "loss": 0.0585,
+      "step": 31160
+    },
+    {
+      "epoch": 2.6929888980085535,
+      "grad_norm": 0.6539524109515839,
+      "learning_rate": 2.719349261394705e-07,
+      "loss": 0.0596,
+      "step": 31170
+    },
+    {
+      "epoch": 2.6938528662145234,
+      "grad_norm": 0.6501080679911405,
+      "learning_rate": 2.7041985071688257e-07,
+      "loss": 0.0562,
+      "step": 31180
+    },
+    {
+      "epoch": 2.6947168344204933,
+      "grad_norm": 0.680056582828761,
+      "learning_rate": 2.6890889037292255e-07,
+      "loss": 0.0589,
+      "step": 31190
+    },
+    {
+      "epoch": 2.6955808026264636,
+      "grad_norm": 0.6577180730454457,
+      "learning_rate": 2.674020464222349e-07,
+      "loss": 0.0573,
+      "step": 31200
+    },
+    {
+      "epoch": 2.6964447708324335,
+      "grad_norm": 0.6778652955187711,
+      "learning_rate": 2.658993201758864e-07,
+      "loss": 0.0557,
+      "step": 31210
+    },
+    {
+      "epoch": 2.6973087390384034,
+      "grad_norm": 0.6401257423410647,
+      "learning_rate": 2.6440071294135504e-07,
+      "loss": 0.0572,
+      "step": 31220
+    },
+    {
+      "epoch": 2.6981727072443733,
+      "grad_norm": 0.6664094217724789,
+      "learning_rate": 2.629062260225396e-07,
+      "loss": 0.0564,
+      "step": 31230
+    },
+    {
+      "epoch": 2.699036675450343,
+      "grad_norm": 0.7071794832969107,
+      "learning_rate": 2.6141586071975146e-07,
+      "loss": 0.0577,
+      "step": 31240
+    },
+    {
+      "epoch": 2.6999006436563135,
+      "grad_norm": 0.6731307049372357,
+      "learning_rate": 2.5992961832971897e-07,
+      "loss": 0.0587,
+      "step": 31250
+    },
+    {
+      "epoch": 2.7007646118622834,
+      "grad_norm": 0.7126246677195216,
+      "learning_rate": 2.584475001455783e-07,
+      "loss": 0.0567,
+      "step": 31260
+    },
+    {
+      "epoch": 2.7016285800682533,
+      "grad_norm": 0.6443499323094868,
+      "learning_rate": 2.5696950745688175e-07,
+      "loss": 0.0607,
+      "step": 31270
+    },
+    {
+      "epoch": 2.7024925482742237,
+      "grad_norm": 0.6613790530887171,
+      "learning_rate": 2.554956415495902e-07,
+      "loss": 0.056,
+      "step": 31280
+    },
+    {
+      "epoch": 2.7033565164801936,
+      "grad_norm": 0.6526390569684382,
+      "learning_rate": 2.540259037060738e-07,
+      "loss": 0.0558,
+      "step": 31290
+    },
+    {
+      "epoch": 2.7042204846861635,
+      "grad_norm": 0.6491466055505781,
+      "learning_rate": 2.525602952051115e-07,
+      "loss": 0.0565,
+      "step": 31300
+    },
+    {
+      "epoch": 2.705084452892134,
+      "grad_norm": 0.6477798234747194,
+      "learning_rate": 2.510988173218881e-07,
+      "loss": 0.0566,
+      "step": 31310
+    },
+    {
+      "epoch": 2.7059484210981037,
+      "grad_norm": 0.678937957222222,
+      "learning_rate": 2.496414713279988e-07,
+      "loss": 0.0594,
+      "step": 31320
+    },
+    {
+      "epoch": 2.7068123893040736,
+      "grad_norm": 0.6481312566936305,
+      "learning_rate": 2.481882584914369e-07,
+      "loss": 0.0559,
+      "step": 31330
+    },
+    {
+      "epoch": 2.7076763575100435,
+      "grad_norm": 0.6400669772772863,
+      "learning_rate": 2.467391800766056e-07,
+      "loss": 0.055,
+      "step": 31340
+    },
+    {
+      "epoch": 2.7085403257160134,
+      "grad_norm": 0.6899428068381959,
+      "learning_rate": 2.452942373443068e-07,
+      "loss": 0.0598,
+      "step": 31350
+    },
+    {
+      "epoch": 2.709404293921984,
+      "grad_norm": 0.7131334900456067,
+      "learning_rate": 2.438534315517482e-07,
+      "loss": 0.0579,
+      "step": 31360
+    },
+    {
+      "epoch": 2.7102682621279537,
+      "grad_norm": 0.6774543991643118,
+      "learning_rate": 2.424167639525327e-07,
+      "loss": 0.0574,
+      "step": 31370
+    },
+    {
+      "epoch": 2.7111322303339236,
+      "grad_norm": 0.6601832620044917,
+      "learning_rate": 2.40984235796668e-07,
+      "loss": 0.0566,
+      "step": 31380
+    },
+    {
+      "epoch": 2.711996198539894,
+      "grad_norm": 0.6510071598504059,
+      "learning_rate": 2.3955584833055655e-07,
+      "loss": 0.0571,
+      "step": 31390
+    },
+    {
+      "epoch": 2.712860166745864,
+      "grad_norm": 0.686821661165841,
+      "learning_rate": 2.381316027969993e-07,
+      "loss": 0.0583,
+      "step": 31400
+    },
+    {
+      "epoch": 2.7137241349518337,
+      "grad_norm": 0.6850970538518514,
+      "learning_rate": 2.3671150043519388e-07,
+      "loss": 0.0573,
+      "step": 31410
+    },
+    {
+      "epoch": 2.7145881031578036,
+      "grad_norm": 0.6976938553504183,
+      "learning_rate": 2.3529554248073127e-07,
+      "loss": 0.0571,
+      "step": 31420
+    },
+    {
+      "epoch": 2.715452071363774,
+      "grad_norm": 0.6707899675092523,
+      "learning_rate": 2.3388373016559995e-07,
+      "loss": 0.054,
+      "step": 31430
+    },
+    {
+      "epoch": 2.716316039569744,
+      "grad_norm": 0.6525033797859318,
+      "learning_rate": 2.324760647181784e-07,
+      "loss": 0.0563,
+      "step": 31440
+    },
+    {
+      "epoch": 2.717180007775714,
+      "grad_norm": 0.6712410318989885,
+      "learning_rate": 2.3107254736323747e-07,
+      "loss": 0.0584,
+      "step": 31450
+    },
+    {
+      "epoch": 2.7180439759816837,
+      "grad_norm": 0.6489822837447691,
+      "learning_rate": 2.2967317932193868e-07,
+      "loss": 0.0568,
+      "step": 31460
+    },
+    {
+      "epoch": 2.718907944187654,
+      "grad_norm": 0.6943226651728736,
+      "learning_rate": 2.282779618118358e-07,
+      "loss": 0.0566,
+      "step": 31470
+    },
+    {
+      "epoch": 2.719771912393624,
+      "grad_norm": 0.7089036129137106,
+      "learning_rate": 2.2688689604686899e-07,
+      "loss": 0.0576,
+      "step": 31480
+    },
+    {
+      "epoch": 2.720635880599594,
+      "grad_norm": 0.6762859711461554,
+      "learning_rate": 2.254999832373661e-07,
+      "loss": 0.0571,
+      "step": 31490
+    },
+    {
+      "epoch": 2.721499848805564,
+      "grad_norm": 0.6544346615382407,
+      "learning_rate": 2.2411722459004192e-07,
+      "loss": 0.0566,
+      "step": 31500
+    },
+    {
+      "epoch": 2.722363817011534,
+      "grad_norm": 0.668302454393643,
+      "learning_rate": 2.2273862130799684e-07,
+      "loss": 0.0569,
+      "step": 31510
+    },
+    {
+      "epoch": 2.723227785217504,
+      "grad_norm": 0.6942525287151675,
+      "learning_rate": 2.2136417459071802e-07,
+      "loss": 0.0566,
+      "step": 31520
+    },
+    {
+      "epoch": 2.724091753423474,
+      "grad_norm": 0.6509997629051298,
+      "learning_rate": 2.199938856340711e-07,
+      "loss": 0.0574,
+      "step": 31530
+    },
+    {
+      "epoch": 2.7249557216294438,
+      "grad_norm": 0.6723635409838065,
+      "learning_rate": 2.1862775563030903e-07,
+      "loss": 0.0596,
+      "step": 31540
+    },
+    {
+      "epoch": 2.725819689835414,
+      "grad_norm": 0.670764933007711,
+      "learning_rate": 2.1726578576806324e-07,
+      "loss": 0.0551,
+      "step": 31550
+    },
+    {
+      "epoch": 2.726683658041384,
+      "grad_norm": 0.6629743678908288,
+      "learning_rate": 2.1590797723234802e-07,
+      "loss": 0.0553,
+      "step": 31560
+    },
+    {
+      "epoch": 2.727547626247354,
+      "grad_norm": 0.6327215531773591,
+      "learning_rate": 2.145543312045534e-07,
+      "loss": 0.0566,
+      "step": 31570
+    },
+    {
+      "epoch": 2.7284115944533243,
+      "grad_norm": 0.6781526282634827,
+      "learning_rate": 2.1320484886245163e-07,
+      "loss": 0.0585,
+      "step": 31580
+    },
+    {
+      "epoch": 2.729275562659294,
+      "grad_norm": 0.6622262488922581,
+      "learning_rate": 2.1185953138019022e-07,
+      "loss": 0.0541,
+      "step": 31590
+    },
+    {
+      "epoch": 2.730139530865264,
+      "grad_norm": 0.667475826799447,
+      "learning_rate": 2.1051837992829226e-07,
+      "loss": 0.0548,
+      "step": 31600
+    },
+    {
+      "epoch": 2.7310034990712344,
+      "grad_norm": 0.657088121258104,
+      "learning_rate": 2.091813956736577e-07,
