diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,3088 +3,6140 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 2179,
+  "global_step": 4358,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0004589261128958238,
-      "grad_norm": 8.542048983401715,
-      "learning_rate": 9.174311926605506e-08,
-      "loss": 1.1206,
+      "epoch": 0.0002294630564479119,
+      "grad_norm": 16.915573515504455,
+      "learning_rate": 4.587155963302753e-08,
+      "loss": 1.3608,
       "step": 1
     },
     {
-      "epoch": 0.002294630564479119,
-      "grad_norm": 8.730820381996422,
-      "learning_rate": 4.587155963302753e-07,
-      "loss": 1.1225,
+      "epoch": 0.0011473152822395595,
+      "grad_norm": 16.57824498398554,
+      "learning_rate": 2.2935779816513764e-07,
+      "loss": 1.3475,
       "step": 5
     },
     {
-      "epoch": 0.004589261128958238,
-      "grad_norm": 5.026231881904144,
-      "learning_rate": 9.174311926605506e-07,
-      "loss": 1.1138,
+      "epoch": 0.002294630564479119,
+      "grad_norm": 13.391120630574697,
+      "learning_rate": 4.587155963302753e-07,
+      "loss": 1.3563,
       "step": 10
     },
     {
-      "epoch": 0.006883891693437357,
-      "grad_norm": 3.214436315778165,
-      "learning_rate": 1.3761467889908258e-06,
-      "loss": 1.0308,
+      "epoch": 0.0034419458467186783,
+      "grad_norm": 7.451292451267116,
+      "learning_rate": 6.880733944954129e-07,
+      "loss": 1.2802,
       "step": 15
     },
     {
-      "epoch": 0.009178522257916476,
-      "grad_norm": 2.9262839132141174,
-      "learning_rate": 1.8348623853211011e-06,
-      "loss": 0.9803,
+      "epoch": 0.004589261128958238,
+      "grad_norm": 5.153183243132574,
+      "learning_rate": 9.174311926605506e-07,
+      "loss": 1.2213,
       "step": 20
     },
     {
-      "epoch": 0.011473152822395595,
-      "grad_norm": 2.5816500702829965,
-      "learning_rate": 2.2935779816513764e-06,
-      "loss": 0.9955,
+      "epoch": 0.005736576411197797,
+      "grad_norm": 4.42706429835042,
+      "learning_rate": 1.1467889908256882e-06,
+      "loss": 1.1605,
       "step": 25
     },
     {
-      "epoch": 0.013767783386874713,
-      "grad_norm": 2.62089651849134,
-      "learning_rate": 2.7522935779816517e-06,
-      "loss": 1.0214,
+      "epoch": 0.006883891693437357,
+      "grad_norm": 4.3909586491865,
+      "learning_rate": 1.3761467889908258e-06,
+      "loss": 1.1432,
       "step": 30
     },
     {
-      "epoch": 0.016062413951353834,
-      "grad_norm": 2.522275443963982,
-      "learning_rate": 3.211009174311927e-06,
-      "loss": 0.9751,
+      "epoch": 0.008031206975676917,
+      "grad_norm": 3.635568351265977,
+      "learning_rate": 1.6055045871559635e-06,
+      "loss": 1.1287,
       "step": 35
     },
     {
-      "epoch": 0.018357044515832952,
-      "grad_norm": 2.415617593047094,
-      "learning_rate": 3.6697247706422022e-06,
-      "loss": 0.995,
+      "epoch": 0.009178522257916476,
+      "grad_norm": 3.6782219530388676,
+      "learning_rate": 1.8348623853211011e-06,
+      "loss": 1.1319,
       "step": 40
     },
     {
-      "epoch": 0.02065167508031207,
-      "grad_norm": 2.525030642664337,
-      "learning_rate": 4.128440366972478e-06,
-      "loss": 1.0113,
+      "epoch": 0.010325837540156035,
+      "grad_norm": 3.9153495168079786,
+      "learning_rate": 2.064220183486239e-06,
+      "loss": 1.1259,
       "step": 45
     },
     {
-      "epoch": 0.02294630564479119,
-      "grad_norm": 2.6665037719940954,
-      "learning_rate": 4.587155963302753e-06,
-      "loss": 0.9655,
+      "epoch": 0.011473152822395595,
+      "grad_norm": 3.9184448958021036,
+      "learning_rate": 2.2935779816513764e-06,
+      "loss": 1.1435,
       "step": 50
     },
     {
-      "epoch": 0.025240936209270308,
-      "grad_norm": 3.3232412426313385,
-      "learning_rate": 5.045871559633028e-06,
-      "loss": 0.9874,
+      "epoch": 0.012620468104635154,
+      "grad_norm": 3.686838030224446,
+      "learning_rate": 2.522935779816514e-06,
+      "loss": 1.1297,
       "step": 55
     },
     {
-      "epoch": 0.027535566773749427,
-      "grad_norm": 2.394624410644964,
-      "learning_rate": 5.504587155963303e-06,
-      "loss": 0.9921,
+      "epoch": 0.013767783386874713,
+      "grad_norm": 3.849747711258154,
+      "learning_rate": 2.7522935779816517e-06,
+      "loss": 1.1342,
       "step": 60
     },
     {
-      "epoch": 0.029830197338228545,
-      "grad_norm": 2.6803374112332796,
-      "learning_rate": 5.963302752293578e-06,
-      "loss": 0.9968,
+      "epoch": 0.014915098669114273,
+      "grad_norm": 3.6560526277238443,
+      "learning_rate": 2.981651376146789e-06,
+      "loss": 1.113,
       "step": 65
     },
     {
-      "epoch": 0.03212482790270767,
-      "grad_norm": 2.695577014057411,
-      "learning_rate": 6.422018348623854e-06,
-      "loss": 0.9819,
+      "epoch": 0.016062413951353834,
+      "grad_norm": 4.12272475395816,
+      "learning_rate": 3.211009174311927e-06,
+      "loss": 1.1367,
       "step": 70
     },
     {
-      "epoch": 0.03441945846718678,
-      "grad_norm": 2.5003353553530308,
-      "learning_rate": 6.880733944954129e-06,
-      "loss": 1.0135,
+      "epoch": 0.01720972923359339,
+      "grad_norm": 3.6689121561670825,
+      "learning_rate": 3.4403669724770644e-06,
+      "loss": 1.1354,
       "step": 75
     },
     {
-      "epoch": 0.036714089031665904,
-      "grad_norm": 2.449197716174392,
-      "learning_rate": 7.3394495412844045e-06,
-      "loss": 0.9861,
+      "epoch": 0.018357044515832952,
+      "grad_norm": 3.5167783836221838,
+      "learning_rate": 3.6697247706422022e-06,
+      "loss": 1.1516,
       "step": 80
     },
     {
-      "epoch": 0.03900871959614502,
-      "grad_norm": 2.547680048482618,
-      "learning_rate": 7.79816513761468e-06,
-      "loss": 0.9886,
+      "epoch": 0.01950435979807251,
+      "grad_norm": 3.6409760141175,
+      "learning_rate": 3.89908256880734e-06,
+      "loss": 1.1447,
       "step": 85
     },
     {
-      "epoch": 0.04130335016062414,
-      "grad_norm": 2.738205329791908,
-      "learning_rate": 8.256880733944956e-06,
-      "loss": 0.9849,
+      "epoch": 0.02065167508031207,
+      "grad_norm": 3.5724528502767567,
+      "learning_rate": 4.128440366972478e-06,
+      "loss": 1.1103,
       "step": 90
     },
     {
-      "epoch": 0.04359798072510326,
-      "grad_norm": 2.6040796770347256,
-      "learning_rate": 8.71559633027523e-06,
-      "loss": 1.0033,
+      "epoch": 0.02179899036255163,
+      "grad_norm": 4.387301576122793,
+      "learning_rate": 4.357798165137615e-06,
+      "loss": 1.148,
       "step": 95
     },
     {
-      "epoch": 0.04589261128958238,
-      "grad_norm": 2.6163342737972495,
-      "learning_rate": 9.174311926605506e-06,
-      "loss": 0.9906,
+      "epoch": 0.02294630564479119,
+      "grad_norm": 3.5739499775971795,
+      "learning_rate": 4.587155963302753e-06,
+      "loss": 1.1412,
       "step": 100
     },
     {
-      "epoch": 0.048187241854061494,
-      "grad_norm": 2.4805649474435536,
-      "learning_rate": 9.633027522935781e-06,
-      "loss": 0.9953,
+      "epoch": 0.024093620927030747,
+      "grad_norm": 3.7629714710136244,
+      "learning_rate": 4.816513761467891e-06,
+      "loss": 1.1386,
       "step": 105
     },
     {
-      "epoch": 0.050481872418540616,
-      "grad_norm": 2.5415673257044604,
-      "learning_rate": 1.0091743119266055e-05,
-      "loss": 1.0171,
+      "epoch": 0.025240936209270308,
+      "grad_norm": 3.56307250426267,
+      "learning_rate": 5.045871559633028e-06,
+      "loss": 1.1265,
       "step": 110
     },
     {
-      "epoch": 0.05277650298301973,
-      "grad_norm": 2.4798532274771303,
-      "learning_rate": 1.055045871559633e-05,
-      "loss": 1.0314,
+      "epoch": 0.026388251491509866,
+      "grad_norm": 3.7511673022298617,
+      "learning_rate": 5.275229357798165e-06,
+      "loss": 1.1331,
       "step": 115
     },
     {
-      "epoch": 0.05507113354749885,
-      "grad_norm": 2.8712967477215217,
-      "learning_rate": 1.1009174311926607e-05,
-      "loss": 0.9966,
+      "epoch": 0.027535566773749427,
+      "grad_norm": 3.779191425409107,
+      "learning_rate": 5.504587155963303e-06,
+      "loss": 1.1324,
       "step": 120
     },
     {
-      "epoch": 0.05736576411197797,
-      "grad_norm": 2.40307484944342,
-      "learning_rate": 1.1467889908256882e-05,
-      "loss": 0.9862,
+      "epoch": 0.028682882055988984,
+      "grad_norm": 3.5751301446097252,
+      "learning_rate": 5.733944954128441e-06,
+      "loss": 1.1409,
       "step": 125
     },
     {
-      "epoch": 0.05966039467645709,
-      "grad_norm": 2.531947293972582,
-      "learning_rate": 1.1926605504587156e-05,
-      "loss": 1.0111,
+      "epoch": 0.029830197338228545,
+      "grad_norm": 3.7113311883773368,
+      "learning_rate": 5.963302752293578e-06,
+      "loss": 1.1522,
       "step": 130
     },
     {
-      "epoch": 0.06195502524093621,
-      "grad_norm": 2.5630237672720355,
-      "learning_rate": 1.238532110091743e-05,
-      "loss": 1.0111,
+      "epoch": 0.030977512620468106,
+      "grad_norm": 3.521541660293309,
+      "learning_rate": 6.192660550458715e-06,
+      "loss": 1.1657,
       "step": 135
     },
     {
-      "epoch": 0.06424965580541533,
-      "grad_norm": 3.0796021235291504,
-      "learning_rate": 1.2844036697247708e-05,
-      "loss": 0.996,
+      "epoch": 0.03212482790270767,
+      "grad_norm": 3.439820085380388,
+      "learning_rate": 6.422018348623854e-06,
+      "loss": 1.1607,
       "step": 140
     },
     {
-      "epoch": 0.06654428636989444,
-      "grad_norm": 2.532220737324273,
-      "learning_rate": 1.3302752293577984e-05,
-      "loss": 1.0361,
+      "epoch": 0.03327214318494722,
+      "grad_norm": 3.4447742850994842,
+      "learning_rate": 6.651376146788992e-06,
+      "loss": 1.1363,
       "step": 145
     },
     {
-      "epoch": 0.06883891693437356,
-      "grad_norm": 2.656655575725622,
-      "learning_rate": 1.3761467889908258e-05,
-      "loss": 1.0061,
+      "epoch": 0.03441945846718678,
+      "grad_norm": 3.5236191397811742,
+      "learning_rate": 6.880733944954129e-06,
+      "loss": 1.138,
       "step": 150
     },
     {
-      "epoch": 0.07113354749885269,
-      "grad_norm": 2.4257637087990194,
-      "learning_rate": 1.4220183486238533e-05,
-      "loss": 0.9896,
+      "epoch": 0.03556677374942634,
+      "grad_norm": 3.3663301898503017,
+      "learning_rate": 7.110091743119267e-06,
+      "loss": 1.1574,
       "step": 155
     },
     {
-      "epoch": 0.07342817806333181,
-      "grad_norm": 2.482384770473286,
-      "learning_rate": 1.4678899082568809e-05,
-      "loss": 1.0321,
+      "epoch": 0.036714089031665904,
+      "grad_norm": 3.360850229705706,
+      "learning_rate": 7.3394495412844045e-06,
+      "loss": 1.133,
       "step": 160
     },
     {
-      "epoch": 0.07572280862781092,
-      "grad_norm": 2.9466970045401957,
-      "learning_rate": 1.5137614678899085e-05,
-      "loss": 1.0113,
+      "epoch": 0.03786140431390546,
+      "grad_norm": 3.4791313578433702,
+      "learning_rate": 7.568807339449542e-06,
+      "loss": 1.1543,
       "step": 165
     },
     {
-      "epoch": 0.07801743919229004,
-      "grad_norm": 2.5190350812742364,
-      "learning_rate": 1.559633027522936e-05,
-      "loss": 1.0271,
+      "epoch": 0.03900871959614502,
+      "grad_norm": 3.7004977570629554,
+      "learning_rate": 7.79816513761468e-06,
+      "loss": 1.1934,
       "step": 170
     },
     {
-      "epoch": 0.08031206975676916,
-      "grad_norm": 2.478301477100018,
-      "learning_rate": 1.6055045871559634e-05,
-      "loss": 1.0165,
+      "epoch": 0.04015603487838458,
+      "grad_norm": 3.8223961646506903,
+      "learning_rate": 8.027522935779817e-06,
+      "loss": 1.1659,
       "step": 175
     },
     {
-      "epoch": 0.08260670032124828,
-      "grad_norm": 2.4464230838695826,
-      "learning_rate": 1.6513761467889912e-05,
-      "loss": 1.0646,
+      "epoch": 0.04130335016062414,
+      "grad_norm": 3.55532539649572,
+      "learning_rate": 8.256880733944956e-06,
+      "loss": 1.1128,
       "step": 180
     },
     {
-      "epoch": 0.08490133088572739,
-      "grad_norm": 2.5240574285299315,
-      "learning_rate": 1.6972477064220186e-05,
-      "loss": 1.0555,
+      "epoch": 0.042450665442863696,
+      "grad_norm": 3.6076069307268015,
+      "learning_rate": 8.486238532110093e-06,
+      "loss": 1.1569,
       "step": 185
     },
     {
-      "epoch": 0.08719596145020651,
-      "grad_norm": 2.4185989752236505,
-      "learning_rate": 1.743119266055046e-05,
-      "loss": 1.0451,
+      "epoch": 0.04359798072510326,
+      "grad_norm": 3.429437490724809,
+      "learning_rate": 8.71559633027523e-06,
+      "loss": 1.1521,
       "step": 190
     },
     {
-      "epoch": 0.08949059201468564,
-      "grad_norm": 2.6482287747026954,
-      "learning_rate": 1.7889908256880734e-05,
-      "loss": 1.0325,
+      "epoch": 0.04474529600734282,
+      "grad_norm": 3.234909828239376,
+      "learning_rate": 8.944954128440367e-06,
+      "loss": 1.1438,
       "step": 195
     },
     {
-      "epoch": 0.09178522257916476,
-      "grad_norm": 2.4286821504213814,
-      "learning_rate": 1.834862385321101e-05,
-      "loss": 1.0269,
+      "epoch": 0.04589261128958238,
+      "grad_norm": 3.3130400141238834,
+      "learning_rate": 9.174311926605506e-06,
+      "loss": 1.1706,
       "step": 200
     },
     {
-      "epoch": 0.09407985314364388,
-      "grad_norm": 2.193912154365014,
-      "learning_rate": 1.8807339449541285e-05,
-      "loss": 1.0551,
+      "epoch": 0.04703992657182194,
+      "grad_norm": 3.7263058619472673,
+      "learning_rate": 9.403669724770643e-06,
+      "loss": 1.1876,
       "step": 205
     },
     {
-      "epoch": 0.09637448370812299,
-      "grad_norm": 2.695401581193139,
-      "learning_rate": 1.9266055045871563e-05,
-      "loss": 1.0456,
+      "epoch": 0.048187241854061494,
+      "grad_norm": 3.4658866154835373,
+      "learning_rate": 9.633027522935781e-06,
+      "loss": 1.153,
       "step": 210
     },
     {
-      "epoch": 0.09866911427260211,
-      "grad_norm": 2.6583940178739827,
-      "learning_rate": 1.9724770642201837e-05,
-      "loss": 1.0664,
+      "epoch": 0.049334557136301055,
+      "grad_norm": 3.4639517157270086,
+      "learning_rate": 9.862385321100918e-06,
+      "loss": 1.185,
       "step": 215
     },
     {
-      "epoch": 0.10096374483708123,
-      "grad_norm": 2.371077606257049,
-      "learning_rate": 1.9999948669655127e-05,
-      "loss": 1.0578,
+      "epoch": 0.050481872418540616,
+      "grad_norm": 3.3739187071512937,
+      "learning_rate": 1.0091743119266055e-05,
+      "loss": 1.1948,
       "step": 220
     },
     {
-      "epoch": 0.10325837540156035,
-      "grad_norm": 2.159798506956808,
-      "learning_rate": 1.999937120932709e-05,
-      "loss": 1.0421,
+      "epoch": 0.05162918770078018,
+      "grad_norm": 3.3478939956905363,
+      "learning_rate": 1.0321100917431192e-05,
+      "loss": 1.1694,
       "step": 225
     },
     {
-      "epoch": 0.10555300596603946,
-      "grad_norm": 2.7274671704918516,
-      "learning_rate": 1.9998152162914807e-05,
-      "loss": 1.0513,
+      "epoch": 0.05277650298301973,
+      "grad_norm": 3.4368659867072773,
+      "learning_rate": 1.055045871559633e-05,
+      "loss": 1.1484,
       "step": 230
     },
     {
-      "epoch": 0.10784763653051858,
-      "grad_norm": 2.171023592361791,
-      "learning_rate": 1.9996291608635527e-05,
-      "loss": 1.0566,
+      "epoch": 0.05392381826525929,
+      "grad_norm": 3.2936522686474294,
+      "learning_rate": 1.077981651376147e-05,
+      "loss": 1.199,
       "step": 235
     },
     {
-      "epoch": 0.1101422670949977,
-      "grad_norm": 2.1701858629565445,
-      "learning_rate": 1.9993789665867316e-05,
-      "loss": 1.0371,
+      "epoch": 0.05507113354749885,
+      "grad_norm": 3.0900646767838165,
+      "learning_rate": 1.1009174311926607e-05,
+      "loss": 1.1961,
       "step": 240
     },
     {
-      "epoch": 0.11243689765947683,
-      "grad_norm": 2.2165028686751254,
-      "learning_rate": 1.9990646495141445e-05,
-      "loss": 1.0491,
+      "epoch": 0.056218448829738414,
+      "grad_norm": 3.422687599705482,
+      "learning_rate": 1.1238532110091744e-05,
+      "loss": 1.1988,
       "step": 245
     },
     {
-      "epoch": 0.11473152822395594,
-      "grad_norm": 2.6072713574574466,
-      "learning_rate": 1.998686229813205e-05,
-      "loss": 1.0636,
+      "epoch": 0.05736576411197797,
+      "grad_norm": 3.3144909495939103,
+      "learning_rate": 1.1467889908256882e-05,
+      "loss": 1.195,
       "step": 250
     },
     {
-      "epoch": 0.11702615878843506,
-      "grad_norm": 2.135919480089347,
-      "learning_rate": 1.9982437317643218e-05,
-      "loss": 1.05,
+      "epoch": 0.05851307939421753,
+      "grad_norm": 3.309877252453861,
+      "learning_rate": 1.169724770642202e-05,
+      "loss": 1.1785,
       "step": 255
     },
     {
-      "epoch": 0.11932078935291418,
-      "grad_norm": 2.4909036218466993,
-      "learning_rate": 1.9977371837593382e-05,
-      "loss": 1.0611,
+      "epoch": 0.05966039467645709,
+      "grad_norm": 3.280922712269594,
+      "learning_rate": 1.1926605504587156e-05,
+      "loss": 1.1776,
       "step": 260
     },
     {
-      "epoch": 0.1216154199173933,
-      "grad_norm": 5.355161964355881,
-      "learning_rate": 1.997166618299714e-05,
-      "loss": 1.0253,
+      "epoch": 0.06080770995869665,
+      "grad_norm": 3.341721826593927,
+      "learning_rate": 1.2155963302752293e-05,
+      "loss": 1.1948,
       "step": 265
     },
     {
-      "epoch": 0.12391005048187242,
-      "grad_norm": 2.374369183096664,
-      "learning_rate": 1.9965320719944366e-05,
-      "loss": 1.044,
+      "epoch": 0.06195502524093621,
+      "grad_norm": 3.475868431178915,
+      "learning_rate": 1.238532110091743e-05,
+      "loss": 1.1768,
       "step": 270
     },
     {
-      "epoch": 0.12620468104635155,
-      "grad_norm": 2.5178752868507144,
-      "learning_rate": 1.995833585557674e-05,
-      "loss": 1.0692,
+      "epoch": 0.06310234052317577,
+      "grad_norm": 3.3610034239289153,
+      "learning_rate": 1.261467889908257e-05,
+      "loss": 1.1789,
       "step": 275
     },
     {
-      "epoch": 0.12849931161083067,
-      "grad_norm": 2.560582169592424,
-      "learning_rate": 1.9950712038061617e-05,
-      "loss": 1.0518,
+      "epoch": 0.06424965580541533,
+      "grad_norm": 3.6114556514756755,
+      "learning_rate": 1.2844036697247708e-05,
+      "loss": 1.1839,
       "step": 280
     },
     {
-      "epoch": 0.13079394217530976,
-      "grad_norm": 2.279796271646809,
-      "learning_rate": 1.994244975656328e-05,
-      "loss": 1.0627,
+      "epoch": 0.06539697108765488,
+      "grad_norm": 3.5383966049543765,
+      "learning_rate": 1.3073394495412845e-05,
+      "loss": 1.1732,
       "step": 285
     },
     {
-      "epoch": 0.13308857273978889,
-      "grad_norm": 2.428851471685207,
-      "learning_rate": 1.993354954121155e-05,
-      "loss": 1.0672,
+      "epoch": 0.06654428636989444,
+      "grad_norm": 3.3264744648376805,
+      "learning_rate": 1.3302752293577984e-05,
+      "loss": 1.195,
       "step": 290
     },
     {
-      "epoch": 0.135383203304268,
-      "grad_norm": 2.253932739513235,
-      "learning_rate": 1.9924011963067765e-05,
-      "loss": 1.0572,
+      "epoch": 0.067691601652134,
+      "grad_norm": 3.242818454268948,
+      "learning_rate": 1.353211009174312e-05,
+      "loss": 1.2111,
       "step": 295
     },
     {
-      "epoch": 0.13767783386874713,
-      "grad_norm": 2.2863725772380232,
-      "learning_rate": 1.9913837634088143e-05,
-      "loss": 1.0612,
+      "epoch": 0.06883891693437356,
+      "grad_norm": 3.460258415056196,
+      "learning_rate": 1.3761467889908258e-05,
+      "loss": 1.1993,
       "step": 300
     },
     {
-      "epoch": 0.13997246443322625,
-      "grad_norm": 2.3216601188965433,
-      "learning_rate": 1.9903027207084525e-05,
-      "loss": 1.0711,
+      "epoch": 0.06998623221661313,
+      "grad_norm": 3.695239807114153,
+      "learning_rate": 1.3990825688073395e-05,
+      "loss": 1.1888,
       "step": 305
     },
     {
-      "epoch": 0.14226709499770537,
-      "grad_norm": 2.4619963294994305,
-      "learning_rate": 1.9891581375682472e-05,
-      "loss": 1.091,
+      "epoch": 0.07113354749885269,
+      "grad_norm": 3.107063192268371,
+      "learning_rate": 1.4220183486238533e-05,
+      "loss": 1.1733,
       "step": 310
     },
     {
-      "epoch": 0.1445617255621845,
-      "grad_norm": 2.6730364673716216,
-      "learning_rate": 1.9879500874276788e-05,
-      "loss": 1.0739,
+      "epoch": 0.07228086278109225,
+      "grad_norm": 3.2028941831424036,
+      "learning_rate": 1.4449541284403672e-05,
+      "loss": 1.2139,
       "step": 315
     },
     {
-      "epoch": 0.14685635612666362,
-      "grad_norm": 2.240223628972246,
-      "learning_rate": 1.9866786477984357e-05,
-      "loss": 1.0675,
+      "epoch": 0.07342817806333181,
+      "grad_norm": 3.066486456903709,
+      "learning_rate": 1.4678899082568809e-05,
+      "loss": 1.2,
       "step": 320
     },
     {
-      "epoch": 0.14915098669114274,
-      "grad_norm": 2.3152538012542436,
-      "learning_rate": 1.985343900259446e-05,
-      "loss": 1.0723,
+      "epoch": 0.07457549334557137,
+      "grad_norm": 4.414463221911407,
+      "learning_rate": 1.4908256880733946e-05,
+      "loss": 1.2214,
       "step": 325
     },
     {
-      "epoch": 0.15144561725562183,
-      "grad_norm": 2.4361030726660364,
-      "learning_rate": 1.983945930451639e-05,
-      "loss": 1.0895,
+      "epoch": 0.07572280862781092,
+      "grad_norm": 4.982537045862733,
+      "learning_rate": 1.5137614678899085e-05,
+      "loss": 1.1999,
       "step": 330
     },
     {
-      "epoch": 0.15374024782010096,
-      "grad_norm": 2.1416955206141086,
-      "learning_rate": 1.982484828072452e-05,
-      "loss": 1.0606,
+      "epoch": 0.07687012391005048,
+      "grad_norm": 3.339016451684457,
+      "learning_rate": 1.536697247706422e-05,
+      "loss": 1.2257,
       "step": 335
     },
     {
-      "epoch": 0.15603487838458008,
-      "grad_norm": 2.1520402841348054,
-      "learning_rate": 1.9809606868700755e-05,
-      "loss": 1.0751,
+      "epoch": 0.07801743919229004,
+      "grad_norm": 3.3426471175954022,
+      "learning_rate": 1.559633027522936e-05,
+      "loss": 1.1864,
       "step": 340
     },
     {
-      "epoch": 0.1583295089490592,
-      "grad_norm": 2.198043038406304,
-      "learning_rate": 1.9793736046374375e-05,
-      "loss": 1.0642,
+      "epoch": 0.0791647544745296,
+      "grad_norm": 3.151247165643508,
+      "learning_rate": 1.5825688073394497e-05,
+      "loss": 1.2004,
       "step": 345
     },
     {
-      "epoch": 0.16062413951353832,
-      "grad_norm": 2.310801939427329,
-      "learning_rate": 1.977723683205928e-05,
-      "loss": 1.0922,
+      "epoch": 0.08031206975676916,
+      "grad_norm": 3.398589476409347,
+      "learning_rate": 1.6055045871559634e-05,
+      "loss": 1.2106,
       "step": 350
     },
     {
-      "epoch": 0.16291877007801744,
-      "grad_norm": 2.108829610277771,
-      "learning_rate": 1.9760110284388667e-05,
-      "loss": 1.0738,
+      "epoch": 0.08145938503900872,
+      "grad_norm": 3.4252971557238103,
+      "learning_rate": 1.628440366972477e-05,
+      "loss": 1.2156,
       "step": 355
     },
     {
-      "epoch": 0.16521340064249657,
-      "grad_norm": 2.2185687059059602,
-      "learning_rate": 1.9742357502247104e-05,
-      "loss": 1.0636,
+      "epoch": 0.08260670032124828,
+      "grad_norm": 3.331446919185167,
+      "learning_rate": 1.6513761467889912e-05,
+      "loss": 1.2221,
       "step": 360
     },
     {
-      "epoch": 0.1675080312069757,
-      "grad_norm": 2.3820797912728433,
-      "learning_rate": 1.9723979624700004e-05,
-      "loss": 1.0646,
+      "epoch": 0.08375401560348784,
+      "grad_norm": 3.2290984121569073,
+      "learning_rate": 1.674311926605505e-05,
+      "loss": 1.2287,
       "step": 365
     },
     {
-      "epoch": 0.16980266177145478,
-      "grad_norm": 2.39166890987706,
-      "learning_rate": 1.970497783092057e-05,
-      "loss": 1.0826,
+      "epoch": 0.08490133088572739,
+      "grad_norm": 3.260527699936813,
+      "learning_rate": 1.6972477064220186e-05,
+      "loss": 1.1927,
       "step": 370
     },
     {
-      "epoch": 0.1720972923359339,
-      "grad_norm": 2.067234800314795,
-      "learning_rate": 1.9685353340114104e-05,
-      "loss": 1.0675,
+      "epoch": 0.08604864616796695,
+      "grad_norm": 3.290227145809032,
+      "learning_rate": 1.7201834862385323e-05,
+      "loss": 1.2177,
       "step": 375
     },
     {
-      "epoch": 0.17439192290041303,
-      "grad_norm": 2.16248649556658,
-      "learning_rate": 1.9665107411439805e-05,
-      "loss": 1.0955,
+      "epoch": 0.08719596145020651,
+      "grad_norm": 3.3351224170083884,
+      "learning_rate": 1.743119266055046e-05,
+      "loss": 1.2332,
       "step": 380
     },
     {
-      "epoch": 0.17668655346489215,
-      "grad_norm": 2.290019980298114,
-      "learning_rate": 1.9644241343929966e-05,
-      "loss": 1.0591,
+      "epoch": 0.08834327673244607,
