diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10577 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 752982,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019920794919400463,
+      "grad_norm": 2.0818653106689453,
+      "learning_rate": 4.996679867513434e-05,
+      "loss": 1.7687,
+      "step": 500
+    },
+    {
+      "epoch": 0.003984158983880093,
+      "grad_norm": 2.0961356163024902,
+      "learning_rate": 4.993359735026867e-05,
+      "loss": 1.6725,
+      "step": 1000
+    },
+    {
+      "epoch": 0.005976238475820139,
+      "grad_norm": 2.148750066757202,
+      "learning_rate": 4.9900396025403e-05,
+      "loss": 1.6113,
+      "step": 1500
+    },
+    {
+      "epoch": 0.007968317967760185,
+      "grad_norm": 2.0956828594207764,
+      "learning_rate": 4.986719470053733e-05,
+      "loss": 1.579,
+      "step": 2000
+    },
+    {
+      "epoch": 0.009960397459700232,
+      "grad_norm": 2.015962600708008,
+      "learning_rate": 4.983399337567167e-05,
+      "loss": 1.547,
+      "step": 2500
+    },
+    {
+      "epoch": 0.011952476951640279,
+      "grad_norm": 2.0305891036987305,
+      "learning_rate": 4.9800792050806e-05,
+      "loss": 1.5297,
+      "step": 3000
+    },
+    {
+      "epoch": 0.013944556443580324,
+      "grad_norm": 2.0634851455688477,
+      "learning_rate": 4.976759072594033e-05,
+      "loss": 1.5094,
+      "step": 3500
+    },
+    {
+      "epoch": 0.01593663593552037,
+      "grad_norm": 1.981088399887085,
+      "learning_rate": 4.973438940107466e-05,
+      "loss": 1.4947,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01792871542746042,
+      "grad_norm": 2.009782314300537,
+      "learning_rate": 4.9701188076208996e-05,
+      "loss": 1.4825,
+      "step": 4500
+    },
+    {
+      "epoch": 0.019920794919400464,
+      "grad_norm": 1.9277538061141968,
+      "learning_rate": 4.9667986751343325e-05,
+      "loss": 1.4705,
+      "step": 5000
+    },
+    {
+      "epoch": 0.02191287441134051,
+      "grad_norm": 1.9650582075119019,
+      "learning_rate": 4.963478542647766e-05,
+      "loss": 1.4568,
+      "step": 5500
+    },
+    {
+      "epoch": 0.023904953903280558,
+      "grad_norm": 1.8561341762542725,
+      "learning_rate": 4.960158410161199e-05,
+      "loss": 1.4463,
+      "step": 6000
+    },
+    {
+      "epoch": 0.025897033395220603,
+      "grad_norm": 1.891862154006958,
+      "learning_rate": 4.9568382776746325e-05,
+      "loss": 1.4398,
+      "step": 6500
+    },
+    {
+      "epoch": 0.027889112887160648,
+      "grad_norm": 1.8454586267471313,
+      "learning_rate": 4.953518145188066e-05,
+      "loss": 1.4338,
+      "step": 7000
+    },
+    {
+      "epoch": 0.029881192379100696,
+      "grad_norm": 1.9231008291244507,
+      "learning_rate": 4.950198012701499e-05,
+      "loss": 1.4266,
+      "step": 7500
+    },
+    {
+      "epoch": 0.03187327187104074,
+      "grad_norm": 1.8806790113449097,
+      "learning_rate": 4.9468778802149325e-05,
+      "loss": 1.4136,
+      "step": 8000
+    },
+    {
+      "epoch": 0.033865351362980786,
+      "grad_norm": 1.818979263305664,
+      "learning_rate": 4.9435577477283654e-05,
+      "loss": 1.4037,
+      "step": 8500
+    },
+    {
+      "epoch": 0.03585743085492084,
+      "grad_norm": 1.8734707832336426,
+      "learning_rate": 4.940237615241799e-05,
+      "loss": 1.3966,
+      "step": 9000
+    },
+    {
+      "epoch": 0.03784951034686088,
+      "grad_norm": 1.8471806049346924,
+      "learning_rate": 4.9369174827552325e-05,
+      "loss": 1.3912,
+      "step": 9500
+    },
+    {
+      "epoch": 0.03984158983880093,
+      "grad_norm": 1.7787472009658813,
+      "learning_rate": 4.9335973502686654e-05,
+      "loss": 1.3875,
+      "step": 10000
+    },
+    {
+      "epoch": 0.04183366933074097,
+      "grad_norm": 1.8721691370010376,
+      "learning_rate": 4.930277217782099e-05,
+      "loss": 1.3796,
+      "step": 10500
+    },
+    {
+      "epoch": 0.04382574882268102,
+      "grad_norm": 1.8752695322036743,
+      "learning_rate": 4.926957085295532e-05,
+      "loss": 1.3753,
+      "step": 11000
+    },
+    {
+      "epoch": 0.04581782831462106,
+      "grad_norm": 1.8108872175216675,
+      "learning_rate": 4.923636952808965e-05,
+      "loss": 1.3702,
+      "step": 11500
+    },
+    {
+      "epoch": 0.047809907806561115,
+      "grad_norm": 2.0425238609313965,
+      "learning_rate": 4.920316820322398e-05,
+      "loss": 1.3673,
+      "step": 12000
+    },
+    {
+      "epoch": 0.04980198729850116,
+      "grad_norm": 1.8299578428268433,
+      "learning_rate": 4.916996687835831e-05,
+      "loss": 1.3573,
+      "step": 12500
+    },
+    {
+      "epoch": 0.051794066790441205,
+      "grad_norm": 1.7092092037200928,
+      "learning_rate": 4.913676555349265e-05,
+      "loss": 1.356,
+      "step": 13000
+    },
+    {
+      "epoch": 0.05378614628238125,
+      "grad_norm": 1.8673481941223145,
+      "learning_rate": 4.910356422862698e-05,
+      "loss": 1.3549,
+      "step": 13500
+    },
+    {
+      "epoch": 0.055778225774321295,
+      "grad_norm": 1.7857859134674072,
+      "learning_rate": 4.907036290376131e-05,
+      "loss": 1.3474,
+      "step": 14000
+    },
+    {
+      "epoch": 0.05777030526626135,
+      "grad_norm": 1.7066707611083984,
+      "learning_rate": 4.903716157889565e-05,
+      "loss": 1.3402,
+      "step": 14500
+    },
+    {
+      "epoch": 0.05976238475820139,
+      "grad_norm": 1.8482805490493774,
+      "learning_rate": 4.9003960254029976e-05,
+      "loss": 1.34,
+      "step": 15000
+    },
+    {
+      "epoch": 0.06175446425014144,
+      "grad_norm": 1.8168840408325195,
+      "learning_rate": 4.897075892916431e-05,
+      "loss": 1.3378,
+      "step": 15500
+    },
+    {
+      "epoch": 0.06374654374208148,
+      "grad_norm": 1.7170970439910889,
+      "learning_rate": 4.893755760429865e-05,
+      "loss": 1.3294,
+      "step": 16000
+    },
+    {
+      "epoch": 0.06573862323402153,
+      "grad_norm": 1.8236738443374634,
+      "learning_rate": 4.8904356279432976e-05,
+      "loss": 1.3289,
+      "step": 16500
+    },
+    {
+      "epoch": 0.06773070272596157,
+      "grad_norm": 1.7402244806289673,
+      "learning_rate": 4.887115495456731e-05,
+      "loss": 1.323,
+      "step": 17000
+    },
+    {
+      "epoch": 0.06972278221790162,
+      "grad_norm": 1.6897387504577637,
+      "learning_rate": 4.883795362970164e-05,
+      "loss": 1.3181,
+      "step": 17500
+    },
+    {
+      "epoch": 0.07171486170984168,
+      "grad_norm": 1.6281301975250244,
+      "learning_rate": 4.8804752304835976e-05,
+      "loss": 1.3151,
+      "step": 18000
+    },
+    {
+      "epoch": 0.07370694120178171,
+      "grad_norm": 1.736750841140747,
+      "learning_rate": 4.877155097997031e-05,
+      "loss": 1.3123,
+      "step": 18500
+    },
+    {
+      "epoch": 0.07569902069372177,
+      "grad_norm": 2.108617067337036,
+      "learning_rate": 4.8738349655104634e-05,
+      "loss": 1.3103,
+      "step": 19000
+    },
+    {
+      "epoch": 0.0776911001856618,
+      "grad_norm": 1.7343802452087402,
+      "learning_rate": 4.870514833023897e-05,
+      "loss": 1.2977,
+      "step": 19500
+    },
+    {
+      "epoch": 0.07968317967760186,
+      "grad_norm": 1.7792308330535889,
+      "learning_rate": 4.8671947005373305e-05,
+      "loss": 1.2989,
+      "step": 20000
+    },
+    {
+      "epoch": 0.0816752591695419,
+      "grad_norm": 1.606641173362732,
+      "learning_rate": 4.8638745680507634e-05,
+      "loss": 1.3065,
+      "step": 20500
+    },
+    {
+      "epoch": 0.08366733866148195,
+      "grad_norm": 1.7401373386383057,
+      "learning_rate": 4.860554435564197e-05,
+      "loss": 1.2983,
+      "step": 21000
+    },
+    {
+      "epoch": 0.085659418153422,
+      "grad_norm": 1.6886597871780396,
+      "learning_rate": 4.85723430307763e-05,
+      "loss": 1.2934,
+      "step": 21500
+    },
+    {
+      "epoch": 0.08765149764536204,
+      "grad_norm": 1.6913639307022095,
+      "learning_rate": 4.8539141705910634e-05,
+      "loss": 1.2926,
+      "step": 22000
+    },
+    {
+      "epoch": 0.08964357713730209,
+      "grad_norm": 1.6619659662246704,
+      "learning_rate": 4.850594038104497e-05,
+      "loss": 1.2927,
+      "step": 22500
+    },
+    {
+      "epoch": 0.09163565662924213,
+      "grad_norm": 1.7290693521499634,
+      "learning_rate": 4.84727390561793e-05,
+      "loss": 1.2879,
+      "step": 23000
+    },
+    {
+      "epoch": 0.09362773612118218,
+      "grad_norm": 1.67817223072052,
+      "learning_rate": 4.8439537731313634e-05,
+      "loss": 1.2893,
+      "step": 23500
+    },
+    {
+      "epoch": 0.09561981561312223,
+      "grad_norm": 1.6673177480697632,
+      "learning_rate": 4.840633640644796e-05,
+      "loss": 1.2808,
+      "step": 24000
+    },
+    {
+      "epoch": 0.09761189510506227,
+      "grad_norm": 1.7721155881881714,
+      "learning_rate": 4.83731350815823e-05,
+      "loss": 1.2811,
+      "step": 24500
+    },
+    {
+      "epoch": 0.09960397459700232,
+      "grad_norm": 1.8109252452850342,
+      "learning_rate": 4.8339933756716634e-05,
+      "loss": 1.2796,
+      "step": 25000
+    },
+    {
+      "epoch": 0.10159605408894236,
+      "grad_norm": 1.8570215702056885,
+      "learning_rate": 4.830673243185096e-05,
+      "loss": 1.2807,
+      "step": 25500
+    },
+    {
+      "epoch": 0.10358813358088241,
+      "grad_norm": 1.7140320539474487,
+      "learning_rate": 4.82735311069853e-05,
+      "loss": 1.2701,
+      "step": 26000
+    },
+    {
+      "epoch": 0.10558021307282246,
+      "grad_norm": 1.6948349475860596,
+      "learning_rate": 4.824032978211963e-05,
+      "loss": 1.2688,
+      "step": 26500
+    },
+    {
+      "epoch": 0.1075722925647625,
+      "grad_norm": 1.6990410089492798,
+      "learning_rate": 4.8207128457253956e-05,
+      "loss": 1.2711,
+      "step": 27000
+    },
+    {
+      "epoch": 0.10956437205670255,
+      "grad_norm": 1.7393864393234253,
+      "learning_rate": 4.817392713238829e-05,
+      "loss": 1.2691,
+      "step": 27500
+    },
+    {
+      "epoch": 0.11155645154864259,
+      "grad_norm": 1.6735812425613403,
+      "learning_rate": 4.814072580752262e-05,
+      "loss": 1.2627,
+      "step": 28000
+    },
+    {
+      "epoch": 0.11354853104058264,
+      "grad_norm": 1.7558528184890747,
+      "learning_rate": 4.8107524482656956e-05,
+      "loss": 1.2573,
+      "step": 28500
+    },
+    {
+      "epoch": 0.1155406105325227,
+      "grad_norm": 1.6510869264602661,
+      "learning_rate": 4.807432315779129e-05,
+      "loss": 1.2576,
+      "step": 29000
+    },
+    {
+      "epoch": 0.11753269002446273,
+      "grad_norm": 1.7427338361740112,
+      "learning_rate": 4.804112183292562e-05,
+      "loss": 1.255,
+      "step": 29500
+    },
+    {
+      "epoch": 0.11952476951640278,
+      "grad_norm": 1.7312365770339966,
+      "learning_rate": 4.8007920508059957e-05,
+      "loss": 1.2537,
+      "step": 30000
+    },
+    {
+      "epoch": 0.12151684900834282,
+      "grad_norm": 1.6289525032043457,
+      "learning_rate": 4.797471918319429e-05,
+      "loss": 1.2583,
+      "step": 30500
+    },
+    {
+      "epoch": 0.12350892850028287,
+      "grad_norm": 1.6537457704544067,
+      "learning_rate": 4.794151785832862e-05,
+      "loss": 1.2524,
+      "step": 31000
+    },
+    {
+      "epoch": 0.1255010079922229,
+      "grad_norm": 1.7153490781784058,
+      "learning_rate": 4.790831653346296e-05,
+      "loss": 1.2483,
+      "step": 31500
+    },
+    {
+      "epoch": 0.12749308748416296,
+      "grad_norm": 1.6941293478012085,
+      "learning_rate": 4.7875115208597285e-05,
+      "loss": 1.2501,
+      "step": 32000
+    },
+    {
+      "epoch": 0.12948516697610302,
+      "grad_norm": 1.7295812368392944,
+      "learning_rate": 4.784191388373162e-05,
+      "loss": 1.2479,
+      "step": 32500
+    },
+    {
+      "epoch": 0.13147724646804307,
+      "grad_norm": 1.6068220138549805,
+      "learning_rate": 4.780871255886596e-05,
+      "loss": 1.2472,
+      "step": 33000
+    },
+    {
+      "epoch": 0.13346932595998312,
+      "grad_norm": 1.6310116052627563,
+      "learning_rate": 4.7775511234000286e-05,
+      "loss": 1.24,
+      "step": 33500
+    },
+    {
+      "epoch": 0.13546140545192314,
+      "grad_norm": 1.6823935508728027,
+      "learning_rate": 4.7742309909134614e-05,
+      "loss": 1.2404,
+      "step": 34000
+    },
+    {
+      "epoch": 0.1374534849438632,
+      "grad_norm": 1.7916193008422852,
+      "learning_rate": 4.770910858426895e-05,
+      "loss": 1.2401,
+      "step": 34500
+    },
+    {
+      "epoch": 0.13944556443580325,
+      "grad_norm": 1.7779194116592407,
+      "learning_rate": 4.767590725940328e-05,
+      "loss": 1.2401,
+      "step": 35000
+    },
+    {
+      "epoch": 0.1414376439277433,
+      "grad_norm": 1.6148273944854736,
+      "learning_rate": 4.7642705934537614e-05,
+      "loss": 1.2338,
+      "step": 35500
+    },
+    {
+      "epoch": 0.14342972341968335,
+      "grad_norm": 1.6597368717193604,
+      "learning_rate": 4.760950460967194e-05,
+      "loss": 1.236,
+      "step": 36000
+    },
+    {
+      "epoch": 0.14542180291162338,
+      "grad_norm": 1.629267930984497,
+      "learning_rate": 4.757630328480628e-05,
+      "loss": 1.2341,
+      "step": 36500
+    },
+    {
+      "epoch": 0.14741388240356343,
+      "grad_norm": 1.6153258085250854,
+      "learning_rate": 4.7543101959940615e-05,
+      "loss": 1.2289,
+      "step": 37000
+    },
+    {
+      "epoch": 0.14940596189550348,
+      "grad_norm": 1.6635195016860962,
+      "learning_rate": 4.7509900635074943e-05,
+      "loss": 1.2328,
+      "step": 37500
+    },
+    {
+      "epoch": 0.15139804138744353,
+      "grad_norm": 1.6255178451538086,
+      "learning_rate": 4.747669931020928e-05,
+      "loss": 1.2332,
+      "step": 38000
+    },
+    {
+      "epoch": 0.15339012087938358,
+      "grad_norm": 1.6072956323623657,
+      "learning_rate": 4.744349798534361e-05,
+      "loss": 1.2288,
+      "step": 38500
+    },
+    {
+      "epoch": 0.1553822003713236,
+      "grad_norm": 1.6697778701782227,
+      "learning_rate": 4.7410296660477943e-05,
+      "loss": 1.2236,
+      "step": 39000
+    },
+    {
+      "epoch": 0.15737427986326366,
+      "grad_norm": 1.6340432167053223,
+      "learning_rate": 4.737709533561228e-05,
+      "loss": 1.2244,
+      "step": 39500
+    },
+    {
+      "epoch": 0.1593663593552037,
+      "grad_norm": 1.746995449066162,
+      "learning_rate": 4.734389401074661e-05,
+      "loss": 1.224,
+      "step": 40000
+    },
+    {
+      "epoch": 0.16135843884714376,
+      "grad_norm": 1.644507884979248,
+      "learning_rate": 4.7310692685880944e-05,
+      "loss": 1.2223,
+      "step": 40500
+    },
+    {
+      "epoch": 0.1633505183390838,
+      "grad_norm": 1.6474827527999878,
+      "learning_rate": 4.727749136101527e-05,
+      "loss": 1.2217,
+      "step": 41000
+    },
+    {
+      "epoch": 0.16534259783102384,
+      "grad_norm": 1.8044672012329102,
+      "learning_rate": 4.72442900361496e-05,
+      "loss": 1.2181,
+      "step": 41500
+    },
+    {
+      "epoch": 0.1673346773229639,
+      "grad_norm": 1.650636076927185,
+      "learning_rate": 4.721108871128394e-05,
+      "loss": 1.2182,
+      "step": 42000
+    },
+    {
+      "epoch": 0.16932675681490394,
+      "grad_norm": 1.7140562534332275,
+      "learning_rate": 4.7177887386418266e-05,
+      "loss": 1.2153,
+      "step": 42500
+    },
+    {
+      "epoch": 0.171318836306844,
+      "grad_norm": 2.3554928302764893,
+      "learning_rate": 4.71446860615526e-05,
+      "loss": 1.2146,
+      "step": 43000
+    },
+    {
+      "epoch": 0.17331091579878402,
+      "grad_norm": 1.6242191791534424,
+      "learning_rate": 4.711148473668694e-05,
+      "loss": 1.2078,
+      "step": 43500
+    },
+    {
+      "epoch": 0.17530299529072407,
+      "grad_norm": 1.851409673690796,
+      "learning_rate": 4.7078283411821266e-05,
+      "loss": 1.2139,
+      "step": 44000
+    },
+    {
+      "epoch": 0.17729507478266412,
+      "grad_norm": 1.830284833908081,
+      "learning_rate": 4.70450820869556e-05,
+      "loss": 1.2121,
+      "step": 44500
+    },
+    {
+      "epoch": 0.17928715427460418,
+      "grad_norm": 1.5572370290756226,
+      "learning_rate": 4.701188076208993e-05,
+      "loss": 1.2121,
+      "step": 45000
+    },
+    {
+      "epoch": 0.18127923376654423,
+      "grad_norm": 1.6135003566741943,
+      "learning_rate": 4.6978679437224266e-05,
+      "loss": 1.208,
+      "step": 45500
+    },
+    {
+      "epoch": 0.18327131325848425,
+      "grad_norm": 1.5585676431655884,
+      "learning_rate": 4.69454781123586e-05,
+      "loss": 1.2085,
+      "step": 46000
+    },
+    {
+      "epoch": 0.1852633927504243,
+      "grad_norm": 1.6594507694244385,
+      "learning_rate": 4.691227678749293e-05,
+      "loss": 1.2065,
+      "step": 46500
+    },
+    {
+      "epoch": 0.18725547224236436,
+      "grad_norm": 1.5677218437194824,
+      "learning_rate": 4.6879075462627266e-05,
+      "loss": 1.2006,
+      "step": 47000
+    },
+    {
+      "epoch": 0.1892475517343044,
+      "grad_norm": 1.6618664264678955,
+      "learning_rate": 4.6845874137761595e-05,
+      "loss": 1.2093,
+      "step": 47500
+    },
+    {
+      "epoch": 0.19123963122624446,
+      "grad_norm": 1.7201578617095947,
+      "learning_rate": 4.681267281289593e-05,
+      "loss": 1.2009,
+      "step": 48000
+    },
+    {
+      "epoch": 0.19323171071818449,
+      "grad_norm": 1.6738592386245728,
+      "learning_rate": 4.6779471488030266e-05,
+      "loss": 1.1967,
+      "step": 48500
+    },
+    {
+      "epoch": 0.19522379021012454,
+      "grad_norm": 1.5782294273376465,
+      "learning_rate": 4.674627016316459e-05,
+      "loss": 1.1993,
+      "step": 49000
+    },
+    {
+      "epoch": 0.1972158697020646,
+      "grad_norm": 1.541048288345337,
+      "learning_rate": 4.6713068838298924e-05,
+      "loss": 1.1984,
+      "step": 49500
+    },
+    {
+      "epoch": 0.19920794919400464,
+      "grad_norm": 1.676732063293457,
+      "learning_rate": 4.667986751343326e-05,
+      "loss": 1.2028,
+      "step": 50000
+    },
+    {
+      "epoch": 0.2012000286859447,
+      "grad_norm": 1.5797938108444214,
+      "learning_rate": 4.664666618856759e-05,
+      "loss": 1.1964,
+      "step": 50500
+    },
+    {
+      "epoch": 0.20319210817788472,
+      "grad_norm": 1.6108742952346802,
+      "learning_rate": 4.6613464863701924e-05,
+      "loss": 1.1996,
+      "step": 51000
+    },
+    {
+      "epoch": 0.20518418766982477,
+      "grad_norm": 1.7384244203567505,
+      "learning_rate": 4.658026353883625e-05,
+      "loss": 1.1933,
+      "step": 51500
+    },
+    {
+      "epoch": 0.20717626716176482,
+      "grad_norm": 1.726766586303711,
+      "learning_rate": 4.654706221397059e-05,
+      "loss": 1.1974,
+      "step": 52000
+    },
+    {
+      "epoch": 0.20916834665370487,
+      "grad_norm": 1.5999727249145508,
+      "learning_rate": 4.6513860889104924e-05,
+      "loss": 1.1965,
+      "step": 52500
+    },
+    {
+      "epoch": 0.21116042614564492,
+      "grad_norm": 1.6723779439926147,
+      "learning_rate": 4.648065956423925e-05,
+      "loss": 1.1916,
+      "step": 53000
+    },
+    {
+      "epoch": 0.21315250563758495,
+      "grad_norm": 1.570471167564392,
+      "learning_rate": 4.644745823937359e-05,
+      "loss": 1.1901,
+      "step": 53500
+    },
+    {
+      "epoch": 0.215144585129525,
+      "grad_norm": 1.5597429275512695,
+      "learning_rate": 4.641425691450792e-05,
+      "loss": 1.1884,
+      "step": 54000
+    },
+    {
+      "epoch": 0.21713666462146505,
+      "grad_norm": 1.6004964113235474,
+      "learning_rate": 4.638105558964225e-05,
+      "loss": 1.1882,
+      "step": 54500
+    },
+    {
+      "epoch": 0.2191287441134051,
+      "grad_norm": 1.6097034215927124,
+      "learning_rate": 4.634785426477659e-05,
+      "loss": 1.1859,
+      "step": 55000
+    },
+    {
+      "epoch": 0.22112082360534516,
+      "grad_norm": 1.5970678329467773,
+      "learning_rate": 4.631465293991092e-05,
+      "loss": 1.1885,
+      "step": 55500
+    },
+    {
+      "epoch": 0.22311290309728518,
+      "grad_norm": 1.6068161725997925,
+      "learning_rate": 4.628145161504525e-05,
+      "loss": 1.1839,
+      "step": 56000
+    },
+    {
+      "epoch": 0.22510498258922523,
+      "grad_norm": 1.5946305990219116,
+      "learning_rate": 4.624825029017958e-05,
+      "loss": 1.187,
+      "step": 56500
+    },
+    {
+      "epoch": 0.22709706208116529,
+      "grad_norm": 1.630587100982666,
+      "learning_rate": 4.621504896531391e-05,
+      "loss": 1.181,
+      "step": 57000
+    },
+    {
+      "epoch": 0.22908914157310534,
+      "grad_norm": 1.6598210334777832,
+      "learning_rate": 4.6181847640448246e-05,
+      "loss": 1.1809,
+      "step": 57500
+    },
+    {
+      "epoch": 0.2310812210650454,
+      "grad_norm": 1.6476484537124634,
+      "learning_rate": 4.6148646315582575e-05,
+      "loss": 1.1804,
+      "step": 58000
+    },
+    {
+      "epoch": 0.2330733005569854,
+      "grad_norm": 1.6074743270874023,
+      "learning_rate": 4.611544499071691e-05,
+      "loss": 1.1799,
+      "step": 58500
+    },
+    {
+      "epoch": 0.23506538004892547,
+      "grad_norm": 1.6919032335281372,
+      "learning_rate": 4.6082243665851246e-05,
+      "loss": 1.1785,
+      "step": 59000
+    },
+    {
+      "epoch": 0.23705745954086552,
+      "grad_norm": 1.6016688346862793,
+      "learning_rate": 4.6049042340985575e-05,
+      "loss": 1.1795,
+      "step": 59500
+    },
+    {
+      "epoch": 0.23904953903280557,
+      "grad_norm": 1.6147820949554443,
+      "learning_rate": 4.601584101611991e-05,
+      "loss": 1.1815,
+      "step": 60000
+    },
+    {
+      "epoch": 0.24104161852474562,
+      "grad_norm": 1.6349416971206665,
+      "learning_rate": 4.598263969125424e-05,
+      "loss": 1.1777,
+      "step": 60500
+    },
+    {
+      "epoch": 0.24303369801668565,
+      "grad_norm": 1.5517864227294922,
+      "learning_rate": 4.5949438366388575e-05,
+      "loss": 1.1784,
+      "step": 61000
+    },
+    {
+      "epoch": 0.2450257775086257,
+      "grad_norm": 1.5386013984680176,
+      "learning_rate": 4.591623704152291e-05,
+      "loss": 1.1743,
+      "step": 61500
+    },
+    {
+      "epoch": 0.24701785700056575,
+      "grad_norm": 1.5564523935317993,
+      "learning_rate": 4.588303571665724e-05,
+      "loss": 1.177,
+      "step": 62000
+    },
+    {
+      "epoch": 0.2490099364925058,
+      "grad_norm": 1.653681755065918,
+      "learning_rate": 4.5849834391791575e-05,
+      "loss": 1.1742,
+      "step": 62500
+    },
+    {
+      "epoch": 0.2510020159844458,
+      "grad_norm": 1.7013434171676636,
+      "learning_rate": 4.5816633066925904e-05,
+      "loss": 1.1711,
+      "step": 63000
+    },
+    {
+      "epoch": 0.2529940954763859,
+      "grad_norm": 1.5096421241760254,
+      "learning_rate": 4.578343174206024e-05,
+      "loss": 1.177,
+      "step": 63500
+    },
+    {
+      "epoch": 0.25498617496832593,
+      "grad_norm": 1.5327662229537964,
+      "learning_rate": 4.575023041719457e-05,
+      "loss": 1.1743,
+      "step": 64000
+    },
+    {
+      "epoch": 0.256978254460266,
+      "grad_norm": 1.5677859783172607,
+      "learning_rate": 4.57170290923289e-05,
+      "loss": 1.1713,
+      "step": 64500
+    },
+    {
+      "epoch": 0.25897033395220603,
+      "grad_norm": 1.695319652557373,
+      "learning_rate": 4.568382776746323e-05,
+      "loss": 1.168,
+      "step": 65000
+    },
+    {
+      "epoch": 0.26096241344414606,
+      "grad_norm": 1.5221295356750488,
+      "learning_rate": 4.565062644259757e-05,
+      "loss": 1.1711,
+      "step": 65500
+    },
+    {
+      "epoch": 0.26295449293608614,
+      "grad_norm": 1.6128042936325073,
+      "learning_rate": 4.56174251177319e-05,
+      "loss": 1.1693,
+      "step": 66000
+    },
+    {
+      "epoch": 0.26494657242802616,
+      "grad_norm": 1.570115089416504,
+      "learning_rate": 4.558422379286623e-05,
+      "loss": 1.1682,
+      "step": 66500
+    },
+    {
+      "epoch": 0.26693865191996624,
+      "grad_norm": 1.59817373752594,
+      "learning_rate": 4.555102246800056e-05,
+      "loss": 1.1697,
+      "step": 67000
+    },
+    {
+      "epoch": 0.26893073141190627,
+      "grad_norm": 1.619484543800354,
+      "learning_rate": 4.55178211431349e-05,
+      "loss": 1.1646,
+      "step": 67500
+    },
+    {
+      "epoch": 0.2709228109038463,
+      "grad_norm": 1.588550329208374,
+      "learning_rate": 4.548461981826923e-05,
+      "loss": 1.164,
+      "step": 68000
+    },
+    {
+      "epoch": 0.27291489039578637,
+      "grad_norm": 1.586408257484436,
+      "learning_rate": 4.545141849340356e-05,
+      "loss": 1.1612,
+      "step": 68500
+    },
+    {
+      "epoch": 0.2749069698877264,
+      "grad_norm": 1.7903696298599243,
+      "learning_rate": 4.54182171685379e-05,
+      "loss": 1.162,
+      "step": 69000
+    },
+    {
+      "epoch": 0.2768990493796665,
+      "grad_norm": 1.594846487045288,
+      "learning_rate": 4.5385015843672226e-05,
+      "loss": 1.1659,
+      "step": 69500
+    },
+    {
+      "epoch": 0.2788911288716065,
+      "grad_norm": 1.6955819129943848,
+      "learning_rate": 4.535181451880656e-05,
+      "loss": 1.1628,
+      "step": 70000
+    },
+    {
+      "epoch": 0.2808832083635465,
+      "grad_norm": 1.6978071928024292,
+      "learning_rate": 4.53186131939409e-05,
+      "loss": 1.1602,
+      "step": 70500
+    },
+    {
+      "epoch": 0.2828752878554866,
+      "grad_norm": 1.6447266340255737,
+      "learning_rate": 4.5285411869075227e-05,
+      "loss": 1.1591,
+      "step": 71000
+    },
+    {
+      "epoch": 0.2848673673474266,
+      "grad_norm": 1.6464730501174927,
+      "learning_rate": 4.5252210544209555e-05,
+      "loss": 1.1606,
+      "step": 71500
+    },
+    {
+      "epoch": 0.2868594468393667,
+      "grad_norm": 1.555372714996338,
+      "learning_rate": 4.521900921934389e-05,
+      "loss": 1.1585,
+      "step": 72000
+    },
+    {
+      "epoch": 0.28885152633130673,
+      "grad_norm": 1.6100929975509644,
+      "learning_rate": 4.518580789447822e-05,
+      "loss": 1.1607,
+      "step": 72500
+    },
+    {
+      "epoch": 0.29084360582324675,
+      "grad_norm": 1.629961371421814,
+      "learning_rate": 4.5152606569612555e-05,
+      "loss": 1.1576,
+      "step": 73000
+    },
+    {
+      "epoch": 0.29283568531518683,
+      "grad_norm": 1.6332063674926758,
+      "learning_rate": 4.5119405244746884e-05,
+      "loss": 1.1538,
+      "step": 73500
+    },
+    {
+      "epoch": 0.29482776480712686,
+      "grad_norm": 1.6296850442886353,
+      "learning_rate": 4.508620391988122e-05,
+      "loss": 1.1515,
+      "step": 74000
+    },
+    {
+      "epoch": 0.29681984429906694,
+      "grad_norm": 1.6172661781311035,
+      "learning_rate": 4.5053002595015556e-05,
+      "loss": 1.1564,
+      "step": 74500
+    },
+    {
+      "epoch": 0.29881192379100696,
+      "grad_norm": 1.5800873041152954,
+      "learning_rate": 4.5019801270149884e-05,
+      "loss": 1.1496,
+      "step": 75000
+    },
+    {
+      "epoch": 0.300804003282947,
+      "grad_norm": 1.6038795709609985,
+      "learning_rate": 4.498659994528422e-05,
+      "loss": 1.1539,
+      "step": 75500
+    },
+    {
+      "epoch": 0.30279608277488707,
+      "grad_norm": 1.6404781341552734,
+      "learning_rate": 4.495339862041855e-05,
+      "loss": 1.1535,
+      "step": 76000
+    },
+    {
+      "epoch": 0.3047881622668271,
+      "grad_norm": 1.5564305782318115,
+      "learning_rate": 4.4920197295552885e-05,
+      "loss": 1.1512,
+      "step": 76500
+    },
+    {
+      "epoch": 0.30678024175876717,
+      "grad_norm": 1.5874309539794922,
+      "learning_rate": 4.488699597068722e-05,
+      "loss": 1.1527,
+      "step": 77000
+    },
+    {
+      "epoch": 0.3087723212507072,
+      "grad_norm": 1.5185225009918213,
+      "learning_rate": 4.485379464582155e-05,
+      "loss": 1.1512,
+      "step": 77500
+    },
+    {
+      "epoch": 0.3107644007426472,
+      "grad_norm": 1.5815175771713257,
+      "learning_rate": 4.4820593320955885e-05,
+      "loss": 1.1494,
+      "step": 78000
+    },
+    {
+      "epoch": 0.3127564802345873,
+      "grad_norm": 1.651529312133789,
+      "learning_rate": 4.4787391996090213e-05,
+      "loss": 1.1515,
+      "step": 78500
+    },
+    {
+      "epoch": 0.3147485597265273,
+      "grad_norm": 1.5654505491256714,
+      "learning_rate": 4.475419067122454e-05,
+      "loss": 1.1456,
+      "step": 79000
+    },
+    {
+      "epoch": 0.3167406392184674,
+      "grad_norm": 1.5761442184448242,
+      "learning_rate": 4.472098934635888e-05,
+      "loss": 1.1494,
+      "step": 79500
+    },
+    {
+      "epoch": 0.3187327187104074,
+      "grad_norm": 1.5618621110916138,
+      "learning_rate": 4.468778802149321e-05,
+      "loss": 1.1469,
+      "step": 80000
+    },
+    {
+      "epoch": 0.32072479820234745,
+      "grad_norm": 1.573455810546875,
+      "learning_rate": 4.465458669662754e-05,
+      "loss": 1.1432,
+      "step": 80500
+    },
+    {
+      "epoch": 0.32271687769428753,
+      "grad_norm": 1.6327378749847412,
+      "learning_rate": 4.462138537176188e-05,
+      "loss": 1.1492,
+      "step": 81000
+    },
+    {
+      "epoch": 0.32470895718622755,
+      "grad_norm": 1.584782361984253,
+      "learning_rate": 4.458818404689621e-05,
+      "loss": 1.1456,
+      "step": 81500
+    },
+    {
+      "epoch": 0.3267010366781676,
+      "grad_norm": 1.615024209022522,
+      "learning_rate": 4.455498272203054e-05,
+      "loss": 1.1432,
+      "step": 82000
+    },
+    {
+      "epoch": 0.32869311617010766,
+      "grad_norm": 1.6665747165679932,
+      "learning_rate": 4.452178139716487e-05,
+      "loss": 1.1436,
+      "step": 82500
+    },
+    {
+      "epoch": 0.3306851956620477,
+      "grad_norm": 1.6672935485839844,
+      "learning_rate": 4.448858007229921e-05,
+      "loss": 1.1423,
+      "step": 83000
+    },
+    {
+      "epoch": 0.33267727515398776,
+      "grad_norm": 1.585807204246521,
+      "learning_rate": 4.445537874743354e-05,
+      "loss": 1.1435,
+      "step": 83500
+    },
+    {
+      "epoch": 0.3346693546459278,
+      "grad_norm": 1.60165274143219,
+      "learning_rate": 4.442217742256787e-05,
+      "loss": 1.1447,
+      "step": 84000
+    },
+    {
+      "epoch": 0.3366614341378678,
+      "grad_norm": 1.5934467315673828,
+      "learning_rate": 4.438897609770221e-05,
+      "loss": 1.1402,
+      "step": 84500
+    },
+    {
+      "epoch": 0.3386535136298079,
+      "grad_norm": 1.50346040725708,
+      "learning_rate": 4.4355774772836536e-05,
+      "loss": 1.1429,
+      "step": 85000
+    },
+    {
+      "epoch": 0.3406455931217479,
+      "grad_norm": 1.6559315919876099,
+      "learning_rate": 4.432257344797087e-05,
+      "loss": 1.1382,
+      "step": 85500
+    },
+    {
+      "epoch": 0.342637672613688,
+      "grad_norm": 1.5334893465042114,
+      "learning_rate": 4.428937212310521e-05,
+      "loss": 1.1334,
+      "step": 86000
+    },
+    {
+      "epoch": 0.344629752105628,
+      "grad_norm": 1.5708363056182861,
+      "learning_rate": 4.425617079823953e-05,
+      "loss": 1.1372,
+      "step": 86500
+    },
+    {
+      "epoch": 0.34662183159756804,
+      "grad_norm": 1.5030122995376587,
+      "learning_rate": 4.4222969473373865e-05,
+      "loss": 1.1378,
+      "step": 87000
+    },
+    {
+      "epoch": 0.3486139110895081,
+      "grad_norm": 1.596693992614746,
+      "learning_rate": 4.41897681485082e-05,
+      "loss": 1.1378,
+      "step": 87500
+    },
+    {
+      "epoch": 0.35060599058144815,
+      "grad_norm": 1.5458290576934814,
+      "learning_rate": 4.415656682364253e-05,
+      "loss": 1.1321,
+      "step": 88000
+    },
+    {
+      "epoch": 0.3525980700733882,
+      "grad_norm": 1.5623424053192139,
+      "learning_rate": 4.4123365498776865e-05,
+      "loss": 1.1345,
+      "step": 88500
+    },
+    {
+      "epoch": 0.35459014956532825,
+      "grad_norm": 1.6507903337478638,
+      "learning_rate": 4.4090164173911194e-05,
+      "loss": 1.1356,
+      "step": 89000
+    },
+    {
+      "epoch": 0.3565822290572683,
+      "grad_norm": 1.6864041090011597,
+      "learning_rate": 4.405696284904553e-05,
+      "loss": 1.1378,
+      "step": 89500
+    },
+    {
+      "epoch": 0.35857430854920835,
+      "grad_norm": 1.5543265342712402,
+      "learning_rate": 4.4023761524179865e-05,
+      "loss": 1.1327,
+      "step": 90000
+    },
+    {
+      "epoch": 0.3605663880411484,
+      "grad_norm": 1.5025701522827148,
+      "learning_rate": 4.3990560199314194e-05,
+      "loss": 1.1303,
+      "step": 90500
+    },
+    {
+      "epoch": 0.36255846753308846,
+      "grad_norm": 1.604069709777832,
+      "learning_rate": 4.395735887444853e-05,
+      "loss": 1.1348,
+      "step": 91000
+    },
+    {
+      "epoch": 0.3645505470250285,
+      "grad_norm": 1.5211833715438843,
+      "learning_rate": 4.3924157549582865e-05,
+      "loss": 1.1289,
+      "step": 91500
+    },
+    {
+      "epoch": 0.3665426265169685,
+      "grad_norm": 1.5616035461425781,
+      "learning_rate": 4.3890956224717194e-05,
+      "loss": 1.1336,
+      "step": 92000
+    },
+    {
+      "epoch": 0.3685347060089086,
+      "grad_norm": 1.5745919942855835,
+      "learning_rate": 4.385775489985153e-05,
+      "loss": 1.1324,
+      "step": 92500
+    },
+    {
+      "epoch": 0.3705267855008486,
+      "grad_norm": 1.5644596815109253,
+      "learning_rate": 4.382455357498586e-05,
+      "loss": 1.1311,
+      "step": 93000
+    },
+    {
+      "epoch": 0.3725188649927887,
+      "grad_norm": 1.5604528188705444,
+      "learning_rate": 4.3791352250120194e-05,
+      "loss": 1.1296,
+      "step": 93500
+    },
+    {
+      "epoch": 0.3745109444847287,
+      "grad_norm": 1.656253695487976,
+      "learning_rate": 4.375815092525452e-05,
+      "loss": 1.1296,
+      "step": 94000