+      "loss": 0.0568,
+      "step": 31610
+    },
+    {
+      "epoch": 2.7318674672772043,
+      "grad_norm": 0.6498889038053293,
+      "learning_rate": 2.0784857977955931e-07,
+      "loss": 0.057,
+      "step": 31620
+    },
+    {
+      "epoch": 2.732731435483174,
+      "grad_norm": 0.6733409297948825,
+      "learning_rate": 2.0651993340564614e-07,
+      "loss": 0.0584,
+      "step": 31630
+    },
+    {
+      "epoch": 2.733595403689144,
+      "grad_norm": 0.6852886398578109,
+      "learning_rate": 2.0519545770793348e-07,
+      "loss": 0.0574,
+      "step": 31640
+    },
+    {
+      "epoch": 2.734459371895114,
+      "grad_norm": 0.6792628939751382,
+      "learning_rate": 2.0387515383881452e-07,
+      "loss": 0.0572,
+      "step": 31650
+    },
+    {
+      "epoch": 2.7353233401010844,
+      "grad_norm": 0.6599160690463196,
+      "learning_rate": 2.025590229470481e-07,
+      "loss": 0.0566,
+      "step": 31660
+    },
+    {
+      "epoch": 2.7361873083070543,
+      "grad_norm": 0.629698984875205,
+      "learning_rate": 2.0124706617776546e-07,
+      "loss": 0.0568,
+      "step": 31670
+    },
+    {
+      "epoch": 2.737051276513024,
+      "grad_norm": 0.6599918753484036,
+      "learning_rate": 1.999392846724618e-07,
+      "loss": 0.0568,
+      "step": 31680
+    },
+    {
+      "epoch": 2.7379152447189945,
+      "grad_norm": 0.6543758042897919,
+      "learning_rate": 1.986356795690042e-07,
+      "loss": 0.0556,
+      "step": 31690
+    },
+    {
+      "epoch": 2.7387792129249644,
+      "grad_norm": 0.6784143602920379,
+      "learning_rate": 1.973362520016231e-07,
+      "loss": 0.0582,
+      "step": 31700
+    },
+    {
+      "epoch": 2.7396431811309343,
+      "grad_norm": 0.6903678628469099,
+      "learning_rate": 1.9604100310091523e-07,
+      "loss": 0.0581,
+      "step": 31710
+    },
+    {
+      "epoch": 2.7405071493369046,
+      "grad_norm": 0.6463342523156203,
+      "learning_rate": 1.947499339938408e-07,
+      "loss": 0.0567,
+      "step": 31720
+    },
+    {
+      "epoch": 2.7413711175428745,
+      "grad_norm": 0.7025200315096658,
+      "learning_rate": 1.93463045803724e-07,
+      "loss": 0.0577,
+      "step": 31730
+    },
+    {
+      "epoch": 2.7422350857488444,
+      "grad_norm": 0.6884195548334803,
+      "learning_rate": 1.9218033965025195e-07,
+      "loss": 0.0578,
+      "step": 31740
+    },
+    {
+      "epoch": 2.7430990539548143,
+      "grad_norm": 0.6444704425358405,
+      "learning_rate": 1.9090181664947137e-07,
+      "loss": 0.0575,
+      "step": 31750
+    },
+    {
+      "epoch": 2.7439630221607842,
+      "grad_norm": 0.6725158461214576,
+      "learning_rate": 1.8962747791379023e-07,
+      "loss": 0.0572,
+      "step": 31760
+    },
+    {
+      "epoch": 2.7448269903667546,
+      "grad_norm": 0.6760315296837612,
+      "learning_rate": 1.8835732455197596e-07,
+      "loss": 0.0571,
+      "step": 31770
+    },
+    {
+      "epoch": 2.7456909585727245,
+      "grad_norm": 0.6515448653414146,
+      "learning_rate": 1.870913576691552e-07,
+      "loss": 0.0574,
+      "step": 31780
+    },
+    {
+      "epoch": 2.7465549267786944,
+      "grad_norm": 0.702263272691988,
+      "learning_rate": 1.8582957836681015e-07,
+      "loss": 0.0576,
+      "step": 31790
+    },
+    {
+      "epoch": 2.7474188949846647,
+      "grad_norm": 0.674403943690754,
+      "learning_rate": 1.845719877427815e-07,
+      "loss": 0.0567,
+      "step": 31800
+    },
+    {
+      "epoch": 2.7482828631906346,
+      "grad_norm": 0.6919511526668807,
+      "learning_rate": 1.8331858689126348e-07,
+      "loss": 0.0577,
+      "step": 31810
+    },
+    {
+      "epoch": 2.7491468313966045,
+      "grad_norm": 0.669158032499984,
+      "learning_rate": 1.820693769028059e-07,
+      "loss": 0.057,
+      "step": 31820
+    },
+    {
+      "epoch": 2.750010799602575,
+      "grad_norm": 0.6565133834585644,
+      "learning_rate": 1.8082435886431383e-07,
+      "loss": 0.0549,
+      "step": 31830
+    },
+    {
+      "epoch": 2.750874767808545,
+      "grad_norm": 0.6700558974120085,
+      "learning_rate": 1.7958353385904126e-07,
+      "loss": 0.0581,
+      "step": 31840
+    },
+    {
+      "epoch": 2.7517387360145147,
+      "grad_norm": 0.6799608539206533,
+      "learning_rate": 1.7834690296659852e-07,
+      "loss": 0.0588,
+      "step": 31850
+    },
+    {
+      "epoch": 2.7526027042204846,
+      "grad_norm": 0.6342124927694704,
+      "learning_rate": 1.7711446726294267e-07,
+      "loss": 0.0586,
+      "step": 31860
+    },
+    {
+      "epoch": 2.7534666724264545,
+      "grad_norm": 0.6311819082310265,
+      "learning_rate": 1.7588622782038378e-07,
+      "loss": 0.0562,
+      "step": 31870
+    },
+    {
+      "epoch": 2.754330640632425,
+      "grad_norm": 0.674920565799908,
+      "learning_rate": 1.7466218570757754e-07,
+      "loss": 0.0595,
+      "step": 31880
+    },
+    {
+      "epoch": 2.7551946088383947,
+      "grad_norm": 0.6635343214254527,
+      "learning_rate": 1.7344234198953146e-07,
+      "loss": 0.0574,
+      "step": 31890
+    },
+    {
+      "epoch": 2.7560585770443646,
+      "grad_norm": 0.6525627692019093,
+      "learning_rate": 1.7222669772759716e-07,
+      "loss": 0.0547,
+      "step": 31900
+    },
+    {
+      "epoch": 2.756922545250335,
+      "grad_norm": 0.6896387409372627,
+      "learning_rate": 1.7101525397947406e-07,
+      "loss": 0.0579,
+      "step": 31910
+    },
+    {
+      "epoch": 2.757786513456305,
+      "grad_norm": 0.6720002525440155,
+      "learning_rate": 1.6980801179920681e-07,
+      "loss": 0.0573,
+      "step": 31920
+    },
+    {
+      "epoch": 2.758650481662275,
+      "grad_norm": 0.6655959204666657,
+      "learning_rate": 1.6860497223718188e-07,
+      "loss": 0.056,
+      "step": 31930
+    },
+    {
+      "epoch": 2.759514449868245,
+      "grad_norm": 0.6723784296940033,
+      "learning_rate": 1.6740613634013413e-07,
+      "loss": 0.0564,
+      "step": 31940
+    },
+    {
+      "epoch": 2.760378418074215,
+      "grad_norm": 0.6816159314364688,
+      "learning_rate": 1.6621150515113538e-07,
+      "loss": 0.0548,
+      "step": 31950
+    },
+    {
+      "epoch": 2.761242386280185,
+      "grad_norm": 0.6597189697816395,
+      "learning_rate": 1.6502107970960246e-07,
+      "loss": 0.0581,
+      "step": 31960
+    },
+    {
+      "epoch": 2.762106354486155,
+      "grad_norm": 0.7127449095465125,
+      "learning_rate": 1.6383486105129243e-07,
+      "loss": 0.0564,
+      "step": 31970
+    },
+    {
+      "epoch": 2.7629703226921247,
+      "grad_norm": 0.6623458323343064,
+      "learning_rate": 1.6265285020830245e-07,
+      "loss": 0.0588,
+      "step": 31980
+    },
+    {
+      "epoch": 2.763834290898095,
+      "grad_norm": 0.6524127223074473,
+      "learning_rate": 1.6147504820906546e-07,
+      "loss": 0.0568,
+      "step": 31990
+    },
+    {
+      "epoch": 2.764698259104065,
+      "grad_norm": 0.6452682493379834,
+      "learning_rate": 1.603014560783578e-07,
+      "loss": 0.0574,
+      "step": 32000
+    },
+    {
+      "epoch": 2.765562227310035,
+      "grad_norm": 0.6838304105116794,
+      "learning_rate": 1.5913207483728765e-07,
+      "loss": 0.0565,
+      "step": 32010
+    },
+    {
+      "epoch": 2.766426195516005,
+      "grad_norm": 0.7006017396514543,
+      "learning_rate": 1.579669055033034e-07,
+      "loss": 0.0563,
+      "step": 32020
+    },
+    {
+      "epoch": 2.767290163721975,
+      "grad_norm": 0.63390922464745,
+      "learning_rate": 1.5680594909018575e-07,
+      "loss": 0.0574,
+      "step": 32030
+    },
+    {
+      "epoch": 2.768154131927945,
+      "grad_norm": 0.630174268226932,
+      "learning_rate": 1.5564920660805115e-07,
+      "loss": 0.0563,
+      "step": 32040
+    },
+    {
+      "epoch": 2.769018100133915,
+      "grad_norm": 0.6950850152884743,
+      "learning_rate": 1.544966790633512e-07,
+      "loss": 0.055,
+      "step": 32050
+    },
+    {
+      "epoch": 2.769882068339885,
+      "grad_norm": 0.686069735603795,