+      "grad_norm": 3.0987470046595944,
+      "learning_rate": 1.7660550458715597e-05,
+      "loss": 1.2218,
       "step": 385
     },
     {
-      "epoch": 0.17898118402937127,
-      "grad_norm": 2.2129439226232503,
-      "learning_rate": 1.9622756476406625e-05,
-      "loss": 1.0631,
+      "epoch": 0.08949059201468564,
+      "grad_norm": 3.4005430609570766,
+      "learning_rate": 1.7889908256880734e-05,
+      "loss": 1.2104,
       "step": 390
     },
     {
-      "epoch": 0.1812758145938504,
-      "grad_norm": 2.027387747335278,
-      "learning_rate": 1.9600654187395666e-05,
-      "loss": 1.0497,
+      "epoch": 0.0906379072969252,
+      "grad_norm": 3.2242800996442633,
+      "learning_rate": 1.811926605504587e-05,
+      "loss": 1.2444,
       "step": 395
     },
     {
-      "epoch": 0.18357044515832951,
-      "grad_norm": 2.4021963976761995,
-      "learning_rate": 1.9577935895038363e-05,
-      "loss": 1.0738,
-      "step": 400
+      "epoch": 0.09178522257916476,
+      "grad_norm": 3.465093348681784,
+      "learning_rate": 1.834862385321101e-05,
+      "loss": 1.2207,
+      "step": 400
     },
     {
-      "epoch": 0.18586507572280864,
-      "grad_norm": 2.066698083420903,
-      "learning_rate": 1.9554603057000397e-05,
-      "loss": 1.0708,
+      "epoch": 0.09293253786140432,
+      "grad_norm": 3.273893281103885,
+      "learning_rate": 1.8577981651376148e-05,
+      "loss": 1.2811,
       "step": 405
     },
     {
-      "epoch": 0.18815970628728776,
-      "grad_norm": 2.2419541719123823,
-      "learning_rate": 1.953065717037832e-05,
-      "loss": 1.058,
+      "epoch": 0.09407985314364388,
+      "grad_norm": 3.389197589070483,
+      "learning_rate": 1.8807339449541285e-05,
+      "loss": 1.2325,
       "step": 410
     },
     {
-      "epoch": 0.19045433685176685,
-      "grad_norm": 2.2301364115878823,
-      "learning_rate": 1.9506099771603515e-05,
-      "loss": 1.0654,
+      "epoch": 0.09522716842588343,
+      "grad_norm": 3.289591416306655,
+      "learning_rate": 1.9036697247706422e-05,
+      "loss": 1.2516,
       "step": 415
     },
     {
-      "epoch": 0.19274896741624598,
-      "grad_norm": 2.259635192704883,
-      "learning_rate": 1.9480932436343584e-05,
-      "loss": 1.0608,
+      "epoch": 0.09637448370812299,
+      "grad_norm": 3.0860795978402225,
+      "learning_rate": 1.9266055045871563e-05,
+      "loss": 1.2475,
       "step": 420
     },
     {
-      "epoch": 0.1950435979807251,
-      "grad_norm": 2.1694738965454743,
-      "learning_rate": 1.945515677940127e-05,
-      "loss": 1.0795,
+      "epoch": 0.09752179899036255,
+      "grad_norm": 3.2671393904118475,
+      "learning_rate": 1.94954128440367e-05,
+      "loss": 1.2681,
       "step": 425
     },
     {
-      "epoch": 0.19733822854520422,
-      "grad_norm": 24.332124680220836,
-      "learning_rate": 1.9428774454610845e-05,
-      "loss": 1.09,
+      "epoch": 0.09866911427260211,
+      "grad_norm": 3.2619543784914535,
+      "learning_rate": 1.9724770642201837e-05,
+      "loss": 1.2523,
       "step": 430
     },
     {
-      "epoch": 0.19963285910968334,
-      "grad_norm": 13.049567795778453,
-      "learning_rate": 1.9401787154731993e-05,
-      "loss": 1.6953,
+      "epoch": 0.09981642955484167,
+      "grad_norm": 3.302002070021214,
+      "learning_rate": 1.9954128440366974e-05,
+      "loss": 1.2393,
       "step": 435
     },
     {
-      "epoch": 0.20192748967416246,
-      "grad_norm": 7.723175796818905,
-      "learning_rate": 1.9374196611341212e-05,
-      "loss": 1.1141,
+      "epoch": 0.10096374483708123,
+      "grad_norm": 3.093855168671344,
+      "learning_rate": 1.9999948669655127e-05,
+      "loss": 1.2365,
       "step": 440
     },
     {
-      "epoch": 0.20422212023864159,
-      "grad_norm": 87.6420124445034,
-      "learning_rate": 1.934600459472067e-05,
-      "loss": 1.1117,
+      "epoch": 0.10211106011932079,
+      "grad_norm": 3.123194323251484,
+      "learning_rate": 1.9999740141032216e-05,
+      "loss": 1.2305,
       "step": 445
     },
     {
-      "epoch": 0.2065167508031207,
-      "grad_norm": 2.9869822864961053,
-      "learning_rate": 1.931721291374467e-05,
-      "loss": 1.0953,
+      "epoch": 0.10325837540156035,
+      "grad_norm": 3.0544621364543856,
+      "learning_rate": 1.999937120932709e-05,
+      "loss": 1.2906,
       "step": 450
     },
     {
-      "epoch": 0.20881138136759983,
-      "grad_norm": 2.4448207284406775,
-      "learning_rate": 1.9287823415763552e-05,
-      "loss": 1.0806,
+      "epoch": 0.10440569068379991,
+      "grad_norm": 3.0019401915038273,
+      "learning_rate": 1.9998841880457682e-05,
+      "loss": 1.2699,
       "step": 455
     },
     {
-      "epoch": 0.21110601193207892,
-      "grad_norm": 2.566404692048504,
-      "learning_rate": 1.9257837986485187e-05,
-      "loss": 1.0969,
+      "epoch": 0.10555300596603946,
+      "grad_norm": 3.2375777216613972,
+      "learning_rate": 1.9998152162914807e-05,
+      "loss": 1.2553,
       "step": 460
     },
     {
-      "epoch": 0.21340064249655805,
-      "grad_norm": 4.385856228808105,
-      "learning_rate": 1.922725854985396e-05,
-      "loss": 1.055,
+      "epoch": 0.10670032124827902,
+      "grad_norm": 3.2092571589249737,
+      "learning_rate": 1.9997302067762044e-05,
+      "loss": 1.2313,
       "step": 465
     },
     {
-      "epoch": 0.21569527306103717,
-      "grad_norm": 2.3775711238334507,
-      "learning_rate": 1.9196087067927348e-05,
-      "loss": 1.0698,
+      "epoch": 0.10784763653051858,
+      "grad_norm": 3.017032153972236,
+      "learning_rate": 1.9996291608635527e-05,
+      "loss": 1.2775,
       "step": 470
     },
     {
-      "epoch": 0.2179899036255163,
-      "grad_norm": 2.2620590555710915,
-      "learning_rate": 1.916432554075002e-05,
-      "loss": 1.0825,
+      "epoch": 0.10899495181275815,
+      "grad_norm": 3.324102394140963,
+      "learning_rate": 1.999512080174375e-05,
+      "loss": 1.2837,
       "step": 475
     },
     {
-      "epoch": 0.2202845341899954,
-      "grad_norm": 2.7304207516254997,
-      "learning_rate": 1.913197600622549e-05,
-      "loss": 1.035,
+      "epoch": 0.1101422670949977,
+      "grad_norm": 3.1004727688280744,
+      "learning_rate": 1.9993789665867316e-05,
+      "loss": 1.2867,
       "step": 480
     },
     {
-      "epoch": 0.22257916475447453,
-      "grad_norm": 2.3652995101192382,
-      "learning_rate": 1.9099040539985395e-05,
-      "loss": 1.0598,
+      "epoch": 0.11128958237723727,
+      "grad_norm": 3.011852793277257,
+      "learning_rate": 1.9992298222358603e-05,
+      "loss": 1.258,
       "step": 485
     },
     {
-      "epoch": 0.22487379531895366,
-      "grad_norm": 2.3400675787952627,
-      "learning_rate": 1.90655212552563e-05,
-      "loss": 1.0507,
+      "epoch": 0.11243689765947683,
+      "grad_norm": 2.905138461740028,
+      "learning_rate": 1.9990646495141445e-05,
+      "loss": 1.2602,
       "step": 490
     },
     {
-      "epoch": 0.22716842588343278,
-      "grad_norm": 2.1219643393114493,
-      "learning_rate": 1.9031420302724093e-05,
-      "loss": 1.0549,
+      "epoch": 0.11358421294171639,
+      "grad_norm": 2.9824902308051984,
+      "learning_rate": 1.9988834510710747e-05,
+      "loss": 1.2446,
       "step": 495
     },
     {
-      "epoch": 0.22946305644791187,
-      "grad_norm": 2.0597546067330454,
-      "learning_rate": 1.8996739870396027e-05,
-      "loss": 1.0545,
+      "epoch": 0.11473152822395594,
+      "grad_norm": 3.122225637940991,
+      "learning_rate": 1.998686229813205e-05,
+      "loss": 1.2704,
       "step": 500
     },
     {
-      "epoch": 0.231757687012391,
-      "grad_norm": 2.0948012744577467,
-      "learning_rate": 1.896148218346028e-05,
-      "loss": 1.052,
+      "epoch": 0.1158788435061955,
+      "grad_norm": 3.116874664916445,
+      "learning_rate": 1.9984729889041077e-05,
+      "loss": 1.2559,
       "step": 505
     },
     {
-      "epoch": 0.23405231757687012,
-      "grad_norm": 2.1526817626887227,
-      "learning_rate": 1.8925649504143244e-05,
-      "loss": 1.0462,
+      "epoch": 0.11702615878843506,
+      "grad_norm": 3.280595317938923,
+      "learning_rate": 1.9982437317643218e-05,
+      "loss": 1.2762,
       "step": 510
     },
     {
-      "epoch": 0.23634694814134924,
-      "grad_norm": 2.038989673045541,
-      "learning_rate": 1.888924413156432e-05,
-      "loss": 1.0651,
+      "epoch": 0.11817347407067462,
+      "grad_norm": 3.013891926535922,
+      "learning_rate": 1.9979984620712972e-05,
+      "loss": 1.2655,
       "step": 515
     },
     {
-      "epoch": 0.23864157870582836,
-      "grad_norm": 2.1006533572356183,
-      "learning_rate": 1.885226840158843e-05,
-      "loss": 1.0644,
+      "epoch": 0.11932078935291418,
+      "grad_norm": 3.230616995695979,
+      "learning_rate": 1.9977371837593382e-05,
+      "loss": 1.2794,
       "step": 520
     },
     {
-      "epoch": 0.24093620927030748,
-      "grad_norm": 2.3332747317371596,
-      "learning_rate": 1.8814724686676133e-05,
-      "loss": 1.0722,
+      "epoch": 0.12046810463515374,
+      "grad_norm": 2.985385004989635,
+      "learning_rate": 1.9974599010195384e-05,
+      "loss": 1.2688,
       "step": 525
     },
     {
-      "epoch": 0.2432308398347866,
-      "grad_norm": 1.9851117275060848,
-      "learning_rate": 1.8776615395731398e-05,
-      "loss": 1.0596,
+      "epoch": 0.1216154199173933,
+      "grad_norm": 2.9459280723846955,
+      "learning_rate": 1.997166618299714e-05,
+      "loss": 1.2565,
       "step": 530
     },
     {
-      "epoch": 0.24552547039926573,
-      "grad_norm": 2.030081193927863,
-      "learning_rate": 1.8737942973947062e-05,
-      "loss": 1.0759,
+      "epoch": 0.12276273519963286,
+      "grad_norm": 3.327101250478011,
+      "learning_rate": 1.9968573403043325e-05,
+      "loss": 1.2386,
       "step": 535
     },
     {
-      "epoch": 0.24782010096374485,
-      "grad_norm": 2.2516155399349866,
-      "learning_rate": 1.8698709902647903e-05,
-      "loss": 1.0453,
+      "epoch": 0.12391005048187242,
+      "grad_norm": 3.10050951594509,
+      "learning_rate": 1.9965320719944366e-05,
+      "loss": 1.2785,
       "step": 540
     },
     {
-      "epoch": 0.25011473152822394,
-      "grad_norm": 2.5068515840745067,
-      "learning_rate": 1.865891869913147e-05,
-      "loss": 1.0908,
+      "epoch": 0.12505736576411197,
+      "grad_norm": 3.2120944118445367,
+      "learning_rate": 1.9961908185875662e-05,
+      "loss": 1.2578,
       "step": 545
     },
     {
-      "epoch": 0.2524093620927031,
-      "grad_norm": 2.4364345139539307,
-      "learning_rate": 1.8618571916506548e-05,
-      "loss": 1.0747,
+      "epoch": 0.12620468104635155,
+      "grad_norm": 2.9884869949235204,
+      "learning_rate": 1.995833585557674e-05,
+      "loss": 1.2694,
       "step": 550
     },
     {
-      "epoch": 0.2547039926571822,
-      "grad_norm": 2.174192040006664,
-      "learning_rate": 1.8577672143529337e-05,
-      "loss": 1.0695,
+      "epoch": 0.1273519963285911,
+      "grad_norm": 2.937215349500601,
+      "learning_rate": 1.9954603786350353e-05,
+      "loss": 1.2475,
       "step": 555
     },
     {
-      "epoch": 0.25699862322166134,
-      "grad_norm": 2.1082629086312306,
-      "learning_rate": 1.853622200443737e-05,
-      "loss": 1.0678,
+      "epoch": 0.12849931161083067,
+      "grad_norm": 4.433606116061364,
+      "learning_rate": 1.9950712038061617e-05,
+      "loss": 1.2883,
       "step": 560
     },
     {
-      "epoch": 0.25929325378614043,
-      "grad_norm": 2.026795919074908,
-      "learning_rate": 1.849422415878112e-05,
-      "loss": 1.0615,
+      "epoch": 0.12964662689307022,
+      "grad_norm": 3.081051255447705,
+      "learning_rate": 1.994666067313698e-05,
+      "loss": 1.2919,
       "step": 565
     },
     {
-      "epoch": 0.2615878843506195,
-      "grad_norm": 4.391522020875447,
-      "learning_rate": 1.8451681301253363e-05,
-      "loss": 1.0748,
+      "epoch": 0.13079394217530976,
+      "grad_norm": 3.246318877481652,
+      "learning_rate": 1.994244975656328e-05,
+      "loss": 1.2913,
       "step": 570
     },
     {
-      "epoch": 0.2638825149150987,
-      "grad_norm": 2.2545875042843218,
-      "learning_rate": 1.840859616151627e-05,
-      "loss": 1.0841,
+      "epoch": 0.13194125745754934,
+      "grad_norm": 2.860708074989303,
+      "learning_rate": 1.9938079355886674e-05,
+      "loss": 1.2504,
       "step": 575
     },
     {
-      "epoch": 0.26617714547957777,
-      "grad_norm": 2.4383336966875806,
-      "learning_rate": 1.8364971504026273e-05,
-      "loss": 1.0683,
+      "epoch": 0.13308857273978889,
+      "grad_norm": 3.2926223671952157,
+      "learning_rate": 1.993354954121155e-05,
+      "loss": 1.2938,
       "step": 580
     },
     {
-      "epoch": 0.2684717760440569,
-      "grad_norm": 2.032234747949604,
-      "learning_rate": 1.8320810127856706e-05,
-      "loss": 1.0503,
+      "epoch": 0.13423588802202846,
+      "grad_norm": 2.8372407984509076,
+      "learning_rate": 1.992886038519943e-05,
+      "loss": 1.266,
       "step": 585
     },
     {
-      "epoch": 0.270766406608536,
-      "grad_norm": 2.135201384180716,
-      "learning_rate": 1.827611486651817e-05,
-      "loss": 1.0702,
+      "epoch": 0.135383203304268,
+      "grad_norm": 3.0930378514658816,
+      "learning_rate": 1.9924011963067765e-05,
+      "loss": 1.2736,
       "step": 590
     },
     {
-      "epoch": 0.27306103717301516,
-      "grad_norm": 3.134318242625564,
-      "learning_rate": 1.8230888587776758e-05,
-      "loss": 1.0808,
+      "epoch": 0.13653051858650758,
+      "grad_norm": 3.046343540888883,
+      "learning_rate": 1.9919004352588768e-05,
+      "loss": 1.2603,
       "step": 595
     },
     {
-      "epoch": 0.27535566773749426,
-      "grad_norm": 2.098719263576858,
-      "learning_rate": 1.8185134193470043e-05,
-      "loss": 1.0639,
+      "epoch": 0.13767783386874713,
+      "grad_norm": 3.4947134554407038,
+      "learning_rate": 1.9913837634088143e-05,
+      "loss": 1.2806,
       "step": 600
     },
     {
-      "epoch": 0.2776502983019734,
-      "grad_norm": 2.167620655991022,
-      "learning_rate": 1.8138854619320893e-05,
-      "loss": 1.0679,
+      "epoch": 0.1388251491509867,
+      "grad_norm": 2.9164567885161863,
+      "learning_rate": 1.99085118904438e-05,
+      "loss": 1.2727,
       "step": 605
     },
     {
-      "epoch": 0.2799449288664525,
-      "grad_norm": 2.110572139729922,
-      "learning_rate": 1.8092052834749094e-05,
-      "loss": 1.0481,
+      "epoch": 0.13997246443322625,
+      "grad_norm": 3.2929601550445455,
+      "learning_rate": 1.9903027207084525e-05,
+      "loss": 1.283,
       "step": 610
     },
     {
-      "epoch": 0.2822395594309316,
-      "grad_norm": 2.111753085065974,
-      "learning_rate": 1.804473184268084e-05,
-      "loss": 1.0929,
+      "epoch": 0.1411197797154658,
+      "grad_norm": 3.1329363090645104,
+      "learning_rate": 1.989738367198862e-05,
+      "loss": 1.2742,
       "step": 615
     },
     {
-      "epoch": 0.28453418999541075,
-      "grad_norm": 2.047982339354027,
-      "learning_rate": 1.799689467935604e-05,
-      "loss": 1.0675,
+      "epoch": 0.14226709499770537,
+      "grad_norm": 3.1482798021615905,
+      "learning_rate": 1.9891581375682472e-05,
+      "loss": 1.271,
       "step": 620
     },
     {
-      "epoch": 0.28682882055988984,
-      "grad_norm": 1.9901275601162307,
-      "learning_rate": 1.7948544414133534e-05,
-      "loss": 1.0671,
+      "epoch": 0.14341441027994492,
+      "grad_norm": 2.8188055853541694,
+      "learning_rate": 1.9885620411239134e-05,
+      "loss": 1.2436,
       "step": 625
     },
     {
-      "epoch": 0.289123451124369,
-      "grad_norm": 2.355283701088896,
-      "learning_rate": 1.7899684149294118e-05,
-      "loss": 1.0831,
+      "epoch": 0.1445617255621845,
+      "grad_norm": 2.994182190028967,
+      "learning_rate": 1.9879500874276788e-05,
+      "loss": 1.256,
       "step": 630
     },
     {
-      "epoch": 0.2914180816888481,
-      "grad_norm": 2.034042330707897,
-      "learning_rate": 1.7850317019841514e-05,
-      "loss": 1.0551,
+      "epoch": 0.14570904084442404,
+      "grad_norm": 2.8840122615715407,
+      "learning_rate": 1.9873222862957243e-05,
+      "loss": 1.2759,
       "step": 635
     },
     {
-      "epoch": 0.29371271225332723,
-      "grad_norm": 2.0459885339908337,
-      "learning_rate": 1.7800446193301225e-05,
-      "loss": 1.0656,
+      "epoch": 0.14685635612666362,
+      "grad_norm": 3.216834319956669,
+      "learning_rate": 1.9866786477984357e-05,
+      "loss": 1.2297,
       "step": 640
     },
     {
-      "epoch": 0.29600734281780633,
-      "grad_norm": 1.9809660700340832,
-      "learning_rate": 1.7750074869517285e-05,
-      "loss": 1.0547,
+      "epoch": 0.14800367140890316,
+      "grad_norm": 2.9383340312733575,
+      "learning_rate": 1.9860191822602415e-05,
+      "loss": 1.25,
       "step": 645
     },
     {
-      "epoch": 0.2983019733822855,
-      "grad_norm": 2.1222169648334037,
-      "learning_rate": 1.7699206280446955e-05,
-      "loss": 1.0856,
+      "epoch": 0.14915098669114274,
+      "grad_norm": 2.897291855134717,
+      "learning_rate": 1.985343900259446e-05,
+      "loss": 1.2537,
       "step": 650
     },
     {
-      "epoch": 0.3005966039467646,
-      "grad_norm": 2.033320981208692,
-      "learning_rate": 1.7647843689953352e-05,
-      "loss": 1.0496,
+      "epoch": 0.1502983019733823,
+      "grad_norm": 2.975239778063116,
+      "learning_rate": 1.9846528126280632e-05,
+      "loss": 1.2621,
       "step": 655
     },
     {
-      "epoch": 0.30289123451124367,
-      "grad_norm": 1.9724823263811897,
-      "learning_rate": 1.759599039359603e-05,
-      "loss": 1.058,
+      "epoch": 0.15144561725562183,
+      "grad_norm": 2.978965429770825,
+      "learning_rate": 1.983945930451639e-05,
+      "loss": 1.2632,
       "step": 660
     },
     {
-      "epoch": 0.3051858650757228,
-      "grad_norm": 2.081504542418749,
-      "learning_rate": 1.754364971841952e-05,
-      "loss": 1.0652,
+      "epoch": 0.1525929325378614,
+      "grad_norm": 2.834209330189236,
+      "learning_rate": 1.9832232650690765e-05,
+      "loss": 1.2555,
       "step": 665
     },
     {
-      "epoch": 0.3074804956402019,
-      "grad_norm": 1.9226506132795518,
-      "learning_rate": 1.749082502273988e-05,
-      "loss": 1.0498,
+      "epoch": 0.15374024782010096,
+      "grad_norm": 2.8560141240052497,
+      "learning_rate": 1.982484828072452e-05,
+      "loss": 1.2592,
       "step": 670
     },
     {
-      "epoch": 0.30977512620468106,
-      "grad_norm": 2.093061922783108,
-      "learning_rate": 1.7437519695929194e-05,
-      "loss": 1.0469,
+      "epoch": 0.15488756310234053,
+      "grad_norm": 2.8618117387415456,
+      "learning_rate": 1.981730631306831e-05,
+      "loss": 1.2417,
       "step": 675
     },
     {
-      "epoch": 0.31206975676916016,
-      "grad_norm": 1.9436746042499786,
-      "learning_rate": 1.738373715819811e-05,
-      "loss": 1.0935,
+      "epoch": 0.15603487838458008,
+      "grad_norm": 2.910525869963812,
+      "learning_rate": 1.9809606868700755e-05,
+      "loss": 1.2771,
       "step": 680
     },
     {
-      "epoch": 0.3143643873336393,
-      "grad_norm": 2.141922458996327,
-      "learning_rate": 1.7329480860376392e-05,
-      "loss": 1.0708,
+      "epoch": 0.15718219366681965,
+      "grad_norm": 2.6548556540071266,
+      "learning_rate": 1.9801750071126536e-05,
+      "loss": 1.2352,
       "step": 685
     },
     {
-      "epoch": 0.3166590178981184,
-      "grad_norm": 1.9342237159056526,
-      "learning_rate": 1.7274754283691507e-05,
-      "loss": 1.0576,
+      "epoch": 0.1583295089490592,
+      "grad_norm": 2.867755148956131,
+      "learning_rate": 1.9793736046374375e-05,
+      "loss": 1.2651,
       "step": 690
     },
     {
-      "epoch": 0.3189536484625975,
-      "grad_norm": 2.3671090315479724,
-      "learning_rate": 1.7219560939545246e-05,
-      "loss": 1.0314,
+      "epoch": 0.15947682423129875,
+      "grad_norm": 2.956110626708722,
+      "learning_rate": 1.9785564922995042e-05,
+      "loss": 1.261,
       "step": 695
     },
     {
-      "epoch": 0.32124827902707664,
-      "grad_norm": 2.804036651572821,
-      "learning_rate": 1.7163904369288443e-05,
-      "loss": 1.037,
+      "epoch": 0.16062413951353832,
+      "grad_norm": 3.295040302897522,
+      "learning_rate": 1.977723683205928e-05,
+      "loss": 1.274,
       "step": 700
     },
     {
-      "epoch": 0.32354290959155574,
-      "grad_norm": 2.130875477845555,
-      "learning_rate": 1.710778814399374e-05,
-      "loss": 1.0676,
+      "epoch": 0.16177145479577787,
+      "grad_norm": 2.936931272401806,
+      "learning_rate": 1.9768751907155707e-05,
+      "loss": 1.2776,
       "step": 705
     },
     {
-      "epoch": 0.3258375401560349,
-      "grad_norm": 2.2917073480609145,
-      "learning_rate": 1.705121586422647e-05,
-      "loss": 1.0603,
+      "epoch": 0.16291877007801744,
+      "grad_norm": 2.9114545457536845,
+      "learning_rate": 1.9760110284388667e-05,
+      "loss": 1.3049,
       "step": 710
     },
     {
-      "epoch": 0.328132170720514,
-      "grad_norm": 1.9344196026563947,
-      "learning_rate": 1.699419115981361e-05,
-      "loss": 1.0563,
+      "epoch": 0.164066085360257,
+      "grad_norm": 2.646123204687912,
+      "learning_rate": 1.9751312102376062e-05,
+      "loss": 1.2781,
       "step": 715
     },
     {
-      "epoch": 0.33042680128499313,
-      "grad_norm": 2.588695773647949,
-      "learning_rate": 1.69367176896109e-05,
-      "loss": 1.0383,
+      "epoch": 0.16521340064249657,
+      "grad_norm": 2.774463476806892,
+      "learning_rate": 1.9742357502247104e-05,
+      "loss": 1.2376,
       "step": 720
     },
     {
-      "epoch": 0.3327214318494722,
-      "grad_norm": 1.8338278514413,
-      "learning_rate": 1.6878799141268107e-05,
-      "loss": 1.0367,
+      "epoch": 0.1663607159247361,
+      "grad_norm": 3.092003565928542,
+      "learning_rate": 1.9733246627640072e-05,
+      "loss": 1.2956,
       "step": 725
     },
     {
-      "epoch": 0.3350160624139514,
-      "grad_norm": 1.9232017037694178,
-      "learning_rate": 1.682043923099234e-05,
-      "loss": 1.0609,
+      "epoch": 0.1675080312069757,
+      "grad_norm": 2.888582015827288,
+      "learning_rate": 1.9723979624700004e-05,
+      "loss": 1.2603,
       "step": 730
     },
     {
-      "epoch": 0.33731069297843047,
-      "grad_norm": 2.0707722390620638,
-      "learning_rate": 1.6761641703309702e-05,
-      "loss": 1.0544,
+      "epoch": 0.16865534648921524,
+      "grad_norm": 2.8242754861502744,
+      "learning_rate": 1.9714556642076347e-05,
+      "loss": 1.2572,
       "step": 735
     },
     {
-      "epoch": 0.33960532354290957,
-      "grad_norm": 1.863894286464851,
-      "learning_rate": 1.6702410330824962e-05,
-      "loss": 1.0459,
+      "epoch": 0.16980266177145478,
+      "grad_norm": 2.8564441721897045,
+      "learning_rate": 1.970497783092057e-05,
+      "loss": 1.275,
       "step": 740
     },
     {
-      "epoch": 0.3418999541073887,
-      "grad_norm": 2.295179291530668,
-      "learning_rate": 1.6642748913979515e-05,
-      "loss": 1.0664,
+      "epoch": 0.17094997705369436,
+      "grad_norm": 4.737708384861383,
+      "learning_rate": 1.969524334488375e-05,
+      "loss": 1.2806,
       "step": 745
     },
     {
-      "epoch": 0.3441945846718678,
-      "grad_norm": 1.9748901739443767,
-      "learning_rate": 1.6582661280807553e-05,
-      "loss": 1.0824,
+      "epoch": 0.1720972923359339,
+      "grad_norm": 2.7729048751649685,
+      "learning_rate": 1.9685353340114104e-05,
+      "loss": 1.2555,
       "step": 750
     },
     {
-      "epoch": 0.34648921523634696,
-      "grad_norm": 2.0941843827334976,
-      "learning_rate": 1.652215128669042e-05,
-      "loss": 1.0663,
+      "epoch": 0.17324460761817348,
+      "grad_norm": 3.7322733945617985,
+      "learning_rate": 1.9675307975254478e-05,
+      "loss": 1.2735,
       "step": 755
     },
     {
-      "epoch": 0.34878384580082605,
-      "grad_norm": 2.026330695493712,
-      "learning_rate": 1.646122281410927e-05,
-      "loss": 1.0582,
+      "epoch": 0.17439192290041303,
+      "grad_norm": 6.994325743658938,
+      "learning_rate": 1.9665107411439805e-05,
+      "loss": 1.2726,
       "step": 760
     },
     {
-      "epoch": 0.3510784763653052,
-      "grad_norm": 1.814204512009467,
-      "learning_rate": 1.6399879772395915e-05,
-      "loss": 1.0664,
+      "epoch": 0.1755392381826526,
+      "grad_norm": 3.0013238213848012,
+      "learning_rate": 1.965475181229453e-05,
+      "loss": 1.2633,
       "step": 765
     },
     {
-      "epoch": 0.3533731069297843,
-      "grad_norm": 2.0517724058062794,
-      "learning_rate": 1.633812609748206e-05,
-      "loss": 1.0479,
+      "epoch": 0.17668655346489215,
+      "grad_norm": 2.743017825024535,
+      "learning_rate": 1.9644241343929966e-05,
+      "loss": 1.2657,
       "step": 770
     },
     {
-      "epoch": 0.35566773749426345,
-      "grad_norm": 1.917088322296075,
-      "learning_rate": 1.6275965751646682e-05,
-      "loss": 1.049,
+      "epoch": 0.17783386874713172,
+      "grad_norm": 2.7294372056243126,
