diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7981 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 5000,
+  "global_step": 5666,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0017649135192375574,
+      "grad_norm": 0.9870162796904358,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 2.3057,
+      "step": 5
+    },
+    {
+      "epoch": 0.0035298270384751147,
+      "grad_norm": 1.2696540200103463,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 2.4562,
+      "step": 10
+    },
+    {
+      "epoch": 0.005294740557712672,
+      "grad_norm": 0.23777213433556077,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 2.2577,
+      "step": 15
+    },
+    {
+      "epoch": 0.0070596540769502295,
+      "grad_norm": 0.2920555522596246,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 2.4244,
+      "step": 20
+    },
+    {
+      "epoch": 0.008824567596187787,
+      "grad_norm": 0.2248081862851944,
+      "learning_rate": 7.5e-05,
+      "loss": 2.4041,
+      "step": 25
+    },
+    {
+      "epoch": 0.010589481115425344,
+      "grad_norm": 0.18401943742866847,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 2.204,
+      "step": 30
+    },
+    {
+      "epoch": 0.012354394634662902,
+      "grad_norm": 0.19354324682907367,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 2.3191,
+      "step": 35
+    },
+    {
+      "epoch": 0.014119308153900459,
+      "grad_norm": 0.16059440131333053,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 2.2965,
+      "step": 40
+    },
+    {
+      "epoch": 0.015884221673138015,
+      "grad_norm": 0.16328667697622465,
+      "learning_rate": 0.000135,
+      "loss": 2.195,
+      "step": 45
+    },
+    {
+      "epoch": 0.017649135192375574,
+      "grad_norm": 0.22972784310095215,
+      "learning_rate": 0.00015,
+      "loss": 2.3032,
+      "step": 50
+    },
+    {
+      "epoch": 0.01941404871161313,
+      "grad_norm": 0.15973579864036214,
+      "learning_rate": 0.000165,
+      "loss": 2.1918,
+      "step": 55
+    },
+    {
+      "epoch": 0.02117896223085069,
+      "grad_norm": 0.14556204808144899,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 2.3392,
+      "step": 60
+    },
+    {
+      "epoch": 0.022943875750088244,
+      "grad_norm": 0.15187375699081626,
+      "learning_rate": 0.000195,
+      "loss": 2.2509,
+      "step": 65
+    },
+    {
+      "epoch": 0.024708789269325803,
+      "grad_norm": 0.14583247519304798,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 2.262,
+      "step": 70
+    },
+    {
+      "epoch": 0.02647370278856336,
+      "grad_norm": 0.130571828262095,
+      "learning_rate": 0.000225,
+      "loss": 2.0138,
+      "step": 75
+    },
+    {
+      "epoch": 0.028238616307800918,
+      "grad_norm": 0.12570822735245077,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 2.2382,
+      "step": 80
+    },
+    {
+      "epoch": 0.030003529827038473,
+      "grad_norm": 0.100863621514295,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 2.2271,
+      "step": 85
+    },
+    {
+      "epoch": 0.03176844334627603,
+      "grad_norm": 0.11678822131319391,
+      "learning_rate": 0.00027,
+      "loss": 2.3561,
+      "step": 90
+    },
+    {
+      "epoch": 0.03353335686551359,
+      "grad_norm": 0.4799056072379646,
+      "learning_rate": 0.000285,
+      "loss": 2.2833,
+      "step": 95
+    },
+    {
+      "epoch": 0.03529827038475115,
+      "grad_norm": 0.06140412715809612,
+      "learning_rate": 0.0003,
+      "loss": 2.1165,
+      "step": 100
+    },
+    {
+      "epoch": 0.0370631839039887,
+      "grad_norm": 0.09284059082940595,
+      "learning_rate": 0.00029999940267000255,
+      "loss": 2.0447,
+      "step": 105
+    },
+    {
+      "epoch": 0.03882809742322626,
+      "grad_norm": 0.06391784311117,
+      "learning_rate": 0.00029999761068476765,
+      "loss": 2.4483,
+      "step": 110
+    },
+    {
+      "epoch": 0.04059301094246382,
+      "grad_norm": 0.05488718519971469,
+      "learning_rate": 0.0002999946240585673,
+      "loss": 2.2005,
+      "step": 115
+    },
+    {
+      "epoch": 0.04235792446170138,
+      "grad_norm": 0.05700323093547788,
+      "learning_rate": 0.00029999044281518825,
+      "loss": 2.4502,
+      "step": 120
+    },
+    {
+      "epoch": 0.04412283798093893,
+      "grad_norm": 0.05899689278027543,
+      "learning_rate": 0.0002999850669879316,
+      "loss": 2.2069,
+      "step": 125
+    },
+    {
+      "epoch": 0.04588775150017649,
+      "grad_norm": 0.059534772088159695,
+      "learning_rate": 0.0002999784966196125,
+      "loss": 2.2102,
+      "step": 130
+    },
+    {
+      "epoch": 0.04765266501941405,
+      "grad_norm": 0.0666704275619814,
+      "learning_rate": 0.00029997073176256015,
+      "loss": 2.2909,
+      "step": 135
+    },
+    {
+      "epoch": 0.049417578538651606,
+      "grad_norm": 0.05118624740319015,
+      "learning_rate": 0.00029996177247861684,
+      "loss": 2.0928,
+      "step": 140
+    },
+    {
+      "epoch": 0.05118249205788916,
+      "grad_norm": 0.058912964325059915,
+      "learning_rate": 0.00029995161883913794,
+      "loss": 2.1318,
+      "step": 145
+    },
+    {
+      "epoch": 0.05294740557712672,
+      "grad_norm": 0.054397362904488586,
+      "learning_rate": 0.0002999402709249911,
+      "loss": 2.0537,
+      "step": 150
+    },
+    {
+      "epoch": 0.05471231909636428,
+      "grad_norm": 0.052922745671143676,
+      "learning_rate": 0.00029992772882655566,
+      "loss": 2.1201,
+      "step": 155
+    },
+    {
+      "epoch": 0.056477232615601836,
+      "grad_norm": 0.07482784216874547,
+      "learning_rate": 0.00029991399264372184,
+      "loss": 2.2252,
+      "step": 160
+    },
+    {
+      "epoch": 0.05824214613483939,
+      "grad_norm": 0.06445190734138931,
+      "learning_rate": 0.0002998990624858901,
+      "loss": 2.1946,
+      "step": 165
+    },
+    {
+      "epoch": 0.06000705965407695,
+      "grad_norm": 0.05508339456152396,
+      "learning_rate": 0.00029988293847197034,
+      "loss": 2.1067,
+      "step": 170
+    },
+    {
+      "epoch": 0.06177197317331451,
+      "grad_norm": 0.06227507178040559,
+      "learning_rate": 0.00029986562073038047,
+      "loss": 2.3251,
+      "step": 175
+    },
+    {
+      "epoch": 0.06353688669255206,
+      "grad_norm": 0.05596410282220337,
+      "learning_rate": 0.00029984710939904603,
+      "loss": 2.0071,
+      "step": 180
+    },
+    {
+      "epoch": 0.06530180021178962,
+      "grad_norm": 0.07121635939914675,
+      "learning_rate": 0.0002998274046253986,
+      "loss": 2.2113,
+      "step": 185
+    },
+    {
+      "epoch": 0.06706671373102718,
+      "grad_norm": 0.05498123627050147,
+      "learning_rate": 0.0002998065065663749,
+      "loss": 2.2285,
+      "step": 190
+    },
+    {
+      "epoch": 0.06883162725026473,
+      "grad_norm": 0.05904016772193954,
+      "learning_rate": 0.00029978441538841544,
+      "loss": 2.171,
+      "step": 195
+    },
+    {
+      "epoch": 0.0705965407695023,
+      "grad_norm": 0.10225601868235279,
+      "learning_rate": 0.00029976113126746316,
+      "loss": 2.0817,
+      "step": 200
+    },
+    {
+      "epoch": 0.07236145428873986,
+      "grad_norm": 0.1206878560479012,
+      "learning_rate": 0.00029973665438896217,
+      "loss": 2.2159,
+      "step": 205
+    },
+    {
+      "epoch": 0.0741263678079774,
+      "grad_norm": 0.09697684972242247,
+      "learning_rate": 0.0002997109849478561,
+      "loss": 2.1484,
+      "step": 210
+    },
+    {
+      "epoch": 0.07589128132721497,
+      "grad_norm": 0.08926248535681242,
+      "learning_rate": 0.0002996841231485866,
+      "loss": 2.1547,
+      "step": 215
+    },
+    {
+      "epoch": 0.07765619484645252,
+      "grad_norm": 0.08365430219933052,
+      "learning_rate": 0.00029965606920509183,
+      "loss": 2.1373,
+      "step": 220
+    },
+    {
+      "epoch": 0.07942110836569008,
+      "grad_norm": 0.08751331951687442,
+      "learning_rate": 0.0002996268233408046,
+      "loss": 2.2049,
+      "step": 225
+    },
+    {
+      "epoch": 0.08118602188492764,
+      "grad_norm": 0.08611489013745913,
+      "learning_rate": 0.0002995963857886507,
+      "loss": 2.1362,
+      "step": 230
+    },
+    {
+      "epoch": 0.08295093540416519,
+      "grad_norm": 0.0924452715515503,
+      "learning_rate": 0.0002995647567910469,
+      "loss": 2.055,
+      "step": 235
+    },
+    {
+      "epoch": 0.08471584892340275,
+      "grad_norm": 0.08594549823823565,
+      "learning_rate": 0.0002995319365998993,
+      "loss": 2.0891,
+      "step": 240
+    },
+    {
+      "epoch": 0.08648076244264032,
+      "grad_norm": 0.0878338915817975,
+      "learning_rate": 0.000299497925476601,
+      "loss": 2.2344,
+      "step": 245
+    },
+    {
+      "epoch": 0.08824567596187786,
+      "grad_norm": 0.09203765946832222,
+      "learning_rate": 0.00029946272369203004,
+      "loss": 2.2645,
+      "step": 250
+    },
+    {
+      "epoch": 0.09001058948111543,
+      "grad_norm": 0.09048053309391206,
+      "learning_rate": 0.0002994263315265477,
+      "loss": 2.1878,
+      "step": 255
+    },
+    {
+      "epoch": 0.09177550300035298,
+      "grad_norm": 0.08645372426143949,
+      "learning_rate": 0.0002993887492699956,
+      "loss": 2.3129,
+      "step": 260
+    },
+    {
+      "epoch": 0.09354041651959054,
+      "grad_norm": 0.08218396733511944,
+      "learning_rate": 0.000299349977221694,
+      "loss": 1.9386,
+      "step": 265
+    },
+    {
+      "epoch": 0.0953053300388281,
+      "grad_norm": 0.10833457956923179,
+      "learning_rate": 0.0002993100156904389,
+      "loss": 2.2377,
+      "step": 270
+    },
+    {
+      "epoch": 0.09707024355806565,
+      "grad_norm": 0.09201168120090504,
+      "learning_rate": 0.00029926886499449996,
+      "loss": 2.3703,
+      "step": 275
+    },
+    {
+      "epoch": 0.09883515707730321,
+      "grad_norm": 0.08900707862028383,
+      "learning_rate": 0.0002992265254616177,
+      "loss": 2.0394,
+      "step": 280
+    },
+    {
+      "epoch": 0.10060007059654077,
+      "grad_norm": 0.10412129279413046,
+      "learning_rate": 0.00029918299742900126,
+      "loss": 2.2186,
+      "step": 285
+    },
+    {
+      "epoch": 0.10236498411577832,
+      "grad_norm": 0.0867323431073369,
+      "learning_rate": 0.00029913828124332513,
+      "loss": 2.1372,
+      "step": 290
+    },
+    {
+      "epoch": 0.10412989763501589,
+      "grad_norm": 0.09600642456697991,
+      "learning_rate": 0.000299092377260727,
+      "loss": 2.1068,
+      "step": 295
+    },
+    {
+      "epoch": 0.10589481115425343,
+      "grad_norm": 0.04175588339217386,
+      "learning_rate": 0.0002990452858468045,
+      "loss": 2.2192,
+      "step": 300
+    },
+    {
+      "epoch": 0.107659724673491,
+      "grad_norm": 0.04180583872091638,
+      "learning_rate": 0.0002989970073766125,
+      "loss": 2.1785,
+      "step": 305
+    },
+    {
+      "epoch": 0.10942463819272856,
+      "grad_norm": 0.04077262342802702,
+      "learning_rate": 0.0002989475422346601,
+      "loss": 2.2577,
+      "step": 310
+    },
+    {
+      "epoch": 0.11118955171196611,
+      "grad_norm": 0.03743152652028906,
+      "learning_rate": 0.0002988968908149074,
+      "loss": 2.2192,
+      "step": 315
+    },
+    {
+      "epoch": 0.11295446523120367,
+      "grad_norm": 0.0369669158283002,
+      "learning_rate": 0.00029884505352076264,
+      "loss": 2.2923,
+      "step": 320
+    },
+    {
+      "epoch": 0.11471937875044123,
+      "grad_norm": 0.03766638227324958,
+      "learning_rate": 0.0002987920307650786,
+      "loss": 2.1331,
+      "step": 325
+    },
+    {
+      "epoch": 0.11648429226967878,
+      "grad_norm": 0.036718435576890515,
+      "learning_rate": 0.0002987378229701499,
+      "loss": 2.0629,
+      "step": 330
+    },
+    {
+      "epoch": 0.11824920578891635,
+      "grad_norm": 0.03475908972372543,
+      "learning_rate": 0.0002986824305677091,
+      "loss": 2.0498,
+      "step": 335
+    },
+    {
+      "epoch": 0.1200141193081539,
+      "grad_norm": 0.03413128313818488,
+      "learning_rate": 0.00029862585399892324,
+      "loss": 2.2615,
+      "step": 340
+    },
+    {
+      "epoch": 0.12177903282739146,
+      "grad_norm": 0.036131244030678,
+      "learning_rate": 0.000298568093714391,
+      "loss": 2.1939,
+      "step": 345
+    },
+    {
+      "epoch": 0.12354394634662902,
+      "grad_norm": 0.03559766773507592,
+      "learning_rate": 0.0002985091501741382,
+      "loss": 2.2365,
+      "step": 350
+    },
+    {
+      "epoch": 0.12530885986586657,
+      "grad_norm": 0.036492021233864184,
+      "learning_rate": 0.0002984490238476148,
+      "loss": 2.1711,
+      "step": 355
+    },
+    {
+      "epoch": 0.12707377338510412,
+      "grad_norm": 0.036729294690958515,
+      "learning_rate": 0.000298387715213691,
+      "loss": 2.2582,
+      "step": 360
+    },
+    {
+      "epoch": 0.1288386869043417,
+      "grad_norm": 0.041367703842070995,
+      "learning_rate": 0.00029832522476065323,
+      "loss": 2.1152,
+      "step": 365
+    },
+    {
+      "epoch": 0.13060360042357924,
+      "grad_norm": 0.03815542267761237,
+      "learning_rate": 0.0002982615529862004,
+      "loss": 2.0446,
+      "step": 370
+    },
+    {
+      "epoch": 0.1323685139428168,
+      "grad_norm": 0.036363249569976774,
+      "learning_rate": 0.0002981967003974401,
+      "loss": 2.1343,
+      "step": 375
+    },
+    {
+      "epoch": 0.13413342746205437,
+      "grad_norm": 0.036853678676345404,
+      "learning_rate": 0.00029813066751088426,
+      "loss": 2.2858,
+      "step": 380
+    },
+    {
+      "epoch": 0.13589834098129192,
+      "grad_norm": 0.03813670814046979,
+      "learning_rate": 0.0002980634548524451,
+      "loss": 2.06,
+      "step": 385
+    },
+    {
+      "epoch": 0.13766325450052946,
+      "grad_norm": 0.04065031480766099,
+      "learning_rate": 0.0002979950629574312,
+      "loss": 2.2976,
+      "step": 390
+    },
+    {
+      "epoch": 0.13942816801976704,
+      "grad_norm": 0.03852663679252818,
+      "learning_rate": 0.000297925492370543,
+      "loss": 2.2986,
+      "step": 395
+    },
+    {
+      "epoch": 0.1411930815390046,
+      "grad_norm": 0.03696331701078347,
+      "learning_rate": 0.00029785474364586837,
+      "loss": 2.2626,
+      "step": 400
+    },
+    {
+      "epoch": 0.14295799505824214,
+      "grad_norm": 0.044040908964770184,
+      "learning_rate": 0.00029778281734687856,
+      "loss": 2.0912,
+      "step": 405
+    },
+    {
+      "epoch": 0.14472290857747971,
+      "grad_norm": 0.0372882318058729,
+      "learning_rate": 0.0002977097140464233,
+      "loss": 2.2395,
+      "step": 410
+    },
+    {
+      "epoch": 0.14648782209671726,
+      "grad_norm": 0.03468254500555961,
+      "learning_rate": 0.0002976354343267265,
+      "loss": 2.1716,
+      "step": 415
+    },
+    {
+      "epoch": 0.1482527356159548,
+      "grad_norm": 0.03635593346985757,
+      "learning_rate": 0.00029755997877938163,
+      "loss": 2.1567,
+      "step": 420
+    },
+    {
+      "epoch": 0.1500176491351924,
+      "grad_norm": 0.03470673777171037,
+      "learning_rate": 0.0002974833480053469,
+      "loss": 2.057,
+      "step": 425
+    },
+    {
+      "epoch": 0.15178256265442994,
+      "grad_norm": 0.03519535892047795,
+      "learning_rate": 0.0002974055426149403,
+      "loss": 2.3069,
+      "step": 430
+    },
+    {
+      "epoch": 0.15354747617366749,
+      "grad_norm": 0.03557023333314978,
+      "learning_rate": 0.00029732656322783513,
+      "loss": 2.1936,
+      "step": 435
+    },
+    {
+      "epoch": 0.15531238969290503,
+      "grad_norm": 0.03341152268324352,
+      "learning_rate": 0.0002972464104730549,
+      "loss": 2.3136,
+      "step": 440
+    },
+    {
+      "epoch": 0.1570773032121426,
+      "grad_norm": 0.03390532463348913,
+      "learning_rate": 0.00029716508498896807,
+      "loss": 2.1568,
+      "step": 445
+    },
+    {
+      "epoch": 0.15884221673138016,
+      "grad_norm": 0.035934397571484936,
+      "learning_rate": 0.0002970825874232834,
+      "loss": 2.1949,
+      "step": 450
+    },
+    {
+      "epoch": 0.1606071302506177,
+      "grad_norm": 0.03345542998218734,
+      "learning_rate": 0.0002969989184330445,
+      "loss": 2.1642,
+      "step": 455
+    },
+    {
+      "epoch": 0.16237204376985528,
+      "grad_norm": 0.03226997917878771,
+      "learning_rate": 0.00029691407868462464,
+      "loss": 2.1097,
+      "step": 460
+    },
+    {
+      "epoch": 0.16413695728909283,
+      "grad_norm": 0.03277396003569313,
+      "learning_rate": 0.00029682806885372156,
+      "loss": 2.2478,
+      "step": 465
+    },
+    {
+      "epoch": 0.16590187080833038,
+      "grad_norm": 0.033555291710766426,
+      "learning_rate": 0.00029674088962535194,
+      "loss": 2.0988,
+      "step": 470
+    },
+    {
+      "epoch": 0.16766678432756796,
+      "grad_norm": 0.03455582331158057,
+      "learning_rate": 0.000296652541693846,
+      "loss": 2.1584,
+      "step": 475
+    },
+    {
+      "epoch": 0.1694316978468055,
+      "grad_norm": 0.032840245870870644,
+      "learning_rate": 0.000296563025762842,
+      "loss": 2.2052,
+      "step": 480
+    },
+    {
+      "epoch": 0.17119661136604306,
+      "grad_norm": 0.03649344986848586,
+      "learning_rate": 0.00029647234254528065,
+      "loss": 2.318,
+      "step": 485
+    },
+    {
+      "epoch": 0.17296152488528063,
+      "grad_norm": 0.034228935154978664,
+      "learning_rate": 0.00029638049276339937,
+      "loss": 2.2473,
+      "step": 490
+    },
+    {
+      "epoch": 0.17472643840451818,
+      "grad_norm": 0.033305402111050667,
+      "learning_rate": 0.0002962874771487265,
+      "loss": 2.18,
+      "step": 495
+    },
+    {
+      "epoch": 0.17649135192375573,
+      "grad_norm": 0.034605741747086013,
+      "learning_rate": 0.00029619329644207563,
+      "loss": 2.101,
+      "step": 500
+    },
+    {
+      "epoch": 0.1782562654429933,
+      "grad_norm": 0.05106461176183687,
+      "learning_rate": 0.00029609795139353963,
+      "loss": 2.1359,
+      "step": 505
+    },
+    {
+      "epoch": 0.18002117896223085,
+      "grad_norm": 0.04738316910513329,
+      "learning_rate": 0.0002960014427624845,
+      "loss": 2.1597,
+      "step": 510
+    },
+    {
+      "epoch": 0.1817860924814684,
+      "grad_norm": 0.03672169099683521,
+      "learning_rate": 0.00029590377131754366,
+      "loss": 2.2543,
+      "step": 515
+    },
+    {
+      "epoch": 0.18355100600070595,
+      "grad_norm": 0.03290499509725145,
+      "learning_rate": 0.0002958049378366115,
+      "loss": 2.0531,
+      "step": 520
+    },
+    {
+      "epoch": 0.18531591951994353,
+      "grad_norm": 0.033252522240519874,
+      "learning_rate": 0.0002957049431068375,
+      "loss": 2.0309,
+      "step": 525
+    },
+    {
+      "epoch": 0.18708083303918108,
+      "grad_norm": 0.034546018162165563,
+      "learning_rate": 0.00029560378792461956,
+      "loss": 2.1264,
+      "step": 530
+    },
+    {
+      "epoch": 0.18884574655841863,
+      "grad_norm": 0.03489106895770067,
+      "learning_rate": 0.00029550147309559806,
+      "loss": 2.076,
+      "step": 535
+    },
+    {
+      "epoch": 0.1906106600776562,
+      "grad_norm": 0.03132334443811583,
+      "learning_rate": 0.0002953979994346492,
+      "loss": 2.1084,
+      "step": 540
+    },
+    {
+      "epoch": 0.19237557359689375,
+      "grad_norm": 0.03185171054665557,
+      "learning_rate": 0.00029529336776587864,
+      "loss": 2.2279,
+      "step": 545
+    },
+    {
+      "epoch": 0.1941404871161313,
+      "grad_norm": 0.032835092236925004,
+      "learning_rate": 0.0002951875789226148,
+      "loss": 2.1829,
+      "step": 550
+    },
+    {
+      "epoch": 0.19590540063536888,
+      "grad_norm": 0.035691822767610695,
+      "learning_rate": 0.0002950806337474024,
+      "loss": 1.9802,
+      "step": 555
+    },
+    {
+      "epoch": 0.19767031415460642,
+      "grad_norm": 0.031613767845956355,
+      "learning_rate": 0.0002949725330919955,
+      "loss": 2.0829,
+      "step": 560
+    },
+    {
+      "epoch": 0.19943522767384397,
+      "grad_norm": 0.03314933655274603,
+      "learning_rate": 0.0002948632778173511,
+      "loss": 2.1448,
+      "step": 565
+    },
+    {
+      "epoch": 0.20120014119308155,
+      "grad_norm": 0.034209151222267656,
+      "learning_rate": 0.0002947528687936217,
+      "loss": 2.097,
+      "step": 570
+    },
+    {
+      "epoch": 0.2029650547123191,
+      "grad_norm": 0.03267788675448161,
+      "learning_rate": 0.00029464130690014907,
+      "loss": 2.059,
+      "step": 575
+    },
+    {
+      "epoch": 0.20472996823155665,
+      "grad_norm": 0.036874612694987166,
+      "learning_rate": 0.0002945285930254567,
+      "loss": 2.1296,
+      "step": 580
+    },
+    {
+      "epoch": 0.20649488175079422,
+      "grad_norm": 0.03566767464519494,
+      "learning_rate": 0.00029441472806724305,
+      "loss": 2.1708,
+      "step": 585
+    },
+    {
+      "epoch": 0.20825979527003177,
+      "grad_norm": 0.05009220178943291,
+      "learning_rate": 0.0002942997129323741,
+      "loss": 2.0103,
+      "step": 590
+    },
+    {
+      "epoch": 0.21002470878926932,
+      "grad_norm": 0.03266455531259752,
+      "learning_rate": 0.00029418354853687635,
+      "loss": 2.178,
+      "step": 595
+    },
+    {
+      "epoch": 0.21178962230850687,
+      "grad_norm": 0.03469882604537321,
+      "learning_rate": 0.00029406623580592963,
+      "loss": 2.1516,
+      "step": 600
+    },
+    {
+      "epoch": 0.21355453582774445,
+      "grad_norm": 0.04568627032197128,
+      "learning_rate": 0.00029394777567385937,
+      "loss": 2.0929,
+      "step": 605
+    },
+    {
+      "epoch": 0.215319449346982,
+      "grad_norm": 0.03832075356380359,
+      "learning_rate": 0.00029382816908412945,
+      "loss": 2.2197,
+      "step": 610
+    },
+    {
+      "epoch": 0.21708436286621954,
+      "grad_norm": 0.033676208545798766,
+      "learning_rate": 0.00029370741698933465,
+      "loss": 2.3015,
+      "step": 615
+    },
+    {
+      "epoch": 0.21884927638545712,
+      "grad_norm": 0.03309688830370379,
+      "learning_rate": 0.00029358552035119287,
+      "loss": 2.0767,
+      "step": 620
+    },
+    {
+      "epoch": 0.22061418990469467,
+      "grad_norm": 0.03129253720400337,
+      "learning_rate": 0.0002934624801405377,
+      "loss": 1.9616,
+      "step": 625
+    },
+    {
+      "epoch": 0.22237910342393222,
+      "grad_norm": 0.031270008948313964,
+      "learning_rate": 0.0002933382973373107,
+      "loss": 1.9926,
+      "step": 630
+    },
+    {
+      "epoch": 0.2241440169431698,
+      "grad_norm": 0.03345258541076931,
+      "learning_rate": 0.0002932129729305532,
+      "loss": 2.1212,
+      "step": 635
+    },
+    {
+      "epoch": 0.22590893046240734,
+      "grad_norm": 0.03281266692932466,
+      "learning_rate": 0.000293086507918399,
+      "loss": 2.0959,
+      "step": 640
+    },
+    {
+      "epoch": 0.2276738439816449,
+      "grad_norm": 0.03526934637872925,
+      "learning_rate": 0.00029295890330806605,
+      "loss": 2.1843,
+      "step": 645
+    },
+    {
+      "epoch": 0.22943875750088247,
+      "grad_norm": 0.03368526402489238,
+      "learning_rate": 0.0002928301601158485,
+      "loss": 2.0457,
+      "step": 650
+    },
+    {
+      "epoch": 0.23120367102012002,
+      "grad_norm": 0.030091547447406056,
+      "learning_rate": 0.0002927002793671086,
+      "loss": 2.0672,
+      "step": 655
+    },
+    {
+      "epoch": 0.23296858453935756,
+      "grad_norm": 0.03266116650214598,
+      "learning_rate": 0.0002925692620962686,
+      "loss": 2.2223,
+      "step": 660
+    },
+    {
+      "epoch": 0.23473349805859514,
+      "grad_norm": 0.032737664374382816,
+      "learning_rate": 0.0002924371093468024,
+      "loss": 2.2904,
+      "step": 665
+    },
+    {
+      "epoch": 0.2364984115778327,
+      "grad_norm": 0.034943051249963254,
+      "learning_rate": 0.00029230382217122747,
+      "loss": 2.0731,
+      "step": 670
+    },
+    {
+      "epoch": 0.23826332509707024,
+      "grad_norm": 0.03308629979203581,
+      "learning_rate": 0.00029216940163109606,
+      "loss": 2.2583,
+      "step": 675
+    },
+    {
+      "epoch": 0.2400282386163078,
+      "grad_norm": 0.03383854148212595,
+      "learning_rate": 0.00029203384879698715,
+      "loss": 2.2264,
+      "step": 680
+    },
+    {
+      "epoch": 0.24179315213554536,
+      "grad_norm": 0.03432256937603642,
+      "learning_rate": 0.00029189716474849783,
+      "loss": 1.9708,
+      "step": 685
+    },
+    {
+      "epoch": 0.2435580656547829,
+      "grad_norm": 0.03649297931913595,
+      "learning_rate": 0.0002917593505742344,
+      "loss": 2.1891,
+      "step": 690
+    },
+    {
+      "epoch": 0.24532297917402046,
+      "grad_norm": 0.033460968686979314,
+      "learning_rate": 0.00029162040737180407,
+      "loss": 2.0362,
+      "step": 695
+    },
+    {
+      "epoch": 0.24708789269325804,
+      "grad_norm": 0.0342962322777911,
+      "learning_rate": 0.0002914803362478061,
+      "loss": 2.0718,
+      "step": 700
+    },
+    {
+      "epoch": 0.2488528062124956,
+      "grad_norm": 0.04259200470850871,
+      "learning_rate": 0.00029133913831782307,
+      "loss": 2.1238,
+      "step": 705
+    },
+    {
+      "epoch": 0.25061771973173314,
+      "grad_norm": 0.03443072792452799,
+      "learning_rate": 0.00029119681470641157,
+      "loss": 2.1241,
+      "step": 710
+    },
+    {
+      "epoch": 0.2523826332509707,