+      "learning_rate": 1.5334836745886762e-07,
+      "loss": 0.0555,
+      "step": 32060
+    },
+    {
+      "epoch": 2.770746036545855,
+      "grad_norm": 0.6924302699941641,
+      "learning_rate": 1.5220427279371507e-07,
+      "loss": 0.0579,
+      "step": 32070
+    },
+    {
+      "epoch": 2.771610004751825,
+      "grad_norm": 0.6763504845141571,
+      "learning_rate": 1.51064396063339e-07,
+      "loss": 0.059,
+      "step": 32080
+    },
+    {
+      "epoch": 2.772473972957795,
+      "grad_norm": 0.6855868771736366,
+      "learning_rate": 1.4992873825951548e-07,
+      "loss": 0.0581,
+      "step": 32090
+    },
+    {
+      "epoch": 2.7733379411637653,
+      "grad_norm": 0.6702095553630669,
+      "learning_rate": 1.4879730037034857e-07,
+      "loss": 0.058,
+      "step": 32100
+    },
+    {
+      "epoch": 2.774201909369735,
+      "grad_norm": 0.6798131090156784,
+      "learning_rate": 1.47670083380273e-07,
+      "loss": 0.0562,
+      "step": 32110
+    },
+    {
+      "epoch": 2.775065877575705,
+      "grad_norm": 0.6878985668508981,
+      "learning_rate": 1.465470882700476e-07,
+      "loss": 0.0587,
+      "step": 32120
+    },
+    {
+      "epoch": 2.7759298457816755,
+      "grad_norm": 0.6956681523253542,
+      "learning_rate": 1.4542831601676078e-07,
+      "loss": 0.0569,
+      "step": 32130
+    },
+    {
+      "epoch": 2.7767938139876454,
+      "grad_norm": 0.6749804800774928,
+      "learning_rate": 1.4431376759382664e-07,
+      "loss": 0.0552,
+      "step": 32140
+    },
+    {
+      "epoch": 2.7776577821936153,
+      "grad_norm": 0.6820298790434,
+      "learning_rate": 1.432034439709812e-07,
+      "loss": 0.0601,
+      "step": 32150
+    },
+    {
+      "epoch": 2.778521750399585,
+      "grad_norm": 0.7023866826285801,
+      "learning_rate": 1.4209734611428882e-07,
+      "loss": 0.0588,
+      "step": 32160
+    },
+    {
+      "epoch": 2.779385718605555,
+      "grad_norm": 0.7251283545475257,
+      "learning_rate": 1.4099547498613365e-07,
+      "loss": 0.0611,
+      "step": 32170
+    },
+    {
+      "epoch": 2.7802496868115254,
+      "grad_norm": 0.6794399022502436,
+      "learning_rate": 1.398978315452243e-07,
+      "loss": 0.0586,
+      "step": 32180
+    },
+    {
+      "epoch": 2.7811136550174953,
+      "grad_norm": 0.6829730156934579,
+      "learning_rate": 1.3880441674659017e-07,
+      "loss": 0.0578,
+      "step": 32190
+    },
+    {
+      "epoch": 2.781977623223465,
+      "grad_norm": 0.6780607571604497,
+      "learning_rate": 1.3771523154158194e-07,
+      "loss": 0.0582,
+      "step": 32200
+    },
+    {
+      "epoch": 2.7828415914294355,
+      "grad_norm": 0.6944067360719522,
+      "learning_rate": 1.366302768778699e-07,
+      "loss": 0.0563,
+      "step": 32210
+    },
+    {
+      "epoch": 2.7837055596354054,
+      "grad_norm": 0.6926254817665085,
+      "learning_rate": 1.3554955369944334e-07,
+      "loss": 0.0575,
+      "step": 32220
+    },
+    {
+      "epoch": 2.7845695278413753,
+      "grad_norm": 0.6412247121885293,
+      "learning_rate": 1.3447306294661012e-07,
+      "loss": 0.0589,
+      "step": 32230
+    },
+    {
+      "epoch": 2.7854334960473457,
+      "grad_norm": 0.6625032725187675,
+      "learning_rate": 1.334008055559949e-07,
+      "loss": 0.057,
+      "step": 32240
+    },
+    {
+      "epoch": 2.7862974642533156,
+      "grad_norm": 0.6890514579031258,
+      "learning_rate": 1.3233278246054081e-07,
+      "loss": 0.0575,
+      "step": 32250
+    },
+    {
+      "epoch": 2.7871614324592855,
+      "grad_norm": 0.6749984423954664,
+      "learning_rate": 1.3126899458950403e-07,
+      "loss": 0.0567,
+      "step": 32260
+    },
+    {
+      "epoch": 2.7880254006652554,
+      "grad_norm": 0.6058312889602515,
+      "learning_rate": 1.302094428684586e-07,
+      "loss": 0.0576,
+      "step": 32270
+    },
+    {
+      "epoch": 2.7888893688712253,
+      "grad_norm": 0.6640679135768991,
+      "learning_rate": 1.2915412821929096e-07,
+      "loss": 0.0549,
+      "step": 32280
+    },
+    {
+      "epoch": 2.7897533370771956,
+      "grad_norm": 0.6667923213308632,
+      "learning_rate": 1.281030515602022e-07,
+      "loss": 0.0542,
+      "step": 32290
+    },
+    {
+      "epoch": 2.7906173052831655,
+      "grad_norm": 0.6763644991293586,
+      "learning_rate": 1.2705621380570465e-07,
+      "loss": 0.0564,
+      "step": 32300
+    },
+    {
+      "epoch": 2.7914812734891354,
+      "grad_norm": 0.671158560568952,
+      "learning_rate": 1.2601361586662364e-07,
+      "loss": 0.0585,
+      "step": 32310
+    },
+    {
+      "epoch": 2.792345241695106,
+      "grad_norm": 0.6522339042934434,
+      "learning_rate": 1.249752586500952e-07,
+      "loss": 0.0581,
+      "step": 32320
+    },
+    {
+      "epoch": 2.7932092099010757,
+      "grad_norm": 0.6979271307967363,
+      "learning_rate": 1.239411430595655e-07,
+      "loss": 0.0586,
+      "step": 32330
+    },
+    {
+      "epoch": 2.7940731781070456,
+      "grad_norm": 0.6799894206330298,
+      "learning_rate": 1.229112699947904e-07,
+      "loss": 0.0563,
+      "step": 32340
+    },
+    {
+      "epoch": 2.794937146313016,
+      "grad_norm": 0.677619561225624,
+      "learning_rate": 1.2188564035183425e-07,
+      "loss": 0.0556,
+      "step": 32350
+    },
+    {
+      "epoch": 2.795801114518986,
+      "grad_norm": 0.6297337136303034,
+      "learning_rate": 1.2086425502306986e-07,
+      "loss": 0.0553,
+      "step": 32360
+    },
+    {
+      "epoch": 2.7966650827249557,
+      "grad_norm": 0.6450798875927606,
+      "learning_rate": 1.1984711489717583e-07,
+      "loss": 0.057,
+      "step": 32370
+    },
+    {
+      "epoch": 2.7975290509309256,
+      "grad_norm": 0.6710396032708422,
+      "learning_rate": 1.188342208591392e-07,
+      "loss": 0.0572,
+      "step": 32380
+    },
+    {
+      "epoch": 2.7983930191368955,
+      "grad_norm": 0.6870288667869697,
+      "learning_rate": 1.1782557379025005e-07,
+      "loss": 0.0566,
+      "step": 32390
+    },
+    {
+      "epoch": 2.799256987342866,
+      "grad_norm": 0.6549415421772309,
+      "learning_rate": 1.1682117456810583e-07,
+      "loss": 0.0591,
+      "step": 32400
+    },
+    {
+      "epoch": 2.8001209555488358,
+      "grad_norm": 0.6412098336825817,
+      "learning_rate": 1.1582102406660634e-07,
+      "loss": 0.0572,
+      "step": 32410
+    },
+    {
+      "epoch": 2.8009849237548057,
+      "grad_norm": 0.6752618412945657,
+      "learning_rate": 1.1482512315595607e-07,
+      "loss": 0.058,
+      "step": 32420
+    },
+    {
+      "epoch": 2.801848891960776,
+      "grad_norm": 0.6359954770406422,
+      "learning_rate": 1.1383347270265965e-07,
+      "loss": 0.0553,
+      "step": 32430
+    },
+    {
+      "epoch": 2.802712860166746,
+      "grad_norm": 0.6876557978296493,
+      "learning_rate": 1.1284607356952638e-07,
+      "loss": 0.0575,
+      "step": 32440
+    },
+    {
+      "epoch": 2.803576828372716,
+      "grad_norm": 0.7021487811787359,
+      "learning_rate": 1.1186292661566511e-07,
+      "loss": 0.0566,
+      "step": 32450
+    },
+    {
+      "epoch": 2.804440796578686,
+      "grad_norm": 0.6750019608519389,
+      "learning_rate": 1.108840326964844e-07,
+      "loss": 0.0576,
+      "step": 32460
+    },
+    {
+      "epoch": 2.805304764784656,
+      "grad_norm": 0.6277430752290388,
+      "learning_rate": 1.0990939266369404e-07,
+      "loss": 0.0587,
+      "step": 32470
+    },
+    {
+      "epoch": 2.806168732990626,
+      "grad_norm": 0.6530520547027956,
+      "learning_rate": 1.0893900736530127e-07,
+      "loss": 0.0573,
+      "step": 32480
+    },
+    {
+      "epoch": 2.807032701196596,
+      "grad_norm": 0.6696408985828085,
+      "learning_rate": 1.0797287764561238e-07,
+      "loss": 0.0566,
+      "step": 32490
+    },
+    {
+      "epoch": 2.8078966694025658,
+      "grad_norm": 0.6628118939310904,
+      "learning_rate": 1.0701100434522882e-07,
+      "loss": 0.056,
+      "step": 32500
+    },
+    {
+      "epoch": 2.808760637608536,