+      "learning_rate": 1.963357617494165e-05,
+      "loss": 1.254,
       "step": 775
     },
     {
-      "epoch": 0.35796236805874254,
-      "grad_norm": 1.8734674623090466,
-      "learning_rate": 1.6213402723261852e-05,
-      "loss": 1.0544,
+      "epoch": 0.17898118402937127,
+      "grad_norm": 2.6639405368239806,
+      "learning_rate": 1.9622756476406625e-05,
+      "loss": 1.2555,
       "step": 780
     },
     {
-      "epoch": 0.36025699862322164,
-      "grad_norm": 2.1022753225906268,
+      "epoch": 0.18012849931161082,
+      "grad_norm": 2.7727550754305086,
+      "learning_rate": 1.9611782421880702e-05,
+      "loss": 1.2726,
+      "step": 785
+    },
+    {
+      "epoch": 0.1812758145938504,
+      "grad_norm": 2.8402056215181304,
+      "learning_rate": 1.9600654187395666e-05,
+      "loss": 1.2892,
+      "step": 790
+    },
+    {
+      "epoch": 0.18242312987608994,
+      "grad_norm": 2.783195044162214,
+      "learning_rate": 1.958937195145647e-05,
+      "loss": 1.2551,
+      "step": 795
+    },
+    {
+      "epoch": 0.18357044515832951,
+      "grad_norm": 2.8916825221026303,
+      "learning_rate": 1.9577935895038363e-05,
+      "loss": 1.2669,
+      "step": 800
+    },
+    {
+      "epoch": 0.18471776044056906,
+      "grad_norm": 2.707687158217741,
+      "learning_rate": 1.9566346201583974e-05,
+      "loss": 1.2492,
+      "step": 805
+    },
+    {
+      "epoch": 0.18586507572280864,
+      "grad_norm": 3.158708488021887,
+      "learning_rate": 1.9554603057000397e-05,
+      "loss": 1.2637,
+      "step": 810
+    },
+    {
+      "epoch": 0.18701239100504818,
+      "grad_norm": 2.9465795983656182,
+      "learning_rate": 1.954270664965618e-05,
+      "loss": 1.2449,
+      "step": 815
+    },
+    {
+      "epoch": 0.18815970628728776,
+      "grad_norm": 2.7211490382547883,
+      "learning_rate": 1.953065717037832e-05,
+      "loss": 1.2544,
+      "step": 820
+    },
+    {
+      "epoch": 0.1893070215695273,
+      "grad_norm": 2.8949023742596314,
+      "learning_rate": 1.951845481244921e-05,
+      "loss": 1.2273,
+      "step": 825
+    },
+    {
+      "epoch": 0.19045433685176685,
+      "grad_norm": 2.765474296144506,
+      "learning_rate": 1.9506099771603515e-05,
+      "loss": 1.2493,
+      "step": 830
+    },
+    {
+      "epoch": 0.19160165213400643,
+      "grad_norm": 2.900346678087231,
+      "learning_rate": 1.9493592246025047e-05,
+      "loss": 1.2956,
+      "step": 835
+    },
+    {
+      "epoch": 0.19274896741624598,
+      "grad_norm": 2.845481992620935,
+      "learning_rate": 1.9480932436343584e-05,
+      "loss": 1.2694,
+      "step": 840
+    },
+    {
+      "epoch": 0.19389628269848555,
+      "grad_norm": 2.8255091786274047,
+      "learning_rate": 1.9468120545631647e-05,
+      "loss": 1.2453,
+      "step": 845
+    },
+    {
+      "epoch": 0.1950435979807251,
+      "grad_norm": 2.854290942764917,
+      "learning_rate": 1.945515677940127e-05,
+      "loss": 1.2611,
+      "step": 850
+    },
+    {
+      "epoch": 0.19619091326296467,
+      "grad_norm": 2.7613113548764083,
+      "learning_rate": 1.944204134560064e-05,
+      "loss": 1.2483,
+      "step": 855
+    },
+    {
+      "epoch": 0.19733822854520422,
+      "grad_norm": 2.9238914008425145,
+      "learning_rate": 1.9428774454610845e-05,
+      "loss": 1.2613,
+      "step": 860
+    },
+    {
+      "epoch": 0.1984855438274438,
+      "grad_norm": 2.9258869381539463,
+      "learning_rate": 1.941535631924242e-05,
+      "loss": 1.2924,
+      "step": 865
+    },
+    {
+      "epoch": 0.19963285910968334,
+      "grad_norm": 2.845716729573166,
+      "learning_rate": 1.9401787154731993e-05,
+      "loss": 1.2757,
+      "step": 870
+    },
+    {
+      "epoch": 0.2007801743919229,
+      "grad_norm": 2.7827244136621814,
+      "learning_rate": 1.9388067178738807e-05,
+      "loss": 1.2926,
+      "step": 875
+    },
+    {
+      "epoch": 0.20192748967416246,
+      "grad_norm": 2.8654232792004755,
+      "learning_rate": 1.9374196611341212e-05,
+      "loss": 1.2831,
+      "step": 880
+    },
+    {
+      "epoch": 0.203074804956402,
+      "grad_norm": 2.768678720904829,
+      "learning_rate": 1.936017567503317e-05,
+      "loss": 1.2646,
+      "step": 885
+    },
+    {
+      "epoch": 0.20422212023864159,
+      "grad_norm": 2.7336541760912083,
+      "learning_rate": 1.934600459472067e-05,
+      "loss": 1.2654,
+      "step": 890
+    },
+    {
+      "epoch": 0.20536943552088113,
+      "grad_norm": 2.877666820443206,
+      "learning_rate": 1.933168359771811e-05,
+      "loss": 1.2617,
+      "step": 895
+    },
+    {
+      "epoch": 0.2065167508031207,
+      "grad_norm": 2.77452485009731,
+      "learning_rate": 1.931721291374467e-05,
+      "loss": 1.2537,
+      "step": 900
+    },
+    {
+      "epoch": 0.20766406608536025,
+      "grad_norm": 2.759805605741722,
+      "learning_rate": 1.9302592774920606e-05,
+      "loss": 1.2469,
+      "step": 905
+    },
+    {
+      "epoch": 0.20881138136759983,
+      "grad_norm": 2.633186753874087,
+      "learning_rate": 1.9287823415763552e-05,
+      "loss": 1.2598,
+      "step": 910
+    },
+    {
+      "epoch": 0.20995869664983938,
+      "grad_norm": 2.762895084636893,
+      "learning_rate": 1.9272905073184734e-05,
+      "loss": 1.2776,
+      "step": 915
+    },
+    {
+      "epoch": 0.21110601193207892,
+      "grad_norm": 2.6741191476011523,
+      "learning_rate": 1.9257837986485187e-05,
+      "loss": 1.2636,
+      "step": 920
+    },
+    {
+      "epoch": 0.2122533272143185,
+      "grad_norm": 2.7328410543396293,
+      "learning_rate": 1.92426223973519e-05,
+      "loss": 1.2329,
+      "step": 925
+    },
+    {
+      "epoch": 0.21340064249655805,
+      "grad_norm": 2.758194666346516,
+      "learning_rate": 1.922725854985396e-05,
+      "loss": 1.2447,
+      "step": 930
+    },
+    {
+      "epoch": 0.21454795777879762,
+      "grad_norm": 2.9215747504377796,
+      "learning_rate": 1.921174669043862e-05,
+      "loss": 1.265,
+      "step": 935
+    },
+    {
+      "epoch": 0.21569527306103717,
+      "grad_norm": 2.6957052221176574,
+      "learning_rate": 1.9196087067927348e-05,
+      "loss": 1.273,
+      "step": 940
+    },
+    {
+      "epoch": 0.21684258834327674,
+      "grad_norm": 3.151322186534713,
+      "learning_rate": 1.918027993351185e-05,
+      "loss": 1.2494,
+      "step": 945
+    },
+    {
+      "epoch": 0.2179899036255163,
+      "grad_norm": 2.8468019061200005,
+      "learning_rate": 1.916432554075002e-05,
+      "loss": 1.2601,
+      "step": 950
+    },
+    {
+      "epoch": 0.21913721890775584,
+      "grad_norm": 2.8675388482251565,
+      "learning_rate": 1.9148224145561876e-05,
+      "loss": 1.2694,
+      "step": 955
+    },
+    {
+      "epoch": 0.2202845341899954,
+      "grad_norm": 2.742245998362719,
+      "learning_rate": 1.913197600622549e-05,
+      "loss": 1.208,
+      "step": 960
+    },
+    {
+      "epoch": 0.22143184947223496,
+      "grad_norm": 2.738604567476206,
+      "learning_rate": 1.9115581383372782e-05,
+      "loss": 1.2773,
+      "step": 965
+    },
+    {
+      "epoch": 0.22257916475447453,
+      "grad_norm": 2.8303874276680294,
+      "learning_rate": 1.9099040539985395e-05,
+      "loss": 1.2869,
+      "step": 970
+    },
+    {
+      "epoch": 0.22372648003671408,
+      "grad_norm": 2.6179532721860173,
+      "learning_rate": 1.9082353741390453e-05,
+      "loss": 1.2405,
+      "step": 975
+    },
+    {
+      "epoch": 0.22487379531895366,
+      "grad_norm": 2.6182090072869024,
+      "learning_rate": 1.90655212552563e-05,
+      "loss": 1.2459,
+      "step": 980
+    },
+    {
+      "epoch": 0.2260211106011932,
+      "grad_norm": 2.7666514211794278,
+      "learning_rate": 1.904854335158822e-05,
+      "loss": 1.2752,
+      "step": 985
+    },
+    {
+      "epoch": 0.22716842588343278,
+      "grad_norm": 2.818653838842991,
+      "learning_rate": 1.9031420302724093e-05,
+      "loss": 1.2815,
+      "step": 990
+    },
+    {
+      "epoch": 0.22831574116567233,
+      "grad_norm": 2.9114632131177367,
+      "learning_rate": 1.901415238333005e-05,
+      "loss": 1.2539,
+      "step": 995
+    },
+    {
+      "epoch": 0.22946305644791187,
+      "grad_norm": 2.733983237186035,
+      "learning_rate": 1.8996739870396027e-05,
+      "loss": 1.2919,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23061037173015145,
+      "grad_norm": 2.61730939299338,
+      "learning_rate": 1.897918304323136e-05,
+      "loss": 1.2365,
+      "step": 1005
+    },
+    {
+      "epoch": 0.231757687012391,
+      "grad_norm": 2.7775678613752803,
+      "learning_rate": 1.896148218346028e-05,
+      "loss": 1.2835,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23290500229463057,
+      "grad_norm": 2.6620434707223914,
+      "learning_rate": 1.8943637575017428e-05,
+      "loss": 1.2754,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23405231757687012,
+      "grad_norm": 2.826697714775707,
+      "learning_rate": 1.8925649504143244e-05,
+      "loss": 1.2547,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2351996328591097,
+      "grad_norm": 2.723804846994195,
+      "learning_rate": 1.890751825937944e-05,
+      "loss": 1.2622,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23634694814134924,
+      "grad_norm": 2.59612274862863,
+      "learning_rate": 1.888924413156432e-05,
+      "loss": 1.2766,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23749426342358881,
+      "grad_norm": 2.6494254565525774,
+      "learning_rate": 1.8870827413828148e-05,
+      "loss": 1.242,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23864157870582836,
+      "grad_norm": 2.656174111008909,
+      "learning_rate": 1.885226840158843e-05,
+      "loss": 1.23,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2397888939880679,
+      "grad_norm": 2.763662684655998,
+      "learning_rate": 1.8833567392545177e-05,
+      "loss": 1.2676,
+      "step": 1045
+    },
+    {
+      "epoch": 0.24093620927030748,
+      "grad_norm": 2.641174999509725,
+      "learning_rate": 1.8814724686676133e-05,
+      "loss": 1.2455,
+      "step": 1050
+    },
+    {
+      "epoch": 0.24208352455254703,
+      "grad_norm": 2.6616650480778152,
+      "learning_rate": 1.879574058623196e-05,
+      "loss": 1.2372,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2432308398347866,
+      "grad_norm": 2.573070310310057,
+      "learning_rate": 1.8776615395731398e-05,
+      "loss": 1.2301,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24437815511702615,
+      "grad_norm": 2.7789276102882505,
+      "learning_rate": 1.875734942195637e-05,
+      "loss": 1.256,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24552547039926573,
+      "grad_norm": 2.5198281488260093,
+      "learning_rate": 1.8737942973947062e-05,
+      "loss": 1.2509,
+      "step": 1070
+    },
+    {
+      "epoch": 0.24667278568150527,
+      "grad_norm": 2.6757554347799504,
+      "learning_rate": 1.8718396362996968e-05,
+      "loss": 1.2686,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24782010096374485,
+      "grad_norm": 2.7605588886847907,
+      "learning_rate": 1.8698709902647903e-05,
+      "loss": 1.2404,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2489674162459844,
+      "grad_norm": 2.6067001677094797,
+      "learning_rate": 1.8678883908684964e-05,
+      "loss": 1.243,
+      "step": 1085
+    },
+    {
+      "epoch": 0.25011473152822394,
+      "grad_norm": 2.5935855463654676,
+      "learning_rate": 1.865891869913147e-05,
+      "loss": 1.2629,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2512620468104635,
+      "grad_norm": 2.6846198334888047,
+      "learning_rate": 1.863881459424386e-05,
+      "loss": 1.2455,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2524093620927031,
+      "grad_norm": 2.770064972387501,
+      "learning_rate": 1.8618571916506548e-05,
+      "loss": 1.2702,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25355667737494264,
+      "grad_norm": 7.16414497208806,
+      "learning_rate": 1.8598190990626764e-05,
+      "loss": 1.2746,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2547039926571822,
+      "grad_norm": 2.8981550261204982,
+      "learning_rate": 1.8577672143529337e-05,
+      "loss": 1.2653,
+      "step": 1110
+    },
+    {
+      "epoch": 0.25585130793942173,
+      "grad_norm": 2.8392180307250805,
+      "learning_rate": 1.8557015704351453e-05,
+      "loss": 1.28,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25699862322166134,
+      "grad_norm": 2.8866536206176465,
+      "learning_rate": 1.853622200443737e-05,
+      "loss": 1.2475,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2581459385039009,
+      "grad_norm": 2.7434811297459287,
+      "learning_rate": 1.8515291377333114e-05,
+      "loss": 1.2525,
+      "step": 1125
+    },
+    {
+      "epoch": 0.25929325378614043,
+      "grad_norm": 2.516325346750563,
+      "learning_rate": 1.849422415878112e-05,
+      "loss": 1.2477,
+      "step": 1130
+    },
+    {
+      "epoch": 0.26044056906838,
+      "grad_norm": 2.7285846616073455,
+      "learning_rate": 1.8473020686714847e-05,
+      "loss": 1.293,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2615878843506195,
+      "grad_norm": 2.5584913324504517,
+      "learning_rate": 1.8451681301253363e-05,
+      "loss": 1.2314,
+      "step": 1140
+    },
+    {
+      "epoch": 0.26273519963285913,
+      "grad_norm": 2.833266497213628,
+      "learning_rate": 1.8430206344695875e-05,
+      "loss": 1.2366,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2638825149150987,
+      "grad_norm": 2.6820864779644547,
+      "learning_rate": 1.840859616151627e-05,
+      "loss": 1.2479,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2650298301973382,
+      "grad_norm": 2.6742387500940565,
+      "learning_rate": 1.8386851098357538e-05,
+      "loss": 1.2892,
+      "step": 1155
+    },
+    {
+      "epoch": 0.26617714547957777,
+      "grad_norm": 2.6037489307720936,
+      "learning_rate": 1.8364971504026273e-05,
+      "loss": 1.2493,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2673244607618174,
+      "grad_norm": 2.722204223715423,
+      "learning_rate": 1.834295772948703e-05,
+      "loss": 1.2827,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2684717760440569,
+      "grad_norm": 2.7859216758984555,
+      "learning_rate": 1.8320810127856706e-05,
+      "loss": 1.2338,
+      "step": 1170
+    },
+    {
+      "epoch": 0.26961909132629647,
+      "grad_norm": 2.6379505799749174,
+      "learning_rate": 1.8298529054398896e-05,
+      "loss": 1.2653,
+      "step": 1175
+    },
+    {
+      "epoch": 0.270766406608536,
+      "grad_norm": 2.70334833150815,
+      "learning_rate": 1.827611486651817e-05,
+      "loss": 1.2373,
+      "step": 1180
+    },
+    {
+      "epoch": 0.27191372189077556,
+      "grad_norm": 2.689144836189015,
+      "learning_rate": 1.8253567923754353e-05,
+      "loss": 1.2876,
+      "step": 1185
+    },
+    {
+      "epoch": 0.27306103717301516,
+      "grad_norm": 2.8506131594938497,
+      "learning_rate": 1.8230888587776758e-05,
+      "loss": 1.254,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2742083524552547,
+      "grad_norm": 2.7967969326711333,
+      "learning_rate": 1.8208077222378376e-05,
+      "loss": 1.2268,
+      "step": 1195
+    },
+    {
+      "epoch": 0.27535566773749426,
+      "grad_norm": 2.8211137090561986,
+      "learning_rate": 1.8185134193470043e-05,
+      "loss": 1.3033,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2765029830197338,
+      "grad_norm": 2.6789699271006118,
+      "learning_rate": 1.8162059869074586e-05,
+      "loss": 1.2396,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2776502983019734,
+      "grad_norm": 2.669915646408006,
+      "learning_rate": 1.8138854619320893e-05,
+      "loss": 1.2382,
+      "step": 1210
+    },
+    {
+      "epoch": 0.27879761358421296,
+      "grad_norm": 2.8691949960246697,
+      "learning_rate": 1.8115518816437997e-05,
+      "loss": 1.2627,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2799449288664525,
+      "grad_norm": 2.5198989709998605,
+      "learning_rate": 1.8092052834749094e-05,
+      "loss": 1.2421,
+      "step": 1220
+    },
+    {
+      "epoch": 0.28109224414869205,
+      "grad_norm": 2.623953735014251,
+      "learning_rate": 1.8068457050665547e-05,
+      "loss": 1.2411,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2822395594309316,
+      "grad_norm": 2.701213051257084,
+      "learning_rate": 1.804473184268084e-05,
+      "loss": 1.2664,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2833868747131712,
+      "grad_norm": 2.5467029866920954,
+      "learning_rate": 1.8020877591364508e-05,
+      "loss": 1.2607,
+      "step": 1235
+    },
+    {
+      "epoch": 0.28453418999541075,
+      "grad_norm": 2.5990206825217377,
+      "learning_rate": 1.799689467935604e-05,
+      "loss": 1.2528,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2856815052776503,
+      "grad_norm": 2.7964994357271413,
+      "learning_rate": 1.797278349135874e-05,
+      "loss": 1.2496,
+      "step": 1245
+    },
+    {
+      "epoch": 0.28682882055988984,
+      "grad_norm": 2.6044762278441858,
+      "learning_rate": 1.7948544414133534e-05,
+      "loss": 1.2508,
+      "step": 1250
+    },
+    {
+      "epoch": 0.28797613584212944,
+      "grad_norm": 2.72468981644295,
+      "learning_rate": 1.7924177836492802e-05,
+      "loss": 1.2324,
+      "step": 1255
+    },
+    {
+      "epoch": 0.289123451124369,
+      "grad_norm": 2.596318179164043,
+      "learning_rate": 1.7899684149294118e-05,
+      "loss": 1.2392,
+      "step": 1260
+    },
+    {
+      "epoch": 0.29027076640660854,
+      "grad_norm": 2.681066970647215,
+      "learning_rate": 1.7875063745433978e-05,
+      "loss": 1.2719,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2914180816888481,
+      "grad_norm": 2.7223877723737577,
+      "learning_rate": 1.7850317019841514e-05,
+      "loss": 1.2569,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29256539697108763,
+      "grad_norm": 2.4605258800436474,
+      "learning_rate": 1.7825444369472147e-05,
+      "loss": 1.2495,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29371271225332723,
+      "grad_norm": 2.645246095924411,
+      "learning_rate": 1.7800446193301225e-05,
+      "loss": 1.2494,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2948600275355668,
+      "grad_norm": 2.686411807869381,
+      "learning_rate": 1.7775322892317618e-05,
+      "loss": 1.2565,
+      "step": 1285
+    },
+    {
+      "epoch": 0.29600734281780633,
+      "grad_norm": 2.5616644208417036,
+      "learning_rate": 1.7750074869517285e-05,
+      "loss": 1.2373,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2971546581000459,
+      "grad_norm": 2.752699987412825,
+      "learning_rate": 1.7724702529896824e-05,
+      "loss": 1.2349,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2983019733822855,
+      "grad_norm": 2.6292531808701103,
+      "learning_rate": 1.7699206280446955e-05,
+      "loss": 1.2284,
+      "step": 1300
+    },
+    {
+      "epoch": 0.299449288664525,
+      "grad_norm": 2.4894873410954657,
+      "learning_rate": 1.767358653014601e-05,
+      "loss": 1.2206,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3005966039467646,
+      "grad_norm": 2.4189638669221507,
+      "learning_rate": 1.7647843689953352e-05,
+      "loss": 1.2321,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3017439192290041,
+      "grad_norm": 2.569424733644508,
+      "learning_rate": 1.762197817280281e-05,
+      "loss": 1.2487,
+      "step": 1315
+    },
+    {
+      "epoch": 0.30289123451124367,
+      "grad_norm": 2.581452393958971,
+      "learning_rate": 1.759599039359603e-05,
+      "loss": 1.2311,
+      "step": 1320
+    },
+    {
+      "epoch": 0.30403854979348327,
+      "grad_norm": 2.4458262952088536,
+      "learning_rate": 1.756988076919583e-05,
+      "loss": 1.2615,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3051858650757228,
+      "grad_norm": 2.511753933286628,
+      "learning_rate": 1.754364971841952e-05,
+      "loss": 1.2479,
+      "step": 1330
+    },
+    {
+      "epoch": 0.30633318035796236,
+      "grad_norm": 2.61085957686584,
+      "learning_rate": 1.7517297662032174e-05,
+      "loss": 1.2379,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3074804956402019,
+      "grad_norm": 2.6407993774020206,
+      "learning_rate": 1.749082502273988e-05,
+      "loss": 1.2579,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3086278109224415,
+      "grad_norm": 2.6120216096924977,
+      "learning_rate": 1.746423222518297e-05,
+      "loss": 1.2387,
+      "step": 1345
+    },
+    {
+      "epoch": 0.30977512620468106,
+      "grad_norm": 2.8982851648085504,
+      "learning_rate": 1.7437519695929194e-05,
+      "loss": 1.23,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3109224414869206,
+      "grad_norm": 2.5580235398027793,
+      "learning_rate": 1.741068786346689e-05,
+      "loss": 1.2557,
+      "step": 1355
+    },
+    {
+      "epoch": 0.31206975676916016,
+      "grad_norm": 2.565411769595978,
+      "learning_rate": 1.738373715819811e-05,
+      "loss": 1.2392,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3132170720513997,
+      "grad_norm": 2.617152710791485,
+      "learning_rate": 1.7356668012431705e-05,
+      "loss": 1.2352,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3143643873336393,
+      "grad_norm": 2.503182025195152,
+      "learning_rate": 1.7329480860376392e-05,
+      "loss": 1.2393,
+      "step": 1370
+    },
+    {
+      "epoch": 0.31551170261587885,
+      "grad_norm": 2.504235729550804,
+      "learning_rate": 1.7302176138133814e-05,
+      "loss": 1.2367,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3166590178981184,
+      "grad_norm": 2.589109156050675,
+      "learning_rate": 1.7274754283691507e-05,
+      "loss": 1.2437,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31780633318035795,
+      "grad_norm": 2.5984716155591223,
+      "learning_rate": 1.72472157369159e-05,
+      "loss": 1.2332,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3189536484625975,
+      "grad_norm": 2.5431122123105374,
+      "learning_rate": 1.7219560939545246e-05,
+      "loss": 1.2475,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3201009637448371,
+      "grad_norm": 2.5803698318259682,
+      "learning_rate": 1.719179033518255e-05,
+      "loss": 1.2456,
+      "step": 1395
+    },
+    {
+      "epoch": 0.32124827902707664,
+      "grad_norm": 2.654846032452953,
+      "learning_rate": 1.7163904369288443e-05,
+      "loss": 1.2485,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3223955943093162,
+      "grad_norm": 2.4536340664314955,
+      "learning_rate": 1.7135903489174034e-05,
+      "loss": 1.2536,
+      "step": 1405
+    },
+    {
+      "epoch": 0.32354290959155574,
+      "grad_norm": 2.5841724320770005,
+      "learning_rate": 1.710778814399374e-05,
+      "loss": 1.2392,
+      "step": 1410
+    },
+    {
+      "epoch": 0.32469022487379534,
+      "grad_norm": 2.527960263106006,
+      "learning_rate": 1.7079558784738092e-05,
+      "loss": 1.2336,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3258375401560349,
+      "grad_norm": 2.470359172698634,
+      "learning_rate": 1.705121586422647e-05,
+      "loss": 1.2138,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32698485543827444,
+      "grad_norm": 2.619939129309239,
+      "learning_rate": 1.702275983709987e-05,
+      "loss": 1.2346,
+      "step": 1425
+    },
+    {
+      "epoch": 0.328132170720514,
+      "grad_norm": 2.655269081012586,
+      "learning_rate": 1.699419115981361e-05,
+      "loss": 1.241,
+      "step": 1430
+    },
+    {
+      "epoch": 0.32927948600275353,
+      "grad_norm": 2.5162818627628796,
+      "learning_rate": 1.6965510290629973e-05,
+      "loss": 1.2222,
+      "step": 1435
+    },
+    {
+      "epoch": 0.33042680128499313,
+      "grad_norm": 2.5758937912713447,
+      "learning_rate": 1.69367176896109e-05,
+      "loss": 1.2419,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3315741165672327,
+      "grad_norm": 2.4682531973490263,
+      "learning_rate": 1.6907813818610597e-05,
+      "loss": 1.2358,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3327214318494722,
+      "grad_norm": 2.6517734913980466,
+      "learning_rate": 1.6878799141268107e-05,
+      "loss": 1.2399,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3338687471317118,
+      "grad_norm": 2.5133914405781392,
+      "learning_rate": 1.6849674122999878e-05,
+      "loss": 1.2167,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3350160624139514,
+      "grad_norm": 2.4861956286667852,
+      "learning_rate": 1.682043923099234e-05,
+      "loss": 1.2341,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3361633776961909,
+      "grad_norm": 2.476043042749583,
+      "learning_rate": 1.679109493419435e-05,
+      "loss": 1.2068,
+      "step": 1465
+    },
+    {
+      "epoch": 0.33731069297843047,
+      "grad_norm": 2.45123241603087,
+      "learning_rate": 1.6761641703309702e-05,
+      "loss": 1.2214,
+      "step": 1470
+    },
+    {
+      "epoch": 0.33845800826067,
+      "grad_norm": 2.4841737308673912,
+      "learning_rate": 1.673208001078958e-05,
+      "loss": 1.2608,
+      "step": 1475
