diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 10.0,
   "eval_steps": 500,
   "global_step": 868560,
   "is_hyper_param_search": false,
@@ -10,30564 +10,30484 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.004605323754259924,
-      "grad_norm": 2.7894906997680664,
+      "epoch": 0.002302685074714935,
+      "grad_norm": 0.7734375,
       "learning_rate": 1.3746891406465874e-07,
-      "loss": 0.6038,
+      "loss": 0.9869670867919922,
       "step": 200
     },
     {
-      "epoch": 0.009210647508519848,
-      "grad_norm": 3.5113091468811035,
+      "epoch": 0.00460537014942987,
+      "grad_norm": 0.71484375,
       "learning_rate": 2.756286266924565e-07,
-      "loss": 0.5979,
+      "loss": 0.9922489166259766,
       "step": 400
     },
     {
-      "epoch": 0.013815971262779773,
-      "grad_norm": 3.262929677963257,
+      "epoch": 0.006908055224144804,
+      "grad_norm": 0.470703125,
       "learning_rate": 4.137883393202542e-07,
-      "loss": 0.6018,
+      "loss": 0.9945306396484375,
       "step": 600
     },
     {
-      "epoch": 0.018421295017039697,
-      "grad_norm": 2.736161231994629,
+      "epoch": 0.00921074029885974,
+      "grad_norm": 0.77734375,
       "learning_rate": 5.51948051948052e-07,
-      "loss": 0.5985,
+      "loss": 0.971050033569336,
       "step": 800
     },
     {
-      "epoch": 0.023026618771299624,
-      "grad_norm": 3.1657299995422363,
+      "epoch": 0.011513425373574674,
+      "grad_norm": 0.56640625,
       "learning_rate": 6.901077645758497e-07,
-      "loss": 0.6069,
+      "loss": 0.98829833984375,
       "step": 1000
     },
     {
-      "epoch": 0.027631942525559547,
-      "grad_norm": 2.5964059829711914,
+      "epoch": 0.013816110448289608,
+      "grad_norm": 0.86328125,
       "learning_rate": 8.282674772036474e-07,
-      "loss": 0.6026,
+      "loss": 0.9930474853515625,
       "step": 1200
     },
     {
-      "epoch": 0.032237266279819474,
-      "grad_norm": 2.9468190670013428,
+      "epoch": 0.016118795523004544,
+      "grad_norm": 0.98828125,
       "learning_rate": 9.664271898314452e-07,
-      "loss": 0.6118,
+      "loss": 0.9899032592773438,
       "step": 1400
     },
     {
-      "epoch": 0.036842590034079394,
-      "grad_norm": 3.6232540607452393,
+      "epoch": 0.01842148059771948,
+      "grad_norm": 0.78515625,
       "learning_rate": 1.1045869024592427e-06,
-      "loss": 0.5998,
+      "loss": 0.9930052947998047,
       "step": 1600
     },
     {
-      "epoch": 0.04144791378833932,
-      "grad_norm": 2.8659310340881348,
+      "epoch": 0.020724165672434413,
+      "grad_norm": 1.0546875,
       "learning_rate": 1.2427466150870407e-06,
-      "loss": 0.6092,
+      "loss": 0.9895816802978515,
       "step": 1800
     },
     {
-      "epoch": 0.04605323754259925,
-      "grad_norm": 3.2160274982452393,
+      "epoch": 0.023026850747149347,
+      "grad_norm": 0.734375,
       "learning_rate": 1.3809063277148384e-06,
-      "loss": 0.6177,
+      "loss": 0.9816262817382813,
       "step": 2000
     },
     {
-      "epoch": 0.05065856129685917,
-      "grad_norm": 2.949619770050049,
+      "epoch": 0.025329535821864282,
+      "grad_norm": 0.625,
       "learning_rate": 1.519066040342636e-06,
-      "loss": 0.6005,
+      "loss": 1.0000796508789063,
       "step": 2200
     },
     {
-      "epoch": 0.055263885051119094,
-      "grad_norm": 3.6632375717163086,
+      "epoch": 0.027632220896579216,
+      "grad_norm": 0.71484375,
       "learning_rate": 1.6572257529704338e-06,
-      "loss": 0.6053,
+      "loss": 0.9617407989501953,
       "step": 2400
     },
     {
-      "epoch": 0.05986920880537902,
-      "grad_norm": 3.9618542194366455,
+      "epoch": 0.02993490597129415,
+      "grad_norm": 0.546875,
       "learning_rate": 1.7953854655982317e-06,
-      "loss": 0.6257,
+      "loss": 0.9662563323974609,
       "step": 2600
     },
     {
-      "epoch": 0.06447453255963895,
-      "grad_norm": 3.0995821952819824,
+      "epoch": 0.03223759104600909,
+      "grad_norm": 0.69140625,
       "learning_rate": 1.9335451782260294e-06,
-      "loss": 0.6203,
+      "loss": 1.013211669921875,
       "step": 2800
     },
     {
-      "epoch": 0.06907985631389886,
-      "grad_norm": 3.9655020236968994,
+      "epoch": 0.03454027612072402,
+      "grad_norm": 0.77734375,
       "learning_rate": 2.071704890853827e-06,
-      "loss": 0.6165,
+      "loss": 0.9835720062255859,
       "step": 3000
     },
     {
-      "epoch": 0.07368518006815879,
-      "grad_norm": 2.8039631843566895,
+      "epoch": 0.03684296119543896,
+      "grad_norm": 0.765625,
       "learning_rate": 2.209864603481625e-06,
-      "loss": 0.6196,
+      "loss": 0.9656615447998047,
       "step": 3200
     },
     {
-      "epoch": 0.07829050382241871,
-      "grad_norm": 3.1764214038848877,
-      "learning_rate": 2.3473335175462836e-06,
-      "loss": 0.61,
+      "epoch": 0.03914564627015389,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.3480243161094226e-06,
+      "loss": 0.9996417236328125,
       "step": 3400
     },
     {
-      "epoch": 0.08289582757667864,
-      "grad_norm": 3.0694644451141357,
-      "learning_rate": 2.4854932301740813e-06,
-      "loss": 0.6113,
+      "epoch": 0.041448331344868826,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.4861840287372203e-06,
+      "loss": 0.9792176055908203,
       "step": 3600
     },
     {
-      "epoch": 0.08750115133093857,
-      "grad_norm": 2.9657208919525146,
-      "learning_rate": 2.623652942801879e-06,
-      "loss": 0.6129,
+      "epoch": 0.04375101641958376,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.624343741365018e-06,
+      "loss": 0.9834017944335938,
       "step": 3800
     },
     {
-      "epoch": 0.0921064750851985,
-      "grad_norm": 3.0443339347839355,
-      "learning_rate": 2.7618126554296768e-06,
-      "loss": 0.6021,
+      "epoch": 0.046053701494298695,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.7625034539928157e-06,
+      "loss": 1.0003073883056641,
       "step": 4000
     },
     {
-      "epoch": 0.09671179883945841,
-      "grad_norm": 3.029796838760376,
-      "learning_rate": 2.8992815694943355e-06,
-      "loss": 0.6093,
+      "epoch": 0.04835638656901363,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.9006631666206134e-06,
+      "loss": 0.9826235961914063,
       "step": 4200
     },
     {
-      "epoch": 0.10131712259371833,
-      "grad_norm": 3.724391222000122,
-      "learning_rate": 3.0374412821221332e-06,
-      "loss": 0.61,
+      "epoch": 0.050659071643728563,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.038822879248411e-06,
+      "loss": 0.9803595733642578,
       "step": 4400
     },
     {
-      "epoch": 0.10592244634797826,
-      "grad_norm": 3.0700559616088867,
-      "learning_rate": 3.175600994749931e-06,
-      "loss": 0.612,
+      "epoch": 0.0529617567184435,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1769825918762093e-06,
+      "loss": 0.9817047119140625,
       "step": 4600
     },
     {
-      "epoch": 0.11052777010223819,
-      "grad_norm": 2.6112992763519287,
-      "learning_rate": 3.313760707377729e-06,
-      "loss": 0.6075,
+      "epoch": 0.05526444179315843,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.315142304504007e-06,
+      "loss": 0.9922354125976562,
       "step": 4800
     },
     {
-      "epoch": 0.11513309385649811,
-      "grad_norm": 3.0309786796569824,
-      "learning_rate": 3.4519204200055268e-06,
-      "loss": 0.6075,
+      "epoch": 0.05756712686787337,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4533020171318043e-06,
+      "loss": 0.9926442718505859,
       "step": 5000
     },
     {
-      "epoch": 0.11973841761075804,
-      "grad_norm": 4.39963960647583,
-      "learning_rate": 3.590080132633324e-06,
-      "loss": 0.616,
+      "epoch": 0.0598698119425883,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.591461729759602e-06,
+      "loss": 0.9734688568115234,
       "step": 5200
     },
     {
-      "epoch": 0.12434374136501795,
-      "grad_norm": 3.1912529468536377,
-      "learning_rate": 3.7282398452611218e-06,
-      "loss": 0.607,
+      "epoch": 0.06217249701730324,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.7296214423873997e-06,
+      "loss": 0.983599624633789,
       "step": 5400
     },
     {
-      "epoch": 0.1289490651192779,
-      "grad_norm": 3.3386693000793457,
-      "learning_rate": 3.8663995578889195e-06,
-      "loss": 0.6155,
+      "epoch": 0.06447518209201818,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.8677811550151974e-06,
+      "loss": 0.9984064483642578,
       "step": 5600
     },
     {
-      "epoch": 0.13355438887353782,
-      "grad_norm": 3.854520082473755,
-      "learning_rate": 4.004559270516718e-06,
-      "loss": 0.6154,
+      "epoch": 0.0667778671667331,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.005940867642995e-06,
+      "loss": 0.9748104858398438,
       "step": 5800
     },
     {
-      "epoch": 0.13815971262779772,
-      "grad_norm": 3.730701208114624,
-      "learning_rate": 4.142028184581376e-06,
-      "loss": 0.6088,
+      "epoch": 0.06908055224144805,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.144100580270793e-06,
+      "loss": 1.0019890594482421,
       "step": 6000
     },
     {
-      "epoch": 0.14276503638205765,
-      "grad_norm": 4.679981231689453,
-      "learning_rate": 4.2801878972091745e-06,
-      "loss": 0.6186,
+      "epoch": 0.07138323731616297,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.282260292898591e-06,
+      "loss": 0.978772964477539,
       "step": 6200
     },
     {
-      "epoch": 0.14737036013631757,
-      "grad_norm": 3.9258837699890137,
-      "learning_rate": 4.418347609836971e-06,
-      "loss": 0.6166,
+      "epoch": 0.07368592239087791,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.420420005526388e-06,
+      "loss": 0.9969493103027344,
       "step": 6400
     },
     {
-      "epoch": 0.1519756838905775,
-      "grad_norm": 3.088724136352539,
-      "learning_rate": 4.556507322464769e-06,
-      "loss": 0.6173,
+      "epoch": 0.07598860746559284,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.5585797181541864e-06,
+      "loss": 0.9867162322998047,
       "step": 6600
     },
     {
-      "epoch": 0.15658100764483743,
-      "grad_norm": 3.408620595932007,
-      "learning_rate": 4.694667035092567e-06,
-      "loss": 0.6005,
+      "epoch": 0.07829129254030778,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.696739430781984e-06,
+      "loss": 0.9982801818847656,
       "step": 6800
     },
     {
-      "epoch": 0.16118633139909735,
-      "grad_norm": 3.2880802154541016,
-      "learning_rate": 4.8328267477203645e-06,
-      "loss": 0.614,
+      "epoch": 0.08059397761502271,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.834899143409782e-06,
+      "loss": 0.9909251403808593,
       "step": 7000
     },
     {
-      "epoch": 0.16579165515335728,
-      "grad_norm": 2.6477365493774414,
-      "learning_rate": 4.970986460348163e-06,
-      "loss": 0.6111,
+      "epoch": 0.08289666268973765,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.97305885603758e-06,
+      "loss": 1.0040445709228516,
       "step": 7200
     },
     {
-      "epoch": 0.1703969789076172,
-      "grad_norm": 3.5173377990722656,
-      "learning_rate": 5.10914617297596e-06,
-      "loss": 0.6173,
+      "epoch": 0.0851993477644526,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.111218568665377e-06,
+      "loss": 1.0038322448730468,
       "step": 7400
     },
     {
-      "epoch": 0.17500230266187713,
-      "grad_norm": 2.5049519538879395,
-      "learning_rate": 5.247305885603758e-06,
-      "loss": 0.6016,
+      "epoch": 0.08750203283916752,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.2493782812931754e-06,
+      "loss": 0.9839054107666015,
       "step": 7600
     },
     {
-      "epoch": 0.17960762641613706,
-      "grad_norm": 4.8469157218933105,
-      "learning_rate": 5.385465598231556e-06,
-      "loss": 0.6142,
+      "epoch": 0.08980471791388246,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.3875379939209736e-06,
+      "loss": 0.9923789978027344,
       "step": 7800
     },
     {
-      "epoch": 0.184212950170397,
-      "grad_norm": 3.006009101867676,
-      "learning_rate": 5.5236253108593535e-06,
-      "loss": 0.6147,
+      "epoch": 0.09210740298859739,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.52569770654877e-06,
+      "loss": 0.9952265930175781,
       "step": 8000
     },
     {
-      "epoch": 0.18881827392465692,
-      "grad_norm": 3.286186456680298,
-      "learning_rate": 5.661785023487152e-06,
-      "loss": 0.6037,
+      "epoch": 0.09441008806331233,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.663857419176568e-06,
+      "loss": 0.9856996154785156,
       "step": 8200
     },
     {
-      "epoch": 0.19342359767891681,
-      "grad_norm": 2.788668394088745,
-      "learning_rate": 5.799944736114949e-06,
-      "loss": 0.6148,
+      "epoch": 0.09671277313802726,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.8020171318043654e-06,
+      "loss": 1.0003697204589843,
       "step": 8400
     },
     {
-      "epoch": 0.19802892143317674,
-      "grad_norm": 3.9425911903381348,
-      "learning_rate": 5.938104448742747e-06,
-      "loss": 0.6246,
+      "epoch": 0.0990154582127422,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.940176844432164e-06,
+      "loss": 0.9982233428955078,
       "step": 8600
     },
     {
-      "epoch": 0.20263424518743667,
-      "grad_norm": 3.2533152103424072,
-      "learning_rate": 6.076264161370544e-06,
-      "loss": 0.6109,
+      "epoch": 0.10131814328745713,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.078336557059962e-06,
+      "loss": 0.9870623016357422,
       "step": 8800
     },
     {
-      "epoch": 0.2072395689416966,
-      "grad_norm": 3.5510547161102295,
-      "learning_rate": 6.214423873998342e-06,
-      "loss": 0.6226,
+      "epoch": 0.10362082836217207,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.216496269687759e-06,
+      "loss": 1.0076547241210938,
       "step": 9000
     },
     {
-      "epoch": 0.21184489269595652,
-      "grad_norm": 2.908961057662964,
-      "learning_rate": 6.35258358662614e-06,
-      "loss": 0.6152,
+      "epoch": 0.105923513436887,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.354655982315557e-06,
+      "loss": 1.013364028930664,
       "step": 9200
     },
     {
-      "epoch": 0.21645021645021645,
-      "grad_norm": 3.2206296920776367,
-      "learning_rate": 6.490743299253938e-06,
-      "loss": 0.6119,
+      "epoch": 0.10822619851160194,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.4928156949433544e-06,
+      "loss": 0.9872899627685547,
       "step": 9400
     },
     {
-      "epoch": 0.22105554020447638,
-      "grad_norm": 2.8051042556762695,
-      "learning_rate": 6.628212213318597e-06,
-      "loss": 0.6112,
+      "epoch": 0.11052888358631686,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.630975407571153e-06,
+      "loss": 0.9925249481201172,
       "step": 9600
     },
     {
-      "epoch": 0.2256608639587363,
-      "grad_norm": 3.4578733444213867,
-      "learning_rate": 6.766371925946394e-06,
-      "loss": 0.5983,
+      "epoch": 0.1128315686610318,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.769135120198951e-06,
+      "loss": 0.9940272521972656,
       "step": 9800
     },
     {
-      "epoch": 0.23026618771299623,
-      "grad_norm": 3.29227614402771,
-      "learning_rate": 6.904531638574192e-06,
-      "loss": 0.6135,
+      "epoch": 0.11513425373574673,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.907294832826748e-06,
+      "loss": 0.9853199005126954,
       "step": 10000
     },
     {
-      "epoch": 0.23487151146725616,
-      "grad_norm": 3.175769567489624,
-      "learning_rate": 7.04269135120199e-06,
-      "loss": 0.604,
+      "epoch": 0.11743693881046167,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.045454545454546e-06,
+      "loss": 0.9513562774658203,
       "step": 10200
     },
     {
-      "epoch": 0.23947683522151608,
-      "grad_norm": 2.6275556087493896,
-      "learning_rate": 7.1808510638297875e-06,
-      "loss": 0.6145,
+      "epoch": 0.1197396238851766,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.183614258082343e-06,
+      "loss": 0.9889139556884765,
       "step": 10400
     },
     {
-      "epoch": 0.244082158975776,
-      "grad_norm": 3.551853895187378,
-      "learning_rate": 7.319010776457585e-06,
-      "loss": 0.6302,
+      "epoch": 0.12204230895989154,
+      "grad_norm": 2.375,
+      "learning_rate": 7.321773970710141e-06,
+      "loss": 0.9937071990966797,
       "step": 10600
     },
     {
-      "epoch": 0.2486874827300359,
-      "grad_norm": 2.5962212085723877,
-      "learning_rate": 7.457170489085383e-06,
-      "loss": 0.6079,
+      "epoch": 0.12434499403460648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.459933683337939e-06,
+      "loss": 0.9788626098632812,
       "step": 10800
     },
     {
-      "epoch": 0.25329280648429586,
-      "grad_norm": 3.54152774810791,
-      "learning_rate": 7.59533020171318e-06,
-      "loss": 0.626,
+      "epoch": 0.12664767910932143,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.598093395965737e-06,
+      "loss": 0.9843457794189453,
       "step": 11000
     },
     {
-      "epoch": 0.2578981302385558,
-      "grad_norm": 3.380774974822998,
-      "learning_rate": 7.733489914340978e-06,
-      "loss": 0.6205,
+      "epoch": 0.12895036418403635,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.736253108593534e-06,
+      "loss": 0.9941126251220703,
       "step": 11200
     },
     {
-      "epoch": 0.2625034539928157,
-      "grad_norm": 3.3264219760894775,
-      "learning_rate": 7.871649626968777e-06,
-      "loss": 0.6121,
+      "epoch": 0.13125304925875128,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.874412821221332e-06,
+      "loss": 0.9951610565185547,
       "step": 11400
     },
     {
-      "epoch": 0.26710877774707564,
-      "grad_norm": 2.831803798675537,
-      "learning_rate": 8.009809339596573e-06,
-      "loss": 0.5994,
+      "epoch": 0.1335557343334662,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.012572533849129e-06,
+      "loss": 0.9955122375488281,
       "step": 11600
     },
     {
-      "epoch": 0.27171410150133557,
-      "grad_norm": 3.301412582397461,
-      "learning_rate": 8.147969052224371e-06,
-      "loss": 0.6236,
+      "epoch": 0.13585841940818116,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.150732246476927e-06,
+      "loss": 1.0030738830566406,
       "step": 11800
     },
     {
-      "epoch": 0.27631942525559544,
-      "grad_norm": 3.9409468173980713,
-      "learning_rate": 8.28612876485217e-06,
-      "loss": 0.6208,
+      "epoch": 0.1381611044828961,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.288891959104725e-06,
+      "loss": 0.9774797058105469,
       "step": 12000
     },
     {
-      "epoch": 0.28092474900985537,
-      "grad_norm": 2.723642349243164,
-      "learning_rate": 8.424288477479967e-06,
-      "loss": 0.6137,
+      "epoch": 0.14046378955761102,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.427051671732523e-06,
+      "loss": 0.9933738708496094,
       "step": 12200
     },
     {
-      "epoch": 0.2855300727641153,
-      "grad_norm": 3.9278056621551514,
-      "learning_rate": 8.562448190107766e-06,
-      "loss": 0.6156,
+      "epoch": 0.14276647463232595,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.56521138436032e-06,
+      "loss": 0.9964907073974609,
       "step": 12400
     },
     {
-      "epoch": 0.2901353965183752,
-      "grad_norm": 3.5097107887268066,
-      "learning_rate": 8.700607902735564e-06,
-      "loss": 0.618,
+      "epoch": 0.1450691597070409,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.703371096988118e-06,
+      "loss": 1.0063589477539063,
       "step": 12600
     },
     {
-      "epoch": 0.29474072027263515,
-      "grad_norm": 3.150387763977051,
-      "learning_rate": 8.83876761536336e-06,
-      "loss": 0.6165,
+      "epoch": 0.14737184478175583,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.841530809615916e-06,
+      "loss": 0.9975699615478516,
       "step": 12800
     },
     {
-      "epoch": 0.2993460440268951,
-      "grad_norm": 3.5158212184906006,
-      "learning_rate": 8.976927327991158e-06,
-      "loss": 0.6156,
+      "epoch": 0.14967452985647076,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.979690522243714e-06,
+      "loss": 1.0073341369628905,
       "step": 13000
     },
     {
-      "epoch": 0.303951367781155,
-      "grad_norm": 3.6976168155670166,
-      "learning_rate": 9.115087040618956e-06,
-      "loss": 0.611,
+      "epoch": 0.15197721493118568,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.117850234871512e-06,
+      "loss": 0.9989344024658203,
       "step": 13200
     },
     {
-      "epoch": 0.30855669153541493,
-      "grad_norm": 2.7496871948242188,
-      "learning_rate": 9.253246753246755e-06,
-      "loss": 0.6202,
+      "epoch": 0.15427990000590064,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.256009947499309e-06,
+      "loss": 0.9917288970947266,
       "step": 13400
     },
     {
-      "epoch": 0.31316201528967486,
-      "grad_norm": 2.973073959350586,
-      "learning_rate": 9.391406465874553e-06,
-      "loss": 0.6151,
+      "epoch": 0.15658258508061557,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.394169660127107e-06,
+      "loss": 0.9956332397460937,
       "step": 13600
     },
     {
-      "epoch": 0.3177673390439348,
-      "grad_norm": 4.366604804992676,
-      "learning_rate": 9.529566178502349e-06,
-      "loss": 0.6264,
+      "epoch": 0.1588852701553305,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.532329372754905e-06,
+      "loss": 0.993236312866211,
       "step": 13800
     },
     {
-      "epoch": 0.3223726627981947,
-      "grad_norm": 2.6890974044799805,
-      "learning_rate": 9.667725891130147e-06,
-      "loss": 0.6149,
+      "epoch": 0.16118795523004542,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.670489085382703e-06,
+      "loss": 1.0043616485595703,
       "step": 14000
     },
     {
-      "epoch": 0.32697798655245464,
-      "grad_norm": 2.9724693298339844,
-      "learning_rate": 9.805885603757944e-06,
-      "loss": 0.6301,
+      "epoch": 0.16349064030476038,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.808648798010501e-06,
+      "loss": 0.9785184478759765,
       "step": 14200
     },
     {
-      "epoch": 0.33158331030671456,
-      "grad_norm": 2.7939202785491943,
-      "learning_rate": 9.944045316385742e-06,
-      "loss": 0.6275,
+      "epoch": 0.1657933253794753,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.946808510638298e-06,
+      "loss": 1.0018584442138672,
       "step": 14400
     },
     {
-      "epoch": 0.3361886340609745,
-      "grad_norm": 3.3362252712249756,
-      "learning_rate": 1.008220502901354e-05,
-      "loss": 0.6207,
+      "epoch": 0.16809601045419023,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0084968223266096e-05,
+      "loss": 0.9883576202392578,
       "step": 14600
     },
     {
-      "epoch": 0.3407939578152344,
-      "grad_norm": 3.060162305831909,
-      "learning_rate": 1.0220364741641336e-05,
-      "loss": 0.6169,
+      "epoch": 0.1703986955289052,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0223127935893894e-05,
+      "loss": 0.9930864715576172,
       "step": 14800
     },
     {
-      "epoch": 0.34539928156949434,
-      "grad_norm": 2.850121021270752,
-      "learning_rate": 1.0358524454269135e-05,
-      "loss": 0.6269,
+      "epoch": 0.1727013806036201,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0361287648521692e-05,
+      "loss": 0.9924915313720704,
       "step": 15000
     },
     {
-      "epoch": 0.35000460532375427,
-      "grad_norm": 3.7358944416046143,
-      "learning_rate": 1.0496684166896933e-05,
-      "loss": 0.6223,
+      "epoch": 0.17500406567833504,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.049944736114949e-05,
+      "loss": 0.9670263671875,
       "step": 15200
     },
     {
-      "epoch": 0.3546099290780142,
-      "grad_norm": 3.362393379211426,
-      "learning_rate": 1.063484387952473e-05,
-      "loss": 0.6271,
+      "epoch": 0.17730675075304997,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0637607073777287e-05,
+      "loss": 0.9942636108398437,
       "step": 15400
     },
     {
-      "epoch": 0.3592152528322741,
-      "grad_norm": 2.7046902179718018,
-      "learning_rate": 1.0773003592152529e-05,
-      "loss": 0.6194,
+      "epoch": 0.17960943582776492,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0775766786405085e-05,
+      "loss": 1.0009183502197265,
       "step": 15600
     },
     {
-      "epoch": 0.36382057658653405,
-      "grad_norm": 3.0208888053894043,
-      "learning_rate": 1.0911163304780325e-05,
-      "loss": 0.6249,
+      "epoch": 0.18191212090247985,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0913926499032883e-05,
+      "loss": 0.9625146484375,
       "step": 15800
     },
     {
-      "epoch": 0.368425900340794,
-      "grad_norm": 3.113605260848999,
-      "learning_rate": 1.1049323017408124e-05,
-      "loss": 0.6144,
+      "epoch": 0.18421480597719478,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1052086211660681e-05,
+      "loss": 0.9849910736083984,
       "step": 16000
     },
     {
-      "epoch": 0.3730312240950539,
-      "grad_norm": 3.3108925819396973,
-      "learning_rate": 1.1186791931472782e-05,
-      "loss": 0.6209,
+      "epoch": 0.1865174910519097,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.119024592428848e-05,
+      "loss": 1.0009237670898437,
       "step": 16200
     },
     {
-      "epoch": 0.37763654784931383,
-      "grad_norm": 2.8616628646850586,
-      "learning_rate": 1.132495164410058e-05,
-      "loss": 0.6106,
+      "epoch": 0.18882017612662466,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1328405636916274e-05,
+      "loss": 0.980335464477539,
       "step": 16400
     },
     {
-      "epoch": 0.38224187160357376,
-      "grad_norm": 3.016742706298828,
-      "learning_rate": 1.1463111356728379e-05,
-      "loss": 0.6179,
+      "epoch": 0.1911228612013396,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1466565349544072e-05,
+      "loss": 1.0018857574462892,
       "step": 16600
     },
     {
-      "epoch": 0.38684719535783363,
-      "grad_norm": 3.3050427436828613,
-      "learning_rate": 1.1601271069356177e-05,
-      "loss": 0.6225,
+      "epoch": 0.19342554627605452,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.160472506217187e-05,
+      "loss": 1.0179067993164062,
       "step": 16800
     },
     {
-      "epoch": 0.39145251911209356,
-      "grad_norm": 2.2596371173858643,
-      "learning_rate": 1.1739430781983975e-05,
-      "loss": 0.6282,
+      "epoch": 0.19572823135076944,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.1742884774799668e-05,
+      "loss": 0.9830726623535156,
       "step": 17000
     },
     {
-      "epoch": 0.3960578428663535,
-      "grad_norm": 3.506793260574341,
-      "learning_rate": 1.1877590494611771e-05,
-      "loss": 0.6213,
+      "epoch": 0.1980309164254844,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1881044487427467e-05,
+      "loss": 1.0245218658447266,
       "step": 17200
     },
     {
-      "epoch": 0.4006631666206134,
-      "grad_norm": 3.41317081451416,
-      "learning_rate": 1.2015059408676432e-05,
-      "loss": 0.6143,
+      "epoch": 0.20033360150019933,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2019204200055263e-05,
+      "loss": 0.9935623168945312,
       "step": 17400
     },
     {
-      "epoch": 0.40526849037487334,
-      "grad_norm": 2.875344753265381,
-      "learning_rate": 1.2153219121304228e-05,
-      "loss": 0.6113,
+      "epoch": 0.20263628657491425,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2157363912683061e-05,
+      "loss": 0.9736779022216797,
       "step": 17600
     },
     {
-      "epoch": 0.40987381412913326,
-      "grad_norm": 2.982757568359375,
-      "learning_rate": 1.2291378833932026e-05,
-      "loss": 0.6151,
+      "epoch": 0.2049389716496292,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.229552362531086e-05,
+      "loss": 0.9839263153076172,
       "step": 17800
     },
     {
-      "epoch": 0.4144791378833932,
-      "grad_norm": 3.633155107498169,
-      "learning_rate": 1.2429538546559823e-05,
-      "loss": 0.6338,
+      "epoch": 0.20724165672434414,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.2433683337938657e-05,
+      "loss": 0.9675028228759766,
       "step": 18000
     },
     {
-      "epoch": 0.4190844616376531,
-      "grad_norm": 2.6880545616149902,
-      "learning_rate": 1.2567698259187621e-05,
-      "loss": 0.6159,
+      "epoch": 0.20954434179905906,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2571843050566456e-05,
+      "loss": 0.9875794982910157,
       "step": 18200
     },
     {
-      "epoch": 0.42368978539191304,
-      "grad_norm": 2.981095790863037,
-      "learning_rate": 1.2705857971815419e-05,
-      "loss": 0.6312,
+      "epoch": 0.211847026873774,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2710002763194252e-05,
+      "loss": 0.9895236968994141,
       "step": 18400
     },
     {
-      "epoch": 0.42829510914617297,
-      "grad_norm": 2.658784866333008,
-      "learning_rate": 1.2844017684443215e-05,
-      "loss": 0.6183,
+      "epoch": 0.21414971194848895,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.284816247582205e-05,
+      "loss": 0.9853239440917969,
       "step": 18600
     },
     {
-      "epoch": 0.4329004329004329,
-      "grad_norm": 2.975275754928589,
-      "learning_rate": 1.2982177397071014e-05,
-      "loss": 0.6246,
+      "epoch": 0.21645239702320387,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2986322188449848e-05,
+      "loss": 0.974911880493164,
       "step": 18800
     },
     {
-      "epoch": 0.4375057566546928,
-      "grad_norm": 2.6521835327148438,
-      "learning_rate": 1.3120337109698812e-05,
-      "loss": 0.6213,
+      "epoch": 0.2187550820979188,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3124481901077646e-05,
+      "loss": 0.9828662109375,
       "step": 19000
     },
     {
-      "epoch": 0.44211108040895275,
-      "grad_norm": 3.5868492126464844,
-      "learning_rate": 1.325849682232661e-05,
-      "loss": 0.6221,
+      "epoch": 0.22105776717263373,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3262641613705445e-05,
+      "loss": 0.9706605529785156,
       "step": 19200
     },
     {
-      "epoch": 0.4467164041632127,
-      "grad_norm": 2.928968667984009,
-      "learning_rate": 1.3396656534954408e-05,
-      "loss": 0.6291,
+      "epoch": 0.22336045224734868,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3400801326333241e-05,
+      "loss": 0.9889375305175782,
       "step": 19400
     },
     {
-      "epoch": 0.4513217279174726,
-      "grad_norm": 2.7386858463287354,
-      "learning_rate": 1.3534816247582204e-05,
-      "loss": 0.6154,
+      "epoch": 0.2256631373220636,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3538961038961039e-05,
+      "loss": 0.9776881408691406,
       "step": 19600
     },
     {
-      "epoch": 0.45592705167173253,
-      "grad_norm": 2.974900245666504,
-      "learning_rate": 1.3672975960210003e-05,
-      "loss": 0.6147,
+      "epoch": 0.22796582239677854,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3677120751588837e-05,
+      "loss": 0.9969490814208984,
       "step": 19800
     },
     {
-      "epoch": 0.46053237542599246,
-      "grad_norm": 3.2995572090148926,
-      "learning_rate": 1.38111356728378e-05,
-      "loss": 0.6335,
+      "epoch": 0.23026850747149347,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3815280464216635e-05,
+      "loss": 0.9805326080322265,
       "step": 20000
     },
     {
-      "epoch": 0.4651376991802524,
-      "grad_norm": 2.7065649032592773,
-      "learning_rate": 1.3949295385465599e-05,
-      "loss": 0.6236,
+      "epoch": 0.23257119254620842,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3953440176844434e-05,
+      "loss": 0.9756855773925781,
       "step": 20200
     },
     {
-      "epoch": 0.4697430229345123,
-      "grad_norm": 3.2409005165100098,
-      "learning_rate": 1.4086764299530258e-05,
-      "loss": 0.6245,
+      "epoch": 0.23487387762092335,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4091599889472232e-05,
+      "loss": 0.9778380584716797,
       "step": 20400
     },
     {
-      "epoch": 0.47434834668877224,
-      "grad_norm": 3.296063184738159,
-      "learning_rate": 1.4224924012158056e-05,
-      "loss": 0.6198,
+      "epoch": 0.23717656269563828,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.4229759602100028e-05,
+      "loss": 0.9680982208251954,
       "step": 20600
     },
     {
-      "epoch": 0.47895367044303216,
-      "grad_norm": 3.3391969203948975,
-      "learning_rate": 1.4363083724785854e-05,
-      "loss": 0.6339,
+      "epoch": 0.2394792477703532,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4367919314727826e-05,
+      "loss": 1.0011354827880858,
       "step": 20800
     },
     {
-      "epoch": 0.4835589941972921,
-      "grad_norm": 3.536306858062744,
-      "learning_rate": 1.450124343741365e-05,
-      "loss": 0.6264,
+      "epoch": 0.24178193284506816,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4506079027355623e-05,
+      "loss": 0.982271499633789,
       "step": 21000
     },
     {
-      "epoch": 0.488164317951552,
-      "grad_norm": 2.2294647693634033,
-      "learning_rate": 1.4639403150041448e-05,
-      "loss": 0.6182,
+      "epoch": 0.2440846179197831,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4644238739983421e-05,
+      "loss": 0.978791732788086,
       "step": 21200
     },
     {
-      "epoch": 0.49276964170581194,
-      "grad_norm": 2.4579017162323,
-      "learning_rate": 1.4777562862669247e-05,
-      "loss": 0.6287,
+      "epoch": 0.24638730299449801,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4782398452611219e-05,
+      "loss": 0.9688619995117187,
       "step": 21400
     },
     {
-      "epoch": 0.4973749654600718,
-      "grad_norm": 3.0496151447296143,
-      "learning_rate": 1.4915722575297045e-05,
-      "loss": 0.6272,
+      "epoch": 0.24868998806921297,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4920558165239015e-05,
+      "loss": 0.9980213928222657,
       "step": 21600
     },
     {
-      "epoch": 0.5019802892143318,
-      "grad_norm": 2.970477819442749,
-      "learning_rate": 1.5053882287924841e-05,
-      "loss": 0.6364,
+      "epoch": 0.25099267314392787,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5058717877866815e-05,
+      "loss": 1.0045819854736329,
       "step": 21800
     },
     {
-      "epoch": 0.5065856129685917,
-      "grad_norm": 3.1261672973632812,
-      "learning_rate": 1.5192042000552638e-05,
-      "loss": 0.6387,
+      "epoch": 0.25329535821864285,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.5196877590494613e-05,
+      "loss": 0.9890023040771484,
       "step": 22000
     },
     {
-      "epoch": 0.5111909367228517,
-      "grad_norm": 3.239424228668213,
-      "learning_rate": 1.5330201713180437e-05,
-      "loss": 0.6235,
+      "epoch": 0.2555980432933578,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.533503730312241e-05,
+      "loss": 0.9958658599853516,
       "step": 22200
     },
     {
-      "epoch": 0.5157962604771116,
-      "grad_norm": 3.2637672424316406,
-      "learning_rate": 1.5468361425808236e-05,
-      "loss": 0.6302,
+      "epoch": 0.2579007283680727,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.547319701575021e-05,
+      "loss": 1.0080494689941406,
       "step": 22400
     },
     {
-      "epoch": 0.5204015842313715,
-      "grad_norm": 3.604191541671753,
-      "learning_rate": 1.560652113843603e-05,
-      "loss": 0.6193,
+      "epoch": 0.26020341344278763,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5611356728378008e-05,
+      "loss": 0.974377670288086,
       "step": 22600
     },
     {
-      "epoch": 0.5250069079856314,
-      "grad_norm": 3.1682140827178955,
-      "learning_rate": 1.574468085106383e-05,
-      "loss": 0.6231,
+      "epoch": 0.26250609851750256,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5749516441005806e-05,
+      "loss": 0.9929348754882813,
       "step": 22800
     },
     {
-      "epoch": 0.5296122317398914,
-      "grad_norm": 2.781855583190918,
-      "learning_rate": 1.5882840563691627e-05,
-      "loss": 0.6283,
+      "epoch": 0.2648087835922175,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.58876761536336e-05,
+      "loss": 0.9754251098632812,
       "step": 23000
     },
     {
-      "epoch": 0.5342175554941513,
-      "grad_norm": 2.9197840690612793,
-      "learning_rate": 1.6021000276319425e-05,
-      "loss": 0.6377,
+      "epoch": 0.2671114686669324,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.60258358662614e-05,
+      "loss": 0.9828868103027344,
       "step": 23200
     },
     {
-      "epoch": 0.5388228792484112,
-      "grad_norm": 3.9863662719726562,
-      "learning_rate": 1.6159159988947223e-05,
-      "loss": 0.6231,
+      "epoch": 0.26941415374164734,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6163995578889194e-05,
+      "loss": 0.9940708923339844,
       "step": 23400
     },
     {
-      "epoch": 0.5434282030026711,
-      "grad_norm": 2.8455443382263184,
-      "learning_rate": 1.629731970157502e-05,
-      "loss": 0.6227,
+      "epoch": 0.2717168388163623,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6302155291516992e-05,
+      "loss": 0.996038818359375,
       "step": 23600
     },
     {
-      "epoch": 0.5480335267569311,
-      "grad_norm": 2.6754982471466064,
-      "learning_rate": 1.643547941420282e-05,
-      "loss": 0.6314,
+      "epoch": 0.27401952389107725,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.644031500414479e-05,
+      "loss": 0.9999778747558594,
       "step": 23800
     },
     {
-      "epoch": 0.5526388505111909,
-      "grad_norm": 3.931835651397705,
-      "learning_rate": 1.6573639126830617e-05,
-      "loss": 0.6298,
+      "epoch": 0.2763222089657922,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.6578474716772588e-05,
+      "loss": 0.9920552825927734,
       "step": 24000
     },
     {
-      "epoch": 0.5572441742654508,
-      "grad_norm": 3.3282952308654785,
-      "learning_rate": 1.6711798839458415e-05,
-      "loss": 0.6149,
+      "epoch": 0.2786248940405071,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6716634429400386e-05,
+      "loss": 0.9706204223632813,
       "step": 24200
     },
     {
-      "epoch": 0.5618494980197107,
-      "grad_norm": 3.1311309337615967,
-      "learning_rate": 1.6849958552086214e-05,
-      "loss": 0.6408,
+      "epoch": 0.28092757911522204,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6854794142028184e-05,
+      "loss": 0.9838566589355469,
       "step": 24400
     },
     {
-      "epoch": 0.5664548217739707,
-      "grad_norm": 4.448089122772217,
-      "learning_rate": 1.698811826471401e-05,
-      "loss": 0.6316,
+      "epoch": 0.28323026418993696,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6992953854655982e-05,
+      "loss": 0.9927280426025391,
       "step": 24600
     },
     {
-      "epoch": 0.5710601455282306,
-      "grad_norm": 3.4819765090942383,
-      "learning_rate": 1.7126277977341806e-05,
-      "loss": 0.6361,
+      "epoch": 0.2855329492646519,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.713111356728378e-05,
+      "loss": 0.9818917083740234,
       "step": 24800
     },
     {
-      "epoch": 0.5756654692824905,
-      "grad_norm": 3.762315034866333,
-      "learning_rate": 1.7264437689969605e-05,
-      "loss": 0.6267,
+      "epoch": 0.2878356343393669,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.726927327991158e-05,
+      "loss": 1.0006220245361328,
       "step": 25000
     },
     {
-      "epoch": 0.5802707930367504,
-      "grad_norm": 3.5128722190856934,
-      "learning_rate": 1.7402597402597403e-05,
-      "loss": 0.6363,
+      "epoch": 0.2901383194140818,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7407432992539373e-05,
+      "loss": 0.9794861602783204,
       "step": 25200
     },
     {
-      "epoch": 0.5848761167910104,
-      "grad_norm": 2.8912715911865234,
-      "learning_rate": 1.75407571152252e-05,
-      "loss": 0.6381,
+      "epoch": 0.29244100448879673,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.754559270516717e-05,
+      "loss": 1.0012014770507813,
       "step": 25400
     },
     {
-      "epoch": 0.5894814405452703,
-      "grad_norm": 3.6335270404815674,
-      "learning_rate": 1.767753523072672e-05,
-      "loss": 0.6278,
+      "epoch": 0.29474368956351166,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.768375241779497e-05,
+      "loss": 0.9826255035400391,
       "step": 25600
     },
     {
-      "epoch": 0.5940867642995302,
-      "grad_norm": 3.1860055923461914,
-      "learning_rate": 1.781569494335452e-05,
-      "loss": 0.6321,
+      "epoch": 0.2970463746382266,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7821912130422768e-05,
+      "loss": 0.9798987579345703,
       "step": 25800
     },
     {
-      "epoch": 0.5986920880537902,
-      "grad_norm": 2.7040374279022217,
-      "learning_rate": 1.7953854655982317e-05,
-      "loss": 0.636,
+      "epoch": 0.2993490597129415,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.7960071843050566e-05,
+      "loss": 0.9695772552490234,
       "step": 26000
     },
     {
-      "epoch": 0.6032974118080501,
-      "grad_norm": 3.322178840637207,
-      "learning_rate": 1.8092014368610115e-05,
-      "loss": 0.636,
+      "epoch": 0.30165174478765644,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8098231555678364e-05,
+      "loss": 0.9748085784912109,
       "step": 26200
     },
     {
-      "epoch": 0.60790273556231,
-      "grad_norm": 3.2014033794403076,
-      "learning_rate": 1.8230174081237913e-05,
-      "loss": 0.6369,
+      "epoch": 0.30395442986237137,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8236391268306162e-05,
+      "loss": 0.9792256927490235,
       "step": 26400
     },
     {
-      "epoch": 0.6125080593165699,
-      "grad_norm": 3.6202449798583984,
-      "learning_rate": 1.836833379386571e-05,
-      "loss": 0.6269,
+      "epoch": 0.30625711493708635,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.837455098093396e-05,
+      "loss": 0.9950776672363282,
       "step": 26600
     },
     {
-      "epoch": 0.6171133830708299,
-      "grad_norm": 3.62076735496521,
-      "learning_rate": 1.850649350649351e-05,
-      "loss": 0.6307,
+      "epoch": 0.3085598000118013,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.851271069356176e-05,
+      "loss": 0.9882295989990234,
       "step": 26800
     },
     {
-      "epoch": 0.6217187068250898,
-      "grad_norm": 2.5848195552825928,
-      "learning_rate": 1.8644653219121307e-05,
-      "loss": 0.6311,
+      "epoch": 0.3108624850865162,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8650870406189557e-05,
+      "loss": 0.9964749908447266,
       "step": 27000
     },
     {
-      "epoch": 0.6263240305793497,
-      "grad_norm": 3.199153184890747,
-      "learning_rate": 1.8782812931749105e-05,
-      "loss": 0.639,
+      "epoch": 0.31316517016123113,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.878903011881735e-05,
+      "loss": 0.9571038818359375,
       "step": 27200
     },
     {
-      "epoch": 0.6309293543336096,
-      "grad_norm": 2.6851189136505127,
-      "learning_rate": 1.89209726443769e-05,
-      "loss": 0.6422,
+      "epoch": 0.31546785523594606,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.892718983144515e-05,
+      "loss": 0.9926722717285156,
       "step": 27400
     },
     {
-      "epoch": 0.6355346780878696,
-      "grad_norm": 3.650444507598877,
-      "learning_rate": 1.9059132357004698e-05,
-      "loss": 0.6412,
+      "epoch": 0.317770540310661,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9065349544072948e-05,
+      "loss": 1.017679443359375,
       "step": 27600
     },
     {
-      "epoch": 0.6401400018421295,
-      "grad_norm": 2.841759204864502,
-      "learning_rate": 1.9197292069632496e-05,
-      "loss": 0.6419,
+      "epoch": 0.3200732253853759,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9203509256700746e-05,
+      "loss": 0.9835760498046875,
       "step": 27800
     },
     {
-      "epoch": 0.6447453255963894,
-      "grad_norm": 3.3673157691955566,
-      "learning_rate": 1.9335451782260294e-05,
-      "loss": 0.6237,
+      "epoch": 0.32237591046009084,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9341668969328544e-05,
+      "loss": 1.0087098693847656,
       "step": 28000
     },
     {
-      "epoch": 0.6493506493506493,
-      "grad_norm": 3.8618879318237305,
-      "learning_rate": 1.9473611494888093e-05,
-      "loss": 0.6369,
+      "epoch": 0.3246785955348058,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9479828681956342e-05,
+      "loss": 0.9972708892822265,
       "step": 28200
     },
     {
-      "epoch": 0.6539559731049093,
-      "grad_norm": 2.8501274585723877,
-      "learning_rate": 1.9611771207515887e-05,
-      "loss": 0.6509,
+      "epoch": 0.32698128060952075,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.961798839458414e-05,
+      "loss": 1.008561553955078,
       "step": 28400
     },
     {
-      "epoch": 0.6585612968591692,
-      "grad_norm": 2.839573621749878,
-      "learning_rate": 1.9749930920143686e-05,
-      "loss": 0.6326,
+      "epoch": 0.3292839656842357,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.975614810721194e-05,
+      "loss": 0.9927769470214843,
       "step": 28600
     },
     {
-      "epoch": 0.6631666206134291,
-      "grad_norm": 2.609498977661133,
-      "learning_rate": 1.9888090632771484e-05,
-      "loss": 0.6436,
+      "epoch": 0.3315866507589506,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.9894307819839737e-05,
+      "loss": 0.9703852844238281,
       "step": 28800
     },
     {
-      "epoch": 0.667771944367689,
-      "grad_norm": 3.8914794921875,
-      "learning_rate": 2.0026250345399282e-05,
-      "loss": 0.6317,
+      "epoch": 0.33388933583366553,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0032467532467535e-05,
+      "loss": 0.9893107604980469,
       "step": 29000
     },
     {
-      "epoch": 0.672377268121949,
-      "grad_norm": 3.074779987335205,
-      "learning_rate": 2.016441005802708e-05,
-      "loss": 0.6473,
+      "epoch": 0.33619202090838046,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.017062724509533e-05,
+      "loss": 1.0032604217529297,
       "step": 29200
     },
     {
-      "epoch": 0.6769825918762089,
-      "grad_norm": 3.1600515842437744,
-      "learning_rate": 2.0302569770654875e-05,
-      "loss": 0.6418,
+      "epoch": 0.3384947059830954,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.0308786957723128e-05,
+      "loss": 0.9759930419921875,
       "step": 29400
     },
     {
-      "epoch": 0.6815879156304688,
-      "grad_norm": 3.190645694732666,
-      "learning_rate": 2.0440729483282673e-05,
-      "loss": 0.6289,
+      "epoch": 0.3407973910578104,
+      "grad_norm": 0.75,
+      "learning_rate": 2.0446946670350926e-05,
+      "loss": 0.9818559265136719,
       "step": 29600
     },
     {
-      "epoch": 0.6861932393847288,
-      "grad_norm": 3.1118228435516357,
-      "learning_rate": 2.057888919591047e-05,
-      "loss": 0.6495,
+      "epoch": 0.3431000761325253,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.0585106382978724e-05,
+      "loss": 0.9869992065429688,
       "step": 29800
     },
     {
-      "epoch": 0.6907985631389887,
-      "grad_norm": 3.027270555496216,
-      "learning_rate": 2.071635810997513e-05,
-      "loss": 0.6393,
+      "epoch": 0.3454027612072402,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.0723266095606522e-05,
+      "loss": 1.0209011077880858,
       "step": 30000
     },
     {
-      "epoch": 0.6954038868932486,
-      "grad_norm": 2.9312691688537598,
-      "learning_rate": 2.085451782260293e-05,
-      "loss": 0.645,
+      "epoch": 0.34770544628195516,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.086142580823432e-05,
+      "loss": 0.9973705291748047,
       "step": 30200
     },
     {
-      "epoch": 0.7000092106475085,
-      "grad_norm": 3.4496731758117676,
-      "learning_rate": 2.0992677535230728e-05,
-      "loss": 0.6522,
+      "epoch": 0.3500081313566701,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0999585520862118e-05,
+      "loss": 0.9845616149902344,
       "step": 30400
     },
     {
-      "epoch": 0.7046145344017685,
-      "grad_norm": 3.7810165882110596,
-      "learning_rate": 2.1130837247858526e-05,
-      "loss": 0.653,
+      "epoch": 0.352310816431385,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.1137745233489916e-05,
+      "loss": 0.9949858093261719,
       "step": 30600
     },
     {
-      "epoch": 0.7092198581560284,
-      "grad_norm": 2.8673808574676514,
-      "learning_rate": 2.126899696048632e-05,
-      "loss": 0.6249,
+      "epoch": 0.35461350150609994,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.1275904946117715e-05,
+      "loss": 1.0123253631591798,
       "step": 30800
     },
     {
-      "epoch": 0.7138251819102883,
-      "grad_norm": 2.933802843093872,
-      "learning_rate": 2.140715667311412e-05,
-      "loss": 0.6381,
+      "epoch": 0.35691618658081486,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.1414064658745513e-05,
+      "loss": 0.9973126220703125,
       "step": 31000
     },
     {
-      "epoch": 0.7184305056645482,
-      "grad_norm": 2.904672861099243,
-      "learning_rate": 2.1545316385741917e-05,
-      "loss": 0.643,
+      "epoch": 0.35921887165552985,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.155222437137331e-05,
+      "loss": 1.0002081298828125,
       "step": 31200
     },
     {
-      "epoch": 0.7230358294188082,
-      "grad_norm": 3.2614898681640625,
-      "learning_rate": 2.1683476098369715e-05,
-      "loss": 0.6401,
+      "epoch": 0.3615215567302448,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.1690384084001106e-05,
+      "loss": 1.0208023071289063,
       "step": 31400
     },
     {
-      "epoch": 0.7276411531730681,
-      "grad_norm": 4.081573963165283,
-      "learning_rate": 2.1821635810997513e-05,
-      "loss": 0.6401,
+      "epoch": 0.3638242418049597,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1828543796628904e-05,
+      "loss": 0.994631118774414,
       "step": 31600
     },
     {
-      "epoch": 0.732246476927328,
-      "grad_norm": 2.736985206604004,
-      "learning_rate": 2.195979552362531e-05,
-      "loss": 0.6373,
+      "epoch": 0.36612692687967463,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.1966703509256702e-05,
+      "loss": 1.0063703918457032,
       "step": 31800
     },
     {
-      "epoch": 0.736851800681588,
-      "grad_norm": 3.4728550910949707,
-      "learning_rate": 2.209795523625311e-05,
-      "loss": 0.6488,
+      "epoch": 0.36842961195438956,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.21048632218845e-05,
+      "loss": 0.9737712097167969,
       "step": 32000
     },
     {
-      "epoch": 0.7414571244358479,
-      "grad_norm": 2.758512496948242,
-      "learning_rate": 2.2236114948880908e-05,
-      "loss": 0.6492,
+      "epoch": 0.3707322970291045,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.2243022934512298e-05,
+      "loss": 0.9741205596923828,
       "step": 32200
     },
     {
-      "epoch": 0.7460624481901078,
-      "grad_norm": 3.4446616172790527,
-      "learning_rate": 2.2374274661508706e-05,
-      "loss": 0.6468,
+      "epoch": 0.3730349821038194,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.2381182647140096e-05,
+      "loss": 0.9894085693359375,
       "step": 32400
     },
     {
-      "epoch": 0.7506677719443677,
-      "grad_norm": 3.721409320831299,
-      "learning_rate": 2.2512434374136504e-05,
-      "loss": 0.6364,
+      "epoch": 0.3753376671785344,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.2519342359767894e-05,
+      "loss": 0.9812338256835937,
       "step": 32600
     },
     {
-      "epoch": 0.7552730956986277,
-      "grad_norm": 2.9497594833374023,
-      "learning_rate": 2.26505940867643e-05,
-      "loss": 0.6337,
+      "epoch": 0.3776403522532493,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.2657502072395693e-05,
+      "loss": 0.9896620178222656,
       "step": 32800
     },
     {
-      "epoch": 0.7598784194528876,
-      "grad_norm": 4.462852954864502,
-      "learning_rate": 2.2788753799392097e-05,
-      "loss": 0.6592,
+      "epoch": 0.37994303732796425,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.2795661785023487e-05,
+      "loss": 1.001156997680664,
       "step": 33000
     },
     {
-      "epoch": 0.7644837432071475,
-      "grad_norm": 2.867154836654663,
-      "learning_rate": 2.2926913512019895e-05,
-      "loss": 0.6472,
+      "epoch": 0.3822457224026792,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2933821497651285e-05,
+      "loss": 0.9897576141357421,
       "step": 33200
     },
     {
-      "epoch": 0.7690890669614074,
-      "grad_norm": 3.1840884685516357,
-      "learning_rate": 2.3065073224647693e-05,
-      "loss": 0.6454,
+      "epoch": 0.3845484074773941,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.307198121027908e-05,
+      "loss": 0.9884554290771485,
       "step": 33400
     },
     {
-      "epoch": 0.7736943907156673,
-      "grad_norm": 2.719569206237793,
-      "learning_rate": 2.320323293727549e-05,
-      "loss": 0.6565,
+      "epoch": 0.38685109255210903,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.3210140922906878e-05,
+      "loss": 0.9887247467041016,
       "step": 33600
     },
     {
-      "epoch": 0.7782997144699272,
-      "grad_norm": 3.1424355506896973,
-      "learning_rate": 2.334139264990329e-05,
-      "loss": 0.6333,
+      "epoch": 0.38915377762682396,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.3348300635534676e-05,
+      "loss": 1.0124942779541015,
       "step": 33800
     },
     {
-      "epoch": 0.7829050382241871,
-      "grad_norm": 3.147237539291382,
-      "learning_rate": 2.347886156396795e-05,
-      "loss": 0.6462,
+      "epoch": 0.3914564627015389,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.3486460348162475e-05,
+      "loss": 0.9673490905761719,
       "step": 34000
     },
     {
-      "epoch": 0.787510361978447,
-      "grad_norm": 3.6933629512786865,
-      "learning_rate": 2.3617021276595744e-05,
-      "loss": 0.642,
+      "epoch": 0.39375914777625387,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3624620060790273e-05,
+      "loss": 0.9783273315429688,
       "step": 34200
     },
     {
-      "epoch": 0.792115685732707,
-      "grad_norm": 3.1719932556152344,
-      "learning_rate": 2.3755180989223543e-05,
-      "loss": 0.6591,
+      "epoch": 0.3960618328509688,
+      "grad_norm": 0.625,
+      "learning_rate": 2.376277977341807e-05,
+      "loss": 1.0001426696777345,
       "step": 34400
     },
     {
-      "epoch": 0.7967210094869669,
-      "grad_norm": 4.069468021392822,
-      "learning_rate": 2.389334070185134e-05,
-      "loss": 0.6558,
+      "epoch": 0.3983645179256837,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.390093948604587e-05,
+      "loss": 0.9712574768066407,
       "step": 34600
     },
     {
-      "epoch": 0.8013263332412268,
-      "grad_norm": 3.7244644165039062,
-      "learning_rate": 2.403150041447914e-05,
-      "loss": 0.6586,
+      "epoch": 0.40066720300039865,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4039099198673667e-05,
+      "loss": 0.9945822143554688,
       "step": 34800
     },
     {
-      "epoch": 0.8059316569954867,
-      "grad_norm": 2.9359726905822754,
-      "learning_rate": 2.41689693285438e-05,
-      "loss": 0.6425,
+      "epoch": 0.4029698880751136,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4177258911301465e-05,
+      "loss": 0.9821370697021484,
       "step": 35000
     },
     {
-      "epoch": 0.8105369807497467,
-      "grad_norm": 3.2560675144195557,
-      "learning_rate": 2.4307129041171597e-05,
-      "loss": 0.6535,
+      "epoch": 0.4052725731498285,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4315418623929263e-05,
+      "loss": 0.9928865814208985,
       "step": 35200
     },
     {
-      "epoch": 0.8151423045040066,
-      "grad_norm": 3.1228187084198,
-      "learning_rate": 2.4445288753799396e-05,
-      "loss": 0.6346,
+      "epoch": 0.40757525822454344,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4453578336557058e-05,
+      "loss": 0.9809159851074218,
       "step": 35400
     },
     {
-      "epoch": 0.8197476282582665,
-      "grad_norm": 2.620872974395752,
-      "learning_rate": 2.458344846642719e-05,
-      "loss": 0.6412,
+      "epoch": 0.4098779432992584,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4591738049184856e-05,
+      "loss": 0.9863732147216797,
       "step": 35600
     },
     {
-      "epoch": 0.8243529520125265,
-      "grad_norm": 3.165461540222168,
-      "learning_rate": 2.472160817905499e-05,
-      "loss": 0.643,
+      "epoch": 0.41218062837397335,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4729897761812654e-05,
+      "loss": 1.0015038299560546,
       "step": 35800
     },
     {
-      "epoch": 0.8289582757667864,
-      "grad_norm": 3.4835033416748047,
-      "learning_rate": 2.4859767891682787e-05,
-      "loss": 0.6503,
+      "epoch": 0.4144833134486883,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4868057474440453e-05,
+      "loss": 0.9992890167236328,
       "step": 36000
     },
     {
-      "epoch": 0.8335635995210463,
-      "grad_norm": 3.1745879650115967,
-      "learning_rate": 2.4997927604310585e-05,
-      "loss": 0.6572,
+      "epoch": 0.4167859985234032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.500621718706825e-05,
+      "loss": 0.9928932952880859,
       "step": 36200
     },
     {
-      "epoch": 0.8381689232753062,
-      "grad_norm": 2.9764230251312256,
-      "learning_rate": 2.513608731693838e-05,
-      "loss": 0.6426,
+      "epoch": 0.41908868359811813,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.514437689969605e-05,
+      "loss": 0.9791075897216797,
       "step": 36400
     },
     {
-      "epoch": 0.8427742470295662,
-      "grad_norm": 2.6247684955596924,
-      "learning_rate": 2.5274247029566178e-05,
-      "loss": 0.6483,
+      "epoch": 0.42139136867283306,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.5282536612323847e-05,
+      "loss": 0.9951918792724609,
       "step": 36600
     },
     {
-      "epoch": 0.8473795707838261,
-      "grad_norm": 2.999863386154175,
-      "learning_rate": 2.5412406742193976e-05,
-      "loss": 0.6487,
+      "epoch": 0.423694053747548,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5420696324951645e-05,
+      "loss": 0.9908084869384766,
       "step": 36800
     },
     {
-      "epoch": 0.851984894538086,
-      "grad_norm": 6.160555839538574,
-      "learning_rate": 2.5550566454821774e-05,
-      "loss": 0.6596,
+      "epoch": 0.4259967388222629,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.5558856037579443e-05,
+      "loss": 0.979951171875,
       "step": 37000
     },
     {
-      "epoch": 0.8565902182923459,
-      "grad_norm": 3.588473320007324,
-      "learning_rate": 2.5688726167449572e-05,
-      "loss": 0.6562,
+      "epoch": 0.4282994238969779,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.569701575020724e-05,
+      "loss": 0.9870540618896484,
       "step": 37200
     },
     {
-      "epoch": 0.8611955420466059,
-      "grad_norm": 2.841158628463745,
-      "learning_rate": 2.582688588007737e-05,
-      "loss": 0.6634,
+      "epoch": 0.4306021089716928,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.583517546283504e-05,
+      "loss": 0.9819743347167968,
       "step": 37400
     },
     {
-      "epoch": 0.8658008658008658,
-      "grad_norm": 3.2244651317596436,
-      "learning_rate": 2.5965045592705168e-05,
-      "loss": 0.6436,
+      "epoch": 0.43290479404640775,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5973335175462834e-05,
+      "loss": 0.9861982727050781,
       "step": 37600
     },
     {
-      "epoch": 0.8704061895551257,
-      "grad_norm": 2.9466798305511475,
-      "learning_rate": 2.6103205305332963e-05,
-      "loss": 0.6551,
+      "epoch": 0.4352074791211227,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.6111494888090632e-05,
+      "loss": 0.9916728973388672,
       "step": 37800
     },
     {
-      "epoch": 0.8750115133093856,
-      "grad_norm": 3.141784191131592,
-      "learning_rate": 2.624136501796076e-05,
-      "loss": 0.6607,
+      "epoch": 0.4375101641958376,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.624965460071843e-05,
+      "loss": 0.9688997650146485,
       "step": 38000
     },
     {
-      "epoch": 0.8796168370636456,
-      "grad_norm": 3.590473175048828,
-      "learning_rate": 2.637952473058856e-05,
-      "loss": 0.6604,
+      "epoch": 0.43981284927055253,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.638781431334623e-05,
+      "loss": 0.9839357757568359,
       "step": 38200
     },
     {
-      "epoch": 0.8842221608179055,
-      "grad_norm": 3.251824378967285,
-      "learning_rate": 2.6517684443216357e-05,
-      "loss": 0.6656,
+      "epoch": 0.44211553434526746,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.6525974025974027e-05,
+      "loss": 0.9878810119628906,
       "step": 38400
     },
     {
-      "epoch": 0.8888274845721654,
-      "grad_norm": 2.8062353134155273,
-      "learning_rate": 2.6655844155844156e-05,
-      "loss": 0.6494,
+      "epoch": 0.44441821941998244,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.6664133738601825e-05,
+      "loss": 0.9786476135253906,
       "step": 38600
     },
     {
-      "epoch": 0.8934328083264254,
-      "grad_norm": 2.897364616394043,
-      "learning_rate": 2.6794003868471954e-05,
-      "loss": 0.6474,
+      "epoch": 0.44672090449469737,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.6802293451229623e-05,
+      "loss": 0.9965245056152344,
       "step": 38800
     },
     {
-      "epoch": 0.8980381320806853,
-      "grad_norm": 3.172089099884033,
-      "learning_rate": 2.6932163581099752e-05,
-      "loss": 0.6601,
+      "epoch": 0.4490235895694123,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.694045316385742e-05,
+      "loss": 0.9912705993652344,
       "step": 39000
     },
     {
-      "epoch": 0.9026434558349452,
-      "grad_norm": 3.3217780590057373,
-      "learning_rate": 2.707032329372755e-05,
-      "loss": 0.6426,
+      "epoch": 0.4513262746441272,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.707861287648522e-05,
+      "loss": 0.986620864868164,
       "step": 39200
     },
     {
-      "epoch": 0.9072487795892051,
-      "grad_norm": 3.3799307346343994,
-      "learning_rate": 2.7208483006355348e-05,
-      "loss": 0.6614,
+      "epoch": 0.45362895971884215,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7216772589113018e-05,
+      "loss": 0.9704518127441406,
       "step": 39400
     },
     {
-      "epoch": 0.9118541033434651,
-      "grad_norm": 3.2434334754943848,
-      "learning_rate": 2.7346642718983146e-05,
-      "loss": 0.6546,
+      "epoch": 0.4559316447935571,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7354932301740812e-05,
+      "loss": 0.9987767791748047,
       "step": 39600
     },
     {
-      "epoch": 0.916459427097725,
-      "grad_norm": 3.017298936843872,
-      "learning_rate": 2.7483420834484666e-05,
-      "loss": 0.6403,
+      "epoch": 0.458234329868272,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.749309201436861e-05,
+      "loss": 1.0047703552246094,
       "step": 39800
     },
     {
-      "epoch": 0.9210647508519849,
-      "grad_norm": 2.9930150508880615,
-      "learning_rate": 2.7621580547112464e-05,
-      "loss": 0.6527,
+      "epoch": 0.46053701494298693,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.763125172699641e-05,
+      "loss": 0.9702092742919922,
       "step": 40000
     },
     {
-      "epoch": 0.9256700746062448,
-      "grad_norm": 2.474975347518921,
-      "learning_rate": 2.7759740259740262e-05,
-      "loss": 0.6559,
+      "epoch": 0.4628397000177019,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.7769411439624207e-05,
+      "loss": 0.9662624359130859,
       "step": 40200
     },
     {
-      "epoch": 0.9302753983605048,
-      "grad_norm": 3.3437626361846924,
-      "learning_rate": 2.789789997236806e-05,
-      "loss": 0.6479,
+      "epoch": 0.46514238509241684,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.7907571152252005e-05,
+      "loss": 0.9837642669677734,
       "step": 40400
     },
     {
-      "epoch": 0.9348807221147647,
-      "grad_norm": 3.64098858833313,
-      "learning_rate": 2.8036059684995855e-05,
-      "loss": 0.6671,
+      "epoch": 0.46744507016713177,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.8045730864879803e-05,
+      "loss": 0.9509716796875,
       "step": 40600
     },
     {
-      "epoch": 0.9394860458690246,
-      "grad_norm": 3.154405355453491,
-      "learning_rate": 2.8174219397623653e-05,
-      "loss": 0.6498,
+      "epoch": 0.4697477552418467,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.81838905775076e-05,
+      "loss": 0.9896253204345703,
       "step": 40800
     },
     {
-      "epoch": 0.9440913696232845,
-      "grad_norm": 3.1763336658477783,
-      "learning_rate": 2.831237911025145e-05,
-      "loss": 0.6581,
+      "epoch": 0.4720504403165616,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.83220502901354e-05,
+      "loss": 0.9683509063720703,
       "step": 41000
     },
     {
-      "epoch": 0.9486966933775445,
-      "grad_norm": 3.8115978240966797,
-      "learning_rate": 2.845053882287925e-05,
-      "loss": 0.6544,
+      "epoch": 0.47435312539127655,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.8460210002763197e-05,
+      "loss": 0.9861078643798828,
       "step": 41200
     },
     {
-      "epoch": 0.9533020171318044,
-      "grad_norm": 3.476557731628418,
-      "learning_rate": 2.8588698535507047e-05,
-      "loss": 0.6607,
+      "epoch": 0.4766558104659915,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.8598369715390995e-05,
+      "loss": 1.0022505950927734,
       "step": 41400
     },
     {
-      "epoch": 0.9579073408860643,
-      "grad_norm": 3.1274263858795166,
-      "learning_rate": 2.8726858248134845e-05,
-      "loss": 0.6557,
+      "epoch": 0.4789584955407064,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.873652942801879e-05,
+      "loss": 0.9984149932861328,
       "step": 41600
     },
     {
-      "epoch": 0.9625126646403243,
-      "grad_norm": 3.3021888732910156,
-      "learning_rate": 2.8865017960762644e-05,
-      "loss": 0.6529,
+      "epoch": 0.4812611806154214,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.887468914064659e-05,
+      "loss": 0.990358657836914,
       "step": 41800
     },
     {
-      "epoch": 0.9671179883945842,
-      "grad_norm": 2.258256435394287,
-      "learning_rate": 2.9003177673390442e-05,
-      "loss": 0.6537,
+      "epoch": 0.4835638656901363,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9012848853274387e-05,
+      "loss": 0.9875827026367188,
       "step": 42000
     },
     {
-      "epoch": 0.9717233121488441,
-      "grad_norm": 3.9193155765533447,
-      "learning_rate": 2.914133738601824e-05,
-      "loss": 0.6521,
+      "epoch": 0.48586655076485125,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.9151008565902185e-05,
+      "loss": 0.9853507995605468,
       "step": 42200
     },
     {
-      "epoch": 0.976328635903104,
-      "grad_norm": 2.55462384223938,
-      "learning_rate": 2.9279497098646038e-05,
-      "loss": 0.661,
+      "epoch": 0.4881692358395662,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.9289168278529983e-05,
+      "loss": 0.9948759460449219,
       "step": 42400
     },
     {
-      "epoch": 0.980933959657364,
-      "grad_norm": 2.79758358001709,
-      "learning_rate": 2.9417656811273833e-05,
-      "loss": 0.6601,
+      "epoch": 0.4904719209142811,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.9427327991157778e-05,
+      "loss": 0.9923246765136718,
       "step": 42600
     },
     {
-      "epoch": 0.9855392834116239,
-      "grad_norm": 3.391768217086792,
-      "learning_rate": 2.955581652390163e-05,
-      "loss": 0.6502,
+      "epoch": 0.49277460598899603,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.9565487703785576e-05,
+      "loss": 0.9764729309082031,
       "step": 42800
     },
     {
-      "epoch": 0.9901446071658838,
-      "grad_norm": 2.6248672008514404,
-      "learning_rate": 2.969397623652943e-05,
-      "loss": 0.6678,
+      "epoch": 0.49507729106371096,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.9703647416413374e-05,
+      "loss": 0.9748880004882813,
       "step": 43000
     },
     {
-      "epoch": 0.9947499309201436,
-      "grad_norm": 2.8915598392486572,
-      "learning_rate": 2.9832135949157227e-05,
-      "loss": 0.6473,
+      "epoch": 0.49737997613842594,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9841807129041172e-05,
+      "loss": 1.002592544555664,
       "step": 43200
     },
     {
-      "epoch": 0.9993552546744036,
-      "grad_norm": 4.923104286193848,
-      "learning_rate": 2.9970295661785025e-05,
-      "loss": 0.6451,
+      "epoch": 0.49968266121314087,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.997996684166897e-05,
+      "loss": 1.0050338745117187,
       "step": 43400
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 0.6161314845085144,
-      "eval_runtime": 146.2574,
-      "eval_samples_per_second": 193.911,
-      "eval_steps_per_second": 12.122,
-      "step": 43428
-    },
-    {
-      "epoch": 1.0039605784286636,
-      "grad_norm": 3.111292600631714,
-      "learning_rate": 2.9994291822399326e-05,
-      "loss": 0.6622,
+      "epoch": 0.5019853462878557,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.999378281293175e-05,
+      "loss": 1.0200328826904297,
       "step": 43600
     },
     {
-      "epoch": 1.0085659021829234,
-      "grad_norm": 3.6699461936950684,
-      "learning_rate": 2.998702025857681e-05,
-      "loss": 0.6542,
+      "epoch": 0.5042880313625707,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9986511249109234e-05,
+      "loss": 0.998750228881836,
       "step": 43800
     },
     {
-      "epoch": 1.0131712259371835,
-      "grad_norm": 3.563108205795288,
-      "learning_rate": 2.997978505257341e-05,
-      "loss": 0.6444,
+      "epoch": 0.5065907164372857,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9979239685286716e-05,
+      "loss": 0.9793384552001954,
       "step": 44000
     },
     {
-      "epoch": 1.0177765496914433,
-      "grad_norm": 3.8684511184692383,
-      "learning_rate": 2.997251348875089e-05,
-      "loss": 0.6534,
+      "epoch": 0.5088934015120006,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.9971968121464202e-05,
+      "loss": 0.9925625610351563,
       "step": 44200
     },
     {
-      "epoch": 1.0223818734457033,
-      "grad_norm": 2.8675427436828613,
-      "learning_rate": 2.9965241924928375e-05,
-      "loss": 0.65,
+      "epoch": 0.5111960865867156,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.9964696557641688e-05,
+      "loss": 0.9737508392333984,
       "step": 44400
     },
     {
-      "epoch": 1.0269871971999631,
-      "grad_norm": 3.130798101425171,
-      "learning_rate": 2.995797036110586e-05,
-      "loss": 0.6566,
+      "epoch": 0.5134987716614304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.9957424993819174e-05,
+      "loss": 0.987296142578125,
       "step": 44600
     },
     {
-      "epoch": 1.0315925209542232,
-      "grad_norm": 3.477407932281494,
-      "learning_rate": 2.9950698797283344e-05,
-      "loss": 0.6701,
+      "epoch": 0.5158014567361454,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.9950153429996657e-05,
+      "loss": 0.9984680938720704,
       "step": 44800
     },
     {
-      "epoch": 1.036197844708483,
-      "grad_norm": 3.613179922103882,
-      "learning_rate": 2.994342723346083e-05,
-      "loss": 0.6636,
+      "epoch": 0.5181041418108603,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.994288186617414e-05,
+      "loss": 0.9554759979248046,
       "step": 45000
     },
     {
-      "epoch": 1.040803168462743,
-      "grad_norm": 4.168652057647705,
-      "learning_rate": 2.9936155669638315e-05,
-      "loss": 0.656,
+      "epoch": 0.5204068268855753,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.9935610302351625e-05,
+      "loss": 1.0087938690185547,
       "step": 45200
     },
     {
-      "epoch": 1.0454084922170028,
-      "grad_norm": 3.4714102745056152,
-      "learning_rate": 2.992892046363491e-05,
-      "loss": 0.666,
+      "epoch": 0.5227095119602901,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9928338738529108e-05,
+      "loss": 0.9874359130859375,
       "step": 45400
     },
     {
-      "epoch": 1.0500138159712629,
-      "grad_norm": 3.0046637058258057,
-      "learning_rate": 2.9921648899812396e-05,
-      "loss": 0.6547,
+      "epoch": 0.5250121970350051,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9921067174706593e-05,
+      "loss": 0.9804747009277344,
       "step": 45600
     },
     {
-      "epoch": 1.0546191397255227,
-      "grad_norm": 2.514378070831299,
-      "learning_rate": 2.991437733598988e-05,
-      "loss": 0.6592,
+      "epoch": 0.5273148821097201,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.991379561088408e-05,
+      "loss": 0.9834337615966797,
       "step": 45800
     },
     {
-      "epoch": 1.0592244634797827,
-      "grad_norm": 3.263707160949707,
-      "learning_rate": 2.990710577216736e-05,
-      "loss": 0.6544,
+      "epoch": 0.529617567184435,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.9906524047061562e-05,
+      "loss": 0.9804586791992187,
       "step": 46000
     },
     {
-      "epoch": 1.0638297872340425,
-      "grad_norm": 2.668816566467285,
-      "learning_rate": 2.9899834208344847e-05,
-      "loss": 0.6631,
+      "epoch": 0.53192025225915,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.9899252483239044e-05,
+      "loss": 0.9947004699707032,
       "step": 46200
     },
     {
-      "epoch": 1.0684351109883026,
-      "grad_norm": 3.3855276107788086,
-      "learning_rate": 2.9892562644522333e-05,
-      "loss": 0.6581,
+      "epoch": 0.5342229373338648,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.989198091941653e-05,
+      "loss": 0.9872706604003906,
       "step": 46400
     },
     {
-      "epoch": 1.0730404347425624,
-      "grad_norm": 3.417602062225342,
-      "learning_rate": 2.9885291080699816e-05,
-      "loss": 0.6686,
+      "epoch": 0.5365256224085798,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.9884709355594013e-05,
+      "loss": 0.9967252349853516,
       "step": 46600
     },
     {
-      "epoch": 1.0776457584968224,
-      "grad_norm": 3.3175718784332275,
-      "learning_rate": 2.9878019516877302e-05,
-      "loss": 0.6322,
+      "epoch": 0.5388283074832947,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.98774377917715e-05,
+      "loss": 0.9788603210449218,
       "step": 46800
     },
     {
-      "epoch": 1.0822510822510822,
-      "grad_norm": 3.263648748397827,
-      "learning_rate": 2.9870747953054784e-05,
-      "loss": 0.6482,
+      "epoch": 0.5411309925580097,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9870166227948985e-05,
+      "loss": 1.0005274963378907,
       "step": 47000
     },
     {
-      "epoch": 1.0868564060053423,
-      "grad_norm": 3.151909589767456,
-      "learning_rate": 2.9863476389232267e-05,
-      "loss": 0.6526,
+      "epoch": 0.5434336776327247,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9862894664126467e-05,
+      "loss": 0.9859010314941407,
       "step": 47200
     },
     {
-      "epoch": 1.091461729759602,
-      "grad_norm": 3.5228278636932373,
-      "learning_rate": 2.9856241183228865e-05,
-      "loss": 0.6459,
+      "epoch": 0.5457363627074395,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.9855623100303953e-05,
+      "loss": 0.997774429321289,
       "step": 47400
     },
     {
-      "epoch": 1.096067053513862,
-      "grad_norm": 2.767185926437378,
-      "learning_rate": 2.984896961940635e-05,
-      "loss": 0.6595,
+      "epoch": 0.5480390477821545,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.9848351536481436e-05,
+      "loss": 0.9882254028320312,
       "step": 47600
     },
     {
-      "epoch": 1.100672377268122,
-      "grad_norm": 3.287398099899292,
-      "learning_rate": 2.9841698055583834e-05,
-      "loss": 0.6625,
+      "epoch": 0.5503417328568694,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.984107997265892e-05,
+      "loss": 0.9773269653320312,
       "step": 47800
     },
     {
-      "epoch": 1.105277701022382,
-      "grad_norm": 3.132371425628662,
-      "learning_rate": 2.983442649176132e-05,
-      "loss": 0.6505,
+      "epoch": 0.5526444179315844,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9833808408836404e-05,
+      "loss": 0.9920169830322265,
       "step": 48000
     },
     {
-      "epoch": 1.1098830247766418,
-      "grad_norm": 2.7118616104125977,
-      "learning_rate": 2.9827154927938806e-05,
-      "loss": 0.6747,
+      "epoch": 0.5549471030062992,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.982653684501389e-05,
+      "loss": 0.9774995422363282,
       "step": 48200
     },
     {
-      "epoch": 1.1144883485309016,
-      "grad_norm": 2.996696949005127,
-      "learning_rate": 2.9819883364116285e-05,
-      "loss": 0.6633,
+      "epoch": 0.5572497880810142,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9819265281191376e-05,
+      "loss": 0.9765099334716797,
       "step": 48400
     },
     {
-      "epoch": 1.1190936722851617,
-      "grad_norm": 3.6714489459991455,
-      "learning_rate": 2.981261180029377e-05,
-      "loss": 0.6554,
+      "epoch": 0.5595524731557292,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.981199371736886e-05,
+      "loss": 0.9840308380126953,
       "step": 48600
     },
     {
-      "epoch": 1.1236989960394215,
-      "grad_norm": 2.80454421043396,
-      "learning_rate": 2.9805340236471257e-05,
-      "loss": 0.6641,
+      "epoch": 0.5618551582304441,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.980472215354634e-05,
+      "loss": 1.0011785125732422,
       "step": 48800
     },
     {
-      "epoch": 1.1283043197936815,
-      "grad_norm": 3.167177200317383,
-      "learning_rate": 2.979806867264874e-05,
-      "loss": 0.6508,
+      "epoch": 0.5641578433051591,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.9797450589723827e-05,
+      "loss": 0.9817601013183593,
       "step": 49000
     },
     {
-      "epoch": 1.1329096435479413,
-      "grad_norm": 3.2425365447998047,
-      "learning_rate": 2.9790797108826225e-05,
-      "loss": 0.6508,
+      "epoch": 0.5664605283798739,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.979017902590131e-05,
+      "loss": 1.0076644134521484,
       "step": 49200
     },
     {
-      "epoch": 1.1375149673022014,
-      "grad_norm": 2.834266424179077,
-      "learning_rate": 2.978352554500371e-05,
-      "loss": 0.6558,
+      "epoch": 0.5687632134545889,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.9782907462078795e-05,
+      "loss": 0.9878813171386719,
       "step": 49400
     },
     {
-      "epoch": 1.1421202910564612,
-      "grad_norm": 2.795851707458496,
-      "learning_rate": 2.9776253981181194e-05,
-      "loss": 0.6567,
+      "epoch": 0.5710658985293038,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.977563589825628e-05,
+      "loss": 0.9952576446533203,
       "step": 49600
     },
     {
-      "epoch": 1.1467256148107212,
-      "grad_norm": 2.940406084060669,
-      "learning_rate": 2.9768982417358676e-05,
-      "loss": 0.6605,
+      "epoch": 0.5733685836040188,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.9768364334433764e-05,
+      "loss": 0.9908863067626953,
       "step": 49800
     },
     {
-      "epoch": 1.151330938564981,
-      "grad_norm": 3.899711847305298,
-      "learning_rate": 2.9761710853536162e-05,
-      "loss": 0.6639,
+      "epoch": 0.5756712686787337,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.9761092770611246e-05,
+      "loss": 0.9856517791748047,
       "step": 50000
     },
     {
-      "epoch": 1.155936262319241,
-      "grad_norm": 2.7121920585632324,
-      "learning_rate": 2.9754475647532757e-05,
-      "loss": 0.6759,
+      "epoch": 0.5779739537534486,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9753821206788732e-05,
+      "loss": 1.0019198608398439,
       "step": 50200
     },
     {
-      "epoch": 1.1605415860735009,
-      "grad_norm": 3.220299005508423,
-      "learning_rate": 2.9747204083710243e-05,
-      "loss": 0.6514,
+      "epoch": 0.5802766388281636,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.9746549642966218e-05,
+      "loss": 0.9981635284423828,
       "step": 50400
     },
     {
-      "epoch": 1.165146909827761,
-      "grad_norm": 3.94079852104187,
-      "learning_rate": 2.973993251988773e-05,
-      "loss": 0.661,
+      "epoch": 0.5825793239028785,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.97392780791437e-05,
+      "loss": 1.0029985046386718,
       "step": 50600
     },
     {
-      "epoch": 1.1697522335820207,
-      "grad_norm": 3.2445309162139893,
-      "learning_rate": 2.973266095606521e-05,
-      "loss": 0.6682,
+      "epoch": 0.5848820089775935,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.9732006515321187e-05,
+      "loss": 1.0053339385986328,
       "step": 50800
     },
     {
-      "epoch": 1.1743575573362808,
-      "grad_norm": 3.0553812980651855,
-      "learning_rate": 2.9725389392242697e-05,
-      "loss": 0.6687,
+      "epoch": 0.5871846940523083,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.9724734951498672e-05,
+      "loss": 0.9716258239746094,
       "step": 51000
     },
     {
-      "epoch": 1.1789628810905406,
-      "grad_norm": 4.24676513671875,
-      "learning_rate": 2.971811782842018e-05,
-      "loss": 0.656,
+      "epoch": 0.5894873791270233,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.9717463387676155e-05,
+      "loss": 0.9866657257080078,
       "step": 51200
     },
     {
-      "epoch": 1.1835682048448006,
-      "grad_norm": 3.447143316268921,
-      "learning_rate": 2.9710882622416775e-05,
-      "loss": 0.6562,
+      "epoch": 0.5917900642017382,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.9710191823853638e-05,
+      "loss": 0.9880976104736328,
       "step": 51400
     },
     {
-      "epoch": 1.1881735285990604,
-      "grad_norm": 2.6866791248321533,
-      "learning_rate": 2.970361105859426e-05,
-      "loss": 0.6684,
+      "epoch": 0.5940927492764532,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.9702920260031123e-05,
+      "loss": 0.9747943878173828,
       "step": 51600
     },
     {
-      "epoch": 1.1927788523533205,
-      "grad_norm": 3.1234331130981445,
-      "learning_rate": 2.9696339494771747e-05,
-      "loss": 0.658,
+      "epoch": 0.5963954343511682,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9695648696208606e-05,
+      "loss": 0.9976082611083984,
       "step": 51800
     },
     {
-      "epoch": 1.1973841761075803,
-      "grad_norm": 3.407177686691284,
-      "learning_rate": 2.968906793094923e-05,
-      "loss": 0.6628,
+      "epoch": 0.598698119425883,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.9688377132386092e-05,
+      "loss": 1.0007020568847655,
       "step": 52000
     },
     {
-      "epoch": 1.2019894998618403,
-      "grad_norm": 3.587676763534546,
-      "learning_rate": 2.9681796367126715e-05,
-      "loss": 0.6493,
+      "epoch": 0.601000804500598,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.9681105568563578e-05,
+      "loss": 0.9923267364501953,
       "step": 52200
     },
     {
-      "epoch": 1.2065948236161002,
-      "grad_norm": 2.697775363922119,
-      "learning_rate": 2.96745248033042e-05,
-      "loss": 0.6637,
+      "epoch": 0.6033034895753129,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.967383400474106e-05,
+      "loss": 1.0014637756347655,
       "step": 52400
     },
     {
-      "epoch": 1.2112001473703602,
-      "grad_norm": 3.507350206375122,
-      "learning_rate": 2.9667253239481684e-05,
-      "loss": 0.6528,
+      "epoch": 0.6056061746500279,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9666562440918543e-05,
+      "loss": 0.9647525024414062,
       "step": 52600
     },
     {
-      "epoch": 1.21580547112462,
-      "grad_norm": 3.143338203430176,
-      "learning_rate": 2.9659981675659166e-05,
-      "loss": 0.6521,
+      "epoch": 0.6079088597247427,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.965929087709603e-05,
+      "loss": 0.9859085083007812,
       "step": 52800
     },
     {
-      "epoch": 1.22041079487888,
-      "grad_norm": 2.8852453231811523,
-      "learning_rate": 2.9652710111836652e-05,
-      "loss": 0.6595,
+      "epoch": 0.6102115447994577,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.965201931327351e-05,
+      "loss": 0.9939171600341797,
       "step": 53000
     },
     {
-      "epoch": 1.2250161186331399,
-      "grad_norm": 3.039896249771118,
-      "learning_rate": 2.9645438548014135e-05,
-      "loss": 0.6542,
+      "epoch": 0.6125142298741727,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.9644747749450997e-05,
+      "loss": 0.9924463653564453,
       "step": 53200
     },
     {
-      "epoch": 1.2296214423874,
-      "grad_norm": 2.8866536617279053,
-      "learning_rate": 2.963816698419162e-05,
-      "loss": 0.6535,
+      "epoch": 0.6148169149488876,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.9637476185628483e-05,
+      "loss": 0.9631312561035156,
       "step": 53400
     },
     {
-      "epoch": 1.2342267661416597,
-      "grad_norm": 2.5456418991088867,
-      "learning_rate": 2.9630895420369107e-05,
-      "loss": 0.664,
+      "epoch": 0.6171196000236026,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.963020462180597e-05,
+      "loss": 0.9812735748291016,
       "step": 53600
     },
     {
-      "epoch": 1.2388320898959198,
-      "grad_norm": 2.5279481410980225,
-      "learning_rate": 2.9623623856546593e-05,
-      "loss": 0.6603,
+      "epoch": 0.6194222850983174,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.962293305798345e-05,
+      "loss": 1.0052700805664063,
       "step": 53800
     },
     {
-      "epoch": 1.2434374136501796,
-      "grad_norm": 3.0540003776550293,
-      "learning_rate": 2.9616352292724072e-05,
-      "loss": 0.6626,
+      "epoch": 0.6217249701730324,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.9615661494160934e-05,
+      "loss": 0.97032958984375,
       "step": 54000
     },
     {
-      "epoch": 1.2480427374044396,
-      "grad_norm": 2.9297878742218018,
-      "learning_rate": 2.9609080728901558e-05,
-      "loss": 0.6612,
+      "epoch": 0.6240276552477473,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.960838993033842e-05,
+      "loss": 1.013146286010742,
       "step": 54200
     },
     {
-      "epoch": 1.2526480611586994,
-      "grad_norm": 3.4049458503723145,
-      "learning_rate": 2.9601809165079044e-05,
-      "loss": 0.6509,
+      "epoch": 0.6263303403224623,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.9601118366515903e-05,
+      "loss": 0.9846820831298828,
       "step": 54400
     },
     {
-      "epoch": 1.2572533849129595,
-      "grad_norm": 4.160104274749756,
-      "learning_rate": 2.9594537601256526e-05,
-      "loss": 0.6533,
+      "epoch": 0.6286330253971772,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.959384680269339e-05,
+      "loss": 0.9867259979248046,
       "step": 54600
     },
     {
-      "epoch": 1.2618587086672193,
-      "grad_norm": 3.6435911655426025,
-      "learning_rate": 2.9587266037434012e-05,
-      "loss": 0.6703,
+      "epoch": 0.6309357104718921,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.9586575238870874e-05,
+      "loss": 0.9823355102539062,
       "step": 54800
     },
     {
-      "epoch": 1.2664640324214793,
-      "grad_norm": 4.371586322784424,
-      "learning_rate": 2.9579994473611498e-05,
-      "loss": 0.6507,
+      "epoch": 0.6332383955466071,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9579303675048357e-05,
+      "loss": 0.9936563873291016,
       "step": 55000
     },
     {
-      "epoch": 1.2710693561757391,
-      "grad_norm": 2.840867757797241,
-      "learning_rate": 2.9572722909788977e-05,
-      "loss": 0.6644,
+      "epoch": 0.635541080621322,
+      "grad_norm": 0.5,
+      "learning_rate": 2.957203211122584e-05,
+      "loss": 0.9837480926513672,
       "step": 55200
     },
     {
-      "epoch": 1.275674679929999,
-      "grad_norm": 4.047842502593994,
-      "learning_rate": 2.9565451345966463e-05,
-      "loss": 0.6752,
+      "epoch": 0.637843765696037,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.9564760547403325e-05,
+      "loss": 0.9884334564208984,
       "step": 55400
     },
     {
-      "epoch": 1.280280003684259,
-      "grad_norm": 4.21028470993042,
-      "learning_rate": 2.955817978214395e-05,
-      "loss": 0.6582,
+      "epoch": 0.6401464507707518,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9557488983580808e-05,
+      "loss": 0.9931166839599609,
       "step": 55600
     },
     {
-      "epoch": 1.284885327438519,
-      "grad_norm": 2.9886422157287598,
-      "learning_rate": 2.955090821832143e-05,
-      "loss": 0.6634,
+      "epoch": 0.6424491358454668,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.9550217419758294e-05,
+      "loss": 0.9690689086914063,
       "step": 55800
     },
     {
-      "epoch": 1.2894906511927788,
-      "grad_norm": 3.3771917819976807,
-      "learning_rate": 2.9543636654498917e-05,
-      "loss": 0.6485,
+      "epoch": 0.6447518209201817,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.954294585593578e-05,
+      "loss": 0.9920832824707031,
       "step": 56000
     },
     {
-      "epoch": 1.2940959749470387,
-      "grad_norm": 3.0930802822113037,
-      "learning_rate": 2.9536365090676403e-05,
-      "loss": 0.6665,
+      "epoch": 0.6470545059948967,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.9535674292113266e-05,
+      "loss": 0.9990323638916015,
       "step": 56200
     },
     {
-      "epoch": 1.2987012987012987,
-      "grad_norm": 3.460557222366333,
-      "learning_rate": 2.952909352685389e-05,
-      "loss": 0.6703,
+      "epoch": 0.6493571910696117,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.9528402728290745e-05,
+      "loss": 0.9787998199462891,
       "step": 56400
     },
     {
-      "epoch": 1.3033066224555587,
-      "grad_norm": 3.846440076828003,
-      "learning_rate": 2.9521858320850484e-05,
-      "loss": 0.6626,
+      "epoch": 0.6516598761443265,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.952113116446823e-05,
+      "loss": 1.0044151306152345,
       "step": 56600
     },
     {
-      "epoch": 1.3079119462098185,
-      "grad_norm": 3.103189468383789,
-      "learning_rate": 2.9514586757027967e-05,
-      "loss": 0.6732,
+      "epoch": 0.6539625612190415,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.9513859600645717e-05,
+      "loss": 0.9847600555419922,
       "step": 56800
     },
     {
-      "epoch": 1.3125172699640784,
-      "grad_norm": 3.5162618160247803,
-      "learning_rate": 2.950731519320545e-05,
-      "loss": 0.6659,
+      "epoch": 0.6562652462937564,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.95065880368232e-05,
+      "loss": 0.9836011505126954,
       "step": 57000
     },
     {
-      "epoch": 1.3171225937183384,
-      "grad_norm": 3.511310338973999,
-      "learning_rate": 2.9500079987202048e-05,
-      "loss": 0.668,
+      "epoch": 0.6585679313684714,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.9499316473000685e-05,
+      "loss": 0.9912583923339844,
       "step": 57200
     },
     {
-      "epoch": 1.3217279174725984,
-      "grad_norm": 3.5012547969818115,
-      "learning_rate": 2.9492808423379534e-05,
-      "loss": 0.6693,
+      "epoch": 0.6608706164431862,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.949204490917817e-05,
+      "loss": 0.9929079437255859,
       "step": 57400
     },
     {
-      "epoch": 1.3263332412268583,
-      "grad_norm": 3.130657434463501,
-      "learning_rate": 2.9485536859557016e-05,
-      "loss": 0.6678,
+      "epoch": 0.6631733015179012,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.9484773345355653e-05,
+      "loss": 0.9952792358398438,
       "step": 57600
     },
     {
-      "epoch": 1.330938564981118,
-      "grad_norm": 3.0127837657928467,
-      "learning_rate": 2.9478265295734502e-05,
-      "loss": 0.6606,
+      "epoch": 0.6654759865926162,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.9477501781533136e-05,
+      "loss": 0.9805123138427735,
       "step": 57800
     },
     {
-      "epoch": 1.335543888735378,
-      "grad_norm": 2.949445962905884,
-      "learning_rate": 2.947099373191199e-05,
-      "loss": 0.6663,
+      "epoch": 0.6677786716673311,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.9470230217710622e-05,
+      "loss": 0.989420166015625,
       "step": 58000
     },
     {
-      "epoch": 1.3401492124896381,
-      "grad_norm": 3.239060640335083,
-      "learning_rate": 2.9463722168089467e-05,
-      "loss": 0.659,
+      "epoch": 0.670081356742046,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.9462958653888104e-05,
+      "loss": 0.9806766510009766,
       "step": 58200
     },
     {
-      "epoch": 1.344754536243898,
-      "grad_norm": 3.4041852951049805,
-      "learning_rate": 2.9456450604266953e-05,
-      "loss": 0.6652,
+      "epoch": 0.6723840418167609,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.945568709006559e-05,
+      "loss": 1.0109115600585938,
       "step": 58400
     },
     {
-      "epoch": 1.3493598599981578,
-      "grad_norm": 2.7172391414642334,
-      "learning_rate": 2.944917904044444e-05,
-      "loss": 0.643,
+      "epoch": 0.6746867268914759,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9448415526243076e-05,
+      "loss": 0.9819956207275391,
       "step": 58600
     },
     {
-      "epoch": 1.3539651837524178,
-      "grad_norm": 3.234712600708008,
-      "learning_rate": 2.9441907476621922e-05,
-      "loss": 0.6552,
+      "epoch": 0.6769894119661908,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.944114396242056e-05,
+      "loss": 0.9844336700439453,
       "step": 58800
     },
     {
-      "epoch": 1.3585705075066778,
-      "grad_norm": 2.3830983638763428,
-      "learning_rate": 2.9434635912799408e-05,
-      "loss": 0.6704,
+      "epoch": 0.6792920970409058,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.943387239859804e-05,
+      "loss": 0.9926036071777343,
       "step": 59000
     },
     {
-      "epoch": 1.3631758312609377,
-      "grad_norm": 3.203972578048706,
-      "learning_rate": 2.9427364348976894e-05,
-      "loss": 0.6609,
+      "epoch": 0.6815947821156207,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9426600834775527e-05,
+      "loss": 0.9849551391601562,
       "step": 59200
     },
     {
-      "epoch": 1.3677811550151975,
-      "grad_norm": 3.8763749599456787,
-      "learning_rate": 2.9420092785154373e-05,
-      "loss": 0.6641,
+      "epoch": 0.6838974671903356,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.9419329270953013e-05,
+      "loss": 1.0088782501220703,
       "step": 59400
     },
     {
-      "epoch": 1.3723864787694575,
-      "grad_norm": 3.5821003913879395,
-      "learning_rate": 2.941282122133186e-05,
-      "loss": 0.6577,
+      "epoch": 0.6862001522650506,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.9412057707130496e-05,
+      "loss": 0.9990327453613281,
       "step": 59600
     },
     {
-      "epoch": 1.3769918025237173,
-      "grad_norm": 3.558241844177246,
-      "learning_rate": 2.9405549657509345e-05,
-      "loss": 0.6564,
+      "epoch": 0.6885028373397655,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.940478614330798e-05,
+      "loss": 0.99803955078125,
       "step": 59800
     },
     {
-      "epoch": 1.3815971262779774,
-      "grad_norm": 2.6265041828155518,
-      "learning_rate": 2.939827809368683e-05,
-      "loss": 0.6497,
+      "epoch": 0.6908055224144805,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.9397514579485467e-05,
+      "loss": 0.9917265319824219,
       "step": 60000
     },
     {
-      "epoch": 1.3862024500322372,
-      "grad_norm": 3.8873655796051025,
-      "learning_rate": 2.9391006529864313e-05,
-      "loss": 0.6618,
+      "epoch": 0.6931082074891953,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.9390243015662947e-05,
+      "loss": 0.9893582916259765,
       "step": 60200
     },
     {
-      "epoch": 1.3908077737864972,
-      "grad_norm": 3.367459774017334,
-      "learning_rate": 2.93837349660418e-05,
-      "loss": 0.6427,
+      "epoch": 0.6954108925639103,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.9382971451840432e-05,
+      "loss": 0.9903961944580079,
       "step": 60400
     },
     {
-      "epoch": 1.395413097540757,
-      "grad_norm": 3.5943410396575928,
-      "learning_rate": 2.9376463402219285e-05,
-      "loss": 0.6767,
+      "epoch": 0.6977135776386253,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.937569988801792e-05,
+      "loss": 0.99945556640625,
       "step": 60600
     },
     {
-      "epoch": 1.400018421295017,
-      "grad_norm": 3.536818742752075,
-      "learning_rate": 2.9369191838396764e-05,
-      "loss": 0.6474,
+      "epoch": 0.7000162627133402,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.93684283241954e-05,
+      "loss": 1.0101787567138671,
       "step": 60800
     },
     {
-      "epoch": 1.404623745049277,
-      "grad_norm": 3.1541073322296143,
-      "learning_rate": 2.9361956632393363e-05,
-      "loss": 0.657,
+      "epoch": 0.7023189477880551,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.9361156760372887e-05,
+      "loss": 1.004135513305664,
       "step": 61000
     },
     {
-      "epoch": 1.409229068803537,
-      "grad_norm": 3.130194902420044,
-      "learning_rate": 2.9354685068570845e-05,
-      "loss": 0.6623,
+      "epoch": 0.70462163286277,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.9353885196550373e-05,
+      "loss": 0.9818244934082031,
       "step": 61200
     },
     {
-      "epoch": 1.4138343925577967,
-      "grad_norm": 3.6014018058776855,
-      "learning_rate": 2.934741350474833e-05,
-      "loss": 0.6544,
+      "epoch": 0.706924317937485,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9346613632727855e-05,
+      "loss": 0.9904888916015625,
       "step": 61400
     },
     {
-      "epoch": 1.4184397163120568,
-      "grad_norm": 3.4205713272094727,
-      "learning_rate": 2.9340141940925817e-05,
-      "loss": 0.663,
+      "epoch": 0.7092270030121999,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.9339342068905338e-05,
+      "loss": 0.966075668334961,
       "step": 61600
     },
     {
-      "epoch": 1.4230450400663166,
-      "grad_norm": 2.7789885997772217,
-      "learning_rate": 2.9332870377103303e-05,
-      "loss": 0.6607,
+      "epoch": 0.7115296880869149,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.9332070505082824e-05,
+      "loss": 0.981036376953125,
       "step": 61800
     },
     {
-      "epoch": 1.4276503638205766,
-      "grad_norm": 2.774611234664917,
-      "learning_rate": 2.9325598813280785e-05,
-      "loss": 0.6559,
+      "epoch": 0.7138323731616297,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.9324798941260306e-05,
+      "loss": 0.9956706237792968,
       "step": 62000
     },
     {
-      "epoch": 1.4322556875748365,
-      "grad_norm": 3.5750908851623535,
-      "learning_rate": 2.9318327249458268e-05,
-      "loss": 0.6546,
+      "epoch": 0.7161350582363447,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9317527377437792e-05,
+      "loss": 0.9858502197265625,
       "step": 62200
     },
     {
-      "epoch": 1.4368610113290965,
-      "grad_norm": 3.758307456970215,
-      "learning_rate": 2.9311055685635754e-05,
-      "loss": 0.6681,
+      "epoch": 0.7184377433110597,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9310255813615278e-05,
+      "loss": 0.99573974609375,
       "step": 62400
     },
     {
-      "epoch": 1.4414663350833563,
-      "grad_norm": 2.9125349521636963,
-      "learning_rate": 2.9303784121813236e-05,
-      "loss": 0.6504,
+      "epoch": 0.7207404283857746,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.9302984249792764e-05,
+      "loss": 0.9763796234130859,
       "step": 62600
     },
     {
-      "epoch": 1.4460716588376163,
-      "grad_norm": 3.2125773429870605,
-      "learning_rate": 2.9296512557990722e-05,
-      "loss": 0.6535,
+      "epoch": 0.7230431134604896,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.9295712685970243e-05,
+      "loss": 0.9572287750244141,
       "step": 62800
     },
     {
-      "epoch": 1.4506769825918762,
-      "grad_norm": 3.11696457862854,
-      "learning_rate": 2.9289240994168208e-05,
-      "loss": 0.6425,
+      "epoch": 0.7253457985352044,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.928844112214773e-05,
+      "loss": 1.0051393127441406,
       "step": 63000
     },
     {
-      "epoch": 1.4552823063461362,
-      "grad_norm": 4.079625606536865,
-      "learning_rate": 2.928196943034569e-05,
-      "loss": 0.6641,
+      "epoch": 0.7276484836099194,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9281169558325215e-05,
+      "loss": 0.9663025665283204,
       "step": 63200
     },
     {
-      "epoch": 1.459887630100396,
-      "grad_norm": 2.697766065597534,
-      "learning_rate": 2.9274697866523177e-05,
-      "loss": 0.6568,
+      "epoch": 0.7299511686846343,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.9273897994502697e-05,
+      "loss": 0.9826088714599609,
       "step": 63400
     },
     {
-      "epoch": 1.464492953854656,
-      "grad_norm": 2.6162514686584473,
-      "learning_rate": 2.926742630270066e-05,
-      "loss": 0.659,
+      "epoch": 0.7322538537593493,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9266626430680183e-05,
+      "loss": 0.9884089660644532,
       "step": 63600
     },
     {
-      "epoch": 1.4690982776089159,
-      "grad_norm": 3.08160138130188,
-      "learning_rate": 2.926015473887814e-05,
-      "loss": 0.6527,
+      "epoch": 0.7345565388340642,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.925935486685767e-05,
+      "loss": 0.9830115509033203,
       "step": 63800
     },
     {
-      "epoch": 1.473703601363176,
-      "grad_norm": 3.38775897026062,
-      "learning_rate": 2.9252883175055628e-05,
-      "loss": 0.6614,
+      "epoch": 0.7368592239087791,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.9252083303035152e-05,
+      "loss": 0.9935911560058593,
       "step": 64000
     },
     {
-      "epoch": 1.4783089251174357,
-      "grad_norm": 3.066603183746338,
-      "learning_rate": 2.9245611611233113e-05,
-      "loss": 0.6607,
+      "epoch": 0.7391619089834941,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.9244811739212634e-05,
+      "loss": 0.9883312225341797,
       "step": 64200
     },
     {
-      "epoch": 1.4829142488716958,
-      "grad_norm": 2.781545400619507,
-      "learning_rate": 2.92383400474106e-05,
-      "loss": 0.6529,
+      "epoch": 0.741464594058209,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.923754017539012e-05,
+      "loss": 0.9756713104248047,
       "step": 64400
     },
     {
-      "epoch": 1.4875195726259556,
-      "grad_norm": 3.7093193531036377,
-      "learning_rate": 2.9231068483588082e-05,
-      "loss": 0.6507,
+      "epoch": 0.743767279132924,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9230268611567603e-05,
+      "loss": 0.996288070678711,
       "step": 64600
     },
     {
-      "epoch": 1.4921248963802154,
-      "grad_norm": 3.037850856781006,
-      "learning_rate": 2.9223796919765564e-05,
-      "loss": 0.6593,
+      "epoch": 0.7460699642076388,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.922299704774509e-05,
+      "loss": 0.9647946166992187,
       "step": 64800
     },
     {
-      "epoch": 1.4967302201344754,
-      "grad_norm": 3.192322254180908,
-      "learning_rate": 2.921652535594305e-05,
-      "loss": 0.6561,
+      "epoch": 0.7483726492823538,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.9215725483922575e-05,
+      "loss": 0.9827190399169922,
       "step": 65000
     },
     {
-      "epoch": 1.5013355438887355,
-      "grad_norm": 3.157817840576172,
-      "learning_rate": 2.9209253792120533e-05,
-      "loss": 0.6488,
+      "epoch": 0.7506753343570688,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.920845392010006e-05,
+      "loss": 0.9791514587402343,
       "step": 65200
     },
     {
-      "epoch": 1.5059408676429953,
-      "grad_norm": 3.132276773452759,
-      "learning_rate": 2.920198222829802e-05,
-      "loss": 0.6549,
+      "epoch": 0.7529780194317837,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.920118235627754e-05,
+      "loss": 0.9852659606933594,
       "step": 65400
     },
     {
-      "epoch": 1.510546191397255,
-      "grad_norm": 3.7865426540374756,
-      "learning_rate": 2.9194710664475505e-05,
-      "loss": 0.6534,
+      "epoch": 0.7552807045064986,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.9193910792455026e-05,
+      "loss": 0.9933679962158203,
       "step": 65600
     },
     {
-      "epoch": 1.5151515151515151,
-      "grad_norm": 3.1854352951049805,
-      "learning_rate": 2.9187439100652987e-05,
-      "loss": 0.6592,
+      "epoch": 0.7575833895812135,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.918663922863251e-05,
+      "loss": 1.0145327758789062,
       "step": 65800
     },
     {
-      "epoch": 1.5197568389057752,
-      "grad_norm": 2.9768927097320557,
-      "learning_rate": 2.918016753683047e-05,
-      "loss": 0.6581,
+      "epoch": 0.7598860746559285,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.9179367664809994e-05,
+      "loss": 0.9841469573974609,
       "step": 66000
     },
     {
-      "epoch": 1.524362162660035,
-      "grad_norm": 3.467859983444214,
-      "learning_rate": 2.9172895973007956e-05,
-      "loss": 0.6599,
+      "epoch": 0.7621887597306434,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.917209610098748e-05,
+      "loss": 1.0026850128173828,
       "step": 66200
     },
     {
-      "epoch": 1.5289674864142948,
-      "grad_norm": 3.409672975540161,
-      "learning_rate": 2.9165624409185438e-05,
-      "loss": 0.6589,
+      "epoch": 0.7644914448053584,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.9164824537164966e-05,
+      "loss": 0.9971443939208985,
       "step": 66400
     },
     {
-      "epoch": 1.5335728101685548,
-      "grad_norm": 3.304276943206787,
-      "learning_rate": 2.9158352845362924e-05,
-      "loss": 0.6619,
+      "epoch": 0.7667941298800732,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.9157552973342445e-05,
+      "loss": 0.9800926971435547,
       "step": 66600
     },
     {
-      "epoch": 1.5381781339228149,
-      "grad_norm": 2.9806504249572754,
-      "learning_rate": 2.915108128154041e-05,
-      "loss": 0.6686,
+      "epoch": 0.7690968149547882,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.915028140951993e-05,
+      "loss": 0.9816303253173828,
       "step": 66800
     },
     {
-      "epoch": 1.5427834576770747,
-      "grad_norm": 3.9188714027404785,
-      "learning_rate": 2.9143846075537005e-05,
-      "loss": 0.6599,
+      "epoch": 0.7713995000295032,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.9143009845697417e-05,
+      "loss": 0.9932497406005859,
       "step": 67000
     },
     {
-      "epoch": 1.5473887814313345,
-      "grad_norm": 3.01717209815979,
-      "learning_rate": 2.913657451171449e-05,
-      "loss": 0.6603,
+      "epoch": 0.7737021851042181,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.91357382818749e-05,
+      "loss": 0.9947916412353516,
       "step": 67200
     },
     {
-      "epoch": 1.5519941051855946,
-      "grad_norm": 3.1575047969818115,
-      "learning_rate": 2.9129302947891977e-05,
-      "loss": 0.6532,
+      "epoch": 0.776004870178933,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.9128466718052385e-05,
+      "loss": 0.9930197143554688,
       "step": 67400
     },
     {
-      "epoch": 1.5565994289398546,
-      "grad_norm": 2.840865135192871,
-      "learning_rate": 2.9122031384069456e-05,
-      "loss": 0.6537,
+      "epoch": 0.7783075552536479,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.912119515422987e-05,
+      "loss": 0.9941978454589844,
       "step": 67600
     },
     {
-      "epoch": 1.5612047526941144,
-      "grad_norm": 3.181452989578247,
-      "learning_rate": 2.9114759820246942e-05,
-      "loss": 0.6605,
+      "epoch": 0.7806102403283629,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9113923590407354e-05,
+      "loss": 0.9890399932861328,
       "step": 67800
     },
     {
-      "epoch": 1.5658100764483742,
-      "grad_norm": 3.727302074432373,
-      "learning_rate": 2.9107488256424428e-05,
-      "loss": 0.6553,
+      "epoch": 0.7829129254030778,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.9106652026584836e-05,
+      "loss": 0.983370361328125,
       "step": 68000
     },
     {
-      "epoch": 1.5704154002026343,
-      "grad_norm": 3.4746217727661133,
-      "learning_rate": 2.910021669260191e-05,
-      "loss": 0.6603,
+      "epoch": 0.7852156104777928,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.9099380462762322e-05,
+      "loss": 0.9984974670410156,
       "step": 68200
     },
     {
-      "epoch": 1.5750207239568943,
-      "grad_norm": 3.0356459617614746,
-      "learning_rate": 2.9092945128779396e-05,
-      "loss": 0.6705,
+      "epoch": 0.7875182955525077,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.9092108898939808e-05,
+      "loss": 0.9801548004150391,
       "step": 68400
     },
     {
-      "epoch": 1.5796260477111541,
-      "grad_norm": 4.231008529663086,
-      "learning_rate": 2.9085673564956882e-05,
-      "loss": 0.6509,
+      "epoch": 0.7898209806272226,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.908483733511729e-05,
+      "loss": 0.989987564086914,
       "step": 68600
     },
     {
-      "epoch": 1.584231371465414,
-      "grad_norm": 2.5000898838043213,
-      "learning_rate": 2.907840200113436e-05,
-      "loss": 0.6577,
+      "epoch": 0.7921236657019376,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.9077565771294776e-05,
+      "loss": 0.9926038360595704,
       "step": 68800
     },
     {
-      "epoch": 1.588836695219674,
-      "grad_norm": 3.202843427658081,
-      "learning_rate": 2.9071130437311847e-05,
-      "loss": 0.6515,
+      "epoch": 0.7944263507766525,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.9070294207472262e-05,
+      "loss": 0.9818599700927735,
       "step": 69000
     },
     {
-      "epoch": 1.593442018973934,
-      "grad_norm": 3.5553150177001953,
-      "learning_rate": 2.9063858873489333e-05,
-      "loss": 0.6569,
+      "epoch": 0.7967290358513675,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.906302264364974e-05,
+      "loss": 0.9795211029052734,
       "step": 69200
     },
     {
-      "epoch": 1.5980473427281938,
-      "grad_norm": 3.006439208984375,
-      "learning_rate": 2.905658730966682e-05,
-      "loss": 0.6556,
+      "epoch": 0.7990317209260823,
+      "grad_norm": 1.0,
+      "learning_rate": 2.9055751079827227e-05,
+      "loss": 0.9876821899414062,
       "step": 69400
     },
     {
-      "epoch": 1.6026526664824536,
-      "grad_norm": 2.8993167877197266,
-      "learning_rate": 2.9049352103663414e-05,
-      "loss": 0.6602,
+      "epoch": 0.8013344060007973,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.9048479516004713e-05,
+      "loss": 0.9914775085449219,
       "step": 69600
     },
     {
-      "epoch": 1.6072579902367137,
-      "grad_norm": 3.5758256912231445,
-      "learning_rate": 2.90420805398409e-05,
-      "loss": 0.6512,
+      "epoch": 0.8036370910755123,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.9041207952182196e-05,
+      "loss": 0.9936875152587891,
       "step": 69800
     },
     {
-      "epoch": 1.6118633139909737,
-      "grad_norm": 2.4730918407440186,
-      "learning_rate": 2.9034808976018383e-05,
-      "loss": 0.6479,
+      "epoch": 0.8059397761502272,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.9033936388359682e-05,
+      "loss": 0.9944598388671875,
       "step": 70000
     },
     {
-      "epoch": 1.6164686377452335,
-      "grad_norm": 3.3662991523742676,
-      "learning_rate": 2.902757377001498e-05,
-      "loss": 0.6519,
+      "epoch": 0.8082424612249421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.9026664824537168e-05,
+      "loss": 0.9848135375976562,
       "step": 70200
     },
     {
-      "epoch": 1.6210739614994933,
-      "grad_norm": 2.8330750465393066,
-      "learning_rate": 2.9020302206192467e-05,
-      "loss": 0.6492,
+      "epoch": 0.810545146299657,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.9019393260714647e-05,
+      "loss": 0.9835179138183594,
       "step": 70400
     },
     {
-      "epoch": 1.6256792852537534,
-      "grad_norm": 3.55230975151062,
-      "learning_rate": 2.9013030642369947e-05,
-      "loss": 0.654,
+      "epoch": 0.812847831374372,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9012121696892133e-05,
+      "loss": 0.9890862274169921,
       "step": 70600
     },
     {
-      "epoch": 1.6302846090080134,
-      "grad_norm": 3.134399890899658,
-      "learning_rate": 2.9005795436366545e-05,
-      "loss": 0.657,
+      "epoch": 0.8151505164490869,
+      "grad_norm": 1.75,
+      "learning_rate": 2.900485013306962e-05,
+      "loss": 0.9792247009277344,
       "step": 70800
     },
     {
-      "epoch": 1.6348899327622732,
-      "grad_norm": 3.899355888366699,
-      "learning_rate": 2.8998523872544028e-05,
-      "loss": 0.6565,
+      "epoch": 0.8174532015238019,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.8997578569247105e-05,
+      "loss": 0.9930565643310547,
       "step": 71000
     },
     {
-      "epoch": 1.639495256516533,
-      "grad_norm": 3.356893301010132,
-      "learning_rate": 2.8991252308721514e-05,
-      "loss": 0.6465,
+      "epoch": 0.8197558865985168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8990307005424587e-05,
+      "loss": 1.0037332153320313,
       "step": 71200
     },
     {
-      "epoch": 1.644100580270793,
-      "grad_norm": 3.021803855895996,
-      "learning_rate": 2.8983980744899e-05,
-      "loss": 0.6409,
+      "epoch": 0.8220585716732317,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.8983035441602073e-05,
+      "loss": 0.9568766021728515,
       "step": 71400
     },
     {
-      "epoch": 1.6487059040250531,
-      "grad_norm": 3.1984188556671143,
-      "learning_rate": 2.8976709181076485e-05,
-      "loss": 0.6457,
+      "epoch": 0.8243612567479467,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.897576387777956e-05,
+      "loss": 1.0169525146484375,
       "step": 71600
     },
     {
-      "epoch": 1.6533112277793127,
-      "grad_norm": 2.6747777462005615,
-      "learning_rate": 2.8969437617253968e-05,
-      "loss": 0.6497,
+      "epoch": 0.8266639418226616,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.8968492313957038e-05,
+      "loss": 0.9906223297119141,
       "step": 71800
     },
     {
-      "epoch": 1.6579165515335728,
-      "grad_norm": 3.9369540214538574,
-      "learning_rate": 2.896216605343145e-05,
-      "loss": 0.6556,
+      "epoch": 0.8289666268973765,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.8961220750134524e-05,
+      "loss": 1.0080840301513672,
       "step": 72000
     },
     {
-      "epoch": 1.6625218752878328,
-      "grad_norm": 3.03513503074646,
-      "learning_rate": 2.8954894489608936e-05,
-      "loss": 0.652,
+      "epoch": 0.8312693119720914,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.895394918631201e-05,
+      "loss": 0.9585226440429687,
       "step": 72200
     },
     {
-      "epoch": 1.6671271990420926,
-      "grad_norm": 2.8434908390045166,
-      "learning_rate": 2.894762292578642e-05,
-      "loss": 0.6575,
+      "epoch": 0.8335719970468064,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.8946677622489492e-05,
+      "loss": 0.9908672332763672,
       "step": 72400
     },
     {
-      "epoch": 1.6717325227963524,
-      "grad_norm": 3.372441291809082,
-      "learning_rate": 2.8940351361963905e-05,
-      "loss": 0.6576,
+      "epoch": 0.8358746821215213,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.893940605866698e-05,
+      "loss": 0.9811976623535156,
       "step": 72600
     },
     {
-      "epoch": 1.6763378465506125,
-      "grad_norm": 2.8065624237060547,
-      "learning_rate": 2.893307979814139e-05,
-      "loss": 0.6379,
+      "epoch": 0.8381773671962363,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.8932134494844464e-05,
+      "loss": 0.9929940795898438,
       "step": 72800
     },
     {
-      "epoch": 1.6809431703048725,
-      "grad_norm": 4.157696723937988,
-      "learning_rate": 2.8925808234318873e-05,
-      "loss": 0.6482,
+      "epoch": 0.8404800522709512,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8924862931021943e-05,
+      "loss": 0.9645259857177735,
       "step": 73000
     },
     {
-      "epoch": 1.6855484940591323,
-      "grad_norm": 2.8976545333862305,
-      "learning_rate": 2.8918536670496356e-05,
-      "loss": 0.6596,
+      "epoch": 0.8427827373456661,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.891759136719943e-05,
+      "loss": 0.9708584594726563,
       "step": 73200
     },
     {
-      "epoch": 1.6901538178133921,
-      "grad_norm": 3.67669415473938,
-      "learning_rate": 2.8911265106673842e-05,
-      "loss": 0.6556,
+      "epoch": 0.8450854224203811,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8910319803376915e-05,
+      "loss": 0.981541748046875,
       "step": 73400
     },
     {
-      "epoch": 1.6947591415676522,
-      "grad_norm": 3.2437984943389893,
-      "learning_rate": 2.8903993542851324e-05,
-      "loss": 0.6519,
+      "epoch": 0.847388107495096,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.8903048239554398e-05,
+      "loss": 0.9799684143066406,
       "step": 73600
     },
     {
-      "epoch": 1.6993644653219122,
-      "grad_norm": 2.789106607437134,
-      "learning_rate": 2.889672197902881e-05,
-      "loss": 0.673,
+      "epoch": 0.849690792569811,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.8895776675731884e-05,
+      "loss": 0.9951773834228516,
       "step": 73800
     },
     {
-      "epoch": 1.703969789076172,
-      "grad_norm": 3.525022268295288,
-      "learning_rate": 2.8889450415206296e-05,
-      "loss": 0.6559,
+      "epoch": 0.8519934776445258,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.888850511190937e-05,
+      "loss": 1.0009957122802735,
       "step": 74000
     },
     {
-      "epoch": 1.7085751128304318,
-      "grad_norm": 3.419024705886841,
-      "learning_rate": 2.8882178851383782e-05,
-      "loss": 0.6539,
+      "epoch": 0.8542961627192408,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.8881233548086855e-05,
+      "loss": 0.9798103332519531,
       "step": 74200
     },
     {
-      "epoch": 1.7131804365846919,
-      "grad_norm": 3.462752103805542,
-      "learning_rate": 2.8874907287561265e-05,
-      "loss": 0.6531,
+      "epoch": 0.8565988477939558,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.8873961984264335e-05,
+      "loss": 0.9798514556884765,
       "step": 74400
     },
     {
-      "epoch": 1.717785760338952,
-      "grad_norm": 3.5559234619140625,
-      "learning_rate": 2.8867635723738747e-05,
-      "loss": 0.6499,
+      "epoch": 0.8589015328686707,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.886669042044182e-05,
+      "loss": 0.9884852600097657,
       "step": 74600
     },
     {
-      "epoch": 1.7223910840932117,
-      "grad_norm": 2.969327211380005,
-      "learning_rate": 2.8860364159916233e-05,
-      "loss": 0.6348,
+      "epoch": 0.8612042179433856,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.8859418856619306e-05,
+      "loss": 1.0256126403808594,
       "step": 74800
     },
     {
-      "epoch": 1.7269964078474715,
-      "grad_norm": 2.5927562713623047,
-      "learning_rate": 2.8853128953912828e-05,
-      "loss": 0.6491,
+      "epoch": 0.8635069030181005,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.885214729279679e-05,
+      "loss": 0.99552734375,
       "step": 75000
     },
     {
-      "epoch": 1.7316017316017316,
-      "grad_norm": 3.60489821434021,
-      "learning_rate": 2.8845857390090314e-05,
-      "loss": 0.6627,
+      "epoch": 0.8658095880928155,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.8844875728974275e-05,
+      "loss": 0.9917625427246094,
       "step": 75200
     },
     {
-      "epoch": 1.7362070553559916,
-      "grad_norm": 3.0263278484344482,
-      "learning_rate": 2.8838585826267797e-05,
-      "loss": 0.6649,
+      "epoch": 0.8681122731675304,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.883760416515176e-05,
+      "loss": 0.9782889556884765,
       "step": 75400
     },
     {
-      "epoch": 1.7408123791102514,
-      "grad_norm": 3.406975507736206,
-      "learning_rate": 2.8831314262445283e-05,
-      "loss": 0.6602,
+      "epoch": 0.8704149582422454,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.883033260132924e-05,
+      "loss": 0.9814031219482422,
       "step": 75600
     },
     {
-      "epoch": 1.7454177028645113,
-      "grad_norm": 3.032499313354492,
-      "learning_rate": 2.882404269862277e-05,
-      "loss": 0.6548,
+      "epoch": 0.8727176433169603,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.8823061037506726e-05,
+      "loss": 0.9945686340332032,
       "step": 75800
     },
     {
-      "epoch": 1.7500230266187713,
-      "grad_norm": 3.3158748149871826,
-      "learning_rate": 2.881677113480025e-05,
-      "loss": 0.6661,
+      "epoch": 0.8750203283916752,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8815789473684212e-05,
+      "loss": 0.9448981475830078,
       "step": 76000
     },
     {
-      "epoch": 1.7546283503730313,
-      "grad_norm": 2.534254312515259,
-      "learning_rate": 2.8809499570977733e-05,
-      "loss": 0.645,
+      "epoch": 0.8773230134663902,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.8808517909861694e-05,
+      "loss": 1.0142716979980468,
       "step": 76200
     },
     {
-      "epoch": 1.7592336741272911,
-      "grad_norm": 2.951493978500366,
-      "learning_rate": 2.880222800715522e-05,
-      "loss": 0.6594,
+      "epoch": 0.8796256985411051,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.880124634603918e-05,
+      "loss": 0.9929308319091796,
       "step": 76400
     },
     {
-      "epoch": 1.763838997881551,
-      "grad_norm": 2.681861162185669,
-      "learning_rate": 2.8794956443332705e-05,
-      "loss": 0.652,
+      "epoch": 0.88192838361582,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.8793974782216666e-05,
+      "loss": 0.9916824340820313,
       "step": 76600
     },
     {
-      "epoch": 1.768444321635811,
-      "grad_norm": 3.9448912143707275,
-      "learning_rate": 2.8787684879510188e-05,
-      "loss": 0.659,
+      "epoch": 0.8842310686905349,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.878670321839415e-05,
+      "loss": 1.0056848907470703,
       "step": 76800
     },
     {
-      "epoch": 1.773049645390071,
-      "grad_norm": 3.7418482303619385,
-      "learning_rate": 2.8780413315687674e-05,
-      "loss": 0.655,
+      "epoch": 0.8865337537652499,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.877943165457163e-05,
+      "loss": 0.9842034149169921,
       "step": 77000
     },
     {
-      "epoch": 1.7776549691443309,
-      "grad_norm": 3.9162464141845703,
-      "learning_rate": 2.877314175186516e-05,
-      "loss": 0.6591,
+      "epoch": 0.8888364388399649,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.8772160090749117e-05,
+      "loss": 1.0175214385986329,
       "step": 77200
     },
     {
-      "epoch": 1.7822602928985907,
-      "grad_norm": 3.8449208736419678,
-      "learning_rate": 2.876587018804264e-05,
-      "loss": 0.6717,
+      "epoch": 0.8911391239146798,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.8764888526926603e-05,
+      "loss": 0.9752666473388671,
       "step": 77400
     },
     {
-      "epoch": 1.7868656166528507,
-      "grad_norm": 2.90519118309021,
-      "learning_rate": 2.8758598624220125e-05,
-      "loss": 0.6559,
+      "epoch": 0.8934418089893947,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.8757616963104086e-05,
+      "loss": 0.9827954864501953,
       "step": 77600
     },
     {
-      "epoch": 1.7914709404071107,
-      "grad_norm": 3.301666259765625,
-      "learning_rate": 2.875132706039761e-05,
-      "loss": 0.6533,
+      "epoch": 0.8957444940641096,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.875034539928157e-05,
+      "loss": 0.9853438568115235,
       "step": 77800
     },
     {
-      "epoch": 1.7960762641613706,
-      "grad_norm": 3.6207275390625,
-      "learning_rate": 2.8744091854394206e-05,
-      "loss": 0.6582,
+      "epoch": 0.8980471791388246,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.8743073835459057e-05,
+      "loss": 0.9848164367675781,
       "step": 78000
     },
     {
-      "epoch": 1.8006815879156304,
-      "grad_norm": 3.3020622730255127,
-      "learning_rate": 2.8736856648390804e-05,
-      "loss": 0.6606,
+      "epoch": 0.9003498642135395,
+      "grad_norm": 0.875,
+      "learning_rate": 2.8735802271636536e-05,
+      "loss": 0.9895773315429688,
       "step": 78200
     },
     {
-      "epoch": 1.8052869116698904,
-      "grad_norm": 3.674409866333008,
-      "learning_rate": 2.8729585084568287e-05,
-      "loss": 0.6581,
+      "epoch": 0.9026525492882544,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.8728530707814022e-05,
+      "loss": 0.9734499359130859,
       "step": 78400
     },
     {
-      "epoch": 1.8098922354241505,
-      "grad_norm": 2.18864369392395,
-      "learning_rate": 2.8722313520745773e-05,
-      "loss": 0.6641,
+      "epoch": 0.9049552343629693,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.8721259143991508e-05,
+      "loss": 0.9732881927490235,
       "step": 78600
     },
     {
-      "epoch": 1.8144975591784103,
-      "grad_norm": 3.165595769882202,
-      "learning_rate": 2.871504195692326e-05,
-      "loss": 0.6568,
+      "epoch": 0.9072579194376843,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.871398758016899e-05,
+      "loss": 0.9826437377929688,
       "step": 78800
     },
     {
-      "epoch": 1.81910288293267,
-      "grad_norm": 2.2597508430480957,
-      "learning_rate": 2.8707770393100738e-05,
-      "loss": 0.651,
+      "epoch": 0.9095606045123993,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.8706716016346477e-05,
+      "loss": 0.9796221160888672,
       "step": 79000
     },
     {
-      "epoch": 1.8237082066869301,
-      "grad_norm": 3.7777042388916016,
-      "learning_rate": 2.8700498829278224e-05,
-      "loss": 0.6773,
+      "epoch": 0.9118632895871142,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8699444452523963e-05,
+      "loss": 1.0010478973388672,
       "step": 79200
     },
     {
-      "epoch": 1.8283135304411902,
-      "grad_norm": 3.0962727069854736,
-      "learning_rate": 2.869322726545571e-05,
-      "loss": 0.6643,
+      "epoch": 0.9141659746618291,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.8692172888701442e-05,
+      "loss": 0.9749209594726562,
       "step": 79400
     },
     {
-      "epoch": 1.83291885419545,
-      "grad_norm": 3.4903509616851807,
-      "learning_rate": 2.8685955701633196e-05,
-      "loss": 0.6636,
+      "epoch": 0.916468659736544,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.8684901324878928e-05,
+      "loss": 0.9853527069091796,
       "step": 79600
     },
     {
-      "epoch": 1.8375241779497098,
-      "grad_norm": 3.2742772102355957,
-      "learning_rate": 2.8678684137810678e-05,
-      "loss": 0.6497,
+      "epoch": 0.918771344811259,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.8677629761056414e-05,
+      "loss": 1.0083032989501952,
       "step": 79800
     },
     {
-      "epoch": 1.8421295017039698,
-      "grad_norm": 3.5722882747650146,
-      "learning_rate": 2.8671412573988164e-05,
-      "loss": 0.6536,
+      "epoch": 0.9210740298859739,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.86703581972339e-05,
+      "loss": 0.9884980010986328,
       "step": 80000
     },
     {
-      "epoch": 1.8467348254582299,
-      "grad_norm": 4.832566738128662,
-      "learning_rate": 2.8664141010165647e-05,
-      "loss": 0.6584,
+      "epoch": 0.9233767149606888,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.8663086633411382e-05,
+      "loss": 0.9641397857666015,
       "step": 80200
     },
     {
-      "epoch": 1.8513401492124897,
-      "grad_norm": 3.033071756362915,
-      "learning_rate": 2.865686944634313e-05,
-      "loss": 0.6542,
+      "epoch": 0.9256794000354038,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.8655815069588868e-05,
+      "loss": 0.9726267242431641,
       "step": 80400
     },
     {
-      "epoch": 1.8559454729667495,
-      "grad_norm": 3.029519557952881,
-      "learning_rate": 2.8649597882520615e-05,
-      "loss": 0.6627,
+      "epoch": 0.9279820851101187,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.864854350576635e-05,
+      "loss": 1.001600570678711,
       "step": 80600
     },
     {
-      "epoch": 1.8605507967210095,
-      "grad_norm": 3.6355273723602295,
-      "learning_rate": 2.86423263186981e-05,
-      "loss": 0.6568,
+      "epoch": 0.9302847701848337,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.8641271941943833e-05,
+      "loss": 0.9985619354248046,
       "step": 80800
     },
     {
-      "epoch": 1.8651561204752696,
-      "grad_norm": 2.547201633453369,
-      "learning_rate": 2.8635054754875584e-05,
-      "loss": 0.6582,
+      "epoch": 0.9325874552595486,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.863400037812132e-05,
+      "loss": 0.967567138671875,
       "step": 81000
     },
     {
-      "epoch": 1.8697614442295294,
-      "grad_norm": 3.6909263134002686,
-      "learning_rate": 2.862778319105307e-05,
-      "loss": 0.6528,
+      "epoch": 0.9348901403342635,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.8626728814298805e-05,
+      "loss": 1.0003462219238282,
       "step": 81200
     },
     {
-      "epoch": 1.8743667679837892,
-      "grad_norm": 3.107825994491577,
-      "learning_rate": 2.8620511627230555e-05,
-      "loss": 0.6452,
+      "epoch": 0.9371928254089784,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.8619457250476287e-05,
+      "loss": 0.9940373229980469,
       "step": 81400
     },
     {
-      "epoch": 1.8789720917380492,
-      "grad_norm": 2.624311685562134,
-      "learning_rate": 2.8613240063408034e-05,
-      "loss": 0.6588,
+      "epoch": 0.9394955104836934,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.8612185686653773e-05,
+      "loss": 0.9728090667724609,
       "step": 81600
     },
     {
-      "epoch": 1.883577415492309,
-      "grad_norm": 3.7440786361694336,
-      "learning_rate": 2.8606004857404633e-05,
-      "loss": 0.6489,
+      "epoch": 0.9417981955584084,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.860491412283126e-05,
+      "loss": 0.9660067749023438,
       "step": 81800
     },
     {
-      "epoch": 1.8881827392465689,
-      "grad_norm": 3.2564656734466553,
-      "learning_rate": 2.859873329358212e-05,
-      "loss": 0.6599,
+      "epoch": 0.9441008806331233,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.859764255900874e-05,
+      "loss": 0.9793297576904297,
       "step": 82000
     },
     {
-      "epoch": 1.892788063000829,
-      "grad_norm": 3.0685300827026367,
-      "learning_rate": 2.85914617297596e-05,
-      "loss": 0.6504,
+      "epoch": 0.9464035657078382,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.8590370995186224e-05,
+      "loss": 0.9641354370117188,
       "step": 82200
     },
     {
-      "epoch": 1.897393386755089,
-      "grad_norm": 3.587435245513916,
-      "learning_rate": 2.8584190165937087e-05,
-      "loss": 0.6499,
+      "epoch": 0.9487062507825531,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.858309943136371e-05,
+      "loss": 0.9971208953857422,
       "step": 82400
     },
     {
-      "epoch": 1.9019987105093488,
-      "grad_norm": 2.8492074012756348,
-      "learning_rate": 2.8576918602114573e-05,
-      "loss": 0.6353,
+      "epoch": 0.9510089358572681,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.8575827867541193e-05,
+      "loss": 0.9889365386962891,
       "step": 82600
     },
     {
-      "epoch": 1.9066040342636086,
-      "grad_norm": 3.6821560859680176,
-      "learning_rate": 2.8569647038292056e-05,
-      "loss": 0.6529,
+      "epoch": 0.953311620931983,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.856855630371868e-05,
+      "loss": 0.995948715209961,
       "step": 82800
     },
     {
-      "epoch": 1.9112093580178686,
-      "grad_norm": 4.206520080566406,
-      "learning_rate": 2.856237547446954e-05,
-      "loss": 0.6524,
+      "epoch": 0.9556143060066979,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.8561284739896165e-05,
+      "loss": 0.9824695587158203,
       "step": 83000
     },
     {
-      "epoch": 1.9158146817721287,
-      "grad_norm": 3.277606725692749,
-      "learning_rate": 2.8555103910647024e-05,
-      "loss": 0.6579,
+      "epoch": 0.9579169910814128,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.8554013176073647e-05,
+      "loss": 0.97533935546875,
       "step": 83200
     },
     {
-      "epoch": 1.9204200055263885,
-      "grad_norm": 3.900179624557495,
-      "learning_rate": 2.8547832346824507e-05,
-      "loss": 0.6627,
+      "epoch": 0.9602196761561278,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.854674161225113e-05,
+      "loss": 0.9821726226806641,
       "step": 83400
     },
     {
-      "epoch": 1.9250253292806483,
-      "grad_norm": 2.804596185684204,
-      "learning_rate": 2.8540560783001993e-05,
-      "loss": 0.6655,
+      "epoch": 0.9625223612308428,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.8539470048428615e-05,
+      "loss": 0.9993029022216797,
       "step": 83600
     },
     {
-      "epoch": 1.9296306530349083,
-      "grad_norm": 3.212975263595581,
-      "learning_rate": 2.853328921917948e-05,
-      "loss": 0.6614,
+      "epoch": 0.9648250463055577,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.85321984846061e-05,
+      "loss": 0.9740554046630859,
       "step": 83800
     },
     {
-      "epoch": 1.9342359767891684,
-      "grad_norm": 4.128197193145752,
-      "learning_rate": 2.8526017655356965e-05,
-      "loss": 0.6546,
+      "epoch": 0.9671277313802726,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.8524926920783584e-05,
+      "loss": 0.9690258026123046,
       "step": 84000
     },
     {
-      "epoch": 1.9388413005434282,
-      "grad_norm": 3.0182225704193115,
-      "learning_rate": 2.8518746091534447e-05,
-      "loss": 0.659,
+      "epoch": 0.9694304164549875,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.851765535696107e-05,
+      "loss": 0.9684633636474609,
       "step": 84200
     },
     {
-      "epoch": 1.943446624297688,
-      "grad_norm": 2.6237552165985107,
-      "learning_rate": 2.851147452771193e-05,
-      "loss": 0.6467,
+      "epoch": 0.9717331015297025,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.8510383793138556e-05,
+      "loss": 0.9727975463867188,
       "step": 84400
     },
     {
-      "epoch": 1.948051948051948,
-      "grad_norm": 2.700956106185913,
-      "learning_rate": 2.8504202963889416e-05,
-      "loss": 0.6597,
+      "epoch": 0.9740357866044174,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.8503112229316035e-05,
+      "loss": 0.9977262878417968,
       "step": 84600
     },
     {
-      "epoch": 1.952657271806208,
-      "grad_norm": 3.318554162979126,
-      "learning_rate": 2.8496931400066898e-05,
-      "loss": 0.649,
+      "epoch": 0.9763384716791323,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.849584066549352e-05,
+      "loss": 0.9849166107177735,
       "step": 84800
     },
     {
-      "epoch": 1.9572625955604679,
-      "grad_norm": 2.75907301902771,
-      "learning_rate": 2.8489659836244384e-05,
-      "loss": 0.6713,
+      "epoch": 0.9786411567538473,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.8488569101671007e-05,
+      "loss": 1.0033737182617188,
       "step": 85000
     },
     {
-      "epoch": 1.9618679193147277,
-      "grad_norm": 3.922351837158203,
-      "learning_rate": 2.848238827242187e-05,
-      "loss": 0.6495,
+      "epoch": 0.9809438418285622,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.848129753784849e-05,
+      "loss": 0.9663671875,
       "step": 85200
     },
     {
-      "epoch": 1.9664732430689877,
-      "grad_norm": 2.883132219314575,
-      "learning_rate": 2.8475116708599352e-05,
-      "loss": 0.6609,
+      "epoch": 0.9832465269032772,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.8474025974025975e-05,
+      "loss": 0.9906986236572266,
       "step": 85400
     },
     {
-      "epoch": 1.9710785668232478,
-      "grad_norm": 2.974393844604492,
-      "learning_rate": 2.8467845144776835e-05,
-      "loss": 0.6435,
+      "epoch": 0.9855492119779921,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.846675441020346e-05,
+      "loss": 0.9876493072509765,
       "step": 85600
     },
     {
-      "epoch": 1.9756838905775076,
-      "grad_norm": 3.174490451812744,
-      "learning_rate": 2.846057358095432e-05,
-      "loss": 0.6514,
+      "epoch": 0.987851897052707,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.8459482846380944e-05,
+      "loss": 0.997586898803711,
       "step": 85800
     },
     {
-      "epoch": 1.9802892143317674,
-      "grad_norm": 3.1419320106506348,
-      "learning_rate": 2.8453302017131803e-05,
-      "loss": 0.6652,
+      "epoch": 0.9901545821274219,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.8452211282558426e-05,
+      "loss": 0.995333480834961,
       "step": 86000
     },
     {
-      "epoch": 1.9848945380860274,
-      "grad_norm": 3.508948802947998,
-      "learning_rate": 2.844603045330929e-05,
-      "loss": 0.6562,
+      "epoch": 0.9924572672021369,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.8444939718735912e-05,
+      "loss": 0.99697509765625,
       "step": 86200
     },
     {
-      "epoch": 1.9894998618402875,
-      "grad_norm": 3.572232484817505,
-      "learning_rate": 2.8438758889486775e-05,
-      "loss": 0.6547,
+      "epoch": 0.9947599522768519,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.8437668154913398e-05,
+      "loss": 0.9929368591308594,
       "step": 86400
     },
     {
-      "epoch": 1.9941051855945473,
-      "grad_norm": 2.923038959503174,
-      "learning_rate": 2.8431487325664258e-05,
-      "loss": 0.6555,
+      "epoch": 0.9970626373515667,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.843039659109088e-05,
+      "loss": 1.0003439331054687,
       "step": 86600
     },
     {
-      "epoch": 1.9987105093488071,
-      "grad_norm": 2.6537113189697266,
-      "learning_rate": 2.8424252119660856e-05,
-      "loss": 0.6481,
+      "epoch": 0.9993653224262817,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.8423125027268366e-05,
+      "loss": 0.9839323425292968,
       "step": 86800
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 0.6096945405006409,
-      "eval_runtime": 145.8772,
-      "eval_samples_per_second": 194.417,
-      "eval_steps_per_second": 12.154,
+      "epoch": 1.0,
+      "eval_loss": NaN,
+      "eval_runtime": 466.187,
+      "eval_samples_per_second": 60.836,
+      "eval_steps_per_second": 30.419,
       "step": 86856
     },
     {
-      "epoch": 2.003315833103067,
-      "grad_norm": 2.6319403648376465,
-      "learning_rate": 2.841698055583834e-05,
-      "loss": 0.6447,
+      "epoch": 1.0016579332537947,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.841585346344585e-05,
+      "loss": 0.9767461395263672,
       "step": 87000
     },
     {
-      "epoch": 2.007921156857327,
-      "grad_norm": 3.3144571781158447,
-      "learning_rate": 2.840970899201582e-05,
-      "loss": 0.6489,
+      "epoch": 1.0039606183285097,
+      "grad_norm": 0.875,
+      "learning_rate": 2.840858189962333e-05,
+      "loss": 0.9855145263671875,
       "step": 87200
     },
     {
-      "epoch": 2.012526480611587,
-      "grad_norm": 3.0572547912597656,
-      "learning_rate": 2.8402437428193307e-05,
-      "loss": 0.6458,
+      "epoch": 1.0062633034032247,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.8401310335800817e-05,
+      "loss": 0.9700740051269531,
       "step": 87400
     },
     {
-      "epoch": 2.017131804365847,
-      "grad_norm": 3.151933431625366,
-      "learning_rate": 2.8395165864370793e-05,
-      "loss": 0.6484,
+      "epoch": 1.0085659884779397,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.8394038771978303e-05,
+      "loss": 0.9850736236572266,
       "step": 87600
     },
     {
-      "epoch": 2.021737128120107,
-      "grad_norm": 3.794235944747925,
-      "learning_rate": 2.8387894300548276e-05,
-      "loss": 0.6493,
+      "epoch": 1.0108686735526544,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.8386767208155786e-05,
+      "loss": 0.9656169128417968,
       "step": 87800
     },
     {
-      "epoch": 2.026342451874367,
-      "grad_norm": 2.9801411628723145,
-      "learning_rate": 2.838062273672576e-05,
-      "loss": 0.6399,
+      "epoch": 1.0131713586273694,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.837949564433327e-05,
+      "loss": 0.9913868713378906,
       "step": 88000
     },
     {
-      "epoch": 2.0309477756286265,
-      "grad_norm": 3.094648599624634,
-      "learning_rate": 2.8373351172903248e-05,
-      "loss": 0.6488,
+      "epoch": 1.0154740437020844,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.8372224080510758e-05,
+      "loss": 0.9808352661132812,
       "step": 88200
     },
     {
-      "epoch": 2.0355530993828865,
-      "grad_norm": 2.9568779468536377,
-      "learning_rate": 2.8366079609080727e-05,
-      "loss": 0.6476,
+      "epoch": 1.0177767287767994,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.8364952516688237e-05,
+      "loss": 0.9822716522216797,
       "step": 88400
     },
     {
-      "epoch": 2.0401584231371466,
-      "grad_norm": 3.226010799407959,
-      "learning_rate": 2.8358808045258213e-05,
-      "loss": 0.6439,
+      "epoch": 1.0200794138515141,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.8357680952865723e-05,
+      "loss": 0.9840770721435547,
       "step": 88600
     },
     {
-      "epoch": 2.0447637468914066,
-      "grad_norm": 3.10718035697937,
-      "learning_rate": 2.83515364814357e-05,
-      "loss": 0.6421,
+      "epoch": 1.022382098926229,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.835040938904321e-05,
+      "loss": 0.9758528137207031,
       "step": 88800
     },
     {
-      "epoch": 2.049369070645666,
-      "grad_norm": 3.8462493419647217,
-      "learning_rate": 2.8344264917613184e-05,
-      "loss": 0.6497,
+      "epoch": 1.024684784000944,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.8343137825220694e-05,
+      "loss": 1.005013427734375,
       "step": 89000
     },
     {
-      "epoch": 2.0539743943999262,
-      "grad_norm": 2.297658920288086,
-      "learning_rate": 2.8336993353790667e-05,
-      "loss": 0.6425,
+      "epoch": 1.026987469075659,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8335866261398177e-05,
+      "loss": 1.0009622955322266,
       "step": 89200
     },
     {
-      "epoch": 2.0585797181541863,
-      "grad_norm": 3.4975883960723877,
-      "learning_rate": 2.8329721789968153e-05,
-      "loss": 0.6387,
+      "epoch": 1.029290154150374,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8328594697575663e-05,
+      "loss": 0.9745167541503906,
       "step": 89400
     },
     {
-      "epoch": 2.0631850419084463,
-      "grad_norm": 2.6451213359832764,
-      "learning_rate": 2.8322450226145635e-05,
-      "loss": 0.6507,
+      "epoch": 1.0315928392250888,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.8321323133753145e-05,
+      "loss": 0.9853366088867187,
       "step": 89600
     },
     {
-      "epoch": 2.067790365662706,
-      "grad_norm": 4.4015679359436035,
-      "learning_rate": 2.8315178662323118e-05,
-      "loss": 0.6549,
+      "epoch": 1.0338955242998038,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.8314051569930628e-05,
+      "loss": 0.9802220916748047,
       "step": 89800
     },
     {
-      "epoch": 2.072395689416966,
-      "grad_norm": 3.2496938705444336,
-      "learning_rate": 2.8307907098500604e-05,
-      "loss": 0.6443,
+      "epoch": 1.0361982093745188,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8306780006108114e-05,
+      "loss": 0.9878942108154297,
       "step": 90000
     },
     {
-      "epoch": 2.077001013171226,
-      "grad_norm": 3.8921310901641846,
-      "learning_rate": 2.830063553467809e-05,
-      "loss": 0.6563,
+      "epoch": 1.0385008944492338,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.82995084422856e-05,
+      "loss": 1.0209328460693359,
       "step": 90200
     },
     {
-      "epoch": 2.081606336925486,
-      "grad_norm": 2.6001060009002686,
-      "learning_rate": 2.8293363970855572e-05,
-      "loss": 0.6561,
+      "epoch": 1.0408035795239488,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.8292236878463082e-05,
+      "loss": 0.9749903869628906,
       "step": 90400
     },
     {
-      "epoch": 2.0862116606797456,
-      "grad_norm": 2.925668716430664,
-      "learning_rate": 2.8286092407033058e-05,
-      "loss": 0.6395,
+      "epoch": 1.0431062645986635,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.8284965314640568e-05,
+      "loss": 0.964295654296875,
       "step": 90600
     },
     {
-      "epoch": 2.0908169844340057,
-      "grad_norm": 2.9956302642822266,
-      "learning_rate": 2.8278857201029657e-05,
-      "loss": 0.6584,
+      "epoch": 1.0454089496733785,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.8277693750818054e-05,
+      "loss": 1.0175611877441406,
       "step": 90800
     },
     {
-      "epoch": 2.0954223081882657,
-      "grad_norm": 3.5268077850341797,
-      "learning_rate": 2.8271585637207136e-05,
-      "loss": 0.648,
+      "epoch": 1.0477116347480935,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.8270422186995533e-05,
+      "loss": 1.0039064025878905,
       "step": 91000
     },
     {
-      "epoch": 2.1000276319425257,
-      "grad_norm": 4.238087177276611,
-      "learning_rate": 2.8264314073384622e-05,
-      "loss": 0.6387,
+      "epoch": 1.0500143198228085,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.826315062317302e-05,
+      "loss": 0.9765067291259766,
       "step": 91200
     },
     {
-      "epoch": 2.1046329556967853,
-      "grad_norm": 2.673576593399048,
-      "learning_rate": 2.8257042509562108e-05,
-      "loss": 0.6379,
+      "epoch": 1.0523170048975232,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8255879059350505e-05,
+      "loss": 1.002488555908203,
       "step": 91400
     },
     {
-      "epoch": 2.1092382794510454,
-      "grad_norm": 2.9278106689453125,
-      "learning_rate": 2.824977094573959e-05,
-      "loss": 0.6527,
+      "epoch": 1.0546196899722382,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.824860749552799e-05,
+      "loss": 0.9849620819091797,
       "step": 91600
     },
     {
-      "epoch": 2.1138436032053054,
-      "grad_norm": 3.109639883041382,
-      "learning_rate": 2.824253573973619e-05,
-      "loss": 0.6437,
+      "epoch": 1.0569223750469532,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.8241335931705474e-05,
+      "loss": 0.9739988708496093,
       "step": 91800
     },
     {
-      "epoch": 2.1184489269595654,
-      "grad_norm": 3.2861876487731934,
-      "learning_rate": 2.8235264175913675e-05,
-      "loss": 0.6269,
+      "epoch": 1.0592250601216682,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.823406436788296e-05,
+      "loss": 0.9917415618896485,
       "step": 92000
     },
     {
-      "epoch": 2.123054250713825,
-      "grad_norm": 3.4922659397125244,
-      "learning_rate": 2.8227992612091157e-05,
-      "loss": 0.647,
+      "epoch": 1.0615277451963832,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.8226792804060442e-05,
+      "loss": 0.9903607177734375,
       "step": 92200
     },
     {
-      "epoch": 2.127659574468085,
-      "grad_norm": 3.1858551502227783,
-      "learning_rate": 2.8220721048268643e-05,
-      "loss": 0.6438,
+      "epoch": 1.063830430271098,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.8219521240237925e-05,
+      "loss": 0.9716462707519531,
       "step": 92400
     },
     {
-      "epoch": 2.132264898222345,
-      "grad_norm": 3.2486298084259033,
-      "learning_rate": 2.8213449484446126e-05,
-      "loss": 0.6403,
+      "epoch": 1.066133115345813,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.821224967641541e-05,
+      "loss": 0.9898148345947265,
       "step": 92600
     },
     {
-      "epoch": 2.136870221976605,
-      "grad_norm": 2.3722951412200928,
-      "learning_rate": 2.8206177920623608e-05,
-      "loss": 0.6346,
+      "epoch": 1.0684358004205279,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.8204978112592896e-05,
+      "loss": 0.9770774841308594,
       "step": 92800
     },
     {
-      "epoch": 2.1414755457308647,
-      "grad_norm": 3.3336052894592285,
-      "learning_rate": 2.8198906356801094e-05,
-      "loss": 0.6514,
+      "epoch": 1.0707384854952429,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.819770654877038e-05,
+      "loss": 1.016856918334961,
       "step": 93000
     },
     {
-      "epoch": 2.1460808694851248,
-      "grad_norm": 2.755908727645874,
-      "learning_rate": 2.819163479297858e-05,
-      "loss": 0.6402,
+      "epoch": 1.0730411705699576,
+      "grad_norm": 1.125,
+      "learning_rate": 2.8190434984947865e-05,
+      "loss": 0.9860330200195313,
       "step": 93200
     },
     {
-      "epoch": 2.150686193239385,
-      "grad_norm": 3.960623264312744,
-      "learning_rate": 2.8184363229156063e-05,
-      "loss": 0.6398,
+      "epoch": 1.0753438556446726,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.8183163421125347e-05,
+      "loss": 0.9857374572753906,
       "step": 93400
     },
     {
-      "epoch": 2.155291516993645,
-      "grad_norm": 3.4139673709869385,
-      "learning_rate": 2.817709166533355e-05,
-      "loss": 0.6282,
+      "epoch": 1.0776465407193876,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.817589185730283e-05,
+      "loss": 0.9712320709228516,
       "step": 93600
     },
     {
-      "epoch": 2.1598968407479044,
-      "grad_norm": 3.090846300125122,
-      "learning_rate": 2.816982010151103e-05,
-      "loss": 0.6634,
+      "epoch": 1.0799492257941026,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.8168620293480316e-05,
+      "loss": 0.9812728881835937,
       "step": 93800
     },
     {
-      "epoch": 2.1645021645021645,
-      "grad_norm": 2.835188627243042,
-      "learning_rate": 2.8162584895507626e-05,
-      "loss": 0.6342,
+      "epoch": 1.0822519108688176,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.81613487296578e-05,
+      "loss": 0.9942698669433594,
       "step": 94000
     },
     {
-      "epoch": 2.1691074882564245,
-      "grad_norm": 2.9350528717041016,
-      "learning_rate": 2.8155313331685112e-05,
-      "loss": 0.6412,
+      "epoch": 1.0845545959435323,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8154077165835284e-05,
+      "loss": 0.9869801330566407,
       "step": 94200
     },
     {
-      "epoch": 2.1737128120106846,
-      "grad_norm": 3.2040562629699707,
-      "learning_rate": 2.8148041767862598e-05,
-      "loss": 0.6519,
+      "epoch": 1.0868572810182473,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.814680560201277e-05,
+      "loss": 0.9874109649658203,
       "step": 94400
     },
     {
-      "epoch": 2.178318135764944,
-      "grad_norm": 2.805971384048462,
-      "learning_rate": 2.814077020404008e-05,
-      "loss": 0.6438,
+      "epoch": 1.0891599660929623,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.8139534038190256e-05,
+      "loss": 0.974925537109375,
       "step": 94600
     },
     {
-      "epoch": 2.182923459519204,
-      "grad_norm": 2.857898712158203,
-      "learning_rate": 2.8133498640217567e-05,
-      "loss": 0.6487,
+      "epoch": 1.0914626511676773,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.813226247436774e-05,
+      "loss": 0.9871249389648438,
       "step": 94800
     },
     {
-      "epoch": 2.1875287832734642,
-      "grad_norm": 2.6172449588775635,
-      "learning_rate": 2.8126227076395052e-05,
-      "loss": 0.6422,
+      "epoch": 1.093765336242392,
+      "grad_norm": 0.75,
+      "learning_rate": 2.812499091054522e-05,
+      "loss": 0.9713893127441406,
       "step": 95000
     },
     {
-      "epoch": 2.192134107027724,
-      "grad_norm": 4.178059101104736,
-      "learning_rate": 2.8118955512572535e-05,
-      "loss": 0.6477,
+      "epoch": 1.096068021317107,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.8117719346722707e-05,
+      "loss": 0.9837332916259766,
       "step": 95200
     },
     {
-      "epoch": 2.196739430781984,
-      "grad_norm": 3.011091947555542,
-      "learning_rate": 2.8111683948750018e-05,
-      "loss": 0.6463,
+      "epoch": 1.098370706391822,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.8110447782900193e-05,
+      "loss": 0.9709980010986328,
       "step": 95400
     },
     {
-      "epoch": 2.201344754536244,
-      "grad_norm": 3.2300426959991455,
-      "learning_rate": 2.8104412384927503e-05,
-      "loss": 0.6409,
+      "epoch": 1.100673391466537,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.8103176219077675e-05,
+      "loss": 0.9841094207763672,
       "step": 95600
     },
     {
-      "epoch": 2.205950078290504,
-      "grad_norm": 3.7736730575561523,
-      "learning_rate": 2.8097140821104986e-05,
-      "loss": 0.6504,
+      "epoch": 1.102976076541252,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.809590465525516e-05,
+      "loss": 0.989020767211914,
       "step": 95800
     },
     {
-      "epoch": 2.210555402044764,
-      "grad_norm": 3.110412120819092,
-      "learning_rate": 2.8089869257282472e-05,
-      "loss": 0.6378,
+      "epoch": 1.1052787616159667,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.8088633091432644e-05,
+      "loss": 0.9772124481201172,
       "step": 96000
     },
     {
-      "epoch": 2.2151607257990236,
-      "grad_norm": 4.121431350708008,
-      "learning_rate": 2.8082597693459958e-05,
-      "loss": 0.653,
+      "epoch": 1.1075814466906817,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8081361527610126e-05,
+      "loss": 0.9723876953125,
       "step": 96200
     },
     {
-      "epoch": 2.2197660495532836,
-      "grad_norm": 3.8190181255340576,
-      "learning_rate": 2.807532612963744e-05,
-      "loss": 0.6441,
+      "epoch": 1.1098841317653967,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8074089963787612e-05,
+      "loss": 0.9917980194091797,
       "step": 96400
     },
     {
-      "epoch": 2.2243713733075436,
-      "grad_norm": 2.9528844356536865,
-      "learning_rate": 2.8068054565814923e-05,
-      "loss": 0.6276,
+      "epoch": 1.1121868168401117,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.8066818399965098e-05,
+      "loss": 0.9809168243408203,
       "step": 96600
     },
     {
-      "epoch": 2.2289766970618032,
-      "grad_norm": 3.1562676429748535,
-      "learning_rate": 2.806078300199241e-05,
-      "loss": 0.6367,
+      "epoch": 1.1144895019148267,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.805954683614258e-05,
+      "loss": 0.9748997497558594,
       "step": 96800
     },
     {
-      "epoch": 2.2335820208160633,
-      "grad_norm": 3.0100133419036865,
-      "learning_rate": 2.8053511438169895e-05,
-      "loss": 0.6336,
+      "epoch": 1.1167921869895414,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.8052275272320067e-05,
+      "loss": 0.9917316436767578,
       "step": 97000
     },
     {
-      "epoch": 2.2381873445703233,
-      "grad_norm": 2.758850336074829,
-      "learning_rate": 2.8046239874347377e-05,
-      "loss": 0.6509,
+      "epoch": 1.1190948720642564,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.804500370849755e-05,
+      "loss": 0.9702124786376953,
       "step": 97200
     },
     {
-      "epoch": 2.2427926683245833,
-      "grad_norm": 3.6952168941497803,
-      "learning_rate": 2.8038968310524863e-05,
-      "loss": 0.6601,
+      "epoch": 1.1213975571389714,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.8037732144675035e-05,
+      "loss": 0.9899410247802735,
       "step": 97400
     },
     {
-      "epoch": 2.247397992078843,
-      "grad_norm": 3.7031092643737793,
-      "learning_rate": 2.803173310452146e-05,
-      "loss": 0.6531,
+      "epoch": 1.1237002422136864,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.8030460580852518e-05,
+      "loss": 0.9802442169189454,
       "step": 97600
     },
     {
-      "epoch": 2.252003315833103,
-      "grad_norm": 3.341907024383545,
-      "learning_rate": 2.8024461540698944e-05,
-      "loss": 0.6614,
+      "epoch": 1.1260029272884013,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.8023189017030004e-05,
+      "loss": 0.9868735504150391,
       "step": 97800
     },
     {
-      "epoch": 2.256608639587363,
-      "grad_norm": 3.172600746154785,
-      "learning_rate": 2.8017189976876427e-05,
-      "loss": 0.6426,
+      "epoch": 1.128305612363116,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.801591745320749e-05,
+      "loss": 0.9881065368652344,
       "step": 98000
     },
     {
-      "epoch": 2.261213963341623,
-      "grad_norm": 3.3580822944641113,
-      "learning_rate": 2.800991841305391e-05,
-      "loss": 0.6453,
+      "epoch": 1.130608297437831,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8008645889384972e-05,
+      "loss": 0.9748045349121094,
       "step": 98200
     },
     {
-      "epoch": 2.2658192870958827,
-      "grad_norm": 4.135528087615967,
-      "learning_rate": 2.8002646849231395e-05,
-      "loss": 0.6386,
+      "epoch": 1.132910982512546,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8001374325562458e-05,
+      "loss": 0.9920476531982422,
       "step": 98400
     },
     {
-      "epoch": 2.2704246108501427,
-      "grad_norm": 3.5923011302948,
-      "learning_rate": 2.799537528540888e-05,
-      "loss": 0.6651,
+      "epoch": 1.135213667587261,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.799410276173994e-05,
+      "loss": 1.0023683166503907,
       "step": 98600
     },
     {
-      "epoch": 2.2750299346044027,
-      "grad_norm": 3.149178981781006,
-      "learning_rate": 2.7988103721586367e-05,
-      "loss": 0.6435,
+      "epoch": 1.1375163526619758,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.7986831197917423e-05,
+      "loss": 0.9839882659912109,
       "step": 98800
     },
     {
-      "epoch": 2.2796352583586628,
-      "grad_norm": 3.2348053455352783,
-      "learning_rate": 2.798083215776385e-05,
-      "loss": 0.6424,
+      "epoch": 1.1398190377366908,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.797955963409491e-05,
+      "loss": 0.9793102264404296,
       "step": 99000
     },
     {
-      "epoch": 2.2842405821129224,
-      "grad_norm": 3.580576181411743,
-      "learning_rate": 2.7973560593941335e-05,
-      "loss": 0.6572,
+      "epoch": 1.1421217228114058,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.7972288070272395e-05,
+      "loss": 0.9812483215332031,
       "step": 99200
     },
     {
-      "epoch": 2.2888459058671824,
-      "grad_norm": 3.2298011779785156,
-      "learning_rate": 2.7966289030118818e-05,
-      "loss": 0.666,
+      "epoch": 1.1444244078861208,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.7965016506449877e-05,
+      "loss": 0.9835822296142578,
       "step": 99400
     },
     {
-      "epoch": 2.2934512296214424,
-      "grad_norm": 3.1280014514923096,
-      "learning_rate": 2.79590174662963e-05,
-      "loss": 0.6499,
+      "epoch": 1.1467270929608357,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.7957744942627363e-05,
+      "loss": 0.9863402557373047,
       "step": 99600
     },
     {
-      "epoch": 2.2980565533757025,
-      "grad_norm": 3.244581937789917,
-      "learning_rate": 2.79517822602929e-05,
-      "loss": 0.6428,
+      "epoch": 1.1490297780355505,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7950473378804846e-05,
+      "loss": 0.9773790740966797,
       "step": 99800
     },
     {
-      "epoch": 2.302661877129962,
-      "grad_norm": 3.160811424255371,
-      "learning_rate": 2.7944510696470385e-05,
-      "loss": 0.639,
+      "epoch": 1.1513324631102655,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.7943201814982328e-05,
+      "loss": 0.9791139984130859,
       "step": 100000
     },
     {
-      "epoch": 2.307267200884222,
-      "grad_norm": 3.2800822257995605,
-      "learning_rate": 2.793727549046698e-05,
-      "loss": 0.6504,
+      "epoch": 1.1536351481849805,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7935930251159814e-05,
+      "loss": 1.0102542114257813,
       "step": 100200
     },
     {
-      "epoch": 2.311872524638482,
-      "grad_norm": 2.948641300201416,
-      "learning_rate": 2.7930003926644466e-05,
-      "loss": 0.6455,
+      "epoch": 1.1559378332596955,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.79286586873373e-05,
+      "loss": 0.9698152923583985,
       "step": 100400
     },
     {
-      "epoch": 2.316477848392742,
-      "grad_norm": 3.270315408706665,
-      "learning_rate": 2.792273236282195e-05,
-      "loss": 0.6498,
+      "epoch": 1.1582405183344102,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7921387123514786e-05,
+      "loss": 0.9856640625,
       "step": 100600
     },
     {
-      "epoch": 2.3210831721470018,
-      "grad_norm": 3.389112710952759,
-      "learning_rate": 2.7915460798999435e-05,
-      "loss": 0.6392,
+      "epoch": 1.1605432034091252,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.791411555969227e-05,
+      "loss": 0.9970908355712891,
       "step": 100800
     },
     {
-      "epoch": 2.325688495901262,
-      "grad_norm": 3.453878164291382,
-      "learning_rate": 2.7908189235176917e-05,
-      "loss": 0.6538,
+      "epoch": 1.1628458884838402,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.7906843995869754e-05,
+      "loss": 0.9615906524658203,
       "step": 101000
     },
     {
-      "epoch": 2.330293819655522,
-      "grad_norm": 2.7830543518066406,
-      "learning_rate": 2.79009176713544e-05,
-      "loss": 0.6402,
+      "epoch": 1.1651485735585552,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7899572432047237e-05,
+      "loss": 0.9927642822265625,
       "step": 101200
     },
     {
-      "epoch": 2.334899143409782,
-      "grad_norm": 3.6506121158599854,
-      "learning_rate": 2.7893646107531886e-05,
-      "loss": 0.6448,
+      "epoch": 1.1674512586332702,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.789230086822472e-05,
+      "loss": 0.9984485626220703,
       "step": 101400
     },
     {
-      "epoch": 2.3395044671640415,
-      "grad_norm": 2.926161050796509,
-      "learning_rate": 2.788637454370937e-05,
-      "loss": 0.6506,
+      "epoch": 1.169753943707985,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7885029304402205e-05,
+      "loss": 0.9728903198242187,
       "step": 101600
     },
     {
-      "epoch": 2.3441097909183015,
-      "grad_norm": 2.95794677734375,
-      "learning_rate": 2.7879102979886857e-05,
-      "loss": 0.6472,
+      "epoch": 1.1720566287827,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.787775774057969e-05,
+      "loss": 0.968542709350586,
       "step": 101800
     },
     {
-      "epoch": 2.3487151146725616,
-      "grad_norm": 3.9702541828155518,
-      "learning_rate": 2.787183141606434e-05,
-      "loss": 0.645,
+      "epoch": 1.1743593138574149,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.7870486176757174e-05,
+      "loss": 0.9618368530273438,
       "step": 102000
     },
     {
-      "epoch": 2.3533204384268216,
-      "grad_norm": 2.927553176879883,
-      "learning_rate": 2.7864559852241826e-05,
-      "loss": 0.6354,
+      "epoch": 1.1766619989321299,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.786321461293466e-05,
+      "loss": 0.9763688659667968,
       "step": 102200
     },
     {
-      "epoch": 2.357925762181081,
-      "grad_norm": 3.047414779663086,
-      "learning_rate": 2.785728828841931e-05,
-      "loss": 0.6475,
+      "epoch": 1.1789646840068446,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.7855943049112142e-05,
+      "loss": 0.9664765167236328,
       "step": 102400
     },
     {
-      "epoch": 2.3625310859353412,
-      "grad_norm": 2.788905382156372,
-      "learning_rate": 2.785001672459679e-05,
-      "loss": 0.6542,
+      "epoch": 1.1812673690815596,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7848671485289625e-05,
+      "loss": 1.0016697692871093,
       "step": 102600
     },
     {
-      "epoch": 2.3671364096896013,
-      "grad_norm": 2.729799509048462,
-      "learning_rate": 2.7842745160774277e-05,
-      "loss": 0.6384,
+      "epoch": 1.1835700541562746,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.784139992146711e-05,
+      "loss": 0.982967529296875,
       "step": 102800
     },
     {
-      "epoch": 2.3717417334438613,
-      "grad_norm": 3.3632562160491943,
-      "learning_rate": 2.7835473596951763e-05,
-      "loss": 0.6314,
+      "epoch": 1.1858727392309896,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7834128357644597e-05,
+      "loss": 0.9826297760009766,
       "step": 103000
     },
     {
-      "epoch": 2.376347057198121,
-      "grad_norm": 3.1274969577789307,
-      "learning_rate": 2.7828202033129245e-05,
-      "loss": 0.6456,
+      "epoch": 1.1881754243057046,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.782685679382208e-05,
+      "loss": 1.0046814727783202,
       "step": 103200
     },
     {
-      "epoch": 2.380952380952381,
-      "grad_norm": 3.8563055992126465,
-      "learning_rate": 2.782093046930673e-05,
-      "loss": 0.6393,
+      "epoch": 1.1904781093804193,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.7819585229999565e-05,
+      "loss": 0.9985332489013672,
       "step": 103400
     },
     {
-      "epoch": 2.385557704706641,
-      "grad_norm": 3.3758862018585205,
-      "learning_rate": 2.7813658905484214e-05,
-      "loss": 0.6386,
+      "epoch": 1.1927807944551343,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7812313666177048e-05,
+      "loss": 0.9772327423095704,
       "step": 103600
     },
     {
-      "epoch": 2.3901630284609006,
-      "grad_norm": 3.2293407917022705,
-      "learning_rate": 2.7806387341661696e-05,
-      "loss": 0.6505,
+      "epoch": 1.1950834795298493,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.7805042102354533e-05,
+      "loss": 0.9572280883789063,
       "step": 103800
     },
     {
-      "epoch": 2.3947683522151606,
-      "grad_norm": 3.1914443969726562,
-      "learning_rate": 2.7799115777839182e-05,
-      "loss": 0.643,
+      "epoch": 1.1973861646045643,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7797770538532016e-05,
+      "loss": 1.0030010223388672,
       "step": 104000
     },
     {
-      "epoch": 2.3993736759694206,
-      "grad_norm": 2.7224652767181396,
-      "learning_rate": 2.7791844214016668e-05,
-      "loss": 0.6374,
+      "epoch": 1.199688849679279,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.7790498974709502e-05,
+      "loss": 1.004496841430664,
       "step": 104200
     },
     {
-      "epoch": 2.4039789997236807,
-      "grad_norm": 3.057840347290039,
-      "learning_rate": 2.778457265019415e-05,
-      "loss": 0.6343,
+      "epoch": 1.201991534753994,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7783227410886988e-05,
+      "loss": 0.9863777160644531,
       "step": 104400
     },
     {
-      "epoch": 2.4085843234779407,
-      "grad_norm": 2.8306033611297607,
-      "learning_rate": 2.7777301086371636e-05,
-      "loss": 0.645,
+      "epoch": 1.204294219828709,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.777595584706447e-05,
+      "loss": 0.9835929870605469,
       "step": 104600
     },
     {
-      "epoch": 2.4131896472322003,
-      "grad_norm": 3.3138833045959473,
-      "learning_rate": 2.777002952254912e-05,
-      "loss": 0.6502,
+      "epoch": 1.206596904903424,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.7768684283241956e-05,
+      "loss": 0.9846414947509765,
       "step": 104800
     },
     {
-      "epoch": 2.4177949709864603,
-      "grad_norm": 2.700263738632202,
-      "learning_rate": 2.7762757958726605e-05,
-      "loss": 0.6443,
+      "epoch": 1.208899589978139,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.776141271941944e-05,
+      "loss": 0.9859349060058594,
       "step": 105000
     },
     {
-      "epoch": 2.4224002947407204,
-      "grad_norm": 4.0957932472229,
-      "learning_rate": 2.7755486394904087e-05,
-      "loss": 0.6615,
+      "epoch": 1.211202275052854,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.775414115559692e-05,
+      "loss": 0.9805864715576171,
       "step": 105200
     },
     {
-      "epoch": 2.42700561849498,
-      "grad_norm": 2.5739316940307617,
-      "learning_rate": 2.7748214831081573e-05,
-      "loss": 0.6456,
+      "epoch": 1.2135049601275687,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.7746869591774407e-05,
+      "loss": 1.0142818450927735,
       "step": 105400
     },
     {
-      "epoch": 2.43161094224924,
-      "grad_norm": 3.833193302154541,
-      "learning_rate": 2.774094326725906e-05,
-      "loss": 0.664,
+      "epoch": 1.2158076452022837,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.7739598027951893e-05,
+      "loss": 0.9921229553222656,
       "step": 105600
     },
     {
-      "epoch": 2.4362162660035,
-      "grad_norm": 2.990009307861328,
-      "learning_rate": 2.7733671703436542e-05,
-      "loss": 0.6308,
+      "epoch": 1.2181103302769987,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7732326464129376e-05,
+      "loss": 0.9776945495605469,
       "step": 105800
     },
     {
-      "epoch": 2.44082158975776,
-      "grad_norm": 3.57989501953125,
-      "learning_rate": 2.7726400139614028e-05,
-      "loss": 0.6434,
+      "epoch": 1.2204130153517136,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.772505490030686e-05,
+      "loss": 0.9756576538085937,
       "step": 106000
     },
     {
-      "epoch": 2.44542691351202,
-      "grad_norm": 2.9661567211151123,
-      "learning_rate": 2.771912857579151e-05,
-      "loss": 0.6467,
+      "epoch": 1.2227157004264284,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.7717783336484344e-05,
+      "loss": 0.9888297271728516,
       "step": 106200
     },
     {
-      "epoch": 2.4500322372662797,
-      "grad_norm": 2.5463151931762695,
-      "learning_rate": 2.7711857011968993e-05,
-      "loss": 0.6579,
+      "epoch": 1.2250183855011434,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.771051177266183e-05,
+      "loss": 0.9722689819335938,
       "step": 106400
     },
     {
-      "epoch": 2.4546375610205398,
-      "grad_norm": 3.589879274368286,
-      "learning_rate": 2.770462180596559e-05,
-      "loss": 0.6557,
+      "epoch": 1.2273210705758584,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.7703240208839313e-05,
+      "loss": 0.976254653930664,
       "step": 106600
     },
     {
-      "epoch": 2.4592428847748,
-      "grad_norm": 2.7275781631469727,
-      "learning_rate": 2.7697386599962187e-05,
-      "loss": 0.6576,
+      "epoch": 1.2296237556505734,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.76959686450168e-05,
+      "loss": 0.9966973114013672,
       "step": 106800
     },
     {
-      "epoch": 2.4638482085290594,
-      "grad_norm": 3.2831528186798096,
-      "learning_rate": 2.7690115036139672e-05,
-      "loss": 0.645,
+      "epoch": 1.2319264407252883,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.7688697081194284e-05,
+      "loss": 0.9784175109863281,
       "step": 107000
     },
     {
-      "epoch": 2.4684535322833194,
-      "grad_norm": 2.328716993331909,
-      "learning_rate": 2.768284347231716e-05,
-      "loss": 0.6388,
+      "epoch": 1.234229125800003,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.7681425517371767e-05,
+      "loss": 0.9850876617431641,
       "step": 107200
     },
     {
-      "epoch": 2.4730588560375795,
-      "grad_norm": 3.319284439086914,
-      "learning_rate": 2.767557190849464e-05,
-      "loss": 0.6465,
+      "epoch": 1.236531810874718,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.767415395354925e-05,
+      "loss": 0.9829400634765625,
       "step": 107400
     },
     {
-      "epoch": 2.4776641797918395,
-      "grad_norm": 3.1933634281158447,
-      "learning_rate": 2.766833670249124e-05,
-      "loss": 0.6414,
+      "epoch": 1.238834495949433,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.7666882389726735e-05,
+      "loss": 0.9864478302001953,
       "step": 107600
     },
     {
-      "epoch": 2.482269503546099,
-      "grad_norm": 3.2471187114715576,
-      "learning_rate": 2.7661065138668725e-05,
-      "loss": 0.6406,
+      "epoch": 1.241137181024148,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.7659610825904218e-05,
+      "loss": 1.0001952362060547,
       "step": 107800
     },
     {
-      "epoch": 2.486874827300359,
-      "grad_norm": 3.0858373641967773,
-      "learning_rate": 2.7653793574846205e-05,
-      "loss": 0.6537,
+      "epoch": 1.2434398660988628,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7652339262081704e-05,
+      "loss": 0.9876364135742187,
       "step": 108000
     },
     {
-      "epoch": 2.491480151054619,
-      "grad_norm": 2.9696667194366455,
-      "learning_rate": 2.7646558368842803e-05,
-      "loss": 0.6347,
+      "epoch": 1.2457425511735778,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.764506769825919e-05,
+      "loss": 1.0017476654052735,
       "step": 108200
     },
     {
-      "epoch": 2.496085474808879,
-      "grad_norm": 3.244499683380127,
-      "learning_rate": 2.7639286805020286e-05,
-      "loss": 0.6431,
+      "epoch": 1.2480452362482928,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.7637796134436672e-05,
+      "loss": 0.9773336029052735,
       "step": 108400
     },
     {
-      "epoch": 2.500690798563139,
-      "grad_norm": 3.8631808757781982,
-      "learning_rate": 2.763201524119777e-05,
-      "loss": 0.6439,
+      "epoch": 1.2503479213230078,
+      "grad_norm": 0.5,
+      "learning_rate": 2.7630524570614158e-05,
+      "loss": 0.9803909301757813,
       "step": 108600
     },
     {
-      "epoch": 2.505296122317399,
-      "grad_norm": 3.0365819931030273,
-      "learning_rate": 2.7624743677375258e-05,
-      "loss": 0.6466,