+      "grad_norm": 0.6358518359835514,
+      "learning_rate": 1.0605338830105283e-07,
+      "loss": 0.0567,
+      "step": 32510
+    },
+    {
+      "epoch": 2.809624605814506,
+      "grad_norm": 0.6552015344909092,
+      "learning_rate": 1.0510003034627848e-07,
+      "loss": 0.0558,
+      "step": 32520
+    },
+    {
+      "epoch": 2.810488574020476,
+      "grad_norm": 0.6786865741009189,
+      "learning_rate": 1.0415093131039666e-07,
+      "loss": 0.0542,
+      "step": 32530
+    },
+    {
+      "epoch": 2.8113525422264463,
+      "grad_norm": 0.6708262449590945,
+      "learning_rate": 1.0320609201919241e-07,
+      "loss": 0.0546,
+      "step": 32540
+    },
+    {
+      "epoch": 2.812216510432416,
+      "grad_norm": 0.6871889215796761,
+      "learning_rate": 1.0226551329474477e-07,
+      "loss": 0.0591,
+      "step": 32550
+    },
+    {
+      "epoch": 2.813080478638386,
+      "grad_norm": 0.6796148770244584,
+      "learning_rate": 1.0132919595542634e-07,
+      "loss": 0.0572,
+      "step": 32560
+    },
+    {
+      "epoch": 2.813944446844356,
+      "grad_norm": 0.6448854547531903,
+      "learning_rate": 1.0039714081589991e-07,
+      "loss": 0.057,
+      "step": 32570
+    },
+    {
+      "epoch": 2.8148084150503263,
+      "grad_norm": 0.7214571409269863,
+      "learning_rate": 9.946934868712176e-08,
+      "loss": 0.0566,
+      "step": 32580
+    },
+    {
+      "epoch": 2.815672383256296,
+      "grad_norm": 0.6864676012136871,
+      "learning_rate": 9.854582037633953e-08,
+      "loss": 0.0559,
+      "step": 32590
+    },
+    {
+      "epoch": 2.816536351462266,
+      "grad_norm": 0.6875452157455118,
+      "learning_rate": 9.762655668708876e-08,
+      "loss": 0.0574,
+      "step": 32600
+    },
+    {
+      "epoch": 2.817400319668236,
+      "grad_norm": 0.6471250629240326,
+      "learning_rate": 9.671155841919577e-08,
+      "loss": 0.0574,
+      "step": 32610
+    },
+    {
+      "epoch": 2.8182642878742064,
+      "grad_norm": 0.6830004863111572,
+      "learning_rate": 9.580082636877652e-08,
+      "loss": 0.0569,
+      "step": 32620
+    },
+    {
+      "epoch": 2.8191282560801763,
+      "grad_norm": 0.6419472467128323,
+      "learning_rate": 9.489436132823326e-08,
+      "loss": 0.057,
+      "step": 32630
+    },
+    {
+      "epoch": 2.819992224286146,
+      "grad_norm": 0.657306933977155,
+      "learning_rate": 9.399216408625678e-08,
+      "loss": 0.0585,
+      "step": 32640
+    },
+    {
+      "epoch": 2.8208561924921165,
+      "grad_norm": 0.6432005216885947,
+      "learning_rate": 9.309423542782414e-08,
+      "loss": 0.0573,
+      "step": 32650
+    },
+    {
+      "epoch": 2.8217201606980864,
+      "grad_norm": 0.7002865300654664,
+      "learning_rate": 9.22005761341982e-08,
+      "loss": 0.0588,
+      "step": 32660
+    },
+    {
+      "epoch": 2.8225841289040563,
+      "grad_norm": 0.6612868667391867,
+      "learning_rate": 9.131118698292863e-08,
+      "loss": 0.0566,
+      "step": 32670
+    },
+    {
+      "epoch": 2.823448097110026,
+      "grad_norm": 0.6761202883755489,
+      "learning_rate": 9.042606874784809e-08,
+      "loss": 0.059,
+      "step": 32680
+    },
+    {
+      "epoch": 2.824312065315996,
+      "grad_norm": 0.6862611791080319,
+      "learning_rate": 8.954522219907392e-08,
+      "loss": 0.0547,
+      "step": 32690
+    },
+    {
+      "epoch": 2.8251760335219664,
+      "grad_norm": 0.6254683022957856,
+      "learning_rate": 8.866864810300579e-08,
+      "loss": 0.0564,
+      "step": 32700
+    },
+    {
+      "epoch": 2.8260400017279363,
+      "grad_norm": 0.6541988728358744,
+      "learning_rate": 8.77963472223281e-08,
+      "loss": 0.0576,
+      "step": 32710
+    },
+    {
+      "epoch": 2.8269039699339062,
+      "grad_norm": 0.7137230691053326,
+      "learning_rate": 8.692832031600539e-08,
+      "loss": 0.0568,
+      "step": 32720
+    },
+    {
+      "epoch": 2.8277679381398766,
+      "grad_norm": 0.6744846313172326,
+      "learning_rate": 8.60645681392841e-08,
+      "loss": 0.0563,
+      "step": 32730
+    },
+    {
+      "epoch": 2.8286319063458465,
+      "grad_norm": 0.7290262081800398,
+      "learning_rate": 8.520509144369194e-08,
+      "loss": 0.0582,
+      "step": 32740
+    },
+    {
+      "epoch": 2.8294958745518164,
+      "grad_norm": 0.6928234087420588,
+      "learning_rate": 8.434989097703572e-08,
+      "loss": 0.0565,
+      "step": 32750
+    },
+    {
+      "epoch": 2.8303598427577867,
+      "grad_norm": 0.6537822863085431,
+      "learning_rate": 8.349896748340191e-08,
+      "loss": 0.0576,
+      "step": 32760
+    },
+    {
+      "epoch": 2.8312238109637566,
+      "grad_norm": 0.6664152071722941,
+      "learning_rate": 8.265232170315607e-08,
+      "loss": 0.0584,
+      "step": 32770
+    },
+    {
+      "epoch": 2.8320877791697265,
+      "grad_norm": 0.6543148772010279,
+      "learning_rate": 8.180995437294226e-08,
+      "loss": 0.0564,
+      "step": 32780
+    },
+    {
+      "epoch": 2.8329517473756964,
+      "grad_norm": 0.6469114237292819,
+      "learning_rate": 8.097186622568032e-08,
+      "loss": 0.0566,
+      "step": 32790
+    },
+    {
+      "epoch": 2.8338157155816663,
+      "grad_norm": 0.6486680998913869,
+      "learning_rate": 8.013805799056806e-08,
+      "loss": 0.0559,
+      "step": 32800
+    },
+    {
+      "epoch": 2.8346796837876367,
+      "grad_norm": 0.6528582438625301,
+      "learning_rate": 7.930853039307962e-08,
+      "loss": 0.0569,
+      "step": 32810
+    },
+    {
+      "epoch": 2.8355436519936066,
+      "grad_norm": 0.6647853143387031,
+      "learning_rate": 7.848328415496375e-08,
+      "loss": 0.0584,
+      "step": 32820
+    },
+    {
+      "epoch": 2.8364076201995765,
+      "grad_norm": 0.6857635854830215,
+      "learning_rate": 7.766231999424556e-08,
+      "loss": 0.0569,
+      "step": 32830
+    },
+    {
+      "epoch": 2.837271588405547,
+      "grad_norm": 0.6888594940329839,
+      "learning_rate": 7.684563862522254e-08,
+      "loss": 0.0581,
+      "step": 32840
+    },
+    {
+      "epoch": 2.8381355566115167,
+      "grad_norm": 0.6483498063174805,
+      "learning_rate": 7.603324075846741e-08,
+      "loss": 0.0548,
+      "step": 32850
+    },
+    {
+      "epoch": 2.8389995248174866,
+      "grad_norm": 0.6961287471485743,
+      "learning_rate": 7.522512710082531e-08,
+      "loss": 0.0572,
+      "step": 32860
+    },
+    {
+      "epoch": 2.839863493023457,
+      "grad_norm": 0.6713877134589054,
+      "learning_rate": 7.442129835541433e-08,
+      "loss": 0.0558,
+      "step": 32870
+    },
+    {
+      "epoch": 2.840727461229427,
+      "grad_norm": 0.6447047687390621,
+      "learning_rate": 7.362175522162284e-08,
+      "loss": 0.0558,
+      "step": 32880
+    },
+    {
+      "epoch": 2.8415914294353968,
+      "grad_norm": 0.6950714605507896,
+      "learning_rate": 7.282649839511269e-08,
+      "loss": 0.0562,
+      "step": 32890
+    },
+    {
+      "epoch": 2.8424553976413667,
+      "grad_norm": 0.662844122297223,
+      "learning_rate": 7.203552856781482e-08,
+      "loss": 0.057,
+      "step": 32900
+    },
+    {
+      "epoch": 2.8433193658473366,
+      "grad_norm": 0.6764003819400013,
+      "learning_rate": 7.124884642793039e-08,
+      "loss": 0.0576,
+      "step": 32910
+    },
+    {
+      "epoch": 2.844183334053307,
+      "grad_norm": 0.6665223091511387,
+      "learning_rate": 7.046645265992968e-08,
+      "loss": 0.0546,
+      "step": 32920
+    },
+    {
+      "epoch": 2.845047302259277,
+      "grad_norm": 0.7105587069690044,
+      "learning_rate": 6.968834794455203e-08,
+      "loss": 0.0583,
+      "step": 32930
+    },
+    {
+      "epoch": 2.8459112704652467,
+      "grad_norm": 0.6579588776400764,
+      "learning_rate": 6.891453295880591e-08,
+      "loss": 0.0572,
+      "step": 32940
+    },
+    {
+      "epoch": 2.846775238671217,
+      "grad_norm": 0.6500028386919731,
+      "learning_rate": 6.81450083759666e-08,
+      "loss": 0.0572,
+      "step": 32950
+    },
+    {
+      "epoch": 2.847639206877187,