+    },
+    {
+      "epoch": 0.33960532354290957,
+      "grad_norm": 2.559450448851183,
+      "learning_rate": 1.6702410330824962e-05,
+      "loss": 1.2409,
+      "step": 1480
+    },
+    {
+      "epoch": 0.34075263882514917,
+      "grad_norm": 2.7859811126370433,
+      "learning_rate": 1.6672633139339028e-05,
+      "loss": 1.2102,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3418999541073887,
+      "grad_norm": 2.57977983935283,
+      "learning_rate": 1.6642748913979515e-05,
+      "loss": 1.2228,
+      "step": 1490
+    },
+    {
+      "epoch": 0.34304726938962826,
+      "grad_norm": 2.533334211060345,
+      "learning_rate": 1.6612758134111072e-05,
+      "loss": 1.2459,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3441945846718678,
+      "grad_norm": 2.502270363127228,
+      "learning_rate": 1.6582661280807553e-05,
+      "loss": 1.2251,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3453418999541074,
+      "grad_norm": 2.6568753367931963,
+      "learning_rate": 1.65524588368443e-05,
+      "loss": 1.232,
+      "step": 1505
+    },
+    {
+      "epoch": 0.34648921523634696,
+      "grad_norm": 2.4840738066226775,
+      "learning_rate": 1.652215128669042e-05,
+      "loss": 1.2391,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3476365305185865,
+      "grad_norm": 2.473433983671793,
+      "learning_rate": 1.649173911650099e-05,
+      "loss": 1.2104,
+      "step": 1515
+    },
+    {
+      "epoch": 0.34878384580082605,
+      "grad_norm": 2.5021088399168177,
+      "learning_rate": 1.646122281410927e-05,
+      "loss": 1.2164,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3499311610830656,
+      "grad_norm": 2.4403773517332654,
+      "learning_rate": 1.6430602869018867e-05,
+      "loss": 1.2525,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3510784763653052,
+      "grad_norm": 2.601036367621894,
+      "learning_rate": 1.6399879772395915e-05,
+      "loss": 1.2246,
+      "step": 1530
+    },
+    {
+      "epoch": 0.35222579164754475,
+      "grad_norm": 2.526793579521016,
+      "learning_rate": 1.636905401706116e-05,
+      "loss": 1.2336,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3533731069297843,
+      "grad_norm": 2.4874250571730174,
+      "learning_rate": 1.633812609748206e-05,
+      "loss": 1.2335,
+      "step": 1540
+    },
+    {
+      "epoch": 0.35452042221202384,
+      "grad_norm": 2.4642069605568433,
+      "learning_rate": 1.630709650976487e-05,
+      "loss": 1.2266,
+      "step": 1545
+    },
+    {
+      "epoch": 0.35566773749426345,
+      "grad_norm": 2.941877356467088,
+      "learning_rate": 1.6275965751646682e-05,
+      "loss": 1.24,
+      "step": 1550
+    },
+    {
+      "epoch": 0.356815052776503,
+      "grad_norm": 2.906262225664714,
+      "learning_rate": 1.6244734322487415e-05,
+      "loss": 1.2295,
+      "step": 1555
+    },
+    {
+      "epoch": 0.35796236805874254,
+      "grad_norm": 2.501934363393508,
+      "learning_rate": 1.6213402723261852e-05,
+      "loss": 1.2005,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3591096833409821,
+      "grad_norm": 2.6600548799060095,
+      "learning_rate": 1.618197145655155e-05,
+      "loss": 1.2381,
+      "step": 1565
+    },
+    {
+      "epoch": 0.36025699862322164,
+      "grad_norm": 2.443840455534142,
       "learning_rate": 1.6150441026536827e-05,
-      "loss": 1.0487,
-      "step": 785
+      "loss": 1.2189,
+      "step": 1570
+    },
+    {
+      "epoch": 0.36140431390546124,
+      "grad_norm": 2.5147950136244126,
+      "learning_rate": 1.6118811938988632e-05,
+      "loss": 1.2204,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3625516291877008,
+      "grad_norm": 2.695659707022499,
+      "learning_rate": 1.6087084701260468e-05,
+      "loss": 1.2358,
+      "step": 1580
     },
     {
-      "epoch": 0.3625516291877008,
-      "grad_norm": 1.8722897359622224,
-      "learning_rate": 1.6087084701260468e-05,
-      "loss": 1.0349,
-      "step": 790
+      "epoch": 0.36369894446994033,
+      "grad_norm": 2.6213066605333943,
+      "learning_rate": 1.605525982228023e-05,
+      "loss": 1.2266,
+      "step": 1585
     },
     {
       "epoch": 0.3648462597521799,
-      "grad_norm": 4.170483033000324,
+      "grad_norm": 2.522457456754264,
       "learning_rate": 1.6023337812542048e-05,
-      "loss": 1.0416,
-      "step": 795
+      "loss": 1.2281,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3659935750344195,
+      "grad_norm": 2.611040744359311,
+      "learning_rate": 1.5991319184098107e-05,
+      "loss": 1.2363,
+      "step": 1595
     },
     {
       "epoch": 0.36714089031665903,
-      "grad_norm": 1.9386838285813406,
+      "grad_norm": 2.6055726423625605,
       "learning_rate": 1.5959204450550427e-05,
-      "loss": 1.0588,
-      "step": 800
+      "loss": 1.1846,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3682882055988986,
+      "grad_norm": 2.7334871548994233,
+      "learning_rate": 1.5926994127042615e-05,
+      "loss": 1.2297,
+      "step": 1605
     },
     {
       "epoch": 0.3694355208811381,
-      "grad_norm": 2.0117934404635136,
+      "grad_norm": 3.065228106081381,
       "learning_rate": 1.5894688730251613e-05,
-      "loss": 1.0454,
-      "step": 805
+      "loss": 1.2259,
+      "step": 1610
+    },
+    {
+      "epoch": 0.37058283616337767,
+      "grad_norm": 2.53275476586628,
+      "learning_rate": 1.586228877837941e-05,
+      "loss": 1.2223,
+      "step": 1615
     },
     {
       "epoch": 0.3717301514456173,
-      "grad_norm": 2.121087436021138,
+      "grad_norm": 2.6450492194453754,
       "learning_rate": 1.5829794791144723e-05,
-      "loss": 1.0388,
-      "step": 810
+      "loss": 1.2324,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3728774667278568,
+      "grad_norm": 2.6505407805972925,
+      "learning_rate": 1.5797207289774668e-05,
+      "loss": 1.206,
+      "step": 1625
     },
     {
       "epoch": 0.37402478201009637,
-      "grad_norm": 2.0084458470640394,
+      "grad_norm": 2.4512098548124084,
       "learning_rate": 1.57645267969964e-05,
-      "loss": 1.0425,
-      "step": 815
+      "loss": 1.2094,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3751720972923359,
+      "grad_norm": 2.4974996247628565,
+      "learning_rate": 1.5731753837028714e-05,
+      "loss": 1.2384,
+      "step": 1635
     },
     {
       "epoch": 0.3763194125745755,
-      "grad_norm": 1.9502413688336317,
+      "grad_norm": 2.5336333006362772,
       "learning_rate": 1.569888893557365e-05,
-      "loss": 1.0352,
-      "step": 820
+      "loss": 1.2179,
+      "step": 1640
+    },
+    {
+      "epoch": 0.37746672785681507,
+      "grad_norm": 2.480462659266847,
+      "learning_rate": 1.5665932619808058e-05,
+      "loss": 1.2137,
+      "step": 1645
     },
     {
       "epoch": 0.3786140431390546,
-      "grad_norm": 1.9753852329094403,
+      "grad_norm": 2.5575394813805388,
       "learning_rate": 1.5632885418375136e-05,
-      "loss": 1.0562,
-      "step": 825
+      "loss": 1.2099,
+      "step": 1650
+    },
+    {
+      "epoch": 0.37976135842129416,
+      "grad_norm": 2.638731671620944,
+      "learning_rate": 1.5599747861375957e-05,
+      "loss": 1.2067,
+      "step": 1655
     },
     {
       "epoch": 0.3809086737035337,
-      "grad_norm": 2.0333895336233745,
+      "grad_norm": 2.7046597751215704,
       "learning_rate": 1.556652048036096e-05,
-      "loss": 1.0571,
-      "step": 830
+      "loss": 1.2139,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3820559889857733,
+      "grad_norm": 2.434498118282716,
+      "learning_rate": 1.553320380832143e-05,
+      "loss": 1.2299,
+      "step": 1665
     },
     {
       "epoch": 0.38320330426801286,
-      "grad_norm": 2.2000920597322464,
+      "grad_norm": 2.5520040709852148,
       "learning_rate": 1.549979837968094e-05,
-      "loss": 1.0504,
-      "step": 835
+      "loss": 1.1987,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3843506195502524,
+      "grad_norm": 2.4129325989146735,
+      "learning_rate": 1.5466304730286795e-05,
+      "loss": 1.2333,
+      "step": 1675
     },
     {
       "epoch": 0.38549793483249195,
-      "grad_norm": 2.0041860947851777,
+      "grad_norm": 2.3734664430568286,
       "learning_rate": 1.5432723397401406e-05,
-      "loss": 1.0483,
-      "step": 840
+      "loss": 1.2118,
+      "step": 1680
+    },
+    {
+      "epoch": 0.38664525011473155,
+      "grad_norm": 2.489082024919344,
+      "learning_rate": 1.5399054919693704e-05,
+      "loss": 1.2249,
+      "step": 1685
     },
     {
       "epoch": 0.3877925653969711,
-      "grad_norm": 2.063158568485873,
+      "grad_norm": 2.632893560679943,
       "learning_rate": 1.5365299837230483e-05,
-      "loss": 1.0479,
-      "step": 845
+      "loss": 1.2447,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38893988067921065,
+      "grad_norm": 2.502360122889922,
+      "learning_rate": 1.5331458691467742e-05,
+      "loss": 1.2039,
+      "step": 1695
     },
     {
       "epoch": 0.3900871959614502,
-      "grad_norm": 2.0282677973594017,
+      "grad_norm": 2.56570123550657,
       "learning_rate": 1.5297532025241993e-05,
-      "loss": 1.03,
-      "step": 850
+      "loss": 1.2,
+      "step": 1700
+    },
+    {
+      "epoch": 0.39123451124368974,
+      "grad_norm": 2.4982474943791706,
+      "learning_rate": 1.5263520382761563e-05,
+      "loss": 1.2182,
+      "step": 1705
     },
     {
       "epoch": 0.39238182652592934,
-      "grad_norm": 1.8181342854668368,
+      "grad_norm": 2.5057101433951776,
       "learning_rate": 1.5229424309597853e-05,
-      "loss": 1.0461,
-      "step": 855
+      "loss": 1.2199,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3935291418081689,
+      "grad_norm": 2.5257720554796403,
+      "learning_rate": 1.5195244352676606e-05,
+      "loss": 1.2072,
+      "step": 1715
     },
     {
       "epoch": 0.39467645709040844,
-      "grad_norm": 1.9188404459309296,
+      "grad_norm": 2.5247477276803516,
       "learning_rate": 1.5160981060269107e-05,
-      "loss": 1.0926,
-      "step": 860
+      "loss": 1.2059,
+      "step": 1720
+    },
+    {
+      "epoch": 0.395823772372648,
+      "grad_norm": 2.3583351198103784,
+      "learning_rate": 1.5126634981983412e-05,
+      "loss": 1.215,
+      "step": 1725
     },
     {
       "epoch": 0.3969710876548876,
-      "grad_norm": 1.8587214491979347,
+      "grad_norm": 2.459251655659526,
       "learning_rate": 1.5092206668755518e-05,
-      "loss": 1.023,
-      "step": 865
+      "loss": 1.203,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39811840293712714,
+      "grad_norm": 2.5334007418963753,
+      "learning_rate": 1.5057696672840529e-05,
+      "loss": 1.2473,
+      "step": 1735
     },
     {
       "epoch": 0.3992657182193667,
-      "grad_norm": 2.0702564004428736,
+      "grad_norm": 2.3391171431662108,
       "learning_rate": 1.5023105547803807e-05,
-      "loss": 1.0497,
-      "step": 870
+      "loss": 1.2226,
+      "step": 1740
+    },
+    {
+      "epoch": 0.40041303350160623,
+      "grad_norm": 2.3984874089782573,
+      "learning_rate": 1.4988433848512074e-05,
+      "loss": 1.1848,
+      "step": 1745
     },
     {
       "epoch": 0.4015603487838458,
-      "grad_norm": 1.866283622621383,
+      "grad_norm": 2.3056101239425986,
       "learning_rate": 1.4953682131124527e-05,
-      "loss": 1.0548,
-      "step": 875
+      "loss": 1.2188,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4027076640660854,
+      "grad_norm": 2.556190876138143,
+      "learning_rate": 1.491885095308391e-05,
+      "loss": 1.1984,
+      "step": 1755
     },
     {
       "epoch": 0.4038549793483249,
-      "grad_norm": 2.3748260414549613,
+      "grad_norm": 2.4518910879213895,
       "learning_rate": 1.4883940873107572e-05,
-      "loss": 1.0354,
-      "step": 880
+      "loss": 1.2238,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4050022946305645,
+      "grad_norm": 2.473122618546395,
+      "learning_rate": 1.4848952451178508e-05,
+      "loss": 1.207,
+      "step": 1765
     },
     {
       "epoch": 0.406149609912804,
-      "grad_norm": 1.8622352558527133,
+      "grad_norm": 2.359534310542907,
       "learning_rate": 1.4813886248536376e-05,
-      "loss": 1.0272,
-      "step": 885
+      "loss": 1.1977,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4072969251950436,
+      "grad_norm": 2.477818274819773,
+      "learning_rate": 1.4778742827668484e-05,
+      "loss": 1.2391,
+      "step": 1775
     },
     {
       "epoch": 0.40844424047728317,
-      "grad_norm": 1.8451716042859445,
+      "grad_norm": 2.530388893111552,
       "learning_rate": 1.4743522752300793e-05,
-      "loss": 1.0346,
-      "step": 890
+      "loss": 1.2163,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4095915557595227,
+      "grad_norm": 2.5519663146982605,
+      "learning_rate": 1.4708226587388845e-05,
+      "loss": 1.2532,
+      "step": 1785
     },
     {
       "epoch": 0.41073887104176227,
-      "grad_norm": 1.7791032929700201,
+      "grad_norm": 2.394616840944208,
       "learning_rate": 1.467285489910872e-05,
-      "loss": 1.0445,
-      "step": 895
+      "loss": 1.2155,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4118861863240018,
+      "grad_norm": 2.412230711815024,
+      "learning_rate": 1.4637408254847936e-05,
+      "loss": 1.1907,
+      "step": 1795
     },
     {
       "epoch": 0.4130335016062414,
-      "grad_norm": 1.9142844645885038,
+      "grad_norm": 2.5078430973460817,
       "learning_rate": 1.4601887223196374e-05,
-      "loss": 1.0367,
-      "step": 900
+      "loss": 1.1998,
+      "step": 1800
+    },
+    {
+      "epoch": 0.41418081688848096,
+      "grad_norm": 2.4576469102426373,
+      "learning_rate": 1.4566292373937133e-05,
+      "loss": 1.1996,
+      "step": 1805
     },
     {
       "epoch": 0.4153281321707205,
-      "grad_norm": 1.9821960797296039,
+      "grad_norm": 2.7831279786324137,
       "learning_rate": 1.4530624278037406e-05,
-      "loss": 1.0459,
-      "step": 905
+      "loss": 1.2164,
+      "step": 1810
+    },
+    {
+      "epoch": 0.41647544745296006,
+      "grad_norm": 2.478975911392516,
+      "learning_rate": 1.449488350763931e-05,
+      "loss": 1.2199,
+      "step": 1815
     },
     {
       "epoch": 0.41762276273519966,
-      "grad_norm": 1.8030416883238183,
+      "grad_norm": 2.446888200491552,
       "learning_rate": 1.4459070636050721e-05,
-      "loss": 1.0395,
-      "step": 910
+      "loss": 1.1929,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4187700780174392,
+      "grad_norm": 2.456196149028385,
+      "learning_rate": 1.4423186237736063e-05,
+      "loss": 1.2198,
+      "step": 1825
     },
     {
       "epoch": 0.41991739329967875,
-      "grad_norm": 1.8541049154816385,
+      "grad_norm": 2.4671601504916296,
       "learning_rate": 1.4387230888307098e-05,
-      "loss": 1.0297,
-      "step": 915
+      "loss": 1.2093,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4210647085819183,
+      "grad_norm": 2.469445463788116,
+      "learning_rate": 1.4351205164513708e-05,
+      "loss": 1.2115,
+      "step": 1835
     },
     {
       "epoch": 0.42221202386415785,
-      "grad_norm": 1.882570932670042,
+      "grad_norm": 2.412944791501572,
       "learning_rate": 1.4315109644234619e-05,
-      "loss": 1.0526,
-      "step": 920
+      "loss": 1.1899,
+      "step": 1840
+    },
+    {
+      "epoch": 0.42335933914639745,
+      "grad_norm": 2.537897011913316,
+      "learning_rate": 1.427894490646815e-05,
+      "loss": 1.1931,
+      "step": 1845
     },
     {
       "epoch": 0.424506654428637,
-      "grad_norm": 2.104659835262144,
+      "grad_norm": 2.4797926920153275,
       "learning_rate": 1.4242711531322912e-05,
-      "loss": 1.0602,
-      "step": 925
+      "loss": 1.2071,
+      "step": 1850
+    },
+    {
+      "epoch": 0.42565396971087655,
+      "grad_norm": 2.4620300685296095,
+      "learning_rate": 1.420641010000852e-05,
+      "loss": 1.1901,
+      "step": 1855
     },
     {
       "epoch": 0.4268012849931161,
-      "grad_norm": 1.9520034799103085,
+      "grad_norm": 2.5654139268387905,
       "learning_rate": 1.4170041194826247e-05,
-      "loss": 1.0513,
-      "step": 930
+      "loss": 1.1672,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4279486002753557,
+      "grad_norm": 2.349473278321755,
+      "learning_rate": 1.4133605399159706e-05,
+      "loss": 1.1863,
+      "step": 1865
     },
     {
       "epoch": 0.42909591555759524,
-      "grad_norm": 1.8636499197241878,
+      "grad_norm": 2.4170921121837603,
       "learning_rate": 1.4097103297465471e-05,
-      "loss": 1.0313,
-      "step": 935
+      "loss": 1.1997,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4302432308398348,
+      "grad_norm": 2.5029515745014277,
+      "learning_rate": 1.4060535475263725e-05,
+      "loss": 1.2134,
+      "step": 1875
     },
     {
       "epoch": 0.43139054612207434,
-      "grad_norm": 1.866901038371879,
+      "grad_norm": 2.3135883184156136,
       "learning_rate": 1.402390251912885e-05,
-      "loss": 1.0492,
-      "step": 940
+      "loss": 1.1844,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4325378614043139,
+      "grad_norm": 2.404470246112949,
+      "learning_rate": 1.398720501668002e-05,
+      "loss": 1.202,
+      "step": 1885
     },
     {
       "epoch": 0.4336851766865535,
-      "grad_norm": 1.9095505177623704,
+      "grad_norm": 2.5569577276828035,
       "learning_rate": 1.395044355657178e-05,
-      "loss": 1.0427,
-      "step": 945
+      "loss": 1.2208,
+      "step": 1890
+    },
+    {
+      "epoch": 0.43483249196879303,
+      "grad_norm": 2.3371018844193108,
+      "learning_rate": 1.391361872848461e-05,
+      "loss": 1.1737,
+      "step": 1895
     },
     {
       "epoch": 0.4359798072510326,
-      "grad_norm": 1.8833835923493178,
+      "grad_norm": 2.660539570330875,
       "learning_rate": 1.387673112311545e-05,
-      "loss": 1.0062,
-      "step": 950
+      "loss": 1.1778,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4371271225332721,
+      "grad_norm": 2.3038070568716793,
+      "learning_rate": 1.3839781332168236e-05,
+      "loss": 1.1921,
+      "step": 1905
     },
     {
       "epoch": 0.4382744378155117,
-      "grad_norm": 1.8503750464774162,
+      "grad_norm": 2.4670441595509995,
       "learning_rate": 1.3802769948344406e-05,
-      "loss": 1.0584,
-      "step": 955
+      "loss": 1.1833,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4394217530977513,
+      "grad_norm": 2.4330880664690864,
+      "learning_rate": 1.3765697565333387e-05,
+      "loss": 1.1835,
+      "step": 1915
     },
     {
       "epoch": 0.4405690683799908,
-      "grad_norm": 1.9850425042063173,
+      "grad_norm": 2.5476719771012886,
       "learning_rate": 1.3728564777803089e-05,
-      "loss": 1.0299,
-      "step": 960
+      "loss": 1.1809,
+      "step": 1920
+    },
+    {
+      "epoch": 0.44171638366223037,
+      "grad_norm": 2.4245419288540906,
+      "learning_rate": 1.369137218139034e-05,
+      "loss": 1.216,
+      "step": 1925
     },
     {
       "epoch": 0.4428636989444699,
-      "grad_norm": 2.0470542749046308,
+      "grad_norm": 2.48980878325673,
       "learning_rate": 1.3654120372691361e-05,
-      "loss": 1.0211,
-      "step": 965
+      "loss": 1.2043,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4440110142267095,
+      "grad_norm": 2.309423158584073,
+      "learning_rate": 1.3616809949252168e-05,
+      "loss": 1.2084,
+      "step": 1935
     },
     {
       "epoch": 0.44515832950894907,
-      "grad_norm": 1.963916772949296,
+      "grad_norm": 2.4292416714345526,
       "learning_rate": 1.3579441509559007e-05,
-      "loss": 1.0453,
-      "step": 970
+      "loss": 1.2251,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4463056447911886,
+      "grad_norm": 2.5052554766414996,
+      "learning_rate": 1.3542015653028742e-05,
+      "loss": 1.2122,
+      "step": 1945
     },
     {
       "epoch": 0.44745296007342816,
-      "grad_norm": 1.889454130597838,
+      "grad_norm": 2.5664797047962433,
       "learning_rate": 1.350453297999925e-05,
-      "loss": 1.0441,
-      "step": 975
+      "loss": 1.2101,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4486002753556677,
+      "grad_norm": 2.5032873925292973,
+      "learning_rate": 1.3466994091719782e-05,
+      "loss": 1.2125,
+      "step": 1955
     },
     {
       "epoch": 0.4497475906379073,
-      "grad_norm": 1.8300661545935728,
+      "grad_norm": 2.4259701176858126,
       "learning_rate": 1.3429399590341325e-05,
-      "loss": 1.0109,
-      "step": 980
+      "loss": 1.1872,
+      "step": 1960
+    },
+    {
+      "epoch": 0.45089490592014686,
+      "grad_norm": 2.4344258356766844,
+      "learning_rate": 1.3391750078906939e-05,
+      "loss": 1.1538,
+      "step": 1965
     },
     {
       "epoch": 0.4520422212023864,
-      "grad_norm": 1.8927209237187554,
+      "grad_norm": 2.473246644054591,
       "learning_rate": 1.3354046161342087e-05,
-      "loss": 1.0153,
-      "step": 985
+      "loss": 1.1941,
+      "step": 1970
+    },
+    {
+      "epoch": 0.45318953648462595,
+      "grad_norm": 2.4762521196200233,
+      "learning_rate": 1.3316288442444943e-05,
+      "loss": 1.2287,
+      "step": 1975
     },
     {
       "epoch": 0.45433685176686556,
-      "grad_norm": 1.8419785511286724,
+      "grad_norm": 2.5747496155427925,
       "learning_rate": 1.327847752787669e-05,
-      "loss": 1.0234,
-      "step": 990
+      "loss": 1.1946,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4554841670491051,
+      "grad_norm": 2.4957625591645396,
+      "learning_rate": 1.324061402415182e-05,
+      "loss": 1.1943,
+      "step": 1985
     },
     {
       "epoch": 0.45663148233134465,
-      "grad_norm": 1.8479631910892635,
+      "grad_norm": 2.4803686926028354,
       "learning_rate": 1.3202698538628376e-05,
-      "loss": 1.0515,
-      "step": 995
+      "loss": 1.1958,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4577787976135842,
+      "grad_norm": 2.515928887088066,
+      "learning_rate": 1.3164731679498249e-05,
+      "loss": 1.1914,
+      "step": 1995
     },
     {
       "epoch": 0.45892611289582375,
-      "grad_norm": 1.8454383680403843,
+      "grad_norm": 2.40416499995082,
       "learning_rate": 1.3126714055777378e-05,
-      "loss": 1.0283,
-      "step": 1000
+      "loss": 1.2007,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46007342817806335,
+      "grad_norm": 2.4071385033932726,
+      "learning_rate": 1.3088646277296018e-05,
+      "loss": 1.205,
+      "step": 2005
     },
     {
       "epoch": 0.4612207434603029,
-      "grad_norm": 1.8614938060407424,
+      "grad_norm": 2.3336952244417857,
       "learning_rate": 1.3050528954688932e-05,
-      "loss": 1.028,
-      "step": 1005
+      "loss": 1.1807,
+      "step": 2010
+    },
+    {
+      "epoch": 0.46236805874254244,
+      "grad_norm": 2.570415975161917,
+      "learning_rate": 1.3012362699385616e-05,
+      "loss": 1.1677,
+      "step": 2015
     },
     {
       "epoch": 0.463515374024782,
-      "grad_norm": 2.138120924499351,
+      "grad_norm": 2.4434180063843396,
       "learning_rate": 1.2974148123600477e-05,
-      "loss": 1.0154,
-      "step": 1010
+      "loss": 1.1938,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4646626893070216,
+      "grad_norm": 2.388368907016271,
+      "learning_rate": 1.2935885840323015e-05,
+      "loss": 1.1938,
+      "step": 2025
     },
     {
       "epoch": 0.46581000458926114,
-      "grad_norm": 1.8576229088325686,
+      "grad_norm": 2.4093448614688953,
       "learning_rate": 1.2897576463307999e-05,
-      "loss": 1.036,
-      "step": 1015
+      "loss": 1.1814,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4669573198715007,
+      "grad_norm": 2.5376955340059646,
+      "learning_rate": 1.285922060706561e-05,
+      "loss": 1.1668,
+      "step": 2035
     },
     {
       "epoch": 0.46810463515374023,
-      "grad_norm": 2.0045877966342776,
+      "grad_norm": 2.417127573905711,
       "learning_rate": 1.2820818886851599e-05,
-      "loss": 1.0382,
-      "step": 1020
+      "loss": 1.1821,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4692519504359798,
+      "grad_norm": 2.298184877749985,
+      "learning_rate": 1.2782371918657393e-05,
+      "loss": 1.1954,
+      "step": 2045
     },
     {
       "epoch": 0.4703992657182194,
-      "grad_norm": 1.8120186612041058,
+      "grad_norm": 2.417757124526873,
       "learning_rate": 1.2743880319200241e-05,
-      "loss": 1.0532,
-      "step": 1025
+      "loss": 1.166,
+      "step": 2050
+    },
+    {
+      "epoch": 0.47154658100045893,
+      "grad_norm": 2.371659919793289,
+      "learning_rate": 1.270534470591331e-05,
+      "loss": 1.1936,
+      "step": 2055
     },
     {
       "epoch": 0.4726938962826985,
-      "grad_norm": 1.8202058192088733,
+      "grad_norm": 2.4662501454162062,
       "learning_rate": 1.2666765696935773e-05,
-      "loss": 1.0223,
-      "step": 1030
+      "loss": 1.2021,
+      "step": 2060
+    },
+    {
+      "epoch": 0.473841211564938,
+      "grad_norm": 2.4729202869839115,
+      "learning_rate": 1.2628143911102905e-05,
+      "loss": 1.1952,
+      "step": 2065
     },
     {
       "epoch": 0.47498852684717763,
-      "grad_norm": 1.8341249835092734,