+      "grad_norm": 0.03463099769615107,
+      "learning_rate": 0.0002910533665470939,
+      "loss": 2.1697,
+      "step": 715
+    },
+    {
+      "epoch": 0.25414754677020823,
+      "grad_norm": 0.03309668511207264,
+      "learning_rate": 0.0002909087949823485,
+      "loss": 2.2821,
+      "step": 720
+    },
+    {
+      "epoch": 0.2559124602894458,
+      "grad_norm": 0.033170681828503336,
+      "learning_rate": 0.0002907631011636012,
+      "loss": 2.2057,
+      "step": 725
+    },
+    {
+      "epoch": 0.2576773738086834,
+      "grad_norm": 0.03379275053343088,
+      "learning_rate": 0.00029061628625121576,
+      "loss": 2.1832,
+      "step": 730
+    },
+    {
+      "epoch": 0.2594422873279209,
+      "grad_norm": 0.03349202991869069,
+      "learning_rate": 0.00029046835141448497,
+      "loss": 1.9967,
+      "step": 735
+    },
+    {
+      "epoch": 0.2612072008471585,
+      "grad_norm": 0.03284345125933767,
+      "learning_rate": 0.00029031929783162095,
+      "loss": 2.2037,
+      "step": 740
+    },
+    {
+      "epoch": 0.26297211436639606,
+      "grad_norm": 0.03234900272024827,
+      "learning_rate": 0.0002901691266897461,
+      "loss": 2.1817,
+      "step": 745
+    },
+    {
+      "epoch": 0.2647370278856336,
+      "grad_norm": 0.0343158509752639,
+      "learning_rate": 0.0002900178391848834,
+      "loss": 2.2453,
+      "step": 750
+    },
+    {
+      "epoch": 0.26650194140487116,
+      "grad_norm": 0.031395689877245656,
+      "learning_rate": 0.00028986543652194716,
+      "loss": 2.1835,
+      "step": 755
+    },
+    {
+      "epoch": 0.26826685492410873,
+      "grad_norm": 0.0341025406000622,
+      "learning_rate": 0.00028971191991473304,
+      "loss": 2.2215,
+      "step": 760
+    },
+    {
+      "epoch": 0.27003176844334625,
+      "grad_norm": 0.030903977573464605,
+      "learning_rate": 0.00028955729058590876,
+      "loss": 2.0813,
+      "step": 765
+    },
+    {
+      "epoch": 0.27179668196258383,
+      "grad_norm": 0.03450506880366533,
+      "learning_rate": 0.00028940154976700406,
+      "loss": 2.1511,
+      "step": 770
+    },
+    {
+      "epoch": 0.2735615954818214,
+      "grad_norm": 0.032745788369762274,
+      "learning_rate": 0.00028924469869840124,
+      "loss": 2.0847,
+      "step": 775
+    },
+    {
+      "epoch": 0.27532650900105893,
+      "grad_norm": 0.03414446799000706,
+      "learning_rate": 0.00028908673862932483,
+      "loss": 2.1602,
+      "step": 780
+    },
+    {
+      "epoch": 0.2770914225202965,
+      "grad_norm": 0.033537805234877066,
+      "learning_rate": 0.000288927670817832,
+      "loss": 2.1164,
+      "step": 785
+    },
+    {
+      "epoch": 0.2788563360395341,
+      "grad_norm": 0.032928697653069765,
+      "learning_rate": 0.0002887674965308025,
+      "loss": 2.2996,
+      "step": 790
+    },
+    {
+      "epoch": 0.2806212495587716,
+      "grad_norm": 0.03173309369231481,
+      "learning_rate": 0.0002886062170439284,
+      "loss": 2.1099,
+      "step": 795
+    },
+    {
+      "epoch": 0.2823861630780092,
+      "grad_norm": 0.03249924246847984,
+      "learning_rate": 0.00028844383364170405,
+      "loss": 2.2077,
+      "step": 800
+    },
+    {
+      "epoch": 0.28415107659724675,
+      "grad_norm": 0.03526224130536068,
+      "learning_rate": 0.0002882803476174157,
+      "loss": 2.1466,
+      "step": 805
+    },
+    {
+      "epoch": 0.2859159901164843,
+      "grad_norm": 0.036605813414546357,
+      "learning_rate": 0.0002881157602731316,
+      "loss": 2.1529,
+      "step": 810
+    },
+    {
+      "epoch": 0.28768090363572185,
+      "grad_norm": 0.03589663428281995,
+      "learning_rate": 0.00028795007291969104,
+      "loss": 2.1245,
+      "step": 815
+    },
+    {
+      "epoch": 0.28944581715495943,
+      "grad_norm": 0.036121970182539424,
+      "learning_rate": 0.0002877832868766945,
+      "loss": 2.2219,
+      "step": 820
+    },
+    {
+      "epoch": 0.29121073067419695,
+      "grad_norm": 0.03437654036658201,
+      "learning_rate": 0.00028761540347249263,
+      "loss": 2.012,
+      "step": 825
+    },
+    {
+      "epoch": 0.2929756441934345,
+      "grad_norm": 0.032381897241935326,
+      "learning_rate": 0.0002874464240441761,
+      "loss": 2.2461,
+      "step": 830
+    },
+    {
+      "epoch": 0.2947405577126721,
+      "grad_norm": 0.031558627469942674,
+      "learning_rate": 0.0002872763499375646,
+      "loss": 2.086,
+      "step": 835
+    },
+    {
+      "epoch": 0.2965054712319096,
+      "grad_norm": 0.03283164503679855,
+      "learning_rate": 0.0002871051825071964,
+      "loss": 2.1209,
+      "step": 840
+    },
+    {
+      "epoch": 0.2982703847511472,
+      "grad_norm": 0.03264520962815425,
+      "learning_rate": 0.0002869329231163172,
+      "loss": 2.0898,
+      "step": 845
+    },
+    {
+      "epoch": 0.3000352982703848,
+      "grad_norm": 0.03136286807195927,
+      "learning_rate": 0.0002867595731368699,
+      "loss": 2.0256,
+      "step": 850
+    },
+    {
+      "epoch": 0.3018002117896223,
+      "grad_norm": 0.030906771854311548,
+      "learning_rate": 0.0002865851339494831,
+      "loss": 2.1845,
+      "step": 855
+    },
+    {
+      "epoch": 0.3035651253088599,
+      "grad_norm": 0.03280850863917648,
+      "learning_rate": 0.00028640960694346,
+      "loss": 2.0018,
+      "step": 860
+    },
+    {
+      "epoch": 0.30533003882809745,
+      "grad_norm": 0.033589188584267925,
+      "learning_rate": 0.0002862329935167681,
+      "loss": 2.3826,
+      "step": 865
+    },
+    {
+      "epoch": 0.30709495234733497,
+      "grad_norm": 0.032363586423996546,
+      "learning_rate": 0.0002860552950760272,
+      "loss": 2.1097,
+      "step": 870
+    },
+    {
+      "epoch": 0.30885986586657255,
+      "grad_norm": 0.0317130458760469,
+      "learning_rate": 0.000285876513036499,
+      "loss": 2.106,
+      "step": 875
+    },
+    {
+      "epoch": 0.31062477938581007,
+      "grad_norm": 0.03192850313391572,
+      "learning_rate": 0.00028569664882207495,
+      "loss": 2.1243,
+      "step": 880
+    },
+    {
+      "epoch": 0.31238969290504764,
+      "grad_norm": 0.030787316418390272,
+      "learning_rate": 0.0002855157038652656,
+      "loss": 2.0857,
+      "step": 885
+    },
+    {
+      "epoch": 0.3141546064242852,
+      "grad_norm": 0.031589314818130196,
+      "learning_rate": 0.00028533367960718913,
+      "loss": 2.1684,
+      "step": 890
+    },
+    {
+      "epoch": 0.31591951994352274,
+      "grad_norm": 0.03164938589075929,
+      "learning_rate": 0.0002851505774975594,
+      "loss": 2.0861,
+      "step": 895
+    },
+    {
+      "epoch": 0.3176844334627603,
+      "grad_norm": 0.03496517145140474,
+      "learning_rate": 0.00028496639899467486,
+      "loss": 2.1541,
+      "step": 900
+    },
+    {
+      "epoch": 0.3194493469819979,
+      "grad_norm": 0.03669965554282724,
+      "learning_rate": 0.0002847811455654068,
+      "loss": 2.0715,
+      "step": 905
+    },
+    {
+      "epoch": 0.3212142605012354,
+      "grad_norm": 0.036077066352649886,
+      "learning_rate": 0.0002845948186851876,
+      "loss": 2.2075,
+      "step": 910
+    },
+    {
+      "epoch": 0.322979174020473,
+      "grad_norm": 0.032675231869251295,
+      "learning_rate": 0.0002844074198379991,
+      "loss": 2.0532,
+      "step": 915
+    },
+    {
+      "epoch": 0.32474408753971057,
+      "grad_norm": 0.03498146054855494,
+      "learning_rate": 0.00028421895051636064,
+      "loss": 2.212,
+      "step": 920
+    },
+    {
+      "epoch": 0.3265090010589481,
+      "grad_norm": 0.0344843689407788,
+      "learning_rate": 0.0002840294122213173,
+      "loss": 2.2024,
+      "step": 925
+    },
+    {
+      "epoch": 0.32827391457818567,
+      "grad_norm": 0.03237165573136892,
+      "learning_rate": 0.0002838388064624278,
+      "loss": 2.215,
+      "step": 930
+    },
+    {
+      "epoch": 0.33003882809742324,
+      "grad_norm": 0.03445765046406074,
+      "learning_rate": 0.0002836471347577527,
+      "loss": 2.1015,
+      "step": 935
+    },
+    {
+      "epoch": 0.33180374161666076,
+      "grad_norm": 0.033095612728491826,
+      "learning_rate": 0.00028345439863384207,
+      "loss": 2.0577,
+      "step": 940
+    },
+    {
+      "epoch": 0.33356865513589834,
+      "grad_norm": 0.03363037214322298,
+      "learning_rate": 0.0002832605996257236,
+      "loss": 2.1047,
+      "step": 945
+    },
+    {
+      "epoch": 0.3353335686551359,
+      "grad_norm": 0.031353729957200864,
+      "learning_rate": 0.00028306573927689007,
+      "loss": 2.0823,
+      "step": 950
+    },
+    {
+      "epoch": 0.33709848217437344,
+      "grad_norm": 0.033728463876960685,
+      "learning_rate": 0.0002828698191392871,
+      "loss": 2.2383,
+      "step": 955
+    },
+    {
+      "epoch": 0.338863395693611,
+      "grad_norm": 0.031957978011313046,
+      "learning_rate": 0.0002826728407733012,
+      "loss": 2.1302,
+      "step": 960
+    },
+    {
+      "epoch": 0.3406283092128486,
+      "grad_norm": 0.03212775871087497,
+      "learning_rate": 0.00028247480574774674,
+      "loss": 2.2524,
+      "step": 965
+    },
+    {
+      "epoch": 0.3423932227320861,
+      "grad_norm": 0.034989168862386295,
+      "learning_rate": 0.0002822757156398539,
+      "loss": 1.9753,
+      "step": 970
+    },
+    {
+      "epoch": 0.3441581362513237,
+      "grad_norm": 0.03385051882383798,
+      "learning_rate": 0.00028207557203525604,
+      "loss": 1.9766,
+      "step": 975
+    },
+    {
+      "epoch": 0.34592304977056126,
+      "grad_norm": 0.03273252857741976,
+      "learning_rate": 0.0002818743765279767,
+      "loss": 2.0491,
+      "step": 980
+    },
+    {
+      "epoch": 0.3476879632897988,
+      "grad_norm": 0.030850980074499237,
+      "learning_rate": 0.0002816721307204175,
+      "loss": 2.0357,
+      "step": 985
+    },
+    {
+      "epoch": 0.34945287680903636,
+      "grad_norm": 0.031865527362074424,
+      "learning_rate": 0.0002814688362233449,
+      "loss": 2.0826,
+      "step": 990
+    },
+    {
+      "epoch": 0.35121779032827394,
+      "grad_norm": 0.031495060621332706,
+      "learning_rate": 0.0002812644946558776,
+      "loss": 2.1531,
+      "step": 995
+    },
+    {
+      "epoch": 0.35298270384751146,
+      "grad_norm": 0.03477014326283163,
+      "learning_rate": 0.0002810591076454735,
+      "loss": 2.1923,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35474761736674904,
+      "grad_norm": 0.036670064592913755,
+      "learning_rate": 0.00028085267682791695,
+      "loss": 2.0221,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3565125308859866,
+      "grad_norm": 0.033691898464027425,
+      "learning_rate": 0.00028064520384730557,
+      "loss": 1.9867,
+      "step": 1010
+    },
+    {
+      "epoch": 0.35827744440522413,
+      "grad_norm": 0.035095823707589985,
+      "learning_rate": 0.00028043669035603716,
+      "loss": 2.1052,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3600423579244617,
+      "grad_norm": 0.03512742369207502,
+      "learning_rate": 0.00028022713801479656,
+      "loss": 2.0658,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3618072714436993,
+      "grad_norm": 0.0345518953155344,
+      "learning_rate": 0.0002800165484925424,
+      "loss": 2.0377,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3635721849629368,
+      "grad_norm": 0.03586824307019526,
+      "learning_rate": 0.00027980492346649385,
+      "loss": 2.0155,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3653370984821744,
+      "grad_norm": 0.03237762834207819,
+      "learning_rate": 0.0002795922646221173,
+      "loss": 2.1023,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3671020120014119,
+      "grad_norm": 0.03297821626751376,
+      "learning_rate": 0.00027937857365311285,
+      "loss": 2.0894,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3688669255206495,
+      "grad_norm": 0.03328373844207128,
+      "learning_rate": 0.0002791638522614008,
+      "loss": 2.1916,
+      "step": 1045
+    },
+    {
+      "epoch": 0.37063183903988706,
+      "grad_norm": 0.032890869295347853,
+      "learning_rate": 0.00027894810215710824,
+      "loss": 2.0841,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3723967525591246,
+      "grad_norm": 0.03210559071060243,
+      "learning_rate": 0.00027873132505855527,
+      "loss": 2.0588,
+      "step": 1055
+    },
+    {
+      "epoch": 0.37416166607836215,
+      "grad_norm": 0.033366455037701755,
+      "learning_rate": 0.0002785135226922414,
+      "loss": 2.0739,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37592657959759973,
+      "grad_norm": 0.03374599112721655,
+      "learning_rate": 0.00027829469679283177,
+      "loss": 2.1457,
+      "step": 1065
+    },
+    {
+      "epoch": 0.37769149311683725,
+      "grad_norm": 0.03285848677061954,
+      "learning_rate": 0.0002780748491031435,
+      "loss": 2.0158,
+      "step": 1070
+    },
+    {
+      "epoch": 0.37945640663607483,
+      "grad_norm": 0.034298649570843436,
+      "learning_rate": 0.0002778539813741313,
+      "loss": 2.0311,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3812213201553124,
+      "grad_norm": 0.03493328500218814,
+      "learning_rate": 0.0002776320953648743,
+      "loss": 2.1531,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3829862336745499,
+      "grad_norm": 0.03418377528359241,
+      "learning_rate": 0.0002774091928425613,
+      "loss": 2.1248,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3847511471937875,
+      "grad_norm": 0.03190825476068291,
+      "learning_rate": 0.0002771852755824772,
+      "loss": 2.2637,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3865160607130251,
+      "grad_norm": 0.03475039484489621,
+      "learning_rate": 0.00027696034536798863,
+      "loss": 2.0331,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3882809742322626,
+      "grad_norm": 0.03722899073079216,
+      "learning_rate": 0.0002767344039905297,
+      "loss": 2.0959,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3900458877515002,
+      "grad_norm": 0.04206335956043232,
+      "learning_rate": 0.000276507453249588,
+      "loss": 2.1869,
+      "step": 1105
+    },
+    {
+      "epoch": 0.39181080127073775,
+      "grad_norm": 0.03741766702570404,
+      "learning_rate": 0.00027627949495269004,
+      "loss": 2.1399,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3935757147899753,
+      "grad_norm": 0.035987744465726425,
+      "learning_rate": 0.00027605053091538665,
+      "loss": 2.085,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39534062830921285,
+      "grad_norm": 0.036512922759248466,
+      "learning_rate": 0.0002758205629612393,
+      "loss": 2.0886,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3971055418284504,
+      "grad_norm": 0.03632362257347427,
+      "learning_rate": 0.0002755895929218045,
+      "loss": 2.0927,
+      "step": 1125
+    },
+    {
+      "epoch": 0.39887045534768795,
+      "grad_norm": 0.03423191738091953,
+      "learning_rate": 0.0002753576226366202,
+      "loss": 2.1753,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4006353688669255,
+      "grad_norm": 0.03615517054444847,
+      "learning_rate": 0.0002751246539531905,
+      "loss": 2.0145,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4024002823861631,
+      "grad_norm": 0.03480923996115515,
+      "learning_rate": 0.00027489068872697104,
+      "loss": 2.2001,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4041651959054006,
+      "grad_norm": 0.040212739205729986,
+      "learning_rate": 0.0002746557288213546,
+      "loss": 2.1179,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4059301094246382,
+      "grad_norm": 0.035285244575101765,
+      "learning_rate": 0.0002744197761076558,
+      "loss": 2.0346,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4076950229438758,
+      "grad_norm": 0.03479285940674813,
+      "learning_rate": 0.0002741828324650964,
+      "loss": 2.0631,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4094599364631133,
+      "grad_norm": 0.03478212799072865,
+      "learning_rate": 0.00027394489978079043,
+      "loss": 2.0288,
+      "step": 1160
+    },
+    {
+      "epoch": 0.41122484998235087,
+      "grad_norm": 0.03442721873390001,
+      "learning_rate": 0.00027370597994972885,
+      "loss": 2.1027,
+      "step": 1165
+    },
+    {
+      "epoch": 0.41298976350158845,
+      "grad_norm": 0.035013108215072315,
+      "learning_rate": 0.00027346607487476485,
+      "loss": 2.122,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41475467702082597,
+      "grad_norm": 0.03685488507083505,
+      "learning_rate": 0.0002732251864665983,
+      "loss": 2.0809,
+      "step": 1175
+    },
+    {
+      "epoch": 0.41651959054006354,
+      "grad_norm": 0.0374851905968305,
+      "learning_rate": 0.0002729833166437609,
+      "loss": 2.0885,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4182845040593011,
+      "grad_norm": 0.03513428701010772,
+      "learning_rate": 0.0002727404673326005,
+      "loss": 2.1349,
+      "step": 1185
+    },
+    {
+      "epoch": 0.42004941757853864,
+      "grad_norm": 0.03593527810394835,
+      "learning_rate": 0.00027249664046726646,
+      "loss": 1.9886,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4218143310977762,
+      "grad_norm": 0.036875791448744456,
+      "learning_rate": 0.00027225183798969315,
+      "loss": 2.0894,
+      "step": 1195
+    },
+    {
+      "epoch": 0.42357924461701374,
+      "grad_norm": 0.03262167438974139,
+      "learning_rate": 0.0002720060618495856,
+      "loss": 2.0764,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4253441581362513,
+      "grad_norm": 0.033675860134706456,
+      "learning_rate": 0.0002717593140044032,
+      "loss": 2.233,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4271090716554889,
+      "grad_norm": 0.03387281917057669,
+      "learning_rate": 0.00027151159641934457,
+      "loss": 2.25,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4288739851747264,
+      "grad_norm": 0.03250899436211034,
+      "learning_rate": 0.00027126291106733153,
+      "loss": 2.0767,
+      "step": 1215
+    },
+    {
+      "epoch": 0.430638898693964,
+      "grad_norm": 0.03233745826410541,
+      "learning_rate": 0.0002710132599289937,
+      "loss": 2.0126,
+      "step": 1220
+    },
+    {
+      "epoch": 0.43240381221320157,
+      "grad_norm": 0.03180368800275238,
+      "learning_rate": 0.00027076264499265273,
+      "loss": 2.0627,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4341687257324391,
+      "grad_norm": 0.0322734260973177,
+      "learning_rate": 0.0002705110682543061,
+      "loss": 2.0426,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43593363925167666,
+      "grad_norm": 0.03102410121289385,
+      "learning_rate": 0.0002702585317176115,
+      "loss": 2.1374,
+      "step": 1235
+    },
+    {
+      "epoch": 0.43769855277091424,
+      "grad_norm": 0.03200479656396118,
+      "learning_rate": 0.0002700050373938711,
+      "loss": 2.0922,
+      "step": 1240
+    },
+    {
+      "epoch": 0.43946346629015176,
+      "grad_norm": 0.031177007128579348,
+      "learning_rate": 0.00026975058730201493,
+      "loss": 2.0486,
+      "step": 1245
+    },
+    {
+      "epoch": 0.44122837980938934,
+      "grad_norm": 0.03139238732614584,
+      "learning_rate": 0.00026949518346858534,
+      "loss": 1.9698,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4429932933286269,
+      "grad_norm": 0.030907168779273565,
+      "learning_rate": 0.0002692388279277207,
+      "loss": 1.9343,
+      "step": 1255
+    },
+    {
+      "epoch": 0.44475820684786443,
+      "grad_norm": 0.03082907795184397,
+      "learning_rate": 0.000268981522721139,
+      "loss": 2.0842,
+      "step": 1260
+    },
+    {
+      "epoch": 0.446523120367102,
+      "grad_norm": 0.03075293960047849,
+      "learning_rate": 0.0002687232698981217,
+      "loss": 2.1045,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4482880338863396,
+      "grad_norm": 0.031408364123282295,
+      "learning_rate": 0.0002684640715154978,
+      "loss": 2.0529,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4500529474055771,
+      "grad_norm": 0.03147517230687401,
+      "learning_rate": 0.0002682039296376267,
+      "loss": 2.0879,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4518178609248147,
+      "grad_norm": 0.0323233734261303,
+      "learning_rate": 0.0002679428463363825,
+      "loss": 2.1399,
+      "step": 1280
+    },
+    {
+      "epoch": 0.45358277444405226,
+      "grad_norm": 0.031263782667529036,
+      "learning_rate": 0.00026768082369113696,
+      "loss": 2.0434,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4553476879632898,
+      "grad_norm": 0.031713538307666836,
+      "learning_rate": 0.0002674178637887432,
+      "loss": 2.1592,
+      "step": 1290
+    },
+    {
+      "epoch": 0.45711260148252736,
+      "grad_norm": 0.03140672686733592,
+      "learning_rate": 0.00026715396872351916,
+      "loss": 1.9876,
+      "step": 1295
+    },
+    {
+      "epoch": 0.45887751500176494,
+      "grad_norm": 0.034399309547844266,
+      "learning_rate": 0.0002668891405972306,
+      "loss": 2.1152,
+      "step": 1300
+    },
+    {
+      "epoch": 0.46064242852100246,
+      "grad_norm": 0.03920278097306655,
+      "learning_rate": 0.0002666233815190747,
+      "loss": 2.0441,
+      "step": 1305
+    },
+    {
+      "epoch": 0.46240734204024003,
+      "grad_norm": 0.03679690406460357,
+      "learning_rate": 0.00026635669360566296,
+      "loss": 2.0483,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4641722555594776,
+      "grad_norm": 0.03218013320330528,
+      "learning_rate": 0.0002660890789810047,
+      "loss": 2.1069,
+      "step": 1315
+    },
+    {
+      "epoch": 0.46593716907871513,
+      "grad_norm": 0.03305246974531592,
+      "learning_rate": 0.0002658205397764897,
+      "loss": 1.974,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4677020825979527,
+      "grad_norm": 0.03182030180023369,
+      "learning_rate": 0.0002655510781308718,
+      "loss": 2.125,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4694669961171903,
+      "grad_norm": 0.030945200388846374,
+      "learning_rate": 0.0002652806961902511,
+      "loss": 2.0957,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4712319096364278,
+      "grad_norm": 0.030658171577616956,
+      "learning_rate": 0.0002650093961080577,
+      "loss": 1.8768,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4729968231556654,
+      "grad_norm": 0.030998377047293,
+      "learning_rate": 0.0002647371800450337,
+      "loss": 2.2316,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47476173667490296,
+      "grad_norm": 0.03161221396323837,
+      "learning_rate": 0.00026446405016921704,
+      "loss": 2.0952,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4765266501941405,
+      "grad_norm": 0.03004654129424578,
+      "learning_rate": 0.000264190008655923,
+      "loss": 1.9688,
+      "step": 1350
+    },
+    {
+      "epoch": 0.47829156371337805,
+      "grad_norm": 0.03276315991633029,
+      "learning_rate": 0.00026391505768772806,
+      "loss": 2.1006,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4800564772326156,
+      "grad_norm": 0.03230843220270764,
+      "learning_rate": 0.0002636391994544515,
+      "loss": 2.1486,
+      "step": 1360
+    },
+    {
+      "epoch": 0.48182139075185315,
+      "grad_norm": 0.03077879401754805,
+      "learning_rate": 0.00026336243615313873,
+      "loss": 1.9104,
+      "step": 1365
+    },
+    {
+      "epoch": 0.48358630427109073,
+      "grad_norm": 0.032324649121534106,
+      "learning_rate": 0.00026308476998804334,
+      "loss": 2.0982,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48535121779032825,
+      "grad_norm": 0.03234316342989001,
+      "learning_rate": 0.0002628062031706097,
+      "loss": 2.0679,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4871161313095658,
+      "grad_norm": 0.031221867182909337,
+      "learning_rate": 0.0002625267379194554,
+      "loss": 2.0957,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4888810448288034,
+      "grad_norm": 0.03268103407525422,
+      "learning_rate": 0.00026224637646035345,
+      "loss": 2.0209,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4906459583480409,
+      "grad_norm": 0.031993982186490015,
+      "learning_rate": 0.00026196512102621473,
+      "loss": 2.1004,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4924108718672785,
+      "grad_norm": 0.03077913399535889,
+      "learning_rate": 0.00026168297385706986,
+      "loss": 1.867,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4941757853865161,
+      "grad_norm": 0.034698623632773436,
+      "learning_rate": 0.0002613999372000519,
+      "loss": 2.1669,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4959406989057536,
+      "grad_norm": 0.04476846282510472,
+      "learning_rate": 0.00026111601330937784,
+      "loss": 2.0645,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4977056124249912,
+      "grad_norm": 0.035739813315660666,
+      "learning_rate": 0.0002608312044463312,
+      "loss": 2.0385,
+      "step": 1410
+    },
+    {
+      "epoch": 0.49947052594422875,
+      "grad_norm": 0.03573054273851387,
+      "learning_rate": 0.0002605455128792437,
+      "loss": 2.1873,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5012354394634663,
+      "grad_norm": 0.03597947385937176,
+      "learning_rate": 0.0002602589408834772,
+      "loss": 2.0318,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5030003529827038,