+      "grad_norm": 0.7149086521713711,
+      "learning_rate": 6.737977486557578e-08,
+      "loss": 0.0543,
+      "step": 32960
+    },
+    {
+      "epoch": 2.848503175083157,
+      "grad_norm": 0.6856207174938993,
+      "learning_rate": 6.661883309344253e-08,
+      "loss": 0.0568,
+      "step": 32970
+    },
+    {
+      "epoch": 2.849367143289127,
+      "grad_norm": 0.6891817002441347,
+      "learning_rate": 6.586218372164166e-08,
+      "loss": 0.0571,
+      "step": 32980
+    },
+    {
+      "epoch": 2.850231111495097,
+      "grad_norm": 0.6777405340120837,
+      "learning_rate": 6.510982740851269e-08,
+      "loss": 0.0564,
+      "step": 32990
+    },
+    {
+      "epoch": 2.851095079701067,
+      "grad_norm": 0.6673915333144583,
+      "learning_rate": 6.436176480866141e-08,
+      "loss": 0.0569,
+      "step": 33000
+    },
+    {
+      "epoch": 2.851959047907037,
+      "grad_norm": 0.6734418146878922,
+      "learning_rate": 6.361799657295552e-08,
+      "loss": 0.0586,
+      "step": 33010
+    },
+    {
+      "epoch": 2.852823016113007,
+      "grad_norm": 0.6904229506397264,
+      "learning_rate": 6.287852334852795e-08,
+      "loss": 0.06,
+      "step": 33020
+    },
+    {
+      "epoch": 2.853686984318977,
+      "grad_norm": 0.6830159237337381,
+      "learning_rate": 6.214334577877401e-08,
+      "loss": 0.0554,
+      "step": 33030
+    },
+    {
+      "epoch": 2.854550952524947,
+      "grad_norm": 0.7220761887506169,
+      "learning_rate": 6.141246450335148e-08,
+      "loss": 0.0586,
+      "step": 33040
+    },
+    {
+      "epoch": 2.855414920730917,
+      "grad_norm": 0.6791045442446241,
+      "learning_rate": 6.068588015818055e-08,
+      "loss": 0.0579,
+      "step": 33050
+    },
+    {
+      "epoch": 2.8562788889368873,
+      "grad_norm": 0.6860422169888754,
+      "learning_rate": 5.996359337544277e-08,
+      "loss": 0.0554,
+      "step": 33060
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.6926889419675112,
+      "learning_rate": 5.924560478357988e-08,
+      "loss": 0.0572,
+      "step": 33070
+    },
+    {
+      "epoch": 2.858006825348827,
+      "grad_norm": 0.6572432697985694,
+      "learning_rate": 5.853191500729327e-08,
+      "loss": 0.0589,
+      "step": 33080
+    },
+    {
+      "epoch": 2.8588707935547975,
+      "grad_norm": 0.6959692528825726,
+      "learning_rate": 5.782252466754623e-08,
+      "loss": 0.0577,
+      "step": 33090
+    },
+    {
+      "epoch": 2.8597347617607674,
+      "grad_norm": 0.6563165677197141,
+      "learning_rate": 5.711743438155892e-08,
+      "loss": 0.0583,
+      "step": 33100
+    },
+    {
+      "epoch": 2.8605987299667373,
+      "grad_norm": 0.6539622559768795,
+      "learning_rate": 5.6416644762812277e-08,
+      "loss": 0.0578,
+      "step": 33110
+    },
+    {
+      "epoch": 2.861462698172707,
+      "grad_norm": 0.6369086566201649,
+      "learning_rate": 5.5720156421043556e-08,
+      "loss": 0.0564,
+      "step": 33120
+    },
+    {
+      "epoch": 2.862326666378677,
+      "grad_norm": 0.6420979588342911,
+      "learning_rate": 5.502796996224746e-08,
+      "loss": 0.0563,
+      "step": 33130
+    },
+    {
+      "epoch": 2.8631906345846474,
+      "grad_norm": 0.6695727199391603,
+      "learning_rate": 5.434008598867835e-08,
+      "loss": 0.0554,
+      "step": 33140
+    },
+    {
+      "epoch": 2.8640546027906173,
+      "grad_norm": 0.6600310210773601,
+      "learning_rate": 5.365650509884357e-08,
+      "loss": 0.0586,
+      "step": 33150
+    },
+    {
+      "epoch": 2.864918570996587,
+      "grad_norm": 0.6484929852915966,
+      "learning_rate": 5.297722788750958e-08,
+      "loss": 0.0556,
+      "step": 33160
+    },
+    {
+      "epoch": 2.8657825392025575,
+      "grad_norm": 0.7134033307243018,
+      "learning_rate": 5.230225494569585e-08,
+      "loss": 0.0543,
+      "step": 33170
+    },
+    {
+      "epoch": 2.8666465074085274,
+      "grad_norm": 0.6946163701427388,
+      "learning_rate": 5.1631586860678703e-08,
+      "loss": 0.0557,
+      "step": 33180
+    },
+    {
+      "epoch": 2.8675104756144973,
+      "grad_norm": 0.6743236939216565,
+      "learning_rate": 5.0965224215987484e-08,
+      "loss": 0.0557,
+      "step": 33190
+    },
+    {
+      "epoch": 2.8683744438204672,
+      "grad_norm": 0.6385782451925563,
+      "learning_rate": 5.030316759140674e-08,
+      "loss": 0.0543,
+      "step": 33200
+    },
+    {
+      "epoch": 2.8692384120264376,
+      "grad_norm": 0.6894700813928674,
+      "learning_rate": 4.9645417562973476e-08,
+      "loss": 0.056,
+      "step": 33210
+    },
+    {
+      "epoch": 2.8701023802324075,
+      "grad_norm": 0.6675629300734072,
+      "learning_rate": 4.899197470297823e-08,
+      "loss": 0.056,
+      "step": 33220
+    },
+    {
+      "epoch": 2.8709663484383774,
+      "grad_norm": 0.668713281409164,
+      "learning_rate": 4.834283957996344e-08,
+      "loss": 0.0547,
+      "step": 33230
+    },
+    {
+      "epoch": 2.8718303166443473,
+      "grad_norm": 0.6665885336321217,
+      "learning_rate": 4.7698012758724547e-08,
+      "loss": 0.0559,
+      "step": 33240
+    },
+    {
+      "epoch": 2.8726942848503176,
+      "grad_norm": 0.6821105530534564,
+      "learning_rate": 4.705749480030719e-08,
+      "loss": 0.0572,
+      "step": 33250
+    },
+    {
+      "epoch": 2.8735582530562875,
+      "grad_norm": 0.6708658577242343,
+      "learning_rate": 4.6421286262008924e-08,
+      "loss": 0.0572,
+      "step": 33260
+    },
+    {
+      "epoch": 2.8744222212622574,
+      "grad_norm": 0.647767125351178,
+      "learning_rate": 4.578938769737751e-08,
+      "loss": 0.0584,
+      "step": 33270
+    },
+    {
+      "epoch": 2.875286189468228,
+      "grad_norm": 0.6780529409748565,
+      "learning_rate": 4.5161799656209814e-08,
+      "loss": 0.0559,
+      "step": 33280
+    },
+    {
+      "epoch": 2.8761501576741977,
+      "grad_norm": 0.649348742633913,
+      "learning_rate": 4.453852268455461e-08,
+      "loss": 0.0593,
+      "step": 33290
+    },
+    {
+      "epoch": 2.8770141258801676,
+      "grad_norm": 0.6676144330672698,
+      "learning_rate": 4.391955732470643e-08,
+      "loss": 0.0566,
+      "step": 33300
+    },
+    {
+      "epoch": 2.8778780940861375,
+      "grad_norm": 0.686248518816948,
+      "learning_rate": 4.330490411521116e-08,
+      "loss": 0.0577,
+      "step": 33310
+    },
+    {
+      "epoch": 2.8787420622921074,
+      "grad_norm": 0.6842785616361624,
+      "learning_rate": 4.2694563590861546e-08,
+      "loss": 0.0562,
+      "step": 33320
+    },
+    {
+      "epoch": 2.8796060304980777,
+      "grad_norm": 0.6437126649798809,
+      "learning_rate": 4.2088536282698376e-08,
+      "loss": 0.057,
+      "step": 33330
+    },
+    {
+      "epoch": 2.8804699987040476,
+      "grad_norm": 0.6942606890668203,
+      "learning_rate": 4.14868227180093e-08,
+      "loss": 0.0588,
+      "step": 33340
+    },
+    {
+      "epoch": 2.8813339669100175,
+      "grad_norm": 0.6577488295499587,
+      "learning_rate": 4.0889423420328866e-08,
+      "loss": 0.0559,
+      "step": 33350
+    },
+    {
+      "epoch": 2.882197935115988,
+      "grad_norm": 0.6658072223920517,
+      "learning_rate": 4.029633890943796e-08,
+      "loss": 0.0557,
+      "step": 33360
+    },
+    {
+      "epoch": 2.8830619033219578,
+      "grad_norm": 0.6897837859797347,
+      "learning_rate": 3.97075697013638e-08,
+      "loss": 0.0577,
+      "step": 33370
+    },
+    {
+      "epoch": 2.8839258715279277,
+      "grad_norm": 0.6755459921707765,
+      "learning_rate": 3.912311630837717e-08,
+      "loss": 0.0582,
+      "step": 33380
+    },
+    {
+      "epoch": 2.884789839733898,
+      "grad_norm": 0.688782499527184,
+      "learning_rate": 3.8542979238995746e-08,
+      "loss": 0.0589,
+      "step": 33390
+    },
+    {
+      "epoch": 2.885653807939868,
+      "grad_norm": 0.6655676381591664,
+      "learning_rate": 3.7967158997981315e-08,
+      "loss": 0.0576,
+      "step": 33400
+    },
+    {