+      "grad_norm": 2.3449678946057113,
       "learning_rate": 1.2589479967936163e-05,
-      "loss": 1.0327,
-      "step": 1035
+      "loss": 1.1959,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4761358421294172,
+      "grad_norm": 2.439109075077548,
+      "learning_rate": 1.2550774487633218e-05,
+      "loss": 1.1864,
+      "step": 2075
     },
     {
       "epoch": 0.4772831574116567,
-      "grad_norm": 1.890017754125909,
+      "grad_norm": 2.529102082715549,
       "learning_rate": 1.2512028091058044e-05,
-      "loss": 1.0282,
-      "step": 1040
+      "loss": 1.186,
+      "step": 2080
+    },
+    {
+      "epoch": 0.47843047269389627,
+      "grad_norm": 2.376604159825014,
+      "learning_rate": 1.2473241399730931e-05,
+      "loss": 1.1701,
+      "step": 2085
     },
     {
       "epoch": 0.4795777879761358,
-      "grad_norm": 1.7963761077816227,
+      "grad_norm": 2.4169774686564263,
       "learning_rate": 1.2434415035818535e-05,
-      "loss": 1.015,
-      "step": 1045
+      "loss": 1.2191,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4807251032583754,
+      "grad_norm": 2.532981770965977,
+      "learning_rate": 1.239554962212388e-05,
+      "loss": 1.2286,
+      "step": 2095
     },
     {
       "epoch": 0.48187241854061497,
-      "grad_norm": 1.89045217328483,
+      "grad_norm": 2.378557479987113,
       "learning_rate": 1.2356645782076384e-05,
-      "loss": 1.0189,
-      "step": 1050
+      "loss": 1.1677,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4830197338228545,
+      "grad_norm": 2.709648710575487,
+      "learning_rate": 1.2317704139721847e-05,
+      "loss": 1.1751,
+      "step": 2105
     },
     {
       "epoch": 0.48416704910509406,
-      "grad_norm": 1.8184075309069563,
+      "grad_norm": 2.25659549412137,
       "learning_rate": 1.2278725319712449e-05,
-      "loss": 1.0388,
-      "step": 1055
+      "loss": 1.1706,
+      "step": 2110
+    },
+    {
+      "epoch": 0.48531436438733366,
+      "grad_norm": 2.410190860485096,
+      "learning_rate": 1.2239709947296722e-05,
+      "loss": 1.1681,
+      "step": 2115
     },
     {
       "epoch": 0.4864616796695732,
-      "grad_norm": 1.7960523305438247,
+      "grad_norm": 2.5109950886921233,
       "learning_rate": 1.2200658648309531e-05,
-      "loss": 1.0447,
-      "step": 1060
+      "loss": 1.1655,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48760899495181276,
+      "grad_norm": 2.4008584174966243,
+      "learning_rate": 1.2161572049162027e-05,
+      "loss": 1.1665,
+      "step": 2125
     },
     {
       "epoch": 0.4887563102340523,
-      "grad_norm": 1.8467823592474473,
+      "grad_norm": 2.3827246030282194,
       "learning_rate": 1.2122450776831593e-05,
-      "loss": 1.0122,
-      "step": 1065
+      "loss": 1.1662,
+      "step": 2130
+    },
+    {
+      "epoch": 0.48990362551629185,
+      "grad_norm": 2.424648067070357,
+      "learning_rate": 1.208329545885181e-05,
+      "loss": 1.2112,
+      "step": 2135
     },
     {
       "epoch": 0.49105094079853145,
-      "grad_norm": 1.7814611078259526,
+      "grad_norm": 2.412168453903357,
       "learning_rate": 1.2044106723302364e-05,
-      "loss": 0.998,
-      "step": 1070
+      "loss": 1.18,
+      "step": 2140
+    },
+    {
+      "epoch": 0.492198256080771,
+      "grad_norm": 2.5463621491353865,
+      "learning_rate": 1.200488519879899e-05,
+      "loss": 1.1426,
+      "step": 2145
     },
     {
       "epoch": 0.49334557136301055,
-      "grad_norm": 1.8655763524159674,
+      "grad_norm": 3.2345084391862216,
       "learning_rate": 1.1965631514483376e-05,
-      "loss": 1.0051,
-      "step": 1075
+      "loss": 1.1739,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4944928866452501,
+      "grad_norm": 2.3242325394764176,
+      "learning_rate": 1.1926346300013078e-05,
+      "loss": 1.1795,
+      "step": 2155
     },
     {
       "epoch": 0.4956402019274897,
-      "grad_norm": 1.8504586712684936,
+      "grad_norm": 2.4343577481110334,
       "learning_rate": 1.1887030185551427e-05,
-      "loss": 1.0173,
-      "step": 1080
+      "loss": 1.1751,
+      "step": 2160
+    },
+    {
+      "epoch": 0.49678751720972925,
+      "grad_norm": 2.536903854244761,
+      "learning_rate": 1.18476838017574e-05,
+      "loss": 1.1886,
+      "step": 2165
     },
     {
       "epoch": 0.4979348324919688,
-      "grad_norm": 1.8553598830430822,
+      "grad_norm": 2.518861365226075,
       "learning_rate": 1.1808307779775518e-05,
-      "loss": 1.0276,
-      "step": 1085
+      "loss": 1.1822,
+      "step": 2170
+    },
+    {
+      "epoch": 0.49908214777420834,
+      "grad_norm": 2.33985336402769,
+      "learning_rate": 1.176890275122573e-05,
+      "loss": 1.1852,
+      "step": 2175
     },
     {
       "epoch": 0.5002294630564479,
-      "grad_norm": 1.7545364104117196,
+      "grad_norm": 2.695318454608909,
       "learning_rate": 1.1729469348193263e-05,
-      "loss": 1.0079,
-      "step": 1090
+      "loss": 1.164,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5013767783386874,
+      "grad_norm": 2.4114946936382835,
+      "learning_rate": 1.1690008203218493e-05,
+      "loss": 1.1793,
+      "step": 2185
     },
     {
       "epoch": 0.502524093620927,
-      "grad_norm": 1.7459754103129477,
+      "grad_norm": 2.414381546692464,
       "learning_rate": 1.1650519949286797e-05,
-      "loss": 1.0049,
-      "step": 1095
+      "loss": 1.1716,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5036714089031666,
+      "grad_norm": 2.3096954634729525,
+      "learning_rate": 1.1611005219818392e-05,
+      "loss": 1.1771,
+      "step": 2195
     },
     {
       "epoch": 0.5048187241854062,
-      "grad_norm": 1.8061298693213486,
+      "grad_norm": 2.377639229729014,
       "learning_rate": 1.1571464648658201e-05,
-      "loss": 0.9806,
-      "step": 1100
+      "loss": 1.1594,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5059660394676457,
+      "grad_norm": 2.413947191443181,
+      "learning_rate": 1.1531898870065645e-05,
+      "loss": 1.1377,
+      "step": 2205
     },
     {
       "epoch": 0.5071133547498853,
-      "grad_norm": 1.7062918430841572,
+      "grad_norm": 2.413078162601223,
       "learning_rate": 1.1492308518704507e-05,
-      "loss": 1.0133,
-      "step": 1105
+      "loss": 1.1643,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5082606700321248,
+      "grad_norm": 2.3703796014060083,
+      "learning_rate": 1.145269422963272e-05,
+      "loss": 1.1672,
+      "step": 2215
     },
     {
       "epoch": 0.5094079853143644,
-      "grad_norm": 1.744436514313244,
+      "grad_norm": 2.52185413660041,
       "learning_rate": 1.1413056638292215e-05,
-      "loss": 0.9962,
-      "step": 1110
+      "loss": 1.1482,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5105553005966039,
+      "grad_norm": 2.5494899458552704,
+      "learning_rate": 1.1373396380498683e-05,
+      "loss": 1.1855,
+      "step": 2225
     },
     {
       "epoch": 0.5117026158788435,
-      "grad_norm": 1.7695374210335713,
+      "grad_norm": 2.4759934453822687,
       "learning_rate": 1.1333714092431423e-05,
-      "loss": 1.0085,
-      "step": 1115
+      "loss": 1.183,
+      "step": 2230
+    },
+    {
+      "epoch": 0.512849931161083,
+      "grad_norm": 2.739893863248377,
+      "learning_rate": 1.1294010410623107e-05,
+      "loss": 1.1619,
+      "step": 2235
     },
     {
       "epoch": 0.5139972464433227,
-      "grad_norm": 1.9055267822840887,
+      "grad_norm": 2.3307913465594736,
       "learning_rate": 1.1254285971949574e-05,
-      "loss": 1.0198,
-      "step": 1120
+      "loss": 1.1752,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5151445617255622,
+      "grad_norm": 2.407069188869029,
+      "learning_rate": 1.1214541413619628e-05,
+      "loss": 1.1682,
+      "step": 2245
     },
     {
       "epoch": 0.5162918770078018,
-      "grad_norm": 1.7470563158478891,
+      "grad_norm": 2.38595550036874,
       "learning_rate": 1.1174777373164797e-05,
-      "loss": 1.005,
-      "step": 1125
+      "loss": 1.1804,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5174391922900413,
+      "grad_norm": 2.3691735593741865,
+      "learning_rate": 1.1134994488429128e-05,
+      "loss": 1.1982,
+      "step": 2255
     },
     {
       "epoch": 0.5185865075722809,
-      "grad_norm": 1.8402758989282078,
+      "grad_norm": 2.3715490304848696,
       "learning_rate": 1.109519339755893e-05,
-      "loss": 1.0225,
-      "step": 1130
+      "loss": 1.18,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5197338228545204,
+      "grad_norm": 2.452149269203383,
+      "learning_rate": 1.1055374738992561e-05,
+      "loss": 1.1726,
+      "step": 2265
     },
     {
       "epoch": 0.52088113813676,
-      "grad_norm": 1.9185683165429528,
+      "grad_norm": 2.3884283160593136,
       "learning_rate": 1.1015539151450172e-05,
-      "loss": 1.0202,
-      "step": 1135
+      "loss": 1.1706,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5220284534189995,
+      "grad_norm": 2.3927663987307253,
+      "learning_rate": 1.0975687273923474e-05,
+      "loss": 1.1737,
+      "step": 2275
     },
     {
       "epoch": 0.523175768701239,
-      "grad_norm": 1.7862218928745999,
+      "grad_norm": 2.480726870506535,
       "learning_rate": 1.0935819745665477e-05,
-      "loss": 0.9938,
-      "step": 1140
+      "loss": 1.177,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5243230839834787,
+      "grad_norm": 2.6254719429941,
+      "learning_rate": 1.0895937206180243e-05,
+      "loss": 1.2074,
+      "step": 2285
     },
     {
       "epoch": 0.5254703992657183,
-      "grad_norm": 1.7946257670444443,
+      "grad_norm": 2.4621930223006436,
       "learning_rate": 1.0856040295212614e-05,
-      "loss": 1.0203,
-      "step": 1145
+      "loss": 1.1801,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5266177145479578,
+      "grad_norm": 2.411669222534401,
+      "learning_rate": 1.0816129652737976e-05,
+      "loss": 1.1549,
+      "step": 2295
     },
     {
       "epoch": 0.5277650298301974,
-      "grad_norm": 1.8084692996883154,
+      "grad_norm": 2.3766352222254876,
       "learning_rate": 1.077620591895197e-05,
-      "loss": 1.0132,
-      "step": 1150
+      "loss": 1.1882,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5289123451124369,
+      "grad_norm": 2.387743064914442,
+      "learning_rate": 1.0736269734260232e-05,
+      "loss": 1.1474,
+      "step": 2305
     },
     {
       "epoch": 0.5300596603946764,
-      "grad_norm": 1.8496851492898159,
+      "grad_norm": 2.336284695991881,
       "learning_rate": 1.069632173926812e-05,
-      "loss": 1.009,
-      "step": 1155
+      "loss": 1.145,
+      "step": 2310
+    },
+    {
+      "epoch": 0.531206975676916,
+      "grad_norm": 2.234183118079691,
+      "learning_rate": 1.0656362574770442e-05,
+      "loss": 1.1274,
+      "step": 2315
     },
     {
       "epoch": 0.5323542909591555,
-      "grad_norm": 1.7965406450823929,
+      "grad_norm": 2.494121033628584,
       "learning_rate": 1.0616392881741166e-05,
-      "loss": 0.9898,
-      "step": 1160
+      "loss": 1.1803,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5335016062413951,
+      "grad_norm": 2.4656108820538294,
+      "learning_rate": 1.0576413301323148e-05,
+      "loss": 1.1594,
+      "step": 2325
     },
     {
       "epoch": 0.5346489215236347,
-      "grad_norm": 1.8696271364197514,
+      "grad_norm": 2.3529378211004017,
       "learning_rate": 1.0536424474817848e-05,
-      "loss": 0.9949,
-      "step": 1165
+      "loss": 1.1669,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5357962368058743,
+      "grad_norm": 2.447756807412706,
+      "learning_rate": 1.0496427043675032e-05,
+      "loss": 1.1577,
+      "step": 2335
     },
     {
       "epoch": 0.5369435520881138,
-      "grad_norm": 1.8512029165438884,
+      "grad_norm": 2.4126468040306768,
       "learning_rate": 1.0456421649482502e-05,
-      "loss": 1.0065,
-      "step": 1170
+      "loss": 1.1351,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5380908673703534,
+      "grad_norm": 2.582646994016727,
+      "learning_rate": 1.041640893395578e-05,
+      "loss": 1.1625,
+      "step": 2345
     },
     {
       "epoch": 0.5392381826525929,
-      "grad_norm": 1.833426494602823,
+      "grad_norm": 2.424869345038524,
       "learning_rate": 1.0376389538927841e-05,
-      "loss": 1.0158,
-      "step": 1175
+      "loss": 1.1444,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5403854979348325,
+      "grad_norm": 2.3759407482791186,
+      "learning_rate": 1.0336364106338793e-05,
+      "loss": 1.1473,
+      "step": 2355
     },
     {
       "epoch": 0.541532813217072,
-      "grad_norm": 1.8287651297802663,
+      "grad_norm": 2.3954926124121982,
       "learning_rate": 1.0296333278225599e-05,
-      "loss": 1.0064,
-      "step": 1180
+      "loss": 1.15,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5426801284993116,
+      "grad_norm": 2.3048166671918517,
+      "learning_rate": 1.0256297696711764e-05,
+      "loss": 1.1688,
+      "step": 2365
     },
     {
       "epoch": 0.5438274437815511,
-      "grad_norm": 1.8964046455895252,
+      "grad_norm": 2.455044790822921,
       "learning_rate": 1.0216258003997044e-05,
-      "loss": 1.0198,
-      "step": 1185
+      "loss": 1.1499,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5449747590637908,
+      "grad_norm": 2.440410779702323,
+      "learning_rate": 1.0176214842347143e-05,
+      "loss": 1.1883,
+      "step": 2375
     },
     {
       "epoch": 0.5461220743460303,
-      "grad_norm": 1.777279291853422,
+      "grad_norm": 2.469069553035829,
       "learning_rate": 1.0136168854083401e-05,
-      "loss": 0.9712,
-      "step": 1190
+      "loss": 1.1564,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5472693896282699,
+      "grad_norm": 2.36749378237334,
+      "learning_rate": 1.0096120681572513e-05,
+      "loss": 1.1581,
+      "step": 2385
     },
     {
       "epoch": 0.5484167049105094,
-      "grad_norm": 1.8585809331304046,
+      "grad_norm": 2.4927612686073446,
       "learning_rate": 1.0056070967216199e-05,
-      "loss": 0.9954,
-      "step": 1195
+      "loss": 1.1604,
+      "step": 2390
+    },
+    {
+      "epoch": 0.549564020192749,
+      "grad_norm": 2.282784095648045,
+      "learning_rate": 1.0016020353440916e-05,
+      "loss": 1.1495,
+      "step": 2395
     },
     {
       "epoch": 0.5507113354749885,
-      "grad_norm": 1.9725460354320736,
+      "grad_norm": 2.319998209280646,
       "learning_rate": 9.975969482687547e-06,
-      "loss": 0.9835,
-      "step": 1200
+      "loss": 1.1709,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5518586507572281,
+      "grad_norm": 2.4648551075271348,
+      "learning_rate": 9.935918997401104e-06,
+      "loss": 1.2095,
+      "step": 2405
     },
     {
       "epoch": 0.5530059660394676,
-      "grad_norm": 1.7245123642288822,
+      "grad_norm": 2.401099714183983,
       "learning_rate": 9.8958695400204e-06,
-      "loss": 1.0004,
-      "step": 1205
+      "loss": 1.1616,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5541532813217072,
+      "grad_norm": 2.427714048915411,
+      "learning_rate": 9.855821752967779e-06,
+      "loss": 1.1622,
+      "step": 2415
     },
     {
       "epoch": 0.5553005966039468,
-      "grad_norm": 1.8366846045184153,
+      "grad_norm": 2.3990234519386924,
       "learning_rate": 9.815776278638772e-06,
-      "loss": 1.0265,
-      "step": 1210
+      "loss": 1.1666,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5564479118861864,
+      "grad_norm": 2.5621471996893592,
+      "learning_rate": 9.775733759391833e-06,
+      "loss": 1.1379,
+      "step": 2425
     },
     {
       "epoch": 0.5575952271684259,
-      "grad_norm": 1.8649330489030282,
+      "grad_norm": 2.343484566189912,
       "learning_rate": 9.735694837537993e-06,
-      "loss": 0.9572,
-      "step": 1215
+      "loss": 1.1399,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5587425424506655,
+      "grad_norm": 2.612209603264774,
+      "learning_rate": 9.695660155330598e-06,
+      "loss": 1.165,
+      "step": 2435
     },
     {
       "epoch": 0.559889857732905,
-      "grad_norm": 1.7668993960491821,
+      "grad_norm": 2.3356646935305254,
       "learning_rate": 9.655630354954974e-06,
-      "loss": 0.9922,
-      "step": 1220
+      "loss": 1.1964,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5610371730151446,
+      "grad_norm": 2.5695540299579402,
+      "learning_rate": 9.615606078518143e-06,
+      "loss": 1.159,
+      "step": 2445
     },
     {
       "epoch": 0.5621844882973841,
-      "grad_norm": 1.9573923418916903,
+      "grad_norm": 2.3599515222260705,
       "learning_rate": 9.57558796803852e-06,
-      "loss": 1.0109,
-      "step": 1225
+      "loss": 1.1316,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5633318035796236,
+      "grad_norm": 2.2759018372704802,
+      "learning_rate": 9.535576665435606e-06,
+      "loss": 1.1426,
+      "step": 2455
     },
     {
       "epoch": 0.5644791188618632,
-      "grad_norm": 1.711437620336517,
+      "grad_norm": 2.483770091141199,
       "learning_rate": 9.495572812519718e-06,
-      "loss": 0.9946,
-      "step": 1230
+      "loss": 1.1813,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5656264341441029,
+      "grad_norm": 2.2856329803598565,
+      "learning_rate": 9.455577050981648e-06,
+      "loss": 1.1421,
+      "step": 2465
     },
     {
       "epoch": 0.5667737494263424,
-      "grad_norm": 1.7691621774728759,
+      "grad_norm": 2.3740281936243446,
       "learning_rate": 9.41559002238242e-06,
-      "loss": 1.0061,
-      "step": 1235
+      "loss": 1.1607,
+      "step": 2470
+    },
+    {
+      "epoch": 0.567921064708582,
+      "grad_norm": 2.3093025719761475,
+      "learning_rate": 9.375612368142962e-06,
+      "loss": 1.1422,
+      "step": 2475
     },
     {
       "epoch": 0.5690683799908215,
-      "grad_norm": 1.8386891074186076,
+      "grad_norm": 2.360053549045114,
       "learning_rate": 9.33564472953383e-06,
-      "loss": 0.9886,
-      "step": 1240
+      "loss": 1.1494,
+      "step": 2480
+    },
+    {
+      "epoch": 0.570215695273061,
+      "grad_norm": 2.521839899431946,
+      "learning_rate": 9.295687747664935e-06,
+      "loss": 1.1842,
+      "step": 2485
     },
     {
       "epoch": 0.5713630105553006,
-      "grad_norm": 1.8757816038935415,
+      "grad_norm": 2.3908456398757916,
       "learning_rate": 9.255742063475228e-06,
-      "loss": 0.9952,
-      "step": 1245
+      "loss": 1.1461,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5725103258375401,
+      "grad_norm": 2.3888342198222663,
+      "learning_rate": 9.215808317722453e-06,
+      "loss": 1.1216,
+      "step": 2495
     },
     {
       "epoch": 0.5736576411197797,
-      "grad_norm": 1.7204727826291393,
+      "grad_norm": 2.1979418755669724,
       "learning_rate": 9.175887150972841e-06,
-      "loss": 0.9922,
-      "step": 1250
+      "loss": 1.1301,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5748049564020192,
+      "grad_norm": 2.3114444692539076,
+      "learning_rate": 9.135979203590852e-06,
+      "loss": 1.1469,
+      "step": 2505
     },
     {
       "epoch": 0.5759522716842589,
-      "grad_norm": 1.7936118442403097,
+      "grad_norm": 2.410614961263144,
       "learning_rate": 9.096085115728902e-06,
-      "loss": 0.982,
-      "step": 1255
+      "loss": 1.1377,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5770995869664984,
+      "grad_norm": 2.4404399337721054,
+      "learning_rate": 9.056205527317082e-06,
+      "loss": 1.1678,
+      "step": 2515
     },
     {
       "epoch": 0.578246902248738,
-      "grad_norm": 1.6703981680215847,
+      "grad_norm": 2.3903862678162513,
       "learning_rate": 9.016341078052908e-06,
-      "loss": 0.9828,
-      "step": 1260
+      "loss": 1.1566,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5793942175309775,
+      "grad_norm": 2.487113987482187,
+      "learning_rate": 8.976492407391046e-06,
+      "loss": 1.1502,
+      "step": 2525
     },
     {
       "epoch": 0.5805415328132171,
-      "grad_norm": 1.836996443217227,
+      "grad_norm": 2.265788508857309,
       "learning_rate": 8.93666015453307e-06,
-      "loss": 0.9882,
-      "step": 1265
+      "loss": 1.1359,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5816888480954566,
+      "grad_norm": 2.361564602393008,
+      "learning_rate": 8.89684495841719e-06,
+      "loss": 1.1664,
+      "step": 2535
     },
     {
       "epoch": 0.5828361633776962,
-      "grad_norm": 2.001392851330759,
+      "grad_norm": 2.4240744893246156,
       "learning_rate": 8.857047457708023e-06,
-      "loss": 0.9923,
-      "step": 1270
+      "loss": 1.1523,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5839834786599357,
+      "grad_norm": 2.338487255417482,
+      "learning_rate": 8.817268290786343e-06,
+      "loss": 1.1125,
+      "step": 2545
     },
     {
       "epoch": 0.5851307939421753,
-      "grad_norm": 1.779792680703507,
+      "grad_norm": 2.3267291458801727,
       "learning_rate": 8.777508095738818e-06,
-      "loss": 0.9821,
-      "step": 1275
+      "loss": 1.123,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5862781092244149,
+      "grad_norm": 2.3955672934820966,
+      "learning_rate": 8.737767510347816e-06,
+      "loss": 1.1247,
+      "step": 2555
     },
     {
       "epoch": 0.5874254245066545,
-      "grad_norm": 1.7551844873505347,
+      "grad_norm": 2.463831836478037,
       "learning_rate": 8.698047172081129e-06,
-      "loss": 0.9905,
-      "step": 1280
+      "loss": 1.148,
+      "step": 2560
+    },
+    {
+      "epoch": 0.588572739788894,
+      "grad_norm": 2.370560723719375,
+      "learning_rate": 8.658347718081791e-06,
+      "loss": 1.1158,
+      "step": 2565
     },
     {
       "epoch": 0.5897200550711336,
-      "grad_norm": 1.934345593929013,
+      "grad_norm": 2.387926689961274,
       "learning_rate": 8.618669785157825e-06,
-      "loss": 0.9796,
-      "step": 1285
+      "loss": 1.1733,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5908673703533731,
+      "grad_norm": 2.383333743021682,
+      "learning_rate": 8.579014009772045e-06,
+      "loss": 1.1675,
+      "step": 2575
     },
     {
       "epoch": 0.5920146856356127,
-      "grad_norm": 1.8648278829036111,
+      "grad_norm": 2.3291743526646633,
       "learning_rate": 8.539381028031838e-06,
-      "loss": 1.0053,
-      "step": 1290
+      "loss": 1.1252,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5931620009178522,
+      "grad_norm": 2.1671129717844213,
+      "learning_rate": 8.499771475678968e-06,
+      "loss": 1.1349,
+      "step": 2585
     },
     {
       "epoch": 0.5943093162000918,
-      "grad_norm": 1.7236526799565135,
+      "grad_norm": 2.3005770803883885,
       "learning_rate": 8.46018598807938e-06,
-      "loss": 1.0005,
-      "step": 1295
+      "loss": 1.1272,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5954566314823313,
+      "grad_norm": 2.387101372210533,
+      "learning_rate": 8.420625200212985e-06,
+      "loss": 1.1567,
+      "step": 2595
     },
     {
       "epoch": 0.596603946764571,
-      "grad_norm": 1.8020519118969505,
+      "grad_norm": 2.4728108936932736,
       "learning_rate": 8.381089746663517e-06,
-      "loss": 0.9855,
-      "step": 1300
+      "loss": 1.1681,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5977512620468105,
+      "grad_norm": 2.4007708346991357,
+      "learning_rate": 8.341580261608305e-06,
+      "loss": 1.1507,
+      "step": 2605
     },
     {
       "epoch": 0.59889857732905,
-      "grad_norm": 1.8759050818870677,
+      "grad_norm": 2.4457553394851774,
       "learning_rate": 8.302097378808147e-06,
-      "loss": 0.9884,
-      "step": 1305
+      "loss": 1.1337,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6000458926112896,
+      "grad_norm": 2.315516541044818,
+      "learning_rate": 8.262641731597097e-06,
+      "loss": 1.1388,
+      "step": 2615
     },
     {
       "epoch": 0.6011932078935291,
-      "grad_norm": 1.7101070470722577,
+      "grad_norm": 2.479154641664564,
       "learning_rate": 8.223213952872353e-06,
-      "loss": 0.9939,
-      "step": 1310
+      "loss": 1.1541,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6023405231757687,
+      "grad_norm": 2.352645397604372,
+      "learning_rate": 8.183814675084074e-06,
+      "loss": 1.1265,
+      "step": 2625
     },
     {
       "epoch": 0.6034878384580082,
-      "grad_norm": 1.8363414696856302,
+      "grad_norm": 2.3289347257902775,
       "learning_rate": 8.144444530225237e-06,
-      "loss": 1.006,
-      "step": 1315
+      "loss": 1.1359,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6046351537402478,
+      "grad_norm": 2.4536927501335444,
+      "learning_rate": 8.105104149821515e-06,
+      "loss": 1.1372,
+      "step": 2635
     },
     {
       "epoch": 0.6057824690224873,
-      "grad_norm": 1.8189880317647966,
+      "grad_norm": 2.254438191832268,
       "learning_rate": 8.065794164921128e-06,
-      "loss": 1.0103,
-      "step": 1320
+      "loss": 1.1621,
+      "step": 2640
+    },
+    {
+      "epoch": 0.606929784304727,
+      "grad_norm": 2.385368404798516,
+      "learning_rate": 8.026515206084744e-06,
+      "loss": 1.1321,
+      "step": 2645