+      "epoch": 1.2526506063977227,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.762325300679164e-05,
+      "loss": 0.9660771942138672,
       "step": 108800
     },
     {
-      "epoch": 2.509901446071659,
-      "grad_norm": 3.1504974365234375,
-      "learning_rate": 2.7617472113552743e-05,
-      "loss": 0.6623,
+      "epoch": 1.2549532914724375,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.7615981442969123e-05,
+      "loss": 0.9591829681396484,
       "step": 109000
     },
     {
-      "epoch": 2.514506769825919,
-      "grad_norm": 2.797166109085083,
-      "learning_rate": 2.7610200549730226e-05,
-      "loss": 0.6378,
+      "epoch": 1.2572559765471525,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.760870987914661e-05,
+      "loss": 1.0150857543945313,
       "step": 109200
     },
     {
-      "epoch": 2.519112093580179,
-      "grad_norm": 6.695431232452393,
-      "learning_rate": 2.7602928985907712e-05,
-      "loss": 0.6388,
+      "epoch": 1.2595586616218675,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.7601438315324095e-05,
+      "loss": 1.011795425415039,
       "step": 109400
     },
     {
-      "epoch": 2.5237174173344386,
-      "grad_norm": 2.692798614501953,
-      "learning_rate": 2.7595657422085194e-05,
-      "loss": 0.6556,
+      "epoch": 1.2618613466965825,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.759416675150158e-05,
+      "loss": 1.0024031066894532,
       "step": 109600
     },
     {
-      "epoch": 2.5283227410886986,
-      "grad_norm": 4.388221740722656,
-      "learning_rate": 2.7588385858262677e-05,
-      "loss": 0.6547,
+      "epoch": 1.2641640317712972,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.7586895187679063e-05,
+      "loss": 0.9753433990478516,
       "step": 109800
     },
     {
-      "epoch": 2.5329280648429586,
-      "grad_norm": 6.0723676681518555,
-      "learning_rate": 2.7581114294440163e-05,
-      "loss": 0.642,
+      "epoch": 1.2664667168460122,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.7579623623856546e-05,
+      "loss": 0.9753445434570313,
       "step": 110000
     },
     {
-      "epoch": 2.537533388597218,
-      "grad_norm": 3.7598865032196045,
-      "learning_rate": 2.757384273061765e-05,
-      "loss": 0.637,
+      "epoch": 1.2687694019207272,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.7572352060034032e-05,
+      "loss": 0.9854216003417968,
       "step": 110200
     },
     {
-      "epoch": 2.5421387123514783,
-      "grad_norm": 4.075303077697754,
-      "learning_rate": 2.756657116679513e-05,
-      "loss": 0.6458,
+      "epoch": 1.2710720869954422,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7565080496211514e-05,
+      "loss": 0.9756188201904297,
       "step": 110400
     },
     {
-      "epoch": 2.5467440361057383,
-      "grad_norm": 2.6612720489501953,
-      "learning_rate": 2.7559299602972617e-05,
-      "loss": 0.6478,
+      "epoch": 1.2733747720701571,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.7557808932389e-05,
+      "loss": 0.9979515075683594,
       "step": 110600
     },
     {
-      "epoch": 2.551349359859998,
-      "grad_norm": 3.044689416885376,
-      "learning_rate": 2.75520280391501e-05,
-      "loss": 0.6586,
+      "epoch": 1.2756774571448721,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.7550537368566486e-05,
+      "loss": 1.0006640625,
       "step": 110800
     },
     {
-      "epoch": 2.555954683614258,
-      "grad_norm": 3.982969284057617,
-      "learning_rate": 2.7544756475327582e-05,
-      "loss": 0.6457,
+      "epoch": 1.277980142219587,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.754326580474397e-05,
+      "loss": 0.9963182067871094,
       "step": 111000
     },
     {
-      "epoch": 2.560560007368518,
-      "grad_norm": 3.1461875438690186,
-      "learning_rate": 2.7537484911505068e-05,
-      "loss": 0.649,
+      "epoch": 1.2802828272943019,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.7535994240921455e-05,
+      "loss": 0.9819471740722656,
       "step": 111200
     },
     {
-      "epoch": 2.565165331122778,
-      "grad_norm": 3.544240951538086,
-      "learning_rate": 2.7530213347682554e-05,
-      "loss": 0.6479,
+      "epoch": 1.2825855123690169,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7528722677098937e-05,
+      "loss": 0.9869737243652343,
       "step": 111400
     },
     {
-      "epoch": 2.569770654877038,
-      "grad_norm": 3.3684897422790527,
-      "learning_rate": 2.752294178386004e-05,
-      "loss": 0.6437,
+      "epoch": 1.2848881974437316,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.752145111327642e-05,
+      "loss": 1.0044972229003906,
       "step": 111600
     },
     {
-      "epoch": 2.5743759786312976,
-      "grad_norm": 3.4466545581817627,
-      "learning_rate": 2.7515670220037523e-05,
-      "loss": 0.6518,
+      "epoch": 1.2871908825184466,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7514179549453906e-05,
+      "loss": 0.9748899078369141,
       "step": 111800
     },
     {
-      "epoch": 2.5789813023855577,
-      "grad_norm": 3.449298858642578,
-      "learning_rate": 2.7508398656215005e-05,
-      "loss": 0.6504,
+      "epoch": 1.2894935675931616,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.750690798563139e-05,
+      "loss": 0.9800602722167969,
       "step": 112000
     },
     {
-      "epoch": 2.5835866261398177,
-      "grad_norm": 2.8273544311523438,
-      "learning_rate": 2.750112709239249e-05,
-      "loss": 0.6395,
+      "epoch": 1.2917962526678766,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.7499636421808877e-05,
+      "loss": 0.9784731292724609,
       "step": 112200
     },
     {
-      "epoch": 2.5881919498940773,
-      "grad_norm": 2.8043863773345947,
-      "learning_rate": 2.7493855528569973e-05,
-      "loss": 0.6562,
+      "epoch": 1.2940989377425915,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.749236485798636e-05,
+      "loss": 0.9836808013916015,
       "step": 112400
     },
     {
-      "epoch": 2.5927972736483373,
-      "grad_norm": 3.3060178756713867,
-      "learning_rate": 2.748658396474746e-05,
-      "loss": 0.637,
+      "epoch": 1.2964016228173065,
+      "grad_norm": 2.5,
+      "learning_rate": 2.7485093294163843e-05,
+      "loss": 0.991956558227539,
       "step": 112600
     },
     {
-      "epoch": 2.5974025974025974,
-      "grad_norm": 3.410897970199585,
-      "learning_rate": 2.7479312400924945e-05,
-      "loss": 0.6376,
+      "epoch": 1.2987043078920213,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.747782173034133e-05,
+      "loss": 0.9863188934326171,
       "step": 112800
     },
     {
-      "epoch": 2.6020079211568574,
-      "grad_norm": 3.279798984527588,
-      "learning_rate": 2.7472040837102428e-05,
-      "loss": 0.6449,
+      "epoch": 1.3010069929667363,
+      "grad_norm": 1.0,
+      "learning_rate": 2.747055016651881e-05,
+      "loss": 1.0016487121582032,
       "step": 113000
     },
     {
-      "epoch": 2.6066132449111175,
-      "grad_norm": 3.1562256813049316,
-      "learning_rate": 2.7464769273279914e-05,
-      "loss": 0.6447,
+      "epoch": 1.3033096780414513,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.7463278602696297e-05,
+      "loss": 0.9983952331542969,
       "step": 113200
     },
     {
-      "epoch": 2.611218568665377,
-      "grad_norm": 2.973257303237915,
-      "learning_rate": 2.7457534067276512e-05,
-      "loss": 0.6406,
+      "epoch": 1.305612363116166,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.7456007038873783e-05,
+      "loss": 1.014336395263672,
       "step": 113400
     },
     {
-      "epoch": 2.615823892419637,
-      "grad_norm": 3.439020872116089,
-      "learning_rate": 2.745026250345399e-05,
-      "loss": 0.6391,
+      "epoch": 1.307915048190881,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7448735475051265e-05,
+      "loss": 0.970293197631836,
       "step": 113600
     },
     {
-      "epoch": 2.620429216173897,
-      "grad_norm": 2.99761700630188,
-      "learning_rate": 2.7442990939631477e-05,
-      "loss": 0.6486,
+      "epoch": 1.310217733265596,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.7441463911228748e-05,
+      "loss": 0.9937506866455078,
       "step": 113800
     },
     {
-      "epoch": 2.6250345399281567,
-      "grad_norm": 3.7341833114624023,
-      "learning_rate": 2.7435719375808963e-05,
-      "loss": 0.6371,
+      "epoch": 1.312520418340311,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.7434192347406234e-05,
+      "loss": 0.9721630859375,
       "step": 114000
     },
     {
-      "epoch": 2.6296398636824168,
-      "grad_norm": 3.088653326034546,
-      "learning_rate": 2.7428447811986446e-05,
-      "loss": 0.6513,
+      "epoch": 1.314823103415026,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.7426920783583716e-05,
+      "loss": 0.9970010375976562,
       "step": 114200
     },
     {
-      "epoch": 2.634245187436677,
-      "grad_norm": 3.4004712104797363,
-      "learning_rate": 2.7421176248163932e-05,
-      "loss": 0.6321,
+      "epoch": 1.317125788489741,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.7419649219761202e-05,
+      "loss": 1.006850814819336,
       "step": 114400
     },
     {
-      "epoch": 2.638850511190937,
-      "grad_norm": 2.652799129486084,
-      "learning_rate": 2.7413904684341418e-05,
-      "loss": 0.6364,
+      "epoch": 1.3194284735644557,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7412377655938688e-05,
+      "loss": 1.0042724609375,
       "step": 114600
     },
     {
-      "epoch": 2.643455834945197,
-      "grad_norm": 3.6918208599090576,
-      "learning_rate": 2.7406633120518897e-05,
-      "loss": 0.6341,
+      "epoch": 1.3217311586391707,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.740510609211617e-05,
+      "loss": 0.9619822692871094,
       "step": 114800
     },
     {
-      "epoch": 2.6480611586994565,
-      "grad_norm": 3.6925747394561768,
-      "learning_rate": 2.7399361556696383e-05,
-      "loss": 0.6422,
+      "epoch": 1.3240338437138857,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.7397834528293657e-05,
+      "loss": 0.9862764739990234,
       "step": 115000
     },
     {
-      "epoch": 2.6526664824537165,
-      "grad_norm": 3.1433355808258057,
-      "learning_rate": 2.739208999287387e-05,
-      "loss": 0.6372,
+      "epoch": 1.3263365287886004,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.739056296447114e-05,
+      "loss": 0.9918037414550781,
       "step": 115200
     },
     {
-      "epoch": 2.6572718062079765,
-      "grad_norm": 2.9344465732574463,
-      "learning_rate": 2.738481842905135e-05,
-      "loss": 0.6478,
+      "epoch": 1.3286392138633154,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.7383291400648625e-05,
+      "loss": 1.0145311737060547,
       "step": 115400
     },
     {
-      "epoch": 2.661877129962236,
-      "grad_norm": 3.018880605697632,
-      "learning_rate": 2.7377546865228837e-05,
-      "loss": 0.6395,
+      "epoch": 1.3309418989380304,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.7376019836826108e-05,
+      "loss": 0.9897518920898437,
       "step": 115600
     },
     {
-      "epoch": 2.666482453716496,
-      "grad_norm": 2.8171603679656982,
-      "learning_rate": 2.7370275301406323e-05,
-      "loss": 0.6492,
+      "epoch": 1.3332445840127454,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.7368748273003593e-05,
+      "loss": 0.9979398345947266,
       "step": 115800
     },
     {
-      "epoch": 2.671087777470756,
-      "grad_norm": 2.8728506565093994,
-      "learning_rate": 2.7363003737583802e-05,
-      "loss": 0.6405,
+      "epoch": 1.3355472690874604,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.736147670918108e-05,
+      "loss": 0.9715717315673829,
       "step": 116000
     },
     {
-      "epoch": 2.6756931012250162,
-      "grad_norm": 3.09126877784729,
-      "learning_rate": 2.7355732173761288e-05,
-      "loss": 0.6398,
+      "epoch": 1.3378499541621753,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.7354205145358562e-05,
+      "loss": 0.9838251495361328,
       "step": 116200
     },
     {
-      "epoch": 2.6802984249792763,
-      "grad_norm": 3.089311122894287,
-      "learning_rate": 2.7348460609938774e-05,
-      "loss": 0.6494,
+      "epoch": 1.34015263923689,
+      "grad_norm": 0.875,
+      "learning_rate": 2.7346933581536044e-05,
+      "loss": 0.9873313140869141,
       "step": 116400
     },
     {
-      "epoch": 2.684903748733536,
-      "grad_norm": 2.3543293476104736,
-      "learning_rate": 2.734118904611626e-05,
-      "loss": 0.6376,
+      "epoch": 1.342455324311605,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.733966201771353e-05,
+      "loss": 0.9789092254638672,
       "step": 116600
     },
     {
-      "epoch": 2.689509072487796,
-      "grad_norm": 2.459444046020508,
-      "learning_rate": 2.7333917482293742e-05,
-      "loss": 0.6337,
+      "epoch": 1.34475800938632,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7332390453891013e-05,
+      "loss": 0.9955917358398437,
       "step": 116800
     },
     {
-      "epoch": 2.694114396242056,
-      "grad_norm": 3.3463871479034424,
-      "learning_rate": 2.7326645918471228e-05,
-      "loss": 0.6359,
+      "epoch": 1.347060694461035,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.73251188900685e-05,
+      "loss": 0.985167007446289,
       "step": 117000
     },
     {
-      "epoch": 2.6987197199963155,
-      "grad_norm": 2.9327232837677,
-      "learning_rate": 2.7319374354648714e-05,
-      "loss": 0.6264,
+      "epoch": 1.3493633795357498,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7317847326245985e-05,
+      "loss": 0.9772294616699219,
       "step": 117200
     },
     {
-      "epoch": 2.7033250437505756,
-      "grad_norm": 3.052452802658081,
-      "learning_rate": 2.7312102790826193e-05,
-      "loss": 0.6466,
+      "epoch": 1.3516660646104648,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7310575762423467e-05,
+      "loss": 0.9946612548828125,
       "step": 117400
     },
     {
-      "epoch": 2.7079303675048356,
-      "grad_norm": 3.294747829437256,
-      "learning_rate": 2.730483122700368e-05,
-      "loss": 0.6525,
+      "epoch": 1.3539687496851798,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.730330419860095e-05,
+      "loss": 0.97852783203125,
       "step": 117600
     },
     {
-      "epoch": 2.7125356912590957,
-      "grad_norm": 3.948997735977173,
-      "learning_rate": 2.7297596021000278e-05,
-      "loss": 0.6457,
+      "epoch": 1.3562714347598948,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.7296032634778436e-05,
+      "loss": 1.0050527191162109,
       "step": 117800
     },
     {
-      "epoch": 2.7171410150133557,
-      "grad_norm": 2.7049851417541504,
-      "learning_rate": 2.729032445717776e-05,
-      "loss": 0.6435,
+      "epoch": 1.3585741198346097,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.728876107095592e-05,
+      "loss": 0.9625944519042968,
       "step": 118000
     },
     {
-      "epoch": 2.7217463387676153,
-      "grad_norm": 3.9702322483062744,
-      "learning_rate": 2.7283052893355246e-05,
-      "loss": 0.6516,
+      "epoch": 1.3608768049093245,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.7281489507133404e-05,
+      "loss": 0.9805034637451172,
       "step": 118200
     },
     {
-      "epoch": 2.7263516625218753,
-      "grad_norm": 2.606865882873535,
-      "learning_rate": 2.7275781329532732e-05,
-      "loss": 0.6551,
+      "epoch": 1.3631794899840395,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.727421794331089e-05,
+      "loss": 0.991854248046875,
       "step": 118400
     },
     {
-      "epoch": 2.7309569862761354,
-      "grad_norm": 3.387070894241333,
-      "learning_rate": 2.7268509765710215e-05,
-      "loss": 0.6403,
+      "epoch": 1.3654821750587545,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.7266946379488376e-05,
+      "loss": 0.9891043853759766,
       "step": 118600
     },
     {
-      "epoch": 2.735562310030395,
-      "grad_norm": 3.625943899154663,
-      "learning_rate": 2.7261238201887697e-05,
-      "loss": 0.6269,
+      "epoch": 1.3677848601334694,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.725967481566586e-05,
+      "loss": 0.9937184143066407,
       "step": 118800
     },
     {
-      "epoch": 2.740167633784655,
-      "grad_norm": 2.9141345024108887,
-      "learning_rate": 2.7254002995884292e-05,
-      "loss": 0.6501,
+      "epoch": 1.3700875452081842,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.725240325184334e-05,
+      "loss": 0.9900875854492187,
       "step": 119000
     },
     {
-      "epoch": 2.744772957538915,
-      "grad_norm": 3.324512004852295,
-      "learning_rate": 2.724673143206178e-05,
-      "loss": 0.6393,
+      "epoch": 1.3723902302828992,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.7245131688020827e-05,
+      "loss": 0.956904296875,
       "step": 119200
     },
     {
-      "epoch": 2.749378281293175,
-      "grad_norm": 3.5145280361175537,
-      "learning_rate": 2.7239459868239264e-05,
-      "loss": 0.6319,
+      "epoch": 1.3746929153576142,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.723786012419831e-05,
+      "loss": 0.989427719116211,
       "step": 119400
     },
     {
-      "epoch": 2.7539836050474347,
-      "grad_norm": 2.751305103302002,
-      "learning_rate": 2.723218830441675e-05,
-      "loss": 0.6362,
+      "epoch": 1.3769956004323292,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.7230588560375795e-05,
+      "loss": 0.9846376800537109,
       "step": 119600
     },
     {
-      "epoch": 2.7585889288016947,
-      "grad_norm": 3.314880132675171,
-      "learning_rate": 2.7224916740594233e-05,
-      "loss": 0.635,
+      "epoch": 1.3792982855070441,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.722331699655328e-05,
+      "loss": 0.9800340270996094,
       "step": 119800
     },
     {
-      "epoch": 2.7631942525559547,
-      "grad_norm": 3.0874931812286377,
-      "learning_rate": 2.721764517677172e-05,
-      "loss": 0.6568,
+      "epoch": 1.3816009705817591,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.7216045432730764e-05,
+      "loss": 0.9806147003173828,
       "step": 120000
     },
     {
-      "epoch": 2.767799576310215,
-      "grad_norm": 3.5240249633789062,
-      "learning_rate": 2.7210373612949205e-05,
-      "loss": 0.6494,
+      "epoch": 1.3839036556564739,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.7208773868908246e-05,
+      "loss": 0.978929672241211,
       "step": 120200
     },
     {
-      "epoch": 2.7724049000644744,
-      "grad_norm": 2.875325918197632,
-      "learning_rate": 2.7203102049126684e-05,
-      "loss": 0.6355,
+      "epoch": 1.3862063407311889,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.7201502305085732e-05,
+      "loss": 0.9661943054199219,
       "step": 120400
     },
     {
-      "epoch": 2.7770102238187344,
-      "grad_norm": 3.846749782562256,
-      "learning_rate": 2.719583048530417e-05,
-      "loss": 0.6338,
+      "epoch": 1.3885090258059039,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.7194230741263215e-05,
+      "loss": 0.9726935577392578,
       "step": 120600
     },
     {
-      "epoch": 2.7816155475729945,
-      "grad_norm": 3.331397533416748,
-      "learning_rate": 2.7188558921481656e-05,
-      "loss": 0.6511,
+      "epoch": 1.3908117108806186,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.71869591774407e-05,
+      "loss": 0.9841280364990235,
       "step": 120800
     },
     {
-      "epoch": 2.786220871327254,
-      "grad_norm": 3.6676218509674072,
-      "learning_rate": 2.7181287357659138e-05,
-      "loss": 0.6326,
+      "epoch": 1.3931143959553336,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.7179687613618187e-05,
+      "loss": 1.0080772399902345,
       "step": 121000
     },
     {
-      "epoch": 2.790826195081514,
-      "grad_norm": 3.7938387393951416,
-      "learning_rate": 2.7174015793836624e-05,
-      "loss": 0.6354,
+      "epoch": 1.3954170810300486,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.7172416049795672e-05,
+      "loss": 0.9903179931640625,
       "step": 121200
     },
     {
-      "epoch": 2.795431518835774,
-      "grad_norm": 3.1966028213500977,
-      "learning_rate": 2.716674423001411e-05,
-      "loss": 0.6457,
+      "epoch": 1.3977197661047636,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.7165144485973155e-05,
+      "loss": 0.97442626953125,
       "step": 121400
     },
     {
-      "epoch": 2.800036842590034,
-      "grad_norm": 2.660951852798462,
-      "learning_rate": 2.715947266619159e-05,
-      "loss": 0.6536,
+      "epoch": 1.4000224511794785,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.7157872922150637e-05,
+      "loss": 0.9792997741699219,
       "step": 121600
     },
     {
-      "epoch": 2.804642166344294,
-      "grad_norm": 3.02982234954834,
-      "learning_rate": 2.7152201102369075e-05,
-      "loss": 0.6507,
+      "epoch": 1.4023251362541935,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7150601358328123e-05,
+      "loss": 0.9764759063720703,
       "step": 121800
     },
     {
-      "epoch": 2.809247490098554,
-      "grad_norm": 3.0377113819122314,
-      "learning_rate": 2.714492953854656e-05,
-      "loss": 0.645,
+      "epoch": 1.4046278213289083,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.7143329794505606e-05,
+      "loss": 1.0077233123779297,
       "step": 122000
     },
     {
-      "epoch": 2.813852813852814,
-      "grad_norm": 3.3548905849456787,
-      "learning_rate": 2.7137657974724043e-05,
-      "loss": 0.636,
+      "epoch": 1.4069305064036233,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.7136058230683092e-05,
+      "loss": 0.9881072235107422,
       "step": 122200
     },
     {
-      "epoch": 2.818458137607074,
-      "grad_norm": 4.323446750640869,
-      "learning_rate": 2.713038641090153e-05,
-      "loss": 0.6529,
+      "epoch": 1.4092331914783383,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.7128786666860578e-05,
+      "loss": 0.9633769989013672,
       "step": 122400
     },
     {
-      "epoch": 2.8230634613613335,
-      "grad_norm": 3.245694637298584,
-      "learning_rate": 2.7123114847079015e-05,
-      "loss": 0.6364,
+      "epoch": 1.411535876553053,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.712151510303806e-05,
+      "loss": 0.9723685455322265,
       "step": 122600
     },
     {
-      "epoch": 2.8276687851155935,
-      "grad_norm": 2.8766958713531494,
-      "learning_rate": 2.7115843283256498e-05,
-      "loss": 0.6388,
+      "epoch": 1.413838561627768,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.7114243539215543e-05,
+      "loss": 0.9701836395263672,
       "step": 122800
     },
     {
-      "epoch": 2.8322741088698535,
-      "grad_norm": 2.433682441711426,
-      "learning_rate": 2.7108608077253093e-05,
-      "loss": 0.6383,
+      "epoch": 1.416141246702483,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.710697197539303e-05,
+      "loss": 0.9833367919921875,
       "step": 123000
     },
     {
-      "epoch": 2.8368794326241136,
-      "grad_norm": 3.1630749702453613,
-      "learning_rate": 2.710137287124969e-05,
-      "loss": 0.6312,
+      "epoch": 1.418443931777198,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.709970041157051e-05,
+      "loss": 0.9934508514404297,
       "step": 123200
     },
     {
-      "epoch": 2.8414847563783736,
-      "grad_norm": 3.1013870239257812,
-      "learning_rate": 2.7094101307427174e-05,
-      "loss": 0.6377,
+      "epoch": 1.420746616851913,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7092428847747997e-05,
+      "loss": 0.9717855072021484,
       "step": 123400
     },
     {
-      "epoch": 2.846090080132633,
-      "grad_norm": 3.5772645473480225,
-      "learning_rate": 2.708682974360466e-05,
-      "loss": 0.6404,
+      "epoch": 1.423049301926628,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7085157283925483e-05,
+      "loss": 1.0004611206054688,
       "step": 123600
     },
     {
-      "epoch": 2.8506954038868932,
-      "grad_norm": 2.970681667327881,
-      "learning_rate": 2.7079558179782146e-05,
-      "loss": 0.6472,
+      "epoch": 1.4253519870013427,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7077885720102966e-05,
+      "loss": 0.991996841430664,
       "step": 123800
     },
     {
-      "epoch": 2.8553007276411533,
-      "grad_norm": 2.7734873294830322,
-      "learning_rate": 2.707228661595963e-05,
-      "loss": 0.6329,
+      "epoch": 1.4276546720760577,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.7070614156280448e-05,
+      "loss": 0.9875720977783203,
       "step": 124000
     },
     {
-      "epoch": 2.859906051395413,
-      "grad_norm": 3.264719247817993,
-      "learning_rate": 2.7065015052137114e-05,
-      "loss": 0.6518,
+      "epoch": 1.4299573571507727,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.7063342592457934e-05,
+      "loss": 0.9992472076416016,
       "step": 124200
     },
     {
-      "epoch": 2.864511375149673,
-      "grad_norm": 3.1306731700897217,
-      "learning_rate": 2.70577434883146e-05,
-      "loss": 0.652,
+      "epoch": 1.4322600422254876,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.705607102863542e-05,
+      "loss": 0.9760001373291015,
       "step": 124400
     },
     {
-      "epoch": 2.869116698903933,
-      "grad_norm": 3.149149179458618,
-      "learning_rate": 2.705047192449208e-05,
-      "loss": 0.6429,
+      "epoch": 1.4345627273002024,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.7048799464812902e-05,
+      "loss": 1.0076248168945312,
       "step": 124600
     },
     {
-      "epoch": 2.873722022658193,
-      "grad_norm": 3.6126046180725098,
-      "learning_rate": 2.7043200360669565e-05,
-      "loss": 0.6355,
+      "epoch": 1.4368654123749174,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.704152790099039e-05,
+      "loss": 0.986727066040039,
       "step": 124800
     },
     {
-      "epoch": 2.878327346412453,
-      "grad_norm": 2.957371234893799,
-      "learning_rate": 2.703592879684705e-05,
-      "loss": 0.6425,
+      "epoch": 1.4391680974496324,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.7034256337167874e-05,
+      "loss": 0.9612811279296875,
       "step": 125000
     },
     {
-      "epoch": 2.8829326701667126,
-      "grad_norm": 2.859186887741089,
-      "learning_rate": 2.7028693590843646e-05,
-      "loss": 0.6445,
+      "epoch": 1.4414707825243473,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7026984773345357e-05,
+      "loss": 0.9965091705322265,
       "step": 125200
     },
     {
-      "epoch": 2.8875379939209727,
-      "grad_norm": 3.0626325607299805,
-      "learning_rate": 2.7021422027021132e-05,
-      "loss": 0.6572,
+      "epoch": 1.4437734675990623,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.701971320952284e-05,
+      "loss": 0.984040756225586,
       "step": 125400
     },
     {
-      "epoch": 2.8921433176752327,
-      "grad_norm": 2.7346231937408447,
-      "learning_rate": 2.7014150463198618e-05,
-      "loss": 0.6421,
+      "epoch": 1.446076152673777,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.7012441645700325e-05,
+      "loss": 0.9824740600585937,
       "step": 125600
     },
     {
-      "epoch": 2.8967486414294923,
-      "grad_norm": 2.984370470046997,
-      "learning_rate": 2.70068788993761e-05,
-      "loss": 0.6436,
+      "epoch": 1.448378837748492,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.7005170081877808e-05,
+      "loss": 0.9734669494628906,
       "step": 125800
     },
     {
-      "epoch": 2.9013539651837523,
-      "grad_norm": 3.117213726043701,
-      "learning_rate": 2.6999607335553583e-05,
-      "loss": 0.6493,
+      "epoch": 1.450681522823207,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.6997898518055294e-05,
+      "loss": 0.9760218811035156,
       "step": 126000
     },
     {
-      "epoch": 2.9059592889380124,
-      "grad_norm": 3.779122829437256,
-      "learning_rate": 2.699233577173107e-05,
-      "loss": 0.6394,
+      "epoch": 1.452984207897922,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.699062695423278e-05,
+      "loss": 0.979968032836914,
       "step": 126200
     },
     {
-      "epoch": 2.9105646126922724,
-      "grad_norm": 3.12115740776062,
-      "learning_rate": 2.6985064207908552e-05,
-      "loss": 0.6344,
+      "epoch": 1.4552868929726368,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.6983355390410262e-05,
+      "loss": 0.9945246887207031,
       "step": 126400
     },
     {
-      "epoch": 2.9151699364465324,
-      "grad_norm": 3.0919904708862305,
-      "learning_rate": 2.6977792644086038e-05,
-      "loss": 0.6405,
+      "epoch": 1.4575895780473518,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.6976083826587745e-05,
+      "loss": 0.9984708404541016,
       "step": 126600
     },
     {
-      "epoch": 2.919775260200792,
-      "grad_norm": 2.8004214763641357,
-      "learning_rate": 2.6970521080263524e-05,
-      "loss": 0.6489,
+      "epoch": 1.4598922631220668,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.696881226276523e-05,
+      "loss": 0.9899350738525391,
       "step": 126800
     },
     {
-      "epoch": 2.924380583955052,
-      "grad_norm": 2.920830726623535,
-      "learning_rate": 2.6963249516441006e-05,
-      "loss": 0.6394,
+      "epoch": 1.4621949481967818,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.6961540698942716e-05,
+      "loss": 1.0038150787353515,
       "step": 127000
     },
     {
-      "epoch": 2.928985907709312,
-      "grad_norm": 3.278106451034546,
-      "learning_rate": 2.6955977952618492e-05,
-      "loss": 0.6434,
+      "epoch": 1.4644976332714967,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.69542691351202e-05,
+      "loss": 0.9827015686035157,
       "step": 127200
     },
     {
-      "epoch": 2.9335912314635717,
-      "grad_norm": 3.6461238861083984,
-      "learning_rate": 2.6948706388795975e-05,
-      "loss": 0.646,
+      "epoch": 1.4668003183462117,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6946997571297685e-05,
+      "loss": 0.9892060089111329,
       "step": 127400
     },
     {
-      "epoch": 2.9381965552178317,
-      "grad_norm": 3.31779146194458,
-      "learning_rate": 2.694143482497346e-05,
-      "loss": 0.6361,
+      "epoch": 1.4691030034209265,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.693972600747517e-05,
+      "loss": 0.9666407012939453,
       "step": 127600
     },
     {
-      "epoch": 2.942801878972092,
-      "grad_norm": 2.7438087463378906,
-      "learning_rate": 2.6934199618970056e-05,
-      "loss": 0.6437,
+      "epoch": 1.4714056884956415,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.693245444365265e-05,
+      "loss": 0.973760986328125,
       "step": 127800
     },
     {
-      "epoch": 2.947407202726352,
-      "grad_norm": 2.8473029136657715,
-      "learning_rate": 2.692692805514754e-05,
-      "loss": 0.6338,
+      "epoch": 1.4737083735703564,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6925182879830136e-05,
+      "loss": 0.9950855255126954,
       "step": 128000
     },
     {
-      "epoch": 2.952012526480612,
-      "grad_norm": 2.5815622806549072,
-      "learning_rate": 2.6919656491325024e-05,
-      "loss": 0.6371,
+      "epoch": 1.4760110586450712,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.6917911316007622e-05,
+      "loss": 0.9772869873046875,
       "step": 128200
     },
     {
-      "epoch": 2.9566178502348714,
-      "grad_norm": 3.202380657196045,
-      "learning_rate": 2.691238492750251e-05,
-      "loss": 0.6469,
+      "epoch": 1.4783137437197862,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.6910639752185104e-05,
+      "loss": 1.0049010467529298,
       "step": 128400
     },
     {
-      "epoch": 2.9612231739891315,
-      "grad_norm": 3.2944741249084473,
-      "learning_rate": 2.6905113363679996e-05,
-      "loss": 0.6515,
+      "epoch": 1.4806164287945012,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.690336818836259e-05,
+      "loss": 0.9666265106201172,
       "step": 128600
     },
     {
-      "epoch": 2.9658284977433915,
-      "grad_norm": 2.853166103363037,
-      "learning_rate": 2.6897841799857475e-05,
-      "loss": 0.6392,
+      "epoch": 1.4829191138692162,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.6896096624540076e-05,
+      "loss": 0.9786612701416015,
       "step": 128800
     },
     {
-      "epoch": 2.970433821497651,
-      "grad_norm": 3.1663033962249756,
-      "learning_rate": 2.689057023603496e-05,
-      "loss": 0.6454,
+      "epoch": 1.4852217989439311,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.688882506071756e-05,
+      "loss": 1.0051888275146483,
       "step": 129000
     },
     {
-      "epoch": 2.975039145251911,
-      "grad_norm": 2.5806081295013428,
-      "learning_rate": 2.6883298672212447e-05,
-      "loss": 0.6411,
+      "epoch": 1.4875244840186461,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.688155349689504e-05,
+      "loss": 0.9594706726074219,
       "step": 129200
     },
     {
-      "epoch": 2.979644469006171,
-      "grad_norm": 2.8585219383239746,
-      "learning_rate": 2.6876027108389933e-05,
-      "loss": 0.6475,
+      "epoch": 1.4898271690933609,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.6874281933072527e-05,
+      "loss": 0.9975163269042969,
       "step": 129400
     },
     {
-      "epoch": 2.984249792760431,
-      "grad_norm": 2.955064058303833,
-      "learning_rate": 2.6868755544567415e-05,
-      "loss": 0.6548,
+      "epoch": 1.4921298541680759,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.686701036925001e-05,
+      "loss": 0.9801680755615234,
       "step": 129600
     },
     {
-      "epoch": 2.988855116514691,
-      "grad_norm": 4.072856426239014,
-      "learning_rate": 2.68614839807449e-05,
-      "loss": 0.6538,
+      "epoch": 1.4944325392427908,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.6859738805427496e-05,
+      "loss": 0.9652777862548828,
       "step": 129800
     },
     {
-      "epoch": 2.993460440268951,
-      "grad_norm": 3.6334667205810547,
-      "learning_rate": 2.6854212416922384e-05,
-      "loss": 0.6406,
+      "epoch": 1.4967352243175056,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.685246724160498e-05,
+      "loss": 0.9953791809082031,
       "step": 130000
     },
     {
-      "epoch": 2.998065764023211,
-      "grad_norm": 2.8243021965026855,
-      "learning_rate": 2.6846940853099866e-05,
-      "loss": 0.6339,
+      "epoch": 1.4990379093922206,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6845195677782467e-05,
+      "loss": 0.9773629760742187,
       "step": 130200
     },
     {
-      "epoch": 3.0,
-      "eval_loss": 0.5988962650299072,
-      "eval_runtime": 145.7025,
-      "eval_samples_per_second": 194.65,
-      "eval_steps_per_second": 12.169,
-      "step": 130284
-    },
-    {
-      "epoch": 3.002671087777471,
-      "grad_norm": 2.527531385421753,
-      "learning_rate": 2.6839669289277352e-05,
-      "loss": 0.6279,
+      "epoch": 1.5013405944669356,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6837924113959947e-05,
+      "loss": 0.9973490142822266,
       "step": 130400
     },
     {
-      "epoch": 3.0072764115317305,
-      "grad_norm": 3.933335304260254,
-      "learning_rate": 2.6832397725454838e-05,
-      "loss": 0.6282,
+      "epoch": 1.5036432795416506,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.6830652550137432e-05,
+      "loss": 0.9864173126220703,
       "step": 130600
     },
     {
-      "epoch": 3.0118817352859906,
-      "grad_norm": 3.1796278953552246,
-      "learning_rate": 2.682512616163232e-05,
-      "loss": 0.6296,
+      "epoch": 1.5059459646163655,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.682338098631492e-05,
+      "loss": 0.9872917175292969,
       "step": 130800
     },
     {
-      "epoch": 3.0164870590402506,
-      "grad_norm": 2.6977293491363525,
-      "learning_rate": 2.681789095562892e-05,
-      "loss": 0.6249,
+      "epoch": 1.5082486496910805,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.68161094224924e-05,
+      "loss": 0.9996871185302735,
       "step": 131000
     },
     {
-      "epoch": 3.0210923827945106,
-      "grad_norm": 3.7220981121063232,
-      "learning_rate": 2.6810619391806405e-05,
-      "loss": 0.6298,
+      "epoch": 1.5105513347657955,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.6808837858669887e-05,
+      "loss": 1.005075912475586,
       "step": 131200
     },
     {
-      "epoch": 3.0256977065487702,
-      "grad_norm": 3.126789093017578,
-      "learning_rate": 2.6803347827983888e-05,
-      "loss": 0.6473,
+      "epoch": 1.5128540198405103,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.6801566294847373e-05,
+      "loss": 0.9876535034179688,
       "step": 131400
     },
     {
-      "epoch": 3.0303030303030303,
-      "grad_norm": 2.9881398677825928,
-      "learning_rate": 2.679607626416137e-05,
-      "loss": 0.6445,
+      "epoch": 1.5151567049152252,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.6794294731024855e-05,
+      "loss": 0.9811707305908203,
       "step": 131600
     },
     {
-      "epoch": 3.0349083540572903,
-      "grad_norm": 4.632811069488525,
-      "learning_rate": 2.6788804700338856e-05,
-      "loss": 0.6415,
+      "epoch": 1.51745938998994,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.6787023167202338e-05,
+      "loss": 0.9567855072021484,
       "step": 131800
     },
     {
-      "epoch": 3.0395136778115504,
-      "grad_norm": 3.6200356483459473,
-      "learning_rate": 2.678153313651634e-05,
-      "loss": 0.6242,
+      "epoch": 1.519762075064655,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.6779751603379824e-05,
+      "loss": 0.9803247833251953,
       "step": 132000
     },
     {
-      "epoch": 3.04411900156581,
-      "grad_norm": 3.1809945106506348,
-      "learning_rate": 2.6774261572693825e-05,
-      "loss": 0.638,
+      "epoch": 1.52206476013937,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.6772480039557306e-05,
+      "loss": 0.9724285125732421,
       "step": 132200
     },
     {
-      "epoch": 3.04872432532007,
-      "grad_norm": 4.0169806480407715,
-      "learning_rate": 2.676699000887131e-05,
-      "loss": 0.6335,
+      "epoch": 1.524367445214085,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6765208475734792e-05,
+      "loss": 0.9653998565673828,
       "step": 132400
     },
     {
-      "epoch": 3.05332964907433,
-      "grad_norm": 3.9464833736419678,
-      "learning_rate": 2.6759718445048793e-05,
-      "loss": 0.6376,
+      "epoch": 1.5266701302888,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.6757936911912278e-05,
+      "loss": 0.9880249786376953,
       "step": 132600
     },
     {
-      "epoch": 3.05793497282859,
-      "grad_norm": 2.9559237957000732,
-      "learning_rate": 2.6752446881226276e-05,
-      "loss": 0.6343,
+      "epoch": 1.528972815363515,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6750665348089764e-05,
+      "loss": 0.9710153198242187,
       "step": 132800
     },
     {
-      "epoch": 3.0625402965828497,
-      "grad_norm": 3.0488362312316895,
-      "learning_rate": 2.674517531740376e-05,
-      "loss": 0.6426,
+      "epoch": 1.53127550043823,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.6743393784267243e-05,
+      "loss": 0.9638467407226563,
       "step": 133000
     },
     {
-      "epoch": 3.0671456203371097,
-      "grad_norm": 3.531463861465454,
-      "learning_rate": 2.6737903753581244e-05,
-      "loss": 0.6438,
+      "epoch": 1.5335781855129447,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.673612222044473e-05,
+      "loss": 0.9837078094482422,
       "step": 133200
     },
     {
-      "epoch": 3.0717509440913697,
-      "grad_norm": 3.7633442878723145,
-      "learning_rate": 2.673063218975873e-05,
-      "loss": 0.6383,
+      "epoch": 1.5358808705876597,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.6728850656622215e-05,
+      "loss": 0.9877970886230468,
       "step": 133400
     },
     {
-      "epoch": 3.0763562678456298,
-      "grad_norm": 2.790278196334839,
-      "learning_rate": 2.6723360625936216e-05,
-      "loss": 0.6376,
+      "epoch": 1.5381835556623744,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.6721579092799697e-05,
+      "loss": 0.9861704254150391,
       "step": 133600
     },
     {
-      "epoch": 3.0809615915998894,
-      "grad_norm": 3.4008514881134033,
-      "learning_rate": 2.67160890621137e-05,
-      "loss": 0.6293,
+      "epoch": 1.5404862407370894,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6714307528977183e-05,
+      "loss": 0.9800881958007812,
       "step": 133800
     },
     {
-      "epoch": 3.0855669153541494,
-      "grad_norm": 2.998616933822632,
-      "learning_rate": 2.6708817498291184e-05,
-      "loss": 0.6387,
+      "epoch": 1.5427889258118044,
+      "grad_norm": 1.0,
+      "learning_rate": 2.670703596515467e-05,
+      "loss": 0.9763209533691406,
       "step": 134000
     },
     {
-      "epoch": 3.0901722391084094,
-      "grad_norm": 3.212735891342163,
-      "learning_rate": 2.6701545934468667e-05,
-      "loss": 0.6208,
+      "epoch": 1.5450916108865194,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.669976440133215e-05,
+      "loss": 0.9747321319580078,
       "step": 134200
     },
     {
-      "epoch": 3.094777562862669,
-      "grad_norm": 3.259173631668091,
-      "learning_rate": 2.6694310728465262e-05,
-      "loss": 0.6398,
+      "epoch": 1.5473942959612343,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.6692492837509634e-05,
+      "loss": 0.9830812072753906,
       "step": 134400
     },
     {
-      "epoch": 3.099382886616929,
-      "grad_norm": 3.346163511276245,
-      "learning_rate": 2.6687039164642748e-05,
-      "loss": 0.6488,
+      "epoch": 1.5496969810359493,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.668522127368712e-05,
+      "loss": 0.967867660522461,
       "step": 134600
     },
     {
-      "epoch": 3.103988210371189,
-      "grad_norm": 3.005929708480835,
-      "learning_rate": 2.6679767600820234e-05,
-      "loss": 0.6339,
+      "epoch": 1.5519996661106643,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.6677949709864603e-05,
+      "loss": 0.9955473327636719,
       "step": 134800
     },
     {
-      "epoch": 3.108593534125449,
-      "grad_norm": 3.2009568214416504,
-      "learning_rate": 2.6672496036997716e-05,
-      "loss": 0.63,
+      "epoch": 1.554302351185379,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.667067814604209e-05,
+      "loss": 0.9771687316894532,
       "step": 135000
     },
     {
-      "epoch": 3.1131988578797087,
-      "grad_norm": 2.799675226211548,
-      "learning_rate": 2.6665224473175202e-05,
-      "loss": 0.6275,
+      "epoch": 1.556605036260094,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.6663406582219575e-05,
+      "loss": 0.9583516693115235,
       "step": 135200
     },
     {
-      "epoch": 3.1178041816339688,
-      "grad_norm": 3.211223840713501,
-      "learning_rate": 2.6657952909352688e-05,
-      "loss": 0.6436,
+      "epoch": 1.5589077213348088,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.6656135018397057e-05,
+      "loss": 0.9967800140380859,
       "step": 135400
     },
     {
-      "epoch": 3.122409505388229,
-      "grad_norm": 2.964587926864624,
-      "learning_rate": 2.665068134553017e-05,
-      "loss": 0.6512,
+      "epoch": 1.5612104064095238,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.664886345457454e-05,
+      "loss": 0.9840859985351562,
       "step": 135600
     },
     {
-      "epoch": 3.127014829142489,
-      "grad_norm": 3.677194595336914,
-      "learning_rate": 2.6643409781707653e-05,
-      "loss": 0.6396,
+      "epoch": 1.5635130914842388,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.6641591890752026e-05,
+      "loss": 0.9754515075683594,
       "step": 135800
     },
     {
-      "epoch": 3.1316201528967484,
-      "grad_norm": 3.1306989192962646,
-      "learning_rate": 2.663613821788514e-05,
-      "loss": 0.6295,
+      "epoch": 1.5658157765589538,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.663432032692951e-05,
+      "loss": 0.9733502960205078,
       "step": 136000
     },
     {
-      "epoch": 3.1362254766510085,
-      "grad_norm": 2.744112968444824,
-      "learning_rate": 2.6628866654062625e-05,
-      "loss": 0.6328,
+      "epoch": 1.5681184616336687,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.6627048763106994e-05,
+      "loss": 0.9950527954101562,
       "step": 136200
     },
     {
-      "epoch": 3.1408308004052685,
-      "grad_norm": 3.3055408000946045,
-      "learning_rate": 2.6621595090240108e-05,
-      "loss": 0.649,
+      "epoch": 1.5704211467083837,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.661977719928448e-05,
+      "loss": 0.9778156280517578,
       "step": 136400
     },
     {
-      "epoch": 3.1454361241595286,
-      "grad_norm": 2.8039541244506836,
-      "learning_rate": 2.6614323526417593e-05,
-      "loss": 0.6382,
+      "epoch": 1.5727238317830987,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.6612505635461966e-05,
+      "loss": 0.9937000274658203,
       "step": 136600
     },
     {
-      "epoch": 3.150041447913788,
-      "grad_norm": 4.089864730834961,
-      "learning_rate": 2.6607051962595076e-05,
-      "loss": 0.642,
+      "epoch": 1.5750265168578135,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.6605234071639445e-05,
+      "loss": 0.9916617584228515,
       "step": 136800
     },
     {
-      "epoch": 3.154646771668048,
-      "grad_norm": 4.057522296905518,
-      "learning_rate": 2.659978039877256e-05,
-      "loss": 0.6349,
+      "epoch": 1.5773292019325285,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.659796250781693e-05,
+      "loss": 0.9857948303222657,
       "step": 137000
     },
     {
-      "epoch": 3.1592520954223082,
-      "grad_norm": 3.3105990886688232,
-      "learning_rate": 2.6592508834950044e-05,
-      "loss": 0.6389,
+      "epoch": 1.5796318870072432,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.6590690943994417e-05,
+      "loss": 0.9857929992675781,
       "step": 137200
     },
     {
-      "epoch": 3.1638574191765683,
-      "grad_norm": 3.4372615814208984,
-      "learning_rate": 2.6585273628946643e-05,
-      "loss": 0.6244,
+      "epoch": 1.5819345720819582,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.65834193801719e-05,
+      "loss": 0.9621226501464843,
       "step": 137400
     },
     {
-      "epoch": 3.168462742930828,
-      "grad_norm": 2.707204818725586,
-      "learning_rate": 2.6578002065124126e-05,
-      "loss": 0.6341,
+      "epoch": 1.5842372571566732,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.6576147816349385e-05,
+      "loss": 0.98962158203125,
       "step": 137600
     },
     {
-      "epoch": 3.173068066685088,
-      "grad_norm": 2.621185064315796,
-      "learning_rate": 2.657073050130161e-05,
-      "loss": 0.6464,
+      "epoch": 1.5865399422313882,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.656887625252687e-05,
+      "loss": 0.981746826171875,
       "step": 137800
     },
     {
-      "epoch": 3.177673390439348,
-      "grad_norm": 3.090057611465454,
-      "learning_rate": 2.6563458937479097e-05,
-      "loss": 0.6257,
+      "epoch": 1.5888426273061032,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.656160468870435e-05,
+      "loss": 0.982820053100586,
       "step": 138000
     },
     {
-      "epoch": 3.182278714193608,
-      "grad_norm": 2.8256921768188477,
-      "learning_rate": 2.655618737365658e-05,
-      "loss": 0.6349,
+      "epoch": 1.5911453123808181,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.6554333124881836e-05,
+      "loss": 0.9794469451904297,
       "step": 138200
     },
     {
-      "epoch": 3.1868840379478676,
-      "grad_norm": 4.003274917602539,
-      "learning_rate": 2.6548915809834062e-05,
-      "loss": 0.6321,
+      "epoch": 1.5934479974555331,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.6547061561059322e-05,
+      "loss": 0.9850052642822266,
       "step": 138400
     },
     {
-      "epoch": 3.1914893617021276,
-      "grad_norm": 3.346566677093506,
-      "learning_rate": 2.654164424601155e-05,
-      "loss": 0.636,
+      "epoch": 1.5957506825302479,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.6539789997236808e-05,
+      "loss": 0.9982978820800781,
       "step": 138600
     },
     {
-      "epoch": 3.1960946854563876,
-      "grad_norm": 3.386808156967163,
-      "learning_rate": 2.653437268218903e-05,
-      "loss": 0.6287,
+      "epoch": 1.5980533676049629,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.653251843341429e-05,
+      "loss": 0.9814372253417969,
       "step": 138800
     },
     {
-      "epoch": 3.2007000092106477,
-      "grad_norm": 2.970651149749756,
-      "learning_rate": 2.6527101118366517e-05,
-      "loss": 0.6299,
+      "epoch": 1.6003560526796778,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.6525246869591776e-05,
+      "loss": 0.9815452575683594,
       "step": 139000
     },
     {
-      "epoch": 3.2053053329649073,
-      "grad_norm": 3.1046876907348633,
-      "learning_rate": 2.6519865912363115e-05,
-      "loss": 0.6245,
+      "epoch": 1.6026587377543926,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.6517975305769262e-05,
+      "loss": 0.9756555938720703,
       "step": 139200
     },
     {
-      "epoch": 3.2099106567191673,
-      "grad_norm": 3.6521196365356445,
-      "learning_rate": 2.6512594348540598e-05,
-      "loss": 0.6315,
+      "epoch": 1.6049614228291076,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.651070374194674e-05,
+      "loss": 0.97823974609375,
       "step": 139400
     },
     {
-      "epoch": 3.2145159804734273,
-      "grad_norm": 2.994075298309326,
-      "learning_rate": 2.6505322784718084e-05,
-      "loss": 0.6459,
+      "epoch": 1.6072641079038226,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.6503432178124227e-05,
+      "loss": 0.9964341735839843,
       "step": 139600
     },
     {
-      "epoch": 3.2191213042276874,
-      "grad_norm": 3.732825517654419,
-      "learning_rate": 2.6498051220895566e-05,
-      "loss": 0.6445,
+      "epoch": 1.6095667929785376,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.6496160614301713e-05,
+      "loss": 0.9912190246582031,
       "step": 139800
     },
     {
-      "epoch": 3.223726627981947,
-      "grad_norm": 3.046517848968506,
-      "learning_rate": 2.649077965707305e-05,
-      "loss": 0.6352,
+      "epoch": 1.6118694780532525,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.6488889050479196e-05,
+      "loss": 0.9697745513916015,
       "step": 140000
     },
     {
-      "epoch": 3.228331951736207,
-      "grad_norm": 3.441150426864624,
-      "learning_rate": 2.6483508093250535e-05,
-      "loss": 0.6337,
+      "epoch": 1.6141721631279675,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.6481617486656682e-05,
+      "loss": 0.9726390838623047,
       "step": 140200
     },
     {
-      "epoch": 3.232937275490467,
-      "grad_norm": 4.240359306335449,
-      "learning_rate": 2.647623652942802e-05,
-      "loss": 0.6317,
+      "epoch": 1.6164748482026825,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6474345922834168e-05,
+      "loss": 0.9908740997314454,
       "step": 140400
     },
     {
-      "epoch": 3.237542599244727,
-      "grad_norm": 3.5375964641571045,
-      "learning_rate": 2.6468964965605503e-05,
-      "loss": 0.6309,
+      "epoch": 1.6187775332773973,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.6467074359011647e-05,
+      "loss": 0.9849369049072265,
       "step": 140600
     },
     {
-      "epoch": 3.2421479229989867,
-      "grad_norm": 3.648287773132324,
-      "learning_rate": 2.646169340178299e-05,
-      "loss": 0.6336,
+      "epoch": 1.6210802183521122,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.6459802795189133e-05,
+      "loss": 0.9715933227539062,
       "step": 140800
     },
     {
-      "epoch": 3.2467532467532467,
-      "grad_norm": 3.4129178524017334,
-      "learning_rate": 2.6454421837960475e-05,
-      "loss": 0.6226,
+      "epoch": 1.623382903426827,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.645253123136662e-05,
+      "loss": 0.9683143615722656,
       "step": 141000
     },
     {
-      "epoch": 3.2513585705075068,
-      "grad_norm": 3.6263647079467773,
-      "learning_rate": 2.6447150274137954e-05,
-      "loss": 0.6289,
+      "epoch": 1.625685588501542,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.64452596675441e-05,
+      "loss": 1.0068618774414062,
       "step": 141200
     },
     {
-      "epoch": 3.2559638942617664,
-      "grad_norm": 2.798513889312744,
-      "learning_rate": 2.643987871031544e-05,
-      "loss": 0.631,
+      "epoch": 1.627988273576257,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6437988103721587e-05,
+      "loss": 0.9583765411376953,
       "step": 141400
     },
     {
-      "epoch": 3.2605692180160264,
-      "grad_norm": 3.113395929336548,
-      "learning_rate": 2.6432607146492926e-05,
-      "loss": 0.6316,
+      "epoch": 1.630290958650972,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.6430716539899073e-05,
+      "loss": 0.9848359680175781,
       "step": 141600
     },
     {
-      "epoch": 3.2651745417702864,
-      "grad_norm": 2.998884916305542,
-      "learning_rate": 2.642533558267041e-05,
-      "loss": 0.6218,
+      "epoch": 1.632593643725687,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.642344497607656e-05,
+      "loss": 0.9773321533203125,
       "step": 141800
     },
     {
-      "epoch": 3.2697798655245465,
-      "grad_norm": 3.264549493789673,
-      "learning_rate": 2.6418064018847894e-05,
-      "loss": 0.6306,
+      "epoch": 1.634896328800402,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.6416173412254038e-05,
+      "loss": 0.9831598663330078,
       "step": 142000
     },
     {
-      "epoch": 3.2743851892788065,
-      "grad_norm": 2.9390909671783447,
-      "learning_rate": 2.641079245502538e-05,
-      "loss": 0.6357,
+      "epoch": 1.637199013875117,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.6408901848431524e-05,
+      "loss": 0.9870728302001953,
       "step": 142200
     },
     {
-      "epoch": 3.278990513033066,
-      "grad_norm": 2.7646069526672363,
-      "learning_rate": 2.6403520891202863e-05,
-      "loss": 0.6232,
+      "epoch": 1.6395016989498317,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.640163028460901e-05,
+      "loss": 0.9868105316162109,
       "step": 142400
     },
     {
-      "epoch": 3.283595836787326,
-      "grad_norm": 3.1988773345947266,
-      "learning_rate": 2.6396249327380345e-05,
-      "loss": 0.6246,
+      "epoch": 1.6418043840245466,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6394358720786492e-05,
+      "loss": 1.0004991912841796,
       "step": 142600
     },
     {
-      "epoch": 3.288201160541586,
-      "grad_norm": 2.912640333175659,
-      "learning_rate": 2.638897776355783e-05,
-      "loss": 0.6361,
+      "epoch": 1.6441070690992614,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.6387087156963978e-05,
+      "loss": 0.9994534301757813,
       "step": 142800
     },
     {
-      "epoch": 3.2928064842958458,
-      "grad_norm": 3.98209810256958,
-      "learning_rate": 2.6381706199735317e-05,
-      "loss": 0.6234,
+      "epoch": 1.6464097541739764,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.6379815593141464e-05,
+      "loss": 0.9692447662353516,
       "step": 143000
     },
     {
-      "epoch": 3.297411808050106,
-      "grad_norm": 3.3420772552490234,
-      "learning_rate": 2.63744346359128e-05,
-      "loss": 0.6344,
+      "epoch": 1.6487124392486914,
+      "grad_norm": 5.09375,
+      "learning_rate": 2.6372544029318943e-05,
+      "loss": 0.9961146545410157,
       "step": 143200
     },
     {
-      "epoch": 3.302017131804366,
-      "grad_norm": 4.18234395980835,
-      "learning_rate": 2.6367163072090286e-05,
-      "loss": 0.6178,
+      "epoch": 1.6510151243234064,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.636527246549643e-05,
+      "loss": 0.9730229187011719,
       "step": 143400
     },
     {
-      "epoch": 3.306622455558626,
-      "grad_norm": 2.9908995628356934,
-      "learning_rate": 2.6359891508267768e-05,
-      "loss": 0.6181,
+      "epoch": 1.6533178093981213,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.6358000901673915e-05,
+      "loss": 0.9857634735107422,
       "step": 143600
     },
     {
-      "epoch": 3.311227779312886,
-      "grad_norm": 3.504472255706787,
-      "learning_rate": 2.635261994444525e-05,
-      "loss": 0.6403,
+      "epoch": 1.6556204944728363,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.6350729337851398e-05,
+      "loss": 0.9707545471191407,
       "step": 143800
     },
     {
-      "epoch": 3.3158331030671455,
-      "grad_norm": 3.0107550621032715,
-      "learning_rate": 2.6345348380622737e-05,
-      "loss": 0.6338,
+      "epoch": 1.6579231795475513,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.6343457774028884e-05,
+      "loss": 0.9886442565917969,
       "step": 144000
     },
     {
-      "epoch": 3.3204384268214056,
-      "grad_norm": 3.5131990909576416,
-      "learning_rate": 2.6338076816800223e-05,
-      "loss": 0.6412,
+      "epoch": 1.660225864622266,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.633618621020637e-05,
+      "loss": 0.9728403472900391,
       "step": 144200
     },
     {
-      "epoch": 3.3250437505756656,
-      "grad_norm": 3.315256357192993,
-      "learning_rate": 2.6330805252977705e-05,
-      "loss": 0.6329,
+      "epoch": 1.662528549696981,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6328914646383852e-05,
+      "loss": 0.9904885864257813,
       "step": 144400
     },
     {
-      "epoch": 3.329649074329925,
-      "grad_norm": 3.0303072929382324,
-      "learning_rate": 2.632353368915519e-05,
-      "loss": 0.6312,
+      "epoch": 1.6648312347716958,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.6321643082561335e-05,
+      "loss": 0.9826412963867187,
       "step": 144600
     },
     {
-      "epoch": 3.3342543980841852,
-      "grad_norm": 3.3184735774993896,
-      "learning_rate": 2.6316262125332677e-05,
-      "loss": 0.647,
+      "epoch": 1.6671339198464108,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.631437151873882e-05,
+      "loss": 0.977677001953125,
       "step": 144800
     },
     {
-      "epoch": 3.3388597218384453,
-      "grad_norm": 2.842087745666504,
-      "learning_rate": 2.630899056151016e-05,
-      "loss": 0.622,
+      "epoch": 1.6694366049211258,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.6307099954916306e-05,
+      "loss": 0.9630680847167968,
       "step": 145000
     },
     {
-      "epoch": 3.3434650455927053,
-      "grad_norm": 2.741712808609009,
-      "learning_rate": 2.6301718997687642e-05,
-      "loss": 0.635,
+      "epoch": 1.6717392899958408,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.629982839109379e-05,
+      "loss": 0.9929466247558594,
       "step": 145200
     },
     {
-      "epoch": 3.3480703693469653,
-      "grad_norm": 2.9912500381469727,
-      "learning_rate": 2.6294447433865128e-05,
-      "loss": 0.6426,
+      "epoch": 1.6740419750705557,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.6292556827271275e-05,
+      "loss": 0.97507080078125,
       "step": 145400
     },
     {
-      "epoch": 3.352675693101225,
-      "grad_norm": 3.2431087493896484,
-      "learning_rate": 2.6287175870042614e-05,
-      "loss": 0.6314,
+      "epoch": 1.6763446601452707,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.628528526344876e-05,
+      "loss": 0.9925634765625,
       "step": 145600
     },
     {
-      "epoch": 3.357281016855485,
-      "grad_norm": 2.5913467407226562,
-      "learning_rate": 2.6279904306220096e-05,
-      "loss": 0.6282,
+      "epoch": 1.6786473452199857,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.627801369962624e-05,
+      "loss": 0.9717263031005859,
       "step": 145800
     },
     {
-      "epoch": 3.361886340609745,
-      "grad_norm": 2.8884477615356445,
-      "learning_rate": 2.6272632742397582e-05,
-      "loss": 0.6446,
+      "epoch": 1.6809500302947005,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6270742135803726e-05,
+      "loss": 0.9731719207763672,
       "step": 146000
     },
     {
-      "epoch": 3.3664916643640046,
-      "grad_norm": 3.9637069702148438,
-      "learning_rate": 2.6265361178575065e-05,
-      "loss": 0.6327,
+      "epoch": 1.6832527153694155,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6263470571981212e-05,
+      "loss": 0.9792316436767579,
       "step": 146200
     },
     {
-      "epoch": 3.3710969881182646,
-      "grad_norm": 3.9412338733673096,
-      "learning_rate": 2.6258089614752547e-05,
-      "loss": 0.6294,
+      "epoch": 1.6855554004441304,
+      "grad_norm": 0.5,
+      "learning_rate": 2.6256199008158694e-05,
+      "loss": 0.9754666900634765,
       "step": 146400
     },
     {
-      "epoch": 3.3757023118725247,
-      "grad_norm": 3.3022706508636475,
-      "learning_rate": 2.6250818050930033e-05,
-      "loss": 0.6326,
+      "epoch": 1.6878580855188452,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.624892744433618e-05,
+      "loss": 0.9769933319091797,
       "step": 146600
     },
     {
-      "epoch": 3.3803076356267847,
-      "grad_norm": 2.921814441680908,
-      "learning_rate": 2.624354648710752e-05,
-      "loss": 0.6312,
+      "epoch": 1.6901607705935602,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6241655880513666e-05,
+      "loss": 0.9838820648193359,
       "step": 146800
     },
     {
-      "epoch": 3.3849129593810443,
-      "grad_norm": 3.0949013233184814,
-      "learning_rate": 2.6236274923285e-05,
-      "loss": 0.6192,
+      "epoch": 1.6924634556682752,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6234384316691145e-05,
+      "loss": 0.9945056915283204,
       "step": 147000
     },
     {
-      "epoch": 3.3895182831353043,
-      "grad_norm": 4.004492282867432,
-      "learning_rate": 2.6229003359462488e-05,
-      "loss": 0.6298,
+      "epoch": 1.6947661407429901,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.622711275286863e-05,
+      "loss": 0.9864218139648437,
       "step": 147200
     },
     {
-      "epoch": 3.3941236068895644,
-      "grad_norm": 3.4825210571289062,
-      "learning_rate": 2.622173179563997e-05,
-      "loss": 0.6333,
+      "epoch": 1.6970688258177051,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.6219841189046117e-05,
+      "loss": 0.9742472839355468,
       "step": 147400
     },
     {
-      "epoch": 3.3987289306438244,
-      "grad_norm": 3.0541634559631348,
-      "learning_rate": 2.6214460231817453e-05,
-      "loss": 0.6347,
+      "epoch": 1.6993715108924201,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.6212569625223603e-05,
+      "loss": 0.9858525085449219,
       "step": 147600
     },
     {
-      "epoch": 3.403334254398084,
-      "grad_norm": 4.5528693199157715,
-      "learning_rate": 2.6207261383633164e-05,
-      "loss": 0.6381,
+      "epoch": 1.7016741959671349,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.6205298061401085e-05,
+      "loss": 0.9894911956787109,
       "step": 147800
     },
     {
-      "epoch": 3.407939578152344,
-      "grad_norm": 3.09186053276062,
-      "learning_rate": 2.6199989819810646e-05,
-      "loss": 0.6424,
+      "epoch": 1.7039768810418499,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.619802649757857e-05,
+      "loss": 0.9979306030273437,
       "step": 148000
     },
     {
-      "epoch": 3.412544901906604,
-      "grad_norm": 4.061122894287109,
-      "learning_rate": 2.6192718255988132e-05,
-      "loss": 0.6292,
+      "epoch": 1.7062795661165648,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6190754933756057e-05,
+      "loss": 0.9839297485351562,
       "step": 148200
     },
     {
-      "epoch": 3.417150225660864,
-      "grad_norm": 3.7221198081970215,
-      "learning_rate": 2.6185446692165618e-05,
-      "loss": 0.6446,
+      "epoch": 1.7085822511912796,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.6183483369933536e-05,
+      "loss": 0.9801649475097656,
       "step": 148400
     },
     {
-      "epoch": 3.4217555494151237,
-      "grad_norm": 2.8360812664031982,
-      "learning_rate": 2.6178175128343104e-05,
-      "loss": 0.6492,
+      "epoch": 1.7108849362659946,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6176211806111022e-05,
+      "loss": 0.9773529815673828,
       "step": 148600
     },
     {
-      "epoch": 3.4263608731693838,
-      "grad_norm": 3.033463954925537,
-      "learning_rate": 2.6170903564520587e-05,
-      "loss": 0.6273,
+      "epoch": 1.7131876213407096,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.6168940242288508e-05,
+      "loss": 0.9774404907226563,
       "step": 148800
     },
     {
-      "epoch": 3.430966196923644,
-      "grad_norm": 5.270578384399414,
-      "learning_rate": 2.6163632000698073e-05,
-      "loss": 0.6235,
+      "epoch": 1.7154903064154245,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.616166867846599e-05,
+      "loss": 0.9896322631835938,
       "step": 149000
     },
     {
-      "epoch": 3.435571520677904,
-      "grad_norm": 2.8622264862060547,
-      "learning_rate": 2.6156360436875555e-05,
-      "loss": 0.6325,
+      "epoch": 1.7177929914901395,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.6154397114643477e-05,
+      "loss": 0.9660794067382813,
       "step": 149200
     },
     {
-      "epoch": 3.4401768444321634,
-      "grad_norm": 3.450784206390381,
-      "learning_rate": 2.6149088873053038e-05,
-      "loss": 0.6373,
+      "epoch": 1.7200956765648545,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6147125550820963e-05,
+      "loss": 0.9710120391845704,
       "step": 149400
     },
     {
-      "epoch": 3.4447821681864235,
-      "grad_norm": 3.269690990447998,
-      "learning_rate": 2.6141817309230524e-05,
-      "loss": 0.6274,
+      "epoch": 1.7223983616395695,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.6139853986998442e-05,
+      "loss": 0.9865705871582031,
       "step": 149600
     },
     {
-      "epoch": 3.4493874919406835,
-      "grad_norm": 3.7646055221557617,
-      "learning_rate": 2.613454574540801e-05,
-      "loss": 0.6287,
+      "epoch": 1.7247010467142843,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.6132582423175928e-05,
+      "loss": 0.9635546875,
       "step": 149800
     },
     {
-      "epoch": 3.4539928156949435,
-      "grad_norm": 3.5802409648895264,
-      "learning_rate": 2.6127274181585492e-05,
-      "loss": 0.6322,
+      "epoch": 1.7270037317889992,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.6125310859353414e-05,
+      "loss": 0.9851200866699219,
       "step": 150000
     },
     {
-      "epoch": 3.458598139449203,
-      "grad_norm": 3.1698813438415527,
-      "learning_rate": 2.6120002617762978e-05,
-      "loss": 0.6339,
+      "epoch": 1.729306416863714,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.6118039295530896e-05,
+      "loss": 0.9768991088867187,
       "step": 150200
     },
     {
-      "epoch": 3.463203463203463,
-      "grad_norm": 2.7178969383239746,
-      "learning_rate": 2.611273105394046e-05,
-      "loss": 0.6382,
+      "epoch": 1.731609101938429,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6110767731708382e-05,
+      "loss": 0.9886154174804688,
       "step": 150400
     },
     {
-      "epoch": 3.467808786957723,
-      "grad_norm": 3.088505506515503,
-      "learning_rate": 2.6105459490117943e-05,
-      "loss": 0.6256,
+      "epoch": 1.733911787013144,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6103496167885868e-05,
+      "loss": 0.977115249633789,
       "step": 150600
     },
     {
-      "epoch": 3.4724141107119832,
-      "grad_norm": 2.7076618671417236,
-      "learning_rate": 2.609818792629543e-05,
-      "loss": 0.6312,
+      "epoch": 1.736214472087859,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.609622460406335e-05,
+      "loss": 0.9892723846435547,
       "step": 150800
     },
     {
-      "epoch": 3.477019434466243,
-      "grad_norm": 2.320732831954956,
-      "learning_rate": 2.6090916362472915e-05,
-      "loss": 0.6289,
+      "epoch": 1.738517157162574,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.6088953040240833e-05,
+      "loss": 1.0145347595214844,
       "step": 151000
     },
     {
-      "epoch": 3.481624758220503,
-      "grad_norm": 3.147030830383301,
-      "learning_rate": 2.6083644798650397e-05,
-      "loss": 0.641,
+      "epoch": 1.740819842237289,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.608168147641832e-05,
+      "loss": 0.9738581848144531,
       "step": 151200
     },
     {
-      "epoch": 3.486230081974763,
-      "grad_norm": 3.3665804862976074,
-      "learning_rate": 2.6076373234827883e-05,
-      "loss": 0.633,
+      "epoch": 1.743122527312004,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.6074409912595805e-05,
+      "loss": 0.9952706146240234,
       "step": 151400
     },
     {
-      "epoch": 3.4908354057290225,
-      "grad_norm": 3.705167531967163,
-      "learning_rate": 2.606910167100537e-05,
-      "loss": 0.6372,
+      "epoch": 1.7454252123867187,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.6067138348773287e-05,
+      "loss": 0.9953517913818359,
       "step": 151600
     },
     {
-      "epoch": 3.4954407294832825,
-      "grad_norm": 2.7983779907226562,
-      "learning_rate": 2.606183010718285e-05,
-      "loss": 0.6277,
+      "epoch": 1.7477278974614336,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.6059866784950773e-05,
+      "loss": 0.9769023895263672,
       "step": 151800
     },
     {
-      "epoch": 3.5000460532375426,
-      "grad_norm": 3.0177760124206543,
-      "learning_rate": 2.6054594901179447e-05,
-      "loss": 0.6277,
+      "epoch": 1.7500305825361484,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.605259522112826e-05,
+      "loss": 1.0000345611572266,
       "step": 152000
     },
     {
-      "epoch": 3.5046513769918026,
-      "grad_norm": 3.7722854614257812,
-      "learning_rate": 2.6047323337356933e-05,
-      "loss": 0.6278,
+      "epoch": 1.7523332676108634,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6045323657305738e-05,
+      "loss": 0.9734282684326172,
       "step": 152200
     },
     {
-      "epoch": 3.5092567007460627,
-      "grad_norm": 2.6710495948791504,
-      "learning_rate": 2.6040051773534415e-05,
-      "loss": 0.6202,
+      "epoch": 1.7546359526855784,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6038052093483224e-05,
+      "loss": 0.983162841796875,
       "step": 152400
     },
     {
-      "epoch": 3.5138620245003223,
-      "grad_norm": 3.0754504203796387,
-      "learning_rate": 2.60327802097119e-05,
-      "loss": 0.6317,
+      "epoch": 1.7569386377602934,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.603078052966071e-05,
+      "loss": 0.9792951202392578,
       "step": 152600
     },
     {
-      "epoch": 3.5184673482545823,
-      "grad_norm": 3.376631498336792,
-      "learning_rate": 2.60255450037085e-05,
-      "loss": 0.6422,
+      "epoch": 1.7592413228350083,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.6023508965838193e-05,
+      "loss": 0.9730397033691406,
       "step": 152800
     },
     {
-      "epoch": 3.5230726720088423,
-      "grad_norm": 2.970846652984619,
-      "learning_rate": 2.6018273439885982e-05,
-      "loss": 0.6315,
+      "epoch": 1.7615440079097233,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.601623740201568e-05,
+      "loss": 0.9625283050537109,
       "step": 153000
     },
     {
-      "epoch": 3.527677995763102,
-      "grad_norm": 2.7988736629486084,
-      "learning_rate": 2.6011001876063468e-05,
-      "loss": 0.6321,
+      "epoch": 1.7638466929844383,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.6008965838193164e-05,
+      "loss": 0.9787934112548828,
       "step": 153200
     },
     {
-      "epoch": 3.532283319517362,
-      "grad_norm": 2.94100022315979,
-      "learning_rate": 2.600373031224095e-05,
-      "loss": 0.6253,
+      "epoch": 1.766149378059153,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.6001694274370647e-05,
+      "loss": 0.9870194244384766,
       "step": 153400
     },
     {
-      "epoch": 3.536888643271622,
-      "grad_norm": 3.08990216255188,
-      "learning_rate": 2.5996458748418433e-05,
-      "loss": 0.6337,
+      "epoch": 1.768452063133868,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.599442271054813e-05,
+      "loss": 0.9825399017333984,
       "step": 153600
     },
     {
-      "epoch": 3.541493967025882,
-      "grad_norm": 2.9995460510253906,
-      "learning_rate": 2.598918718459592e-05,
-      "loss": 0.634,
+      "epoch": 1.7707547482085828,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.5987151146725615e-05,
+      "loss": 0.9877588653564453,
       "step": 153800
     },
     {
-      "epoch": 3.546099290780142,
-      "grad_norm": 3.2981879711151123,
-      "learning_rate": 2.5981915620773405e-05,
-      "loss": 0.6263,
+      "epoch": 1.7730574332832978,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.59798795829031e-05,
+      "loss": 0.981055679321289,
       "step": 154000
     },
     {
-      "epoch": 3.5507046145344017,
-      "grad_norm": 3.200422525405884,
-      "learning_rate": 2.5974644056950888e-05,
-      "loss": 0.6185,
+      "epoch": 1.7753601183580128,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.5972608019080584e-05,
+      "loss": 0.9929129028320313,
       "step": 154200
     },
     {
-      "epoch": 3.5553099382886617,
-      "grad_norm": 2.9688191413879395,
-      "learning_rate": 2.5967372493128374e-05,
-      "loss": 0.6213,
+      "epoch": 1.7776628034327278,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.596533645525807e-05,
+      "loss": 0.975859375,
       "step": 154400
     },
     {
-      "epoch": 3.5599152620429217,
-      "grad_norm": 2.5572378635406494,
-      "learning_rate": 2.5960100929305856e-05,
-      "loss": 0.6365,
+      "epoch": 1.7799654885074427,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.5958064891435552e-05,
+      "loss": 0.9847544860839844,
       "step": 154600
     },
     {
-      "epoch": 3.5645205857971813,
-      "grad_norm": 2.690509557723999,
-      "learning_rate": 2.5952829365483342e-05,
-      "loss": 0.6428,
+      "epoch": 1.7822681735821577,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5950793327613035e-05,
+      "loss": 0.9796549224853516,
       "step": 154800
     },
     {
-      "epoch": 3.5691259095514414,
-      "grad_norm": 5.786709785461426,
-      "learning_rate": 2.5945594159479937e-05,
-      "loss": 0.6448,
+      "epoch": 1.7845708586568727,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.594352176379052e-05,
+      "loss": 0.9731486511230468,
       "step": 155000
     },
     {
-      "epoch": 3.5737312333057014,
-      "grad_norm": 2.30430269241333,
-      "learning_rate": 2.5938322595657423e-05,
-      "loss": 0.6192,
+      "epoch": 1.7868735437315875,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.5936250199968007e-05,
+      "loss": 0.9948165893554688,
       "step": 155200
     },
     {
-      "epoch": 3.5783365570599615,
-      "grad_norm": 3.188671588897705,
-      "learning_rate": 2.5931051031834906e-05,
-      "loss": 0.6335,
+      "epoch": 1.7891762288063024,
+      "grad_norm": 0.625,
+      "learning_rate": 2.592897863614549e-05,
+      "loss": 0.9961344909667968,
       "step": 155400
     },
     {
-      "epoch": 3.5829418808142215,
-      "grad_norm": 2.945617198944092,
-      "learning_rate": 2.592377946801239e-05,
-      "loss": 0.6265,
+      "epoch": 1.7914789138810174,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.5921707072322975e-05,
+      "loss": 0.9734591674804688,
       "step": 155600
     },
     {
-      "epoch": 3.587547204568481,
-      "grad_norm": 3.1751980781555176,
-      "learning_rate": 2.5916507904189878e-05,
-      "loss": 0.6199,
+      "epoch": 1.7937815989557322,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.591443550850046e-05,
+      "loss": 0.956961669921875,
       "step": 155800
     },
     {
-      "epoch": 3.592152528322741,
-      "grad_norm": 2.6950361728668213,
-      "learning_rate": 2.590923634036736e-05,
-      "loss": 0.6283,
+      "epoch": 1.7960842840304472,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.590716394467794e-05,
+      "loss": 0.9658924865722657,
       "step": 156000
     },
     {
-      "epoch": 3.596757852077001,
-      "grad_norm": 3.544735908508301,
-      "learning_rate": 2.5901964776544843e-05,
-      "loss": 0.6281,
+      "epoch": 1.7983869691051622,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.5899892380855426e-05,
+      "loss": 0.9590577697753906,
       "step": 156200
     },
     {
-      "epoch": 3.6013631758312608,
-      "grad_norm": 2.9812283515930176,
-      "learning_rate": 2.589469321272233e-05,
-      "loss": 0.6275,
+      "epoch": 1.8006896541798771,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.5892620817032912e-05,
+      "loss": 1.0172974395751953,
       "step": 156400
     },
     {
-      "epoch": 3.605968499585521,
-      "grad_norm": 3.3013553619384766,
-      "learning_rate": 2.5887421648899814e-05,
-      "loss": 0.6292,
+      "epoch": 1.8029923392545921,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.5885349253210398e-05,
+      "loss": 0.9836770629882813,
       "step": 156600
     },
     {
-      "epoch": 3.610573823339781,
-      "grad_norm": 3.0218076705932617,
-      "learning_rate": 2.5880150085077297e-05,
-      "loss": 0.6279,
+      "epoch": 1.805295024329307,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.587807768938788e-05,
+      "loss": 0.990052719116211,
       "step": 156800
     },
     {
-      "epoch": 3.615179147094041,
-      "grad_norm": 2.9628562927246094,
-      "learning_rate": 2.5872878521254783e-05,
-      "loss": 0.6444,
+      "epoch": 1.807597709404022,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5870806125565366e-05,
+      "loss": 0.9714091491699218,
       "step": 157000
     },
     {
-      "epoch": 3.619784470848301,
-      "grad_norm": 3.080231189727783,
-      "learning_rate": 2.586560695743227e-05,
-      "loss": 0.6305,
+      "epoch": 1.8099003944787369,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.586353456174285e-05,
+      "loss": 0.9685219573974609,
       "step": 157200
     },
     {
-      "epoch": 3.6243897946025605,
-      "grad_norm": 3.3439502716064453,
-      "learning_rate": 2.5858335393609748e-05,
-      "loss": 0.6354,
+      "epoch": 1.8122030795534518,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.585626299792033e-05,
+      "loss": 0.9915869140625,
       "step": 157400
     },
     {
-      "epoch": 3.6289951183568205,
-      "grad_norm": 3.6231765747070312,
-      "learning_rate": 2.5851063829787234e-05,
-      "loss": 0.622,
+      "epoch": 1.8145057646281666,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.5848991434097817e-05,
+      "loss": 0.9873688507080078,
       "step": 157600
     },
     {
-      "epoch": 3.6336004421110806,
-      "grad_norm": 4.266796588897705,
-      "learning_rate": 2.584379226596472e-05,
-      "loss": 0.6385,
+      "epoch": 1.8168084497028816,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.5841719870275303e-05,
+      "loss": 0.9832802581787109,
       "step": 157800
     },
     {
-      "epoch": 3.63820576586534,
-      "grad_norm": 2.9517550468444824,
-      "learning_rate": 2.5836520702142202e-05,
-      "loss": 0.6319,
+      "epoch": 1.8191111347775966,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.5834448306452786e-05,
+      "loss": 0.9880329895019532,
       "step": 158000
     },
     {
-      "epoch": 3.6428110896196,
-      "grad_norm": 3.549234390258789,
-      "learning_rate": 2.5829249138319688e-05,
-      "loss": 0.6233,
+      "epoch": 1.8214138198523115,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.582717674263027e-05,
+      "loss": 0.9875634765625,
       "step": 158200
     },
     {
-      "epoch": 3.6474164133738602,
-      "grad_norm": 2.957629442214966,
-      "learning_rate": 2.5821977574497174e-05,
-      "loss": 0.628,
+      "epoch": 1.8237165049270265,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.5819905178807758e-05,
+      "loss": 0.9814795684814454,
       "step": 158400
     },
     {
-      "epoch": 3.65202173712812,
-      "grad_norm": 2.4832262992858887,
-      "learning_rate": 2.581474236849377e-05,
-      "loss": 0.6252,
+      "epoch": 1.8260191900017415,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.5812633614985237e-05,
+      "loss": 0.9618077850341797,
       "step": 158600
     },
     {
-      "epoch": 3.65662706088238,
-      "grad_norm": 3.6070032119750977,
-      "learning_rate": 2.5807470804671255e-05,
-      "loss": 0.6317,
+      "epoch": 1.8283218750764565,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.5805362051162723e-05,
+      "loss": 0.9954774475097656,
       "step": 158800
     },
     {
-      "epoch": 3.66123238463664,
-      "grad_norm": 3.4381484985351562,
-      "learning_rate": 2.5800199240848738e-05,
-      "loss": 0.6174,
+      "epoch": 1.8306245601511713,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.579809048734021e-05,
+      "loss": 0.992311782836914,
       "step": 159000
     },
     {
-      "epoch": 3.6658377083909,
-      "grad_norm": 2.8952486515045166,
-      "learning_rate": 2.579292767702622e-05,
-      "loss": 0.6334,
+      "epoch": 1.8329272452258862,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5790818923517694e-05,
+      "loss": 0.9816593933105469,
       "step": 159200
     },
     {
-      "epoch": 3.67044303214516,
-      "grad_norm": 3.704712390899658,
-      "learning_rate": 2.5785656113203706e-05,
-      "loss": 0.6371,
+      "epoch": 1.835229930300601,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5783547359695177e-05,
+      "loss": 0.9977922821044922,
       "step": 159400
     },
     {
-      "epoch": 3.6750483558994196,
-      "grad_norm": 3.430786609649658,
-      "learning_rate": 2.5778384549381192e-05,
-      "loss": 0.6369,
+      "epoch": 1.837532615375316,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5776275795872663e-05,
+      "loss": 0.9930438995361328,
       "step": 159600
     },
     {
-      "epoch": 3.6796536796536796,
-      "grad_norm": 3.201442003250122,
-      "learning_rate": 2.5771112985558675e-05,
-      "loss": 0.6274,
+      "epoch": 1.839835300450031,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5769004232050145e-05,
+      "loss": 0.9736163330078125,
       "step": 159800
     },
     {
-      "epoch": 3.6842590034079397,
-      "grad_norm": 3.125777244567871,
-      "learning_rate": 2.576384142173616e-05,
-      "loss": 0.6318,
+      "epoch": 1.842137985524746,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5761732668227628e-05,
+      "loss": 0.9892105865478515,
       "step": 160000
     },
     {
-      "epoch": 3.6888643271621993,
-      "grad_norm": 2.6248064041137695,
-      "learning_rate": 2.5756569857913643e-05,
-      "loss": 0.6344,
+      "epoch": 1.844440670599461,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5754461104405114e-05,
+      "loss": 0.9836799621582031,
       "step": 160200
     },
     {
-      "epoch": 3.6934696509164593,
-      "grad_norm": 3.5570108890533447,
-      "learning_rate": 2.5749298294091126e-05,
-      "loss": 0.6293,
+      "epoch": 1.846743355674176,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.57471895405826e-05,
+      "loss": 0.9932373809814453,
       "step": 160400
     },
     {
-      "epoch": 3.6980749746707193,
-      "grad_norm": 3.13261342048645,
-      "learning_rate": 2.5742063088087724e-05,
-      "loss": 0.634,
+      "epoch": 1.849046040748891,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.5739917976760082e-05,
+      "loss": 0.9650992584228516,
       "step": 160600
     },
     {
-      "epoch": 3.7026802984249794,
-      "grad_norm": 3.070614814758301,
-      "learning_rate": 2.573479152426521e-05,
-      "loss": 0.6255,
+      "epoch": 1.8513487258236057,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.5732646412937568e-05,
+      "loss": 0.979215087890625,
       "step": 160800
     },
     {
-      "epoch": 3.7072856221792394,
-      "grad_norm": 2.835911750793457,
-      "learning_rate": 2.5727519960442693e-05,
-      "loss": 0.6224,
+      "epoch": 1.8536514108983206,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.572537484911505e-05,
+      "loss": 0.9821718597412109,
       "step": 161000
     },
     {
-      "epoch": 3.711890945933499,
-      "grad_norm": 2.9913852214813232,
-      "learning_rate": 2.572024839662018e-05,
-      "loss": 0.6335,
+      "epoch": 1.8559540959730354,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.5718103285292533e-05,
+      "loss": 0.9929822540283203,
       "step": 161200
     },
     {
-      "epoch": 3.716496269687759,
-      "grad_norm": 3.0731775760650635,
-      "learning_rate": 2.5712976832797664e-05,
-      "loss": 0.6479,
+      "epoch": 1.8582567810477504,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.571083172147002e-05,
+      "loss": 0.9742303466796876,
       "step": 161400
     },
     {
-      "epoch": 3.721101593442019,
-      "grad_norm": 3.5844905376434326,
-      "learning_rate": 2.5705705268975144e-05,
-      "loss": 0.6201,
+      "epoch": 1.8605594661224654,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5703560157647505e-05,
+      "loss": 0.9975980377197265,
       "step": 161600
     },
     {
-      "epoch": 3.7257069171962787,
-      "grad_norm": 3.9543886184692383,
-      "learning_rate": 2.569843370515263e-05,
-      "loss": 0.6318,
+      "epoch": 1.8628621511971803,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5696288593824988e-05,
+      "loss": 0.9791291809082031,
       "step": 161800
     },
     {
-      "epoch": 3.7303122409505387,
-      "grad_norm": 2.6492578983306885,
-      "learning_rate": 2.5691162141330115e-05,
-      "loss": 0.6319,
+      "epoch": 1.8651648362718953,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.5689017030002474e-05,
+      "loss": 0.9771806335449219,
       "step": 162000
     },
     {
-      "epoch": 3.7349175647047987,
-      "grad_norm": 3.418893814086914,
-      "learning_rate": 2.5683890577507598e-05,
-      "loss": 0.637,
+      "epoch": 1.8674675213466103,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.568174546617996e-05,
+      "loss": 0.9976700592041016,
       "step": 162200
     },
     {
-      "epoch": 3.739522888459059,
-      "grad_norm": 2.8401198387145996,
-      "learning_rate": 2.5676619013685084e-05,
-      "loss": 0.6223,
+      "epoch": 1.8697702064213253,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.5674473902357442e-05,
+      "loss": 0.9705537414550781,
       "step": 162400
     },
     {
-      "epoch": 3.744128212213319,
-      "grad_norm": 3.19801664352417,
-      "learning_rate": 2.566934744986257e-05,
-      "loss": 0.6277,
+      "epoch": 1.87207289149604,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.5667202338534924e-05,
+      "loss": 1.0070448303222657,
       "step": 162600
     },
     {
-      "epoch": 3.7487335359675784,
-      "grad_norm": 2.9500930309295654,
-      "learning_rate": 2.5662075886040052e-05,
-      "loss": 0.6275,
+      "epoch": 1.874375576570755,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.565993077471241e-05,
+      "loss": 0.96590087890625,
       "step": 162800
     },
     {
-      "epoch": 3.7533388597218384,
-      "grad_norm": 2.7662055492401123,
-      "learning_rate": 2.5654804322217535e-05,
-      "loss": 0.619,
+      "epoch": 1.87667826164547,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.5652659210889896e-05,
+      "loss": 0.9896265411376953,
       "step": 163000
     },
     {
-      "epoch": 3.7579441834760985,
-      "grad_norm": 2.61580753326416,
-      "learning_rate": 2.564753275839502e-05,
-      "loss": 0.6305,
+      "epoch": 1.8789809467201848,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.564538764706738e-05,
+      "loss": 0.9813925170898438,
       "step": 163200
     },
     {
-      "epoch": 3.762549507230358,
-      "grad_norm": 3.7837417125701904,
-      "learning_rate": 2.5640261194572507e-05,
-      "loss": 0.6273,
+      "epoch": 1.8812836317948998,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.5638116083244865e-05,
+      "loss": 0.9730363464355469,
       "step": 163400
     },
     {
-      "epoch": 3.767154830984618,
-      "grad_norm": 3.516418695449829,
-      "learning_rate": 2.563298963074999e-05,
-      "loss": 0.638,
+      "epoch": 1.8835863168696148,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.5630844519422347e-05,
+      "loss": 0.9983138275146485,
       "step": 163600
     },
     {
-      "epoch": 3.771760154738878,
-      "grad_norm": 2.6570441722869873,
-      "learning_rate": 2.5625718066927475e-05,
-      "loss": 0.6319,
+      "epoch": 1.8858890019443297,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.562357295559983e-05,
+      "loss": 0.9798451232910156,
       "step": 163800
     },
     {
-      "epoch": 3.776365478493138,
-      "grad_norm": 2.759058952331543,
-      "learning_rate": 2.561844650310496e-05,
-      "loss": 0.6406,
+      "epoch": 1.8881916870190447,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.5616301391777316e-05,
+      "loss": 0.9604888916015625,
       "step": 164000
     },
     {
-      "epoch": 3.7809708022473982,
-      "grad_norm": 2.742692232131958,
-      "learning_rate": 2.561117493928244e-05,
-      "loss": 0.6278,
+      "epoch": 1.8904943720937597,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.56090298279548e-05,
+      "loss": 0.9828205871582031,
       "step": 164200
     },
     {
-      "epoch": 3.785576126001658,
-      "grad_norm": 3.729440212249756,
-      "learning_rate": 2.5603903375459926e-05,
-      "loss": 0.6289,
+      "epoch": 1.8927970571684745,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.5601758264132284e-05,
+      "loss": 0.9900843811035156,
       "step": 164400
     },
     {
-      "epoch": 3.790181449755918,
-      "grad_norm": 3.219895601272583,
-      "learning_rate": 2.5596631811637412e-05,
-      "loss": 0.6275,
+      "epoch": 1.8950997422431894,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.559448670030977e-05,
+      "loss": 0.9711660003662109,
       "step": 164600
     },
     {
-      "epoch": 3.794786773510178,
-      "grad_norm": 3.222836971282959,
-      "learning_rate": 2.5589360247814894e-05,
-      "loss": 0.6331,
+      "epoch": 1.8974024273179044,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.5587215136487253e-05,
+      "loss": 0.9848336791992187,
       "step": 164800
     },
     {
-      "epoch": 3.7993920972644375,
-      "grad_norm": 2.4223623275756836,
-      "learning_rate": 2.558208868399238e-05,
-      "loss": 0.623,
+      "epoch": 1.8997051123926192,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.557994357266474e-05,
+      "loss": 1.0130525207519532,
       "step": 165000
     },
     {
-      "epoch": 3.8039974210186975,
-      "grad_norm": 3.0702037811279297,
-      "learning_rate": 2.5574817120169866e-05,
-      "loss": 0.6285,
+      "epoch": 1.9020077974673342,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.557267200884222e-05,
+      "loss": 0.969343032836914,
       "step": 165200
     },
     {
-      "epoch": 3.8086027447729576,
-      "grad_norm": 2.8104872703552246,
-      "learning_rate": 2.556758191416646e-05,
-      "loss": 0.6353,
+      "epoch": 1.9043104825420492,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5565400445019707e-05,
+      "loss": 0.9901512908935547,
       "step": 165400
     },
     {
-      "epoch": 3.8132080685272176,
-      "grad_norm": 3.1389353275299072,
-      "learning_rate": 2.556034670816306e-05,
-      "loss": 0.6363,
+      "epoch": 1.9066131676167641,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.5558128881197193e-05,
+      "loss": 0.966179428100586,
       "step": 165600
     },
     {
-      "epoch": 3.8178133922814776,
-      "grad_norm": 3.318134307861328,
-      "learning_rate": 2.5553111502159655e-05,
-      "loss": 0.6259,
+      "epoch": 1.9089158526914791,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.5550857317374675e-05,
+      "loss": 0.983575439453125,
       "step": 165800
     },
     {
-      "epoch": 3.8224187160357372,
-      "grad_norm": 3.075765609741211,
-      "learning_rate": 2.554583993833714e-05,
-      "loss": 0.6325,
+      "epoch": 1.911218537766194,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.554358575355216e-05,
+      "loss": 0.9980619812011718,
       "step": 166000
     },
     {
-      "epoch": 3.8270240397899973,
-      "grad_norm": 3.2406067848205566,
-      "learning_rate": 2.5538568374514624e-05,
-      "loss": 0.6303,
+      "epoch": 1.913521222840909,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.5536314189729644e-05,
+      "loss": 0.9869134521484375,
       "step": 166200
     },
     {
-      "epoch": 3.8316293635442573,
-      "grad_norm": 3.4027702808380127,
-      "learning_rate": 2.5531296810692106e-05,
-      "loss": 0.634,
+      "epoch": 1.9158239079156238,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.5529042625907126e-05,
+      "loss": 0.9981081390380859,
       "step": 166400
     },
     {
-      "epoch": 3.836234687298517,
-      "grad_norm": 3.2992939949035645,
-      "learning_rate": 2.5524025246869592e-05,
-      "loss": 0.6285,
+      "epoch": 1.9181265929903388,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.5521771062084612e-05,
+      "loss": 0.9734851837158203,
       "step": 166600
     },
     {
-      "epoch": 3.840840011052777,
-      "grad_norm": 3.0388951301574707,
-      "learning_rate": 2.5516753683047078e-05,
-      "loss": 0.631,
+      "epoch": 1.9204292780650536,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.5514499498262098e-05,
+      "loss": 0.9968444061279297,
       "step": 166800
     },
     {
-      "epoch": 3.845445334807037,
-      "grad_norm": 3.0130348205566406,
-      "learning_rate": 2.550948211922456e-05,
-      "loss": 0.6239,
+      "epoch": 1.9227319631397686,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.550722793443958e-05,
+      "loss": 0.9921951293945312,
       "step": 167000
     },
     {
-      "epoch": 3.850050658561297,
-      "grad_norm": 3.0870108604431152,
-      "learning_rate": 2.5502210555402047e-05,
-      "loss": 0.6318,
+      "epoch": 1.9250346482144836,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.5499956370617067e-05,
+      "loss": 0.9820545959472656,
       "step": 167200
     },
     {
-      "epoch": 3.8546559823155566,
-      "grad_norm": 3.3640689849853516,
-      "learning_rate": 2.549493899157953e-05,
-      "loss": 0.6381,
+      "epoch": 1.9273373332891985,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.549268480679455e-05,
+      "loss": 0.9863339233398437,
       "step": 167400
     },
     {
-      "epoch": 3.8592613060698167,
-      "grad_norm": 3.3157622814178467,
-      "learning_rate": 2.548766742775701e-05,
-      "loss": 0.6208,
+      "epoch": 1.9296400183639135,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.548541324297203e-05,
+      "loss": 0.963076400756836,
       "step": 167600
     },
     {
-      "epoch": 3.8638666298240767,
-      "grad_norm": 3.0967700481414795,
-      "learning_rate": 2.5480395863934498e-05,
-      "loss": 0.6294,
+      "epoch": 1.9319427034386285,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.5478141679149518e-05,
+      "loss": 0.9745853424072266,
       "step": 167800
     },
     {
-      "epoch": 3.8684719535783367,
-      "grad_norm": 2.9279990196228027,
-      "learning_rate": 2.5473124300111983e-05,
-      "loss": 0.629,
+      "epoch": 1.9342453885133435,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5470870115327003e-05,
+      "loss": 0.9626667022705078,
       "step": 168000
     },
     {
-      "epoch": 3.8730772773325963,
-      "grad_norm": 3.182149648666382,
-      "learning_rate": 2.546585273628947e-05,
-      "loss": 0.62,
+      "epoch": 1.9365480735880582,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.546359855150449e-05,
+      "loss": 0.9721767425537109,
       "step": 168200
     },
     {
-      "epoch": 3.8776826010868564,
-      "grad_norm": 2.424578905105591,
-      "learning_rate": 2.5458581172466952e-05,
-      "loss": 0.6178,
+      "epoch": 1.9388507586627732,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5456326987681972e-05,
+      "loss": 0.9810872650146485,
       "step": 168400
     },
     {
-      "epoch": 3.8822879248411164,
-      "grad_norm": 3.4072725772857666,
-      "learning_rate": 2.5451309608644434e-05,
-      "loss": 0.6325,
+      "epoch": 1.941153443737488,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5449055423859458e-05,
+      "loss": 0.9774066162109375,
       "step": 168600
     },
     {
-      "epoch": 3.886893248595376,
-      "grad_norm": 3.33803653717041,
-      "learning_rate": 2.544407440264103e-05,
-      "loss": 0.6347,
+      "epoch": 1.943456128812203,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.544178386003694e-05,
+      "loss": 0.98538818359375,
       "step": 168800
     },
     {
-      "epoch": 3.891498572349636,
-      "grad_norm": 3.434117555618286,
-      "learning_rate": 2.5436802838818516e-05,
-      "loss": 0.6278,
+      "epoch": 1.945758813886918,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5434512296214423e-05,
+      "loss": 0.9620954895019531,
       "step": 169000
     },
     {
-      "epoch": 3.896103896103896,
-      "grad_norm": 3.191801071166992,
-      "learning_rate": 2.5429531274996e-05,
-      "loss": 0.6292,
+      "epoch": 1.948061498961633,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.542724073239191e-05,
+      "loss": 0.9863482666015625,
       "step": 169200
     },
     {
-      "epoch": 3.900709219858156,
-      "grad_norm": 3.666332483291626,
-      "learning_rate": 2.5422259711173484e-05,
-      "loss": 0.6279,
+      "epoch": 1.950364184036348,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5419969168569395e-05,
+      "loss": 0.9762551879882813,
       "step": 169400
     },
     {
-      "epoch": 3.905314543612416,
-      "grad_norm": 4.157709121704102,
-      "learning_rate": 2.541498814735097e-05,
-      "loss": 0.6244,
+      "epoch": 1.952666869111063,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5412697604746877e-05,
+      "loss": 0.966393051147461,
       "step": 169600
     },
     {
-      "epoch": 3.9099198673666757,
-      "grad_norm": 2.783106565475464,
-      "learning_rate": 2.5407716583528456e-05,
-      "loss": 0.6278,
+      "epoch": 1.954969554185778,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.5405426040924363e-05,
+      "loss": 1.0133824157714844,
       "step": 169800
     },
     {
-      "epoch": 3.9145251911209358,
-      "grad_norm": 3.1697680950164795,
-      "learning_rate": 2.5400445019705942e-05,
-      "loss": 0.6312,
+      "epoch": 1.9572722392604927,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.5398154477101846e-05,
+      "loss": 0.9961969757080078,
       "step": 170000
     },
     {
-      "epoch": 3.919130514875196,
-      "grad_norm": 2.8394782543182373,
-      "learning_rate": 2.539317345588342e-05,
-      "loss": 0.6282,
+      "epoch": 1.9595749243352076,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.5390882913279328e-05,
+      "loss": 0.9705589294433594,
       "step": 170200
     },
     {
-      "epoch": 3.9237358386294554,
-      "grad_norm": 2.653749704360962,
-      "learning_rate": 2.5385901892060907e-05,
-      "loss": 0.6185,
+      "epoch": 1.9618776094099224,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.5383611349456814e-05,
+      "loss": 0.9754343414306641,
       "step": 170400
     },
     {
-      "epoch": 3.9283411623837154,
-      "grad_norm": 2.6118738651275635,
-      "learning_rate": 2.5378630328238393e-05,
-      "loss": 0.6259,
+      "epoch": 1.9641802944846374,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.53763397856343e-05,
+      "loss": 1.0036404418945313,
       "step": 170600
     },
     {
-      "epoch": 3.9329464861379755,
-      "grad_norm": 4.355154991149902,
-      "learning_rate": 2.5371358764415875e-05,
-      "loss": 0.6394,
+      "epoch": 1.9664829795593524,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.5369068221811783e-05,
+      "loss": 0.9673897552490235,
       "step": 170800
     },
     {
-      "epoch": 3.9375518098922355,
-      "grad_norm": 2.658700466156006,
-      "learning_rate": 2.536408720059336e-05,
-      "loss": 0.6277,
+      "epoch": 1.9687856646340673,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.536179665798927e-05,
+      "loss": 0.9643704223632813,
       "step": 171000
     },
     {
-      "epoch": 3.9421571336464956,
-      "grad_norm": 3.9519739151000977,
-      "learning_rate": 2.5356815636770847e-05,
-      "loss": 0.6268,
+      "epoch": 1.9710883497087823,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.535452509416675e-05,
+      "loss": 0.9760980987548828,
       "step": 171200
     },
     {
-      "epoch": 3.946762457400755,
-      "grad_norm": 2.6966614723205566,
-      "learning_rate": 2.5349544072948326e-05,
-      "loss": 0.6368,
+      "epoch": 1.9733910347834973,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.5347253530344237e-05,
+      "loss": 1.0071582794189453,
       "step": 171400
     },
     {
-      "epoch": 3.951367781155015,
-      "grad_norm": 2.29526686668396,
-      "learning_rate": 2.5342272509125812e-05,
-      "loss": 0.6327,
+      "epoch": 1.9756937198582123,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.533998196652172e-05,
+      "loss": 0.9611239624023438,
       "step": 171600
     },
     {
-      "epoch": 3.9559731049092752,
-      "grad_norm": 2.904367208480835,
-      "learning_rate": 2.5335000945303298e-05,
-      "loss": 0.6201,
+      "epoch": 1.977996404932927,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5332710402699205e-05,
+      "loss": 1.0153284454345703,
       "step": 171800
     },
     {
-      "epoch": 3.960578428663535,
-      "grad_norm": 3.2330174446105957,
-      "learning_rate": 2.532772938148078e-05,
-      "loss": 0.6297,
+      "epoch": 1.980299090007642,
+      "grad_norm": 1.5,
+      "learning_rate": 2.532543883887669e-05,
+      "loss": 0.9687615966796875,
       "step": 172000
     },
     {
-      "epoch": 3.965183752417795,
-      "grad_norm": 2.654149293899536,
-      "learning_rate": 2.5320457817658266e-05,
-      "loss": 0.6334,
+      "epoch": 1.982601775082357,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.5318167275054174e-05,
+      "loss": 0.9706464385986329,
       "step": 172200
     },
     {
-      "epoch": 3.969789076172055,
-      "grad_norm": 2.883559465408325,
-      "learning_rate": 2.5313186253835752e-05,
-      "loss": 0.6175,
+      "epoch": 1.9849044601570718,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.531089571123166e-05,
+      "loss": 0.9883045959472656,
       "step": 172400
     },
     {
-      "epoch": 3.974394399926315,
-      "grad_norm": 2.8193156719207764,
-      "learning_rate": 2.5305914690013235e-05,
-      "loss": 0.6293,
+      "epoch": 1.9872071452317868,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.5303624147409142e-05,
+      "loss": 1.0035950469970703,
       "step": 172600
     },
     {
-      "epoch": 3.978999723680575,
-      "grad_norm": 3.1371946334838867,
-      "learning_rate": 2.5298643126190717e-05,
-      "loss": 0.6256,
+      "epoch": 1.9895098303065017,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.5296352583586625e-05,
+      "loss": 0.9956815338134766,
       "step": 172800
     },
     {
-      "epoch": 3.9836050474348346,
-      "grad_norm": 3.378344774246216,
-      "learning_rate": 2.5291371562368203e-05,
-      "loss": 0.6241,
+      "epoch": 1.9918125153812167,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.528908101976411e-05,
+      "loss": 0.9859121704101562,
       "step": 173000
     },
     {
-      "epoch": 3.9882103711890946,
-      "grad_norm": 3.4124484062194824,
-      "learning_rate": 2.52841363563648e-05,
-      "loss": 0.6255,
+      "epoch": 1.9941152004559317,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.5281809455941597e-05,
+      "loss": 0.9893330383300781,
       "step": 173200
     },
     {
-      "epoch": 3.9928156949433546,
-      "grad_norm": 2.8864808082580566,
-      "learning_rate": 2.5276901150361397e-05,
-      "loss": 0.6253,
+      "epoch": 1.9964178855306467,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.527453789211908e-05,
+      "loss": 0.9664070892333985,
       "step": 173400
     },
     {
-      "epoch": 3.9974210186976142,
-      "grad_norm": 2.594573736190796,
-      "learning_rate": 2.5269629586538883e-05,
-      "loss": 0.6265,
+      "epoch": 1.9987205706053617,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5267266328296565e-05,
+      "loss": 0.9660589599609375,
       "step": 173600
     },
     {
-      "epoch": 4.0,
-      "eval_loss": 0.5913488864898682,
-      "eval_runtime": 146.1973,
-      "eval_samples_per_second": 193.991,
-      "eval_steps_per_second": 12.127,
+      "epoch": 2.0,
+      "eval_loss": NaN,
+      "eval_runtime": 429.7683,
+      "eval_samples_per_second": 65.991,
+      "eval_steps_per_second": 32.997,
       "step": 173712
     },
     {
-      "epoch": 4.002026342451875,
-      "grad_norm": 4.0110249519348145,
-      "learning_rate": 2.5262358022716366e-05,
-      "loss": 0.6277,
+      "epoch": 2.0010131814328744,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5259994764474048e-05,
+      "loss": 0.9948661804199219,
       "step": 173800
     },
     {
-      "epoch": 4.006631666206134,
-      "grad_norm": 3.5719571113586426,
-      "learning_rate": 2.525508645889385e-05,
-      "loss": 0.6222,
+      "epoch": 2.0033158665075894,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.5252723200651533e-05,
+      "loss": 0.974677505493164,
       "step": 174000
     },
     {
-      "epoch": 4.011236989960394,
-      "grad_norm": 3.5779929161071777,
-      "learning_rate": 2.5247814895071337e-05,
-      "loss": 0.6147,
+      "epoch": 2.0056185515823044,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.5245451636829016e-05,
+      "loss": 0.9709995269775391,
       "step": 174200
     },
     {
-      "epoch": 4.015842313714654,
-      "grad_norm": 2.7468457221984863,
-      "learning_rate": 2.5240543331248817e-05,
-      "loss": 0.6193,
+      "epoch": 2.0079212366570194,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.5238180073006502e-05,
+      "loss": 0.9784161376953125,
       "step": 174400
     },
     {
-      "epoch": 4.020447637468914,
-      "grad_norm": 3.6863133907318115,
-      "learning_rate": 2.5233271767426302e-05,
-      "loss": 0.627,
+      "epoch": 2.0102239217317344,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5230908509183988e-05,
+      "loss": 0.9698686218261718,
       "step": 174600
     },
     {
-      "epoch": 4.025052961223174,
-      "grad_norm": 3.8438150882720947,
-      "learning_rate": 2.522600020360379e-05,
-      "loss": 0.6304,
+      "epoch": 2.0125266068064493,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.522363694536147e-05,
+      "loss": 0.9923031616210938,
       "step": 174800
     },
     {
-      "epoch": 4.029658284977434,
-      "grad_norm": 3.4153921604156494,
-      "learning_rate": 2.521872863978127e-05,
-      "loss": 0.6209,
+      "epoch": 2.0148292918811643,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.5216365381538953e-05,
+      "loss": 0.9968160247802734,
       "step": 175000
     },
     {
-      "epoch": 4.034263608731694,
-      "grad_norm": 3.0748400688171387,
-      "learning_rate": 2.5211457075958757e-05,
-      "loss": 0.6252,
+      "epoch": 2.0171319769558793,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.520909381771644e-05,
+      "loss": 0.986888656616211,
       "step": 175200
     },
     {
-      "epoch": 4.038868932485954,
-      "grad_norm": 4.052780628204346,
-      "learning_rate": 2.5204185512136243e-05,
-      "loss": 0.6123,
+      "epoch": 2.019434662030594,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.520182225389392e-05,
+      "loss": 0.9929759979248047,
       "step": 175400
     },
     {
-      "epoch": 4.043474256240214,
-      "grad_norm": 3.1808793544769287,
-      "learning_rate": 2.5196913948313722e-05,
-      "loss": 0.6148,
+      "epoch": 2.021737347105309,
+      "grad_norm": 0.75,
+      "learning_rate": 2.5194550690071407e-05,
+      "loss": 0.9717521667480469,
       "step": 175600
     },
     {
-      "epoch": 4.048079579994473,
-      "grad_norm": 2.881800413131714,
-      "learning_rate": 2.5189642384491208e-05,
-      "loss": 0.6156,
+      "epoch": 2.024040032180024,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.5187279126248893e-05,
+      "loss": 0.9931996917724609,
       "step": 175800
     },
     {
-      "epoch": 4.052684903748734,
-      "grad_norm": 2.8981564044952393,
-      "learning_rate": 2.5182370820668694e-05,
-      "loss": 0.6211,
+      "epoch": 2.026342717254739,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5180007562426376e-05,
+      "loss": 0.9702210998535157,
       "step": 176000
     },
     {
-      "epoch": 4.057290227502993,
-      "grad_norm": 3.5495710372924805,
-      "learning_rate": 2.517509925684618e-05,
-      "loss": 0.617,
+      "epoch": 2.028645402329454,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.517273599860386e-05,
+      "loss": 0.9971956634521484,
       "step": 176200
     },
     {
-      "epoch": 4.061895551257253,
-      "grad_norm": 2.4835290908813477,
-      "learning_rate": 2.5167827693023662e-05,
-      "loss": 0.6297,
+      "epoch": 2.0309480874041688,
+      "grad_norm": 0.625,
+      "learning_rate": 2.5165464434781344e-05,
+      "loss": 0.971846923828125,
       "step": 176400
     },
     {
-      "epoch": 4.0665008750115135,
-      "grad_norm": 2.574183225631714,
-      "learning_rate": 2.5160556129201148e-05,
-      "loss": 0.6176,
+      "epoch": 2.0332507724788837,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.5158192870958827e-05,
+      "loss": 0.9721129608154296,
       "step": 176600
     },
     {
-      "epoch": 4.071106198765773,
-      "grad_norm": 3.0550966262817383,
-      "learning_rate": 2.5153284565378634e-05,
-      "loss": 0.6242,
+      "epoch": 2.0355534575535987,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.5150921307136313e-05,
+      "loss": 0.9739250183105469,
       "step": 176800
     },
     {
-      "epoch": 4.0757115225200335,
-      "grad_norm": 3.120473861694336,
-      "learning_rate": 2.5146013001556113e-05,
-      "loss": 0.6201,
+      "epoch": 2.0378561426283137,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.51436497433138e-05,
+      "loss": 0.9803957366943359,
       "step": 177000
     },
     {
-      "epoch": 4.080316846274293,
-      "grad_norm": 2.9657979011535645,
-      "learning_rate": 2.51387414377336e-05,
-      "loss": 0.6219,
+      "epoch": 2.0401588277030283,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.5136378179491284e-05,
+      "loss": 0.9933158874511718,
       "step": 177200
     },
     {
-      "epoch": 4.084922170028553,
-      "grad_norm": 2.9161338806152344,
-      "learning_rate": 2.5131469873911085e-05,
-      "loss": 0.623,
+      "epoch": 2.0424615127777432,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5129106615668767e-05,
+      "loss": 1.0045361328125,
       "step": 177400
     },
     {
-      "epoch": 4.089527493782813,
-      "grad_norm": 2.9412291049957275,
-      "learning_rate": 2.5124198310088567e-05,
-      "loss": 0.6353,
+      "epoch": 2.044764197852458,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.512183505184625e-05,
+      "loss": 0.9811223602294922,
       "step": 177600
     },
     {
-      "epoch": 4.094132817537073,
-      "grad_norm": 3.1718437671661377,
-      "learning_rate": 2.5116926746266053e-05,
-      "loss": 0.6205,
+      "epoch": 2.047066882927173,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.5114563488023735e-05,
+      "loss": 0.9679241943359375,
       "step": 177800
     },
     {
-      "epoch": 4.098738141291332,
-      "grad_norm": 4.523801326751709,
-      "learning_rate": 2.510965518244354e-05,
-      "loss": 0.6243,
+      "epoch": 2.049369568001888,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.5107291924201218e-05,
+      "loss": 0.9778891754150391,
       "step": 178000
     },
     {
-      "epoch": 4.103343465045593,
-      "grad_norm": 2.9366707801818848,
-      "learning_rate": 2.510238361862102e-05,
-      "loss": 0.619,
+      "epoch": 2.051672253076603,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.5100020360378704e-05,
+      "loss": 0.9766893768310547,
       "step": 178200
     },
     {
-      "epoch": 4.1079487887998525,
-      "grad_norm": 2.822934865951538,
-      "learning_rate": 2.5095148412617617e-05,
-      "loss": 0.6183,
+      "epoch": 2.053974938151318,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.509274879655619e-05,
+      "loss": 0.974738998413086,
       "step": 178400
     },
     {
-      "epoch": 4.112554112554113,
-      "grad_norm": 4.115916728973389,
-      "learning_rate": 2.5087876848795103e-05,
-      "loss": 0.6153,
+      "epoch": 2.056277623226033,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.5085477232733672e-05,
+      "loss": 0.9749568176269531,
       "step": 178600
     },
     {
-      "epoch": 4.1171594363083726,
-      "grad_norm": 3.5023276805877686,
-      "learning_rate": 2.5080605284972585e-05,
-      "loss": 0.6262,
+      "epoch": 2.058580308300748,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.5078205668911158e-05,
+      "loss": 0.9782068634033203,
       "step": 178800
     },
     {
-      "epoch": 4.121764760062632,
-      "grad_norm": 2.906297206878662,
-      "learning_rate": 2.507333372115007e-05,
-      "loss": 0.6307,
+      "epoch": 2.0608829933754627,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.507093410508864e-05,
+      "loss": 0.9693759155273437,
       "step": 179000
     },
     {
-      "epoch": 4.126370083816893,
-      "grad_norm": 2.6169967651367188,
-      "learning_rate": 2.5066062157327557e-05,
-      "loss": 0.625,
+      "epoch": 2.0631856784501776,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.5063662541266123e-05,
+      "loss": 0.9943150329589844,
       "step": 179200
     },
     {
-      "epoch": 4.130975407571152,
-      "grad_norm": 3.4697892665863037,
-      "learning_rate": 2.505879059350504e-05,
-      "loss": 0.631,
+      "epoch": 2.0654883635248926,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.505639097744361e-05,
+      "loss": 0.9823705291748047,
       "step": 179400
     },
     {
-      "epoch": 4.135580731325412,
-      "grad_norm": 3.589235305786133,
-      "learning_rate": 2.5051519029682522e-05,
-      "loss": 0.6238,
+      "epoch": 2.0677910485996076,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5049119413621095e-05,
+      "loss": 0.989997787475586,
       "step": 179600
     },
     {
-      "epoch": 4.140186055079672,
-      "grad_norm": 3.2380635738372803,
-      "learning_rate": 2.5044247465860008e-05,
-      "loss": 0.6222,
+      "epoch": 2.0700937336743226,
+      "grad_norm": 1.25,
+      "learning_rate": 2.504184784979858e-05,
+      "loss": 0.9868365478515625,
       "step": 179800
     },
     {
-      "epoch": 4.144791378833932,
-      "grad_norm": 2.568429946899414,
-      "learning_rate": 2.503697590203749e-05,
-      "loss": 0.6218,
+      "epoch": 2.0723964187490376,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5034576285976063e-05,
+      "loss": 0.981561050415039,
       "step": 180000
     },
     {
-      "epoch": 4.149396702588192,
-      "grad_norm": 3.3458428382873535,
-      "learning_rate": 2.5029704338214977e-05,
-      "loss": 0.6222,
+      "epoch": 2.0746991038237526,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5027304722153546e-05,
+      "loss": 0.9862503814697265,
       "step": 180200
     },
     {
-      "epoch": 4.154002026342452,
-      "grad_norm": 3.1617233753204346,
-      "learning_rate": 2.5022432774392463e-05,
-      "loss": 0.6284,
+      "epoch": 2.0770017888984675,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5020033158331032e-05,
+      "loss": 0.9771318054199218,
       "step": 180400
     },
     {
-      "epoch": 4.158607350096712,
-      "grad_norm": 2.6967251300811768,
-      "learning_rate": 2.5015197568389058e-05,
-      "loss": 0.6211,
+      "epoch": 2.0793044739731825,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5012761594508514e-05,
+      "loss": 0.9810347747802735,
       "step": 180600
     },
     {
-      "epoch": 4.163212673850972,
-      "grad_norm": 4.2213568687438965,
-      "learning_rate": 2.5007926004566544e-05,
-      "loss": 0.6242,
+      "epoch": 2.0816071590478975,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.5005490030686e-05,
+      "loss": 0.9873623657226562,
       "step": 180800
     },
     {
-      "epoch": 4.167817997605232,
-      "grad_norm": 3.8117053508758545,
-      "learning_rate": 2.500065444074403e-05,
-      "loss": 0.6182,
+      "epoch": 2.083909844122612,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4998218466863486e-05,
+      "loss": 0.9683272552490234,
       "step": 181000
     },
     {
-      "epoch": 4.172423321359491,
-      "grad_norm": 3.316926956176758,
-      "learning_rate": 2.499338287692151e-05,
-      "loss": 0.6174,
+      "epoch": 2.086212529197327,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.499094690304097e-05,
+      "loss": 0.9873902130126954,
       "step": 181200
     },
     {
-      "epoch": 4.177028645113752,
-      "grad_norm": 3.363097667694092,
-      "learning_rate": 2.4986111313098995e-05,
-      "loss": 0.6207,
+      "epoch": 2.088515214272042,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.498367533921845e-05,
+      "loss": 0.966907958984375,
       "step": 181400
     },
     {
-      "epoch": 4.181633968868011,
-      "grad_norm": 3.064410924911499,
-      "learning_rate": 2.497883974927648e-05,
-      "loss": 0.6218,
+      "epoch": 2.090817899346757,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.4976403775395937e-05,
+      "loss": 0.9876545715332031,
       "step": 181600
     },
     {
-      "epoch": 4.186239292622272,
-      "grad_norm": 2.6976685523986816,
-      "learning_rate": 2.4971568185453963e-05,
-      "loss": 0.6128,
+      "epoch": 2.093120584421472,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.496913221157342e-05,
+      "loss": 0.9763893890380859,
       "step": 181800
     },
     {
-      "epoch": 4.190844616376531,
-      "grad_norm": 3.532400608062744,
-      "learning_rate": 2.496429662163145e-05,
-      "loss": 0.6185,
+      "epoch": 2.095423269496187,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.4961860647750906e-05,
+      "loss": 0.9983731842041016,
       "step": 182000
     },
     {
-      "epoch": 4.195449940130791,
-      "grad_norm": 2.791353702545166,
-      "learning_rate": 2.4957025057808935e-05,
-      "loss": 0.6279,
+      "epoch": 2.097725954570902,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.495458908392839e-05,
+      "loss": 0.978978271484375,
       "step": 182200
     },
     {
-      "epoch": 4.2000552638850515,
-      "grad_norm": 2.433746814727783,
-      "learning_rate": 2.4949753493986417e-05,
-      "loss": 0.6216,
+      "epoch": 2.100028639645617,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.4947317520105874e-05,
+      "loss": 0.965992431640625,
       "step": 182400
     },
     {
-      "epoch": 4.204660587639311,
-      "grad_norm": 3.122479200363159,
-      "learning_rate": 2.49424819301639e-05,
-      "loss": 0.6206,
+      "epoch": 2.102331324720332,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.494004595628336e-05,
+      "loss": 0.9818271636962891,
       "step": 182600
     },
     {
-      "epoch": 4.209265911393571,
-      "grad_norm": 3.7672858238220215,
-      "learning_rate": 2.4935210366341386e-05,
-      "loss": 0.6327,
+      "epoch": 2.1046340097950464,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4932774392460842e-05,
+      "loss": 0.9539933013916015,
       "step": 182800
     },
     {
-      "epoch": 4.213871235147831,
-      "grad_norm": 2.832796812057495,
-      "learning_rate": 2.4927938802518872e-05,
-      "loss": 0.6363,
+      "epoch": 2.1069366948697614,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.492550282863833e-05,
+      "loss": 1.0096530151367187,
       "step": 183000
     },
     {
-      "epoch": 4.218476558902091,
-      "grad_norm": 2.60891056060791,
-      "learning_rate": 2.4920667238696354e-05,
-      "loss": 0.6279,
+      "epoch": 2.1092393799444764,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.491823126481581e-05,
+      "loss": 0.9892758178710938,
       "step": 183200
     },
     {
-      "epoch": 4.22308188265635,
-      "grad_norm": 2.9585559368133545,
-      "learning_rate": 2.491339567487384e-05,
-      "loss": 0.6306,
+      "epoch": 2.1115420650191914,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.4910959700993297e-05,
+      "loss": 0.9912564086914063,
       "step": 183400
     },
     {
-      "epoch": 4.227687206410611,
-      "grad_norm": 2.4493677616119385,
-      "learning_rate": 2.4906124111051326e-05,
-      "loss": 0.6151,
+      "epoch": 2.1138447500939064,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4903688137170783e-05,
+      "loss": 0.9664544677734375,
       "step": 183600
     },
     {
-      "epoch": 4.23229253016487,
-      "grad_norm": 3.008601188659668,
-      "learning_rate": 2.4898852547228805e-05,
-      "loss": 0.61,
+      "epoch": 2.1161474351686214,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.4896416573348265e-05,
+      "loss": 0.9713141632080078,
       "step": 183800
     },
     {
-      "epoch": 4.236897853919131,
-      "grad_norm": 3.0694305896759033,
-      "learning_rate": 2.489158098340629e-05,
-      "loss": 0.6369,
+      "epoch": 2.1184501202433363,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4889145009525748e-05,
+      "loss": 0.9756626892089844,
       "step": 184000
     },
     {
-      "epoch": 4.2415031776733905,
-      "grad_norm": 3.306530714035034,
-      "learning_rate": 2.4884309419583777e-05,
-      "loss": 0.6271,
+      "epoch": 2.1207528053180513,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.4881873445703234e-05,
+      "loss": 0.97317626953125,
       "step": 184200
     },
     {
-      "epoch": 4.24610850142765,
-      "grad_norm": 3.441470146179199,
-      "learning_rate": 2.487703785576126e-05,
-      "loss": 0.6163,
+      "epoch": 2.1230554903927663,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.4874601881880716e-05,
+      "loss": 0.9787543487548828,
       "step": 184400
     },
     {
-      "epoch": 4.2507138251819105,
-      "grad_norm": 2.845263957977295,
-      "learning_rate": 2.4869766291938746e-05,
-      "loss": 0.6207,
+      "epoch": 2.125358175467481,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.4867330318058202e-05,
+      "loss": 0.9660845184326172,
       "step": 184600
     },
     {
-      "epoch": 4.25531914893617,
-      "grad_norm": 3.332775354385376,
-      "learning_rate": 2.4862531085935344e-05,
-      "loss": 0.6221,
+      "epoch": 2.127660860542196,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.4860058754235688e-05,
+      "loss": 0.9829474639892578,
       "step": 184800
     },
     {
-      "epoch": 4.25992447269043,
-      "grad_norm": 2.332599639892578,
-      "learning_rate": 2.4855259522112827e-05,
-      "loss": 0.6207,
+      "epoch": 2.129963545616911,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.485278719041317e-05,
+      "loss": 0.9686710357666015,
       "step": 185000
     },
     {
-      "epoch": 4.26452979644469,
-      "grad_norm": 4.441643238067627,
-      "learning_rate": 2.4848024316109425e-05,
-      "loss": 0.6276,
+      "epoch": 2.132266230691626,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.4845515626590653e-05,
+      "loss": 0.9558204650878906,
       "step": 185200
     },
     {
-      "epoch": 4.26913512019895,
-      "grad_norm": 3.2958297729492188,
-      "learning_rate": 2.4840752752286904e-05,
-      "loss": 0.6209,
+      "epoch": 2.134568915766341,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.483824406276814e-05,
+      "loss": 0.9875593566894532,
       "step": 185400
     },
     {
-      "epoch": 4.27374044395321,
-      "grad_norm": 2.6615874767303467,
-      "learning_rate": 2.483348118846439e-05,
-      "loss": 0.6227,
+      "epoch": 2.1368716008410558,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.4830972498945625e-05,
+      "loss": 1.0049014282226563,
       "step": 185600
     },
     {
-      "epoch": 4.27834576770747,
-      "grad_norm": 2.7402963638305664,
-      "learning_rate": 2.4826209624641876e-05,
-      "loss": 0.613,
+      "epoch": 2.1391742859157707,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.4823700935123107e-05,
+      "loss": 0.9857121276855468,
       "step": 185800
     },
     {
-      "epoch": 4.2829510914617295,
-      "grad_norm": 3.0777735710144043,
-      "learning_rate": 2.4818938060819362e-05,
-      "loss": 0.6105,
+      "epoch": 2.1414769709904857,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4816429371300593e-05,
+      "loss": 0.9936911010742188,
       "step": 186000
     },
     {
-      "epoch": 4.28755641521599,
-      "grad_norm": 3.108518123626709,
-      "learning_rate": 2.4811702854815957e-05,
-      "loss": 0.6155,
+      "epoch": 2.1437796560652007,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.480915780747808e-05,
+      "loss": 0.9867401885986328,
       "step": 186200
     },
     {
-      "epoch": 4.2921617389702496,
-      "grad_norm": 2.7684409618377686,
-      "learning_rate": 2.4804431290993443e-05,
-      "loss": 0.6132,
+      "epoch": 2.1460823411399153,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4801886243655562e-05,
+      "loss": 0.9839734649658203,
       "step": 186400
     },
     {
-      "epoch": 4.296767062724509,
-      "grad_norm": 3.1115176677703857,
-      "learning_rate": 2.4797159727170926e-05,
-      "loss": 0.6306,
+      "epoch": 2.1483850262146302,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4794614679833044e-05,
+      "loss": 0.9782057189941407,
       "step": 186600
     },
     {
-      "epoch": 4.30137238647877,
-      "grad_norm": 3.601163387298584,
-      "learning_rate": 2.478988816334841e-05,
-      "loss": 0.6142,
+      "epoch": 2.150687711289345,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.478734311601053e-05,
+      "loss": 0.977763671875,
       "step": 186800
     },
     {
-      "epoch": 4.305977710233029,
-      "grad_norm": 3.0134575366973877,
-      "learning_rate": 2.4782616599525894e-05,
-      "loss": 0.6248,
+      "epoch": 2.15299039636406,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4780071552188013e-05,
+      "loss": 0.9595011901855469,
       "step": 187000
     },
     {
-      "epoch": 4.31058303398729,
-      "grad_norm": 2.950279951095581,
-      "learning_rate": 2.4775345035703377e-05,
-      "loss": 0.6184,
+      "epoch": 2.155293081438775,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.47727999883655e-05,
+      "loss": 0.97533203125,
       "step": 187200
     },
     {
-      "epoch": 4.315188357741549,
-      "grad_norm": 2.8289947509765625,
-      "learning_rate": 2.4768073471880863e-05,
-      "loss": 0.6292,
+      "epoch": 2.15759576651349,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.4765528424542985e-05,
+      "loss": 0.9841778564453125,
       "step": 187400
     },
     {
-      "epoch": 4.319793681495809,
-      "grad_norm": 3.4812533855438232,
-      "learning_rate": 2.476080190805835e-05,
-      "loss": 0.6284,
+      "epoch": 2.159898451588205,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4758256860720467e-05,
+      "loss": 1.0073722839355468,
       "step": 187600
     },
     {
-      "epoch": 4.324399005250069,
-      "grad_norm": 4.80519437789917,
-      "learning_rate": 2.4753530344235835e-05,
-      "loss": 0.6278,
+      "epoch": 2.16220113666292,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.475098529689795e-05,
+      "loss": 0.9862171936035157,
       "step": 187800
     },
     {
-      "epoch": 4.329004329004329,
-      "grad_norm": 2.8514554500579834,
-      "learning_rate": 2.4746258780413317e-05,
-      "loss": 0.6139,
+      "epoch": 2.164503821737635,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4743713733075436e-05,
+      "loss": 0.9807298278808594,
       "step": 188000
     },
     {
-      "epoch": 4.333609652758589,
-      "grad_norm": 2.441417694091797,
-      "learning_rate": 2.47389872165908e-05,
-      "loss": 0.6142,
+      "epoch": 2.1668065068123497,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4736442169252918e-05,
+      "loss": 0.9974630737304687,
       "step": 188200
     },
     {
-      "epoch": 4.338214976512849,
-      "grad_norm": 3.1779348850250244,
-      "learning_rate": 2.4731715652768285e-05,
-      "loss": 0.6245,
+      "epoch": 2.1691091918870646,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.4729170605430404e-05,
+      "loss": 0.9857058715820313,
       "step": 188400
     },
     {
-      "epoch": 4.342820300267109,
-      "grad_norm": 3.2313997745513916,
-      "learning_rate": 2.4724444088945768e-05,
-      "loss": 0.631,
+      "epoch": 2.1714118769617796,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.472189904160789e-05,
+      "loss": 0.9878272247314454,
       "step": 188600
     },
     {
-      "epoch": 4.347425624021369,
-      "grad_norm": 3.2089149951934814,
-      "learning_rate": 2.4717172525123254e-05,
-      "loss": 0.6131,
+      "epoch": 2.1737145620364946,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.4714627477785376e-05,
+      "loss": 0.9819071960449218,
       "step": 188800
     },
     {
-      "epoch": 4.352030947775629,
-      "grad_norm": 3.3893136978149414,
-      "learning_rate": 2.470990096130074e-05,
-      "loss": 0.6144,
+      "epoch": 2.1760172471112096,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.470735591396286e-05,
+      "loss": 0.9822796630859375,
       "step": 189000
     },
     {
-      "epoch": 4.356636271529888,
-      "grad_norm": 3.192901611328125,
-      "learning_rate": 2.4702629397478222e-05,
-      "loss": 0.62,
+      "epoch": 2.1783199321859246,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.470008435014034e-05,
+      "loss": 0.9743334197998047,
       "step": 189200
     },
     {
-      "epoch": 4.361241595284149,
-      "grad_norm": 3.3176310062408447,
-      "learning_rate": 2.4695357833655705e-05,
-      "loss": 0.623,
+      "epoch": 2.1806226172606396,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.4692812786317827e-05,
+      "loss": 0.9917398071289063,
       "step": 189400
     },
     {
-      "epoch": 4.365846919038408,
-      "grad_norm": 3.1651062965393066,
-      "learning_rate": 2.468808626983319e-05,
-      "loss": 0.6313,
+      "epoch": 2.1829253023353545,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.468554122249531e-05,
+      "loss": 0.9816542053222657,
       "step": 189600
     },
     {
-      "epoch": 4.370452242792668,
-      "grad_norm": 4.049862861633301,
-      "learning_rate": 2.4680814706010673e-05,
-      "loss": 0.6236,
+      "epoch": 2.1852279874100695,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.4678269658672795e-05,
+      "loss": 0.9726030731201172,
       "step": 189800
     },
     {
-      "epoch": 4.3750575665469285,
-      "grad_norm": 3.2649123668670654,
-      "learning_rate": 2.467354314218816e-05,
-      "loss": 0.614,
+      "epoch": 2.187530672484784,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.467099809485028e-05,
+      "loss": 0.986077880859375,
       "step": 190000
     },
     {
-      "epoch": 4.379662890301188,
-      "grad_norm": 2.824345350265503,
-      "learning_rate": 2.4666271578365645e-05,
-      "loss": 0.6238,
+      "epoch": 2.189833357559499,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.4663726531027764e-05,
+      "loss": 0.991075439453125,
       "step": 190200
     },
     {
-      "epoch": 4.384268214055448,
-      "grad_norm": 3.1499876976013184,
-      "learning_rate": 2.4659000014543128e-05,
-      "loss": 0.6226,
+      "epoch": 2.192136042634214,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.4656454967205246e-05,
+      "loss": 0.9752172088623047,
       "step": 190400
     },
     {
-      "epoch": 4.388873537809708,
-      "grad_norm": 3.3163814544677734,
-      "learning_rate": 2.4651728450720614e-05,
-      "loss": 0.6126,
+      "epoch": 2.194438727708929,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.4649183403382732e-05,
+      "loss": 1.0028966522216798,
       "step": 190600
     },
     {
-      "epoch": 4.393478861563968,
-      "grad_norm": 2.8196558952331543,
-      "learning_rate": 2.4644493244717212e-05,
-      "loss": 0.6226,
+      "epoch": 2.196741412783644,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.4641911839560215e-05,
+      "loss": 0.97932373046875,
       "step": 190800
     },
     {
-      "epoch": 4.398084185318228,
-      "grad_norm": 2.6953351497650146,
-      "learning_rate": 2.463722168089469e-05,
-      "loss": 0.62,
+      "epoch": 2.199044097858359,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.46346402757377e-05,
+      "loss": 0.9712058258056641,
       "step": 191000
     },
     {
-      "epoch": 4.402689509072488,
-      "grad_norm": 2.942456007003784,
-      "learning_rate": 2.4629950117072177e-05,
-      "loss": 0.6265,
+      "epoch": 2.201346782933074,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.4627368711915186e-05,
+      "loss": 0.985837631225586,
       "step": 191200
     },
     {
-      "epoch": 4.407294832826747,
-      "grad_norm": 2.7047855854034424,
-      "learning_rate": 2.4622678553249663e-05,
-      "loss": 0.6177,
+      "epoch": 2.203649468007789,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.462009714809267e-05,
+      "loss": 0.9744161224365234,
       "step": 191400
     },
     {
-      "epoch": 4.411900156581008,
-      "grad_norm": 3.093517541885376,
-      "learning_rate": 2.4615406989427146e-05,
-      "loss": 0.6225,
+      "epoch": 2.205952153082504,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.461282558427015e-05,
+      "loss": 0.9757750701904296,
       "step": 191600
     },
     {
-      "epoch": 4.4165054803352675,
-      "grad_norm": 3.8072400093078613,
-      "learning_rate": 2.4608171783423744e-05,
-      "loss": 0.6177,
+      "epoch": 2.208254838157219,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4605554020447637e-05,
+      "loss": 0.9795994567871094,
       "step": 191800
     },
     {
-      "epoch": 4.421110804089528,
-      "grad_norm": 2.6595051288604736,
-      "learning_rate": 2.460090021960123e-05,
-      "loss": 0.6163,
+      "epoch": 2.2105575232319334,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.4598282456625123e-05,
+      "loss": 0.9868009948730468,
       "step": 192000
     },
     {
-      "epoch": 4.4257161278437875,
-      "grad_norm": 3.101194143295288,
-      "learning_rate": 2.4593628655778713e-05,
-      "loss": 0.6157,
+      "epoch": 2.2128602083066484,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4591010892802606e-05,
+      "loss": 0.9895838928222657,
       "step": 192200
     },
     {
-      "epoch": 4.430321451598047,
-      "grad_norm": 2.7680416107177734,
-      "learning_rate": 2.4586357091956195e-05,
-      "loss": 0.6209,
+      "epoch": 2.2151628933813634,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4583739328980092e-05,
+      "loss": 0.9988584899902344,
       "step": 192400
     },
     {
-      "epoch": 4.434926775352308,
-      "grad_norm": 3.988497734069824,
-      "learning_rate": 2.457908552813368e-05,
-      "loss": 0.6246,
+      "epoch": 2.2174655784560784,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.4576467765157578e-05,
+      "loss": 0.9978138732910157,
       "step": 192600
     },
     {
-      "epoch": 4.439532099106567,
-      "grad_norm": 3.0158700942993164,
-      "learning_rate": 2.4571813964311164e-05,
-      "loss": 0.6225,
+      "epoch": 2.2197682635307934,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.456919620133506e-05,
+      "loss": 0.992156982421875,
       "step": 192800
     },
     {
-      "epoch": 4.444137422860827,
-      "grad_norm": 3.032444953918457,
-      "learning_rate": 2.456454240048865e-05,
-      "loss": 0.6229,
+      "epoch": 2.2220709486055084,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.4561924637512543e-05,
+      "loss": 0.9902561187744141,
       "step": 193000
     },
     {
-      "epoch": 4.448742746615087,
-      "grad_norm": 2.915055274963379,
-      "learning_rate": 2.4557270836666136e-05,
-      "loss": 0.629,
+      "epoch": 2.2243736336802233,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.455465307369003e-05,
+      "loss": 0.9733341979980469,
       "step": 193200
     },
     {
-      "epoch": 4.453348070369347,
-      "grad_norm": 3.404008388519287,
-      "learning_rate": 2.4549999272843618e-05,
-      "loss": 0.6233,
+      "epoch": 2.2266763187549383,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.454738150986751e-05,
+      "loss": 0.9912680816650391,
       "step": 193400
     },
     {
-      "epoch": 4.4579533941236065,
-      "grad_norm": 2.661874294281006,
-      "learning_rate": 2.45427277090211e-05,
-      "loss": 0.6216,
+      "epoch": 2.2289790038296533,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.4540109946044997e-05,
+      "loss": 0.9892380523681641,
       "step": 193600
     },
     {
-      "epoch": 4.462558717877867,
-      "grad_norm": 3.4487946033477783,
-      "learning_rate": 2.4535456145198586e-05,
-      "loss": 0.6312,
+      "epoch": 2.231281688904368,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.4532838382222483e-05,
+      "loss": 0.9895990753173828,
       "step": 193800
     },
     {
-      "epoch": 4.4671640416321265,
-      "grad_norm": 2.926490068435669,
-      "learning_rate": 2.4528184581376072e-05,
-      "loss": 0.6134,
+      "epoch": 2.233584373979083,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.4525566818399966e-05,
+      "loss": 0.9855638885498047,
       "step": 194000
     },
     {
-      "epoch": 4.471769365386387,
-      "grad_norm": 3.7607150077819824,
-      "learning_rate": 2.4520913017553555e-05,
-      "loss": 0.621,
+      "epoch": 2.235887059053798,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.4518295254577448e-05,
+      "loss": 0.9776099395751953,
       "step": 194200
     },
     {
-      "epoch": 4.476374689140647,
-      "grad_norm": 4.96255350112915,
-      "learning_rate": 2.451364145373104e-05,
-      "loss": 0.622,
+      "epoch": 2.238189744128513,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4511023690754934e-05,
+      "loss": 0.9914614868164062,
       "step": 194400
     },
     {
-      "epoch": 4.480980012894906,
-      "grad_norm": 3.3920106887817383,
-      "learning_rate": 2.4506369889908527e-05,
-      "loss": 0.6184,
+      "epoch": 2.2404924292032278,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.450375212693242e-05,
+      "loss": 0.9876277160644531,
       "step": 194600
     },
     {
-      "epoch": 4.485585336649167,
-      "grad_norm": 3.106212854385376,
-      "learning_rate": 2.449909832608601e-05,
-      "loss": 0.624,
+      "epoch": 2.2427951142779428,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.4496480563109902e-05,
+      "loss": 0.9671038055419922,
       "step": 194800
     },
     {
-      "epoch": 4.490190660403426,
-      "grad_norm": 3.0986621379852295,
-      "learning_rate": 2.4491826762263492e-05,
-      "loss": 0.6199,
+      "epoch": 2.2450977993526577,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.448920899928739e-05,
+      "loss": 0.9613436889648438,
       "step": 195000
     },
     {
-      "epoch": 4.494795984157686,
-      "grad_norm": 2.7702393531799316,
-      "learning_rate": 2.4484555198440978e-05,
-      "loss": 0.6219,
+      "epoch": 2.2474004844273727,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.4481937435464874e-05,
+      "loss": 0.9555902099609375,
       "step": 195200
     },
     {
-      "epoch": 4.499401307911946,
-      "grad_norm": 2.7158069610595703,
-      "learning_rate": 2.447728363461846e-05,
-      "loss": 0.6228,
+      "epoch": 2.2497031695020877,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.4474665871642353e-05,
+      "loss": 1.0098719787597656,
       "step": 195400
     },
     {
-      "epoch": 4.504006631666206,
-      "grad_norm": 2.6804583072662354,
-      "learning_rate": 2.4470012070795946e-05,
-      "loss": 0.6264,
+      "epoch": 2.2520058545768027,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.446739430781984e-05,
+      "loss": 0.9947319793701171,
       "step": 195600
     },
     {
-      "epoch": 4.508611955420466,
-      "grad_norm": 3.254936695098877,
-      "learning_rate": 2.4462740506973432e-05,
-      "loss": 0.6178,
+      "epoch": 2.2543085396515172,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4460122743997325e-05,
+      "loss": 0.9614073944091797,
       "step": 195800
     },
     {
-      "epoch": 4.513217279174726,
-      "grad_norm": 2.8413448333740234,
-      "learning_rate": 2.4455468943150915e-05,
-      "loss": 0.6221,
+      "epoch": 2.256611224726232,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.4452851180174808e-05,
+      "loss": 0.9944358062744141,
       "step": 196000
     },
     {
-      "epoch": 4.517822602928986,
-      "grad_norm": 3.7565321922302246,
-      "learning_rate": 2.4448197379328397e-05,
-      "loss": 0.6167,
+      "epoch": 2.258913909800947,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.4445579616352294e-05,
+      "loss": 0.9826112365722657,
       "step": 196200
     },
     {
-      "epoch": 4.522427926683246,
-      "grad_norm": 2.5034019947052,
-      "learning_rate": 2.4440925815505883e-05,
-      "loss": 0.6137,
+      "epoch": 2.261216594875662,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.443830805252978e-05,
+      "loss": 0.9536886596679688,
       "step": 196400
     },
     {
-      "epoch": 4.527033250437506,
-      "grad_norm": 3.1947107315063477,
-      "learning_rate": 2.4433654251683366e-05,
-      "loss": 0.621,
+      "epoch": 2.263519279950377,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.4431036488707262e-05,
+      "loss": 0.9850600433349609,
       "step": 196600
     },
     {
-      "epoch": 4.531638574191765,
-      "grad_norm": 2.7699062824249268,
-      "learning_rate": 2.4426419045679964e-05,
-      "loss": 0.6192,
+      "epoch": 2.265821965025092,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.4423764924884745e-05,
+      "loss": 0.9687165832519531,
       "step": 196800
     },
     {
-      "epoch": 4.536243897946026,
-      "grad_norm": 3.791555166244507,
-      "learning_rate": 2.441914748185745e-05,
-      "loss": 0.633,
+      "epoch": 2.268124650099807,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.441649336106223e-05,
+      "loss": 0.9938833618164062,
       "step": 197000
     },
     {
-      "epoch": 4.540849221700285,
-      "grad_norm": 2.6840012073516846,
-      "learning_rate": 2.4411875918034933e-05,
-      "loss": 0.6252,
+      "epoch": 2.270427335174522,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4409221797239713e-05,
+      "loss": 0.9914147186279297,
       "step": 197200
     },
     {
-      "epoch": 4.545454545454545,
-      "grad_norm": 3.7765004634857178,
-      "learning_rate": 2.440460435421242e-05,
-      "loss": 0.6171,
+      "epoch": 2.272730020249237,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.44019502334172e-05,
+      "loss": 0.9896980285644531,
       "step": 197400
     },
     {
-      "epoch": 4.5500598692088055,
-      "grad_norm": 2.8564860820770264,
-      "learning_rate": 2.4397332790389904e-05,
-      "loss": 0.6235,
+      "epoch": 2.2750327053239516,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4394678669594685e-05,
+      "loss": 1.000285873413086,
       "step": 197600
     },
     {
-      "epoch": 4.554665192963065,
-      "grad_norm": 3.1641488075256348,
-      "learning_rate": 2.4390061226567384e-05,
-      "loss": 0.6197,
+      "epoch": 2.2773353903986666,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.438740710577217e-05,
+      "loss": 0.9800799560546875,
       "step": 197800
     },
     {
-      "epoch": 4.5592705167173255,
-      "grad_norm": 3.0477914810180664,
-      "learning_rate": 2.438278966274487e-05,
-      "loss": 0.6269,
+      "epoch": 2.2796380754733816,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.438013554194965e-05,
+      "loss": 0.9833818054199219,
       "step": 198000
     },
     {
-      "epoch": 4.563875840471585,
-      "grad_norm": 3.0091304779052734,
-      "learning_rate": 2.4375518098922355e-05,
-      "loss": 0.622,
+      "epoch": 2.2819407605480966,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.4372863978127136e-05,
+      "loss": 0.9885999298095703,
       "step": 198200
     },
     {
-      "epoch": 4.568481164225845,
-      "grad_norm": 3.7036385536193848,
-      "learning_rate": 2.4368246535099838e-05,
-      "loss": 0.6151,
+      "epoch": 2.2842434456228116,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.4365592414304622e-05,
+      "loss": 0.9842572784423829,
       "step": 198400
     },
     {
-      "epoch": 4.573086487980105,
-      "grad_norm": 2.7939934730529785,
-      "learning_rate": 2.4360974971277324e-05,
-      "loss": 0.6295,
+      "epoch": 2.2865461306975265,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.4358320850482104e-05,
+      "loss": 1.0083957672119142,
       "step": 198600
     },
     {
-      "epoch": 4.577691811734365,
-      "grad_norm": 3.725222110748291,
-      "learning_rate": 2.435370340745481e-05,
-      "loss": 0.6212,
+      "epoch": 2.2888488157722415,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.435104928665959e-05,
+      "loss": 0.9839616394042969,
       "step": 198800
     },
     {
-      "epoch": 4.582297135488625,
-      "grad_norm": 3.55149245262146,
-      "learning_rate": 2.4346431843632292e-05,
-      "loss": 0.6143,
+      "epoch": 2.2911515008469565,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.4343777722837076e-05,
+      "loss": 0.9807469940185547,
       "step": 199000
     },
     {
-      "epoch": 4.586902459242885,
-      "grad_norm": 3.657155990600586,
-      "learning_rate": 2.4339160279809775e-05,
-      "loss": 0.6245,
+      "epoch": 2.2934541859216715,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.433650615901456e-05,
+      "loss": 0.9862178039550781,
       "step": 199200
     },
     {
-      "epoch": 4.5915077829971445,
-      "grad_norm": 2.774144172668457,
-      "learning_rate": 2.433188871598726e-05,
-      "loss": 0.6263,
+      "epoch": 2.295756870996386,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.432923459519204e-05,
+      "loss": 0.9967860412597657,
       "step": 199400
     },
     {
-      "epoch": 4.596113106751405,
-      "grad_norm": 2.516934633255005,
-      "learning_rate": 2.4324617152164747e-05,
-      "loss": 0.6096,
+      "epoch": 2.298059556071101,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4321963031369527e-05,
+      "loss": 0.9825498199462891,
       "step": 199600
     },
     {
-      "epoch": 4.6007184305056645,
-      "grad_norm": 3.243980646133423,
-      "learning_rate": 2.431734558834223e-05,
-      "loss": 0.6136,
+      "epoch": 2.300362241145816,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.431469146754701e-05,
+      "loss": 0.9612421417236328,
       "step": 199800
     },
     {
-      "epoch": 4.605323754259924,
-      "grad_norm": 3.0042223930358887,
-      "learning_rate": 2.4310110382338828e-05,
-      "loss": 0.6219,
+      "epoch": 2.302664926220531,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.4307419903724496e-05,
+      "loss": 0.9812924194335938,
       "step": 200000
     },
     {
-      "epoch": 4.609929078014185,
-      "grad_norm": 2.626081943511963,
-      "learning_rate": 2.430283881851631e-05,
-      "loss": 0.6093,
+      "epoch": 2.304967611295246,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.430014833990198e-05,
+      "loss": 0.9764970397949219,
       "step": 200200
     },
     {
-      "epoch": 4.614534401768444,
-      "grad_norm": 3.4706902503967285,
-      "learning_rate": 2.4295567254693793e-05,
-      "loss": 0.6217,
+      "epoch": 2.307270296369961,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.4292876776079467e-05,
+      "loss": 0.983216781616211,
       "step": 200400
     },
     {
-      "epoch": 4.619139725522704,
-      "grad_norm": 3.2710297107696533,
-      "learning_rate": 2.428829569087128e-05,
-      "loss": 0.617,
+      "epoch": 2.309572981444676,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4285605212256946e-05,
+      "loss": 0.9808189392089843,
       "step": 200600
     },
     {
-      "epoch": 4.623745049276964,
-      "grad_norm": 2.6531553268432617,
-      "learning_rate": 2.4281024127048765e-05,
-      "loss": 0.636,
+      "epoch": 2.311875666519391,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.4278333648434432e-05,
+      "loss": 0.9695719909667969,
       "step": 200800
     },
     {
-      "epoch": 4.628350373031224,
-      "grad_norm": 2.9135398864746094,
-      "learning_rate": 2.4273752563226247e-05,
-      "loss": 0.6269,
+      "epoch": 2.314178351594106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.427106208461192e-05,
+      "loss": 0.9838567352294922,
       "step": 201000
     },
     {
-      "epoch": 4.632955696785484,
-      "grad_norm": 2.918564558029175,
-      "learning_rate": 2.4266480999403733e-05,
-      "loss": 0.6139,
+      "epoch": 2.3164810366688204,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.42637905207894e-05,
+      "loss": 0.9900138092041015,
       "step": 201200
     },
     {
-      "epoch": 4.637561020539744,
-      "grad_norm": 3.0329713821411133,
-      "learning_rate": 2.425920943558122e-05,
-      "loss": 0.619,
+      "epoch": 2.3187837217435354,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.4256518956966887e-05,
+      "loss": 0.9634355926513671,
       "step": 201400
     },
     {
-      "epoch": 4.6421663442940035,
-      "grad_norm": 2.717499017715454,
-      "learning_rate": 2.42519378717587e-05,
-      "loss": 0.6141,
+      "epoch": 2.3210864068182504,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.4249247393144373e-05,
+      "loss": 0.9662844848632812,
       "step": 201600
     },
     {
-      "epoch": 4.646771668048264,
-      "grad_norm": 3.2214903831481934,
-      "learning_rate": 2.4244666307936184e-05,
-      "loss": 0.614,
+      "epoch": 2.3233890918929654,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4241975829321852e-05,
+      "loss": 0.9637671661376953,
       "step": 201800
     },
     {
-      "epoch": 4.651376991802524,
-      "grad_norm": 2.94065260887146,
-      "learning_rate": 2.423739474411367e-05,
-      "loss": 0.6206,
+      "epoch": 2.3256917769676804,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4234704265499338e-05,
+      "loss": 0.9792532348632812,
       "step": 202000
     },
     {
-      "epoch": 4.655982315556784,
-      "grad_norm": 3.6057605743408203,
-      "learning_rate": 2.4230123180291152e-05,
-      "loss": 0.6097,
+      "epoch": 2.3279944620423954,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.4227432701676824e-05,
+      "loss": 0.9898047637939453,
       "step": 202200
     },
     {
-      "epoch": 4.660587639311044,
-      "grad_norm": 3.165396213531494,
-      "learning_rate": 2.422285161646864e-05,
-      "loss": 0.6283,
+      "epoch": 2.3302971471171103,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.4220161137854306e-05,
+      "loss": 0.974963150024414,
       "step": 202400
     },
     {
-      "epoch": 4.665192963065303,
-      "grad_norm": 3.9540810585021973,
-      "learning_rate": 2.4215580052646124e-05,
-      "loss": 0.6247,
+      "epoch": 2.3325998321918253,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.4212889574031792e-05,
+      "loss": 0.987943344116211,
       "step": 202600
     },
     {
-      "epoch": 4.669798286819564,
-      "grad_norm": 2.56862735748291,
-      "learning_rate": 2.4208308488823607e-05,
-      "loss": 0.6188,
+      "epoch": 2.3349025172665403,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.4205618010209278e-05,
+      "loss": 0.9598580169677734,
       "step": 202800
     },
     {
-      "epoch": 4.674403610573823,
-      "grad_norm": 2.8995814323425293,
-      "learning_rate": 2.420103692500109e-05,
-      "loss": 0.6152,
+      "epoch": 2.337205202341255,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.419834644638676e-05,
+      "loss": 0.9809665679931641,
       "step": 203000
     },
     {
-      "epoch": 4.679008934328083,
-      "grad_norm": 3.891038656234741,
-      "learning_rate": 2.4193765361178575e-05,
-      "loss": 0.6129,
+      "epoch": 2.33950788741597,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4191074882564243e-05,
+      "loss": 0.9566979217529297,
       "step": 203200
     },
     {
-      "epoch": 4.683614258082343,
-      "grad_norm": 3.055830478668213,
-      "learning_rate": 2.418649379735606e-05,
-      "loss": 0.6204,
+      "epoch": 2.341810572490685,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.418380331874173e-05,
+      "loss": 0.979805908203125,
       "step": 203400
     },
     {
-      "epoch": 4.688219581836603,
-      "grad_norm": 2.924553394317627,
-      "learning_rate": 2.4179222233533544e-05,
-      "loss": 0.6245,
+      "epoch": 2.3441132575654,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.4176531754919215e-05,
+      "loss": 0.9806376647949219,
       "step": 203600
     },
     {
-      "epoch": 4.692824905590863,
-      "grad_norm": 3.2216031551361084,
-      "learning_rate": 2.4171987027530142e-05,
-      "loss": 0.6289,
+      "epoch": 2.3464159426401148,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4169260191096697e-05,
+      "loss": 0.9827217864990234,
       "step": 203800
     },
     {
-      "epoch": 4.697430229345123,
-      "grad_norm": 3.0679335594177246,
-      "learning_rate": 2.4164715463707625e-05,
-      "loss": 0.6114,
+      "epoch": 2.3487186277148298,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4161988627274183e-05,
+      "loss": 0.9710164642333985,
       "step": 204000
     },
     {
-      "epoch": 4.702035553099383,
-      "grad_norm": 2.5067596435546875,
-      "learning_rate": 2.415744389988511e-05,
-      "loss": 0.6313,
+      "epoch": 2.3510213127895447,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.415471706345167e-05,
+      "loss": 0.9706101989746094,
       "step": 204200
     },
     {
-      "epoch": 4.706640876853643,
-      "grad_norm": 3.141855239868164,
-      "learning_rate": 2.4150172336062597e-05,
-      "loss": 0.6201,
+      "epoch": 2.3533239978642597,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.414744549962915e-05,
+      "loss": 0.9726792907714844,
       "step": 204400
     },
     {
-      "epoch": 4.711246200607903,
-      "grad_norm": 2.854891300201416,
-      "learning_rate": 2.4142900772240076e-05,
-      "loss": 0.6298,
+      "epoch": 2.3556266829389747,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.4140173935806634e-05,
+      "loss": 0.989774169921875,
       "step": 204600
     },
     {
-      "epoch": 4.715851524362162,
-      "grad_norm": 3.0763156414031982,
-      "learning_rate": 2.413562920841756e-05,
-      "loss": 0.6246,
+      "epoch": 2.3579293680136892,
+      "grad_norm": 0.625,
+      "learning_rate": 2.413290237198412e-05,
+      "loss": 0.9620243835449219,
       "step": 204800
     },
     {
-      "epoch": 4.720456848116423,
-      "grad_norm": 3.858271837234497,
-      "learning_rate": 2.4128357644595048e-05,
-      "loss": 0.6288,
+      "epoch": 2.3602320530884042,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4125630808161603e-05,
+      "loss": 0.9820114135742187,
       "step": 205000
     },
     {
-      "epoch": 4.7250621718706824,
-      "grad_norm": 3.183344602584839,
-      "learning_rate": 2.4121086080772534e-05,
-      "loss": 0.614,
+      "epoch": 2.362534738163119,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.411835924433909e-05,
+      "loss": 0.981342544555664,
       "step": 205200
     },
     {
-      "epoch": 4.729667495624943,
-      "grad_norm": 3.646907329559326,
-      "learning_rate": 2.4113814516950016e-05,
-      "loss": 0.6259,
+      "epoch": 2.364837423237834,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4111087680516575e-05,
+      "loss": 0.9816950225830078,
       "step": 205400
     },
     {
-      "epoch": 4.7342728193792025,
-      "grad_norm": 3.3939154148101807,
-      "learning_rate": 2.4106542953127502e-05,
-      "loss": 0.6315,
+      "epoch": 2.367140108312549,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.4103816116694057e-05,
+      "loss": 0.9805652618408203,
       "step": 205600
     },
     {
-      "epoch": 4.738878143133462,
-      "grad_norm": 3.364961862564087,
-      "learning_rate": 2.4099271389304984e-05,
-      "loss": 0.6232,
+      "epoch": 2.369442793387264,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.409654455287154e-05,
+      "loss": 0.9725960540771484,
       "step": 205800
     },
     {
-      "epoch": 4.743483466887723,
-      "grad_norm": 2.7885661125183105,
-      "learning_rate": 2.4091999825482467e-05,
-      "loss": 0.6219,
+      "epoch": 2.371745478461979,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.4089272989049025e-05,
+      "loss": 0.9859681701660157,
       "step": 206000
     },
     {
-      "epoch": 4.748088790641982,
-      "grad_norm": 3.6182515621185303,
-      "learning_rate": 2.4084728261659953e-05,
-      "loss": 0.6302,
+      "epoch": 2.374048163536694,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.408200142522651e-05,
+      "loss": 0.9848648834228516,
       "step": 206200
     },
     {
-      "epoch": 4.752694114396242,
-      "grad_norm": 2.98297381401062,
-      "learning_rate": 2.407745669783744e-05,
-      "loss": 0.6217,
+      "epoch": 2.376350848611409,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.4074729861403994e-05,
+      "loss": 0.9847093200683594,
       "step": 206400
     },
     {
-      "epoch": 4.757299438150502,
-      "grad_norm": 4.925945281982422,
-      "learning_rate": 2.407018513401492e-05,
-      "loss": 0.6299,
+      "epoch": 2.3786535336861236,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.406745829758148e-05,
+      "loss": 0.986251220703125,
       "step": 206600
     },
     {
-      "epoch": 4.761904761904762,
-      "grad_norm": 2.90513277053833,
-      "learning_rate": 2.406294992801152e-05,
-      "loss": 0.6254,
+      "epoch": 2.3809562187608386,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.4060186733758966e-05,
+      "loss": 0.9802590179443359,
       "step": 206800
     },
     {
-      "epoch": 4.7665100856590215,
-      "grad_norm": 2.9922261238098145,
-      "learning_rate": 2.4055714722008115e-05,
-      "loss": 0.6204,
+      "epoch": 2.3832589038355536,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.4052915169936445e-05,
+      "loss": 0.9795805358886719,
       "step": 207000
     },
     {
-      "epoch": 4.771115409413282,
-      "grad_norm": 3.269758939743042,
-      "learning_rate": 2.40484431581856e-05,
-      "loss": 0.6091,
+      "epoch": 2.3855615889102686,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.404564360611393e-05,
+      "loss": 0.9767790985107422,
       "step": 207200
     },
     {
-      "epoch": 4.7757207331675415,
-      "grad_norm": 2.7046399116516113,
-      "learning_rate": 2.4041171594363084e-05,
-      "loss": 0.6204,
+      "epoch": 2.3878642739849836,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.4038372042291417e-05,
+      "loss": 0.9922294616699219,
       "step": 207400
     },
     {
-      "epoch": 4.780326056921801,
-      "grad_norm": 3.494405508041382,
-      "learning_rate": 2.4033900030540566e-05,
-      "loss": 0.6361,
+      "epoch": 2.3901669590596986,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.40311004784689e-05,
+      "loss": 0.9741712951660156,
       "step": 207600
     },
     {
-      "epoch": 4.784931380676062,
-      "grad_norm": 2.91849684715271,
-      "learning_rate": 2.4026628466718052e-05,
-      "loss": 0.623,
+      "epoch": 2.3924696441344135,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4023828914646385e-05,
+      "loss": 0.9773464965820312,
       "step": 207800
     },
     {
-      "epoch": 4.789536704430321,
-      "grad_norm": 2.500559091567993,
-      "learning_rate": 2.4019356902895538e-05,
-      "loss": 0.615,
+      "epoch": 2.3947723292091285,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.401655735082387e-05,
+      "loss": 0.9808747863769531,
       "step": 208000
     },
     {
-      "epoch": 4.794142028184582,
-      "grad_norm": 3.2106189727783203,
-      "learning_rate": 2.401208533907302e-05,
-      "loss": 0.6345,
+      "epoch": 2.3970750142838435,
+      "grad_norm": 0.875,
+      "learning_rate": 2.400928578700135e-05,
+      "loss": 0.9907157135009765,
       "step": 208200
     },
     {
-      "epoch": 4.798747351938841,
-      "grad_norm": 3.0638959407806396,
-      "learning_rate": 2.4004813775250506e-05,
-      "loss": 0.6176,
+      "epoch": 2.399377699358558,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4002014223178836e-05,
+      "loss": 0.9853793334960937,
       "step": 208400
     },
     {
-      "epoch": 4.803352675693101,
-      "grad_norm": 3.2531685829162598,
-      "learning_rate": 2.3997542211427992e-05,
-      "loss": 0.628,
+      "epoch": 2.401680384433273,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3994742659356322e-05,
+      "loss": 0.978099365234375,
       "step": 208600
     },
     {
-      "epoch": 4.807957999447361,
-      "grad_norm": 2.51589035987854,
-      "learning_rate": 2.3990270647605475e-05,
-      "loss": 0.6179,
+      "epoch": 2.403983069507988,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.3987471095533805e-05,
+      "loss": 0.9827703857421874,
       "step": 208800
     },
     {
-      "epoch": 4.812563323201621,
-      "grad_norm": 4.157649040222168,
-      "learning_rate": 2.3982999083782957e-05,
-      "loss": 0.6069,
+      "epoch": 2.406285754582703,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.398019953171129e-05,
+      "loss": 0.9785765838623047,
       "step": 209000
     },
     {
-      "epoch": 4.817168646955881,
-      "grad_norm": 2.5439136028289795,
-      "learning_rate": 2.3975727519960443e-05,
-      "loss": 0.6243,
+      "epoch": 2.408588439657418,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.3972927967888776e-05,
+      "loss": 0.9798247528076172,
       "step": 209200
     },
     {
-      "epoch": 4.821773970710141,
-      "grad_norm": 2.936598300933838,
-      "learning_rate": 2.396845595613793e-05,
-      "loss": 0.6232,
+      "epoch": 2.410891124732133,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3965656404066262e-05,
+      "loss": 0.9880730438232422,
       "step": 209400
     },
     {
-      "epoch": 4.826379294464401,
-      "grad_norm": 3.0496158599853516,
-      "learning_rate": 2.3961184392315412e-05,
-      "loss": 0.6308,
+      "epoch": 2.413193809806848,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.395838484024374e-05,
+      "loss": 0.9864563751220703,
       "step": 209600
     },
     {
-      "epoch": 4.830984618218661,
-      "grad_norm": 3.6222925186157227,
-      "learning_rate": 2.3953912828492898e-05,
-      "loss": 0.6168,
+      "epoch": 2.415496494881563,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.3951113276421227e-05,
+      "loss": 0.9851638031005859,
       "step": 209800
     },
     {
-      "epoch": 4.835589941972921,
-      "grad_norm": 3.679816722869873,
-      "learning_rate": 2.394664126467038e-05,
-      "loss": 0.6095,
+      "epoch": 2.417799179956278,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.3943841712598713e-05,
+      "loss": 0.9939958953857422,
       "step": 210000
     },
     {
-      "epoch": 4.84019526572718,
-      "grad_norm": 3.8149142265319824,
-      "learning_rate": 2.3939369700847863e-05,
-      "loss": 0.616,
+      "epoch": 2.420101865030993,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.3936570148776196e-05,
+      "loss": 0.9842073059082032,
       "step": 210200
     },
     {
-      "epoch": 4.844800589481441,
-      "grad_norm": 3.1322624683380127,
-      "learning_rate": 2.393209813702535e-05,
-      "loss": 0.6137,
+      "epoch": 2.422404550105708,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.3929298584953682e-05,
+      "loss": 0.9827200317382813,
       "step": 210400
     },
     {
-      "epoch": 4.8494059132357,
-      "grad_norm": 2.972074031829834,
-      "learning_rate": 2.3924826573202835e-05,
-      "loss": 0.6372,
+      "epoch": 2.4247072351804224,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.3922027021131168e-05,
+      "loss": 0.9755634307861328,
       "step": 210600
     },
     {
-      "epoch": 4.85401123698996,
-      "grad_norm": 2.327326774597168,
-      "learning_rate": 2.3917555009380317e-05,
-      "loss": 0.6182,
+      "epoch": 2.4270099202551374,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.3914755457308647e-05,
+      "loss": 0.9901658630371094,
       "step": 210800
     },
     {
-      "epoch": 4.85861656074422,
-      "grad_norm": 3.573911666870117,
-      "learning_rate": 2.3910283445557803e-05,
-      "loss": 0.6278,
+      "epoch": 2.4293126053298524,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.3907483893486133e-05,
+      "loss": 0.979815444946289,
       "step": 211000
     },
     {
-      "epoch": 4.86322188449848,
-      "grad_norm": 3.1551811695098877,
-      "learning_rate": 2.39030482395544e-05,
-      "loss": 0.6174,
+      "epoch": 2.4316152904045674,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.390021232966362e-05,
+      "loss": 0.9816466522216797,
       "step": 211200
     },
     {
-      "epoch": 4.8678272082527405,
-      "grad_norm": 3.1177220344543457,
-      "learning_rate": 2.389577667573188e-05,
-      "loss": 0.6203,
+      "epoch": 2.4339179754792823,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.38929407658411e-05,
+      "loss": 0.9903483581542969,
       "step": 211400
     },
     {
-      "epoch": 4.872432532007,
-      "grad_norm": 3.517178773880005,
-      "learning_rate": 2.3888505111909367e-05,
-      "loss": 0.6165,
+      "epoch": 2.4362206605539973,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3885669202018587e-05,
+      "loss": 0.9710890197753906,
       "step": 211600
     },
     {
-      "epoch": 4.87703785576126,
-      "grad_norm": 2.932919502258301,
-      "learning_rate": 2.3881233548086853e-05,
-      "loss": 0.6154,
+      "epoch": 2.4385233456287123,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.3878397638196073e-05,
+      "loss": 0.9668023681640625,
       "step": 211800
     },
     {
-      "epoch": 4.88164317951552,
-      "grad_norm": 3.162429094314575,
-      "learning_rate": 2.3873961984264335e-05,
-      "loss": 0.6215,
+      "epoch": 2.4408260307034273,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.3871126074373552e-05,
+      "loss": 1.0000997161865235,
       "step": 212000
     },
     {
-      "epoch": 4.88624850326978,
-      "grad_norm": 3.3734543323516846,
-      "learning_rate": 2.386669042044182e-05,
-      "loss": 0.6068,
+      "epoch": 2.4431287157781423,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.3863854510551038e-05,
+      "loss": 0.9645013427734375,
       "step": 212200
     },
     {
-      "epoch": 4.89085382702404,
-      "grad_norm": 3.567715644836426,
-      "learning_rate": 2.3859418856619307e-05,
-      "loss": 0.6183,
+      "epoch": 2.445431400852857,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.3856582946728524e-05,
+      "loss": 0.9931552886962891,
       "step": 212400
     },
     {
-      "epoch": 4.8954591507783,
-      "grad_norm": 3.8765225410461426,
-      "learning_rate": 2.385214729279679e-05,
-      "loss": 0.6122,
+      "epoch": 2.447734085927572,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.384931138290601e-05,
+      "loss": 0.9747293853759765,
       "step": 212600
     },
     {
-      "epoch": 4.9000644745325594,
-      "grad_norm": 3.328627109527588,
-      "learning_rate": 2.3844875728974272e-05,
-      "loss": 0.6196,
+      "epoch": 2.450036771002287,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.3842039819083492e-05,
+      "loss": 0.9917967224121094,
       "step": 212800
     },
     {
-      "epoch": 4.90466979828682,
-      "grad_norm": 3.3088150024414062,
-      "learning_rate": 2.3837604165151758e-05,
-      "loss": 0.6223,
+      "epoch": 2.4523394560770018,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.3834768255260978e-05,
+      "loss": 0.9693154144287109,
       "step": 213000
     },
     {
-      "epoch": 4.9092751220410795,
-      "grad_norm": 3.936145305633545,
-      "learning_rate": 2.3830368959148353e-05,
-      "loss": 0.6145,
+      "epoch": 2.4546421411517167,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.3827496691438464e-05,
+      "loss": 1.0230400848388672,
       "step": 213200
     },
     {
-      "epoch": 4.913880445795339,
-      "grad_norm": 4.120792388916016,
-      "learning_rate": 2.382309739532584e-05,
-      "loss": 0.6168,
+      "epoch": 2.4569448262264317,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3820225127615943e-05,
+      "loss": 0.9706953430175781,
       "step": 213400
     },
     {
-      "epoch": 4.9184857695496,
-      "grad_norm": 2.7531352043151855,
-      "learning_rate": 2.3815825831503325e-05,
-      "loss": 0.608,
+      "epoch": 2.4592475113011467,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.381295356379343e-05,
+      "loss": 1.0089253997802734,
       "step": 213600
     },
     {
-      "epoch": 4.923091093303859,
-      "grad_norm": 3.97782301902771,
-      "learning_rate": 2.3808554267680807e-05,
-      "loss": 0.6283,
+      "epoch": 2.4615501963758617,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3805681999970915e-05,
+      "loss": 1.0018858337402343,
       "step": 213800
     },
     {
-      "epoch": 4.927696417058119,
-      "grad_norm": 2.840651035308838,
-      "learning_rate": 2.3801282703858293e-05,
-      "loss": 0.608,
+      "epoch": 2.4638528814505767,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.3798410436148398e-05,
+      "loss": 0.9635964965820313,
       "step": 214000
     },
     {
-      "epoch": 4.932301740812379,
-      "grad_norm": 2.7774887084960938,
-      "learning_rate": 2.3794011140035776e-05,
-      "loss": 0.6015,
+      "epoch": 2.466155566525291,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.3791138872325884e-05,
+      "loss": 0.9803479766845703,
       "step": 214200
     },
     {
-      "epoch": 4.936907064566639,
-      "grad_norm": 2.750030040740967,
-      "learning_rate": 2.378677593403237e-05,
-      "loss": 0.5992,
+      "epoch": 2.468458251600006,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.378386730850337e-05,
+      "loss": 0.9677488708496094,
       "step": 214400
     },
     {
-      "epoch": 4.941512388320899,
-      "grad_norm": 3.340794563293457,
-      "learning_rate": 2.3779504370209857e-05,
-      "loss": 0.6196,
+      "epoch": 2.470760936674721,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.377659574468085e-05,
+      "loss": 0.9796437072753906,
       "step": 214600
     },
     {
-      "epoch": 4.946117712075159,
-      "grad_norm": 2.626340866088867,
-      "learning_rate": 2.3772232806387343e-05,
-      "loss": 0.6199,
+      "epoch": 2.473063621749436,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3769324180858335e-05,
+      "loss": 0.9855581665039063,
       "step": 214800
     },
     {
-      "epoch": 4.9507230358294185,
-      "grad_norm": 2.984501361846924,
-      "learning_rate": 2.3764961242564825e-05,
-      "loss": 0.6095,
+      "epoch": 2.475366306824151,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.376205261703582e-05,
+      "loss": 0.9737202453613282,
       "step": 215000
     },
     {
-      "epoch": 4.955328359583679,
-      "grad_norm": 3.6400811672210693,
-      "learning_rate": 2.375768967874231e-05,
-      "loss": 0.6172,
+      "epoch": 2.477668991898866,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.3754781053213306e-05,
+      "loss": 0.9882557678222657,
       "step": 215200
     },
     {
-      "epoch": 4.959933683337939,
-      "grad_norm": 3.16477632522583,
-      "learning_rate": 2.3750418114919797e-05,
-      "loss": 0.6134,
+      "epoch": 2.479971676973581,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.374750948939079e-05,
+      "loss": 1.0013471221923829,
       "step": 215400
     },
     {
-      "epoch": 4.964539007092198,
-      "grad_norm": 3.852839708328247,
-      "learning_rate": 2.374314655109728e-05,
-      "loss": 0.6247,
+      "epoch": 2.482274362048296,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3740237925568275e-05,
+      "loss": 0.9897122192382812,
       "step": 215600
     },
     {
-      "epoch": 4.969144330846459,
-      "grad_norm": 3.3444173336029053,
-      "learning_rate": 2.3735874987274762e-05,
-      "loss": 0.6085,
+      "epoch": 2.484577047123011,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.373296636174576e-05,
+      "loss": 0.9997686767578124,
       "step": 215800
     },
     {
-      "epoch": 4.973749654600718,
-      "grad_norm": 4.497110366821289,
-      "learning_rate": 2.3728603423452248e-05,
-      "loss": 0.6162,
+      "epoch": 2.4868797321977256,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.372569479792324e-05,
+      "loss": 0.958477783203125,
       "step": 216000
     },
     {
-      "epoch": 4.978354978354979,
-      "grad_norm": 2.6496007442474365,
-      "learning_rate": 2.372133185962973e-05,
-      "loss": 0.6196,
+      "epoch": 2.4891824172724406,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.3718423234100726e-05,
+      "loss": 0.9921128845214844,
       "step": 216200
     },
     {
-      "epoch": 4.982960302109238,
-      "grad_norm": 3.1037495136260986,
-      "learning_rate": 2.371409665362633e-05,
-      "loss": 0.6156,
+      "epoch": 2.4914851023471556,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.371115167027821e-05,
+      "loss": 0.9742437744140625,
       "step": 216400
     },
     {
-      "epoch": 4.987565625863498,
-      "grad_norm": 2.7602591514587402,
-      "learning_rate": 2.3706825089803815e-05,
-      "loss": 0.628,
+      "epoch": 2.4937877874218706,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.3703880106455694e-05,
+      "loss": 0.9888323211669922,
       "step": 216600
     },
     {
-      "epoch": 4.992170949617758,
-      "grad_norm": 4.069486141204834,
-      "learning_rate": 2.3699553525981298e-05,
-      "loss": 0.6232,
+      "epoch": 2.4960904724965856,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.369660854263318e-05,
+      "loss": 0.9792591857910157,
       "step": 216800
     },
     {
-      "epoch": 4.996776273372018,
-      "grad_norm": 2.9536936283111572,
-      "learning_rate": 2.3692281962158784e-05,
-      "loss": 0.6129,
+      "epoch": 2.4983931575713005,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3689336978810666e-05,
+      "loss": 0.9883199310302735,
       "step": 217000
     },
     {
-      "epoch": 5.0,
-      "eval_loss": 0.5846441388130188,
-      "eval_runtime": 145.6578,
-      "eval_samples_per_second": 194.71,
-      "eval_steps_per_second": 12.172,
-      "step": 217140
-    },
-    {