+    },
+    {
+      "epoch": 0.37650302397666874,
+      "grad_norm": 1.557961106300354,
+      "learning_rate": 4.372494960038885e-05,
+      "loss": 1.1283,
+      "step": 94500
+    },
+    {
+      "epoch": 0.3784951034686088,
+      "grad_norm": 1.6911364793777466,
+      "learning_rate": 4.369174827552319e-05,
+      "loss": 1.1289,
+      "step": 95000
+    },
+    {
+      "epoch": 0.38048718296054884,
+      "grad_norm": 1.645554542541504,
+      "learning_rate": 4.365854695065752e-05,
+      "loss": 1.1282,
+      "step": 95500
+    },
+    {
+      "epoch": 0.3824792624524889,
+      "grad_norm": 1.6432863473892212,
+      "learning_rate": 4.362534562579185e-05,
+      "loss": 1.1246,
+      "step": 96000
+    },
+    {
+      "epoch": 0.38447134194442895,
+      "grad_norm": 1.6187262535095215,
+      "learning_rate": 4.359214430092619e-05,
+      "loss": 1.1276,
+      "step": 96500
+    },
+    {
+      "epoch": 0.38646342143636897,
+      "grad_norm": 1.5006526708602905,
+      "learning_rate": 4.3558942976060516e-05,
+      "loss": 1.1252,
+      "step": 97000
+    },
+    {
+      "epoch": 0.38845550092830905,
+      "grad_norm": 1.534436821937561,
+      "learning_rate": 4.352574165119485e-05,
+      "loss": 1.1262,
+      "step": 97500
+    },
+    {
+      "epoch": 0.3904475804202491,
+      "grad_norm": 1.6795482635498047,
+      "learning_rate": 4.349254032632919e-05,
+      "loss": 1.1219,
+      "step": 98000
+    },
+    {
+      "epoch": 0.39243965991218915,
+      "grad_norm": 1.4863214492797852,
+      "learning_rate": 4.3459339001463516e-05,
+      "loss": 1.1189,
+      "step": 98500
+    },
+    {
+      "epoch": 0.3944317394041292,
+      "grad_norm": 1.5260475873947144,
+      "learning_rate": 4.342613767659785e-05,
+      "loss": 1.1257,
+      "step": 99000
+    },
+    {
+      "epoch": 0.3964238188960692,
+      "grad_norm": 1.539004921913147,
+      "learning_rate": 4.339293635173218e-05,
+      "loss": 1.123,
+      "step": 99500
+    },
+    {
+      "epoch": 0.3984158983880093,
+      "grad_norm": 1.566393256187439,
+      "learning_rate": 4.3359735026866516e-05,
+      "loss": 1.1193,
+      "step": 100000
+    },
+    {
+      "epoch": 0.4004079778799493,
+      "grad_norm": 1.639440655708313,
+      "learning_rate": 4.332653370200085e-05,
+      "loss": 1.12,
+      "step": 100500
+    },
+    {
+      "epoch": 0.4024000573718894,
+      "grad_norm": 1.5583209991455078,
+      "learning_rate": 4.329333237713518e-05,
+      "loss": 1.1208,
+      "step": 101000
+    },
+    {
+      "epoch": 0.4043921368638294,
+      "grad_norm": 1.5819141864776611,
+      "learning_rate": 4.326013105226951e-05,
+      "loss": 1.1193,
+      "step": 101500
+    },
+    {
+      "epoch": 0.40638421635576943,
+      "grad_norm": 1.5694029331207275,
+      "learning_rate": 4.3226929727403845e-05,
+      "loss": 1.1217,
+      "step": 102000
+    },
+    {
+      "epoch": 0.4083762958477095,
+      "grad_norm": 1.5695774555206299,
+      "learning_rate": 4.3193728402538174e-05,
+      "loss": 1.1142,
+      "step": 102500
+    },
+    {
+      "epoch": 0.41036837533964954,
+      "grad_norm": 1.4776016473770142,
+      "learning_rate": 4.316052707767251e-05,
+      "loss": 1.1203,
+      "step": 103000
+    },
+    {
+      "epoch": 0.4123604548315896,
+      "grad_norm": 1.6063282489776611,
+      "learning_rate": 4.312732575280684e-05,
+      "loss": 1.1226,
+      "step": 103500
+    },
+    {
+      "epoch": 0.41435253432352964,
+      "grad_norm": 1.6728895902633667,
+      "learning_rate": 4.3094124427941174e-05,
+      "loss": 1.1133,
+      "step": 104000
+    },
+    {
+      "epoch": 0.41634461381546967,
+      "grad_norm": 1.5661665201187134,
+      "learning_rate": 4.306092310307551e-05,
+      "loss": 1.1152,
+      "step": 104500
+    },
+    {
+      "epoch": 0.41833669330740975,
+      "grad_norm": 1.5983315706253052,
+      "learning_rate": 4.302772177820984e-05,
+      "loss": 1.1173,
+      "step": 105000
+    },
+    {
+      "epoch": 0.42032877279934977,
+      "grad_norm": 1.5964795351028442,
+      "learning_rate": 4.2994520453344174e-05,
+      "loss": 1.1168,
+      "step": 105500
+    },
+    {
+      "epoch": 0.42232085229128985,
+      "grad_norm": 1.6091669797897339,
+      "learning_rate": 4.29613191284785e-05,
+      "loss": 1.1122,
+      "step": 106000
+    },
+    {
+      "epoch": 0.4243129317832299,
+      "grad_norm": 1.5613433122634888,
+      "learning_rate": 4.292811780361284e-05,
+      "loss": 1.1154,
+      "step": 106500
+    },
+    {
+      "epoch": 0.4263050112751699,
+      "grad_norm": 1.507009506225586,
+      "learning_rate": 4.2894916478747174e-05,
+      "loss": 1.1175,
+      "step": 107000
+    },
+    {
+      "epoch": 0.42829709076711,
+      "grad_norm": 1.5644526481628418,
+      "learning_rate": 4.28617151538815e-05,
+      "loss": 1.1143,
+      "step": 107500
+    },
+    {
+      "epoch": 0.43028917025905,
+      "grad_norm": 1.5901525020599365,
+      "learning_rate": 4.282851382901584e-05,
+      "loss": 1.1083,
+      "step": 108000
+    },
+    {
+      "epoch": 0.4322812497509901,
+      "grad_norm": 1.5851898193359375,
+      "learning_rate": 4.279531250415017e-05,
+      "loss": 1.1166,
+      "step": 108500
+    },
+    {
+      "epoch": 0.4342733292429301,
+      "grad_norm": 1.5360541343688965,
+      "learning_rate": 4.2762111179284496e-05,
+      "loss": 1.112,
+      "step": 109000
+    },
+    {
+      "epoch": 0.43626540873487013,
+      "grad_norm": 1.7155137062072754,
+      "learning_rate": 4.272890985441883e-05,
+      "loss": 1.1066,
+      "step": 109500
+    },
+    {
+      "epoch": 0.4382574882268102,
+      "grad_norm": 1.5425488948822021,
+      "learning_rate": 4.269570852955316e-05,
+      "loss": 1.1119,
+      "step": 110000
+    },
+    {
+      "epoch": 0.44024956771875023,
+      "grad_norm": 1.579074740409851,
+      "learning_rate": 4.2662507204687497e-05,
+      "loss": 1.1136,
+      "step": 110500
+    },
+    {
+      "epoch": 0.4422416472106903,
+      "grad_norm": 1.6529029607772827,
+      "learning_rate": 4.262930587982183e-05,
+      "loss": 1.1158,
+      "step": 111000
+    },
+    {
+      "epoch": 0.44423372670263034,
+      "grad_norm": 2.756675958633423,
+      "learning_rate": 4.259610455495616e-05,
+      "loss": 1.1104,
+      "step": 111500
+    },
+    {
+      "epoch": 0.44622580619457036,
+      "grad_norm": 1.631072759628296,
+      "learning_rate": 4.2562903230090497e-05,
+      "loss": 1.108,
+      "step": 112000
+    },
+    {
+      "epoch": 0.44821788568651044,
+      "grad_norm": 1.621834397315979,
+      "learning_rate": 4.2529701905224825e-05,
+      "loss": 1.1116,
+      "step": 112500
+    },
+    {
+      "epoch": 0.45020996517845047,
+      "grad_norm": 1.569144606590271,
+      "learning_rate": 4.249650058035916e-05,
+      "loss": 1.1097,
+      "step": 113000
+    },
+    {
+      "epoch": 0.45220204467039055,
+      "grad_norm": 1.7324265241622925,
+      "learning_rate": 4.24632992554935e-05,
+      "loss": 1.1076,
+      "step": 113500
+    },
+    {
+      "epoch": 0.45419412416233057,
+      "grad_norm": 1.5229979753494263,
+      "learning_rate": 4.2430097930627826e-05,
+      "loss": 1.1065,
+      "step": 114000
+    },
+    {
+      "epoch": 0.4561862036542706,
+      "grad_norm": 1.59482741355896,
+      "learning_rate": 4.239689660576216e-05,
+      "loss": 1.1102,
+      "step": 114500
+    },
+    {
+      "epoch": 0.4581782831462107,
+      "grad_norm": 1.5474742650985718,
+      "learning_rate": 4.236369528089649e-05,
+      "loss": 1.0994,
+      "step": 115000
+    },
+    {
+      "epoch": 0.4601703626381507,
+      "grad_norm": 1.5477596521377563,
+      "learning_rate": 4.2330493956030826e-05,
+      "loss": 1.1073,
+      "step": 115500
+    },
+    {
+      "epoch": 0.4621624421300908,
+      "grad_norm": 1.5370910167694092,
+      "learning_rate": 4.229729263116516e-05,
+      "loss": 1.1048,
+      "step": 116000
+    },
+    {
+      "epoch": 0.4641545216220308,
+      "grad_norm": 1.4928206205368042,
+      "learning_rate": 4.226409130629948e-05,
+      "loss": 1.1086,
+      "step": 116500
+    },
+    {
+      "epoch": 0.4661466011139708,
+      "grad_norm": 1.4816399812698364,
+      "learning_rate": 4.223088998143382e-05,
+      "loss": 1.1046,
+      "step": 117000
+    },
+    {
+      "epoch": 0.4681386806059109,
+      "grad_norm": 1.5415599346160889,
+      "learning_rate": 4.2197688656568155e-05,
+      "loss": 1.1052,
+      "step": 117500
+    },
+    {
+      "epoch": 0.47013076009785093,
+      "grad_norm": 1.4426968097686768,
+      "learning_rate": 4.216448733170248e-05,
+      "loss": 1.1074,
+      "step": 118000
+    },
+    {
+      "epoch": 0.472122839589791,
+      "grad_norm": 1.6103301048278809,
+      "learning_rate": 4.213128600683682e-05,
+      "loss": 1.1023,
+      "step": 118500
+    },
+    {
+      "epoch": 0.47411491908173103,
+      "grad_norm": 1.628383755683899,
+      "learning_rate": 4.209808468197115e-05,
+      "loss": 1.1043,
+      "step": 119000
+    },
+    {
+      "epoch": 0.47610699857367106,
+      "grad_norm": 1.577172040939331,
+      "learning_rate": 4.2064883357105483e-05,
+      "loss": 1.1028,
+      "step": 119500
+    },
+    {
+      "epoch": 0.47809907806561114,
+      "grad_norm": 1.5424474477767944,
+      "learning_rate": 4.203168203223982e-05,
+      "loss": 1.1051,
+      "step": 120000
+    },
+    {
+      "epoch": 0.48009115755755116,
+      "grad_norm": 1.576210618019104,
+      "learning_rate": 4.199848070737415e-05,
+      "loss": 1.1013,
+      "step": 120500
+    },
+    {
+      "epoch": 0.48208323704949124,
+      "grad_norm": 1.5756324529647827,
+      "learning_rate": 4.1965279382508484e-05,
+      "loss": 1.1009,
+      "step": 121000
+    },
+    {
+      "epoch": 0.48407531654143127,
+      "grad_norm": 1.5778107643127441,
+      "learning_rate": 4.193207805764281e-05,
+      "loss": 1.1059,
+      "step": 121500
+    },
+    {
+      "epoch": 0.4860673960333713,
+      "grad_norm": 1.581160306930542,
+      "learning_rate": 4.189887673277715e-05,
+      "loss": 1.1008,
+      "step": 122000
+    },
+    {
+      "epoch": 0.48805947552531137,
+      "grad_norm": 1.5310883522033691,
+      "learning_rate": 4.1865675407911484e-05,
+      "loss": 1.1007,
+      "step": 122500
+    },
+    {
+      "epoch": 0.4900515550172514,
+      "grad_norm": 1.5881164073944092,
+      "learning_rate": 4.183247408304581e-05,
+      "loss": 1.1046,
+      "step": 123000
+    },
+    {
+      "epoch": 0.4920436345091915,
+      "grad_norm": 1.5284779071807861,
+      "learning_rate": 4.179927275818015e-05,
+      "loss": 1.1008,
+      "step": 123500
+    },
+    {
+      "epoch": 0.4940357140011315,
+      "grad_norm": 1.6251980066299438,
+      "learning_rate": 4.176607143331448e-05,
+      "loss": 1.1004,
+      "step": 124000
+    },
+    {
+      "epoch": 0.4960277934930715,
+      "grad_norm": 1.6742424964904785,
+      "learning_rate": 4.1732870108448806e-05,
+      "loss": 1.0973,
+      "step": 124500
+    },
+    {
+      "epoch": 0.4980198729850116,
+      "grad_norm": 1.5259268283843994,
+      "learning_rate": 4.169966878358314e-05,
+      "loss": 1.0988,
+      "step": 125000
+    },
+    {
+      "epoch": 0.5000119524769516,
+      "grad_norm": 1.6057409048080444,
+      "learning_rate": 4.166646745871747e-05,
+      "loss": 1.1004,
+      "step": 125500
+    },
+    {
+      "epoch": 0.5020040319688917,
+      "grad_norm": 1.8888304233551025,
+      "learning_rate": 4.1633266133851806e-05,
+      "loss": 1.0969,
+      "step": 126000
+    },
+    {
+      "epoch": 0.5039961114608318,
+      "grad_norm": 1.5599915981292725,
+      "learning_rate": 4.160006480898614e-05,
+      "loss": 1.0969,
+      "step": 126500
+    },
+    {
+      "epoch": 0.5059881909527718,
+      "grad_norm": 1.552174687385559,
+      "learning_rate": 4.156686348412047e-05,
+      "loss": 1.0956,
+      "step": 127000
+    },
+    {
+      "epoch": 0.5079802704447118,
+      "grad_norm": 1.6370817422866821,
+      "learning_rate": 4.1533662159254806e-05,
+      "loss": 1.0965,
+      "step": 127500
+    },
+    {
+      "epoch": 0.5099723499366519,
+      "grad_norm": 1.5194529294967651,
+      "learning_rate": 4.1500460834389135e-05,
+      "loss": 1.0943,
+      "step": 128000
+    },
+    {
+      "epoch": 0.5119644294285919,
+      "grad_norm": 1.601110577583313,
+      "learning_rate": 4.146725950952347e-05,
+      "loss": 1.0952,
+      "step": 128500
+    },
+    {
+      "epoch": 0.513956508920532,
+      "grad_norm": 1.6505762338638306,
+      "learning_rate": 4.1434058184657806e-05,
+      "loss": 1.0924,
+      "step": 129000
+    },
+    {
+      "epoch": 0.515948588412472,
+      "grad_norm": 1.7002733945846558,
+      "learning_rate": 4.1400856859792135e-05,
+      "loss": 1.0984,
+      "step": 129500
+    },
+    {
+      "epoch": 0.5179406679044121,
+      "grad_norm": 1.6136902570724487,
+      "learning_rate": 4.136765553492647e-05,
+      "loss": 1.0938,
+      "step": 130000
+    },
+    {
+      "epoch": 0.5199327473963521,
+      "grad_norm": 1.5523444414138794,
+      "learning_rate": 4.13344542100608e-05,
+      "loss": 1.0913,
+      "step": 130500
+    },
+    {
+      "epoch": 0.5219248268882921,
+      "grad_norm": 1.5890451669692993,
+      "learning_rate": 4.1301252885195135e-05,
+      "loss": 1.0935,
+      "step": 131000
+    },
+    {
+      "epoch": 0.5239169063802323,
+      "grad_norm": 1.60792076587677,
+      "learning_rate": 4.1268051560329464e-05,
+      "loss": 1.0906,
+      "step": 131500
+    },
+    {
+      "epoch": 0.5259089858721723,
+      "grad_norm": 1.6133267879486084,
+      "learning_rate": 4.123485023546379e-05,
+      "loss": 1.0951,
+      "step": 132000
+    },
+    {
+      "epoch": 0.5279010653641123,
+      "grad_norm": 1.528605341911316,
+      "learning_rate": 4.120164891059813e-05,
+      "loss": 1.0907,
+      "step": 132500
+    },
+    {
+      "epoch": 0.5298931448560523,
+      "grad_norm": 1.612090826034546,
+      "learning_rate": 4.1168447585732464e-05,
+      "loss": 1.0905,
+      "step": 133000
+    },
+    {
+      "epoch": 0.5318852243479923,
+      "grad_norm": 1.6218080520629883,
+      "learning_rate": 4.113524626086679e-05,
+      "loss": 1.09,
+      "step": 133500
+    },
+    {
+      "epoch": 0.5338773038399325,
+      "grad_norm": 1.6004669666290283,
+      "learning_rate": 4.110204493600113e-05,
+      "loss": 1.0894,
+      "step": 134000
+    },
+    {
+      "epoch": 0.5358693833318725,
+      "grad_norm": 1.555478572845459,
+      "learning_rate": 4.106884361113546e-05,
+      "loss": 1.0888,
+      "step": 134500
+    },
+    {
+      "epoch": 0.5378614628238125,
+      "grad_norm": 1.5299959182739258,
+      "learning_rate": 4.103564228626979e-05,
+      "loss": 1.096,
+      "step": 135000
+    },
+    {
+      "epoch": 0.5398535423157526,
+      "grad_norm": 1.5377782583236694,
+      "learning_rate": 4.100244096140413e-05,
+      "loss": 1.0895,
+      "step": 135500
+    },
+    {
+      "epoch": 0.5418456218076926,
+      "grad_norm": 1.5002425909042358,
+      "learning_rate": 4.096923963653846e-05,
+      "loss": 1.0899,
+      "step": 136000
+    },
+    {
+      "epoch": 0.5438377012996327,
+      "grad_norm": 1.5155582427978516,
+      "learning_rate": 4.093603831167279e-05,
+      "loss": 1.0906,
+      "step": 136500
+    },
+    {
+      "epoch": 0.5458297807915727,
+      "grad_norm": 1.553161382675171,
+      "learning_rate": 4.090283698680712e-05,
+      "loss": 1.0876,
+      "step": 137000
+    },
+    {
+      "epoch": 0.5478218602835128,
+      "grad_norm": 1.5937870740890503,
+      "learning_rate": 4.086963566194146e-05,
+      "loss": 1.0889,
+      "step": 137500
+    },
+    {
+      "epoch": 0.5498139397754528,
+      "grad_norm": 1.5624568462371826,
+      "learning_rate": 4.083643433707579e-05,
+      "loss": 1.0941,
+      "step": 138000
+    },
+    {
+      "epoch": 0.5518060192673928,
+      "grad_norm": 1.6195722818374634,
+      "learning_rate": 4.080323301221012e-05,
+      "loss": 1.0887,
+      "step": 138500
+    },
+    {
+      "epoch": 0.553798098759333,
+      "grad_norm": 1.5802372694015503,
+      "learning_rate": 4.077003168734445e-05,
+      "loss": 1.0861,
+      "step": 139000
+    },
+    {
+      "epoch": 0.555790178251273,
+      "grad_norm": 1.5672920942306519,
+      "learning_rate": 4.0736830362478786e-05,
+      "loss": 1.0857,
+      "step": 139500
+    },
+    {
+      "epoch": 0.557782257743213,
+      "grad_norm": 1.5582618713378906,
+      "learning_rate": 4.0703629037613115e-05,
+      "loss": 1.0865,
+      "step": 140000
+    },
+    {
+      "epoch": 0.559774337235153,
+      "grad_norm": 1.4463245868682861,
+      "learning_rate": 4.067042771274745e-05,
+      "loss": 1.0879,
+      "step": 140500
+    },
+    {
+      "epoch": 0.561766416727093,
+      "grad_norm": 1.4612869024276733,
+      "learning_rate": 4.063722638788178e-05,
+      "loss": 1.0837,
+      "step": 141000
+    },
+    {
+      "epoch": 0.5637584962190332,
+      "grad_norm": 1.6285316944122314,
+      "learning_rate": 4.0604025063016115e-05,
+      "loss": 1.0822,
+      "step": 141500
+    },
+    {
+      "epoch": 0.5657505757109732,
+      "grad_norm": 1.5200395584106445,
+      "learning_rate": 4.057082373815045e-05,
+      "loss": 1.0849,
+      "step": 142000
+    },
+    {
+      "epoch": 0.5677426552029132,
+      "grad_norm": 1.5283201932907104,
+      "learning_rate": 4.053762241328478e-05,
+      "loss": 1.082,
+      "step": 142500
+    },
+    {
+      "epoch": 0.5697347346948533,
+      "grad_norm": 1.5992571115493774,
+      "learning_rate": 4.0504421088419115e-05,
+      "loss": 1.0877,
+      "step": 143000
+    },
+    {
+      "epoch": 0.5717268141867933,
+      "grad_norm": 1.6001373529434204,
+      "learning_rate": 4.0471219763553444e-05,
+      "loss": 1.0788,
+      "step": 143500
+    },
+    {
+      "epoch": 0.5737188936787334,
+      "grad_norm": 1.6298253536224365,
+      "learning_rate": 4.043801843868778e-05,
+      "loss": 1.0882,
+      "step": 144000
+    },
+    {
+      "epoch": 0.5757109731706734,
+      "grad_norm": 1.633254051208496,
+      "learning_rate": 4.0404817113822115e-05,
+      "loss": 1.0839,
+      "step": 144500
+    },
+    {
+      "epoch": 0.5777030526626135,
+      "grad_norm": 1.5562070608139038,
+      "learning_rate": 4.0371615788956444e-05,
+      "loss": 1.0845,
+      "step": 145000
+    },
+    {
+      "epoch": 0.5796951321545535,
+      "grad_norm": 1.602396011352539,
+      "learning_rate": 4.033841446409078e-05,
+      "loss": 1.0805,
+      "step": 145500
+    },
+    {
+      "epoch": 0.5816872116464935,
+      "grad_norm": 1.5106642246246338,
+      "learning_rate": 4.0305213139225115e-05,
+      "loss": 1.0824,
+      "step": 146000
+    },
+    {
+      "epoch": 0.5836792911384336,
+      "grad_norm": 1.5491997003555298,
+      "learning_rate": 4.027201181435944e-05,
+      "loss": 1.0821,
+      "step": 146500
+    },
+    {
+      "epoch": 0.5856713706303737,
+      "grad_norm": 1.6149988174438477,
+      "learning_rate": 4.023881048949377e-05,
+      "loss": 1.0784,
+      "step": 147000
+    },
+    {
+      "epoch": 0.5876634501223137,
+      "grad_norm": 1.49197518825531,
+      "learning_rate": 4.02056091646281e-05,
+      "loss": 1.0804,
+      "step": 147500
+    },
+    {
+      "epoch": 0.5896555296142537,
+      "grad_norm": 1.6360849142074585,
+      "learning_rate": 4.017240783976244e-05,
+      "loss": 1.0798,
+      "step": 148000
+    },
+    {
+      "epoch": 0.5916476091061937,
+      "grad_norm": 1.596742033958435,
+      "learning_rate": 4.013920651489677e-05,
+      "loss": 1.0798,
+      "step": 148500
+    },
+    {
+      "epoch": 0.5936396885981339,
+      "grad_norm": 1.626608967781067,
+      "learning_rate": 4.01060051900311e-05,
+      "loss": 1.081,
+      "step": 149000
+    },
+    {
+      "epoch": 0.5956317680900739,
+      "grad_norm": 1.4785698652267456,
+      "learning_rate": 4.007280386516544e-05,
+      "loss": 1.0805,
+      "step": 149500
+    },
+    {
+      "epoch": 0.5976238475820139,
+      "grad_norm": 1.623596429824829,
+      "learning_rate": 4.003960254029977e-05,
+      "loss": 1.0791,
+      "step": 150000
+    },
+    {
+      "epoch": 0.599615927073954,
+      "grad_norm": 1.6670762300491333,
+      "learning_rate": 4.00064012154341e-05,
+      "loss": 1.0767,
+      "step": 150500
+    },
+    {
+      "epoch": 0.601608006565894,
+      "grad_norm": 1.6499557495117188,
+      "learning_rate": 3.997319989056844e-05,
+      "loss": 1.0764,
+      "step": 151000
+    },
+    {
+      "epoch": 0.6036000860578341,
+      "grad_norm": 1.5722723007202148,
+      "learning_rate": 3.9939998565702767e-05,
+      "loss": 1.0812,
+      "step": 151500
+    },
+    {
+      "epoch": 0.6055921655497741,
+      "grad_norm": 1.5595028400421143,
+      "learning_rate": 3.99067972408371e-05,
+      "loss": 1.0776,
+      "step": 152000
+    },
+    {
+      "epoch": 0.6075842450417142,
+      "grad_norm": 1.5210609436035156,
+      "learning_rate": 3.987359591597144e-05,
+      "loss": 1.0784,
+      "step": 152500
+    },
+    {
+      "epoch": 0.6095763245336542,
+      "grad_norm": 1.6691687107086182,
+      "learning_rate": 3.9840394591105767e-05,
+      "loss": 1.0829,
+      "step": 153000
+    },
+    {
+      "epoch": 0.6115684040255942,
+      "grad_norm": 1.574648141860962,
+      "learning_rate": 3.98071932662401e-05,
+      "loss": 1.0798,
+      "step": 153500
+    },
+    {
+      "epoch": 0.6135604835175343,
+      "grad_norm": 1.460086464881897,
+      "learning_rate": 3.977399194137443e-05,
+      "loss": 1.0797,
+      "step": 154000
+    },
+    {
+      "epoch": 0.6155525630094744,
+      "grad_norm": 1.6174976825714111,
+      "learning_rate": 3.974079061650876e-05,
+      "loss": 1.0724,
+      "step": 154500
+    },
+    {
+      "epoch": 0.6175446425014144,
+      "grad_norm": 1.5996955633163452,
+      "learning_rate": 3.9707589291643096e-05,
+      "loss": 1.0743,
+      "step": 155000
+    },
+    {
+      "epoch": 0.6195367219933544,
+      "grad_norm": 1.547351360321045,
+      "learning_rate": 3.9674387966777424e-05,
+      "loss": 1.0713,
+      "step": 155500
+    },
+    {
+      "epoch": 0.6215288014852944,
+      "grad_norm": 1.567685604095459,
+      "learning_rate": 3.964118664191176e-05,
+      "loss": 1.0715,
+      "step": 156000
+    },
+    {
+      "epoch": 0.6235208809772346,
+      "grad_norm": 1.5260729789733887,
+      "learning_rate": 3.9607985317046096e-05,
+      "loss": 1.0757,
+      "step": 156500
+    },
+    {
+      "epoch": 0.6255129604691746,
+      "grad_norm": 1.5618181228637695,
+      "learning_rate": 3.9574783992180424e-05,
+      "loss": 1.0783,
+      "step": 157000
+    },
+    {
+      "epoch": 0.6275050399611146,
+      "grad_norm": 1.4893778562545776,
+      "learning_rate": 3.954158266731476e-05,
+      "loss": 1.069,
+      "step": 157500
+    },
+    {
+      "epoch": 0.6294971194530546,
+      "grad_norm": 1.4940786361694336,
+      "learning_rate": 3.950838134244909e-05,
+      "loss": 1.071,
+      "step": 158000
+    },
+    {
+      "epoch": 0.6314891989449947,
+      "grad_norm": 1.5145844221115112,
+      "learning_rate": 3.9475180017583425e-05,
+      "loss": 1.0758,
+      "step": 158500
+    },
+    {
+      "epoch": 0.6334812784369348,
+      "grad_norm": 1.5686500072479248,
+      "learning_rate": 3.944197869271776e-05,
+      "loss": 1.0692,
+      "step": 159000
+    },
+    {
+      "epoch": 0.6354733579288748,
+      "grad_norm": 1.560360312461853,
+      "learning_rate": 3.940877736785209e-05,
+      "loss": 1.0717,
+      "step": 159500
+    },
+    {
+      "epoch": 0.6374654374208149,
+      "grad_norm": 1.5164998769760132,
+      "learning_rate": 3.9375576042986425e-05,
+      "loss": 1.0739,
+      "step": 160000
+    },
+    {
+      "epoch": 0.6394575169127549,
+      "grad_norm": 1.527099370956421,
+      "learning_rate": 3.9342374718120753e-05,
+      "loss": 1.0771,
+      "step": 160500
+    },
+    {
+      "epoch": 0.6414495964046949,
+      "grad_norm": 1.6750723123550415,
+      "learning_rate": 3.930917339325509e-05,
+      "loss": 1.0719,
+      "step": 161000
+    },
+    {
+      "epoch": 0.643441675896635,
+      "grad_norm": 1.5235224962234497,
+      "learning_rate": 3.927597206838942e-05,
+      "loss": 1.072,
+      "step": 161500
+    },
+    {
+      "epoch": 0.6454337553885751,
+      "grad_norm": 1.527896761894226,
+      "learning_rate": 3.924277074352375e-05,
+      "loss": 1.0701,
+      "step": 162000
+    },
+    {
+      "epoch": 0.6474258348805151,
+      "grad_norm": 1.6444692611694336,
+      "learning_rate": 3.920956941865808e-05,
+      "loss": 1.0725,
+      "step": 162500
+    },
+    {
+      "epoch": 0.6494179143724551,
+      "grad_norm": 1.4681565761566162,
+      "learning_rate": 3.917636809379242e-05,
+      "loss": 1.0691,
+      "step": 163000
+    },
+    {
+      "epoch": 0.6514099938643951,
+      "grad_norm": 1.7241281270980835,
+      "learning_rate": 3.914316676892675e-05,
+      "loss": 1.0717,
+      "step": 163500
+    },
+    {
+      "epoch": 0.6534020733563352,
+      "grad_norm": 1.576015591621399,
+      "learning_rate": 3.910996544406108e-05,
+      "loss": 1.0682,
+      "step": 164000
+    },
+    {
+      "epoch": 0.6553941528482753,
+      "grad_norm": 1.602088451385498,
+      "learning_rate": 3.907676411919541e-05,
+      "loss": 1.0703,
+      "step": 164500
+    },
+    {
+      "epoch": 0.6573862323402153,
+      "grad_norm": 1.5285007953643799,
+      "learning_rate": 3.904356279432975e-05,
+      "loss": 1.066,
+      "step": 165000
+    },
+    {
+      "epoch": 0.6593783118321553,
+      "grad_norm": 1.5032992362976074,
+      "learning_rate": 3.901036146946408e-05,
+      "loss": 1.0721,
+      "step": 165500
+    },
+    {
+      "epoch": 0.6613703913240954,
+      "grad_norm": 1.6137770414352417,
+      "learning_rate": 3.897716014459841e-05,
+      "loss": 1.0651,
+      "step": 166000
+    },
+    {
+      "epoch": 0.6633624708160354,
+      "grad_norm": 1.6281675100326538,
+      "learning_rate": 3.894395881973275e-05,
+      "loss": 1.0673,
+      "step": 166500
+    },
+    {
+      "epoch": 0.6653545503079755,
+      "grad_norm": 1.5388950109481812,
+      "learning_rate": 3.8910757494867076e-05,
+      "loss": 1.0725,
+      "step": 167000
+    },
+    {
+      "epoch": 0.6673466297999155,
+      "grad_norm": 1.5771666765213013,
+      "learning_rate": 3.887755617000141e-05,
+      "loss": 1.0664,
+      "step": 167500
+    },
+    {
+      "epoch": 0.6693387092918556,
+      "grad_norm": 1.6227260828018188,
+      "learning_rate": 3.884435484513575e-05,
+      "loss": 1.07,
+      "step": 168000
+    },
+    {
+      "epoch": 0.6713307887837956,
+      "grad_norm": 1.601846694946289,
+      "learning_rate": 3.8811153520270076e-05,
+      "loss": 1.0743,
+      "step": 168500
+    },
+    {
+      "epoch": 0.6733228682757356,
+      "grad_norm": 1.5994261503219604,
+      "learning_rate": 3.877795219540441e-05,
+      "loss": 1.0661,
+      "step": 169000
+    },
+    {
+      "epoch": 0.6753149477676758,
+      "grad_norm": 1.5916807651519775,
+      "learning_rate": 3.874475087053874e-05,
+      "loss": 1.0679,
+      "step": 169500
+    },
+    {
+      "epoch": 0.6773070272596158,
+      "grad_norm": 1.5930758714675903,
+      "learning_rate": 3.871154954567307e-05,
+      "loss": 1.0663,
+      "step": 170000
+    },
+    {
+      "epoch": 0.6792991067515558,
+      "grad_norm": 1.5338292121887207,
+      "learning_rate": 3.8678348220807405e-05,
+      "loss": 1.0699,
+      "step": 170500
+    },
+    {
+      "epoch": 0.6812911862434958,
+      "grad_norm": 1.5095643997192383,
+      "learning_rate": 3.8645146895941734e-05,
+      "loss": 1.0669,
+      "step": 171000
+    },
+    {
+      "epoch": 0.6832832657354359,
+      "grad_norm": 1.5679811239242554,
+      "learning_rate": 3.861194557107607e-05,
+      "loss": 1.0616,
+      "step": 171500
+    },
+    {
+      "epoch": 0.685275345227376,
+      "grad_norm": 1.5606422424316406,
+      "learning_rate": 3.8578744246210405e-05,
+      "loss": 1.0629,
+      "step": 172000
+    },
+    {
+      "epoch": 0.687267424719316,
+      "grad_norm": 1.6172137260437012,
+      "learning_rate": 3.8545542921344734e-05,
+      "loss": 1.0622,
+      "step": 172500
+    },
+    {
+      "epoch": 0.689259504211256,
+      "grad_norm": 1.5312018394470215,
+      "learning_rate": 3.851234159647907e-05,
+      "loss": 1.0614,
+      "step": 173000
+    },
+    {
+      "epoch": 0.6912515837031961,
+      "grad_norm": 1.6117522716522217,
+      "learning_rate": 3.84791402716134e-05,
+      "loss": 1.0651,
+      "step": 173500
+    },
+    {
+      "epoch": 0.6932436631951361,
+      "grad_norm": 1.588781476020813,
+      "learning_rate": 3.8445938946747734e-05,
+      "loss": 1.0612,
+      "step": 174000
+    },
+    {
+      "epoch": 0.6952357426870762,
+      "grad_norm": 1.4916971921920776,
+      "learning_rate": 3.841273762188207e-05,
+      "loss": 1.0588,
+      "step": 174500
+    },
+    {
+      "epoch": 0.6972278221790162,
+      "grad_norm": 1.645673155784607,
+      "learning_rate": 3.83795362970164e-05,
+      "loss": 1.0615,
+      "step": 175000
+    },
+    {
+      "epoch": 0.6992199016709563,
+      "grad_norm": 1.6432006359100342,
+      "learning_rate": 3.8346334972150734e-05,
+      "loss": 1.0605,
+      "step": 175500
+    },
+    {
+      "epoch": 0.7012119811628963,
+      "grad_norm": 1.5961941480636597,
+      "learning_rate": 3.831313364728506e-05,
+      "loss": 1.0553,
+      "step": 176000
+    },
+    {
+      "epoch": 0.7032040606548363,
+      "grad_norm": 1.6069000959396362,
+      "learning_rate": 3.82799323224194e-05,
+      "loss": 1.0637,
+      "step": 176500
+    },
+    {
+      "epoch": 0.7051961401467765,
+      "grad_norm": 1.5077065229415894,
+      "learning_rate": 3.824673099755373e-05,
+      "loss": 1.0641,
+      "step": 177000
+    },
+    {
+      "epoch": 0.7071882196387165,
+      "grad_norm": 1.5742119550704956,
+      "learning_rate": 3.8213529672688056e-05,
+      "loss": 1.0611,
+      "step": 177500
+    },
+    {
+      "epoch": 0.7091802991306565,
+      "grad_norm": 1.5928370952606201,
+      "learning_rate": 3.818032834782239e-05,
+      "loss": 1.0607,
+      "step": 178000
+    },
+    {
+      "epoch": 0.7111723786225965,
+      "grad_norm": 1.6845070123672485,
+      "learning_rate": 3.814712702295673e-05,
+      "loss": 1.0599,
+      "step": 178500
+    },
+    {
+      "epoch": 0.7131644581145365,
+      "grad_norm": 1.5199116468429565,
+      "learning_rate": 3.8113925698091056e-05,
+      "loss": 1.0587,
+      "step": 179000
+    },
+    {
+      "epoch": 0.7151565376064767,
+      "grad_norm": 1.5391650199890137,
+      "learning_rate": 3.808072437322539e-05,
+      "loss": 1.06,
+      "step": 179500
+    },
+    {
+      "epoch": 0.7171486170984167,
+      "grad_norm": 1.6307920217514038,
+      "learning_rate": 3.804752304835972e-05,
+      "loss": 1.0577,
+      "step": 180000
+    },
+    {
+      "epoch": 0.7191406965903567,
+      "grad_norm": 1.6464365720748901,
+      "learning_rate": 3.8014321723494056e-05,
+      "loss": 1.0598,
+      "step": 180500
+    },
+    {
+      "epoch": 0.7211327760822968,
+      "grad_norm": 1.5811011791229248,
+      "learning_rate": 3.798112039862839e-05,
+      "loss": 1.0608,
+      "step": 181000
+    },
+    {
+      "epoch": 0.7231248555742368,
+      "grad_norm": 1.6433868408203125,
+      "learning_rate": 3.794791907376272e-05,
+      "loss": 1.0567,
+      "step": 181500
+    },
+    {
+      "epoch": 0.7251169350661769,
+      "grad_norm": 1.604437232017517,
+      "learning_rate": 3.7914717748897056e-05,
+      "loss": 1.0582,
+      "step": 182000
+    },
+    {
+      "epoch": 0.7271090145581169,
+      "grad_norm": 1.4986939430236816,
+      "learning_rate": 3.7881516424031385e-05,
+      "loss": 1.059,
+      "step": 182500
+    },
+    {
+      "epoch": 0.729101094050057,
+      "grad_norm": 1.5265394449234009,
+      "learning_rate": 3.784831509916572e-05,
+      "loss": 1.0614,
+      "step": 183000
+    },
+    {
+      "epoch": 0.731093173541997,
+      "grad_norm": 1.5469063520431519,
+      "learning_rate": 3.7815113774300056e-05,
+      "loss": 1.0612,
+      "step": 183500
+    },
+    {
+      "epoch": 0.733085253033937,
+      "grad_norm": 1.613490104675293,
+      "learning_rate": 3.7781912449434385e-05,
+      "loss": 1.0593,
+      "step": 184000
+    },
+    {
+      "epoch": 0.7350773325258771,
+      "grad_norm": 1.559168815612793,
+      "learning_rate": 3.7748711124568714e-05,
+      "loss": 1.059,
+      "step": 184500
+    },
+    {
+      "epoch": 0.7370694120178172,
+      "grad_norm": 1.5357121229171753,
+      "learning_rate": 3.771550979970305e-05,
+      "loss": 1.0606,
+      "step": 185000
+    },
+    {
+      "epoch": 0.7390614915097572,
+      "grad_norm": 1.659907341003418,
+      "learning_rate": 3.768230847483738e-05,
+      "loss": 1.0544,
+      "step": 185500
+    },
+    {
+      "epoch": 0.7410535710016972,
+      "grad_norm": 1.5671064853668213,
+      "learning_rate": 3.7649107149971714e-05,
+      "loss": 1.0557,
+      "step": 186000
+    },
+    {
+      "epoch": 0.7430456504936372,
+      "grad_norm": 1.5508842468261719,
+      "learning_rate": 3.761590582510604e-05,
+      "loss": 1.0486,
+      "step": 186500
+    },
+    {
+      "epoch": 0.7450377299855774,
+      "grad_norm": 1.5046502351760864,
+      "learning_rate": 3.758270450024038e-05,
+      "loss": 1.0573,
+      "step": 187000
+    },
+    {
+      "epoch": 0.7470298094775174,
+      "grad_norm": 1.5921417474746704,
+      "learning_rate": 3.7549503175374714e-05,
+      "loss": 1.0598,
+      "step": 187500
+    },
+    {
+      "epoch": 0.7490218889694574,
+      "grad_norm": 1.5565036535263062,
+      "learning_rate": 3.751630185050904e-05,
+      "loss": 1.0568,
+      "step": 188000
+    },
+    {
+      "epoch": 0.7510139684613975,
+      "grad_norm": 1.4704923629760742,