+      "epoch": 2.886517776145838,
+      "grad_norm": 0.6986677921599695,
+      "learning_rate": 3.739565608633866e-08,
+      "loss": 0.0558,
+      "step": 33410
+    },
+    {
+      "epoch": 2.8873817443518077,
+      "grad_norm": 0.663380414330348,
+      "learning_rate": 3.6828471001317254e-08,
+      "loss": 0.0577,
+      "step": 33420
+    },
+    {
+      "epoch": 2.8882457125577776,
+      "grad_norm": 0.6936533010310243,
+      "learning_rate": 3.626560423640957e-08,
+      "loss": 0.0589,
+      "step": 33430
+    },
+    {
+      "epoch": 2.889109680763748,
+      "grad_norm": 0.6477438704961663,
+      "learning_rate": 3.570705628134996e-08,
+      "loss": 0.0577,
+      "step": 33440
+    },
+    {
+      "epoch": 2.889973648969718,
+      "grad_norm": 0.6871551024773646,
+      "learning_rate": 3.5152827622116356e-08,
+      "loss": 0.0571,
+      "step": 33450
+    },
+    {
+      "epoch": 2.8908376171756878,
+      "grad_norm": 0.6764162456797573,
+      "learning_rate": 3.460291874092747e-08,
+      "loss": 0.0584,
+      "step": 33460
+    },
+    {
+      "epoch": 2.891701585381658,
+      "grad_norm": 0.6676655057360613,
+      "learning_rate": 3.405733011624446e-08,
+      "loss": 0.0578,
+      "step": 33470
+    },
+    {
+      "epoch": 2.892565553587628,
+      "grad_norm": 0.662029257328337,
+      "learning_rate": 3.3516062222768706e-08,
+      "loss": 0.0547,
+      "step": 33480
+    },
+    {
+      "epoch": 2.893429521793598,
+      "grad_norm": 0.7215604812631449,
+      "learning_rate": 3.297911553144295e-08,
+      "loss": 0.0572,
+      "step": 33490
+    },
+    {
+      "epoch": 2.8942934899995683,
+      "grad_norm": 0.6623796053756021,
+      "learning_rate": 3.2446490509450145e-08,
+      "loss": 0.0557,
+      "step": 33500
+    },
+    {
+      "epoch": 2.895157458205538,
+      "grad_norm": 0.6771863576032046,
+      "learning_rate": 3.1918187620211816e-08,
+      "loss": 0.0571,
+      "step": 33510
+    },
+    {
+      "epoch": 2.896021426411508,
+      "grad_norm": 0.6760263824674185,
+      "learning_rate": 3.1394207323390826e-08,
+      "loss": 0.0572,
+      "step": 33520
+    },
+    {
+      "epoch": 2.896885394617478,
+      "grad_norm": 0.664126918625583,
+      "learning_rate": 3.087455007488804e-08,
+      "loss": 0.0575,
+      "step": 33530
+    },
+    {
+      "epoch": 2.897749362823448,
+      "grad_norm": 0.695653549487125,
+      "learning_rate": 3.035921632684236e-08,
+      "loss": 0.0578,
+      "step": 33540
+    },
+    {
+      "epoch": 2.898613331029418,
+      "grad_norm": 0.6849873662405368,
+      "learning_rate": 2.984820652763232e-08,
+      "loss": 0.0565,
+      "step": 33550
+    },
+    {
+      "epoch": 2.899477299235388,
+      "grad_norm": 0.6401444792075645,
+      "learning_rate": 2.9341521121873385e-08,
+      "loss": 0.0587,
+      "step": 33560
+    },
+    {
+      "epoch": 2.900341267441358,
+      "grad_norm": 0.6562321741503842,
+      "learning_rate": 2.8839160550418466e-08,
+      "loss": 0.0578,
+      "step": 33570
+    },
+    {
+      "epoch": 2.9012052356473284,
+      "grad_norm": 0.7165707023965725,
+      "learning_rate": 2.8341125250357928e-08,
+      "loss": 0.056,
+      "step": 33580
+    },
+    {
+      "epoch": 2.9020692038532983,
+      "grad_norm": 0.6535414446662702,
+      "learning_rate": 2.7847415655019026e-08,
+      "loss": 0.0562,
+      "step": 33590
+    },
+    {
+      "epoch": 2.902933172059268,
+      "grad_norm": 0.6590199643709393,
+      "learning_rate": 2.7358032193964824e-08,
+      "loss": 0.0568,
+      "step": 33600
+    },
+    {
+      "epoch": 2.9037971402652385,
+      "grad_norm": 0.7038395574717661,
+      "learning_rate": 2.6872975292994153e-08,
+      "loss": 0.0574,
+      "step": 33610
+    },
+    {
+      "epoch": 2.9046611084712084,
+      "grad_norm": 0.6603831559878955,
+      "learning_rate": 2.63922453741422e-08,
+      "loss": 0.0537,
+      "step": 33620
+    },
+    {
+      "epoch": 2.9055250766771783,
+      "grad_norm": 0.6809259145907269,
+      "learning_rate": 2.5915842855678276e-08,
+      "loss": 0.0561,
+      "step": 33630
+    },
+    {
+      "epoch": 2.906389044883148,
+      "grad_norm": 0.6675281253432127,
+      "learning_rate": 2.5443768152108583e-08,
+      "loss": 0.0545,
+      "step": 33640
+    },
+    {
+      "epoch": 2.907253013089118,
+      "grad_norm": 0.6681893173881671,
+      "learning_rate": 2.4976021674171236e-08,
+      "loss": 0.0547,
+      "step": 33650
+    },
+    {
+      "epoch": 2.9081169812950884,
+      "grad_norm": 0.6592809611380828,
+      "learning_rate": 2.4512603828839575e-08,
+      "loss": 0.0561,
+      "step": 33660
+    },
+    {
+      "epoch": 2.9089809495010583,
+      "grad_norm": 0.6490028684571275,
+      "learning_rate": 2.4053515019322184e-08,
+      "loss": 0.0565,
+      "step": 33670
+    },
+    {
+      "epoch": 2.9098449177070282,
+      "grad_norm": 0.656601858138169,
+      "learning_rate": 2.3598755645057868e-08,
+      "loss": 0.0571,
+      "step": 33680
+    },
+    {
+      "epoch": 2.9107088859129986,
+      "grad_norm": 0.6941640271804507,
+      "learning_rate": 2.3148326101722352e-08,
+      "loss": 0.0574,
+      "step": 33690
+    },
+    {
+      "epoch": 2.9115728541189685,
+      "grad_norm": 0.6603126085653375,
+      "learning_rate": 2.2702226781219916e-08,
+      "loss": 0.0581,
+      "step": 33700
+    },
+    {
+      "epoch": 2.9124368223249384,
+      "grad_norm": 0.668590890656347,
+      "learning_rate": 2.2260458071691194e-08,
+      "loss": 0.0568,
+      "step": 33710
+    },
+    {
+      "epoch": 2.9133007905309087,
+      "grad_norm": 0.6493466506588934,
+      "learning_rate": 2.1823020357506497e-08,
+      "loss": 0.0569,
+      "step": 33720
+    },
+    {
+      "epoch": 2.9141647587368786,
+      "grad_norm": 0.6539909184550178,
+      "learning_rate": 2.1389914019268045e-08,
+      "loss": 0.0574,
+      "step": 33730
+    },
+    {
+      "epoch": 2.9150287269428485,
+      "grad_norm": 0.6527098662158097,
+      "learning_rate": 2.0961139433810508e-08,
+      "loss": 0.0556,
+      "step": 33740
+    },
+    {
+      "epoch": 2.9158926951488184,
+      "grad_norm": 0.7010404064747029,
+      "learning_rate": 2.0536696974198245e-08,
+      "loss": 0.0556,
+      "step": 33750
+    },
+    {
+      "epoch": 2.9167566633547883,
+      "grad_norm": 0.6286359295433286,
+      "learning_rate": 2.0116587009727516e-08,
+      "loss": 0.0565,
+      "step": 33760
+    },
+    {
+      "epoch": 2.9176206315607587,
+      "grad_norm": 0.6462394665804463,
+      "learning_rate": 1.9700809905924267e-08,
+      "loss": 0.0551,
+      "step": 33770
+    },
+    {
+      "epoch": 2.9184845997667286,
+      "grad_norm": 0.6467936095668956,
+      "learning_rate": 1.928936602454523e-08,
+      "loss": 0.0566,
+      "step": 33780
+    },
+    {
+      "epoch": 2.9193485679726985,
+      "grad_norm": 0.6431999964981738,
+      "learning_rate": 1.888225572357627e-08,
+      "loss": 0.0527,
+      "step": 33790
+    },
+    {
+      "epoch": 2.920212536178669,
+      "grad_norm": 0.677044077943376,
+      "learning_rate": 1.8479479357232933e-08,
+      "loss": 0.0584,
+      "step": 33800
+    },
+    {
+      "epoch": 2.9210765043846387,
+      "grad_norm": 0.6490379738984648,
+      "learning_rate": 1.808103727595989e-08,
+      "loss": 0.0556,
+      "step": 33810
+    },
+    {
+      "epoch": 2.9219404725906086,
+      "grad_norm": 0.6365858333329392,
+      "learning_rate": 1.7686929826430944e-08,
+      "loss": 0.0579,
+      "step": 33820
+    },
+    {
+      "epoch": 2.9228044407965785,
+      "grad_norm": 0.6736372234260222,
+      "learning_rate": 1.7297157351547912e-08,
+      "loss": 0.055,
+      "step": 33830
+    },
+    {
+      "epoch": 2.9236684090025484,
+      "grad_norm": 0.6796407556520421,
+      "learning_rate": 1.6911720190441183e-08,
+      "loss": 0.0543,
+      "step": 33840
+    },
+    {
+      "epoch": 2.9245323772085188,
+      "grad_norm": 0.6906670089068052,
+      "learning_rate": 1.6530618678469723e-08,
+      "loss": 0.0585,
+      "step": 33850