+      "grad_norm": 0.0336954959804502,
+      "learning_rate": 0.0002599714907414057,
+      "loss": 2.0639,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5047652665019414,
+      "grad_norm": 0.0336673399376456,
+      "learning_rate": 0.0002596831647423971,
+      "loss": 1.9877,
+      "step": 1430
+    },
+    {
+      "epoch": 0.506530180021179,
+      "grad_norm": 0.03204639666250037,
+      "learning_rate": 0.000259393965182795,
+      "loss": 2.1717,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5082950935404165,
+      "grad_norm": 0.03477364359496535,
+      "learning_rate": 0.0002591038943659003,
+      "loss": 2.1521,
+      "step": 1440
+    },
+    {
+      "epoch": 0.510060007059654,
+      "grad_norm": 0.03344648625259787,
+      "learning_rate": 0.00025881295460195307,
+      "loss": 2.0867,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5118249205788916,
+      "grad_norm": 0.03356190309860851,
+      "learning_rate": 0.0002585211482081139,
+      "loss": 2.2151,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5135898340981292,
+      "grad_norm": 0.035250829956586675,
+      "learning_rate": 0.00025822847750844576,
+      "loss": 2.1646,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5153547476173668,
+      "grad_norm": 0.033468807619869284,
+      "learning_rate": 0.00025793494483389496,
+      "loss": 2.0988,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5171196611366043,
+      "grad_norm": 0.033283971857736974,
+      "learning_rate": 0.0002576405525222732,
+      "loss": 2.0323,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5188845746558418,
+      "grad_norm": 0.034389937250349646,
+      "learning_rate": 0.0002573453029182387,
+      "loss": 2.2742,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5206494881750794,
+      "grad_norm": 0.032300315453966,
+      "learning_rate": 0.0002570491983732773,
+      "loss": 2.0879,
+      "step": 1475
+    },
+    {
+      "epoch": 0.522414401694317,
+      "grad_norm": 0.03539641068412446,
+      "learning_rate": 0.000256752241245684,
+      "loss": 2.1998,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5241793152135545,
+      "grad_norm": 0.03461192321620284,
+      "learning_rate": 0.0002564544339005442,
+      "loss": 2.2052,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5259442287327921,
+      "grad_norm": 0.03402386654170651,
+      "learning_rate": 0.0002561557787097147,
+      "loss": 2.2198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5277091422520297,
+      "grad_norm": 0.03622145830784721,
+      "learning_rate": 0.0002558562780518049,
+      "loss": 2.0614,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5294740557712672,
+      "grad_norm": 0.0371718258385898,
+      "learning_rate": 0.0002555559343121577,
+      "loss": 2.2121,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5312389692905047,
+      "grad_norm": 0.04180893521708982,
+      "learning_rate": 0.00025525474988283097,
+      "loss": 2.0787,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5330038828097423,
+      "grad_norm": 0.04036241283975166,
+      "learning_rate": 0.00025495272716257784,
+      "loss": 1.9809,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5347687963289799,
+      "grad_norm": 0.04209950499131354,
+      "learning_rate": 0.00025464986855682815,
+      "loss": 2.0939,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5365337098482175,
+      "grad_norm": 0.03829236470820007,
+      "learning_rate": 0.00025434617647766887,
+      "loss": 1.9998,
+      "step": 1520
+    },
+    {
+      "epoch": 0.538298623367455,
+      "grad_norm": 0.03794497094740352,
+      "learning_rate": 0.00025404165334382516,
+      "loss": 1.9074,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5400635368866925,
+      "grad_norm": 0.03760553300906864,
+      "learning_rate": 0.0002537363015806412,
+      "loss": 2.0548,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5418284504059301,
+      "grad_norm": 0.03499570362354962,
+      "learning_rate": 0.00025343012362006044,
+      "loss": 1.9747,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5435933639251677,
+      "grad_norm": 0.037916636045701606,
+      "learning_rate": 0.0002531231219006067,
+      "loss": 2.0348,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5453582774444052,
+      "grad_norm": 0.03739455905733242,
+      "learning_rate": 0.0002528152988673644,
+      "loss": 2.1421,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5471231909636428,
+      "grad_norm": 0.03663858296682295,
+      "learning_rate": 0.00025250665697195936,
+      "loss": 2.0219,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5488881044828804,
+      "grad_norm": 0.035717893252210364,
+      "learning_rate": 0.000252197198672539,
+      "loss": 2.1173,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5506530180021179,
+      "grad_norm": 0.03573303516125351,
+      "learning_rate": 0.00025188692643375306,
+      "loss": 2.1262,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5524179315213554,
+      "grad_norm": 0.033283116033722755,
+      "learning_rate": 0.0002515758427267337,
+      "loss": 2.1402,
+      "step": 1565
+    },
+    {
+      "epoch": 0.554182845040593,
+      "grad_norm": 0.03670333464658142,
+      "learning_rate": 0.00025126395002907597,
+      "loss": 2.0886,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5559477585598306,
+      "grad_norm": 0.03696742889718139,
+      "learning_rate": 0.0002509512508248182,
+      "loss": 2.1521,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5577126720790682,
+      "grad_norm": 0.035448258104787154,
+      "learning_rate": 0.0002506377476044217,
+      "loss": 2.0807,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5594775855983057,
+      "grad_norm": 0.035479788995991794,
+      "learning_rate": 0.0002503234428647517,
+      "loss": 2.0106,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5612424991175432,
+      "grad_norm": 0.03797297560579901,
+      "learning_rate": 0.0002500083391090567,
+      "loss": 2.0087,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5630074126367808,
+      "grad_norm": 0.03608249672999952,
+      "learning_rate": 0.00024969243884694916,
+      "loss": 1.9312,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5647723261560184,
+      "grad_norm": 0.03681238195116478,
+      "learning_rate": 0.00024937574459438515,
+      "loss": 2.2043,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5665372396752559,
+      "grad_norm": 0.045285409749272164,
+      "learning_rate": 0.00024905825887364423,
+      "loss": 2.2137,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5683021531944935,
+      "grad_norm": 0.03884372126912092,
+      "learning_rate": 0.00024873998421330975,
+      "loss": 1.9218,
+      "step": 1610
+    },
+    {
+      "epoch": 0.570067066713731,
+      "grad_norm": 0.03463932955320792,
+      "learning_rate": 0.0002484209231482484,
+      "loss": 2.2093,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5718319802329686,
+      "grad_norm": 0.03448296034126532,
+      "learning_rate": 0.00024810107821959007,
+      "loss": 2.0617,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5735968937522061,
+      "grad_norm": 0.03664801323301665,
+      "learning_rate": 0.0002477804519747077,
+      "loss": 2.1588,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5753618072714437,
+      "grad_norm": 0.036742037704079834,
+      "learning_rate": 0.000247459046967197,
+      "loss": 2.0023,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5771267207906813,
+      "grad_norm": 0.03437854224178402,
+      "learning_rate": 0.000247136865756856,
+      "loss": 2.0192,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5788916343099189,
+      "grad_norm": 0.035954464814617046,
+      "learning_rate": 0.0002468139109096646,
+      "loss": 2.1808,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5806565478291563,
+      "grad_norm": 0.036463863652557084,
+      "learning_rate": 0.00024649018499776454,
+      "loss": 1.9863,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5824214613483939,
+      "grad_norm": 0.034047333314064704,
+      "learning_rate": 0.0002461656905994384,
+      "loss": 2.1675,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5841863748676315,
+      "grad_norm": 0.034802884332088065,
+      "learning_rate": 0.0002458404302990893,
+      "loss": 2.1104,
+      "step": 1655
+    },
+    {
+      "epoch": 0.585951288386869,
+      "grad_norm": 0.03484770231742836,
+      "learning_rate": 0.0002455144066872204,
+      "loss": 1.9928,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5877162019061066,
+      "grad_norm": 0.03485844317241166,
+      "learning_rate": 0.0002451876223604142,
+      "loss": 2.0685,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5894811154253442,
+      "grad_norm": 0.03517892166758923,
+      "learning_rate": 0.0002448600799213117,
+      "loss": 2.339,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5912460289445817,
+      "grad_norm": 0.03391289871055977,
+      "learning_rate": 0.0002445317819785919,
+      "loss": 2.251,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5930109424638192,
+      "grad_norm": 0.03509293856751318,
+      "learning_rate": 0.000244202731146951,
+      "loss": 2.0822,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5947758559830568,
+      "grad_norm": 0.034195274147096895,
+      "learning_rate": 0.00024387293004708135,
+      "loss": 2.2187,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5965407695022944,
+      "grad_norm": 0.03587758206809719,
+      "learning_rate": 0.00024354238130565087,
+      "loss": 2.0735,
+      "step": 1690
+    },
+    {
+      "epoch": 0.598305683021532,
+      "grad_norm": 0.034772307118060036,
+      "learning_rate": 0.00024321108755528186,
+      "loss": 1.999,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6000705965407696,
+      "grad_norm": 0.03473687411277052,
+      "learning_rate": 0.00024287905143453037,
+      "loss": 1.9622,
+      "step": 1700
+    },
+    {
+      "epoch": 0.601835510060007,
+      "grad_norm": 0.0441149655652353,
+      "learning_rate": 0.00024254627558786482,
+      "loss": 1.9453,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6036004235792446,
+      "grad_norm": 0.04178910688626766,
+      "learning_rate": 0.00024221276266564513,
+      "loss": 2.0796,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6053653370984822,
+      "grad_norm": 0.036570587487241873,
+      "learning_rate": 0.0002418785153241016,
+      "loss": 2.1459,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6071302506177197,
+      "grad_norm": 0.03684621099065397,
+      "learning_rate": 0.0002415435362253138,
+      "loss": 2.0044,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6088951641369573,
+      "grad_norm": 0.03771251891950647,
+      "learning_rate": 0.0002412078280371892,
+      "loss": 2.0982,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6106600776561949,
+      "grad_norm": 0.037107840799768615,
+      "learning_rate": 0.00024087139343344213,
+      "loss": 2.1232,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6124249911754324,
+      "grad_norm": 0.037674437327224926,
+      "learning_rate": 0.0002405342350935723,
+      "loss": 2.0146,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6141899046946699,
+      "grad_norm": 0.03692512652029206,
+      "learning_rate": 0.0002401963557028436,
+      "loss": 2.1244,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6159548182139075,
+      "grad_norm": 0.03689340020991997,
+      "learning_rate": 0.00023985775795226256,
+      "loss": 2.0166,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6177197317331451,
+      "grad_norm": 0.03582171552708508,
+      "learning_rate": 0.00023951844453855723,
+      "loss": 2.0037,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6194846452523827,
+      "grad_norm": 0.03591063581394303,
+      "learning_rate": 0.00023917841816415523,
+      "loss": 2.0314,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6212495587716201,
+      "grad_norm": 0.037387473204399134,
+      "learning_rate": 0.00023883768153716274,
+      "loss": 2.2235,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6230144722908577,
+      "grad_norm": 0.03463512713478883,
+      "learning_rate": 0.00023849623737134238,
+      "loss": 2.1317,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6247793858100953,
+      "grad_norm": 0.0335651964435421,
+      "learning_rate": 0.00023815408838609217,
+      "loss": 2.1241,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6265442993293329,
+      "grad_norm": 0.03501784016654201,
+      "learning_rate": 0.00023781123730642342,
+      "loss": 1.9438,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6283092128485704,
+      "grad_norm": 0.03382438076413039,
+      "learning_rate": 0.00023746768686293927,
+      "loss": 1.9621,
+      "step": 1780
+    },
+    {
+      "epoch": 0.630074126367808,
+      "grad_norm": 0.03593096091497881,
+      "learning_rate": 0.00023712343979181288,
+      "loss": 2.1609,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6318390398870455,
+      "grad_norm": 0.036296264367783077,
+      "learning_rate": 0.00023677849883476556,
+      "loss": 2.2099,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6336039534062831,
+      "grad_norm": 0.03660540857678081,
+      "learning_rate": 0.00023643286673904513,
+      "loss": 2.0158,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6353688669255206,
+      "grad_norm": 0.034431328952848395,
+      "learning_rate": 0.00023608654625740377,
+      "loss": 2.0521,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6371337804447582,
+      "grad_norm": 0.036458655720471446,
+      "learning_rate": 0.0002357395401480764,
+      "loss": 2.1523,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6388986939639958,
+      "grad_norm": 0.036077257402277484,
+      "learning_rate": 0.00023539185117475836,
+      "loss": 2.337,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6406636074832334,
+      "grad_norm": 0.03801124880836332,
+      "learning_rate": 0.00023504348210658376,
+      "loss": 1.8912,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6424285210024708,
+      "grad_norm": 0.03455758965007696,
+      "learning_rate": 0.0002346944357181032,
+      "loss": 1.9723,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6441934345217084,
+      "grad_norm": 0.03661742218081978,
+      "learning_rate": 0.00023434471478926165,
+      "loss": 2.0922,
+      "step": 1825
+    },
+    {
+      "epoch": 0.645958348040946,
+      "grad_norm": 0.03435967613102213,
+      "learning_rate": 0.0002339943221053766,
+      "loss": 1.9364,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6477232615601836,
+      "grad_norm": 0.035667133768118635,
+      "learning_rate": 0.0002336432604571154,
+      "loss": 1.9636,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6494881750794211,
+      "grad_norm": 0.03486785641377008,
+      "learning_rate": 0.00023329153264047348,
+      "loss": 2.0558,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6512530885986587,
+      "grad_norm": 0.03668403831416794,
+      "learning_rate": 0.00023293914145675184,
+      "loss": 2.0943,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6530180021178962,
+      "grad_norm": 0.03476943300951533,
+      "learning_rate": 0.00023258608971253486,
+      "loss": 1.9965,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6547829156371338,
+      "grad_norm": 0.03443774753477751,
+      "learning_rate": 0.00023223238021966776,
+      "loss": 2.0822,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6565478291563713,
+      "grad_norm": 0.03665607563086681,
+      "learning_rate": 0.00023187801579523446,
+      "loss": 2.0828,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6583127426756089,
+      "grad_norm": 0.033787887951323356,
+      "learning_rate": 0.000231522999261535,
+      "loss": 1.8954,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6600776561948465,
+      "grad_norm": 0.03530902707709943,
+      "learning_rate": 0.00023116733344606297,
+      "loss": 2.1379,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6618425697140841,
+      "grad_norm": 0.038298495977867164,
+      "learning_rate": 0.00023081102118148325,
+      "loss": 1.9456,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6636074832333215,
+      "grad_norm": 0.03611598379196364,
+      "learning_rate": 0.0002304540653056092,
+      "loss": 1.9262,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6653723967525591,
+      "grad_norm": 0.03586878959495951,
+      "learning_rate": 0.00023009646866138016,
+      "loss": 2.0516,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6671373102717967,
+      "grad_norm": 0.034677572851205714,
+      "learning_rate": 0.0002297382340968389,
+      "loss": 2.0916,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6689022237910343,
+      "grad_norm": 0.036032611808584275,
+      "learning_rate": 0.00022937936446510866,
+      "loss": 1.9777,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6706671373102718,
+      "grad_norm": 0.042360784395910844,
+      "learning_rate": 0.00022901986262437087,
+      "loss": 2.0911,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6724320508295094,
+      "grad_norm": 0.04159007223098105,
+      "learning_rate": 0.00022865973143784183,
+      "loss": 2.1849,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6741969643487469,
+      "grad_norm": 0.04009396168279417,
+      "learning_rate": 0.00022829897377375042,
+      "loss": 1.9319,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6759618778679845,
+      "grad_norm": 0.042567972668609885,
+      "learning_rate": 0.00022793759250531502,
+      "loss": 2.1074,
+      "step": 1915
+    },
+    {
+      "epoch": 0.677726791387222,
+      "grad_norm": 0.0425895160699235,
+      "learning_rate": 0.00022757559051072052,
+      "loss": 1.9762,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6794917049064596,
+      "grad_norm": 0.03948452206527659,
+      "learning_rate": 0.00022721297067309558,
+      "loss": 1.9072,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6812566184256972,
+      "grad_norm": 0.03939288726626579,
+      "learning_rate": 0.0002268497358804897,
+      "loss": 1.8813,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6830215319449346,
+      "grad_norm": 0.040208933214084225,
+      "learning_rate": 0.00022648588902585,
+      "loss": 2.0787,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6847864454641722,
+      "grad_norm": 0.04093700935422431,
+      "learning_rate": 0.0002261214330069983,
+      "loss": 1.9761,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6865513589834098,
+      "grad_norm": 0.039488426920537424,
+      "learning_rate": 0.0002257563707266082,
+      "loss": 2.0439,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6883162725026474,
+      "grad_norm": 0.03745905909736412,
+      "learning_rate": 0.00022539070509218178,
+      "loss": 1.9896,
+      "step": 1950
+    },
+    {
+      "epoch": 0.690081186021885,
+      "grad_norm": 0.040472557552296826,
+      "learning_rate": 0.00022502443901602622,
+      "loss": 2.0269,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6918460995411225,
+      "grad_norm": 0.040086135550804675,
+      "learning_rate": 0.00022465757541523105,
+      "loss": 1.961,
+      "step": 1960
+    },
+    {
+      "epoch": 0.69361101306036,
+      "grad_norm": 0.04202801113128649,
+      "learning_rate": 0.00022429011721164489,
+      "loss": 2.1577,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6953759265795976,
+      "grad_norm": 0.04086062088028195,
+      "learning_rate": 0.00022392206733185175,
+      "loss": 2.0868,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6971408400988351,
+      "grad_norm": 0.041900709891816414,
+      "learning_rate": 0.00022355342870714807,
+      "loss": 2.1132,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6989057536180727,
+      "grad_norm": 0.037571214161659414,
+      "learning_rate": 0.00022318420427351932,
+      "loss": 2.0171,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7006706671373103,
+      "grad_norm": 0.04060263617550077,
+      "learning_rate": 0.00022281439697161655,
+      "loss": 2.1087,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7024355806565479,
+      "grad_norm": 0.0383963400418068,
+      "learning_rate": 0.00022244400974673302,
+      "loss": 1.9833,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7042004941757853,
+      "grad_norm": 0.037842473816367274,
+      "learning_rate": 0.00022207304554878077,
+      "loss": 2.0165,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7059654076950229,
+      "grad_norm": 0.03609698670438139,
+      "learning_rate": 0.00022170150733226703,
+      "loss": 2.1203,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7077303212142605,
+      "grad_norm": 0.03577157081311834,
+      "learning_rate": 0.00022132939805627068,
+      "loss": 2.0111,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7094952347334981,
+      "grad_norm": 0.03868543135349071,
+      "learning_rate": 0.00022095672068441895,
+      "loss": 2.0769,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7112601482527356,
+      "grad_norm": 0.034883283436284496,
+      "learning_rate": 0.00022058347818486342,
+      "loss": 1.9393,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7130250617719732,
+      "grad_norm": 0.034465723226893064,
+      "learning_rate": 0.00022020967353025668,
+      "loss": 2.166,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7147899752912107,
+      "grad_norm": 0.03719847154644835,
+      "learning_rate": 0.0002198353096977284,
+      "loss": 2.083,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7165548888104483,
+      "grad_norm": 0.03470780961600924,
+      "learning_rate": 0.00021946038966886203,
+      "loss": 2.2259,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7183198023296858,
+      "grad_norm": 0.035688275285025675,
+      "learning_rate": 0.00021908491642967052,
+      "loss": 2.079,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7200847158489234,
+      "grad_norm": 0.035337313477990455,
+      "learning_rate": 0.00021870889297057294,
+      "loss": 2.1396,
+      "step": 2040
+    },
+    {
+      "epoch": 0.721849629368161,
+      "grad_norm": 0.03752828984704112,
+      "learning_rate": 0.0002183323222863706,
+      "loss": 1.9795,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7236145428873986,
+      "grad_norm": 0.03410015621832168,
+      "learning_rate": 0.00021795520737622295,
+      "loss": 2.0263,
+      "step": 2050
+    },
+    {
+      "epoch": 0.725379456406636,
+      "grad_norm": 0.03551846169959177,
+      "learning_rate": 0.00021757755124362398,
+      "loss": 2.068,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7271443699258736,
+      "grad_norm": 0.035228372159733866,
+      "learning_rate": 0.0002171993568963783,
+      "loss": 2.0887,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7289092834451112,
+      "grad_norm": 0.03487008300055342,
+      "learning_rate": 0.00021682062734657673,
+      "loss": 2.0686,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7306741969643488,
+      "grad_norm": 0.036293036439221084,
+      "learning_rate": 0.00021644136561057304,
+      "loss": 2.0438,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7324391104835863,
+      "grad_norm": 0.03740037018924331,
+      "learning_rate": 0.00021606157470895935,
+      "loss": 2.0746,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7342040240028238,
+      "grad_norm": 0.036022863459193324,
+      "learning_rate": 0.00021568125766654236,
+      "loss": 2.0029,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7359689375220614,
+      "grad_norm": 0.03506624236920992,
+      "learning_rate": 0.00021530041751231904,
+      "loss": 2.0028,
+      "step": 2085
+    },
+    {
+      "epoch": 0.737733851041299,
+      "grad_norm": 0.03524684985350237,
+      "learning_rate": 0.00021491905727945274,
+      "loss": 1.9794,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7394987645605365,
+      "grad_norm": 0.03571714051432255,
+      "learning_rate": 0.0002145371800052488,
+      "loss": 2.0336,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7412636780797741,
+      "grad_norm": 0.03672168161281596,
+      "learning_rate": 0.00021415478873113066,
+      "loss": 1.917,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7430285915990117,
+      "grad_norm": 0.037967852002981854,
+      "learning_rate": 0.00021377188650261537,
+      "loss": 1.9402,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7447935051182492,
+      "grad_norm": 0.03820955824316396,
+      "learning_rate": 0.0002133884763692894,
+      "loss": 2.0277,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7465584186374867,
+      "grad_norm": 0.038723757739905926,
+      "learning_rate": 0.0002130045613847844,
+      "loss": 2.0032,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7483233321567243,
+      "grad_norm": 0.04043199045168578,
+      "learning_rate": 0.00021262014460675285,
+      "loss": 1.9176,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7500882456759619,
+      "grad_norm": 0.037523305034100646,
+      "learning_rate": 0.00021223522909684375,
+      "loss": 1.9901,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7518531591951995,
+      "grad_norm": 0.03514012539236237,
+      "learning_rate": 0.00021184981792067814,
+      "loss": 2.0598,
+      "step": 2130
+    },
+    {
+      "epoch": 0.753618072714437,
+      "grad_norm": 0.03710882914586427,
+      "learning_rate": 0.00021146391414782483,