+      "learning_rate": 3.748310052564338e-05,
+      "loss": 1.05,
+      "step": 188500
+    },
+    {
+      "epoch": 0.7530060479533375,
+      "grad_norm": 1.5152666568756104,
+      "learning_rate": 3.744989920077771e-05,
+      "loss": 1.0553,
+      "step": 189000
+    },
+    {
+      "epoch": 0.7549981274452776,
+      "grad_norm": 1.44629967212677,
+      "learning_rate": 3.741669787591204e-05,
+      "loss": 1.0558,
+      "step": 189500
+    },
+    {
+      "epoch": 0.7569902069372176,
+      "grad_norm": 1.5390712022781372,
+      "learning_rate": 3.738349655104638e-05,
+      "loss": 1.0519,
+      "step": 190000
+    },
+    {
+      "epoch": 0.7589822864291577,
+      "grad_norm": 1.575546145439148,
+      "learning_rate": 3.735029522618071e-05,
+      "loss": 1.0527,
+      "step": 190500
+    },
+    {
+      "epoch": 0.7609743659210977,
+      "grad_norm": 1.548821210861206,
+      "learning_rate": 3.731709390131504e-05,
+      "loss": 1.0568,
+      "step": 191000
+    },
+    {
+      "epoch": 0.7629664454130377,
+      "grad_norm": 1.494032382965088,
+      "learning_rate": 3.728389257644937e-05,
+      "loss": 1.0568,
+      "step": 191500
+    },
+    {
+      "epoch": 0.7649585249049778,
+      "grad_norm": 1.5374219417572021,
+      "learning_rate": 3.72506912515837e-05,
+      "loss": 1.0534,
+      "step": 192000
+    },
+    {
+      "epoch": 0.7669506043969179,
+      "grad_norm": 1.4390031099319458,
+      "learning_rate": 3.7217489926718037e-05,
+      "loss": 1.0523,
+      "step": 192500
+    },
+    {
+      "epoch": 0.7689426838888579,
+      "grad_norm": 1.5236526727676392,
+      "learning_rate": 3.7184288601852365e-05,
+      "loss": 1.0496,
+      "step": 193000
+    },
+    {
+      "epoch": 0.7709347633807979,
+      "grad_norm": 1.5415639877319336,
+      "learning_rate": 3.71510872769867e-05,
+      "loss": 1.054,
+      "step": 193500
+    },
+    {
+      "epoch": 0.7729268428727379,
+      "grad_norm": 1.5501717329025269,
+      "learning_rate": 3.711788595212104e-05,
+      "loss": 1.0513,
+      "step": 194000
+    },
+    {
+      "epoch": 0.7749189223646781,
+      "grad_norm": 1.505335807800293,
+      "learning_rate": 3.7084684627255365e-05,
+      "loss": 1.0514,
+      "step": 194500
+    },
+    {
+      "epoch": 0.7769110018566181,
+      "grad_norm": 1.5714764595031738,
+      "learning_rate": 3.70514833023897e-05,
+      "loss": 1.0569,
+      "step": 195000
+    },
+    {
+      "epoch": 0.7789030813485581,
+      "grad_norm": 1.623746395111084,
+      "learning_rate": 3.701828197752403e-05,
+      "loss": 1.0476,
+      "step": 195500
+    },
+    {
+      "epoch": 0.7808951608404981,
+      "grad_norm": 1.5616533756256104,
+      "learning_rate": 3.6985080652658366e-05,
+      "loss": 1.0472,
+      "step": 196000
+    },
+    {
+      "epoch": 0.7828872403324382,
+      "grad_norm": 1.6897337436676025,
+      "learning_rate": 3.69518793277927e-05,
+      "loss": 1.05,
+      "step": 196500
+    },
+    {
+      "epoch": 0.7848793198243783,
+      "grad_norm": 1.518666386604309,
+      "learning_rate": 3.691867800292703e-05,
+      "loss": 1.0531,
+      "step": 197000
+    },
+    {
+      "epoch": 0.7868713993163183,
+      "grad_norm": 1.5739809274673462,
+      "learning_rate": 3.6885476678061366e-05,
+      "loss": 1.0485,
+      "step": 197500
+    },
+    {
+      "epoch": 0.7888634788082584,
+      "grad_norm": 1.656569242477417,
+      "learning_rate": 3.6852275353195694e-05,
+      "loss": 1.0497,
+      "step": 198000
+    },
+    {
+      "epoch": 0.7908555583001984,
+      "grad_norm": 1.6614595651626587,
+      "learning_rate": 3.681907402833003e-05,
+      "loss": 1.05,
+      "step": 198500
+    },
+    {
+      "epoch": 0.7928476377921384,
+      "grad_norm": 1.518548607826233,
+      "learning_rate": 3.6785872703464366e-05,
+      "loss": 1.0493,
+      "step": 199000
+    },
+    {
+      "epoch": 0.7948397172840785,
+      "grad_norm": 1.5620031356811523,
+      "learning_rate": 3.675267137859869e-05,
+      "loss": 1.0464,
+      "step": 199500
+    },
+    {
+      "epoch": 0.7968317967760186,
+      "grad_norm": 1.646616816520691,
+      "learning_rate": 3.671947005373302e-05,
+      "loss": 1.046,
+      "step": 200000
+    },
+    {
+      "epoch": 0.7988238762679586,
+      "grad_norm": 1.6432793140411377,
+      "learning_rate": 3.668626872886736e-05,
+      "loss": 1.0519,
+      "step": 200500
+    },
+    {
+      "epoch": 0.8008159557598986,
+      "grad_norm": 1.6187670230865479,
+      "learning_rate": 3.665306740400169e-05,
+      "loss": 1.0493,
+      "step": 201000
+    },
+    {
+      "epoch": 0.8028080352518386,
+      "grad_norm": 1.6005345582962036,
+      "learning_rate": 3.6619866079136023e-05,
+      "loss": 1.0469,
+      "step": 201500
+    },
+    {
+      "epoch": 0.8048001147437788,
+      "grad_norm": 1.48025643825531,
+      "learning_rate": 3.658666475427035e-05,
+      "loss": 1.0467,
+      "step": 202000
+    },
+    {
+      "epoch": 0.8067921942357188,
+      "grad_norm": 1.5883903503417969,
+      "learning_rate": 3.655346342940469e-05,
+      "loss": 1.0455,
+      "step": 202500
+    },
+    {
+      "epoch": 0.8087842737276588,
+      "grad_norm": 1.576514720916748,
+      "learning_rate": 3.6520262104539024e-05,
+      "loss": 1.0468,
+      "step": 203000
+    },
+    {
+      "epoch": 0.8107763532195988,
+      "grad_norm": 1.4558312892913818,
+      "learning_rate": 3.648706077967335e-05,
+      "loss": 1.0444,
+      "step": 203500
+    },
+    {
+      "epoch": 0.8127684327115389,
+      "grad_norm": 1.4881688356399536,
+      "learning_rate": 3.645385945480769e-05,
+      "loss": 1.0487,
+      "step": 204000
+    },
+    {
+      "epoch": 0.814760512203479,
+      "grad_norm": 1.4842990636825562,
+      "learning_rate": 3.642065812994202e-05,
+      "loss": 1.0448,
+      "step": 204500
+    },
+    {
+      "epoch": 0.816752591695419,
+      "grad_norm": 1.562641978263855,
+      "learning_rate": 3.638745680507635e-05,
+      "loss": 1.0479,
+      "step": 205000
+    },
+    {
+      "epoch": 0.818744671187359,
+      "grad_norm": 1.5246717929840088,
+      "learning_rate": 3.635425548021069e-05,
+      "loss": 1.0478,
+      "step": 205500
+    },
+    {
+      "epoch": 0.8207367506792991,
+      "grad_norm": 1.5847394466400146,
+      "learning_rate": 3.632105415534502e-05,
+      "loss": 1.048,
+      "step": 206000
+    },
+    {
+      "epoch": 0.8227288301712391,
+      "grad_norm": 1.9202566146850586,
+      "learning_rate": 3.628785283047935e-05,
+      "loss": 1.0468,
+      "step": 206500
+    },
+    {
+      "epoch": 0.8247209096631792,
+      "grad_norm": 1.6928787231445312,
+      "learning_rate": 3.625465150561368e-05,
+      "loss": 1.0424,
+      "step": 207000
+    },
+    {
+      "epoch": 0.8267129891551193,
+      "grad_norm": 1.565021276473999,
+      "learning_rate": 3.622145018074801e-05,
+      "loss": 1.0435,
+      "step": 207500
+    },
+    {
+      "epoch": 0.8287050686470593,
+      "grad_norm": 1.5331549644470215,
+      "learning_rate": 3.6188248855882346e-05,
+      "loss": 1.0453,
+      "step": 208000
+    },
+    {
+      "epoch": 0.8306971481389993,
+      "grad_norm": 1.6887444257736206,
+      "learning_rate": 3.6155047531016675e-05,
+      "loss": 1.0417,
+      "step": 208500
+    },
+    {
+      "epoch": 0.8326892276309393,
+      "grad_norm": 1.5474079847335815,
+      "learning_rate": 3.612184620615101e-05,
+      "loss": 1.0413,
+      "step": 209000
+    },
+    {
+      "epoch": 0.8346813071228795,
+      "grad_norm": 1.5392274856567383,
+      "learning_rate": 3.6088644881285346e-05,
+      "loss": 1.0447,
+      "step": 209500
+    },
+    {
+      "epoch": 0.8366733866148195,
+      "grad_norm": 1.5742685794830322,
+      "learning_rate": 3.6055443556419675e-05,
+      "loss": 1.0482,
+      "step": 210000
+    },
+    {
+      "epoch": 0.8386654661067595,
+      "grad_norm": 1.6905549764633179,
+      "learning_rate": 3.602224223155401e-05,
+      "loss": 1.0441,
+      "step": 210500
+    },
+    {
+      "epoch": 0.8406575455986995,
+      "grad_norm": 2.097541093826294,
+      "learning_rate": 3.5989040906688346e-05,
+      "loss": 1.0434,
+      "step": 211000
+    },
+    {
+      "epoch": 0.8426496250906396,
+      "grad_norm": 1.5824739933013916,
+      "learning_rate": 3.5955839581822675e-05,
+      "loss": 1.0415,
+      "step": 211500
+    },
+    {
+      "epoch": 0.8446417045825797,
+      "grad_norm": 1.6093156337738037,
+      "learning_rate": 3.592263825695701e-05,
+      "loss": 1.0398,
+      "step": 212000
+    },
+    {
+      "epoch": 0.8466337840745197,
+      "grad_norm": 1.5098934173583984,
+      "learning_rate": 3.588943693209134e-05,
+      "loss": 1.0426,
+      "step": 212500
+    },
+    {
+      "epoch": 0.8486258635664597,
+      "grad_norm": 1.5104949474334717,
+      "learning_rate": 3.5856235607225675e-05,
+      "loss": 1.0436,
+      "step": 213000
+    },
+    {
+      "epoch": 0.8506179430583998,
+      "grad_norm": 1.578851342201233,
+      "learning_rate": 3.582303428236001e-05,
+      "loss": 1.0426,
+      "step": 213500
+    },
+    {
+      "epoch": 0.8526100225503398,
+      "grad_norm": 1.5221812725067139,
+      "learning_rate": 3.578983295749434e-05,
+      "loss": 1.0369,
+      "step": 214000
+    },
+    {
+      "epoch": 0.8546021020422799,
+      "grad_norm": 1.5251935720443726,
+      "learning_rate": 3.575663163262867e-05,
+      "loss": 1.0444,
+      "step": 214500
+    },
+    {
+      "epoch": 0.85659418153422,
+      "grad_norm": 1.626284122467041,
+      "learning_rate": 3.5723430307763004e-05,
+      "loss": 1.0469,
+      "step": 215000
+    },
+    {
+      "epoch": 0.85858626102616,
+      "grad_norm": 1.607362151145935,
+      "learning_rate": 3.569022898289733e-05,
+      "loss": 1.0414,
+      "step": 215500
+    },
+    {
+      "epoch": 0.8605783405181,
+      "grad_norm": 1.4563950300216675,
+      "learning_rate": 3.565702765803167e-05,
+      "loss": 1.0414,
+      "step": 216000
+    },
+    {
+      "epoch": 0.86257042001004,
+      "grad_norm": 1.5460984706878662,
+      "learning_rate": 3.5623826333166e-05,
+      "loss": 1.0401,
+      "step": 216500
+    },
+    {
+      "epoch": 0.8645624995019802,
+      "grad_norm": 1.5916367769241333,
+      "learning_rate": 3.559062500830033e-05,
+      "loss": 1.0412,
+      "step": 217000
+    },
+    {
+      "epoch": 0.8665545789939202,
+      "grad_norm": 1.640368938446045,
+      "learning_rate": 3.555742368343467e-05,
+      "loss": 1.0385,
+      "step": 217500
+    },
+    {
+      "epoch": 0.8685466584858602,
+      "grad_norm": 1.5175517797470093,
+      "learning_rate": 3.5524222358569e-05,
+      "loss": 1.0352,
+      "step": 218000
+    },
+    {
+      "epoch": 0.8705387379778002,
+      "grad_norm": 1.5440324544906616,
+      "learning_rate": 3.549102103370333e-05,
+      "loss": 1.0372,
+      "step": 218500
+    },
+    {
+      "epoch": 0.8725308174697403,
+      "grad_norm": 1.5476912260055542,
+      "learning_rate": 3.545781970883766e-05,
+      "loss": 1.0414,
+      "step": 219000
+    },
+    {
+      "epoch": 0.8745228969616804,
+      "grad_norm": 1.5171074867248535,
+      "learning_rate": 3.5424618383972e-05,
+      "loss": 1.0403,
+      "step": 219500
+    },
+    {
+      "epoch": 0.8765149764536204,
+      "grad_norm": 1.551775336265564,
+      "learning_rate": 3.539141705910633e-05,
+      "loss": 1.0399,
+      "step": 220000
+    },
+    {
+      "epoch": 0.8785070559455604,
+      "grad_norm": 1.6885789632797241,
+      "learning_rate": 3.535821573424066e-05,
+      "loss": 1.0376,
+      "step": 220500
+    },
+    {
+      "epoch": 0.8804991354375005,
+      "grad_norm": 1.513099193572998,
+      "learning_rate": 3.5325014409375e-05,
+      "loss": 1.0341,
+      "step": 221000
+    },
+    {
+      "epoch": 0.8824912149294405,
+      "grad_norm": 1.6219054460525513,
+      "learning_rate": 3.5291813084509326e-05,
+      "loss": 1.0378,
+      "step": 221500
+    },
+    {
+      "epoch": 0.8844832944213806,
+      "grad_norm": 1.571273684501648,
+      "learning_rate": 3.5258611759643655e-05,
+      "loss": 1.0378,
+      "step": 222000
+    },
+    {
+      "epoch": 0.8864753739133207,
+      "grad_norm": 1.6003777980804443,
+      "learning_rate": 3.522541043477799e-05,
+      "loss": 1.0392,
+      "step": 222500
+    },
+    {
+      "epoch": 0.8884674534052607,
+      "grad_norm": 1.5567518472671509,
+      "learning_rate": 3.519220910991232e-05,
+      "loss": 1.0373,
+      "step": 223000
+    },
+    {
+      "epoch": 0.8904595328972007,
+      "grad_norm": 1.5490124225616455,
+      "learning_rate": 3.5159007785046655e-05,
+      "loss": 1.0403,
+      "step": 223500
+    },
+    {
+      "epoch": 0.8924516123891407,
+      "grad_norm": 1.5508618354797363,
+      "learning_rate": 3.512580646018099e-05,
+      "loss": 1.0378,
+      "step": 224000
+    },
+    {
+      "epoch": 0.8944436918810809,
+      "grad_norm": 1.6086506843566895,
+      "learning_rate": 3.509260513531532e-05,
+      "loss": 1.0329,
+      "step": 224500
+    },
+    {
+      "epoch": 0.8964357713730209,
+      "grad_norm": 1.516716718673706,
+      "learning_rate": 3.5059403810449655e-05,
+      "loss": 1.0345,
+      "step": 225000
+    },
+    {
+      "epoch": 0.8984278508649609,
+      "grad_norm": 1.5782248973846436,
+      "learning_rate": 3.5026202485583984e-05,
+      "loss": 1.0377,
+      "step": 225500
+    },
+    {
+      "epoch": 0.9004199303569009,
+      "grad_norm": 1.6613645553588867,
+      "learning_rate": 3.499300116071832e-05,
+      "loss": 1.0324,
+      "step": 226000
+    },
+    {
+      "epoch": 0.902412009848841,
+      "grad_norm": 1.5320236682891846,
+      "learning_rate": 3.4959799835852655e-05,
+      "loss": 1.0364,
+      "step": 226500
+    },
+    {
+      "epoch": 0.9044040893407811,
+      "grad_norm": 1.6433411836624146,
+      "learning_rate": 3.4926598510986984e-05,
+      "loss": 1.0363,
+      "step": 227000
+    },
+    {
+      "epoch": 0.9063961688327211,
+      "grad_norm": 1.4934873580932617,
+      "learning_rate": 3.489339718612132e-05,
+      "loss": 1.0349,
+      "step": 227500
+    },
+    {
+      "epoch": 0.9083882483246611,
+      "grad_norm": 1.5197460651397705,
+      "learning_rate": 3.486019586125565e-05,
+      "loss": 1.0314,
+      "step": 228000
+    },
+    {
+      "epoch": 0.9103803278166012,
+      "grad_norm": 1.5315827131271362,
+      "learning_rate": 3.4826994536389984e-05,
+      "loss": 1.0374,
+      "step": 228500
+    },
+    {
+      "epoch": 0.9123724073085412,
+      "grad_norm": 1.6260908842086792,
+      "learning_rate": 3.479379321152432e-05,
+      "loss": 1.0402,
+      "step": 229000
+    },
+    {
+      "epoch": 0.9143644868004813,
+      "grad_norm": 1.6497081518173218,
+      "learning_rate": 3.476059188665864e-05,
+      "loss": 1.0349,
+      "step": 229500
+    },
+    {
+      "epoch": 0.9163565662924213,
+      "grad_norm": 1.5770347118377686,
+      "learning_rate": 3.472739056179298e-05,
+      "loss": 1.038,
+      "step": 230000
+    },
+    {
+      "epoch": 0.9183486457843614,
+      "grad_norm": 1.6462286710739136,
+      "learning_rate": 3.469418923692731e-05,
+      "loss": 1.028,
+      "step": 230500
+    },
+    {
+      "epoch": 0.9203407252763014,
+      "grad_norm": 1.5420483350753784,
+      "learning_rate": 3.466098791206164e-05,
+      "loss": 1.0352,
+      "step": 231000
+    },
+    {
+      "epoch": 0.9223328047682414,
+      "grad_norm": 1.5735697746276855,
+      "learning_rate": 3.462778658719598e-05,
+      "loss": 1.034,
+      "step": 231500
+    },
+    {
+      "epoch": 0.9243248842601816,
+      "grad_norm": 1.6684339046478271,
+      "learning_rate": 3.4594585262330306e-05,
+      "loss": 1.0331,
+      "step": 232000
+    },
+    {
+      "epoch": 0.9263169637521216,
+      "grad_norm": 1.6409918069839478,
+      "learning_rate": 3.456138393746464e-05,
+      "loss": 1.036,
+      "step": 232500
+    },
+    {
+      "epoch": 0.9283090432440616,
+      "grad_norm": 1.4565562009811401,
+      "learning_rate": 3.452818261259898e-05,
+      "loss": 1.0313,
+      "step": 233000
+    },
+    {
+      "epoch": 0.9303011227360016,
+      "grad_norm": 1.4914257526397705,
+      "learning_rate": 3.4494981287733307e-05,
+      "loss": 1.0351,
+      "step": 233500
+    },
+    {
+      "epoch": 0.9322932022279417,
+      "grad_norm": 1.5664403438568115,
+      "learning_rate": 3.446177996286764e-05,
+      "loss": 1.0327,
+      "step": 234000
+    },
+    {
+      "epoch": 0.9342852817198818,
+      "grad_norm": 1.6254390478134155,
+      "learning_rate": 3.442857863800197e-05,
+      "loss": 1.0327,
+      "step": 234500
+    },
+    {
+      "epoch": 0.9362773612118218,
+      "grad_norm": 1.4953609704971313,
+      "learning_rate": 3.439537731313631e-05,
+      "loss": 1.0288,
+      "step": 235000
+    },
+    {
+      "epoch": 0.9382694407037618,
+      "grad_norm": 1.5648387670516968,
+      "learning_rate": 3.436217598827064e-05,
+      "loss": 1.0332,
+      "step": 235500
+    },
+    {
+      "epoch": 0.9402615201957019,
+      "grad_norm": 1.5644855499267578,
+      "learning_rate": 3.432897466340497e-05,
+      "loss": 1.0344,
+      "step": 236000
+    },
+    {
+      "epoch": 0.9422535996876419,
+      "grad_norm": 1.6307463645935059,
+      "learning_rate": 3.429577333853931e-05,
+      "loss": 1.0306,
+      "step": 236500
+    },
+    {
+      "epoch": 0.944245679179582,
+      "grad_norm": 1.5409446954727173,
+      "learning_rate": 3.4262572013673636e-05,
+      "loss": 1.0332,
+      "step": 237000
+    },
+    {
+      "epoch": 0.946237758671522,
+      "grad_norm": 1.6127350330352783,
+      "learning_rate": 3.4229370688807964e-05,
+      "loss": 1.0309,
+      "step": 237500
+    },
+    {
+      "epoch": 0.9482298381634621,
+      "grad_norm": 1.5565184354782104,
+      "learning_rate": 3.41961693639423e-05,
+      "loss": 1.0286,
+      "step": 238000
+    },
+    {
+      "epoch": 0.9502219176554021,
+      "grad_norm": 1.4683204889297485,
+      "learning_rate": 3.416296803907663e-05,
+      "loss": 1.0305,
+      "step": 238500
+    },
+    {
+      "epoch": 0.9522139971473421,
+      "grad_norm": 1.5222786664962769,
+      "learning_rate": 3.4129766714210964e-05,
+      "loss": 1.03,
+      "step": 239000
+    },
+    {
+      "epoch": 0.9542060766392823,
+      "grad_norm": 1.5132941007614136,
+      "learning_rate": 3.40965653893453e-05,
+      "loss": 1.0324,
+      "step": 239500
+    },
+    {
+      "epoch": 0.9561981561312223,
+      "grad_norm": 1.5569360256195068,
+      "learning_rate": 3.406336406447963e-05,
+      "loss": 1.0297,
+      "step": 240000
+    },
+    {
+      "epoch": 0.9581902356231623,
+      "grad_norm": 1.5893915891647339,
+      "learning_rate": 3.4030162739613965e-05,
+      "loss": 1.0328,
+      "step": 240500
+    },
+    {
+      "epoch": 0.9601823151151023,
+      "grad_norm": 1.6012451648712158,
+      "learning_rate": 3.3996961414748293e-05,
+      "loss": 1.029,
+      "step": 241000
+    },
+    {
+      "epoch": 0.9621743946070423,
+      "grad_norm": 2.4129886627197266,
+      "learning_rate": 3.396376008988263e-05,
+      "loss": 1.0285,
+      "step": 241500
+    },
+    {
+      "epoch": 0.9641664740989825,
+      "grad_norm": 1.6554327011108398,
+      "learning_rate": 3.3930558765016965e-05,
+      "loss": 1.0287,
+      "step": 242000
+    },
+    {
+      "epoch": 0.9661585535909225,
+      "grad_norm": 1.4973152875900269,
+      "learning_rate": 3.3897357440151293e-05,
+      "loss": 1.0305,
+      "step": 242500
+    },
+    {
+      "epoch": 0.9681506330828625,
+      "grad_norm": 1.467353343963623,
+      "learning_rate": 3.386415611528563e-05,
+      "loss": 1.0302,
+      "step": 243000
+    },
+    {
+      "epoch": 0.9701427125748026,
+      "grad_norm": 1.482872724533081,
+      "learning_rate": 3.383095479041996e-05,
+      "loss": 1.0267,
+      "step": 243500
+    },
+    {
+      "epoch": 0.9721347920667426,
+      "grad_norm": 1.4883670806884766,
+      "learning_rate": 3.3797753465554294e-05,
+      "loss": 1.0277,
+      "step": 244000
+    },
+    {
+      "epoch": 0.9741268715586827,
+      "grad_norm": 1.637110948562622,
+      "learning_rate": 3.376455214068862e-05,
+      "loss": 1.0278,
+      "step": 244500
+    },
+    {
+      "epoch": 0.9761189510506227,
+      "grad_norm": 1.5941038131713867,
+      "learning_rate": 3.373135081582295e-05,
+      "loss": 1.0291,
+      "step": 245000
+    },
+    {
+      "epoch": 0.9781110305425628,
+      "grad_norm": 1.5783106088638306,
+      "learning_rate": 3.369814949095729e-05,
+      "loss": 1.0234,
+      "step": 245500
+    },
+    {
+      "epoch": 0.9801031100345028,
+      "grad_norm": 1.5791685581207275,
+      "learning_rate": 3.366494816609162e-05,
+      "loss": 1.0285,
+      "step": 246000
+    },
+    {
+      "epoch": 0.9820951895264428,
+      "grad_norm": 1.4624321460723877,
+      "learning_rate": 3.363174684122595e-05,
+      "loss": 1.0266,
+      "step": 246500
+    },
+    {
+      "epoch": 0.984087269018383,
+      "grad_norm": 1.5991438627243042,
+      "learning_rate": 3.359854551636029e-05,
+      "loss": 1.026,
+      "step": 247000
+    },
+    {
+      "epoch": 0.986079348510323,
+      "grad_norm": 1.692667007446289,
+      "learning_rate": 3.3565344191494616e-05,
+      "loss": 1.0279,
+      "step": 247500
+    },
+    {
+      "epoch": 0.988071428002263,
+      "grad_norm": 1.5520540475845337,
+      "learning_rate": 3.353214286662895e-05,
+      "loss": 1.0233,
+      "step": 248000
+    },
+    {
+      "epoch": 0.990063507494203,
+      "grad_norm": 1.617071509361267,
+      "learning_rate": 3.349894154176329e-05,
+      "loss": 1.0265,
+      "step": 248500
+    },
+    {
+      "epoch": 0.992055586986143,
+      "grad_norm": 1.6249128580093384,
+      "learning_rate": 3.3465740216897616e-05,
+      "loss": 1.0224,
+      "step": 249000
+    },
+    {
+      "epoch": 0.9940476664780832,
+      "grad_norm": 1.5957164764404297,
+      "learning_rate": 3.343253889203195e-05,
+      "loss": 1.0215,
+      "step": 249500
+    },
+    {
+      "epoch": 0.9960397459700232,
+      "grad_norm": 1.5286188125610352,
+      "learning_rate": 3.339933756716628e-05,
+      "loss": 1.0222,
+      "step": 250000
+    },
+    {
+      "epoch": 0.9980318254619632,
+      "grad_norm": 1.6094142198562622,
+      "learning_rate": 3.3366136242300616e-05,
+      "loss": 1.0245,
+      "step": 250500
+    },
+    {
+      "epoch": 1.0000239049539033,
+      "grad_norm": 1.4799585342407227,
+      "learning_rate": 3.333293491743495e-05,
+      "loss": 1.0224,
+      "step": 251000
+    },
+    {
+      "epoch": 1.0020159844458434,
+      "grad_norm": 1.60854172706604,
+      "learning_rate": 3.329973359256928e-05,
+      "loss": 1.0227,
+      "step": 251500
+    },
+    {
+      "epoch": 1.0040080639377833,
+      "grad_norm": 1.5131163597106934,
+      "learning_rate": 3.326653226770361e-05,
+      "loss": 1.0195,
+      "step": 252000
+    },
+    {
+      "epoch": 1.0060001434297234,
+      "grad_norm": 1.4940989017486572,
+      "learning_rate": 3.3233330942837945e-05,
+      "loss": 1.0212,
+      "step": 252500
+    },
+    {
+      "epoch": 1.0079922229216636,
+      "grad_norm": 1.641563892364502,
+      "learning_rate": 3.3200129617972274e-05,
+      "loss": 1.0242,
+      "step": 253000
+    },
+    {
+      "epoch": 1.0099843024136035,
+      "grad_norm": 1.5117000341415405,
+      "learning_rate": 3.316692829310661e-05,
+      "loss": 1.0205,
+      "step": 253500
+    },
+    {
+      "epoch": 1.0119763819055436,
+      "grad_norm": 1.6550041437149048,
+      "learning_rate": 3.313372696824094e-05,
+      "loss": 1.0229,
+      "step": 254000
+    },
+    {
+      "epoch": 1.0139684613974835,
+      "grad_norm": 1.5466336011886597,
+      "learning_rate": 3.3100525643375274e-05,
+      "loss": 1.0208,
+      "step": 254500
+    },
+    {
+      "epoch": 1.0159605408894237,
+      "grad_norm": 1.6535435914993286,
+      "learning_rate": 3.306732431850961e-05,
+      "loss": 1.0214,
+      "step": 255000
+    },
+    {
+      "epoch": 1.0179526203813638,
+      "grad_norm": 1.5536569356918335,
+      "learning_rate": 3.303412299364394e-05,
+      "loss": 1.0232,
+      "step": 255500
+    },
+    {
+      "epoch": 1.0199446998733037,
+      "grad_norm": 1.4865309000015259,
+      "learning_rate": 3.3000921668778274e-05,
+      "loss": 1.0191,
+      "step": 256000
+    },
+    {
+      "epoch": 1.0219367793652439,
+      "grad_norm": 1.5492526292800903,
+      "learning_rate": 3.29677203439126e-05,
+      "loss": 1.0254,
+      "step": 256500
+    },
+    {
+      "epoch": 1.0239288588571838,
+      "grad_norm": 1.5980515480041504,
+      "learning_rate": 3.293451901904694e-05,
+      "loss": 1.02,
+      "step": 257000
+    },
+    {
+      "epoch": 1.025920938349124,
+      "grad_norm": 1.5227771997451782,
+      "learning_rate": 3.2901317694181274e-05,
+      "loss": 1.0216,
+      "step": 257500
+    },
+    {
+      "epoch": 1.027913017841064,
+      "grad_norm": 1.5084716081619263,
+      "learning_rate": 3.28681163693156e-05,
+      "loss": 1.0226,
+      "step": 258000
+    },
+    {
+      "epoch": 1.029905097333004,
+      "grad_norm": 1.5168397426605225,
+      "learning_rate": 3.283491504444994e-05,
+      "loss": 1.0177,
+      "step": 258500
+    },
+    {
+      "epoch": 1.031897176824944,
+      "grad_norm": 1.4984822273254395,
+      "learning_rate": 3.280171371958427e-05,
+      "loss": 1.0205,
+      "step": 259000
+    },
+    {
+      "epoch": 1.033889256316884,
+      "grad_norm": 1.500741958618164,
+      "learning_rate": 3.2768512394718596e-05,
+      "loss": 1.0222,
+      "step": 259500
+    },
+    {
+      "epoch": 1.0358813358088241,
+      "grad_norm": 1.562975525856018,
+      "learning_rate": 3.273531106985293e-05,
+      "loss": 1.0175,
+      "step": 260000
+    },
+    {
+      "epoch": 1.0378734153007643,
+      "grad_norm": 1.649877905845642,
+      "learning_rate": 3.270210974498726e-05,
+      "loss": 1.019,
+      "step": 260500
+    },
+    {
+      "epoch": 1.0398654947927042,
+      "grad_norm": 1.5648247003555298,
+      "learning_rate": 3.2668908420121596e-05,
+      "loss": 1.0244,
+      "step": 261000
+    },
+    {
+      "epoch": 1.0418575742846443,
+      "grad_norm": 1.6014691591262817,
+      "learning_rate": 3.263570709525593e-05,
+      "loss": 1.0221,
+      "step": 261500
+    },
+    {
+      "epoch": 1.0438496537765842,
+      "grad_norm": 1.706006646156311,
+      "learning_rate": 3.260250577039026e-05,
+      "loss": 1.0193,
+      "step": 262000
+    },
+    {
+      "epoch": 1.0458417332685244,
+      "grad_norm": 1.6342687606811523,
+      "learning_rate": 3.2569304445524596e-05,
+      "loss": 1.0154,
+      "step": 262500
+    },
+    {
+      "epoch": 1.0478338127604645,
+      "grad_norm": 1.567576289176941,
+      "learning_rate": 3.2536103120658925e-05,
+      "loss": 1.0204,
+      "step": 263000
+    },
+    {
+      "epoch": 1.0498258922524044,
+      "grad_norm": 1.586421012878418,
+      "learning_rate": 3.250290179579326e-05,
+      "loss": 1.0213,
+      "step": 263500
+    },
+    {
+      "epoch": 1.0518179717443445,
+      "grad_norm": 1.6086459159851074,
+      "learning_rate": 3.2469700470927596e-05,
+      "loss": 1.0207,
+      "step": 264000
+    },
+    {
+      "epoch": 1.0538100512362845,
+      "grad_norm": 1.5718351602554321,
+      "learning_rate": 3.2436499146061925e-05,
+      "loss": 1.0265,
+      "step": 264500
+    },
+    {
+      "epoch": 1.0558021307282246,
+      "grad_norm": 1.6731963157653809,
+      "learning_rate": 3.240329782119626e-05,
+      "loss": 1.0173,
+      "step": 265000
+    },
+    {
+      "epoch": 1.0577942102201647,
+      "grad_norm": 1.597184181213379,
+      "learning_rate": 3.2370096496330596e-05,
+      "loss": 1.0141,
+      "step": 265500
+    },
+    {
+      "epoch": 1.0597862897121046,
+      "grad_norm": 1.5035780668258667,
+      "learning_rate": 3.2336895171464925e-05,
+      "loss": 1.0179,
+      "step": 266000
+    },
+    {
+      "epoch": 1.0617783692040448,
+      "grad_norm": 1.5219330787658691,
+      "learning_rate": 3.230369384659926e-05,
+      "loss": 1.019,
+      "step": 266500
+    },
+    {
+      "epoch": 1.0637704486959847,
+      "grad_norm": 1.5741368532180786,
+      "learning_rate": 3.227049252173358e-05,
+      "loss": 1.0191,
+      "step": 267000
+    },
+    {
+      "epoch": 1.0657625281879248,
+      "grad_norm": 1.5799455642700195,
+      "learning_rate": 3.223729119686792e-05,
+      "loss": 1.0149,
+      "step": 267500
+    },
+    {
+      "epoch": 1.067754607679865,
+      "grad_norm": 1.6045303344726562,
+      "learning_rate": 3.2204089872002254e-05,
+      "loss": 1.0212,
+      "step": 268000
+    },
+    {
+      "epoch": 1.0697466871718049,
+      "grad_norm": 1.5570663213729858,
+      "learning_rate": 3.217088854713658e-05,
+      "loss": 1.0122,
+      "step": 268500
+    },
+    {
+      "epoch": 1.071738766663745,
+      "grad_norm": 1.6131365299224854,
+      "learning_rate": 3.213768722227092e-05,
+      "loss": 1.0143,
+      "step": 269000
+    },
+    {
+      "epoch": 1.073730846155685,
+      "grad_norm": 1.6053640842437744,
+      "learning_rate": 3.2104485897405254e-05,
+      "loss": 1.0167,
+      "step": 269500
+    },
+    {
+      "epoch": 1.075722925647625,
+      "grad_norm": 1.6846861839294434,
+      "learning_rate": 3.207128457253958e-05,
+      "loss": 1.0219,
+      "step": 270000
+    },
+    {
+      "epoch": 1.0777150051395652,
+      "grad_norm": 1.5112779140472412,
+      "learning_rate": 3.203808324767392e-05,
+      "loss": 1.0178,
+      "step": 270500
+    },
+    {
+      "epoch": 1.079707084631505,
+      "grad_norm": 1.5776927471160889,
+      "learning_rate": 3.200488192280825e-05,
+      "loss": 1.0158,
+      "step": 271000
+    },
+    {
+      "epoch": 1.0816991641234452,
+      "grad_norm": 1.5890170335769653,
+      "learning_rate": 3.197168059794258e-05,
+      "loss": 1.0199,
+      "step": 271500
+    },
+    {
+      "epoch": 1.0836912436153852,
+      "grad_norm": 1.6621835231781006,
+      "learning_rate": 3.193847927307692e-05,
+      "loss": 1.0153,
+      "step": 272000
+    },
+    {
+      "epoch": 1.0856833231073253,
+      "grad_norm": 1.5698069334030151,
+      "learning_rate": 3.190527794821125e-05,
+      "loss": 1.0145,
+      "step": 272500
+    },
+    {
+      "epoch": 1.0876754025992654,
+      "grad_norm": 1.5621885061264038,
+      "learning_rate": 3.187207662334558e-05,
+      "loss": 1.0129,
+      "step": 273000
+    },
+    {
+      "epoch": 1.0896674820912053,
+      "grad_norm": 1.5177123546600342,
+      "learning_rate": 3.183887529847991e-05,
+      "loss": 1.0157,
+      "step": 273500
+    },
+    {
+      "epoch": 1.0916595615831455,
+      "grad_norm": 1.5849789381027222,
+      "learning_rate": 3.180567397361425e-05,
+      "loss": 1.0195,
+      "step": 274000
+    },
+    {
+      "epoch": 1.0936516410750854,
+      "grad_norm": 1.5286483764648438,
+      "learning_rate": 3.1772472648748577e-05,
+      "loss": 1.0161,
+      "step": 274500
+    },
+    {
+      "epoch": 1.0956437205670255,
+      "grad_norm": 1.6150707006454468,
+      "learning_rate": 3.1739271323882905e-05,
+      "loss": 1.0131,
+      "step": 275000
+    },
+    {
+      "epoch": 1.0976358000589657,
+      "grad_norm": 1.701650857925415,
+      "learning_rate": 3.170606999901724e-05,
+      "loss": 1.0133,
+      "step": 275500
+    },
+    {
+      "epoch": 1.0996278795509056,
+      "grad_norm": 1.6005955934524536,
+      "learning_rate": 3.167286867415158e-05,
+      "loss": 1.013,
+      "step": 276000
+    },
+    {
+      "epoch": 1.1016199590428457,
+      "grad_norm": 1.658247709274292,
+      "learning_rate": 3.1639667349285905e-05,
+      "loss": 1.0161,
+      "step": 276500
+    },
+    {
+      "epoch": 1.1036120385347856,
+      "grad_norm": 1.581350326538086,
+      "learning_rate": 3.160646602442024e-05,
+      "loss": 1.0171,
+      "step": 277000
+    },
+    {
+      "epoch": 1.1056041180267258,
+      "grad_norm": 1.5592961311340332,
+      "learning_rate": 3.157326469955457e-05,
+      "loss": 1.0188,
+      "step": 277500
+    },
+    {
+      "epoch": 1.107596197518666,
+      "grad_norm": 1.6156282424926758,
+      "learning_rate": 3.1540063374688906e-05,
+      "loss": 1.0138,
+      "step": 278000
+    },
+    {
+      "epoch": 1.1095882770106058,
+      "grad_norm": 1.6247326135635376,
+      "learning_rate": 3.150686204982324e-05,
+      "loss": 1.0119,
+      "step": 278500
+    },
+    {
+      "epoch": 1.111580356502546,
+      "grad_norm": 1.6090619564056396,
+      "learning_rate": 3.147366072495757e-05,
+      "loss": 1.0163,
+      "step": 279000
+    },
+    {
+      "epoch": 1.1135724359944859,
+      "grad_norm": 1.677801489830017,
+      "learning_rate": 3.1440459400091906e-05,
+      "loss": 1.01,
+      "step": 279500
+    },
+    {
+      "epoch": 1.115564515486426,
+      "grad_norm": 1.5337806940078735,
+      "learning_rate": 3.1407258075226234e-05,
+      "loss": 1.0109,
+      "step": 280000
+    },
+    {
+      "epoch": 1.1175565949783661,
+      "grad_norm": 1.6475934982299805,
+      "learning_rate": 3.137405675036057e-05,
+      "loss": 1.0103,
+      "step": 280500
+    },
+    {
+      "epoch": 1.119548674470306,
+      "grad_norm": 1.528709053993225,
+      "learning_rate": 3.1340855425494906e-05,
+      "loss": 1.0114,
+      "step": 281000
+    },
+    {
+      "epoch": 1.1215407539622462,
+      "grad_norm": 1.5790069103240967,
+      "learning_rate": 3.1307654100629235e-05,
+      "loss": 1.0126,
+      "step": 281500
+    },
+    {
+      "epoch": 1.123532833454186,
+      "grad_norm": 1.6203980445861816,
+      "learning_rate": 3.1274452775763563e-05,
+      "loss": 1.015,
+      "step": 282000
+    },
+    {
+      "epoch": 1.1255249129461262,
+      "grad_norm": 1.5446475744247437,
+      "learning_rate": 3.12412514508979e-05,
+      "loss": 1.0124,
+      "step": 282500