-      "epoch": 5.001381597126278,
-      "grad_norm": 3.7548468112945557,
-      "learning_rate": 2.3685010398336266e-05,
-      "loss": 0.6154,
+      "epoch": 2.5006958426460155,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.3682065414988145e-05,
+      "loss": 0.9837226867675781,
       "step": 217200
     },
     {
-      "epoch": 5.005986920880538,
-      "grad_norm": 3.002516269683838,
-      "learning_rate": 2.367773883451375e-05,
-      "loss": 0.6113,
+      "epoch": 2.5029985277207305,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.367479385116563e-05,
+      "loss": 0.9776473999023437,
       "step": 217400
     },
     {
-      "epoch": 5.010592244634798,
-      "grad_norm": 2.816727638244629,
-      "learning_rate": 2.3670467270691235e-05,
-      "loss": 0.6102,
+      "epoch": 2.5053012127954455,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.3667522287343117e-05,
+      "loss": 0.9647300720214844,
       "step": 217600
     },
     {
-      "epoch": 5.015197568389058,
-      "grad_norm": 2.5023694038391113,
-      "learning_rate": 2.366319570686872e-05,
-      "loss": 0.6157,
+      "epoch": 2.50760389787016,
+      "grad_norm": 0.5,
+      "learning_rate": 2.36602507235206e-05,
+      "loss": 0.9693010711669922,
       "step": 217800
     },
     {
-      "epoch": 5.019802892143318,
-      "grad_norm": 3.309704542160034,
-      "learning_rate": 2.3655924143046203e-05,
-      "loss": 0.6056,
+      "epoch": 2.509906582944875,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3652979159698085e-05,
+      "loss": 0.9758069610595703,
       "step": 218000
     },
     {
-      "epoch": 5.024408215897577,
-      "grad_norm": 2.3339779376983643,
-      "learning_rate": 2.364865257922369e-05,
-      "loss": 0.6094,
+      "epoch": 2.51220926801959,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.364570759587557e-05,
+      "loss": 0.9871754455566406,
       "step": 218200
     },
     {
-      "epoch": 5.029013539651838,
-      "grad_norm": 3.4080960750579834,
-      "learning_rate": 2.364138101540117e-05,
-      "loss": 0.6022,
+      "epoch": 2.514511953094305,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3638436032053054e-05,
+      "loss": 0.9758554840087891,
       "step": 218400
     },
     {
-      "epoch": 5.033618863406097,
-      "grad_norm": 3.769949197769165,
-      "learning_rate": 2.3634109451578657e-05,
-      "loss": 0.6158,
+      "epoch": 2.51681463816902,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3631164468230536e-05,
+      "loss": 0.9900711822509766,
       "step": 218600
     },
     {
-      "epoch": 5.038224187160357,
-      "grad_norm": 3.2221176624298096,
-      "learning_rate": 2.362683788775614e-05,
-      "loss": 0.6077,
+      "epoch": 2.519117323243735,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3623892904408022e-05,
+      "loss": 0.985194091796875,
       "step": 218800
     },
     {
-      "epoch": 5.0428295109146175,
-      "grad_norm": 2.9638614654541016,
-      "learning_rate": 2.3619566323933626e-05,
-      "loss": 0.6134,
+      "epoch": 2.52142000831845,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3616621340585508e-05,
+      "loss": 0.9648825073242188,
       "step": 219000
     },
     {
-      "epoch": 5.047434834668877,
-      "grad_norm": 3.3254809379577637,
-      "learning_rate": 2.3612294760111112e-05,
-      "loss": 0.6121,
+      "epoch": 2.523722693393165,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.360934977676299e-05,
+      "loss": 0.9976405334472657,
       "step": 219200
     },
     {
-      "epoch": 5.052040158423138,
-      "grad_norm": 3.3986082077026367,
-      "learning_rate": 2.3605023196288594e-05,
-      "loss": 0.6231,
+      "epoch": 2.52602537846788,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.3602078212940477e-05,
+      "loss": 0.9568039703369141,
       "step": 219400
     },
     {
-      "epoch": 5.056645482177397,
-      "grad_norm": 3.187117099761963,
-      "learning_rate": 2.359775163246608e-05,
-      "loss": 0.6075,
+      "epoch": 2.5283280635425944,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.3594806649117963e-05,
+      "loss": 1.014922103881836,
       "step": 219600
     },
     {
-      "epoch": 5.061250805931657,
-      "grad_norm": 3.4964001178741455,
-      "learning_rate": 2.3590480068643563e-05,
-      "loss": 0.6022,
+      "epoch": 2.5306307486173094,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3587535085295442e-05,
+      "loss": 0.9895259857177734,
       "step": 219800
     },
     {
-      "epoch": 5.065856129685917,
-      "grad_norm": 2.966295003890991,
-      "learning_rate": 2.3583208504821045e-05,
-      "loss": 0.6103,
+      "epoch": 2.5329334336920244,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3580263521472928e-05,
+      "loss": 0.9885545349121094,
       "step": 220000
     },
     {
-      "epoch": 5.070461453440177,
-      "grad_norm": 2.458843946456909,
-      "learning_rate": 2.357593694099853e-05,
-      "loss": 0.6164,
+      "epoch": 2.5352361187667394,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3572991957650414e-05,
+      "loss": 0.9912796783447265,
       "step": 220200
     },
     {
-      "epoch": 5.075066777194436,
-      "grad_norm": 2.4280691146850586,
-      "learning_rate": 2.3568665377176017e-05,
-      "loss": 0.6103,
+      "epoch": 2.5375388038414544,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.3565720393827896e-05,
+      "loss": 0.9948096466064453,
       "step": 220400
     },
     {
-      "epoch": 5.079672100948697,
-      "grad_norm": 2.904318332672119,
-      "learning_rate": 2.35613938133535e-05,
-      "loss": 0.6018,
+      "epoch": 2.5398414889161693,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3558448830005382e-05,
+      "loss": 0.9950077056884765,
       "step": 220600
     },
     {
-      "epoch": 5.0842774247029565,
-      "grad_norm": 3.2932612895965576,
-      "learning_rate": 2.3554122249530986e-05,
-      "loss": 0.6065,
+      "epoch": 2.5421441739908843,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.3551177266182868e-05,
+      "loss": 0.9848309326171875,
       "step": 220800
     },
     {
-      "epoch": 5.088882748457216,
-      "grad_norm": 3.404580593109131,
-      "learning_rate": 2.3546850685708468e-05,
-      "loss": 0.6202,
+      "epoch": 2.5444468590655993,
+      "grad_norm": 1.9921875,
+      "learning_rate": 2.354390570236035e-05,
+      "loss": 0.9799929809570312,
       "step": 221000
     },
     {
-      "epoch": 5.093488072211477,
-      "grad_norm": 3.106785535812378,
-      "learning_rate": 2.3539579121885954e-05,
-      "loss": 0.6062,
+      "epoch": 2.5467495441403143,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.3536634138537833e-05,
+      "loss": 0.9768532562255859,
       "step": 221200
     },
     {
-      "epoch": 5.098093395965736,
-      "grad_norm": 3.391956329345703,
-      "learning_rate": 2.3532307558063437e-05,
-      "loss": 0.6091,
+      "epoch": 2.549052229215029,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.352936257471532e-05,
+      "loss": 0.9896360015869141,
       "step": 221400
     },
     {
-      "epoch": 5.102698719719997,
-      "grad_norm": 2.8422746658325195,
-      "learning_rate": 2.3525035994240922e-05,
-      "loss": 0.6021,
+      "epoch": 2.5513549142897443,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.3522091010892805e-05,
+      "loss": 0.9728896331787109,
       "step": 221600
     },
     {
-      "epoch": 5.107304043474256,
-      "grad_norm": 2.9199986457824707,
-      "learning_rate": 2.351776443041841e-05,
-      "loss": 0.6146,
+      "epoch": 2.553657599364459,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3514819447070287e-05,
+      "loss": 1.011695098876953,
       "step": 221800
     },
     {
-      "epoch": 5.111909367228516,
-      "grad_norm": 3.0510222911834717,
-      "learning_rate": 2.351049286659589e-05,
-      "loss": 0.6042,
+      "epoch": 2.555960284439174,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.3507547883247773e-05,
+      "loss": 1.0155657958984374,
       "step": 222000
     },
     {
-      "epoch": 5.116514690982776,
-      "grad_norm": 4.642153263092041,
-      "learning_rate": 2.350325766059249e-05,
-      "loss": 0.6086,
+      "epoch": 2.5582629695138888,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.3500276319425256e-05,
+      "loss": 0.9898668670654297,
       "step": 222200
     },
     {
-      "epoch": 5.121120014737036,
-      "grad_norm": 4.9221086502075195,
-      "learning_rate": 2.3495986096769972e-05,
-      "loss": 0.6069,
+      "epoch": 2.5605656545886037,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.3493004755602738e-05,
+      "loss": 0.9848436737060546,
       "step": 222400
     },
     {
-      "epoch": 5.1257253384912955,
-      "grad_norm": 3.0352025032043457,
-      "learning_rate": 2.3488714532947455e-05,
-      "loss": 0.6003,
+      "epoch": 2.5628683396633187,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.3485733191780224e-05,
+      "loss": 0.9848309326171875,
       "step": 222600
     },
     {
-      "epoch": 5.130330662245556,
-      "grad_norm": 3.675919771194458,
-      "learning_rate": 2.348144296912494e-05,
-      "loss": 0.6157,
+      "epoch": 2.5651710247380337,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.347846162795771e-05,
+      "loss": 0.9747190856933594,
       "step": 222800
     },
     {
-      "epoch": 5.134935985999816,
-      "grad_norm": 3.621706247329712,
-      "learning_rate": 2.3474171405302426e-05,
-      "loss": 0.6081,
+      "epoch": 2.5674737098127487,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3471190064135193e-05,
+      "loss": 0.9759510803222656,
       "step": 223000
     },
     {
-      "epoch": 5.139541309754076,
-      "grad_norm": 3.0481488704681396,
-      "learning_rate": 2.346689984147991e-05,
-      "loss": 0.6155,
+      "epoch": 2.5697763948874632,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.346391850031268e-05,
+      "loss": 0.9792971038818359,
       "step": 223200
     },
     {
-      "epoch": 5.144146633508336,
-      "grad_norm": 3.5414528846740723,
-      "learning_rate": 2.3459628277657395e-05,
-      "loss": 0.6054,
+      "epoch": 2.5720790799621787,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3456646936490164e-05,
+      "loss": 0.9957722473144531,
       "step": 223400
     },
     {
-      "epoch": 5.148751957262595,
-      "grad_norm": 3.2819435596466064,
-      "learning_rate": 2.345235671383488e-05,
-      "loss": 0.6131,
+      "epoch": 2.574381765036893,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.3449375372667644e-05,
+      "loss": 0.9673760223388672,
       "step": 223600
     },
     {
-      "epoch": 5.153357281016856,
-      "grad_norm": 2.693819761276245,
-      "learning_rate": 2.344508515001236e-05,
-      "loss": 0.6082,
+      "epoch": 2.576684450111608,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.344210380884513e-05,
+      "loss": 1.0127267456054687,
       "step": 223800
     },
     {
-      "epoch": 5.157962604771115,
-      "grad_norm": 2.757230520248413,
-      "learning_rate": 2.3437813586189846e-05,
-      "loss": 0.6084,
+      "epoch": 2.578987135186323,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3434832245022615e-05,
+      "loss": 0.9820223236083985,
       "step": 224000
     },
     {
-      "epoch": 5.162567928525375,
-      "grad_norm": 3.3219823837280273,
-      "learning_rate": 2.343054202236733e-05,
-      "loss": 0.6068,
+      "epoch": 2.581289820261038,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.34275606812001e-05,
+      "loss": 0.9654301452636719,
       "step": 224200
     },
     {
-      "epoch": 5.167173252279635,
-      "grad_norm": 2.8286936283111572,
-      "learning_rate": 2.3423270458544814e-05,
-      "loss": 0.6043,
+      "epoch": 2.583592505335753,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.3420289117377584e-05,
+      "loss": 0.9646720123291016,
       "step": 224400
     },
     {
-      "epoch": 5.171778576033895,
-      "grad_norm": 3.0710113048553467,
-      "learning_rate": 2.34159988947223e-05,
-      "loss": 0.6132,
+      "epoch": 2.585895190410468,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.341301755355507e-05,
+      "loss": 0.9920122528076172,
       "step": 224600
     },
     {
-      "epoch": 5.1763838997881555,
-      "grad_norm": 3.501431465148926,
-      "learning_rate": 2.3408727330899786e-05,
-      "loss": 0.6119,
+      "epoch": 2.588197875485183,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.3405745989732552e-05,
+      "loss": 0.9855150604248046,
       "step": 224800
     },
     {
-      "epoch": 5.180989223542415,
-      "grad_norm": 2.9645442962646484,
-      "learning_rate": 2.3401455767077265e-05,
-      "loss": 0.6027,
+      "epoch": 2.5905005605598976,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3398474425910035e-05,
+      "loss": 0.9884321594238281,
       "step": 225000
     },
     {
-      "epoch": 5.185594547296675,
-      "grad_norm": 3.057513475418091,
-      "learning_rate": 2.3394220561073864e-05,
-      "loss": 0.6105,
+      "epoch": 2.592803245634613,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.339120286208752e-05,
+      "loss": 0.9909938049316406,
       "step": 225200
     },
     {
-      "epoch": 5.190199871050935,
-      "grad_norm": 3.0226759910583496,
-      "learning_rate": 2.338694899725135e-05,
-      "loss": 0.6052,
+      "epoch": 2.5951059307093276,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.3383931298265007e-05,
+      "loss": 0.9761519622802735,
       "step": 225400
     },
     {
-      "epoch": 5.194805194805195,
-      "grad_norm": 2.725600481033325,
-      "learning_rate": 2.3379677433428832e-05,
-      "loss": 0.6066,
+      "epoch": 2.5974086157840426,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.337665973444249e-05,
+      "loss": 0.9695050048828125,
       "step": 225600
     },
     {
-      "epoch": 5.199410518559454,
-      "grad_norm": 3.8302340507507324,
-      "learning_rate": 2.3372405869606318e-05,
-      "loss": 0.6098,
+      "epoch": 2.5997113008587576,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3369388170619975e-05,
+      "loss": 0.9951475524902343,
       "step": 225800
     },
     {
-      "epoch": 5.204015842313715,
-      "grad_norm": 2.8420770168304443,
-      "learning_rate": 2.3365134305783804e-05,
-      "loss": 0.6164,
+      "epoch": 2.6020139859334726,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.336211660679746e-05,
+      "loss": 0.9993846893310547,
       "step": 226000
     },
     {
-      "epoch": 5.208621166067974,
-      "grad_norm": 3.4650795459747314,
-      "learning_rate": 2.3357862741961287e-05,
-      "loss": 0.6102,
+      "epoch": 2.6043166710081875,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.335484504297494e-05,
+      "loss": 0.9837513732910156,
       "step": 226200
     },
     {
-      "epoch": 5.213226489822235,
-      "grad_norm": 3.013132095336914,
-      "learning_rate": 2.3350591178138772e-05,
-      "loss": 0.6133,
+      "epoch": 2.6066193560829025,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.3347573479152426e-05,
+      "loss": 0.9643393707275391,
       "step": 226400
     },
     {
-      "epoch": 5.2178318135764945,
-      "grad_norm": 3.517982244491577,
-      "learning_rate": 2.3343319614316255e-05,
-      "loss": 0.6042,
+      "epoch": 2.6089220411576175,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3340301915329912e-05,
+      "loss": 0.9859303283691406,
       "step": 226600
     },
     {
-      "epoch": 5.222437137330754,
-      "grad_norm": 3.570617914199829,
-      "learning_rate": 2.3336048050493737e-05,
-      "loss": 0.6194,
+      "epoch": 2.611224726232332,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3333030351507394e-05,
+      "loss": 0.9887308502197265,
       "step": 226800
     },
     {
-      "epoch": 5.227042461085015,
-      "grad_norm": 3.553957223892212,
-      "learning_rate": 2.3328776486671223e-05,
-      "loss": 0.6028,
+      "epoch": 2.6135274113070475,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.332575878768488e-05,
+      "loss": 0.9941259002685547,
       "step": 227000
     },
     {
-      "epoch": 5.231647784839274,
-      "grad_norm": 3.234729051589966,
-      "learning_rate": 2.332150492284871e-05,
-      "loss": 0.6069,
+      "epoch": 2.615830096381762,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.3318487223862366e-05,
+      "loss": 0.9827572631835938,
       "step": 227200
     },
     {
-      "epoch": 5.236253108593534,
-      "grad_norm": 2.738168478012085,
-      "learning_rate": 2.3314233359026195e-05,
-      "loss": 0.6122,
+      "epoch": 2.618132781456477,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.331121566003985e-05,
+      "loss": 0.9632863616943359,
       "step": 227400
     },
     {
-      "epoch": 5.240858432347794,
-      "grad_norm": 3.4680685997009277,
-      "learning_rate": 2.3306961795203678e-05,
-      "loss": 0.6166,
+      "epoch": 2.620435466531192,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.330394409621733e-05,
+      "loss": 0.9896905517578125,
       "step": 227600
     },
     {
-      "epoch": 5.245463756102054,
-      "grad_norm": 3.349167823791504,
-      "learning_rate": 2.329969023138116e-05,
-      "loss": 0.6182,
+      "epoch": 2.622738151605907,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3296672532394817e-05,
+      "loss": 0.9736611938476563,
       "step": 227800
     },
     {
-      "epoch": 5.250069079856313,
-      "grad_norm": 2.7940080165863037,
-      "learning_rate": 2.3292418667558646e-05,
-      "loss": 0.6073,
+      "epoch": 2.625040836680622,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.3289400968572303e-05,
+      "loss": 0.991059799194336,
       "step": 228000
     },
     {
-      "epoch": 5.254674403610574,
-      "grad_norm": 3.263004779815674,
-      "learning_rate": 2.328514710373613e-05,
-      "loss": 0.6102,
+      "epoch": 2.627343521755337,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3282129404749786e-05,
+      "loss": 0.9731089019775391,
       "step": 228200
     },
     {
-      "epoch": 5.2592797273648335,
-      "grad_norm": 3.2109506130218506,
-      "learning_rate": 2.3277875539913615e-05,
-      "loss": 0.6147,
+      "epoch": 2.629646206830052,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.327485784092727e-05,
+      "loss": 0.9963571166992188,
       "step": 228400
     },
     {
-      "epoch": 5.263885051119094,
-      "grad_norm": 3.7838523387908936,
-      "learning_rate": 2.32706039760911e-05,
-      "loss": 0.61,
+      "epoch": 2.6319488919047664,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.3267586277104754e-05,
+      "loss": 0.9739570617675781,
       "step": 228600
     },
     {
-      "epoch": 5.268490374873354,
-      "grad_norm": 2.4661216735839844,
-      "learning_rate": 2.3263332412268583e-05,
-      "loss": 0.6123,
+      "epoch": 2.634251576979482,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3260314713282237e-05,
+      "loss": 0.9853155517578125,
       "step": 228800
     },
     {
-      "epoch": 5.273095698627613,
-      "grad_norm": 3.483590841293335,
-      "learning_rate": 2.325606084844607e-05,
-      "loss": 0.6215,
+      "epoch": 2.6365542620541964,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3253043149459723e-05,
+      "loss": 0.9822714233398437,
       "step": 229000
     },
     {
-      "epoch": 5.277701022381874,
-      "grad_norm": 2.8300187587738037,
-      "learning_rate": 2.324878928462355e-05,
-      "loss": 0.6036,
+      "epoch": 2.6388569471289114,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.324577158563721e-05,
+      "loss": 1.0132391357421875,
       "step": 229200
     },
     {
-      "epoch": 5.282306346136133,
-      "grad_norm": 3.444559097290039,
-      "learning_rate": 2.3241517720801034e-05,
-      "loss": 0.6136,
+      "epoch": 2.6411596322036264,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.323850002181469e-05,
+      "loss": 0.9821296691894531,
       "step": 229400
     },
     {
-      "epoch": 5.286911669890394,
-      "grad_norm": 3.6756949424743652,
-      "learning_rate": 2.323424615697852e-05,
-      "loss": 0.621,
+      "epoch": 2.6434623172783414,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.3231228457992177e-05,
+      "loss": 0.9867617797851562,
       "step": 229600
     },
     {
-      "epoch": 5.291516993644653,
-      "grad_norm": 3.188176393508911,
-      "learning_rate": 2.322701095097512e-05,
-      "loss": 0.6068,
+      "epoch": 2.6457650023530563,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.3223956894169663e-05,
+      "loss": 0.9906488037109376,
       "step": 229800
     },
     {
-      "epoch": 5.296122317398913,
-      "grad_norm": 3.3259594440460205,
-      "learning_rate": 2.32197393871526e-05,
-      "loss": 0.6159,
+      "epoch": 2.6480676874277713,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.3216685330347145e-05,
+      "loss": 0.9868508911132813,
       "step": 230000
     },
     {
-      "epoch": 5.300727641153173,
-      "grad_norm": 2.9103612899780273,
-      "learning_rate": 2.3212467823330087e-05,
-      "loss": 0.6131,
+      "epoch": 2.6503703725024863,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.3209413766524628e-05,
+      "loss": 0.9740293884277343,
       "step": 230200
     },
     {
-      "epoch": 5.305332964907433,
-      "grad_norm": 2.856694459915161,
-      "learning_rate": 2.3205196259507573e-05,
-      "loss": 0.608,
+      "epoch": 2.652673057577201,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.3202142202702114e-05,
+      "loss": 0.9676800537109375,
       "step": 230400
     },
     {
-      "epoch": 5.309938288661693,
-      "grad_norm": 3.170351982116699,
-      "learning_rate": 2.3197924695685052e-05,
-      "loss": 0.6211,
+      "epoch": 2.6549757426519163,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.31948706388796e-05,
+      "loss": 0.9770725250244141,
       "step": 230600
     },
     {
-      "epoch": 5.314543612415953,
-      "grad_norm": 2.481973648071289,
-      "learning_rate": 2.3190653131862538e-05,
-      "loss": 0.6183,
+      "epoch": 2.657278427726631,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.3187599075057082e-05,
+      "loss": 0.9891566467285157,
       "step": 230800
     },
     {
-      "epoch": 5.319148936170213,
-      "grad_norm": 2.819699287414551,
-      "learning_rate": 2.3183381568040024e-05,
-      "loss": 0.6043,
+      "epoch": 2.659581112801346,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.3180327511234568e-05,
+      "loss": 0.9634536743164063,
       "step": 231000
     },
     {
-      "epoch": 5.323754259924472,
-      "grad_norm": 2.6565167903900146,
-      "learning_rate": 2.3176110004217506e-05,
-      "loss": 0.6098,
+      "epoch": 2.6618837978760608,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.317305594741205e-05,
+      "loss": 0.9817213439941406,
       "step": 231200
     },
     {
-      "epoch": 5.328359583678733,
-      "grad_norm": 3.6145877838134766,
-      "learning_rate": 2.3168838440394992e-05,
-      "loss": 0.6108,
+      "epoch": 2.6641864829507758,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.3165784383589533e-05,
+      "loss": 0.9655717468261719,
       "step": 231400
     },
     {
-      "epoch": 5.332964907432992,
-      "grad_norm": 3.620488166809082,
-      "learning_rate": 2.3161566876572478e-05,
-      "loss": 0.6066,
+      "epoch": 2.6664891680254907,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.315851281976702e-05,
+      "loss": 1.0008053588867187,
       "step": 231600
     },
     {
-      "epoch": 5.337570231187253,
-      "grad_norm": 3.417673110961914,
-      "learning_rate": 2.3154331670569073e-05,
-      "loss": 0.6115,
+      "epoch": 2.6687918531002057,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3151241255944505e-05,
+      "loss": 1.0027259826660155,
       "step": 231800
     },
     {
-      "epoch": 5.342175554941512,
-      "grad_norm": 3.6562862396240234,
-      "learning_rate": 2.3147060106746556e-05,
-      "loss": 0.6219,
+      "epoch": 2.6710945381749207,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3143969692121988e-05,
+      "loss": 0.9779088592529297,
       "step": 232000
     },
     {
-      "epoch": 5.346780878695772,
-      "grad_norm": 3.1641769409179688,
-      "learning_rate": 2.3139788542924042e-05,
-      "loss": 0.5932,
+      "epoch": 2.6733972232496357,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3136698128299473e-05,
+      "loss": 0.9831785583496093,
       "step": 232200
     },
     {
-      "epoch": 5.3513862024500325,
-      "grad_norm": 3.7529780864715576,
-      "learning_rate": 2.3132516979101524e-05,
-      "loss": 0.6114,
+      "epoch": 2.6756999083243507,
+      "grad_norm": 0.5,
+      "learning_rate": 2.3129426564476956e-05,
+      "loss": 0.975287094116211,
       "step": 232400
     },
     {
-      "epoch": 5.355991526204292,
-      "grad_norm": 4.237635612487793,
-      "learning_rate": 2.3125281773098123e-05,
-      "loss": 0.6147,
+      "epoch": 2.678002593399065,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.312215500065444e-05,
+      "loss": 1.011480178833008,
       "step": 232600
     },
     {
-      "epoch": 5.360596849958552,
-      "grad_norm": 3.604637861251831,
-      "learning_rate": 2.311801020927561e-05,
-      "loss": 0.6077,
+      "epoch": 2.68030527847378,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.3114883436831924e-05,
+      "loss": 0.9821878814697266,
       "step": 232800
     },
     {
-      "epoch": 5.365202173712812,
-      "grad_norm": 3.4843764305114746,
-      "learning_rate": 2.311073864545309e-05,
-      "loss": 0.6114,
+      "epoch": 2.682607963548495,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.310761187300941e-05,
+      "loss": 0.9614496612548828,
       "step": 233000
     },
     {
-      "epoch": 5.369807497467072,
-      "grad_norm": 2.583153486251831,
-      "learning_rate": 2.3103467081630577e-05,
-      "loss": 0.5862,
+      "epoch": 2.68491064862321,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.3100340309186896e-05,
+      "loss": 0.9810744476318359,
       "step": 233200
     },
     {
-      "epoch": 5.374412821221332,
-      "grad_norm": 2.859898328781128,
-      "learning_rate": 2.3096195517808063e-05,
-      "loss": 0.6113,
+      "epoch": 2.687213333697925,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.309306874536438e-05,
+      "loss": 0.9930284881591797,
       "step": 233400
     },
     {
-      "epoch": 5.379018144975592,
-      "grad_norm": 3.472050666809082,
-      "learning_rate": 2.3088923953985542e-05,
-      "loss": 0.6087,
+      "epoch": 2.68951601877264,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.3085797181541865e-05,
+      "loss": 0.9711058807373046,
       "step": 233600
     },
     {
-      "epoch": 5.383623468729851,
-      "grad_norm": 3.197916030883789,
-      "learning_rate": 2.308168874798214e-05,
-      "loss": 0.602,
+      "epoch": 2.691818703847355,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.3078525617719347e-05,
+      "loss": 0.9531632232666015,
       "step": 233800
     },
     {
-      "epoch": 5.388228792484112,
-      "grad_norm": 3.002330780029297,
-      "learning_rate": 2.3074417184159624e-05,
-      "loss": 0.6079,
+      "epoch": 2.69412138892207,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.307125405389683e-05,
+      "loss": 0.9774398040771485,
       "step": 234000
     },
     {
-      "epoch": 5.3928341162383715,
-      "grad_norm": 3.3833117485046387,
-      "learning_rate": 2.306714562033711e-05,
-      "loss": 0.6113,
+      "epoch": 2.696424073996785,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3063982490074316e-05,
+      "loss": 0.9752250671386719,
       "step": 234200
     },
     {
-      "epoch": 5.397439439992631,
-      "grad_norm": 2.8529210090637207,
-      "learning_rate": 2.3059874056514595e-05,
-      "loss": 0.6106,
+      "epoch": 2.6987267590714996,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.30567109262518e-05,
+      "loss": 0.9602703857421875,
       "step": 234400
     },
     {
-      "epoch": 5.402044763746892,
-      "grad_norm": 3.600402593612671,
-      "learning_rate": 2.305260249269208e-05,
-      "loss": 0.6042,
+      "epoch": 2.7010294441462146,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3049439362429284e-05,
+      "loss": 0.9857449340820312,
       "step": 234600
     },
     {
-      "epoch": 5.406650087501151,
-      "grad_norm": 2.895305633544922,
-      "learning_rate": 2.3045330928869564e-05,
-      "loss": 0.6136,
+      "epoch": 2.7033321292209296,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.304216779860677e-05,
+      "loss": 0.9743757629394532,
       "step": 234800
     },
     {
-      "epoch": 5.411255411255412,
-      "grad_norm": 2.833522319793701,
-      "learning_rate": 2.3038059365047046e-05,
-      "loss": 0.6148,
+      "epoch": 2.7056348142956446,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.3034896234784253e-05,
+      "loss": 0.9862078857421875,
       "step": 235000
     },
     {
-      "epoch": 5.415860735009671,
-      "grad_norm": 3.2785439491271973,
-      "learning_rate": 2.3030787801224532e-05,
-      "loss": 0.5955,
+      "epoch": 2.7079374993703595,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3027624670961735e-05,
+      "loss": 1.021921615600586,
       "step": 235200
     },
     {
-      "epoch": 5.420466058763931,
-      "grad_norm": 2.5901570320129395,
-      "learning_rate": 2.3023516237402015e-05,
-      "loss": 0.6243,
+      "epoch": 2.7102401844450745,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.302035310713922e-05,
+      "loss": 0.9676741790771485,
       "step": 235400
     },
     {
-      "epoch": 5.425071382518191,
-      "grad_norm": 2.593280076980591,
-      "learning_rate": 2.30162446735795e-05,
-      "loss": 0.6095,
+      "epoch": 2.7125428695197895,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.3013081543316707e-05,
+      "loss": 0.96970947265625,
       "step": 235600
     },
     {
-      "epoch": 5.429676706272451,
-      "grad_norm": 2.4974257946014404,
-      "learning_rate": 2.3008973109756987e-05,
-      "loss": 0.6196,
+      "epoch": 2.7148455545945045,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.3005809979494193e-05,
+      "loss": 0.9862263488769532,
       "step": 235800
     },
     {
-      "epoch": 5.4342820300267105,
-      "grad_norm": 2.7232484817504883,
-      "learning_rate": 2.3001737903753582e-05,
-      "loss": 0.6088,
+      "epoch": 2.7171482396692195,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.2998538415671675e-05,
+      "loss": 0.9895063781738281,
       "step": 236000
     },
     {
-      "epoch": 5.438887353780971,
-      "grad_norm": 3.5613925457000732,
-      "learning_rate": 2.2994466339931068e-05,
-      "loss": 0.6062,
+      "epoch": 2.719450924743934,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.299126685184916e-05,
+      "loss": 0.9968424987792969,
       "step": 236200
     },
     {
-      "epoch": 5.443492677535231,
-      "grad_norm": 3.134225368499756,
-      "learning_rate": 2.2987194776108554e-05,
-      "loss": 0.6045,
+      "epoch": 2.721753609818649,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.2983995288026644e-05,
+      "loss": 0.9946724700927735,
       "step": 236400
     },
     {
-      "epoch": 5.448098001289491,
-      "grad_norm": 3.1845412254333496,
-      "learning_rate": 2.2979923212286033e-05,
-      "loss": 0.6135,
+      "epoch": 2.724056294893364,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2976723724204126e-05,
+      "loss": 0.9951986694335937,
       "step": 236600
     },
     {
-      "epoch": 5.452703325043751,
-      "grad_norm": 3.119331121444702,
-      "learning_rate": 2.297265164846352e-05,
-      "loss": 0.6089,
+      "epoch": 2.726358979968079,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.2969452160381612e-05,
+      "loss": 1.0003690338134765,
       "step": 236800
     },
     {
-      "epoch": 5.45730864879801,
-      "grad_norm": 3.175262928009033,
-      "learning_rate": 2.2965380084641005e-05,
-      "loss": 0.6185,
+      "epoch": 2.728661665042794,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2962180596559098e-05,
+      "loss": 1.0061814880371094,
       "step": 237000
     },
     {
-      "epoch": 5.461913972552271,
-      "grad_norm": 2.737478733062744,
-      "learning_rate": 2.2958108520818487e-05,
-      "loss": 0.6136,
+      "epoch": 2.730964350117509,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.295490903273658e-05,
+      "loss": 0.9780216217041016,
       "step": 237200
     },
     {
-      "epoch": 5.46651929630653,
-      "grad_norm": 3.8264358043670654,
-      "learning_rate": 2.2950836956995973e-05,
-      "loss": 0.5989,
+      "epoch": 2.733267035192224,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2947637468914067e-05,
+      "loss": 0.9820411682128907,
       "step": 237400
     },
     {
-      "epoch": 5.47112462006079,
-      "grad_norm": 2.678365468978882,
-      "learning_rate": 2.294356539317346e-05,
-      "loss": 0.6062,
+      "epoch": 2.735569720266939,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.294036590509155e-05,
+      "loss": 0.9854095458984375,
       "step": 237600
     },
     {
-      "epoch": 5.47572994381505,
-      "grad_norm": 3.2151529788970947,
-      "learning_rate": 2.2936293829350938e-05,
-      "loss": 0.6091,
+      "epoch": 2.737872405341654,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.293309434126903e-05,
+      "loss": 0.9723403167724609,
       "step": 237800
     },
     {
-      "epoch": 5.48033526756931,
-      "grad_norm": 2.8425002098083496,
-      "learning_rate": 2.2929022265528424e-05,
-      "loss": 0.6134,
+      "epoch": 2.7401750904163684,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.2925822777446518e-05,
+      "loss": 0.9777320098876953,
       "step": 238000
     },
     {
-      "epoch": 5.48494059132357,
-      "grad_norm": 2.8937878608703613,
-      "learning_rate": 2.292175070170591e-05,
-      "loss": 0.6041,
+      "epoch": 2.742477775491084,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2918551213624003e-05,
+      "loss": 0.9932432556152344,
       "step": 238200
     },
     {
-      "epoch": 5.48954591507783,
-      "grad_norm": 3.9438226222991943,
-      "learning_rate": 2.2914479137883392e-05,
-      "loss": 0.6061,
+      "epoch": 2.7447804605657984,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.2911279649801486e-05,
+      "loss": 0.9813819122314453,
       "step": 238400
     },
     {
-      "epoch": 5.49415123883209,
-      "grad_norm": 2.902695655822754,
-      "learning_rate": 2.290720757406088e-05,
-      "loss": 0.6239,
+      "epoch": 2.7470831456405134,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.2904008085978972e-05,
+      "loss": 0.9843547821044922,
       "step": 238600
     },
     {
-      "epoch": 5.49875656258635,
-      "grad_norm": 2.9250173568725586,
-      "learning_rate": 2.2899936010238364e-05,
-      "loss": 0.6094,
+      "epoch": 2.7493858307152284,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.2896736522156454e-05,
+      "loss": 0.9740335083007813,
       "step": 238800
     },
     {
-      "epoch": 5.50336188634061,
-      "grad_norm": 3.668975591659546,
-      "learning_rate": 2.2892664446415847e-05,
-      "loss": 0.61,
+      "epoch": 2.7516885157899433,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.288946495833394e-05,
+      "loss": 0.9769838714599609,
       "step": 239000
     },
     {
-      "epoch": 5.507967210094869,
-      "grad_norm": 3.020686626434326,
-      "learning_rate": 2.288539288259333e-05,
-      "loss": 0.621,
+      "epoch": 2.7539912008646583,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.2882193394511423e-05,
+      "loss": 0.9714900970458984,
       "step": 239200
     },
     {
-      "epoch": 5.51257253384913,
-      "grad_norm": 3.6001780033111572,
-      "learning_rate": 2.2878121318770815e-05,
-      "loss": 0.6037,
+      "epoch": 2.7562938859393733,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.287492183068891e-05,
+      "loss": 0.9776052093505859,
       "step": 239400
     },
     {
-      "epoch": 5.517177857603389,
-      "grad_norm": 2.8009440898895264,
-      "learning_rate": 2.28708497549483e-05,
-      "loss": 0.6158,
+      "epoch": 2.7585965710140883,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.2867650266866395e-05,
+      "loss": 0.9773331451416015,
       "step": 239600
     },
     {
-      "epoch": 5.52178318135765,
-      "grad_norm": 2.9138927459716797,
-      "learning_rate": 2.2863578191125784e-05,
-      "loss": 0.6091,
+      "epoch": 2.760899256088803,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.2860378703043877e-05,
+      "loss": 0.9772712707519531,
       "step": 239800
     },
     {
-      "epoch": 5.5263885051119095,
-      "grad_norm": 3.2366840839385986,
-      "learning_rate": 2.285630662730327e-05,
-      "loss": 0.61,
+      "epoch": 2.7632019411635182,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2853107139221363e-05,
+      "loss": 0.9649584197998047,
       "step": 240000
     },
     {
-      "epoch": 5.530993828866169,
-      "grad_norm": 3.211121082305908,
-      "learning_rate": 2.2849071421299865e-05,
-      "loss": 0.6203,
+      "epoch": 2.765504626238233,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.2845835575398846e-05,
+      "loss": 0.966458969116211,
       "step": 240200
     },
     {
-      "epoch": 5.53559915262043,
-      "grad_norm": 2.9791855812072754,
-      "learning_rate": 2.284179985747735e-05,
-      "loss": 0.611,
+      "epoch": 2.7678073113129478,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.2838564011576328e-05,
+      "loss": 0.986024398803711,
       "step": 240400
     },
     {
-      "epoch": 5.540204476374689,
-      "grad_norm": 2.9468283653259277,
-      "learning_rate": 2.2834528293654833e-05,
-      "loss": 0.5963,
+      "epoch": 2.7701099963876628,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.2831292447753814e-05,
+      "loss": 0.9941361236572266,
       "step": 240600
     },
     {
-      "epoch": 5.544809800128949,
-      "grad_norm": 2.6335391998291016,
-      "learning_rate": 2.282725672983232e-05,
-      "loss": 0.5982,
+      "epoch": 2.7724126814623777,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.28240208839313e-05,
+      "loss": 0.9598740386962891,
       "step": 240800
     },
     {
-      "epoch": 5.549415123883209,
-      "grad_norm": 3.262544870376587,
-      "learning_rate": 2.28199851660098e-05,
-      "loss": 0.6005,
+      "epoch": 2.7747153665370927,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.2816749320108783e-05,
+      "loss": 0.9616059875488281,
       "step": 241000
     },
     {
-      "epoch": 5.554020447637469,
-      "grad_norm": 2.9877734184265137,
-      "learning_rate": 2.28127499600064e-05,
-      "loss": 0.6062,
+      "epoch": 2.7770180516118077,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.280947775628627e-05,
+      "loss": 0.9962806701660156,
       "step": 241200
     },
     {
-      "epoch": 5.558625771391728,
-      "grad_norm": 3.008165121078491,
-      "learning_rate": 2.2805478396183883e-05,
-      "loss": 0.6056,
+      "epoch": 2.7793207366865227,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.280220619246375e-05,
+      "loss": 0.9695357513427735,
       "step": 241400
     },
     {
-      "epoch": 5.563231095145989,
-      "grad_norm": 2.6136744022369385,
-      "learning_rate": 2.279820683236137e-05,
-      "loss": 0.604,
+      "epoch": 2.7816234217612372,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2794934628641237e-05,
+      "loss": 0.9759367370605468,
       "step": 241600
     },
     {
-      "epoch": 5.5678364189002485,
-      "grad_norm": 3.0362038612365723,
-      "learning_rate": 2.2790935268538855e-05,
-      "loss": 0.6016,
+      "epoch": 2.7839261068359527,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.278766306481872e-05,
+      "loss": 0.9805625915527344,
       "step": 241800
     },
     {
-      "epoch": 5.572441742654509,
-      "grad_norm": 3.641286849975586,
-      "learning_rate": 2.2783663704716334e-05,
-      "loss": 0.6089,
+      "epoch": 2.786228791910667,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.2780391500996205e-05,
+      "loss": 0.9758683776855469,
       "step": 242000
     },
     {
-      "epoch": 5.577047066408769,
-      "grad_norm": 3.9004762172698975,
-      "learning_rate": 2.277639214089382e-05,
-      "loss": 0.6049,
+      "epoch": 2.788531476985382,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.277311993717369e-05,
+      "loss": 0.9969324493408203,
       "step": 242200
     },
     {
-      "epoch": 5.581652390163028,
-      "grad_norm": 3.441751003265381,
-      "learning_rate": 2.2769120577071306e-05,
-      "loss": 0.6185,
+      "epoch": 2.790834162060097,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.2765848373351174e-05,
+      "loss": 0.9669373321533203,
       "step": 242400
     },
     {
-      "epoch": 5.586257713917289,
-      "grad_norm": 2.6894123554229736,
-      "learning_rate": 2.276184901324879e-05,
-      "loss": 0.614,
+      "epoch": 2.793136847134812,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.2758576809528656e-05,
+      "loss": 0.9705430603027344,
       "step": 242600
     },
     {
-      "epoch": 5.590863037671548,
-      "grad_norm": 3.961691379547119,
-      "learning_rate": 2.2754577449426274e-05,
-      "loss": 0.6081,
+      "epoch": 2.795439532209527,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.2751305245706142e-05,
+      "loss": 0.9699090576171875,
       "step": 242800
     },
     {
-      "epoch": 5.595468361425809,
-      "grad_norm": 4.229228496551514,
-      "learning_rate": 2.274730588560376e-05,
-      "loss": 0.6125,
+      "epoch": 2.797742217284242,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2744033681883625e-05,
+      "loss": 0.998017807006836,
       "step": 243000
     },
     {
-      "epoch": 5.600073685180068,
-      "grad_norm": 2.8456718921661377,
-      "learning_rate": 2.2740034321781242e-05,
-      "loss": 0.6013,
+      "epoch": 2.800044902358957,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.273676211806111e-05,
+      "loss": 0.9731108093261719,
       "step": 243200
     },
     {
-      "epoch": 5.604679008934328,
-      "grad_norm": 2.563215494155884,
-      "learning_rate": 2.2732762757958725e-05,
-      "loss": 0.6055,
+      "epoch": 2.8023475874336716,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.2729490554238597e-05,
+      "loss": 0.9816190338134766,
       "step": 243400
     },
     {
-      "epoch": 5.609284332688588,
-      "grad_norm": 3.1517744064331055,
-      "learning_rate": 2.272549119413621e-05,
-      "loss": 0.598,
+      "epoch": 2.804650272508387,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.272221899041608e-05,
+      "loss": 0.9914408874511719,
       "step": 243600
     },
     {
-      "epoch": 5.613889656442848,
-      "grad_norm": 3.084362030029297,
-      "learning_rate": 2.2718219630313697e-05,
-      "loss": 0.6114,
+      "epoch": 2.8069529575831016,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2714947426593565e-05,
+      "loss": 0.9849854278564453,
       "step": 243800
     },
     {
-      "epoch": 5.618494980197108,
-      "grad_norm": 2.9705121517181396,
-      "learning_rate": 2.271094806649118e-05,
-      "loss": 0.6105,
+      "epoch": 2.8092556426578166,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.2707675862771048e-05,
+      "loss": 1.00310791015625,
       "step": 244000
     },
     {
-      "epoch": 5.623100303951368,
-      "grad_norm": 2.46872615814209,
-      "learning_rate": 2.2703676502668665e-05,
-      "loss": 0.6021,
+      "epoch": 2.8115583277325316,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.270040429894853e-05,
+      "loss": 0.9826456451416016,
       "step": 244200
     },
     {
-      "epoch": 5.627705627705628,
-      "grad_norm": 3.1614882946014404,
-      "learning_rate": 2.269640493884615e-05,
-      "loss": 0.6155,
+      "epoch": 2.8138610128072465,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.2693132735126016e-05,
+      "loss": 0.9897075653076172,
       "step": 244400
     },
     {
-      "epoch": 5.632310951459887,
-      "grad_norm": 3.2688674926757812,
-      "learning_rate": 2.268913337502363e-05,
-      "loss": 0.617,
+      "epoch": 2.8161636978819615,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2685861171303502e-05,
+      "loss": 0.9984790802001953,
       "step": 244600
     },
     {
-      "epoch": 5.636916275214148,
-      "grad_norm": 2.7544074058532715,
-      "learning_rate": 2.2681861811201116e-05,
-      "loss": 0.6218,
+      "epoch": 2.8184663829566765,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.2678589607480988e-05,
+      "loss": 0.9899990081787109,
       "step": 244800
     },
     {
-      "epoch": 5.641521598968407,
-      "grad_norm": 3.150360345840454,
-      "learning_rate": 2.2674590247378602e-05,
-      "loss": 0.6118,
+      "epoch": 2.8207690680313915,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.267131804365847e-05,
+      "loss": 0.9996802520751953,
       "step": 245000
     },
     {
-      "epoch": 5.646126922722667,
-      "grad_norm": 3.299985647201538,
-      "learning_rate": 2.2667318683556088e-05,
-      "loss": 0.6018,
+      "epoch": 2.823071753106106,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.2664046479835953e-05,
+      "loss": 0.9771870422363281,
       "step": 245200
     },
     {
-      "epoch": 5.650732246476927,
-      "grad_norm": 3.0777273178100586,
-      "learning_rate": 2.266004711973357e-05,
-      "loss": 0.6136,
+      "epoch": 2.8253744381808215,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.265677491601344e-05,
+      "loss": 0.9960012817382813,
       "step": 245400
     },
     {
-      "epoch": 5.655337570231187,
-      "grad_norm": 3.37021803855896,
-      "learning_rate": 2.2652775555911056e-05,
-      "loss": 0.6124,
+      "epoch": 2.827677123255536,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.264950335219092e-05,
+      "loss": 0.9986069488525391,
       "step": 245600
     },
     {
-      "epoch": 5.6599428939854475,
-      "grad_norm": 3.1448745727539062,
-      "learning_rate": 2.264550399208854e-05,
-      "loss": 0.6185,
+      "epoch": 2.829979808330251,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2642231788368407e-05,
+      "loss": 0.9607802581787109,
       "step": 245800
     },
     {
-      "epoch": 5.664548217739707,
-      "grad_norm": 3.1662933826446533,
-      "learning_rate": 2.263823242826602e-05,
-      "loss": 0.6087,
+      "epoch": 2.832282493404966,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2634960224545893e-05,
+      "loss": 1.0018367767333984,
       "step": 246000
     },
     {
-      "epoch": 5.669153541493967,
-      "grad_norm": 2.839693546295166,
-      "learning_rate": 2.2630960864443507e-05,
-      "loss": 0.6074,
+      "epoch": 2.834585178479681,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.2627688660723376e-05,
+      "loss": 0.9637592315673829,
       "step": 246200
     },
     {
-      "epoch": 5.673758865248227,
-      "grad_norm": 3.3215548992156982,
-      "learning_rate": 2.2623689300620993e-05,
-      "loss": 0.6169,
+      "epoch": 2.836887863554396,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.262041709690086e-05,
+      "loss": 0.9718500518798828,
       "step": 246400
     },
     {
-      "epoch": 5.678364189002487,
-      "grad_norm": 3.0180182456970215,
-      "learning_rate": 2.2616417736798476e-05,
-      "loss": 0.6126,
+      "epoch": 2.839190548629111,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2613145533078344e-05,
+      "loss": 0.9744422912597657,
       "step": 246600
     },
     {
-      "epoch": 5.682969512756747,
-      "grad_norm": 2.791381597518921,
-      "learning_rate": 2.2609146172975962e-05,
-      "loss": 0.6059,
+      "epoch": 2.841493233703826,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.2605873969255827e-05,
+      "loss": 0.9885763549804687,
       "step": 246800
     },
     {
-      "epoch": 5.687574836511007,
-      "grad_norm": 3.137742280960083,
-      "learning_rate": 2.2601874609153448e-05,
-      "loss": 0.6082,
+      "epoch": 2.8437959187785404,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.2598602405433312e-05,
+      "loss": 0.9664224243164062,
       "step": 247000
     },
     {
-      "epoch": 5.692180160265266,
-      "grad_norm": 3.0345163345336914,
-      "learning_rate": 2.2594603045330927e-05,
-      "loss": 0.5985,
+      "epoch": 2.846098603853256,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.25913308416108e-05,
+      "loss": 1.005243453979492,
       "step": 247200
     },
     {
-      "epoch": 5.696785484019527,
-      "grad_norm": 3.281728744506836,
-      "learning_rate": 2.2587331481508413e-05,
-      "loss": 0.5944,
+      "epoch": 2.8484012889279704,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.258405927778828e-05,
+      "loss": 0.979427490234375,
       "step": 247400
     },
     {
-      "epoch": 5.7013908077737865,
-      "grad_norm": 2.6144189834594727,
-      "learning_rate": 2.258009627550501e-05,
-      "loss": 0.6033,
+      "epoch": 2.8507039740026854,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.2576787713965767e-05,
+      "loss": 0.9953087615966797,
       "step": 247600
     },
     {
-      "epoch": 5.705996131528046,
-      "grad_norm": 3.488960027694702,
-      "learning_rate": 2.2572824711682494e-05,
-      "loss": 0.6041,
+      "epoch": 2.8530066590774004,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.256951615014325e-05,
+      "loss": 0.9799519348144531,
       "step": 247800
     },
     {
-      "epoch": 5.710601455282307,
-      "grad_norm": 2.470059871673584,
-      "learning_rate": 2.256555314785998e-05,
-      "loss": 0.5939,
+      "epoch": 2.8553093441521153,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.2562244586320735e-05,
+      "loss": 0.9886309051513672,
       "step": 248000
     },
     {
-      "epoch": 5.715206779036566,
-      "grad_norm": 2.9853994846343994,
-      "learning_rate": 2.2558281584037466e-05,
-      "loss": 0.6016,
+      "epoch": 2.8576120292268303,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.2554973022498218e-05,
+      "loss": 0.9763843536376953,
       "step": 248200
     },
     {
-      "epoch": 5.719812102790826,
-      "grad_norm": 2.8882532119750977,
-      "learning_rate": 2.2551010020214948e-05,
-      "loss": 0.614,
+      "epoch": 2.8599147143015453,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.2547701458675704e-05,
+      "loss": 0.9831291198730469,
       "step": 248400
     },
     {
-      "epoch": 5.724417426545086,
-      "grad_norm": 2.4324746131896973,
-      "learning_rate": 2.254373845639243e-05,
-      "loss": 0.6169,
+      "epoch": 2.8622173993762603,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.254042989485319e-05,
+      "loss": 0.9907274627685547,
       "step": 248600
     },
     {
-      "epoch": 5.729022750299346,
-      "grad_norm": 3.518960952758789,
-      "learning_rate": 2.253650325038903e-05,
-      "loss": 0.6053,
+      "epoch": 2.8645200844509753,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.2533158331030672e-05,
+      "loss": 0.9687870788574219,
       "step": 248800
     },
     {
-      "epoch": 5.733628074053606,
-      "grad_norm": 3.073248863220215,
-      "learning_rate": 2.2529231686566512e-05,
-      "loss": 0.6129,
+      "epoch": 2.8668227695256903,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.2525886767208155e-05,
+      "loss": 0.951844482421875,
       "step": 249000
     },
     {
-      "epoch": 5.738233397807866,
-      "grad_norm": 3.165026903152466,
-      "learning_rate": 2.2521960122743998e-05,
-      "loss": 0.6069,
+      "epoch": 2.869125454600405,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.251861520338564e-05,
+      "loss": 0.9599909973144531,
       "step": 249200
     },
     {
-      "epoch": 5.7428387215621255,
-      "grad_norm": 3.6242356300354004,
-      "learning_rate": 2.2514688558921484e-05,
-      "loss": 0.617,
+      "epoch": 2.87142813967512,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.2511343639563123e-05,
+      "loss": 0.9898688507080078,
       "step": 249400
     },
     {
-      "epoch": 5.747444045316386,
-      "grad_norm": 2.7420730590820312,
-      "learning_rate": 2.2507416995098966e-05,
-      "loss": 0.6056,
+      "epoch": 2.8737308247498348,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.250407207574061e-05,
+      "loss": 0.9829541015625,
       "step": 249600
     },
     {
-      "epoch": 5.752049369070646,
-      "grad_norm": 3.044351577758789,
-      "learning_rate": 2.2500145431276452e-05,
-      "loss": 0.6047,
+      "epoch": 2.8760335098245497,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.2496800511918095e-05,
+      "loss": 0.9819578552246093,
       "step": 249800
     },
     {
-      "epoch": 5.756654692824906,
-      "grad_norm": 3.59561824798584,
-      "learning_rate": 2.2492873867453935e-05,
-      "loss": 0.5947,
+      "epoch": 2.8783361948992647,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2489528948095577e-05,
+      "loss": 1.0021160125732422,
       "step": 250000
     },
     {
-      "epoch": 5.761260016579166,
-      "grad_norm": 2.7053959369659424,
-      "learning_rate": 2.2485602303631417e-05,
-      "loss": 0.6167,
+      "epoch": 2.8806388799739797,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.2482257384273063e-05,
+      "loss": 0.9732303619384766,
       "step": 250200
     },
     {
-      "epoch": 5.765865340333425,
-      "grad_norm": 3.2808427810668945,
-      "learning_rate": 2.2478367097628016e-05,
-      "loss": 0.6096,
+      "epoch": 2.8829415650486947,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.2474985820450546e-05,
+      "loss": 0.9762882995605469,
       "step": 250400
     },
     {
-      "epoch": 5.770470664087686,
-      "grad_norm": 3.0669496059417725,
-      "learning_rate": 2.2471095533805502e-05,
-      "loss": 0.6172,
+      "epoch": 2.8852442501234097,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.2467714256628032e-05,
+      "loss": 0.9967890930175781,
       "step": 250600
     },
     {
-      "epoch": 5.775075987841945,
-      "grad_norm": 3.3072104454040527,
-      "learning_rate": 2.2463823969982984e-05,
-      "loss": 0.6153,
+      "epoch": 2.8875469351981247,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.2460442692805514e-05,
+      "loss": 0.9803618621826172,
       "step": 250800
     },
     {
-      "epoch": 5.779681311596205,
-      "grad_norm": 3.2189414501190186,
-      "learning_rate": 2.245655240616047e-05,
-      "loss": 0.6051,
+      "epoch": 2.889849620272839,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.2453171128983e-05,
+      "loss": 0.9854412078857422,
       "step": 251000
     },
     {
-      "epoch": 5.784286635350465,
-      "grad_norm": 3.9628310203552246,
-      "learning_rate": 2.2449280842337956e-05,
-      "loss": 0.6128,
+      "epoch": 2.892152305347554,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2445899565160486e-05,
+      "loss": 0.9635372161865234,
       "step": 251200
     },
     {
-      "epoch": 5.788891959104725,
-      "grad_norm": 2.860816717147827,
-      "learning_rate": 2.244200927851544e-05,
-      "loss": 0.6007,
+      "epoch": 2.894454990422269,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.243862800133797e-05,
+      "loss": 0.9794617462158203,
       "step": 251400
     },
     {
-      "epoch": 5.793497282858985,
-      "grad_norm": 2.6828348636627197,
-      "learning_rate": 2.243473771469292e-05,
-      "loss": 0.5999,
+      "epoch": 2.896757675496984,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.243135643751545e-05,
+      "loss": 0.9850530242919922,
       "step": 251600
     },
     {
-      "epoch": 5.798102606613245,
-      "grad_norm": 3.44728946685791,
-      "learning_rate": 2.2427466150870407e-05,
-      "loss": 0.6162,
+      "epoch": 2.899060360571699,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2424084873692937e-05,
+      "loss": 0.9839181518554687,
       "step": 251800
     },
     {
-      "epoch": 5.802707930367505,
-      "grad_norm": 2.5375330448150635,
-      "learning_rate": 2.242019458704789e-05,
-      "loss": 0.6101,
+      "epoch": 2.901363045646414,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.241681330987042e-05,
+      "loss": 1.0048432922363282,
       "step": 252000
     },
     {
-      "epoch": 5.807313254121765,
-      "grad_norm": 3.269181489944458,
-      "learning_rate": 2.2412923023225375e-05,
-      "loss": 0.6227,
+      "epoch": 2.903665730721129,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.2409541746047906e-05,
+      "loss": 0.9703173828125,
       "step": 252200
     },
     {
-      "epoch": 5.811918577876025,
-      "grad_norm": 3.957916498184204,
-      "learning_rate": 2.240565145940286e-05,
-      "loss": 0.6036,
+      "epoch": 2.905968415795844,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.240227018222539e-05,
+      "loss": 0.9863452911376953,
       "step": 252400
     },
     {
-      "epoch": 5.816523901630284,
-      "grad_norm": 3.4399096965789795,
-      "learning_rate": 2.2398379895580344e-05,
-      "loss": 0.6127,
+      "epoch": 2.908271100870559,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.2394998618402874e-05,
+      "loss": 0.9932093048095703,
       "step": 252600
     },
     {
-      "epoch": 5.821129225384545,
-      "grad_norm": 3.4735524654388428,
-      "learning_rate": 2.2391108331757826e-05,
-      "loss": 0.6012,
+      "epoch": 2.9105737859452736,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.2387727054580357e-05,
+      "loss": 0.9997026062011719,
       "step": 252800
     },
     {
-      "epoch": 5.825734549138804,
-      "grad_norm": 2.868939161300659,
-      "learning_rate": 2.2383836767935312e-05,
-      "loss": 0.6138,
+      "epoch": 2.9128764710199886,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.2380455490757842e-05,
+      "loss": 0.9846231079101563,
       "step": 253000
     },
     {
-      "epoch": 5.830339872893065,
-      "grad_norm": 3.8807289600372314,
-      "learning_rate": 2.2376565204112798e-05,
-      "loss": 0.6036,
+      "epoch": 2.9151791560947036,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.2373183926935325e-05,
+      "loss": 0.9628493499755859,
       "step": 253200
     },
     {
-      "epoch": 5.8349451966473245,
-      "grad_norm": 3.3366916179656982,
-      "learning_rate": 2.236929364029028e-05,
-      "loss": 0.6151,
+      "epoch": 2.9174818411694186,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.236591236311281e-05,
+      "loss": 0.9956505584716797,
       "step": 253400
     },
     {
-      "epoch": 5.839550520401584,
-      "grad_norm": 3.0559029579162598,
-      "learning_rate": 2.2362022076467767e-05,
-      "loss": 0.6081,
+      "epoch": 2.9197845262441335,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.2358640799290297e-05,
+      "loss": 0.9899761199951171,
       "step": 253600
     },
     {
-      "epoch": 5.8441558441558445,
-      "grad_norm": 2.922211170196533,
-      "learning_rate": 2.2354750512645253e-05,
-      "loss": 0.611,
+      "epoch": 2.9220872113188485,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.2351369235467783e-05,
+      "loss": 0.986265869140625,
       "step": 253800
     },
     {
-      "epoch": 5.848761167910104,
-      "grad_norm": 3.6084461212158203,
-      "learning_rate": 2.2347478948822735e-05,
-      "loss": 0.6,
+      "epoch": 2.9243898963935635,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2344097671645265e-05,
+      "loss": 0.9733154296875,
       "step": 254000
     },
     {
-      "epoch": 5.853366491664364,
-      "grad_norm": 3.366588592529297,
-      "learning_rate": 2.2340207385000218e-05,
-      "loss": 0.6079,
+      "epoch": 2.9266925814682785,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.2336826107822748e-05,
+      "loss": 0.9507086181640625,
       "step": 254200
     },
     {
-      "epoch": 5.857971815418624,
-      "grad_norm": 4.779687404632568,
-      "learning_rate": 2.2332935821177704e-05,
-      "loss": 0.5983,
+      "epoch": 2.9289952665429935,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2329554544000234e-05,
+      "loss": 0.9880396270751953,
       "step": 254400
     },
     {
-      "epoch": 5.862577139172884,
-      "grad_norm": 2.4821720123291016,
-      "learning_rate": 2.2325664257355186e-05,
-      "loss": 0.5977,
+      "epoch": 2.931297951617708,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2322282980177716e-05,
+      "loss": 0.9833847808837891,
       "step": 254600
     },
     {
-      "epoch": 5.867182462927143,
-      "grad_norm": 3.079432487487793,
-      "learning_rate": 2.2318392693532672e-05,
-      "loss": 0.6135,
+      "epoch": 2.9336006366924234,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.2315011416355202e-05,
+      "loss": 0.9833303833007813,
       "step": 254800
     },
     {
-      "epoch": 5.871787786681404,
-      "grad_norm": 2.8621842861175537,
-      "learning_rate": 2.2311121129710158e-05,
-      "loss": 0.615,
+      "epoch": 2.935903321767138,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.2307739852532688e-05,
+      "loss": 0.9785620880126953,
       "step": 255000
     },
     {
-      "epoch": 5.8763931104356635,
-      "grad_norm": 3.1682121753692627,
-      "learning_rate": 2.230384956588764e-05,
-      "loss": 0.6029,
+      "epoch": 2.938206006841853,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.230046828871017e-05,
+      "loss": 0.9954475402832031,
       "step": 255200
     },
     {
-      "epoch": 5.880998434189923,
-      "grad_norm": 3.040579080581665,
-      "learning_rate": 2.2296578002065123e-05,
-      "loss": 0.6091,
+      "epoch": 2.940508691916568,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.2293196724887653e-05,
+      "loss": 0.991054916381836,
       "step": 255400
     },
     {
-      "epoch": 5.885603757944184,
-      "grad_norm": 3.665220260620117,
-      "learning_rate": 2.228930643824261e-05,
-      "loss": 0.6065,
+      "epoch": 2.942811376991283,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.228592516106514e-05,
+      "loss": 1.0015912628173829,
       "step": 255600
     },
     {
-      "epoch": 5.890209081698443,
-      "grad_norm": 2.9831740856170654,
-      "learning_rate": 2.228203487442009e-05,
-      "loss": 0.6027,
+      "epoch": 2.945114062065998,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.227865359724262e-05,
+      "loss": 0.983523178100586,
       "step": 255800
     },
     {
-      "epoch": 5.894814405452704,
-      "grad_norm": 2.8328495025634766,
-      "learning_rate": 2.2274763310597577e-05,
-      "loss": 0.6086,
+      "epoch": 2.947416747140713,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.2271382033420107e-05,
+      "loss": 0.9683066558837891,
       "step": 256000
     },
     {
-      "epoch": 5.899419729206963,
-      "grad_norm": 3.341989755630493,
-      "learning_rate": 2.2267528104594176e-05,
-      "loss": 0.6206,
+      "epoch": 2.949719432215428,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.2264110469597593e-05,
+      "loss": 0.998266830444336,
       "step": 256200
     },
     {
-      "epoch": 5.904025052961223,
-      "grad_norm": 2.885848045349121,
-      "learning_rate": 2.226025654077166e-05,
-      "loss": 0.6136,
+      "epoch": 2.9520221172901424,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.225683890577508e-05,
+      "loss": 0.9820155334472657,
       "step": 256400
     },
     {
-      "epoch": 5.908630376715483,
-      "grad_norm": 2.7301769256591797,
-      "learning_rate": 2.2252984976949144e-05,
-      "loss": 0.6032,
+      "epoch": 2.954324802364858,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.2249567341952562e-05,
+      "loss": 1.0017244720458984,
       "step": 256600
     },
     {
-      "epoch": 5.913235700469743,
-      "grad_norm": 3.254574775695801,
-      "learning_rate": 2.2245713413126627e-05,
-      "loss": 0.6161,
+      "epoch": 2.9566274874395724,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.2242295778130044e-05,
+      "loss": 0.9865045166015625,
       "step": 256800
     },
     {
-      "epoch": 5.917841024224003,
-      "grad_norm": 3.9567196369171143,
-      "learning_rate": 2.223844184930411e-05,
-      "loss": 0.6057,
+      "epoch": 2.9589301725142874,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.223502421430753e-05,
+      "loss": 0.9958966827392578,
       "step": 257000
     },
     {
-      "epoch": 5.922446347978263,
-      "grad_norm": 2.612321615219116,
-      "learning_rate": 2.2231170285481595e-05,
-      "loss": 0.6112,
+      "epoch": 2.9612328575890023,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.2227752650485013e-05,
+      "loss": 1.001144561767578,
       "step": 257200
     },
     {
-      "epoch": 5.927051671732523,
-      "grad_norm": 2.945931911468506,
-      "learning_rate": 2.222389872165908e-05,
-      "loss": 0.6134,
+      "epoch": 2.9635355426637173,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.22204810866625e-05,
+      "loss": 0.9909696197509765,
       "step": 257400
     },
     {
-      "epoch": 5.931656995486783,
-      "grad_norm": 3.260687828063965,
-      "learning_rate": 2.2216627157836564e-05,
-      "loss": 0.5979,
+      "epoch": 2.9658382277384323,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.2213209522839985e-05,
+      "loss": 0.9749507141113282,
       "step": 257600
     },
     {
-      "epoch": 5.936262319241043,
-      "grad_norm": 2.9708778858184814,
-      "learning_rate": 2.220935559401405e-05,
-      "loss": 0.6135,
+      "epoch": 2.9681409128131473,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.2205937959017467e-05,
+      "loss": 0.9850035858154297,
       "step": 257800
     },
     {
-      "epoch": 5.940867642995302,
-      "grad_norm": 3.1571850776672363,
-      "learning_rate": 2.2202084030191536e-05,
-      "loss": 0.615,
+      "epoch": 2.9704435978878623,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.219866639519495e-05,
+      "loss": 0.99532470703125,
       "step": 258000
     },
     {
-      "epoch": 5.945472966749563,
-      "grad_norm": 3.849186658859253,
-      "learning_rate": 2.2194812466369018e-05,
-      "loss": 0.62,
+      "epoch": 2.972746282962577,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.2191394831372436e-05,
+      "loss": 1.015971450805664,
       "step": 258200
     },
     {
-      "epoch": 5.950078290503822,
-      "grad_norm": 3.6725234985351562,
-      "learning_rate": 2.21875409025465e-05,
-      "loss": 0.6047,
+      "epoch": 2.9750489680372922,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2184123267549918e-05,
+      "loss": 0.9804122161865234,
       "step": 258400
     },
     {
-      "epoch": 5.954683614258082,
-      "grad_norm": 2.586672306060791,
-      "learning_rate": 2.2180269338723987e-05,
-      "loss": 0.6092,
+      "epoch": 2.977351653112007,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2176851703727404e-05,
+      "loss": 0.9570524597167969,
       "step": 258600
     },
     {
-      "epoch": 5.959288938012342,
-      "grad_norm": 3.1655867099761963,
-      "learning_rate": 2.2172997774901473e-05,
-      "loss": 0.609,
+      "epoch": 2.9796543381867218,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.216958013990489e-05,
+      "loss": 0.9921315002441407,
       "step": 258800
     },
     {
-      "epoch": 5.963894261766602,
-      "grad_norm": 2.586534023284912,
-      "learning_rate": 2.2165726211078955e-05,
-      "loss": 0.6168,
+      "epoch": 2.9819570232614367,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2162308576082372e-05,
+      "loss": 0.9853153228759766,
       "step": 259000
     },
     {
-      "epoch": 5.9684995855208625,
-      "grad_norm": 2.9593117237091064,
-      "learning_rate": 2.215845464725644e-05,
-      "loss": 0.6058,
+      "epoch": 2.9842597083361517,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.2155037012259855e-05,
+      "loss": 0.9744070434570312,
       "step": 259200
     },
     {
-      "epoch": 5.973104909275122,
-      "grad_norm": 3.1021628379821777,
-      "learning_rate": 2.2151183083433923e-05,
-      "loss": 0.6136,
+      "epoch": 2.9865623934108667,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.214776544843734e-05,
+      "loss": 1.001054458618164,
       "step": 259400
     },
     {
-      "epoch": 5.977710233029382,
-      "grad_norm": 3.3624014854431152,
-      "learning_rate": 2.2143911519611406e-05,
-      "loss": 0.6066,
+      "epoch": 2.9888650784855817,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.2140493884614827e-05,
+      "loss": 1.0041928100585937,
       "step": 259600
     },
     {
-      "epoch": 5.982315556783642,
-      "grad_norm": 3.2505760192871094,
-      "learning_rate": 2.2136639955788892e-05,
-      "loss": 0.6154,
+      "epoch": 2.9911677635602967,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.213322232079231e-05,
+      "loss": 0.9652430725097656,
       "step": 259800
     },
     {
-      "epoch": 5.986920880537902,
-      "grad_norm": 3.521606683731079,
-      "learning_rate": 2.2129368391966378e-05,
-      "loss": 0.6109,
+      "epoch": 2.993470448635011,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2125950756969795e-05,
+      "loss": 0.98406982421875,
       "step": 260000
     },
     {
-      "epoch": 5.991526204292162,
-      "grad_norm": 3.341179370880127,
-      "learning_rate": 2.212209682814386e-05,
-      "loss": 0.6179,
+      "epoch": 2.9957731337097266,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.211867919314728e-05,
+      "loss": 0.9826647186279297,
       "step": 260200
     },
     {
-      "epoch": 5.996131528046422,
-      "grad_norm": 3.8948066234588623,
-      "learning_rate": 2.211486162214046e-05,
-      "loss": 0.6183,
+      "epoch": 2.998075818784441,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.2111407629324764e-05,
+      "loss": 0.9724476623535157,
       "step": 260400
     },
     {
-      "epoch": 6.0,
-      "eval_loss": 0.5793450474739075,
-      "eval_runtime": 146.4894,
-      "eval_samples_per_second": 193.604,
-      "eval_steps_per_second": 12.103,
+      "epoch": 3.0,
+      "eval_loss": NaN,
+      "eval_runtime": 430.4459,
+      "eval_samples_per_second": 65.887,
+      "eval_steps_per_second": 32.945,
       "step": 260568
     },
     {
-      "epoch": 6.000736851800681,
-      "grad_norm": 3.0538671016693115,
-      "learning_rate": 2.2107626416137054e-05,
-      "loss": 0.6025,
+      "epoch": 3.0003684296119544,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2104136065502246e-05,
+      "loss": 0.9871656036376953,
       "step": 260600
     },
     {
-      "epoch": 6.005342175554942,
-      "grad_norm": 2.459566354751587,
-      "learning_rate": 2.210035485231454e-05,
-      "loss": 0.6142,
+      "epoch": 3.0026711146866694,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.2096864501679732e-05,
+      "loss": 0.9829486083984375,
       "step": 260800
     },
     {
-      "epoch": 6.0099474993092015,
-      "grad_norm": 3.1511268615722656,
-      "learning_rate": 2.2093083288492026e-05,
-      "loss": 0.6067,
+      "epoch": 3.0049737997613843,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.2089592937857215e-05,
+      "loss": 0.9937672424316406,
       "step": 261000
     },
     {
-      "epoch": 6.014552823063461,
-      "grad_norm": 2.9049158096313477,
-      "learning_rate": 2.2085811724669505e-05,
-      "loss": 0.6013,
+      "epoch": 3.0072764848360993,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.20823213740347e-05,
+      "loss": 0.9738970947265625,
       "step": 261200
     },
     {
-      "epoch": 6.0191581468177215,
-      "grad_norm": 2.9899065494537354,
-      "learning_rate": 2.207854016084699e-05,
-      "loss": 0.5913,
+      "epoch": 3.0095791699108143,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2075049810212186e-05,
+      "loss": 0.9735186004638672,
       "step": 261400
     },
     {
-      "epoch": 6.023763470571981,
-      "grad_norm": 3.2502787113189697,
-      "learning_rate": 2.2071268597024477e-05,
-      "loss": 0.5923,
+      "epoch": 3.011881854985529,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.206777824638967e-05,
+      "loss": 1.0058969116210938,
       "step": 261600
     },
     {
-      "epoch": 6.028368794326241,
-      "grad_norm": 3.313624382019043,
-      "learning_rate": 2.2063997033201963e-05,
-      "loss": 0.6092,
+      "epoch": 3.014184540060244,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.206050668256715e-05,
+      "loss": 1.0027965545654296,
       "step": 261800
     },
     {
-      "epoch": 6.032974118080501,
-      "grad_norm": 3.1583757400512695,
-      "learning_rate": 2.2056725469379445e-05,
-      "loss": 0.6048,
+      "epoch": 3.016487225134959,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2053235118744637e-05,
+      "loss": 0.9587387084960938,
       "step": 262000
     },
     {
-      "epoch": 6.037579441834761,
-      "grad_norm": 2.4791722297668457,
-      "learning_rate": 2.204945390555693e-05,
-      "loss": 0.6054,
+      "epoch": 3.018789910209674,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.2045963554922123e-05,
+      "loss": 0.952236557006836,
       "step": 262200
     },
     {
-      "epoch": 6.042184765589021,
-      "grad_norm": 3.2906877994537354,
-      "learning_rate": 2.2042182341734414e-05,
-      "loss": 0.6054,
+      "epoch": 3.021092595284389,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.2038691991099606e-05,
+      "loss": 0.9858063507080078,
       "step": 262400
     },
     {
-      "epoch": 6.046790089343281,
-      "grad_norm": 3.4483656883239746,
-      "learning_rate": 2.2034910777911896e-05,
-      "loss": 0.5909,
+      "epoch": 3.0233952803591038,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.2031420427277092e-05,
+      "loss": 1.0064692687988281,
       "step": 262600
     },
     {
-      "epoch": 6.0513954130975405,
-      "grad_norm": 3.3106260299682617,
-      "learning_rate": 2.2027639214089382e-05,
-      "loss": 0.6066,
+      "epoch": 3.0256979654338187,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.2024148863454578e-05,
+      "loss": 0.9844532012939453,
       "step": 262800
     },
     {
-      "epoch": 6.056000736851801,
-      "grad_norm": 3.107494354248047,
-      "learning_rate": 2.2020367650266868e-05,
-      "loss": 0.5997,
+      "epoch": 3.0280006505085337,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.201687729963206e-05,
+      "loss": 0.9752964782714844,
       "step": 263000
     },
     {
-      "epoch": 6.0606060606060606,
-      "grad_norm": 3.537893295288086,
-      "learning_rate": 2.2013132444263463e-05,
-      "loss": 0.5907,
+      "epoch": 3.0303033355832487,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.2009605735809543e-05,
+      "loss": 0.970207748413086,
       "step": 263200
     },
     {
-      "epoch": 6.06521138436032,
-      "grad_norm": 4.185168266296387,
-      "learning_rate": 2.200586088044095e-05,
-      "loss": 0.5958,
+      "epoch": 3.0326060206579637,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.200233417198703e-05,
+      "loss": 0.9757695007324219,
       "step": 263400
     },
     {
-      "epoch": 6.069816708114581,
-      "grad_norm": 2.7345850467681885,
-      "learning_rate": 2.1998589316618435e-05,
-      "loss": 0.6022,
+      "epoch": 3.0349087057326782,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.199506260816451e-05,
+      "loss": 0.9893525695800781,
       "step": 263600
     },
     {
-      "epoch": 6.07442203186884,
-      "grad_norm": 2.91339111328125,
-      "learning_rate": 2.1991317752795914e-05,
-      "loss": 0.6085,
+      "epoch": 3.037211390807393,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.1987791044341997e-05,
+      "loss": 0.9798577880859375,
       "step": 263800
     },
     {
-      "epoch": 6.079027355623101,
-      "grad_norm": 3.6052587032318115,
-      "learning_rate": 2.19840461889734e-05,
-      "loss": 0.5946,
+      "epoch": 3.039514075882108,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.1980519480519483e-05,
+      "loss": 0.9850680541992187,
       "step": 264000
     },
     {
-      "epoch": 6.08363267937736,
-      "grad_norm": 3.44474196434021,
-      "learning_rate": 2.1976774625150886e-05,
-      "loss": 0.6043,
+      "epoch": 3.041816760956823,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1973247916696966e-05,
+      "loss": 0.958910140991211,
       "step": 264200
     },
     {
-      "epoch": 6.08823800313162,
-      "grad_norm": 2.796043634414673,
-      "learning_rate": 2.196950306132837e-05,
-      "loss": 0.5965,
+      "epoch": 3.044119446031538,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1965976352874448e-05,
+      "loss": 0.9847620391845703,
       "step": 264400
     },
     {
-      "epoch": 6.09284332688588,
-      "grad_norm": 2.4689533710479736,
-      "learning_rate": 2.1962231497505855e-05,
-      "loss": 0.6029,
+      "epoch": 3.046422131106253,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1958704789051934e-05,
+      "loss": 0.9931247711181641,
       "step": 264600
     },
     {
-      "epoch": 6.09744865064014,
-      "grad_norm": 4.061243057250977,
-      "learning_rate": 2.195495993368334e-05,
-      "loss": 0.6063,
+      "epoch": 3.048724816180968,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.1951433225229416e-05,
+      "loss": 0.9987555694580078,
       "step": 264800
     },
     {
-      "epoch": 6.1020539743944,
-      "grad_norm": 3.009557008743286,
-      "learning_rate": 2.1947688369860823e-05,
-      "loss": 0.6208,
+      "epoch": 3.051027501255683,
+      "grad_norm": 0.5,
+      "learning_rate": 2.1944161661406902e-05,
+      "loss": 0.9749890899658203,
       "step": 265000
     },
     {
-      "epoch": 6.10665929814866,
-      "grad_norm": 3.3373091220855713,
-      "learning_rate": 2.1940416806038306e-05,
-      "loss": 0.609,
+      "epoch": 3.053330186330398,
+      "grad_norm": 0.875,
+      "learning_rate": 2.193689009758439e-05,
+      "loss": 0.9770771026611328,
       "step": 265200
     },
     {
-      "epoch": 6.11126462190292,
-      "grad_norm": 2.8920810222625732,
-      "learning_rate": 2.193314524221579e-05,
-      "loss": 0.5935,
+      "epoch": 3.0556328714051126,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.1929618533761874e-05,
+      "loss": 0.9705675506591797,
       "step": 265400
     },
     {
-      "epoch": 6.11586994565718,
-      "grad_norm": 4.441521644592285,
-      "learning_rate": 2.1925873678393274e-05,
-      "loss": 0.6031,
+      "epoch": 3.0579355564798276,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1922346969939353e-05,
+      "loss": 0.985162124633789,
       "step": 265600
     },
     {
-      "epoch": 6.12047526941144,
-      "grad_norm": 3.1611151695251465,
-      "learning_rate": 2.191860211457076e-05,
-      "loss": 0.6097,
+      "epoch": 3.0602382415545426,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.191507540611684e-05,
+      "loss": 0.9938688659667969,
       "step": 265800
     },
     {
-      "epoch": 6.125080593165699,
-      "grad_norm": 2.591564178466797,
-      "learning_rate": 2.1911330550748246e-05,
-      "loss": 0.5988,
+      "epoch": 3.0625409266292576,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.1907803842294325e-05,
+      "loss": 0.9857105255126953,
       "step": 266000
     },
     {
-      "epoch": 6.12968591691996,
-      "grad_norm": 2.5633347034454346,
-      "learning_rate": 2.1904058986925732e-05,
-      "loss": 0.6033,
+      "epoch": 3.0648436117039726,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.1900532278471808e-05,
+      "loss": 0.9707794952392578,
       "step": 266200
     },
     {
-      "epoch": 6.134291240674219,
-      "grad_norm": 2.846238851547241,
-      "learning_rate": 2.189678742310321e-05,
-      "loss": 0.6011,
+      "epoch": 3.0671462967786876,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1893260714649294e-05,
+      "loss": 0.9963924407958984,
       "step": 266400
     },
     {
-      "epoch": 6.138896564428479,
-      "grad_norm": 2.9226114749908447,
-      "learning_rate": 2.1889515859280697e-05,
-      "loss": 0.6115,
+      "epoch": 3.0694489818534025,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.188598915082678e-05,
+      "loss": 0.98943603515625,
       "step": 266600
     },
     {
-      "epoch": 6.1435018881827395,
-      "grad_norm": 3.3836491107940674,
-      "learning_rate": 2.1882244295458183e-05,
-      "loss": 0.6041,
+      "epoch": 3.0717516669281175,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.1878717587004262e-05,
+      "loss": 0.9711357879638672,
       "step": 266800
     },
     {
-      "epoch": 6.148107211936999,
-      "grad_norm": 3.141162157058716,
-      "learning_rate": 2.1874972731635665e-05,
-      "loss": 0.5929,
+      "epoch": 3.0740543520028325,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.1871446023181745e-05,
+      "loss": 1.0122907257080078,
       "step": 267000
     },
     {
-      "epoch": 6.1527125356912595,
-      "grad_norm": 3.0072133541107178,
-      "learning_rate": 2.186770116781315e-05,
-      "loss": 0.6038,
+      "epoch": 3.076357037077547,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.186417445935923e-05,
+      "loss": 0.9749220275878906,
       "step": 267200
     },
     {
-      "epoch": 6.157317859445519,
-      "grad_norm": 3.065700054168701,
-      "learning_rate": 2.1860465961809746e-05,
-      "loss": 0.5946,
+      "epoch": 3.078659722152262,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.1856902895536713e-05,
+      "loss": 0.9898588562011719,
       "step": 267400
     },
     {
-      "epoch": 6.161923183199779,
-      "grad_norm": 2.743119239807129,
-      "learning_rate": 2.1853194397987232e-05,
-      "loss": 0.615,
+      "epoch": 3.080962407226977,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.18496313317142e-05,
+      "loss": 0.9629341888427735,
       "step": 267600
     },
     {
-      "epoch": 6.166528506954039,
-      "grad_norm": 3.0159752368927,
-      "learning_rate": 2.184595919198383e-05,
-      "loss": 0.5905,
+      "epoch": 3.083265092301692,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1842359767891685e-05,
+      "loss": 0.9805014038085937,
       "step": 267800
     },
     {
-      "epoch": 6.171133830708299,
-      "grad_norm": 3.077939033508301,
-      "learning_rate": 2.1838687628161313e-05,
-      "loss": 0.6043,
+      "epoch": 3.085567777376407,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1835088204069167e-05,
+      "loss": 0.9865693664550781,
       "step": 268000
     },
     {
-      "epoch": 6.175739154462558,
-      "grad_norm": 3.3442816734313965,
-      "learning_rate": 2.1831416064338796e-05,
-      "loss": 0.606,
+      "epoch": 3.087870462451122,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.182781664024665e-05,
+      "loss": 0.957706298828125,
       "step": 268200
     },
     {
-      "epoch": 6.180344478216819,
-      "grad_norm": 2.753549337387085,
-      "learning_rate": 2.1824144500516282e-05,
-      "loss": 0.6119,
+      "epoch": 3.090173147525837,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.1820545076424136e-05,
+      "loss": 0.96357177734375,
       "step": 268400
     },
     {
-      "epoch": 6.1849498019710785,
-      "grad_norm": 2.6323747634887695,
-      "learning_rate": 2.1816872936693764e-05,
-      "loss": 0.5995,
+      "epoch": 3.092475832600552,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1813273512601622e-05,
+      "loss": 1.0093704223632813,
       "step": 268600
     },
     {
-      "epoch": 6.189555125725338,
-      "grad_norm": 3.0418403148651123,
-      "learning_rate": 2.180960137287125e-05,
-      "loss": 0.6196,
+      "epoch": 3.094778517675267,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.1806001948779104e-05,
+      "loss": 1.0027647399902344,
       "step": 268800
     },
     {
-      "epoch": 6.1941604494795985,
-      "grad_norm": 2.880768060684204,
-      "learning_rate": 2.1802329809048736e-05,
-      "loss": 0.5948,
+      "epoch": 3.0970812027499814,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.179873038495659e-05,
+      "loss": 0.990602798461914,
       "step": 269000
     },
     {
-      "epoch": 6.198765773233858,
-      "grad_norm": 3.229389190673828,
-      "learning_rate": 2.179505824522622e-05,
-      "loss": 0.6088,
+      "epoch": 3.0993838878246964,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.1791458821134076e-05,
+      "loss": 0.9695339202880859,
       "step": 269200
     },
     {
-      "epoch": 6.203371096988119,
-      "grad_norm": 2.6277809143066406,
-      "learning_rate": 2.17877866814037e-05,
-      "loss": 0.6097,
+      "epoch": 3.1016865728994114,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1784187257311555e-05,
+      "loss": 0.9737651062011718,
       "step": 269400
     },
     {
-      "epoch": 6.207976420742378,
-      "grad_norm": 2.8957467079162598,
-      "learning_rate": 2.1780515117581187e-05,
-      "loss": 0.6027,
+      "epoch": 3.1039892579741264,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.177691569348904e-05,
+      "loss": 0.9789590454101562,
       "step": 269600
     },
     {
-      "epoch": 6.212581744496638,
-      "grad_norm": 3.4677155017852783,
-      "learning_rate": 2.1773243553758673e-05,
-      "loss": 0.6016,
+      "epoch": 3.1062919430488414,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.1769644129666527e-05,
+      "loss": 0.9908535003662109,
       "step": 269800
     },
     {
-      "epoch": 6.217187068250898,
-      "grad_norm": 2.457707405090332,
-      "learning_rate": 2.1765971989936156e-05,
-      "loss": 0.6114,
+      "epoch": 3.1085946281235564,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.176237256584401e-05,
+      "loss": 0.9667711639404297,
       "step": 270000
     },
     {
-      "epoch": 6.221792392005158,
-      "grad_norm": 3.322401762008667,
-      "learning_rate": 2.1758773141751867e-05,
-      "loss": 0.6114,
+      "epoch": 3.1108973131982713,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1755101002021495e-05,
+      "loss": 0.9762010192871093,
       "step": 270200
     },
     {
-      "epoch": 6.2263977157594175,
-      "grad_norm": 2.784977912902832,
-      "learning_rate": 2.175150157792935e-05,
-      "loss": 0.5959,
+      "epoch": 3.1131999982729863,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.174782943819898e-05,
+      "loss": 0.9766085815429687,
       "step": 270400
     },
     {
-      "epoch": 6.231003039513678,
-      "grad_norm": 2.8627588748931885,
-      "learning_rate": 2.1744230014106835e-05,
-      "loss": 0.6067,
+      "epoch": 3.1155026833477013,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.1740557874376464e-05,
+      "loss": 0.9565920257568359,
       "step": 270600
     },
     {
-      "epoch": 6.2356083632679375,
-      "grad_norm": 3.0779459476470947,
-      "learning_rate": 2.173695845028432e-05,
-      "loss": 0.6007,
+      "epoch": 3.117805368422416,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1733286310553946e-05,
+      "loss": 1.0026136016845704,
       "step": 270800
     },
     {
-      "epoch": 6.240213687022198,
-      "grad_norm": 3.0918633937835693,
-      "learning_rate": 2.17296868864618e-05,
-      "loss": 0.6034,
+      "epoch": 3.120108053497131,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.1726014746731432e-05,
+      "loss": 0.9930970001220704,
       "step": 271000
     },
     {
-      "epoch": 6.244819010776458,
-      "grad_norm": 3.245352029800415,
-      "learning_rate": 2.1722415322639286e-05,
-      "loss": 0.6013,
+      "epoch": 3.122410738571846,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1718743182908918e-05,
+      "loss": 0.9859510803222656,
       "step": 271200
     },
     {
-      "epoch": 6.249424334530717,
-      "grad_norm": 3.179417610168457,
-      "learning_rate": 2.1715143758816772e-05,
-      "loss": 0.6043,
+      "epoch": 3.124713423646561,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.17114716190864e-05,
+      "loss": 0.9907965087890624,
       "step": 271400
     },
     {
-      "epoch": 6.254029658284978,
-      "grad_norm": 3.051330804824829,
-      "learning_rate": 2.1707872194994255e-05,
-      "loss": 0.5994,
+      "epoch": 3.1270161087212758,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.1704200055263887e-05,
+      "loss": 0.9711041259765625,
       "step": 271600
     },
     {
-      "epoch": 6.258634982039237,
-      "grad_norm": 3.0867514610290527,
-      "learning_rate": 2.170060063117174e-05,
-      "loss": 0.607,
+      "epoch": 3.1293187937959908,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.1696928491441373e-05,
+      "loss": 0.9887767028808594,
       "step": 271800
     },
     {
-      "epoch": 6.263240305793497,
-      "grad_norm": 3.2159993648529053,
-      "learning_rate": 2.1693329067349227e-05,
-      "loss": 0.592,
+      "epoch": 3.1316214788707057,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1689656927618852e-05,
+      "loss": 0.9776232147216797,
       "step": 272000
     },
     {
-      "epoch": 6.267845629547757,
-      "grad_norm": 2.641268253326416,
-      "learning_rate": 2.168605750352671e-05,
-      "loss": 0.6019,
+      "epoch": 3.1339241639454207,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.1682385363796338e-05,
+      "loss": 0.9863359069824219,
       "step": 272200
     },
     {
-      "epoch": 6.272450953302017,
-      "grad_norm": 3.2037994861602783,
-      "learning_rate": 2.167878593970419e-05,
-      "loss": 0.5966,
+      "epoch": 3.1362268490201357,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.1675113799973824e-05,
+      "loss": 0.9684203338623046,
       "step": 272400
     },
     {
-      "epoch": 6.2770562770562774,
-      "grad_norm": 2.531846523284912,
-      "learning_rate": 2.1671514375881678e-05,
-      "loss": 0.598,
+      "epoch": 3.1385295340948502,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.1667842236151306e-05,
+      "loss": 0.9655731964111328,
       "step": 272600
     },
     {
-      "epoch": 6.281661600810537,
-      "grad_norm": 3.13655161857605,
-      "learning_rate": 2.1664242812059163e-05,
-      "loss": 0.6102,
+      "epoch": 3.1408322191695652,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.1660570672328792e-05,
+      "loss": 0.9733708953857422,
       "step": 272800
     },
     {
-      "epoch": 6.286266924564797,
-      "grad_norm": 3.5509562492370605,
-      "learning_rate": 2.1656971248236646e-05,
-      "loss": 0.6071,
+      "epoch": 3.14313490424428,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.1653299108506278e-05,
+      "loss": 0.9558826446533203,
       "step": 273000
     },
     {
-      "epoch": 6.290872248319057,
-      "grad_norm": 4.013696193695068,
-      "learning_rate": 2.1649699684414132e-05,
-      "loss": 0.6048,
+      "epoch": 3.145437589318995,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.164602754468376e-05,
+      "loss": 0.9840922546386719,
       "step": 273200
     },
     {
-      "epoch": 6.295477572073317,
-      "grad_norm": 3.690094232559204,
-      "learning_rate": 2.1642428120591618e-05,
-      "loss": 0.6037,
+      "epoch": 3.14774027439371,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1638755980861243e-05,
+      "loss": 0.9617237854003906,
       "step": 273400
     },
     {
-      "epoch": 6.300082895827576,
-      "grad_norm": 2.8810877799987793,
-      "learning_rate": 2.1635156556769097e-05,
-      "loss": 0.5939,
+      "epoch": 3.150042959468425,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.163148441703873e-05,
+      "loss": 0.9859131622314453,
       "step": 273600
     },
     {
-      "epoch": 6.304688219581837,
-      "grad_norm": 2.392094612121582,
-      "learning_rate": 2.1627884992946583e-05,
-      "loss": 0.5881,
+      "epoch": 3.15234564454314,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.162421285321621e-05,
+      "loss": 0.9815999603271485,
       "step": 273800
     },
     {
-      "epoch": 6.309293543336096,
-      "grad_norm": 2.9791033267974854,
-      "learning_rate": 2.162061342912407e-05,
-      "loss": 0.6063,
+      "epoch": 3.154648329617855,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1616941289393697e-05,
+      "loss": 0.9714360809326172,
       "step": 274000
     },
     {
-      "epoch": 6.313898867090357,
-      "grad_norm": 3.6105103492736816,
-      "learning_rate": 2.161334186530155e-05,
-      "loss": 0.5895,
+      "epoch": 3.15695101469257,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.1609669725571183e-05,
+      "loss": 0.9890179443359375,
       "step": 274200
     },
     {
-      "epoch": 6.3185041908446165,
-      "grad_norm": 3.0321569442749023,
-      "learning_rate": 2.1606070301479037e-05,
-      "loss": 0.5868,
+      "epoch": 3.159253699767285,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.160239816174867e-05,
+      "loss": 0.9887825012207031,
       "step": 274400
     },
     {
-      "epoch": 6.323109514598876,
-      "grad_norm": 2.3522133827209473,
-      "learning_rate": 2.1598798737656523e-05,
-      "loss": 0.5985,
+      "epoch": 3.1615563848419996,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.159512659792615e-05,
+      "loss": 0.9828462219238281,
       "step": 274600
     },
     {
-      "epoch": 6.3277148383531365,
-      "grad_norm": 3.172877550125122,
-      "learning_rate": 2.1591527173834002e-05,
-      "loss": 0.6006,
+      "epoch": 3.1638590699167146,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.1587855034103634e-05,
+      "loss": 0.9640022277832031,
       "step": 274800
     },
     {
-      "epoch": 6.332320162107396,
-      "grad_norm": 3.2130961418151855,
-      "learning_rate": 2.1584255610011488e-05,
-      "loss": 0.6014,
+      "epoch": 3.1661617549914296,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.158058347028112e-05,
+      "loss": 0.9707381439208984,
       "step": 275000
     },
     {
-      "epoch": 6.336925485861656,
-      "grad_norm": 3.1269421577453613,
-      "learning_rate": 2.1576984046188974e-05,
-      "loss": 0.5893,
+      "epoch": 3.1684644400661446,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.1573311906458603e-05,
+      "loss": 0.9819551849365235,
       "step": 275200
     },
     {
-      "epoch": 6.341530809615916,
-      "grad_norm": 2.8266472816467285,
-      "learning_rate": 2.1569712482366457e-05,
-      "loss": 0.6025,
+      "epoch": 3.1707671251408596,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.156604034263609e-05,
+      "loss": 0.970483169555664,
       "step": 275400
     },
     {
-      "epoch": 6.346136133370176,
-      "grad_norm": 3.0055863857269287,
-      "learning_rate": 2.1562440918543943e-05,
-      "loss": 0.5972,
+      "epoch": 3.1730698102155745,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1558768778813574e-05,
+      "loss": 0.9733103942871094,
       "step": 275600
     },
     {
-      "epoch": 6.350741457124435,
-      "grad_norm": 3.1242921352386475,
-      "learning_rate": 2.155516935472143e-05,
-      "loss": 0.603,
+      "epoch": 3.1753724952902895,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1551497214991054e-05,
+      "loss": 0.9875724029541015,
       "step": 275800
     },
     {
-      "epoch": 6.355346780878696,
-      "grad_norm": 3.971442222595215,
-      "learning_rate": 2.1547897790898914e-05,
-      "loss": 0.6083,
+      "epoch": 3.1776751803650045,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.154422565116854e-05,
+      "loss": 0.9920774078369141,
       "step": 276000
     },
     {
-      "epoch": 6.3599521046329555,
-      "grad_norm": 3.2091002464294434,
-      "learning_rate": 2.154066258489551e-05,
-      "loss": 0.6112,
+      "epoch": 3.1799778654397195,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.1536954087346025e-05,
+      "loss": 0.9924172210693359,
       "step": 276200
     },
     {
-      "epoch": 6.364557428387216,
-      "grad_norm": 3.378394842147827,
-      "learning_rate": 2.1533391021072992e-05,
-      "loss": 0.5943,
+      "epoch": 3.182280550514434,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.1529682523523508e-05,
+      "loss": 0.9932424926757812,
       "step": 276400
     },
     {
-      "epoch": 6.3691627521414755,
-      "grad_norm": 3.663804292678833,
-      "learning_rate": 2.1526119457250475e-05,
-      "loss": 0.5942,
+      "epoch": 3.184583235589149,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.1522410959700994e-05,
+      "loss": 0.9929659271240234,
       "step": 276600
     },
     {
-      "epoch": 6.373768075895735,
-      "grad_norm": 2.897817611694336,
-      "learning_rate": 2.151884789342796e-05,
-      "loss": 0.5977,
+      "epoch": 3.186885920663864,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.151513939587848e-05,
+      "loss": 0.9686992645263672,
       "step": 276800
     },
     {
-      "epoch": 6.378373399649996,
-      "grad_norm": 3.405097484588623,
-      "learning_rate": 2.1511576329605446e-05,
-      "loss": 0.5935,
+      "epoch": 3.189188605738579,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.1507867832055966e-05,
+      "loss": 0.9910945129394532,
       "step": 277000
     },
     {
-      "epoch": 6.382978723404255,
-      "grad_norm": 3.117182970046997,
-      "learning_rate": 2.150430476578293e-05,
-      "loss": 0.6087,
+      "epoch": 3.191491290813294,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.1500596268233445e-05,
+      "loss": 0.9778562164306641,
       "step": 277200
     },
     {
-      "epoch": 6.387584047158516,
-      "grad_norm": 2.3870809078216553,
-      "learning_rate": 2.1497033201960415e-05,
-      "loss": 0.6,
+      "epoch": 3.193793975888009,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.149332470441093e-05,
+      "loss": 0.9815302276611328,
       "step": 277400
     },
     {
-      "epoch": 6.392189370912775,
-      "grad_norm": 3.2303497791290283,
-      "learning_rate": 2.1489761638137897e-05,
-      "loss": 0.6026,
+      "epoch": 3.196096660962724,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.1486053140588417e-05,
+      "loss": 0.9855834197998047,
       "step": 277600
     },
     {
-      "epoch": 6.396794694667035,
-      "grad_norm": 3.261338710784912,
-      "learning_rate": 2.1482490074315383e-05,
-      "loss": 0.6004,
+      "epoch": 3.198399346037439,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.14787815767659e-05,
+      "loss": 0.9841216278076171,
       "step": 277800
     },
     {
-      "epoch": 6.401400018421295,
-      "grad_norm": 2.4772233963012695,
-      "learning_rate": 2.1475218510492866e-05,
-      "loss": 0.6047,
+      "epoch": 3.200702031112154,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1471510012943385e-05,
+      "loss": 0.9852655029296875,
       "step": 278000
     },
     {
-      "epoch": 6.406005342175555,
-      "grad_norm": 4.6650590896606445,
-      "learning_rate": 2.1467946946670352e-05,
-      "loss": 0.5949,
+      "epoch": 3.203004716186869,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.146423844912087e-05,
+      "loss": 0.9748995971679687,
       "step": 278200
     },
     {
-      "epoch": 6.4106106659298145,
-      "grad_norm": 2.938955783843994,
-      "learning_rate": 2.1460675382847838e-05,
-      "loss": 0.6107,
+      "epoch": 3.2053074012615834,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.145696688529835e-05,
+      "loss": 0.9632807922363281,
       "step": 278400
     },
     {
-      "epoch": 6.415215989684075,
-      "grad_norm": 2.6229448318481445,
-      "learning_rate": 2.145340381902532e-05,
-      "loss": 0.6161,
+      "epoch": 3.2076100863362984,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.1449695321475836e-05,
+      "loss": 0.9814723205566406,
       "step": 278600
     },
     {
-      "epoch": 6.419821313438335,
-      "grad_norm": 2.3591392040252686,
-      "learning_rate": 2.1446132255202806e-05,
-      "loss": 0.5965,
+      "epoch": 3.2099127714110134,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.1442423757653322e-05,
+      "loss": 0.9979678344726562,
       "step": 278800
     },
     {
-      "epoch": 6.424426637192594,
-      "grad_norm": 3.7696409225463867,
-      "learning_rate": 2.143886069138029e-05,
-      "loss": 0.6014,
+      "epoch": 3.2122154564857284,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.1435152193830805e-05,
+      "loss": 0.9872224426269531,
       "step": 279000
     },
     {
-      "epoch": 6.429031960946855,
-      "grad_norm": 3.477569818496704,
-      "learning_rate": 2.143158912755777e-05,
-      "loss": 0.5994,
+      "epoch": 3.2145181415604434,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.142788063000829e-05,
+      "loss": 0.987719497680664,
       "step": 279200
     },
     {
-      "epoch": 6.433637284701114,
-      "grad_norm": 3.7880353927612305,
-      "learning_rate": 2.1424317563735257e-05,
-      "loss": 0.6106,
+      "epoch": 3.2168208266351583,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.1420609066185776e-05,
+      "loss": 0.9706076049804687,
       "step": 279400
     },
     {
-      "epoch": 6.438242608455375,
-      "grad_norm": 2.632434844970703,
-      "learning_rate": 2.1417045999912743e-05,
-      "loss": 0.6119,
+      "epoch": 3.2191235117098733,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.1413337502363255e-05,
+      "loss": 1.0041499328613281,
       "step": 279600
     },
     {
-      "epoch": 6.442847932209634,
-      "grad_norm": 2.5677168369293213,
-      "learning_rate": 2.1409774436090226e-05,
-      "loss": 0.5988,
+      "epoch": 3.2214261967845883,
+      "grad_norm": 0.875,
+      "learning_rate": 2.140606593854074e-05,
+      "loss": 0.9960893249511719,
       "step": 279800
     },
     {
-      "epoch": 6.447453255963894,
-      "grad_norm": 3.0066206455230713,
-      "learning_rate": 2.140250287226771e-05,
-      "loss": 0.6077,
+      "epoch": 3.2237288818593033,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.1398794374718227e-05,
+      "loss": 0.9895648956298828,
       "step": 280000
     },
     {
-      "epoch": 6.452058579718154,
-      "grad_norm": 3.006856679916382,
-      "learning_rate": 2.1395231308445194e-05,
-      "loss": 0.5956,
+      "epoch": 3.226031566934018,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.1391522810895713e-05,
+      "loss": 0.9820411682128907,
       "step": 280200
     },
     {
-      "epoch": 6.456663903472414,
-      "grad_norm": 3.6395933628082275,
-      "learning_rate": 2.138799610244179e-05,
-      "loss": 0.6028,
+      "epoch": 3.228334252008733,
+      "grad_norm": 1.125,
+      "learning_rate": 2.1384251247073196e-05,
+      "loss": 0.993807144165039,
       "step": 280400
     },
     {
-      "epoch": 6.461269227226674,
-      "grad_norm": 2.876479148864746,
-      "learning_rate": 2.1380724538619275e-05,
-      "loss": 0.6037,
+      "epoch": 3.230636937083448,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.137697968325068e-05,
+      "loss": 0.9687660980224609,
       "step": 280600
     },
     {
-      "epoch": 6.465874550980934,
-      "grad_norm": 2.5705347061157227,
-      "learning_rate": 2.137345297479676e-05,
-      "loss": 0.6006,
+      "epoch": 3.2329396221581628,
+      "grad_norm": 1.7265625,
+      "learning_rate": 2.1369708119428168e-05,
+      "loss": 0.9756636810302735,
       "step": 280800
     },
     {
-      "epoch": 6.470479874735194,
-      "grad_norm": 3.720731496810913,
-      "learning_rate": 2.1366181410974244e-05,
-      "loss": 0.599,
+      "epoch": 3.2352423072328778,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.1362436555605647e-05,
+      "loss": 0.9740959167480469,
       "step": 281000
     },
     {
-      "epoch": 6.475085198489454,
-      "grad_norm": 4.109841823577881,
-      "learning_rate": 2.135890984715173e-05,
-      "loss": 0.6118,
+      "epoch": 3.2375449923075927,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1355164991783133e-05,
+      "loss": 0.9740967559814453,
       "step": 281200
     },
     {
-      "epoch": 6.479690522243714,
-      "grad_norm": 3.3773958683013916,
-      "learning_rate": 2.1351638283329215e-05,
-      "loss": 0.6039,
+      "epoch": 3.2398476773823077,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.134789342796062e-05,
+      "loss": 0.9790853881835937,
       "step": 281400
     },
     {
-      "epoch": 6.484295845997973,
-      "grad_norm": 2.8852438926696777,
-      "learning_rate": 2.1344366719506694e-05,
-      "loss": 0.6062,
+      "epoch": 3.2421503624570227,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.13406218641381e-05,
+      "loss": 0.9897017669677735,
       "step": 281600
     },
     {
-      "epoch": 6.488901169752234,
-      "grad_norm": 2.8208401203155518,
-      "learning_rate": 2.1337131513503293e-05,
-      "loss": 0.6038,
+      "epoch": 3.2444530475317377,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.1333350300315587e-05,
+      "loss": 0.9742921447753906,
       "step": 281800
     },
     {
-      "epoch": 6.4935064935064934,
-      "grad_norm": 2.8988304138183594,
-      "learning_rate": 2.132985994968078e-05,
-      "loss": 0.6174,
+      "epoch": 3.2467557326064522,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.1326078736493073e-05,
+      "loss": 0.9642566680908203,
       "step": 282000
     },
     {
-      "epoch": 6.498111817260753,
-      "grad_norm": 3.9682488441467285,
-      "learning_rate": 2.132258838585826e-05,
-      "loss": 0.5891,
+      "epoch": 3.249058417681167,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.1318807172670552e-05,
+      "loss": 0.9810607147216797,
       "step": 282200
     },
     {
-      "epoch": 6.5027171410150135,
-      "grad_norm": 2.687206745147705,
-      "learning_rate": 2.1315316822035747e-05,
-      "loss": 0.5942,
+      "epoch": 3.251361102755882,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.1311535608848038e-05,
+      "loss": 0.9737554168701172,
       "step": 282400
     },
     {
-      "epoch": 6.507322464769273,
-      "grad_norm": 2.6542367935180664,
-      "learning_rate": 2.1308045258213233e-05,
-      "loss": 0.6074,
+      "epoch": 3.253663787830597,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1304264045025524e-05,
+      "loss": 0.9845205688476563,
       "step": 282600
     },
     {
-      "epoch": 6.511927788523533,
-      "grad_norm": 4.3083906173706055,
-      "learning_rate": 2.1300773694390716e-05,
-      "loss": 0.6039,
+      "epoch": 3.255966472905312,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.129699248120301e-05,
+      "loss": 0.974234619140625,
       "step": 282800
     },
     {
-      "epoch": 6.516533112277793,
-      "grad_norm": 3.527921199798584,
-      "learning_rate": 2.1293502130568202e-05,
-      "loss": 0.6026,
+      "epoch": 3.258269157980027,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.1289720917380492e-05,
+      "loss": 0.9663818359375,
       "step": 283000
     },
     {
-      "epoch": 6.521138436032053,
-      "grad_norm": 2.9603312015533447,
-      "learning_rate": 2.1286230566745684e-05,
-      "loss": 0.6139,
+      "epoch": 3.260571843054742,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.1282449353557978e-05,
+      "loss": 0.988833236694336,
       "step": 283200
     },
     {
-      "epoch": 6.525743759786313,
-      "grad_norm": 3.2545838356018066,
-      "learning_rate": 2.1278959002923167e-05,
-      "loss": 0.6099,
+      "epoch": 3.262874528129457,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.1275177789735464e-05,
+      "loss": 0.9499282836914062,
       "step": 283400
     },
     {
-      "epoch": 6.530349083540573,
-      "grad_norm": 3.675568103790283,
-      "learning_rate": 2.1271687439100653e-05,
-      "loss": 0.6022,
+      "epoch": 3.265177213204172,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.1267906225912943e-05,
+      "loss": 0.9761683654785156,
       "step": 283600
     },
     {
-      "epoch": 6.5349544072948325,
-      "grad_norm": 2.5842385292053223,
-      "learning_rate": 2.126441587527814e-05,
-      "loss": 0.608,
+      "epoch": 3.2674798982788866,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.126063466209043e-05,
+      "loss": 0.9818894195556641,
       "step": 283800
     },
     {
-      "epoch": 6.539559731049093,
-      "grad_norm": 3.1550087928771973,
-      "learning_rate": 2.1257144311455625e-05,
-      "loss": 0.5932,
+      "epoch": 3.2697825833536016,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.1253363098267915e-05,
+      "loss": 0.99319091796875,
       "step": 284000
     },
     {
-      "epoch": 6.5441650548033525,
-      "grad_norm": 2.6331937313079834,
-      "learning_rate": 2.1249872747633107e-05,
-      "loss": 0.5988,
+      "epoch": 3.2720852684283166,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.1246091534445398e-05,
+      "loss": 0.992230453491211,
       "step": 284200
     },
     {
-      "epoch": 6.548770378557613,
-      "grad_norm": 3.004626750946045,
-      "learning_rate": 2.124260118381059e-05,
-      "loss": 0.6025,
+      "epoch": 3.2743879535030316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.1238819970622884e-05,
+      "loss": 0.9897457122802734,
       "step": 284400
     },
     {
-      "epoch": 6.553375702311873,
-      "grad_norm": 2.480337142944336,
-      "learning_rate": 2.1235329619988076e-05,
-      "loss": 0.5956,
+      "epoch": 3.2766906385777466,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.123154840680037e-05,
+      "loss": 0.9627342224121094,
       "step": 284600
     },
     {
-      "epoch": 6.557981026066132,
-      "grad_norm": 3.277716875076294,
-      "learning_rate": 2.1228058056165558e-05,
-      "loss": 0.5961,
+      "epoch": 3.2789933236524615,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.122427684297785e-05,
+      "loss": 0.9894837951660156,
       "step": 284800
     },
     {
-      "epoch": 6.562586349820393,
-      "grad_norm": 2.759143590927124,
-      "learning_rate": 2.1220786492343044e-05,
-      "loss": 0.5875,
+      "epoch": 3.2812960087271765,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.1217005279155334e-05,
+      "loss": 0.9815957641601563,
       "step": 285000
     },
     {
-      "epoch": 6.567191673574652,
-      "grad_norm": 2.480003833770752,
-      "learning_rate": 2.121351492852053e-05,
-      "loss": 0.5972,
+      "epoch": 3.2835986938018915,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.120973371533282e-05,
+      "loss": 0.9799363708496094,
       "step": 285200
     },
     {
-      "epoch": 6.571796997328912,
-      "grad_norm": 3.330686569213867,
-      "learning_rate": 2.1206279722517125e-05,
-      "loss": 0.5915,
+      "epoch": 3.2859013788766065,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.1202462151510303e-05,
+      "loss": 0.9863890075683593,
       "step": 285400
     },
     {
-      "epoch": 6.576402321083172,
-      "grad_norm": 2.3041577339172363,
-      "learning_rate": 2.119900815869461e-05,
-      "loss": 0.6091,
+      "epoch": 3.288204063951321,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.119519058768779e-05,
+      "loss": 0.9854249572753906,
       "step": 285600
     },
     {
-      "epoch": 6.581007644837432,
-      "grad_norm": 3.41249680519104,
-      "learning_rate": 2.1191736594872097e-05,
-      "loss": 0.5871,
+      "epoch": 3.290506749026036,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.1187919023865275e-05,
+      "loss": 0.9754698944091796,
       "step": 285800
     },
     {
-      "epoch": 6.5856129685916915,
-      "grad_norm": 2.979804277420044,
-      "learning_rate": 2.1184465031049576e-05,
-      "loss": 0.593,
+      "epoch": 3.292809434100751,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.1180647460042757e-05,
+      "loss": 0.9730270385742188,
       "step": 286000
     },
     {
-      "epoch": 6.590218292345952,
-      "grad_norm": 3.0623714923858643,
-      "learning_rate": 2.1177193467227062e-05,
-      "loss": 0.6024,
+      "epoch": 3.295112119175466,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.117337589622024e-05,
+      "loss": 0.9700435638427735,
       "step": 286200
     },
     {
-      "epoch": 6.594823616100212,
-      "grad_norm": 3.924044132232666,
-      "learning_rate": 2.1169921903404548e-05,
-      "loss": 0.6108,
+      "epoch": 3.297414804250181,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.1166104332397726e-05,
+      "loss": 0.9972874450683594,
       "step": 286400
     },
     {
-      "epoch": 6.599428939854472,
-      "grad_norm": 2.970489501953125,
-      "learning_rate": 2.116265033958203e-05,
-      "loss": 0.5991,
+      "epoch": 3.299717489324896,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.115883276857521e-05,
+      "loss": 0.9965068054199219,
       "step": 286600
     },
     {
-      "epoch": 6.604034263608732,
-      "grad_norm": 2.641425371170044,
-      "learning_rate": 2.1155378775759516e-05,
-      "loss": 0.6112,
+      "epoch": 3.302020174399611,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.1151561204752694e-05,
+      "loss": 0.9861847686767579,
       "step": 286800
     },
     {
-      "epoch": 6.608639587362991,
-      "grad_norm": 2.8824870586395264,
-      "learning_rate": 2.1148107211937002e-05,
-      "loss": 0.5837,
+      "epoch": 3.304322859474326,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.114428964093018e-05,
+      "loss": 0.9861808013916016,
       "step": 287000
     },
     {
-      "epoch": 6.613244911117252,
-      "grad_norm": 2.681509256362915,
-      "learning_rate": 2.114083564811448e-05,
-      "loss": 0.6031,
+      "epoch": 3.306625544549041,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.1137018077107666e-05,
+      "loss": 0.9840087127685547,
       "step": 287200
     },
     {
-      "epoch": 6.617850234871511,
-      "grad_norm": 3.070089101791382,
-      "learning_rate": 2.1133564084291967e-05,
-      "loss": 0.5896,
+      "epoch": 3.3089282296237554,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.1129746513285145e-05,
+      "loss": 0.979336166381836,
       "step": 287400
     },
     {
-      "epoch": 6.622455558625772,
-      "grad_norm": 3.1326851844787598,
-      "learning_rate": 2.1126328878288566e-05,
-      "loss": 0.5909,
+      "epoch": 3.3112309146984704,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.112247494946263e-05,
+      "loss": 0.9942705535888672,
       "step": 287600
     },
     {
-      "epoch": 6.627060882380031,
-      "grad_norm": 3.5522382259368896,
-      "learning_rate": 2.111905731446605e-05,
-      "loss": 0.601,
+      "epoch": 3.3135335997731854,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.1115203385640117e-05,
+      "loss": 0.9899281311035156,
       "step": 287800
     },
     {
-      "epoch": 6.631666206134291,
-      "grad_norm": 3.211503028869629,
-      "learning_rate": 2.1111785750643534e-05,
-      "loss": 0.6018,
+      "epoch": 3.3158362848479004,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.11079318218176e-05,
+      "loss": 0.9654022216796875,
       "step": 288000
     },
     {
-      "epoch": 6.6362715298885515,
-      "grad_norm": 2.8858470916748047,
-      "learning_rate": 2.110451418682102e-05,
-      "loss": 0.6202,
+      "epoch": 3.3181389699226154,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.1100660257995085e-05,
+      "loss": 0.9817692565917969,
       "step": 288200
     },
     {
-      "epoch": 6.640876853642811,
-      "grad_norm": 3.317497730255127,
-      "learning_rate": 2.1097242622998503e-05,
-      "loss": 0.5961,
+      "epoch": 3.3204416549973303,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.109338869417257e-05,
+      "loss": 0.9879948425292969,
       "step": 288400
     },
     {
-      "epoch": 6.645482177397071,
-      "grad_norm": 2.4950830936431885,
-      "learning_rate": 2.1089971059175985e-05,
-      "loss": 0.6002,
+      "epoch": 3.3227443400720453,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.1086117130350054e-05,
+      "loss": 0.9916354370117187,
       "step": 288600
     },
     {
-      "epoch": 6.650087501151331,
-      "grad_norm": 3.110621929168701,
-      "learning_rate": 2.108269949535347e-05,
-      "loss": 0.6046,
+      "epoch": 3.3250470251467603,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.1078845566527536e-05,
+      "loss": 0.9782781982421875,
       "step": 288800
     },
     {
-      "epoch": 6.654692824905591,
-      "grad_norm": 3.673153877258301,
-      "learning_rate": 2.1075427931530954e-05,
-      "loss": 0.5994,
+      "epoch": 3.3273497102214753,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.1071574002705022e-05,
+      "loss": 0.9925672912597656,
       "step": 289000
     },
     {
-      "epoch": 6.65929814865985,
-      "grad_norm": 3.486978530883789,
-      "learning_rate": 2.106815636770844e-05,
-      "loss": 0.5981,
+      "epoch": 3.32965239529619,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.1064302438882508e-05,
+      "loss": 0.971702651977539,
       "step": 289200
     },
     {
-      "epoch": 6.663903472414111,
-      "grad_norm": 3.4665493965148926,
-      "learning_rate": 2.1060884803885926e-05,
-      "loss": 0.6127,
+      "epoch": 3.331955080370905,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.105703087505999e-05,
+      "loss": 1.0004792022705078,
       "step": 289400
     },
     {
-      "epoch": 6.6685087961683704,
-      "grad_norm": 3.6639597415924072,
-      "learning_rate": 2.1053613240063408e-05,
-      "loss": 0.6067,
+      "epoch": 3.33425776544562,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.1049759311237477e-05,
+      "loss": 0.9968527221679687,
       "step": 289600
     },
     {
-      "epoch": 6.673114119922631,
-      "grad_norm": 3.012031316757202,
-      "learning_rate": 2.1046341676240894e-05,
-      "loss": 0.596,
+      "epoch": 3.336560450520335,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.104248774741496e-05,
+      "loss": 0.983421630859375,
       "step": 289800
     },
     {
-      "epoch": 6.6777194436768905,
-      "grad_norm": 2.717411518096924,
-      "learning_rate": 2.1039070112418377e-05,
-      "loss": 0.6002,
+      "epoch": 3.3388631355950498,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.103521618359244e-05,
+      "loss": 0.9886469268798828,
       "step": 290000
     },
     {
-      "epoch": 6.68232476743115,
-      "grad_norm": 2.571988105773926,
-      "learning_rate": 2.1031798548595862e-05,
-      "loss": 0.5986,
+      "epoch": 3.3411658206697648,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.1027944619769928e-05,
+      "loss": 0.9767284393310547,
       "step": 290200
     },
     {
-      "epoch": 6.686930091185411,
-      "grad_norm": 2.4127187728881836,
-      "learning_rate": 2.1024526984773345e-05,
-      "loss": 0.594,
+      "epoch": 3.3434685057444797,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1020673055947414e-05,
+      "loss": 0.9893907165527344,
       "step": 290400
     },
     {
-      "epoch": 6.69153541493967,
-      "grad_norm": 3.4181158542633057,
-      "learning_rate": 2.1017291778769944e-05,
-      "loss": 0.6023,
+      "epoch": 3.3457711908191947,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1013401492124896e-05,
+      "loss": 0.9892060852050781,
       "step": 290600
     },
     {
-      "epoch": 6.696140738693931,
-      "grad_norm": 3.2288289070129395,
-      "learning_rate": 2.101005657276654e-05,
-      "loss": 0.6004,
+      "epoch": 3.3480738758939097,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.1006129928302382e-05,
+      "loss": 0.9863589477539062,
       "step": 290800
     },
     {
-      "epoch": 6.70074606244819,
-      "grad_norm": 3.2876336574554443,
-      "learning_rate": 2.1002785008944025e-05,
-      "loss": 0.6054,
+      "epoch": 3.3503765609686242,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0998858364479868e-05,
+      "loss": 0.9872593688964844,
       "step": 291000
     },
     {
-      "epoch": 6.70535138620245,
-      "grad_norm": 2.789076328277588,
-      "learning_rate": 2.099551344512151e-05,
-      "loss": 0.6041,
+      "epoch": 3.3526792460433392,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.0991586800657347e-05,
+      "loss": 0.9884869384765625,
       "step": 291200
     },
     {
-      "epoch": 6.70995670995671,
-      "grad_norm": 2.86173939704895,
-      "learning_rate": 2.0988241881298993e-05,
-      "loss": 0.5944,
+      "epoch": 3.354981931118054,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.0984315236834833e-05,
+      "loss": 0.9901111602783204,
       "step": 291400
     },
     {
-      "epoch": 6.71456203371097,
-      "grad_norm": 2.8522419929504395,
-      "learning_rate": 2.0980970317476476e-05,
-      "loss": 0.5996,
+      "epoch": 3.357284616192769,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.097704367301232e-05,
+      "loss": 0.9891106414794922,
       "step": 291600
     },
     {
-      "epoch": 6.7191673574652295,
-      "grad_norm": 3.7493839263916016,
-      "learning_rate": 2.097369875365396e-05,
-      "loss": 0.5959,
+      "epoch": 3.359587301267484,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.0969772109189805e-05,
+      "loss": 0.9849647521972656,
       "step": 291800
     },
     {
-      "epoch": 6.72377268121949,
-      "grad_norm": 3.5623674392700195,
-      "learning_rate": 2.0966427189831444e-05,
-      "loss": 0.5982,
+      "epoch": 3.361889986342199,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.0962500545367287e-05,
+      "loss": 0.9863345336914062,
       "step": 292000
     },
     {
-      "epoch": 6.72837800497375,
-      "grad_norm": 3.4356765747070312,
-      "learning_rate": 2.095915562600893e-05,
-      "loss": 0.5966,
+      "epoch": 3.364192671416914,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0955228981544773e-05,
+      "loss": 1.0030756378173828,
       "step": 292200
     },
     {
-      "epoch": 6.732983328728009,
-      "grad_norm": 3.1786386966705322,
-      "learning_rate": 2.0951884062186416e-05,
-      "loss": 0.6007,
+      "epoch": 3.366495356491629,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.0947957417722256e-05,
+      "loss": 0.9938818359375,
       "step": 292400
     },
     {
-      "epoch": 6.73758865248227,
-      "grad_norm": 2.424635410308838,
-      "learning_rate": 2.09446124983639e-05,
-      "loss": 0.6155,
+      "epoch": 3.368798041566344,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0940685853899738e-05,
+      "loss": 0.9862526702880859,
       "step": 292600
     },
     {
-      "epoch": 6.742193976236529,
-      "grad_norm": 3.4390957355499268,
-      "learning_rate": 2.0937340934541384e-05,
-      "loss": 0.5994,
+      "epoch": 3.3711007266410586,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0933414290077224e-05,
+      "loss": 0.9963729095458984,
       "step": 292800
     },
     {
-      "epoch": 6.746799299990789,
-      "grad_norm": 4.376504421234131,
-      "learning_rate": 2.0930069370718867e-05,
-      "loss": 0.5985,
+      "epoch": 3.373403411715774,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.092614272625471e-05,
+      "loss": 0.9868894195556641,
       "step": 293000
     },
     {
-      "epoch": 6.751404623745049,
-      "grad_norm": 2.7441396713256836,
-      "learning_rate": 2.092279780689635e-05,
-      "loss": 0.6037,
+      "epoch": 3.3757060967904886,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.0918871162432193e-05,
+      "loss": 0.992561264038086,
       "step": 293200
     },
     {
-      "epoch": 6.756009947499309,
-      "grad_norm": 2.771723985671997,
-      "learning_rate": 2.0915562600892948e-05,
-      "loss": 0.5969,
+      "epoch": 3.3780087818652036,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.091159959860968e-05,
+      "loss": 0.9652907562255859,
       "step": 293400
     },
     {
-      "epoch": 6.760615271253569,
-      "grad_norm": 3.066767692565918,
-      "learning_rate": 2.0908291037070434e-05,
-      "loss": 0.5907,
+      "epoch": 3.3803114669399186,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0904328034787164e-05,
+      "loss": 0.9668193817138672,
       "step": 293600
     },
     {
-      "epoch": 6.765220595007829,
-      "grad_norm": 3.7000675201416016,
-      "learning_rate": 2.0901019473247916e-05,
-      "loss": 0.5968,
+      "epoch": 3.3826141520146336,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.0897056470964644e-05,
+      "loss": 0.962864990234375,
       "step": 293800
     },
     {
-      "epoch": 6.769825918762089,
-      "grad_norm": 2.6257290840148926,
-      "learning_rate": 2.0893747909425402e-05,
-      "loss": 0.6015,
+      "epoch": 3.3849168370893485,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.088978490714213e-05,
+      "loss": 0.9714779663085937,
       "step": 294000
     },
     {
-      "epoch": 6.774431242516349,
-      "grad_norm": 2.972223997116089,
-      "learning_rate": 2.0886476345602888e-05,
-      "loss": 0.6104,
+      "epoch": 3.3872195221640635,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0882513343319615e-05,
+      "loss": 0.9760708618164062,
       "step": 294200
     },
     {
-      "epoch": 6.779036566270609,
-      "grad_norm": 2.6368885040283203,
-      "learning_rate": 2.0879204781780367e-05,
-      "loss": 0.6009,
+      "epoch": 3.3895222072387785,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.0875241779497098e-05,
+      "loss": 0.9788286590576172,
       "step": 294400
     },
     {
-      "epoch": 6.783641890024869,
-      "grad_norm": 2.8268954753875732,
-      "learning_rate": 2.0871933217957853e-05,
-      "loss": 0.5894,
+      "epoch": 3.3918248923134935,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.0867970215674584e-05,
+      "loss": 0.9956980133056641,
       "step": 294600
     },
     {
-      "epoch": 6.788247213779129,
-      "grad_norm": 2.7802374362945557,
-      "learning_rate": 2.086466165413534e-05,
-      "loss": 0.6079,
+      "epoch": 3.3941275773882085,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.086069865185207e-05,
+      "loss": 0.9827316284179688,
       "step": 294800
     },
     {
-      "epoch": 6.792852537533388,
-      "grad_norm": 3.244723320007324,
-      "learning_rate": 2.0857390090312822e-05,
-      "loss": 0.5933,
+      "epoch": 3.396430262462923,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.0853427088029552e-05,
+      "loss": 0.9808500671386718,
       "step": 295000
     },
     {
-      "epoch": 6.797457861287649,
-      "grad_norm": 2.8554670810699463,
-      "learning_rate": 2.0850118526490308e-05,
-      "loss": 0.5893,
+      "epoch": 3.398732947537638,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.0846155524207035e-05,
+      "loss": 0.9899409484863281,
       "step": 295200
     },
     {
-      "epoch": 6.802063185041908,
-      "grad_norm": 3.411585807800293,
-      "learning_rate": 2.0842846962667794e-05,
-      "loss": 0.5874,
+      "epoch": 3.401035632612353,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.083888396038452e-05,
+      "loss": 0.9769796752929687,
       "step": 295400
     },
     {
-      "epoch": 6.806668508796168,
-      "grad_norm": 3.076263904571533,
-      "learning_rate": 2.0835575398845276e-05,
-      "loss": 0.5866,
+      "epoch": 3.403338317687068,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.0831612396562007e-05,
+      "loss": 0.9801354217529297,
       "step": 295600
     },
     {
-      "epoch": 6.8112738325504285,
-      "grad_norm": 3.0649490356445312,
-      "learning_rate": 2.082830383502276e-05,
-      "loss": 0.6006,
+      "epoch": 3.405641002761783,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.082434083273949e-05,
+      "loss": 1.0167056274414064,
       "step": 295800
     },
     {
-      "epoch": 6.815879156304688,
-      "grad_norm": 3.0832936763763428,
-      "learning_rate": 2.0821032271200245e-05,
-      "loss": 0.5896,
+      "epoch": 3.407943687836498,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.0817069268916975e-05,
+      "loss": 0.981565933227539,
       "step": 296000
     },
     {
-      "epoch": 6.820484480058948,
-      "grad_norm": 4.413976192474365,
-      "learning_rate": 2.081376070737773e-05,
-      "loss": 0.5995,
+      "epoch": 3.410246372911213,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.0809797705094458e-05,
+      "loss": 0.9864985656738281,
       "step": 296200
     },
     {
-      "epoch": 6.825089803813208,
-      "grad_norm": 2.9547083377838135,
-      "learning_rate": 2.0806489143555213e-05,
-      "loss": 0.588,
+      "epoch": 3.412549057985928,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.080252614127194e-05,
+      "loss": 0.9858273315429688,
       "step": 296400
     },
     {
-      "epoch": 6.829695127567468,
-      "grad_norm": 3.1978492736816406,
-      "learning_rate": 2.07992175797327e-05,
-      "loss": 0.605,
+      "epoch": 3.414851743060643,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0795254577449426e-05,
+      "loss": 0.9845752716064453,
       "step": 296600
     },
     {
-      "epoch": 6.834300451321728,
-      "grad_norm": 2.5067381858825684,
-      "learning_rate": 2.0791946015910185e-05,
-      "loss": 0.5945,
+      "epoch": 3.4171544281353574,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0787983013626912e-05,
+      "loss": 0.979270248413086,
       "step": 296800
     },
     {
-      "epoch": 6.838905775075988,
-      "grad_norm": 2.9180538654327393,
-      "learning_rate": 2.0784674452087664e-05,
-      "loss": 0.6033,
+      "epoch": 3.4194571132100724,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.0780711449804394e-05,
+      "loss": 1.0109344482421876,
       "step": 297000
     },
     {
-      "epoch": 6.843511098830247,
-      "grad_norm": 2.9533698558807373,
-      "learning_rate": 2.077740288826515e-05,
-      "loss": 0.5826,
+      "epoch": 3.4217597982847874,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.077343988598188e-05,
+      "loss": 0.9867710876464844,
       "step": 297200
     },
     {
-      "epoch": 6.848116422584508,
-      "grad_norm": 3.087790012359619,
-      "learning_rate": 2.0770131324442636e-05,
-      "loss": 0.6035,
+      "epoch": 3.4240624833595024,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0766168322159366e-05,
+      "loss": 0.9730811309814453,
       "step": 297400
     },
     {
-      "epoch": 6.8527217463387675,
-      "grad_norm": 3.130922317504883,
-      "learning_rate": 2.076285976062012e-05,
-      "loss": 0.599,
+      "epoch": 3.4263651684342173,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.075889675833685e-05,
+      "loss": 0.9847515869140625,
       "step": 297600
     },
     {
-      "epoch": 6.857327070093028,
-      "grad_norm": 3.0297648906707764,
-      "learning_rate": 2.0755588196797604e-05,
-      "loss": 0.6053,
+      "epoch": 3.4286678535089323,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.075162519451433e-05,
+      "loss": 0.9762758636474609,
       "step": 297800
     },
     {
-      "epoch": 6.861932393847288,
-      "grad_norm": 2.8300116062164307,
-      "learning_rate": 2.0748352990794203e-05,
-      "loss": 0.6033,
+      "epoch": 3.4309705385836473,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0744353630691817e-05,
+      "loss": 0.9684622192382812,
       "step": 298000
     },
     {
-      "epoch": 6.866537717601547,
-      "grad_norm": 3.1619298458099365,
-      "learning_rate": 2.0741081426971685e-05,
-      "loss": 0.5995,
+      "epoch": 3.4332732236583623,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.0737082066869303e-05,
+      "loss": 0.9942131042480469,
       "step": 298200
     },
     {
-      "epoch": 6.871143041355808,
-      "grad_norm": 3.3405187129974365,
-      "learning_rate": 2.0733809863149168e-05,
-      "loss": 0.5918,
+      "epoch": 3.4355759087330773,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.0729810503046786e-05,
+      "loss": 0.9833889770507812,
       "step": 298400
     },
     {
-      "epoch": 6.875748365110067,
-      "grad_norm": 2.8210740089416504,
-      "learning_rate": 2.0726538299326654e-05,
-      "loss": 0.599,
+      "epoch": 3.437878593807792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.072253893922427e-05,
+      "loss": 0.9864968109130859,
       "step": 298600
     },
     {
-      "epoch": 6.880353688864327,
-      "grad_norm": 3.2792325019836426,
-      "learning_rate": 2.0719266735504136e-05,
-      "loss": 0.5899,
+      "epoch": 3.440181278882507,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.0715267375401754e-05,
+      "loss": 0.9911534881591797,
       "step": 298800
     },
     {
-      "epoch": 6.884959012618587,
-      "grad_norm": 2.485522508621216,
-      "learning_rate": 2.0711995171681622e-05,
-      "loss": 0.5864,
+      "epoch": 3.442483963957222,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.0707995811579237e-05,
+      "loss": 0.9858901977539063,
       "step": 299000
     },
     {
-      "epoch": 6.889564336372847,
-      "grad_norm": 2.585327625274658,
-      "learning_rate": 2.0704723607859108e-05,
-      "loss": 0.5904,
+      "epoch": 3.4447866490319368,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.0700724247756723e-05,
+      "loss": 0.9796755981445312,
       "step": 299200
     },
     {
-      "epoch": 6.8941696601271065,
-      "grad_norm": 2.3402791023254395,
-      "learning_rate": 2.069745204403659e-05,
-      "loss": 0.5845,
+      "epoch": 3.4470893341066517,
+      "grad_norm": 0.875,
+      "learning_rate": 2.069345268393421e-05,
+      "loss": 0.981687240600586,
       "step": 299400
     },
     {
-      "epoch": 6.898774983881367,
-      "grad_norm": 2.9729385375976562,
-      "learning_rate": 2.0690180480214073e-05,
-      "loss": 0.5963,
+      "epoch": 3.4493920191813667,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.068618112011169e-05,
+      "loss": 0.9910392761230469,
       "step": 299600
     },
     {
-      "epoch": 6.903380307635627,
-      "grad_norm": 3.0972092151641846,
-      "learning_rate": 2.068290891639156e-05,
-      "loss": 0.6085,
+      "epoch": 3.4516947042560817,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0678909556289177e-05,
+      "loss": 0.9705625915527344,
       "step": 299800
     },
     {
-      "epoch": 6.907985631389887,
-      "grad_norm": 3.688617467880249,
-      "learning_rate": 2.0675637352569045e-05,
-      "loss": 0.6013,
+      "epoch": 3.4539973893307967,
+      "grad_norm": 0.75,
+      "learning_rate": 2.067163799246666e-05,
+      "loss": 0.9778738403320313,
       "step": 300000
     },
     {
-      "epoch": 6.912590955144147,
-      "grad_norm": 2.644268751144409,
-      "learning_rate": 2.0668365788746528e-05,
-      "loss": 0.5893,
+      "epoch": 3.4563000744055117,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.0664366428644142e-05,
+      "loss": 0.9827516937255859,
       "step": 300200
     },
     {
-      "epoch": 6.917196278898406,
-      "grad_norm": 3.49151873588562,
-      "learning_rate": 2.0661094224924013e-05,
-      "loss": 0.5896,
+      "epoch": 3.458602759480226,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.0657094864821628e-05,
+      "loss": 0.9912580108642578,
       "step": 300400
     },
     {
-      "epoch": 6.921801602652667,
-      "grad_norm": 3.661813974380493,
-      "learning_rate": 2.06538226611015e-05,
-      "loss": 0.5972,
+      "epoch": 3.460905444554941,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.0649823300999114e-05,
+      "loss": 1.0024603271484376,
       "step": 300600
     },
     {
-      "epoch": 6.926406926406926,
-      "grad_norm": 3.322665214538574,
-      "learning_rate": 2.0646551097278982e-05,
-      "loss": 0.5989,
+      "epoch": 3.463208129629656,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.06425517371766e-05,
+      "loss": 0.9889586639404296,
       "step": 300800
     },
     {
-      "epoch": 6.931012250161187,
-      "grad_norm": 3.0434811115264893,
-      "learning_rate": 2.0639279533456464e-05,
-      "loss": 0.5886,
+      "epoch": 3.465510814704371,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.0635280173354082e-05,
+      "loss": 0.97587158203125,
       "step": 301000
     },
     {
-      "epoch": 6.935617573915446,
-      "grad_norm": 3.6743383407592773,
-      "learning_rate": 2.063204432745306e-05,
-      "loss": 0.6096,
+      "epoch": 3.467813499779086,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.0628008609531568e-05,
+      "loss": 0.9978359985351563,
       "step": 301200
     },
     {
-      "epoch": 6.940222897669706,
-      "grad_norm": 3.656269073486328,
-      "learning_rate": 2.0624772763630546e-05,
-      "loss": 0.5816,
+      "epoch": 3.470116184853801,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.062073704570905e-05,
+      "loss": 0.9813436889648437,
       "step": 301400
     },
     {
-      "epoch": 6.9448282214239665,
-      "grad_norm": 2.7301547527313232,
-      "learning_rate": 2.061750119980803e-05,
-      "loss": 0.5909,
+      "epoch": 3.472418869928516,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0613465481886533e-05,
+      "loss": 0.9656988525390625,
       "step": 301600
     },
     {
-      "epoch": 6.949433545178226,
-      "grad_norm": 2.843813896179199,
-      "learning_rate": 2.0610229635985517e-05,
-      "loss": 0.5987,
+      "epoch": 3.474721555003231,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.060619391806402e-05,
+      "loss": 0.9734279632568359,
       "step": 301800
     },
     {
-      "epoch": 6.954038868932486,
-      "grad_norm": 3.3553197383880615,
-      "learning_rate": 2.0602958072163e-05,
-      "loss": 0.5974,
+      "epoch": 3.477024240077946,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.0598922354241505e-05,
+      "loss": 0.9843352508544921,
       "step": 302000
     },
     {
-      "epoch": 6.958644192686746,
-      "grad_norm": 2.880885601043701,
-      "learning_rate": 2.0595686508340486e-05,
-      "loss": 0.6037,
+      "epoch": 3.4793269251526606,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.0591650790418988e-05,
+      "loss": 0.9777169036865234,
       "step": 302200
     },
     {
-      "epoch": 6.963249516441006,
-      "grad_norm": 4.332859039306641,
-      "learning_rate": 2.058841494451797e-05,
-      "loss": 0.5863,
+      "epoch": 3.4816296102273756,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0584379226596473e-05,
+      "loss": 0.9745001220703124,
       "step": 302400
     },
     {
-      "epoch": 6.967854840195265,
-      "grad_norm": 2.701078176498413,
-      "learning_rate": 2.058114338069545e-05,
-      "loss": 0.6154,
+      "epoch": 3.4839322953020906,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0577107662773956e-05,
+      "loss": 0.9934710693359375,
       "step": 302600
     },
     {
-      "epoch": 6.972460163949526,
-      "grad_norm": 3.1130638122558594,
-      "learning_rate": 2.0573871816872937e-05,
-      "loss": 0.5895,
+      "epoch": 3.4862349803768056,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.056983609895144e-05,
+      "loss": 1.0010362243652344,
       "step": 302800
     },
     {
-      "epoch": 6.977065487703785,
-      "grad_norm": 3.547954559326172,
-      "learning_rate": 2.0566600253050423e-05,
-      "loss": 0.6219,
+      "epoch": 3.4885376654515206,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.0562564535128924e-05,
+      "loss": 0.9597031402587891,
       "step": 303000
     },
     {
-      "epoch": 6.981670811458045,
-      "grad_norm": 2.777357816696167,
-      "learning_rate": 2.0559328689227905e-05,
-      "loss": 0.5969,
+      "epoch": 3.4908403505262355,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.055529297130641e-05,
+      "loss": 0.9741526031494141,
       "step": 303200
     },
     {
-      "epoch": 6.9862761352123055,
-      "grad_norm": 3.05338978767395,
-      "learning_rate": 2.055205712540539e-05,
-      "loss": 0.5945,
+      "epoch": 3.4931430356009505,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.0548021407483896e-05,
+      "loss": 0.9581996917724609,
       "step": 303400
     },
     {
-      "epoch": 6.990881458966565,
-      "grad_norm": 2.9107885360717773,
-      "learning_rate": 2.0544785561582877e-05,
-      "loss": 0.5978,
+      "epoch": 3.4954457206756655,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.054074984366138e-05,
+      "loss": 0.982119369506836,
       "step": 303600
     },
     {
-      "epoch": 6.995486782720826,
-      "grad_norm": 3.5367624759674072,
-      "learning_rate": 2.0537550355579472e-05,
-      "loss": 0.6023,
+      "epoch": 3.4977484057503805,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.0533478279838865e-05,
+      "loss": 0.9806587219238281,
       "step": 303800
     },
     {
-      "epoch": 7.0,
-      "eval_loss": 0.567062258720398,
-      "eval_runtime": 145.2794,
-      "eval_samples_per_second": 195.217,
-      "eval_steps_per_second": 12.204,
-      "step": 303996
-    },
-    {
-      "epoch": 7.000092106475085,
-      "grad_norm": 4.047206878662109,
-      "learning_rate": 2.0530278791756955e-05,
-      "loss": 0.5904,
+      "epoch": 3.500051090825095,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0526206716016347e-05,
+      "loss": 0.989941635131836,
       "step": 304000
     },
     {
-      "epoch": 7.004697430229345,
-      "grad_norm": 2.5840604305267334,
-      "learning_rate": 2.052300722793444e-05,
-      "loss": 0.5862,
+      "epoch": 3.50235377589981,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.051893515219383e-05,
+      "loss": 1.0165106201171874,
       "step": 304200
     },
     {
-      "epoch": 7.009302753983605,
-      "grad_norm": 3.08518385887146,
-      "learning_rate": 2.0515735664111923e-05,
-      "loss": 0.5918,
+      "epoch": 3.504656460974525,
+      "grad_norm": 1.0,
+      "learning_rate": 2.0511663588371316e-05,
+      "loss": 0.9855970764160156,
       "step": 304400
     },
     {
-      "epoch": 7.013908077737865,
-      "grad_norm": 2.9152956008911133,
-      "learning_rate": 2.050846410028941e-05,
-      "loss": 0.5858,
+      "epoch": 3.50695914604924,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.05043920245488e-05,
+      "loss": 0.9602165222167969,
       "step": 304600
     },
     {
-      "epoch": 7.018513401492125,
-      "grad_norm": 3.358140468597412,
-      "learning_rate": 2.0501192536466895e-05,
-      "loss": 0.5958,
+      "epoch": 3.509261831123955,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.0497120460726284e-05,
+      "loss": 0.976239013671875,
       "step": 304800
     },
     {
-      "epoch": 7.023118725246385,
-      "grad_norm": 2.961735248565674,
-      "learning_rate": 2.0493920972644378e-05,
-      "loss": 0.5929,
+      "epoch": 3.51156451619867,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.048984889690377e-05,
+      "loss": 0.9843711853027344,
       "step": 305000
     },
     {
-      "epoch": 7.0277240490006445,
-      "grad_norm": 2.8176980018615723,
-      "learning_rate": 2.048664940882186e-05,
-      "loss": 0.5862,
+      "epoch": 3.513867201273385,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.0482577333081253e-05,
+      "loss": 0.9723324584960937,
       "step": 305200
     },
     {
-      "epoch": 7.032329372754905,
-      "grad_norm": 3.275162696838379,
-      "learning_rate": 2.0479377844999346e-05,
-      "loss": 0.5934,
+      "epoch": 3.5161698863481,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0475305769258735e-05,
+      "loss": 0.9799480438232422,
       "step": 305400
     },
     {
-      "epoch": 7.036934696509165,
-      "grad_norm": 3.5453879833221436,
-      "learning_rate": 2.047210628117683e-05,
-      "loss": 0.588,
+      "epoch": 3.518472571422815,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.046803420543622e-05,
+      "loss": 1.0050213623046875,
       "step": 305600
     },
     {
-      "epoch": 7.041540020263424,
-      "grad_norm": 2.8470346927642822,
-      "learning_rate": 2.0464834717354314e-05,
-      "loss": 0.597,
+      "epoch": 3.5207752564975294,
+      "grad_norm": 1.125,
+      "learning_rate": 2.0460762641613707e-05,
+      "loss": 0.969159164428711,
       "step": 305800
     },
     {
-      "epoch": 7.046145344017685,
-      "grad_norm": 2.6055123805999756,
-      "learning_rate": 2.0457599511350913e-05,
-      "loss": 0.591,
+      "epoch": 3.523077941572245,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.045349107779119e-05,
+      "loss": 0.9573507690429688,
       "step": 306000
     },
     {
-      "epoch": 7.050750667771944,
-      "grad_norm": 2.6110992431640625,
-      "learning_rate": 2.0450327947528396e-05,
-      "loss": 0.5906,
+      "epoch": 3.5253806266469594,
+      "grad_norm": 0.75,
+      "learning_rate": 2.0446219513968675e-05,
+      "loss": 1.0036588287353516,
       "step": 306200
     },
     {
-      "epoch": 7.055355991526204,
-      "grad_norm": 3.3166911602020264,
-      "learning_rate": 2.044305638370588e-05,
-      "loss": 0.582,
+      "epoch": 3.5276833117216744,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.0438947950146158e-05,
+      "loss": 0.9807686614990234,
       "step": 306400
     },
     {
-      "epoch": 7.059961315280464,
-      "grad_norm": 2.7513551712036133,
-      "learning_rate": 2.0435784819883364e-05,
-      "loss": 0.5985,
+      "epoch": 3.5299859967963894,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.0431676386323644e-05,
+      "loss": 0.9778697204589843,
       "step": 306600
     },
     {
-      "epoch": 7.064566639034724,
-      "grad_norm": 3.547184944152832,
-      "learning_rate": 2.0428513256060847e-05,
-      "loss": 0.6041,
+      "epoch": 3.5322886818711043,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.0424404822501126e-05,
+      "loss": 0.9704287719726562,
       "step": 306800
     },
     {
-      "epoch": 7.069171962788984,
-      "grad_norm": 3.074101686477661,
-      "learning_rate": 2.0421241692238332e-05,
-      "loss": 0.5915,
+      "epoch": 3.5345913669458193,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.0417133258678612e-05,
+      "loss": 0.9505149841308593,
       "step": 307000
     },
     {
-      "epoch": 7.073777286543244,
-      "grad_norm": 3.347473621368408,
-      "learning_rate": 2.041397012841582e-05,
-      "loss": 0.5927,
+      "epoch": 3.5368940520205343,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0409861694856098e-05,
+      "loss": 0.9952140045166016,
       "step": 307200
     },
     {
-      "epoch": 7.078382610297504,
-      "grad_norm": 3.3471734523773193,
-      "learning_rate": 2.04066985645933e-05,
-      "loss": 0.5954,
+      "epoch": 3.5391967370952493,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.040259013103358e-05,
+      "loss": 0.9753850555419922,
       "step": 307400
     },
     {
-      "epoch": 7.082987934051764,
-      "grad_norm": 2.5890464782714844,
-      "learning_rate": 2.0399427000770787e-05,
-      "loss": 0.6003,
+      "epoch": 3.541499422169964,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0395318567211067e-05,
+      "loss": 0.9734141540527343,
       "step": 307600
     },
     {
-      "epoch": 7.087593257806024,
-      "grad_norm": 2.5832886695861816,
-      "learning_rate": 2.0392155436948273e-05,
-      "loss": 0.5964,
+      "epoch": 3.5438021072446793,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.038804700338855e-05,
+      "loss": 0.9796247100830078,
       "step": 307800
     },
     {
-      "epoch": 7.092198581560283,
-      "grad_norm": 3.1692888736724854,
-      "learning_rate": 2.0384883873125755e-05,
-      "loss": 0.5985,
+      "epoch": 3.546104792319394,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.038077543956603e-05,
+      "loss": 0.984339599609375,
       "step": 308000
     },
     {
-      "epoch": 7.096803905314544,
-      "grad_norm": 2.8189423084259033,
-      "learning_rate": 2.0377612309303238e-05,
-      "loss": 0.5802,
+      "epoch": 3.5484074773941088,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.0373503875743517e-05,
+      "loss": 0.9877261352539063,
       "step": 308200
     },
     {
-      "epoch": 7.101409229068803,
-      "grad_norm": 2.9584109783172607,
-      "learning_rate": 2.0370340745480724e-05,
-      "loss": 0.5949,
+      "epoch": 3.5507101624688238,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0366232311921003e-05,
+      "loss": 0.9739342498779296,
       "step": 308400
     },
     {
-      "epoch": 7.106014552823064,
-      "grad_norm": 2.7427165508270264,
-      "learning_rate": 2.036306918165821e-05,
-      "loss": 0.59,
+      "epoch": 3.5530128475435387,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.0358960748098486e-05,
+      "loss": 0.9697119140625,
       "step": 308600
     },
     {
-      "epoch": 7.110619876577323,
-      "grad_norm": 3.3998868465423584,
-      "learning_rate": 2.0355797617835692e-05,
-      "loss": 0.5907,
+      "epoch": 3.5553155326182537,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.0351689184275972e-05,
+      "loss": 0.9906707763671875,
       "step": 308800
     },
     {
-      "epoch": 7.115225200331583,
-      "grad_norm": 3.8071186542510986,
-      "learning_rate": 2.0348526054013178e-05,
-      "loss": 0.5865,
+      "epoch": 3.5576182176929687,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0344417620453454e-05,
+      "loss": 0.986513671875,
       "step": 309000
     },
     {
-      "epoch": 7.1198305240858435,
-      "grad_norm": 2.8909687995910645,
-      "learning_rate": 2.0341290848009773e-05,
-      "loss": 0.5883,
+      "epoch": 3.5599209027676837,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.033714605663094e-05,
+      "loss": 0.9764996337890625,
       "step": 309200
     },
     {
-      "epoch": 7.124435847840103,
-      "grad_norm": 2.901247262954712,
-      "learning_rate": 2.0334019284187256e-05,
-      "loss": 0.6004,
+      "epoch": 3.5622235878423982,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.0329874492808423e-05,
+      "loss": 0.990147705078125,
       "step": 309400
     },
     {
-      "epoch": 7.129041171594363,
-      "grad_norm": 3.5564959049224854,
-      "learning_rate": 2.0326747720364742e-05,
-      "loss": 0.5866,
+      "epoch": 3.5645262729171137,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.032260292898591e-05,
+      "loss": 0.9730393981933594,
       "step": 309600
     },
     {
-      "epoch": 7.133646495348623,
-      "grad_norm": 3.2295401096343994,
-      "learning_rate": 2.0319476156542228e-05,
-      "loss": 0.6012,
+      "epoch": 3.566828957991828,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.0315331365163395e-05,
+      "loss": 0.9966936492919922,
       "step": 309800
     },
     {
-      "epoch": 7.138251819102883,
-      "grad_norm": 2.9399068355560303,
-      "learning_rate": 2.031220459271971e-05,
-      "loss": 0.6112,
+      "epoch": 3.569131643066543,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.0308059801340877e-05,
+      "loss": 0.9686882781982422,
       "step": 310000
     },
     {
-      "epoch": 7.142857142857143,
-      "grad_norm": 3.2517316341400146,
-      "learning_rate": 2.0304933028897196e-05,
-      "loss": 0.5686,
+      "epoch": 3.571434328141258,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.030078823751836e-05,
+      "loss": 0.9749455261230469,
       "step": 310200
     },
     {
-      "epoch": 7.147462466611403,
-      "grad_norm": 2.8699967861175537,
-      "learning_rate": 2.0297661465074682e-05,
-      "loss": 0.5826,
+      "epoch": 3.573737013215973,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.0293516673695846e-05,
+      "loss": 0.9812612152099609,
       "step": 310400
     },
     {
-      "epoch": 7.152067790365662,
-      "grad_norm": 3.1824159622192383,
-      "learning_rate": 2.0290389901252165e-05,
-      "loss": 0.5878,
+      "epoch": 3.576039698290688,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0286245109873328e-05,
+      "loss": 0.9691918182373047,
       "step": 310600
     },
     {
-      "epoch": 7.156673114119923,
-      "grad_norm": 2.710801839828491,
-      "learning_rate": 2.0283118337429647e-05,
-      "loss": 0.5989,
+      "epoch": 3.578342383365403,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.0278973546050814e-05,
+      "loss": 0.9892355346679688,
       "step": 310800
     },
     {
-      "epoch": 7.1612784378741825,
-      "grad_norm": 3.2563531398773193,
-      "learning_rate": 2.0275846773607133e-05,
-      "loss": 0.5956,
+      "epoch": 3.580645068440118,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.02717019822283e-05,
+      "loss": 0.9764984130859375,
       "step": 311000
     },
     {
-      "epoch": 7.165883761628442,
-      "grad_norm": 3.0747923851013184,
-      "learning_rate": 2.0268575209784615e-05,
-      "loss": 0.5924,
+      "epoch": 3.5829477535148326,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0264430418405782e-05,
+      "loss": 0.9705098724365234,
       "step": 311200
     },
     {
-      "epoch": 7.170489085382703,
-      "grad_norm": 3.374802589416504,
-      "learning_rate": 2.02613036459621e-05,
-      "loss": 0.594,
+      "epoch": 3.585250438589548,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.025715885458327e-05,
+      "loss": 0.9755960845947266,
       "step": 311400
     },
     {
-      "epoch": 7.175094409136962,
-      "grad_norm": 3.170912504196167,
-      "learning_rate": 2.0254032082139587e-05,
-      "loss": 0.5821,
+      "epoch": 3.5875531236642626,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.024988729076075e-05,
+      "loss": 0.9599433898925781,
       "step": 311600
     },
     {
-      "epoch": 7.179699732891223,
-      "grad_norm": 2.800078868865967,
-      "learning_rate": 2.024676051831707e-05,
-      "loss": 0.5837,
+      "epoch": 3.5898558087389776,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.0242615726938233e-05,
+      "loss": 0.9769892883300781,
       "step": 311800
     },
     {
-      "epoch": 7.184305056645482,
-      "grad_norm": 4.168891906738281,
-      "learning_rate": 2.0239488954494552e-05,
-      "loss": 0.5977,
+      "epoch": 3.5921584938136926,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.023534416311572e-05,
+      "loss": 1.0197854614257813,
       "step": 312000
     },
     {
-      "epoch": 7.188910380399742,
-      "grad_norm": 4.314828872680664,
-      "learning_rate": 2.023225374849115e-05,
-      "loss": 0.5871,
+      "epoch": 3.5944611788884075,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0228072599293205e-05,
+      "loss": 0.9848535919189453,
       "step": 312200
     },
     {
-      "epoch": 7.193515704154002,
-      "grad_norm": 2.661975383758545,
-      "learning_rate": 2.0224982184668633e-05,
-      "loss": 0.6009,
+      "epoch": 3.5967638639631225,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.022080103547069e-05,
+      "loss": 0.9727433776855469,
       "step": 312400
     },
     {
-      "epoch": 7.198121027908262,
-      "grad_norm": 2.217339277267456,
-      "learning_rate": 2.021771062084612e-05,
-      "loss": 0.5916,
+      "epoch": 3.5990665490378375,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.0213529471648174e-05,
+      "loss": 0.9948325347900391,
       "step": 312600
     },
     {
-      "epoch": 7.2027263516625215,
-      "grad_norm": 3.5877535343170166,
-      "learning_rate": 2.0210439057023605e-05,
-      "loss": 0.5889,
+      "epoch": 3.6013692341125525,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.0206257907825656e-05,
+      "loss": 0.9774640655517578,
       "step": 312800
     },
     {
-      "epoch": 7.207331675416782,
-      "grad_norm": 3.6511406898498535,
-      "learning_rate": 2.0203167493201088e-05,
-      "loss": 0.5906,
+      "epoch": 3.603671919187267,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.0198986344003142e-05,
+      "loss": 0.9790972900390625,
       "step": 313000
     },
     {
-      "epoch": 7.211936999171042,
-      "grad_norm": 3.414182186126709,
-      "learning_rate": 2.0195895929378574e-05,
-      "loss": 0.5968,
+      "epoch": 3.6059746042619825,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.0191714780180625e-05,
+      "loss": 0.9803644561767578,
       "step": 313200
     },
     {
-      "epoch": 7.216542322925302,
-      "grad_norm": 3.0940256118774414,
-      "learning_rate": 2.0188624365556056e-05,
-      "loss": 0.5972,
+      "epoch": 3.608277289336697,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.018444321635811e-05,
+      "loss": 0.9680404663085938,
       "step": 313400
     },
     {
-      "epoch": 7.221147646679562,
-      "grad_norm": 2.9856083393096924,
-      "learning_rate": 2.018135280173354e-05,
-      "loss": 0.5918,
+      "epoch": 3.610579974411412,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.0177171652535597e-05,
+      "loss": 0.9820728302001953,
       "step": 313600
     },
     {
-      "epoch": 7.225752970433821,
-      "grad_norm": 3.411344528198242,
-      "learning_rate": 2.0174081237911025e-05,
-      "loss": 0.5816,
+      "epoch": 3.612882659486127,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.016990008871308e-05,
+      "loss": 1.0057897186279297,
       "step": 313800
     },
     {
-      "epoch": 7.230358294188082,
-      "grad_norm": 3.6355364322662354,
-      "learning_rate": 2.016680967408851e-05,
-      "loss": 0.6008,
+      "epoch": 3.615185344560842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0162628524890565e-05,
+      "loss": 0.9696613311767578,
       "step": 314000
     },
     {
-      "epoch": 7.234963617942341,
-      "grad_norm": 3.124440908432007,
-      "learning_rate": 2.0159538110265993e-05,
-      "loss": 0.5892,
+      "epoch": 3.617488029635557,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.0155356961068047e-05,
+      "loss": 0.9803009033203125,
       "step": 314200
     },
     {
-      "epoch": 7.239568941696601,
-      "grad_norm": 3.7005481719970703,
-      "learning_rate": 2.015226654644348e-05,
-      "loss": 0.5834,
+      "epoch": 3.619790714710272,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.014808539724553e-05,
+      "loss": 0.9822318267822265,
       "step": 314400
     },
     {
-      "epoch": 7.244174265450861,
-      "grad_norm": 2.2314252853393555,
-      "learning_rate": 2.0144994982620965e-05,
-      "loss": 0.5888,
+      "epoch": 3.622093399784987,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.0140813833423016e-05,
+      "loss": 0.985562744140625,
       "step": 314600
     },
     {
-      "epoch": 7.248779589205121,
-      "grad_norm": 3.4673757553100586,
-      "learning_rate": 2.0137723418798448e-05,
-      "loss": 0.6033,
+      "epoch": 3.6243960848597014,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.0133542269600502e-05,
+      "loss": 0.9913748168945312,
       "step": 314800
     },
     {
-      "epoch": 7.2533849129593815,
-      "grad_norm": 3.7593634128570557,
-      "learning_rate": 2.013045185497593e-05,
-      "loss": 0.6022,
+      "epoch": 3.626698769934417,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.0126270705777984e-05,
+      "loss": 0.974439697265625,
       "step": 315000
     },
     {
-      "epoch": 7.257990236713641,
-      "grad_norm": 3.1982200145721436,
-      "learning_rate": 2.0123180291153416e-05,
-      "loss": 0.5823,
+      "epoch": 3.6290014550091314,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.011899914195547e-05,
+      "loss": 0.972523193359375,
       "step": 315200
     },
     {
-      "epoch": 7.262595560467901,
-      "grad_norm": 2.9602959156036377,
-      "learning_rate": 2.0115908727330902e-05,
-      "loss": 0.5856,
+      "epoch": 3.6313041400838464,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0111727578132953e-05,
+      "loss": 0.9973404693603516,
       "step": 315400
     },
     {
-      "epoch": 7.267200884222161,
-      "grad_norm": 3.3312976360321045,
-      "learning_rate": 2.0108637163508384e-05,
-      "loss": 0.5826,
+      "epoch": 3.6336068251585614,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.010445601431044e-05,
+      "loss": 0.9800202941894531,
       "step": 315600
     },
     {
-      "epoch": 7.271806207976421,
-      "grad_norm": 3.437633514404297,
-      "learning_rate": 2.010136559968587e-05,
-      "loss": 0.5889,
+      "epoch": 3.6359095102332764,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.009718445048792e-05,
+      "loss": 0.9633809661865235,
       "step": 315800
     },
     {
-      "epoch": 7.27641153173068,
-      "grad_norm": 4.305769443511963,
-      "learning_rate": 2.0094094035863353e-05,
-      "loss": 0.5868,
+      "epoch": 3.6382121953079913,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0089912886665407e-05,
+      "loss": 0.9927108001708984,
       "step": 316000
     },
     {
-      "epoch": 7.281016855484941,
-      "grad_norm": 3.3740241527557373,
-      "learning_rate": 2.0086822472040835e-05,
-      "loss": 0.5826,
+      "epoch": 3.6405148803827063,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0082641322842893e-05,
+      "loss": 0.9627687072753907,
       "step": 316200
     },
     {
-      "epoch": 7.2856221792392,
-      "grad_norm": 2.570251226425171,
-      "learning_rate": 2.0079623623856547e-05,
-      "loss": 0.5914,
+      "epoch": 3.6428175654574213,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.0075369759020376e-05,
+      "loss": 0.9577132415771484,
       "step": 316400
     },
     {
-      "epoch": 7.29022750299346,
-      "grad_norm": 4.058407783508301,
-      "learning_rate": 2.007235206003403e-05,
-      "loss": 0.6015,
+      "epoch": 3.6451202505321363,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0068098195197858e-05,
+      "loss": 0.9758236694335938,
       "step": 316600
     },
     {
-      "epoch": 7.2948328267477205,
-      "grad_norm": 3.7470202445983887,
-      "learning_rate": 2.0065080496211515e-05,
-      "loss": 0.5866,
+      "epoch": 3.6474229356068513,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.0060826631375344e-05,
+      "loss": 0.9763569641113281,
       "step": 316800
     },
     {
-      "epoch": 7.29943815050198,
-      "grad_norm": 3.243685007095337,
-      "learning_rate": 2.0057808932389e-05,
-      "loss": 0.5829,
+      "epoch": 3.649725620681566,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.0053555067552827e-05,
+      "loss": 0.9699946594238281,
       "step": 317000
     },
     {
-      "epoch": 7.304043474256241,
-      "grad_norm": 3.5165421962738037,
-      "learning_rate": 2.0050537368566484e-05,
-      "loss": 0.5883,
+      "epoch": 3.652028305756281,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.0046283503730312e-05,
+      "loss": 0.988985824584961,
       "step": 317200
     },
     {
-      "epoch": 7.3086487980105,
-      "grad_norm": 2.5565083026885986,
-      "learning_rate": 2.004326580474397e-05,
-      "loss": 0.5944,
+      "epoch": 3.6543309908309958,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.00390119399078e-05,
+      "loss": 0.9836240386962891,
       "step": 317400
     },
     {
-      "epoch": 7.31325412176476,
-      "grad_norm": 2.588308095932007,
-      "learning_rate": 2.0035994240921455e-05,
-      "loss": 0.5903,
+      "epoch": 3.6566336759057108,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.003174037608528e-05,
+      "loss": 0.9703248596191406,
       "step": 317600
     },
     {
-      "epoch": 7.31785944551902,
-      "grad_norm": 3.13081431388855,
-      "learning_rate": 2.0028722677098938e-05,
-      "loss": 0.5962,
+      "epoch": 3.6589363609804257,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.0024468812262767e-05,
+      "loss": 0.9764913940429687,
       "step": 317800
     },
     {
-      "epoch": 7.32246476927328,
-      "grad_norm": 3.2789556980133057,
-      "learning_rate": 2.002145111327642e-05,
-      "loss": 0.5912,
+      "epoch": 3.6612390460551407,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.001719724844025e-05,
+      "loss": 0.9827780914306641,
       "step": 318000
     },
     {
-      "epoch": 7.327070093027539,
-      "grad_norm": 2.885380744934082,
-      "learning_rate": 2.0014179549453906e-05,
-      "loss": 0.5992,
+      "epoch": 3.6635417311298557,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.0009925684617735e-05,
+      "loss": 0.9800835418701171,
       "step": 318200
     },
     {
-      "epoch": 7.3316754167818,
-      "grad_norm": 2.445709705352783,
-      "learning_rate": 2.0006907985631392e-05,
-      "loss": 0.5915,
+      "epoch": 3.6658444162045707,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.0002654120795218e-05,
+      "loss": 0.9786171722412109,
       "step": 318400
     },
     {
-      "epoch": 7.3362807405360595,
-      "grad_norm": 4.696218967437744,
-      "learning_rate": 1.9999636421808875e-05,
-      "loss": 0.602,
+      "epoch": 3.6681471012792857,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9995382556972704e-05,
+      "loss": 0.99401611328125,
       "step": 318600
     },
     {
-      "epoch": 7.34088606429032,
-      "grad_norm": 2.6832456588745117,
-      "learning_rate": 1.999236485798636e-05,
-      "loss": 0.5936,
+      "epoch": 3.670449786354,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.998811099315019e-05,
+      "loss": 0.9821390533447265,
       "step": 318800
     },
     {
-      "epoch": 7.34549138804458,
-      "grad_norm": 2.7151763439178467,
-      "learning_rate": 1.9985093294163843e-05,
-      "loss": 0.5932,
+      "epoch": 3.672752471428715,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9980839429327672e-05,
+      "loss": 0.9792152404785156,
       "step": 319000
     },
     {
-      "epoch": 7.350096711798839,
-      "grad_norm": 2.803915500640869,
-      "learning_rate": 1.9977821730341326e-05,
-      "loss": 0.6016,
+      "epoch": 3.67505515650343,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9973567865505155e-05,
+      "loss": 0.9921223449707032,
       "step": 319200
     },
     {
-      "epoch": 7.3547020355531,
-      "grad_norm": 2.600085973739624,
-      "learning_rate": 1.997055016651881e-05,
-      "loss": 0.5952,
+      "epoch": 3.677357841578145,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.996629630168264e-05,
+      "loss": 0.984896011352539,
       "step": 319400
     },
     {
-      "epoch": 7.359307359307359,
-      "grad_norm": 3.053616523742676,
-      "learning_rate": 1.9963278602696298e-05,
-      "loss": 0.5924,
+      "epoch": 3.67966052665286,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9959024737860123e-05,
+      "loss": 0.9692173004150391,
       "step": 319600
     },
     {
-      "epoch": 7.363912683061619,
-      "grad_norm": 2.9440255165100098,
-      "learning_rate": 1.995600703887378e-05,
-      "loss": 0.5995,
+      "epoch": 3.681963211727575,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.995175317403761e-05,
+      "loss": 0.9897268676757812,
       "step": 319800
     },
     {
-      "epoch": 7.368518006815879,
-      "grad_norm": 3.2415082454681396,
-      "learning_rate": 1.9948735475051266e-05,
-      "loss": 0.5919,
+      "epoch": 3.68426589680229,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9944481610215095e-05,
+      "loss": 0.9933697509765625,
       "step": 320000
     },
     {
-      "epoch": 7.373123330570139,
-      "grad_norm": 2.619912624359131,
-      "learning_rate": 1.994146391122875e-05,
-      "loss": 0.5929,
+      "epoch": 3.686568581877005,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9937210046392577e-05,
+      "loss": 0.9815640258789062,
       "step": 320200
     },
     {
-      "epoch": 7.377728654324399,
-      "grad_norm": 3.2725462913513184,
-      "learning_rate": 1.993419234740623e-05,
-      "loss": 0.5849,
+      "epoch": 3.68887126695172,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9929938482570063e-05,
+      "loss": 0.9922060394287109,
       "step": 320400
     },
     {
-      "epoch": 7.382333978078659,
-      "grad_norm": 4.5242743492126465,
-      "learning_rate": 1.9926920783583717e-05,
-      "loss": 0.5891,
+      "epoch": 3.6911739520264346,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9922666918747546e-05,
+      "loss": 1.0016986083984376,
       "step": 320600
     },
     {
-      "epoch": 7.386939301832919,
-      "grad_norm": 3.2568304538726807,
-      "learning_rate": 1.9919649219761203e-05,
-      "loss": 0.5892,
+      "epoch": 3.6934766371011496,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.991539535492503e-05,
+      "loss": 0.9736055755615234,
       "step": 320800
     },
     {
-      "epoch": 7.391544625587179,
-      "grad_norm": 3.1149260997772217,
-      "learning_rate": 1.9912414013757798e-05,
-      "loss": 0.5955,
+      "epoch": 3.6957793221758646,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9908123791102514e-05,
+      "loss": 0.9745178985595703,
       "step": 321000
     },
     {
-      "epoch": 7.396149949341439,
-      "grad_norm": 3.272343158721924,
-      "learning_rate": 1.9905142449935284e-05,
-      "loss": 0.5938,
+      "epoch": 3.6980820072505796,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.990085222728e-05,
+      "loss": 0.9692243194580078,
       "step": 321200
     },
     {
-      "epoch": 7.400755273095698,
-      "grad_norm": 2.8861818313598633,
-      "learning_rate": 1.989787088611277e-05,
-      "loss": 0.5973,
+      "epoch": 3.7003846923252945,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9893580663457486e-05,
+      "loss": 0.9790972137451172,
       "step": 321400
     },
     {
-      "epoch": 7.405360596849959,
-      "grad_norm": 4.322711944580078,
-      "learning_rate": 1.9890599322290252e-05,
-      "loss": 0.5979,
+      "epoch": 3.7026873774000095,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.988630909963497e-05,
+      "loss": 0.9805580139160156,
       "step": 321600
     },
     {
-      "epoch": 7.409965920604218,
-      "grad_norm": 3.2034659385681152,
-      "learning_rate": 1.9883327758467735e-05,
-      "loss": 0.5897,
+      "epoch": 3.7049900624747245,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.987903753581245e-05,
+      "loss": 0.9620713806152343,
       "step": 321800
     },
     {
-      "epoch": 7.414571244358479,
-      "grad_norm": 2.6252517700195312,
-      "learning_rate": 1.987605619464522e-05,
-      "loss": 0.593,
+      "epoch": 3.7072927475494395,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9871765971989937e-05,
+      "loss": 0.9783279418945312,
       "step": 322000
     },
     {
-      "epoch": 7.419176568112738,
-      "grad_norm": 2.79892897605896,
-      "learning_rate": 1.9868784630822703e-05,
-      "loss": 0.5988,
+      "epoch": 3.7095954326241545,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.986449440816742e-05,
+      "loss": 0.9842681884765625,
       "step": 322200
     },
     {
-      "epoch": 7.423781891866998,
-      "grad_norm": 4.176153659820557,
-      "learning_rate": 1.986151306700019e-05,
-      "loss": 0.5828,
+      "epoch": 3.711898117698869,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9857222844344906e-05,
+      "loss": 0.9877159881591797,
       "step": 322400
     },
     {
-      "epoch": 7.4283872156212585,
-      "grad_norm": 4.084776878356934,
-      "learning_rate": 1.9854241503177675e-05,
-      "loss": 0.5949,
+      "epoch": 3.7142008027735844,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.984995128052239e-05,
+      "loss": 0.9558129119873047,
       "step": 322600
     },
     {
-      "epoch": 7.432992539375518,
-      "grad_norm": 3.2946360111236572,
-      "learning_rate": 1.984696993935516e-05,
-      "loss": 0.5911,
+      "epoch": 3.716503487848299,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9842679716699874e-05,
+      "loss": 0.964300537109375,
       "step": 322800
     },
     {
-      "epoch": 7.437597863129778,
-      "grad_norm": 2.784147262573242,
-      "learning_rate": 1.983969837553264e-05,
-      "loss": 0.5864,
+      "epoch": 3.718806172923014,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9835408152877357e-05,
+      "loss": 0.9851215362548829,
       "step": 323000
     },
     {
-      "epoch": 7.442203186884038,
-      "grad_norm": 3.3848652839660645,
-      "learning_rate": 1.9832426811710126e-05,
-      "loss": 0.579,
+      "epoch": 3.721108857997729,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9828136589054842e-05,
+      "loss": 0.9854248046875,
       "step": 323200
     },
     {
-      "epoch": 7.446808510638298,
-      "grad_norm": 3.4433703422546387,
-      "learning_rate": 1.9825155247887612e-05,
-      "loss": 0.5887,
+      "epoch": 3.723411543072444,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9820865025232325e-05,
+      "loss": 0.981217041015625,
       "step": 323400
     },
     {
-      "epoch": 7.451413834392557,
-      "grad_norm": 3.3405020236968994,
-      "learning_rate": 1.9817920041884207e-05,
-      "loss": 0.5876,
+      "epoch": 3.725714228147159,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.981359346140981e-05,
+      "loss": 0.9648824310302735,
       "step": 323600
     },
     {
-      "epoch": 7.456019158146818,
-      "grad_norm": 3.7312204837799072,
-      "learning_rate": 1.9810648478061693e-05,
-      "loss": 0.5955,
+      "epoch": 3.728016913221874,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9806321897587297e-05,
+      "loss": 0.9728410339355469,
       "step": 323800
     },
     {
-      "epoch": 7.460624481901077,
-      "grad_norm": 4.358093738555908,
-      "learning_rate": 1.9803376914239176e-05,
-      "loss": 0.5866,
+      "epoch": 3.730319598296589,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9799050333764783e-05,
+      "loss": 0.9918624114990234,
       "step": 324000
     },
     {
-      "epoch": 7.465229805655338,
-      "grad_norm": 2.68795108795166,
-      "learning_rate": 1.979610535041666e-05,
-      "loss": 0.6004,
+      "epoch": 3.7326222833713034,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9791778769942265e-05,
+      "loss": 0.9825603485107421,
       "step": 324200
     },
     {
-      "epoch": 7.4698351294095975,
-      "grad_norm": 2.8480021953582764,
-      "learning_rate": 1.978887014441326e-05,
-      "loss": 0.5983,
+      "epoch": 3.734924968446019,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.9784507206119748e-05,
+      "loss": 0.987041015625,
       "step": 324400
     },
     {
-      "epoch": 7.474440453163857,
-      "grad_norm": 4.532684803009033,
-      "learning_rate": 1.9781598580590743e-05,
-      "loss": 0.5917,
+      "epoch": 3.7372276535207334,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9777235642297234e-05,
+      "loss": 0.9773162841796875,
       "step": 324600
     },
     {
-      "epoch": 7.479045776918118,
-      "grad_norm": 3.242600202560425,
-      "learning_rate": 1.9774327016768225e-05,
-      "loss": 0.5981,
+      "epoch": 3.7395303385954484,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9769964078474716e-05,
+      "loss": 0.980873031616211,
       "step": 324800
     },
     {
-      "epoch": 7.483651100672377,
-      "grad_norm": 3.071572780609131,
-      "learning_rate": 1.976705545294571e-05,
-      "loss": 0.5892,
+      "epoch": 3.7418330236701633,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9762692514652202e-05,
+      "loss": 0.9816900634765625,
       "step": 325000
     },
     {
-      "epoch": 7.488256424426638,
-      "grad_norm": 3.3507227897644043,
-      "learning_rate": 1.9759783889123194e-05,
-      "loss": 0.593,
+      "epoch": 3.7441357087448783,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9755420950829688e-05,
+      "loss": 0.9802275085449219,
       "step": 325200
     },
     {
-      "epoch": 7.492861748180897,
-      "grad_norm": 3.0197222232818604,
-      "learning_rate": 1.975251232530068e-05,
-      "loss": 0.598,
+      "epoch": 3.7464383938195933,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.974814938700717e-05,
+      "loss": 0.9866002655029297,
       "step": 325400
     },
     {
-      "epoch": 7.497467071935157,
-      "grad_norm": 2.7512412071228027,
-      "learning_rate": 1.9745240761478166e-05,
-      "loss": 0.5942,
+      "epoch": 3.7487410788943083,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9740877823184653e-05,