+      "loss": 1.9927,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7553829862336745,
+      "grad_norm": 0.03813665346750089,
+      "learning_rate": 0.00021107752085177577,
+      "loss": 2.1043,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7571478997529121,
+      "grad_norm": 0.039150624412499356,
+      "learning_rate": 0.00021069064110992172,
+      "loss": 1.9659,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7589128132721497,
+      "grad_norm": 0.03752106406943299,
+      "learning_rate": 0.00021030327800352773,
+      "loss": 1.9538,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7606777267913872,
+      "grad_norm": 0.03805789209082946,
+      "learning_rate": 0.00020991543461770844,
+      "loss": 2.0388,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7624426403106248,
+      "grad_norm": 0.03699737329194902,
+      "learning_rate": 0.00020952711404140372,
+      "loss": 2.0764,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7642075538298624,
+      "grad_norm": 0.03690775474820472,
+      "learning_rate": 0.00020913831936735398,
+      "loss": 1.9851,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7659724673490999,
+      "grad_norm": 0.03834568659341274,
+      "learning_rate": 0.00020874905369207554,
+      "loss": 2.1236,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7677373808683374,
+      "grad_norm": 0.03920184789874272,
+      "learning_rate": 0.00020835932011583586,
+      "loss": 2.0835,
+      "step": 2175
+    },
+    {
+      "epoch": 0.769502294387575,
+      "grad_norm": 0.03665360666533061,
+      "learning_rate": 0.00020796912174262899,
+      "loss": 1.9415,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7712672079068126,
+      "grad_norm": 0.03511969420514865,
+      "learning_rate": 0.00020757846168015093,
+      "loss": 2.0053,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7730321214260502,
+      "grad_norm": 0.03804817661262571,
+      "learning_rate": 0.00020718734303977468,
+      "loss": 1.959,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7747970349452877,
+      "grad_norm": 0.035085437498550955,
+      "learning_rate": 0.00020679576893652547,
+      "loss": 2.0356,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7765619484645252,
+      "grad_norm": 0.0347594960806743,
+      "learning_rate": 0.00020640374248905605,
+      "loss": 1.9538,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7783268619837628,
+      "grad_norm": 0.040667419330315806,
+      "learning_rate": 0.00020601126681962187,
+      "loss": 2.1364,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7800917755030004,
+      "grad_norm": 0.03379498766650798,
+      "learning_rate": 0.00020561834505405618,
+      "loss": 2.0756,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7818566890222379,
+      "grad_norm": 0.03453855024872286,
+      "learning_rate": 0.00020522498032174506,
+      "loss": 2.0517,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7836216025414755,
+      "grad_norm": 0.035065080415837925,
+      "learning_rate": 0.00020483117575560255,
+      "loss": 2.0356,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7853865160607131,
+      "grad_norm": 0.03365493432369374,
+      "learning_rate": 0.0002044369344920457,
+      "loss": 2.0107,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7871514295799505,
+      "grad_norm": 0.03388885656441201,
+      "learning_rate": 0.00020404225967096963,
+      "loss": 1.9988,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7889163430991881,
+      "grad_norm": 0.037128285258714915,
+      "learning_rate": 0.00020364715443572256,
+      "loss": 1.9492,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7906812566184257,
+      "grad_norm": 0.03559719180193601,
+      "learning_rate": 0.0002032516219330805,
+      "loss": 1.9293,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7924461701376633,
+      "grad_norm": 0.03462919459607743,
+      "learning_rate": 0.00020285566531322253,
+      "loss": 2.1581,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7942110836569009,
+      "grad_norm": 0.03348918512022428,
+      "learning_rate": 0.0002024592877297056,
+      "loss": 1.9208,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7959759971761383,
+      "grad_norm": 0.03360398227983138,
+      "learning_rate": 0.00020206249233943921,
+      "loss": 1.8854,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7977409106953759,
+      "grad_norm": 0.0350715526817266,
+      "learning_rate": 0.00020166528230266065,
+      "loss": 2.0812,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7995058242146135,
+      "grad_norm": 0.03440017066419089,
+      "learning_rate": 0.00020126766078290944,
+      "loss": 1.984,
+      "step": 2265
+    },
+    {
+      "epoch": 0.801270737733851,
+      "grad_norm": 0.03526836975981266,
+      "learning_rate": 0.00020086963094700247,
+      "loss": 1.9724,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8030356512530886,
+      "grad_norm": 0.034070172140250064,
+      "learning_rate": 0.00020047119596500854,
+      "loss": 2.2208,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8048005647723262,
+      "grad_norm": 0.03377996637204823,
+      "learning_rate": 0.00020007235901022322,
+      "loss": 2.0375,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8065654782915637,
+      "grad_norm": 0.03369303597609998,
+      "learning_rate": 0.00019967312325914346,
+      "loss": 2.0931,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8083303918108012,
+      "grad_norm": 0.03390741571878263,
+      "learning_rate": 0.0001992734918914425,
+      "loss": 1.9972,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8100953053300388,
+      "grad_norm": 0.0332681271049445,
+      "learning_rate": 0.00019887346808994446,
+      "loss": 2.0413,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8118602188492764,
+      "grad_norm": 0.03438053218088771,
+      "learning_rate": 0.00019847305504059888,
+      "loss": 2.0348,
+      "step": 2300
+    },
+    {
+      "epoch": 0.813625132368514,
+      "grad_norm": 0.03599754490402073,
+      "learning_rate": 0.00019807225593245533,
+      "loss": 2.1065,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8153900458877515,
+      "grad_norm": 0.03383294367019318,
+      "learning_rate": 0.00019767107395763826,
+      "loss": 2.1208,
+      "step": 2310
+    },
+    {
+      "epoch": 0.817154959406989,
+      "grad_norm": 0.036187693760982076,
+      "learning_rate": 0.0001972695123113214,
+      "loss": 2.0006,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8189198729262266,
+      "grad_norm": 0.03471902617434609,
+      "learning_rate": 0.00019686757419170236,
+      "loss": 1.8923,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8206847864454642,
+      "grad_norm": 0.035905691842104126,
+      "learning_rate": 0.00019646526279997698,
+      "loss": 2.1583,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8224496999647017,
+      "grad_norm": 0.034476087820152795,
+      "learning_rate": 0.00019606258134031415,
+      "loss": 2.0751,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8242146134839393,
+      "grad_norm": 0.03516357709469974,
+      "learning_rate": 0.0001956595330198301,
+      "loss": 2.0197,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8259795270031769,
+      "grad_norm": 0.03560768952994699,
+      "learning_rate": 0.00019525612104856274,
+      "loss": 2.2069,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8277444405224144,
+      "grad_norm": 0.03183591056028483,
+      "learning_rate": 0.00019485234863944657,
+      "loss": 2.1048,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8295093540416519,
+      "grad_norm": 0.031784734200597624,
+      "learning_rate": 0.00019444821900828637,
+      "loss": 1.8618,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8312742675608895,
+      "grad_norm": 0.03550619688076879,
+      "learning_rate": 0.00019404373537373226,
+      "loss": 2.0147,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8330391810801271,
+      "grad_norm": 0.03487649669359907,
+      "learning_rate": 0.00019363890095725362,
+      "loss": 1.8918,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8348040945993647,
+      "grad_norm": 0.03498002275247159,
+      "learning_rate": 0.00019323371898311367,
+      "loss": 2.0407,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8365690081186022,
+      "grad_norm": 0.036977436069902204,
+      "learning_rate": 0.00019282819267834377,
+      "loss": 2.0526,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8383339216378397,
+      "grad_norm": 0.033619636803852736,
+      "learning_rate": 0.00019242232527271754,
+      "loss": 1.9411,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8400988351570773,
+      "grad_norm": 0.03556834017261921,
+      "learning_rate": 0.00019201611999872538,
+      "loss": 2.1269,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8418637486763149,
+      "grad_norm": 0.0355931462629283,
+      "learning_rate": 0.00019160958009154853,
+      "loss": 2.1054,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8436286621955524,
+      "grad_norm": 0.03291016332021864,
+      "learning_rate": 0.00019120270878903345,
+      "loss": 1.9528,
+      "step": 2390
+    },
+    {
+      "epoch": 0.84539357571479,
+      "grad_norm": 0.03481457166438036,
+      "learning_rate": 0.00019079550933166585,
+      "loss": 2.0606,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8471584892340275,
+      "grad_norm": 0.03114909436222026,
+      "learning_rate": 0.00019038798496254518,
+      "loss": 2.0169,
+      "step": 2400
+    },
+    {
+      "epoch": 0.848923402753265,
+      "grad_norm": 0.03679426721444016,
+      "learning_rate": 0.00018998013892735838,
+      "loss": 2.0958,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8506883162725026,
+      "grad_norm": 0.0328534645099942,
+      "learning_rate": 0.00018957197447435457,
+      "loss": 1.9126,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8524532297917402,
+      "grad_norm": 0.03457712400451724,
+      "learning_rate": 0.00018916349485431858,
+      "loss": 1.9843,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8542181433109778,
+      "grad_norm": 0.03465174290508125,
+      "learning_rate": 0.00018875470332054548,
+      "loss": 1.9633,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8559830568302154,
+      "grad_norm": 0.03326982643658357,
+      "learning_rate": 0.00018834560312881458,
+      "loss": 2.1546,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8577479703494528,
+      "grad_norm": 0.032709197694297136,
+      "learning_rate": 0.0001879361975373635,
+      "loss": 1.877,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8595128838686904,
+      "grad_norm": 0.033755386616952134,
+      "learning_rate": 0.000187526489806862,
+      "loss": 2.0058,
+      "step": 2435
+    },
+    {
+      "epoch": 0.861277797387928,
+      "grad_norm": 0.03408280435423345,
+      "learning_rate": 0.00018711648320038636,
+      "loss": 1.9685,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8630427109071656,
+      "grad_norm": 0.03297097026193091,
+      "learning_rate": 0.00018670618098339323,
+      "loss": 2.0074,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8648076244264031,
+      "grad_norm": 0.034810103755827146,
+      "learning_rate": 0.00018629558642369346,
+      "loss": 1.9922,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8665725379456407,
+      "grad_norm": 0.03283833360547938,
+      "learning_rate": 0.00018588470279142652,
+      "loss": 1.9966,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8683374514648782,
+      "grad_norm": 0.03328550884584844,
+      "learning_rate": 0.00018547353335903382,
+      "loss": 1.9703,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8701023649841158,
+      "grad_norm": 0.03514416651889243,
+      "learning_rate": 0.0001850620814012333,
+      "loss": 2.2116,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8718672785033533,
+      "grad_norm": 0.033837741207024985,
+      "learning_rate": 0.0001846503501949928,
+      "loss": 2.1126,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8736321920225909,
+      "grad_norm": 0.03345433810476732,
+      "learning_rate": 0.00018423834301950444,
+      "loss": 2.1365,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8753971055418285,
+      "grad_norm": 0.032283686969351404,
+      "learning_rate": 0.0001838260631561581,
+      "loss": 2.0955,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8771620190610661,
+      "grad_norm": 0.03310144974854262,
+      "learning_rate": 0.00018341351388851554,
+      "loss": 2.0701,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8789269325803035,
+      "grad_norm": 0.032555566440577645,
+      "learning_rate": 0.00018300069850228407,
+      "loss": 2.1663,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8806918460995411,
+      "grad_norm": 0.03382832633148891,
+      "learning_rate": 0.0001825876202852906,
+      "loss": 1.9167,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8824567596187787,
+      "grad_norm": 0.033750270287439134,
+      "learning_rate": 0.00018217428252745527,
+      "loss": 2.0097,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8842216731380163,
+      "grad_norm": 0.03366446552723328,
+      "learning_rate": 0.00018176068852076522,
+      "loss": 2.0314,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8859865866572538,
+      "grad_norm": 0.035426688583260535,
+      "learning_rate": 0.00018134684155924862,
+      "loss": 2.0366,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8877515001764914,
+      "grad_norm": 0.033512709822249624,
+      "learning_rate": 0.00018093274493894822,
+      "loss": 1.929,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8895164136957289,
+      "grad_norm": 0.03177934822675584,
+      "learning_rate": 0.00018051840195789506,
+      "loss": 1.9234,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8912813272149664,
+      "grad_norm": 0.034336311090997225,
+      "learning_rate": 0.00018010381591608243,
+      "loss": 1.8586,
+      "step": 2525
+    },
+    {
+      "epoch": 0.893046240734204,
+      "grad_norm": 0.03310698686089758,
+      "learning_rate": 0.00017968899011543927,
+      "loss": 2.0306,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8948111542534416,
+      "grad_norm": 0.03268017246092975,
+      "learning_rate": 0.00017927392785980426,
+      "loss": 2.0074,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8965760677726792,
+      "grad_norm": 0.033657568041629796,
+      "learning_rate": 0.0001788586324548993,
+      "loss": 2.0077,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8983409812919168,
+      "grad_norm": 0.031741650322905525,
+      "learning_rate": 0.00017844310720830294,
+      "loss": 2.1483,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9001058948111542,
+      "grad_norm": 0.03217688749396154,
+      "learning_rate": 0.0001780273554294245,
+      "loss": 1.8316,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9018708083303918,
+      "grad_norm": 0.033548107692924885,
+      "learning_rate": 0.0001776113804294774,
+      "loss": 1.8651,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9036357218496294,
+      "grad_norm": 0.0331892007290859,
+      "learning_rate": 0.00017719518552145302,
+      "loss": 2.0009,
+      "step": 2560
+    },
+    {
+      "epoch": 0.905400635368867,
+      "grad_norm": 0.034152098471257095,
+      "learning_rate": 0.000176778774020094,
+      "loss": 2.0482,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9071655488881045,
+      "grad_norm": 0.03362407935533296,
+      "learning_rate": 0.00017636214924186805,
+      "loss": 2.0247,
+      "step": 2570
+    },
+    {
+      "epoch": 0.908930462407342,
+      "grad_norm": 0.03433353712344028,
+      "learning_rate": 0.00017594531450494162,
+      "loss": 2.025,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9106953759265796,
+      "grad_norm": 0.03186469418749604,
+      "learning_rate": 0.0001755282731291532,
+      "loss": 1.941,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9124602894458171,
+      "grad_norm": 0.0336514980864186,
+      "learning_rate": 0.0001751110284359872,
+      "loss": 1.9883,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9142252029650547,
+      "grad_norm": 0.033575507709247374,
+      "learning_rate": 0.00017469358374854713,
+      "loss": 1.9776,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9159901164842923,
+      "grad_norm": 0.03143527009840965,
+      "learning_rate": 0.0001742759423915295,
+      "loss": 1.9624,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9177550300035299,
+      "grad_norm": 0.04017674349332411,
+      "learning_rate": 0.00017385810769119715,
+      "loss": 1.8998,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9195199435227673,
+      "grad_norm": 0.10147298136890572,
+      "learning_rate": 0.00017344008297535268,
+      "loss": 2.0375,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9212848570420049,
+      "grad_norm": 0.04356273722033763,
+      "learning_rate": 0.00017302187157331218,
+      "loss": 1.9619,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9230497705612425,
+      "grad_norm": 0.0769652189955819,
+      "learning_rate": 0.00017260347681587852,
+      "loss": 2.0921,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9248146840804801,
+      "grad_norm": 0.08439587443987893,
+      "learning_rate": 0.00017218490203531488,
+      "loss": 1.9965,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9265795975997176,
+      "grad_norm": 0.05233213242900874,
+      "learning_rate": 0.00017176615056531825,
+      "loss": 1.8418,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9283445111189552,
+      "grad_norm": 0.043658805194593545,
+      "learning_rate": 0.00017134722574099276,
+      "loss": 2.1117,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9301094246381927,
+      "grad_norm": 0.059882971090714214,
+      "learning_rate": 0.00017092813089882334,
+      "loss": 2.1308,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9318743381574303,
+      "grad_norm": 0.04838485493831148,
+      "learning_rate": 0.0001705088693766489,
+      "loss": 2.1006,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9336392516766678,
+      "grad_norm": 0.043350536710871405,
+      "learning_rate": 0.00017008944451363586,
+      "loss": 2.0499,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9354041651959054,
+      "grad_norm": 0.04180685858976042,
+      "learning_rate": 0.0001696698596502517,
+      "loss": 1.8886,
+      "step": 2650
+    },
+    {
+      "epoch": 0.937169078715143,
+      "grad_norm": 0.03942140123822123,
+      "learning_rate": 0.00016925011812823795,
+      "loss": 2.0256,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9389339922343806,
+      "grad_norm": 0.04710574410726406,
+      "learning_rate": 0.00016883022329058407,
+      "loss": 1.9635,
+      "step": 2660
+    },
+    {
+      "epoch": 0.940698905753618,
+      "grad_norm": 0.04215487259771187,
+      "learning_rate": 0.00016841017848150043,
+      "loss": 2.0084,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9424638192728556,
+      "grad_norm": 0.040521663541122854,
+      "learning_rate": 0.00016798998704639195,
+      "loss": 1.994,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9442287327920932,
+      "grad_norm": 0.044354025040441655,
+      "learning_rate": 0.00016756965233183127,
+      "loss": 2.147,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9459936463113308,
+      "grad_norm": 0.05221156356902017,
+      "learning_rate": 0.00016714917768553212,
+      "loss": 2.1125,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9477585598305683,
+      "grad_norm": 0.04190748044489629,
+      "learning_rate": 0.00016672856645632281,
+      "loss": 1.9652,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9495234733498059,
+      "grad_norm": 0.04030418410423746,
+      "learning_rate": 0.0001663078219941194,
+      "loss": 1.9353,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9512883868690434,
+      "grad_norm": 0.04454384662570148,
+      "learning_rate": 0.00016588694764989905,
+      "loss": 2.0403,
+      "step": 2695
+    },
+    {
+      "epoch": 0.953053300388281,
+      "grad_norm": 0.09832702780934671,
+      "learning_rate": 0.00016546594677567338,
+      "loss": 1.9162,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9548182139075185,
+      "grad_norm": 1.4856672814856848,
+      "learning_rate": 0.00016504482272446173,
+      "loss": 2.0594,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9565831274267561,
+      "grad_norm": 0.38210054110528996,
+      "learning_rate": 0.00016462357885026451,
+      "loss": 1.8185,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9583480409459937,
+      "grad_norm": 0.4326804710954502,
+      "learning_rate": 0.00016420221850803636,
+      "loss": 2.1007,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9601129544652311,
+      "grad_norm": 0.48021086351840253,
+      "learning_rate": 0.00016378074505365967,
+      "loss": 1.9731,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9618778679844687,
+      "grad_norm": 0.3357665550559798,
+      "learning_rate": 0.00016335916184391752,
+      "loss": 2.0191,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9636427815037063,
+      "grad_norm": 0.23898496061823835,
+      "learning_rate": 0.0001629374722364672,
+      "loss": 2.1089,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9654076950229439,
+      "grad_norm": 0.12476100829729746,
+      "learning_rate": 0.00016251567958981347,
+      "loss": 2.1129,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9671726085421815,
+      "grad_norm": 0.11357000703735999,
+      "learning_rate": 0.00016209378726328167,
+      "loss": 2.0357,
+      "step": 2740
+    },
+    {
+      "epoch": 0.968937522061419,
+      "grad_norm": 0.14699484083424258,
+      "learning_rate": 0.00016167179861699096,
+      "loss": 1.9399,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9707024355806565,
+      "grad_norm": 0.13780379525973888,
+      "learning_rate": 0.00016124971701182776,
+      "loss": 2.0116,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9724673490998941,
+      "grad_norm": 0.09763739291197564,
+      "learning_rate": 0.00016082754580941876,
+      "loss": 2.0625,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9742322626191317,
+      "grad_norm": 0.14516048025484143,
+      "learning_rate": 0.00016040528837210437,
+      "loss": 2.0045,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9759971761383692,
+      "grad_norm": 0.13517590987858727,
+      "learning_rate": 0.00015998294806291154,
+      "loss": 1.8777,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9777620896576068,
+      "grad_norm": 0.09063769862045969,
+      "learning_rate": 0.00015956052824552753,
+      "loss": 2.0501,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9795270031768444,
+      "grad_norm": 0.11526559679330342,
+      "learning_rate": 0.00015913803228427266,
+      "loss": 2.0181,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9812919166960818,
+      "grad_norm": 0.13568430711331458,
+      "learning_rate": 0.0001587154635440738,
+      "loss": 2.0077,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9830568302153194,
+      "grad_norm": 0.13134226205246313,
+      "learning_rate": 0.00015829282539043743,
+      "loss": 1.9184,
+      "step": 2785
+    },
+    {
+      "epoch": 0.984821743734557,
+      "grad_norm": 0.10245093105196082,
+      "learning_rate": 0.00015787012118942276,
+      "loss": 1.9212,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9865866572537946,
+      "grad_norm": 0.11313742423392072,
+      "learning_rate": 0.00015744735430761513,
+      "loss": 2.0666,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9883515707730322,
+      "grad_norm": 0.08210158446547428,
+      "learning_rate": 0.00015702452811209913,
+      "loss": 1.9224,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9901164842922697,
+      "grad_norm": 0.1620180154663213,
+      "learning_rate": 0.0001566016459704317,
+      "loss": 1.9829,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9918813978115072,
+      "grad_norm": 0.11204318038889231,
+      "learning_rate": 0.0001561787112506153,
+      "loss": 2.097,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9936463113307448,
+      "grad_norm": 0.0785993303696052,
+      "learning_rate": 0.00015575572732107122,
+      "loss": 2.057,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9954112248499823,
+      "grad_norm": 0.06867530974191595,
+      "learning_rate": 0.00015533269755061267,
+      "loss": 1.869,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9971761383692199,
+      "grad_norm": 0.0733417713820362,
+      "learning_rate": 0.00015490962530841795,
+      "loss": 1.8386,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9989410518884575,
+      "grad_norm": 0.0760102082218872,
+      "learning_rate": 0.00015448651396400358,
+      "loss": 1.7059,
+      "step": 2830
+    },
+    {
+      "epoch": 1.000705965407695,
+      "grad_norm": 0.07337667273344894,
+      "learning_rate": 0.00015406336688719754,
+      "loss": 1.9036,
+      "step": 2835
+    },
+    {
+      "epoch": 1.0024708789269325,
+      "grad_norm": 0.07589223109110338,
+      "learning_rate": 0.00015364018744811242,
+      "loss": 2.0621,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0042357924461702,
+      "grad_norm": 0.07037428635896245,
+      "learning_rate": 0.0001532169790171185,
+      "loss": 2.0563,