+    },
+    {
+      "epoch": 1.1275169924380664,
+      "grad_norm": 1.571184515953064,
+      "learning_rate": 3.120805012603223e-05,
+      "loss": 1.0105,
+      "step": 283000
+    },
+    {
+      "epoch": 1.1295090719300063,
+      "grad_norm": 1.4736685752868652,
+      "learning_rate": 3.1174848801166563e-05,
+      "loss": 1.0117,
+      "step": 283500
+    },
+    {
+      "epoch": 1.1315011514219464,
+      "grad_norm": 1.4694329500198364,
+      "learning_rate": 3.114164747630089e-05,
+      "loss": 1.0113,
+      "step": 284000
+    },
+    {
+      "epoch": 1.1334932309138863,
+      "grad_norm": 1.5552548170089722,
+      "learning_rate": 3.110844615143523e-05,
+      "loss": 1.0122,
+      "step": 284500
+    },
+    {
+      "epoch": 1.1354853104058265,
+      "grad_norm": 1.5717132091522217,
+      "learning_rate": 3.1075244826569564e-05,
+      "loss": 1.0181,
+      "step": 285000
+    },
+    {
+      "epoch": 1.1374773898977666,
+      "grad_norm": 1.6472405195236206,
+      "learning_rate": 3.104204350170389e-05,
+      "loss": 1.0107,
+      "step": 285500
+    },
+    {
+      "epoch": 1.1394694693897065,
+      "grad_norm": 1.4941881895065308,
+      "learning_rate": 3.100884217683823e-05,
+      "loss": 1.0152,
+      "step": 286000
+    },
+    {
+      "epoch": 1.1414615488816466,
+      "grad_norm": 1.6063954830169678,
+      "learning_rate": 3.097564085197256e-05,
+      "loss": 1.0084,
+      "step": 286500
+    },
+    {
+      "epoch": 1.1434536283735866,
+      "grad_norm": 1.5692435503005981,
+      "learning_rate": 3.094243952710689e-05,
+      "loss": 1.0128,
+      "step": 287000
+    },
+    {
+      "epoch": 1.1454457078655267,
+      "grad_norm": 1.7007684707641602,
+      "learning_rate": 3.090923820224123e-05,
+      "loss": 1.0087,
+      "step": 287500
+    },
+    {
+      "epoch": 1.1474377873574668,
+      "grad_norm": 1.551627516746521,
+      "learning_rate": 3.087603687737556e-05,
+      "loss": 1.016,
+      "step": 288000
+    },
+    {
+      "epoch": 1.1494298668494067,
+      "grad_norm": 1.489047646522522,
+      "learning_rate": 3.084283555250989e-05,
+      "loss": 1.0158,
+      "step": 288500
+    },
+    {
+      "epoch": 1.1514219463413469,
+      "grad_norm": 1.66600501537323,
+      "learning_rate": 3.080963422764422e-05,
+      "loss": 1.0059,
+      "step": 289000
+    },
+    {
+      "epoch": 1.1534140258332868,
+      "grad_norm": 1.5344865322113037,
+      "learning_rate": 3.077643290277855e-05,
+      "loss": 1.0092,
+      "step": 289500
+    },
+    {
+      "epoch": 1.155406105325227,
+      "grad_norm": 1.5712891817092896,
+      "learning_rate": 3.0743231577912886e-05,
+      "loss": 1.0075,
+      "step": 290000
+    },
+    {
+      "epoch": 1.157398184817167,
+      "grad_norm": 1.6189531087875366,
+      "learning_rate": 3.0710030253047215e-05,
+      "loss": 1.0093,
+      "step": 290500
+    },
+    {
+      "epoch": 1.159390264309107,
+      "grad_norm": 1.5747156143188477,
+      "learning_rate": 3.067682892818155e-05,
+      "loss": 1.0087,
+      "step": 291000
+    },
+    {
+      "epoch": 1.161382343801047,
+      "grad_norm": 1.5633164644241333,
+      "learning_rate": 3.0643627603315886e-05,
+      "loss": 1.0088,
+      "step": 291500
+    },
+    {
+      "epoch": 1.163374423292987,
+      "grad_norm": 1.6010435819625854,
+      "learning_rate": 3.0610426278450215e-05,
+      "loss": 1.0085,
+      "step": 292000
+    },
+    {
+      "epoch": 1.1653665027849271,
+      "grad_norm": 1.5415773391723633,
+      "learning_rate": 3.057722495358455e-05,
+      "loss": 1.0061,
+      "step": 292500
+    },
+    {
+      "epoch": 1.1673585822768673,
+      "grad_norm": 1.591548204421997,
+      "learning_rate": 3.054402362871888e-05,
+      "loss": 1.0059,
+      "step": 293000
+    },
+    {
+      "epoch": 1.1693506617688072,
+      "grad_norm": 1.470170497894287,
+      "learning_rate": 3.0510822303853215e-05,
+      "loss": 1.0077,
+      "step": 293500
+    },
+    {
+      "epoch": 1.1713427412607473,
+      "grad_norm": 1.5852841138839722,
+      "learning_rate": 3.0477620978987547e-05,
+      "loss": 1.0107,
+      "step": 294000
+    },
+    {
+      "epoch": 1.1733348207526872,
+      "grad_norm": 1.6136342287063599,
+      "learning_rate": 3.044441965412188e-05,
+      "loss": 1.0035,
+      "step": 294500
+    },
+    {
+      "epoch": 1.1753269002446274,
+      "grad_norm": 1.5250244140625,
+      "learning_rate": 3.0411218329256215e-05,
+      "loss": 1.0075,
+      "step": 295000
+    },
+    {
+      "epoch": 1.1773189797365675,
+      "grad_norm": 1.5861165523529053,
+      "learning_rate": 3.0378017004390547e-05,
+      "loss": 1.003,
+      "step": 295500
+    },
+    {
+      "epoch": 1.1793110592285074,
+      "grad_norm": 1.5056827068328857,
+      "learning_rate": 3.034481567952488e-05,
+      "loss": 1.0066,
+      "step": 296000
+    },
+    {
+      "epoch": 1.1813031387204476,
+      "grad_norm": 1.5771924257278442,
+      "learning_rate": 3.031161435465921e-05,
+      "loss": 1.0082,
+      "step": 296500
+    },
+    {
+      "epoch": 1.1832952182123875,
+      "grad_norm": 1.5771832466125488,
+      "learning_rate": 3.027841302979354e-05,
+      "loss": 1.003,
+      "step": 297000
+    },
+    {
+      "epoch": 1.1852872977043276,
+      "grad_norm": 1.4951648712158203,
+      "learning_rate": 3.0245211704927873e-05,
+      "loss": 1.0051,
+      "step": 297500
+    },
+    {
+      "epoch": 1.1872793771962677,
+      "grad_norm": 1.6125141382217407,
+      "learning_rate": 3.0212010380062205e-05,
+      "loss": 1.0055,
+      "step": 298000
+    },
+    {
+      "epoch": 1.1892714566882077,
+      "grad_norm": 1.6001217365264893,
+      "learning_rate": 3.0178809055196537e-05,
+      "loss": 1.0043,
+      "step": 298500
+    },
+    {
+      "epoch": 1.1912635361801478,
+      "grad_norm": 1.5025718212127686,
+      "learning_rate": 3.0145607730330873e-05,
+      "loss": 1.0042,
+      "step": 299000
+    },
+    {
+      "epoch": 1.1932556156720877,
+      "grad_norm": 1.527305006980896,
+      "learning_rate": 3.0112406405465205e-05,
+      "loss": 1.0076,
+      "step": 299500
+    },
+    {
+      "epoch": 1.1952476951640278,
+      "grad_norm": 1.4677504301071167,
+      "learning_rate": 3.0079205080599537e-05,
+      "loss": 1.0089,
+      "step": 300000
+    },
+    {
+      "epoch": 1.197239774655968,
+      "grad_norm": 1.558254361152649,
+      "learning_rate": 3.004600375573387e-05,
+      "loss": 1.0067,
+      "step": 300500
+    },
+    {
+      "epoch": 1.199231854147908,
+      "grad_norm": 1.5117969512939453,
+      "learning_rate": 3.0012802430868202e-05,
+      "loss": 1.0032,
+      "step": 301000
+    },
+    {
+      "epoch": 1.201223933639848,
+      "grad_norm": 1.5806593894958496,
+      "learning_rate": 2.9979601106002537e-05,
+      "loss": 1.0089,
+      "step": 301500
+    },
+    {
+      "epoch": 1.203216013131788,
+      "grad_norm": 1.6340941190719604,
+      "learning_rate": 2.994639978113687e-05,
+      "loss": 1.004,
+      "step": 302000
+    },
+    {
+      "epoch": 1.205208092623728,
+      "grad_norm": 1.775723934173584,
+      "learning_rate": 2.9913198456271202e-05,
+      "loss": 1.0003,
+      "step": 302500
+    },
+    {
+      "epoch": 1.2072001721156682,
+      "grad_norm": 1.5807640552520752,
+      "learning_rate": 2.9879997131405534e-05,
+      "loss": 1.0002,
+      "step": 303000
+    },
+    {
+      "epoch": 1.2091922516076081,
+      "grad_norm": 1.6332534551620483,
+      "learning_rate": 2.9846795806539866e-05,
+      "loss": 1.0059,
+      "step": 303500
+    },
+    {
+      "epoch": 1.2111843310995483,
+      "grad_norm": 1.4831372499465942,
+      "learning_rate": 2.9813594481674202e-05,
+      "loss": 1.0046,
+      "step": 304000
+    },
+    {
+      "epoch": 1.2131764105914882,
+      "grad_norm": 1.5793821811676025,
+      "learning_rate": 2.9780393156808527e-05,
+      "loss": 1.0065,
+      "step": 304500
+    },
+    {
+      "epoch": 1.2151684900834283,
+      "grad_norm": 1.6091269254684448,
+      "learning_rate": 2.974719183194286e-05,
+      "loss": 1.0022,
+      "step": 305000
+    },
+    {
+      "epoch": 1.2171605695753684,
+      "grad_norm": 1.5727423429489136,
+      "learning_rate": 2.9713990507077195e-05,
+      "loss": 1.0015,
+      "step": 305500
+    },
+    {
+      "epoch": 1.2191526490673084,
+      "grad_norm": 1.5554064512252808,
+      "learning_rate": 2.9680789182211527e-05,
+      "loss": 1.0051,
+      "step": 306000
+    },
+    {
+      "epoch": 1.2211447285592485,
+      "grad_norm": 1.5735516548156738,
+      "learning_rate": 2.964758785734586e-05,
+      "loss": 1.0003,
+      "step": 306500
+    },
+    {
+      "epoch": 1.2231368080511884,
+      "grad_norm": 1.6825499534606934,
+      "learning_rate": 2.9614386532480192e-05,
+      "loss": 1.0002,
+      "step": 307000
+    },
+    {
+      "epoch": 1.2251288875431285,
+      "grad_norm": 1.6207140684127808,
+      "learning_rate": 2.9581185207614524e-05,
+      "loss": 0.9968,
+      "step": 307500
+    },
+    {
+      "epoch": 1.2271209670350687,
+      "grad_norm": 1.586870789527893,
+      "learning_rate": 2.954798388274886e-05,
+      "loss": 1.0011,
+      "step": 308000
+    },
+    {
+      "epoch": 1.2291130465270086,
+      "grad_norm": 1.608620524406433,
+      "learning_rate": 2.9514782557883192e-05,
+      "loss": 1.0035,
+      "step": 308500
+    },
+    {
+      "epoch": 1.2311051260189487,
+      "grad_norm": 1.5619418621063232,
+      "learning_rate": 2.9481581233017524e-05,
+      "loss": 1.0012,
+      "step": 309000
+    },
+    {
+      "epoch": 1.2330972055108886,
+      "grad_norm": 1.5945076942443848,
+      "learning_rate": 2.9448379908151856e-05,
+      "loss": 1.0113,
+      "step": 309500
+    },
+    {
+      "epoch": 1.2350892850028288,
+      "grad_norm": 1.5399153232574463,
+      "learning_rate": 2.9415178583286192e-05,
+      "loss": 1.0041,
+      "step": 310000
+    },
+    {
+      "epoch": 1.237081364494769,
+      "grad_norm": 1.6591147184371948,
+      "learning_rate": 2.9381977258420524e-05,
+      "loss": 1.001,
+      "step": 310500
+    },
+    {
+      "epoch": 1.2390734439867088,
+      "grad_norm": 1.611533522605896,
+      "learning_rate": 2.9348775933554857e-05,
+      "loss": 1.0097,
+      "step": 311000
+    },
+    {
+      "epoch": 1.241065523478649,
+      "grad_norm": 1.5552842617034912,
+      "learning_rate": 2.931557460868919e-05,
+      "loss": 0.9987,
+      "step": 311500
+    },
+    {
+      "epoch": 1.2430576029705889,
+      "grad_norm": 1.6259946823120117,
+      "learning_rate": 2.9282373283823518e-05,
+      "loss": 0.9997,
+      "step": 312000
+    },
+    {
+      "epoch": 1.245049682462529,
+      "grad_norm": 1.5336946249008179,
+      "learning_rate": 2.924917195895785e-05,
+      "loss": 1.0026,
+      "step": 312500
+    },
+    {
+      "epoch": 1.2470417619544691,
+      "grad_norm": 1.4883583784103394,
+      "learning_rate": 2.9215970634092182e-05,
+      "loss": 0.9988,
+      "step": 313000
+    },
+    {
+      "epoch": 1.249033841446409,
+      "grad_norm": 1.5403988361358643,
+      "learning_rate": 2.9182769309226514e-05,
+      "loss": 1.0006,
+      "step": 313500
+    },
+    {
+      "epoch": 1.2510259209383492,
+      "grad_norm": 1.6766672134399414,
+      "learning_rate": 2.914956798436085e-05,
+      "loss": 1.003,
+      "step": 314000
+    },
+    {
+      "epoch": 1.253018000430289,
+      "grad_norm": 1.55039381980896,
+      "learning_rate": 2.9116366659495182e-05,
+      "loss": 0.9983,
+      "step": 314500
+    },
+    {
+      "epoch": 1.2550100799222292,
+      "grad_norm": 1.5901780128479004,
+      "learning_rate": 2.9083165334629514e-05,
+      "loss": 0.9979,
+      "step": 315000
+    },
+    {
+      "epoch": 1.2570021594141694,
+      "grad_norm": 1.558403730392456,
+      "learning_rate": 2.9049964009763847e-05,
+      "loss": 1.0039,
+      "step": 315500
+    },
+    {
+      "epoch": 1.2589942389061093,
+      "grad_norm": 1.4940515756607056,
+      "learning_rate": 2.901676268489818e-05,
+      "loss": 0.9984,
+      "step": 316000
+    },
+    {
+      "epoch": 1.2609863183980494,
+      "grad_norm": 1.560308575630188,
+      "learning_rate": 2.8983561360032514e-05,
+      "loss": 1.0013,
+      "step": 316500
+    },
+    {
+      "epoch": 1.2629783978899893,
+      "grad_norm": 1.6117875576019287,
+      "learning_rate": 2.8950360035166847e-05,
+      "loss": 0.9943,
+      "step": 317000
+    },
+    {
+      "epoch": 1.2649704773819295,
+      "grad_norm": 1.603472352027893,
+      "learning_rate": 2.891715871030118e-05,
+      "loss": 0.9997,
+      "step": 317500
+    },
+    {
+      "epoch": 1.2669625568738696,
+      "grad_norm": 1.6008336544036865,
+      "learning_rate": 2.888395738543551e-05,
+      "loss": 1.0041,
+      "step": 318000
+    },
+    {
+      "epoch": 1.2689546363658095,
+      "grad_norm": 1.5423409938812256,
+      "learning_rate": 2.8850756060569843e-05,
+      "loss": 1.0021,
+      "step": 318500
+    },
+    {
+      "epoch": 1.2709467158577497,
+      "grad_norm": 1.4314305782318115,
+      "learning_rate": 2.881755473570418e-05,
+      "loss": 0.9979,
+      "step": 319000
+    },
+    {
+      "epoch": 1.2729387953496896,
+      "grad_norm": 1.5170232057571411,
+      "learning_rate": 2.8784353410838504e-05,
+      "loss": 1.0031,
+      "step": 319500
+    },
+    {
+      "epoch": 1.2749308748416297,
+      "grad_norm": 1.6106996536254883,
+      "learning_rate": 2.8751152085972837e-05,
+      "loss": 1.0004,
+      "step": 320000
+    },
+    {
+      "epoch": 1.2769229543335698,
+      "grad_norm": 1.5303128957748413,
+      "learning_rate": 2.8717950761107172e-05,
+      "loss": 0.9985,
+      "step": 320500
+    },
+    {
+      "epoch": 1.2789150338255098,
+      "grad_norm": 1.5119428634643555,
+      "learning_rate": 2.8684749436241505e-05,
+      "loss": 0.9971,
+      "step": 321000
+    },
+    {
+      "epoch": 1.2809071133174499,
+      "grad_norm": 1.5663682222366333,
+      "learning_rate": 2.8651548111375837e-05,
+      "loss": 0.9983,
+      "step": 321500
+    },
+    {
+      "epoch": 1.2828991928093898,
+      "grad_norm": 1.5699875354766846,
+      "learning_rate": 2.861834678651017e-05,
+      "loss": 0.9958,
+      "step": 322000
+    },
+    {
+      "epoch": 1.28489127230133,
+      "grad_norm": 1.4993077516555786,
+      "learning_rate": 2.85851454616445e-05,
+      "loss": 0.9917,
+      "step": 322500
+    },
+    {
+      "epoch": 1.28688335179327,
+      "grad_norm": 1.6096224784851074,
+      "learning_rate": 2.8551944136778837e-05,
+      "loss": 0.9988,
+      "step": 323000
+    },
+    {
+      "epoch": 1.28887543128521,
+      "grad_norm": 1.6216344833374023,
+      "learning_rate": 2.851874281191317e-05,
+      "loss": 0.9939,
+      "step": 323500
+    },
+    {
+      "epoch": 1.2908675107771501,
+      "grad_norm": 1.5237042903900146,
+      "learning_rate": 2.84855414870475e-05,
+      "loss": 1.0003,
+      "step": 324000
+    },
+    {
+      "epoch": 1.29285959026909,
+      "grad_norm": 1.592710256576538,
+      "learning_rate": 2.8452340162181834e-05,
+      "loss": 0.9977,
+      "step": 324500
+    },
+    {
+      "epoch": 1.2948516697610302,
+      "grad_norm": 1.6038693189620972,
+      "learning_rate": 2.8419138837316166e-05,
+      "loss": 0.9964,
+      "step": 325000
+    },
+    {
+      "epoch": 1.2968437492529703,
+      "grad_norm": 1.5696605443954468,
+      "learning_rate": 2.83859375124505e-05,
+      "loss": 0.9954,
+      "step": 325500
+    },
+    {
+      "epoch": 1.2988358287449102,
+      "grad_norm": 1.6246914863586426,
+      "learning_rate": 2.8352736187584834e-05,
+      "loss": 0.9987,
+      "step": 326000
+    },
+    {
+      "epoch": 1.3008279082368503,
+      "grad_norm": 1.5583215951919556,
+      "learning_rate": 2.8319534862719166e-05,
+      "loss": 0.9964,
+      "step": 326500
+    },
+    {
+      "epoch": 1.3028199877287903,
+      "grad_norm": 1.5545244216918945,
+      "learning_rate": 2.8286333537853495e-05,
+      "loss": 0.9952,
+      "step": 327000
+    },
+    {
+      "epoch": 1.3048120672207304,
+      "grad_norm": 1.5580933094024658,
+      "learning_rate": 2.8253132212987827e-05,
+      "loss": 0.9996,
+      "step": 327500
+    },
+    {
+      "epoch": 1.3068041467126705,
+      "grad_norm": 1.5481173992156982,
+      "learning_rate": 2.821993088812216e-05,
+      "loss": 0.9959,
+      "step": 328000
+    },
+    {
+      "epoch": 1.3087962262046104,
+      "grad_norm": 1.5840754508972168,
+      "learning_rate": 2.818672956325649e-05,
+      "loss": 0.9986,
+      "step": 328500
+    },
+    {
+      "epoch": 1.3107883056965506,
+      "grad_norm": 1.557962417602539,
+      "learning_rate": 2.8153528238390824e-05,
+      "loss": 0.998,
+      "step": 329000
+    },
+    {
+      "epoch": 1.3127803851884905,
+      "grad_norm": 1.4909387826919556,
+      "learning_rate": 2.812032691352516e-05,
+      "loss": 0.9957,
+      "step": 329500
+    },
+    {
+      "epoch": 1.3147724646804306,
+      "grad_norm": 1.598558783531189,
+      "learning_rate": 2.808712558865949e-05,
+      "loss": 1.0,
+      "step": 330000
+    },
+    {
+      "epoch": 1.3167645441723708,
+      "grad_norm": 1.6057049036026,
+      "learning_rate": 2.8053924263793824e-05,
+      "loss": 0.9958,
+      "step": 330500
+    },
+    {
+      "epoch": 1.3187566236643107,
+      "grad_norm": 1.5668245553970337,
+      "learning_rate": 2.8020722938928156e-05,
+      "loss": 0.995,
+      "step": 331000
+    },
+    {
+      "epoch": 1.3207487031562508,
+      "grad_norm": 1.7019222974777222,
+      "learning_rate": 2.7987521614062488e-05,
+      "loss": 0.9897,
+      "step": 331500
+    },
+    {
+      "epoch": 1.3227407826481907,
+      "grad_norm": 1.5326474905014038,
+      "learning_rate": 2.7954320289196824e-05,
+      "loss": 0.9998,
+      "step": 332000
+    },
+    {
+      "epoch": 1.3247328621401309,
+      "grad_norm": 1.7239539623260498,
+      "learning_rate": 2.7921118964331156e-05,
+      "loss": 0.9934,
+      "step": 332500
+    },
+    {
+      "epoch": 1.326724941632071,
+      "grad_norm": 1.5657050609588623,
+      "learning_rate": 2.7887917639465488e-05,
+      "loss": 0.9945,
+      "step": 333000
+    },
+    {
+      "epoch": 1.328717021124011,
+      "grad_norm": 1.6078459024429321,
+      "learning_rate": 2.785471631459982e-05,
+      "loss": 0.9948,
+      "step": 333500
+    },
+    {
+      "epoch": 1.330709100615951,
+      "grad_norm": 1.5506103038787842,
+      "learning_rate": 2.7821514989734153e-05,
+      "loss": 0.995,
+      "step": 334000
+    },
+    {
+      "epoch": 1.332701180107891,
+      "grad_norm": 1.6498535871505737,
+      "learning_rate": 2.778831366486848e-05,
+      "loss": 0.9974,
+      "step": 334500
+    },
+    {
+      "epoch": 1.334693259599831,
+      "grad_norm": 1.6052839756011963,
+      "learning_rate": 2.7755112340002814e-05,
+      "loss": 0.9942,
+      "step": 335000
+    },
+    {
+      "epoch": 1.3366853390917712,
+      "grad_norm": 1.5408515930175781,
+      "learning_rate": 2.7721911015137146e-05,
+      "loss": 0.9956,
+      "step": 335500
+    },
+    {
+      "epoch": 1.3386774185837111,
+      "grad_norm": 1.5408698320388794,
+      "learning_rate": 2.768870969027148e-05,
+      "loss": 0.9921,
+      "step": 336000
+    },
+    {
+      "epoch": 1.3406694980756513,
+      "grad_norm": 1.613200068473816,
+      "learning_rate": 2.7655508365405814e-05,
+      "loss": 0.9955,
+      "step": 336500
+    },
+    {
+      "epoch": 1.3426615775675912,
+      "grad_norm": 1.6325806379318237,
+      "learning_rate": 2.7622307040540146e-05,
+      "loss": 0.9924,
+      "step": 337000
+    },
+    {
+      "epoch": 1.3446536570595313,
+      "grad_norm": 1.5414601564407349,
+      "learning_rate": 2.7589105715674478e-05,
+      "loss": 0.9929,
+      "step": 337500
+    },
+    {
+      "epoch": 1.3466457365514715,
+      "grad_norm": 1.608492136001587,
+      "learning_rate": 2.755590439080881e-05,
+      "loss": 0.9931,
+      "step": 338000
+    },
+    {
+      "epoch": 1.3486378160434114,
+      "grad_norm": 1.540972352027893,
+      "learning_rate": 2.7522703065943146e-05,
+      "loss": 0.9913,
+      "step": 338500
+    },
+    {
+      "epoch": 1.3506298955353515,
+      "grad_norm": 1.568434238433838,
+      "learning_rate": 2.748950174107748e-05,
+      "loss": 0.9971,
+      "step": 339000
+    },
+    {
+      "epoch": 1.3526219750272914,
+      "grad_norm": 1.5180193185806274,
+      "learning_rate": 2.745630041621181e-05,
+      "loss": 0.9937,
+      "step": 339500
+    },
+    {
+      "epoch": 1.3546140545192316,
+      "grad_norm": 1.5868713855743408,
+      "learning_rate": 2.7423099091346143e-05,
+      "loss": 0.9937,
+      "step": 340000
+    },
+    {
+      "epoch": 1.3566061340111717,
+      "grad_norm": 1.5642015933990479,
+      "learning_rate": 2.738989776648048e-05,
+      "loss": 0.9919,
+      "step": 340500
+    },
+    {
+      "epoch": 1.3585982135031116,
+      "grad_norm": 1.5025302171707153,
+      "learning_rate": 2.735669644161481e-05,
+      "loss": 0.9911,
+      "step": 341000
+    },
+    {
+      "epoch": 1.3605902929950517,
+      "grad_norm": 1.5415868759155273,
+      "learning_rate": 2.7323495116749143e-05,
+      "loss": 0.9898,
+      "step": 341500
+    },
+    {
+      "epoch": 1.3625823724869917,
+      "grad_norm": 1.644484043121338,
+      "learning_rate": 2.729029379188347e-05,
+      "loss": 0.9873,
+      "step": 342000
+    },
+    {
+      "epoch": 1.3645744519789318,
+      "grad_norm": 1.5348944664001465,
+      "learning_rate": 2.7257092467017804e-05,
+      "loss": 0.9918,
+      "step": 342500
+    },
+    {
+      "epoch": 1.366566531470872,
+      "grad_norm": 1.5913807153701782,
+      "learning_rate": 2.7223891142152136e-05,
+      "loss": 0.9909,
+      "step": 343000
+    },
+    {
+      "epoch": 1.3685586109628118,
+      "grad_norm": 1.5665234327316284,
+      "learning_rate": 2.719068981728647e-05,
+      "loss": 0.9927,
+      "step": 343500
+    },
+    {
+      "epoch": 1.370550690454752,
+      "grad_norm": 1.5880101919174194,
+      "learning_rate": 2.71574884924208e-05,
+      "loss": 0.9903,
+      "step": 344000
+    },
+    {
+      "epoch": 1.3725427699466919,
+      "grad_norm": 1.5287762880325317,
+      "learning_rate": 2.7124287167555136e-05,
+      "loss": 0.994,
+      "step": 344500
+    },
+    {
+      "epoch": 1.374534849438632,
+      "grad_norm": 1.5047370195388794,
+      "learning_rate": 2.709108584268947e-05,
+      "loss": 0.9935,
+      "step": 345000
+    },
+    {
+      "epoch": 1.3765269289305722,
+      "grad_norm": 1.5519222021102905,
+      "learning_rate": 2.70578845178238e-05,
+      "loss": 0.9916,
+      "step": 345500
+    },
+    {
+      "epoch": 1.378519008422512,
+      "grad_norm": 1.5660203695297241,
+      "learning_rate": 2.7024683192958133e-05,
+      "loss": 0.9873,
+      "step": 346000
+    },
+    {
+      "epoch": 1.3805110879144522,
+      "grad_norm": 1.5806609392166138,
+      "learning_rate": 2.6991481868092465e-05,
+      "loss": 0.9917,
+      "step": 346500
+    },
+    {
+      "epoch": 1.3825031674063921,
+      "grad_norm": 1.6304339170455933,
+      "learning_rate": 2.69582805432268e-05,
+      "loss": 0.9882,
+      "step": 347000
+    },
+    {
+      "epoch": 1.3844952468983323,
+      "grad_norm": 1.5563949346542358,
+      "learning_rate": 2.6925079218361133e-05,
+      "loss": 0.9863,
+      "step": 347500
+    },
+    {
+      "epoch": 1.3864873263902724,
+      "grad_norm": 1.5538074970245361,
+      "learning_rate": 2.6891877893495465e-05,
+      "loss": 0.9897,
+      "step": 348000
+    },
+    {
+      "epoch": 1.3884794058822123,
+      "grad_norm": 1.6218252182006836,
+      "learning_rate": 2.6858676568629798e-05,
+      "loss": 0.9917,
+      "step": 348500
+    },
+    {
+      "epoch": 1.3904714853741524,
+      "grad_norm": 1.5162501335144043,
+      "learning_rate": 2.682547524376413e-05,
+      "loss": 0.9944,
+      "step": 349000
+    },
+    {
+      "epoch": 1.3924635648660924,
+      "grad_norm": 1.6467870473861694,
+      "learning_rate": 2.679227391889846e-05,
+      "loss": 0.989,
+      "step": 349500
+    },
+    {
+      "epoch": 1.3944556443580325,
+      "grad_norm": 1.5588877201080322,
+      "learning_rate": 2.675907259403279e-05,
+      "loss": 0.9924,
+      "step": 350000
+    },
+    {
+      "epoch": 1.3964477238499726,
+      "grad_norm": 1.647695779800415,
+      "learning_rate": 2.6725871269167123e-05,
+      "loss": 0.9918,
+      "step": 350500
+    },
+    {
+      "epoch": 1.3984398033419125,
+      "grad_norm": 1.47678804397583,
+      "learning_rate": 2.669266994430146e-05,
+      "loss": 0.9889,
+      "step": 351000
+    },
+    {
+      "epoch": 1.4004318828338527,
+      "grad_norm": 1.512880802154541,
+      "learning_rate": 2.665946861943579e-05,
+      "loss": 0.9892,
+      "step": 351500
+    },
+    {
+      "epoch": 1.4024239623257926,
+      "grad_norm": 1.5763474702835083,
+      "learning_rate": 2.6626267294570123e-05,
+      "loss": 0.989,
+      "step": 352000
+    },
+    {
+      "epoch": 1.4044160418177327,
+      "grad_norm": 1.5481244325637817,
+      "learning_rate": 2.6593065969704455e-05,
+      "loss": 0.9907,
+      "step": 352500
+    },
+    {
+      "epoch": 1.4064081213096729,
+      "grad_norm": 1.5645235776901245,
+      "learning_rate": 2.6559864644838788e-05,
+      "loss": 0.9909,
+      "step": 353000
+    },
+    {
+      "epoch": 1.4084002008016128,
+      "grad_norm": 1.629308819770813,
+      "learning_rate": 2.6526663319973123e-05,
+      "loss": 0.9892,
+      "step": 353500
+    },
+    {
+      "epoch": 1.410392280293553,
+      "grad_norm": 1.5493535995483398,
+      "learning_rate": 2.6493461995107455e-05,
+      "loss": 0.9872,
+      "step": 354000
+    },
+    {
+      "epoch": 1.4123843597854928,
+      "grad_norm": 1.4820462465286255,
+      "learning_rate": 2.6460260670241788e-05,
+      "loss": 0.9873,
+      "step": 354500
+    },
+    {
+      "epoch": 1.414376439277433,
+      "grad_norm": 1.6087406873703003,
+      "learning_rate": 2.642705934537612e-05,
+      "loss": 0.99,
+      "step": 355000
+    },
+    {
+      "epoch": 1.416368518769373,
+      "grad_norm": 1.5790669918060303,
+      "learning_rate": 2.6393858020510452e-05,
+      "loss": 0.9827,
+      "step": 355500
+    },
+    {
+      "epoch": 1.418360598261313,
+      "grad_norm": 1.5915374755859375,
+      "learning_rate": 2.6360656695644788e-05,
+      "loss": 0.991,
+      "step": 356000
+    },
+    {
+      "epoch": 1.4203526777532531,
+      "grad_norm": 1.638120174407959,
+      "learning_rate": 2.632745537077912e-05,
+      "loss": 0.9888,
+      "step": 356500
+    },
+    {
+      "epoch": 1.422344757245193,
+      "grad_norm": 1.627324104309082,
+      "learning_rate": 2.6294254045913445e-05,
+      "loss": 0.988,
+      "step": 357000
+    },
+    {
+      "epoch": 1.4243368367371332,
+      "grad_norm": 1.5350537300109863,
+      "learning_rate": 2.626105272104778e-05,
+      "loss": 0.9923,
+      "step": 357500
+    },
+    {
+      "epoch": 1.4263289162290733,
+      "grad_norm": 1.4739323854446411,
+      "learning_rate": 2.6227851396182113e-05,
+      "loss": 0.9849,
+      "step": 358000
+    },
+    {
+      "epoch": 1.4283209957210132,
+      "grad_norm": 1.6218150854110718,
+      "learning_rate": 2.6194650071316446e-05,
+      "loss": 0.9847,
+      "step": 358500
+    },
+    {
+      "epoch": 1.4303130752129534,
+      "grad_norm": 1.5591546297073364,
+      "learning_rate": 2.6161448746450778e-05,
+      "loss": 0.9882,
+      "step": 359000
+    },
+    {
+      "epoch": 1.4323051547048933,
+      "grad_norm": 1.6043953895568848,
+      "learning_rate": 2.612824742158511e-05,
+      "loss": 0.9899,
+      "step": 359500
+    },
+    {
+      "epoch": 1.4342972341968334,
+      "grad_norm": 1.6079070568084717,
+      "learning_rate": 2.6095046096719446e-05,
+      "loss": 0.9832,
+      "step": 360000
+    },
+    {
+      "epoch": 1.4362893136887736,
+      "grad_norm": 1.6043310165405273,
+      "learning_rate": 2.6061844771853778e-05,
+      "loss": 0.9853,
+      "step": 360500
+    },
+    {
+      "epoch": 1.4382813931807135,
+      "grad_norm": 1.4828139543533325,
+      "learning_rate": 2.602864344698811e-05,
+      "loss": 0.9816,
+      "step": 361000
+    },
+    {
+      "epoch": 1.4402734726726536,
+      "grad_norm": 1.5727781057357788,
+      "learning_rate": 2.5995442122122442e-05,
+      "loss": 0.9864,
+      "step": 361500
+    },
+    {
+      "epoch": 1.4422655521645935,
+      "grad_norm": 1.588922142982483,
+      "learning_rate": 2.5962240797256775e-05,
+      "loss": 0.9822,
+      "step": 362000
+    },
+    {
+      "epoch": 1.4442576316565336,
+      "grad_norm": 1.5049546957015991,
+      "learning_rate": 2.592903947239111e-05,
+      "loss": 0.9838,
+      "step": 362500
+    },
+    {
+      "epoch": 1.4462497111484738,
+      "grad_norm": 1.6511805057525635,
+      "learning_rate": 2.5895838147525442e-05,
+      "loss": 0.988,
+      "step": 363000
+    },
+    {
+      "epoch": 1.4482417906404137,
+      "grad_norm": 1.6251769065856934,
+      "learning_rate": 2.5862636822659775e-05,
+      "loss": 0.9887,
+      "step": 363500
+    },
+    {
+      "epoch": 1.4502338701323538,
+      "grad_norm": 1.5443693399429321,
+      "learning_rate": 2.5829435497794107e-05,
+      "loss": 0.9851,
+      "step": 364000
+    },
+    {
+      "epoch": 1.4522259496242937,
+      "grad_norm": 1.5328843593597412,
+      "learning_rate": 2.5796234172928436e-05,
+      "loss": 0.9854,
+      "step": 364500
+    },
+    {
+      "epoch": 1.4542180291162339,
+      "grad_norm": 1.6825765371322632,
+      "learning_rate": 2.5763032848062768e-05,
+      "loss": 0.9843,
+      "step": 365000
+    },
+    {
+      "epoch": 1.456210108608174,
+      "grad_norm": 1.5050028562545776,
+      "learning_rate": 2.57298315231971e-05,
+      "loss": 0.9858,
+      "step": 365500
+    },
+    {
+      "epoch": 1.458202188100114,
+      "grad_norm": 1.544442057609558,
+      "learning_rate": 2.5696630198331432e-05,
+      "loss": 0.9828,
+      "step": 366000
+    },
+    {
+      "epoch": 1.460194267592054,
+      "grad_norm": 1.7254536151885986,
+      "learning_rate": 2.5663428873465768e-05,
+      "loss": 0.9825,
+      "step": 366500
+    },
+    {
+      "epoch": 1.462186347083994,
+      "grad_norm": 1.6863619089126587,
+      "learning_rate": 2.56302275486001e-05,
+      "loss": 0.9868,
+      "step": 367000
+    },
+    {
+      "epoch": 1.4641784265759341,
+      "grad_norm": 1.5243964195251465,
+      "learning_rate": 2.5597026223734432e-05,
+      "loss": 0.9819,
+      "step": 367500
+    },
+    {
+      "epoch": 1.4661705060678742,
+      "grad_norm": 1.493849515914917,
+      "learning_rate": 2.5563824898868765e-05,
+      "loss": 0.9899,
+      "step": 368000
+    },
+    {
+      "epoch": 1.4681625855598142,
+      "grad_norm": 1.5080227851867676,
+      "learning_rate": 2.5530623574003097e-05,
+      "loss": 0.9871,
+      "step": 368500
+    },
+    {
+      "epoch": 1.4701546650517543,
+      "grad_norm": 1.5969493389129639,
+      "learning_rate": 2.5497422249137433e-05,
+      "loss": 0.9826,
+      "step": 369000
+    },
+    {
+      "epoch": 1.4721467445436942,
+      "grad_norm": 1.541832447052002,
+      "learning_rate": 2.5464220924271765e-05,
+      "loss": 0.9891,
+      "step": 369500
+    },
+    {
+      "epoch": 1.4741388240356343,
+      "grad_norm": 1.548997163772583,
+      "learning_rate": 2.5431019599406097e-05,
+      "loss": 0.9854,
+      "step": 370000
+    },
+    {
+      "epoch": 1.4761309035275745,
+      "grad_norm": 1.577426552772522,
+      "learning_rate": 2.539781827454043e-05,
+      "loss": 0.9818,
+      "step": 370500
+    },
+    {
+      "epoch": 1.4781229830195144,
+      "grad_norm": 1.4438782930374146,
+      "learning_rate": 2.5364616949674765e-05,
+      "loss": 0.9819,
+      "step": 371000
+    },
+    {
+      "epoch": 1.4801150625114545,
+      "grad_norm": 1.5754557847976685,
+      "learning_rate": 2.5331415624809097e-05,
+      "loss": 0.9765,
+      "step": 371500
+    },
+    {
+      "epoch": 1.4821071420033944,
+      "grad_norm": 1.5382401943206787,
+      "learning_rate": 2.5298214299943423e-05,
+      "loss": 0.9795,
+      "step": 372000
+    },
+    {
+      "epoch": 1.4840992214953346,
+      "grad_norm": 1.5432796478271484,
+      "learning_rate": 2.5265012975077755e-05,
+      "loss": 0.9827,
+      "step": 372500
+    },
+    {
+      "epoch": 1.4860913009872747,
+      "grad_norm": 1.5150165557861328,
+      "learning_rate": 2.523181165021209e-05,
+      "loss": 0.984,
+      "step": 373000
+    },
+    {
+      "epoch": 1.4880833804792146,
+      "grad_norm": 1.4984188079833984,
+      "learning_rate": 2.5198610325346423e-05,
+      "loss": 0.9809,
+      "step": 373500
+    },
+    {
+      "epoch": 1.4900754599711548,
+      "grad_norm": 1.5726864337921143,
+      "learning_rate": 2.5165409000480755e-05,
+      "loss": 0.9804,
+      "step": 374000
+    },
+    {
+      "epoch": 1.4920675394630947,
+      "grad_norm": 1.594090461730957,
+      "learning_rate": 2.5132207675615087e-05,
+      "loss": 0.9846,
+      "step": 374500
+    },
+    {
+      "epoch": 1.4940596189550348,
+      "grad_norm": 1.6440484523773193,
+      "learning_rate": 2.5099006350749423e-05,
+      "loss": 0.9866,
+      "step": 375000
+    },
+    {
+      "epoch": 1.496051698446975,
+      "grad_norm": 1.8777843713760376,
+      "learning_rate": 2.5065805025883755e-05,
+      "loss": 0.9797,
+      "step": 375500
+    },
+    {
+      "epoch": 1.4980437779389149,
+      "grad_norm": 1.5775259733200073,
+      "learning_rate": 2.5032603701018087e-05,
+      "loss": 0.9801,
+      "step": 376000
+    },
+    {
+      "epoch": 1.5000358574308548,
+      "grad_norm": 1.5753065347671509,
+      "learning_rate": 2.499940237615242e-05,
+      "loss": 0.9801,
+      "step": 376500
+    },
+    {
+      "epoch": 1.502027936922795,