+    },
+    {
+      "epoch": 2.9253963454144887,
+      "grad_norm": 0.6885910897171846,
+      "learning_rate": 1.6153853147218846e-08,
+      "loss": 0.058,
+      "step": 33860
+    },
+    {
+      "epoch": 2.9262603136204586,
+      "grad_norm": 0.6780701321572589,
+      "learning_rate": 1.5781423924502438e-08,
+      "loss": 0.0557,
+      "step": 33870
+    },
+    {
+      "epoch": 2.927124281826429,
+      "grad_norm": 0.6588324510313825,
+      "learning_rate": 1.541333133436018e-08,
+      "loss": 0.058,
+      "step": 33880
+    },
+    {
+      "epoch": 2.927988250032399,
+      "grad_norm": 0.6491208370689091,
+      "learning_rate": 1.5049575697060337e-08,
+      "loss": 0.0569,
+      "step": 33890
+    },
+    {
+      "epoch": 2.9288522182383687,
+      "grad_norm": 0.6517897930275939,
+      "learning_rate": 1.4690157329096399e-08,
+      "loss": 0.0555,
+      "step": 33900
+    },
+    {
+      "epoch": 2.929716186444339,
+      "grad_norm": 0.7042225264467061,
+      "learning_rate": 1.4335076543188776e-08,
+      "loss": 0.0555,
+      "step": 33910
+    },
+    {
+      "epoch": 2.930580154650309,
+      "grad_norm": 0.6580352339265412,
+      "learning_rate": 1.398433364828311e-08,
+      "loss": 0.0579,
+      "step": 33920
+    },
+    {
+      "epoch": 2.931444122856279,
+      "grad_norm": 0.6628295724247529,
+      "learning_rate": 1.3637928949551404e-08,
+      "loss": 0.0563,
+      "step": 33930
+    },
+    {
+      "epoch": 2.9323080910622488,
+      "grad_norm": 0.6664636967953322,
+      "learning_rate": 1.3295862748390898e-08,
+      "loss": 0.0564,
+      "step": 33940
+    },
+    {
+      "epoch": 2.9331720592682187,
+      "grad_norm": 0.684321235610932,
+      "learning_rate": 1.2958135342424071e-08,
+      "loss": 0.0549,
+      "step": 33950
+    },
+    {
+      "epoch": 2.934036027474189,
+      "grad_norm": 0.678915128766411,
+      "learning_rate": 1.262474702549865e-08,
+      "loss": 0.0569,
+      "step": 33960
+    },
+    {
+      "epoch": 2.934899995680159,
+      "grad_norm": 0.6801977159105094,
+      "learning_rate": 1.2295698087685937e-08,
+      "loss": 0.0582,
+      "step": 33970
+    },
+    {
+      "epoch": 2.935763963886129,
+      "grad_norm": 0.630776425498901,
+      "learning_rate": 1.1970988815283025e-08,
+      "loss": 0.0563,
+      "step": 33980
+    },
+    {
+      "epoch": 2.936627932092099,
+      "grad_norm": 0.6857107028827767,
+      "learning_rate": 1.1650619490810588e-08,
+      "loss": 0.0551,
+      "step": 33990
+    },
+    {
+      "epoch": 2.937491900298069,
+      "grad_norm": 0.6480468923534133,
+      "learning_rate": 1.1334590393013434e-08,
+      "loss": 0.0559,
+      "step": 34000
+    },
+    {
+      "epoch": 2.938355868504039,
+      "grad_norm": 0.6734319616204011,
+      "learning_rate": 1.1022901796858832e-08,
+      "loss": 0.0559,
+      "step": 34010
+    },
+    {
+      "epoch": 2.9392198367100093,
+      "grad_norm": 0.6603543423140258,
+      "learning_rate": 1.0715553973539849e-08,
+      "loss": 0.0588,
+      "step": 34020
+    },
+    {
+      "epoch": 2.940083804915979,
+      "grad_norm": 0.7978284748396589,
+      "learning_rate": 1.0412547190470912e-08,
+      "loss": 0.0564,
+      "step": 34030
+    },
+    {
+      "epoch": 2.940947773121949,
+      "grad_norm": 0.681328579342505,
+      "learning_rate": 1.0113881711289464e-08,
+      "loss": 0.0581,
+      "step": 34040
+    },
+    {
+      "epoch": 2.941811741327919,
+      "grad_norm": 0.6893855973861117,
+      "learning_rate": 9.819557795857082e-09,
+      "loss": 0.058,
+      "step": 34050
+    },
+    {
+      "epoch": 2.942675709533889,
+      "grad_norm": 0.6615292039101736,
+      "learning_rate": 9.52957570025559e-09,
+      "loss": 0.0558,
+      "step": 34060
+    },
+    {
+      "epoch": 2.9435396777398593,
+      "grad_norm": 0.6897433502814333,
+      "learning_rate": 9.243935676792603e-09,
+      "loss": 0.0573,
+      "step": 34070
+    },
+    {
+      "epoch": 2.944403645945829,
+      "grad_norm": 0.6570832913653472,
+      "learning_rate": 8.96263797399377e-09,
+      "loss": 0.0556,
+      "step": 34080
+    },
+    {
+      "epoch": 2.945267614151799,
+      "grad_norm": 0.6792304176789006,
+      "learning_rate": 8.685682836609421e-09,
+      "loss": 0.0543,
+      "step": 34090
+    },
+    {
+      "epoch": 2.9461315823577694,
+      "grad_norm": 0.6579739271767348,
+      "learning_rate": 8.413070505610132e-09,
+      "loss": 0.058,
+      "step": 34100
+    },
+    {
+      "epoch": 2.9469955505637393,
+      "grad_norm": 0.669809562415206,
+      "learning_rate": 8.144801218189502e-09,
+      "loss": 0.0578,
+      "step": 34110
+    },
+    {
+      "epoch": 2.947859518769709,
+      "grad_norm": 0.6799320416148782,
+      "learning_rate": 7.880875207760264e-09,
+      "loss": 0.0567,
+      "step": 34120
+    },
+    {
+      "epoch": 2.9487234869756795,
+      "grad_norm": 0.6629897134562075,
+      "learning_rate": 7.621292703957617e-09,
+      "loss": 0.0578,
+      "step": 34130
+    },
+    {
+      "epoch": 2.9495874551816494,
+      "grad_norm": 0.6552021162116146,
+      "learning_rate": 7.366053932637563e-09,
+      "loss": 0.0538,
+      "step": 34140
+    },
+    {
+      "epoch": 2.9504514233876193,
+      "grad_norm": 0.683065500854779,
+      "learning_rate": 7.115159115875792e-09,
+      "loss": 0.0553,
+      "step": 34150
+    },
+    {
+      "epoch": 2.9513153915935892,
+      "grad_norm": 0.6874739280286147,
+      "learning_rate": 6.8686084719693515e-09,
+      "loss": 0.0597,
+      "step": 34160
+    },
+    {
+      "epoch": 2.952179359799559,
+      "grad_norm": 0.6422584639658296,
+      "learning_rate": 6.626402215434979e-09,
+      "loss": 0.0569,
+      "step": 34170
+    },
+    {
+      "epoch": 2.9530433280055295,
+      "grad_norm": 0.7158995024434536,
+      "learning_rate": 6.388540557010214e-09,
+      "loss": 0.0574,
+      "step": 34180
+    },
+    {
+      "epoch": 2.9539072962114994,
+      "grad_norm": 0.6680913688006366,
+      "learning_rate": 6.15502370365173e-09,
+      "loss": 0.0564,
+      "step": 34190
+    },
+    {
+      "epoch": 2.9547712644174693,
+      "grad_norm": 0.7093620408628543,
+      "learning_rate": 5.925851858535891e-09,
+      "loss": 0.0567,
+      "step": 34200
+    },
+    {
+      "epoch": 2.9556352326234396,
+      "grad_norm": 0.6286177131295109,
+      "learning_rate": 5.7010252210593085e-09,
+      "loss": 0.0561,
+      "step": 34210
+    },
+    {
+      "epoch": 2.9564992008294095,
+      "grad_norm": 0.6348257609589285,
+      "learning_rate": 5.480543986837172e-09,
+      "loss": 0.0525,
+      "step": 34220
+    },
+    {
+      "epoch": 2.9573631690353794,
+      "grad_norm": 0.7138653645631411,
+      "learning_rate": 5.264408347704364e-09,
+      "loss": 0.0563,
+      "step": 34230
+    },
+    {
+      "epoch": 2.95822713724135,
+      "grad_norm": 0.6958979162027398,
+      "learning_rate": 5.0526184917149e-09,
+      "loss": 0.0545,
+      "step": 34240
+    },
+    {
+      "epoch": 2.9590911054473197,
+      "grad_norm": 0.6673351506192208,
+      "learning_rate": 4.845174603140823e-09,
+      "loss": 0.0573,
+      "step": 34250
+    },
+    {
+      "epoch": 2.9599550736532896,
+      "grad_norm": 0.6597776756569177,
+      "learning_rate": 4.6420768624738654e-09,
+      "loss": 0.0543,
+      "step": 34260
+    },
+    {
+      "epoch": 2.9608190418592595,
+      "grad_norm": 0.6907853197848071,
+      "learning_rate": 4.44332544642323e-09,
+      "loss": 0.0561,
+      "step": 34270
+    },
+    {
+      "epoch": 2.9616830100652294,
+      "grad_norm": 0.6367120740998353,
+      "learning_rate": 4.2489205279172555e-09,
+      "loss": 0.0556,
+      "step": 34280
+    },
+    {
+      "epoch": 2.9625469782711997,
+      "grad_norm": 0.6835548939710773,
+      "learning_rate": 4.05886227610286e-09,
+      "loss": 0.0563,
+      "step": 34290
+    },
+    {
+      "epoch": 2.9634109464771696,
+      "grad_norm": 0.6656890164686422,
+      "learning_rate": 3.873150856344432e-09,
+      "loss": 0.0568,