+      "step": 2845
+    },
+    {
+      "epoch": 1.0060007059654077,
+      "grad_norm": 0.07590285997350149,
+      "learning_rate": 0.00015279374496481707,
+      "loss": 2.0112,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0077656194846452,
+      "grad_norm": 0.08362917378727402,
+      "learning_rate": 0.0001523704886620132,
+      "loss": 2.0247,
+      "step": 2855
+    },
+    {
+      "epoch": 1.0095305330038828,
+      "grad_norm": 0.08222846980057047,
+      "learning_rate": 0.00015194721347968961,
+      "loss": 2.0245,
+      "step": 2860
+    },
+    {
+      "epoch": 1.0112954465231203,
+      "grad_norm": 0.06725630529893803,
+      "learning_rate": 0.00015152392278897902,
+      "loss": 1.9894,
+      "step": 2865
+    },
+    {
+      "epoch": 1.013060360042358,
+      "grad_norm": 0.07028812765292143,
+      "learning_rate": 0.0001511006199611379,
+      "loss": 1.9291,
+      "step": 2870
+    },
+    {
+      "epoch": 1.0148252735615955,
+      "grad_norm": 0.06574980065749114,
+      "learning_rate": 0.00015067730836751912,
+      "loss": 1.7575,
+      "step": 2875
+    },
+    {
+      "epoch": 1.016590187080833,
+      "grad_norm": 0.07910079864748733,
+      "learning_rate": 0.0001502539913795456,
+      "loss": 1.9924,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0183551006000706,
+      "grad_norm": 0.07082133613408269,
+      "learning_rate": 0.00014983067236868313,
+      "loss": 1.9584,
+      "step": 2885
+    },
+    {
+      "epoch": 1.020120014119308,
+      "grad_norm": 0.08000004125294279,
+      "learning_rate": 0.00014940735470641368,
+      "loss": 1.954,
+      "step": 2890
+    },
+    {
+      "epoch": 1.0218849276385458,
+      "grad_norm": 0.06966307606190847,
+      "learning_rate": 0.0001489840417642084,
+      "loss": 1.8815,
+      "step": 2895
+    },
+    {
+      "epoch": 1.0236498411577832,
+      "grad_norm": 0.04989765544226569,
+      "learning_rate": 0.0001485607369135008,
+      "loss": 1.8912,
+      "step": 2900
+    },
+    {
+      "epoch": 1.025414754677021,
+      "grad_norm": 0.09896136422307783,
+      "learning_rate": 0.0001481374435256601,
+      "loss": 1.8818,
+      "step": 2905
+    },
+    {
+      "epoch": 1.0271796681962584,
+      "grad_norm": 0.049086160158615395,
+      "learning_rate": 0.00014771416497196417,
+      "loss": 2.0263,
+      "step": 2910
+    },
+    {
+      "epoch": 1.0289445817154959,
+      "grad_norm": 0.04879201475104746,
+      "learning_rate": 0.0001472909046235726,
+      "loss": 2.1125,
+      "step": 2915
+    },
+    {
+      "epoch": 1.0307094952347335,
+      "grad_norm": 0.05475936091574238,
+      "learning_rate": 0.00014686766585150017,
+      "loss": 2.0275,
+      "step": 2920
+    },
+    {
+      "epoch": 1.032474408753971,
+      "grad_norm": 0.048771161093180304,
+      "learning_rate": 0.00014644445202658968,
+      "loss": 1.9635,
+      "step": 2925
+    },
+    {
+      "epoch": 1.0342393222732087,
+      "grad_norm": 0.05410279348686431,
+      "learning_rate": 0.00014602126651948544,
+      "loss": 1.9654,
+      "step": 2930
+    },
+    {
+      "epoch": 1.0360042357924462,
+      "grad_norm": 0.04439304624772019,
+      "learning_rate": 0.00014559811270060594,
+      "loss": 2.0137,
+      "step": 2935
+    },
+    {
+      "epoch": 1.0377691493116836,
+      "grad_norm": 0.04946084465786224,
+      "learning_rate": 0.00014517499394011755,
+      "loss": 2.0591,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0395340628309213,
+      "grad_norm": 0.04628087419374212,
+      "learning_rate": 0.00014475191360790717,
+      "loss": 2.1446,
+      "step": 2945
+    },
+    {
+      "epoch": 1.0412989763501588,
+      "grad_norm": 0.04893099442976669,
+      "learning_rate": 0.00014432887507355587,
+      "loss": 1.893,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0430638898693965,
+      "grad_norm": 0.05605667282142996,
+      "learning_rate": 0.00014390588170631183,
+      "loss": 2.0861,
+      "step": 2955
+    },
+    {
+      "epoch": 1.044828803388634,
+      "grad_norm": 0.06758114575265582,
+      "learning_rate": 0.0001434829368750633,
+      "loss": 1.8338,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0465937169078716,
+      "grad_norm": 0.056420006212586714,
+      "learning_rate": 0.0001430600439483121,
+      "loss": 1.8763,
+      "step": 2965
+    },
+    {
+      "epoch": 1.048358630427109,
+      "grad_norm": 0.04262420319352538,
+      "learning_rate": 0.00014263720629414657,
+      "loss": 1.9167,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0501235439463465,
+      "grad_norm": 0.046142351015556984,
+      "learning_rate": 0.00014221442728021505,
+      "loss": 1.9689,
+      "step": 2975
+    },
+    {
+      "epoch": 1.0518884574655842,
+      "grad_norm": 0.05043607137322254,
+      "learning_rate": 0.0001417917102736986,
+      "loss": 2.0867,
+      "step": 2980
+    },
+    {
+      "epoch": 1.0536533709848217,
+      "grad_norm": 0.052163209459506144,
+      "learning_rate": 0.00014136905864128453,
+      "loss": 1.8988,
+      "step": 2985
+    },
+    {
+      "epoch": 1.0554182845040594,
+      "grad_norm": 0.04968349194670495,
+      "learning_rate": 0.0001409464757491396,
+      "loss": 1.898,
+      "step": 2990
+    },
+    {
+      "epoch": 1.0571831980232969,
+      "grad_norm": 0.04719855223158864,
+      "learning_rate": 0.00014052396496288286,
+      "loss": 2.0711,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0589481115425343,
+      "grad_norm": 0.07332811857742193,
+      "learning_rate": 0.00014010152964755927,
+      "loss": 1.9845,
+      "step": 3000
+    },
+    {
+      "epoch": 1.060713025061772,
+      "grad_norm": 0.11425369275007886,
+      "learning_rate": 0.0001396791731676126,
+      "loss": 2.1484,
+      "step": 3005
+    },
+    {
+      "epoch": 1.0624779385810095,
+      "grad_norm": 0.0779084772321061,
+      "learning_rate": 0.00013925689888685878,
+      "loss": 1.9079,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0642428521002472,
+      "grad_norm": 0.07947882876171825,
+      "learning_rate": 0.00013883471016845915,
+      "loss": 2.0101,
+      "step": 3015
+    },
+    {
+      "epoch": 1.0660077656194846,
+      "grad_norm": 0.06968763507879955,
+      "learning_rate": 0.00013841261037489343,
+      "loss": 1.8759,
+      "step": 3020
+    },
+    {
+      "epoch": 1.0677726791387223,
+      "grad_norm": 0.07592590476818559,
+      "learning_rate": 0.00013799060286793328,
+      "loss": 1.8954,
+      "step": 3025
+    },
+    {
+      "epoch": 1.0695375926579598,
+      "grad_norm": 0.07000611325204889,
+      "learning_rate": 0.00013756869100861518,
+      "loss": 2.003,
+      "step": 3030
+    },
+    {
+      "epoch": 1.0713025061771972,
+      "grad_norm": 0.07669248125153308,
+      "learning_rate": 0.00013714687815721398,
+      "loss": 2.1761,
+      "step": 3035
+    },
+    {
+      "epoch": 1.073067419696435,
+      "grad_norm": 0.06539172555990422,
+      "learning_rate": 0.000136725167673216,
+      "loss": 1.9865,
+      "step": 3040
+    },
+    {
+      "epoch": 1.0748323332156724,
+      "grad_norm": 0.07103495425034416,
+      "learning_rate": 0.00013630356291529215,
+      "loss": 2.0329,
+      "step": 3045
+    },
+    {
+      "epoch": 1.07659724673491,
+      "grad_norm": 0.06866142907770165,
+      "learning_rate": 0.00013588206724127135,
+      "loss": 1.9419,
+      "step": 3050
+    },
+    {
+      "epoch": 1.0783621602541476,
+      "grad_norm": 0.07688750830223512,
+      "learning_rate": 0.00013546068400811367,
+      "loss": 1.8413,
+      "step": 3055
+    },
+    {
+      "epoch": 1.080127073773385,
+      "grad_norm": 0.07877883407388588,
+      "learning_rate": 0.00013503941657188378,
+      "loss": 2.0163,
+      "step": 3060
+    },
+    {
+      "epoch": 1.0818919872926227,
+      "grad_norm": 0.08220186465541197,
+      "learning_rate": 0.0001346182682877241,
+      "loss": 1.9333,
+      "step": 3065
+    },
+    {
+      "epoch": 1.0836569008118602,
+      "grad_norm": 0.08626523157026107,
+      "learning_rate": 0.00013419724250982794,
+      "loss": 1.7206,
+      "step": 3070
+    },
+    {
+      "epoch": 1.0854218143310979,
+      "grad_norm": 0.07290793596211015,
+      "learning_rate": 0.00013377634259141298,
+      "loss": 1.9189,
+      "step": 3075
+    },
+    {
+      "epoch": 1.0871867278503353,
+      "grad_norm": 0.07971198025121311,
+      "learning_rate": 0.00013335557188469448,
+      "loss": 2.0069,
+      "step": 3080
+    },
+    {
+      "epoch": 1.0889516413695728,
+      "grad_norm": 0.07505840980640857,
+      "learning_rate": 0.00013293493374085872,
+      "loss": 1.9691,
+      "step": 3085
+    },
+    {
+      "epoch": 1.0907165548888105,
+      "grad_norm": 0.07036482223862744,
+      "learning_rate": 0.00013251443151003612,
+      "loss": 1.8643,
+      "step": 3090
+    },
+    {
+      "epoch": 1.092481468408048,
+      "grad_norm": 0.07634026063823118,
+      "learning_rate": 0.00013209406854127457,
+      "loss": 2.0899,
+      "step": 3095
+    },
+    {
+      "epoch": 1.0942463819272856,
+      "grad_norm": 0.03251069346182019,
+      "learning_rate": 0.00013167384818251295,
+      "loss": 1.7938,
+      "step": 3100
+    },
+    {
+      "epoch": 1.096011295446523,
+      "grad_norm": 0.038723980730320934,
+      "learning_rate": 0.00013125377378055418,
+      "loss": 1.993,
+      "step": 3105
+    },
+    {
+      "epoch": 1.0977762089657608,
+      "grad_norm": 0.03588108226931451,
+      "learning_rate": 0.00013083384868103892,
+      "loss": 1.8665,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0995411224849982,
+      "grad_norm": 0.03534782908949122,
+      "learning_rate": 0.00013041407622841857,
+      "loss": 1.9102,
+      "step": 3115
+    },
+    {
+      "epoch": 1.1013060360042357,
+      "grad_norm": 0.03394423715517006,
+      "learning_rate": 0.00012999445976592878,
+      "loss": 2.018,
+      "step": 3120
+    },
+    {
+      "epoch": 1.1030709495234734,
+      "grad_norm": 0.033174070244108365,
+      "learning_rate": 0.00012957500263556304,
+      "loss": 2.1041,
+      "step": 3125
+    },
+    {
+      "epoch": 1.1048358630427109,
+      "grad_norm": 0.0352196315248866,
+      "learning_rate": 0.00012915570817804552,
+      "loss": 1.8694,
+      "step": 3130
+    },
+    {
+      "epoch": 1.1066007765619486,
+      "grad_norm": 0.031168756944625068,
+      "learning_rate": 0.00012873657973280515,
+      "loss": 2.06,
+      "step": 3135
+    },
+    {
+      "epoch": 1.108365690081186,
+      "grad_norm": 0.03339282397455765,
+      "learning_rate": 0.00012831762063794837,
+      "loss": 2.2009,
+      "step": 3140
+    },
+    {
+      "epoch": 1.1101306036004235,
+      "grad_norm": 0.033375354612009904,
+      "learning_rate": 0.000127898834230233,
+      "loss": 2.0763,
+      "step": 3145
+    },
+    {
+      "epoch": 1.1118955171196612,
+      "grad_norm": 0.03418125212359994,
+      "learning_rate": 0.00012748022384504162,
+      "loss": 2.0429,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1136604306388986,
+      "grad_norm": 0.03421647597656662,
+      "learning_rate": 0.00012706179281635465,
+      "loss": 1.9824,
+      "step": 3155
+    },
+    {
+      "epoch": 1.1154253441581363,
+      "grad_norm": 0.034398741808789754,
+      "learning_rate": 0.00012664354447672422,
+      "loss": 1.9851,
+      "step": 3160
+    },
+    {
+      "epoch": 1.1171902576773738,
+      "grad_norm": 0.03666933867331459,
+      "learning_rate": 0.00012622548215724728,
+      "loss": 2.0281,
+      "step": 3165
+    },
+    {
+      "epoch": 1.1189551711966113,
+      "grad_norm": 0.03523282657003033,
+      "learning_rate": 0.00012580760918753944,
+      "loss": 2.0809,
+      "step": 3170
+    },
+    {
+      "epoch": 1.120720084715849,
+      "grad_norm": 0.03387976748553185,
+      "learning_rate": 0.0001253899288957082,
+      "loss": 1.9259,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1224849982350864,
+      "grad_norm": 0.034136896604200544,
+      "learning_rate": 0.00012497244460832641,
+      "loss": 2.0725,
+      "step": 3180
+    },
+    {
+      "epoch": 1.124249911754324,
+      "grad_norm": 0.032174670370331206,
+      "learning_rate": 0.00012455515965040598,
+      "loss": 1.9467,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1260148252735616,
+      "grad_norm": 0.03221364698070383,
+      "learning_rate": 0.00012413807734537106,
+      "loss": 1.9903,
+      "step": 3190
+    },
+    {
+      "epoch": 1.1277797387927992,
+      "grad_norm": 0.036476137419521416,
+      "learning_rate": 0.0001237212010150321,
+      "loss": 2.111,
+      "step": 3195
+    },
+    {
+      "epoch": 1.1295446523120367,
+      "grad_norm": 0.031399463423770896,
+      "learning_rate": 0.0001233045339795589,
+      "loss": 1.9945,
+      "step": 3200
+    },
+    {
+      "epoch": 1.1313095658312742,
+      "grad_norm": 0.030824431922845962,
+      "learning_rate": 0.0001228880795574543,
+      "loss": 1.9593,
+      "step": 3205
+    },
+    {
+      "epoch": 1.1330744793505119,
+      "grad_norm": 0.03211287501208764,
+      "learning_rate": 0.000122471841065528,
+      "loss": 1.9419,
+      "step": 3210
+    },
+    {
+      "epoch": 1.1348393928697493,
+      "grad_norm": 0.03229559162774337,
+      "learning_rate": 0.0001220558218188697,
+      "loss": 2.1435,
+      "step": 3215
+    },
+    {
+      "epoch": 1.136604306388987,
+      "grad_norm": 0.028865250894228946,
+      "learning_rate": 0.00012164002513082318,
+      "loss": 2.0217,
+      "step": 3220
+    },
+    {
+      "epoch": 1.1383692199082245,
+      "grad_norm": 0.030763799222927924,
+      "learning_rate": 0.00012122445431295958,
+      "loss": 1.8215,
+      "step": 3225
+    },
+    {
+      "epoch": 1.1401341334274622,
+      "grad_norm": 0.03514777669043741,
+      "learning_rate": 0.00012080911267505103,
+      "loss": 1.9606,
+      "step": 3230
+    },
+    {
+      "epoch": 1.1418990469466996,
+      "grad_norm": 0.03145844309266733,
+      "learning_rate": 0.00012039400352504451,
+      "loss": 2.1079,
+      "step": 3235
+    },
+    {
+      "epoch": 1.143663960465937,
+      "grad_norm": 0.030806559968599934,
+      "learning_rate": 0.0001199791301690353,
+      "loss": 1.8784,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1454288739851748,
+      "grad_norm": 0.03823663786497356,
+      "learning_rate": 0.00011956449591124077,
+      "loss": 2.1676,
+      "step": 3245
+    },
+    {
+      "epoch": 1.1471937875044123,
+      "grad_norm": 0.03009831829614018,
+      "learning_rate": 0.00011915010405397389,
+      "loss": 1.9355,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1489587010236497,
+      "grad_norm": 0.035210465103021234,
+      "learning_rate": 0.00011873595789761723,
+      "loss": 2.0074,
+      "step": 3255
+    },
+    {
+      "epoch": 1.1507236145428874,
+      "grad_norm": 0.030012800086470587,
+      "learning_rate": 0.00011832206074059648,
+      "loss": 1.8919,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1524885280621249,
+      "grad_norm": 0.03595767439905811,
+      "learning_rate": 0.00011790841587935405,
+      "loss": 1.8595,
+      "step": 3265
+    },
+    {
+      "epoch": 1.1542534415813626,
+      "grad_norm": 0.030822132960768998,
+      "learning_rate": 0.00011749502660832313,
+      "loss": 2.058,
+      "step": 3270
+    },
+    {
+      "epoch": 1.1560183551006,
+      "grad_norm": 0.03154968815500067,
+      "learning_rate": 0.00011708189621990112,
+      "loss": 2.0532,
+      "step": 3275
+    },
+    {
+      "epoch": 1.1577832686198377,
+      "grad_norm": 0.040080406157402514,
+      "learning_rate": 0.00011666902800442375,
+      "loss": 2.018,
+      "step": 3280
+    },
+    {
+      "epoch": 1.1595481821390752,
+      "grad_norm": 0.029961190157627198,
+      "learning_rate": 0.0001162564252501386,
+      "loss": 2.0849,
+      "step": 3285
+    },
+    {
+      "epoch": 1.1613130956583126,
+      "grad_norm": 0.03196867255519621,
+      "learning_rate": 0.00011584409124317904,
+      "loss": 1.8444,
+      "step": 3290
+    },
+    {
+      "epoch": 1.1630780091775503,
+      "grad_norm": 0.030014312506980993,
+      "learning_rate": 0.00011543202926753803,
+      "loss": 1.9923,
+      "step": 3295
+    },
+    {
+      "epoch": 1.1648429226967878,
+      "grad_norm": 0.033147821664570046,
+      "learning_rate": 0.0001150202426050419,
+      "loss": 1.9986,
+      "step": 3300
+    },
+    {
+      "epoch": 1.1666078362160255,
+      "grad_norm": 0.0389611966917462,
+      "learning_rate": 0.00011460873453532438,
+      "loss": 1.9281,
+      "step": 3305
+    },
+    {
+      "epoch": 1.168372749735263,
+      "grad_norm": 0.031935670371534816,
+      "learning_rate": 0.00011419750833580035,
+      "loss": 1.9358,
+      "step": 3310
+    },
+    {
+      "epoch": 1.1701376632545006,
+      "grad_norm": 0.030428834410082803,
+      "learning_rate": 0.00011378656728163966,
+      "loss": 1.8651,
+      "step": 3315
+    },
+    {
+      "epoch": 1.171902576773738,
+      "grad_norm": 0.03281604951134294,
+      "learning_rate": 0.00011337591464574131,
+      "loss": 1.8506,
+      "step": 3320
+    },
+    {
+      "epoch": 1.1736674902929756,
+      "grad_norm": 0.03421610885263129,
+      "learning_rate": 0.00011296555369870706,
+      "loss": 1.9455,
+      "step": 3325
+    },
+    {
+      "epoch": 1.1754324038122133,
+      "grad_norm": 0.03131134010133921,
+      "learning_rate": 0.00011255548770881568,
+      "loss": 2.0772,
+      "step": 3330
+    },
+    {
+      "epoch": 1.1771973173314507,
+      "grad_norm": 0.032843412553030034,
+      "learning_rate": 0.00011214571994199674,
+      "loss": 1.8982,
+      "step": 3335
+    },
+    {
+      "epoch": 1.1789622308506884,
+      "grad_norm": 0.030727755899184493,
+      "learning_rate": 0.00011173625366180455,
+      "loss": 1.9232,
+      "step": 3340
+    },
+    {
+      "epoch": 1.1807271443699259,
+      "grad_norm": 0.029541004866353728,
+      "learning_rate": 0.00011132709212939245,
+      "loss": 1.9702,
+      "step": 3345
+    },
+    {
+      "epoch": 1.1824920578891633,
+      "grad_norm": 0.030726562855779436,
+      "learning_rate": 0.00011091823860348643,
+      "loss": 1.8255,
+      "step": 3350
+    },
+    {
+      "epoch": 1.184256971408401,
+      "grad_norm": 0.03156401680159205,
+      "learning_rate": 0.00011050969634035949,
+      "loss": 1.8687,
+      "step": 3355
+    },
+    {
+      "epoch": 1.1860218849276385,
+      "grad_norm": 0.03127613884082788,
+      "learning_rate": 0.00011010146859380575,
+      "loss": 2.0063,
+      "step": 3360
+    },
+    {
+      "epoch": 1.1877867984468762,
+      "grad_norm": 0.03308661953307029,
+      "learning_rate": 0.00010969355861511411,
+      "loss": 2.1353,
+      "step": 3365
+    },
+    {
+      "epoch": 1.1895517119661136,
+      "grad_norm": 0.030159409543144285,
+      "learning_rate": 0.0001092859696530429,
+      "loss": 1.885,
+      "step": 3370
+    },
+    {
+      "epoch": 1.191316625485351,
+      "grad_norm": 0.029702148333993476,
+      "learning_rate": 0.00010887870495379358,
+      "loss": 1.9941,
+      "step": 3375
+    },
+    {
+      "epoch": 1.1930815390045888,
+      "grad_norm": 0.031102469318039554,
+      "learning_rate": 0.00010847176776098514,
+      "loss": 1.9189,
+      "step": 3380
+    },
+    {
+      "epoch": 1.1948464525238263,
+      "grad_norm": 0.030535639390433026,
+      "learning_rate": 0.00010806516131562807,
+      "loss": 1.9432,
+      "step": 3385
+    },
+    {
+      "epoch": 1.196611366043064,
+      "grad_norm": 0.03407485096416553,
+      "learning_rate": 0.0001076588888560988,
+      "loss": 1.9876,
+      "step": 3390
+    },
+    {
+      "epoch": 1.1983762795623014,
+      "grad_norm": 0.032959257552623265,
+      "learning_rate": 0.0001072529536181137,
+      "loss": 2.0636,
+      "step": 3395
+    },
+    {
+      "epoch": 1.200141193081539,
+      "grad_norm": 0.03032018419400235,
+      "learning_rate": 0.00010684735883470331,
+      "loss": 1.9318,
+      "step": 3400
+    },
+    {
+      "epoch": 1.2019061066007766,
+      "grad_norm": 0.03158701784367694,
+      "learning_rate": 0.00010644210773618679,
+      "loss": 1.9487,
+      "step": 3405
+    },
+    {
+      "epoch": 1.203671020120014,
+      "grad_norm": 0.032495071860565446,
+      "learning_rate": 0.00010603720355014586,
+      "loss": 1.8043,
+      "step": 3410
+    },
+    {
+      "epoch": 1.2054359336392517,
+      "grad_norm": 0.030415017614548217,
+      "learning_rate": 0.00010563264950139949,
+      "loss": 2.0779,
+      "step": 3415
+    },
+    {
+      "epoch": 1.2072008471584892,
+      "grad_norm": 0.030994232739364636,
+      "learning_rate": 0.00010522844881197796,
+      "loss": 1.8511,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2089657606777269,
+      "grad_norm": 0.02990713438925777,
+      "learning_rate": 0.00010482460470109711,
+      "loss": 2.0744,
+      "step": 3425
+    },
+    {
+      "epoch": 1.2107306741969643,
+      "grad_norm": 0.029954678305008517,
+      "learning_rate": 0.00010442112038513308,
+      "loss": 1.8954,
+      "step": 3430
+    },
+    {
+      "epoch": 1.212495587716202,
+      "grad_norm": 0.030866059429023793,
+      "learning_rate": 0.00010401799907759626,
+      "loss": 1.9577,
+      "step": 3435
+    },
+    {
+      "epoch": 1.2142605012354395,
+      "grad_norm": 0.030394062939301546,
+      "learning_rate": 0.00010361524398910602,
+      "loss": 1.9895,
+      "step": 3440
+    },
+    {
+      "epoch": 1.216025414754677,
+      "grad_norm": 0.029794056256019365,
+      "learning_rate": 0.00010321285832736501,
+      "loss": 2.0698,
+      "step": 3445
+    },
+    {
+      "epoch": 1.2177903282739146,
+      "grad_norm": 0.030935648080681492,
+      "learning_rate": 0.00010281084529713346,
+      "loss": 2.0168,
+      "step": 3450
+    },
+    {
+      "epoch": 1.2195552417931521,
+      "grad_norm": 0.03000407036964225,
+      "learning_rate": 0.00010240920810020408,
+      "loss": 2.0395,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2213201553123896,
+      "grad_norm": 0.030470268414440352,
+      "learning_rate": 0.00010200794993537605,
+      "loss": 1.9029,
+      "step": 3460
+    },
+    {
+      "epoch": 1.2230850688316273,
+      "grad_norm": 0.02990718014055773,
+      "learning_rate": 0.00010160707399842987,
+      "loss": 1.8046,
+      "step": 3465
+    },
+    {
+      "epoch": 1.2248499823508647,
+      "grad_norm": 0.029866689308828343,
+      "learning_rate": 0.00010120658348210194,
+      "loss": 1.8879,
+      "step": 3470
+    },
+    {
+      "epoch": 1.2266148958701024,
+      "grad_norm": 0.028836964079095424,
+      "learning_rate": 0.00010080648157605881,
+      "loss": 2.0476,
+      "step": 3475
+    },
+    {
+      "epoch": 1.2283798093893399,
+      "grad_norm": 0.03234687909376445,
+      "learning_rate": 0.00010040677146687218,
+      "loss": 2.0518,
+      "step": 3480
+    },
+    {
+      "epoch": 1.2301447229085776,
+      "grad_norm": 0.028337352370958524,
+      "learning_rate": 0.00010000745633799317,
+      "loss": 1.9547,
+      "step": 3485
+    },
+    {
+      "epoch": 1.231909636427815,
+      "grad_norm": 0.029306606507423914,
+      "learning_rate": 9.96085393697272e-05,
+      "loss": 1.9264,
+      "step": 3490
+    },
+    {
+      "epoch": 1.2336745499470525,
+      "grad_norm": 0.029473809687637992,
+      "learning_rate": 9.921002373920848e-05,
+      "loss": 1.9587,
+      "step": 3495
+    },
+    {
+      "epoch": 1.2354394634662902,
+      "grad_norm": 0.03198727980706444,
+      "learning_rate": 9.881191262037501e-05,
+      "loss": 1.8541,
+      "step": 3500
+    },
+    {
+      "epoch": 1.2372043769855277,
+      "grad_norm": 0.031184096631820344,
+      "learning_rate": 9.841420918394294e-05,
+      "loss": 1.9058,
+      "step": 3505
+    },
+    {
+      "epoch": 1.2389692905047653,
+      "grad_norm": 0.029830305044107316,
+      "learning_rate": 9.801691659738139e-05,
+      "loss": 1.9184,
+      "step": 3510
+    },
+    {
+      "epoch": 1.2407342040240028,
+      "grad_norm": 0.030750065456757465,
+      "learning_rate": 9.76200380248876e-05,
+      "loss": 1.9931,
+      "step": 3515
+    },
+    {
+      "epoch": 1.2424991175432405,
+      "grad_norm": 0.02917211898358977,
+      "learning_rate": 9.722357662736111e-05,
+      "loss": 2.0504,
+      "step": 3520
+    },
+    {
+      "epoch": 1.244264031062478,
+      "grad_norm": 0.030044198407593453,
+      "learning_rate": 9.682753556237914e-05,
+      "loss": 1.9253,
+      "step": 3525
+    },
+    {
+      "epoch": 1.2460289445817154,
+      "grad_norm": 0.031228420729122578,
+      "learning_rate": 9.643191798417115e-05,
+      "loss": 2.0314,
+      "step": 3530
+    },
+    {
+      "epoch": 1.2477938581009531,
+      "grad_norm": 0.030825577828092175,
+      "learning_rate": 9.603672704359368e-05,
+      "loss": 1.9675,
+      "step": 3535
+    },
+    {
+      "epoch": 1.2495587716201906,
+      "grad_norm": 0.030786626122809125,
+      "learning_rate": 9.564196588810556e-05,
+      "loss": 1.9578,
+      "step": 3540
+    },
+    {
+      "epoch": 1.251323685139428,
+      "grad_norm": 0.029261797690061857,
+      "learning_rate": 9.524763766174238e-05,
+      "loss": 1.8483,
+      "step": 3545
+    },
+    {
+      "epoch": 1.2530885986586657,
+      "grad_norm": 0.031687445857207994,
+      "learning_rate": 9.485374550509197e-05,
+      "loss": 2.1405,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2548535121779034,
+      "grad_norm": 0.030789757512965626,