+      "grad_norm": 1.6271437406539917,
+      "learning_rate": 2.496620105128675e-05,
+      "loss": 0.9796,
+      "step": 377000
+    },
+    {
+      "epoch": 1.504020016414735,
+      "grad_norm": 1.6890838146209717,
+      "learning_rate": 2.4932999726421084e-05,
+      "loss": 0.9793,
+      "step": 377500
+    },
+    {
+      "epoch": 1.5060120959066752,
+      "grad_norm": 1.5349699258804321,
+      "learning_rate": 2.4899798401555416e-05,
+      "loss": 0.979,
+      "step": 378000
+    },
+    {
+      "epoch": 1.508004175398615,
+      "grad_norm": 1.5833008289337158,
+      "learning_rate": 2.4866597076689748e-05,
+      "loss": 0.9794,
+      "step": 378500
+    },
+    {
+      "epoch": 1.509996254890555,
+      "grad_norm": 1.6600587368011475,
+      "learning_rate": 2.483339575182408e-05,
+      "loss": 0.9818,
+      "step": 379000
+    },
+    {
+      "epoch": 1.5119883343824951,
+      "grad_norm": 1.506956934928894,
+      "learning_rate": 2.4800194426958416e-05,
+      "loss": 0.9845,
+      "step": 379500
+    },
+    {
+      "epoch": 1.5139804138744353,
+      "grad_norm": 1.5452346801757812,
+      "learning_rate": 2.476699310209275e-05,
+      "loss": 0.9803,
+      "step": 380000
+    },
+    {
+      "epoch": 1.5159724933663754,
+      "grad_norm": 1.6241753101348877,
+      "learning_rate": 2.473379177722708e-05,
+      "loss": 0.9773,
+      "step": 380500
+    },
+    {
+      "epoch": 1.5179645728583153,
+      "grad_norm": 1.5774579048156738,
+      "learning_rate": 2.4700590452361413e-05,
+      "loss": 0.9802,
+      "step": 381000
+    },
+    {
+      "epoch": 1.5199566523502552,
+      "grad_norm": 1.6061701774597168,
+      "learning_rate": 2.4667389127495745e-05,
+      "loss": 0.9802,
+      "step": 381500
+    },
+    {
+      "epoch": 1.5219487318421954,
+      "grad_norm": 1.5194026231765747,
+      "learning_rate": 2.4634187802630077e-05,
+      "loss": 0.9833,
+      "step": 382000
+    },
+    {
+      "epoch": 1.5239408113341355,
+      "grad_norm": 1.6393102407455444,
+      "learning_rate": 2.460098647776441e-05,
+      "loss": 0.981,
+      "step": 382500
+    },
+    {
+      "epoch": 1.5259328908260756,
+      "grad_norm": 1.563338279724121,
+      "learning_rate": 2.4567785152898742e-05,
+      "loss": 0.9814,
+      "step": 383000
+    },
+    {
+      "epoch": 1.5279249703180156,
+      "grad_norm": 1.5839451551437378,
+      "learning_rate": 2.4534583828033074e-05,
+      "loss": 0.9804,
+      "step": 383500
+    },
+    {
+      "epoch": 1.5299170498099555,
+      "grad_norm": 1.7216163873672485,
+      "learning_rate": 2.450138250316741e-05,
+      "loss": 0.9792,
+      "step": 384000
+    },
+    {
+      "epoch": 1.5319091293018956,
+      "grad_norm": 1.611944556236267,
+      "learning_rate": 2.4468181178301742e-05,
+      "loss": 0.9763,
+      "step": 384500
+    },
+    {
+      "epoch": 1.5339012087938357,
+      "grad_norm": 1.585035800933838,
+      "learning_rate": 2.443497985343607e-05,
+      "loss": 0.9779,
+      "step": 385000
+    },
+    {
+      "epoch": 1.5358932882857759,
+      "grad_norm": 1.6227517127990723,
+      "learning_rate": 2.4401778528570403e-05,
+      "loss": 0.9789,
+      "step": 385500
+    },
+    {
+      "epoch": 1.5378853677777158,
+      "grad_norm": 1.5471761226654053,
+      "learning_rate": 2.436857720370474e-05,
+      "loss": 0.9794,
+      "step": 386000
+    },
+    {
+      "epoch": 1.5398774472696557,
+      "grad_norm": 1.513098120689392,
+      "learning_rate": 2.433537587883907e-05,
+      "loss": 0.9826,
+      "step": 386500
+    },
+    {
+      "epoch": 1.5418695267615958,
+      "grad_norm": 1.5211966037750244,
+      "learning_rate": 2.4302174553973403e-05,
+      "loss": 0.9823,
+      "step": 387000
+    },
+    {
+      "epoch": 1.543861606253536,
+      "grad_norm": 1.5719540119171143,
+      "learning_rate": 2.4268973229107735e-05,
+      "loss": 0.9744,
+      "step": 387500
+    },
+    {
+      "epoch": 1.545853685745476,
+      "grad_norm": 1.751381516456604,
+      "learning_rate": 2.4235771904242067e-05,
+      "loss": 0.9742,
+      "step": 388000
+    },
+    {
+      "epoch": 1.547845765237416,
+      "grad_norm": 1.5535237789154053,
+      "learning_rate": 2.4202570579376403e-05,
+      "loss": 0.9759,
+      "step": 388500
+    },
+    {
+      "epoch": 1.549837844729356,
+      "grad_norm": 1.6061205863952637,
+      "learning_rate": 2.4169369254510732e-05,
+      "loss": 0.9753,
+      "step": 389000
+    },
+    {
+      "epoch": 1.551829924221296,
+      "grad_norm": 1.5686087608337402,
+      "learning_rate": 2.4136167929645064e-05,
+      "loss": 0.9758,
+      "step": 389500
+    },
+    {
+      "epoch": 1.5538220037132362,
+      "grad_norm": 1.520169973373413,
+      "learning_rate": 2.4102966604779396e-05,
+      "loss": 0.9741,
+      "step": 390000
+    },
+    {
+      "epoch": 1.5558140832051763,
+      "grad_norm": 1.6067639589309692,
+      "learning_rate": 2.4069765279913732e-05,
+      "loss": 0.9782,
+      "step": 390500
+    },
+    {
+      "epoch": 1.5578061626971162,
+      "grad_norm": 1.5374830961227417,
+      "learning_rate": 2.4036563955048064e-05,
+      "loss": 0.9774,
+      "step": 391000
+    },
+    {
+      "epoch": 1.5597982421890562,
+      "grad_norm": 1.6808209419250488,
+      "learning_rate": 2.4003362630182396e-05,
+      "loss": 0.9791,
+      "step": 391500
+    },
+    {
+      "epoch": 1.5617903216809963,
+      "grad_norm": 1.5406571626663208,
+      "learning_rate": 2.397016130531673e-05,
+      "loss": 0.9765,
+      "step": 392000
+    },
+    {
+      "epoch": 1.5637824011729364,
+      "grad_norm": 1.6367825269699097,
+      "learning_rate": 2.393695998045106e-05,
+      "loss": 0.9794,
+      "step": 392500
+    },
+    {
+      "epoch": 1.5657744806648766,
+      "grad_norm": 1.5081055164337158,
+      "learning_rate": 2.3903758655585393e-05,
+      "loss": 0.9736,
+      "step": 393000
+    },
+    {
+      "epoch": 1.5677665601568165,
+      "grad_norm": 1.5345408916473389,
+      "learning_rate": 2.3870557330719725e-05,
+      "loss": 0.9753,
+      "step": 393500
+    },
+    {
+      "epoch": 1.5697586396487564,
+      "grad_norm": 1.5150047540664673,
+      "learning_rate": 2.3837356005854058e-05,
+      "loss": 0.974,
+      "step": 394000
+    },
+    {
+      "epoch": 1.5717507191406965,
+      "grad_norm": 1.5795561075210571,
+      "learning_rate": 2.380415468098839e-05,
+      "loss": 0.9769,
+      "step": 394500
+    },
+    {
+      "epoch": 1.5737427986326367,
+      "grad_norm": 1.5274839401245117,
+      "learning_rate": 2.3770953356122725e-05,
+      "loss": 0.9768,
+      "step": 395000
+    },
+    {
+      "epoch": 1.5757348781245768,
+      "grad_norm": 1.5731112957000732,
+      "learning_rate": 2.3737752031257058e-05,
+      "loss": 0.9773,
+      "step": 395500
+    },
+    {
+      "epoch": 1.5777269576165167,
+      "grad_norm": 1.52711021900177,
+      "learning_rate": 2.370455070639139e-05,
+      "loss": 0.9785,
+      "step": 396000
+    },
+    {
+      "epoch": 1.5797190371084566,
+      "grad_norm": 1.5392756462097168,
+      "learning_rate": 2.367134938152572e-05,
+      "loss": 0.9749,
+      "step": 396500
+    },
+    {
+      "epoch": 1.5817111166003968,
+      "grad_norm": 1.520892858505249,
+      "learning_rate": 2.3638148056660054e-05,
+      "loss": 0.9756,
+      "step": 397000
+    },
+    {
+      "epoch": 1.583703196092337,
+      "grad_norm": 1.5895448923110962,
+      "learning_rate": 2.3604946731794387e-05,
+      "loss": 0.9776,
+      "step": 397500
+    },
+    {
+      "epoch": 1.585695275584277,
+      "grad_norm": 1.7636686563491821,
+      "learning_rate": 2.357174540692872e-05,
+      "loss": 0.9758,
+      "step": 398000
+    },
+    {
+      "epoch": 1.587687355076217,
+      "grad_norm": 1.5240639448165894,
+      "learning_rate": 2.353854408206305e-05,
+      "loss": 0.9729,
+      "step": 398500
+    },
+    {
+      "epoch": 1.5896794345681569,
+      "grad_norm": 1.4982706308364868,
+      "learning_rate": 2.3505342757197383e-05,
+      "loss": 0.9736,
+      "step": 399000
+    },
+    {
+      "epoch": 1.591671514060097,
+      "grad_norm": 1.6463743448257446,
+      "learning_rate": 2.347214143233172e-05,
+      "loss": 0.9695,
+      "step": 399500
+    },
+    {
+      "epoch": 1.5936635935520371,
+      "grad_norm": 1.6215156316757202,
+      "learning_rate": 2.3438940107466048e-05,
+      "loss": 0.9724,
+      "step": 400000
+    },
+    {
+      "epoch": 1.5956556730439773,
+      "grad_norm": 1.6194143295288086,
+      "learning_rate": 2.340573878260038e-05,
+      "loss": 0.9698,
+      "step": 400500
+    },
+    {
+      "epoch": 1.5976477525359172,
+      "grad_norm": 1.5803414583206177,
+      "learning_rate": 2.3372537457734716e-05,
+      "loss": 0.9726,
+      "step": 401000
+    },
+    {
+      "epoch": 1.599639832027857,
+      "grad_norm": 1.5295658111572266,
+      "learning_rate": 2.3339336132869048e-05,
+      "loss": 0.9721,
+      "step": 401500
+    },
+    {
+      "epoch": 1.6016319115197972,
+      "grad_norm": 1.7031586170196533,
+      "learning_rate": 2.330613480800338e-05,
+      "loss": 0.9708,
+      "step": 402000
+    },
+    {
+      "epoch": 1.6036239910117374,
+      "grad_norm": 1.6370879411697388,
+      "learning_rate": 2.3272933483137712e-05,
+      "loss": 0.9692,
+      "step": 402500
+    },
+    {
+      "epoch": 1.6056160705036775,
+      "grad_norm": 1.6956123113632202,
+      "learning_rate": 2.3239732158272045e-05,
+      "loss": 0.9756,
+      "step": 403000
+    },
+    {
+      "epoch": 1.6076081499956174,
+      "grad_norm": 1.568358063697815,
+      "learning_rate": 2.320653083340638e-05,
+      "loss": 0.9741,
+      "step": 403500
+    },
+    {
+      "epoch": 1.6096002294875573,
+      "grad_norm": 1.5767632722854614,
+      "learning_rate": 2.317332950854071e-05,
+      "loss": 0.9811,
+      "step": 404000
+    },
+    {
+      "epoch": 1.6115923089794975,
+      "grad_norm": 1.5962797403335571,
+      "learning_rate": 2.314012818367504e-05,
+      "loss": 0.9677,
+      "step": 404500
+    },
+    {
+      "epoch": 1.6135843884714376,
+      "grad_norm": 1.6314454078674316,
+      "learning_rate": 2.3106926858809373e-05,
+      "loss": 0.9745,
+      "step": 405000
+    },
+    {
+      "epoch": 1.6155764679633777,
+      "grad_norm": 1.5715343952178955,
+      "learning_rate": 2.307372553394371e-05,
+      "loss": 0.9749,
+      "step": 405500
+    },
+    {
+      "epoch": 1.6175685474553176,
+      "grad_norm": 1.6156021356582642,
+      "learning_rate": 2.304052420907804e-05,
+      "loss": 0.9724,
+      "step": 406000
+    },
+    {
+      "epoch": 1.6195606269472576,
+      "grad_norm": 1.5854839086532593,
+      "learning_rate": 2.3007322884212374e-05,
+      "loss": 0.971,
+      "step": 406500
+    },
+    {
+      "epoch": 1.6215527064391977,
+      "grad_norm": 1.605729579925537,
+      "learning_rate": 2.2974121559346706e-05,
+      "loss": 0.967,
+      "step": 407000
+    },
+    {
+      "epoch": 1.6235447859311378,
+      "grad_norm": 1.528767466545105,
+      "learning_rate": 2.2940920234481038e-05,
+      "loss": 0.9711,
+      "step": 407500
+    },
+    {
+      "epoch": 1.625536865423078,
+      "grad_norm": 1.5413762331008911,
+      "learning_rate": 2.290771890961537e-05,
+      "loss": 0.9735,
+      "step": 408000
+    },
+    {
+      "epoch": 1.6275289449150179,
+      "grad_norm": 1.5808117389678955,
+      "learning_rate": 2.2874517584749702e-05,
+      "loss": 0.9728,
+      "step": 408500
+    },
+    {
+      "epoch": 1.6295210244069578,
+      "grad_norm": 1.4834887981414795,
+      "learning_rate": 2.2841316259884035e-05,
+      "loss": 0.9767,
+      "step": 409000
+    },
+    {
+      "epoch": 1.631513103898898,
+      "grad_norm": 1.6740790605545044,
+      "learning_rate": 2.2808114935018367e-05,
+      "loss": 0.9744,
+      "step": 409500
+    },
+    {
+      "epoch": 1.633505183390838,
+      "grad_norm": 1.5696837902069092,
+      "learning_rate": 2.2774913610152703e-05,
+      "loss": 0.9741,
+      "step": 410000
+    },
+    {
+      "epoch": 1.6354972628827782,
+      "grad_norm": 1.5815658569335938,
+      "learning_rate": 2.2741712285287035e-05,
+      "loss": 0.9732,
+      "step": 410500
+    },
+    {
+      "epoch": 1.637489342374718,
+      "grad_norm": 1.6248801946640015,
+      "learning_rate": 2.2708510960421367e-05,
+      "loss": 0.9704,
+      "step": 411000
+    },
+    {
+      "epoch": 1.639481421866658,
+      "grad_norm": 1.5978702306747437,
+      "learning_rate": 2.2675309635555696e-05,
+      "loss": 0.9715,
+      "step": 411500
+    },
+    {
+      "epoch": 1.6414735013585982,
+      "grad_norm": 1.6880117654800415,
+      "learning_rate": 2.264210831069003e-05,
+      "loss": 0.9731,
+      "step": 412000
+    },
+    {
+      "epoch": 1.6434655808505383,
+      "grad_norm": 1.6841530799865723,
+      "learning_rate": 2.2608906985824364e-05,
+      "loss": 0.9691,
+      "step": 412500
+    },
+    {
+      "epoch": 1.6454576603424784,
+      "grad_norm": 1.569221019744873,
+      "learning_rate": 2.2575705660958696e-05,
+      "loss": 0.9711,
+      "step": 413000
+    },
+    {
+      "epoch": 1.6474497398344183,
+      "grad_norm": 1.7104644775390625,
+      "learning_rate": 2.2542504336093028e-05,
+      "loss": 0.9771,
+      "step": 413500
+    },
+    {
+      "epoch": 1.6494418193263582,
+      "grad_norm": 1.6464779376983643,
+      "learning_rate": 2.250930301122736e-05,
+      "loss": 0.9687,
+      "step": 414000
+    },
+    {
+      "epoch": 1.6514338988182984,
+      "grad_norm": 1.5476428270339966,
+      "learning_rate": 2.2476101686361696e-05,
+      "loss": 0.973,
+      "step": 414500
+    },
+    {
+      "epoch": 1.6534259783102385,
+      "grad_norm": 1.6614503860473633,
+      "learning_rate": 2.2442900361496028e-05,
+      "loss": 0.9677,
+      "step": 415000
+    },
+    {
+      "epoch": 1.6554180578021787,
+      "grad_norm": 1.6797555685043335,
+      "learning_rate": 2.2409699036630357e-05,
+      "loss": 0.9707,
+      "step": 415500
+    },
+    {
+      "epoch": 1.6574101372941186,
+      "grad_norm": 1.7716996669769287,
+      "learning_rate": 2.237649771176469e-05,
+      "loss": 0.9693,
+      "step": 416000
+    },
+    {
+      "epoch": 1.6594022167860585,
+      "grad_norm": 1.6168303489685059,
+      "learning_rate": 2.2343296386899025e-05,
+      "loss": 0.9688,
+      "step": 416500
+    },
+    {
+      "epoch": 1.6613942962779986,
+      "grad_norm": 1.6495788097381592,
+      "learning_rate": 2.2310095062033357e-05,
+      "loss": 0.9697,
+      "step": 417000
+    },
+    {
+      "epoch": 1.6633863757699388,
+      "grad_norm": 1.667330265045166,
+      "learning_rate": 2.227689373716769e-05,
+      "loss": 0.9703,
+      "step": 417500
+    },
+    {
+      "epoch": 1.6653784552618789,
+      "grad_norm": 1.6251211166381836,
+      "learning_rate": 2.224369241230202e-05,
+      "loss": 0.9683,
+      "step": 418000
+    },
+    {
+      "epoch": 1.6673705347538188,
+      "grad_norm": 1.5647958517074585,
+      "learning_rate": 2.2210491087436354e-05,
+      "loss": 0.9723,
+      "step": 418500
+    },
+    {
+      "epoch": 1.6693626142457587,
+      "grad_norm": 1.624349594116211,
+      "learning_rate": 2.2177289762570686e-05,
+      "loss": 0.9673,
+      "step": 419000
+    },
+    {
+      "epoch": 1.6713546937376988,
+      "grad_norm": 1.736549735069275,
+      "learning_rate": 2.214408843770502e-05,
+      "loss": 0.9733,
+      "step": 419500
+    },
+    {
+      "epoch": 1.673346773229639,
+      "grad_norm": 1.6103734970092773,
+      "learning_rate": 2.211088711283935e-05,
+      "loss": 0.9701,
+      "step": 420000
+    },
+    {
+      "epoch": 1.6753388527215791,
+      "grad_norm": 1.5810284614562988,
+      "learning_rate": 2.2077685787973683e-05,
+      "loss": 0.9696,
+      "step": 420500
+    },
+    {
+      "epoch": 1.677330932213519,
+      "grad_norm": 1.570515751838684,
+      "learning_rate": 2.204448446310802e-05,
+      "loss": 0.9644,
+      "step": 421000
+    },
+    {
+      "epoch": 1.679323011705459,
+      "grad_norm": 1.66584312915802,
+      "learning_rate": 2.201128313824235e-05,
+      "loss": 0.9695,
+      "step": 421500
+    },
+    {
+      "epoch": 1.681315091197399,
+      "grad_norm": 1.5663527250289917,
+      "learning_rate": 2.1978081813376683e-05,
+      "loss": 0.9655,
+      "step": 422000
+    },
+    {
+      "epoch": 1.6833071706893392,
+      "grad_norm": 1.656043529510498,
+      "learning_rate": 2.1944880488511015e-05,
+      "loss": 0.9716,
+      "step": 422500
+    },
+    {
+      "epoch": 1.6852992501812794,
+      "grad_norm": 1.638369083404541,
+      "learning_rate": 2.1911679163645347e-05,
+      "loss": 0.972,
+      "step": 423000
+    },
+    {
+      "epoch": 1.6872913296732193,
+      "grad_norm": 1.6632630825042725,
+      "learning_rate": 2.187847783877968e-05,
+      "loss": 0.9713,
+      "step": 423500
+    },
+    {
+      "epoch": 1.6892834091651592,
+      "grad_norm": 1.6088122129440308,
+      "learning_rate": 2.1845276513914012e-05,
+      "loss": 0.9646,
+      "step": 424000
+    },
+    {
+      "epoch": 1.6912754886570993,
+      "grad_norm": 1.5877692699432373,
+      "learning_rate": 2.1812075189048344e-05,
+      "loss": 0.9624,
+      "step": 424500
+    },
+    {
+      "epoch": 1.6932675681490394,
+      "grad_norm": 1.5771279335021973,
+      "learning_rate": 2.1778873864182676e-05,
+      "loss": 0.9712,
+      "step": 425000
+    },
+    {
+      "epoch": 1.6952596476409796,
+      "grad_norm": 1.5921114683151245,
+      "learning_rate": 2.1745672539317012e-05,
+      "loss": 0.9662,
+      "step": 425500
+    },
+    {
+      "epoch": 1.6972517271329195,
+      "grad_norm": 1.4950190782546997,
+      "learning_rate": 2.1712471214451344e-05,
+      "loss": 0.9647,
+      "step": 426000
+    },
+    {
+      "epoch": 1.6992438066248594,
+      "grad_norm": 1.5348923206329346,
+      "learning_rate": 2.1679269889585673e-05,
+      "loss": 0.9708,
+      "step": 426500
+    },
+    {
+      "epoch": 1.7012358861167995,
+      "grad_norm": 1.6691009998321533,
+      "learning_rate": 2.1646068564720005e-05,
+      "loss": 0.965,
+      "step": 427000
+    },
+    {
+      "epoch": 1.7032279656087397,
+      "grad_norm": 1.6985533237457275,
+      "learning_rate": 2.161286723985434e-05,
+      "loss": 0.9713,
+      "step": 427500
+    },
+    {
+      "epoch": 1.7052200451006798,
+      "grad_norm": 1.498036503791809,
+      "learning_rate": 2.1579665914988673e-05,
+      "loss": 0.9681,
+      "step": 428000
+    },
+    {
+      "epoch": 1.7072121245926197,
+      "grad_norm": 1.563021183013916,
+      "learning_rate": 2.1546464590123005e-05,
+      "loss": 0.9654,
+      "step": 428500
+    },
+    {
+      "epoch": 1.7092042040845596,
+      "grad_norm": 1.5391950607299805,
+      "learning_rate": 2.1513263265257337e-05,
+      "loss": 0.9658,
+      "step": 429000
+    },
+    {
+      "epoch": 1.7111962835764998,
+      "grad_norm": 1.5777959823608398,
+      "learning_rate": 2.1480061940391673e-05,
+      "loss": 0.969,
+      "step": 429500
+    },
+    {
+      "epoch": 1.71318836306844,
+      "grad_norm": 1.5898194313049316,
+      "learning_rate": 2.1446860615526005e-05,
+      "loss": 0.9635,
+      "step": 430000
+    },
+    {
+      "epoch": 1.71518044256038,
+      "grad_norm": 1.6636534929275513,
+      "learning_rate": 2.1413659290660334e-05,
+      "loss": 0.9666,
+      "step": 430500
+    },
+    {
+      "epoch": 1.71717252205232,
+      "grad_norm": 1.6843503713607788,
+      "learning_rate": 2.1380457965794666e-05,
+      "loss": 0.9616,
+      "step": 431000
+    },
+    {
+      "epoch": 1.7191646015442599,
+      "grad_norm": 1.664065957069397,
+      "learning_rate": 2.1347256640929002e-05,
+      "loss": 0.9664,
+      "step": 431500
+    },
+    {
+      "epoch": 1.7211566810362,
+      "grad_norm": 1.7058095932006836,
+      "learning_rate": 2.1314055316063334e-05,
+      "loss": 0.9708,
+      "step": 432000
+    },
+    {
+      "epoch": 1.7231487605281401,
+      "grad_norm": 1.55594003200531,
+      "learning_rate": 2.1280853991197666e-05,
+      "loss": 0.9667,
+      "step": 432500
+    },
+    {
+      "epoch": 1.7251408400200803,
+      "grad_norm": 1.5780198574066162,
+      "learning_rate": 2.1247652666332e-05,
+      "loss": 0.9618,
+      "step": 433000
+    },
+    {
+      "epoch": 1.7271329195120202,
+      "grad_norm": 1.650241732597351,
+      "learning_rate": 2.121445134146633e-05,
+      "loss": 0.9645,
+      "step": 433500
+    },
+    {
+      "epoch": 1.72912499900396,
+      "grad_norm": 1.5503370761871338,
+      "learning_rate": 2.1181250016600663e-05,
+      "loss": 0.9642,
+      "step": 434000
+    },
+    {
+      "epoch": 1.7311170784959002,
+      "grad_norm": 1.5153306722640991,
+      "learning_rate": 2.1148048691734995e-05,
+      "loss": 0.9705,
+      "step": 434500
+    },
+    {
+      "epoch": 1.7331091579878404,
+      "grad_norm": 1.6421469449996948,
+      "learning_rate": 2.1114847366869328e-05,
+      "loss": 0.9671,
+      "step": 435000
+    },
+    {
+      "epoch": 1.7351012374797805,
+      "grad_norm": 1.5826116800308228,
+      "learning_rate": 2.108164604200366e-05,
+      "loss": 0.9664,
+      "step": 435500
+    },
+    {
+      "epoch": 1.7370933169717204,
+      "grad_norm": 1.5960677862167358,
+      "learning_rate": 2.1048444717137995e-05,
+      "loss": 0.9701,
+      "step": 436000
+    },
+    {
+      "epoch": 1.7390853964636603,
+      "grad_norm": 1.5060874223709106,
+      "learning_rate": 2.1015243392272328e-05,
+      "loss": 0.968,
+      "step": 436500
+    },
+    {
+      "epoch": 1.7410774759556005,
+      "grad_norm": 1.5233402252197266,
+      "learning_rate": 2.098204206740666e-05,
+      "loss": 0.973,
+      "step": 437000
+    },
+    {
+      "epoch": 1.7430695554475406,
+      "grad_norm": 1.7586143016815186,
+      "learning_rate": 2.0948840742540992e-05,
+      "loss": 0.9689,
+      "step": 437500
+    },
+    {
+      "epoch": 1.7450616349394807,
+      "grad_norm": 1.6264207363128662,
+      "learning_rate": 2.0915639417675324e-05,
+      "loss": 0.9629,
+      "step": 438000
+    },
+    {
+      "epoch": 1.7470537144314207,
+      "grad_norm": 1.6368554830551147,
+      "learning_rate": 2.0882438092809657e-05,
+      "loss": 0.9649,
+      "step": 438500
+    },
+    {
+      "epoch": 1.7490457939233606,
+      "grad_norm": 1.648601770401001,
+      "learning_rate": 2.084923676794399e-05,
+      "loss": 0.9679,
+      "step": 439000
+    },
+    {
+      "epoch": 1.7510378734153007,
+      "grad_norm": 1.6422462463378906,
+      "learning_rate": 2.081603544307832e-05,
+      "loss": 0.9631,
+      "step": 439500
+    },
+    {
+      "epoch": 1.7530299529072408,
+      "grad_norm": 1.6329069137573242,
+      "learning_rate": 2.0782834118212653e-05,
+      "loss": 0.9624,
+      "step": 440000
+    },
+    {
+      "epoch": 1.755022032399181,
+      "grad_norm": 1.5675632953643799,
+      "learning_rate": 2.074963279334699e-05,
+      "loss": 0.9605,
+      "step": 440500
+    },
+    {
+      "epoch": 1.7570141118911209,
+      "grad_norm": 1.5548125505447388,
+      "learning_rate": 2.071643146848132e-05,
+      "loss": 0.9669,
+      "step": 441000
+    },
+    {
+      "epoch": 1.7590061913830608,
+      "grad_norm": 1.6204962730407715,
+      "learning_rate": 2.068323014361565e-05,
+      "loss": 0.9652,
+      "step": 441500
+    },
+    {
+      "epoch": 1.760998270875001,
+      "grad_norm": 1.6343318223953247,
+      "learning_rate": 2.0650028818749982e-05,
+      "loss": 0.9642,
+      "step": 442000
+    },
+    {
+      "epoch": 1.762990350366941,
+      "grad_norm": 1.6250325441360474,
+      "learning_rate": 2.0616827493884318e-05,
+      "loss": 0.9641,
+      "step": 442500
+    },
+    {
+      "epoch": 1.7649824298588812,
+      "grad_norm": 1.6133662462234497,
+      "learning_rate": 2.058362616901865e-05,
+      "loss": 0.9628,
+      "step": 443000
+    },
+    {
+      "epoch": 1.7669745093508211,
+      "grad_norm": 1.5805374383926392,
+      "learning_rate": 2.0550424844152982e-05,
+      "loss": 0.9591,
+      "step": 443500
+    },
+    {
+      "epoch": 1.768966588842761,
+      "grad_norm": 1.56540048122406,
+      "learning_rate": 2.0517223519287315e-05,
+      "loss": 0.9599,
+      "step": 444000
+    },
+    {
+      "epoch": 1.7709586683347012,
+      "grad_norm": 1.555434226989746,
+      "learning_rate": 2.0484022194421647e-05,
+      "loss": 0.9632,
+      "step": 444500
+    },
+    {
+      "epoch": 1.7729507478266413,
+      "grad_norm": 1.570823311805725,
+      "learning_rate": 2.0450820869555982e-05,
+      "loss": 0.9596,
+      "step": 445000
+    },
+    {
+      "epoch": 1.7749428273185814,
+      "grad_norm": 1.6308107376098633,
+      "learning_rate": 2.041761954469031e-05,
+      "loss": 0.9572,
+      "step": 445500
+    },
+    {
+      "epoch": 1.7769349068105214,
+      "grad_norm": 1.536621332168579,
+      "learning_rate": 2.0384418219824643e-05,
+      "loss": 0.9637,
+      "step": 446000
+    },
+    {
+      "epoch": 1.7789269863024613,
+      "grad_norm": 1.5725562572479248,
+      "learning_rate": 2.0351216894958976e-05,
+      "loss": 0.9644,
+      "step": 446500
+    },
+    {
+      "epoch": 1.7809190657944014,
+      "grad_norm": 1.6269222497940063,
+      "learning_rate": 2.031801557009331e-05,
+      "loss": 0.9642,
+      "step": 447000
+    },
+    {
+      "epoch": 1.7829111452863415,
+      "grad_norm": 1.4896668195724487,
+      "learning_rate": 2.0284814245227644e-05,
+      "loss": 0.9635,
+      "step": 447500
+    },
+    {
+      "epoch": 1.7849032247782817,
+      "grad_norm": 1.6548491716384888,
+      "learning_rate": 2.0251612920361976e-05,
+      "loss": 0.9637,
+      "step": 448000
+    },
+    {
+      "epoch": 1.7868953042702216,
+      "grad_norm": 1.5710760354995728,
+      "learning_rate": 2.0218411595496308e-05,
+      "loss": 0.9596,
+      "step": 448500
+    },
+    {
+      "epoch": 1.7888873837621615,
+      "grad_norm": 1.679811954498291,
+      "learning_rate": 2.018521027063064e-05,
+      "loss": 0.962,
+      "step": 449000
+    },
+    {
+      "epoch": 1.7908794632541016,
+      "grad_norm": 1.6168251037597656,
+      "learning_rate": 2.0152008945764972e-05,
+      "loss": 0.9616,
+      "step": 449500
+    },
+    {
+      "epoch": 1.7928715427460418,
+      "grad_norm": 1.6792786121368408,
+      "learning_rate": 2.0118807620899305e-05,
+      "loss": 0.9621,
+      "step": 450000
+    },
+    {
+      "epoch": 1.794863622237982,
+      "grad_norm": 1.604726791381836,
+      "learning_rate": 2.0085606296033637e-05,
+      "loss": 0.9616,
+      "step": 450500
+    },
+    {
+      "epoch": 1.7968557017299218,
+      "grad_norm": 1.6965233087539673,
+      "learning_rate": 2.005240497116797e-05,
+      "loss": 0.9593,
+      "step": 451000
+    },
+    {
+      "epoch": 1.7988477812218617,
+      "grad_norm": 1.4541702270507812,
+      "learning_rate": 2.0019203646302305e-05,
+      "loss": 0.9611,
+      "step": 451500
+    },
+    {
+      "epoch": 1.8008398607138019,
+      "grad_norm": 1.5673445463180542,
+      "learning_rate": 1.9986002321436637e-05,
+      "loss": 0.9602,
+      "step": 452000
+    },
+    {
+      "epoch": 1.802831940205742,
+      "grad_norm": 1.534553050994873,
+      "learning_rate": 1.995280099657097e-05,
+      "loss": 0.9571,
+      "step": 452500
+    },
+    {
+      "epoch": 1.8048240196976821,
+      "grad_norm": 1.6114728450775146,
+      "learning_rate": 1.9919599671705298e-05,
+      "loss": 0.9645,
+      "step": 453000
+    },
+    {
+      "epoch": 1.806816099189622,
+      "grad_norm": 1.5884662866592407,
+      "learning_rate": 1.9886398346839634e-05,
+      "loss": 0.9613,
+      "step": 453500
+    },
+    {
+      "epoch": 1.808808178681562,
+      "grad_norm": 1.635599136352539,
+      "learning_rate": 1.9853197021973966e-05,
+      "loss": 0.9604,
+      "step": 454000
+    },
+    {
+      "epoch": 1.810800258173502,
+      "grad_norm": 1.5859266519546509,
+      "learning_rate": 1.9819995697108298e-05,
+      "loss": 0.9568,
+      "step": 454500
+    },
+    {
+      "epoch": 1.8127923376654422,
+      "grad_norm": 1.6607033014297485,
+      "learning_rate": 1.978679437224263e-05,
+      "loss": 0.958,
+      "step": 455000
+    },
+    {
+      "epoch": 1.8147844171573824,
+      "grad_norm": 1.6127053499221802,
+      "learning_rate": 1.9753593047376963e-05,
+      "loss": 0.9576,
+      "step": 455500
+    },
+    {
+      "epoch": 1.8167764966493223,
+      "grad_norm": 1.5654854774475098,
+      "learning_rate": 1.9720391722511298e-05,
+      "loss": 0.961,
+      "step": 456000
+    },
+    {
+      "epoch": 1.8187685761412622,
+      "grad_norm": 1.8537358045578003,
+      "learning_rate": 1.9687190397645627e-05,
+      "loss": 0.9643,
+      "step": 456500
+    },
+    {
+      "epoch": 1.8207606556332023,
+      "grad_norm": 1.5898334980010986,
+      "learning_rate": 1.965398907277996e-05,
+      "loss": 0.9598,
+      "step": 457000
+    },
+    {
+      "epoch": 1.8227527351251425,
+      "grad_norm": 1.6024376153945923,
+      "learning_rate": 1.962078774791429e-05,
+      "loss": 0.9585,
+      "step": 457500
+    },
+    {
+      "epoch": 1.8247448146170826,
+      "grad_norm": 1.5592955350875854,
+      "learning_rate": 1.9587586423048627e-05,
+      "loss": 0.9597,
+      "step": 458000
+    },
+    {
+      "epoch": 1.8267368941090225,
+      "grad_norm": 1.57802152633667,
+      "learning_rate": 1.955438509818296e-05,
+      "loss": 0.9579,
+      "step": 458500
+    },
+    {
+      "epoch": 1.8287289736009624,
+      "grad_norm": 1.6363270282745361,
+      "learning_rate": 1.952118377331729e-05,
+      "loss": 0.9615,
+      "step": 459000
+    },
+    {
+      "epoch": 1.8307210530929026,
+      "grad_norm": 1.699605107307434,
+      "learning_rate": 1.9487982448451624e-05,
+      "loss": 0.9622,
+      "step": 459500
+    },
+    {
+      "epoch": 1.8327131325848427,
+      "grad_norm": 1.573757290840149,
+      "learning_rate": 1.945478112358596e-05,
+      "loss": 0.961,
+      "step": 460000
+    },
+    {
+      "epoch": 1.8347052120767828,
+      "grad_norm": 1.5768842697143555,
+      "learning_rate": 1.942157979872029e-05,
+      "loss": 0.9603,
+      "step": 460500
+    },
+    {
+      "epoch": 1.8366972915687227,
+      "grad_norm": 1.681291937828064,
+      "learning_rate": 1.938837847385462e-05,
+      "loss": 0.9572,
+      "step": 461000
+    },
+    {
+      "epoch": 1.8386893710606627,
+      "grad_norm": 1.6499990224838257,
+      "learning_rate": 1.9355177148988953e-05,
+      "loss": 0.9616,
+      "step": 461500
+    },
+    {
+      "epoch": 1.8406814505526028,
+      "grad_norm": 1.5527372360229492,
+      "learning_rate": 1.932197582412329e-05,
+      "loss": 0.9589,
+      "step": 462000
+    },
+    {
+      "epoch": 1.842673530044543,
+      "grad_norm": 1.6306767463684082,
+      "learning_rate": 1.928877449925762e-05,
+      "loss": 0.9582,
+      "step": 462500
+    },
+    {
+      "epoch": 1.844665609536483,
+      "grad_norm": 1.5439594984054565,
+      "learning_rate": 1.9255573174391953e-05,
+      "loss": 0.9586,
+      "step": 463000
+    },
+    {
+      "epoch": 1.846657689028423,
+      "grad_norm": 1.5823874473571777,
+      "learning_rate": 1.9222371849526285e-05,
+      "loss": 0.9573,
+      "step": 463500
+    },
+    {
+      "epoch": 1.848649768520363,
+      "grad_norm": 1.624509334564209,
+      "learning_rate": 1.9189170524660617e-05,
+      "loss": 0.9564,
+      "step": 464000
+    },
+    {
+      "epoch": 1.850641848012303,
+      "grad_norm": 1.6275229454040527,
+      "learning_rate": 1.915596919979495e-05,
+      "loss": 0.9557,
+      "step": 464500
+    },
+    {
+      "epoch": 1.8526339275042432,
+      "grad_norm": 1.6211669445037842,
+      "learning_rate": 1.9122767874929282e-05,
+      "loss": 0.9596,
+      "step": 465000
+    },
+    {
+      "epoch": 1.8546260069961833,
+      "grad_norm": 1.7692539691925049,
+      "learning_rate": 1.9089566550063614e-05,
+      "loss": 0.9587,
+      "step": 465500
+    },
+    {
+      "epoch": 1.8566180864881232,
+      "grad_norm": 1.5504094362258911,
+      "learning_rate": 1.9056365225197946e-05,
+      "loss": 0.9566,
+      "step": 466000
+    },
+    {
+      "epoch": 1.8586101659800631,
+      "grad_norm": 1.6282809972763062,
+      "learning_rate": 1.9023163900332282e-05,
+      "loss": 0.9579,
+      "step": 466500
+    },
+    {
+      "epoch": 1.8606022454720033,
+      "grad_norm": 1.5149763822555542,
+      "learning_rate": 1.8989962575466614e-05,
+      "loss": 0.9601,
+      "step": 467000
+    },
+    {
+      "epoch": 1.8625943249639434,
+      "grad_norm": 1.5261211395263672,
+      "learning_rate": 1.8956761250600946e-05,
+      "loss": 0.9584,
+      "step": 467500
+    },
+    {
+      "epoch": 1.8645864044558835,
+      "grad_norm": 1.616898536682129,
+      "learning_rate": 1.8923559925735275e-05,
+      "loss": 0.9568,
+      "step": 468000
+    },
+    {
+      "epoch": 1.8665784839478234,
+      "grad_norm": 1.6234068870544434,
+      "learning_rate": 1.889035860086961e-05,
+      "loss": 0.9513,
+      "step": 468500
+    },
+    {
+      "epoch": 1.8685705634397634,
+      "grad_norm": 1.6470742225646973,
+      "learning_rate": 1.8857157276003943e-05,
+      "loss": 0.9579,
+      "step": 469000
+    },
+    {
+      "epoch": 1.8705626429317035,
+      "grad_norm": 1.5634570121765137,
+      "learning_rate": 1.8823955951138275e-05,
+      "loss": 0.9565,
+      "step": 469500
+    },
+    {
+      "epoch": 1.8725547224236436,
+      "grad_norm": 1.5136973857879639,
+      "learning_rate": 1.8790754626272607e-05,
+      "loss": 0.9556,
+      "step": 470000
+    },
+    {
+      "epoch": 1.8745468019155838,
+      "grad_norm": 1.6878010034561157,
+      "learning_rate": 1.875755330140694e-05,
+      "loss": 0.9545,
+      "step": 470500
+    },
+    {
+      "epoch": 1.8765388814075237,
+      "grad_norm": 1.588257074356079,
