diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.7499691218246943,
+  "epoch": 0.9999588290995924,
   "eval_steps": 759,
-  "global_step": 2277,
+  "global_step": 3036,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -15978,6 +15978,5327 @@
       "eval_samples_per_second": 3.34,
       "eval_steps_per_second": 1.671,
       "step": 2277
+    },
+    {
+      "epoch": 0.750298489027955,
+      "grad_norm": 2.1856529712677,
+      "learning_rate": 1.5075010885937351e-05,
+      "loss": 2.0107,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7506278562312158,
+      "grad_norm": 2.1313235759735107,
+      "learning_rate": 1.5037385280283556e-05,
+      "loss": 2.5175,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7509572234344765,
+      "grad_norm": 2.2917118072509766,
+      "learning_rate": 1.4999798375791584e-05,
+      "loss": 2.2043,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7512865906377373,
+      "grad_norm": 2.0116426944732666,
+      "learning_rate": 1.4962250214067525e-05,
+      "loss": 2.001,
+      "step": 2281
+    },
+    {
+      "epoch": 0.751615957840998,
+      "grad_norm": 2.135652542114258,
+      "learning_rate": 1.4924740836674523e-05,
+      "loss": 2.2404,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7519453250442587,
+      "grad_norm": 2.7450170516967773,
+      "learning_rate": 1.4887270285132849e-05,
+      "loss": 2.1799,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7522746922475194,
+      "grad_norm": 2.9586498737335205,
+      "learning_rate": 1.484983860091977e-05,
+      "loss": 2.5526,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7526040594507802,
+      "grad_norm": 2.5501980781555176,
+      "learning_rate": 1.4812445825469534e-05,
+      "loss": 2.2452,
+      "step": 2285
+    },
+    {
+      "epoch": 0.752933426654041,
+      "grad_norm": 2.3167929649353027,
+      "learning_rate": 1.4775092000173323e-05,
+      "loss": 2.1382,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7532627938573017,
+      "grad_norm": 2.4865753650665283,
+      "learning_rate": 1.473777716637922e-05,
+      "loss": 2.2168,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7535921610605624,
+      "grad_norm": 2.432619333267212,
+      "learning_rate": 1.4700501365392089e-05,
+      "loss": 2.033,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7539215282638231,
+      "grad_norm": 2.7285501956939697,
+      "learning_rate": 1.4663264638473644e-05,
+      "loss": 2.0842,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7542508954670839,
+      "grad_norm": 2.7809548377990723,
+      "learning_rate": 1.4626067026842317e-05,
+      "loss": 2.2073,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7545802626703446,
+      "grad_norm": 2.550999879837036,
+      "learning_rate": 1.4588908571673287e-05,
+      "loss": 1.9711,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7549096298736053,
+      "grad_norm": 2.92107892036438,
+      "learning_rate": 1.4551789314098342e-05,
+      "loss": 2.2896,
+      "step": 2292
+    },
+    {
+      "epoch": 0.755238997076866,
+      "grad_norm": 2.7185511589050293,
+      "learning_rate": 1.4514709295205908e-05,
+      "loss": 2.1395,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7555683642801269,
+      "grad_norm": 3.085444211959839,
+      "learning_rate": 1.4477668556040936e-05,
+      "loss": 1.8752,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7558977314833876,
+      "grad_norm": 2.5131072998046875,
+      "learning_rate": 1.4440667137604936e-05,
+      "loss": 2.0242,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7562270986866483,
+      "grad_norm": 2.6334311962127686,
+      "learning_rate": 1.440370508085589e-05,
+      "loss": 2.0597,
+      "step": 2296
+    },
+    {
+      "epoch": 0.756556465889909,
+      "grad_norm": 2.336954355239868,
+      "learning_rate": 1.4366782426708197e-05,
+      "loss": 1.7077,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7568858330931697,
+      "grad_norm": 2.7365548610687256,
+      "learning_rate": 1.432989921603265e-05,
+      "loss": 1.9901,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7572152002964305,
+      "grad_norm": 3.0975027084350586,
+      "learning_rate": 1.4293055489656388e-05,
+      "loss": 1.7814,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7575445674996912,
+      "grad_norm": 3.5924556255340576,
+      "learning_rate": 1.4256251288362792e-05,
+      "loss": 1.5984,
+      "step": 2300
+    },
+    {
+      "epoch": 0.757873934702952,
+      "grad_norm": 2.4012112617492676,
+      "learning_rate": 1.4219486652891568e-05,
+      "loss": 2.4546,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7582033019062127,
+      "grad_norm": 2.0408058166503906,
+      "learning_rate": 1.4182761623938568e-05,
+      "loss": 1.8476,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7585326691094735,
+      "grad_norm": 2.051454782485962,
+      "learning_rate": 1.4146076242155831e-05,
+      "loss": 2.1982,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7588620363127342,
+      "grad_norm": 2.249690294265747,
+      "learning_rate": 1.4109430548151498e-05,
+      "loss": 2.0641,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7591914035159949,
+      "grad_norm": 2.241851329803467,
+      "learning_rate": 1.4072824582489785e-05,
+      "loss": 2.1071,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7595207707192556,
+      "grad_norm": 2.264829635620117,
+      "learning_rate": 1.4036258385690948e-05,
+      "loss": 2.2489,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7598501379225163,
+      "grad_norm": 2.8766753673553467,
+      "learning_rate": 1.3999731998231158e-05,
+      "loss": 2.4871,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7601795051257771,
+      "grad_norm": 2.8360114097595215,
+      "learning_rate": 1.396324546054259e-05,
+      "loss": 2.3806,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7605088723290379,
+      "grad_norm": 2.4339394569396973,
+      "learning_rate": 1.3926798813013275e-05,
+      "loss": 2.1888,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7608382395322986,
+      "grad_norm": 2.5512523651123047,
+      "learning_rate": 1.3890392095987098e-05,
+      "loss": 2.1649,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7611676067355593,
+      "grad_norm": 2.7157366275787354,
+      "learning_rate": 1.3854025349763744e-05,
+      "loss": 2.3705,
+      "step": 2311
+    },
+    {
+      "epoch": 0.76149697393882,
+      "grad_norm": 2.924532890319824,
+      "learning_rate": 1.3817698614598663e-05,
+      "loss": 2.4532,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7618263411420808,
+      "grad_norm": 2.370842456817627,
+      "learning_rate": 1.3781411930702965e-05,
+      "loss": 1.979,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7621557083453415,
+      "grad_norm": 2.5691492557525635,
+      "learning_rate": 1.3745165338243488e-05,
+      "loss": 2.0097,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7624850755486022,
+      "grad_norm": 2.5397489070892334,
+      "learning_rate": 1.370895887734266e-05,
+      "loss": 1.9783,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7628144427518629,
+      "grad_norm": 2.7631609439849854,
+      "learning_rate": 1.3672792588078497e-05,
+      "loss": 2.0406,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7631438099551238,
+      "grad_norm": 2.517392873764038,
+      "learning_rate": 1.3636666510484548e-05,
+      "loss": 1.8141,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7634731771583845,
+      "grad_norm": 2.7305877208709717,
+      "learning_rate": 1.3600580684549841e-05,
+      "loss": 2.1282,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7638025443616452,
+      "grad_norm": 2.475735902786255,
+      "learning_rate": 1.3564535150218872e-05,
+      "loss": 2.1104,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7641319115649059,
+      "grad_norm": 2.88720703125,
+      "learning_rate": 1.3528529947391466e-05,
+      "loss": 1.9543,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7644612787681666,
+      "grad_norm": 2.685481548309326,
+      "learning_rate": 1.3492565115922911e-05,
+      "loss": 2.1007,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7647906459714274,
+      "grad_norm": 3.0710129737854004,
+      "learning_rate": 1.3456640695623735e-05,
+      "loss": 1.9479,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7651200131746881,
+      "grad_norm": 3.278609037399292,
+      "learning_rate": 1.3420756726259754e-05,
+      "loss": 2.3666,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7654493803779489,
+      "grad_norm": 2.645240068435669,
+      "learning_rate": 1.3384913247552011e-05,
+      "loss": 1.8713,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7657787475812096,
+      "grad_norm": 3.5083346366882324,
+      "learning_rate": 1.3349110299176704e-05,
+      "loss": 2.1438,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7661081147844704,
+      "grad_norm": 2.030167579650879,
+      "learning_rate": 1.331334792076519e-05,
+      "loss": 2.4294,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7664374819877311,
+      "grad_norm": 2.1392946243286133,
+      "learning_rate": 1.3277626151903917e-05,
+      "loss": 2.3092,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7667668491909918,
+      "grad_norm": 2.1700847148895264,
+      "learning_rate": 1.324194503213438e-05,
+      "loss": 2.0788,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7670962163942525,
+      "grad_norm": 2.3623058795928955,
+      "learning_rate": 1.3206304600953068e-05,
+      "loss": 2.4574,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7674255835975132,
+      "grad_norm": 2.407792806625366,
+      "learning_rate": 1.3170704897811448e-05,
+      "loss": 2.4819,
+      "step": 2330
+    },
+    {
+      "epoch": 0.767754950800774,
+      "grad_norm": 2.397343397140503,
+      "learning_rate": 1.3135145962115902e-05,
+      "loss": 2.1706,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7680843180040348,
+      "grad_norm": 2.4522149562835693,
+      "learning_rate": 1.3099627833227652e-05,
+      "loss": 2.4389,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7684136852072955,
+      "grad_norm": 2.521545648574829,
+      "learning_rate": 1.3064150550462783e-05,
+      "loss": 2.5801,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7687430524105562,
+      "grad_norm": 2.5713517665863037,
+      "learning_rate": 1.3028714153092163e-05,
+      "loss": 2.1845,
+      "step": 2334
+    },
+    {
+      "epoch": 0.769072419613817,
+      "grad_norm": 2.2203288078308105,
+      "learning_rate": 1.2993318680341399e-05,
+      "loss": 1.7644,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7694017868170777,
+      "grad_norm": 2.6416571140289307,
+      "learning_rate": 1.2957964171390796e-05,
+      "loss": 2.2982,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7697311540203384,
+      "grad_norm": 2.398996591567993,
+      "learning_rate": 1.292265066537533e-05,
+      "loss": 2.0765,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7700605212235991,
+      "grad_norm": 2.396077871322632,
+      "learning_rate": 1.2887378201384542e-05,
+      "loss": 2.2264,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7703898884268598,
+      "grad_norm": 2.5502443313598633,
+      "learning_rate": 1.285214681846259e-05,
+      "loss": 1.9957,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7707192556301207,
+      "grad_norm": 2.315697431564331,
+      "learning_rate": 1.281695655560815e-05,
+      "loss": 1.898,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7710486228333814,
+      "grad_norm": 2.28928804397583,
+      "learning_rate": 1.278180745177437e-05,
+      "loss": 1.9005,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7713779900366421,
+      "grad_norm": 2.6291017532348633,
+      "learning_rate": 1.274669954586884e-05,
+      "loss": 2.1606,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7717073572399028,
+      "grad_norm": 2.399324417114258,
+      "learning_rate": 1.2711632876753549e-05,
+      "loss": 1.884,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7720367244431636,
+      "grad_norm": 2.780337333679199,
+      "learning_rate": 1.2676607483244862e-05,
+      "loss": 2.2213,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7723660916464243,
+      "grad_norm": 2.4328701496124268,
+      "learning_rate": 1.2641623404113396e-05,
+      "loss": 1.7746,
+      "step": 2345
+    },
+    {
+      "epoch": 0.772695458849685,
+      "grad_norm": 2.7607338428497314,
+      "learning_rate": 1.2606680678084087e-05,
+      "loss": 2.1983,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7730248260529458,
+      "grad_norm": 2.7279250621795654,
+      "learning_rate": 1.2571779343836082e-05,
+      "loss": 1.6927,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7733541932562065,
+      "grad_norm": 2.8370895385742188,
+      "learning_rate": 1.2536919440002715e-05,
+      "loss": 1.7246,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7736835604594673,
+      "grad_norm": 2.732421875,
+      "learning_rate": 1.2502101005171446e-05,
+      "loss": 2.0061,
+      "step": 2349
+    },
+    {
+      "epoch": 0.774012927662728,
+      "grad_norm": 3.144651412963867,
+      "learning_rate": 1.2467324077883864e-05,
+      "loss": 1.6209,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7743422948659887,
+      "grad_norm": 2.1023104190826416,
+      "learning_rate": 1.2432588696635533e-05,
+      "loss": 2.5392,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7746716620692494,
+      "grad_norm": 2.024048328399658,
+      "learning_rate": 1.2397894899876133e-05,
+      "loss": 2.1375,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7750010292725102,
+      "grad_norm": 2.4644556045532227,
+      "learning_rate": 1.2363242726009256e-05,
+      "loss": 2.4582,
+      "step": 2353
+    },
+    {
+      "epoch": 0.775330396475771,
+      "grad_norm": 2.37516450881958,
+      "learning_rate": 1.232863221339241e-05,
+      "loss": 1.9603,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7756597636790317,
+      "grad_norm": 2.3742220401763916,
+      "learning_rate": 1.2294063400337036e-05,
+      "loss": 2.2916,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7759891308822924,
+      "grad_norm": 2.2466628551483154,
+      "learning_rate": 1.2259536325108345e-05,
+      "loss": 2.0996,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7763184980855531,
+      "grad_norm": 2.6852376461029053,
+      "learning_rate": 1.2225051025925405e-05,
+      "loss": 2.3539,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7766478652888139,
+      "grad_norm": 2.818572521209717,
+      "learning_rate": 1.2190607540961025e-05,
+      "loss": 2.0653,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7769772324920746,
+      "grad_norm": 2.7895278930664062,
+      "learning_rate": 1.2156205908341728e-05,
+      "loss": 2.3245,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7773065996953353,
+      "grad_norm": 2.3493518829345703,
+      "learning_rate": 1.2121846166147699e-05,
+      "loss": 1.9228,
+      "step": 2360
+    },
+    {
+      "epoch": 0.777635966898596,
+      "grad_norm": 2.814150810241699,
+      "learning_rate": 1.2087528352412775e-05,
+      "loss": 2.0494,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7779653341018568,
+      "grad_norm": 2.2821226119995117,
+      "learning_rate": 1.2053252505124368e-05,
+      "loss": 1.8652,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7782947013051176,
+      "grad_norm": 2.916224479675293,
+      "learning_rate": 1.2019018662223425e-05,
+      "loss": 2.3056,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7786240685083783,
+      "grad_norm": 2.693669080734253,
+      "learning_rate": 1.198482686160441e-05,
+      "loss": 2.1136,
+      "step": 2364
+    },
+    {
+      "epoch": 0.778953435711639,
+      "grad_norm": 2.8778226375579834,
+      "learning_rate": 1.1950677141115246e-05,
+      "loss": 2.3815,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7792828029148997,
+      "grad_norm": 2.426630735397339,
+      "learning_rate": 1.1916569538557287e-05,
+      "loss": 2.1077,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7796121701181605,
+      "grad_norm": 2.658508539199829,
+      "learning_rate": 1.1882504091685253e-05,
+      "loss": 1.8335,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7799415373214212,
+      "grad_norm": 3.0452957153320312,
+      "learning_rate": 1.184848083820721e-05,
+      "loss": 1.9977,
+      "step": 2368
+    },
+    {
+      "epoch": 0.780270904524682,
+      "grad_norm": 2.9064278602600098,
+      "learning_rate": 1.1814499815784513e-05,
+      "loss": 2.1386,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7806002717279427,
+      "grad_norm": 2.525879144668579,
+      "learning_rate": 1.1780561062031753e-05,
+      "loss": 1.5748,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7809296389312034,