+      "loss": 0.9880412292480468,
       "step": 325600
     },
     {
-      "epoch": 7.502072395689417,
-      "grad_norm": 3.136456251144409,
-      "learning_rate": 1.9737969197655648e-05,
-      "loss": 0.5828,
+      "epoch": 3.7510437639690233,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.973360625936214e-05,
+      "loss": 0.9862410736083984,
       "step": 325800
     },
     {
-      "epoch": 7.506677719443677,
-      "grad_norm": 4.431463241577148,
-      "learning_rate": 1.973069763383313e-05,
-      "loss": 0.6002,
+      "epoch": 3.753346449043738,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.972633469553962e-05,
+      "loss": 0.9755754852294922,
       "step": 326000
     },
     {
-      "epoch": 7.5112830431979365,
-      "grad_norm": 2.8027350902557373,
-      "learning_rate": 1.9723426070010617e-05,
-      "loss": 0.5935,
+      "epoch": 3.7556491341184532,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9719063131717107e-05,
+      "loss": 1.0151551055908203,
       "step": 326200
     },
     {
-      "epoch": 7.515888366952197,
-      "grad_norm": 4.174149990081787,
-      "learning_rate": 1.9716154506188102e-05,
-      "loss": 0.5837,
+      "epoch": 3.757951819193168,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9711791567894593e-05,
+      "loss": 0.9837895202636718,
       "step": 326400
     },
     {
-      "epoch": 7.520493690706457,
-      "grad_norm": 4.200664043426514,
-      "learning_rate": 1.9708919300184698e-05,
-      "loss": 0.6054,
+      "epoch": 3.7602545042678828,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9704520004072076e-05,
+      "loss": 0.9870108795166016,
       "step": 326600
     },
     {
-      "epoch": 7.525099014460716,
-      "grad_norm": 2.414337158203125,
-      "learning_rate": 1.9701647736362184e-05,
-      "loss": 0.5915,
+      "epoch": 3.7625571893425978,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.969724844024956e-05,
+      "loss": 0.9984638977050782,
       "step": 326800
     },
     {
-      "epoch": 7.529704338214977,
-      "grad_norm": 2.9279587268829346,
-      "learning_rate": 1.9694376172539666e-05,
-      "loss": 0.5921,
+      "epoch": 3.7648598744173127,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9689976876427044e-05,
+      "loss": 0.9689566040039063,
       "step": 327000
     },
     {
-      "epoch": 7.534309661969236,
-      "grad_norm": 3.7588605880737305,
-      "learning_rate": 1.9687104608717152e-05,
-      "loss": 0.5839,
+      "epoch": 3.7671625594920277,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.968270531260453e-05,
+      "loss": 0.9929909515380859,
       "step": 327200
     },
     {
-      "epoch": 7.538914985723497,
-      "grad_norm": 3.5354812145233154,
-      "learning_rate": 1.9679833044894635e-05,
-      "loss": 0.5954,
+      "epoch": 3.7694652445667427,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9675433748782013e-05,
+      "loss": 1.0006656646728516,
       "step": 327400
     },
     {
-      "epoch": 7.543520309477756,
-      "grad_norm": 3.489088773727417,
-      "learning_rate": 1.967256148107212e-05,
-      "loss": 0.6005,
+      "epoch": 3.7717679296414577,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.96681621849595e-05,
+      "loss": 0.9882119750976562,
       "step": 327600
     },
     {
-      "epoch": 7.548125633232016,
-      "grad_norm": 3.0802366733551025,
-      "learning_rate": 1.9665289917249603e-05,
-      "loss": 0.5928,
+      "epoch": 3.7740706147161722,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9660890621136985e-05,
+      "loss": 0.9780780029296875,
       "step": 327800
     },
     {
-      "epoch": 7.552730956986276,
-      "grad_norm": 3.6488418579101562,
-      "learning_rate": 1.965801835342709e-05,
-      "loss": 0.5977,
+      "epoch": 3.7763732997908876,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.9653619057314467e-05,
+      "loss": 0.9838511657714843,
       "step": 328000
     },
     {
-      "epoch": 7.557336280740536,
-      "grad_norm": 3.7551910877227783,
-      "learning_rate": 1.9650746789604575e-05,
-      "loss": 0.5832,
+      "epoch": 3.778675984865602,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.964634749349195e-05,
+      "loss": 0.9679155731201172,
       "step": 328200
     },
     {
-      "epoch": 7.5619416044947965,
-      "grad_norm": 2.494746685028076,
-      "learning_rate": 1.9643475225782057e-05,
-      "loss": 0.5867,
+      "epoch": 3.780978669940317,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9639075929669436e-05,
+      "loss": 0.9874655151367188,
       "step": 328400
     },
     {
-      "epoch": 7.566546928249056,
-      "grad_norm": 3.4806318283081055,
-      "learning_rate": 1.9636240019778656e-05,
-      "loss": 0.5818,
+      "epoch": 3.783281355015032,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9631804365846918e-05,
+      "loss": 0.9853670501708984,
       "step": 328600
     },
     {
-      "epoch": 7.571152252003316,
-      "grad_norm": 3.278502941131592,
-      "learning_rate": 1.962896845595614e-05,
-      "loss": 0.5864,
+      "epoch": 3.785584040089747,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9624532802024404e-05,
+      "loss": 0.973040771484375,
       "step": 328800
     },
     {
-      "epoch": 7.575757575757576,
-      "grad_norm": 3.046190023422241,
-      "learning_rate": 1.962169689213362e-05,
-      "loss": 0.5904,
+      "epoch": 3.787886725164462,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.961726123820189e-05,
+      "loss": 1.0069470977783204,
       "step": 329000
     },
     {
-      "epoch": 7.580362899511836,
-      "grad_norm": 2.7647550106048584,
-      "learning_rate": 1.9614425328311107e-05,
-      "loss": 0.6045,
+      "epoch": 3.790189410239177,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9609989674379372e-05,
+      "loss": 1.0185445404052735,
       "step": 329200
     },
     {
-      "epoch": 7.584968223266095,
-      "grad_norm": 3.201040506362915,
-      "learning_rate": 1.9607153764488593e-05,
-      "loss": 0.5932,
+      "epoch": 3.792492095313892,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9602718110556855e-05,
+      "loss": 0.9744924926757812,
       "step": 329400
     },
     {
-      "epoch": 7.589573547020356,
-      "grad_norm": 4.180473327636719,
-      "learning_rate": 1.9599882200666075e-05,
-      "loss": 0.5883,
+      "epoch": 3.7947947803886066,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.959544654673434e-05,
+      "loss": 0.955069808959961,
       "step": 329600
     },
     {
-      "epoch": 7.594178870774615,
-      "grad_norm": 2.905902862548828,
-      "learning_rate": 1.959261063684356e-05,
-      "loss": 0.593,
+      "epoch": 3.797097465463322,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9588174982911827e-05,
+      "loss": 0.9998871612548829,
       "step": 329800
     },
     {
-      "epoch": 7.598784194528875,
-      "grad_norm": 3.984675884246826,
-      "learning_rate": 1.9585339073021047e-05,
-      "loss": 0.6001,
+      "epoch": 3.7994001505380366,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.958090341908931e-05,
+      "loss": 0.9712278747558594,
       "step": 330000
     },
     {
-      "epoch": 7.6033895182831355,
-      "grad_norm": 3.31915545463562,
-      "learning_rate": 1.9578067509198526e-05,
-      "loss": 0.5868,
+      "epoch": 3.8017028356127516,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9573631855266795e-05,
+      "loss": 0.9825238037109375,
       "step": 330200
     },
     {
-      "epoch": 7.607994842037395,
-      "grad_norm": 2.907419204711914,
-      "learning_rate": 1.9570795945376012e-05,
-      "loss": 0.5885,
+      "epoch": 3.8040055206874666,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.956636029144428e-05,
+      "loss": 0.9737995910644531,
       "step": 330400
     },
     {
-      "epoch": 7.612600165791655,
-      "grad_norm": 3.3257858753204346,
-      "learning_rate": 1.9563524381553498e-05,
-      "loss": 0.5821,
+      "epoch": 3.8063082057621815,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9559088727621764e-05,
+      "loss": 0.9713996124267578,
       "step": 330600
     },
     {
-      "epoch": 7.617205489545915,
-      "grad_norm": 3.4570491313934326,
-      "learning_rate": 1.955625281773098e-05,
-      "loss": 0.5999,
+      "epoch": 3.8086108908368965,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9551817163799246e-05,
+      "loss": 0.9742137145996094,
       "step": 330800
     },
     {
-      "epoch": 7.621810813300175,
-      "grad_norm": 3.399859666824341,
-      "learning_rate": 1.9548981253908467e-05,
-      "loss": 0.5977,
+      "epoch": 3.8109135759116115,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9544545599976732e-05,
+      "loss": 0.9718436431884766,
       "step": 331000
     },
     {
-      "epoch": 7.626416137054435,
-      "grad_norm": 3.104308605194092,
-      "learning_rate": 1.9541709690085952e-05,
-      "loss": 0.5842,
+      "epoch": 3.8132162609863265,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9537274036154215e-05,
+      "loss": 0.9649051666259766,
       "step": 331200
     },
     {
-      "epoch": 7.631021460808695,
-      "grad_norm": 3.126985788345337,
-      "learning_rate": 1.953443812626343e-05,
-      "loss": 0.6003,
+      "epoch": 3.815518946061041,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.95300024723317e-05,
+      "loss": 0.9786886596679687,
       "step": 331400
     },
     {
-      "epoch": 7.635626784562954,
-      "grad_norm": 2.7562856674194336,
-      "learning_rate": 1.9527166562440918e-05,
-      "loss": 0.5831,
+      "epoch": 3.8178216311357565,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9522730908509186e-05,
+      "loss": 0.9678800964355468,
       "step": 331600
     },
     {
-      "epoch": 7.640232108317215,
-      "grad_norm": 2.559391975402832,
-      "learning_rate": 1.9519894998618403e-05,
-      "loss": 0.5963,
+      "epoch": 3.820124316210471,
+      "grad_norm": 0.5,
+      "learning_rate": 1.951545934468667e-05,
+      "loss": 0.9907022094726563,
       "step": 331800
     },
     {
-      "epoch": 7.6448374320714745,
-      "grad_norm": 3.1173386573791504,
-      "learning_rate": 1.9512659792615e-05,
-      "loss": 0.6023,
+      "epoch": 3.822427001285186,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.950818778086415e-05,
+      "loss": 0.9843609619140625,
       "step": 332000
     },
     {
-      "epoch": 7.649442755825735,
-      "grad_norm": 2.7219719886779785,
-      "learning_rate": 1.9505388228792485e-05,
-      "loss": 0.5869,
+      "epoch": 3.824729686359901,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9500916217041637e-05,
+      "loss": 0.9689896392822266,
       "step": 332200
     },
     {
-      "epoch": 7.654048079579995,
-      "grad_norm": 2.787108898162842,
-      "learning_rate": 1.949811666496997e-05,
-      "loss": 0.6002,
+      "epoch": 3.827032371434616,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.949364465321912e-05,
+      "loss": 0.9918228149414062,
       "step": 332400
     },
     {
-      "epoch": 7.658653403334254,
-      "grad_norm": 2.7488784790039062,
-      "learning_rate": 1.9490845101147453e-05,
-      "loss": 0.5946,
+      "epoch": 3.829335056509331,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9486373089396606e-05,
+      "loss": 0.9822836303710938,
       "step": 332600
     },
     {
-      "epoch": 7.663258727088515,
-      "grad_norm": 3.056337833404541,
-      "learning_rate": 1.948357353732494e-05,
-      "loss": 0.5933,
+      "epoch": 3.831637741584046,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9479101525574092e-05,
+      "loss": 0.9748810577392578,
       "step": 332800
     },
     {
-      "epoch": 7.667864050842774,
-      "grad_norm": 3.6845805644989014,
-      "learning_rate": 1.947630197350242e-05,
-      "loss": 0.5947,
+      "epoch": 3.833940426658761,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9471829961751578e-05,
+      "loss": 0.9820301055908203,
       "step": 333000
     },
     {
-      "epoch": 7.672469374597034,
-      "grad_norm": 2.5756258964538574,
-      "learning_rate": 1.9469030409679904e-05,
-      "loss": 0.5984,
+      "epoch": 3.836243111733476,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9464558397929057e-05,
+      "loss": 0.9797927856445312,
       "step": 333200
     },
     {
-      "epoch": 7.677074698351294,
-      "grad_norm": 3.2758781909942627,
-      "learning_rate": 1.946175884585739e-05,
-      "loss": 0.5955,
+      "epoch": 3.838545796808191,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.9457286834106543e-05,
+      "loss": 0.9862161254882813,
       "step": 333400
     },
     {
-      "epoch": 7.681680022105554,
-      "grad_norm": 3.465355634689331,
-      "learning_rate": 1.9454487282034876e-05,
-      "loss": 0.5903,
+      "epoch": 3.8408484818829054,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.945001527028403e-05,
+      "loss": 0.9810741424560547,
       "step": 333600
     },
     {
-      "epoch": 7.6862853458598135,
-      "grad_norm": 3.2135512828826904,
-      "learning_rate": 1.944721571821236e-05,
-      "loss": 0.5864,
+      "epoch": 3.8431511669576204,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.944274370646151e-05,
+      "loss": 0.9708782196044922,
       "step": 333800
     },
     {
-      "epoch": 7.690890669614074,
-      "grad_norm": 2.3275580406188965,
-      "learning_rate": 1.9439944154389844e-05,
-      "loss": 0.5797,
+      "epoch": 3.8454538520323354,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9435472142638997e-05,
+      "loss": 0.9710533905029297,
       "step": 334000
     },
     {
-      "epoch": 7.695495993368334,
-      "grad_norm": 2.834540843963623,
-      "learning_rate": 1.9432672590567327e-05,
-      "loss": 0.5891,
+      "epoch": 3.8477565371070503,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9428200578816483e-05,
+      "loss": 0.9628467559814453,
       "step": 334200
     },
     {
-      "epoch": 7.700101317122594,
-      "grad_norm": 2.680785894393921,
-      "learning_rate": 1.9425401026744813e-05,
-      "loss": 0.5916,
+      "epoch": 3.8500592221817653,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9420929014993965e-05,
+      "loss": 1.0163386535644532,
       "step": 334400
     },
     {
-      "epoch": 7.704706640876854,
-      "grad_norm": 4.272080421447754,
-      "learning_rate": 1.9418129462922295e-05,
-      "loss": 0.5999,
+      "epoch": 3.8523619072564803,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9413657451171448e-05,
+      "loss": 0.9830091857910156,
       "step": 334600
     },
     {
-      "epoch": 7.709311964631113,
-      "grad_norm": 2.508699417114258,
-      "learning_rate": 1.941085789909978e-05,
-      "loss": 0.5903,
+      "epoch": 3.8546645923311953,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9406385887348934e-05,
+      "loss": 0.9770928192138671,
       "step": 334800
     },
     {
-      "epoch": 7.713917288385374,
-      "grad_norm": 4.05361270904541,
-      "learning_rate": 1.9403586335277267e-05,
-      "loss": 0.5947,
+      "epoch": 3.8569672774059103,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9399114323526416e-05,
+      "loss": 0.9688341522216797,
       "step": 335000
     },
     {
-      "epoch": 7.718522612139633,
-      "grad_norm": 2.704547882080078,
-      "learning_rate": 1.939631477145475e-05,
-      "loss": 0.5867,
+      "epoch": 3.8592699624806253,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9391842759703902e-05,
+      "loss": 0.977500991821289,
       "step": 335200
     },
     {
-      "epoch": 7.723127935893894,
-      "grad_norm": 2.987114667892456,
-      "learning_rate": 1.9389043207632235e-05,
-      "loss": 0.5883,
+      "epoch": 3.86157264755534,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9384571195881388e-05,
+      "loss": 0.9860353088378906,
       "step": 335400
     },
     {
-      "epoch": 7.727733259648153,
-      "grad_norm": 2.5852272510528564,
-      "learning_rate": 1.9381771643809718e-05,
-      "loss": 0.6022,
+      "epoch": 3.863875332630055,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.937729963205887e-05,
+      "loss": 0.9698558807373047,
       "step": 335600
     },
     {
-      "epoch": 7.732338583402413,
-      "grad_norm": 2.52404522895813,
-      "learning_rate": 1.93745000799872e-05,
-      "loss": 0.5791,
+      "epoch": 3.8661780177047698,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9370028068236353e-05,
+      "loss": 0.974271469116211,
       "step": 335800
     },
     {
-      "epoch": 7.7369439071566735,
-      "grad_norm": 2.9356820583343506,
-      "learning_rate": 1.9367228516164686e-05,
-      "loss": 0.5922,
+      "epoch": 3.8684807027794847,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.936275650441384e-05,
+      "loss": 0.9913460540771485,
       "step": 336000
     },
     {
-      "epoch": 7.741549230910933,
-      "grad_norm": 4.60886287689209,
-      "learning_rate": 1.9359956952342172e-05,
-      "loss": 0.583,
+      "epoch": 3.8707833878541997,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9355484940591325e-05,
+      "loss": 0.9814734649658203,
       "step": 336200
     },
     {
-      "epoch": 7.746154554665193,
-      "grad_norm": 4.0157856941223145,
-      "learning_rate": 1.9352685388519655e-05,
-      "loss": 0.5979,
+      "epoch": 3.8730860729289147,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9348213376768808e-05,
+      "loss": 0.9771896362304687,
       "step": 336400
     },
     {
-      "epoch": 7.750759878419453,
-      "grad_norm": 2.8278868198394775,
-      "learning_rate": 1.934541382469714e-05,
-      "loss": 0.603,
+      "epoch": 3.8753887580036297,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9340941812946294e-05,
+      "loss": 0.9758899688720704,
       "step": 336600
     },
     {
-      "epoch": 7.755365202173713,
-      "grad_norm": 2.852208375930786,
-      "learning_rate": 1.9338142260874623e-05,
-      "loss": 0.5775,
+      "epoch": 3.8776914430783447,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.933367024912378e-05,
+      "loss": 0.9744023895263672,
       "step": 336800
     },
     {
-      "epoch": 7.759970525927972,
-      "grad_norm": 3.623089075088501,
-      "learning_rate": 1.933087069705211e-05,
-      "loss": 0.589,
+      "epoch": 3.8799941281530597,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.932639868530126e-05,
+      "loss": 0.9712601470947265,
       "step": 337000
     },
     {
-      "epoch": 7.764575849682233,
-      "grad_norm": 2.7371058464050293,
-      "learning_rate": 1.9323635491048704e-05,
-      "loss": 0.5884,
+      "epoch": 3.882296813227774,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9319127121478745e-05,
+      "loss": 0.9955208587646485,
       "step": 337200
     },
     {
-      "epoch": 7.769181173436492,
-      "grad_norm": 4.192208290100098,
-      "learning_rate": 1.931636392722619e-05,
-      "loss": 0.5906,
+      "epoch": 3.884599498302489,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.931185555765623e-05,
+      "loss": 0.9782139587402344,
       "step": 337400
     },
     {
-      "epoch": 7.773786497190753,
-      "grad_norm": 4.169963836669922,
-      "learning_rate": 1.9309128721222786e-05,
-      "loss": 0.6003,
+      "epoch": 3.886902183377204,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9304583993833713e-05,
+      "loss": 0.9548270416259765,
       "step": 337600
     },
     {
-      "epoch": 7.7783918209450125,
-      "grad_norm": 3.134167432785034,
-      "learning_rate": 1.930185715740027e-05,
-      "loss": 0.6032,
+      "epoch": 3.889204868451919,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.92973124300112e-05,
+      "loss": 0.9846508026123046,
       "step": 337800
     },
     {
-      "epoch": 7.782997144699272,
-      "grad_norm": 3.1043667793273926,
-      "learning_rate": 1.9294585593577757e-05,
-      "loss": 0.5872,
+      "epoch": 3.891507553526634,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9290040866188685e-05,
+      "loss": 0.9758512878417969,
       "step": 338000
     },
     {
-      "epoch": 7.7876024684535325,
-      "grad_norm": 2.8810665607452393,
-      "learning_rate": 1.928731402975524e-05,
-      "loss": 0.5848,
+      "epoch": 3.893810238601349,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9282769302366167e-05,
+      "loss": 0.9999530029296875,
       "step": 338200
     },
     {
-      "epoch": 7.792207792207792,
-      "grad_norm": 2.3954150676727295,
-      "learning_rate": 1.9280042465932722e-05,
-      "loss": 0.587,
+      "epoch": 3.896112923676064,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.927549773854365e-05,
+      "loss": 0.9721403503417969,
       "step": 338400
     },
     {
-      "epoch": 7.796813115962053,
-      "grad_norm": 3.5883309841156006,
-      "learning_rate": 1.927277090211021e-05,
-      "loss": 0.5954,
+      "epoch": 3.898415608750779,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9268226174721136e-05,
+      "loss": 0.9728392791748047,
       "step": 338600
     },
     {
-      "epoch": 7.801418439716312,
-      "grad_norm": 2.6461853981018066,
-      "learning_rate": 1.926549933828769e-05,
-      "loss": 0.5901,
+      "epoch": 3.900718293825494,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9260954610898622e-05,
+      "loss": 0.9802152252197266,
       "step": 338800
     },
     {
-      "epoch": 7.806023763470572,
-      "grad_norm": 3.625852108001709,
-      "learning_rate": 1.9258227774465177e-05,
-      "loss": 0.5784,
+      "epoch": 3.9030209789002086,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9253683047076104e-05,
+      "loss": 0.9853790283203125,
       "step": 339000
     },
     {
-      "epoch": 7.810629087224832,
-      "grad_norm": 3.1065762042999268,
-      "learning_rate": 1.9250956210642663e-05,
-      "loss": 0.5917,
+      "epoch": 3.905323663974924,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.924641148325359e-05,
+      "loss": 0.9751505279541015,
       "step": 339200
     },
     {
-      "epoch": 7.815234410979092,
-      "grad_norm": 2.77862286567688,
-      "learning_rate": 1.9243684646820145e-05,
-      "loss": 0.5922,
+      "epoch": 3.9076263490496386,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9239139919431076e-05,
+      "loss": 0.9823059844970703,
       "step": 339400
     },
     {
-      "epoch": 7.8198397347333515,
-      "grad_norm": 4.303182601928711,
-      "learning_rate": 1.923641308299763e-05,
-      "loss": 0.5842,
+      "epoch": 3.9099290341243536,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9231868355608555e-05,
+      "loss": 0.9839841461181641,
       "step": 339600
     },
     {
-      "epoch": 7.824445058487612,
-      "grad_norm": 3.4562771320343018,
-      "learning_rate": 1.9229141519175114e-05,
-      "loss": 0.6,
+      "epoch": 3.9122317191990685,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.922459679178604e-05,
+      "loss": 0.9967166137695312,
       "step": 339800
     },
     {
-      "epoch": 7.8290503822418716,
-      "grad_norm": 2.983295440673828,
-      "learning_rate": 1.9221869955352596e-05,
-      "loss": 0.6038,
+      "epoch": 3.9145344042737835,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9217325227963527e-05,
+      "loss": 0.9998323822021484,
       "step": 340000
     },
     {
-      "epoch": 7.833655705996131,
-      "grad_norm": 3.2224607467651367,
-      "learning_rate": 1.9214598391530082e-05,
-      "loss": 0.5834,
+      "epoch": 3.9168370893484985,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.921005366414101e-05,
+      "loss": 0.992779541015625,
       "step": 340200
     },
     {
-      "epoch": 7.838261029750392,
-      "grad_norm": 3.168325185775757,
-      "learning_rate": 1.9207326827707568e-05,
-      "loss": 0.5772,
+      "epoch": 3.9191397744232135,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9202782100318495e-05,
+      "loss": 1.0167972564697265,
       "step": 340400
     },
     {
-      "epoch": 7.842866353504651,
-      "grad_norm": 2.240220546722412,
-      "learning_rate": 1.9200055263885054e-05,
-      "loss": 0.592,
+      "epoch": 3.9214424594979285,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.919551053649598e-05,
+      "loss": 0.9844973754882812,
       "step": 340600
     },
     {
-      "epoch": 7.847471677258911,
-      "grad_norm": 3.125988245010376,
-      "learning_rate": 1.9192783700062536e-05,
-      "loss": 0.579,
+      "epoch": 3.923745144572643,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9188238972673464e-05,
+      "loss": 0.9800909423828125,
       "step": 340800
     },
     {
-      "epoch": 7.852077001013171,
-      "grad_norm": 3.029654026031494,
-      "learning_rate": 1.918551213624002e-05,
-      "loss": 0.6008,
+      "epoch": 3.9260478296473584,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9180967408850946e-05,
+      "loss": 0.9858707427978516,
       "step": 341000
     },
     {
-      "epoch": 7.856682324767431,
-      "grad_norm": 3.496350049972534,
-      "learning_rate": 1.9178240572417505e-05,
-      "loss": 0.584,
+      "epoch": 3.928350514722073,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9173695845028432e-05,
+      "loss": 0.9715339660644531,
       "step": 341200
     },
     {
-      "epoch": 7.861287648521691,
-      "grad_norm": 3.7270963191986084,
-      "learning_rate": 1.9170969008594987e-05,
-      "loss": 0.5942,
+      "epoch": 3.930653199796788,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9166424281205915e-05,
+      "loss": 0.9722483825683593,
       "step": 341400
     },
     {
-      "epoch": 7.865892972275951,
-      "grad_norm": 3.109341621398926,
-      "learning_rate": 1.9163697444772473e-05,
-      "loss": 0.5729,
+      "epoch": 3.932955884871503,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.91591527173834e-05,
+      "loss": 0.9885195159912109,
       "step": 341600
     },
     {
-      "epoch": 7.870498296030211,
-      "grad_norm": 2.5632474422454834,
-      "learning_rate": 1.915642588094996e-05,
-      "loss": 0.5945,
+      "epoch": 3.935258569946218,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9151881153560887e-05,
+      "loss": 0.9816614532470703,
       "step": 341800
     },
     {
-      "epoch": 7.875103619784471,
-      "grad_norm": 2.6529979705810547,
-      "learning_rate": 1.9149154317127442e-05,
-      "loss": 0.5764,
+      "epoch": 3.937561255020933,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9144609589738373e-05,
+      "loss": 0.9931729125976563,
       "step": 342000
     },
     {
-      "epoch": 7.879708943538731,
-      "grad_norm": 2.466392993927002,
-      "learning_rate": 1.9141882753304928e-05,
-      "loss": 0.5857,
+      "epoch": 3.939863940095648,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9137338025915852e-05,
+      "loss": 0.9719671630859374,
       "step": 342200
     },
     {
-      "epoch": 7.884314267292991,
-      "grad_norm": 4.523742198944092,
-      "learning_rate": 1.913461118948241e-05,
-      "loss": 0.5771,
+      "epoch": 3.942166625170363,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9130066462093338e-05,
+      "loss": 0.9845899963378906,
       "step": 342400
     },
     {
-      "epoch": 7.888919591047251,
-      "grad_norm": 3.0389695167541504,
-      "learning_rate": 1.9127339625659893e-05,
-      "loss": 0.587,
+      "epoch": 3.9444693102450774,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9122794898270824e-05,
+      "loss": 1.0178986358642579,
       "step": 342600
     },
     {
-      "epoch": 7.89352491480151,
-      "grad_norm": 3.1406664848327637,
-      "learning_rate": 1.912006806183738e-05,
-      "loss": 0.5812,
+      "epoch": 3.946771995319793,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9115523334448306e-05,
+      "loss": 0.9837549591064453,
       "step": 342800
     },
     {
-      "epoch": 7.898130238555771,
-      "grad_norm": 2.3682878017425537,
-      "learning_rate": 1.9112796498014865e-05,
-      "loss": 0.5873,
+      "epoch": 3.9490746803945074,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9108251770625792e-05,
+      "loss": 0.9918838500976562,
       "step": 343000
     },
     {
-      "epoch": 7.90273556231003,
-      "grad_norm": 2.9673991203308105,
-      "learning_rate": 1.910552493419235e-05,
-      "loss": 0.5781,
+      "epoch": 3.9513773654692224,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9100980206803278e-05,
+      "loss": 0.9758218383789062,
       "step": 343200
     },
     {
-      "epoch": 7.90734088606429,
-      "grad_norm": 3.378129243850708,
-      "learning_rate": 1.9098253370369833e-05,
-      "loss": 0.5848,
+      "epoch": 3.9536800505439373,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.9093708642980757e-05,
+      "loss": 0.9743316650390625,
       "step": 343400
     },
     {
-      "epoch": 7.9119462098185505,
-      "grad_norm": 3.566913366317749,
-      "learning_rate": 1.9090981806547316e-05,
-      "loss": 0.5836,
+      "epoch": 3.9559827356186523,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9086437079158243e-05,
+      "loss": 0.9659014129638672,
       "step": 343600
     },
     {
-      "epoch": 7.91655153357281,
-      "grad_norm": 3.525033712387085,
-      "learning_rate": 1.90837102427248e-05,
-      "loss": 0.594,
+      "epoch": 3.9582854206933673,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.907916551533573e-05,
+      "loss": 0.9813739776611328,
       "step": 343800
     },
     {
-      "epoch": 7.92115685732707,
-      "grad_norm": 2.8252060413360596,
-      "learning_rate": 1.9076438678902284e-05,
-      "loss": 0.5913,
+      "epoch": 3.9605881057680823,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.907189395151321e-05,
+      "loss": 0.9784278106689454,
       "step": 344000
     },
     {
-      "epoch": 7.92576218108133,
-      "grad_norm": 4.159696102142334,
-      "learning_rate": 1.9069203472898883e-05,
-      "loss": 0.5845,
+      "epoch": 3.9628907908427973,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9064622387690697e-05,
+      "loss": 0.9834820556640625,
       "step": 344200
     },
     {
-      "epoch": 7.93036750483559,
-      "grad_norm": 3.0841197967529297,
-      "learning_rate": 1.9061968266895478e-05,
-      "loss": 0.5936,
+      "epoch": 3.965193475917512,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9057350823868183e-05,
+      "loss": 0.9798778533935547,
       "step": 344400
     },
     {
-      "epoch": 7.93497282858985,
-      "grad_norm": 2.3615498542785645,
-      "learning_rate": 1.9054696703072964e-05,
-      "loss": 0.5897,
+      "epoch": 3.9674961609922272,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.905007926004567e-05,
+      "loss": 0.9758098602294922,
       "step": 344600
     },
     {
-      "epoch": 7.93957815234411,
-      "grad_norm": 2.4957964420318604,
-      "learning_rate": 1.904742513925045e-05,
-      "loss": 0.5839,
+      "epoch": 3.9697988460669418,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9042807696223148e-05,
+      "loss": 0.9676461029052734,
       "step": 344800
     },
     {
-      "epoch": 7.944183476098369,
-      "grad_norm": 3.6795175075531006,
-      "learning_rate": 1.9040153575427932e-05,
-      "loss": 0.5876,
+      "epoch": 3.9721015311416568,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9035536132400634e-05,
+      "loss": 0.9825784301757813,
       "step": 345000
     },
     {
-      "epoch": 7.94878879985263,
-      "grad_norm": 2.6692540645599365,
-      "learning_rate": 1.9032882011605415e-05,
-      "loss": 0.5847,
+      "epoch": 3.9744042162163717,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.902826456857812e-05,
+      "loss": 0.9833172607421875,
       "step": 345200
     },
     {
-      "epoch": 7.9533941236068895,
-      "grad_norm": 3.4338459968566895,
-      "learning_rate": 1.90256104477829e-05,
-      "loss": 0.5909,
+      "epoch": 3.9767069012910867,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9020993004755603e-05,
+      "loss": 0.9863948059082032,
       "step": 345400
     },
     {
-      "epoch": 7.95799944736115,
-      "grad_norm": 2.6297929286956787,
-      "learning_rate": 1.9018375241779496e-05,
-      "loss": 0.5925,
+      "epoch": 3.9790095863658017,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.901372144093309e-05,
+      "loss": 1.005031509399414,
       "step": 345600
     },
     {
-      "epoch": 7.9626047711154095,
-      "grad_norm": 3.2749218940734863,
-      "learning_rate": 1.9011103677956982e-05,
-      "loss": 0.586,
+      "epoch": 3.9813122714405167,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9006449877110574e-05,
+      "loss": 0.9735250091552734,
       "step": 345800
     },
     {
-      "epoch": 7.967210094869669,
-      "grad_norm": 2.956953763961792,
-      "learning_rate": 1.9003832114134468e-05,
-      "loss": 0.5992,
+      "epoch": 3.9836149565152317,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.8999178313288054e-05,
+      "loss": 0.9725241851806641,
       "step": 346000
     },
     {
-      "epoch": 7.97181541862393,
-      "grad_norm": 3.2748615741729736,
-      "learning_rate": 1.899656055031195e-05,
-      "loss": 0.5948,
+      "epoch": 3.985917641589946,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.899190674946554e-05,
+      "loss": 0.9731990814208984,
       "step": 346200
     },
     {
-      "epoch": 7.976420742378189,
-      "grad_norm": 3.0381953716278076,
-      "learning_rate": 1.8989288986489436e-05,
-      "loss": 0.5946,
+      "epoch": 3.9882203266646616,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8984635185643025e-05,
+      "loss": 0.9702101135253907,
       "step": 346400
     },
     {
-      "epoch": 7.981026066132449,
-      "grad_norm": 2.761564254760742,
-      "learning_rate": 1.8982017422666922e-05,
-      "loss": 0.5925,
+      "epoch": 3.990523011739376,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8977363621820508e-05,
+      "loss": 0.9851368713378906,
       "step": 346600
     },
     {
-      "epoch": 7.985631389886709,
-      "grad_norm": 2.955965042114258,
-      "learning_rate": 1.89747458588444e-05,
-      "loss": 0.5809,
+      "epoch": 3.992825696814091,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.8970092057997994e-05,
+      "loss": 0.977271728515625,
       "step": 346800
     },
     {
-      "epoch": 7.990236713640969,
-      "grad_norm": 3.134218454360962,
-      "learning_rate": 1.8967474295021887e-05,
-      "loss": 0.5939,
+      "epoch": 3.995128381888806,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.896282049417548e-05,
+      "loss": 0.9866420745849609,
       "step": 347000
     },
     {
-      "epoch": 7.9948420373952285,
-      "grad_norm": 2.840864896774292,
-      "learning_rate": 1.8960202731199373e-05,
-      "loss": 0.5862,
+      "epoch": 3.997431066963521,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.895554893035296e-05,
+      "loss": 0.9789151763916015,
       "step": 347200
     },
     {
-      "epoch": 7.999447361149489,
-      "grad_norm": 2.9396276473999023,
-      "learning_rate": 1.8952931167376855e-05,
-      "loss": 0.588,
+      "epoch": 3.999733752038236,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8948277366530445e-05,
+      "loss": 0.9966899108886719,
       "step": 347400
     },
     {
-      "epoch": 8.0,
-      "eval_loss": 0.5577627420425415,
-      "eval_runtime": 145.2006,
-      "eval_samples_per_second": 195.323,
-      "eval_steps_per_second": 12.211,
+      "epoch": 4.0,
+      "eval_loss": NaN,
+      "eval_runtime": 431.4871,
+      "eval_samples_per_second": 65.728,
+      "eval_steps_per_second": 32.865,
       "step": 347424
     },
     {
-      "epoch": 8.00405268490375,
-      "grad_norm": 2.531428098678589,
-      "learning_rate": 1.894565960355434e-05,
-      "loss": 0.5814,
+      "epoch": 4.002026362865749,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.894100580270793e-05,
+      "loss": 0.9862317657470703,
       "step": 347600
     },
     {
-      "epoch": 8.008658008658008,
-      "grad_norm": 3.6581764221191406,
-      "learning_rate": 1.8938388039731827e-05,
-      "loss": 0.5853,
+      "epoch": 4.004329047940464,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8933734238885417e-05,
+      "loss": 0.9764509582519532,
       "step": 347800
     },
     {
-      "epoch": 8.013263332412269,
-      "grad_norm": 2.494723081588745,
-      "learning_rate": 1.8931116475909306e-05,
-      "loss": 0.5866,
+      "epoch": 4.006631733015179,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.89264626750629e-05,
+      "loss": 1.006572265625,
       "step": 348000
     },
     {
-      "epoch": 8.017868656166529,
-      "grad_norm": 2.6664531230926514,
-      "learning_rate": 1.8923844912086792e-05,
-      "loss": 0.5971,
+      "epoch": 4.008934418089894,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8919191111240385e-05,
+      "loss": 0.9794659423828125,
       "step": 348200
     },
     {
-      "epoch": 8.022473979920788,
-      "grad_norm": 2.837907552719116,
-      "learning_rate": 1.8916573348264278e-05,
-      "loss": 0.583,
+      "epoch": 4.011237103164609,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.891191954741787e-05,
+      "loss": 0.9678721618652344,
       "step": 348400
     },
     {
-      "epoch": 8.027079303675048,
-      "grad_norm": 3.521794557571411,
-      "learning_rate": 1.8909301784441764e-05,
-      "loss": 0.5916,
+      "epoch": 4.013539788239324,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.890464798359535e-05,
+      "loss": 0.9642930603027344,
       "step": 348600
     },
     {
-      "epoch": 8.031684627429309,
-      "grad_norm": 2.823016881942749,
-      "learning_rate": 1.8902030220619247e-05,
-      "loss": 0.5879,
+      "epoch": 4.015842473314039,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8897376419772836e-05,
+      "loss": 0.9878228759765625,
       "step": 348800
     },
     {
-      "epoch": 8.036289951183567,
-      "grad_norm": 2.7679924964904785,
-      "learning_rate": 1.8894758656796733e-05,
-      "loss": 0.5776,
+      "epoch": 4.018145158388753,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8890104855950322e-05,
+      "loss": 0.9811653900146484,
       "step": 349000
     },
     {
-      "epoch": 8.040895274937828,
-      "grad_norm": 4.573503017425537,
-      "learning_rate": 1.888748709297422e-05,
-      "loss": 0.5819,
+      "epoch": 4.020447843463469,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8882833292127804e-05,
+      "loss": 0.9771403503417969,
       "step": 349200
     },
     {
-      "epoch": 8.045500598692088,
-      "grad_norm": 2.8802812099456787,
-      "learning_rate": 1.8880215529151698e-05,
-      "loss": 0.5816,
+      "epoch": 4.022750528538183,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.887556172830529e-05,
+      "loss": 1.0004103088378906,
       "step": 349400
     },
     {
-      "epoch": 8.050105922446347,
-      "grad_norm": 2.96124005317688,
-      "learning_rate": 1.8872943965329184e-05,
-      "loss": 0.6029,
+      "epoch": 4.025053213612899,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8868290164482776e-05,
+      "loss": 1.0042034912109374,
       "step": 349600
     },
     {
-      "epoch": 8.054711246200608,
-      "grad_norm": 2.664660930633545,
-      "learning_rate": 1.886567240150667e-05,
-      "loss": 0.5707,
+      "epoch": 4.027355898687613,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8861018600660255e-05,
+      "loss": 0.9859686279296875,
       "step": 349800
     },
     {
-      "epoch": 8.059316569954868,
-      "grad_norm": 3.073604106903076,
-      "learning_rate": 1.8858400837684152e-05,
-      "loss": 0.5928,
+      "epoch": 4.029658583762329,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.885374703683774e-05,
+      "loss": 0.9745539855957032,
       "step": 350000
     },
     {
-      "epoch": 8.063921893709129,
-      "grad_norm": 3.5625150203704834,
-      "learning_rate": 1.8851129273861638e-05,
-      "loss": 0.5908,
+      "epoch": 4.031961268837043,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8846475473015227e-05,
+      "loss": 0.9789989471435547,
       "step": 350200
     },
     {
-      "epoch": 8.068527217463387,
-      "grad_norm": 2.9049417972564697,
-      "learning_rate": 1.8843857710039124e-05,
-      "loss": 0.5835,
+      "epoch": 4.034263953911759,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8839203909192713e-05,
+      "loss": 0.9792119598388672,
       "step": 350400
     },
     {
-      "epoch": 8.073132541217648,
-      "grad_norm": 3.092867851257324,
-      "learning_rate": 1.8836586146216603e-05,
-      "loss": 0.5788,
+      "epoch": 4.036566638986473,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8831932345370196e-05,
+      "loss": 1.007418441772461,
       "step": 350600
     },
     {
-      "epoch": 8.077737864971908,
-      "grad_norm": 2.719918727874756,
-      "learning_rate": 1.88293509402132e-05,
-      "loss": 0.5823,
+      "epoch": 4.038869324061188,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.882466078154768e-05,
+      "loss": 0.988331298828125,
       "step": 350800
     },
     {
-      "epoch": 8.082343188726167,
-      "grad_norm": 4.278101444244385,
-      "learning_rate": 1.8822115734209797e-05,
-      "loss": 0.5818,
+      "epoch": 4.041172009135903,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8817389217725168e-05,
+      "loss": 0.9866494750976562,
       "step": 351000
     },
     {
-      "epoch": 8.086948512480427,
-      "grad_norm": 2.67177677154541,
-      "learning_rate": 1.8814844170387283e-05,
-      "loss": 0.5893,
+      "epoch": 4.043474694210618,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8810117653902647e-05,
+      "loss": 0.9749382781982422,
       "step": 351200
     },
     {
-      "epoch": 8.091553836234688,
-      "grad_norm": 3.964667558670044,
-      "learning_rate": 1.880757260656477e-05,
-      "loss": 0.5855,
+      "epoch": 4.045777379285333,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8802846090080133e-05,
+      "loss": 0.969979019165039,
       "step": 351400
     },
     {
-      "epoch": 8.096159159988947,
-      "grad_norm": 3.183964252471924,
-      "learning_rate": 1.880030104274225e-05,
-      "loss": 0.5794,
+      "epoch": 4.048080064360048,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.879557452625762e-05,
+      "loss": 0.9858515167236328,
       "step": 351600
     },
     {
-      "epoch": 8.100764483743207,
-      "grad_norm": 2.9570157527923584,
-      "learning_rate": 1.8793029478919737e-05,
-      "loss": 0.5979,
+      "epoch": 4.050382749434763,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.87883029624351e-05,
+      "loss": 0.9740243530273438,
       "step": 351800
     },
     {
-      "epoch": 8.105369807497468,
-      "grad_norm": 3.8420159816741943,
-      "learning_rate": 1.8785757915097223e-05,
-      "loss": 0.5792,
+      "epoch": 4.052685434509478,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8781031398612587e-05,
+      "loss": 0.9444357299804688,
       "step": 352000
     },
     {
-      "epoch": 8.109975131251726,
-      "grad_norm": 4.49518346786499,
-      "learning_rate": 1.8778486351274705e-05,
-      "loss": 0.5845,
+      "epoch": 4.054988119584193,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8773759834790073e-05,
+      "loss": 0.9807145690917969,
       "step": 352200
     },
     {
-      "epoch": 8.114580455005987,
-      "grad_norm": 2.9264657497406006,
-      "learning_rate": 1.8771214787452188e-05,
-      "loss": 0.5866,
+      "epoch": 4.057290804658908,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8766488270967552e-05,
+      "loss": 0.9957177734375,
       "step": 352400
     },
     {
-      "epoch": 8.119185778760247,
-      "grad_norm": 2.69858455657959,
-      "learning_rate": 1.8763943223629674e-05,
-      "loss": 0.5859,
+      "epoch": 4.059593489733622,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8759216707145038e-05,
+      "loss": 0.9883721160888672,
       "step": 352600
     },
     {
-      "epoch": 8.123791102514506,
-      "grad_norm": 3.450650453567505,
-      "learning_rate": 1.875667165980716e-05,
-      "loss": 0.5838,
+      "epoch": 4.0618961748083375,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.8751945143322524e-05,
+      "loss": 0.9880260467529297,
       "step": 352800
     },
     {
-      "epoch": 8.128396426268766,
-      "grad_norm": 2.7137413024902344,
-      "learning_rate": 1.8749400095984642e-05,
-      "loss": 0.5824,
+      "epoch": 4.064198859883052,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8744673579500006e-05,
+      "loss": 0.9933139038085937,
       "step": 353000
     },
     {
-      "epoch": 8.133001750023027,
-      "grad_norm": 3.0916788578033447,
-      "learning_rate": 1.8742128532162128e-05,
-      "loss": 0.5959,
+      "epoch": 4.0665015449577675,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8737402015677492e-05,
+      "loss": 0.9660495758056641,
       "step": 353200
     },
     {
-      "epoch": 8.137607073777286,
-      "grad_norm": 3.159905433654785,
-      "learning_rate": 1.8734856968339614e-05,
-      "loss": 0.5792,
+      "epoch": 4.068804230032482,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8730130451854978e-05,
+      "loss": 0.991396713256836,
       "step": 353400
     },
     {
-      "epoch": 8.142212397531546,
-      "grad_norm": 3.4873342514038086,
-      "learning_rate": 1.8727585404517093e-05,
-      "loss": 0.5834,
+      "epoch": 4.0711069151071975,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.872285888803246e-05,
+      "loss": 0.9717893981933594,
       "step": 353600
     },
     {
-      "epoch": 8.146817721285807,
-      "grad_norm": 3.2853617668151855,
-      "learning_rate": 1.872031384069458e-05,
-      "loss": 0.5818,
+      "epoch": 4.073409600181912,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8715587324209943e-05,
+      "loss": 0.9627035522460937,
       "step": 353800
     },
     {
-      "epoch": 8.151423045040067,
-      "grad_norm": 4.945338726043701,
-      "learning_rate": 1.8713042276872065e-05,
-      "loss": 0.579,
+      "epoch": 4.075712285256627,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.870831576038743e-05,
+      "loss": 1.0016319274902343,
       "step": 354000
     },
     {
-      "epoch": 8.156028368794326,
-      "grad_norm": 2.7136645317077637,
-      "learning_rate": 1.8705770713049548e-05,
-      "loss": 0.5743,
+      "epoch": 4.078014970331342,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8701044196564915e-05,
+      "loss": 0.9791947937011719,
       "step": 354200
     },
     {
-      "epoch": 8.160633692548586,
-      "grad_norm": 3.545706272125244,
-      "learning_rate": 1.8698535507046146e-05,
-      "loss": 0.5861,
+      "epoch": 4.0803176554060565,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8693772632742398e-05,
+      "loss": 0.9732559204101563,
       "step": 354400
     },
     {
-      "epoch": 8.165239016302847,
-      "grad_norm": 2.9433093070983887,
-      "learning_rate": 1.8691263943223632e-05,
-      "loss": 0.5775,
+      "epoch": 4.082620340480772,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8686501068919884e-05,
+      "loss": 0.9835048675537109,
       "step": 354600
     },
     {
-      "epoch": 8.169844340057105,
-      "grad_norm": 3.4373202323913574,
-      "learning_rate": 1.8683992379401115e-05,
-      "loss": 0.589,
+      "epoch": 4.0849230255554865,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.867922950509737e-05,
+      "loss": 0.9886447143554687,
       "step": 354800
     },
     {
-      "epoch": 8.174449663811366,
-      "grad_norm": 3.6623363494873047,
-      "learning_rate": 1.8676720815578597e-05,
-      "loss": 0.5737,
+      "epoch": 4.087225710630202,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.867195794127485e-05,
+      "loss": 0.9782583618164062,
       "step": 355000
     },
     {
-      "epoch": 8.179054987565626,
-      "grad_norm": 3.4433813095092773,
-      "learning_rate": 1.8669449251756083e-05,
-      "loss": 0.5852,
+      "epoch": 4.089528395704916,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8664686377452334e-05,
+      "loss": 0.9842738342285157,
       "step": 355200
     },
     {
-      "epoch": 8.183660311319885,
-      "grad_norm": 3.2509543895721436,
-      "learning_rate": 1.8662177687933566e-05,
-      "loss": 0.5787,
+      "epoch": 4.091831080779632,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.865741481362982e-05,
+      "loss": 0.980291976928711,
       "step": 355400
     },
     {
-      "epoch": 8.188265635074146,
-      "grad_norm": 2.7837679386138916,
-      "learning_rate": 1.865490612411105e-05,
-      "loss": 0.5808,
+      "epoch": 4.094133765854346,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8650143249807303e-05,
+      "loss": 0.978897705078125,
       "step": 355600
     },
     {
-      "epoch": 8.192870958828406,
-      "grad_norm": 3.2710013389587402,
-      "learning_rate": 1.8647634560288538e-05,
-      "loss": 0.5764,
+      "epoch": 4.096436450929062,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.864287168598479e-05,
+      "loss": 0.9819556427001953,
       "step": 355800
     },
     {
-      "epoch": 8.197476282582665,
-      "grad_norm": 3.4434151649475098,
-      "learning_rate": 1.864036299646602e-05,
-      "loss": 0.5755,
+      "epoch": 4.098739136003776,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.8635600122162275e-05,
+      "loss": 0.9883599853515626,
       "step": 356000
     },
     {
-      "epoch": 8.202081606336925,
-      "grad_norm": 3.4400975704193115,
-      "learning_rate": 1.8633091432643503e-05,
-      "loss": 0.5874,
+      "epoch": 4.101041821078491,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8628328558339757e-05,
+      "loss": 0.9936543273925781,
       "step": 356200
     },
     {
-      "epoch": 8.206686930091186,
-      "grad_norm": 3.25714111328125,
-      "learning_rate": 1.862581986882099e-05,
-      "loss": 0.5868,
+      "epoch": 4.103344506153206,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.862105699451724e-05,
+      "loss": 0.9793272399902344,
       "step": 356400
     },
     {
-      "epoch": 8.211292253845444,
-      "grad_norm": 5.997415542602539,
-      "learning_rate": 1.8618548304998474e-05,
-      "loss": 0.5689,
+      "epoch": 4.105647191227921,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8613785430694726e-05,
+      "loss": 0.9574286651611328,
       "step": 356600
     },
     {
-      "epoch": 8.215897577599705,
-      "grad_norm": 3.3374671936035156,
-      "learning_rate": 1.861131309899507e-05,
-      "loss": 0.5931,
+      "epoch": 4.107949876302636,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.860651386687221e-05,
+      "loss": 0.9973934173583985,
       "step": 356800
     },
     {
-      "epoch": 8.220502901353965,
-      "grad_norm": 2.680866241455078,
-      "learning_rate": 1.8604041535172556e-05,
-      "loss": 0.5855,
+      "epoch": 4.110252561377351,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8599242303049694e-05,
+      "loss": 0.9751262664794922,
       "step": 357000
     },
     {
-      "epoch": 8.225108225108226,
-      "grad_norm": 3.1401114463806152,
-      "learning_rate": 1.8596769971350038e-05,
-      "loss": 0.5864,
+      "epoch": 4.112555246452066,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.859197073922718e-05,
+      "loss": 0.9817977142333985,
       "step": 357200
     },
     {
-      "epoch": 8.229713548862485,
-      "grad_norm": 2.7683680057525635,
-      "learning_rate": 1.8589498407527524e-05,
-      "loss": 0.578,
+      "epoch": 4.114857931526781,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8584699175404663e-05,
+      "loss": 0.9779124450683594,
       "step": 357400
     },
     {
-      "epoch": 8.234318872616745,
-      "grad_norm": 3.4666707515716553,
-      "learning_rate": 1.858222684370501e-05,
-      "loss": 0.5788,
+      "epoch": 4.117160616601496,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8577427611582145e-05,
+      "loss": 0.9914723205566406,
       "step": 357600
     },
     {
-      "epoch": 8.238924196371006,
-      "grad_norm": 2.9666600227355957,
-      "learning_rate": 1.857495527988249e-05,
-      "loss": 0.5948,
+      "epoch": 4.119463301676211,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.857015604775963e-05,
+      "loss": 0.9791415405273437,
       "step": 357800
     },
     {
-      "epoch": 8.243529520125264,
-      "grad_norm": 3.1708505153656006,
-      "learning_rate": 1.8567683716059975e-05,
-      "loss": 0.5724,
+      "epoch": 4.121765986750925,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8562884483937117e-05,
+      "loss": 0.9917333984375,
       "step": 358000
     },
     {
-      "epoch": 8.248134843879525,
-      "grad_norm": 3.3768389225006104,
-      "learning_rate": 1.856041215223746e-05,
-      "loss": 0.5721,
+      "epoch": 4.124068671825641,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.85556129201146e-05,
+      "loss": 0.978380126953125,
       "step": 358200
     },
     {
-      "epoch": 8.252740167633785,
-      "grad_norm": 2.644984722137451,
-      "learning_rate": 1.8553140588414947e-05,
-      "loss": 0.5743,
+      "epoch": 4.126371356900355,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8548341356292085e-05,
+      "loss": 0.9702176666259765,
       "step": 358400
     },
     {
-      "epoch": 8.257345491388044,
-      "grad_norm": 3.155850648880005,
-      "learning_rate": 1.854586902459243e-05,
-      "loss": 0.5792,
+      "epoch": 4.128674041975071,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.854106979246957e-05,
+      "loss": 1.0004293060302734,
       "step": 358600
     },
     {
-      "epoch": 8.261950815142304,
-      "grad_norm": 2.7669639587402344,
-      "learning_rate": 1.8538597460769915e-05,
-      "loss": 0.5915,
+      "epoch": 4.130976727049785,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.853379822864705e-05,
+      "loss": 0.9715544128417969,
       "step": 358800
     },
     {
-      "epoch": 8.266556138896565,
-      "grad_norm": 3.183368682861328,
-      "learning_rate": 1.8531325896947398e-05,
-      "loss": 0.5899,
+      "epoch": 4.133279412124501,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8526526664824536e-05,
+      "loss": 0.9909072875976562,
       "step": 359000
     },
     {
-      "epoch": 8.271161462650824,
-      "grad_norm": 2.635927677154541,
-      "learning_rate": 1.852405433312488e-05,
-      "loss": 0.588,
+      "epoch": 4.135582097199215,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8519255101002022e-05,
+      "loss": 0.9791348266601563,
       "step": 359200
     },
     {
-      "epoch": 8.275766786405084,
-      "grad_norm": 3.761402130126953,
-      "learning_rate": 1.8516782769302366e-05,
-      "loss": 0.5888,
+      "epoch": 4.137884782273931,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.8511983537179508e-05,
+      "loss": 1.0011709594726563,
       "step": 359400
     },
     {
-      "epoch": 8.280372110159345,
-      "grad_norm": 2.46097469329834,
-      "learning_rate": 1.850954756329896e-05,
-      "loss": 0.5802,
+      "epoch": 4.140187467348645,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.850471197335699e-05,
+      "loss": 0.994935302734375,
       "step": 359600
     },
     {
-      "epoch": 8.284977433913603,
-      "grad_norm": 2.9108171463012695,
-      "learning_rate": 1.8502275999476447e-05,
-      "loss": 0.5836,
+      "epoch": 4.14249015242336,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8497440409534477e-05,
+      "loss": 0.9711431121826172,
       "step": 359800
     },
     {
-      "epoch": 8.289582757667864,
-      "grad_norm": 3.4605486392974854,
-      "learning_rate": 1.8495004435653933e-05,
-      "loss": 0.5948,
+      "epoch": 4.144792837498075,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.849016884571196e-05,
+      "loss": 0.9803479766845703,
       "step": 360000
     },
     {
-      "epoch": 8.294188081422124,
-      "grad_norm": 3.1106951236724854,
-      "learning_rate": 1.848773287183142e-05,
-      "loss": 0.5734,
+      "epoch": 4.14709552257279,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.848289728188944e-05,
+      "loss": 0.9712246704101563,
       "step": 360200
     },
     {
-      "epoch": 8.298793405176385,
-      "grad_norm": 2.4587388038635254,
-      "learning_rate": 1.84804613080089e-05,
-      "loss": 0.5776,
+      "epoch": 4.149398207647505,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8475625718066928e-05,
+      "loss": 0.9849492645263672,
       "step": 360400
     },
     {
-      "epoch": 8.303398728930643,
-      "grad_norm": 2.9624781608581543,
-      "learning_rate": 1.8473189744186384e-05,
-      "loss": 0.5843,
+      "epoch": 4.15170089272222,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8468354154244413e-05,
+      "loss": 1.0022724151611329,
       "step": 360600
     },
     {
-      "epoch": 8.308004052684904,
-      "grad_norm": 2.9920835494995117,
-      "learning_rate": 1.846591818036387e-05,
-      "loss": 0.5915,
+      "epoch": 4.154003577796935,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8461082590421896e-05,
+      "loss": 0.9713355255126953,
       "step": 360800
     },
     {
-      "epoch": 8.312609376439164,
-      "grad_norm": 2.7066526412963867,
-      "learning_rate": 1.8458646616541353e-05,
-      "loss": 0.5826,
+      "epoch": 4.15630626287165,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8453811026599382e-05,
+      "loss": 1.0081776428222655,
       "step": 361000
     },
     {
-      "epoch": 8.317214700193423,
-      "grad_norm": 3.3276214599609375,
-      "learning_rate": 1.845141141053795e-05,
-      "loss": 0.5697,
+      "epoch": 4.158608947946365,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8446539462776868e-05,
+      "loss": 0.9850017547607421,
       "step": 361200
     },
     {
-      "epoch": 8.321820023947684,
-      "grad_norm": 3.875354290008545,
-      "learning_rate": 1.8444139846715434e-05,
-      "loss": 0.5921,
+      "epoch": 4.16091163302108,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8439267898954347e-05,
+      "loss": 0.9687409210205078,
       "step": 361400
     },
     {
-      "epoch": 8.326425347701944,
-      "grad_norm": 2.708665132522583,
-      "learning_rate": 1.8436904640712032e-05,
-      "loss": 0.5803,
+      "epoch": 4.163214318095795,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8431996335131833e-05,
+      "loss": 0.9744286346435547,
       "step": 361600
     },
     {
-      "epoch": 8.331030671456203,
-      "grad_norm": 3.221430540084839,
-      "learning_rate": 1.8429633076889518e-05,
-      "loss": 0.5827,
+      "epoch": 4.1655170031705095,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.842472477130932e-05,
+      "loss": 0.9651099395751953,
       "step": 361800
     },
     {
-      "epoch": 8.335635995210463,
-      "grad_norm": 2.921086549758911,
-      "learning_rate": 1.8422361513067e-05,
-      "loss": 0.5774,
+      "epoch": 4.167819688245224,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.84174532074868e-05,
+      "loss": 0.9873933410644531,
       "step": 362000
     },
     {
-      "epoch": 8.340241318964724,
-      "grad_norm": 3.2136096954345703,
-      "learning_rate": 1.8415089949244483e-05,
-      "loss": 0.5872,
+      "epoch": 4.1701223733199395,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8410181643664287e-05,
+      "loss": 0.9850807952880859,
       "step": 362200
     },
     {
-      "epoch": 8.344846642718982,
-      "grad_norm": 2.5168769359588623,
-      "learning_rate": 1.840781838542197e-05,
-      "loss": 0.5836,
+      "epoch": 4.172425058394654,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.8402910079841773e-05,
+      "loss": 0.9752694702148438,
       "step": 362400
     },
     {
-      "epoch": 8.349451966473243,
-      "grad_norm": 2.8332345485687256,
-      "learning_rate": 1.8400546821599452e-05,
-      "loss": 0.5748,
+      "epoch": 4.1747277434693695,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.8395638516019256e-05,
+      "loss": 0.9639392852783203,
       "step": 362600
     },
     {
-      "epoch": 8.354057290227503,
-      "grad_norm": 2.76617431640625,
-      "learning_rate": 1.8393275257776938e-05,
-      "loss": 0.5802,
+      "epoch": 4.177030428544084,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8388366952196738e-05,
+      "loss": 0.9980973052978516,
       "step": 362800
     },
     {
-      "epoch": 8.358662613981762,
-      "grad_norm": 3.170368194580078,
-      "learning_rate": 1.8386003693954424e-05,
-      "loss": 0.5819,
+      "epoch": 4.1793331136187994,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8381095388374224e-05,
+      "loss": 0.9663079833984375,
       "step": 363000
     },
     {
-      "epoch": 8.363267937736023,
-      "grad_norm": 3.0185418128967285,
-      "learning_rate": 1.8378732130131906e-05,
-      "loss": 0.5835,
+      "epoch": 4.181635798693514,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.837382382455171e-05,
+      "loss": 0.9786151123046875,
       "step": 363200
     },
     {
-      "epoch": 8.367873261490283,
-      "grad_norm": 3.179088830947876,
-      "learning_rate": 1.837146056630939e-05,
-      "loss": 0.5827,
+      "epoch": 4.183938483768229,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8366552260729193e-05,
+      "loss": 0.9864177703857422,
       "step": 363400
     },
     {
-      "epoch": 8.372478585244544,
-      "grad_norm": 2.731213331222534,
-      "learning_rate": 1.8364189002486875e-05,
-      "loss": 0.5866,
+      "epoch": 4.186241168842944,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.835928069690668e-05,
+      "loss": 0.9760264587402344,
       "step": 363600
     },
     {
-      "epoch": 8.377083908998802,
-      "grad_norm": 2.9844472408294678,
-      "learning_rate": 1.835691743866436e-05,
-      "loss": 0.5795,
+      "epoch": 4.1885438539176585,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.835200913308416e-05,
+      "loss": 0.9764144134521484,
       "step": 363800
     },
     {
-      "epoch": 8.381689232753063,
-      "grad_norm": 3.1267082691192627,
-      "learning_rate": 1.8349645874841843e-05,
-      "loss": 0.5794,
+      "epoch": 4.190846538992374,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.8344737569261644e-05,
+      "loss": 0.9865864562988281,
       "step": 364000
     },
     {
-      "epoch": 8.386294556507323,
-      "grad_norm": 2.6979808807373047,
-      "learning_rate": 1.834237431101933e-05,
-      "loss": 0.5956,
+      "epoch": 4.1931492240670885,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.833746600543913e-05,
+      "loss": 0.962984848022461,
       "step": 364200
     },
     {
-      "epoch": 8.390899880261582,
-      "grad_norm": 2.9630825519561768,
-      "learning_rate": 1.8335102747196815e-05,
-      "loss": 0.5796,
+      "epoch": 4.195451909141804,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8330194441616615e-05,
+      "loss": 0.9840062713623047,
       "step": 364400
     },
     {
-      "epoch": 8.395505204015842,
-      "grad_norm": 3.268383741378784,
-      "learning_rate": 1.8327831183374297e-05,
-      "loss": 0.5943,
+      "epoch": 4.197754594216518,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8322922877794098e-05,
+      "loss": 0.9727967834472656,
       "step": 364600
     },
     {
-      "epoch": 8.400110527770103,
-      "grad_norm": 2.594475507736206,
-      "learning_rate": 1.832055961955178e-05,
-      "loss": 0.5741,
+      "epoch": 4.200057279291234,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8315651313971584e-05,
+      "loss": 0.9823301696777343,
       "step": 364800
     },
     {
-      "epoch": 8.404715851524362,
-      "grad_norm": 2.8668203353881836,
-      "learning_rate": 1.8313288055729266e-05,
-      "loss": 0.5843,
+      "epoch": 4.202359964365948,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.830837975014907e-05,
+      "loss": 0.9683138275146485,
       "step": 365000
     },
     {
-      "epoch": 8.409321175278622,
-      "grad_norm": 2.9619016647338867,
-      "learning_rate": 1.8306016491906748e-05,
-      "loss": 0.5845,
+      "epoch": 4.204662649440664,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.8301108186326552e-05,
+      "loss": 0.9668745422363281,
       "step": 365200
     },
     {
-      "epoch": 8.413926499032883,
-      "grad_norm": 3.0649566650390625,
-      "learning_rate": 1.8298744928084234e-05,
-      "loss": 0.6053,
+      "epoch": 4.206965334515378,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8293836622504035e-05,
+      "loss": 0.978956069946289,
       "step": 365400
     },
     {
-      "epoch": 8.418531822787141,
-      "grad_norm": 3.129210948944092,
-      "learning_rate": 1.829147336426172e-05,
-      "loss": 0.5795,
+      "epoch": 4.209268019590093,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.828656505868152e-05,
+      "loss": 1.00272216796875,
       "step": 365600
     },
     {
-      "epoch": 8.423137146541402,
-      "grad_norm": 3.2703466415405273,
-      "learning_rate": 1.8284201800439203e-05,
-      "loss": 0.5817,
+      "epoch": 4.211570704664808,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8279293494859007e-05,
+      "loss": 0.9836425018310547,
       "step": 365800
     },
     {
-      "epoch": 8.427742470295662,
-      "grad_norm": 3.3453192710876465,
-      "learning_rate": 1.8276930236616685e-05,
-      "loss": 0.5837,
+      "epoch": 4.213873389739523,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.827202193103649e-05,
+      "loss": 0.9879397583007813,
       "step": 366000
     },
     {
-      "epoch": 8.432347794049921,
-      "grad_norm": 3.5036461353302,
-      "learning_rate": 1.826965867279417e-05,
-      "loss": 0.588,
+      "epoch": 4.216176074814238,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8264750367213975e-05,
+      "loss": 0.9587268829345703,
       "step": 366200
     },
     {
-      "epoch": 8.436953117804181,
-      "grad_norm": 3.5221807956695557,
-      "learning_rate": 1.8262387108971657e-05,
-      "loss": 0.5866,
+      "epoch": 4.218478759888953,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8257478803391458e-05,
+      "loss": 0.9883473205566407,
       "step": 366400
     },
     {
-      "epoch": 8.441558441558442,
-      "grad_norm": 3.2333667278289795,
-      "learning_rate": 1.825511554514914e-05,
-      "loss": 0.5848,
+      "epoch": 4.220781444963668,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.825020723956894e-05,
+      "loss": 0.9879270172119141,
       "step": 366600
     },
     {
-      "epoch": 8.4461637653127,
-      "grad_norm": 3.349844455718994,
-      "learning_rate": 1.8247843981326625e-05,
-      "loss": 0.5864,
+      "epoch": 4.223084130038383,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8242935675746426e-05,
+      "loss": 0.9905947875976563,
       "step": 366800
     },
     {
-      "epoch": 8.450769089066961,
-      "grad_norm": 3.0328314304351807,
-      "learning_rate": 1.824057241750411e-05,
-      "loss": 0.5713,
+      "epoch": 4.225386815113098,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8235664111923912e-05,
+      "loss": 0.9815054321289063,
       "step": 367000
     },
     {
-      "epoch": 8.455374412821222,
-      "grad_norm": 3.0258967876434326,
-      "learning_rate": 1.8233300853681594e-05,
-      "loss": 0.5747,
+      "epoch": 4.227689500187813,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8228392548101394e-05,
+      "loss": 0.9851663208007813,
       "step": 367200
     },
     {
-      "epoch": 8.459979736575482,
-      "grad_norm": 2.960977792739868,
-      "learning_rate": 1.8226029289859076e-05,
-      "loss": 0.5685,
+      "epoch": 4.229992185262527,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.822112098427888e-05,
+      "loss": 0.9802882385253906,
       "step": 367400
     },
     {
-      "epoch": 8.46458506032974,
-      "grad_norm": 3.144864797592163,
-      "learning_rate": 1.8218757726036562e-05,
-      "loss": 0.5863,
+      "epoch": 4.232294870337243,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.8213849420456363e-05,
+      "loss": 0.9949236297607422,
       "step": 367600
     },
     {
-      "epoch": 8.469190384084001,
-      "grad_norm": 2.93837833404541,
-      "learning_rate": 1.8211486162214045e-05,
-      "loss": 0.5778,
+      "epoch": 4.234597555411957,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8206577856633845e-05,
+      "loss": 0.9930872344970703,
       "step": 367800
     },
     {
-      "epoch": 8.473795707838262,
-      "grad_norm": 3.0120725631713867,
-      "learning_rate": 1.8204250956210643e-05,
-      "loss": 0.5898,
+      "epoch": 4.236900240486673,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.819930629281133e-05,
+      "loss": 0.9888762664794922,
       "step": 368000
     },
     {
-      "epoch": 8.47840103159252,
-      "grad_norm": 2.907320976257324,
-      "learning_rate": 1.819697939238813e-05,
-      "loss": 0.58,
+      "epoch": 4.239202925561387,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8192034728988817e-05,
+      "loss": 0.9819789886474609,
       "step": 368200
     },
     {
-      "epoch": 8.483006355346781,
-      "grad_norm": 2.7009003162384033,
-      "learning_rate": 1.8189744186384725e-05,
-      "loss": 0.5916,
+      "epoch": 4.241505610636103,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8184763165166303e-05,
+      "loss": 0.99475341796875,
       "step": 368400
     },
     {
-      "epoch": 8.487611679101041,
-      "grad_norm": 2.902514696121216,
-      "learning_rate": 1.818247262256221e-05,
-      "loss": 0.5683,
+      "epoch": 4.243808295710817,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8177491601343786e-05,
+      "loss": 0.9711346435546875,
       "step": 368600
     },
     {
-      "epoch": 8.4922170028553,
-      "grad_norm": 3.5447628498077393,
-      "learning_rate": 1.8175201058739693e-05,
-      "loss": 0.5765,
+      "epoch": 4.246110980785533,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.817022003752127e-05,
+      "loss": 0.9897885894775391,
       "step": 368800
     },
     {
-      "epoch": 8.49682232660956,
-      "grad_norm": 3.0981380939483643,
-      "learning_rate": 1.8167929494917176e-05,
-      "loss": 0.5716,
+      "epoch": 4.248413665860247,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8162948473698754e-05,
+      "loss": 0.9859911346435547,
       "step": 369000
     },
     {
-      "epoch": 8.501427650363821,
-      "grad_norm": 2.814976453781128,
-      "learning_rate": 1.816065793109466e-05,
-      "loss": 0.5864,
+      "epoch": 4.250716350934962,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8155676909876237e-05,
+      "loss": 0.9784740447998047,
       "step": 369200
     },
     {
-      "epoch": 8.50603297411808,
-      "grad_norm": 3.170222043991089,
-      "learning_rate": 1.8153386367272144e-05,
-      "loss": 0.5754,
+      "epoch": 4.253019036009677,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8148405346053723e-05,
+      "loss": 1.0029774475097657,
       "step": 369400
     },
     {
-      "epoch": 8.51063829787234,
-      "grad_norm": 2.6524722576141357,
-      "learning_rate": 1.814611480344963e-05,
-      "loss": 0.5859,
+      "epoch": 4.255321721084392,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.814113378223121e-05,
+      "loss": 0.9826509857177734,
       "step": 369600
     },
     {
-      "epoch": 8.5152436216266,
-      "grad_norm": 3.715766191482544,
-      "learning_rate": 1.8138843239627116e-05,
-      "loss": 0.5889,
+      "epoch": 4.257624406159107,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.813386221840869e-05,
+      "loss": 0.9854049682617188,
       "step": 369800
     },
     {
-      "epoch": 8.51984894538086,
-      "grad_norm": 3.317150592803955,
-      "learning_rate": 1.8131571675804602e-05,
-      "loss": 0.5823,
+      "epoch": 4.259927091233822,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8126590654586177e-05,
+      "loss": 0.9904918670654297,
       "step": 370000
     },
     {
-      "epoch": 8.52445426913512,
-      "grad_norm": 3.7694554328918457,
-      "learning_rate": 1.812430011198208e-05,
-      "loss": 0.5776,
+      "epoch": 4.262229776308537,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.811931909076366e-05,
+      "loss": 0.978302001953125,
       "step": 370200
     },
     {
-      "epoch": 8.52905959288938,
-      "grad_norm": 3.328153371810913,
-      "learning_rate": 1.8117028548159567e-05,
-      "loss": 0.5895,
+      "epoch": 4.264532461383252,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.8112047526941142e-05,
+      "loss": 0.9758035278320313,
       "step": 370400
     },
     {
-      "epoch": 8.533664916643641,
-      "grad_norm": 3.546926498413086,
-      "learning_rate": 1.8109756984337053e-05,
-      "loss": 0.5798,
+      "epoch": 4.266835146457967,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8104775963118628e-05,
+      "loss": 0.9813452911376953,
       "step": 370600
     },
     {
-      "epoch": 8.5382702403979,
-      "grad_norm": 3.3275558948516846,
-      "learning_rate": 1.8102485420514535e-05,
-      "loss": 0.5787,
+      "epoch": 4.269137831532682,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8097504399296114e-05,
+      "loss": 0.9689815521240235,
       "step": 370800
     },
     {
-      "epoch": 8.54287556415216,
-      "grad_norm": 2.915025234222412,
-      "learning_rate": 1.809521385669202e-05,
-      "loss": 0.5742,
+      "epoch": 4.271440516607396,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.80902328354736e-05,
+      "loss": 0.985603256225586,
       "step": 371000
     },
     {
-      "epoch": 8.54748088790642,
-      "grad_norm": 2.634218454360962,
-      "learning_rate": 1.8087942292869507e-05,
-      "loss": 0.5869,
+      "epoch": 4.2737432016821115,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8082961271651082e-05,
+      "loss": 0.9839646148681641,
       "step": 371200
     },
     {
-      "epoch": 8.55208621166068,
-      "grad_norm": 3.156991481781006,
-      "learning_rate": 1.808067072904699e-05,
-      "loss": 0.582,
+      "epoch": 4.276045886756826,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8075689707828568e-05,
+      "loss": 0.9936179351806641,
       "step": 371400
     },
     {
-      "epoch": 8.55669153541494,
-      "grad_norm": 3.4311141967773438,
-      "learning_rate": 1.8073399165224472e-05,
-      "loss": 0.5955,
+      "epoch": 4.2783485718315415,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.806841814400605e-05,
+      "loss": 0.981746826171875,
       "step": 371600
     },
     {
-      "epoch": 8.5612968591692,
-      "grad_norm": 2.9205784797668457,
-      "learning_rate": 1.8066127601401958e-05,
-      "loss": 0.5826,
+      "epoch": 4.280651256906256,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8061146580183533e-05,
+      "loss": 0.954793701171875,
       "step": 371800
     },
     {
-      "epoch": 8.565902182923459,
-      "grad_norm": 3.045046329498291,
-      "learning_rate": 1.805885603757944e-05,
-      "loss": 0.5948,
+      "epoch": 4.2829539419809715,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.805387501636102e-05,
+      "loss": 0.9726139068603515,
       "step": 372000
     },
     {
-      "epoch": 8.57050750667772,
-      "grad_norm": 3.506364583969116,
-      "learning_rate": 1.8051584473756926e-05,
-      "loss": 0.5762,
+      "epoch": 4.285256627055686,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8046603452538505e-05,
+      "loss": 0.985401382446289,
       "step": 372200
     },
     {
-      "epoch": 8.57511283043198,
-      "grad_norm": 3.3249881267547607,
-      "learning_rate": 1.8044312909934412e-05,
-      "loss": 0.5784,
+      "epoch": 4.287559312130401,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8039331888715987e-05,
+      "loss": 0.9654906463623046,
       "step": 372400
     },
     {
-      "epoch": 8.579718154186239,
-      "grad_norm": 3.3042402267456055,
-      "learning_rate": 1.8037041346111898e-05,
-      "loss": 0.5907,
+      "epoch": 4.289861997205116,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8032060324893473e-05,
+      "loss": 0.9556643676757812,
       "step": 372600
     },
     {
-      "epoch": 8.584323477940499,
-      "grad_norm": 2.80226469039917,
-      "learning_rate": 1.8029769782289377e-05,
-      "loss": 0.5917,
+      "epoch": 4.2921646822798305,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8024788761070956e-05,
+      "loss": 0.9781275939941406,
       "step": 372800
     },
     {
-      "epoch": 8.58892880169476,
-      "grad_norm": 3.677151679992676,
-      "learning_rate": 1.8022498218466863e-05,
-      "loss": 0.5947,
+      "epoch": 4.294467367354546,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.801751719724844e-05,
+      "loss": 0.9931547546386719,
       "step": 373000
     },
     {
-      "epoch": 8.593534125449018,
-      "grad_norm": 2.8977012634277344,
-      "learning_rate": 1.801522665464435e-05,
-      "loss": 0.5859,
+      "epoch": 4.2967700524292605,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8010245633425924e-05,
+      "loss": 0.9784978485107422,
       "step": 373200
     },
     {
-      "epoch": 8.598139449203279,
-      "grad_norm": 3.019442558288574,
-      "learning_rate": 1.8007955090821832e-05,
-      "loss": 0.5862,
+      "epoch": 4.299072737503976,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.800297406960341e-05,
+      "loss": 0.9716792297363281,
       "step": 373400
     },
     {
-      "epoch": 8.60274477295754,
-      "grad_norm": 3.2348787784576416,
-      "learning_rate": 1.8000683526999318e-05,
-      "loss": 0.5909,
+      "epoch": 4.30137542257869,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7995702505780893e-05,
+      "loss": 0.976788330078125,
       "step": 373600
     },
     {
-      "epoch": 8.607350096711798,
-      "grad_norm": 3.2550971508026123,
-      "learning_rate": 1.7993411963176804e-05,
-      "loss": 0.5851,
+      "epoch": 4.303678107653406,
+      "grad_norm": 1.375,
+      "learning_rate": 1.798843094195838e-05,
+      "loss": 0.9814914703369141,
       "step": 373800
     },
     {
-      "epoch": 8.611955420466058,
-      "grad_norm": 2.6115951538085938,
-      "learning_rate": 1.7986140399354286e-05,
-      "loss": 0.5935,
+      "epoch": 4.30598079272812,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.798115937813586e-05,
+      "loss": 0.9726478576660156,
       "step": 374000
     },
     {
-      "epoch": 8.616560744220319,
-      "grad_norm": 3.2661094665527344,
-      "learning_rate": 1.797886883553177e-05,
-      "loss": 0.581,
+      "epoch": 4.308283477802836,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7973887814313347e-05,
+      "loss": 0.996259765625,
       "step": 374200
     },
     {
-      "epoch": 8.62116606797458,
-      "grad_norm": 3.3261373043060303,
-      "learning_rate": 1.7971597271709255e-05,
-      "loss": 0.5674,
+      "epoch": 4.31058616287755,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.796661625049083e-05,
+      "loss": 0.9984632873535156,
       "step": 374400
     },
     {
-      "epoch": 8.625771391728838,
-      "grad_norm": 3.81058669090271,
-      "learning_rate": 1.7964325707886737e-05,
-      "loss": 0.5717,
+      "epoch": 4.312888847952266,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.7959344686668316e-05,
+      "loss": 0.988074722290039,
       "step": 374600
     },
     {
-      "epoch": 8.630376715483099,
-      "grad_norm": 3.5395662784576416,
-      "learning_rate": 1.7957090501883336e-05,
-      "loss": 0.5872,
+      "epoch": 4.31519153302698,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.79520731228458e-05,
+      "loss": 0.9838126373291015,
       "step": 374800
     },
     {
-      "epoch": 8.634982039237359,
-      "grad_norm": 3.262457847595215,
-      "learning_rate": 1.794981893806082e-05,
-      "loss": 0.5776,
+      "epoch": 4.317494218101695,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7944801559023284e-05,
+      "loss": 0.9583352661132812,
       "step": 375000
     },
     {
-      "epoch": 8.639587362991618,
-      "grad_norm": 3.360276460647583,
-      "learning_rate": 1.7942547374238304e-05,
-      "loss": 0.5762,
+      "epoch": 4.31979690317641,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.793752999520077e-05,
+      "loss": 0.988289794921875,
       "step": 375200
     },
     {
-      "epoch": 8.644192686745878,
-      "grad_norm": 3.0414352416992188,
-      "learning_rate": 1.793527581041579e-05,
-      "loss": 0.5866,
+      "epoch": 4.322099588251125,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7930258431378252e-05,
+      "loss": 0.9876296234130859,
       "step": 375400
     },
     {
-      "epoch": 8.648798010500139,
-      "grad_norm": 3.696171283721924,
-      "learning_rate": 1.7928004246593273e-05,
-      "loss": 0.5764,
+      "epoch": 4.32440227332584,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.7922986867555735e-05,
+      "loss": 0.9777294921875,
       "step": 375600
     },
     {
-      "epoch": 8.653403334254397,
-      "grad_norm": 3.294072389602661,
-      "learning_rate": 1.7920732682770755e-05,
-      "loss": 0.5772,
+      "epoch": 4.326704958400555,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.791571530373322e-05,
+      "loss": 0.9797138977050781,
       "step": 375800
     },
     {
-      "epoch": 8.658008658008658,
-      "grad_norm": 3.592628240585327,
-      "learning_rate": 1.791346111894824e-05,
-      "loss": 0.5712,
+      "epoch": 4.32900764347527,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7908443739910707e-05,
+      "loss": 0.9727606964111328,
       "step": 376000
     },
     {
-      "epoch": 8.662613981762918,
-      "grad_norm": 2.2415127754211426,
-      "learning_rate": 1.7906189555125727e-05,
-      "loss": 0.5692,
+      "epoch": 4.331310328549985,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.790117217608819e-05,
+      "loss": 0.9707086944580078,
       "step": 376200
     },
     {
-      "epoch": 8.667219305517177,
-      "grad_norm": 3.072105646133423,
-      "learning_rate": 1.7898954349122322e-05,
-      "loss": 0.5747,
+      "epoch": 4.333613013624699,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7893900612265675e-05,
+      "loss": 0.9848200225830078,
       "step": 376400
     },
     {
-      "epoch": 8.671824629271438,
-      "grad_norm": 3.045968770980835,
-      "learning_rate": 1.7891682785299808e-05,
-      "loss": 0.5738,
+      "epoch": 4.335915698699415,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7886629048443158e-05,
+      "loss": 0.9899727630615235,
       "step": 376600
     },
     {
-      "epoch": 8.676429953025698,
-      "grad_norm": 3.102426052093506,
-      "learning_rate": 1.7884411221477294e-05,
-      "loss": 0.5619,
+      "epoch": 4.338218383774129,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.787935748462064e-05,
+      "loss": 1.0073524475097657,
       "step": 376800
     },
     {
-      "epoch": 8.681035276779959,
-      "grad_norm": 2.7489142417907715,
-      "learning_rate": 1.7877139657654773e-05,
-      "loss": 0.5673,
+      "epoch": 4.340521068848845,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7872085920798126e-05,
+      "loss": 0.9903134155273438,
       "step": 377000
     },
     {
-      "epoch": 8.685640600534217,
-      "grad_norm": 2.9566493034362793,
-      "learning_rate": 1.786986809383226e-05,
-      "loss": 0.5868,
+      "epoch": 4.342823753923559,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7864814356975612e-05,
+      "loss": 0.9808146667480468,
       "step": 377200
     },
     {
-      "epoch": 8.690245924288478,
-      "grad_norm": 3.6835222244262695,
-      "learning_rate": 1.7862596530009745e-05,
-      "loss": 0.5878,
+      "epoch": 4.345126438998275,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7857542793153098e-05,
+      "loss": 0.989017105102539,
       "step": 377400
     },
     {
-      "epoch": 8.694851248042738,
-      "grad_norm": 3.0927047729492188,
-      "learning_rate": 1.7855324966187227e-05,
-      "loss": 0.59,
+      "epoch": 4.347429124072989,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.785027122933058e-05,
+      "loss": 0.95684326171875,
       "step": 377600
     },
     {
-      "epoch": 8.699456571796997,
-      "grad_norm": 3.5249454975128174,
-      "learning_rate": 1.7848053402364713e-05,
-      "loss": 0.5731,
+      "epoch": 4.349731809147705,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7842999665508067e-05,
+      "loss": 0.9856446075439453,
       "step": 377800
     },
     {
-      "epoch": 8.704061895551257,
-      "grad_norm": 2.711740732192993,
-      "learning_rate": 1.78407818385422e-05,
-      "loss": 0.587,
+      "epoch": 4.352034494222419,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.783572810168555e-05,
+      "loss": 0.9768900299072265,
       "step": 378000
     },
     {
-      "epoch": 8.708667219305518,
-      "grad_norm": 3.440530776977539,
-      "learning_rate": 1.7833510274719682e-05,
-      "loss": 0.582,
+      "epoch": 4.354337179297135,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.782845653786303e-05,
+      "loss": 0.9693342590332031,
       "step": 378200
     },
     {
-      "epoch": 8.713272543059777,
-      "grad_norm": 3.404754400253296,
-      "learning_rate": 1.7826238710897164e-05,
-      "loss": 0.585,
+      "epoch": 4.356639864371849,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.7821184974040517e-05,
+      "loss": 0.9808509826660157,
       "step": 378400
     },
     {
-      "epoch": 8.717877866814037,
-      "grad_norm": 3.556629180908203,
-      "learning_rate": 1.781896714707465e-05,
-      "loss": 0.5731,
+      "epoch": 4.358942549446564,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7813913410218003e-05,
+      "loss": 1.0005860137939453,
       "step": 378600
     },
     {
-      "epoch": 8.722483190568298,
-      "grad_norm": 2.7652220726013184,
-      "learning_rate": 1.7811695583252133e-05,
-      "loss": 0.576,
+      "epoch": 4.361245234521279,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7806641846395486e-05,
+      "loss": 1.0117581176757813,
       "step": 378800
     },
     {
-      "epoch": 8.727088514322556,
-      "grad_norm": 3.707284450531006,
-      "learning_rate": 1.780442401942962e-05,
-      "loss": 0.5766,
+      "epoch": 4.363547919595994,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7799370282572972e-05,
+      "loss": 0.9801927185058594,
       "step": 379000
     },
     {
-      "epoch": 8.731693838076817,
-      "grad_norm": 2.5321168899536133,
-      "learning_rate": 1.7797152455607105e-05,
-      "loss": 0.5838,
+      "epoch": 4.365850604670709,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7792098718750454e-05,
+      "loss": 0.9840756225585937,
       "step": 379200
     },
     {
-      "epoch": 8.736299161831077,
-      "grad_norm": 2.458303213119507,
-      "learning_rate": 1.778988089178459e-05,
-      "loss": 0.5953,
+      "epoch": 4.368153289745424,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7784827154927937e-05,
+      "loss": 0.9926817321777344,
       "step": 379400
     },
     {
-      "epoch": 8.740904485585336,
-      "grad_norm": 3.712862253189087,
-      "learning_rate": 1.778260932796207e-05,
-      "loss": 0.5852,
+      "epoch": 4.370455974820139,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.7777555591105423e-05,
+      "loss": 0.9944758605957031,
       "step": 379600
     },
     {
-      "epoch": 8.745509809339596,
-      "grad_norm": 3.2254207134246826,
-      "learning_rate": 1.7775337764139556e-05,
-      "loss": 0.5876,
+      "epoch": 4.372758659894854,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.777028402728291e-05,
+      "loss": 0.9793479919433594,
       "step": 379800
     },
     {
-      "epoch": 8.750115133093857,
-      "grad_norm": 2.612973690032959,
-      "learning_rate": 1.776806620031704e-05,
-      "loss": 0.5762,
+      "epoch": 4.375061344969568,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7763012463460395e-05,
+      "loss": 0.9661212921142578,
       "step": 380000
     },
     {
-      "epoch": 8.754720456848116,
-      "grad_norm": 3.134018898010254,
-      "learning_rate": 1.7760794636494524e-05,
-      "loss": 0.573,
+      "epoch": 4.3773640300442835,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7755740899637877e-05,
+      "loss": 0.9963236999511719,
       "step": 380200
     },
     {
-      "epoch": 8.759325780602376,
-      "grad_norm": 3.392620801925659,
-      "learning_rate": 1.7753559430491123e-05,
-      "loss": 0.5713,
+      "epoch": 4.379666715118998,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.774846933581536e-05,
+      "loss": 0.9838841247558594,
       "step": 380400
     },
     {
-      "epoch": 8.763931104356637,
-      "grad_norm": 2.867246627807617,
-      "learning_rate": 1.774628786666861e-05,
-      "loss": 0.5774,
+      "epoch": 4.3819694001937135,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7741197771992846e-05,
+      "loss": 0.9869212341308594,
       "step": 380600
     },
     {
-      "epoch": 8.768536428110895,
-      "grad_norm": 2.381554126739502,
-      "learning_rate": 1.7739052660665204e-05,
-      "loss": 0.5897,
+      "epoch": 4.384272085268428,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7733926208170328e-05,
+      "loss": 0.9717589569091797,
       "step": 380800
     },
     {
-      "epoch": 8.773141751865156,
-      "grad_norm": 2.7664806842803955,
-      "learning_rate": 1.773178109684269e-05,
-      "loss": 0.5864,
+      "epoch": 4.3865747703431435,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7726654644347814e-05,
+      "loss": 0.9582798767089844,
       "step": 381000
     },
     {
-      "epoch": 8.777747075619416,
-      "grad_norm": 3.2918121814727783,
-      "learning_rate": 1.7724509533020172e-05,
-      "loss": 0.5768,
+      "epoch": 4.388877455417858,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.77193830805253e-05,
+      "loss": 0.9778752899169922,
       "step": 381200
     },
     {
-      "epoch": 8.782352399373677,
-      "grad_norm": 3.4607903957366943,
-      "learning_rate": 1.7717237969197655e-05,
-      "loss": 0.5853,
+      "epoch": 4.391180140492573,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7712111516702782e-05,
+      "loss": 1.0042454528808593,
       "step": 381400
     },
     {
-      "epoch": 8.786957723127935,
-      "grad_norm": 3.931750535964966,
-      "learning_rate": 1.770996640537514e-05,
-      "loss": 0.5778,
+      "epoch": 4.393482825567288,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.770483995288027e-05,
+      "loss": 0.9691288757324219,
       "step": 381600
     },
     {
-      "epoch": 8.791563046882196,
-      "grad_norm": 2.7144854068756104,
-      "learning_rate": 1.7702694841552623e-05,
-      "loss": 0.5793,
+      "epoch": 4.395785510642003,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.769756838905775e-05,
+      "loss": 0.9935425567626953,
       "step": 381800
     },
     {
-      "epoch": 8.796168370636456,
-      "grad_norm": 3.4606308937072754,
-      "learning_rate": 1.769542327773011e-05,
-      "loss": 0.5813,
+      "epoch": 4.398088195716718,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7690296825235233e-05,
+      "loss": 0.9775373840332031,
       "step": 382000
     },
     {
-      "epoch": 8.800773694390715,
-      "grad_norm": 2.6973564624786377,
-      "learning_rate": 1.7688151713907595e-05,
-      "loss": 0.587,
+      "epoch": 4.4003908807914325,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.768302526141272e-05,
+      "loss": 0.9748054504394531,
       "step": 382200
     },
     {
-      "epoch": 8.805379018144976,
-      "grad_norm": 2.9456260204315186,
-      "learning_rate": 1.768088015008508e-05,
-      "loss": 0.5725,
+      "epoch": 4.402693565866148,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7675753697590205e-05,
+      "loss": 0.9840547943115234,
       "step": 382400
     },
     {
-      "epoch": 8.809984341899236,
-      "grad_norm": 3.2708475589752197,
-      "learning_rate": 1.767360858626256e-05,
-      "loss": 0.5781,
+      "epoch": 4.4049962509408624,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.7668482133767688e-05,
+      "loss": 0.9826657867431641,
       "step": 382600
     },
     {
-      "epoch": 8.814589665653495,
-      "grad_norm": 3.055715799331665,
-      "learning_rate": 1.7666337022440046e-05,
-      "loss": 0.5829,
+      "epoch": 4.407298936015578,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7661210569945174e-05,
+      "loss": 0.9580550384521485,
       "step": 382800
     },
     {
-      "epoch": 8.819194989407755,
-      "grad_norm": 2.2745134830474854,
-      "learning_rate": 1.7659065458617532e-05,
-      "loss": 0.5836,
+      "epoch": 4.409601621090292,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7653939006122656e-05,
+      "loss": 0.963132553100586,
       "step": 383000
     },
     {
-      "epoch": 8.823800313162016,
-      "grad_norm": 2.850818157196045,
-      "learning_rate": 1.7651793894795014e-05,
-      "loss": 0.5843,
+      "epoch": 4.411904306165008,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7646667442300142e-05,
+      "loss": 0.9853313446044922,
       "step": 383200
     },
     {
-      "epoch": 8.828405636916274,
-      "grad_norm": 3.06870436668396,
-      "learning_rate": 1.76445223309725e-05,
-      "loss": 0.5718,
+      "epoch": 4.414206991239722,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7639395878477625e-05,
+      "loss": 0.9874549865722656,
       "step": 383400
     },
     {
-      "epoch": 8.833010960670535,
-      "grad_norm": 3.3040614128112793,
-      "learning_rate": 1.7637250767149986e-05,
-      "loss": 0.5822,
+      "epoch": 4.416509676314438,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.763212431465511e-05,
+      "loss": 0.9777555847167969,
       "step": 383600
     },
     {
-      "epoch": 8.837616284424795,
-      "grad_norm": 2.4403414726257324,
-      "learning_rate": 1.7629979203327465e-05,
-      "loss": 0.5899,
+      "epoch": 4.418812361389152,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7624852750832596e-05,
+      "loss": 0.974601058959961,
       "step": 383800
     },
     {
-      "epoch": 8.842221608179056,
-      "grad_norm": 2.8603031635284424,
-      "learning_rate": 1.762270763950495e-05,
-      "loss": 0.5811,
+      "epoch": 4.421115046463867,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.761758118701008e-05,
+      "loss": 0.9725660705566406,
       "step": 384000
     },
     {
-      "epoch": 8.846826931933315,
-      "grad_norm": 2.7087676525115967,
-      "learning_rate": 1.7615436075682437e-05,
-      "loss": 0.5904,
+      "epoch": 4.423417731538582,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.761030962318756e-05,
+      "loss": 0.9747721862792968,
       "step": 384200
     },
     {
-      "epoch": 8.851432255687575,
-      "grad_norm": 2.9180452823638916,
-      "learning_rate": 1.760816451185992e-05,
-      "loss": 0.5789,
+      "epoch": 4.425720416613297,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.7603038059365047e-05,
+      "loss": 0.98368896484375,
       "step": 384400
     },
     {
-      "epoch": 8.856037579441836,
-      "grad_norm": 3.6480114459991455,
-      "learning_rate": 1.7600892948037406e-05,
-      "loss": 0.5775,
+      "epoch": 4.428023101688012,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.759576649554253e-05,
+      "loss": 0.9807275390625,
       "step": 384600
     },
     {
-      "epoch": 8.860642903196094,
-      "grad_norm": 3.7235474586486816,
-      "learning_rate": 1.7593657742034004e-05,
-      "loss": 0.5756,
+      "epoch": 4.430325786762727,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7588494931720016e-05,
+      "loss": 0.9803495788574219,
       "step": 384800
     },
     {
-      "epoch": 8.865248226950355,
-      "grad_norm": 2.3067126274108887,
-      "learning_rate": 1.7586386178211487e-05,
-      "loss": 0.5848,
+      "epoch": 4.432628471837442,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7581223367897502e-05,
+      "loss": 0.9786897277832032,
       "step": 385000
     },
     {
-      "epoch": 8.869853550704615,
-      "grad_norm": 3.3306350708007812,
-      "learning_rate": 1.7579114614388973e-05,
-      "loss": 0.5807,
+      "epoch": 4.434931156912157,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.7573951804074984e-05,
+      "loss": 1.00064453125,
       "step": 385200
     },
     {
-      "epoch": 8.874458874458874,
-      "grad_norm": 3.0547428131103516,
-      "learning_rate": 1.7571879408385568e-05,
-      "loss": 0.5596,
+      "epoch": 4.437233841986872,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.756668024025247e-05,
+      "loss": 0.9689328002929688,
       "step": 385400
     },
     {
-      "epoch": 8.879064198213134,
-      "grad_norm": 3.389777660369873,
-      "learning_rate": 1.756460784456305e-05,
-      "loss": 0.5688,
+      "epoch": 4.439536527061587,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7559408676429953e-05,
+      "loss": 0.9944960021972656,
       "step": 385600
     },
     {
-      "epoch": 8.883669521967395,
-      "grad_norm": 3.0648694038391113,
-      "learning_rate": 1.7557336280740536e-05,
-      "loss": 0.5934,
+      "epoch": 4.441839212136301,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.755213711260744e-05,
+      "loss": 0.9811604309082032,
       "step": 385800
     },
     {
-      "epoch": 8.888274845721654,
-      "grad_norm": 2.877584457397461,
-      "learning_rate": 1.7550064716918022e-05,
-      "loss": 0.5749,
+      "epoch": 4.444141897211017,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.754486554878492e-05,
+      "loss": 0.9793187713623047,
       "step": 386000
     },
     {
-      "epoch": 8.892880169475914,
-      "grad_norm": 3.3926568031311035,
-      "learning_rate": 1.7542793153095505e-05,
-      "loss": 0.5716,
+      "epoch": 4.446444582285731,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.7537593984962407e-05,
+      "loss": 1.0037923431396485,
       "step": 386200
     },
     {
-      "epoch": 8.897485493230175,
-      "grad_norm": 2.8832592964172363,
-      "learning_rate": 1.753552158927299e-05,
-      "loss": 0.577,
+      "epoch": 4.448747267360447,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.7530322421139893e-05,
+      "loss": 0.9684130096435547,
       "step": 386400
     },
     {
-      "epoch": 8.902090816984433,
-      "grad_norm": 3.812521457672119,
-      "learning_rate": 1.7528250025450477e-05,
-      "loss": 0.5931,
+      "epoch": 4.451049952435161,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.7523050857317376e-05,
+      "loss": 0.9824868774414063,
       "step": 386600
     },
     {
-      "epoch": 8.906696140738694,
-      "grad_norm": 3.376232624053955,
-      "learning_rate": 1.7520978461627956e-05,
-      "loss": 0.5758,
+      "epoch": 4.453352637509877,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.7515779293494858e-05,
+      "loss": 0.9914022827148438,
       "step": 386800
     },
     {
-      "epoch": 8.911301464492954,
-      "grad_norm": 3.098276138305664,
-      "learning_rate": 1.751370689780544e-05,
-      "loss": 0.579,
+      "epoch": 4.455655322584591,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.7508507729672344e-05,
+      "loss": 0.9939438629150391,
       "step": 387000
     },
     {
-      "epoch": 8.915906788247213,
-      "grad_norm": 3.7928085327148438,
-      "learning_rate": 1.7506435333982927e-05,
-      "loss": 0.5814,
+      "epoch": 4.457958007659307,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7501236165849827e-05,
+      "loss": 0.978121109008789,
       "step": 387200
     },
     {
-      "epoch": 8.920512112001473,
-      "grad_norm": 2.6792008876800537,
-      "learning_rate": 1.749916377016041e-05,
-      "loss": 0.5924,
+      "epoch": 4.460260692734021,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7493964602027312e-05,
+      "loss": 1.0017222595214843,
       "step": 387400
     },
     {
-      "epoch": 8.925117435755734,
-      "grad_norm": 3.9660255908966064,
-      "learning_rate": 1.7491892206337896e-05,
-      "loss": 0.5773,
+      "epoch": 4.462563377808736,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.74866930382048e-05,
+      "loss": 1.0084938049316405,
       "step": 387600
     },
     {
-      "epoch": 8.929722759509993,
-      "grad_norm": 3.670153856277466,
-      "learning_rate": 1.7484657000334495e-05,
-      "loss": 0.5863,
+      "epoch": 4.464866062883451,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.747942147438228e-05,
+      "loss": 0.9803755187988281,
       "step": 387800
     },
     {
-      "epoch": 8.934328083264253,
-      "grad_norm": 2.9507553577423096,
-      "learning_rate": 1.7477385436511977e-05,
-      "loss": 0.5849,
+      "epoch": 4.467168747958166,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7472149910559767e-05,
+      "loss": 0.9992539978027344,
       "step": 388000
     },
     {
-      "epoch": 8.938933407018514,
-      "grad_norm": 3.5913052558898926,
-      "learning_rate": 1.7470113872689463e-05,
-      "loss": 0.5853,
+      "epoch": 4.469471433032881,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.746487834673725e-05,
+      "loss": 0.9823341369628906,
       "step": 388200
     },
     {
-      "epoch": 8.943538730772774,
-      "grad_norm": 3.083280086517334,
-      "learning_rate": 1.7462842308866945e-05,
-      "loss": 0.5752,
+      "epoch": 4.471774118107596,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7457606782914732e-05,
+      "loss": 0.9724301147460938,
       "step": 388400
     },
     {
-      "epoch": 8.948144054527033,
-      "grad_norm": 3.377192497253418,
-      "learning_rate": 1.7455570745044428e-05,
-      "loss": 0.5705,
+      "epoch": 4.474076803182311,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7450335219092218e-05,
+      "loss": 0.98064697265625,
       "step": 388600
     },
     {
-      "epoch": 8.952749378281293,
-      "grad_norm": 3.040677785873413,
-      "learning_rate": 1.7448299181221914e-05,
-      "loss": 0.5838,
+      "epoch": 4.476379488257026,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7443063655269704e-05,
+      "loss": 0.9858995056152344,
       "step": 388800
     },
     {
-      "epoch": 8.957354702035554,
-      "grad_norm": 2.774031400680542,
-      "learning_rate": 1.74410276173994e-05,
-      "loss": 0.5814,
+      "epoch": 4.478682173331741,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.743579209144719e-05,
+      "loss": 0.9684735870361328,
       "step": 389000
     },
     {
-      "epoch": 8.961960025789812,
-      "grad_norm": 3.28867244720459,
-      "learning_rate": 1.7433756053576882e-05,
-      "loss": 0.5915,
+      "epoch": 4.4809848584064556,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.7428520527624672e-05,
+      "loss": 0.982616195678711,
       "step": 389200
     },
     {
-      "epoch": 8.966565349544073,
-      "grad_norm": 2.8730006217956543,
-      "learning_rate": 1.7426484489754368e-05,
-      "loss": 0.579,
+      "epoch": 4.48328754348117,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7421248963802155e-05,
+      "loss": 0.9835112762451171,
       "step": 389400
     },
     {
-      "epoch": 8.971170673298333,
-      "grad_norm": 3.185356855392456,
-      "learning_rate": 1.741921292593185e-05,
-      "loss": 0.5694,
+      "epoch": 4.4855902285558855,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.741397739997964e-05,
+      "loss": 0.9883460998535156,
       "step": 389600
     },
     {
-      "epoch": 8.975775997052592,
-      "grad_norm": 3.238297700881958,
-      "learning_rate": 1.7411941362109333e-05,
-      "loss": 0.5869,
+      "epoch": 4.4878929136306,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7406705836157123e-05,
+      "loss": 0.9911152648925782,
       "step": 389800
     },
     {
-      "epoch": 8.980381320806853,
-      "grad_norm": 2.7840898036956787,
-      "learning_rate": 1.740466979828682e-05,
-      "loss": 0.5765,
+      "epoch": 4.4901955987053155,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.739943427233461e-05,
+      "loss": 0.9605591583251953,
       "step": 390000
     },
     {
-      "epoch": 8.984986644561113,
-      "grad_norm": 3.177186965942383,
-      "learning_rate": 1.7397398234464305e-05,
-      "loss": 0.587,
+      "epoch": 4.49249828378003,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7392162708512095e-05,
+      "loss": 0.9712622833251953,
       "step": 390200
     },
     {
-      "epoch": 8.989591968315372,
-      "grad_norm": 2.8901917934417725,
-      "learning_rate": 1.739012667064179e-05,
-      "loss": 0.5862,
+      "epoch": 4.4948009688547454,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.7384891144689577e-05,
+      "loss": 0.9833880615234375,
       "step": 390400
     },
     {
-      "epoch": 8.994197292069632,
-      "grad_norm": 2.768673896789551,
-      "learning_rate": 1.7382855106819274e-05,
-      "loss": 0.584,
+      "epoch": 4.49710365392946,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.737761958086706e-05,
+      "loss": 0.9929757690429688,
       "step": 390600
     },
     {
-      "epoch": 8.998802615823893,
-      "grad_norm": 2.805500030517578,
-      "learning_rate": 1.7375583542996756e-05,
-      "loss": 0.5783,
+      "epoch": 4.499406339004175,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7370348017044546e-05,
+      "loss": 0.9671168518066406,
       "step": 390800
     },
     {
-      "epoch": 9.0,
-      "eval_loss": 0.5523322224617004,
-      "eval_runtime": 146.0681,
-      "eval_samples_per_second": 194.163,
-      "eval_steps_per_second": 12.138,
-      "step": 390852
-    },
-    {
-      "epoch": 9.003407939578153,
-      "grad_norm": 4.481179714202881,
-      "learning_rate": 1.736834833699335e-05,
-      "loss": 0.5692,
+      "epoch": 4.50170902407889,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.736307645322203e-05,
+      "loss": 0.9829707336425781,
       "step": 391000
     },
     {
-      "epoch": 9.008013263332412,
-      "grad_norm": 3.07963228225708,
-      "learning_rate": 1.7361076773170837e-05,
-      "loss": 0.5692,
+      "epoch": 4.504011709153605,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7355804889399514e-05,
+      "loss": 0.9815752410888672,
       "step": 391200
     },
     {
-      "epoch": 9.012618587086672,
-      "grad_norm": 3.5437726974487305,
-      "learning_rate": 1.7353805209348323e-05,
-      "loss": 0.5569,
+      "epoch": 4.50631439422832,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7348533325577e-05,
+      "loss": 0.9918634796142578,
       "step": 391400
     },
     {
-      "epoch": 9.017223910840933,
-      "grad_norm": 4.162001609802246,
-      "learning_rate": 1.7346533645525806e-05,
-      "loss": 0.5681,
+      "epoch": 4.5086170793030345,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7341261761754483e-05,
+      "loss": 0.993114013671875,
       "step": 391600
     },
     {
-      "epoch": 9.021829234595192,
-      "grad_norm": 3.081472873687744,
-      "learning_rate": 1.733926208170329e-05,
-      "loss": 0.5798,
+      "epoch": 4.51091976437775,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.733399019793197e-05,
+      "loss": 0.9865582275390625,
       "step": 391800
     },
     {
-      "epoch": 9.026434558349452,
-      "grad_norm": 2.833381175994873,
-      "learning_rate": 1.7331990517880778e-05,
-      "loss": 0.5837,
+      "epoch": 4.513222449452464,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.732671863410945e-05,
+      "loss": 0.9887853240966797,
       "step": 392000
     },
     {
-      "epoch": 9.031039882103713,
-      "grad_norm": 3.1221764087677,
-      "learning_rate": 1.7324718954058263e-05,
-      "loss": 0.5657,
+      "epoch": 4.51552513452718,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.7319447070286937e-05,
+      "loss": 0.987093734741211,
       "step": 392200
     },
     {
-      "epoch": 9.035645205857971,
-      "grad_norm": 3.7719027996063232,
-      "learning_rate": 1.7317447390235743e-05,
-      "loss": 0.5752,
+      "epoch": 4.517827819601894,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.731217550646442e-05,
+      "loss": 0.9806004333496093,
       "step": 392400
     },
     {
-      "epoch": 9.040250529612232,
-      "grad_norm": 3.361680507659912,
-      "learning_rate": 1.731017582641323e-05,
-      "loss": 0.5688,
+      "epoch": 4.52013050467661,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.7304903942641906e-05,
+      "loss": 0.9670136260986328,
       "step": 392600
     },
     {
-      "epoch": 9.044855853366492,
-      "grad_norm": 3.1694486141204834,
-      "learning_rate": 1.7302904262590714e-05,
-      "loss": 0.5802,
+      "epoch": 4.522433189751324,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.729763237881939e-05,
+      "loss": 0.9552702331542968,
       "step": 392800
     },
     {
-      "epoch": 9.049461177120751,
-      "grad_norm": 3.455324649810791,
-      "learning_rate": 1.7295632698768197e-05,
-      "loss": 0.5688,
+      "epoch": 4.524735874826039,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.7290360814996874e-05,
+      "loss": 0.9844831085205078,
       "step": 393000
     },
     {
-      "epoch": 9.054066500875011,
-      "grad_norm": 3.718388795852661,
-      "learning_rate": 1.7288361134945683e-05,
-      "loss": 0.5823,
+      "epoch": 4.527038559900754,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7283089251174356e-05,
+      "loss": 0.9598838806152343,
       "step": 393200
     },
     {
-      "epoch": 9.058671824629272,
-      "grad_norm": 3.018453359603882,
-      "learning_rate": 1.728108957112317e-05,
-      "loss": 0.5785,
+      "epoch": 4.529341244975469,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7275817687351842e-05,
+      "loss": 0.9726866912841797,
       "step": 393400
     },
     {
-      "epoch": 9.06327714838353,
-      "grad_norm": 2.9057810306549072,
-      "learning_rate": 1.7273818007300648e-05,
-      "loss": 0.5837,
+      "epoch": 4.531643930050184,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7268546123529325e-05,
+      "loss": 0.9710126495361329,
       "step": 393600
     },
     {
-      "epoch": 9.067882472137791,
-      "grad_norm": 3.081549644470215,
-      "learning_rate": 1.7266546443478134e-05,
-      "loss": 0.5735,
+      "epoch": 4.533946615124899,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.726127455970681e-05,
+      "loss": 0.9930452728271484,
       "step": 393800
     },
     {
-      "epoch": 9.072487795892052,
-      "grad_norm": 3.056722640991211,
-      "learning_rate": 1.725927487965562e-05,
-      "loss": 0.5877,
+      "epoch": 4.536249300199614,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7254002995884297e-05,
+      "loss": 0.9848808288574219,
       "step": 394000
     },
     {
-      "epoch": 9.07709311964631,
-      "grad_norm": 2.4788143634796143,
-      "learning_rate": 1.7252003315833102e-05,
-      "loss": 0.562,
+      "epoch": 4.538551985274329,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.724673143206178e-05,
+      "loss": 0.9978121185302734,
       "step": 394200
     },
     {
-      "epoch": 9.08169844340057,
-      "grad_norm": 2.9332191944122314,
-      "learning_rate": 1.7244731752010588e-05,
-      "loss": 0.575,
+      "epoch": 4.540854670349044,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7239459868239262e-05,
+      "loss": 0.9887928009033203,
       "step": 394400
     },
     {
-      "epoch": 9.086303767154831,
-      "grad_norm": 2.4606101512908936,
-      "learning_rate": 1.7237460188188074e-05,
-      "loss": 0.579,
+      "epoch": 4.543157355423759,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7232188304416748e-05,
+      "loss": 0.9775355529785156,
       "step": 394600
     },
     {
-      "epoch": 9.090909090909092,
-      "grad_norm": 2.8705406188964844,
-      "learning_rate": 1.7230188624365553e-05,
-      "loss": 0.5667,
+      "epoch": 4.545460040498474,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7224916740594234e-05,
+      "loss": 0.9925399017333985,
       "step": 394800
     },
     {
-      "epoch": 9.09551441466335,
-      "grad_norm": 3.100910186767578,
-      "learning_rate": 1.722291706054304e-05,
-      "loss": 0.5774,
+      "epoch": 4.547762725573189,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7217645176771716e-05,
+      "loss": 0.9753104400634766,
       "step": 395000
     },
     {
-      "epoch": 9.100119738417611,
-      "grad_norm": 3.324007034301758,
-      "learning_rate": 1.7215645496720525e-05,
-      "loss": 0.579,
+      "epoch": 4.550065410647903,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7210373612949202e-05,
+      "loss": 0.9818225860595703,
       "step": 395200
     },
     {
-      "epoch": 9.104725062171871,
-      "grad_norm": 3.0006039142608643,
-      "learning_rate": 1.720837393289801e-05,
-      "loss": 0.5762,
+      "epoch": 4.552368095722619,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7203102049126688e-05,
+      "loss": 0.9720571899414062,
       "step": 395400
     },
     {
-      "epoch": 9.10933038592613,
-      "grad_norm": 3.3208751678466797,
-      "learning_rate": 1.7201102369075493e-05,
-      "loss": 0.5756,
+      "epoch": 4.554670780797333,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.719583048530417e-05,
+      "loss": 0.9696012115478516,
       "step": 395600
     },
     {
-      "epoch": 9.11393570968039,
-      "grad_norm": 2.7627346515655518,
-      "learning_rate": 1.7193867163072092e-05,
-      "loss": 0.5721,
+      "epoch": 4.556973465872049,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7188558921481653e-05,
+      "loss": 0.9840654754638671,
       "step": 395800
     },
     {
-      "epoch": 9.118541033434651,
-      "grad_norm": 2.2746336460113525,
-      "learning_rate": 1.7186595599249575e-05,
-      "loss": 0.5729,
+      "epoch": 4.559276150946763,
+      "grad_norm": 1.0,
+      "learning_rate": 1.718128735765914e-05,
+      "loss": 0.9878695678710937,
       "step": 396000
     },
     {
-      "epoch": 9.12314635718891,
-      "grad_norm": 3.013868808746338,
-      "learning_rate": 1.717932403542706e-05,
-      "loss": 0.5684,
+      "epoch": 4.561578836021479,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.717401579383662e-05,
+      "loss": 0.9655953216552734,
       "step": 396200
     },
     {
-      "epoch": 9.12775168094317,
-      "grad_norm": 3.658040761947632,
-      "learning_rate": 1.7172052471604543e-05,
-      "loss": 0.5754,
+      "epoch": 4.563881521096193,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7166744230014107e-05,
+      "loss": 0.9800299835205079,
       "step": 396400
     },
     {
-      "epoch": 9.13235700469743,
-      "grad_norm": 3.7302823066711426,
-      "learning_rate": 1.7164780907782026e-05,
-      "loss": 0.5768,
+      "epoch": 4.566184206170908,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7159472666191593e-05,
+      "loss": 0.971158447265625,
       "step": 396600
     },
     {
-      "epoch": 9.13696232845169,
-      "grad_norm": 3.326700448989868,
-      "learning_rate": 1.715750934395951e-05,
-      "loss": 0.5704,
+      "epoch": 4.568486891245623,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.7152201102369076e-05,
+      "loss": 0.9807333374023437,
       "step": 396800
     },
     {
-      "epoch": 9.14156765220595,
-      "grad_norm": 2.6242971420288086,
-      "learning_rate": 1.7150237780136997e-05,
-      "loss": 0.5749,
+      "epoch": 4.570789576320338,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.714492953854656e-05,
+      "loss": 0.9847396087646484,
       "step": 397000
     },
     {
-      "epoch": 9.14617297596021,
-      "grad_norm": 3.557640552520752,
-      "learning_rate": 1.7142966216314483e-05,
-      "loss": 0.5707,
+      "epoch": 4.573092261395053,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7137657974724044e-05,
+      "loss": 0.9942626190185547,
       "step": 397200
     },
     {
-      "epoch": 9.150778299714469,
-      "grad_norm": 4.010790824890137,
-      "learning_rate": 1.7135694652491966e-05,
-      "loss": 0.5678,
+      "epoch": 4.575394946469768,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7130386410901527e-05,
+      "loss": 0.9864826965332031,
       "step": 397400
     },
     {
-      "epoch": 9.15538362346873,
-      "grad_norm": 2.8452792167663574,
-      "learning_rate": 1.7128459446488564e-05,
-      "loss": 0.5773,
+      "epoch": 4.577697631544483,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7123114847079013e-05,
+      "loss": 0.9885680389404297,
       "step": 397600
     },
     {
-      "epoch": 9.15998894722299,
-      "grad_norm": 3.276362895965576,
-      "learning_rate": 1.7121187882666044e-05,
-      "loss": 0.5768,
+      "epoch": 4.580000316619198,
+      "grad_norm": 0.5,
+      "learning_rate": 1.71158432832565e-05,
+      "loss": 1.004285888671875,
       "step": 397800
     },
     {
-      "epoch": 9.16459427097725,
-      "grad_norm": 2.479546546936035,
-      "learning_rate": 1.711391631884353e-05,
-      "loss": 0.5864,
+      "epoch": 4.582303001693913,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.7108571719433985e-05,
+      "loss": 0.9720408630371093,
       "step": 398000
     },
     {
-      "epoch": 9.16919959473151,
-      "grad_norm": 2.4583699703216553,
-      "learning_rate": 1.7106644755021015e-05,
-      "loss": 0.5769,
+      "epoch": 4.584605686768628,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7101300155611467e-05,
+      "loss": 0.982972412109375,
       "step": 398200
     },
     {
-      "epoch": 9.17380491848577,
-      "grad_norm": 3.225541830062866,
-      "learning_rate": 1.70993731911985e-05,
-      "loss": 0.5712,
+      "epoch": 4.586908371843343,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.709402859178895e-05,
+      "loss": 0.9912247467041015,
       "step": 398400
     },
     {
-      "epoch": 9.17841024224003,
-      "grad_norm": 2.6798887252807617,
-      "learning_rate": 1.7092101627375984e-05,
-      "loss": 0.5776,
+      "epoch": 4.5892110569180575,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7086757027966435e-05,
+      "loss": 0.9738587188720703,
       "step": 398600
     },
     {
-      "epoch": 9.183015565994289,
-      "grad_norm": 4.995772838592529,
-      "learning_rate": 1.708483006355347e-05,
-      "loss": 0.5714,
+      "epoch": 4.591513741992772,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7079485464143918e-05,
+      "loss": 0.9671392059326172,
       "step": 398800
     },
     {
-      "epoch": 9.18762088974855,
-      "grad_norm": 3.2529380321502686,
-      "learning_rate": 1.7077558499730956e-05,
-      "loss": 0.5678,
+      "epoch": 4.5938164270674875,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7072213900321404e-05,
+      "loss": 0.9823818206787109,
       "step": 399000
     },
     {
-      "epoch": 9.19222621350281,
-      "grad_norm": 2.858065366744995,
-      "learning_rate": 1.7070286935908435e-05,
-      "loss": 0.5693,
+      "epoch": 4.596119112142202,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.706494233649889e-05,
+      "loss": 0.9794180297851562,
       "step": 399200
     },
     {
-      "epoch": 9.196831537257069,
-      "grad_norm": 2.9561386108398438,
-      "learning_rate": 1.706301537208592e-05,
-      "loss": 0.5748,
+      "epoch": 4.5984217972169175,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7057670772676372e-05,
+      "loss": 0.9763394927978516,
       "step": 399400
     },
     {
-      "epoch": 9.201436861011329,
-      "grad_norm": 3.378155469894409,
-      "learning_rate": 1.7055743808263407e-05,
-      "loss": 0.5608,
+      "epoch": 4.600724482291632,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7050399208853855e-05,
+      "loss": 0.9823768615722657,
       "step": 399600
     },
     {
-      "epoch": 9.20604218476559,
-      "grad_norm": 3.1857070922851562,
-      "learning_rate": 1.704847224444089e-05,
-      "loss": 0.5818,
+      "epoch": 4.603027167366347,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.704312764503134e-05,
+      "loss": 0.9815471649169922,
       "step": 399800
     },
     {
-      "epoch": 9.210647508519848,
-      "grad_norm": 4.287872791290283,
-      "learning_rate": 1.7041200680618375e-05,
-      "loss": 0.5751,
+      "epoch": 4.605329852441062,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.7035856081208823e-05,
+      "loss": 1.0154531860351563,
       "step": 400000
     },
     {
-      "epoch": 9.215252832274109,
-      "grad_norm": 3.0986175537109375,
-      "learning_rate": 1.703392911679586e-05,
-      "loss": 0.5718,
+      "epoch": 4.6076325375157765,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.702858451738631e-05,
+      "loss": 0.9649147796630859,
       "step": 400200
     },
     {
-      "epoch": 9.21985815602837,
-      "grad_norm": 3.245264768600464,
-      "learning_rate": 1.702665755297334e-05,
-      "loss": 0.5748,
+      "epoch": 4.609935222590492,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.7021312953563795e-05,
+      "loss": 0.9521435546875,
       "step": 400400
     },
     {
-      "epoch": 9.224463479782628,
-      "grad_norm": 3.5556859970092773,
-      "learning_rate": 1.7019385989150826e-05,
-      "loss": 0.5736,
+      "epoch": 4.612237907665207,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.701404138974128e-05,
+      "loss": 0.9710287475585937,
       "step": 400600
     },
     {
-      "epoch": 9.229068803536888,
-      "grad_norm": 3.7929844856262207,
-      "learning_rate": 1.7012114425328312e-05,
-      "loss": 0.5812,
+      "epoch": 4.614540592739922,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.700676982591876e-05,
+      "loss": 0.9978025817871093,
       "step": 400800
     },
     {
-      "epoch": 9.233674127291149,
-      "grad_norm": 3.4577155113220215,
-      "learning_rate": 1.7004879219324907e-05,
-      "loss": 0.5725,
+      "epoch": 4.616843277814636,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6999498262096246e-05,
+      "loss": 0.9523931121826172,
       "step": 401000
     },
     {
-      "epoch": 9.238279451045408,
-      "grad_norm": 3.2667465209960938,
-      "learning_rate": 1.6997607655502393e-05,
-      "loss": 0.5845,
+      "epoch": 4.619145962889352,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6992226698273732e-05,
+      "loss": 0.968920669555664,
       "step": 401200
     },
     {
-      "epoch": 9.242884774799668,
-      "grad_norm": 3.0420265197753906,
-      "learning_rate": 1.699033609167988e-05,
-      "loss": 0.5881,
+      "epoch": 4.621448647964066,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6984955134451215e-05,
+      "loss": 0.9653475952148437,
       "step": 401400
     },
     {
-      "epoch": 9.247490098553929,
-      "grad_norm": 2.9382212162017822,
-      "learning_rate": 1.698306452785736e-05,
-      "loss": 0.5904,
+      "epoch": 4.623751333038782,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.69776835706287e-05,
+      "loss": 1.0043842315673828,
       "step": 401600
     },
     {
-      "epoch": 9.252095422308189,
-      "grad_norm": 3.358121395111084,
-      "learning_rate": 1.6975792964034844e-05,
-      "loss": 0.5678,
+      "epoch": 4.626054018113496,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6970412006806186e-05,
+      "loss": 0.9865167236328125,
       "step": 401800
     },
     {
-      "epoch": 9.256700746062448,
-      "grad_norm": 6.984818458557129,
-      "learning_rate": 1.696852140021233e-05,
-      "loss": 0.5678,
+      "epoch": 4.628356703188212,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.696314044298367e-05,
+      "loss": 0.9743873596191406,
       "step": 402000
     },
     {
-      "epoch": 9.261306069816708,
-      "grad_norm": 2.786919355392456,
-      "learning_rate": 1.6961249836389812e-05,
-      "loss": 0.5801,
+      "epoch": 4.630659388262926,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.695586887916115e-05,
+      "loss": 0.972658920288086,
       "step": 402200
     },
     {
-      "epoch": 9.265911393570969,
-      "grad_norm": 2.3352646827697754,
-      "learning_rate": 1.69539782725673e-05,
-      "loss": 0.5722,
+      "epoch": 4.632962073337641,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6948597315338637e-05,
+      "loss": 0.9776019287109375,
       "step": 402400
     },
     {
-      "epoch": 9.270516717325227,
-      "grad_norm": 3.7076637744903564,
-      "learning_rate": 1.6946706708744784e-05,
-      "loss": 0.5773,
+      "epoch": 4.635264758412356,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.694132575151612e-05,
+      "loss": 1.0061843872070313,
       "step": 402600
     },
     {
-      "epoch": 9.275122041079488,
-      "grad_norm": 2.3307411670684814,
-      "learning_rate": 1.6939435144922267e-05,
-      "loss": 0.559,
+      "epoch": 4.637567443487071,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6934054187693606e-05,
+      "loss": 0.9649028015136719,
       "step": 402800
     },
     {
-      "epoch": 9.279727364833748,
-      "grad_norm": 3.8850185871124268,
-      "learning_rate": 1.6932163581099753e-05,
-      "loss": 0.5656,
+      "epoch": 4.639870128561786,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6926782623871092e-05,
+      "loss": 0.9808583068847656,
       "step": 403000
     },
     {
-      "epoch": 9.284332688588007,
-      "grad_norm": 3.246624231338501,
-      "learning_rate": 1.6924892017277235e-05,
-      "loss": 0.5893,
+      "epoch": 4.642172813636501,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6919511060048574e-05,
+      "loss": 0.9756576538085937,
       "step": 403200
     },
     {
-      "epoch": 9.288938012342268,
-      "grad_norm": 2.8703010082244873,
-      "learning_rate": 1.691762045345472e-05,
-      "loss": 0.5618,
+      "epoch": 4.644475498711216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6912239496226057e-05,
+      "loss": 0.9983255767822266,
       "step": 403400
     },
     {
-      "epoch": 9.293543336096528,
-      "grad_norm": 3.0698583126068115,
-      "learning_rate": 1.6910385247451316e-05,
-      "loss": 0.5576,
+      "epoch": 4.646778183785931,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6904967932403543e-05,
+      "loss": 0.964236068725586,
       "step": 403600
     },
     {
-      "epoch": 9.298148659850787,
-      "grad_norm": 2.5422704219818115,
-      "learning_rate": 1.6903113683628802e-05,
-      "loss": 0.5816,
+      "epoch": 4.649080868860646,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.689769636858103e-05,
+      "loss": 0.9836663818359375,
       "step": 403800
     },
     {
-      "epoch": 9.302753983605047,
-      "grad_norm": 3.1674082279205322,
-      "learning_rate": 1.6895842119806285e-05,
-      "loss": 0.586,
+      "epoch": 4.651383553935361,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.689042480475851e-05,
+      "loss": 0.9943917083740235,
       "step": 404000
     },
     {
-      "epoch": 9.307359307359308,
-      "grad_norm": 3.551118850708008,
-      "learning_rate": 1.688857055598377e-05,
-      "loss": 0.5837,
+      "epoch": 4.653686239010076,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6883153240935997e-05,
+      "loss": 0.988160400390625,
       "step": 404200
     },
     {
-      "epoch": 9.311964631113566,
-      "grad_norm": 2.9174249172210693,
-      "learning_rate": 1.6881298992161257e-05,
-      "loss": 0.5796,
+      "epoch": 4.655988924084791,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6875881677113483e-05,
+      "loss": 0.9667311096191407,
       "step": 404400
     },
     {
-      "epoch": 9.316569954867827,
-      "grad_norm": 3.165064573287964,
-      "learning_rate": 1.6874027428338736e-05,
-      "loss": 0.5811,
+      "epoch": 4.658291609159505,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6868610113290962e-05,
+      "loss": 0.9998202514648438,
       "step": 404600
     },
     {
-      "epoch": 9.321175278622087,
-      "grad_norm": 3.249115228652954,
-      "learning_rate": 1.6866755864516222e-05,
-      "loss": 0.5701,
+      "epoch": 4.660594294234221,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.6861338549468448e-05,
+      "loss": 0.9816936492919922,
       "step": 404800
     },
     {
-      "epoch": 9.325780602376348,
-      "grad_norm": 3.103433847427368,
-      "learning_rate": 1.6859484300693708e-05,
-      "loss": 0.5762,
+      "epoch": 4.662896979308935,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6854066985645934e-05,
+      "loss": 0.9878934478759765,
       "step": 405000
     },
     {
-      "epoch": 9.330385926130607,
-      "grad_norm": 3.612250328063965,
-      "learning_rate": 1.6852212736871194e-05,
-      "loss": 0.5697,
+      "epoch": 4.665199664383651,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.6846795421823416e-05,
+      "loss": 0.9956065368652344,
       "step": 405200
     },
     {
-      "epoch": 9.334991249884867,
-      "grad_norm": 4.119315147399902,
-      "learning_rate": 1.6844941173048676e-05,
-      "loss": 0.5787,
+      "epoch": 4.667502349458365,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6839523858000902e-05,
+      "loss": 0.9837393951416016,
       "step": 405400
     },
     {
-      "epoch": 9.339596573639128,
-      "grad_norm": 4.298466205596924,
-      "learning_rate": 1.6837669609226162e-05,
-      "loss": 0.563,
+      "epoch": 4.669805034533081,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6832252294178388e-05,
+      "loss": 0.9827626800537109,
       "step": 405600
     },
     {
-      "epoch": 9.344201897393386,
-      "grad_norm": 2.9533872604370117,
-      "learning_rate": 1.6830398045403648e-05,
-      "loss": 0.5874,
+      "epoch": 4.672107719607795,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.682498073035587e-05,
+      "loss": 0.9554264831542969,
       "step": 405800
     },
     {
-      "epoch": 9.348807221147647,
-      "grad_norm": 3.106743812561035,
-      "learning_rate": 1.6823126481581127e-05,
-      "loss": 0.5742,
+      "epoch": 4.67441040468251,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6817709166533353e-05,
+      "loss": 0.9837741851806641,
       "step": 406000
     },
     {
-      "epoch": 9.353412544901907,
-      "grad_norm": 2.491485118865967,
-      "learning_rate": 1.6815854917758613e-05,
-      "loss": 0.5726,
+      "epoch": 4.676713089757225,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.681043760271084e-05,
+      "loss": 0.9775701141357422,
       "step": 406200
     },
     {
-      "epoch": 9.358017868656166,
-      "grad_norm": 3.787768602371216,
-      "learning_rate": 1.680861971175521e-05,
-      "loss": 0.5682,
+      "epoch": 4.67901577483194,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6803166038888325e-05,
+      "loss": 0.9809413146972656,
       "step": 406400
     },
     {
-      "epoch": 9.362623192410426,
-      "grad_norm": 3.0979113578796387,
-      "learning_rate": 1.6801348147932694e-05,
-      "loss": 0.5698,
+      "epoch": 4.681318459906655,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.6795894475065808e-05,
+      "loss": 0.9923878479003906,
       "step": 406600
     },
     {
-      "epoch": 9.367228516164687,
-      "grad_norm": 2.7929654121398926,
-      "learning_rate": 1.679407658411018e-05,
-      "loss": 0.5644,
+      "epoch": 4.68362114498137,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6788622911243294e-05,
+      "loss": 0.9792401885986328,
       "step": 406800
     },
     {
-      "epoch": 9.371833839918946,
-      "grad_norm": 2.568842887878418,
-      "learning_rate": 1.6786805020287666e-05,
-      "loss": 0.5827,
+      "epoch": 4.685923830056085,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.678135134742078e-05,
+      "loss": 0.9748235321044922,
       "step": 407000
     },
     {
-      "epoch": 9.376439163673206,
-      "grad_norm": 3.327233076095581,
-      "learning_rate": 1.677953345646515e-05,
-      "loss": 0.5885,
+      "epoch": 4.6882265151308,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.677407978359826e-05,
+      "loss": 0.9866321563720704,
       "step": 407200
     },
     {
-      "epoch": 9.381044487427467,
-      "grad_norm": 2.4469523429870605,
-      "learning_rate": 1.677226189264263e-05,
-      "loss": 0.5692,
+      "epoch": 4.690529200205515,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6766808219775745e-05,
+      "loss": 1.0059503173828126,
       "step": 407400
     },
     {
-      "epoch": 9.385649811181725,
-      "grad_norm": 2.9710397720336914,
-      "learning_rate": 1.6764990328820117e-05,
-      "loss": 0.5651,
+      "epoch": 4.6928318852802295,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.675953665595323e-05,
+      "loss": 0.9742549896240235,
       "step": 407600
     },
     {
-      "epoch": 9.390255134935986,
-      "grad_norm": 3.184701442718506,
-      "learning_rate": 1.67577187649976e-05,
-      "loss": 0.5642,
+      "epoch": 4.695134570354945,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6752265092130713e-05,
+      "loss": 0.988902587890625,
       "step": 407800
     },
     {
-      "epoch": 9.394860458690246,
-      "grad_norm": 2.20974063873291,
-      "learning_rate": 1.6750447201175085e-05,
-      "loss": 0.5758,
+      "epoch": 4.6974372554296595,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.67449935283082e-05,
+      "loss": 1.0010816192626952,
       "step": 408000
     },
     {
-      "epoch": 9.399465782444507,
-      "grad_norm": 2.655507802963257,
-      "learning_rate": 1.674317563735257e-05,
-      "loss": 0.5758,
+      "epoch": 4.699739940504374,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6737721964485685e-05,
+      "loss": 0.9866725158691406,
       "step": 408200
     },
     {
-      "epoch": 9.404071106198765,
-      "grad_norm": 2.7525951862335205,
-      "learning_rate": 1.6735904073530054e-05,
-      "loss": 0.5656,
+      "epoch": 4.7020426255790895,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6730450400663167e-05,
+      "loss": 0.9774224090576172,
       "step": 408400
     },
     {
-      "epoch": 9.408676429953026,
-      "grad_norm": 3.151881694793701,
-      "learning_rate": 1.6728632509707536e-05,
-      "loss": 0.5701,
+      "epoch": 4.704345310653804,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.672317883684065e-05,
+      "loss": 0.9602320098876953,
       "step": 408600
     },
     {
-      "epoch": 9.413281753707286,
-      "grad_norm": 2.949288845062256,
-      "learning_rate": 1.6721360945885022e-05,
-      "loss": 0.5764,
+      "epoch": 4.706647995728519,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6715907273018136e-05,
+      "loss": 0.9782334899902344,
       "step": 408800
     },
     {
-      "epoch": 9.417887077461545,
-      "grad_norm": 3.4469354152679443,
-      "learning_rate": 1.6714089382062505e-05,
-      "loss": 0.577,
+      "epoch": 4.708950680803234,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6708635709195618e-05,
+      "loss": 0.9678709411621094,
       "step": 409000
     },
     {
-      "epoch": 9.422492401215806,
-      "grad_norm": 3.0155844688415527,
-      "learning_rate": 1.670681781823999e-05,
-      "loss": 0.5771,
+      "epoch": 4.711253365877949,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6701364145373104e-05,
+      "loss": 1.002803192138672,
       "step": 409200
     },
     {
-      "epoch": 9.427097724970066,
-      "grad_norm": 3.070678472518921,
-      "learning_rate": 1.669958261223659e-05,
-      "loss": 0.5845,
+      "epoch": 4.713556050952664,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.669409258155059e-05,
+      "loss": 0.9881256866455078,
       "step": 409400
     },
     {
-      "epoch": 9.431703048724325,
-      "grad_norm": 2.2283005714416504,
-      "learning_rate": 1.6692311048414072e-05,
-      "loss": 0.5636,
+      "epoch": 4.7158587360273785,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6686821017728076e-05,
+      "loss": 0.9696332550048828,
       "step": 409600
     },
     {
-      "epoch": 9.436308372478585,
-      "grad_norm": 2.9284727573394775,
-      "learning_rate": 1.6685039484591558e-05,
-      "loss": 0.581,
+      "epoch": 4.718161421102094,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6679549453905555e-05,
+      "loss": 0.976830825805664,
       "step": 409800
     },
     {
-      "epoch": 9.440913696232846,
-      "grad_norm": 3.039658308029175,
-      "learning_rate": 1.6677767920769044e-05,
-      "loss": 0.568,
+      "epoch": 4.7204641061768084,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.667227789008304e-05,
+      "loss": 1.0149959564208983,
       "step": 410000
     },
     {
-      "epoch": 9.445519019987104,
-      "grad_norm": 2.3291382789611816,
-      "learning_rate": 1.6670496356946523e-05,
-      "loss": 0.5628,
+      "epoch": 4.722766791251524,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6665006326260527e-05,
+      "loss": 1.0042466735839843,
       "step": 410200
     },
     {
-      "epoch": 9.450124343741365,
-      "grad_norm": 2.804224967956543,
-      "learning_rate": 1.666322479312401e-05,
-      "loss": 0.5728,
+      "epoch": 4.725069476326238,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.665773476243801e-05,
+      "loss": 0.9852333831787109,
       "step": 410400
     },
     {
-      "epoch": 9.454729667495625,
-      "grad_norm": 3.0607471466064453,
-      "learning_rate": 1.6655989587120607e-05,
-      "loss": 0.5655,
+      "epoch": 4.727372161400954,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6650463198615495e-05,
+      "loss": 0.9823546600341797,
       "step": 410600
     },
     {
-      "epoch": 9.459334991249884,
-      "grad_norm": 2.556657075881958,
-      "learning_rate": 1.664871802329809e-05,
-      "loss": 0.5723,
+      "epoch": 4.729674846475668,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.664319163479298e-05,
+      "loss": 0.9931096649169922,
       "step": 410800
     },
     {
-      "epoch": 9.463940315004145,
-      "grad_norm": 2.7533397674560547,
-      "learning_rate": 1.6641446459475576e-05,
-      "loss": 0.574,
+      "epoch": 4.731977531550384,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.663592007097046e-05,
+      "loss": 0.9812409210205079,
       "step": 411000
     },
     {
-      "epoch": 9.468545638758405,
-      "grad_norm": 2.0265374183654785,
-      "learning_rate": 1.663417489565306e-05,
-      "loss": 0.5751,
+      "epoch": 4.734280216625098,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6628648507147946e-05,
+      "loss": 0.9841607666015625,
       "step": 411200
     },
     {
-      "epoch": 9.473150962512666,
-      "grad_norm": 3.6705782413482666,
-      "learning_rate": 1.6626903331830544e-05,
-      "loss": 0.5787,
+      "epoch": 4.736582901699814,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6621376943325432e-05,
+      "loss": 0.9859600067138672,
       "step": 411400
     },
     {
-      "epoch": 9.477756286266924,
-      "grad_norm": 2.9727859497070312,
-      "learning_rate": 1.6619631768008027e-05,
-      "loss": 0.5658,
+      "epoch": 4.738885586774528,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6614105379502915e-05,
+      "loss": 0.9925346374511719,
       "step": 411600
     },
     {
-      "epoch": 9.482361610021185,
-      "grad_norm": 2.764256477355957,
-      "learning_rate": 1.6612360204185513e-05,
-      "loss": 0.5674,
+      "epoch": 4.741188271849243,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.66068338156804e-05,
+      "loss": 0.9583389282226562,
       "step": 411800
     },
     {
-      "epoch": 9.486966933775445,
-      "grad_norm": 4.042858123779297,
-      "learning_rate": 1.6605088640362995e-05,
-      "loss": 0.5726,
+      "epoch": 4.743490956923958,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6599562251857887e-05,
+      "loss": 0.9840799713134766,
       "step": 412000
     },
     {
-      "epoch": 9.491572257529704,
-      "grad_norm": 3.3755908012390137,
-      "learning_rate": 1.659781707654048e-05,
-      "loss": 0.5881,
+      "epoch": 4.745793641998673,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.659229068803537e-05,
+      "loss": 0.9803349304199219,
       "step": 412200
     },
     {
-      "epoch": 9.496177581283964,
-      "grad_norm": 3.193547487258911,
-      "learning_rate": 1.6590545512717967e-05,
-      "loss": 0.5649,
+      "epoch": 4.748096327073388,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6585019124212852e-05,
+      "loss": 0.9900364685058594,
       "step": 412400
     },
     {
-      "epoch": 9.500782905038225,
-      "grad_norm": 3.653513193130493,
-      "learning_rate": 1.658327394889545e-05,
-      "loss": 0.5807,
+      "epoch": 4.750399012148103,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6577747560390338e-05,
+      "loss": 0.9938496398925781,
       "step": 412600
     },
     {
-      "epoch": 9.505388228792484,
-      "grad_norm": 2.6081292629241943,
-      "learning_rate": 1.6576002385072935e-05,
-      "loss": 0.5756,
+      "epoch": 4.752701697222818,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6570475996567824e-05,
+      "loss": 0.971194076538086,
       "step": 412800
     },
     {
-      "epoch": 9.509993552546744,
-      "grad_norm": 3.2407031059265137,
-      "learning_rate": 1.6568730821250418e-05,
-      "loss": 0.5617,
+      "epoch": 4.755004382297533,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.6563204432745306e-05,
+      "loss": 0.9769140625,
       "step": 413000
     },
     {
-      "epoch": 9.514598876301005,
-      "grad_norm": 3.620607614517212,
-      "learning_rate": 1.6561459257427904e-05,
-      "loss": 0.5697,
+      "epoch": 4.757307067372247,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.6555932868922792e-05,
+      "loss": 0.9798918151855469,
       "step": 413200
     },
     {
-      "epoch": 9.519204200055263,
-      "grad_norm": 3.521552085876465,
-      "learning_rate": 1.6554187693605386e-05,
-      "loss": 0.5804,
+      "epoch": 4.759609752446963,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6548661305100278e-05,
+      "loss": 1.015190887451172,
       "step": 413400
     },
     {
-      "epoch": 9.523809523809524,
-      "grad_norm": 3.074605941772461,
-      "learning_rate": 1.6546916129782872e-05,
-      "loss": 0.5629,
+      "epoch": 4.761912437521677,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6541389741277757e-05,
+      "loss": 0.9922569274902344,
       "step": 413600
     },
     {
-      "epoch": 9.528414847563784,
-      "grad_norm": 3.518260955810547,
-      "learning_rate": 1.6539644565960358e-05,
-      "loss": 0.5772,
+      "epoch": 4.764215122596393,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6534118177455243e-05,
+      "loss": 0.9750289916992188,
       "step": 413800
     },
     {
-      "epoch": 9.533020171318043,
-      "grad_norm": 4.271399974822998,
-      "learning_rate": 1.653237300213784e-05,
-      "loss": 0.5713,
+      "epoch": 4.766517807671107,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.652684661363273e-05,
+      "loss": 0.9908672332763672,
       "step": 414000
     },
     {
-      "epoch": 9.537625495072303,
-      "grad_norm": 2.3440682888031006,
-      "learning_rate": 1.6525101438315323e-05,
-      "loss": 0.5715,
+      "epoch": 4.768820492745823,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.651957504981021e-05,
+      "loss": 0.982567138671875,
       "step": 414200
     },
     {
-      "epoch": 9.542230818826564,
-      "grad_norm": 3.0813117027282715,
-      "learning_rate": 1.651782987449281e-05,
-      "loss": 0.5761,
+      "epoch": 4.771123177820537,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6512303485987697e-05,
+      "loss": 1.0064954376220703,
       "step": 414400
     },
     {
-      "epoch": 9.546836142580823,
-      "grad_norm": 3.250023126602173,
-      "learning_rate": 1.651055831067029e-05,
-      "loss": 0.5737,
+      "epoch": 4.773425862895253,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6505031922165183e-05,
+      "loss": 0.9864546966552734,
       "step": 414600
     },
     {
-      "epoch": 9.551441466335083,
-      "grad_norm": 3.33170485496521,
-      "learning_rate": 1.6503286746847778e-05,
-      "loss": 0.5758,
+      "epoch": 4.775728547969967,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6497760358342662e-05,
+      "loss": 0.9798056030273438,
       "step": 414800
     },
     {
-      "epoch": 9.556046790089344,
-      "grad_norm": 3.526592254638672,
-      "learning_rate": 1.6496015183025263e-05,
-      "loss": 0.5632,
+      "epoch": 4.778031233044683,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.6490488794520148e-05,
+      "loss": 1.0037300872802735,
       "step": 415000
     },
     {
-      "epoch": 9.560652113843604,
-      "grad_norm": 3.5739214420318604,
-      "learning_rate": 1.6488743619202746e-05,
-      "loss": 0.5862,
+      "epoch": 4.780333918119397,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6483217230697634e-05,
+      "loss": 1.0045165252685546,
       "step": 415200
     },
     {
-      "epoch": 9.565257437597863,
-      "grad_norm": 3.0339114665985107,
-      "learning_rate": 1.648147205538023e-05,
-      "loss": 0.5733,
+      "epoch": 4.782636603194112,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.647594566687512e-05,
+      "loss": 0.9751959991455078,
       "step": 415400
     },
     {
-      "epoch": 9.569862761352123,
-      "grad_norm": 2.7391164302825928,
-      "learning_rate": 1.6474200491557714e-05,
-      "loss": 0.5671,
+      "epoch": 4.784939288268827,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.6468674103052603e-05,
+      "loss": 0.995721435546875,
       "step": 415600
     },
     {
-      "epoch": 9.574468085106384,
-      "grad_norm": 2.494318962097168,
-      "learning_rate": 1.64669289277352e-05,
-      "loss": 0.586,
+      "epoch": 4.787241973343542,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.646140253923009e-05,
+      "loss": 0.9940896606445313,
       "step": 415800
     },
     {
-      "epoch": 9.579073408860642,
-      "grad_norm": 2.9275710582733154,
-      "learning_rate": 1.6459657363912683e-05,
-      "loss": 0.5868,
+      "epoch": 4.789544658418257,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6454130975407574e-05,
+      "loss": 0.9893470764160156,
       "step": 416000
     },
     {
-      "epoch": 9.583678732614903,
-      "grad_norm": 2.980813980102539,
-      "learning_rate": 1.645238580009017e-05,
-      "loss": 0.5784,
+      "epoch": 4.791847343492972,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6446859411585054e-05,
+      "loss": 0.9823805236816406,
       "step": 416200
     },
     {
-      "epoch": 9.588284056369163,
-      "grad_norm": 2.879387617111206,
-      "learning_rate": 1.6445114236267655e-05,
-      "loss": 0.5534,
+      "epoch": 4.794150028567687,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.643958784776254e-05,
+      "loss": 0.9910964965820312,
       "step": 416400
     },
     {
-      "epoch": 9.592889380123422,
-      "grad_norm": 2.6000664234161377,
-      "learning_rate": 1.643787903026425e-05,
-      "loss": 0.5756,
+      "epoch": 4.7964527136424016,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.6432316283940025e-05,
+      "loss": 0.9669364929199219,
       "step": 416600
     },
     {
-      "epoch": 9.597494703877683,
-      "grad_norm": 2.941392660140991,
-      "learning_rate": 1.6430607466441736e-05,
-      "loss": 0.5787,
+      "epoch": 4.798755398717116,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6425044720117508e-05,
+      "loss": 0.9694680023193359,
       "step": 416800
     },
     {
-      "epoch": 9.602100027631943,
-      "grad_norm": 3.4506592750549316,
-      "learning_rate": 1.6423335902619215e-05,
-      "loss": 0.5694,
+      "epoch": 4.8010580837918315,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.6417773156294994e-05,
+      "loss": 0.993616943359375,
       "step": 417000
     },
     {
-      "epoch": 9.606705351386202,
-      "grad_norm": 3.014477014541626,
-      "learning_rate": 1.64160643387967e-05,
-      "loss": 0.5785,
+      "epoch": 4.803360768866546,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.641050159247248e-05,
+      "loss": 0.987774658203125,
       "step": 417200
     },
     {
-      "epoch": 9.611310675140462,
-      "grad_norm": 3.4871177673339844,
-      "learning_rate": 1.6408792774974187e-05,
-      "loss": 0.58,
+      "epoch": 4.8056634539412615,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.640323002864996e-05,
+      "loss": 0.9848440551757812,
       "step": 417400
     },
     {
-      "epoch": 9.615915998894723,
-      "grad_norm": 3.0921521186828613,
-      "learning_rate": 1.6401521211151673e-05,
-      "loss": 0.5571,
+      "epoch": 4.807966139015976,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6395958464827445e-05,
+      "loss": 0.9658927154541016,
       "step": 417600
     },
     {
-      "epoch": 9.620521322648981,
-      "grad_norm": 3.4543838500976562,
-      "learning_rate": 1.6394249647329155e-05,
-      "loss": 0.5689,
+      "epoch": 4.8102688240906915,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.638868690100493e-05,
+      "loss": 0.9894432830810547,
       "step": 417800
     },
     {
-      "epoch": 9.625126646403242,
-      "grad_norm": 3.799140453338623,
-      "learning_rate": 1.638697808350664e-05,
-      "loss": 0.5898,
+      "epoch": 4.812571509165406,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.6381415337182413e-05,
+      "loss": 0.9847721099853516,
       "step": 418000
     },
     {
-      "epoch": 9.629731970157502,
-      "grad_norm": 2.701723575592041,
-      "learning_rate": 1.6379706519684124e-05,
-      "loss": 0.584,
+      "epoch": 4.814874194240121,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.63741437733599e-05,
+      "loss": 0.9953361511230469,
       "step": 418200
     },
     {
-      "epoch": 9.634337293911763,
-      "grad_norm": 3.6430578231811523,
-      "learning_rate": 1.6372434955861606e-05,
-      "loss": 0.5649,
+      "epoch": 4.817176879314836,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6366872209537385e-05,
+      "loss": 0.9875911712646485,
       "step": 418400
     },
     {
-      "epoch": 9.638942617666022,
-      "grad_norm": 3.381641387939453,
-      "learning_rate": 1.6365163392039092e-05,
-      "loss": 0.5755,
+      "epoch": 4.819479564389551,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.635960064571487e-05,
+      "loss": 0.9725639343261718,
       "step": 418600
     },
     {
-      "epoch": 9.643547941420282,
-      "grad_norm": 3.1972997188568115,
-      "learning_rate": 1.6357891828216578e-05,
-      "loss": 0.5767,
+      "epoch": 4.821782249464266,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.635232908189235e-05,
+      "loss": 0.9967562866210937,
       "step": 418800
     },
     {
-      "epoch": 9.648153265174543,
-      "grad_norm": 2.762204647064209,
-      "learning_rate": 1.635062026439406e-05,
-      "loss": 0.5659,
+      "epoch": 4.8240849345389805,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.6345057518069836e-05,
+      "loss": 0.9764330291748047,
       "step": 419000
     },
     {
-      "epoch": 9.652758588928801,
-      "grad_norm": 2.9068100452423096,
-      "learning_rate": 1.6343348700571546e-05,
-      "loss": 0.5822,
+      "epoch": 4.826387619613696,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.6337785954247322e-05,
+      "loss": 0.9936328125,
       "step": 419200
     },
     {
-      "epoch": 9.657363912683062,
-      "grad_norm": 3.140155553817749,
-      "learning_rate": 1.633607713674903e-05,
-      "loss": 0.5617,
+      "epoch": 4.82869030468841,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6330514390424804e-05,
+      "loss": 0.9871342468261719,
       "step": 419400
     },
     {
-      "epoch": 9.661969236437322,
-      "grad_norm": 3.393486738204956,
-      "learning_rate": 1.632880557292651e-05,
-      "loss": 0.59,
+      "epoch": 4.830992989763126,
+      "grad_norm": 0.75,
+      "learning_rate": 1.632324282660229e-05,
+      "loss": 0.9947419738769532,
       "step": 419600
     },
     {
-      "epoch": 9.666574560191581,
-      "grad_norm": 3.3374438285827637,
-      "learning_rate": 1.6321534009103997e-05,
-      "loss": 0.5863,
+      "epoch": 4.83329567483784,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6315971262779776e-05,
+      "loss": 0.9876214599609375,
       "step": 419800
     },
     {
-      "epoch": 9.671179883945841,
-      "grad_norm": 2.8088021278381348,
-      "learning_rate": 1.6314262445281483e-05,
-      "loss": 0.5748,
+      "epoch": 4.835598359912556,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.6308699698957255e-05,
+      "loss": 0.9614956665039063,
       "step": 420000
     },
     {
-      "epoch": 9.675785207700102,
-      "grad_norm": 3.9555068016052246,
-      "learning_rate": 1.630702723927808e-05,
-      "loss": 0.5809,
+      "epoch": 4.83790104498727,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.630142813513474e-05,
+      "loss": 0.981507797241211,
       "step": 420200
     },
     {
-      "epoch": 9.68039053145436,
-      "grad_norm": 3.5197412967681885,
-      "learning_rate": 1.6299755675455564e-05,
-      "loss": 0.5796,
+      "epoch": 4.840203730061986,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6294156571312227e-05,
+      "loss": 0.9904711151123047,
       "step": 420400
     },
     {
-      "epoch": 9.684995855208621,
-      "grad_norm": 2.782156229019165,
-      "learning_rate": 1.629248411163305e-05,
-      "loss": 0.5738,
+      "epoch": 4.8425064151367,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.628688500748971e-05,
+      "loss": 0.981631851196289,
       "step": 420600
     },
     {
-      "epoch": 9.689601178962882,
-      "grad_norm": 3.482156276702881,
-      "learning_rate": 1.6285212547810533e-05,
-      "loss": 0.5766,
+      "epoch": 4.844809100211416,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6279613443667196e-05,
+      "loss": 0.9610891723632813,
       "step": 420800
     },
     {
-      "epoch": 9.69420650271714,
-      "grad_norm": 2.7525062561035156,
-      "learning_rate": 1.6277940983988015e-05,
-      "loss": 0.5774,
+      "epoch": 4.84711178528613,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.627234187984468e-05,
+      "loss": 1.0071526336669923,
       "step": 421000
     },
     {
-      "epoch": 9.6988118264714,
-      "grad_norm": 3.4421098232269287,
-      "learning_rate": 1.62706694201655e-05,
-      "loss": 0.5709,
+      "epoch": 4.849414470360845,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.6265070316022164e-05,
+      "loss": 0.9685559844970704,
       "step": 421200
     },
     {
-      "epoch": 9.703417150225661,
-      "grad_norm": 2.6279489994049072,
-      "learning_rate": 1.6263397856342984e-05,
-      "loss": 0.5647,
+      "epoch": 4.85171715543556,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6257798752199647e-05,
+      "loss": 0.9827532196044921,
       "step": 421400
     },
     {
-      "epoch": 9.70802247397992,
-      "grad_norm": 2.9035749435424805,
-      "learning_rate": 1.625612629252047e-05,
-      "loss": 0.5715,
+      "epoch": 4.854019840510275,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6250527188377133e-05,
+      "loss": 0.9905502319335937,
       "step": 421600
     },
     {
-      "epoch": 9.71262779773418,
-      "grad_norm": 3.324402093887329,
-      "learning_rate": 1.6248854728697956e-05,
-      "loss": 0.5828,
+      "epoch": 4.85632252558499,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.624325562455462e-05,
+      "loss": 0.9697259521484375,
       "step": 421800
     },
     {
-      "epoch": 9.71723312148844,
-      "grad_norm": 3.438358783721924,
-      "learning_rate": 1.6241583164875438e-05,
-      "loss": 0.5718,
+      "epoch": 4.858625210659705,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.62359840607321e-05,
+      "loss": 0.9974535369873047,
       "step": 422000
     },
     {
-      "epoch": 9.721838445242701,
-      "grad_norm": 3.8631820678710938,
-      "learning_rate": 1.623431160105292e-05,
-      "loss": 0.5787,
+      "epoch": 4.86092789573442,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6228712496909587e-05,
+      "loss": 0.9746042633056641,
       "step": 422200
     },
     {
-      "epoch": 9.72644376899696,
-      "grad_norm": 3.755342483520508,
-      "learning_rate": 1.6227040037230407e-05,
-      "loss": 0.5799,
+      "epoch": 4.863230580809135,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6221440933087073e-05,
+      "loss": 0.9784107971191406,
       "step": 422400
     },
     {
-      "epoch": 9.73104909275122,
-      "grad_norm": 4.419692039489746,
-      "learning_rate": 1.6219768473407893e-05,
-      "loss": 0.572,
+      "epoch": 4.865533265883849,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6214169369264552e-05,
+      "loss": 0.9914553833007812,
       "step": 422600
     },
     {
-      "epoch": 9.735654416505481,
-      "grad_norm": 3.5479319095611572,
-      "learning_rate": 1.6212496909585375e-05,
-      "loss": 0.5864,
+      "epoch": 4.867835950958565,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6206897805442038e-05,
+      "loss": 0.9690621948242187,
       "step": 422800
     },
     {
-      "epoch": 9.74025974025974,
-      "grad_norm": 2.9561350345611572,
-      "learning_rate": 1.620522534576286e-05,
-      "loss": 0.5806,
+      "epoch": 4.870138636033279,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6199626241619524e-05,
+      "loss": 0.9820099639892578,
       "step": 423000
     },
     {
-      "epoch": 9.744865064014,
-      "grad_norm": 2.8360917568206787,
-      "learning_rate": 1.6197953781940347e-05,
-      "loss": 0.5824,
+      "epoch": 4.872441321107995,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6192354677797006e-05,
+      "loss": 0.9805403137207032,
       "step": 423200
     },
     {
-      "epoch": 9.74947038776826,
-      "grad_norm": 3.090214967727661,
-      "learning_rate": 1.619068221811783e-05,
-      "loss": 0.5732,
+      "epoch": 4.874744006182709,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.6185083113974492e-05,
+      "loss": 0.9757598876953125,
       "step": 423400
     },
     {
-      "epoch": 9.75407571152252,
-      "grad_norm": 3.477980613708496,
-      "learning_rate": 1.6183447012114428e-05,
-      "loss": 0.5688,
+      "epoch": 4.877046691257425,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6177811550151978e-05,
+      "loss": 0.9821089172363281,
       "step": 423600
     },
     {
-      "epoch": 9.75868103527678,
-      "grad_norm": 2.7006947994232178,
-      "learning_rate": 1.617617544829191e-05,
-      "loss": 0.5744,
+      "epoch": 4.879349376332139,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6170539986329457e-05,
+      "loss": 0.9766545867919922,
       "step": 423800
     },
     {
-      "epoch": 9.76328635903104,
-      "grad_norm": 3.080148696899414,
-      "learning_rate": 1.6168903884469393e-05,
-      "loss": 0.5751,
+      "epoch": 4.881652061406855,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6163268422506943e-05,
+      "loss": 0.9831093597412109,
       "step": 424000
     },
     {
-      "epoch": 9.767891682785299,
-      "grad_norm": 3.5952165126800537,
-      "learning_rate": 1.616163232064688e-05,
-      "loss": 0.5706,
+      "epoch": 4.883954746481569,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.615599685868443e-05,
+      "loss": 0.9682130432128906,
       "step": 424200
     },
     {
-      "epoch": 9.77249700653956,
-      "grad_norm": 3.141179323196411,
-      "learning_rate": 1.6154360756824365e-05,
-      "loss": 0.5758,
+      "epoch": 4.886257431556285,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6148725294861915e-05,
+      "loss": 0.9696171569824219,
       "step": 424400
     },
     {
-      "epoch": 9.77710233029382,
-      "grad_norm": 2.961484909057617,
-      "learning_rate": 1.6147089193001847e-05,
-      "loss": 0.5813,
+      "epoch": 4.888560116630999,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.6141453731039398e-05,
+      "loss": 0.9785321807861328,
       "step": 424600
     },
     {
-      "epoch": 9.78170765404808,
-      "grad_norm": 3.8218839168548584,
-      "learning_rate": 1.6139817629179333e-05,
-      "loss": 0.5677,
+      "epoch": 4.890862801705714,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6134182167216883e-05,
+      "loss": 0.9684022521972656,
       "step": 424800
     },
     {
-      "epoch": 9.78631297780234,
-      "grad_norm": 2.3090195655822754,
-      "learning_rate": 1.6132546065356816e-05,
-      "loss": 0.5806,
+      "epoch": 4.893165486780429,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6126910603394366e-05,
+      "loss": 0.9904294586181641,
       "step": 425000
     },
     {
-      "epoch": 9.7909183015566,
-      "grad_norm": 2.9039793014526367,
-      "learning_rate": 1.61252745015343e-05,
-      "loss": 0.5748,
+      "epoch": 4.895468171855144,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.611963903957185e-05,
+      "loss": 0.9782858276367188,
       "step": 425200
     },
     {
-      "epoch": 9.79552362531086,
-      "grad_norm": 3.138223171234131,
-      "learning_rate": 1.6118002937711784e-05,
-      "loss": 0.5711,
+      "epoch": 4.897770856929859,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6112367475749334e-05,
+      "loss": 0.995420913696289,
       "step": 425400
     },
     {
-      "epoch": 9.800128949065119,
-      "grad_norm": 3.0506985187530518,
-      "learning_rate": 1.611073137388927e-05,
-      "loss": 0.5705,
+      "epoch": 4.900073542004574,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.610509591192682e-05,
+      "loss": 0.9884532928466797,
       "step": 425600
     },
     {
-      "epoch": 9.80473427281938,
-      "grad_norm": 2.8752284049987793,
-      "learning_rate": 1.6103496167885865e-05,
-      "loss": 0.5763,
+      "epoch": 4.902376227079289,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6097824348104303e-05,
+      "loss": 0.9923461151123046,
       "step": 425800
     },
     {
-      "epoch": 9.80933959657364,
-      "grad_norm": 3.2525858879089355,
-      "learning_rate": 1.609622460406335e-05,
-      "loss": 0.5703,
+      "epoch": 4.9046789121540035,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.609055278428179e-05,
+      "loss": 0.9905690002441406,
       "step": 426000
     },
     {
-      "epoch": 9.813944920327899,
-      "grad_norm": 3.1471686363220215,
-      "learning_rate": 1.6088953040240837e-05,
-      "loss": 0.5665,
+      "epoch": 4.906981597228718,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6083281220459275e-05,
+      "loss": 1.0047775268554688,
       "step": 426200
     },
     {
-      "epoch": 9.818550244082159,
-      "grad_norm": 2.207984685897827,
-      "learning_rate": 1.6081681476418316e-05,
-      "loss": 0.5853,
+      "epoch": 4.9092842823034335,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6076009656636754e-05,
+      "loss": 0.9881675720214844,
       "step": 426400
     },
     {
-      "epoch": 9.82315556783642,
-      "grad_norm": 3.0319111347198486,
-      "learning_rate": 1.6074409912595802e-05,
-      "loss": 0.5681,
+      "epoch": 4.911586967378148,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.606873809281424e-05,
+      "loss": 0.9722449493408203,
       "step": 426600
     },
     {
-      "epoch": 9.827760891590678,
-      "grad_norm": 2.59087872505188,
-      "learning_rate": 1.6067138348773288e-05,
-      "loss": 0.5699,
+      "epoch": 4.9138896524528635,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6061466528991726e-05,
+      "loss": 0.9992510223388672,
       "step": 426800
     },
     {
-      "epoch": 9.832366215344939,
-      "grad_norm": 2.8990542888641357,
-      "learning_rate": 1.605986678495077e-05,
-      "loss": 0.5683,
+      "epoch": 4.916192337527578,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.605419496516921e-05,
+      "loss": 0.9681453704833984,
       "step": 427000
     },
     {
-      "epoch": 9.8369715390992,
-      "grad_norm": 2.702047109603882,
-      "learning_rate": 1.6052595221128257e-05,
-      "loss": 0.5793,
+      "epoch": 4.918495022602293,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6046923401346694e-05,
+      "loss": 0.9711811828613282,
       "step": 427200
     },
     {
-      "epoch": 9.841576862853458,
-      "grad_norm": 3.4502487182617188,
-      "learning_rate": 1.6045323657305743e-05,
-      "loss": 0.5725,
+      "epoch": 4.920797707677008,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.603965183752418e-05,
+      "loss": 0.9840038299560547,
       "step": 427400
     },
     {
-      "epoch": 9.846182186607718,
-      "grad_norm": 2.59299635887146,
-      "learning_rate": 1.6038052093483225e-05,
-      "loss": 0.5812,
+      "epoch": 4.923100392751723,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.6032380273701663e-05,
+      "loss": 0.9930586242675781,
       "step": 427600
     },
     {
-      "epoch": 9.850787510361979,
-      "grad_norm": 3.2523722648620605,
-      "learning_rate": 1.6030780529660708e-05,
-      "loss": 0.5743,
+      "epoch": 4.925403077826438,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6025108709879145e-05,
+      "loss": 0.9876455688476562,
       "step": 427800
     },
     {
-      "epoch": 9.855392834116238,
-      "grad_norm": 2.8594510555267334,
-      "learning_rate": 1.6023508965838194e-05,
-      "loss": 0.5623,
+      "epoch": 4.927705762901153,
+      "grad_norm": 0.5,
+      "learning_rate": 1.601783714605663e-05,
+      "loss": 0.9847328186035156,
       "step": 428000
     },
     {
-      "epoch": 9.859998157870498,
-      "grad_norm": 3.209397792816162,
-      "learning_rate": 1.6016237402015676e-05,
-      "loss": 0.5766,
+      "epoch": 4.930008447975868,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6010565582234117e-05,
+      "loss": 0.980836181640625,
       "step": 428200
     },
     {
-      "epoch": 9.864603481624759,
-      "grad_norm": 3.1015477180480957,
-      "learning_rate": 1.6008965838193162e-05,
-      "loss": 0.5793,
+      "epoch": 4.932311133050582,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.60032940184116e-05,
+      "loss": 0.9755804443359375,
       "step": 428400
     },
     {
-      "epoch": 9.869208805379017,
-      "grad_norm": 2.951237440109253,
-      "learning_rate": 1.600173063218976e-05,
-      "loss": 0.5745,
+      "epoch": 4.934613818125298,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5996022454589085e-05,
+      "loss": 0.9845207214355469,
       "step": 428600
     },
     {
-      "epoch": 9.873814129133278,
-      "grad_norm": 3.060861825942993,
-      "learning_rate": 1.5994459068367243e-05,
-      "loss": 0.58,
+      "epoch": 4.936916503200012,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.598875089076657e-05,
+      "loss": 0.9741409301757813,
       "step": 428800
     },
     {
-      "epoch": 9.878419452887538,
-      "grad_norm": 2.895648956298828,
-      "learning_rate": 1.598718750454473e-05,
-      "loss": 0.5711,
+      "epoch": 4.939219188274728,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.598147932694405e-05,
+      "loss": 0.9703875732421875,
       "step": 429000
     },
     {
-      "epoch": 9.883024776641799,
-      "grad_norm": 3.3986594676971436,
-      "learning_rate": 1.597991594072221e-05,
-      "loss": 0.5773,
+      "epoch": 4.941521873349442,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5974207763121536e-05,
+      "loss": 0.9845467376708984,
       "step": 429200
     },
     {
-      "epoch": 9.887630100396057,
-      "grad_norm": 3.4484481811523438,
-      "learning_rate": 1.5972644376899694e-05,
-      "loss": 0.5774,
+      "epoch": 4.943824558424158,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5966936199299022e-05,
+      "loss": 0.9741230010986328,
       "step": 429400
     },
     {
-      "epoch": 9.892235424150318,
-      "grad_norm": 3.4718239307403564,
-      "learning_rate": 1.596537281307718e-05,
-      "loss": 0.577,
+      "epoch": 4.946127243498872,
+      "grad_norm": 0.5,
+      "learning_rate": 1.5959664635476505e-05,
+      "loss": 0.9744194793701172,
       "step": 429600
     },
     {
-      "epoch": 9.896840747904578,
-      "grad_norm": 2.872313976287842,
-      "learning_rate": 1.5958101249254666e-05,
-      "loss": 0.5752,
+      "epoch": 4.948429928573587,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.595239307165399e-05,
+      "loss": 0.9622555541992187,
       "step": 429800
     },
     {
-      "epoch": 9.901446071658837,
-      "grad_norm": 3.031860589981079,
-      "learning_rate": 1.595082968543215e-05,
-      "loss": 0.5822,
+      "epoch": 4.950732613648302,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5945121507831477e-05,
+      "loss": 0.9699842071533203,
       "step": 430000
     },
     {
-      "epoch": 9.906051395413098,
-      "grad_norm": 3.6577560901641846,
-      "learning_rate": 1.5943558121609634e-05,
-      "loss": 0.5665,
+      "epoch": 4.953035298723017,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.593784994400896e-05,
+      "loss": 0.9580533599853516,
       "step": 430200
     },
     {
-      "epoch": 9.910656719167358,
-      "grad_norm": 2.5794246196746826,
-      "learning_rate": 1.593628655778712e-05,
-      "loss": 0.584,
+      "epoch": 4.955337983797732,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.593057838018644e-05,
+      "loss": 0.9905766296386719,
       "step": 430400
     },
     {
-      "epoch": 9.915262042921617,
-      "grad_norm": 3.3950586318969727,
-      "learning_rate": 1.5929014993964603e-05,
-      "loss": 0.5548,
+      "epoch": 4.957640668872447,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5923306816363928e-05,
+      "loss": 0.977527847290039,
       "step": 430600
     },
     {
-      "epoch": 9.919867366675877,
-      "grad_norm": 2.942765474319458,
-      "learning_rate": 1.5921743430142085e-05,
-      "loss": 0.5763,
+      "epoch": 4.959943353947162,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5916035252541413e-05,
+      "loss": 0.9803998565673828,
       "step": 430800
     },
     {
-      "epoch": 9.924472690430138,
-      "grad_norm": 2.925250291824341,
-      "learning_rate": 1.591447186631957e-05,
-      "loss": 0.5572,
+      "epoch": 4.962246039021877,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5908763688718896e-05,
+      "loss": 1.0080958557128907,
       "step": 431000
     },
     {
-      "epoch": 9.929078014184396,
-      "grad_norm": 2.779973030090332,
-      "learning_rate": 1.5907200302497057e-05,
-      "loss": 0.5651,
+      "epoch": 4.964548724096592,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5901492124896382e-05,
+      "loss": 0.9899720001220703,
       "step": 431200
     },
     {
-      "epoch": 9.933683337938657,
-      "grad_norm": 2.9650769233703613,
-      "learning_rate": 1.589992873867454e-05,
-      "loss": 0.5765,
+      "epoch": 4.966851409171307,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5894220561073864e-05,
+      "loss": 0.96768798828125,
       "step": 431400
     },
     {
-      "epoch": 9.938288661692917,
-      "grad_norm": 2.562681198120117,
-      "learning_rate": 1.5892657174852026e-05,
-      "loss": 0.5653,
+      "epoch": 4.969154094246022,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.5886948997251347e-05,
+      "loss": 0.9832350158691406,
       "step": 431600
     },
     {
-      "epoch": 9.942893985447178,
-      "grad_norm": 3.5627505779266357,
-      "learning_rate": 1.5885385611029508e-05,
-      "loss": 0.5617,
+      "epoch": 4.971456779320737,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.5879677433428833e-05,
+      "loss": 0.973259506225586,
       "step": 431800
     },
     {
-      "epoch": 9.947499309201437,
-      "grad_norm": 3.4282848834991455,
-      "learning_rate": 1.5878150405026103e-05,
-      "loss": 0.5673,
+      "epoch": 4.973759464395451,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.587240586960632e-05,
+      "loss": 0.9868125915527344,
       "step": 432000
     },
     {
-      "epoch": 9.952104632955697,
-      "grad_norm": 3.556774139404297,
-      "learning_rate": 1.587087884120359e-05,
-      "loss": 0.5806,
+      "epoch": 4.976062149470167,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.58651343057838e-05,
+      "loss": 1.0016079711914063,
       "step": 432200
     },
     {
-      "epoch": 9.956709956709958,
-      "grad_norm": 3.773305654525757,
-      "learning_rate": 1.5863607277381075e-05,
-      "loss": 0.5731,
+      "epoch": 4.978364834544881,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.5857862741961287e-05,
+      "loss": 0.997337875366211,
       "step": 432400
     },
     {
-      "epoch": 9.961315280464216,
-      "grad_norm": 2.997891902923584,
-      "learning_rate": 1.5856335713558558e-05,
-      "loss": 0.5713,
+      "epoch": 4.980667519619597,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.5850591178138773e-05,
+      "loss": 0.9590795135498047,
       "step": 432600
     },
     {
-      "epoch": 9.965920604218477,
-      "grad_norm": 3.1296310424804688,
-      "learning_rate": 1.5849064149736044e-05,
-      "loss": 0.5618,
+      "epoch": 4.982970204694311,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5843319614316256e-05,
+      "loss": 0.9921549224853515,
       "step": 432800
     },
     {
-      "epoch": 9.970525927972737,
-      "grad_norm": 3.2844841480255127,
-      "learning_rate": 1.584179258591353e-05,
-      "loss": 0.5774,
+      "epoch": 4.985272889769027,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.5836048050493738e-05,
+      "loss": 0.9745076751708984,
       "step": 433000
     },
     {
-      "epoch": 9.975131251726996,
-      "grad_norm": 2.3057990074157715,
-      "learning_rate": 1.583452102209101e-05,
-      "loss": 0.5783,
+      "epoch": 4.987575574843741,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5828776486671224e-05,
+      "loss": 0.972037582397461,
       "step": 433200
     },
     {
-      "epoch": 9.979736575481256,
-      "grad_norm": 2.7492430210113525,
-      "learning_rate": 1.5827249458268495e-05,
-      "loss": 0.5757,
+      "epoch": 4.989878259918456,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.582150492284871e-05,
+      "loss": 0.973899154663086,
       "step": 433400
     },
     {
-      "epoch": 9.984341899235517,
-      "grad_norm": 2.937319278717041,
-      "learning_rate": 1.581997789444598e-05,
-      "loss": 0.5814,
+      "epoch": 4.992180944993171,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5814233359026193e-05,
+      "loss": 0.9822638702392578,
       "step": 433600
     },
     {
-      "epoch": 9.988947222989776,
-      "grad_norm": 2.906705141067505,
-      "learning_rate": 1.5812706330623463e-05,
-      "loss": 0.5783,
+      "epoch": 4.994483630067886,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.580696179520368e-05,
+      "loss": 0.9719387054443359,
       "step": 433800
     },
     {
-      "epoch": 9.993552546744036,
-      "grad_norm": 2.5989561080932617,
-      "learning_rate": 1.580543476680095e-05,
-      "loss": 0.5865,
+      "epoch": 4.996786315142601,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.579969023138116e-05,
+      "loss": 0.9759081268310547,
       "step": 434000
     },
     {
-      "epoch": 9.998157870498297,
-      "grad_norm": 3.9318482875823975,
-      "learning_rate": 1.5798163202978435e-05,
-      "loss": 0.5763,
+      "epoch": 4.999089000217316,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5792418667558643e-05,
+      "loss": 0.9978753662109375,
       "step": 434200
     },
     {
-      "epoch": 10.0,
-      "eval_loss": 0.5503791570663452,
-      "eval_runtime": 162.7661,
-      "eval_samples_per_second": 174.244,
-      "eval_steps_per_second": 10.893,
+      "epoch": 5.0,
+      "eval_loss": NaN,
+      "eval_runtime": 432.1192,
+      "eval_samples_per_second": 65.632,
+      "eval_steps_per_second": 32.817,