+      "learning_rate": 1.8724351976541275e-05,
+      "loss": 0.9572,
+      "step": 471000
+    },
+    {
+      "epoch": 1.8785309608994636,
+      "grad_norm": 1.6008977890014648,
+      "learning_rate": 1.8691150651675604e-05,
+      "loss": 0.9518,
+      "step": 471500
+    },
+    {
+      "epoch": 1.8805230403914037,
+      "grad_norm": 1.7102978229522705,
+      "learning_rate": 1.8657949326809936e-05,
+      "loss": 0.956,
+      "step": 472000
+    },
+    {
+      "epoch": 1.8825151198833439,
+      "grad_norm": 1.5778944492340088,
+      "learning_rate": 1.862474800194427e-05,
+      "loss": 0.9533,
+      "step": 472500
+    },
+    {
+      "epoch": 1.884507199375284,
+      "grad_norm": 1.668061375617981,
+      "learning_rate": 1.8591546677078604e-05,
+      "loss": 0.9568,
+      "step": 473000
+    },
+    {
+      "epoch": 1.886499278867224,
+      "grad_norm": 1.66815185546875,
+      "learning_rate": 1.8558345352212936e-05,
+      "loss": 0.9573,
+      "step": 473500
+    },
+    {
+      "epoch": 1.8884913583591638,
+      "grad_norm": 1.5782902240753174,
+      "learning_rate": 1.852514402734727e-05,
+      "loss": 0.9542,
+      "step": 474000
+    },
+    {
+      "epoch": 1.890483437851104,
+      "grad_norm": 1.5876054763793945,
+      "learning_rate": 1.84919427024816e-05,
+      "loss": 0.9619,
+      "step": 474500
+    },
+    {
+      "epoch": 1.892475517343044,
+      "grad_norm": 1.6155978441238403,
+      "learning_rate": 1.8458741377615933e-05,
+      "loss": 0.9606,
+      "step": 475000
+    },
+    {
+      "epoch": 1.8944675968349842,
+      "grad_norm": 1.5506558418273926,
+      "learning_rate": 1.8425540052750265e-05,
+      "loss": 0.9532,
+      "step": 475500
+    },
+    {
+      "epoch": 1.8964596763269241,
+      "grad_norm": 1.5561200380325317,
+      "learning_rate": 1.8392338727884598e-05,
+      "loss": 0.9595,
+      "step": 476000
+    },
+    {
+      "epoch": 1.898451755818864,
+      "grad_norm": 1.6783262491226196,
+      "learning_rate": 1.835913740301893e-05,
+      "loss": 0.9626,
+      "step": 476500
+    },
+    {
+      "epoch": 1.9004438353108042,
+      "grad_norm": 1.6589523553848267,
+      "learning_rate": 1.8325936078153262e-05,
+      "loss": 0.9552,
+      "step": 477000
+    },
+    {
+      "epoch": 1.9024359148027443,
+      "grad_norm": 1.5693058967590332,
+      "learning_rate": 1.8292734753287598e-05,
+      "loss": 0.9532,
+      "step": 477500
+    },
+    {
+      "epoch": 1.9044279942946845,
+      "grad_norm": 1.60469651222229,
+      "learning_rate": 1.825953342842193e-05,
+      "loss": 0.9594,
+      "step": 478000
+    },
+    {
+      "epoch": 1.9064200737866244,
+      "grad_norm": 1.5541704893112183,
+      "learning_rate": 1.8226332103556262e-05,
+      "loss": 0.953,
+      "step": 478500
+    },
+    {
+      "epoch": 1.9084121532785643,
+      "grad_norm": 1.6770862340927124,
+      "learning_rate": 1.819313077869059e-05,
+      "loss": 0.9594,
+      "step": 479000
+    },
+    {
+      "epoch": 1.9104042327705044,
+      "grad_norm": 1.64461350440979,
+      "learning_rate": 1.8159929453824927e-05,
+      "loss": 0.9571,
+      "step": 479500
+    },
+    {
+      "epoch": 1.9123963122624446,
+      "grad_norm": 1.7355711460113525,
+      "learning_rate": 1.812672812895926e-05,
+      "loss": 0.9546,
+      "step": 480000
+    },
+    {
+      "epoch": 1.9143883917543847,
+      "grad_norm": 1.5793958902359009,
+      "learning_rate": 1.809352680409359e-05,
+      "loss": 0.9543,
+      "step": 480500
+    },
+    {
+      "epoch": 1.9163804712463246,
+      "grad_norm": 1.613983392715454,
+      "learning_rate": 1.8060325479227923e-05,
+      "loss": 0.9522,
+      "step": 481000
+    },
+    {
+      "epoch": 1.9183725507382645,
+      "grad_norm": 1.5817432403564453,
+      "learning_rate": 1.8027124154362256e-05,
+      "loss": 0.9514,
+      "step": 481500
+    },
+    {
+      "epoch": 1.9203646302302047,
+      "grad_norm": 1.719308614730835,
+      "learning_rate": 1.799392282949659e-05,
+      "loss": 0.9533,
+      "step": 482000
+    },
+    {
+      "epoch": 1.9223567097221448,
+      "grad_norm": 1.599969744682312,
+      "learning_rate": 1.7960721504630923e-05,
+      "loss": 0.9496,
+      "step": 482500
+    },
+    {
+      "epoch": 1.924348789214085,
+      "grad_norm": 1.6354597806930542,
+      "learning_rate": 1.7927520179765252e-05,
+      "loss": 0.9567,
+      "step": 483000
+    },
+    {
+      "epoch": 1.9263408687060248,
+      "grad_norm": 1.6391348838806152,
+      "learning_rate": 1.7894318854899584e-05,
+      "loss": 0.9592,
+      "step": 483500
+    },
+    {
+      "epoch": 1.9283329481979647,
+      "grad_norm": 1.6589632034301758,
+      "learning_rate": 1.786111753003392e-05,
+      "loss": 0.9507,
+      "step": 484000
+    },
+    {
+      "epoch": 1.9303250276899049,
+      "grad_norm": 1.595211386680603,
+      "learning_rate": 1.7827916205168252e-05,
+      "loss": 0.9571,
+      "step": 484500
+    },
+    {
+      "epoch": 1.932317107181845,
+      "grad_norm": 1.587093472480774,
+      "learning_rate": 1.7794714880302585e-05,
+      "loss": 0.9529,
+      "step": 485000
+    },
+    {
+      "epoch": 1.9343091866737852,
+      "grad_norm": 1.666335940361023,
+      "learning_rate": 1.7761513555436917e-05,
+      "loss": 0.9559,
+      "step": 485500
+    },
+    {
+      "epoch": 1.936301266165725,
+      "grad_norm": 1.5669444799423218,
+      "learning_rate": 1.772831223057125e-05,
+      "loss": 0.9529,
+      "step": 486000
+    },
+    {
+      "epoch": 1.938293345657665,
+      "grad_norm": 1.6388435363769531,
+      "learning_rate": 1.769511090570558e-05,
+      "loss": 0.9529,
+      "step": 486500
+    },
+    {
+      "epoch": 1.9402854251496051,
+      "grad_norm": 1.6335638761520386,
+      "learning_rate": 1.7661909580839913e-05,
+      "loss": 0.9546,
+      "step": 487000
+    },
+    {
+      "epoch": 1.9422775046415452,
+      "grad_norm": 1.6439802646636963,
+      "learning_rate": 1.7628708255974246e-05,
+      "loss": 0.9593,
+      "step": 487500
+    },
+    {
+      "epoch": 1.9442695841334854,
+      "grad_norm": 1.6448768377304077,
+      "learning_rate": 1.7595506931108578e-05,
+      "loss": 0.9561,
+      "step": 488000
+    },
+    {
+      "epoch": 1.9462616636254253,
+      "grad_norm": 1.5907398462295532,
+      "learning_rate": 1.7562305606242914e-05,
+      "loss": 0.9556,
+      "step": 488500
+    },
+    {
+      "epoch": 1.9482537431173652,
+      "grad_norm": 1.6046091318130493,
+      "learning_rate": 1.7529104281377246e-05,
+      "loss": 0.9504,
+      "step": 489000
+    },
+    {
+      "epoch": 1.9502458226093053,
+      "grad_norm": 1.6411865949630737,
+      "learning_rate": 1.7495902956511578e-05,
+      "loss": 0.9546,
+      "step": 489500
+    },
+    {
+      "epoch": 1.9522379021012455,
+      "grad_norm": 1.5086331367492676,
+      "learning_rate": 1.746270163164591e-05,
+      "loss": 0.9497,
+      "step": 490000
+    },
+    {
+      "epoch": 1.9542299815931856,
+      "grad_norm": 1.5817451477050781,
+      "learning_rate": 1.7429500306780242e-05,
+      "loss": 0.9517,
+      "step": 490500
+    },
+    {
+      "epoch": 1.9562220610851255,
+      "grad_norm": 1.6611523628234863,
+      "learning_rate": 1.7396298981914575e-05,
+      "loss": 0.9518,
+      "step": 491000
+    },
+    {
+      "epoch": 1.9582141405770654,
+      "grad_norm": 1.708668828010559,
+      "learning_rate": 1.7363097657048907e-05,
+      "loss": 0.956,
+      "step": 491500
+    },
+    {
+      "epoch": 1.9602062200690056,
+      "grad_norm": 1.6335152387619019,
+      "learning_rate": 1.732989633218324e-05,
+      "loss": 0.9459,
+      "step": 492000
+    },
+    {
+      "epoch": 1.9621982995609457,
+      "grad_norm": 1.6695103645324707,
+      "learning_rate": 1.7296695007317575e-05,
+      "loss": 0.9497,
+      "step": 492500
+    },
+    {
+      "epoch": 1.9641903790528858,
+      "grad_norm": 1.5089513063430786,
+      "learning_rate": 1.7263493682451907e-05,
+      "loss": 0.9534,
+      "step": 493000
+    },
+    {
+      "epoch": 1.9661824585448258,
+      "grad_norm": 1.6098097562789917,
+      "learning_rate": 1.723029235758624e-05,
+      "loss": 0.9517,
+      "step": 493500
+    },
+    {
+      "epoch": 1.9681745380367657,
+      "grad_norm": 1.707329273223877,
+      "learning_rate": 1.7197091032720568e-05,
+      "loss": 0.9501,
+      "step": 494000
+    },
+    {
+      "epoch": 1.9701666175287058,
+      "grad_norm": 1.5940375328063965,
+      "learning_rate": 1.7163889707854904e-05,
+      "loss": 0.9539,
+      "step": 494500
+    },
+    {
+      "epoch": 1.972158697020646,
+      "grad_norm": 1.602586269378662,
+      "learning_rate": 1.7130688382989236e-05,
+      "loss": 0.956,
+      "step": 495000
+    },
+    {
+      "epoch": 1.974150776512586,
+      "grad_norm": 1.6518926620483398,
+      "learning_rate": 1.7097487058123568e-05,
+      "loss": 0.9477,
+      "step": 495500
+    },
+    {
+      "epoch": 1.976142856004526,
+      "grad_norm": 1.679147720336914,
+      "learning_rate": 1.70642857332579e-05,
+      "loss": 0.9489,
+      "step": 496000
+    },
+    {
+      "epoch": 1.978134935496466,
+      "grad_norm": 1.6897059679031372,
+      "learning_rate": 1.7031084408392233e-05,
+      "loss": 0.9494,
+      "step": 496500
+    },
+    {
+      "epoch": 1.980127014988406,
+      "grad_norm": 1.5620797872543335,
+      "learning_rate": 1.6997883083526568e-05,
+      "loss": 0.9455,
+      "step": 497000
+    },
+    {
+      "epoch": 1.9821190944803462,
+      "grad_norm": 1.6151255369186401,
+      "learning_rate": 1.69646817586609e-05,
+      "loss": 0.9543,
+      "step": 497500
+    },
+    {
+      "epoch": 1.9841111739722863,
+      "grad_norm": 1.5484846830368042,
+      "learning_rate": 1.693148043379523e-05,
+      "loss": 0.954,
+      "step": 498000
+    },
+    {
+      "epoch": 1.9861032534642262,
+      "grad_norm": 1.518297553062439,
+      "learning_rate": 1.689827910892956e-05,
+      "loss": 0.95,
+      "step": 498500
+    },
+    {
+      "epoch": 1.9880953329561661,
+      "grad_norm": 1.5891717672348022,
+      "learning_rate": 1.6865077784063897e-05,
+      "loss": 0.9533,
+      "step": 499000
+    },
+    {
+      "epoch": 1.9900874124481063,
+      "grad_norm": 1.697347640991211,
+      "learning_rate": 1.683187645919823e-05,
+      "loss": 0.9481,
+      "step": 499500
+    },
+    {
+      "epoch": 1.9920794919400464,
+      "grad_norm": 1.6892317533493042,
+      "learning_rate": 1.679867513433256e-05,
+      "loss": 0.9522,
+      "step": 500000
+    },
+    {
+      "epoch": 1.9940715714319865,
+      "grad_norm": 1.5398199558258057,
+      "learning_rate": 1.6765473809466894e-05,
+      "loss": 0.951,
+      "step": 500500
+    },
+    {
+      "epoch": 1.9960636509239265,
+      "grad_norm": 1.6792997121810913,
+      "learning_rate": 1.6732272484601226e-05,
+      "loss": 0.9479,
+      "step": 501000
+    },
+    {
+      "epoch": 1.9980557304158664,
+      "grad_norm": 1.5564945936203003,
+      "learning_rate": 1.669907115973556e-05,
+      "loss": 0.9497,
+      "step": 501500
+    },
+    {
+      "epoch": 2.0000478099078065,
+      "grad_norm": 1.6831318140029907,
+      "learning_rate": 1.666586983486989e-05,
+      "loss": 0.9457,
+      "step": 502000
+    },
+    {
+      "epoch": 2.0020398893997466,
+      "grad_norm": 1.609836220741272,
+      "learning_rate": 1.6632668510004223e-05,
+      "loss": 0.9508,
+      "step": 502500
+    },
+    {
+      "epoch": 2.0040319688916868,
+      "grad_norm": 1.5143826007843018,
+      "learning_rate": 1.6599467185138555e-05,
+      "loss": 0.9508,
+      "step": 503000
+    },
+    {
+      "epoch": 2.006024048383627,
+      "grad_norm": 1.6649590730667114,
+      "learning_rate": 1.656626586027289e-05,
+      "loss": 0.9482,
+      "step": 503500
+    },
+    {
+      "epoch": 2.0080161278755666,
+      "grad_norm": 1.6936696767807007,
+      "learning_rate": 1.6533064535407223e-05,
+      "loss": 0.9448,
+      "step": 504000
+    },
+    {
+      "epoch": 2.0100082073675067,
+      "grad_norm": 1.7341660261154175,
+      "learning_rate": 1.6499863210541555e-05,
+      "loss": 0.9473,
+      "step": 504500
+    },
+    {
+      "epoch": 2.012000286859447,
+      "grad_norm": 1.5983432531356812,
+      "learning_rate": 1.6466661885675887e-05,
+      "loss": 0.9478,
+      "step": 505000
+    },
+    {
+      "epoch": 2.013992366351387,
+      "grad_norm": 1.5848313570022583,
+      "learning_rate": 1.643346056081022e-05,
+      "loss": 0.9467,
+      "step": 505500
+    },
+    {
+      "epoch": 2.015984445843327,
+      "grad_norm": 1.6222004890441895,
+      "learning_rate": 1.6400259235944552e-05,
+      "loss": 0.9449,
+      "step": 506000
+    },
+    {
+      "epoch": 2.017976525335267,
+      "grad_norm": 1.6453777551651,
+      "learning_rate": 1.6367057911078884e-05,
+      "loss": 0.9487,
+      "step": 506500
+    },
+    {
+      "epoch": 2.019968604827207,
+      "grad_norm": 1.7078001499176025,
+      "learning_rate": 1.6333856586213216e-05,
+      "loss": 0.9507,
+      "step": 507000
+    },
+    {
+      "epoch": 2.021960684319147,
+      "grad_norm": 1.6671719551086426,
+      "learning_rate": 1.630065526134755e-05,
+      "loss": 0.9512,
+      "step": 507500
+    },
+    {
+      "epoch": 2.0239527638110872,
+      "grad_norm": 1.6074438095092773,
+      "learning_rate": 1.6267453936481884e-05,
+      "loss": 0.9423,
+      "step": 508000
+    },
+    {
+      "epoch": 2.0259448433030274,
+      "grad_norm": 1.6333234310150146,
+      "learning_rate": 1.6234252611616216e-05,
+      "loss": 0.951,
+      "step": 508500
+    },
+    {
+      "epoch": 2.027936922794967,
+      "grad_norm": 1.6678357124328613,
+      "learning_rate": 1.6201051286750545e-05,
+      "loss": 0.9461,
+      "step": 509000
+    },
+    {
+      "epoch": 2.029929002286907,
+      "grad_norm": 1.5896990299224854,
+      "learning_rate": 1.6167849961884877e-05,
+      "loss": 0.9451,
+      "step": 509500
+    },
+    {
+      "epoch": 2.0319210817788473,
+      "grad_norm": 1.6058069467544556,
+      "learning_rate": 1.6134648637019213e-05,
+      "loss": 0.9496,
+      "step": 510000
+    },
+    {
+      "epoch": 2.0339131612707875,
+      "grad_norm": 1.707418441772461,
+      "learning_rate": 1.6101447312153545e-05,
+      "loss": 0.9421,
+      "step": 510500
+    },
+    {
+      "epoch": 2.0359052407627276,
+      "grad_norm": 1.6255860328674316,
+      "learning_rate": 1.6068245987287877e-05,
+      "loss": 0.9483,
+      "step": 511000
+    },
+    {
+      "epoch": 2.0378973202546673,
+      "grad_norm": 1.6119027137756348,
+      "learning_rate": 1.603504466242221e-05,
+      "loss": 0.947,
+      "step": 511500
+    },
+    {
+      "epoch": 2.0398893997466074,
+      "grad_norm": 1.6432864665985107,
+      "learning_rate": 1.6001843337556542e-05,
+      "loss": 0.9476,
+      "step": 512000
+    },
+    {
+      "epoch": 2.0418814792385476,
+      "grad_norm": 1.7091964483261108,
+      "learning_rate": 1.5968642012690878e-05,
+      "loss": 0.9456,
+      "step": 512500
+    },
+    {
+      "epoch": 2.0438735587304877,
+      "grad_norm": 1.5329601764678955,
+      "learning_rate": 1.5935440687825206e-05,
+      "loss": 0.9444,
+      "step": 513000
+    },
+    {
+      "epoch": 2.045865638222428,
+      "grad_norm": 1.580696940422058,
+      "learning_rate": 1.590223936295954e-05,
+      "loss": 0.9507,
+      "step": 513500
+    },
+    {
+      "epoch": 2.0478577177143675,
+      "grad_norm": 1.6027048826217651,
+      "learning_rate": 1.586903803809387e-05,
+      "loss": 0.945,
+      "step": 514000
+    },
+    {
+      "epoch": 2.0498497972063077,
+      "grad_norm": 1.6171530485153198,
+      "learning_rate": 1.5835836713228207e-05,
+      "loss": 0.9473,
+      "step": 514500
+    },
+    {
+      "epoch": 2.051841876698248,
+      "grad_norm": 1.6771458387374878,
+      "learning_rate": 1.580263538836254e-05,
+      "loss": 0.9496,
+      "step": 515000
+    },
+    {
+      "epoch": 2.053833956190188,
+      "grad_norm": 1.6373488903045654,
+      "learning_rate": 1.576943406349687e-05,
+      "loss": 0.9463,
+      "step": 515500
+    },
+    {
+      "epoch": 2.055826035682128,
+      "grad_norm": 1.5845364332199097,
+      "learning_rate": 1.5736232738631203e-05,
+      "loss": 0.9471,
+      "step": 516000
+    },
+    {
+      "epoch": 2.0578181151740678,
+      "grad_norm": 1.6744481325149536,
+      "learning_rate": 1.5703031413765535e-05,
+      "loss": 0.9445,
+      "step": 516500
+    },
+    {
+      "epoch": 2.059810194666008,
+      "grad_norm": 1.7394870519638062,
+      "learning_rate": 1.5669830088899868e-05,
+      "loss": 0.9494,
+      "step": 517000
+    },
+    {
+      "epoch": 2.061802274157948,
+      "grad_norm": 1.6003581285476685,
+      "learning_rate": 1.56366287640342e-05,
+      "loss": 0.9444,
+      "step": 517500
+    },
+    {
+      "epoch": 2.063794353649888,
+      "grad_norm": 1.6030445098876953,
+      "learning_rate": 1.5603427439168532e-05,
+      "loss": 0.9506,
+      "step": 518000
+    },
+    {
+      "epoch": 2.0657864331418283,
+      "grad_norm": 1.6108741760253906,
+      "learning_rate": 1.5570226114302868e-05,
+      "loss": 0.9445,
+      "step": 518500
+    },
+    {
+      "epoch": 2.067778512633768,
+      "grad_norm": 1.6967995166778564,
+      "learning_rate": 1.55370247894372e-05,
+      "loss": 0.9419,
+      "step": 519000
+    },
+    {
+      "epoch": 2.069770592125708,
+      "grad_norm": 1.5767747163772583,
+      "learning_rate": 1.5503823464571532e-05,
+      "loss": 0.944,
+      "step": 519500
+    },
+    {
+      "epoch": 2.0717626716176483,
+      "grad_norm": 1.5738168954849243,
+      "learning_rate": 1.5470622139705864e-05,
+      "loss": 0.949,
+      "step": 520000
+    },
+    {
+      "epoch": 2.0737547511095884,
+      "grad_norm": 1.6936604976654053,
+      "learning_rate": 1.5437420814840197e-05,
+      "loss": 0.9452,
+      "step": 520500
+    },
+    {
+      "epoch": 2.0757468306015285,
+      "grad_norm": 1.6442201137542725,
+      "learning_rate": 1.540421948997453e-05,
+      "loss": 0.9478,
+      "step": 521000
+    },
+    {
+      "epoch": 2.0777389100934682,
+      "grad_norm": 1.6269065141677856,
+      "learning_rate": 1.537101816510886e-05,
+      "loss": 0.9515,
+      "step": 521500
+    },
+    {
+      "epoch": 2.0797309895854084,
+      "grad_norm": 1.637317419052124,
+      "learning_rate": 1.5337816840243193e-05,
+      "loss": 0.9443,
+      "step": 522000
+    },
+    {
+      "epoch": 2.0817230690773485,
+      "grad_norm": 1.615427017211914,
+      "learning_rate": 1.5304615515377526e-05,
+      "loss": 0.9417,
+      "step": 522500
+    },
+    {
+      "epoch": 2.0837151485692886,
+      "grad_norm": 1.6065024137496948,
+      "learning_rate": 1.527141419051186e-05,
+      "loss": 0.9396,
+      "step": 523000
+    },
+    {
+      "epoch": 2.0857072280612288,
+      "grad_norm": 1.6434810161590576,
+      "learning_rate": 1.5238212865646192e-05,
+      "loss": 0.9428,
+      "step": 523500
+    },
+    {
+      "epoch": 2.0876993075531685,
+      "grad_norm": 1.6344958543777466,
+      "learning_rate": 1.5205011540780522e-05,
+      "loss": 0.9442,
+      "step": 524000
+    },
+    {
+      "epoch": 2.0896913870451086,
+      "grad_norm": 1.6828653812408447,
+      "learning_rate": 1.5171810215914856e-05,
+      "loss": 0.9477,
+      "step": 524500
+    },
+    {
+      "epoch": 2.0916834665370487,
+      "grad_norm": 1.6989903450012207,
+      "learning_rate": 1.5138608891049188e-05,
+      "loss": 0.9431,
+      "step": 525000
+    },
+    {
+      "epoch": 2.093675546028989,
+      "grad_norm": 1.7739417552947998,
+      "learning_rate": 1.510540756618352e-05,
+      "loss": 0.9416,
+      "step": 525500
+    },
+    {
+      "epoch": 2.095667625520929,
+      "grad_norm": 1.6136302947998047,
+      "learning_rate": 1.5072206241317855e-05,
+      "loss": 0.9438,
+      "step": 526000
+    },
+    {
+      "epoch": 2.0976597050128687,
+      "grad_norm": 1.5163776874542236,
+      "learning_rate": 1.5039004916452187e-05,
+      "loss": 0.94,
+      "step": 526500
+    },
+    {
+      "epoch": 2.099651784504809,
+      "grad_norm": 1.613546371459961,
+      "learning_rate": 1.500580359158652e-05,
+      "loss": 0.9397,
+      "step": 527000
+    },
+    {
+      "epoch": 2.101643863996749,
+      "grad_norm": 1.6435461044311523,
+      "learning_rate": 1.4972602266720853e-05,
+      "loss": 0.944,
+      "step": 527500
+    },
+    {
+      "epoch": 2.103635943488689,
+      "grad_norm": 1.5988768339157104,
+      "learning_rate": 1.4939400941855184e-05,
+      "loss": 0.9438,
+      "step": 528000
+    },
+    {
+      "epoch": 2.1056280229806292,
+      "grad_norm": 1.619510531425476,
+      "learning_rate": 1.4906199616989516e-05,
+      "loss": 0.9448,
+      "step": 528500
+    },
+    {
+      "epoch": 2.107620102472569,
+      "grad_norm": 1.5073994398117065,
+      "learning_rate": 1.487299829212385e-05,
+      "loss": 0.9451,
+      "step": 529000
+    },
+    {
+      "epoch": 2.109612181964509,
+      "grad_norm": 1.546704888343811,
+      "learning_rate": 1.4839796967258182e-05,
+      "loss": 0.9432,
+      "step": 529500
+    },
+    {
+      "epoch": 2.111604261456449,
+      "grad_norm": 1.5903176069259644,
+      "learning_rate": 1.4806595642392516e-05,
+      "loss": 0.9467,
+      "step": 530000
+    },
+    {
+      "epoch": 2.1135963409483893,
+      "grad_norm": 1.5203909873962402,
+      "learning_rate": 1.4773394317526848e-05,
+      "loss": 0.9482,
+      "step": 530500
+    },
+    {
+      "epoch": 2.1155884204403295,
+      "grad_norm": 1.585738182067871,
+      "learning_rate": 1.474019299266118e-05,
+      "loss": 0.9445,
+      "step": 531000
+    },
+    {
+      "epoch": 2.117580499932269,
+      "grad_norm": 1.668945074081421,
+      "learning_rate": 1.470699166779551e-05,
+      "loss": 0.938,
+      "step": 531500
+    },
+    {
+      "epoch": 2.1195725794242093,
+      "grad_norm": 1.604917049407959,
+      "learning_rate": 1.4673790342929845e-05,
+      "loss": 0.9385,
+      "step": 532000
+    },
+    {
+      "epoch": 2.1215646589161494,
+      "grad_norm": 1.6301393508911133,
+      "learning_rate": 1.4640589018064177e-05,
+      "loss": 0.9471,
+      "step": 532500
+    },
+    {
+      "epoch": 2.1235567384080896,
+      "grad_norm": 1.6021013259887695,
+      "learning_rate": 1.460738769319851e-05,
+      "loss": 0.9474,
+      "step": 533000
+    },
+    {
+      "epoch": 2.1255488179000297,
+      "grad_norm": 1.6153134107589722,
+      "learning_rate": 1.4574186368332843e-05,
+      "loss": 0.9432,
+      "step": 533500
+    },
+    {
+      "epoch": 2.1275408973919694,
+      "grad_norm": 1.721356749534607,
+      "learning_rate": 1.4540985043467175e-05,
+      "loss": 0.9435,
+      "step": 534000
+    },
+    {
+      "epoch": 2.1295329768839095,
+      "grad_norm": 1.6970722675323486,
+      "learning_rate": 1.450778371860151e-05,
+      "loss": 0.9466,
+      "step": 534500
+    },
+    {
+      "epoch": 2.1315250563758497,
+      "grad_norm": 1.6612935066223145,
+      "learning_rate": 1.4474582393735842e-05,
+      "loss": 0.9416,
+      "step": 535000
+    },
+    {
+      "epoch": 2.13351713586779,
+      "grad_norm": 1.6095038652420044,
+      "learning_rate": 1.4441381068870172e-05,
+      "loss": 0.9425,
+      "step": 535500
+    },
+    {
+      "epoch": 2.13550921535973,
+      "grad_norm": 1.621237874031067,
+      "learning_rate": 1.4408179744004504e-05,
+      "loss": 0.9423,
+      "step": 536000
+    },
+    {
+      "epoch": 2.1375012948516696,
+      "grad_norm": 1.6304696798324585,
+      "learning_rate": 1.4374978419138838e-05,
+      "loss": 0.9442,
+      "step": 536500
+    },
+    {
+      "epoch": 2.1394933743436098,
+      "grad_norm": 1.6269677877426147,
+      "learning_rate": 1.434177709427317e-05,
+      "loss": 0.9473,
+      "step": 537000
+    },
+    {
+      "epoch": 2.14148545383555,
+      "grad_norm": 1.6289795637130737,
+      "learning_rate": 1.4308575769407503e-05,
+      "loss": 0.9367,
+      "step": 537500
+    },
+    {
+      "epoch": 2.14347753332749,
+      "grad_norm": 1.5147687196731567,
+      "learning_rate": 1.4275374444541837e-05,
+      "loss": 0.9458,
+      "step": 538000
+    },
+    {
+      "epoch": 2.14546961281943,
+      "grad_norm": 1.5897058248519897,
+      "learning_rate": 1.4242173119676169e-05,
+      "loss": 0.9388,
+      "step": 538500
+    },
+    {
+      "epoch": 2.14746169231137,
+      "grad_norm": 1.6228257417678833,
+      "learning_rate": 1.42089717948105e-05,
+      "loss": 0.9396,
+      "step": 539000
+    },
+    {
+      "epoch": 2.14945377180331,
+      "grad_norm": 1.53834867477417,
+      "learning_rate": 1.4175770469944832e-05,
+      "loss": 0.9385,
+      "step": 539500
+    },
+    {
+      "epoch": 2.15144585129525,
+      "grad_norm": 1.6430636644363403,
+      "learning_rate": 1.4142569145079166e-05,
+      "loss": 0.9425,
+      "step": 540000
+    },
+    {
+      "epoch": 2.1534379307871903,
+      "grad_norm": 1.5667630434036255,
+      "learning_rate": 1.4109367820213498e-05,
+      "loss": 0.9391,
+      "step": 540500
+    },
+    {
+      "epoch": 2.1554300102791304,
+      "grad_norm": 1.549913763999939,
+      "learning_rate": 1.4076166495347832e-05,
+      "loss": 0.9381,
+      "step": 541000
+    },
+    {
+      "epoch": 2.15742208977107,
+      "grad_norm": 1.6583200693130493,
+      "learning_rate": 1.4042965170482164e-05,
+      "loss": 0.9403,
+      "step": 541500
+    },
+    {
+      "epoch": 2.15941416926301,
+      "grad_norm": 1.6687999963760376,
+      "learning_rate": 1.4009763845616498e-05,
+      "loss": 0.9381,
+      "step": 542000
+    },
+    {
+      "epoch": 2.1614062487549504,
+      "grad_norm": 1.5992510318756104,
+      "learning_rate": 1.397656252075083e-05,
+      "loss": 0.9364,
+      "step": 542500
+    },
+    {
+      "epoch": 2.1633983282468905,
+      "grad_norm": 1.7213138341903687,
+      "learning_rate": 1.394336119588516e-05,
+      "loss": 0.9413,
+      "step": 543000
+    },
+    {
+      "epoch": 2.1653904077388306,
+      "grad_norm": 1.6391761302947998,
+      "learning_rate": 1.3910159871019493e-05,
+      "loss": 0.9421,
+      "step": 543500
+    },
+    {
+      "epoch": 2.1673824872307703,
+      "grad_norm": 1.6031770706176758,
+      "learning_rate": 1.3876958546153827e-05,
+      "loss": 0.9326,
+      "step": 544000
+    },
+    {
+      "epoch": 2.1693745667227105,
+      "grad_norm": 1.687530517578125,
+      "learning_rate": 1.3843757221288159e-05,
+      "loss": 0.9362,
+      "step": 544500
+    },
+    {
+      "epoch": 2.1713666462146506,
+      "grad_norm": 1.613240361213684,
+      "learning_rate": 1.3810555896422491e-05,
+      "loss": 0.9376,
+      "step": 545000
+    },
+    {
+      "epoch": 2.1733587257065907,
+      "grad_norm": 1.5415453910827637,
+      "learning_rate": 1.3777354571556825e-05,
+      "loss": 0.9378,
+      "step": 545500
+    },
+    {
+      "epoch": 2.175350805198531,
+      "grad_norm": 1.663482904434204,
+      "learning_rate": 1.3744153246691157e-05,
+      "loss": 0.9411,
+      "step": 546000
+    },
+    {
+      "epoch": 2.1773428846904705,
+      "grad_norm": 1.7156012058258057,
+      "learning_rate": 1.3710951921825488e-05,
+      "loss": 0.9375,
+      "step": 546500
+    },
+    {
+      "epoch": 2.1793349641824107,
+      "grad_norm": 1.7120176553726196,
+      "learning_rate": 1.367775059695982e-05,
+      "loss": 0.936,
+      "step": 547000
+    },
+    {
+      "epoch": 2.181327043674351,
+      "grad_norm": 1.781632661819458,
+      "learning_rate": 1.3644549272094154e-05,
+      "loss": 0.9416,
+      "step": 547500
+    },
+    {
+      "epoch": 2.183319123166291,
+      "grad_norm": 1.6548519134521484,
+      "learning_rate": 1.3611347947228486e-05,
+      "loss": 0.9332,
+      "step": 548000
+    },
+    {
+      "epoch": 2.185311202658231,
+      "grad_norm": 1.773166537284851,
+      "learning_rate": 1.357814662236282e-05,
+      "loss": 0.9375,
+      "step": 548500
+    },
+    {
+      "epoch": 2.187303282150171,
+      "grad_norm": 1.5719444751739502,
+      "learning_rate": 1.3544945297497152e-05,
+      "loss": 0.9398,
+      "step": 549000
+    },
+    {
+      "epoch": 2.189295361642111,
+      "grad_norm": 1.6089649200439453,
+      "learning_rate": 1.3511743972631485e-05,
+      "loss": 0.9328,
+      "step": 549500
+    },
+    {
+      "epoch": 2.191287441134051,
+      "grad_norm": 1.5634949207305908,
+      "learning_rate": 1.3478542647765819e-05,
+      "loss": 0.9392,
+      "step": 550000
+    },
+    {
+      "epoch": 2.193279520625991,
+      "grad_norm": 1.6151635646820068,
+      "learning_rate": 1.3445341322900149e-05,
+      "loss": 0.9392,
+      "step": 550500
+    },
+    {
+      "epoch": 2.1952716001179313,
+      "grad_norm": 1.6121504306793213,
+      "learning_rate": 1.3412139998034481e-05,
+      "loss": 0.9371,
+      "step": 551000
+    },
+    {
+      "epoch": 2.197263679609871,
+      "grad_norm": 1.5835556983947754,
+      "learning_rate": 1.3378938673168814e-05,
+      "loss": 0.934,
+      "step": 551500
+    },
+    {
+      "epoch": 2.199255759101811,
+      "grad_norm": 1.6224946975708008,
+      "learning_rate": 1.3345737348303148e-05,
+      "loss": 0.9388,
+      "step": 552000
+    },
+    {
+      "epoch": 2.2012478385937513,
+      "grad_norm": 1.7343262434005737,
+      "learning_rate": 1.331253602343748e-05,
+      "loss": 0.9421,
+      "step": 552500
+    },
+    {
+      "epoch": 2.2032399180856914,
+      "grad_norm": 1.5879350900650024,
+      "learning_rate": 1.3279334698571814e-05,
+      "loss": 0.9398,
+      "step": 553000
+    },
+    {
+      "epoch": 2.2052319975776316,
+      "grad_norm": 1.6103401184082031,
+      "learning_rate": 1.3246133373706146e-05,
+      "loss": 0.9336,
+      "step": 553500
+    },
+    {
+      "epoch": 2.2072240770695712,
+      "grad_norm": 1.7701284885406494,
+      "learning_rate": 1.3212932048840476e-05,
+      "loss": 0.9389,
+      "step": 554000
+    },
+    {
+      "epoch": 2.2092161565615114,
+      "grad_norm": 1.797418236732483,
+      "learning_rate": 1.3179730723974809e-05,
+      "loss": 0.9301,
+      "step": 554500
+    },
+    {
+      "epoch": 2.2112082360534515,
+      "grad_norm": 1.6516611576080322,
+      "learning_rate": 1.3146529399109143e-05,
+      "loss": 0.9393,
+      "step": 555000
+    },
+    {
+      "epoch": 2.2132003155453916,
+      "grad_norm": 1.5950418710708618,
+      "learning_rate": 1.3113328074243475e-05,
+      "loss": 0.9327,
+      "step": 555500
+    },
+    {
+      "epoch": 2.215192395037332,
+      "grad_norm": 1.5820379257202148,
+      "learning_rate": 1.3080126749377809e-05,
+      "loss": 0.9331,
+      "step": 556000
+    },
+    {
+      "epoch": 2.2171844745292715,
+      "grad_norm": 1.5876384973526,
+      "learning_rate": 1.3046925424512141e-05,
+      "loss": 0.9366,
+      "step": 556500
+    },
+    {
+      "epoch": 2.2191765540212116,
+      "grad_norm": 1.6041828393936157,
+      "learning_rate": 1.3013724099646473e-05,
+      "loss": 0.9395,
+      "step": 557000
+    },
+    {
+      "epoch": 2.2211686335131517,
+      "grad_norm": 1.6151782274246216,
+      "learning_rate": 1.2980522774780807e-05,
+      "loss": 0.9366,
+      "step": 557500
+    },
+    {
+      "epoch": 2.223160713005092,
+      "grad_norm": 1.6997599601745605,
+      "learning_rate": 1.2947321449915138e-05,
+      "loss": 0.9372,
+      "step": 558000
+    },
+    {
+      "epoch": 2.225152792497032,
+      "grad_norm": 1.675376534461975,
+      "learning_rate": 1.291412012504947e-05,
+      "loss": 0.9389,
+      "step": 558500
+    },
+    {
+      "epoch": 2.2271448719889717,
+      "grad_norm": 1.5898339748382568,
+      "learning_rate": 1.2880918800183802e-05,
+      "loss": 0.9397,
+      "step": 559000
+    },
+    {
+      "epoch": 2.229136951480912,
+      "grad_norm": 1.583292007446289,
+      "learning_rate": 1.2847717475318136e-05,
+      "loss": 0.9393,
+      "step": 559500
+    },
+    {
+      "epoch": 2.231129030972852,
+      "grad_norm": 1.5941981077194214,
+      "learning_rate": 1.2814516150452468e-05,
+      "loss": 0.9382,
+      "step": 560000
+    },
+    {
+      "epoch": 2.233121110464792,
+      "grad_norm": 1.6873384714126587,
+      "learning_rate": 1.2781314825586802e-05,
+      "loss": 0.9328,
+      "step": 560500
+    },
+    {
+      "epoch": 2.2351131899567322,
+      "grad_norm": 1.6278514862060547,
+      "learning_rate": 1.2748113500721134e-05,
+      "loss": 0.9361,
+      "step": 561000
+    },
+    {
+      "epoch": 2.237105269448672,
+      "grad_norm": 1.6467139720916748,
+      "learning_rate": 1.2714912175855465e-05,
+      "loss": 0.9348,
+      "step": 561500
+    },
+    {
+      "epoch": 2.239097348940612,
+      "grad_norm": 1.6214808225631714,
+      "learning_rate": 1.2681710850989797e-05,
+      "loss": 0.9332,
+      "step": 562000
+    },
+    {
+      "epoch": 2.241089428432552,
+      "grad_norm": 1.6389514207839966,
+      "learning_rate": 1.2648509526124131e-05,
+      "loss": 0.9395,
+      "step": 562500
+    },
+    {
+      "epoch": 2.2430815079244923,
+      "grad_norm": 1.6445648670196533,
+      "learning_rate": 1.2615308201258463e-05,
+      "loss": 0.9381,
+      "step": 563000
+    },
+    {
+      "epoch": 2.2450735874164325,
+      "grad_norm": 1.6182383298873901,
+      "learning_rate": 1.2582106876392796e-05,
+      "loss": 0.9356,
+      "step": 563500
+    },
+    {
+      "epoch": 2.247065666908372,
+      "grad_norm": 1.6169302463531494,
+      "learning_rate": 1.254890555152713e-05,
+      "loss": 0.932,
+      "step": 564000
+    },
+    {
+      "epoch": 2.2490577464003123,
+      "grad_norm": 1.5797264575958252,
+      "learning_rate": 1.2515704226661462e-05,
+      "loss": 0.935,
+      "step": 564500
+    },
+    {
+      "epoch": 2.2510498258922524,
+      "grad_norm": 1.5935204029083252,
+      "learning_rate": 1.2482502901795794e-05,
+      "loss": 0.9338,