+      "grad_norm": 2.879523277282715,
+      "learning_rate": 1.1746664614516756e-05,
+      "loss": 2.3143,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7812590061344642,
+      "grad_norm": 2.849067211151123,
+      "learning_rate": 1.171281051076052e-05,
+      "loss": 1.8855,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7815883733377249,
+      "grad_norm": 2.7838194370269775,
+      "learning_rate": 1.1678998788237155e-05,
+      "loss": 1.7152,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7819177405409856,
+      "grad_norm": 2.8870933055877686,
+      "learning_rate": 1.164522948437387e-05,
+      "loss": 1.693,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7822471077442463,
+      "grad_norm": 3.9270734786987305,
+      "learning_rate": 1.1611502636550931e-05,
+      "loss": 1.7755,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7825764749475072,
+      "grad_norm": 1.9541338682174683,
+      "learning_rate": 1.157781828210156e-05,
+      "loss": 2.5028,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7829058421507679,
+      "grad_norm": 2.4501519203186035,
+      "learning_rate": 1.1544176458311996e-05,
+      "loss": 2.7866,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7832352093540286,
+      "grad_norm": 2.14868426322937,
+      "learning_rate": 1.1510577202421369e-05,
+      "loss": 2.3476,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7835645765572893,
+      "grad_norm": 2.618394374847412,
+      "learning_rate": 1.1477020551621698e-05,
+      "loss": 2.5685,
+      "step": 2379
+    },
+    {
+      "epoch": 0.78389394376055,
+      "grad_norm": 2.1860313415527344,
+      "learning_rate": 1.1443506543057852e-05,
+      "loss": 1.9295,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7842233109638108,
+      "grad_norm": 2.496148109436035,
+      "learning_rate": 1.141003521382748e-05,
+      "loss": 2.3171,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7845526781670715,
+      "grad_norm": 2.7145514488220215,
+      "learning_rate": 1.1376606600980993e-05,
+      "loss": 2.456,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7848820453703322,
+      "grad_norm": 2.3949971199035645,
+      "learning_rate": 1.1343220741521526e-05,
+      "loss": 2.1141,
+      "step": 2383
+    },
+    {
+      "epoch": 0.785211412573593,
+      "grad_norm": 2.371872901916504,
+      "learning_rate": 1.1309877672404878e-05,
+      "loss": 2.0841,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7855407797768538,
+      "grad_norm": 2.2183234691619873,
+      "learning_rate": 1.1276577430539492e-05,
+      "loss": 2.1255,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7858701469801145,
+      "grad_norm": 2.644172430038452,
+      "learning_rate": 1.12433200527864e-05,
+      "loss": 2.5442,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7861995141833752,
+      "grad_norm": 2.8008968830108643,
+      "learning_rate": 1.1210105575959195e-05,
+      "loss": 2.4805,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7865288813866359,
+      "grad_norm": 3.0009310245513916,
+      "learning_rate": 1.1176934036823949e-05,
+      "loss": 2.406,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7868582485898966,
+      "grad_norm": 2.7110002040863037,
+      "learning_rate": 1.1143805472099245e-05,
+      "loss": 2.2543,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7871876157931574,
+      "grad_norm": 2.5703489780426025,
+      "learning_rate": 1.1110719918456075e-05,
+      "loss": 1.7567,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7875169829964181,
+      "grad_norm": 2.9390273094177246,
+      "learning_rate": 1.1077677412517828e-05,
+      "loss": 2.1317,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7878463501996789,
+      "grad_norm": 2.4334805011749268,
+      "learning_rate": 1.1044677990860236e-05,
+      "loss": 2.1525,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7881757174029396,
+      "grad_norm": 2.6386754512786865,
+      "learning_rate": 1.1011721690011368e-05,
+      "loss": 2.2738,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7885050846062004,
+      "grad_norm": 2.390347957611084,
+      "learning_rate": 1.0978808546451503e-05,
+      "loss": 2.0497,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7888344518094611,
+      "grad_norm": 3.1142003536224365,
+      "learning_rate": 1.0945938596613193e-05,
+      "loss": 2.1531,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7891638190127218,
+      "grad_norm": 3.1742782592773438,
+      "learning_rate": 1.0913111876881166e-05,
+      "loss": 2.3082,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7894931862159825,
+      "grad_norm": 3.1472554206848145,
+      "learning_rate": 1.0880328423592307e-05,
+      "loss": 2.0666,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7898225534192432,
+      "grad_norm": 2.8876795768737793,
+      "learning_rate": 1.0847588273035592e-05,
+      "loss": 1.9119,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7901519206225041,
+      "grad_norm": 3.2102115154266357,
+      "learning_rate": 1.081489146145207e-05,
+      "loss": 2.1106,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7904812878257648,
+      "grad_norm": 3.5183258056640625,
+      "learning_rate": 1.0782238025034835e-05,
+      "loss": 2.0614,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7908106550290255,
+      "grad_norm": 2.160437822341919,
+      "learning_rate": 1.0749627999928923e-05,
+      "loss": 2.2165,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7911400222322862,
+      "grad_norm": 2.1976757049560547,
+      "learning_rate": 1.0717061422231361e-05,
+      "loss": 1.9749,
+      "step": 2402
+    },
+    {
+      "epoch": 0.791469389435547,
+      "grad_norm": 2.3027257919311523,
+      "learning_rate": 1.0684538327991055e-05,
+      "loss": 2.4518,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7917987566388077,
+      "grad_norm": 2.299556255340576,
+      "learning_rate": 1.06520587532088e-05,
+      "loss": 2.4594,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7921281238420684,
+      "grad_norm": 2.538569211959839,
+      "learning_rate": 1.0619622733837198e-05,
+      "loss": 2.3555,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7924574910453291,
+      "grad_norm": 2.4368882179260254,
+      "learning_rate": 1.0587230305780666e-05,
+      "loss": 2.3176,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7927868582485899,
+      "grad_norm": 2.313861131668091,
+      "learning_rate": 1.0554881504895326e-05,
+      "loss": 2.4077,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7931162254518507,
+      "grad_norm": 2.871065139770508,
+      "learning_rate": 1.0522576366989035e-05,
+      "loss": 2.0461,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7934455926551114,
+      "grad_norm": 2.8059747219085693,
+      "learning_rate": 1.0490314927821315e-05,
+      "loss": 2.4673,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7937749598583721,
+      "grad_norm": 2.2953708171844482,
+      "learning_rate": 1.0458097223103331e-05,
+      "loss": 2.0045,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7941043270616328,
+      "grad_norm": 2.41961407661438,
+      "learning_rate": 1.0425923288497801e-05,
+      "loss": 2.2286,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7944336942648935,
+      "grad_norm": 2.5134241580963135,
+      "learning_rate": 1.039379315961903e-05,
+      "loss": 2.2415,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7947630614681543,
+      "grad_norm": 2.3589727878570557,
+      "learning_rate": 1.0361706872032812e-05,
+      "loss": 1.9094,
+      "step": 2413
+    },
+    {
+      "epoch": 0.795092428671415,
+      "grad_norm": 2.9635963439941406,
+      "learning_rate": 1.0329664461256411e-05,
+      "loss": 2.3048,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7954217958746758,
+      "grad_norm": 2.931699275970459,
+      "learning_rate": 1.0297665962758535e-05,
+      "loss": 2.3376,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7957511630779365,
+      "grad_norm": 2.3775904178619385,
+      "learning_rate": 1.0265711411959272e-05,
+      "loss": 1.9315,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7960805302811973,
+      "grad_norm": 3.018141031265259,
+      "learning_rate": 1.0233800844230073e-05,
+      "loss": 2.0433,
+      "step": 2417
+    },
+    {
+      "epoch": 0.796409897484458,
+      "grad_norm": 2.5099568367004395,
+      "learning_rate": 1.0201934294893705e-05,
+      "loss": 2.4133,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7967392646877187,
+      "grad_norm": 2.779721975326538,
+      "learning_rate": 1.0170111799224186e-05,
+      "loss": 2.1728,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7970686318909794,
+      "grad_norm": 2.5381271839141846,
+      "learning_rate": 1.0138333392446786e-05,
+      "loss": 1.7606,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7973979990942401,
+      "grad_norm": 2.808448314666748,
+      "learning_rate": 1.010659910973798e-05,
+      "loss": 1.9098,
+      "step": 2421
+    },
+    {
+      "epoch": 0.797727366297501,
+      "grad_norm": 2.90604829788208,
+      "learning_rate": 1.0074908986225395e-05,
+      "loss": 2.0743,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7980567335007617,
+      "grad_norm": 2.6868839263916016,
+      "learning_rate": 1.0043263056987773e-05,
+      "loss": 2.0367,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7983861007040224,
+      "grad_norm": 3.5458996295928955,
+      "learning_rate": 1.001166135705493e-05,
+      "loss": 2.1114,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7987154679072831,
+      "grad_norm": 3.005037784576416,
+      "learning_rate": 9.980103921407757e-06,
+      "loss": 1.6499,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7990448351105439,
+      "grad_norm": 2.083054542541504,
+      "learning_rate": 9.948590784978085e-06,
+      "loss": 2.5122,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7993742023138046,
+      "grad_norm": 2.21122670173645,
+      "learning_rate": 9.917121982648764e-06,
+      "loss": 2.3761,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7997035695170653,
+      "grad_norm": 1.965364694595337,
+      "learning_rate": 9.885697549253542e-06,
+      "loss": 2.0381,
+      "step": 2428
+    },
+    {
+      "epoch": 0.800032936720326,
+      "grad_norm": 2.5739123821258545,
+      "learning_rate": 9.854317519577077e-06,
+      "loss": 2.4915,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8003623039235868,
+      "grad_norm": 2.479092597961426,
+      "learning_rate": 9.822981928354846e-06,
+      "loss": 2.3367,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8006916711268476,
+      "grad_norm": 2.881108522415161,
+      "learning_rate": 9.791690810273173e-06,
+      "loss": 2.369,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8010210383301083,
+      "grad_norm": 2.5094995498657227,
+      "learning_rate": 9.760444199969104e-06,
+      "loss": 2.4964,
+      "step": 2432
+    },
+    {
+      "epoch": 0.801350405533369,
+      "grad_norm": 2.3330986499786377,
+      "learning_rate": 9.729242132030452e-06,
+      "loss": 1.9987,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8016797727366297,
+      "grad_norm": 2.8625338077545166,
+      "learning_rate": 9.698084640995719e-06,
+      "loss": 2.0433,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8020091399398905,
+      "grad_norm": 2.96150541305542,
+      "learning_rate": 9.666971761354065e-06,
+      "loss": 2.4458,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8023385071431512,
+      "grad_norm": 3.234882354736328,
+      "learning_rate": 9.635903527545264e-06,
+      "loss": 2.2081,
+      "step": 2436
+    },
+    {
+      "epoch": 0.802667874346412,
+      "grad_norm": 2.900803804397583,
+      "learning_rate": 9.604879973959668e-06,
+      "loss": 2.3133,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8029972415496727,
+      "grad_norm": 2.271087646484375,
+      "learning_rate": 9.57390113493819e-06,
+      "loss": 2.052,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8033266087529334,
+      "grad_norm": 2.268280267715454,
+      "learning_rate": 9.542967044772205e-06,
+      "loss": 1.9981,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8036559759561942,
+      "grad_norm": 2.6638338565826416,
+      "learning_rate": 9.512077737703601e-06,
+      "loss": 1.9637,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8039853431594549,
+      "grad_norm": 2.6653592586517334,
+      "learning_rate": 9.481233247924658e-06,
+      "loss": 1.9678,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8043147103627156,
+      "grad_norm": 2.4272871017456055,
+      "learning_rate": 9.450433609578064e-06,
+      "loss": 2.0047,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8046440775659763,
+      "grad_norm": 2.5219974517822266,
+      "learning_rate": 9.419678856756891e-06,
+      "loss": 2.3667,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8049734447692372,
+      "grad_norm": 2.637106418609619,
+      "learning_rate": 9.388969023504457e-06,
+      "loss": 2.0492,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8053028119724979,
+      "grad_norm": 2.4616830348968506,
+      "learning_rate": 9.358304143814401e-06,
+      "loss": 2.2313,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8056321791757586,
+      "grad_norm": 2.7604615688323975,
+      "learning_rate": 9.327684251630597e-06,
+      "loss": 2.0396,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8059615463790193,
+      "grad_norm": 2.6307003498077393,
+      "learning_rate": 9.297109380847119e-06,
+      "loss": 1.9813,
+      "step": 2447
+    },
+    {
+      "epoch": 0.80629091358228,
+      "grad_norm": 3.0449466705322266,
+      "learning_rate": 9.266579565308198e-06,
+      "loss": 2.2011,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8066202807855408,
+      "grad_norm": 3.0230000019073486,
+      "learning_rate": 9.236094838808206e-06,
+      "loss": 1.7364,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8069496479888015,
+      "grad_norm": 3.4453773498535156,
+      "learning_rate": 9.205655235091603e-06,
+      "loss": 2.0221,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8072790151920622,
+      "grad_norm": 1.9125607013702393,
+      "learning_rate": 9.175260787852873e-06,
+      "loss": 2.2074,
+      "step": 2451
+    },
+    {
+      "epoch": 0.807608382395323,
+      "grad_norm": 2.63358736038208,
+      "learning_rate": 9.14491153073655e-06,
+      "loss": 2.5471,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8079377495985838,
+      "grad_norm": 2.265695095062256,
+      "learning_rate": 9.114607497337135e-06,
+      "loss": 2.2188,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8082671168018445,
+      "grad_norm": 2.2626595497131348,
+      "learning_rate": 9.084348721199059e-06,
+      "loss": 2.2639,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8085964840051052,
+      "grad_norm": 2.104132652282715,
+      "learning_rate": 9.054135235816669e-06,
+      "loss": 2.286,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8089258512083659,
+      "grad_norm": 2.395782709121704,
+      "learning_rate": 9.023967074634187e-06,
+      "loss": 2.1671,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8092552184116266,
+      "grad_norm": 2.4479575157165527,
+      "learning_rate": 8.993844271045626e-06,
+      "loss": 2.2154,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8095845856148874,
+      "grad_norm": 2.32531476020813,
+      "learning_rate": 8.963766858394823e-06,
+      "loss": 2.0973,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8099139528181482,
+      "grad_norm": 2.2058169841766357,
+      "learning_rate": 8.933734869975374e-06,
+      "loss": 2.2149,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8102433200214089,
+      "grad_norm": 2.1951229572296143,
+      "learning_rate": 8.903748339030582e-06,
+      "loss": 2.1389,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8105726872246696,
+      "grad_norm": 2.675092935562134,
+      "learning_rate": 8.873807298753422e-06,
+      "loss": 2.2773,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8109020544279303,
+      "grad_norm": 2.5307810306549072,
+      "learning_rate": 8.84391178228654e-06,
+      "loss": 2.0614,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8112314216311911,
+      "grad_norm": 2.554347038269043,
+      "learning_rate": 8.814061822722174e-06,
+      "loss": 2.6156,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8115607888344518,
+      "grad_norm": 2.7959494590759277,
+      "learning_rate": 8.784257453102124e-06,
+      "loss": 1.8778,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8118901560377125,
+      "grad_norm": 3.0916404724121094,
+      "learning_rate": 8.754498706417741e-06,
+      "loss": 2.5136,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8122195232409732,
+      "grad_norm": 2.7095844745635986,