     },
     {
       "epoch": 0.6080770995869665,
-      "grad_norm": 1.7184329357048536,
+      "grad_norm": 2.3736422295023494,
       "learning_rate": 7.987267903375331e-06,
-      "loss": 0.9802,
-      "step": 1325
+      "loss": 1.14,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6092244148692061,
+      "grad_norm": 2.3859764598647235,
+      "learning_rate": 7.948052886348091e-06,
+      "loss": 1.1566,
+      "step": 2655
     },
     {
       "epoch": 0.6103717301514456,
-      "grad_norm": 1.6760569997350725,
+      "grad_norm": 2.3273871754564923,
       "learning_rate": 7.90887078404033e-06,
-      "loss": 0.9864,
-      "step": 1330
+      "loss": 1.1258,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6115190454336852,
+      "grad_norm": 2.3442824667455096,
+      "learning_rate": 7.869722224961372e-06,
+      "loss": 1.1323,
+      "step": 2665
     },
     {
       "epoch": 0.6126663607159247,
-      "grad_norm": 1.759298284823385,
+      "grad_norm": 2.2232728108361033,
       "learning_rate": 7.830607837082494e-06,
-      "loss": 0.979,
-      "step": 1335
+      "loss": 1.1503,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6138136759981643,
+      "grad_norm": 2.265021484521007,
+      "learning_rate": 7.791528247826832e-06,
+      "loss": 1.112,
+      "step": 2675
     },
     {
       "epoch": 0.6149609912804038,
-      "grad_norm": 1.693614807322877,
+      "grad_norm": 2.4203498234905694,
       "learning_rate": 7.75248408405934e-06,
-      "loss": 0.9865,
-      "step": 1340
+      "loss": 1.1267,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6161083065626434,
+      "grad_norm": 2.3786029930756527,
+      "learning_rate": 7.71347597207671e-06,
+      "loss": 1.1377,
+      "step": 2685
     },
     {
       "epoch": 0.617255621844883,
-      "grad_norm": 1.7256883752598309,
+      "grad_norm": 2.3033389080041706,
       "learning_rate": 7.674504537597336e-06,
-      "loss": 0.9883,
-      "step": 1345
+      "loss": 1.12,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6184029371271226,
+      "grad_norm": 2.349456539458806,
+      "learning_rate": 7.635570405751297e-06,
+      "loss": 1.112,
+      "step": 2695
     },
     {
       "epoch": 0.6195502524093621,
-      "grad_norm": 1.7555572901812775,
+      "grad_norm": 2.547343077883558,
       "learning_rate": 7.596674201070282e-06,
-      "loss": 0.9953,
-      "step": 1350
+      "loss": 1.134,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6206975676916017,
+      "grad_norm": 2.3588027590426344,
+      "learning_rate": 7.557816547477627e-06,
+      "loss": 1.1496,
+      "step": 2705
     },
     {
       "epoch": 0.6218448829738412,
-      "grad_norm": 1.7549420037924819,
+      "grad_norm": 2.4201812828437794,
       "learning_rate": 7.518998068278266e-06,
-      "loss": 0.9761,
-      "step": 1355
+      "loss": 1.1345,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6229921982560808,
+      "grad_norm": 2.497692339906205,
+      "learning_rate": 7.480219386148751e-06,
+      "loss": 1.144,
+      "step": 2715
     },
     {
       "epoch": 0.6241395135383203,
-      "grad_norm": 1.7638758955771752,
+      "grad_norm": 2.326917172406579,
       "learning_rate": 7.441481123127257e-06,
-      "loss": 0.9663,
-      "step": 1360
+      "loss": 1.1416,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6252868288205599,
+      "grad_norm": 2.3514950300039628,
+      "learning_rate": 7.402783900603612e-06,
+      "loss": 1.1094,
+      "step": 2725
     },
     {
       "epoch": 0.6264341441027994,
-      "grad_norm": 1.9316469572623007,
+      "grad_norm": 2.4507528851663283,
       "learning_rate": 7.364128339309326e-06,
-      "loss": 0.9912,
-      "step": 1365
+      "loss": 1.1325,
+      "step": 2730
+    },
+    {
+      "epoch": 0.627581459385039,
+      "grad_norm": 2.3290237424744418,
+      "learning_rate": 7.325515059307622e-06,
+      "loss": 1.1292,
+      "step": 2735
     },
     {
       "epoch": 0.6287287746672786,
-      "grad_norm": 1.7094849210033014,
+      "grad_norm": 2.3160967005917077,
       "learning_rate": 7.286944679983521e-06,
-      "loss": 0.9759,
-      "step": 1370
+      "loss": 1.1128,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6298760899495182,
+      "grad_norm": 2.26849447367049,
+      "learning_rate": 7.248417820033857e-06,
+      "loss": 1.1198,
+      "step": 2745
     },
     {
       "epoch": 0.6310234052317577,
-      "grad_norm": 1.6887590729536133,
+      "grad_norm": 2.3127943494175187,
       "learning_rate": 7.209935097457412e-06,
-      "loss": 0.9899,
-      "step": 1375
+      "loss": 1.1356,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6321707205139973,
+      "grad_norm": 2.2785842730855586,
+      "learning_rate": 7.171497129544946e-06,
+      "loss": 1.1264,
+      "step": 2755
     },
     {
       "epoch": 0.6333180357962368,
-      "grad_norm": 1.7727590940094575,
+      "grad_norm": 2.2682813135513937,
       "learning_rate": 7.133104532869342e-06,
-      "loss": 1.0,
-      "step": 1380
+      "loss": 1.1096,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6344653510784763,
+      "grad_norm": 2.331975147659969,
+      "learning_rate": 7.094757923275688e-06,
+      "loss": 1.1119,
+      "step": 2765
     },
     {
       "epoch": 0.6356126663607159,
-      "grad_norm": 1.8172481765220854,
+      "grad_norm": 2.4071880692345116,
       "learning_rate": 7.056457915871399e-06,
-      "loss": 0.9663,
-      "step": 1385
+      "loss": 1.109,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6367599816429554,
+      "grad_norm": 2.3292160849189902,
+      "learning_rate": 7.018205125016369e-06,
+      "loss": 1.1561,
+      "step": 2775
     },
     {
       "epoch": 0.637907296925195,
-      "grad_norm": 1.7974312188733144,
+      "grad_norm": 2.347302425036466,
       "learning_rate": 6.980000164313093e-06,
-      "loss": 0.9586,
-      "step": 1390
+      "loss": 1.1349,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6390546122074346,
+      "grad_norm": 2.4185665904201694,
+      "learning_rate": 6.9418436465968485e-06,
+      "loss": 1.1423,
+      "step": 2785
     },
     {
       "epoch": 0.6402019274896742,
-      "grad_norm": 1.7508328675764726,
+      "grad_norm": 2.462052073994329,
       "learning_rate": 6.903736183925835e-06,
-      "loss": 0.9719,
-      "step": 1395
+      "loss": 1.1228,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6413492427719137,
+      "grad_norm": 2.464606580026005,
+      "learning_rate": 6.865678387571394e-06,
+      "loss": 1.1109,
+      "step": 2795
     },
     {
       "epoch": 0.6424965580541533,
-      "grad_norm": 1.7326843284737148,
+      "grad_norm": 2.3960490934288226,
       "learning_rate": 6.82767086800817e-06,
-      "loss": 0.9961,
-      "step": 1400
+      "loss": 1.1206,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6436438733363928,
+      "grad_norm": 2.307847274056824,
+      "learning_rate": 6.789714234904332e-06,
+      "loss": 1.1223,
+      "step": 2805
     },
     {
       "epoch": 0.6447911886186324,
-      "grad_norm": 1.7352259185492422,
+      "grad_norm": 2.378007781793691,
       "learning_rate": 6.751809097111799e-06,
-      "loss": 0.997,
-      "step": 1405
+      "loss": 1.1254,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6459385039008719,
+      "grad_norm": 2.412595548847994,
+      "learning_rate": 6.71395606265646e-06,
+      "loss": 1.1342,
+      "step": 2815
     },
     {
       "epoch": 0.6470858191831115,
-      "grad_norm": 1.8933368018227825,
+      "grad_norm": 2.376108073746517,
       "learning_rate": 6.676155738728438e-06,
-      "loss": 0.9779,
-      "step": 1410
+      "loss": 1.1388,
+      "step": 2820
+    },
+    {
+      "epoch": 0.648233134465351,
+      "grad_norm": 2.368409721798146,
+      "learning_rate": 6.638408731672332e-06,
+      "loss": 1.1375,
+      "step": 2825
     },
     {
       "epoch": 0.6493804497475907,
-      "grad_norm": 1.6974498786941927,
+      "grad_norm": 2.338962451200504,
       "learning_rate": 6.600715646977503e-06,
-      "loss": 0.9799,
-      "step": 1415
+      "loss": 1.1278,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6505277650298302,
+      "grad_norm": 2.4197757460285767,
+      "learning_rate": 6.5630770892683656e-06,
+      "loss": 1.1146,
+      "step": 2835
     },
     {
       "epoch": 0.6516750803120698,
-      "grad_norm": 1.7806592546745614,
+      "grad_norm": 2.456285922205978,
       "learning_rate": 6.525493662294669e-06,
-      "loss": 0.9718,
-      "step": 1420
+      "loss": 1.1124,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6528223955943093,
+      "grad_norm": 2.377267536143265,
+      "learning_rate": 6.487965968921834e-06,
+      "loss": 1.1435,
+      "step": 2845
     },
     {
       "epoch": 0.6539697108765489,
-      "grad_norm": 1.7658701508098122,
+      "grad_norm": 2.282462616393301,
       "learning_rate": 6.450494611121274e-06,
-      "loss": 0.9852,
-      "step": 1425
+      "loss": 1.1134,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6551170261587884,
+      "grad_norm": 2.364080280114686,
+      "learning_rate": 6.413080189960734e-06,
+      "loss": 1.1174,
+      "step": 2855
     },
     {
       "epoch": 0.656264341441028,
-      "grad_norm": 1.721774620040715,
+      "grad_norm": 3.2795783117457513,
       "learning_rate": 6.375723305594658e-06,
-      "loss": 0.9766,
-      "step": 1430
+      "loss": 1.1226,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6574116567232675,
+      "grad_norm": 2.381603120461069,
+      "learning_rate": 6.338424557254556e-06,
+      "loss": 1.1112,
+      "step": 2865
     },
     {
       "epoch": 0.6585589720055071,
-      "grad_norm": 1.8456782805007963,
+      "grad_norm": 2.471915466738649,
       "learning_rate": 6.301184543239398e-06,
-      "loss": 0.9604,
-      "step": 1435
+      "loss": 1.1375,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6597062872877467,
+      "grad_norm": 2.668918836628048,
+      "learning_rate": 6.264003860906003e-06,
+      "loss": 1.0987,
+      "step": 2875
     },
     {
       "epoch": 0.6608536025699863,
-      "grad_norm": 1.6517747886851708,
+      "grad_norm": 2.415869244791648,
       "learning_rate": 6.2268831066594846e-06,
-      "loss": 0.9978,
-      "step": 1440
+      "loss": 1.0971,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6620009178522258,
+      "grad_norm": 2.4645346416387524,
+      "learning_rate": 6.189822875943644e-06,
+      "loss": 1.1365,
+      "step": 2885
     },
     {
       "epoch": 0.6631482331344654,
-      "grad_norm": 1.7868240633431423,
+      "grad_norm": 2.3243005208081025,
       "learning_rate": 6.152823763231463e-06,
-      "loss": 0.9659,
-      "step": 1445
+      "loss": 1.1058,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6642955484167049,
+      "grad_norm": 2.2736888091121026,
+      "learning_rate": 6.115886362015525e-06,
+      "loss": 1.1106,
+      "step": 2895
     },
     {
       "epoch": 0.6654428636989445,
-      "grad_norm": 1.6864295399871028,
+      "grad_norm": 2.3904086910487616,
       "learning_rate": 6.079011264798534e-06,
-      "loss": 0.97,
-      "step": 1450
+      "loss": 1.0944,
+      "step": 2900
+    },
+    {
+      "epoch": 0.666590178981184,
+      "grad_norm": 2.427732424122635,
+      "learning_rate": 6.042199063083787e-06,
+      "loss": 1.1154,
+      "step": 2905
     },
     {
       "epoch": 0.6677374942634235,
-      "grad_norm": 1.7298790044434966,
+      "grad_norm": 2.238637944143973,
       "learning_rate": 6.005450347365687e-06,
-      "loss": 0.9807,
-      "step": 1455
+      "loss": 1.1079,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6688848095456631,
+      "grad_norm": 2.491540791091301,
+      "learning_rate": 5.96876570712028e-06,
+      "loss": 1.1369,
+      "step": 2915
     },
     {
       "epoch": 0.6700321248279028,
-      "grad_norm": 1.7628376564735946,
+      "grad_norm": 2.388157401107027,
       "learning_rate": 5.932145730795793e-06,
-      "loss": 0.9686,
-      "step": 1460
+      "loss": 1.0972,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6711794401101423,
+      "grad_norm": 2.281687888087642,
+      "learning_rate": 5.895591005803198e-06,
+      "loss": 1.0911,
+      "step": 2925
     },
     {
       "epoch": 0.6723267553923818,
-      "grad_norm": 1.7343262868098874,
+      "grad_norm": 2.435185289282404,
       "learning_rate": 5.859102118506787e-06,
-      "loss": 0.9524,
-      "step": 1465
+      "loss": 1.1201,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6734740706746214,
+      "grad_norm": 2.4699883089278156,
+      "learning_rate": 5.822679654214771e-06,
+      "loss": 1.0911,
+      "step": 2935
     },
     {
       "epoch": 0.6746213859568609,
-      "grad_norm": 1.9084160680148732,
+      "grad_norm": 2.2306849746847264,
       "learning_rate": 5.786324197169887e-06,
-      "loss": 0.9818,
-      "step": 1470
+      "loss": 1.12,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6757687012391005,
+      "grad_norm": 2.3632510883273037,
+      "learning_rate": 5.7500363305400185e-06,
+      "loss": 1.1199,
+      "step": 2945
     },
     {
       "epoch": 0.67691601652134,
-      "grad_norm": 1.735061984384823,
+      "grad_norm": 2.2743440540898257,
       "learning_rate": 5.713816636408871e-06,
-      "loss": 0.9673,
-      "step": 1475
+      "loss": 1.1177,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6780633318035796,
+      "grad_norm": 2.471910084539622,
+      "learning_rate": 5.677665695766581e-06,
+      "loss": 1.1034,
+      "step": 2955
     },
     {
       "epoch": 0.6792106470858191,
-      "grad_norm": 1.7982234436842786,
+      "grad_norm": 2.2906550476388934,
       "learning_rate": 5.641584088500461e-06,
-      "loss": 0.9694,
-      "step": 1480
+      "loss": 1.1142,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6803579623680588,
+      "grad_norm": 2.348585353757232,
+      "learning_rate": 5.605572393385645e-06,
+      "loss": 1.1398,
+      "step": 2965
     },
     {
       "epoch": 0.6815052776502983,
-      "grad_norm": 1.7104886881206862,
+      "grad_norm": 2.300443710441418,
       "learning_rate": 5.569631188075842e-06,
-      "loss": 0.9638,
-      "step": 1485
+      "loss": 1.0944,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6826525929325379,
+      "grad_norm": 2.436421664434051,
+      "learning_rate": 5.5337610490940375e-06,
+      "loss": 1.1178,
+      "step": 2975
     },
     {
       "epoch": 0.6837999082147774,
-      "grad_norm": 1.7069667971132039,
+      "grad_norm": 2.3576525615606716,
       "learning_rate": 5.497962551823266e-06,
-      "loss": 0.9746,
-      "step": 1490
+      "loss": 1.1096,
+      "step": 2980
+    },
+    {
+      "epoch": 0.684947223497017,
+      "grad_norm": 2.3782230649051437,
+      "learning_rate": 5.46223627049739e-06,
+      "loss": 1.1149,
+      "step": 2985
     },
     {
       "epoch": 0.6860945387792565,
-      "grad_norm": 2.380253232983682,
+      "grad_norm": 2.308682218164475,
       "learning_rate": 5.426582778191858e-06,
-      "loss": 0.9898,
-      "step": 1495
+      "loss": 1.0993,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6872418540614961,
+      "grad_norm": 2.357568661419494,
+      "learning_rate": 5.3910026468145384e-06,
+      "loss": 1.0961,
+      "step": 2995
     },
     {
       "epoch": 0.6883891693437356,
-      "grad_norm": 1.789361116223533,
+      "grad_norm": 2.444899732941997,
       "learning_rate": 5.355496447096533e-06,
-      "loss": 0.9813,
-      "step": 1500
+      "loss": 1.1036,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6895364846259752,
+      "grad_norm": 2.3977195078454443,
+      "learning_rate": 5.320064748583031e-06,
+      "loss": 1.1348,
+      "step": 3005
     },
     {
       "epoch": 0.6906837999082148,
-      "grad_norm": 1.7257448134853217,
+      "grad_norm": 2.462475456258121,
       "learning_rate": 5.284708119624173e-06,
-      "loss": 0.9897,
-      "step": 1505
+      "loss": 1.1045,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6918311151904544,
+      "grad_norm": 2.4248136095007564,
+      "learning_rate": 5.249427127365918e-06,
+      "loss": 1.1213,
+      "step": 3015
     },
     {
       "epoch": 0.6929784304726939,
-      "grad_norm": 1.695066389007114,
+      "grad_norm": 2.4415766775762506,
       "learning_rate": 5.2142223377409616e-06,
-      "loss": 0.9662,
-      "step": 1510
+      "loss": 1.0923,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6941257457549335,
+      "grad_norm": 2.2409511706430645,
+      "learning_rate": 5.179094315459652e-06,
+      "loss": 1.1032,
+      "step": 3025
     },
     {
       "epoch": 0.695273061037173,
-      "grad_norm": 1.8265038779545326,
+      "grad_norm": 2.4777655292781247,
       "learning_rate": 5.144043624000944e-06,
-      "loss": 0.9221,
-      "step": 1515
+      "loss": 1.0872,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6964203763194126,
+      "grad_norm": 2.341568313842261,
+      "learning_rate": 5.109070825603338e-06,
+      "loss": 1.1318,
+      "step": 3035
     },
     {
       "epoch": 0.6975676916016521,
-      "grad_norm": 1.682425505945564,
+      "grad_norm": 2.3725651477887397,
       "learning_rate": 5.074176481255873e-06,
-      "loss": 0.9665,
-      "step": 1520
+      "loss": 1.0942,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6987150068838917,
+      "grad_norm": 2.523833216091205,
+      "learning_rate": 5.039361150689141e-06,
+      "loss": 1.0842,
+      "step": 3045
     },
     {
       "epoch": 0.6998623221661312,
-      "grad_norm": 1.6947722881421843,
+      "grad_norm": 2.2418957167374964,
       "learning_rate": 5.00462539236628e-06,
-      "loss": 0.9735,
-      "step": 1525
+      "loss": 1.1216,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7010096374483709,
+      "grad_norm": 2.8693298084928625,
+      "learning_rate": 4.969969763474047e-06,
+      "loss": 1.1108,
+      "step": 3055
     },
     {
       "epoch": 0.7021569527306104,
-      "grad_norm": 2.0862442184881926,
+      "grad_norm": 2.2823674364430167,
       "learning_rate": 4.935394819913849e-06,
-      "loss": 0.9462,
-      "step": 1530
+      "loss": 1.1102,
+      "step": 3060
+    },
+    {
+      "epoch": 0.70330426801285,
+      "grad_norm": 2.470986373278573,
+      "learning_rate": 4.900901116292854e-06,
+      "loss": 1.1494,
+      "step": 3065
     },
     {
       "epoch": 0.7044515832950895,
-      "grad_norm": 1.8473532390280518,
+      "grad_norm": 2.395023784542252,
       "learning_rate": 4.866489205915072e-06,
-      "loss": 0.9699,
-      "step": 1535
+      "loss": 1.082,
+      "step": 3070
+    },
+    {
+      "epoch": 0.705598898577329,
+      "grad_norm": 2.291309312067593,
+      "learning_rate": 4.8321596407725044e-06,
+      "loss": 1.1017,
+      "step": 3075
     },
     {
       "epoch": 0.7067462138595686,
-      "grad_norm": 1.748002794088265,
+      "grad_norm": 2.2843719962521223,
       "learning_rate": 4.7979129715362625e-06,
-      "loss": 1.0002,
-      "step": 1540
+      "loss": 1.1089,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7078935291418081,
+      "grad_norm": 2.2151523970617504,
+      "learning_rate": 4.7637497475477465e-06,
+      "loss": 1.0997,
+      "step": 3085
     },
     {
       "epoch": 0.7090408444240477,
-      "grad_norm": 1.70600748304509,
+      "grad_norm": 2.402700465949013,
       "learning_rate": 4.72967051680985e-06,
-      "loss": 0.9638,
-      "step": 1545
+      "loss": 1.1124,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7101881597062872,
+      "grad_norm": 2.4442267015771466,
+      "learning_rate": 4.695675825978133e-06,
+      "loss": 1.0944,
+      "step": 3095
     },
     {
       "epoch": 0.7113354749885269,
-      "grad_norm": 1.6249584663789558,
+      "grad_norm": 2.388630826645324,
       "learning_rate": 4.661766220352098e-06,
-      "loss": 0.9414,
-      "step": 1550
+      "loss": 1.0959,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7124827902707664,
+      "grad_norm": 2.309795253109109,
+      "learning_rate": 4.627942243866387e-06,
+      "loss": 1.1013,
+      "step": 3105
     },
     {
       "epoch": 0.713630105553006,
-      "grad_norm": 1.80174378572434,
+      "grad_norm": 2.467492297537252,
       "learning_rate": 4.594204439082122e-06,
-      "loss": 0.9727,
-      "step": 1555
+      "loss": 1.0966,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7147774208352455,
+      "grad_norm": 2.4257152563230058,
+      "learning_rate": 4.560553347178144e-06,
+      "loss": 1.0993,
+      "step": 3115
     },
     {
       "epoch": 0.7159247361174851,
-      "grad_norm": 1.8374638729274564,
+      "grad_norm": 2.481936943192826,
       "learning_rate": 4.526989507942374e-06,
-      "loss": 0.9565,
-      "step": 1560
+      "loss": 1.0981,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7170720513997246,
+      "grad_norm": 2.34807669335114,
+      "learning_rate": 4.493513459763126e-06,
+      "loss": 1.0751,
+      "step": 3125
     },
     {
       "epoch": 0.7182193666819642,
-      "grad_norm": 1.6726191265269426,
+      "grad_norm": 2.3700833632481193,
       "learning_rate": 4.460125739620479e-06,
-      "loss": 0.9487,
-      "step": 1565
+      "loss": 1.098,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7193666819642037,
+      "grad_norm": 2.438776471298164,
+      "learning_rate": 4.426826883077681e-06,
+      "loss": 1.1084,
+      "step": 3135
     },
     {
       "epoch": 0.7205139972464433,
-      "grad_norm": 1.8264345398924386,
+      "grad_norm": 2.272516466827752,
       "learning_rate": 4.393617424272527e-06,
-      "loss": 0.9636,
-      "step": 1570
+      "loss": 1.0798,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7216613125286829,
+      "grad_norm": 2.3356663935933293,
+      "learning_rate": 4.360497895908826e-06,
+      "loss": 1.0999,
+      "step": 3145
     },
     {
       "epoch": 0.7228086278109225,
-      "grad_norm": 1.6887334363300377,
+      "grad_norm": 2.3180589970845604,
       "learning_rate": 4.3274688292478105e-06,
-      "loss": 0.961,
-      "step": 1575
+      "loss": 1.1116,
+      "step": 3150
+    },
+    {
+      "epoch": 0.723955943093162,
+      "grad_norm": 2.2901324964266174,
+      "learning_rate": 4.294530754099666e-06,
+      "loss": 1.1108,
+      "step": 3155
     },
     {
       "epoch": 0.7251032583754016,
-      "grad_norm": 1.8470657094730811,
+      "grad_norm": 2.3563145658091993,
       "learning_rate": 4.261684198815004e-06,
-      "loss": 0.9694,
-      "step": 1580
+      "loss": 1.0863,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7262505736576411,
+      "grad_norm": 2.309763606122749,
+      "learning_rate": 4.228929690276381e-06,
+      "loss": 1.0647,
+      "step": 3165
     },
     {
       "epoch": 0.7273978889398807,
-      "grad_norm": 1.7132408618736092,
+      "grad_norm": 2.2777260232984675,
       "learning_rate": 4.196267753889864e-06,
-      "loss": 0.9581,
-      "step": 1585
+      "loss": 1.0915,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7285452042221202,
+      "grad_norm": 2.34894914988906,
+      "learning_rate": 4.163698913576592e-06,
+      "loss": 1.0917,
+      "step": 3175
     },
     {
       "epoch": 0.7296925195043598,
-      "grad_norm": 1.6900030303741553,
+      "grad_norm": 2.289533046571488,
       "learning_rate": 4.131223691764384e-06,
-      "loss": 0.9325,
-      "step": 1590
+      "loss": 1.0791,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7308398347865993,
+      "grad_norm": 2.351419451457308,
+      "learning_rate": 4.098842609379339e-06,
+      "loss": 1.1045,
+      "step": 3185
     },
     {
       "epoch": 0.731987150068839,
-      "grad_norm": 1.7522033167073814,
+      "grad_norm": 2.318863793299261,
       "learning_rate": 4.066556185837494e-06,
-      "loss": 0.9347,
-      "step": 1595
+      "loss": 1.1015,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7331344653510785,
+      "grad_norm": 2.3923461411768847,
+      "learning_rate": 4.0343649390365e-06,
+      "loss": 1.0958,
+      "step": 3195
     },
     {
       "epoch": 0.7342817806333181,
-      "grad_norm": 1.755668968211341,
+      "grad_norm": 2.3929957610837342,
       "learning_rate": 4.002269385347289e-06,
-      "loss": 0.9256,
-      "step": 1600
+      "loss": 1.1092,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7354290959155576,
+      "grad_norm": 2.4257111429359566,
+      "learning_rate": 3.970270039605818e-06,
+      "loss": 1.0883,
+      "step": 3205
     },
     {
       "epoch": 0.7365764111977972,
-      "grad_norm": 1.713678331254988,
+      "grad_norm": 2.2636635937684217,
       "learning_rate": 3.9383674151047936e-06,
-      "loss": 0.975,
-      "step": 1605
+      "loss": 1.0814,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7377237264800367,
+      "grad_norm": 2.382278040921493,
+      "learning_rate": 3.906562023585442e-06,
+      "loss": 1.0917,
+      "step": 3215
     },
     {
       "epoch": 0.7388710417622762,
-      "grad_norm": 1.6865904408801047,
+      "grad_norm": 2.5011209640625407,
       "learning_rate": 3.8748543752293e-06,
-      "loss": 0.9729,
-      "step": 1610
+      "loss": 1.0927,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7400183570445158,
+      "grad_norm": 2.4421075205952527,
+      "learning_rate": 3.843244978650045e-06,