+      "step": 565000
+    },
+    {
+      "epoch": 2.2530419053841926,
+      "grad_norm": 1.563314437866211,
+      "learning_rate": 1.2449301576930126e-05,
+      "loss": 0.9326,
+      "step": 565500
+    },
+    {
+      "epoch": 2.2550339848761327,
+      "grad_norm": 1.7124460935592651,
+      "learning_rate": 1.241610025206446e-05,
+      "loss": 0.9359,
+      "step": 566000
+    },
+    {
+      "epoch": 2.2570260643680724,
+      "grad_norm": 1.6239125728607178,
+      "learning_rate": 1.238289892719879e-05,
+      "loss": 0.9347,
+      "step": 566500
+    },
+    {
+      "epoch": 2.2590181438600125,
+      "grad_norm": 1.7260369062423706,
+      "learning_rate": 1.2349697602333125e-05,
+      "loss": 0.9347,
+      "step": 567000
+    },
+    {
+      "epoch": 2.2610102233519527,
+      "grad_norm": 1.6370229721069336,
+      "learning_rate": 1.2316496277467457e-05,
+      "loss": 0.9352,
+      "step": 567500
+    },
+    {
+      "epoch": 2.263002302843893,
+      "grad_norm": 1.7867411375045776,
+      "learning_rate": 1.2283294952601789e-05,
+      "loss": 0.9347,
+      "step": 568000
+    },
+    {
+      "epoch": 2.264994382335833,
+      "grad_norm": 1.6779757738113403,
+      "learning_rate": 1.2250093627736121e-05,
+      "loss": 0.9342,
+      "step": 568500
+    },
+    {
+      "epoch": 2.2669864618277726,
+      "grad_norm": 1.6026312112808228,
+      "learning_rate": 1.2216892302870454e-05,
+      "loss": 0.9321,
+      "step": 569000
+    },
+    {
+      "epoch": 2.2689785413197128,
+      "grad_norm": 1.6595126390457153,
+      "learning_rate": 1.2183690978004787e-05,
+      "loss": 0.9356,
+      "step": 569500
+    },
+    {
+      "epoch": 2.270970620811653,
+      "grad_norm": 1.825676679611206,
+      "learning_rate": 1.2150489653139118e-05,
+      "loss": 0.9399,
+      "step": 570000
+    },
+    {
+      "epoch": 2.272962700303593,
+      "grad_norm": 1.6996159553527832,
+      "learning_rate": 1.2117288328273452e-05,
+      "loss": 0.936,
+      "step": 570500
+    },
+    {
+      "epoch": 2.274954779795533,
+      "grad_norm": 1.6745915412902832,
+      "learning_rate": 1.2084087003407784e-05,
+      "loss": 0.9398,
+      "step": 571000
+    },
+    {
+      "epoch": 2.276946859287473,
+      "grad_norm": 1.6762967109680176,
+      "learning_rate": 1.2050885678542118e-05,
+      "loss": 0.9331,
+      "step": 571500
+    },
+    {
+      "epoch": 2.278938938779413,
+      "grad_norm": 1.6009882688522339,
+      "learning_rate": 1.2017684353676449e-05,
+      "loss": 0.9341,
+      "step": 572000
+    },
+    {
+      "epoch": 2.280931018271353,
+      "grad_norm": 1.6973503828048706,
+      "learning_rate": 1.1984483028810783e-05,
+      "loss": 0.9343,
+      "step": 572500
+    },
+    {
+      "epoch": 2.2829230977632933,
+      "grad_norm": 1.6619569063186646,
+      "learning_rate": 1.1951281703945115e-05,
+      "loss": 0.9348,
+      "step": 573000
+    },
+    {
+      "epoch": 2.2849151772552334,
+      "grad_norm": 1.7245748043060303,
+      "learning_rate": 1.1918080379079449e-05,
+      "loss": 0.9322,
+      "step": 573500
+    },
+    {
+      "epoch": 2.286907256747173,
+      "grad_norm": 1.719177484512329,
+      "learning_rate": 1.188487905421378e-05,
+      "loss": 0.9344,
+      "step": 574000
+    },
+    {
+      "epoch": 2.2888993362391132,
+      "grad_norm": 1.6953608989715576,
+      "learning_rate": 1.1851677729348113e-05,
+      "loss": 0.9347,
+      "step": 574500
+    },
+    {
+      "epoch": 2.2908914157310534,
+      "grad_norm": 1.6570985317230225,
+      "learning_rate": 1.1818476404482445e-05,
+      "loss": 0.9327,
+      "step": 575000
+    },
+    {
+      "epoch": 2.2928834952229935,
+      "grad_norm": 1.772316813468933,
+      "learning_rate": 1.1785275079616778e-05,
+      "loss": 0.9294,
+      "step": 575500
+    },
+    {
+      "epoch": 2.2948755747149336,
+      "grad_norm": 1.68768310546875,
+      "learning_rate": 1.175207375475111e-05,
+      "loss": 0.9335,
+      "step": 576000
+    },
+    {
+      "epoch": 2.2968676542068733,
+      "grad_norm": 1.6368399858474731,
+      "learning_rate": 1.1718872429885442e-05,
+      "loss": 0.9371,
+      "step": 576500
+    },
+    {
+      "epoch": 2.2988597336988135,
+      "grad_norm": 1.6593579053878784,
+      "learning_rate": 1.1685671105019776e-05,
+      "loss": 0.9376,
+      "step": 577000
+    },
+    {
+      "epoch": 2.3008518131907536,
+      "grad_norm": 1.7236179113388062,
+      "learning_rate": 1.1652469780154107e-05,
+      "loss": 0.9336,
+      "step": 577500
+    },
+    {
+      "epoch": 2.3028438926826937,
+      "grad_norm": 1.7361576557159424,
+      "learning_rate": 1.161926845528844e-05,
+      "loss": 0.9307,
+      "step": 578000
+    },
+    {
+      "epoch": 2.304835972174634,
+      "grad_norm": 1.5797193050384521,
+      "learning_rate": 1.1586067130422773e-05,
+      "loss": 0.9348,
+      "step": 578500
+    },
+    {
+      "epoch": 2.3068280516665736,
+      "grad_norm": 1.773432970046997,
+      "learning_rate": 1.1552865805557107e-05,
+      "loss": 0.932,
+      "step": 579000
+    },
+    {
+      "epoch": 2.3088201311585137,
+      "grad_norm": 1.7023743391036987,
+      "learning_rate": 1.1519664480691437e-05,
+      "loss": 0.939,
+      "step": 579500
+    },
+    {
+      "epoch": 2.310812210650454,
+      "grad_norm": 1.6463055610656738,
+      "learning_rate": 1.1486463155825771e-05,
+      "loss": 0.9316,
+      "step": 580000
+    },
+    {
+      "epoch": 2.312804290142394,
+      "grad_norm": 1.6124160289764404,
+      "learning_rate": 1.1453261830960103e-05,
+      "loss": 0.9357,
+      "step": 580500
+    },
+    {
+      "epoch": 2.314796369634334,
+      "grad_norm": 1.7371443510055542,
+      "learning_rate": 1.1420060506094436e-05,
+      "loss": 0.9368,
+      "step": 581000
+    },
+    {
+      "epoch": 2.316788449126274,
+      "grad_norm": 1.711488127708435,
+      "learning_rate": 1.1386859181228768e-05,
+      "loss": 0.931,
+      "step": 581500
+    },
+    {
+      "epoch": 2.318780528618214,
+      "grad_norm": 1.7853292226791382,
+      "learning_rate": 1.13536578563631e-05,
+      "loss": 0.933,
+      "step": 582000
+    },
+    {
+      "epoch": 2.320772608110154,
+      "grad_norm": 1.6868606805801392,
+      "learning_rate": 1.1320456531497434e-05,
+      "loss": 0.9309,
+      "step": 582500
+    },
+    {
+      "epoch": 2.322764687602094,
+      "grad_norm": 1.7364510297775269,
+      "learning_rate": 1.1287255206631766e-05,
+      "loss": 0.9359,
+      "step": 583000
+    },
+    {
+      "epoch": 2.3247567670940343,
+      "grad_norm": 1.633932113647461,
+      "learning_rate": 1.1254053881766098e-05,
+      "loss": 0.9305,
+      "step": 583500
+    },
+    {
+      "epoch": 2.326748846585974,
+      "grad_norm": 1.6290020942687988,
+      "learning_rate": 1.122085255690043e-05,
+      "loss": 0.9308,
+      "step": 584000
+    },
+    {
+      "epoch": 2.328740926077914,
+      "grad_norm": 1.6390776634216309,
+      "learning_rate": 1.1187651232034765e-05,
+      "loss": 0.936,
+      "step": 584500
+    },
+    {
+      "epoch": 2.3307330055698543,
+      "grad_norm": 1.625496506690979,
+      "learning_rate": 1.1154449907169095e-05,
+      "loss": 0.9338,
+      "step": 585000
+    },
+    {
+      "epoch": 2.3327250850617944,
+      "grad_norm": 1.9022979736328125,
+      "learning_rate": 1.1121248582303429e-05,
+      "loss": 0.9291,
+      "step": 585500
+    },
+    {
+      "epoch": 2.3347171645537346,
+      "grad_norm": 1.647518277168274,
+      "learning_rate": 1.1088047257437761e-05,
+      "loss": 0.9358,
+      "step": 586000
+    },
+    {
+      "epoch": 2.3367092440456743,
+      "grad_norm": 1.5992262363433838,
+      "learning_rate": 1.1054845932572095e-05,
+      "loss": 0.9346,
+      "step": 586500
+    },
+    {
+      "epoch": 2.3387013235376144,
+      "grad_norm": 1.6610697507858276,
+      "learning_rate": 1.1021644607706426e-05,
+      "loss": 0.9307,
+      "step": 587000
+    },
+    {
+      "epoch": 2.3406934030295545,
+      "grad_norm": 1.6619411706924438,
+      "learning_rate": 1.098844328284076e-05,
+      "loss": 0.9311,
+      "step": 587500
+    },
+    {
+      "epoch": 2.3426854825214947,
+      "grad_norm": 1.5342708826065063,
+      "learning_rate": 1.0955241957975092e-05,
+      "loss": 0.9316,
+      "step": 588000
+    },
+    {
+      "epoch": 2.344677562013435,
+      "grad_norm": 1.6005909442901611,
+      "learning_rate": 1.0922040633109424e-05,
+      "loss": 0.93,
+      "step": 588500
+    },
+    {
+      "epoch": 2.3466696415053745,
+      "grad_norm": 1.674174189567566,
+      "learning_rate": 1.0888839308243756e-05,
+      "loss": 0.9312,
+      "step": 589000
+    },
+    {
+      "epoch": 2.3486617209973146,
+      "grad_norm": 1.6399884223937988,
+      "learning_rate": 1.0855637983378089e-05,
+      "loss": 0.9325,
+      "step": 589500
+    },
+    {
+      "epoch": 2.3506538004892548,
+      "grad_norm": 1.6004582643508911,
+      "learning_rate": 1.0822436658512422e-05,
+      "loss": 0.9303,
+      "step": 590000
+    },
+    {
+      "epoch": 2.352645879981195,
+      "grad_norm": 1.7159782648086548,
+      "learning_rate": 1.0789235333646755e-05,
+      "loss": 0.929,
+      "step": 590500
+    },
+    {
+      "epoch": 2.354637959473135,
+      "grad_norm": 1.5917738676071167,
+      "learning_rate": 1.0756034008781087e-05,
+      "loss": 0.9321,
+      "step": 591000
+    },
+    {
+      "epoch": 2.3566300389650747,
+      "grad_norm": 1.5573068857192993,
+      "learning_rate": 1.0722832683915419e-05,
+      "loss": 0.9324,
+      "step": 591500
+    },
+    {
+      "epoch": 2.358622118457015,
+      "grad_norm": 1.6334115266799927,
+      "learning_rate": 1.0689631359049753e-05,
+      "loss": 0.932,
+      "step": 592000
+    },
+    {
+      "epoch": 2.360614197948955,
+      "grad_norm": 1.576357126235962,
+      "learning_rate": 1.0656430034184084e-05,
+      "loss": 0.9302,
+      "step": 592500
+    },
+    {
+      "epoch": 2.362606277440895,
+      "grad_norm": 1.6486066579818726,
+      "learning_rate": 1.0623228709318418e-05,
+      "loss": 0.9332,
+      "step": 593000
+    },
+    {
+      "epoch": 2.3645983569328353,
+      "grad_norm": 1.6311765909194946,
+      "learning_rate": 1.059002738445275e-05,
+      "loss": 0.9285,
+      "step": 593500
+    },
+    {
+      "epoch": 2.366590436424775,
+      "grad_norm": 1.7107651233673096,
+      "learning_rate": 1.0556826059587082e-05,
+      "loss": 0.9277,
+      "step": 594000
+    },
+    {
+      "epoch": 2.368582515916715,
+      "grad_norm": 1.593829870223999,
+      "learning_rate": 1.0523624734721414e-05,
+      "loss": 0.9266,
+      "step": 594500
+    },
+    {
+      "epoch": 2.3705745954086552,
+      "grad_norm": 1.7289330959320068,
+      "learning_rate": 1.0490423409855746e-05,
+      "loss": 0.9307,
+      "step": 595000
+    },
+    {
+      "epoch": 2.3725666749005954,
+      "grad_norm": 1.5663368701934814,
+      "learning_rate": 1.045722208499008e-05,
+      "loss": 0.9256,
+      "step": 595500
+    },
+    {
+      "epoch": 2.3745587543925355,
+      "grad_norm": 1.719089388847351,
+      "learning_rate": 1.0424020760124413e-05,
+      "loss": 0.9295,
+      "step": 596000
+    },
+    {
+      "epoch": 2.376550833884475,
+      "grad_norm": 1.601806879043579,
+      "learning_rate": 1.0390819435258745e-05,
+      "loss": 0.9316,
+      "step": 596500
+    },
+    {
+      "epoch": 2.3785429133764153,
+      "grad_norm": 1.6492115259170532,
+      "learning_rate": 1.0357618110393077e-05,
+      "loss": 0.9243,
+      "step": 597000
+    },
+    {
+      "epoch": 2.3805349928683555,
+      "grad_norm": 1.6850099563598633,
+      "learning_rate": 1.0324416785527411e-05,
+      "loss": 0.93,
+      "step": 597500
+    },
+    {
+      "epoch": 2.3825270723602956,
+      "grad_norm": 1.657905101776123,
+      "learning_rate": 1.0291215460661743e-05,
+      "loss": 0.9322,
+      "step": 598000
+    },
+    {
+      "epoch": 2.3845191518522357,
+      "grad_norm": 1.6386810541152954,
+      "learning_rate": 1.0258014135796075e-05,
+      "loss": 0.9282,
+      "step": 598500
+    },
+    {
+      "epoch": 2.3865112313441754,
+      "grad_norm": 1.6624351739883423,
+      "learning_rate": 1.0224812810930408e-05,
+      "loss": 0.9268,
+      "step": 599000
+    },
+    {
+      "epoch": 2.3885033108361156,
+      "grad_norm": 1.6481283903121948,
+      "learning_rate": 1.019161148606474e-05,
+      "loss": 0.9277,
+      "step": 599500
+    },
+    {
+      "epoch": 2.3904953903280557,
+      "grad_norm": 1.7030024528503418,
+      "learning_rate": 1.0158410161199072e-05,
+      "loss": 0.926,
+      "step": 600000
+    },
+    {
+      "epoch": 2.392487469819996,
+      "grad_norm": 1.6452306509017944,
+      "learning_rate": 1.0125208836333406e-05,
+      "loss": 0.9346,
+      "step": 600500
+    },
+    {
+      "epoch": 2.394479549311936,
+      "grad_norm": 1.5868730545043945,
+      "learning_rate": 1.0092007511467738e-05,
+      "loss": 0.9276,
+      "step": 601000
+    },
+    {
+      "epoch": 2.3964716288038757,
+      "grad_norm": 1.5723228454589844,
+      "learning_rate": 1.005880618660207e-05,
+      "loss": 0.9273,
+      "step": 601500
+    },
+    {
+      "epoch": 2.398463708295816,
+      "grad_norm": 1.6179591417312622,
+      "learning_rate": 1.0025604861736403e-05,
+      "loss": 0.9293,
+      "step": 602000
+    },
+    {
+      "epoch": 2.400455787787756,
+      "grad_norm": 1.7436559200286865,
+      "learning_rate": 9.992403536870735e-06,
+      "loss": 0.9269,
+      "step": 602500
+    },
+    {
+      "epoch": 2.402447867279696,
+      "grad_norm": 1.5943328142166138,
+      "learning_rate": 9.959202212005069e-06,
+      "loss": 0.9321,
+      "step": 603000
+    },
+    {
+      "epoch": 2.404439946771636,
+      "grad_norm": 1.9251843690872192,
+      "learning_rate": 9.926000887139401e-06,
+      "loss": 0.9334,
+      "step": 603500
+    },
+    {
+      "epoch": 2.406432026263576,
+      "grad_norm": 1.6488077640533447,
+      "learning_rate": 9.892799562273733e-06,
+      "loss": 0.9282,
+      "step": 604000
+    },
+    {
+      "epoch": 2.408424105755516,
+      "grad_norm": 1.724740743637085,
+      "learning_rate": 9.859598237408066e-06,
+      "loss": 0.9314,
+      "step": 604500
+    },
+    {
+      "epoch": 2.410416185247456,
+      "grad_norm": 1.619619369506836,
+      "learning_rate": 9.8263969125424e-06,
+      "loss": 0.9249,
+      "step": 605000
+    },
+    {
+      "epoch": 2.4124082647393963,
+      "grad_norm": 1.6060360670089722,
+      "learning_rate": 9.793195587676732e-06,
+      "loss": 0.9295,
+      "step": 605500
+    },
+    {
+      "epoch": 2.4144003442313364,
+      "grad_norm": 1.5857383012771606,
+      "learning_rate": 9.759994262811064e-06,
+      "loss": 0.925,
+      "step": 606000
+    },
+    {
+      "epoch": 2.416392423723276,
+      "grad_norm": 1.6881599426269531,
+      "learning_rate": 9.726792937945396e-06,
+      "loss": 0.9302,
+      "step": 606500
+    },
+    {
+      "epoch": 2.4183845032152163,
+      "grad_norm": 1.7167803049087524,
+      "learning_rate": 9.693591613079728e-06,
+      "loss": 0.9311,
+      "step": 607000
+    },
+    {
+      "epoch": 2.4203765827071564,
+      "grad_norm": 1.6533732414245605,
+      "learning_rate": 9.66039028821406e-06,
+      "loss": 0.9302,
+      "step": 607500
+    },
+    {
+      "epoch": 2.4223686621990965,
+      "grad_norm": 1.625589370727539,
+      "learning_rate": 9.627188963348393e-06,
+      "loss": 0.9266,
+      "step": 608000
+    },
+    {
+      "epoch": 2.4243607416910367,
+      "grad_norm": 1.6718621253967285,
+      "learning_rate": 9.593987638482727e-06,
+      "loss": 0.9332,
+      "step": 608500
+    },
+    {
+      "epoch": 2.4263528211829763,
+      "grad_norm": 1.60472571849823,
+      "learning_rate": 9.560786313617059e-06,
+      "loss": 0.9299,
+      "step": 609000
+    },
+    {
+      "epoch": 2.4283449006749165,
+      "grad_norm": 1.7404916286468506,
+      "learning_rate": 9.527584988751391e-06,
+      "loss": 0.925,
+      "step": 609500
+    },
+    {
+      "epoch": 2.4303369801668566,
+      "grad_norm": 1.6216133832931519,
+      "learning_rate": 9.494383663885724e-06,
+      "loss": 0.9239,
+      "step": 610000
+    },
+    {
+      "epoch": 2.4323290596587968,
+      "grad_norm": 1.7234395742416382,
+      "learning_rate": 9.461182339020057e-06,
+      "loss": 0.9239,
+      "step": 610500
+    },
+    {
+      "epoch": 2.434321139150737,
+      "grad_norm": 1.6947396993637085,
+      "learning_rate": 9.42798101415439e-06,
+      "loss": 0.9287,
+      "step": 611000
+    },
+    {
+      "epoch": 2.4363132186426766,
+      "grad_norm": 1.6728731393814087,
+      "learning_rate": 9.394779689288722e-06,
+      "loss": 0.9251,
+      "step": 611500
+    },
+    {
+      "epoch": 2.4383052981346167,
+      "grad_norm": 1.6504451036453247,
+      "learning_rate": 9.361578364423054e-06,
+      "loss": 0.9292,
+      "step": 612000
+    },
+    {
+      "epoch": 2.440297377626557,
+      "grad_norm": 1.7249456644058228,
+      "learning_rate": 9.328377039557386e-06,
+      "loss": 0.923,
+      "step": 612500
+    },
+    {
+      "epoch": 2.442289457118497,
+      "grad_norm": 1.6237436532974243,
+      "learning_rate": 9.29517571469172e-06,
+      "loss": 0.9314,
+      "step": 613000
+    },
+    {
+      "epoch": 2.444281536610437,
+      "grad_norm": 1.7261240482330322,
+      "learning_rate": 9.26197438982605e-06,
+      "loss": 0.9292,
+      "step": 613500
+    },
+    {
+      "epoch": 2.446273616102377,
+      "grad_norm": 1.7471133470535278,
+      "learning_rate": 9.228773064960385e-06,
+      "loss": 0.925,
+      "step": 614000
+    },
+    {
+      "epoch": 2.448265695594317,
+      "grad_norm": 1.6441329717636108,
+      "learning_rate": 9.195571740094717e-06,
+      "loss": 0.9232,
+      "step": 614500
+    },
+    {
+      "epoch": 2.450257775086257,
+      "grad_norm": 1.6200135946273804,
+      "learning_rate": 9.162370415229051e-06,
+      "loss": 0.9244,
+      "step": 615000
+    },
+    {
+      "epoch": 2.452249854578197,
+      "grad_norm": 1.6879627704620361,
+      "learning_rate": 9.129169090363381e-06,
+      "loss": 0.9313,
+      "step": 615500
+    },
+    {
+      "epoch": 2.4542419340701374,
+      "grad_norm": 1.7343822717666626,
+      "learning_rate": 9.095967765497715e-06,
+      "loss": 0.9278,
+      "step": 616000
+    },
+    {
+      "epoch": 2.456234013562077,
+      "grad_norm": 1.747880458831787,
+      "learning_rate": 9.062766440632048e-06,
+      "loss": 0.9257,
+      "step": 616500
+    },
+    {
+      "epoch": 2.458226093054017,
+      "grad_norm": 1.7319437265396118,
+      "learning_rate": 9.02956511576638e-06,
+      "loss": 0.9276,
+      "step": 617000
+    },
+    {
+      "epoch": 2.4602181725459573,
+      "grad_norm": 1.5755479335784912,
+      "learning_rate": 8.996363790900712e-06,
+      "loss": 0.9298,
+      "step": 617500
+    },
+    {
+      "epoch": 2.4622102520378975,
+      "grad_norm": 1.6048145294189453,
+      "learning_rate": 8.963162466035046e-06,
+      "loss": 0.9272,
+      "step": 618000
+    },
+    {
+      "epoch": 2.4642023315298376,
+      "grad_norm": 1.631492257118225,
+      "learning_rate": 8.929961141169378e-06,
+      "loss": 0.9252,
+      "step": 618500
+    },
+    {
+      "epoch": 2.4661944110217773,
+      "grad_norm": 1.6088603734970093,
+      "learning_rate": 8.89675981630371e-06,
+      "loss": 0.9248,
+      "step": 619000
+    },
+    {
+      "epoch": 2.4681864905137174,
+      "grad_norm": 1.5782322883605957,
+      "learning_rate": 8.863558491438043e-06,
+      "loss": 0.928,
+      "step": 619500
+    },
+    {
+      "epoch": 2.4701785700056575,
+      "grad_norm": 1.699977993965149,
+      "learning_rate": 8.830357166572375e-06,
+      "loss": 0.9247,
+      "step": 620000
+    },
+    {
+      "epoch": 2.4721706494975977,
+      "grad_norm": 1.6532070636749268,
+      "learning_rate": 8.797155841706709e-06,
+      "loss": 0.9251,
+      "step": 620500
+    },
+    {
+      "epoch": 2.474162728989538,
+      "grad_norm": 1.6223140954971313,
+      "learning_rate": 8.76395451684104e-06,
+      "loss": 0.9292,
+      "step": 621000
+    },
+    {
+      "epoch": 2.4761548084814775,
+      "grad_norm": 1.8070526123046875,
+      "learning_rate": 8.730753191975373e-06,
+      "loss": 0.924,
+      "step": 621500
+    },
+    {
+      "epoch": 2.4781468879734176,
+      "grad_norm": 1.5933187007904053,
+      "learning_rate": 8.697551867109706e-06,
+      "loss": 0.9258,
+      "step": 622000
+    },
+    {
+      "epoch": 2.480138967465358,
+      "grad_norm": 1.6845451593399048,
+      "learning_rate": 8.66435054224404e-06,
+      "loss": 0.9259,
+      "step": 622500
+    },
+    {
+      "epoch": 2.482131046957298,
+      "grad_norm": 1.7925593852996826,
+      "learning_rate": 8.63114921737837e-06,
+      "loss": 0.9232,
+      "step": 623000
+    },
+    {
+      "epoch": 2.484123126449238,
+      "grad_norm": 1.6180541515350342,
+      "learning_rate": 8.597947892512704e-06,
+      "loss": 0.9255,
+      "step": 623500
+    },
+    {
+      "epoch": 2.4861152059411777,
+      "grad_norm": 1.732258677482605,
+      "learning_rate": 8.564746567647036e-06,
+      "loss": 0.929,
+      "step": 624000
+    },
+    {
+      "epoch": 2.488107285433118,
+      "grad_norm": 1.5710930824279785,
+      "learning_rate": 8.531545242781368e-06,
+      "loss": 0.9301,
+      "step": 624500
+    },
+    {
+      "epoch": 2.490099364925058,
+      "grad_norm": 1.5921260118484497,
+      "learning_rate": 8.4983439179157e-06,
+      "loss": 0.9256,
+      "step": 625000
+    },
+    {
+      "epoch": 2.492091444416998,
+      "grad_norm": 1.640980839729309,
+      "learning_rate": 8.465142593050033e-06,
+      "loss": 0.9236,
+      "step": 625500
+    },
+    {
+      "epoch": 2.4940835239089383,
+      "grad_norm": 1.5668790340423584,
+      "learning_rate": 8.431941268184367e-06,
+      "loss": 0.9242,
+      "step": 626000
+    },
+    {
+      "epoch": 2.496075603400878,
+      "grad_norm": 1.5718388557434082,
+      "learning_rate": 8.398739943318697e-06,
+      "loss": 0.9246,
+      "step": 626500
+    },
+    {
+      "epoch": 2.498067682892818,
+      "grad_norm": 1.5446245670318604,
+      "learning_rate": 8.365538618453031e-06,
+      "loss": 0.9265,
+      "step": 627000
+    },
+    {
+      "epoch": 2.5000597623847582,
+      "grad_norm": 1.6311447620391846,
+      "learning_rate": 8.332337293587363e-06,
+      "loss": 0.9244,
+      "step": 627500
+    },
+    {
+      "epoch": 2.5020518418766984,
+      "grad_norm": 1.5544483661651611,
+      "learning_rate": 8.299135968721697e-06,
+      "loss": 0.927,
+      "step": 628000
+    },
+    {
+      "epoch": 2.5040439213686385,
+      "grad_norm": 1.6825876235961914,
+      "learning_rate": 8.265934643856028e-06,
+      "loss": 0.9249,
+      "step": 628500
+    },
+    {
+      "epoch": 2.506036000860578,
+      "grad_norm": 1.6947489976882935,
+      "learning_rate": 8.232733318990362e-06,
+      "loss": 0.9255,
+      "step": 629000
+    },
+    {
+      "epoch": 2.5080280803525183,
+      "grad_norm": 1.69216787815094,
+      "learning_rate": 8.199531994124694e-06,
+      "loss": 0.9246,
+      "step": 629500
+    },
+    {
+      "epoch": 2.5100201598444585,
+      "grad_norm": 1.7156099081039429,
+      "learning_rate": 8.166330669259028e-06,
+      "loss": 0.9249,
+      "step": 630000
+    },
+    {
+      "epoch": 2.5120122393363986,
+      "grad_norm": 1.7315220832824707,
+      "learning_rate": 8.133129344393359e-06,
+      "loss": 0.9228,
+      "step": 630500
+    },
+    {
+      "epoch": 2.5140043188283387,
+      "grad_norm": 1.5498874187469482,
+      "learning_rate": 8.099928019527692e-06,
+      "loss": 0.9282,
+      "step": 631000
+    },
+    {
+      "epoch": 2.5159963983202784,
+      "grad_norm": 1.7208738327026367,
+      "learning_rate": 8.066726694662025e-06,
+      "loss": 0.9275,
+      "step": 631500
+    },
+    {
+      "epoch": 2.5179884778122186,
+      "grad_norm": 1.7096562385559082,
+      "learning_rate": 8.033525369796357e-06,
+      "loss": 0.9241,
+      "step": 632000
+    },
+    {
+      "epoch": 2.5199805573041587,
+      "grad_norm": 1.6687158346176147,
+      "learning_rate": 8.00032404493069e-06,
+      "loss": 0.9206,
+      "step": 632500
+    },
+    {
+      "epoch": 2.521972636796099,
+      "grad_norm": 1.6677367687225342,
+      "learning_rate": 7.967122720065021e-06,
+      "loss": 0.9222,
+      "step": 633000
+    },
+    {
+      "epoch": 2.523964716288039,
+      "grad_norm": 1.7346562147140503,
+      "learning_rate": 7.933921395199355e-06,
+      "loss": 0.9226,
+      "step": 633500
+    },
+    {
+      "epoch": 2.5259567957799787,
+      "grad_norm": 1.6944063901901245,
+      "learning_rate": 7.900720070333686e-06,
+      "loss": 0.9286,
+      "step": 634000
+    },
+    {
+      "epoch": 2.527948875271919,
+      "grad_norm": 1.638764500617981,
+      "learning_rate": 7.86751874546802e-06,
+      "loss": 0.9236,
+      "step": 634500
+    },
+    {
+      "epoch": 2.529940954763859,
+      "grad_norm": 1.6230340003967285,
+      "learning_rate": 7.834317420602352e-06,
+      "loss": 0.9201,
+      "step": 635000
+    },
+    {
+      "epoch": 2.531933034255799,
+      "grad_norm": 1.7226732969284058,
+      "learning_rate": 7.801116095736686e-06,
+      "loss": 0.9261,
+      "step": 635500
+    },
+    {
+      "epoch": 2.533925113747739,
+      "grad_norm": 1.712620735168457,
+      "learning_rate": 7.767914770871016e-06,
+      "loss": 0.9253,
+      "step": 636000
+    },
+    {
+      "epoch": 2.535917193239679,
+      "grad_norm": 1.6912479400634766,
+      "learning_rate": 7.73471344600535e-06,
+      "loss": 0.9233,
+      "step": 636500
+    },
+    {
+      "epoch": 2.537909272731619,
+      "grad_norm": 1.7039637565612793,
+      "learning_rate": 7.701512121139683e-06,
+      "loss": 0.9223,
+      "step": 637000
+    },
+    {
+      "epoch": 2.539901352223559,
+      "grad_norm": 1.7137672901153564,
+      "learning_rate": 7.668310796274015e-06,
+      "loss": 0.9228,
+      "step": 637500
+    },
+    {
+      "epoch": 2.5418934317154993,
+      "grad_norm": 1.7645823955535889,
+      "learning_rate": 7.635109471408347e-06,
+      "loss": 0.9252,
+      "step": 638000
+    },
+    {
+      "epoch": 2.5438855112074394,
+      "grad_norm": 1.6571165323257446,
+      "learning_rate": 7.60190814654268e-06,
+      "loss": 0.9222,
+      "step": 638500
+    },
+    {
+      "epoch": 2.545877590699379,
+      "grad_norm": 1.7159883975982666,
+      "learning_rate": 7.568706821677013e-06,
+      "loss": 0.9247,
+      "step": 639000
+    },
+    {
+      "epoch": 2.5478696701913193,
+      "grad_norm": 1.6923420429229736,
+      "learning_rate": 7.535505496811345e-06,
+      "loss": 0.9263,
+      "step": 639500
+    },
+    {
+      "epoch": 2.5498617496832594,
+      "grad_norm": 1.6385585069656372,
+      "learning_rate": 7.502304171945678e-06,
+      "loss": 0.9251,
+      "step": 640000
+    },
+    {
+      "epoch": 2.5518538291751995,
+      "grad_norm": 1.5766503810882568,
+      "learning_rate": 7.46910284708001e-06,
+      "loss": 0.9217,
+      "step": 640500
+    },
+    {
+      "epoch": 2.5538459086671397,
+      "grad_norm": 1.662268042564392,
+      "learning_rate": 7.435901522214343e-06,
+      "loss": 0.9218,
+      "step": 641000
+    },
+    {
+      "epoch": 2.5558379881590794,
+      "grad_norm": 1.6351548433303833,
+      "learning_rate": 7.402700197348675e-06,
+      "loss": 0.9226,
+      "step": 641500
+    },
+    {
+      "epoch": 2.5578300676510195,
+      "grad_norm": 1.6451557874679565,
+      "learning_rate": 7.3694988724830075e-06,
+      "loss": 0.922,
+      "step": 642000
+    },
+    {
+      "epoch": 2.5598221471429596,
+      "grad_norm": 1.7144174575805664,
+      "learning_rate": 7.3362975476173405e-06,
+      "loss": 0.9261,
+      "step": 642500
+    },
+    {
+      "epoch": 2.5618142266348998,
+      "grad_norm": 1.6228238344192505,
+      "learning_rate": 7.303096222751674e-06,
+      "loss": 0.9234,
+      "step": 643000
+    },
+    {
+      "epoch": 2.56380630612684,
+      "grad_norm": 1.6820931434631348,
+      "learning_rate": 7.269894897886005e-06,
+      "loss": 0.9245,
+      "step": 643500
+    },
+    {
+      "epoch": 2.5657983856187796,
+      "grad_norm": 1.6930865049362183,
+      "learning_rate": 7.236693573020338e-06,
+      "loss": 0.9197,
+      "step": 644000
+    },
+    {
+      "epoch": 2.5677904651107197,
+      "grad_norm": 1.6372424364089966,
+      "learning_rate": 7.203492248154671e-06,
+      "loss": 0.9182,
+      "step": 644500
+    },
+    {
+      "epoch": 2.56978254460266,
+      "grad_norm": 1.7624679803848267,
+      "learning_rate": 7.170290923289004e-06,
+      "loss": 0.9231,
+      "step": 645000
+    },
+    {
+      "epoch": 2.5717746240946,
+      "grad_norm": 1.5923954248428345,
+      "learning_rate": 7.137089598423336e-06,
+      "loss": 0.9235,
+      "step": 645500
+    },
+    {
+      "epoch": 2.57376670358654,
+      "grad_norm": 1.723497748374939,
+      "learning_rate": 7.103888273557669e-06,
+      "loss": 0.9214,
+      "step": 646000
+    },
+    {
+      "epoch": 2.57575878307848,
+      "grad_norm": 1.7304041385650635,
+      "learning_rate": 7.070686948692001e-06,
+      "loss": 0.9193,
+      "step": 646500
+    },
+    {
+      "epoch": 2.57775086257042,
+      "grad_norm": 1.6448661088943481,
+      "learning_rate": 7.037485623826333e-06,
+      "loss": 0.921,
+      "step": 647000
+    },
+    {
+      "epoch": 2.57974294206236,
+      "grad_norm": 1.667236566543579,
+      "learning_rate": 7.004284298960665e-06,
+      "loss": 0.923,
+      "step": 647500
+    },
+    {
+      "epoch": 2.5817350215543002,
+      "grad_norm": 1.5769813060760498,
+      "learning_rate": 6.9710829740949985e-06,
+      "loss": 0.9217,
+      "step": 648000
+    },
+    {
+      "epoch": 2.5837271010462404,
+      "grad_norm": 1.629429817199707,
+      "learning_rate": 6.9378816492293315e-06,
+      "loss": 0.9259,
+      "step": 648500
+    },
+    {
+      "epoch": 2.58571918053818,
+      "grad_norm": 1.6282541751861572,
+      "learning_rate": 6.904680324363663e-06,
+      "loss": 0.9235,
+      "step": 649000
+    },
+    {
+      "epoch": 2.58771126003012,
+      "grad_norm": 1.6371887922286987,
+      "learning_rate": 6.871478999497996e-06,
+      "loss": 0.9216,
+      "step": 649500
+    },
+    {
+      "epoch": 2.5897033395220603,
+      "grad_norm": 1.6698936223983765,
+      "learning_rate": 6.838277674632329e-06,
+      "loss": 0.9196,
+      "step": 650000
+    },
+    {
+      "epoch": 2.5916954190140005,
+      "grad_norm": 1.615806221961975,
+      "learning_rate": 6.805076349766662e-06,
+      "loss": 0.9235,
+      "step": 650500
+    },
+    {
+      "epoch": 2.5936874985059406,
+      "grad_norm": 1.7115705013275146,
+      "learning_rate": 6.7718750249009935e-06,
+      "loss": 0.924,
+      "step": 651000
+    },
+    {
+      "epoch": 2.5956795779978803,
+      "grad_norm": 1.6222649812698364,
+      "learning_rate": 6.738673700035327e-06,
+      "loss": 0.9225,
+      "step": 651500
+    },
+    {
+      "epoch": 2.5976716574898204,
+      "grad_norm": 1.761478066444397,
+      "learning_rate": 6.70547237516966e-06,
+      "loss": 0.925,
+      "step": 652000
+    },
+    {
+      "epoch": 2.5996637369817606,
+      "grad_norm": 1.6382819414138794,
+      "learning_rate": 6.672271050303992e-06,
+      "loss": 0.9231,
+      "step": 652500
+    },
+    {
+      "epoch": 2.6016558164737007,
+      "grad_norm": 1.710344910621643,
+      "learning_rate": 6.639069725438324e-06,
+      "loss": 0.9234,
+      "step": 653000
+    },
+    {
+      "epoch": 2.603647895965641,
+      "grad_norm": 1.6353403329849243,
+      "learning_rate": 6.605868400572656e-06,
+      "loss": 0.922,
+      "step": 653500
+    },
+    {
+      "epoch": 2.6056399754575805,
+      "grad_norm": 1.6766170263290405,
+      "learning_rate": 6.5726670757069895e-06,
+      "loss": 0.9153,
+      "step": 654000
+    },
+    {
+      "epoch": 2.6076320549495207,
+      "grad_norm": 1.7408475875854492,
+      "learning_rate": 6.539465750841321e-06,
+      "loss": 0.9213,
+      "step": 654500
+    },
+    {
+      "epoch": 2.609624134441461,
+      "grad_norm": 1.653783917427063,
+      "learning_rate": 6.506264425975654e-06,
+      "loss": 0.921,
+      "step": 655000
+    },
+    {
+      "epoch": 2.611616213933401,
+      "grad_norm": 1.6450515985488892,
+      "learning_rate": 6.473063101109987e-06,
+      "loss": 0.9176,
+      "step": 655500
+    },
+    {
+      "epoch": 2.613608293425341,
+      "grad_norm": 1.6264305114746094,
+      "learning_rate": 6.43986177624432e-06,
+      "loss": 0.9186,
+      "step": 656000
+    },
+    {
+      "epoch": 2.6156003729172808,
+      "grad_norm": 1.6926147937774658,
+      "learning_rate": 6.4066604513786515e-06,
+      "loss": 0.9207,
+      "step": 656500
+    },
+    {
+      "epoch": 2.617592452409221,
+      "grad_norm": 1.749976396560669,
+      "learning_rate": 6.3734591265129845e-06,
+      "loss": 0.9211,
+      "step": 657000
+    },
+    {
+      "epoch": 2.619584531901161,
+      "grad_norm": 1.6186459064483643,
+      "learning_rate": 6.340257801647318e-06,
+      "loss": 0.9151,
+      "step": 657500
+    },
+    {
+      "epoch": 2.621576611393101,
+      "grad_norm": 1.5388000011444092,
+      "learning_rate": 6.307056476781651e-06,
+      "loss": 0.9197,
+      "step": 658000
+    },
+    {
+      "epoch": 2.6235686908850413,
+      "grad_norm": 1.5867704153060913,
+      "learning_rate": 6.273855151915982e-06,
+      "loss": 0.9231,
+      "step": 658500
+    },
+    {
+      "epoch": 2.625560770376981,
+      "grad_norm": 1.7041345834732056,
+      "learning_rate": 6.240653827050315e-06,
+      "loss": 0.9211,
+      "step": 659000
+    },
+    {
+      "epoch": 2.627552849868921,
+      "grad_norm": 1.6285433769226074,
+      "learning_rate": 6.207452502184647e-06,