       "step": 434280
     },
     {
-      "epoch": 10.002763194252555,
-      "grad_norm": 3.2857744693756104,
-      "learning_rate": 1.579092799697503e-05,
-      "loss": 0.584,
+      "epoch": 5.001381611044829,
+      "grad_norm": 0.625,
+      "learning_rate": 1.578514710373613e-05,
+      "loss": 0.9861835479736328,
       "step": 434400
     },
     {
-      "epoch": 10.007368518006816,
-      "grad_norm": 2.8815836906433105,
-      "learning_rate": 1.578369279097163e-05,
-      "loss": 0.5577,
+      "epoch": 5.003684296119544,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5777875539913615e-05,
+      "loss": 0.9894441986083984,
       "step": 434600
     },
     {
-      "epoch": 10.011973841761076,
-      "grad_norm": 2.8259472846984863,
-      "learning_rate": 1.577642122714911e-05,
-      "loss": 0.5706,
+      "epoch": 5.005986981194259,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5770603976091098e-05,
+      "loss": 0.9802412414550781,
       "step": 434800
     },
     {
-      "epoch": 10.016579165515335,
-      "grad_norm": 2.6986682415008545,
-      "learning_rate": 1.5769149663326594e-05,
-      "loss": 0.5662,
+      "epoch": 5.008289666268974,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5763332412268584e-05,
+      "loss": 0.9692536163330078,
       "step": 435000
     },
     {
-      "epoch": 10.021184489269595,
-      "grad_norm": 3.565535545349121,
-      "learning_rate": 1.576187809950408e-05,
-      "loss": 0.5678,
+      "epoch": 5.010592351343688,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.575606084844607e-05,
+      "loss": 0.9791400909423829,
       "step": 435200
     },
     {
-      "epoch": 10.025789813023856,
-      "grad_norm": 2.950361967086792,
-      "learning_rate": 1.5754606535681565e-05,
-      "loss": 0.5743,
+      "epoch": 5.012895036418404,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.574878928462355e-05,
+      "loss": 0.98474609375,
       "step": 435400
     },
     {
-      "epoch": 10.030395136778116,
-      "grad_norm": 2.67488694190979,
-      "learning_rate": 1.5747334971859048e-05,
-      "loss": 0.5798,
+      "epoch": 5.015197721493118,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5741517720801035e-05,
+      "loss": 0.9832347869873047,
       "step": 435600
     },
     {
-      "epoch": 10.035000460532375,
-      "grad_norm": 2.76383113861084,
-      "learning_rate": 1.5740063408036534e-05,
-      "loss": 0.5616,
+      "epoch": 5.017500406567834,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.573424615697852e-05,
+      "loss": 0.983828353881836,
       "step": 435800
     },
     {
-      "epoch": 10.039605784286636,
-      "grad_norm": 2.7756881713867188,
-      "learning_rate": 1.573279184421402e-05,
-      "loss": 0.582,
+      "epoch": 5.019803091642548,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5726974593156007e-05,
+      "loss": 0.9714756011962891,
       "step": 436000
     },
     {
-      "epoch": 10.044211108040896,
-      "grad_norm": 3.5233657360076904,
-      "learning_rate": 1.57255202803915e-05,
-      "loss": 0.5697,
+      "epoch": 5.022105776717264,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.571970302933349e-05,
+      "loss": 0.9881432342529297,
       "step": 436200
     },
     {
-      "epoch": 10.048816431795155,
-      "grad_norm": 2.4349958896636963,
-      "learning_rate": 1.5718248716568985e-05,
-      "loss": 0.565,
+      "epoch": 5.024408461791978,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5712431465510975e-05,
+      "loss": 0.9834969329833985,
       "step": 436400
     },
     {
-      "epoch": 10.053421755549415,
-      "grad_norm": 2.8675363063812256,
-      "learning_rate": 1.571097715274647e-05,
-      "loss": 0.5669,
+      "epoch": 5.026711146866694,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5705159901688457e-05,
+      "loss": 0.9566364288330078,
       "step": 436600
     },
     {
-      "epoch": 10.058027079303676,
-      "grad_norm": 2.97993540763855,
-      "learning_rate": 1.5703705588923953e-05,
-      "loss": 0.5759,
+      "epoch": 5.029013831941408,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.569788833786594e-05,
+      "loss": 0.9873866271972657,
       "step": 436800
     },
     {
-      "epoch": 10.062632403057934,
-      "grad_norm": 3.144256353378296,
-      "learning_rate": 1.569643402510144e-05,
-      "loss": 0.5606,
+      "epoch": 5.031316517016123,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5690616774043426e-05,
+      "loss": 0.9714003753662109,
       "step": 437000
     },
     {
-      "epoch": 10.067237726812195,
-      "grad_norm": 2.9724130630493164,
-      "learning_rate": 1.5689162461278925e-05,
-      "loss": 0.5811,
+      "epoch": 5.033619202090838,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5683345210220912e-05,
+      "loss": 0.9779801940917969,
       "step": 437200
     },
     {
-      "epoch": 10.071843050566455,
-      "grad_norm": 3.0126194953918457,
-      "learning_rate": 1.5681890897456408e-05,
-      "loss": 0.5646,
+      "epoch": 5.035921887165553,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5676073646398394e-05,
+      "loss": 0.9925363159179688,
       "step": 437400
     },
     {
-      "epoch": 10.076448374320714,
-      "grad_norm": 2.4599967002868652,
-      "learning_rate": 1.567461933363389e-05,
-      "loss": 0.582,
+      "epoch": 5.038224572240268,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.566880208257588e-05,
+      "loss": 0.962649917602539,
       "step": 437600
     },
     {
-      "epoch": 10.081053698074975,
-      "grad_norm": 2.899773120880127,
-      "learning_rate": 1.5667347769811376e-05,
-      "loss": 0.5706,
+      "epoch": 5.040527257314983,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5661530518753363e-05,
+      "loss": 0.9825556182861328,
       "step": 437800
     },
     {
-      "epoch": 10.085659021829235,
-      "grad_norm": 2.810063600540161,
-      "learning_rate": 1.566007620598886e-05,
-      "loss": 0.5678,
+      "epoch": 5.042829942389698,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5654258954930845e-05,
+      "loss": 0.9778278350830079,
       "step": 438000
     },
     {
-      "epoch": 10.090264345583494,
-      "grad_norm": 2.5231738090515137,
-      "learning_rate": 1.5652804642166345e-05,
-      "loss": 0.5784,
+      "epoch": 5.045132627464413,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.564698739110833e-05,
+      "loss": 0.9800408935546875,
       "step": 438200
     },
     {
-      "epoch": 10.094869669337754,
-      "grad_norm": 2.406407594680786,
-      "learning_rate": 1.564553307834383e-05,
-      "loss": 0.5653,
+      "epoch": 5.047435312539128,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.5639715827285817e-05,
+      "loss": 0.9869258117675781,
       "step": 438400
     },
     {
-      "epoch": 10.099474993092015,
-      "grad_norm": 3.0736372470855713,
-      "learning_rate": 1.5638261514521316e-05,
-      "loss": 0.5703,
+      "epoch": 5.049737997613843,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.56324442634633e-05,
+      "loss": 1.0070904541015624,
       "step": 438600
     },
     {
-      "epoch": 10.104080316846275,
-      "grad_norm": 2.5752131938934326,
-      "learning_rate": 1.5630989950698796e-05,
-      "loss": 0.5667,
+      "epoch": 5.052040682688557,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5625172699640786e-05,
+      "loss": 0.9841986846923828,
       "step": 438800
     },
     {
-      "epoch": 10.108685640600534,
-      "grad_norm": 3.5142014026641846,
-      "learning_rate": 1.562371838687628e-05,
-      "loss": 0.5829,
+      "epoch": 5.0543433677632725,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.561790113581827e-05,
+      "loss": 0.9610589599609375,
       "step": 439000
     },
     {
-      "epoch": 10.113290964354794,
-      "grad_norm": 3.4527642726898193,
-      "learning_rate": 1.5616446823053767e-05,
-      "loss": 0.5663,
+      "epoch": 5.056646052837987,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.5610629571995754e-05,
+      "loss": 0.9943707275390625,
       "step": 439200
     },
     {
-      "epoch": 10.117896288109055,
-      "grad_norm": 3.4308338165283203,
-      "learning_rate": 1.560917525923125e-05,
-      "loss": 0.5725,
+      "epoch": 5.0589487379127025,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5603358008173237e-05,
+      "loss": 0.9718830108642578,
       "step": 439400
     },
     {
-      "epoch": 10.122501611863314,
-      "grad_norm": 2.8218986988067627,
-      "learning_rate": 1.5601903695408736e-05,
-      "loss": 0.5723,
+      "epoch": 5.061251422987417,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5596086444350722e-05,
+      "loss": 0.9629135131835938,
       "step": 439600
     },
     {
-      "epoch": 10.127106935617574,
-      "grad_norm": 3.292224407196045,
-      "learning_rate": 1.559463213158622e-05,
-      "loss": 0.5753,
+      "epoch": 5.0635541080621325,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.558881488052821e-05,
+      "loss": 0.988746337890625,
       "step": 439800
     },
     {
-      "epoch": 10.131712259371835,
-      "grad_norm": 2.748878240585327,
-      "learning_rate": 1.5587396925582817e-05,
-      "loss": 0.5636,
+      "epoch": 5.065856793136847,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.558154331670569e-05,
+      "loss": 0.9679503631591797,
       "step": 440000
     },
     {
-      "epoch": 10.136317583126093,
-      "grad_norm": 3.109516143798828,
-      "learning_rate": 1.55801253617603e-05,
-      "loss": 0.5534,
+      "epoch": 5.068159478211562,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.5574271752883177e-05,
+      "loss": 0.9796177673339844,
       "step": 440200
     },
     {
-      "epoch": 10.140922906880354,
-      "grad_norm": 3.0865907669067383,
-      "learning_rate": 1.5572853797937785e-05,
-      "loss": 0.5619,
+      "epoch": 5.070462163286277,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.556700018906066e-05,
+      "loss": 0.9999834442138672,
       "step": 440400
     },
     {
-      "epoch": 10.145528230634614,
-      "grad_norm": 3.90614914894104,
-      "learning_rate": 1.5565582234115268e-05,
-      "loss": 0.5706,
+      "epoch": 5.0727648483609915,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5559728625238142e-05,
+      "loss": 0.9751573181152344,
       "step": 440600
     },
     {
-      "epoch": 10.150133554388873,
-      "grad_norm": 3.342874526977539,
-      "learning_rate": 1.5558310670292754e-05,
-      "loss": 0.5666,
+      "epoch": 5.075067533435707,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5552457061415628e-05,
+      "loss": 0.9943019866943359,
       "step": 440800
     },
     {
-      "epoch": 10.154738878143133,
-      "grad_norm": 3.8713204860687256,
-      "learning_rate": 1.555103910647024e-05,
-      "loss": 0.57,
+      "epoch": 5.0773702185104215,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5545185497593114e-05,
+      "loss": 0.9995348358154297,
       "step": 441000
     },
     {
-      "epoch": 10.159344201897394,
-      "grad_norm": 3.356518507003784,
-      "learning_rate": 1.5543767542647722e-05,
-      "loss": 0.5671,
+      "epoch": 5.079672903585137,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5537913933770596e-05,
+      "loss": 1.0012002563476563,
       "step": 441200
     },
     {
-      "epoch": 10.163949525651653,
-      "grad_norm": 3.4761784076690674,
-      "learning_rate": 1.5536495978825208e-05,
-      "loss": 0.5756,
+      "epoch": 5.081975588659851,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5530642369948082e-05,
+      "loss": 0.9686441802978516,
       "step": 441400
     },
     {
-      "epoch": 10.168554849405913,
-      "grad_norm": 3.0958569049835205,
-      "learning_rate": 1.552922441500269e-05,
-      "loss": 0.5545,
+      "epoch": 5.084278273734567,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5523370806125565e-05,
+      "loss": 0.9707472991943359,
       "step": 441600
     },
     {
-      "epoch": 10.173160173160174,
-      "grad_norm": 3.5366578102111816,
-      "learning_rate": 1.5521952851180173e-05,
-      "loss": 0.5569,
+      "epoch": 5.086580958809281,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.551609924230305e-05,
+      "loss": 0.991561279296875,
       "step": 441800
     },
     {
-      "epoch": 10.177765496914432,
-      "grad_norm": 3.0976240634918213,
-      "learning_rate": 1.551468128735766e-05,
-      "loss": 0.5787,
+      "epoch": 5.088883643883997,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5508827678480533e-05,
+      "loss": 0.9646013641357422,
       "step": 442000
     },
     {
-      "epoch": 10.182370820668693,
-      "grad_norm": 2.9487667083740234,
-      "learning_rate": 1.5507409723535145e-05,
-      "loss": 0.5719,
+      "epoch": 5.091186328958711,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.550155611465802e-05,
+      "loss": 0.9842584991455078,
       "step": 442200
     },
     {
-      "epoch": 10.186976144422953,
-      "grad_norm": 3.9748101234436035,
-      "learning_rate": 1.5500138159712628e-05,
-      "loss": 0.5772,
+      "epoch": 5.093489014033427,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5494284550835505e-05,
+      "loss": 0.9831058502197265,
       "step": 442400
     },
     {
-      "epoch": 10.191581468177214,
-      "grad_norm": 3.862164258956909,
-      "learning_rate": 1.5492866595890113e-05,
-      "loss": 0.5625,
+      "epoch": 5.095791699108141,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5487012987012987e-05,
+      "loss": 0.9710477447509765,
       "step": 442600
     },
     {
-      "epoch": 10.196186791931472,
-      "grad_norm": 3.5025646686553955,
-      "learning_rate": 1.5485595032067596e-05,
-      "loss": 0.5744,
+      "epoch": 5.098094384182856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5479741423190473e-05,
+      "loss": 0.986195068359375,
       "step": 442800
     },
     {
-      "epoch": 10.200792115685733,
-      "grad_norm": 2.911168336868286,
-      "learning_rate": 1.5478323468245082e-05,
-      "loss": 0.5606,
+      "epoch": 5.100397069257571,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5472469859367956e-05,
+      "loss": 0.9769602966308594,
       "step": 443000
     },
     {
-      "epoch": 10.205397439439993,
-      "grad_norm": 3.7616069316864014,
-      "learning_rate": 1.5471051904422564e-05,
-      "loss": 0.5661,
+      "epoch": 5.102699754332286,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.546519829554544e-05,
+      "loss": 0.982730484008789,
       "step": 443200
     },
     {
-      "epoch": 10.210002763194252,
-      "grad_norm": 2.7809619903564453,
-      "learning_rate": 1.546378034060005e-05,
-      "loss": 0.5583,
+      "epoch": 5.105002439407001,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.5457926731722924e-05,
+      "loss": 0.9746049499511719,
       "step": 443400
     },
     {
-      "epoch": 10.214608086948513,
-      "grad_norm": 4.306921005249023,
-      "learning_rate": 1.5456508776777536e-05,
-      "loss": 0.5722,
+      "epoch": 5.107305124481716,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.545065516790041e-05,
+      "loss": 1.0082730102539061,
       "step": 443600
     },
     {
-      "epoch": 10.219213410702773,
-      "grad_norm": 2.9962635040283203,
-      "learning_rate": 1.544923721295502e-05,
-      "loss": 0.5667,
+      "epoch": 5.109607809556431,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5443383604077893e-05,
+      "loss": 0.9730638885498046,
       "step": 443800
     },
     {
-      "epoch": 10.223818734457032,
-      "grad_norm": 2.8903093338012695,
-      "learning_rate": 1.54419656491325e-05,
-      "loss": 0.5655,
+      "epoch": 5.111910494631146,
+      "grad_norm": 0.625,
+      "learning_rate": 1.543611204025538e-05,
+      "loss": 0.9843389129638672,
       "step": 444000
     },
     {
-      "epoch": 10.228424058211292,
-      "grad_norm": 3.401038885116577,
-      "learning_rate": 1.5434694085309987e-05,
-      "loss": 0.5708,
+      "epoch": 5.114213179705861,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.542884047643286e-05,
+      "loss": 0.9781219482421875,
       "step": 444200
     },
     {
-      "epoch": 10.233029381965553,
-      "grad_norm": 2.52870774269104,
-      "learning_rate": 1.542742252148747e-05,
-      "loss": 0.564,
+      "epoch": 5.116515864780576,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.5421568912610344e-05,
+      "loss": 0.9844216156005859,
       "step": 444400
     },
     {
-      "epoch": 10.237634705719811,
-      "grad_norm": 3.640711784362793,
-      "learning_rate": 1.542018731548407e-05,
-      "loss": 0.5723,
+      "epoch": 5.11881854985529,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.541429734878783e-05,
+      "loss": 0.9864201354980469,
       "step": 444600
     },
     {
-      "epoch": 10.242240029474072,
-      "grad_norm": 2.800980806350708,
-      "learning_rate": 1.5412915751661554e-05,
-      "loss": 0.5613,
+      "epoch": 5.121121234930006,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5407025784965316e-05,
+      "loss": 0.9955843353271484,
       "step": 444800
     },
     {
-      "epoch": 10.246845353228332,
-      "grad_norm": 3.3499021530151367,
-      "learning_rate": 1.5405644187839037e-05,
-      "loss": 0.5739,
+      "epoch": 5.12342392000472,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.53997542211428e-05,
+      "loss": 0.9566255950927735,
       "step": 445000
     },
     {
-      "epoch": 10.251450676982591,
-      "grad_norm": 2.415259838104248,
-      "learning_rate": 1.5398408981835635e-05,
-      "loss": 0.566,
+      "epoch": 5.125726605079436,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5392482657320284e-05,
+      "loss": 0.9770396423339843,
       "step": 445200
     },
     {
-      "epoch": 10.256056000736852,
-      "grad_norm": 3.0290565490722656,
-      "learning_rate": 1.5391137418013118e-05,
-      "loss": 0.5834,
+      "epoch": 5.12802929015415,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.538521109349777e-05,
+      "loss": 0.9763878631591797,
       "step": 445400
     },
     {
-      "epoch": 10.260661324491112,
-      "grad_norm": 2.8429722785949707,
-      "learning_rate": 1.5383865854190604e-05,
-      "loss": 0.5733,
+      "epoch": 5.130331975228866,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5377939529675252e-05,
+      "loss": 0.97738037109375,
       "step": 445600
     },
     {
-      "epoch": 10.265266648245372,
-      "grad_norm": 2.889953136444092,
-      "learning_rate": 1.5376594290368086e-05,
-      "loss": 0.5635,
+      "epoch": 5.13263466030358,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.5370667965852735e-05,
+      "loss": 0.9798085021972657,
       "step": 445800
     },
     {
-      "epoch": 10.269871971999631,
-      "grad_norm": 3.5304625034332275,
-      "learning_rate": 1.536932272654557e-05,
-      "loss": 0.5689,
+      "epoch": 5.134937345378296,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.536339640203022e-05,
+      "loss": 0.9778961944580078,
       "step": 446000
     },
     {
-      "epoch": 10.274477295753892,
-      "grad_norm": 3.2472240924835205,
-      "learning_rate": 1.5362051162723055e-05,
-      "loss": 0.5626,
+      "epoch": 5.13724003045301,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5356124838207707e-05,
+      "loss": 0.9781346130371094,
       "step": 446200
     },
     {
-      "epoch": 10.279082619508152,
-      "grad_norm": 3.443258047103882,
-      "learning_rate": 1.535477959890054e-05,
-      "loss": 0.5686,
+      "epoch": 5.139542715527725,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.534885327438519e-05,
+      "loss": 1.0051431274414062,
       "step": 446400
     },
     {
-      "epoch": 10.28368794326241,
-      "grad_norm": 3.0054471492767334,
-      "learning_rate": 1.5347508035078027e-05,
-      "loss": 0.563,
+      "epoch": 5.14184540060244,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5341581710562675e-05,
+      "loss": 0.9866223907470704,
       "step": 446600
     },
     {
-      "epoch": 10.288293267016671,
-      "grad_norm": 3.115903854370117,
-      "learning_rate": 1.534023647125551e-05,
-      "loss": 0.5618,
+      "epoch": 5.144148085677155,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.5334310146740158e-05,
+      "loss": 0.979072265625,
       "step": 446800
     },
     {
-      "epoch": 10.292898590770932,
-      "grad_norm": 3.3110814094543457,
-      "learning_rate": 1.533296490743299e-05,
-      "loss": 0.5567,
+      "epoch": 5.14645077075187,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.532703858291764e-05,
+      "loss": 1.0071894073486327,
       "step": 447000
     },
     {
-      "epoch": 10.29750391452519,
-      "grad_norm": 2.7418558597564697,
-      "learning_rate": 1.5325693343610478e-05,
-      "loss": 0.5526,
+      "epoch": 5.148753455826585,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5319767019095126e-05,
+      "loss": 0.9757035064697266,
       "step": 447200
     },
     {
-      "epoch": 10.302109238279451,
-      "grad_norm": 3.7277021408081055,
-      "learning_rate": 1.531842177978796e-05,
-      "loss": 0.5549,
+      "epoch": 5.1510561409013,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5312495455272612e-05,
+      "loss": 0.9830738067626953,
       "step": 447400
     },
     {
-      "epoch": 10.306714562033712,
-      "grad_norm": 3.4062893390655518,
-      "learning_rate": 1.5311150215965446e-05,
-      "loss": 0.5749,
+      "epoch": 5.153358825976015,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5305223891450098e-05,
+      "loss": 0.9869244384765625,
       "step": 447600
     },
     {
-      "epoch": 10.31131988578797,
-      "grad_norm": 3.7285375595092773,
-      "learning_rate": 1.5303878652142932e-05,
-      "loss": 0.5635,
+      "epoch": 5.15566151105073,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.529795232762758e-05,
+      "loss": 0.9828427124023438,
       "step": 447800
     },
     {
-      "epoch": 10.31592520954223,
-      "grad_norm": 3.451294183731079,
-      "learning_rate": 1.5296643446139527e-05,
-      "loss": 0.5615,
+      "epoch": 5.1579641961254445,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5290680763805063e-05,
+      "loss": 0.9620997619628906,
       "step": 448000
     },
     {
-      "epoch": 10.320530533296491,
-      "grad_norm": 2.3311824798583984,
-      "learning_rate": 1.5289371882317013e-05,
-      "loss": 0.5668,
+      "epoch": 5.160266881200159,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.528340919998255e-05,
+      "loss": 0.9812986755371094,
       "step": 448200
     },
     {
-      "epoch": 10.32513585705075,
-      "grad_norm": 3.451138734817505,
-      "learning_rate": 1.52821003184945e-05,
-      "loss": 0.5659,
+      "epoch": 5.1625695662748745,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.527613763616003e-05,
+      "loss": 0.9796999359130859,
       "step": 448400
     },
     {
-      "epoch": 10.32974118080501,
-      "grad_norm": 2.754631757736206,
-      "learning_rate": 1.5274828754671978e-05,
-      "loss": 0.5746,
+      "epoch": 5.164872251349589,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5268866072337517e-05,
+      "loss": 0.9977180480957031,
       "step": 448600
     },
     {
-      "epoch": 10.33434650455927,
-      "grad_norm": 3.7678756713867188,
-      "learning_rate": 1.5267557190849464e-05,
-      "loss": 0.5647,
+      "epoch": 5.1671749364243045,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5261594508515003e-05,
+      "loss": 0.9861807250976562,
       "step": 448800
     },
     {
-      "epoch": 10.33895182831353,
-      "grad_norm": 2.977546215057373,
-      "learning_rate": 1.526028562702695e-05,
-      "loss": 0.5675,
+      "epoch": 5.169477621499019,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5254322944692488e-05,
+      "loss": 0.9859234619140625,
       "step": 449000
     },
     {
-      "epoch": 10.34355715206779,
-      "grad_norm": 2.2833847999572754,
-      "learning_rate": 1.5253014063204434e-05,
-      "loss": 0.5571,
+      "epoch": 5.171780306573734,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5247051380869972e-05,
+      "loss": 0.9808116912841797,
       "step": 449200
     },
     {
-      "epoch": 10.34816247582205,
-      "grad_norm": 3.352088689804077,
-      "learning_rate": 1.5245742499381918e-05,
-      "loss": 0.5803,
+      "epoch": 5.174082991648449,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5239779817047453e-05,
+      "loss": 0.9652620697021485,
       "step": 449400
     },
     {
-      "epoch": 10.352767799576311,
-      "grad_norm": 3.5555100440979004,
-      "learning_rate": 1.5238470935559403e-05,
-      "loss": 0.5733,
+      "epoch": 5.176385676723164,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5232508253224939e-05,
+      "loss": 0.9635680389404296,
       "step": 449600
     },
     {
-      "epoch": 10.35737312333057,
-      "grad_norm": 2.3247013092041016,
-      "learning_rate": 1.5231199371736885e-05,
-      "loss": 0.5573,
+      "epoch": 5.178688361797879,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5225236689402423e-05,
+      "loss": 1.0047270202636718,
       "step": 449800
     },
     {
-      "epoch": 10.36197844708483,
-      "grad_norm": 2.8674051761627197,
-      "learning_rate": 1.5223964165733482e-05,
-      "loss": 0.5689,
+      "epoch": 5.1809910468725935,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5217965125579909e-05,
+      "loss": 0.9865448760986328,
       "step": 450000
     },
     {
-      "epoch": 10.36658377083909,
-      "grad_norm": 3.7046046257019043,
-      "learning_rate": 1.5216692601910966e-05,
-      "loss": 0.5753,
+      "epoch": 5.183293731947309,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.5210693561757393e-05,
+      "loss": 0.9875264739990235,
       "step": 450200
     },
     {
-      "epoch": 10.37118909459335,
-      "grad_norm": 3.3615944385528564,
-      "learning_rate": 1.520942103808845e-05,
-      "loss": 0.5823,
+      "epoch": 5.1855964170220235,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5203421997934877e-05,
+      "loss": 0.9982109069824219,
       "step": 450400
     },
     {
-      "epoch": 10.37579441834761,
-      "grad_norm": 4.1015238761901855,
-      "learning_rate": 1.5202149474265936e-05,
-      "loss": 0.5746,
+      "epoch": 5.187899102096739,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.519615043411236e-05,
+      "loss": 0.983709945678711,
       "step": 450600
     },
     {
-      "epoch": 10.38039974210187,
-      "grad_norm": 2.8031654357910156,
-      "learning_rate": 1.519487791044342e-05,
-      "loss": 0.5707,
+      "epoch": 5.190201787171453,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5188878870289844e-05,
+      "loss": 0.9661166381835937,
       "step": 450800
     },
     {
-      "epoch": 10.385005065856129,
-      "grad_norm": 3.370823860168457,
-      "learning_rate": 1.5187606346620907e-05,
-      "loss": 0.5652,
+      "epoch": 5.192504472246169,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.5181607306467328e-05,
+      "loss": 0.9745197296142578,
       "step": 451000
     },
     {
-      "epoch": 10.38961038961039,
-      "grad_norm": 3.259626626968384,
-      "learning_rate": 1.5180371140617503e-05,
-      "loss": 0.5666,
+      "epoch": 5.194807157320883,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5174335742644814e-05,
+      "loss": 0.9809587097167969,
       "step": 451200
     },
     {
-      "epoch": 10.39421571336465,
-      "grad_norm": 4.263233184814453,
-      "learning_rate": 1.5173099576794984e-05,
-      "loss": 0.5697,
+      "epoch": 5.197109842395599,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5167064178822298e-05,
+      "loss": 0.9783151245117188,
       "step": 451400
     },
     {
-      "epoch": 10.398821037118909,
-      "grad_norm": 2.861125946044922,
-      "learning_rate": 1.5165828012972468e-05,
-      "loss": 0.5782,
+      "epoch": 5.199412527470313,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5159792614999784e-05,
+      "loss": 0.9720295715332031,
       "step": 451600
     },
     {
-      "epoch": 10.40342636087317,
-      "grad_norm": 3.5409388542175293,
-      "learning_rate": 1.5158556449149954e-05,
-      "loss": 0.5651,
+      "epoch": 5.201715212545028,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5152521051177265e-05,
+      "loss": 0.9959677124023437,
       "step": 451800
     },
     {
-      "epoch": 10.40803168462743,
-      "grad_norm": 3.4147825241088867,
-      "learning_rate": 1.5151284885327439e-05,
-      "loss": 0.5646,
+      "epoch": 5.204017897619743,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.514524948735475e-05,
+      "loss": 0.9739774322509765,
       "step": 452000
     },
     {
-      "epoch": 10.412637008381688,
-      "grad_norm": 2.7869393825531006,
-      "learning_rate": 1.5144013321504925e-05,
-      "loss": 0.5703,
+      "epoch": 5.206320582694458,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5137977923532235e-05,
+      "loss": 0.9681526947021485,
       "step": 452200
     },
     {
-      "epoch": 10.417242332135949,
-      "grad_norm": 3.2392425537109375,
-      "learning_rate": 1.5136741757682409e-05,
-      "loss": 0.5604,
+      "epoch": 5.208623267769173,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.513070635970972e-05,
+      "loss": 0.9602870941162109,
       "step": 452400
     },
     {
-      "epoch": 10.42184765589021,
-      "grad_norm": 3.0997209548950195,
-      "learning_rate": 1.5129470193859893e-05,
-      "loss": 0.5643,
+      "epoch": 5.210925952843888,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5123434795887204e-05,
+      "loss": 1.0015093994140625,
       "step": 452600
     },
     {
-      "epoch": 10.42645297964447,
-      "grad_norm": 2.82871150970459,
-      "learning_rate": 1.5122198630037375e-05,
-      "loss": 0.5727,
+      "epoch": 5.213228637918603,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.511616323206469e-05,
+      "loss": 0.96810302734375,
       "step": 452800
     },
     {
-      "epoch": 10.431058303398729,
-      "grad_norm": 3.5015945434570312,
-      "learning_rate": 1.511492706621486e-05,
-      "loss": 0.566,
+      "epoch": 5.215531322993318,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.5108891668242174e-05,
+      "loss": 0.9807380676269531,
       "step": 453000
     },
     {
-      "epoch": 10.435663627152989,
-      "grad_norm": 2.741576671600342,
-      "learning_rate": 1.5107655502392344e-05,
-      "loss": 0.5739,
+      "epoch": 5.217834008068033,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5101620104419656e-05,
+      "loss": 0.9774137115478516,
       "step": 453200
     },
     {
-      "epoch": 10.44026895090725,
-      "grad_norm": 3.6131269931793213,
-      "learning_rate": 1.510038393856983e-05,
-      "loss": 0.5577,
+      "epoch": 5.220136693142748,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.509434854059714e-05,
+      "loss": 0.9768766021728515,
       "step": 453400
     },
     {
-      "epoch": 10.444874274661508,
-      "grad_norm": 3.369382858276367,
-      "learning_rate": 1.5093112374747314e-05,
-      "loss": 0.5676,
+      "epoch": 5.222439378217462,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.5087076976774625e-05,
+      "loss": 1.003528823852539,
       "step": 453600
     },
     {
-      "epoch": 10.449479598415769,
-      "grad_norm": 2.8960883617401123,
-      "learning_rate": 1.50858408109248e-05,
-      "loss": 0.5564,
+      "epoch": 5.224742063292178,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.507980541295211e-05,
+      "loss": 0.9844652557373047,
       "step": 453800
     },
     {
-      "epoch": 10.45408492217003,
-      "grad_norm": 3.1670303344726562,
-      "learning_rate": 1.507856924710228e-05,
-      "loss": 0.5777,
+      "epoch": 5.227044748366892,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5072533849129595e-05,
+      "loss": 0.9844374084472656,
       "step": 454000
     },
     {
-      "epoch": 10.458690245924288,
-      "grad_norm": 2.9549036026000977,
-      "learning_rate": 1.5071297683279765e-05,
-      "loss": 0.5686,
+      "epoch": 5.229347433441608,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.506526228530708e-05,
+      "loss": 0.9929544830322266,
       "step": 454200
     },
     {
-      "epoch": 10.463295569678548,
-      "grad_norm": 3.2000997066497803,
-      "learning_rate": 1.5064026119457251e-05,
-      "loss": 0.5699,
+      "epoch": 5.231650118516322,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5057990721484561e-05,
+      "loss": 0.9618767547607422,
       "step": 454400
     },
     {
-      "epoch": 10.467900893432809,
-      "grad_norm": 3.4854366779327393,
-      "learning_rate": 1.5056754555634735e-05,
-      "loss": 0.5676,
+      "epoch": 5.233952803591038,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5050719157662046e-05,
+      "loss": 0.9770437622070313,
       "step": 454600
     },
     {
-      "epoch": 10.472506217187068,
-      "grad_norm": 3.309793710708618,
-      "learning_rate": 1.504948299181222e-05,
-      "loss": 0.5569,
+      "epoch": 5.236255488665752,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5043447593839532e-05,
+      "loss": 1.0074984741210937,
       "step": 454800
     },
     {
-      "epoch": 10.477111540941328,
-      "grad_norm": 2.8395392894744873,
-      "learning_rate": 1.5042211427989705e-05,
-      "loss": 0.569,
+      "epoch": 5.238558173740468,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.5036176030017016e-05,
+      "loss": 0.9927102661132813,
       "step": 455000
     },
     {
-      "epoch": 10.481716864695588,
-      "grad_norm": 3.1417343616485596,
-      "learning_rate": 1.503493986416719e-05,
-      "loss": 0.5696,
+      "epoch": 5.240860858815182,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.50289044661945e-05,
+      "loss": 0.9880549621582031,
       "step": 455200
     },
     {
-      "epoch": 10.486322188449847,
-      "grad_norm": 3.1475086212158203,
-      "learning_rate": 1.5027668300344672e-05,
-      "loss": 0.5647,
+      "epoch": 5.243163543889897,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5021632902371986e-05,
+      "loss": 0.9959085845947265,
       "step": 455400
     },
     {
-      "epoch": 10.490927512204108,
-      "grad_norm": 3.277684450149536,
-      "learning_rate": 1.5020396736522156e-05,
-      "loss": 0.58,
+      "epoch": 5.245466228964612,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.501436133854947e-05,
+      "loss": 0.993404541015625,
       "step": 455600
     },
     {
-      "epoch": 10.495532835958368,
-      "grad_norm": 3.4847934246063232,
-      "learning_rate": 1.501312517269964e-05,
-      "loss": 0.5613,
+      "epoch": 5.247768914039327,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.5007089774726953e-05,
+      "loss": 0.9818459320068359,
       "step": 455800
     },
     {
-      "epoch": 10.500138159712627,
-      "grad_norm": 2.999443292617798,
-      "learning_rate": 1.5005853608877126e-05,
-      "loss": 0.5617,
+      "epoch": 5.250071599114042,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4999818210904437e-05,
+      "loss": 0.9703733062744141,
       "step": 456000
     },
     {
-      "epoch": 10.504743483466887,
-      "grad_norm": 2.7356905937194824,
-      "learning_rate": 1.4998582045054609e-05,
-      "loss": 0.5679,
+      "epoch": 5.252374284188757,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4992546647081921e-05,
+      "loss": 0.9827092742919922,
       "step": 456200
     },
     {
-      "epoch": 10.509348807221148,
-      "grad_norm": 3.4266111850738525,
-      "learning_rate": 1.4991310481232093e-05,
-      "loss": 0.5498,
+      "epoch": 5.254676969263472,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4985275083259407e-05,
+      "loss": 0.9929819488525391,
       "step": 456400
     },
     {
-      "epoch": 10.513954130975408,
-      "grad_norm": 3.2002484798431396,
-      "learning_rate": 1.4984038917409579e-05,
-      "loss": 0.5548,
+      "epoch": 5.256979654338187,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.497800351943689e-05,
+      "loss": 0.9914261627197266,
       "step": 456600
     },
     {
-      "epoch": 10.518559454729667,
-      "grad_norm": 2.8026347160339355,
-      "learning_rate": 1.4976767353587063e-05,
-      "loss": 0.5744,
+      "epoch": 5.259282339412902,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4970731955614374e-05,
+      "loss": 0.9822914123535156,
       "step": 456800
     },
     {
-      "epoch": 10.523164778483928,
-      "grad_norm": 3.1921815872192383,
-      "learning_rate": 1.4969495789764547e-05,
-      "loss": 0.5578,
+      "epoch": 5.261585024487617,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.496346039179186e-05,
+      "loss": 0.9853893280029297,
       "step": 457000
     },
     {
-      "epoch": 10.527770102238188,
-      "grad_norm": 3.4548532962799072,
-      "learning_rate": 1.4962224225942032e-05,
-      "loss": 0.5696,
+      "epoch": 5.263887709562331,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4956188827969344e-05,
+      "loss": 0.9783312225341797,
       "step": 457200
     },
     {
-      "epoch": 10.532375425992447,
-      "grad_norm": 2.9152286052703857,
-      "learning_rate": 1.4954952662119516e-05,
-      "loss": 0.5592,
+      "epoch": 5.2661903946370465,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4948917264146828e-05,
+      "loss": 0.9891160583496094,
       "step": 457400
     },
     {
-      "epoch": 10.536980749746707,
-      "grad_norm": 2.514460802078247,
-      "learning_rate": 1.4947681098297e-05,
-      "loss": 0.5629,
+      "epoch": 5.268493079711761,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4941645700324312e-05,
+      "loss": 0.9707080841064453,
       "step": 457600
     },
     {
-      "epoch": 10.541586073500968,
-      "grad_norm": 3.525857448577881,
-      "learning_rate": 1.4940409534474484e-05,
-      "loss": 0.5682,
+      "epoch": 5.2707957647864765,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.4934374136501797e-05,
+      "loss": 0.9764946746826172,
       "step": 457800
     },
     {
-      "epoch": 10.546191397255226,
-      "grad_norm": 3.7510156631469727,
-      "learning_rate": 1.4933137970651969e-05,
-      "loss": 0.5724,
+      "epoch": 5.273098449861191,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.492710257267928e-05,
+      "loss": 0.9986644744873047,
       "step": 458000
     },
     {
-      "epoch": 10.550796721009487,
-      "grad_norm": 3.3571460247039795,
-      "learning_rate": 1.4925902764648565e-05,
-      "loss": 0.568,
+      "epoch": 5.2754011349359065,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4919831008856765e-05,
+      "loss": 0.984770278930664,
       "step": 458200
     },
     {
-      "epoch": 10.555402044763747,
-      "grad_norm": 2.8560373783111572,
-      "learning_rate": 1.491863120082605e-05,
-      "loss": 0.5629,
+      "epoch": 5.277703820010621,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.491255944503425e-05,
+      "loss": 0.9775885772705079,
       "step": 458400
     },
     {
-      "epoch": 10.560007368518006,
-      "grad_norm": 3.2522172927856445,
-      "learning_rate": 1.4911359637003534e-05,
-      "loss": 0.5643,
+      "epoch": 5.280006505085336,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4905287881211733e-05,
+      "loss": 0.9861846923828125,
       "step": 458600
     },
     {
-      "epoch": 10.564612692272267,
-      "grad_norm": 3.196812152862549,
-      "learning_rate": 1.490408807318102e-05,
-      "loss": 0.5603,
+      "epoch": 5.282309190160051,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.4898016317389218e-05,
+      "loss": 0.9947913360595703,
       "step": 458800
     },
     {
-      "epoch": 10.569218016026527,
-      "grad_norm": 3.4622724056243896,
-      "learning_rate": 1.4896816509358502e-05,
-      "loss": 0.5625,
+      "epoch": 5.284611875234766,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4890744753566704e-05,
+      "loss": 0.9821308898925781,
       "step": 459000
     },
     {
-      "epoch": 10.573823339780787,
-      "grad_norm": 2.8027701377868652,
-      "learning_rate": 1.4889544945535987e-05,
-      "loss": 0.5714,
+      "epoch": 5.286914560309481,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4883473189744186e-05,
+      "loss": 0.991390609741211,
       "step": 459200
     },
     {
-      "epoch": 10.578428663535046,
-      "grad_norm": 2.9170937538146973,
-      "learning_rate": 1.4882273381713472e-05,
-      "loss": 0.5723,
+      "epoch": 5.2892172453841955,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.487620162592167e-05,
+      "loss": 0.9703672027587891,
       "step": 459400
     },
     {
-      "epoch": 10.583033987289307,
-      "grad_norm": 3.9101593494415283,
-      "learning_rate": 1.4875001817890955e-05,
-      "loss": 0.5641,
+      "epoch": 5.291519930458911,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4868930062099156e-05,
+      "loss": 0.9787174224853515,
       "step": 459600
     },
     {
-      "epoch": 10.587639311043567,
-      "grad_norm": 3.0541505813598633,
-      "learning_rate": 1.4867730254068441e-05,
-      "loss": 0.5576,
+      "epoch": 5.293822615533625,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4861658498276639e-05,
+      "loss": 0.9987583923339843,
       "step": 459800
     },
     {
-      "epoch": 10.592244634797826,
-      "grad_norm": 2.2873919010162354,
-      "learning_rate": 1.4860458690245925e-05,
-      "loss": 0.5682,
+      "epoch": 5.296125300608341,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4854386934454125e-05,
+      "loss": 0.9813053131103515,
       "step": 460000
     },
     {
-      "epoch": 10.596849958552086,
-      "grad_norm": 4.237215518951416,
-      "learning_rate": 1.485318712642341e-05,
-      "loss": 0.5709,
+      "epoch": 5.298427985683055,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4847115370631609e-05,
+      "loss": 1.0133036041259766,
       "step": 460200
     },
     {
-      "epoch": 10.601455282306347,
-      "grad_norm": 3.327172040939331,
-      "learning_rate": 1.4845915562600894e-05,
-      "loss": 0.5712,
+      "epoch": 5.300730670757771,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4839843806809091e-05,
+      "loss": 1.000592269897461,
       "step": 460400
     },
     {
-      "epoch": 10.606060606060606,
-      "grad_norm": 3.6685967445373535,
-      "learning_rate": 1.4838643998778378e-05,
-      "loss": 0.5654,
+      "epoch": 5.303033355832485,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4832572242986577e-05,
+      "loss": 0.9804405212402344,
       "step": 460600
     },
     {
-      "epoch": 10.610665929814866,
-      "grad_norm": 3.0738847255706787,
-      "learning_rate": 1.4831372434955862e-05,
-      "loss": 0.5676,
+      "epoch": 5.3053360409072,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4825300679164062e-05,
+      "loss": 0.9746564483642578,
       "step": 460800
     },
     {
-      "epoch": 10.615271253569126,
-      "grad_norm": 2.7393765449523926,
-      "learning_rate": 1.4824100871133346e-05,
-      "loss": 0.5645,
+      "epoch": 5.307638725981915,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4818029115341546e-05,
+      "loss": 0.9901017761230468,
       "step": 461000
     },
     {
-      "epoch": 10.619876577323385,
-      "grad_norm": 2.6728312969207764,
-      "learning_rate": 1.481682930731083e-05,
-      "loss": 0.5598,
+      "epoch": 5.30994141105663,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.481075755151903e-05,
+      "loss": 0.9895965576171875,
       "step": 461200
     },
     {
-      "epoch": 10.624481901077646,
-      "grad_norm": 2.915443181991577,
-      "learning_rate": 1.4809557743488316e-05,
-      "loss": 0.5659,
+      "epoch": 5.312244096131345,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4803485987696514e-05,
+      "loss": 0.9866258239746094,
       "step": 461400
     },
     {
-      "epoch": 10.629087224831906,
-      "grad_norm": 3.617243528366089,
-      "learning_rate": 1.4802286179665799e-05,
-      "loss": 0.5706,
+      "epoch": 5.31454678120606,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4796214423874e-05,
+      "loss": 0.9936985778808594,
       "step": 461600
     },
     {
-      "epoch": 10.633692548586165,
-      "grad_norm": 3.1753506660461426,
-      "learning_rate": 1.4795014615843283e-05,
-      "loss": 0.5551,
+      "epoch": 5.316849466280775,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4788942860051483e-05,
+      "loss": 0.988270034790039,
       "step": 461800
     },
     {
-      "epoch": 10.638297872340425,
-      "grad_norm": 3.585939407348633,
-      "learning_rate": 1.4787743052020769e-05,
-      "loss": 0.5634,
+      "epoch": 5.31915215135549,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4781671296228967e-05,
+      "loss": 0.9763542938232422,
       "step": 462000
     },
     {
-      "epoch": 10.642903196094686,
-      "grad_norm": 2.8085319995880127,
-      "learning_rate": 1.4780471488198252e-05,
-      "loss": 0.572,
+      "epoch": 5.321454836430205,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4774399732406453e-05,
+      "loss": 0.980770263671875,
       "step": 462200
     },
     {
-      "epoch": 10.647508519848945,
-      "grad_norm": 3.3687798976898193,
-      "learning_rate": 1.4773236282194848e-05,
-      "loss": 0.5583,
+      "epoch": 5.32375752150492,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4767128168583935e-05,
+      "loss": 0.9823168182373047,
       "step": 462400
     },
     {
-      "epoch": 10.652113843603205,
-      "grad_norm": 3.3423590660095215,
-      "learning_rate": 1.4765964718372333e-05,
-      "loss": 0.572,
+      "epoch": 5.326060206579635,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.475985660476142e-05,
+      "loss": 0.9989656829833984,
       "step": 462600
     },
     {
-      "epoch": 10.656719167357465,
-      "grad_norm": 2.939920425415039,
-      "learning_rate": 1.4758693154549819e-05,
-      "loss": 0.5757,
+      "epoch": 5.32836289165435,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4752585040938905e-05,
+      "loss": 1.0028610992431641,
       "step": 462800
     },
     {
-      "epoch": 10.661324491111726,
-      "grad_norm": 3.0086398124694824,
-      "learning_rate": 1.4751457948546416e-05,
-      "loss": 0.5608,
+      "epoch": 5.330665576729064,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4745313477116388e-05,
+      "loss": 1.0008536529541017,
       "step": 463000
     },
     {
-      "epoch": 10.665929814865985,
-      "grad_norm": 3.477020740509033,
-      "learning_rate": 1.4744186384723898e-05,
-      "loss": 0.5704,
+      "epoch": 5.33296826180378,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4738041913293874e-05,
+      "loss": 0.958477783203125,
       "step": 463200
     },
     {
-      "epoch": 10.670535138620245,
-      "grad_norm": 3.078016996383667,
-      "learning_rate": 1.4736914820901384e-05,
-      "loss": 0.5596,
+      "epoch": 5.335270946878494,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4730770349471358e-05,
+      "loss": 1.0013047027587891,
       "step": 463400
     },
     {
-      "epoch": 10.675140462374506,
-      "grad_norm": 3.607464075088501,
-      "learning_rate": 1.4729643257078868e-05,
-      "loss": 0.5584,
+      "epoch": 5.33757363195321,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.472349878564884e-05,
+      "loss": 0.9811349487304688,
       "step": 463600
     },
     {
-      "epoch": 10.679745786128764,
-      "grad_norm": 2.997453212738037,
-      "learning_rate": 1.4722371693256352e-05,
-      "loss": 0.5566,
+      "epoch": 5.339876317027924,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4716227221826327e-05,
+      "loss": 0.9876372528076172,
       "step": 463800
     },
     {
-      "epoch": 10.684351109883025,
-      "grad_norm": 2.907045364379883,
-      "learning_rate": 1.4715100129433837e-05,
-      "loss": 0.5699,
+      "epoch": 5.34217900210264,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.470895565800381e-05,
+      "loss": 0.987093505859375,
       "step": 464000
     },
     {
-      "epoch": 10.688956433637285,
-      "grad_norm": 2.692117929458618,
-      "learning_rate": 1.470782856561132e-05,
-      "loss": 0.5607,
+      "epoch": 5.344481687177354,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.4701684094181295e-05,
+      "loss": 0.9624461364746094,
       "step": 464200
     },
     {
-      "epoch": 10.693561757391544,
-      "grad_norm": 3.213514804840088,
-      "learning_rate": 1.4700557001788805e-05,
-      "loss": 0.5622,
+      "epoch": 5.346784372252069,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.469441253035878e-05,
+      "loss": 0.972740249633789,
       "step": 464400
     },
     {
-      "epoch": 10.698167081145805,
-      "grad_norm": 2.8752119541168213,
-      "learning_rate": 1.469328543796629e-05,
-      "loss": 0.5759,
+      "epoch": 5.349087057326784,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4687140966536263e-05,
+      "loss": 1.0043175506591797,
       "step": 464600
     },
     {
-      "epoch": 10.702772404900065,
-      "grad_norm": 2.7570505142211914,
-      "learning_rate": 1.4686013874143773e-05,
-      "loss": 0.5732,
+      "epoch": 5.351389742401499,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.467986940271375e-05,
+      "loss": 0.9737282562255859,
       "step": 464800
     },
     {
-      "epoch": 10.707377728654324,
-      "grad_norm": 2.800178050994873,
-      "learning_rate": 1.467877866814037e-05,
-      "loss": 0.5693,
+      "epoch": 5.353692427476214,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4672597838891232e-05,
+      "loss": 0.9622456359863282,
       "step": 465000
     },
     {
-      "epoch": 10.711983052408584,
-      "grad_norm": 2.608539342880249,
-      "learning_rate": 1.4671507104317856e-05,
-      "loss": 0.5683,
+      "epoch": 5.355995112550929,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4665326275068716e-05,
+      "loss": 0.9822862243652344,
       "step": 465200
     },
     {
-      "epoch": 10.716588376162845,
-      "grad_norm": 3.1758687496185303,
-      "learning_rate": 1.4664235540495339e-05,
-      "loss": 0.571,
+      "epoch": 5.358297797625644,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4658054711246202e-05,
+      "loss": 0.9863087463378907,
       "step": 465400
     },
     {
-      "epoch": 10.721193699917103,
-      "grad_norm": 2.868126392364502,
-      "learning_rate": 1.4656963976672823e-05,
-      "loss": 0.5706,
+      "epoch": 5.360600482700359,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4650783147423685e-05,
+      "loss": 1.0010931396484375,
       "step": 465600
     },
     {
-      "epoch": 10.725799023671364,
-      "grad_norm": 3.5965654850006104,
-      "learning_rate": 1.4649692412850309e-05,
-      "loss": 0.5794,
+      "epoch": 5.362903167775074,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4643511583601169e-05,
+      "loss": 0.9757025146484375,
       "step": 465800
     },
     {
-      "epoch": 10.730404347425624,
-      "grad_norm": 3.169706106185913,
-      "learning_rate": 1.4642420849027791e-05,
-      "loss": 0.5607,
+      "epoch": 5.365205852849789,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.4636240019778655e-05,
+      "loss": 0.9824378204345703,
       "step": 466000
     },
     {
-      "epoch": 10.735009671179885,
-      "grad_norm": 3.449915647506714,
-      "learning_rate": 1.4635149285205276e-05,
-      "loss": 0.5568,
+      "epoch": 5.367508537924504,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4628968455956137e-05,
+      "loss": 0.9661509704589843,
       "step": 466200
     },
     {
-      "epoch": 10.739614994934144,
-      "grad_norm": 2.98413348197937,
-      "learning_rate": 1.4627877721382762e-05,
-      "loss": 0.5568,
+      "epoch": 5.3698112229992185,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4621696892133623e-05,
+      "loss": 0.967217025756836,
       "step": 466400
     },
     {
-      "epoch": 10.744220318688404,
-      "grad_norm": 3.158750534057617,
-      "learning_rate": 1.4620606157560244e-05,
-      "loss": 0.5663,
+      "epoch": 5.372113908073933,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.4614425328311107e-05,
+      "loss": 0.9927802276611328,
       "step": 466600
     },
     {
-      "epoch": 10.748825642442664,
-      "grad_norm": 3.3182640075683594,
-      "learning_rate": 1.461333459373773e-05,
-      "loss": 0.547,
+      "epoch": 5.3744165931486485,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.460715376448859e-05,
+      "loss": 0.9893026733398438,
       "step": 466800
     },
     {
-      "epoch": 10.753430966196923,
-      "grad_norm": 2.8168752193450928,
-      "learning_rate": 1.4606063029915214e-05,
-      "loss": 0.5632,
+      "epoch": 5.376719278223363,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4599882200666076e-05,
+      "loss": 0.9661732482910156,
       "step": 467000
     },
     {
-      "epoch": 10.758036289951184,
-      "grad_norm": 2.551912546157837,
-      "learning_rate": 1.4598791466092697e-05,
-      "loss": 0.564,
+      "epoch": 5.3790219632980785,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.459261063684356e-05,
+      "loss": 0.9615573120117188,
       "step": 467200
     },
     {
-      "epoch": 10.762641613705444,
-      "grad_norm": 3.1398086547851562,
-      "learning_rate": 1.4591519902270183e-05,
-      "loss": 0.5656,
+      "epoch": 5.381324648372793,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4585339073021046e-05,
+      "loss": 0.9748114776611329,
       "step": 467400
     },
     {
-      "epoch": 10.767246937459703,
-      "grad_norm": 2.806156873703003,
-      "learning_rate": 1.4584248338447667e-05,
-      "loss": 0.5712,
+      "epoch": 5.383627333447508,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4578067509198528e-05,
+      "loss": 0.9690747833251954,
       "step": 467600
     },
     {
-      "epoch": 10.771852261213963,
-      "grad_norm": 4.065160751342773,
-      "learning_rate": 1.4576976774625151e-05,
-      "loss": 0.5717,
+      "epoch": 5.385930018522223,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4570795945376013e-05,
+      "loss": 0.9754686737060547,
       "step": 467800
     },
     {
-      "epoch": 10.776457584968224,
-      "grad_norm": 2.9523870944976807,
-      "learning_rate": 1.4569705210802635e-05,
-      "loss": 0.5639,
+      "epoch": 5.388232703596938,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4563524381553499e-05,
+      "loss": 0.9902306365966796,
       "step": 468000
     },
     {
-      "epoch": 10.781062908722483,
-      "grad_norm": 2.7149343490600586,
-      "learning_rate": 1.456243364698012e-05,
-      "loss": 0.5587,
+      "epoch": 5.390535388671653,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4556252817730981e-05,
+      "loss": 0.9710675811767578,
       "step": 468200
     },
     {
-      "epoch": 10.785668232476743,
-      "grad_norm": 3.6111390590667725,
-      "learning_rate": 1.4555162083157606e-05,
-      "loss": 0.5648,
+      "epoch": 5.3928380737463675,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4548981253908465e-05,
+      "loss": 0.9841913604736328,
       "step": 468400
     },
     {
-      "epoch": 10.790273556231003,
-      "grad_norm": 3.606182098388672,
-      "learning_rate": 1.4547890519335088e-05,
-      "loss": 0.5726,
+      "epoch": 5.395140758821083,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4541709690085951e-05,
+      "loss": 0.9748066711425781,
       "step": 468600
     },
     {
-      "epoch": 10.794878879985262,
-      "grad_norm": 2.920095443725586,
-      "learning_rate": 1.4540618955512572e-05,
-      "loss": 0.5633,
+      "epoch": 5.397443443895797,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4534438126263434e-05,
+      "loss": 0.9825104522705078,
       "step": 468800
     },
     {
-      "epoch": 10.799484203739523,
-      "grad_norm": 3.206716299057007,
-      "learning_rate": 1.4533347391690058e-05,
-      "loss": 0.5613,
+      "epoch": 5.399746128970513,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.452716656244092e-05,
+      "loss": 0.9869859313964844,
       "step": 469000
     },
     {
-      "epoch": 10.804089527493783,
-      "grad_norm": 2.6802709102630615,
-      "learning_rate": 1.452607582786754e-05,
-      "loss": 0.5627,
+      "epoch": 5.402048814045227,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.4519894998618404e-05,
+      "loss": 0.9971955108642578,
       "step": 469200
     },
     {
-      "epoch": 10.808694851248042,
-      "grad_norm": 2.581068754196167,
-      "learning_rate": 1.4518804264045027e-05,
-      "loss": 0.5645,
+      "epoch": 5.404351499119943,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4512623434795886e-05,
+      "loss": 0.9768145751953125,
       "step": 469400
     },
     {
-      "epoch": 10.813300175002302,
-      "grad_norm": 2.9014952182769775,
-      "learning_rate": 1.451153270022251e-05,
-      "loss": 0.5717,
+      "epoch": 5.406654184194657,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.4505351870973372e-05,
+      "loss": 0.9668540954589844,
       "step": 469600
     },
     {
-      "epoch": 10.817905498756563,
-      "grad_norm": 3.2331104278564453,
-      "learning_rate": 1.4504261136399993e-05,
-      "loss": 0.5561,
+      "epoch": 5.408956869269373,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4498080307150857e-05,
+      "loss": 0.968299560546875,
       "step": 469800
     },
     {
-      "epoch": 10.822510822510823,
-      "grad_norm": 3.1601359844207764,
-      "learning_rate": 1.449698957257748e-05,
-      "loss": 0.5606,
+      "epoch": 5.411259554344087,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4490808743328339e-05,
+      "loss": 1.008473892211914,
       "step": 470000
     },
     {
-      "epoch": 10.827116146265082,
-      "grad_norm": 2.7113850116729736,
-      "learning_rate": 1.4489718008754963e-05,
-      "loss": 0.5678,
+      "epoch": 5.413562239418802,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4483537179505825e-05,
+      "loss": 0.9806656646728515,
       "step": 470200
     },
     {
-      "epoch": 10.831721470019342,
-      "grad_norm": 3.2414093017578125,
-      "learning_rate": 1.4482446444932446e-05,
-      "loss": 0.5614,
+      "epoch": 5.415864924493517,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.447626561568331e-05,
+      "loss": 0.9698346710205078,
       "step": 470400
     },
     {
-      "epoch": 10.836326793773603,
-      "grad_norm": 2.453719139099121,
-      "learning_rate": 1.4475174881109932e-05,
-      "loss": 0.5608,
+      "epoch": 5.418167609568232,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4468994051860795e-05,
+      "loss": 0.9919847869873046,
       "step": 470600
     },
     {
-      "epoch": 10.840932117527862,
-      "grad_norm": 3.3981575965881348,
-      "learning_rate": 1.4467903317287416e-05,
-      "loss": 0.5644,
+      "epoch": 5.420470294642947,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4461722488038278e-05,
+      "loss": 0.9748435211181641,
       "step": 470800
     },
     {
-      "epoch": 10.845537441282122,
-      "grad_norm": 3.9584078788757324,
-      "learning_rate": 1.4460631753464902e-05,
-      "loss": 0.5649,
+      "epoch": 5.422772979717662,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4454450924215762e-05,
+      "loss": 0.9667948150634765,
       "step": 471000
     },
     {
-      "epoch": 10.850142765036383,
-      "grad_norm": 2.609250545501709,
-      "learning_rate": 1.4453360189642385e-05,
-      "loss": 0.5737,
+      "epoch": 5.425075664792377,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4447179360393248e-05,
+      "loss": 1.006150360107422,
       "step": 471200
     },
     {
-      "epoch": 10.854748088790641,
-      "grad_norm": 2.888627767562866,
-      "learning_rate": 1.4446088625819869e-05,
-      "loss": 0.5763,
+      "epoch": 5.427378349867092,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.443990779657073e-05,
+      "loss": 0.9815165710449218,
       "step": 471400
     },
     {
-      "epoch": 10.859353412544902,
-      "grad_norm": 3.1995785236358643,
-      "learning_rate": 1.4438817061997355e-05,
-      "loss": 0.5639,
+      "epoch": 5.429681034941807,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4432636232748215e-05,
+      "loss": 0.9920414733886719,
       "step": 471600
     },
     {
-      "epoch": 10.863958736299162,
-      "grad_norm": 3.3186826705932617,
-      "learning_rate": 1.4431545498174837e-05,
-      "loss": 0.5643,
+      "epoch": 5.431983720016522,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.44253646689257e-05,
+      "loss": 0.9791954040527344,
       "step": 471800
     },
     {
-      "epoch": 10.868564060053421,
-      "grad_norm": 3.200171709060669,
-      "learning_rate": 1.4424310292171434e-05,
-      "loss": 0.5615,
+      "epoch": 5.434286405091237,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4418093105103183e-05,
+      "loss": 0.9826209259033203,
       "step": 472000
     },
     {
-      "epoch": 10.873169383807681,
-      "grad_norm": 3.0748026371002197,
-      "learning_rate": 1.4417038728348918e-05,
-      "loss": 0.5585,
+      "epoch": 5.436589090165952,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4410821541280669e-05,
+      "loss": 0.9753388214111328,
       "step": 472200
     },
     {
-      "epoch": 10.877774707561942,
-      "grad_norm": 2.5233187675476074,
-      "learning_rate": 1.4409767164526404e-05,
-      "loss": 0.5693,
+      "epoch": 5.438891775240666,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4403549977458153e-05,
+      "loss": 0.9879838562011719,
       "step": 472400
     },
     {
-      "epoch": 10.882380031316202,
-      "grad_norm": 3.0431416034698486,
-      "learning_rate": 1.4402531958523001e-05,
-      "loss": 0.5627,
+      "epoch": 5.441194460315382,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.4396278413635636e-05,
+      "loss": 0.9852864837646484,
       "step": 472600
     },
     {
-      "epoch": 10.886985355070461,
-      "grad_norm": 2.8230044841766357,
-      "learning_rate": 1.4395260394700484e-05,
-      "loss": 0.5725,
+      "epoch": 5.443497145390096,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4389006849813122e-05,
+      "loss": 0.9750405883789063,
       "step": 472800
     },
     {
-      "epoch": 10.891590678824722,
-      "grad_norm": 2.9966495037078857,
-      "learning_rate": 1.438798883087797e-05,
-      "loss": 0.566,
+      "epoch": 5.445799830464812,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4381735285990606e-05,
+      "loss": 0.9863434600830078,
       "step": 473000
     },
     {
-      "epoch": 10.896196002578982,
-      "grad_norm": 3.4623706340789795,
-      "learning_rate": 1.4380717267055454e-05,
-      "loss": 0.5664,
+      "epoch": 5.448102515539526,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.437446372216809e-05,
+      "loss": 0.9665998077392578,
       "step": 473200
     },
     {
-      "epoch": 10.90080132633324,
-      "grad_norm": 2.6742422580718994,
-      "learning_rate": 1.4373445703232936e-05,
-      "loss": 0.5632,
+      "epoch": 5.450405200614242,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4367192158345574e-05,
+      "loss": 0.9747258758544922,
       "step": 473400
     },
     {
-      "epoch": 10.905406650087501,
-      "grad_norm": 3.300477981567383,
-      "learning_rate": 1.4366174139410422e-05,
-      "loss": 0.5797,
+      "epoch": 5.452707885688956,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4359920594523058e-05,
+      "loss": 0.9988768005371094,
       "step": 473600
     },
     {
-      "epoch": 10.910011973841762,
-      "grad_norm": 2.9274837970733643,
-      "learning_rate": 1.4358902575587907e-05,
-      "loss": 0.5604,
+      "epoch": 5.455010570763671,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4352649030700543e-05,
+      "loss": 0.9905022430419922,
       "step": 473800
     },
     {
-      "epoch": 10.91461729759602,
-      "grad_norm": 2.8111343383789062,
-      "learning_rate": 1.4351631011765389e-05,
-      "loss": 0.5643,
+      "epoch": 5.457313255838386,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4345377466878027e-05,
+      "loss": 0.9991416168212891,
       "step": 474000
     },
     {
-      "epoch": 10.919222621350281,
-      "grad_norm": 2.678849458694458,
-      "learning_rate": 1.4344359447942875e-05,
-      "loss": 0.5646,
+      "epoch": 5.459615940913101,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4338105903055511e-05,
+      "loss": 0.9977909851074219,
       "step": 474200
     },
     {
-      "epoch": 10.923827945104541,
-      "grad_norm": 2.9185190200805664,
-      "learning_rate": 1.433708788412036e-05,
-      "loss": 0.5691,
+      "epoch": 5.461918625987816,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4330834339232997e-05,
+      "loss": 0.9725360107421875,
       "step": 474400
     },
     {
-      "epoch": 10.9284332688588,
-      "grad_norm": 3.45511794090271,
-      "learning_rate": 1.4329852678116956e-05,
-      "loss": 0.568,
+      "epoch": 5.464221311062531,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.432356277541048e-05,
+      "loss": 0.9939174652099609,
       "step": 474600
     },
     {
-      "epoch": 10.93303859261306,
-      "grad_norm": 3.163696527481079,
-      "learning_rate": 1.4322617472113553e-05,
-      "loss": 0.551,
+      "epoch": 5.466523996137246,
+      "grad_norm": 1.0,
+      "learning_rate": 1.4316291211587965e-05,
+      "loss": 0.9697264862060547,
       "step": 474800
     },
     {
-      "epoch": 10.937643916367321,
-      "grad_norm": 3.591355323791504,
-      "learning_rate": 1.4315345908291039e-05,
-      "loss": 0.5525,
+      "epoch": 5.468826681211961,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.430901964776545e-05,
+      "loss": 0.9911696624755859,
       "step": 475000
     },
     {
-      "epoch": 10.94224924012158,
-      "grad_norm": 3.392062187194824,
-      "learning_rate": 1.4308074344468521e-05,
-      "loss": 0.562,
+      "epoch": 5.471129366286676,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4301748083942932e-05,
+      "loss": 0.9740504455566407,
       "step": 475200
     },
     {
-      "epoch": 10.94685456387584,
-      "grad_norm": 3.2929904460906982,
-      "learning_rate": 1.4300802780646006e-05,
-      "loss": 0.5711,
+      "epoch": 5.4734320513613905,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4294476520120418e-05,
+      "loss": 0.948077163696289,
       "step": 475400
     },
     {
-      "epoch": 10.9514598876301,
-      "grad_norm": 3.0882444381713867,
-      "learning_rate": 1.4293531216823492e-05,
-      "loss": 0.5573,
+      "epoch": 5.475734736436106,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4287204956297902e-05,
+      "loss": 0.9827384185791016,
       "step": 475600
     },
     {
-      "epoch": 10.95606521138436,
-      "grad_norm": 3.073636054992676,
-      "learning_rate": 1.4286259653000974e-05,
-      "loss": 0.5738,
+      "epoch": 5.4780374215108205,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4279933392475385e-05,
+      "loss": 0.9965239715576172,
       "step": 475800
     },
     {
-      "epoch": 10.96067053513862,
-      "grad_norm": 3.676265001296997,
-      "learning_rate": 1.4278988089178458e-05,
-      "loss": 0.5551,
+      "epoch": 5.480340106585535,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.427266182865287e-05,
+      "loss": 0.980682373046875,
       "step": 476000
     },
     {
-      "epoch": 10.96527585889288,
-      "grad_norm": 3.5196962356567383,
-      "learning_rate": 1.4271716525355944e-05,
-      "loss": 0.565,
+      "epoch": 5.4826427916602505,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.4265390264830355e-05,
+      "loss": 0.9934986877441406,
       "step": 476200
     },
     {
-      "epoch": 10.96988118264714,
-      "grad_norm": 3.1859378814697266,
-      "learning_rate": 1.4264444961533427e-05,
-      "loss": 0.565,
+      "epoch": 5.484945476734965,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.425811870100784e-05,
+      "loss": 0.979002456665039,
       "step": 476400
     },
     {
-      "epoch": 10.9744865064014,
-      "grad_norm": 2.839648962020874,
-      "learning_rate": 1.4257173397710913e-05,
-      "loss": 0.5656,
+      "epoch": 5.4872481618096804,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4250847137185323e-05,
+      "loss": 0.9670918273925782,
       "step": 476600
     },
     {
-      "epoch": 10.97909183015566,
-      "grad_norm": 3.086836338043213,
-      "learning_rate": 1.4249901833888397e-05,
-      "loss": 0.5653,
+      "epoch": 5.489550846884395,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4243575573362808e-05,
+      "loss": 0.9969381713867187,
       "step": 476800
     },
     {
-      "epoch": 10.98369715390992,
-      "grad_norm": 3.1683554649353027,
-      "learning_rate": 1.424263027006588e-05,
-      "loss": 0.5679,
+      "epoch": 5.49185353195911,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4236304009540292e-05,
+      "loss": 0.9819905090332032,
       "step": 477000
     },
     {
-      "epoch": 10.98830247766418,
-      "grad_norm": 3.869631052017212,
-      "learning_rate": 1.4235358706243365e-05,
-      "loss": 0.5605,
+      "epoch": 5.494156217033825,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4229032445717776e-05,
+      "loss": 0.9679257202148438,
       "step": 477200
     },
     {
-      "epoch": 10.99290780141844,
-      "grad_norm": 3.0082695484161377,
-      "learning_rate": 1.422808714242085e-05,
-      "loss": 0.5558,
+      "epoch": 5.4964589021085395,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.422176088189526e-05,
+      "loss": 0.9745060729980469,
       "step": 477400
     },
     {
-      "epoch": 10.9975131251727,
-      "grad_norm": 2.398848295211792,
-      "learning_rate": 1.4220815578598334e-05,
-      "loss": 0.5681,
+      "epoch": 5.498761587183255,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4214489318072746e-05,
+      "loss": 0.9809858703613281,
       "step": 477600
     },
     {
-      "epoch": 11.0,
-      "eval_loss": 0.5423869490623474,
-      "eval_runtime": 166.7579,
-      "eval_samples_per_second": 170.073,
-      "eval_steps_per_second": 10.632,
-      "step": 477708
-    },
-    {
-      "epoch": 11.002118448926959,
-      "grad_norm": 3.4873993396759033,
-      "learning_rate": 1.4213544014775818e-05,
-      "loss": 0.5714,
+      "epoch": 5.5010642722579695,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4207217754250229e-05,
+      "loss": 0.9731687927246093,
       "step": 477800
     },
     {
-      "epoch": 11.00672377268122,
-      "grad_norm": 2.91593337059021,
-      "learning_rate": 1.4206272450953302e-05,
-      "loss": 0.558,
+      "epoch": 5.503366957332685,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4199946190427715e-05,
+      "loss": 0.9807541656494141,
       "step": 478000
     },
     {
-      "epoch": 11.01132909643548,
-      "grad_norm": 3.278059959411621,
-      "learning_rate": 1.4199037244949899e-05,
-      "loss": 0.5672,
+      "epoch": 5.505669642407399,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4192674626605199e-05,
+      "loss": 1.0058878326416016,
       "step": 478200
     },
     {
-      "epoch": 11.015934420189739,
-      "grad_norm": 2.8523497581481934,
-      "learning_rate": 1.4191765681127385e-05,
-      "loss": 0.5712,
+      "epoch": 5.507972327482115,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4185403062782681e-05,
+      "loss": 1.0039275360107422,
       "step": 478400
     },
     {
-      "epoch": 11.020539743944,
-      "grad_norm": 2.9555845260620117,
-      "learning_rate": 1.4184494117304868e-05,
-      "loss": 0.5571,
+      "epoch": 5.510275012556829,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4178131498960167e-05,
+      "loss": 0.9918733978271485,
       "step": 478600
     },
     {
-      "epoch": 11.02514506769826,
-      "grad_norm": 3.0521042346954346,
-      "learning_rate": 1.4177222553482352e-05,
-      "loss": 0.5565,
+      "epoch": 5.512577697631545,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4170859935137651e-05,
+      "loss": 0.9834480285644531,
       "step": 478800
     },
     {
-      "epoch": 11.029750391452518,
-      "grad_norm": 3.1665375232696533,
-      "learning_rate": 1.4169950989659838e-05,
-      "loss": 0.5573,
+      "epoch": 5.514880382706259,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4163588371315134e-05,
+      "loss": 1.0038261413574219,
       "step": 479000
     },
     {
-      "epoch": 11.034355715206779,
-      "grad_norm": 2.929694890975952,
-      "learning_rate": 1.416267942583732e-05,
-      "loss": 0.5592,
+      "epoch": 5.517183067780975,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.415631680749262e-05,
+      "loss": 0.9893887329101563,
       "step": 479200
     },
     {
-      "epoch": 11.03896103896104,
-      "grad_norm": 3.597895860671997,
-      "learning_rate": 1.4155407862014806e-05,
-      "loss": 0.5658,
+      "epoch": 5.519485752855689,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4149045243670104e-05,
+      "loss": 0.9794209289550782,
       "step": 479400
     },
     {
-      "epoch": 11.043566362715298,
-      "grad_norm": 2.8076977729797363,
-      "learning_rate": 1.414813629819229e-05,
-      "loss": 0.554,
+      "epoch": 5.521788437930404,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4141773679847588e-05,
+      "loss": 0.96505126953125,
       "step": 479600
     },
     {
-      "epoch": 11.048171686469558,
-      "grad_norm": 3.267301559448242,
-      "learning_rate": 1.4140864734369773e-05,
-      "loss": 0.568,
+      "epoch": 5.524091123005119,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4134502116025073e-05,
+      "loss": 0.9843000030517578,
       "step": 479800
     },
     {
-      "epoch": 11.052777010223819,
-      "grad_norm": 3.1478617191314697,
-      "learning_rate": 1.4133593170547259e-05,
-      "loss": 0.5702,
+      "epoch": 5.526393808079834,
+      "grad_norm": 1.125,
+      "learning_rate": 1.4127230552202557e-05,
+      "loss": 0.9835649871826172,
       "step": 480000
     },
     {
-      "epoch": 11.05738233397808,
-      "grad_norm": 2.9337422847747803,
-      "learning_rate": 1.4126321606724743e-05,
-      "loss": 0.5565,
+      "epoch": 5.528696493154549,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4119958988380041e-05,
+      "loss": 1.0410501098632812,
       "step": 480200
     },
     {
-      "epoch": 11.061987657732338,
-      "grad_norm": 2.9447715282440186,
-      "learning_rate": 1.4119050042902226e-05,
-      "loss": 0.5574,
+      "epoch": 5.530999178229264,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4112687424557525e-05,
+      "loss": 0.9889697265625,
       "step": 480400
     },
     {
-      "epoch": 11.066592981486599,
-      "grad_norm": 3.1373238563537598,
-      "learning_rate": 1.4111778479079711e-05,
-      "loss": 0.5539,
+      "epoch": 5.533301863303979,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4105415860735011e-05,
+      "loss": 0.9981174468994141,
       "step": 480600
     },
     {
-      "epoch": 11.07119830524086,
-      "grad_norm": 3.812917947769165,
-      "learning_rate": 1.4104506915257196e-05,
-      "loss": 0.5531,
+      "epoch": 5.535604548378694,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4098144296912495e-05,
+      "loss": 0.9815375518798828,
       "step": 480800
     },
     {
-      "epoch": 11.075803628995118,
-      "grad_norm": 2.356658935546875,
-      "learning_rate": 1.409723535143468e-05,
-      "loss": 0.5652,
+      "epoch": 5.537907233453408,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4090872733089978e-05,
+      "loss": 0.977080078125,
       "step": 481000
     },
     {
-      "epoch": 11.080408952749378,
-      "grad_norm": 3.141037940979004,
-      "learning_rate": 1.4089963787612164e-05,
-      "loss": 0.552,
+      "epoch": 5.540209918528124,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4083601169267464e-05,
+      "loss": 0.9870880126953125,
       "step": 481200
     },
     {
-      "epoch": 11.085014276503639,
-      "grad_norm": 3.6183485984802246,
-      "learning_rate": 1.4082692223789648e-05,
-      "loss": 0.5601,
+      "epoch": 5.542512603602838,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4076329605444948e-05,
+      "loss": 0.9716754150390625,
       "step": 481400
     },
     {
-      "epoch": 11.089619600257898,
-      "grad_norm": 2.416588306427002,
-      "learning_rate": 1.4075420659967134e-05,
-      "loss": 0.5571,
+      "epoch": 5.544815288677554,
+      "grad_norm": 0.75,
+      "learning_rate": 1.406905804162243e-05,
+      "loss": 0.986975326538086,
       "step": 481600
     },
     {
-      "epoch": 11.094224924012158,
-      "grad_norm": 3.343844175338745,
-      "learning_rate": 1.4068149096144617e-05,
-      "loss": 0.5558,
+      "epoch": 5.547117973752268,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.4061786477799916e-05,
+      "loss": 0.9707216644287109,
       "step": 481800
     },
     {
-      "epoch": 11.098830247766418,
-      "grad_norm": 2.7808258533477783,
-      "learning_rate": 1.4060877532322101e-05,
-      "loss": 0.5705,
+      "epoch": 5.549420658826984,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.40545149139774e-05,
+      "loss": 0.9838639068603515,
       "step": 482000
     },
     {
-      "epoch": 11.103435571520677,
-      "grad_norm": 2.9986512660980225,
-      "learning_rate": 1.4053605968499587e-05,
-      "loss": 0.5589,
+      "epoch": 5.551723343901698,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4047243350154885e-05,
+      "loss": 0.978685302734375,
       "step": 482200
     },
     {
-      "epoch": 11.108040895274938,
-      "grad_norm": 3.053314685821533,
-      "learning_rate": 1.404633440467707e-05,
-      "loss": 0.5586,
+      "epoch": 5.554026028976414,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4039971786332369e-05,
+      "loss": 1.0009799957275392,
       "step": 482400
     },
     {
-      "epoch": 11.112646219029198,
-      "grad_norm": 2.5446858406066895,
-      "learning_rate": 1.4039062840854555e-05,
-      "loss": 0.5498,
+      "epoch": 5.556328714051128,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4032700222509853e-05,
+      "loss": 0.9997296905517579,
       "step": 482600
     },
     {
-      "epoch": 11.117251542783457,
-      "grad_norm": 3.446873664855957,
-      "learning_rate": 1.403179127703204e-05,
-      "loss": 0.5492,
+      "epoch": 5.558631399125844,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4025428658687338e-05,
+      "loss": 0.9815716552734375,
       "step": 482800
     },
     {
-      "epoch": 11.121856866537717,
-      "grad_norm": 2.8828375339508057,
-      "learning_rate": 1.4024519713209522e-05,
-      "loss": 0.5557,
+      "epoch": 5.560934084200558,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4018157094864822e-05,
+      "loss": 0.9789281463623047,
       "step": 483000
     },
     {
-      "epoch": 11.126462190291978,
-      "grad_norm": 2.941978693008423,
-      "learning_rate": 1.4017248149387008e-05,
-      "loss": 0.5594,
+      "epoch": 5.563236769275273,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4010885531042306e-05,
+      "loss": 0.9825881958007813,
       "step": 483200
     },
     {
-      "epoch": 11.131067514046238,
-      "grad_norm": 3.2115888595581055,
-      "learning_rate": 1.4009976585564492e-05,
-      "loss": 0.5628,
+      "epoch": 5.565539454349988,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.400361396721979e-05,
+      "loss": 0.9777683258056641,
       "step": 483400
     },
     {
-      "epoch": 11.135672837800497,
-      "grad_norm": 3.323775291442871,
-      "learning_rate": 1.4002705021741975e-05,
-      "loss": 0.5574,
+      "epoch": 5.567842139424703,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3996342403397274e-05,
+      "loss": 0.9617288208007813,
       "step": 483600
     },
     {
-      "epoch": 11.140278161554757,
-      "grad_norm": 3.129638195037842,
-      "learning_rate": 1.399543345791946e-05,
-      "loss": 0.5684,
+      "epoch": 5.570144824499418,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.398907083957476e-05,
+      "loss": 0.9682086944580078,
       "step": 483800
     },
     {
-      "epoch": 11.144883485309018,
-      "grad_norm": 3.986924886703491,
-      "learning_rate": 1.3988161894096945e-05,
-      "loss": 0.5628,
+      "epoch": 5.572447509574133,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3981799275752243e-05,
+      "loss": 0.9663679504394531,
       "step": 484000
     },
     {
-      "epoch": 11.149488809063277,
-      "grad_norm": 3.6282601356506348,
-      "learning_rate": 1.3980890330274429e-05,
-      "loss": 0.563,
+      "epoch": 5.574750194648848,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3974527711929727e-05,
+      "loss": 0.9882929992675781,
       "step": 484200
     },
     {
-      "epoch": 11.154094132817537,
-      "grad_norm": 3.6448795795440674,
-      "learning_rate": 1.3973618766451913e-05,
-      "loss": 0.5604,
+      "epoch": 5.577052879723563,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3967256148107213e-05,
+      "loss": 0.9890274810791015,
       "step": 484400
     },
     {
-      "epoch": 11.158699456571798,
-      "grad_norm": 4.1878662109375,
-      "learning_rate": 1.3966347202629397e-05,
-      "loss": 0.5612,
+      "epoch": 5.579355564798277,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3959984584284697e-05,
+      "loss": 0.995350341796875,
       "step": 484600
     },
     {
-      "epoch": 11.163304780326056,
-      "grad_norm": 3.1330618858337402,
-      "learning_rate": 1.3959075638806883e-05,
-      "loss": 0.5677,
+      "epoch": 5.5816582498729925,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.395271302046218e-05,
+      "loss": 0.9710835266113281,
       "step": 484800
     },
     {
-      "epoch": 11.167910104080317,
-      "grad_norm": 3.0046327114105225,
-      "learning_rate": 1.3951804074984366e-05,
-      "loss": 0.5619,
+      "epoch": 5.583960934947708,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3945441456639666e-05,
+      "loss": 0.9732130432128906,
       "step": 485000
     },
     {
-      "epoch": 11.172515427834577,
-      "grad_norm": 2.964937925338745,
-      "learning_rate": 1.3944532511161852e-05,
-      "loss": 0.5613,
+      "epoch": 5.5862636200224225,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.393816989281715e-05,
+      "loss": 0.9844066619873046,
       "step": 485200
     },
     {
-      "epoch": 11.177120751588836,
-      "grad_norm": 3.570549249649048,
-      "learning_rate": 1.3937260947339336e-05,
-      "loss": 0.556,
+      "epoch": 5.588566305097137,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3930898328994634e-05,
+      "loss": 0.9600486755371094,
       "step": 485400
     },
     {
-      "epoch": 11.181726075343096,
-      "grad_norm": 2.8079946041107178,
-      "learning_rate": 1.3930025741335933e-05,
-      "loss": 0.5513,
+      "epoch": 5.5908689901718525,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3923626765172118e-05,
+      "loss": 0.9797865295410156,
       "step": 485600
     },
     {
-      "epoch": 11.186331399097357,
-      "grad_norm": 2.9145305156707764,
-      "learning_rate": 1.3922754177513416e-05,
-      "loss": 0.5556,
+      "epoch": 5.593171675246567,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3916355201349603e-05,
+      "loss": 0.9919526672363281,
       "step": 485800
     },
     {
-      "epoch": 11.190936722851616,
-      "grad_norm": 3.74453067779541,
-      "learning_rate": 1.3915482613690901e-05,
-      "loss": 0.5637,
+      "epoch": 5.595474360321282,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3909083637527087e-05,
+      "loss": 0.9721823120117188,
       "step": 486000
     },
     {
-      "epoch": 11.195542046605876,
-      "grad_norm": 3.717133045196533,
-      "learning_rate": 1.3908211049868386e-05,
-      "loss": 0.5556,
+      "epoch": 5.597777045395997,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3901812073704571e-05,
+      "loss": 0.9831861114501953,
       "step": 486200
     },
     {
-      "epoch": 11.200147370360137,
-      "grad_norm": 3.242644786834717,
-      "learning_rate": 1.3900939486045868e-05,
-      "loss": 0.5659,
+      "epoch": 5.600079730470712,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3894540509882055e-05,
+      "loss": 0.9493144226074218,
       "step": 486400
     },
     {
-      "epoch": 11.204752694114397,
-      "grad_norm": 3.173064708709717,
-      "learning_rate": 1.3893667922223354e-05,
-      "loss": 0.5541,
+      "epoch": 5.602382415545427,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.388726894605954e-05,
+      "loss": 0.9714206695556641,
       "step": 486600
     },
     {
-      "epoch": 11.209358017868656,
-      "grad_norm": 2.6633315086364746,
-      "learning_rate": 1.3886396358400838e-05,
-      "loss": 0.575,
+      "epoch": 5.6046851006201415,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3879997382237024e-05,
+      "loss": 0.9733635711669922,
       "step": 486800
     },
     {
-      "epoch": 11.213963341622916,
-      "grad_norm": 3.097576856613159,
-      "learning_rate": 1.3879124794578323e-05,
-      "loss": 0.5574,
+      "epoch": 5.606987785694857,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.387272581841451e-05,
+      "loss": 0.9818306732177734,
       "step": 487000
     },
     {
-      "epoch": 11.218568665377177,
-      "grad_norm": 3.0718281269073486,
-      "learning_rate": 1.3871853230755807e-05,
-      "loss": 0.5525,
+      "epoch": 5.609290470769571,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.3865454254591992e-05,
+      "loss": 0.9948311614990234,
       "step": 487200
     },
     {
-      "epoch": 11.223173989131435,
-      "grad_norm": 4.379641056060791,
-      "learning_rate": 1.3864581666933291e-05,
-      "loss": 0.5663,
+      "epoch": 5.611593155844287,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3858182690769476e-05,
+      "loss": 0.9782450866699218,
       "step": 487400
     },
     {
-      "epoch": 11.227779312885696,
-      "grad_norm": 2.682002305984497,
-      "learning_rate": 1.3857310103110775e-05,
-      "loss": 0.5677,
+      "epoch": 5.613895840919001,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3850911126946962e-05,
+      "loss": 1.0036625671386719,
       "step": 487600
     },
     {
-      "epoch": 11.232384636639956,
-      "grad_norm": 3.9355664253234863,
-      "learning_rate": 1.385003853928826e-05,
-      "loss": 0.5645,
+      "epoch": 5.616198525993717,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3843639563124446e-05,
+      "loss": 0.9743613433837891,
       "step": 487800
     },
     {
-      "epoch": 11.236989960394215,
-      "grad_norm": 3.0530924797058105,
-      "learning_rate": 1.3842766975465744e-05,
-      "loss": 0.5584,
+      "epoch": 5.618501211068431,
+      "grad_norm": 0.75,
+      "learning_rate": 1.383636799930193e-05,
+      "loss": 0.9842996978759766,
       "step": 488000
     },
     {
-      "epoch": 11.241595284148476,
-      "grad_norm": 3.311056613922119,
-      "learning_rate": 1.383549541164323e-05,
-      "loss": 0.5569,
+      "epoch": 5.620803896143147,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3829096435479415e-05,
+      "loss": 0.9787957000732422,
       "step": 488200
     },
     {
-      "epoch": 11.246200607902736,
-      "grad_norm": 3.7261691093444824,
-      "learning_rate": 1.3828260205639826e-05,
-      "loss": 0.5502,
+      "epoch": 5.623106581217861,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3821824871656899e-05,
+      "loss": 0.9686898040771484,
       "step": 488400
     },
     {
-      "epoch": 11.250805931656995,
-      "grad_norm": 2.7717695236206055,
-      "learning_rate": 1.3820988641817309e-05,
-      "loss": 0.566,
+      "epoch": 5.625409266292577,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3814553307834383e-05,
+      "loss": 0.9968467712402344,
       "step": 488600
     },
     {
-      "epoch": 11.255411255411255,
-      "grad_norm": 2.931807041168213,
-      "learning_rate": 1.3813717077994795e-05,
-      "loss": 0.5555,
+      "epoch": 5.627711951367291,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3807281744011868e-05,
+      "loss": 0.9966775512695313,
       "step": 488800
     },
     {
-      "epoch": 11.260016579165516,
-      "grad_norm": 3.2359492778778076,
-      "learning_rate": 1.3806445514172279e-05,
-      "loss": 0.5608,
+      "epoch": 5.630014636442006,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3800010180189352e-05,
+      "loss": 1.007319107055664,
       "step": 489000
     },
     {
-      "epoch": 11.264621902919774,
-      "grad_norm": 2.8134891986846924,
-      "learning_rate": 1.3799173950349762e-05,
-      "loss": 0.5694,
+      "epoch": 5.632317321516721,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3792738616366836e-05,
+      "loss": 1.000706024169922,
       "step": 489200
     },
     {
-      "epoch": 11.269227226674035,
-      "grad_norm": 3.166933059692383,
-      "learning_rate": 1.3791902386527248e-05,
-      "loss": 0.554,
+      "epoch": 5.634620006591436,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.378546705254432e-05,
+      "loss": 0.9947959899902343,
       "step": 489400
     },
     {
-      "epoch": 11.273832550428295,
-      "grad_norm": 3.2484281063079834,
-      "learning_rate": 1.3784630822704732e-05,
-      "loss": 0.5577,
+      "epoch": 5.636922691666151,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3778195488721806e-05,
+      "loss": 0.9884097290039062,
       "step": 489600
     },
     {
-      "epoch": 11.278437874182554,
-      "grad_norm": 3.1392292976379395,
-      "learning_rate": 1.3777359258882214e-05,
-      "loss": 0.5552,
+      "epoch": 5.639225376740866,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3770923924899289e-05,
+      "loss": 0.9712710571289063,
       "step": 489800
     },
     {
-      "epoch": 11.283043197936815,
-      "grad_norm": 3.2946557998657227,
-      "learning_rate": 1.37700876950597e-05,
-      "loss": 0.5578,
+      "epoch": 5.641528061815581,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3763652361076773e-05,
+      "loss": 0.9973468017578125,
       "step": 490000
     },
     {
-      "epoch": 11.287648521691075,
-      "grad_norm": 3.325127124786377,
-      "learning_rate": 1.3762816131237184e-05,
-      "loss": 0.546,
+      "epoch": 5.643830746890296,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3756380797254259e-05,
+      "loss": 0.9909369659423828,
       "step": 490200
     },
     {
-      "epoch": 11.292253845445336,
-      "grad_norm": 2.8645405769348145,
-      "learning_rate": 1.3755544567414669e-05,
-      "loss": 0.5588,
+      "epoch": 5.64613343196501,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3749109233431741e-05,
+      "loss": 0.9876738739013672,
       "step": 490400
     },
     {
-      "epoch": 11.296859169199594,
-      "grad_norm": 4.306154727935791,
-      "learning_rate": 1.3748273003592153e-05,
-      "loss": 0.5669,
+      "epoch": 5.648436117039726,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3741837669609226e-05,
+      "loss": 0.999638671875,
       "step": 490600
     },
     {
-      "epoch": 11.301464492953855,
-      "grad_norm": 3.181588649749756,
-      "learning_rate": 1.3741001439769637e-05,
-      "loss": 0.5579,
+      "epoch": 5.65073880211444,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3734566105786711e-05,
+      "loss": 0.9692737579345703,
       "step": 490800
     },
     {
-      "epoch": 11.306069816708115,
-      "grad_norm": 2.690722942352295,
-      "learning_rate": 1.3733729875947121e-05,
-      "loss": 0.5648,
+      "epoch": 5.653041487189156,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3727294541964196e-05,
+      "loss": 0.9833884429931641,
       "step": 491000
     },
     {
-      "epoch": 11.310675140462374,
-      "grad_norm": 3.0725932121276855,
-      "learning_rate": 1.3726458312124606e-05,
-      "loss": 0.552,
+      "epoch": 5.65534417226387,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.372002297814168e-05,
+      "loss": 1.0041567230224608,
       "step": 491200
     },
     {
-      "epoch": 11.315280464216634,
-      "grad_norm": 3.268697500228882,
-      "learning_rate": 1.371918674830209e-05,
-      "loss": 0.5616,
+      "epoch": 5.657646857338586,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3712751414319164e-05,
+      "loss": 0.9968643951416015,
       "step": 491400
     },
     {
-      "epoch": 11.319885787970895,
-      "grad_norm": 3.456531524658203,
-      "learning_rate": 1.3711915184479574e-05,
-      "loss": 0.5647,
+      "epoch": 5.6599495424133,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3705479850496648e-05,
+      "loss": 0.9939844512939453,
       "step": 491600
     },
     {
-      "epoch": 11.324491111725154,
-      "grad_norm": 2.9139323234558105,
-      "learning_rate": 1.3704643620657058e-05,
-      "loss": 0.5547,
+      "epoch": 5.662252227488016,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3698208286674133e-05,
+      "loss": 0.9795943450927734,
       "step": 491800
     },
     {
-      "epoch": 11.329096435479414,
-      "grad_norm": 3.0424976348876953,
-      "learning_rate": 1.3697372056834544e-05,
-      "loss": 0.5564,
+      "epoch": 5.66455491256273,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.3690936722851617e-05,
+      "loss": 0.9953731536865235,
       "step": 492000
     },
     {
-      "epoch": 11.333701759233675,
-      "grad_norm": 3.656421661376953,
-      "learning_rate": 1.3690100493012028e-05,
-      "loss": 0.5547,
+      "epoch": 5.666857597637446,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3683665159029101e-05,
+      "loss": 0.9865888214111328,
       "step": 492200
     },
     {
-      "epoch": 11.338307082987933,
-      "grad_norm": 2.804675340652466,
-      "learning_rate": 1.368282892918951e-05,
-      "loss": 0.5504,
+      "epoch": 5.66916028271216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3676393595206585e-05,
+      "loss": 0.9924819946289063,
       "step": 492400
     },
     {
-      "epoch": 11.342912406742194,
-      "grad_norm": 2.907458543777466,
-      "learning_rate": 1.3675557365366997e-05,
-      "loss": 0.5557,
+      "epoch": 5.671462967786875,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.366912203138407e-05,
+      "loss": 0.99431396484375,
       "step": 492600
     },
     {
-      "epoch": 11.347517730496454,
-      "grad_norm": 3.368783712387085,
-      "learning_rate": 1.3668322159363594e-05,
-      "loss": 0.5629,
+      "epoch": 5.67376565286159,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3661850467561555e-05,
+      "loss": 0.9882295989990234,
       "step": 492800
     },
     {
-      "epoch": 11.352123054250713,
-      "grad_norm": 2.7372817993164062,
-      "learning_rate": 1.366108695336019e-05,
-      "loss": 0.5667,
+      "epoch": 5.676068337936305,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.3654578903739038e-05,
+      "loss": 0.9925931549072265,
       "step": 493000
     },
     {
-      "epoch": 11.356728378004973,
-      "grad_norm": 3.26176381111145,
-      "learning_rate": 1.3653815389537675e-05,
-      "loss": 0.5581,
+      "epoch": 5.67837102301102,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3647307339916522e-05,
+      "loss": 0.98350830078125,
       "step": 493200
     },
     {
-      "epoch": 11.361333701759234,
-      "grad_norm": 3.8855667114257812,
-      "learning_rate": 1.3646543825715157e-05,
-      "loss": 0.5589,
+      "epoch": 5.680673708085735,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3640035776094008e-05,
+      "loss": 0.9921775054931641,
       "step": 493400
     },
     {
-      "epoch": 11.365939025513494,
-      "grad_norm": 2.7255947589874268,
-      "learning_rate": 1.3639272261892643e-05,
-      "loss": 0.5483,
+      "epoch": 5.68297639316045,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.363276421227149e-05,
+      "loss": 0.9786495971679687,
       "step": 493600
     },
     {
-      "epoch": 11.370544349267753,
-      "grad_norm": 3.3710615634918213,
-      "learning_rate": 1.3632000698070127e-05,
-      "loss": 0.5443,
+      "epoch": 5.6852790782351645,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3625492648448976e-05,
+      "loss": 0.9776170349121094,
       "step": 493800
     },
     {
-      "epoch": 11.375149673022014,
-      "grad_norm": 3.4504647254943848,
-      "learning_rate": 1.3624729134247612e-05,
-      "loss": 0.5645,
+      "epoch": 5.687581763309879,
+      "grad_norm": 1.0,
+      "learning_rate": 1.361822108462646e-05,
+      "loss": 0.9838382720947265,
       "step": 494000
     },
     {
-      "epoch": 11.379754996776274,
-      "grad_norm": 2.8603811264038086,
-      "learning_rate": 1.3617493928244209e-05,
-      "loss": 0.5617,
+      "epoch": 5.6898844483845945,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.3610949520803943e-05,
+      "loss": 0.9766007995605469,
       "step": 494200
     },
     {
-      "epoch": 11.384360320530533,
-      "grad_norm": 3.494835376739502,
-      "learning_rate": 1.3610222364421693e-05,
-      "loss": 0.5618,
+      "epoch": 5.692187133459309,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3603677956981429e-05,
+      "loss": 0.9869244384765625,
       "step": 494400
     },
     {
-      "epoch": 11.388965644284793,
-      "grad_norm": 2.891660213470459,
-      "learning_rate": 1.3602950800599177e-05,
-      "loss": 0.5602,
+      "epoch": 5.6944898185340245,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3596406393158913e-05,
+      "loss": 0.9463243865966797,
       "step": 494600
     },
     {
-      "epoch": 11.393570968039054,
-      "grad_norm": 3.5556230545043945,
-      "learning_rate": 1.3595679236776661e-05,
-      "loss": 0.5636,
+      "epoch": 5.696792503608739,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3589134829336398e-05,
+      "loss": 0.9932901000976563,
       "step": 494800
     },
     {
-      "epoch": 11.398176291793312,
-      "grad_norm": 2.6204235553741455,
-      "learning_rate": 1.3588407672954145e-05,
-      "loss": 0.5588,
+      "epoch": 5.699095188683454,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3581863265513882e-05,
+      "loss": 0.9636652374267578,
       "step": 495000
     },
     {
-      "epoch": 11.402781615547573,
-      "grad_norm": 3.596064805984497,
-      "learning_rate": 1.3581136109131631e-05,
-      "loss": 0.5591,
+      "epoch": 5.701397873758169,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3574591701691366e-05,
+      "loss": 0.9758798217773438,
       "step": 495200
     },
     {
-      "epoch": 11.407386939301833,
-      "grad_norm": 3.3406505584716797,
-      "learning_rate": 1.3573864545309114e-05,
-      "loss": 0.5631,
+      "epoch": 5.703700558832884,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3567320137868852e-05,
+      "loss": 0.968555679321289,
       "step": 495400
     },
     {
-      "epoch": 11.411992263056092,
-      "grad_norm": 3.357797145843506,
-      "learning_rate": 1.3566592981486598e-05,
-      "loss": 0.5515,
+      "epoch": 5.706003243907599,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.3560048574046334e-05,
+      "loss": 0.9713514709472656,
       "step": 495600
     },
     {
-      "epoch": 11.416597586810353,
-      "grad_norm": 3.499816656112671,
-      "learning_rate": 1.3559321417664084e-05,
-      "loss": 0.5618,
+      "epoch": 5.708305928982314,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3552777010223819e-05,
+      "loss": 0.9624390411376953,
       "step": 495800
     },
     {
-      "epoch": 11.421202910564613,
-      "grad_norm": 2.6165995597839355,
-      "learning_rate": 1.3552049853841568e-05,
-      "loss": 0.5569,
+      "epoch": 5.710608614057029,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3545505446401305e-05,
+      "loss": 0.9755973052978516,
       "step": 496000
     },
     {
-      "epoch": 11.425808234318872,
-      "grad_norm": 2.801710844039917,
-      "learning_rate": 1.354477829001905e-05,
-      "loss": 0.5481,
+      "epoch": 5.7129112991317434,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3538233882578787e-05,
+      "loss": 1.0029659271240234,
       "step": 496200
     },
     {
-      "epoch": 11.430413558073132,
-      "grad_norm": 3.2075862884521484,
-      "learning_rate": 1.3537506726196537e-05,
-      "loss": 0.5672,
+      "epoch": 5.715213984206459,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3530962318756271e-05,
+      "loss": 0.958271484375,
       "step": 496400
     },
     {
-      "epoch": 11.435018881827393,
-      "grad_norm": 2.7603681087493896,
-      "learning_rate": 1.3530235162374021e-05,
-      "loss": 0.5631,
+      "epoch": 5.717516669281173,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3523690754933757e-05,
+      "loss": 1.0108000183105468,
       "step": 496600
     },
     {
-      "epoch": 11.439624205581651,
-      "grad_norm": 2.9977965354919434,
-      "learning_rate": 1.3522963598551505e-05,
-      "loss": 0.5488,
+      "epoch": 5.719819354355889,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.351641919111124e-05,
+      "loss": 0.9738809204101563,
       "step": 496800
     },
     {
-      "epoch": 11.444229529335912,
-      "grad_norm": 2.803744316101074,
-      "learning_rate": 1.351569203472899e-05,
-      "loss": 0.5578,
+      "epoch": 5.722122039430603,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3509147627288726e-05,
+      "loss": 0.99912109375,
       "step": 497000
     },
     {
-      "epoch": 11.448834853090172,
-      "grad_norm": 3.423523187637329,
-      "learning_rate": 1.3508420470906474e-05,
-      "loss": 0.563,
+      "epoch": 5.724424724505319,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.350187606346621e-05,
+      "loss": 0.99430419921875,
       "step": 497200
     },
     {
-      "epoch": 11.453440176844433,
-      "grad_norm": 3.3576598167419434,
-      "learning_rate": 1.3501148907083958e-05,
-      "loss": 0.5555,
+      "epoch": 5.726727409580033,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.3494604499643692e-05,
+      "loss": 0.973975601196289,
       "step": 497400
     },
     {
-      "epoch": 11.458045500598692,
-      "grad_norm": 3.5315699577331543,
-      "learning_rate": 1.3493877343261442e-05,
-      "loss": 0.564,
+      "epoch": 5.729030094654748,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3487332935821178e-05,
+      "loss": 0.9921266937255859,
       "step": 497600
     },
     {
-      "epoch": 11.462650824352952,
-      "grad_norm": 2.96769642829895,
-      "learning_rate": 1.3486605779438926e-05,
-      "loss": 0.5566,
+      "epoch": 5.731332779729463,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3480061371998663e-05,
+      "loss": 0.9895915222167969,
       "step": 497800
     },
     {
-      "epoch": 11.467256148107213,
-      "grad_norm": 2.768280267715454,
-      "learning_rate": 1.347933421561641e-05,
-      "loss": 0.5495,
+      "epoch": 5.733635464804178,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3472789808176147e-05,
+      "loss": 0.9898014831542968,
       "step": 498000
     },
     {
-      "epoch": 11.471861471861471,
-      "grad_norm": 3.71108078956604,
-      "learning_rate": 1.3472062651793895e-05,
-      "loss": 0.561,
+      "epoch": 5.735938149878893,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3465518244353631e-05,
+      "loss": 0.9841976928710937,
       "step": 498200
     },
     {
-      "epoch": 11.476466795615732,
-      "grad_norm": 3.0207583904266357,
-      "learning_rate": 1.346479108797138e-05,
-      "loss": 0.5592,
+      "epoch": 5.738240834953608,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3458246680531115e-05,
+      "loss": 0.9864323425292969,
       "step": 498400
     },
     {
-      "epoch": 11.481072119369992,
-      "grad_norm": 3.7575089931488037,
-      "learning_rate": 1.3457519524148863e-05,
-      "loss": 0.556,
+      "epoch": 5.740543520028323,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3450975116708601e-05,
+      "loss": 0.9855081939697266,
       "step": 498600
     },
     {
-      "epoch": 11.485677443124251,
-      "grad_norm": 3.050069570541382,
-      "learning_rate": 1.3450247960326347e-05,
-      "loss": 0.5618,
+      "epoch": 5.742846205103038,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3443703552886084e-05,
+      "loss": 0.977813720703125,
       "step": 498800
     },
     {
-      "epoch": 11.490282766878511,
-      "grad_norm": 3.386132001876831,
-      "learning_rate": 1.3443012754322944e-05,
-      "loss": 0.5526,
+      "epoch": 5.745148890177753,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3436431989063568e-05,
+      "loss": 1.0014332580566405,
       "step": 499000
     },
     {
-      "epoch": 11.494888090632772,
-      "grad_norm": 2.874272584915161,
-      "learning_rate": 1.343574119050043e-05,
-      "loss": 0.561,
+      "epoch": 5.747451575252468,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3429160425241054e-05,
+      "loss": 0.9798948669433594,
       "step": 499200
     },
     {
-      "epoch": 11.49949341438703,
-      "grad_norm": 2.706022262573242,
-      "learning_rate": 1.3428469626677914e-05,
-      "loss": 0.5654,
+      "epoch": 5.749754260327183,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3421888861418536e-05,
+      "loss": 0.9854695129394532,
       "step": 499400
     },
     {
-      "epoch": 11.504098738141291,
-      "grad_norm": 3.2156574726104736,
-      "learning_rate": 1.3421234420674511e-05,
-      "loss": 0.561,
+      "epoch": 5.752056945401898,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.341461729759602e-05,
+      "loss": 0.9640402221679687,
       "step": 499600
     },
     {
-      "epoch": 11.508704061895552,
-      "grad_norm": 2.3544604778289795,
-      "learning_rate": 1.3413962856851994e-05,
-      "loss": 0.5725,
+      "epoch": 5.754359630476612,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3407345733773506e-05,
+      "loss": 0.9828694915771484,
       "step": 499800
     },
     {
-      "epoch": 11.513309385649812,
-      "grad_norm": 3.362711191177368,
-      "learning_rate": 1.340669129302948e-05,
-      "loss": 0.5669,
+      "epoch": 5.756662315551328,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3400074169950989e-05,
+      "loss": 0.9606679534912109,
       "step": 500000
     },
     {
-      "epoch": 11.51791470940407,
-      "grad_norm": 3.2059872150421143,
-      "learning_rate": 1.3399419729206964e-05,
-      "loss": 0.5644,
+      "epoch": 5.758965000626042,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3392802606128475e-05,
+      "loss": 0.9945602416992188,
       "step": 500200
     },
     {
-      "epoch": 11.522520033158331,
-      "grad_norm": 2.8236091136932373,
-      "learning_rate": 1.3392148165384448e-05,
-      "loss": 0.5625,
+      "epoch": 5.761267685700758,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3385531042305959e-05,
+      "loss": 1.0002516937255859,
       "step": 500400
     },
     {
-      "epoch": 11.527125356912592,
-      "grad_norm": 3.4916694164276123,
-      "learning_rate": 1.3384912959381045e-05,
-      "loss": 0.5556,
+      "epoch": 5.763570370775472,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.3378259478483442e-05,
+      "loss": 1.0048902893066407,
       "step": 500600
     },
     {
-      "epoch": 11.53173068066685,
-      "grad_norm": 2.976130962371826,
-      "learning_rate": 1.337764139555853e-05,
-      "loss": 0.5611,
+      "epoch": 5.765873055850188,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3370987914660927e-05,
+      "loss": 0.978136215209961,
       "step": 500800
     },
     {
-      "epoch": 11.536336004421111,
-      "grad_norm": 3.8852763175964355,
-      "learning_rate": 1.3370369831736013e-05,
-      "loss": 0.5676,
+      "epoch": 5.768175740924902,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3363716350838412e-05,
+      "loss": 0.9923693084716797,
       "step": 501000
     },
     {
-      "epoch": 11.540941328175371,
-      "grad_norm": 3.322741985321045,
-      "learning_rate": 1.3363098267913498e-05,
-      "loss": 0.5669,
+      "epoch": 5.770478425999617,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3356444787015898e-05,
+      "loss": 0.970635757446289,
       "step": 501200
     },
     {
-      "epoch": 11.54554665192963,
-      "grad_norm": 2.873216390609741,
-      "learning_rate": 1.3355826704090982e-05,
-      "loss": 0.5711,
+      "epoch": 5.772781111074332,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.334917322319338e-05,
+      "loss": 0.9800533294677735,
       "step": 501400
     },
     {
-      "epoch": 11.55015197568389,
-      "grad_norm": 2.740243911743164,
-      "learning_rate": 1.3348555140268466e-05,
-      "loss": 0.5584,
+      "epoch": 5.7750837961490475,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.3341901659370864e-05,
+      "loss": 1.0008333587646485,
       "step": 501600
     },
     {
-      "epoch": 11.554757299438151,
-      "grad_norm": 2.4768307209014893,
-      "learning_rate": 1.334128357644595e-05,
-      "loss": 0.5652,
+      "epoch": 5.777386481223762,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.333463009554835e-05,
+      "loss": 0.9850815582275391,
       "step": 501800
     },
     {
-      "epoch": 11.55936262319241,
-      "grad_norm": 2.8483810424804688,
-      "learning_rate": 1.3334012012623435e-05,
-      "loss": 0.5571,
+      "epoch": 5.779689166298477,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3327358531725833e-05,
+      "loss": 0.96795166015625,
       "step": 502000
     },
     {
-      "epoch": 11.56396794694667,
-      "grad_norm": 2.515305280685425,
-      "learning_rate": 1.332674044880092e-05,
-      "loss": 0.5515,
+      "epoch": 5.781991851373192,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3320086967903317e-05,
+      "loss": 0.9840054321289062,
       "step": 502200
     },
     {
-      "epoch": 11.56857327070093,
-      "grad_norm": 2.972041606903076,
-      "learning_rate": 1.3319468884978403e-05,
-      "loss": 0.5558,
+      "epoch": 5.784294536447907,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3312815404080803e-05,
+      "loss": 0.9985717010498046,
       "step": 502400
     },
     {
-      "epoch": 11.57317859445519,
-      "grad_norm": 2.9375104904174805,
-      "learning_rate": 1.3312197321155887e-05,
-      "loss": 0.5514,
+      "epoch": 5.786597221522622,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3305543840258285e-05,
+      "loss": 0.9868987274169921,
       "step": 502600
     },
     {
-      "epoch": 11.57778391820945,
-      "grad_norm": 2.7295291423797607,
-      "learning_rate": 1.3304925757333373e-05,
-      "loss": 0.5623,
+      "epoch": 5.7888999065973366,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3298272276435771e-05,
+      "loss": 0.9638268280029297,
       "step": 502800
     },
     {
-      "epoch": 11.58238924196371,
-      "grad_norm": 3.959212064743042,
-      "learning_rate": 1.3297654193510857e-05,
-      "loss": 0.5642,
+      "epoch": 5.791202591672052,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3291000712613256e-05,
+      "loss": 0.9871996307373047,
       "step": 503000
     },
     {
-      "epoch": 11.58699456571797,
-      "grad_norm": 3.5699644088745117,
-      "learning_rate": 1.329038262968834e-05,
-      "loss": 0.553,
+      "epoch": 5.7935052767467665,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3283729148790738e-05,
+      "loss": 0.982093276977539,
       "step": 503200
     },
     {
-      "epoch": 11.59159988947223,
-      "grad_norm": 2.7130260467529297,
-      "learning_rate": 1.3283111065865826e-05,
-      "loss": 0.5518,
+      "epoch": 5.795807961821481,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.3276457584968224e-05,
+      "loss": 0.9666788482666016,
       "step": 503400
     },
     {
-      "epoch": 11.59620521322649,
-      "grad_norm": 2.8701913356781006,
-      "learning_rate": 1.327583950204331e-05,
-      "loss": 0.5529,
+      "epoch": 5.7981106468961965,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3269186021145708e-05,
+      "loss": 0.9756722259521484,
       "step": 503600
     },
     {
-      "epoch": 11.600810536980749,
-      "grad_norm": 2.8965420722961426,
-      "learning_rate": 1.3268567938220794e-05,
-      "loss": 0.556,
+      "epoch": 5.800413331970911,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.326191445732319e-05,
+      "loss": 0.9657235717773438,
       "step": 503800
     },
     {
-      "epoch": 11.60541586073501,
-      "grad_norm": 2.749356269836426,
-      "learning_rate": 1.3261296374398278e-05,
-      "loss": 0.5555,
+      "epoch": 5.8027160170456265,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3254642893500677e-05,
+      "loss": 0.9707088470458984,
       "step": 504000
     },
     {
-      "epoch": 11.61002118448927,
-      "grad_norm": 3.4139721393585205,
-      "learning_rate": 1.3254024810575763e-05,
-      "loss": 0.5527,
+      "epoch": 5.805018702120341,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3247371329678161e-05,
+      "loss": 0.9648799133300782,
       "step": 504200
     },
     {
-      "epoch": 11.61462650824353,
-      "grad_norm": 2.8031933307647705,
-      "learning_rate": 1.3246753246753247e-05,
-      "loss": 0.5462,
+      "epoch": 5.807321387195056,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3240099765855647e-05,
+      "loss": 0.9958235168457031,
       "step": 504400
     },
     {
-      "epoch": 11.619231831997789,
-      "grad_norm": 3.3376245498657227,
-      "learning_rate": 1.3239481682930731e-05,
-      "loss": 0.563,
+      "epoch": 5.809624072269771,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.323282820203313e-05,
+      "loss": 0.9823985290527344,
       "step": 504600
     },
     {
-      "epoch": 11.62383715575205,
-      "grad_norm": 2.577552318572998,
-      "learning_rate": 1.3232210119108217e-05,
-      "loss": 0.551,
+      "epoch": 5.811926757344486,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.3225556638210614e-05,
+      "loss": 0.9838928985595703,
       "step": 504800
     },
     {
-      "epoch": 11.62844247950631,
-      "grad_norm": 2.8429970741271973,
-      "learning_rate": 1.32249385552857e-05,
-      "loss": 0.5543,
+      "epoch": 5.814229442419201,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.32182850743881e-05,
+      "loss": 0.9789728546142578,
       "step": 505000
     },
     {
-      "epoch": 11.633047803260569,
-      "grad_norm": 2.562479019165039,
-      "learning_rate": 1.3217666991463184e-05,
-      "loss": 0.5515,
+      "epoch": 5.816532127493916,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3211013510565582e-05,
+      "loss": 0.9647856140136719,
       "step": 505200
     },
     {
-      "epoch": 11.63765312701483,
-      "grad_norm": 3.0546891689300537,
-      "learning_rate": 1.321039542764067e-05,
-      "loss": 0.5623,
+      "epoch": 5.818834812568631,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3203741946743066e-05,
+      "loss": 0.9774513244628906,
       "step": 505400
     },
     {
-      "epoch": 11.64225845076909,
-      "grad_norm": 3.4961397647857666,
-      "learning_rate": 1.3203123863818152e-05,
-      "loss": 0.552,
+      "epoch": 5.821137497643345,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3196470382920552e-05,
+      "loss": 0.9803495788574219,
       "step": 505600
     },
     {
-      "epoch": 11.646863774523348,
-      "grad_norm": 2.9656572341918945,
-      "learning_rate": 1.3195852299995636e-05,
-      "loss": 0.5574,
+      "epoch": 5.823440182718061,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3189198819098035e-05,
+      "loss": 0.9838663482666016,
       "step": 505800
     },
     {
-      "epoch": 11.651469098277609,
-      "grad_norm": 2.4916203022003174,
-      "learning_rate": 1.3188617093992233e-05,
-      "loss": 0.5648,
+      "epoch": 5.825742867792775,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.318192725527552e-05,
+      "loss": 0.9702992248535156,
       "step": 506000
     },
     {
-      "epoch": 11.65607442203187,
-      "grad_norm": 3.366502046585083,
-      "learning_rate": 1.318134553016972e-05,
-      "loss": 0.5664,
+      "epoch": 5.828045552867491,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3174655691453005e-05,
+      "loss": 0.9721846771240235,
       "step": 506200
     },
     {
-      "epoch": 11.660679745786128,
-      "grad_norm": 2.6675634384155273,
-      "learning_rate": 1.3174073966347203e-05,
-      "loss": 0.561,
+      "epoch": 5.830348237942205,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3167384127630487e-05,
+      "loss": 0.9824070739746094,
       "step": 506400
     },
     {
-      "epoch": 11.665285069540388,
-      "grad_norm": 2.798330068588257,
-      "learning_rate": 1.3166802402524688e-05,
-      "loss": 0.5609,
+      "epoch": 5.832650923016921,
+      "grad_norm": 1.0,
+      "learning_rate": 1.3160112563807973e-05,
+      "loss": 0.9909092712402344,
       "step": 506600
     },
     {
-      "epoch": 11.669890393294649,
-      "grad_norm": 3.1210689544677734,
-      "learning_rate": 1.3159530838702172e-05,
-      "loss": 0.5702,
+      "epoch": 5.834953608091635,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3152840999985457e-05,
+      "loss": 0.981254653930664,
       "step": 506800
     },
     {
-      "epoch": 11.67449571704891,
-      "grad_norm": 4.939085006713867,
-      "learning_rate": 1.3152259274879656e-05,
-      "loss": 0.559,
+      "epoch": 5.83725629316635,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.314556943616294e-05,
+      "loss": 0.9601033782958984,
       "step": 507000
     },
     {
-      "epoch": 11.679101040803168,
-      "grad_norm": 3.4388821125030518,
-      "learning_rate": 1.314498771105714e-05,
-      "loss": 0.5616,
+      "epoch": 5.839558978241065,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3138297872340426e-05,
+      "loss": 0.9732923126220703,
       "step": 507200
     },
     {
-      "epoch": 11.683706364557429,
-      "grad_norm": 2.974487066268921,
-      "learning_rate": 1.3137716147234625e-05,
-      "loss": 0.5539,
+      "epoch": 5.84186166331578,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.313102630851791e-05,
+      "loss": 0.9881997680664063,
       "step": 507400
     },
     {
-      "epoch": 11.688311688311689,
-      "grad_norm": 2.9296836853027344,
-      "learning_rate": 1.3130444583412109e-05,
-      "loss": 0.5657,
+      "epoch": 5.844164348390495,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3123754744695394e-05,
+      "loss": 0.9774773406982422,
       "step": 507600
     },
     {
-      "epoch": 11.692917012065948,
-      "grad_norm": 3.680119276046753,
-      "learning_rate": 1.3123173019589593e-05,
-      "loss": 0.5624,
+      "epoch": 5.84646703346521,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3116483180872879e-05,
+      "loss": 0.9892032623291016,
       "step": 507800
     },
     {
-      "epoch": 11.697522335820208,
-      "grad_norm": 2.4661529064178467,
-      "learning_rate": 1.3115901455767077e-05,
-      "loss": 0.5654,
+      "epoch": 5.848769718539925,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3109211617050363e-05,
+      "loss": 0.9897892761230469,
       "step": 508000
     },
     {
-      "epoch": 11.702127659574469,
-      "grad_norm": 3.046780586242676,
-      "learning_rate": 1.3108629891944563e-05,
-      "loss": 0.5738,
+      "epoch": 5.85107240361464,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3101940053227849e-05,
+      "loss": 0.9681523132324219,
       "step": 508200
     },
     {
-      "epoch": 11.706732983328727,
-      "grad_norm": 2.5584285259246826,
-      "learning_rate": 1.3101358328122046e-05,
-      "loss": 0.554,
+      "epoch": 5.853375088689355,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3094668489405331e-05,
+      "loss": 0.966031494140625,
       "step": 508400
     },
     {
-      "epoch": 11.711338307082988,
-      "grad_norm": 3.345762014389038,
-      "learning_rate": 1.309408676429953e-05,
-      "loss": 0.5675,
+      "epoch": 5.85567777376407,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3087396925582817e-05,
+      "loss": 0.9670277404785156,
       "step": 508600
     },
     {
-      "epoch": 11.715943630837248,
-      "grad_norm": 2.661505937576294,
-      "learning_rate": 1.3086815200477016e-05,
-      "loss": 0.5649,
+      "epoch": 5.857980458838785,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3080125361760301e-05,
+      "loss": 0.9803645324707031,
       "step": 508800
     },
     {
-      "epoch": 11.720548954591507,
-      "grad_norm": 3.2517101764678955,
-      "learning_rate": 1.3079543636654498e-05,
-      "loss": 0.5518,
+      "epoch": 5.8602831439135,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3072853797937784e-05,
+      "loss": 0.9849859619140625,
       "step": 509000
     },
     {
-      "epoch": 11.725154278345768,
-      "grad_norm": 2.780080556869507,
-      "learning_rate": 1.3072272072831983e-05,
-      "loss": 0.5524,
+      "epoch": 5.862585828988214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.306558223411527e-05,
+      "loss": 0.9650568389892578,
       "step": 509200
     },
     {
-      "epoch": 11.729759602100028,
-      "grad_norm": 3.5682613849639893,
-      "learning_rate": 1.3065000509009468e-05,
-      "loss": 0.5642,
+      "epoch": 5.86488851406293,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3058310670292754e-05,
+      "loss": 0.9649927520751953,
       "step": 509400
     },
     {
-      "epoch": 11.734364925854287,
-      "grad_norm": 2.8819103240966797,
-      "learning_rate": 1.3057728945186953e-05,
-      "loss": 0.5553,
+      "epoch": 5.867191199137644,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3051039106470237e-05,
+      "loss": 1.0007953643798828,
       "step": 509600
     },
     {
-      "epoch": 11.738970249608547,
-      "grad_norm": 3.0837063789367676,
-      "learning_rate": 1.3050457381364437e-05,
-      "loss": 0.5659,
+      "epoch": 5.86949388421236,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3043767542647722e-05,
+      "loss": 0.9851982879638672,
       "step": 509800
     },
     {
-      "epoch": 11.743575573362808,
-      "grad_norm": 2.94468355178833,
-      "learning_rate": 1.3043185817541921e-05,
-      "loss": 0.5447,
+      "epoch": 5.871796569287074,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3036495978825207e-05,
+      "loss": 0.982063980102539,
       "step": 510000
     },
     {
-      "epoch": 11.748180897117066,
-      "grad_norm": 2.4969544410705566,
-      "learning_rate": 1.3035914253719405e-05,
-      "loss": 0.5503,
+      "epoch": 5.87409925436179,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3029224415002691e-05,
+      "loss": 0.9685086822509765,
       "step": 510200
     },
     {
-      "epoch": 11.752786220871327,
-      "grad_norm": 2.9021387100219727,
-      "learning_rate": 1.302864268989689e-05,
-      "loss": 0.5498,
+      "epoch": 5.876401939436504,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3021952851180175e-05,
+      "loss": 0.9721865844726563,
       "step": 510400
     },
     {
-      "epoch": 11.757391544625587,
-      "grad_norm": 3.434443473815918,
-      "learning_rate": 1.3021371126074374e-05,
-      "loss": 0.5611,
+      "epoch": 5.878704624511219,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.301468128735766e-05,
+      "loss": 0.985322265625,
       "step": 510600
     },
     {
-      "epoch": 11.761996868379848,
-      "grad_norm": 2.887711763381958,
-      "learning_rate": 1.3014099562251858e-05,
-      "loss": 0.5348,
+      "epoch": 5.881007309585934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3007409723535144e-05,
+      "loss": 0.9845337677001953,
       "step": 510800
     },
     {
-      "epoch": 11.766602192134107,
-      "grad_norm": 3.3442883491516113,
-      "learning_rate": 1.3006827998429342e-05,
-      "loss": 0.5537,
+      "epoch": 5.883309994660649,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.3000138159712628e-05,
+      "loss": 0.9656434631347657,
       "step": 511000
     },
     {
-      "epoch": 11.771207515888367,
-      "grad_norm": 2.8919873237609863,
-      "learning_rate": 1.2999592792425939e-05,
-      "loss": 0.5731,
+      "epoch": 5.885612679735364,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2992866595890112e-05,
+      "loss": 0.9908671569824219,
       "step": 511200
     },
     {
-      "epoch": 11.775812839642628,
-      "grad_norm": 3.013439178466797,
-      "learning_rate": 1.2992321228603423e-05,
-      "loss": 0.5496,
+      "epoch": 5.887915364810079,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2985595032067598e-05,
+      "loss": 0.9956317901611328,
       "step": 511400
     },
     {
-      "epoch": 11.780418163396886,
-      "grad_norm": 2.499289035797119,
-      "learning_rate": 1.298508602260002e-05,
-      "loss": 0.5678,
+      "epoch": 5.890218049884794,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.297832346824508e-05,
+      "loss": 0.9613990783691406,
       "step": 511600
     },
     {
-      "epoch": 11.785023487151147,
-      "grad_norm": 3.30153226852417,
-      "learning_rate": 1.2977814458777506e-05,
-      "loss": 0.557,
+      "epoch": 5.892520734959509,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.2971051904422566e-05,
+      "loss": 0.988326416015625,
       "step": 511800
     },
     {
-      "epoch": 11.789628810905407,
-      "grad_norm": 3.1825666427612305,
-      "learning_rate": 1.2970542894954989e-05,
-      "loss": 0.5488,
+      "epoch": 5.894823420034224,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.296378034060005e-05,
+      "loss": 0.9974417877197266,
       "step": 512000
     },
     {
-      "epoch": 11.794234134659666,
-      "grad_norm": 3.229902744293213,
-      "learning_rate": 1.2963271331132473e-05,
-      "loss": 0.567,
+      "epoch": 5.8971261051089385,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2956508776777533e-05,
+      "loss": 0.9984935760498047,
       "step": 512200
     },
     {
-      "epoch": 11.798839458413926,
-      "grad_norm": 2.68745493888855,
-      "learning_rate": 1.2955999767309959e-05,
-      "loss": 0.5674,
+      "epoch": 5.899428790183654,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2949237212955019e-05,
+      "loss": 0.9851997375488282,
       "step": 512400
     },
     {
-      "epoch": 11.803444782168187,
-      "grad_norm": 3.270493507385254,
-      "learning_rate": 1.2948728203487441e-05,
-      "loss": 0.5487,
+      "epoch": 5.9017314752583685,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2941965649132503e-05,
+      "loss": 0.9762357330322265,
       "step": 512600
     },
     {
-      "epoch": 11.808050105922446,
-      "grad_norm": 3.161329507827759,
-      "learning_rate": 1.2941456639664927e-05,
-      "loss": 0.5625,
+      "epoch": 5.904034160333083,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.2934694085309986e-05,
+      "loss": 0.9988518524169921,
       "step": 512800
     },
     {
-      "epoch": 11.812655429676706,
-      "grad_norm": 2.8435397148132324,
-      "learning_rate": 1.2934185075842411e-05,
-      "loss": 0.5453,
+      "epoch": 5.9063368454077985,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2927422521487472e-05,
+      "loss": 0.9867639923095703,
       "step": 513000
     },
     {
-      "epoch": 11.817260753430967,
-      "grad_norm": 2.739443778991699,
-      "learning_rate": 1.2926913512019896e-05,
-      "loss": 0.5546,
+      "epoch": 5.908639530482513,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2920150957664956e-05,
+      "loss": 0.9704194641113282,
       "step": 513200
     },
     {
-      "epoch": 11.821866077185225,
-      "grad_norm": 2.649695634841919,
-      "learning_rate": 1.291964194819738e-05,
-      "loss": 0.5519,
+      "epoch": 5.910942215557228,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.291287939384244e-05,
+      "loss": 0.967171630859375,
       "step": 513400
     },
     {
-      "epoch": 11.826471400939486,
-      "grad_norm": 3.265084981918335,
-      "learning_rate": 1.2912370384374864e-05,
-      "loss": 0.5485,
+      "epoch": 5.913244900631943,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.2905607830019924e-05,
+      "loss": 0.9707279968261718,
       "step": 513600
     },
     {
-      "epoch": 11.831076724693746,
-      "grad_norm": 3.4886581897735596,
-      "learning_rate": 1.2905098820552348e-05,
-      "loss": 0.5591,
+      "epoch": 5.915547585706658,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2898336266197409e-05,
+      "loss": 0.9839116668701172,
       "step": 513800
     },
     {
-      "epoch": 11.835682048448007,
-      "grad_norm": 3.9588351249694824,
-      "learning_rate": 1.2897827256729833e-05,
-      "loss": 0.5473,
+      "epoch": 5.917850270781373,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2891064702374893e-05,
+      "loss": 0.9717050933837891,
       "step": 514000
     },
     {
-      "epoch": 11.840287372202265,
-      "grad_norm": 3.0174105167388916,
-      "learning_rate": 1.2890555692907317e-05,
-      "loss": 0.5637,
+      "epoch": 5.9201529558560875,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2883793138552377e-05,
+      "loss": 1.015107421875,
       "step": 514200
     },
     {
-      "epoch": 11.844892695956526,
-      "grad_norm": 2.9916231632232666,
-      "learning_rate": 1.2883284129084803e-05,
-      "loss": 0.5631,
+      "epoch": 5.922455640930803,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2876521574729861e-05,
+      "loss": 0.9687767028808594,
       "step": 514400
     },
     {
-      "epoch": 11.849498019710786,
-      "grad_norm": 3.1581456661224365,
-      "learning_rate": 1.2876012565262285e-05,
-      "loss": 0.5553,
+      "epoch": 5.924758326005517,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2869250010907347e-05,
+      "loss": 1.0059048461914062,
       "step": 514600
     },
     {
-      "epoch": 11.854103343465045,
-      "grad_norm": 2.6253857612609863,
-      "learning_rate": 1.286874100143977e-05,
-      "loss": 0.5628,
+      "epoch": 5.927061011080233,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.286197844708483e-05,
+      "loss": 0.9638690948486328,
       "step": 514800
     },
     {
-      "epoch": 11.858708667219306,
-      "grad_norm": 3.4651215076446533,
-      "learning_rate": 1.2861469437617255e-05,
-      "loss": 0.555,
+      "epoch": 5.929363696154947,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2854706883262316e-05,
+      "loss": 0.9991024780273438,
       "step": 515000
     },
     {
-      "epoch": 11.863313990973566,
-      "grad_norm": 5.004696369171143,
-      "learning_rate": 1.2854197873794738e-05,
-      "loss": 0.5577,
+      "epoch": 5.931666381229663,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.28474353194398e-05,
+      "loss": 0.9884070587158204,
       "step": 515200
     },
     {
-      "epoch": 11.867919314727825,
-      "grad_norm": 3.6759226322174072,
-      "learning_rate": 1.2846926309972222e-05,
-      "loss": 0.5608,
+      "epoch": 5.933969066304377,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2840163755617282e-05,
+      "loss": 0.9854502868652344,
       "step": 515400
     },
     {
-      "epoch": 11.872524638482085,
-      "grad_norm": 4.682463645935059,
-      "learning_rate": 1.2839654746149708e-05,
-      "loss": 0.5639,
+      "epoch": 5.936271751379093,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2832892191794768e-05,
+      "loss": 1.0004030609130858,
       "step": 515600
     },
     {
-      "epoch": 11.877129962236346,
-      "grad_norm": 3.1118075847625732,
-      "learning_rate": 1.2832419540146305e-05,
-      "loss": 0.5593,
+      "epoch": 5.938574436453807,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.2825620627972252e-05,
+      "loss": 0.9818889617919921,
       "step": 515800
     },
     {
-      "epoch": 11.881735285990604,
-      "grad_norm": 3.632077693939209,
-      "learning_rate": 1.2825147976323787e-05,
-      "loss": 0.5623,
+      "epoch": 5.940877121528523,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2818349064149737e-05,
+      "loss": 0.9734959411621094,
       "step": 516000
     },
     {
-      "epoch": 11.886340609744865,
-      "grad_norm": 3.2938430309295654,
-      "learning_rate": 1.2817876412501273e-05,
-      "loss": 0.5567,
+      "epoch": 5.943179806603237,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.2811077500327221e-05,
+      "loss": 0.9703239440917969,
       "step": 516200
     },
     {
-      "epoch": 11.890945933499125,
-      "grad_norm": 3.3671703338623047,
-      "learning_rate": 1.2810604848678758e-05,
-      "loss": 0.558,
+      "epoch": 5.945482491677952,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2803805936504705e-05,
+      "loss": 0.9912371063232421,
       "step": 516400
     },
     {
-      "epoch": 11.895551257253384,
-      "grad_norm": 2.796440839767456,
-      "learning_rate": 1.2803333284856242e-05,
-      "loss": 0.5565,
+      "epoch": 5.947785176752667,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.279653437268219e-05,
+      "loss": 0.9781558227539062,
       "step": 516600
     },
     {
-      "epoch": 11.900156581007645,
-      "grad_norm": 4.450211524963379,
-      "learning_rate": 1.2796061721033726e-05,
-      "loss": 0.551,
+      "epoch": 5.950087861827382,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2789262808859674e-05,
+      "loss": 0.9869168090820313,
       "step": 516800
     },
     {
-      "epoch": 11.904761904761905,
-      "grad_norm": 2.8839023113250732,
-      "learning_rate": 1.278879015721121e-05,
-      "loss": 0.5542,
+      "epoch": 5.952390546902097,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2781991245037158e-05,
+      "loss": 0.9716383361816406,
       "step": 517000
     },
     {
-      "epoch": 11.909367228516164,
-      "grad_norm": 3.0486433506011963,
-      "learning_rate": 1.2781518593388694e-05,
-      "loss": 0.5539,
+      "epoch": 5.954693231976812,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2774719681214642e-05,
+      "loss": 0.9640602111816406,
       "step": 517200
     },
     {
-      "epoch": 11.913972552270424,
-      "grad_norm": 3.802643299102783,
-      "learning_rate": 1.2774247029566179e-05,
-      "loss": 0.552,
+      "epoch": 5.956995917051527,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2767448117392126e-05,
+      "loss": 0.9790910339355469,
       "step": 517400
     },
     {
-      "epoch": 11.918577876024685,
-      "grad_norm": 3.0999131202697754,
-      "learning_rate": 1.2766975465743663e-05,
-      "loss": 0.548,
+      "epoch": 5.959298602126242,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2760176553569612e-05,
+      "loss": 0.9806784057617187,
       "step": 517600
     },
     {
-      "epoch": 11.923183199778945,
-      "grad_norm": 2.016890525817871,
-      "learning_rate": 1.2759703901921149e-05,
-      "loss": 0.5554,
+      "epoch": 5.961601287200956,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2752904989747095e-05,
+      "loss": 0.984098892211914,
       "step": 517800
     },
     {
-      "epoch": 11.927788523533204,
-      "grad_norm": 3.3377721309661865,
-      "learning_rate": 1.2752432338098631e-05,
-      "loss": 0.5529,
+      "epoch": 5.963903972275672,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2745633425924579e-05,
+      "loss": 0.9858362579345703,
       "step": 518000
     },
     {
-      "epoch": 11.932393847287464,
-      "grad_norm": 3.55546498298645,
-      "learning_rate": 1.2745197132095228e-05,
-      "loss": 0.5649,
+      "epoch": 5.966206657350387,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2738361862102065e-05,
+      "loss": 0.9572196960449219,
       "step": 518200
     },
     {
-      "epoch": 11.936999171041725,
-      "grad_norm": 2.818472146987915,
-      "learning_rate": 1.2737961926091825e-05,
-      "loss": 0.5533,
+      "epoch": 5.968509342425102,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2731090298279549e-05,
+      "loss": 0.9759970092773438,
       "step": 518400
     },
     {
-      "epoch": 11.941604494795984,
-      "grad_norm": 3.24600887298584,
-      "learning_rate": 1.273069036226931e-05,
-      "loss": 0.5611,
+      "epoch": 5.970812027499816,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2723818734457031e-05,
+      "loss": 0.9888639831542969,
       "step": 518600
     },
     {
-      "epoch": 11.946209818550244,
-      "grad_norm": 2.4776716232299805,
-      "learning_rate": 1.2723418798446795e-05,
-      "loss": 0.5607,
+      "epoch": 5.973114712574532,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2716547170634517e-05,
+      "loss": 0.9922422790527343,
       "step": 518800
     },
     {
-      "epoch": 11.950815142304505,
-      "grad_norm": 3.1430325508117676,
-      "learning_rate": 1.2716147234624278e-05,
-      "loss": 0.5527,
+      "epoch": 5.975417397649246,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2709275606812002e-05,
+      "loss": 0.9752543640136718,
       "step": 519000
     },
     {
-      "epoch": 11.955420466058763,
-      "grad_norm": 3.204774856567383,
-      "learning_rate": 1.2708875670801762e-05,
-      "loss": 0.5664,
+      "epoch": 5.977720082723962,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.2702004042989486e-05,
+      "loss": 0.9682593536376953,
       "step": 519200
     },
     {
-      "epoch": 11.960025789813024,
-      "grad_norm": 4.431719779968262,
-      "learning_rate": 1.2701604106979248e-05,
-      "loss": 0.5497,
+      "epoch": 5.980022767798676,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.269473247916697e-05,
+      "loss": 1.0047833251953124,
       "step": 519400
     },
     {
-      "epoch": 11.964631113567284,
-      "grad_norm": 3.592142343521118,
-      "learning_rate": 1.269433254315673e-05,
-      "loss": 0.5578,
+      "epoch": 5.982325452873392,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2687460915344454e-05,
+      "loss": 0.9858283996582031,
       "step": 519600
     },
     {
-      "epoch": 11.969236437321543,
-      "grad_norm": 3.99924373626709,
-      "learning_rate": 1.2687060979334216e-05,
-      "loss": 0.5628,
+      "epoch": 5.984628137948106,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2680189351521938e-05,
+      "loss": 0.9966690826416016,
       "step": 519800
     },
     {
-      "epoch": 11.973841761075803,
-      "grad_norm": 3.2276337146759033,
-      "learning_rate": 1.26797894155117e-05,
-      "loss": 0.5689,
+      "epoch": 5.986930823022821,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2672917787699423e-05,
+      "loss": 0.9701039123535157,
       "step": 520000
     },
     {
-      "epoch": 11.978447084830064,
-      "grad_norm": 3.4805285930633545,
-      "learning_rate": 1.2672517851689185e-05,
-      "loss": 0.5545,
+      "epoch": 5.989233508097536,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2665646223876907e-05,
+      "loss": 0.9833212280273438,
       "step": 520200
     },
     {
-      "epoch": 11.983052408584324,
-      "grad_norm": 3.0665435791015625,
-      "learning_rate": 1.2665246287866669e-05,
-      "loss": 0.5564,
+      "epoch": 5.991536193172251,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2658374660054391e-05,
+      "loss": 0.9889899444580078,
       "step": 520400
     },
     {
-      "epoch": 11.987657732338583,
-      "grad_norm": 3.2864396572113037,
-      "learning_rate": 1.2657974724044153e-05,
-      "loss": 0.5689,
+      "epoch": 5.993838878246966,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2651103096231875e-05,
+      "loss": 0.9911026000976563,
       "step": 520600
     },
     {
-      "epoch": 11.992263056092844,
-      "grad_norm": 4.101680278778076,
-      "learning_rate": 1.2650703160221637e-05,
-      "loss": 0.5586,
+      "epoch": 5.996141563321681,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2643831532409361e-05,
+      "loss": 0.9813486480712891,
       "step": 520800
     },
     {
-      "epoch": 11.996868379847104,
-      "grad_norm": 3.1322414875030518,
-      "learning_rate": 1.2643431596399122e-05,
-      "loss": 0.5665,
+      "epoch": 5.998444248396396,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2636559968586844e-05,
+      "loss": 0.9834247589111328,
       "step": 521000
     },
     {
-      "epoch": 12.0,
-      "eval_loss": 0.5357550382614136,
-      "eval_runtime": 168.6634,
-      "eval_samples_per_second": 168.152,
-      "eval_steps_per_second": 10.512,
+      "epoch": 6.0,
+      "eval_loss": NaN,
+      "eval_runtime": 429.1974,
+      "eval_samples_per_second": 66.079,
+      "eval_steps_per_second": 33.041,
       "step": 521136
     },
     {
-      "epoch": 12.001473703601363,
-      "grad_norm": 3.8854141235351562,
-      "learning_rate": 1.2636160032576606e-05,
-      "loss": 0.5461,
+      "epoch": 6.000736859223909,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2629288404764328e-05,
+      "loss": 0.9674974060058594,
       "step": 521200
     },
     {
-      "epoch": 12.006079027355623,
-      "grad_norm": 3.1888136863708496,
-      "learning_rate": 1.2628888468754092e-05,
-      "loss": 0.5459,
+      "epoch": 6.003039544298623,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2622016840941814e-05,
+      "loss": 0.9826670837402344,
       "step": 521400
     },
     {
-      "epoch": 12.010684351109884,
-      "grad_norm": 2.910918951034546,
-      "learning_rate": 1.2621653262750689e-05,
-      "loss": 0.5549,
+      "epoch": 6.005342229373339,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2614745277119298e-05,
+      "loss": 0.9674659729003906,
       "step": 521600
     },
     {
-      "epoch": 12.015289674864142,
-      "grad_norm": 3.348071813583374,
-      "learning_rate": 1.2614381698928171e-05,
-      "loss": 0.5604,
+      "epoch": 6.007644914448053,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2607473713296782e-05,
+      "loss": 0.9796168518066406,
       "step": 521800
     },
     {
-      "epoch": 12.019894998618403,
-      "grad_norm": 2.9339218139648438,
-      "learning_rate": 1.2607110135105655e-05,
-      "loss": 0.5526,
+      "epoch": 6.009947599522769,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2600202149474267e-05,
+      "loss": 0.9959929656982421,
       "step": 522000
     },
     {
-      "epoch": 12.024500322372663,
-      "grad_norm": 2.8683624267578125,
-      "learning_rate": 1.2599838571283141e-05,
-      "loss": 0.5467,
+      "epoch": 6.012250284597483,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.259293058565175e-05,
+      "loss": 0.9869330596923828,
       "step": 522200
     },
     {
-      "epoch": 12.029105646126922,
-      "grad_norm": 3.5013625621795654,
-      "learning_rate": 1.2592567007460624e-05,
-      "loss": 0.5531,
+      "epoch": 6.014552969672199,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2585659021829235e-05,
+      "loss": 0.9866191101074219,
       "step": 522400
     },
     {
-      "epoch": 12.033710969881183,
-      "grad_norm": 2.941629409790039,
-      "learning_rate": 1.258529544363811e-05,
-      "loss": 0.553,
+      "epoch": 6.016855654746913,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.257838745800672e-05,
+      "loss": 0.9863572692871094,
       "step": 522600
     },
     {
-      "epoch": 12.038316293635443,
-      "grad_norm": 4.11761999130249,
-      "learning_rate": 1.2578023879815594e-05,
-      "loss": 0.5439,
+      "epoch": 6.019158339821629,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2571115894184203e-05,
+      "loss": 0.9905815887451171,
       "step": 522800
     },
     {
-      "epoch": 12.042921617389702,
-      "grad_norm": 2.471980094909668,
-      "learning_rate": 1.2570752315993077e-05,
-      "loss": 0.5569,
+      "epoch": 6.021461024896343,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2563844330361688e-05,
+      "loss": 0.9822950744628907,
       "step": 523000
     },
     {
-      "epoch": 12.047526941143962,
-      "grad_norm": 3.2188100814819336,
-      "learning_rate": 1.2563480752170562e-05,
-      "loss": 0.5494,
+      "epoch": 6.023763709971058,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2556572766539172e-05,
+      "loss": 0.9872956085205078,
       "step": 523200
     },
     {
-      "epoch": 12.052132264898223,
-      "grad_norm": 3.172586679458618,
-      "learning_rate": 1.2556209188348047e-05,
-      "loss": 0.5461,
+      "epoch": 6.026066395045773,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2549301202716658e-05,
+      "loss": 0.9910843658447266,
       "step": 523400
     },
     {
-      "epoch": 12.056737588652481,
-      "grad_norm": 2.9670534133911133,
-      "learning_rate": 1.254893762452553e-05,
-      "loss": 0.5479,
+      "epoch": 6.028369080120488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.254202963889414e-05,
+      "loss": 1.0010240173339844,
       "step": 523600
     },
     {
-      "epoch": 12.061342912406742,
-      "grad_norm": 2.948775053024292,
-      "learning_rate": 1.2541666060703015e-05,
-      "loss": 0.5428,
+      "epoch": 6.030671765195203,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.2534758075071625e-05,
+      "loss": 0.9774690246582032,
       "step": 523800
     },
     {
-      "epoch": 12.065948236161002,
-      "grad_norm": 2.3510968685150146,
-      "learning_rate": 1.25343944968805e-05,
-      "loss": 0.5526,
+      "epoch": 6.032974450269918,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.252748651124911e-05,
+      "loss": 0.9842904663085937,
       "step": 524000
     },
     {
-      "epoch": 12.070553559915263,
-      "grad_norm": 4.567888259887695,
-      "learning_rate": 1.2527122933057985e-05,
-      "loss": 0.5531,
+      "epoch": 6.035277135344633,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2520214947426593e-05,
+      "loss": 0.98850830078125,
       "step": 524200
     },
     {
-      "epoch": 12.075158883669522,
-      "grad_norm": 3.3408639430999756,
-      "learning_rate": 1.2519851369235468e-05,
-      "loss": 0.545,
+      "epoch": 6.037579820419348,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2512943383604077e-05,
+      "loss": 0.9948532867431641,
       "step": 524400
     },
     {
-      "epoch": 12.079764207423782,
-      "grad_norm": 3.8344717025756836,
-      "learning_rate": 1.2512579805412952e-05,
-      "loss": 0.5609,
+      "epoch": 6.039882505494063,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2505671819781563e-05,
+      "loss": 0.9795927429199218,
       "step": 524600
     },
     {
-      "epoch": 12.084369531178043,
-      "grad_norm": 2.317721366882324,
-      "learning_rate": 1.2505308241590438e-05,
-      "loss": 0.5464,
+      "epoch": 6.042185190568778,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2498400255959047e-05,
+      "loss": 0.9803174591064453,
       "step": 524800
     },
     {
-      "epoch": 12.088974854932301,
-      "grad_norm": 2.980634927749634,
-      "learning_rate": 1.249803667776792e-05,
-      "loss": 0.5394,
+      "epoch": 6.044487875643492,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2491128692136532e-05,
+      "loss": 0.9761277770996094,
       "step": 525000
     },
     {
-      "epoch": 12.093580178686562,
-      "grad_norm": 3.3348686695098877,
-      "learning_rate": 1.2490765113945405e-05,
-      "loss": 0.5551,
+      "epoch": 6.0467905607182075,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2483857128314016e-05,
+      "loss": 0.9698514556884765,
       "step": 525200
     },
     {
-      "epoch": 12.098185502440822,
-      "grad_norm": 3.0347678661346436,
-      "learning_rate": 1.248349355012289e-05,
-      "loss": 0.551,
+      "epoch": 6.049093245792922,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.24765855644915e-05,
+      "loss": 0.9775434875488281,
       "step": 525400
     },
     {
-      "epoch": 12.102790826195081,
-      "grad_norm": 3.286358118057251,
-      "learning_rate": 1.2476221986300373e-05,
-      "loss": 0.5652,
+      "epoch": 6.0513959308676375,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2469314000668984e-05,
+      "loss": 0.9901437377929687,
       "step": 525600
     },
     {
-      "epoch": 12.107396149949341,
-      "grad_norm": 3.339553117752075,
-      "learning_rate": 1.2468950422477859e-05,
-      "loss": 0.5548,
+      "epoch": 6.053698615942352,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2462042436846468e-05,
+      "loss": 0.993725814819336,
       "step": 525800
     },
     {
-      "epoch": 12.112001473703602,
-      "grad_norm": 2.7706658840179443,
-      "learning_rate": 1.2461715216474456e-05,
-      "loss": 0.5553,
+      "epoch": 6.0560013010170675,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2454770873023953e-05,
+      "loss": 0.9937523651123047,
       "step": 526000
     },
     {
-      "epoch": 12.11660679745786,
-      "grad_norm": 2.956782817840576,
-      "learning_rate": 1.245444365265194e-05,
-      "loss": 0.5499,
+      "epoch": 6.058303986091782,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2447499309201437e-05,
+      "loss": 0.9665583038330078,
       "step": 526200
     },
     {
-      "epoch": 12.121212121212121,
-      "grad_norm": 3.1089818477630615,
-      "learning_rate": 1.2447172088829423e-05,
-      "loss": 0.5438,
+      "epoch": 6.060606671166497,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.2440227745378921e-05,
+      "loss": 0.9925328063964843,
       "step": 526400
     },
     {
-      "epoch": 12.125817444966382,
-      "grad_norm": 2.3799941539764404,
-      "learning_rate": 1.2439900525006909e-05,
-      "loss": 0.5455,
+      "epoch": 6.062909356241212,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2432956181556407e-05,
+      "loss": 0.9762063598632813,
       "step": 526600
     },
     {
-      "epoch": 12.13042276872064,
-      "grad_norm": 4.066889762878418,
-      "learning_rate": 1.2432628961184393e-05,
-      "loss": 0.554,
+      "epoch": 6.065212041315927,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.242568461773389e-05,
+      "loss": 0.9671697998046875,
       "step": 526800
     },
     {
-      "epoch": 12.1350280924749,
-      "grad_norm": 3.080200672149658,
-      "learning_rate": 1.2425357397361875e-05,
-      "loss": 0.5558,
+      "epoch": 6.067514726390642,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2418413053911374e-05,
+      "loss": 0.9953849792480469,
       "step": 527000
     },
     {
-      "epoch": 12.139633416229161,
-      "grad_norm": 2.6959545612335205,
-      "learning_rate": 1.2418085833539361e-05,
-      "loss": 0.5523,
+      "epoch": 6.0698174114653565,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.241114149008886e-05,
+      "loss": 0.9750828552246094,
       "step": 527200
     },
     {
-      "epoch": 12.14423873998342,
-      "grad_norm": 2.727076292037964,
-      "learning_rate": 1.2410814269716845e-05,
-      "loss": 0.5532,
+      "epoch": 6.072120096540072,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2403869926266342e-05,
+      "loss": 0.9784751129150391,
       "step": 527400
     },
     {
-      "epoch": 12.14884406373768,
-      "grad_norm": 4.091317653656006,
-      "learning_rate": 1.2403579063713442e-05,
-      "loss": 0.5506,
+      "epoch": 6.074422781614786,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2396598362443826e-05,
+      "loss": 0.99782958984375,
       "step": 527600
     },
     {
-      "epoch": 12.153449387491941,
-      "grad_norm": 3.140193462371826,
-      "learning_rate": 1.2396307499890928e-05,
-      "loss": 0.5585,
+      "epoch": 6.076725466689502,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2389326798621312e-05,
+      "loss": 0.9635990142822266,
       "step": 527800
     },
     {
-      "epoch": 12.158054711246201,
-      "grad_norm": 3.9469101428985596,
-      "learning_rate": 1.238903593606841e-05,
-      "loss": 0.5526,
+      "epoch": 6.079028151764216,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.2382055234798797e-05,
+      "loss": 0.9723565673828125,
       "step": 528000
     },
     {
-      "epoch": 12.16266003500046,
-      "grad_norm": 3.1913647651672363,
-      "learning_rate": 1.2381764372245895e-05,
-      "loss": 0.5434,
+      "epoch": 6.081330836838932,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.237478367097628e-05,
+      "loss": 0.9830453491210938,
       "step": 528200
     },
     {
-      "epoch": 12.16726535875472,
-      "grad_norm": 2.5850043296813965,
-      "learning_rate": 1.2374492808423381e-05,
-      "loss": 0.5574,
+      "epoch": 6.083633521913646,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2367512107153765e-05,
+      "loss": 0.9581756591796875,
       "step": 528400
     },
     {
-      "epoch": 12.171870682508981,
-      "grad_norm": 2.807478904724121,
-      "learning_rate": 1.2367221244600863e-05,
-      "loss": 0.5513,
+      "epoch": 6.085936206988362,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.236024054333125e-05,
+      "loss": 0.9884815216064453,
       "step": 528600
     },
     {
-      "epoch": 12.17647600626324,
-      "grad_norm": 3.168511390686035,
-      "learning_rate": 1.2359949680778348e-05,
-      "loss": 0.5539,
+      "epoch": 6.088238892063076,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2352968979508733e-05,
+      "loss": 0.9800252532958984,
       "step": 528800
     },
     {
-      "epoch": 12.1810813300175,
-      "grad_norm": 3.547654867172241,
-      "learning_rate": 1.2352678116955834e-05,
-      "loss": 0.5485,
+      "epoch": 6.090541577137791,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2345697415686218e-05,
+      "loss": 0.9709556579589844,
       "step": 529000
     },
     {
-      "epoch": 12.18568665377176,
-      "grad_norm": 2.912477493286133,
-      "learning_rate": 1.2345406553133316e-05,
-      "loss": 0.5397,
+      "epoch": 6.092844262212506,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2338425851863704e-05,
+      "loss": 1.0069298553466797,
       "step": 529200
     },
     {
-      "epoch": 12.19029197752602,
-      "grad_norm": 3.6767055988311768,
-      "learning_rate": 1.2338134989310802e-05,
-      "loss": 0.5581,
+      "epoch": 6.095146947287221,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2331154288041186e-05,
+      "loss": 0.9980406951904297,
       "step": 529400
     },
     {
-      "epoch": 12.19489730128028,
-      "grad_norm": 3.366831064224243,
-      "learning_rate": 1.2330863425488286e-05,
-      "loss": 0.5556,
+      "epoch": 6.097449632361936,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.232388272421867e-05,
+      "loss": 0.9931936645507813,
       "step": 529600
     },
     {
-      "epoch": 12.19950262503454,
-      "grad_norm": 3.2023255825042725,
-      "learning_rate": 1.2323628219484883e-05,
-      "loss": 0.5535,
+      "epoch": 6.099752317436651,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2316611160396156e-05,
+      "loss": 0.9748165130615234,
       "step": 529800
     },
     {
-      "epoch": 12.2041079487888,
-      "grad_norm": 2.8660030364990234,
-      "learning_rate": 1.2316356655662366e-05,
-      "loss": 0.5565,
+      "epoch": 6.102055002511366,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2309339596573639e-05,
+      "loss": 0.98735107421875,
       "step": 530000
     },
     {
-      "epoch": 12.20871327254306,
-      "grad_norm": 2.7351365089416504,
-      "learning_rate": 1.2309085091839852e-05,
-      "loss": 0.5524,
+      "epoch": 6.104357687586081,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2302068032751123e-05,
+      "loss": 0.976137924194336,
       "step": 530200
     },
     {
-      "epoch": 12.21331859629732,
-      "grad_norm": 3.0572710037231445,
-      "learning_rate": 1.2301813528017336e-05,
-      "loss": 0.5538,
+      "epoch": 6.106660372660796,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2294796468928609e-05,
+      "loss": 0.9755792236328125,
       "step": 530400
     },
     {
-      "epoch": 12.217923920051579,
-      "grad_norm": 3.4272561073303223,
-      "learning_rate": 1.229454196419482e-05,
-      "loss": 0.5479,
+      "epoch": 6.108963057735511,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2287524905106091e-05,
+      "loss": 0.9658280181884765,
       "step": 530600
     },
     {
-      "epoch": 12.22252924380584,
-      "grad_norm": 2.7951786518096924,
-      "learning_rate": 1.2287270400372304e-05,
-      "loss": 0.5439,
+      "epoch": 6.111265742810225,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2280253341283577e-05,
+      "loss": 0.954208984375,
       "step": 530800
     },
     {
-      "epoch": 12.2271345675601,
-      "grad_norm": 3.222205638885498,
-      "learning_rate": 1.2279998836549789e-05,
-      "loss": 0.5448,
+      "epoch": 6.113568427884941,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2272981777461062e-05,
+      "loss": 0.9845800018310547,
       "step": 531000
     },
     {
-      "epoch": 12.23173989131436,
-      "grad_norm": 2.7026097774505615,
-      "learning_rate": 1.2272727272727274e-05,
-      "loss": 0.5442,
+      "epoch": 6.115871112959655,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.2265710213638544e-05,
+      "loss": 0.9828291320800782,
       "step": 531200
     },
     {
-      "epoch": 12.236345215068619,
-      "grad_norm": 3.955737590789795,
-      "learning_rate": 1.2265455708904757e-05,
-      "loss": 0.5558,
+      "epoch": 6.118173798034371,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.225843864981603e-05,
+      "loss": 0.9656654357910156,
       "step": 531400
     },
     {
-      "epoch": 12.24095053882288,
-      "grad_norm": 3.2188639640808105,
-      "learning_rate": 1.2258184145082241e-05,
-      "loss": 0.5434,
+      "epoch": 6.120476483109085,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2251167085993514e-05,
+      "loss": 0.9733367156982422,
       "step": 531600
     },
     {
-      "epoch": 12.24555586257714,
-      "grad_norm": 2.410057783126831,
-      "learning_rate": 1.2250912581259727e-05,
-      "loss": 0.5655,
+      "epoch": 6.122779168183801,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2243895522170998e-05,
+      "loss": 0.9854462432861328,
       "step": 531800
     },
     {
-      "epoch": 12.250161186331399,
-      "grad_norm": 2.9907710552215576,
-      "learning_rate": 1.224364101743721e-05,
-      "loss": 0.5488,
+      "epoch": 6.125081853258515,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2236623958348483e-05,
+      "loss": 0.9575458526611328,
       "step": 532000
     },
     {
-      "epoch": 12.254766510085659,
-      "grad_norm": 3.5246376991271973,
-      "learning_rate": 1.2236369453614696e-05,
-      "loss": 0.5616,
+      "epoch": 6.127384538333231,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2229352394525967e-05,
+      "loss": 0.9649818420410157,
       "step": 532200
     },
     {
-      "epoch": 12.25937183383992,
-      "grad_norm": 3.297806978225708,
-      "learning_rate": 1.222909788979218e-05,
-      "loss": 0.5688,
+      "epoch": 6.129687223407945,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2222080830703453e-05,
+      "loss": 0.9760279083251953,
       "step": 532400
     },
     {
-      "epoch": 12.263977157594178,
-      "grad_norm": 2.848536729812622,
-      "learning_rate": 1.2221826325969662e-05,
-      "loss": 0.5417,
+      "epoch": 6.13198990848266,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2214809266880935e-05,
+      "loss": 1.0074124145507812,
       "step": 532600
     },
     {
-      "epoch": 12.268582481348439,
-      "grad_norm": 2.740184783935547,
-      "learning_rate": 1.2214554762147148e-05,
-      "loss": 0.558,
+      "epoch": 6.134292593557375,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.220753770305842e-05,
+      "loss": 0.9710540008544922,
       "step": 532800
     },
     {
-      "epoch": 12.2731878051027,
-      "grad_norm": 3.5397231578826904,
-      "learning_rate": 1.2207283198324632e-05,
-      "loss": 0.5495,
+      "epoch": 6.13659527863209,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2200266139235905e-05,
+      "loss": 0.9692363739013672,
       "step": 533000
     },
     {
-      "epoch": 12.277793128856958,
-      "grad_norm": 2.4408740997314453,
-      "learning_rate": 1.220004799232123e-05,
-      "loss": 0.5511,
+      "epoch": 6.138897963706805,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2192994575413388e-05,
+      "loss": 0.9787443542480468,
       "step": 533200
     },
     {
-      "epoch": 12.282398452611218,
-      "grad_norm": 3.476637363433838,
-      "learning_rate": 1.2192776428498712e-05,
-      "loss": 0.5498,
+      "epoch": 6.14120064878152,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2185723011590872e-05,
+      "loss": 0.9511615753173828,
       "step": 533400
     },
     {
-      "epoch": 12.287003776365479,
-      "grad_norm": 3.668010711669922,
-      "learning_rate": 1.2185504864676198e-05,
-      "loss": 0.549,
+      "epoch": 6.143503333856235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2178451447768358e-05,
+      "loss": 0.9663115692138672,
       "step": 533600
     },
     {
-      "epoch": 12.291609100119738,
-      "grad_norm": 3.334040641784668,
-      "learning_rate": 1.2178233300853682e-05,
-      "loss": 0.5352,
+      "epoch": 6.14580601893095,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.217117988394584e-05,
+      "loss": 0.9776692962646485,
       "step": 533800
     },
     {
-      "epoch": 12.296214423873998,
-      "grad_norm": 2.8002822399139404,
-      "learning_rate": 1.2170961737031166e-05,
-      "loss": 0.5501,
+      "epoch": 6.148108704005665,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2163908320123327e-05,
+      "loss": 0.983729248046875,
       "step": 534000
     },
     {
-      "epoch": 12.300819747628259,
-      "grad_norm": 2.5833489894866943,
-      "learning_rate": 1.216369017320865e-05,
-      "loss": 0.5483,
+      "epoch": 6.1504113890803795,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.215663675630081e-05,
+      "loss": 0.9854757690429687,
       "step": 534200
     },
     {
-      "epoch": 12.305425071382519,
-      "grad_norm": 3.02528977394104,
-      "learning_rate": 1.2156418609386135e-05,
-      "loss": 0.5439,
+      "epoch": 6.152714074155094,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2149365192478293e-05,
+      "loss": 0.9797981262207032,
       "step": 534400
     },
     {
-      "epoch": 12.310030395136778,
-      "grad_norm": 3.139838933944702,
-      "learning_rate": 1.214914704556362e-05,
-      "loss": 0.5504,
+      "epoch": 6.1550167592298095,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.214209362865578e-05,
+      "loss": 0.9856068420410157,
       "step": 534600
     },
     {
-      "epoch": 12.314635718891038,
-      "grad_norm": 2.3408186435699463,
-      "learning_rate": 1.2141875481741103e-05,
-      "loss": 0.5431,
+      "epoch": 6.157319444304524,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2134822064833263e-05,
+      "loss": 0.9999673461914063,
       "step": 534800
     },
     {
-      "epoch": 12.319241042645299,
-      "grad_norm": 3.896301746368408,
-      "learning_rate": 1.2134603917918587e-05,
-      "loss": 0.5497,
+      "epoch": 6.1596221293792395,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2127550501010748e-05,
+      "loss": 0.9946244049072266,
       "step": 535000
     },
     {
-      "epoch": 12.323846366399557,
-      "grad_norm": 3.0924932956695557,
-      "learning_rate": 1.2127332354096073e-05,
-      "loss": 0.5636,
+      "epoch": 6.161924814453954,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2120278937188232e-05,
+      "loss": 0.9824993133544921,
       "step": 535200
     },
     {
-      "epoch": 12.328451690153818,
-      "grad_norm": 3.347902297973633,
-      "learning_rate": 1.2120060790273556e-05,
-      "loss": 0.5617,
+      "epoch": 6.164227499528669,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.2113007373365716e-05,
+      "loss": 0.9499840545654297,
       "step": 535400
     },
     {
-      "epoch": 12.333057013908078,
-      "grad_norm": 3.351632833480835,
-      "learning_rate": 1.2112789226451042e-05,
-      "loss": 0.5495,
+      "epoch": 6.166530184603384,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2105735809543202e-05,
+      "loss": 0.9955195617675782,
       "step": 535600
     },
     {
-      "epoch": 12.337662337662337,
-      "grad_norm": 3.4726133346557617,
-      "learning_rate": 1.2105517662628526e-05,
-      "loss": 0.563,
+      "epoch": 6.168832869678099,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.2098464245720685e-05,
+      "loss": 0.9737226104736328,
       "step": 535800
     },
     {
-      "epoch": 12.342267661416598,
-      "grad_norm": 3.684715509414673,
-      "learning_rate": 1.2098246098806008e-05,
-      "loss": 0.548,
+      "epoch": 6.171135554752814,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.2091192681898169e-05,
+      "loss": 0.9871718597412109,
       "step": 536000
     },
     {
-      "epoch": 12.346872985170858,
-      "grad_norm": 2.476365327835083,
-      "learning_rate": 1.2091010892802605e-05,
-      "loss": 0.5503,
+      "epoch": 6.1734382398275285,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2083921118075655e-05,
+      "loss": 0.9826788330078124,
       "step": 536200
     },
     {
-      "epoch": 12.351478308925117,
-      "grad_norm": 3.3055906295776367,
-      "learning_rate": 1.2083739328980091e-05,
-      "loss": 0.5412,
+      "epoch": 6.175740924902244,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2076649554253137e-05,
+      "loss": 1.0013955688476563,
       "step": 536400
     },
     {
-      "epoch": 12.356083632679377,
-      "grad_norm": 3.4834065437316895,
-      "learning_rate": 1.2076467765157575e-05,
-      "loss": 0.5416,
+      "epoch": 6.1780436099769585,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2069377990430623e-05,
+      "loss": 0.9916542053222657,
       "step": 536600
     },
     {
-      "epoch": 12.360688956433638,
-      "grad_norm": 3.004270553588867,
-      "learning_rate": 1.2069196201335058e-05,
-      "loss": 0.5557,
+      "epoch": 6.180346295051674,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2062106426608107e-05,
+      "loss": 0.9936505889892578,
       "step": 536800
     },
     {
-      "epoch": 12.365294280187896,
-      "grad_norm": 2.7978525161743164,
-      "learning_rate": 1.2061924637512544e-05,
-      "loss": 0.5507,
+      "epoch": 6.182648980126388,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.205483486278559e-05,
+      "loss": 0.9823659515380859,
       "step": 537000
     },
     {
-      "epoch": 12.369899603942157,
-      "grad_norm": 3.2909445762634277,
-      "learning_rate": 1.2054653073690028e-05,
-      "loss": 0.5573,
+      "epoch": 6.184951665201104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2047563298963076e-05,
+      "loss": 0.974571762084961,
       "step": 537200
     },
     {
-      "epoch": 12.374504927696417,
-      "grad_norm": 3.068385124206543,
-      "learning_rate": 1.2047381509867512e-05,
-      "loss": 0.561,
+      "epoch": 6.187254350275818,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.204029173514056e-05,
+      "loss": 0.9920982360839844,
       "step": 537400
     },
     {
-      "epoch": 12.379110251450676,
-      "grad_norm": 3.6887574195861816,
-      "learning_rate": 1.2040109946044997e-05,
-      "loss": 0.5581,
+      "epoch": 6.189557035350534,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2033020171318042e-05,
+      "loss": 0.9927661895751954,
       "step": 537600
     },
     {
-      "epoch": 12.383715575204937,
-      "grad_norm": 3.1257615089416504,
-      "learning_rate": 1.203283838222248e-05,
-      "loss": 0.5471,
+      "epoch": 6.191859720425248,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2025748607495528e-05,
+      "loss": 0.9779354858398438,
       "step": 537800
     },
     {
-      "epoch": 12.388320898959197,
-      "grad_norm": 2.885606527328491,
-      "learning_rate": 1.2025566818399967e-05,
-      "loss": 0.558,
+      "epoch": 6.194162405499963,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2018477043673013e-05,
+      "loss": 0.9786163330078125,
       "step": 538000
     },
     {
-      "epoch": 12.392926222713458,
-      "grad_norm": 4.5534138679504395,
-      "learning_rate": 1.201829525457745e-05,
-      "loss": 0.5434,
+      "epoch": 6.196465090574678,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2011205479850499e-05,
+      "loss": 0.9637989807128906,
       "step": 538200
     },
     {
-      "epoch": 12.397531546467716,
-      "grad_norm": 4.204901695251465,
-      "learning_rate": 1.2011023690754933e-05,
-      "loss": 0.5512,
+      "epoch": 6.198767775649393,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2003933916027981e-05,
+      "loss": 0.9681629943847656,
       "step": 538400
     },
     {
-      "epoch": 12.402136870221977,
-      "grad_norm": 3.7075207233428955,
-      "learning_rate": 1.200375212693242e-05,
-      "loss": 0.5446,
+      "epoch": 6.201070460724108,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1996662352205465e-05,
+      "loss": 0.9763223266601563,
       "step": 538600
     },
     {
-      "epoch": 12.406742193976237,
-      "grad_norm": 2.795241117477417,
-      "learning_rate": 1.1996480563109902e-05,
-      "loss": 0.5551,
+      "epoch": 6.203373145798823,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1989390788382951e-05,
+      "loss": 0.9757502746582031,
       "step": 538800
     },
     {
-      "epoch": 12.411347517730496,
-      "grad_norm": 3.3998024463653564,
-      "learning_rate": 1.1989208999287388e-05,
-      "loss": 0.5498,
+      "epoch": 6.205675830873538,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1982119224560434e-05,
+      "loss": 1.003903350830078,
       "step": 539000
     },
     {
-      "epoch": 12.415952841484756,
-      "grad_norm": 2.8881683349609375,
-      "learning_rate": 1.1981937435464872e-05,
-      "loss": 0.5478,
+      "epoch": 6.207978515948253,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1974847660737918e-05,
+      "loss": 0.9690301513671875,
       "step": 539200
     },
     {
-      "epoch": 12.420558165239017,
-      "grad_norm": 3.143094062805176,
-      "learning_rate": 1.1974665871642354e-05,
-      "loss": 0.5445,
+      "epoch": 6.210281201022968,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1967576096915404e-05,
+      "loss": 0.9842018127441406,
       "step": 539400
     },
     {
-      "epoch": 12.425163488993276,
-      "grad_norm": 3.2156903743743896,
-      "learning_rate": 1.196739430781984e-05,
-      "loss": 0.5572,
+      "epoch": 6.212583886097683,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1960304533092886e-05,
+      "loss": 0.9938206481933594,
       "step": 539600
     },
     {
-      "epoch": 12.429768812747536,
-      "grad_norm": 2.0139708518981934,
-      "learning_rate": 1.1960122743997325e-05,
-      "loss": 0.5578,
+      "epoch": 6.214886571172397,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1953032969270372e-05,
+      "loss": 0.9942153930664063,
       "step": 539800
     },
     {
-      "epoch": 12.434374136501797,
-      "grad_norm": 2.9531643390655518,
-      "learning_rate": 1.1952851180174809e-05,
-      "loss": 0.5555,
+      "epoch": 6.217189256247113,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1945761405447857e-05,
+      "loss": 0.9966661834716797,
       "step": 540000
     },
     {
-      "epoch": 12.438979460256055,
-      "grad_norm": 2.786808490753174,
-      "learning_rate": 1.1945579616352293e-05,
-      "loss": 0.5339,
+      "epoch": 6.219491941321827,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1938489841625339e-05,
+      "loss": 0.972867660522461,
       "step": 540200
     },
     {
-      "epoch": 12.443584784010316,
-      "grad_norm": 3.243690013885498,
-      "learning_rate": 1.1938308052529777e-05,
-      "loss": 0.5459,
+      "epoch": 6.221794626396543,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.1931218277802825e-05,
+      "loss": 0.9939933776855469,
       "step": 540400
     },
     {
-      "epoch": 12.448190107764576,
-      "grad_norm": 3.303994655609131,
-      "learning_rate": 1.1931036488707262e-05,
-      "loss": 0.5634,
+      "epoch": 6.224097311471257,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.192394671398031e-05,
+      "loss": 0.9651322174072265,
       "step": 540600
     },
     {
-      "epoch": 12.452795431518835,
-      "grad_norm": 4.073940277099609,
-      "learning_rate": 1.1923764924884746e-05,
-      "loss": 0.5491,
+      "epoch": 6.226399996545973,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1916675150157792e-05,
+      "loss": 0.9611105346679687,
       "step": 540800
     },
     {
-      "epoch": 12.457400755273095,
-      "grad_norm": 2.980377435684204,
-      "learning_rate": 1.191649336106223e-05,
-      "loss": 0.5501,
+      "epoch": 6.228702681620687,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1909403586335278e-05,
+      "loss": 0.9767247009277343,
       "step": 541000
     },
     {
-      "epoch": 12.462006079027356,
-      "grad_norm": 3.772876501083374,
-      "learning_rate": 1.1909221797239716e-05,
-      "loss": 0.5487,
+      "epoch": 6.231005366695403,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1902132022512762e-05,
+      "loss": 0.9699462127685546,
       "step": 541200
     },
     {
-      "epoch": 12.466611402781616,
-      "grad_norm": 2.804576873779297,
-      "learning_rate": 1.1901950233417198e-05,
-      "loss": 0.5479,
+      "epoch": 6.233308051770117,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.1894860458690246e-05,
+      "loss": 0.9935227966308594,
       "step": 541400
     },
     {
-      "epoch": 12.471216726535875,
-      "grad_norm": 3.1938371658325195,
-      "learning_rate": 1.1894678669594684e-05,
-      "loss": 0.5541,
+      "epoch": 6.235610736844832,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.188758889486773e-05,
+      "loss": 0.9764249420166016,
       "step": 541600
     },
     {
-      "epoch": 12.475822050290136,
-      "grad_norm": 2.5504708290100098,
-      "learning_rate": 1.1887407105772169e-05,
-      "loss": 0.5465,
+      "epoch": 6.237913421919547,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1880317331045214e-05,
+      "loss": 0.9721925354003906,
       "step": 541800
     },
     {
-      "epoch": 12.480427374044396,
-      "grad_norm": 2.7534172534942627,
-      "learning_rate": 1.1880135541949651e-05,
-      "loss": 0.5542,
+      "epoch": 6.240216106994262,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.18730457672227e-05,
+      "loss": 0.9650587463378906,
       "step": 542000
     },
     {
-      "epoch": 12.485032697798655,
-      "grad_norm": 3.2549421787261963,
-      "learning_rate": 1.1872863978127137e-05,
-      "loss": 0.5521,
+      "epoch": 6.242518792068977,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1865774203400183e-05,
+      "loss": 0.9611381530761719,
       "step": 542200
     },
     {
-      "epoch": 12.489638021552915,
-      "grad_norm": 3.0159831047058105,
-      "learning_rate": 1.1865592414304621e-05,
-      "loss": 0.5503,
+      "epoch": 6.244821477143692,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1858502639577669e-05,
+      "loss": 0.9796166229248047,
       "step": 542400
     },
     {
-      "epoch": 12.494243345307176,
-      "grad_norm": 2.7797956466674805,
-      "learning_rate": 1.1858357208301218e-05,
-      "loss": 0.5564,
+      "epoch": 6.247124162218407,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1851231075755153e-05,
+      "loss": 0.9999110412597656,
       "step": 542600
     },
     {
-      "epoch": 12.498848669061434,
-      "grad_norm": 3.071063995361328,
-      "learning_rate": 1.18510856444787e-05,
-      "loss": 0.5411,