+      "learning_rate": 9.4460292555269e-05,
+      "loss": 1.9838,
+      "step": 3555
+    },
+    {
+      "epoch": 1.2566184256971409,
+      "grad_norm": 0.03085385786781096,
+      "learning_rate": 9.406728194589003e-05,
+      "loss": 1.9995,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2583833392163783,
+      "grad_norm": 0.028945859623109978,
+      "learning_rate": 9.367471680704891e-05,
+      "loss": 1.8116,
+      "step": 3565
+    },
+    {
+      "epoch": 1.260148252735616,
+      "grad_norm": 0.030155495415531395,
+      "learning_rate": 9.328260026529128e-05,
+      "loss": 2.0901,
+      "step": 3570
+    },
+    {
+      "epoch": 1.2619131662548535,
+      "grad_norm": 0.030128890240650772,
+      "learning_rate": 9.289093544359012e-05,
+      "loss": 1.8706,
+      "step": 3575
+    },
+    {
+      "epoch": 1.263678079774091,
+      "grad_norm": 0.03075311496322448,
+      "learning_rate": 9.24997254613208e-05,
+      "loss": 2.0057,
+      "step": 3580
+    },
+    {
+      "epoch": 1.2654429932933287,
+      "grad_norm": 0.029777633208128304,
+      "learning_rate": 9.210897343423602e-05,
+      "loss": 1.9914,
+      "step": 3585
+    },
+    {
+      "epoch": 1.2672079068125661,
+      "grad_norm": 0.029045166460438053,
+      "learning_rate": 9.171868247444128e-05,
+      "loss": 2.1123,
+      "step": 3590
+    },
+    {
+      "epoch": 1.2689728203318038,
+      "grad_norm": 0.031084357387276943,
+      "learning_rate": 9.132885569036983e-05,
+      "loss": 2.0011,
+      "step": 3595
+    },
+    {
+      "epoch": 1.2707377338510413,
+      "grad_norm": 0.030998383255735507,
+      "learning_rate": 9.093949618675805e-05,
+      "loss": 2.0393,
+      "step": 3600
+    },
+    {
+      "epoch": 1.272502647370279,
+      "grad_norm": 0.034012219980734545,
+      "learning_rate": 9.055060706462092e-05,
+      "loss": 2.0263,
+      "step": 3605
+    },
+    {
+      "epoch": 1.2742675608895164,
+      "grad_norm": 0.03162666935245559,
+      "learning_rate": 9.016219142122679e-05,
+      "loss": 1.9792,
+      "step": 3610
+    },
+    {
+      "epoch": 1.276032474408754,
+      "grad_norm": 0.03158373469449993,
+      "learning_rate": 8.977425235007329e-05,
+      "loss": 2.1153,
+      "step": 3615
+    },
+    {
+      "epoch": 1.2777973879279916,
+      "grad_norm": 0.02987208569419512,
+      "learning_rate": 8.938679294086225e-05,
+      "loss": 2.0119,
+      "step": 3620
+    },
+    {
+      "epoch": 1.279562301447229,
+      "grad_norm": 0.030889178819891533,
+      "learning_rate": 8.899981627947551e-05,
+      "loss": 1.9112,
+      "step": 3625
+    },
+    {
+      "epoch": 1.2813272149664665,
+      "grad_norm": 0.030529433425646905,
+      "learning_rate": 8.861332544794984e-05,
+      "loss": 2.1946,
+      "step": 3630
+    },
+    {
+      "epoch": 1.2830921284857042,
+      "grad_norm": 0.031019883722968555,
+      "learning_rate": 8.82273235244529e-05,
+      "loss": 2.2375,
+      "step": 3635
+    },
+    {
+      "epoch": 1.2848570420049419,
+      "grad_norm": 0.029064623912400923,
+      "learning_rate": 8.784181358325838e-05,
+      "loss": 2.0459,
+      "step": 3640
+    },
+    {
+      "epoch": 1.2866219555241794,
+      "grad_norm": 0.030009701908391445,
+      "learning_rate": 8.745679869472159e-05,
+      "loss": 2.0742,
+      "step": 3645
+    },
+    {
+      "epoch": 1.2883868690434168,
+      "grad_norm": 0.03129026866184981,
+      "learning_rate": 8.707228192525518e-05,
+      "loss": 1.9052,
+      "step": 3650
+    },
+    {
+      "epoch": 1.2901517825626545,
+      "grad_norm": 0.030710848052605294,
+      "learning_rate": 8.66882663373044e-05,
+      "loss": 1.8587,
+      "step": 3655
+    },
+    {
+      "epoch": 1.291916696081892,
+      "grad_norm": 0.029438103608201007,
+      "learning_rate": 8.630475498932303e-05,
+      "loss": 2.0541,
+      "step": 3660
+    },
+    {
+      "epoch": 1.2936816096011294,
+      "grad_norm": 0.02823666819685662,
+      "learning_rate": 8.59217509357489e-05,
+      "loss": 1.8486,
+      "step": 3665
+    },
+    {
+      "epoch": 1.2954465231203671,
+      "grad_norm": 0.030894190341993692,
+      "learning_rate": 8.553925722697944e-05,
+      "loss": 2.0947,
+      "step": 3670
+    },
+    {
+      "epoch": 1.2972114366396046,
+      "grad_norm": 0.02914617905324265,
+      "learning_rate": 8.51572769093475e-05,
+      "loss": 1.9047,
+      "step": 3675
+    },
+    {
+      "epoch": 1.2989763501588423,
+      "grad_norm": 0.03072247064606092,
+      "learning_rate": 8.477581302509708e-05,
+      "loss": 2.1091,
+      "step": 3680
+    },
+    {
+      "epoch": 1.3007412636780797,
+      "grad_norm": 0.030161184552864203,
+      "learning_rate": 8.439486861235916e-05,
+      "loss": 2.1205,
+      "step": 3685
+    },
+    {
+      "epoch": 1.3025061771973174,
+      "grad_norm": 0.029381067607062526,
+      "learning_rate": 8.401444670512748e-05,
+      "loss": 1.9346,
+      "step": 3690
+    },
+    {
+      "epoch": 1.304271090716555,
+      "grad_norm": 0.028806108562055628,
+      "learning_rate": 8.363455033323412e-05,
+      "loss": 1.9801,
+      "step": 3695
+    },
+    {
+      "epoch": 1.3060360042357924,
+      "grad_norm": 0.030292456550096224,
+      "learning_rate": 8.325518252232591e-05,
+      "loss": 2.1433,
+      "step": 3700
+    },
+    {
+      "epoch": 1.30780091775503,
+      "grad_norm": 0.031342730199109825,
+      "learning_rate": 8.287634629383963e-05,
+      "loss": 1.8015,
+      "step": 3705
+    },
+    {
+      "epoch": 1.3095658312742675,
+      "grad_norm": 0.030742138431772023,
+      "learning_rate": 8.249804466497874e-05,
+      "loss": 2.0684,
+      "step": 3710
+    },
+    {
+      "epoch": 1.3113307447935052,
+      "grad_norm": 0.031609591109127325,
+      "learning_rate": 8.212028064868857e-05,
+      "loss": 1.973,
+      "step": 3715
+    },
+    {
+      "epoch": 1.3130956583127427,
+      "grad_norm": 0.03156145121381424,
+      "learning_rate": 8.174305725363283e-05,
+      "loss": 1.8917,
+      "step": 3720
+    },
+    {
+      "epoch": 1.3148605718319804,
+      "grad_norm": 0.03241603868534105,
+      "learning_rate": 8.136637748416962e-05,
+      "loss": 1.9114,
+      "step": 3725
+    },
+    {
+      "epoch": 1.3166254853512178,
+      "grad_norm": 0.030631325425011652,
+      "learning_rate": 8.099024434032717e-05,
+      "loss": 1.9653,
+      "step": 3730
+    },
+    {
+      "epoch": 1.3183903988704553,
+      "grad_norm": 0.03186459486256166,
+      "learning_rate": 8.061466081778035e-05,
+      "loss": 1.8212,
+      "step": 3735
+    },
+    {
+      "epoch": 1.320155312389693,
+      "grad_norm": 0.03202163541066731,
+      "learning_rate": 8.023962990782657e-05,
+      "loss": 2.044,
+      "step": 3740
+    },
+    {
+      "epoch": 1.3219202259089304,
+      "grad_norm": 0.030803329235101193,
+      "learning_rate": 7.986515459736193e-05,
+      "loss": 1.9966,
+      "step": 3745
+    },
+    {
+      "epoch": 1.323685139428168,
+      "grad_norm": 0.030128829450787764,
+      "learning_rate": 7.949123786885764e-05,
+      "loss": 2.0253,
+      "step": 3750
+    },
+    {
+      "epoch": 1.3254500529474056,
+      "grad_norm": 0.03299000453503951,
+      "learning_rate": 7.911788270033608e-05,
+      "loss": 1.8925,
+      "step": 3755
+    },
+    {
+      "epoch": 1.327214966466643,
+      "grad_norm": 0.0324077675613907,
+      "learning_rate": 7.874509206534709e-05,
+      "loss": 1.9373,
+      "step": 3760
+    },
+    {
+      "epoch": 1.3289798799858807,
+      "grad_norm": 0.03250609524308436,
+      "learning_rate": 7.837286893294438e-05,
+      "loss": 1.9891,
+      "step": 3765
+    },
+    {
+      "epoch": 1.3307447935051182,
+      "grad_norm": 0.031241675070779008,
+      "learning_rate": 7.800121626766189e-05,
+      "loss": 1.79,
+      "step": 3770
+    },
+    {
+      "epoch": 1.332509707024356,
+      "grad_norm": 0.03147293740440876,
+      "learning_rate": 7.763013702949015e-05,
+      "loss": 2.0676,
+      "step": 3775
+    },
+    {
+      "epoch": 1.3342746205435934,
+      "grad_norm": 0.029915200474572905,
+      "learning_rate": 7.72596341738525e-05,
+      "loss": 1.915,
+      "step": 3780
+    },
+    {
+      "epoch": 1.3360395340628308,
+      "grad_norm": 0.030334178625772805,
+      "learning_rate": 7.6889710651582e-05,
+      "loss": 1.9927,
+      "step": 3785
+    },
+    {
+      "epoch": 1.3378044475820685,
+      "grad_norm": 0.030683396349377345,
+      "learning_rate": 7.652036940889745e-05,
+      "loss": 1.8792,
+      "step": 3790
+    },
+    {
+      "epoch": 1.339569361101306,
+      "grad_norm": 0.031111190639847297,
+      "learning_rate": 7.61516133873803e-05,
+      "loss": 1.889,
+      "step": 3795
+    },
+    {
+      "epoch": 1.3413342746205437,
+      "grad_norm": 0.03232671453012522,
+      "learning_rate": 7.578344552395097e-05,
+      "loss": 1.9541,
+      "step": 3800
+    },
+    {
+      "epoch": 1.3430991881397811,
+      "grad_norm": 0.03309394950753326,
+      "learning_rate": 7.541586875084551e-05,
+      "loss": 2.038,
+      "step": 3805
+    },
+    {
+      "epoch": 1.3448641016590188,
+      "grad_norm": 0.034339421262217736,
+      "learning_rate": 7.504888599559246e-05,
+      "loss": 1.8775,
+      "step": 3810
+    },
+    {
+      "epoch": 1.3466290151782563,
+      "grad_norm": 0.03498688064044485,
+      "learning_rate": 7.468250018098918e-05,
+      "loss": 1.9517,
+      "step": 3815
+    },
+    {
+      "epoch": 1.3483939286974937,
+      "grad_norm": 0.0336286559029592,
+      "learning_rate": 7.431671422507888e-05,
+      "loss": 1.8343,
+      "step": 3820
+    },
+    {
+      "epoch": 1.3501588422167314,
+      "grad_norm": 0.03356286834610317,
+      "learning_rate": 7.395153104112727e-05,
+      "loss": 2.0133,
+      "step": 3825
+    },
+    {
+      "epoch": 1.351923755735969,
+      "grad_norm": 0.0318699107480801,
+      "learning_rate": 7.358695353759919e-05,
+      "loss": 1.8356,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3536886692552064,
+      "grad_norm": 0.033391373631519,
+      "learning_rate": 7.322298461813577e-05,
+      "loss": 2.0353,
+      "step": 3835
+    },
+    {
+      "epoch": 1.355453582774444,
+      "grad_norm": 0.0326616476909793,
+      "learning_rate": 7.285962718153098e-05,
+      "loss": 2.0202,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3572184962936817,
+      "grad_norm": 0.0310991326358504,
+      "learning_rate": 7.249688412170889e-05,
+      "loss": 2.0012,
+      "step": 3845
+    },
+    {
+      "epoch": 1.3589834098129192,
+      "grad_norm": 0.033557843179260465,
+      "learning_rate": 7.213475832770023e-05,
+      "loss": 2.086,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3607483233321567,
+      "grad_norm": 0.03220485942446892,
+      "learning_rate": 7.177325268361965e-05,
+      "loss": 2.0772,
+      "step": 3855
+    },
+    {
+      "epoch": 1.3625132368513944,
+      "grad_norm": 0.03358709639504059,
+      "learning_rate": 7.141237006864275e-05,
+      "loss": 1.8464,
+      "step": 3860
+    },
+    {
+      "epoch": 1.3642781503706318,
+      "grad_norm": 0.0325469361019868,
+      "learning_rate": 7.105211335698294e-05,
+      "loss": 2.0004,
+      "step": 3865
+    },
+    {
+      "epoch": 1.3660430638898693,
+      "grad_norm": 0.03261608562751164,
+      "learning_rate": 7.069248541786887e-05,
+      "loss": 2.0437,
+      "step": 3870
+    },
+    {
+      "epoch": 1.367807977409107,
+      "grad_norm": 0.033757726003366594,
+      "learning_rate": 7.033348911552115e-05,
+      "loss": 1.9603,
+      "step": 3875
+    },
+    {
+      "epoch": 1.3695728909283444,
+      "grad_norm": 0.031547260146207665,
+      "learning_rate": 6.997512730912999e-05,
+      "loss": 1.9865,
+      "step": 3880
+    },
+    {
+      "epoch": 1.3713378044475821,
+      "grad_norm": 0.032842126200468585,
+      "learning_rate": 6.961740285283221e-05,
+      "loss": 1.85,
+      "step": 3885
+    },
+    {
+      "epoch": 1.3731027179668196,
+      "grad_norm": 0.0318719539843843,
+      "learning_rate": 6.92603185956884e-05,
+      "loss": 2.0996,
+      "step": 3890
+    },
+    {
+      "epoch": 1.3748676314860573,
+      "grad_norm": 0.03496883372091561,
+      "learning_rate": 6.890387738166041e-05,
+      "loss": 1.9583,
+      "step": 3895
+    },
+    {
+      "epoch": 1.3766325450052947,
+      "grad_norm": 0.03505976503622303,
+      "learning_rate": 6.854808204958857e-05,
+      "loss": 1.9905,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3783974585245322,
+      "grad_norm": 0.035881630271846314,
+      "learning_rate": 6.819293543316925e-05,
+      "loss": 1.9506,
+      "step": 3905
+    },
+    {
+      "epoch": 1.38016237204377,
+      "grad_norm": 0.034924419367974774,
+      "learning_rate": 6.783844036093222e-05,
+      "loss": 1.9496,
+      "step": 3910
+    },
+    {
+      "epoch": 1.3819272855630074,
+      "grad_norm": 0.03475186352790621,
+      "learning_rate": 6.748459965621793e-05,
+      "loss": 2.0638,
+      "step": 3915
+    },
+    {
+      "epoch": 1.3836921990822448,
+      "grad_norm": 0.033898408666595395,
+      "learning_rate": 6.713141613715535e-05,
+      "loss": 1.8883,
+      "step": 3920
+    },
+    {
+      "epoch": 1.3854571126014825,
+      "grad_norm": 0.03502085144782307,
+      "learning_rate": 6.67788926166392e-05,
+      "loss": 2.0215,
+      "step": 3925
+    },
+    {
+      "epoch": 1.3872220261207202,
+      "grad_norm": 0.03333086617300456,
+      "learning_rate": 6.64270319023079e-05,
+      "loss": 2.0483,
+      "step": 3930
+    },
+    {
+      "epoch": 1.3889869396399577,
+      "grad_norm": 0.03589974377080066,
+      "learning_rate": 6.607583679652088e-05,
+      "loss": 1.8987,
+      "step": 3935
+    },
+    {
+      "epoch": 1.3907518531591951,
+      "grad_norm": 0.034050944674780946,
+      "learning_rate": 6.57253100963363e-05,
+      "loss": 1.8423,
+      "step": 3940
+    },
+    {
+      "epoch": 1.3925167666784328,
+      "grad_norm": 0.03480061387848505,
+      "learning_rate": 6.537545459348921e-05,
+      "loss": 1.9584,
+      "step": 3945
+    },
+    {
+      "epoch": 1.3942816801976703,
+      "grad_norm": 0.034558372317336446,
+      "learning_rate": 6.502627307436856e-05,
+      "loss": 2.0189,
+      "step": 3950
+    },
+    {
+      "epoch": 1.3960465937169078,
+      "grad_norm": 0.03401303820618363,
+      "learning_rate": 6.467776831999575e-05,
+      "loss": 1.898,
+      "step": 3955
+    },
+    {
+      "epoch": 1.3978115072361454,
+      "grad_norm": 0.03464719311174923,
+      "learning_rate": 6.432994310600208e-05,
+      "loss": 1.9088,
+      "step": 3960
+    },
+    {
+      "epoch": 1.399576420755383,
+      "grad_norm": 0.03484561221357701,
+      "learning_rate": 6.398280020260651e-05,
+      "loss": 1.9583,
+      "step": 3965
+    },
+    {
+      "epoch": 1.4013413342746206,
+      "grad_norm": 0.03446400293747441,
+      "learning_rate": 6.363634237459414e-05,
+      "loss": 1.8288,
+      "step": 3970
+    },
+    {
+      "epoch": 1.403106247793858,
+      "grad_norm": 0.03423945281481812,
+      "learning_rate": 6.329057238129362e-05,
+      "loss": 1.9538,
+      "step": 3975
+    },
+    {
+      "epoch": 1.4048711613130958,
+      "grad_norm": 0.03444048263380152,
+      "learning_rate": 6.294549297655538e-05,
+      "loss": 1.9491,
+      "step": 3980
+    },
+    {
+      "epoch": 1.4066360748323332,
+      "grad_norm": 0.03478853271551262,
+      "learning_rate": 6.260110690872995e-05,
+      "loss": 1.982,
+      "step": 3985
+    },
+    {
+      "epoch": 1.4084009883515707,
+      "grad_norm": 0.03407339866745072,
+      "learning_rate": 6.225741692064561e-05,
+      "loss": 2.0367,
+      "step": 3990
+    },
+    {
+      "epoch": 1.4101659018708084,
+      "grad_norm": 0.03310925378500857,
+      "learning_rate": 6.191442574958701e-05,
+      "loss": 1.9349,
+      "step": 3995
+    },
+    {
+      "epoch": 1.4119308153900458,
+      "grad_norm": 0.030211259163328414,
+      "learning_rate": 6.15721361272729e-05,
+      "loss": 1.9296,
+      "step": 4000
+    },
+    {
+      "epoch": 1.4136957289092835,
+      "grad_norm": 0.032131019426815836,
+      "learning_rate": 6.123055077983485e-05,
+      "loss": 1.9549,
+      "step": 4005
+    },
+    {
+      "epoch": 1.415460642428521,
+      "grad_norm": 0.02996955515322764,
+      "learning_rate": 6.088967242779506e-05,
+      "loss": 2.0469,
+      "step": 4010
+    },
+    {
+      "epoch": 1.4172255559477587,
+      "grad_norm": 0.030949648160448658,
+      "learning_rate": 6.0549503786045194e-05,
+      "loss": 1.9037,
+      "step": 4015
+    },
+    {
+      "epoch": 1.4189904694669961,
+      "grad_norm": 0.02962549902014697,
+      "learning_rate": 6.021004756382432e-05,
+      "loss": 1.7788,
+      "step": 4020
+    },
+    {
+      "epoch": 1.4207553829862336,
+      "grad_norm": 0.03127216284630837,
+      "learning_rate": 5.987130646469749e-05,
+      "loss": 1.9419,
+      "step": 4025
+    },
+    {
+      "epoch": 1.4225202965054713,
+      "grad_norm": 0.03210803533770451,
+      "learning_rate": 5.9533283186534424e-05,
+      "loss": 1.9051,
+      "step": 4030
+    },
+    {
+      "epoch": 1.4242852100247088,
+      "grad_norm": 0.03087351458876162,
+      "learning_rate": 5.91959804214876e-05,
+      "loss": 1.9196,
+      "step": 4035
+    },
+    {
+      "epoch": 1.4260501235439462,
+      "grad_norm": 0.030645441745246662,
+      "learning_rate": 5.88594008559712e-05,
+      "loss": 2.0086,
+      "step": 4040
+    },
+    {
+      "epoch": 1.427815037063184,
+      "grad_norm": 0.03629929969527182,
+      "learning_rate": 5.852354717063954e-05,
+      "loss": 1.9886,
+      "step": 4045
+    },
+    {
+      "epoch": 1.4295799505824216,
+      "grad_norm": 0.029703626987045696,
+      "learning_rate": 5.8188422040365626e-05,
+      "loss": 2.0076,
+      "step": 4050
+    },
+    {
+      "epoch": 1.431344864101659,
+      "grad_norm": 0.030105257577454053,
+      "learning_rate": 5.785402813422012e-05,
+      "loss": 1.8839,
+      "step": 4055
+    },
+    {
+      "epoch": 1.4331097776208965,
+      "grad_norm": 0.03261338616480212,
+      "learning_rate": 5.752036811544973e-05,
+      "loss": 2.1015,
+      "step": 4060
+    },
+    {
+      "epoch": 1.4348746911401342,
+      "grad_norm": 0.030698727898042278,
+      "learning_rate": 5.718744464145641e-05,
+      "loss": 2.109,
+      "step": 4065
+    },
+    {
+      "epoch": 1.4366396046593717,
+      "grad_norm": 0.030767091745341205,
+      "learning_rate": 5.685526036377582e-05,
+      "loss": 1.869,
+      "step": 4070
+    },
+    {
+      "epoch": 1.4384045181786091,
+      "grad_norm": 0.030075523350369757,
+      "learning_rate": 5.652381792805632e-05,
+      "loss": 2.0871,
+      "step": 4075
+    },
+    {
+      "epoch": 1.4401694316978468,
+      "grad_norm": 0.03102482240023936,
+      "learning_rate": 5.619311997403819e-05,
+      "loss": 2.0683,
+      "step": 4080
+    },
+    {
+      "epoch": 1.4419343452170843,
+      "grad_norm": 0.030590176064959243,
+      "learning_rate": 5.5863169135532074e-05,
+      "loss": 2.0001,
+      "step": 4085
+    },
+    {
+      "epoch": 1.443699258736322,
+      "grad_norm": 0.030473850542287392,
+      "learning_rate": 5.553396804039847e-05,
+      "loss": 1.8997,
+      "step": 4090
+    },
+    {
+      "epoch": 1.4454641722555595,
+      "grad_norm": 0.030529190063898,
+      "learning_rate": 5.5205519310526606e-05,
+      "loss": 1.9751,
+      "step": 4095
+    },
+    {
+      "epoch": 1.4472290857747971,
+      "grad_norm": 0.030011620191175203,
+      "learning_rate": 5.4877825561813474e-05,
+      "loss": 1.9433,
+      "step": 4100
+    },
+    {
+      "epoch": 1.4489939992940346,
+      "grad_norm": 0.02892031793205635,
+      "learning_rate": 5.4550889404143214e-05,
+      "loss": 1.7073,
+      "step": 4105
+    },
+    {
+      "epoch": 1.450758912813272,
+      "grad_norm": 0.029717790708532206,
+      "learning_rate": 5.422471344136611e-05,
+      "loss": 1.9333,
+      "step": 4110
+    },
+    {
+      "epoch": 1.4525238263325098,
+      "grad_norm": 0.02930954097067733,
+      "learning_rate": 5.3899300271277976e-05,
+      "loss": 1.9659,
+      "step": 4115
+    },
+    {
+      "epoch": 1.4542887398517472,
+      "grad_norm": 0.03000545192877199,
+      "learning_rate": 5.3574652485599485e-05,
+      "loss": 2.1037,
+      "step": 4120
+    },
+    {
+      "epoch": 1.4560536533709847,
+      "grad_norm": 0.031217563643557062,
+      "learning_rate": 5.325077266995541e-05,
+      "loss": 1.9793,
+      "step": 4125
+    },
+    {
+      "epoch": 1.4578185668902224,
+      "grad_norm": 0.03149908382304411,
+      "learning_rate": 5.2927663403854207e-05,
+      "loss": 2.0097,
+      "step": 4130
+    },
+    {
+      "epoch": 1.45958348040946,
+      "grad_norm": 0.029283485300683372,
+      "learning_rate": 5.260532726066723e-05,
+      "loss": 2.0911,
+      "step": 4135
+    },
+    {
+      "epoch": 1.4613483939286975,
+      "grad_norm": 0.031013403887131315,
+      "learning_rate": 5.2283766807608527e-05,
+      "loss": 2.1522,
+      "step": 4140
+    },
+    {
+      "epoch": 1.463113307447935,
+      "grad_norm": 0.029849094261517904,
+      "learning_rate": 5.196298460571406e-05,
+      "loss": 1.9244,
+      "step": 4145
+    },
+    {
+      "epoch": 1.4648782209671727,
+      "grad_norm": 0.031995056189921926,
+      "learning_rate": 5.164298320982169e-05,
+      "loss": 2.0307,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4666431344864101,
+      "grad_norm": 0.030535109935981897,
+      "learning_rate": 5.13237651685505e-05,
+      "loss": 2.075,
+      "step": 4155
+    },
+    {
+      "epoch": 1.4684080480056476,
+      "grad_norm": 0.029861382483526627,
+      "learning_rate": 5.100533302428056e-05,
+      "loss": 2.0797,
+      "step": 4160
+    },
+    {
+      "epoch": 1.4701729615248853,
+      "grad_norm": 0.03113422332401389,
+      "learning_rate": 5.068768931313293e-05,
+      "loss": 2.0316,
+      "step": 4165
+    },
+    {
+      "epoch": 1.4719378750441228,
+      "grad_norm": 0.030371057421138728,
+      "learning_rate": 5.037083656494909e-05,
+      "loss": 2.0134,
+      "step": 4170
+    },
+    {
+      "epoch": 1.4737027885633605,
+      "grad_norm": 0.030195355576018514,
+      "learning_rate": 5.005477730327112e-05,
+      "loss": 2.005,
+      "step": 4175
+    },
+    {
+      "epoch": 1.475467702082598,
+      "grad_norm": 0.031642143263772855,
+      "learning_rate": 4.97395140453214e-05,
+      "loss": 2.0198,
+      "step": 4180
+    },
+    {
+      "epoch": 1.4772326156018356,
+      "grad_norm": 0.030783067250674048,
+      "learning_rate": 4.942504930198254e-05,
+      "loss": 1.9722,
+      "step": 4185
+    },
+    {
+      "epoch": 1.478997529121073,
+      "grad_norm": 0.03147018031393304,
+      "learning_rate": 4.9111385577777644e-05,
+      "loss": 1.8712,
+      "step": 4190
+    },
+    {
+      "epoch": 1.4807624426403105,
+      "grad_norm": 0.02988972329672921,
+      "learning_rate": 4.8798525370849976e-05,
+      "loss": 1.7412,
+      "step": 4195
+    },
+    {
+      "epoch": 1.4825273561595482,
+      "grad_norm": 0.031335254889202926,
+      "learning_rate": 4.8486471172943336e-05,
+      "loss": 1.8953,
+      "step": 4200
+    },
+    {
+      "epoch": 1.4842922696787857,
+      "grad_norm": 0.04292576923872797,
+      "learning_rate": 4.8175225469382265e-05,
+      "loss": 1.9579,
+      "step": 4205
+    },
+    {
+      "epoch": 1.4860571831980234,
+      "grad_norm": 0.033075868159661063,
+      "learning_rate": 4.7864790739051936e-05,
+      "loss": 1.9982,
+      "step": 4210
+    },
+    {
+      "epoch": 1.4878220967172608,
+      "grad_norm": 0.03093078048486634,
+      "learning_rate": 4.755516945437878e-05,
+      "loss": 1.9502,
+      "step": 4215
+    },
+    {
+      "epoch": 1.4895870102364985,
+      "grad_norm": 0.032565005325328994,
+      "learning_rate": 4.724636408131048e-05,
+      "loss": 2.0328,
+      "step": 4220
+    },
+    {
+      "epoch": 1.491351923755736,
+      "grad_norm": 0.03382235424474912,
+      "learning_rate": 4.6938377079296577e-05,
+      "loss": 1.7977,
+      "step": 4225
+    },
+    {
+      "epoch": 1.4931168372749735,
+      "grad_norm": 0.03122599356098033,
+      "learning_rate": 4.663121090126878e-05,
+      "loss": 2.0124,
+      "step": 4230
+    },
+    {
+      "epoch": 1.4948817507942112,
+      "grad_norm": 0.03293457498046121,
+      "learning_rate": 4.632486799362133e-05,
+      "loss": 2.0137,
+      "step": 4235
+    },
+    {
+      "epoch": 1.4966466643134486,
+      "grad_norm": 0.03274112503582576,
+      "learning_rate": 4.60193507961917e-05,
+      "loss": 1.905,
+      "step": 4240
+    },
+    {
+      "epoch": 1.498411577832686,
+      "grad_norm": 0.03121910772804433,
+      "learning_rate": 4.5714661742240956e-05,
+      "loss": 1.7853,
+      "step": 4245
+    },
+    {
+      "epoch": 1.5001764913519238,
+      "grad_norm": 0.03127952953510779,
+      "learning_rate": 4.541080325843472e-05,
+      "loss": 1.996,
+      "step": 4250
+    },
+    {
+      "epoch": 1.5019414048711615,
+      "grad_norm": 0.03388019026652588,
+      "learning_rate": 4.51077777648234e-05,
+      "loss": 1.8462,
+      "step": 4255
+    },
+    {
+      "epoch": 1.503706318390399,
+      "grad_norm": 0.031815745456058475,
+      "learning_rate": 4.480558767482324e-05,
+      "loss": 1.9531,
+      "step": 4260
+    },
+    {
+      "epoch": 1.5054712319096364,
+      "grad_norm": 0.032317129873010504,
+      "learning_rate": 4.4504235395197054e-05,