+      "learning_rate": 8.724785615609865e-06,
+      "loss": 2.1074,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8125488904442341,
+      "grad_norm": 2.490854263305664,
+      "learning_rate": 8.695118213568816e-06,
+      "loss": 2.132,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8128782576474948,
+      "grad_norm": 2.495551824569702,
+      "learning_rate": 8.665496533134315e-06,
+      "loss": 1.9723,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8132076248507555,
+      "grad_norm": 2.5706124305725098,
+      "learning_rate": 8.635920607095504e-06,
+      "loss": 2.1114,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8135369920540162,
+      "grad_norm": 3.406837224960327,
+      "learning_rate": 8.606390468190833e-06,
+      "loss": 2.4752,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8138663592572769,
+      "grad_norm": 2.7835068702697754,
+      "learning_rate": 8.576906149108104e-06,
+      "loss": 1.7684,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8141957264605377,
+      "grad_norm": 2.842190742492676,
+      "learning_rate": 8.547467682484389e-06,
+      "loss": 1.7345,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8145250936637984,
+      "grad_norm": 3.7220804691314697,
+      "learning_rate": 8.518075100905992e-06,
+      "loss": 2.0355,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8148544608670591,
+      "grad_norm": 3.133759021759033,
+      "learning_rate": 8.488728436908483e-06,
+      "loss": 1.9687,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8151838280703199,
+      "grad_norm": 3.0628323554992676,
+      "learning_rate": 8.45942772297652e-06,
+      "loss": 1.7354,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8155131952735807,
+      "grad_norm": 2.066558599472046,
+      "learning_rate": 8.430172991543944e-06,
+      "loss": 2.533,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8158425624768414,
+      "grad_norm": 2.2155539989471436,
+      "learning_rate": 8.4009642749937e-06,
+      "loss": 2.5765,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8161719296801021,
+      "grad_norm": 2.202932834625244,
+      "learning_rate": 8.371801605657781e-06,
+      "loss": 2.2965,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8165012968833628,
+      "grad_norm": 2.7432045936584473,
+      "learning_rate": 8.34268501581722e-06,
+      "loss": 2.5274,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8168306640866235,
+      "grad_norm": 2.2428295612335205,
+      "learning_rate": 8.313614537702042e-06,
+      "loss": 2.0618,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8171600312898843,
+      "grad_norm": 2.3830106258392334,
+      "learning_rate": 8.284590203491232e-06,
+      "loss": 2.2108,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8174893984931451,
+      "grad_norm": 2.368865966796875,
+      "learning_rate": 8.255612045312667e-06,
+      "loss": 2.1992,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8178187656964058,
+      "grad_norm": 2.4614439010620117,
+      "learning_rate": 8.226680095243155e-06,
+      "loss": 2.1652,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8181481328996665,
+      "grad_norm": 2.3410730361938477,
+      "learning_rate": 8.197794385308333e-06,
+      "loss": 1.9139,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8184775001029273,
+      "grad_norm": 2.664398431777954,
+      "learning_rate": 8.168954947482654e-06,
+      "loss": 2.5085,
+      "step": 2485
+    },
+    {
+      "epoch": 0.818806867306188,
+      "grad_norm": 2.8934290409088135,
+      "learning_rate": 8.140161813689352e-06,
+      "loss": 2.2249,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8191362345094487,
+      "grad_norm": 2.2794628143310547,
+      "learning_rate": 8.111415015800406e-06,
+      "loss": 1.8115,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8194656017127094,
+      "grad_norm": 2.8373513221740723,
+      "learning_rate": 8.08271458563652e-06,
+      "loss": 2.2524,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8197949689159701,
+      "grad_norm": 3.0686585903167725,
+      "learning_rate": 8.054060554967024e-06,
+      "loss": 2.0289,
+      "step": 2489
+    },
+    {
+      "epoch": 0.820124336119231,
+      "grad_norm": 2.923983097076416,
+      "learning_rate": 8.025452955509943e-06,
+      "loss": 2.1504,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8204537033224917,
+      "grad_norm": 2.634913682937622,
+      "learning_rate": 7.996891818931879e-06,
+      "loss": 2.0277,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8207830705257524,
+      "grad_norm": 2.4638595581054688,
+      "learning_rate": 7.968377176848003e-06,
+      "loss": 2.1846,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8211124377290131,
+      "grad_norm": 2.867859363555908,
+      "learning_rate": 7.939909060822025e-06,
+      "loss": 2.0976,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8214418049322739,
+      "grad_norm": 2.6945831775665283,
+      "learning_rate": 7.911487502366166e-06,
+      "loss": 2.3166,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8217711721355346,
+      "grad_norm": 2.430236577987671,
+      "learning_rate": 7.883112532941072e-06,
+      "loss": 2.1648,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8221005393387953,
+      "grad_norm": 2.43375563621521,
+      "learning_rate": 7.85478418395586e-06,
+      "loss": 2.2089,
+      "step": 2496
+    },
+    {
+      "epoch": 0.822429906542056,
+      "grad_norm": 2.8123996257781982,
+      "learning_rate": 7.826502486768018e-06,
+      "loss": 2.2756,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8227592737453168,
+      "grad_norm": 2.7135322093963623,
+      "learning_rate": 7.798267472683407e-06,
+      "loss": 2.0788,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8230886409485776,
+      "grad_norm": 2.929687261581421,
+      "learning_rate": 7.770079172956201e-06,
+      "loss": 1.9092,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8234180081518383,
+      "grad_norm": 3.2018990516662598,
+      "learning_rate": 7.741937618788875e-06,
+      "loss": 1.9998,
+      "step": 2500
+    },
+    {
+      "epoch": 0.823747375355099,
+      "grad_norm": 2.0544869899749756,
+      "learning_rate": 7.713842841332164e-06,
+      "loss": 2.3543,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8240767425583597,
+      "grad_norm": 2.4761202335357666,
+      "learning_rate": 7.685794871684998e-06,
+      "loss": 2.6125,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8244061097616204,
+      "grad_norm": 2.063749074935913,
+      "learning_rate": 7.657793740894504e-06,
+      "loss": 2.2784,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8247354769648813,
+      "grad_norm": 2.2412142753601074,
+      "learning_rate": 7.629839479955998e-06,
+      "loss": 2.3776,
+      "step": 2504
+    },
+    {
+      "epoch": 0.825064844168142,
+      "grad_norm": 2.3341736793518066,
+      "learning_rate": 7.6019321198128655e-06,
+      "loss": 2.4814,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8253942113714027,
+      "grad_norm": 2.342841625213623,
+      "learning_rate": 7.574071691356621e-06,
+      "loss": 2.3653,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8257235785746634,
+      "grad_norm": 2.213564157485962,
+      "learning_rate": 7.546258225426778e-06,
+      "loss": 2.2279,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8260529457779242,
+      "grad_norm": 2.4937024116516113,
+      "learning_rate": 7.518491752810897e-06,
+      "loss": 2.4755,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8263823129811849,
+      "grad_norm": 2.3210904598236084,
+      "learning_rate": 7.49077230424452e-06,
+      "loss": 1.9111,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8267116801844456,
+      "grad_norm": 2.904571771621704,
+      "learning_rate": 7.463099910411137e-06,
+      "loss": 2.7039,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8270410473877063,
+      "grad_norm": 2.4512391090393066,
+      "learning_rate": 7.435474601942133e-06,
+      "loss": 2.1503,
+      "step": 2511
+    },
+    {
+      "epoch": 0.827370414590967,
+      "grad_norm": 2.329566240310669,
+      "learning_rate": 7.407896409416809e-06,
+      "loss": 2.0849,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8276997817942279,
+      "grad_norm": 2.870210886001587,
+      "learning_rate": 7.380365363362263e-06,
+      "loss": 1.9724,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8280291489974886,
+      "grad_norm": 2.983409881591797,
+      "learning_rate": 7.3528814942534445e-06,
+      "loss": 2.4121,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8283585162007493,
+      "grad_norm": 2.7085869312286377,
+      "learning_rate": 7.3254448325130675e-06,
+      "loss": 1.9919,
+      "step": 2515
+    },
+    {
+      "epoch": 0.82868788340401,
+      "grad_norm": 2.715024948120117,
+      "learning_rate": 7.298055408511595e-06,
+      "loss": 2.0174,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8290172506072708,
+      "grad_norm": 2.552612066268921,
+      "learning_rate": 7.270713252567191e-06,
+      "loss": 2.3856,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8293466178105315,
+      "grad_norm": 2.6823573112487793,
+      "learning_rate": 7.243418394945711e-06,
+      "loss": 1.9564,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8296759850137922,
+      "grad_norm": 2.3931643962860107,
+      "learning_rate": 7.216170865860655e-06,
+      "loss": 1.6479,
+      "step": 2519
+    },
+    {
+      "epoch": 0.830005352217053,
+      "grad_norm": 2.824610710144043,
+      "learning_rate": 7.188970695473107e-06,
+      "loss": 2.1261,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8303347194203137,
+      "grad_norm": 2.481971502304077,
+      "learning_rate": 7.161817913891755e-06,
+      "loss": 1.9109,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8306640866235745,
+      "grad_norm": 2.6014606952667236,
+      "learning_rate": 7.134712551172823e-06,
+      "loss": 2.1093,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8309934538268352,
+      "grad_norm": 3.1851959228515625,
+      "learning_rate": 7.107654637320049e-06,
+      "loss": 2.0577,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8313228210300959,
+      "grad_norm": 3.0487592220306396,
+      "learning_rate": 7.080644202284648e-06,
+      "loss": 1.911,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8316521882333566,
+      "grad_norm": 3.059511423110962,
+      "learning_rate": 7.0536812759652684e-06,
+      "loss": 2.0921,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8319815554366174,
+      "grad_norm": 2.1465368270874023,
+      "learning_rate": 7.0267658882079976e-06,
+      "loss": 2.5033,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8323109226398782,
+      "grad_norm": 2.2625226974487305,
+      "learning_rate": 6.999898068806254e-06,
+      "loss": 2.1521,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8326402898431389,
+      "grad_norm": 2.0414772033691406,
+      "learning_rate": 6.973077847500842e-06,
+      "loss": 2.1878,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8329696570463996,
+      "grad_norm": 2.8698439598083496,
+      "learning_rate": 6.946305253979862e-06,
+      "loss": 2.1674,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8332990242496603,
+      "grad_norm": 2.2601876258850098,
+      "learning_rate": 6.919580317878705e-06,
+      "loss": 2.2299,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8336283914529211,
+      "grad_norm": 2.2976324558258057,
+      "learning_rate": 6.892903068779993e-06,
+      "loss": 2.0683,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8339577586561818,
+      "grad_norm": 2.5393729209899902,
+      "learning_rate": 6.866273536213586e-06,
+      "loss": 2.0545,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8342871258594425,
+      "grad_norm": 2.5304923057556152,
+      "learning_rate": 6.83969174965648e-06,
+      "loss": 2.1273,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8346164930627032,
+      "grad_norm": 2.707395076751709,
+      "learning_rate": 6.813157738532855e-06,
+      "loss": 2.2007,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8349458602659641,
+      "grad_norm": 2.2538604736328125,
+      "learning_rate": 6.7866715322140175e-06,
+      "loss": 2.1026,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8352752274692248,
+      "grad_norm": 2.6380133628845215,
+      "learning_rate": 6.760233160018326e-06,
+      "loss": 2.3277,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8356045946724855,
+      "grad_norm": 2.7567970752716064,
+      "learning_rate": 6.73384265121122e-06,
+      "loss": 2.1349,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8359339618757462,
+      "grad_norm": 2.683866024017334,
+      "learning_rate": 6.707500035005115e-06,
+      "loss": 2.1828,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8362633290790069,
+      "grad_norm": 2.5327346324920654,
+      "learning_rate": 6.68120534055946e-06,
+      "loss": 2.1023,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8365926962822677,
+      "grad_norm": 2.509141206741333,
+      "learning_rate": 6.654958596980626e-06,
+      "loss": 2.2115,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8369220634855284,
+      "grad_norm": 2.387683629989624,
+      "learning_rate": 6.628759833321918e-06,
+      "loss": 2.0021,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8372514306887892,
+      "grad_norm": 2.726041078567505,
+      "learning_rate": 6.602609078583538e-06,
+      "loss": 2.3552,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8375807978920499,
+      "grad_norm": 2.4618892669677734,
+      "learning_rate": 6.576506361712526e-06,
+      "loss": 1.9296,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8379101650953107,
+      "grad_norm": 2.628038167953491,
+      "learning_rate": 6.550451711602778e-06,
+      "loss": 1.9976,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8382395322985714,
+      "grad_norm": 2.565180540084839,
+      "learning_rate": 6.524445157094938e-06,
+      "loss": 2.2082,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8385688995018321,
+      "grad_norm": 2.320673942565918,
+      "learning_rate": 6.4984867269764545e-06,
+      "loss": 1.7515,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8388982667050928,
+      "grad_norm": 2.666262626647949,
+      "learning_rate": 6.472576449981477e-06,
+      "loss": 1.8173,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8392276339083535,
+      "grad_norm": 2.7199695110321045,
+      "learning_rate": 6.446714354790873e-06,
+      "loss": 1.7931,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8395570011116144,
+      "grad_norm": 2.7355263233184814,
+      "learning_rate": 6.420900470032165e-06,
+      "loss": 1.698,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8398863683148751,
+      "grad_norm": 3.240067481994629,
+      "learning_rate": 6.395134824279525e-06,
+      "loss": 1.9727,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8402157355181358,
+      "grad_norm": 3.454439640045166,
+      "learning_rate": 6.369417446053694e-06,
+      "loss": 2.2517,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8405451027213965,
+      "grad_norm": 2.324172019958496,
+      "learning_rate": 6.343748363822011e-06,
+      "loss": 2.3305,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8408744699246572,
+      "grad_norm": 2.483888626098633,
+      "learning_rate": 6.3181276059983586e-06,
+      "loss": 2.2321,
+      "step": 2553
+    },
+    {
+      "epoch": 0.841203837127918,
+      "grad_norm": 2.10294508934021,
+      "learning_rate": 6.292555200943107e-06,
+      "loss": 2.2044,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8415332043311787,
+      "grad_norm": 2.234780788421631,
+      "learning_rate": 6.267031176963123e-06,
+      "loss": 2.0835,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8418625715344394,
+      "grad_norm": 2.515031576156616,
+      "learning_rate": 6.241555562311712e-06,
+      "loss": 2.3331,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8421919387377002,
+      "grad_norm": 2.2735471725463867,
+      "learning_rate": 6.2161283851885966e-06,
+      "loss": 2.2742,
+      "step": 2557
+    },
+    {
+      "epoch": 0.842521305940961,
+      "grad_norm": 3.118337869644165,
+      "learning_rate": 6.1907496737398595e-06,
+      "loss": 2.5658,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8428506731442217,
+      "grad_norm": 2.6256306171417236,
+      "learning_rate": 6.165419456057964e-06,
+      "loss": 2.1742,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8431800403474824,
+      "grad_norm": 2.4760029315948486,
+      "learning_rate": 6.140137760181686e-06,
+      "loss": 2.1152,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8435094075507431,
+      "grad_norm": 2.9300849437713623,
+      "learning_rate": 6.11490461409609e-06,
+      "loss": 2.5476,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8438387747540038,
+      "grad_norm": 2.8328449726104736,
+      "learning_rate": 6.089720045732495e-06,
+      "loss": 2.1809,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8441681419572646,
+      "grad_norm": 2.6135897636413574,
+      "learning_rate": 6.06458408296845e-06,
+      "loss": 2.5915,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8444975091605254,
+      "grad_norm": 2.574871301651001,
+      "learning_rate": 6.039496753627711e-06,
+      "loss": 1.9764,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8448268763637861,
+      "grad_norm": 2.5949325561523438,
+      "learning_rate": 6.014458085480179e-06,
+      "loss": 2.1941,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8451562435670468,
+      "grad_norm": 2.195402145385742,
+      "learning_rate": 5.989468106241919e-06,
+      "loss": 2.0671,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8454856107703076,
+      "grad_norm": 2.7210750579833984,
+      "learning_rate": 5.964526843575069e-06,
+      "loss": 1.946,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8458149779735683,
+      "grad_norm": 2.7602195739746094,
+      "learning_rate": 5.9396343250878676e-06,
+      "loss": 1.8686,
+      "step": 2568
+    },
+    {
+      "epoch": 0.846144345176829,
+      "grad_norm": 2.3747804164886475,
+      "learning_rate": 5.914790578334595e-06,
+      "loss": 2.1268,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8464737123800897,
+      "grad_norm": 2.572371006011963,
+      "learning_rate": 5.889995630815515e-06,
+      "loss": 1.9882,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8468030795833504,
+      "grad_norm": 2.7990000247955322,
+      "learning_rate": 5.8652495099769e-06,
+      "loss": 2.1318,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8471324467866113,
+      "grad_norm": 2.5445713996887207,
+      "learning_rate": 5.840552243210978e-06,
+      "loss": 1.7079,
+      "step": 2572
+    },
+    {
+      "epoch": 0.847461813989872,
+      "grad_norm": 2.591435432434082,
+      "learning_rate": 5.8159038578558866e-06,
+      "loss": 1.7843,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8477911811931327,
+      "grad_norm": 2.6687839031219482,
+      "learning_rate": 5.791304381195662e-06,
+      "loss": 1.8852,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8481205483963934,
+      "grad_norm": 3.511777877807617,
+      "learning_rate": 5.766753840460204e-06,
+      "loss": 1.7108,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8484499155996542,
+      "grad_norm": 1.7083003520965576,
+      "learning_rate": 5.742252262825221e-06,
+      "loss": 2.2711,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8487792828029149,
+      "grad_norm": 2.8398375511169434,
+      "learning_rate": 5.7177996754122496e-06,
+      "loss": 2.6802,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8491086500061756,
+      "grad_norm": 2.1514852046966553,
+      "learning_rate": 5.693396105288595e-06,
+      "loss": 2.4099,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8494380172094363,
+      "grad_norm": 2.346045970916748,
+      "learning_rate": 5.669041579467293e-06,
+      "loss": 2.4432,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8497673844126971,
+      "grad_norm": 2.4979259967803955,
+      "learning_rate": 5.644736124907097e-06,
+      "loss": 2.3092,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8500967516159579,
+      "grad_norm": 2.5898611545562744,
+      "learning_rate": 5.620479768512432e-06,
+      "loss": 2.5798,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8504261188192186,
+      "grad_norm": 2.240121364593506,
+      "learning_rate": 5.596272537133402e-06,
+      "loss": 2.3491,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8507554860224793,
+      "grad_norm": 2.3242783546447754,
+      "learning_rate": 5.572114457565692e-06,
+      "loss": 2.1093,
+      "step": 2583
+    },
+    {
+      "epoch": 0.85108485322574,
+      "grad_norm": 2.1993753910064697,
+      "learning_rate": 5.5480055565506116e-06,
+      "loss": 2.0687,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8514142204290008,
+      "grad_norm": 2.7326152324676514,
+      "learning_rate": 5.52394586077502e-06,
+      "loss": 2.3334,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8517435876322615,
+      "grad_norm": 2.5875439643859863,
+      "learning_rate": 5.499935396871314e-06,
+      "loss": 2.3656,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8520729548355223,
+      "grad_norm": 2.634803533554077,
+      "learning_rate": 5.475974191417399e-06,
+      "loss": 2.3282,
+      "step": 2587
+    },
+    {
+      "epoch": 0.852402322038783,
+      "grad_norm": 2.4338207244873047,
+      "learning_rate": 5.452062270936648e-06,
+      "loss": 2.0295,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8527316892420437,
+      "grad_norm": 2.489284038543701,
+      "learning_rate": 5.4281996618978646e-06,
+      "loss": 1.9639,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8530610564453045,
+      "grad_norm": 2.8681483268737793,
+      "learning_rate": 5.404386390715293e-06,
+      "loss": 2.1408,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8533904236485652,
+      "grad_norm": 2.817866086959839,
+      "learning_rate": 5.3806224837485545e-06,
+      "loss": 2.255,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8537197908518259,
+      "grad_norm": 2.494019031524658,
+      "learning_rate": 5.356907967302627e-06,
+      "loss": 1.9273,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8540491580550866,
+      "grad_norm": 2.7966413497924805,
+      "learning_rate": 5.333242867627814e-06,
+      "loss": 2.5284,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8543785252583475,
+      "grad_norm": 2.6782631874084473,
+      "learning_rate": 5.309627210919721e-06,
+      "loss": 2.0938,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8547078924616082,
+      "grad_norm": 2.4979779720306396,
+      "learning_rate": 5.286061023319228e-06,
+      "loss": 2.2774,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8550372596648689,
+      "grad_norm": 2.979830026626587,
+      "learning_rate": 5.262544330912445e-06,
+      "loss": 2.1949,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8553666268681296,
+      "grad_norm": 2.6444127559661865,
+      "learning_rate": 5.23907715973071e-06,
+      "loss": 1.8533,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8556959940713903,
+      "grad_norm": 2.309861183166504,
+      "learning_rate": 5.215659535750522e-06,
+      "loss": 1.8244,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8560253612746511,
+      "grad_norm": 2.973512649536133,
+      "learning_rate": 5.192291484893563e-06,
+      "loss": 1.8739,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8563547284779118,
+      "grad_norm": 3.404737710952759,
+      "learning_rate": 5.16897303302662e-06,
+      "loss": 1.8598,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8566840956811725,
+      "grad_norm": 2.0394158363342285,
+      "learning_rate": 5.145704205961577e-06,
+      "loss": 2.415,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8570134628844333,
+      "grad_norm": 1.9012643098831177,
+      "learning_rate": 5.122485029455398e-06,
+      "loss": 2.1571,
+      "step": 2602
+    },
+    {
+      "epoch": 0.857342830087694,
+      "grad_norm": 2.2303390502929688,
+      "learning_rate": 5.099315529210081e-06,
+      "loss": 2.3392,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8576721972909548,
+      "grad_norm": 2.5212888717651367,
+      "learning_rate": 5.076195730872641e-06,
+      "loss": 2.338,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8580015644942155,
+      "grad_norm": 2.758668899536133,
+      "learning_rate": 5.053125660035068e-06,
+      "loss": 2.2201,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8583309316974762,
+      "grad_norm": 2.390583038330078,
+      "learning_rate": 5.030105342234314e-06,
+      "loss": 2.4438,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8586602989007369,
+      "grad_norm": 2.5289313793182373,
+      "learning_rate": 5.007134802952274e-06,
+      "loss": 2.1411,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8589896661039977,
+      "grad_norm": 2.3819549083709717,
+      "learning_rate": 4.984214067615689e-06,
+      "loss": 2.1277,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8593190333072585,
+      "grad_norm": 2.2808547019958496,
+      "learning_rate": 4.961343161596227e-06,
+      "loss": 2.2532,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8596484005105192,
+      "grad_norm": 2.3810598850250244,
+      "learning_rate": 4.938522110210369e-06,
+      "loss": 2.3429,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8599777677137799,
+      "grad_norm": 2.7011477947235107,
+      "learning_rate": 4.915750938719421e-06,
+      "loss": 2.31,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8603071349170406,
+      "grad_norm": 2.4542653560638428,
+      "learning_rate": 4.893029672329474e-06,
+      "loss": 2.3009,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8606365021203014,
+      "grad_norm": 2.194711923599243,
+      "learning_rate": 4.870358336191383e-06,
+      "loss": 2.0109,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8609658693235621,
+      "grad_norm": 2.693532943725586,
+      "learning_rate": 4.8477369554007055e-06,
+      "loss": 2.3555,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8612952365268228,
+      "grad_norm": 2.6600592136383057,
+      "learning_rate": 4.825165554997741e-06,
+      "loss": 2.2698,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8616246037300835,
+      "grad_norm": 2.6437413692474365,
+      "learning_rate": 4.802644159967435e-06,
+      "loss": 2.0872,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8619539709333444,
+      "grad_norm": 2.502390146255493,
+      "learning_rate": 4.780172795239401e-06,
+      "loss": 2.0786,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8622833381366051,
+      "grad_norm": 2.4297420978546143,
+      "learning_rate": 4.7577514856878616e-06,
+      "loss": 2.0086,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8626127053398658,
+      "grad_norm": 2.603891134262085,
+      "learning_rate": 4.735380256131627e-06,
+      "loss": 2.0819,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8629420725431265,
+      "grad_norm": 2.4713308811187744,
+      "learning_rate": 4.713059131334102e-06,
+      "loss": 2.1809,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8632714397463872,
+      "grad_norm": 2.54862642288208,
+      "learning_rate": 4.6907881360031715e-06,
+      "loss": 1.937,
+      "step": 2621
+    },
+    {
+      "epoch": 0.863600806949648,
+      "grad_norm": 2.82242488861084,
+      "learning_rate": 4.668567294791282e-06,
+      "loss": 1.9862,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8639301741529087,
+      "grad_norm": 3.0480566024780273,
+      "learning_rate": 4.646396632295347e-06,
+      "loss": 2.105,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8642595413561694,
+      "grad_norm": 2.951061248779297,
+      "learning_rate": 4.624276173056735e-06,
+      "loss": 1.9754,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8645889085594302,
+      "grad_norm": 2.960705041885376,
+      "learning_rate": 4.602205941561238e-06,
+      "loss": 1.648,
+      "step": 2625
+    },
+    {
+      "epoch": 0.864918275762691,
+      "grad_norm": 1.7927899360656738,
+      "learning_rate": 4.5801859622390584e-06,
+      "loss": 2.4304,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8652476429659517,
+      "grad_norm": 2.601215124130249,
+      "learning_rate": 4.558216259464765e-06,
+      "loss": 2.4331,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8655770101692124,
+      "grad_norm": 2.0217134952545166,
+      "learning_rate": 4.536296857557282e-06,
+      "loss": 2.1572,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8659063773724731,
+      "grad_norm": 2.2973663806915283,
+      "learning_rate": 4.514427780779845e-06,
+      "loss": 2.0926,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8662357445757338,
+      "grad_norm": 2.2276058197021484,
+      "learning_rate": 4.492609053339991e-06,
+      "loss": 2.1517,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8665651117789946,
+      "grad_norm": 2.29813814163208,
+      "learning_rate": 4.470840699389528e-06,
+      "loss": 2.0991,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8668944789822554,
+      "grad_norm": 2.494429588317871,
+      "learning_rate": 4.4491227430244804e-06,
+      "loss": 2.1413,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8672238461855161,
+      "grad_norm": 2.598921775817871,
+      "learning_rate": 4.4274552082851084e-06,
+      "loss": 2.2659,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8675532133887768,
+      "grad_norm": 2.702181339263916,
+      "learning_rate": 4.4058381191558525e-06,
+      "loss": 2.4259,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8678825805920376,
+      "grad_norm": 2.753028154373169,
+      "learning_rate": 4.384271499565323e-06,
+      "loss": 2.4114,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8682119477952983,
+      "grad_norm": 2.4215359687805176,
+      "learning_rate": 4.362755373386246e-06,
+      "loss": 2.3793,
+      "step": 2636
+    },
+    {
+      "epoch": 0.868541314998559,
+      "grad_norm": 2.8576810359954834,
+      "learning_rate": 4.341289764435463e-06,
+      "loss": 2.3855,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8688706822018197,
+      "grad_norm": 2.6904537677764893,
+      "learning_rate": 4.319874696473908e-06,
+      "loss": 2.1941,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8692000494050804,
+      "grad_norm": 2.96549916267395,
+      "learning_rate": 4.298510193206545e-06,
+      "loss": 2.1925,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8695294166083413,
+      "grad_norm": 2.366469144821167,
+      "learning_rate": 4.277196278282386e-06,
+      "loss": 2.0381,
+      "step": 2640
+    },
+    {
+      "epoch": 0.869858783811602,
+      "grad_norm": 2.5164620876312256,
+      "learning_rate": 4.255932975294441e-06,
+      "loss": 2.0369,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8701881510148627,
+      "grad_norm": 3.4024715423583984,
+      "learning_rate": 4.234720307779693e-06,
+      "loss": 2.0745,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8705175182181234,
+      "grad_norm": 2.9264469146728516,
+      "learning_rate": 4.213558299219078e-06,
+      "loss": 2.1802,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8708468854213842,
+      "grad_norm": 2.655367612838745,
+      "learning_rate": 4.192446973037451e-06,
+      "loss": 2.1807,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8711762526246449,
+      "grad_norm": 2.6134543418884277,
+      "learning_rate": 4.171386352603585e-06,
+      "loss": 2.1425,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8715056198279056,
+      "grad_norm": 2.909407615661621,
+      "learning_rate": 4.150376461230087e-06,
+      "loss": 2.1832,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8718349870311664,
+      "grad_norm": 2.7573821544647217,
+      "learning_rate": 4.129417322173446e-06,
+      "loss": 1.8164,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8721643542344271,
+      "grad_norm": 3.42594051361084,
+      "learning_rate": 4.108508958633955e-06,
+      "loss": 2.0909,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8724937214376879,
+      "grad_norm": 2.7599639892578125,
+      "learning_rate": 4.087651393755703e-06,
+      "loss": 1.6129,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8728230886409486,
+      "grad_norm": 3.7333033084869385,
+      "learning_rate": 4.066844650626556e-06,
+      "loss": 2.2567,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8731524558442093,
+      "grad_norm": 2.1264333724975586,
+      "learning_rate": 4.046088752278127e-06,
+      "loss": 2.4013,
+      "step": 2651
+    },
+    {
+      "epoch": 0.87348182304747,
+      "grad_norm": 2.2733492851257324,
+      "learning_rate": 4.025383721685722e-06,
+      "loss": 2.4881,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8738111902507307,
+      "grad_norm": 2.3039214611053467,
+      "learning_rate": 4.004729581768363e-06,
+      "loss": 2.0644,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8741405574539916,
+      "grad_norm": 2.163909912109375,
+      "learning_rate": 3.9841263553887395e-06,
+      "loss": 2.3237,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8744699246572523,
+      "grad_norm": 2.235494613647461,
+      "learning_rate": 3.96357406535317e-06,
+      "loss": 2.2701,
+      "step": 2655
+    },
+    {
+      "epoch": 0.874799291860513,
+      "grad_norm": 2.8736233711242676,
+      "learning_rate": 3.943072734411607e-06,
+      "loss": 2.4804,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8751286590637737,
+      "grad_norm": 2.4093754291534424,
+      "learning_rate": 3.922622385257579e-06,
+      "loss": 2.1099,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8754580262670345,
+      "grad_norm": 3.4634957313537598,
+      "learning_rate": 3.902223040528185e-06,
+      "loss": 2.7003,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8757873934702952,
+      "grad_norm": 2.40132474899292,
+      "learning_rate": 3.8818747228040795e-06,
+      "loss": 2.3205,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8761167606735559,
+      "grad_norm": 2.1787571907043457,
+      "learning_rate": 3.861577454609417e-06,
+      "loss": 1.9683,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8764461278768166,
+      "grad_norm": 2.4341671466827393,
+      "learning_rate": 3.841331258411846e-06,
+      "loss": 1.9743,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8767754950800773,
+      "grad_norm": 2.984081745147705,
+      "learning_rate": 3.8211361566224905e-06,
+      "loss": 2.2883,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8771048622833382,
+      "grad_norm": 2.602597236633301,
+      "learning_rate": 3.8009921715959184e-06,
+      "loss": 2.2132,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8774342294865989,
+      "grad_norm": 2.761525869369507,
+      "learning_rate": 3.7808993256300884e-06,
+      "loss": 1.8199,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8777635966898596,
+      "grad_norm": 2.7709145545959473,
+      "learning_rate": 3.7608576409663866e-06,
+      "loss": 2.265,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8780929638931203,
+      "grad_norm": 2.5388407707214355,
+      "learning_rate": 3.7408671397895444e-06,
+      "loss": 1.9894,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8784223310963811,
+      "grad_norm": 2.6022965908050537,
+      "learning_rate": 3.7209278442276507e-06,
+      "loss": 2.1244,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8787516982996418,
+      "grad_norm": 2.502260208129883,
+      "learning_rate": 3.7010397763521e-06,
+      "loss": 2.1518,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8790810655029025,
+      "grad_norm": 2.6602087020874023,
+      "learning_rate": 3.681202958177604e-06,
+      "loss": 2.2532,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8794104327061633,
+      "grad_norm": 2.9550154209136963,
+      "learning_rate": 3.661417411662105e-06,
+      "loss": 2.0059,
+      "step": 2670
+    },
+    {
+      "epoch": 0.879739799909424,
+      "grad_norm": 2.7335567474365234,
+      "learning_rate": 3.641683158706827e-06,
+      "loss": 2.4175,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8800691671126848,
+      "grad_norm": 2.868046522140503,
+      "learning_rate": 3.6220002211562023e-06,
+      "loss": 1.9707,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8803985343159455,
+      "grad_norm": 3.0974791049957275,
+      "learning_rate": 3.602368620797869e-06,
+      "loss": 1.9962,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8807279015192062,
+      "grad_norm": 2.911442518234253,
+      "learning_rate": 3.5827883793626225e-06,
+      "loss": 1.8045,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8810572687224669,
+      "grad_norm": 3.6394643783569336,
+      "learning_rate": 3.5632595185244143e-06,
+      "loss": 1.8825,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8813866359257277,
+      "grad_norm": 1.845200538635254,
+      "learning_rate": 3.543782059900347e-06,
+      "loss": 2.3288,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8817160031289885,
+      "grad_norm": 2.175915002822876,
+      "learning_rate": 3.524356025050568e-06,
+      "loss": 2.1404,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8820453703322492,
+      "grad_norm": 2.516981363296509,
+      "learning_rate": 3.5049814354783496e-06,
+      "loss": 2.4723,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8823747375355099,
+      "grad_norm": 2.449856758117676,
+      "learning_rate": 3.4856583126300078e-06,
+      "loss": 2.1453,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8827041047387706,
+      "grad_norm": 2.3987417221069336,
+      "learning_rate": 3.4663866778948728e-06,
+      "loss": 2.2724,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8830334719420314,
+      "grad_norm": 2.567919969558716,
+      "learning_rate": 3.4471665526053075e-06,
+      "loss": 2.4146,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8833628391452921,
+      "grad_norm": 2.4699082374572754,
+      "learning_rate": 3.4279979580366274e-06,
+      "loss": 2.422,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8836922063485528,
+      "grad_norm": 2.8117380142211914,
+      "learning_rate": 3.408880915407142e-06,
+      "loss": 2.5623,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8840215735518135,
+      "grad_norm": 2.675783634185791,
+      "learning_rate": 3.3898154458780527e-06,
+      "loss": 2.4708,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8843509407550744,
+      "grad_norm": 2.3889050483703613,
+      "learning_rate": 3.3708015705535055e-06,
+      "loss": 1.9012,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8846803079583351,
+      "grad_norm": 2.9828548431396484,
+      "learning_rate": 3.35183931048052e-06,
+      "loss": 2.7445,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8850096751615958,
+      "grad_norm": 2.40610933303833,
+      "learning_rate": 3.3329286866490064e-06,
+      "loss": 2.0757,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8853390423648565,
+      "grad_norm": 2.6663005352020264,
+      "learning_rate": 3.314069719991697e-06,
+      "loss": 2.161,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8856684095681172,
+      "grad_norm": 3.467353582382202,
+      "learning_rate": 3.2952624313841275e-06,
+      "loss": 2.5355,
+      "step": 2689
+    },
+    {
+      "epoch": 0.885997776771378,
+      "grad_norm": 2.527111053466797,
+      "learning_rate": 3.276506841644655e-06,
+      "loss": 2.1816,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8863271439746387,
+      "grad_norm": 2.746145009994507,
+      "learning_rate": 3.257802971534402e-06,
+      "loss": 2.1686,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8866565111778995,
+      "grad_norm": 2.5353686809539795,
+      "learning_rate": 3.2391508417572438e-06,
+      "loss": 2.2508,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8869858783811602,
+      "grad_norm": 2.6209967136383057,
+      "learning_rate": 3.2205504729597714e-06,
+      "loss": 2.0796,
+      "step": 2693
+    },
+    {
+      "epoch": 0.887315245584421,
+      "grad_norm": 2.5521199703216553,
+      "learning_rate": 3.2020018857313096e-06,
+      "loss": 2.2875,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8876446127876817,
+      "grad_norm": 2.6525521278381348,
+      "learning_rate": 3.183505100603812e-06,
+      "loss": 2.0905,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8879739799909424,
+      "grad_norm": 3.0146303176879883,
+      "learning_rate": 3.1650601380519397e-06,
+      "loss": 2.0727,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8883033471942031,
+      "grad_norm": 3.2892794609069824,
+      "learning_rate": 3.1466670184929714e-06,
+      "loss": 2.0854,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8886327143974638,
+      "grad_norm": 2.7834742069244385,
+      "learning_rate": 3.128325762286799e-06,
+      "loss": 1.9612,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8889620816007247,
+      "grad_norm": 2.958502769470215,
+      "learning_rate": 3.110036389735904e-06,
+      "loss": 1.9484,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8892914488039854,
+      "grad_norm": 2.957582712173462,
+      "learning_rate": 3.091798921085343e-06,
+      "loss": 1.6971,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8896208160072461,
+      "grad_norm": 2.2949178218841553,
+      "learning_rate": 3.0736133765227213e-06,
+      "loss": 2.2748,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8899501832105068,
+      "grad_norm": 2.3045599460601807,
+      "learning_rate": 3.055479776178144e-06,
+      "loss": 2.2503,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8902795504137675,
+      "grad_norm": 2.671010732650757,
+      "learning_rate": 3.0373981401242423e-06,
+      "loss": 2.5059,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8906089176170283,
+      "grad_norm": 2.3719635009765625,
+      "learning_rate": 3.0193684883761177e-06,
+      "loss": 2.3507,
+      "step": 2704
+    },
+    {
+      "epoch": 0.890938284820289,
+      "grad_norm": 2.4885711669921875,
+      "learning_rate": 3.0013908408913328e-06,
+      "loss": 2.4243,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8912676520235497,
+      "grad_norm": 2.3072006702423096,
+      "learning_rate": 2.9834652175698806e-06,
+      "loss": 2.1482,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8915970192268104,
+      "grad_norm": 2.133578300476074,
+      "learning_rate": 2.965591638254178e-06,
+      "loss": 1.98,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8919263864300713,
+      "grad_norm": 2.2153220176696777,
+      "learning_rate": 2.9477701227290046e-06,
+      "loss": 2.1145,
+      "step": 2708
+    },
+    {
+      "epoch": 0.892255753633332,
+      "grad_norm": 2.178952932357788,
+      "learning_rate": 2.9300006907215404e-06,
+      "loss": 2.1486,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8925851208365927,
+      "grad_norm": 3.1489574909210205,
+      "learning_rate": 2.912283361901297e-06,
+      "loss": 2.6655,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8929144880398534,
+      "grad_norm": 2.3635940551757812,
+      "learning_rate": 2.8946181558801135e-06,
+      "loss": 2.1683,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8932438552431141,
+      "grad_norm": 3.0581488609313965,
+      "learning_rate": 2.8770050922121318e-06,
+      "loss": 2.2093,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8935732224463749,
+      "grad_norm": 2.5770411491394043,
+      "learning_rate": 2.8594441903937895e-06,
+      "loss": 2.0562,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8939025896496356,
+      "grad_norm": 2.952260732650757,
+      "learning_rate": 2.8419354698637657e-06,
+      "loss": 2.3356,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8942319568528964,
+      "grad_norm": 2.414741039276123,
+      "learning_rate": 2.824478950002979e-06,
+      "loss": 2.1069,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8945613240561571,
+      "grad_norm": 2.4834258556365967,
+      "learning_rate": 2.807074650134578e-06,
+      "loss": 1.668,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8948906912594179,
+      "grad_norm": 3.139636516571045,
+      "learning_rate": 2.7897225895238965e-06,
+      "loss": 2.1568,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8952200584626786,
+      "grad_norm": 2.8860831260681152,
+      "learning_rate": 2.772422787378459e-06,
+      "loss": 1.9833,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8955494256659393,
+      "grad_norm": 2.7193524837493896,
+      "learning_rate": 2.755175262847931e-06,
+      "loss": 2.0764,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8958787928692,
+      "grad_norm": 2.3650174140930176,
+      "learning_rate": 2.737980035024107e-06,
+      "loss": 1.6969,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8962081600724607,
+      "grad_norm": 2.5429396629333496,
+      "learning_rate": 2.720837122940895e-06,
+      "loss": 1.9325,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8965375272757216,
+      "grad_norm": 2.7163567543029785,
+      "learning_rate": 2.703746545574304e-06,
+      "loss": 2.2487,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8968668944789823,
+      "grad_norm": 2.875310182571411,
+      "learning_rate": 2.6867083218424014e-06,
+      "loss": 1.6736,
+      "step": 2723
+    },
+    {
+      "epoch": 0.897196261682243,
+      "grad_norm": 3.140059232711792,
+      "learning_rate": 2.669722470605307e-06,
+      "loss": 1.938,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8975256288855037,
+      "grad_norm": 3.804771661758423,
+      "learning_rate": 2.6527890106651576e-06,
+      "loss": 2.3505,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8978549960887645,
+      "grad_norm": 2.1390380859375,
+      "learning_rate": 2.635907960766126e-06,
+      "loss": 2.4625,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8981843632920252,
+      "grad_norm": 2.6339430809020996,
+      "learning_rate": 2.6190793395943268e-06,
+      "loss": 2.6437,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8985137304952859,
+      "grad_norm": 2.0819966793060303,
+      "learning_rate": 2.6023031657778752e-06,
+      "loss": 2.1295,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8988430976985466,
+      "grad_norm": 2.2953975200653076,
+      "learning_rate": 2.585579457886811e-06,
+      "loss": 2.131,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8991724649018074,
+      "grad_norm": 2.132822275161743,
+      "learning_rate": 2.568908234433115e-06,
+      "loss": 2.114,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8995018321050682,
+      "grad_norm": 2.3440139293670654,
+      "learning_rate": 2.5522895138706593e-06,
+      "loss": 2.3049,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8998311993083289,
+      "grad_norm": 2.4961066246032715,
+      "learning_rate": 2.5357233145952065e-06,
+      "loss": 2.3662,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9001605665115896,
+      "grad_norm": 2.0596587657928467,
+      "learning_rate": 2.5192096549443655e-06,
+      "loss": 1.7812,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9004899337148503,
+      "grad_norm": 2.2845394611358643,
+      "learning_rate": 2.5027485531976035e-06,
+      "loss": 2.0744,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9008193009181111,
+      "grad_norm": 2.4581706523895264,
+      "learning_rate": 2.4863400275762063e-06,
+      "loss": 2.3042,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9011486681213718,
+      "grad_norm": 2.356445074081421,
+      "learning_rate": 2.4699840962432606e-06,
+      "loss": 2.3077,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9014780353246326,
+      "grad_norm": 2.9179089069366455,
+      "learning_rate": 2.453680777303641e-06,
+      "loss": 2.4067,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9018074025278933,
+      "grad_norm": 2.447138547897339,
+      "learning_rate": 2.437430088803966e-06,
+      "loss": 2.2238,
+      "step": 2738
+    },
+    {
+      "epoch": 0.902136769731154,
+      "grad_norm": 2.1415274143218994,
+      "learning_rate": 2.4212320487326246e-06,
+      "loss": 1.9145,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9024661369344148,
+      "grad_norm": 2.2996044158935547,
+      "learning_rate": 2.405086675019691e-06,
+      "loss": 2.2053,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9027955041376755,
+      "grad_norm": 2.8865909576416016,
+      "learning_rate": 2.3889939855369693e-06,
+      "loss": 2.4935,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9031248713409362,
+      "grad_norm": 2.9362375736236572,
+      "learning_rate": 2.372953998097943e-06,
+      "loss": 2.0295,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9034542385441969,
+      "grad_norm": 2.538898468017578,
+      "learning_rate": 2.356966730457749e-06,
+      "loss": 1.9073,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9037836057474578,
+      "grad_norm": 2.7134170532226562,
+      "learning_rate": 2.34103220031317e-06,
+      "loss": 2.0446,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9041129729507185,
+      "grad_norm": 2.733494758605957,
+      "learning_rate": 2.325150425302619e-06,
+      "loss": 2.2672,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9044423401539792,
+      "grad_norm": 2.400979995727539,
+      "learning_rate": 2.3093214230061e-06,
+      "loss": 2.0618,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9047717073572399,
+      "grad_norm": 2.6966586112976074,
+      "learning_rate": 2.293545210945203e-06,
+      "loss": 1.9108,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9051010745605006,
+      "grad_norm": 2.9036622047424316,
+      "learning_rate": 2.2778218065830968e-06,
+      "loss": 2.2464,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9054304417637614,
+      "grad_norm": 2.609510898590088,
+      "learning_rate": 2.262151227324488e-06,
+      "loss": 1.8424,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9057598089670221,
+      "grad_norm": 2.7685904502868652,
+      "learning_rate": 2.2465334905156065e-06,
+      "loss": 1.7066,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9060891761702828,
+      "grad_norm": 2.4640228748321533,
+      "learning_rate": 2.230968613444201e-06,
+      "loss": 2.6138,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9064185433735436,
+      "grad_norm": 2.218118190765381,
+      "learning_rate": 2.2154566133394784e-06,
+      "loss": 2.2352,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9067479105768043,
+      "grad_norm": 2.2658257484436035,
+      "learning_rate": 2.1999975073721423e-06,
+      "loss": 2.4639,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9070772777800651,
+      "grad_norm": 2.1403121948242188,
+      "learning_rate": 2.184591312654344e-06,
+      "loss": 2.1344,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9074066449833258,
+      "grad_norm": 2.4829635620117188,
+      "learning_rate": 2.169238046239658e-06,
+      "loss": 2.414,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9077360121865865,
+      "grad_norm": 2.313821315765381,
+      "learning_rate": 2.1539377251230673e-06,
+      "loss": 1.9432,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9080653793898472,
+      "grad_norm": 2.121927261352539,
+      "learning_rate": 2.13869036624097e-06,
+      "loss": 2.0372,
+      "step": 2757
+    },
+    {
+      "epoch": 0.908394746593108,
+      "grad_norm": 2.498476266860962,
+      "learning_rate": 2.1234959864711025e-06,
+      "loss": 2.1289,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9087241137963687,
+      "grad_norm": 2.445488691329956,
+      "learning_rate": 2.1083546026325895e-06,
+      "loss": 2.3582,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9090534809996295,
+      "grad_norm": 2.175457239151001,
+      "learning_rate": 2.093266231485874e-06,
+      "loss": 1.9455,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9093828482028902,
+      "grad_norm": 2.539600372314453,
+      "learning_rate": 2.0782308897327297e-06,
+      "loss": 2.4663,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9097122154061509,
+      "grad_norm": 2.7965781688690186,
+      "learning_rate": 2.063248594016226e-06,
+      "loss": 2.4572,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9100415826094117,
+      "grad_norm": 2.535855293273926,
+      "learning_rate": 2.0483193609207086e-06,
+      "loss": 2.1825,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9103709498126724,
+      "grad_norm": 2.506638526916504,
+      "learning_rate": 2.0334432069718023e-06,
+      "loss": 2.2343,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9107003170159331,
+      "grad_norm": 2.5606038570404053,
+      "learning_rate": 2.0186201486363465e-06,
+      "loss": 2.016,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9110296842191938,
+      "grad_norm": 2.3744375705718994,
+      "learning_rate": 2.003850202322444e-06,
+      "loss": 1.8271,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9113590514224547,
+      "grad_norm": 2.7433090209960938,
+      "learning_rate": 1.989133384379377e-06,
+      "loss": 2.1174,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9116884186257154,
+      "grad_norm": 2.728475570678711,
+      "learning_rate": 1.9744697110976373e-06,
+      "loss": 1.9861,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9120177858289761,
+      "grad_norm": 2.9022865295410156,
+      "learning_rate": 1.9598591987088853e-06,
+      "loss": 2.2861,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9123471530322368,
+      "grad_norm": 2.443715810775757,
+      "learning_rate": 1.9453018633859344e-06,
+      "loss": 1.8767,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9126765202354975,
+      "grad_norm": 3.8203299045562744,
+      "learning_rate": 1.930797721242722e-06,
+      "loss": 1.98,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9130058874387583,
+      "grad_norm": 2.870041608810425,
+      "learning_rate": 1.9163467883343223e-06,
+      "loss": 2.0627,
+      "step": 2772
+    },
+    {
+      "epoch": 0.913335254642019,
+      "grad_norm": 2.7346694469451904,
+      "learning_rate": 1.9019490806569063e-06,
+      "loss": 2.3428,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9136646218452797,
+      "grad_norm": 3.569988489151001,
+      "learning_rate": 1.88760461414772e-06,
+      "loss": 2.3849,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9139939890485405,
+      "grad_norm": 3.2586612701416016,
+      "learning_rate": 1.8733134046850898e-06,
+      "loss": 1.825,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9143233562518013,
+      "grad_norm": 2.39789080619812,
+      "learning_rate": 1.8590754680883782e-06,
+      "loss": 2.234,
+      "step": 2776
+    },
+    {
+      "epoch": 0.914652723455062,
+      "grad_norm": 2.7146317958831787,
+      "learning_rate": 1.8448908201179892e-06,
+      "loss": 2.5098,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9149820906583227,
+      "grad_norm": 2.7325544357299805,
+      "learning_rate": 1.830759476475319e-06,
+      "loss": 2.3236,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9153114578615834,
+      "grad_norm": 2.257678747177124,
+      "learning_rate": 1.816681452802782e-06,
+      "loss": 2.1115,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9156408250648441,
+      "grad_norm": 2.2593493461608887,
+      "learning_rate": 1.8026567646837633e-06,
+      "loss": 2.4698,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9159701922681049,
+      "grad_norm": 2.5541627407073975,
+      "learning_rate": 1.7886854276426057e-06,
+      "loss": 2.3059,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9162995594713657,
+      "grad_norm": 2.6000893115997314,
+      "learning_rate": 1.774767457144605e-06,
+      "loss": 2.2396,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9166289266746264,
+      "grad_norm": 2.629995107650757,
+      "learning_rate": 1.7609028685959704e-06,
+      "loss": 2.0903,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9169582938778871,
+      "grad_norm": 2.4334161281585693,
+      "learning_rate": 1.7470916773438373e-06,
+      "loss": 2.1372,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9172876610811479,
+      "grad_norm": 2.6131606101989746,
+      "learning_rate": 1.733333898676215e-06,
+      "loss": 2.5038,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9176170282844086,
+      "grad_norm": 2.5102593898773193,
+      "learning_rate": 1.7196295478220048e-06,
+      "loss": 1.9959,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9179463954876693,
+      "grad_norm": 2.299788236618042,
+      "learning_rate": 1.705978639950967e-06,
+      "loss": 1.9553,
+      "step": 2787
+    },
+    {
+      "epoch": 0.91827576269093,
+      "grad_norm": 2.4039552211761475,
+      "learning_rate": 1.692381190173692e-06,
+      "loss": 2.1673,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9186051298941907,
+      "grad_norm": 2.8299825191497803,
+      "learning_rate": 1.6788372135416063e-06,
+      "loss": 2.0858,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9189344970974516,
+      "grad_norm": 2.670093059539795,
+      "learning_rate": 1.6653467250469402e-06,
+      "loss": 2.2654,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9192638643007123,
+      "grad_norm": 2.81526255607605,
+      "learning_rate": 1.65190973962272e-06,
+      "loss": 1.9366,
+      "step": 2791
+    },
+    {
+      "epoch": 0.919593231503973,
+      "grad_norm": 2.22592830657959,
+      "learning_rate": 1.6385262721427374e-06,
+      "loss": 1.7114,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9199225987072337,
+      "grad_norm": 2.677290916442871,
+      "learning_rate": 1.625196337421564e-06,
+      "loss": 2.3907,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9202519659104944,
+      "grad_norm": 2.487548828125,
+      "learning_rate": 1.6119199502144966e-06,
+      "loss": 1.9388,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9205813331137552,
+      "grad_norm": 3.054719924926758,
+      "learning_rate": 1.5986971252175686e-06,
+      "loss": 2.0144,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9209107003170159,
+      "grad_norm": 2.6823740005493164,
+      "learning_rate": 1.5855278770675108e-06,
+      "loss": 1.9984,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9212400675202767,
+      "grad_norm": 2.6444766521453857,
+      "learning_rate": 1.5724122203417679e-06,
+      "loss": 1.9557,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9215694347235374,
+      "grad_norm": 3.096118211746216,
+      "learning_rate": 1.5593501695584433e-06,
+      "loss": 1.9882,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9218988019267982,
+      "grad_norm": 3.212019681930542,
+      "learning_rate": 1.5463417391763101e-06,
+      "loss": 2.1233,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9222281691300589,
+      "grad_norm": 2.8375954627990723,
+      "learning_rate": 1.5333869435948e-06,
+      "loss": 1.7124,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9225575363333196,
+      "grad_norm": 2.4758059978485107,
+      "learning_rate": 1.5204857971539477e-06,
+      "loss": 2.2537,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9228869035365803,
+      "grad_norm": 2.129150629043579,
+      "learning_rate": 1.5076383141344352e-06,
+      "loss": 2.0797,
+      "step": 2802
+    },
+    {
+      "epoch": 0.923216270739841,
+      "grad_norm": 2.272265672683716,
+      "learning_rate": 1.4948445087575091e-06,
+      "loss": 2.1409,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9235456379431018,
+      "grad_norm": 2.231919288635254,
+      "learning_rate": 1.4821043951850133e-06,
+      "loss": 2.1279,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9238750051463626,
+      "grad_norm": 2.3199281692504883,
+      "learning_rate": 1.4694179875193726e-06,
+      "loss": 2.208,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9242043723496233,
+      "grad_norm": 2.0582661628723145,
+      "learning_rate": 1.4567852998035426e-06,
+      "loss": 2.0938,
+      "step": 2806
+    },
+    {
+      "epoch": 0.924533739552884,
+      "grad_norm": 2.4887518882751465,
+      "learning_rate": 1.4442063460210263e-06,
+      "loss": 2.3684,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9248631067561448,
+      "grad_norm": 2.5550124645233154,
+      "learning_rate": 1.4316811400958362e-06,
+      "loss": 2.0291,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9251924739594055,
+      "grad_norm": 2.2390754222869873,
+      "learning_rate": 1.4192096958925038e-06,
+      "loss": 1.8898,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9255218411626662,
+      "grad_norm": 2.5573508739471436,
+      "learning_rate": 1.4067920272160362e-06,
+      "loss": 2.3696,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9258512083659269,
+      "grad_norm": 2.908836841583252,
+      "learning_rate": 1.394428147811927e-06,
+      "loss": 2.4529,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9261805755691876,
+      "grad_norm": 2.5992095470428467,
+      "learning_rate": 1.382118071366123e-06,
+      "loss": 1.9448,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9265099427724485,
+      "grad_norm": 2.3026111125946045,
+      "learning_rate": 1.3698618115050244e-06,
+      "loss": 2.0529,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9268393099757092,
+      "grad_norm": 2.8297219276428223,
+      "learning_rate": 1.3576593817954286e-06,
+      "loss": 2.1613,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9271686771789699,
+      "grad_norm": 2.73018479347229,
+      "learning_rate": 1.345510795744581e-06,
+      "loss": 2.0114,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9274980443822306,
+      "grad_norm": 3.0178444385528564,
+      "learning_rate": 1.3334160668001195e-06,
+      "loss": 2.2001,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9278274115854914,
+      "grad_norm": 3.0993990898132324,
+      "learning_rate": 1.3213752083500508e-06,
+      "loss": 2.2696,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9281567787887521,
+      "grad_norm": 2.804314136505127,
+      "learning_rate": 1.3093882337227693e-06,
+      "loss": 2.0842,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9284861459920128,
+      "grad_norm": 2.4682846069335938,
+      "learning_rate": 1.2974551561870107e-06,
+      "loss": 1.8982,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9288155131952736,
+      "grad_norm": 2.5155272483825684,
+      "learning_rate": 1.2855759889518594e-06,
+      "loss": 2.1026,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9291448803985343,
+      "grad_norm": 2.995867967605591,
+      "learning_rate": 1.2737507451667074e-06,
+      "loss": 2.0615,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9294742476017951,
+      "grad_norm": 2.7689971923828125,
+      "learning_rate": 1.2619794379212845e-06,
+      "loss": 2.1408,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9298036148050558,
+      "grad_norm": 2.611830234527588,
+      "learning_rate": 1.2502620802455955e-06,
+      "loss": 1.7406,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9301329820083165,
+      "grad_norm": 2.6598401069641113,
+      "learning_rate": 1.2385986851099318e-06,
+      "loss": 1.5781,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9304623492115772,
+      "grad_norm": 4.383054733276367,
+      "learning_rate": 1.226989265424855e-06,
+      "loss": 1.6148,
+      "step": 2825
+    },
+    {
+      "epoch": 0.930791716414838,
+      "grad_norm": 1.9361492395401,
+      "learning_rate": 1.2154338340411797e-06,
+      "loss": 2.2058,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9311210836180988,
+      "grad_norm": 2.430833101272583,
+      "learning_rate": 1.2039324037499578e-06,
+      "loss": 2.2598,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9314504508213595,
+      "grad_norm": 2.149946928024292,
+      "learning_rate": 1.192484987282466e-06,
+      "loss": 2.2,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9317798180246202,
+      "grad_norm": 2.194275140762329,
+      "learning_rate": 1.181091597310191e-06,
+      "loss": 2.4223,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9321091852278809,
+      "grad_norm": 2.4965078830718994,
+      "learning_rate": 1.1697522464448107e-06,
+      "loss": 2.3354,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9324385524311417,
+      "grad_norm": 2.4370152950286865,
+      "learning_rate": 1.1584669472382014e-06,
+      "loss": 2.2227,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9327679196344024,
+      "grad_norm": 2.3684237003326416,
+      "learning_rate": 1.1472357121823873e-06,
+      "loss": 2.0301,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9330972868376631,
+      "grad_norm": 2.8566548824310303,
+      "learning_rate": 1.136058553709568e-06,
+      "loss": 2.6412,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9334266540409238,
+      "grad_norm": 2.3684935569763184,
+      "learning_rate": 1.124935484192069e-06,
+      "loss": 2.0108,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9337560212441847,
+      "grad_norm": 2.404863119125366,
+      "learning_rate": 1.1138665159423411e-06,
+      "loss": 2.2514,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9340853884474454,
+      "grad_norm": 2.4747257232666016,
+      "learning_rate": 1.102851661212967e-06,
+      "loss": 2.1483,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9344147556507061,
+      "grad_norm": 2.7829651832580566,
+      "learning_rate": 1.0918909321966097e-06,
+      "loss": 2.513,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9347441228539668,
+      "grad_norm": 2.477625608444214,
+      "learning_rate": 1.0809843410260312e-06,
+      "loss": 2.0257,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9350734900572275,
+      "grad_norm": 3.145524024963379,
+      "learning_rate": 1.0701318997740629e-06,
+      "loss": 2.3148,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9354028572604883,
+      "grad_norm": 3.496830701828003,
+      "learning_rate": 1.0593336204535952e-06,
+      "loss": 2.4637,
+      "step": 2840
+    },
+    {
+      "epoch": 0.935732224463749,
+      "grad_norm": 2.429844856262207,
+      "learning_rate": 1.0485895150175672e-06,
+      "loss": 1.8108,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9360615916670098,
+      "grad_norm": 2.853318452835083,
+      "learning_rate": 1.0378995953589488e-06,
+      "loss": 2.4304,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9363909588702705,
+      "grad_norm": 2.5983059406280518,
+      "learning_rate": 1.0272638733107298e-06,
+      "loss": 2.0046,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9367203260735312,
+      "grad_norm": 2.3808035850524902,
+      "learning_rate": 1.0166823606459097e-06,
+      "loss": 1.8917,
+      "step": 2844
+    },
+    {
+      "epoch": 0.937049693276792,
+      "grad_norm": 2.6919662952423096,
+      "learning_rate": 1.0061550690774857e-06,
+      "loss": 1.9641,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9373790604800527,
+      "grad_norm": 2.5325708389282227,
+      "learning_rate": 9.956820102584253e-07,
+      "loss": 1.8109,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9377084276833134,
+      "grad_norm": 3.0093464851379395,
+      "learning_rate": 9.85263195781666e-07,
+      "loss": 2.1708,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9380377948865741,
+      "grad_norm": 2.9539554119110107,
+      "learning_rate": 9.748986371801106e-07,
+      "loss": 2.1125,
+      "step": 2848
+    },
+    {
+      "epoch": 0.938367162089835,
+      "grad_norm": 2.907458543777466,
+      "learning_rate": 9.645883459265982e-07,
+      "loss": 1.437,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9386965292930957,
+      "grad_norm": 3.331385374069214,
+      "learning_rate": 9.543323334338939e-07,
+      "loss": 1.8898,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9390258964963564,
+      "grad_norm": 1.8661000728607178,
+      "learning_rate": 9.441306110546944e-07,
+      "loss": 2.2593,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9393552636996171,
+      "grad_norm": 2.427933692932129,
+      "learning_rate": 9.339831900815776e-07,
+      "loss": 2.6372,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9396846309028778,
+      "grad_norm": 2.3956634998321533,
+      "learning_rate": 9.238900817470308e-07,
+      "loss": 2.1686,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9400139981061386,
+      "grad_norm": 2.860553503036499,
+      "learning_rate": 9.138512972234225e-07,
+      "loss": 1.998,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9403433653093993,
+      "grad_norm": 2.519784927368164,
+      "learning_rate": 9.03866847622975e-07,
+      "loss": 2.4247,
+      "step": 2855
+    },
+    {
+      "epoch": 0.94067273251266,
+      "grad_norm": 2.6663100719451904,
+      "learning_rate": 8.939367439977808e-07,
+      "loss": 2.4213,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9410020997159207,
+      "grad_norm": 2.3024487495422363,
+      "learning_rate": 8.840609973397585e-07,
+      "loss": 2.4378,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9413314669191816,
+      "grad_norm": 2.864820957183838,
+      "learning_rate": 8.7423961858068e-07,
+      "loss": 2.4476,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9416608341224423,
+      "grad_norm": 2.920145034790039,
+      "learning_rate": 8.644726185921159e-07,
+      "loss": 2.2515,
+      "step": 2859
+    },
+    {
+      "epoch": 0.941990201325703,
+      "grad_norm": 2.6233792304992676,
+      "learning_rate": 8.547600081854456e-07,
+      "loss": 2.4181,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9423195685289637,
+      "grad_norm": 2.389942169189453,
+      "learning_rate": 8.451017981118525e-07,
+      "loss": 2.3034,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9426489357322244,
+      "grad_norm": 2.579540252685547,
+      "learning_rate": 8.354979990622957e-07,
+      "loss": 2.5177,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9429783029354852,
+      "grad_norm": 2.6074090003967285,
+      "learning_rate": 8.259486216675105e-07,
+      "loss": 1.9105,
+      "step": 2863
+    },
+    {
+      "epoch": 0.943307670138746,
+      "grad_norm": 2.744189977645874,
+      "learning_rate": 8.164536764979857e-07,
+      "loss": 2.1191,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9436370373420067,
+      "grad_norm": 2.729266405105591,
+      "learning_rate": 8.070131740639586e-07,
+      "loss": 2.1855,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9439664045452674,
+      "grad_norm": 2.6576309204101562,
+      "learning_rate": 7.976271248154033e-07,
+      "loss": 2.0916,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9442957717485282,
+      "grad_norm": 2.8636012077331543,
+      "learning_rate": 7.882955391420143e-07,
+      "loss": 2.3881,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9446251389517889,
+      "grad_norm": 2.3858633041381836,
+      "learning_rate": 7.790184273732071e-07,
+      "loss": 1.817,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9449545061550496,
+      "grad_norm": 2.9734864234924316,
+      "learning_rate": 7.697957997780947e-07,
+      "loss": 2.0361,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9452838733583103,
+      "grad_norm": 2.881843090057373,
+      "learning_rate": 7.606276665654777e-07,
+      "loss": 2.2603,
+      "step": 2870
+    },
+    {
+      "epoch": 0.945613240561571,
+      "grad_norm": 2.7740421295166016,
+      "learning_rate": 7.515140378838381e-07,
+      "loss": 2.1547,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9459426077648319,
+      "grad_norm": 2.4186999797821045,
+      "learning_rate": 7.424549238213174e-07,
+      "loss": 1.6122,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9462719749680926,
+      "grad_norm": 2.858855724334717,
+      "learning_rate": 7.33450334405722e-07,
+      "loss": 1.7438,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9466013421713533,
+      "grad_norm": 2.9262588024139404,
+      "learning_rate": 7.245002796045008e-07,
+      "loss": 1.896,
+      "step": 2874
+    },
+    {
+      "epoch": 0.946930709374614,
+      "grad_norm": 4.089500427246094,
+      "learning_rate": 7.156047693247403e-07,
+      "loss": 2.6594,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9472600765778748,
+      "grad_norm": 2.226041078567505,
+      "learning_rate": 7.067638134131472e-07,
+      "loss": 2.405,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9475894437811355,
+      "grad_norm": 2.344412326812744,
+      "learning_rate": 6.97977421656032e-07,
+      "loss": 2.4191,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9479188109843962,
+      "grad_norm": 2.1592891216278076,
+      "learning_rate": 6.892456037793204e-07,
+      "loss": 2.4441,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9482481781876569,
+      "grad_norm": 2.270063877105713,
+      "learning_rate": 6.805683694485143e-07,
+      "loss": 1.9226,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9485775453909177,
+      "grad_norm": 2.3790276050567627,
+      "learning_rate": 6.719457282687136e-07,
+      "loss": 2.3371,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9489069125941785,
+      "grad_norm": 2.7500150203704834,
+      "learning_rate": 6.633776897845667e-07,
+      "loss": 2.4479,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9492362797974392,
+      "grad_norm": 2.5941100120544434,
+      "learning_rate": 6.548642634803037e-07,
+      "loss": 2.2519,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9495656470006999,
+      "grad_norm": 2.158876895904541,
+      "learning_rate": 6.464054587796809e-07,
+      "loss": 2.4172,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9498950142039606,
+      "grad_norm": 2.5841729640960693,
+      "learning_rate": 6.380012850460082e-07,
+      "loss": 2.297,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9502243814072214,
+      "grad_norm": 2.719069004058838,
+      "learning_rate": 6.296517515821055e-07,
+      "loss": 2.2869,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9505537486104821,
+      "grad_norm": 2.4023020267486572,
+      "learning_rate": 6.213568676303294e-07,
+      "loss": 2.2206,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9508831158137429,
+      "grad_norm": 2.7940988540649414,
+      "learning_rate": 6.131166423725354e-07,
+      "loss": 2.0593,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9512124830170036,
+      "grad_norm": 2.5038106441497803,
+      "learning_rate": 6.04931084930066e-07,
+      "loss": 2.0466,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9515418502202643,
+      "grad_norm": 2.42012619972229,
+      "learning_rate": 5.968002043637733e-07,
+      "loss": 2.0527,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9518712174235251,
+      "grad_norm": 2.5355756282806396,
+      "learning_rate": 5.887240096739633e-07,
+      "loss": 2.2696,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9522005846267858,
+      "grad_norm": 2.5182111263275146,
+      "learning_rate": 5.807025098004127e-07,
+      "loss": 1.7523,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9525299518300465,
+      "grad_norm": 2.754927396774292,
+      "learning_rate": 5.727357136223633e-07,
+      "loss": 2.0462,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9528593190333072,
+      "grad_norm": 2.4149563312530518,
+      "learning_rate": 5.648236299584997e-07,
+      "loss": 2.0588,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9531886862365679,
+      "grad_norm": 2.708524703979492,
+      "learning_rate": 5.56966267566944e-07,
+      "loss": 2.0039,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9535180534398288,
+      "grad_norm": 2.535885810852051,
+      "learning_rate": 5.491636351452445e-07,
+      "loss": 2.0345,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9538474206430895,
+      "grad_norm": 3.0200557708740234,
+      "learning_rate": 5.414157413303755e-07,
+      "loss": 1.7782,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9541767878463502,
+      "grad_norm": 2.9520061016082764,
+      "learning_rate": 5.33722594698699e-07,
+      "loss": 2.0286,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9545061550496109,
+      "grad_norm": 2.945594549179077,
+      "learning_rate": 5.260842037659919e-07,
+      "loss": 1.8728,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9548355222528717,
+      "grad_norm": 3.269399642944336,
+      "learning_rate": 5.185005769874185e-07,
+      "loss": 2.2312,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9551648894561324,
+      "grad_norm": 3.335346221923828,
+      "learning_rate": 5.10971722757525e-07,
+      "loss": 2.1155,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9554942566593931,
+      "grad_norm": 1.8741716146469116,
+      "learning_rate": 5.034976494102284e-07,
+      "loss": 2.2083,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9558236238626538,
+      "grad_norm": 2.057013511657715,
+      "learning_rate": 4.960783652187939e-07,
+      "loss": 2.1859,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9561529910659146,
+      "grad_norm": 2.220473051071167,
+      "learning_rate": 4.887138783958467e-07,
+      "loss": 2.3862,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9564823582691754,
+      "grad_norm": 2.3261795043945312,
+      "learning_rate": 4.814041970933713e-07,
+      "loss": 2.317,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9568117254724361,
+      "grad_norm": 2.2345638275146484,
+      "learning_rate": 4.7414932940265664e-07,
+      "loss": 2.5307,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9571410926756968,
+      "grad_norm": 2.4309141635894775,
+      "learning_rate": 4.669492833543454e-07,
+      "loss": 2.2482,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9574704598789575,
+      "grad_norm": 2.4680659770965576,
+      "learning_rate": 4.598040669183734e-07,
+      "loss": 2.6021,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9577998270822183,
+      "grad_norm": 2.261176347732544,
+      "learning_rate": 4.5271368800400283e-07,
+      "loss": 2.277,
+      "step": 2908
+    },
+    {
+      "epoch": 0.958129194285479,
+      "grad_norm": 2.5138943195343018,
+      "learning_rate": 4.4567815445977767e-07,
+      "loss": 2.2765,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9584585614887398,
+      "grad_norm": 2.699733257293701,
+      "learning_rate": 4.386974740735461e-07,
+      "loss": 2.098,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9587879286920005,
+      "grad_norm": 2.301736831665039,
+      "learning_rate": 4.317716545724215e-07,
+      "loss": 2.1506,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9591172958952612,
+      "grad_norm": 2.6759886741638184,
+      "learning_rate": 4.2490070362281587e-07,
+      "loss": 2.0459,
+      "step": 2912
+    },
+    {
+      "epoch": 0.959446663098522,
+      "grad_norm": 2.393205404281616,
+      "learning_rate": 4.180846288303786e-07,
+      "loss": 2.05,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9597760303017827,
+      "grad_norm": 2.5351758003234863,
+      "learning_rate": 4.113234377400299e-07,
+      "loss": 1.8458,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9601053975050434,
+      "grad_norm": 2.2176060676574707,
+      "learning_rate": 4.0461713783593313e-07,
+      "loss": 1.9287,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9604347647083041,
+      "grad_norm": 3.161897659301758,
+      "learning_rate": 3.9796573654148905e-07,
+      "loss": 2.5525,
+      "step": 2916
+    },
+    {
+      "epoch": 0.960764131911565,
+      "grad_norm": 2.861020803451538,
+      "learning_rate": 3.9136924121933594e-07,
+      "loss": 1.9952,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9610934991148257,
+      "grad_norm": 2.7215144634246826,
+      "learning_rate": 3.848276591713329e-07,
+      "loss": 2.1098,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9614228663180864,
+      "grad_norm": 2.677161455154419,
+      "learning_rate": 3.783409976385488e-07,
+      "loss": 2.0806,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9617522335213471,
+      "grad_norm": 3.068319082260132,
+      "learning_rate": 3.719092638012622e-07,
+      "loss": 1.9768,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9620816007246078,
+      "grad_norm": 3.429572820663452,
+      "learning_rate": 3.655324647789615e-07,
+      "loss": 1.8987,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9624109679278686,
+      "grad_norm": 3.0230207443237305,
+      "learning_rate": 3.5921060763030033e-07,
+      "loss": 2.1049,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9627403351311293,
+      "grad_norm": 3.3592071533203125,
+      "learning_rate": 3.5294369935314207e-07,
+      "loss": 2.118,
+      "step": 2923
+    },
+    {
+      "epoch": 0.96306970233439,
+      "grad_norm": 2.8637266159057617,
+      "learning_rate": 3.4673174688450994e-07,
+      "loss": 2.0448,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9633990695376508,
+      "grad_norm": 3.2536802291870117,
+      "learning_rate": 3.4057475710059796e-07,
+      "loss": 1.8276,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9637284367409116,
+      "grad_norm": 2.2297792434692383,
+      "learning_rate": 3.3447273681676547e-07,
+      "loss": 2.4885,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9640578039441723,
+      "grad_norm": 2.3141520023345947,
+      "learning_rate": 3.284256927875262e-07,
+      "loss": 2.5599,
+      "step": 2927
+    },
+    {
+      "epoch": 0.964387171147433,
+      "grad_norm": 2.275432586669922,
+      "learning_rate": 3.224336317065202e-07,
+      "loss": 2.1016,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9647165383506937,
+      "grad_norm": 2.278019666671753,
+      "learning_rate": 3.1649656020654726e-07,
+      "loss": 2.2134,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9650459055539544,
+      "grad_norm": 2.311898946762085,
+      "learning_rate": 3.106144848595283e-07,
+      "loss": 2.386,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9653752727572152,
+      "grad_norm": 2.683455228805542,
+      "learning_rate": 3.047874121765049e-07,
+      "loss": 2.6858,
+      "step": 2931
+    },
+    {
+      "epoch": 0.965704639960476,
+      "grad_norm": 2.5970003604888916,
+      "learning_rate": 2.9901534860764524e-07,
+      "loss": 2.3024,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9660340071637367,
+      "grad_norm": 2.268040418624878,
+      "learning_rate": 2.9329830054221076e-07,
+      "loss": 2.1205,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9663633743669974,
+      "grad_norm": 2.2917988300323486,
+      "learning_rate": 2.876362743085781e-07,
+      "loss": 2.0411,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9666927415702582,
+      "grad_norm": 2.5120959281921387,
+      "learning_rate": 2.820292761742116e-07,
+      "loss": 2.1063,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9670221087735189,
+      "grad_norm": 2.403350353240967,
+      "learning_rate": 2.7647731234565767e-07,
+      "loss": 2.4898,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9673514759767796,
+      "grad_norm": 2.1802988052368164,
+      "learning_rate": 2.7098038896856136e-07,
+      "loss": 2.0262,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9676808431800403,
+      "grad_norm": 2.6515746116638184,
+      "learning_rate": 2.655385121276277e-07,
+      "loss": 2.0875,
+      "step": 2938
+    },
+    {
+      "epoch": 0.968010210383301,
+      "grad_norm": 2.75669527053833,
+      "learning_rate": 2.6015168784663255e-07,
+      "loss": 1.963,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9683395775865619,
+      "grad_norm": 2.94972562789917,
+      "learning_rate": 2.5481992208841175e-07,
+      "loss": 2.2923,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9686689447898226,
+      "grad_norm": 2.706496477127075,
+      "learning_rate": 2.4954322075485536e-07,
+      "loss": 2.4328,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9689983119930833,
+      "grad_norm": 3.0476579666137695,
+      "learning_rate": 2.4432158968689666e-07,
+      "loss": 2.018,
+      "step": 2942
+    },
+    {
+      "epoch": 0.969327679196344,
+      "grad_norm": 2.800839900970459,
+      "learning_rate": 2.391550346645288e-07,
+      "loss": 2.0205,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9696570463996047,
+      "grad_norm": 2.6107017993927,
+      "learning_rate": 2.3404356140675488e-07,
+      "loss": 2.1034,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9699864136028655,
+      "grad_norm": 2.8112082481384277,
+      "learning_rate": 2.2898717557161554e-07,
+      "loss": 2.3968,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9703157808061262,
+      "grad_norm": 2.9608657360076904,
+      "learning_rate": 2.2398588275618914e-07,
+      "loss": 2.1058,
+      "step": 2946
+    },
+    {
+      "epoch": 0.970645148009387,
+      "grad_norm": 2.786487340927124,
+      "learning_rate": 2.1903968849653623e-07,
+      "loss": 1.85,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9709745152126477,
+      "grad_norm": 2.775702714920044,
+      "learning_rate": 2.1414859826776045e-07,
+      "loss": 1.7929,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9713038824159085,
+      "grad_norm": 2.9629390239715576,
+      "learning_rate": 2.0931261748395326e-07,
+      "loss": 2.2975,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9716332496191692,
+      "grad_norm": 2.814885139465332,
+      "learning_rate": 2.045317514982048e-07,
+      "loss": 1.4608,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9719626168224299,
+      "grad_norm": 1.88387131690979,
+      "learning_rate": 1.9980600560259854e-07,
+      "loss": 2.3342,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9722919840256906,
+      "grad_norm": 2.205899238586426,
+      "learning_rate": 1.9513538502820562e-07,
+      "loss": 2.3497,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9726213512289513,
+      "grad_norm": 2.299445867538452,
+      "learning_rate": 1.9051989494506817e-07,
+      "loss": 2.3363,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9729507184322121,
+      "grad_norm": 2.064168930053711,
+      "learning_rate": 1.8595954046222165e-07,
+      "loss": 2.2905,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9732800856354729,
+      "grad_norm": 2.4672629833221436,
+      "learning_rate": 1.814543266276447e-07,
+      "loss": 2.4163,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9736094528387336,
+      "grad_norm": 2.261767864227295,
+      "learning_rate": 1.7700425842830382e-07,
+      "loss": 2.3594,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9739388200419943,
+      "grad_norm": 2.067182779312134,
+      "learning_rate": 1.726093407901086e-07,
+      "loss": 2.0535,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9742681872452551,
+      "grad_norm": 2.835082769393921,
+      "learning_rate": 1.682695785779287e-07,
+      "loss": 2.2098,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9745975544485158,
+      "grad_norm": 2.3787896633148193,
+      "learning_rate": 1.639849765955659e-07,
+      "loss": 2.0753,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9749269216517765,
+      "grad_norm": 2.4231419563293457,
+      "learning_rate": 1.5975553958578193e-07,
+      "loss": 2.0069,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9752562888550372,
+      "grad_norm": 2.6847152709960938,
+      "learning_rate": 1.5558127223027075e-07,
+      "loss": 2.346,
+      "step": 2961
+    },
+    {
+      "epoch": 0.975585656058298,
+      "grad_norm": 2.533125638961792,
+      "learning_rate": 1.514621791496418e-07,
+      "loss": 2.0978,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9759150232615588,
+      "grad_norm": 2.544414520263672,
+      "learning_rate": 1.4739826490345332e-07,
+      "loss": 1.8396,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9762443904648195,
+      "grad_norm": 2.6801774501800537,
+      "learning_rate": 1.4338953399016808e-07,
+      "loss": 2.0108,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9765737576680802,
+      "grad_norm": 2.523963212966919,
+      "learning_rate": 1.3943599084716984e-07,
+      "loss": 2.0665,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9769031248713409,
+      "grad_norm": 2.6162335872650146,
+      "learning_rate": 1.3553763985075795e-07,
+      "loss": 2.0451,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9772324920746017,
+      "grad_norm": 2.9109017848968506,
+      "learning_rate": 1.3169448531612506e-07,
+      "loss": 1.9372,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9775618592778624,
+      "grad_norm": 2.5702192783355713,
+      "learning_rate": 1.279065314973793e-07,
+      "loss": 2.0518,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9778912264811231,
+      "grad_norm": 2.58215069770813,
+      "learning_rate": 1.2417378258751665e-07,
+      "loss": 1.9024,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9782205936843839,
+      "grad_norm": 2.6025025844573975,
+      "learning_rate": 1.2049624271843752e-07,
+      "loss": 1.7492,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9785499608876446,
+      "grad_norm": 3.015721559524536,
+      "learning_rate": 1.1687391596090779e-07,
+      "loss": 2.3415,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9788793280909054,
+      "grad_norm": 2.538717746734619,
+      "learning_rate": 1.133068063245979e-07,
+      "loss": 1.6529,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9792086952941661,
+      "grad_norm": 3.052213191986084,
+      "learning_rate": 1.0979491775804373e-07,
+      "loss": 2.429,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9795380624974268,
+      "grad_norm": 3.415714740753174,
+      "learning_rate": 1.0633825414865794e-07,
+      "loss": 1.8479,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9798674297006875,
+      "grad_norm": 2.9711825847625732,
+      "learning_rate": 1.029368193227298e-07,
+      "loss": 1.6421,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9801967969039483,
+      "grad_norm": 2.3700506687164307,
+      "learning_rate": 9.959061704540862e-08,
+      "loss": 2.5403,
+      "step": 2976
+    },
+    {
+      "epoch": 0.980526164107209,
+      "grad_norm": 2.2373926639556885,
+      "learning_rate": 9.629965102070371e-08,
+      "loss": 2.4728,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9808555313104698,
+      "grad_norm": 2.5330634117126465,
+      "learning_rate": 9.306392489147886e-08,
+      "loss": 2.3086,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9811848985137305,
+      "grad_norm": 2.203402042388916,
+      "learning_rate": 8.988344223946343e-08,
+      "loss": 2.0723,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9815142657169912,
+      "grad_norm": 2.7258424758911133,
+      "learning_rate": 8.67582065852246e-08,
+      "loss": 2.3731,
+      "step": 2980
+    },
+    {
+      "epoch": 0.981843632920252,
+      "grad_norm": 2.4026784896850586,
+      "learning_rate": 8.3688221388184e-08,
+      "loss": 1.9317,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9821730001235127,
+      "grad_norm": 2.536128044128418,
+      "learning_rate": 8.067349004658998e-08,
+      "loss": 2.2797,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9825023673267734,
+      "grad_norm": 2.5913748741149902,
+      "learning_rate": 7.77140158975509e-08,
+      "loss": 2.2726,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9828317345300341,
+      "grad_norm": 2.1963465213775635,
+      "learning_rate": 7.480980221699075e-08,
+      "loss": 1.9628,
+      "step": 2984
+    },
+    {
+      "epoch": 0.983161101733295,
+      "grad_norm": 2.823356866836548,
+      "learning_rate": 7.196085221966575e-08,
+      "loss": 2.0853,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9834904689365557,
+      "grad_norm": 2.5668678283691406,
+      "learning_rate": 6.916716905916998e-08,
+      "loss": 2.3201,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9838198361398164,
+      "grad_norm": 2.500898599624634,
+      "learning_rate": 6.642875582791863e-08,
+      "loss": 2.2272,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9841492033430771,
+      "grad_norm": 2.5262458324432373,
+      "learning_rate": 6.374561555713143e-08,
+      "loss": 2.0321,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9844785705463378,
+      "grad_norm": 2.6859130859375,
+      "learning_rate": 6.111775121686037e-08,
+      "loss": 2.1567,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9848079377495986,
+      "grad_norm": 2.619617462158203,
+      "learning_rate": 5.854516571597857e-08,
+      "loss": 1.9454,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9851373049528593,
+      "grad_norm": 3.04783034324646,
+      "learning_rate": 5.602786190214149e-08,
+      "loss": 2.4391,
+      "step": 2991
+    },
+    {
+      "epoch": 0.98546667215612,
+      "grad_norm": 2.7410717010498047,
+      "learning_rate": 5.3565842561831283e-08,
+      "loss": 2.2959,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9857960393593808,
+      "grad_norm": 2.6359169483184814,
+      "learning_rate": 5.115911042034016e-08,
+      "loss": 1.8954,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9861254065626415,
+      "grad_norm": 2.6837198734283447,
+      "learning_rate": 4.880766814174265e-08,
+      "loss": 2.0915,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9864547737659023,
+      "grad_norm": 2.4030003547668457,
+      "learning_rate": 4.651151832892886e-08,
+      "loss": 1.6885,
+      "step": 2995
+    },
+    {
+      "epoch": 0.986784140969163,
+      "grad_norm": 2.9259960651397705,
+      "learning_rate": 4.427066352356568e-08,
+      "loss": 2.1477,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9871135081724237,
+      "grad_norm": 3.3638339042663574,
+      "learning_rate": 4.208510620612449e-08,
+      "loss": 2.1854,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9874428753756844,
+      "grad_norm": 3.0315308570861816,
+      "learning_rate": 3.995484879587008e-08,
+      "loss": 1.9279,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9877722425789452,
+      "grad_norm": 3.3093647956848145,
+      "learning_rate": 3.787989365083844e-08,
+      "loss": 1.8074,
+      "step": 2999
+    },
+    {
+      "epoch": 0.988101609782206,
+      "grad_norm": 3.9945030212402344,
+      "learning_rate": 3.5860243067870056e-08,
+      "loss": 1.9964,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9884309769854667,
+      "grad_norm": 1.813816785812378,
+      "learning_rate": 3.3895899282571083e-08,
+      "loss": 2.3394,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9887603441887274,
+      "grad_norm": 1.9609613418579102,
+      "learning_rate": 3.198686446932997e-08,
+      "loss": 2.1147,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9890897113919881,
+      "grad_norm": 2.0877554416656494,
+      "learning_rate": 3.013314074131746e-08,
+      "loss": 2.2933,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9894190785952489,
+      "grad_norm": 2.293480396270752,
+      "learning_rate": 2.8334730150475498e-08,
+      "loss": 2.3962,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9897484457985096,
+      "grad_norm": 2.9149646759033203,
+      "learning_rate": 2.659163468751724e-08,
+      "loss": 2.2562,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9900778130017703,
+      "grad_norm": 2.495413064956665,
+      "learning_rate": 2.490385628192704e-08,
+      "loss": 2.2912,
+      "step": 3006
+    },
+    {
+      "epoch": 0.990407180205031,
+      "grad_norm": 2.356163501739502,
+      "learning_rate": 2.3271396801960442e-08,
+      "loss": 2.107,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9907365474082919,
+      "grad_norm": 2.2497031688690186,
+      "learning_rate": 2.1694258054633097e-08,
+      "loss": 2.5133,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9910659146115526,
+      "grad_norm": 2.5045933723449707,
+      "learning_rate": 2.0172441785726302e-08,
+      "loss": 2.1748,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9913952818148133,
+      "grad_norm": 2.6602587699890137,
+      "learning_rate": 1.8705949679781454e-08,
+      "loss": 1.9129,
+      "step": 3010
+    },
+    {
+      "epoch": 0.991724649018074,
+      "grad_norm": 2.7580525875091553,
+      "learning_rate": 1.7294783360105593e-08,
+      "loss": 2.117,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9920540162213347,
+      "grad_norm": 2.4535562992095947,
+      "learning_rate": 1.5938944388765865e-08,
+      "loss": 2.0893,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9923833834245955,
+      "grad_norm": 2.4854702949523926,
+      "learning_rate": 1.4638434266572853e-08,
+      "loss": 1.9715,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9927127506278562,
+      "grad_norm": 2.8677029609680176,
+      "learning_rate": 1.3393254433102798e-08,
+      "loss": 2.3892,
+      "step": 3014
+    },
+    {
+      "epoch": 0.993042117831117,
+      "grad_norm": 2.844385862350464,
+      "learning_rate": 1.2203406266680927e-08,
+      "loss": 2.2286,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9933714850343777,
+      "grad_norm": 2.4216244220733643,
+      "learning_rate": 1.1068891084392574e-08,
+      "loss": 2.004,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9937008522376385,
+      "grad_norm": 2.4584288597106934,
+      "learning_rate": 9.989710142055408e-09,
+      "loss": 2.2331,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9940302194408992,
+      "grad_norm": 2.5853490829467773,
+      "learning_rate": 8.965864634252752e-09,
+      "loss": 1.6233,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9943595866441599,
+      "grad_norm": 2.8570823669433594,
+      "learning_rate": 7.997355694311371e-09,
+      "loss": 2.0152,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9946889538474206,
+      "grad_norm": 2.8787717819213867,
+      "learning_rate": 7.084184394307025e-09,
+      "loss": 1.9575,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9950183210506813,
+      "grad_norm": 2.658804416656494,
+      "learning_rate": 6.226351745042269e-09,
+      "loss": 2.1558,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9953476882539422,
+      "grad_norm": 2.7765133380889893,
+      "learning_rate": 5.423858696090855e-09,
+      "loss": 1.8984,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9956770554572029,
+      "grad_norm": 2.8934073448181152,
+      "learning_rate": 4.676706135747777e-09,
+      "loss": 2.0635,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9960064226604636,
+      "grad_norm": 2.6557319164276123,
+      "learning_rate": 3.984894891062574e-09,
+      "loss": 1.7287,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9963357898637243,
+      "grad_norm": 3.173056125640869,
+      "learning_rate": 3.3484257278226795e-09,
+      "loss": 1.834,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9966651570669851,
+      "grad_norm": 2.187110424041748,
+      "learning_rate": 2.7672993505534206e-09,
+      "loss": 2.694,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9969945242702458,
+      "grad_norm": 2.553576707839966,
+      "learning_rate": 2.2415164025235693e-09,
+      "loss": 2.3846,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9973238914735065,
+      "grad_norm": 2.6559970378875732,
+      "learning_rate": 1.7710774657286876e-09,
+      "loss": 2.021,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9976532586767672,
+      "grad_norm": 2.1335747241973877,
+      "learning_rate": 1.3559830609244373e-09,
+      "loss": 2.1005,
+      "step": 3029
+    },
+    {
+      "epoch": 0.997982625880028,
+      "grad_norm": 2.4700515270233154,
+      "learning_rate": 9.962336475821677e-10,
+      "loss": 2.14,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9983119930832888,
+      "grad_norm": 2.239428758621216,
+      "learning_rate": 6.918296239222244e-10,
+      "loss": 1.8967,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9986413602865495,
+      "grad_norm": 3.131831169128418,
+      "learning_rate": 4.427713268972955e-10,
+      "loss": 1.9217,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9989707274898102,
+      "grad_norm": 2.660928726196289,
+      "learning_rate": 2.4905903219796244e-10,
+      "loss": 2.2264,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9993000946930709,
+      "grad_norm": 2.3439199924468994,
+      "learning_rate": 1.1069295425270021e-10,
+      "loss": 1.7058,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9996294618963317,
+      "grad_norm": 2.770085573196411,
+      "learning_rate": 2.7673246222326144e-11,
+      "loss": 1.6823,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9999588290995924,
+      "grad_norm": 3.2446770668029785,
+      "learning_rate": 0.0,
+      "loss": 2.2193,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9999588290995924,
+      "eval_loss": 2.1753156185150146,
+      "eval_runtime": 761.5568,
+      "eval_samples_per_second": 3.358,
+      "eval_steps_per_second": 1.679,
+      "step": 3036
     }
   ],
   "logging_steps": 1,
@@ -15992,12 +21313,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 3.492436769658372e+18,
+  "total_flos": 4.656582359544496e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null