+      "epoch": 6.249426847293122,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1843959511932636e-05,
+      "loss": 0.9900628662109375,
       "step": 542800
     },
     {
-      "epoch": 12.503453992815695,
-      "grad_norm": 4.068755626678467,
-      "learning_rate": 1.1843814080656187e-05,
-      "loss": 0.5572,
+      "epoch": 6.251729532367837,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1836687948110121e-05,
+      "loss": 0.9753203582763672,
       "step": 543000
     },
     {
-      "epoch": 12.508059316569955,
-      "grad_norm": 3.225297689437866,
-      "learning_rate": 1.183654251683367e-05,
-      "loss": 0.5508,
+      "epoch": 6.2540322174425516,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.1829416384287606e-05,
+      "loss": 0.96914306640625,
       "step": 543200
     },
     {
-      "epoch": 12.512664640324214,
-      "grad_norm": 2.821578025817871,
-      "learning_rate": 1.1829307310830268e-05,
-      "loss": 0.5474,
+      "epoch": 6.256334902517267,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1822144820465088e-05,
+      "loss": 0.9859267425537109,
       "step": 543400
     },
     {
-      "epoch": 12.517269964078475,
-      "grad_norm": 5.399388313293457,
-      "learning_rate": 1.1822035747007752e-05,
-      "loss": 0.5504,
+      "epoch": 6.2586375875919815,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.1814873256642574e-05,
+      "loss": 0.9566590881347656,
       "step": 543600
     },
     {
-      "epoch": 12.521875287832735,
-      "grad_norm": 3.5580897331237793,
-      "learning_rate": 1.1814764183185236e-05,
-      "loss": 0.5641,
+      "epoch": 6.260940272666696,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1807601692820058e-05,
+      "loss": 0.9894088745117188,
       "step": 543800
     },
     {
-      "epoch": 12.526480611586994,
-      "grad_norm": 3.240264892578125,
-      "learning_rate": 1.180749261936272e-05,
-      "loss": 0.5511,
+      "epoch": 6.2632429577414115,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1800330128997543e-05,
+      "loss": 0.9496869659423828,
       "step": 544000
     },
     {
-      "epoch": 12.531085935341254,
-      "grad_norm": 3.2633004188537598,
-      "learning_rate": 1.1800221055540205e-05,
-      "loss": 0.5504,
+      "epoch": 6.265545642816126,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1793058565175027e-05,
+      "loss": 0.9789578247070313,
       "step": 544200
     },
     {
-      "epoch": 12.535691259095515,
-      "grad_norm": 3.4117608070373535,
-      "learning_rate": 1.1792949491717689e-05,
-      "loss": 0.5429,
+      "epoch": 6.2678483278908415,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1785787001352511e-05,
+      "loss": 0.9891857147216797,
       "step": 544400
     },
     {
-      "epoch": 12.540296582849773,
-      "grad_norm": 3.1878907680511475,
-      "learning_rate": 1.1785677927895173e-05,
-      "loss": 0.5544,
+      "epoch": 6.270151012965556,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.1778515437529995e-05,
+      "loss": 0.9948699188232422,
       "step": 544600
     },
     {
-      "epoch": 12.544901906604034,
-      "grad_norm": 3.49251389503479,
-      "learning_rate": 1.1778406364072659e-05,
-      "loss": 0.5402,
+      "epoch": 6.272453698040271,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.177124387370748e-05,
+      "loss": 0.9798444366455078,
       "step": 544800
     },
     {
-      "epoch": 12.549507230358294,
-      "grad_norm": 3.136068820953369,
-      "learning_rate": 1.1771134800250141e-05,
-      "loss": 0.5617,
+      "epoch": 6.274756383114986,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1763972309884964e-05,
+      "loss": 0.9842873382568359,
       "step": 545000
     },
     {
-      "epoch": 12.554112554112555,
-      "grad_norm": 3.5267179012298584,
-      "learning_rate": 1.1763863236427627e-05,
-      "loss": 0.5583,
+      "epoch": 6.2770590681897005,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.175670074606245e-05,
+      "loss": 0.9862162017822266,
       "step": 545200
     },
     {
-      "epoch": 12.558717877866814,
-      "grad_norm": 2.639244794845581,
-      "learning_rate": 1.1756591672605112e-05,
-      "loss": 0.5556,
+      "epoch": 6.279361753264416,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1749429182239932e-05,
+      "loss": 1.006587600708008,
       "step": 545400
     },
     {
-      "epoch": 12.563323201621074,
-      "grad_norm": 2.4130160808563232,
-      "learning_rate": 1.1749356466601708e-05,
-      "loss": 0.5378,
+      "epoch": 6.2816644383391305,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1742157618417418e-05,
+      "loss": 0.9883873748779297,
       "step": 545600
     },
     {
-      "epoch": 12.567928525375335,
-      "grad_norm": 3.0723562240600586,
-      "learning_rate": 1.1742084902779191e-05,
-      "loss": 0.5493,
+      "epoch": 6.283967123413846,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.1734886054594902e-05,
+      "loss": 0.982315673828125,
       "step": 545800
     },
     {
-      "epoch": 12.572533849129593,
-      "grad_norm": 3.0826873779296875,
-      "learning_rate": 1.1734813338956677e-05,
-      "loss": 0.5478,
+      "epoch": 6.28626980848856,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1727614490772385e-05,
+      "loss": 0.9863622283935547,
       "step": 546000
     },
     {
-      "epoch": 12.577139172883854,
-      "grad_norm": 3.6943721771240234,
-      "learning_rate": 1.1727541775134161e-05,
-      "loss": 0.552,
+      "epoch": 6.288572493563276,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.172034292694987e-05,
+      "loss": 0.9699191284179688,
       "step": 546200
     },
     {
-      "epoch": 12.581744496638114,
-      "grad_norm": 3.1983044147491455,
-      "learning_rate": 1.1720270211311644e-05,
-      "loss": 0.5533,
+      "epoch": 6.29087517863799,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1713071363127355e-05,
+      "loss": 0.9829322052001953,
       "step": 546400
     },
     {
-      "epoch": 12.586349820392373,
-      "grad_norm": 2.8163883686065674,
-      "learning_rate": 1.171299864748913e-05,
-      "loss": 0.5551,
+      "epoch": 6.293177863712706,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1705799799304837e-05,
+      "loss": 0.9939772796630859,
       "step": 546600
     },
     {
-      "epoch": 12.590955144146633,
-      "grad_norm": 3.1819756031036377,
-      "learning_rate": 1.1705727083666614e-05,
-      "loss": 0.5637,
+      "epoch": 6.29548054878742,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1698528235482323e-05,
+      "loss": 0.9853797149658203,
       "step": 546800
     },
     {
-      "epoch": 12.595560467900894,
-      "grad_norm": 3.5395870208740234,
-      "learning_rate": 1.1698455519844098e-05,
-      "loss": 0.5517,
+      "epoch": 6.297783233862136,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1691256671659808e-05,
+      "loss": 0.9782255554199218,
       "step": 547000
     },
     {
-      "epoch": 12.600165791655153,
-      "grad_norm": 2.773688793182373,
-      "learning_rate": 1.1691183956021582e-05,
-      "loss": 0.5692,
+      "epoch": 6.30008591893685,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1683985107837292e-05,
+      "loss": 0.9772499847412109,
       "step": 547200
     },
     {
-      "epoch": 12.604771115409413,
-      "grad_norm": 3.0060763359069824,
-      "learning_rate": 1.1683912392199066e-05,
-      "loss": 0.5571,
+      "epoch": 6.302388604011565,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1676713544014776e-05,
+      "loss": 0.967388916015625,
       "step": 547400
     },
     {
-      "epoch": 12.609376439163674,
-      "grad_norm": 3.25728702545166,
-      "learning_rate": 1.167664082837655e-05,
-      "loss": 0.5608,
+      "epoch": 6.30469128908628,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.166944198019226e-05,
+      "loss": 0.9973257446289062,
       "step": 547600
     },
     {
-      "epoch": 12.613981762917934,
-      "grad_norm": 3.356100559234619,
-      "learning_rate": 1.1669369264554035e-05,
-      "loss": 0.5368,
+      "epoch": 6.306993974160995,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1662170416369744e-05,
+      "loss": 1.0004740142822266,
       "step": 547800
     },
     {
-      "epoch": 12.618587086672193,
-      "grad_norm": 3.625145435333252,
-      "learning_rate": 1.166209770073152e-05,
-      "loss": 0.538,
+      "epoch": 6.30929665923571,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1654898852547229e-05,
+      "loss": 0.9807405090332031,
       "step": 548000
     },
     {
-      "epoch": 12.623192410426453,
-      "grad_norm": 2.9204094409942627,
-      "learning_rate": 1.1654826136909003e-05,
-      "loss": 0.546,
+      "epoch": 6.311599344310425,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1647627288724713e-05,
+      "loss": 0.9859786987304687,
       "step": 548200
     },
     {
-      "epoch": 12.627797734180714,
-      "grad_norm": 3.6132349967956543,
-      "learning_rate": 1.1647590930905602e-05,
-      "loss": 0.5592,
+      "epoch": 6.31390202938514,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1640355724902199e-05,
+      "loss": 0.969800796508789,
       "step": 548400
     },
     {
-      "epoch": 12.632403057934972,
-      "grad_norm": 3.495785713195801,
-      "learning_rate": 1.1640319367083084e-05,
-      "loss": 0.539,
+      "epoch": 6.316204714459855,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1633084161079681e-05,
+      "loss": 0.9897576141357421,
       "step": 548600
     },
     {
-      "epoch": 12.637008381689233,
-      "grad_norm": 3.632181167602539,
-      "learning_rate": 1.163304780326057e-05,
-      "loss": 0.5363,
+      "epoch": 6.31850739953457,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.1625812597257167e-05,
+      "loss": 0.9738155364990234,
       "step": 548800
     },
     {
-      "epoch": 12.641613705443493,
-      "grad_norm": 3.3912816047668457,
-      "learning_rate": 1.1625776239438055e-05,
-      "loss": 0.554,
+      "epoch": 6.320810084609285,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1618541033434651e-05,
+      "loss": 1.0006730651855469,
       "step": 549000
     },
     {
-      "epoch": 12.646219029197752,
-      "grad_norm": 3.4610326290130615,
-      "learning_rate": 1.1618504675615537e-05,
-      "loss": 0.5537,
+      "epoch": 6.323112769683999,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.1611269469612134e-05,
+      "loss": 1.010643768310547,
       "step": 549200
     },
     {
-      "epoch": 12.650824352952013,
-      "grad_norm": 3.4539434909820557,
-      "learning_rate": 1.1611233111793023e-05,
-      "loss": 0.5621,
+      "epoch": 6.325415454758715,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.160399790578962e-05,
+      "loss": 0.9987123870849609,
       "step": 549400
     },
     {
-      "epoch": 12.655429676706273,
-      "grad_norm": 3.00600528717041,
-      "learning_rate": 1.160399790578962e-05,
-      "loss": 0.5558,
+      "epoch": 6.327718139833429,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1596726341967104e-05,
+      "loss": 0.9761300659179688,
       "step": 549600
     },
     {
-      "epoch": 12.660035000460532,
-      "grad_norm": 3.060939311981201,
-      "learning_rate": 1.1596726341967104e-05,
-      "loss": 0.561,
+      "epoch": 6.330020824908145,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1589454778144588e-05,
+      "loss": 0.9792920684814453,
       "step": 549800
     },
     {
-      "epoch": 12.664640324214792,
-      "grad_norm": 3.2902770042419434,
-      "learning_rate": 1.1589454778144588e-05,
-      "loss": 0.5428,
+      "epoch": 6.332323509982859,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1582183214322073e-05,
+      "loss": 0.9785138702392578,
       "step": 550000
     },
     {
-      "epoch": 12.669245647969053,
-      "grad_norm": 3.1374452114105225,
-      "learning_rate": 1.1582183214322073e-05,
-      "loss": 0.5523,
+      "epoch": 6.334626195057575,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1574911650499557e-05,
+      "loss": 0.9807654571533203,
       "step": 550200
     },
     {
-      "epoch": 12.673850971723311,
-      "grad_norm": 2.6760544776916504,
-      "learning_rate": 1.1574911650499557e-05,
-      "loss": 0.555,
+      "epoch": 6.336928880132289,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1567640086677041e-05,
+      "loss": 0.9899097442626953,
       "step": 550400
     },
     {
-      "epoch": 12.678456295477572,
-      "grad_norm": 3.863650321960449,
-      "learning_rate": 1.1567640086677041e-05,
-      "loss": 0.5525,
+      "epoch": 6.339231565207005,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.1560368522854525e-05,
+      "loss": 0.9844416046142578,
       "step": 550600
     },
     {
-      "epoch": 12.683061619231832,
-      "grad_norm": 2.675981044769287,
-      "learning_rate": 1.1560368522854525e-05,
-      "loss": 0.5486,
+      "epoch": 6.341534250281719,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.155309695903201e-05,
+      "loss": 0.986839370727539,
       "step": 550800
     },
     {
-      "epoch": 12.687666942986091,
-      "grad_norm": 3.3591291904449463,
-      "learning_rate": 1.155309695903201e-05,
-      "loss": 0.5461,
+      "epoch": 6.343836935356434,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1545825395209494e-05,
+      "loss": 1.0087448120117188,
       "step": 551000
     },
     {
-      "epoch": 12.692272266740352,
-      "grad_norm": 2.852849006652832,
-      "learning_rate": 1.1545825395209494e-05,
-      "loss": 0.5527,
+      "epoch": 6.346139620431149,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1538553831386978e-05,
+      "loss": 0.9610017395019531,
       "step": 551200
     },
     {
-      "epoch": 12.696877590494612,
-      "grad_norm": 2.850482940673828,
-      "learning_rate": 1.1538553831386978e-05,
-      "loss": 0.5474,
+      "epoch": 6.348442305505864,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1531282267564464e-05,
+      "loss": 1.00368896484375,
       "step": 551400
     },
     {
-      "epoch": 12.70148291424887,
-      "grad_norm": 3.2086989879608154,
-      "learning_rate": 1.1531282267564464e-05,
-      "loss": 0.549,
+      "epoch": 6.350744990580579,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.1524010703741946e-05,
+      "loss": 0.9645812225341797,
       "step": 551600
     },
     {
-      "epoch": 12.706088238003131,
-      "grad_norm": 3.074154853820801,
-      "learning_rate": 1.152404706156106e-05,
-      "loss": 0.559,
+      "epoch": 6.353047675655294,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.151673913991943e-05,
+      "loss": 0.9886111450195313,
       "step": 551800
     },
     {
-      "epoch": 12.710693561757392,
-      "grad_norm": 3.1534781455993652,
-      "learning_rate": 1.1516775497738545e-05,
-      "loss": 0.5647,
+      "epoch": 6.355350360730009,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1509467576096916e-05,
+      "loss": 1.0107182312011718,
       "step": 552000
     },
     {
-      "epoch": 12.715298885511652,
-      "grad_norm": 3.174724817276001,
-      "learning_rate": 1.1509503933916027e-05,
-      "loss": 0.5527,
+      "epoch": 6.357653045804724,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.15021960122744e-05,
+      "loss": 0.9821375274658203,
       "step": 552200
     },
     {
-      "epoch": 12.719904209265911,
-      "grad_norm": 2.7721035480499268,
-      "learning_rate": 1.1502232370093513e-05,
-      "loss": 0.5488,
+      "epoch": 6.359955730879439,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.1494924448451883e-05,
+      "loss": 0.9769761657714844,
       "step": 552400
     },
     {
-      "epoch": 12.724509533020171,
-      "grad_norm": 3.360368251800537,
-      "learning_rate": 1.1494960806270998e-05,
-      "loss": 0.5643,
+      "epoch": 6.3622584159541535,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1487652884629369e-05,
+      "loss": 0.9906520080566407,
       "step": 552600
     },
     {
-      "epoch": 12.729114856774432,
-      "grad_norm": 3.283878803253174,
-      "learning_rate": 1.148768924244848e-05,
-      "loss": 0.5432,
+      "epoch": 6.364561101028868,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1480381320806853e-05,
+      "loss": 0.9958131408691406,
       "step": 552800
     },
     {
-      "epoch": 12.73372018052869,
-      "grad_norm": 3.110537052154541,
-      "learning_rate": 1.1480417678625966e-05,
-      "loss": 0.555,
+      "epoch": 6.3668637861035835,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.1473109756984338e-05,
+      "loss": 0.9820102691650391,
       "step": 553000
     },
     {
-      "epoch": 12.738325504282951,
-      "grad_norm": 4.010375499725342,
-      "learning_rate": 1.147314611480345e-05,
-      "loss": 0.5433,
+      "epoch": 6.369166471178298,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1465838193161822e-05,
+      "loss": 1.0095059204101562,
       "step": 553200
     },
     {
-      "epoch": 12.742930828037212,
-      "grad_norm": 3.154646873474121,
-      "learning_rate": 1.1465874550980934e-05,
-      "loss": 0.5519,
+      "epoch": 6.3714691562530135,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1458566629339306e-05,
+      "loss": 0.9783484649658203,
       "step": 553400
     },
     {
-      "epoch": 12.74753615179147,
-      "grad_norm": 3.806518077850342,
-      "learning_rate": 1.1458602987158419e-05,
-      "loss": 0.5567,
+      "epoch": 6.373771841327728,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.145129506551679e-05,
+      "loss": 0.9865557098388672,
       "step": 553600
     },
     {
-      "epoch": 12.75214147554573,
-      "grad_norm": 3.772310972213745,
-      "learning_rate": 1.1451331423335903e-05,
-      "loss": 0.5504,
+      "epoch": 6.376074526402443,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.1444023501694274e-05,
+      "loss": 1.0026549530029296,
       "step": 553800
     },
     {
-      "epoch": 12.756746799299991,
-      "grad_norm": 3.2461366653442383,
-      "learning_rate": 1.1444059859513387e-05,
-      "loss": 0.5575,
+      "epoch": 6.378377211477158,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1436751937871759e-05,
+      "loss": 0.9793618774414062,
       "step": 554000
     },
     {
-      "epoch": 12.76135212305425,
-      "grad_norm": 3.1284291744232178,
-      "learning_rate": 1.1436788295690871e-05,
-      "loss": 0.5541,
+      "epoch": 6.380679896551873,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1429480374049243e-05,
+      "loss": 0.9879568481445312,
       "step": 554200
     },
     {
-      "epoch": 12.76595744680851,
-      "grad_norm": 3.069737434387207,
-      "learning_rate": 1.1429516731868356e-05,
-      "loss": 0.5529,
+      "epoch": 6.382982581626588,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.1422208810226727e-05,
+      "loss": 1.0095554351806642,
       "step": 554400
     },
     {
-      "epoch": 12.770562770562771,
-      "grad_norm": 3.285787582397461,
-      "learning_rate": 1.142224516804584e-05,
-      "loss": 0.5531,
+      "epoch": 6.3852852667013025,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1414937246404213e-05,
+      "loss": 0.9665096282958985,
       "step": 554600
     },
     {
-      "epoch": 12.775168094317031,
-      "grad_norm": 2.772050142288208,
-      "learning_rate": 1.1415009962042437e-05,
-      "loss": 0.5438,
+      "epoch": 6.387587951776018,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1407665682581696e-05,
+      "loss": 0.99393798828125,
       "step": 554800
     },
     {
-      "epoch": 12.77977341807129,
-      "grad_norm": 2.2192680835723877,
-      "learning_rate": 1.1407738398219921e-05,
-      "loss": 0.5319,
+      "epoch": 6.389890636850732,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.140039411875918e-05,
+      "loss": 0.9509799194335937,
       "step": 555000
     },
     {
-      "epoch": 12.78437874182555,
-      "grad_norm": 3.2888576984405518,
-      "learning_rate": 1.1400466834397407e-05,
-      "loss": 0.5462,
+      "epoch": 6.392193321925448,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.1393122554936666e-05,
+      "loss": 0.9921061706542968,
       "step": 555200
     },
     {
-      "epoch": 12.788984065579811,
-      "grad_norm": 3.075023889541626,
-      "learning_rate": 1.139319527057489e-05,
-      "loss": 0.5449,
+      "epoch": 6.394496007000162,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.138585099111415e-05,
+      "loss": 0.9750463104248047,
       "step": 555400
     },
     {
-      "epoch": 12.79358938933407,
-      "grad_norm": 3.2394826412200928,
-      "learning_rate": 1.1385923706752374e-05,
-      "loss": 0.5461,
+      "epoch": 6.396798692074878,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1378579427291634e-05,
+      "loss": 0.9920413970947266,
       "step": 555600
     },
     {
-      "epoch": 12.79819471308833,
-      "grad_norm": 3.1171531677246094,
-      "learning_rate": 1.137865214292986e-05,
-      "loss": 0.5433,
+      "epoch": 6.399101377149592,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1371307863469118e-05,
+      "loss": 1.0085044860839845,
       "step": 555800
     },
     {
-      "epoch": 12.80280003684259,
-      "grad_norm": 2.5984578132629395,
-      "learning_rate": 1.1371380579107344e-05,
-      "loss": 0.5531,
+      "epoch": 6.401404062224308,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1364036299646603e-05,
+      "loss": 0.9668448638916015,
       "step": 556000
     },
     {
-      "epoch": 12.80740536059685,
-      "grad_norm": 4.881665229797363,
-      "learning_rate": 1.1364109015284826e-05,
-      "loss": 0.5432,
+      "epoch": 6.403706747299022,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1356764735824087e-05,
+      "loss": 0.9831298828125,
       "step": 556200
     },
     {
-      "epoch": 12.81201068435111,
-      "grad_norm": 2.8773512840270996,
-      "learning_rate": 1.1356837451462312e-05,
-      "loss": 0.5667,
+      "epoch": 6.406009432373738,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.1349493172001571e-05,
+      "loss": 0.9696995544433594,
       "step": 556400
     },
     {
-      "epoch": 12.81661600810537,
-      "grad_norm": 3.3617618083953857,
-      "learning_rate": 1.1349565887639796e-05,
-      "loss": 0.5393,
+      "epoch": 6.408312117448452,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1342221608179055e-05,
+      "loss": 0.9708090209960938,
       "step": 556600
     },
     {
-      "epoch": 12.821221331859629,
-      "grad_norm": 2.5510413646698,
-      "learning_rate": 1.134229432381728e-05,
-      "loss": 0.5449,
+      "epoch": 6.410614802523167,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.133495004435654e-05,
+      "loss": 0.9818246459960938,
       "step": 556800
     },
     {
-      "epoch": 12.82582665561389,
-      "grad_norm": 3.1804873943328857,
-      "learning_rate": 1.1335022759994765e-05,
-      "loss": 0.5515,
+      "epoch": 6.412917487597882,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1327678480534024e-05,
+      "loss": 0.9953689575195312,
       "step": 557000
     },
     {
-      "epoch": 12.83043197936815,
-      "grad_norm": 3.242882490158081,
-      "learning_rate": 1.1327751196172249e-05,
-      "loss": 0.5713,
+      "epoch": 6.415220172672597,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.132040691671151e-05,
+      "loss": 0.9716912841796875,
       "step": 557200
     },
     {
-      "epoch": 12.835037303122409,
-      "grad_norm": 2.9455573558807373,
-      "learning_rate": 1.1320479632349733e-05,
-      "loss": 0.5588,
+      "epoch": 6.417522857747312,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1313135352888992e-05,
+      "loss": 0.9794044494628906,
       "step": 557400
     },
     {
-      "epoch": 12.83964262687667,
-      "grad_norm": 3.158345937728882,
-      "learning_rate": 1.131324442634633e-05,
-      "loss": 0.5454,
+      "epoch": 6.419825542822027,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1305863789066476e-05,
+      "loss": 0.9842906951904297,
       "step": 557600
     },
     {
-      "epoch": 12.84424795063093,
-      "grad_norm": 3.215794563293457,
-      "learning_rate": 1.1305972862523814e-05,
-      "loss": 0.543,
+      "epoch": 6.422128227896742,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1298592225243962e-05,
+      "loss": 0.9960566711425781,
       "step": 557800
     },
     {
-      "epoch": 12.848853274385188,
-      "grad_norm": 3.158463954925537,
-      "learning_rate": 1.12987012987013e-05,
-      "loss": 0.5543,
+      "epoch": 6.424430912971457,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1291320661421445e-05,
+      "loss": 0.9707205200195312,
       "step": 558000
     },
     {
-      "epoch": 12.853458598139449,
-      "grad_norm": 2.367264747619629,
-      "learning_rate": 1.1291429734878783e-05,
-      "loss": 0.5444,
+      "epoch": 6.426733598046171,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.1284049097598929e-05,
+      "loss": 0.9869296264648437,
       "step": 558200
     },
     {
-      "epoch": 12.85806392189371,
-      "grad_norm": 2.9616682529449463,
-      "learning_rate": 1.1284158171056267e-05,
-      "loss": 0.5505,
+      "epoch": 6.429036283120887,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.1276777533776415e-05,
+      "loss": 0.9972148895263672,
       "step": 558400
     },
     {
-      "epoch": 12.86266924564797,
-      "grad_norm": 4.787532329559326,
-      "learning_rate": 1.1276886607233753e-05,
-      "loss": 0.542,
+      "epoch": 6.431338968195601,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1269505969953899e-05,
+      "loss": 0.9682855224609375,
       "step": 558600
     },
     {
-      "epoch": 12.867274569402229,
-      "grad_norm": 2.9759278297424316,
-      "learning_rate": 1.1269615043411235e-05,
-      "loss": 0.5531,
+      "epoch": 6.433641653270317,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1262234406131383e-05,
+      "loss": 1.0019464874267578,
       "step": 558800
     },
     {
-      "epoch": 12.871879893156489,
-      "grad_norm": 3.0211524963378906,
-      "learning_rate": 1.126234347958872e-05,
-      "loss": 0.5526,
+      "epoch": 6.435944338345031,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1254962842308868e-05,
+      "loss": 0.9992044067382813,
       "step": 559000
     },
     {
-      "epoch": 12.87648521691075,
-      "grad_norm": 3.0512332916259766,
-      "learning_rate": 1.1255071915766206e-05,
-      "loss": 0.5651,
+      "epoch": 6.438247023419747,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1247691278486352e-05,
+      "loss": 0.9800476837158203,
       "step": 559200
     },
     {
-      "epoch": 12.881090540665008,
-      "grad_norm": 3.1640422344207764,
-      "learning_rate": 1.124780035194369e-05,
-      "loss": 0.5473,
+      "epoch": 6.440549708494461,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.1240419714663836e-05,
+      "loss": 0.9893782043457031,
       "step": 559400
     },
     {
-      "epoch": 12.885695864419269,
-      "grad_norm": 2.9496026039123535,
-      "learning_rate": 1.1240528788121174e-05,
-      "loss": 0.5571,
+      "epoch": 6.442852393569177,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.123314815084132e-05,
+      "loss": 0.9816644287109375,
       "step": 559600
     },
     {
-      "epoch": 12.89030118817353,
-      "grad_norm": 2.9672327041625977,
-      "learning_rate": 1.1233293582117771e-05,
-      "loss": 0.5536,
+      "epoch": 6.445155078643891,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1225876587018804e-05,
+      "loss": 0.9631548309326172,
       "step": 559800
     },
     {
-      "epoch": 12.894906511927788,
-      "grad_norm": 3.0200912952423096,
-      "learning_rate": 1.1226022018295255e-05,
-      "loss": 0.5478,
+      "epoch": 6.447457763718607,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1218605023196289e-05,
+      "loss": 0.9729603576660156,
       "step": 560000
     },
     {
-      "epoch": 12.899511835682048,
-      "grad_norm": 3.494211196899414,
-      "learning_rate": 1.121875045447274e-05,
-      "loss": 0.5472,
+      "epoch": 6.449760448793321,
+      "grad_norm": 1.0,
+      "learning_rate": 1.1211333459373773e-05,
+      "loss": 0.9714144134521484,
       "step": 560200
     },
     {
-      "epoch": 12.904117159436309,
-      "grad_norm": 3.695265293121338,
-      "learning_rate": 1.1211478890650224e-05,
-      "loss": 0.5525,
+      "epoch": 6.452063133868036,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1204061895551259e-05,
+      "loss": 0.9656848907470703,
       "step": 560400
     },
     {
-      "epoch": 12.908722483190568,
-      "grad_norm": 2.6711409091949463,
-      "learning_rate": 1.1204207326827708e-05,
-      "loss": 0.5473,
+      "epoch": 6.454365818942751,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1196790331728741e-05,
+      "loss": 0.9788755798339843,
       "step": 560600
     },
     {
-      "epoch": 12.913327806944828,
-      "grad_norm": 3.254509925842285,
-      "learning_rate": 1.1196935763005192e-05,
-      "loss": 0.5604,
+      "epoch": 6.456668504017466,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1189518767906225e-05,
+      "loss": 0.9863886260986328,
       "step": 560800
     },
     {
-      "epoch": 12.917933130699089,
-      "grad_norm": 3.699026107788086,
-      "learning_rate": 1.1189664199182676e-05,
-      "loss": 0.5427,
+      "epoch": 6.458971189092181,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1182247204083711e-05,
+      "loss": 0.9713378143310547,
       "step": 561000
     },
     {
-      "epoch": 12.922538454453347,
-      "grad_norm": 3.1557929515838623,
-      "learning_rate": 1.118239263536016e-05,
-      "loss": 0.5558,
+      "epoch": 6.461273874166896,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1174975640261194e-05,
+      "loss": 0.9846022033691406,
       "step": 561200
     },
     {
-      "epoch": 12.927143778207608,
-      "grad_norm": 3.200070381164551,
-      "learning_rate": 1.1175121071537646e-05,
-      "loss": 0.5462,
+      "epoch": 6.463576559241611,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1167704076438678e-05,
+      "loss": 0.9709203338623047,
       "step": 561400
     },
     {
-      "epoch": 12.931749101961868,
-      "grad_norm": 2.912868022918701,
-      "learning_rate": 1.1167849507715129e-05,
-      "loss": 0.5445,
+      "epoch": 6.4658792443163255,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1160432512616164e-05,
+      "loss": 0.9927424621582032,
       "step": 561600
     },
     {
-      "epoch": 12.936354425716129,
-      "grad_norm": 3.3618550300598145,
-      "learning_rate": 1.1160577943892613e-05,
-      "loss": 0.5478,
+      "epoch": 6.46818192939104,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1153160948793648e-05,
+      "loss": 0.982939224243164,
       "step": 561800
     },
     {
-      "epoch": 12.940959749470387,
-      "grad_norm": 2.8849663734436035,
-      "learning_rate": 1.1153306380070099e-05,
-      "loss": 0.5437,
+      "epoch": 6.4704846144657555,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1145889384971132e-05,
+      "loss": 0.9967118835449219,
       "step": 562000
     },
     {
-      "epoch": 12.945565073224648,
-      "grad_norm": 2.9502649307250977,
-      "learning_rate": 1.1146034816247582e-05,
-      "loss": 0.5529,
+      "epoch": 6.47278729954047,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1138617821148617e-05,
+      "loss": 0.9820690155029297,
       "step": 562200
     },
     {
-      "epoch": 12.950170396978908,
-      "grad_norm": 3.4335813522338867,
-      "learning_rate": 1.1138763252425066e-05,
-      "loss": 0.5609,
+      "epoch": 6.4750899846151855,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1131346257326101e-05,
+      "loss": 1.004434585571289,
       "step": 562400
     },
     {
-      "epoch": 12.954775720733167,
-      "grad_norm": 2.878819227218628,
-      "learning_rate": 1.1131491688602552e-05,
-      "loss": 0.5495,
+      "epoch": 6.4773926696899,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1124074693503585e-05,
+      "loss": 1.005102081298828,
       "step": 562600
     },
     {
-      "epoch": 12.959381044487428,
-      "grad_norm": 3.1835694313049316,
-      "learning_rate": 1.1124220124780036e-05,
-      "loss": 0.5511,
+      "epoch": 6.4796953547646154,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.111680312968107e-05,
+      "loss": 0.9745596313476562,
       "step": 562800
     },
     {
-      "epoch": 12.963986368241688,
-      "grad_norm": 3.6071760654449463,
-      "learning_rate": 1.111694856095752e-05,
-      "loss": 0.5501,
+      "epoch": 6.48199803983933,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1109531565858555e-05,
+      "loss": 0.9957767486572265,
       "step": 563000
     },
     {
-      "epoch": 12.968591691995947,
-      "grad_norm": 2.51938533782959,
-      "learning_rate": 1.1109676997135004e-05,
-      "loss": 0.5579,
+      "epoch": 6.484300724914045,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1102260002036038e-05,
+      "loss": 0.996214599609375,
       "step": 563200
     },
     {
-      "epoch": 12.973197015750207,
-      "grad_norm": 3.228456497192383,
-      "learning_rate": 1.1102405433312489e-05,
-      "loss": 0.5607,
+      "epoch": 6.48660340998876,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.1094988438213522e-05,
+      "loss": 0.9917230224609375,
       "step": 563400
     },
     {
-      "epoch": 12.977802339504468,
-      "grad_norm": 3.919980764389038,
-      "learning_rate": 1.1095133869489973e-05,
-      "loss": 0.5575,
+      "epoch": 6.488906095063475,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1087716874391008e-05,
+      "loss": 0.9812522888183594,
       "step": 563600
     },
     {
-      "epoch": 12.982407663258726,
-      "grad_norm": 3.511781930923462,
-      "learning_rate": 1.1087862305667457e-05,
-      "loss": 0.5527,
+      "epoch": 6.49120878013819,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.108044531056849e-05,
+      "loss": 0.9831491851806641,
       "step": 563800
     },
     {
-      "epoch": 12.987012987012987,
-      "grad_norm": 3.6616077423095703,
-      "learning_rate": 1.1080590741844941e-05,
-      "loss": 0.5486,
+      "epoch": 6.4935114652129045,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1073173746745975e-05,
+      "loss": 0.9870780181884765,
       "step": 564000
     },
     {
-      "epoch": 12.991618310767247,
-      "grad_norm": 3.4475338459014893,
-      "learning_rate": 1.1073355535841538e-05,
-      "loss": 0.5548,
+      "epoch": 6.49581415028762,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.106590218292346e-05,
+      "loss": 0.960843505859375,
       "step": 564200
     },
     {
-      "epoch": 12.996223634521506,
-      "grad_norm": 3.6094648838043213,
-      "learning_rate": 1.1066083972019022e-05,
-      "loss": 0.5447,
+      "epoch": 6.498116835362334,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1058630619100943e-05,
+      "loss": 0.9944307708740234,
       "step": 564400
     },
     {
-      "epoch": 13.0,
-      "eval_loss": 0.535080075263977,
-      "eval_runtime": 174.1646,
-      "eval_samples_per_second": 162.84,
-      "eval_steps_per_second": 10.18,
-      "step": 564564
-    },
-    {
-      "epoch": 13.000828958275767,
-      "grad_norm": 3.2227447032928467,
-      "learning_rate": 1.1058812408196507e-05,
-      "loss": 0.5408,
+      "epoch": 6.50041952043705,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1051359055278429e-05,
+      "loss": 0.9977530670166016,
       "step": 564600
     },
     {
-      "epoch": 13.005434282030027,
-      "grad_norm": 2.8370840549468994,
-      "learning_rate": 1.1051540844373992e-05,
-      "loss": 0.5382,
+      "epoch": 6.502722205511764,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1044087491455913e-05,
+      "loss": 0.9751404571533203,
       "step": 564800
     },
     {
-      "epoch": 13.010039605784286,
-      "grad_norm": 3.57401442527771,
-      "learning_rate": 1.1044269280551475e-05,
-      "loss": 0.5435,
+      "epoch": 6.50502489058648,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1036815927633396e-05,
+      "loss": 1.0031847381591796,
       "step": 565000
     },
     {
-      "epoch": 13.014644929538546,
-      "grad_norm": 3.077897310256958,
-      "learning_rate": 1.1037034074548072e-05,
-      "loss": 0.5478,
+      "epoch": 6.507327575661194,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1029544363810882e-05,
+      "loss": 0.9857213592529297,
       "step": 565200
     },
     {
-      "epoch": 13.019250253292807,
-      "grad_norm": 3.1777496337890625,
-      "learning_rate": 1.1029762510725556e-05,
-      "loss": 0.5281,
+      "epoch": 6.509630260735909,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1022272799988366e-05,
+      "loss": 0.9579655456542969,
       "step": 565400
     },
     {
-      "epoch": 13.023855577047067,
-      "grad_norm": 3.8741352558135986,
-      "learning_rate": 1.1022490946903042e-05,
-      "loss": 0.5499,
+      "epoch": 6.511932945810624,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.101500123616585e-05,
+      "loss": 0.970751724243164,
       "step": 565600
     },
     {
-      "epoch": 13.028460900801326,
-      "grad_norm": 2.772387742996216,
-      "learning_rate": 1.1015219383080525e-05,
-      "loss": 0.5341,
+      "epoch": 6.514235630885339,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1007729672343334e-05,
+      "loss": 0.9810092163085937,
       "step": 565800
     },
     {
-      "epoch": 13.033066224555586,
-      "grad_norm": 2.9487922191619873,
-      "learning_rate": 1.1007947819258009e-05,
-      "loss": 0.5334,
+      "epoch": 6.516538315960054,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.1000458108520819e-05,
+      "loss": 0.9883064270019531,
       "step": 566000
     },
     {
-      "epoch": 13.037671548309847,
-      "grad_norm": 2.795191764831543,
-      "learning_rate": 1.1000676255435495e-05,
-      "loss": 0.55,
+      "epoch": 6.518841001034769,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0993186544698304e-05,
+      "loss": 0.9706736755371094,
       "step": 566200
     },
     {
-      "epoch": 13.042276872064106,
-      "grad_norm": 2.9274868965148926,
-      "learning_rate": 1.0993404691612979e-05,
-      "loss": 0.5449,
+      "epoch": 6.521143686109484,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0985914980875787e-05,
+      "loss": 0.9820281982421875,
       "step": 566400
     },
     {
-      "epoch": 13.046882195818366,
-      "grad_norm": 2.8535702228546143,
-      "learning_rate": 1.0986133127790463e-05,
-      "loss": 0.5465,
+      "epoch": 6.523446371184199,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0978643417053271e-05,
+      "loss": 0.9836884307861328,
       "step": 566600
     },
     {
-      "epoch": 13.051487519572627,
-      "grad_norm": 2.9377224445343018,
-      "learning_rate": 1.0978861563967947e-05,
-      "loss": 0.5542,
+      "epoch": 6.525749056258914,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.0971371853230757e-05,
+      "loss": 1.001689682006836,
       "step": 566800
     },
     {
-      "epoch": 13.056092843326885,
-      "grad_norm": 3.308588743209839,
-      "learning_rate": 1.0971590000145432e-05,
-      "loss": 0.5451,
+      "epoch": 6.528051741333629,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.096410028940824e-05,
+      "loss": 0.9856477355957032,
       "step": 567000
     },
     {
-      "epoch": 13.060698167081146,
-      "grad_norm": 3.0085132122039795,
-      "learning_rate": 1.0964354794142028e-05,
-      "loss": 0.5511,
+      "epoch": 6.530354426408344,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0956828725585724e-05,
+      "loss": 0.9702422332763672,
       "step": 567200
     },
     {
-      "epoch": 13.065303490835406,
-      "grad_norm": 2.9090094566345215,
-      "learning_rate": 1.0957083230319513e-05,
-      "loss": 0.5345,
+      "epoch": 6.532657111483059,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.094955716176321e-05,
+      "loss": 0.990765609741211,
       "step": 567400
     },
     {
-      "epoch": 13.069908814589665,
-      "grad_norm": 2.8537399768829346,
-      "learning_rate": 1.0949811666496997e-05,
-      "loss": 0.5553,
+      "epoch": 6.534959796557773,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0942285597940692e-05,
+      "loss": 1.006974563598633,
       "step": 567600
     },
     {
-      "epoch": 13.074514138343925,
-      "grad_norm": 2.471475124359131,
-      "learning_rate": 1.0942540102674483e-05,
-      "loss": 0.5526,
+      "epoch": 6.537262481632489,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0935014034118178e-05,
+      "loss": 0.9736676788330079,
       "step": 567800
     },
     {
-      "epoch": 13.079119462098186,
-      "grad_norm": 3.00661301612854,
-      "learning_rate": 1.0935268538851965e-05,
-      "loss": 0.5402,
+      "epoch": 6.539565166707203,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0927742470295662e-05,
+      "loss": 0.9964972686767578,
       "step": 568000
     },
     {
-      "epoch": 13.083724785852445,
-      "grad_norm": 3.291968822479248,
-      "learning_rate": 1.092799697502945e-05,
-      "loss": 0.5389,
+      "epoch": 6.541867851781919,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0920470906473145e-05,
+      "loss": 0.9812130737304687,
       "step": 568200
     },
     {
-      "epoch": 13.088330109606705,
-      "grad_norm": 5.072547912597656,
-      "learning_rate": 1.0920725411206936e-05,
-      "loss": 0.5422,
+      "epoch": 6.544170536856633,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.0913199342650631e-05,
+      "loss": 0.9940660858154297,
       "step": 568400
     },
     {
-      "epoch": 13.092935433360966,
-      "grad_norm": 2.9961328506469727,
-      "learning_rate": 1.0913453847384418e-05,
-      "loss": 0.5345,
+      "epoch": 6.546473221931349,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.0905927778828115e-05,
+      "loss": 0.9886941528320312,
       "step": 568600
     },
     {
-      "epoch": 13.097540757115226,
-      "grad_norm": 2.3409829139709473,
-      "learning_rate": 1.0906182283561902e-05,
-      "loss": 0.5516,
+      "epoch": 6.548775907006063,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.08986562150056e-05,
+      "loss": 0.9685357666015625,
       "step": 568800
     },
     {
-      "epoch": 13.102146080869485,
-      "grad_norm": 2.485668897628784,
-      "learning_rate": 1.0898910719739388e-05,
-      "loss": 0.563,
+      "epoch": 6.551078592080778,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0891384651183084e-05,
+      "loss": 0.9673428344726562,
       "step": 569000
     },
     {
-      "epoch": 13.106751404623745,
-      "grad_norm": 2.9506261348724365,
-      "learning_rate": 1.089163915591687e-05,
-      "loss": 0.5452,
+      "epoch": 6.553381277155493,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0884113087360568e-05,
+      "loss": 0.9898664855957031,
       "step": 569200
     },
     {
-      "epoch": 13.111356728378006,
-      "grad_norm": 3.2650866508483887,
-      "learning_rate": 1.0884367592094357e-05,
-      "loss": 0.547,
+      "epoch": 6.5556839622302086,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0876841523538054e-05,
+      "loss": 0.9710381317138672,
       "step": 569400
     },
     {
-      "epoch": 13.115962052132264,
-      "grad_norm": 4.075747489929199,
-      "learning_rate": 1.087709602827184e-05,
-      "loss": 0.5352,
+      "epoch": 6.557986647304923,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0869569959715536e-05,
+      "loss": 0.957591552734375,
       "step": 569600
     },
     {
-      "epoch": 13.120567375886525,
-      "grad_norm": 3.905189275741577,
-      "learning_rate": 1.0869824464449325e-05,
-      "loss": 0.5458,
+      "epoch": 6.560289332379638,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.086229839589302e-05,
+      "loss": 0.9589216613769531,
       "step": 569800
     },
     {
-      "epoch": 13.125172699640785,
-      "grad_norm": 3.0449321269989014,
-      "learning_rate": 1.086255290062681e-05,
-      "loss": 0.5476,
+      "epoch": 6.562592017454353,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0855026832070506e-05,
+      "loss": 0.9881840515136718,
       "step": 570000
     },
     {
-      "epoch": 13.129778023395044,
-      "grad_norm": 5.308364391326904,
-      "learning_rate": 1.0855281336804293e-05,
-      "loss": 0.5482,
+      "epoch": 6.564894702529068,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0847755268247989e-05,
+      "loss": 0.9679238891601563,
       "step": 570200
     },
     {
-      "epoch": 13.134383347149305,
-      "grad_norm": 3.4296975135803223,
-      "learning_rate": 1.0848009772981778e-05,
-      "loss": 0.5381,
+      "epoch": 6.567197387603783,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0840483704425475e-05,
+      "loss": 0.9674862670898438,
       "step": 570400
     },
     {
-      "epoch": 13.138988670903565,
-      "grad_norm": 3.1632046699523926,
-      "learning_rate": 1.0840738209159262e-05,
-      "loss": 0.5482,
+      "epoch": 6.569500072678498,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0833212140602959e-05,
+      "loss": 0.9502196502685547,
       "step": 570600
     },
     {
-      "epoch": 13.143593994657824,
-      "grad_norm": 2.9227309226989746,
-      "learning_rate": 1.0833466645336746e-05,
-      "loss": 0.5479,
+      "epoch": 6.571802757753213,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0825940576780442e-05,
+      "loss": 0.9787293243408203,
       "step": 570800
     },
     {
-      "epoch": 13.148199318412084,
-      "grad_norm": 2.676102638244629,
-      "learning_rate": 1.0826195081514232e-05,
-      "loss": 0.5539,
+      "epoch": 6.5741054428279275,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0818669012957927e-05,
+      "loss": 1.0087420654296875,
       "step": 571000
     },
     {
-      "epoch": 13.152804642166345,
-      "grad_norm": 3.266191244125366,
-      "learning_rate": 1.0818923517691715e-05,
-      "loss": 0.5409,
+      "epoch": 6.576408127902642,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0811397449135412e-05,
+      "loss": 0.9817219543457031,
       "step": 571200
     },
     {
-      "epoch": 13.157409965920603,
-      "grad_norm": 2.1656315326690674,
-      "learning_rate": 1.0811651953869199e-05,
-      "loss": 0.5465,
+      "epoch": 6.5787108129773575,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0804125885312894e-05,
+      "loss": 0.9803526306152344,
       "step": 571400
     },
     {
-      "epoch": 13.162015289674864,
-      "grad_norm": 2.710066318511963,
-      "learning_rate": 1.0804380390046685e-05,
-      "loss": 0.5401,
+      "epoch": 6.581013498052072,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.079685432149038e-05,
+      "loss": 0.9712112426757813,
       "step": 571600
     },
     {
-      "epoch": 13.166620613429124,
-      "grad_norm": 2.9499149322509766,
-      "learning_rate": 1.0797108826224167e-05,
-      "loss": 0.5369,
+      "epoch": 6.5833161831267875,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0789582757667864e-05,
+      "loss": 0.9745654296875,
       "step": 571800
     },
     {
-      "epoch": 13.171225937183385,
-      "grad_norm": 3.0565083026885986,
-      "learning_rate": 1.0789837262401651e-05,
-      "loss": 0.5431,
+      "epoch": 6.585618868201502,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.078231119384535e-05,
+      "loss": 0.9731366729736328,
       "step": 572000
     },
     {
-      "epoch": 13.175831260937644,
-      "grad_norm": 2.4222073554992676,
-      "learning_rate": 1.0782602056398248e-05,
-      "loss": 0.5354,
+      "epoch": 6.587921553276217,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0775039630022833e-05,
+      "loss": 0.9887467193603515,
       "step": 572200
     },
     {
-      "epoch": 13.180436584691904,
-      "grad_norm": 3.2243409156799316,
-      "learning_rate": 1.0775330492575734e-05,
-      "loss": 0.5388,
+      "epoch": 6.590224238350932,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0767768066200317e-05,
+      "loss": 0.9958180999755859,
       "step": 572400
     },
     {
-      "epoch": 13.185041908446165,
-      "grad_norm": 3.4836127758026123,
-      "learning_rate": 1.0768058928753217e-05,
-      "loss": 0.5449,
+      "epoch": 6.592526923425647,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0760496502377803e-05,
+      "loss": 0.9899420166015624,
       "step": 572600
     },
     {
-      "epoch": 13.189647232200423,
-      "grad_norm": 2.60473370552063,
-      "learning_rate": 1.0760787364930703e-05,
-      "loss": 0.5426,
+      "epoch": 6.594829608500362,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0753224938555285e-05,
+      "loss": 0.9905350494384766,
       "step": 572800
     },
     {
-      "epoch": 13.194252555954684,
-      "grad_norm": 2.716947317123413,
-      "learning_rate": 1.0753515801108187e-05,
-      "loss": 0.5478,
+      "epoch": 6.597132293575077,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.074595337473277e-05,
+      "loss": 0.9972842407226562,
       "step": 573000
     },
     {
-      "epoch": 13.198857879708944,
-      "grad_norm": 3.650529623031616,
-      "learning_rate": 1.0746244237285671e-05,
-      "loss": 0.5549,
+      "epoch": 6.599434978649792,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0738681810910256e-05,
+      "loss": 0.9325897216796875,
       "step": 573200
     },
     {
-      "epoch": 13.203463203463203,
-      "grad_norm": 2.8756980895996094,
-      "learning_rate": 1.0738972673463155e-05,
-      "loss": 0.5316,
+      "epoch": 6.601737663724506,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.0731410247087738e-05,
+      "loss": 0.9847271728515625,
       "step": 573400
     },
     {
-      "epoch": 13.208068527217463,
-      "grad_norm": 2.6897635459899902,
-      "learning_rate": 1.0731737467459752e-05,
-      "loss": 0.5562,
+      "epoch": 6.604040348799222,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0724138683265224e-05,
+      "loss": 0.9817544555664063,
       "step": 573600
     },
     {
-      "epoch": 13.212673850971724,
-      "grad_norm": 3.219444751739502,
-      "learning_rate": 1.0724465903637237e-05,
-      "loss": 0.5417,
+      "epoch": 6.606343033873936,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0716867119442708e-05,
+      "loss": 0.9679446411132813,
       "step": 573800
     },
     {
-      "epoch": 13.217279174725983,
-      "grad_norm": 3.117840051651001,
-      "learning_rate": 1.071719433981472e-05,
-      "loss": 0.5398,
+      "epoch": 6.608645718948652,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.070959555562019e-05,
+      "loss": 0.9627344512939453,
       "step": 574000
     },
     {
-      "epoch": 13.221884498480243,
-      "grad_norm": 4.030891418457031,
-      "learning_rate": 1.0709922775992205e-05,
-      "loss": 0.5512,
+      "epoch": 6.610948404023366,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0702323991797677e-05,
+      "loss": 1.007912139892578,
       "step": 574200
     },
     {
-      "epoch": 13.226489822234504,
-      "grad_norm": 3.657562494277954,
-      "learning_rate": 1.070265121216969e-05,
-      "loss": 0.5485,
+      "epoch": 6.613251089098082,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0695052427975161e-05,
+      "loss": 0.9748046112060547,
       "step": 574400
     },
     {
-      "epoch": 13.231095145988762,
-      "grad_norm": 2.986097574234009,
-      "learning_rate": 1.0695379648347175e-05,
-      "loss": 0.5443,
+      "epoch": 6.615553774172796,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0687780864152643e-05,
+      "loss": 0.971591567993164,
       "step": 574600
     },
     {
-      "epoch": 13.235700469743023,
-      "grad_norm": 3.578052043914795,
-      "learning_rate": 1.0688108084524658e-05,
-      "loss": 0.5537,
+      "epoch": 6.617856459247511,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.068050930033013e-05,
+      "loss": 0.9806672668457032,
       "step": 574800
     },
     {
-      "epoch": 13.240305793497283,
-      "grad_norm": 3.427138328552246,
-      "learning_rate": 1.0680836520702142e-05,
-      "loss": 0.5402,
+      "epoch": 6.620159144322226,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.0673237736507614e-05,
+      "loss": 0.9956932830810546,
       "step": 575000
     },
     {
-      "epoch": 13.244911117251542,
-      "grad_norm": 3.066641330718994,
-      "learning_rate": 1.0673564956879628e-05,
-      "loss": 0.5465,
+      "epoch": 6.622461829396941,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.0665966172685098e-05,
+      "loss": 0.9696024322509765,
       "step": 575200
     },
     {
-      "epoch": 13.249516441005802,
-      "grad_norm": 2.710906505584717,
-      "learning_rate": 1.066629339305711e-05,
-      "loss": 0.5569,
+      "epoch": 6.624764514471656,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0658694608862582e-05,
+      "loss": 0.96540283203125,
       "step": 575400
     },
     {
-      "epoch": 13.254121764760063,
-      "grad_norm": 2.721404790878296,
-      "learning_rate": 1.0659021829234596e-05,
-      "loss": 0.5504,
+      "epoch": 6.627067199546371,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0651423045040066e-05,
+      "loss": 0.9723766326904297,
       "step": 575600
     },
     {
-      "epoch": 13.258727088514323,
-      "grad_norm": 4.378937244415283,
-      "learning_rate": 1.065175026541208e-05,
-      "loss": 0.5419,
+      "epoch": 6.629369884621086,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0644151481217552e-05,
+      "loss": 1.006049346923828,
       "step": 575800
     },
     {
-      "epoch": 13.263332412268582,
-      "grad_norm": 2.9746248722076416,
-      "learning_rate": 1.0644478701589563e-05,
-      "loss": 0.5399,
+      "epoch": 6.631672569695801,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0636879917395035e-05,
+      "loss": 0.9848546600341797,
       "step": 576000
     },
     {
-      "epoch": 13.267937736022843,
-      "grad_norm": 3.093886613845825,
-      "learning_rate": 1.0637207137767049e-05,
-      "loss": 0.5506,
+      "epoch": 6.633975254770516,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.062960835357252e-05,
+      "loss": 1.018324203491211,
       "step": 576200
     },
     {
-      "epoch": 13.272543059777103,
-      "grad_norm": 3.421165943145752,
-      "learning_rate": 1.0629935573944533e-05,
-      "loss": 0.5569,
+      "epoch": 6.636277939845231,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0622336789750005e-05,
+      "loss": 0.9924317169189453,
       "step": 576400
     },
     {
-      "epoch": 13.277148383531362,
-      "grad_norm": 2.879319906234741,
-      "learning_rate": 1.0622664010122017e-05,
-      "loss": 0.5379,
+      "epoch": 6.638580624919946,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0615065225927487e-05,
+      "loss": 0.9519772338867187,
       "step": 576600
     },
     {
-      "epoch": 13.281753707285622,
-      "grad_norm": 4.0503363609313965,
-      "learning_rate": 1.0615392446299501e-05,
-      "loss": 0.5518,
+      "epoch": 6.640883309994661,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0607793662104973e-05,
+      "loss": 0.9751395416259766,
       "step": 576800
     },
     {
-      "epoch": 13.286359031039883,
-      "grad_norm": 3.7506306171417236,
-      "learning_rate": 1.0608120882476986e-05,
-      "loss": 0.5454,
+      "epoch": 6.643185995069375,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0600522098282457e-05,
+      "loss": 0.9836444854736328,
       "step": 577000
     },
     {
-      "epoch": 13.290964354794141,
-      "grad_norm": 3.531977891921997,
-      "learning_rate": 1.0600849318654472e-05,
-      "loss": 0.5499,
+      "epoch": 6.645488680144091,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.059325053445994e-05,
+      "loss": 0.9723554229736329,
       "step": 577200
     },
     {
-      "epoch": 13.295569678548402,
-      "grad_norm": 3.0157346725463867,
-      "learning_rate": 1.0593614112651069e-05,
-      "loss": 0.5529,
+      "epoch": 6.647791365218805,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0585978970637426e-05,
+      "loss": 1.0011803436279296,
       "step": 577400
     },
     {
-      "epoch": 13.300175002302662,
-      "grad_norm": 2.793652057647705,
-      "learning_rate": 1.0586342548828551e-05,
-      "loss": 0.5458,
+      "epoch": 6.650094050293521,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.057870740681491e-05,
+      "loss": 0.9975678253173829,
       "step": 577600
     },
     {
-      "epoch": 13.304780326056921,
-      "grad_norm": 2.5756351947784424,
-      "learning_rate": 1.0579070985006035e-05,
-      "loss": 0.5509,
+      "epoch": 6.652396735368235,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.0571435842992394e-05,
+      "loss": 0.9962368774414062,
       "step": 577800
     },
     {
-      "epoch": 13.309385649811182,
-      "grad_norm": 2.810288906097412,
-      "learning_rate": 1.0571799421183521e-05,
-      "loss": 0.5426,
+      "epoch": 6.654699420442951,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.0564164279169879e-05,
+      "loss": 0.9701191711425782,
       "step": 578000
     },
     {
-      "epoch": 13.313990973565442,
-      "grad_norm": 2.495398759841919,
-      "learning_rate": 1.0564527857361004e-05,
-      "loss": 0.5526,
+      "epoch": 6.657002105517665,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0556892715347363e-05,
+      "loss": 0.9803343200683594,
       "step": 578200
     },
     {
-      "epoch": 13.3185962973197,
-      "grad_norm": 4.127913475036621,
-      "learning_rate": 1.0557256293538488e-05,
-      "loss": 0.5498,
+      "epoch": 6.65930479059238,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.0549621151524847e-05,
+      "loss": 0.9848207855224609,
       "step": 578400
     },
     {
-      "epoch": 13.323201621073961,
-      "grad_norm": 3.074737071990967,
-      "learning_rate": 1.0549984729715974e-05,
-      "loss": 0.543,
+      "epoch": 6.661607475667095,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0542349587702331e-05,
+      "loss": 0.9811007690429687,
       "step": 578600
     },
     {
-      "epoch": 13.327806944828222,
-      "grad_norm": 2.928070306777954,
-      "learning_rate": 1.0542713165893456e-05,
-      "loss": 0.5449,
+      "epoch": 6.66391016074181,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0535078023879815e-05,
+      "loss": 0.9867758941650391,
       "step": 578800
     },
     {
-      "epoch": 13.332412268582482,
-      "grad_norm": 3.301532745361328,
-      "learning_rate": 1.0535441602070942e-05,
-      "loss": 0.5471,
+      "epoch": 6.666212845816525,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0527806460057301e-05,
+      "loss": 0.9914443206787109,
       "step": 579000
     },
     {
-      "epoch": 13.337017592336741,
-      "grad_norm": 3.355616331100464,
-      "learning_rate": 1.0528170038248427e-05,
-      "loss": 0.5447,
+      "epoch": 6.66851553089124,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0520534896234784e-05,
+      "loss": 0.9794180297851562,
       "step": 579200
     },
     {
-      "epoch": 13.341622916091001,
-      "grad_norm": 2.8991310596466064,
-      "learning_rate": 1.0520898474425909e-05,
-      "loss": 0.5405,
+      "epoch": 6.670818215965955,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.051326333241227e-05,
+      "loss": 0.9820137023925781,
       "step": 579400
     },
     {
-      "epoch": 13.346228239845262,
-      "grad_norm": 3.3360657691955566,
-      "learning_rate": 1.0513626910603395e-05,
-      "loss": 0.5585,
+      "epoch": 6.67312090104067,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0505991768589754e-05,
+      "loss": 0.990504150390625,
       "step": 579600
     },
     {
-      "epoch": 13.35083356359952,
-      "grad_norm": 4.818231105804443,
-      "learning_rate": 1.050635534678088e-05,
-      "loss": 0.5512,
+      "epoch": 6.675423586115385,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0498720204767236e-05,
+      "loss": 0.9704412078857422,
       "step": 579800
     },
     {
-      "epoch": 13.355438887353781,
-      "grad_norm": 4.0637078285217285,
-      "learning_rate": 1.0499083782958362e-05,
-      "loss": 0.5499,
+      "epoch": 6.6777262711900995,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0491448640944722e-05,
+      "loss": 0.9653994750976562,
       "step": 580000
     },
     {
-      "epoch": 13.360044211108042,
-      "grad_norm": 4.462865352630615,
-      "learning_rate": 1.0491812219135848e-05,
-      "loss": 0.5425,
+      "epoch": 6.680028956264815,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0484177077122207e-05,
+      "loss": 0.9979000091552734,
       "step": 580200
     },
     {
-      "epoch": 13.3646495348623,
-      "grad_norm": 2.5351686477661133,
-      "learning_rate": 1.0484540655313332e-05,
-      "loss": 0.5363,
+      "epoch": 6.6823316413395295,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0476905513299689e-05,
+      "loss": 0.9706903076171876,
       "step": 580400
     },
     {
-      "epoch": 13.36925485861656,
-      "grad_norm": 3.533302068710327,
-      "learning_rate": 1.0477269091490818e-05,
-      "loss": 0.5424,
+      "epoch": 6.684634326414244,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0469633949477175e-05,
+      "loss": 0.9929607391357422,
       "step": 580600
     },
     {
-      "epoch": 13.373860182370821,
-      "grad_norm": 3.593005418777466,
-      "learning_rate": 1.04699975276683e-05,
-      "loss": 0.5491,
+      "epoch": 6.6869370114889595,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.046236238565466e-05,
+      "loss": 0.9664354705810547,
       "step": 580800
     },
     {
-      "epoch": 13.37846550612508,
-      "grad_norm": 3.0207858085632324,
-      "learning_rate": 1.0462725963845784e-05,
-      "loss": 0.5451,
+      "epoch": 6.689239696563674,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.0455090821832144e-05,
+      "loss": 0.9774636840820312,
       "step": 581000
     },
     {
-      "epoch": 13.38307082987934,
-      "grad_norm": 3.600573778152466,
-      "learning_rate": 1.045545440002327e-05,
-      "loss": 0.5409,
+      "epoch": 6.691542381638389,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.0447819258009628e-05,
+      "loss": 0.9902906799316407,
       "step": 581200
     },
     {
-      "epoch": 13.3876761536336,
-      "grad_norm": 3.4477622509002686,
-      "learning_rate": 1.0448182836200753e-05,
-      "loss": 0.5515,
+      "epoch": 6.693845066713104,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0440547694187112e-05,
+      "loss": 1.0105450439453125,
       "step": 581400
     },
     {
-      "epoch": 13.39228147738786,
-      "grad_norm": 4.029882431030273,
-      "learning_rate": 1.044094763019735e-05,
-      "loss": 0.557,
+      "epoch": 6.696147751787819,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0433276130364596e-05,
+      "loss": 0.9908228302001953,
       "step": 581600
     },
     {
-      "epoch": 13.39688680114212,
-      "grad_norm": 3.970107316970825,
-      "learning_rate": 1.0433676066374834e-05,
-      "loss": 0.545,
+      "epoch": 6.698450436862534,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.042600456654208e-05,
+      "loss": 0.9905318450927735,
       "step": 581800
     },
     {
-      "epoch": 13.40149212489638,
-      "grad_norm": 3.111429214477539,
-      "learning_rate": 1.042640450255232e-05,
-      "loss": 0.5443,
+      "epoch": 6.7007531219372485,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0418733002719565e-05,
+      "loss": 0.9779328155517578,
       "step": 582000
     },
     {
-      "epoch": 13.406097448650641,
-      "grad_norm": 3.304234027862549,
-      "learning_rate": 1.0419132938729802e-05,
-      "loss": 0.5425,
+      "epoch": 6.703055807011964,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.041146143889705e-05,
+      "loss": 0.9844884490966797,
       "step": 582200
     },
     {
-      "epoch": 13.4107027724049,
-      "grad_norm": 3.118661642074585,
-      "learning_rate": 1.0411861374907288e-05,
-      "loss": 0.5426,
+      "epoch": 6.7053584920866784,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0404189875074533e-05,
+      "loss": 0.9839838409423828,
       "step": 582400
     },
     {
-      "epoch": 13.41530809615916,
-      "grad_norm": 3.274920701980591,
-      "learning_rate": 1.0404589811084773e-05,
-      "loss": 0.5526,
+      "epoch": 6.707661177161394,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0396918311252019e-05,
+      "loss": 0.9807382202148438,
       "step": 582600
     },
     {
-      "epoch": 13.41991341991342,
-      "grad_norm": 2.8108534812927246,
-      "learning_rate": 1.0397318247262255e-05,
-      "loss": 0.5496,
+      "epoch": 6.709963862236108,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0389646747429503e-05,
+      "loss": 0.9836902618408203,
       "step": 582800
     },
     {
-      "epoch": 13.42451874366768,
-      "grad_norm": 3.452916383743286,
-      "learning_rate": 1.0390046683439741e-05,
-      "loss": 0.5546,
+      "epoch": 6.712266547310824,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0382375183606986e-05,
+      "loss": 0.9829158020019532,
       "step": 583000
     },
     {
-      "epoch": 13.42912406742194,
-      "grad_norm": 2.889641284942627,
-      "learning_rate": 1.0382775119617225e-05,
-      "loss": 0.5399,
+      "epoch": 6.714569232385538,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0375103619784472e-05,
+      "loss": 0.9916680908203125,
       "step": 583200
     },
     {
-      "epoch": 13.4337293911762,
-      "grad_norm": 2.9781227111816406,
-      "learning_rate": 1.037550355579471e-05,
-      "loss": 0.5442,
+      "epoch": 6.716871917460254,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0367832055961956e-05,
+      "loss": 0.9769731903076172,
       "step": 583400
     },
     {
-      "epoch": 13.438334714930459,
-      "grad_norm": 2.3209664821624756,
-      "learning_rate": 1.0368231991972194e-05,
-      "loss": 0.5563,
+      "epoch": 6.719174602534968,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.036056049213944e-05,
+      "loss": 0.9810326385498047,
       "step": 583600
     },
     {
-      "epoch": 13.44294003868472,
-      "grad_norm": 3.272317409515381,
-      "learning_rate": 1.036099678596879e-05,
-      "loss": 0.5424,
+      "epoch": 6.721477287609684,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0353288928316924e-05,
+      "loss": 1.0014809417724608,
       "step": 583800
     },
     {
-      "epoch": 13.44754536243898,
-      "grad_norm": 3.984229803085327,
-      "learning_rate": 1.0353725222146275e-05,
-      "loss": 0.5377,
+      "epoch": 6.723779972684398,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0346017364494408e-05,
+      "loss": 0.9822134399414062,
       "step": 584000
     },
     {
-      "epoch": 13.452150686193239,
-      "grad_norm": 2.723288059234619,
-      "learning_rate": 1.034645365832376e-05,
-      "loss": 0.5353,
+      "epoch": 6.726082657759113,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.0338745800671893e-05,
+      "loss": 0.9904820251464844,
       "step": 584200
     },
     {
-      "epoch": 13.4567560099475,
-      "grad_norm": 2.800933361053467,
-      "learning_rate": 1.0339218452320358e-05,
-      "loss": 0.5375,
+      "epoch": 6.728385342833828,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0331474236849377e-05,
+      "loss": 0.9892137145996094,
       "step": 584400
     },
     {
-      "epoch": 13.46136133370176,
-      "grad_norm": 8.197662353515625,
-      "learning_rate": 1.033194688849784e-05,
-      "loss": 0.5404,
+      "epoch": 6.730688027908543,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.0324202673026861e-05,
+      "loss": 0.980720443725586,
       "step": 584600
     },
     {
-      "epoch": 13.465966657456018,
-      "grad_norm": 2.963289737701416,
-      "learning_rate": 1.0324675324675324e-05,
-      "loss": 0.5413,
+      "epoch": 6.732990712983258,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0316931109204345e-05,
+      "loss": 0.9904631805419922,
       "step": 584800
     },
     {
-      "epoch": 13.470571981210279,
-      "grad_norm": 3.9458508491516113,
-      "learning_rate": 1.031740376085281e-05,
-      "loss": 0.5431,
+      "epoch": 6.735293398057973,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.030965954538183e-05,
+      "loss": 0.9949741363525391,
       "step": 585000
     },
     {
-      "epoch": 13.47517730496454,
-      "grad_norm": 2.880087375640869,
-      "learning_rate": 1.0310132197030293e-05,
-      "loss": 0.5559,
+      "epoch": 6.737596083132688,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.0302387981559315e-05,
+      "loss": 1.012327880859375,
       "step": 585200
     },
     {
-      "epoch": 13.479782628718798,
-      "grad_norm": 2.817814826965332,
-      "learning_rate": 1.0302860633207779e-05,
-      "loss": 0.5403,
+      "epoch": 6.739898768207403,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.02951164177368e-05,
+      "loss": 0.9787694549560547,
       "step": 585400
     },
     {
-      "epoch": 13.484387952473059,
-      "grad_norm": 3.2126080989837646,
-      "learning_rate": 1.0295589069385263e-05,
-      "loss": 0.536,
+      "epoch": 6.742201453282117,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0287844853914282e-05,
+      "loss": 0.9854193878173828,
       "step": 585600
     },
     {
-      "epoch": 13.488993276227319,
-      "grad_norm": 2.991262674331665,
-      "learning_rate": 1.0288317505562746e-05,
-      "loss": 0.5428,
+      "epoch": 6.744504138356833,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0280573290091768e-05,
+      "loss": 0.9754302215576172,
       "step": 585800
     },
     {
-      "epoch": 13.49359859998158,
-      "grad_norm": 2.65507435798645,
-      "learning_rate": 1.0281045941740231e-05,
-      "loss": 0.5384,
+      "epoch": 6.746806823431548,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.0273301726269252e-05,
+      "loss": 0.9814857482910156,
       "step": 586000
     },
     {
-      "epoch": 13.498203923735838,
-      "grad_norm": 3.388256788253784,
-      "learning_rate": 1.0273774377917716e-05,
-      "loss": 0.5431,
+      "epoch": 6.749109508506263,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0266030162446735e-05,
+      "loss": 0.9915824890136719,
       "step": 586200
     },
     {
-      "epoch": 13.502809247490099,
-      "grad_norm": 3.2963063716888428,
-      "learning_rate": 1.0266502814095198e-05,
-      "loss": 0.5506,
+      "epoch": 6.751412193580977,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.025875859862422e-05,
+      "loss": 0.9872344970703125,
       "step": 586400
     },
     {
-      "epoch": 13.50741457124436,
-      "grad_norm": 3.947801351547241,
-      "learning_rate": 1.0259231250272684e-05,
-      "loss": 0.5419,
+      "epoch": 6.753714878655693,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0251487034801705e-05,
+      "loss": 0.9672900390625,
       "step": 586600
     },
     {
-      "epoch": 13.512019894998618,
-      "grad_norm": 3.5557472705841064,
-      "learning_rate": 1.0251959686450168e-05,
-      "loss": 0.5538,
+      "epoch": 6.756017563730407,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.024421547097919e-05,
+      "loss": 0.9766867065429687,
       "step": 586800
     },
     {
-      "epoch": 13.516625218752878,
-      "grad_norm": 3.8408093452453613,
-      "learning_rate": 1.0244688122627653e-05,
-      "loss": 0.5394,
+      "epoch": 6.758320248805123,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.0236943907156673e-05,
+      "loss": 1.0006546783447265,
       "step": 587000
     },
     {
-      "epoch": 13.521230542507139,
-      "grad_norm": 3.790393352508545,
-      "learning_rate": 1.0237416558805137e-05,
-      "loss": 0.5567,
+      "epoch": 6.760622933879837,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0229672343334158e-05,
+      "loss": 0.9878829193115234,
       "step": 587200
     },
     {
-      "epoch": 13.525835866261398,
-      "grad_norm": 2.450249195098877,
-      "learning_rate": 1.0230144994982621e-05,
-      "loss": 0.5475,
+      "epoch": 6.762925618954553,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0222400779511642e-05,
+      "loss": 0.9849906921386719,
       "step": 587400
     },
     {
-      "epoch": 13.530441190015658,
-      "grad_norm": 3.272376537322998,
-      "learning_rate": 1.0222873431160107e-05,
-      "loss": 0.5481,
+      "epoch": 6.765228304029267,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0215129215689126e-05,
+      "loss": 0.9837782287597656,
       "step": 587600
     },
     {
-      "epoch": 13.535046513769919,
-      "grad_norm": 3.2694225311279297,
-      "learning_rate": 1.021560186733759e-05,
-      "loss": 0.5467,
+      "epoch": 6.767530989103982,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.020785765186661e-05,
+      "loss": 0.978974609375,
       "step": 587800
     },
     {
-      "epoch": 13.539651837524177,
-      "grad_norm": 2.739555835723877,
-      "learning_rate": 1.0208330303515074e-05,
-      "loss": 0.541,
+      "epoch": 6.769833674178697,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0200586088044095e-05,
+      "loss": 0.9979602813720703,
       "step": 588000
     },
     {
-      "epoch": 13.544257161278438,
-      "grad_norm": 3.357004404067993,
-      "learning_rate": 1.020105873969256e-05,
-      "loss": 0.5549,
+      "epoch": 6.772136359253412,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0193314524221579e-05,
+      "loss": 0.9683854675292969,
       "step": 588200
     },
     {
-      "epoch": 13.548862485032698,
-      "grad_norm": 3.012406349182129,
-      "learning_rate": 1.0193823533689156e-05,
-      "loss": 0.5503,
+      "epoch": 6.774439044328127,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0186042960399065e-05,
+      "loss": 0.9971164703369141,
       "step": 588400
     },
     {
-      "epoch": 13.553467808786957,
-      "grad_norm": 2.7790162563323975,
-      "learning_rate": 1.0186551969866639e-05,
-      "loss": 0.5447,
+      "epoch": 6.776741729402842,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0178771396576547e-05,
+      "loss": 0.9966475677490234,
       "step": 588600
     },
     {
-      "epoch": 13.558073132541217,
-      "grad_norm": 3.0401885509490967,
-      "learning_rate": 1.0179280406044125e-05,
-      "loss": 0.5352,
+      "epoch": 6.779044414477557,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.0171499832754031e-05,
+      "loss": 0.9796976470947265,
       "step": 588800
     },
     {
-      "epoch": 13.562678456295478,
-      "grad_norm": 3.2460315227508545,
-      "learning_rate": 1.0172008842221609e-05,
-      "loss": 0.5387,
+      "epoch": 6.7813470995522716,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.0164228268931517e-05,
+      "loss": 0.9383587646484375,
       "step": 589000
     },
     {
-      "epoch": 13.567283780049738,
-      "grad_norm": 3.198957681655884,
-      "learning_rate": 1.0164737278399092e-05,
-      "loss": 0.5492,
+      "epoch": 6.783649784626987,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0156956705109002e-05,
+      "loss": 0.9837628936767578,
       "step": 589200
     },
     {
-      "epoch": 13.571889103803997,
-      "grad_norm": 3.1791601181030273,
-      "learning_rate": 1.0157465714576578e-05,
-      "loss": 0.5488,
+      "epoch": 6.7859524697017015,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0149685141286484e-05,
+      "loss": 0.9929660034179687,
       "step": 589400
     },
     {
-      "epoch": 13.576494427558258,
-      "grad_norm": 3.927855968475342,
-      "learning_rate": 1.0150194150754062e-05,
-      "loss": 0.5459,
+      "epoch": 6.788255154776417,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.014241357746397e-05,
+      "loss": 0.993055191040039,
       "step": 589600
     },
     {
-      "epoch": 13.581099751312518,
-      "grad_norm": 3.2088983058929443,
-      "learning_rate": 1.0142922586931544e-05,
-      "loss": 0.5425,
+      "epoch": 6.7905578398511315,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0135142013641454e-05,
+      "loss": 1.001646957397461,
       "step": 589800
     },
     {
-      "epoch": 13.585705075066777,
-      "grad_norm": 3.1823244094848633,
-      "learning_rate": 1.013565102310903e-05,
-      "loss": 0.5521,
+      "epoch": 6.792860524925846,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0127870449818938e-05,
+      "loss": 0.9851081085205078,
       "step": 590000
     },
     {
-      "epoch": 13.590310398821037,
-      "grad_norm": 3.0992443561553955,
-      "learning_rate": 1.0128379459286514e-05,
-      "loss": 0.5455,
+      "epoch": 6.7951632100005614,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0120598885996423e-05,
+      "loss": 0.9693524932861328,
       "step": 590200
     },
     {
-      "epoch": 13.594915722575298,
-      "grad_norm": 3.4294822216033936,
-      "learning_rate": 1.0121107895463999e-05,
-      "loss": 0.54,
+      "epoch": 6.797465895075276,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.0113327322173907e-05,
+      "loss": 0.9681202697753907,
       "step": 590400
     },
     {
-      "epoch": 13.599521046329556,
-      "grad_norm": 3.575033187866211,
-      "learning_rate": 1.0113836331641483e-05,
-      "loss": 0.553,
+      "epoch": 6.799768580149991,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.0106055758351391e-05,
+      "loss": 0.9943393707275391,
       "step": 590600
     },
     {
-      "epoch": 13.604126370083817,
-      "grad_norm": 2.823061466217041,
-      "learning_rate": 1.0106564767818967e-05,
-      "loss": 0.5422,
+      "epoch": 6.802071265224706,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0098784194528875e-05,
+      "loss": 0.9649395751953125,
       "step": 590800
     },
     {
-      "epoch": 13.608731693838077,
-      "grad_norm": 3.334573268890381,
-      "learning_rate": 1.0099293203996453e-05,
-      "loss": 0.5412,
+      "epoch": 6.804373950299421,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0091512630706361e-05,
+      "loss": 0.9603765869140625,
       "step": 591000
     },
     {
-      "epoch": 13.613337017592336,
-      "grad_norm": 3.7110838890075684,
-      "learning_rate": 1.0092021640173936e-05,
-      "loss": 0.5439,
+      "epoch": 6.806676635374136,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0084241066883844e-05,
+      "loss": 0.9838564300537109,
       "step": 591200
     },
     {
-      "epoch": 13.617942341346597,
-      "grad_norm": 2.9446306228637695,
-      "learning_rate": 1.008475007635142e-05,
-      "loss": 0.551,
+      "epoch": 6.8089793204488505,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0076969503061328e-05,
+      "loss": 0.9951017761230468,
       "step": 591400
     },
     {
-      "epoch": 13.622547665100857,
-      "grad_norm": 3.2592220306396484,
-      "learning_rate": 1.0077478512528906e-05,
-      "loss": 0.546,
+      "epoch": 6.811282005523566,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0069697939238814e-05,
+      "loss": 0.9828236389160157,
       "step": 591600
     },
     {
-      "epoch": 13.627152988855116,
-      "grad_norm": 2.9190750122070312,
-      "learning_rate": 1.0070206948706388e-05,
-      "loss": 0.5515,
+      "epoch": 6.81358469059828,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0062426375416296e-05,
+      "loss": 0.9695124053955078,
       "step": 591800
     },
     {
-      "epoch": 13.631758312609376,
-      "grad_norm": 3.803069829940796,
-      "learning_rate": 1.0062935384883874e-05,
-      "loss": 0.5438,
+      "epoch": 6.815887375672996,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.005515481159378e-05,
+      "loss": 0.979861831665039,
       "step": 592000
     },
     {
-      "epoch": 13.636363636363637,
-      "grad_norm": 3.252239942550659,
-      "learning_rate": 1.0055663821061358e-05,
-      "loss": 0.547,
+      "epoch": 6.81819006074771,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0047883247771267e-05,
+      "loss": 0.9814624786376953,
       "step": 592200
     },
     {
-      "epoch": 13.640968960117895,
-      "grad_norm": 3.070131778717041,
-      "learning_rate": 1.004839225723884e-05,
-      "loss": 0.5504,
+      "epoch": 6.820492745822426,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.004061168394875e-05,
+      "loss": 0.9897750091552734,
       "step": 592400
     },
     {
-      "epoch": 13.645574283872156,
-      "grad_norm": 3.4407927989959717,
-      "learning_rate": 1.0041120693416327e-05,
-      "loss": 0.5491,
+      "epoch": 6.82279543089714,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0033340120126235e-05,
+      "loss": 0.9900749206542969,
       "step": 592600
     },
     {
-      "epoch": 13.650179607626416,
-      "grad_norm": 3.2702760696411133,
-      "learning_rate": 1.0033849129593811e-05,
-      "loss": 0.5437,
+      "epoch": 6.825098115971856,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.002606855630372e-05,
+      "loss": 1.0019539642333983,
       "step": 592800
     },
     {
-      "epoch": 13.654784931380677,
-      "grad_norm": 3.0138094425201416,
-      "learning_rate": 1.0026577565771295e-05,
-      "loss": 0.5406,
+      "epoch": 6.82740080104657,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.0018796992481203e-05,
+      "loss": 0.9855136108398438,
       "step": 593000
     },
     {
-      "epoch": 13.659390255134936,
-      "grad_norm": 3.4657397270202637,
-      "learning_rate": 1.0019342359767892e-05,
-      "loss": 0.5364,
+      "epoch": 6.829703486121286,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0011525428658688e-05,
+      "loss": 0.9888275909423828,
       "step": 593200
     },
     {
-      "epoch": 13.663995578889196,
-      "grad_norm": 2.7210536003112793,
-      "learning_rate": 1.0012070795945376e-05,
-      "loss": 0.5348,
+      "epoch": 6.832006171196,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.0004253864836172e-05,
+      "loss": 0.9735780334472657,
       "step": 593400
     },
     {
-      "epoch": 13.668600902643457,
-      "grad_norm": 2.796694278717041,
-      "learning_rate": 1.000479923212286e-05,
-      "loss": 0.5356,
+      "epoch": 6.834308856270715,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.996982301013656e-06,
+      "loss": 0.9800492858886719,
       "step": 593600
     },
     {
-      "epoch": 13.673206226397715,
-      "grad_norm": 3.0048344135284424,
-      "learning_rate": 9.997527668300345e-06,
-      "loss": 0.547,
+      "epoch": 6.83661154134543,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.98971073719114e-06,
+      "loss": 0.9859529113769532,
       "step": 593800
     },
     {
-      "epoch": 13.677811550151976,
-      "grad_norm": 3.452331781387329,
-      "learning_rate": 9.990292462296942e-06,
-      "loss": 0.5483,
+      "epoch": 6.838914226420145,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.982439173368625e-06,
+      "loss": 0.9810272979736329,
       "step": 594000
     },
     {
-      "epoch": 13.682416873906236,
-      "grad_norm": 2.8550798892974854,
-      "learning_rate": 9.983020898474426e-06,
-      "loss": 0.5466,
+      "epoch": 6.84121691149486,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.97516760954611e-06,
+      "loss": 0.9831668090820312,
       "step": 594200
     },
     {
-      "epoch": 13.687022197660495,
-      "grad_norm": 3.329484701156616,
-      "learning_rate": 9.97574933465191e-06,
-      "loss": 0.5404,
+      "epoch": 6.843519596569575,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.967896045723593e-06,
+      "loss": 0.9732088470458984,
       "step": 594400
     },
     {
-      "epoch": 13.691627521414755,
-      "grad_norm": 2.8781020641326904,
-      "learning_rate": 9.968477770829396e-06,
-      "loss": 0.5518,
+      "epoch": 6.84582228164429,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.960624481901077e-06,
+      "loss": 0.9884004974365235,
       "step": 594600
     },
     {
-      "epoch": 13.696232845169016,
-      "grad_norm": 3.1248257160186768,
-      "learning_rate": 9.961206207006879e-06,
-      "loss": 0.5421,
+      "epoch": 6.848124966719005,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.953352918078563e-06,
+      "loss": 0.9899817657470703,
       "step": 594800
     },
     {
-      "epoch": 13.700838168923275,
-      "grad_norm": 3.2857155799865723,
-      "learning_rate": 9.953934643184364e-06,
-      "loss": 0.5479,
+      "epoch": 6.850427651793719,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.946081354256046e-06,
+      "loss": 0.981840591430664,
       "step": 595000
     },
     {
-      "epoch": 13.705443492677535,
-      "grad_norm": 3.693159818649292,
-      "learning_rate": 9.946663079361849e-06,
-      "loss": 0.5427,
+      "epoch": 6.852730336868435,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.93880979043353e-06,
+      "loss": 0.9764828491210937,
       "step": 595200
     },
     {
-      "epoch": 13.710048816431796,
-      "grad_norm": 3.3811700344085693,
-      "learning_rate": 9.939391515539331e-06,
-      "loss": 0.5495,
+      "epoch": 6.855033021943149,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.931538226611016e-06,
+      "loss": 0.9888670349121094,
       "step": 595400
     },
     {
-      "epoch": 13.714654140186056,
-      "grad_norm": 3.110826253890991,
-      "learning_rate": 9.932119951716817e-06,
-      "loss": 0.5327,
+      "epoch": 6.857335707017865,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.9242666627885e-06,
+      "loss": 0.99365966796875,
       "step": 595600
     },
     {
-      "epoch": 13.719259463940315,
-      "grad_norm": 3.164827346801758,
-      "learning_rate": 9.924848387894301e-06,
-      "loss": 0.5373,
+      "epoch": 6.859638392092579,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.916995098965984e-06,
+      "loss": 0.9660527038574219,
       "step": 595800
     },
     {
-      "epoch": 13.723864787694575,
-      "grad_norm": 3.971219062805176,
-      "learning_rate": 9.917576824071784e-06,
-      "loss": 0.5416,
+      "epoch": 6.861941077167295,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.909723535143468e-06,
+      "loss": 0.9956941986083985,
       "step": 596000
     },
     {
-      "epoch": 13.728470111448836,
-      "grad_norm": 3.429321527481079,
-      "learning_rate": 9.91030526024927e-06,
-      "loss": 0.5397,
+      "epoch": 6.864243762242009,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.902451971320953e-06,
+      "loss": 0.9861712646484375,
       "step": 596200
     },
     {
-      "epoch": 13.733075435203094,
-      "grad_norm": 2.479889392852783,
-      "learning_rate": 9.903033696426754e-06,
-      "loss": 0.5494,
+      "epoch": 6.866546447316725,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.895180407498437e-06,
+      "loss": 0.9897097778320313,
       "step": 596400
     },
     {
-      "epoch": 13.737680758957355,
-      "grad_norm": 2.7341201305389404,
-      "learning_rate": 9.895762132604238e-06,
-      "loss": 0.5354,
+      "epoch": 6.868849132391439,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.887908843675921e-06,
+      "loss": 0.9684822845458985,
       "step": 596600
     },
     {
-      "epoch": 13.742286082711615,
-      "grad_norm": 3.743123769760132,
-      "learning_rate": 9.888490568781722e-06,
-      "loss": 0.5502,
+      "epoch": 6.8711518174661546,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.880637279853405e-06,
+      "loss": 0.9750865936279297,
       "step": 596800
     },
     {
-      "epoch": 13.746891406465874,
-      "grad_norm": 2.8174068927764893,
-      "learning_rate": 9.88125536277832e-06,
-      "loss": 0.5447,
+      "epoch": 6.873454502540869,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.87336571603089e-06,
+      "loss": 0.9999384307861328,
       "step": 597000
     },
     {
-      "epoch": 13.751496730220135,
-      "grad_norm": 2.5518455505371094,
-      "learning_rate": 9.873983798955804e-06,
-      "loss": 0.5406,
+      "epoch": 6.875757187615584,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.866094152208374e-06,
+      "loss": 0.9845417785644531,
       "step": 597200
     },
     {
-      "epoch": 13.756102053974395,
-      "grad_norm": 2.6613545417785645,
-      "learning_rate": 9.866712235133288e-06,
-      "loss": 0.5436,
+      "epoch": 6.878059872690299,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.85882258838586e-06,
+      "loss": 0.9779905700683593,
       "step": 597400
     },
     {
-      "epoch": 13.760707377728654,
-      "grad_norm": 3.2584588527679443,
-      "learning_rate": 9.859440671310772e-06,
-      "loss": 0.5454,
+      "epoch": 6.880362557765014,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.851551024563342e-06,
+      "loss": 0.9748655700683594,
       "step": 597600
     },
     {
-      "epoch": 13.765312701482914,
-      "grad_norm": 3.6804134845733643,
-      "learning_rate": 9.852169107488256e-06,
-      "loss": 0.5415,
+      "epoch": 6.882665242839729,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.844279460740826e-06,
+      "loss": 0.9606812286376953,
       "step": 597800
     },
     {
-      "epoch": 13.769918025237175,
-      "grad_norm": 3.93707013130188,
-      "learning_rate": 9.844933901484853e-06,
-      "loss": 0.5405,
+      "epoch": 6.884967927914444,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.837007896918312e-06,
+      "loss": 0.9712443542480469,
       "step": 598000
     },
     {
-      "epoch": 13.774523348991433,
-      "grad_norm": 2.835728645324707,
-      "learning_rate": 9.837662337662339e-06,
-      "loss": 0.5355,
+      "epoch": 6.887270612989159,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.829736333095795e-06,
+      "loss": 0.9712157440185547,
       "step": 598200
     },
     {
-      "epoch": 13.779128672745694,
-      "grad_norm": 2.6848623752593994,
-      "learning_rate": 9.830390773839822e-06,
-      "loss": 0.5337,
+      "epoch": 6.8895732980638735,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.82246476927328e-06,
+      "loss": 0.9641321563720703,
       "step": 598400
     },
     {
-      "epoch": 13.783733996499954,
-      "grad_norm": 3.4841365814208984,
-      "learning_rate": 9.823119210017307e-06,
-      "loss": 0.531,
+      "epoch": 6.891875983138588,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.815193205450765e-06,
+      "loss": 0.9719064331054688,
       "step": 598600
     },
     {
-      "epoch": 13.788339320254213,
-      "grad_norm": 2.816138982772827,
-      "learning_rate": 9.815847646194792e-06,
-      "loss": 0.5315,
+      "epoch": 6.8941786682133035,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.807921641628247e-06,
+      "loss": 0.9778326416015625,
       "step": 598800
     },
     {
-      "epoch": 13.792944644008474,
-      "grad_norm": 2.922630786895752,
-      "learning_rate": 9.808576082372274e-06,
-      "loss": 0.5528,
+      "epoch": 6.896481353288018,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.800650077805733e-06,
+      "loss": 0.9895073699951172,
       "step": 599000
     },
     {
-      "epoch": 13.797549967762734,
-      "grad_norm": 3.093204975128174,
-      "learning_rate": 9.80130451854976e-06,
-      "loss": 0.5525,
+      "epoch": 6.8987840383627335,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.793378513983218e-06,
+      "loss": 0.9691558074951172,
       "step": 599200
     },
     {
-      "epoch": 13.802155291516993,
-      "grad_norm": 3.1738221645355225,
-      "learning_rate": 9.794032954727244e-06,
-      "loss": 0.5269,
+      "epoch": 6.901086723437448,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.786106950160702e-06,
+      "loss": 0.9776676940917969,
       "step": 599400
     },
     {
-      "epoch": 13.806760615271253,
-      "grad_norm": 2.908632755279541,
-      "learning_rate": 9.786761390904727e-06,
-      "loss": 0.5307,
+      "epoch": 6.903389408512163,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.778835386338186e-06,
+      "loss": 0.9621530914306641,
       "step": 599600
     },
     {
-      "epoch": 13.811365939025514,
-      "grad_norm": 3.4265527725219727,
-      "learning_rate": 9.779489827082213e-06,
-      "loss": 0.5327,
+      "epoch": 6.905692093586878,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.77156382251567e-06,
+      "loss": 0.9924739074707031,
       "step": 599800
     },
     {
-      "epoch": 13.815971262779774,
-      "grad_norm": 3.252572774887085,
-      "learning_rate": 9.772218263259697e-06,
-      "loss": 0.5453,
+      "epoch": 6.907994778661593,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.764292258693156e-06,
+      "loss": 0.9811209869384766,
       "step": 600000
     },
     {
-      "epoch": 13.820576586534033,
-      "grad_norm": 3.13742733001709,
-      "learning_rate": 9.764946699437181e-06,
-      "loss": 0.546,
+      "epoch": 6.910297463736308,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.757020694870639e-06,
+      "loss": 0.9884233856201172,
       "step": 600200
     },
     {
-      "epoch": 13.825181910288293,
-      "grad_norm": 2.9839015007019043,
-      "learning_rate": 9.757675135614665e-06,
-      "loss": 0.5444,
+      "epoch": 6.912600148811023,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.749749131048123e-06,
+      "loss": 0.9793074798583984,
       "step": 600400
     },
     {
-      "epoch": 13.829787234042554,
-      "grad_norm": 2.5981061458587646,
-      "learning_rate": 9.75040357179215e-06,
-      "loss": 0.5294,
+      "epoch": 6.914902833885738,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.742477567225609e-06,
+      "loss": 0.9784436798095704,
       "step": 600600
     },
     {
-      "epoch": 13.834392557796813,
-      "grad_norm": 2.8224525451660156,
-      "learning_rate": 9.743132007969634e-06,
-      "loss": 0.5433,
+      "epoch": 6.917205518960452,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.735206003403091e-06,
+      "loss": 0.9627828979492188,
       "step": 600800
     },
     {
-      "epoch": 13.838997881551073,
-      "grad_norm": 2.841012477874756,
-      "learning_rate": 9.735860444147118e-06,
-      "loss": 0.5346,
+      "epoch": 6.919508204035168,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.727934439580576e-06,
+      "loss": 0.9826019287109375,
       "step": 601000
     },
     {
-      "epoch": 13.843603205305334,
-      "grad_norm": 5.748844146728516,
-      "learning_rate": 9.728588880324602e-06,
-      "loss": 0.5509,
+      "epoch": 6.921810889109882,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.720662875758062e-06,
+      "loss": 0.9585595703125,
       "step": 601200
     },
     {
-      "epoch": 13.848208529059592,
-      "grad_norm": 2.9748475551605225,
-      "learning_rate": 9.721317316502088e-06,
-      "loss": 0.5386,
+      "epoch": 6.924113574184598,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.713391311935544e-06,
+      "loss": 0.9860497283935546,
       "step": 601400
     },
     {
-      "epoch": 13.852813852813853,
-      "grad_norm": 3.0845344066619873,
-      "learning_rate": 9.71404575267957e-06,
-      "loss": 0.5288,
+      "epoch": 6.926416259259312,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.70611974811303e-06,
+      "loss": 0.9735778045654296,
       "step": 601600
     },
     {
-      "epoch": 13.857419176568113,
-      "grad_norm": 2.9533259868621826,
-      "learning_rate": 9.706774188857057e-06,
-      "loss": 0.5563,
+      "epoch": 6.928718944334028,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.698848184290514e-06,
+      "loss": 0.9859268951416016,
       "step": 601800
     },
     {
-      "epoch": 13.862024500322372,
-      "grad_norm": 4.201800346374512,
-      "learning_rate": 9.699502625034541e-06,
-      "loss": 0.5445,
+      "epoch": 6.931021629408742,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.691576620467997e-06,
+      "loss": 0.9492156219482422,
       "step": 602000
     },
     {
-      "epoch": 13.866629824076632,
-      "grad_norm": 2.9285476207733154,
-      "learning_rate": 9.692231061212023e-06,
-      "loss": 0.5451,
+      "epoch": 6.933324314483457,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.684305056645483e-06,
+      "loss": 0.988839111328125,
       "step": 602200
     },
     {
-      "epoch": 13.871235147830893,
-      "grad_norm": 4.119074821472168,
-      "learning_rate": 9.68495949738951e-06,
-      "loss": 0.5433,
+      "epoch": 6.935626999558172,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.677033492822967e-06,
+      "loss": 1.00001220703125,
       "step": 602400
     },
     {
-      "epoch": 13.875840471585153,
-      "grad_norm": 4.315847396850586,
-      "learning_rate": 9.677687933566994e-06,
-      "loss": 0.5429,
+      "epoch": 6.937929684632888,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.669761929000451e-06,
+      "loss": 0.9814157104492187,
       "step": 602600
     },
     {
-      "epoch": 13.880445795339412,
-      "grad_norm": 3.370650053024292,
-      "learning_rate": 9.670416369744478e-06,
-      "loss": 0.538,
+      "epoch": 6.940232369707602,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.662490365177935e-06,
+      "loss": 0.9747137451171874,
       "step": 602800
     },
     {
-      "epoch": 13.885051119093673,
-      "grad_norm": 3.967122793197632,
-      "learning_rate": 9.663144805921962e-06,
-      "loss": 0.5346,
+      "epoch": 6.942535054782317,
+      "grad_norm": 1.125,
+      "learning_rate": 9.65521880135542e-06,
+      "loss": 0.9681507873535157,
       "step": 603000
     },
     {
-      "epoch": 13.889656442847933,
-      "grad_norm": 3.766850233078003,
-      "learning_rate": 9.655873242099446e-06,
-      "loss": 0.5345,
+      "epoch": 6.944837739857032,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.647947237532905e-06,
+      "loss": 0.9693973541259766,
       "step": 603200
     },
     {
-      "epoch": 13.894261766602192,
-      "grad_norm": 2.5367722511291504,
-      "learning_rate": 9.64860167827693e-06,
-      "loss": 0.5518,
+      "epoch": 6.947140424931747,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.640675673710388e-06,
+      "loss": 0.9954480743408203,
       "step": 603400
     },
     {
-      "epoch": 13.898867090356452,
-      "grad_norm": 3.3279225826263428,
-      "learning_rate": 9.641330114454415e-06,
-      "loss": 0.5508,
+      "epoch": 6.949443110006462,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.633404109887872e-06,
+      "loss": 0.9874771881103516,
       "step": 603600
     },
     {
-      "epoch": 13.903472414110713,
-      "grad_norm": 2.8402557373046875,
-      "learning_rate": 9.634058550631899e-06,
-      "loss": 0.5419,
+      "epoch": 6.951745795081177,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.626132546065358e-06,
+      "loss": 0.9629541778564453,
       "step": 603800
     },
     {
-      "epoch": 13.908077737864971,
-      "grad_norm": 2.7564351558685303,
-      "learning_rate": 9.626786986809383e-06,
-      "loss": 0.545,
+      "epoch": 6.954048480155892,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.61886098224284e-06,
+      "loss": 0.977274169921875,
       "step": 604000
     },
     {
-      "epoch": 13.912683061619232,
-      "grad_norm": 2.57075572013855,
-      "learning_rate": 9.61955178080598e-06,
-      "loss": 0.5444,
+      "epoch": 6.956351165230607,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.611589418420327e-06,
+      "loss": 0.996618881225586,
       "step": 604200
     },
     {
-      "epoch": 13.917288385373492,
-      "grad_norm": 3.027113676071167,
-      "learning_rate": 9.612280216983464e-06,
-      "loss": 0.5438,
+      "epoch": 6.958653850305321,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.60431785459781e-06,
+      "loss": 0.9640678405761719,
       "step": 604400
     },
     {
-      "epoch": 13.921893709127751,
-      "grad_norm": 3.521167516708374,
-      "learning_rate": 9.60500865316095e-06,
-      "loss": 0.5461,
+      "epoch": 6.960956535380037,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.597046290775293e-06,
+      "loss": 0.9609552001953126,
       "step": 604600
     },
     {
-      "epoch": 13.926499032882012,
-      "grad_norm": 3.0881049633026123,
-      "learning_rate": 9.597737089338434e-06,
-      "loss": 0.5481,
+      "epoch": 6.963259220454751,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.58977472695278e-06,
+      "loss": 0.9898723602294922,
       "step": 604800
     },
     {
-      "epoch": 13.931104356636272,
-      "grad_norm": 2.756943941116333,
-      "learning_rate": 9.590465525515917e-06,
-      "loss": 0.551,
+      "epoch": 6.965561905529467,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.582503163130263e-06,
+      "loss": 0.9942790985107421,
       "step": 605000
     },
     {
-      "epoch": 13.93570968039053,
-      "grad_norm": 2.7377870082855225,
-      "learning_rate": 9.583193961693403e-06,
-      "loss": 0.5483,
+      "epoch": 6.967864590604181,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.575231599307746e-06,
+      "loss": 0.9800389099121094,
       "step": 605200
     },
     {
-      "epoch": 13.940315004144791,
-      "grad_norm": 3.1983232498168945,
-      "learning_rate": 9.575922397870887e-06,
-      "loss": 0.5511,
+      "epoch": 6.970167275678897,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.567960035485232e-06,
+      "loss": 0.9767919158935547,
       "step": 605400
     },
     {
-      "epoch": 13.944920327899052,
-      "grad_norm": 2.8796114921569824,
-      "learning_rate": 9.56865083404837e-06,
-      "loss": 0.5355,
+      "epoch": 6.972469960753611,
+      "grad_norm": 1.390625,
+      "learning_rate": 9.560688471662716e-06,
+      "loss": 0.9678658294677734,
       "step": 605600
     },
     {
-      "epoch": 13.94952565165331,
-      "grad_norm": 3.955488443374634,
-      "learning_rate": 9.561379270225855e-06,
-      "loss": 0.5489,
+      "epoch": 6.974772645828327,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.553416907840202e-06,
+      "loss": 0.9765434265136719,
       "step": 605800
     },
     {
-      "epoch": 13.95413097540757,
-      "grad_norm": 3.366757869720459,
-      "learning_rate": 9.55410770640334e-06,
-      "loss": 0.5402,
+      "epoch": 6.977075330903041,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.546145344017684e-06,
+      "loss": 0.9870706939697266,
       "step": 606000
     },
     {
-      "epoch": 13.958736299161831,
-      "grad_norm": 3.5042965412139893,
-      "learning_rate": 9.546836142580824e-06,
-      "loss": 0.5414,
+      "epoch": 6.9793780159777565,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.538873780195169e-06,
+      "loss": 0.9717103576660157,
       "step": 606200
     },
     {
-      "epoch": 13.963341622916092,
-      "grad_norm": 4.029479026794434,
-      "learning_rate": 9.539564578758308e-06,
-      "loss": 0.5524,
+      "epoch": 6.981680701052471,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.531602216372655e-06,
+      "loss": 0.9808879089355469,
       "step": 606400
     },
     {
-      "epoch": 13.96794694667035,
-      "grad_norm": 3.883596897125244,
-      "learning_rate": 9.532293014935792e-06,
-      "loss": 0.5432,
+      "epoch": 6.983983386127186,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.524330652550137e-06,
+      "loss": 1.0108739471435546,
       "step": 606600
     },
     {
-      "epoch": 13.972552270424611,
-      "grad_norm": 3.188150644302368,
-      "learning_rate": 9.52505780893239e-06,
-      "loss": 0.5396,
+      "epoch": 6.986286071201901,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.517059088727621e-06,
+      "loss": 0.9731607818603516,
       "step": 606800
     },
     {
-      "epoch": 13.977157594178871,
-      "grad_norm": 2.3923470973968506,
-      "learning_rate": 9.517786245109873e-06,
-      "loss": 0.5358,
+      "epoch": 6.988588756276616,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.509787524905107e-06,
+      "loss": 0.9771558380126953,
       "step": 607000
     },
     {
-      "epoch": 13.98176291793313,
-      "grad_norm": 2.5199859142303467,
-      "learning_rate": 9.510514681287358e-06,
-      "loss": 0.5378,
+      "epoch": 6.990891441351331,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.50251596108259e-06,
+      "loss": 0.9768960571289063,
       "step": 607200
     },
     {
-      "epoch": 13.98636824168739,
-      "grad_norm": 3.514549970626831,
-      "learning_rate": 9.503279475283955e-06,
-      "loss": 0.5471,
+      "epoch": 6.9931941264260455,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.495244397260076e-06,
+      "loss": 0.986895523071289,
       "step": 607400
     },
     {
-      "epoch": 13.990973565441651,
-      "grad_norm": 3.0256316661834717,
-      "learning_rate": 9.496007911461439e-06,
-      "loss": 0.5447,
+      "epoch": 6.995496811500761,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.48797283343756e-06,
+      "loss": 0.9975414276123047,
       "step": 607600
     },
     {
-      "epoch": 13.99557888919591,
-      "grad_norm": 2.8945305347442627,
-      "learning_rate": 9.488736347638923e-06,
-      "loss": 0.5516,
+      "epoch": 6.9977994965754755,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.480701269615042e-06,
+      "loss": 1.000497055053711,
       "step": 607800
     },
     {
-      "epoch": 14.0,
-      "eval_loss": 0.5243151783943176,
-      "eval_runtime": 171.7907,
-      "eval_samples_per_second": 165.09,
-      "eval_steps_per_second": 10.321,
+      "epoch": 7.0,
+      "eval_loss": NaN,
+      "eval_runtime": 430.5016,
+      "eval_samples_per_second": 65.879,
+      "eval_steps_per_second": 32.941,
       "step": 607992
     },
     {
-      "epoch": 14.00018421295017,
-      "grad_norm": 3.234140634536743,
-      "learning_rate": 9.481464783816407e-06,
-      "loss": 0.5465,
+      "epoch": 7.000092107402988,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.473429705792528e-06,
+      "loss": 0.9825310516357422,
       "step": 608000
     },
     {
-      "epoch": 14.00478953670443,
-      "grad_norm": 3.9970383644104004,
-      "learning_rate": 9.474193219993893e-06,
-      "loss": 0.5352,
+      "epoch": 7.002394792477704,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.466158141970013e-06,
+      "loss": 0.9799756622314453,
       "step": 608200
     },
     {
-      "epoch": 14.00939486045869,
-      "grad_norm": 2.708507537841797,
-      "learning_rate": 9.466921656171377e-06,
-      "loss": 0.5386,
+      "epoch": 7.004697477552418,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.458886578147495e-06,
+      "loss": 0.9800210571289063,
       "step": 608400
     },
     {
-      "epoch": 14.01400018421295,
-      "grad_norm": 3.3150854110717773,
-      "learning_rate": 9.45965009234886e-06,
-      "loss": 0.5356,
+      "epoch": 7.007000162627134,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.451615014324981e-06,
+      "loss": 0.9912975311279297,
       "step": 608600
     },
     {
-      "epoch": 14.01860550796721,
-      "grad_norm": 2.4977152347564697,
-      "learning_rate": 9.452378528526346e-06,
-      "loss": 0.5303,
+      "epoch": 7.009302847701848,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.444343450502465e-06,
+      "loss": 0.9743272399902344,
       "step": 608800
     },
     {
-      "epoch": 14.02321083172147,
-      "grad_norm": 2.6058592796325684,
-      "learning_rate": 9.44510696470383e-06,
-      "loss": 0.5452,
+      "epoch": 7.011605532776564,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.43707188667995e-06,
+      "loss": 0.9799197387695312,
       "step": 609000
     },
     {
-      "epoch": 14.02781615547573,
-      "grad_norm": 3.491718292236328,
-      "learning_rate": 9.437835400881313e-06,
-      "loss": 0.5431,
+      "epoch": 7.013908217851278,
+      "grad_norm": 0.5,
+      "learning_rate": 9.429800322857434e-06,
+      "loss": 0.9837110137939453,
       "step": 609200
     },
     {
-      "epoch": 14.03242147922999,
-      "grad_norm": 3.100562810897827,
-      "learning_rate": 9.430563837058798e-06,
-      "loss": 0.534,
+      "epoch": 7.016210902925993,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.422528759034918e-06,
+      "loss": 1.004596939086914,
       "step": 609400
     },
     {
-      "epoch": 14.03702680298425,
-      "grad_norm": 3.014887809753418,
-      "learning_rate": 9.423292273236283e-06,
-      "loss": 0.5268,
+      "epoch": 7.018513588000708,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.415257195212404e-06,
+      "loss": 0.9808485412597656,
       "step": 609600
     },
     {
-      "epoch": 14.04163212673851,
-      "grad_norm": 4.120047569274902,
-      "learning_rate": 9.416020709413767e-06,
-      "loss": 0.5492,
+      "epoch": 7.020816273075423,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.407985631389886e-06,
+      "loss": 0.9921746826171876,
       "step": 609800
     },
     {
-      "epoch": 14.04623745049277,
-      "grad_norm": 3.292614698410034,
-      "learning_rate": 9.408749145591251e-06,
-      "loss": 0.5406,
+      "epoch": 7.023118958150138,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.40071406756737e-06,
+      "loss": 0.9875336456298828,
       "step": 610000
     },
     {
-      "epoch": 14.05084277424703,
-      "grad_norm": 2.325287342071533,
-      "learning_rate": 9.401477581768735e-06,
-      "loss": 0.5308,
+      "epoch": 7.025421643224853,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.393442503744856e-06,
+      "loss": 0.9784149932861328,
       "step": 610200
     },
     {
-      "epoch": 14.055448098001289,
-      "grad_norm": 2.72371244430542,
-      "learning_rate": 9.394242375765332e-06,
-      "loss": 0.5306,
+      "epoch": 7.027724328299568,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.386170939922339e-06,
+      "loss": 0.9732196044921875,
       "step": 610400
     },
     {
-      "epoch": 14.06005342175555,
-      "grad_norm": 2.5140206813812256,
-      "learning_rate": 9.386970811942816e-06,
-      "loss": 0.5412,
+      "epoch": 7.030027013374283,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.378899376099825e-06,
+      "loss": 1.0045655822753907,
       "step": 610600
     },
     {
-      "epoch": 14.06465874550981,
-      "grad_norm": 2.95316481590271,
-      "learning_rate": 9.3796992481203e-06,
-      "loss": 0.5408,
+      "epoch": 7.032329698448998,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.371627812277309e-06,
+      "loss": 0.9965348052978515,
       "step": 610800
     },
     {
-      "epoch": 14.069264069264069,
-      "grad_norm": 3.183243989944458,
-      "learning_rate": 9.372427684297785e-06,
-      "loss": 0.5371,
+      "epoch": 7.034632383523713,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.364356248454792e-06,
+      "loss": 0.9734232330322266,
       "step": 611000
     },
     {
-      "epoch": 14.07386939301833,
-      "grad_norm": 2.708509683609009,
-      "learning_rate": 9.365156120475269e-06,
-      "loss": 0.5479,
+      "epoch": 7.036935068598428,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.357084684632278e-06,
+      "loss": 0.9870662689208984,
       "step": 611200
     },
     {
-      "epoch": 14.07847471677259,
-      "grad_norm": 3.455313205718994,
-      "learning_rate": 9.357884556652753e-06,
-      "loss": 0.5429,
+      "epoch": 7.0392377536731425,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.349813120809762e-06,
+      "loss": 0.9682694244384765,
       "step": 611400
     },
     {
-      "epoch": 14.083080040526848,
-      "grad_norm": 3.839599609375,
-      "learning_rate": 9.35061299283024e-06,
-      "loss": 0.5455,
+      "epoch": 7.041540438747857,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.342541556987246e-06,
+      "loss": 0.9698184967041016,
       "step": 611600
     },
     {
-      "epoch": 14.087685364281109,
-      "grad_norm": 3.0706214904785156,
-      "learning_rate": 9.343341429007722e-06,
-      "loss": 0.5382,
+      "epoch": 7.0438431238225725,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.33526999316473e-06,
+      "loss": 0.9657843780517578,
       "step": 611800
     },
     {
-      "epoch": 14.09229068803537,
-      "grad_norm": 2.855938673019409,
-      "learning_rate": 9.336069865185206e-06,
-      "loss": 0.5346,
+      "epoch": 7.046145808897287,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.327998429342214e-06,
+      "loss": 0.9849095153808594,
       "step": 612000
     },
     {
-      "epoch": 14.096896011789628,
-      "grad_norm": 2.6702816486358643,
-      "learning_rate": 9.328798301362692e-06,
-      "loss": 0.5304,
+      "epoch": 7.0484484939720025,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.320726865519699e-06,
+      "loss": 0.9951589965820312,
       "step": 612200
     },
     {
-      "epoch": 14.101501335543889,
-      "grad_norm": 2.650740146636963,
-      "learning_rate": 9.321526737540176e-06,
-      "loss": 0.549,
+      "epoch": 7.050751179046717,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.313455301697183e-06,
+      "loss": 0.9628787231445313,
       "step": 612400
     },
     {
-      "epoch": 14.106106659298149,
-      "grad_norm": 4.2644124031066895,
-      "learning_rate": 9.31425517371766e-06,
-      "loss": 0.5448,
+      "epoch": 7.053053864121432,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.306183737874667e-06,
+      "loss": 0.9930134582519531,
       "step": 612600
     },
     {
-      "epoch": 14.110711983052408,
-      "grad_norm": 4.463499546051025,
-      "learning_rate": 9.306983609895145e-06,
-      "loss": 0.5472,
+      "epoch": 7.055356549196147,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.298912174052153e-06,
+      "loss": 0.9693550872802734,
       "step": 612800
     },
     {
-      "epoch": 14.115317306806668,
-      "grad_norm": 3.099343776702881,
-      "learning_rate": 9.299712046072629e-06,
-      "loss": 0.5316,
+      "epoch": 7.057659234270862,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.291640610229636e-06,
+      "loss": 0.9798488616943359,
       "step": 613000
     },
     {
-      "epoch": 14.119922630560929,
-      "grad_norm": 3.3200466632843018,
-      "learning_rate": 9.292440482250113e-06,
-      "loss": 0.5424,
+      "epoch": 7.059961919345577,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.284369046407121e-06,
+      "loss": 0.9596015167236328,
       "step": 613200
     },
     {
-      "epoch": 14.12452795431519,
-      "grad_norm": 4.582034587860107,
-      "learning_rate": 9.285168918427597e-06,
-      "loss": 0.5322,
+      "epoch": 7.0622646044202915,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.277097482584606e-06,
+      "loss": 0.9790519714355469,
       "step": 613400
     },
     {
-      "epoch": 14.129133278069448,
-      "grad_norm": 2.7270848751068115,
-      "learning_rate": 9.277897354605081e-06,
-      "loss": 0.5405,
+      "epoch": 7.064567289495007,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.269825918762088e-06,
+      "loss": 0.9601467895507813,
       "step": 613600
     },
     {
-      "epoch": 14.133738601823708,
-      "grad_norm": 3.561002731323242,
-      "learning_rate": 9.270625790782566e-06,
-      "loss": 0.5403,
+      "epoch": 7.066869974569721,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.262554354939574e-06,
+      "loss": 0.9998540496826172,
       "step": 613800
     },
     {
-      "epoch": 14.138343925577969,
-      "grad_norm": 2.916008710861206,
-      "learning_rate": 9.26335422696005e-06,
-      "loss": 0.5363,
+      "epoch": 7.069172659644437,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.255282791117058e-06,
+      "loss": 0.9718901824951172,
       "step": 614000
     },
     {
-      "epoch": 14.142949249332228,
-      "grad_norm": 3.4459245204925537,
-      "learning_rate": 9.256082663137536e-06,
-      "loss": 0.5467,
+      "epoch": 7.071475344719151,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.248011227294541e-06,
+      "loss": 0.9607735443115234,
       "step": 614200
     },
     {
-      "epoch": 14.147554573086488,
-      "grad_norm": 4.279280662536621,
-      "learning_rate": 9.248811099315018e-06,
-      "loss": 0.5397,
+      "epoch": 7.073778029793867,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.240739663472027e-06,
+      "loss": 0.9820717620849609,
       "step": 614400
     },
     {
-      "epoch": 14.152159896840748,
-      "grad_norm": 3.383220672607422,
-      "learning_rate": 9.241539535492503e-06,
-      "loss": 0.5442,
+      "epoch": 7.076080714868581,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.233468099649511e-06,
+      "loss": 0.9817266845703125,
       "step": 614600
     },
     {
-      "epoch": 14.156765220595007,
-      "grad_norm": 2.799022912979126,
-      "learning_rate": 9.234267971669988e-06,
-      "loss": 0.5455,
+      "epoch": 7.078383399943297,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.226196535826995e-06,
+      "loss": 0.9754546356201171,
       "step": 614800
     },
     {
-      "epoch": 14.161370544349268,
-      "grad_norm": 3.1966896057128906,
-      "learning_rate": 9.226996407847471e-06,
-      "loss": 0.5399,
+      "epoch": 7.080686085018011,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.21892497200448e-06,
+      "loss": 0.9722274780273438,
       "step": 615000
     },
     {
-      "epoch": 14.165975868103528,
-      "grad_norm": 3.2406864166259766,
-      "learning_rate": 9.219724844024955e-06,
-      "loss": 0.543,
+      "epoch": 7.082988770092726,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.211653408181964e-06,
+      "loss": 0.9742633056640625,
       "step": 615200
     },
     {
-      "epoch": 14.170581191857787,
-      "grad_norm": 2.572612762451172,
-      "learning_rate": 9.212453280202441e-06,
-      "loss": 0.5397,
+      "epoch": 7.085291455167441,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.204381844359448e-06,
+      "loss": 0.961524658203125,
       "step": 615400
     },
     {
-      "epoch": 14.175186515612047,
-      "grad_norm": 2.9783430099487305,
-      "learning_rate": 9.205181716379925e-06,
-      "loss": 0.5372,
+      "epoch": 7.087594140242156,
+      "grad_norm": 0.75,
+      "learning_rate": 9.197110280536932e-06,
+      "loss": 0.9746446990966797,
       "step": 615600
     },
     {
-      "epoch": 14.179791839366308,
-      "grad_norm": 3.033306837081909,
-      "learning_rate": 9.19791015255741e-06,
-      "loss": 0.5424,
+      "epoch": 7.089896825316871,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.189838716714416e-06,
+      "loss": 1.0018827056884765,
       "step": 615800
     },
     {
-      "epoch": 14.184397163120567,
-      "grad_norm": 3.275662660598755,
-      "learning_rate": 9.190638588734894e-06,
-      "loss": 0.5333,
+      "epoch": 7.092199510391586,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.182567152891902e-06,
+      "loss": 0.9909403991699218,
       "step": 616000
     },
     {
-      "epoch": 14.189002486874827,
-      "grad_norm": 2.4978132247924805,
-      "learning_rate": 9.183367024912378e-06,
-      "loss": 0.5341,
+      "epoch": 7.094502195466301,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.175295589069385e-06,
+      "loss": 0.9678092193603516,
       "step": 616200
     },
     {
-      "epoch": 14.193607810629087,
-      "grad_norm": 2.4442851543426514,
-      "learning_rate": 9.176095461089862e-06,
-      "loss": 0.5327,
+      "epoch": 7.096804880541016,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.16802402524687e-06,
+      "loss": 0.9697416687011718,
       "step": 616400
     },
     {
-      "epoch": 14.198213134383348,
-      "grad_norm": 3.3033714294433594,
-      "learning_rate": 9.168823897267346e-06,
-      "loss": 0.527,
+      "epoch": 7.099107565615731,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.160752461424355e-06,
+      "loss": 0.9911090850830078,
       "step": 616600
     },
     {
-      "epoch": 14.202818458137607,
-      "grad_norm": 3.127436399459839,
-      "learning_rate": 9.16155233344483e-06,
-      "loss": 0.5371,
+      "epoch": 7.101410250690446,
+      "grad_norm": 0.75,
+      "learning_rate": 9.153480897601837e-06,
+      "loss": 0.9620626068115234,
       "step": 616800
     },
     {
-      "epoch": 14.207423781891867,
-      "grad_norm": 3.2301642894744873,
-      "learning_rate": 9.154280769622315e-06,
-      "loss": 0.537,
+      "epoch": 7.10371293576516,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.146209333779323e-06,
+      "loss": 0.9994487762451172,
       "step": 617000
     },
     {
-      "epoch": 14.212029105646128,
-      "grad_norm": 3.059586763381958,
-      "learning_rate": 9.147009205799799e-06,
-      "loss": 0.5396,
+      "epoch": 7.106015620839876,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.138937769956808e-06,
+      "loss": 0.9509933471679688,
       "step": 617200
     },
     {
-      "epoch": 14.216634429400386,
-      "grad_norm": 3.4187586307525635,
-      "learning_rate": 9.139773999796396e-06,
-      "loss": 0.5429,
+      "epoch": 7.10831830591459,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.131666206134292e-06,
+      "loss": 0.9687966918945312,
       "step": 617400
     },
     {
-      "epoch": 14.221239753154647,
-      "grad_norm": 2.5766658782958984,
-      "learning_rate": 9.132502435973882e-06,
-      "loss": 0.5241,
+      "epoch": 7.110620990989306,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.124394642311776e-06,
+      "loss": 0.9788529968261719,
       "step": 617600
     },
     {
-      "epoch": 14.225845076908907,
-      "grad_norm": 3.2574641704559326,
-      "learning_rate": 9.125230872151364e-06,
-      "loss": 0.5498,
+      "epoch": 7.11292367606402,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.11712307848926e-06,
+      "loss": 0.9835548400878906,
       "step": 617800
     },
     {
-      "epoch": 14.230450400663166,
-      "grad_norm": 3.0240838527679443,
-      "learning_rate": 9.117995666147961e-06,
-      "loss": 0.5397,
+      "epoch": 7.115226361138736,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.109851514666744e-06,
+      "loss": 1.001695556640625,
       "step": 618000
     },
     {
-      "epoch": 14.235055724417427,
-      "grad_norm": 3.250147819519043,
-      "learning_rate": 9.110724102325446e-06,
-      "loss": 0.5319,
+      "epoch": 7.11752904621345,
+      "grad_norm": 1.125,
+      "learning_rate": 9.102579950844229e-06,
+      "loss": 0.989010238647461,
       "step": 618200
     },
     {
-      "epoch": 14.239661048171687,
-      "grad_norm": 3.770249128341675,
-      "learning_rate": 9.103488896322042e-06,
-      "loss": 0.5388,
+      "epoch": 7.119831731288166,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.095308387021713e-06,
+      "loss": 0.9796863555908203,
       "step": 618400
     },
     {
-      "epoch": 14.244266371925946,
-      "grad_norm": 3.1490142345428467,
-      "learning_rate": 9.096217332499528e-06,
-      "loss": 0.5409,
+      "epoch": 7.12213441636288,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.088036823199197e-06,
+      "loss": 1.0017696380615235,
       "step": 618600
     },
     {
-      "epoch": 14.248871695680206,
-      "grad_norm": 3.5395240783691406,
-      "learning_rate": 9.088945768677011e-06,
-      "loss": 0.5395,
+      "epoch": 7.124437101437595,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.080765259376681e-06,
+      "loss": 0.9877196502685547,
       "step": 618800
     },
     {
-      "epoch": 14.253477019434467,
-      "grad_norm": 3.340034246444702,
-      "learning_rate": 9.081674204854495e-06,
-      "loss": 0.5388,
+      "epoch": 7.12673978651231,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.073493695554167e-06,
+      "loss": 0.9870804595947266,
       "step": 619000
     },
     {
-      "epoch": 14.258082343188725,
-      "grad_norm": 2.908026933670044,
-      "learning_rate": 9.074402641031981e-06,
-      "loss": 0.5424,
+      "epoch": 7.129042471587025,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.066222131731651e-06,
+      "loss": 0.9838044738769531,
       "step": 619200
     },
     {
-      "epoch": 14.262687666942986,
-      "grad_norm": 2.7248940467834473,
-      "learning_rate": 9.067131077209465e-06,
-      "loss": 0.5236,
+      "epoch": 7.13134515666174,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.058950567909134e-06,
+      "loss": 0.9858052825927734,
       "step": 619400
     },
     {
-      "epoch": 14.267292990697246,
-      "grad_norm": 3.4366588592529297,
-      "learning_rate": 9.05985951338695e-06,
-      "loss": 0.5526,
+      "epoch": 7.133647841736455,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.05167900408662e-06,
+      "loss": 0.9669033050537109,
       "step": 619600
     },
     {
-      "epoch": 14.271898314451505,
-      "grad_norm": 3.0255722999572754,
-      "learning_rate": 9.052587949564434e-06,
-      "loss": 0.5282,
+      "epoch": 7.13595052681117,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.044407440264104e-06,
+      "loss": 1.0087360382080077,
       "step": 619800
     },
     {
-      "epoch": 14.276503638205766,
-      "grad_norm": 3.2476325035095215,
-      "learning_rate": 9.045316385741918e-06,
-      "loss": 0.5384,
+      "epoch": 7.138253211885885,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.037135876441587e-06,
+      "loss": 0.995894775390625,
       "step": 620000
     },
     {
-      "epoch": 14.281108961960026,
-      "grad_norm": 3.1371731758117676,
-      "learning_rate": 9.038044821919402e-06,
-      "loss": 0.539,
+      "epoch": 7.1405558969606,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.029864312619073e-06,
+      "loss": 0.9510137176513672,
       "step": 620200
     },
     {
-      "epoch": 14.285714285714286,
-      "grad_norm": 2.5684289932250977,
-      "learning_rate": 9.030773258096886e-06,
-      "loss": 0.5288,
+      "epoch": 7.1428585820353145,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.022592748796557e-06,
+      "loss": 0.9642363739013672,
       "step": 620400
     },
     {
-      "epoch": 14.290319609468545,
-      "grad_norm": 3.3922548294067383,
-      "learning_rate": 9.023501694274372e-06,
-      "loss": 0.5344,
+      "epoch": 7.145161267110029,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.015321184974041e-06,
+      "loss": 0.9794168090820312,
       "step": 620600
     },
     {
-      "epoch": 14.294924933222806,
-      "grad_norm": 2.9498867988586426,
-      "learning_rate": 9.016230130451855e-06,
-      "loss": 0.5327,
+      "epoch": 7.1474639521847445,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.008049621151525e-06,
+      "loss": 0.9878856658935546,
       "step": 620800
     },
     {
-      "epoch": 14.299530256977066,
-      "grad_norm": 3.503101110458374,
-      "learning_rate": 9.008958566629339e-06,
-      "loss": 0.5408,
+      "epoch": 7.149766637259459,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.00077805732901e-06,
+      "loss": 0.982391357421875,
       "step": 621000
     },
     {
-      "epoch": 14.304135580731325,
-      "grad_norm": 2.9323580265045166,
-      "learning_rate": 9.001687002806825e-06,
-      "loss": 0.5407,
+      "epoch": 7.1520693223341745,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.993506493506494e-06,
+      "loss": 0.9797698211669922,
       "step": 621200
     },
     {
-      "epoch": 14.308740904485585,
-      "grad_norm": 3.0948574542999268,
-      "learning_rate": 8.994415438984307e-06,
-      "loss": 0.543,
+      "epoch": 7.154372007408889,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.986234929683978e-06,
+      "loss": 0.9885771942138671,
       "step": 621400
     },
     {
-      "epoch": 14.313346228239846,
-      "grad_norm": 2.787778854370117,
-      "learning_rate": 8.987180232980904e-06,
-      "loss": 0.5334,
+      "epoch": 7.156674692483604,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.978963365861462e-06,
+      "loss": 0.9900411224365234,
       "step": 621600
     },
     {
-      "epoch": 14.317951551994105,
-      "grad_norm": 2.2373883724212646,
-      "learning_rate": 8.979908669158389e-06,
-      "loss": 0.5355,
+      "epoch": 7.158977377558319,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.971691802038946e-06,
+      "loss": 0.9649449157714843,
       "step": 621800
     },
     {
-      "epoch": 14.322556875748365,
-      "grad_norm": 3.0317771434783936,
-      "learning_rate": 8.972637105335874e-06,
-      "loss": 0.5372,
+      "epoch": 7.161280062633034,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.96442023821643e-06,
+      "loss": 0.9854252624511719,
       "step": 622000
     },
     {
-      "epoch": 14.327162199502625,
-      "grad_norm": 3.486945152282715,
-      "learning_rate": 8.965365541513357e-06,
-      "loss": 0.5294,
+      "epoch": 7.163582747707749,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.957148674393916e-06,
+      "loss": 0.9867980194091797,
       "step": 622200
     },
     {
-      "epoch": 14.331767523256884,
-      "grad_norm": 3.084419012069702,
-      "learning_rate": 8.958093977690843e-06,
-      "loss": 0.545,
+      "epoch": 7.1658854327824635,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.949877110571399e-06,
+      "loss": 0.9880463409423829,
       "step": 622400
     },
     {
-      "epoch": 14.336372847011145,
-      "grad_norm": 3.271340847015381,
-      "learning_rate": 8.950822413868327e-06,
-      "loss": 0.5436,
+      "epoch": 7.168188117857179,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.942605546748883e-06,
+      "loss": 1.0098218536376953,
       "step": 622600
     },
     {
-      "epoch": 14.340978170765405,
-      "grad_norm": 2.878502607345581,
-      "learning_rate": 8.943550850045811e-06,
-      "loss": 0.5316,
+      "epoch": 7.1704908029318934,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.935333982926369e-06,
+      "loss": 0.964631576538086,
       "step": 622800
     },
     {
-      "epoch": 14.345583494519666,
-      "grad_norm": 3.714315414428711,
-      "learning_rate": 8.936279286223296e-06,
-      "loss": 0.5538,
+      "epoch": 7.172793488006609,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.928062419103853e-06,
+      "loss": 0.9773719787597657,
       "step": 623000
     },
     {
-      "epoch": 14.350188818273924,
-      "grad_norm": 3.77945876121521,
-      "learning_rate": 8.92900772240078e-06,
-      "loss": 0.5426,
+      "epoch": 7.175096173081323,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.920790855281336e-06,
+      "loss": 0.9826641082763672,
       "step": 623200
     },
     {
-      "epoch": 14.354794142028185,
-      "grad_norm": 3.177495241165161,
-      "learning_rate": 8.921736158578264e-06,
-      "loss": 0.5407,
+      "epoch": 7.177398858156039,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.913519291458822e-06,
+      "loss": 0.9746383666992188,
       "step": 623400
     },
     {
-      "epoch": 14.359399465782445,
-      "grad_norm": 2.679957151412964,
-      "learning_rate": 8.914464594755748e-06,
-      "loss": 0.5439,
+      "epoch": 7.179701543230753,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.906247727636306e-06,
+      "loss": 0.9711588287353515,
       "step": 623600
     },
     {
-      "epoch": 14.364004789536704,
-      "grad_norm": 3.1106204986572266,
-      "learning_rate": 8.907193030933232e-06,
-      "loss": 0.5319,
+      "epoch": 7.182004228305469,
+      "grad_norm": 1.0,
+      "learning_rate": 8.89897616381379e-06,
+      "loss": 0.9825574493408203,
       "step": 623800
     },
     {
-      "epoch": 14.368610113290964,
-      "grad_norm": 4.358025550842285,
-      "learning_rate": 8.899921467110718e-06,
-      "loss": 0.5272,
+      "epoch": 7.184306913380183,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.891704599991274e-06,
+      "loss": 0.9776390075683594,
       "step": 624000
     },
     {
-      "epoch": 14.373215437045225,
-      "grad_norm": 3.5645134449005127,
-      "learning_rate": 8.892649903288201e-06,
-      "loss": 0.5265,
+      "epoch": 7.186609598454898,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.884433036168759e-06,
+      "loss": 0.9564485168457031,
       "step": 624200
     },
     {
-      "epoch": 14.377820760799484,
-      "grad_norm": 3.1196436882019043,
-      "learning_rate": 8.885378339465685e-06,
-      "loss": 0.5261,
+      "epoch": 7.188912283529613,
+      "grad_norm": 8.25,
+      "learning_rate": 8.877161472346243e-06,
+      "loss": 0.9756488037109375,
       "step": 624400
     },
     {
-      "epoch": 14.382426084553744,
-      "grad_norm": 3.206526517868042,
-      "learning_rate": 8.878106775643171e-06,
-      "loss": 0.5461,
+      "epoch": 7.191214968604328,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.869889908523727e-06,
+      "loss": 0.9909695434570313,
       "step": 624600
     },
     {
-      "epoch": 14.387031408308005,
-      "grad_norm": 3.2251393795013428,
-      "learning_rate": 8.870835211820654e-06,
-      "loss": 0.5388,
+      "epoch": 7.193517653679043,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.862618344701213e-06,
+      "loss": 0.9954541015625,
       "step": 624800
     },
     {
-      "epoch": 14.391636732062263,
-      "grad_norm": 2.812065362930298,
-      "learning_rate": 8.86360000581725e-06,
-      "loss": 0.5345,
+      "epoch": 7.195820338753758,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.855346780878695e-06,
+      "loss": 0.9796237945556641,
       "step": 625000
     },
     {
-      "epoch": 14.396242055816524,
-      "grad_norm": 3.157151460647583,
-      "learning_rate": 8.856328441994735e-06,
-      "loss": 0.5436,
+      "epoch": 7.198123023828473,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.84807521705618e-06,
+      "loss": 0.9632915496826172,
       "step": 625200
     },
     {
-      "epoch": 14.400847379570784,
-      "grad_norm": 3.392270088195801,
-      "learning_rate": 8.84905687817222e-06,
-      "loss": 0.527,
+      "epoch": 7.200425708903188,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.840803653233666e-06,
+      "loss": 0.9457245635986328,
       "step": 625400
     },
     {
-      "epoch": 14.405452703325043,
-      "grad_norm": 2.6218199729919434,
-      "learning_rate": 8.841785314349703e-06,
-      "loss": 0.5309,
+      "epoch": 7.202728393977903,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.833532089411148e-06,
+      "loss": 0.9884097290039062,
       "step": 625600
     },
     {
-      "epoch": 14.410058027079303,
-      "grad_norm": 2.878702402114868,
-      "learning_rate": 8.834513750527189e-06,
-      "loss": 0.5339,
+      "epoch": 7.205031079052618,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.826260525588632e-06,
+      "loss": 0.9882837677001953,
       "step": 625800
     },
     {
-      "epoch": 14.414663350833564,
-      "grad_norm": 3.8115878105163574,
-      "learning_rate": 8.827242186704673e-06,
-      "loss": 0.5313,
+      "epoch": 7.207333764127332,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.818988961766118e-06,
+      "loss": 0.9795391845703125,
       "step": 626000
     },
     {
-      "epoch": 14.419268674587823,
-      "grad_norm": 3.6281392574310303,
-      "learning_rate": 8.819970622882157e-06,
-      "loss": 0.5352,
+      "epoch": 7.209636449202048,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.811717397943602e-06,
+      "loss": 0.9771727752685547,
       "step": 626200
     },
     {
-      "epoch": 14.423873998342083,
-      "grad_norm": 3.010791301727295,
-      "learning_rate": 8.812699059059642e-06,
-      "loss": 0.5449,
+      "epoch": 7.211939134276762,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.804445834121087e-06,
+      "loss": 0.9845864868164063,
       "step": 626400
     },
     {
-      "epoch": 14.428479322096344,
-      "grad_norm": 3.1626334190368652,
-      "learning_rate": 8.805427495237126e-06,
-      "loss": 0.5427,
+      "epoch": 7.214241819351478,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.797174270298571e-06,
+      "loss": 0.9764852905273438,
       "step": 626600
     },
     {
-      "epoch": 14.433084645850604,
-      "grad_norm": 4.617162704467773,
-      "learning_rate": 8.79815593141461e-06,
-      "loss": 0.5296,
+      "epoch": 7.216544504426192,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.789902706476055e-06,
+      "loss": 0.9779194641113281,
       "step": 626800
     },
     {
-      "epoch": 14.437689969604863,
-      "grad_norm": 2.7026705741882324,
-      "learning_rate": 8.790884367592094e-06,
-      "loss": 0.5386,
+      "epoch": 7.218847189500908,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.78263114265354e-06,
+      "loss": 0.9794198608398438,
       "step": 627000
     },
     {
-      "epoch": 14.442295293359123,
-      "grad_norm": 2.913951873779297,
-      "learning_rate": 8.783612803769579e-06,
-      "loss": 0.5436,
+      "epoch": 7.221149874575622,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.775359578831024e-06,
+      "loss": 0.9964470672607422,
       "step": 627200
     },
     {
-      "epoch": 14.446900617113384,
-      "grad_norm": 3.2576162815093994,
-      "learning_rate": 8.776341239947064e-06,
-      "loss": 0.5406,
+      "epoch": 7.223452559650338,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.768088015008508e-06,
+      "loss": 0.9815599822998047,
       "step": 627400
     },
     {
-      "epoch": 14.451505940867643,
-      "grad_norm": 2.7531659603118896,
-      "learning_rate": 8.769069676124547e-06,
-      "loss": 0.5552,
+      "epoch": 7.225755244725052,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.760816451185992e-06,
+      "loss": 0.9903423309326171,
       "step": 627600
     },
     {
-      "epoch": 14.456111264621903,
-      "grad_norm": 3.288334608078003,
-      "learning_rate": 8.761798112302031e-06,
-      "loss": 0.5285,
+      "epoch": 7.228057929799768,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.753544887363476e-06,
+      "loss": 0.9563906097412109,
       "step": 627800
     },
     {
-      "epoch": 14.460716588376163,
-      "grad_norm": 2.9675936698913574,
-      "learning_rate": 8.754526548479517e-06,
-      "loss": 0.5303,
+      "epoch": 7.230360614874482,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.746273323540962e-06,
+      "loss": 0.9850975036621094,
       "step": 628000
     },
     {
-      "epoch": 14.465321912130422,
-      "grad_norm": 3.810732364654541,
-      "learning_rate": 8.747254984657e-06,
-      "loss": 0.5309,
+      "epoch": 7.232663299949197,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.739001759718445e-06,
+      "loss": 0.9879227447509765,
       "step": 628200
     },
     {
-      "epoch": 14.469927235884683,
-      "grad_norm": 3.084245204925537,
-      "learning_rate": 8.739983420834486e-06,
-      "loss": 0.53,
+      "epoch": 7.234965985023912,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.731730195895929e-06,
+      "loss": 0.9770551300048829,
       "step": 628400
     },
     {
-      "epoch": 14.474532559638943,
-      "grad_norm": 3.4375557899475098,
-      "learning_rate": 8.73274821483108e-06,
-      "loss": 0.5417,
+      "epoch": 7.237268670098627,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.724458632073415e-06,
+      "loss": 0.991946029663086,
       "step": 628600
     },
     {
-      "epoch": 14.479137883393202,
-      "grad_norm": 4.032174587249756,
-      "learning_rate": 8.725476651008567e-06,
-      "loss": 0.5392,
+      "epoch": 7.239571355173342,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.717187068250897e-06,
+      "loss": 0.9801811218261719,
       "step": 628800
     },
     {
-      "epoch": 14.483743207147462,
-      "grad_norm": 3.3805575370788574,
-      "learning_rate": 8.71820508718605e-06,
-      "loss": 0.5307,
+      "epoch": 7.241874040248057,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.709915504428382e-06,
+      "loss": 0.9778077697753906,
       "step": 629000
     },
     {
-      "epoch": 14.488348530901723,
-      "grad_norm": 3.0205650329589844,
-      "learning_rate": 8.710933523363535e-06,
-      "loss": 0.5383,
+      "epoch": 7.244176725322772,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.702643940605867e-06,
+      "loss": 0.9597901153564453,
       "step": 629200
     },
     {
-      "epoch": 14.492953854655982,
-      "grad_norm": 2.994614601135254,
-      "learning_rate": 8.70366195954102e-06,
-      "loss": 0.543,
+      "epoch": 7.2464794103974866,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.695372376783352e-06,
+      "loss": 0.9801737976074218,
       "step": 629400
     },
     {
-      "epoch": 14.497559178410242,
-      "grad_norm": 3.1769533157348633,
-      "learning_rate": 8.696390395718504e-06,
-      "loss": 0.5316,
+      "epoch": 7.248782095472202,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.688100812960836e-06,
+      "loss": 0.9846310424804687,
       "step": 629600
     },
     {
-      "epoch": 14.502164502164502,
-      "grad_norm": 2.6771903038024902,
-      "learning_rate": 8.689118831895988e-06,
-      "loss": 0.5426,
+      "epoch": 7.2510847805469165,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.68082924913832e-06,
+      "loss": 0.9861672973632812,
       "step": 629800
     },
     {
-      "epoch": 14.506769825918763,
-      "grad_norm": 2.680912494659424,
-      "learning_rate": 8.681847268073472e-06,
-      "loss": 0.5232,
+      "epoch": 7.253387465621631,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.673557685315804e-06,
+      "loss": 1.0003617095947266,
       "step": 630000
     },
     {
-      "epoch": 14.511375149673022,
-      "grad_norm": 2.9476630687713623,
-      "learning_rate": 8.674575704250958e-06,
-      "loss": 0.5455,
+      "epoch": 7.2556901506963465,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.666286121493289e-06,
+      "loss": 0.9874579620361328,
       "step": 630200
     },
     {
-      "epoch": 14.515980473427282,
-      "grad_norm": 4.284652233123779,
-      "learning_rate": 8.66730414042844e-06,
-      "loss": 0.5425,
+      "epoch": 7.257992835771061,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.659014557670773e-06,
+      "loss": 0.9611412811279297,
       "step": 630400
     },
     {
-      "epoch": 14.520585797181543,
-      "grad_norm": 3.3897085189819336,
-      "learning_rate": 8.660032576605925e-06,
-      "loss": 0.5245,
+      "epoch": 7.2602955208457765,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.651742993848257e-06,
+      "loss": 0.98235107421875,
       "step": 630600
     },
     {
-      "epoch": 14.525191120935801,
-      "grad_norm": 2.838503122329712,
-      "learning_rate": 8.65276101278341e-06,
-      "loss": 0.5334,
+      "epoch": 7.262598205920491,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.644471430025741e-06,
+      "loss": 0.9877023315429687,
       "step": 630800
     },
     {
-      "epoch": 14.529796444690062,
-      "grad_norm": 2.7507781982421875,
-      "learning_rate": 8.645489448960893e-06,
-      "loss": 0.5292,
+      "epoch": 7.264900890995206,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.637199866203225e-06,
+      "loss": 0.9970263671875,
       "step": 631000
     },
     {
-      "epoch": 14.534401768444322,
-      "grad_norm": 4.586523532867432,
-      "learning_rate": 8.638217885138377e-06,
-      "loss": 0.543,
+      "epoch": 7.267203576069921,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.629928302380711e-06,
+      "loss": 0.9533394622802734,
       "step": 631200
     },
     {
-      "epoch": 14.539007092198581,
-      "grad_norm": 2.6941652297973633,
-      "learning_rate": 8.630946321315863e-06,
-      "loss": 0.5372,
+      "epoch": 7.269506261144636,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.622656738558194e-06,
+      "loss": 0.9910484313964844,
       "step": 631400
     },
     {
-      "epoch": 14.543612415952841,
-      "grad_norm": 2.9810523986816406,
-      "learning_rate": 8.62371111531246e-06,
-      "loss": 0.5382,
+      "epoch": 7.271808946219351,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.615385174735678e-06,
+      "loss": 0.9794743347167969,
       "step": 631600
     },
     {
-      "epoch": 14.548217739707102,
-      "grad_norm": 2.956925868988037,
-      "learning_rate": 8.616439551489943e-06,
-      "loss": 0.5463,
+      "epoch": 7.2741116312940655,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.608113610913164e-06,
+      "loss": 0.9812853240966797,
       "step": 631800
     },
     {
-      "epoch": 14.55282306346136,
-      "grad_norm": 2.96683669090271,
-      "learning_rate": 8.609167987667429e-06,
-      "loss": 0.5295,
+      "epoch": 7.276414316368781,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.600842047090647e-06,
+      "loss": 0.9823606109619141,
       "step": 632000
     },
     {
-      "epoch": 14.557428387215621,
-      "grad_norm": 3.31773042678833,
-      "learning_rate": 8.601896423844913e-06,
-      "loss": 0.5294,
+      "epoch": 7.278717001443495,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.593570483268132e-06,
+      "loss": 0.979590835571289,
       "step": 632200
     },
     {
-      "epoch": 14.562033710969882,
-      "grad_norm": 3.2458245754241943,
-      "learning_rate": 8.594624860022395e-06,
-      "loss": 0.5335,
+      "epoch": 7.281019686518211,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.586298919445617e-06,
+      "loss": 0.988263168334961,
       "step": 632400
     },
     {
-      "epoch": 14.56663903472414,
-      "grad_norm": 3.0680081844329834,
-      "learning_rate": 8.587353296199881e-06,
-      "loss": 0.547,
+      "epoch": 7.283322371592925,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.5790273556231e-06,
+      "loss": 0.979410400390625,
       "step": 632600
     },
     {
-      "epoch": 14.5712443584784,
-      "grad_norm": 3.1158411502838135,
-      "learning_rate": 8.580081732377365e-06,
-      "loss": 0.555,
+      "epoch": 7.285625056667641,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.571755791800585e-06,
+      "loss": 0.9735614776611328,
       "step": 632800
     },
     {
-      "epoch": 14.575849682232661,
-      "grad_norm": 4.015247821807861,
-      "learning_rate": 8.57281016855485e-06,
-      "loss": 0.5307,
+      "epoch": 7.287927741742355,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.56448422797807e-06,
+      "loss": 1.0004599761962891,
       "step": 633000
     },
     {
-      "epoch": 14.58045500598692,
-      "grad_norm": 3.309248924255371,
-      "learning_rate": 8.565538604732334e-06,
-      "loss": 0.5387,
+      "epoch": 7.290230426817071,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.557212664155554e-06,
+      "loss": 0.9967330932617188,
       "step": 633200
     },
     {
-      "epoch": 14.58506032974118,
-      "grad_norm": 3.6629464626312256,
-      "learning_rate": 8.558267040909818e-06,
-      "loss": 0.5384,
+      "epoch": 7.292533111891785,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.549941100333038e-06,
+      "loss": 0.9675569915771485,
       "step": 633400
     },
     {
-      "epoch": 14.589665653495441,
-      "grad_norm": 3.6036899089813232,
-      "learning_rate": 8.550995477087304e-06,
-      "loss": 0.5417,
+      "epoch": 7.2948357969665,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.542669536510522e-06,
+      "loss": 0.9942715454101563,
       "step": 633600
     },
     {
-      "epoch": 14.594270977249701,
-      "grad_norm": 2.716386556625366,
-      "learning_rate": 8.543723913264787e-06,
-      "loss": 0.5361,
+      "epoch": 7.297138482041215,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.535397972688008e-06,
+      "loss": 0.9726704406738281,
       "step": 633800
     },
     {
-      "epoch": 14.59887630100396,
-      "grad_norm": 2.9629690647125244,
-      "learning_rate": 8.53645234944227e-06,
-      "loss": 0.5406,
+      "epoch": 7.29944116711593,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.52812640886549e-06,
+      "loss": 0.9834327697753906,
       "step": 634000
     },
     {
-      "epoch": 14.60348162475822,
-      "grad_norm": 3.4902899265289307,
-      "learning_rate": 8.529180785619757e-06,
-      "loss": 0.5288,
+      "epoch": 7.301743852190645,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.520854845042975e-06,
+      "loss": 0.9955378723144531,
       "step": 634200
     },
     {
-      "epoch": 14.608086948512481,
-      "grad_norm": 3.307722568511963,
-      "learning_rate": 8.52190922179724e-06,
-      "loss": 0.54,
+      "epoch": 7.30404653726536,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.51358328122046e-06,
+      "loss": 0.9764264678955078,
       "step": 634400
     },
     {
-      "epoch": 14.61269227226674,
-      "grad_norm": 2.7474825382232666,
-      "learning_rate": 8.514637657974723e-06,
-      "loss": 0.5514,
+      "epoch": 7.306349222340075,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.506311717397943e-06,
+      "loss": 0.9716084289550782,
       "step": 634600
     },
     {
-      "epoch": 14.617297596021,
-      "grad_norm": 3.100318431854248,
-      "learning_rate": 8.50736609415221e-06,
-      "loss": 0.5373,
+      "epoch": 7.30865190741479,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.499040153575427e-06,
+      "loss": 0.9564635467529297,
       "step": 634800
     },
     {
-      "epoch": 14.62190291977526,
-      "grad_norm": 3.0019752979278564,
-      "learning_rate": 8.500094530329692e-06,
-      "loss": 0.5312,
+      "epoch": 7.310954592489505,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.491768589752913e-06,
+      "loss": 0.9855192565917968,
       "step": 635000
     },
     {
-      "epoch": 14.62650824352952,
-      "grad_norm": 3.607322931289673,
-      "learning_rate": 8.492822966507178e-06,
-      "loss": 0.5222,
+      "epoch": 7.31325727756422,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.484497025930396e-06,
+      "loss": 0.9694292449951172,
       "step": 635200
     },
     {
-      "epoch": 14.63111356728378,
-      "grad_norm": 2.8540308475494385,
-      "learning_rate": 8.485551402684662e-06,
-      "loss": 0.5392,
+      "epoch": 7.315559962638934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.477225462107882e-06,
+      "loss": 0.9754814910888672,
       "step": 635400
     },
     {
-      "epoch": 14.63571889103804,
-      "grad_norm": 2.7987473011016846,
-      "learning_rate": 8.478279838862145e-06,
-      "loss": 0.5433,
+      "epoch": 7.31786264771365,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.469953898285366e-06,
+      "loss": 0.995964126586914,
       "step": 635600
     },
     {
-      "epoch": 14.6403242147923,
-      "grad_norm": 3.4120616912841797,
-      "learning_rate": 8.471044632858741e-06,
-      "loss": 0.5306,
+      "epoch": 7.320165332788364,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.462682334462848e-06,
+      "loss": 0.9831593322753907,
       "step": 635800
     },
     {
-      "epoch": 14.64492953854656,
-      "grad_norm": 2.813546657562256,
-      "learning_rate": 8.463773069036227e-06,
-      "loss": 0.5319,
+      "epoch": 7.32246801786308,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.455410770640334e-06,
+      "loss": 0.994800796508789,
       "step": 636000
     },
     {
-      "epoch": 14.64953486230082,
-      "grad_norm": 3.185399293899536,
-      "learning_rate": 8.456501505213712e-06,
-      "loss": 0.5474,
+      "epoch": 7.324770702937794,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.448139206817819e-06,
+      "loss": 0.9779727172851562,
       "step": 636200
     },
     {
-      "epoch": 14.654140186055079,
-      "grad_norm": 3.0868401527404785,
-      "learning_rate": 8.449229941391194e-06,
-      "loss": 0.5404,
+      "epoch": 7.32707338801251,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.440867642995303e-06,
+      "loss": 0.9931973266601563,
       "step": 636400
     },
     {
-      "epoch": 14.65874550980934,
-      "grad_norm": 3.2732975482940674,
-      "learning_rate": 8.44195837756868e-06,
-      "loss": 0.536,
+      "epoch": 7.329376073087224,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.433596079172787e-06,
+      "loss": 0.9613636779785156,
       "step": 636600
     },
     {
-      "epoch": 14.6633508335636,
-      "grad_norm": 3.1363329887390137,
-      "learning_rate": 8.434686813746164e-06,
-      "loss": 0.5395,
+      "epoch": 7.33167875816194,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.426324515350271e-06,
+      "loss": 0.9925965118408203,
       "step": 636800
     },
     {
-      "epoch": 14.66795615731786,
-      "grad_norm": 3.2439472675323486,
-      "learning_rate": 8.42741524992365e-06,
-      "loss": 0.5405,
+      "epoch": 7.333981443236654,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.419052951527757e-06,
+      "loss": 0.978579330444336,
       "step": 637000
     },
     {
-      "epoch": 14.672561481072119,
-      "grad_norm": 2.4419994354248047,
-      "learning_rate": 8.420180043920247e-06,
-      "loss": 0.5228,
+      "epoch": 7.336284128311369,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.41178138770524e-06,
+      "loss": 0.972872314453125,
       "step": 637200
     },
     {
-      "epoch": 14.67716680482638,
-      "grad_norm": 4.0651044845581055,
-      "learning_rate": 8.41290848009773e-06,
-      "loss": 0.5441,
+      "epoch": 7.338586813386084,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.404509823882724e-06,
+      "loss": 0.9567903900146484,
       "step": 637400
     },
     {
-      "epoch": 14.68177212858064,
-      "grad_norm": 3.520622730255127,
-      "learning_rate": 8.405636916275214e-06,
-      "loss": 0.5334,
+      "epoch": 7.340889498460799,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.39723826006021e-06,
+      "loss": 0.9756443786621094,
       "step": 637600
     },
     {
-      "epoch": 14.686377452334899,
-      "grad_norm": 2.7496581077575684,
-      "learning_rate": 8.3983653524527e-06,
-      "loss": 0.5408,
+      "epoch": 7.343192183535514,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.389966696237692e-06,
+      "loss": 1.0001806640625,
       "step": 637800
     },
     {
-      "epoch": 14.69098277608916,
-      "grad_norm": 2.8176567554473877,
-      "learning_rate": 8.391093788630182e-06,
-      "loss": 0.5329,
+      "epoch": 7.345494868610229,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.382695132415178e-06,
+      "loss": 0.9706695556640625,
       "step": 638000
     },
     {
-      "epoch": 14.69558809984342,
-      "grad_norm": 3.117527484893799,
-      "learning_rate": 8.383822224807668e-06,
-      "loss": 0.5327,
+      "epoch": 7.347797553684944,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.375423568592662e-06,
+      "loss": 0.9860546112060546,
       "step": 638200
     },
     {
-      "epoch": 14.700193423597678,
-      "grad_norm": 3.2691383361816406,
-      "learning_rate": 8.376550660985152e-06,
-      "loss": 0.5375,
+      "epoch": 7.350100238759659,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.368152004770145e-06,
+      "loss": 0.9711141204833984,
       "step": 638400
     },
     {
-      "epoch": 14.704798747351939,
-      "grad_norm": 3.023637533187866,
-      "learning_rate": 8.369279097162635e-06,
-      "loss": 0.535,
+      "epoch": 7.352402923834374,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.360880440947631e-06,
+      "loss": 0.9637662506103516,
       "step": 638600
     },
     {
-      "epoch": 14.7094040711062,
-      "grad_norm": 4.642063140869141,
-      "learning_rate": 8.36200753334012e-06,
-      "loss": 0.5324,
+      "epoch": 7.3547056089090885,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.353608877125115e-06,
+      "loss": 1.0081533813476562,
       "step": 638800
     },
     {
-      "epoch": 14.714009394860458,
-      "grad_norm": 3.2362873554229736,
-      "learning_rate": 8.354735969517605e-06,
-      "loss": 0.5345,
+      "epoch": 7.357008293983803,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.346337313302598e-06,
+      "loss": 0.9800515747070313,
       "step": 639000
     },
     {
-      "epoch": 14.718614718614718,
-      "grad_norm": 3.0052409172058105,
-      "learning_rate": 8.347464405695088e-06,
-      "loss": 0.5431,
+      "epoch": 7.3593109790585185,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.339065749480084e-06,
+      "loss": 0.9915232849121094,
       "step": 639200
     },
     {
-      "epoch": 14.723220042368979,
-      "grad_norm": 2.905585765838623,
-      "learning_rate": 8.340229199691684e-06,
-      "loss": 0.5339,
+      "epoch": 7.361613664133233,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.331794185657568e-06,
+      "loss": 0.9679240417480469,
       "step": 639400
     },
     {
-      "epoch": 14.727825366123238,
-      "grad_norm": 3.1975319385528564,
-      "learning_rate": 8.33295763586917e-06,
-      "loss": 0.5351,
+      "epoch": 7.3639163492079485,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.324522621835054e-06,
+      "loss": 0.9997870635986328,
       "step": 639600
     },
     {
-      "epoch": 14.732430689877498,
-      "grad_norm": 3.070645809173584,
-      "learning_rate": 8.325686072046655e-06,
-      "loss": 0.5395,
+      "epoch": 7.366219034282663,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.317251058012536e-06,
+      "loss": 0.9761660766601562,
       "step": 639800
     },
     {
-      "epoch": 14.737036013631759,
-      "grad_norm": 4.0564351081848145,
-      "learning_rate": 8.318414508224139e-06,
-      "loss": 0.542,
+      "epoch": 7.368521719357378,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.30997949419002e-06,
+      "loss": 0.9725517272949219,
       "step": 640000
     },
     {
-      "epoch": 14.741641337386017,
-      "grad_norm": 3.2501471042633057,
-      "learning_rate": 8.311142944401623e-06,
-      "loss": 0.5381,
+      "epoch": 7.370824404432093,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.302707930367506e-06,
+      "loss": 0.9768034362792969,
       "step": 640200
     },
     {
-      "epoch": 14.746246661140278,
-      "grad_norm": 2.8298182487487793,
-      "learning_rate": 8.303871380579107e-06,
-      "loss": 0.5309,
+      "epoch": 7.373127089506808,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.295436366544989e-06,
+      "loss": 0.997316665649414,
       "step": 640400
     },
     {
-      "epoch": 14.750851984894538,
-      "grad_norm": 2.915964365005493,
-      "learning_rate": 8.296599816756593e-06,
-      "loss": 0.5458,
+      "epoch": 7.375429774581523,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.288164802722473e-06,
+      "loss": 0.9917243957519531,
       "step": 640600
     },
     {
-      "epoch": 14.755457308648799,
-      "grad_norm": 3.227104425430298,
-      "learning_rate": 8.289328252934076e-06,
-      "loss": 0.5375,
+      "epoch": 7.377732459656238,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.280893238899959e-06,
+      "loss": 0.9770442962646484,
       "step": 640800
     },
     {
-      "epoch": 14.760062632403057,
-      "grad_norm": 3.489304304122925,
-      "learning_rate": 8.28205668911156e-06,
-      "loss": 0.5386,
+      "epoch": 7.380035144730953,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.273621675077442e-06,
+      "loss": 0.9819647979736328,
       "step": 641000
     },
     {
-      "epoch": 14.764667956157318,
-      "grad_norm": 2.84167218208313,
-      "learning_rate": 8.274785125289046e-06,
-      "loss": 0.5288,
+      "epoch": 7.382337829805667,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.266350111254927e-06,
+      "loss": 0.9610335540771484,
       "step": 641200
     },
     {
-      "epoch": 14.769273279911578,
-      "grad_norm": 3.9559543132781982,
-      "learning_rate": 8.267513561466528e-06,
-      "loss": 0.5479,
+      "epoch": 7.384640514880383,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.259078547432412e-06,
+      "loss": 0.9789620971679688,
       "step": 641400
     },
     {
-      "epoch": 14.773878603665837,
-      "grad_norm": 3.0262818336486816,
-      "learning_rate": 8.260241997644014e-06,
-      "loss": 0.5233,
+      "epoch": 7.386943199955097,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.251806983609894e-06,
+      "loss": 0.9713826751708985,
       "step": 641600
     },
     {
-      "epoch": 14.778483927420098,
-      "grad_norm": 2.9113521575927734,
-      "learning_rate": 8.252970433821499e-06,
-      "loss": 0.545,
+      "epoch": 7.389245885029813,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.24453541978738e-06,
+      "loss": 0.9683750152587891,
       "step": 641800
     },
     {
-      "epoch": 14.783089251174358,
-      "grad_norm": 3.577045440673828,
-      "learning_rate": 8.245698869998981e-06,
-      "loss": 0.5376,
+      "epoch": 7.391548570104527,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.237263855964864e-06,
+      "loss": 0.9706134796142578,
       "step": 642000
     },
     {
-      "epoch": 14.787694574928617,
-      "grad_norm": 2.9530036449432373,
-      "learning_rate": 8.238427306176467e-06,
-      "loss": 0.5402,
+      "epoch": 7.393851255179243,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.229992292142347e-06,
+      "loss": 0.9962574768066407,
       "step": 642200
     },
     {
-      "epoch": 14.792299898682877,
-      "grad_norm": 2.98638916015625,
-      "learning_rate": 8.231155742353951e-06,
-      "loss": 0.5326,
+      "epoch": 7.396153940253957,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.222720728319833e-06,
+      "loss": 1.0055622100830077,
       "step": 642400
     },
     {
-      "epoch": 14.796905222437138,
-      "grad_norm": 3.20383620262146,
-      "learning_rate": 8.223884178531434e-06,
-      "loss": 0.527,
+      "epoch": 7.398456625328672,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.215449164497317e-06,
+      "loss": 0.973365478515625,
       "step": 642600
     },
     {
-      "epoch": 14.801510546191397,
-      "grad_norm": 3.5670669078826904,
-      "learning_rate": 8.21661261470892e-06,
-      "loss": 0.54,
+      "epoch": 7.400759310403387,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.208177600674803e-06,
+      "loss": 0.9849382019042969,
       "step": 642800
     },
     {
-      "epoch": 14.806115869945657,
-      "grad_norm": 3.084299087524414,
-      "learning_rate": 8.209377408705517e-06,
-      "loss": 0.5414,
+      "epoch": 7.403061995478102,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.200906036852285e-06,
+      "loss": 0.9878323364257813,
       "step": 643000
     },
     {
-      "epoch": 14.810721193699917,
-      "grad_norm": 3.610342264175415,
-      "learning_rate": 8.202105844883e-06,
-      "loss": 0.547,
+      "epoch": 7.405364680552817,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.19363447302977e-06,
+      "loss": 1.0094932556152343,
       "step": 643200
     },
     {
-      "epoch": 14.815326517454178,
-      "grad_norm": 3.762600898742676,
-      "learning_rate": 8.194834281060485e-06,
-      "loss": 0.5448,
+      "epoch": 7.407667365627532,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.186362909207256e-06,
+      "loss": 0.9762681579589844,
       "step": 643400
     },
     {
-      "epoch": 14.819931841208437,
-      "grad_norm": 2.416067600250244,
-      "learning_rate": 8.18756271723797e-06,
-      "loss": 0.5423,
+      "epoch": 7.409970050702247,
+      "grad_norm": 2.25,
+      "learning_rate": 8.179091345384738e-06,
+      "loss": 0.9903445434570313,
       "step": 643600
     },
     {
-      "epoch": 14.824537164962697,
-      "grad_norm": 3.5616204738616943,
-      "learning_rate": 8.180291153415453e-06,
-      "loss": 0.5233,
+      "epoch": 7.412272735776962,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.171819781562222e-06,
+      "loss": 0.9842520904541016,
       "step": 643800
     },
     {
-      "epoch": 14.829142488716958,
-      "grad_norm": 2.5570333003997803,
-      "learning_rate": 8.17301958959294e-06,
-      "loss": 0.537,
+      "epoch": 7.414575420851677,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.164548217739708e-06,
+      "loss": 1.0014985656738282,
       "step": 644000
     },
     {
-      "epoch": 14.833747812471216,
-      "grad_norm": 3.0077576637268066,
-      "learning_rate": 8.165784383589536e-06,
-      "loss": 0.546,
+      "epoch": 7.416878105926392,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.15727665391719e-06,
+      "loss": 0.98802490234375,
       "step": 644200
     },
     {
-      "epoch": 14.838353136225477,
-      "grad_norm": 2.393003463745117,
-      "learning_rate": 8.158512819767019e-06,
-      "loss": 0.5329,
+      "epoch": 7.419180791001107,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.150005090094677e-06,
+      "loss": 0.9854653930664062,
       "step": 644400
     },
     {
-      "epoch": 14.842958459979737,
-      "grad_norm": 3.575730085372925,
-      "learning_rate": 8.151241255944503e-06,
-      "loss": 0.5481,
+      "epoch": 7.421483476075822,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.14273352627216e-06,
+      "loss": 0.9909928131103516,
       "step": 644600
     },
     {
-      "epoch": 14.847563783733996,
-      "grad_norm": 3.534426689147949,
-      "learning_rate": 8.143969692121989e-06,
-      "loss": 0.5373,
+      "epoch": 7.423786161150536,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.135461962449643e-06,
+      "loss": 0.9801638031005859,
       "step": 644800
     },
     {
-      "epoch": 14.852169107488256,
-      "grad_norm": 2.990290880203247,
-      "learning_rate": 8.136698128299471e-06,
-      "loss": 0.5358,
+      "epoch": 7.426088846225252,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.12819039862713e-06,
+      "loss": 0.994657974243164,
       "step": 645000
     },
     {
-      "epoch": 14.856774431242517,
-      "grad_norm": 2.6674997806549072,
-      "learning_rate": 8.129426564476957e-06,
-      "loss": 0.545,
+      "epoch": 7.428391531299966,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.120918834804613e-06,
+      "loss": 0.9822552490234375,
       "step": 645200
     },
     {
-      "epoch": 14.861379754996776,
-      "grad_norm": 2.798583745956421,
-      "learning_rate": 8.122155000654442e-06,
-      "loss": 0.5366,
+      "epoch": 7.430694216374682,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.113647270982098e-06,
+      "loss": 0.965282211303711,
       "step": 645400
     },
     {
-      "epoch": 14.865985078751036,
-      "grad_norm": 2.496399164199829,
-      "learning_rate": 8.114883436831924e-06,
-      "loss": 0.5419,
+      "epoch": 7.432996901449396,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.106375707159582e-06,
+      "loss": 0.9767619323730469,
       "step": 645600
     },
     {
-      "epoch": 14.870590402505297,
-      "grad_norm": 2.99106502532959,
-      "learning_rate": 8.10761187300941e-06,
-      "loss": 0.5414,
+      "epoch": 7.435299586524112,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.099104143337066e-06,
+      "loss": 0.9821694946289062,
       "step": 645800
     },
     {
-      "epoch": 14.875195726259555,
-      "grad_norm": 3.370936393737793,
-      "learning_rate": 8.100340309186894e-06,
-      "loss": 0.5246,
+      "epoch": 7.437602271598826,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.09183257951455e-06,
+      "loss": 0.9880028533935546,
       "step": 646000
     },
     {
-      "epoch": 14.879801050013816,
-      "grad_norm": 2.571732521057129,
-      "learning_rate": 8.093068745364378e-06,
-      "loss": 0.5323,
+      "epoch": 7.439904956673541,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.084561015692035e-06,
+      "loss": 1.0000632476806641,
       "step": 646200
     },
     {
-      "epoch": 14.884406373768076,
-      "grad_norm": 4.192697525024414,
-      "learning_rate": 8.085797181541863e-06,
-      "loss": 0.5265,
+      "epoch": 7.442207641748256,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.077289451869519e-06,
+      "loss": 0.9520333862304687,
       "step": 646400
     },
     {
-      "epoch": 14.889011697522335,
-      "grad_norm": 3.2478342056274414,
-      "learning_rate": 8.078525617719347e-06,
-      "loss": 0.5266,
+      "epoch": 7.444510326822971,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.070017888047005e-06,
+      "loss": 0.962379150390625,
       "step": 646600
     },
     {
-      "epoch": 14.893617021276595,
-      "grad_norm": 3.7629292011260986,
-      "learning_rate": 8.071254053896831e-06,
-      "loss": 0.5455,
+      "epoch": 7.446813011897686,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.062746324224487e-06,
+      "loss": 0.9810494995117187,
       "step": 646800
     },
     {
-      "epoch": 14.898222345030856,
-      "grad_norm": 3.1777334213256836,
-      "learning_rate": 8.063982490074315e-06,
-      "loss": 0.5255,
+      "epoch": 7.449115696972401,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.055474760401973e-06,
+      "loss": 0.9787494659423828,
       "step": 647000
     },
     {
-      "epoch": 14.902827668785115,
-      "grad_norm": 3.3058884143829346,
-      "learning_rate": 8.0567109262518e-06,
-      "loss": 0.5519,
+      "epoch": 7.451418382047116,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.048203196579457e-06,
+      "loss": 0.9649417877197266,
       "step": 647200
     },
     {
-      "epoch": 14.907432992539375,
-      "grad_norm": 3.7439780235290527,
-      "learning_rate": 8.049439362429285e-06,
-      "loss": 0.545,
+      "epoch": 7.453721067121831,
+      "grad_norm": 1.0,
+      "learning_rate": 8.04093163275694e-06,
+      "loss": 0.9953223419189453,
       "step": 647400
     },
     {
-      "epoch": 14.912038316293636,
-      "grad_norm": 3.925713062286377,
-      "learning_rate": 8.042167798606768e-06,
-      "loss": 0.5236,
+      "epoch": 7.456023752196546,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.033660068934426e-06,
+      "loss": 0.9852490997314454,
       "step": 647600
     },
     {
-      "epoch": 14.916643640047896,
-      "grad_norm": 3.196169376373291,
-      "learning_rate": 8.034896234784254e-06,
-      "loss": 0.541,
+      "epoch": 7.4583264372712605,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.02638850511191e-06,
+      "loss": 0.9794223022460937,
       "step": 647800
     },
     {
-      "epoch": 14.921248963802155,
-      "grad_norm": 2.940185308456421,
-      "learning_rate": 8.027624670961738e-06,
-      "loss": 0.5409,
+      "epoch": 7.460629122345976,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.019116941289393e-06,
+      "loss": 0.9743696594238281,
       "step": 648000
     },
     {
-      "epoch": 14.925854287556415,
-      "grad_norm": 3.2344326972961426,
-      "learning_rate": 8.02035310713922e-06,
-      "loss": 0.5228,
+      "epoch": 7.4629318074206905,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.011845377466878e-06,
+      "loss": 0.9630597686767578,
       "step": 648200
     },
     {
-      "epoch": 14.930459611310676,
-      "grad_norm": 2.691751718521118,
-      "learning_rate": 8.013081543316707e-06,
-      "loss": 0.5243,
+      "epoch": 7.465234492495405,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.004573813644363e-06,
+      "loss": 0.9866986846923829,
       "step": 648400
     },
     {
-      "epoch": 14.935064935064934,
-      "grad_norm": 2.5507559776306152,
-      "learning_rate": 8.00580997949419e-06,
-      "loss": 0.5326,
+      "epoch": 7.4675371775701205,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.997302249821847e-06,
+      "loss": 0.996558837890625,
       "step": 648600
     },
     {
-      "epoch": 14.939670258819195,
-      "grad_norm": 3.069321870803833,
-      "learning_rate": 7.998538415671673e-06,
-      "loss": 0.5364,
+      "epoch": 7.469839862644835,
+      "grad_norm": 0.625,
+      "learning_rate": 7.990030685999331e-06,
+      "loss": 0.9859928894042969,
       "step": 648800
     },
     {
-      "epoch": 14.944275582573455,
-      "grad_norm": 2.938601016998291,
-      "learning_rate": 7.99126685184916e-06,
-      "loss": 0.5257,
+      "epoch": 7.47214254771955,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.982759122176815e-06,
+      "loss": 0.9754070281982422,
       "step": 649000
     },
     {
-      "epoch": 14.948880906327714,
-      "grad_norm": 2.886845588684082,
-      "learning_rate": 7.983995288026643e-06,
-      "loss": 0.5415,
+      "epoch": 7.474445232794265,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.9754875583543e-06,
+      "loss": 1.0027849578857422,
       "step": 649200
     },
     {
-      "epoch": 14.953486230081975,
-      "grad_norm": 2.869224786758423,
-      "learning_rate": 7.976723724204128e-06,
-      "loss": 0.5408,
+      "epoch": 7.47674791786898,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.968215994531784e-06,
+      "loss": 0.968758316040039,
       "step": 649400
     },
     {
-      "epoch": 14.958091553836235,
-      "grad_norm": 2.6893749237060547,
-      "learning_rate": 7.969452160381612e-06,
-      "loss": 0.5284,
+      "epoch": 7.479050602943695,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.960944430709268e-06,
+      "loss": 0.9768476867675782,
       "step": 649600
     },
     {
-      "epoch": 14.962696877590494,
-      "grad_norm": 3.181957244873047,
-      "learning_rate": 7.962180596559096e-06,
-      "loss": 0.5446,
+      "epoch": 7.48135328801841,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.953672866886754e-06,
+      "loss": 0.965670166015625,
       "step": 649800
     },
     {
-      "epoch": 14.967302201344754,
-      "grad_norm": 2.8158302307128906,
-      "learning_rate": 7.95490903273658e-06,
-      "loss": 0.5428,
+      "epoch": 7.483655973093125,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.946401303064236e-06,
+      "loss": 0.9942027282714844,
       "step": 650000
     },
     {
-      "epoch": 14.971907525099015,
-      "grad_norm": 3.031609296798706,
-      "learning_rate": 7.947673826733177e-06,
-      "loss": 0.5318,
+      "epoch": 7.4859586581678395,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.939129739241722e-06,
+      "loss": 0.9836433410644532,
       "step": 650200
     },
     {
-      "epoch": 14.976512848853275,
-      "grad_norm": 2.578882932662964,
-      "learning_rate": 7.940402262910661e-06,
-      "loss": 0.5252,
+      "epoch": 7.488261343242555,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.931858175419207e-06,
+      "loss": 0.9702881622314453,
       "step": 650400
     },
     {
-      "epoch": 14.981118172607534,
-      "grad_norm": 3.891084671020508,
-      "learning_rate": 7.933130699088146e-06,
-      "loss": 0.5412,
+      "epoch": 7.490564028317269,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.924586611596689e-06,
+      "loss": 0.9784402465820312,
       "step": 650600
     },
     {
-      "epoch": 14.985723496361794,
-      "grad_norm": 4.118905544281006,
-      "learning_rate": 7.925859135265632e-06,
-      "loss": 0.5389,
+      "epoch": 7.492866713391985,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.917315047774175e-06,
+      "loss": 1.0029482269287109,
       "step": 650800
     },
     {
-      "epoch": 14.990328820116055,
-      "grad_norm": 3.9632790088653564,
-      "learning_rate": 7.918587571443114e-06,
-      "loss": 0.527,
+      "epoch": 7.495169398466699,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.91004348395166e-06,
+      "loss": 0.9745333099365234,
       "step": 651000
     },
     {
-      "epoch": 14.994934143870314,
-      "grad_norm": 3.387319564819336,
-      "learning_rate": 7.9113160076206e-06,
-      "loss": 0.5273,
+      "epoch": 7.497472083541415,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.902771920129143e-06,
+      "loss": 0.9536399841308594,
       "step": 651200
     },
     {
-      "epoch": 14.999539467624574,
-      "grad_norm": 2.79263973236084,
-      "learning_rate": 7.904044443798084e-06,
-      "loss": 0.539,
+      "epoch": 7.499774768616129,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.895500356306628e-06,
+      "loss": 0.9742577362060547,
       "step": 651400
     },
     {
-      "epoch": 15.0,
-      "eval_loss": 0.5169693231582642,