+      "loss": 1.937,
+      "step": 4265
+    },
+    {
+      "epoch": 1.507236145428874,
+      "grad_norm": 0.03122000405064531,
+      "learning_rate": 4.420372332603483e-05,
+      "loss": 2.0452,
+      "step": 4270
+    },
+    {
+      "epoch": 1.5090010589481115,
+      "grad_norm": 0.03366683694943424,
+      "learning_rate": 4.390405386073497e-05,
+      "loss": 1.9186,
+      "step": 4275
+    },
+    {
+      "epoch": 1.510765972467349,
+      "grad_norm": 0.032210809084909024,
+      "learning_rate": 4.360522938598491e-05,
+      "loss": 1.9893,
+      "step": 4280
+    },
+    {
+      "epoch": 1.5125308859865867,
+      "grad_norm": 0.031648345467933534,
+      "learning_rate": 4.330725228174225e-05,
+      "loss": 1.9865,
+      "step": 4285
+    },
+    {
+      "epoch": 1.5142957995058244,
+      "grad_norm": 0.03320143068243922,
+      "learning_rate": 4.301012492121589e-05,
+      "loss": 1.7901,
+      "step": 4290
+    },
+    {
+      "epoch": 1.5160607130250616,
+      "grad_norm": 0.032575017922963696,
+      "learning_rate": 4.271384967084691e-05,
+      "loss": 1.8788,
+      "step": 4295
+    },
+    {
+      "epoch": 1.5178256265442993,
+      "grad_norm": 0.033807906266644455,
+      "learning_rate": 4.241842889028999e-05,
+      "loss": 1.8669,
+      "step": 4300
+    },
+    {
+      "epoch": 1.519590540063537,
+      "grad_norm": 0.03544941725879154,
+      "learning_rate": 4.212386493239428e-05,
+      "loss": 1.9998,
+      "step": 4305
+    },
+    {
+      "epoch": 1.5213554535827745,
+      "grad_norm": 0.03352425296231444,
+      "learning_rate": 4.183016014318501e-05,
+      "loss": 1.9388,
+      "step": 4310
+    },
+    {
+      "epoch": 1.523120367102012,
+      "grad_norm": 0.03429388603806833,
+      "learning_rate": 4.153731686184464e-05,
+      "loss": 1.9969,
+      "step": 4315
+    },
+    {
+      "epoch": 1.5248852806212496,
+      "grad_norm": 0.035745263271570446,
+      "learning_rate": 4.124533742069408e-05,
+      "loss": 1.8527,
+      "step": 4320
+    },
+    {
+      "epoch": 1.526650194140487,
+      "grad_norm": 0.03331654703180943,
+      "learning_rate": 4.0954224145174516e-05,
+      "loss": 1.8984,
+      "step": 4325
+    },
+    {
+      "epoch": 1.5284151076597245,
+      "grad_norm": 0.033161562192361986,
+      "learning_rate": 4.066397935382842e-05,
+      "loss": 1.9105,
+      "step": 4330
+    },
+    {
+      "epoch": 1.5301800211789622,
+      "grad_norm": 0.03420793235763417,
+      "learning_rate": 4.0374605358281376e-05,
+      "loss": 1.8085,
+      "step": 4335
+    },
+    {
+      "epoch": 1.5319449346982,
+      "grad_norm": 0.0351341442029064,
+      "learning_rate": 4.0086104463223704e-05,
+      "loss": 1.9777,
+      "step": 4340
+    },
+    {
+      "epoch": 1.5337098482174374,
+      "grad_norm": 0.03518965728758648,
+      "learning_rate": 3.9798478966391845e-05,
+      "loss": 1.8944,
+      "step": 4345
+    },
+    {
+      "epoch": 1.5354747617366749,
+      "grad_norm": 0.03549401557703974,
+      "learning_rate": 3.9511731158550366e-05,
+      "loss": 2.0236,
+      "step": 4350
+    },
+    {
+      "epoch": 1.5372396752559125,
+      "grad_norm": 0.03624716728232057,
+      "learning_rate": 3.922586332347343e-05,
+      "loss": 2.0764,
+      "step": 4355
+    },
+    {
+      "epoch": 1.53900458877515,
+      "grad_norm": 0.03614978489396938,
+      "learning_rate": 3.894087773792686e-05,
+      "loss": 1.8375,
+      "step": 4360
+    },
+    {
+      "epoch": 1.5407695022943875,
+      "grad_norm": 0.03480835954719979,
+      "learning_rate": 3.8656776671649856e-05,
+      "loss": 1.8728,
+      "step": 4365
+    },
+    {
+      "epoch": 1.5425344158136252,
+      "grad_norm": 0.0351064356806685,
+      "learning_rate": 3.837356238733692e-05,
+      "loss": 1.976,
+      "step": 4370
+    },
+    {
+      "epoch": 1.5442993293328628,
+      "grad_norm": 0.03436045015342909,
+      "learning_rate": 3.80912371406199e-05,
+      "loss": 1.8648,
+      "step": 4375
+    },
+    {
+      "epoch": 1.5460642428521,
+      "grad_norm": 0.035825034074144994,
+      "learning_rate": 3.780980318004991e-05,
+      "loss": 1.9092,
+      "step": 4380
+    },
+    {
+      "epoch": 1.5478291563713378,
+      "grad_norm": 0.03618090985329705,
+      "learning_rate": 3.752926274707966e-05,
+      "loss": 2.0864,
+      "step": 4385
+    },
+    {
+      "epoch": 1.5495940698905755,
+      "grad_norm": 0.03382552946679882,
+      "learning_rate": 3.724961807604531e-05,
+      "loss": 1.9425,
+      "step": 4390
+    },
+    {
+      "epoch": 1.551358983409813,
+      "grad_norm": 0.03484827805342798,
+      "learning_rate": 3.6970871394148895e-05,
+      "loss": 2.101,
+      "step": 4395
+    },
+    {
+      "epoch": 1.5531238969290504,
+      "grad_norm": 0.03320640448986934,
+      "learning_rate": 3.669302492144052e-05,
+      "loss": 2.0688,
+      "step": 4400
+    },
+    {
+      "epoch": 1.554888810448288,
+      "grad_norm": 0.03484038306740596,
+      "learning_rate": 3.6416080870800547e-05,
+      "loss": 1.9454,
+      "step": 4405
+    },
+    {
+      "epoch": 1.5566537239675255,
+      "grad_norm": 0.035197623913338666,
+      "learning_rate": 3.6140041447922225e-05,
+      "loss": 1.7624,
+      "step": 4410
+    },
+    {
+      "epoch": 1.558418637486763,
+      "grad_norm": 0.03609254311952255,
+      "learning_rate": 3.5864908851293914e-05,
+      "loss": 2.1283,
+      "step": 4415
+    },
+    {
+      "epoch": 1.5601835510060007,
+      "grad_norm": 0.03481377536297356,
+      "learning_rate": 3.559068527218157e-05,
+      "loss": 1.8186,
+      "step": 4420
+    },
+    {
+      "epoch": 1.5619484645252384,
+      "grad_norm": 0.03396508159901674,
+      "learning_rate": 3.531737289461159e-05,
+      "loss": 1.9517,
+      "step": 4425
+    },
+    {
+      "epoch": 1.5637133780444759,
+      "grad_norm": 0.03349459424542829,
+      "learning_rate": 3.504497389535296e-05,
+      "loss": 1.8817,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5654782915637133,
+      "grad_norm": 0.03354989744772323,
+      "learning_rate": 3.477349044390034e-05,
+      "loss": 1.8772,
+      "step": 4435
+    },
+    {
+      "epoch": 1.567243205082951,
+      "grad_norm": 0.03453484301292485,
+      "learning_rate": 3.450292470245645e-05,
+      "loss": 1.9036,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5690081186021885,
+      "grad_norm": 0.03411400764399191,
+      "learning_rate": 3.42332788259151e-05,
+      "loss": 1.9382,
+      "step": 4445
+    },
+    {
+      "epoch": 1.570773032121426,
+      "grad_norm": 0.033435228514182735,
+      "learning_rate": 3.3964554961843944e-05,
+      "loss": 2.0641,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5725379456406636,
+      "grad_norm": 0.03367024468091004,
+      "learning_rate": 3.369675525046727e-05,
+      "loss": 1.9334,
+      "step": 4455
+    },
+    {
+      "epoch": 1.5743028591599013,
+      "grad_norm": 0.03433997274475782,
+      "learning_rate": 3.342988182464911e-05,
+      "loss": 2.0534,
+      "step": 4460
+    },
+    {
+      "epoch": 1.5760677726791386,
+      "grad_norm": 0.0350617920784125,
+      "learning_rate": 3.316393680987612e-05,
+      "loss": 1.8423,
+      "step": 4465
+    },
+    {
+      "epoch": 1.5778326861983762,
+      "grad_norm": 0.03481307165141278,
+      "learning_rate": 3.289892232424079e-05,
+      "loss": 2.0224,
+      "step": 4470
+    },
+    {
+      "epoch": 1.579597599717614,
+      "grad_norm": 0.03428922222495566,
+      "learning_rate": 3.2634840478424535e-05,
+      "loss": 1.7084,
+      "step": 4475
+    },
+    {
+      "epoch": 1.5813625132368514,
+      "grad_norm": 0.033417882269180514,
+      "learning_rate": 3.237169337568071e-05,
+      "loss": 2.0756,
+      "step": 4480
+    },
+    {
+      "epoch": 1.5831274267560889,
+      "grad_norm": 0.033803883912114696,
+      "learning_rate": 3.210948311181818e-05,
+      "loss": 1.9036,
+      "step": 4485
+    },
+    {
+      "epoch": 1.5848923402753265,
+      "grad_norm": 0.03634601411926046,
+      "learning_rate": 3.184821177518428e-05,
+      "loss": 1.9282,
+      "step": 4490
+    },
+    {
+      "epoch": 1.586657253794564,
+      "grad_norm": 0.033508534289896266,
+      "learning_rate": 3.1587881446648514e-05,
+      "loss": 1.954,
+      "step": 4495
+    },
+    {
+      "epoch": 1.5884221673138015,
+      "grad_norm": 0.035590160411160615,
+      "learning_rate": 3.13284941995857e-05,
+      "loss": 1.9688,
+      "step": 4500
+    },
+    {
+      "epoch": 1.5901870808330392,
+      "grad_norm": 0.03265422079734215,
+      "learning_rate": 3.107005209985958e-05,
+      "loss": 2.0097,
+      "step": 4505
+    },
+    {
+      "epoch": 1.5919519943522769,
+      "grad_norm": 0.03296117328913444,
+      "learning_rate": 3.0812557205806475e-05,
+      "loss": 1.9257,
+      "step": 4510
+    },
+    {
+      "epoch": 1.5937169078715143,
+      "grad_norm": 0.034587592714518715,
+      "learning_rate": 3.055601156821865e-05,
+      "loss": 1.8154,
+      "step": 4515
+    },
+    {
+      "epoch": 1.5954818213907518,
+      "grad_norm": 0.03376921264284581,
+      "learning_rate": 3.0300417230328244e-05,
+      "loss": 1.9254,
+      "step": 4520
+    },
+    {
+      "epoch": 1.5972467349099895,
+      "grad_norm": 0.0343184167137067,
+      "learning_rate": 3.0045776227790697e-05,
+      "loss": 2.0497,
+      "step": 4525
+    },
+    {
+      "epoch": 1.599011648429227,
+      "grad_norm": 0.03622464562386491,
+      "learning_rate": 2.9792090588668876e-05,
+      "loss": 1.9846,
+      "step": 4530
+    },
+    {
+      "epoch": 1.6007765619484644,
+      "grad_norm": 0.03575303468534114,
+      "learning_rate": 2.953936233341667e-05,
+      "loss": 1.8907,
+      "step": 4535
+    },
+    {
+      "epoch": 1.602541475467702,
+      "grad_norm": 0.032723961319037265,
+      "learning_rate": 2.9287593474862986e-05,
+      "loss": 2.0426,
+      "step": 4540
+    },
+    {
+      "epoch": 1.6043063889869398,
+      "grad_norm": 0.03377216791919303,
+      "learning_rate": 2.9036786018195668e-05,
+      "loss": 2.0135,
+      "step": 4545
+    },
+    {
+      "epoch": 1.6060713025061772,
+      "grad_norm": 0.03208908689321701,
+      "learning_rate": 2.8786941960945674e-05,
+      "loss": 1.9142,
+      "step": 4550
+    },
+    {
+      "epoch": 1.6078362160254147,
+      "grad_norm": 0.03308820380791096,
+      "learning_rate": 2.8538063292970935e-05,
+      "loss": 1.9797,
+      "step": 4555
+    },
+    {
+      "epoch": 1.6096011295446524,
+      "grad_norm": 0.032953914742919574,
+      "learning_rate": 2.829015199644078e-05,
+      "loss": 1.9388,
+      "step": 4560
+    },
+    {
+      "epoch": 1.6113660430638899,
+      "grad_norm": 0.03240639314147682,
+      "learning_rate": 2.8043210045819865e-05,
+      "loss": 1.965,
+      "step": 4565
+    },
+    {
+      "epoch": 1.6131309565831273,
+      "grad_norm": 0.03187279733830787,
+      "learning_rate": 2.7797239407852722e-05,
+      "loss": 1.8038,
+      "step": 4570
+    },
+    {
+      "epoch": 1.614895870102365,
+      "grad_norm": 0.03500866531045439,
+      "learning_rate": 2.755224204154783e-05,
+      "loss": 1.9147,
+      "step": 4575
+    },
+    {
+      "epoch": 1.6166607836216027,
+      "grad_norm": 0.03266894651679398,
+      "learning_rate": 2.730821989816222e-05,
+      "loss": 2.055,
+      "step": 4580
+    },
+    {
+      "epoch": 1.61842569714084,
+      "grad_norm": 0.0320641736799915,
+      "learning_rate": 2.70651749211859e-05,
+      "loss": 1.9506,
+      "step": 4585
+    },
+    {
+      "epoch": 1.6201906106600776,
+      "grad_norm": 0.03369471922644863,
+      "learning_rate": 2.6823109046326213e-05,
+      "loss": 2.0466,
+      "step": 4590
+    },
+    {
+      "epoch": 1.6219555241793153,
+      "grad_norm": 0.03244818698525006,
+      "learning_rate": 2.6582024201492624e-05,
+      "loss": 1.9494,
+      "step": 4595
+    },
+    {
+      "epoch": 1.6237204376985528,
+      "grad_norm": 0.03364220990923876,
+      "learning_rate": 2.634192230678123e-05,
+      "loss": 2.055,
+      "step": 4600
+    },
+    {
+      "epoch": 1.6254853512177903,
+      "grad_norm": 0.03559180939610663,
+      "learning_rate": 2.610280527445958e-05,
+      "loss": 1.9353,
+      "step": 4605
+    },
+    {
+      "epoch": 1.627250264737028,
+      "grad_norm": 0.03374331211309215,
+      "learning_rate": 2.586467500895139e-05,
+      "loss": 1.9115,
+      "step": 4610
+    },
+    {
+      "epoch": 1.6290151782562654,
+      "grad_norm": 0.03427240045585054,
+      "learning_rate": 2.5627533406821283e-05,
+      "loss": 1.9418,
+      "step": 4615
+    },
+    {
+      "epoch": 1.6307800917755029,
+      "grad_norm": 0.03484851275376693,
+      "learning_rate": 2.5391382356759875e-05,
+      "loss": 1.8673,
+      "step": 4620
+    },
+    {
+      "epoch": 1.6325450052947406,
+      "grad_norm": 0.033056791723209876,
+      "learning_rate": 2.5156223739568537e-05,
+      "loss": 2.0146,
+      "step": 4625
+    },
+    {
+      "epoch": 1.6343099188139782,
+      "grad_norm": 0.032684067257381516,
+      "learning_rate": 2.49220594281446e-05,
+      "loss": 1.9037,
+      "step": 4630
+    },
+    {
+      "epoch": 1.6360748323332157,
+      "grad_norm": 0.03524316145001109,
+      "learning_rate": 2.4688891287466278e-05,
+      "loss": 1.9452,
+      "step": 4635
+    },
+    {
+      "epoch": 1.6378397458524532,
+      "grad_norm": 0.03267041417057933,
+      "learning_rate": 2.4456721174577864e-05,
+      "loss": 2.0948,
+      "step": 4640
+    },
+    {
+      "epoch": 1.6396046593716909,
+      "grad_norm": 0.03305088136172108,
+      "learning_rate": 2.4225550938575054e-05,
+      "loss": 1.9246,
+      "step": 4645
+    },
+    {
+      "epoch": 1.6413695728909283,
+      "grad_norm": 0.033692619829589356,
+      "learning_rate": 2.3995382420590014e-05,
+      "loss": 2.0536,
+      "step": 4650
+    },
+    {
+      "epoch": 1.6431344864101658,
+      "grad_norm": 0.03367884203170107,
+      "learning_rate": 2.3766217453776935e-05,
+      "loss": 1.982,
+      "step": 4655
+    },
+    {
+      "epoch": 1.6448993999294035,
+      "grad_norm": 0.03360602357947156,
+      "learning_rate": 2.3538057863297227e-05,
+      "loss": 1.9125,
+      "step": 4660
+    },
+    {
+      "epoch": 1.6466643134486412,
+      "grad_norm": 0.03295215993407256,
+      "learning_rate": 2.331090546630513e-05,
+      "loss": 1.7465,
+      "step": 4665
+    },
+    {
+      "epoch": 1.6484292269678784,
+      "grad_norm": 0.03510990884133024,
+      "learning_rate": 2.308476207193324e-05,
+      "loss": 1.9727,
+      "step": 4670
+    },
+    {
+      "epoch": 1.650194140487116,
+      "grad_norm": 0.03249544603841909,
+      "learning_rate": 2.2859629481277957e-05,
+      "loss": 1.7661,
+      "step": 4675
+    },
+    {
+      "epoch": 1.6519590540063538,
+      "grad_norm": 0.033302650260135465,
+      "learning_rate": 2.2635509487385284e-05,
+      "loss": 1.8944,
+      "step": 4680
+    },
+    {
+      "epoch": 1.6537239675255913,
+      "grad_norm": 0.03358669674623887,
+      "learning_rate": 2.2412403875236445e-05,
+      "loss": 2.0577,
+      "step": 4685
+    },
+    {
+      "epoch": 1.6554888810448287,
+      "grad_norm": 0.0332853305537157,
+      "learning_rate": 2.2190314421733816e-05,
+      "loss": 1.8156,
+      "step": 4690
+    },
+    {
+      "epoch": 1.6572537945640664,
+      "grad_norm": 0.03399706837086618,
+      "learning_rate": 2.1969242895686684e-05,
+      "loss": 1.9527,
+      "step": 4695
+    },
+    {
+      "epoch": 1.6590187080833039,
+      "grad_norm": 0.034468055413874314,
+      "learning_rate": 2.1749191057796995e-05,
+      "loss": 2.1008,
+      "step": 4700
+    },
+    {
+      "epoch": 1.6607836216025413,
+      "grad_norm": 0.03808073137337732,
+      "learning_rate": 2.1530160660645695e-05,
+      "loss": 2.1552,
+      "step": 4705
+    },
+    {
+      "epoch": 1.662548535121779,
+      "grad_norm": 0.03717021529550901,
+      "learning_rate": 2.131215344867841e-05,
+      "loss": 1.8806,
+      "step": 4710
+    },
+    {
+      "epoch": 1.6643134486410167,
+      "grad_norm": 0.03482861043963838,
+      "learning_rate": 2.109517115819181e-05,
+      "loss": 2.0962,
+      "step": 4715
+    },
+    {
+      "epoch": 1.6660783621602542,
+      "grad_norm": 0.037454099036786793,
+      "learning_rate": 2.0879215517319632e-05,
+      "loss": 1.9126,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6678432756794916,
+      "grad_norm": 0.03761764890590753,
+      "learning_rate": 2.066428824601892e-05,
+      "loss": 1.8034,
+      "step": 4725
+    },
+    {
+      "epoch": 1.6696081891987293,
+      "grad_norm": 0.03777633028348587,
+      "learning_rate": 2.0450391056056514e-05,
+      "loss": 1.8725,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6713731027179668,
+      "grad_norm": 0.0383366262825412,
+      "learning_rate": 2.023752565099513e-05,
+      "loss": 1.8807,
+      "step": 4735
+    },
+    {
+      "epoch": 1.6731380162372043,
+      "grad_norm": 0.041395627558972546,
+      "learning_rate": 2.0025693726179992e-05,
+      "loss": 1.8714,
+      "step": 4740
+    },
+    {
+      "epoch": 1.674902929756442,
+      "grad_norm": 0.03776877520097343,
+      "learning_rate": 1.981489696872532e-05,
+      "loss": 1.981,
+      "step": 4745
+    },
+    {
+      "epoch": 1.6766678432756796,
+      "grad_norm": 0.03864560142228677,
+      "learning_rate": 1.960513705750075e-05,
+      "loss": 2.0369,
+      "step": 4750
+    },
+    {
+      "epoch": 1.678432756794917,
+      "grad_norm": 0.03660753652499858,
+      "learning_rate": 1.9396415663118158e-05,
+      "loss": 2.1063,
+      "step": 4755
+    },
+    {
+      "epoch": 1.6801976703141546,
+      "grad_norm": 0.03794168690051479,
+      "learning_rate": 1.9188734447918203e-05,
+      "loss": 2.1145,
+      "step": 4760
+    },
+    {
+      "epoch": 1.6819625838333923,
+      "grad_norm": 0.03689174548789547,
+      "learning_rate": 1.8982095065957126e-05,
+      "loss": 2.0411,
+      "step": 4765
+    },
+    {
+      "epoch": 1.6837274973526297,
+      "grad_norm": 0.038791143862949254,
+      "learning_rate": 1.877649916299359e-05,
+      "loss": 2.0615,
+      "step": 4770
+    },
+    {
+      "epoch": 1.6854924108718672,
+      "grad_norm": 0.03794446708599632,
+      "learning_rate": 1.857194837647561e-05,
+      "loss": 1.9214,
+      "step": 4775
+    },
+    {
+      "epoch": 1.6872573243911049,
+      "grad_norm": 0.0355230849136272,
+      "learning_rate": 1.8368444335527534e-05,
+      "loss": 1.6658,
+      "step": 4780
+    },
+    {
+      "epoch": 1.6890222379103423,
+      "grad_norm": 0.03874174681361104,
+      "learning_rate": 1.8165988660936847e-05,
+      "loss": 1.9385,
+      "step": 4785
+    },
+    {
+      "epoch": 1.6907871514295798,
+      "grad_norm": 0.03677305174678137,
+      "learning_rate": 1.796458296514159e-05,
+      "loss": 2.0329,
+      "step": 4790
+    },
+    {
+      "epoch": 1.6925520649488175,
+      "grad_norm": 0.03788140831824028,
+      "learning_rate": 1.776422885221721e-05,
+      "loss": 1.9731,
+      "step": 4795
+    },
+    {
+      "epoch": 1.6943169784680552,
+      "grad_norm": 0.03571226468199232,
+      "learning_rate": 1.7564927917864037e-05,
+      "loss": 2.0341,
+      "step": 4800
+    },
+    {
+      "epoch": 1.6960818919872926,
+      "grad_norm": 0.035248073343293045,
+      "learning_rate": 1.7366681749394445e-05,
+      "loss": 2.09,
+      "step": 4805
+    },
+    {
+      "epoch": 1.69784680550653,
+      "grad_norm": 0.035564067795887826,
+      "learning_rate": 1.716949192572019e-05,
+      "loss": 2.1418,
+      "step": 4810
+    },
+    {
+      "epoch": 1.6996117190257678,
+      "grad_norm": 0.03388794299753511,
+      "learning_rate": 1.697336001733989e-05,
+      "loss": 2.0008,
+      "step": 4815
+    },
+    {
+      "epoch": 1.7013766325450053,
+      "grad_norm": 0.035929992822541924,
+      "learning_rate": 1.677828758632649e-05,
+      "loss": 2.0369,
+      "step": 4820
+    },
+    {
+      "epoch": 1.7031415460642427,
+      "grad_norm": 0.034283244845831146,
+      "learning_rate": 1.6584276186314865e-05,
+      "loss": 1.9419,
+      "step": 4825
+    },
+    {
+      "epoch": 1.7049064595834804,
+      "grad_norm": 0.03400790973735181,
+      "learning_rate": 1.6391327362489448e-05,
+      "loss": 2.0578,
+      "step": 4830
+    },
+    {
+      "epoch": 1.706671373102718,
+      "grad_norm": 0.03509175665870161,
+      "learning_rate": 1.6199442651571788e-05,
+      "loss": 1.8902,
+      "step": 4835
+    },
+    {
+      "epoch": 1.7084362866219556,
+      "grad_norm": 0.03508523956425158,
+      "learning_rate": 1.6008623581808528e-05,
+      "loss": 2.0424,
+      "step": 4840
+    },
+    {
+      "epoch": 1.710201200141193,
+      "grad_norm": 0.0336206368964309,
+      "learning_rate": 1.581887167295899e-05,
+      "loss": 1.8735,
+      "step": 4845
+    },
+    {
+      "epoch": 1.7119661136604307,
+      "grad_norm": 0.03501249923546285,
+      "learning_rate": 1.5630188436283336e-05,
+      "loss": 2.0012,
+      "step": 4850
+    },
+    {
+      "epoch": 1.7137310271796682,
+      "grad_norm": 0.03434071903450408,
+      "learning_rate": 1.5442575374530297e-05,
+      "loss": 1.9999,
+      "step": 4855
+    },
+    {
+      "epoch": 1.7154959406989057,
+      "grad_norm": 0.03373694298322959,
+      "learning_rate": 1.5256033981925297e-05,
+      "loss": 1.946,
+      "step": 4860
+    },
+    {
+      "epoch": 1.7172608542181433,
+      "grad_norm": 0.03391461878353098,
+      "learning_rate": 1.5070565744158692e-05,
+      "loss": 1.9577,
+      "step": 4865
+    },
+    {
+      "epoch": 1.719025767737381,
+      "grad_norm": 0.03550827603229797,
+      "learning_rate": 1.4886172138373603e-05,
+      "loss": 2.0624,
+      "step": 4870
+    },
+    {
+      "epoch": 1.7207906812566183,
+      "grad_norm": 0.03438548046525696,
+      "learning_rate": 1.4702854633154521e-05,
+      "loss": 1.753,
+      "step": 4875
+    },
+    {
+      "epoch": 1.722555594775856,
+      "grad_norm": 0.03642489961359046,
+      "learning_rate": 1.4520614688515403e-05,
+      "loss": 1.984,
+      "step": 4880
+    },
+    {
+      "epoch": 1.7243205082950936,
+      "grad_norm": 0.03492018291854148,
+      "learning_rate": 1.4339453755887976e-05,
+      "loss": 1.8575,
+      "step": 4885
+    },
+    {
+      "epoch": 1.7260854218143311,
+      "grad_norm": 0.033482674089617574,
+      "learning_rate": 1.4159373278110452e-05,
+      "loss": 1.8484,
+      "step": 4890
+    },
+    {
+      "epoch": 1.7278503353335686,
+      "grad_norm": 0.03445439341727427,
+      "learning_rate": 1.3980374689415735e-05,
+      "loss": 1.7615,
+      "step": 4895
+    },
+    {
+      "epoch": 1.7296152488528063,
+      "grad_norm": 0.03472204442938287,
+      "learning_rate": 1.3802459415420197e-05,
+      "loss": 1.8561,
+      "step": 4900
+    },
+    {
+      "epoch": 1.7313801623720437,
+      "grad_norm": 0.03535791081181799,
+      "learning_rate": 1.3625628873112182e-05,
+      "loss": 2.0643,
+      "step": 4905
+    },
+    {
+      "epoch": 1.7331450758912812,
+      "grad_norm": 0.0349559723312278,
+      "learning_rate": 1.344988447084091e-05,
+      "loss": 1.9739,
+      "step": 4910
+    },
+    {
+      "epoch": 1.7349099894105189,
+      "grad_norm": 0.0344026445785495,
+      "learning_rate": 1.3275227608305122e-05,
+      "loss": 1.9237,
+      "step": 4915
+    },
+    {
+      "epoch": 1.7366749029297566,
+      "grad_norm": 0.03523023241778303,
+      "learning_rate": 1.3101659676541837e-05,
+      "loss": 1.998,
+      "step": 4920
+    },
+    {
+      "epoch": 1.738439816448994,
+      "grad_norm": 0.037054044254085525,
+      "learning_rate": 1.2929182057915577e-05,
+      "loss": 1.9682,
+      "step": 4925
+    },
+    {
+      "epoch": 1.7402047299682315,
+      "grad_norm": 0.03607678344856152,
+      "learning_rate": 1.2757796126107028e-05,
+      "loss": 1.8698,
+      "step": 4930
+    },
+    {
+      "epoch": 1.7419696434874692,
+      "grad_norm": 0.03461436106601157,
+      "learning_rate": 1.2587503246102332e-05,
+      "loss": 1.9138,
+      "step": 4935
+    },
+    {
+      "epoch": 1.7437345570067067,
+      "grad_norm": 0.03539240913339504,
+      "learning_rate": 1.2418304774182075e-05,
+      "loss": 1.7903,
+      "step": 4940
+    },
+    {
+      "epoch": 1.7454994705259441,
+      "grad_norm": 0.0353244956045996,
+      "learning_rate": 1.2250202057910535e-05,
+      "loss": 1.9706,
+      "step": 4945
+    },
+    {
+      "epoch": 1.7472643840451818,
+      "grad_norm": 0.03474009794468743,
+      "learning_rate": 1.2083196436125037e-05,
+      "loss": 1.9454,
+      "step": 4950
+    },
+    {
+      "epoch": 1.7490292975644195,
+      "grad_norm": 0.034766879117241446,
+      "learning_rate": 1.1917289238925092e-05,
+      "loss": 1.937,
+      "step": 4955
+    },
+    {
+      "epoch": 1.7507942110836567,
+      "grad_norm": 0.03418572125679393,
+      "learning_rate": 1.1752481787662015e-05,
+      "loss": 1.867,
+      "step": 4960
+    },
+    {
+      "epoch": 1.7525591246028944,
+      "grad_norm": 0.03534656952176172,
+      "learning_rate": 1.1588775394928273e-05,
+      "loss": 2.0516,
+      "step": 4965
+    },
+    {
+      "epoch": 1.7543240381221321,
+      "grad_norm": 0.035690207687914664,
+      "learning_rate": 1.1426171364547015e-05,
+      "loss": 1.9852,
+      "step": 4970
+    },
+    {
+      "epoch": 1.7560889516413696,
+      "grad_norm": 0.035577783905641425,