+      "step": 34300
+    },
+    {
+      "epoch": 2.9642749146831395,
+      "grad_norm": 0.6556734265025637,
+      "learning_rate": 3.6917864302238315e-09,
+      "loss": 0.0564,
+      "step": 34310
+    },
+    {
+      "epoch": 2.96513888288911,
+      "grad_norm": 0.7248595471350836,
+      "learning_rate": 3.5147691555420526e-09,
+      "loss": 0.057,
+      "step": 34320
+    },
+    {
+      "epoch": 2.9660028510950798,
+      "grad_norm": 0.662123530396553,
+      "learning_rate": 3.34209918631645e-09,
+      "loss": 0.0583,
+      "step": 34330
+    },
+    {
+      "epoch": 2.9668668193010497,
+      "grad_norm": 0.6509452814103437,
+      "learning_rate": 3.1737766727824028e-09,
+      "loss": 0.0582,
+      "step": 34340
+    },
+    {
+      "epoch": 2.96773078750702,
+      "grad_norm": 0.7050813393203283,
+      "learning_rate": 3.0098017613933163e-09,
+      "loss": 0.0564,
+      "step": 34350
+    },
+    {
+      "epoch": 2.96859475571299,
+      "grad_norm": 0.6654292686780983,
+      "learning_rate": 2.8501745948184e-09,
+      "loss": 0.0559,
+      "step": 34360
+    },
+    {
+      "epoch": 2.96945872391896,
+      "grad_norm": 0.6821160988216692,
+      "learning_rate": 2.6948953119459997e-09,
+      "loss": 0.0571,
+      "step": 34370
+    },
+    {
+      "epoch": 2.9703226921249297,
+      "grad_norm": 0.684418049892257,
+      "learning_rate": 2.54396404787971e-09,
+      "loss": 0.0555,
+      "step": 34380
+    },
+    {
+      "epoch": 2.9711866603308996,
+      "grad_norm": 0.6814190481168771,
+      "learning_rate": 2.3973809339405963e-09,
+      "loss": 0.0571,
+      "step": 34390
+    },
+    {
+      "epoch": 2.97205062853687,
+      "grad_norm": 0.6795433479173877,
+      "learning_rate": 2.2551460976666385e-09,
+      "loss": 0.0565,
+      "step": 34400
+    },
+    {
+      "epoch": 2.97291459674284,
+      "grad_norm": 0.6785623293407695,
+      "learning_rate": 2.1172596628127316e-09,
+      "loss": 0.0567,
+      "step": 34410
+    },
+    {
+      "epoch": 2.9737785649488098,
+      "grad_norm": 0.6809051694327195,
+      "learning_rate": 1.9837217493501313e-09,
+      "loss": 0.0555,
+      "step": 34420
+    },
+    {
+      "epoch": 2.97464253315478,
+      "grad_norm": 0.6369055902211925,
+      "learning_rate": 1.8545324734664526e-09,
+      "loss": 0.0569,
+      "step": 34430
+    },
+    {
+      "epoch": 2.97550650136075,
+      "grad_norm": 0.6910310396003512,
+      "learning_rate": 1.7296919475662254e-09,
+      "loss": 0.0562,
+      "step": 34440
+    },
+    {
+      "epoch": 2.97637046956672,
+      "grad_norm": 0.7202616011304915,
+      "learning_rate": 1.6092002802686747e-09,
+      "loss": 0.0578,
+      "step": 34450
+    },
+    {
+      "epoch": 2.97723443777269,
+      "grad_norm": 0.6564232420344026,
+      "learning_rate": 1.49305757641105e-09,
+      "loss": 0.0561,
+      "step": 34460
+    },
+    {
+      "epoch": 2.9780984059786597,
+      "grad_norm": 0.6755704437504937,
+      "learning_rate": 1.3812639370458514e-09,
+      "loss": 0.0579,
+      "step": 34470
+    },
+    {
+      "epoch": 2.97896237418463,
+      "grad_norm": 0.6502855806426288,
+      "learning_rate": 1.2738194594419384e-09,
+      "loss": 0.0557,
+      "step": 34480
+    },
+    {
+      "epoch": 2.9798263423906,
+      "grad_norm": 0.6439414003128155,
+      "learning_rate": 1.1707242370834204e-09,
+      "loss": 0.0575,
+      "step": 34490
+    },
+    {
+      "epoch": 2.98069031059657,
+      "grad_norm": 0.6340317895115861,
+      "learning_rate": 1.0719783596707666e-09,
+      "loss": 0.0564,
+      "step": 34500
+    },
+    {
+      "epoch": 2.98155427880254,
+      "grad_norm": 0.6753564648972211,
+      "learning_rate": 9.775819131202513e-10,
+      "loss": 0.0556,
+      "step": 34510
+    },
+    {
+      "epoch": 2.98241824700851,
+      "grad_norm": 0.6455724063217956,
+      "learning_rate": 8.87534979562843e-10,
+      "loss": 0.0559,
+      "step": 34520
+    },
+    {
+      "epoch": 2.98328221521448,
+      "grad_norm": 0.6909467453570447,
+      "learning_rate": 8.018376373469805e-10,
+      "loss": 0.0557,
+      "step": 34530
+    },
+    {
+      "epoch": 2.9841461834204503,
+      "grad_norm": 0.6744291133600068,
+      "learning_rate": 7.204899610352423e-10,
+      "loss": 0.0538,
+      "step": 34540
+    },
+    {
+      "epoch": 2.9850101516264202,
+      "grad_norm": 0.6400069292789241,
+      "learning_rate": 6.434920214060114e-10,
+      "loss": 0.0565,
+      "step": 34550
+    },
+    {
+      "epoch": 2.98587411983239,
+      "grad_norm": 0.6660665422309995,
+      "learning_rate": 5.708438854523656e-10,
+      "loss": 0.0567,
+      "step": 34560
+    },
+    {
+      "epoch": 2.98673808803836,
+      "grad_norm": 0.6877542654558588,
+      "learning_rate": 5.02545616384853e-10,
+      "loss": 0.0589,
+      "step": 34570
+    },
+    {
+      "epoch": 2.98760205624433,
+      "grad_norm": 0.6685372840389431,
+      "learning_rate": 4.385972736264954e-10,
+      "loss": 0.0567,
+      "step": 34580
+    },
+    {
+      "epoch": 2.9884660244503003,
+      "grad_norm": 0.6579606458219801,
+      "learning_rate": 3.7899891281834025e-10,
+      "loss": 0.0568,
+      "step": 34590
+    },
+    {
+      "epoch": 2.98932999265627,
+      "grad_norm": 0.6810902715442537,
+      "learning_rate": 3.2375058581446407e-10,
+      "loss": 0.0585,
+      "step": 34600
+    },
+    {
+      "epoch": 2.99019396086224,
+      "grad_norm": 0.6969850812240518,
+      "learning_rate": 2.728523406847483e-10,
+      "loss": 0.0569,
+      "step": 34610
+    },
+    {
+      "epoch": 2.9910579290682104,
+      "grad_norm": 0.6943564708722185,
+      "learning_rate": 2.263042217148792e-10,
+      "loss": 0.0569,
+      "step": 34620
+    },
+    {
+      "epoch": 2.9919218972741803,
+      "grad_norm": 0.7012718186319877,
+      "learning_rate": 1.8410626940523757e-10,
+      "loss": 0.0564,
+      "step": 34630
+    },
+    {
+      "epoch": 2.9927858654801502,
+      "grad_norm": 0.6768713971350516,
+      "learning_rate": 1.4625852047034373e-10,
+      "loss": 0.0591,
+      "step": 34640
+    },
+    {
+      "epoch": 2.9936498336861206,
+      "grad_norm": 0.6859084073565257,
+      "learning_rate": 1.1276100784163302e-10,
+      "loss": 0.0576,
+      "step": 34650
+    },
+    {
+      "epoch": 2.9945138018920905,
+      "grad_norm": 0.6851362984991908,
+      "learning_rate": 8.361376066357007e-11,
+      "loss": 0.0553,
+      "step": 34660
+    },
+    {
+      "epoch": 2.9953777700980604,
+      "grad_norm": 0.6529437264727286,
+      "learning_rate": 5.881680429586922e-11,
+      "loss": 0.0546,
+      "step": 34670
+    },
+    {
+      "epoch": 2.9962417383040303,
+      "grad_norm": 0.6852860443468494,
+      "learning_rate": 3.8370160315159834e-11,
+      "loss": 0.0563,
+      "step": 34680
+    },
+    {
+      "epoch": 2.99710570651,
+      "grad_norm": 0.6575050306681262,
+      "learning_rate": 2.2273846509990314e-11,
+      "loss": 0.0579,
+      "step": 34690
+    },
+    {
+      "epoch": 2.9979696747159705,
+      "grad_norm": 0.6579715634072224,
+      "learning_rate": 1.052787688637924e-11,
+      "loss": 0.0549,
+      "step": 34700
+    },
+    {
+      "epoch": 2.9988336429219404,
+      "grad_norm": 0.7080004725681195,
+      "learning_rate": 3.132261663929548e-12,
+      "loss": 0.06,
+      "step": 34710
+    },
+    {
+      "epoch": 2.9996976111279103,
+      "grad_norm": 0.6662699930555676,
+      "learning_rate": 8.700727749388904e-14,
+      "loss": 0.058,
+      "step": 34720
+    },
+    {
+      "epoch": 2.9998704047691045,
+      "step": 34722,
+      "total_flos": 7939707523891200.0,
+      "train_loss": 0.11055064558756254,
+      "train_runtime": 74830.8535,
+      "train_samples_per_second": 59.395,
+      "train_steps_per_second": 0.464
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 34722,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50000000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7939707523891200.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}