-      "eval_runtime": 168.9237,
-      "eval_samples_per_second": 167.892,
-      "eval_steps_per_second": 10.496,
-      "step": 651420
-    },
-    {
-      "epoch": 15.004144791378835,
-      "grad_norm": 3.222738742828369,
-      "learning_rate": 7.896772879975567e-06,
-      "loss": 0.5368,
+      "epoch": 7.502077453690845,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.888228792484112e-06,
+      "loss": 0.9795113372802734,
       "step": 651600
     },
     {
-      "epoch": 15.008750115133093,
-      "grad_norm": 3.698350191116333,
-      "learning_rate": 7.889501316153053e-06,
-      "loss": 0.5365,
+      "epoch": 7.504380138765559,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.880957228661596e-06,
+      "loss": 1.003966293334961,
       "step": 651800
     },
     {
-      "epoch": 15.013355438887354,
-      "grad_norm": 3.877946376800537,
-      "learning_rate": 7.882229752330537e-06,
-      "loss": 0.5331,
+      "epoch": 7.506682823840274,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.87368566483908e-06,
+      "loss": 0.97415283203125,
       "step": 652000
     },
     {
-      "epoch": 15.017960762641614,
-      "grad_norm": 3.619192600250244,
-      "learning_rate": 7.87495818850802e-06,
-      "loss": 0.5337,
+      "epoch": 7.508985508914989,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.866414101016565e-06,
+      "loss": 0.9731611633300781,
       "step": 652200
     },
     {
-      "epoch": 15.022566086395873,
-      "grad_norm": 2.9322800636291504,
-      "learning_rate": 7.867686624685505e-06,
-      "loss": 0.5217,
+      "epoch": 7.511288193989704,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.859142537194049e-06,
+      "loss": 0.983914566040039,
       "step": 652400
     },
     {
-      "epoch": 15.027171410150133,
-      "grad_norm": 2.9945201873779297,
-      "learning_rate": 7.86041506086299e-06,
-      "loss": 0.5297,
+      "epoch": 7.513590879064419,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.851870973371533e-06,
+      "loss": 0.9763735961914063,
       "step": 652600
     },
     {
-      "epoch": 15.031776733904394,
-      "grad_norm": 2.359719753265381,
-      "learning_rate": 7.853143497040474e-06,
-      "loss": 0.5316,
+      "epoch": 7.515893564139134,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.844599409549019e-06,
+      "loss": 0.969703140258789,
       "step": 652800
     },
     {
-      "epoch": 15.036382057658653,
-      "grad_norm": 3.135897636413574,
-      "learning_rate": 7.845871933217958e-06,
-      "loss": 0.534,
+      "epoch": 7.518196249213849,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.837327845726503e-06,
+      "loss": 0.9964493560791016,
       "step": 653000
     },
     {
-      "epoch": 15.040987381412913,
-      "grad_norm": 3.2203567028045654,
-      "learning_rate": 7.838600369395442e-06,
-      "loss": 0.5371,
+      "epoch": 7.520498934288564,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.830056281903986e-06,
+      "loss": 1.0019598388671875,
       "step": 653200
     },
     {
-      "epoch": 15.045592705167174,
-      "grad_norm": 3.355985164642334,
-      "learning_rate": 7.831328805572926e-06,
-      "loss": 0.5294,
+      "epoch": 7.522801619363279,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.822784718081472e-06,
+      "loss": 0.9844602966308593,
       "step": 653400
     },
     {
-      "epoch": 15.050198028921432,
-      "grad_norm": 3.7770304679870605,
-      "learning_rate": 7.82405724175041e-06,
-      "loss": 0.5409,
+      "epoch": 7.525104304437994,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.815513154258956e-06,
+      "loss": 0.9833013153076172,
       "step": 653600
     },
     {
-      "epoch": 15.054803352675693,
-      "grad_norm": 9.291865348815918,
-      "learning_rate": 7.816785677927895e-06,
-      "loss": 0.5363,
+      "epoch": 7.527406989512709,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.808241590436438e-06,
+      "loss": 0.9813348388671875,
       "step": 653800
     },
     {
-      "epoch": 15.059408676429953,
-      "grad_norm": 2.6774072647094727,
-      "learning_rate": 7.80951411410538e-06,
-      "loss": 0.5297,
+      "epoch": 7.529709674587424,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.800970026613924e-06,
+      "loss": 0.9650066375732422,
       "step": 654000
     },
     {
-      "epoch": 15.064014000184214,
-      "grad_norm": 3.5901434421539307,
-      "learning_rate": 7.802242550282863e-06,
-      "loss": 0.527,
+      "epoch": 7.532012359662138,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.793698462791408e-06,
+      "loss": 0.9807217407226563,
       "step": 654200
     },
     {
-      "epoch": 15.068619323938472,
-      "grad_norm": 3.092885971069336,
-      "learning_rate": 7.79497098646035e-06,
-      "loss": 0.5316,
+      "epoch": 7.534315044736854,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.786426898968893e-06,
+      "loss": 0.9870914459228516,
       "step": 654400
     },
     {
-      "epoch": 15.073224647692733,
-      "grad_norm": 2.6505744457244873,
-      "learning_rate": 7.787735780456946e-06,
-      "loss": 0.523,
+      "epoch": 7.536617729811568,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.779155335146377e-06,
+      "loss": 0.9670613098144532,
       "step": 654600
     },
     {
-      "epoch": 15.077829971446993,
-      "grad_norm": 2.4430651664733887,
-      "learning_rate": 7.78046421663443e-06,
-      "loss": 0.5332,
+      "epoch": 7.538920414886284,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.771883771323861e-06,
+      "loss": 0.9968564605712891,
       "step": 654800
     },
     {
-      "epoch": 15.082435295201252,
-      "grad_norm": 3.4239096641540527,
-      "learning_rate": 7.773192652811913e-06,
-      "loss": 0.5329,
+      "epoch": 7.541223099960998,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.764612207501345e-06,
+      "loss": 1.0076090240478515,
       "step": 655000
     },
     {
-      "epoch": 15.087040618955513,
-      "grad_norm": 3.358081340789795,
-      "learning_rate": 7.765921088989399e-06,
-      "loss": 0.5242,
+      "epoch": 7.543525785035714,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.75734064367883e-06,
+      "loss": 0.987034912109375,
       "step": 655200
     },
     {
-      "epoch": 15.091645942709773,
-      "grad_norm": 3.774402141571045,
-      "learning_rate": 7.758685882985996e-06,
-      "loss": 0.5316,
+      "epoch": 7.545828470110428,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.750069079856314e-06,
+      "loss": 0.9792374420166016,
       "step": 655400
     },
     {
-      "epoch": 15.096251266464032,
-      "grad_norm": 3.171030282974243,
-      "learning_rate": 7.75141431916348e-06,
-      "loss": 0.545,
+      "epoch": 7.548131155185143,
+      "grad_norm": 0.625,
+      "learning_rate": 7.742797516033798e-06,
+      "loss": 1.0022062683105468,
       "step": 655600
     },
     {
-      "epoch": 15.100856590218292,
-      "grad_norm": 2.9989264011383057,
-      "learning_rate": 7.744142755340964e-06,
-      "loss": 0.5374,
+      "epoch": 7.550433840259858,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.735525952211282e-06,
+      "loss": 0.9784234619140625,
       "step": 655800
     },
     {
-      "epoch": 15.105461913972553,
-      "grad_norm": 2.5736443996429443,
-      "learning_rate": 7.736871191518448e-06,
-      "loss": 0.5254,
+      "epoch": 7.552736525334573,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.728254388388768e-06,
+      "loss": 0.9708540344238281,
       "step": 656000
     },
     {
-      "epoch": 15.110067237726811,
-      "grad_norm": 3.9181089401245117,
-      "learning_rate": 7.729599627695933e-06,
-      "loss": 0.5276,
+      "epoch": 7.555039210409288,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.72098282456625e-06,
+      "loss": 0.9731727600097656,
       "step": 656200
     },
     {
-      "epoch": 15.114672561481072,
-      "grad_norm": 2.8905580043792725,
-      "learning_rate": 7.722328063873417e-06,
-      "loss": 0.5338,
+      "epoch": 7.557341895484003,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.713711260743735e-06,
+      "loss": 0.9630989074707031,
       "step": 656400
     },
     {
-      "epoch": 15.119277885235332,
-      "grad_norm": 3.0359580516815186,
-      "learning_rate": 7.715056500050901e-06,
-      "loss": 0.5302,
+      "epoch": 7.559644580558718,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.70643969692122e-06,
+      "loss": 0.9803575134277344,
       "step": 656600
     },
     {
-      "epoch": 15.123883208989591,
-      "grad_norm": 3.0906269550323486,
-      "learning_rate": 7.707784936228385e-06,
-      "loss": 0.5233,
+      "epoch": 7.561947265633433,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.699168133098705e-06,
+      "loss": 0.9948880004882813,
       "step": 656800
     },
     {
-      "epoch": 15.128488532743852,
-      "grad_norm": 2.517125129699707,
-      "learning_rate": 7.70051337240587e-06,
-      "loss": 0.5382,
+      "epoch": 7.564249950708148,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.691896569276188e-06,
+      "loss": 0.9783583068847657,
       "step": 657000
     },
     {
-      "epoch": 15.133093856498112,
-      "grad_norm": 3.945773124694824,
-      "learning_rate": 7.693241808583354e-06,
-      "loss": 0.5329,
+      "epoch": 7.5665526357828625,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.684625005453673e-06,
+      "loss": 0.9461773681640625,
       "step": 657200
     },
     {
-      "epoch": 15.137699180252373,
-      "grad_norm": 3.2435407638549805,
-      "learning_rate": 7.68597024476084e-06,
-      "loss": 0.5289,
+      "epoch": 7.568855320857578,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.677353441631158e-06,
+      "loss": 0.981177978515625,
       "step": 657400
     },
     {
-      "epoch": 15.142304504006631,
-      "grad_norm": 2.7934975624084473,
-      "learning_rate": 7.678698680938324e-06,
-      "loss": 0.5373,
+      "epoch": 7.5711580059322925,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.670081877808642e-06,
+      "loss": 0.9690900421142579,
       "step": 657600
     },
     {
-      "epoch": 15.146909827760892,
-      "grad_norm": 2.8237783908843994,
-      "learning_rate": 7.671427117115806e-06,
-      "loss": 0.5276,
+      "epoch": 7.573460691007007,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.662810313986126e-06,
+      "loss": 0.9845625305175781,
       "step": 657800
     },
     {
-      "epoch": 15.151515151515152,
-      "grad_norm": 4.111272811889648,
-      "learning_rate": 7.664155553293292e-06,
-      "loss": 0.5264,
+      "epoch": 7.5757633760817225,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.65553875016361e-06,
+      "loss": 0.9893363952636719,
       "step": 658000
     },
     {
-      "epoch": 15.156120475269411,
-      "grad_norm": 3.4282681941986084,
-      "learning_rate": 7.656883989470776e-06,
-      "loss": 0.5343,
+      "epoch": 7.578066061156437,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.648267186341095e-06,
+      "loss": 0.9911522674560547,
       "step": 658200
     },
     {
-      "epoch": 15.160725799023671,
-      "grad_norm": 2.7267274856567383,
-      "learning_rate": 7.649612425648259e-06,
-      "loss": 0.5311,
+      "epoch": 7.580368746231152,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.640995622518579e-06,
+      "loss": 0.9923857116699218,
       "step": 658400
     },
     {
-      "epoch": 15.165331122777932,
-      "grad_norm": 3.3111186027526855,
-      "learning_rate": 7.642340861825745e-06,
-      "loss": 0.5329,
+      "epoch": 7.582671431305867,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.633724058696065e-06,
+      "loss": 0.9609416961669922,
       "step": 658600
     },
     {
-      "epoch": 15.16993644653219,
-      "grad_norm": 2.9832489490509033,
-      "learning_rate": 7.635069298003229e-06,
-      "loss": 0.5277,
+      "epoch": 7.584974116380582,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.626452494873547e-06,
+      "loss": 0.9927304077148438,
       "step": 658800
     },
     {
-      "epoch": 15.174541770286451,
-      "grad_norm": 3.5788393020629883,
-      "learning_rate": 7.6277977341807125e-06,
-      "loss": 0.5254,
+      "epoch": 7.587276801455297,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.619180931051032e-06,
+      "loss": 0.9676020812988281,
       "step": 659000
     },
     {
-      "epoch": 15.179147094040712,
-      "grad_norm": 2.7909393310546875,
-      "learning_rate": 7.6205261703581975e-06,
-      "loss": 0.5332,
+      "epoch": 7.5895794865300115,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.6119093672285165e-06,
+      "loss": 0.991962890625,
       "step": 659200
     },
     {
-      "epoch": 15.18375241779497,
-      "grad_norm": 3.0156285762786865,
-      "learning_rate": 7.613254606535683e-06,
-      "loss": 0.5396,
+      "epoch": 7.591882171604727,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.604637803406e-06,
+      "loss": 0.98235107421875,
       "step": 659400
     },
     {
-      "epoch": 15.18835774154923,
-      "grad_norm": 3.3758177757263184,
-      "learning_rate": 7.605983042713165e-06,
-      "loss": 0.5257,
+      "epoch": 7.594184856679441,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.597366239583485e-06,
+      "loss": 0.9749604034423828,
       "step": 659600
     },
     {
-      "epoch": 15.192963065303491,
-      "grad_norm": 3.161837100982666,
-      "learning_rate": 7.59871147889065e-06,
-      "loss": 0.5314,
+      "epoch": 7.596487541754157,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.59009467576097e-06,
+      "loss": 0.9732862091064454,
       "step": 659800
     },
     {
-      "epoch": 15.19756838905775,
-      "grad_norm": 4.073904037475586,
-      "learning_rate": 7.591476272887247e-06,
-      "loss": 0.5275,
+      "epoch": 7.598790226828871,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.582823111938454e-06,
+      "loss": 0.9805200958251953,
       "step": 660000
     },
     {
-      "epoch": 15.20217371281201,
-      "grad_norm": 3.1027348041534424,
-      "learning_rate": 7.584204709064732e-06,
-      "loss": 0.5367,
+      "epoch": 7.601092911903587,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.5755515481159376e-06,
+      "loss": 0.9868666839599609,
       "step": 660200
     },
     {
-      "epoch": 15.206779036566271,
-      "grad_norm": 2.5127763748168945,
-      "learning_rate": 7.5769331452422155e-06,
-      "loss": 0.5321,
+      "epoch": 7.603395596978301,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.568279984293423e-06,
+      "loss": 1.0004676055908204,
       "step": 660400
     },
     {
-      "epoch": 15.21138436032053,
-      "grad_norm": 3.317880153656006,
-      "learning_rate": 7.5696615814197e-06,
-      "loss": 0.535,
+      "epoch": 7.605698282053017,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.561008420470908e-06,
+      "loss": 0.9936216735839843,
       "step": 660600
     },
     {
-      "epoch": 15.21598968407479,
-      "grad_norm": 3.18585467338562,
-      "learning_rate": 7.562390017597185e-06,
-      "loss": 0.5364,
+      "epoch": 7.608000967127731,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.55373685664839e-06,
+      "loss": 0.9566283416748047,
       "step": 660800
     },
     {
-      "epoch": 15.22059500782905,
-      "grad_norm": 3.1549441814422607,
-      "learning_rate": 7.555118453774668e-06,
-      "loss": 0.5288,
+      "epoch": 7.610303652202447,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.546465292825875e-06,
+      "loss": 0.9897992706298828,
       "step": 661000
     },
     {
-      "epoch": 15.225200331583311,
-      "grad_norm": 3.0933032035827637,
-      "learning_rate": 7.547846889952153e-06,
-      "loss": 0.5315,
+      "epoch": 7.612606337277161,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.53919372900336e-06,
+      "loss": 0.9802947998046875,
       "step": 661200
     },
     {
-      "epoch": 15.22980565533757,
-      "grad_norm": 3.4483559131622314,
-      "learning_rate": 7.5405753261296375e-06,
-      "loss": 0.5372,
+      "epoch": 7.614909022351876,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.531922165180844e-06,
+      "loss": 0.9826642608642578,
       "step": 661400
     },
     {
-      "epoch": 15.23441097909183,
-      "grad_norm": 3.425278663635254,
-      "learning_rate": 7.5333037623071225e-06,
-      "loss": 0.5268,
+      "epoch": 7.617211707426591,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.524650601358328e-06,
+      "loss": 0.9850830841064453,
       "step": 661600
     },
     {
-      "epoch": 15.23901630284609,
-      "grad_norm": 2.674301862716675,
-      "learning_rate": 7.526032198484606e-06,
-      "loss": 0.523,
+      "epoch": 7.619514392501306,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.517379037535813e-06,
+      "loss": 0.9891217803955078,
       "step": 661800
     },
     {
-      "epoch": 15.24362162660035,
-      "grad_norm": 4.0348358154296875,
-      "learning_rate": 7.518760634662091e-06,
-      "loss": 0.5256,
+      "epoch": 7.621817077576021,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.510107473713296e-06,
+      "loss": 0.996153335571289,
       "step": 662000
     },
     {
-      "epoch": 15.24822695035461,
-      "grad_norm": 2.689831018447876,
-      "learning_rate": 7.511489070839575e-06,
-      "loss": 0.5198,
+      "epoch": 7.624119762650736,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.5028359098907815e-06,
+      "loss": 0.9904628753662109,
       "step": 662200
     },
     {
-      "epoch": 15.25283227410887,
-      "grad_norm": 3.238309860229492,
-      "learning_rate": 7.5042175070170586e-06,
-      "loss": 0.5295,
+      "epoch": 7.626422447725451,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.495564346068266e-06,
+      "loss": 0.9812112426757813,
       "step": 662400
     },
     {
-      "epoch": 15.25743759786313,
-      "grad_norm": 2.907236337661743,
-      "learning_rate": 7.496945943194544e-06,
-      "loss": 0.5316,
+      "epoch": 7.628725132800166,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.48829278224575e-06,
+      "loss": 0.9918553161621094,
       "step": 662600
     },
     {
-      "epoch": 15.26204292161739,
-      "grad_norm": 2.706756114959717,
-      "learning_rate": 7.4897107371911405e-06,
-      "loss": 0.522,
+      "epoch": 7.63102781787488,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.481021218423234e-06,
+      "loss": 0.9850691986083985,
       "step": 662800
     },
     {
-      "epoch": 15.26664824537165,
-      "grad_norm": 3.2358510494232178,
-      "learning_rate": 7.482439173368625e-06,
-      "loss": 0.5382,
+      "epoch": 7.633330502949596,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.473749654600719e-06,
+      "loss": 0.9833558654785156,
       "step": 663000
     },
     {
-      "epoch": 15.271253569125909,
-      "grad_norm": 2.505164861679077,
-      "learning_rate": 7.475167609546109e-06,
-      "loss": 0.5304,
+      "epoch": 7.63563318802431,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.4664780907782026e-06,
+      "loss": 0.980711669921875,
       "step": 663200
     },
     {
-      "epoch": 15.27585889288017,
-      "grad_norm": 2.7011849880218506,
-      "learning_rate": 7.467896045723593e-06,
-      "loss": 0.5381,
+      "epoch": 7.637935873099026,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.459206526955687e-06,
+      "loss": 0.9675973510742187,
       "step": 663400
     },
     {
-      "epoch": 15.28046421663443,
-      "grad_norm": 4.1736159324646,
-      "learning_rate": 7.460624481901077e-06,
-      "loss": 0.53,
+      "epoch": 7.64023855817374,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.451934963133172e-06,
+      "loss": 0.9979064178466797,
       "step": 663600
     },
     {
-      "epoch": 15.285069540388688,
-      "grad_norm": 2.8597121238708496,
-      "learning_rate": 7.4533529180785625e-06,
-      "loss": 0.5273,
+      "epoch": 7.642541243248456,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.444663399310656e-06,
+      "loss": 0.996081314086914,
       "step": 663800
     },
     {
-      "epoch": 15.289674864142949,
-      "grad_norm": 4.301202297210693,
-      "learning_rate": 7.446081354256047e-06,
-      "loss": 0.5306,
+      "epoch": 7.64484392832317,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.43739183548814e-06,
+      "loss": 0.9812240600585938,
       "step": 664000
     },
     {
-      "epoch": 15.29428018789721,
-      "grad_norm": 3.257692575454712,
-      "learning_rate": 7.438809790433531e-06,
-      "loss": 0.5249,
+      "epoch": 7.647146613397886,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.4301202716656245e-06,
+      "loss": 0.986427993774414,
       "step": 664200
     },
     {
-      "epoch": 15.29888551165147,
-      "grad_norm": 4.112732887268066,
-      "learning_rate": 7.431538226611015e-06,
-      "loss": 0.5303,
+      "epoch": 7.6494492984726,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.422848707843109e-06,
+      "loss": 0.9837487030029297,
       "step": 664400
     },
     {
-      "epoch": 15.303490835405729,
-      "grad_norm": 3.026270627975464,
-      "learning_rate": 7.424266662788499e-06,
-      "loss": 0.5309,
+      "epoch": 7.651751983547316,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.415577144020594e-06,
+      "loss": 0.9797900390625,
       "step": 664600
     },
     {
-      "epoch": 15.30809615915999,
-      "grad_norm": 3.302058219909668,
-      "learning_rate": 7.417031456785096e-06,
-      "loss": 0.5272,
+      "epoch": 7.65405466862203,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.408305580198077e-06,
+      "loss": 1.005549087524414,
       "step": 664800
     },
     {
-      "epoch": 15.31270148291425,
-      "grad_norm": 3.1343607902526855,
-      "learning_rate": 7.409759892962581e-06,
-      "loss": 0.543,
+      "epoch": 7.656357353696745,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.401034016375561e-06,
+      "loss": 0.9849571228027344,
       "step": 665000
     },
     {
-      "epoch": 15.317306806668508,
-      "grad_norm": 3.4846322536468506,
-      "learning_rate": 7.402488329140065e-06,
-      "loss": 0.5345,
+      "epoch": 7.65866003877146,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.3937624525530464e-06,
+      "loss": 0.96781494140625,
       "step": 665200
     },
     {
-      "epoch": 15.321912130422769,
-      "grad_norm": 2.883300304412842,
-      "learning_rate": 7.395216765317549e-06,
-      "loss": 0.5179,
+      "epoch": 7.660962723846175,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.386490888730531e-06,
+      "loss": 1.0096734619140626,
       "step": 665400
     },
     {
-      "epoch": 15.32651745417703,
-      "grad_norm": 3.417581081390381,
-      "learning_rate": 7.387945201495034e-06,
-      "loss": 0.5259,
+      "epoch": 7.66326540892089,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.379219324908016e-06,
+      "loss": 0.9794087219238281,
       "step": 665600
     },
     {
-      "epoch": 15.331122777931288,
-      "grad_norm": 2.906959056854248,
-      "learning_rate": 7.380673637672518e-06,
-      "loss": 0.5253,
+      "epoch": 7.665568093995605,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.371947761085499e-06,
+      "loss": 0.992833251953125,
       "step": 665800
     },
     {
-      "epoch": 15.335728101685548,
-      "grad_norm": 2.635720729827881,
-      "learning_rate": 7.3734020738500024e-06,
-      "loss": 0.5321,
+      "epoch": 7.66787077907032,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.364676197262983e-06,
+      "loss": 0.9783699035644531,
       "step": 666000
     },
     {
-      "epoch": 15.340333425439809,
-      "grad_norm": 2.8805277347564697,
-      "learning_rate": 7.366130510027487e-06,
-      "loss": 0.5336,
+      "epoch": 7.6701734641450345,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.357404633440468e-06,
+      "loss": 1.0216586303710937,
       "step": 666200
     },
     {
-      "epoch": 15.344938749194068,
-      "grad_norm": 2.8946115970611572,
-      "learning_rate": 7.358858946204971e-06,
-      "loss": 0.5299,
+      "epoch": 7.672476149219749,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.350133069617953e-06,
+      "loss": 0.9817404174804687,
       "step": 666400
     },
     {
-      "epoch": 15.349544072948328,
-      "grad_norm": 3.4212236404418945,
-      "learning_rate": 7.351587382382456e-06,
-      "loss": 0.5307,
+      "epoch": 7.6747788342944645,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.342861505795436e-06,
+      "loss": 0.9726050567626953,
       "step": 666600
     },
     {
-      "epoch": 15.354149396702589,
-      "grad_norm": 3.027736186981201,
-      "learning_rate": 7.344315818559939e-06,
-      "loss": 0.5367,
+      "epoch": 7.677081519369179,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.335589941972921e-06,
+      "loss": 1.0005302429199219,
       "step": 666800
     },
     {
-      "epoch": 15.358754720456847,
-      "grad_norm": 2.96783185005188,
-      "learning_rate": 7.3370442547374235e-06,
-      "loss": 0.5288,
+      "epoch": 7.6793842044438945,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.328318378150405e-06,
+      "loss": 0.976294937133789,
       "step": 667000
     },
     {
-      "epoch": 15.363360044211108,
-      "grad_norm": 3.2669103145599365,
-      "learning_rate": 7.329772690914909e-06,
-      "loss": 0.5318,
+      "epoch": 7.681686889518609,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.3210468143278895e-06,
+      "loss": 0.9913665771484375,
       "step": 667200
     },
     {
-      "epoch": 15.367965367965368,
-      "grad_norm": 3.050448417663574,
-      "learning_rate": 7.322501127092393e-06,
-      "loss": 0.534,
+      "epoch": 7.683989574593324,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.313775250505374e-06,
+      "loss": 0.9836793518066407,
       "step": 667400
     },
     {
-      "epoch": 15.372570691719627,
-      "grad_norm": 4.087128162384033,
-      "learning_rate": 7.315229563269877e-06,
-      "loss": 0.5263,
+      "epoch": 7.686292259668039,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.306503686682858e-06,
+      "loss": 0.9918992614746094,
       "step": 667600
     },
     {
-      "epoch": 15.377176015473887,
-      "grad_norm": 3.5883004665374756,
-      "learning_rate": 7.307957999447361e-06,
-      "loss": 0.5277,
+      "epoch": 7.688594944742754,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.299232122860343e-06,
+      "loss": 0.9827339935302735,
       "step": 667800
     },
     {
-      "epoch": 15.381781339228148,
-      "grad_norm": 3.088322162628174,
-      "learning_rate": 7.3006864356248455e-06,
-      "loss": 0.5414,
+      "epoch": 7.690897629817469,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.291960559037827e-06,
+      "loss": 0.979901351928711,
       "step": 668000
     },
     {
-      "epoch": 15.386386662982408,
-      "grad_norm": 3.5304698944091797,
-      "learning_rate": 7.2934148718023305e-06,
-      "loss": 0.5338,
+      "epoch": 7.693200314892184,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.284688995215311e-06,
+      "loss": 0.992957992553711,
       "step": 668200
     },
     {
-      "epoch": 15.390991986736667,
-      "grad_norm": 3.154963731765747,
-      "learning_rate": 7.286143307979814e-06,
-      "loss": 0.5417,
+      "epoch": 7.695502999966899,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.277417431392796e-06,
+      "loss": 0.981059799194336,
       "step": 668400
     },
     {
-      "epoch": 15.395597310490928,
-      "grad_norm": 3.262460231781006,
-      "learning_rate": 7.278871744157298e-06,
-      "loss": 0.5163,
+      "epoch": 7.697805685041613,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.27014586757028e-06,
+      "loss": 0.9890800476074219,
       "step": 668600
     },
     {
-      "epoch": 15.400202634245188,
-      "grad_norm": 3.046994209289551,
-      "learning_rate": 7.271600180334783e-06,
-      "loss": 0.5291,
+      "epoch": 7.700108370116329,
+      "grad_norm": 0.625,
+      "learning_rate": 7.262874303747764e-06,
+      "loss": 0.9820852661132813,
       "step": 668800
     },
     {
-      "epoch": 15.404807957999447,
-      "grad_norm": 3.0019094944000244,
-      "learning_rate": 7.264328616512267e-06,
-      "loss": 0.5308,
+      "epoch": 7.702411055191043,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.255602739925248e-06,
+      "loss": 0.9690668487548828,
       "step": 669000
     },
     {
-      "epoch": 15.409413281753707,
-      "grad_norm": 2.577397584915161,
-      "learning_rate": 7.257057052689752e-06,
-      "loss": 0.5328,
+      "epoch": 7.704713740265759,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.2483311761027325e-06,
+      "loss": 0.9957647705078125,
       "step": 669200
     },
     {
-      "epoch": 15.414018605507968,
-      "grad_norm": 3.689873218536377,
-      "learning_rate": 7.249785488867236e-06,
-      "loss": 0.5362,
+      "epoch": 7.707016425340473,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.241059612280218e-06,
+      "loss": 1.0044274139404297,
       "step": 669400
     },
     {
-      "epoch": 15.418623929262226,
-      "grad_norm": 3.1702044010162354,
-      "learning_rate": 7.24251392504472e-06,
-      "loss": 0.5278,
+      "epoch": 7.709319110415189,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.233788048457702e-06,
+      "loss": 1.0007498168945312,
       "step": 669600
     },
     {
-      "epoch": 15.423229253016487,
-      "grad_norm": 2.942192316055298,
-      "learning_rate": 7.235278719041317e-06,
-      "loss": 0.5438,
+      "epoch": 7.711621795489903,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.226516484635185e-06,
+      "loss": 0.98791748046875,
       "step": 669800
     },
     {
-      "epoch": 15.427834576770747,
-      "grad_norm": 3.0847814083099365,
-      "learning_rate": 7.228007155218802e-06,
-      "loss": 0.5425,
+      "epoch": 7.713924480564619,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.21924492081267e-06,
+      "loss": 0.9789481353759766,
       "step": 670000
     },
     {
-      "epoch": 15.432439900525006,
-      "grad_norm": 2.7735490798950195,
-      "learning_rate": 7.220735591396285e-06,
-      "loss": 0.5362,
+      "epoch": 7.716227165639333,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.2119733569901545e-06,
+      "loss": 0.978783187866211,
       "step": 670200
     },
     {
-      "epoch": 15.437045224279267,
-      "grad_norm": 3.166234254837036,
-      "learning_rate": 7.21346402757377e-06,
-      "loss": 0.528,
+      "epoch": 7.718529850714049,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.204701793167639e-06,
+      "loss": 0.9658313751220703,
       "step": 670400
     },
     {
-      "epoch": 15.441650548033527,
-      "grad_norm": 4.274187088012695,
-      "learning_rate": 7.206192463751255e-06,
-      "loss": 0.5324,
+      "epoch": 7.720832535788763,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.197430229345123e-06,
+      "loss": 0.9662199401855469,
       "step": 670600
     },
     {
-      "epoch": 15.446255871787788,
-      "grad_norm": 3.542017698287964,
-      "learning_rate": 7.198920899928739e-06,
-      "loss": 0.5365,
+      "epoch": 7.723135220863478,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.190158665522607e-06,
+      "loss": 0.9699585723876953,
       "step": 670800
     },
     {
-      "epoch": 15.450861195542046,
-      "grad_norm": 3.5981078147888184,
-      "learning_rate": 7.191649336106223e-06,
-      "loss": 0.5345,
+      "epoch": 7.725437905938193,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.182887101700092e-06,
+      "loss": 0.9652238464355469,
       "step": 671000
     },
     {
-      "epoch": 15.455466519296307,
-      "grad_norm": 2.7863192558288574,
-      "learning_rate": 7.184377772283707e-06,
-      "loss": 0.5415,
+      "epoch": 7.727740591012908,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.175615537877576e-06,
+      "loss": 1.00942138671875,
       "step": 671200
     },
     {
-      "epoch": 15.460071843050567,
-      "grad_norm": 3.475003957748413,
-      "learning_rate": 7.177106208461192e-06,
-      "loss": 0.5261,
+      "epoch": 7.730043276087623,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.16834397405506e-06,
+      "loss": 0.9876091003417968,
       "step": 671400
     },
     {
-      "epoch": 15.464677166804826,
-      "grad_norm": 2.9916470050811768,
-      "learning_rate": 7.169834644638677e-06,
-      "loss": 0.5382,
+      "epoch": 7.732345961162338,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.161072410232545e-06,
+      "loss": 0.9880516052246093,
       "step": 671600
     },
     {
-      "epoch": 15.469282490559086,
-      "grad_norm": 3.3957395553588867,
-      "learning_rate": 7.162563080816161e-06,
-      "loss": 0.5401,
+      "epoch": 7.734648646237053,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.153800846410029e-06,
+      "loss": 0.9689411926269531,
       "step": 671800
     },
     {
-      "epoch": 15.473887814313347,
-      "grad_norm": 3.301769495010376,
-      "learning_rate": 7.155291516993644e-06,
-      "loss": 0.529,
+      "epoch": 7.736951331311768,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.146529282587513e-06,
+      "loss": 0.9719631958007813,
       "step": 672000
     },
     {
-      "epoch": 15.478493138067606,
-      "grad_norm": 3.8502883911132812,
-      "learning_rate": 7.148019953171129e-06,
-      "loss": 0.5284,
+      "epoch": 7.739254016386482,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.139257718764998e-06,
+      "loss": 0.9924961853027344,
       "step": 672200
     },
     {
-      "epoch": 15.483098461821866,
-      "grad_norm": 2.8070068359375,
-      "learning_rate": 7.1407483893486135e-06,
-      "loss": 0.5286,
+      "epoch": 7.741556701461198,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.131986154942482e-06,
+      "loss": 0.9704615020751953,
       "step": 672400
     },
     {
-      "epoch": 15.487703785576127,
-      "grad_norm": 3.680961847305298,
-      "learning_rate": 7.133476825526099e-06,
-      "loss": 0.5374,
+      "epoch": 7.743859386535912,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.124714591119967e-06,
+      "loss": 0.9846534729003906,
       "step": 672600
     },
     {
-      "epoch": 15.492309109330385,
-      "grad_norm": 3.303067922592163,
-      "learning_rate": 7.126205261703582e-06,
-      "loss": 0.5306,
+      "epoch": 7.746162071610628,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.117443027297451e-06,
+      "loss": 0.9816999053955078,
       "step": 672800
     },
     {
-      "epoch": 15.496914433084646,
-      "grad_norm": 3.2984859943389893,
-      "learning_rate": 7.118933697881066e-06,
-      "loss": 0.5258,
+      "epoch": 7.748464756685342,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.110171463474935e-06,
+      "loss": 0.9916375732421875,
       "step": 673000
     },
     {
-      "epoch": 15.501519756838906,
-      "grad_norm": 3.2153737545013428,
-      "learning_rate": 7.111662134058551e-06,
-      "loss": 0.5297,
+      "epoch": 7.750767441760058,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.1028998996524195e-06,
+      "loss": 1.0035631561279297,
       "step": 673200
     },
     {
-      "epoch": 15.506125080593165,
-      "grad_norm": 2.8928987979888916,
-      "learning_rate": 7.1043905702360355e-06,
-      "loss": 0.5369,
+      "epoch": 7.753070126834772,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.095628335829904e-06,
+      "loss": 0.9929666900634766,
       "step": 673400
     },
     {
-      "epoch": 15.510730404347425,
-      "grad_norm": 3.484666109085083,
-      "learning_rate": 7.097119006413519e-06,
-      "loss": 0.5319,
+      "epoch": 7.755372811909488,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.088356772007388e-06,
+      "loss": 0.9711143493652343,
       "step": 673600
     },
     {
-      "epoch": 15.515335728101686,
-      "grad_norm": 3.3581833839416504,
-      "learning_rate": 7.089883800410116e-06,
-      "loss": 0.5299,
+      "epoch": 7.757675496984202,
+      "grad_norm": 0.875,
+      "learning_rate": 7.081085208184873e-06,
+      "loss": 0.9745367431640625,
       "step": 673800
     },
     {
-      "epoch": 15.519941051855945,
-      "grad_norm": 3.2200675010681152,
-      "learning_rate": 7.082612236587601e-06,
-      "loss": 0.5264,
+      "epoch": 7.7599781820589175,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.073813644362356e-06,
+      "loss": 0.982524642944336,
       "step": 674000
     },
     {
-      "epoch": 15.524546375610205,
-      "grad_norm": 3.4437928199768066,
-      "learning_rate": 7.075340672765085e-06,
-      "loss": 0.5304,
+      "epoch": 7.762280867133632,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.066542080539841e-06,
+      "loss": 0.984613037109375,
       "step": 674200
     },
     {
-      "epoch": 15.529151699364466,
-      "grad_norm": 2.679928779602051,
-      "learning_rate": 7.06806910894257e-06,
-      "loss": 0.5364,
+      "epoch": 7.764583552208347,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.059270516717326e-06,
+      "loss": 0.9818270874023437,
       "step": 674400
     },
     {
-      "epoch": 15.533757023118724,
-      "grad_norm": 3.410003662109375,
-      "learning_rate": 7.0607975451200535e-06,
-      "loss": 0.5329,
+      "epoch": 7.766886237283062,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.05199895289481e-06,
+      "loss": 0.9911299896240234,
       "step": 674600
     },
     {
-      "epoch": 15.538362346872985,
-      "grad_norm": 2.5245554447174072,
-      "learning_rate": 7.053525981297538e-06,
-      "loss": 0.5294,
+      "epoch": 7.769188922357777,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.044727389072294e-06,
+      "loss": 0.9584814453125,
       "step": 674800
     },
     {
-      "epoch": 15.542967670627245,
-      "grad_norm": 3.400965452194214,
-      "learning_rate": 7.046290775294135e-06,
-      "loss": 0.5289,
+      "epoch": 7.771491607432492,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.037455825249778e-06,
+      "loss": 0.984047622680664,
       "step": 675000
     },
     {
-      "epoch": 15.547572994381506,
-      "grad_norm": 3.5481555461883545,
-      "learning_rate": 7.03901921147162e-06,
-      "loss": 0.534,
+      "epoch": 7.7737942925072065,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.0301842614272625e-06,
+      "loss": 0.968690414428711,
       "step": 675200
     },
     {
-      "epoch": 15.552178318135764,
-      "grad_norm": 3.425252914428711,
-      "learning_rate": 7.031747647649104e-06,
-      "loss": 0.5249,
+      "epoch": 7.776096977581922,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.0229126976047476e-06,
+      "loss": 1.0027845764160157,
       "step": 675400
     },
     {
-      "epoch": 15.556783641890025,
-      "grad_norm": 3.6418352127075195,
-      "learning_rate": 7.024476083826587e-06,
-      "loss": 0.5178,
+      "epoch": 7.7783996626566365,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.015641133782231e-06,
+      "loss": 0.9877825164794922,
       "step": 675600
     },
     {
-      "epoch": 15.561388965644285,
-      "grad_norm": 2.9544789791107178,
-      "learning_rate": 7.017204520004072e-06,
-      "loss": 0.5337,
+      "epoch": 7.780702347731351,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.008369569959716e-06,
+      "loss": 0.9784912109375,
       "step": 675800
     },
     {
-      "epoch": 15.565994289398544,
-      "grad_norm": 2.8568692207336426,
-      "learning_rate": 7.0099329561815565e-06,
-      "loss": 0.5304,
+      "epoch": 7.7830050328060665,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.0010980061372e-06,
+      "loss": 0.9643284606933594,
       "step": 676000
     },
     {
-      "epoch": 15.570599613152805,
-      "grad_norm": 3.2933669090270996,
-      "learning_rate": 7.002661392359042e-06,
-      "loss": 0.5202,
+      "epoch": 7.785307717880781,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.9938264423146844e-06,
+      "loss": 0.9686238098144532,
       "step": 676200
     },
     {
-      "epoch": 15.575204936907065,
-      "grad_norm": 2.9382436275482178,
-      "learning_rate": 6.995389828536525e-06,
-      "loss": 0.5256,
+      "epoch": 7.7876104029554964,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.986554878492169e-06,
+      "loss": 0.9636341857910157,
       "step": 676400
     },
     {
-      "epoch": 15.579810260661324,
-      "grad_norm": 2.4080753326416016,
-      "learning_rate": 6.988118264714009e-06,
-      "loss": 0.5245,
+      "epoch": 7.789913088030211,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.979283314669653e-06,
+      "loss": 0.9681974029541016,
       "step": 676600
     },
     {
-      "epoch": 15.584415584415584,
-      "grad_norm": 2.816089630126953,
-      "learning_rate": 6.980846700891494e-06,
-      "loss": 0.5256,
+      "epoch": 7.792215773104926,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.972011750847137e-06,
+      "loss": 0.9814032745361329,
       "step": 676800
     },
     {
-      "epoch": 15.589020908169845,
-      "grad_norm": 2.963587999343872,
-      "learning_rate": 6.9735751370689785e-06,
-      "loss": 0.5329,
+      "epoch": 7.794518458179641,
+      "grad_norm": 0.625,
+      "learning_rate": 6.964740187024622e-06,
+      "loss": 0.9747649383544922,
       "step": 677000
     },
     {
-      "epoch": 15.593626231924103,
-      "grad_norm": 3.300342559814453,
-      "learning_rate": 6.966303573246462e-06,
-      "loss": 0.5376,
+      "epoch": 7.796821143254356,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.9574686232021055e-06,
+      "loss": 0.9751251220703125,
       "step": 677200
     },
     {
-      "epoch": 15.598231555678364,
-      "grad_norm": 3.2802958488464355,
-      "learning_rate": 6.959032009423947e-06,
-      "loss": 0.528,
+      "epoch": 7.799123828329071,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.95019705937959e-06,
+      "loss": 1.0097690582275392,
       "step": 677400
     },
     {
-      "epoch": 15.602836879432624,
-      "grad_norm": 2.6798970699310303,
-      "learning_rate": 6.951760445601431e-06,
-      "loss": 0.5237,
+      "epoch": 7.801426513403786,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.942925495557075e-06,
+      "loss": 0.97711669921875,
       "step": 677600
     },
     {
-      "epoch": 15.607442203186885,
-      "grad_norm": 3.517439603805542,
-      "learning_rate": 6.944488881778916e-06,
-      "loss": 0.5408,
+      "epoch": 7.803729198478501,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.935653931734559e-06,
+      "loss": 0.9708983612060547,
       "step": 677800
     },
     {
-      "epoch": 15.612047526941144,
-      "grad_norm": 3.87048602104187,
-      "learning_rate": 6.9372173179564e-06,
-      "loss": 0.534,
+      "epoch": 7.806031883553215,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.928382367912043e-06,
+      "loss": 0.9763180541992188,
       "step": 678000
     },
     {
-      "epoch": 15.616652850695404,
-      "grad_norm": 2.775428295135498,
-      "learning_rate": 6.929945754133884e-06,
-      "loss": 0.5248,
+      "epoch": 7.808334568627931,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.9211108040895275e-06,
+      "loss": 0.9944866180419922,
       "step": 678200
     },
     {
-      "epoch": 15.621258174449665,
-      "grad_norm": 2.7201571464538574,
-      "learning_rate": 6.922710548130481e-06,
-      "loss": 0.5338,
+      "epoch": 7.810637253702645,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.913839240267012e-06,
+      "loss": 0.9518258666992188,
       "step": 678400
     },
     {
-      "epoch": 15.625863498203923,
-      "grad_norm": 2.8548715114593506,
-      "learning_rate": 6.915438984307966e-06,
-      "loss": 0.5275,
+      "epoch": 7.812939938777361,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.906567676444497e-06,
+      "loss": 0.9801985931396484,
       "step": 678600
     },
     {
-      "epoch": 15.630468821958184,
-      "grad_norm": 2.6835644245147705,
-      "learning_rate": 6.90816742048545e-06,
-      "loss": 0.532,
+      "epoch": 7.815242623852075,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.899296112621981e-06,
+      "loss": 1.0015017700195312,
       "step": 678800
     },
     {
-      "epoch": 15.635074145712444,
-      "grad_norm": 2.5075621604919434,
-      "learning_rate": 6.900895856662933e-06,
-      "loss": 0.5155,
+      "epoch": 7.817545308926791,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.892024548799464e-06,
+      "loss": 0.9830785369873047,
       "step": 679000
     },
     {
-      "epoch": 15.639679469466703,
-      "grad_norm": 3.476208448410034,
-      "learning_rate": 6.89366065065953e-06,
-      "loss": 0.5243,
+      "epoch": 7.819847994001505,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.8847529849769494e-06,
+      "loss": 0.962328872680664,
       "step": 679200
     },
     {
-      "epoch": 15.644284793220963,
-      "grad_norm": 2.8655552864074707,
-      "learning_rate": 6.886389086837015e-06,
-      "loss": 0.5279,
+      "epoch": 7.82215067907622,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.877481421154434e-06,
+      "loss": 0.9914726257324219,
       "step": 679400
     },
     {
-      "epoch": 15.648890116975224,
-      "grad_norm": 2.7669894695281982,
-      "learning_rate": 6.8791175230144996e-06,
-      "loss": 0.5381,
+      "epoch": 7.824453364150935,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.870209857331919e-06,
+      "loss": 0.995595703125,
       "step": 679600
     },
     {
-      "epoch": 15.653495440729483,
-      "grad_norm": 4.442414283752441,
-      "learning_rate": 6.871845959191985e-06,
-      "loss": 0.5195,
+      "epoch": 7.82675604922565,
+      "grad_norm": 0.5,
+      "learning_rate": 6.862938293509402e-06,
+      "loss": 1.0089399719238281,
       "step": 679800
     },
     {
-      "epoch": 15.658100764483743,
-      "grad_norm": 2.7395546436309814,
-      "learning_rate": 6.864574395369468e-06,
-      "loss": 0.5274,
+      "epoch": 7.829058734300365,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.855666729686886e-06,
+      "loss": 0.975849609375,
       "step": 680000
     },
     {
-      "epoch": 15.662706088238004,
-      "grad_norm": 3.1513760089874268,
-      "learning_rate": 6.857302831546952e-06,
-      "loss": 0.5258,
+      "epoch": 7.83136141937508,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.848395165864371e-06,
+      "loss": 0.9750611114501954,
       "step": 680200
     },
     {
-      "epoch": 15.667311411992262,
-      "grad_norm": 3.7221457958221436,
-      "learning_rate": 6.850031267724437e-06,
-      "loss": 0.5325,
+      "epoch": 7.833664104449795,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.841123602041856e-06,
+      "loss": 0.9694293212890625,
       "step": 680400
     },
     {
-      "epoch": 15.671916735746523,
-      "grad_norm": 2.821197271347046,
-      "learning_rate": 6.8427597039019215e-06,
-      "loss": 0.5167,
+      "epoch": 7.83596678952451,
+      "grad_norm": 0.5,
+      "learning_rate": 6.833852038219339e-06,
+      "loss": 0.992059326171875,
       "step": 680600
     },
     {
-      "epoch": 15.676522059500783,
-      "grad_norm": 3.9420833587646484,
-      "learning_rate": 6.835488140079406e-06,
-      "loss": 0.5301,
+      "epoch": 7.838269474599225,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.826580474396824e-06,
+      "loss": 0.9736553192138672,
       "step": 680800
     },
     {
-      "epoch": 15.681127383255042,
-      "grad_norm": 3.045525312423706,
-      "learning_rate": 6.82821657625689e-06,
-      "loss": 0.5269,
+      "epoch": 7.84057215967394,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.819308910574308e-06,
+      "loss": 0.9563919830322266,
       "step": 681000
     },
     {
-      "epoch": 15.685732707009302,
-      "grad_norm": 3.1435394287109375,
-      "learning_rate": 6.820945012434374e-06,
-      "loss": 0.5343,
+      "epoch": 7.842874844748655,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.812037346751793e-06,
+      "loss": 0.9833192443847656,
       "step": 681200
     },
     {
-      "epoch": 15.690338030763563,
-      "grad_norm": 3.5429775714874268,
-      "learning_rate": 6.813673448611859e-06,
-      "loss": 0.5335,
+      "epoch": 7.84517752982337,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.804765782929277e-06,
+      "loss": 0.9806179809570312,
       "step": 681400
     },
     {
-      "epoch": 15.694943354517823,
-      "grad_norm": 3.320526361465454,
-      "learning_rate": 6.8064018847893435e-06,
-      "loss": 0.5305,
+      "epoch": 7.847480214898084,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.797494219106761e-06,
+      "loss": 0.9624887847900391,
       "step": 681600
     },
     {
-      "epoch": 15.699548678272082,
-      "grad_norm": 3.505725860595703,
-      "learning_rate": 6.799130320966827e-06,
-      "loss": 0.5378,
+      "epoch": 7.8497828999728,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.790222655284246e-06,
+      "loss": 1.0056896209716797,
       "step": 681800
     },
     {
-      "epoch": 15.704154002026343,
-      "grad_norm": 3.6867668628692627,
-      "learning_rate": 6.791858757144312e-06,
-      "loss": 0.5328,
+      "epoch": 7.852085585047514,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.78295109146173e-06,
+      "loss": 0.9849835968017578,
       "step": 682000
     },
     {
-      "epoch": 15.708759325780603,
-      "grad_norm": 3.264688491821289,
-      "learning_rate": 6.784587193321796e-06,
-      "loss": 0.5392,
+      "epoch": 7.85438827012223,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.7756795276392136e-06,
+      "loss": 0.9797569274902344,
       "step": 682200
     },
     {
-      "epoch": 15.713364649534862,
-      "grad_norm": 2.858060121536255,
-      "learning_rate": 6.77731562949928e-06,
-      "loss": 0.5387,
+      "epoch": 7.856690955196944,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.768407963816699e-06,
+      "loss": 0.9842268371582031,
       "step": 682400
     },
     {
-      "epoch": 15.717969973289122,
-      "grad_norm": 3.1449389457702637,
-      "learning_rate": 6.7700440656767646e-06,
-      "loss": 0.5224,
+      "epoch": 7.85899364027166,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.761136399994183e-06,
+      "loss": 0.9827518463134766,
       "step": 682600
     },
     {
-      "epoch": 15.722575297043383,
-      "grad_norm": 4.474485397338867,
-      "learning_rate": 6.762772501854249e-06,
-      "loss": 0.5496,
+      "epoch": 7.861296325346374,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.753864836171668e-06,
+      "loss": 0.9835359954833984,
       "step": 682800
     },
     {
-      "epoch": 15.727180620797641,
-      "grad_norm": 3.6581742763519287,
-      "learning_rate": 6.755500938031734e-06,
-      "loss": 0.5347,
+      "epoch": 7.863599010421089,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.746593272349151e-06,
+      "loss": 0.9827082061767578,
       "step": 683000
     },
     {
-      "epoch": 15.731785944551902,
-      "grad_norm": 3.4204602241516113,
-      "learning_rate": 6.748229374209218e-06,
-      "loss": 0.5319,
+      "epoch": 7.865901695495804,
+      "grad_norm": 1.0,
+      "learning_rate": 6.7393217085266355e-06,
+      "loss": 0.9873677825927735,
       "step": 683200
     },
     {
-      "epoch": 15.736391268306162,
-      "grad_norm": 3.710782289505005,
-      "learning_rate": 6.7409578103867014e-06,
-      "loss": 0.5362,
+      "epoch": 7.868204380570519,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.7320501447041206e-06,
+      "loss": 0.9825727081298828,
       "step": 683400
     },
     {
-      "epoch": 15.740996592060421,
-      "grad_norm": 3.196833610534668,
-      "learning_rate": 6.7336862465641865e-06,
-      "loss": 0.5305,
+      "epoch": 7.870507065645234,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.724778580881605e-06,
+      "loss": 0.9821214294433593,
       "step": 683600
     },
     {
-      "epoch": 15.745601915814682,
-      "grad_norm": 3.187162160873413,
-      "learning_rate": 6.726414682741671e-06,
-      "loss": 0.5248,
+      "epoch": 7.872809750719949,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.717507017059088e-06,
+      "loss": 0.9567387390136719,
       "step": 683800
     },
     {
-      "epoch": 15.750207239568942,
-      "grad_norm": 3.471301794052124,
-      "learning_rate": 6.719143118919155e-06,
-      "loss": 0.5289,
+      "epoch": 7.875112435794664,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.710235453236573e-06,
+      "loss": 0.9816065979003906,
       "step": 684000
     },
     {
-      "epoch": 15.7548125633232,
-      "grad_norm": 3.092719316482544,
-      "learning_rate": 6.711871555096639e-06,
-      "loss": 0.5483,
+      "epoch": 7.877415120869379,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.7029638894140575e-06,
+      "loss": 1.004992218017578,
       "step": 684200
     },
     {
-      "epoch": 15.759417887077461,
-      "grad_norm": 2.3404221534729004,
-      "learning_rate": 6.704599991274123e-06,
-      "loss": 0.5327,
+      "epoch": 7.879717805944094,
+      "grad_norm": 0.625,
+      "learning_rate": 6.6956923255915425e-06,
+      "loss": 0.9947642517089844,
       "step": 684400
     },
     {
-      "epoch": 15.764023210831722,
-      "grad_norm": 3.6259121894836426,
-      "learning_rate": 6.69736478527072e-06,
-      "loss": 0.53,
+      "epoch": 7.8820204910188085,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.688420761769026e-06,
+      "loss": 0.9833740234375,
       "step": 684600
     },
     {
-      "epoch": 15.768628534585982,
-      "grad_norm": 3.0392374992370605,
-      "learning_rate": 6.690129579267317e-06,
-      "loss": 0.5328,
+      "epoch": 7.884323176093524,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.68114919794651e-06,
+      "loss": 0.978486328125,
       "step": 684800
     },
     {
-      "epoch": 15.773233858340241,
-      "grad_norm": 3.2242770195007324,
-      "learning_rate": 6.682858015444802e-06,
-      "loss": 0.5337,
+      "epoch": 7.8866258611682385,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.673877634123995e-06,
+      "loss": 0.9901490020751953,
       "step": 685000
     },
     {
-      "epoch": 15.777839182094501,
-      "grad_norm": 2.7651524543762207,
-      "learning_rate": 6.6755864516222865e-06,
-      "loss": 0.5286,
+      "epoch": 7.888928546242953,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.666606070301479e-06,
+      "loss": 0.9926892852783203,
       "step": 685200
     },
     {
-      "epoch": 15.782444505848762,
-      "grad_norm": 2.7037301063537598,
-      "learning_rate": 6.66831488779977e-06,
-      "loss": 0.5247,
+      "epoch": 7.8912312313176685,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.659334506478963e-06,
+      "loss": 0.9790725708007812,
       "step": 685400
     },
     {
-      "epoch": 15.78704982960302,
-      "grad_norm": 3.071185350418091,
-      "learning_rate": 6.661043323977255e-06,
-      "loss": 0.5305,
+      "epoch": 7.893533916392383,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.652062942656448e-06,
+      "loss": 0.9824659729003906,
       "step": 685600
     },
     {
-      "epoch": 15.791655153357281,
-      "grad_norm": 3.1269609928131104,
-      "learning_rate": 6.653771760154739e-06,
-      "loss": 0.523,
+      "epoch": 7.895836601467098,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.644791378833932e-06,
+      "loss": 0.9952230834960938,
       "step": 685800
     },
     {
-      "epoch": 15.796260477111542,
-      "grad_norm": 2.9973862171173096,
-      "learning_rate": 6.646500196332223e-06,
-      "loss": 0.5328,
+      "epoch": 7.898139286541813,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.637519815011417e-06,
+      "loss": 0.9813288116455078,
       "step": 686000
     },
     {
-      "epoch": 15.8008658008658,
-      "grad_norm": 2.5962016582489014,
-      "learning_rate": 6.6392286325097076e-06,
-      "loss": 0.5325,
+      "epoch": 7.900441971616528,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.630248251188901e-06,
+      "loss": 0.9603262329101563,
       "step": 686200
     },
     {
-      "epoch": 15.80547112462006,
-      "grad_norm": 3.33072566986084,
-      "learning_rate": 6.631957068687192e-06,
-      "loss": 0.5284,
+      "epoch": 7.902744656691243,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.622976687366385e-06,
+      "loss": 0.9719887542724609,
       "step": 686400
     },
     {
-      "epoch": 15.810076448374321,
-      "grad_norm": 2.7956624031066895,
-      "learning_rate": 6.624685504864677e-06,
-      "loss": 0.5206,
+      "epoch": 7.9050473417659575,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.61570512354387e-06,
+      "loss": 0.9816300964355469,
       "step": 686600
     },
     {
-      "epoch": 15.81468177212858,
-      "grad_norm": 3.8837435245513916,
-      "learning_rate": 6.617413941042161e-06,
-      "loss": 0.5238,
+      "epoch": 7.907350026840673,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.608433559721354e-06,
+      "loss": 0.9870523834228515,
       "step": 686800
     },
     {
-      "epoch": 15.81928709588284,
-      "grad_norm": 3.072110652923584,
-      "learning_rate": 6.6101423772196445e-06,
-      "loss": 0.5346,
+      "epoch": 7.909652711915388,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.601161995898838e-06,
+      "loss": 0.9839715576171875,
       "step": 687000
     },
     {
-      "epoch": 15.823892419637101,
-      "grad_norm": 2.895686388015747,
-      "learning_rate": 6.6028708133971295e-06,
-      "loss": 0.5327,
+      "epoch": 7.911955396990103,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.5938904320763224e-06,
+      "loss": 0.9607354736328125,
       "step": 687200
     },
     {
-      "epoch": 15.82849774339136,
-      "grad_norm": 3.0027430057525635,
-      "learning_rate": 6.595599249574614e-06,
-      "loss": 0.5186,
+      "epoch": 7.914258082064817,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.586618868253807e-06,
+      "loss": 0.9836206817626953,
       "step": 687400
     },
     {
-      "epoch": 15.83310306714562,
-      "grad_norm": 2.7608888149261475,
-      "learning_rate": 6.588327685752098e-06,
-      "loss": 0.5347,
+      "epoch": 7.916560767139533,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.579347304431292e-06,
+      "loss": 0.9973263549804687,
       "step": 687600
     },
     {
-      "epoch": 15.83770839089988,
-      "grad_norm": 3.104914903640747,
-      "learning_rate": 6.581056121929582e-06,
-      "loss": 0.5289,
+      "epoch": 7.918863452214247,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.572075740608776e-06,
+      "loss": 0.9725442504882813,
       "step": 687800
     },
     {
-      "epoch": 15.84231371465414,
-      "grad_norm": 4.210239410400391,
-      "learning_rate": 6.573784558107066e-06,
-      "loss": 0.542,
+      "epoch": 7.921166137288963,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.564804176786259e-06,
+      "loss": 0.9735817718505859,
       "step": 688000
     },
     {
-      "epoch": 15.8469190384084,
-      "grad_norm": 2.743356704711914,
-      "learning_rate": 6.566512994284551e-06,
-      "loss": 0.5315,
+      "epoch": 7.923468822363677,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.557532612963744e-06,
+      "loss": 0.9831144714355469,
       "step": 688200
     },
     {
-      "epoch": 15.85152436216266,
-      "grad_norm": 3.4717984199523926,
-      "learning_rate": 6.559241430462036e-06,
-      "loss": 0.5337,
+      "epoch": 7.925771507438393,
+      "grad_norm": 0.625,
+      "learning_rate": 6.550261049141229e-06,
+      "loss": 0.9740536499023438,
       "step": 688400
     },
     {
-      "epoch": 15.85612968591692,
-      "grad_norm": 3.4075236320495605,
-      "learning_rate": 6.551969866639519e-06,
-      "loss": 0.5363,
+      "epoch": 7.928074192513107,
+      "grad_norm": 0.75,
+      "learning_rate": 6.542989485318713e-06,
+      "loss": 0.9716651153564453,
       "step": 688600
     },
     {
-      "epoch": 15.86073500967118,
-      "grad_norm": 4.167804718017578,
-      "learning_rate": 6.544698302817004e-06,
-      "loss": 0.5274,
+      "epoch": 7.930376877587822,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.535717921496197e-06,
+      "loss": 0.9694881439208984,
       "step": 688800
     },
     {
-      "epoch": 15.86534033342544,
-      "grad_norm": 3.3863329887390137,
-      "learning_rate": 6.537426738994488e-06,
-      "loss": 0.5284,
+      "epoch": 7.932679562662537,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.528446357673681e-06,
+      "loss": 0.9982528686523438,
       "step": 689000
     },
     {
-      "epoch": 15.8699456571797,
-      "grad_norm": 3.17704439163208,
-      "learning_rate": 6.5301551751719726e-06,
-      "loss": 0.5336,
+      "epoch": 7.934982247737252,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.5211747938511655e-06,
+      "loss": 0.9639457702636719,
       "step": 689200
     },
     {
-      "epoch": 15.874550980933959,
-      "grad_norm": 2.546211004257202,
-      "learning_rate": 6.522883611349458e-06,
-      "loss": 0.5344,
+      "epoch": 7.937284932811967,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.5139032300286505e-06,
+      "loss": 0.9916056823730469,
       "step": 689400
     },
     {
-      "epoch": 15.87915630468822,
-      "grad_norm": 3.9471852779388428,
-      "learning_rate": 6.515612047526941e-06,
-      "loss": 0.5266,
+      "epoch": 7.939587617886682,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.506631666206134e-06,
+      "loss": 0.964366683959961,
       "step": 689600
     },
     {
-      "epoch": 15.88376162844248,
-      "grad_norm": 2.803940773010254,
-      "learning_rate": 6.508340483704425e-06,
-      "loss": 0.5422,
+      "epoch": 7.941890302961397,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.499360102383619e-06,
+      "loss": 0.998613510131836,
       "step": 689800
     },
     {
-      "epoch": 15.888366952196739,
-      "grad_norm": 3.0068013668060303,
-      "learning_rate": 6.50106891988191e-06,
-      "loss": 0.5318,
+      "epoch": 7.944192988036112,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.492088538561103e-06,
+      "loss": 0.9819116973876953,
       "step": 690000
     },
     {
-      "epoch": 15.892972275951,
-      "grad_norm": 3.8956100940704346,
-      "learning_rate": 6.4937973560593945e-06,
-      "loss": 0.5329,
+      "epoch": 7.946495673110827,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.484816974738587e-06,
+      "loss": 0.992962875366211,
       "step": 690200
     },
     {
-      "epoch": 15.89757759970526,
-      "grad_norm": 2.6682162284851074,
-      "learning_rate": 6.486562150055991e-06,
-      "loss": 0.5214,
+      "epoch": 7.948798358185542,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.477545410916072e-06,
+      "loss": 0.9869437408447266,
       "step": 690400
     },
     {
-      "epoch": 15.902182923459518,
-      "grad_norm": 3.1031453609466553,
-      "learning_rate": 6.479290586233476e-06,
-      "loss": 0.5215,
+      "epoch": 7.951101043260257,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.470273847093556e-06,
+      "loss": 0.9729832458496094,
       "step": 690600
     },
     {
-      "epoch": 15.906788247213779,
-      "grad_norm": 3.070067882537842,
-      "learning_rate": 6.47201902241096e-06,
-      "loss": 0.5341,
+      "epoch": 7.953403728334972,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.46300228327104e-06,
+      "loss": 0.9928602600097656,
       "step": 690800
     },
     {
-      "epoch": 15.91139357096804,
-      "grad_norm": 3.4854178428649902,
-      "learning_rate": 6.464747458588444e-06,
-      "loss": 0.5161,
+      "epoch": 7.955706413409686,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.455730719448525e-06,
+      "loss": 0.9930166625976562,
       "step": 691000
     },
     {
-      "epoch": 15.9159988947223,
-      "grad_norm": 2.887986421585083,
-      "learning_rate": 6.457475894765929e-06,
-      "loss": 0.5282,
+      "epoch": 7.958009098484402,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.4484591556260085e-06,
+      "loss": 0.9928379058837891,
       "step": 691200
     },
     {
-      "epoch": 15.920604218476559,
-      "grad_norm": 3.0503897666931152,
-      "learning_rate": 6.4502043309434125e-06,
-      "loss": 0.5358,
+      "epoch": 7.960311783559116,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.441187591803494e-06,
+      "loss": 0.9737590026855468,
       "step": 691400
     },
     {
-      "epoch": 15.925209542230819,
-      "grad_norm": 2.886718988418579,
-      "learning_rate": 6.442932767120897e-06,
-      "loss": 0.5237,
+      "epoch": 7.962614468633832,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.433916027980978e-06,
+      "loss": 0.9859085845947265,
       "step": 691600
     },
     {
-      "epoch": 15.92981486598508,
-      "grad_norm": 3.217623710632324,
-      "learning_rate": 6.435661203298382e-06,
-      "loss": 0.5236,
+      "epoch": 7.964917153708546,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.426644464158462e-06,
+      "loss": 0.9696751403808593,
       "step": 691800
     },
     {
-      "epoch": 15.934420189739338,
-      "grad_norm": 3.3502800464630127,
-      "learning_rate": 6.428389639475866e-06,
-      "loss": 0.531,
+      "epoch": 7.967219838783262,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.419372900335946e-06,
+      "loss": 0.9876094818115234,
       "step": 692000
     },
     {
-      "epoch": 15.939025513493599,
-      "grad_norm": 2.8795714378356934,
-      "learning_rate": 6.42111807565335e-06,
-      "loss": 0.5261,
+      "epoch": 7.969522523857976,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.4121013365134305e-06,
+      "loss": 0.9785052490234375,
       "step": 692200
     },
     {
-      "epoch": 15.94363083724786,
-      "grad_norm": 3.2952957153320312,
-      "learning_rate": 6.4138465118308345e-06,
-      "loss": 0.5382,
+      "epoch": 7.971825208932691,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.404829772690915e-06,
+      "loss": 0.9876641082763672,
       "step": 692400
     },
     {
-      "epoch": 15.948236161002118,
-      "grad_norm": 3.00901460647583,
-      "learning_rate": 6.406574948008319e-06,
-      "loss": 0.5271,
+      "epoch": 7.974127894007406,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.3975582088684e-06,
+      "loss": 0.9774613189697265,
       "step": 692600
     },
     {
-      "epoch": 15.952841484756378,
-      "grad_norm": 2.900832414627075,
-      "learning_rate": 6.399303384185804e-06,
-      "loss": 0.5227,
+      "epoch": 7.976430579082121,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.390286645045884e-06,
+      "loss": 1.0123577880859376,
       "step": 692800
     },
     {
-      "epoch": 15.957446808510639,
-      "grad_norm": 3.579869031906128,
-      "learning_rate": 6.392031820363287e-06,
-      "loss": 0.5219,
+      "epoch": 7.978733264156836,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.383015081223368e-06,
+      "loss": 0.9731853485107422,
       "step": 693000
     },
     {
-      "epoch": 15.962052132264898,
-      "grad_norm": 2.8938448429107666,
-      "learning_rate": 6.384760256540771e-06,
-      "loss": 0.5369,
+      "epoch": 7.981035949231551,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.375743517400852e-06,
+      "loss": 0.9835764312744141,
       "step": 693200
     },
     {
-      "epoch": 15.966657456019158,
-      "grad_norm": 3.0627634525299072,
-      "learning_rate": 6.377488692718256e-06,
-      "loss": 0.5327,
+      "epoch": 7.983338634306266,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.368471953578337e-06,
+      "loss": 0.9932479095458985,
       "step": 693400
     },
     {
-      "epoch": 15.971262779773419,
-      "grad_norm": 3.0290048122406006,
-      "learning_rate": 6.370217128895741e-06,
-      "loss": 0.5283,
+      "epoch": 7.9856413193809805,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.361200389755822e-06,
+      "loss": 0.950868911743164,
       "step": 693600
     },
     {
-      "epoch": 15.975868103527677,
-      "grad_norm": 2.843632698059082,
-      "learning_rate": 6.362945565073225e-06,
-      "loss": 0.5388,
+      "epoch": 7.987944004455696,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.353928825933305e-06,
+      "loss": 0.9951741027832032,
       "step": 693800
     },
     {
-      "epoch": 15.980473427281938,
-      "grad_norm": 3.2812376022338867,
-      "learning_rate": 6.355674001250709e-06,
-      "loss": 0.5199,
+      "epoch": 7.9902466895304105,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.346657262110789e-06,
+      "loss": 1.0014451599121095,
       "step": 694000
     },
     {
-      "epoch": 15.985078751036198,
-      "grad_norm": 3.616701126098633,
-      "learning_rate": 6.348438795247306e-06,
-      "loss": 0.5359,
+      "epoch": 7.992549374605126,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.339385698288274e-06,
+      "loss": 0.989116439819336,
       "step": 694200
     },
     {
-      "epoch": 15.989684074790457,
-      "grad_norm": 3.0894737243652344,
-      "learning_rate": 6.34116723142479e-06,
-      "loss": 0.5268,
+      "epoch": 7.9948520596798405,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.3321141344657586e-06,
+      "loss": 0.9893257904052735,
       "step": 694400
     },
     {
-      "epoch": 15.994289398544717,
-      "grad_norm": 3.021001100540161,
-      "learning_rate": 6.333895667602275e-06,
-      "loss": 0.5346,
+      "epoch": 7.997154744754555,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.324842570643243e-06,
+      "loss": 0.9822025299072266,
       "step": 694600
     },
     {
-      "epoch": 15.998894722298978,
-      "grad_norm": 3.486922264099121,
-      "learning_rate": 6.326624103779759e-06,
-      "loss": 0.5314,
+      "epoch": 7.99945742982927,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.317571006820727e-06,
+      "loss": 0.9817815399169922,
       "step": 694800
     },
     {
-      "epoch": 16.0,
-      "eval_loss": 0.5152611136436462,
-      "eval_runtime": 161.1159,
-      "eval_samples_per_second": 176.029,
-      "eval_steps_per_second": 11.005,
+      "epoch": 8.0,
+      "eval_loss": NaN,
+      "eval_runtime": 428.7338,
+      "eval_samples_per_second": 66.151,
+      "eval_steps_per_second": 33.076,
       "step": 694848
     },
     {
-      "epoch": 16.003500046053237,
-      "grad_norm": 3.235459566116333,
-      "learning_rate": 6.319352539957243e-06,
-      "loss": 0.5261,
+      "epoch": 8.001750040656784,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.310299442998211e-06,
+      "loss": 0.9696829223632812,
       "step": 695000
     },
     {
-      "epoch": 16.0081053698075,
-      "grad_norm": 2.7268731594085693,
-      "learning_rate": 6.312080976134728e-06,
-      "loss": 0.5294,
+      "epoch": 8.004052725731498,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.303027879175696e-06,
+      "loss": 0.9977623748779297,
       "step": 695200
     },
     {
-      "epoch": 16.012710693561758,
-      "grad_norm": 2.6637632846832275,
-      "learning_rate": 6.304809412312212e-06,
-      "loss": 0.5276,
+      "epoch": 8.006355410806213,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.29575631535318e-06,
+      "loss": 0.9823947143554688,
       "step": 695400
     },
     {
-      "epoch": 16.017316017316016,
-      "grad_norm": 2.3737680912017822,
-      "learning_rate": 6.297537848489696e-06,
-      "loss": 0.5162,
+      "epoch": 8.008658095880929,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.288484751530664e-06,
+      "loss": 0.9951921081542969,
       "step": 695600
     },
     {
-      "epoch": 16.02192134107028,
-      "grad_norm": 2.3051598072052,
-      "learning_rate": 6.2902662846671806e-06,
-      "loss": 0.5141,
+      "epoch": 8.010960780955642,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.281213187708149e-06,
+      "loss": 0.9983968353271484,
       "step": 695800
     },
     {
-      "epoch": 16.026526664824537,
-      "grad_norm": 3.3139994144439697,
-      "learning_rate": 6.282994720844665e-06,
-      "loss": 0.5175,
+      "epoch": 8.013263466030358,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.273941623885633e-06,
+      "loss": 0.9798602294921875,
       "step": 696000
     },
     {
-      "epoch": 16.031131988578796,
-      "grad_norm": 2.9257097244262695,
-      "learning_rate": 6.27572315702215e-06,
-      "loss": 0.5121,
+      "epoch": 8.015566151105073,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.266670060063117e-06,
+      "loss": 1.0001656341552734,
       "step": 696200
     },
     {
-      "epoch": 16.035737312333058,
-      "grad_norm": 3.4968149662017822,
-      "learning_rate": 6.268451593199633e-06,
-      "loss": 0.5392,
+      "epoch": 8.017868836179789,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.259398496240602e-06,
+      "loss": 0.9662248229980469,
       "step": 696400
     },
     {
-      "epoch": 16.040342636087317,
-      "grad_norm": 2.8969199657440186,
-      "learning_rate": 6.2611800293771174e-06,
-      "loss": 0.5296,
+      "epoch": 8.020171521254502,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.252126932418086e-06,
+      "loss": 0.9670308685302734,
       "step": 696600
     },
     {
-      "epoch": 16.044947959841576,
-      "grad_norm": 2.9978652000427246,
-      "learning_rate": 6.253944823373714e-06,
-      "loss": 0.5304,
+      "epoch": 8.022474206329218,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.244855368595571e-06,
+      "loss": 0.9894591522216797,
       "step": 696800
     },
     {
-      "epoch": 16.049553283595838,
-      "grad_norm": 3.812542676925659,
-      "learning_rate": 6.246673259551199e-06,
-      "loss": 0.526,
+      "epoch": 8.024776891403933,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.237583804773054e-06,
+      "loss": 0.9984318542480469,
       "step": 697000
     },
     {
-      "epoch": 16.054158607350097,
-      "grad_norm": 3.203511953353882,
-      "learning_rate": 6.239401695728684e-06,
-      "loss": 0.5188,
+      "epoch": 8.027079576478648,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.2303122409505385e-06,
+      "loss": 0.9907484436035157,
       "step": 697200
     },
     {
-      "epoch": 16.058763931104355,
-      "grad_norm": 3.406111478805542,
-      "learning_rate": 6.232130131906168e-06,
-      "loss": 0.5199,
+      "epoch": 8.029382261553362,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.2230406771280236e-06,
+      "loss": 0.9838898468017578,
       "step": 697400
     },
     {
-      "epoch": 16.063369254858618,
-      "grad_norm": 3.4208409786224365,
-      "learning_rate": 6.224858568083652e-06,
-      "loss": 0.5317,
+      "epoch": 8.031684946628078,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.215769113305508e-06,
+      "loss": 0.9810810852050781,
       "step": 697600
     },
     {
-      "epoch": 16.067974578612876,
-      "grad_norm": 3.2451865673065186,
-      "learning_rate": 6.217587004261136e-06,
-      "loss": 0.5201,
+      "epoch": 8.033987631702793,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.208497549482992e-06,
+      "loss": 0.968192138671875,
       "step": 697800
     },
     {
-      "epoch": 16.072579902367135,
-      "grad_norm": 2.8705670833587646,
-      "learning_rate": 6.210315440438621e-06,
-      "loss": 0.53,
+      "epoch": 8.036290316777507,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.201225985660476e-06,
+      "loss": 0.9575815582275391,
       "step": 698000
     },
     {
-      "epoch": 16.077185226121397,
-      "grad_norm": 3.8681085109710693,
-      "learning_rate": 6.203043876616106e-06,
-      "loss": 0.5249,
+      "epoch": 8.038593001852222,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.1939544218379604e-06,
+      "loss": 0.9703910827636719,
       "step": 698200
     },
     {
-      "epoch": 16.081790549875656,
-      "grad_norm": 4.016938209533691,
-      "learning_rate": 6.195772312793589e-06,
-      "loss": 0.5138,
+      "epoch": 8.040895686926937,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.1866828580154455e-06,
+      "loss": 0.989373779296875,
       "step": 698400
     },
     {
-      "epoch": 16.086395873629915,
-      "grad_norm": 2.935976266860962,
-      "learning_rate": 6.188500748971074e-06,
-      "loss": 0.5238,
+      "epoch": 8.043198372001653,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.179411294192929e-06,
+      "loss": 0.9677561950683594,
       "step": 698600
     },
     {
-      "epoch": 16.091001197384177,
-      "grad_norm": 3.2641563415527344,
-      "learning_rate": 6.181229185148558e-06,
-      "loss": 0.5174,
+      "epoch": 8.045501057076367,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.172139730370413e-06,
+      "loss": 0.9914716339111328,
       "step": 698800
     },
     {
-      "epoch": 16.095606521138436,
-      "grad_norm": 3.6100635528564453,
-      "learning_rate": 6.173993979145155e-06,
-      "loss": 0.5299,
+      "epoch": 8.047803742151082,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.164868166547898e-06,
+      "loss": 0.9957130432128907,
       "step": 699000
     },
     {
-      "epoch": 16.100211844892694,
-      "grad_norm": 3.5034003257751465,
-      "learning_rate": 6.16672241532264e-06,
-      "loss": 0.5293,
+      "epoch": 8.050106427225797,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.157596602725382e-06,
+      "loss": 0.9571633148193359,
       "step": 699200
     },
     {
-      "epoch": 16.104817168646957,
-      "grad_norm": 2.800861358642578,
-      "learning_rate": 6.159450851500124e-06,
-      "loss": 0.5214,
+      "epoch": 8.052409112300511,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.150325038902867e-06,
+      "loss": 0.9698536682128907,
       "step": 699400
     },
     {
-      "epoch": 16.109422492401215,
-      "grad_norm": 3.0961642265319824,
-      "learning_rate": 6.152179287677608e-06,
-      "loss": 0.5254,
+      "epoch": 8.054711797375226,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.143053475080351e-06,
+      "loss": 0.965628662109375,
       "step": 699600
     },
     {
-      "epoch": 16.114027816155474,
-      "grad_norm": 2.8042900562286377,
-      "learning_rate": 6.144907723855093e-06,
-      "loss": 0.5155,
+      "epoch": 8.057014482449942,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.135781911257835e-06,
+      "loss": 0.9815982818603516,
       "step": 699800
     },
     {
-      "epoch": 16.118633139909736,
-      "grad_norm": 3.172004461288452,
-      "learning_rate": 6.137636160032577e-06,
-      "loss": 0.5195,
+      "epoch": 8.059317167524657,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.12851034743532e-06,
+      "loss": 0.9958517456054687,
       "step": 700000
     },
     {
-      "epoch": 16.123238463663995,
-      "grad_norm": 3.4041786193847656,
-      "learning_rate": 6.1303645962100605e-06,
-      "loss": 0.5156,
+      "epoch": 8.061619852599371,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.121238783612804e-06,
+      "loss": 0.9758916473388672,
       "step": 700200
     },
     {
-      "epoch": 16.127843787418257,
-      "grad_norm": 3.219374895095825,
-      "learning_rate": 6.1230930323875455e-06,
-      "loss": 0.5299,
+      "epoch": 8.063922537674086,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.113967219790288e-06,
+      "loss": 0.9970526123046874,
       "step": 700400
     },
     {
-      "epoch": 16.132449111172516,
-      "grad_norm": 2.8847038745880127,
-      "learning_rate": 6.11582146856503e-06,
-      "loss": 0.5128,
+      "epoch": 8.066225222748802,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.106695655967773e-06,
+      "loss": 0.9790088653564453,
       "step": 700600
     },
     {
-      "epoch": 16.137054434926775,
-      "grad_norm": 2.9731578826904297,
-      "learning_rate": 6.108622620380739e-06,
-      "loss": 0.5165,
+      "epoch": 8.068527907823517,
+      "grad_norm": 0.5,
+      "learning_rate": 6.099424092145257e-06,
+      "loss": 0.9624451446533203,
       "step": 700800
     },
     {
-      "epoch": 16.141659758681037,
-      "grad_norm": 3.0595650672912598,
-      "learning_rate": 6.1013510565582236e-06,
-      "loss": 0.5284,
+      "epoch": 8.070830592898231,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.092152528322741e-06,
+      "loss": 0.9581060028076172,
       "step": 701000
     },
     {
-      "epoch": 16.146265082435296,
-      "grad_norm": 2.8716259002685547,
-      "learning_rate": 6.094079492735709e-06,
-      "loss": 0.519,
+      "epoch": 8.073133277972946,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.084880964500225e-06,
+      "loss": 0.9783612060546875,
       "step": 701200
     },
     {
-      "epoch": 16.150870406189554,
-      "grad_norm": 3.1690585613250732,
-      "learning_rate": 6.086807928913192e-06,
-      "loss": 0.529,
+      "epoch": 8.075435963047662,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.07760940067771e-06,
+      "loss": 1.001886444091797,
       "step": 701400
     },
     {
-      "epoch": 16.155475729943817,
-      "grad_norm": 2.8554840087890625,
-      "learning_rate": 6.079536365090676e-06,
-      "loss": 0.5153,
+      "epoch": 8.077738648122375,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.070337836855195e-06,
+      "loss": 0.9688577270507812,
       "step": 701600
     },
     {
-      "epoch": 16.160081053698075,
-      "grad_norm": 2.701791524887085,
-      "learning_rate": 6.072264801268161e-06,
-      "loss": 0.5385,
+      "epoch": 8.08004133319709,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.063066273032679e-06,
+      "loss": 0.96413330078125,
       "step": 701800
     },
     {
-      "epoch": 16.164686377452334,
-      "grad_norm": 2.9476139545440674,
-      "learning_rate": 6.0649932374456455e-06,
-      "loss": 0.5344,
+      "epoch": 8.082344018271806,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.055794709210162e-06,
+      "loss": 0.9906436157226562,
       "step": 702000
     },
     {
-      "epoch": 16.169291701206596,
-      "grad_norm": 3.677595853805542,
-      "learning_rate": 6.057721673623129e-06,
-      "loss": 0.5302,
+      "epoch": 8.084646703346522,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.048523145387647e-06,
+      "loss": 0.9788544464111328,
       "step": 702200
     },
     {
-      "epoch": 16.173897024960855,
-      "grad_norm": 3.322723388671875,
-      "learning_rate": 6.050450109800614e-06,
-      "loss": 0.5301,
+      "epoch": 8.086949388421235,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.041251581565132e-06,
+      "loss": 0.9843678283691406,
       "step": 702400
     },
     {
-      "epoch": 16.178502348715114,
-      "grad_norm": 3.177597761154175,
-      "learning_rate": 6.043178545978098e-06,
-      "loss": 0.5293,
+      "epoch": 8.08925207349595,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.033980017742616e-06,
+      "loss": 0.9825135040283203,
       "step": 702600
     },
     {
-      "epoch": 16.183107672469376,
-      "grad_norm": 2.9423975944519043,
-      "learning_rate": 6.035906982155583e-06,
-      "loss": 0.5263,
+      "epoch": 8.091554758570666,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.0267084539201e-06,
+      "loss": 0.991893310546875,
       "step": 702800
     },
     {
-      "epoch": 16.187712996223635,
-      "grad_norm": 3.1583240032196045,
-      "learning_rate": 6.028635418333067e-06,
-      "loss": 0.5101,
+      "epoch": 8.093857443645382,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.019436890097584e-06,
+      "loss": 0.987342300415039,
       "step": 703000
     },
     {
-      "epoch": 16.192318319977893,
-      "grad_norm": 2.8584578037261963,
-      "learning_rate": 6.021363854510551e-06,
-      "loss": 0.523,
+      "epoch": 8.096160128720095,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.012165326275069e-06,
+      "loss": 1.0043768310546874,
       "step": 703200
     },
     {
-      "epoch": 16.196923643732156,
-      "grad_norm": 2.663959503173828,
-      "learning_rate": 6.014092290688036e-06,
-      "loss": 0.5341,
+      "epoch": 8.09846281379481,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.0048937624525535e-06,
+      "loss": 0.979031753540039,
       "step": 703400
     },
     {
-      "epoch": 16.201528967486414,
-      "grad_norm": 3.1101975440979004,
-      "learning_rate": 6.00682072686552e-06,
-      "loss": 0.5324,
+      "epoch": 8.100765498869526,
+      "grad_norm": 0.75,
+      "learning_rate": 5.997622198630037e-06,
+      "loss": 0.9823381805419922,
       "step": 703600
     },
     {
-      "epoch": 16.206134291240673,
-      "grad_norm": 3.0388479232788086,
-      "learning_rate": 5.9995491630430035e-06,
-      "loss": 0.5287,
+      "epoch": 8.10306818394424,
+      "grad_norm": 1.625,
+      "learning_rate": 5.990350634807522e-06,
+      "loss": 0.9878911590576172,
       "step": 703800
     },
     {
-      "epoch": 16.210739614994935,
-      "grad_norm": 2.731867790222168,
-      "learning_rate": 5.9922775992204886e-06,
-      "loss": 0.5305,
+      "epoch": 8.105370869018955,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.983079070985006e-06,
+      "loss": 0.981031265258789,
       "step": 704000
     },
     {
-      "epoch": 16.215344938749194,
-      "grad_norm": 3.142996311187744,
-      "learning_rate": 5.985006035397973e-06,
-      "loss": 0.5323,
+      "epoch": 8.10767355409367,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.97580750716249e-06,
+      "loss": 0.9635255432128906,
       "step": 704200
     },
     {
-      "epoch": 16.219950262503453,
-      "grad_norm": 3.147834062576294,
-      "learning_rate": 5.977734471575458e-06,
-      "loss": 0.5253,
+      "epoch": 8.109976239168386,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.968535943339975e-06,
+      "loss": 0.9721682739257812,
       "step": 704400
     },
     {
-      "epoch": 16.224555586257715,
-      "grad_norm": 2.3879802227020264,
-      "learning_rate": 5.970462907752941e-06,
-      "loss": 0.5245,
+      "epoch": 8.1122789242431,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.961264379517459e-06,
+      "loss": 0.987343521118164,
       "step": 704600
     },
     {
-      "epoch": 16.229160910011974,
-      "grad_norm": 2.7794997692108154,
-      "learning_rate": 5.9631913439304254e-06,
-      "loss": 0.5124,
+      "epoch": 8.114581609317815,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.953992815694944e-06,
+      "loss": 0.9890452575683594,
       "step": 704800
     },
     {
-      "epoch": 16.233766233766232,
-      "grad_norm": 3.1171607971191406,
-      "learning_rate": 5.9559197801079105e-06,
-      "loss": 0.5293,
+      "epoch": 8.11688429439253,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.946721251872428e-06,
+      "loss": 0.9851099395751953,
       "step": 705000
     },
     {
-      "epoch": 16.238371557520495,
-      "grad_norm": 3.1548678874969482,
-      "learning_rate": 5.948648216285395e-06,
-      "loss": 0.523,
+      "epoch": 8.119186979467244,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.9394496880499115e-06,
+      "loss": 0.9777668762207031,
       "step": 705200
     },
     {
-      "epoch": 16.242976881274753,
-      "grad_norm": 3.2496373653411865,
-      "learning_rate": 5.941376652462878e-06,
-      "loss": 0.5198,
+      "epoch": 8.12148966454196,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.9321781242273966e-06,
+      "loss": 0.9775521087646485,
       "step": 705400
     },
     {
-      "epoch": 16.247582205029012,
-      "grad_norm": 3.0726122856140137,
-      "learning_rate": 5.934105088640363e-06,
-      "loss": 0.5216,
+      "epoch": 8.123792349616675,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.924906560404881e-06,
+      "loss": 0.9734044647216797,
       "step": 705600
     },
     {
-      "epoch": 16.252187528783274,
-      "grad_norm": 3.4089269638061523,
-      "learning_rate": 5.92686988263696e-06,
-      "loss": 0.5391,
+      "epoch": 8.12609503469139,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.917634996582365e-06,
+      "loss": 0.9939882659912109,
       "step": 705800
     },
     {
-      "epoch": 16.256792852537533,
-      "grad_norm": 2.813822031021118,
-      "learning_rate": 5.919598318814444e-06,
-      "loss": 0.5194,
+      "epoch": 8.128397719766104,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.910363432759849e-06,
+      "loss": 0.9817446899414063,
       "step": 706000
     },
     {
-      "epoch": 16.26139817629179,
-      "grad_norm": 3.2995693683624268,
-      "learning_rate": 5.912326754991929e-06,
-      "loss": 0.5242,
+      "epoch": 8.13070040484082,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.9030918689373334e-06,
+      "loss": 0.9877121734619141,
       "step": 706200
     },
     {
-      "epoch": 16.266003500046054,
-      "grad_norm": 3.632739543914795,
-      "learning_rate": 5.905055191169413e-06,
-      "loss": 0.5327,
+      "epoch": 8.133003089915535,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.8958203051148185e-06,
+      "loss": 0.9951105499267578,
       "step": 706400
     },
     {
-      "epoch": 16.270608823800313,
-      "grad_norm": 2.6692042350769043,
-      "learning_rate": 5.897783627346897e-06,
-      "loss": 0.5227,
+      "epoch": 8.13530577499025,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.888548741292303e-06,
+      "loss": 0.9965312194824218,
       "step": 706600
     },
     {
-      "epoch": 16.27521414755457,
-      "grad_norm": 3.4040117263793945,
-      "learning_rate": 5.890512063524382e-06,
-      "loss": 0.524,
+      "epoch": 8.137608460064964,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.881277177469787e-06,
+      "loss": 0.9727510833740234,
       "step": 706800
     },
     {
-      "epoch": 16.279819471308834,
-      "grad_norm": 2.4502665996551514,
-      "learning_rate": 5.883240499701866e-06,
-      "loss": 0.5203,
+      "epoch": 8.13991114513968,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.874005613647271e-06,
+      "loss": 0.9715268707275391,
       "step": 707000
     },
     {
-      "epoch": 16.284424795063092,
-      "grad_norm": 3.1462149620056152,
-      "learning_rate": 5.8759689358793504e-06,
-      "loss": 0.5262,
+      "epoch": 8.142213830214395,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.866734049824755e-06,
+      "loss": 0.9857926940917969,
       "step": 707200
     },
     {
-      "epoch": 16.289030118817355,
-      "grad_norm": 3.5388402938842773,
-      "learning_rate": 5.868697372056835e-06,
-      "loss": 0.5278,
+      "epoch": 8.144516515289109,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.85946248600224e-06,
+      "loss": 0.9734233093261718,
       "step": 707400
     },
     {
-      "epoch": 16.293635442571613,
-      "grad_norm": 2.5811402797698975,
-      "learning_rate": 5.861425808234319e-06,
-      "loss": 0.5255,
+      "epoch": 8.146819200363824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.852190922179725e-06,
+      "loss": 0.9755318450927735,
       "step": 707600
     },
     {
-      "epoch": 16.298240766325872,
-      "grad_norm": 2.9123117923736572,
-      "learning_rate": 5.854154244411804e-06,
-      "loss": 0.5326,
+      "epoch": 8.14912188543854,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.844919358357208e-06,
+      "loss": 0.9818879699707032,
       "step": 707800
     },
     {
-      "epoch": 16.302846090080134,
-      "grad_norm": 3.3094441890716553,
-      "learning_rate": 5.846919038408401e-06,
-      "loss": 0.5216,
+      "epoch": 8.151424570513255,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.837647794534693e-06,
+      "loss": 0.9886809539794922,
       "step": 708000
     },
     {
-      "epoch": 16.307451413834393,
-      "grad_norm": 3.1968655586242676,
-      "learning_rate": 5.839647474585885e-06,
-      "loss": 0.5238,
+      "epoch": 8.153727255587969,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.830376230712177e-06,
+      "loss": 0.981470947265625,
       "step": 708200
     },
     {
-      "epoch": 16.31205673758865,
-      "grad_norm": 3.054832935333252,
-      "learning_rate": 5.8323759107633685e-06,
-      "loss": 0.5256,
+      "epoch": 8.156029940662684,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.8231046668896616e-06,
+      "loss": 0.9918637084960937,
       "step": 708400
     },
     {
-      "epoch": 16.316662061342914,
-      "grad_norm": 2.9398386478424072,
-      "learning_rate": 5.8251043469408535e-06,
-      "loss": 0.5221,
+      "epoch": 8.1583326257374,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.815833103067146e-06,
+      "loss": 1.0077804565429687,
       "step": 708600
     },
     {
-      "epoch": 16.321267385097173,
-      "grad_norm": 2.4585120677948,
-      "learning_rate": 5.817832783118338e-06,
-      "loss": 0.5256,
+      "epoch": 8.160635310812113,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.80856153924463e-06,
+      "loss": 1.0214208984375,
       "step": 708800
     },
     {
-      "epoch": 16.32587270885143,
-      "grad_norm": 3.5368127822875977,
-      "learning_rate": 5.810561219295822e-06,
-      "loss": 0.5297,
+      "epoch": 8.162937995886828,
+      "grad_norm": 1.3359375,
+      "learning_rate": 5.801289975422114e-06,
+      "loss": 0.9646558380126953,
       "step": 709000
     },
     {
-      "epoch": 16.330478032605694,
-      "grad_norm": 3.043287515640259,
-      "learning_rate": 5.803289655473306e-06,
-      "loss": 0.5319,
+      "epoch": 8.165240680961544,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.794018411599599e-06,
+      "loss": 0.9757888031005859,
       "step": 709200
     },
     {
-      "epoch": 16.335083356359952,
-      "grad_norm": 3.1237127780914307,
-      "learning_rate": 5.79601809165079e-06,
-      "loss": 0.5316,
+      "epoch": 8.16754336603626,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.786746847777083e-06,
+      "loss": 0.9735916900634766,
       "step": 709400
     },
     {
-      "epoch": 16.33968868011421,
-      "grad_norm": 3.545804500579834,
-      "learning_rate": 5.7887465278282755e-06,
-      "loss": 0.5328,
+      "epoch": 8.169846051110973,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.779475283954568e-06,
+      "loss": 1.012173843383789,
       "step": 709600
     },
     {
-      "epoch": 16.344294003868473,
-      "grad_norm": 3.1963136196136475,
-      "learning_rate": 5.78147496400576e-06,
-      "loss": 0.5225,
+      "epoch": 8.172148736185688,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.772203720132052e-06,
+      "loss": 0.9869450378417969,
       "step": 709800
     },
     {
-      "epoch": 16.348899327622732,
-      "grad_norm": 3.5646724700927734,
-      "learning_rate": 5.774203400183243e-06,
-      "loss": 0.5219,
+      "epoch": 8.174451421260404,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.764932156309536e-06,
+      "loss": 0.9860933685302734,
       "step": 710000
     },
     {
-      "epoch": 16.35350465137699,
-      "grad_norm": 3.244006395339966,
-      "learning_rate": 5.766931836360728e-06,
-      "loss": 0.5275,
+      "epoch": 8.17675410633512,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.75766059248702e-06,
+      "loss": 0.9633110809326172,
       "step": 710200
     },
     {
-      "epoch": 16.358109975131253,
-      "grad_norm": 3.042506456375122,
-      "learning_rate": 5.759660272538212e-06,
-      "loss": 0.5267,
+      "epoch": 8.179056791409833,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.750389028664505e-06,
+      "loss": 0.9767103576660157,
       "step": 710400
     },
     {
-      "epoch": 16.36271529888551,
-      "grad_norm": 3.204932689666748,
-      "learning_rate": 5.752425066534809e-06,
-      "loss": 0.5237,
+      "epoch": 8.181359476484548,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.743117464841989e-06,
+      "loss": 0.9685611724853516,
       "step": 710600
     },
     {
-      "epoch": 16.36732062263977,
-      "grad_norm": 4.271747589111328,
-      "learning_rate": 5.7451535027122935e-06,
-      "loss": 0.5332,
+      "epoch": 8.183662161559264,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.735845901019474e-06,
+      "loss": 0.9771729278564453,
       "step": 710800
     },
     {
-      "epoch": 16.371925946394033,
-      "grad_norm": 2.5158495903015137,
-      "learning_rate": 5.737881938889778e-06,
-      "loss": 0.5368,
+      "epoch": 8.185964846633977,
+      "grad_norm": 1.0,
+      "learning_rate": 5.728574337196957e-06,
+      "loss": 0.9966614532470703,
       "step": 711000
     },
     {
-      "epoch": 16.37653127014829,
-      "grad_norm": 3.7071948051452637,
-      "learning_rate": 5.730610375067262e-06,
-      "loss": 0.5206,
+      "epoch": 8.188267531708693,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.721302773374442e-06,
+      "loss": 0.9915567779541016,
       "step": 711200
     },
     {
-      "epoch": 16.38113659390255,
-      "grad_norm": 3.480548858642578,
-      "learning_rate": 5.723338811244747e-06,
-      "loss": 0.5189,
+      "epoch": 8.190570216783408,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.7140312095519265e-06,
+      "loss": 0.9702359771728516,
       "step": 711400
     },
     {
-      "epoch": 16.385741917656812,
-      "grad_norm": 3.0035040378570557,
-      "learning_rate": 5.716067247422231e-06,
-      "loss": 0.527,
+      "epoch": 8.192872901858124,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.706759645729411e-06,
+      "loss": 0.9778258514404297,
       "step": 711600
     },
     {
-      "epoch": 16.39034724141107,
-      "grad_norm": 3.1353304386138916,
-      "learning_rate": 5.7087956835997146e-06,
-      "loss": 0.5278,
+      "epoch": 8.195175586932837,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.699488081906895e-06,
+      "loss": 0.9641788482666016,
       "step": 711800
     },
     {
-      "epoch": 16.39495256516533,
-      "grad_norm": 3.6932437419891357,
-      "learning_rate": 5.7015241197772e-06,
-      "loss": 0.5242,
+      "epoch": 8.197478272007553,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.692216518084379e-06,
+      "loss": 0.9780941009521484,
       "step": 712000
     },
     {
-      "epoch": 16.399557888919592,
-      "grad_norm": 3.175496816635132,
-      "learning_rate": 5.694252555954684e-06,
-      "loss": 0.5197,
+      "epoch": 8.199780957082268,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.684944954261863e-06,
+      "loss": 0.9986826324462891,
       "step": 712200
     },
     {
-      "epoch": 16.40416321267385,
-      "grad_norm": 3.2287960052490234,
-      "learning_rate": 5.686980992132168e-06,
-      "loss": 0.5237,
+      "epoch": 8.202083642156982,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.6776733904393485e-06,
+      "loss": 0.980475082397461,
       "step": 712400
     },
     {
-      "epoch": 16.40876853642811,
-      "grad_norm": 3.57425594329834,
-      "learning_rate": 5.679709428309652e-06,
-      "loss": 0.5292,
+      "epoch": 8.204386327231697,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.670401826616832e-06,
+      "loss": 0.9819741058349609,
       "step": 712600
     },
     {
-      "epoch": 16.41337386018237,
-      "grad_norm": 3.088691234588623,
-      "learning_rate": 5.6724378644871365e-06,
-      "loss": 0.5241,
+      "epoch": 8.206689012306413,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.663130262794316e-06,
+      "loss": 0.9808096313476562,
       "step": 712800
     },
     {
-      "epoch": 16.41797918393663,
-      "grad_norm": 2.952932119369507,
-      "learning_rate": 5.665202658483733e-06,
-      "loss": 0.5209,
+      "epoch": 8.208991697381128,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.655858698971801e-06,
+      "loss": 0.9706039428710938,
       "step": 713000
     },
     {
-      "epoch": 16.42258450769089,
-      "grad_norm": 3.005553960800171,
-      "learning_rate": 5.6579310946612185e-06,
-      "loss": 0.5168,
+      "epoch": 8.211294382455842,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.648587135149285e-06,
+      "loss": 0.9784176635742188,
       "step": 713200
     },
     {
-      "epoch": 16.42718983144515,
-      "grad_norm": 3.226658821105957,
-      "learning_rate": 5.650659530838703e-06,
-      "loss": 0.525,
+      "epoch": 8.213597067530557,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.64131557132677e-06,
+      "loss": 0.9678997039794922,
       "step": 713400
     },
     {
-      "epoch": 16.43179515519941,
-      "grad_norm": 2.5592401027679443,
-      "learning_rate": 5.643387967016186e-06,
-      "loss": 0.5251,
+      "epoch": 8.215899752605273,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.634044007504254e-06,
+      "loss": 0.9841072082519531,
       "step": 713600
     },
     {
-      "epoch": 16.436400478953672,
-      "grad_norm": 3.2084832191467285,
-      "learning_rate": 5.636116403193671e-06,
-      "loss": 0.5257,
+      "epoch": 8.218202437679988,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.626772443681738e-06,
+      "loss": 0.9815675354003907,
       "step": 713800
     },
     {
-      "epoch": 16.44100580270793,
-      "grad_norm": 3.55430006980896,
-      "learning_rate": 5.628844839371155e-06,
-      "loss": 0.5308,
+      "epoch": 8.220505122754702,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.619500879859223e-06,
+      "loss": 0.9935374450683594,
       "step": 714000
     },
     {
-      "epoch": 16.44561112646219,
-      "grad_norm": 3.086090564727783,
-      "learning_rate": 5.62157327554864e-06,
-      "loss": 0.5316,
+      "epoch": 8.222807807829417,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.612229316036707e-06,
+      "loss": 0.9976025390625,
       "step": 714200
     },
     {
-      "epoch": 16.450216450216452,
-      "grad_norm": 3.0008442401885986,
-      "learning_rate": 5.614301711726124e-06,
-      "loss": 0.5259,
+      "epoch": 8.225110492904133,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.604957752214191e-06,
+      "loss": 0.9853916931152343,
       "step": 714400
     },
     {
-      "epoch": 16.45482177397071,
-      "grad_norm": 2.9424831867218018,
-      "learning_rate": 5.607030147903608e-06,
-      "loss": 0.5265,
+      "epoch": 8.227413177978846,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.597686188391676e-06,
+      "loss": 0.9812276458740234,
       "step": 714600
     },
     {
-      "epoch": 16.45942709772497,
-      "grad_norm": 3.0486032962799072,
-      "learning_rate": 5.599758584081093e-06,
-      "loss": 0.5277,
+      "epoch": 8.229715863053562,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.59041462456916e-06,
+      "loss": 0.9623724365234375,
       "step": 714800
     },
     {
-      "epoch": 16.46403242147923,
-      "grad_norm": 2.7830045223236084,
-      "learning_rate": 5.592487020258577e-06,
-      "loss": 0.5324,
+      "epoch": 8.232018548128277,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.583143060746645e-06,
+      "loss": 0.9900458526611328,
       "step": 715000
     },
     {
-      "epoch": 16.46863774523349,
-      "grad_norm": 3.4023308753967285,
-      "learning_rate": 5.585215456436061e-06,
-      "loss": 0.5357,
+      "epoch": 8.234321233202992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.575871496924128e-06,
+      "loss": 0.995158920288086,
       "step": 715200
     },
     {
-      "epoch": 16.47324306898775,
-      "grad_norm": 3.0721235275268555,
-      "learning_rate": 5.577943892613546e-06,
-      "loss": 0.5222,
+      "epoch": 8.236623918277706,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.568599933101613e-06,
+      "loss": 0.9949196624755859,
       "step": 715400
     },
     {
-      "epoch": 16.47784839274201,
-      "grad_norm": 3.580204486846924,
-      "learning_rate": 5.57067232879103e-06,
-      "loss": 0.5206,
+      "epoch": 8.238926603352422,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.561328369279098e-06,
+      "loss": 0.9860065460205079,
       "step": 715600
     },
     {
-      "epoch": 16.48245371649627,
-      "grad_norm": 3.269157648086548,
-      "learning_rate": 5.563400764968514e-06,
-      "loss": 0.5242,
+      "epoch": 8.241229288427137,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.554056805456582e-06,
+      "loss": 0.972602767944336,
       "step": 715800
     },
     {
-      "epoch": 16.48705904025053,
-      "grad_norm": 4.157374858856201,
-      "learning_rate": 5.556129201145998e-06,
-      "loss": 0.5103,
+      "epoch": 8.24353197350185,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.546785241634065e-06,
+      "loss": 0.9766617584228515,
       "step": 716000
     },
     {
-      "epoch": 16.49166436400479,
-      "grad_norm": 3.5531797409057617,
-      "learning_rate": 5.548857637323483e-06,
-      "loss": 0.5272,
+      "epoch": 8.245834658576566,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.53951367781155e-06,
+      "loss": 0.9807737731933593,
       "step": 716200
     },
     {
-      "epoch": 16.49626968775905,
-      "grad_norm": 3.210127353668213,
-      "learning_rate": 5.541586073500967e-06,
-      "loss": 0.5267,
+      "epoch": 8.248137343651281,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.5322421139890346e-06,
+      "loss": 0.9466493988037109,
       "step": 716400
     },
     {
-      "epoch": 16.50087501151331,
-      "grad_norm": 3.227973461151123,
-      "learning_rate": 5.534314509678452e-06,
-      "loss": 0.5219,
+      "epoch": 8.250440028725997,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.52497055016652e-06,
+      "loss": 0.9830075836181641,
       "step": 716600
     },
     {
-      "epoch": 16.50548033526757,
-      "grad_norm": 3.346400499343872,
-      "learning_rate": 5.527042945855936e-06,
-      "loss": 0.532,
+      "epoch": 8.25274271380071,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.517698986344003e-06,
+      "loss": 0.9652910614013672,
       "step": 716800
     },
     {
-      "epoch": 16.51008565902183,
-      "grad_norm": 3.1097564697265625,
-      "learning_rate": 5.51977138203342e-06,
-      "loss": 0.5206,
+      "epoch": 8.255045398875426,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.510427422521487e-06,
+      "loss": 0.9870015716552735,
       "step": 717000
     },
     {
-      "epoch": 16.514690982776088,
-      "grad_norm": 3.079420328140259,
-      "learning_rate": 5.5124998182109046e-06,
-      "loss": 0.5197,
+      "epoch": 8.257348083950141,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.503155858698972e-06,
+      "loss": 0.9680076599121094,
       "step": 717200
     },
     {
-      "epoch": 16.51929630653035,
-      "grad_norm": 2.728698968887329,
-      "learning_rate": 5.505228254388389e-06,
-      "loss": 0.5102,
+      "epoch": 8.259650769024857,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.4958842948764565e-06,
+      "loss": 0.9882659912109375,
       "step": 717400
     },
     {
-      "epoch": 16.52390163028461,
-      "grad_norm": 2.969097137451172,
-      "learning_rate": 5.497993048384986e-06,
-      "loss": 0.517,
+      "epoch": 8.26195345409957,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.48861273105394e-06,
+      "loss": 0.98557373046875,
       "step": 717600
     },
     {
-      "epoch": 16.528506954038868,
-      "grad_norm": 4.207714557647705,
-      "learning_rate": 5.490721484562471e-06,
-      "loss": 0.5329,
+      "epoch": 8.264256139174286,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.481341167231425e-06,
+      "loss": 0.9993340301513672,
       "step": 717800
     },
     {
-      "epoch": 16.53311227779313,
-      "grad_norm": 3.510308027267456,
-      "learning_rate": 5.483449920739954e-06,
-      "loss": 0.5141,
+      "epoch": 8.266558824249001,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.474069603408909e-06,
+      "loss": 0.9662551879882812,
       "step": 718000
     },
     {
-      "epoch": 16.53771760154739,
-      "grad_norm": 3.0636706352233887,
-      "learning_rate": 5.476178356917438e-06,
-      "loss": 0.5215,
+      "epoch": 8.268861509323715,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.466798039586394e-06,
+      "loss": 0.9752616119384766,
       "step": 718200
     },
     {
-      "epoch": 16.542322925301647,
-      "grad_norm": 3.4108800888061523,
-      "learning_rate": 5.468943150914035e-06,
-      "loss": 0.5338,
+      "epoch": 8.27116419439843,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.459526475763878e-06,
+      "loss": 0.9850730133056641,
       "step": 718400
     },
     {
-      "epoch": 16.54692824905591,
-      "grad_norm": 3.3463311195373535,
-      "learning_rate": 5.46167158709152e-06,
-      "loss": 0.5261,
+      "epoch": 8.273466879473146,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.452254911941362e-06,
+      "loss": 0.9840998840332031,
       "step": 718600
     },
     {
-      "epoch": 16.55153357281017,
-      "grad_norm": 3.2347726821899414,
-      "learning_rate": 5.4544000232690045e-06,
-      "loss": 0.5086,
+      "epoch": 8.275769564547861,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.444983348118847e-06,
+      "loss": 0.9916070556640625,
       "step": 718800
     },
     {
-      "epoch": 16.556138896564427,
-      "grad_norm": 3.8457565307617188,
-      "learning_rate": 5.447128459446489e-06,
-      "loss": 0.5202,
+      "epoch": 8.278072249622575,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.437711784296331e-06,
+      "loss": 0.969988021850586,
       "step": 719000
     },
     {
-      "epoch": 16.56074422031869,
-      "grad_norm": 2.905299425125122,
-      "learning_rate": 5.439856895623973e-06,
-      "loss": 0.5187,
+      "epoch": 8.28037493469729,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.4304402204738145e-06,
+      "loss": 0.9657275390625,
       "step": 719200
     },
     {
-      "epoch": 16.565349544072948,
-      "grad_norm": 3.7381389141082764,
-      "learning_rate": 5.432585331801457e-06,
-      "loss": 0.5302,
+      "epoch": 8.282677619772006,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.4231686566512995e-06,
+      "loss": 1.0001145172119141,
       "step": 719400
     },
     {
-      "epoch": 16.569954867827207,
-      "grad_norm": 2.6707417964935303,
-      "learning_rate": 5.425313767978942e-06,
-      "loss": 0.534,
+      "epoch": 8.28498030484672,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.415897092828784e-06,
+      "loss": 0.9815138244628906,
       "step": 719600
     },
     {
-      "epoch": 16.57456019158147,
-      "grad_norm": 2.757823944091797,
-      "learning_rate": 5.418042204156426e-06,
-      "loss": 0.5162,
+      "epoch": 8.287282989921435,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.408625529006269e-06,
+      "loss": 0.9896107482910156,
       "step": 719800
     },
     {
-      "epoch": 16.579165515335728,
-      "grad_norm": 2.5773825645446777,
-      "learning_rate": 5.41077064033391e-06,
-      "loss": 0.527,
+      "epoch": 8.28958567499615,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.401353965183753e-06,
+      "loss": 0.9785402679443359,
       "step": 720000
     },
     {
-      "epoch": 16.583770839089986,
-      "grad_norm": 3.3267204761505127,
-      "learning_rate": 5.403499076511395e-06,
-      "loss": 0.5241,
+      "epoch": 8.291888360070866,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.394082401361236e-06,
+      "loss": 0.9890270233154297,
       "step": 720200
     },
     {
-      "epoch": 16.58837616284425,
-      "grad_norm": 2.8863368034362793,
-      "learning_rate": 5.396227512688879e-06,
-      "loss": 0.5274,
+      "epoch": 8.29419104514558,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.3868108375387215e-06,
+      "loss": 0.9779497528076172,
       "step": 720400
     },
     {
-      "epoch": 16.592981486598507,
-      "grad_norm": 3.51645827293396,
-      "learning_rate": 5.388955948866363e-06,
-      "loss": 0.5319,
+      "epoch": 8.296493730220295,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.379539273716206e-06,
+      "loss": 0.9936688995361328,
       "step": 720600
     },
     {
-      "epoch": 16.59758681035277,
-      "grad_norm": 3.1848063468933105,
-      "learning_rate": 5.381684385043848e-06,
-      "loss": 0.5271,
+      "epoch": 8.29879641529501,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.37226770989369e-06,
+      "loss": 0.9549921417236328,
       "step": 720800
     },
     {
-      "epoch": 16.602192134107028,
-      "grad_norm": 3.438476085662842,
-      "learning_rate": 5.374412821221332e-06,
-      "loss": 0.5299,
+      "epoch": 8.301099100369726,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.364996146071174e-06,
+      "loss": 0.9940463256835937,
       "step": 721000
     },
     {
-      "epoch": 16.606797457861287,
-      "grad_norm": 3.2890145778656006,
-      "learning_rate": 5.367141257398817e-06,
-      "loss": 0.5313,
+      "epoch": 8.30340178544444,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.357724582248658e-06,
+      "loss": 0.9849036407470703,
       "step": 721200
     },
     {
-      "epoch": 16.61140278161555,
-      "grad_norm": 3.603616952896118,
-      "learning_rate": 5.3598696935763e-06,
-      "loss": 0.5249,
+      "epoch": 8.305704470519155,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.3504530184261434e-06,
+      "loss": 0.9826818084716797,
       "step": 721400
     },
     {
-      "epoch": 16.616008105369808,
-      "grad_norm": 3.5745153427124023,
-      "learning_rate": 5.3525981297537845e-06,
-      "loss": 0.5234,
+      "epoch": 8.30800715559387,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.343181454603628e-06,
+      "loss": 0.9915721130371093,
       "step": 721600
     },
     {
-      "epoch": 16.620613429124067,
-      "grad_norm": 2.906506061553955,
-      "learning_rate": 5.3453265659312695e-06,
-      "loss": 0.5255,
+      "epoch": 8.310309840668584,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.335909890781111e-06,
+      "loss": 0.9841053771972657,
       "step": 721800
     },
     {
-      "epoch": 16.62521875287833,
-      "grad_norm": 3.5672848224639893,
-      "learning_rate": 5.338055002108754e-06,
-      "loss": 0.5171,
+      "epoch": 8.3126125257433,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.328638326958596e-06,
+      "loss": 0.9658685302734376,
       "step": 722000
     },
     {
-      "epoch": 16.629824076632588,
-      "grad_norm": 2.794722080230713,
-      "learning_rate": 5.330783438286238e-06,
-      "loss": 0.528,
+      "epoch": 8.314915210818015,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.32136676313608e-06,
+      "loss": 0.9773105621337891,
       "step": 722200
     },
     {
-      "epoch": 16.634429400386846,
-      "grad_norm": 2.808533191680908,
-      "learning_rate": 5.323511874463722e-06,
-      "loss": 0.5162,
+      "epoch": 8.31721789589273,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.3140951993135645e-06,
+      "loss": 0.9766971588134765,
       "step": 722400
     },
     {
-      "epoch": 16.63903472414111,
-      "grad_norm": 2.948071241378784,
-      "learning_rate": 5.316240310641206e-06,
-      "loss": 0.522,
+      "epoch": 8.319520580967444,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.306823635491049e-06,
+      "loss": 0.9907508850097656,
       "step": 722600
     },
     {
-      "epoch": 16.643640047895367,
-      "grad_norm": 3.144495964050293,
-      "learning_rate": 5.3089687468186915e-06,
-      "loss": 0.5295,
+      "epoch": 8.32182326604216,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.299552071668533e-06,
+      "loss": 0.9931290435791016,
       "step": 722800
     },
     {
-      "epoch": 16.648245371649626,
-      "grad_norm": 3.648850440979004,
-      "learning_rate": 5.301697182996175e-06,
-      "loss": 0.5313,
+      "epoch": 8.324125951116875,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.292280507846018e-06,
+      "loss": 1.004654541015625,
       "step": 723000
     },
     {
-      "epoch": 16.652850695403888,
-      "grad_norm": 3.740377902984619,
-      "learning_rate": 5.294425619173659e-06,
-      "loss": 0.5188,
+      "epoch": 8.32642863619159,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.285008944023502e-06,
+      "loss": 0.9720559692382813,
       "step": 723200
     },
     {
-      "epoch": 16.657456019158147,
-      "grad_norm": 3.226428270339966,
-      "learning_rate": 5.287154055351144e-06,
-      "loss": 0.5244,
+      "epoch": 8.328731321266304,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.277737380200986e-06,
+      "loss": 0.9656230163574219,
       "step": 723400
     },
     {
-      "epoch": 16.662061342912406,
-      "grad_norm": 3.131943941116333,
-      "learning_rate": 5.279882491528628e-06,
-      "loss": 0.5309,
+      "epoch": 8.331034006341019,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.270465816378471e-06,
+      "loss": 0.9970842742919922,
       "step": 723600
     },
     {
-      "epoch": 16.666666666666668,
-      "grad_norm": 3.345491409301758,
-      "learning_rate": 5.2726109277061126e-06,
-      "loss": 0.5319,
+      "epoch": 8.333336691415735,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.263194252555955e-06,
+      "loss": 0.9868692016601562,
       "step": 723800
     },
     {
-      "epoch": 16.671271990420927,
-      "grad_norm": 2.5959975719451904,
-      "learning_rate": 5.265339363883597e-06,
-      "loss": 0.5249,
+      "epoch": 8.335639376490448,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.255922688733439e-06,
+      "loss": 0.9869068145751954,
       "step": 724000
     },
     {
-      "epoch": 16.675877314175185,
-      "grad_norm": 2.993213415145874,
-      "learning_rate": 5.258067800061081e-06,
-      "loss": 0.5219,
+      "epoch": 8.337942061565164,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.248651124910923e-06,
+      "loss": 0.9860624694824218,
       "step": 724200
     },
     {
-      "epoch": 16.680482637929448,
-      "grad_norm": 3.5764873027801514,
-      "learning_rate": 5.250796236238566e-06,
-      "loss": 0.5264,
+      "epoch": 8.340244746639879,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.2413795610884076e-06,
+      "loss": 0.9691709899902343,
       "step": 724400
     },
     {
-      "epoch": 16.685087961683706,
-      "grad_norm": 3.0356667041778564,
-      "learning_rate": 5.24352467241605e-06,
-      "loss": 0.5139,
+      "epoch": 8.342547431714594,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.234107997265892e-06,
+      "loss": 0.9740015411376953,
       "step": 724600
     },
     {
-      "epoch": 16.689693285437965,
-      "grad_norm": 2.6434290409088135,
-      "learning_rate": 5.236253108593534e-06,
-      "loss": 0.5276,
+      "epoch": 8.344850116789308,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.226836433443377e-06,
+      "loss": 1.0019692230224608,
       "step": 724800
     },
     {
-      "epoch": 16.694298609192227,
-      "grad_norm": 3.0571916103363037,
-      "learning_rate": 5.228981544771019e-06,
-      "loss": 0.5153,
+      "epoch": 8.347152801864024,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.21956486962086e-06,
+      "loss": 0.9928291320800782,
       "step": 725000
     },
     {
-      "epoch": 16.698903932946486,
-      "grad_norm": 3.077500104904175,
-      "learning_rate": 5.221709980948503e-06,
-      "loss": 0.5254,
+      "epoch": 8.349455486938739,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.212293305798345e-06,
+      "loss": 0.9705638885498047,
       "step": 725200
     },
     {
-      "epoch": 16.703509256700745,
-      "grad_norm": 3.5419986248016357,
-      "learning_rate": 5.214438417125988e-06,
-      "loss": 0.5232,
+      "epoch": 8.351758172013453,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.2050217419758295e-06,
+      "loss": 0.9696398162841797,
       "step": 725400
     },
     {
-      "epoch": 16.708114580455007,
-      "grad_norm": 2.9213297367095947,
-      "learning_rate": 5.207166853303471e-06,
-      "loss": 0.5284,
+      "epoch": 8.354060857088168,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.197750178153314e-06,
+      "loss": 0.9995099639892578,
       "step": 725600
     },
     {
-      "epoch": 16.712719904209266,
-      "grad_norm": 3.5840189456939697,
-      "learning_rate": 5.199895289480956e-06,
-      "loss": 0.532,
+      "epoch": 8.356363542162883,
+      "grad_norm": 1.3515625,
+      "learning_rate": 5.190478614330798e-06,
+      "loss": 0.9892217254638672,
       "step": 725800
     },
     {
-      "epoch": 16.717325227963524,
-      "grad_norm": 3.3302419185638428,
-      "learning_rate": 5.1926600834775525e-06,
-      "loss": 0.5257,
+      "epoch": 8.358666227237599,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.183207050508282e-06,
+      "loss": 0.983095474243164,
       "step": 726000
     },
     {
-      "epoch": 16.721930551717787,
-      "grad_norm": 3.81793212890625,
-      "learning_rate": 5.1854248774741494e-06,
-      "loss": 0.5225,
+      "epoch": 8.360968912312313,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.175935486685766e-06,
+      "loss": 0.9840006256103515,
       "step": 726200
     },
     {
-      "epoch": 16.726535875472045,
-      "grad_norm": 3.0137243270874023,
-      "learning_rate": 5.1781533136516345e-06,
-      "loss": 0.5172,
+      "epoch": 8.363271597387028,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.1686639228632515e-06,
+      "loss": 1.0099681854248046,
       "step": 726400
     },
     {
-      "epoch": 16.731141199226304,
-      "grad_norm": 2.965055465698242,
-      "learning_rate": 5.170881749829119e-06,
-      "loss": 0.5248,
+      "epoch": 8.365574282461743,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.161392359040735e-06,
+      "loss": 0.9920525360107422,
       "step": 726600
     },
     {
-      "epoch": 16.735746522980566,
-      "grad_norm": 2.7242608070373535,
-      "learning_rate": 5.163610186006602e-06,
-      "loss": 0.5156,
+      "epoch": 8.367876967536459,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.15412079521822e-06,
+      "loss": 0.9998994445800782,
       "step": 726800
     },
     {
-      "epoch": 16.740351846734825,
-      "grad_norm": 2.787821054458618,
-      "learning_rate": 5.156338622184087e-06,
-      "loss": 0.5251,
+      "epoch": 8.370179652611172,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.146849231395704e-06,
+      "loss": 0.9936015319824218,
       "step": 727000
     },
     {
-      "epoch": 16.744957170489087,
-      "grad_norm": 3.265836238861084,
-      "learning_rate": 5.149067058361571e-06,
-      "loss": 0.5232,
+      "epoch": 8.372482337685888,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.139577667573188e-06,
+      "loss": 0.9726751708984375,
       "step": 727200
     },
     {
-      "epoch": 16.749562494243346,
-      "grad_norm": 2.891186475753784,
-      "learning_rate": 5.1417954945390564e-06,
-      "loss": 0.5294,
+      "epoch": 8.374785022760603,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.132306103750673e-06,
+      "loss": 0.9772026824951172,
       "step": 727400
     },
     {
-      "epoch": 16.754167817997605,
-      "grad_norm": 4.678770065307617,
-      "learning_rate": 5.13452393071654e-06,
-      "loss": 0.5194,
+      "epoch": 8.377087707835317,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.125034539928157e-06,
+      "loss": 0.9934384918212891,
       "step": 727600
     },
     {
-      "epoch": 16.758773141751867,
-      "grad_norm": 3.1750643253326416,
-      "learning_rate": 5.127252366894024e-06,
-      "loss": 0.5256,
+      "epoch": 8.379390392910032,
+      "grad_norm": 0.5,
+      "learning_rate": 5.117762976105641e-06,
+      "loss": 0.9848506927490235,
       "step": 727800
     },
     {
-      "epoch": 16.763378465506126,
-      "grad_norm": 4.042088508605957,
-      "learning_rate": 5.119980803071509e-06,
-      "loss": 0.5231,
+      "epoch": 8.381693077984748,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.110491412283126e-06,
+      "loss": 0.970640869140625,
       "step": 728000
     },
     {
-      "epoch": 16.767983789260384,
-      "grad_norm": 3.6129207611083984,
-      "learning_rate": 5.112709239248993e-06,
-      "loss": 0.5343,
+      "epoch": 8.383995763059463,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.10321984846061e-06,
+      "loss": 0.9867308807373046,
       "step": 728200
     },
     {
-      "epoch": 16.772589113014647,
-      "grad_norm": 2.856450319290161,
-      "learning_rate": 5.105437675426477e-06,
-      "loss": 0.5233,
+      "epoch": 8.386298448134177,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.0959482846380945e-06,
+      "loss": 0.9939826965332031,
       "step": 728400
     },
     {
-      "epoch": 16.777194436768905,
-      "grad_norm": 3.1779849529266357,
-      "learning_rate": 5.098166111603962e-06,
-      "loss": 0.5195,
+      "epoch": 8.388601133208892,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.088676720815579e-06,
+      "loss": 1.0018857574462892,
       "step": 728600
     },
     {
-      "epoch": 16.781799760523164,
-      "grad_norm": 3.1687657833099365,
-      "learning_rate": 5.090894547781446e-06,
-      "loss": 0.5315,
+      "epoch": 8.390903818283608,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.081405156993063e-06,
+      "loss": 0.9786152648925781,
       "step": 728800
     },
     {
-      "epoch": 16.786405084277426,
-      "grad_norm": 2.3541252613067627,
-      "learning_rate": 5.083622983958931e-06,
-      "loss": 0.5206,
+      "epoch": 8.393206503358321,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.074133593170548e-06,
+      "loss": 0.9753307342529297,
       "step": 729000
     },
     {
-      "epoch": 16.791010408031685,
-      "grad_norm": 3.6288070678710938,
-      "learning_rate": 5.076351420136414e-06,
-      "loss": 0.5189,
+      "epoch": 8.395509188433037,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.066862029348031e-06,
+      "loss": 0.9845266723632813,
       "step": 729200
     },
     {
-      "epoch": 16.795615731785944,
-      "grad_norm": 3.1729352474212646,
-      "learning_rate": 5.069079856313899e-06,
-      "loss": 0.5242,
+      "epoch": 8.397811873507752,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.059590465525516e-06,
+      "loss": 0.9853861236572266,
       "step": 729400
     },
     {
-      "epoch": 16.800221055540206,
-      "grad_norm": 3.039484739303589,
-      "learning_rate": 5.061808292491384e-06,
-      "loss": 0.5298,
+      "epoch": 8.400114558582468,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.052318901703001e-06,
+      "loss": 0.9771392822265625,
       "step": 729600
     },
     {
-      "epoch": 16.804826379294465,
-      "grad_norm": 3.6854867935180664,
-      "learning_rate": 5.0546094443070924e-06,
-      "loss": 0.5296,
+      "epoch": 8.402417243657181,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.045047337880485e-06,
+      "loss": 0.9864144134521484,
       "step": 729800
     },
     {
-      "epoch": 16.809431703048723,
-      "grad_norm": 3.062591791152954,
-      "learning_rate": 5.0473378804845775e-06,
-      "loss": 0.5335,
+      "epoch": 8.404719928731897,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.037775774057969e-06,
+      "loss": 0.9801872253417969,
       "step": 730000
     },
     {
-      "epoch": 16.814037026802986,
-      "grad_norm": 3.3723227977752686,
-      "learning_rate": 5.040066316662062e-06,
-      "loss": 0.5266,
+      "epoch": 8.407022613806612,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.030504210235453e-06,
+      "loss": 0.9925270843505859,
       "step": 730200
     },
     {
-      "epoch": 16.818642350557244,
-      "grad_norm": 2.7272493839263916,
-      "learning_rate": 5.032794752839545e-06,
-      "loss": 0.5222,
+      "epoch": 8.409325298881328,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.0232326464129375e-06,
+      "loss": 0.9811373138427735,
       "step": 730400
     },
     {
-      "epoch": 16.823247674311503,
-      "grad_norm": 3.0595741271972656,
-      "learning_rate": 5.02552318901703e-06,
-      "loss": 0.5252,
+      "epoch": 8.411627983956041,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.015961082590423e-06,
+      "loss": 0.9967536926269531,
       "step": 730600
     },
     {
-      "epoch": 16.827852998065765,
-      "grad_norm": 3.748103380203247,
-      "learning_rate": 5.018251625194514e-06,
-      "loss": 0.5266,
+      "epoch": 8.413930669030757,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.008689518767906e-06,
+      "loss": 1.0077542877197265,
       "step": 730800
     },
     {
-      "epoch": 16.832458321820024,
-      "grad_norm": 2.6202964782714844,
-      "learning_rate": 5.0109800613719995e-06,
-      "loss": 0.5449,
+      "epoch": 8.416233354105472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.00141795494539e-06,
+      "loss": 0.9968529510498046,
       "step": 731000
     },
     {
-      "epoch": 16.837063645574283,
-      "grad_norm": 2.5717170238494873,
-      "learning_rate": 5.003708497549483e-06,
-      "loss": 0.5277,
+      "epoch": 8.418536039180186,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.994146391122875e-06,
+      "loss": 0.9632941436767578,
       "step": 731200
     },
     {
-      "epoch": 16.841668969328545,
-      "grad_norm": 2.716874361038208,
-      "learning_rate": 4.996436933726967e-06,
-      "loss": 0.5235,
+      "epoch": 8.420838724254901,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.9868748273003595e-06,
+      "loss": 0.9997820281982421,
       "step": 731400
     },
     {
-      "epoch": 16.846274293082804,
-      "grad_norm": 3.9380040168762207,
-      "learning_rate": 4.989165369904452e-06,
-      "loss": 0.5278,
+      "epoch": 8.423141409329617,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.979603263477844e-06,
+      "loss": 0.9839828491210938,
       "step": 731600
     },
     {
-      "epoch": 16.850879616837062,
-      "grad_norm": 3.0047407150268555,
-      "learning_rate": 4.981893806081936e-06,
-      "loss": 0.5331,
+      "epoch": 8.425444094404332,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.972331699655328e-06,
+      "loss": 0.9684250640869141,
       "step": 731800
     },
     {
-      "epoch": 16.855484940591325,
-      "grad_norm": 3.2942941188812256,
-      "learning_rate": 4.97462224225942e-06,
-      "loss": 0.5394,
+      "epoch": 8.427746779479046,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.965060135832812e-06,
+      "loss": 0.9693086242675781,
       "step": 732000
     },
     {
-      "epoch": 16.860090264345583,
-      "grad_norm": 2.960007905960083,
-      "learning_rate": 4.967350678436905e-06,
-      "loss": 0.5284,
+      "epoch": 8.430049464553761,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.957788572010297e-06,
+      "loss": 0.9741654205322265,
       "step": 732200
     },
     {
-      "epoch": 16.864695588099842,
-      "grad_norm": 4.325229167938232,
-      "learning_rate": 4.960079114614389e-06,
-      "loss": 0.5142,
+      "epoch": 8.432352149628477,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.950517008187781e-06,
+      "loss": 0.9970919799804687,
       "step": 732400
     },
     {
-      "epoch": 16.869300911854104,
-      "grad_norm": 3.9659183025360107,
-      "learning_rate": 4.952807550791874e-06,
-      "loss": 0.5277,
+      "epoch": 8.434654834703192,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.943245444365265e-06,
+      "loss": 0.9704073333740234,
       "step": 732600
     },
     {
-      "epoch": 16.873906235608363,
-      "grad_norm": 4.010354042053223,
-      "learning_rate": 4.9455359869693574e-06,
-      "loss": 0.545,
+      "epoch": 8.436957519777906,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.93597388054275e-06,
+      "loss": 1.0041982269287109,
       "step": 732800
     },
     {
-      "epoch": 16.87851155936262,
-      "grad_norm": 2.861638307571411,
-      "learning_rate": 4.938264423146842e-06,
-      "loss": 0.5232,
+      "epoch": 8.439260204852621,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.928702316720234e-06,
+      "loss": 0.9678970336914062,
       "step": 733000
     },
     {
-      "epoch": 16.883116883116884,
-      "grad_norm": 3.198781967163086,
-      "learning_rate": 4.930992859324327e-06,
-      "loss": 0.5184,
+      "epoch": 8.441562889927336,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.921430752897718e-06,
+      "loss": 0.986545181274414,
       "step": 733200
     },
     {
-      "epoch": 16.887722206871143,
-      "grad_norm": 3.7329859733581543,
-      "learning_rate": 4.923721295501811e-06,
-      "loss": 0.5179,
+      "epoch": 8.44386557500205,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.9141591890752025e-06,
+      "loss": 0.9558411407470703,
       "step": 733400
     },
     {
-      "epoch": 16.8923275306254,
-      "grad_norm": 3.9860944747924805,
-      "learning_rate": 4.916449731679295e-06,
-      "loss": 0.5298,
+      "epoch": 8.446168260076766,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.906887625252687e-06,
+      "loss": 0.9802081298828125,
       "step": 733600
     },
     {
-      "epoch": 16.896932854379664,
-      "grad_norm": 3.4595282077789307,
-      "learning_rate": 4.909178167856779e-06,
-      "loss": 0.5261,
+      "epoch": 8.448470945151481,
+      "grad_norm": 0.75,
+      "learning_rate": 4.899616061430172e-06,
+      "loss": 0.9869933319091797,
       "step": 733800
     },
     {
-      "epoch": 16.901538178133922,
-      "grad_norm": 3.319841146469116,
-      "learning_rate": 4.901906604034264e-06,
-      "loss": 0.5225,
+      "epoch": 8.450773630226196,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.892344497607656e-06,
+      "loss": 0.9878524780273438,
       "step": 734000
     },
     {
-      "epoch": 16.90614350188818,
-      "grad_norm": 3.025320529937744,
-      "learning_rate": 4.894635040211749e-06,
-      "loss": 0.5423,
+      "epoch": 8.45307631530091,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.885072933785139e-06,
+      "loss": 0.9893035125732422,
       "step": 734200
     },
     {
-      "epoch": 16.910748825642443,
-      "grad_norm": 3.0882341861724854,
-      "learning_rate": 4.887363476389233e-06,
-      "loss": 0.5239,
+      "epoch": 8.455379000375626,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.8778013699626245e-06,
+      "loss": 0.9825079345703125,
       "step": 734400
     },
     {
-      "epoch": 16.915354149396702,
-      "grad_norm": 3.2497994899749756,
-      "learning_rate": 4.880091912566716e-06,
-      "loss": 0.5168,
+      "epoch": 8.457681685450341,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.870529806140109e-06,
+      "loss": 0.9930937194824219,
       "step": 734600
     },
     {
-      "epoch": 16.919959473150964,
-      "grad_norm": 3.428805351257324,
-      "learning_rate": 4.872820348744201e-06,
-      "loss": 0.5209,
+      "epoch": 8.459984370525055,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.863258242317593e-06,
+      "loss": 0.961860580444336,
       "step": 734800
     },
     {
-      "epoch": 16.924564796905223,
-      "grad_norm": 3.8149008750915527,
-      "learning_rate": 4.8655487849216855e-06,
-      "loss": 0.5226,
+      "epoch": 8.46228705559977,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.855986678495077e-06,
+      "loss": 1.0016260528564453,
       "step": 735000
     },
     {
-      "epoch": 16.92917012065948,
-      "grad_norm": 2.430985450744629,
-      "learning_rate": 4.85827722109917e-06,
-      "loss": 0.5272,
+      "epoch": 8.464589740674485,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.848715114672561e-06,
+      "loss": 0.9752246856689453,
       "step": 735200
     },
     {
-      "epoch": 16.933775444413744,
-      "grad_norm": 3.142275333404541,
-      "learning_rate": 4.851005657276654e-06,
-      "loss": 0.5294,
+      "epoch": 8.4668924257492,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.841443550850046e-06,
+      "loss": 0.9766324615478515,
       "step": 735400
     },
     {
-      "epoch": 16.938380768168003,
-      "grad_norm": 3.180042028427124,
-      "learning_rate": 4.843734093454138e-06,
-      "loss": 0.5203,
+      "epoch": 8.469195110823915,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.834171987027531e-06,
+      "loss": 0.9969155883789063,
       "step": 735600
     },
     {
-      "epoch": 16.94298609192226,
-      "grad_norm": 3.004331350326538,
-      "learning_rate": 4.836462529631623e-06,
-      "loss": 0.5244,
+      "epoch": 8.47149779589863,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.826900423205014e-06,
+      "loss": 0.9860860443115235,
       "step": 735800
     },
     {
-      "epoch": 16.947591415676523,
-      "grad_norm": 2.585542678833008,
-      "learning_rate": 4.8291909658091075e-06,
-      "loss": 0.5289,
+      "epoch": 8.473800480973345,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.819628859382499e-06,
+      "loss": 0.9745457458496094,
       "step": 736000
     },
     {
-      "epoch": 16.952196739430782,
-      "grad_norm": 3.3415260314941406,
-      "learning_rate": 4.821919401986591e-06,
-      "loss": 0.5253,
+      "epoch": 8.476103166048059,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.812357295559983e-06,
+      "loss": 1.0002130889892578,
       "step": 736200
     },
     {
-      "epoch": 16.95680206318504,
-      "grad_norm": 3.3344316482543945,
-      "learning_rate": 4.814647838164076e-06,
-      "loss": 0.5234,
+      "epoch": 8.478405851122774,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.8050857317374675e-06,
+      "loss": 0.98840087890625,
       "step": 736400
     },
     {
-      "epoch": 16.961407386939303,
-      "grad_norm": 2.819533348083496,
-      "learning_rate": 4.807412632160673e-06,
-      "loss": 0.5313,
+      "epoch": 8.48070853619749,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.797814167914952e-06,
+      "loss": 1.0051963806152344,
       "step": 736600
     },
     {
-      "epoch": 16.966012710693562,
-      "grad_norm": 3.417186737060547,
-      "learning_rate": 4.80017742615727e-06,
-      "loss": 0.5333,
+      "epoch": 8.483011221272205,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.790542604092436e-06,
+      "loss": 1.006302032470703,
       "step": 736800
     },
     {
-      "epoch": 16.97061803444782,
-      "grad_norm": 3.4736814498901367,
-      "learning_rate": 4.792905862334754e-06,
-      "loss": 0.5133,
+      "epoch": 8.485313906346919,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.783271040269921e-06,
+      "loss": 0.9928652954101562,
       "step": 737000
     },
     {
-      "epoch": 16.975223358202083,
-      "grad_norm": 2.8336431980133057,
-      "learning_rate": 4.785634298512238e-06,
-      "loss": 0.5268,
+      "epoch": 8.487616591421634,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.775999476447405e-06,
+      "loss": 0.9688255310058593,
       "step": 737200
     },
     {
-      "epoch": 16.97982868195634,
-      "grad_norm": 3.896120071411133,
-      "learning_rate": 4.778362734689722e-06,
-      "loss": 0.5304,
+      "epoch": 8.48991927649635,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.768727912624889e-06,
+      "loss": 0.9724116516113281,
       "step": 737400
     },
     {
-      "epoch": 16.9844340057106,
-      "grad_norm": 3.212686777114868,
-      "learning_rate": 4.771127528686319e-06,
-      "loss": 0.5216,
+      "epoch": 8.492221961571065,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.761456348802374e-06,
+      "loss": 0.9784508514404296,
       "step": 737600
     },
     {
-      "epoch": 16.989039329464863,
-      "grad_norm": 2.8021464347839355,
-      "learning_rate": 4.7638559648638035e-06,
-      "loss": 0.5258,
+      "epoch": 8.494524646645779,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.754184784979858e-06,
+      "loss": 0.9876103973388672,
       "step": 737800
     },
     {
-      "epoch": 16.99364465321912,
-      "grad_norm": 2.3054471015930176,
-      "learning_rate": 4.756584401041289e-06,
-      "loss": 0.5119,
+      "epoch": 8.496827331720494,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.746913221157342e-06,
+      "loss": 0.9536474609375,
       "step": 738000
     },
     {
-      "epoch": 16.99824997697338,
-      "grad_norm": 3.4670674800872803,
-      "learning_rate": 4.749312837218773e-06,
-      "loss": 0.5228,
+      "epoch": 8.49913001679521,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.739641657334826e-06,
+      "loss": 1.0035865020751953,
       "step": 738200
     },
     {
-      "epoch": 17.0,
-      "eval_loss": 0.513902485370636,
-      "eval_runtime": 169.6655,
-      "eval_samples_per_second": 167.158,
-      "eval_steps_per_second": 10.45,
-      "step": 738276
-    },
-    {
-      "epoch": 17.002855300727642,
-      "grad_norm": 4.795275688171387,
-      "learning_rate": 4.74207763121537e-06,
-      "loss": 0.5208,
+      "epoch": 8.501432701869923,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.7323700935123106e-06,
+      "loss": 0.9735746765136719,
       "step": 738400
     },
     {
-      "epoch": 17.0074606244819,
-      "grad_norm": 2.918687582015991,
-      "learning_rate": 4.734806067392853e-06,
-      "loss": 0.5252,
+      "epoch": 8.503735386944639,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.725098529689796e-06,
+      "loss": 0.9690022277832031,
       "step": 738600
     },
     {
-      "epoch": 17.01206594823616,
-      "grad_norm": 2.214686632156372,
-      "learning_rate": 4.727534503570338e-06,
-      "loss": 0.5238,
+      "epoch": 8.506038072019354,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.71782696586728e-06,
+      "loss": 0.9983604431152344,
       "step": 738800
     },
     {
-      "epoch": 17.016671271990422,
-      "grad_norm": 3.4804065227508545,
-      "learning_rate": 4.720262939747822e-06,
-      "loss": 0.5244,
+      "epoch": 8.50834075709407,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.710555402044763e-06,
+      "loss": 0.9797876739501953,
       "step": 739000
     },
     {
-      "epoch": 17.02127659574468,
-      "grad_norm": 2.5966029167175293,
-      "learning_rate": 4.712991375925307e-06,
-      "loss": 0.5301,
+      "epoch": 8.510643442168783,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.703283838222248e-06,
+      "loss": 0.9583721923828125,
       "step": 739200
     },
     {
-      "epoch": 17.02588191949894,
-      "grad_norm": 2.7424144744873047,
-      "learning_rate": 4.705719812102791e-06,
-      "loss": 0.5181,
+      "epoch": 8.512946127243499,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.6960122743997325e-06,
+      "loss": 0.9663965606689453,
       "step": 739400
     },
     {
-      "epoch": 17.0304872432532,
-      "grad_norm": 2.9380509853363037,
-      "learning_rate": 4.698448248280275e-06,
-      "loss": 0.5324,
+      "epoch": 8.515248812318214,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.688740710577217e-06,
+      "loss": 0.9920121002197265,
       "step": 739600
     },
     {
-      "epoch": 17.03509256700746,
-      "grad_norm": 2.6877858638763428,
-      "learning_rate": 4.69117668445776e-06,
-      "loss": 0.5186,
+      "epoch": 8.51755149739293,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.681469146754701e-06,
+      "loss": 0.9966871643066406,
       "step": 739800
     },
     {
-      "epoch": 17.03969789076172,
-      "grad_norm": 2.7260663509368896,
-      "learning_rate": 4.683905120635244e-06,
-      "loss": 0.5212,
+      "epoch": 8.519854182467643,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.674197582932185e-06,
+      "loss": 0.9799186706542968,
       "step": 740000
     },
     {
-      "epoch": 17.04430321451598,
-      "grad_norm": 3.156268358230591,
-      "learning_rate": 4.676633556812728e-06,
-      "loss": 0.5262,
+      "epoch": 8.522156867542359,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.66692601910967e-06,
+      "loss": 1.0063607788085938,
       "step": 740200
     },
     {
-      "epoch": 17.04890853827024,
-      "grad_norm": 4.205811500549316,
-      "learning_rate": 4.669361992990213e-06,
-      "loss": 0.5204,
+      "epoch": 8.524459552617074,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.6596544552871544e-06,
+      "loss": 0.976033935546875,
       "step": 740400
     },
     {
-      "epoch": 17.0535138620245,
-      "grad_norm": 3.210843563079834,
-      "learning_rate": 4.662090429167697e-06,
-      "loss": 0.5324,
+      "epoch": 8.526762237691788,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.652382891464639e-06,
+      "loss": 0.964898681640625,
       "step": 740600
     },
     {
-      "epoch": 17.05811918577876,
-      "grad_norm": 2.8761239051818848,
-      "learning_rate": 4.654818865345181e-06,
-      "loss": 0.5184,
+      "epoch": 8.529064922766503,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.645111327642123e-06,
+      "loss": 0.9735550689697265,
       "step": 740800
     },
     {
-      "epoch": 17.06272450953302,
-      "grad_norm": 2.9451096057891846,
-      "learning_rate": 4.647547301522665e-06,
-      "loss": 0.5132,
+      "epoch": 8.531367607841219,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.637839763819607e-06,
+      "loss": 0.9649169921875,
       "step": 741000
     },
     {
-      "epoch": 17.067329833287282,
-      "grad_norm": 2.4379804134368896,
-      "learning_rate": 4.64027573770015e-06,
-      "loss": 0.5156,
+      "epoch": 8.533670292915934,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.630568199997091e-06,
+      "loss": 0.96186279296875,
       "step": 741200
     },
     {
-      "epoch": 17.07193515704154,
-      "grad_norm": 2.7910115718841553,
-      "learning_rate": 4.633004173877635e-06,
-      "loss": 0.5234,
+      "epoch": 8.535972977990648,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.623296636174576e-06,
+      "loss": 1.0031502532958985,
       "step": 741400
     },
     {
-      "epoch": 17.0765404807958,
-      "grad_norm": 3.084688663482666,
-      "learning_rate": 4.625732610055119e-06,
-      "loss": 0.5324,
+      "epoch": 8.538275663065363,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.61602507235206e-06,
+      "loss": 0.9780646514892578,
       "step": 741600
     },
     {
-      "epoch": 17.08114580455006,
-      "grad_norm": 2.9508092403411865,
-      "learning_rate": 4.618461046232602e-06,
-      "loss": 0.512,
+      "epoch": 8.540578348140079,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.608753508529545e-06,
+      "loss": 0.9804470825195313,
       "step": 741800
     },
     {
-      "epoch": 17.08575112830432,
-      "grad_norm": 4.045099258422852,
-      "learning_rate": 4.611225840229199e-06,
-      "loss": 0.5115,
+      "epoch": 8.542881033214792,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.601481944707029e-06,
+      "loss": 0.9616146087646484,
       "step": 742000
     },
     {
-      "epoch": 17.09035645205858,
-      "grad_norm": 3.059542655944824,
-      "learning_rate": 4.603954276406684e-06,
-      "loss": 0.5239,
+      "epoch": 8.545183718289508,
+      "grad_norm": 0.875,
+      "learning_rate": 4.594210380884513e-06,
+      "loss": 0.9716793060302734,
       "step": 742200
     },
     {
-      "epoch": 17.09496177581284,
-      "grad_norm": 3.1609458923339844,
-      "learning_rate": 4.5966827125841685e-06,
-      "loss": 0.5215,
+      "epoch": 8.547486403364223,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.5869388170619975e-06,
+      "loss": 0.9809304809570313,
       "step": 742400
     },
     {
-      "epoch": 17.0995670995671,
-      "grad_norm": 3.208566427230835,
-      "learning_rate": 4.589411148761653e-06,
-      "loss": 0.5125,
+      "epoch": 8.549789088438938,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.579667253239482e-06,
+      "loss": 0.9804471588134765,
       "step": 742600
     },
     {
-      "epoch": 17.10417242332136,
-      "grad_norm": 3.126858949661255,
-      "learning_rate": 4.582139584939137e-06,
-      "loss": 0.5145,
+      "epoch": 8.552091773513652,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.572395689416966e-06,
+      "loss": 0.9672725677490235,
       "step": 742800
     },
     {
-      "epoch": 17.10877774707562,
-      "grad_norm": 2.690805673599243,
-      "learning_rate": 4.574868021116621e-06,
-      "loss": 0.533,
+      "epoch": 8.554394458588368,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.565124125594451e-06,
+      "loss": 1.0129902648925782,
       "step": 743000
     },
     {
-      "epoch": 17.11338307082988,
-      "grad_norm": 3.323164939880371,
-      "learning_rate": 4.567596457294106e-06,
-      "loss": 0.5194,
+      "epoch": 8.556697143663083,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.557852561771934e-06,
+      "loss": 0.9989830017089844,
       "step": 743200
     },
     {
-      "epoch": 17.11798839458414,
-      "grad_norm": 2.800487995147705,
-      "learning_rate": 4.5603248934715904e-06,
-      "loss": 0.5217,
+      "epoch": 8.558999828737797,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.550580997949419e-06,
+      "loss": 0.9873731994628906,
       "step": 743400
     },
     {
-      "epoch": 17.1225937183384,
-      "grad_norm": 3.2171924114227295,
-      "learning_rate": 4.553053329649074e-06,
-      "loss": 0.5251,
+      "epoch": 8.561302513812512,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.543309434126904e-06,
+      "loss": 0.9832039642333984,
       "step": 743600
     },
     {
-      "epoch": 17.12719904209266,
-      "grad_norm": 2.8016812801361084,
-      "learning_rate": 4.545781765826559e-06,
-      "loss": 0.516,
+      "epoch": 8.563605198887227,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.536037870304388e-06,
+      "loss": 0.9728511810302735,
       "step": 743800
     },
     {
-      "epoch": 17.131804365846918,
-      "grad_norm": 2.8154873847961426,
-      "learning_rate": 4.538510202004043e-06,
-      "loss": 0.5167,
+      "epoch": 8.565907883961943,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.528766306481872e-06,
+      "loss": 0.9821251678466797,
       "step": 744000
     },
     {
-      "epoch": 17.13640968960118,
-      "grad_norm": 3.5362391471862793,
-      "learning_rate": 4.531238638181527e-06,
-      "loss": 0.5388,
+      "epoch": 8.568210569036657,
+      "grad_norm": 0.75,
+      "learning_rate": 4.521494742659356e-06,
+      "loss": 0.9702672576904297,
       "step": 744200
     },
     {
-      "epoch": 17.14101501335544,
-      "grad_norm": 3.001676321029663,
-      "learning_rate": 4.523967074359012e-06,
-      "loss": 0.5284,
+      "epoch": 8.570513254111372,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.5142231788368405e-06,
+      "loss": 0.9961418151855469,
       "step": 744400
     },
     {
-      "epoch": 17.145620337109698,
-      "grad_norm": 2.9126548767089844,
-      "learning_rate": 4.516695510536496e-06,
-      "loss": 0.5217,
+      "epoch": 8.572815939186087,
+      "grad_norm": 1.125,
+      "learning_rate": 4.506951615014326e-06,
+      "loss": 0.9817313385009766,
       "step": 744600
     },
     {
-      "epoch": 17.15022566086396,
-      "grad_norm": 2.922006368637085,
-      "learning_rate": 4.509423946713981e-06,
-      "loss": 0.5176,
+      "epoch": 8.575118624260803,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.499680051191809e-06,
+      "loss": 0.985406494140625,
       "step": 744800
     },
     {
-      "epoch": 17.15483098461822,
-      "grad_norm": 3.5556321144104004,
-      "learning_rate": 4.502152382891465e-06,
-      "loss": 0.5301,
+      "epoch": 8.577421309335516,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.492408487369294e-06,
+      "loss": 0.9726136779785156,
       "step": 745000
     },
     {
-      "epoch": 17.159436308372477,
-      "grad_norm": 2.726541519165039,
-      "learning_rate": 4.494880819068949e-06,
-      "loss": 0.5158,
+      "epoch": 8.579723994410232,
+      "grad_norm": 1.3671875,
+      "learning_rate": 4.485136923546778e-06,
+      "loss": 0.9817514038085937,
       "step": 745200
     },
     {
-      "epoch": 17.16404163212674,
-      "grad_norm": 2.8062782287597656,
-      "learning_rate": 4.4876092552464335e-06,
-      "loss": 0.5243,
+      "epoch": 8.582026679484947,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.4778653597242625e-06,
+      "loss": 0.9874824523925781,
       "step": 745400
     },
     {
-      "epoch": 17.168646955880998,
-      "grad_norm": 2.6974618434906006,
-      "learning_rate": 4.480337691423918e-06,
-      "loss": 0.5189,
+      "epoch": 8.584329364559661,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.470593795901747e-06,
+      "loss": 0.9865628051757812,
       "step": 745600
     },
     {
-      "epoch": 17.173252279635257,
-      "grad_norm": 4.060593128204346,
-      "learning_rate": 4.473066127601402e-06,
-      "loss": 0.514,
+      "epoch": 8.586632049634376,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.463322232079231e-06,
+      "loss": 0.9929932403564453,
       "step": 745800
     },
     {
-      "epoch": 17.17785760338952,
-      "grad_norm": 2.953789472579956,
-      "learning_rate": 4.465794563778887e-06,
-      "loss": 0.5274,
+      "epoch": 8.588934734709092,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.456050668256715e-06,
+      "loss": 1.007845687866211,
       "step": 746000
     },
     {
-      "epoch": 17.182462927143778,
-      "grad_norm": 2.9467928409576416,
-      "learning_rate": 4.45852299995637e-06,
-      "loss": 0.5094,
+      "epoch": 8.591237419783807,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.4487791044342e-06,
+      "loss": 0.9824122619628907,
       "step": 746200
     },
     {
-      "epoch": 17.187068250898037,
-      "grad_norm": 3.19561505317688,
-      "learning_rate": 4.4512514361338546e-06,
-      "loss": 0.5341,
+      "epoch": 8.593540104858521,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.4415075406116836e-06,
+      "loss": 0.9818960571289063,
       "step": 746400
     },
     {
-      "epoch": 17.1916735746523,
-      "grad_norm": 2.521677017211914,
-      "learning_rate": 4.44397987231134e-06,
-      "loss": 0.525,
+      "epoch": 8.595842789933236,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.434235976789168e-06,
+      "loss": 0.9791338348388672,
       "step": 746600
     },
     {
-      "epoch": 17.196278898406558,
-      "grad_norm": 3.030155897140503,
-      "learning_rate": 4.436708308488824e-06,
-      "loss": 0.5327,
+      "epoch": 8.598145475007952,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.426964412966653e-06,
+      "loss": 0.9939133453369141,
       "step": 746800
     },
     {
-      "epoch": 17.200884222160816,
-      "grad_norm": 3.0446267127990723,
-      "learning_rate": 4.429436744666308e-06,
-      "loss": 0.5185,
+      "epoch": 8.600448160082667,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.419692849144137e-06,
+      "loss": 0.9964035797119141,
       "step": 747000
     },
     {
-      "epoch": 17.20548954591508,
-      "grad_norm": 3.0444605350494385,
-      "learning_rate": 4.422165180843792e-06,
-      "loss": 0.5256,
+      "epoch": 8.60275084515738,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.412421285321621e-06,
+      "loss": 0.9719769287109375,
       "step": 747200
     },
     {
-      "epoch": 17.210094869669337,
-      "grad_norm": 3.9111011028289795,
-      "learning_rate": 4.4148936170212765e-06,
-      "loss": 0.5247,
+      "epoch": 8.605053530232096,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.4051497214991055e-06,
+      "loss": 0.9511328887939453,
       "step": 747400
     },
     {
-      "epoch": 17.214700193423596,
-      "grad_norm": 3.1988956928253174,
-      "learning_rate": 4.407622053198762e-06,
-      "loss": 0.5177,
+      "epoch": 8.607356215306812,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.39787815767659e-06,
+      "loss": 0.987777099609375,
       "step": 747600
     },
     {
-      "epoch": 17.219305517177858,
-      "grad_norm": 3.0281171798706055,
-      "learning_rate": 4.400350489376245e-06,
-      "loss": 0.5212,
+      "epoch": 8.609658900381525,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.390606593854075e-06,
+      "loss": 0.99900634765625,
       "step": 747800
     },
     {
-      "epoch": 17.223910840932117,
-      "grad_norm": 3.074202060699463,
-      "learning_rate": 4.393078925553729e-06,
-      "loss": 0.5171,
+      "epoch": 8.61196158545624,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.383335030031559e-06,
+      "loss": 0.9555909729003906,
       "step": 748000
     },
     {
-      "epoch": 17.22851616468638,
-      "grad_norm": 3.3265304565429688,
-      "learning_rate": 4.385807361731214e-06,
-      "loss": 0.5188,
+      "epoch": 8.614264270530956,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.376063466209042e-06,
+      "loss": 0.9895748901367187,
       "step": 748200
     },
     {
-      "epoch": 17.233121488440638,
-      "grad_norm": 3.696972370147705,
-      "learning_rate": 4.378572155727811e-06,
-      "loss": 0.5163,
+      "epoch": 8.616566955605672,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.3687919023865274e-06,
+      "loss": 0.9897019958496094,
       "step": 748400
     },
     {
-      "epoch": 17.237726812194897,
-      "grad_norm": 2.553654193878174,
-      "learning_rate": 4.371300591905295e-06,
-      "loss": 0.5148,
+      "epoch": 8.618869640680385,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.361520338564012e-06,
+      "loss": 0.9821299743652344,
       "step": 748600
     },
     {
-      "epoch": 17.24233213594916,
-      "grad_norm": 3.0239148139953613,
-      "learning_rate": 4.36402902808278e-06,
-      "loss": 0.5086,
+      "epoch": 8.6211723257551,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.354248774741497e-06,
+      "loss": 0.9836980438232422,
       "step": 748800
     },
     {
-      "epoch": 17.246937459703418,
-      "grad_norm": 3.5661909580230713,
-      "learning_rate": 4.356757464260264e-06,
-      "loss": 0.5203,
+      "epoch": 8.623475010829816,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.34697721091898e-06,
+      "loss": 0.9656773376464843,
       "step": 749000
     },
     {
-      "epoch": 17.251542783457676,
-      "grad_norm": 3.1638379096984863,
-      "learning_rate": 4.349485900437748e-06,
-      "loss": 0.5089,
+      "epoch": 8.625777695904532,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.339705647096464e-06,
+      "loss": 0.9764061737060546,
       "step": 749200
     },
     {
-      "epoch": 17.25614810721194,
-      "grad_norm": 3.8323419094085693,
-      "learning_rate": 4.342214336615233e-06,
-      "loss": 0.527,
+      "epoch": 8.628080380979245,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.332434083273949e-06,
+      "loss": 0.9941240692138672,
       "step": 749400
     },
     {
-      "epoch": 17.260753430966197,
-      "grad_norm": 2.8561997413635254,
-      "learning_rate": 4.3349427727927165e-06,
-      "loss": 0.5227,
+      "epoch": 8.63038306605396,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.325162519451434e-06,
+      "loss": 0.9832437896728515,
       "step": 749600
     },
     {
-      "epoch": 17.265358754720456,
-      "grad_norm": 2.703752040863037,
-      "learning_rate": 4.327671208970201e-06,
-      "loss": 0.5293,
+      "epoch": 8.632685751128676,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.317890955628917e-06,
+      "loss": 0.9863040161132812,
       "step": 749800
     },
     {
-      "epoch": 17.269964078474718,
-      "grad_norm": 2.9282515048980713,
-      "learning_rate": 4.320399645147686e-06,
-      "loss": 0.5195,
+      "epoch": 8.63498843620339,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.310619391806402e-06,
+      "loss": 0.9784996795654297,
       "step": 750000
     },
     {
-      "epoch": 17.274569402228977,
-      "grad_norm": 3.2129111289978027,
-      "learning_rate": 4.31312808132517e-06,
-      "loss": 0.5199,
+      "epoch": 8.637291121278105,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.303347827983886e-06,
+      "loss": 0.9644253540039063,
       "step": 750200
     },
     {
-      "epoch": 17.279174725983236,
-      "grad_norm": 3.368345260620117,
-      "learning_rate": 4.305856517502654e-06,
-      "loss": 0.5246,
+      "epoch": 8.63959380635282,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.296076264161371e-06,
+      "loss": 0.9938349151611328,
       "step": 750400
     },
     {
-      "epoch": 17.283780049737498,
-      "grad_norm": 3.3532567024230957,
-      "learning_rate": 4.298621311499251e-06,
-      "loss": 0.5149,
+      "epoch": 8.641896491427536,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.288804700338855e-06,
+      "loss": 0.9900470733642578,
       "step": 750600
     },
     {
-      "epoch": 17.288385373491757,
-      "grad_norm": 3.7765021324157715,
-      "learning_rate": 4.291349747676735e-06,
-      "loss": 0.5282,
+      "epoch": 8.64419917650225,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.281533136516339e-06,
+      "loss": 0.992398452758789,
       "step": 750800
     },
     {
-      "epoch": 17.292990697246015,
-      "grad_norm": 2.6797635555267334,
-      "learning_rate": 4.2840781838542195e-06,
-      "loss": 0.5261,
+      "epoch": 8.646501861576965,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.274261572693824e-06,
+      "loss": 0.9773881530761719,
       "step": 751000
     },
     {
-      "epoch": 17.297596021000277,
-      "grad_norm": 2.490145683288574,
-      "learning_rate": 4.276806620031705e-06,
-      "loss": 0.5254,
+      "epoch": 8.64880454665168,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.266990008871308e-06,
+      "loss": 0.9816563415527344,
       "step": 751200
     },
     {
-      "epoch": 17.302201344754536,
-      "grad_norm": 2.785992383956909,
-      "learning_rate": 4.269535056209188e-06,
-      "loss": 0.5196,
+      "epoch": 8.651107231726394,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.259718445048792e-06,
+      "loss": 0.9742191314697266,
       "step": 751400
     },
     {
-      "epoch": 17.306806668508795,
-      "grad_norm": 3.1607396602630615,
-      "learning_rate": 4.262263492386672e-06,
-      "loss": 0.5217,
+      "epoch": 8.65340991680111,
+      "grad_norm": 0.75,
+      "learning_rate": 4.252446881226277e-06,
+      "loss": 0.983070297241211,
       "step": 751600
     },
     {
-      "epoch": 17.311411992263057,
-      "grad_norm": 3.664815902709961,
-      "learning_rate": 4.254991928564157e-06,
-      "loss": 0.5251,
+      "epoch": 8.655712601875825,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.245175317403761e-06,
+      "loss": 0.9570191955566406,
       "step": 751800
     },
     {
-      "epoch": 17.316017316017316,
-      "grad_norm": 3.4818427562713623,
-      "learning_rate": 4.247756722560754e-06,
-      "loss": 0.5282,
+      "epoch": 8.65801528695054,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.237903753581246e-06,
+      "loss": 0.9718711853027344,
       "step": 752000
     },
     {
-      "epoch": 17.320622639771575,
-      "grad_norm": 3.277512311935425,
-      "learning_rate": 4.240521516557351e-06,
-      "loss": 0.5343,
+      "epoch": 8.660317972025254,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.230632189758729e-06,
+      "loss": 0.9831970977783203,
       "step": 752200
     },
     {
-      "epoch": 17.325227963525837,
-      "grad_norm": 3.601278305053711,
-      "learning_rate": 4.233249952734835e-06,
-      "loss": 0.527,
+      "epoch": 8.66262065709997,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.2233606259362135e-06,
+      "loss": 0.9732750701904297,
       "step": 752400
     },
     {
-      "epoch": 17.329833287280096,
-      "grad_norm": 3.438324451446533,
-      "learning_rate": 4.2259783889123195e-06,
-      "loss": 0.5108,
+      "epoch": 8.664923342174685,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.216089062113699e-06,
+      "loss": 0.9819364166259765,
       "step": 752600
     },
     {
-      "epoch": 17.334438611034354,
-      "grad_norm": 3.622999429702759,
-      "learning_rate": 4.218706825089804e-06,
-      "loss": 0.522,
+      "epoch": 8.667226027249399,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.208817498291183e-06,
+      "loss": 1.0118328094482423,
       "step": 752800
     },
     {
-      "epoch": 17.339043934788616,
-      "grad_norm": 2.718357563018799,
-      "learning_rate": 4.211435261267288e-06,
-      "loss": 0.5287,
+      "epoch": 8.669528712324114,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.201545934468666e-06,
+      "loss": 0.9904637145996094,
       "step": 753000
     },
     {
-      "epoch": 17.343649258542875,
-      "grad_norm": 3.1246633529663086,
-      "learning_rate": 4.204163697444773e-06,
-      "loss": 0.5293,
+      "epoch": 8.67183139739883,
+      "grad_norm": 0.75,
+      "learning_rate": 4.194274370646151e-06,
+      "loss": 0.972322769165039,
       "step": 753200
     },
     {
-      "epoch": 17.348254582297134,
-      "grad_norm": 2.551449775695801,
-      "learning_rate": 4.196892133622257e-06,
-      "loss": 0.5222,
+      "epoch": 8.674134082473545,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.1870028068236355e-06,
+      "loss": 0.9820072174072265,
       "step": 753400
     },
     {
-      "epoch": 17.352859906051396,
-      "grad_norm": 2.530979633331299,
-      "learning_rate": 4.189620569799741e-06,
-      "loss": 0.5185,
+      "epoch": 8.676436767548259,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.1797312430011205e-06,
+      "loss": 0.9733422088623047,
       "step": 753600
     },
     {
-      "epoch": 17.357465229805655,
-      "grad_norm": 3.4675188064575195,
-      "learning_rate": 4.182349005977226e-06,
-      "loss": 0.5193,
+      "epoch": 8.678739452622974,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.172459679178604e-06,
+      "loss": 0.9668608856201172,
       "step": 753800
     },
     {
-      "epoch": 17.362070553559914,
-      "grad_norm": 3.597015619277954,
-      "learning_rate": 4.17507744215471e-06,
-      "loss": 0.5124,
+      "epoch": 8.68104213769769,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.165188115356088e-06,
+      "loss": 0.9491535186767578,
       "step": 754000
     },
     {
-      "epoch": 17.366675877314176,
-      "grad_norm": 3.48524808883667,
-      "learning_rate": 4.167805878332195e-06,
-      "loss": 0.5167,
+      "epoch": 8.683344822772405,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.157916551533573e-06,
+      "loss": 0.972203369140625,
       "step": 754200
     },
     {
-      "epoch": 17.371281201068435,
-      "grad_norm": 2.2207281589508057,
-      "learning_rate": 4.160534314509678e-06,
-      "loss": 0.5343,
+      "epoch": 8.685647507847118,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.150644987711057e-06,
+      "loss": 0.983939208984375,
       "step": 754400
     },
     {
-      "epoch": 17.375886524822697,
-      "grad_norm": 3.222555160522461,
-      "learning_rate": 4.1532627506871626e-06,
-      "loss": 0.5166,
+      "epoch": 8.687950192921834,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.143373423888542e-06,
+      "loss": 0.9967646789550781,
       "step": 754600
     },
     {
-      "epoch": 17.380491848576956,
-      "grad_norm": 3.4512622356414795,
-      "learning_rate": 4.145991186864648e-06,
-      "loss": 0.5148,
+      "epoch": 8.69025287799655,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.136101860066026e-06,
+      "loss": 0.9793635559082031,
       "step": 754800
     },
     {
-      "epoch": 17.385097172331214,
-      "grad_norm": 2.870035171508789,
-      "learning_rate": 4.138719623042132e-06,
-      "loss": 0.5195,
+      "epoch": 8.692555563071263,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.12883029624351e-06,
+      "loss": 0.9769644927978516,
       "step": 755000
     },
     {
-      "epoch": 17.389702496085476,
-      "grad_norm": 3.166290521621704,
-      "learning_rate": 4.131448059219615e-06,
-      "loss": 0.5147,
+      "epoch": 8.694858248145978,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.121558732420995e-06,
+      "loss": 0.9760430145263672,
       "step": 755200
     },
     {
-      "epoch": 17.394307819839735,
-      "grad_norm": 3.1299068927764893,
-      "learning_rate": 4.1241764953971e-06,
-      "loss": 0.5193,
+      "epoch": 8.697160933220694,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.114287168598479e-06,
+      "loss": 0.975523681640625,
       "step": 755400
     },
     {
-      "epoch": 17.398913143593994,
-      "grad_norm": 2.6893653869628906,
-      "learning_rate": 4.1169049315745845e-06,
-      "loss": 0.5246,
+      "epoch": 8.69946361829541,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.107015604775963e-06,
+      "loss": 0.9740261840820312,
       "step": 755600
     },
     {
-      "epoch": 17.403518467348256,
-      "grad_norm": 2.988776206970215,
-      "learning_rate": 4.1096333677520696e-06,
-      "loss": 0.5223,
+      "epoch": 8.701766303370123,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.099744040953448e-06,
+      "loss": 0.9879347229003906,
       "step": 755800
     },
     {
-      "epoch": 17.408123791102515,
-      "grad_norm": 3.235541820526123,
-      "learning_rate": 4.102361803929553e-06,
-      "loss": 0.5116,
+      "epoch": 8.704068988444838,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.092472477130932e-06,
+      "loss": 0.9873762512207032,
       "step": 756000
     },
     {
-      "epoch": 17.412729114856774,
-      "grad_norm": 3.031001091003418,
-      "learning_rate": 4.09512659792615e-06,
-      "loss": 0.5167,
+      "epoch": 8.706371673519554,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.085200913308416e-06,
+      "loss": 0.9697098541259765,
       "step": 756200
     },
     {
-      "epoch": 17.417334438611036,
-      "grad_norm": 2.9991252422332764,
-      "learning_rate": 4.087855034103634e-06,
-      "loss": 0.5125,
+      "epoch": 8.70867435859427,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.0779293494859005e-06,
+      "loss": 0.9750542449951172,
       "step": 756400
     },
     {
-      "epoch": 17.421939762365295,
-      "grad_norm": 3.0221939086914062,
-      "learning_rate": 4.080583470281119e-06,
-      "loss": 0.5213,
+      "epoch": 8.710977043668983,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.070657785663385e-06,
+      "loss": 0.9895989227294922,
       "step": 756600
     },
     {
-      "epoch": 17.426545086119553,
-      "grad_norm": 2.6054844856262207,
-      "learning_rate": 4.073311906458603e-06,
-      "loss": 0.5255,
+      "epoch": 8.713279728743698,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.06338622184087e-06,
+      "loss": 0.9877436828613281,
       "step": 756800
     },
     {
-      "epoch": 17.431150409873815,
-      "grad_norm": 2.908795118331909,
-      "learning_rate": 4.066040342636087e-06,
-      "loss": 0.5223,
+      "epoch": 8.715582413818414,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.056114658018354e-06,
+      "loss": 0.9801091766357422,
       "step": 757000
     },
     {
-      "epoch": 17.435755733628074,
-      "grad_norm": 3.4502451419830322,
-      "learning_rate": 4.058768778813572e-06,
-      "loss": 0.5242,
+      "epoch": 8.717885098893127,
+      "grad_norm": 0.5,
+      "learning_rate": 4.048843094195837e-06,
+      "loss": 0.9831630706787109,
       "step": 757200
     },
     {
-      "epoch": 17.440361057382333,
-      "grad_norm": 3.335181474685669,
-      "learning_rate": 4.051497214991056e-06,
-      "loss": 0.5198,
+      "epoch": 8.720187783967843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.041571530373322e-06,
+      "loss": 0.9879035949707031,
       "step": 757400
     },
     {
-      "epoch": 17.444966381136595,
-      "grad_norm": 3.5651845932006836,
-      "learning_rate": 4.044225651168541e-06,
-      "loss": 0.5176,
+      "epoch": 8.722490469042558,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.034299966550807e-06,
+      "loss": 0.9739523315429688,
       "step": 757600
     },
     {
-      "epoch": 17.449571704890854,
-      "grad_norm": 2.8494350910186768,
-      "learning_rate": 4.0369540873460245e-06,
-      "loss": 0.52,
+      "epoch": 8.724793154117274,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.027028402728291e-06,
+      "loss": 0.9841912078857422,
       "step": 757800
     },
     {
-      "epoch": 17.454177028645113,
-      "grad_norm": 2.9451777935028076,
-      "learning_rate": 4.029682523523509e-06,
-      "loss": 0.5285,
+      "epoch": 8.727095839191987,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.019756838905775e-06,
+      "loss": 0.9674013519287109,
       "step": 758000
     },
     {
-      "epoch": 17.458782352399375,
-      "grad_norm": 3.2299864292144775,
-      "learning_rate": 4.022410959700994e-06,
-      "loss": 0.5242,
+      "epoch": 8.729398524266703,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.012485275083259e-06,
+      "loss": 1.0047560119628907,
       "step": 758200
     },
     {
-      "epoch": 17.463387676153634,
-      "grad_norm": 2.959676742553711,
-      "learning_rate": 4.015139395878478e-06,
-      "loss": 0.5261,
+      "epoch": 8.731701209341418,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.0052137112607435e-06,
+      "loss": 0.9761631774902344,
       "step": 758400
     },
     {
-      "epoch": 17.467992999907892,
-      "grad_norm": 3.600370407104492,
-      "learning_rate": 4.007867832055961e-06,
-      "loss": 0.5173,
+      "epoch": 8.734003894416132,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.9979421474382286e-06,
+      "loss": 0.9929995727539063,
       "step": 758600
     },
     {
-      "epoch": 17.472598323662154,
-      "grad_norm": 2.899125814437866,
-      "learning_rate": 4.000596268233446e-06,
-      "loss": 0.5236,
+      "epoch": 8.736306579490847,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.990670583615712e-06,
+      "loss": 0.9686499786376953,
       "step": 758800
     },
     {
-      "epoch": 17.477203647416413,
-      "grad_norm": 3.0333738327026367,
-      "learning_rate": 3.993324704410931e-06,
-      "loss": 0.5209,
+      "epoch": 8.738609264565563,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.983399019793197e-06,
+      "loss": 0.9847555541992188,
       "step": 759000
     },
     {
-      "epoch": 17.481808971170672,
-      "grad_norm": 2.875760078430176,
-      "learning_rate": 3.986053140588416e-06,
-      "loss": 0.5118,
+      "epoch": 8.740911949640278,
+      "grad_norm": 0.625,
+      "learning_rate": 3.976127455970681e-06,
+      "loss": 0.9836019134521484,
       "step": 759200
     },
     {
-      "epoch": 17.486414294924934,
-      "grad_norm": 2.935258150100708,
-      "learning_rate": 3.978781576765899e-06,
-      "loss": 0.5157,
+      "epoch": 8.743214634714992,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.9688558921481654e-06,
+      "loss": 0.98874755859375,
       "step": 759400
     },
     {
-      "epoch": 17.491019618679193,
-      "grad_norm": 3.5687882900238037,
-      "learning_rate": 3.971510012943383e-06,
-      "loss": 0.5255,
+      "epoch": 8.745517319789707,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.96158432832565e-06,
+      "loss": 0.9745330810546875,
       "step": 759600
     },
     {
-      "epoch": 17.49562494243345,
-      "grad_norm": 3.5541269779205322,
-      "learning_rate": 3.96427480693998e-06,
-      "loss": 0.5209,
+      "epoch": 8.747820004864423,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.954312764503134e-06,
+      "loss": 0.9955284881591797,
       "step": 759800
     },
     {
-      "epoch": 17.500230266187714,
-      "grad_norm": 2.5105998516082764,
-      "learning_rate": 3.957003243117465e-06,
-      "loss": 0.5237,
+      "epoch": 8.750122689939136,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.947041200680618e-06,
+      "loss": 0.9717284393310547,
       "step": 760000
     },
     {
-      "epoch": 17.504835589941973,
-      "grad_norm": 3.4563724994659424,
-      "learning_rate": 3.9497316792949495e-06,
-      "loss": 0.5289,
+      "epoch": 8.752425375013852,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.939769636858103e-06,