+      "loss": 0.9163,
+      "step": 659500
+    },
+    {
+      "epoch": 2.6295449293608613,
+      "grad_norm": 1.6007583141326904,
+      "learning_rate": 6.17425117731898e-06,
+      "loss": 0.9155,
+      "step": 660000
+    },
+    {
+      "epoch": 2.6315370088528014,
+      "grad_norm": 1.7431002855300903,
+      "learning_rate": 6.141049852453313e-06,
+      "loss": 0.9166,
+      "step": 660500
+    },
+    {
+      "epoch": 2.6335290883447415,
+      "grad_norm": 1.6533669233322144,
+      "learning_rate": 6.107848527587645e-06,
+      "loss": 0.9185,
+      "step": 661000
+    },
+    {
+      "epoch": 2.6355211678366812,
+      "grad_norm": 1.7119617462158203,
+      "learning_rate": 6.074647202721977e-06,
+      "loss": 0.9221,
+      "step": 661500
+    },
+    {
+      "epoch": 2.6375132473286214,
+      "grad_norm": 1.6701066493988037,
+      "learning_rate": 6.04144587785631e-06,
+      "loss": 0.916,
+      "step": 662000
+    },
+    {
+      "epoch": 2.6395053268205615,
+      "grad_norm": 1.9097412824630737,
+      "learning_rate": 6.0082445529906425e-06,
+      "loss": 0.9197,
+      "step": 662500
+    },
+    {
+      "epoch": 2.6414974063125016,
+      "grad_norm": 1.7103551626205444,
+      "learning_rate": 5.9750432281249756e-06,
+      "loss": 0.9211,
+      "step": 663000
+    },
+    {
+      "epoch": 2.6434894858044418,
+      "grad_norm": 1.700982689857483,
+      "learning_rate": 5.941841903259308e-06,
+      "loss": 0.917,
+      "step": 663500
+    },
+    {
+      "epoch": 2.6454815652963815,
+      "grad_norm": 1.5500088930130005,
+      "learning_rate": 5.908640578393641e-06,
+      "loss": 0.9218,
+      "step": 664000
+    },
+    {
+      "epoch": 2.6474736447883216,
+      "grad_norm": 1.614115834236145,
+      "learning_rate": 5.875439253527973e-06,
+      "loss": 0.9158,
+      "step": 664500
+    },
+    {
+      "epoch": 2.6494657242802617,
+      "grad_norm": 1.6373786926269531,
+      "learning_rate": 5.842237928662306e-06,
+      "loss": 0.9204,
+      "step": 665000
+    },
+    {
+      "epoch": 2.651457803772202,
+      "grad_norm": 1.6152838468551636,
+      "learning_rate": 5.809036603796638e-06,
+      "loss": 0.9164,
+      "step": 665500
+    },
+    {
+      "epoch": 2.653449883264142,
+      "grad_norm": 1.836296558380127,
+      "learning_rate": 5.775835278930971e-06,
+      "loss": 0.9199,
+      "step": 666000
+    },
+    {
+      "epoch": 2.6554419627560817,
+      "grad_norm": 1.6997501850128174,
+      "learning_rate": 5.742633954065303e-06,
+      "loss": 0.9193,
+      "step": 666500
+    },
+    {
+      "epoch": 2.657434042248022,
+      "grad_norm": 1.6507128477096558,
+      "learning_rate": 5.709432629199636e-06,
+      "loss": 0.9222,
+      "step": 667000
+    },
+    {
+      "epoch": 2.659426121739962,
+      "grad_norm": 1.6997631788253784,
+      "learning_rate": 5.676231304333968e-06,
+      "loss": 0.9159,
+      "step": 667500
+    },
+    {
+      "epoch": 2.661418201231902,
+      "grad_norm": 1.7037001848220825,
+      "learning_rate": 5.643029979468301e-06,
+      "loss": 0.9196,
+      "step": 668000
+    },
+    {
+      "epoch": 2.6634102807238422,
+      "grad_norm": 1.6910909414291382,
+      "learning_rate": 5.6098286546026335e-06,
+      "loss": 0.9158,
+      "step": 668500
+    },
+    {
+      "epoch": 2.665402360215782,
+      "grad_norm": 1.608192801475525,
+      "learning_rate": 5.576627329736966e-06,
+      "loss": 0.9221,
+      "step": 669000
+    },
+    {
+      "epoch": 2.667394439707722,
+      "grad_norm": 1.7195848226547241,
+      "learning_rate": 5.543426004871299e-06,
+      "loss": 0.9191,
+      "step": 669500
+    },
+    {
+      "epoch": 2.669386519199662,
+      "grad_norm": 1.7445330619812012,
+      "learning_rate": 5.510224680005631e-06,
+      "loss": 0.9189,
+      "step": 670000
+    },
+    {
+      "epoch": 2.6713785986916023,
+      "grad_norm": 1.6905847787857056,
+      "learning_rate": 5.477023355139964e-06,
+      "loss": 0.918,
+      "step": 670500
+    },
+    {
+      "epoch": 2.6733706781835425,
+      "grad_norm": 1.6639331579208374,
+      "learning_rate": 5.443822030274296e-06,
+      "loss": 0.9215,
+      "step": 671000
+    },
+    {
+      "epoch": 2.675362757675482,
+      "grad_norm": 1.6434050798416138,
+      "learning_rate": 5.410620705408629e-06,
+      "loss": 0.9168,
+      "step": 671500
+    },
+    {
+      "epoch": 2.6773548371674223,
+      "grad_norm": 1.684760570526123,
+      "learning_rate": 5.377419380542962e-06,
+      "loss": 0.92,
+      "step": 672000
+    },
+    {
+      "epoch": 2.6793469166593624,
+      "grad_norm": 1.7400184869766235,
+      "learning_rate": 5.344218055677294e-06,
+      "loss": 0.9174,
+      "step": 672500
+    },
+    {
+      "epoch": 2.6813389961513026,
+      "grad_norm": 1.6050313711166382,
+      "learning_rate": 5.311016730811626e-06,
+      "loss": 0.9187,
+      "step": 673000
+    },
+    {
+      "epoch": 2.6833310756432427,
+      "grad_norm": 1.6722040176391602,
+      "learning_rate": 5.277815405945959e-06,
+      "loss": 0.9196,
+      "step": 673500
+    },
+    {
+      "epoch": 2.6853231551351824,
+      "grad_norm": 1.6967012882232666,
+      "learning_rate": 5.244614081080291e-06,
+      "loss": 0.9184,
+      "step": 674000
+    },
+    {
+      "epoch": 2.6873152346271225,
+      "grad_norm": 1.8413442373275757,
+      "learning_rate": 5.2114127562146245e-06,
+      "loss": 0.9197,
+      "step": 674500
+    },
+    {
+      "epoch": 2.6893073141190627,
+      "grad_norm": 1.6100513935089111,
+      "learning_rate": 5.178211431348957e-06,
+      "loss": 0.9152,
+      "step": 675000
+    },
+    {
+      "epoch": 2.691299393611003,
+      "grad_norm": 1.6033347845077515,
+      "learning_rate": 5.14501010648329e-06,
+      "loss": 0.9173,
+      "step": 675500
+    },
+    {
+      "epoch": 2.693291473102943,
+      "grad_norm": 1.6931695938110352,
+      "learning_rate": 5.111808781617622e-06,
+      "loss": 0.9172,
+      "step": 676000
+    },
+    {
+      "epoch": 2.6952835525948826,
+      "grad_norm": 1.7061998844146729,
+      "learning_rate": 5.078607456751955e-06,
+      "loss": 0.9171,
+      "step": 676500
+    },
+    {
+      "epoch": 2.6972756320868227,
+      "grad_norm": 1.630188226699829,
+      "learning_rate": 5.045406131886287e-06,
+      "loss": 0.9158,
+      "step": 677000
+    },
+    {
+      "epoch": 2.699267711578763,
+      "grad_norm": 1.6837302446365356,
+      "learning_rate": 5.0122048070206196e-06,
+      "loss": 0.9181,
+      "step": 677500
+    },
+    {
+      "epoch": 2.701259791070703,
+      "grad_norm": 1.6485655307769775,
+      "learning_rate": 4.979003482154953e-06,
+      "loss": 0.9138,
+      "step": 678000
+    },
+    {
+      "epoch": 2.703251870562643,
+      "grad_norm": 1.706154227256775,
+      "learning_rate": 4.945802157289285e-06,
+      "loss": 0.9171,
+      "step": 678500
+    },
+    {
+      "epoch": 2.705243950054583,
+      "grad_norm": 1.7595213651657104,
+      "learning_rate": 4.912600832423617e-06,
+      "loss": 0.9191,
+      "step": 679000
+    },
+    {
+      "epoch": 2.707236029546523,
+      "grad_norm": 1.6348868608474731,
+      "learning_rate": 4.879399507557949e-06,
+      "loss": 0.9225,
+      "step": 679500
+    },
+    {
+      "epoch": 2.709228109038463,
+      "grad_norm": 1.9276422262191772,
+      "learning_rate": 4.846198182692282e-06,
+      "loss": 0.9139,
+      "step": 680000
+    },
+    {
+      "epoch": 2.7112201885304033,
+      "grad_norm": 1.7091796398162842,
+      "learning_rate": 4.812996857826615e-06,
+      "loss": 0.9159,
+      "step": 680500
+    },
+    {
+      "epoch": 2.7132122680223434,
+      "grad_norm": 1.6942174434661865,
+      "learning_rate": 4.779795532960948e-06,
+      "loss": 0.9198,
+      "step": 681000
+    },
+    {
+      "epoch": 2.715204347514283,
+      "grad_norm": 1.5856685638427734,
+      "learning_rate": 4.74659420809528e-06,
+      "loss": 0.9182,
+      "step": 681500
+    },
+    {
+      "epoch": 2.717196427006223,
+      "grad_norm": 1.7248765230178833,
+      "learning_rate": 4.713392883229613e-06,
+      "loss": 0.9189,
+      "step": 682000
+    },
+    {
+      "epoch": 2.7191885064981633,
+      "grad_norm": 1.717492699623108,
+      "learning_rate": 4.680191558363945e-06,
+      "loss": 0.9158,
+      "step": 682500
+    },
+    {
+      "epoch": 2.7211805859901035,
+      "grad_norm": 1.5787744522094727,
+      "learning_rate": 4.646990233498278e-06,
+      "loss": 0.9143,
+      "step": 683000
+    },
+    {
+      "epoch": 2.7231726654820436,
+      "grad_norm": 1.7174098491668701,
+      "learning_rate": 4.6137889086326106e-06,
+      "loss": 0.9149,
+      "step": 683500
+    },
+    {
+      "epoch": 2.7251647449739833,
+      "grad_norm": 1.6954288482666016,
+      "learning_rate": 4.580587583766943e-06,
+      "loss": 0.914,
+      "step": 684000
+    },
+    {
+      "epoch": 2.7271568244659234,
+      "grad_norm": 1.6374183893203735,
+      "learning_rate": 4.547386258901275e-06,
+      "loss": 0.9165,
+      "step": 684500
+    },
+    {
+      "epoch": 2.7291489039578636,
+      "grad_norm": 1.691360592842102,
+      "learning_rate": 4.514184934035607e-06,
+      "loss": 0.9139,
+      "step": 685000
+    },
+    {
+      "epoch": 2.7311409834498037,
+      "grad_norm": 1.669922947883606,
+      "learning_rate": 4.48098360916994e-06,
+      "loss": 0.9175,
+      "step": 685500
+    },
+    {
+      "epoch": 2.733133062941744,
+      "grad_norm": 1.6336150169372559,
+      "learning_rate": 4.4477822843042726e-06,
+      "loss": 0.9198,
+      "step": 686000
+    },
+    {
+      "epoch": 2.7351251424336835,
+      "grad_norm": 1.7089776992797852,
+      "learning_rate": 4.414580959438606e-06,
+      "loss": 0.9167,
+      "step": 686500
+    },
+    {
+      "epoch": 2.7371172219256237,
+      "grad_norm": 1.8231812715530396,
+      "learning_rate": 4.381379634572938e-06,
+      "loss": 0.9181,
+      "step": 687000
+    },
+    {
+      "epoch": 2.739109301417564,
+      "grad_norm": 1.7484509944915771,
+      "learning_rate": 4.348178309707271e-06,
+      "loss": 0.9162,
+      "step": 687500
+    },
+    {
+      "epoch": 2.741101380909504,
+      "grad_norm": 1.6094883680343628,
+      "learning_rate": 4.314976984841603e-06,
+      "loss": 0.9108,
+      "step": 688000
+    },
+    {
+      "epoch": 2.743093460401444,
+      "grad_norm": 1.6755807399749756,
+      "learning_rate": 4.281775659975936e-06,
+      "loss": 0.915,
+      "step": 688500
+    },
+    {
+      "epoch": 2.7450855398933838,
+      "grad_norm": 1.7994046211242676,
+      "learning_rate": 4.2485743351102685e-06,
+      "loss": 0.9156,
+      "step": 689000
+    },
+    {
+      "epoch": 2.747077619385324,
+      "grad_norm": 1.6629095077514648,
+      "learning_rate": 4.2153730102446016e-06,
+      "loss": 0.9126,
+      "step": 689500
+    },
+    {
+      "epoch": 2.749069698877264,
+      "grad_norm": 1.7381521463394165,
+      "learning_rate": 4.182171685378934e-06,
+      "loss": 0.9151,
+      "step": 690000
+    },
+    {
+      "epoch": 2.751061778369204,
+      "grad_norm": 1.7194076776504517,
+      "learning_rate": 4.148970360513266e-06,
+      "loss": 0.9141,
+      "step": 690500
+    },
+    {
+      "epoch": 2.7530538578611443,
+      "grad_norm": 1.6452442407608032,
+      "learning_rate": 4.115769035647598e-06,
+      "loss": 0.9124,
+      "step": 691000
+    },
+    {
+      "epoch": 2.755045937353084,
+      "grad_norm": 1.6988683938980103,
+      "learning_rate": 4.082567710781931e-06,
+      "loss": 0.9153,
+      "step": 691500
+    },
+    {
+      "epoch": 2.757038016845024,
+      "grad_norm": 1.7242997884750366,
+      "learning_rate": 4.0493663859162636e-06,
+      "loss": 0.9138,
+      "step": 692000
+    },
+    {
+      "epoch": 2.7590300963369643,
+      "grad_norm": 1.6954371929168701,
+      "learning_rate": 4.016165061050596e-06,
+      "loss": 0.9129,
+      "step": 692500
+    },
+    {
+      "epoch": 2.7610221758289044,
+      "grad_norm": 1.7216899394989014,
+      "learning_rate": 3.982963736184929e-06,
+      "loss": 0.9186,
+      "step": 693000
+    },
+    {
+      "epoch": 2.7630142553208445,
+      "grad_norm": 1.586531639099121,
+      "learning_rate": 3.949762411319261e-06,
+      "loss": 0.9152,
+      "step": 693500
+    },
+    {
+      "epoch": 2.7650063348127842,
+      "grad_norm": 1.5927194356918335,
+      "learning_rate": 3.916561086453594e-06,
+      "loss": 0.916,
+      "step": 694000
+    },
+    {
+      "epoch": 2.7669984143047244,
+      "grad_norm": 1.607954978942871,
+      "learning_rate": 3.883359761587926e-06,
+      "loss": 0.9195,
+      "step": 694500
+    },
+    {
+      "epoch": 2.7689904937966645,
+      "grad_norm": 1.6423405408859253,
+      "learning_rate": 3.8501584367222595e-06,
+      "loss": 0.9179,
+      "step": 695000
+    },
+    {
+      "epoch": 2.7709825732886046,
+      "grad_norm": 1.5982316732406616,
+      "learning_rate": 3.816957111856592e-06,
+      "loss": 0.9182,
+      "step": 695500
+    },
+    {
+      "epoch": 2.772974652780545,
+      "grad_norm": 1.735586166381836,
+      "learning_rate": 3.7837557869909244e-06,
+      "loss": 0.9123,
+      "step": 696000
+    },
+    {
+      "epoch": 2.7749667322724845,
+      "grad_norm": 1.7213603258132935,
+      "learning_rate": 3.7505544621252566e-06,
+      "loss": 0.9133,
+      "step": 696500
+    },
+    {
+      "epoch": 2.7769588117644246,
+      "grad_norm": 1.6382230520248413,
+      "learning_rate": 3.7173531372595897e-06,
+      "loss": 0.9162,
+      "step": 697000
+    },
+    {
+      "epoch": 2.7789508912563647,
+      "grad_norm": 1.647326946258545,
+      "learning_rate": 3.684151812393922e-06,
+      "loss": 0.9143,
+      "step": 697500
+    },
+    {
+      "epoch": 2.780942970748305,
+      "grad_norm": 1.7565851211547852,
+      "learning_rate": 3.6509504875282546e-06,
+      "loss": 0.9164,
+      "step": 698000
+    },
+    {
+      "epoch": 2.782935050240245,
+      "grad_norm": 1.6492735147476196,
+      "learning_rate": 3.6177491626625872e-06,
+      "loss": 0.9133,
+      "step": 698500
+    },
+    {
+      "epoch": 2.7849271297321847,
+      "grad_norm": 1.7115992307662964,
+      "learning_rate": 3.58454783779692e-06,
+      "loss": 0.9138,
+      "step": 699000
+    },
+    {
+      "epoch": 2.786919209224125,
+      "grad_norm": 1.6319609880447388,
+      "learning_rate": 3.551346512931252e-06,
+      "loss": 0.9103,
+      "step": 699500
+    },
+    {
+      "epoch": 2.788911288716065,
+      "grad_norm": 1.7215720415115356,
+      "learning_rate": 3.5181451880655843e-06,
+      "loss": 0.9152,
+      "step": 700000
+    },
+    {
+      "epoch": 2.790903368208005,
+      "grad_norm": 1.7807778120040894,
+      "learning_rate": 3.4849438631999174e-06,
+      "loss": 0.9175,
+      "step": 700500
+    },
+    {
+      "epoch": 2.7928954476999452,
+      "grad_norm": 1.6248434782028198,
+      "learning_rate": 3.4517425383342496e-06,
+      "loss": 0.9151,
+      "step": 701000
+    },
+    {
+      "epoch": 2.794887527191885,
+      "grad_norm": 1.6597427129745483,
+      "learning_rate": 3.4185412134685823e-06,
+      "loss": 0.9169,
+      "step": 701500
+    },
+    {
+      "epoch": 2.796879606683825,
+      "grad_norm": 1.630365252494812,
+      "learning_rate": 3.3853398886029145e-06,
+      "loss": 0.9108,
+      "step": 702000
+    },
+    {
+      "epoch": 2.798871686175765,
+      "grad_norm": 1.6231971979141235,
+      "learning_rate": 3.3521385637372476e-06,
+      "loss": 0.9111,
+      "step": 702500
+    },
+    {
+      "epoch": 2.8008637656677053,
+      "grad_norm": 1.7414125204086304,
+      "learning_rate": 3.31893723887158e-06,
+      "loss": 0.9143,
+      "step": 703000
+    },
+    {
+      "epoch": 2.8028558451596455,
+      "grad_norm": 1.5917302370071411,
+      "learning_rate": 3.285735914005913e-06,
+      "loss": 0.9165,
+      "step": 703500
+    },
+    {
+      "epoch": 2.804847924651585,
+      "grad_norm": 1.6118699312210083,
+      "learning_rate": 3.252534589140245e-06,
+      "loss": 0.914,
+      "step": 704000
+    },
+    {
+      "epoch": 2.8068400041435253,
+      "grad_norm": 1.638920783996582,
+      "learning_rate": 3.219333264274578e-06,
+      "loss": 0.9131,
+      "step": 704500
+    },
+    {
+      "epoch": 2.8088320836354654,
+      "grad_norm": 1.7069677114486694,
+      "learning_rate": 3.18613193940891e-06,
+      "loss": 0.9155,
+      "step": 705000
+    },
+    {
+      "epoch": 2.8108241631274056,
+      "grad_norm": 1.7749594449996948,
+      "learning_rate": 3.152930614543243e-06,
+      "loss": 0.9153,
+      "step": 705500
+    },
+    {
+      "epoch": 2.8128162426193457,
+      "grad_norm": 1.6838828325271606,
+      "learning_rate": 3.1197292896775753e-06,
+      "loss": 0.9149,
+      "step": 706000
+    },
+    {
+      "epoch": 2.8148083221112854,
+      "grad_norm": 1.6687116622924805,
+      "learning_rate": 3.086527964811908e-06,
+      "loss": 0.9102,
+      "step": 706500
+    },
+    {
+      "epoch": 2.8168004016032255,
+      "grad_norm": 1.686370611190796,
+      "learning_rate": 3.0533266399462406e-06,
+      "loss": 0.9208,
+      "step": 707000
+    },
+    {
+      "epoch": 2.8187924810951657,
+      "grad_norm": 1.7233439683914185,
+      "learning_rate": 3.0201253150805733e-06,
+      "loss": 0.9118,
+      "step": 707500
+    },
+    {
+      "epoch": 2.820784560587106,
+      "grad_norm": 1.6567538976669312,
+      "learning_rate": 2.9869239902149055e-06,
+      "loss": 0.9104,
+      "step": 708000
+    },
+    {
+      "epoch": 2.822776640079046,
+      "grad_norm": 1.6241168975830078,
+      "learning_rate": 2.953722665349238e-06,
+      "loss": 0.9155,
+      "step": 708500
+    },
+    {
+      "epoch": 2.8247687195709856,
+      "grad_norm": 1.6660428047180176,
+      "learning_rate": 2.920521340483571e-06,
+      "loss": 0.9133,
+      "step": 709000
+    },
+    {
+      "epoch": 2.8267607990629258,
+      "grad_norm": 1.617960810661316,
+      "learning_rate": 2.8873200156179035e-06,
+      "loss": 0.9161,
+      "step": 709500
+    },
+    {
+      "epoch": 2.828752878554866,
+      "grad_norm": 1.7162351608276367,
+      "learning_rate": 2.854118690752236e-06,
+      "loss": 0.9151,
+      "step": 710000
+    },
+    {
+      "epoch": 2.830744958046806,
+      "grad_norm": 1.6791682243347168,
+      "learning_rate": 2.820917365886569e-06,
+      "loss": 0.916,
+      "step": 710500
+    },
+    {
+      "epoch": 2.832737037538746,
+      "grad_norm": 1.703940749168396,
+      "learning_rate": 2.787716041020901e-06,
+      "loss": 0.9177,
+      "step": 711000
+    },
+    {
+      "epoch": 2.834729117030686,
+      "grad_norm": 1.7650041580200195,
+      "learning_rate": 2.7545147161552333e-06,
+      "loss": 0.9151,
+      "step": 711500
+    },
+    {
+      "epoch": 2.836721196522626,
+      "grad_norm": 1.5798845291137695,
+      "learning_rate": 2.721313391289566e-06,
+      "loss": 0.9175,
+      "step": 712000
+    },
+    {
+      "epoch": 2.838713276014566,
+      "grad_norm": 1.5189650058746338,
+      "learning_rate": 2.6881120664238986e-06,
+      "loss": 0.9124,
+      "step": 712500
+    },
+    {
+      "epoch": 2.8407053555065063,
+      "grad_norm": 1.6569610834121704,
+      "learning_rate": 2.6549107415582312e-06,
+      "loss": 0.9141,
+      "step": 713000
+    },
+    {
+      "epoch": 2.8426974349984464,
+      "grad_norm": 1.686488389968872,
+      "learning_rate": 2.621709416692564e-06,
+      "loss": 0.9123,
+      "step": 713500
+    },
+    {
+      "epoch": 2.844689514490386,
+      "grad_norm": 1.6028521060943604,
+      "learning_rate": 2.5885080918268965e-06,
+      "loss": 0.9117,
+      "step": 714000
+    },
+    {
+      "epoch": 2.8466815939823262,
+      "grad_norm": 1.7710810899734497,
+      "learning_rate": 2.5553067669612288e-06,
+      "loss": 0.9154,
+      "step": 714500
+    },
+    {
+      "epoch": 2.8486736734742664,
+      "grad_norm": 1.671952247619629,
+      "learning_rate": 2.5221054420955614e-06,
+      "loss": 0.9157,
+      "step": 715000
+    },
+    {
+      "epoch": 2.8506657529662065,
+      "grad_norm": 1.6095607280731201,
+      "learning_rate": 2.488904117229894e-06,
+      "loss": 0.914,
+      "step": 715500
+    },
+    {
+      "epoch": 2.8526578324581466,
+      "grad_norm": 1.6664694547653198,
+      "learning_rate": 2.4557027923642267e-06,
+      "loss": 0.9093,
+      "step": 716000
+    },
+    {
+      "epoch": 2.8546499119500863,
+      "grad_norm": 1.6126105785369873,
+      "learning_rate": 2.4225014674985594e-06,
+      "loss": 0.9077,
+      "step": 716500
+    },
+    {
+      "epoch": 2.8566419914420265,
+      "grad_norm": 1.6313860416412354,
+      "learning_rate": 2.389300142632892e-06,
+      "loss": 0.9172,
+      "step": 717000
+    },
+    {
+      "epoch": 2.8586340709339666,
+      "grad_norm": 1.65628182888031,
+      "learning_rate": 2.3560988177672243e-06,
+      "loss": 0.9117,
+      "step": 717500
+    },
+    {
+      "epoch": 2.8606261504259067,
+      "grad_norm": 1.621504306793213,
+      "learning_rate": 2.322897492901557e-06,
+      "loss": 0.9131,
+      "step": 718000
+    },
+    {
+      "epoch": 2.862618229917847,
+      "grad_norm": 1.6166635751724243,
+      "learning_rate": 2.2896961680358896e-06,
+      "loss": 0.9091,
+      "step": 718500
+    },
+    {
+      "epoch": 2.8646103094097866,
+      "grad_norm": 1.732576847076416,
+      "learning_rate": 2.256494843170222e-06,
+      "loss": 0.9131,
+      "step": 719000
+    },
+    {
+      "epoch": 2.8666023889017267,
+      "grad_norm": 1.671106219291687,
+      "learning_rate": 2.2232935183045545e-06,
+      "loss": 0.9126,
+      "step": 719500
+    },
+    {
+      "epoch": 2.868594468393667,
+      "grad_norm": 1.6212049722671509,
+      "learning_rate": 2.190092193438887e-06,
+      "loss": 0.9082,
+      "step": 720000
+    },
+    {
+      "epoch": 2.870586547885607,
+      "grad_norm": 1.747321605682373,
+      "learning_rate": 2.1568908685732193e-06,
+      "loss": 0.914,
+      "step": 720500
+    },
+    {
+      "epoch": 2.872578627377547,
+      "grad_norm": 1.6470304727554321,
+      "learning_rate": 2.123689543707552e-06,
+      "loss": 0.9078,
+      "step": 721000
+    },
+    {
+      "epoch": 2.874570706869487,
+      "grad_norm": 1.6893097162246704,
+      "learning_rate": 2.0904882188418846e-06,
+      "loss": 0.9071,
+      "step": 721500
+    },
+    {
+      "epoch": 2.876562786361427,
+      "grad_norm": 1.6795344352722168,
+      "learning_rate": 2.0572868939762173e-06,
+      "loss": 0.9086,
+      "step": 722000
+    },
+    {
+      "epoch": 2.878554865853367,
+      "grad_norm": 1.7055758237838745,
+      "learning_rate": 2.02408556911055e-06,
+      "loss": 0.9091,
+      "step": 722500
+    },
+    {
+      "epoch": 2.880546945345307,
+      "grad_norm": 1.667500615119934,
+      "learning_rate": 1.9908842442448826e-06,
+      "loss": 0.9094,
+      "step": 723000
+    },
+    {
+      "epoch": 2.8825390248372473,
+      "grad_norm": 1.6084439754486084,
+      "learning_rate": 1.957682919379215e-06,
+      "loss": 0.9115,
+      "step": 723500
+    },
+    {
+      "epoch": 2.884531104329187,
+      "grad_norm": 1.7185031175613403,
+      "learning_rate": 1.9244815945135475e-06,
+      "loss": 0.9148,
+      "step": 724000
+    },
+    {
+      "epoch": 2.886523183821127,
+      "grad_norm": 1.758104920387268,
+      "learning_rate": 1.8912802696478802e-06,
+      "loss": 0.9123,
+      "step": 724500
+    },
+    {
+      "epoch": 2.8885152633130673,
+      "grad_norm": 1.6831014156341553,
+      "learning_rate": 1.8580789447822128e-06,
+      "loss": 0.9103,
+      "step": 725000
+    },
+    {
+      "epoch": 2.8905073428050074,
+      "grad_norm": 1.7254362106323242,
+      "learning_rate": 1.8248776199165455e-06,
+      "loss": 0.914,
+      "step": 725500
+    },
+    {
+      "epoch": 2.8924994222969476,
+      "grad_norm": 1.7268210649490356,
+      "learning_rate": 1.791676295050878e-06,
+      "loss": 0.9105,
+      "step": 726000
+    },
+    {
+      "epoch": 2.8944915017888873,
+      "grad_norm": 1.7437758445739746,
+      "learning_rate": 1.7584749701852106e-06,
+      "loss": 0.9097,
+      "step": 726500
+    },
+    {
+      "epoch": 2.8964835812808274,
+      "grad_norm": 1.6752897500991821,
+      "learning_rate": 1.7252736453195428e-06,
+      "loss": 0.9106,
+      "step": 727000
+    },
+    {
+      "epoch": 2.8984756607727675,
+      "grad_norm": 1.7121684551239014,
+      "learning_rate": 1.6920723204538754e-06,
+      "loss": 0.9149,
+      "step": 727500
+    },
+    {
+      "epoch": 2.9004677402647077,
+      "grad_norm": 1.6758103370666504,
+      "learning_rate": 1.6588709955882079e-06,
+      "loss": 0.9118,
+      "step": 728000
+    },
+    {
+      "epoch": 2.902459819756648,
+      "grad_norm": 1.573196291923523,
+      "learning_rate": 1.6256696707225405e-06,
+      "loss": 0.914,
+      "step": 728500
+    },
+    {
+      "epoch": 2.9044518992485875,
+      "grad_norm": 1.6836477518081665,
+      "learning_rate": 1.5924683458568732e-06,
+      "loss": 0.9028,
+      "step": 729000
+    },
+    {
+      "epoch": 2.9064439787405276,
+      "grad_norm": 1.6682339906692505,
+      "learning_rate": 1.5592670209912056e-06,
+      "loss": 0.9135,
+      "step": 729500
+    },
+    {
+      "epoch": 2.9084360582324678,
+      "grad_norm": 1.7347893714904785,
+      "learning_rate": 1.5260656961255383e-06,
+      "loss": 0.9142,
+      "step": 730000
+    },
+    {
+      "epoch": 2.910428137724408,
+      "grad_norm": 1.740942358970642,
+      "learning_rate": 1.4928643712598707e-06,
+      "loss": 0.9097,
+      "step": 730500
+    },
+    {
+      "epoch": 2.912420217216348,
+      "grad_norm": 1.715808629989624,
+      "learning_rate": 1.4596630463942034e-06,
+      "loss": 0.909,
+      "step": 731000
+    },
+    {
+      "epoch": 2.9144122967082877,
+      "grad_norm": 1.6233854293823242,
+      "learning_rate": 1.426461721528536e-06,
+      "loss": 0.9103,
+      "step": 731500
+    },
+    {
+      "epoch": 2.916404376200228,
+      "grad_norm": 1.6787744760513306,
+      "learning_rate": 1.3932603966628685e-06,
+      "loss": 0.914,
+      "step": 732000
+    },
+    {
+      "epoch": 2.918396455692168,
+      "grad_norm": 1.6370259523391724,
+      "learning_rate": 1.360059071797201e-06,
+      "loss": 0.9112,
+      "step": 732500
+    },
+    {
+      "epoch": 2.920388535184108,
+      "grad_norm": 1.7400423288345337,
+      "learning_rate": 1.3268577469315336e-06,
+      "loss": 0.9152,
+      "step": 733000
+    },
+    {
+      "epoch": 2.9223806146760483,
+      "grad_norm": 1.6786566972732544,
+      "learning_rate": 1.2936564220658662e-06,
+      "loss": 0.9106,
+      "step": 733500
+    },
+    {
+      "epoch": 2.924372694167988,
+      "grad_norm": 1.6905065774917603,
+      "learning_rate": 1.2604550972001987e-06,
+      "loss": 0.9115,
+      "step": 734000
+    },
+    {
+      "epoch": 2.926364773659928,
+      "grad_norm": 1.7250709533691406,
+      "learning_rate": 1.2272537723345313e-06,
+      "loss": 0.912,
+      "step": 734500
+    },
+    {
+      "epoch": 2.9283568531518682,
+      "grad_norm": 1.6732368469238281,
+      "learning_rate": 1.194052447468864e-06,
+      "loss": 0.9077,
+      "step": 735000
+    },
+    {
+      "epoch": 2.9303489326438084,
+      "grad_norm": 1.6451321840286255,
+      "learning_rate": 1.1608511226031964e-06,
+      "loss": 0.9096,
+      "step": 735500
+    },
+    {
+      "epoch": 2.9323410121357485,
+      "grad_norm": 1.693692922592163,
+      "learning_rate": 1.127649797737529e-06,
+      "loss": 0.903,
+      "step": 736000
+    },
+    {
+      "epoch": 2.934333091627688,
+      "grad_norm": 1.722675085067749,
+      "learning_rate": 1.0944484728718615e-06,
+      "loss": 0.9053,
+      "step": 736500
+    },
+    {
+      "epoch": 2.9363251711196283,
+      "grad_norm": 1.7938162088394165,
+      "learning_rate": 1.061247148006194e-06,
+      "loss": 0.9142,
+      "step": 737000
+    },
+    {
+      "epoch": 2.9383172506115685,
+      "grad_norm": 1.6571873426437378,
+      "learning_rate": 1.0280458231405266e-06,
+      "loss": 0.9161,
+      "step": 737500
+    },
+    {
+      "epoch": 2.9403093301035086,
+      "grad_norm": 1.6131647825241089,
+      "learning_rate": 9.948444982748593e-07,
+      "loss": 0.9118,
+      "step": 738000
+    },
+    {
+      "epoch": 2.9423014095954487,
+      "grad_norm": 1.6702022552490234,
+      "learning_rate": 9.616431734091917e-07,
+      "loss": 0.9125,
+      "step": 738500
+    },
+    {
+      "epoch": 2.9442934890873884,
+      "grad_norm": 1.7070099115371704,
+      "learning_rate": 9.284418485435244e-07,
+      "loss": 0.9102,
+      "step": 739000
+    },
+    {
+      "epoch": 2.9462855685793286,
+      "grad_norm": 1.6868396997451782,
+      "learning_rate": 8.952405236778569e-07,
+      "loss": 0.9107,
+      "step": 739500
+    },
+    {
+      "epoch": 2.9482776480712687,
+      "grad_norm": 1.6585192680358887,
+      "learning_rate": 8.620391988121894e-07,
+      "loss": 0.9071,
+      "step": 740000
+    },
+    {
+      "epoch": 2.950269727563209,
+      "grad_norm": 1.6128840446472168,
+      "learning_rate": 8.288378739465219e-07,
+      "loss": 0.9115,
+      "step": 740500
+    },
+    {
+      "epoch": 2.952261807055149,
+      "grad_norm": 1.65033757686615,
+      "learning_rate": 7.956365490808546e-07,
+      "loss": 0.9122,
+      "step": 741000
+    },
+    {
+      "epoch": 2.9542538865470886,
+      "grad_norm": 1.7108838558197021,
+      "learning_rate": 7.624352242151871e-07,
+      "loss": 0.9065,
+      "step": 741500
+    },
+    {
+      "epoch": 2.956245966039029,
+      "grad_norm": 1.7979404926300049,
+      "learning_rate": 7.292338993495197e-07,
+      "loss": 0.9091,
+      "step": 742000
+    },
+    {
+      "epoch": 2.958238045530969,
+      "grad_norm": 1.7434169054031372,
+      "learning_rate": 6.960325744838523e-07,
+      "loss": 0.9093,
+      "step": 742500
+    },
+    {
+      "epoch": 2.960230125022909,
+      "grad_norm": 1.80579674243927,
+      "learning_rate": 6.628312496181848e-07,
+      "loss": 0.9077,
+      "step": 743000
+    },
+    {
+      "epoch": 2.962222204514849,
+      "grad_norm": 1.6637142896652222,
+      "learning_rate": 6.296299247525173e-07,
+      "loss": 0.9085,
+      "step": 743500
+    },
+    {
+      "epoch": 2.964214284006789,
+      "grad_norm": 1.6318128108978271,
+      "learning_rate": 5.9642859988685e-07,
+      "loss": 0.9076,
+      "step": 744000
+    },
+    {
+      "epoch": 2.966206363498729,
+      "grad_norm": 1.7178044319152832,
+      "learning_rate": 5.632272750211825e-07,
+      "loss": 0.9129,
+      "step": 744500
+    },
+    {
+      "epoch": 2.968198442990669,
+      "grad_norm": 1.7807506322860718,
+      "learning_rate": 5.300259501555149e-07,
+      "loss": 0.9156,
+      "step": 745000
+    },
+    {
+      "epoch": 2.9701905224826093,
+      "grad_norm": 1.7373597621917725,
+      "learning_rate": 4.968246252898476e-07,
+      "loss": 0.9127,
+      "step": 745500
+    },
+    {
+      "epoch": 2.9721826019745494,
+      "grad_norm": 1.755888819694519,
+      "learning_rate": 4.6362330042418015e-07,
+      "loss": 0.9059,
+      "step": 746000
+    },
+    {
+      "epoch": 2.974174681466489,
+      "grad_norm": 1.7097597122192383,
+      "learning_rate": 4.3042197555851275e-07,
+      "loss": 0.912,
+      "step": 746500
+    },
+    {
+      "epoch": 2.9761667609584292,
+      "grad_norm": 1.635866641998291,
+      "learning_rate": 3.9722065069284525e-07,
+      "loss": 0.911,
+      "step": 747000
+    },
+    {
+      "epoch": 2.9781588404503694,
+      "grad_norm": 1.703961968421936,
+      "learning_rate": 3.640193258271778e-07,
+      "loss": 0.9112,
+      "step": 747500
+    },
+    {
+      "epoch": 2.9801509199423095,
+      "grad_norm": 1.7729219198226929,
+      "learning_rate": 3.308180009615104e-07,
+      "loss": 0.9114,
+      "step": 748000
+    },
+    {
+      "epoch": 2.9821429994342497,
+      "grad_norm": 1.7777047157287598,
+      "learning_rate": 2.9761667609584294e-07,
+      "loss": 0.9071,
+      "step": 748500
+    },
+    {
+      "epoch": 2.9841350789261893,
+      "grad_norm": 1.7533235549926758,
+      "learning_rate": 2.6441535123017554e-07,
+      "loss": 0.9086,
+      "step": 749000
+    },
+    {
+      "epoch": 2.9861271584181295,
+      "grad_norm": 1.6186124086380005,
+      "learning_rate": 2.3121402636450807e-07,
+      "loss": 0.9134,
+      "step": 749500
+    },
+    {
+      "epoch": 2.9881192379100696,
+      "grad_norm": 1.7477833032608032,
+      "learning_rate": 1.9801270149884064e-07,
+      "loss": 0.908,
+      "step": 750000
+    },
+    {
+      "epoch": 2.9901113174020097,
+      "grad_norm": 1.656666874885559,
+      "learning_rate": 1.6481137663317316e-07,
+      "loss": 0.9091,
+      "step": 750500
+    },
+    {
+      "epoch": 2.99210339689395,
+      "grad_norm": 1.680490255355835,
+      "learning_rate": 1.3161005176750574e-07,
+      "loss": 0.9098,
+      "step": 751000
+    },
+    {
+      "epoch": 2.9940954763858896,
+      "grad_norm": 1.6490333080291748,
+      "learning_rate": 9.84087269018383e-08,
+      "loss": 0.9091,
+      "step": 751500
+    },
+    {
+      "epoch": 2.9960875558778297,
+      "grad_norm": 1.7370842695236206,
+      "learning_rate": 6.520740203617086e-08,
+      "loss": 0.9097,
+      "step": 752000
+    },
+    {
+      "epoch": 2.99807963536977,
+      "grad_norm": 1.711270809173584,
+      "learning_rate": 3.2006077170503415e-08,
+      "loss": 0.9158,
+      "step": 752500
+    },
+    {
+      "epoch": 3.0,
+      "step": 752982,
+      "total_flos": 1.2695212826516791e+19,
+      "train_loss": 1.0139643951673536,
+      "train_runtime": 533830.6605,
+      "train_samples_per_second": 90.274,
+      "train_steps_per_second": 1.411
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 752982,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2695212826516791e+19,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}