+      "loss": 1.1204,
+      "step": 3225
     },
     {
       "epoch": 0.7411656723267553,
-      "grad_norm": 1.6694830205166722,
+      "grad_norm": 2.298007752819784,
       "learning_rate": 3.8117343408853124e-06,
-      "loss": 0.9554,
-      "step": 1615
+      "loss": 1.0542,
+      "step": 3230
+    },
+    {
+      "epoch": 0.742312987608995,
+      "grad_norm": 2.3328586721249733,
+      "learning_rate": 3.780322967388577e-06,
+      "loss": 1.0953,
+      "step": 3235
     },
     {
       "epoch": 0.7434603028912345,
-      "grad_norm": 1.6703972771704265,
+      "grad_norm": 2.2089269310188726,
       "learning_rate": 3.7490113620210487e-06,
-      "loss": 0.9297,
-      "step": 1620
+      "loss": 1.1072,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7446076181734741,
+      "grad_norm": 2.4006844079537206,
+      "learning_rate": 3.7178000270435765e-06,
+      "loss": 1.1153,
+      "step": 3245
     },
     {
       "epoch": 0.7457549334557136,
-      "grad_norm": 1.7865359308285997,
+      "grad_norm": 2.3450482883357897,
       "learning_rate": 3.686689463108608e-06,
-      "loss": 0.952,
-      "step": 1625
+      "loss": 1.0809,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7469022487379532,
+      "grad_norm": 2.3120419044499614,
+      "learning_rate": 3.6556801692521426e-06,
+      "loss": 1.1197,
+      "step": 3255
     },
     {
       "epoch": 0.7480495640201927,
-      "grad_norm": 1.7448768340578766,
+      "grad_norm": 2.24362750024479,
       "learning_rate": 3.6247726428857344e-06,
-      "loss": 0.9286,
-      "step": 1630
+      "loss": 1.0798,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7491968793024323,
+      "grad_norm": 2.2810362806921476,
+      "learning_rate": 3.593967379788522e-06,
+      "loss": 1.1028,
+      "step": 3265
     },
     {
       "epoch": 0.7503441945846718,
-      "grad_norm": 1.7530030106253895,
+      "grad_norm": 2.2340180118435122,
       "learning_rate": 3.563264874099258e-06,
-      "loss": 0.9524,
-      "step": 1635
+      "loss": 1.1028,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7514915098669114,
+      "grad_norm": 2.3620284704853787,
+      "learning_rate": 3.532665618308395e-06,
+      "loss": 1.0889,
+      "step": 3275
     },
     {
       "epoch": 0.752638825149151,
-      "grad_norm": 1.653777443656037,
+      "grad_norm": 2.4078730171290834,
       "learning_rate": 3.5021701032501777e-06,
-      "loss": 0.9218,
-      "step": 1640
+      "loss": 1.0841,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7537861404313906,
+      "grad_norm": 2.4065328343083214,
+      "learning_rate": 3.4717788180947855e-06,
+      "loss": 1.0774,
+      "step": 3285
     },
     {
       "epoch": 0.7549334557136301,
-      "grad_norm": 3.331790675502453,
+      "grad_norm": 2.2828217399348074,
       "learning_rate": 3.441492250340461e-06,
-      "loss": 0.9453,
-      "step": 1645
+      "loss": 1.0648,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7560807709958697,
+      "grad_norm": 2.390793459374678,
+      "learning_rate": 3.4113108858057175e-06,
+      "loss": 1.0954,
+      "step": 3295
     },
     {
       "epoch": 0.7572280862781092,
-      "grad_norm": 1.7244365474975045,
+      "grad_norm": 2.322350920799196,
       "learning_rate": 3.3812352086215216e-06,
-      "loss": 0.9374,
-      "step": 1650
+      "loss": 1.079,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7583754015603488,
+      "grad_norm": 2.357540821713843,
+      "learning_rate": 3.3512657012235396e-06,
+      "loss": 1.1212,
+      "step": 3305
     },
     {
       "epoch": 0.7595227168425883,
-      "grad_norm": 1.766304020059291,
+      "grad_norm": 2.278921399044481,
       "learning_rate": 3.3214028443444034e-06,
-      "loss": 0.9394,
-      "step": 1655
+      "loss": 1.0907,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7606700321248279,
+      "grad_norm": 2.43748599532908,
+      "learning_rate": 3.2916471170059895e-06,
+      "loss": 1.1205,
+      "step": 3315
     },
     {
       "epoch": 0.7618173474070674,
-      "grad_norm": 1.7263700470016823,
+      "grad_norm": 2.4397124395924314,
       "learning_rate": 3.261998996511736e-06,
-      "loss": 0.9371,
-      "step": 1660
+      "loss": 1.1171,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7629646626893071,
+      "grad_norm": 2.334365795517474,
+      "learning_rate": 3.232458958438992e-06,
+      "loss": 1.0866,
+      "step": 3325
     },
     {
       "epoch": 0.7641119779715466,
-      "grad_norm": 1.7511251400040453,
+      "grad_norm": 2.285744606391934,
       "learning_rate": 3.203027476631386e-06,
-      "loss": 0.9535,
-      "step": 1665
+      "loss": 1.075,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7652592932537862,
+      "grad_norm": 2.221051450915321,
+      "learning_rate": 3.1737050231912324e-06,
+      "loss": 1.0806,
+      "step": 3335
     },
     {
       "epoch": 0.7664066085360257,
-      "grad_norm": 1.7194345747840845,
+      "grad_norm": 2.4305076526822655,
       "learning_rate": 3.1444920684719394e-06,
-      "loss": 0.9399,
-      "step": 1670
+      "loss": 1.0797,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7675539238182653,
+      "grad_norm": 2.294796028036383,
+      "learning_rate": 3.115389081070481e-06,
+      "loss": 1.0972,
+      "step": 3345
     },
     {
       "epoch": 0.7687012391005048,
-      "grad_norm": 1.7100670269940494,
+      "grad_norm": 2.4013023682687593,
       "learning_rate": 3.086396527819876e-06,
-      "loss": 0.9311,
-      "step": 1675
+      "loss": 1.0907,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7698485543827444,
+      "grad_norm": 2.380235464496965,
+      "learning_rate": 3.057514873781703e-06,
+      "loss": 1.0727,
+      "step": 3355
     },
     {
       "epoch": 0.7709958696649839,
-      "grad_norm": 1.7008517044769462,
+      "grad_norm": 2.3192527318279113,
       "learning_rate": 3.028744582238633e-06,
-      "loss": 0.9418,
-      "step": 1680
+      "loss": 1.0798,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7721431849472234,
+      "grad_norm": 2.4421045687535194,
+      "learning_rate": 3.0000861146869963e-06,
+      "loss": 1.1253,
+      "step": 3365
     },
     {
       "epoch": 0.7732905002294631,
-      "grad_norm": 1.7289960064005703,
+      "grad_norm": 2.299556073481683,
       "learning_rate": 2.9715399308294003e-06,
-      "loss": 0.9748,
-      "step": 1685
+      "loss": 1.0641,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7744378155117027,
+      "grad_norm": 2.39327873552009,
+      "learning_rate": 2.9431064885673245e-06,
+      "loss": 1.0821,
+      "step": 3375
     },
     {
       "epoch": 0.7755851307939422,
-      "grad_norm": 1.7377650207700148,
+      "grad_norm": 2.5227837528194694,
       "learning_rate": 2.914786243993808e-06,
-      "loss": 0.936,
-      "step": 1690
+      "loss": 1.0753,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7767324460761817,
+      "grad_norm": 2.313573533828395,
+      "learning_rate": 2.8865796513860933e-06,
+      "loss": 1.1004,
+      "step": 3385
     },
     {
       "epoch": 0.7778797613584213,
-      "grad_norm": 1.8858822369634092,
+      "grad_norm": 2.3172251611325922,
       "learning_rate": 2.858487163198389e-06,
-      "loss": 0.9358,
-      "step": 1695
+      "loss": 1.069,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7790270766406608,
+      "grad_norm": 2.4697157072807534,
+      "learning_rate": 2.8305092300545668e-06,
+      "loss": 1.0763,
+      "step": 3395
     },
     {
       "epoch": 0.7801743919229004,
-      "grad_norm": 1.7588532610850607,
+      "grad_norm": 2.2944498736872,
       "learning_rate": 2.8026463007409665e-06,
-      "loss": 0.9399,
-      "step": 1700
+      "loss": 1.1025,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7813217072051399,
+      "grad_norm": 2.40801000676571,
+      "learning_rate": 2.7748988221991722e-06,
+      "loss": 1.0889,
+      "step": 3405
     },
     {
       "epoch": 0.7824690224873795,
-      "grad_norm": 1.8795956958488933,
+      "grad_norm": 2.3225590835236005,
       "learning_rate": 2.747267239518857e-06,
-      "loss": 0.9642,
-      "step": 1705
+      "loss": 1.1041,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7836163377696191,
+      "grad_norm": 2.3739143205259725,
+      "learning_rate": 2.719751995930645e-06,
+      "loss": 1.0998,
+      "step": 3415
     },
     {
       "epoch": 0.7847636530518587,
-      "grad_norm": 1.733996923587813,
+      "grad_norm": 2.1994328046650975,
       "learning_rate": 2.6923535327989925e-06,
-      "loss": 0.9447,
-      "step": 1710
+      "loss": 1.0643,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7859109683340982,
+      "grad_norm": 2.4392403454609513,
+      "learning_rate": 2.6650722896151126e-06,
+      "loss": 1.0657,
+      "step": 3425
     },
     {
       "epoch": 0.7870582836163378,
-      "grad_norm": 1.7231127716770163,
+      "grad_norm": 2.243421023055753,
       "learning_rate": 2.637908703989924e-06,
-      "loss": 0.9615,
-      "step": 1715
+      "loss": 1.0965,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7882055988985773,
+      "grad_norm": 2.5872874083756052,
+      "learning_rate": 2.610863211647038e-06,
+      "loss": 1.068,
+      "step": 3435
     },
     {
       "epoch": 0.7893529141808169,
-      "grad_norm": 1.7382322874229845,
+      "grad_norm": 2.3302790860000777,
       "learning_rate": 2.5839362464157635e-06,
-      "loss": 0.9523,
-      "step": 1720
+      "loss": 1.0682,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7905002294630564,
+      "grad_norm": 2.314204825191404,
+      "learning_rate": 2.5571282402241435e-06,
+      "loss": 1.0691,
+      "step": 3445
     },
     {
       "epoch": 0.791647544745296,
-      "grad_norm": 1.76018582887562,
+      "grad_norm": 5.3426573880056845,
       "learning_rate": 2.5304396230920346e-06,
-      "loss": 0.9661,
-      "step": 1725
+      "loss": 1.0978,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7927948600275355,
+      "grad_norm": 2.583278253386985,
+      "learning_rate": 2.5038708231242047e-06,
+      "loss": 1.0788,
+      "step": 3455
     },
     {
       "epoch": 0.7939421753097752,
-      "grad_norm": 1.6962974121939103,
+      "grad_norm": 2.374823385658736,
       "learning_rate": 2.477422266503473e-06,
-      "loss": 0.9491,
-      "step": 1730
+      "loss": 1.0795,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7950894905920147,
+      "grad_norm": 2.416491427084334,
+      "learning_rate": 2.4510943774838624e-06,
+      "loss": 1.0881,
+      "step": 3465
     },
     {
       "epoch": 0.7962368058742543,
-      "grad_norm": 1.6864869043430553,
+      "grad_norm": 2.340810846568605,
       "learning_rate": 2.424887578383799e-06,
-      "loss": 0.9495,
-      "step": 1735
+      "loss": 1.1142,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7973841211564938,
+      "grad_norm": 2.310980554847478,
+      "learning_rate": 2.398802289579347e-06,
+      "loss": 1.0724,
+      "step": 3475
     },
     {
       "epoch": 0.7985314364387334,
-      "grad_norm": 1.7487461110618132,
+      "grad_norm": 2.4718474311759424,
       "learning_rate": 2.3728389294974472e-06,
-      "loss": 0.9497,
-      "step": 1740
+      "loss": 1.0926,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7996787517209729,
+      "grad_norm": 2.381651829008113,
+      "learning_rate": 2.346997914609226e-06,
+      "loss": 1.119,
+      "step": 3485
     },
     {
       "epoch": 0.8008260670032125,
-      "grad_norm": 3.804635705666014,
+      "grad_norm": 2.3463971161415587,
       "learning_rate": 2.3212796594232947e-06,
-      "loss": 0.9633,
-      "step": 1745
+      "loss": 1.0799,
+      "step": 3490
+    },
+    {
+      "epoch": 0.801973382285452,
+      "grad_norm": 2.3280363845921923,
+      "learning_rate": 2.2956845764791126e-06,
+      "loss": 1.0642,
+      "step": 3495
     },
     {
       "epoch": 0.8031206975676916,
-      "grad_norm": 1.7104227286473583,
+      "grad_norm": 2.4526295488857124,
       "learning_rate": 2.2702130763403674e-06,
-      "loss": 0.9402,
-      "step": 1750
+      "loss": 1.0997,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8042680128499312,
+      "grad_norm": 2.354907631567458,
+      "learning_rate": 2.2448655675883936e-06,
+      "loss": 1.0713,
+      "step": 3505
     },
     {
       "epoch": 0.8054153281321708,
-      "grad_norm": 1.703649935060647,
+      "grad_norm": 2.275567450615706,
       "learning_rate": 2.2196424568156073e-06,
-      "loss": 0.9493,
-      "step": 1755
+      "loss": 1.0788,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8065626434144103,
+      "grad_norm": 2.592545388172047,
+      "learning_rate": 2.1945441486189913e-06,
+      "loss": 1.1048,
+      "step": 3515
     },
     {
       "epoch": 0.8077099586966499,
-      "grad_norm": 1.731876264748493,
+      "grad_norm": 2.3536440458369747,
       "learning_rate": 2.1695710455936115e-06,
-      "loss": 0.949,
-      "step": 1760
+      "loss": 1.0987,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8088572739788894,
+      "grad_norm": 2.3512904218054813,
+      "learning_rate": 2.144723548326142e-06,
+      "loss": 1.074,
+      "step": 3525
     },
     {
       "epoch": 0.810004589261129,
-      "grad_norm": 1.6008487931711857,
+      "grad_norm": 2.395880729511023,
       "learning_rate": 2.1200020553884603e-06,
-      "loss": 0.931,
-      "step": 1765
+      "loss": 1.0636,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8111519045433685,
+      "grad_norm": 2.2521164104118996,
+      "learning_rate": 2.095406963331236e-06,
+      "loss": 1.0795,
+      "step": 3535
     },
     {
       "epoch": 0.812299219825608,
-      "grad_norm": 1.7485750666616973,
+      "grad_norm": 2.337428316331845,
       "learning_rate": 2.0709386666775732e-06,
-      "loss": 0.9473,
-      "step": 1770
+      "loss": 1.0928,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8134465351078476,
+      "grad_norm": 2.2804754836362666,
+      "learning_rate": 2.0465975579166984e-06,
+      "loss": 1.077,
+      "step": 3545
     },
     {
       "epoch": 0.8145938503900872,
-      "grad_norm": 1.907625017544001,
+      "grad_norm": 2.3835852120695584,
       "learning_rate": 2.0223840274976413e-06,
-      "loss": 0.9921,
-      "step": 1775
+      "loss": 1.0681,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8157411656723268,
+      "grad_norm": 2.3451403504700408,
+      "learning_rate": 1.998298463822986e-06,
+      "loss": 1.0379,
+      "step": 3555
     },
     {
       "epoch": 0.8168884809545663,
-      "grad_norm": 1.7187402463869024,
+      "grad_norm": 2.391585746462503,
       "learning_rate": 1.9743412532426355e-06,
-      "loss": 0.9147,
-      "step": 1780
+      "loss": 1.0713,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8180357962368059,
+      "grad_norm": 2.565408804237695,
+      "learning_rate": 1.950512780047622e-06,
+      "loss": 1.0727,
+      "step": 3565
     },
     {
       "epoch": 0.8191831115190454,
-      "grad_norm": 1.6617657428032502,
+      "grad_norm": 2.2127377834416313,
       "learning_rate": 1.9268134264639273e-06,
-      "loss": 0.9457,
-      "step": 1785
+      "loss": 1.0681,
+      "step": 3570
+    },
+    {
+      "epoch": 0.820330426801285,
+      "grad_norm": 2.4867018259226623,
+      "learning_rate": 1.9032435726463716e-06,
+      "loss": 1.0865,
+      "step": 3575
     },
     {
       "epoch": 0.8214777420835245,
-      "grad_norm": 1.7007750210249686,
+      "grad_norm": 2.589421332937048,
       "learning_rate": 1.879803596672497e-06,
-      "loss": 0.9665,
-      "step": 1790
+      "loss": 1.0573,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8226250573657641,
+      "grad_norm": 2.7149409401173474,
+      "learning_rate": 1.8564938745365102e-06,
+      "loss": 1.0896,
+      "step": 3585
     },
     {
       "epoch": 0.8237723726480036,
-      "grad_norm": 1.7101753911626782,
+      "grad_norm": 2.433136107961442,
       "learning_rate": 1.8333147801432616e-06,
-      "loss": 0.929,
-      "step": 1795
+      "loss": 1.0915,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8249196879302433,
+      "grad_norm": 2.5073579970693887,
+      "learning_rate": 1.8102666853022277e-06,
+      "loss": 1.0904,
+      "step": 3595
     },
     {
       "epoch": 0.8260670032124828,
-      "grad_norm": 1.774895067675658,
+      "grad_norm": 2.3242486986861093,
       "learning_rate": 1.7873499597215604e-06,
-      "loss": 0.9205,
-      "step": 1800
+      "loss": 1.0629,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8272143184947224,
+      "grad_norm": 2.3675402875292035,
+      "learning_rate": 1.7645649710021528e-06,
+      "loss": 1.0695,
+      "step": 3605
     },
     {
       "epoch": 0.8283616337769619,
-      "grad_norm": 1.7257248036119928,
+      "grad_norm": 2.469832693689684,
       "learning_rate": 1.7419120846317462e-06,
-      "loss": 0.9482,
-      "step": 1805
+      "loss": 1.067,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8295089490592015,
+      "grad_norm": 2.2836763919133776,
+      "learning_rate": 1.7193916639790665e-06,
+      "loss": 1.0598,
+      "step": 3615
     },
     {
       "epoch": 0.830656264341441,
-      "grad_norm": 1.7250389140740945,
+      "grad_norm": 2.3357605323364665,
       "learning_rate": 1.697004070287982e-06,
-      "loss": 0.907,
-      "step": 1810
+      "loss": 1.0814,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8318035796236806,
+      "grad_norm": 2.419230420031552,
+      "learning_rate": 1.6747496626717318e-06,
+      "loss": 1.0622,
+      "step": 3625
     },
     {
       "epoch": 0.8329508949059201,
-      "grad_norm": 1.6576292112238928,
+      "grad_norm": 2.2377807288438363,
       "learning_rate": 1.6526287981071477e-06,
-      "loss": 0.9169,
-      "step": 1815
+      "loss": 1.0619,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8340982101881597,
+      "grad_norm": 2.313360528415539,
+      "learning_rate": 1.6306418314289408e-06,
+      "loss": 1.0754,
+      "step": 3635
     },
     {
       "epoch": 0.8352455254703993,
-      "grad_norm": 1.6244943528548752,
+      "grad_norm": 2.388054570280702,
       "learning_rate": 1.6087891153239932e-06,
-      "loss": 0.9707,
-      "step": 1820
+      "loss": 1.0828,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8363928407526389,
+      "grad_norm": 2.3362823868458267,
+      "learning_rate": 1.5870710003257162e-06,
+      "loss": 1.049,
+      "step": 3645
     },
     {
       "epoch": 0.8375401560348784,
-      "grad_norm": 1.7684196725742007,
+      "grad_norm": 2.3694652165740466,
       "learning_rate": 1.5654878348084246e-06,
-      "loss": 0.9597,
-      "step": 1825
+      "loss": 1.0882,
+      "step": 3650
+    },
+    {
+      "epoch": 0.838687471317118,
+      "grad_norm": 2.4438443231900355,
+      "learning_rate": 1.5440399649817384e-06,
+      "loss": 1.0493,
+      "step": 3655
     },
     {
       "epoch": 0.8398347865993575,
-      "grad_norm": 1.7498743654009963,
+      "grad_norm": 2.479186138101549,
       "learning_rate": 1.5227277348850466e-06,
-      "loss": 0.9438,
-      "step": 1830
+      "loss": 1.0997,
+      "step": 3660
+    },
+    {
+      "epoch": 0.840982101881597,
+      "grad_norm": 2.298179774902708,
+      "learning_rate": 1.5015514863819625e-06,
+      "loss": 1.0802,
+      "step": 3665
     },
     {
       "epoch": 0.8421294171638366,
-      "grad_norm": 1.666533412682978,
+      "grad_norm": 2.384294974826556,
       "learning_rate": 1.4805115591548746e-06,
-      "loss": 0.9234,
-      "step": 1835
+      "loss": 1.1026,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8432767324460761,
+      "grad_norm": 2.2772068888876635,
+      "learning_rate": 1.4596082906994658e-06,
+      "loss": 1.0767,
+      "step": 3675
     },
     {
       "epoch": 0.8444240477283157,
-      "grad_norm": 1.6784896812471775,
+      "grad_norm": 2.539754475914315,
       "learning_rate": 1.4388420163193217e-06,
-      "loss": 0.9228,
-      "step": 1840
+      "loss": 1.0786,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8455713630105554,
+      "grad_norm": 2.2681491099219464,
+      "learning_rate": 1.4182130691205399e-06,
+      "loss": 1.0681,
+      "step": 3685
     },
     {
       "epoch": 0.8467186782927949,
-      "grad_norm": 1.782913468726149,
+      "grad_norm": 2.328527473494352,
       "learning_rate": 1.3977217800063847e-06,
-      "loss": 0.9525,
-      "step": 1845
+      "loss": 1.0636,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8478659935750344,
+      "grad_norm": 2.3756645866176567,
+      "learning_rate": 1.3773684776719987e-06,
+      "loss": 1.0834,
+      "step": 3695
     },
     {
       "epoch": 0.849013308857274,
-      "grad_norm": 1.8714100640537485,
+      "grad_norm": 2.3382653231993684,
       "learning_rate": 1.3571534885991044e-06,
-      "loss": 0.9202,
-      "step": 1850
+      "loss": 1.0656,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8501606241395135,
+      "grad_norm": 2.3242086558835173,
+      "learning_rate": 1.337077137050784e-06,
+      "loss": 1.0451,
+      "step": 3705
     },
     {
       "epoch": 0.8513079394217531,
-      "grad_norm": 1.6341170338063304,
+      "grad_norm": 2.269434671317697,
       "learning_rate": 1.3171397450662716e-06,
-      "loss": 0.9324,
-      "step": 1855
+      "loss": 1.0571,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8524552547039926,
+      "grad_norm": 2.2469307076032483,
+      "learning_rate": 1.297341632455793e-06,
+      "loss": 1.0557,
+      "step": 3715
     },
     {
       "epoch": 0.8536025699862322,
-      "grad_norm": 1.7569629604181838,
+      "grad_norm": 2.420410716316704,
       "learning_rate": 1.2776831167954252e-06,
-      "loss": 0.92,
-      "step": 1860
+      "loss": 1.0919,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8547498852684717,
+      "grad_norm": 2.406358800823869,
+      "learning_rate": 1.258164513422019e-06,
+      "loss": 1.0696,
+      "step": 3725
     },
     {
       "epoch": 0.8558972005507114,
-      "grad_norm": 1.7383544701565197,
+      "grad_norm": 2.312198841285599,
       "learning_rate": 1.2387861354281194e-06,
-      "loss": 0.9415,
-      "step": 1865
+      "loss": 1.0645,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8570445158329509,
+      "grad_norm": 2.337798774099059,
+      "learning_rate": 1.2195482936569603e-06,
+      "loss": 1.0698,
+      "step": 3735
     },
     {
       "epoch": 0.8581918311151905,
-      "grad_norm": 1.6322549841517964,
+      "grad_norm": 2.3239212258602424,
       "learning_rate": 1.2004512966974746e-06,
-      "loss": 0.9377,
-      "step": 1870
+      "loss": 1.0672,
+      "step": 3740
+    },
+    {
+      "epoch": 0.85933914639743,
+      "grad_norm": 2.317821805872963,
+      "learning_rate": 1.1814954508793397e-06,
+      "loss": 1.0875,
+      "step": 3745
     },
     {
       "epoch": 0.8604864616796696,
-      "grad_norm": 1.6131611632532987,
+      "grad_norm": 2.34518455218088,
       "learning_rate": 1.162681060268065e-06,
-      "loss": 0.9339,
-      "step": 1875
+      "loss": 1.0615,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8616337769619091,
+      "grad_norm": 2.3102053326026226,
+      "learning_rate": 1.1440084266601148e-06,
+      "loss": 1.0692,
+      "step": 3755
     },
     {
       "epoch": 0.8627810922441487,
-      "grad_norm": 1.825918080886602,
+      "grad_norm": 2.2988208279144744,
       "learning_rate": 1.1254778495780749e-06,
-      "loss": 0.9435,
-      "step": 1880
+      "loss": 1.0365,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8639284075263882,
+      "grad_norm": 2.3343105163956603,
+      "learning_rate": 1.1070896262658381e-06,
+      "loss": 1.0526,
+      "step": 3765
     },
     {
       "epoch": 0.8650757228086278,
-      "grad_norm": 1.7362620117149066,
+      "grad_norm": 2.484363397400702,
       "learning_rate": 1.0888440516838373e-06,
-      "loss": 0.9375,
-      "step": 1885
+      "loss": 1.0601,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8662230380908674,
+      "grad_norm": 2.3781607794246287,
+      "learning_rate": 1.0707414185043163e-06,
+      "loss": 1.0706,
+      "step": 3775
     },
     {
       "epoch": 0.867370353373107,
-      "grad_norm": 1.6924752721498857,
+      "grad_norm": 2.3738365009035376,
       "learning_rate": 1.0527820171066372e-06,
-      "loss": 0.9518,
-      "step": 1890
+      "loss": 1.0825,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8685176686553465,
+      "grad_norm": 2.407144887683863,
+      "learning_rate": 1.0349661355726215e-06,
+      "loss": 1.0824,
+      "step": 3785
     },
     {
       "epoch": 0.8696649839375861,
-      "grad_norm": 1.7645375725424164,
+      "grad_norm": 2.3740565527233812,
       "learning_rate": 1.0172940596819258e-06,
-      "loss": 0.9105,
-      "step": 1895
+      "loss": 1.0609,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8708122992198256,
+      "grad_norm": 2.4197584148785265,
+      "learning_rate": 9.997660729074587e-07,
+      "loss": 1.0759,
+      "step": 3795
     },
     {
       "epoch": 0.8719596145020652,
-      "grad_norm": 1.7445241378126046,
+      "grad_norm": 2.4001709433977165,
       "learning_rate": 9.823824564108408e-07,
-      "loss": 0.9534,
-      "step": 1900
+      "loss": 1.0529,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8731069297843047,