+      "loss": 0.9807109832763672,
       "step": 760200
     },
     {
-      "epoch": 17.50944091369623,
-      "grad_norm": 4.190462112426758,
-      "learning_rate": 3.942460115472433e-06,
-      "loss": 0.5144,
+      "epoch": 8.754728060088567,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.9324980730355865e-06,
+      "loss": 0.9712101745605469,
       "step": 760400
     },
     {
-      "epoch": 17.514046237450493,
-      "grad_norm": 3.6691291332244873,
-      "learning_rate": 3.935188551649918e-06,
-      "loss": 0.5129,
+      "epoch": 8.757030745163283,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.925226509213072e-06,
+      "loss": 0.9616006469726562,
       "step": 760600
     },
     {
-      "epoch": 17.518651561204752,
-      "grad_norm": 3.044640064239502,
-      "learning_rate": 3.927916987827402e-06,
-      "loss": 0.5197,
+      "epoch": 8.759333430237996,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.917954945390556e-06,
+      "loss": 0.9621672058105468,
       "step": 760800
     },
     {
-      "epoch": 17.52325688495901,
-      "grad_norm": 3.2396230697631836,
-      "learning_rate": 3.920645424004887e-06,
-      "loss": 0.5099,
+      "epoch": 8.761636115312712,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.91068338156804e-06,
+      "loss": 0.9885343170166015,
       "step": 761000
     },
     {
-      "epoch": 17.527862208713273,
-      "grad_norm": 3.3681046962738037,
-      "learning_rate": 3.9133738601823706e-06,
-      "loss": 0.5258,
+      "epoch": 8.763938800387427,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.903411817745525e-06,
+      "loss": 0.9749143981933593,
       "step": 761200
     },
     {
-      "epoch": 17.532467532467532,
-      "grad_norm": 3.473139524459839,
-      "learning_rate": 3.906102296359855e-06,
-      "loss": 0.5112,
+      "epoch": 8.766241485462142,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.8961402539230085e-06,
+      "loss": 0.9841162872314453,
       "step": 761400
     },
     {
-      "epoch": 17.53707285622179,
-      "grad_norm": 3.2254271507263184,
-      "learning_rate": 3.89883073253734e-06,
-      "loss": 0.5194,
+      "epoch": 8.768544170536856,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.888868690100493e-06,
+      "loss": 0.9840245056152344,
       "step": 761600
     },
     {
-      "epoch": 17.541678179976053,
-      "grad_norm": 3.007357120513916,
-      "learning_rate": 3.891559168714824e-06,
-      "loss": 0.5226,
+      "epoch": 8.770846855611572,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.881597126277978e-06,
+      "loss": 0.9939464569091797,
       "step": 761800
     },
     {
-      "epoch": 17.54628350373031,
-      "grad_norm": 3.1616954803466797,
-      "learning_rate": 3.884287604892308e-06,
-      "loss": 0.5195,
+      "epoch": 8.773149540686287,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.874325562455462e-06,
+      "loss": 0.9811724090576172,
       "step": 762000
     },
     {
-      "epoch": 17.550888827484574,
-      "grad_norm": 3.4159657955169678,
-      "learning_rate": 3.8770160410697925e-06,
-      "loss": 0.5196,
+      "epoch": 8.775452225761,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.867053998632946e-06,
+      "loss": 0.9716242980957032,
       "step": 762200
     },
     {
-      "epoch": 17.555494151238833,
-      "grad_norm": 2.448462724685669,
-      "learning_rate": 3.869744477247277e-06,
-      "loss": 0.5221,
+      "epoch": 8.777754910835716,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.8597824348104304e-06,
+      "loss": 0.9963446807861328,
       "step": 762400
     },
     {
-      "epoch": 17.56009947499309,
-      "grad_norm": 3.6879563331604004,
-      "learning_rate": 3.862472913424762e-06,
-      "loss": 0.5182,
+      "epoch": 8.780057595910431,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.852510870987915e-06,
+      "loss": 0.9831389617919922,
       "step": 762600
     },
     {
-      "epoch": 17.564704798747353,
-      "grad_norm": 3.100130796432495,
-      "learning_rate": 3.855201349602246e-06,
-      "loss": 0.5108,
+      "epoch": 8.782360280985147,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.8452393071654e-06,
+      "loss": 0.9837663269042969,
       "step": 762800
     },
     {
-      "epoch": 17.569310122501612,
-      "grad_norm": 3.8761160373687744,
-      "learning_rate": 3.847929785779729e-06,
-      "loss": 0.529,
+      "epoch": 8.78466296605986,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.837967743342883e-06,
+      "loss": 0.9864227294921875,
       "step": 763000
     },
     {
-      "epoch": 17.57391544625587,
-      "grad_norm": 3.630197525024414,
-      "learning_rate": 3.8406582219572145e-06,
-      "loss": 0.5229,
+      "epoch": 8.786965651134576,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.830696179520367e-06,
+      "loss": 1.002994384765625,
       "step": 763200
     },
     {
-      "epoch": 17.578520770010133,
-      "grad_norm": 3.126621961593628,
-      "learning_rate": 3.833386658134699e-06,
-      "loss": 0.5199,
+      "epoch": 8.789268336209291,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.823424615697852e-06,
+      "loss": 0.9955747222900391,
       "step": 763400
     },
     {
-      "epoch": 17.583126093764392,
-      "grad_norm": 2.9933745861053467,
-      "learning_rate": 3.826115094312183e-06,
-      "loss": 0.5093,
+      "epoch": 8.791571021284007,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.816153051875337e-06,
+      "loss": 0.975970458984375,
       "step": 763600
     },
     {
-      "epoch": 17.58773141751865,
-      "grad_norm": 3.0438787937164307,
-      "learning_rate": 3.818843530489667e-06,
-      "loss": 0.5211,
+      "epoch": 8.79387370635872,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.8088814880528212e-06,
+      "loss": 0.9739787292480468,
       "step": 763800
     },
     {
-      "epoch": 17.592336741272913,
-      "grad_norm": 3.802643299102783,
-      "learning_rate": 3.8115719666671513e-06,
-      "loss": 0.5248,
+      "epoch": 8.796176391433436,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.801609924230305e-06,
+      "loss": 0.9904090881347656,
       "step": 764000
     },
     {
-      "epoch": 17.59694206502717,
-      "grad_norm": 4.053820610046387,
-      "learning_rate": 3.8043004028446364e-06,
-      "loss": 0.5208,
+      "epoch": 8.798479076508151,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.7943383604077892e-06,
+      "loss": 0.9814543914794922,
       "step": 764200
     },
     {
-      "epoch": 17.60154738878143,
-      "grad_norm": 2.8927087783813477,
-      "learning_rate": 3.7970651968412333e-06,
-      "loss": 0.5216,
+      "epoch": 8.800781761582865,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.787066796585274e-06,
+      "loss": 0.9868421936035157,
       "step": 764400
     },
     {
-      "epoch": 17.606152712535692,
-      "grad_norm": 3.9510834217071533,
-      "learning_rate": 3.789793633018717e-06,
-      "loss": 0.5174,
+      "epoch": 8.80308444665758,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.779795232762758e-06,
+      "loss": 0.9769282531738281,
       "step": 764600
     },
     {
-      "epoch": 17.61075803628995,
-      "grad_norm": 2.498469114303589,
-      "learning_rate": 3.7825220691962013e-06,
-      "loss": 0.5162,
+      "epoch": 8.805387131732296,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.772523668940242e-06,
+      "loss": 0.9739444732666016,
       "step": 764800
     },
     {
-      "epoch": 17.61536336004421,
-      "grad_norm": 3.019011974334717,
-      "learning_rate": 3.775250505373686e-06,
-      "loss": 0.525,
+      "epoch": 8.807689816807011,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.765252105117727e-06,
+      "loss": 0.9981250762939453,
       "step": 765000
     },
     {
-      "epoch": 17.619968683798472,
-      "grad_norm": 3.5621178150177,
-      "learning_rate": 3.76797894155117e-06,
-      "loss": 0.5176,
+      "epoch": 8.809992501881725,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.7579805412952108e-06,
+      "loss": 0.9787905120849609,
       "step": 765200
     },
     {
-      "epoch": 17.62457400755273,
-      "grad_norm": 3.7816877365112305,
-      "learning_rate": 3.760707377728654e-06,
-      "loss": 0.5132,
+      "epoch": 8.81229518695644,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.750708977472696e-06,
+      "loss": 0.9816758728027344,
       "step": 765400
     },
     {
-      "epoch": 17.62917933130699,
-      "grad_norm": 2.997817277908325,
-      "learning_rate": 3.753435813906139e-06,
-      "loss": 0.5202,
+      "epoch": 8.814597872031156,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.7434374136501796e-06,
+      "loss": 0.9940653228759766,
       "step": 765600
     },
     {
-      "epoch": 17.633784655061252,
-      "grad_norm": 3.401996374130249,
-      "learning_rate": 3.746164250083623e-06,
-      "loss": 0.513,
+      "epoch": 8.816900557105871,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.7361658498276643e-06,
+      "loss": 0.9731235504150391,
       "step": 765800
     },
     {
-      "epoch": 17.63838997881551,
-      "grad_norm": 3.3747670650482178,
-      "learning_rate": 3.7388926862611075e-06,
-      "loss": 0.5162,
+      "epoch": 8.819203242180585,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.728894286005148e-06,
+      "loss": 0.9854316711425781,
       "step": 766000
     },
     {
-      "epoch": 17.64299530256977,
-      "grad_norm": 3.419926404953003,
-      "learning_rate": 3.7316211224385917e-06,
-      "loss": 0.5122,
+      "epoch": 8.8215059272553,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.7216227221826327e-06,
+      "loss": 0.9703307342529297,
       "step": 766200
     },
     {
-      "epoch": 17.64760062632403,
-      "grad_norm": 3.317556381225586,
-      "learning_rate": 3.7243859164351886e-06,
-      "loss": 0.5277,
+      "epoch": 8.823808612330016,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.714351158360117e-06,
+      "loss": 0.9908853149414063,
       "step": 766400
     },
     {
-      "epoch": 17.65220595007829,
-      "grad_norm": 3.193814516067505,
-      "learning_rate": 3.7171143526126733e-06,
-      "loss": 0.5075,
+      "epoch": 8.82611129740473,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7070795945376016e-06,
+      "loss": 0.9777758026123047,
       "step": 766600
     },
     {
-      "epoch": 17.65681127383255,
-      "grad_norm": 2.69360089302063,
-      "learning_rate": 3.709842788790157e-06,
-      "loss": 0.519,
+      "epoch": 8.828413982479445,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.6998080307150854e-06,
+      "loss": 0.9859429931640625,
       "step": 766800
     },
     {
-      "epoch": 17.66141659758681,
-      "grad_norm": 2.8568925857543945,
-      "learning_rate": 3.7025712249676417e-06,
-      "loss": 0.5266,
+      "epoch": 8.83071666755416,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.69253646689257e-06,
+      "loss": 0.9925715637207031,
       "step": 767000
     },
     {
-      "epoch": 17.66602192134107,
-      "grad_norm": 3.275151491165161,
-      "learning_rate": 3.695299661145126e-06,
-      "loss": 0.522,
+      "epoch": 8.833019352628876,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.6852649030700542e-06,
+      "loss": 0.9771989440917969,
       "step": 767200
     },
     {
-      "epoch": 17.67062724509533,
-      "grad_norm": 3.3356289863586426,
-      "learning_rate": 3.6880280973226106e-06,
-      "loss": 0.5257,
+      "epoch": 8.83532203770359,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.677993339247539e-06,
+      "loss": 0.9750688934326172,
       "step": 767400
     },
     {
-      "epoch": 17.67523256884959,
-      "grad_norm": 3.011615514755249,
-      "learning_rate": 3.6807565335000943e-06,
-      "loss": 0.5226,
+      "epoch": 8.837624722778305,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.6707217754250227e-06,
+      "loss": 0.9724152374267578,
       "step": 767600
     },
     {
-      "epoch": 17.67983789260385,
-      "grad_norm": 2.5866901874542236,
-      "learning_rate": 3.673484969677579e-06,
-      "loss": 0.5196,
+      "epoch": 8.83992740785302,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6634502116025073e-06,
+      "loss": 0.982885971069336,
       "step": 767800
     },
     {
-      "epoch": 17.684443216358112,
-      "grad_norm": 3.7336740493774414,
-      "learning_rate": 3.666213405855063e-06,
-      "loss": 0.5119,
+      "epoch": 8.842230092927734,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.656178647779992e-06,
+      "loss": 0.9649812316894532,
       "step": 768000
     },
     {
-      "epoch": 17.68904854011237,
-      "grad_norm": 2.9361438751220703,
-      "learning_rate": 3.658941842032548e-06,
-      "loss": 0.5093,
+      "epoch": 8.84453277800245,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.648907083957476e-06,
+      "loss": 1.0030159759521484,
       "step": 768200
     },
     {
-      "epoch": 17.69365386386663,
-      "grad_norm": 3.3662490844726562,
-      "learning_rate": 3.6516702782100317e-06,
-      "loss": 0.5179,
+      "epoch": 8.846835463077165,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.6416355201349604e-06,
+      "loss": 0.9977754211425781,
       "step": 768400
     },
     {
-      "epoch": 17.69825918762089,
-      "grad_norm": 3.797208547592163,
-      "learning_rate": 3.6444350722066286e-06,
-      "loss": 0.5157,
+      "epoch": 8.84913814815188,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.6343639563124446e-06,
+      "loss": 0.9839167785644531,
       "step": 768600
     },
     {
-      "epoch": 17.70286451137515,
-      "grad_norm": 3.147529363632202,
-      "learning_rate": 3.637163508384113e-06,
-      "loss": 0.5324,
+      "epoch": 8.851440833226594,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.6270923924899293e-06,
+      "loss": 0.9640473175048828,
       "step": 768800
     },
     {
-      "epoch": 17.70746983512941,
-      "grad_norm": 3.4828877449035645,
-      "learning_rate": 3.62992830238071e-06,
-      "loss": 0.5217,
+      "epoch": 8.85374351830131,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.6198208286674135e-06,
+      "loss": 0.9928032684326172,
       "step": 769000
     },
     {
-      "epoch": 17.71207515888367,
-      "grad_norm": 3.616196393966675,
-      "learning_rate": 3.6226567385581943e-06,
-      "loss": 0.5199,
+      "epoch": 8.856046203376025,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.6125492648448977e-06,
+      "loss": 0.9882096099853516,
       "step": 769200
     },
     {
-      "epoch": 17.71668048263793,
-      "grad_norm": 2.796513319015503,
-      "learning_rate": 3.615385174735679e-06,
-      "loss": 0.5121,
+      "epoch": 8.858348888450738,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.605277701022382e-06,
+      "loss": 0.9933953857421876,
       "step": 769400
     },
     {
-      "epoch": 17.72128580639219,
-      "grad_norm": 2.718034029006958,
-      "learning_rate": 3.6081136109131628e-06,
-      "loss": 0.5194,
+      "epoch": 8.860651573525454,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5980061371998666e-06,
+      "loss": 0.9875212097167969,
       "step": 769600
     },
     {
-      "epoch": 17.72589113014645,
-      "grad_norm": 3.4165828227996826,
-      "learning_rate": 3.6008784049097597e-06,
-      "loss": 0.5348,
+      "epoch": 8.862954258600169,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.5907345733773504e-06,
+      "loss": 0.9890708160400391,
       "step": 769800
     },
     {
-      "epoch": 17.73049645390071,
-      "grad_norm": 3.0737781524658203,
-      "learning_rate": 3.5936068410872443e-06,
-      "loss": 0.517,
+      "epoch": 8.865256943674884,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.583463009554835e-06,
+      "loss": 0.9977067565917969,
       "step": 770000
     },
     {
-      "epoch": 17.735101777654968,
-      "grad_norm": 3.283372163772583,
-      "learning_rate": 3.5863352772647285e-06,
-      "loss": 0.5262,
+      "epoch": 8.867559628749598,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.5761914457323192e-06,
+      "loss": 0.9735486602783203,
       "step": 770200
     },
     {
-      "epoch": 17.73970710140923,
-      "grad_norm": 2.618170738220215,
-      "learning_rate": 3.579063713442213e-06,
-      "loss": 0.5252,
+      "epoch": 8.869862313824314,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.568919881909804e-06,
+      "loss": 0.9656082916259766,
       "step": 770400
     },
     {
-      "epoch": 17.74431242516349,
-      "grad_norm": 3.1445319652557373,
-      "learning_rate": 3.571792149619697e-06,
-      "loss": 0.5196,
+      "epoch": 8.872164998899029,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.5616483180872877e-06,
+      "loss": 0.9706471252441407,
       "step": 770600
     },
     {
-      "epoch": 17.748917748917748,
-      "grad_norm": 3.0439717769622803,
-      "learning_rate": 3.5645205857971816e-06,
-      "loss": 0.5282,
+      "epoch": 8.874467683973744,
+      "grad_norm": 0.75,
+      "learning_rate": 3.5543767542647723e-06,
+      "loss": 1.0011856079101562,
       "step": 770800
     },
     {
-      "epoch": 17.75352307267201,
-      "grad_norm": 2.830927610397339,
-      "learning_rate": 3.557249021974666e-06,
-      "loss": 0.517,
+      "epoch": 8.876770369048458,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.5471051904422565e-06,
+      "loss": 0.9597758483886719,
       "step": 771000
     },
     {
-      "epoch": 17.75812839642627,
-      "grad_norm": 2.9426324367523193,
-      "learning_rate": 3.5499774581521505e-06,
-      "loss": 0.5185,
+      "epoch": 8.879073054123173,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.539833626619741e-06,
+      "loss": 0.9833696746826172,
       "step": 771200
     },
     {
-      "epoch": 17.762733720180528,
-      "grad_norm": 3.1209681034088135,
-      "learning_rate": 3.5427058943296343e-06,
-      "loss": 0.5221,
+      "epoch": 8.881375739197889,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.532562062797225e-06,
+      "loss": 0.9873859405517578,
       "step": 771400
     },
     {
-      "epoch": 17.76733904393479,
-      "grad_norm": 2.9420995712280273,
-      "learning_rate": 3.535434330507119e-06,
-      "loss": 0.5119,
+      "epoch": 8.883678424272603,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.5252904989747096e-06,
+      "loss": 0.9804158020019531,
       "step": 771600
     },
     {
-      "epoch": 17.77194436768905,
-      "grad_norm": 2.995070219039917,
-      "learning_rate": 3.5281627666846036e-06,
-      "loss": 0.5254,
+      "epoch": 8.885981109347318,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.518018935152194e-06,
+      "loss": 0.9784156036376953,
       "step": 771800
     },
     {
-      "epoch": 17.776549691443307,
-      "grad_norm": 3.6873717308044434,
-      "learning_rate": 3.5208912028620878e-06,
-      "loss": 0.527,
+      "epoch": 8.888283794422033,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.5107473713296785e-06,
+      "loss": 0.9803633117675781,
       "step": 772000
     },
     {
-      "epoch": 17.78115501519757,
-      "grad_norm": 2.82792592048645,
-      "learning_rate": 3.513619639039572e-06,
-      "loss": 0.5208,
+      "epoch": 8.890586479496749,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.5034758075071623e-06,
+      "loss": 0.9750108337402343,
       "step": 772200
     },
     {
-      "epoch": 17.785760338951828,
-      "grad_norm": 2.409423351287842,
-      "learning_rate": 3.5063480752170562e-06,
-      "loss": 0.526,
+      "epoch": 8.892889164571463,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.496204243684647e-06,
+      "loss": 0.9871961975097656,
       "step": 772400
     },
     {
-      "epoch": 17.790365662706087,
-      "grad_norm": 2.7182860374450684,
-      "learning_rate": 3.499076511394541e-06,
-      "loss": 0.5197,
+      "epoch": 8.895191849646178,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.488932679862131e-06,
+      "loss": 0.9883628082275391,
       "step": 772600
     },
     {
-      "epoch": 17.79497098646035,
-      "grad_norm": 4.183868885040283,
-      "learning_rate": 3.491804947572025e-06,
-      "loss": 0.5077,
+      "epoch": 8.897494534720893,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.4816611160396158e-06,
+      "loss": 0.9762785339355469,
       "step": 772800
     },
     {
-      "epoch": 17.799576310214608,
-      "grad_norm": 3.585371971130371,
-      "learning_rate": 3.4845333837495093e-06,
-      "loss": 0.5102,
+      "epoch": 8.899797219795609,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.4743895522170996e-06,
+      "loss": 0.979869384765625,
       "step": 773000
     },
     {
-      "epoch": 17.804181633968867,
-      "grad_norm": 3.362762451171875,
-      "learning_rate": 3.4772618199269935e-06,
-      "loss": 0.5208,
+      "epoch": 8.902099904870322,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.467117988394584e-06,
+      "loss": 0.9955022430419922,
       "step": 773200
     },
     {
-      "epoch": 17.80878695772313,
-      "grad_norm": 3.437537431716919,
-      "learning_rate": 3.469990256104478e-06,
-      "loss": 0.531,
+      "epoch": 8.904402589945038,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.4598464245720684e-06,
+      "loss": 0.9861965179443359,
       "step": 773400
     },
     {
-      "epoch": 17.813392281477388,
-      "grad_norm": 3.5331273078918457,
-      "learning_rate": 3.4627186922819624e-06,
-      "loss": 0.504,
+      "epoch": 8.906705275019753,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.452574860749553e-06,
+      "loss": 0.9948319244384766,
       "step": 773600
     },
     {
-      "epoch": 17.817997605231646,
-      "grad_norm": 2.810429811477661,
-      "learning_rate": 3.4554471284594466e-06,
-      "loss": 0.5231,
+      "epoch": 8.909007960094467,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.445303296927037e-06,
+      "loss": 0.9853091430664063,
       "step": 773800
     },
     {
-      "epoch": 17.82260292898591,
-      "grad_norm": 2.8518621921539307,
-      "learning_rate": 3.448175564636931e-06,
-      "loss": 0.5173,
+      "epoch": 8.911310645169182,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.4380317331045215e-06,
+      "loss": 0.9895594787597656,
       "step": 774000
     },
     {
-      "epoch": 17.827208252740167,
-      "grad_norm": 4.182199478149414,
-      "learning_rate": 3.4409040008144155e-06,
-      "loss": 0.5206,
+      "epoch": 8.913613330243898,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.4307601692820057e-06,
+      "loss": 0.9816792297363282,
       "step": 774200
     },
     {
-      "epoch": 17.831813576494426,
-      "grad_norm": 2.786973237991333,
-      "learning_rate": 3.4336324369918997e-06,
-      "loss": 0.5256,
+      "epoch": 8.915916015318613,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.4234886054594904e-06,
+      "loss": 0.984228515625,
       "step": 774400
     },
     {
-      "epoch": 17.836418900248688,
-      "grad_norm": 4.052456855773926,
-      "learning_rate": 3.426360873169384e-06,
-      "loss": 0.5245,
+      "epoch": 8.918218700393327,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.4162170416369746e-06,
+      "loss": 0.9908686828613281,
       "step": 774600
     },
     {
-      "epoch": 17.841024224002947,
-      "grad_norm": 3.5962717533111572,
-      "learning_rate": 3.419089309346868e-06,
-      "loss": 0.5331,
+      "epoch": 8.920521385468042,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.408945477814459e-06,
+      "loss": 0.9926894378662109,
       "step": 774800
     },
     {
-      "epoch": 17.845629547757206,
-      "grad_norm": 3.3177649974823,
-      "learning_rate": 3.4118177455243528e-06,
-      "loss": 0.5154,
+      "epoch": 8.922824070542758,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.4016739139919434e-06,
+      "loss": 0.9694911956787109,
       "step": 775000
     },
     {
-      "epoch": 17.850234871511468,
-      "grad_norm": 2.6768603324890137,
-      "learning_rate": 3.4045461817018366e-06,
-      "loss": 0.5156,
+      "epoch": 8.925126755617471,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.3944023501694277e-06,
+      "loss": 0.9806173706054687,
       "step": 775200
     },
     {
-      "epoch": 17.854840195265727,
-      "grad_norm": 3.424499034881592,
-      "learning_rate": 3.397274617879321e-06,
-      "loss": 0.5172,
+      "epoch": 8.927429440692187,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.387130786346912e-06,
+      "loss": 0.9553947448730469,
       "step": 775400
     },
     {
-      "epoch": 17.85944551901999,
-      "grad_norm": 2.6786439418792725,
-      "learning_rate": 3.3900030540568054e-06,
-      "loss": 0.5164,
+      "epoch": 8.929732125766902,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.379859222524396e-06,
+      "loss": 0.9826571655273437,
       "step": 775600
     },
     {
-      "epoch": 17.864050842774247,
-      "grad_norm": 4.423825740814209,
-      "learning_rate": 3.38273149023429e-06,
-      "loss": 0.5096,
+      "epoch": 8.932034810841618,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.3725876587018808e-06,
+      "loss": 0.9717965698242188,
       "step": 775800
     },
     {
-      "epoch": 17.868656166528506,
-      "grad_norm": 3.1935222148895264,
-      "learning_rate": 3.375459926411774e-06,
-      "loss": 0.5238,
+      "epoch": 8.934337495916331,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.365316094879365e-06,
+      "loss": 1.011261215209961,
       "step": 776000
     },
     {
-      "epoch": 17.87326149028277,
-      "grad_norm": 3.05393648147583,
-      "learning_rate": 3.3681883625892585e-06,
-      "loss": 0.5086,
+      "epoch": 8.936640180991047,
+      "grad_norm": 0.875,
+      "learning_rate": 3.358044531056849e-06,
+      "loss": 0.99462158203125,
       "step": 776200
     },
     {
-      "epoch": 17.877866814037027,
-      "grad_norm": 2.8868277072906494,
-      "learning_rate": 3.3609167987667427e-06,
-      "loss": 0.5192,
+      "epoch": 8.938942866065762,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.3507729672343334e-06,
+      "loss": 0.9862718200683593,
       "step": 776400
     },
     {
-      "epoch": 17.882472137791286,
-      "grad_norm": 3.3218915462493896,
-      "learning_rate": 3.3536815927633396e-06,
-      "loss": 0.5108,
+      "epoch": 8.941245551140476,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.343501403411818e-06,
+      "loss": 0.958528823852539,
       "step": 776600
     },
     {
-      "epoch": 17.887077461545548,
-      "grad_norm": 3.428025007247925,
-      "learning_rate": 3.3464100289408243e-06,
-      "loss": 0.5158,
+      "epoch": 8.943548236215191,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.3362298395893023e-06,
+      "loss": 0.996722183227539,
       "step": 776800
     },
     {
-      "epoch": 17.891682785299807,
-      "grad_norm": 2.7553398609161377,
-      "learning_rate": 3.339138465118308e-06,
-      "loss": 0.512,
+      "epoch": 8.945850921289907,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.3289582757667865e-06,
+      "loss": 0.9624080657958984,
       "step": 777000
     },
     {
-      "epoch": 17.896288109054066,
-      "grad_norm": 3.779360294342041,
-      "learning_rate": 3.3318669012957927e-06,
-      "loss": 0.5345,
+      "epoch": 8.948153606364622,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.3216867119442707e-06,
+      "loss": 0.9863372039794922,
       "step": 777200
     },
     {
-      "epoch": 17.900893432808328,
-      "grad_norm": 4.041235446929932,
-      "learning_rate": 3.324595337473277e-06,
-      "loss": 0.5183,
+      "epoch": 8.950456291439336,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.3144151481217554e-06,
+      "loss": 0.9881964111328125,
       "step": 777400
     },
     {
-      "epoch": 17.905498756562586,
-      "grad_norm": 3.0583138465881348,
-      "learning_rate": 3.3173237736507616e-06,
-      "loss": 0.5256,
+      "epoch": 8.952758976514051,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.3071435842992396e-06,
+      "loss": 0.9761305999755859,
       "step": 777600
     },
     {
-      "epoch": 17.910104080316845,
-      "grad_norm": 2.8636226654052734,
-      "learning_rate": 3.3100522098282454e-06,
-      "loss": 0.5244,
+      "epoch": 8.955061661588767,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.299872020476724e-06,
+      "loss": 0.9757279968261718,
       "step": 777800
     },
     {
-      "epoch": 17.914709404071107,
-      "grad_norm": 3.2224411964416504,
-      "learning_rate": 3.30278064600573e-06,
-      "loss": 0.5183,
+      "epoch": 8.957364346663482,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.292600456654208e-06,
+      "loss": 0.971686782836914,
       "step": 778000
     },
     {
-      "epoch": 17.919314727825366,
-      "grad_norm": 3.5364627838134766,
-      "learning_rate": 3.2955090821832142e-06,
-      "loss": 0.5119,
+      "epoch": 8.959667031738196,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.2853288928316927e-06,
+      "loss": 0.9756494903564453,
       "step": 778200
     },
     {
-      "epoch": 17.923920051579625,
-      "grad_norm": 3.234290361404419,
-      "learning_rate": 3.288237518360699e-06,
-      "loss": 0.5245,
+      "epoch": 8.961969716812911,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.278057329009177e-06,
+      "loss": 0.9671916198730469,
       "step": 778400
     },
     {
-      "epoch": 17.928525375333887,
-      "grad_norm": 2.7013673782348633,
-      "learning_rate": 3.2809659545381827e-06,
-      "loss": 0.5182,
+      "epoch": 8.964272401887627,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.270785765186661e-06,
+      "loss": 0.9700312042236328,
       "step": 778600
     },
     {
-      "epoch": 17.933130699088146,
-      "grad_norm": 3.9952573776245117,
-      "learning_rate": 3.2736943907156673e-06,
-      "loss": 0.5145,
+      "epoch": 8.96657508696234,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.2635142013641453e-06,
+      "loss": 1.0189725494384765,
       "step": 778800
     },
     {
-      "epoch": 17.937736022842405,
-      "grad_norm": 3.0978848934173584,
-      "learning_rate": 3.2664228268931515e-06,
-      "loss": 0.5202,
+      "epoch": 8.968877772037056,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.25624263754163e-06,
+      "loss": 0.9961846923828125,
       "step": 779000
     },
     {
-      "epoch": 17.942341346596667,
-      "grad_norm": 3.4453442096710205,
-      "learning_rate": 3.2591876208897484e-06,
-      "loss": 0.5161,
+      "epoch": 8.971180457111771,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.248971073719114e-06,
+      "loss": 0.9650535583496094,
       "step": 779200
     },
     {
-      "epoch": 17.946946670350926,
-      "grad_norm": 2.99537992477417,
-      "learning_rate": 3.251916057067233e-06,
-      "loss": 0.5138,
+      "epoch": 8.973483142186486,
+      "grad_norm": 0.625,
+      "learning_rate": 3.2416995098965984e-06,
+      "loss": 1.0100181579589844,
       "step": 779400
     },
     {
-      "epoch": 17.951551994105184,
-      "grad_norm": 2.7019853591918945,
-      "learning_rate": 3.244644493244717e-06,
-      "loss": 0.5056,
+      "epoch": 8.9757858272612,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.2344279460740826e-06,
+      "loss": 0.9652468872070312,
       "step": 779600
     },
     {
-      "epoch": 17.956157317859446,
-      "grad_norm": 3.723837375640869,
-      "learning_rate": 3.2373729294222015e-06,
-      "loss": 0.5151,
+      "epoch": 8.978088512335916,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.2271563822515673e-06,
+      "loss": 0.9727928161621093,
       "step": 779800
     },
     {
-      "epoch": 17.960762641613705,
-      "grad_norm": 2.7109344005584717,
-      "learning_rate": 3.230101365599686e-06,
-      "loss": 0.5082,
+      "epoch": 8.980391197410631,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.2198848184290515e-06,
+      "loss": 0.9928997802734375,
       "step": 780000
     },
     {
-      "epoch": 17.965367965367964,
-      "grad_norm": 3.6867780685424805,
-      "learning_rate": 3.2228298017771704e-06,
-      "loss": 0.5128,
+      "epoch": 8.982693882485346,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.2126132546065357e-06,
+      "loss": 1.0174883270263673,
       "step": 780200
     },
     {
-      "epoch": 17.969973289122226,
-      "grad_norm": 3.721771478652954,
-      "learning_rate": 3.2155945957737673e-06,
-      "loss": 0.5192,
+      "epoch": 8.98499656756006,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.20534169078402e-06,
+      "loss": 0.994881591796875,
       "step": 780400
     },
     {
-      "epoch": 17.974578612876485,
-      "grad_norm": 2.7441351413726807,
-      "learning_rate": 3.2083230319512515e-06,
-      "loss": 0.5246,
+      "epoch": 8.987299252634775,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.1980701269615046e-06,
+      "loss": 0.9750624847412109,
       "step": 780600
     },
     {
-      "epoch": 17.979183936630744,
-      "grad_norm": 3.119499921798706,
-      "learning_rate": 3.2010514681287357e-06,
-      "loss": 0.519,
+      "epoch": 8.989601937709491,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.1907985631389884e-06,
+      "loss": 0.998040771484375,
       "step": 780800
     },
     {
-      "epoch": 17.983789260385006,
-      "grad_norm": 5.068978786468506,
-      "learning_rate": 3.1937799043062204e-06,
-      "loss": 0.5247,
+      "epoch": 8.991904622784205,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.183526999316473e-06,
+      "loss": 0.9677017974853516,
       "step": 781000
     },
     {
-      "epoch": 17.988394584139265,
-      "grad_norm": 2.891235828399658,
-      "learning_rate": 3.1865083404837046e-06,
-      "loss": 0.5135,
+      "epoch": 8.99420730785892,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.1762554354939572e-06,
+      "loss": 0.9724453735351563,
       "step": 781200
     },
     {
-      "epoch": 17.992999907893523,
-      "grad_norm": 3.192763566970825,
-      "learning_rate": 3.179236776661189e-06,
-      "loss": 0.5187,
+      "epoch": 8.996509992933635,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.168983871671442e-06,
+      "loss": 0.9718405151367188,
       "step": 781400
     },
     {
-      "epoch": 17.997605231647785,
-      "grad_norm": 3.4980311393737793,
-      "learning_rate": 3.171965212838673e-06,
-      "loss": 0.5234,
+      "epoch": 8.99881267800835,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.161712307848926e-06,
+      "loss": 0.9822624969482422,
       "step": 781600
     },
     {
-      "epoch": 18.0,
-      "eval_loss": 0.5081140398979187,
-      "eval_runtime": 170.1048,
-      "eval_samples_per_second": 166.727,
-      "eval_steps_per_second": 10.423,
+      "epoch": 9.0,
+      "eval_loss": NaN,
+      "eval_runtime": 429.8804,
+      "eval_samples_per_second": 65.974,
+      "eval_steps_per_second": 32.988,
       "step": 781704
     },
     {
-      "epoch": 18.002210555402044,
-      "grad_norm": 2.9518814086914062,
-      "learning_rate": 3.1646936490161577e-06,
-      "loss": 0.5216,
+      "epoch": 9.001105288835863,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.1544407440264103e-06,
+      "loss": 0.9588357543945313,
       "step": 781800
     },
     {
-      "epoch": 18.006815879156306,
-      "grad_norm": 2.917123794555664,
-      "learning_rate": 3.157422085193642e-06,
-      "loss": 0.5168,
+      "epoch": 9.003407973910578,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.147169180203895e-06,
+      "loss": 0.9717916870117187,
       "step": 782000
     },
     {
-      "epoch": 18.011421202910565,
-      "grad_norm": 3.122274875640869,
-      "learning_rate": 3.150150521371126e-06,
-      "loss": 0.511,
+      "epoch": 9.005710658985294,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.139897616381379e-06,
+      "loss": 0.9799015045166015,
       "step": 782200
     },
     {
-      "epoch": 18.016026526664824,
-      "grad_norm": 3.571920394897461,
-      "learning_rate": 3.1428789575486103e-06,
-      "loss": 0.5183,
+      "epoch": 9.008013344060007,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.1326260525588634e-06,
+      "loss": 0.9933222198486328,
       "step": 782400
     },
     {
-      "epoch": 18.020631850419086,
-      "grad_norm": 3.9012715816497803,
-      "learning_rate": 3.135607393726095e-06,
-      "loss": 0.5222,
+      "epoch": 9.010316029134723,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1253544887363476e-06,
+      "loss": 0.972984390258789,
       "step": 782600
     },
     {
-      "epoch": 18.025237174173345,
-      "grad_norm": 3.0531198978424072,
-      "learning_rate": 3.128335829903579e-06,
-      "loss": 0.5064,
+      "epoch": 9.012618714209438,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.1180829249138322e-06,
+      "loss": 0.9755500030517578,
       "step": 782800
     },
     {
-      "epoch": 18.029842497927604,
-      "grad_norm": 3.265467405319214,
-      "learning_rate": 3.121100623900176e-06,
-      "loss": 0.5119,
+      "epoch": 9.014921399284153,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.1108113610913165e-06,
+      "loss": 0.9731624603271485,
       "step": 783000
     },
     {
-      "epoch": 18.034447821681866,
-      "grad_norm": 2.407761335372925,
-      "learning_rate": 3.1138290600776603e-06,
-      "loss": 0.5109,
+      "epoch": 9.017224084358867,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.1035397972688007e-06,
+      "loss": 0.9951696014404297,
       "step": 783200
     },
     {
-      "epoch": 18.039053145436124,
-      "grad_norm": 3.1489205360412598,
-      "learning_rate": 3.1065574962551445e-06,
-      "loss": 0.509,
+      "epoch": 9.019526769433583,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.096268233446285e-06,
+      "loss": 0.9952273559570313,
       "step": 783400
     },
     {
-      "epoch": 18.043658469190383,
-      "grad_norm": 3.7624034881591797,
-      "learning_rate": 3.099285932432629e-06,
-      "loss": 0.5208,
+      "epoch": 9.021829454508298,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.0889966696237695e-06,
+      "loss": 0.989826431274414,
       "step": 783600
     },
     {
-      "epoch": 18.048263792944645,
-      "grad_norm": 3.1477630138397217,
-      "learning_rate": 3.0920143686101134e-06,
-      "loss": 0.5228,
+      "epoch": 9.024132139583012,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.0817251058012538e-06,
+      "loss": 0.9967062377929687,
       "step": 783800
     },
     {
-      "epoch": 18.052869116698904,
-      "grad_norm": 3.3497445583343506,
-      "learning_rate": 3.0847428047875976e-06,
-      "loss": 0.5191,
+      "epoch": 9.026434824657727,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.074453541978738e-06,
+      "loss": 0.9919381713867188,
       "step": 784000
     },
     {
-      "epoch": 18.057474440453163,
-      "grad_norm": 2.5725488662719727,
-      "learning_rate": 3.077471240965082e-06,
-      "loss": 0.5346,
+      "epoch": 9.028737509732442,
+      "grad_norm": 0.5,
+      "learning_rate": 3.067181978156222e-06,
+      "loss": 0.9788978576660157,
       "step": 784200
     },
     {
-      "epoch": 18.062079764207425,
-      "grad_norm": 3.2691822052001953,
-      "learning_rate": 3.0701996771425665e-06,
-      "loss": 0.5029,
+      "epoch": 9.031040194807158,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.059910414333707e-06,
+      "loss": 0.9845250701904297,
       "step": 784400
     },
     {
-      "epoch": 18.066685087961684,
-      "grad_norm": 3.542738914489746,
-      "learning_rate": 3.0629281133200507e-06,
-      "loss": 0.5097,
+      "epoch": 9.033342879881872,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.052638850511191e-06,
+      "loss": 0.9908609008789062,
       "step": 784600
     },
     {
-      "epoch": 18.071290411715943,
-      "grad_norm": 3.2178702354431152,
-      "learning_rate": 3.055656549497535e-06,
-      "loss": 0.5153,
+      "epoch": 9.035645564956587,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.0453672866886753e-06,
+      "loss": 0.979962387084961,
       "step": 784800
     },
     {
-      "epoch": 18.075895735470205,
-      "grad_norm": 3.4788858890533447,
-      "learning_rate": 3.048384985675019e-06,
-      "loss": 0.5136,
+      "epoch": 9.037948250031302,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.0380957228661595e-06,
+      "loss": 0.9913197326660156,
       "step": 785000
     },
     {
-      "epoch": 18.080501059224463,
-      "grad_norm": 2.703796625137329,
-      "learning_rate": 3.041113421852504e-06,
-      "loss": 0.525,
+      "epoch": 9.040250935106018,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.030824159043644e-06,
+      "loss": 0.9811695098876954,
       "step": 785200
     },
     {
-      "epoch": 18.085106382978722,
-      "grad_norm": 3.2644095420837402,
-      "learning_rate": 3.033841858029988e-06,
-      "loss": 0.5144,
+      "epoch": 9.042553620180731,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.0235525952211284e-06,
+      "loss": 0.9808423614501953,
       "step": 785400
     },
     {
-      "epoch": 18.089711706732984,
-      "grad_norm": 3.0157806873321533,
-      "learning_rate": 3.0265702942074722e-06,
-      "loss": 0.5259,
+      "epoch": 9.044856305255447,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.0162810313986126e-06,
+      "loss": 0.9801983642578125,
       "step": 785600
     },
     {
-      "epoch": 18.094317030487243,
-      "grad_norm": 3.0737905502319336,
-      "learning_rate": 3.0192987303849565e-06,
-      "loss": 0.5122,
+      "epoch": 9.047158990330162,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.009009467576097e-06,
+      "loss": 0.9870307159423828,
       "step": 785800
     },
     {
-      "epoch": 18.098922354241502,
-      "grad_norm": 2.9035747051239014,
-      "learning_rate": 3.012027166562441e-06,
-      "loss": 0.5151,
+      "epoch": 9.049461675404876,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.0017379037535814e-06,
+      "loss": 0.9723702239990234,
       "step": 786000
     },
     {
-      "epoch": 18.103527677995764,
-      "grad_norm": 2.6687934398651123,
-      "learning_rate": 3.0047556027399253e-06,
-      "loss": 0.516,
+      "epoch": 9.051764360479591,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.9944663399310657e-06,
+      "loss": 0.9945685577392578,
       "step": 786200
     },
     {
-      "epoch": 18.108133001750023,
-      "grad_norm": 3.1634881496429443,
-      "learning_rate": 2.9974840389174095e-06,
-      "loss": 0.5203,
+      "epoch": 9.054067045554307,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.98719477610855e-06,
+      "loss": 0.9849127197265625,
       "step": 786400
     },
     {
-      "epoch": 18.11273832550428,
-      "grad_norm": 2.5920820236206055,
-      "learning_rate": 2.9902124750948938e-06,
-      "loss": 0.5052,
+      "epoch": 9.056369730629022,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.979923212286034e-06,
+      "loss": 0.9685233306884765,
       "step": 786600
     },
     {
-      "epoch": 18.117343649258544,
-      "grad_norm": 3.145585536956787,
-      "learning_rate": 2.9829409112723784e-06,
-      "loss": 0.509,
+      "epoch": 9.058672415703736,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.9726516484635187e-06,
+      "loss": 0.9885980987548828,
       "step": 786800
     },
     {
-      "epoch": 18.121948973012802,
-      "grad_norm": 3.541355609893799,
-      "learning_rate": 2.9756693474498626e-06,
-      "loss": 0.5237,
+      "epoch": 9.060975100778451,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.965380084641003e-06,
+      "loss": 0.9755864715576172,
       "step": 787000
     },
     {
-      "epoch": 18.12655429676706,
-      "grad_norm": 2.4166955947875977,
-      "learning_rate": 2.968397783627347e-06,
-      "loss": 0.508,
+      "epoch": 9.063277785853167,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.958108520818487e-06,
+      "loss": 0.9793693542480468,
       "step": 787200
     },
     {
-      "epoch": 18.131159620521323,
-      "grad_norm": 3.0432989597320557,
-      "learning_rate": 2.961126219804831e-06,
-      "loss": 0.5131,
+      "epoch": 9.065580470927882,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9508369569959714e-06,
+      "loss": 0.9743875885009765,
       "step": 787400
     },
     {
-      "epoch": 18.135764944275582,
-      "grad_norm": 2.8021647930145264,
-      "learning_rate": 2.9538546559823157e-06,
-      "loss": 0.5154,
+      "epoch": 9.067883156002596,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.943565393173456e-06,
+      "loss": 0.987215576171875,
       "step": 787600
     },
     {
-      "epoch": 18.14037026802984,
-      "grad_norm": 4.511200428009033,
-      "learning_rate": 2.9465830921598e-06,
-      "loss": 0.5155,
+      "epoch": 9.070185841077311,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9362938293509403e-06,
+      "loss": 0.9882207489013672,
       "step": 787800
     },
     {
-      "epoch": 18.144975591784103,
-      "grad_norm": 3.441237449645996,
-      "learning_rate": 2.939311528337284e-06,
-      "loss": 0.5131,
+      "epoch": 9.072488526152027,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.9290222655284245e-06,
+      "loss": 0.9898771667480468,
       "step": 788000
     },
     {
-      "epoch": 18.149580915538362,
-      "grad_norm": 3.7599685192108154,
-      "learning_rate": 2.9320399645147688e-06,
-      "loss": 0.5169,
+      "epoch": 9.07479121122674,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.9217507017059087e-06,
+      "loss": 0.9595680999755859,
       "step": 788200
     },
     {
-      "epoch": 18.15418623929262,
-      "grad_norm": 3.4300057888031006,
-      "learning_rate": 2.9248047585113653e-06,
-      "loss": 0.5213,
+      "epoch": 9.077093896301456,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.9144791378833933e-06,
+      "loss": 0.9669092559814453,
       "step": 788400
     },
     {
-      "epoch": 18.158791563046883,
-      "grad_norm": 3.6189639568328857,
-      "learning_rate": 2.91753319468885e-06,
-      "loss": 0.5173,
+      "epoch": 9.079396581376171,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.907207574060878e-06,
+      "loss": 0.9772784423828125,
       "step": 788600
     },
     {
-      "epoch": 18.16339688680114,
-      "grad_norm": 2.8627123832702637,
-      "learning_rate": 2.9102616308663346e-06,
-      "loss": 0.5261,
+      "epoch": 9.081699266450887,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.8999360102383618e-06,
+      "loss": 0.9737366485595703,
       "step": 788800
     },
     {
-      "epoch": 18.168002210555404,
-      "grad_norm": 3.8576009273529053,
-      "learning_rate": 2.9029900670438183e-06,
-      "loss": 0.52,
+      "epoch": 9.0840019515256,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.8926644464158464e-06,
+      "loss": 0.9656417083740234,
       "step": 789000
     },
     {
-      "epoch": 18.172607534309662,
-      "grad_norm": 3.4420907497406006,
-      "learning_rate": 2.895718503221303e-06,
-      "loss": 0.5262,
+      "epoch": 9.086304636600316,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8853928825933306e-06,
+      "loss": 0.9802579498291015,
       "step": 789200
     },
     {
-      "epoch": 18.17721285806392,
-      "grad_norm": 3.3978118896484375,
-      "learning_rate": 2.888446939398787e-06,
-      "loss": 0.5168,
+      "epoch": 9.088607321675031,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.8781213187708153e-06,
+      "loss": 0.9718341827392578,
       "step": 789400
     },
     {
-      "epoch": 18.181818181818183,
-      "grad_norm": 2.9462685585021973,
-      "learning_rate": 2.881175375576272e-06,
-      "loss": 0.5105,
+      "epoch": 9.090910006749745,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.870849754948299e-06,
+      "loss": 0.9539443206787109,
       "step": 789600
     },
     {
-      "epoch": 18.186423505572442,
-      "grad_norm": 2.9186863899230957,
-      "learning_rate": 2.8739038117537556e-06,
-      "loss": 0.5146,
+      "epoch": 9.09321269182446,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8635781911257837e-06,
+      "loss": 0.9921604919433594,
       "step": 789800
     },
     {
-      "epoch": 18.1910288293267,
-      "grad_norm": 2.7235682010650635,
-      "learning_rate": 2.8666322479312403e-06,
-      "loss": 0.5206,
+      "epoch": 9.095515376899176,
+      "grad_norm": 0.75,
+      "learning_rate": 2.856306627303268e-06,
+      "loss": 0.9690369415283203,
       "step": 790000
     },
     {
-      "epoch": 18.195634153080963,
-      "grad_norm": 3.430112361907959,
-      "learning_rate": 2.8593606841087245e-06,
-      "loss": 0.5105,
+      "epoch": 9.097818061973891,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.8490350634807526e-06,
+      "loss": 0.9706679534912109,
       "step": 790200
     },
     {
-      "epoch": 18.200239476835222,
-      "grad_norm": 3.3343279361724854,
-      "learning_rate": 2.852089120286209e-06,
-      "loss": 0.5144,
+      "epoch": 9.100120747048605,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.8417634996582364e-06,
+      "loss": 1.0010890197753906,
       "step": 790400
     },
     {
-      "epoch": 18.20484480058948,
-      "grad_norm": 3.0267772674560547,
-      "learning_rate": 2.844817556463693e-06,
-      "loss": 0.5233,
+      "epoch": 9.10242343212332,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.834491935835721e-06,
+      "loss": 0.9859672546386719,
       "step": 790600
     },
     {
-      "epoch": 18.209450124343743,
-      "grad_norm": 3.3748552799224854,
-      "learning_rate": 2.8375459926411776e-06,
-      "loss": 0.5131,
+      "epoch": 9.104726117198036,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.8272203720132053e-06,
+      "loss": 0.9713872528076172,
       "step": 790800
     },
     {
-      "epoch": 18.214055448098,
-      "grad_norm": 2.5174360275268555,
-      "learning_rate": 2.830274428818662e-06,
-      "loss": 0.5182,
+      "epoch": 9.107028802272751,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.81994880819069e-06,
+      "loss": 0.9894600677490234,
       "step": 791000
     },
     {
-      "epoch": 18.21866077185226,
-      "grad_norm": 3.363555669784546,
-      "learning_rate": 2.8230028649961465e-06,
-      "loss": 0.531,
+      "epoch": 9.109331487347465,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.8126772443681737e-06,
+      "loss": 0.9755573272705078,
       "step": 791200
     },
     {
-      "epoch": 18.223266095606522,
-      "grad_norm": 2.635622024536133,
-      "learning_rate": 2.8157313011736302e-06,
-      "loss": 0.5156,
+      "epoch": 9.11163417242218,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.8054056805456583e-06,
+      "loss": 0.9854010772705079,
       "step": 791400
     },
     {
-      "epoch": 18.22787141936078,
-      "grad_norm": 3.1038053035736084,
-      "learning_rate": 2.808459737351115e-06,
-      "loss": 0.52,
+      "epoch": 9.113936857496896,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.7981341167231426e-06,
+      "loss": 0.9738497161865234,
       "step": 791600
     },
     {
-      "epoch": 18.23247674311504,
-      "grad_norm": 3.0769765377044678,
-      "learning_rate": 2.801188173528599e-06,
-      "loss": 0.5156,
+      "epoch": 9.11623954257161,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.790862552900627e-06,
+      "loss": 0.9824601745605469,
       "step": 791800
     },
     {
-      "epoch": 18.237082066869302,
-      "grad_norm": 2.5016119480133057,
-      "learning_rate": 2.7939166097060838e-06,
-      "loss": 0.5227,
+      "epoch": 9.118542227646325,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.783590989078111e-06,
+      "loss": 0.9630702209472656,
       "step": 792000
     },
     {
-      "epoch": 18.24168739062356,
-      "grad_norm": 3.1567609310150146,
-      "learning_rate": 2.7866450458835676e-06,
-      "loss": 0.5215,
+      "epoch": 9.12084491272104,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7763194252555956e-06,
+      "loss": 0.9746056365966796,
       "step": 792200
     },
     {
-      "epoch": 18.24629271437782,
-      "grad_norm": 3.602930784225464,
-      "learning_rate": 2.779373482061052e-06,
-      "loss": 0.5226,
+      "epoch": 9.123147597795755,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.76904786143308e-06,
+      "loss": 0.9801070404052734,
       "step": 792400
     },
     {
-      "epoch": 18.250898038132082,
-      "grad_norm": 2.699720859527588,
-      "learning_rate": 2.7721019182385364e-06,
-      "loss": 0.5195,
+      "epoch": 9.125450282870469,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.761776297610564e-06,
+      "loss": 0.9643087005615234,
       "step": 792600
     },
     {
-      "epoch": 18.25550336188634,
-      "grad_norm": 2.6667487621307373,
-      "learning_rate": 2.764830354416021e-06,
-      "loss": 0.5111,
+      "epoch": 9.127752967945185,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.7545047337880483e-06,
+      "loss": 0.9930847930908203,
       "step": 792800
     },
     {
-      "epoch": 18.2601086856406,
-      "grad_norm": 3.1403419971466064,
-      "learning_rate": 2.757558790593505e-06,
-      "loss": 0.5139,
+      "epoch": 9.1300556530199,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.747233169965533e-06,
+      "loss": 0.9791233825683594,
       "step": 793000
     },
     {
-      "epoch": 18.26471400939486,
-      "grad_norm": 3.1445913314819336,
-      "learning_rate": 2.7502872267709895e-06,
-      "loss": 0.5169,
+      "epoch": 9.132358338094614,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.739961606143017e-06,
+      "loss": 0.9894448852539063,
       "step": 793200
     },
     {
-      "epoch": 18.26931933314912,
-      "grad_norm": 3.3420560359954834,
-      "learning_rate": 2.7430156629484737e-06,
-      "loss": 0.5188,
+      "epoch": 9.134661023169329,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.7326900423205014e-06,
+      "loss": 0.9905043792724609,
       "step": 793400
     },
     {
-      "epoch": 18.27392465690338,
-      "grad_norm": 3.481611490249634,
-      "learning_rate": 2.7357804569450706e-06,
-      "loss": 0.51,
+      "epoch": 9.136963708244044,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.7254184784979856e-06,
+      "loss": 0.9656953430175781,
       "step": 793600
     },
     {
-      "epoch": 18.27852998065764,
-      "grad_norm": 3.8973641395568848,
-      "learning_rate": 2.7285088931225553e-06,
-      "loss": 0.5136,
+      "epoch": 9.13926639331876,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7181469146754702e-06,
+      "loss": 0.9717328643798828,
       "step": 793800
     },
     {
-      "epoch": 18.2831353044119,
-      "grad_norm": 2.5481739044189453,
-      "learning_rate": 2.721237329300039e-06,
-      "loss": 0.5212,
+      "epoch": 9.141569078393474,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7108753508529545e-06,
+      "loss": 0.9814446258544922,
       "step": 794000
     },
     {
-      "epoch": 18.28774062816616,
-      "grad_norm": 2.917550802230835,
-      "learning_rate": 2.7139657654775237e-06,
-      "loss": 0.5216,
+      "epoch": 9.143871763468189,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.7036037870304387e-06,
+      "loss": 0.9948249816894531,
       "step": 794200
     },
     {
-      "epoch": 18.29234595192042,
-      "grad_norm": 3.176706314086914,
-      "learning_rate": 2.706694201655008e-06,
-      "loss": 0.5028,
+      "epoch": 9.146174448542904,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.696332223207923e-06,
+      "loss": 0.9781787109375,
       "step": 794400
     },
     {
-      "epoch": 18.29695127567468,
-      "grad_norm": 3.2188303470611572,
-      "learning_rate": 2.699458995651605e-06,
-      "loss": 0.52,
+      "epoch": 9.14847713361762,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.6890606593854075e-06,
+      "loss": 0.9804864501953126,
       "step": 794600
     },
     {
-      "epoch": 18.301556599428938,
-      "grad_norm": 3.91021466255188,
-      "learning_rate": 2.6921874318290895e-06,
-      "loss": 0.5127,
+      "epoch": 9.150779818692333,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.6817890955628918e-06,
+      "loss": 0.9924549102783203,
       "step": 794800
     },
     {
-      "epoch": 18.3061619231832,
-      "grad_norm": 3.8334686756134033,
-      "learning_rate": 2.6849158680065733e-06,
-      "loss": 0.5062,
+      "epoch": 9.153082503767049,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.674517531740376e-06,
+      "loss": 0.9628752899169922,
       "step": 795000
     },
     {
-      "epoch": 18.31076724693746,
-      "grad_norm": 3.018327236175537,
-      "learning_rate": 2.677644304184058e-06,
-      "loss": 0.5228,
+      "epoch": 9.155385188841764,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.6672459679178606e-06,
+      "loss": 0.994937744140625,
       "step": 795200
     },
     {
-      "epoch": 18.31537257069172,
-      "grad_norm": 4.303511619567871,
-      "learning_rate": 2.670372740361542e-06,
-      "loss": 0.5138,
+      "epoch": 9.157687873916478,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.659974404095345e-06,
+      "loss": 0.9820407104492187,
       "step": 795400
     },
     {
-      "epoch": 18.31997789444598,
-      "grad_norm": 3.589290142059326,
-      "learning_rate": 2.6631011765390268e-06,
-      "loss": 0.5281,
+      "epoch": 9.159990558991193,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.6527028402728295e-06,
+      "loss": 1.0036585998535157,
       "step": 795600
     },
     {
-      "epoch": 18.32458321820024,
-      "grad_norm": 2.7188146114349365,
-      "learning_rate": 2.6558296127165106e-06,
-      "loss": 0.5093,
+      "epoch": 9.162293244065909,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.6454312764503133e-06,
+      "loss": 0.9978278350830078,
       "step": 795800
     },
     {
-      "epoch": 18.3291885419545,
-      "grad_norm": 3.7704238891601562,
-      "learning_rate": 2.6485580488939952e-06,
-      "loss": 0.5106,
+      "epoch": 9.164595929140624,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.638159712627798e-06,
+      "loss": 0.9959989929199219,
       "step": 796000
     },
     {
-      "epoch": 18.33379386570876,
-      "grad_norm": 2.9038262367248535,
-      "learning_rate": 2.6412864850714794e-06,
-      "loss": 0.5047,
+      "epoch": 9.166898614215338,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.630888148805282e-06,
+      "loss": 0.9977657318115234,
       "step": 796200
     },
     {
-      "epoch": 18.33839918946302,
-      "grad_norm": 3.1424877643585205,
-      "learning_rate": 2.634014921248964e-06,
-      "loss": 0.5227,
+      "epoch": 9.169201299290053,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6236165849827668e-06,
+      "loss": 0.9684986877441406,
       "step": 796400
     },
     {
-      "epoch": 18.34300451321728,
-      "grad_norm": 3.6267952919006348,
-      "learning_rate": 2.626743357426448e-06,
-      "loss": 0.5094,
+      "epoch": 9.171503984364769,
+      "grad_norm": 1.0,
+      "learning_rate": 2.6163450211602506e-06,
+      "loss": 1.0095248413085938,
       "step": 796600
     },
     {
-      "epoch": 18.34760983697154,
-      "grad_norm": 3.1225740909576416,
-      "learning_rate": 2.6194717936039325e-06,
-      "loss": 0.5222,
+      "epoch": 9.173806669439482,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.6090734573377352e-06,
+      "loss": 0.9658311462402344,
       "step": 796800
     },
     {
-      "epoch": 18.352215160725798,
-      "grad_norm": 3.5148890018463135,
-      "learning_rate": 2.612200229781417e-06,
-      "loss": 0.5134,
+      "epoch": 9.176109354514198,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.6018018935152194e-06,
+      "loss": 0.9893347930908203,
       "step": 797000
     },
     {
-      "epoch": 18.35682048448006,
-      "grad_norm": 3.0749411582946777,
-      "learning_rate": 2.6049286659589014e-06,
-      "loss": 0.5055,
+      "epoch": 9.178412039588913,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.594530329692704e-06,
+      "loss": 0.9784815216064453,
       "step": 797200
     },
     {
-      "epoch": 18.36142580823432,
-      "grad_norm": 3.1104702949523926,
-      "learning_rate": 2.5976571021363856e-06,
-      "loss": 0.5224,
+      "epoch": 9.180714724663629,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.587258765870188e-06,
+      "loss": 0.984940185546875,
       "step": 797400
     },
     {
-      "epoch": 18.366031131988578,
-      "grad_norm": 4.1002020835876465,
-      "learning_rate": 2.59038553831387e-06,
-      "loss": 0.5125,
+      "epoch": 9.183017409738342,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.5799872020476725e-06,
+      "loss": 0.9787003326416016,
       "step": 797600
     },
     {
-      "epoch": 18.37063645574284,
-      "grad_norm": 3.385798692703247,
-      "learning_rate": 2.5831139744913545e-06,
-      "loss": 0.5238,
+      "epoch": 9.185320094813058,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.5727156382251567e-06,
+      "loss": 0.9814534759521485,
       "step": 797800
     },
     {
-      "epoch": 18.3752417794971,
-      "grad_norm": 2.667778253555298,
-      "learning_rate": 2.5758424106688383e-06,
-      "loss": 0.519,
+      "epoch": 9.187622779887773,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.5654440744026414e-06,
+      "loss": 0.9823260498046875,
       "step": 798000
     },
     {
-      "epoch": 18.379847103251358,
-      "grad_norm": 3.486220121383667,
-      "learning_rate": 2.568570846846323e-06,
-      "loss": 0.5123,
+      "epoch": 9.189925464962489,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.558172510580125e-06,
+      "loss": 0.9531006622314453,
       "step": 798200
     },
     {
-      "epoch": 18.38445242700562,
-      "grad_norm": 2.9387404918670654,
-      "learning_rate": 2.56133564084292e-06,
-      "loss": 0.5075,
+      "epoch": 9.192228150037202,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.55090094675761e-06,
+      "loss": 0.9803681182861328,
       "step": 798400
     },
     {
-      "epoch": 18.38905775075988,
-      "grad_norm": 3.13200044631958,
-      "learning_rate": 2.554064077020404e-06,
-      "loss": 0.5107,
+      "epoch": 9.194530835111918,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.543629382935094e-06,
+      "loss": 0.962587890625,
       "step": 798600
     },
     {
-      "epoch": 18.393663074514137,
-      "grad_norm": 2.374018669128418,
-      "learning_rate": 2.5467925131978887e-06,
-      "loss": 0.5245,
+      "epoch": 9.196833520186633,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.5363578191125787e-06,
+      "loss": 0.9822435760498047,
       "step": 798800
     },
     {
-      "epoch": 18.3982683982684,
-      "grad_norm": 3.3281686305999756,
-      "learning_rate": 2.539520949375373e-06,
-      "loss": 0.5111,
+      "epoch": 9.199136205261347,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5290862552900625e-06,
+      "loss": 0.9760208129882812,
       "step": 799000
     },
     {
-      "epoch": 18.402873722022658,
-      "grad_norm": 2.587111711502075,
-      "learning_rate": 2.532249385552857e-06,
-      "loss": 0.5144,
+      "epoch": 9.201438890336062,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.521814691467547e-06,
+      "loss": 0.9792575836181641,
       "step": 799200
     },
     {
-      "epoch": 18.407479045776917,
-      "grad_norm": 2.7880938053131104,
-      "learning_rate": 2.5249778217303413e-06,
-      "loss": 0.5129,
+      "epoch": 9.203741575410778,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5145431276450313e-06,
+      "loss": 0.9941600036621093,
       "step": 799400
     },
     {
-      "epoch": 18.41208436953118,
-      "grad_norm": 2.612704038619995,
-      "learning_rate": 2.517706257907826e-06,
-      "loss": 0.5169,
+      "epoch": 9.206044260485493,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.507271563822516e-06,
+      "loss": 0.9764888763427735,
       "step": 799600
     },
     {
-      "epoch": 18.416689693285438,
-      "grad_norm": 3.7261786460876465,
-      "learning_rate": 2.5104346940853098e-06,
-      "loss": 0.5128,
+      "epoch": 9.208346945560207,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 0.9621796417236328,
       "step": 799800
     },
     {
-      "epoch": 18.421295017039697,
-      "grad_norm": 3.7250654697418213,
-      "learning_rate": 2.5031631302627944e-06,
-      "loss": 0.5108,
+      "epoch": 9.210649630634922,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.4927284361774844e-06,
+      "loss": 0.9800115203857422,
       "step": 800000
     },
     {
-      "epoch": 18.42590034079396,
-      "grad_norm": 3.5736753940582275,
-      "learning_rate": 2.4958915664402786e-06,
-      "loss": 0.5307,
+      "epoch": 9.212952315709638,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.4854568723549686e-06,
+      "loss": 0.9602899169921875,
       "step": 800200
     },
     {
-      "epoch": 18.430505664548217,
-      "grad_norm": 2.80611515045166,
-      "learning_rate": 2.4886200026177633e-06,
-      "loss": 0.5141,
+      "epoch": 9.215255000784353,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4781853085324533e-06,
+      "loss": 0.9630294036865235,
       "step": 800400
     },
     {
-      "epoch": 18.435110988302476,
-      "grad_norm": 3.6763579845428467,
-      "learning_rate": 2.481348438795247e-06,
-      "loss": 0.5067,
+      "epoch": 9.217557685859067,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.470913744709937e-06,
+      "loss": 0.9817597961425781,
       "step": 800600
     },
     {
-      "epoch": 18.43971631205674,
-      "grad_norm": 2.915642023086548,
-      "learning_rate": 2.4740768749727317e-06,
-      "loss": 0.5219,
+      "epoch": 9.219860370933782,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4636421808874217e-06,
+      "loss": 0.9933039093017578,
       "step": 800800
     },
     {
-      "epoch": 18.444321635810997,
-      "grad_norm": 4.065909385681152,
-      "learning_rate": 2.466805311150216e-06,
-      "loss": 0.5083,
+      "epoch": 9.222163056008498,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.456370617064906e-06,
+      "loss": 0.9749951171875,
       "step": 801000
     },
     {
-      "epoch": 18.448926959565256,
-      "grad_norm": 3.1720340251922607,
-      "learning_rate": 2.4595337473277006e-06,
-      "loss": 0.5222,
+      "epoch": 9.224465741083211,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.4490990532423906e-06,
+      "loss": 0.9645835113525391,
       "step": 801200
     },
     {
-      "epoch": 18.453532283319518,
-      "grad_norm": 2.6975619792938232,
-      "learning_rate": 2.4522621835051844e-06,
-      "loss": 0.5067,
+      "epoch": 9.226768426157927,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.4418274894198744e-06,
+      "loss": 0.9951029968261719,
       "step": 801400
     },
     {
-      "epoch": 18.458137607073777,
-      "grad_norm": 2.7786128520965576,
-      "learning_rate": 2.444990619682669e-06,
-      "loss": 0.5181,
+      "epoch": 9.229071111232642,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.434555925597359e-06,
+      "loss": 0.9582917785644531,
       "step": 801600
     },
     {
-      "epoch": 18.462742930828036,
-      "grad_norm": 2.5183842182159424,
-      "learning_rate": 2.4377190558601532e-06,
-      "loss": 0.5166,
+      "epoch": 9.231373796307357,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.4272843617748432e-06,
+      "loss": 0.9795014953613281,
       "step": 801800
     },
     {
-      "epoch": 18.467348254582298,
-      "grad_norm": 2.8333635330200195,
-      "learning_rate": 2.430447492037638e-06,
-      "loss": 0.5163,
+      "epoch": 9.233676481382071,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.420012797952328e-06,
+      "loss": 0.9806991577148437,
       "step": 802000
     },
     {
-      "epoch": 18.471953578336556,
-      "grad_norm": 3.31788969039917,
-      "learning_rate": 2.4231759282151217e-06,
-      "loss": 0.518,
+      "epoch": 9.235979166456787,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.412741234129812e-06,
+      "loss": 0.9899074554443359,
       "step": 802200
     },
     {
-      "epoch": 18.476558902090815,
-      "grad_norm": 3.1232898235321045,
-      "learning_rate": 2.4159043643926063e-06,
-      "loss": 0.5137,
+      "epoch": 9.238281851531502,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4054696703072963e-06,
+      "loss": 0.9765318298339843,
       "step": 802400
     },
     {
-      "epoch": 18.481164225845077,
-      "grad_norm": 3.5303895473480225,
-      "learning_rate": 2.4086691583892032e-06,
-      "loss": 0.5241,
+      "epoch": 9.240584536606216,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.398198106484781e-06,
+      "loss": 1.0049002075195312,
       "step": 802600
     },
     {
-      "epoch": 18.485769549599336,
-      "grad_norm": 2.972893714904785,
-      "learning_rate": 2.4013975945666874e-06,
-      "loss": 0.5078,
+      "epoch": 9.242887221680931,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.390926542662265e-06,
+      "loss": 1.0022148895263672,
       "step": 802800
     },
     {
-      "epoch": 18.4903748733536,
-      "grad_norm": 2.6240293979644775,
-      "learning_rate": 2.394126030744172e-06,
-      "loss": 0.5166,
+      "epoch": 9.245189906755646,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3836549788397494e-06,
+      "loss": 0.9906528472900391,
       "step": 803000
     },
     {
-      "epoch": 18.494980197107857,
-      "grad_norm": 2.6254336833953857,
-      "learning_rate": 2.386890824740769e-06,
-      "loss": 0.5195,
+      "epoch": 9.247492591830362,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.3763834150172336e-06,
+      "loss": 0.9908769989013672,
       "step": 803200
     },
     {
-      "epoch": 18.499585520862116,
-      "grad_norm": 3.5183451175689697,
-      "learning_rate": 2.3796192609182528e-06,
-      "loss": 0.5125,
+      "epoch": 9.249795276905076,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3691118511947183e-06,
+      "loss": 0.9784947967529297,
       "step": 803400
     },
     {
-      "epoch": 18.504190844616378,
-      "grad_norm": 2.951514720916748,
-      "learning_rate": 2.3723476970957374e-06,
-      "loss": 0.515,
+      "epoch": 9.252097961979791,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.3618402873722025e-06,
+      "loss": 0.9808582305908203,
       "step": 803600
     },
     {
-      "epoch": 18.508796168370637,
-      "grad_norm": 3.723506450653076,
-      "learning_rate": 2.3650761332732216e-06,
-      "loss": 0.5108,
+      "epoch": 9.254400647054506,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.3545687235496867e-06,
+      "loss": 0.9718049621582031,
       "step": 803800
     },
     {
-      "epoch": 18.513401492124895,
-      "grad_norm": 2.8375089168548584,
-      "learning_rate": 2.3578045694507063e-06,
-      "loss": 0.5099,
+      "epoch": 9.25670333212922,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.347297159727171e-06,
+      "loss": 0.999382095336914,
       "step": 804000
     },
     {
-      "epoch": 18.518006815879158,
-      "grad_norm": 2.5292952060699463,
-      "learning_rate": 2.35053300562819e-06,
-      "loss": 0.5176,
+      "epoch": 9.259006017203935,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.3400255959046556e-06,
+      "loss": 0.9925857543945312,
       "step": 804200
     },
     {
-      "epoch": 18.522612139633416,
-      "grad_norm": 2.64688777923584,
-      "learning_rate": 2.3432614418056747e-06,
-      "loss": 0.5067,
+      "epoch": 9.26130870227865,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.3327540320821394e-06,
+      "loss": 0.9597567749023438,
       "step": 804400
     },
     {
-      "epoch": 18.527217463387675,
-      "grad_norm": 3.6783273220062256,
-      "learning_rate": 2.335989877983159e-06,
-      "loss": 0.5204,
+      "epoch": 9.263611387353366,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.325482468259624e-06,
+      "loss": 0.9851140594482422,
       "step": 804600
     },
     {
-      "epoch": 18.531822787141937,
-      "grad_norm": 3.5552361011505127,
-      "learning_rate": 2.3287183141606436e-06,
-      "loss": 0.5043,
+      "epoch": 9.26591407242808,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.3182109044371082e-06,
+      "loss": 0.9718254089355469,
       "step": 804800
     },
     {
-      "epoch": 18.536428110896196,
-      "grad_norm": 3.2236578464508057,
-      "learning_rate": 2.3214467503381274e-06,
-      "loss": 0.5089,
+      "epoch": 9.268216757502795,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.310939340614593e-06,
+      "loss": 0.9673877716064453,
       "step": 805000
     },
     {
-      "epoch": 18.541033434650455,
-      "grad_norm": 3.2282896041870117,
-      "learning_rate": 2.314175186515612e-06,
-      "loss": 0.5058,
+      "epoch": 9.27051944257751,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.3036677767920767e-06,
+      "loss": 0.9953539276123047,
       "step": 805200
     },
     {
-      "epoch": 18.545638758404717,
-      "grad_norm": 2.5554604530334473,
-      "learning_rate": 2.3069036226930963e-06,
-      "loss": 0.5093,
+      "epoch": 9.272822127652226,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.2963962129695613e-06,
+      "loss": 0.9628050231933594,
       "step": 805400
     },
     {
-      "epoch": 18.550244082158976,
-      "grad_norm": 3.2304649353027344,
-      "learning_rate": 2.299632058870581e-06,
-      "loss": 0.5146,
+      "epoch": 9.27512481272694,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.2891246491470455e-06,
+      "loss": 0.9810200500488281,
       "step": 805600
     },
     {
-      "epoch": 18.554849405913235,
-      "grad_norm": 4.473756790161133,
-      "learning_rate": 2.292360495048065e-06,
-      "loss": 0.505,
+      "epoch": 9.277427497801655,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.28185308532453e-06,
+      "loss": 0.9924951171875,
       "step": 805800
     },
     {
-      "epoch": 18.559454729667497,
-      "grad_norm": 3.0843703746795654,
-      "learning_rate": 2.2850889312255493e-06,
-      "loss": 0.5167,
+      "epoch": 9.27973018287637,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.274581521502014e-06,
+      "loss": 0.9816727447509765,
       "step": 806000
     },
     {
-      "epoch": 18.564060053421755,
-      "grad_norm": 3.773874044418335,
-      "learning_rate": 2.277817367403034e-06,
-      "loss": 0.5102,
+      "epoch": 9.282032867951084,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.2673099576794986e-06,
+      "loss": 0.9924248504638672,
       "step": 806200
     },
     {
-      "epoch": 18.568665377176014,
-      "grad_norm": 3.845970630645752,
-      "learning_rate": 2.270545803580518e-06,
-      "loss": 0.5235,
+      "epoch": 9.2843355530258,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.260038393856983e-06,
+      "loss": 0.9997078704833985,
       "step": 806400
     },
     {
-      "epoch": 18.573270700930276,
-      "grad_norm": 3.6182453632354736,
-      "learning_rate": 2.2632742397580024e-06,
-      "loss": 0.5057,
+      "epoch": 9.286638238100515,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2527668300344675e-06,
+      "loss": 0.9612949371337891,
       "step": 806600
     },
     {
-      "epoch": 18.577876024684535,
-      "grad_norm": 3.738835573196411,
-      "learning_rate": 2.2560026759354866e-06,
-      "loss": 0.5207,
+      "epoch": 9.28894092317523,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.2454952662119513e-06,
+      "loss": 0.9775658416748046,
       "step": 806800
     },
     {
-      "epoch": 18.582481348438794,
-      "grad_norm": 3.5557377338409424,
-      "learning_rate": 2.2487311121129713e-06,
-      "loss": 0.529,
+      "epoch": 9.291243608249944,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.238223702389436e-06,
+      "loss": 0.9638466644287109,
       "step": 807000
     },
     {
-      "epoch": 18.587086672193056,
-      "grad_norm": 3.059429168701172,
-      "learning_rate": 2.2414595482904555e-06,
-      "loss": 0.5075,
+      "epoch": 9.29354629332466,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.23095213856692e-06,
+      "loss": 0.9751538848876953,
       "step": 807200
     },
     {
-      "epoch": 18.591691995947315,
-      "grad_norm": 3.12056303024292,
-      "learning_rate": 2.2342243422870524e-06,
-      "loss": 0.5114,
+      "epoch": 9.295848978399375,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.2236805747444048e-06,
+      "loss": 0.986815185546875,
       "step": 807400
     },
     {
-      "epoch": 18.596297319701574,
-      "grad_norm": 3.1706645488739014,
-      "learning_rate": 2.2269527784645366e-06,
-      "loss": 0.5217,
+      "epoch": 9.29815166347409,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.2164090109218886e-06,
+      "loss": 0.9854920959472656,
       "step": 807600
     },
     {
-      "epoch": 18.600902643455836,
-      "grad_norm": 3.850172281265259,
-      "learning_rate": 2.219681214642021e-06,
-      "loss": 0.5137,
+      "epoch": 9.300454348548804,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.2091374470993732e-06,
+      "loss": 0.997869873046875,
       "step": 807800
     },
     {
-      "epoch": 18.605507967210094,
-      "grad_norm": 3.080021858215332,
-      "learning_rate": 2.2124096508195055e-06,
-      "loss": 0.5139,
+      "epoch": 9.30275703362352,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.2018658832768574e-06,
+      "loss": 0.975609130859375,
       "step": 808000
     },
     {
-      "epoch": 18.610113290964353,
-      "grad_norm": 2.29182505607605,
-      "learning_rate": 2.2051380869969897e-06,
-      "loss": 0.5026,
+      "epoch": 9.305059718698235,
+      "grad_norm": 0.75,
+      "learning_rate": 2.194594319454342e-06,
+      "loss": 1.0007555389404297,
       "step": 808200
     },
     {
-      "epoch": 18.614718614718615,
-      "grad_norm": 3.1314103603363037,
-      "learning_rate": 2.197866523174474e-06,
-      "loss": 0.5123,
+      "epoch": 9.307362403772949,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.187322755631826e-06,
+      "loss": 0.9574895477294922,
       "step": 808400
     },
     {
-      "epoch": 18.619323938472874,
-      "grad_norm": 3.334752082824707,
-      "learning_rate": 2.190594959351958e-06,
-      "loss": 0.5101,
+      "epoch": 9.309665088847664,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1800511918093105e-06,
+      "loss": 0.9769926452636719,
       "step": 808600
     },
     {
-      "epoch": 18.623929262227133,
-      "grad_norm": 3.3216257095336914,
-      "learning_rate": 2.1833233955294428e-06,
-      "loss": 0.5094,
+      "epoch": 9.31196777392238,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.1727796279867947e-06,
+      "loss": 0.9835977935791016,
       "step": 808800
     },
     {
-      "epoch": 18.628534585981395,
-      "grad_norm": 2.6570966243743896,
-      "learning_rate": 2.176051831706927e-06,
-      "loss": 0.5118,
+      "epoch": 9.314270458997095,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.1655080641642794e-06,
+      "loss": 0.98232421875,
       "step": 809000
     },
     {
-      "epoch": 18.633139909735654,
-      "grad_norm": 3.783985137939453,
-      "learning_rate": 2.1687802678844112e-06,
-      "loss": 0.5226,
+      "epoch": 9.316573144071809,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.1582365003417636e-06,
+      "loss": 0.9904391479492187,
       "step": 809200
     },
     {
-      "epoch": 18.637745233489916,
-      "grad_norm": 3.1615185737609863,
-      "learning_rate": 2.1615087040618954e-06,
-      "loss": 0.5194,
+      "epoch": 9.318875829146524,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.150964936519248e-06,
+      "loss": 0.977032470703125,
       "step": 809400
     },
     {
-      "epoch": 18.642350557244175,
-      "grad_norm": 3.571136951446533,
-      "learning_rate": 2.15423714023938e-06,
-      "loss": 0.5248,
+      "epoch": 9.32117851422124,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1436933726967325e-06,
+      "loss": 0.9892449188232422,
       "step": 809600
     },
     {
-      "epoch": 18.646955880998433,
-      "grad_norm": 2.769298553466797,
-      "learning_rate": 2.147001934235977e-06,
-      "loss": 0.5104,
+      "epoch": 9.323481199295953,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1364218088742167e-06,
+      "loss": 0.9981741333007812,
       "step": 809800
     },
     {
-      "epoch": 18.651561204752696,
-      "grad_norm": 2.678037166595459,
-      "learning_rate": 2.1397303704134612e-06,
-      "loss": 0.5246,
+      "epoch": 9.325783884370669,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.129150245051701e-06,
+      "loss": 0.9746675109863281,
       "step": 810000
     },
     {
-      "epoch": 18.656166528506954,
-      "grad_norm": 3.228646993637085,
-      "learning_rate": 2.1324588065909454e-06,
-      "loss": 0.502,
+      "epoch": 9.328086569445384,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.121878681229185e-06,
+      "loss": 0.9943876647949219,
       "step": 810200
     },
     {
-      "epoch": 18.660771852261213,
-      "grad_norm": 2.839290142059326,
-      "learning_rate": 2.1252236005875423e-06,
-      "loss": 0.5132,
+      "epoch": 9.3303892545201,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.1146071174066698e-06,
+      "loss": 0.9812979888916016,
       "step": 810400
     },
     {
-      "epoch": 18.665377176015475,
-      "grad_norm": 2.865377187728882,
-      "learning_rate": 2.1179520367650266e-06,
-      "loss": 0.5063,
+      "epoch": 9.332691939594813,
+      "grad_norm": 0.625,
+      "learning_rate": 2.107335553584154e-06,
+      "loss": 0.9795059967041015,
       "step": 810600
     },
     {
-      "epoch": 18.669982499769734,
-      "grad_norm": 4.583608150482178,
-      "learning_rate": 2.110680472942511e-06,
-      "loss": 0.5155,
+      "epoch": 9.334994624669529,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.100063989761638e-06,
+      "loss": 0.9666686248779297,
       "step": 810800
     },
     {
-      "epoch": 18.674587823523993,
-      "grad_norm": 2.7023823261260986,
-      "learning_rate": 2.1034089091199954e-06,
-      "loss": 0.5033,
+      "epoch": 9.337297309744244,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.0927924259391224e-06,
+      "loss": 0.9806160736083984,
       "step": 811000
     },
     {
-      "epoch": 18.679193147278255,
-      "grad_norm": 2.9620158672332764,
-      "learning_rate": 2.0961373452974796e-06,
-      "loss": 0.5184,
+      "epoch": 9.33959999481896,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.085520862116607e-06,
+      "loss": 0.9875475311279297,
       "step": 811200
     },
     {
-      "epoch": 18.683798471032514,
-      "grad_norm": 2.519291639328003,
-      "learning_rate": 2.088865781474964e-06,
-      "loss": 0.5126,
+      "epoch": 9.341902679893673,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.0782492982940913e-06,
+      "loss": 0.9789799499511719,
       "step": 811400
     },
     {
-      "epoch": 18.688403794786772,
-      "grad_norm": 2.743424415588379,
-      "learning_rate": 2.0815942176524485e-06,
-      "loss": 0.509,
+      "epoch": 9.344205364968388,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0709777344715755e-06,
+      "loss": 0.9655042266845704,
       "step": 811600
     },
     {
-      "epoch": 18.693009118541035,
-      "grad_norm": 3.654670000076294,
-      "learning_rate": 2.0743226538299327e-06,
-      "loss": 0.5155,
+      "epoch": 9.346508050043104,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.0637061706490597e-06,
+      "loss": 0.9765595245361328,
       "step": 811800
     },
     {
-      "epoch": 18.697614442295293,
-      "grad_norm": 3.285017251968384,
-      "learning_rate": 2.067051090007417e-06,
-      "loss": 0.5126,
+      "epoch": 9.348810735117818,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.0564346068265444e-06,
+      "loss": 0.9863420867919922,
       "step": 812000
     },
     {
-      "epoch": 18.702219766049552,
-      "grad_norm": 2.268007755279541,
-      "learning_rate": 2.059779526184901e-06,
-      "loss": 0.5104,
+      "epoch": 9.351113420192533,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0491630430040286e-06,
+      "loss": 0.987475357055664,
       "step": 812200
     },
     {
-      "epoch": 18.706825089803814,
-      "grad_norm": 3.560014247894287,
-      "learning_rate": 2.052507962362386e-06,
-      "loss": 0.5294,
+      "epoch": 9.353416105267248,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.041891479181513e-06,
+      "loss": 0.9715730285644532,
       "step": 812400
     },
     {
-      "epoch": 18.711430413558073,
-      "grad_norm": 2.835283041000366,
-      "learning_rate": 2.04523639853987e-06,
-      "loss": 0.5132,
+      "epoch": 9.355718790341964,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.034619915358997e-06,
+      "loss": 0.9703129577636719,
       "step": 812600
     },
     {
-      "epoch": 18.716035737312332,
-      "grad_norm": 2.768789291381836,
-      "learning_rate": 2.0379648347173542e-06,
-      "loss": 0.5094,
+      "epoch": 9.358021475416678,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.0273483515364817e-06,
+      "loss": 0.9737438201904297,
       "step": 812800
     },
     {
-      "epoch": 18.720641061066594,
-      "grad_norm": 2.965252637863159,
-      "learning_rate": 2.0306932708948385e-06,
-      "loss": 0.5031,
+      "epoch": 9.360324160491393,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.020076787713966e-06,
+      "loss": 0.9923829650878906,
       "step": 813000
     },
     {
-      "epoch": 18.725246384820853,
-      "grad_norm": 3.548333168029785,
-      "learning_rate": 2.023421707072323e-06,
-      "loss": 0.5152,
+      "epoch": 9.362626845566108,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.01280522389145e-06,
+      "loss": 0.9827593994140625,
       "step": 813200
     },
     {
-      "epoch": 18.72985170857511,
-      "grad_norm": 2.307636260986328,
-      "learning_rate": 2.0161501432498073e-06,
-      "loss": 0.5115,
+      "epoch": 9.364929530640822,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0055336600689343e-06,
+      "loss": 0.9942121887207032,
       "step": 813400
     },
     {
-      "epoch": 18.734457032329374,
-      "grad_norm": 3.0482430458068848,
-      "learning_rate": 2.0088785794272915e-06,
-      "loss": 0.5234,
+      "epoch": 9.367232215715537,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.998262096246419e-06,
+      "loss": 0.9751979064941406,
       "step": 813600
     },
     {
-      "epoch": 18.739062356083632,
-      "grad_norm": 2.7281389236450195,
-      "learning_rate": 2.0016070156047758e-06,
-      "loss": 0.5183,
+      "epoch": 9.369534900790253,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.990990532423903e-06,
+      "loss": 0.9989577484130859,
       "step": 813800
     },
     {
-      "epoch": 18.74366767983789,
-      "grad_norm": 3.274071216583252,
-      "learning_rate": 1.9943354517822604e-06,
-      "loss": 0.5144,
+      "epoch": 9.371837585864968,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9837189686013874e-06,
+      "loss": 0.9913412475585938,
       "step": 814000
     },
     {
-      "epoch": 18.748273003592153,
-      "grad_norm": 4.017879486083984,
-      "learning_rate": 1.9870638879597446e-06,
-      "loss": 0.5162,
+      "epoch": 9.374140270939682,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9764474047788716e-06,
+      "loss": 0.99336669921875,
       "step": 814200
     },
     {
-      "epoch": 18.752878327346412,
-      "grad_norm": 3.111361265182495,
-      "learning_rate": 1.9798286819563415e-06,
-      "loss": 0.5183,
+      "epoch": 9.376442956014397,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9691758409563563e-06,
+      "loss": 0.9956826782226562,
       "step": 814400
     },
     {
-      "epoch": 18.75748365110067,
-      "grad_norm": 2.9756317138671875,
-      "learning_rate": 1.9725571181338258e-06,
-      "loss": 0.516,
+      "epoch": 9.378745641089113,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9619042771338405e-06,
+      "loss": 0.9798301696777344,
       "step": 814600
     },
     {
-      "epoch": 18.762088974854933,
-      "grad_norm": 2.925050973892212,
-      "learning_rate": 1.96528555431131e-06,
-      "loss": 0.5216,
+      "epoch": 9.381048326163828,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9546327133113247e-06,
+      "loss": 0.9882535552978515,
       "step": 814800
     },
     {
-      "epoch": 18.766694298609192,
-      "grad_norm": 3.605431079864502,
-      "learning_rate": 1.9580139904887946e-06,
-      "loss": 0.5169,
+      "epoch": 9.383351011238542,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.947361149488809e-06,
+      "loss": 0.9871403503417969,
       "step": 815000
     },
     {
-      "epoch": 18.77129962236345,
-      "grad_norm": 4.39565372467041,
-      "learning_rate": 1.9507424266662793e-06,
-      "loss": 0.5192,
+      "epoch": 9.385653696313257,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9400895856662936e-06,
+      "loss": 0.9676121520996094,
       "step": 815200
     },
     {
-      "epoch": 18.775904946117713,
-      "grad_norm": 2.9125242233276367,
-      "learning_rate": 1.943470862843763e-06,
-      "loss": 0.5228,
+      "epoch": 9.387956381387973,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9328180218437774e-06,
+      "loss": 0.9921231842041016,
       "step": 815400
     },
     {
-      "epoch": 18.78051026987197,
-      "grad_norm": 2.8098058700561523,
-      "learning_rate": 1.9361992990212477e-06,
-      "loss": 0.5089,
+      "epoch": 9.390259066462686,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.925546458021262e-06,
+      "loss": 0.9892926025390625,
       "step": 815600
     },
     {
-      "epoch": 18.78511559362623,
-      "grad_norm": 3.1895477771759033,
-      "learning_rate": 1.928927735198732e-06,
-      "loss": 0.5062,
+      "epoch": 9.392561751537402,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9182748941987466e-06,
+      "loss": 1.0030435943603515,
       "step": 815800
     },
     {
-      "epoch": 18.789720917380492,
-      "grad_norm": 3.4902753829956055,
-      "learning_rate": 1.9216561713762166e-06,
-      "loss": 0.5108,
+      "epoch": 9.394864436612117,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.911003330376231e-06,
+      "loss": 0.9686942291259766,
       "step": 816000
     },
     {
-      "epoch": 18.79432624113475,
-      "grad_norm": 2.913203239440918,
-      "learning_rate": 1.9143846075537004e-06,
-      "loss": 0.5183,
+      "epoch": 9.397167121686833,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9037317665537149e-06,
+      "loss": 0.9933390808105469,
       "step": 816200
     },
     {
-      "epoch": 18.798931564889013,
-      "grad_norm": 3.108332872390747,
-      "learning_rate": 1.9071130437311848e-06,
-      "loss": 0.514,
+      "epoch": 9.399469806761546,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8964602027311993e-06,
+      "loss": 0.9900911712646484,
       "step": 816400
     },
     {
-      "epoch": 18.803536888643272,
-      "grad_norm": 3.0031442642211914,
-      "learning_rate": 1.8998778377277817e-06,
-      "loss": 0.5184,
+      "epoch": 9.401772491836262,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.8891886389086837e-06,
+      "loss": 0.9867049407958984,
       "step": 816600
     },
     {
-      "epoch": 18.80814221239753,
-      "grad_norm": 3.7675793170928955,
-      "learning_rate": 1.8926062739052661e-06,
-      "loss": 0.5141,
+      "epoch": 9.404075176910977,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.8819170750861682e-06,
+      "loss": 0.9971127319335937,
       "step": 816800
     },
     {
-      "epoch": 18.812747536151793,
-      "grad_norm": 3.3338944911956787,
-      "learning_rate": 1.8853347100827506e-06,
-      "loss": 0.5124,
+      "epoch": 9.406377861985693,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8746455112636524e-06,
+      "loss": 0.9814208221435546,
       "step": 817000
     },
     {
-      "epoch": 18.81735285990605,
-      "grad_norm": 4.226090908050537,
-      "learning_rate": 1.8780631462602346e-06,
-      "loss": 0.5115,
+      "epoch": 9.408680547060406,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8673739474411366e-06,
+      "loss": 0.9804402160644531,
       "step": 817200
     },
     {
-      "epoch": 18.82195818366031,
-      "grad_norm": 3.7023580074310303,
-      "learning_rate": 1.870791582437719e-06,
-      "loss": 0.5124,
+      "epoch": 9.410983232135122,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.860102383618621e-06,
+      "loss": 0.9882728576660156,
       "step": 817400
     },
     {
-      "epoch": 18.826563507414573,
-      "grad_norm": 2.838956832885742,
-      "learning_rate": 1.8635200186152034e-06,
-      "loss": 0.524,
+      "epoch": 9.413285917209837,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8528308197961053e-06,
+      "loss": 0.9980832672119141,
       "step": 817600
     },
     {
-      "epoch": 18.83116883116883,
-      "grad_norm": 3.4643635749816895,
-      "learning_rate": 1.8562484547926876e-06,
-      "loss": 0.5137,
+      "epoch": 9.41558860228455,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8455592559735897e-06,
+      "loss": 0.984637222290039,
       "step": 817800
     },
     {
-      "epoch": 18.83577415492309,
-      "grad_norm": 3.38789963722229,
-      "learning_rate": 1.848976890970172e-06,
-      "loss": 0.5007,
+      "epoch": 9.417891287359266,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.838287692151074e-06,
+      "loss": 0.9878707885742187,
       "step": 818000
     },
     {
-      "epoch": 18.840379478677352,
-      "grad_norm": 3.213732957839966,
-      "learning_rate": 1.8417053271476563e-06,
-      "loss": 0.5106,
+      "epoch": 9.420193972433982,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8310161283285583e-06,
+      "loss": 0.9839839935302734,
       "step": 818200
     },
     {
-      "epoch": 18.84498480243161,
-      "grad_norm": 3.5286900997161865,
-      "learning_rate": 1.8344337633251407e-06,
-      "loss": 0.5138,
+      "epoch": 9.422496657508697,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.8237445645060426e-06,
+      "loss": 0.9818657684326172,
       "step": 818400
     },
     {
-      "epoch": 18.84959012618587,
-      "grad_norm": 3.5109329223632812,
-      "learning_rate": 1.827162199502625e-06,
-      "loss": 0.5178,
+      "epoch": 9.42479934258341,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.816473000683527e-06,
+      "loss": 0.982414779663086,
       "step": 818600
     },
     {
-      "epoch": 18.854195449940132,
-      "grad_norm": 2.604468822479248,
-      "learning_rate": 1.8198906356801094e-06,
-      "loss": 0.5086,
+      "epoch": 9.427102027658126,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8092014368610112e-06,
+      "loss": 0.9848974609375,
       "step": 818800
     },
     {
-      "epoch": 18.85880077369439,
-      "grad_norm": 3.2741763591766357,
-      "learning_rate": 1.8126190718575936e-06,
-      "loss": 0.5152,
+      "epoch": 9.429404712732842,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8019298730384956e-06,
+      "loss": 0.9848900604248046,
       "step": 819000
     },
     {
-      "epoch": 18.86340609744865,
-      "grad_norm": 3.35343337059021,
-      "learning_rate": 1.8053475080350782e-06,
-      "loss": 0.5163,
+      "epoch": 9.431707397807555,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.79465830921598e-06,
+      "loss": 0.9777683258056641,
       "step": 819200
     },
     {
-      "epoch": 18.86801142120291,
-      "grad_norm": 3.6704483032226562,
-      "learning_rate": 1.7980759442125625e-06,
-      "loss": 0.5255,
+      "epoch": 9.43401008288227,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7873867453934645e-06,
+      "loss": 0.9755431365966797,
       "step": 819400
     },
     {
-      "epoch": 18.87261674495717,
-      "grad_norm": 3.315521240234375,
-      "learning_rate": 1.790804380390047e-06,
-      "loss": 0.5222,
+      "epoch": 9.436312767956986,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7801151815709487e-06,
+      "loss": 1.0023635864257812,
       "step": 819600
     },
     {
-      "epoch": 18.87722206871143,
-      "grad_norm": 2.896103858947754,
-      "learning_rate": 1.7835328165675311e-06,
-      "loss": 0.5166,
+      "epoch": 9.438615453031701,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7728436177484332e-06,
+      "loss": 0.9739385223388672,
       "step": 819800
     },
     {
-      "epoch": 18.88182739246569,
-      "grad_norm": 2.6083381175994873,
-      "learning_rate": 1.7762612527450155e-06,
-      "loss": 0.5098,
+      "epoch": 9.440918138106415,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7655720539259174e-06,
+      "loss": 0.9773667907714844,
       "step": 820000
     },
     {
-      "epoch": 18.88643271621995,
-      "grad_norm": 3.2741894721984863,
-      "learning_rate": 1.7689896889224998e-06,
-      "loss": 0.5153,
+      "epoch": 9.44322082318113,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.7583004901034018e-06,
+      "loss": 0.9598074340820313,
       "step": 820200
     },
     {
-      "epoch": 18.89103803997421,
-      "grad_norm": 2.9880030155181885,
-      "learning_rate": 1.7617181250999842e-06,
-      "loss": 0.5124,
+      "epoch": 9.445523508255846,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.751028926280886e-06,
+      "loss": 0.9942584228515625,
       "step": 820400
     },
     {
-      "epoch": 18.89564336372847,
-      "grad_norm": 3.405291795730591,
-      "learning_rate": 1.7544465612774684e-06,
-      "loss": 0.518,
+      "epoch": 9.44782619333056,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7437573624583705e-06,
+      "loss": 0.9610377502441406,
       "step": 820600
     },
     {
-      "epoch": 18.90024868748273,
-      "grad_norm": 2.607799768447876,
-      "learning_rate": 1.7472113552740653e-06,
-      "loss": 0.5123,
+      "epoch": 9.450128878405275,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7364857986358547e-06,
+      "loss": 0.984643325805664,
       "step": 820800
     },
     {
-      "epoch": 18.90485401123699,
-      "grad_norm": 3.2214035987854004,
-      "learning_rate": 1.7399397914515498e-06,
-      "loss": 0.5064,
+      "epoch": 9.45243156347999,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.729214234813339e-06,
+      "loss": 0.9934852600097657,
       "step": 821000
     },
     {
-      "epoch": 18.90945933499125,
-      "grad_norm": 3.4306139945983887,
-      "learning_rate": 1.7327045854481467e-06,
-      "loss": 0.5062,
+      "epoch": 9.454734248554706,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7219426709908233e-06,
+      "loss": 0.9818549346923828,
       "step": 821200
     },
     {
-      "epoch": 18.91406465874551,
-      "grad_norm": 2.69059681892395,
-      "learning_rate": 1.7254330216256309e-06,
-      "loss": 0.5194,
+      "epoch": 9.45703693362942,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7146711071683078e-06,
+      "loss": 0.9730079650878907,
       "step": 821400
     },
     {
-      "epoch": 18.918669982499768,
-      "grad_norm": 3.644313097000122,
-      "learning_rate": 1.7181614578031153e-06,
-      "loss": 0.5043,
+      "epoch": 9.459339618704135,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.707399543345792e-06,
+      "loss": 0.9771785736083984,
       "step": 821600
     },
     {
-      "epoch": 18.92327530625403,
-      "grad_norm": 3.1063473224639893,
-      "learning_rate": 1.7108898939805995e-06,
-      "loss": 0.5124,
+      "epoch": 9.46164230377885,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7001279795232764e-06,
+      "loss": 0.9868184661865235,
       "step": 821800
     },
     {
-      "epoch": 18.92788063000829,
-      "grad_norm": 3.852067708969116,
-      "learning_rate": 1.703618330158084e-06,
-      "loss": 0.5106,
+      "epoch": 9.463944988853566,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.6928564157007606e-06,
+      "loss": 0.9857994079589844,
       "step": 822000
     },
     {
-      "epoch": 18.932485953762548,
-      "grad_norm": 2.8518900871276855,
-      "learning_rate": 1.6963467663355682e-06,
-      "loss": 0.5175,
+      "epoch": 9.46624767392828,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.685584851878245e-06,
+      "loss": 0.9687263488769531,
       "step": 822200
     },
     {
-      "epoch": 18.93709127751681,
-      "grad_norm": 2.894487142562866,
-      "learning_rate": 1.6890752025130526e-06,
-      "loss": 0.5137,
+      "epoch": 9.468550359002995,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6783132880557293e-06,
+      "loss": 0.9900286102294922,
       "step": 822400
     },
     {
-      "epoch": 18.94169660127107,
-      "grad_norm": 3.1942451000213623,
-      "learning_rate": 1.6818036386905368e-06,
-      "loss": 0.5127,
+      "epoch": 9.47085304407771,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6710417242332137e-06,
+      "loss": 0.9791419219970703,
       "step": 822600
     },
     {
-      "epoch": 18.94630192502533,
-      "grad_norm": 3.005305051803589,
-      "learning_rate": 1.6745320748680213e-06,
-      "loss": 0.5046,
+      "epoch": 9.473155729152424,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.663770160410698e-06,
+      "loss": 0.9795015716552734,
       "step": 822800
     },
     {
-      "epoch": 18.95090724877959,
-      "grad_norm": 4.5873284339904785,
-      "learning_rate": 1.6672605110455055e-06,
-      "loss": 0.5146,
+      "epoch": 9.47545841422714,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6564985965881824e-06,
+      "loss": 0.9941027069091797,
       "step": 823000
     },
     {
-      "epoch": 18.95551257253385,
-      "grad_norm": 2.7985775470733643,
-      "learning_rate": 1.65998894722299e-06,
-      "loss": 0.5285,
+      "epoch": 9.477761099301855,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6492270327656666e-06,
+      "loss": 0.9751151275634765,
       "step": 823200
     },
     {
-      "epoch": 18.96011789628811,
-      "grad_norm": 3.6909046173095703,
-      "learning_rate": 1.6527173834004741e-06,
-      "loss": 0.522,
+      "epoch": 9.48006378437657,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.641955468943151e-06,
+      "loss": 0.974825668334961,
       "step": 823400
     },
     {
-      "epoch": 18.96472322004237,
-      "grad_norm": 2.9598708152770996,
-      "learning_rate": 1.6454458195779586e-06,
-      "loss": 0.5164,
+      "epoch": 9.482366469451284,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6346839051206352e-06,
+      "loss": 0.9874494934082031,
       "step": 823600
     },
     {
-      "epoch": 18.969328543796628,
-      "grad_norm": 3.4591517448425293,
-      "learning_rate": 1.6381742557554428e-06,
-      "loss": 0.5118,
+      "epoch": 9.484669154526,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.6274123412981197e-06,
+      "loss": 0.9932949066162109,
       "step": 823800
     },
     {
-      "epoch": 18.97393386755089,
-      "grad_norm": 3.0720736980438232,
-      "learning_rate": 1.6309026919329272e-06,
-      "loss": 0.527,
+      "epoch": 9.486971839600715,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6201407774756039e-06,
+      "loss": 0.9929389953613281,
       "step": 824000
     },
     {
-      "epoch": 18.97853919130515,
-      "grad_norm": 4.1211018562316895,
-      "learning_rate": 1.6236311281104114e-06,
-      "loss": 0.5107,
+      "epoch": 9.48927452467543,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6128692136530883e-06,
+      "loss": 0.9867328643798828,
       "step": 824200
     },
     {
-      "epoch": 18.983144515059408,
-      "grad_norm": 2.694967746734619,
-      "learning_rate": 1.6163595642878959e-06,
-      "loss": 0.5063,
+      "epoch": 9.491577209750144,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6055976498305725e-06,
+      "loss": 0.9870306396484375,
       "step": 824400
     },
     {
-      "epoch": 18.98774983881367,
-      "grad_norm": 2.379441022872925,
-      "learning_rate": 1.60908800046538e-06,
-      "loss": 0.5177,
+      "epoch": 9.49387989482486,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.598326086008057e-06,
+      "loss": 0.9809159088134766,
       "step": 824600
     },
     {
-      "epoch": 18.99235516256793,
-      "grad_norm": 2.991995096206665,
-      "learning_rate": 1.6018164366428645e-06,
-      "loss": 0.5018,
+      "epoch": 9.496182579899575,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5910545221855412e-06,
+      "loss": 0.9663616943359375,
       "step": 824800
     },
     {
-      "epoch": 18.996960486322187,
-      "grad_norm": 3.4043898582458496,
-      "learning_rate": 1.5945448728203487e-06,
-      "loss": 0.5125,
+      "epoch": 9.498485264974288,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.5837829583630256e-06,
+      "loss": 0.976966323852539,
       "step": 825000
     },
     {
-      "epoch": 19.0,
-      "eval_loss": 0.5035088062286377,
-      "eval_runtime": 162.1896,
-      "eval_samples_per_second": 174.863,
-      "eval_steps_per_second": 10.932,
-      "step": 825132
-    },
-    {
-      "epoch": 19.00156581007645,
-      "grad_norm": 2.809699296951294,
-      "learning_rate": 1.5873096668169456e-06,
-      "loss": 0.5089,
+      "epoch": 9.500787950049004,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5765113945405098e-06,
+      "loss": 0.9969239807128907,
       "step": 825200
     },
     {
-      "epoch": 19.00617113383071,
-      "grad_norm": 3.992568254470825,
-      "learning_rate": 1.58003810299443e-06,
-      "loss": 0.5059,
+      "epoch": 9.50309063512372,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5692398307179943e-06,
+      "loss": 0.9940318298339844,
       "step": 825400
     },
     {
-      "epoch": 19.010776457584967,
-      "grad_norm": 2.535313129425049,
-      "learning_rate": 1.5727665391719143e-06,
-      "loss": 0.5081,
+      "epoch": 9.505393320198435,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5619682668954785e-06,
+      "loss": 0.9804867553710938,
       "step": 825600
     },
     {
-      "epoch": 19.01538178133923,
-      "grad_norm": 2.8412795066833496,
-      "learning_rate": 1.5654949753493987e-06,
-      "loss": 0.521,
+      "epoch": 9.507696005273148,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.554696703072963e-06,
+      "loss": 0.979859848022461,
       "step": 825800
     },
     {
-      "epoch": 19.019987105093488,
-      "grad_norm": 3.4555296897888184,
-      "learning_rate": 1.558223411526883e-06,
-      "loss": 0.5154,
+      "epoch": 9.509998690347864,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5474251392504471e-06,
+      "loss": 0.9958170318603515,
       "step": 826000
     },
     {
-      "epoch": 19.024592428847747,
-      "grad_norm": 3.3591201305389404,
-      "learning_rate": 1.5509518477043674e-06,
-      "loss": 0.5027,
+      "epoch": 9.51230137542258,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5401535754279318e-06,
+      "loss": 0.9820123291015626,
       "step": 826200
     },
     {
-      "epoch": 19.02919775260201,
-      "grad_norm": 2.9588282108306885,
-      "learning_rate": 1.5436802838818516e-06,
-      "loss": 0.516,
+      "epoch": 9.514604060497295,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.532882011605416e-06,
+      "loss": 0.9946317291259765,
       "step": 826400
     },
     {
-      "epoch": 19.033803076356268,
-      "grad_norm": 3.250680446624756,
-      "learning_rate": 1.536408720059336e-06,
-      "loss": 0.5068,
+      "epoch": 9.516906745572008,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.5256104477829002e-06,
+      "loss": 0.9823373413085937,
       "step": 826600
     },
     {
-      "epoch": 19.038408400110526,
-      "grad_norm": 2.998945713043213,
-      "learning_rate": 1.5291371562368202e-06,
-      "loss": 0.5137,
+      "epoch": 9.519209430646724,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5183388839603846e-06,
+      "loss": 0.97361328125,
       "step": 826800
     },
     {
-      "epoch": 19.04301372386479,
-      "grad_norm": 2.949064254760742,
-      "learning_rate": 1.5218655924143047e-06,
-      "loss": 0.4984,
+      "epoch": 9.521512115721439,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5110673201378689e-06,
+      "loss": 0.9896958160400391,
       "step": 827000
     },
     {
-      "epoch": 19.047619047619047,
-      "grad_norm": 3.1367082595825195,
-      "learning_rate": 1.5146303864109016e-06,
-      "loss": 0.5117,
+      "epoch": 9.523814800796153,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.5037957563153533e-06,
+      "loss": 0.9637728881835937,
       "step": 827200
     },
     {
-      "epoch": 19.052224371373306,
-      "grad_norm": 2.956386089324951,
-      "learning_rate": 1.5073588225883858e-06,
-      "loss": 0.5142,
+      "epoch": 9.526117485870868,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4965241924928375e-06,
+      "loss": 0.9915892028808594,
       "step": 827400
     },
     {
-      "epoch": 19.05682969512757,
-      "grad_norm": 3.038003444671631,
-      "learning_rate": 1.5000872587658702e-06,
-      "loss": 0.4999,
+      "epoch": 9.528420170945584,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.489252628670322e-06,
+      "loss": 0.9733876800537109,
       "step": 827600
     },
     {
-      "epoch": 19.061435018881827,
-      "grad_norm": 3.545823335647583,
-      "learning_rate": 1.4928156949433545e-06,
-      "loss": 0.5041,
+      "epoch": 9.530722856020297,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.4819810648478062e-06,
+      "loss": 0.9952828216552735,
       "step": 827800
     },
     {
-      "epoch": 19.066040342636086,
-      "grad_norm": 2.507877826690674,
-      "learning_rate": 1.4855804889399514e-06,
-      "loss": 0.5173,
+      "epoch": 9.533025541095013,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4747095010252906e-06,
+      "loss": 1.0074019622802735,
       "step": 828000
     },
     {
-      "epoch": 19.070645666390348,
-      "grad_norm": 2.498446464538574,
-      "learning_rate": 1.4783089251174358e-06,
-      "loss": 0.5251,
+      "epoch": 9.535328226169728,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.4674379372027748e-06,
+      "loss": 0.9821954345703126,
       "step": 828200
     },
     {
-      "epoch": 19.075250990144607,
-      "grad_norm": 3.057725429534912,
-      "learning_rate": 1.47103736129492e-06,
-      "loss": 0.511,
+      "epoch": 9.537630911244444,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4601663733802592e-06,
+      "loss": 0.9739497375488281,
       "step": 828400
     },
     {
-      "epoch": 19.079856313898865,
-      "grad_norm": 2.1749329566955566,
-      "learning_rate": 1.4637657974724044e-06,
-      "loss": 0.49,
+      "epoch": 9.539933596319157,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4528948095577435e-06,
+      "loss": 0.9812117004394532,
       "step": 828600
     },
     {
-      "epoch": 19.084461637653128,
-      "grad_norm": 2.94429874420166,
-      "learning_rate": 1.4564942336498887e-06,
-      "loss": 0.5181,
+      "epoch": 9.542236281393873,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.445623245735228e-06,
+      "loss": 1.0017292022705078,
       "step": 828800
     },
     {
-      "epoch": 19.089066961407386,
-      "grad_norm": 2.4639828205108643,
-      "learning_rate": 1.449222669827373e-06,
-      "loss": 0.5149,
+      "epoch": 9.544538966468588,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4383516819127121e-06,
+      "loss": 0.9692825317382813,
       "step": 829000
     },
     {
-      "epoch": 19.093672285161645,
-      "grad_norm": 3.422637462615967,
-      "learning_rate": 1.4419511060048575e-06,
-      "loss": 0.5154,
+      "epoch": 9.546841651543303,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4310801180901965e-06,
+      "loss": 0.9848232269287109,
       "step": 829200
     },
     {
-      "epoch": 19.098277608915907,
-      "grad_norm": 2.8384313583374023,
-      "learning_rate": 1.434679542182342e-06,
-      "loss": 0.505,
+      "epoch": 9.549144336618017,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.4238085542676808e-06,
+      "loss": 0.9496469116210937,
       "step": 829400
     },
     {
-      "epoch": 19.102882932670166,
-      "grad_norm": 2.945739984512329,
-      "learning_rate": 1.4274079783598262e-06,
-      "loss": 0.5124,
+      "epoch": 9.551447021692733,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.4165369904451652e-06,
+      "loss": 0.9722937774658204,
       "step": 829600
     },
     {
-      "epoch": 19.10748825642443,
-      "grad_norm": 2.952714204788208,
-      "learning_rate": 1.4201364145373106e-06,
-      "loss": 0.5124,
+      "epoch": 9.553749706767448,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.4092654266226494e-06,
+      "loss": 0.9781289672851563,
       "step": 829800
     },
     {
-      "epoch": 19.112093580178687,
-      "grad_norm": 4.0392680168151855,
-      "learning_rate": 1.4128648507147948e-06,
-      "loss": 0.5164,
+      "epoch": 9.556052391842162,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4019938628001338e-06,
+      "loss": 1.0032569885253906,
       "step": 830000
     },
     {
-      "epoch": 19.116698903932946,
-      "grad_norm": 3.5149545669555664,
-      "learning_rate": 1.4055932868922793e-06,
-      "loss": 0.511,
+      "epoch": 9.558355076916877,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.394722298977618e-06,
+      "loss": 0.9749464416503906,
       "step": 830200
     },
     {
-      "epoch": 19.121304227687208,
-      "grad_norm": 2.5348634719848633,
-      "learning_rate": 1.3983217230697635e-06,
-      "loss": 0.5078,
+      "epoch": 9.560657761991592,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3874507351551025e-06,
+      "loss": 0.9686476135253906,
       "step": 830400
     },
     {
-      "epoch": 19.125909551441467,
-      "grad_norm": 3.143413782119751,
-      "learning_rate": 1.3910865170663604e-06,
-      "loss": 0.5182,
+      "epoch": 9.562960447066308,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3801791713325867e-06,
+      "loss": 0.9786281585693359,
       "step": 830600
     },
     {
-      "epoch": 19.130514875195725,
-      "grad_norm": 2.6804118156433105,
-      "learning_rate": 1.3838149532438448e-06,
-      "loss": 0.5154,
+      "epoch": 9.565263132141022,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.3729076075100711e-06,
+      "loss": 0.9751205444335938,
       "step": 830800
     },
     {
-      "epoch": 19.135120198949988,
-      "grad_norm": 3.680730104446411,
-      "learning_rate": 1.376543389421329e-06,
-      "loss": 0.5083,
+      "epoch": 9.567565817215737,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3656360436875554e-06,
+      "loss": 0.9922896575927734,
       "step": 831000
     },
     {
-      "epoch": 19.139725522704246,
-      "grad_norm": 3.507612943649292,
-      "learning_rate": 1.369308183417926e-06,
-      "loss": 0.5128,
+      "epoch": 9.569868502290452,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.3583644798650398e-06,
+      "loss": 0.9657626342773438,
       "step": 831200
     },
     {
-      "epoch": 19.144330846458505,
-      "grad_norm": 2.8536086082458496,
-      "learning_rate": 1.3620366195954104e-06,
-      "loss": 0.4996,
+      "epoch": 9.572171187365168,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.351092916042524e-06,
+      "loss": 0.969854736328125,
       "step": 831400
     },
     {
-      "epoch": 19.148936170212767,
-      "grad_norm": 3.133923053741455,
-      "learning_rate": 1.3547650557728946e-06,
-      "loss": 0.5172,
+      "epoch": 9.574473872439881,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3438213522200084e-06,
+      "loss": 0.998516845703125,
       "step": 831600
     },
     {
-      "epoch": 19.153541493967026,
-      "grad_norm": 3.9035439491271973,
-      "learning_rate": 1.347493491950379e-06,
-      "loss": 0.5043,
+      "epoch": 9.576776557514597,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3365497883974927e-06,
+      "loss": 0.99427001953125,
       "step": 831800
     },
     {
-      "epoch": 19.158146817721285,
-      "grad_norm": 2.622859477996826,
-      "learning_rate": 1.3402219281278632e-06,
-      "loss": 0.5147,
+      "epoch": 9.579079242589312,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.329278224574977e-06,
+      "loss": 0.9744255065917968,
       "step": 832000
     },
     {
-      "epoch": 19.162752141475547,
-      "grad_norm": 3.7109525203704834,
-      "learning_rate": 1.3329503643053477e-06,
-      "loss": 0.5115,
+      "epoch": 9.581381927664026,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3220066607524613e-06,
+      "loss": 0.9698799133300782,
       "step": 832200
     },
     {
-      "epoch": 19.167357465229806,
-      "grad_norm": 3.2106027603149414,
-      "learning_rate": 1.325678800482832e-06,
-      "loss": 0.5089,
+      "epoch": 9.583684612738741,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3147350969299458e-06,
+      "loss": 1.000227508544922,
       "step": 832400
     },
     {
-      "epoch": 19.171962788984064,
-      "grad_norm": 2.890758514404297,
-      "learning_rate": 1.3184072366603163e-06,
-      "loss": 0.5142,
+      "epoch": 9.585987297813457,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.30746353310743e-06,
+      "loss": 1.012973861694336,
       "step": 832600
     },
     {
-      "epoch": 19.176568112738327,
-      "grad_norm": 3.477639675140381,
-      "learning_rate": 1.3111356728378005e-06,
-      "loss": 0.5041,
+      "epoch": 9.588289982888172,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3001919692849144e-06,
+      "loss": 0.9542025756835938,
       "step": 832800
     },
     {
-      "epoch": 19.181173436492585,
-      "grad_norm": 3.0593950748443604,
-      "learning_rate": 1.303864109015285e-06,
-      "loss": 0.5304,
+      "epoch": 9.590592667962886,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2929204054623988e-06,
+      "loss": 1.001008071899414,
       "step": 833000
     },
     {
-      "epoch": 19.185778760246844,
-      "grad_norm": 2.967453956604004,
-      "learning_rate": 1.2965925451927692e-06,
-      "loss": 0.5035,
+      "epoch": 9.592895353037601,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2856488416398833e-06,
+      "loss": 0.9920815277099609,
       "step": 833200
     },
     {
-      "epoch": 19.190384084001106,
-      "grad_norm": 3.582881450653076,
-      "learning_rate": 1.2893209813702536e-06,
-      "loss": 0.5136,
+      "epoch": 9.595198038112317,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2783772778173675e-06,
+      "loss": 0.9846733856201172,
       "step": 833400
     },
     {
-      "epoch": 19.194989407755365,
-      "grad_norm": 3.2862138748168945,
-      "learning_rate": 1.2820494175477379e-06,
-      "loss": 0.5052,
+      "epoch": 9.597500723187032,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.271105713994852e-06,
+      "loss": 0.9644556427001953,
       "step": 833600
     },
     {
-      "epoch": 19.199594731509624,
-      "grad_norm": 3.457026720046997,
-      "learning_rate": 1.2747778537252223e-06,
-      "loss": 0.5166,
+      "epoch": 9.599803408261746,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2638341501723361e-06,
+      "loss": 0.9778141021728516,
       "step": 833800
     },
     {
-      "epoch": 19.204200055263886,
-      "grad_norm": 3.373370885848999,
-      "learning_rate": 1.2675062899027065e-06,
-      "loss": 0.5075,
+      "epoch": 9.602106093336461,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2565625863498206e-06,
+      "loss": 0.9689984130859375,
       "step": 834000
     },
     {
-      "epoch": 19.208805379018145,
-      "grad_norm": 3.317833185195923,
-      "learning_rate": 1.260234726080191e-06,
-      "loss": 0.5078,
+      "epoch": 9.604408778411177,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.2492910225273048e-06,
+      "loss": 0.9787627410888672,
       "step": 834200
     },
     {
-      "epoch": 19.213410702772403,
-      "grad_norm": 2.9720406532287598,
-      "learning_rate": 1.2529631622576752e-06,
-      "loss": 0.5074,
+      "epoch": 9.60671146348589,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2420194587047892e-06,
+      "loss": 1.011881103515625,
       "step": 834400
     },
     {
-      "epoch": 19.218016026526666,
-      "grad_norm": 3.7691690921783447,
-      "learning_rate": 1.245727956254272e-06,
-      "loss": 0.5205,
+      "epoch": 9.609014148560606,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2347478948822734e-06,
+      "loss": 0.9822531127929688,
       "step": 834600
     },
     {
-      "epoch": 19.222621350280924,
-      "grad_norm": 3.1545891761779785,
-      "learning_rate": 1.2384563924317565e-06,
-      "loss": 0.5151,
+      "epoch": 9.611316833635321,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2274763310597579e-06,
+      "loss": 0.9833524322509766,
       "step": 834800
     },
     {
-      "epoch": 19.227226674035183,
-      "grad_norm": 2.8010830879211426,
-      "learning_rate": 1.2311848286092407e-06,
-      "loss": 0.515,
+      "epoch": 9.613619518710037,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.220204767237242e-06,
+      "loss": 0.9715459442138672,
       "step": 835000
     },
     {
-      "epoch": 19.231831997789445,
-      "grad_norm": 2.3824315071105957,
-      "learning_rate": 1.2239132647867251e-06,
-      "loss": 0.5145,
+      "epoch": 9.61592220378475,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2129332034147265e-06,
+      "loss": 0.9895572662353516,
       "step": 835200
     },
     {
-      "epoch": 19.236437321543704,
-      "grad_norm": 3.0972163677215576,
-      "learning_rate": 1.2166417009642094e-06,
-      "loss": 0.5138,
+      "epoch": 9.618224888859466,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2056616395922107e-06,
+      "loss": 0.9693475341796876,
       "step": 835400
     },
     {
-      "epoch": 19.241042645297963,
-      "grad_norm": 2.7618439197540283,
-      "learning_rate": 1.2093701371416938e-06,
-      "loss": 0.5125,
+      "epoch": 9.620527573934181,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.1983900757696952e-06,
+      "loss": 0.9663648223876953,
       "step": 835600
     },
     {
-      "epoch": 19.245647969052225,
-      "grad_norm": 3.029303789138794,
-      "learning_rate": 1.202098573319178e-06,
-      "loss": 0.5098,
+      "epoch": 9.622830259008895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.1911185119471794e-06,
+      "loss": 0.999940414428711,
       "step": 835800
     },
     {
-      "epoch": 19.250253292806484,
-      "grad_norm": 3.3938510417938232,
-      "learning_rate": 1.1948270094966622e-06,
-      "loss": 0.5101,
+      "epoch": 9.62513294408361,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1838469481246638e-06,
+      "loss": 0.9695598602294921,
       "step": 836000
     },
     {
-      "epoch": 19.254858616560742,
-      "grad_norm": 3.0954368114471436,
-      "learning_rate": 1.1875554456741467e-06,
-      "loss": 0.5129,
+      "epoch": 9.627435629158326,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.176575384302148e-06,
+      "loss": 0.9636432647705078,
       "step": 836200
     },
     {
-      "epoch": 19.259463940315005,
-      "grad_norm": 2.748145818710327,
-      "learning_rate": 1.1802838818516309e-06,
-      "loss": 0.517,
+      "epoch": 9.629738314233041,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1693038204796325e-06,
+      "loss": 0.963818588256836,
       "step": 836400
     },
     {
-      "epoch": 19.264069264069263,
-      "grad_norm": 3.037444829940796,
-      "learning_rate": 1.1730123180291153e-06,
-      "loss": 0.5031,
+      "epoch": 9.632040999307755,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1620322566571167e-06,
+      "loss": 0.9796644592285156,
       "step": 836600
     },
     {
-      "epoch": 19.268674587823526,
-      "grad_norm": 2.792266845703125,
-      "learning_rate": 1.1657407542065995e-06,
-      "loss": 0.5045,
+      "epoch": 9.63434368438247,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1547606928346011e-06,
+      "loss": 0.9926264190673828,
       "step": 836800
     },
     {
-      "epoch": 19.273279911577784,
-      "grad_norm": 2.9860918521881104,
-      "learning_rate": 1.158469190384084e-06,
-      "loss": 0.5285,
+      "epoch": 9.636646369457186,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1474891290120853e-06,
+      "loss": 0.982134017944336,
       "step": 837000
     },
     {
-      "epoch": 19.277885235332043,
-      "grad_norm": 3.5693347454071045,
-      "learning_rate": 1.1511976265615682e-06,
-      "loss": 0.5165,
+      "epoch": 9.6389490545319,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1402175651895698e-06,
+      "loss": 0.9629191589355469,
       "step": 837200
     },
     {
-      "epoch": 19.282490559086305,
-      "grad_norm": 2.5332469940185547,
-      "learning_rate": 1.1439260627390526e-06,
-      "loss": 0.5243,
+      "epoch": 9.641251739606615,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.132946001367054e-06,
+      "loss": 0.9919446563720703,
       "step": 837400
     },
     {
-      "epoch": 19.287095882840564,
-      "grad_norm": 3.3224828243255615,
-      "learning_rate": 1.1366544989165368e-06,
-      "loss": 0.5106,
+      "epoch": 9.64355442468133,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1256744375445384e-06,
+      "loss": 0.9850608825683593,
       "step": 837600
     },
     {
-      "epoch": 19.291701206594823,
-      "grad_norm": 3.539848804473877,
-      "learning_rate": 1.1293829350940213e-06,
-      "loss": 0.516,
+      "epoch": 9.645857109756045,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1184028737220226e-06,
+      "loss": 0.9781081390380859,
       "step": 837800
     },
     {
-      "epoch": 19.296306530349085,
-      "grad_norm": 4.043239116668701,
-      "learning_rate": 1.1221113712715057e-06,
-      "loss": 0.5165,
+      "epoch": 9.64815979483076,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.111131309899507e-06,
+      "loss": 0.993724365234375,
       "step": 838000
     },
     {
-      "epoch": 19.300911854103344,
-      "grad_norm": 3.046630620956421,
-      "learning_rate": 1.1148398074489901e-06,
-      "loss": 0.5079,
+      "epoch": 9.650462479905475,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1038597460769913e-06,
+      "loss": 0.9846406555175782,
       "step": 838200
     },
     {
-      "epoch": 19.305517177857602,
-      "grad_norm": 2.7998993396759033,
-      "learning_rate": 1.1075682436264743e-06,
-      "loss": 0.5248,
+      "epoch": 9.65276516498019,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0965881822544755e-06,
+      "loss": 0.9925849151611328,
       "step": 838400
     },
     {
-      "epoch": 19.310122501611865,
-      "grad_norm": 3.3862979412078857,
-      "learning_rate": 1.1002966798039588e-06,
-      "loss": 0.5173,
+      "epoch": 9.655067850054905,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.08931661843196e-06,
+      "loss": 0.9630934143066406,
       "step": 838600
     },
     {
-      "epoch": 19.314727825366123,
-      "grad_norm": 3.7672178745269775,
-      "learning_rate": 1.093025115981443e-06,
-      "loss": 0.5157,
+      "epoch": 9.657370535129619,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0820450546094442e-06,
+      "loss": 0.9670571899414062,
       "step": 838800
     },
     {
-      "epoch": 19.319333149120382,
-      "grad_norm": 3.1289238929748535,
-      "learning_rate": 1.0857535521589274e-06,
-      "loss": 0.5193,
+      "epoch": 9.659673220204334,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0747734907869286e-06,
+      "loss": 0.9866361236572265,
       "step": 839000
     },
     {
-      "epoch": 19.323938472874644,
-      "grad_norm": 3.5175318717956543,
-      "learning_rate": 1.0784819883364116e-06,
-      "loss": 0.5246,
+      "epoch": 9.66197590527905,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.0675019269644128e-06,
+      "loss": 0.9836607360839844,
       "step": 839200
     },
     {
-      "epoch": 19.328543796628903,
-      "grad_norm": 2.3767151832580566,
-      "learning_rate": 1.071210424513896e-06,
-      "loss": 0.5068,
+      "epoch": 9.664278590353764,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.0602303631418972e-06,
+      "loss": 0.9988703155517578,
       "step": 839400
     },
     {
-      "epoch": 19.333149120383162,
-      "grad_norm": 2.1531851291656494,
-      "learning_rate": 1.0639388606913803e-06,
-      "loss": 0.5172,
+      "epoch": 9.666581275428479,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0529587993193815e-06,
+      "loss": 0.9767841339111328,
       "step": 839600
     },
     {
-      "epoch": 19.337754444137424,
-      "grad_norm": 3.3230369091033936,
-      "learning_rate": 1.0566672968688647e-06,
-      "loss": 0.5105,
+      "epoch": 9.668883960503194,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.045687235496866e-06,
+      "loss": 0.9695176696777343,
       "step": 839800
     },
     {
-      "epoch": 19.342359767891683,
-      "grad_norm": 3.470402956008911,
-      "learning_rate": 1.049395733046349e-06,
-      "loss": 0.5121,
+      "epoch": 9.67118664557791,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.0384156716743503e-06,
+      "loss": 0.9947758483886718,
       "step": 840000
     },
     {
-      "epoch": 19.34696509164594,
-      "grad_norm": 3.0232300758361816,
-      "learning_rate": 1.0421241692238334e-06,
-      "loss": 0.5038,
+      "epoch": 9.673489330652624,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.0311441078518348e-06,
+      "loss": 0.9972378540039063,
       "step": 840200
     },
     {
-      "epoch": 19.351570415400204,
-      "grad_norm": 2.391409397125244,
-      "learning_rate": 1.0348526054013176e-06,
-      "loss": 0.5104,
+      "epoch": 9.675792015727339,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.023872544029319e-06,
+      "loss": 1.0000531768798828,
       "step": 840400
     },
     {
-      "epoch": 19.356175739154462,
-      "grad_norm": 2.891075849533081,
-      "learning_rate": 1.027581041578802e-06,
-      "loss": 0.5088,
+      "epoch": 9.678094700802054,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0166009802068034e-06,
+      "loss": 0.9775434875488281,
       "step": 840600
     },
     {
-      "epoch": 19.36078106290872,
-      "grad_norm": 3.1132750511169434,
-      "learning_rate": 1.0203094777562862e-06,
-      "loss": 0.5115,
+      "epoch": 9.68039738587677,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0093294163842876e-06,
+      "loss": 0.9801913452148437,
       "step": 840800
     },
     {
-      "epoch": 19.365386386662983,
-      "grad_norm": 2.8549110889434814,
-      "learning_rate": 1.0130742717528832e-06,
-      "loss": 0.5142,
+      "epoch": 9.682700070951483,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.002057852561772e-06,
+      "loss": 1.0049404144287108,
       "step": 841000
     },
     {
-      "epoch": 19.369991710417242,
-      "grad_norm": 3.4312000274658203,
-      "learning_rate": 1.0058027079303676e-06,
-      "loss": 0.5001,
+      "epoch": 9.685002756026199,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.947862887392563e-07,
+      "loss": 0.9907779693603516,
       "step": 841200
     },
     {
-      "epoch": 19.3745970341715,
-      "grad_norm": 3.5737624168395996,
-      "learning_rate": 9.985675019269645e-07,
-      "loss": 0.5036,
+      "epoch": 9.687305441100914,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.875147249167407e-07,
+      "loss": 0.9752469635009766,
       "step": 841400
     },
     {
-      "epoch": 19.379202357925763,
-      "grad_norm": 2.611541748046875,
-      "learning_rate": 9.912959381044487e-07,
-      "loss": 0.5061,
+      "epoch": 9.689608126175628,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.80243161094225e-07,
+      "loss": 0.9838815307617188,
       "step": 841600
     },
     {
-      "epoch": 19.38380768168002,
-      "grad_norm": 3.1176815032958984,
-      "learning_rate": 9.840243742819331e-07,
-      "loss": 0.5011,
+      "epoch": 9.691910811250343,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.729715972717094e-07,
+      "loss": 0.9889019012451172,
       "step": 841800
     },
     {
-      "epoch": 19.38841300543428,
-      "grad_norm": 3.4241085052490234,
-      "learning_rate": 9.767528104594174e-07,
-      "loss": 0.5101,
+      "epoch": 9.694213496325059,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.657000334491936e-07,
+      "loss": 0.9789157104492188,
       "step": 842000
     },
     {
-      "epoch": 19.393018329188543,
-      "grad_norm": 3.769061803817749,
-      "learning_rate": 9.694812466369018e-07,
-      "loss": 0.5212,
+      "epoch": 9.696516181399774,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.58428469626678e-07,
+      "loss": 0.9792619323730469,
       "step": 842200
     },
     {
-      "epoch": 19.3976236529428,
-      "grad_norm": 3.25093674659729,
-      "learning_rate": 9.62209682814386e-07,
-      "loss": 0.5043,
+      "epoch": 9.698818866474488,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.511569058041622e-07,
+      "loss": 0.9853229522705078,
       "step": 842400
     },
     {
-      "epoch": 19.40222897669706,
-      "grad_norm": 3.0781145095825195,
-      "learning_rate": 9.549381189918704e-07,
-      "loss": 0.5031,
+      "epoch": 9.701121551549203,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.438853419816467e-07,
+      "loss": 0.9715174102783203,
       "step": 842600
     },
     {
-      "epoch": 19.406834300451322,
-      "grad_norm": 2.676129102706909,
-      "learning_rate": 9.476665551693547e-07,
-      "loss": 0.5187,
+      "epoch": 9.703424236623919,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.36613778159131e-07,
+      "loss": 0.9615935516357422,
       "step": 842800
     },
     {
-      "epoch": 19.41143962420558,
-      "grad_norm": 3.2870798110961914,
-      "learning_rate": 9.403949913468391e-07,
-      "loss": 0.5144,
+      "epoch": 9.705726921698632,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.293422143366153e-07,
+      "loss": 0.9686674499511718,
       "step": 843000
     },
     {
-      "epoch": 19.41604494795984,
-      "grad_norm": 2.994854211807251,
-      "learning_rate": 9.331234275243234e-07,
-      "loss": 0.5111,
+      "epoch": 9.708029606773348,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.220706505140996e-07,
+      "loss": 0.9705300140380859,
       "step": 843200
     },
     {
-      "epoch": 19.420650271714102,
-      "grad_norm": 4.02811336517334,
-      "learning_rate": 9.258882215209203e-07,
-      "loss": 0.5123,
+      "epoch": 9.710332291848063,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.14799086691584e-07,
+      "loss": 0.9872032165527344,
       "step": 843400
     },
     {
-      "epoch": 19.42525559546836,
-      "grad_norm": 2.6425375938415527,
-      "learning_rate": 9.186166576984047e-07,
-      "loss": 0.5134,
+      "epoch": 9.712634976922779,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.075275228690683e-07,
+      "loss": 0.9720492553710938,
       "step": 843600
     },
     {
-      "epoch": 19.429860919222623,
-      "grad_norm": 3.219266891479492,
-      "learning_rate": 9.11345093875889e-07,
-      "loss": 0.5197,
+      "epoch": 9.714937661997492,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.002559590465526e-07,
+      "loss": 0.9982711791992187,
       "step": 843800
     },
     {
-      "epoch": 19.43446624297688,
-      "grad_norm": 3.178544282913208,
-      "learning_rate": 9.040735300533733e-07,
-      "loss": 0.5238,
+      "epoch": 9.717240347072208,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.929843952240368e-07,
+      "loss": 0.9764456939697266,
       "step": 844000
     },
     {
-      "epoch": 19.43907156673114,
-      "grad_norm": 2.9460504055023193,
-      "learning_rate": 8.968019662308576e-07,
-      "loss": 0.5171,
+      "epoch": 9.719543032146923,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.857128314015212e-07,
+      "loss": 0.9976146697998047,
       "step": 844200
     },
     {
-      "epoch": 19.443676890485403,
-      "grad_norm": 2.947618246078491,
-      "learning_rate": 8.89530402408342e-07,
-      "loss": 0.5113,
+      "epoch": 9.721845717221637,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.784412675790055e-07,
+      "loss": 0.9691112518310547,
       "step": 844400
     },
     {
-      "epoch": 19.44828221423966,
-      "grad_norm": 3.1434247493743896,
-      "learning_rate": 8.822588385858263e-07,
-      "loss": 0.5121,
+      "epoch": 9.724148402296352,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.711697037564898e-07,
+      "loss": 0.9843009948730469,
       "step": 844600
     },
     {
-      "epoch": 19.45288753799392,
-      "grad_norm": 4.052412986755371,
-      "learning_rate": 8.749872747633106e-07,
-      "loss": 0.5132,
+      "epoch": 9.726451087371068,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.638981399339741e-07,
+      "loss": 0.9929389190673829,
       "step": 844800
     },
     {
-      "epoch": 19.457492861748182,
-      "grad_norm": 2.8195929527282715,
-      "learning_rate": 8.677157109407949e-07,
-      "loss": 0.5209,
+      "epoch": 9.728753772445783,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.566265761114586e-07,
+      "loss": 0.95454345703125,
       "step": 845000
     },
     {
-      "epoch": 19.46209818550244,
-      "grad_norm": 2.5590524673461914,
-      "learning_rate": 8.604805049373918e-07,
-      "loss": 0.5102,
+      "epoch": 9.731056457520497,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.493550122889429e-07,
+      "loss": 0.9873353576660157,
       "step": 845200
     },
     {
-      "epoch": 19.4667035092567,
-      "grad_norm": 2.98577618598938,
-      "learning_rate": 8.532089411148762e-07,
-      "loss": 0.5053,
+      "epoch": 9.733359142595212,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.420834484664272e-07,
+      "loss": 0.985777587890625,
       "step": 845400
     },
     {
-      "epoch": 19.471308833010962,
-      "grad_norm": 3.019172430038452,
-      "learning_rate": 8.459373772923605e-07,
-      "loss": 0.5076,
+      "epoch": 9.735661827669928,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.348118846439115e-07,
+      "loss": 1.007764129638672,
       "step": 845600
     },
     {
-      "epoch": 19.47591415676522,
-      "grad_norm": 3.0066978931427,
-      "learning_rate": 8.386658134698449e-07,
-      "loss": 0.5134,
+      "epoch": 9.737964512744643,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.275403208213959e-07,
+      "loss": 0.9851364135742188,
       "step": 845800
     },
     {
-      "epoch": 19.48051948051948,
-      "grad_norm": 2.8860740661621094,
-      "learning_rate": 8.313942496473292e-07,
-      "loss": 0.5171,
+      "epoch": 9.740267197819357,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.202687569988802e-07,
+      "loss": 0.9813011169433594,
       "step": 846000
     },
     {
-      "epoch": 19.48512480427374,
-      "grad_norm": 3.4044198989868164,
-      "learning_rate": 8.241226858248136e-07,
-      "loss": 0.5019,
+      "epoch": 9.742569882894072,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.129971931763645e-07,
+      "loss": 0.9733602905273437,
       "step": 846200
     },
     {
-      "epoch": 19.489730128028,
-      "grad_norm": 2.7292261123657227,
-      "learning_rate": 8.168511220022979e-07,
-      "loss": 0.5093,
+      "epoch": 9.744872567968788,
+      "grad_norm": 0.75,
+      "learning_rate": 8.057256293538488e-07,
+      "loss": 0.9741603851318359,
       "step": 846400
     },
     {
-      "epoch": 19.49433545178226,
-      "grad_norm": 3.8156752586364746,
-      "learning_rate": 8.095795581797822e-07,
-      "loss": 0.5105,
+      "epoch": 9.747175253043501,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.984540655313332e-07,
+      "loss": 0.9832783508300781,
       "step": 846600
     },
     {
-      "epoch": 19.49894077553652,
-      "grad_norm": 3.2243752479553223,
-      "learning_rate": 8.023079943572666e-07,
-      "loss": 0.5104,
+      "epoch": 9.749477938118217,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.911825017088175e-07,
+      "loss": 1.0015650177001953,
       "step": 846800
     },
     {
-      "epoch": 19.50354609929078,
-      "grad_norm": 3.429675340652466,
-      "learning_rate": 7.950364305347509e-07,
-      "loss": 0.5064,
+      "epoch": 9.751780623192932,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.839109378863018e-07,
+      "loss": 0.9919761657714844,
       "step": 847000
     },
     {
-      "epoch": 19.50815142304504,
-      "grad_norm": 2.856170415878296,
-      "learning_rate": 7.877648667122352e-07,
-      "loss": 0.5156,
+      "epoch": 9.754083308267647,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.766393740637861e-07,
+      "loss": 0.9678952789306641,
       "step": 847200
     },
     {
-      "epoch": 19.5127567467993,
-      "grad_norm": 3.27316951751709,
-      "learning_rate": 7.804933028897195e-07,
-      "loss": 0.4983,
+      "epoch": 9.756385993342361,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.693678102412705e-07,
+      "loss": 0.9774272918701172,
       "step": 847400
     },
     {
-      "epoch": 19.51736207055356,
-      "grad_norm": 3.1368207931518555,
-      "learning_rate": 7.732217390672039e-07,
-      "loss": 0.5118,
+      "epoch": 9.758688678417077,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.620962464187548e-07,
+      "loss": 0.9884217071533203,
       "step": 847600
     },
     {
-      "epoch": 19.52196739430782,
-      "grad_norm": 2.677781581878662,
-      "learning_rate": 7.659501752446882e-07,
-      "loss": 0.5106,
+      "epoch": 9.760991363491792,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.548246825962391e-07,
+      "loss": 0.9622210693359375,
       "step": 847800
     },
     {
-      "epoch": 19.52657271806208,
-      "grad_norm": 2.9507906436920166,
-      "learning_rate": 7.586786114221725e-07,
-      "loss": 0.5203,
+      "epoch": 9.763294048566507,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.475531187737234e-07,
+      "loss": 0.9797222900390625,
       "step": 848000
     },
     {
-      "epoch": 19.53117804181634,
-      "grad_norm": 3.28304123878479,
-      "learning_rate": 7.514070475996568e-07,
-      "loss": 0.5062,
+      "epoch": 9.765596733641221,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.402815549512078e-07,
+      "loss": 0.9862091827392578,
       "step": 848200
     },
     {
-      "epoch": 19.535783365570598,
-      "grad_norm": 3.1453018188476562,
-      "learning_rate": 7.44135483777141e-07,
-      "loss": 0.5085,
+      "epoch": 9.767899418715936,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.330099911286922e-07,
+      "loss": 0.9772564697265625,
       "step": 848400
     },
     {
-      "epoch": 19.54038868932486,
-      "grad_norm": 2.7345168590545654,
-      "learning_rate": 7.368639199546254e-07,
-      "loss": 0.5128,
+      "epoch": 9.770202103790652,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.257384273061765e-07,
+      "loss": 0.9882775115966796,
       "step": 848600
     },
     {
-      "epoch": 19.54499401307912,
-      "grad_norm": 3.114023447036743,
-      "learning_rate": 7.295923561321097e-07,
-      "loss": 0.5089,
+      "epoch": 9.772504788865366,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.184668634836608e-07,
+      "loss": 0.9722277069091797,
       "step": 848800
     },
     {
-      "epoch": 19.549599336833378,
-      "grad_norm": 2.9676406383514404,
-      "learning_rate": 7.22320792309594e-07,
-      "loss": 0.5033,
+      "epoch": 9.774807473940081,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.111952996611452e-07,
+      "loss": 1.0243209075927735,
       "step": 849000
     },
     {
-      "epoch": 19.55420466058764,
-      "grad_norm": 3.178053855895996,
-      "learning_rate": 7.150492284870785e-07,
-      "loss": 0.5015,
+      "epoch": 9.777110159014796,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.039237358386295e-07,
+      "loss": 0.9980621337890625,
       "step": 849200
     },
     {
-      "epoch": 19.5588099843419,
-      "grad_norm": 2.661820411682129,
-      "learning_rate": 7.077776646645628e-07,
-      "loss": 0.5122,
+      "epoch": 9.779412844089512,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.966521720161138e-07,
+      "loss": 1.0021202850341797,
       "step": 849400
     },
     {
-      "epoch": 19.563415308096157,
-      "grad_norm": 3.9830830097198486,
-      "learning_rate": 7.005061008420471e-07,
-      "loss": 0.5087,
+      "epoch": 9.781715529164225,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.893806081935981e-07,
+      "loss": 0.9711225128173828,
       "step": 849600
     },
     {
-      "epoch": 19.56802063185042,
-      "grad_norm": 2.1384353637695312,
-      "learning_rate": 6.932345370195314e-07,
-      "loss": 0.5082,
+      "epoch": 9.784018214238941,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.821090443710825e-07,
+      "loss": 0.97218505859375,
       "step": 849800
     },
     {
-      "epoch": 19.57262595560468,
-      "grad_norm": 3.7364590167999268,
-      "learning_rate": 6.859629731970158e-07,
-      "loss": 0.4953,
+      "epoch": 9.786320899313656,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.748374805485668e-07,
+      "loss": 0.9849751281738282,
       "step": 850000
     },
     {
-      "epoch": 19.57723127935894,
-      "grad_norm": 2.6958413124084473,
-      "learning_rate": 6.786914093745001e-07,
-      "loss": 0.5113,
+      "epoch": 9.788623584388372,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.675659167260511e-07,
+      "loss": 0.9899353790283203,
       "step": 850200
     },
     {
-      "epoch": 19.5818366031132,
-      "grad_norm": 3.8758256435394287,
-      "learning_rate": 6.714198455519844e-07,
-      "loss": 0.507,
+      "epoch": 9.790926269463085,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.602943529035354e-07,
+      "loss": 0.9618595886230469,
       "step": 850400
     },
     {
-      "epoch": 19.586441926867458,
-      "grad_norm": 2.9180338382720947,
-      "learning_rate": 6.641482817294687e-07,
-      "loss": 0.519,
+      "epoch": 9.7932289545378,
+      "grad_norm": 0.75,
+      "learning_rate": 6.530227890810198e-07,
+      "loss": 0.9787592315673828,
       "step": 850600
     },
     {
-      "epoch": 19.59104725062172,
-      "grad_norm": 2.6355738639831543,
-      "learning_rate": 6.568767179069531e-07,
-      "loss": 0.4985,
+      "epoch": 9.795531639612516,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.457512252585041e-07,
+      "loss": 0.980878677368164,
       "step": 850800
     },
     {
-      "epoch": 19.59565257437598,
-      "grad_norm": 2.67168927192688,
-      "learning_rate": 6.4964151190355e-07,
-      "loss": 0.5045,
+      "epoch": 9.79783432468723,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.384796614359884e-07,
+      "loss": 0.994683837890625,
       "step": 851000
     },
     {
-      "epoch": 19.600257898130238,
-      "grad_norm": 4.244370460510254,
-      "learning_rate": 6.423699480810343e-07,
-      "loss": 0.5072,
+      "epoch": 9.800137009761945,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.312080976134727e-07,
+      "loss": 0.9969705200195312,
       "step": 851200
     },
     {
-      "epoch": 19.6048632218845,
-      "grad_norm": 3.0310142040252686,
-      "learning_rate": 6.350983842585186e-07,
-      "loss": 0.5097,
+      "epoch": 9.80243969483666,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.239365337909571e-07,
+      "loss": 0.9933558654785156,
       "step": 851400
     },
     {
-      "epoch": 19.60946854563876,
-      "grad_norm": 3.2820241451263428,
-      "learning_rate": 6.278268204360029e-07,
-      "loss": 0.5038,
+      "epoch": 9.804742379911376,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.166649699684414e-07,
+      "loss": 0.9854070281982422,
       "step": 851600
     },
     {
-      "epoch": 19.614073869393017,
-      "grad_norm": 3.4763691425323486,
-      "learning_rate": 6.205552566134873e-07,
-      "loss": 0.5262,
+      "epoch": 9.80704506498609,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.093934061459258e-07,
+      "loss": 0.986405029296875,
       "step": 851800
     },
     {
-      "epoch": 19.61867919314728,
-      "grad_norm": 3.8789632320404053,
-      "learning_rate": 6.132836927909716e-07,
-      "loss": 0.5077,
+      "epoch": 9.809347750060805,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.021218423234102e-07,
+      "loss": 0.9843788909912109,
       "step": 852000
     },
     {
-      "epoch": 19.62328451690154,
-      "grad_norm": 2.8242287635803223,
-      "learning_rate": 6.060121289684559e-07,
-      "loss": 0.5073,
+      "epoch": 9.81165043513552,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.948502785008945e-07,
+      "loss": 0.9770623779296875,
       "step": 852200
     },
     {
-      "epoch": 19.627889840655797,
-      "grad_norm": 3.055107593536377,
-      "learning_rate": 5.987405651459402e-07,
-      "loss": 0.5079,
+      "epoch": 9.813953120210234,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.875787146783788e-07,
+      "loss": 0.9609633636474609,
       "step": 852400
     },
     {
-      "epoch": 19.63249516441006,
-      "grad_norm": 3.6184749603271484,
-      "learning_rate": 5.914690013234246e-07,
-      "loss": 0.5184,
+      "epoch": 9.81625580528495,
+      "grad_norm": 1.90625,
+      "learning_rate": 5.803071508558631e-07,
+      "loss": 1.0110892486572265,
       "step": 852600
     },
     {
-      "epoch": 19.637100488164318,
-      "grad_norm": 2.965026378631592,
-      "learning_rate": 5.841974375009089e-07,
-      "loss": 0.5054,
+      "epoch": 9.818558490359665,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.730355870333475e-07,
+      "loss": 0.9992935180664062,
       "step": 852800
     },
     {
-      "epoch": 19.641705811918577,
-      "grad_norm": 2.952613115310669,
-      "learning_rate": 5.769258736783932e-07,
-      "loss": 0.5095,
+      "epoch": 9.82086117543438,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.657640232108318e-07,
+      "loss": 0.9737750244140625,
       "step": 853000
     },
     {
-      "epoch": 19.64631113567284,
-      "grad_norm": 3.1526806354522705,
-      "learning_rate": 5.696543098558776e-07,
-      "loss": 0.501,
+      "epoch": 9.823163860509094,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.584924593883161e-07,
+      "loss": 0.9821039581298828,
       "step": 853200
     },
     {
-      "epoch": 19.650916459427098,
-      "grad_norm": 2.6503217220306396,
-      "learning_rate": 5.62382746033362e-07,
-      "loss": 0.5078,
+      "epoch": 9.82546654558381,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.512208955658004e-07,
+      "loss": 0.9867521667480469,
       "step": 853400
     },
     {
-      "epoch": 19.655521783181356,
-      "grad_norm": 3.481039524078369,
-      "learning_rate": 5.551111822108463e-07,
-      "loss": 0.5219,
+      "epoch": 9.827769230658525,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.439493317432848e-07,
+      "loss": 0.9823966979980469,
       "step": 853600
     },
     {
-      "epoch": 19.66012710693562,
-      "grad_norm": 2.944347858428955,
-      "learning_rate": 5.478396183883306e-07,
-      "loss": 0.5095,
+      "epoch": 9.830071915733239,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.366777679207691e-07,
+      "loss": 0.9842733764648437,
       "step": 853800
     },
     {
-      "epoch": 19.664732430689877,
-      "grad_norm": 3.2116689682006836,
-      "learning_rate": 5.40568054565815e-07,
-      "loss": 0.5008,
+      "epoch": 9.832374600807954,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.294062040982534e-07,
+      "loss": 0.9998354339599609,
       "step": 854000
     },
     {
-      "epoch": 19.669337754444136,
-      "grad_norm": 2.333085775375366,
-      "learning_rate": 5.332964907432993e-07,
-      "loss": 0.5028,
+      "epoch": 9.83467728588267,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.221346402757377e-07,
+      "loss": 0.998682861328125,
       "step": 854200
     },
     {
-      "epoch": 19.6739430781984,
-      "grad_norm": 3.2551276683807373,
-      "learning_rate": 5.260249269207836e-07,
-      "loss": 0.5081,
+      "epoch": 9.836979970957385,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.148630764532221e-07,
+      "loss": 0.9682182312011719,
       "step": 854400
     },
     {
-      "epoch": 19.678548401952657,
-      "grad_norm": 2.8687589168548584,
-      "learning_rate": 5.187533630982679e-07,
-      "loss": 0.5078,
+      "epoch": 9.839282656032099,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.075915126307064e-07,
+      "loss": 0.9853717041015625,
       "step": 854600
     },
     {
-      "epoch": 19.683153725706916,
-      "grad_norm": 2.7332775592803955,
-      "learning_rate": 5.114817992757523e-07,
-      "loss": 0.5065,
+      "epoch": 9.841585341106814,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.003199488081907e-07,
+      "loss": 0.9746234130859375,
       "step": 854800
     },
     {
-      "epoch": 19.687759049461178,
-      "grad_norm": 3.2149274349212646,
-      "learning_rate": 5.042465932723492e-07,
-      "loss": 0.5121,
+      "epoch": 9.84388802618153,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.93048384985675e-07,
+      "loss": 0.976443099975586,
       "step": 855000
     },
     {
-      "epoch": 19.692364373215437,
-      "grad_norm": 3.2051103115081787,
-      "learning_rate": 4.969750294498335e-07,
-      "loss": 0.5176,
+      "epoch": 9.846190711256245,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.857768211631595e-07,
+      "loss": 0.9782569122314453,
       "step": 855200
     },
     {
-      "epoch": 19.696969696969695,
-      "grad_norm": 4.147640228271484,
-      "learning_rate": 4.897034656273178e-07,
-      "loss": 0.4981,
+      "epoch": 9.848493396330959,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.785052573406437e-07,
+      "loss": 0.9826511383056641,
       "step": 855400
     },
     {
-      "epoch": 19.701575020723958,
-      "grad_norm": 2.6071434020996094,
-      "learning_rate": 4.824319018048021e-07,
-      "loss": 0.519,
+      "epoch": 9.850796081405674,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.7123369351812796e-07,
+      "loss": 0.9919064331054688,
       "step": 855600
     },
     {
-      "epoch": 19.706180344478216,
-      "grad_norm": 2.8884124755859375,
-      "learning_rate": 4.7516033798228646e-07,
-      "loss": 0.5043,
+      "epoch": 9.85309876648039,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.6396212969561234e-07,
+      "loss": 0.979754409790039,
       "step": 855800
     },
     {
-      "epoch": 19.710785668232475,
-      "grad_norm": 3.735039234161377,
-      "learning_rate": 4.678887741597708e-07,
-      "loss": 0.5042,
+      "epoch": 9.855401451555103,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.5669056587309666e-07,
+      "loss": 0.9738344573974609,
       "step": 856000
     },
     {
-      "epoch": 19.715390991986737,
-      "grad_norm": 2.957726001739502,
-      "learning_rate": 4.606172103372551e-07,
-      "loss": 0.5158,
+      "epoch": 9.857704136629819,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.4941900205058104e-07,
+      "loss": 0.9660050964355469,
       "step": 856200
     },
     {
-      "epoch": 19.719996315740996,
-      "grad_norm": 2.8788440227508545,
-      "learning_rate": 4.533456465147395e-07,
-      "loss": 0.5084,
+      "epoch": 9.860006821704534,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.4214743822806536e-07,
+      "loss": 0.9862245941162109,
       "step": 856400
     },
     {
-      "epoch": 19.724601639495255,
-      "grad_norm": 3.8605153560638428,
-      "learning_rate": 4.460740826922238e-07,
-      "loss": 0.5201,
+      "epoch": 9.86230950677925,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.348758744055497e-07,
+      "loss": 0.9857429504394531,
       "step": 856600
     },
     {
-      "epoch": 19.729206963249517,
-      "grad_norm": 3.0567197799682617,
-      "learning_rate": 4.3880251886970814e-07,
-      "loss": 0.4976,
+      "epoch": 9.864612191853963,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.27604310583034e-07,
+      "loss": 0.9800594329833985,
       "step": 856800
     },
     {
-      "epoch": 19.733812287003776,
-      "grad_norm": 3.3743460178375244,
-      "learning_rate": 4.3153095504719246e-07,
-      "loss": 0.509,
+      "epoch": 9.866914876928679,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.2033274676051834e-07,
+      "loss": 0.9967246246337891,
       "step": 857000
     },
     {
-      "epoch": 19.738417610758034,
-      "grad_norm": 2.7658159732818604,
-      "learning_rate": 4.242593912246768e-07,
-      "loss": 0.5259,
+      "epoch": 9.869217562003394,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.1306118293800267e-07,
+      "loss": 0.9913703918457031,
       "step": 857200
     },
     {
-      "epoch": 19.743022934512297,
-      "grad_norm": 3.7195098400115967,
-      "learning_rate": 4.170241852212737e-07,
-      "loss": 0.5088,
+      "epoch": 9.87152024707811,
+      "grad_norm": 1.359375,
+      "learning_rate": 4.05789619115487e-07,
+      "loss": 0.9892777252197266,
       "step": 857400
     },
     {
-      "epoch": 19.747628258266555,
-      "grad_norm": 2.6873762607574463,
-      "learning_rate": 4.09752621398758e-07,
-      "loss": 0.5168,
+      "epoch": 9.873822932152823,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.985180552929713e-07,
+      "loss": 1.000076675415039,
       "step": 857600
     },
     {
-      "epoch": 19.752233582020818,
-      "grad_norm": 2.6738898754119873,
-      "learning_rate": 4.0248105757624235e-07,
-      "loss": 0.5081,
+      "epoch": 9.876125617227538,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.912464914704557e-07,
+      "loss": 1.00410888671875,
       "step": 857800
     },
     {
-      "epoch": 19.756838905775076,
-      "grad_norm": 2.91037917137146,
-      "learning_rate": 3.9520949375372667e-07,
-      "loss": 0.5066,
+      "epoch": 9.878428302302254,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.8397492764794e-07,
+      "loss": 0.971823959350586,
       "step": 858000
     },
     {
-      "epoch": 19.761444229529335,
-      "grad_norm": 2.883091926574707,
-      "learning_rate": 3.87937929931211e-07,
-      "loss": 0.5125,
+      "epoch": 9.880730987376968,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.767033638254243e-07,
+      "loss": 0.9962834930419922,
       "step": 858200
     },
     {
-      "epoch": 19.766049553283597,
-      "grad_norm": 2.677788257598877,
-      "learning_rate": 3.806663661086953e-07,
-      "loss": 0.5051,
+      "epoch": 9.883033672451683,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.694318000029086e-07,
+      "loss": 0.9885250854492188,
       "step": 858400
     },
     {
-      "epoch": 19.770654877037856,
-      "grad_norm": 3.4007511138916016,
-      "learning_rate": 3.7339480228617965e-07,
-      "loss": 0.5061,
+      "epoch": 9.885336357526398,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.6216023618039294e-07,
+      "loss": 0.9724180603027344,
       "step": 858600
     },
     {
-      "epoch": 19.775260200792115,
-      "grad_norm": 3.597402572631836,
-      "learning_rate": 3.6612323846366403e-07,
-      "loss": 0.5014,
+      "epoch": 9.887639042601114,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.5488867235787727e-07,
+      "loss": 0.9764485931396485,
       "step": 858800
     },
     {
-      "epoch": 19.779865524546377,
-      "grad_norm": 3.0766663551330566,
-      "learning_rate": 3.5885167464114835e-07,
-      "loss": 0.5205,
+      "epoch": 9.889941727675827,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.476171085353616e-07,
+      "loss": 0.9844834899902344,
       "step": 859000
     },
     {
-      "epoch": 19.784470848300636,
-      "grad_norm": 3.5148487091064453,
-      "learning_rate": 3.516164686377452e-07,
-      "loss": 0.5238,
+      "epoch": 9.892244412750543,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.403455447128459e-07,
+      "loss": 0.9847768402099609,
       "step": 859200
     },
     {
-      "epoch": 19.789076172054894,
-      "grad_norm": 2.736527681350708,
-      "learning_rate": 3.4434490481522953e-07,
-      "loss": 0.5033,
+      "epoch": 9.894547097825258,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.3307398089033024e-07,
+      "loss": 0.9868364715576172,
       "step": 859400
     },
     {
-      "epoch": 19.793681495809157,
-      "grad_norm": 3.351614236831665,
-      "learning_rate": 3.370733409927139e-07,
-      "loss": 0.5144,
+      "epoch": 9.896849782899972,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.258024170678146e-07,
+      "loss": 0.9970125579833984,
       "step": 859600
     },
     {
-      "epoch": 19.798286819563415,
-      "grad_norm": 2.9245457649230957,
-      "learning_rate": 3.2980177717019824e-07,
-      "loss": 0.5079,
+      "epoch": 9.899152467974687,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.1853085324529895e-07,
+      "loss": 0.9947715759277344,
       "step": 859800
     },
     {
-      "epoch": 19.802892143317674,
-      "grad_norm": 2.770165205001831,
-      "learning_rate": 3.2253021334768256e-07,
-      "loss": 0.5054,
+      "epoch": 9.901455153049403,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.1125928942278327e-07,
+      "loss": 0.995891342163086,
       "step": 860000
     },
     {
-      "epoch": 19.807497467071936,
-      "grad_norm": 3.2325165271759033,
-      "learning_rate": 3.152586495251669e-07,
-      "loss": 0.5162,
+      "epoch": 9.903757838124118,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.039877256002676e-07,
+      "loss": 1.0195358276367188,
       "step": 860200
     },
     {
-      "epoch": 19.812102790826195,
-      "grad_norm": 2.7553768157958984,
-      "learning_rate": 3.079870857026512e-07,
-      "loss": 0.5005,
+      "epoch": 9.906060523198832,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.967161617777519e-07,
+      "loss": 0.9633563232421875,
       "step": 860400
     },
     {
-      "epoch": 19.816708114580454,
-      "grad_norm": 3.5754334926605225,
-      "learning_rate": 3.0071552188013554e-07,
-      "loss": 0.5289,
+      "epoch": 9.908363208273547,
+      "grad_norm": 0.75,
+      "learning_rate": 2.8944459795523625e-07,
+      "loss": 0.9851063537597656,
       "step": 860600
     },
     {
-      "epoch": 19.821313438334716,
-      "grad_norm": 3.362692356109619,
-      "learning_rate": 2.9344395805761986e-07,
-      "loss": 0.5098,
+      "epoch": 9.910665893348263,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8217303413272057e-07,
+      "loss": 0.9972560119628906,
       "step": 860800
     },
     {
-      "epoch": 19.825918762088975,
-      "grad_norm": 2.9253594875335693,
-      "learning_rate": 2.861723942351042e-07,
-      "loss": 0.5011,
+      "epoch": 9.912968578422976,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.749014703102049e-07,
+      "loss": 0.965716552734375,
       "step": 861000
     },
     {
-      "epoch": 19.830524085843233,
-      "grad_norm": 3.8785996437072754,
-      "learning_rate": 2.7890083041258857e-07,
-      "loss": 0.5125,
+      "epoch": 9.915271263497692,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.676299064876893e-07,
+      "loss": 0.9895931243896484,
       "step": 861200
     },
     {
-      "epoch": 19.835129409597496,
-      "grad_norm": 3.061610698699951,
-      "learning_rate": 2.716292665900729e-07,
-      "loss": 0.5266,
+      "epoch": 9.917573948572407,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.603583426651736e-07,
+      "loss": 0.9821286010742187,
       "step": 861400
     },
     {
-      "epoch": 19.839734733351754,
-      "grad_norm": 3.2867627143859863,
-      "learning_rate": 2.643577027675572e-07,
-      "loss": 0.5109,
+      "epoch": 9.919876633647123,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.5308677884265793e-07,
+      "loss": 0.998079833984375,
       "step": 861600
     },
     {
-      "epoch": 19.844340057106013,
-      "grad_norm": 3.554579496383667,
-      "learning_rate": 2.5708613894504154e-07,
-      "loss": 0.5218,
+      "epoch": 9.922179318721836,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.4581521502014225e-07,
+      "loss": 0.9605708312988281,
       "step": 861800
     },
     {
-      "epoch": 19.848945380860275,
-      "grad_norm": 3.178513288497925,
-      "learning_rate": 2.4981457512252587e-07,
-      "loss": 0.5147,
+      "epoch": 9.924482003796552,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.385436511976266e-07,
+      "loss": 0.9791954040527344,
       "step": 862000
     },
     {
-      "epoch": 19.853550704614534,
-      "grad_norm": 3.356754779815674,
-      "learning_rate": 2.425430113000102e-07,
-      "loss": 0.5141,
+      "epoch": 9.926784688871267,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.312720873751109e-07,
+      "loss": 0.9803323364257812,
       "step": 862200
     },
     {
-      "epoch": 19.858156028368793,
-      "grad_norm": 2.666679859161377,
-      "learning_rate": 2.3527144747749452e-07,
-      "loss": 0.5183,
+      "epoch": 9.929087373945983,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.2400052355259523e-07,
+      "loss": 0.977833251953125,
       "step": 862400
     },
     {
-      "epoch": 19.862761352123055,
-      "grad_norm": 3.7374143600463867,
-      "learning_rate": 2.2799988365497884e-07,
-      "loss": 0.5166,
+      "epoch": 9.931390059020696,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.1672895973007955e-07,
+      "loss": 0.9808238220214843,
       "step": 862600
     },
     {
-      "epoch": 19.867366675877314,
-      "grad_norm": 3.064572334289551,
-      "learning_rate": 2.2072831983246317e-07,
-      "loss": 0.5002,
+      "epoch": 9.933692744095412,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.0945739590756388e-07,
+      "loss": 0.9963713073730469,
       "step": 862800
     },
     {
-      "epoch": 19.871971999631572,
-      "grad_norm": 3.1594460010528564,
-      "learning_rate": 2.134567560099475e-07,
-      "loss": 0.5128,
+      "epoch": 9.935995429170127,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.021858320850482e-07,
+      "loss": 0.9976137542724609,
       "step": 863000
     },
     {
-      "epoch": 19.876577323385835,
-      "grad_norm": 3.091594934463501,
-      "learning_rate": 2.0618519218743182e-07,
-      "loss": 0.5092,
+      "epoch": 9.93829811424484,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9491426826253253e-07,
+      "loss": 0.971026611328125,
       "step": 863200
     },
     {
-      "epoch": 19.881182647140093,
-      "grad_norm": 3.6168246269226074,
-      "learning_rate": 1.9891362836491617e-07,
-      "loss": 0.5154,
+      "epoch": 9.940600799319556,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8764270444001688e-07,
+      "loss": 0.9874187469482422,
       "step": 863400
     },
     {
-      "epoch": 19.885787970894356,
-      "grad_norm": 3.3390212059020996,
-      "learning_rate": 1.916420645424005e-07,
-      "loss": 0.5118,
+      "epoch": 9.942903484394272,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.803711406175012e-07,
+      "loss": 0.97000732421875,
       "step": 863600
     },
     {
-      "epoch": 19.890393294648614,
-      "grad_norm": 2.9384098052978516,
-      "learning_rate": 1.8437050071988482e-07,
-      "loss": 0.4973,
+      "epoch": 9.945206169468987,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7309957679498553e-07,
+      "loss": 0.9777810668945313,
       "step": 863800
     },
     {
-      "epoch": 19.894998618402873,
-      "grad_norm": 3.3351097106933594,
-      "learning_rate": 1.7713529471648173e-07,
-      "loss": 0.5152,
+      "epoch": 9.9475088545437,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.6582801297246986e-07,
+      "loss": 0.9669603729248046,
       "step": 864000
     },
     {
-      "epoch": 19.899603942157135,
-      "grad_norm": 3.4556448459625244,
-      "learning_rate": 1.6986373089396606e-07,
-      "loss": 0.5134,
+      "epoch": 9.949811539618416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.585564491499542e-07,
+      "loss": 0.9518377685546875,
       "step": 864200
     },
     {
-      "epoch": 19.904209265911394,
-      "grad_norm": 3.008500337600708,
-      "learning_rate": 1.6262852489056296e-07,
-      "loss": 0.5085,
+      "epoch": 9.952114224693132,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5128488532743853e-07,
+      "loss": 0.994540786743164,
       "step": 864400
     },
     {
-      "epoch": 19.908814589665653,
-      "grad_norm": 2.8039371967315674,
-      "learning_rate": 1.5535696106804732e-07,
-      "loss": 0.513,
+      "epoch": 9.954416909767847,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4401332150492286e-07,
+      "loss": 0.9761511993408203,
       "step": 864600
     },
     {
-      "epoch": 19.913419913419915,
-      "grad_norm": 4.19392204284668,
-      "learning_rate": 1.4808539724553164e-07,
-      "loss": 0.5077,
+      "epoch": 9.95671959484256,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3674175768240718e-07,
+      "loss": 0.9886344909667969,
       "step": 864800
     },
     {
-      "epoch": 19.918025237174174,
-      "grad_norm": 3.3566884994506836,
-      "learning_rate": 1.4081383342301597e-07,
-      "loss": 0.5156,
+      "epoch": 9.959022279917276,
+      "grad_norm": 0.75,
+      "learning_rate": 1.294701938598915e-07,
+      "loss": 0.9843460083007812,
       "step": 865000
     },
     {
-      "epoch": 19.922630560928432,
-      "grad_norm": 3.012193202972412,
-      "learning_rate": 1.335422696005003e-07,
-      "loss": 0.5019,
+      "epoch": 9.961324964991991,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.2219863003737586e-07,
+      "loss": 0.9988361358642578,
       "step": 865200
     },
     {
-      "epoch": 19.927235884682695,
-      "grad_norm": 3.2162373065948486,
-      "learning_rate": 1.2627070577798462e-07,
-      "loss": 0.5135,
+      "epoch": 9.963627650066705,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1492706621486017e-07,
+      "loss": 0.9724899291992187,
       "step": 865400
     },
     {
-      "epoch": 19.931841208436953,
-      "grad_norm": 3.0207014083862305,
-      "learning_rate": 1.1903549977458152e-07,
-      "loss": 0.506,
+      "epoch": 9.96593033514142,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.076555023923445e-07,
+      "loss": 0.9676179504394531,
       "step": 865600
     },
     {
-      "epoch": 19.936446532191212,
-      "grad_norm": 3.044567823410034,
-      "learning_rate": 1.1176393595206585e-07,
-      "loss": 0.5098,
+      "epoch": 9.968233020216136,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0038393856982884e-07,
+      "loss": 0.9617569732666016,
       "step": 865800
     },
     {
-      "epoch": 19.941051855945474,
-      "grad_norm": 2.884814977645874,
-      "learning_rate": 1.0449237212955019e-07,
-      "loss": 0.5142,
+      "epoch": 9.970535705290851,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.311237474731316e-08,
+      "loss": 0.9744757843017579,
       "step": 866000
     },
     {
-      "epoch": 19.945657179699733,
-      "grad_norm": 3.6238908767700195,
-      "learning_rate": 9.722080830703451e-08,
-      "loss": 0.5142,
+      "epoch": 9.972838390365565,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.584081092479749e-08,
+      "loss": 1.0132231140136718,
       "step": 866200
     },
     {
-      "epoch": 19.95026250345399,
-      "grad_norm": 2.9918384552001953,
-      "learning_rate": 8.994924448451885e-08,
-      "loss": 0.518,
+      "epoch": 9.97514107544028,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.856924710228181e-08,
+      "loss": 0.9756494903564453,
       "step": 866400
     },
     {
-      "epoch": 19.954867827208254,
-      "grad_norm": 3.4772942066192627,
-      "learning_rate": 8.267768066200316e-08,
-      "loss": 0.5172,
+      "epoch": 9.977443760514996,
+      "grad_norm": 1.4296875,
+      "learning_rate": 7.129768327976615e-08,
+      "loss": 0.9711592102050781,
       "step": 866600
     },
     {
-      "epoch": 19.959473150962513,
-      "grad_norm": 3.1812825202941895,
-      "learning_rate": 7.54061168394875e-08,
-      "loss": 0.5074,
+      "epoch": 9.979746445589711,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.402611945725048e-08,
+      "loss": 1.0018580627441407,
       "step": 866800
     },
     {
-      "epoch": 19.96407847471677,
-      "grad_norm": 3.3182106018066406,
-      "learning_rate": 6.813455301697183e-08,
-      "loss": 0.5054,
+      "epoch": 9.982049130664425,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.675455563473481e-08,
+      "loss": 1.0008620452880859,
       "step": 867000
     },
     {
-      "epoch": 19.968683798471034,
-      "grad_norm": 3.577573299407959,
-      "learning_rate": 6.086298919445617e-08,
-      "loss": 0.5108,
+      "epoch": 9.98435181573914,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.948299181221913e-08,
+      "loss": 0.9791983795166016,
       "step": 867200
     },
     {
-      "epoch": 19.973289122225292,
-      "grad_norm": 3.277371644973755,
-      "learning_rate": 5.359142537194049e-08,
-      "loss": 0.5053,
+      "epoch": 9.986654500813856,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.2211427989703465e-08,
+      "loss": 0.9899475860595703,
       "step": 867400
     },
     {
-      "epoch": 19.97789444597955,
-      "grad_norm": 2.689061164855957,
-      "learning_rate": 4.631986154942482e-08,
-      "loss": 0.516,
+      "epoch": 9.98895718588857,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.49398641671878e-08,
+      "loss": 0.9756233215332031,
       "step": 867600
     },
     {
-      "epoch": 19.982499769733813,
-      "grad_norm": 2.8559155464172363,
-      "learning_rate": 3.904829772690915e-08,
-      "loss": 0.5057,
+      "epoch": 9.991259870963285,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.7668300344672125e-08,
+      "loss": 0.9905976867675781,
       "step": 867800
     },
     {
-      "epoch": 19.987105093488072,
-      "grad_norm": 3.4477171897888184,
-      "learning_rate": 3.177673390439348e-08,
-      "loss": 0.5065,
+      "epoch": 9.993562556038,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.0396736522156457e-08,
+      "loss": 0.9811268615722656,
       "step": 868000
     },
     {
-      "epoch": 19.99171041724233,
-      "grad_norm": 3.181300163269043,
-      "learning_rate": 2.450517008187781e-08,
-      "loss": 0.5202,
+      "epoch": 9.995865241112716,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.3125172699640784e-08,
+      "loss": 0.9911077880859375,
       "step": 868200
     },
     {
-      "epoch": 19.996315740996593,
-      "grad_norm": 2.9424209594726562,
-      "learning_rate": 1.723360625936214e-08,
-      "loss": 0.5106,
+      "epoch": 9.99816792618743,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.853608877125114e-09,
+      "loss": 0.989971923828125,
       "step": 868400
     },
     {
-      "epoch": 20.0,
-      "eval_loss": 0.5036933422088623,
-      "eval_runtime": 162.2891,
-      "eval_samples_per_second": 174.756,
-      "eval_steps_per_second": 10.925,
+      "epoch": 10.0,
+      "eval_loss": NaN,
+      "eval_runtime": 431.6936,
+      "eval_samples_per_second": 65.697,
+      "eval_steps_per_second": 32.85,
       "step": 868560
     }
   ],
   "logging_steps": 200,
   "max_steps": 868560,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 20,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -30581,8 +30501,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.661507840870656e+18,
-  "train_batch_size": 16,
+  "total_flos": 9.814236186142986e+18,
+  "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null
 }