+      "learning_rate": 1.1264670991561813e-05,
+      "loss": 2.0095,
+      "step": 4975
+    },
+    {
+      "epoch": 1.757853865160607,
+      "grad_norm": 0.034858555694979386,
+      "learning_rate": 1.110427556222624e-05,
+      "loss": 1.9413,
+      "step": 4980
+    },
+    {
+      "epoch": 1.7596187786798447,
+      "grad_norm": 0.03539053113137391,
+      "learning_rate": 1.0944986353993584e-05,
+      "loss": 1.9451,
+      "step": 4985
+    },
+    {
+      "epoch": 1.7613836921990822,
+      "grad_norm": 0.03726042259223971,
+      "learning_rate": 1.0786804635506874e-05,
+      "loss": 1.9751,
+      "step": 4990
+    },
+    {
+      "epoch": 1.7631486057183197,
+      "grad_norm": 0.03557920147425404,
+      "learning_rate": 1.062973166658852e-05,
+      "loss": 2.0106,
+      "step": 4995
+    },
+    {
+      "epoch": 1.7649135192375573,
+      "grad_norm": 0.03372002449364942,
+      "learning_rate": 1.047376869823055e-05,
+      "loss": 1.8719,
+      "step": 5000
+    },
+    {
+      "epoch": 1.7649135192375573,
+      "eval_loss": 1.9466089010238647,
+      "eval_runtime": 5.9863,
+      "eval_samples_per_second": 30.403,
+      "eval_steps_per_second": 2.005,
+      "step": 5000
+    },
+    {
+      "epoch": 1.766678432756795,
+      "grad_norm": 0.03388234508287682,
+      "learning_rate": 1.0318916972584363e-05,
+      "loss": 1.9646,
+      "step": 5005
+    },
+    {
+      "epoch": 1.7684433462760325,
+      "grad_norm": 0.03173844540441306,
+      "learning_rate": 1.0165177722951051e-05,
+      "loss": 2.0326,
+      "step": 5010
+    },
+    {
+      "epoch": 1.77020825979527,
+      "grad_norm": 0.03342440602825524,
+      "learning_rate": 1.0012552173771515e-05,
+      "loss": 1.9593,
+      "step": 5015
+    },
+    {
+      "epoch": 1.7719731733145077,
+      "grad_norm": 0.032979318906046794,
+      "learning_rate": 9.861041540616666e-06,
+      "loss": 2.109,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7737380868337451,
+      "grad_norm": 0.032750277229710106,
+      "learning_rate": 9.710647030177732e-06,
+      "loss": 1.9195,
+      "step": 5025
+    },
+    {
+      "epoch": 1.7755030003529826,
+      "grad_norm": 0.034019726750770585,
+      "learning_rate": 9.561369840256822e-06,
+      "loss": 2.0177,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7772679138722203,
+      "grad_norm": 0.034567983065093455,
+      "learning_rate": 9.413211159757144e-06,
+      "loss": 1.8478,
+      "step": 5035
+    },
+    {
+      "epoch": 1.779032827391458,
+      "grad_norm": 0.032952913636075415,
+      "learning_rate": 9.266172168673668e-06,
+      "loss": 1.893,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7807977409106954,
+      "grad_norm": 0.03456253901205636,
+      "learning_rate": 9.120254038083762e-06,
+      "loss": 1.9549,
+      "step": 5045
+    },
+    {
+      "epoch": 1.782562654429933,
+      "grad_norm": 0.031152620981820957,
+      "learning_rate": 8.975457930137786e-06,
+      "loss": 1.7915,
+      "step": 5050
+    },
+    {
+      "epoch": 1.7843275679491706,
+      "grad_norm": 0.03186519152464313,
+      "learning_rate": 8.831784998049828e-06,
+      "loss": 1.8485,
+      "step": 5055
+    },
+    {
+      "epoch": 1.786092481468408,
+      "grad_norm": 0.032729806647655305,
+      "learning_rate": 8.689236386088622e-06,
+      "loss": 1.9637,
+      "step": 5060
+    },
+    {
+      "epoch": 1.7878573949876455,
+      "grad_norm": 0.0329128133725629,
+      "learning_rate": 8.547813229568274e-06,
+      "loss": 1.8866,
+      "step": 5065
+    },
+    {
+      "epoch": 1.7896223085068832,
+      "grad_norm": 0.032989449010666604,
+      "learning_rate": 8.407516654839431e-06,
+      "loss": 2.1032,
+      "step": 5070
+    },
+    {
+      "epoch": 1.7913872220261209,
+      "grad_norm": 0.032982273579494535,
+      "learning_rate": 8.268347779280081e-06,
+      "loss": 1.9345,
+      "step": 5075
+    },
+    {
+      "epoch": 1.7931521355453581,
+      "grad_norm": 0.031480972712878474,
+      "learning_rate": 8.130307711286805e-06,
+      "loss": 1.9074,
+      "step": 5080
+    },
+    {
+      "epoch": 1.7949170490645958,
+      "grad_norm": 0.03285840220088487,
+      "learning_rate": 7.993397550265963e-06,
+      "loss": 1.9735,
+      "step": 5085
+    },
+    {
+      "epoch": 1.7966819625838335,
+      "grad_norm": 0.03301267055686261,
+      "learning_rate": 7.857618386624787e-06,
+      "loss": 1.8525,
+      "step": 5090
+    },
+    {
+      "epoch": 1.798446876103071,
+      "grad_norm": 0.03229186047709807,
+      "learning_rate": 7.722971301762881e-06,
+      "loss": 2.0201,
+      "step": 5095
+    },
+    {
+      "epoch": 1.8002117896223084,
+      "grad_norm": 0.03311208018629292,
+      "learning_rate": 7.589457368063462e-06,
+      "loss": 1.9816,
+      "step": 5100
+    },
+    {
+      "epoch": 1.8019767031415461,
+      "grad_norm": 0.033904337774243884,
+      "learning_rate": 7.457077648884913e-06,
+      "loss": 2.0992,
+      "step": 5105
+    },
+    {
+      "epoch": 1.8037416166607836,
+      "grad_norm": 0.03283954085938039,
+      "learning_rate": 7.325833198552261e-06,
+      "loss": 1.8921,
+      "step": 5110
+    },
+    {
+      "epoch": 1.805506530180021,
+      "grad_norm": 0.03321362379399172,
+      "learning_rate": 7.195725062348801e-06,
+      "loss": 1.9042,
+      "step": 5115
+    },
+    {
+      "epoch": 1.8072714436992587,
+      "grad_norm": 0.03372338045578344,
+      "learning_rate": 7.06675427650773e-06,
+      "loss": 1.9033,
+      "step": 5120
+    },
+    {
+      "epoch": 1.8090363572184964,
+      "grad_norm": 0.03182286061460464,
+      "learning_rate": 6.938921868204045e-06,
+      "loss": 1.9908,
+      "step": 5125
+    },
+    {
+      "epoch": 1.810801270737734,
+      "grad_norm": 0.034275817692474446,
+      "learning_rate": 6.812228855546109e-06,
+      "loss": 1.8681,
+      "step": 5130
+    },
+    {
+      "epoch": 1.8125661842569714,
+      "grad_norm": 0.03270753521457761,
+      "learning_rate": 6.686676247567796e-06,
+      "loss": 1.9336,
+      "step": 5135
+    },
+    {
+      "epoch": 1.814331097776209,
+      "grad_norm": 0.0339513917580001,
+      "learning_rate": 6.562265044220244e-06,
+      "loss": 1.9743,
+      "step": 5140
+    },
+    {
+      "epoch": 1.8160960112954465,
+      "grad_norm": 0.03382092291300359,
+      "learning_rate": 6.438996236364097e-06,
+      "loss": 2.0854,
+      "step": 5145
+    },
+    {
+      "epoch": 1.817860924814684,
+      "grad_norm": 0.03270997373745948,
+      "learning_rate": 6.316870805761376e-06,
+      "loss": 2.0476,
+      "step": 5150
+    },
+    {
+      "epoch": 1.8196258383339217,
+      "grad_norm": 0.03421471377316621,
+      "learning_rate": 6.19588972506792e-06,
+      "loss": 2.0668,
+      "step": 5155
+    },
+    {
+      "epoch": 1.8213907518531594,
+      "grad_norm": 0.03393510749774542,
+      "learning_rate": 6.076053957825411e-06,
+      "loss": 1.985,
+      "step": 5160
+    },
+    {
+      "epoch": 1.8231556653723966,
+      "grad_norm": 0.03176836254857722,
+      "learning_rate": 5.957364458453806e-06,
+      "loss": 1.8655,
+      "step": 5165
+    },
+    {
+      "epoch": 1.8249205788916343,
+      "grad_norm": 0.0325259215466001,
+      "learning_rate": 5.839822172243802e-06,
+      "loss": 2.0488,
+      "step": 5170
+    },
+    {
+      "epoch": 1.826685492410872,
+      "grad_norm": 0.032014400627162105,
+      "learning_rate": 5.723428035349137e-06,
+      "loss": 1.9829,
+      "step": 5175
+    },
+    {
+      "epoch": 1.8284504059301094,
+      "grad_norm": 0.03351955788227276,
+      "learning_rate": 5.60818297477928e-06,
+      "loss": 2.0452,
+      "step": 5180
+    },
+    {
+      "epoch": 1.830215319449347,
+      "grad_norm": 0.03486740191285143,
+      "learning_rate": 5.49408790839202e-06,
+      "loss": 2.0024,
+      "step": 5185
+    },
+    {
+      "epoch": 1.8319802329685846,
+      "grad_norm": 0.0340211315454655,
+      "learning_rate": 5.381143744886057e-06,
+      "loss": 1.8852,
+      "step": 5190
+    },
+    {
+      "epoch": 1.833745146487822,
+      "grad_norm": 0.03312009988956621,
+      "learning_rate": 5.269351383793924e-06,
+      "loss": 2.0005,
+      "step": 5195
+    },
+    {
+      "epoch": 1.8355100600070595,
+      "grad_norm": 0.03129193004017765,
+      "learning_rate": 5.158711715474673e-06,
+      "loss": 2.0418,
+      "step": 5200
+    },
+    {
+      "epoch": 1.8372749735262972,
+      "grad_norm": 0.032671808615875125,
+      "learning_rate": 5.04922562110685e-06,
+      "loss": 1.9405,
+      "step": 5205
+    },
+    {
+      "epoch": 1.839039887045535,
+      "grad_norm": 0.03295012459437104,
+      "learning_rate": 4.940893972681553e-06,
+      "loss": 2.0745,
+      "step": 5210
+    },
+    {
+      "epoch": 1.8408048005647724,
+      "grad_norm": 0.031320115174951424,
+      "learning_rate": 4.833717632995315e-06,
+      "loss": 1.8338,
+      "step": 5215
+    },
+    {
+      "epoch": 1.8425697140840098,
+      "grad_norm": 0.033457783481520245,
+      "learning_rate": 4.727697455643414e-06,
+      "loss": 1.9601,
+      "step": 5220
+    },
+    {
+      "epoch": 1.8443346276032475,
+      "grad_norm": 0.032468690157388756,
+      "learning_rate": 4.622834285012894e-06,
+      "loss": 1.8938,
+      "step": 5225
+    },
+    {
+      "epoch": 1.846099541122485,
+      "grad_norm": 0.03148126778843055,
+      "learning_rate": 4.519128956276019e-06,
+      "loss": 1.9992,
+      "step": 5230
+    },
+    {
+      "epoch": 1.8478644546417224,
+      "grad_norm": 0.03319909789233191,
+      "learning_rate": 4.416582295383531e-06,
+      "loss": 1.9458,
+      "step": 5235
+    },
+    {
+      "epoch": 1.8496293681609601,
+      "grad_norm": 0.03231655503711228,
+      "learning_rate": 4.315195119058018e-06,
+      "loss": 1.8073,
+      "step": 5240
+    },
+    {
+      "epoch": 1.8513942816801978,
+      "grad_norm": 0.031586637342813084,
+      "learning_rate": 4.214968234787508e-06,
+      "loss": 2.006,
+      "step": 5245
+    },
+    {
+      "epoch": 1.8531591951994353,
+      "grad_norm": 0.03228558244480731,
+      "learning_rate": 4.115902440818985e-06,
+      "loss": 1.9002,
+      "step": 5250
+    },
+    {
+      "epoch": 1.8549241087186727,
+      "grad_norm": 0.030397134471291578,
+      "learning_rate": 4.017998526152066e-06,
+      "loss": 1.9321,
+      "step": 5255
+    },
+    {
+      "epoch": 1.8566890222379104,
+      "grad_norm": 0.031956548072100195,
+      "learning_rate": 3.921257270532718e-06,
+      "loss": 1.799,
+      "step": 5260
+    },
+    {
+      "epoch": 1.858453935757148,
+      "grad_norm": 0.03194458267746179,
+      "learning_rate": 3.825679444446933e-06,
+      "loss": 1.8247,
+      "step": 5265
+    },
+    {
+      "epoch": 1.8602188492763854,
+      "grad_norm": 0.03035443049757921,
+      "learning_rate": 3.7312658091147972e-06,
+      "loss": 1.8936,
+      "step": 5270
+    },
+    {
+      "epoch": 1.861983762795623,
+      "grad_norm": 0.031679179750616765,
+      "learning_rate": 3.6380171164842153e-06,
+      "loss": 2.0703,
+      "step": 5275
+    },
+    {
+      "epoch": 1.8637486763148605,
+      "grad_norm": 0.03244756676740826,
+      "learning_rate": 3.5459341092251293e-06,
+      "loss": 1.9064,
+      "step": 5280
+    },
+    {
+      "epoch": 1.865513589834098,
+      "grad_norm": 0.03235174239807699,
+      "learning_rate": 3.455017520723391e-06,
+      "loss": 1.9654,
+      "step": 5285
+    },
+    {
+      "epoch": 1.8672785033533357,
+      "grad_norm": 0.03332835363031365,
+      "learning_rate": 3.365268075075117e-06,
+      "loss": 1.8933,
+      "step": 5290
+    },
+    {
+      "epoch": 1.8690434168725734,
+      "grad_norm": 0.03219195010701653,
+      "learning_rate": 3.276686487080743e-06,
+      "loss": 1.8224,
+      "step": 5295
+    },
+    {
+      "epoch": 1.8708083303918108,
+      "grad_norm": 0.03042162941322287,
+      "learning_rate": 3.1892734622394945e-06,
+      "loss": 1.7724,
+      "step": 5300
+    },
+    {
+      "epoch": 1.8725732439110483,
+      "grad_norm": 0.032089385918014524,
+      "learning_rate": 3.1030296967436584e-06,
+      "loss": 1.9049,
+      "step": 5305
+    },
+    {
+      "epoch": 1.874338157430286,
+      "grad_norm": 0.03230796321056136,
+      "learning_rate": 3.017955877473072e-06,
+      "loss": 1.9944,
+      "step": 5310
+    },
+    {
+      "epoch": 1.8761030709495234,
+      "grad_norm": 0.03207319746609541,
+      "learning_rate": 2.934052681989657e-06,
+      "loss": 2.0412,
+      "step": 5315
+    },
+    {
+      "epoch": 1.877867984468761,
+      "grad_norm": 0.03105215654531922,
+      "learning_rate": 2.851320778532046e-06,
+      "loss": 2.0387,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8796328979879986,
+      "grad_norm": 0.030533134995743266,
+      "learning_rate": 2.769760826010181e-06,
+      "loss": 1.9938,
+      "step": 5325
+    },
+    {
+      "epoch": 1.8813978115072363,
+      "grad_norm": 0.031945692271731635,
+      "learning_rate": 2.689373474000156e-06,
+      "loss": 1.9957,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8831627250264737,
+      "grad_norm": 0.032079035497907214,
+      "learning_rate": 2.610159362738984e-06,
+      "loss": 2.1756,
+      "step": 5335
+    },
+    {
+      "epoch": 1.8849276385457112,
+      "grad_norm": 0.03229954383181674,
+      "learning_rate": 2.5321191231195357e-06,
+      "loss": 1.9178,
+      "step": 5340
+    },
+    {
+      "epoch": 1.886692552064949,
+      "grad_norm": 0.032668607498753596,
+      "learning_rate": 2.455253376685512e-06,
+      "loss": 1.86,
+      "step": 5345
+    },
+    {
+      "epoch": 1.8884574655841864,
+      "grad_norm": 0.031925102017349695,
+      "learning_rate": 2.3795627356264114e-06,
+      "loss": 2.1114,
+      "step": 5350
+    },
+    {
+      "epoch": 1.8902223791034238,
+      "grad_norm": 0.030449498377627754,
+      "learning_rate": 2.3050478027728358e-06,
+      "loss": 1.8879,
+      "step": 5355
+    },
+    {
+      "epoch": 1.8919872926226615,
+      "grad_norm": 0.031246546160277756,
+      "learning_rate": 2.2317091715914616e-06,
+      "loss": 2.0636,
+      "step": 5360
+    },
+    {
+      "epoch": 1.8937522061418992,
+      "grad_norm": 0.03262026506147617,
+      "learning_rate": 2.159547426180508e-06,
+      "loss": 1.8624,
+      "step": 5365
+    },
+    {
+      "epoch": 1.8955171196611365,
+      "grad_norm": 0.03286335639207764,
+      "learning_rate": 2.088563141264993e-06,
+      "loss": 2.0062,
+      "step": 5370
+    },
+    {
+      "epoch": 1.8972820331803741,
+      "grad_norm": 0.03150066662386577,
+      "learning_rate": 2.0187568821921186e-06,
+      "loss": 1.8553,
+      "step": 5375
+    },
+    {
+      "epoch": 1.8990469466996118,
+      "grad_norm": 0.03081444435815705,
+      "learning_rate": 1.9501292049268593e-06,
+      "loss": 2.0964,
+      "step": 5380
+    },
+    {
+      "epoch": 1.9008118602188493,
+      "grad_norm": 0.032909666575650864,
+      "learning_rate": 1.8826806560474973e-06,
+      "loss": 2.008,
+      "step": 5385
+    },
+    {
+      "epoch": 1.9025767737380868,
+      "grad_norm": 0.03199505668910896,
+      "learning_rate": 1.8164117727412275e-06,
+      "loss": 2.0443,
+      "step": 5390
+    },
+    {
+      "epoch": 1.9043416872573244,
+      "grad_norm": 0.031219065243428327,
+      "learning_rate": 1.7513230827999936e-06,
+      "loss": 1.8145,
+      "step": 5395
+    },
+    {
+      "epoch": 1.906106600776562,
+      "grad_norm": 0.03872583004998154,
+      "learning_rate": 1.687415104616091e-06,
+      "loss": 1.9071,
+      "step": 5400
+    },
+    {
+      "epoch": 1.9078715142957994,
+      "grad_norm": 0.03591829349011323,
+      "learning_rate": 1.6246883471782713e-06,
+      "loss": 1.999,
+      "step": 5405
+    },
+    {
+      "epoch": 1.909636427815037,
+      "grad_norm": 0.037920342576348166,
+      "learning_rate": 1.563143310067494e-06,
+      "loss": 2.0992,
+      "step": 5410
+    },
+    {
+      "epoch": 1.9114013413342748,
+      "grad_norm": 0.038195692231333386,
+      "learning_rate": 1.5027804834530644e-06,
+      "loss": 1.9252,
+      "step": 5415
+    },
+    {
+      "epoch": 1.9131662548535122,
+      "grad_norm": 0.03563838653456414,
+      "learning_rate": 1.4436003480886693e-06,
+      "loss": 2.0734,
+      "step": 5420
+    },
+    {
+      "epoch": 1.9149311683727497,
+      "grad_norm": 0.036955792282102086,
+      "learning_rate": 1.3856033753085637e-06,
+      "loss": 1.9187,
+      "step": 5425
+    },
+    {
+      "epoch": 1.9166960818919874,
+      "grad_norm": 0.03838198594517415,
+      "learning_rate": 1.3287900270238571e-06,
+      "loss": 1.8917,
+      "step": 5430
+    },
+    {
+      "epoch": 1.9184609954112248,
+      "grad_norm": 0.037612537561979596,
+      "learning_rate": 1.273160755718766e-06,
+      "loss": 2.1228,
+      "step": 5435
+    },
+    {
+      "epoch": 1.9202259089304623,
+      "grad_norm": 0.03755711160928464,
+      "learning_rate": 1.218716004447068e-06,
+      "loss": 2.0537,
+      "step": 5440
+    },
+    {
+      "epoch": 1.9219908224497,
+      "grad_norm": 0.038302665572445466,
+      "learning_rate": 1.165456206828519e-06,
+      "loss": 1.92,
+      "step": 5445
+    },
+    {
+      "epoch": 1.9237557359689377,
+      "grad_norm": 0.03928986374724929,
+      "learning_rate": 1.1133817870454587e-06,
+      "loss": 2.1021,
+      "step": 5450
+    },
+    {
+      "epoch": 1.925520649488175,
+      "grad_norm": 0.037195745863024564,
+      "learning_rate": 1.0624931598394116e-06,
+      "loss": 1.8913,
+      "step": 5455
+    },
+    {
+      "epoch": 1.9272855630074126,
+      "grad_norm": 0.03750351670343573,
+      "learning_rate": 1.0127907305077398e-06,
+      "loss": 1.8375,
+      "step": 5460
+    },
+    {
+      "epoch": 1.9290504765266503,
+      "grad_norm": 0.03819519190720331,
+      "learning_rate": 9.642748949004632e-07,
+      "loss": 1.8955,
+      "step": 5465
+    },
+    {
+      "epoch": 1.9308153900458878,
+      "grad_norm": 0.036198910084744025,
+      "learning_rate": 9.169460394171103e-07,
+      "loss": 1.9922,
+      "step": 5470
+    },
+    {
+      "epoch": 1.9325803035651252,
+      "grad_norm": 0.03603030409058373,
+      "learning_rate": 8.708045410036224e-07,
+      "loss": 1.915,
+      "step": 5475
+    },
+    {
+      "epoch": 1.934345217084363,
+      "grad_norm": 0.036998818891040314,
+      "learning_rate": 8.258507671493386e-07,
+      "loss": 1.9352,
+      "step": 5480
+    },
+    {
+      "epoch": 1.9361101306036004,
+      "grad_norm": 0.03741688587725777,
+      "learning_rate": 7.82085075884098e-07,
+      "loss": 2.0887,
+      "step": 5485
+    },
+    {
+      "epoch": 1.9378750441228378,
+      "grad_norm": 0.03677701116119859,
+      "learning_rate": 7.39507815775392e-07,
+      "loss": 2.1482,
+      "step": 5490
+    },
+    {
+      "epoch": 1.9396399576420755,
+      "grad_norm": 0.03576471011591952,
+      "learning_rate": 6.981193259255002e-07,
+      "loss": 1.9899,
+      "step": 5495
+    },
+    {
+      "epoch": 1.9414048711613132,
+      "grad_norm": 0.08637960426386758,
+      "learning_rate": 6.579199359689425e-07,
+      "loss": 2.0606,
+      "step": 5500
+    },
+    {
+      "epoch": 1.9431697846805507,
+      "grad_norm": 0.09931887278527518,
+      "learning_rate": 6.189099660697472e-07,
+      "loss": 2.0384,
+      "step": 5505
+    },
+    {
+      "epoch": 1.9449346981997881,
+      "grad_norm": 0.08731081738947606,
+      "learning_rate": 5.810897269188874e-07,
+      "loss": 1.8498,
+      "step": 5510
+    },
+    {
+      "epoch": 1.9466996117190258,
+      "grad_norm": 0.08756193374420392,
+      "learning_rate": 5.444595197319157e-07,
+      "loss": 1.839,
+      "step": 5515
+    },
+    {
+      "epoch": 1.9484645252382633,
+      "grad_norm": 0.08911094605704886,
+      "learning_rate": 5.090196362463994e-07,
+      "loss": 1.8862,
+      "step": 5520
+    },
+    {
+      "epoch": 1.9502294387575008,
+      "grad_norm": 0.09095727018188401,
+      "learning_rate": 4.747703587197893e-07,
+      "loss": 2.126,
+      "step": 5525
+    },
+    {
+      "epoch": 1.9519943522767385,
+      "grad_norm": 0.09306232896526734,
+      "learning_rate": 4.4171195992702135e-07,
+      "loss": 1.9376,
+      "step": 5530
+    },
+    {
+      "epoch": 1.9537592657959761,
+      "grad_norm": 0.08321472301728682,
+      "learning_rate": 4.0984470315838514e-07,
+      "loss": 1.9896,
+      "step": 5535
+    },
+    {
+      "epoch": 1.9555241793152136,
+      "grad_norm": 0.081724168774306,
+      "learning_rate": 3.7916884221747545e-07,
+      "loss": 1.8946,
+      "step": 5540
+    },
+    {
+      "epoch": 1.957289092834451,
+      "grad_norm": 0.0859236498261215,
+      "learning_rate": 3.496846214191107e-07,
+      "loss": 2.0307,
+      "step": 5545
+    },
+    {
+      "epoch": 1.9590540063536888,
+      "grad_norm": 0.08329989946133373,
+      "learning_rate": 3.2139227558745095e-07,
+      "loss": 2.0076,
+      "step": 5550
+    },
+    {
+      "epoch": 1.9608189198729262,
+      "grad_norm": 0.08761147165058103,
+      "learning_rate": 2.942920300539997e-07,
+      "loss": 2.0952,
+      "step": 5555
+    },
+    {
+      "epoch": 1.9625838333921637,
+      "grad_norm": 0.08731812196465744,
+      "learning_rate": 2.6838410065598834e-07,
+      "loss": 2.1741,
+      "step": 5560
+    },
+    {
+      "epoch": 1.9643487469114014,
+      "grad_norm": 0.09085626534065844,
+      "learning_rate": 2.436686937345278e-07,
+      "loss": 1.872,
+      "step": 5565
+    },
+    {
+      "epoch": 1.966113660430639,
+      "grad_norm": 0.08073060092367672,
+      "learning_rate": 2.2014600613297627e-07,
+      "loss": 1.8811,
+      "step": 5570
+    },
+    {
+      "epoch": 1.9678785739498763,
+      "grad_norm": 0.08978745617817076,
+      "learning_rate": 1.9781622519545738e-07,
+      "loss": 1.9227,
+      "step": 5575
+    },
+    {
+      "epoch": 1.969643487469114,
+      "grad_norm": 0.09079926227330753,
+      "learning_rate": 1.766795287652778e-07,
+      "loss": 1.8001,
+      "step": 5580
+    },
+    {
+      "epoch": 1.9714084009883517,
+      "grad_norm": 0.08758121621555381,
+      "learning_rate": 1.56736085183512e-07,
+      "loss": 1.9557,
+      "step": 5585
+    },
+    {
+      "epoch": 1.9731733145075891,
+      "grad_norm": 0.09174925659720469,
+      "learning_rate": 1.3798605328775304e-07,
+      "loss": 1.9966,
+      "step": 5590
+    },
+    {
+      "epoch": 1.9749382280268266,
+      "grad_norm": 0.0853879503570275,
+      "learning_rate": 1.2042958241073042e-07,
+      "loss": 2.072,
+      "step": 5595
+    },
+    {
+      "epoch": 1.9767031415460643,
+      "grad_norm": 0.0642092738731648,
+      "learning_rate": 1.04066812379211e-07,
+      "loss": 2.0203,
+      "step": 5600
+    },
+    {
+      "epoch": 1.9784680550653018,
+      "grad_norm": 0.06428299227475578,
+      "learning_rate": 8.889787351283317e-08,
+      "loss": 2.0843,
+      "step": 5605
+    },
+    {
+      "epoch": 1.9802329685845392,
+      "grad_norm": 0.05749298640268323,
+      "learning_rate": 7.492288662310774e-08,
+      "loss": 2.0887,
+      "step": 5610
+    },
+    {
+      "epoch": 1.981997882103777,
+      "grad_norm": 0.05782532238953428,
+      "learning_rate": 6.21419630124187e-08,
+      "loss": 2.0708,
+      "step": 5615
+    },
+    {
+      "epoch": 1.9837627956230146,
+      "grad_norm": 0.06279948058227888,
+      "learning_rate": 5.0555204473140545e-08,
+      "loss": 2.0345,
+      "step": 5620
+    },
+    {
+      "epoch": 1.985527709142252,
+      "grad_norm": 0.06278897677484054,
+      "learning_rate": 4.016270328685567e-08,
+      "loss": 1.9149,
+      "step": 5625
+    },
+    {
+      "epoch": 1.9872926226614895,
+      "grad_norm": 0.05776467482344734,
+      "learning_rate": 3.096454222358824e-08,
+      "loss": 1.9564,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9890575361807272,
+      "grad_norm": 0.06807533084790415,
+      "learning_rate": 2.2960794541204697e-08,
+      "loss": 2.0475,
+      "step": 5635
+    },
+    {
+      "epoch": 1.9908224496999647,
+      "grad_norm": 0.05905181864978403,
+      "learning_rate": 1.6151523984730962e-08,
+      "loss": 1.8153,
+      "step": 5640
+    },
+    {
+      "epoch": 1.9925873632192022,
+      "grad_norm": 0.05828133137542685,
+      "learning_rate": 1.0536784785936114e-08,
+      "loss": 1.8298,
+      "step": 5645
+    },
+    {
+      "epoch": 1.9943522767384398,
+      "grad_norm": 0.06835898566768721,
+      "learning_rate": 6.1166216628161234e-09,
+      "loss": 1.9217,
+      "step": 5650
+    },
+    {
+      "epoch": 1.9961171902576775,
+      "grad_norm": 0.06034261919044672,
+      "learning_rate": 2.8910698193440607e-09,
+      "loss": 1.9689,
+      "step": 5655
+    },
+    {
+      "epoch": 1.9978821037769148,
+      "grad_norm": 0.06218030936990563,
+      "learning_rate": 8.601549451203726e-10,
+      "loss": 1.7275,
+      "step": 5660
+    },
+    {
+      "epoch": 1.9996470172961525,
+      "grad_norm": 0.06211761036387624,
+      "learning_rate": 2.3893215123083065e-11,
+      "loss": 2.046,
+      "step": 5665
+    },
+    {
+      "epoch": 2.0,
+      "step": 5666,
+      "total_flos": 1.317227096577147e+18,
+      "train_loss": 2.02712393619128,
+      "train_runtime": 22983.8323,
+      "train_samples_per_second": 15.776,
+      "train_steps_per_second": 0.247
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 5666,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.317227096577147e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}