+      "grad_norm": 2.3761802261416745,
+      "learning_rate": 9.651434890378797e-07,
+      "loss": 1.0864,
+      "step": 3805
     },
     {
       "epoch": 0.8742542450665443,
-      "grad_norm": 1.6812050382224737,
+      "grad_norm": 2.390851465404339,
       "learning_rate": 9.480494473141189e-07,
-      "loss": 0.9513,
-      "step": 1905
+      "loss": 1.0641,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8754015603487838,
+      "grad_norm": 2.2302978594213583,
+      "learning_rate": 9.311006054403726e-07,
+      "loss": 1.0548,
+      "step": 3815
     },
     {
       "epoch": 0.8765488756310233,
-      "grad_norm": 1.7180889119146323,
+      "grad_norm": 2.375126839515668,
       "learning_rate": 9.142972352883595e-07,
-      "loss": 0.9352,
-      "step": 1910
+      "loss": 1.0781,
+      "step": 3820
+    },
+    {
+      "epoch": 0.877696190913263,
+      "grad_norm": 2.3054748603808237,
+      "learning_rate": 8.976396063963156e-07,
+      "loss": 1.0679,
+      "step": 3825
     },
     {
       "epoch": 0.8788435061955026,
-      "grad_norm": 1.593136362005407,
+      "grad_norm": 2.6101738491562783,
       "learning_rate": 8.811279859646915e-07,
-      "loss": 0.9135,
-      "step": 1915
+      "loss": 1.038,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8799908214777421,
+      "grad_norm": 2.2926250243874575,
+      "learning_rate": 8.647626388518471e-07,
+      "loss": 1.0416,
+      "step": 3835
     },
     {
       "epoch": 0.8811381367599816,
-      "grad_norm": 1.732265054877238,
+      "grad_norm": 2.3466635198769112,
       "learning_rate": 8.485438275698154e-07,
-      "loss": 0.9325,
-      "step": 1920
+      "loss": 1.0915,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8822854520422212,
+      "grad_norm": 2.346242158668616,
+      "learning_rate": 8.324718122800912e-07,
+      "loss": 1.0658,
+      "step": 3845
     },
     {
       "epoch": 0.8834327673244607,
-      "grad_norm": 1.6465162040573431,
+      "grad_norm": 2.2828753234624672,
       "learning_rate": 8.165468507894514e-07,
-      "loss": 0.9363,
-      "step": 1925
+      "loss": 1.0691,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8845800826067003,
+      "grad_norm": 2.2381710090655362,
+      "learning_rate": 8.007691985458277e-07,
+      "loss": 1.0695,
+      "step": 3855
     },
     {
       "epoch": 0.8857273978889398,
-      "grad_norm": 1.5926308744021644,
+      "grad_norm": 2.3043691660140637,
       "learning_rate": 7.851391086341953e-07,
-      "loss": 0.9251,
-      "step": 1930
+      "loss": 1.0627,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8868747131711794,
+      "grad_norm": 2.3258581770015176,
+      "learning_rate": 7.696568317725339e-07,
+      "loss": 1.0805,
+      "step": 3865
     },
     {
       "epoch": 0.888022028453419,
-      "grad_norm": 1.7096508943146629,
+      "grad_norm": 2.3844984378244263,
       "learning_rate": 7.543226163077899e-07,
-      "loss": 0.9471,
-      "step": 1935
+      "loss": 1.0669,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8891693437356586,
+      "grad_norm": 2.4109038157393012,
+      "learning_rate": 7.391367082118961e-07,
+      "loss": 1.0861,
+      "step": 3875
     },
     {
       "epoch": 0.8903166590178981,
-      "grad_norm": 1.803335599460816,
+      "grad_norm": 2.355682219549452,
       "learning_rate": 7.240993510778304e-07,
-      "loss": 0.9328,
-      "step": 1940
+      "loss": 1.0607,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8914639743001377,
+      "grad_norm": 2.3034428538054232,
+      "learning_rate": 7.092107861157004e-07,
+      "loss": 1.0718,
+      "step": 3885
     },
     {
       "epoch": 0.8926112895823772,
-      "grad_norm": 1.7212998187307902,
+      "grad_norm": 2.3594432455984116,
       "learning_rate": 6.944712521488884e-07,
-      "loss": 0.9479,
-      "step": 1945
+      "loss": 1.058,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8937586048646168,
+      "grad_norm": 2.384920183351963,
+      "learning_rate": 6.798809856102028e-07,
+      "loss": 1.0624,
+      "step": 3895
     },
     {
       "epoch": 0.8949059201468563,
-      "grad_norm": 1.735482839392802,
+      "grad_norm": 2.315792099017849,
       "learning_rate": 6.654402205380961e-07,
-      "loss": 0.9541,
-      "step": 1950
+      "loss": 1.0824,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8960532354290959,
+      "grad_norm": 2.354796149058742,
+      "learning_rate": 6.511491885729149e-07,
+      "loss": 1.0613,
+      "step": 3905
     },
     {
       "epoch": 0.8972005507113354,
-      "grad_norm": 1.65097416177555,
+      "grad_norm": 2.397012765505873,
       "learning_rate": 6.370081189531707e-07,
-      "loss": 0.9132,
-      "step": 1955
+      "loss": 1.0625,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8983478659935751,
+      "grad_norm": 2.400999165531199,
+      "learning_rate": 6.230172385118738e-07,
+      "loss": 1.0912,
+      "step": 3915
     },
     {
       "epoch": 0.8994951812758146,
-      "grad_norm": 1.7196384780834688,
+      "grad_norm": 2.2856810018900036,
       "learning_rate": 6.091767716728924e-07,
-      "loss": 0.9341,
-      "step": 1960
+      "loss": 1.0627,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9006424965580542,
+      "grad_norm": 2.4240133866264273,
+      "learning_rate": 5.954869404473473e-07,
+      "loss": 1.1025,
+      "step": 3925
     },
     {
       "epoch": 0.9017898118402937,
-      "grad_norm": 1.8341416005690085,
+      "grad_norm": 2.3880077625070313,
       "learning_rate": 5.819479644300563e-07,
-      "loss": 0.928,
-      "step": 1965
+      "loss": 1.0495,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9029371271225333,
+      "grad_norm": 2.316683568122776,
+      "learning_rate": 5.685600607960129e-07,
+      "loss": 1.0508,
+      "step": 3935
     },
     {
       "epoch": 0.9040844424047728,
-      "grad_norm": 1.7371545663059045,
+      "grad_norm": 2.2964133731561525,
       "learning_rate": 5.553234442969014e-07,
-      "loss": 0.9311,
-      "step": 1970
+      "loss": 1.0649,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9052317576870124,
+      "grad_norm": 2.385202578198628,
+      "learning_rate": 5.422383272576426e-07,
+      "loss": 1.0615,
+      "step": 3945
     },
     {
       "epoch": 0.9063790729692519,
-      "grad_norm": 1.6781199455045805,
+      "grad_norm": 2.3105288018374512,
       "learning_rate": 5.293049195730038e-07,
-      "loss": 0.9394,
-      "step": 1975
+      "loss": 1.0376,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9075263882514915,
+      "grad_norm": 2.3047993770301893,
+      "learning_rate": 5.165234287042198e-07,
+      "loss": 1.0253,
+      "step": 3955
     },
     {
       "epoch": 0.9086737035337311,
-      "grad_norm": 1.7582513777384394,
+      "grad_norm": 2.350917765719763,
       "learning_rate": 5.038940596756747e-07,
-      "loss": 0.9495,
-      "step": 1980
+      "loss": 1.0706,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9098210188159707,
+      "grad_norm": 2.4392692899669117,
+      "learning_rate": 4.914170150716024e-07,
+      "loss": 1.0762,
+      "step": 3965
     },
     {
       "epoch": 0.9109683340982102,
-      "grad_norm": 1.6554350150475086,
+      "grad_norm": 2.3371470208418916,
       "learning_rate": 4.790924950328435e-07,
-      "loss": 0.9491,
-      "step": 1985
+      "loss": 1.0453,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9121156493804498,
+      "grad_norm": 2.204682457809935,
+      "learning_rate": 4.6692069725363887e-07,
+      "loss": 1.0761,
+      "step": 3975
     },
     {
       "epoch": 0.9132629646626893,
-      "grad_norm": 1.5696628957382233,
+      "grad_norm": 2.290854793691128,
       "learning_rate": 4.5490181697844916e-07,
-      "loss": 0.9047,
-      "step": 1990
+      "loss": 1.052,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9144102799449288,
+      "grad_norm": 2.2988253093129174,
+      "learning_rate": 4.4303604699882594e-07,
+      "loss": 1.0451,
+      "step": 3985
     },
     {
       "epoch": 0.9155575952271684,
-      "grad_norm": 1.660006052531741,
+      "grad_norm": 2.26124979946759,
       "learning_rate": 4.313235776503244e-07,
-      "loss": 0.9266,
-      "step": 1995
+      "loss": 1.0555,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9167049105094079,
+      "grad_norm": 2.274647741800421,
+      "learning_rate": 4.197645968094466e-07,
+      "loss": 1.0421,
+      "step": 3995
     },
     {
       "epoch": 0.9178522257916475,
-      "grad_norm": 1.6789601308304245,
+      "grad_norm": 2.3957337901708953,
       "learning_rate": 4.08359289890623e-07,
-      "loss": 0.9356,
-      "step": 2000
+      "loss": 1.0559,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9189995410738871,
+      "grad_norm": 2.3442250173869827,
+      "learning_rate": 3.971078398432482e-07,
+      "loss": 1.0783,
+      "step": 4005
     },
     {
       "epoch": 0.9201468563561267,
-      "grad_norm": 1.6938274156881534,
+      "grad_norm": 2.3642424317720954,
       "learning_rate": 3.860104271487397e-07,
-      "loss": 0.9501,
-      "step": 2005
+      "loss": 1.0403,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9212941716383662,
+      "grad_norm": 2.2488142508080893,
+      "learning_rate": 3.750672298176405e-07,
+      "loss": 1.0504,
+      "step": 4015
     },
     {
       "epoch": 0.9224414869206058,
-      "grad_norm": 1.6613669924815342,
+      "grad_norm": 2.3517485020535576,
       "learning_rate": 3.6427842338677353e-07,
-      "loss": 0.9178,
-      "step": 2010
+      "loss": 1.0618,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9235888022028453,
+      "grad_norm": 2.3231400273137033,
+      "learning_rate": 3.5364418091641374e-07,
+      "loss": 1.0808,
+      "step": 4025
     },
     {
       "epoch": 0.9247361174850849,
-      "grad_norm": 1.6863633601389076,
+      "grad_norm": 2.3960144408780715,
       "learning_rate": 3.4316467298752264e-07,
-      "loss": 0.9283,
-      "step": 2015
+      "loss": 1.0768,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9258834327673244,
+      "grad_norm": 2.481754354110254,
+      "learning_rate": 3.328400676990029e-07,
+      "loss": 1.0819,
+      "step": 4035
     },
     {
       "epoch": 0.927030748049564,
-      "grad_norm": 1.6704738495679314,
+      "grad_norm": 2.261698190389203,
       "learning_rate": 3.226705306650113e-07,
-      "loss": 0.9397,
-      "step": 2020
+      "loss": 1.0653,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9281780633318035,
+      "grad_norm": 2.347018729974591,
+      "learning_rate": 3.1265622501229554e-07,
+      "loss": 1.0485,
+      "step": 4045
     },
     {
       "epoch": 0.9293253786140432,
-      "grad_norm": 1.6915649797171424,
+      "grad_norm": 2.3350732569344985,
       "learning_rate": 3.027973113775795e-07,
-      "loss": 0.9162,
-      "step": 2025
+      "loss": 1.0678,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9304726938962827,
+      "grad_norm": 2.1947433426596867,
+      "learning_rate": 2.9309394790498547e-07,
+      "loss": 1.057,
+      "step": 4055
     },
     {
       "epoch": 0.9316200091785223,
-      "grad_norm": 1.624516748205876,
+      "grad_norm": 2.3905173803456816,
       "learning_rate": 2.835462902434971e-07,
-      "loss": 0.9088,
-      "step": 2030
+      "loss": 1.0724,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9327673244607618,
+      "grad_norm": 2.4899822063351524,
+      "learning_rate": 2.741544915444694e-07,
+      "loss": 1.0759,
+      "step": 4065
     },
     {
       "epoch": 0.9339146397430014,
-      "grad_norm": 1.7321228286245276,
+      "grad_norm": 2.3105967151836206,
       "learning_rate": 2.649187024591604e-07,
-      "loss": 0.9291,
-      "step": 2035
+      "loss": 1.0325,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9350619550252409,
+      "grad_norm": 2.3253855249232105,
+      "learning_rate": 2.5583907113632456e-07,
+      "loss": 1.0736,
+      "step": 4075
     },
     {
       "epoch": 0.9362092703074805,
-      "grad_norm": 1.6981737432031856,
+      "grad_norm": 2.5115169921624334,
       "learning_rate": 2.4691574321983216e-07,
-      "loss": 0.9455,
-      "step": 2040
+      "loss": 1.0662,
+      "step": 4080
+    },
+    {
+      "epoch": 0.93735658558972,
+      "grad_norm": 2.240118179487899,
+      "learning_rate": 2.3814886184633012e-07,
+      "loss": 1.0598,
+      "step": 4085
     },
     {
       "epoch": 0.9385039008719596,
-      "grad_norm": 1.6209611667412176,
+      "grad_norm": 2.4398202590323885,
       "learning_rate": 2.2953856764295623e-07,
-      "loss": 0.9427,
-      "step": 2045
+      "loss": 1.0763,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9396512161541992,
+      "grad_norm": 2.384051442170586,
+      "learning_rate": 2.210849987250685e-07,
+      "loss": 1.0578,
+      "step": 4095
     },
     {
       "epoch": 0.9407985314364388,
-      "grad_norm": 1.6075798817603826,
+      "grad_norm": 2.521911651181548,
       "learning_rate": 2.1278829069404483e-07,
-      "loss": 0.8938,
-      "step": 2050
+      "loss": 1.0507,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9419458467186783,
+      "grad_norm": 2.312630644718287,
+      "learning_rate": 2.0464857663509473e-07,
+      "loss": 1.0632,
+      "step": 4105
     },
     {
       "epoch": 0.9430931620009179,
-      "grad_norm": 1.6694215956258427,
+      "grad_norm": 2.28669658207446,
       "learning_rate": 1.9666598711513663e-07,
-      "loss": 0.9425,
-      "step": 2055
+      "loss": 1.0546,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9442404772831574,
+      "grad_norm": 2.4280041176246336,
+      "learning_rate": 1.8884065018069165e-07,
+      "loss": 1.0682,
+      "step": 4115
     },
     {
       "epoch": 0.945387792565397,
-      "grad_norm": 1.6380060844501796,
+      "grad_norm": 2.3965039181313172,
       "learning_rate": 1.811726913558387e-07,
-      "loss": 0.9344,
-      "step": 2060
+      "loss": 1.053,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9465351078476365,
+      "grad_norm": 2.317349780596125,
+      "learning_rate": 1.736622336401983e-07,
+      "loss": 1.0586,
+      "step": 4125
     },
     {
       "epoch": 0.947682423129876,
-      "grad_norm": 1.7361661814337723,
+      "grad_norm": 2.4267122952778695,
       "learning_rate": 1.663093975069552e-07,
-      "loss": 0.9264,
-      "step": 2065
+      "loss": 1.0677,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9488297384121156,
+      "grad_norm": 2.267834335988602,
+      "learning_rate": 1.5911430090093437e-07,
+      "loss": 1.0375,
+      "step": 4135
     },
     {
       "epoch": 0.9499770536943553,
-      "grad_norm": 1.72352480478006,
+      "grad_norm": 2.2507286853587347,
       "learning_rate": 1.5207705923670158e-07,
-      "loss": 0.9274,
-      "step": 2070
+      "loss": 1.0486,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9511243689765948,
+      "grad_norm": 2.351735220333465,
+      "learning_rate": 1.451977853967146e-07,
+      "loss": 1.0634,
+      "step": 4145
     },
     {
       "epoch": 0.9522716842588343,
-      "grad_norm": 1.7260384890222744,
+      "grad_norm": 2.354995051605583,
       "learning_rate": 1.3847658972951482e-07,
-      "loss": 0.9416,
-      "step": 2075
+      "loss": 1.059,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9534189995410739,
+      "grad_norm": 2.4112199312162312,
+      "learning_rate": 1.319135800479543e-07,
+      "loss": 1.0663,
+      "step": 4155
     },
     {
       "epoch": 0.9545663148233134,
-      "grad_norm": 1.678840542944468,
+      "grad_norm": 2.224163787960051,
       "learning_rate": 1.2550886162746468e-07,
-      "loss": 0.9476,
-      "step": 2080
+      "loss": 1.0321,
+      "step": 4160
+    },
+    {
+      "epoch": 0.955713630105553,
+      "grad_norm": 2.3832043966523475,
+      "learning_rate": 1.192625372043754e-07,
+      "loss": 1.0558,
+      "step": 4165
     },
     {
       "epoch": 0.9568609453877925,
-      "grad_norm": 1.7327761448558552,
+      "grad_norm": 2.573113863913223,
       "learning_rate": 1.1317470697425837e-07,
-      "loss": 0.9433,
-      "step": 2085
+      "loss": 1.0453,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9580082606700321,
+      "grad_norm": 2.36325059791216,
+      "learning_rate": 1.072454685903257e-07,
+      "loss": 1.0619,
+      "step": 4175
     },
     {
       "epoch": 0.9591555759522716,
-      "grad_norm": 1.6188397369323666,
+      "grad_norm": 2.3121307241018036,
       "learning_rate": 1.0147491716185675e-07,
-      "loss": 0.9306,
-      "step": 2090
+      "loss": 1.0611,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9603028912345113,
+      "grad_norm": 2.400560326114734,
+      "learning_rate": 9.586314525268369e-08,
+      "loss": 1.0828,
+      "step": 4185
     },
     {
       "epoch": 0.9614502065167508,
-      "grad_norm": 1.7303211630716537,
+      "grad_norm": 2.422338514530241,
       "learning_rate": 9.041024287969491e-08,
-      "loss": 0.951,
-      "step": 2095
+      "loss": 1.062,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9625975217989904,
+      "grad_norm": 2.3711025333563263,
+      "learning_rate": 8.511629751139949e-08,
+      "loss": 1.046,
+      "step": 4195
     },
     {
       "epoch": 0.9637448370812299,
-      "grad_norm": 1.6406456484197742,
+      "grad_norm": 2.3607300247468586,
       "learning_rate": 7.99813940665195e-08,
-      "loss": 0.9414,
-      "step": 2100
+      "loss": 1.0718,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9648921523634695,
+      "grad_norm": 2.2844036833072745,
+      "learning_rate": 7.50056149126277e-08,
+      "loss": 1.0338,
+      "step": 4205
     },
     {
       "epoch": 0.966039467645709,
-      "grad_norm": 1.742920472563737,
+      "grad_norm": 2.412561145835342,
       "learning_rate": 7.018903986483083e-08,
-      "loss": 0.9331,
-      "step": 2105
+      "loss": 1.0733,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9671867829279486,
+      "grad_norm": 2.3903032571937444,
+      "learning_rate": 6.553174618448399e-08,
+      "loss": 1.0629,
+      "step": 4215
     },
     {
       "epoch": 0.9683340982101881,
-      "grad_norm": 1.7699514892789145,
+      "grad_norm": 2.3221221391589038,
       "learning_rate": 6.103380857795604e-08,
-      "loss": 0.9458,
-      "step": 2110
+      "loss": 1.0822,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9694814134924277,
+      "grad_norm": 2.285250340291083,
+      "learning_rate": 5.6695299195425045e-08,
+      "loss": 1.0854,
+      "step": 4225
     },
     {
       "epoch": 0.9706287287746673,
-      "grad_norm": 1.6207392453935818,
+      "grad_norm": 2.2592180518969434,
       "learning_rate": 5.251628762972916e-08,
-      "loss": 0.9286,
-      "step": 2115
+      "loss": 1.0692,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9717760440569069,
+      "grad_norm": 2.31137882081886,
+      "learning_rate": 4.84968409152442e-08,
+      "loss": 1.076,
+      "step": 4235
     },
     {
       "epoch": 0.9729233593391464,
-      "grad_norm": 1.6654831370842644,
+      "grad_norm": 2.3345498965147993,
       "learning_rate": 4.4637023526807875e-08,
-      "loss": 0.9207,
-      "step": 2120
+      "loss": 1.0913,
+      "step": 4240
+    },
+    {
+      "epoch": 0.974070674621386,
+      "grad_norm": 2.306918943368033,
+      "learning_rate": 4.0936897378691664e-08,
+      "loss": 1.0631,
+      "step": 4245
     },
     {
       "epoch": 0.9752179899036255,
-      "grad_norm": 1.6652739914078956,
+      "grad_norm": 2.3923389986430137,
       "learning_rate": 3.739652182360054e-08,
-      "loss": 0.9153,
-      "step": 2125
+      "loss": 1.0489,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9763653051858651,
+      "grad_norm": 2.3531625372494007,
+      "learning_rate": 3.401595365172483e-08,
+      "loss": 1.0733,
+      "step": 4255
     },
     {
       "epoch": 0.9775126204681046,
-      "grad_norm": 1.7380457873464283,
+      "grad_norm": 2.436621801398954,
       "learning_rate": 3.079524708983095e-08,
-      "loss": 0.9193,
-      "step": 2130
+      "loss": 1.0576,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9786599357503442,
+      "grad_norm": 2.3786365675633108,
+      "learning_rate": 2.773445380038653e-08,
+      "loss": 1.0554,
+      "step": 4265
     },
     {
       "epoch": 0.9798072510325837,
-      "grad_norm": 1.7928242828367038,
+      "grad_norm": 2.3509821615607454,
       "learning_rate": 2.483362288073443e-08,
-      "loss": 0.9258,
-      "step": 2135
+      "loss": 1.0824,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9809545663148234,
+      "grad_norm": 2.3808539729525844,
+      "learning_rate": 2.2092800862305587e-08,
+      "loss": 1.0991,
+      "step": 4275
     },
     {
       "epoch": 0.9821018815970629,
-      "grad_norm": 1.6588826041799296,
+      "grad_norm": 2.249305593751699,
       "learning_rate": 1.9512031709874037e-08,
-      "loss": 0.9275,
-      "step": 2140
+      "loss": 1.0636,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9832491968793025,
+      "grad_norm": 2.3078218741343544,
+      "learning_rate": 1.7091356820848616e-08,
+      "loss": 1.0602,
+      "step": 4285
     },
     {
       "epoch": 0.984396512161542,
-      "grad_norm": 1.763891394623808,
+      "grad_norm": 2.529969042443467,
       "learning_rate": 1.4830815024606815e-08,
-      "loss": 0.9211,
-      "step": 2145
+      "loss": 1.0555,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9855438274437816,
+      "grad_norm": 2.3368971664389084,
+      "learning_rate": 1.2730442581879721e-08,
+      "loss": 1.0576,
+      "step": 4295
     },
     {
       "epoch": 0.9866911427260211,
-      "grad_norm": 1.782131811620528,
+      "grad_norm": 2.478745371368321,
       "learning_rate": 1.0790273184164701e-08,
-      "loss": 0.8983,
-      "step": 2150
+      "loss": 1.0713,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9878384580082606,
+      "grad_norm": 2.347771751748443,
+      "learning_rate": 9.010337953185843e-09,
+      "loss": 1.0794,
+      "step": 4305
     },
     {
       "epoch": 0.9889857732905002,
-      "grad_norm": 2.0241090268732402,
+      "grad_norm": 2.3670492207343177,
       "learning_rate": 7.390665440393241e-09,
-      "loss": 0.9335,
-      "step": 2155
+      "loss": 1.0731,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9901330885727397,
+      "grad_norm": 2.301292219008886,
+      "learning_rate": 5.931281626508911e-09,
+      "loss": 1.0511,
+      "step": 4315
     },
     {
       "epoch": 0.9912804038549794,
-      "grad_norm": 1.6899996297251758,
+      "grad_norm": 2.3578534915978784,
       "learning_rate": 4.632209921107133e-09,
-      "loss": 0.9444,
-      "step": 2160
+      "loss": 1.0602,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9924277191372189,
+      "grad_norm": 2.2665225995048828,
+      "learning_rate": 3.493471162241413e-09,
+      "loss": 1.0423,
+      "step": 4325
     },
     {
       "epoch": 0.9935750344194585,
-      "grad_norm": 1.6477330125103407,
+      "grad_norm": 2.3787177342094417,
       "learning_rate": 2.5150836161058624e-09,
-      "loss": 0.9399,
-      "step": 2165
+      "loss": 1.0538,
+      "step": 4330
+    },
+    {
+      "epoch": 0.994722349701698,
+      "grad_norm": 2.308731224538364,
+      "learning_rate": 1.6970629767465441e-09,
+      "loss": 1.0852,
+      "step": 4335
     },
     {
       "epoch": 0.9958696649839376,
-      "grad_norm": 1.7301015675243123,
+      "grad_norm": 2.402013691289646,
       "learning_rate": 1.03942236580723e-09,
-      "loss": 0.9332,
-      "step": 2170
+      "loss": 1.0744,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9970169802661771,
+      "grad_norm": 2.4162986481825364,
+      "learning_rate": 5.421723323195682e-10,
+      "loss": 1.0759,
+      "step": 4345
     },
     {
       "epoch": 0.9981642955484167,
-      "grad_norm": 1.7014766333026237,
+      "grad_norm": 2.445243310213381,
       "learning_rate": 2.053208525365502e-10,
-      "loss": 0.9409,
-      "step": 2175
+      "loss": 1.0729,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9993116108306562,
+      "grad_norm": 2.3681587132053297,
+      "learning_rate": 2.8873329798173588e-11,
+      "loss": 1.0697,
+      "step": 4355
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.9490520358085632,
-      "eval_runtime": 758.2182,
-      "eval_samples_per_second": 20.352,
-      "eval_steps_per_second": 0.637,
-      "step": 2179
+      "eval_loss": 1.1125566959381104,
+      "eval_runtime": 852.54,
+      "eval_samples_per_second": 72.392,
+      "eval_steps_per_second": 1.132,
+      "step": 4358
     },
     {
       "epoch": 1.0,
-      "step": 2179,
+      "step": 4358,
       "total_flos": 456238269726720.0,
-      "train_loss": 1.0045825210718344,
-      "train_runtime": 27286.1988,
-      "train_samples_per_second": 5.11,
-      "train_steps_per_second": 0.08
+      "train_loss": 1.1632422284009862,
+      "train_runtime": 29234.3114,
+      "train_samples_per_second": 19.077,
+      "train_steps_per_second": 0.149
     }
   ],
   "logging_steps": 5,
-  "max_steps": 2179,
+  "max_steps": 4358,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 100,
   "total_flos": 456238269726720.0,
-  "train_batch_size": 16,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }