diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.24999189937074312,
+  "epoch": 0.9999675974829725,
   "eval_steps": 2411,
-  "global_step": 2411,
+  "global_step": 9644,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -16900,6 +16900,50661 @@
       "eval_samples_per_second": 339.196,
       "eval_steps_per_second": 15.935,
       "step": 2411
+    },
+    {
+      "epoch": 0.2500955874252312,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019661550413017746,
+      "loss": 4.5607,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25019927547971926,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019661270150023348,
+      "loss": 4.5456,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25030296353420733,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019660989773036174,
+      "loss": 4.5299,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2504066515886954,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019660709282059532,
+      "loss": 4.4934,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2505103396431835,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019660428677096734,
+      "loss": 4.5475,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25061402769767155,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001966014795815109,
+      "loss": 4.5659,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2507177157521596,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001965986712522591,
+      "loss": 4.5103,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2508214038066477,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019659586178324506,
+      "loss": 4.5531,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25092509186113576,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000196593051174502,
+      "loss": 4.5444,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25102877991562383,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019659023942606303,
+      "loss": 4.554,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2511324679701119,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019658742653796133,
+      "loss": 4.5388,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2512361560246,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019658461251023012,
+      "loss": 4.554,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25133984407908805,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019658179734290254,
+      "loss": 4.5397,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2514435321335761,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019657898103601185,
+      "loss": 4.5455,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2515472201880642,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019657616358959126,
+      "loss": 4.5079,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25165090824255226,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019657334500367406,
+      "loss": 4.5894,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25175459629704033,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019657052527829346,
+      "loss": 4.4915,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2518582843515284,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019656770441348273,
+      "loss": 4.5516,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2519619724060165,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019656488240927516,
+      "loss": 4.5341,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25206566046050455,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019656205926570406,
+      "loss": 4.5455,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2521693485149926,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019655923498280273,
+      "loss": 4.5265,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2522730365694807,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001965564095606045,
+      "loss": 4.5306,
+      "step": 2433
+    },
+    {
+      "epoch": 0.25237672462396876,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001965535829991427,
+      "loss": 4.5759,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2524804126784569,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001965507552984507,
+      "loss": 4.5264,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25258410073294496,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001965479264585618,
+      "loss": 4.5116,
+      "step": 2436
+    },
+    {
+      "epoch": 0.25268778878743303,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019654509647950945,
+      "loss": 4.4965,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2527914768419211,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000196542265361327,
+      "loss": 4.5255,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2528951648964092,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001965394331040479,
+      "loss": 4.5525,
+      "step": 2439
+    },
+    {
+      "epoch": 0.25299885295089725,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001965365997077055,
+      "loss": 4.5402,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2531025410053853,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019653376517233327,
+      "loss": 4.5721,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2532062290598734,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019653092949796467,
+      "loss": 4.538,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25330991711436146,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019652809268463315,
+      "loss": 4.5701,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25341360516884953,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019652525473237215,
+      "loss": 4.5366,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2535172932233376,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019652241564121518,
+      "loss": 4.5249,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2536209812778257,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019651957541119575,
+      "loss": 4.5277,
+      "step": 2446
+    },
+    {
+      "epoch": 0.25372466933231375,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019651673404234732,
+      "loss": 4.5749,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2538283573868018,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019651389153470348,
+      "loss": 4.5396,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2539320454412899,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019651104788829775,
+      "loss": 4.5754,
+      "step": 2449
+    },
+    {
+      "epoch": 0.25403573349577796,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019650820310316362,
+      "loss": 4.5094,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25413942155026603,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019650535717933475,
+      "loss": 4.5006,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2542431096047541,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001965025101168447,
+      "loss": 4.5539,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2543467976592422,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019649966191572697,
+      "loss": 4.4919,
+      "step": 2453
+    },
+    {
+      "epoch": 0.25445048571373025,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001964968125760153,
+      "loss": 4.5391,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2545541737682183,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001964939620977432,
+      "loss": 4.5305,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2546578618227064,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019649111048094435,
+      "loss": 4.5178,
+      "step": 2456
+    },
+    {
+      "epoch": 0.25476154987719446,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019648825772565238,
+      "loss": 4.545,
+      "step": 2457
+    },
+    {
+      "epoch": 0.25486523793168253,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019648540383190098,
+      "loss": 4.5473,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2549689259861706,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001964825487997238,
+      "loss": 4.5182,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2550726140406587,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019647969262915455,
+      "loss": 4.517,
+      "step": 2460
+    },
+    {
+      "epoch": 0.25517630209514675,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019647683532022692,
+      "loss": 4.5324,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2552799901496348,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019647397687297456,
+      "loss": 4.5349,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2553836782041229,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001964711172874313,
+      "loss": 4.5047,
+      "step": 2463
+    },
+    {
+      "epoch": 0.25548736625861096,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001964682565636308,
+      "loss": 4.5019,
+      "step": 2464
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019646539470160684,
+      "loss": 4.5242,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2556947423675871,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001964625317013932,
+      "loss": 4.5476,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2557984304220752,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019645966756302367,
+      "loss": 4.5218,
+      "step": 2467
+    },
+    {
+      "epoch": 0.25590211847656325,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019645680228653204,
+      "loss": 4.5236,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2560058065310513,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019645393587195204,
+      "loss": 4.5162,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2561094945855394,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001964510683193176,
+      "loss": 4.5434,
+      "step": 2470
+    },
+    {
+      "epoch": 0.25621318264002746,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001964481996286625,
+      "loss": 4.4977,
+      "step": 2471
+    },
+    {
+      "epoch": 0.25631687069451553,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019644532980002058,
+      "loss": 4.5071,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2564205587490036,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019644245883342572,
+      "loss": 4.4715,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2565242468034917,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019643958672891181,
+      "loss": 4.5657,
+      "step": 2474
+    },
+    {
+      "epoch": 0.25662793485797974,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019643671348651268,
+      "loss": 4.518,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2567316229124678,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001964338391062623,
+      "loss": 4.5594,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2568353109669559,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019643096358819455,
+      "loss": 4.5379,
+      "step": 2477
+    },
+    {
+      "epoch": 0.25693899902144396,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019642808693234333,
+      "loss": 4.4973,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2570426870759321,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019642520913874264,
+      "loss": 4.5271,
+      "step": 2479
+    },
+    {
+      "epoch": 0.25714637513042016,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001964223302074264,
+      "loss": 4.4797,
+      "step": 2480
+    },
+    {
+      "epoch": 0.25725006318490823,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019641945013842862,
+      "loss": 4.5094,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2573537512393963,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001964165689317832,
+      "loss": 4.5076,
+      "step": 2482
+    },
+    {
+      "epoch": 0.25745743929388437,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001964136865875242,
+      "loss": 4.5485,
+      "step": 2483
+    },
+    {
+      "epoch": 0.25756112734837244,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019641080310568563,
+      "loss": 4.55,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2576648154028605,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019640791848630148,
+      "loss": 4.4889,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2577685034573486,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001964050327294058,
+      "loss": 4.5108,
+      "step": 2486
+    },
+    {
+      "epoch": 0.25787219151183666,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019640214583503264,
+      "loss": 4.5616,
+      "step": 2487
+    },
+    {
+      "epoch": 0.25797587956632473,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019639925780321607,
+      "loss": 4.5208,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2580795676208128,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001963963686339901,
+      "loss": 4.5417,
+      "step": 2489
+    },
+    {
+      "epoch": 0.25818325567530087,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019639347832738896,
+      "loss": 4.5336,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25828694372978894,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019639058688344663,
+      "loss": 4.5381,
+      "step": 2491
+    },
+    {
+      "epoch": 0.258390631784277,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019638769430219727,
+      "loss": 4.5333,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2584943198387651,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019638480058367498,
+      "loss": 4.5055,
+      "step": 2493
+    },
+    {
+      "epoch": 0.25859800789325316,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019638190572791392,
+      "loss": 4.5351,
+      "step": 2494
+    },
+    {
+      "epoch": 0.25870169594774123,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019637900973494828,
+      "loss": 4.5316,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2588053840022293,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019637611260481223,
+      "loss": 4.545,
+      "step": 2496
+    },
+    {
+      "epoch": 0.25890907205671737,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019637321433753986,
+      "loss": 4.5428,
+      "step": 2497
+    },
+    {
+      "epoch": 0.25901276011120544,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019637031493316548,
+      "loss": 4.5547,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2591164481656935,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019636741439172322,
+      "loss": 4.5579,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2592201362201816,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019636451271324735,
+      "loss": 4.5162,
+      "step": 2500
+    },
+    {
+      "epoch": 0.25932382427466966,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019636160989777208,
+      "loss": 4.5121,
+      "step": 2501
+    },
+    {
+      "epoch": 0.25942751232915773,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001963587059453317,
+      "loss": 4.5258,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2595312003836458,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001963558008559604,
+      "loss": 4.5086,
+      "step": 2503
+    },
+    {
+      "epoch": 0.25963488843813387,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019635289462969255,
+      "loss": 4.5275,
+      "step": 2504
+    },
+    {
+      "epoch": 0.25973857649262194,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019634998726656238,
+      "loss": 4.5625,
+      "step": 2505
+    },
+    {
+      "epoch": 0.25984226454711,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001963470787666042,
+      "loss": 4.5497,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2599459526015981,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019634416912985234,
+      "loss": 4.4894,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26004964065608616,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001963412583563411,
+      "loss": 4.4685,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2601533287105742,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019633834644610488,
+      "loss": 4.5355,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2602570167650623,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000196335433399178,
+      "loss": 4.5377,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26036070481955037,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019633251921559482,
+      "loss": 4.5568,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26046439287403844,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019632960389538975,
+      "loss": 4.5713,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2605680809285265,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019632668743859718,
+      "loss": 4.5268,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2606717689830146,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019632376984525155,
+      "loss": 4.5509,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26077545703750266,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001963208511153872,
+      "loss": 4.5079,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2608791450919907,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019631793124903863,
+      "loss": 4.551,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2609828331464788,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019631501024624032,
+      "loss": 4.515,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26108652120096687,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019631208810702667,
+      "loss": 4.498,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26119020925545494,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001963091648314322,
+      "loss": 4.5372,
+      "step": 2519
+    },
+    {
+      "epoch": 0.261293897309943,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001963062404194914,
+      "loss": 4.513,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2613975853644311,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019630331487123872,
+      "loss": 4.497,
+      "step": 2521
+    },
+    {
+      "epoch": 0.26150127341891916,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019630038818670874,
+      "loss": 4.4994,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2616049614734072,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.000196297460365936,
+      "loss": 4.4981,
+      "step": 2523
+    },
+    {
+      "epoch": 0.26170864952789535,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000196294531408955,
+      "loss": 4.5548,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2618123375823834,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019629160131580032,
+      "loss": 4.5412,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2619160256368715,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019628867008650652,
+      "loss": 4.4908,
+      "step": 2526
+    },
+    {
+      "epoch": 0.26201971369135957,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019628573772110822,
+      "loss": 4.4642,
+      "step": 2527
+    },
+    {
+      "epoch": 0.26212340174584764,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019628280421963995,
+      "loss": 4.4949,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2622270898003357,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001962798695821364,
+      "loss": 4.4916,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2623307778548238,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019627693380863215,
+      "loss": 4.5176,
+      "step": 2530
+    },
+    {
+      "epoch": 0.26243446590931185,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019627399689916186,
+      "loss": 4.555,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2625381539637999,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019627105885376017,
+      "loss": 4.5436,
+      "step": 2532
+    },
+    {
+      "epoch": 0.262641842018288,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019626811967246173,
+      "loss": 4.509,
+      "step": 2533
+    },
+    {
+      "epoch": 0.26274553007277607,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019626517935530125,
+      "loss": 4.5512,
+      "step": 2534
+    },
+    {
+      "epoch": 0.26284921812726414,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001962622379023134,
+      "loss": 4.5567,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2629529061817522,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001962592953135329,
+      "loss": 4.5253,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2630565942362403,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001962563515889945,
+      "loss": 4.5262,
+      "step": 2537
+    },
+    {
+      "epoch": 0.26316028229072835,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019625340672873285,
+      "loss": 4.5034,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2632639703452164,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019625046073278276,
+      "loss": 4.5212,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2633676583997045,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019624751360117898,
+      "loss": 4.521,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26347134645419257,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019624456533395628,
+      "loss": 4.5292,
+      "step": 2541
+    },
+    {
+      "epoch": 0.26357503450868064,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019624161593114945,
+      "loss": 4.5043,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2636787225631687,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019623866539279327,
+      "loss": 4.4988,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2637824106176568,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019623571371892257,
+      "loss": 4.5159,
+      "step": 2544
+    },
+    {
+      "epoch": 0.26388609867214485,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019623276090957218,
+      "loss": 4.5345,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2639897867266329,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019622980696477692,
+      "loss": 4.5314,
+      "step": 2546
+    },
+    {
+      "epoch": 0.264093474781121,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019622685188457167,
+      "loss": 4.5287,
+      "step": 2547
+    },
+    {
+      "epoch": 0.26419716283560907,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001962238956689913,
+      "loss": 4.4609,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26430085089009714,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019622093831807064,
+      "loss": 4.5249,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2644045389445852,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019621797983184464,
+      "loss": 4.5609,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2645082269990733,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001962150202103482,
+      "loss": 4.526,
+      "step": 2551
+    },
+    {
+      "epoch": 0.26461191505356135,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019621205945361618,
+      "loss": 4.4839,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2647156031080494,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019620909756168356,
+      "loss": 4.487,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2648192911625375,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001962061345345853,
+      "loss": 4.5395,
+      "step": 2554
+    },
+    {
+      "epoch": 0.26492297921702557,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019620317037235638,
+      "loss": 4.5355,
+      "step": 2555
+    },
+    {
+      "epoch": 0.26502666727151364,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019620020507503174,
+      "loss": 4.4756,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2651303553260017,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001961972386426463,
+      "loss": 4.5216,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2652340433804898,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019619427107523521,
+      "loss": 4.4505,
+      "step": 2558
+    },
+    {
+      "epoch": 0.26533773143497785,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019619130237283336,
+      "loss": 4.5255,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2654414194894659,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019618833253547583,
+      "loss": 4.52,
+      "step": 2560
+    },
+    {
+      "epoch": 0.265545107543954,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019618536156319766,
+      "loss": 4.5316,
+      "step": 2561
+    },
+    {
+      "epoch": 0.26564879559844207,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019618238945603387,
+      "loss": 4.5438,
+      "step": 2562
+    },
+    {
+      "epoch": 0.26575248365293014,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019617941621401957,
+      "loss": 4.4668,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2658561717074182,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019617644183718983,
+      "loss": 4.5624,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2659598597619063,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019617346632557972,
+      "loss": 4.4884,
+      "step": 2565
+    },
+    {
+      "epoch": 0.26606354781639435,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019617048967922438,
+      "loss": 4.5099,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2661672358708824,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019616751189815892,
+      "loss": 4.5287,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2662709239253705,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019616453298241845,
+      "loss": 4.5201,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2663746119798586,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019616155293203816,
+      "loss": 4.5195,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2664783000343467,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001961585717470532,
+      "loss": 4.5095,
+      "step": 2570
+    },
+    {
+      "epoch": 0.26658198808883476,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019615558942749872,
+      "loss": 4.4663,
+      "step": 2571
+    },
+    {
+      "epoch": 0.26668567614332284,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019615260597340996,
+      "loss": 4.5458,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2667893641978109,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019614962138482205,
+      "loss": 4.5008,
+      "step": 2573
+    },
+    {
+      "epoch": 0.266893052252299,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019614663566177028,
+      "loss": 4.4962,
+      "step": 2574
+    },
+    {
+      "epoch": 0.26699674030678705,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001961436488042898,
+      "loss": 4.5274,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2671004283612751,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001961406608124159,
+      "loss": 4.5169,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2672041164157632,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019613767168618384,
+      "loss": 4.498,
+      "step": 2577
+    },
+    {
+      "epoch": 0.26730780447025126,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019613468142562888,
+      "loss": 4.5055,
+      "step": 2578
+    },
+    {
+      "epoch": 0.26741149252473934,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019613169003078628,
+      "loss": 4.4896,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2675151805792274,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001961286975016914,
+      "loss": 4.5204,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2676188686337155,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019612570383837943,
+      "loss": 4.4656,
+      "step": 2581
+    },
+    {
+      "epoch": 0.26772255668820355,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019612270904088578,
+      "loss": 4.5293,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2678262447426916,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001961197131092458,
+      "loss": 4.5334,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2679299327971797,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019611671604349482,
+      "loss": 4.5137,
+      "step": 2584
+    },
+    {
+      "epoch": 0.26803362085166776,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019611371784366815,
+      "loss": 4.5113,
+      "step": 2585
+    },
+    {
+      "epoch": 0.26813730890615584,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019611071850980123,
+      "loss": 4.5013,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2682409969606439,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019610771804192943,
+      "loss": 4.498,
+      "step": 2587
+    },
+    {
+      "epoch": 0.268344685015132,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019610471644008811,
+      "loss": 4.5094,
+      "step": 2588
+    },
+    {
+      "epoch": 0.26844837306962005,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019610171370431276,
+      "loss": 4.5432,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2685520611241081,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019609870983463876,
+      "loss": 4.5388,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2686557491785962,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019609570483110158,
+      "loss": 4.5287,
+      "step": 2591
+    },
+    {
+      "epoch": 0.26875943723308426,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019609269869373663,
+      "loss": 4.4932,
+      "step": 2592
+    },
+    {
+      "epoch": 0.26886312528757234,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019608969142257941,
+      "loss": 4.4847,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2689668133420604,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019608668301766545,
+      "loss": 4.5315,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2690705013965485,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019608367347903017,
+      "loss": 4.5097,
+      "step": 2595
+    },
+    {
+      "epoch": 0.26917418945103655,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001960806628067091,
+      "loss": 4.501,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2692778775055246,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001960776510007378,
+      "loss": 4.4971,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2693815655600127,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001960746380611517,
+      "loss": 4.5445,
+      "step": 2598
+    },
+    {
+      "epoch": 0.26948525361450076,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001960716239879865,
+      "loss": 4.511,
+      "step": 2599
+    },
+    {
+      "epoch": 0.26958894166898884,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019606860878127767,
+      "loss": 4.5239,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2696926297234769,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001960655924410608,
+      "loss": 4.5162,
+      "step": 2601
+    },
+    {
+      "epoch": 0.269796317777965,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019606257496737148,
+      "loss": 4.511,
+      "step": 2602
+    },
+    {
+      "epoch": 0.26990000583245305,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019605955636024534,
+      "loss": 4.4955,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2700036938869411,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019605653661971796,
+      "loss": 4.5352,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2701073819414292,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019605351574582497,
+      "loss": 4.5061,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27021106999591726,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019605049373860205,
+      "loss": 4.4833,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27031475805040533,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001960474705980848,
+      "loss": 4.5036,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2704184461048934,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019604444632430895,
+      "loss": 4.4686,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2705221341593815,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019604142091731016,
+      "loss": 4.5039,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27062582221386955,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019603839437712413,
+      "loss": 4.4556,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2707295102683576,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019603536670378656,
+      "loss": 4.4963,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2708331983228457,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019603233789733317,
+      "loss": 4.4963,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2709368863773338,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019602930795779969,
+      "loss": 4.5181,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2710405744318219,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019602627688522191,
+      "loss": 4.5117,
+      "step": 2614
+    },
+    {
+      "epoch": 0.27114426248630996,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019602324467963555,
+      "loss": 4.5349,
+      "step": 2615
+    },
+    {
+      "epoch": 0.27124795054079803,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001960202113410764,
+      "loss": 4.532,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2713516385952861,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019601717686958025,
+      "loss": 4.4892,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2714553266497742,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019601414126518293,
+      "loss": 4.468,
+      "step": 2618
+    },
+    {
+      "epoch": 0.27155901470426225,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019601110452792023,
+      "loss": 4.4757,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2716627027587503,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019600806665782795,
+      "loss": 4.5217,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2717663908132384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019600502765494202,
+      "loss": 4.4974,
+      "step": 2621
+    },
+    {
+      "epoch": 0.27187007886772646,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019600198751929822,
+      "loss": 4.4916,
+      "step": 2622
+    },
+    {
+      "epoch": 0.27197376692221453,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019599894625093244,
+      "loss": 4.5144,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2720774549767026,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019599590384988057,
+      "loss": 4.5205,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2721811430311907,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019599286031617852,
+      "loss": 4.5329,
+      "step": 2625
+    },
+    {
+      "epoch": 0.27228483108567875,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019598981564986217,
+      "loss": 4.5093,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2723885191401668,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019598676985096747,
+      "loss": 4.4862,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2724922071946549,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019598372291953036,
+      "loss": 4.5171,
+      "step": 2628
+    },
+    {
+      "epoch": 0.27259589524914296,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019598067485558675,
+      "loss": 4.5127,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27269958330363103,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019597762565917265,
+      "loss": 4.4947,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2728032713581191,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.000195974575330324,
+      "loss": 4.4842,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2729069594126072,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019597152386907683,
+      "loss": 4.5063,
+      "step": 2632
+    },
+    {
+      "epoch": 0.27301064746709525,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019596847127546717,
+      "loss": 4.4661,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2731143355215833,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001959654175495309,
+      "loss": 4.5253,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2732180235760714,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001959623626913042,
+      "loss": 4.5194,
+      "step": 2635
+    },
+    {
+      "epoch": 0.27332171163055946,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019595930670082305,
+      "loss": 4.5654,
+      "step": 2636
+    },
+    {
+      "epoch": 0.27342539968504753,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019595624957812353,
+      "loss": 4.5322,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2735290877395356,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019595319132324166,
+      "loss": 4.5138,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2736327757940237,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019595013193621357,
+      "loss": 4.4681,
+      "step": 2639
+    },
+    {
+      "epoch": 0.27373646384851175,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019594707141707535,
+      "loss": 4.5471,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2738401519029998,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019594400976586315,
+      "loss": 4.5668,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2739438399574879,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019594094698261298,
+      "loss": 4.5293,
+      "step": 2642
+    },
+    {
+      "epoch": 0.27404752801197596,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001959378830673611,
+      "loss": 4.5093,
+      "step": 2643
+    },
+    {
+      "epoch": 0.27415121606646403,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019593481802014358,
+      "loss": 4.5032,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2742549041209521,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019593175184099662,
+      "loss": 4.5133,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2743585921754402,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001959286845299564,
+      "loss": 4.5058,
+      "step": 2646
+    },
+    {
+      "epoch": 0.27446228022992825,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019592561608705911,
+      "loss": 4.5501,
+      "step": 2647
+    },
+    {
+      "epoch": 0.2745659682844163,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019592254651234093,
+      "loss": 4.5248,
+      "step": 2648
+    },
+    {
+      "epoch": 0.2746696563389044,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001959194758058381,
+      "loss": 4.4747,
+      "step": 2649
+    },
+    {
+      "epoch": 0.27477334439339246,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019591640396758687,
+      "loss": 4.5561,
+      "step": 2650
+    },
+    {
+      "epoch": 0.27487703244788053,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019591333099762344,
+      "loss": 4.5369,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2749807205023686,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019591025689598407,
+      "loss": 4.5263,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2750844085568567,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001959071816627051,
+      "loss": 4.5292,
+      "step": 2653
+    },
+    {
+      "epoch": 0.27518809661134475,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019590410529782273,
+      "loss": 4.5482,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2752917846658328,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019590102780137333,
+      "loss": 4.5152,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2753954727203209,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019589794917339312,
+      "loss": 4.5202,
+      "step": 2656
+    },
+    {
+      "epoch": 0.27549916077480896,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001958948694139185,
+      "loss": 4.5526,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2756028488292971,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001958917885229858,
+      "loss": 4.4628,
+      "step": 2658
+    },
+    {
+      "epoch": 0.27570653688378516,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019588870650063135,
+      "loss": 4.4477,
+      "step": 2659
+    },
+    {
+      "epoch": 0.27581022493827323,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001958856233468915,
+      "loss": 4.4959,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2759139129927613,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019588253906180266,
+      "loss": 4.4946,
+      "step": 2661
+    },
+    {
+      "epoch": 0.2760176010472494,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001958794536454012,
+      "loss": 4.4696,
+      "step": 2662
+    },
+    {
+      "epoch": 0.27612128910173744,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019587636709772358,
+      "loss": 4.522,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2762249771562255,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019587327941880615,
+      "loss": 4.483,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2763286652107136,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019587019060868536,
+      "loss": 4.5078,
+      "step": 2665
+    },
+    {
+      "epoch": 0.27643235326520166,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001958671006673977,
+      "loss": 4.4836,
+      "step": 2666
+    },
+    {
+      "epoch": 0.27653604131968973,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019586400959497954,
+      "loss": 4.4048,
+      "step": 2667
+    },
+    {
+      "epoch": 0.2766397293741778,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019586091739146738,
+      "loss": 4.5028,
+      "step": 2668
+    },
+    {
+      "epoch": 0.27674341742866587,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001958578240568978,
+      "loss": 4.4601,
+      "step": 2669
+    },
+    {
+      "epoch": 0.27684710548315394,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019585472959130715,
+      "loss": 4.4928,
+      "step": 2670
+    },
+    {
+      "epoch": 0.276950793537642,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019585163399473204,
+      "loss": 4.4804,
+      "step": 2671
+    },
+    {
+      "epoch": 0.2770544815921301,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019584853726720898,
+      "loss": 4.4653,
+      "step": 2672
+    },
+    {
+      "epoch": 0.27715816964661816,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001958454394087745,
+      "loss": 4.5314,
+      "step": 2673
+    },
+    {
+      "epoch": 0.27726185770110623,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001958423404194651,
+      "loss": 4.4906,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2773655457555943,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019583924029931745,
+      "loss": 4.4875,
+      "step": 2675
+    },
+    {
+      "epoch": 0.27746923381008237,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019583613904836805,
+      "loss": 4.4854,
+      "step": 2676
+    },
+    {
+      "epoch": 0.27757292186457044,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001958330366666535,
+      "loss": 4.5304,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2776766099190585,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019582993315421044,
+      "loss": 4.5095,
+      "step": 2678
+    },
+    {
+      "epoch": 0.2777802979735466,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019582682851107547,
+      "loss": 4.4651,
+      "step": 2679
+    },
+    {
+      "epoch": 0.27788398602803466,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019582372273728517,
+      "loss": 4.5261,
+      "step": 2680
+    },
+    {
+      "epoch": 0.27798767408252273,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019582061583287626,
+      "loss": 4.5165,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2780913621370108,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019581750779788538,
+      "loss": 4.511,
+      "step": 2682
+    },
+    {
+      "epoch": 0.27819505019149887,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019581439863234918,
+      "loss": 4.5259,
+      "step": 2683
+    },
+    {
+      "epoch": 0.27829873824598694,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019581128833630435,
+      "loss": 4.4464,
+      "step": 2684
+    },
+    {
+      "epoch": 0.278402426300475,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001958081769097876,
+      "loss": 4.5156,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2785061143549631,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019580506435283566,
+      "loss": 4.4865,
+      "step": 2686
+    },
+    {
+      "epoch": 0.27860980240945116,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019580195066548522,
+      "loss": 4.4432,
+      "step": 2687
+    },
+    {
+      "epoch": 0.27871349046393923,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000195798835847773,
+      "loss": 4.476,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2788171785184273,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019579571989973582,
+      "loss": 4.5375,
+      "step": 2689
+    },
+    {
+      "epoch": 0.27892086657291537,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019579260282141038,
+      "loss": 4.5139,
+      "step": 2690
+    },
+    {
+      "epoch": 0.27902455462740344,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001957894846128335,
+      "loss": 4.5164,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2791282426818915,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019578636527404193,
+      "loss": 4.4868,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2792319307363796,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019578324480507253,
+      "loss": 4.4798,
+      "step": 2693
+    },
+    {
+      "epoch": 0.27933561879086766,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001957801232059621,
+      "loss": 4.4708,
+      "step": 2694
+    },
+    {
+      "epoch": 0.27943930684535573,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019577700047674742,
+      "loss": 4.5079,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2795429948998438,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019577387661746538,
+      "loss": 4.508,
+      "step": 2696
+    },
+    {
+      "epoch": 0.27964668295433187,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019577075162815284,
+      "loss": 4.5315,
+      "step": 2697
+    },
+    {
+      "epoch": 0.27975037100881994,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019576762550884666,
+      "loss": 4.4953,
+      "step": 2698
+    },
+    {
+      "epoch": 0.279854059063308,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019576449825958373,
+      "loss": 4.4543,
+      "step": 2699
+    },
+    {
+      "epoch": 0.2799577471177961,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019576136988040096,
+      "loss": 4.4965,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28006143517228416,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019575824037133525,
+      "loss": 4.4783,
+      "step": 2701
+    },
+    {
+      "epoch": 0.28016512322677223,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019575510973242355,
+      "loss": 4.5024,
+      "step": 2702
+    },
+    {
+      "epoch": 0.28026881128126035,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019575197796370273,
+      "loss": 4.5101,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2803724993357484,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001957488450652098,
+      "loss": 4.469,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2804761873902365,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019574571103698172,
+      "loss": 4.533,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28057987544472457,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019574257587905543,
+      "loss": 4.4898,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28068356349921264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019573943959146797,
+      "loss": 4.5083,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2807872515537007,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001957363021742563,
+      "loss": 4.5129,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2808909396081888,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001957331636274575,
+      "loss": 4.4601,
+      "step": 2709
+    },
+    {
+      "epoch": 0.28099462766267685,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019573002395110854,
+      "loss": 4.4549,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2810983157171649,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019572688314524648,
+      "loss": 4.4636,
+      "step": 2711
+    },
+    {
+      "epoch": 0.281202003771653,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001957237412099084,
+      "loss": 4.5096,
+      "step": 2712
+    },
+    {
+      "epoch": 0.28130569182614107,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019572059814513136,
+      "loss": 4.4603,
+      "step": 2713
+    },
+    {
+      "epoch": 0.28140937988062914,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019571745395095244,
+      "loss": 4.4586,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2815130679351172,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019571430862740875,
+      "loss": 4.4631,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2816167559896053,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019571116217453736,
+      "loss": 4.528,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28172044404409335,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019570801459237543,
+      "loss": 4.5021,
+      "step": 2717
+    },
+    {
+      "epoch": 0.2818241320985814,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019570486588096014,
+      "loss": 4.4842,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2819278201530695,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019570171604032857,
+      "loss": 4.4669,
+      "step": 2719
+    },
+    {
+      "epoch": 0.28203150820755757,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019569856507051792,
+      "loss": 4.5322,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28213519626204564,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019569541297156535,
+      "loss": 4.4856,
+      "step": 2721
+    },
+    {
+      "epoch": 0.2822388843165337,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019569225974350806,
+      "loss": 4.4897,
+      "step": 2722
+    },
+    {
+      "epoch": 0.2823425723710218,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019568910538638327,
+      "loss": 4.5146,
+      "step": 2723
+    },
+    {
+      "epoch": 0.28244626042550985,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019568594990022816,
+      "loss": 4.5252,
+      "step": 2724
+    },
+    {
+      "epoch": 0.2825499484799979,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019568279328508,
+      "loss": 4.5108,
+      "step": 2725
+    },
+    {
+      "epoch": 0.282653636534486,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000195679635540976,
+      "loss": 4.4586,
+      "step": 2726
+    },
+    {
+      "epoch": 0.28275732458897407,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019567647666795347,
+      "loss": 4.4596,
+      "step": 2727
+    },
+    {
+      "epoch": 0.28286101264346214,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019567331666604963,
+      "loss": 4.506,
+      "step": 2728
+    },
+    {
+      "epoch": 0.2829647006979502,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019567015553530182,
+      "loss": 4.4883,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2830683887524383,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001956669932757473,
+      "loss": 4.4837,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28317207680692635,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019566382988742332,
+      "loss": 4.4863,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2832757648614144,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019566066537036734,
+      "loss": 4.5071,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2833794529159025,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019565749972461657,
+      "loss": 4.5131,
+      "step": 2733
+    },
+    {
+      "epoch": 0.28348314097039057,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019565433295020844,
+      "loss": 4.4927,
+      "step": 2734
+    },
+    {
+      "epoch": 0.28358682902487864,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001956511650471803,
+      "loss": 4.4971,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2836905170793667,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019564799601556955,
+      "loss": 4.4516,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2837942051338548,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019564482585541351,
+      "loss": 4.4599,
+      "step": 2737
+    },
+    {
+      "epoch": 0.28389789318834285,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001956416545667496,
+      "loss": 4.4894,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2840015812428309,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001956384821496153,
+      "loss": 4.5148,
+      "step": 2739
+    },
+    {
+      "epoch": 0.284105269297319,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019563530860404802,
+      "loss": 4.5061,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28420895735180707,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019563213393008515,
+      "loss": 4.4999,
+      "step": 2741
+    },
+    {
+      "epoch": 0.28431264540629514,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001956289581277642,
+      "loss": 4.5199,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2844163334607832,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019562578119712264,
+      "loss": 4.5523,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2845200215152713,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019562260313819795,
+      "loss": 4.4767,
+      "step": 2744
+    },
+    {
+      "epoch": 0.28462370956975935,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019561942395102762,
+      "loss": 4.5022,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2847273976242474,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019561624363564914,
+      "loss": 4.4705,
+      "step": 2746
+    },
+    {
+      "epoch": 0.28483108567873555,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019561306219210005,
+      "loss": 4.5097,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2849347737332236,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019560987962041792,
+      "loss": 4.4899,
+      "step": 2748
+    },
+    {
+      "epoch": 0.2850384617877117,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019560669592064026,
+      "loss": 4.4619,
+      "step": 2749
+    },
+    {
+      "epoch": 0.28514214984219977,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019560351109280467,
+      "loss": 4.4725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.28524583789668784,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019560032513694865,
+      "loss": 4.4604,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2853495259511759,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001955971380531099,
+      "loss": 4.5025,
+      "step": 2752
+    },
+    {
+      "epoch": 0.285453214005664,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019559394984132596,
+      "loss": 4.5058,
+      "step": 2753
+    },
+    {
+      "epoch": 0.28555690206015205,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019559076050163445,
+      "loss": 4.4853,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2856605901146401,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019558757003407303,
+      "loss": 4.4615,
+      "step": 2755
+    },
+    {
+      "epoch": 0.2857642781691282,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019558437843867932,
+      "loss": 4.4716,
+      "step": 2756
+    },
+    {
+      "epoch": 0.28586796622361627,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019558118571549096,
+      "loss": 4.4484,
+      "step": 2757
+    },
+    {
+      "epoch": 0.28597165427810434,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019557799186454566,
+      "loss": 4.5193,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2860753423325924,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019557479688588108,
+      "loss": 4.4887,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2861790303870805,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019557160077953491,
+      "loss": 4.4622,
+      "step": 2760
+    },
+    {
+      "epoch": 0.28628271844156855,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001955684035455449,
+      "loss": 4.4998,
+      "step": 2761
+    },
+    {
+      "epoch": 0.2863864064960566,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019556520518394875,
+      "loss": 4.5327,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2864900945505447,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019556200569478417,
+      "loss": 4.4883,
+      "step": 2763
+    },
+    {
+      "epoch": 0.28659378260503277,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019555880507808894,
+      "loss": 4.5244,
+      "step": 2764
+    },
+    {
+      "epoch": 0.28669747065952084,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019555560333390083,
+      "loss": 4.4726,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2868011587140089,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019555240046225763,
+      "loss": 4.457,
+      "step": 2766
+    },
+    {
+      "epoch": 0.286904846768497,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019554919646319708,
+      "loss": 4.4616,
+      "step": 2767
+    },
+    {
+      "epoch": 0.28700853482298505,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.000195545991336757,
+      "loss": 4.4876,
+      "step": 2768
+    },
+    {
+      "epoch": 0.2871122228774731,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019554278508297524,
+      "loss": 4.4903,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2872159109319612,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001955395777018896,
+      "loss": 4.5038,
+      "step": 2770
+    },
+    {
+      "epoch": 0.28731959898644926,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019553636919353794,
+      "loss": 4.4924,
+      "step": 2771
+    },
+    {
+      "epoch": 0.28742328704093734,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001955331595579581,
+      "loss": 4.4793,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2875269750954254,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019552994879518798,
+      "loss": 4.524,
+      "step": 2773
+    },
+    {
+      "epoch": 0.2876306631499135,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019552673690526544,
+      "loss": 4.5274,
+      "step": 2774
+    },
+    {
+      "epoch": 0.28773435120440155,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019552352388822834,
+      "loss": 4.4656,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2878380392588896,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001955203097441147,
+      "loss": 4.5221,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2879417273133777,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001955170944729623,
+      "loss": 4.4798,
+      "step": 2777
+    },
+    {
+      "epoch": 0.28804541536786576,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019551387807480918,
+      "loss": 4.425,
+      "step": 2778
+    },
+    {
+      "epoch": 0.28814910342235384,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001955106605496933,
+      "loss": 4.4931,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2882527914768419,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019550744189765254,
+      "loss": 4.5125,
+      "step": 2780
+    },
+    {
+      "epoch": 0.28835647953133,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019550422211872493,
+      "loss": 4.4673,
+      "step": 2781
+    },
+    {
+      "epoch": 0.28846016758581805,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019550100121294844,
+      "loss": 4.4965,
+      "step": 2782
+    },
+    {
+      "epoch": 0.2885638556403061,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019549777918036112,
+      "loss": 4.5221,
+      "step": 2783
+    },
+    {
+      "epoch": 0.2886675436947942,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019549455602100094,
+      "loss": 4.5056,
+      "step": 2784
+    },
+    {
+      "epoch": 0.28877123174928226,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019549133173490593,
+      "loss": 4.4957,
+      "step": 2785
+    },
+    {
+      "epoch": 0.28887491980377034,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019548810632211413,
+      "loss": 4.5048,
+      "step": 2786
+    },
+    {
+      "epoch": 0.2889786078582584,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019548487978266363,
+      "loss": 4.462,
+      "step": 2787
+    },
+    {
+      "epoch": 0.2890822959127465,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019548165211659248,
+      "loss": 4.4877,
+      "step": 2788
+    },
+    {
+      "epoch": 0.28918598396723455,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019547842332393874,
+      "loss": 4.513,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2892896720217226,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019547519340474057,
+      "loss": 4.4637,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2893933600762107,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019547196235903603,
+      "loss": 4.4883,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2894970481306988,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019546873018686322,
+      "loss": 4.4691,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2896007361851869,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019546549688826038,
+      "loss": 4.4894,
+      "step": 2793
+    },
+    {
+      "epoch": 0.28970442423967496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001954622624632655,
+      "loss": 4.5193,
+      "step": 2794
+    },
+    {
+      "epoch": 0.28980811229416303,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019545902691191688,
+      "loss": 4.5056,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2899118003486511,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019545579023425263,
+      "loss": 4.4285,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2900154884031392,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019545255243031098,
+      "loss": 4.4977,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29011917645762725,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001954493135001301,
+      "loss": 4.442,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2902228645121153,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001954460734437482,
+      "loss": 4.4989,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2903265525666034,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019544283226120352,
+      "loss": 4.5369,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29043024062109146,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019543958995253433,
+      "loss": 4.4906,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29053392867557953,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019543634651777886,
+      "loss": 4.489,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2906376167300676,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019543310195697537,
+      "loss": 4.472,
+      "step": 2803
+    },
+    {
+      "epoch": 0.2907413047845557,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019542985627016214,
+      "loss": 4.464,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29084499283904375,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019542660945737753,
+      "loss": 4.4578,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2909486808935318,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019542336151865975,
+      "loss": 4.5206,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2910523689480199,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019542011245404716,
+      "loss": 4.4985,
+      "step": 2807
+    },
+    {
+      "epoch": 0.29115605700250796,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019541686226357813,
+      "loss": 4.4695,
+      "step": 2808
+    },
+    {
+      "epoch": 0.29125974505699603,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019541361094729102,
+      "loss": 4.4816,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2913634331114841,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019541035850522412,
+      "loss": 4.4805,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2914671211659722,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019540710493741586,
+      "loss": 4.5198,
+      "step": 2811
+    },
+    {
+      "epoch": 0.29157080922046025,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019540385024390458,
+      "loss": 4.4809,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2916744972749483,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019540059442472874,
+      "loss": 4.4642,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2917781853294364,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00019539733747992672,
+      "loss": 4.5016,
+      "step": 2814
+    },
+    {
+      "epoch": 0.29188187338392446,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019539407940953696,
+      "loss": 4.4307,
+      "step": 2815
+    },
+    {
+      "epoch": 0.29198556143841253,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001953908202135979,
+      "loss": 4.4877,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2920892494929006,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000195387559892148,
+      "loss": 4.4839,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2921929375473887,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001953842984452257,
+      "loss": 4.4767,
+      "step": 2818
+    },
+    {
+      "epoch": 0.29229662560187675,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001953810358728695,
+      "loss": 4.4613,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2924003136563648,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019537777217511795,
+      "loss": 4.4937,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2925040017108529,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019537450735200947,
+      "loss": 4.4661,
+      "step": 2821
+    },
+    {
+      "epoch": 0.29260768976534096,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001953712414035826,
+      "loss": 4.5173,
+      "step": 2822
+    },
+    {
+      "epoch": 0.29271137781982903,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019536797432987592,
+      "loss": 4.5015,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2928150658743171,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019536470613092796,
+      "loss": 4.4875,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2929187539288052,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019536143680677726,
+      "loss": 4.464,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29302244198329325,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019535816635746241,
+      "loss": 4.4916,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2931261300377813,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019535489478302197,
+      "loss": 4.491,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2932298180922694,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001953516220834946,
+      "loss": 4.5309,
+      "step": 2828
+    },
+    {
+      "epoch": 0.29333350614675746,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019534834825891886,
+      "loss": 4.4935,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29343719420124553,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019534507330933344,
+      "loss": 4.447,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2935408822557336,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001953417972347769,
+      "loss": 4.4797,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2936445703102217,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019533852003528792,
+      "loss": 4.4689,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29374825836470975,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001953352417109052,
+      "loss": 4.5026,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2938519464191978,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019533196226166738,
+      "loss": 4.4699,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2939556344736859,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019532868168761322,
+      "loss": 4.5083,
+      "step": 2835
+    },
+    {
+      "epoch": 0.29405932252817396,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019532539998878137,
+      "loss": 4.4928,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2941630105826621,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019532211716521053,
+      "loss": 4.492,
+      "step": 2837
+    },
+    {
+      "epoch": 0.29426669863715016,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001953188332169395,
+      "loss": 4.4951,
+      "step": 2838
+    },
+    {
+      "epoch": 0.29437038669163823,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.000195315548144007,
+      "loss": 4.4724,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2944740747461263,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019531226194645176,
+      "loss": 4.4568,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2945777628006144,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019530897462431256,
+      "loss": 4.4589,
+      "step": 2841
+    },
+    {
+      "epoch": 0.29468145085510244,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019530568617762825,
+      "loss": 4.4916,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2947851389095905,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019530239660643756,
+      "loss": 4.4933,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2948888269640786,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019529910591077932,
+      "loss": 4.4446,
+      "step": 2844
+    },
+    {
+      "epoch": 0.29499251501856666,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001952958140906924,
+      "loss": 4.4681,
+      "step": 2845
+    },
+    {
+      "epoch": 0.29509620307305473,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019529252114621558,
+      "loss": 4.4875,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2951998911275428,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019528922707738776,
+      "loss": 4.4859,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2953035791820309,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019528593188424779,
+      "loss": 4.4818,
+      "step": 2848
+    },
+    {
+      "epoch": 0.29540726723651894,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001952826355668345,
+      "loss": 4.4778,
+      "step": 2849
+    },
+    {
+      "epoch": 0.295510955291007,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019527933812518686,
+      "loss": 4.4353,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2956146433454951,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019527603955934376,
+      "loss": 4.5151,
+      "step": 2851
+    },
+    {
+      "epoch": 0.29571833139998316,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001952727398693441,
+      "loss": 4.508,
+      "step": 2852
+    },
+    {
+      "epoch": 0.29582201945447123,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019526943905522678,
+      "loss": 4.5373,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2959257075089593,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019526613711703082,
+      "loss": 4.5151,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2960293955634474,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019526283405479512,
+      "loss": 4.49,
+      "step": 2855
+    },
+    {
+      "epoch": 0.29613308361793544,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019525952986855868,
+      "loss": 4.4924,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2962367716724235,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001952562245583605,
+      "loss": 4.4681,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2963404597269116,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019525291812423954,
+      "loss": 4.4637,
+      "step": 2858
+    },
+    {
+      "epoch": 0.29644414778139966,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019524961056623482,
+      "loss": 4.4826,
+      "step": 2859
+    },
+    {
+      "epoch": 0.29654783583588773,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001952463018843854,
+      "loss": 4.4995,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2966515238903758,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001952429920787303,
+      "loss": 4.4589,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2967552119448639,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019523968114930858,
+      "loss": 4.49,
+      "step": 2862
+    },
+    {
+      "epoch": 0.29685889999935194,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019523636909615926,
+      "loss": 4.4875,
+      "step": 2863
+    },
+    {
+      "epoch": 0.29696258805384,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019523305591932148,
+      "loss": 4.48,
+      "step": 2864
+    },
+    {
+      "epoch": 0.2970662761083281,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019522974161883432,
+      "loss": 4.5249,
+      "step": 2865
+    },
+    {
+      "epoch": 0.29716996416281616,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019522642619473684,
+      "loss": 4.4829,
+      "step": 2866
+    },
+    {
+      "epoch": 0.29727365221730423,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001952231096470682,
+      "loss": 4.4608,
+      "step": 2867
+    },
+    {
+      "epoch": 0.2973773402717923,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001952197919758675,
+      "loss": 4.4793,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2974810283262804,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019521647318117394,
+      "loss": 4.4322,
+      "step": 2869
+    },
+    {
+      "epoch": 0.29758471638076844,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019521315326302664,
+      "loss": 4.4672,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2976884044352565,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019520983222146476,
+      "loss": 4.4932,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2977920924897446,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001952065100565275,
+      "loss": 4.4657,
+      "step": 2872
+    },
+    {
+      "epoch": 0.29789578054423266,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001952031867682541,
+      "loss": 4.4955,
+      "step": 2873
+    },
+    {
+      "epoch": 0.29799946859872073,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001951998623566837,
+      "loss": 4.487,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2981031566532088,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019519653682185552,
+      "loss": 4.4835,
+      "step": 2875
+    },
+    {
+      "epoch": 0.29820684470769687,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019519321016380888,
+      "loss": 4.4324,
+      "step": 2876
+    },
+    {
+      "epoch": 0.29831053276218494,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019518988238258298,
+      "loss": 4.5012,
+      "step": 2877
+    },
+    {
+      "epoch": 0.298414220816673,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019518655347821709,
+      "loss": 4.4762,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2985179088711611,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00019518322345075047,
+      "loss": 4.4632,
+      "step": 2879
+    },
+    {
+      "epoch": 0.29862159692564916,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019517989230022242,
+      "loss": 4.437,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2987252849801373,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019517656002667226,
+      "loss": 4.4664,
+      "step": 2881
+    },
+    {
+      "epoch": 0.29882897303462536,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019517322663013928,
+      "loss": 4.496,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2989326610891134,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019516989211066285,
+      "loss": 4.4569,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2990363491436015,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001951665564682823,
+      "loss": 4.4752,
+      "step": 2884
+    },
+    {
+      "epoch": 0.29914003719808957,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019516321970303695,
+      "loss": 4.5071,
+      "step": 2885
+    },
+    {
+      "epoch": 0.29924372525257764,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019515988181496624,
+      "loss": 4.4836,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2993474133070657,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019515654280410945,
+      "loss": 4.4671,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2994511013615538,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019515320267050613,
+      "loss": 4.4976,
+      "step": 2888
+    },
+    {
+      "epoch": 0.29955478941604186,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019514986141419552,
+      "loss": 4.4499,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2996584774705299,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019514651903521717,
+      "loss": 4.5257,
+      "step": 2890
+    },
+    {
+      "epoch": 0.299762165525018,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019514317553361043,
+      "loss": 4.4424,
+      "step": 2891
+    },
+    {
+      "epoch": 0.29986585357950607,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019513983090941483,
+      "loss": 4.4534,
+      "step": 2892
+    },
+    {
+      "epoch": 0.29996954163399414,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019513648516266975,
+      "loss": 4.4862,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3000732296884822,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019513313829341473,
+      "loss": 4.5139,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3001769177429703,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019512979030168927,
+      "loss": 4.4577,
+      "step": 2895
+    },
+    {
+      "epoch": 0.30028060579745836,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019512644118753277,
+      "loss": 4.4544,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3003842938519464,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001951230909509849,
+      "loss": 4.4591,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3004879819064345,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019511973959208506,
+      "loss": 4.5211,
+      "step": 2898
+    },
+    {
+      "epoch": 0.30059166996092257,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019511638711087282,
+      "loss": 4.4828,
+      "step": 2899
+    },
+    {
+      "epoch": 0.30069535801541064,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001951130335073878,
+      "loss": 4.4544,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3007990460698987,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019510967878166947,
+      "loss": 4.4558,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3009027341243868,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019510632293375752,
+      "loss": 4.4773,
+      "step": 2902
+    },
+    {
+      "epoch": 0.30100642217887486,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00019510296596369147,
+      "loss": 4.4952,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3011101102333629,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019509960787151095,
+      "loss": 4.5134,
+      "step": 2904
+    },
+    {
+      "epoch": 0.301213798287851,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019509624865725558,
+      "loss": 4.4672,
+      "step": 2905
+    },
+    {
+      "epoch": 0.30131748634233907,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019509288832096497,
+      "loss": 4.5035,
+      "step": 2906
+    },
+    {
+      "epoch": 0.30142117439682714,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019508952686267885,
+      "loss": 4.42,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3015248624513152,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00019508616428243677,
+      "loss": 4.465,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3016285505058033,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001950828005802785,
+      "loss": 4.4657,
+      "step": 2909
+    },
+    {
+      "epoch": 0.30173223856029135,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001950794357562437,
+      "loss": 4.484,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3018359266147794,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019507606981037203,
+      "loss": 4.4433,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3019396146692675,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019507270274270324,
+      "loss": 4.4762,
+      "step": 2912
+    },
+    {
+      "epoch": 0.30204330272375557,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019506933455327706,
+      "loss": 4.5061,
+      "step": 2913
+    },
+    {
+      "epoch": 0.30214699077824364,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019506596524213325,
+      "loss": 4.4683,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3022506788327317,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001950625948093115,
+      "loss": 4.4668,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3023543668872198,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019505922325485165,
+      "loss": 4.4326,
+      "step": 2916
+    },
+    {
+      "epoch": 0.30245805494170785,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019505585057879343,
+      "loss": 4.4932,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3025617429961959,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019505247678117663,
+      "loss": 4.4537,
+      "step": 2918
+    },
+    {
+      "epoch": 0.302665431050684,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001950491018620411,
+      "loss": 4.4999,
+      "step": 2919
+    },
+    {
+      "epoch": 0.30276911910517207,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019504572582142667,
+      "loss": 4.4697,
+      "step": 2920
+    },
+    {
+      "epoch": 0.30287280715966014,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001950423486593731,
+      "loss": 4.4829,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3029764952141482,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001950389703759203,
+      "loss": 4.4401,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3030801832686363,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019503559097110807,
+      "loss": 4.4518,
+      "step": 2923
+    },
+    {
+      "epoch": 0.30318387132312435,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019503221044497637,
+      "loss": 4.4695,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3032875593776124,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019502882879756503,
+      "loss": 4.4743,
+      "step": 2925
+    },
+    {
+      "epoch": 0.30339124743210055,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019502544602891395,
+      "loss": 4.5007,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3034949354865886,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019502206213906306,
+      "loss": 4.4697,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3035986235410767,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001950186771280523,
+      "loss": 4.4931,
+      "step": 2928
+    },
+    {
+      "epoch": 0.30370231159556477,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019501529099592155,
+      "loss": 4.4786,
+      "step": 2929
+    },
+    {
+      "epoch": 0.30380599965005284,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001950119037427108,
+      "loss": 4.4402,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3039096877045409,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019500851536846008,
+      "loss": 4.4227,
+      "step": 2931
+    },
+    {
+      "epoch": 0.304013375759029,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019500512587320926,
+      "loss": 4.4456,
+      "step": 2932
+    },
+    {
+      "epoch": 0.30411706381351705,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001950017352569984,
+      "loss": 4.4523,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3042207518680051,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019499834351986746,
+      "loss": 4.453,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3043244399224932,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019499495066185646,
+      "loss": 4.4963,
+      "step": 2935
+    },
+    {
+      "epoch": 0.30442812797698127,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001949915566830055,
+      "loss": 4.4715,
+      "step": 2936
+    },
+    {
+      "epoch": 0.30453181603146934,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019498816158335458,
+      "loss": 4.4423,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3046355040859574,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019498476536294375,
+      "loss": 4.4536,
+      "step": 2938
+    },
+    {
+      "epoch": 0.3047391921404455,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001949813680218131,
+      "loss": 4.4166,
+      "step": 2939
+    },
+    {
+      "epoch": 0.30484288019493355,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001949779695600027,
+      "loss": 4.4335,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3049465682494216,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019497456997755264,
+      "loss": 4.4783,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3050502563039097,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019497116927450305,
+      "loss": 4.474,
+      "step": 2942
+    },
+    {
+      "epoch": 0.30515394435839777,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019496776745089406,
+      "loss": 4.4656,
+      "step": 2943
+    },
+    {
+      "epoch": 0.30525763241288584,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001949643645067658,
+      "loss": 4.4925,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3053613204673739,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019496096044215847,
+      "loss": 4.4422,
+      "step": 2945
+    },
+    {
+      "epoch": 0.305465008521862,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019495755525711212,
+      "loss": 4.4618,
+      "step": 2946
+    },
+    {
+      "epoch": 0.30556869657635005,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.000194954148951667,
+      "loss": 4.4711,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3056723846308381,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001949507415258633,
+      "loss": 4.45,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3057760726853262,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019494733297974125,
+      "loss": 4.428,
+      "step": 2949
+    },
+    {
+      "epoch": 0.30587976073981427,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.000194943923313341,
+      "loss": 4.451,
+      "step": 2950
+    },
+    {
+      "epoch": 0.30598344879430234,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001949405125267028,
+      "loss": 4.4983,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3060871368487904,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019493710061986694,
+      "loss": 4.4957,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3061908249032785,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019493368759287361,
+      "loss": 4.4796,
+      "step": 2953
+    },
+    {
+      "epoch": 0.30629451295776655,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019493027344576316,
+      "loss": 4.4595,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3063982010122546,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001949268581785758,
+      "loss": 4.4666,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3065018890667427,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019492344179135188,
+      "loss": 4.498,
+      "step": 2956
+    },
+    {
+      "epoch": 0.30660557712123077,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019492002428413168,
+      "loss": 4.4927,
+      "step": 2957
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001949166056569555,
+      "loss": 4.4527,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3068129532302069,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001949131859098637,
+      "loss": 4.4533,
+      "step": 2959
+    },
+    {
+      "epoch": 0.306916641284695,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00019490976504289668,
+      "loss": 4.4634,
+      "step": 2960
+    },
+    {
+      "epoch": 0.30702032933918305,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019490634305609471,
+      "loss": 4.5015,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3071240173936711,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019490291994949828,
+      "loss": 4.4517,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3072277054481592,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019489949572314765,
+      "loss": 4.4849,
+      "step": 2963
+    },
+    {
+      "epoch": 0.30733139350264727,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001948960703770833,
+      "loss": 4.4918,
+      "step": 2964
+    },
+    {
+      "epoch": 0.30743508155713534,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019489264391134563,
+      "loss": 4.4788,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3075387696116234,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019488921632597505,
+      "loss": 4.5025,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3076424576661115,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019488578762101203,
+      "loss": 4.467,
+      "step": 2967
+    },
+    {
+      "epoch": 0.30774614572059955,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019488235779649703,
+      "loss": 4.4541,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3078498337750876,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001948789268524705,
+      "loss": 4.4652,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3079535218295757,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001948754947889729,
+      "loss": 4.4786,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3080572098840638,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001948720616060448,
+      "loss": 4.4873,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3081608979385519,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001948686273037266,
+      "loss": 4.4651,
+      "step": 2972
+    },
+    {
+      "epoch": 0.30826458599303996,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019486519188205892,
+      "loss": 4.5188,
+      "step": 2973
+    },
+    {
+      "epoch": 0.30836827404752803,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019486175534108225,
+      "loss": 4.5097,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3084719621020161,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019485831768083713,
+      "loss": 4.4577,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3085756501565042,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019485487890136412,
+      "loss": 4.4674,
+      "step": 2976
+    },
+    {
+      "epoch": 0.30867933821099225,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019485143900270383,
+      "loss": 4.4575,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3087830262654803,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019484799798489676,
+      "loss": 4.5353,
+      "step": 2978
+    },
+    {
+      "epoch": 0.3088867143199684,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00019484455584798361,
+      "loss": 4.4304,
+      "step": 2979
+    },
+    {
+      "epoch": 0.30899040237445646,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.000194841112592005,
+      "loss": 4.4927,
+      "step": 2980
+    },
+    {
+      "epoch": 0.30909409042894453,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019483766821700146,
+      "loss": 4.483,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3091977784834326,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001948342227230137,
+      "loss": 4.4858,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3093014665379207,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019483077611008235,
+      "loss": 4.4839,
+      "step": 2983
+    },
+    {
+      "epoch": 0.30940515459240875,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001948273283782481,
+      "loss": 4.4471,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3095088426468968,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001948238795275516,
+      "loss": 4.4844,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3096125307013849,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019482042955803355,
+      "loss": 4.4448,
+      "step": 2986
+    },
+    {
+      "epoch": 0.30971621875587296,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019481697846973465,
+      "loss": 4.4542,
+      "step": 2987
+    },
+    {
+      "epoch": 0.30981990681036103,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019481352626269565,
+      "loss": 4.4617,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3099235948648491,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019481007293695727,
+      "loss": 4.458,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3100272829193372,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019480661849256023,
+      "loss": 4.4468,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31013097097382525,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001948031629295453,
+      "loss": 4.4601,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3102346590283133,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019479970624795327,
+      "loss": 4.4764,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3103383470828014,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001947962484478249,
+      "loss": 4.4696,
+      "step": 2993
+    },
+    {
+      "epoch": 0.31044203513728946,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019479278952920102,
+      "loss": 4.477,
+      "step": 2994
+    },
+    {
+      "epoch": 0.31054572319177753,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019478932949212245,
+      "loss": 4.4776,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3106494112462656,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019478586833662995,
+      "loss": 4.4574,
+      "step": 2996
+    },
+    {
+      "epoch": 0.3107530993007537,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019478240606276442,
+      "loss": 4.4864,
+      "step": 2997
+    },
+    {
+      "epoch": 0.31085678735524175,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019477894267056666,
+      "loss": 4.368,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3109604754097298,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019477547816007756,
+      "loss": 4.4526,
+      "step": 2999
+    },
+    {
+      "epoch": 0.3110641634642179,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000194772012531338,
+      "loss": 4.4717,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31116785151870596,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001947685457843889,
+      "loss": 4.4854,
+      "step": 3001
+    },
+    {
+      "epoch": 0.31127153957319403,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019476507791927112,
+      "loss": 4.4169,
+      "step": 3002
+    },
+    {
+      "epoch": 0.3113752276276821,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001947616089360256,
+      "loss": 4.4557,
+      "step": 3003
+    },
+    {
+      "epoch": 0.3114789156821702,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019475813883469326,
+      "loss": 4.4646,
+      "step": 3004
+    },
+    {
+      "epoch": 0.31158260373665825,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019475466761531505,
+      "loss": 4.4866,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3116862917911463,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001947511952779319,
+      "loss": 4.4938,
+      "step": 3006
+    },
+    {
+      "epoch": 0.3117899798456344,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001947477218225848,
+      "loss": 4.4514,
+      "step": 3007
+    },
+    {
+      "epoch": 0.31189366790012246,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019474424724931475,
+      "loss": 4.4562,
+      "step": 3008
+    },
+    {
+      "epoch": 0.31199735595461053,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019474077155816276,
+      "loss": 4.4845,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3121010440090986,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019473729474916976,
+      "loss": 4.4958,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3122047320635867,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019473381682237685,
+      "loss": 4.5092,
+      "step": 3011
+    },
+    {
+      "epoch": 0.31230842011807475,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019473033777782503,
+      "loss": 4.4751,
+      "step": 3012
+    },
+    {
+      "epoch": 0.3124121081725628,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019472685761555536,
+      "loss": 4.4743,
+      "step": 3013
+    },
+    {
+      "epoch": 0.3125157962270509,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0001947233763356089,
+      "loss": 4.491,
+      "step": 3014
+    },
+    {
+      "epoch": 0.312619484281539,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019471989393802673,
+      "loss": 4.4783,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3127231723360271,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019471641042284992,
+      "loss": 4.4875,
+      "step": 3016
+    },
+    {
+      "epoch": 0.31282686039051516,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001947129257901196,
+      "loss": 4.4885,
+      "step": 3017
+    },
+    {
+      "epoch": 0.31293054844500323,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019470944003987687,
+      "loss": 4.4643,
+      "step": 3018
+    },
+    {
+      "epoch": 0.3130342364994913,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019470595317216288,
+      "loss": 4.4748,
+      "step": 3019
+    },
+    {
+      "epoch": 0.3131379245539794,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001947024651870187,
+      "loss": 4.4483,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31324161260846745,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001946989760844856,
+      "loss": 4.4574,
+      "step": 3021
+    },
+    {
+      "epoch": 0.3133453006629555,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019469548586460464,
+      "loss": 4.4674,
+      "step": 3022
+    },
+    {
+      "epoch": 0.3134489887174436,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019469199452741705,
+      "loss": 4.5057,
+      "step": 3023
+    },
+    {
+      "epoch": 0.31355267677193166,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019468850207296403,
+      "loss": 4.4566,
+      "step": 3024
+    },
+    {
+      "epoch": 0.31365636482641973,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001946850085012868,
+      "loss": 4.4284,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3137600528809078,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019468151381242649,
+      "loss": 4.4086,
+      "step": 3026
+    },
+    {
+      "epoch": 0.3138637409353959,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019467801800642444,
+      "loss": 4.4613,
+      "step": 3027
+    },
+    {
+      "epoch": 0.31396742898988395,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019467452108332185,
+      "loss": 4.438,
+      "step": 3028
+    },
+    {
+      "epoch": 0.314071117044372,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019467102304316,
+      "loss": 4.4498,
+      "step": 3029
+    },
+    {
+      "epoch": 0.3141748050988601,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019466752388598013,
+      "loss": 4.5043,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31427849315334816,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019466402361182356,
+      "loss": 4.4541,
+      "step": 3031
+    },
+    {
+      "epoch": 0.31438218120783623,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019466052222073157,
+      "loss": 4.4769,
+      "step": 3032
+    },
+    {
+      "epoch": 0.3144858692623243,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019465701971274548,
+      "loss": 4.489,
+      "step": 3033
+    },
+    {
+      "epoch": 0.3145895573168124,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001946535160879066,
+      "loss": 4.4899,
+      "step": 3034
+    },
+    {
+      "epoch": 0.31469324537130045,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001946500113462563,
+      "loss": 4.4684,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3147969334257885,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019464650548783592,
+      "loss": 4.4623,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3149006214802766,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001946429985126868,
+      "loss": 4.4772,
+      "step": 3037
+    },
+    {
+      "epoch": 0.31500430953476466,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019463949042085036,
+      "loss": 4.4399,
+      "step": 3038
+    },
+    {
+      "epoch": 0.31510799758925273,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019463598121236797,
+      "loss": 4.3842,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3152116856437408,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019463247088728102,
+      "loss": 4.462,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3153153736982289,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019462895944563098,
+      "loss": 4.4954,
+      "step": 3041
+    },
+    {
+      "epoch": 0.31541906175271694,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001946254468874592,
+      "loss": 4.4741,
+      "step": 3042
+    },
+    {
+      "epoch": 0.315522749807205,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001946219332128072,
+      "loss": 4.4557,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3156264378616931,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001946184184217164,
+      "loss": 4.4779,
+      "step": 3044
+    },
+    {
+      "epoch": 0.31573012591618116,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019461490251422827,
+      "loss": 4.4799,
+      "step": 3045
+    },
+    {
+      "epoch": 0.31583381397066923,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001946113854903843,
+      "loss": 4.476,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3159375020251573,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000194607867350226,
+      "loss": 4.4456,
+      "step": 3047
+    },
+    {
+      "epoch": 0.3160411900796454,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019460434809379486,
+      "loss": 4.4778,
+      "step": 3048
+    },
+    {
+      "epoch": 0.31614487813413344,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019460082772113245,
+      "loss": 4.4231,
+      "step": 3049
+    },
+    {
+      "epoch": 0.3162485661886215,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019459730623228022,
+      "loss": 4.4713,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3163522542431096,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001945937836272798,
+      "loss": 4.4625,
+      "step": 3051
+    },
+    {
+      "epoch": 0.31645594229759766,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019459025990617272,
+      "loss": 4.4777,
+      "step": 3052
+    },
+    {
+      "epoch": 0.31655963035208573,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019458673506900052,
+      "loss": 4.4678,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3166633184065738,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001945832091158049,
+      "loss": 4.4341,
+      "step": 3054
+    },
+    {
+      "epoch": 0.3167670064610619,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00019457968204662733,
+      "loss": 4.4406,
+      "step": 3055
+    },
+    {
+      "epoch": 0.31687069451554994,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019457615386150954,
+      "loss": 4.4484,
+      "step": 3056
+    },
+    {
+      "epoch": 0.316974382570038,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019457262456049307,
+      "loss": 4.4671,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3170780706245261,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019456909414361962,
+      "loss": 4.4402,
+      "step": 3058
+    },
+    {
+      "epoch": 0.31718175867901416,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001945655626109308,
+      "loss": 4.4295,
+      "step": 3059
+    },
+    {
+      "epoch": 0.3172854467335023,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001945620299624683,
+      "loss": 4.414,
+      "step": 3060
+    },
+    {
+      "epoch": 0.31738913478799036,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019455849619827382,
+      "loss": 4.4506,
+      "step": 3061
+    },
+    {
+      "epoch": 0.31749282284247843,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.000194554961318389,
+      "loss": 4.4642,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3175965108969665,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019455142532285563,
+      "loss": 4.4586,
+      "step": 3063
+    },
+    {
+      "epoch": 0.31770019895145457,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019454788821171538,
+      "loss": 4.4649,
+      "step": 3064
+    },
+    {
+      "epoch": 0.31780388700594264,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00019454434998501,
+      "loss": 4.4466,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3179075750604307,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001945408106427812,
+      "loss": 4.4542,
+      "step": 3066
+    },
+    {
+      "epoch": 0.3180112631149188,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019453727018507077,
+      "loss": 4.5246,
+      "step": 3067
+    },
+    {
+      "epoch": 0.31811495116940686,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001945337286119205,
+      "loss": 4.5043,
+      "step": 3068
+    },
+    {
+      "epoch": 0.31821863922389493,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019453018592337213,
+      "loss": 4.4265,
+      "step": 3069
+    },
+    {
+      "epoch": 0.318322327278383,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019452664211946753,
+      "loss": 4.5128,
+      "step": 3070
+    },
+    {
+      "epoch": 0.31842601533287107,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019452309720024844,
+      "loss": 4.4348,
+      "step": 3071
+    },
+    {
+      "epoch": 0.31852970338735914,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019451955116575674,
+      "loss": 4.4831,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3186333914418472,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019451600401603422,
+      "loss": 4.4866,
+      "step": 3073
+    },
+    {
+      "epoch": 0.3187370794963353,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019451245575112278,
+      "loss": 4.3868,
+      "step": 3074
+    },
+    {
+      "epoch": 0.31884076755082336,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019450890637106428,
+      "loss": 4.4373,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3189444556053114,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019450535587590056,
+      "loss": 4.4538,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3190481436597995,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019450180426567354,
+      "loss": 4.4422,
+      "step": 3077
+    },
+    {
+      "epoch": 0.31915183171428757,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019449825154042513,
+      "loss": 4.4561,
+      "step": 3078
+    },
+    {
+      "epoch": 0.31925551976877564,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019449469770019723,
+      "loss": 4.5059,
+      "step": 3079
+    },
+    {
+      "epoch": 0.3193592078232637,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019449114274503178,
+      "loss": 4.4227,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3194628958777518,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019448758667497075,
+      "loss": 4.4908,
+      "step": 3081
+    },
+    {
+      "epoch": 0.31956658393223986,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019448402949005607,
+      "loss": 4.4663,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3196702719867279,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001944804711903297,
+      "loss": 4.4608,
+      "step": 3083
+    },
+    {
+      "epoch": 0.319773960041216,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019447691177583364,
+      "loss": 4.4182,
+      "step": 3084
+    },
+    {
+      "epoch": 0.31987764809570407,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019447335124660992,
+      "loss": 4.4701,
+      "step": 3085
+    },
+    {
+      "epoch": 0.31998133615019214,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019446978960270048,
+      "loss": 4.4572,
+      "step": 3086
+    },
+    {
+      "epoch": 0.3200850242046802,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019446622684414738,
+      "loss": 4.4828,
+      "step": 3087
+    },
+    {
+      "epoch": 0.3201887122591683,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001944626629709927,
+      "loss": 4.4391,
+      "step": 3088
+    },
+    {
+      "epoch": 0.32029240031365636,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001944590979832784,
+      "loss": 4.4336,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3203960883681444,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019445553188104665,
+      "loss": 4.4397,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3204997764226325,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001944519646643394,
+      "loss": 4.4129,
+      "step": 3091
+    },
+    {
+      "epoch": 0.32060346447712057,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019444839633319885,
+      "loss": 4.4109,
+      "step": 3092
+    },
+    {
+      "epoch": 0.32070715253160864,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019444482688766703,
+      "loss": 4.4595,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3208108405860967,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019444125632778612,
+      "loss": 4.4086,
+      "step": 3094
+    },
+    {
+      "epoch": 0.3209145286405848,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001944376846535982,
+      "loss": 4.4644,
+      "step": 3095
+    },
+    {
+      "epoch": 0.32101821669507286,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019443411186514543,
+      "loss": 4.4432,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3211219047495609,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019443053796246992,
+      "loss": 4.4119,
+      "step": 3097
+    },
+    {
+      "epoch": 0.321225592804049,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019442696294561394,
+      "loss": 4.4644,
+      "step": 3098
+    },
+    {
+      "epoch": 0.32132928085853707,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019442338681461958,
+      "loss": 4.4355,
+      "step": 3099
+    },
+    {
+      "epoch": 0.32143296891302514,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019441980956952905,
+      "loss": 4.4222,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3215366569675132,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019441623121038462,
+      "loss": 4.4843,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3216403450220013,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019441265173722843,
+      "loss": 4.4442,
+      "step": 3102
+    },
+    {
+      "epoch": 0.32174403307648936,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019440907115010275,
+      "loss": 4.503,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3218477211309774,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019440548944904985,
+      "loss": 4.475,
+      "step": 3104
+    },
+    {
+      "epoch": 0.32195140918546555,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019440190663411194,
+      "loss": 4.4357,
+      "step": 3105
+    },
+    {
+      "epoch": 0.3220550972399536,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019439832270533132,
+      "loss": 4.4239,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3221587852944417,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019439473766275027,
+      "loss": 4.4629,
+      "step": 3107
+    },
+    {
+      "epoch": 0.32226247334892977,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001943911515064111,
+      "loss": 4.4815,
+      "step": 3108
+    },
+    {
+      "epoch": 0.32236616140341784,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019438756423635615,
+      "loss": 4.4407,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3224698494579059,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019438397585262767,
+      "loss": 4.467,
+      "step": 3110
+    },
+    {
+      "epoch": 0.322573537512394,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019438038635526806,
+      "loss": 4.4749,
+      "step": 3111
+    },
+    {
+      "epoch": 0.32267722556688205,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019437679574431965,
+      "loss": 4.4781,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3227809136213701,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019437320401982481,
+      "loss": 4.4409,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3228846016758582,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019436961118182592,
+      "loss": 4.4956,
+      "step": 3114
+    },
+    {
+      "epoch": 0.32298828973034627,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001943660172303654,
+      "loss": 4.4362,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32309197778483434,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001943624221654856,
+      "loss": 4.4844,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3231956658393224,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019435882598722897,
+      "loss": 4.4239,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3232993538938105,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019435522869563793,
+      "loss": 4.4157,
+      "step": 3118
+    },
+    {
+      "epoch": 0.32340304194829855,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019435163029075491,
+      "loss": 4.4409,
+      "step": 3119
+    },
+    {
+      "epoch": 0.3235067300027866,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019434803077262244,
+      "loss": 4.4128,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3236104180572747,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019434443014128288,
+      "loss": 4.4665,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32371410611176277,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00019434082839677879,
+      "loss": 4.4656,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32381779416625084,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019433722553915267,
+      "loss": 4.4474,
+      "step": 3123
+    },
+    {
+      "epoch": 0.3239214822207389,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019433362156844698,
+      "loss": 4.4509,
+      "step": 3124
+    },
+    {
+      "epoch": 0.324025170275227,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019433001648470427,
+      "loss": 4.415,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32412885832971505,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001943264102879671,
+      "loss": 4.4765,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3242325463842031,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019432280297827797,
+      "loss": 4.4392,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3243362344386912,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001943191945556795,
+      "loss": 4.4432,
+      "step": 3128
+    },
+    {
+      "epoch": 0.32443992249317927,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001943155850202142,
+      "loss": 4.4573,
+      "step": 3129
+    },
+    {
+      "epoch": 0.32454361054766734,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019431197437192471,
+      "loss": 4.444,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3246472986021554,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019430836261085364,
+      "loss": 4.4296,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3247509866566435,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019430474973704354,
+      "loss": 4.458,
+      "step": 3132
+    },
+    {
+      "epoch": 0.32485467471113155,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019430113575053708,
+      "loss": 4.4591,
+      "step": 3133
+    },
+    {
+      "epoch": 0.3249583627656196,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001942975206513769,
+      "loss": 4.4305,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3250620508201077,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019429390443960568,
+      "loss": 4.4696,
+      "step": 3135
+    },
+    {
+      "epoch": 0.32516573887459577,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019429028711526604,
+      "loss": 4.4581,
+      "step": 3136
+    },
+    {
+      "epoch": 0.32526942692908384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019428666867840066,
+      "loss": 4.4581,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3253731149835719,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001942830491290523,
+      "loss": 4.4528,
+      "step": 3138
+    },
+    {
+      "epoch": 0.32547680303806,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001942794284672636,
+      "loss": 4.4092,
+      "step": 3139
+    },
+    {
+      "epoch": 0.32558049109254805,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001942758066930773,
+      "loss": 4.4827,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3256841791470361,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019427218380653613,
+      "loss": 4.5186,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3257878672015242,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019426855980768287,
+      "loss": 4.4642,
+      "step": 3142
+    },
+    {
+      "epoch": 0.32589155525601227,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001942649346965602,
+      "loss": 4.4665,
+      "step": 3143
+    },
+    {
+      "epoch": 0.32599524331050034,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019426130847321097,
+      "loss": 4.4525,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3260989313649884,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019425768113767795,
+      "loss": 4.4475,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3262026194194765,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001942540526900039,
+      "loss": 4.4347,
+      "step": 3146
+    },
+    {
+      "epoch": 0.32630630747396455,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001942504231302317,
+      "loss": 4.4461,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3264099955284526,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001942467924584041,
+      "loss": 4.4618,
+      "step": 3148
+    },
+    {
+      "epoch": 0.32651368358294075,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019424316067456396,
+      "loss": 4.4542,
+      "step": 3149
+    },
+    {
+      "epoch": 0.3266173716374288,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019423952777875418,
+      "loss": 4.3964,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3267210596919169,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019423589377101758,
+      "loss": 4.4189,
+      "step": 3151
+    },
+    {
+      "epoch": 0.32682474774640496,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019423225865139703,
+      "loss": 4.503,
+      "step": 3152
+    },
+    {
+      "epoch": 0.32692843580089304,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019422862241993545,
+      "loss": 4.4763,
+      "step": 3153
+    },
+    {
+      "epoch": 0.3270321238553811,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019422498507667572,
+      "loss": 4.4683,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3271358119098692,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019422134662166077,
+      "loss": 4.4992,
+      "step": 3155
+    },
+    {
+      "epoch": 0.32723949996435725,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019421770705493354,
+      "loss": 4.4608,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3273431880188453,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019421406637653692,
+      "loss": 4.482,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3274468760733334,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019421042458651395,
+      "loss": 4.4546,
+      "step": 3158
+    },
+    {
+      "epoch": 0.32755056412782146,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019420678168490755,
+      "loss": 4.4352,
+      "step": 3159
+    },
+    {
+      "epoch": 0.32765425218230954,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019420313767176065,
+      "loss": 4.4403,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3277579402367976,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019419949254711636,
+      "loss": 4.4726,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3278616282912857,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001941958463110176,
+      "loss": 4.4651,
+      "step": 3162
+    },
+    {
+      "epoch": 0.32796531634577375,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019419219896350747,
+      "loss": 4.515,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3280690044002618,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001941885505046289,
+      "loss": 4.4578,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3281726924547499,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019418490093442504,
+      "loss": 4.4244,
+      "step": 3165
+    },
+    {
+      "epoch": 0.32827638050923796,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019418125025293887,
+      "loss": 4.4859,
+      "step": 3166
+    },
+    {
+      "epoch": 0.32838006856372604,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019417759846021354,
+      "loss": 4.4438,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3284837566182141,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00019417394555629208,
+      "loss": 4.4728,
+      "step": 3168
+    },
+    {
+      "epoch": 0.3285874446727022,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019417029154121757,
+      "loss": 4.4602,
+      "step": 3169
+    },
+    {
+      "epoch": 0.32869113272719025,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019416663641503323,
+      "loss": 4.423,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3287948207816783,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019416298017778207,
+      "loss": 4.4466,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3288985088361664,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00019415932282950728,
+      "loss": 4.4669,
+      "step": 3172
+    },
+    {
+      "epoch": 0.32900219689065446,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019415566437025206,
+      "loss": 4.4427,
+      "step": 3173
+    },
+    {
+      "epoch": 0.32910588494514253,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019415200480005947,
+      "loss": 4.4346,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3292095729996306,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001941483441189728,
+      "loss": 4.4512,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3293132610541187,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019414468232703517,
+      "loss": 4.4743,
+      "step": 3176
+    },
+    {
+      "epoch": 0.32941694910860675,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019414101942428978,
+      "loss": 4.4539,
+      "step": 3177
+    },
+    {
+      "epoch": 0.3295206371630948,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019413735541077987,
+      "loss": 4.4291,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3296243252175829,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019413369028654868,
+      "loss": 4.4172,
+      "step": 3179
+    },
+    {
+      "epoch": 0.32972801327207096,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019413002405163944,
+      "loss": 4.4088,
+      "step": 3180
+    },
+    {
+      "epoch": 0.32983170132655903,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00019412635670609544,
+      "loss": 4.4669,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3299353893810471,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019412268824995992,
+      "loss": 4.4327,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3300390774355352,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019411901868327617,
+      "loss": 4.4497,
+      "step": 3183
+    },
+    {
+      "epoch": 0.33014276549002325,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001941153480060875,
+      "loss": 4.4268,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3302464535445113,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001941116762184372,
+      "loss": 4.4297,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3303501415989994,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001941080033203686,
+      "loss": 4.4364,
+      "step": 3186
+    },
+    {
+      "epoch": 0.33045382965348746,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019410432931192504,
+      "loss": 4.4927,
+      "step": 3187
+    },
+    {
+      "epoch": 0.33055751770797553,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00019410065419314985,
+      "loss": 4.4977,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3306612057624636,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00019409697796408641,
+      "loss": 4.4403,
+      "step": 3189
+    },
+    {
+      "epoch": 0.3307648938169517,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001940933006247781,
+      "loss": 4.467,
+      "step": 3190
+    },
+    {
+      "epoch": 0.33086858187143975,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00019408962217526833,
+      "loss": 4.4594,
+      "step": 3191
+    },
+    {
+      "epoch": 0.3309722699259278,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019408594261560044,
+      "loss": 4.4412,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3310759579804159,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001940822619458179,
+      "loss": 4.4959,
+      "step": 3193
+    },
+    {
+      "epoch": 0.331179646034904,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001940785801659641,
+      "loss": 4.4427,
+      "step": 3194
+    },
+    {
+      "epoch": 0.3312833340893921,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001940748972760825,
+      "loss": 4.4779,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33138702214388016,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019407121327621657,
+      "loss": 4.4233,
+      "step": 3196
+    },
+    {
+      "epoch": 0.33149071019836823,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019406752816640977,
+      "loss": 4.443,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3315943982528563,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019406384194670554,
+      "loss": 4.4023,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3316980863073444,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019406015461714745,
+      "loss": 4.4119,
+      "step": 3199
+    },
+    {
+      "epoch": 0.33180177436183245,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019405646617777893,
+      "loss": 4.4525,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3319054624163205,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001940527766286435,
+      "loss": 4.4334,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3320091504708086,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00019404908596978477,
+      "loss": 4.4435,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33211283852529666,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019404539420124622,
+      "loss": 4.4833,
+      "step": 3203
+    },
+    {
+      "epoch": 0.33221652657978473,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019404170132307144,
+      "loss": 4.4724,
+      "step": 3204
+    },
+    {
+      "epoch": 0.3323202146342728,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000194038007335304,
+      "loss": 4.4751,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3324239026887609,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019403431223798747,
+      "loss": 4.4767,
+      "step": 3206
+    },
+    {
+      "epoch": 0.33252759074324895,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019403061603116543,
+      "loss": 4.4561,
+      "step": 3207
+    },
+    {
+      "epoch": 0.332631278797737,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019402691871488154,
+      "loss": 4.4188,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3327349668522251,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001940232202891794,
+      "loss": 4.4851,
+      "step": 3209
+    },
+    {
+      "epoch": 0.33283865490671316,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019401952075410263,
+      "loss": 4.4548,
+      "step": 3210
+    },
+    {
+      "epoch": 0.33294234296120123,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019401582010969494,
+      "loss": 4.405,
+      "step": 3211
+    },
+    {
+      "epoch": 0.3330460310156893,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019401211835599989,
+      "loss": 4.439,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3331497190701774,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019400841549306125,
+      "loss": 4.4627,
+      "step": 3213
+    },
+    {
+      "epoch": 0.33325340712466545,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019400471152092267,
+      "loss": 4.4395,
+      "step": 3214
+    },
+    {
+      "epoch": 0.3333570951791535,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019400100643962787,
+      "loss": 4.4621,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3334607832336416,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019399730024922057,
+      "loss": 4.4411,
+      "step": 3216
+    },
+    {
+      "epoch": 0.33356447128812966,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00019399359294974445,
+      "loss": 4.4311,
+      "step": 3217
+    },
+    {
+      "epoch": 0.33366815934261773,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00019398988454124333,
+      "loss": 4.4267,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3337718473971058,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019398617502376092,
+      "loss": 4.4588,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3338755354515939,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019398246439734096,
+      "loss": 4.4264,
+      "step": 3220
+    },
+    {
+      "epoch": 0.33397922350608195,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019397875266202727,
+      "loss": 4.4731,
+      "step": 3221
+    },
+    {
+      "epoch": 0.33408291156057,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019397503981786365,
+      "loss": 4.4846,
+      "step": 3222
+    },
+    {
+      "epoch": 0.3341865996150581,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001939713258648939,
+      "loss": 4.3899,
+      "step": 3223
+    },
+    {
+      "epoch": 0.33429028766954616,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019396761080316184,
+      "loss": 4.4881,
+      "step": 3224
+    },
+    {
+      "epoch": 0.33439397572403423,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019396389463271127,
+      "loss": 4.4176,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3344976637785223,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001939601773535861,
+      "loss": 4.409,
+      "step": 3226
+    },
+    {
+      "epoch": 0.3346013518330104,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019395645896583017,
+      "loss": 4.4441,
+      "step": 3227
+    },
+    {
+      "epoch": 0.33470503988749845,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001939527394694873,
+      "loss": 4.4747,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3348087279419865,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019394901886460143,
+      "loss": 4.4516,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3349124159964746,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019394529715121645,
+      "loss": 4.4232,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33501610405096266,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019394157432937629,
+      "loss": 4.4171,
+      "step": 3231
+    },
+    {
+      "epoch": 0.33511979210545073,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019393785039912483,
+      "loss": 4.4399,
+      "step": 3232
+    },
+    {
+      "epoch": 0.3352234801599388,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019393412536050604,
+      "loss": 4.4708,
+      "step": 3233
+    },
+    {
+      "epoch": 0.3353271682144269,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019393039921356385,
+      "loss": 4.4745,
+      "step": 3234
+    },
+    {
+      "epoch": 0.33543085626891495,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001939266719583422,
+      "loss": 4.4292,
+      "step": 3235
+    },
+    {
+      "epoch": 0.335534544323403,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001939229435948852,
+      "loss": 4.4479,
+      "step": 3236
+    },
+    {
+      "epoch": 0.3356382323778911,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001939192141232367,
+      "loss": 4.4192,
+      "step": 3237
+    },
+    {
+      "epoch": 0.33574192043237916,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019391548354344074,
+      "loss": 4.4338,
+      "step": 3238
+    },
+    {
+      "epoch": 0.3358456084868673,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019391175185554135,
+      "loss": 4.4788,
+      "step": 3239
+    },
+    {
+      "epoch": 0.33594929654135536,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019390801905958257,
+      "loss": 4.4474,
+      "step": 3240
+    },
+    {
+      "epoch": 0.33605298459584343,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019390428515560841,
+      "loss": 4.4906,
+      "step": 3241
+    },
+    {
+      "epoch": 0.3361566726503315,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019390055014366296,
+      "loss": 4.4659,
+      "step": 3242
+    },
+    {
+      "epoch": 0.33626036070481957,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019389681402379028,
+      "loss": 4.4293,
+      "step": 3243
+    },
+    {
+      "epoch": 0.33636404875930764,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019389307679603442,
+      "loss": 4.4542,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3364677368137957,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019388933846043954,
+      "loss": 4.4636,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3365714248682838,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001938855990170497,
+      "loss": 4.4272,
+      "step": 3246
+    },
+    {
+      "epoch": 0.33667511292277186,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019388185846590903,
+      "loss": 4.4256,
+      "step": 3247
+    },
+    {
+      "epoch": 0.33677880097725993,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019387811680706167,
+      "loss": 4.4195,
+      "step": 3248
+    },
+    {
+      "epoch": 0.336882489031748,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019387437404055175,
+      "loss": 4.4388,
+      "step": 3249
+    },
+    {
+      "epoch": 0.33698617708623607,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019387063016642345,
+      "loss": 4.4516,
+      "step": 3250
+    },
+    {
+      "epoch": 0.33708986514072414,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00019386688518472096,
+      "loss": 4.4562,
+      "step": 3251
+    },
+    {
+      "epoch": 0.3371935531952122,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019386313909548842,
+      "loss": 4.4315,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3372972412497003,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019385939189877008,
+      "loss": 4.4577,
+      "step": 3253
+    },
+    {
+      "epoch": 0.33740092930418836,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00019385564359461013,
+      "loss": 4.4885,
+      "step": 3254
+    },
+    {
+      "epoch": 0.33750461735867643,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019385189418305275,
+      "loss": 4.4272,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3376083054131645,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00019384814366414231,
+      "loss": 4.4444,
+      "step": 3256
+    },
+    {
+      "epoch": 0.33771199346765257,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019384439203792291,
+      "loss": 4.4779,
+      "step": 3257
+    },
+    {
+      "epoch": 0.33781568152214064,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019384063930443887,
+      "loss": 4.4524,
+      "step": 3258
+    },
+    {
+      "epoch": 0.3379193695766287,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019383688546373453,
+      "loss": 4.4602,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3380230576311168,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001938331305158541,
+      "loss": 4.4119,
+      "step": 3260
+    },
+    {
+      "epoch": 0.33812674568560486,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019382937446084194,
+      "loss": 4.4518,
+      "step": 3261
+    },
+    {
+      "epoch": 0.33823043374009293,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019382561729874232,
+      "loss": 4.4152,
+      "step": 3262
+    },
+    {
+      "epoch": 0.338334121794581,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019382185902959962,
+      "loss": 4.4262,
+      "step": 3263
+    },
+    {
+      "epoch": 0.33843780984906907,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019381809965345813,
+      "loss": 4.4535,
+      "step": 3264
+    },
+    {
+      "epoch": 0.33854149790355714,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019381433917036223,
+      "loss": 4.4597,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3386451859580452,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001938105775803563,
+      "loss": 4.4852,
+      "step": 3266
+    },
+    {
+      "epoch": 0.3387488740125333,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019380681488348473,
+      "loss": 4.4376,
+      "step": 3267
+    },
+    {
+      "epoch": 0.33885256206702136,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019380305107979191,
+      "loss": 4.4622,
+      "step": 3268
+    },
+    {
+      "epoch": 0.33895625012150943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001937992861693222,
+      "loss": 4.4553,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3390599381759975,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001937955201521201,
+      "loss": 4.4485,
+      "step": 3270
+    },
+    {
+      "epoch": 0.33916362623048557,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019379175302823,
+      "loss": 4.4232,
+      "step": 3271
+    },
+    {
+      "epoch": 0.33926731428497364,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019378798479769636,
+      "loss": 4.4852,
+      "step": 3272
+    },
+    {
+      "epoch": 0.3393710023394617,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019378421546056363,
+      "loss": 4.4148,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3394746903939498,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001937804450168763,
+      "loss": 4.4718,
+      "step": 3274
+    },
+    {
+      "epoch": 0.33957837844843786,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019377667346667885,
+      "loss": 4.4646,
+      "step": 3275
+    },
+    {
+      "epoch": 0.33968206650292593,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019377290081001576,
+      "loss": 4.4659,
+      "step": 3276
+    },
+    {
+      "epoch": 0.339785754557414,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001937691270469316,
+      "loss": 4.4167,
+      "step": 3277
+    },
+    {
+      "epoch": 0.33988944261190207,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001937653521774708,
+      "loss": 4.4329,
+      "step": 3278
+    },
+    {
+      "epoch": 0.33999313066639014,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.000193761576201678,
+      "loss": 4.4596,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3400968187208782,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001937577991195977,
+      "loss": 4.4565,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3402005067753663,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001937540209312745,
+      "loss": 4.4684,
+      "step": 3281
+    },
+    {
+      "epoch": 0.34030419482985436,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019375024163675292,
+      "loss": 4.4353,
+      "step": 3282
+    },
+    {
+      "epoch": 0.3404078828843425,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019374646123607764,
+      "loss": 4.444,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34051157093883055,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00019374267972929317,
+      "loss": 4.4671,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3406152589933186,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019373889711644417,
+      "loss": 4.4257,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3407189470478067,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001937351133975753,
+      "loss": 4.4502,
+      "step": 3286
+    },
+    {
+      "epoch": 0.34082263510229477,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019373132857273114,
+      "loss": 4.4666,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34092632315678284,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001937275426419564,
+      "loss": 4.4717,
+      "step": 3288
+    },
+    {
+      "epoch": 0.3410300112112709,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019372375560529567,
+      "loss": 4.4035,
+      "step": 3289
+    },
+    {
+      "epoch": 0.341133699265759,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019371996746279376,
+      "loss": 4.4463,
+      "step": 3290
+    },
+    {
+      "epoch": 0.34123738732024705,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001937161782144953,
+      "loss": 4.4601,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3413410753747351,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019371238786044498,
+      "loss": 4.4237,
+      "step": 3292
+    },
+    {
+      "epoch": 0.3414447634292232,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019370859640068755,
+      "loss": 4.4393,
+      "step": 3293
+    },
+    {
+      "epoch": 0.34154845148371127,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00019370480383526774,
+      "loss": 4.461,
+      "step": 3294
+    },
+    {
+      "epoch": 0.34165213953819934,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019370101016423028,
+      "loss": 4.4328,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3417558275926874,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00019369721538761996,
+      "loss": 4.4228,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3418595156471755,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019369341950548153,
+      "loss": 4.4922,
+      "step": 3297
+    },
+    {
+      "epoch": 0.34196320370166355,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001936896225178598,
+      "loss": 4.443,
+      "step": 3298
+    },
+    {
+      "epoch": 0.3420668917561516,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019368582442479953,
+      "loss": 4.4383,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3421705798106397,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001936820252263456,
+      "loss": 4.4737,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34227426786512777,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001936782249225428,
+      "loss": 4.4784,
+      "step": 3301
+    },
+    {
+      "epoch": 0.34237795591961584,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00019367442351343593,
+      "loss": 4.4926,
+      "step": 3302
+    },
+    {
+      "epoch": 0.3424816439741039,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001936706209990699,
+      "loss": 4.4373,
+      "step": 3303
+    },
+    {
+      "epoch": 0.342585332028592,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019366681737948956,
+      "loss": 4.4738,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34268902008308005,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019366301265473978,
+      "loss": 4.4515,
+      "step": 3305
+    },
+    {
+      "epoch": 0.3427927081375681,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00019365920682486547,
+      "loss": 4.4219,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3428963961920562,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001936553998899115,
+      "loss": 4.4172,
+      "step": 3307
+    },
+    {
+      "epoch": 0.34300008424654427,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019365159184992284,
+      "loss": 4.4943,
+      "step": 3308
+    },
+    {
+      "epoch": 0.34310377230103234,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019364778270494437,
+      "loss": 4.4547,
+      "step": 3309
+    },
+    {
+      "epoch": 0.3432074603555204,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00019364397245502107,
+      "loss": 4.4844,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3433111484100085,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019364016110019785,
+      "loss": 4.4454,
+      "step": 3311
+    },
+    {
+      "epoch": 0.34341483646449655,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001936363486405197,
+      "loss": 4.4257,
+      "step": 3312
+    },
+    {
+      "epoch": 0.3435185245189846,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019363253507603164,
+      "loss": 4.4496,
+      "step": 3313
+    },
+    {
+      "epoch": 0.3436222125734727,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019362872040677866,
+      "loss": 4.4632,
+      "step": 3314
+    },
+    {
+      "epoch": 0.34372590062796077,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019362490463280572,
+      "loss": 4.4439,
+      "step": 3315
+    },
+    {
+      "epoch": 0.34382958868244884,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001936210877541579,
+      "loss": 4.5136,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3439332767369369,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019361726977088018,
+      "loss": 4.435,
+      "step": 3317
+    },
+    {
+      "epoch": 0.344036964791425,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019361345068301763,
+      "loss": 4.4213,
+      "step": 3318
+    },
+    {
+      "epoch": 0.34414065284591305,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019360963049061533,
+      "loss": 4.4375,
+      "step": 3319
+    },
+    {
+      "epoch": 0.3442443409004011,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019360580919371834,
+      "loss": 4.449,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3443480289548892,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019360198679237172,
+      "loss": 4.4631,
+      "step": 3321
+    },
+    {
+      "epoch": 0.34445171700937727,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019359816328662065,
+      "loss": 4.4259,
+      "step": 3322
+    },
+    {
+      "epoch": 0.34455540506386534,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019359433867651018,
+      "loss": 4.4103,
+      "step": 3323
+    },
+    {
+      "epoch": 0.3446590931183534,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001935905129620854,
+      "loss": 4.4288,
+      "step": 3324
+    },
+    {
+      "epoch": 0.3447627811728415,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019358668614339152,
+      "loss": 4.47,
+      "step": 3325
+    },
+    {
+      "epoch": 0.34486646922732955,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001935828582204737,
+      "loss": 4.4248,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3449701572818176,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019357902919337706,
+      "loss": 4.3995,
+      "step": 3327
+    },
+    {
+      "epoch": 0.34507384533630575,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019357519906214676,
+      "loss": 4.4162,
+      "step": 3328
+    },
+    {
+      "epoch": 0.3451775333907938,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019357136782682804,
+      "loss": 4.4012,
+      "step": 3329
+    },
+    {
+      "epoch": 0.3452812214452819,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019356753548746612,
+      "loss": 4.4357,
+      "step": 3330
+    },
+    {
+      "epoch": 0.34538490949976997,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019356370204410615,
+      "loss": 4.441,
+      "step": 3331
+    },
+    {
+      "epoch": 0.34548859755425804,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019355986749679342,
+      "loss": 4.4196,
+      "step": 3332
+    },
+    {
+      "epoch": 0.3455922856087461,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019355603184557314,
+      "loss": 4.4693,
+      "step": 3333
+    },
+    {
+      "epoch": 0.3456959736632342,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019355219509049058,
+      "loss": 4.481,
+      "step": 3334
+    },
+    {
+      "epoch": 0.34579966171772225,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.000193548357231591,
+      "loss": 4.4777,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3459033497722103,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019354451826891967,
+      "loss": 4.4111,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3460070378266984,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019354067820252194,
+      "loss": 4.4399,
+      "step": 3337
+    },
+    {
+      "epoch": 0.34611072588118647,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019353683703244307,
+      "loss": 4.4452,
+      "step": 3338
+    },
+    {
+      "epoch": 0.34621441393567454,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001935329947587284,
+      "loss": 4.4372,
+      "step": 3339
+    },
+    {
+      "epoch": 0.3463181019901626,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019352915138142325,
+      "loss": 4.4171,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3464217900446507,
+      "grad_norm": 1.5,
+      "learning_rate": 0.000193525306900573,
+      "loss": 4.4561,
+      "step": 3341
+    },
+    {
+      "epoch": 0.34652547809913875,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019352146131622298,
+      "loss": 4.4123,
+      "step": 3342
+    },
+    {
+      "epoch": 0.3466291661536268,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019351761462841857,
+      "loss": 4.4771,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3467328542081149,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019351376683720515,
+      "loss": 4.4315,
+      "step": 3344
+    },
+    {
+      "epoch": 0.34683654226260296,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019350991794262813,
+      "loss": 4.4284,
+      "step": 3345
+    },
+    {
+      "epoch": 0.34694023031709104,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019350606794473293,
+      "loss": 4.4584,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3470439183715791,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.000193502216843565,
+      "loss": 4.3793,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3471476064260672,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001934983646391697,
+      "loss": 4.4539,
+      "step": 3348
+    },
+    {
+      "epoch": 0.34725129448055525,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019349451133159255,
+      "loss": 4.458,
+      "step": 3349
+    },
+    {
+      "epoch": 0.3473549825350433,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.000193490656920879,
+      "loss": 4.4281,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3474586705895314,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001934868014070745,
+      "loss": 4.4403,
+      "step": 3351
+    },
+    {
+      "epoch": 0.34756235864401946,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019348294479022457,
+      "loss": 4.4345,
+      "step": 3352
+    },
+    {
+      "epoch": 0.34766604669850754,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001934790870703747,
+      "loss": 4.4486,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3477697347529956,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019347522824757042,
+      "loss": 4.4779,
+      "step": 3354
+    },
+    {
+      "epoch": 0.3478734228074837,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019347136832185727,
+      "loss": 4.4556,
+      "step": 3355
+    },
+    {
+      "epoch": 0.34797711086197175,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019346750729328077,
+      "loss": 4.4484,
+      "step": 3356
+    },
+    {
+      "epoch": 0.3480807989164598,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019346364516188648,
+      "loss": 4.4024,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3481844869709479,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019345978192772,
+      "loss": 4.4108,
+      "step": 3358
+    },
+    {
+      "epoch": 0.34828817502543596,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00019345591759082684,
+      "loss": 4.4214,
+      "step": 3359
+    },
+    {
+      "epoch": 0.34839186307992404,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00019345205215125265,
+      "loss": 4.4519,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3484955511344121,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019344818560904306,
+      "loss": 4.4294,
+      "step": 3361
+    },
+    {
+      "epoch": 0.3485992391889002,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019344431796424364,
+      "loss": 4.4023,
+      "step": 3362
+    },
+    {
+      "epoch": 0.34870292724338825,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001934404492169,
+      "loss": 4.4062,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3488066152978763,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001934365793670579,
+      "loss": 4.4425,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3489103033523644,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001934327084147629,
+      "loss": 4.4259,
+      "step": 3365
+    },
+    {
+      "epoch": 0.34901399140685246,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001934288363600607,
+      "loss": 4.4137,
+      "step": 3366
+    },
+    {
+      "epoch": 0.34911767946134054,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.000193424963202997,
+      "loss": 4.4226,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3492213675158286,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001934210889436175,
+      "loss": 4.4478,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3493250555703167,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019341721358196785,
+      "loss": 4.4312,
+      "step": 3369
+    },
+    {
+      "epoch": 0.34942874362480475,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019341333711809386,
+      "loss": 4.4476,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3495324316792928,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019340945955204121,
+      "loss": 4.4335,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3496361197337809,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001934055808838557,
+      "loss": 4.4462,
+      "step": 3372
+    },
+    {
+      "epoch": 0.349739807788269,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001934017011135831,
+      "loss": 4.4577,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3498434958427571,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019339782024126908,
+      "loss": 4.3599,
+      "step": 3374
+    },
+    {
+      "epoch": 0.34994718389724516,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019339393826695958,
+      "loss": 4.4257,
+      "step": 3375
+    },
+    {
+      "epoch": 0.35005087195173323,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019339005519070028,
+      "loss": 4.443,
+      "step": 3376
+    },
+    {
+      "epoch": 0.3501545600062213,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001933861710125371,
+      "loss": 4.4858,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3502582480607094,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019338228573251575,
+      "loss": 4.4624,
+      "step": 3378
+    },
+    {
+      "epoch": 0.35036193611519745,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019337839935068218,
+      "loss": 4.4479,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3504656241696855,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019337451186708218,
+      "loss": 4.4735,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3505693122241736,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019337062328176165,
+      "loss": 4.4579,
+      "step": 3381
+    },
+    {
+      "epoch": 0.35067300027866166,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019336673359476647,
+      "loss": 4.4533,
+      "step": 3382
+    },
+    {
+      "epoch": 0.35077668833314973,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001933628428061425,
+      "loss": 4.4427,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3508803763876378,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019335895091593573,
+      "loss": 4.4462,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3509840644421259,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019335505792419198,
+      "loss": 4.4379,
+      "step": 3385
+    },
+    {
+      "epoch": 0.35108775249661395,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019335116383095724,
+      "loss": 4.4643,
+      "step": 3386
+    },
+    {
+      "epoch": 0.351191440551102,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019334726863627744,
+      "loss": 4.4121,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3512951286055901,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019334337234019856,
+      "loss": 4.4721,
+      "step": 3388
+    },
+    {
+      "epoch": 0.35139881666007816,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001933394749427665,
+      "loss": 4.4425,
+      "step": 3389
+    },
+    {
+      "epoch": 0.35150250471456623,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019333557644402735,
+      "loss": 4.4795,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3516061927690543,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019333167684402704,
+      "loss": 4.4531,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3517098808235424,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00019332777614281162,
+      "loss": 4.4808,
+      "step": 3392
+    },
+    {
+      "epoch": 0.35181356887803045,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019332387434042706,
+      "loss": 4.4205,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3519172569325185,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019331997143691947,
+      "loss": 4.429,
+      "step": 3394
+    },
+    {
+      "epoch": 0.3520209449870066,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019331606743233483,
+      "loss": 4.4505,
+      "step": 3395
+    },
+    {
+      "epoch": 0.35212463304149466,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019331216232671924,
+      "loss": 4.4677,
+      "step": 3396
+    },
+    {
+      "epoch": 0.35222832109598273,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019330825612011877,
+      "loss": 4.428,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3523320091504708,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001933043488125795,
+      "loss": 4.4564,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3524356972049589,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019330044040414754,
+      "loss": 4.4115,
+      "step": 3399
+    },
+    {
+      "epoch": 0.35253938525944695,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019329653089486903,
+      "loss": 4.4736,
+      "step": 3400
+    },
+    {
+      "epoch": 0.352643073313935,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019329262028479005,
+      "loss": 4.4048,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3527467613684231,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001932887085739568,
+      "loss": 4.4026,
+      "step": 3402
+    },
+    {
+      "epoch": 0.35285044942291116,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019328479576241535,
+      "loss": 4.4926,
+      "step": 3403
+    },
+    {
+      "epoch": 0.35295413747739923,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019328088185021196,
+      "loss": 4.4422,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3530578255318873,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019327696683739274,
+      "loss": 4.4698,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3531615135863754,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019327305072400393,
+      "loss": 4.4035,
+      "step": 3406
+    },
+    {
+      "epoch": 0.35326520164086345,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019326913351009172,
+      "loss": 4.4461,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3533688896953515,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001932652151957023,
+      "loss": 4.4329,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3534725777498396,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019326129578088193,
+      "loss": 4.4534,
+      "step": 3409
+    },
+    {
+      "epoch": 0.35357626580432766,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019325737526567683,
+      "loss": 4.457,
+      "step": 3410
+    },
+    {
+      "epoch": 0.35367995385881573,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019325345365013333,
+      "loss": 4.4506,
+      "step": 3411
+    },
+    {
+      "epoch": 0.3537836419133038,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001932495309342976,
+      "loss": 4.4096,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3538873299677919,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.000193245607118216,
+      "loss": 4.4292,
+      "step": 3413
+    },
+    {
+      "epoch": 0.35399101802227995,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001932416822019348,
+      "loss": 4.4447,
+      "step": 3414
+    },
+    {
+      "epoch": 0.354094706076768,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019323775618550033,
+      "loss": 4.4365,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3541983941312561,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019323382906895887,
+      "loss": 4.4433,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3543020821857442,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019322990085235677,
+      "loss": 4.3989,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3544057702402323,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019322597153574041,
+      "loss": 4.4656,
+      "step": 3418
+    },
+    {
+      "epoch": 0.35450945829472036,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019322204111915612,
+      "loss": 4.4606,
+      "step": 3419
+    },
+    {
+      "epoch": 0.35461314634920843,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019321810960265027,
+      "loss": 4.3985,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3547168344036965,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001932141769862693,
+      "loss": 4.4343,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3548205224581846,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00019321024327005953,
+      "loss": 4.4955,
+      "step": 3422
+    },
+    {
+      "epoch": 0.35492421051267264,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019320630845406743,
+      "loss": 4.4026,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3550278985671607,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019320237253833944,
+      "loss": 4.4636,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3551315866216488,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019319843552292193,
+      "loss": 4.4602,
+      "step": 3425
+    },
+    {
+      "epoch": 0.35523527467613686,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019319449740786145,
+      "loss": 4.4292,
+      "step": 3426
+    },
+    {
+      "epoch": 0.35533896273062493,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019319055819320437,
+      "loss": 4.4169,
+      "step": 3427
+    },
+    {
+      "epoch": 0.355442650785113,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019318661787899724,
+      "loss": 4.4037,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3555463388396011,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001931826764652865,
+      "loss": 4.4441,
+      "step": 3429
+    },
+    {
+      "epoch": 0.35565002689408914,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019317873395211868,
+      "loss": 4.406,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3557537149485772,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001931747903395403,
+      "loss": 4.4601,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3558574030030653,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019317084562759786,
+      "loss": 4.4696,
+      "step": 3432
+    },
+    {
+      "epoch": 0.35596109105755336,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019316689981633796,
+      "loss": 4.4326,
+      "step": 3433
+    },
+    {
+      "epoch": 0.35606477911204143,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019316295290580708,
+      "loss": 4.443,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3561684671665295,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019315900489605186,
+      "loss": 4.4226,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3562721552210176,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019315505578711888,
+      "loss": 4.4494,
+      "step": 3436
+    },
+    {
+      "epoch": 0.35637584327550564,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001931511055790547,
+      "loss": 4.4322,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3564795313299937,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019314715427190592,
+      "loss": 4.3908,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3565832193844818,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019314320186571919,
+      "loss": 4.4296,
+      "step": 3439
+    },
+    {
+      "epoch": 0.35668690743896986,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001931392483605411,
+      "loss": 4.4108,
+      "step": 3440
+    },
+    {
+      "epoch": 0.35679059549345793,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001931352937564184,
+      "loss": 4.4312,
+      "step": 3441
+    },
+    {
+      "epoch": 0.356894283547946,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019313133805339762,
+      "loss": 4.4593,
+      "step": 3442
+    },
+    {
+      "epoch": 0.35699797160243407,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019312738125152554,
+      "loss": 4.3779,
+      "step": 3443
+    },
+    {
+      "epoch": 0.35710165965692214,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019312342335084878,
+      "loss": 4.3964,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3572053477114102,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019311946435141407,
+      "loss": 4.4928,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3573090357658983,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001931155042532681,
+      "loss": 4.4779,
+      "step": 3446
+    },
+    {
+      "epoch": 0.35741272382038636,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019311154305645762,
+      "loss": 4.4363,
+      "step": 3447
+    },
+    {
+      "epoch": 0.35751641187487443,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019310758076102933,
+      "loss": 4.4306,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3576200999293625,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019310361736703003,
+      "loss": 4.4224,
+      "step": 3449
+    },
+    {
+      "epoch": 0.35772378798385057,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019309965287450645,
+      "loss": 4.4059,
+      "step": 3450
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019309568728350537,
+      "loss": 4.506,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3579311640928267,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001930917205940736,
+      "loss": 4.4283,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3580348521473148,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019308775280625794,
+      "loss": 4.4682,
+      "step": 3453
+    },
+    {
+      "epoch": 0.35813854020180286,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001930837839201052,
+      "loss": 4.4591,
+      "step": 3454
+    },
+    {
+      "epoch": 0.35824222825629093,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019307981393566217,
+      "loss": 4.4646,
+      "step": 3455
+    },
+    {
+      "epoch": 0.358345916310779,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019307584285297572,
+      "loss": 4.3935,
+      "step": 3456
+    },
+    {
+      "epoch": 0.35844960436526707,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019307187067209273,
+      "loss": 4.469,
+      "step": 3457
+    },
+    {
+      "epoch": 0.35855329241975514,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019306789739306006,
+      "loss": 4.4591,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3586569804742432,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019306392301592456,
+      "loss": 4.4508,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3587606685287313,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019305994754073314,
+      "loss": 4.3816,
+      "step": 3460
+    },
+    {
+      "epoch": 0.35886435658321936,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019305597096753272,
+      "loss": 4.4473,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3589680446377075,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019305199329637017,
+      "loss": 4.4264,
+      "step": 3462
+    },
+    {
+      "epoch": 0.35907173269219556,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001930480145272925,
+      "loss": 4.4503,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3591754207466836,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001930440346603466,
+      "loss": 4.4796,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3592791088011717,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019304005369557946,
+      "loss": 4.4133,
+      "step": 3465
+    },
+    {
+      "epoch": 0.35938279685565977,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019303607163303804,
+      "loss": 4.4165,
+      "step": 3466
+    },
+    {
+      "epoch": 0.35948648491014784,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019303208847276928,
+      "loss": 4.4471,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3595901729646359,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019302810421482021,
+      "loss": 4.3831,
+      "step": 3468
+    },
+    {
+      "epoch": 0.359693861019124,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019302411885923788,
+      "loss": 4.4317,
+      "step": 3469
+    },
+    {
+      "epoch": 0.35979754907361206,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019302013240606928,
+      "loss": 4.4411,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3599012371281001,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019301614485536144,
+      "loss": 4.4209,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3600049251825882,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001930121562071614,
+      "loss": 4.4295,
+      "step": 3472
+    },
+    {
+      "epoch": 0.36010861323707627,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019300816646151622,
+      "loss": 4.4055,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36021230129156434,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.000193004175618473,
+      "loss": 4.4217,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3603159893460524,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019300018367807885,
+      "loss": 4.4355,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3604196774005405,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001929961906403808,
+      "loss": 4.425,
+      "step": 3476
+    },
+    {
+      "epoch": 0.36052336545502855,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019299219650542604,
+      "loss": 4.4662,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3606270535095166,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001929882012732616,
+      "loss": 4.4452,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3607307415640047,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019298420494393473,
+      "loss": 4.4603,
+      "step": 3479
+    },
+    {
+      "epoch": 0.36083442961849277,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019298020751749251,
+      "loss": 4.39,
+      "step": 3480
+    },
+    {
+      "epoch": 0.36093811767298084,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019297620899398212,
+      "loss": 4.4747,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3610418057274689,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019297220937345078,
+      "loss": 4.4274,
+      "step": 3482
+    },
+    {
+      "epoch": 0.361145493781957,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019296820865594562,
+      "loss": 4.4059,
+      "step": 3483
+    },
+    {
+      "epoch": 0.36124918183644505,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019296420684151384,
+      "loss": 4.3839,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3613528698909331,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019296020393020272,
+      "loss": 4.4561,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3614565579454212,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019295619992205944,
+      "loss": 4.45,
+      "step": 3486
+    },
+    {
+      "epoch": 0.36156024599990927,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001929521948171313,
+      "loss": 4.4206,
+      "step": 3487
+    },
+    {
+      "epoch": 0.36166393405439734,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019294818861546547,
+      "loss": 4.4488,
+      "step": 3488
+    },
+    {
+      "epoch": 0.3617676221088854,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019294418131710926,
+      "loss": 4.4317,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3618713101633735,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019294017292211,
+      "loss": 4.4465,
+      "step": 3490
+    },
+    {
+      "epoch": 0.36197499821786155,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019293616343051491,
+      "loss": 4.4293,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3620786862723496,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019293215284237134,
+      "loss": 4.4282,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3621823743268377,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001929281411577266,
+      "loss": 4.4317,
+      "step": 3493
+    },
+    {
+      "epoch": 0.36228606238132577,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.000192924128376628,
+      "loss": 4.4273,
+      "step": 3494
+    },
+    {
+      "epoch": 0.36238975043581384,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019292011449912295,
+      "loss": 4.422,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3624934384903019,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019291609952525876,
+      "loss": 4.441,
+      "step": 3496
+    },
+    {
+      "epoch": 0.36259712654479,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019291208345508277,
+      "loss": 4.4409,
+      "step": 3497
+    },
+    {
+      "epoch": 0.36270081459927805,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019290806628864245,
+      "loss": 4.4563,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3628045026537661,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019290404802598516,
+      "loss": 4.4164,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3629081907082542,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019290002866715827,
+      "loss": 4.4176,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36301187876274227,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019289600821220928,
+      "loss": 4.455,
+      "step": 3501
+    },
+    {
+      "epoch": 0.36311556681723034,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019289198666118558,
+      "loss": 4.4429,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3632192548717184,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019288796401413462,
+      "loss": 4.4002,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3633229429262065,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019288394027110387,
+      "loss": 4.413,
+      "step": 3504
+    },
+    {
+      "epoch": 0.36342663098069455,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019287991543214084,
+      "loss": 4.4612,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3635303190351826,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019287588949729295,
+      "loss": 4.4533,
+      "step": 3506
+    },
+    {
+      "epoch": 0.36363400708967075,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019287186246660774,
+      "loss": 4.4742,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3637376951441588,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001928678343401327,
+      "loss": 4.4463,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3638413831986469,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001928638051179154,
+      "loss": 4.41,
+      "step": 3509
+    },
+    {
+      "epoch": 0.36394507125313497,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001928597748000034,
+      "loss": 4.4391,
+      "step": 3510
+    },
+    {
+      "epoch": 0.36404875930762304,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001928557433864442,
+      "loss": 4.4267,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3641524473621111,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019285171087728536,
+      "loss": 4.4649,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3642561354165992,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019284767727257445,
+      "loss": 4.4497,
+      "step": 3513
+    },
+    {
+      "epoch": 0.36435982347108725,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019284364257235917,
+      "loss": 4.479,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3644635115255753,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.000192839606776687,
+      "loss": 4.4417,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3645671995800634,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019283556988560562,
+      "loss": 4.3996,
+      "step": 3516
+    },
+    {
+      "epoch": 0.36467088763455147,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001928315318991626,
+      "loss": 4.4304,
+      "step": 3517
+    },
+    {
+      "epoch": 0.36477457568903954,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001928274928174057,
+      "loss": 4.4209,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3648782637435276,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019282345264038246,
+      "loss": 4.4319,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3649819517980157,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001928194113681406,
+      "loss": 4.4293,
+      "step": 3520
+    },
+    {
+      "epoch": 0.36508563985250375,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019281536900072783,
+      "loss": 4.4231,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3651893279069918,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019281132553819182,
+      "loss": 4.435,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3652930159614799,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019280728098058022,
+      "loss": 4.444,
+      "step": 3523
+    },
+    {
+      "epoch": 0.36539670401596797,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019280323532794085,
+      "loss": 4.4257,
+      "step": 3524
+    },
+    {
+      "epoch": 0.36550039207045604,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001927991885803214,
+      "loss": 4.39,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3656040801249441,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019279514073776958,
+      "loss": 4.4168,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3657077681794322,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019279109180033322,
+      "loss": 4.4334,
+      "step": 3527
+    },
+    {
+      "epoch": 0.36581145623392025,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019278704176806008,
+      "loss": 4.4292,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3659151442884083,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001927829906409979,
+      "loss": 4.4186,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3660188323428964,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019277893841919451,
+      "loss": 4.4232,
+      "step": 3530
+    },
+    {
+      "epoch": 0.36612252039738447,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019277488510269773,
+      "loss": 4.4418,
+      "step": 3531
+    },
+    {
+      "epoch": 0.36622620845187254,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019277083069155535,
+      "loss": 4.4193,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3663298965063606,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019276677518581523,
+      "loss": 4.4612,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3664335845608487,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019276271858552523,
+      "loss": 4.4345,
+      "step": 3534
+    },
+    {
+      "epoch": 0.36653727261533675,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001927586608907332,
+      "loss": 4.4532,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3666409606698248,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019275460210148705,
+      "loss": 4.456,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3667446487243129,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019275054221783462,
+      "loss": 4.4461,
+      "step": 3537
+    },
+    {
+      "epoch": 0.36684833677880097,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019274648123982383,
+      "loss": 4.393,
+      "step": 3538
+    },
+    {
+      "epoch": 0.36695202483328904,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001927424191675026,
+      "loss": 4.4317,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3670557128877771,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019273835600091887,
+      "loss": 4.4409,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3671594009422652,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019273429174012057,
+      "loss": 4.4772,
+      "step": 3541
+    },
+    {
+      "epoch": 0.36726308899675325,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019273022638515564,
+      "loss": 4.4135,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3673667770512413,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019272615993607206,
+      "loss": 4.5041,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3674704651057294,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001927220923929178,
+      "loss": 4.449,
+      "step": 3544
+    },
+    {
+      "epoch": 0.36757415316021746,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019271802375574087,
+      "loss": 4.4044,
+      "step": 3545
+    },
+    {
+      "epoch": 0.36767784121470554,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019271395402458926,
+      "loss": 4.5054,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3677815292691936,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019270988319951103,
+      "loss": 4.4233,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3678852173236817,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019270581128055412,
+      "loss": 4.466,
+      "step": 3548
+    },
+    {
+      "epoch": 0.36798890537816975,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019270173826776667,
+      "loss": 4.4506,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3680925934326578,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001926976641611967,
+      "loss": 4.4769,
+      "step": 3550
+    },
+    {
+      "epoch": 0.36819628148714595,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019269358896089226,
+      "loss": 4.4227,
+      "step": 3551
+    },
+    {
+      "epoch": 0.368299969541634,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019268951266690146,
+      "loss": 4.4437,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3684036575961221,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019268543527927237,
+      "loss": 4.4569,
+      "step": 3553
+    },
+    {
+      "epoch": 0.36850734565061016,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019268135679805312,
+      "loss": 4.4919,
+      "step": 3554
+    },
+    {
+      "epoch": 0.36861103370509823,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001926772772232919,
+      "loss": 4.4385,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3687147217595863,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019267319655503665,
+      "loss": 4.4652,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3688184098140744,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019266911479333572,
+      "loss": 4.3996,
+      "step": 3557
+    },
+    {
+      "epoch": 0.36892209786856245,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019266503193823717,
+      "loss": 4.4644,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3690257859230505,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019266094798978922,
+      "loss": 4.4592,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3691294739775386,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019265686294804,
+      "loss": 4.444,
+      "step": 3560
+    },
+    {
+      "epoch": 0.36923316203202666,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019265277681303775,
+      "loss": 4.4513,
+      "step": 3561
+    },
+    {
+      "epoch": 0.36933685008651473,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019264868958483066,
+      "loss": 4.435,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3694405381410028,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019264460126346697,
+      "loss": 4.4655,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3695442261954909,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019264051184899494,
+      "loss": 4.4762,
+      "step": 3564
+    },
+    {
+      "epoch": 0.36964791424997895,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019263642134146277,
+      "loss": 4.4287,
+      "step": 3565
+    },
+    {
+      "epoch": 0.369751602304467,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019263232974091877,
+      "loss": 4.4286,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3698552903589551,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019262823704741119,
+      "loss": 4.4731,
+      "step": 3567
+    },
+    {
+      "epoch": 0.36995897841344316,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019262414326098832,
+      "loss": 4.4016,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37006266646793123,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019262004838169845,
+      "loss": 4.3934,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3701663545224193,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019261595240958993,
+      "loss": 4.4368,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3702700425769074,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019261185534471108,
+      "loss": 4.4612,
+      "step": 3571
+    },
+    {
+      "epoch": 0.37037373063139545,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001926077571871102,
+      "loss": 4.4292,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3704774186858835,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019260365793683572,
+      "loss": 4.421,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3705811067403716,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019259955759393593,
+      "loss": 4.4411,
+      "step": 3574
+    },
+    {
+      "epoch": 0.37068479479485966,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019259545615845925,
+      "loss": 4.4327,
+      "step": 3575
+    },
+    {
+      "epoch": 0.37078848284934773,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019259135363045406,
+      "loss": 4.4329,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3708921709038358,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019258725000996881,
+      "loss": 4.4206,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3709958589583239,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001925831452970518,
+      "loss": 4.4502,
+      "step": 3578
+    },
+    {
+      "epoch": 0.37109954701281195,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001925790394917516,
+      "loss": 4.4349,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3712032350673,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019257493259411659,
+      "loss": 4.4454,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3713069231217881,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001925708246041952,
+      "loss": 4.4163,
+      "step": 3581
+    },
+    {
+      "epoch": 0.37141061117627616,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019256671552203596,
+      "loss": 4.4385,
+      "step": 3582
+    },
+    {
+      "epoch": 0.37151429923076423,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019256260534768733,
+      "loss": 4.4556,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3716179872852523,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019255849408119778,
+      "loss": 4.4005,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3717216753397404,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019255438172261586,
+      "loss": 4.4442,
+      "step": 3585
+    },
+    {
+      "epoch": 0.37182536339422845,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019255026827199006,
+      "loss": 4.4232,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3719290514487165,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001925461537293689,
+      "loss": 4.4578,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3720327395032046,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019254203809480097,
+      "loss": 4.4158,
+      "step": 3588
+    },
+    {
+      "epoch": 0.37213642755769266,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019253792136833482,
+      "loss": 4.4179,
+      "step": 3589
+    },
+    {
+      "epoch": 0.37224011561218073,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.000192533803550019,
+      "loss": 4.4066,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3723438036666688,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001925296846399021,
+      "loss": 4.4247,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3724474917211569,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019252556463803279,
+      "loss": 4.4401,
+      "step": 3592
+    },
+    {
+      "epoch": 0.37255117977564495,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019252144354445957,
+      "loss": 4.4367,
+      "step": 3593
+    },
+    {
+      "epoch": 0.372654867830133,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001925173213592311,
+      "loss": 4.3991,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3727585558846211,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019251319808239609,
+      "loss": 4.4147,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3728622439391092,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019250907371400308,
+      "loss": 4.4117,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3729659319935973,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001925049482541008,
+      "loss": 4.4487,
+      "step": 3597
+    },
+    {
+      "epoch": 0.37306962004808536,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019250082170273793,
+      "loss": 4.4308,
+      "step": 3598
+    },
+    {
+      "epoch": 0.37317330810257343,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001924966940599631,
+      "loss": 4.4148,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3732769961570615,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001924925653258251,
+      "loss": 4.4425,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3733806842115496,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001924884355003726,
+      "loss": 4.4362,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37348437226603765,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001924843045836543,
+      "loss": 4.4168,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3735880603205257,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000192480172575719,
+      "loss": 4.4619,
+      "step": 3603
+    },
+    {
+      "epoch": 0.3736917483750138,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019247603947661535,
+      "loss": 4.424,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37379543642950186,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019247190528639223,
+      "loss": 4.4545,
+      "step": 3605
+    },
+    {
+      "epoch": 0.37389912448398993,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019246777000509838,
+      "loss": 4.4457,
+      "step": 3606
+    },
+    {
+      "epoch": 0.374002812538478,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019246363363278259,
+      "loss": 4.4527,
+      "step": 3607
+    },
+    {
+      "epoch": 0.3741065005929661,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019245949616949367,
+      "loss": 4.4568,
+      "step": 3608
+    },
+    {
+      "epoch": 0.37421018864745415,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019245535761528037,
+      "loss": 4.4389,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3743138767019422,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019245121797019165,
+      "loss": 4.411,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3744175647564303,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019244707723427623,
+      "loss": 4.3858,
+      "step": 3611
+    },
+    {
+      "epoch": 0.37452125281091836,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019244293540758304,
+      "loss": 4.4078,
+      "step": 3612
+    },
+    {
+      "epoch": 0.37462494086540643,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019243879249016094,
+      "loss": 4.4748,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3747286289198945,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001924346484820588,
+      "loss": 4.439,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3748323169743826,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001924305033833255,
+      "loss": 4.3731,
+      "step": 3615
+    },
+    {
+      "epoch": 0.37493600502887064,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019242635719400996,
+      "loss": 4.4164,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3750396930833587,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019242220991416112,
+      "loss": 4.4429,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3751433811378468,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019241806154382792,
+      "loss": 4.3936,
+      "step": 3618
+    },
+    {
+      "epoch": 0.37524706919233486,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019241391208305926,
+      "loss": 4.4168,
+      "step": 3619
+    },
+    {
+      "epoch": 0.37535075724682293,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001924097615319041,
+      "loss": 4.4512,
+      "step": 3620
+    },
+    {
+      "epoch": 0.375454445301311,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019240560989041146,
+      "loss": 4.4483,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3755581333557991,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001924014571586303,
+      "loss": 4.4152,
+      "step": 3622
+    },
+    {
+      "epoch": 0.37566182141028714,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001923973033366096,
+      "loss": 4.4342,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3757655094647752,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001923931484243984,
+      "loss": 4.3984,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3758691975192633,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001923889924220457,
+      "loss": 4.433,
+      "step": 3625
+    },
+    {
+      "epoch": 0.37597288557375136,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019238483532960058,
+      "loss": 4.4543,
+      "step": 3626
+    },
+    {
+      "epoch": 0.37607657362823943,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019238067714711203,
+      "loss": 4.4161,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3761802616827275,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019237651787462916,
+      "loss": 4.4185,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3762839497372156,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019237235751220103,
+      "loss": 4.4455,
+      "step": 3629
+    },
+    {
+      "epoch": 0.37638763779170364,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001923681960598767,
+      "loss": 4.409,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3764913258461917,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001923640335177053,
+      "loss": 4.4611,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3765950139006798,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019235986988573595,
+      "loss": 4.4378,
+      "step": 3632
+    },
+    {
+      "epoch": 0.37669870195516786,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019235570516401776,
+      "loss": 4.4345,
+      "step": 3633
+    },
+    {
+      "epoch": 0.37680239000965593,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019235153935259986,
+      "loss": 4.458,
+      "step": 3634
+    },
+    {
+      "epoch": 0.376906078064144,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019234737245153142,
+      "loss": 4.4146,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3770097661186321,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001923432044608616,
+      "loss": 4.4413,
+      "step": 3636
+    },
+    {
+      "epoch": 0.37711345417312014,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019233903538063958,
+      "loss": 4.4489,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3772171422276082,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019233486521091458,
+      "loss": 4.4059,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3773208302820963,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019233069395173573,
+      "loss": 4.4205,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3774245183365844,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001923265216031523,
+      "loss": 4.4273,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3775282063910725,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019232234816521352,
+      "loss": 4.4103,
+      "step": 3641
+    },
+    {
+      "epoch": 0.37763189444556056,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001923181736379686,
+      "loss": 4.44,
+      "step": 3642
+    },
+    {
+      "epoch": 0.37773558250004863,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019231399802146685,
+      "loss": 4.4276,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3778392705545367,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019230982131575747,
+      "loss": 4.4077,
+      "step": 3644
+    },
+    {
+      "epoch": 0.37794295860902477,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019230564352088977,
+      "loss": 4.4196,
+      "step": 3645
+    },
+    {
+      "epoch": 0.37804664666351284,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019230146463691307,
+      "loss": 4.4645,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3781503347180009,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019229728466387664,
+      "loss": 4.4093,
+      "step": 3647
+    },
+    {
+      "epoch": 0.378254022772489,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019229310360182984,
+      "loss": 4.4172,
+      "step": 3648
+    },
+    {
+      "epoch": 0.37835771082697706,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019228892145082193,
+      "loss": 4.4356,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3784613988814651,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019228473821090235,
+      "loss": 4.4533,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3785650869359532,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001922805538821204,
+      "loss": 4.4694,
+      "step": 3651
+    },
+    {
+      "epoch": 0.37866877499044127,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019227636846452542,
+      "loss": 4.4129,
+      "step": 3652
+    },
+    {
+      "epoch": 0.37877246304492934,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019227218195816685,
+      "loss": 4.436,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3788761510994174,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001922679943630941,
+      "loss": 4.4356,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3789798391539055,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001922638056793565,
+      "loss": 4.4148,
+      "step": 3655
+    },
+    {
+      "epoch": 0.37908352720839356,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019225961590700353,
+      "loss": 4.4184,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3791872152628816,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019225542504608465,
+      "loss": 4.4467,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3792909033173697,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019225123309664924,
+      "loss": 4.4668,
+      "step": 3658
+    },
+    {
+      "epoch": 0.37939459137185777,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019224704005874684,
+      "loss": 4.4578,
+      "step": 3659
+    },
+    {
+      "epoch": 0.37949827942634584,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001922428459324268,
+      "loss": 4.4322,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3796019674808339,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019223865071773874,
+      "loss": 4.4463,
+      "step": 3661
+    },
+    {
+      "epoch": 0.379705655535322,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019223445441473208,
+      "loss": 4.425,
+      "step": 3662
+    },
+    {
+      "epoch": 0.37980934358981006,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019223025702345639,
+      "loss": 4.4147,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3799130316442981,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019222605854396111,
+      "loss": 4.4286,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3800167196987862,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019222185897629584,
+      "loss": 4.4276,
+      "step": 3665
+    },
+    {
+      "epoch": 0.38012040775327427,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019221765832051012,
+      "loss": 4.4228,
+      "step": 3666
+    },
+    {
+      "epoch": 0.38022409580776234,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019221345657665352,
+      "loss": 4.4041,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3803277838622504,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001922092537447756,
+      "loss": 4.447,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3804314719167385,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019220504982492596,
+      "loss": 4.4779,
+      "step": 3669
+    },
+    {
+      "epoch": 0.38053515997122656,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019220084481715418,
+      "loss": 4.4111,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3806388480257146,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001921966387215099,
+      "loss": 4.4375,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3807425360802027,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019219243153804273,
+      "loss": 4.4184,
+      "step": 3672
+    },
+    {
+      "epoch": 0.38084622413469077,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019218822326680235,
+      "loss": 4.4274,
+      "step": 3673
+    },
+    {
+      "epoch": 0.38094991218917884,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019218401390783834,
+      "loss": 4.4228,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3810536002436669,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019217980346120044,
+      "loss": 4.4196,
+      "step": 3675
+    },
+    {
+      "epoch": 0.381157288298155,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019217559192693831,
+      "loss": 4.4682,
+      "step": 3676
+    },
+    {
+      "epoch": 0.38126097635264306,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001921713793051016,
+      "loss": 4.4386,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3813646644071311,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019216716559574003,
+      "loss": 4.3986,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3814683524616192,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019216295079890338,
+      "loss": 4.4526,
+      "step": 3679
+    },
+    {
+      "epoch": 0.38157204051610727,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001921587349146413,
+      "loss": 4.4512,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38167572857059534,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019215451794300355,
+      "loss": 4.456,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3817794166250834,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019215029988403992,
+      "loss": 4.4242,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3818831046795715,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019214608073780015,
+      "loss": 4.4795,
+      "step": 3683
+    },
+    {
+      "epoch": 0.38198679273405955,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019214186050433405,
+      "loss": 4.427,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3820904807885477,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019213763918369137,
+      "loss": 4.4174,
+      "step": 3685
+    },
+    {
+      "epoch": 0.38219416884303575,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019213341677592197,
+      "loss": 4.4468,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3822978568975238,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019212919328107564,
+      "loss": 4.4515,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3824015449520119,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019212496869920218,
+      "loss": 4.4227,
+      "step": 3688
+    },
+    {
+      "epoch": 0.38250523300649997,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001921207430303515,
+      "loss": 4.457,
+      "step": 3689
+    },
+    {
+      "epoch": 0.38260892106098804,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001921165162745734,
+      "loss": 4.4166,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3827126091154761,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001921122884319178,
+      "loss": 4.3881,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3828162971699642,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019210805950243455,
+      "loss": 4.3417,
+      "step": 3692
+    },
+    {
+      "epoch": 0.38291998522445225,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019210382948617357,
+      "loss": 4.4258,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3830236732789403,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019209959838318476,
+      "loss": 4.4543,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3831273613334284,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019209536619351804,
+      "loss": 4.4249,
+      "step": 3695
+    },
+    {
+      "epoch": 0.38323104938791647,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019209113291722334,
+      "loss": 4.4234,
+      "step": 3696
+    },
+    {
+      "epoch": 0.38333473744240454,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001920868985543506,
+      "loss": 4.4391,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3834384254968926,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019208266310494985,
+      "loss": 4.3681,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3835421135513807,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019207842656907099,
+      "loss": 4.4259,
+      "step": 3699
+    },
+    {
+      "epoch": 0.38364580160586875,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000192074188946764,
+      "loss": 4.4409,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3837494896603568,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019206995023807893,
+      "loss": 4.4505,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3838531777148449,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019206571044306578,
+      "loss": 4.4179,
+      "step": 3702
+    },
+    {
+      "epoch": 0.38395686576933297,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019206146956177454,
+      "loss": 4.4102,
+      "step": 3703
+    },
+    {
+      "epoch": 0.38406055382382104,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019205722759425527,
+      "loss": 4.4051,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3841642418783091,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019205298454055806,
+      "loss": 4.4391,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3842679299327972,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001920487404007329,
+      "loss": 4.4612,
+      "step": 3706
+    },
+    {
+      "epoch": 0.38437161798728525,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019204449517482993,
+      "loss": 4.4197,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3844753060417733,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019204024886289919,
+      "loss": 4.4594,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3845789940962614,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001920360014649908,
+      "loss": 4.4012,
+      "step": 3709
+    },
+    {
+      "epoch": 0.38468268215074947,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019203175298115492,
+      "loss": 4.4191,
+      "step": 3710
+    },
+    {
+      "epoch": 0.38478637020523754,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001920275034114416,
+      "loss": 4.3976,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3848900582597256,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019202325275590102,
+      "loss": 4.3799,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3849937463142137,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019201900101458333,
+      "loss": 4.4567,
+      "step": 3713
+    },
+    {
+      "epoch": 0.38509743436870175,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001920147481875387,
+      "loss": 4.3867,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3852011224231898,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001920104942748173,
+      "loss": 4.4149,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3853048104776779,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019200623927646934,
+      "loss": 4.4151,
+      "step": 3716
+    },
+    {
+      "epoch": 0.38540849853216597,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.000192001983192545,
+      "loss": 4.4309,
+      "step": 3717
+    },
+    {
+      "epoch": 0.38551218658665404,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019199772602309452,
+      "loss": 4.37,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3856158746411421,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019199346776816812,
+      "loss": 4.4452,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3857195626956302,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019198920842781604,
+      "loss": 4.4276,
+      "step": 3720
+    },
+    {
+      "epoch": 0.38582325075011825,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019198494800208853,
+      "loss": 4.4077,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3859269388046063,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019198068649103585,
+      "loss": 4.3955,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3860306268590944,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019197642389470832,
+      "loss": 4.4649,
+      "step": 3723
+    },
+    {
+      "epoch": 0.38613431491358247,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019197216021315622,
+      "loss": 4.4707,
+      "step": 3724
+    },
+    {
+      "epoch": 0.38623800296807054,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019196789544642981,
+      "loss": 4.4139,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3863416910225586,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019196362959457948,
+      "loss": 4.4305,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3864453790770467,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019195936265765552,
+      "loss": 4.4144,
+      "step": 3727
+    },
+    {
+      "epoch": 0.38654906713153475,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019195509463570831,
+      "loss": 4.3813,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3866527551860228,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019195082552878815,
+      "loss": 4.4248,
+      "step": 3729
+    },
+    {
+      "epoch": 0.38675644324051095,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019194655533694545,
+      "loss": 4.4362,
+      "step": 3730
+    },
+    {
+      "epoch": 0.386860131294999,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001919422840602306,
+      "loss": 4.4018,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3869638193494871,
+      "grad_norm": 1.25,
+      "learning_rate": 0.000191938011698694,
+      "loss": 4.4026,
+      "step": 3732
+    },
+    {
+      "epoch": 0.38706750740397516,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019193373825238602,
+      "loss": 4.46,
+      "step": 3733
+    },
+    {
+      "epoch": 0.38717119545846324,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019192946372135713,
+      "loss": 4.3934,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3872748835129513,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019192518810565772,
+      "loss": 4.4224,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3873785715674394,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019192091140533824,
+      "loss": 4.4006,
+      "step": 3736
+    },
+    {
+      "epoch": 0.38748225962192745,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001919166336204492,
+      "loss": 4.4281,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3875859476764155,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019191235475104104,
+      "loss": 4.4234,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3876896357309036,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019190807479716425,
+      "loss": 4.4203,
+      "step": 3739
+    },
+    {
+      "epoch": 0.38779332378539166,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019190379375886934,
+      "loss": 4.4245,
+      "step": 3740
+    },
+    {
+      "epoch": 0.38789701183987974,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019189951163620678,
+      "loss": 4.4599,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3880006998943678,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019189522842922714,
+      "loss": 4.3865,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3881043879488559,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00019189094413798094,
+      "loss": 4.4373,
+      "step": 3743
+    },
+    {
+      "epoch": 0.38820807600334395,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019188665876251874,
+      "loss": 4.4681,
+      "step": 3744
+    },
+    {
+      "epoch": 0.388311764057832,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001918823723028911,
+      "loss": 4.4097,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3884154521123201,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019187808475914855,
+      "loss": 4.4054,
+      "step": 3746
+    },
+    {
+      "epoch": 0.38851914016680816,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019187379613134174,
+      "loss": 4.394,
+      "step": 3747
+    },
+    {
+      "epoch": 0.38862282822129623,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019186950641952128,
+      "loss": 4.4425,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3887265162757843,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019186521562373774,
+      "loss": 4.4016,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3888302043302724,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019186092374404176,
+      "loss": 4.3771,
+      "step": 3750
+    },
+    {
+      "epoch": 0.38893389238476045,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019185663078048396,
+      "loss": 4.4122,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3890375804392485,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019185233673311506,
+      "loss": 4.4324,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3891412684937366,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001918480416019856,
+      "loss": 4.4282,
+      "step": 3753
+    },
+    {
+      "epoch": 0.38924495654822466,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001918437453871464,
+      "loss": 4.4434,
+      "step": 3754
+    },
+    {
+      "epoch": 0.38934864460271273,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019183944808864807,
+      "loss": 4.4197,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3894523326572008,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019183514970654136,
+      "loss": 4.3956,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3895560207116889,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001918308502408769,
+      "loss": 4.4278,
+      "step": 3757
+    },
+    {
+      "epoch": 0.38965970876617695,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019182654969170551,
+      "loss": 4.4023,
+      "step": 3758
+    },
+    {
+      "epoch": 0.389763396820665,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001918222480590779,
+      "loss": 4.4272,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3898670848751531,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019181794534304484,
+      "loss": 4.4145,
+      "step": 3760
+    },
+    {
+      "epoch": 0.38997077292964116,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019181364154365705,
+      "loss": 4.4161,
+      "step": 3761
+    },
+    {
+      "epoch": 0.39007446098412923,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019180933666096536,
+      "loss": 4.4161,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3901781490386173,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019180503069502052,
+      "loss": 4.4283,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3902818370931054,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001918007236458734,
+      "loss": 4.4056,
+      "step": 3764
+    },
+    {
+      "epoch": 0.39038552514759345,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019179641551357474,
+      "loss": 4.4236,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3904892132020815,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019179210629817543,
+      "loss": 4.3897,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3905929012565696,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001917877959997263,
+      "loss": 4.3884,
+      "step": 3767
+    },
+    {
+      "epoch": 0.39069658931105766,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019178348461827818,
+      "loss": 4.4123,
+      "step": 3768
+    },
+    {
+      "epoch": 0.39080027736554573,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019177917215388196,
+      "loss": 4.4079,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3909039654200338,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019177485860658852,
+      "loss": 4.4478,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3910076534745219,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019177054397644878,
+      "loss": 4.4614,
+      "step": 3771
+    },
+    {
+      "epoch": 0.39111134152900995,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019176622826351362,
+      "loss": 4.416,
+      "step": 3772
+    },
+    {
+      "epoch": 0.391215029583498,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019176191146783394,
+      "loss": 4.4209,
+      "step": 3773
+    },
+    {
+      "epoch": 0.39131871763798615,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019175759358946076,
+      "loss": 4.4549,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3914224056924742,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019175327462844495,
+      "loss": 4.411,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3915260937469623,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019174895458483744,
+      "loss": 4.4674,
+      "step": 3776
+    },
+    {
+      "epoch": 0.39162978180145036,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001917446334586893,
+      "loss": 4.4098,
+      "step": 3777
+    },
+    {
+      "epoch": 0.39173346985593843,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019174031125005145,
+      "loss": 4.4235,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3918371579104265,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001917359879589749,
+      "loss": 4.4081,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3919408459649146,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019173166358551065,
+      "loss": 4.4185,
+      "step": 3780
+    },
+    {
+      "epoch": 0.39204453401940265,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019172733812970974,
+      "loss": 4.4354,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3921482220738907,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019172301159162318,
+      "loss": 4.4443,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3922519101283788,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019171868397130205,
+      "loss": 4.4231,
+      "step": 3783
+    },
+    {
+      "epoch": 0.39235559818286686,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019171435526879743,
+      "loss": 4.452,
+      "step": 3784
+    },
+    {
+      "epoch": 0.39245928623735493,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019171002548416035,
+      "loss": 4.3952,
+      "step": 3785
+    },
+    {
+      "epoch": 0.392562974291843,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001917056946174419,
+      "loss": 4.37,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3926666623463311,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019170136266869319,
+      "loss": 4.4024,
+      "step": 3787
+    },
+    {
+      "epoch": 0.39277035040081915,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019169702963796534,
+      "loss": 4.3948,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3928740384553072,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019169269552530947,
+      "loss": 4.3902,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3929777265097953,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019168836033077672,
+      "loss": 4.4408,
+      "step": 3790
+    },
+    {
+      "epoch": 0.39308141456428336,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001916840240544182,
+      "loss": 4.4168,
+      "step": 3791
+    },
+    {
+      "epoch": 0.39318510261877143,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019167968669628517,
+      "loss": 4.3858,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3932887906732595,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001916753482564287,
+      "loss": 4.4098,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3933924787277476,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019167100873490004,
+      "loss": 4.4368,
+      "step": 3794
+    },
+    {
+      "epoch": 0.39349616678223565,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019166666813175037,
+      "loss": 4.4083,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3935998548367237,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019166232644703092,
+      "loss": 4.3992,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3937035428912118,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019165798368079293,
+      "loss": 4.4001,
+      "step": 3797
+    },
+    {
+      "epoch": 0.39380723094569986,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001916536398330876,
+      "loss": 4.3935,
+      "step": 3798
+    },
+    {
+      "epoch": 0.39391091900018793,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001916492949039662,
+      "loss": 4.4262,
+      "step": 3799
+    },
+    {
+      "epoch": 0.394014607054676,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019164494889348,
+      "loss": 4.4222,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3941182951091641,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019164060180168027,
+      "loss": 4.4313,
+      "step": 3801
+    },
+    {
+      "epoch": 0.39422198316365215,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001916362536286183,
+      "loss": 4.3749,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3943256712181402,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019163190437434543,
+      "loss": 4.4329,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3944293592726283,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019162755403891293,
+      "loss": 4.4685,
+      "step": 3804
+    },
+    {
+      "epoch": 0.39453304732711636,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019162320262237215,
+      "loss": 4.4298,
+      "step": 3805
+    },
+    {
+      "epoch": 0.39463673538160443,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019161885012477443,
+      "loss": 4.4126,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3947404234360925,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019161449654617112,
+      "loss": 4.4101,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3948441114905806,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001916101418866136,
+      "loss": 4.3926,
+      "step": 3808
+    },
+    {
+      "epoch": 0.39494779954506865,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019160578614615321,
+      "loss": 4.4267,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3950514875995567,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019160142932484142,
+      "loss": 4.4472,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3951551756540448,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019159707142272952,
+      "loss": 4.4031,
+      "step": 3811
+    },
+    {
+      "epoch": 0.39525886370853286,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001915927124398691,
+      "loss": 4.4008,
+      "step": 3812
+    },
+    {
+      "epoch": 0.39536255176302093,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019158835237631142,
+      "loss": 4.3912,
+      "step": 3813
+    },
+    {
+      "epoch": 0.395466239817509,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019158399123210803,
+      "loss": 4.3948,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3955699278719971,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001915796290073103,
+      "loss": 4.3967,
+      "step": 3815
+    },
+    {
+      "epoch": 0.39567361592648514,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019157526570196982,
+      "loss": 4.4215,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3957773039809732,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019157090131613797,
+      "loss": 4.4453,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3958809920354613,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019156653584986627,
+      "loss": 4.4207,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3959846800899494,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019156216930320624,
+      "loss": 4.372,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3960883681444375,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001915578016762094,
+      "loss": 4.4139,
+      "step": 3820
+    },
+    {
+      "epoch": 0.39619205619892556,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001915534329689273,
+      "loss": 4.4365,
+      "step": 3821
+    },
+    {
+      "epoch": 0.39629574425341363,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019154906318141147,
+      "loss": 4.3953,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3963994323079017,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019154469231371345,
+      "loss": 4.4415,
+      "step": 3823
+    },
+    {
+      "epoch": 0.39650312036238977,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001915403203658848,
+      "loss": 4.4293,
+      "step": 3824
+    },
+    {
+      "epoch": 0.39660680841687784,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019153594733797716,
+      "loss": 4.4074,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3967104964713659,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019153157323004212,
+      "loss": 4.4193,
+      "step": 3826
+    },
+    {
+      "epoch": 0.396814184525854,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019152719804213124,
+      "loss": 4.4267,
+      "step": 3827
+    },
+    {
+      "epoch": 0.39691787258034206,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001915228217742962,
+      "loss": 4.4383,
+      "step": 3828
+    },
+    {
+      "epoch": 0.39702156063483013,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019151844442658858,
+      "loss": 4.3961,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3971252486893182,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019151406599906005,
+      "loss": 4.4261,
+      "step": 3830
+    },
+    {
+      "epoch": 0.39722893674380627,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019150968649176228,
+      "loss": 4.4638,
+      "step": 3831
+    },
+    {
+      "epoch": 0.39733262479829434,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019150530590474694,
+      "loss": 4.4539,
+      "step": 3832
+    },
+    {
+      "epoch": 0.3974363128527824,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019150092423806574,
+      "loss": 4.4545,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3975400009072705,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019149654149177035,
+      "loss": 4.4372,
+      "step": 3834
+    },
+    {
+      "epoch": 0.39764368896175856,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00019149215766591247,
+      "loss": 4.401,
+      "step": 3835
+    },
+    {
+      "epoch": 0.39774737701624663,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019148777276054385,
+      "loss": 4.4252,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3978510650707347,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001914833867757162,
+      "loss": 4.417,
+      "step": 3837
+    },
+    {
+      "epoch": 0.39795475312522277,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001914789997114813,
+      "loss": 4.428,
+      "step": 3838
+    },
+    {
+      "epoch": 0.39805844117971084,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019147461156789088,
+      "loss": 4.401,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3981621292341989,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019147022234499675,
+      "loss": 4.4456,
+      "step": 3840
+    },
+    {
+      "epoch": 0.398265817288687,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019146583204285073,
+      "loss": 4.425,
+      "step": 3841
+    },
+    {
+      "epoch": 0.39836950534317506,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019146144066150452,
+      "loss": 4.456,
+      "step": 3842
+    },
+    {
+      "epoch": 0.39847319339766313,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019145704820101,
+      "loss": 4.3935,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3985768814521512,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019145265466141898,
+      "loss": 4.4166,
+      "step": 3844
+    },
+    {
+      "epoch": 0.39868056950663927,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019144826004278333,
+      "loss": 4.4599,
+      "step": 3845
+    },
+    {
+      "epoch": 0.39878425756112734,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019144386434515487,
+      "loss": 4.4153,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3988879456156154,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019143946756858547,
+      "loss": 4.3922,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3989916336701035,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019143506971312698,
+      "loss": 4.4292,
+      "step": 3848
+    },
+    {
+      "epoch": 0.39909532172459156,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019143067077883138,
+      "loss": 4.4366,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3991990097790796,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019142627076575047,
+      "loss": 4.3678,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3993026978335677,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001914218696739362,
+      "loss": 4.3722,
+      "step": 3851
+    },
+    {
+      "epoch": 0.39940638588805577,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001914174675034405,
+      "loss": 4.4099,
+      "step": 3852
+    },
+    {
+      "epoch": 0.39951007394254384,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019141306425431533,
+      "loss": 4.4257,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3996137619970319,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00019140865992661263,
+      "loss": 4.4804,
+      "step": 3854
+    },
+    {
+      "epoch": 0.39971745005152,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019140425452038437,
+      "loss": 4.4406,
+      "step": 3855
+    },
+    {
+      "epoch": 0.39982113810600806,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001913998480356825,
+      "loss": 4.4184,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3999248261604961,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019139544047255908,
+      "loss": 4.3869,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4000285142149842,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019139103183106606,
+      "loss": 4.3996,
+      "step": 3858
+    },
+    {
+      "epoch": 0.40013220226947227,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019138662211125546,
+      "loss": 4.3957,
+      "step": 3859
+    },
+    {
+      "epoch": 0.40023589032396034,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001913822113131793,
+      "loss": 4.4421,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4003395783784484,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019137779943688966,
+      "loss": 4.4283,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4004432664329365,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019137338648243855,
+      "loss": 4.4069,
+      "step": 3862
+    },
+    {
+      "epoch": 0.40054695448742456,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019136897244987814,
+      "loss": 4.4076,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4006506425419127,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019136455733926035,
+      "loss": 4.4231,
+      "step": 3864
+    },
+    {
+      "epoch": 0.40075433059640075,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001913601411506374,
+      "loss": 4.4225,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4008580186508888,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019135572388406135,
+      "loss": 4.4317,
+      "step": 3866
+    },
+    {
+      "epoch": 0.4009617067053769,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019135130553958433,
+      "loss": 4.4126,
+      "step": 3867
+    },
+    {
+      "epoch": 0.40106539475986497,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001913468861172585,
+      "loss": 4.4147,
+      "step": 3868
+    },
+    {
+      "epoch": 0.40116908281435304,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001913424656171359,
+      "loss": 4.407,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4012727708688411,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019133804403926883,
+      "loss": 4.4141,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4013764589233292,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019133362138370935,
+      "loss": 4.3927,
+      "step": 3871
+    },
+    {
+      "epoch": 0.40148014697781725,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019132919765050969,
+      "loss": 4.3875,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4015838350323053,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000191324772839722,
+      "loss": 4.4231,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4016875230867934,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019132034695139858,
+      "loss": 4.3975,
+      "step": 3874
+    },
+    {
+      "epoch": 0.40179121114128147,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019131591998559157,
+      "loss": 4.3899,
+      "step": 3875
+    },
+    {
+      "epoch": 0.40189489919576954,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019131149194235323,
+      "loss": 4.3924,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4019985872502576,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001913070628217358,
+      "loss": 4.4521,
+      "step": 3877
+    },
+    {
+      "epoch": 0.4021022753047457,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001913026326237916,
+      "loss": 4.4667,
+      "step": 3878
+    },
+    {
+      "epoch": 0.40220596335923375,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019129820134857278,
+      "loss": 4.4182,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4023096514137218,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001912937689961317,
+      "loss": 4.4307,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4024133394682099,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019128933556652067,
+      "loss": 4.4365,
+      "step": 3881
+    },
+    {
+      "epoch": 0.40251702752269797,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019128490105979195,
+      "loss": 4.4601,
+      "step": 3882
+    },
+    {
+      "epoch": 0.40262071557718604,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019128046547599791,
+      "loss": 4.4408,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4027244036316741,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019127602881519086,
+      "loss": 4.4029,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4028280916861622,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019127159107742316,
+      "loss": 4.4356,
+      "step": 3885
+    },
+    {
+      "epoch": 0.40293177974065025,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019126715226274715,
+      "loss": 4.4414,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4030354677951383,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019126271237121523,
+      "loss": 4.4243,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4031391558496264,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019125827140287975,
+      "loss": 4.3972,
+      "step": 3888
+    },
+    {
+      "epoch": 0.40324284390411447,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019125382935779314,
+      "loss": 4.3953,
+      "step": 3889
+    },
+    {
+      "epoch": 0.40334653195860254,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001912493862360078,
+      "loss": 4.4277,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4034502200130906,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001912449420375762,
+      "loss": 4.4313,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4035539080675787,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019124049676255068,
+      "loss": 4.4294,
+      "step": 3892
+    },
+    {
+      "epoch": 0.40365759612206675,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019123605041098377,
+      "loss": 4.4509,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4037612841765548,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019123160298292792,
+      "loss": 4.4061,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4038649722310429,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019122715447843555,
+      "loss": 4.4158,
+      "step": 3895
+    },
+    {
+      "epoch": 0.40396866028553097,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019122270489755922,
+      "loss": 4.3951,
+      "step": 3896
+    },
+    {
+      "epoch": 0.40407234834001904,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019121825424035137,
+      "loss": 4.3816,
+      "step": 3897
+    },
+    {
+      "epoch": 0.4041760363945071,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001912138025068646,
+      "loss": 4.4206,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4042797244489952,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019120934969715133,
+      "loss": 4.4204,
+      "step": 3899
+    },
+    {
+      "epoch": 0.40438341250348325,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019120489581126414,
+      "loss": 4.4165,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4044871005579713,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001912004408492556,
+      "loss": 4.4463,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4045907886124594,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019119598481117826,
+      "loss": 4.4039,
+      "step": 3902
+    },
+    {
+      "epoch": 0.40469447666694747,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019119152769708474,
+      "loss": 4.4281,
+      "step": 3903
+    },
+    {
+      "epoch": 0.40479816472143554,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019118706950702752,
+      "loss": 4.4457,
+      "step": 3904
+    },
+    {
+      "epoch": 0.4049018527759236,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001911826102410593,
+      "loss": 4.4312,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4050055408304117,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019117814989923267,
+      "loss": 4.4275,
+      "step": 3906
+    },
+    {
+      "epoch": 0.40510922888489975,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019117368848160023,
+      "loss": 4.4088,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4052129169393879,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019116922598821464,
+      "loss": 4.4433,
+      "step": 3908
+    },
+    {
+      "epoch": 0.40531660499387595,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00019116476241912858,
+      "loss": 4.4212,
+      "step": 3909
+    },
+    {
+      "epoch": 0.405420293048364,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019116029777439467,
+      "loss": 4.3693,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4055239811028521,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001911558320540656,
+      "loss": 4.416,
+      "step": 3911
+    },
+    {
+      "epoch": 0.40562766915734016,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019115136525819408,
+      "loss": 4.4272,
+      "step": 3912
+    },
+    {
+      "epoch": 0.40573135721182824,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019114689738683281,
+      "loss": 4.3669,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4058350452663163,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001911424284400345,
+      "loss": 4.4228,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4059387333208044,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019113795841785183,
+      "loss": 4.415,
+      "step": 3915
+    },
+    {
+      "epoch": 0.40604242137529245,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019113348732033764,
+      "loss": 4.4196,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4061461094297805,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001911290151475446,
+      "loss": 4.3969,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4062497974842686,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001911245418995255,
+      "loss": 4.3875,
+      "step": 3918
+    },
+    {
+      "epoch": 0.40635348553875666,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019112006757633317,
+      "loss": 4.3719,
+      "step": 3919
+    },
+    {
+      "epoch": 0.40645717359324474,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019111559217802033,
+      "loss": 4.3987,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4065608616477328,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019111111570463982,
+      "loss": 4.434,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4066645497022209,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019110663815624448,
+      "loss": 4.4007,
+      "step": 3922
+    },
+    {
+      "epoch": 0.40676823775670895,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019110215953288708,
+      "loss": 4.3783,
+      "step": 3923
+    },
+    {
+      "epoch": 0.406871925811197,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001910976798346205,
+      "loss": 4.4016,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4069756138656851,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001910931990614976,
+      "loss": 4.4789,
+      "step": 3925
+    },
+    {
+      "epoch": 0.40707930192017316,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019108871721357126,
+      "loss": 4.419,
+      "step": 3926
+    },
+    {
+      "epoch": 0.40718298997466124,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019108423429089433,
+      "loss": 4.389,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4072866780291493,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001910797502935197,
+      "loss": 4.4432,
+      "step": 3928
+    },
+    {
+      "epoch": 0.4073903660836374,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001910752652215003,
+      "loss": 4.4265,
+      "step": 3929
+    },
+    {
+      "epoch": 0.40749405413812545,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019107077907488905,
+      "loss": 4.3687,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4075977421926135,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019106629185373887,
+      "loss": 4.4174,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4077014302471016,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001910618035581027,
+      "loss": 4.4237,
+      "step": 3932
+    },
+    {
+      "epoch": 0.40780511830158966,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001910573141880335,
+      "loss": 4.4584,
+      "step": 3933
+    },
+    {
+      "epoch": 0.40790880635607774,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019105282374358424,
+      "loss": 4.4041,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4080124944105658,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00019104833222480792,
+      "loss": 4.4426,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4081161824650539,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019104383963175755,
+      "loss": 4.4512,
+      "step": 3936
+    },
+    {
+      "epoch": 0.40821987051954195,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019103934596448606,
+      "loss": 4.4471,
+      "step": 3937
+    },
+    {
+      "epoch": 0.40832355857403,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019103485122304653,
+      "loss": 4.432,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4084272466285181,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.000191030355407492,
+      "loss": 4.3946,
+      "step": 3939
+    },
+    {
+      "epoch": 0.40853093468300616,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001910258585178755,
+      "loss": 4.4321,
+      "step": 3940
+    },
+    {
+      "epoch": 0.40863462273749424,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019102136055425007,
+      "loss": 4.4022,
+      "step": 3941
+    },
+    {
+      "epoch": 0.4087383107919823,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001910168615166688,
+      "loss": 4.4023,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4088419988464704,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001910123614051848,
+      "loss": 4.4107,
+      "step": 3943
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019100786021985112,
+      "loss": 4.3875,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4090493749554465,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019100335796072088,
+      "loss": 4.4059,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4091530630099346,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019099885462784723,
+      "loss": 4.4487,
+      "step": 3946
+    },
+    {
+      "epoch": 0.40925675106442266,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019099435022128328,
+      "loss": 4.3937,
+      "step": 3947
+    },
+    {
+      "epoch": 0.40936043911891073,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001909898447410822,
+      "loss": 4.4254,
+      "step": 3948
+    },
+    {
+      "epoch": 0.4094641271733988,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001909853381872971,
+      "loss": 4.4152,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4095678152278869,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001909808305599812,
+      "loss": 4.4128,
+      "step": 3950
+    },
+    {
+      "epoch": 0.40967150328237495,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019097632185918767,
+      "loss": 4.3842,
+      "step": 3951
+    },
+    {
+      "epoch": 0.409775191336863,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001909718120849697,
+      "loss": 4.4393,
+      "step": 3952
+    },
+    {
+      "epoch": 0.40987887939135115,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019096730123738053,
+      "loss": 4.4265,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4099825674458392,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019096278931647333,
+      "loss": 4.4435,
+      "step": 3954
+    },
+    {
+      "epoch": 0.4100862555003273,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001909582763223014,
+      "loss": 4.3978,
+      "step": 3955
+    },
+    {
+      "epoch": 0.41018994355481536,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019095376225491793,
+      "loss": 4.4641,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41029363160930343,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019094924711437623,
+      "loss": 4.4192,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4103973196637915,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019094473090072955,
+      "loss": 4.4066,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4105010077182796,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019094021361403115,
+      "loss": 4.4015,
+      "step": 3959
+    },
+    {
+      "epoch": 0.41060469577276765,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019093569525433437,
+      "loss": 4.4408,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4107083838272557,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00019093117582169252,
+      "loss": 4.4277,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4108120718817438,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019092665531615893,
+      "loss": 4.3769,
+      "step": 3962
+    },
+    {
+      "epoch": 0.41091575993623186,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001909221337377869,
+      "loss": 4.4123,
+      "step": 3963
+    },
+    {
+      "epoch": 0.41101944799071993,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019091761108662979,
+      "loss": 4.4338,
+      "step": 3964
+    },
+    {
+      "epoch": 0.411123136045208,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019091308736274098,
+      "loss": 4.4592,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4112268240996961,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019090856256617384,
+      "loss": 4.449,
+      "step": 3966
+    },
+    {
+      "epoch": 0.41133051215418415,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019090403669698178,
+      "loss": 4.4338,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4114342002086722,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019089950975521814,
+      "loss": 4.4226,
+      "step": 3968
+    },
+    {
+      "epoch": 0.4115378882631603,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001908949817409364,
+      "loss": 4.3943,
+      "step": 3969
+    },
+    {
+      "epoch": 0.41164157631764836,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001908904526541899,
+      "loss": 4.3946,
+      "step": 3970
+    },
+    {
+      "epoch": 0.41174526437213643,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001908859224950322,
+      "loss": 4.4524,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4118489524266245,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019088139126351665,
+      "loss": 4.4208,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4119526404811126,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019087685895969675,
+      "loss": 4.3971,
+      "step": 3973
+    },
+    {
+      "epoch": 0.41205632853560065,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.000190872325583626,
+      "loss": 4.389,
+      "step": 3974
+    },
+    {
+      "epoch": 0.4121600165900887,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019086779113535785,
+      "loss": 4.3749,
+      "step": 3975
+    },
+    {
+      "epoch": 0.4122637046445768,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00019086325561494581,
+      "loss": 4.4153,
+      "step": 3976
+    },
+    {
+      "epoch": 0.41236739269906486,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019085871902244344,
+      "loss": 4.4229,
+      "step": 3977
+    },
+    {
+      "epoch": 0.41247108075355293,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019085418135790416,
+      "loss": 4.4027,
+      "step": 3978
+    },
+    {
+      "epoch": 0.412574768808041,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019084964262138163,
+      "loss": 4.387,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4126784568625291,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019084510281292932,
+      "loss": 4.4023,
+      "step": 3980
+    },
+    {
+      "epoch": 0.41278214491701715,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019084056193260086,
+      "loss": 4.4215,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4128858329715052,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019083601998044975,
+      "loss": 4.4307,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4129895210259933,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00019083147695652964,
+      "loss": 4.4602,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41309320908048136,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001908269328608941,
+      "loss": 4.4425,
+      "step": 3984
+    },
+    {
+      "epoch": 0.41319689713496943,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019082238769359676,
+      "loss": 4.3991,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4133005851894575,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019081784145469127,
+      "loss": 4.3821,
+      "step": 3986
+    },
+    {
+      "epoch": 0.4134042732439456,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019081329414423123,
+      "loss": 4.4002,
+      "step": 3987
+    },
+    {
+      "epoch": 0.41350796129843365,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019080874576227034,
+      "loss": 4.3869,
+      "step": 3988
+    },
+    {
+      "epoch": 0.4136116493529217,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019080419630886222,
+      "loss": 4.4355,
+      "step": 3989
+    },
+    {
+      "epoch": 0.4137153374074098,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019079964578406057,
+      "loss": 4.4083,
+      "step": 3990
+    },
+    {
+      "epoch": 0.41381902546189786,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019079509418791908,
+      "loss": 4.4459,
+      "step": 3991
+    },
+    {
+      "epoch": 0.41392271351638593,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019079054152049146,
+      "loss": 4.4067,
+      "step": 3992
+    },
+    {
+      "epoch": 0.414026401570874,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019078598778183138,
+      "loss": 4.4076,
+      "step": 3993
+    },
+    {
+      "epoch": 0.4141300896253621,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019078143297199268,
+      "loss": 4.4011,
+      "step": 3994
+    },
+    {
+      "epoch": 0.41423377767985015,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.000190776877091029,
+      "loss": 4.3641,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4143374657343382,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019077232013899409,
+      "loss": 4.3824,
+      "step": 3996
+    },
+    {
+      "epoch": 0.4144411537888263,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019076776211594178,
+      "loss": 4.3664,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4145448418433144,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019076320302192584,
+      "loss": 4.442,
+      "step": 3998
+    },
+    {
+      "epoch": 0.4146485298978025,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019075864285700002,
+      "loss": 4.4118,
+      "step": 3999
+    },
+    {
+      "epoch": 0.41475221795229056,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019075408162121815,
+      "loss": 4.4445,
+      "step": 4000
+    },
+    {
+      "epoch": 0.41485590600677863,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019074951931463406,
+      "loss": 4.375,
+      "step": 4001
+    },
+    {
+      "epoch": 0.4149595940612667,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019074495593730158,
+      "loss": 4.4621,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4150632821157548,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019074039148927452,
+      "loss": 4.354,
+      "step": 4003
+    },
+    {
+      "epoch": 0.41516697017024284,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019073582597060678,
+      "loss": 4.3905,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4152706582247309,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019073125938135217,
+      "loss": 4.4419,
+      "step": 4005
+    },
+    {
+      "epoch": 0.415374346279219,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001907266917215646,
+      "loss": 4.4289,
+      "step": 4006
+    },
+    {
+      "epoch": 0.41547803433370706,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019072212299129802,
+      "loss": 4.4418,
+      "step": 4007
+    },
+    {
+      "epoch": 0.41558172238819513,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019071755319060628,
+      "loss": 4.3701,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4156854104426832,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019071298231954325,
+      "loss": 4.3819,
+      "step": 4009
+    },
+    {
+      "epoch": 0.41578909849717127,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019070841037816297,
+      "loss": 4.4376,
+      "step": 4010
+    },
+    {
+      "epoch": 0.41589278655165934,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001907038373665193,
+      "loss": 4.4225,
+      "step": 4011
+    },
+    {
+      "epoch": 0.4159964746061474,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019069926328466624,
+      "loss": 4.394,
+      "step": 4012
+    },
+    {
+      "epoch": 0.4161001626606355,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019069468813265775,
+      "loss": 4.4233,
+      "step": 4013
+    },
+    {
+      "epoch": 0.41620385071512356,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019069011191054784,
+      "loss": 4.3972,
+      "step": 4014
+    },
+    {
+      "epoch": 0.41630753876961163,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019068553461839042,
+      "loss": 4.4305,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4164112268240997,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001906809562562396,
+      "loss": 4.41,
+      "step": 4016
+    },
+    {
+      "epoch": 0.41651491487858777,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001906763768241493,
+      "loss": 4.4322,
+      "step": 4017
+    },
+    {
+      "epoch": 0.41661860293307584,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019067179632217363,
+      "loss": 4.4129,
+      "step": 4018
+    },
+    {
+      "epoch": 0.4167222909875639,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019066721475036657,
+      "loss": 4.4214,
+      "step": 4019
+    },
+    {
+      "epoch": 0.416825979042052,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019066263210878225,
+      "loss": 4.4172,
+      "step": 4020
+    },
+    {
+      "epoch": 0.41692966709654006,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001906580483974747,
+      "loss": 4.3925,
+      "step": 4021
+    },
+    {
+      "epoch": 0.41703335515102813,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000190653463616498,
+      "loss": 4.4388,
+      "step": 4022
+    },
+    {
+      "epoch": 0.4171370432055162,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001906488777659063,
+      "loss": 4.3655,
+      "step": 4023
+    },
+    {
+      "epoch": 0.41724073126000427,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001906442908457536,
+      "loss": 4.3968,
+      "step": 4024
+    },
+    {
+      "epoch": 0.41734441931449234,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019063970285609412,
+      "loss": 4.3332,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4174481073689804,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019063511379698196,
+      "loss": 4.4032,
+      "step": 4026
+    },
+    {
+      "epoch": 0.4175517954234685,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019063052366847125,
+      "loss": 4.4094,
+      "step": 4027
+    },
+    {
+      "epoch": 0.41765548347795656,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019062593247061614,
+      "loss": 4.4239,
+      "step": 4028
+    },
+    {
+      "epoch": 0.41775917153244463,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019062134020347084,
+      "loss": 4.4453,
+      "step": 4029
+    },
+    {
+      "epoch": 0.4178628595869327,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019061674686708953,
+      "loss": 4.3898,
+      "step": 4030
+    },
+    {
+      "epoch": 0.41796654764142077,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019061215246152637,
+      "loss": 4.3921,
+      "step": 4031
+    },
+    {
+      "epoch": 0.41807023569590884,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019060755698683561,
+      "loss": 4.4481,
+      "step": 4032
+    },
+    {
+      "epoch": 0.4181739237503969,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019060296044307146,
+      "loss": 4.4255,
+      "step": 4033
+    },
+    {
+      "epoch": 0.418277611804885,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019059836283028813,
+      "loss": 4.4163,
+      "step": 4034
+    },
+    {
+      "epoch": 0.41838129985937306,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019059376414853988,
+      "loss": 4.4293,
+      "step": 4035
+    },
+    {
+      "epoch": 0.41848498791386113,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019058916439788098,
+      "loss": 4.431,
+      "step": 4036
+    },
+    {
+      "epoch": 0.4185886759683492,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019058456357836571,
+      "loss": 4.3368,
+      "step": 4037
+    },
+    {
+      "epoch": 0.41869236402283727,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019057996169004833,
+      "loss": 4.4211,
+      "step": 4038
+    },
+    {
+      "epoch": 0.41879605207732534,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001905753587329832,
+      "loss": 4.4138,
+      "step": 4039
+    },
+    {
+      "epoch": 0.4188997401318134,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001905707547072245,
+      "loss": 4.373,
+      "step": 4040
+    },
+    {
+      "epoch": 0.4190034281863015,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019056614961282666,
+      "loss": 4.3839,
+      "step": 4041
+    },
+    {
+      "epoch": 0.4191071162407896,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000190561543449844,
+      "loss": 4.4151,
+      "step": 4042
+    },
+    {
+      "epoch": 0.4192108042952777,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019055693621833087,
+      "loss": 4.3428,
+      "step": 4043
+    },
+    {
+      "epoch": 0.41931449234976576,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019055232791834157,
+      "loss": 4.3879,
+      "step": 4044
+    },
+    {
+      "epoch": 0.4194181804042538,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019054771854993055,
+      "loss": 4.3823,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4195218684587419,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019054310811315213,
+      "loss": 4.411,
+      "step": 4046
+    },
+    {
+      "epoch": 0.41962555651322997,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019053849660806077,
+      "loss": 4.4158,
+      "step": 4047
+    },
+    {
+      "epoch": 0.41972924456771804,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019053388403471086,
+      "loss": 4.4367,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4198329326222061,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001905292703931568,
+      "loss": 4.4272,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4199366206766942,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019052465568345306,
+      "loss": 4.3869,
+      "step": 4050
+    },
+    {
+      "epoch": 0.42004030873118225,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019052003990565407,
+      "loss": 4.3906,
+      "step": 4051
+    },
+    {
+      "epoch": 0.4201439967856703,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001905154230598143,
+      "loss": 4.4166,
+      "step": 4052
+    },
+    {
+      "epoch": 0.4202476848401584,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001905108051459882,
+      "loss": 4.423,
+      "step": 4053
+    },
+    {
+      "epoch": 0.42035137289464647,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019050618616423025,
+      "loss": 4.4247,
+      "step": 4054
+    },
+    {
+      "epoch": 0.42045506094913454,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019050156611459502,
+      "loss": 4.4092,
+      "step": 4055
+    },
+    {
+      "epoch": 0.4205587490036226,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019049694499713695,
+      "loss": 4.4116,
+      "step": 4056
+    },
+    {
+      "epoch": 0.4206624370581107,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001904923228119106,
+      "loss": 4.4161,
+      "step": 4057
+    },
+    {
+      "epoch": 0.42076612511259875,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019048769955897047,
+      "loss": 4.4182,
+      "step": 4058
+    },
+    {
+      "epoch": 0.4208698131670868,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019048307523837114,
+      "loss": 4.3953,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4209735012215749,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019047844985016718,
+      "loss": 4.455,
+      "step": 4060
+    },
+    {
+      "epoch": 0.42107718927606297,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019047382339441315,
+      "loss": 4.4332,
+      "step": 4061
+    },
+    {
+      "epoch": 0.42118087733055104,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019046919587116366,
+      "loss": 4.456,
+      "step": 4062
+    },
+    {
+      "epoch": 0.4212845653850391,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019046456728047327,
+      "loss": 4.421,
+      "step": 4063
+    },
+    {
+      "epoch": 0.4213882534395272,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001904599376223966,
+      "loss": 4.393,
+      "step": 4064
+    },
+    {
+      "epoch": 0.42149194149401525,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019045530689698828,
+      "loss": 4.391,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4215956295485033,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019045067510430297,
+      "loss": 4.3836,
+      "step": 4066
+    },
+    {
+      "epoch": 0.4216993176029914,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001904460422443953,
+      "loss": 4.4405,
+      "step": 4067
+    },
+    {
+      "epoch": 0.42180300565747947,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019044140831731997,
+      "loss": 4.418,
+      "step": 4068
+    },
+    {
+      "epoch": 0.42190669371196754,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001904367733231316,
+      "loss": 4.4034,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4220103817664556,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019043213726188486,
+      "loss": 4.3926,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4221140698209437,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019042750013363452,
+      "loss": 4.4189,
+      "step": 4071
+    },
+    {
+      "epoch": 0.42221775787543175,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019042286193843527,
+      "loss": 4.4032,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4223214459299198,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019041822267634184,
+      "loss": 4.4009,
+      "step": 4073
+    },
+    {
+      "epoch": 0.4224251339844079,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019041358234740896,
+      "loss": 4.4393,
+      "step": 4074
+    },
+    {
+      "epoch": 0.42252882203889597,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019040894095169138,
+      "loss": 4.3988,
+      "step": 4075
+    },
+    {
+      "epoch": 0.42263251009338404,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019040429848924385,
+      "loss": 4.4107,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4227361981478721,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019039965496012116,
+      "loss": 4.4161,
+      "step": 4077
+    },
+    {
+      "epoch": 0.4228398862023602,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019039501036437808,
+      "loss": 4.3999,
+      "step": 4078
+    },
+    {
+      "epoch": 0.42294357425684825,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019039036470206946,
+      "loss": 4.4145,
+      "step": 4079
+    },
+    {
+      "epoch": 0.4230472623113363,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019038571797325005,
+      "loss": 4.4175,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4231509503658244,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019038107017797472,
+      "loss": 4.3827,
+      "step": 4081
+    },
+    {
+      "epoch": 0.42325463842031247,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001903764213162983,
+      "loss": 4.3996,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42335832647480054,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019037177138827568,
+      "loss": 4.4007,
+      "step": 4083
+    },
+    {
+      "epoch": 0.4234620145292886,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019036712039396166,
+      "loss": 4.4087,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4235657025837767,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019036246833341113,
+      "loss": 4.4105,
+      "step": 4085
+    },
+    {
+      "epoch": 0.42366939063826475,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019035781520667898,
+      "loss": 4.4509,
+      "step": 4086
+    },
+    {
+      "epoch": 0.4237730786927529,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019035316101382015,
+      "loss": 4.3775,
+      "step": 4087
+    },
+    {
+      "epoch": 0.42387676674724095,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001903485057548895,
+      "loss": 4.4293,
+      "step": 4088
+    },
+    {
+      "epoch": 0.423980454801729,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000190343849429942,
+      "loss": 4.4232,
+      "step": 4089
+    },
+    {
+      "epoch": 0.4240841428562171,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019033919203903258,
+      "loss": 4.3965,
+      "step": 4090
+    },
+    {
+      "epoch": 0.42418783091070517,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001903345335822162,
+      "loss": 4.4255,
+      "step": 4091
+    },
+    {
+      "epoch": 0.42429151896519324,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001903298740595478,
+      "loss": 4.4579,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4243952070196813,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019032521347108237,
+      "loss": 4.4192,
+      "step": 4093
+    },
+    {
+      "epoch": 0.4244988950741694,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019032055181687486,
+      "loss": 4.4378,
+      "step": 4094
+    },
+    {
+      "epoch": 0.42460258312865745,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019031588909698037,
+      "loss": 4.4109,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4247062711831455,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019031122531145382,
+      "loss": 4.3871,
+      "step": 4096
+    },
+    {
+      "epoch": 0.4248099592376336,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019030656046035028,
+      "loss": 4.3683,
+      "step": 4097
+    },
+    {
+      "epoch": 0.42491364729212167,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001903018945437248,
+      "loss": 4.3856,
+      "step": 4098
+    },
+    {
+      "epoch": 0.42501733534660974,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019029722756163243,
+      "loss": 4.3411,
+      "step": 4099
+    },
+    {
+      "epoch": 0.4251210234010978,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001902925595141282,
+      "loss": 4.3784,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4252247114555859,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019028789040126718,
+      "loss": 4.4135,
+      "step": 4101
+    },
+    {
+      "epoch": 0.42532839951007395,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019028322022310453,
+      "loss": 4.443,
+      "step": 4102
+    },
+    {
+      "epoch": 0.425432087564562,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001902785489796953,
+      "loss": 4.3813,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4255357756190501,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019027387667109463,
+      "loss": 4.4411,
+      "step": 4104
+    },
+    {
+      "epoch": 0.42563946367353817,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019026920329735764,
+      "loss": 4.4123,
+      "step": 4105
+    },
+    {
+      "epoch": 0.42574315172802624,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019026452885853947,
+      "loss": 4.4107,
+      "step": 4106
+    },
+    {
+      "epoch": 0.4258468397825143,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019025985335469528,
+      "loss": 4.4329,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4259505278370024,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019025517678588022,
+      "loss": 4.4004,
+      "step": 4108
+    },
+    {
+      "epoch": 0.42605421589149045,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019025049915214948,
+      "loss": 4.356,
+      "step": 4109
+    },
+    {
+      "epoch": 0.4261579039459785,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019024582045355825,
+      "loss": 4.3517,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4262615920004666,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019024114069016173,
+      "loss": 4.4328,
+      "step": 4111
+    },
+    {
+      "epoch": 0.42636528005495467,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019023645986201513,
+      "loss": 4.3555,
+      "step": 4112
+    },
+    {
+      "epoch": 0.42646896810944274,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001902317779691737,
+      "loss": 4.4117,
+      "step": 4113
+    },
+    {
+      "epoch": 0.4265726561639308,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019022709501169267,
+      "loss": 4.4447,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4266763442184189,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019022241098962727,
+      "loss": 4.4215,
+      "step": 4115
+    },
+    {
+      "epoch": 0.42678003227290695,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001902177259030328,
+      "loss": 4.4491,
+      "step": 4116
+    },
+    {
+      "epoch": 0.426883720327395,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019021303975196453,
+      "loss": 4.4125,
+      "step": 4117
+    },
+    {
+      "epoch": 0.4269874083818831,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019020835253647777,
+      "loss": 4.433,
+      "step": 4118
+    },
+    {
+      "epoch": 0.42709109643637116,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001902036642566278,
+      "loss": 4.3897,
+      "step": 4119
+    },
+    {
+      "epoch": 0.42719478449085924,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001901989749124699,
+      "loss": 4.4053,
+      "step": 4120
+    },
+    {
+      "epoch": 0.4272984725453473,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001901942845040595,
+      "loss": 4.4107,
+      "step": 4121
+    },
+    {
+      "epoch": 0.4274021605998354,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019018959303145186,
+      "loss": 4.4075,
+      "step": 4122
+    },
+    {
+      "epoch": 0.42750584865432345,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019018490049470236,
+      "loss": 4.3824,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4276095367088115,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019018020689386632,
+      "loss": 4.4052,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4277132247632996,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001901755122289992,
+      "loss": 4.3915,
+      "step": 4125
+    },
+    {
+      "epoch": 0.42781691281778766,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001901708165001564,
+      "loss": 4.4056,
+      "step": 4126
+    },
+    {
+      "epoch": 0.42792060087227574,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001901661197073932,
+      "loss": 4.424,
+      "step": 4127
+    },
+    {
+      "epoch": 0.4280242889267638,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019016142185076512,
+      "loss": 4.3931,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4281279769812519,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001901567229303276,
+      "loss": 4.4321,
+      "step": 4129
+    },
+    {
+      "epoch": 0.42823166503573995,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019015202294613603,
+      "loss": 4.4057,
+      "step": 4130
+    },
+    {
+      "epoch": 0.428335353090228,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019014732189824587,
+      "loss": 4.3677,
+      "step": 4131
+    },
+    {
+      "epoch": 0.42843904114471615,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019014261978671262,
+      "loss": 4.4088,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4285427291992042,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019013791661159172,
+      "loss": 4.4014,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4286464172536923,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001901332123729387,
+      "loss": 4.4197,
+      "step": 4134
+    },
+    {
+      "epoch": 0.42875010530818036,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019012850707080905,
+      "loss": 4.4187,
+      "step": 4135
+    },
+    {
+      "epoch": 0.42885379336266843,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019012380070525828,
+      "loss": 4.4087,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4289574814171565,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019011909327634193,
+      "loss": 4.4242,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4290611694716446,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019011438478411553,
+      "loss": 4.4295,
+      "step": 4138
+    },
+    {
+      "epoch": 0.42916485752613265,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019010967522863466,
+      "loss": 4.3558,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4292685455806207,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019010496460995487,
+      "loss": 4.3369,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4293722336351088,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019010025292813174,
+      "loss": 4.3923,
+      "step": 4141
+    },
+    {
+      "epoch": 0.42947592168959686,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019009554018322087,
+      "loss": 4.4044,
+      "step": 4142
+    },
+    {
+      "epoch": 0.42957960974408493,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019009082637527785,
+      "loss": 4.3502,
+      "step": 4143
+    },
+    {
+      "epoch": 0.429683297798573,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001900861115043583,
+      "loss": 4.3683,
+      "step": 4144
+    },
+    {
+      "epoch": 0.4297869858530611,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001900813955705179,
+      "loss": 4.3883,
+      "step": 4145
+    },
+    {
+      "epoch": 0.42989067390754915,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019007667857381223,
+      "loss": 4.4058,
+      "step": 4146
+    },
+    {
+      "epoch": 0.4299943619620372,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019007196051429694,
+      "loss": 4.4143,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4300980500165253,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019006724139202774,
+      "loss": 4.4171,
+      "step": 4148
+    },
+    {
+      "epoch": 0.43020173807101336,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019006252120706032,
+      "loss": 4.4089,
+      "step": 4149
+    },
+    {
+      "epoch": 0.43030542612550143,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019005779995945034,
+      "loss": 4.4398,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4304091141799895,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019005307764925351,
+      "loss": 4.382,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4305128022344776,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019004835427652555,
+      "loss": 4.3662,
+      "step": 4152
+    },
+    {
+      "epoch": 0.43061649028896565,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001900436298413222,
+      "loss": 4.3927,
+      "step": 4153
+    },
+    {
+      "epoch": 0.4307201783434537,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001900389043436992,
+      "loss": 4.4008,
+      "step": 4154
+    },
+    {
+      "epoch": 0.4308238663979418,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019003417778371232,
+      "loss": 4.4127,
+      "step": 4155
+    },
+    {
+      "epoch": 0.43092755445242986,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019002945016141726,
+      "loss": 4.4235,
+      "step": 4156
+    },
+    {
+      "epoch": 0.43103124250691793,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019002472147686993,
+      "loss": 4.3804,
+      "step": 4157
+    },
+    {
+      "epoch": 0.431134930561406,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019001999173012596,
+      "loss": 4.371,
+      "step": 4158
+    },
+    {
+      "epoch": 0.4312386186158941,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019001526092124132,
+      "loss": 4.3823,
+      "step": 4159
+    },
+    {
+      "epoch": 0.43134230667038215,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001900105290502717,
+      "loss": 4.4178,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4314459947248702,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019000579611727302,
+      "loss": 4.3976,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4315496827793583,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019000106212230106,
+      "loss": 4.3859,
+      "step": 4162
+    },
+    {
+      "epoch": 0.43165337083384636,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001899963270654117,
+      "loss": 4.4141,
+      "step": 4163
+    },
+    {
+      "epoch": 0.43175705888833443,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018999159094666083,
+      "loss": 4.4009,
+      "step": 4164
+    },
+    {
+      "epoch": 0.4318607469428225,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001899868537661043,
+      "loss": 4.3679,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4319644349973106,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018998211552379802,
+      "loss": 4.4297,
+      "step": 4166
+    },
+    {
+      "epoch": 0.43206812305179865,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001899773762197979,
+      "loss": 4.4041,
+      "step": 4167
+    },
+    {
+      "epoch": 0.4321718111062867,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018997263585415986,
+      "loss": 4.3974,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4322754991607748,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018996789442693977,
+      "loss": 4.4237,
+      "step": 4169
+    },
+    {
+      "epoch": 0.43237918721526286,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018996315193819369,
+      "loss": 4.3515,
+      "step": 4170
+    },
+    {
+      "epoch": 0.43248287526975093,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018995840838797746,
+      "loss": 4.3885,
+      "step": 4171
+    },
+    {
+      "epoch": 0.432586563324239,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018995366377634712,
+      "loss": 4.3844,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4326902513787271,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018994891810335864,
+      "loss": 4.4171,
+      "step": 4173
+    },
+    {
+      "epoch": 0.43279393943321515,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018994417136906802,
+      "loss": 4.4068,
+      "step": 4174
+    },
+    {
+      "epoch": 0.4328976274877032,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018993942357353123,
+      "loss": 4.3834,
+      "step": 4175
+    },
+    {
+      "epoch": 0.43300131554219135,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001899346747168043,
+      "loss": 4.3841,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4331050035966794,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018992992479894332,
+      "loss": 4.3845,
+      "step": 4177
+    },
+    {
+      "epoch": 0.4332086916511675,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018992517382000426,
+      "loss": 4.4253,
+      "step": 4178
+    },
+    {
+      "epoch": 0.43331237970565556,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018992042178004315,
+      "loss": 4.3755,
+      "step": 4179
+    },
+    {
+      "epoch": 0.43341606776014363,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018991566867911617,
+      "loss": 4.4059,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4335197558146317,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018991091451727937,
+      "loss": 4.4073,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4336234438691198,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001899061592945888,
+      "loss": 4.3329,
+      "step": 4182
+    },
+    {
+      "epoch": 0.43372713192360784,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018990140301110056,
+      "loss": 4.3915,
+      "step": 4183
+    },
+    {
+      "epoch": 0.4338308199780959,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001898966456668708,
+      "loss": 4.4172,
+      "step": 4184
+    },
+    {
+      "epoch": 0.433934508032584,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018989188726195567,
+      "loss": 4.4165,
+      "step": 4185
+    },
+    {
+      "epoch": 0.43403819608707206,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018988712779641128,
+      "loss": 4.402,
+      "step": 4186
+    },
+    {
+      "epoch": 0.43414188414156013,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001898823672702938,
+      "loss": 4.4,
+      "step": 4187
+    },
+    {
+      "epoch": 0.4342455721960482,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018987760568365937,
+      "loss": 4.4032,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4343492602505363,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001898728430365642,
+      "loss": 4.4163,
+      "step": 4189
+    },
+    {
+      "epoch": 0.43445294830502434,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001898680793290645,
+      "loss": 4.3809,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4345566363595124,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018986331456121648,
+      "loss": 4.3944,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4346603244140005,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018985854873307628,
+      "loss": 4.3181,
+      "step": 4192
+    },
+    {
+      "epoch": 0.43476401246848856,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018985378184470023,
+      "loss": 4.4519,
+      "step": 4193
+    },
+    {
+      "epoch": 0.43486770052297663,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018984901389614453,
+      "loss": 4.3376,
+      "step": 4194
+    },
+    {
+      "epoch": 0.4349713885774647,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018984424488746544,
+      "loss": 4.3662,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4350750766319528,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018983947481871922,
+      "loss": 4.4071,
+      "step": 4196
+    },
+    {
+      "epoch": 0.43517876468644084,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001898347036899622,
+      "loss": 4.4058,
+      "step": 4197
+    },
+    {
+      "epoch": 0.4352824527409289,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018982993150125057,
+      "loss": 4.4075,
+      "step": 4198
+    },
+    {
+      "epoch": 0.435386140795417,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018982515825264072,
+      "loss": 4.4165,
+      "step": 4199
+    },
+    {
+      "epoch": 0.43548982884990506,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000189820383944189,
+      "loss": 4.4017,
+      "step": 4200
+    },
+    {
+      "epoch": 0.43559351690439313,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018981560857595167,
+      "loss": 4.4161,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4356972049588812,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018981083214798507,
+      "loss": 4.4522,
+      "step": 4202
+    },
+    {
+      "epoch": 0.4358008930133693,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001898060546603456,
+      "loss": 4.4195,
+      "step": 4203
+    },
+    {
+      "epoch": 0.43590458106785734,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001898012761130896,
+      "loss": 4.4096,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4360082691223454,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001897964965062735,
+      "loss": 4.3819,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4361119571768335,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018979171583995366,
+      "loss": 4.3974,
+      "step": 4206
+    },
+    {
+      "epoch": 0.43621564523132156,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018978693411418648,
+      "loss": 4.4042,
+      "step": 4207
+    },
+    {
+      "epoch": 0.43631933328580963,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018978215132902838,
+      "loss": 4.3937,
+      "step": 4208
+    },
+    {
+      "epoch": 0.4364230213402977,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018977736748453577,
+      "loss": 4.3937,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4365267093947858,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018977258258076515,
+      "loss": 4.4096,
+      "step": 4210
+    },
+    {
+      "epoch": 0.43663039744927384,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018976779661777297,
+      "loss": 4.3821,
+      "step": 4211
+    },
+    {
+      "epoch": 0.4367340855037619,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018976300959561564,
+      "loss": 4.3688,
+      "step": 4212
+    },
+    {
+      "epoch": 0.43683777355825,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018975822151434972,
+      "loss": 4.3522,
+      "step": 4213
+    },
+    {
+      "epoch": 0.43694146161273806,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018975343237403162,
+      "loss": 4.3861,
+      "step": 4214
+    },
+    {
+      "epoch": 0.43704514966722613,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018974864217471792,
+      "loss": 4.3924,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4371488377217142,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001897438509164651,
+      "loss": 4.3727,
+      "step": 4216
+    },
+    {
+      "epoch": 0.43725252577620227,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018973905859932973,
+      "loss": 4.4095,
+      "step": 4217
+    },
+    {
+      "epoch": 0.43735621383069034,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018973426522336827,
+      "loss": 4.4024,
+      "step": 4218
+    },
+    {
+      "epoch": 0.4374599018851784,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018972947078863738,
+      "loss": 4.399,
+      "step": 4219
+    },
+    {
+      "epoch": 0.4375635899396665,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018972467529519357,
+      "loss": 4.3364,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4376672779941546,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018971987874309343,
+      "loss": 4.4305,
+      "step": 4221
+    },
+    {
+      "epoch": 0.4377709660486427,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018971508113239358,
+      "loss": 4.3965,
+      "step": 4222
+    },
+    {
+      "epoch": 0.43787465410313076,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018971028246315058,
+      "loss": 4.435,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4379783421576188,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018970548273542108,
+      "loss": 4.3942,
+      "step": 4224
+    },
+    {
+      "epoch": 0.4380820302121069,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001897006819492617,
+      "loss": 4.3815,
+      "step": 4225
+    },
+    {
+      "epoch": 0.43818571826659497,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018969588010472915,
+      "loss": 4.4046,
+      "step": 4226
+    },
+    {
+      "epoch": 0.43828940632108304,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018969107720187995,
+      "loss": 4.4266,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4383930943755711,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018968627324077088,
+      "loss": 4.3653,
+      "step": 4228
+    },
+    {
+      "epoch": 0.4384967824300592,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018968146822145857,
+      "loss": 4.4342,
+      "step": 4229
+    },
+    {
+      "epoch": 0.43860047048454726,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018967666214399976,
+      "loss": 4.4118,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4387041585390353,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018967185500845112,
+      "loss": 4.3927,
+      "step": 4231
+    },
+    {
+      "epoch": 0.4388078465935234,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018966704681486936,
+      "loss": 4.3668,
+      "step": 4232
+    },
+    {
+      "epoch": 0.43891153464801147,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018966223756331125,
+      "loss": 4.3601,
+      "step": 4233
+    },
+    {
+      "epoch": 0.43901522270249954,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001896574272538335,
+      "loss": 4.4051,
+      "step": 4234
+    },
+    {
+      "epoch": 0.4391189107569876,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001896526158864929,
+      "loss": 4.3614,
+      "step": 4235
+    },
+    {
+      "epoch": 0.4392225988114757,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018964780346134618,
+      "loss": 4.414,
+      "step": 4236
+    },
+    {
+      "epoch": 0.43932628686596376,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018964298997845017,
+      "loss": 4.4161,
+      "step": 4237
+    },
+    {
+      "epoch": 0.4394299749204518,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001896381754378616,
+      "loss": 4.4016,
+      "step": 4238
+    },
+    {
+      "epoch": 0.4395336629749399,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001896333598396373,
+      "loss": 4.3883,
+      "step": 4239
+    },
+    {
+      "epoch": 0.43963735102942797,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018962854318383413,
+      "loss": 4.4013,
+      "step": 4240
+    },
+    {
+      "epoch": 0.43974103908391604,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018962372547050887,
+      "loss": 4.4009,
+      "step": 4241
+    },
+    {
+      "epoch": 0.4398447271384041,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001896189066997184,
+      "loss": 4.4045,
+      "step": 4242
+    },
+    {
+      "epoch": 0.4399484151928922,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018961408687151959,
+      "loss": 4.3849,
+      "step": 4243
+    },
+    {
+      "epoch": 0.44005210324738026,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018960926598596926,
+      "loss": 4.3956,
+      "step": 4244
+    },
+    {
+      "epoch": 0.4401557913018683,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001896044440431243,
+      "loss": 4.4034,
+      "step": 4245
+    },
+    {
+      "epoch": 0.4402594793563564,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018959962104304163,
+      "loss": 4.3885,
+      "step": 4246
+    },
+    {
+      "epoch": 0.44036316741084447,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018959479698577814,
+      "loss": 4.3715,
+      "step": 4247
+    },
+    {
+      "epoch": 0.44046685546533254,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018958997187139073,
+      "loss": 4.4171,
+      "step": 4248
+    },
+    {
+      "epoch": 0.4405705435198206,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018958514569993636,
+      "loss": 4.3923,
+      "step": 4249
+    },
+    {
+      "epoch": 0.4406742315743087,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000189580318471472,
+      "loss": 4.3919,
+      "step": 4250
+    },
+    {
+      "epoch": 0.44077791962879675,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018957549018605455,
+      "loss": 4.3884,
+      "step": 4251
+    },
+    {
+      "epoch": 0.4408816076832848,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000189570660843741,
+      "loss": 4.3758,
+      "step": 4252
+    },
+    {
+      "epoch": 0.4409852957377729,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018956583044458833,
+      "loss": 4.3725,
+      "step": 4253
+    },
+    {
+      "epoch": 0.44108898379226097,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018956099898865354,
+      "loss": 4.4041,
+      "step": 4254
+    },
+    {
+      "epoch": 0.44119267184674904,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018955616647599365,
+      "loss": 4.4445,
+      "step": 4255
+    },
+    {
+      "epoch": 0.4412963599012371,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018955133290666563,
+      "loss": 4.376,
+      "step": 4256
+    },
+    {
+      "epoch": 0.4414000479557252,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018954649828072654,
+      "loss": 4.4271,
+      "step": 4257
+    },
+    {
+      "epoch": 0.44150373601021325,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018954166259823346,
+      "loss": 4.3593,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4416074240647013,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018953682585924337,
+      "loss": 4.4018,
+      "step": 4259
+    },
+    {
+      "epoch": 0.4417111121191894,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001895319880638134,
+      "loss": 4.4092,
+      "step": 4260
+    },
+    {
+      "epoch": 0.44181480017367747,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018952714921200063,
+      "loss": 4.3735,
+      "step": 4261
+    },
+    {
+      "epoch": 0.44191848822816554,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001895223093038621,
+      "loss": 4.4423,
+      "step": 4262
+    },
+    {
+      "epoch": 0.4420221762826536,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018951746833945497,
+      "loss": 4.3934,
+      "step": 4263
+    },
+    {
+      "epoch": 0.4421258643371417,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001895126263188363,
+      "loss": 4.3609,
+      "step": 4264
+    },
+    {
+      "epoch": 0.44222955239162975,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001895077832420633,
+      "loss": 4.3961,
+      "step": 4265
+    },
+    {
+      "epoch": 0.4423332404461179,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018950293910919305,
+      "loss": 4.4182,
+      "step": 4266
+    },
+    {
+      "epoch": 0.44243692850060595,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018949809392028276,
+      "loss": 4.3712,
+      "step": 4267
+    },
+    {
+      "epoch": 0.442540616555094,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018949324767538955,
+      "loss": 4.3777,
+      "step": 4268
+    },
+    {
+      "epoch": 0.4426443046095821,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018948840037457057,
+      "loss": 4.3713,
+      "step": 4269
+    },
+    {
+      "epoch": 0.44274799266407017,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018948355201788312,
+      "loss": 4.4353,
+      "step": 4270
+    },
+    {
+      "epoch": 0.44285168071855824,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001894787026053843,
+      "loss": 4.4071,
+      "step": 4271
+    },
+    {
+      "epoch": 0.4429553687730463,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001894738521371314,
+      "loss": 4.3544,
+      "step": 4272
+    },
+    {
+      "epoch": 0.4430590568275344,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018946900061318162,
+      "loss": 4.3539,
+      "step": 4273
+    },
+    {
+      "epoch": 0.44316274488202245,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018946414803359222,
+      "loss": 4.4077,
+      "step": 4274
+    },
+    {
+      "epoch": 0.4432664329365105,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001894592943984204,
+      "loss": 4.3844,
+      "step": 4275
+    },
+    {
+      "epoch": 0.4433701209909986,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001894544397077235,
+      "loss": 4.3638,
+      "step": 4276
+    },
+    {
+      "epoch": 0.44347380904548667,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018944958396155878,
+      "loss": 4.3844,
+      "step": 4277
+    },
+    {
+      "epoch": 0.44357749709997474,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001894447271599835,
+      "loss": 4.4215,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4436811851544628,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018943986930305498,
+      "loss": 4.38,
+      "step": 4279
+    },
+    {
+      "epoch": 0.4437848732089509,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018943501039083056,
+      "loss": 4.391,
+      "step": 4280
+    },
+    {
+      "epoch": 0.44388856126343895,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018943015042336754,
+      "loss": 4.4133,
+      "step": 4281
+    },
+    {
+      "epoch": 0.443992249317927,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018942528940072329,
+      "loss": 4.3592,
+      "step": 4282
+    },
+    {
+      "epoch": 0.4440959373724151,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018942042732295514,
+      "loss": 4.3841,
+      "step": 4283
+    },
+    {
+      "epoch": 0.44419962542690317,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018941556419012047,
+      "loss": 4.362,
+      "step": 4284
+    },
+    {
+      "epoch": 0.44430331348139124,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018941070000227667,
+      "loss": 4.3791,
+      "step": 4285
+    },
+    {
+      "epoch": 0.4444070015358793,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001894058347594811,
+      "loss": 4.3727,
+      "step": 4286
+    },
+    {
+      "epoch": 0.4445106895903674,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018940096846179123,
+      "loss": 4.3808,
+      "step": 4287
+    },
+    {
+      "epoch": 0.44461437764485545,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018939610110926437,
+      "loss": 4.4176,
+      "step": 4288
+    },
+    {
+      "epoch": 0.4447180656993435,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018939123270195806,
+      "loss": 4.3873,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4448217537538316,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001893863632399297,
+      "loss": 4.4036,
+      "step": 4290
+    },
+    {
+      "epoch": 0.44492544180831967,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018938149272323672,
+      "loss": 4.4184,
+      "step": 4291
+    },
+    {
+      "epoch": 0.44502912986280774,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001893766211519366,
+      "loss": 4.3895,
+      "step": 4292
+    },
+    {
+      "epoch": 0.4451328179172958,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018937174852608682,
+      "loss": 4.4001,
+      "step": 4293
+    },
+    {
+      "epoch": 0.4452365059717839,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001893668748457449,
+      "loss": 4.4087,
+      "step": 4294
+    },
+    {
+      "epoch": 0.44534019402627195,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001893620001109683,
+      "loss": 4.3992,
+      "step": 4295
+    },
+    {
+      "epoch": 0.44544388208076,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018935712432181459,
+      "loss": 4.3913,
+      "step": 4296
+    },
+    {
+      "epoch": 0.4455475701352481,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018935224747834123,
+      "loss": 4.4231,
+      "step": 4297
+    },
+    {
+      "epoch": 0.44565125818973617,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001893473695806058,
+      "loss": 4.4124,
+      "step": 4298
+    },
+    {
+      "epoch": 0.44575494624422424,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001893424906286659,
+      "loss": 4.4123,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4458586342987123,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000189337610622579,
+      "loss": 4.4489,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4459623223532004,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018933272956240277,
+      "loss": 4.3554,
+      "step": 4301
+    },
+    {
+      "epoch": 0.44606601040768845,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018932784744819472,
+      "loss": 4.4026,
+      "step": 4302
+    },
+    {
+      "epoch": 0.4461696984621765,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018932296428001252,
+      "loss": 4.3834,
+      "step": 4303
+    },
+    {
+      "epoch": 0.4462733865166646,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018931808005791373,
+      "loss": 4.4165,
+      "step": 4304
+    },
+    {
+      "epoch": 0.44637707457115267,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018931319478195606,
+      "loss": 4.3488,
+      "step": 4305
+    },
+    {
+      "epoch": 0.44648076262564074,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018930830845219706,
+      "loss": 4.4097,
+      "step": 4306
+    },
+    {
+      "epoch": 0.4465844506801288,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018930342106869444,
+      "loss": 4.3845,
+      "step": 4307
+    },
+    {
+      "epoch": 0.4466881387346169,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018929853263150584,
+      "loss": 4.3781,
+      "step": 4308
+    },
+    {
+      "epoch": 0.44679182678910495,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018929364314068897,
+      "loss": 4.4109,
+      "step": 4309
+    },
+    {
+      "epoch": 0.4468955148435931,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018928875259630146,
+      "loss": 4.3925,
+      "step": 4310
+    },
+    {
+      "epoch": 0.44699920289808115,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018928386099840107,
+      "loss": 4.3674,
+      "step": 4311
+    },
+    {
+      "epoch": 0.4471028909525692,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018927896834704548,
+      "loss": 4.4255,
+      "step": 4312
+    },
+    {
+      "epoch": 0.4472065790070573,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018927407464229247,
+      "loss": 4.3883,
+      "step": 4313
+    },
+    {
+      "epoch": 0.44731026706154536,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018926917988419973,
+      "loss": 4.4199,
+      "step": 4314
+    },
+    {
+      "epoch": 0.44741395511603343,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018926428407282502,
+      "loss": 4.4236,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4475176431705215,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018925938720822612,
+      "loss": 4.4018,
+      "step": 4316
+    },
+    {
+      "epoch": 0.4476213312250096,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018925448929046082,
+      "loss": 4.3846,
+      "step": 4317
+    },
+    {
+      "epoch": 0.44772501927949765,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018924959031958686,
+      "loss": 4.3673,
+      "step": 4318
+    },
+    {
+      "epoch": 0.4478287073339857,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001892446902956621,
+      "loss": 4.4087,
+      "step": 4319
+    },
+    {
+      "epoch": 0.4479323953884738,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001892397892187443,
+      "loss": 4.4178,
+      "step": 4320
+    },
+    {
+      "epoch": 0.44803608344296186,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018923488708889137,
+      "loss": 4.3939,
+      "step": 4321
+    },
+    {
+      "epoch": 0.44813977149744993,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018922998390616104,
+      "loss": 4.4221,
+      "step": 4322
+    },
+    {
+      "epoch": 0.448243459551938,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018922507967061126,
+      "loss": 4.3855,
+      "step": 4323
+    },
+    {
+      "epoch": 0.4483471476064261,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018922017438229984,
+      "loss": 4.3899,
+      "step": 4324
+    },
+    {
+      "epoch": 0.44845083566091415,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018921526804128468,
+      "loss": 4.4192,
+      "step": 4325
+    },
+    {
+      "epoch": 0.4485545237154022,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018921036064762365,
+      "loss": 4.4315,
+      "step": 4326
+    },
+    {
+      "epoch": 0.4486582117698903,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018920545220137467,
+      "loss": 4.4587,
+      "step": 4327
+    },
+    {
+      "epoch": 0.44876189982437836,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018920054270259566,
+      "loss": 4.3906,
+      "step": 4328
+    },
+    {
+      "epoch": 0.44886558787886643,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018919563215134453,
+      "loss": 4.3824,
+      "step": 4329
+    },
+    {
+      "epoch": 0.4489692759333545,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001891907205476792,
+      "loss": 4.4012,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4490729639878426,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018918580789165765,
+      "loss": 4.4116,
+      "step": 4331
+    },
+    {
+      "epoch": 0.44917665204233065,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018918089418333786,
+      "loss": 4.4315,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4492803400968187,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018917597942277777,
+      "loss": 4.4087,
+      "step": 4333
+    },
+    {
+      "epoch": 0.4493840281513068,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001891710636100354,
+      "loss": 4.3834,
+      "step": 4334
+    },
+    {
+      "epoch": 0.44948771620579486,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018916614674516875,
+      "loss": 4.4191,
+      "step": 4335
+    },
+    {
+      "epoch": 0.44959140426028293,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018916122882823582,
+      "loss": 4.4078,
+      "step": 4336
+    },
+    {
+      "epoch": 0.449695092314771,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001891563098592946,
+      "loss": 4.3853,
+      "step": 4337
+    },
+    {
+      "epoch": 0.4497987803692591,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018915138983840318,
+      "loss": 4.366,
+      "step": 4338
+    },
+    {
+      "epoch": 0.44990246842374715,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018914646876561962,
+      "loss": 4.3723,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4500061564782352,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018914154664100195,
+      "loss": 4.4083,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4501098445327233,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018913662346460824,
+      "loss": 4.3813,
+      "step": 4341
+    },
+    {
+      "epoch": 0.45021353258721136,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018913169923649663,
+      "loss": 4.3809,
+      "step": 4342
+    },
+    {
+      "epoch": 0.45031722064169943,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018912677395672514,
+      "loss": 4.3945,
+      "step": 4343
+    },
+    {
+      "epoch": 0.4504209086961875,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018912184762535195,
+      "loss": 4.3716,
+      "step": 4344
+    },
+    {
+      "epoch": 0.4505245967506756,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018911692024243517,
+      "loss": 4.3868,
+      "step": 4345
+    },
+    {
+      "epoch": 0.45062828480516365,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018911199180803294,
+      "loss": 4.3969,
+      "step": 4346
+    },
+    {
+      "epoch": 0.4507319728596517,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018910706232220338,
+      "loss": 4.4198,
+      "step": 4347
+    },
+    {
+      "epoch": 0.4508356609141398,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001891021317850047,
+      "loss": 4.4407,
+      "step": 4348
+    },
+    {
+      "epoch": 0.45093934896862786,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.000189097200196495,
+      "loss": 4.3974,
+      "step": 4349
+    },
+    {
+      "epoch": 0.45104303702311593,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018909226755673257,
+      "loss": 4.3658,
+      "step": 4350
+    },
+    {
+      "epoch": 0.451146725077604,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018908733386577552,
+      "loss": 4.3892,
+      "step": 4351
+    },
+    {
+      "epoch": 0.4512504131320921,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018908239912368213,
+      "loss": 4.4091,
+      "step": 4352
+    },
+    {
+      "epoch": 0.45135410118658015,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001890774633305106,
+      "loss": 4.3467,
+      "step": 4353
+    },
+    {
+      "epoch": 0.4514577892410682,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018907252648631912,
+      "loss": 4.3949,
+      "step": 4354
+    },
+    {
+      "epoch": 0.45156147729555635,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018906758859116602,
+      "loss": 4.371,
+      "step": 4355
+    },
+    {
+      "epoch": 0.4516651653500444,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001890626496451095,
+      "loss": 4.4002,
+      "step": 4356
+    },
+    {
+      "epoch": 0.4517688534045325,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001890577096482079,
+      "loss": 4.3438,
+      "step": 4357
+    },
+    {
+      "epoch": 0.45187254145902056,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001890527686005194,
+      "loss": 4.4444,
+      "step": 4358
+    },
+    {
+      "epoch": 0.45197622951350863,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018904782650210243,
+      "loss": 4.4304,
+      "step": 4359
+    },
+    {
+      "epoch": 0.4520799175679967,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018904288335301516,
+      "loss": 4.3794,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4521836056224848,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018903793915331604,
+      "loss": 4.4026,
+      "step": 4361
+    },
+    {
+      "epoch": 0.45228729367697285,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018903299390306334,
+      "loss": 4.4174,
+      "step": 4362
+    },
+    {
+      "epoch": 0.4523909817314609,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018902804760231545,
+      "loss": 4.4262,
+      "step": 4363
+    },
+    {
+      "epoch": 0.452494669785949,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018902310025113066,
+      "loss": 4.3765,
+      "step": 4364
+    },
+    {
+      "epoch": 0.45259835784043706,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018901815184956742,
+      "loss": 4.3958,
+      "step": 4365
+    },
+    {
+      "epoch": 0.45270204589492513,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018901320239768406,
+      "loss": 4.3625,
+      "step": 4366
+    },
+    {
+      "epoch": 0.4528057339494132,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018900825189553904,
+      "loss": 4.3915,
+      "step": 4367
+    },
+    {
+      "epoch": 0.4529094220039013,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018900330034319073,
+      "loss": 4.3817,
+      "step": 4368
+    },
+    {
+      "epoch": 0.45301311005838935,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018899834774069752,
+      "loss": 4.4392,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4531167981128774,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001889933940881179,
+      "loss": 4.3706,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4532204861673655,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018898843938551028,
+      "loss": 4.4122,
+      "step": 4371
+    },
+    {
+      "epoch": 0.45332417422185356,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018898348363293317,
+      "loss": 4.3939,
+      "step": 4372
+    },
+    {
+      "epoch": 0.45342786227634163,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018897852683044498,
+      "loss": 4.3972,
+      "step": 4373
+    },
+    {
+      "epoch": 0.4535315503308297,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018897356897810426,
+      "loss": 4.4002,
+      "step": 4374
+    },
+    {
+      "epoch": 0.4536352383853178,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018896861007596948,
+      "loss": 4.3926,
+      "step": 4375
+    },
+    {
+      "epoch": 0.45373892643980585,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018896365012409914,
+      "loss": 4.4,
+      "step": 4376
+    },
+    {
+      "epoch": 0.4538426144942939,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018895868912255175,
+      "loss": 4.3701,
+      "step": 4377
+    },
+    {
+      "epoch": 0.453946302548782,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018895372707138585,
+      "loss": 4.3929,
+      "step": 4378
+    },
+    {
+      "epoch": 0.45404999060327006,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018894876397066002,
+      "loss": 4.4144,
+      "step": 4379
+    },
+    {
+      "epoch": 0.45415367865775813,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001889437998204328,
+      "loss": 4.4537,
+      "step": 4380
+    },
+    {
+      "epoch": 0.4542573667122462,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018893883462076273,
+      "loss": 4.3803,
+      "step": 4381
+    },
+    {
+      "epoch": 0.4543610547667343,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018893386837170849,
+      "loss": 4.3838,
+      "step": 4382
+    },
+    {
+      "epoch": 0.45446474282122235,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018892890107332857,
+      "loss": 4.382,
+      "step": 4383
+    },
+    {
+      "epoch": 0.4545684308757104,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001889239327256816,
+      "loss": 4.3865,
+      "step": 4384
+    },
+    {
+      "epoch": 0.4546721189301985,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018891896332882624,
+      "loss": 4.3959,
+      "step": 4385
+    },
+    {
+      "epoch": 0.45477580698468656,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001889139928828211,
+      "loss": 4.3629,
+      "step": 4386
+    },
+    {
+      "epoch": 0.45487949503917463,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018890902138772483,
+      "loss": 4.3834,
+      "step": 4387
+    },
+    {
+      "epoch": 0.4549831830936627,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018890404884359613,
+      "loss": 4.3842,
+      "step": 4388
+    },
+    {
+      "epoch": 0.4550868711481508,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018889907525049355,
+      "loss": 4.4223,
+      "step": 4389
+    },
+    {
+      "epoch": 0.45519055920263884,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001888941006084759,
+      "loss": 4.3973,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4552942472571269,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018888912491760182,
+      "loss": 4.3702,
+      "step": 4391
+    },
+    {
+      "epoch": 0.455397935311615,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018888414817793003,
+      "loss": 4.413,
+      "step": 4392
+    },
+    {
+      "epoch": 0.45550162336610306,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018887917038951926,
+      "loss": 4.3807,
+      "step": 4393
+    },
+    {
+      "epoch": 0.45560531142059113,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001888741915524282,
+      "loss": 4.4024,
+      "step": 4394
+    },
+    {
+      "epoch": 0.4557089994750792,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018886921166671563,
+      "loss": 4.3569,
+      "step": 4395
+    },
+    {
+      "epoch": 0.4558126875295673,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018886423073244033,
+      "loss": 4.3702,
+      "step": 4396
+    },
+    {
+      "epoch": 0.45591637558405534,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018885924874966102,
+      "loss": 4.3712,
+      "step": 4397
+    },
+    {
+      "epoch": 0.4560200636385434,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001888542657184365,
+      "loss": 4.4314,
+      "step": 4398
+    },
+    {
+      "epoch": 0.4561237516930315,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018884928163882556,
+      "loss": 4.3574,
+      "step": 4399
+    },
+    {
+      "epoch": 0.4562274397475196,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018884429651088702,
+      "loss": 4.3811,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4563311278020077,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001888393103346797,
+      "loss": 4.3547,
+      "step": 4401
+    },
+    {
+      "epoch": 0.45643481585649576,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018883432311026242,
+      "loss": 4.4052,
+      "step": 4402
+    },
+    {
+      "epoch": 0.45653850391098383,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018882933483769403,
+      "loss": 4.3587,
+      "step": 4403
+    },
+    {
+      "epoch": 0.4566421919654719,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001888243455170334,
+      "loss": 4.3803,
+      "step": 4404
+    },
+    {
+      "epoch": 0.45674588001995997,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018881935514833935,
+      "loss": 4.3837,
+      "step": 4405
+    },
+    {
+      "epoch": 0.45684956807444804,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018881436373167083,
+      "loss": 4.3864,
+      "step": 4406
+    },
+    {
+      "epoch": 0.4569532561289361,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001888093712670867,
+      "loss": 4.3537,
+      "step": 4407
+    },
+    {
+      "epoch": 0.4570569441834242,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018880437775464583,
+      "loss": 4.3655,
+      "step": 4408
+    },
+    {
+      "epoch": 0.45716063223791226,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001887993831944072,
+      "loss": 4.4028,
+      "step": 4409
+    },
+    {
+      "epoch": 0.45726432029240033,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018879438758642972,
+      "loss": 4.3589,
+      "step": 4410
+    },
+    {
+      "epoch": 0.4573680083468884,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018878939093077232,
+      "loss": 4.3918,
+      "step": 4411
+    },
+    {
+      "epoch": 0.45747169640137647,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018878439322749392,
+      "loss": 4.3798,
+      "step": 4412
+    },
+    {
+      "epoch": 0.45757538445586454,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018877939447665358,
+      "loss": 4.3469,
+      "step": 4413
+    },
+    {
+      "epoch": 0.4576790725103526,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018877439467831022,
+      "loss": 4.3823,
+      "step": 4414
+    },
+    {
+      "epoch": 0.4577827605648407,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018876939383252284,
+      "loss": 4.3694,
+      "step": 4415
+    },
+    {
+      "epoch": 0.45788644861932876,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018876439193935042,
+      "loss": 4.3747,
+      "step": 4416
+    },
+    {
+      "epoch": 0.45799013667381683,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018875938899885202,
+      "loss": 4.4072,
+      "step": 4417
+    },
+    {
+      "epoch": 0.4580938247283049,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018875438501108667,
+      "loss": 4.395,
+      "step": 4418
+    },
+    {
+      "epoch": 0.45819751278279297,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018874937997611336,
+      "loss": 4.3889,
+      "step": 4419
+    },
+    {
+      "epoch": 0.45830120083728104,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018874437389399123,
+      "loss": 4.372,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4584048888917691,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018873936676477927,
+      "loss": 4.3597,
+      "step": 4421
+    },
+    {
+      "epoch": 0.4585085769462572,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018873435858853655,
+      "loss": 4.3904,
+      "step": 4422
+    },
+    {
+      "epoch": 0.45861226500074526,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001887293493653222,
+      "loss": 4.3842,
+      "step": 4423
+    },
+    {
+      "epoch": 0.4587159530552333,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018872433909519537,
+      "loss": 4.3949,
+      "step": 4424
+    },
+    {
+      "epoch": 0.4588196411097214,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018871932777821509,
+      "loss": 4.4121,
+      "step": 4425
+    },
+    {
+      "epoch": 0.45892332916420947,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018871431541444053,
+      "loss": 4.3992,
+      "step": 4426
+    },
+    {
+      "epoch": 0.45902701721869754,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018870930200393083,
+      "loss": 4.3586,
+      "step": 4427
+    },
+    {
+      "epoch": 0.4591307052731856,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001887042875467451,
+      "loss": 4.3472,
+      "step": 4428
+    },
+    {
+      "epoch": 0.4592343933276737,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018869927204294258,
+      "loss": 4.3902,
+      "step": 4429
+    },
+    {
+      "epoch": 0.45933808138216176,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001886942554925824,
+      "loss": 4.4539,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4594417694366498,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018868923789572376,
+      "loss": 4.3847,
+      "step": 4431
+    },
+    {
+      "epoch": 0.4595454574911379,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018868421925242586,
+      "loss": 4.3835,
+      "step": 4432
+    },
+    {
+      "epoch": 0.45964914554562597,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001886791995627479,
+      "loss": 4.3715,
+      "step": 4433
+    },
+    {
+      "epoch": 0.45975283360011404,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018867417882674915,
+      "loss": 4.3309,
+      "step": 4434
+    },
+    {
+      "epoch": 0.4598565216546021,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001886691570444888,
+      "loss": 4.3963,
+      "step": 4435
+    },
+    {
+      "epoch": 0.4599602097090902,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018866413421602613,
+      "loss": 4.4098,
+      "step": 4436
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018865911034142042,
+      "loss": 4.3952,
+      "step": 4437
+    },
+    {
+      "epoch": 0.4601675858180663,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018865408542073089,
+      "loss": 4.4235,
+      "step": 4438
+    },
+    {
+      "epoch": 0.4602712738725544,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018864905945401687,
+      "loss": 4.3633,
+      "step": 4439
+    },
+    {
+      "epoch": 0.46037496192704247,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018864403244133767,
+      "loss": 4.395,
+      "step": 4440
+    },
+    {
+      "epoch": 0.46047864998153054,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018863900438275256,
+      "loss": 4.3403,
+      "step": 4441
+    },
+    {
+      "epoch": 0.4605823380360186,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018863397527832095,
+      "loss": 4.3693,
+      "step": 4442
+    },
+    {
+      "epoch": 0.4606860260905067,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018862894512810207,
+      "loss": 4.4441,
+      "step": 4443
+    },
+    {
+      "epoch": 0.4607897141449948,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018862391393215534,
+      "loss": 4.3215,
+      "step": 4444
+    },
+    {
+      "epoch": 0.4608934021994829,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018861888169054012,
+      "loss": 4.3838,
+      "step": 4445
+    },
+    {
+      "epoch": 0.46099709025397095,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018861384840331575,
+      "loss": 4.3675,
+      "step": 4446
+    },
+    {
+      "epoch": 0.461100778308459,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018860881407054163,
+      "loss": 4.3918,
+      "step": 4447
+    },
+    {
+      "epoch": 0.4612044663629471,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001886037786922772,
+      "loss": 4.3753,
+      "step": 4448
+    },
+    {
+      "epoch": 0.46130815441743517,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001885987422685818,
+      "loss": 4.4139,
+      "step": 4449
+    },
+    {
+      "epoch": 0.46141184247192324,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018859370479951492,
+      "loss": 4.4179,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4615155305264113,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018858866628513598,
+      "loss": 4.32,
+      "step": 4451
+    },
+    {
+      "epoch": 0.4616192185808994,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001885836267255044,
+      "loss": 4.4127,
+      "step": 4452
+    },
+    {
+      "epoch": 0.46172290663538745,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018857858612067967,
+      "loss": 4.3886,
+      "step": 4453
+    },
+    {
+      "epoch": 0.4618265946898755,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018857354447072123,
+      "loss": 4.3949,
+      "step": 4454
+    },
+    {
+      "epoch": 0.4619302827443636,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018856850177568864,
+      "loss": 4.3957,
+      "step": 4455
+    },
+    {
+      "epoch": 0.46203397079885167,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018856345803564133,
+      "loss": 4.3744,
+      "step": 4456
+    },
+    {
+      "epoch": 0.46213765885333974,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018855841325063883,
+      "loss": 4.3692,
+      "step": 4457
+    },
+    {
+      "epoch": 0.4622413469078278,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018855336742074066,
+      "loss": 4.3896,
+      "step": 4458
+    },
+    {
+      "epoch": 0.4623450349623159,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018854832054600635,
+      "loss": 4.3986,
+      "step": 4459
+    },
+    {
+      "epoch": 0.46244872301680395,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018854327262649546,
+      "loss": 4.377,
+      "step": 4460
+    },
+    {
+      "epoch": 0.462552411071292,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018853822366226756,
+      "loss": 4.3758,
+      "step": 4461
+    },
+    {
+      "epoch": 0.4626560991257801,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018853317365338218,
+      "loss": 4.4131,
+      "step": 4462
+    },
+    {
+      "epoch": 0.46275978718026817,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000188528122599899,
+      "loss": 4.3954,
+      "step": 4463
+    },
+    {
+      "epoch": 0.46286347523475624,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018852307050187749,
+      "loss": 4.3832,
+      "step": 4464
+    },
+    {
+      "epoch": 0.4629671632892443,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018851801735937732,
+      "loss": 4.3479,
+      "step": 4465
+    },
+    {
+      "epoch": 0.4630708513437324,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018851296317245816,
+      "loss": 4.3886,
+      "step": 4466
+    },
+    {
+      "epoch": 0.46317453939822045,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018850790794117957,
+      "loss": 4.3542,
+      "step": 4467
+    },
+    {
+      "epoch": 0.4632782274527085,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001885028516656012,
+      "loss": 4.3936,
+      "step": 4468
+    },
+    {
+      "epoch": 0.4633819155071966,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018849779434578276,
+      "loss": 4.4084,
+      "step": 4469
+    },
+    {
+      "epoch": 0.46348560356168467,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001884927359817839,
+      "loss": 4.4057,
+      "step": 4470
+    },
+    {
+      "epoch": 0.46358929161617274,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001884876765736643,
+      "loss": 4.3766,
+      "step": 4471
+    },
+    {
+      "epoch": 0.4636929796706608,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001884826161214836,
+      "loss": 4.3469,
+      "step": 4472
+    },
+    {
+      "epoch": 0.4637966677251489,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018847755462530162,
+      "loss": 4.4156,
+      "step": 4473
+    },
+    {
+      "epoch": 0.46390035577963695,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.000188472492085178,
+      "loss": 4.4327,
+      "step": 4474
+    },
+    {
+      "epoch": 0.464004043834125,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018846742850117248,
+      "loss": 4.4199,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4641077318886131,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001884623638733448,
+      "loss": 4.3494,
+      "step": 4476
+    },
+    {
+      "epoch": 0.46421141994310117,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018845729820175477,
+      "loss": 4.3995,
+      "step": 4477
+    },
+    {
+      "epoch": 0.46431510799758924,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001884522314864621,
+      "loss": 4.329,
+      "step": 4478
+    },
+    {
+      "epoch": 0.4644187960520773,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018844716372752663,
+      "loss": 4.4106,
+      "step": 4479
+    },
+    {
+      "epoch": 0.4645224841065654,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018844209492500813,
+      "loss": 4.4032,
+      "step": 4480
+    },
+    {
+      "epoch": 0.46462617216105345,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018843702507896634,
+      "loss": 4.4141,
+      "step": 4481
+    },
+    {
+      "epoch": 0.4647298602155415,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018843195418946117,
+      "loss": 4.4069,
+      "step": 4482
+    },
+    {
+      "epoch": 0.4648335482700296,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018842688225655243,
+      "loss": 4.3916,
+      "step": 4483
+    },
+    {
+      "epoch": 0.46493723632451767,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018842180928029992,
+      "loss": 4.3536,
+      "step": 4484
+    },
+    {
+      "epoch": 0.46504092437900574,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018841673526076355,
+      "loss": 4.3959,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4651446124334938,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018841166019800315,
+      "loss": 4.4139,
+      "step": 4486
+    },
+    {
+      "epoch": 0.4652483004879819,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018840658409207862,
+      "loss": 4.31,
+      "step": 4487
+    },
+    {
+      "epoch": 0.46535198854246995,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018840150694304986,
+      "loss": 4.3708,
+      "step": 4488
+    },
+    {
+      "epoch": 0.4654556765969581,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018839642875097674,
+      "loss": 4.3896,
+      "step": 4489
+    },
+    {
+      "epoch": 0.46555936465144615,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001883913495159192,
+      "loss": 4.374,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4656630527059342,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001883862692379372,
+      "loss": 4.3423,
+      "step": 4491
+    },
+    {
+      "epoch": 0.4657667407604223,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018838118791709063,
+      "loss": 4.4045,
+      "step": 4492
+    },
+    {
+      "epoch": 0.46587042881491036,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018837610555343947,
+      "loss": 4.3357,
+      "step": 4493
+    },
+    {
+      "epoch": 0.46597411686939844,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018837102214704367,
+      "loss": 4.3608,
+      "step": 4494
+    },
+    {
+      "epoch": 0.4660778049238865,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001883659376979632,
+      "loss": 4.3956,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4661814929783746,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018836085220625814,
+      "loss": 4.4046,
+      "step": 4496
+    },
+    {
+      "epoch": 0.46628518103286265,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018835576567198837,
+      "loss": 4.3886,
+      "step": 4497
+    },
+    {
+      "epoch": 0.4663888690873507,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000188350678095214,
+      "loss": 4.363,
+      "step": 4498
+    },
+    {
+      "epoch": 0.4664925571418388,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018834558947599498,
+      "loss": 4.321,
+      "step": 4499
+    },
+    {
+      "epoch": 0.46659624519632686,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001883404998143914,
+      "loss": 4.3768,
+      "step": 4500
+    },
+    {
+      "epoch": 0.46669993325081494,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001883354091104633,
+      "loss": 4.369,
+      "step": 4501
+    },
+    {
+      "epoch": 0.466803621305303,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018833031736427075,
+      "loss": 4.3661,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4669073093597911,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018832522457587386,
+      "loss": 4.3625,
+      "step": 4503
+    },
+    {
+      "epoch": 0.46701099741427915,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018832013074533265,
+      "loss": 4.3675,
+      "step": 4504
+    },
+    {
+      "epoch": 0.4671146854687672,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018831503587270727,
+      "loss": 4.3922,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4672183735232553,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018830993995805782,
+      "loss": 4.385,
+      "step": 4506
+    },
+    {
+      "epoch": 0.46732206157774336,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001883048430014444,
+      "loss": 4.3942,
+      "step": 4507
+    },
+    {
+      "epoch": 0.46742574963223144,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018829974500292717,
+      "loss": 4.3906,
+      "step": 4508
+    },
+    {
+      "epoch": 0.4675294376867195,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018829464596256632,
+      "loss": 4.3429,
+      "step": 4509
+    },
+    {
+      "epoch": 0.4676331257412076,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018828954588042196,
+      "loss": 4.4135,
+      "step": 4510
+    },
+    {
+      "epoch": 0.46773681379569565,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001882844447565543,
+      "loss": 4.3824,
+      "step": 4511
+    },
+    {
+      "epoch": 0.4678405018501837,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018827934259102352,
+      "loss": 4.3978,
+      "step": 4512
+    },
+    {
+      "epoch": 0.4679441899046718,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018827423938388977,
+      "loss": 4.3683,
+      "step": 4513
+    },
+    {
+      "epoch": 0.46804787795915986,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018826913513521335,
+      "loss": 4.4184,
+      "step": 4514
+    },
+    {
+      "epoch": 0.46815156601364794,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001882640298450544,
+      "loss": 4.3855,
+      "step": 4515
+    },
+    {
+      "epoch": 0.468255254068136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001882589235134732,
+      "loss": 4.3802,
+      "step": 4516
+    },
+    {
+      "epoch": 0.4683589421226241,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018825381614053004,
+      "loss": 4.3976,
+      "step": 4517
+    },
+    {
+      "epoch": 0.46846263017711215,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018824870772628512,
+      "loss": 4.3788,
+      "step": 4518
+    },
+    {
+      "epoch": 0.4685663182316002,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018824359827079873,
+      "loss": 4.391,
+      "step": 4519
+    },
+    {
+      "epoch": 0.4686700062860883,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018823848777413114,
+      "loss": 4.3887,
+      "step": 4520
+    },
+    {
+      "epoch": 0.46877369434057636,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018823337623634267,
+      "loss": 4.407,
+      "step": 4521
+    },
+    {
+      "epoch": 0.46887738239506443,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018822826365749365,
+      "loss": 4.4155,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4689810704495525,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018822315003764434,
+      "loss": 4.4231,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4690847585040406,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018821803537685515,
+      "loss": 4.3997,
+      "step": 4524
+    },
+    {
+      "epoch": 0.46918844655852865,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018821291967518637,
+      "loss": 4.3025,
+      "step": 4525
+    },
+    {
+      "epoch": 0.4692921346130167,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001882078029326984,
+      "loss": 4.4274,
+      "step": 4526
+    },
+    {
+      "epoch": 0.4693958226675048,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001882026851494516,
+      "loss": 4.3967,
+      "step": 4527
+    },
+    {
+      "epoch": 0.46949951072199286,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018819756632550635,
+      "loss": 4.3833,
+      "step": 4528
+    },
+    {
+      "epoch": 0.46960319877648093,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018819244646092303,
+      "loss": 4.3692,
+      "step": 4529
+    },
+    {
+      "epoch": 0.469706886830969,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018818732555576207,
+      "loss": 4.3743,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4698105748854571,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001881822036100839,
+      "loss": 4.3747,
+      "step": 4531
+    },
+    {
+      "epoch": 0.46991426293994515,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001881770806239489,
+      "loss": 4.3656,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4700179509944332,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001881719565974176,
+      "loss": 4.3915,
+      "step": 4533
+    },
+    {
+      "epoch": 0.47012163904892135,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001881668315305504,
+      "loss": 4.3953,
+      "step": 4534
+    },
+    {
+      "epoch": 0.4702253271034094,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001881617054234078,
+      "loss": 4.3876,
+      "step": 4535
+    },
+    {
+      "epoch": 0.4703290151578975,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018815657827605023,
+      "loss": 4.3467,
+      "step": 4536
+    },
+    {
+      "epoch": 0.47043270321238556,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018815145008853823,
+      "loss": 4.3883,
+      "step": 4537
+    },
+    {
+      "epoch": 0.47053639126687363,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018814632086093234,
+      "loss": 4.3748,
+      "step": 4538
+    },
+    {
+      "epoch": 0.4706400793213617,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000188141190593293,
+      "loss": 4.3926,
+      "step": 4539
+    },
+    {
+      "epoch": 0.4707437673758498,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001881360592856808,
+      "loss": 4.3796,
+      "step": 4540
+    },
+    {
+      "epoch": 0.47084745543033785,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018813092693815623,
+      "loss": 4.357,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4709511434848259,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018812579355077992,
+      "loss": 4.3667,
+      "step": 4542
+    },
+    {
+      "epoch": 0.471054831539314,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001881206591236124,
+      "loss": 4.3696,
+      "step": 4543
+    },
+    {
+      "epoch": 0.47115851959380206,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018811552365671422,
+      "loss": 4.4268,
+      "step": 4544
+    },
+    {
+      "epoch": 0.47126220764829013,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000188110387150146,
+      "loss": 4.372,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4713658957027782,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018810524960396837,
+      "loss": 4.3459,
+      "step": 4546
+    },
+    {
+      "epoch": 0.4714695837572663,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018810011101824193,
+      "loss": 4.3908,
+      "step": 4547
+    },
+    {
+      "epoch": 0.47157327181175435,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001880949713930273,
+      "loss": 4.4018,
+      "step": 4548
+    },
+    {
+      "epoch": 0.4716769598662424,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001880898307283851,
+      "loss": 4.3963,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4717806479207305,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018808468902437606,
+      "loss": 4.4004,
+      "step": 4550
+    },
+    {
+      "epoch": 0.47188433597521856,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018807954628106076,
+      "loss": 4.3346,
+      "step": 4551
+    },
+    {
+      "epoch": 0.47198802402970663,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018807440249849996,
+      "loss": 4.3979,
+      "step": 4552
+    },
+    {
+      "epoch": 0.4720917120841947,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018806925767675425,
+      "loss": 4.3794,
+      "step": 4553
+    },
+    {
+      "epoch": 0.4721954001386828,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018806411181588443,
+      "loss": 4.3476,
+      "step": 4554
+    },
+    {
+      "epoch": 0.47229908819317085,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018805896491595116,
+      "loss": 4.3423,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4724027762476589,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018805381697701524,
+      "loss": 4.3548,
+      "step": 4556
+    },
+    {
+      "epoch": 0.472506464302147,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001880486679991373,
+      "loss": 4.3531,
+      "step": 4557
+    },
+    {
+      "epoch": 0.47261015235663506,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018804351798237817,
+      "loss": 4.3784,
+      "step": 4558
+    },
+    {
+      "epoch": 0.47271384041112313,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018803836692679856,
+      "loss": 4.3495,
+      "step": 4559
+    },
+    {
+      "epoch": 0.4728175284656112,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018803321483245932,
+      "loss": 4.3597,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4729212165200993,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018802806169942123,
+      "loss": 4.3687,
+      "step": 4561
+    },
+    {
+      "epoch": 0.47302490457458735,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.000188022907527745,
+      "loss": 4.3541,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4731285926290754,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018801775231749152,
+      "loss": 4.412,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4732322806835635,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001880125960687216,
+      "loss": 4.359,
+      "step": 4564
+    },
+    {
+      "epoch": 0.47333596873805156,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018800743878149613,
+      "loss": 4.3814,
+      "step": 4565
+    },
+    {
+      "epoch": 0.47343965679253963,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018800228045587586,
+      "loss": 4.3545,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4735433448470277,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001879971210919217,
+      "loss": 4.377,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4736470329015158,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018799196068969453,
+      "loss": 4.3649,
+      "step": 4568
+    },
+    {
+      "epoch": 0.47375072095600385,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018798679924925525,
+      "loss": 4.3682,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4738544090104919,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018798163677066475,
+      "loss": 4.3693,
+      "step": 4570
+    },
+    {
+      "epoch": 0.47395809706498,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018797647325398392,
+      "loss": 4.3592,
+      "step": 4571
+    },
+    {
+      "epoch": 0.47406178511946806,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001879713086992737,
+      "loss": 4.3882,
+      "step": 4572
+    },
+    {
+      "epoch": 0.47416547317395613,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018796614310659506,
+      "loss": 4.3977,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4742691612284442,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018796097647600887,
+      "loss": 4.3923,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4743728492829323,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018795580880757618,
+      "loss": 4.3809,
+      "step": 4575
+    },
+    {
+      "epoch": 0.47447653733742035,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018795064010135787,
+      "loss": 4.403,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4745802253919084,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.000187945470357415,
+      "loss": 4.3187,
+      "step": 4577
+    },
+    {
+      "epoch": 0.47468391344639654,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018794029957580857,
+      "loss": 4.3486,
+      "step": 4578
+    },
+    {
+      "epoch": 0.4747876015008846,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001879351277565995,
+      "loss": 4.3955,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4748912895553727,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018792995489984893,
+      "loss": 4.3646,
+      "step": 4580
+    },
+    {
+      "epoch": 0.47499497760986076,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001879247810056178,
+      "loss": 4.3669,
+      "step": 4581
+    },
+    {
+      "epoch": 0.47509866566434883,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018791960607396723,
+      "loss": 4.3681,
+      "step": 4582
+    },
+    {
+      "epoch": 0.4752023537188369,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001879144301049582,
+      "loss": 4.3619,
+      "step": 4583
+    },
+    {
+      "epoch": 0.47530604177332497,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001879092530986519,
+      "loss": 4.3535,
+      "step": 4584
+    },
+    {
+      "epoch": 0.47540972982781304,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018790407505510929,
+      "loss": 4.33,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4755134178823011,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018789889597439148,
+      "loss": 4.4277,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4756171059367892,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018789371585655964,
+      "loss": 4.3586,
+      "step": 4587
+    },
+    {
+      "epoch": 0.47572079399127726,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018788853470167488,
+      "loss": 4.3724,
+      "step": 4588
+    },
+    {
+      "epoch": 0.47582448204576533,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018788335250979828,
+      "loss": 4.3605,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4759281701002534,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018787816928099102,
+      "loss": 4.3503,
+      "step": 4590
+    },
+    {
+      "epoch": 0.47603185815474147,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018787298501531428,
+      "loss": 4.4126,
+      "step": 4591
+    },
+    {
+      "epoch": 0.47613554620922954,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018786779971282917,
+      "loss": 4.3549,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4762392342637176,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001878626133735969,
+      "loss": 4.3635,
+      "step": 4593
+    },
+    {
+      "epoch": 0.4763429223182057,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018785742599767872,
+      "loss": 4.3495,
+      "step": 4594
+    },
+    {
+      "epoch": 0.47644661037269376,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018785223758513575,
+      "loss": 4.3772,
+      "step": 4595
+    },
+    {
+      "epoch": 0.47655029842718183,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018784704813602925,
+      "loss": 4.3902,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4766539864816699,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018784185765042042,
+      "loss": 4.3868,
+      "step": 4597
+    },
+    {
+      "epoch": 0.47675767453615797,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018783666612837054,
+      "loss": 4.3831,
+      "step": 4598
+    },
+    {
+      "epoch": 0.47686136259064604,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018783147356994084,
+      "loss": 4.3242,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4769650506451341,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001878262799751926,
+      "loss": 4.3894,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4770687386996222,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018782108534418708,
+      "loss": 4.3856,
+      "step": 4601
+    },
+    {
+      "epoch": 0.47717242675411026,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018781588967698557,
+      "loss": 4.3688,
+      "step": 4602
+    },
+    {
+      "epoch": 0.47727611480859833,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018781069297364944,
+      "loss": 4.3715,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4773798028630864,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001878054952342399,
+      "loss": 4.3521,
+      "step": 4604
+    },
+    {
+      "epoch": 0.47748349091757447,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018780029645881836,
+      "loss": 4.3793,
+      "step": 4605
+    },
+    {
+      "epoch": 0.47758717897206254,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001877950966474461,
+      "loss": 4.3731,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4776908670265506,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018778989580018455,
+      "loss": 4.3969,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4777945550810387,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.000187784693917095,
+      "loss": 4.3838,
+      "step": 4608
+    },
+    {
+      "epoch": 0.47789824313552676,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018777949099823887,
+      "loss": 4.3661,
+      "step": 4609
+    },
+    {
+      "epoch": 0.47800193119001483,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018777428704367752,
+      "loss": 4.3812,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4781056192445029,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018776908205347237,
+      "loss": 4.3726,
+      "step": 4611
+    },
+    {
+      "epoch": 0.47820930729899097,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018776387602768483,
+      "loss": 4.3957,
+      "step": 4612
+    },
+    {
+      "epoch": 0.47831299535347904,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001877586689663763,
+      "loss": 4.3695,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4784166834079671,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018775346086960827,
+      "loss": 4.3326,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4785203714624552,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018774825173744212,
+      "loss": 4.394,
+      "step": 4615
+    },
+    {
+      "epoch": 0.47862405951694326,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001877430415699394,
+      "loss": 4.3301,
+      "step": 4616
+    },
+    {
+      "epoch": 0.47872774757143133,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018773783036716153,
+      "loss": 4.3479,
+      "step": 4617
+    },
+    {
+      "epoch": 0.4788314356259194,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018773261812916997,
+      "loss": 4.3828,
+      "step": 4618
+    },
+    {
+      "epoch": 0.47893512368040747,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018772740485602628,
+      "loss": 4.3921,
+      "step": 4619
+    },
+    {
+      "epoch": 0.47903881173489554,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018772219054779193,
+      "loss": 4.3895,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4791424997893836,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001877169752045285,
+      "loss": 4.3958,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4792461878438717,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018771175882629744,
+      "loss": 4.3306,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4793498758983598,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018770654141316037,
+      "loss": 4.3752,
+      "step": 4623
+    },
+    {
+      "epoch": 0.4794535639528479,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001877013229651788,
+      "loss": 4.3535,
+      "step": 4624
+    },
+    {
+      "epoch": 0.47955725200733595,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018769610348241434,
+      "loss": 4.3461,
+      "step": 4625
+    },
+    {
+      "epoch": 0.479660940061824,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018769088296492854,
+      "loss": 4.3541,
+      "step": 4626
+    },
+    {
+      "epoch": 0.4797646281163121,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000187685661412783,
+      "loss": 4.3587,
+      "step": 4627
+    },
+    {
+      "epoch": 0.47986831617080017,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018768043882603935,
+      "loss": 4.3785,
+      "step": 4628
+    },
+    {
+      "epoch": 0.47997200422528824,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018767521520475925,
+      "loss": 4.3602,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4800756922797763,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018766999054900424,
+      "loss": 4.403,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4801793803342644,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018766476485883603,
+      "loss": 4.3446,
+      "step": 4631
+    },
+    {
+      "epoch": 0.48028306838875245,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018765953813431628,
+      "loss": 4.3514,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4803867564432405,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018765431037550662,
+      "loss": 4.3735,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4804904444977286,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018764908158246875,
+      "loss": 4.3809,
+      "step": 4634
+    },
+    {
+      "epoch": 0.48059413255221667,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018764385175526436,
+      "loss": 4.3671,
+      "step": 4635
+    },
+    {
+      "epoch": 0.48069782060670474,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018763862089395515,
+      "loss": 4.3401,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4808015086611928,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018763338899860287,
+      "loss": 4.3297,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4809051967156809,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018762815606926926,
+      "loss": 4.3903,
+      "step": 4638
+    },
+    {
+      "epoch": 0.48100888477016895,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.000187622922106016,
+      "loss": 4.3983,
+      "step": 4639
+    },
+    {
+      "epoch": 0.481112572824657,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001876176871089049,
+      "loss": 4.3711,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4812162608791451,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018761245107799769,
+      "loss": 4.3311,
+      "step": 4641
+    },
+    {
+      "epoch": 0.48131994893363317,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001876072140133562,
+      "loss": 4.3567,
+      "step": 4642
+    },
+    {
+      "epoch": 0.48142363698812124,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018760197591504213,
+      "loss": 4.3462,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4815273250426093,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001875967367831174,
+      "loss": 4.3802,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4816310130970974,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018759149661764374,
+      "loss": 4.3746,
+      "step": 4645
+    },
+    {
+      "epoch": 0.48173470115158545,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018758625541868303,
+      "loss": 4.3654,
+      "step": 4646
+    },
+    {
+      "epoch": 0.4818383892060735,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018758101318629706,
+      "loss": 4.3534,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4819420772605616,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018757576992054772,
+      "loss": 4.388,
+      "step": 4648
+    },
+    {
+      "epoch": 0.48204576531504967,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001875705256214969,
+      "loss": 4.3819,
+      "step": 4649
+    },
+    {
+      "epoch": 0.48214945336953774,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018756528028920642,
+      "loss": 4.3552,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4822531414240258,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018756003392373817,
+      "loss": 4.394,
+      "step": 4651
+    },
+    {
+      "epoch": 0.4823568294785139,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018755478652515407,
+      "loss": 4.3123,
+      "step": 4652
+    },
+    {
+      "epoch": 0.48246051753300195,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018754953809351608,
+      "loss": 4.3751,
+      "step": 4653
+    },
+    {
+      "epoch": 0.48256420558749,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018754428862888606,
+      "loss": 4.3575,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4826678936419781,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000187539038131326,
+      "loss": 4.3291,
+      "step": 4655
+    },
+    {
+      "epoch": 0.48277158169646617,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001875337866008978,
+      "loss": 4.3593,
+      "step": 4656
+    },
+    {
+      "epoch": 0.48287526975095424,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018752853403766344,
+      "loss": 4.3647,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4829789578054423,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018752328044168492,
+      "loss": 4.3949,
+      "step": 4658
+    },
+    {
+      "epoch": 0.4830826458599304,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001875180258130242,
+      "loss": 4.3514,
+      "step": 4659
+    },
+    {
+      "epoch": 0.48318633391441845,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018751277015174327,
+      "loss": 4.3909,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4832900219689065,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018750751345790416,
+      "loss": 4.3897,
+      "step": 4661
+    },
+    {
+      "epoch": 0.4833937100233946,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018750225573156893,
+      "loss": 4.3775,
+      "step": 4662
+    },
+    {
+      "epoch": 0.48349739807788267,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018749699697279953,
+      "loss": 4.3591,
+      "step": 4663
+    },
+    {
+      "epoch": 0.48360108613237074,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018749173718165805,
+      "loss": 4.3801,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4837047741868588,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018748647635820657,
+      "loss": 4.3897,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4838084622413469,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018748121450250715,
+      "loss": 4.3256,
+      "step": 4666
+    },
+    {
+      "epoch": 0.48391215029583495,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001874759516146219,
+      "loss": 4.3901,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4840158383503231,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018747068769461284,
+      "loss": 4.3546,
+      "step": 4668
+    },
+    {
+      "epoch": 0.48411952640481115,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018746542274254214,
+      "loss": 4.3735,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4842232144592992,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001874601567584719,
+      "loss": 4.3351,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4843269025137873,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018745488974246431,
+      "loss": 4.3622,
+      "step": 4671
+    },
+    {
+      "epoch": 0.48443059056827537,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001874496216945814,
+      "loss": 4.3652,
+      "step": 4672
+    },
+    {
+      "epoch": 0.48453427862276344,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018744435261488541,
+      "loss": 4.3184,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4846379666772515,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018743908250343848,
+      "loss": 4.3506,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4847416547317396,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018743381136030284,
+      "loss": 4.372,
+      "step": 4675
+    },
+    {
+      "epoch": 0.48484534278622765,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018742853918554065,
+      "loss": 4.3609,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4849490308407157,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018742326597921406,
+      "loss": 4.3369,
+      "step": 4677
+    },
+    {
+      "epoch": 0.4850527188952038,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018741799174138538,
+      "loss": 4.425,
+      "step": 4678
+    },
+    {
+      "epoch": 0.48515640694969187,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001874127164721168,
+      "loss": 4.3741,
+      "step": 4679
+    },
+    {
+      "epoch": 0.48526009500417994,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018740744017147056,
+      "loss": 4.3473,
+      "step": 4680
+    },
+    {
+      "epoch": 0.485363783058668,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018740216283950895,
+      "loss": 4.3419,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4854674711131561,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018739688447629415,
+      "loss": 4.3533,
+      "step": 4682
+    },
+    {
+      "epoch": 0.48557115916764415,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018739160508188853,
+      "loss": 4.3423,
+      "step": 4683
+    },
+    {
+      "epoch": 0.4856748472221322,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018738632465635434,
+      "loss": 4.2913,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4857785352766203,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001873810431997539,
+      "loss": 4.3524,
+      "step": 4685
+    },
+    {
+      "epoch": 0.48588222333110836,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018737576071214948,
+      "loss": 4.377,
+      "step": 4686
+    },
+    {
+      "epoch": 0.48598591138559644,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018737047719360347,
+      "loss": 4.3887,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4860895994400845,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018736519264417822,
+      "loss": 4.4016,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4861932874945726,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018735990706393599,
+      "loss": 4.3848,
+      "step": 4689
+    },
+    {
+      "epoch": 0.48629697554906065,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018735462045293923,
+      "loss": 4.3491,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4864006636035487,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018734933281125028,
+      "loss": 4.3585,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4865043516580368,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001873440441389315,
+      "loss": 4.3719,
+      "step": 4692
+    },
+    {
+      "epoch": 0.48660803971252486,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018733875443604538,
+      "loss": 4.3889,
+      "step": 4693
+    },
+    {
+      "epoch": 0.48671172776701294,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018733346370265427,
+      "loss": 4.382,
+      "step": 4694
+    },
+    {
+      "epoch": 0.486815415821501,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018732817193882058,
+      "loss": 4.3575,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4869191038759891,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018732287914460677,
+      "loss": 4.3879,
+      "step": 4696
+    },
+    {
+      "epoch": 0.48702279193047715,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001873175853200753,
+      "loss": 4.3682,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4871264799849652,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018731229046528865,
+      "loss": 4.3852,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4872301680394533,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001873069945803092,
+      "loss": 4.3816,
+      "step": 4699
+    },
+    {
+      "epoch": 0.48733385609394136,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018730169766519955,
+      "loss": 4.3754,
+      "step": 4700
+    },
+    {
+      "epoch": 0.48743754414842944,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018729639972002215,
+      "loss": 4.3726,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4875412322029175,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018729110074483952,
+      "loss": 4.3491,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4876449202574056,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018728580073971413,
+      "loss": 4.4031,
+      "step": 4703
+    },
+    {
+      "epoch": 0.48774860831189365,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018728049970470858,
+      "loss": 4.3821,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4878522963663817,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001872751976398854,
+      "loss": 4.3892,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4879559844208698,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018726989454530713,
+      "loss": 4.3552,
+      "step": 4706
+    },
+    {
+      "epoch": 0.48805967247535786,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018726459042103634,
+      "loss": 4.3492,
+      "step": 4707
+    },
+    {
+      "epoch": 0.48816336052984594,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018725928526713566,
+      "loss": 4.3485,
+      "step": 4708
+    },
+    {
+      "epoch": 0.488267048584334,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018725397908366762,
+      "loss": 4.3352,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4883707366388221,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018724867187069487,
+      "loss": 4.3679,
+      "step": 4710
+    },
+    {
+      "epoch": 0.48847442469331015,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018724336362828002,
+      "loss": 4.3396,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4885781127477983,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001872380543564857,
+      "loss": 4.3776,
+      "step": 4712
+    },
+    {
+      "epoch": 0.48868180080228635,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018723274405537454,
+      "loss": 4.3836,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4887854888567744,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018722743272500921,
+      "loss": 4.352,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4888891769112625,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001872221203654524,
+      "loss": 4.3951,
+      "step": 4715
+    },
+    {
+      "epoch": 0.48899286496575056,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018721680697676675,
+      "loss": 4.4221,
+      "step": 4716
+    },
+    {
+      "epoch": 0.48909655302023863,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018721149255901499,
+      "loss": 4.3905,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4892002410747267,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001872061771122598,
+      "loss": 4.3255,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4893039291292148,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018720086063656388,
+      "loss": 4.371,
+      "step": 4719
+    },
+    {
+      "epoch": 0.48940761718370285,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018719554313198996,
+      "loss": 4.3986,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4895113052381909,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018719022459860084,
+      "loss": 4.4115,
+      "step": 4721
+    },
+    {
+      "epoch": 0.489614993292679,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018718490503645923,
+      "loss": 4.3245,
+      "step": 4722
+    },
+    {
+      "epoch": 0.48971868134716706,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001871795844456279,
+      "loss": 4.4211,
+      "step": 4723
+    },
+    {
+      "epoch": 0.48982236940165513,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001871742628261696,
+      "loss": 4.407,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4899260574561432,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018716894017814718,
+      "loss": 4.3849,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4900297455106313,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018716361650162336,
+      "loss": 4.3758,
+      "step": 4726
+    },
+    {
+      "epoch": 0.49013343356511935,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018715829179666104,
+      "loss": 4.3135,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4902371216196074,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018715296606332303,
+      "loss": 4.427,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4903408096740955,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001871476393016721,
+      "loss": 4.3769,
+      "step": 4729
+    },
+    {
+      "epoch": 0.49044449772858356,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018714231151177116,
+      "loss": 4.3326,
+      "step": 4730
+    },
+    {
+      "epoch": 0.49054818578307163,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018713698269368306,
+      "loss": 4.3383,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4906518738375597,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018713165284747069,
+      "loss": 4.39,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4907555618920478,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018712632197319688,
+      "loss": 4.38,
+      "step": 4733
+    },
+    {
+      "epoch": 0.49085924994653585,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001871209900709246,
+      "loss": 4.359,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4909629380010239,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001871156571407167,
+      "loss": 4.3808,
+      "step": 4735
+    },
+    {
+      "epoch": 0.491066626055512,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018711032318263616,
+      "loss": 4.3714,
+      "step": 4736
+    },
+    {
+      "epoch": 0.49117031411000006,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018710498819674584,
+      "loss": 4.3578,
+      "step": 4737
+    },
+    {
+      "epoch": 0.49127400216448813,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001870996521831088,
+      "loss": 4.3482,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4913776902189762,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001870943151417879,
+      "loss": 4.4213,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4914813782734643,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018708897707284613,
+      "loss": 4.3695,
+      "step": 4740
+    },
+    {
+      "epoch": 0.49158506632795235,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001870836379763465,
+      "loss": 4.3712,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4916887543824404,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018707829785235197,
+      "loss": 4.3614,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4917924424369285,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001870729567009256,
+      "loss": 4.3272,
+      "step": 4743
+    },
+    {
+      "epoch": 0.49189613049141656,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018706761452213038,
+      "loss": 4.3603,
+      "step": 4744
+    },
+    {
+      "epoch": 0.49199981854590463,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018706227131602934,
+      "loss": 4.384,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4921035066003927,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018705692708268553,
+      "loss": 4.3522,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4922071946548808,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000187051581822162,
+      "loss": 4.3933,
+      "step": 4747
+    },
+    {
+      "epoch": 0.49231088270936885,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001870462355345218,
+      "loss": 4.3596,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4924145707638569,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018704088821982806,
+      "loss": 4.3722,
+      "step": 4749
+    },
+    {
+      "epoch": 0.492518258818345,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018703553987814382,
+      "loss": 4.388,
+      "step": 4750
+    },
+    {
+      "epoch": 0.49262194687283306,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018703019050953223,
+      "loss": 4.3509,
+      "step": 4751
+    },
+    {
+      "epoch": 0.49272563492732113,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001870248401140564,
+      "loss": 4.3761,
+      "step": 4752
+    },
+    {
+      "epoch": 0.4928293229818092,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018701948869177942,
+      "loss": 4.3907,
+      "step": 4753
+    },
+    {
+      "epoch": 0.4929330110362973,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018701413624276446,
+      "loss": 4.3387,
+      "step": 4754
+    },
+    {
+      "epoch": 0.49303669909078535,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001870087827670747,
+      "loss": 4.3632,
+      "step": 4755
+    },
+    {
+      "epoch": 0.4931403871452734,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018700342826477324,
+      "loss": 4.3249,
+      "step": 4756
+    },
+    {
+      "epoch": 0.49324407519976154,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001869980727359233,
+      "loss": 4.3865,
+      "step": 4757
+    },
+    {
+      "epoch": 0.4933477632542496,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001869927161805881,
+      "loss": 4.3384,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4934514513087377,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018698735859883076,
+      "loss": 4.3589,
+      "step": 4759
+    },
+    {
+      "epoch": 0.49355513936322576,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018698199999071455,
+      "loss": 4.388,
+      "step": 4760
+    },
+    {
+      "epoch": 0.49365882741771383,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001869766403563027,
+      "loss": 4.3556,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4937625154722019,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001869712796956584,
+      "loss": 4.3647,
+      "step": 4762
+    },
+    {
+      "epoch": 0.49386620352669,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018696591800884495,
+      "loss": 4.3689,
+      "step": 4763
+    },
+    {
+      "epoch": 0.49396989158117804,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018696055529592557,
+      "loss": 4.3974,
+      "step": 4764
+    },
+    {
+      "epoch": 0.4940735796356661,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001869551915569636,
+      "loss": 4.359,
+      "step": 4765
+    },
+    {
+      "epoch": 0.4941772676901542,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018694982679202227,
+      "loss": 4.3794,
+      "step": 4766
+    },
+    {
+      "epoch": 0.49428095574464226,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001869444610011649,
+      "loss": 4.3763,
+      "step": 4767
+    },
+    {
+      "epoch": 0.49438464379913033,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018693909418445478,
+      "loss": 4.3788,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4944883318536184,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018693372634195527,
+      "loss": 4.3416,
+      "step": 4769
+    },
+    {
+      "epoch": 0.4945920199081065,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018692835747372965,
+      "loss": 4.3486,
+      "step": 4770
+    },
+    {
+      "epoch": 0.49469570796259454,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018692298757984132,
+      "loss": 4.3836,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4947993960170826,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001869176166603536,
+      "loss": 4.3463,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4949030840715707,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018691224471532988,
+      "loss": 4.3993,
+      "step": 4773
+    },
+    {
+      "epoch": 0.49500677212605876,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001869068717448336,
+      "loss": 4.35,
+      "step": 4774
+    },
+    {
+      "epoch": 0.49511046018054683,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018690149774892802,
+      "loss": 4.3844,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4952141482350349,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018689612272767666,
+      "loss": 4.3631,
+      "step": 4776
+    },
+    {
+      "epoch": 0.495317836289523,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018689074668114294,
+      "loss": 4.3669,
+      "step": 4777
+    },
+    {
+      "epoch": 0.49542152434401104,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018688536960939018,
+      "loss": 4.3282,
+      "step": 4778
+    },
+    {
+      "epoch": 0.4955252123984991,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018687999151248197,
+      "loss": 4.3554,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4956289004529872,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018687461239048165,
+      "loss": 4.3386,
+      "step": 4780
+    },
+    {
+      "epoch": 0.49573258850747526,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018686923224345273,
+      "loss": 4.3446,
+      "step": 4781
+    },
+    {
+      "epoch": 0.49583627656196333,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018686385107145875,
+      "loss": 4.3566,
+      "step": 4782
+    },
+    {
+      "epoch": 0.4959399646164514,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001868584688745631,
+      "loss": 4.3457,
+      "step": 4783
+    },
+    {
+      "epoch": 0.49604365267093947,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018685308565282935,
+      "loss": 4.3423,
+      "step": 4784
+    },
+    {
+      "epoch": 0.49614734072542754,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018684770140632099,
+      "loss": 4.3752,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4962510287799156,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018684231613510152,
+      "loss": 4.4074,
+      "step": 4786
+    },
+    {
+      "epoch": 0.4963547168344037,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018683692983923456,
+      "loss": 4.3825,
+      "step": 4787
+    },
+    {
+      "epoch": 0.49645840488889176,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001868315425187836,
+      "loss": 4.3474,
+      "step": 4788
+    },
+    {
+      "epoch": 0.49656209294337983,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018682615417381224,
+      "loss": 4.3531,
+      "step": 4789
+    },
+    {
+      "epoch": 0.4966657809978679,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018682076480438405,
+      "loss": 4.3531,
+      "step": 4790
+    },
+    {
+      "epoch": 0.49676946905235597,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001868153744105626,
+      "loss": 4.3875,
+      "step": 4791
+    },
+    {
+      "epoch": 0.49687315710684404,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018680998299241147,
+      "loss": 4.3805,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4969768451613321,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018680459054999432,
+      "loss": 4.3138,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4970805332158202,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018679919708337476,
+      "loss": 4.3827,
+      "step": 4794
+    },
+    {
+      "epoch": 0.49718422127030826,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018679380259261646,
+      "loss": 4.3639,
+      "step": 4795
+    },
+    {
+      "epoch": 0.49728790932479633,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018678840707778298,
+      "loss": 4.391,
+      "step": 4796
+    },
+    {
+      "epoch": 0.4973915973792844,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018678301053893804,
+      "loss": 4.3497,
+      "step": 4797
+    },
+    {
+      "epoch": 0.49749528543377247,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018677761297614532,
+      "loss": 4.4161,
+      "step": 4798
+    },
+    {
+      "epoch": 0.49759897348826054,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018677221438946853,
+      "loss": 4.3588,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4977026615427486,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001867668147789713,
+      "loss": 4.3636,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4978063495972367,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018676141414471738,
+      "loss": 4.3342,
+      "step": 4801
+    },
+    {
+      "epoch": 0.4979100376517248,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001867560124867705,
+      "loss": 4.3903,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4980137257062129,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001867506098051943,
+      "loss": 4.3743,
+      "step": 4803
+    },
+    {
+      "epoch": 0.49811741376070096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001867452061000527,
+      "loss": 4.3445,
+      "step": 4804
+    },
+    {
+      "epoch": 0.498221101815189,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001867398013714093,
+      "loss": 4.3474,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4983247898696771,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018673439561932797,
+      "loss": 4.3979,
+      "step": 4806
+    },
+    {
+      "epoch": 0.49842847792416517,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018672898884387244,
+      "loss": 4.3501,
+      "step": 4807
+    },
+    {
+      "epoch": 0.49853216597865324,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001867235810451065,
+      "loss": 4.4264,
+      "step": 4808
+    },
+    {
+      "epoch": 0.4986358540331413,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.000186718172223094,
+      "loss": 4.4047,
+      "step": 4809
+    },
+    {
+      "epoch": 0.4987395420876294,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018671276237789872,
+      "loss": 4.3913,
+      "step": 4810
+    },
+    {
+      "epoch": 0.49884323014211746,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018670735150958453,
+      "loss": 4.3599,
+      "step": 4811
+    },
+    {
+      "epoch": 0.4989469181966055,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001867019396182152,
+      "loss": 4.3785,
+      "step": 4812
+    },
+    {
+      "epoch": 0.4990506062510936,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018669652670385463,
+      "loss": 4.3586,
+      "step": 4813
+    },
+    {
+      "epoch": 0.49915429430558167,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018669111276656673,
+      "loss": 4.3503,
+      "step": 4814
+    },
+    {
+      "epoch": 0.49925798236006974,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00018668569780641532,
+      "loss": 4.3362,
+      "step": 4815
+    },
+    {
+      "epoch": 0.4993616704145578,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001866802818234643,
+      "loss": 4.3846,
+      "step": 4816
+    },
+    {
+      "epoch": 0.4994653584690459,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001866748648177776,
+      "loss": 4.3617,
+      "step": 4817
+    },
+    {
+      "epoch": 0.49956904652353396,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018666944678941909,
+      "loss": 4.3697,
+      "step": 4818
+    },
+    {
+      "epoch": 0.499672734578022,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018666402773845272,
+      "loss": 4.3483,
+      "step": 4819
+    },
+    {
+      "epoch": 0.4997764226325101,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018665860766494245,
+      "loss": 4.399,
+      "step": 4820
+    },
+    {
+      "epoch": 0.49988011068699817,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00018665318656895219,
+      "loss": 4.4077,
+      "step": 4821
+    },
+    {
+      "epoch": 0.49998379874148624,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018664776445054596,
+      "loss": 4.4045,
+      "step": 4822
+    },
+    {
+      "epoch": 0.49998379874148624,
+      "eval_loss": 4.3779425621032715,
+      "eval_runtime": 0.4398,
+      "eval_samples_per_second": 338.822,
+      "eval_steps_per_second": 15.918,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5000874867959743,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018664234130978766,
+      "loss": 4.3838,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5001911748504624,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00018663691714674132,
+      "loss": 4.3242,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5002948629049505,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018663149196147095,
+      "loss": 4.3391,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5003985509594385,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018662606575404054,
+      "loss": 4.3059,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5005022390139267,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00018662063852451413,
+      "loss": 4.3825,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5006059270684147,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018661521027295573,
+      "loss": 4.3923,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5007096151229028,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00018660978099942945,
+      "loss": 4.3616,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5008133031773908,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018660435070399925,
+      "loss": 4.3542,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5009169912318789,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001865989193867293,
+      "loss": 4.4224,
+      "step": 4831
+    },
+    {
+      "epoch": 0.501020679286367,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018659348704768363,
+      "loss": 4.3345,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5011243673408551,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018658805368692636,
+      "loss": 4.3521,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5012280553953431,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018658261930452153,
+      "loss": 4.4144,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5013317434498312,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018657718390053336,
+      "loss": 4.378,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5014354315043192,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018657174747502593,
+      "loss": 4.318,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5015391195588074,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018656631002806338,
+      "loss": 4.3585,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5016428076132954,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001865608715597099,
+      "loss": 4.3919,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5017464956677835,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018655543207002958,
+      "loss": 4.3366,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5018501837222715,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001865499915590867,
+      "loss": 4.3427,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5019538717767597,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001865445500269454,
+      "loss": 4.3433,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5020575598312477,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018653910747366988,
+      "loss": 4.393,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5021612478857358,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001865336638993244,
+      "loss": 4.3742,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5022649359402238,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001865282193039731,
+      "loss": 4.3722,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5023686239947119,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018652277368768033,
+      "loss": 4.3399,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5024723120492,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018651732705051024,
+      "loss": 4.3586,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5025760001036881,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018651187939252716,
+      "loss": 4.37,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5026796881581761,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018650643071379538,
+      "loss": 4.3741,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5027833762126642,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018650098101437914,
+      "loss": 4.3927,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5028870642671522,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018649553029434274,
+      "loss": 4.3703,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5029907523216404,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018649007855375054,
+      "loss": 4.3834,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5030944403761284,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018648462579266684,
+      "loss": 4.3607,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5031981284306165,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018647917201115597,
+      "loss": 4.3626,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5033018164851045,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018647371720928227,
+      "loss": 4.3501,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5034055045395927,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001864682613871101,
+      "loss": 4.2801,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5035091925940807,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001864628045447039,
+      "loss": 4.3684,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5036128806485688,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018645734668212795,
+      "loss": 4.3729,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5037165687030568,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018645188779944672,
+      "loss": 4.3496,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5038202567575449,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018644642789672456,
+      "loss": 4.3412,
+      "step": 4859
+    },
+    {
+      "epoch": 0.503923944812033,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018644096697402598,
+      "loss": 4.3277,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5040276328665211,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018643550503141533,
+      "loss": 4.412,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5041313209210091,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001864300420689571,
+      "loss": 4.3538,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5042350089754972,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018642457808671573,
+      "loss": 4.3333,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5043386970299852,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00018641911308475573,
+      "loss": 4.3947,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5044423850844734,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018641364706314148,
+      "loss": 4.3542,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5045460731389614,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001864081800219376,
+      "loss": 4.3132,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5046497611934495,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018640271196120848,
+      "loss": 4.3881,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5047534492479375,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001863972428810187,
+      "loss": 4.3479,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5048571373024257,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0001863917727814328,
+      "loss": 4.3538,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5049608253569138,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001863863016625153,
+      "loss": 4.4002,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5050645134114018,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00018638082952433072,
+      "loss": 4.3441,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5051682014658899,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001863753563669437,
+      "loss": 4.3473,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5052718895203779,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00018636988219041878,
+      "loss": 4.3742,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5053755775748661,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018636440699482053,
+      "loss": 4.3611,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5054792656293541,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018635893078021356,
+      "loss": 4.3787,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5055829536838422,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018635345354666252,
+      "loss": 4.3462,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5056866417383302,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000186347975294232,
+      "loss": 4.3807,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5057903297928183,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001863424960229866,
+      "loss": 4.3645,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5058940178473064,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018633701573299107,
+      "loss": 4.3672,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5059977059017945,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018633153442430998,
+      "loss": 4.3876,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5061013939562825,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018632605209700808,
+      "loss": 4.3838,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5062050820107706,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018632056875115,
+      "loss": 4.3444,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5063087700652587,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018631508438680047,
+      "loss": 4.3192,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5064124581197468,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018630959900402412,
+      "loss": 4.3604,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5065161461742348,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001863041126028858,
+      "loss": 4.3894,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5066198342287229,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018629862518345017,
+      "loss": 4.3822,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5067235222832109,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018629313674578196,
+      "loss": 4.3275,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5068272103376991,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018628764728994594,
+      "loss": 4.3346,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5069308983921871,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018628215681600692,
+      "loss": 4.3965,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5070345864466752,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018627666532402962,
+      "loss": 4.3549,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5071382745011632,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018627117281407892,
+      "loss": 4.3535,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5072419625556513,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018626567928621955,
+      "loss": 4.3435,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5073456506101394,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018626018474051634,
+      "loss": 4.295,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5074493386646275,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018625468917703414,
+      "loss": 4.3831,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5075530267191155,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001862491925958378,
+      "loss": 4.3633,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5076567147736036,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001862436949969921,
+      "loss": 4.3539,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5077604028280917,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018623819638056204,
+      "loss": 4.3631,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5078640908825798,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018623269674661238,
+      "loss": 4.3494,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5079677789370678,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00018622719609520804,
+      "loss": 4.3768,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5080714669915559,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018622169442641395,
+      "loss": 4.3804,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5081751550460439,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.000186216191740295,
+      "loss": 4.377,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5082788431005321,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018621068803691612,
+      "loss": 4.3685,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5083825311550201,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018620518331634225,
+      "loss": 4.3558,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5084862192095082,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018619967757863836,
+      "loss": 4.3696,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5085899072639962,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018619417082386936,
+      "loss": 4.3299,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5086935953184843,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018618866305210029,
+      "loss": 4.3453,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5087972833729724,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018618315426339606,
+      "loss": 4.373,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5089009714274605,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018617764445782173,
+      "loss": 4.374,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5090046594819485,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018617213363544228,
+      "loss": 4.3881,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5091083475364366,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018616662179632277,
+      "loss": 4.3841,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5092120355909246,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018616110894052818,
+      "loss": 4.3558,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5093157236454128,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001861555950681236,
+      "loss": 4.2965,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5094194116999009,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018615008017917406,
+      "loss": 4.3539,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5095230997543889,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018614456427374463,
+      "loss": 4.3157,
+      "step": 4914
+    },
+    {
+      "epoch": 0.509626787808877,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018613904735190038,
+      "loss": 4.3303,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5097304758633651,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018613352941370645,
+      "loss": 4.3523,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5098341639178532,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00018612801045922792,
+      "loss": 4.3532,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5099378519723412,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001861224904885299,
+      "loss": 4.3548,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5100415400268293,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00018611696950167754,
+      "loss": 4.3741,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5101452280813173,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018611144749873594,
+      "loss": 4.3321,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5102489161358055,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018610592447977035,
+      "loss": 4.387,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5103526041902935,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0001861004004448458,
+      "loss": 4.364,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5104562922447816,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018609487539402757,
+      "loss": 4.3479,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5105599802992696,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001860893493273808,
+      "loss": 4.3506,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5106636683537578,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001860838222449707,
+      "loss": 4.3584,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5107673564082458,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018607829414686252,
+      "loss": 4.3778,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5108710444627339,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018607276503312145,
+      "loss": 4.3566,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5109747325172219,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018606723490381273,
+      "loss": 4.3141,
+      "step": 4928
+    },
+    {
+      "epoch": 0.51107842057171,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001860617037590016,
+      "loss": 4.3806,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018605617159875336,
+      "loss": 4.3295,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5112857966806862,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00018605063842313326,
+      "loss": 4.3544,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5113894847351742,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018604510423220658,
+      "loss": 4.3995,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5114931727896623,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018603956902603862,
+      "loss": 4.353,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5115968608441503,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018603403280469469,
+      "loss": 4.3788,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5117005488986385,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018602849556824015,
+      "loss": 4.322,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5118042369531265,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018602295731674025,
+      "loss": 4.3518,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5119079250076146,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018601741805026042,
+      "loss": 4.3241,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5120116130621026,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018601187776886595,
+      "loss": 4.3811,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5121153011165908,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018600633647262226,
+      "loss": 4.347,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5122189891710788,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018600079416159472,
+      "loss": 4.3929,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5123226772255669,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001859952508358487,
+      "loss": 4.3574,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5124263652800549,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001859897064954496,
+      "loss": 4.3482,
+      "step": 4942
+    },
+    {
+      "epoch": 0.512530053334543,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001859841611404629,
+      "loss": 4.3844,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5126337413890311,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018597861477095398,
+      "loss": 4.3551,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5127374294435192,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018597306738698827,
+      "loss": 4.3954,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5128411174980072,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00018596751898863128,
+      "loss": 4.3693,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5129448055524953,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018596196957594837,
+      "loss": 4.4051,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5130484936069833,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018595641914900514,
+      "loss": 4.3488,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5131521816614715,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.000185950867707867,
+      "loss": 4.3103,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5132558697159595,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001859453152525995,
+      "loss": 4.3381,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5133595577704476,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001859397617832681,
+      "loss": 4.3591,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5134632458249356,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018593420729993838,
+      "loss": 4.3626,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5135669338794238,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018592865180267582,
+      "loss": 4.3803,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5136706219339118,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.000185923095291546,
+      "loss": 4.3524,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5137743099883999,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001859175377666145,
+      "loss": 4.3688,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5138779980428879,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001859119792279469,
+      "loss": 4.346,
+      "step": 4956
+    },
+    {
+      "epoch": 0.513981686097376,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018590641967560872,
+      "loss": 4.4135,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5140853741518642,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001859008591096656,
+      "loss": 4.3789,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5141890622063522,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018589529753018313,
+      "loss": 4.3745,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5142927502608403,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018588973493722697,
+      "loss": 4.3575,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5143964383153283,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001858841713308627,
+      "loss": 4.3291,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5145001263698165,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018587860671115603,
+      "loss": 4.3675,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5146038144243045,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018587304107817255,
+      "loss": 4.3248,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5147075024787926,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018586747443197796,
+      "loss": 4.3748,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5148111905332806,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00018586190677263793,
+      "loss": 4.3217,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5149148785877687,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018585633810021818,
+      "loss": 4.3518,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5150185666422568,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018585076841478438,
+      "loss": 4.3605,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5151222546967449,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018584519771640227,
+      "loss": 4.3512,
+      "step": 4968
+    },
+    {
+      "epoch": 0.5152259427512329,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018583962600513755,
+      "loss": 4.3504,
+      "step": 4969
+    },
+    {
+      "epoch": 0.515329630805721,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.000185834053281056,
+      "loss": 4.3637,
+      "step": 4970
+    },
+    {
+      "epoch": 0.515433318860209,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018582847954422337,
+      "loss": 4.4037,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5155370069146972,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018582290479470537,
+      "loss": 4.3579,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5156406949691852,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001858173290325678,
+      "loss": 4.3613,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5157443830236733,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00018581175225787652,
+      "loss": 4.3607,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5158480710781613,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018580617447069722,
+      "loss": 4.3722,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5159517591326495,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018580059567109577,
+      "loss": 4.4164,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5160554471871375,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018579501585913802,
+      "loss": 4.3352,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5161591352416256,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018578943503488974,
+      "loss": 4.3437,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5162628232961136,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001857838531984168,
+      "loss": 4.3684,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5163665113506017,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001857782703497851,
+      "loss": 4.3773,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5164701994050898,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00018577268648906046,
+      "loss": 4.3865,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5165738874595779,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001857671016163088,
+      "loss": 4.3972,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5166775755140659,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.000185761515731596,
+      "loss": 4.3545,
+      "step": 4983
+    },
+    {
+      "epoch": 0.516781263568554,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018575592883498794,
+      "loss": 4.3675,
+      "step": 4984
+    },
+    {
+      "epoch": 0.516884951623042,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001857503409265506,
+      "loss": 4.3399,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5169886396775302,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018574475200634985,
+      "loss": 4.3809,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5170923277320182,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001857391620744517,
+      "loss": 4.3294,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5171960157865063,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018573357113092203,
+      "loss": 4.3773,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5172997038409943,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001857279791758269,
+      "loss": 4.3379,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5174033918954825,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018572238620923217,
+      "loss": 4.3642,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5175070799499705,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018571679223120392,
+      "loss": 4.3593,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5176107680044586,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018571119724180814,
+      "loss": 4.3882,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5177144560589466,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018570560124111084,
+      "loss": 4.3503,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5178181441134347,
+      "grad_norm": 0.75,
+      "learning_rate": 0.000185700004229178,
+      "loss": 4.3396,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5179218321679228,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018569440620607572,
+      "loss": 4.3554,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5180255202224109,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018568880717187005,
+      "loss": 4.3116,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5181292082768989,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.000185683207126627,
+      "loss": 4.3576,
+      "step": 4997
+    },
+    {
+      "epoch": 0.518232896331387,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001856776060704127,
+      "loss": 4.375,
+      "step": 4998
+    },
+    {
+      "epoch": 0.518336584385875,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018567200400329317,
+      "loss": 4.3738,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5184402724403632,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001856664009253346,
+      "loss": 4.3946,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5185439604948512,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018566079683660306,
+      "loss": 4.3456,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5186476485493393,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018565519173716465,
+      "loss": 4.3619,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5187513366038274,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001856495856270855,
+      "loss": 4.3406,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5188550246583155,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018564397850643182,
+      "loss": 4.37,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5189587127128036,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018563837037526967,
+      "loss": 4.367,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5190624007672916,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001856327612336653,
+      "loss": 4.3462,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5191660888217797,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018562715108168485,
+      "loss": 4.3543,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5192697768762677,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018562153991939453,
+      "loss": 4.3434,
+      "step": 5008
+    },
+    {
+      "epoch": 0.5193734649307559,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018561592774686057,
+      "loss": 4.3605,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5194771529852439,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018561031456414915,
+      "loss": 4.3769,
+      "step": 5010
+    },
+    {
+      "epoch": 0.519580841039732,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018560470037132652,
+      "loss": 4.3688,
+      "step": 5011
+    },
+    {
+      "epoch": 0.51968452909422,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001855990851684589,
+      "loss": 4.3712,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5197882171487082,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018559346895561253,
+      "loss": 4.3923,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5198919052031962,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018558785173285376,
+      "loss": 4.3656,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5199955932576843,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018558223350024876,
+      "loss": 4.3326,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5200992813121723,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001855766142578639,
+      "loss": 4.3647,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5202029693666604,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018557099400576545,
+      "loss": 4.392,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5203066574211485,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001855653727440197,
+      "loss": 4.3832,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5204103454756366,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.000185559750472693,
+      "loss": 4.3736,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5205140335301246,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018555412719185172,
+      "loss": 4.2913,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5206177215846127,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00018554850290156214,
+      "loss": 4.3555,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5207214096391007,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00018554287760189067,
+      "loss": 4.3643,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5208250976935889,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018553725129290366,
+      "loss": 4.37,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5209287857480769,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001855316239746675,
+      "loss": 4.3919,
+      "step": 5024
+    },
+    {
+      "epoch": 0.521032473802565,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001855259956472486,
+      "loss": 4.3377,
+      "step": 5025
+    },
+    {
+      "epoch": 0.521136161857053,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018552036631071334,
+      "loss": 4.3495,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5212398499115412,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018551473596512818,
+      "loss": 4.3484,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5213435379660292,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018550910461055952,
+      "loss": 4.3541,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5214472260205173,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018550347224707378,
+      "loss": 4.3259,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5215509140750053,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018549783887473749,
+      "loss": 4.3419,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5216546021294934,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018549220449361707,
+      "loss": 4.3365,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5217582901839815,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000185486569103779,
+      "loss": 4.3261,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5218619782384696,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018548093270528976,
+      "loss": 4.3539,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5219656662929576,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001854752952982159,
+      "loss": 4.3635,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5220693543474457,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018546965688262388,
+      "loss": 4.3569,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5221730424019337,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018546401745858027,
+      "loss": 4.3369,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5222767304564219,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001854583770261516,
+      "loss": 4.3832,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5223804185109099,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001854527355854044,
+      "loss": 4.3824,
+      "step": 5038
+    },
+    {
+      "epoch": 0.522484106565398,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018544709313640526,
+      "loss": 4.4,
+      "step": 5039
+    },
+    {
+      "epoch": 0.522587794619886,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018544144967922073,
+      "loss": 4.3677,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5226914826743742,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001854358052139174,
+      "loss": 4.327,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5227951707288622,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001854301597405619,
+      "loss": 4.4058,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5228988587833503,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001854245132592208,
+      "loss": 4.3692,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5230025468378383,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018541886576996076,
+      "loss": 4.4042,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5231062348923264,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001854132172728484,
+      "loss": 4.3645,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5232099229468145,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018540756776795034,
+      "loss": 4.3312,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5233136110013026,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018540191725533326,
+      "loss": 4.3549,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5234172990557907,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018539626573506386,
+      "loss": 4.3363,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5235209871102787,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018539061320720874,
+      "loss": 4.3412,
+      "step": 5049
+    },
+    {
+      "epoch": 0.5236246751647669,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001853849596718347,
+      "loss": 4.3746,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5237283632192549,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018537930512900835,
+      "loss": 4.343,
+      "step": 5051
+    },
+    {
+      "epoch": 0.523832051273743,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018537364957879648,
+      "loss": 4.3576,
+      "step": 5052
+    },
+    {
+      "epoch": 0.523935739328231,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018536799302126578,
+      "loss": 4.3926,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5240394273827191,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018536233545648302,
+      "loss": 4.3628,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5241431154372072,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018535667688451487,
+      "loss": 4.3296,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5242468034916953,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018535101730542822,
+      "loss": 4.3907,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5243504915461833,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018534535671928978,
+      "loss": 4.3648,
+      "step": 5057
+    },
+    {
+      "epoch": 0.5244541796006714,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018533969512616634,
+      "loss": 4.3272,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5245578676551594,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018533403252612467,
+      "loss": 4.3601,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5246615557096476,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018532836891923166,
+      "loss": 4.3732,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5247652437641356,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018532270430555406,
+      "loss": 4.372,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5248689318186237,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00018531703868515878,
+      "loss": 4.3324,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5249726198731117,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00018531137205811258,
+      "loss": 4.3788,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5250763079275999,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001853057044244824,
+      "loss": 4.3657,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5251799959820879,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018530003578433508,
+      "loss": 4.331,
+      "step": 5065
+    },
+    {
+      "epoch": 0.525283684036576,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001852943661377375,
+      "loss": 4.4065,
+      "step": 5066
+    },
+    {
+      "epoch": 0.525387372091064,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00018528869548475654,
+      "loss": 4.3171,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5254910601455521,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001852830238254591,
+      "loss": 4.35,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5255947482000402,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001852773511599122,
+      "loss": 4.3805,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5256984362545283,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00018527167748818267,
+      "loss": 4.3788,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5258021243090163,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018526600281033748,
+      "loss": 4.3321,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5259058123635044,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018526032712644358,
+      "loss": 4.4017,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5260095004179924,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018525465043656795,
+      "loss": 4.335,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5261131884724806,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018524897274077757,
+      "loss": 4.3608,
+      "step": 5074
+    },
+    {
+      "epoch": 0.5262168765269686,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001852432940391394,
+      "loss": 4.2927,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5263205645814567,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001852376143317205,
+      "loss": 4.366,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5264242526359447,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018523193361858785,
+      "loss": 4.3713,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5265279406904328,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018522625189980844,
+      "loss": 4.3462,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5266316287449209,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018522056917544939,
+      "loss": 4.3247,
+      "step": 5079
+    },
+    {
+      "epoch": 0.526735316799409,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018521488544557765,
+      "loss": 4.3743,
+      "step": 5080
+    },
+    {
+      "epoch": 0.526839004853897,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018520920071026038,
+      "loss": 4.4079,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5269426929083851,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001852035149695646,
+      "loss": 4.3398,
+      "step": 5082
+    },
+    {
+      "epoch": 0.5270463809628732,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018519782822355743,
+      "loss": 4.3662,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5271500690173613,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00018519214047230592,
+      "loss": 4.3139,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5272537570718493,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018518645171587719,
+      "loss": 4.3445,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5273574451263374,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001851807619543384,
+      "loss": 4.3736,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5274611331808254,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018517507118775666,
+      "loss": 4.3472,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5275648212353136,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001851693794161991,
+      "loss": 4.3841,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5276685092898016,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001851636866397329,
+      "loss": 4.3624,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5277721973442897,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00018515799285842522,
+      "loss": 4.3591,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5278758853987777,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00018515229807234323,
+      "loss": 4.3731,
+      "step": 5091
+    },
+    {
+      "epoch": 0.5279795734532658,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018514660228155413,
+      "loss": 4.3403,
+      "step": 5092
+    },
+    {
+      "epoch": 0.528083261507754,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018514090548612516,
+      "loss": 4.3769,
+      "step": 5093
+    },
+    {
+      "epoch": 0.528186949562242,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018513520768612346,
+      "loss": 4.315,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5282906376167301,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018512950888161634,
+      "loss": 4.3687,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5283943256712181,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00018512380907267098,
+      "loss": 4.3596,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5284980137257063,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00018511810825935467,
+      "loss": 4.3683,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5286017017801943,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018511240644173462,
+      "loss": 4.3502,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5287053898346824,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00018510670361987817,
+      "loss": 4.3894,
+      "step": 5099
+    },
+    {
+      "epoch": 0.5288090778891704,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018510099979385255,
+      "loss": 4.4088,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5289127659436585,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00018509529496372513,
+      "loss": 4.3636,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5290164539981466,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018508958912956316,
+      "loss": 4.3658,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5291201420526347,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018508388229143398,
+      "loss": 4.3711,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5292238301071227,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001850781744494049,
+      "loss": 4.3279,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5293275181616108,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018507246560354334,
+      "loss": 4.3678,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5294312062160988,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00018506675575391662,
+      "loss": 4.3717,
+      "step": 5106
+    },
+    {
+      "epoch": 0.529534894270587,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018506104490059207,
+      "loss": 4.3604,
+      "step": 5107
+    },
+    {
+      "epoch": 0.529638582325075,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001850553330436371,
+      "loss": 4.3514,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5297422703795631,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018504962018311912,
+      "loss": 4.3815,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5298459584340511,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001850439063191055,
+      "loss": 4.3414,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5299496464885393,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001850381914516637,
+      "loss": 4.3528,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5300533345430273,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018503247558086114,
+      "loss": 4.3303,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5301570225975154,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00018502675870676522,
+      "loss": 4.3694,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5302607106520034,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018502104082944345,
+      "loss": 4.339,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5303643987064915,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018501532194896327,
+      "loss": 4.3463,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5304680867609796,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018500960206539215,
+      "loss": 4.331,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5305717748154677,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018500388117879756,
+      "loss": 4.3701,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5306754628699557,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001849981592892471,
+      "loss": 4.3665,
+      "step": 5118
+    },
+    {
+      "epoch": 0.5307791509244438,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018499243639680813,
+      "loss": 4.3249,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5308828389789318,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018498671250154826,
+      "loss": 4.3709,
+      "step": 5120
+    },
+    {
+      "epoch": 0.53098652703342,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018498098760353507,
+      "loss": 4.3713,
+      "step": 5121
+    },
+    {
+      "epoch": 0.531090215087908,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.000184975261702836,
+      "loss": 4.3346,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5311939031423961,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018496953479951869,
+      "loss": 4.3565,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5312975911968841,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001849638068936507,
+      "loss": 4.3247,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5314012792513723,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018495807798529957,
+      "loss": 4.3497,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5315049673058603,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018495234807453293,
+      "loss": 4.4022,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5316086553603484,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001849466171614184,
+      "loss": 4.3508,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5317123434148364,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018494088524602354,
+      "loss": 4.3337,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5318160314693245,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018493515232841605,
+      "loss": 4.37,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5319197195238126,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018492941840866353,
+      "loss": 4.3041,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5320234075783007,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018492368348683368,
+      "loss": 4.3602,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5321270956327887,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001849179475629941,
+      "loss": 4.349,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5322307836872768,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001849122106372125,
+      "loss": 4.3643,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5323344717417648,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001849064727095566,
+      "loss": 4.3504,
+      "step": 5134
+    },
+    {
+      "epoch": 0.532438159796253,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018490073378009405,
+      "loss": 4.4057,
+      "step": 5135
+    },
+    {
+      "epoch": 0.532541847850741,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001848949938488926,
+      "loss": 4.3717,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5326455359052291,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00018488925291601995,
+      "loss": 4.3567,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5327492239597172,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018488351098154383,
+      "loss": 4.3428,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5328529120142053,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018487776804553203,
+      "loss": 4.3578,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5329566000686934,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001848720241080523,
+      "loss": 4.3299,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5330602881231814,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018486627916917235,
+      "loss": 4.334,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5331639761776695,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018486053322896005,
+      "loss": 4.3096,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5332676642321575,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018485478628748315,
+      "loss": 4.3155,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5333713522866457,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018484903834480946,
+      "loss": 4.3713,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5334750403411337,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001848432894010068,
+      "loss": 4.3835,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5335787283956218,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018483753945614304,
+      "loss": 4.415,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5336824164501098,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018483178851028597,
+      "loss": 4.3042,
+      "step": 5147
+    },
+    {
+      "epoch": 0.533786104504598,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018482603656350347,
+      "loss": 4.3036,
+      "step": 5148
+    },
+    {
+      "epoch": 0.533889792559086,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018482028361586342,
+      "loss": 4.3833,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5339934806135741,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018481452966743363,
+      "loss": 4.338,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5340971686680621,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001848087747182821,
+      "loss": 4.3212,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5342008567225502,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018480301876847667,
+      "loss": 4.3137,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5343045447770383,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018479726181808522,
+      "loss": 4.354,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5344082328315264,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018479150386717575,
+      "loss": 4.3564,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5345119208860144,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018478574491581618,
+      "loss": 4.3444,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5346156089405025,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018477998496407437,
+      "loss": 4.381,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5347192969949905,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018477422401201842,
+      "loss": 4.3283,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5348229850494787,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018476846205971623,
+      "loss": 4.3069,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5349266731039667,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018476269910723577,
+      "loss": 4.3874,
+      "step": 5159
+    },
+    {
+      "epoch": 0.5350303611584548,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018475693515464508,
+      "loss": 4.3665,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5351340492129428,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001847511702020121,
+      "loss": 4.3524,
+      "step": 5161
+    },
+    {
+      "epoch": 0.535237737267431,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00018474540424940497,
+      "loss": 4.4009,
+      "step": 5162
+    },
+    {
+      "epoch": 0.535341425321919,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001847396372968916,
+      "loss": 4.3042,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5354451133764071,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001847338693445401,
+      "loss": 4.3491,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5355488014308951,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018472810039241847,
+      "loss": 4.3576,
+      "step": 5165
+    },
+    {
+      "epoch": 0.5356524894853832,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018472233044059485,
+      "loss": 4.3091,
+      "step": 5166
+    },
+    {
+      "epoch": 0.5357561775398713,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001847165594891373,
+      "loss": 4.368,
+      "step": 5167
+    },
+    {
+      "epoch": 0.5358598655943594,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018471078753811392,
+      "loss": 4.3373,
+      "step": 5168
+    },
+    {
+      "epoch": 0.5359635536488474,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018470501458759273,
+      "loss": 4.3089,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5360672417033355,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00018469924063764192,
+      "loss": 4.3848,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5361709297578235,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001846934656883296,
+      "loss": 4.3728,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5362746178123117,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018468768973972392,
+      "loss": 4.3487,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5363783058667997,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018468191279189303,
+      "loss": 4.3687,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5364819939212878,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018467613484490506,
+      "loss": 4.3389,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5365856819757758,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001846703558988282,
+      "loss": 4.4002,
+      "step": 5175
+    },
+    {
+      "epoch": 0.536689370030264,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001846645759537307,
+      "loss": 4.3487,
+      "step": 5176
+    },
+    {
+      "epoch": 0.536793058084752,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018465879500968065,
+      "loss": 4.3517,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5368967461392401,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018465301306674632,
+      "loss": 4.3955,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5370004341937281,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001846472301249959,
+      "loss": 4.3555,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5371041222482162,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018464144618449764,
+      "loss": 4.2937,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5372078103027044,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001846356612453198,
+      "loss": 4.3372,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5373114983571924,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018462987530753062,
+      "loss": 4.3874,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5374151864116805,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018462408837119837,
+      "loss": 4.3692,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5375188744661685,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018461830043639131,
+      "loss": 4.3388,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5376225625206567,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018461251150317777,
+      "loss": 4.3781,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5377262505751447,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018460672157162602,
+      "loss": 4.3331,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5378299386296328,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001846009306418044,
+      "loss": 4.3527,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5379336266841208,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018459513871378126,
+      "loss": 4.3519,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5380373147386089,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018458934578762483,
+      "loss": 4.3211,
+      "step": 5189
+    },
+    {
+      "epoch": 0.538141002793097,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001845835518634036,
+      "loss": 4.3603,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5382446908475851,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018457775694118582,
+      "loss": 4.368,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5383483789020731,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018457196102103992,
+      "loss": 4.366,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5384520669565612,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001845661641030343,
+      "loss": 4.3691,
+      "step": 5193
+    },
+    {
+      "epoch": 0.5385557550110492,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001845603661872373,
+      "loss": 4.348,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5386594430655374,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018455456727371742,
+      "loss": 4.3388,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5387631311200254,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018454876736254296,
+      "loss": 4.3334,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5388668191745135,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018454296645378245,
+      "loss": 4.3293,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5389705072290015,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018453716454750428,
+      "loss": 4.3296,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5390741952834897,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00018453136164377695,
+      "loss": 4.3375,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5391778833379777,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018452555774266892,
+      "loss": 4.3723,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5392815713924658,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018451975284424862,
+      "loss": 4.3693,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5393852594469538,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001845139469485846,
+      "loss": 4.298,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5394889475014419,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018450814005574532,
+      "loss": 4.3173,
+      "step": 5203
+    },
+    {
+      "epoch": 0.53959263555593,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018450233216579936,
+      "loss": 4.288,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5396963236104181,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018449652327881514,
+      "loss": 4.3271,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5398000116649061,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001844907133948613,
+      "loss": 4.417,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5399036997193942,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018448490251400635,
+      "loss": 4.3407,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5400073877738822,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018447909063631888,
+      "loss": 4.3654,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5401110758283704,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001844732777618674,
+      "loss": 4.3053,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5402147638828584,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018446746389072055,
+      "loss": 4.3725,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5403184519373465,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001844616490229469,
+      "loss": 4.3676,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5404221399918345,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018445583315861508,
+      "loss": 4.3485,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5405258280463227,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001844500162977937,
+      "loss": 4.345,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5406295161008107,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001844441984405514,
+      "loss": 4.3473,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5407332041552988,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018443837958695682,
+      "loss": 4.3451,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5408368922097868,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018443255973707863,
+      "loss": 4.3028,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5409405802642749,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018442673889098546,
+      "loss": 4.3392,
+      "step": 5217
+    },
+    {
+      "epoch": 0.541044268318763,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.000184420917048746,
+      "loss": 4.3519,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5411479563732511,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018441509421042898,
+      "loss": 4.3588,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5412516444277391,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018440927037610306,
+      "loss": 4.3698,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5413553324822272,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.000184403445545837,
+      "loss": 4.3532,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5414590205367152,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018439761971969948,
+      "loss": 4.3166,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5415627085912034,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001843917928977593,
+      "loss": 4.3754,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5416663966456914,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018438596508008512,
+      "loss": 4.3262,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5417700847001795,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018438013626674576,
+      "loss": 4.378,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5418737727546676,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00018437430645781,
+      "loss": 4.3435,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5419774608091557,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00018436847565334662,
+      "loss": 4.368,
+      "step": 5227
+    },
+    {
+      "epoch": 0.5420811488636438,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018436264385342441,
+      "loss": 4.3494,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5421848369181318,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018435681105811216,
+      "loss": 4.3826,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5422885249726199,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001843509772674787,
+      "loss": 4.3497,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5423922130271079,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001843451424815929,
+      "loss": 4.3584,
+      "step": 5231
+    },
+    {
+      "epoch": 0.5424959010815961,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018433930670052354,
+      "loss": 4.3105,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5425995891360841,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018433346992433955,
+      "loss": 4.3521,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5427032771905722,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018432763215310972,
+      "loss": 4.3291,
+      "step": 5234
+    },
+    {
+      "epoch": 0.5428069652450602,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.000184321793386903,
+      "loss": 4.3466,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5429106532995484,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00018431595362578825,
+      "loss": 4.337,
+      "step": 5236
+    },
+    {
+      "epoch": 0.5430143413540364,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018431011286983436,
+      "loss": 4.3597,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5431180294085245,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018430427111911025,
+      "loss": 4.3259,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5432217174630125,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018429842837368486,
+      "loss": 4.357,
+      "step": 5239
+    },
+    {
+      "epoch": 0.5433254055175006,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018429258463362714,
+      "loss": 4.4306,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5434290935719887,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018428673989900603,
+      "loss": 4.3846,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5435327816264768,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018428089416989044,
+      "loss": 4.3188,
+      "step": 5242
+    },
+    {
+      "epoch": 0.5436364696809648,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018427504744634941,
+      "loss": 4.3628,
+      "step": 5243
+    },
+    {
+      "epoch": 0.5437401577354529,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001842691997284519,
+      "loss": 4.338,
+      "step": 5244
+    },
+    {
+      "epoch": 0.5438438457899409,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018426335101626694,
+      "loss": 4.3442,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5439475338444291,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018425750130986347,
+      "loss": 4.3254,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5440512218989171,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018425165060931056,
+      "loss": 4.3504,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5441549099534052,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001842457989146772,
+      "loss": 4.3441,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5442585980078932,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001842399462260325,
+      "loss": 4.3201,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5443622860623814,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001842340925434455,
+      "loss": 4.311,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5444659741168694,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001842282378669852,
+      "loss": 4.4093,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5445696621713575,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018422238219672078,
+      "loss": 4.372,
+      "step": 5252
+    },
+    {
+      "epoch": 0.5446733502258455,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018421652553272125,
+      "loss": 4.3411,
+      "step": 5253
+    },
+    {
+      "epoch": 0.5447770382803336,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018421066787505574,
+      "loss": 4.3217,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5448807263348217,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018420480922379335,
+      "loss": 4.352,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5449844143893098,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001841989495790032,
+      "loss": 4.3746,
+      "step": 5256
+    },
+    {
+      "epoch": 0.5450881024437978,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001841930889407545,
+      "loss": 4.3087,
+      "step": 5257
+    },
+    {
+      "epoch": 0.5451917904982859,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00018418722730911632,
+      "loss": 4.3734,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5452954785527739,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018418136468415785,
+      "loss": 4.3285,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5453991666072621,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018417550106594827,
+      "loss": 4.3549,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5455028546617501,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018416963645455673,
+      "loss": 4.3472,
+      "step": 5261
+    },
+    {
+      "epoch": 0.5456065427162382,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018416377085005248,
+      "loss": 4.3737,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5457102307707262,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018415790425250467,
+      "loss": 4.327,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5458139188252144,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018415203666198257,
+      "loss": 4.3347,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5459176068797024,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018414616807855538,
+      "loss": 4.2823,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5460212949341905,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018414029850229236,
+      "loss": 4.3409,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5461249829886785,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018413442793326277,
+      "loss": 4.3065,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5462286710431666,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018412855637153582,
+      "loss": 4.3402,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5463323590976547,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018412268381718088,
+      "loss": 4.3167,
+      "step": 5269
+    },
+    {
+      "epoch": 0.5464360471521428,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018411681027026715,
+      "loss": 4.3548,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5465397352066309,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.000184110935730864,
+      "loss": 4.34,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5466434232611189,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018410506019904074,
+      "loss": 4.3171,
+      "step": 5272
+    },
+    {
+      "epoch": 0.546747111315607,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018409918367486662,
+      "loss": 4.3388,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5468507993700951,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00018409330615841107,
+      "loss": 4.3309,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5469544874245832,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018408742764974338,
+      "loss": 4.3119,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5470581754790712,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018408154814893296,
+      "loss": 4.3519,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5471618635335593,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018407566765604911,
+      "loss": 4.3591,
+      "step": 5277
+    },
+    {
+      "epoch": 0.5472655515880473,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00018406978617116124,
+      "loss": 4.3208,
+      "step": 5278
+    },
+    {
+      "epoch": 0.5473692396425355,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018406390369433878,
+      "loss": 4.3438,
+      "step": 5279
+    },
+    {
+      "epoch": 0.5474729276970235,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018405802022565114,
+      "loss": 4.3529,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5475766157515116,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00018405213576516772,
+      "loss": 4.3476,
+      "step": 5281
+    },
+    {
+      "epoch": 0.5476803038059996,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001840462503129579,
+      "loss": 4.329,
+      "step": 5282
+    },
+    {
+      "epoch": 0.5477839918604878,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018404036386909117,
+      "loss": 4.3186,
+      "step": 5283
+    },
+    {
+      "epoch": 0.5478876799149758,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00018403447643363703,
+      "loss": 4.3454,
+      "step": 5284
+    },
+    {
+      "epoch": 0.5479913679694639,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018402858800666485,
+      "loss": 4.3464,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5480950560239519,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018402269858824416,
+      "loss": 4.3919,
+      "step": 5286
+    },
+    {
+      "epoch": 0.54819874407844,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00018401680817844445,
+      "loss": 4.3347,
+      "step": 5287
+    },
+    {
+      "epoch": 0.5483024321329281,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018401091677733522,
+      "loss": 4.3367,
+      "step": 5288
+    },
+    {
+      "epoch": 0.5484061201874162,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018400502438498597,
+      "loss": 4.3219,
+      "step": 5289
+    },
+    {
+      "epoch": 0.5485098082419042,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00018399913100146625,
+      "loss": 4.3454,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5486134962963923,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018399323662684553,
+      "loss": 4.3655,
+      "step": 5291
+    },
+    {
+      "epoch": 0.5487171843508803,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018398734126119344,
+      "loss": 4.385,
+      "step": 5292
+    },
+    {
+      "epoch": 0.5488208724053685,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018398144490457946,
+      "loss": 4.3444,
+      "step": 5293
+    },
+    {
+      "epoch": 0.5489245604598565,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018397554755707324,
+      "loss": 4.3266,
+      "step": 5294
+    },
+    {
+      "epoch": 0.5490282485143446,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018396964921874433,
+      "loss": 4.3489,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5491319365688326,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001839637498896623,
+      "loss": 4.2763,
+      "step": 5296
+    },
+    {
+      "epoch": 0.5492356246233208,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018395784956989677,
+      "loss": 4.3416,
+      "step": 5297
+    },
+    {
+      "epoch": 0.5493393126778088,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001839519482595174,
+      "loss": 4.3483,
+      "step": 5298
+    },
+    {
+      "epoch": 0.5494430007322969,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018394604595859372,
+      "loss": 4.3213,
+      "step": 5299
+    },
+    {
+      "epoch": 0.5495466887867849,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018394014266719547,
+      "loss": 4.2955,
+      "step": 5300
+    },
+    {
+      "epoch": 0.549650376841273,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018393423838539225,
+      "loss": 4.3245,
+      "step": 5301
+    },
+    {
+      "epoch": 0.5497540648957611,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018392833311325376,
+      "loss": 4.3408,
+      "step": 5302
+    },
+    {
+      "epoch": 0.5498577529502492,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018392242685084963,
+      "loss": 4.3355,
+      "step": 5303
+    },
+    {
+      "epoch": 0.5499614410047372,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001839165195982496,
+      "loss": 4.3566,
+      "step": 5304
+    },
+    {
+      "epoch": 0.5500651290592253,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018391061135552335,
+      "loss": 4.3398,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5501688171137133,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018390470212274057,
+      "loss": 4.3165,
+      "step": 5306
+    },
+    {
+      "epoch": 0.5502725051682015,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018389879189997097,
+      "loss": 4.353,
+      "step": 5307
+    },
+    {
+      "epoch": 0.5503761932226895,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001838928806872843,
+      "loss": 4.3768,
+      "step": 5308
+    },
+    {
+      "epoch": 0.5504798812771776,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018388696848475034,
+      "loss": 4.3385,
+      "step": 5309
+    },
+    {
+      "epoch": 0.5505835693316656,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018388105529243886,
+      "loss": 4.3456,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5506872573861538,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018387514111041956,
+      "loss": 4.3464,
+      "step": 5311
+    },
+    {
+      "epoch": 0.5507909454406418,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018386922593876224,
+      "loss": 4.3536,
+      "step": 5312
+    },
+    {
+      "epoch": 0.5508946334951299,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018386330977753675,
+      "loss": 4.3569,
+      "step": 5313
+    },
+    {
+      "epoch": 0.5509983215496179,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018385739262681284,
+      "loss": 4.3348,
+      "step": 5314
+    },
+    {
+      "epoch": 0.551102009604106,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018385147448666032,
+      "loss": 4.3592,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5512056976585942,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018384555535714902,
+      "loss": 4.3617,
+      "step": 5316
+    },
+    {
+      "epoch": 0.5513093857130822,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001838396352383488,
+      "loss": 4.3339,
+      "step": 5317
+    },
+    {
+      "epoch": 0.5514130737675703,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018383371413032955,
+      "loss": 4.3743,
+      "step": 5318
+    },
+    {
+      "epoch": 0.5515167618220583,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018382779203316108,
+      "loss": 4.3506,
+      "step": 5319
+    },
+    {
+      "epoch": 0.5516204498765465,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018382186894691326,
+      "loss": 4.3465,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018381594487165597,
+      "loss": 4.3386,
+      "step": 5321
+    },
+    {
+      "epoch": 0.5518278259855226,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00018381001980745916,
+      "loss": 4.3716,
+      "step": 5322
+    },
+    {
+      "epoch": 0.5519315140400106,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00018380409375439268,
+      "loss": 4.3494,
+      "step": 5323
+    },
+    {
+      "epoch": 0.5520352020944987,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001837981667125265,
+      "loss": 4.3477,
+      "step": 5324
+    },
+    {
+      "epoch": 0.5521388901489868,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001837922386819305,
+      "loss": 4.3209,
+      "step": 5325
+    },
+    {
+      "epoch": 0.5522425782034749,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001837863096626747,
+      "loss": 4.3661,
+      "step": 5326
+    },
+    {
+      "epoch": 0.5523462662579629,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018378037965482898,
+      "loss": 4.3566,
+      "step": 5327
+    },
+    {
+      "epoch": 0.552449954312451,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018377444865846337,
+      "loss": 4.3516,
+      "step": 5328
+    },
+    {
+      "epoch": 0.552553642366939,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018376851667364777,
+      "loss": 4.3606,
+      "step": 5329
+    },
+    {
+      "epoch": 0.5526573304214272,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018376258370045228,
+      "loss": 4.3479,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5527610184759152,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001837566497389468,
+      "loss": 4.3693,
+      "step": 5331
+    },
+    {
+      "epoch": 0.5528647065304033,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001837507147892014,
+      "loss": 4.3271,
+      "step": 5332
+    },
+    {
+      "epoch": 0.5529683945848913,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001837447788512861,
+      "loss": 4.3483,
+      "step": 5333
+    },
+    {
+      "epoch": 0.5530720826393795,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018373884192527092,
+      "loss": 4.3756,
+      "step": 5334
+    },
+    {
+      "epoch": 0.5531757706938675,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00018373290401122594,
+      "loss": 4.3084,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5532794587483556,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001837269651092212,
+      "loss": 4.3144,
+      "step": 5336
+    },
+    {
+      "epoch": 0.5533831468028436,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018372102521932676,
+      "loss": 4.3555,
+      "step": 5337
+    },
+    {
+      "epoch": 0.5534868348573317,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001837150843416127,
+      "loss": 4.3477,
+      "step": 5338
+    },
+    {
+      "epoch": 0.5535905229118198,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018370914247614918,
+      "loss": 4.3296,
+      "step": 5339
+    },
+    {
+      "epoch": 0.5536942109663079,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001837031996230062,
+      "loss": 4.3393,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5537978990207959,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00018369725578225402,
+      "loss": 4.3316,
+      "step": 5341
+    },
+    {
+      "epoch": 0.553901587075284,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018369131095396263,
+      "loss": 4.3757,
+      "step": 5342
+    },
+    {
+      "epoch": 0.554005275129772,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00018368536513820228,
+      "loss": 4.3632,
+      "step": 5343
+    },
+    {
+      "epoch": 0.5541089631842602,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00018367941833504308,
+      "loss": 4.354,
+      "step": 5344
+    },
+    {
+      "epoch": 0.5542126512387482,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018367347054455517,
+      "loss": 4.3277,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5543163392932363,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018366752176680878,
+      "loss": 4.3754,
+      "step": 5346
+    },
+    {
+      "epoch": 0.5544200273477243,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018366157200187404,
+      "loss": 4.3197,
+      "step": 5347
+    },
+    {
+      "epoch": 0.5545237154022125,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018365562124982122,
+      "loss": 4.3296,
+      "step": 5348
+    },
+    {
+      "epoch": 0.5546274034567005,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018364966951072047,
+      "loss": 4.3743,
+      "step": 5349
+    },
+    {
+      "epoch": 0.5547310915111886,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018364371678464205,
+      "loss": 4.3345,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5548347795656766,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018363776307165619,
+      "loss": 4.3296,
+      "step": 5351
+    },
+    {
+      "epoch": 0.5549384676201647,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00018363180837183314,
+      "loss": 4.3787,
+      "step": 5352
+    },
+    {
+      "epoch": 0.5550421556746528,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018362585268524314,
+      "loss": 4.2736,
+      "step": 5353
+    },
+    {
+      "epoch": 0.5551458437291409,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001836198960119565,
+      "loss": 4.3756,
+      "step": 5354
+    },
+    {
+      "epoch": 0.5552495317836289,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018361393835204342,
+      "loss": 4.3685,
+      "step": 5355
+    },
+    {
+      "epoch": 0.555353219838117,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001836079797055743,
+      "loss": 4.3604,
+      "step": 5356
+    },
+    {
+      "epoch": 0.555456907892605,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001836020200726194,
+      "loss": 4.3392,
+      "step": 5357
+    },
+    {
+      "epoch": 0.5555605959470932,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00018359605945324903,
+      "loss": 4.3684,
+      "step": 5358
+    },
+    {
+      "epoch": 0.5556642840015812,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001835900978475335,
+      "loss": 4.3533,
+      "step": 5359
+    },
+    {
+      "epoch": 0.5557679720560693,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001835841352555432,
+      "loss": 4.3828,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5558716601105574,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018357817167734844,
+      "loss": 4.3377,
+      "step": 5361
+    },
+    {
+      "epoch": 0.5559753481650455,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018357220711301962,
+      "loss": 4.3375,
+      "step": 5362
+    },
+    {
+      "epoch": 0.5560790362195336,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018356624156262708,
+      "loss": 4.3466,
+      "step": 5363
+    },
+    {
+      "epoch": 0.5561827242740216,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001835602750262412,
+      "loss": 4.3279,
+      "step": 5364
+    },
+    {
+      "epoch": 0.5562864123285097,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018355430750393244,
+      "loss": 4.3026,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5563901003829977,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00018354833899577116,
+      "loss": 4.3357,
+      "step": 5366
+    },
+    {
+      "epoch": 0.5564937884374859,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018354236950182782,
+      "loss": 4.3593,
+      "step": 5367
+    },
+    {
+      "epoch": 0.5565974764919739,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001835363990221728,
+      "loss": 4.3321,
+      "step": 5368
+    },
+    {
+      "epoch": 0.556701164546462,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00018353042755687654,
+      "loss": 4.3385,
+      "step": 5369
+    },
+    {
+      "epoch": 0.55680485260095,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001835244551060096,
+      "loss": 4.3639,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5569085406554382,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018351848166964232,
+      "loss": 4.3301,
+      "step": 5371
+    },
+    {
+      "epoch": 0.5570122287099262,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018351250724784528,
+      "loss": 4.3551,
+      "step": 5372
+    },
+    {
+      "epoch": 0.5571159167644143,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018350653184068892,
+      "loss": 4.3526,
+      "step": 5373
+    },
+    {
+      "epoch": 0.5572196048189023,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001835005554482437,
+      "loss": 4.3247,
+      "step": 5374
+    },
+    {
+      "epoch": 0.5573232928733904,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018349457807058026,
+      "loss": 4.366,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5574269809278785,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018348859970776903,
+      "loss": 4.3393,
+      "step": 5376
+    },
+    {
+      "epoch": 0.5575306689823666,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001834826203598806,
+      "loss": 4.3672,
+      "step": 5377
+    },
+    {
+      "epoch": 0.5576343570368546,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018347664002698545,
+      "loss": 4.3426,
+      "step": 5378
+    },
+    {
+      "epoch": 0.5577380450913427,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00018347065870915423,
+      "loss": 4.3356,
+      "step": 5379
+    },
+    {
+      "epoch": 0.5578417331458307,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018346467640645745,
+      "loss": 4.3529,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5579454212003189,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001834586931189657,
+      "loss": 4.3789,
+      "step": 5381
+    },
+    {
+      "epoch": 0.5580491092548069,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001834527088467496,
+      "loss": 4.3493,
+      "step": 5382
+    },
+    {
+      "epoch": 0.558152797309295,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018344672358987975,
+      "loss": 4.2857,
+      "step": 5383
+    },
+    {
+      "epoch": 0.558256485363783,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018344073734842677,
+      "loss": 4.3451,
+      "step": 5384
+    },
+    {
+      "epoch": 0.5583601734182712,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00018343475012246133,
+      "loss": 4.3336,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5584638614727592,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018342876191205395,
+      "loss": 4.2956,
+      "step": 5386
+    },
+    {
+      "epoch": 0.5585675495272473,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018342277271727543,
+      "loss": 4.2989,
+      "step": 5387
+    },
+    {
+      "epoch": 0.5586712375817353,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018341678253819637,
+      "loss": 4.3567,
+      "step": 5388
+    },
+    {
+      "epoch": 0.5587749256362234,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018341079137488745,
+      "loss": 4.3328,
+      "step": 5389
+    },
+    {
+      "epoch": 0.5588786136907115,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018340479922741932,
+      "loss": 4.3779,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5589823017451996,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018339880609586276,
+      "loss": 4.2602,
+      "step": 5391
+    },
+    {
+      "epoch": 0.5590859897996876,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018339281198028845,
+      "loss": 4.3444,
+      "step": 5392
+    },
+    {
+      "epoch": 0.5591896778541757,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018338681688076709,
+      "loss": 4.3121,
+      "step": 5393
+    },
+    {
+      "epoch": 0.5592933659086637,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001833808207973694,
+      "loss": 4.3673,
+      "step": 5394
+    },
+    {
+      "epoch": 0.5593970539631519,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018337482373016623,
+      "loss": 4.3323,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5595007420176399,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018336882567922822,
+      "loss": 4.3443,
+      "step": 5396
+    },
+    {
+      "epoch": 0.559604430072128,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018336282664462622,
+      "loss": 4.3561,
+      "step": 5397
+    },
+    {
+      "epoch": 0.559708118126616,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018335682662643099,
+      "loss": 4.3319,
+      "step": 5398
+    },
+    {
+      "epoch": 0.5598118061811042,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001833508256247133,
+      "loss": 4.3546,
+      "step": 5399
+    },
+    {
+      "epoch": 0.5599154942355922,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018334482363954396,
+      "loss": 4.3268,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5600191822900803,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018333882067099385,
+      "loss": 4.3379,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5601228703445683,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001833328167191337,
+      "loss": 4.3464,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5602265583990564,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001833268117840344,
+      "loss": 4.3033,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5603302464535445,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001833208058657668,
+      "loss": 4.3792,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5604339345080326,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001833147989644018,
+      "loss": 4.3486,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5605376225625207,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018330879108001023,
+      "loss": 4.3596,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5606413106170087,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00018330278221266298,
+      "loss": 4.3626,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5607449986714969,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018329677236243096,
+      "loss": 4.3301,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5608486867259849,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018329076152938506,
+      "loss": 4.3098,
+      "step": 5409
+    },
+    {
+      "epoch": 0.560952374780473,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018328474971359622,
+      "loss": 4.3482,
+      "step": 5410
+    },
+    {
+      "epoch": 0.561056062834961,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018327873691513537,
+      "loss": 4.3548,
+      "step": 5411
+    },
+    {
+      "epoch": 0.5611597508894491,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018327272313407343,
+      "loss": 4.368,
+      "step": 5412
+    },
+    {
+      "epoch": 0.5612634389439372,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018326670837048143,
+      "loss": 4.3198,
+      "step": 5413
+    },
+    {
+      "epoch": 0.5613671269984253,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018326069262443025,
+      "loss": 4.3299,
+      "step": 5414
+    },
+    {
+      "epoch": 0.5614708150529133,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018325467589599094,
+      "loss": 4.3464,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5615745031074014,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018324865818523445,
+      "loss": 4.3521,
+      "step": 5416
+    },
+    {
+      "epoch": 0.5616781911618894,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018324263949223178,
+      "loss": 4.3357,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5617818792163776,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018323661981705394,
+      "loss": 4.3585,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5618855672708656,
+      "grad_norm": 1.5,
+      "learning_rate": 0.000183230599159772,
+      "loss": 4.3358,
+      "step": 5419
+    },
+    {
+      "epoch": 0.5619892553253537,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018322457752045695,
+      "loss": 4.2999,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5620929433798417,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018321855489917988,
+      "loss": 4.3694,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5621966314343299,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001832125312960118,
+      "loss": 4.3505,
+      "step": 5422
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018320650671102383,
+      "loss": 4.3697,
+      "step": 5423
+    },
+    {
+      "epoch": 0.562404007543306,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018320048114428702,
+      "loss": 4.3777,
+      "step": 5424
+    },
+    {
+      "epoch": 0.562507695597794,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018319445459587247,
+      "loss": 4.3195,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5626113836522821,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001831884270658513,
+      "loss": 4.3189,
+      "step": 5426
+    },
+    {
+      "epoch": 0.5627150717067702,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00018318239855429465,
+      "loss": 4.3681,
+      "step": 5427
+    },
+    {
+      "epoch": 0.5628187597612583,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001831763690612736,
+      "loss": 4.3512,
+      "step": 5428
+    },
+    {
+      "epoch": 0.5629224478157463,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018317033858685934,
+      "loss": 4.3531,
+      "step": 5429
+    },
+    {
+      "epoch": 0.5630261358702344,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.000183164307131123,
+      "loss": 4.3556,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5631298239247224,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001831582746941357,
+      "loss": 4.3025,
+      "step": 5431
+    },
+    {
+      "epoch": 0.5632335119792106,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001831522412759687,
+      "loss": 4.3599,
+      "step": 5432
+    },
+    {
+      "epoch": 0.5633372000336986,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018314620687669316,
+      "loss": 4.3102,
+      "step": 5433
+    },
+    {
+      "epoch": 0.5634408880881867,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018314017149638026,
+      "loss": 4.3832,
+      "step": 5434
+    },
+    {
+      "epoch": 0.5635445761426747,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001831341351351012,
+      "loss": 4.3497,
+      "step": 5435
+    },
+    {
+      "epoch": 0.5636482641971629,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018312809779292722,
+      "loss": 4.3164,
+      "step": 5436
+    },
+    {
+      "epoch": 0.5637519522516509,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018312205946992958,
+      "loss": 4.3224,
+      "step": 5437
+    },
+    {
+      "epoch": 0.563855640306139,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018311602016617945,
+      "loss": 4.3528,
+      "step": 5438
+    },
+    {
+      "epoch": 0.563959328360627,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001831099798817482,
+      "loss": 4.339,
+      "step": 5439
+    },
+    {
+      "epoch": 0.5640630164151151,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.000183103938616707,
+      "loss": 4.3813,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5641667044696032,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001830978963711272,
+      "loss": 4.3508,
+      "step": 5441
+    },
+    {
+      "epoch": 0.5642703925240913,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018309185314508003,
+      "loss": 4.3509,
+      "step": 5442
+    },
+    {
+      "epoch": 0.5643740805785793,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018308580893863685,
+      "loss": 4.3256,
+      "step": 5443
+    },
+    {
+      "epoch": 0.5644777686330674,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018307976375186896,
+      "loss": 4.3288,
+      "step": 5444
+    },
+    {
+      "epoch": 0.5645814566875554,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018307371758484765,
+      "loss": 4.3302,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5646851447420436,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018306767043764429,
+      "loss": 4.3248,
+      "step": 5446
+    },
+    {
+      "epoch": 0.5647888327965316,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001830616223103302,
+      "loss": 4.3706,
+      "step": 5447
+    },
+    {
+      "epoch": 0.5648925208510197,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018305557320297678,
+      "loss": 4.3251,
+      "step": 5448
+    },
+    {
+      "epoch": 0.5649962089055078,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018304952311565544,
+      "loss": 4.3554,
+      "step": 5449
+    },
+    {
+      "epoch": 0.5650998969599959,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018304347204843748,
+      "loss": 4.3653,
+      "step": 5450
+    },
+    {
+      "epoch": 0.565203585014484,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018303742000139433,
+      "loss": 4.321,
+      "step": 5451
+    },
+    {
+      "epoch": 0.565307273068972,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018303136697459736,
+      "loss": 4.3021,
+      "step": 5452
+    },
+    {
+      "epoch": 0.5654109611234601,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001830253129681181,
+      "loss": 4.3297,
+      "step": 5453
+    },
+    {
+      "epoch": 0.5655146491779481,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001830192579820279,
+      "loss": 4.3235,
+      "step": 5454
+    },
+    {
+      "epoch": 0.5656183372324363,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018301320201639815,
+      "loss": 4.3163,
+      "step": 5455
+    },
+    {
+      "epoch": 0.5657220252869243,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001830071450713004,
+      "loss": 4.3826,
+      "step": 5456
+    },
+    {
+      "epoch": 0.5658257133414124,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001830010871468061,
+      "loss": 4.3021,
+      "step": 5457
+    },
+    {
+      "epoch": 0.5659294013959004,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018299502824298668,
+      "loss": 4.2943,
+      "step": 5458
+    },
+    {
+      "epoch": 0.5660330894503885,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018298896835991366,
+      "loss": 4.3351,
+      "step": 5459
+    },
+    {
+      "epoch": 0.5661367775048766,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00018298290749765853,
+      "loss": 4.3269,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5662404655593647,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018297684565629283,
+      "loss": 4.3372,
+      "step": 5461
+    },
+    {
+      "epoch": 0.5663441536138527,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018297078283588804,
+      "loss": 4.3356,
+      "step": 5462
+    },
+    {
+      "epoch": 0.5664478416683408,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018296471903651572,
+      "loss": 4.3262,
+      "step": 5463
+    },
+    {
+      "epoch": 0.5665515297228289,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001829586542582474,
+      "loss": 4.3409,
+      "step": 5464
+    },
+    {
+      "epoch": 0.566655217777317,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001829525885011547,
+      "loss": 4.2715,
+      "step": 5465
+    },
+    {
+      "epoch": 0.566758905831805,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018294652176530912,
+      "loss": 4.325,
+      "step": 5466
+    },
+    {
+      "epoch": 0.5668625938862931,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018294045405078225,
+      "loss": 4.309,
+      "step": 5467
+    },
+    {
+      "epoch": 0.5669662819407811,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001829343853576457,
+      "loss": 4.347,
+      "step": 5468
+    },
+    {
+      "epoch": 0.5670699699952693,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018292831568597105,
+      "loss": 4.3978,
+      "step": 5469
+    },
+    {
+      "epoch": 0.5671736580497573,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018292224503582992,
+      "loss": 4.2942,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5672773461042454,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018291617340729399,
+      "loss": 4.3313,
+      "step": 5471
+    },
+    {
+      "epoch": 0.5673810341587334,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018291010080043483,
+      "loss": 4.344,
+      "step": 5472
+    },
+    {
+      "epoch": 0.5674847222132215,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001829040272153242,
+      "loss": 4.3467,
+      "step": 5473
+    },
+    {
+      "epoch": 0.5675884102677096,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001828979526520336,
+      "loss": 4.3406,
+      "step": 5474
+    },
+    {
+      "epoch": 0.5676920983221977,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001828918771106348,
+      "loss": 4.3034,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5677957863766857,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001828858005911995,
+      "loss": 4.3252,
+      "step": 5476
+    },
+    {
+      "epoch": 0.5678994744311738,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018287972309379932,
+      "loss": 4.3511,
+      "step": 5477
+    },
+    {
+      "epoch": 0.5680031624856618,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018287364461850604,
+      "loss": 4.33,
+      "step": 5478
+    },
+    {
+      "epoch": 0.56810685054015,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018286756516539137,
+      "loss": 4.3105,
+      "step": 5479
+    },
+    {
+      "epoch": 0.568210538594638,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000182861484734527,
+      "loss": 4.2994,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5683142266491261,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001828554033259847,
+      "loss": 4.3377,
+      "step": 5481
+    },
+    {
+      "epoch": 0.5684179147036141,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018284932093983624,
+      "loss": 4.3631,
+      "step": 5482
+    },
+    {
+      "epoch": 0.5685216027581023,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018284323757615335,
+      "loss": 4.3617,
+      "step": 5483
+    },
+    {
+      "epoch": 0.5686252908125903,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018283715323500786,
+      "loss": 4.3284,
+      "step": 5484
+    },
+    {
+      "epoch": 0.5687289788670784,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001828310679164715,
+      "loss": 4.353,
+      "step": 5485
+    },
+    {
+      "epoch": 0.5688326669215664,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001828249816206161,
+      "loss": 4.36,
+      "step": 5486
+    },
+    {
+      "epoch": 0.5689363549760545,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018281889434751346,
+      "loss": 4.372,
+      "step": 5487
+    },
+    {
+      "epoch": 0.5690400430305426,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001828128060972354,
+      "loss": 4.3062,
+      "step": 5488
+    },
+    {
+      "epoch": 0.5691437310850307,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018280671686985377,
+      "loss": 4.3524,
+      "step": 5489
+    },
+    {
+      "epoch": 0.5692474191395187,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018280062666544043,
+      "loss": 4.3408,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5693511071940068,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018279453548406723,
+      "loss": 4.315,
+      "step": 5491
+    },
+    {
+      "epoch": 0.5694547952484948,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018278844332580597,
+      "loss": 4.364,
+      "step": 5492
+    },
+    {
+      "epoch": 0.569558483302983,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018278235019072864,
+      "loss": 4.2915,
+      "step": 5493
+    },
+    {
+      "epoch": 0.5696621713574711,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018277625607890708,
+      "loss": 4.2776,
+      "step": 5494
+    },
+    {
+      "epoch": 0.5697658594119591,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018277016099041318,
+      "loss": 4.3762,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5698695474664472,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018276406492531887,
+      "loss": 4.3403,
+      "step": 5496
+    },
+    {
+      "epoch": 0.5699732355209353,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001827579678836961,
+      "loss": 4.4015,
+      "step": 5497
+    },
+    {
+      "epoch": 0.5700769235754234,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00018275186986561675,
+      "loss": 4.3357,
+      "step": 5498
+    },
+    {
+      "epoch": 0.5701806116299114,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018274577087115285,
+      "loss": 4.3691,
+      "step": 5499
+    },
+    {
+      "epoch": 0.5702842996843995,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001827396709003763,
+      "loss": 4.3375,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5703879877388875,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018273356995335909,
+      "loss": 4.3429,
+      "step": 5501
+    },
+    {
+      "epoch": 0.5704916757933757,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018272746803017323,
+      "loss": 4.3488,
+      "step": 5502
+    },
+    {
+      "epoch": 0.5705953638478637,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001827213651308907,
+      "loss": 4.3569,
+      "step": 5503
+    },
+    {
+      "epoch": 0.5706990519023518,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018271526125558345,
+      "loss": 4.3455,
+      "step": 5504
+    },
+    {
+      "epoch": 0.5708027399568398,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001827091564043236,
+      "loss": 4.3558,
+      "step": 5505
+    },
+    {
+      "epoch": 0.570906428011328,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018270305057718308,
+      "loss": 4.3108,
+      "step": 5506
+    },
+    {
+      "epoch": 0.571010116065816,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018269694377423404,
+      "loss": 4.3547,
+      "step": 5507
+    },
+    {
+      "epoch": 0.5711138041203041,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018269083599554845,
+      "loss": 4.3332,
+      "step": 5508
+    },
+    {
+      "epoch": 0.5712174921747921,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001826847272411984,
+      "loss": 4.3429,
+      "step": 5509
+    },
+    {
+      "epoch": 0.5713211802292802,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.000182678617511256,
+      "loss": 4.3294,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5714248682837683,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018267250680579328,
+      "loss": 4.3342,
+      "step": 5511
+    },
+    {
+      "epoch": 0.5715285563382564,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018266639512488236,
+      "loss": 4.3545,
+      "step": 5512
+    },
+    {
+      "epoch": 0.5716322443927444,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018266028246859538,
+      "loss": 4.3752,
+      "step": 5513
+    },
+    {
+      "epoch": 0.5717359324472325,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018265416883700444,
+      "loss": 4.3253,
+      "step": 5514
+    },
+    {
+      "epoch": 0.5718396205017205,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018264805423018164,
+      "loss": 4.3538,
+      "step": 5515
+    },
+    {
+      "epoch": 0.5719433085562087,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018264193864819922,
+      "loss": 4.3428,
+      "step": 5516
+    },
+    {
+      "epoch": 0.5720469966106967,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018263582209112925,
+      "loss": 4.3088,
+      "step": 5517
+    },
+    {
+      "epoch": 0.5721506846651848,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018262970455904394,
+      "loss": 4.3636,
+      "step": 5518
+    },
+    {
+      "epoch": 0.5722543727196728,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018262358605201546,
+      "loss": 4.3314,
+      "step": 5519
+    },
+    {
+      "epoch": 0.572358060774161,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.000182617466570116,
+      "loss": 4.3425,
+      "step": 5520
+    },
+    {
+      "epoch": 0.572461748828649,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018261134611341774,
+      "loss": 4.3332,
+      "step": 5521
+    },
+    {
+      "epoch": 0.5725654368831371,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018260522468199297,
+      "loss": 4.3129,
+      "step": 5522
+    },
+    {
+      "epoch": 0.5726691249376251,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018259910227591384,
+      "loss": 4.3197,
+      "step": 5523
+    },
+    {
+      "epoch": 0.5727728129921132,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001825929788952526,
+      "loss": 4.3431,
+      "step": 5524
+    },
+    {
+      "epoch": 0.5728765010466013,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018258685454008154,
+      "loss": 4.3203,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5729801891010894,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018258072921047288,
+      "loss": 4.3404,
+      "step": 5526
+    },
+    {
+      "epoch": 0.5730838771555774,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018257460290649889,
+      "loss": 4.3576,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5731875652100655,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001825684756282319,
+      "loss": 4.413,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5732912532645535,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018256234737574415,
+      "loss": 4.3611,
+      "step": 5529
+    },
+    {
+      "epoch": 0.5733949413190417,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.000182556218149108,
+      "loss": 4.377,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5734986293735297,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001825500879483957,
+      "loss": 4.336,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5736023174280178,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018254395677367967,
+      "loss": 4.3242,
+      "step": 5532
+    },
+    {
+      "epoch": 0.5737060054825058,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018253782462503216,
+      "loss": 4.3437,
+      "step": 5533
+    },
+    {
+      "epoch": 0.573809693536994,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018253169150252561,
+      "loss": 4.3471,
+      "step": 5534
+    },
+    {
+      "epoch": 0.573913381591482,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018252555740623234,
+      "loss": 4.358,
+      "step": 5535
+    },
+    {
+      "epoch": 0.5740170696459701,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001825194223362247,
+      "loss": 4.3566,
+      "step": 5536
+    },
+    {
+      "epoch": 0.5741207577004581,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001825132862925751,
+      "loss": 4.3145,
+      "step": 5537
+    },
+    {
+      "epoch": 0.5742244457549462,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018250714927535596,
+      "loss": 4.381,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5743281338094344,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018250101128463965,
+      "loss": 4.3207,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5744318218639224,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001824948723204986,
+      "loss": 4.3251,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5745355099184105,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001824887323830053,
+      "loss": 4.3404,
+      "step": 5541
+    },
+    {
+      "epoch": 0.5746391979728985,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018248259147223215,
+      "loss": 4.3781,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5747428860273867,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001824764495882516,
+      "loss": 4.3385,
+      "step": 5543
+    },
+    {
+      "epoch": 0.5748465740818747,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001824703067311361,
+      "loss": 4.339,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5749502621363628,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001824641629009582,
+      "loss": 4.3373,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5750539501908508,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001824580180977903,
+      "loss": 4.2672,
+      "step": 5546
+    },
+    {
+      "epoch": 0.5751576382453389,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018245187232170497,
+      "loss": 4.3089,
+      "step": 5547
+    },
+    {
+      "epoch": 0.575261326299827,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018244572557277473,
+      "loss": 4.3049,
+      "step": 5548
+    },
+    {
+      "epoch": 0.5753650143543151,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018243957785107204,
+      "loss": 4.3171,
+      "step": 5549
+    },
+    {
+      "epoch": 0.5754687024088031,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018243342915666948,
+      "loss": 4.3574,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5755723904632912,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001824272794896396,
+      "loss": 4.3325,
+      "step": 5551
+    },
+    {
+      "epoch": 0.5756760785177792,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018242112885005494,
+      "loss": 4.3621,
+      "step": 5552
+    },
+    {
+      "epoch": 0.5757797665722674,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001824149772379881,
+      "loss": 4.3562,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5758834546267554,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018240882465351163,
+      "loss": 4.3579,
+      "step": 5554
+    },
+    {
+      "epoch": 0.5759871426812435,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018240267109669814,
+      "loss": 4.3047,
+      "step": 5555
+    },
+    {
+      "epoch": 0.5760908307357315,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018239651656762026,
+      "loss": 4.3361,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5761945187902197,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018239036106635056,
+      "loss": 4.3471,
+      "step": 5557
+    },
+    {
+      "epoch": 0.5762982068447077,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018238420459296167,
+      "loss": 4.3176,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5764018948991958,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018237804714752627,
+      "loss": 4.3554,
+      "step": 5559
+    },
+    {
+      "epoch": 0.5765055829536838,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.000182371888730117,
+      "loss": 4.3582,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5766092710081719,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018236572934080648,
+      "loss": 4.3058,
+      "step": 5561
+    },
+    {
+      "epoch": 0.57671295906266,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018235956897966747,
+      "loss": 4.3419,
+      "step": 5562
+    },
+    {
+      "epoch": 0.5768166471171481,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018235340764677255,
+      "loss": 4.3409,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5769203351716361,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018234724534219453,
+      "loss": 4.3695,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5770240232261242,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000182341082066006,
+      "loss": 4.3446,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5771277112806122,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018233491781827977,
+      "loss": 4.3477,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5772313993351004,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018232875259908854,
+      "loss": 4.3555,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5773350873895884,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018232258640850507,
+      "loss": 4.3217,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5774387754440765,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018231641924660208,
+      "loss": 4.2945,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5775424634985645,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018231025111345233,
+      "loss": 4.3787,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5776461515530527,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018230408200912868,
+      "loss": 4.3239,
+      "step": 5571
+    },
+    {
+      "epoch": 0.5777498396075407,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018229791193370384,
+      "loss": 4.3548,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5778535276620288,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018229174088725062,
+      "loss": 4.3668,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5779572157165168,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018228556886984182,
+      "loss": 4.3516,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5780609037710049,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018227939588155031,
+      "loss": 4.344,
+      "step": 5575
+    },
+    {
+      "epoch": 0.578164591825493,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018227322192244892,
+      "loss": 4.3754,
+      "step": 5576
+    },
+    {
+      "epoch": 0.5782682798799811,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018226704699261047,
+      "loss": 4.3347,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5783719679344691,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001822608710921078,
+      "loss": 4.3517,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5784756559889572,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018225469422101384,
+      "loss": 4.3267,
+      "step": 5579
+    },
+    {
+      "epoch": 0.5785793440434452,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001822485163794014,
+      "loss": 4.3415,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5786830320979334,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018224233756734343,
+      "loss": 4.3439,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5787867201524214,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001822361577849128,
+      "loss": 4.3165,
+      "step": 5582
+    },
+    {
+      "epoch": 0.5788904082069095,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018222997703218246,
+      "loss": 4.3216,
+      "step": 5583
+    },
+    {
+      "epoch": 0.5789940962613976,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001822237953092253,
+      "loss": 4.3286,
+      "step": 5584
+    },
+    {
+      "epoch": 0.5790977843158857,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018221761261611423,
+      "loss": 4.3354,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5792014723703738,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001822114289529223,
+      "loss": 4.3254,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5793051604248618,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018220524431972237,
+      "loss": 4.3203,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5794088484793499,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018219905871658747,
+      "loss": 4.3531,
+      "step": 5588
+    },
+    {
+      "epoch": 0.5795125365338379,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018219287214359055,
+      "loss": 4.3582,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5796162245883261,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018218668460080463,
+      "loss": 4.3153,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5797199126428141,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018218049608830273,
+      "loss": 4.3506,
+      "step": 5591
+    },
+    {
+      "epoch": 0.5798236006973022,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001821743066061578,
+      "loss": 4.3519,
+      "step": 5592
+    },
+    {
+      "epoch": 0.5799272887517902,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018216811615444294,
+      "loss": 4.3325,
+      "step": 5593
+    },
+    {
+      "epoch": 0.5800309768062784,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018216192473323114,
+      "loss": 4.373,
+      "step": 5594
+    },
+    {
+      "epoch": 0.5801346648607664,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001821557323425955,
+      "loss": 4.3325,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5802383529152545,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018214953898260908,
+      "loss": 4.2523,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5803420409697425,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018214334465334488,
+      "loss": 4.3368,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5804457290242306,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001821371493548761,
+      "loss": 4.3503,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5805494170787187,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018213095308727576,
+      "loss": 4.3115,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5806531051332068,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.000182124755850617,
+      "loss": 4.3373,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5807567931876948,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001821185576449729,
+      "loss": 4.3589,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5808604812421829,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018211235847041663,
+      "loss": 4.3404,
+      "step": 5602
+    },
+    {
+      "epoch": 0.5809641692966709,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018210615832702133,
+      "loss": 4.3841,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5810678573511591,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018209995721486016,
+      "loss": 4.3259,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5811715454056471,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018209375513400628,
+      "loss": 4.3524,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5812752334601352,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018208755208453287,
+      "loss": 4.3828,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5813789215146232,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018208134806651312,
+      "loss": 4.3642,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5814826095691114,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018207514308002018,
+      "loss": 4.3149,
+      "step": 5608
+    },
+    {
+      "epoch": 0.5815862976235994,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018206893712512735,
+      "loss": 4.3453,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5816899856780875,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018206273020190782,
+      "loss": 4.343,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5817936737325755,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001820565223104348,
+      "loss": 4.3722,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5818973617870636,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018205031345078156,
+      "loss": 4.3135,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5820010498415517,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018204410362302134,
+      "loss": 4.2871,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5821047378960398,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018203789282722743,
+      "loss": 4.3021,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5822084259505278,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001820316810634731,
+      "loss": 4.338,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5823121140050159,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00018202546833183164,
+      "loss": 4.2786,
+      "step": 5616
+    },
+    {
+      "epoch": 0.5824158020595039,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018201925463237636,
+      "loss": 4.2837,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5825194901139921,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001820130399651806,
+      "loss": 4.2989,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5826231781684801,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018200682433031765,
+      "loss": 4.2944,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5827268662229682,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018200060772786083,
+      "loss": 4.3102,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5828305542774562,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018199439015788356,
+      "loss": 4.3416,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5829342423319444,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018198817162045912,
+      "loss": 4.3264,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5830379303864324,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018198195211566095,
+      "loss": 4.3834,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5831416184409205,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018197573164356238,
+      "loss": 4.3006,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5832453064954085,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018196951020423683,
+      "loss": 4.3308,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5833489945498966,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018196328779775768,
+      "loss": 4.3448,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5834526826043847,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018195706442419843,
+      "loss": 4.3255,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5835563706588728,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001819508400836324,
+      "loss": 4.3572,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5836600587133609,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018194461477613315,
+      "loss": 4.3235,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5837637467678489,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.000181938388501774,
+      "loss": 4.3347,
+      "step": 5630
+    },
+    {
+      "epoch": 0.583867434822337,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018193216126062851,
+      "loss": 4.322,
+      "step": 5631
+    },
+    {
+      "epoch": 0.5839711228768251,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001819259330527701,
+      "loss": 4.363,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5840748109313132,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001819197038782723,
+      "loss": 4.3823,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5841784989858012,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018191347373720858,
+      "loss": 4.3072,
+      "step": 5634
+    },
+    {
+      "epoch": 0.5842821870402893,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018190724262965246,
+      "loss": 4.3529,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5843858750947774,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018190101055567744,
+      "loss": 4.3431,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5844895631492655,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001818947775153571,
+      "loss": 4.3289,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5845932512037535,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018188854350876494,
+      "loss": 4.3134,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5846969392582416,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001818823085359745,
+      "loss": 4.3689,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5848006273127296,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00018187607259705942,
+      "loss": 4.3616,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5849043153672178,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001818698356920932,
+      "loss": 4.3581,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5850080034217058,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018186359782114945,
+      "loss": 4.3455,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5851116914761939,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018185735898430182,
+      "loss": 4.3584,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5852153795306819,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018185111918162384,
+      "loss": 4.3607,
+      "step": 5644
+    },
+    {
+      "epoch": 0.58531906758517,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018184487841318918,
+      "loss": 4.3458,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5854227556396581,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018183863667907147,
+      "loss": 4.3571,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5855264436941462,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018183239397934436,
+      "loss": 4.3665,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5856301317486342,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001818261503140815,
+      "loss": 4.3623,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5857338198031223,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018181990568335657,
+      "loss": 4.3244,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5858375078576104,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018181366008724324,
+      "loss": 4.2745,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5859411959120985,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018180741352581518,
+      "loss": 4.3148,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5860448839665865,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018180116599914614,
+      "loss": 4.3239,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5861485720210746,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018179491750730978,
+      "loss": 4.3469,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5862522600755626,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018178866805037988,
+      "loss": 4.3679,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5863559481300508,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018178241762843014,
+      "loss": 4.3314,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5864596361845388,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018177616624153432,
+      "loss": 4.3647,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5865633242390269,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018176991388976622,
+      "loss": 4.3769,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5866670122935149,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001817636605731995,
+      "loss": 4.3395,
+      "step": 5658
+    },
+    {
+      "epoch": 0.586770700348003,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018175740629190805,
+      "loss": 4.3378,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5868743884024911,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001817511510459656,
+      "loss": 4.3157,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5869780764569792,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018174489483544604,
+      "loss": 4.3298,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5870817645114672,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00018173863766042308,
+      "loss": 4.3223,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5871854525659553,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018173237952097063,
+      "loss": 4.3173,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5872891406204434,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018172612041716246,
+      "loss": 4.3169,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5873928286749315,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001817198603490725,
+      "loss": 4.3136,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5874965167294195,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018171359931677453,
+      "loss": 4.3398,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5876002047839076,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018170733732034248,
+      "loss": 4.3699,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5877038928383956,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018170107435985021,
+      "loss": 4.3358,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5878075808928838,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018169481043537166,
+      "loss": 4.3863,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5879112689473718,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018168854554698064,
+      "loss": 4.3141,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5880149570018599,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018168227969475118,
+      "loss": 4.3231,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5881186450563479,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018167601287875712,
+      "loss": 4.3552,
+      "step": 5672
+    },
+    {
+      "epoch": 0.588222333110836,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018166974509907247,
+      "loss": 4.3648,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5883260211653242,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00018166347635577117,
+      "loss": 4.3125,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5884297092198122,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018165720664892714,
+      "loss": 4.2986,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5885333972743003,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001816509359786144,
+      "loss": 4.3168,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5886370853287883,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018164466434490692,
+      "loss": 4.3265,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5887407733832765,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018163839174787874,
+      "loss": 4.3073,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5888444614377645,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018163211818760379,
+      "loss": 4.3157,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5889481494922526,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018162584366415615,
+      "loss": 4.3241,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5890518375467406,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018161956817760983,
+      "loss": 4.3372,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5891555256012287,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001816132917280389,
+      "loss": 4.3576,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5892592136557168,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018160701431551736,
+      "loss": 4.3108,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5893629017102049,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018160073594011936,
+      "loss": 4.3198,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5894665897646929,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018159445660191888,
+      "loss": 4.3091,
+      "step": 5685
+    },
+    {
+      "epoch": 0.589570277819181,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001815881763009901,
+      "loss": 4.3254,
+      "step": 5686
+    },
+    {
+      "epoch": 0.589673965873669,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018158189503740709,
+      "loss": 4.3508,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5897776539281572,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018157561281124392,
+      "loss": 4.3215,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5898813419826452,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018156932962257475,
+      "loss": 4.2878,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5899850300371333,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018156304547147374,
+      "loss": 4.3404,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5900887180916213,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018155676035801498,
+      "loss": 4.3673,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5901924061461095,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018155047428227268,
+      "loss": 4.3134,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5902960942005975,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000181544187244321,
+      "loss": 4.3343,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5903997822550856,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018153789924423407,
+      "loss": 4.329,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5905034703095736,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018153161028208614,
+      "loss": 4.3077,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5906071583640617,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018152532035795136,
+      "loss": 4.308,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5907108464185498,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018151902947190402,
+      "loss": 4.3066,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5908145344730379,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018151273762401825,
+      "loss": 4.3872,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5909182225275259,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001815064448143684,
+      "loss": 4.3014,
+      "step": 5699
+    },
+    {
+      "epoch": 0.591021910582014,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001815001510430286,
+      "loss": 4.3571,
+      "step": 5700
+    },
+    {
+      "epoch": 0.591125598636502,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018149385631007322,
+      "loss": 4.3611,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5912292866909902,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018148756061557646,
+      "loss": 4.3576,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5913329747454782,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001814812639596126,
+      "loss": 4.3244,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5914366627999663,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018147496634225596,
+      "loss": 4.3498,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5915403508544543,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018146866776358084,
+      "loss": 4.3045,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5916440389089425,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001814623682236616,
+      "loss": 4.398,
+      "step": 5706
+    },
+    {
+      "epoch": 0.5917477269634305,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018145606772257246,
+      "loss": 4.36,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5918514150179186,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018144976626038785,
+      "loss": 4.3212,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5919551030724066,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018144346383718211,
+      "loss": 4.3165,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5920587911268947,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018143716045302956,
+      "loss": 4.3158,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5921624791813828,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001814308561080046,
+      "loss": 4.3757,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5922661672358709,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018142455080218163,
+      "loss": 4.3474,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5923698552903589,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018141824453563504,
+      "loss": 4.3471,
+      "step": 5713
+    },
+    {
+      "epoch": 0.592473543344847,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018141193730843923,
+      "loss": 4.3295,
+      "step": 5714
+    },
+    {
+      "epoch": 0.592577231399335,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018140562912066858,
+      "loss": 4.2494,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5926809194538232,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018139931997239757,
+      "loss": 4.3036,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5927846075083113,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018139300986370064,
+      "loss": 4.3459,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5928882955627993,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018138669879465223,
+      "loss": 4.3585,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5929919836172874,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001813803867653268,
+      "loss": 4.3598,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5930956716717755,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018137407377579883,
+      "loss": 4.3545,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5931993597262636,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018136775982614277,
+      "loss": 4.334,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5933030477807516,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018136144491643318,
+      "loss": 4.3667,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5934067358352397,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018135512904674458,
+      "loss": 4.3182,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5935104238897277,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001813488122171514,
+      "loss": 4.3276,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5936141119442159,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018134249442772825,
+      "loss": 4.2957,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5937177999987039,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018133617567854966,
+      "loss": 4.3233,
+      "step": 5726
+    },
+    {
+      "epoch": 0.593821488053192,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018132985596969013,
+      "loss": 4.3506,
+      "step": 5727
+    },
+    {
+      "epoch": 0.59392517610768,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018132353530122433,
+      "loss": 4.3018,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5940288641621682,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018131721367322672,
+      "loss": 4.3681,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5941325522166562,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00018131089108577197,
+      "loss": 4.3353,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5942362402711443,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018130456753893466,
+      "loss": 4.336,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5943399283256323,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018129824303278938,
+      "loss": 4.3253,
+      "step": 5732
+    },
+    {
+      "epoch": 0.5944436163801204,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018129191756741076,
+      "loss": 4.3457,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5945473044346085,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018128559114287347,
+      "loss": 4.3569,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5946509924890966,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001812792637592521,
+      "loss": 4.3368,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5947546805435846,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018127293541662137,
+      "loss": 4.3151,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5948583685980727,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018126660611505587,
+      "loss": 4.3404,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5949620566525607,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018126027585463038,
+      "loss": 4.3155,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5950657447070489,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018125394463541948,
+      "loss": 4.3546,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5951694327615369,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001812476124574979,
+      "loss": 4.2847,
+      "step": 5740
+    },
+    {
+      "epoch": 0.595273120816025,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001812412793209404,
+      "loss": 4.338,
+      "step": 5741
+    },
+    {
+      "epoch": 0.595376808870513,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001812349452258217,
+      "loss": 4.3296,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5954804969250012,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018122861017221654,
+      "loss": 4.3893,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5955841849794892,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018122227416019957,
+      "loss": 4.3603,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5956878730339773,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018121593718984567,
+      "loss": 4.2856,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5957915610884653,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018120959926122953,
+      "loss": 4.3316,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5958952491429534,
+      "grad_norm": 1.25,
+      "learning_rate": 0.000181203260374426,
+      "loss": 4.3399,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5959989371974415,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018119692052950977,
+      "loss": 4.349,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5961026252519296,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001811905797265558,
+      "loss": 4.345,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5962063133064176,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018118423796563874,
+      "loss": 4.3052,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5963100013609057,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018117789524683348,
+      "loss": 4.2979,
+      "step": 5751
+    },
+    {
+      "epoch": 0.5964136894153937,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001811715515702149,
+      "loss": 4.3457,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5965173774698819,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001811652069358578,
+      "loss": 4.3463,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5966210655243699,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018115886134383705,
+      "loss": 4.299,
+      "step": 5754
+    },
+    {
+      "epoch": 0.596724753578858,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018115251479422755,
+      "loss": 4.324,
+      "step": 5755
+    },
+    {
+      "epoch": 0.596828441633346,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018114616728710415,
+      "loss": 4.3441,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5969321296878342,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018113981882254173,
+      "loss": 4.3491,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5970358177423222,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001811334694006152,
+      "loss": 4.3078,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5971395057968103,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018112711902139954,
+      "loss": 4.3105,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5972431938512983,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001811207676849696,
+      "loss": 4.3357,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5973468819057864,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018111441539140038,
+      "loss": 4.3783,
+      "step": 5761
+    },
+    {
+      "epoch": 0.5974505699602746,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018110806214076676,
+      "loss": 4.3518,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5975542580147626,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018110170793314377,
+      "loss": 4.2929,
+      "step": 5763
+    },
+    {
+      "epoch": 0.5976579460692507,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018109535276860633,
+      "loss": 4.3141,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5977616341237387,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001810889966472295,
+      "loss": 4.3106,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5978653221782269,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001810826395690882,
+      "loss": 4.33,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5979690102327149,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018107628153425745,
+      "loss": 4.3587,
+      "step": 5767
+    },
+    {
+      "epoch": 0.598072698287203,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018106992254281225,
+      "loss": 4.3539,
+      "step": 5768
+    },
+    {
+      "epoch": 0.598176386341691,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001810635625948277,
+      "loss": 4.3211,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5982800743961791,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001810572016903788,
+      "loss": 4.3466,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5983837624506672,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018105083982954058,
+      "loss": 4.3471,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5984874505051553,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018104447701238814,
+      "loss": 4.3135,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5985911385596433,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018103811323899653,
+      "loss": 4.3606,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5986948266141314,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018103174850944085,
+      "loss": 4.3562,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5987985146686194,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018102538282379618,
+      "loss": 4.2699,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5989022027231076,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018101901618213767,
+      "loss": 4.3522,
+      "step": 5776
+    },
+    {
+      "epoch": 0.5990058907775956,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018101264858454036,
+      "loss": 4.3327,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5991095788320837,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018100628003107948,
+      "loss": 4.3317,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5992132668865717,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001809999105218301,
+      "loss": 4.3271,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5993169549410599,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001809935400568674,
+      "loss": 4.3283,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5994206429955479,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001809871686362665,
+      "loss": 4.3304,
+      "step": 5781
+    },
+    {
+      "epoch": 0.599524331050036,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001809807962601027,
+      "loss": 4.3292,
+      "step": 5782
+    },
+    {
+      "epoch": 0.599628019104524,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018097442292845106,
+      "loss": 4.344,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5997317071590121,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001809680486413868,
+      "loss": 4.3426,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5998353952135002,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001809616733989852,
+      "loss": 4.3199,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5999390832679883,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001809552972013214,
+      "loss": 4.3017,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6000427713224763,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018094892004847068,
+      "loss": 4.289,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6001464593769644,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018094254194050827,
+      "loss": 4.3317,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6002501474314524,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018093616287750942,
+      "loss": 4.3347,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6003538354859406,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018092978285954943,
+      "loss": 4.3655,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6004575235404286,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001809234018867035,
+      "loss": 4.3226,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6005612115949167,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.000180917019959047,
+      "loss": 4.3014,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6006648996494047,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001809106370766552,
+      "loss": 4.3476,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6007685877038929,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001809042532396034,
+      "loss": 4.3053,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6008722757583809,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018089786844796693,
+      "loss": 4.3442,
+      "step": 5795
+    },
+    {
+      "epoch": 0.600975963812869,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018089148270182111,
+      "loss": 4.3391,
+      "step": 5796
+    },
+    {
+      "epoch": 0.601079651867357,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018088509600124134,
+      "loss": 4.3439,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6011833399218451,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001808787083463029,
+      "loss": 4.3031,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6012870279763332,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001808723197370812,
+      "loss": 4.3131,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6013907160308213,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018086593017365164,
+      "loss": 4.3422,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6014944040853093,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018085953965608952,
+      "loss": 4.3209,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6015980921397974,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018085314818447036,
+      "loss": 4.3095,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6017017801942854,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018084675575886952,
+      "loss": 4.3618,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6018054682487736,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018084036237936237,
+      "loss": 4.3659,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6019091563032616,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018083396804602443,
+      "loss": 4.3087,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6020128443577497,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018082757275893113,
+      "loss": 4.3818,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6021165324122378,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001808211765181579,
+      "loss": 4.3672,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6022202204667259,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001808147793237802,
+      "loss": 4.3323,
+      "step": 5808
+    },
+    {
+      "epoch": 0.602323908521214,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018080838117587352,
+      "loss": 4.3476,
+      "step": 5809
+    },
+    {
+      "epoch": 0.602427596575702,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001808019820745134,
+      "loss": 4.3119,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6025312846301901,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018079558201977526,
+      "loss": 4.3303,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6026349726846781,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001807891810117347,
+      "loss": 4.3118,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6027386607391663,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001807827790504672,
+      "loss": 4.3289,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6028423487936543,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018077637613604826,
+      "loss": 4.2843,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6029460368481424,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001807699722685535,
+      "loss": 4.3415,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6030497249026304,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018076356744805842,
+      "loss": 4.2981,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6031534129571186,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018075716167463863,
+      "loss": 4.2897,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6032571010116066,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001807507549483697,
+      "loss": 4.3263,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6033607890660947,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001807443472693272,
+      "loss": 4.3334,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6034644771205827,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018073793863758675,
+      "loss": 4.3286,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6035681651750708,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018073152905322397,
+      "loss": 4.3683,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6036718532295589,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018072511851631448,
+      "loss": 4.3507,
+      "step": 5822
+    },
+    {
+      "epoch": 0.603775541284047,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00018071870702693397,
+      "loss": 4.2915,
+      "step": 5823
+    },
+    {
+      "epoch": 0.603879229338535,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000180712294585158,
+      "loss": 4.3375,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6039829173930231,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018070588119106228,
+      "loss": 4.3278,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6040866054475111,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018069946684472248,
+      "loss": 4.3463,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6041902935019993,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018069305154621424,
+      "loss": 4.3182,
+      "step": 5827
+    },
+    {
+      "epoch": 0.6042939815564873,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018068663529561331,
+      "loss": 4.3293,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6043976696109754,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018068021809299536,
+      "loss": 4.3106,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6045013576654634,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018067379993843617,
+      "loss": 4.3078,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6046050457199516,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018066738083201135,
+      "loss": 4.3269,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6047087337744396,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018066096077379675,
+      "loss": 4.321,
+      "step": 5832
+    },
+    {
+      "epoch": 0.6048124218289277,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018065453976386805,
+      "loss": 4.3406,
+      "step": 5833
+    },
+    {
+      "epoch": 0.6049161098834157,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018064811780230103,
+      "loss": 4.3474,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6050197979379038,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001806416948891715,
+      "loss": 4.3533,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6051234859923919,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001806352710245552,
+      "loss": 4.3238,
+      "step": 5836
+    },
+    {
+      "epoch": 0.60522717404688,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018062884620852792,
+      "loss": 4.3164,
+      "step": 5837
+    },
+    {
+      "epoch": 0.605330862101368,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018062242044116552,
+      "loss": 4.3437,
+      "step": 5838
+    },
+    {
+      "epoch": 0.6054345501558561,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018061599372254375,
+      "loss": 4.2953,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6055382382103441,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001806095660527385,
+      "loss": 4.3134,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6056419262648323,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00018060313743182554,
+      "loss": 4.3519,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6057456143193203,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018059670785988075,
+      "loss": 4.355,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6058493023738084,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018059027733698005,
+      "loss": 4.3502,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6059529904282964,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018058384586319926,
+      "loss": 4.3536,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6060566784827845,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018057741343861423,
+      "loss": 4.3558,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6061603665372726,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001805709800633009,
+      "loss": 4.3174,
+      "step": 5846
+    },
+    {
+      "epoch": 0.6062640545917607,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018056454573733518,
+      "loss": 4.364,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6063677426462487,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.000180558110460793,
+      "loss": 4.3445,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6064714307007368,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018055167423375025,
+      "loss": 4.2715,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6065751187552249,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018054523705628292,
+      "loss": 4.305,
+      "step": 5850
+    },
+    {
+      "epoch": 0.606678806809713,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001805387989284669,
+      "loss": 4.3408,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6067824948642011,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001805323598503782,
+      "loss": 4.34,
+      "step": 5852
+    },
+    {
+      "epoch": 0.6068861829186891,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001805259198220928,
+      "loss": 4.3203,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6069898709731772,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018051947884368662,
+      "loss": 4.3286,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6070935590276653,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018051303691523575,
+      "loss": 4.3236,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6071972470821534,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001805065940368161,
+      "loss": 4.335,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6073009351366414,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018050015020850378,
+      "loss": 4.3283,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6074046231911295,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018049370543037475,
+      "loss": 4.3239,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6075083112456175,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018048725970250514,
+      "loss": 4.3298,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6076119993001057,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001804808130249709,
+      "loss": 4.3457,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6077156873545937,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018047436539784812,
+      "loss": 4.3558,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6078193754090818,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018046791682121293,
+      "loss": 4.3571,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6079230634635698,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018046146729514136,
+      "loss": 4.295,
+      "step": 5863
+    },
+    {
+      "epoch": 0.608026751518058,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018045501681970954,
+      "loss": 4.3485,
+      "step": 5864
+    },
+    {
+      "epoch": 0.608130439572546,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018044856539499354,
+      "loss": 4.3523,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6082341276270341,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018044211302106953,
+      "loss": 4.3211,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6083378156815221,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018043565969801359,
+      "loss": 4.3512,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6084415037360102,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018042920542590195,
+      "loss": 4.3781,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6085451917904983,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018042275020481064,
+      "loss": 4.3541,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6086488798449864,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018041629403481592,
+      "loss": 4.3422,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6087525678994744,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018040983691599395,
+      "loss": 4.339,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6088562559539625,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018040337884842086,
+      "loss": 4.2955,
+      "step": 5872
+    },
+    {
+      "epoch": 0.6089599440084505,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018039691983217287,
+      "loss": 4.3055,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6090636320629387,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018039045986732627,
+      "loss": 4.352,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6091673201174267,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001803839989539572,
+      "loss": 4.3229,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6092710081719148,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001803775370921419,
+      "loss": 4.313,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6093746962264028,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018037107428195664,
+      "loss": 4.3309,
+      "step": 5877
+    },
+    {
+      "epoch": 0.609478384280891,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018036461052347766,
+      "loss": 4.3201,
+      "step": 5878
+    },
+    {
+      "epoch": 0.609582072335379,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001803581458167812,
+      "loss": 4.3666,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6096857603898671,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001803516801619436,
+      "loss": 4.3787,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6097894484443551,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00018034521355904108,
+      "loss": 4.3374,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6098931364988432,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018033874600815,
+      "loss": 4.3281,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6099968245533313,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001803322775093466,
+      "loss": 4.3247,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6101005126078194,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018032580806270725,
+      "loss": 4.3396,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6102042006623074,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001803193376683083,
+      "loss": 4.3337,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6103078887167955,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018031286632622603,
+      "loss": 4.347,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6104115767712835,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018030639403653683,
+      "loss": 4.3089,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6105152648257717,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018029992079931711,
+      "loss": 4.3356,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6106189528802597,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018029344661464318,
+      "loss": 4.2903,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6107226409347478,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018028697148259144,
+      "loss": 4.3385,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6108263289892358,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018028049540323832,
+      "loss": 4.3115,
+      "step": 5891
+    },
+    {
+      "epoch": 0.610930017043724,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001802740183766602,
+      "loss": 4.3394,
+      "step": 5892
+    },
+    {
+      "epoch": 0.611033705098212,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018026754040293354,
+      "loss": 4.3097,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6111373931527001,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018026106148213476,
+      "loss": 4.3807,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6112410812071881,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018025458161434027,
+      "loss": 4.3459,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6113447692616762,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018024810079962653,
+      "loss": 4.3674,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6114484573161644,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018024161903807006,
+      "loss": 4.3333,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6115521453706524,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001802351363297473,
+      "loss": 4.3329,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6116558334251405,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018022865267473473,
+      "loss": 4.3412,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6117595214796285,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018022216807310888,
+      "loss": 4.3796,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6118632095341167,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018021568252494624,
+      "loss": 4.3685,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6119668975886047,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00018020919603032334,
+      "loss": 4.3144,
+      "step": 5902
+    },
+    {
+      "epoch": 0.6120705856430928,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018020270858931666,
+      "loss": 4.3126,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6121742736975808,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00018019622020200285,
+      "loss": 4.3694,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6122779617520689,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001801897308684584,
+      "loss": 4.3326,
+      "step": 5905
+    },
+    {
+      "epoch": 0.612381649806557,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018018324058875993,
+      "loss": 4.3275,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6124853378610451,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018017674936298393,
+      "loss": 4.3632,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6125890259155331,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018017025719120703,
+      "loss": 4.3562,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6126927139700212,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018016376407350588,
+      "loss": 4.3318,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6127964020245092,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000180157270009957,
+      "loss": 4.3231,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6129000900789974,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018015077500063714,
+      "loss": 4.3291,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6130037781334854,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001801442790456228,
+      "loss": 4.3329,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6131074661879735,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018013778214499067,
+      "loss": 4.3003,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6132111542424615,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018013128429881747,
+      "loss": 4.3611,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6133148422969497,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001801247855071798,
+      "loss": 4.3372,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018011828577015434,
+      "loss": 4.3413,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6135222184059258,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001801117850878178,
+      "loss": 4.3753,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6136259064604138,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00018010528346024688,
+      "loss": 4.3338,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6137295945149019,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001800987808875183,
+      "loss": 4.3024,
+      "step": 5919
+    },
+    {
+      "epoch": 0.61383328256939,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018009227736970877,
+      "loss": 4.3136,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6139369706238781,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018008577290689503,
+      "loss": 4.318,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6140406586783661,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018007926749915383,
+      "loss": 4.324,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6141443467328542,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001800727611465619,
+      "loss": 4.3304,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6142480347873422,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018006625384919605,
+      "loss": 4.3218,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6143517228418304,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018005974560713305,
+      "loss": 4.3835,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6144554108963184,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018005323642044966,
+      "loss": 4.314,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6145590989508065,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018004672628922267,
+      "loss": 4.3165,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6146627870052945,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018004021521352893,
+      "loss": 4.3353,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6147664750597827,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001800337031934453,
+      "loss": 4.2861,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6148701631142707,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018002719022904855,
+      "loss": 4.3535,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6149738511687588,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018002067632041555,
+      "loss": 4.3158,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6150775392232468,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018001416146762314,
+      "loss": 4.3222,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6151812272777349,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018000764567074822,
+      "loss": 4.3179,
+      "step": 5933
+    },
+    {
+      "epoch": 0.615284915332223,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018000112892986765,
+      "loss": 4.3686,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6153886033867111,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001799946112450583,
+      "loss": 4.3321,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6154922914411991,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00017998809261639712,
+      "loss": 4.3088,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6155959794956872,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.000179981573043961,
+      "loss": 4.3076,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6156996675501752,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017997505252782687,
+      "loss": 4.3049,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6158033556046634,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017996853106807165,
+      "loss": 4.357,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6159070436591514,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017996200866477228,
+      "loss": 4.3391,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6160107317136395,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017995548531800573,
+      "loss": 4.372,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6161144197681276,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.000179948961027849,
+      "loss": 4.2656,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6162181078226157,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017994243579437898,
+      "loss": 4.3352,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6163217958771038,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001799359096176728,
+      "loss": 4.3499,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6164254839315918,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017992938249780733,
+      "loss": 4.3459,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6165291719860799,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017992285443485965,
+      "loss": 4.3025,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6166328600405679,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00017991632542890677,
+      "loss": 4.3394,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6167365480950561,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017990979548002572,
+      "loss": 4.3118,
+      "step": 5948
+    },
+    {
+      "epoch": 0.6168402361495441,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017990326458829355,
+      "loss": 4.2628,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6169439242040322,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001798967327537873,
+      "loss": 4.3008,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6170476122585202,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001798901999765841,
+      "loss": 4.3193,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6171513003130084,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017988366625676098,
+      "loss": 4.3051,
+      "step": 5952
+    },
+    {
+      "epoch": 0.6172549883674964,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017987713159439502,
+      "loss": 4.3361,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6173586764219845,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00017987059598956336,
+      "loss": 4.3399,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6174623644764725,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017986405944234307,
+      "loss": 4.3643,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6175660525309606,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001798575219528113,
+      "loss": 4.3694,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6176697405854487,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001798509835210452,
+      "loss": 4.3734,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6177734286399368,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001798444441471219,
+      "loss": 4.3513,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6178771166944248,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017983790383111856,
+      "loss": 4.3163,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6179808047489129,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017983136257311233,
+      "loss": 4.2875,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6180844928034009,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017982482037318042,
+      "loss": 4.3255,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6181881808578891,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00017981827723140002,
+      "loss": 4.3877,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6182918689123771,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001798117331478483,
+      "loss": 4.3281,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6183955569668652,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001798051881226025,
+      "loss": 4.3363,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6184992450213532,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017979864215573983,
+      "loss": 4.3321,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6186029330758414,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017979209524733754,
+      "loss": 4.2988,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6187066211303294,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017978554739747288,
+      "loss": 4.3493,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6188103091848175,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001797789986062231,
+      "loss": 4.2884,
+      "step": 5968
+    },
+    {
+      "epoch": 0.6189139972393055,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017977244887366545,
+      "loss": 4.3441,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6190176852937936,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00017976589819987724,
+      "loss": 4.2903,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6191213733482817,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017975934658493573,
+      "loss": 4.3436,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6192250614027698,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017975279402891826,
+      "loss": 4.3744,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6193287494572578,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001797462405319021,
+      "loss": 4.3048,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6194324375117459,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001797396860939646,
+      "loss": 4.3472,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6195361255662339,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001797331307151831,
+      "loss": 4.3844,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6196398136207221,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017972657439563493,
+      "loss": 4.3364,
+      "step": 5976
+    },
+    {
+      "epoch": 0.6197435016752101,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017972001713539748,
+      "loss": 4.3088,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6198471897296982,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017971345893454807,
+      "loss": 4.3343,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6199508777841862,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017970689979316412,
+      "loss": 4.3309,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6200545658386744,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017970033971132301,
+      "loss": 4.3378,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6201582538931624,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017969377868910216,
+      "loss": 4.2783,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6202619419476505,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017968721672657892,
+      "loss": 4.3228,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6203656300021385,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017968065382383076,
+      "loss": 4.3054,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6204693180566266,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017967408998093514,
+      "loss": 4.336,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6205730061111148,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00017966752519796945,
+      "loss": 4.3644,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6206766941656028,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017966095947501119,
+      "loss": 4.3458,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6207803822200909,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017965439281213778,
+      "loss": 4.3109,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6208840702745789,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001796478252094268,
+      "loss": 4.3486,
+      "step": 5988
+    },
+    {
+      "epoch": 0.620987758329067,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017964125666695562,
+      "loss": 4.3259,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6210914463835551,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00017963468718480181,
+      "loss": 4.2994,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6211951344380432,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017962811676304285,
+      "loss": 4.3223,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6212988224925312,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017962154540175632,
+      "loss": 4.3949,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6214025105470193,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001796149731010197,
+      "loss": 4.2671,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6215061986015074,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017960839986091057,
+      "loss": 4.3448,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6216098866559955,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00017960182568150642,
+      "loss": 4.3655,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6217135747104835,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001795952505628849,
+      "loss": 4.284,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6218172627649716,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00017958867450512358,
+      "loss": 4.3335,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6219209508194596,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001795820975083,
+      "loss": 4.3414,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6220246388739478,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017957551957249182,
+      "loss": 4.2983,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6221283269284358,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0001795689406977766,
+      "loss": 4.3232,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6222320149829239,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.000179562360884232,
+      "loss": 4.3616,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6223357030374119,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00017955578013193564,
+      "loss": 4.3173,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6224393910919,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017954919844096517,
+      "loss": 4.3492,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6225430791463881,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00017954261581139825,
+      "loss": 4.3315,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6226467672008762,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017953603224331254,
+      "loss": 4.3123,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6227504552553642,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001795294477367857,
+      "loss": 4.332,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6228541433098523,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017952286229189546,
+      "loss": 4.3646,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6229578313643404,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00017951627590871952,
+      "loss": 4.31,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6230615194188285,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00017950968858733557,
+      "loss": 4.3498,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6231652074733165,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00017950310032782132,
+      "loss": 4.3115,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6232688955278046,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017949651113025454,
+      "loss": 4.3336,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6233725835822926,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017948992099471296,
+      "loss": 4.36,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6234762716367808,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017948332992127433,
+      "loss": 4.3652,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6235799596912688,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017947673791001643,
+      "loss": 4.3192,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6236836477457569,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017947014496101703,
+      "loss": 4.3289,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6237873358002449,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00017946355107435391,
+      "loss": 4.3483,
+      "step": 6016
+    },
+    {
+      "epoch": 0.623891023854733,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001794569562501049,
+      "loss": 4.3262,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6239947119092211,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001794503604883478,
+      "loss": 4.3498,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6240983999637092,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017944376378916044,
+      "loss": 4.3238,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6242020880181972,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017943716615262062,
+      "loss": 4.3432,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6243057760726853,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001794305675788062,
+      "loss": 4.3482,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6244094641271734,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00017942396806779507,
+      "loss": 4.3173,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6245131521816615,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017941736761966506,
+      "loss": 4.304,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6246168402361495,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017941076623449406,
+      "loss": 4.3422,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6247205282906376,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017940416391235995,
+      "loss": 4.3239,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6248242163451256,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017939756065334068,
+      "loss": 4.3214,
+      "step": 6026
+    },
+    {
+      "epoch": 0.6249279043996138,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00017939095645751408,
+      "loss": 4.3207,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6250315924541018,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001793843513249581,
+      "loss": 4.3462,
+      "step": 6028
+    },
+    {
+      "epoch": 0.6251352805085899,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00017937774525575073,
+      "loss": 4.3265,
+      "step": 6029
+    },
+    {
+      "epoch": 0.625238968563078,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00017937113824996985,
+      "loss": 4.3543,
+      "step": 6030
+    },
+    {
+      "epoch": 0.625342656617566,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017936453030769346,
+      "loss": 4.3666,
+      "step": 6031
+    },
+    {
+      "epoch": 0.6254463446720542,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00017935792142899948,
+      "loss": 4.3258,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6255500327265422,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017935131161396592,
+      "loss": 4.3247,
+      "step": 6033
+    },
+    {
+      "epoch": 0.6256537207810303,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017934470086267075,
+      "loss": 4.3449,
+      "step": 6034
+    },
+    {
+      "epoch": 0.6257574088355183,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000179338089175192,
+      "loss": 4.3614,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6258610968900065,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00017933147655160766,
+      "loss": 4.3474,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6259647849444945,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017932486299199573,
+      "loss": 4.3562,
+      "step": 6037
+    },
+    {
+      "epoch": 0.6260684729989826,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001793182484964343,
+      "loss": 4.3562,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6261721610534706,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001793116330650013,
+      "loss": 4.3347,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6262758491079587,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017930501669777496,
+      "loss": 4.3492,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6263795371624468,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017929839939483322,
+      "loss": 4.3075,
+      "step": 6041
+    },
+    {
+      "epoch": 0.6264832252169349,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017929178115625417,
+      "loss": 4.3827,
+      "step": 6042
+    },
+    {
+      "epoch": 0.6265869132714229,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017928516198211595,
+      "loss": 4.3436,
+      "step": 6043
+    },
+    {
+      "epoch": 0.626690601325911,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001792785418724966,
+      "loss": 4.3256,
+      "step": 6044
+    },
+    {
+      "epoch": 0.626794289380399,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017927192082747427,
+      "loss": 4.3112,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6268979774348872,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001792652988471271,
+      "loss": 4.3048,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6270016654893752,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017925867593153317,
+      "loss": 4.3716,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6271053535438633,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001792520520807706,
+      "loss": 4.2846,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6272090415983513,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017924542729491765,
+      "loss": 4.3488,
+      "step": 6049
+    },
+    {
+      "epoch": 0.6273127296528395,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017923880157405238,
+      "loss": 4.318,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6274164177073275,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.000179232174918253,
+      "loss": 4.3829,
+      "step": 6051
+    },
+    {
+      "epoch": 0.6275201057618156,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017922554732759775,
+      "loss": 4.3599,
+      "step": 6052
+    },
+    {
+      "epoch": 0.6276237938163036,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017921891880216478,
+      "loss": 4.3358,
+      "step": 6053
+    },
+    {
+      "epoch": 0.6277274818707917,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001792122893420323,
+      "loss": 4.3376,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6278311699252798,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017920565894727854,
+      "loss": 4.3467,
+      "step": 6055
+    },
+    {
+      "epoch": 0.6279348579797679,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017919902761798172,
+      "loss": 4.3125,
+      "step": 6056
+    },
+    {
+      "epoch": 0.6280385460342559,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001791923953542201,
+      "loss": 4.292,
+      "step": 6057
+    },
+    {
+      "epoch": 0.628142234088744,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017918576215607192,
+      "loss": 4.3364,
+      "step": 6058
+    },
+    {
+      "epoch": 0.628245922143232,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017917912802361543,
+      "loss": 4.3216,
+      "step": 6059
+    },
+    {
+      "epoch": 0.6283496101977202,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00017917249295692895,
+      "loss": 4.3368,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6284532982522082,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00017916585695609073,
+      "loss": 4.2879,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6285569863066963,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001791592200211791,
+      "loss": 4.348,
+      "step": 6062
+    },
+    {
+      "epoch": 0.6286606743611843,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017915258215227232,
+      "loss": 4.3623,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6287643624156725,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017914594334944873,
+      "loss": 4.3399,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6288680504701605,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017913930361278672,
+      "loss": 4.325,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6289717385246486,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017913266294236456,
+      "loss": 4.31,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6290754265791366,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001791260213382606,
+      "loss": 4.3083,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6291791146336247,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00017911937880055323,
+      "loss": 4.3409,
+      "step": 6068
+    },
+    {
+      "epoch": 0.6292828026881128,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017911273532932086,
+      "loss": 4.3099,
+      "step": 6069
+    },
+    {
+      "epoch": 0.6293864907426009,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017910609092464181,
+      "loss": 4.3098,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6294901787970889,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001790994455865945,
+      "loss": 4.2889,
+      "step": 6071
+    },
+    {
+      "epoch": 0.629593866851577,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00017909279931525735,
+      "loss": 4.3456,
+      "step": 6072
+    },
+    {
+      "epoch": 0.629697554906065,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017908615211070878,
+      "loss": 4.2922,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6298012429605532,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00017907950397302722,
+      "loss": 4.3142,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6299049310150413,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017907285490229109,
+      "loss": 4.3715,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6300086190695293,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017906620489857887,
+      "loss": 4.3564,
+      "step": 6076
+    },
+    {
+      "epoch": 0.6301123071240174,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000179059553961969,
+      "loss": 4.3453,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6302159951785055,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00017905290209253996,
+      "loss": 4.2826,
+      "step": 6078
+    },
+    {
+      "epoch": 0.6303196832329936,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017904624929037025,
+      "loss": 4.3096,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6304233712874816,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017903959555553832,
+      "loss": 4.3031,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6305270593419697,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017903294088812273,
+      "loss": 4.2875,
+      "step": 6081
+    },
+    {
+      "epoch": 0.6306307473964577,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.000179026285288202,
+      "loss": 4.3089,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6307344354509459,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00017901962875585463,
+      "loss": 4.3298,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6308381235054339,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017901297129115914,
+      "loss": 4.2848,
+      "step": 6084
+    },
+    {
+      "epoch": 0.630941811559922,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017900631289419417,
+      "loss": 4.2869,
+      "step": 6085
+    },
+    {
+      "epoch": 0.63104549961441,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00017899965356503816,
+      "loss": 4.3072,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6311491876688982,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017899299330376977,
+      "loss": 4.3425,
+      "step": 6087
+    },
+    {
+      "epoch": 0.6312528757233862,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017898633211046753,
+      "loss": 4.3198,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6313565637778743,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017897966998521011,
+      "loss": 4.3226,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6314602518323623,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017897300692807603,
+      "loss": 4.3609,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6315639398868504,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00017896634293914398,
+      "loss": 4.2785,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6316676279413385,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00017895967801849253,
+      "loss": 4.2822,
+      "step": 6092
+    },
+    {
+      "epoch": 0.6317713159958266,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00017895301216620032,
+      "loss": 4.2794,
+      "step": 6093
+    },
+    {
+      "epoch": 0.6318750040503146,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017894634538234607,
+      "loss": 4.3705,
+      "step": 6094
+    },
+    {
+      "epoch": 0.6319786921048027,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001789396776670084,
+      "loss": 4.3103,
+      "step": 6095
+    },
+    {
+      "epoch": 0.6320823801592907,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017893300902026594,
+      "loss": 4.2928,
+      "step": 6096
+    },
+    {
+      "epoch": 0.6321860682137789,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00017892633944219743,
+      "loss": 4.3577,
+      "step": 6097
+    },
+    {
+      "epoch": 0.6322897562682669,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017891966893288154,
+      "loss": 4.3371,
+      "step": 6098
+    },
+    {
+      "epoch": 0.632393444322755,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00017891299749239696,
+      "loss": 4.336,
+      "step": 6099
+    },
+    {
+      "epoch": 0.632497132377243,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017890632512082243,
+      "loss": 4.2812,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6326008204317312,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001788996518182367,
+      "loss": 4.2998,
+      "step": 6101
+    },
+    {
+      "epoch": 0.6327045084862192,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017889297758471846,
+      "loss": 4.3527,
+      "step": 6102
+    },
+    {
+      "epoch": 0.6328081965407073,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017888630242034648,
+      "loss": 4.3369,
+      "step": 6103
+    },
+    {
+      "epoch": 0.6329118845951953,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001788796263251995,
+      "loss": 4.3144,
+      "step": 6104
+    },
+    {
+      "epoch": 0.6330155726496834,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017887294929935633,
+      "loss": 4.2951,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6331192607041715,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017886627134289573,
+      "loss": 4.3436,
+      "step": 6106
+    },
+    {
+      "epoch": 0.6332229487586596,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001788595924558965,
+      "loss": 4.3389,
+      "step": 6107
+    },
+    {
+      "epoch": 0.6333266368131476,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001788529126384374,
+      "loss": 4.3352,
+      "step": 6108
+    },
+    {
+      "epoch": 0.6334303248676357,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017884623189059733,
+      "loss": 4.3023,
+      "step": 6109
+    },
+    {
+      "epoch": 0.6335340129221237,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017883955021245505,
+      "loss": 4.2569,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6336377009766119,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001788328676040894,
+      "loss": 4.3254,
+      "step": 6111
+    },
+    {
+      "epoch": 0.6337413890310999,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017882618406557922,
+      "loss": 4.2926,
+      "step": 6112
+    },
+    {
+      "epoch": 0.633845077085588,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017881949959700343,
+      "loss": 4.3386,
+      "step": 6113
+    },
+    {
+      "epoch": 0.633948765140076,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017881281419844088,
+      "loss": 4.3414,
+      "step": 6114
+    },
+    {
+      "epoch": 0.6340524531945642,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017880612786997039,
+      "loss": 4.2778,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6341561412490522,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017879944061167092,
+      "loss": 4.3278,
+      "step": 6116
+    },
+    {
+      "epoch": 0.6342598293035403,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001787927524236213,
+      "loss": 4.3591,
+      "step": 6117
+    },
+    {
+      "epoch": 0.6343635173580283,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017878606330590054,
+      "loss": 4.3536,
+      "step": 6118
+    },
+    {
+      "epoch": 0.6344672054125164,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017877937325858748,
+      "loss": 4.3563,
+      "step": 6119
+    },
+    {
+      "epoch": 0.6345708934670046,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017877268228176112,
+      "loss": 4.3259,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6346745815214926,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017876599037550036,
+      "loss": 4.317,
+      "step": 6121
+    },
+    {
+      "epoch": 0.6347782695759807,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017875929753988416,
+      "loss": 4.3243,
+      "step": 6122
+    },
+    {
+      "epoch": 0.6348819576304687,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017875260377499152,
+      "loss": 4.2985,
+      "step": 6123
+    },
+    {
+      "epoch": 0.6349856456849569,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001787459090809014,
+      "loss": 4.3338,
+      "step": 6124
+    },
+    {
+      "epoch": 0.6350893337394449,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001787392134576928,
+      "loss": 4.3434,
+      "step": 6125
+    },
+    {
+      "epoch": 0.635193021793933,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017873251690544469,
+      "loss": 4.3351,
+      "step": 6126
+    },
+    {
+      "epoch": 0.635296709848421,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001787258194242361,
+      "loss": 4.3537,
+      "step": 6127
+    },
+    {
+      "epoch": 0.6354003979029091,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017871912101414609,
+      "loss": 4.293,
+      "step": 6128
+    },
+    {
+      "epoch": 0.6355040859573972,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001787124216752536,
+      "loss": 4.3296,
+      "step": 6129
+    },
+    {
+      "epoch": 0.6356077740118853,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001787057214076378,
+      "loss": 4.2974,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6357114620663733,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017869902021137765,
+      "loss": 4.3578,
+      "step": 6131
+    },
+    {
+      "epoch": 0.6358151501208614,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017869231808655226,
+      "loss": 4.3119,
+      "step": 6132
+    },
+    {
+      "epoch": 0.6359188381753494,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017868561503324071,
+      "loss": 4.3112,
+      "step": 6133
+    },
+    {
+      "epoch": 0.6360225262298376,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017867891105152205,
+      "loss": 4.3483,
+      "step": 6134
+    },
+    {
+      "epoch": 0.6361262142843256,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017867220614147544,
+      "loss": 4.2971,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6362299023388137,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017866550030317993,
+      "loss": 4.2936,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6363335903933017,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001786587935367147,
+      "loss": 4.294,
+      "step": 6137
+    },
+    {
+      "epoch": 0.6364372784477899,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001786520858421588,
+      "loss": 4.2954,
+      "step": 6138
+    },
+    {
+      "epoch": 0.6365409665022779,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017864537721959148,
+      "loss": 4.3349,
+      "step": 6139
+    },
+    {
+      "epoch": 0.636644654556766,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017863866766909181,
+      "loss": 4.3433,
+      "step": 6140
+    },
+    {
+      "epoch": 0.636748342611254,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017863195719073897,
+      "loss": 4.3523,
+      "step": 6141
+    },
+    {
+      "epoch": 0.6368520306657421,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001786252457846122,
+      "loss": 4.3301,
+      "step": 6142
+    },
+    {
+      "epoch": 0.6369557187202302,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001786185334507906,
+      "loss": 4.3225,
+      "step": 6143
+    },
+    {
+      "epoch": 0.6370594067747183,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017861182018935343,
+      "loss": 4.3684,
+      "step": 6144
+    },
+    {
+      "epoch": 0.6371630948292063,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001786051060003799,
+      "loss": 4.3492,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6372667828836944,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017859839088394915,
+      "loss": 4.3687,
+      "step": 6146
+    },
+    {
+      "epoch": 0.6373704709381824,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017859167484014053,
+      "loss": 4.2824,
+      "step": 6147
+    },
+    {
+      "epoch": 0.6374741589926706,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017858495786903317,
+      "loss": 4.3049,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6375778470471586,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017857823997070643,
+      "loss": 4.2993,
+      "step": 6149
+    },
+    {
+      "epoch": 0.6376815351016467,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017857152114523944,
+      "loss": 4.3058,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6377852231561347,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001785648013927116,
+      "loss": 4.346,
+      "step": 6151
+    },
+    {
+      "epoch": 0.6378889112106229,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017855808071320217,
+      "loss": 4.3001,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6379925992651109,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001785513591067904,
+      "loss": 4.3162,
+      "step": 6153
+    },
+    {
+      "epoch": 0.638096287319599,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017854463657355566,
+      "loss": 4.3288,
+      "step": 6154
+    },
+    {
+      "epoch": 0.638199975374087,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001785379131135772,
+      "loss": 4.3129,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6383036634285751,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001785311887269344,
+      "loss": 4.3475,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6384073514830632,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017852446341370658,
+      "loss": 4.3083,
+      "step": 6157
+    },
+    {
+      "epoch": 0.6385110395375513,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017851773717397307,
+      "loss": 4.3457,
+      "step": 6158
+    },
+    {
+      "epoch": 0.6386147275920393,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001785110100078133,
+      "loss": 4.318,
+      "step": 6159
+    },
+    {
+      "epoch": 0.6387184156465274,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017850428191530657,
+      "loss": 4.365,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6388221037010154,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001784975528965323,
+      "loss": 4.3537,
+      "step": 6161
+    },
+    {
+      "epoch": 0.6389257917555036,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017849082295156988,
+      "loss": 4.369,
+      "step": 6162
+    },
+    {
+      "epoch": 0.6390294798099916,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017848409208049874,
+      "loss": 4.2885,
+      "step": 6163
+    },
+    {
+      "epoch": 0.6391331678644797,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017847736028339824,
+      "loss": 4.3276,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6392368559189678,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017847062756034786,
+      "loss": 4.2959,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6393405439734559,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017846389391142705,
+      "loss": 4.3502,
+      "step": 6166
+    },
+    {
+      "epoch": 0.639444232027944,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001784571593367152,
+      "loss": 4.3225,
+      "step": 6167
+    },
+    {
+      "epoch": 0.639547920082432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001784504238362918,
+      "loss": 4.3102,
+      "step": 6168
+    },
+    {
+      "epoch": 0.6396516081369201,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017844368741023634,
+      "loss": 4.3177,
+      "step": 6169
+    },
+    {
+      "epoch": 0.6397552961914081,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017843695005862828,
+      "loss": 4.3462,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6398589842458963,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017843021178154712,
+      "loss": 4.338,
+      "step": 6171
+    },
+    {
+      "epoch": 0.6399626723003843,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017842347257907237,
+      "loss": 4.2944,
+      "step": 6172
+    },
+    {
+      "epoch": 0.6400663603548724,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017841673245128355,
+      "loss": 4.3055,
+      "step": 6173
+    },
+    {
+      "epoch": 0.6401700484093604,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017840999139826015,
+      "loss": 4.3246,
+      "step": 6174
+    },
+    {
+      "epoch": 0.6402737364638486,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017840324942008175,
+      "loss": 4.2996,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6403774245183366,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001783965065168279,
+      "loss": 4.3144,
+      "step": 6176
+    },
+    {
+      "epoch": 0.6404811125728247,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017838976268857813,
+      "loss": 4.3124,
+      "step": 6177
+    },
+    {
+      "epoch": 0.6405848006273127,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000178383017935412,
+      "loss": 4.3134,
+      "step": 6178
+    },
+    {
+      "epoch": 0.6406884886818008,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001783762722574091,
+      "loss": 4.3734,
+      "step": 6179
+    },
+    {
+      "epoch": 0.6407921767362889,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001783695256546491,
+      "loss": 4.359,
+      "step": 6180
+    },
+    {
+      "epoch": 0.640895864790777,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017836277812721148,
+      "loss": 4.3412,
+      "step": 6181
+    },
+    {
+      "epoch": 0.640999552845265,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017835602967517593,
+      "loss": 4.3363,
+      "step": 6182
+    },
+    {
+      "epoch": 0.6411032408997531,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017834928029862205,
+      "loss": 4.3228,
+      "step": 6183
+    },
+    {
+      "epoch": 0.6412069289542411,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001783425299976295,
+      "loss": 4.3238,
+      "step": 6184
+    },
+    {
+      "epoch": 0.6413106170087293,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017833577877227793,
+      "loss": 4.2995,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6414143050632173,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001783290266226469,
+      "loss": 4.3071,
+      "step": 6186
+    },
+    {
+      "epoch": 0.6415179931177054,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001783222735488162,
+      "loss": 4.3645,
+      "step": 6187
+    },
+    {
+      "epoch": 0.6416216811721934,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017831551955086545,
+      "loss": 4.345,
+      "step": 6188
+    },
+    {
+      "epoch": 0.6417253692266816,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001783087646288744,
+      "loss": 4.3403,
+      "step": 6189
+    },
+    {
+      "epoch": 0.6418290572811696,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017830200878292263,
+      "loss": 4.3157,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6419327453356577,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017829525201308998,
+      "loss": 4.3469,
+      "step": 6191
+    },
+    {
+      "epoch": 0.6420364333901457,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017828849431945608,
+      "loss": 4.3377,
+      "step": 6192
+    },
+    {
+      "epoch": 0.6421401214446338,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017828173570210072,
+      "loss": 4.3062,
+      "step": 6193
+    },
+    {
+      "epoch": 0.6422438094991219,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001782749761611036,
+      "loss": 4.3202,
+      "step": 6194
+    },
+    {
+      "epoch": 0.64234749755361,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017826821569654454,
+      "loss": 4.301,
+      "step": 6195
+    },
+    {
+      "epoch": 0.642451185608098,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017826145430850329,
+      "loss": 4.2708,
+      "step": 6196
+    },
+    {
+      "epoch": 0.6425548736625861,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017825469199705954,
+      "loss": 4.3532,
+      "step": 6197
+    },
+    {
+      "epoch": 0.6426585617170741,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017824792876229318,
+      "loss": 4.2841,
+      "step": 6198
+    },
+    {
+      "epoch": 0.6427622497715623,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017824116460428394,
+      "loss": 4.3245,
+      "step": 6199
+    },
+    {
+      "epoch": 0.6428659378260503,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017823439952311168,
+      "loss": 4.3055,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6429696258805384,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017822763351885623,
+      "loss": 4.308,
+      "step": 6201
+    },
+    {
+      "epoch": 0.6430733139350264,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017822086659159738,
+      "loss": 4.3525,
+      "step": 6202
+    },
+    {
+      "epoch": 0.6431770019895146,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017821409874141497,
+      "loss": 4.3367,
+      "step": 6203
+    },
+    {
+      "epoch": 0.6432806900440026,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001782073299683889,
+      "loss": 4.332,
+      "step": 6204
+    },
+    {
+      "epoch": 0.6433843780984907,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017820056027259895,
+      "loss": 4.3704,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6434880661529787,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001781937896541251,
+      "loss": 4.3155,
+      "step": 6206
+    },
+    {
+      "epoch": 0.6435917542074668,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017818701811304717,
+      "loss": 4.2937,
+      "step": 6207
+    },
+    {
+      "epoch": 0.6436954422619549,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017818024564944507,
+      "loss": 4.2761,
+      "step": 6208
+    },
+    {
+      "epoch": 0.643799130316443,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017817347226339872,
+      "loss": 4.2736,
+      "step": 6209
+    },
+    {
+      "epoch": 0.6439028183709311,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017816669795498805,
+      "loss": 4.309,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6440065064254191,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017815992272429294,
+      "loss": 4.2983,
+      "step": 6211
+    },
+    {
+      "epoch": 0.6441101944799072,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017815314657139336,
+      "loss": 4.315,
+      "step": 6212
+    },
+    {
+      "epoch": 0.6442138825343953,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017814636949636928,
+      "loss": 4.2786,
+      "step": 6213
+    },
+    {
+      "epoch": 0.6443175705888834,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001781395914993006,
+      "loss": 4.2994,
+      "step": 6214
+    },
+    {
+      "epoch": 0.6444212586433714,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001781328125802674,
+      "loss": 4.3356,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6445249466978595,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017812603273934956,
+      "loss": 4.3215,
+      "step": 6216
+    },
+    {
+      "epoch": 0.6446286347523476,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017811925197662714,
+      "loss": 4.3373,
+      "step": 6217
+    },
+    {
+      "epoch": 0.6447323228068357,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001781124702921801,
+      "loss": 4.317,
+      "step": 6218
+    },
+    {
+      "epoch": 0.6448360108613237,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017810568768608848,
+      "loss": 4.3405,
+      "step": 6219
+    },
+    {
+      "epoch": 0.6449396989158118,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017809890415843236,
+      "loss": 4.3034,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6450433869702998,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017809211970929166,
+      "loss": 4.3331,
+      "step": 6221
+    },
+    {
+      "epoch": 0.645147075024788,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017808533433874652,
+      "loss": 4.3232,
+      "step": 6222
+    },
+    {
+      "epoch": 0.645250763079276,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017807854804687695,
+      "loss": 4.3423,
+      "step": 6223
+    },
+    {
+      "epoch": 0.6453544511337641,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001780717608337631,
+      "loss": 4.3282,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6454581391882521,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017806497269948497,
+      "loss": 4.3107,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6455618272427402,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017805818364412268,
+      "loss": 4.3284,
+      "step": 6226
+    },
+    {
+      "epoch": 0.6456655152972283,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001780513936677563,
+      "loss": 4.3306,
+      "step": 6227
+    },
+    {
+      "epoch": 0.6457692033517164,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.000178044602770466,
+      "loss": 4.3592,
+      "step": 6228
+    },
+    {
+      "epoch": 0.6458728914062044,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017803781095233193,
+      "loss": 4.3433,
+      "step": 6229
+    },
+    {
+      "epoch": 0.6459765794606925,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017803101821343411,
+      "loss": 4.3218,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6460802675151806,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017802422455385282,
+      "loss": 4.3262,
+      "step": 6231
+    },
+    {
+      "epoch": 0.6461839555696687,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001780174299736681,
+      "loss": 4.3295,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6462876436241567,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001780106344729602,
+      "loss": 4.3481,
+      "step": 6233
+    },
+    {
+      "epoch": 0.6463913316786448,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001780038380518093,
+      "loss": 4.3348,
+      "step": 6234
+    },
+    {
+      "epoch": 0.6464950197331328,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017799704071029554,
+      "loss": 4.3457,
+      "step": 6235
+    },
+    {
+      "epoch": 0.646598707787621,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017799024244849912,
+      "loss": 4.3509,
+      "step": 6236
+    },
+    {
+      "epoch": 0.646702395842109,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017798344326650032,
+      "loss": 4.3605,
+      "step": 6237
+    },
+    {
+      "epoch": 0.6468060838965971,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017797664316437928,
+      "loss": 4.3658,
+      "step": 6238
+    },
+    {
+      "epoch": 0.6469097719510851,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017796984214221633,
+      "loss": 4.3078,
+      "step": 6239
+    },
+    {
+      "epoch": 0.6470134600055732,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017796304020009163,
+      "loss": 4.3317,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6471171480600613,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017795623733808544,
+      "loss": 4.3386,
+      "step": 6241
+    },
+    {
+      "epoch": 0.6472208361145494,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001779494335562781,
+      "loss": 4.2955,
+      "step": 6242
+    },
+    {
+      "epoch": 0.6473245241690374,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001779426288547498,
+      "loss": 4.3589,
+      "step": 6243
+    },
+    {
+      "epoch": 0.6474282122235255,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017793582323358088,
+      "loss": 4.3091,
+      "step": 6244
+    },
+    {
+      "epoch": 0.6475319002780136,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017792901669285165,
+      "loss": 4.3449,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6476355883325017,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017792220923264237,
+      "loss": 4.3256,
+      "step": 6246
+    },
+    {
+      "epoch": 0.6477392763869897,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001779154008530334,
+      "loss": 4.3174,
+      "step": 6247
+    },
+    {
+      "epoch": 0.6478429644414778,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017790859155410508,
+      "loss": 4.2901,
+      "step": 6248
+    },
+    {
+      "epoch": 0.6479466524959658,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017790178133593768,
+      "loss": 4.3499,
+      "step": 6249
+    },
+    {
+      "epoch": 0.648050340550454,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017789497019861166,
+      "loss": 4.3747,
+      "step": 6250
+    },
+    {
+      "epoch": 0.648154028604942,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017788815814220733,
+      "loss": 4.291,
+      "step": 6251
+    },
+    {
+      "epoch": 0.6482577166594301,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017788134516680504,
+      "loss": 4.2845,
+      "step": 6252
+    },
+    {
+      "epoch": 0.6483614047139182,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017787453127248522,
+      "loss": 4.3211,
+      "step": 6253
+    },
+    {
+      "epoch": 0.6484650927684062,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017786771645932824,
+      "loss": 4.3539,
+      "step": 6254
+    },
+    {
+      "epoch": 0.6485687808228944,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017786090072741448,
+      "loss": 4.329,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6486724688773824,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017785408407682444,
+      "loss": 4.3404,
+      "step": 6256
+    },
+    {
+      "epoch": 0.6487761569318705,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017784726650763846,
+      "loss": 4.3098,
+      "step": 6257
+    },
+    {
+      "epoch": 0.6488798449863585,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017784044801993706,
+      "loss": 4.305,
+      "step": 6258
+    },
+    {
+      "epoch": 0.6489835330408467,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017783362861380065,
+      "loss": 4.3093,
+      "step": 6259
+    },
+    {
+      "epoch": 0.6490872210953347,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017782680828930968,
+      "loss": 4.3413,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6491909091498228,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017781998704654466,
+      "loss": 4.2956,
+      "step": 6261
+    },
+    {
+      "epoch": 0.6492945972043108,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017781316488558603,
+      "loss": 4.3282,
+      "step": 6262
+    },
+    {
+      "epoch": 0.649398285258799,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017780634180651432,
+      "loss": 4.2974,
+      "step": 6263
+    },
+    {
+      "epoch": 0.649501973313287,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017779951780941,
+      "loss": 4.3213,
+      "step": 6264
+    },
+    {
+      "epoch": 0.6496056613677751,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017779269289435365,
+      "loss": 4.3303,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6497093494222631,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017778586706142572,
+      "loss": 4.3236,
+      "step": 6266
+    },
+    {
+      "epoch": 0.6498130374767512,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017777904031070682,
+      "loss": 4.3598,
+      "step": 6267
+    },
+    {
+      "epoch": 0.6499167255312392,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001777722126422774,
+      "loss": 4.3371,
+      "step": 6268
+    },
+    {
+      "epoch": 0.6500204135857274,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017776538405621812,
+      "loss": 4.3058,
+      "step": 6269
+    },
+    {
+      "epoch": 0.6501241016402154,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017775855455260954,
+      "loss": 4.2711,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6502277896947035,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017775172413153216,
+      "loss": 4.3393,
+      "step": 6271
+    },
+    {
+      "epoch": 0.6503314777491915,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017774489279306663,
+      "loss": 4.3036,
+      "step": 6272
+    },
+    {
+      "epoch": 0.6504351658036797,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017773806053729353,
+      "loss": 4.3239,
+      "step": 6273
+    },
+    {
+      "epoch": 0.6505388538581677,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017773122736429353,
+      "loss": 4.3377,
+      "step": 6274
+    },
+    {
+      "epoch": 0.6506425419126558,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001777243932741472,
+      "loss": 4.3155,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6507462299671438,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017771755826693518,
+      "loss": 4.2653,
+      "step": 6276
+    },
+    {
+      "epoch": 0.650849918021632,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017771072234273812,
+      "loss": 4.3229,
+      "step": 6277
+    },
+    {
+      "epoch": 0.65095360607612,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017770388550163667,
+      "loss": 4.3287,
+      "step": 6278
+    },
+    {
+      "epoch": 0.6510572941306081,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001776970477437115,
+      "loss": 4.2721,
+      "step": 6279
+    },
+    {
+      "epoch": 0.6511609821850961,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017769020906904332,
+      "loss": 4.3,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6512646702395842,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017768336947771282,
+      "loss": 4.3336,
+      "step": 6281
+    },
+    {
+      "epoch": 0.6513683582940722,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017767652896980062,
+      "loss": 4.3592,
+      "step": 6282
+    },
+    {
+      "epoch": 0.6514720463485604,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017766968754538755,
+      "loss": 4.3167,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6515757344030484,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001776628452045542,
+      "loss": 4.278,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6516794224575365,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001776560019473814,
+      "loss": 4.3557,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6517831105120245,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017764915777394985,
+      "loss": 4.3205,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6518867985665127,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017764231268434035,
+      "loss": 4.2882,
+      "step": 6287
+    },
+    {
+      "epoch": 0.6519904866210007,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001776354666786336,
+      "loss": 4.3018,
+      "step": 6288
+    },
+    {
+      "epoch": 0.6520941746754888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017762861975691044,
+      "loss": 4.3372,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6521978627299768,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001776217719192516,
+      "loss": 4.342,
+      "step": 6290
+    },
+    {
+      "epoch": 0.652301550784465,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001776149231657379,
+      "loss": 4.2978,
+      "step": 6291
+    },
+    {
+      "epoch": 0.652405238838953,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017760807349645016,
+      "loss": 4.348,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6525089268934411,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017760122291146917,
+      "loss": 4.2836,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6526126149479291,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001775943714108758,
+      "loss": 4.3148,
+      "step": 6294
+    },
+    {
+      "epoch": 0.6527163030024172,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017758751899475087,
+      "loss": 4.304,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6528199910569052,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001775806656631752,
+      "loss": 4.3578,
+      "step": 6296
+    },
+    {
+      "epoch": 0.6529236791113934,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017757381141622971,
+      "loss": 4.3294,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6530273671658815,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017756695625399522,
+      "loss": 4.3513,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6531310552203695,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017756010017655266,
+      "loss": 4.3071,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6532347432748576,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017755324318398288,
+      "loss": 4.326,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6533384313293457,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017754638527636682,
+      "loss": 4.2787,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6534421193838338,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001775395264537854,
+      "loss": 4.2901,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6535458074383218,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001775326667163195,
+      "loss": 4.3453,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6536494954928099,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001775258060640501,
+      "loss": 4.3247,
+      "step": 6304
+    },
+    {
+      "epoch": 0.653753183547298,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017751894449705814,
+      "loss": 4.3338,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6538568716017861,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017751208201542457,
+      "loss": 4.3341,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6539605596562741,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017750521861923036,
+      "loss": 4.334,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6540642477107622,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017749835430855653,
+      "loss": 4.3381,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6541679357652502,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000177491489083484,
+      "loss": 4.3229,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6542716238197384,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017748462294409385,
+      "loss": 4.2798,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6543753118742264,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017747775589046702,
+      "loss": 4.3167,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6544789999287145,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001774708879226846,
+      "loss": 4.3234,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6545826879832025,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017746401904082757,
+      "loss": 4.3201,
+      "step": 6313
+    },
+    {
+      "epoch": 0.6546863760376906,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017745714924497698,
+      "loss": 4.3389,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6547900640921787,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017745027853521394,
+      "loss": 4.2997,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6548937521466668,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017744340691161947,
+      "loss": 4.3189,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6549974402011548,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017743653437427465,
+      "loss": 4.2672,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6551011282556429,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001774296609232606,
+      "loss": 4.3556,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6552048163101309,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017742278655865837,
+      "loss": 4.3441,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6553085043646191,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017741591128054912,
+      "loss": 4.3296,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6554121924191071,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017740903508901395,
+      "loss": 4.3605,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6555158804735952,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017740215798413397,
+      "loss": 4.3189,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6556195685280832,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017739527996599034,
+      "loss": 4.3237,
+      "step": 6323
+    },
+    {
+      "epoch": 0.6557232565825714,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017738840103466422,
+      "loss": 4.3346,
+      "step": 6324
+    },
+    {
+      "epoch": 0.6558269446370594,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001773815211902368,
+      "loss": 4.2839,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6559306326915475,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001773746404327892,
+      "loss": 4.2721,
+      "step": 6326
+    },
+    {
+      "epoch": 0.6560343207460355,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017736775876240265,
+      "loss": 4.2895,
+      "step": 6327
+    },
+    {
+      "epoch": 0.6561380088005236,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001773608761791583,
+      "loss": 4.3336,
+      "step": 6328
+    },
+    {
+      "epoch": 0.6562416968550117,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017735399268313743,
+      "loss": 4.2716,
+      "step": 6329
+    },
+    {
+      "epoch": 0.6563453849094998,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017734710827442118,
+      "loss": 4.2834,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6564490729639878,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001773402229530908,
+      "loss": 4.35,
+      "step": 6331
+    },
+    {
+      "epoch": 0.6565527610184759,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017733333671922756,
+      "loss": 4.3204,
+      "step": 6332
+    },
+    {
+      "epoch": 0.6566564490729639,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001773264495729127,
+      "loss": 4.3374,
+      "step": 6333
+    },
+    {
+      "epoch": 0.6567601371274521,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017731956151422745,
+      "loss": 4.3615,
+      "step": 6334
+    },
+    {
+      "epoch": 0.6568638251819401,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017731267254325316,
+      "loss": 4.3116,
+      "step": 6335
+    },
+    {
+      "epoch": 0.6569675132364282,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017730578266007097,
+      "loss": 4.3278,
+      "step": 6336
+    },
+    {
+      "epoch": 0.6570712012909162,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017729889186476232,
+      "loss": 4.3253,
+      "step": 6337
+    },
+    {
+      "epoch": 0.6571748893454044,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017729200015740844,
+      "loss": 4.2923,
+      "step": 6338
+    },
+    {
+      "epoch": 0.6572785773998924,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017728510753809066,
+      "loss": 4.2721,
+      "step": 6339
+    },
+    {
+      "epoch": 0.6573822654543805,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001772782140068903,
+      "loss": 4.3146,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6574859535088685,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017727131956388873,
+      "loss": 4.3398,
+      "step": 6341
+    },
+    {
+      "epoch": 0.6575896415633566,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017726442420916723,
+      "loss": 4.3239,
+      "step": 6342
+    },
+    {
+      "epoch": 0.6576933296178448,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001772575279428072,
+      "loss": 4.3379,
+      "step": 6343
+    },
+    {
+      "epoch": 0.6577970176723328,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017725063076489003,
+      "loss": 4.3078,
+      "step": 6344
+    },
+    {
+      "epoch": 0.6579007057268209,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017724373267549704,
+      "loss": 4.3641,
+      "step": 6345
+    },
+    {
+      "epoch": 0.6580043937813089,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017723683367470966,
+      "loss": 4.3075,
+      "step": 6346
+    },
+    {
+      "epoch": 0.658108081835797,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017722993376260933,
+      "loss": 4.3449,
+      "step": 6347
+    },
+    {
+      "epoch": 0.6582117698902851,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017722303293927737,
+      "loss": 4.345,
+      "step": 6348
+    },
+    {
+      "epoch": 0.6583154579447732,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017721613120479524,
+      "loss": 4.3421,
+      "step": 6349
+    },
+    {
+      "epoch": 0.6584191459992612,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017720922855924442,
+      "loss": 4.31,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6585228340537493,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001772023250027063,
+      "loss": 4.2911,
+      "step": 6351
+    },
+    {
+      "epoch": 0.6586265221082374,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001771954205352624,
+      "loss": 4.3133,
+      "step": 6352
+    },
+    {
+      "epoch": 0.6587302101627255,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017718851515699407,
+      "loss": 4.2883,
+      "step": 6353
+    },
+    {
+      "epoch": 0.6588338982172135,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017718160886798288,
+      "loss": 4.3507,
+      "step": 6354
+    },
+    {
+      "epoch": 0.6589375862717016,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017717470166831028,
+      "loss": 4.3606,
+      "step": 6355
+    },
+    {
+      "epoch": 0.6590412743261896,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001771677935580578,
+      "loss": 4.3079,
+      "step": 6356
+    },
+    {
+      "epoch": 0.6591449623806778,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017716088453730692,
+      "loss": 4.3126,
+      "step": 6357
+    },
+    {
+      "epoch": 0.6592486504351658,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001771539746061391,
+      "loss": 4.3268,
+      "step": 6358
+    },
+    {
+      "epoch": 0.6593523384896539,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017714706376463602,
+      "loss": 4.3424,
+      "step": 6359
+    },
+    {
+      "epoch": 0.6594560265441419,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017714015201287912,
+      "loss": 4.3265,
+      "step": 6360
+    },
+    {
+      "epoch": 0.65955971459863,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017713323935094995,
+      "loss": 4.3295,
+      "step": 6361
+    },
+    {
+      "epoch": 0.6596634026531181,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001771263257789301,
+      "loss": 4.3506,
+      "step": 6362
+    },
+    {
+      "epoch": 0.6597670907076062,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001771194112969011,
+      "loss": 4.3425,
+      "step": 6363
+    },
+    {
+      "epoch": 0.6598707787620942,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001771124959049446,
+      "loss": 4.3069,
+      "step": 6364
+    },
+    {
+      "epoch": 0.6599744668165823,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017710557960314214,
+      "loss": 4.3119,
+      "step": 6365
+    },
+    {
+      "epoch": 0.6600781548710704,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017709866239157536,
+      "loss": 4.3332,
+      "step": 6366
+    },
+    {
+      "epoch": 0.6601818429255585,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017709174427032583,
+      "loss": 4.3551,
+      "step": 6367
+    },
+    {
+      "epoch": 0.6602855309800465,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017708482523947524,
+      "loss": 4.3009,
+      "step": 6368
+    },
+    {
+      "epoch": 0.6603892190345346,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001770779052991052,
+      "loss": 4.2607,
+      "step": 6369
+    },
+    {
+      "epoch": 0.6604929070890226,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017707098444929732,
+      "loss": 4.277,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6605965951435108,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017706406269013333,
+      "loss": 4.3004,
+      "step": 6371
+    },
+    {
+      "epoch": 0.6607002831979988,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001770571400216948,
+      "loss": 4.325,
+      "step": 6372
+    },
+    {
+      "epoch": 0.6608039712524869,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017705021644406354,
+      "loss": 4.2894,
+      "step": 6373
+    },
+    {
+      "epoch": 0.6609076593069749,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017704329195732113,
+      "loss": 4.3139,
+      "step": 6374
+    },
+    {
+      "epoch": 0.661011347361463,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001770363665615493,
+      "loss": 4.3121,
+      "step": 6375
+    },
+    {
+      "epoch": 0.6611150354159511,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017702944025682981,
+      "loss": 4.3245,
+      "step": 6376
+    },
+    {
+      "epoch": 0.6612187234704392,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017702251304324435,
+      "loss": 4.328,
+      "step": 6377
+    },
+    {
+      "epoch": 0.6613224115249272,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017701558492087463,
+      "loss": 4.329,
+      "step": 6378
+    },
+    {
+      "epoch": 0.6614260995794153,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017700865588980244,
+      "loss": 4.3363,
+      "step": 6379
+    },
+    {
+      "epoch": 0.6615297876339034,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001770017259501095,
+      "loss": 4.2961,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6616334756883915,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001769947951018776,
+      "loss": 4.3053,
+      "step": 6381
+    },
+    {
+      "epoch": 0.6617371637428795,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017698786334518848,
+      "loss": 4.3419,
+      "step": 6382
+    },
+    {
+      "epoch": 0.6618408517973676,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017698093068012398,
+      "loss": 4.3493,
+      "step": 6383
+    },
+    {
+      "epoch": 0.6619445398518556,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017697399710676586,
+      "loss": 4.2962,
+      "step": 6384
+    },
+    {
+      "epoch": 0.6620482279063438,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017696706262519592,
+      "loss": 4.3222,
+      "step": 6385
+    },
+    {
+      "epoch": 0.6621519159608318,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017696012723549602,
+      "loss": 4.2923,
+      "step": 6386
+    },
+    {
+      "epoch": 0.6622556040153199,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017695319093774796,
+      "loss": 4.3247,
+      "step": 6387
+    },
+    {
+      "epoch": 0.662359292069808,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001769462537320336,
+      "loss": 4.2963,
+      "step": 6388
+    },
+    {
+      "epoch": 0.662462980124296,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017693931561843477,
+      "loss": 4.3119,
+      "step": 6389
+    },
+    {
+      "epoch": 0.6625666681787842,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017693237659703335,
+      "loss": 4.3639,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6626703562332722,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017692543666791123,
+      "loss": 4.2999,
+      "step": 6391
+    },
+    {
+      "epoch": 0.6627740442877603,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017691849583115023,
+      "loss": 4.3166,
+      "step": 6392
+    },
+    {
+      "epoch": 0.6628777323422483,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001769115540868323,
+      "loss": 4.3059,
+      "step": 6393
+    },
+    {
+      "epoch": 0.6629814203967365,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017690461143503932,
+      "loss": 4.3199,
+      "step": 6394
+    },
+    {
+      "epoch": 0.6630851084512245,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017689766787585325,
+      "loss": 4.3414,
+      "step": 6395
+    },
+    {
+      "epoch": 0.6631887965057126,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017689072340935596,
+      "loss": 4.2836,
+      "step": 6396
+    },
+    {
+      "epoch": 0.6632924845602006,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017688377803562938,
+      "loss": 4.3232,
+      "step": 6397
+    },
+    {
+      "epoch": 0.6633961726146888,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017687683175475556,
+      "loss": 4.3835,
+      "step": 6398
+    },
+    {
+      "epoch": 0.6634998606691768,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017686988456681632,
+      "loss": 4.3314,
+      "step": 6399
+    },
+    {
+      "epoch": 0.6636035487236649,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017686293647189373,
+      "loss": 4.3314,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6637072367781529,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017685598747006976,
+      "loss": 4.2765,
+      "step": 6401
+    },
+    {
+      "epoch": 0.663810924832641,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017684903756142635,
+      "loss": 4.2805,
+      "step": 6402
+    },
+    {
+      "epoch": 0.663914612887129,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001768420867460455,
+      "loss": 4.3317,
+      "step": 6403
+    },
+    {
+      "epoch": 0.6640183009416172,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001768351350240093,
+      "loss": 4.3102,
+      "step": 6404
+    },
+    {
+      "epoch": 0.6641219889961052,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017682818239539972,
+      "loss": 4.3507,
+      "step": 6405
+    },
+    {
+      "epoch": 0.6642256770505933,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001768212288602988,
+      "loss": 4.2997,
+      "step": 6406
+    },
+    {
+      "epoch": 0.6643293651050813,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017681427441878855,
+      "loss": 4.2939,
+      "step": 6407
+    },
+    {
+      "epoch": 0.6644330531595695,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017680731907095107,
+      "loss": 4.3014,
+      "step": 6408
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017680036281686844,
+      "loss": 4.3149,
+      "step": 6409
+    },
+    {
+      "epoch": 0.6646404292685456,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017679340565662267,
+      "loss": 4.2542,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6647441173230336,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017678644759029592,
+      "loss": 4.333,
+      "step": 6411
+    },
+    {
+      "epoch": 0.6648478053775218,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017677948861797026,
+      "loss": 4.3367,
+      "step": 6412
+    },
+    {
+      "epoch": 0.6649514934320098,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017677252873972776,
+      "loss": 4.3524,
+      "step": 6413
+    },
+    {
+      "epoch": 0.6650551814864979,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017676556795565061,
+      "loss": 4.318,
+      "step": 6414
+    },
+    {
+      "epoch": 0.6651588695409859,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001767586062658209,
+      "loss": 4.3213,
+      "step": 6415
+    },
+    {
+      "epoch": 0.665262557595474,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017675164367032077,
+      "loss": 4.3024,
+      "step": 6416
+    },
+    {
+      "epoch": 0.665366245649962,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017674468016923232,
+      "loss": 4.2829,
+      "step": 6417
+    },
+    {
+      "epoch": 0.6654699337044502,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017673771576263782,
+      "loss": 4.3319,
+      "step": 6418
+    },
+    {
+      "epoch": 0.6655736217589382,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00017673075045061937,
+      "loss": 4.3104,
+      "step": 6419
+    },
+    {
+      "epoch": 0.6656773098134263,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001767237842332592,
+      "loss": 4.3173,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6657809978679143,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00017671681711063945,
+      "loss": 4.3071,
+      "step": 6421
+    },
+    {
+      "epoch": 0.6658846859224025,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017670984908284236,
+      "loss": 4.338,
+      "step": 6422
+    },
+    {
+      "epoch": 0.6659883739768905,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017670288014995014,
+      "loss": 4.3091,
+      "step": 6423
+    },
+    {
+      "epoch": 0.6660920620313786,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.000176695910312045,
+      "loss": 4.3015,
+      "step": 6424
+    },
+    {
+      "epoch": 0.6661957500858666,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001766889395692092,
+      "loss": 4.2442,
+      "step": 6425
+    },
+    {
+      "epoch": 0.6662994381403547,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017668196792152498,
+      "loss": 4.2853,
+      "step": 6426
+    },
+    {
+      "epoch": 0.6664031261948428,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00017667499536907458,
+      "loss": 4.2546,
+      "step": 6427
+    },
+    {
+      "epoch": 0.6665068142493309,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017666802191194033,
+      "loss": 4.3191,
+      "step": 6428
+    },
+    {
+      "epoch": 0.6666105023038189,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001766610475502044,
+      "loss": 4.2634,
+      "step": 6429
+    },
+    {
+      "epoch": 0.666714190358307,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017665407228394916,
+      "loss": 4.3394,
+      "step": 6430
+    },
+    {
+      "epoch": 0.666817878412795,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001766470961132569,
+      "loss": 4.3419,
+      "step": 6431
+    },
+    {
+      "epoch": 0.6669215664672832,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017664011903820994,
+      "loss": 4.2822,
+      "step": 6432
+    },
+    {
+      "epoch": 0.6670252545217713,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017663314105889058,
+      "loss": 4.3083,
+      "step": 6433
+    },
+    {
+      "epoch": 0.6671289425762593,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00017662616217538118,
+      "loss": 4.3549,
+      "step": 6434
+    },
+    {
+      "epoch": 0.6672326306307474,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017661918238776403,
+      "loss": 4.2849,
+      "step": 6435
+    },
+    {
+      "epoch": 0.6673363186852355,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001766122016961215,
+      "loss": 4.3556,
+      "step": 6436
+    },
+    {
+      "epoch": 0.6674400067397236,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017660522010053603,
+      "loss": 4.3536,
+      "step": 6437
+    },
+    {
+      "epoch": 0.6675436947942116,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001765982376010899,
+      "loss": 4.3415,
+      "step": 6438
+    },
+    {
+      "epoch": 0.6676473828486997,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017659125419786556,
+      "loss": 4.2903,
+      "step": 6439
+    },
+    {
+      "epoch": 0.6677510709031877,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017658426989094534,
+      "loss": 4.297,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6678547589576759,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017657728468041173,
+      "loss": 4.2782,
+      "step": 6441
+    },
+    {
+      "epoch": 0.6679584470121639,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017657029856634707,
+      "loss": 4.2852,
+      "step": 6442
+    },
+    {
+      "epoch": 0.668062135066652,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00017656331154883385,
+      "loss": 4.3173,
+      "step": 6443
+    },
+    {
+      "epoch": 0.66816582312114,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017655632362795448,
+      "loss": 4.3095,
+      "step": 6444
+    },
+    {
+      "epoch": 0.6682695111756282,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001765493348037914,
+      "loss": 4.2713,
+      "step": 6445
+    },
+    {
+      "epoch": 0.6683731992301162,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017654234507642711,
+      "loss": 4.3224,
+      "step": 6446
+    },
+    {
+      "epoch": 0.6684768872846043,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017653535444594405,
+      "loss": 4.3031,
+      "step": 6447
+    },
+    {
+      "epoch": 0.6685805753390923,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001765283629124247,
+      "loss": 4.3493,
+      "step": 6448
+    },
+    {
+      "epoch": 0.6686842633935804,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017652137047595155,
+      "loss": 4.3091,
+      "step": 6449
+    },
+    {
+      "epoch": 0.6687879514480685,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017651437713660714,
+      "loss": 4.3148,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6688916395025566,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017650738289447398,
+      "loss": 4.3442,
+      "step": 6451
+    },
+    {
+      "epoch": 0.6689953275570446,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017650038774963452,
+      "loss": 4.352,
+      "step": 6452
+    },
+    {
+      "epoch": 0.6690990156115327,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017649339170217138,
+      "loss": 4.3011,
+      "step": 6453
+    },
+    {
+      "epoch": 0.6692027036660207,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017648639475216705,
+      "loss": 4.3289,
+      "step": 6454
+    },
+    {
+      "epoch": 0.6693063917205089,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017647939689970413,
+      "loss": 4.3572,
+      "step": 6455
+    },
+    {
+      "epoch": 0.6694100797749969,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017647239814486517,
+      "loss": 4.2908,
+      "step": 6456
+    },
+    {
+      "epoch": 0.669513767829485,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017646539848773275,
+      "loss": 4.339,
+      "step": 6457
+    },
+    {
+      "epoch": 0.669617455883973,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017645839792838946,
+      "loss": 4.3345,
+      "step": 6458
+    },
+    {
+      "epoch": 0.6697211439384612,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00017645139646691788,
+      "loss": 4.3301,
+      "step": 6459
+    },
+    {
+      "epoch": 0.6698248319929492,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017644439410340062,
+      "loss": 4.3495,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6699285200474373,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017643739083792034,
+      "loss": 4.2701,
+      "step": 6461
+    },
+    {
+      "epoch": 0.6700322081019253,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017643038667055966,
+      "loss": 4.3344,
+      "step": 6462
+    },
+    {
+      "epoch": 0.6701358961564134,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017642338160140118,
+      "loss": 4.2788,
+      "step": 6463
+    },
+    {
+      "epoch": 0.6702395842109015,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017641637563052756,
+      "loss": 4.3089,
+      "step": 6464
+    },
+    {
+      "epoch": 0.6703432722653896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017640936875802155,
+      "loss": 4.2687,
+      "step": 6465
+    },
+    {
+      "epoch": 0.6704469603198776,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001764023609839657,
+      "loss": 4.303,
+      "step": 6466
+    },
+    {
+      "epoch": 0.6705506483743657,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017639535230844276,
+      "loss": 4.3321,
+      "step": 6467
+    },
+    {
+      "epoch": 0.6706543364288537,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017638834273153546,
+      "loss": 4.3334,
+      "step": 6468
+    },
+    {
+      "epoch": 0.6707580244833419,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001763813322533264,
+      "loss": 4.3177,
+      "step": 6469
+    },
+    {
+      "epoch": 0.6708617125378299,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001763743208738984,
+      "loss": 4.3005,
+      "step": 6470
+    },
+    {
+      "epoch": 0.670965400592318,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017636730859333413,
+      "loss": 4.256,
+      "step": 6471
+    },
+    {
+      "epoch": 0.671069088646806,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017636029541171633,
+      "loss": 4.2535,
+      "step": 6472
+    },
+    {
+      "epoch": 0.6711727767012942,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017635328132912777,
+      "loss": 4.3104,
+      "step": 6473
+    },
+    {
+      "epoch": 0.6712764647557822,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001763462663456512,
+      "loss": 4.3169,
+      "step": 6474
+    },
+    {
+      "epoch": 0.6713801528102703,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001763392504613694,
+      "loss": 4.3292,
+      "step": 6475
+    },
+    {
+      "epoch": 0.6714838408647583,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017633223367636514,
+      "loss": 4.317,
+      "step": 6476
+    },
+    {
+      "epoch": 0.6715875289192464,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017632521599072118,
+      "loss": 4.3153,
+      "step": 6477
+    },
+    {
+      "epoch": 0.6716912169737346,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017631819740452037,
+      "loss": 4.3043,
+      "step": 6478
+    },
+    {
+      "epoch": 0.6717949050282226,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001763111779178455,
+      "loss": 4.3376,
+      "step": 6479
+    },
+    {
+      "epoch": 0.6718985930827107,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001763041575307794,
+      "loss": 4.3351,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6720022811371987,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001762971362434049,
+      "loss": 4.3147,
+      "step": 6481
+    },
+    {
+      "epoch": 0.6721059691916869,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017629011405580482,
+      "loss": 4.3209,
+      "step": 6482
+    },
+    {
+      "epoch": 0.6722096572461749,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017628309096806206,
+      "loss": 4.3264,
+      "step": 6483
+    },
+    {
+      "epoch": 0.672313345300663,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017627606698025945,
+      "loss": 4.2596,
+      "step": 6484
+    },
+    {
+      "epoch": 0.672417033355151,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017626904209247987,
+      "loss": 4.3037,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6725207214096391,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017626201630480622,
+      "loss": 4.3338,
+      "step": 6486
+    },
+    {
+      "epoch": 0.6726244094641272,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017625498961732142,
+      "loss": 4.3313,
+      "step": 6487
+    },
+    {
+      "epoch": 0.6727280975186153,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001762479620301083,
+      "loss": 4.2831,
+      "step": 6488
+    },
+    {
+      "epoch": 0.6728317855731033,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017624093354324987,
+      "loss": 4.3158,
+      "step": 6489
+    },
+    {
+      "epoch": 0.6729354736275914,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000176233904156829,
+      "loss": 4.3364,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6730391616820794,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001762268738709286,
+      "loss": 4.3593,
+      "step": 6491
+    },
+    {
+      "epoch": 0.6731428497365676,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017621984268563173,
+      "loss": 4.3485,
+      "step": 6492
+    },
+    {
+      "epoch": 0.6732465377910556,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017621281060102123,
+      "loss": 4.3344,
+      "step": 6493
+    },
+    {
+      "epoch": 0.6733502258455437,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017620577761718015,
+      "loss": 4.2714,
+      "step": 6494
+    },
+    {
+      "epoch": 0.6734539139000317,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017619874373419144,
+      "loss": 4.3282,
+      "step": 6495
+    },
+    {
+      "epoch": 0.6735576019545199,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017619170895213812,
+      "loss": 4.2696,
+      "step": 6496
+    },
+    {
+      "epoch": 0.6736612900090079,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017618467327110317,
+      "loss": 4.3398,
+      "step": 6497
+    },
+    {
+      "epoch": 0.673764978063496,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017617763669116958,
+      "loss": 4.3381,
+      "step": 6498
+    },
+    {
+      "epoch": 0.673868666117984,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017617059921242042,
+      "loss": 4.3148,
+      "step": 6499
+    },
+    {
+      "epoch": 0.6739723541724721,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001761635608349387,
+      "loss": 4.2602,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6740760422269602,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017615652155880747,
+      "loss": 4.3324,
+      "step": 6501
+    },
+    {
+      "epoch": 0.6741797302814483,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017614948138410975,
+      "loss": 4.341,
+      "step": 6502
+    },
+    {
+      "epoch": 0.6742834183359363,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001761424403109287,
+      "loss": 4.3384,
+      "step": 6503
+    },
+    {
+      "epoch": 0.6743871063904244,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001761353983393473,
+      "loss": 4.2924,
+      "step": 6504
+    },
+    {
+      "epoch": 0.6744907944449124,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001761283554694487,
+      "loss": 4.3056,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6745944824994006,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.000176121311701316,
+      "loss": 4.3578,
+      "step": 6506
+    },
+    {
+      "epoch": 0.6746981705538886,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017611426703503224,
+      "loss": 4.327,
+      "step": 6507
+    },
+    {
+      "epoch": 0.6748018586083767,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001761072214706806,
+      "loss": 4.3091,
+      "step": 6508
+    },
+    {
+      "epoch": 0.6749055466628647,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001761001750083442,
+      "loss": 4.3275,
+      "step": 6509
+    },
+    {
+      "epoch": 0.6750092347173529,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001760931276481062,
+      "loss": 4.316,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6751129227718409,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001760860793900497,
+      "loss": 4.3116,
+      "step": 6511
+    },
+    {
+      "epoch": 0.675216610826329,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017607903023425788,
+      "loss": 4.3461,
+      "step": 6512
+    },
+    {
+      "epoch": 0.675320298880817,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017607198018081396,
+      "loss": 4.3162,
+      "step": 6513
+    },
+    {
+      "epoch": 0.6754239869353051,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017606492922980104,
+      "loss": 4.3131,
+      "step": 6514
+    },
+    {
+      "epoch": 0.6755276749897932,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001760578773813024,
+      "loss": 4.2896,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6756313630442813,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00017605082463540117,
+      "loss": 4.3247,
+      "step": 6516
+    },
+    {
+      "epoch": 0.6757350510987693,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001760437709921806,
+      "loss": 4.336,
+      "step": 6517
+    },
+    {
+      "epoch": 0.6758387391532574,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00017603671645172395,
+      "loss": 4.3142,
+      "step": 6518
+    },
+    {
+      "epoch": 0.6759424272077454,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017602966101411437,
+      "loss": 4.2958,
+      "step": 6519
+    },
+    {
+      "epoch": 0.6760461152622336,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017602260467943517,
+      "loss": 4.301,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6761498033167217,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017601554744776964,
+      "loss": 4.3737,
+      "step": 6521
+    },
+    {
+      "epoch": 0.6762534913712097,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017600848931920098,
+      "loss": 4.3691,
+      "step": 6522
+    },
+    {
+      "epoch": 0.6763571794256978,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017600143029381247,
+      "loss": 4.3371,
+      "step": 6523
+    },
+    {
+      "epoch": 0.6764608674801859,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017599437037168746,
+      "loss": 4.3148,
+      "step": 6524
+    },
+    {
+      "epoch": 0.676564555534674,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017598730955290917,
+      "loss": 4.3323,
+      "step": 6525
+    },
+    {
+      "epoch": 0.676668243589162,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017598024783756095,
+      "loss": 4.3187,
+      "step": 6526
+    },
+    {
+      "epoch": 0.6767719316436501,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017597318522572612,
+      "loss": 4.3184,
+      "step": 6527
+    },
+    {
+      "epoch": 0.6768756196981381,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017596612171748803,
+      "loss": 4.3484,
+      "step": 6528
+    },
+    {
+      "epoch": 0.6769793077526263,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017595905731292998,
+      "loss": 4.3241,
+      "step": 6529
+    },
+    {
+      "epoch": 0.6770829958071143,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017595199201213534,
+      "loss": 4.3255,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6771866838616024,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017594492581518752,
+      "loss": 4.3215,
+      "step": 6531
+    },
+    {
+      "epoch": 0.6772903719160904,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017593785872216982,
+      "loss": 4.3453,
+      "step": 6532
+    },
+    {
+      "epoch": 0.6773940599705786,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017593079073316566,
+      "loss": 4.3257,
+      "step": 6533
+    },
+    {
+      "epoch": 0.6774977480250666,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017592372184825847,
+      "loss": 4.318,
+      "step": 6534
+    },
+    {
+      "epoch": 0.6776014360795547,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001759166520675316,
+      "loss": 4.3414,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6777051241340427,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017590958139106845,
+      "loss": 4.3062,
+      "step": 6536
+    },
+    {
+      "epoch": 0.6778088121885308,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001759025098189525,
+      "loss": 4.3148,
+      "step": 6537
+    },
+    {
+      "epoch": 0.6779125002430189,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017589543735126718,
+      "loss": 4.2926,
+      "step": 6538
+    },
+    {
+      "epoch": 0.678016188297507,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001758883639880959,
+      "loss": 4.3369,
+      "step": 6539
+    },
+    {
+      "epoch": 0.678119876351995,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017588128972952216,
+      "loss": 4.2984,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6782235644064831,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001758742145756294,
+      "loss": 4.3083,
+      "step": 6541
+    },
+    {
+      "epoch": 0.6783272524609711,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001758671385265011,
+      "loss": 4.2944,
+      "step": 6542
+    },
+    {
+      "epoch": 0.6784309405154593,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017586006158222078,
+      "loss": 4.3253,
+      "step": 6543
+    },
+    {
+      "epoch": 0.6785346285699473,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001758529837428719,
+      "loss": 4.3334,
+      "step": 6544
+    },
+    {
+      "epoch": 0.6786383166244354,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017584590500853802,
+      "loss": 4.3115,
+      "step": 6545
+    },
+    {
+      "epoch": 0.6787420046789234,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001758388253793026,
+      "loss": 4.3329,
+      "step": 6546
+    },
+    {
+      "epoch": 0.6788456927334116,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017583174485524925,
+      "loss": 4.2879,
+      "step": 6547
+    },
+    {
+      "epoch": 0.6789493807878996,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017582466343646144,
+      "loss": 4.32,
+      "step": 6548
+    },
+    {
+      "epoch": 0.6790530688423877,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017581758112302276,
+      "loss": 4.375,
+      "step": 6549
+    },
+    {
+      "epoch": 0.6791567568968757,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017581049791501677,
+      "loss": 4.3272,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6792604449513638,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017580341381252703,
+      "loss": 4.2906,
+      "step": 6551
+    },
+    {
+      "epoch": 0.6793641330058519,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001757963288156371,
+      "loss": 4.3129,
+      "step": 6552
+    },
+    {
+      "epoch": 0.67946782106034,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017578924292443066,
+      "loss": 4.2775,
+      "step": 6553
+    },
+    {
+      "epoch": 0.679571509114828,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017578215613899128,
+      "loss": 4.3718,
+      "step": 6554
+    },
+    {
+      "epoch": 0.6796751971693161,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017577506845940254,
+      "loss": 4.2961,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6797788852238041,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017576797988574808,
+      "loss": 4.294,
+      "step": 6556
+    },
+    {
+      "epoch": 0.6798825732782923,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00017576089041811151,
+      "loss": 4.3179,
+      "step": 6557
+    },
+    {
+      "epoch": 0.6799862613327803,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017575380005657658,
+      "loss": 4.2949,
+      "step": 6558
+    },
+    {
+      "epoch": 0.6800899493872684,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00017574670880122687,
+      "loss": 4.3174,
+      "step": 6559
+    },
+    {
+      "epoch": 0.6801936374417564,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017573961665214607,
+      "loss": 4.3556,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6802973254962446,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00017573252360941785,
+      "loss": 4.3147,
+      "step": 6561
+    },
+    {
+      "epoch": 0.6804010135507326,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00017572542967312586,
+      "loss": 4.2997,
+      "step": 6562
+    },
+    {
+      "epoch": 0.6805047016052207,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001757183348433539,
+      "loss": 4.3231,
+      "step": 6563
+    },
+    {
+      "epoch": 0.6806083896597087,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001757112391201856,
+      "loss": 4.2858,
+      "step": 6564
+    },
+    {
+      "epoch": 0.6807120777141968,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017570414250370472,
+      "loss": 4.3065,
+      "step": 6565
+    },
+    {
+      "epoch": 0.680815765768685,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017569704499399496,
+      "loss": 4.2943,
+      "step": 6566
+    },
+    {
+      "epoch": 0.680919453823173,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017568994659114008,
+      "loss": 4.3219,
+      "step": 6567
+    },
+    {
+      "epoch": 0.6810231418776611,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001756828472952239,
+      "loss": 4.3509,
+      "step": 6568
+    },
+    {
+      "epoch": 0.6811268299321491,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017567574710633006,
+      "loss": 4.3214,
+      "step": 6569
+    },
+    {
+      "epoch": 0.6812305179866373,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017566864602454243,
+      "loss": 4.3195,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6813342060411253,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017566154404994471,
+      "loss": 4.3133,
+      "step": 6571
+    },
+    {
+      "epoch": 0.6814378940956134,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017565444118262081,
+      "loss": 4.3164,
+      "step": 6572
+    },
+    {
+      "epoch": 0.6815415821501014,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017564733742265445,
+      "loss": 4.3009,
+      "step": 6573
+    },
+    {
+      "epoch": 0.6816452702045895,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001756402327701295,
+      "loss": 4.3113,
+      "step": 6574
+    },
+    {
+      "epoch": 0.6817489582590776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001756331272251297,
+      "loss": 4.3586,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6818526463135657,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017562602078773898,
+      "loss": 4.3112,
+      "step": 6576
+    },
+    {
+      "epoch": 0.6819563343680537,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017561891345804117,
+      "loss": 4.3211,
+      "step": 6577
+    },
+    {
+      "epoch": 0.6820600224225418,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017561180523612008,
+      "loss": 4.3358,
+      "step": 6578
+    },
+    {
+      "epoch": 0.6821637104770298,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017560469612205965,
+      "loss": 4.3131,
+      "step": 6579
+    },
+    {
+      "epoch": 0.682267398531518,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001755975861159437,
+      "loss": 4.3321,
+      "step": 6580
+    },
+    {
+      "epoch": 0.682371086586006,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017559047521785613,
+      "loss": 4.2691,
+      "step": 6581
+    },
+    {
+      "epoch": 0.6824747746404941,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017558336342788088,
+      "loss": 4.3178,
+      "step": 6582
+    },
+    {
+      "epoch": 0.6825784626949821,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017557625074610185,
+      "loss": 4.3003,
+      "step": 6583
+    },
+    {
+      "epoch": 0.6826821507494703,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001755691371726029,
+      "loss": 4.2979,
+      "step": 6584
+    },
+    {
+      "epoch": 0.6827858388039583,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017556202270746805,
+      "loss": 4.3183,
+      "step": 6585
+    },
+    {
+      "epoch": 0.6828895268584464,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001755549073507812,
+      "loss": 4.2924,
+      "step": 6586
+    },
+    {
+      "epoch": 0.6829932149129344,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001755477911026263,
+      "loss": 4.3058,
+      "step": 6587
+    },
+    {
+      "epoch": 0.6830969029674225,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017554067396308731,
+      "loss": 4.2995,
+      "step": 6588
+    },
+    {
+      "epoch": 0.6832005910219106,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017553355593224822,
+      "loss": 4.3376,
+      "step": 6589
+    },
+    {
+      "epoch": 0.6833042790763987,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017552643701019305,
+      "loss": 4.2894,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6834079671308867,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001755193171970057,
+      "loss": 4.297,
+      "step": 6591
+    },
+    {
+      "epoch": 0.6835116551853748,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017551219649277028,
+      "loss": 4.3418,
+      "step": 6592
+    },
+    {
+      "epoch": 0.6836153432398628,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017550507489757076,
+      "loss": 4.3075,
+      "step": 6593
+    },
+    {
+      "epoch": 0.683719031294351,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017549795241149116,
+      "loss": 4.3406,
+      "step": 6594
+    },
+    {
+      "epoch": 0.683822719348839,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017549082903461552,
+      "loss": 4.2859,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6839264074033271,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001754837047670279,
+      "loss": 4.3241,
+      "step": 6596
+    },
+    {
+      "epoch": 0.6840300954578151,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017547657960881235,
+      "loss": 4.3338,
+      "step": 6597
+    },
+    {
+      "epoch": 0.6841337835123033,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017546945356005294,
+      "loss": 4.3292,
+      "step": 6598
+    },
+    {
+      "epoch": 0.6842374715667913,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017546232662083377,
+      "loss": 4.3014,
+      "step": 6599
+    },
+    {
+      "epoch": 0.6843411596212794,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017545519879123887,
+      "loss": 4.2827,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6844448476757674,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017544807007135243,
+      "loss": 4.2965,
+      "step": 6601
+    },
+    {
+      "epoch": 0.6845485357302555,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001754409404612585,
+      "loss": 4.2946,
+      "step": 6602
+    },
+    {
+      "epoch": 0.6846522237847436,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017543380996104123,
+      "loss": 4.2934,
+      "step": 6603
+    },
+    {
+      "epoch": 0.6847559118392317,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017542667857078472,
+      "loss": 4.3368,
+      "step": 6604
+    },
+    {
+      "epoch": 0.6848595998937197,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017541954629057314,
+      "loss": 4.3014,
+      "step": 6605
+    },
+    {
+      "epoch": 0.6849632879482078,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017541241312049062,
+      "loss": 4.312,
+      "step": 6606
+    },
+    {
+      "epoch": 0.6850669760026958,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017540527906062135,
+      "loss": 4.3363,
+      "step": 6607
+    },
+    {
+      "epoch": 0.685170664057184,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017539814411104949,
+      "loss": 4.2708,
+      "step": 6608
+    },
+    {
+      "epoch": 0.685274352111672,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017539100827185925,
+      "loss": 4.3057,
+      "step": 6609
+    },
+    {
+      "epoch": 0.6853780401661601,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001753838715431348,
+      "loss": 4.3111,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6854817282206482,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001753767339249603,
+      "loss": 4.2907,
+      "step": 6611
+    },
+    {
+      "epoch": 0.6855854162751363,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017536959541742007,
+      "loss": 4.2412,
+      "step": 6612
+    },
+    {
+      "epoch": 0.6856891043296244,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017536245602059827,
+      "loss": 4.2948,
+      "step": 6613
+    },
+    {
+      "epoch": 0.6857927923841124,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017535531573457914,
+      "loss": 4.3325,
+      "step": 6614
+    },
+    {
+      "epoch": 0.6858964804386005,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017534817455944698,
+      "loss": 4.2668,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6860001684930885,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017534103249528595,
+      "loss": 4.3183,
+      "step": 6616
+    },
+    {
+      "epoch": 0.6861038565475767,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001753338895421804,
+      "loss": 4.3117,
+      "step": 6617
+    },
+    {
+      "epoch": 0.6862075446020647,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017532674570021458,
+      "loss": 4.3003,
+      "step": 6618
+    },
+    {
+      "epoch": 0.6863112326565528,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001753196009694728,
+      "loss": 4.3061,
+      "step": 6619
+    },
+    {
+      "epoch": 0.6864149207110408,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017531245535003934,
+      "loss": 4.3236,
+      "step": 6620
+    },
+    {
+      "epoch": 0.686518608765529,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001753053088419985,
+      "loss": 4.3037,
+      "step": 6621
+    },
+    {
+      "epoch": 0.686622296820017,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017529816144543463,
+      "loss": 4.302,
+      "step": 6622
+    },
+    {
+      "epoch": 0.6867259848745051,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017529101316043203,
+      "loss": 4.3131,
+      "step": 6623
+    },
+    {
+      "epoch": 0.6868296729289931,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001752838639870751,
+      "loss": 4.3355,
+      "step": 6624
+    },
+    {
+      "epoch": 0.6869333609834812,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017527671392544812,
+      "loss": 4.3358,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6870370490379692,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001752695629756355,
+      "loss": 4.2912,
+      "step": 6626
+    },
+    {
+      "epoch": 0.6871407370924574,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017526241113772158,
+      "loss": 4.2923,
+      "step": 6627
+    },
+    {
+      "epoch": 0.6872444251469454,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017525525841179077,
+      "loss": 4.3022,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6873481132014335,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017524810479792747,
+      "loss": 4.3173,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6874518012559215,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001752409502962161,
+      "loss": 4.3274,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6875554893104097,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017523379490674102,
+      "loss": 4.3059,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6876591773648977,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017522663862958667,
+      "loss": 4.2875,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6877628654193858,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017521948146483754,
+      "loss": 4.2669,
+      "step": 6633
+    },
+    {
+      "epoch": 0.6878665534738738,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000175212323412578,
+      "loss": 4.3299,
+      "step": 6634
+    },
+    {
+      "epoch": 0.687970241528362,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017520516447289258,
+      "loss": 4.3181,
+      "step": 6635
+    },
+    {
+      "epoch": 0.68807392958285,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017519800464586572,
+      "loss": 4.3058,
+      "step": 6636
+    },
+    {
+      "epoch": 0.6881776176373381,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017519084393158185,
+      "loss": 4.3505,
+      "step": 6637
+    },
+    {
+      "epoch": 0.6882813056918261,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001751836823301255,
+      "loss": 4.3358,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6883849937463142,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017517651984158122,
+      "loss": 4.3308,
+      "step": 6639
+    },
+    {
+      "epoch": 0.6884886818008022,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017516935646603345,
+      "loss": 4.3068,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6885923698552904,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017516219220356673,
+      "loss": 4.3641,
+      "step": 6641
+    },
+    {
+      "epoch": 0.6886960579097784,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001751550270542656,
+      "loss": 4.2864,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6887997459642665,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017514786101821458,
+      "loss": 4.2931,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6889034340187545,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017514069409549823,
+      "loss": 4.2846,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6890071220732427,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017513352628620115,
+      "loss": 4.3249,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6891108101277307,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017512635759040784,
+      "loss": 4.2103,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6892144981822188,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017511918800820292,
+      "loss": 4.3311,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6893181862367068,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017511201753967102,
+      "loss": 4.3075,
+      "step": 6648
+    },
+    {
+      "epoch": 0.689421874291195,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017510484618489668,
+      "loss": 4.323,
+      "step": 6649
+    },
+    {
+      "epoch": 0.689525562345683,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017509767394396458,
+      "loss": 4.3195,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6896292504001711,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017509050081695925,
+      "loss": 4.3298,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6897329384546591,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017508332680396543,
+      "loss": 4.2887,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6898366265091472,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001750761519050677,
+      "loss": 4.3267,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6899403145636352,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001750689761203508,
+      "loss": 4.3083,
+      "step": 6654
+    },
+    {
+      "epoch": 0.6900440026181234,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017506179944989928,
+      "loss": 4.2902,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6901476906726115,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017505462189379783,
+      "loss": 4.3244,
+      "step": 6656
+    },
+    {
+      "epoch": 0.6902513787270995,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017504744345213122,
+      "loss": 4.3278,
+      "step": 6657
+    },
+    {
+      "epoch": 0.6903550667815876,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017504026412498412,
+      "loss": 4.3396,
+      "step": 6658
+    },
+    {
+      "epoch": 0.6904587548360757,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001750330839124412,
+      "loss": 4.2807,
+      "step": 6659
+    },
+    {
+      "epoch": 0.6905624428905638,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001750259028145872,
+      "loss": 4.2965,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6906661309450518,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017501872083150688,
+      "loss": 4.3143,
+      "step": 6661
+    },
+    {
+      "epoch": 0.6907698189995399,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017501153796328493,
+      "loss": 4.314,
+      "step": 6662
+    },
+    {
+      "epoch": 0.690873507054028,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017500435421000611,
+      "loss": 4.2631,
+      "step": 6663
+    },
+    {
+      "epoch": 0.6909771951085161,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017499716957175524,
+      "loss": 4.277,
+      "step": 6664
+    },
+    {
+      "epoch": 0.6910808831630041,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017498998404861702,
+      "loss": 4.3147,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6911845712174922,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017498279764067623,
+      "loss": 4.2968,
+      "step": 6666
+    },
+    {
+      "epoch": 0.6912882592719802,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017497561034801772,
+      "loss": 4.3201,
+      "step": 6667
+    },
+    {
+      "epoch": 0.6913919473264684,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017496842217072626,
+      "loss": 4.2557,
+      "step": 6668
+    },
+    {
+      "epoch": 0.6914956353809564,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017496123310888667,
+      "loss": 4.3024,
+      "step": 6669
+    },
+    {
+      "epoch": 0.6915993234354445,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017495404316258376,
+      "loss": 4.3402,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6917030114899325,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017494685233190234,
+      "loss": 4.3003,
+      "step": 6671
+    },
+    {
+      "epoch": 0.6918066995444206,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001749396606169273,
+      "loss": 4.2904,
+      "step": 6672
+    },
+    {
+      "epoch": 0.6919103875989087,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001749324680177435,
+      "loss": 4.2703,
+      "step": 6673
+    },
+    {
+      "epoch": 0.6920140756533968,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017492527453443578,
+      "loss": 4.3059,
+      "step": 6674
+    },
+    {
+      "epoch": 0.6921177637078848,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00017491808016708899,
+      "loss": 4.2864,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6922214517623729,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017491088491578807,
+      "loss": 4.2779,
+      "step": 6676
+    },
+    {
+      "epoch": 0.692325139816861,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001749036887806179,
+      "loss": 4.3136,
+      "step": 6677
+    },
+    {
+      "epoch": 0.6924288278713491,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017489649176166336,
+      "loss": 4.32,
+      "step": 6678
+    },
+    {
+      "epoch": 0.6925325159258371,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001748892938590094,
+      "loss": 4.3044,
+      "step": 6679
+    },
+    {
+      "epoch": 0.6926362039803252,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017488209507274095,
+      "loss": 4.3141,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6927398920348132,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001748748954029429,
+      "loss": 4.3011,
+      "step": 6681
+    },
+    {
+      "epoch": 0.6928435800893014,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017486769484970026,
+      "loss": 4.2975,
+      "step": 6682
+    },
+    {
+      "epoch": 0.6929472681437894,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001748604934130979,
+      "loss": 4.323,
+      "step": 6683
+    },
+    {
+      "epoch": 0.6930509561982775,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017485329109322089,
+      "loss": 4.3262,
+      "step": 6684
+    },
+    {
+      "epoch": 0.6931546442527655,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00017484608789015418,
+      "loss": 4.2892,
+      "step": 6685
+    },
+    {
+      "epoch": 0.6932583323072536,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001748388838039827,
+      "loss": 4.3314,
+      "step": 6686
+    },
+    {
+      "epoch": 0.6933620203617417,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017483167883479157,
+      "loss": 4.3239,
+      "step": 6687
+    },
+    {
+      "epoch": 0.6934657084162298,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001748244729826657,
+      "loss": 4.2554,
+      "step": 6688
+    },
+    {
+      "epoch": 0.6935693964707178,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017481726624769012,
+      "loss": 4.301,
+      "step": 6689
+    },
+    {
+      "epoch": 0.6936730845252059,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001748100586299499,
+      "loss": 4.2807,
+      "step": 6690
+    },
+    {
+      "epoch": 0.693776772579694,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017480285012953006,
+      "loss": 4.3359,
+      "step": 6691
+    },
+    {
+      "epoch": 0.6938804606341821,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017479564074651568,
+      "loss": 4.3141,
+      "step": 6692
+    },
+    {
+      "epoch": 0.6939841486886701,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017478843048099178,
+      "loss": 4.2948,
+      "step": 6693
+    },
+    {
+      "epoch": 0.6940878367431582,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017478121933304345,
+      "loss": 4.3005,
+      "step": 6694
+    },
+    {
+      "epoch": 0.6941915247976462,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001747740073027558,
+      "loss": 4.3416,
+      "step": 6695
+    },
+    {
+      "epoch": 0.6942952128521344,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001747667943902139,
+      "loss": 4.3027,
+      "step": 6696
+    },
+    {
+      "epoch": 0.6943989009066224,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017475958059550285,
+      "loss": 4.3,
+      "step": 6697
+    },
+    {
+      "epoch": 0.6945025889611105,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001747523659187078,
+      "loss": 4.3291,
+      "step": 6698
+    },
+    {
+      "epoch": 0.6946062770155985,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001747451503599138,
+      "loss": 4.3234,
+      "step": 6699
+    },
+    {
+      "epoch": 0.6947099650700866,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017473793391920608,
+      "loss": 4.3066,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6948136531245748,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001747307165966697,
+      "loss": 4.3168,
+      "step": 6701
+    },
+    {
+      "epoch": 0.6949173411790628,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017472349839238989,
+      "loss": 4.3572,
+      "step": 6702
+    },
+    {
+      "epoch": 0.6950210292335509,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017471627930645175,
+      "loss": 4.2868,
+      "step": 6703
+    },
+    {
+      "epoch": 0.6951247172880389,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017470905933894052,
+      "loss": 4.3121,
+      "step": 6704
+    },
+    {
+      "epoch": 0.6952284053425271,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017470183848994139,
+      "loss": 4.3402,
+      "step": 6705
+    },
+    {
+      "epoch": 0.6953320933970151,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001746946167595395,
+      "loss": 4.301,
+      "step": 6706
+    },
+    {
+      "epoch": 0.6954357814515032,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017468739414782007,
+      "loss": 4.2396,
+      "step": 6707
+    },
+    {
+      "epoch": 0.6955394695059912,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017468017065486836,
+      "loss": 4.2664,
+      "step": 6708
+    },
+    {
+      "epoch": 0.6956431575604793,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017467294628076955,
+      "loss": 4.303,
+      "step": 6709
+    },
+    {
+      "epoch": 0.6957468456149674,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017466572102560894,
+      "loss": 4.3181,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6958505336694555,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001746584948894717,
+      "loss": 4.329,
+      "step": 6711
+    },
+    {
+      "epoch": 0.6959542217239435,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001746512678724432,
+      "loss": 4.2602,
+      "step": 6712
+    },
+    {
+      "epoch": 0.6960579097784316,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001746440399746086,
+      "loss": 4.2684,
+      "step": 6713
+    },
+    {
+      "epoch": 0.6961615978329196,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017463681119605324,
+      "loss": 4.3137,
+      "step": 6714
+    },
+    {
+      "epoch": 0.6962652858874078,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017462958153686243,
+      "loss": 4.3403,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6963689739418958,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017462235099712143,
+      "loss": 4.313,
+      "step": 6716
+    },
+    {
+      "epoch": 0.6964726619963839,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017461511957691554,
+      "loss": 4.2659,
+      "step": 6717
+    },
+    {
+      "epoch": 0.6965763500508719,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017460788727633014,
+      "loss": 4.3121,
+      "step": 6718
+    },
+    {
+      "epoch": 0.69668003810536,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017460065409545053,
+      "loss": 4.2943,
+      "step": 6719
+    },
+    {
+      "epoch": 0.6967837261598481,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017459342003436204,
+      "loss": 4.2951,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6968874142143362,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017458618509315005,
+      "loss": 4.3544,
+      "step": 6721
+    },
+    {
+      "epoch": 0.6969911022688242,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017457894927189996,
+      "loss": 4.286,
+      "step": 6722
+    },
+    {
+      "epoch": 0.6970947903233123,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017457171257069707,
+      "loss": 4.3755,
+      "step": 6723
+    },
+    {
+      "epoch": 0.6971984783778004,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017456447498962678,
+      "loss": 4.3603,
+      "step": 6724
+    },
+    {
+      "epoch": 0.6973021664322885,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001745572365287745,
+      "loss": 4.3034,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6974058544867765,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017454999718822566,
+      "loss": 4.3108,
+      "step": 6726
+    },
+    {
+      "epoch": 0.6975095425412646,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017454275696806567,
+      "loss": 4.2932,
+      "step": 6727
+    },
+    {
+      "epoch": 0.6976132305957526,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001745355158683799,
+      "loss": 4.2865,
+      "step": 6728
+    },
+    {
+      "epoch": 0.6977169186502408,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017452827388925388,
+      "loss": 4.3343,
+      "step": 6729
+    },
+    {
+      "epoch": 0.6978206067047288,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000174521031030773,
+      "loss": 4.2954,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6979242947592169,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017451378729302271,
+      "loss": 4.3043,
+      "step": 6731
+    },
+    {
+      "epoch": 0.6980279828137049,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017450654267608847,
+      "loss": 4.3119,
+      "step": 6732
+    },
+    {
+      "epoch": 0.698131670868193,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017449929718005582,
+      "loss": 4.2998,
+      "step": 6733
+    },
+    {
+      "epoch": 0.6982353589226811,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017449205080501018,
+      "loss": 4.2894,
+      "step": 6734
+    },
+    {
+      "epoch": 0.6983390469771692,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001744848035510371,
+      "loss": 4.2802,
+      "step": 6735
+    },
+    {
+      "epoch": 0.6984427350316572,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017447755541822208,
+      "loss": 4.2742,
+      "step": 6736
+    },
+    {
+      "epoch": 0.6985464230861453,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017447030640665062,
+      "loss": 4.2813,
+      "step": 6737
+    },
+    {
+      "epoch": 0.6986501111406334,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017446305651640825,
+      "loss": 4.2958,
+      "step": 6738
+    },
+    {
+      "epoch": 0.6987537991951215,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017445580574758056,
+      "loss": 4.2717,
+      "step": 6739
+    },
+    {
+      "epoch": 0.6988574872496095,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017444855410025305,
+      "loss": 4.3151,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6989611753040976,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001744413015745113,
+      "loss": 4.3139,
+      "step": 6741
+    },
+    {
+      "epoch": 0.6990648633585856,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017443404817044087,
+      "loss": 4.3068,
+      "step": 6742
+    },
+    {
+      "epoch": 0.6991685514130738,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017442679388812737,
+      "loss": 4.3164,
+      "step": 6743
+    },
+    {
+      "epoch": 0.6992722394675618,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017441953872765638,
+      "loss": 4.2802,
+      "step": 6744
+    },
+    {
+      "epoch": 0.6993759275220499,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017441228268911347,
+      "loss": 4.332,
+      "step": 6745
+    },
+    {
+      "epoch": 0.699479615576538,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017440502577258427,
+      "loss": 4.2957,
+      "step": 6746
+    },
+    {
+      "epoch": 0.699583303631026,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017439776797815445,
+      "loss": 4.3194,
+      "step": 6747
+    },
+    {
+      "epoch": 0.6996869916855142,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00017439050930590958,
+      "loss": 4.2783,
+      "step": 6748
+    },
+    {
+      "epoch": 0.6997906797400022,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017438324975593538,
+      "loss": 4.2853,
+      "step": 6749
+    },
+    {
+      "epoch": 0.6998943677944903,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001743759893283174,
+      "loss": 4.2653,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6999980558489783,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00017436872802314141,
+      "loss": 4.309,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7001017439034665,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017436146584049302,
+      "loss": 4.3477,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7002054319579545,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00017435420278045794,
+      "loss": 4.3174,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7003091200124426,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017434693884312184,
+      "loss": 4.2983,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7004128080669306,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001743396740285705,
+      "loss": 4.3341,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7005164961214188,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017433240833688955,
+      "loss": 4.3021,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7006201841759068,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017432514176816478,
+      "loss": 4.3089,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7007238722303949,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017431787432248188,
+      "loss": 4.3165,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7008275602848829,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017431060599992662,
+      "loss": 4.321,
+      "step": 6759
+    },
+    {
+      "epoch": 0.700931248339371,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017430333680058476,
+      "loss": 4.3034,
+      "step": 6760
+    },
+    {
+      "epoch": 0.701034936393859,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001742960667245421,
+      "loss": 4.2945,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7011386244483472,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00017428879577188435,
+      "loss": 4.2914,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7012423125028352,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017428152394269731,
+      "loss": 4.3338,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7013460005573233,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017427425123706688,
+      "loss": 4.3143,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7014496886118113,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017426697765507876,
+      "loss": 4.2955,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7015533766662995,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017425970319681882,
+      "loss": 4.2712,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7016570647207875,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017425242786237285,
+      "loss": 4.2621,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7017607527752756,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017424515165182674,
+      "loss": 4.3243,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7018644408297636,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017423787456526634,
+      "loss": 4.2579,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7019681288842518,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017423059660277742,
+      "loss": 4.2895,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7020718169387398,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00017422331776444598,
+      "loss": 4.3058,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7021755049932279,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00017421603805035785,
+      "loss": 4.2943,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7022791930477159,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017420875746059893,
+      "loss": 4.2972,
+      "step": 6773
+    },
+    {
+      "epoch": 0.702382881102204,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001742014759952551,
+      "loss": 4.3516,
+      "step": 6774
+    },
+    {
+      "epoch": 0.702486569156692,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017419419365441227,
+      "loss": 4.3031,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7025902572111802,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001741869104381564,
+      "loss": 4.3006,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7026939452656682,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001741796263465734,
+      "loss": 4.3177,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7027976333201563,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017417234137974923,
+      "loss": 4.354,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7029013213746443,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017416505553776983,
+      "loss": 4.3367,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7030050094291325,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017415776882072118,
+      "loss": 4.2742,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7031086974836205,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00017415048122868923,
+      "loss": 4.2843,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7032123855381086,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017414319276175995,
+      "loss": 4.3533,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7033160735925966,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00017413590342001944,
+      "loss": 4.3043,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7034197616470848,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001741286132035536,
+      "loss": 4.3679,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7035234497015728,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017412132211244846,
+      "loss": 4.2908,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7036271377560609,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001741140301467901,
+      "loss": 4.2883,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7037308258105489,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017410673730666452,
+      "loss": 4.3207,
+      "step": 6787
+    },
+    {
+      "epoch": 0.703834513865037,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017409944359215779,
+      "loss": 4.2686,
+      "step": 6788
+    },
+    {
+      "epoch": 0.7039382019195252,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017409214900335592,
+      "loss": 4.3197,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7040418899740132,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000174084853540345,
+      "loss": 4.2796,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7041455780285013,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00017407755720321116,
+      "loss": 4.2983,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7042492660829893,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017407025999204042,
+      "loss": 4.2933,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7043529541374774,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017406296190691892,
+      "loss": 4.3228,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7044566421919655,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017405566294793277,
+      "loss": 4.2952,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7045603302464536,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017404836311516806,
+      "loss": 4.3146,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7046640183009416,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017404106240871093,
+      "loss": 4.316,
+      "step": 6796
+    },
+    {
+      "epoch": 0.7047677063554297,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017403376082864754,
+      "loss": 4.2951,
+      "step": 6797
+    },
+    {
+      "epoch": 0.7048713944099178,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000174026458375064,
+      "loss": 4.2902,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7049750824644059,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017401915504804656,
+      "loss": 4.2828,
+      "step": 6799
+    },
+    {
+      "epoch": 0.7050787705188939,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001740118508476813,
+      "loss": 4.34,
+      "step": 6800
+    },
+    {
+      "epoch": 0.705182458573382,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017400454577405443,
+      "loss": 4.3162,
+      "step": 6801
+    },
+    {
+      "epoch": 0.70528614662787,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017399723982725217,
+      "loss": 4.3065,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7053898346823582,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017398993300736065,
+      "loss": 4.2622,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7054935227368462,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017398262531446616,
+      "loss": 4.3342,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7055972107913343,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001739753167486549,
+      "loss": 4.2971,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7057008988458223,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001739680073100131,
+      "loss": 4.3237,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7058045869003104,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.000173960696998627,
+      "loss": 4.3068,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7059082749547985,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017395338581458286,
+      "loss": 4.2803,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7060119630092866,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017394607375796693,
+      "loss": 4.2956,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7061156510637746,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017393876082886546,
+      "loss": 4.3073,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7062193391182627,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001739314470273648,
+      "loss": 4.2763,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7063230271727508,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017392413235355124,
+      "loss": 4.3287,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7064267152272389,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000173916816807511,
+      "loss": 4.2653,
+      "step": 6813
+    },
+    {
+      "epoch": 0.7065304032817269,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001739095003893305,
+      "loss": 4.3285,
+      "step": 6814
+    },
+    {
+      "epoch": 0.706634091336215,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017390218309909603,
+      "loss": 4.3328,
+      "step": 6815
+    },
+    {
+      "epoch": 0.706737779390703,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017389486493689388,
+      "loss": 4.2716,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7068414674451912,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00017388754590281046,
+      "loss": 4.2941,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7069451554996792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001738802259969321,
+      "loss": 4.3477,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7070488435541673,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00017387290521934517,
+      "loss": 4.2487,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7071525316086553,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017386558357013602,
+      "loss": 4.2578,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7072562196631434,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00017385826104939108,
+      "loss": 4.2883,
+      "step": 6821
+    },
+    {
+      "epoch": 0.7073599077176315,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017385093765719673,
+      "loss": 4.2993,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7074635957721196,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001738436133936394,
+      "loss": 4.2944,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7075672838266076,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017383628825880546,
+      "loss": 4.2836,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7076709718810957,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001738289622527814,
+      "loss": 4.2791,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7077746599355837,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017382163537565357,
+      "loss": 4.2917,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7078783479900719,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001738143076275085,
+      "loss": 4.2644,
+      "step": 6827
+    },
+    {
+      "epoch": 0.7079820360445599,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017380697900843263,
+      "loss": 4.3196,
+      "step": 6828
+    },
+    {
+      "epoch": 0.708085724099048,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017379964951851244,
+      "loss": 4.3053,
+      "step": 6829
+    },
+    {
+      "epoch": 0.708189412153536,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001737923191578344,
+      "loss": 4.2747,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7082931002080242,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017378498792648496,
+      "loss": 4.2809,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7083967882625122,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017377765582455069,
+      "loss": 4.3281,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7085004763170003,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017377032285211804,
+      "loss": 4.3215,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7086041643714884,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017376298900927356,
+      "loss": 4.3022,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7087078524259764,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001737556542961038,
+      "loss": 4.2739,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7088115404804646,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017374831871269528,
+      "loss": 4.3283,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7089152285349526,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017374098225913454,
+      "loss": 4.2811,
+      "step": 6837
+    },
+    {
+      "epoch": 0.7090189165894407,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017373364493550815,
+      "loss": 4.3492,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7091226046439287,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017372630674190274,
+      "loss": 4.249,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7092262926984169,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017371896767840478,
+      "loss": 4.3491,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7093299807529049,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017371162774510097,
+      "loss": 4.3126,
+      "step": 6841
+    },
+    {
+      "epoch": 0.709433668807393,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017370428694207782,
+      "loss": 4.3509,
+      "step": 6842
+    },
+    {
+      "epoch": 0.709537356861881,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017369694526942208,
+      "loss": 4.2861,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7096410449163691,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017368960272722022,
+      "loss": 4.3139,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7097447329708572,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017368225931555892,
+      "loss": 4.3315,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7098484210253453,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001736749150345249,
+      "loss": 4.2999,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7099521090798333,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017366756988420473,
+      "loss": 4.3054,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7100557971343214,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017366022386468513,
+      "loss": 4.2595,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7101594851888094,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017365287697605273,
+      "loss": 4.2923,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7102631732432976,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017364552921839423,
+      "loss": 4.3015,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7103668612977856,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017363818059179634,
+      "loss": 4.3324,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7104705493522737,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017363083109634577,
+      "loss": 4.3018,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7105742374067617,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001736234807321292,
+      "loss": 4.2822,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7106779254612499,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017361612949923344,
+      "loss": 4.3018,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7107816135157379,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001736087773977451,
+      "loss": 4.309,
+      "step": 6855
+    },
+    {
+      "epoch": 0.710885301570226,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017360142442775104,
+      "loss": 4.3042,
+      "step": 6856
+    },
+    {
+      "epoch": 0.710988989624714,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017359407058933793,
+      "loss": 4.3146,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7110926776792021,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017358671588259261,
+      "loss": 4.298,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7111963657336902,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017357936030760183,
+      "loss": 4.3325,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7113000537881783,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017357200386445233,
+      "loss": 4.3342,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7114037418426663,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.000173564646553231,
+      "loss": 4.321,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7115074298971544,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017355728837402458,
+      "loss": 4.274,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7116111179516424,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017354992932691992,
+      "loss": 4.2502,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7117148060061306,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017354256941200385,
+      "loss": 4.2805,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7118184940606186,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017353520862936316,
+      "loss": 4.2811,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7119221821151067,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017352784697908478,
+      "loss": 4.3037,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7120258701695947,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017352048446125551,
+      "loss": 4.2977,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7121295582240829,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017351312107596225,
+      "loss": 4.3676,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7122332462785709,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017350575682329185,
+      "loss": 4.3402,
+      "step": 6869
+    },
+    {
+      "epoch": 0.712336934333059,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001734983917033312,
+      "loss": 4.3002,
+      "step": 6870
+    },
+    {
+      "epoch": 0.712440622387547,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017349102571616727,
+      "loss": 4.3039,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7125443104420351,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001734836588618869,
+      "loss": 4.3131,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7126479984965232,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017347629114057705,
+      "loss": 4.3066,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7127516865510113,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001734689225523246,
+      "loss": 4.2851,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7128553746054993,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017346155309721652,
+      "loss": 4.2915,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7129590626599874,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017345418277533978,
+      "loss": 4.3027,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7130627507144754,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017344681158678132,
+      "loss": 4.3315,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7131664387689636,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001734394395316281,
+      "loss": 4.3191,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7132701268234517,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001734320666099672,
+      "loss": 4.2969,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7133738148779397,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017342469282188546,
+      "loss": 4.3088,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7134775029324278,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017341731816746997,
+      "loss": 4.2874,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7135811909869159,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017340994264680774,
+      "loss": 4.2924,
+      "step": 6882
+    },
+    {
+      "epoch": 0.713684879041404,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001734025662599858,
+      "loss": 4.31,
+      "step": 6883
+    },
+    {
+      "epoch": 0.713788567095892,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017339518900709116,
+      "loss": 4.3129,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7138922551503801,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017338781088821085,
+      "loss": 4.3051,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7139959432048681,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017338043190343196,
+      "loss": 4.3367,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7140996312593563,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017337305205284155,
+      "loss": 4.326,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7142033193138443,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017336567133652668,
+      "loss": 4.3013,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7143070073683324,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017335828975457445,
+      "loss": 4.3218,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7144106954228204,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001733509073070719,
+      "loss": 4.2486,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7145143834773086,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017334352399410623,
+      "loss": 4.285,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7146180715317966,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001733361398157645,
+      "loss": 4.2822,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7147217595862847,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017332875477213378,
+      "loss": 4.3236,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7148254476407727,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017332136886330136,
+      "loss": 4.2788,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7149291356952608,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001733139820893542,
+      "loss": 4.3271,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7150328237497489,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001733065944503796,
+      "loss": 4.2728,
+      "step": 6896
+    },
+    {
+      "epoch": 0.715136511804237,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017329920594646466,
+      "loss": 4.3192,
+      "step": 6897
+    },
+    {
+      "epoch": 0.715240199858725,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017329181657769658,
+      "loss": 4.2962,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7153438879132131,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017328442634416253,
+      "loss": 4.3042,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7154475759677011,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017327703524594971,
+      "loss": 4.3336,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7155512640221893,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017326964328314532,
+      "loss": 4.2664,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001732622504558366,
+      "loss": 4.2763,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7157586401311654,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001732548567641108,
+      "loss": 4.2979,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7158623281856534,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00017324746220805508,
+      "loss": 4.3024,
+      "step": 6904
+    },
+    {
+      "epoch": 0.7159660162401416,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017324006678775674,
+      "loss": 4.3234,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7160697042946296,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017323267050330302,
+      "loss": 4.3004,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7161733923491177,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017322527335478122,
+      "loss": 4.3169,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7162770804036057,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017321787534227862,
+      "loss": 4.3255,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7163807684580938,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017321047646588243,
+      "loss": 4.3095,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7164844565125819,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017320307672568004,
+      "loss": 4.3014,
+      "step": 6910
+    },
+    {
+      "epoch": 0.71658814456707,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017319567612175872,
+      "loss": 4.3117,
+      "step": 6911
+    },
+    {
+      "epoch": 0.716691832621558,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017318827465420577,
+      "loss": 4.273,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7167955206760461,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00017318087232310857,
+      "loss": 4.2808,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7168992087305341,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017317346912855443,
+      "loss": 4.2904,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7170028967850223,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001731660650706307,
+      "loss": 4.3138,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7171065848395103,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017315866014942473,
+      "loss": 4.2877,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7172102728939984,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001731512543650239,
+      "loss": 4.3118,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7173139609484864,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017314384771751563,
+      "loss": 4.3041,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7174176490029746,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017313644020698724,
+      "loss": 4.282,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7175213370574626,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017312903183352616,
+      "loss": 4.266,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7176250251119507,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001731216225972198,
+      "loss": 4.2921,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7177287131664387,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001731142124981556,
+      "loss": 4.2855,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7178324012209268,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017310680153642097,
+      "loss": 4.2506,
+      "step": 6923
+    },
+    {
+      "epoch": 0.717936089275415,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017309938971210337,
+      "loss": 4.3639,
+      "step": 6924
+    },
+    {
+      "epoch": 0.718039777329903,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017309197702529026,
+      "loss": 4.2782,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7181434653843911,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017308456347606905,
+      "loss": 4.332,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7182471534388791,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017307714906452724,
+      "loss": 4.3266,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7183508414933673,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017306973379075234,
+      "loss": 4.3125,
+      "step": 6928
+    },
+    {
+      "epoch": 0.7184545295478553,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001730623176548318,
+      "loss": 4.2936,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7185582176023434,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017305490065685318,
+      "loss": 4.3082,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7186619056568314,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017304748279690392,
+      "loss": 4.3166,
+      "step": 6931
+    },
+    {
+      "epoch": 0.7187655937113195,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001730400640750716,
+      "loss": 4.3189,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7188692817658076,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017303264449144369,
+      "loss": 4.2606,
+      "step": 6933
+    },
+    {
+      "epoch": 0.7189729698202957,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001730252240461078,
+      "loss": 4.28,
+      "step": 6934
+    },
+    {
+      "epoch": 0.7190766578747837,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017301780273915144,
+      "loss": 4.3002,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7191803459292718,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017301038057066222,
+      "loss": 4.3399,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7192840339837598,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017300295754072768,
+      "loss": 4.2772,
+      "step": 6937
+    },
+    {
+      "epoch": 0.719387722038248,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001729955336494354,
+      "loss": 4.319,
+      "step": 6938
+    },
+    {
+      "epoch": 0.719491410092736,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017298810889687297,
+      "loss": 4.3156,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7195950981472241,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017298068328312804,
+      "loss": 4.3124,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7196987862017121,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017297325680828816,
+      "loss": 4.323,
+      "step": 6941
+    },
+    {
+      "epoch": 0.7198024742562003,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.000172965829472441,
+      "loss": 4.3011,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7199061623106883,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017295840127567417,
+      "loss": 4.2752,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7200098503651764,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017295097221807534,
+      "loss": 4.2404,
+      "step": 6944
+    },
+    {
+      "epoch": 0.7201135384196644,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017294354229973214,
+      "loss": 4.2955,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7202172264741525,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017293611152073224,
+      "loss": 4.3117,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7203209145286406,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017292867988116334,
+      "loss": 4.307,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7204246025831287,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017292124738111308,
+      "loss": 4.3032,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7205282906376167,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017291381402066922,
+      "loss": 4.306,
+      "step": 6949
+    },
+    {
+      "epoch": 0.7206319786921048,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001729063797999194,
+      "loss": 4.2715,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7207356667465928,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001728989447189514,
+      "loss": 4.3337,
+      "step": 6951
+    },
+    {
+      "epoch": 0.720839354801081,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00017289150877785282,
+      "loss": 4.2688,
+      "step": 6952
+    },
+    {
+      "epoch": 0.720943042855569,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017288407197671156,
+      "loss": 4.3162,
+      "step": 6953
+    },
+    {
+      "epoch": 0.7210467309100571,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017287663431561528,
+      "loss": 4.2812,
+      "step": 6954
+    },
+    {
+      "epoch": 0.7211504189645451,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017286919579465176,
+      "loss": 4.3023,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7212541070190333,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017286175641390873,
+      "loss": 4.2777,
+      "step": 6956
+    },
+    {
+      "epoch": 0.7213577950735213,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000172854316173474,
+      "loss": 4.3601,
+      "step": 6957
+    },
+    {
+      "epoch": 0.7214614831280094,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001728468750734354,
+      "loss": 4.3323,
+      "step": 6958
+    },
+    {
+      "epoch": 0.7215651711824974,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001728394331138806,
+      "loss": 4.2778,
+      "step": 6959
+    },
+    {
+      "epoch": 0.7216688592369855,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017283199029489752,
+      "loss": 4.3155,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7217725472914736,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017282454661657391,
+      "loss": 4.2807,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7218762353459617,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017281710207899765,
+      "loss": 4.3196,
+      "step": 6962
+    },
+    {
+      "epoch": 0.7219799234004497,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001728096566822566,
+      "loss": 4.3321,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7220836114549378,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001728022104264385,
+      "loss": 4.2515,
+      "step": 6964
+    },
+    {
+      "epoch": 0.7221872995094258,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001727947633116313,
+      "loss": 4.2605,
+      "step": 6965
+    },
+    {
+      "epoch": 0.722290987563914,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017278731533792283,
+      "loss": 4.3116,
+      "step": 6966
+    },
+    {
+      "epoch": 0.722394675618402,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.000172779866505401,
+      "loss": 4.3079,
+      "step": 6967
+    },
+    {
+      "epoch": 0.7224983636728901,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00017277241681415366,
+      "loss": 4.2762,
+      "step": 6968
+    },
+    {
+      "epoch": 0.7226020517273782,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017276496626426874,
+      "loss": 4.2775,
+      "step": 6969
+    },
+    {
+      "epoch": 0.7227057397818663,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001727575148558341,
+      "loss": 4.3017,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7228094278363544,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017275006258893775,
+      "loss": 4.2978,
+      "step": 6971
+    },
+    {
+      "epoch": 0.7229131158908424,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001727426094636675,
+      "loss": 4.2782,
+      "step": 6972
+    },
+    {
+      "epoch": 0.7230168039453305,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017273515548011138,
+      "loss": 4.2956,
+      "step": 6973
+    },
+    {
+      "epoch": 0.7231204919998185,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017272770063835732,
+      "loss": 4.2611,
+      "step": 6974
+    },
+    {
+      "epoch": 0.7232241800543067,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017272024493849325,
+      "loss": 4.2621,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7233278681087947,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017271278838060719,
+      "loss": 4.3214,
+      "step": 6976
+    },
+    {
+      "epoch": 0.7234315561632828,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017270533096478704,
+      "loss": 4.3202,
+      "step": 6977
+    },
+    {
+      "epoch": 0.7235352442177708,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017269787269112084,
+      "loss": 4.2535,
+      "step": 6978
+    },
+    {
+      "epoch": 0.723638932272259,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001726904135596966,
+      "loss": 4.2887,
+      "step": 6979
+    },
+    {
+      "epoch": 0.723742620326747,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017268295357060235,
+      "loss": 4.3077,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7238463083812351,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017267549272392607,
+      "loss": 4.289,
+      "step": 6981
+    },
+    {
+      "epoch": 0.7239499964357231,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017266803101975576,
+      "loss": 4.2946,
+      "step": 6982
+    },
+    {
+      "epoch": 0.7240536844902112,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001726605684581795,
+      "loss": 4.262,
+      "step": 6983
+    },
+    {
+      "epoch": 0.7241573725446993,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001726531050392854,
+      "loss": 4.3065,
+      "step": 6984
+    },
+    {
+      "epoch": 0.7242610605991874,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017264564076316136,
+      "loss": 4.2718,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7243647486536754,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017263817562989563,
+      "loss": 4.326,
+      "step": 6986
+    },
+    {
+      "epoch": 0.7244684367081635,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001726307096395762,
+      "loss": 4.2844,
+      "step": 6987
+    },
+    {
+      "epoch": 0.7245721247626515,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017262324279229113,
+      "loss": 4.3137,
+      "step": 6988
+    },
+    {
+      "epoch": 0.7246758128171397,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001726157750881286,
+      "loss": 4.2779,
+      "step": 6989
+    },
+    {
+      "epoch": 0.7247795008716277,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017260830652717665,
+      "loss": 4.364,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7248831889261158,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017260083710952343,
+      "loss": 4.298,
+      "step": 6991
+    },
+    {
+      "epoch": 0.7249868769806038,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001725933668352571,
+      "loss": 4.3133,
+      "step": 6992
+    },
+    {
+      "epoch": 0.725090565035092,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017258589570446576,
+      "loss": 4.3327,
+      "step": 6993
+    },
+    {
+      "epoch": 0.72519425308958,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001725784237172376,
+      "loss": 4.3136,
+      "step": 6994
+    },
+    {
+      "epoch": 0.7252979411440681,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017257095087366074,
+      "loss": 4.3083,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7254016291985561,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017256347717382338,
+      "loss": 4.2735,
+      "step": 6996
+    },
+    {
+      "epoch": 0.7255053172530442,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001725560026178137,
+      "loss": 4.3301,
+      "step": 6997
+    },
+    {
+      "epoch": 0.7256090053075323,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017254852720571982,
+      "loss": 4.2935,
+      "step": 6998
+    },
+    {
+      "epoch": 0.7257126933620204,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017254105093763008,
+      "loss": 4.2975,
+      "step": 6999
+    },
+    {
+      "epoch": 0.7258163814165084,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001725335738136326,
+      "loss": 4.2722,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7259200694709965,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001725260958338156,
+      "loss": 4.297,
+      "step": 7001
+    },
+    {
+      "epoch": 0.7260237575254845,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00017251861699826737,
+      "loss": 4.2929,
+      "step": 7002
+    },
+    {
+      "epoch": 0.7261274455799727,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017251113730707608,
+      "loss": 4.3054,
+      "step": 7003
+    },
+    {
+      "epoch": 0.7262311336344607,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017250365676033002,
+      "loss": 4.2856,
+      "step": 7004
+    },
+    {
+      "epoch": 0.7263348216889488,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001724961753581175,
+      "loss": 4.2623,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7264385097434368,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001724886931005267,
+      "loss": 4.3196,
+      "step": 7006
+    },
+    {
+      "epoch": 0.726542197797925,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017248120998764591,
+      "loss": 4.3348,
+      "step": 7007
+    },
+    {
+      "epoch": 0.726645885852413,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001724737260195635,
+      "loss": 4.2863,
+      "step": 7008
+    },
+    {
+      "epoch": 0.7267495739069011,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017246624119636773,
+      "loss": 4.2792,
+      "step": 7009
+    },
+    {
+      "epoch": 0.7268532619613891,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017245875551814689,
+      "loss": 4.3015,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7269569500158772,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017245126898498934,
+      "loss": 4.3235,
+      "step": 7011
+    },
+    {
+      "epoch": 0.7270606380703653,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017244378159698344,
+      "loss": 4.2926,
+      "step": 7012
+    },
+    {
+      "epoch": 0.7271643261248534,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017243629335421748,
+      "loss": 4.2849,
+      "step": 7013
+    },
+    {
+      "epoch": 0.7272680141793415,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001724288042567798,
+      "loss": 4.2794,
+      "step": 7014
+    },
+    {
+      "epoch": 0.7273717022338295,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017242131430475878,
+      "loss": 4.2771,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7274753902883176,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017241382349824283,
+      "loss": 4.3026,
+      "step": 7016
+    },
+    {
+      "epoch": 0.7275790783428057,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017240633183732032,
+      "loss": 4.3207,
+      "step": 7017
+    },
+    {
+      "epoch": 0.7276827663972938,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001723988393220796,
+      "loss": 4.3208,
+      "step": 7018
+    },
+    {
+      "epoch": 0.7277864544517818,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001723913459526091,
+      "loss": 4.3088,
+      "step": 7019
+    },
+    {
+      "epoch": 0.7278901425062699,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017238385172899727,
+      "loss": 4.257,
+      "step": 7020
+    },
+    {
+      "epoch": 0.727993830560758,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017237635665133248,
+      "loss": 4.2831,
+      "step": 7021
+    },
+    {
+      "epoch": 0.7280975186152461,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001723688607197032,
+      "loss": 4.2535,
+      "step": 7022
+    },
+    {
+      "epoch": 0.7282012066697341,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017236136393419783,
+      "loss": 4.3271,
+      "step": 7023
+    },
+    {
+      "epoch": 0.7283048947242222,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001723538662949049,
+      "loss": 4.2639,
+      "step": 7024
+    },
+    {
+      "epoch": 0.7284085827787102,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017234636780191282,
+      "loss": 4.2348,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7285122708331984,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017233886845531004,
+      "loss": 4.3681,
+      "step": 7026
+    },
+    {
+      "epoch": 0.7286159588876864,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017233136825518512,
+      "loss": 4.2736,
+      "step": 7027
+    },
+    {
+      "epoch": 0.7287196469421745,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017232386720162648,
+      "loss": 4.3037,
+      "step": 7028
+    },
+    {
+      "epoch": 0.7288233349966625,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017231636529472266,
+      "loss": 4.2867,
+      "step": 7029
+    },
+    {
+      "epoch": 0.7289270230511506,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017230886253456217,
+      "loss": 4.2942,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7290307111056387,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017230135892123358,
+      "loss": 4.3403,
+      "step": 7031
+    },
+    {
+      "epoch": 0.7291343991601268,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00017229385445482532,
+      "loss": 4.3001,
+      "step": 7032
+    },
+    {
+      "epoch": 0.7292380872146148,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017228634913542604,
+      "loss": 4.3063,
+      "step": 7033
+    },
+    {
+      "epoch": 0.7293417752691029,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001722788429631242,
+      "loss": 4.3232,
+      "step": 7034
+    },
+    {
+      "epoch": 0.729445463323591,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017227133593800847,
+      "loss": 4.2879,
+      "step": 7035
+    },
+    {
+      "epoch": 0.7295491513780791,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017226382806016734,
+      "loss": 4.3101,
+      "step": 7036
+    },
+    {
+      "epoch": 0.7296528394325671,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017225631932968945,
+      "loss": 4.2876,
+      "step": 7037
+    },
+    {
+      "epoch": 0.7297565274870552,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017224880974666337,
+      "loss": 4.2636,
+      "step": 7038
+    },
+    {
+      "epoch": 0.7298602155415432,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017224129931117768,
+      "loss": 4.3189,
+      "step": 7039
+    },
+    {
+      "epoch": 0.7299639035960314,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017223378802332098,
+      "loss": 4.3109,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7300675916505194,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.000172226275883182,
+      "loss": 4.2895,
+      "step": 7041
+    },
+    {
+      "epoch": 0.7301712797050075,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001722187628908493,
+      "loss": 4.2554,
+      "step": 7042
+    },
+    {
+      "epoch": 0.7302749677594955,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017221124904641153,
+      "loss": 4.3007,
+      "step": 7043
+    },
+    {
+      "epoch": 0.7303786558139836,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017220373434995737,
+      "loss": 4.2808,
+      "step": 7044
+    },
+    {
+      "epoch": 0.7304823438684717,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017219621880157544,
+      "loss": 4.275,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7305860319229598,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017218870240135446,
+      "loss": 4.3002,
+      "step": 7046
+    },
+    {
+      "epoch": 0.7306897199774478,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017218118514938309,
+      "loss": 4.3309,
+      "step": 7047
+    },
+    {
+      "epoch": 0.7307934080319359,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017217366704575,
+      "loss": 4.3054,
+      "step": 7048
+    },
+    {
+      "epoch": 0.730897096086424,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017216614809054402,
+      "loss": 4.311,
+      "step": 7049
+    },
+    {
+      "epoch": 0.7310007841409121,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001721586282838537,
+      "loss": 4.2963,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7311044721954001,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017215110762576786,
+      "loss": 4.289,
+      "step": 7051
+    },
+    {
+      "epoch": 0.7312081602498882,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017214358611637522,
+      "loss": 4.3274,
+      "step": 7052
+    },
+    {
+      "epoch": 0.7313118483043762,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001721360637557645,
+      "loss": 4.2937,
+      "step": 7053
+    },
+    {
+      "epoch": 0.7314155363588644,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017212854054402451,
+      "loss": 4.2802,
+      "step": 7054
+    },
+    {
+      "epoch": 0.7315192244133524,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017212101648124398,
+      "loss": 4.301,
+      "step": 7055
+    },
+    {
+      "epoch": 0.7316229124678405,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017211349156751168,
+      "loss": 4.2797,
+      "step": 7056
+    },
+    {
+      "epoch": 0.7317266005223286,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017210596580291644,
+      "loss": 4.3053,
+      "step": 7057
+    },
+    {
+      "epoch": 0.7318302885768166,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017209843918754698,
+      "loss": 4.2942,
+      "step": 7058
+    },
+    {
+      "epoch": 0.7319339766313048,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017209091172149215,
+      "loss": 4.3132,
+      "step": 7059
+    },
+    {
+      "epoch": 0.7320376646857928,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001720833834048408,
+      "loss": 4.2949,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7321413527402809,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001720758542376817,
+      "loss": 4.3509,
+      "step": 7061
+    },
+    {
+      "epoch": 0.7322450407947689,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001720683242201037,
+      "loss": 4.3095,
+      "step": 7062
+    },
+    {
+      "epoch": 0.7323487288492571,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017206079335219568,
+      "loss": 4.2731,
+      "step": 7063
+    },
+    {
+      "epoch": 0.7324524169037451,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017205326163404646,
+      "loss": 4.3128,
+      "step": 7064
+    },
+    {
+      "epoch": 0.7325561049582332,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001720457290657449,
+      "loss": 4.3371,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7326597930127212,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001720381956473799,
+      "loss": 4.3332,
+      "step": 7066
+    },
+    {
+      "epoch": 0.7327634810672093,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017203066137904034,
+      "loss": 4.2822,
+      "step": 7067
+    },
+    {
+      "epoch": 0.7328671691216974,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001720231262608151,
+      "loss": 4.305,
+      "step": 7068
+    },
+    {
+      "epoch": 0.7329708571761855,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017201559029279313,
+      "loss": 4.3058,
+      "step": 7069
+    },
+    {
+      "epoch": 0.7330745452306735,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001720080534750633,
+      "loss": 4.3136,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7331782332851616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017200051580771458,
+      "loss": 4.2996,
+      "step": 7071
+    },
+    {
+      "epoch": 0.7332819213396496,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017199297729083584,
+      "loss": 4.2357,
+      "step": 7072
+    },
+    {
+      "epoch": 0.7333856093941378,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017198543792451608,
+      "loss": 4.2651,
+      "step": 7073
+    },
+    {
+      "epoch": 0.7334892974486258,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017197789770884424,
+      "loss": 4.2955,
+      "step": 7074
+    },
+    {
+      "epoch": 0.7335929855031139,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017197035664390936,
+      "loss": 4.308,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7336966735576019,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017196281472980026,
+      "loss": 4.2533,
+      "step": 7076
+    },
+    {
+      "epoch": 0.7338003616120901,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017195527196660603,
+      "loss": 4.3313,
+      "step": 7077
+    },
+    {
+      "epoch": 0.7339040496665781,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001719477283544157,
+      "loss": 4.2987,
+      "step": 7078
+    },
+    {
+      "epoch": 0.7340077377210662,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017194018389331817,
+      "loss": 4.3059,
+      "step": 7079
+    },
+    {
+      "epoch": 0.7341114257755542,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017193263858340254,
+      "loss": 4.2652,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7342151138300423,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017192509242475783,
+      "loss": 4.3037,
+      "step": 7081
+    },
+    {
+      "epoch": 0.7343188018845304,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017191754541747301,
+      "loss": 4.2889,
+      "step": 7082
+    },
+    {
+      "epoch": 0.7344224899390185,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017190999756163723,
+      "loss": 4.3051,
+      "step": 7083
+    },
+    {
+      "epoch": 0.7345261779935065,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017190244885733946,
+      "loss": 4.3293,
+      "step": 7084
+    },
+    {
+      "epoch": 0.7346298660479946,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017189489930466878,
+      "loss": 4.3515,
+      "step": 7085
+    },
+    {
+      "epoch": 0.7347335541024826,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001718873489037143,
+      "loss": 4.309,
+      "step": 7086
+    },
+    {
+      "epoch": 0.7348372421569708,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00017187979765456512,
+      "loss": 4.2999,
+      "step": 7087
+    },
+    {
+      "epoch": 0.7349409302114588,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001718722455573103,
+      "loss": 4.2773,
+      "step": 7088
+    },
+    {
+      "epoch": 0.7350446182659469,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001718646926120389,
+      "loss": 4.3381,
+      "step": 7089
+    },
+    {
+      "epoch": 0.7351483063204349,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017185713881884014,
+      "loss": 4.2982,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7352519943749231,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017184958417780307,
+      "loss": 4.272,
+      "step": 7091
+    },
+    {
+      "epoch": 0.7353556824294111,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017184202868901691,
+      "loss": 4.3178,
+      "step": 7092
+    },
+    {
+      "epoch": 0.7354593704838992,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001718344723525707,
+      "loss": 4.3005,
+      "step": 7093
+    },
+    {
+      "epoch": 0.7355630585383872,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017182691516855368,
+      "loss": 4.3045,
+      "step": 7094
+    },
+    {
+      "epoch": 0.7356667465928753,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017181935713705496,
+      "loss": 4.3003,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7357704346473634,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017181179825816374,
+      "loss": 4.3082,
+      "step": 7096
+    },
+    {
+      "epoch": 0.7358741227018515,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017180423853196923,
+      "loss": 4.3297,
+      "step": 7097
+    },
+    {
+      "epoch": 0.7359778107563395,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001717966779585606,
+      "loss": 4.2416,
+      "step": 7098
+    },
+    {
+      "epoch": 0.7360814988108276,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017178911653802705,
+      "loss": 4.2868,
+      "step": 7099
+    },
+    {
+      "epoch": 0.7361851868653156,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017178155427045782,
+      "loss": 4.3129,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7362888749198038,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017177399115594213,
+      "loss": 4.3162,
+      "step": 7101
+    },
+    {
+      "epoch": 0.7363925629742919,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001717664271945692,
+      "loss": 4.2818,
+      "step": 7102
+    },
+    {
+      "epoch": 0.7364962510287799,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017175886238642832,
+      "loss": 4.3238,
+      "step": 7103
+    },
+    {
+      "epoch": 0.736599939083268,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017175129673160865,
+      "loss": 4.2752,
+      "step": 7104
+    },
+    {
+      "epoch": 0.7367036271377561,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017174373023019958,
+      "loss": 4.3085,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7368073151922442,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001717361628822903,
+      "loss": 4.2717,
+      "step": 7106
+    },
+    {
+      "epoch": 0.7369110032467322,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017172859468797015,
+      "loss": 4.3096,
+      "step": 7107
+    },
+    {
+      "epoch": 0.7370146913012203,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001717210256473284,
+      "loss": 4.3106,
+      "step": 7108
+    },
+    {
+      "epoch": 0.7371183793557083,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017171345576045437,
+      "loss": 4.3068,
+      "step": 7109
+    },
+    {
+      "epoch": 0.7372220674101965,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00017170588502743735,
+      "loss": 4.2665,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7373257554646845,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017169831344836668,
+      "loss": 4.3333,
+      "step": 7111
+    },
+    {
+      "epoch": 0.7374294435191726,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001716907410233317,
+      "loss": 4.2849,
+      "step": 7112
+    },
+    {
+      "epoch": 0.7375331315736606,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017168316775242174,
+      "loss": 4.2733,
+      "step": 7113
+    },
+    {
+      "epoch": 0.7376368196281488,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001716755936357262,
+      "loss": 4.2969,
+      "step": 7114
+    },
+    {
+      "epoch": 0.7377405076826368,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017166801867333443,
+      "loss": 4.3362,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7378441957371249,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017166044286533576,
+      "loss": 4.2884,
+      "step": 7116
+    },
+    {
+      "epoch": 0.7379478837916129,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017165286621181961,
+      "loss": 4.2443,
+      "step": 7117
+    },
+    {
+      "epoch": 0.738051571846101,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001716452887128754,
+      "loss": 4.3151,
+      "step": 7118
+    },
+    {
+      "epoch": 0.7381552599005891,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017163771036859252,
+      "loss": 4.3142,
+      "step": 7119
+    },
+    {
+      "epoch": 0.7382589479550772,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001716301311790604,
+      "loss": 4.2677,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7383626360095652,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017162255114436842,
+      "loss": 4.2915,
+      "step": 7121
+    },
+    {
+      "epoch": 0.7384663240640533,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017161497026460605,
+      "loss": 4.2741,
+      "step": 7122
+    },
+    {
+      "epoch": 0.7385700121185413,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017160738853986272,
+      "loss": 4.2836,
+      "step": 7123
+    },
+    {
+      "epoch": 0.7386737001730295,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001715998059702279,
+      "loss": 4.2986,
+      "step": 7124
+    },
+    {
+      "epoch": 0.7387773882275175,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017159222255579105,
+      "loss": 4.3167,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7388810762820056,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017158463829664169,
+      "loss": 4.3071,
+      "step": 7126
+    },
+    {
+      "epoch": 0.7389847643364936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017157705319286923,
+      "loss": 4.2997,
+      "step": 7127
+    },
+    {
+      "epoch": 0.7390884523909818,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017156946724456321,
+      "loss": 4.3084,
+      "step": 7128
+    },
+    {
+      "epoch": 0.7391921404454698,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017156188045181313,
+      "loss": 4.3035,
+      "step": 7129
+    },
+    {
+      "epoch": 0.7392958284999579,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017155429281470852,
+      "loss": 4.3053,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7393995165544459,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017154670433333887,
+      "loss": 4.2988,
+      "step": 7131
+    },
+    {
+      "epoch": 0.739503204608934,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017153911500779377,
+      "loss": 4.2948,
+      "step": 7132
+    },
+    {
+      "epoch": 0.739606892663422,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017153152483816267,
+      "loss": 4.2857,
+      "step": 7133
+    },
+    {
+      "epoch": 0.7397105807179102,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017152393382453523,
+      "loss": 4.2967,
+      "step": 7134
+    },
+    {
+      "epoch": 0.7398142687723982,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017151634196700097,
+      "loss": 4.3033,
+      "step": 7135
+    },
+    {
+      "epoch": 0.7399179568268863,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017150874926564948,
+      "loss": 4.2786,
+      "step": 7136
+    },
+    {
+      "epoch": 0.7400216448813743,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017150115572057032,
+      "loss": 4.2795,
+      "step": 7137
+    },
+    {
+      "epoch": 0.7401253329358625,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017149356133185312,
+      "loss": 4.2819,
+      "step": 7138
+    },
+    {
+      "epoch": 0.7402290209903505,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017148596609958746,
+      "loss": 4.2671,
+      "step": 7139
+    },
+    {
+      "epoch": 0.7403327090448386,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017147837002386295,
+      "loss": 4.2576,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7404363970993266,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017147077310476923,
+      "loss": 4.2497,
+      "step": 7141
+    },
+    {
+      "epoch": 0.7405400851538148,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017146317534239597,
+      "loss": 4.3075,
+      "step": 7142
+    },
+    {
+      "epoch": 0.7406437732083028,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017145557673683273,
+      "loss": 4.3005,
+      "step": 7143
+    },
+    {
+      "epoch": 0.7407474612627909,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017144797728816928,
+      "loss": 4.3209,
+      "step": 7144
+    },
+    {
+      "epoch": 0.7408511493172789,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001714403769964952,
+      "loss": 4.284,
+      "step": 7145
+    },
+    {
+      "epoch": 0.740954837371767,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017143277586190015,
+      "loss": 4.3253,
+      "step": 7146
+    },
+    {
+      "epoch": 0.7410585254262552,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017142517388447388,
+      "loss": 4.3014,
+      "step": 7147
+    },
+    {
+      "epoch": 0.7411622134807432,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017141757106430605,
+      "loss": 4.2871,
+      "step": 7148
+    },
+    {
+      "epoch": 0.7412659015352313,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017140996740148636,
+      "loss": 4.2668,
+      "step": 7149
+    },
+    {
+      "epoch": 0.7413695895897193,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017140236289610457,
+      "loss": 4.3317,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7414732776442075,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017139475754825037,
+      "loss": 4.279,
+      "step": 7151
+    },
+    {
+      "epoch": 0.7415769656986955,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017138715135801347,
+      "loss": 4.2713,
+      "step": 7152
+    },
+    {
+      "epoch": 0.7416806537531836,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017137954432548365,
+      "loss": 4.271,
+      "step": 7153
+    },
+    {
+      "epoch": 0.7417843418076716,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017137193645075068,
+      "loss": 4.3096,
+      "step": 7154
+    },
+    {
+      "epoch": 0.7418880298621597,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017136432773390427,
+      "loss": 4.3234,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7419917179166478,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00017135671817503426,
+      "loss": 4.3376,
+      "step": 7156
+    },
+    {
+      "epoch": 0.7420954059711359,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001713491077742304,
+      "loss": 4.2886,
+      "step": 7157
+    },
+    {
+      "epoch": 0.7421990940256239,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001713414965315825,
+      "loss": 4.299,
+      "step": 7158
+    },
+    {
+      "epoch": 0.742302782080112,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001713338844471803,
+      "loss": 4.2673,
+      "step": 7159
+    },
+    {
+      "epoch": 0.7424064701346,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017132627152111372,
+      "loss": 4.297,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7425101581890882,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017131865775347249,
+      "loss": 4.2811,
+      "step": 7161
+    },
+    {
+      "epoch": 0.7426138462435762,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017131104314434652,
+      "loss": 4.2929,
+      "step": 7162
+    },
+    {
+      "epoch": 0.7427175342980643,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017130342769382562,
+      "loss": 4.3183,
+      "step": 7163
+    },
+    {
+      "epoch": 0.7428212223525523,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017129581140199962,
+      "loss": 4.2843,
+      "step": 7164
+    },
+    {
+      "epoch": 0.7429249104070405,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017128819426895841,
+      "loss": 4.288,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7430285984615285,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017128057629479186,
+      "loss": 4.2776,
+      "step": 7166
+    },
+    {
+      "epoch": 0.7431322865160166,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001712729574795899,
+      "loss": 4.2689,
+      "step": 7167
+    },
+    {
+      "epoch": 0.7432359745705046,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017126533782344235,
+      "loss": 4.3065,
+      "step": 7168
+    },
+    {
+      "epoch": 0.7433396626249927,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017125771732643915,
+      "loss": 4.3338,
+      "step": 7169
+    },
+    {
+      "epoch": 0.7434433506794808,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017125009598867018,
+      "loss": 4.2874,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7435470387339689,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001712424738102254,
+      "loss": 4.2767,
+      "step": 7171
+    },
+    {
+      "epoch": 0.7436507267884569,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017123485079119477,
+      "loss": 4.3092,
+      "step": 7172
+    },
+    {
+      "epoch": 0.743754414842945,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017122722693166815,
+      "loss": 4.2933,
+      "step": 7173
+    },
+    {
+      "epoch": 0.743858102897433,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017121960223173558,
+      "loss": 4.2672,
+      "step": 7174
+    },
+    {
+      "epoch": 0.7439617909519212,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.000171211976691487,
+      "loss": 4.2863,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7440654790064092,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017120435031101232,
+      "loss": 4.2946,
+      "step": 7176
+    },
+    {
+      "epoch": 0.7441691670608973,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001711967230904016,
+      "loss": 4.2853,
+      "step": 7177
+    },
+    {
+      "epoch": 0.7442728551153853,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017118909502974482,
+      "loss": 4.3053,
+      "step": 7178
+    },
+    {
+      "epoch": 0.7443765431698735,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00017118146612913192,
+      "loss": 4.251,
+      "step": 7179
+    },
+    {
+      "epoch": 0.7444802312243615,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017117383638865302,
+      "loss": 4.2856,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7445839192788496,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017116620580839804,
+      "loss": 4.3232,
+      "step": 7181
+    },
+    {
+      "epoch": 0.7446876073333376,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017115857438845708,
+      "loss": 4.2889,
+      "step": 7182
+    },
+    {
+      "epoch": 0.7447912953878257,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017115094212892017,
+      "loss": 4.2871,
+      "step": 7183
+    },
+    {
+      "epoch": 0.7448949834423138,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017114330902987733,
+      "loss": 4.2325,
+      "step": 7184
+    },
+    {
+      "epoch": 0.7449986714968019,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017113567509141863,
+      "loss": 4.2844,
+      "step": 7185
+    },
+    {
+      "epoch": 0.7451023595512899,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001711280403136342,
+      "loss": 4.2512,
+      "step": 7186
+    },
+    {
+      "epoch": 0.745206047605778,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00017112040469661409,
+      "loss": 4.3283,
+      "step": 7187
+    },
+    {
+      "epoch": 0.745309735660266,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017111276824044832,
+      "loss": 4.2993,
+      "step": 7188
+    },
+    {
+      "epoch": 0.7454134237147542,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017110513094522708,
+      "loss": 4.2461,
+      "step": 7189
+    },
+    {
+      "epoch": 0.7455171117692422,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017109749281104048,
+      "loss": 4.3055,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7456207998237303,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001710898538379786,
+      "loss": 4.2779,
+      "step": 7191
+    },
+    {
+      "epoch": 0.7457244878782184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001710822140261316,
+      "loss": 4.2817,
+      "step": 7192
+    },
+    {
+      "epoch": 0.7458281759327064,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017107457337558958,
+      "loss": 4.3003,
+      "step": 7193
+    },
+    {
+      "epoch": 0.7459318639871946,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017106693188644276,
+      "loss": 4.2541,
+      "step": 7194
+    },
+    {
+      "epoch": 0.7460355520416826,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017105928955878127,
+      "loss": 4.2842,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7461392400961707,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00017105164639269526,
+      "loss": 4.2931,
+      "step": 7196
+    },
+    {
+      "epoch": 0.7462429281506587,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001710440023882749,
+      "loss": 4.3405,
+      "step": 7197
+    },
+    {
+      "epoch": 0.7463466162051469,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00017103635754561045,
+      "loss": 4.3018,
+      "step": 7198
+    },
+    {
+      "epoch": 0.7464503042596349,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017102871186479206,
+      "loss": 4.3131,
+      "step": 7199
+    },
+    {
+      "epoch": 0.746553992314123,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017102106534590993,
+      "loss": 4.3459,
+      "step": 7200
+    },
+    {
+      "epoch": 0.746657680368611,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017101341798905434,
+      "loss": 4.3024,
+      "step": 7201
+    },
+    {
+      "epoch": 0.7467613684230991,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017100576979431543,
+      "loss": 4.2866,
+      "step": 7202
+    },
+    {
+      "epoch": 0.7468650564775872,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017099812076178355,
+      "loss": 4.295,
+      "step": 7203
+    },
+    {
+      "epoch": 0.7469687445320753,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017099047089154886,
+      "loss": 4.2849,
+      "step": 7204
+    },
+    {
+      "epoch": 0.7470724325865633,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017098282018370163,
+      "loss": 4.3026,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7471761206410514,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017097516863833222,
+      "loss": 4.2857,
+      "step": 7206
+    },
+    {
+      "epoch": 0.7472798086955394,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001709675162555308,
+      "loss": 4.2794,
+      "step": 7207
+    },
+    {
+      "epoch": 0.7473834967500276,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001709598630353877,
+      "loss": 4.2753,
+      "step": 7208
+    },
+    {
+      "epoch": 0.7474871848045156,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017095220897799323,
+      "loss": 4.2956,
+      "step": 7209
+    },
+    {
+      "epoch": 0.7475908728590037,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00017094455408343768,
+      "loss": 4.2803,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7476945609134917,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001709368983518114,
+      "loss": 4.2632,
+      "step": 7211
+    },
+    {
+      "epoch": 0.7477982489679799,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001709292417832047,
+      "loss": 4.3143,
+      "step": 7212
+    },
+    {
+      "epoch": 0.7479019370224679,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017092158437770794,
+      "loss": 4.2838,
+      "step": 7213
+    },
+    {
+      "epoch": 0.748005625076956,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017091392613541144,
+      "loss": 4.284,
+      "step": 7214
+    },
+    {
+      "epoch": 0.748109313131444,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017090626705640557,
+      "loss": 4.2753,
+      "step": 7215
+    },
+    {
+      "epoch": 0.7482130011859321,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001708986071407807,
+      "loss": 4.3097,
+      "step": 7216
+    },
+    {
+      "epoch": 0.7483166892404202,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001708909463886272,
+      "loss": 4.2965,
+      "step": 7217
+    },
+    {
+      "epoch": 0.7484203772949083,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017088328480003545,
+      "loss": 4.2942,
+      "step": 7218
+    },
+    {
+      "epoch": 0.7485240653493963,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017087562237509592,
+      "loss": 4.3179,
+      "step": 7219
+    },
+    {
+      "epoch": 0.7486277534038844,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017086795911389895,
+      "loss": 4.289,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7487314414583724,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017086029501653496,
+      "loss": 4.2506,
+      "step": 7221
+    },
+    {
+      "epoch": 0.7488351295128606,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017085263008309438,
+      "loss": 4.3148,
+      "step": 7222
+    },
+    {
+      "epoch": 0.7489388175673486,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017084496431366767,
+      "loss": 4.2945,
+      "step": 7223
+    },
+    {
+      "epoch": 0.7490425056218367,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00017083729770834527,
+      "loss": 4.3027,
+      "step": 7224
+    },
+    {
+      "epoch": 0.7491461936763247,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00017082963026721762,
+      "loss": 4.2845,
+      "step": 7225
+    },
+    {
+      "epoch": 0.7492498817308129,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001708219619903752,
+      "loss": 4.3067,
+      "step": 7226
+    },
+    {
+      "epoch": 0.7493535697853009,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00017081429287790854,
+      "loss": 4.3316,
+      "step": 7227
+    },
+    {
+      "epoch": 0.749457257839789,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00017080662292990803,
+      "loss": 4.309,
+      "step": 7228
+    },
+    {
+      "epoch": 0.749560945894277,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001707989521464642,
+      "loss": 4.2783,
+      "step": 7229
+    },
+    {
+      "epoch": 0.7496646339487651,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017079128052766764,
+      "loss": 4.2817,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7497683220032532,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00017078360807360875,
+      "loss": 4.2768,
+      "step": 7231
+    },
+    {
+      "epoch": 0.7498720100577413,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017077593478437812,
+      "loss": 4.2631,
+      "step": 7232
+    },
+    {
+      "epoch": 0.7499756981122293,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017076826066006627,
+      "loss": 4.2689,
+      "step": 7233
+    },
+    {
+      "epoch": 0.7499756981122293,
+      "eval_loss": 4.309548377990723,
+      "eval_runtime": 0.441,
+      "eval_samples_per_second": 337.873,
+      "eval_steps_per_second": 15.873,
+      "step": 7233
+    },
+    {
+      "epoch": 0.7500793861667174,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00017076058570076374,
+      "loss": 4.2999,
+      "step": 7234
+    },
+    {
+      "epoch": 0.7501830742212054,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001707529099065611,
+      "loss": 4.2887,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7502867622756936,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00017074523327754895,
+      "loss": 4.2617,
+      "step": 7236
+    },
+    {
+      "epoch": 0.7503904503301817,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00017073755581381778,
+      "loss": 4.3132,
+      "step": 7237
+    },
+    {
+      "epoch": 0.7504941383846697,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017072987751545827,
+      "loss": 4.2769,
+      "step": 7238
+    },
+    {
+      "epoch": 0.7505978264391578,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017072219838256092,
+      "loss": 4.3216,
+      "step": 7239
+    },
+    {
+      "epoch": 0.7507015144936459,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001707145184152164,
+      "loss": 4.3209,
+      "step": 7240
+    },
+    {
+      "epoch": 0.750805202548134,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017070683761351532,
+      "loss": 4.3313,
+      "step": 7241
+    },
+    {
+      "epoch": 0.750908890602622,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017069915597754827,
+      "loss": 4.2942,
+      "step": 7242
+    },
+    {
+      "epoch": 0.7510125786571101,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00017069147350740594,
+      "loss": 4.3034,
+      "step": 7243
+    },
+    {
+      "epoch": 0.7511162667115981,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00017068379020317894,
+      "loss": 4.3169,
+      "step": 7244
+    },
+    {
+      "epoch": 0.7512199547660863,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017067610606495795,
+      "loss": 4.3012,
+      "step": 7245
+    },
+    {
+      "epoch": 0.7513236428205743,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00017066842109283358,
+      "loss": 4.3494,
+      "step": 7246
+    },
+    {
+      "epoch": 0.7514273308750624,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017066073528689658,
+      "loss": 4.3039,
+      "step": 7247
+    },
+    {
+      "epoch": 0.7515310189295504,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00017065304864723757,
+      "loss": 4.3302,
+      "step": 7248
+    },
+    {
+      "epoch": 0.7516347069840386,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017064536117394728,
+      "loss": 4.2775,
+      "step": 7249
+    },
+    {
+      "epoch": 0.7517383950385266,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001706376728671164,
+      "loss": 4.2653,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7518420830930147,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00017062998372683566,
+      "loss": 4.3025,
+      "step": 7251
+    },
+    {
+      "epoch": 0.7519457711475027,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017062229375319573,
+      "loss": 4.2539,
+      "step": 7252
+    },
+    {
+      "epoch": 0.7520494592019908,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017061460294628744,
+      "loss": 4.2528,
+      "step": 7253
+    },
+    {
+      "epoch": 0.7521531472564789,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017060691130620144,
+      "loss": 4.318,
+      "step": 7254
+    },
+    {
+      "epoch": 0.752256835310967,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017059921883302853,
+      "loss": 4.3094,
+      "step": 7255
+    },
+    {
+      "epoch": 0.752360523365455,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001705915255268595,
+      "loss": 4.3801,
+      "step": 7256
+    },
+    {
+      "epoch": 0.7524642114199431,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00017058383138778504,
+      "loss": 4.319,
+      "step": 7257
+    },
+    {
+      "epoch": 0.7525678994744311,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00017057613641589597,
+      "loss": 4.2533,
+      "step": 7258
+    },
+    {
+      "epoch": 0.7526715875289193,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017056844061128312,
+      "loss": 4.2892,
+      "step": 7259
+    },
+    {
+      "epoch": 0.7527752755834073,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017056074397403726,
+      "loss": 4.2962,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7528789636378954,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017055304650424923,
+      "loss": 4.3135,
+      "step": 7261
+    },
+    {
+      "epoch": 0.7529826516923834,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017054534820200982,
+      "loss": 4.2944,
+      "step": 7262
+    },
+    {
+      "epoch": 0.7530863397468716,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001705376490674098,
+      "loss": 4.2739,
+      "step": 7263
+    },
+    {
+      "epoch": 0.7531900278013596,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00017052994910054018,
+      "loss": 4.2843,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7532937158558477,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017052224830149166,
+      "loss": 4.3169,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7533974039103357,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017051454667035517,
+      "loss": 4.2898,
+      "step": 7266
+    },
+    {
+      "epoch": 0.7535010919648238,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017050684420722155,
+      "loss": 4.244,
+      "step": 7267
+    },
+    {
+      "epoch": 0.7536047800193119,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017049914091218175,
+      "loss": 4.3089,
+      "step": 7268
+    },
+    {
+      "epoch": 0.7537084680738,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00017049143678532654,
+      "loss": 4.284,
+      "step": 7269
+    },
+    {
+      "epoch": 0.753812156128288,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00017048373182674694,
+      "loss": 4.3093,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7539158441827761,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001704760260365338,
+      "loss": 4.2962,
+      "step": 7271
+    },
+    {
+      "epoch": 0.7540195322372641,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017046831941477803,
+      "loss": 4.3126,
+      "step": 7272
+    },
+    {
+      "epoch": 0.7541232202917523,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001704606119615706,
+      "loss": 4.2415,
+      "step": 7273
+    },
+    {
+      "epoch": 0.7542269083462403,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001704529036770024,
+      "loss": 4.3483,
+      "step": 7274
+    },
+    {
+      "epoch": 0.7543305964007284,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017044519456116443,
+      "loss": 4.3222,
+      "step": 7275
+    },
+    {
+      "epoch": 0.7544342844552164,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017043748461414762,
+      "loss": 4.2517,
+      "step": 7276
+    },
+    {
+      "epoch": 0.7545379725097046,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017042977383604295,
+      "loss": 4.3102,
+      "step": 7277
+    },
+    {
+      "epoch": 0.7546416605641926,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001704220622269414,
+      "loss": 4.3405,
+      "step": 7278
+    },
+    {
+      "epoch": 0.7547453486186807,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017041434978693393,
+      "loss": 4.3233,
+      "step": 7279
+    },
+    {
+      "epoch": 0.7548490366731688,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017040663651611158,
+      "loss": 4.2985,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7549527247276568,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017039892241456537,
+      "loss": 4.2694,
+      "step": 7281
+    },
+    {
+      "epoch": 0.755056412782145,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017039120748238627,
+      "loss": 4.3193,
+      "step": 7282
+    },
+    {
+      "epoch": 0.755160100836633,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001703834917196653,
+      "loss": 4.2847,
+      "step": 7283
+    },
+    {
+      "epoch": 0.7552637888911211,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017037577512649357,
+      "loss": 4.2836,
+      "step": 7284
+    },
+    {
+      "epoch": 0.7553674769456091,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017036805770296207,
+      "loss": 4.2952,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7554711650000973,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017036033944916188,
+      "loss": 4.3062,
+      "step": 7286
+    },
+    {
+      "epoch": 0.7555748530545853,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017035262036518403,
+      "loss": 4.2864,
+      "step": 7287
+    },
+    {
+      "epoch": 0.7556785411090734,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017034490045111963,
+      "loss": 4.2936,
+      "step": 7288
+    },
+    {
+      "epoch": 0.7557822291635614,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001703371797070598,
+      "loss": 4.265,
+      "step": 7289
+    },
+    {
+      "epoch": 0.7558859172180495,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017032945813309555,
+      "loss": 4.3142,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7559896052725376,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017032173572931807,
+      "loss": 4.2957,
+      "step": 7291
+    },
+    {
+      "epoch": 0.7560932933270257,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017031401249581842,
+      "loss": 4.3144,
+      "step": 7292
+    },
+    {
+      "epoch": 0.7561969813815137,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017030628843268776,
+      "loss": 4.2702,
+      "step": 7293
+    },
+    {
+      "epoch": 0.7563006694360018,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001702985635400172,
+      "loss": 4.3049,
+      "step": 7294
+    },
+    {
+      "epoch": 0.7564043574904898,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017029083781789793,
+      "loss": 4.278,
+      "step": 7295
+    },
+    {
+      "epoch": 0.756508045544978,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017028311126642107,
+      "loss": 4.2731,
+      "step": 7296
+    },
+    {
+      "epoch": 0.756611733599466,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00017027538388567778,
+      "loss": 4.3164,
+      "step": 7297
+    },
+    {
+      "epoch": 0.7567154216539541,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00017026765567575923,
+      "loss": 4.3086,
+      "step": 7298
+    },
+    {
+      "epoch": 0.7568191097084421,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001702599266367566,
+      "loss": 4.3135,
+      "step": 7299
+    },
+    {
+      "epoch": 0.7569227977629303,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017025219676876114,
+      "loss": 4.2603,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7570264858174183,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017024446607186402,
+      "loss": 4.2973,
+      "step": 7301
+    },
+    {
+      "epoch": 0.7571301738719064,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017023673454615645,
+      "loss": 4.3145,
+      "step": 7302
+    },
+    {
+      "epoch": 0.7572338619263944,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017022900219172964,
+      "loss": 4.2712,
+      "step": 7303
+    },
+    {
+      "epoch": 0.7573375499808825,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017022126900867484,
+      "loss": 4.313,
+      "step": 7304
+    },
+    {
+      "epoch": 0.7574412380353706,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017021353499708331,
+      "loss": 4.3111,
+      "step": 7305
+    },
+    {
+      "epoch": 0.7575449260898587,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017020580015704627,
+      "loss": 4.3204,
+      "step": 7306
+    },
+    {
+      "epoch": 0.7576486141443467,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017019806448865502,
+      "loss": 4.3063,
+      "step": 7307
+    },
+    {
+      "epoch": 0.7577523021988348,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001701903279920008,
+      "loss": 4.3014,
+      "step": 7308
+    },
+    {
+      "epoch": 0.7578559902533228,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00017018259066717487,
+      "loss": 4.2751,
+      "step": 7309
+    },
+    {
+      "epoch": 0.757959678307811,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001701748525142686,
+      "loss": 4.2672,
+      "step": 7310
+    },
+    {
+      "epoch": 0.758063366362299,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017016711353337325,
+      "loss": 4.2977,
+      "step": 7311
+    },
+    {
+      "epoch": 0.7581670544167871,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00017015937372458009,
+      "loss": 4.2664,
+      "step": 7312
+    },
+    {
+      "epoch": 0.7582707424712751,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001701516330879805,
+      "loss": 4.3269,
+      "step": 7313
+    },
+    {
+      "epoch": 0.7583744305257633,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001701438916236658,
+      "loss": 4.2875,
+      "step": 7314
+    },
+    {
+      "epoch": 0.7584781185802513,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017013614933172733,
+      "loss": 4.3359,
+      "step": 7315
+    },
+    {
+      "epoch": 0.7585818066347394,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017012840621225643,
+      "loss": 4.3346,
+      "step": 7316
+    },
+    {
+      "epoch": 0.7586854946892274,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017012066226534446,
+      "loss": 4.321,
+      "step": 7317
+    },
+    {
+      "epoch": 0.7587891827437155,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017011291749108282,
+      "loss": 4.2992,
+      "step": 7318
+    },
+    {
+      "epoch": 0.7588928707982036,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017010517188956284,
+      "loss": 4.2377,
+      "step": 7319
+    },
+    {
+      "epoch": 0.7589965588526917,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017009742546087594,
+      "loss": 4.2905,
+      "step": 7320
+    },
+    {
+      "epoch": 0.7591002469071797,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017008967820511352,
+      "loss": 4.273,
+      "step": 7321
+    },
+    {
+      "epoch": 0.7592039349616678,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.000170081930122367,
+      "loss": 4.2929,
+      "step": 7322
+    },
+    {
+      "epoch": 0.7593076230161558,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017007418121272775,
+      "loss": 4.3323,
+      "step": 7323
+    },
+    {
+      "epoch": 0.759411311070644,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017006643147628726,
+      "loss": 4.3453,
+      "step": 7324
+    },
+    {
+      "epoch": 0.7595149991251321,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00017005868091313694,
+      "loss": 4.3426,
+      "step": 7325
+    },
+    {
+      "epoch": 0.7596186871796201,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00017005092952336823,
+      "loss": 4.2514,
+      "step": 7326
+    },
+    {
+      "epoch": 0.7597223752341082,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017004317730707263,
+      "loss": 4.2899,
+      "step": 7327
+    },
+    {
+      "epoch": 0.7598260632885963,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001700354242643415,
+      "loss": 4.2561,
+      "step": 7328
+    },
+    {
+      "epoch": 0.7599297513430844,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017002767039526646,
+      "loss": 4.3167,
+      "step": 7329
+    },
+    {
+      "epoch": 0.7600334393975724,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001700199156999389,
+      "loss": 4.3333,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7601371274520605,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017001216017845038,
+      "loss": 4.2794,
+      "step": 7331
+    },
+    {
+      "epoch": 0.7602408155065485,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00017000440383089239,
+      "loss": 4.2827,
+      "step": 7332
+    },
+    {
+      "epoch": 0.7603445035610367,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016999664665735638,
+      "loss": 4.2929,
+      "step": 7333
+    },
+    {
+      "epoch": 0.7604481916155247,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016998888865793396,
+      "loss": 4.3017,
+      "step": 7334
+    },
+    {
+      "epoch": 0.7605518796700128,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001699811298327166,
+      "loss": 4.2976,
+      "step": 7335
+    },
+    {
+      "epoch": 0.7606555677245008,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001699733701817959,
+      "loss": 4.2537,
+      "step": 7336
+    },
+    {
+      "epoch": 0.760759255778989,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001699656097052634,
+      "loss": 4.3015,
+      "step": 7337
+    },
+    {
+      "epoch": 0.760862943833477,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001699578484032107,
+      "loss": 4.2592,
+      "step": 7338
+    },
+    {
+      "epoch": 0.7609666318879651,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016995008627572933,
+      "loss": 4.3026,
+      "step": 7339
+    },
+    {
+      "epoch": 0.7610703199424531,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016994232332291084,
+      "loss": 4.2917,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7611740079969412,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016993455954484687,
+      "loss": 4.2922,
+      "step": 7341
+    },
+    {
+      "epoch": 0.7612776960514293,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016992679494162903,
+      "loss": 4.2461,
+      "step": 7342
+    },
+    {
+      "epoch": 0.7613813841059174,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016991902951334894,
+      "loss": 4.3031,
+      "step": 7343
+    },
+    {
+      "epoch": 0.7614850721604054,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016991126326009818,
+      "loss": 4.2936,
+      "step": 7344
+    },
+    {
+      "epoch": 0.7615887602148935,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016990349618196845,
+      "loss": 4.2804,
+      "step": 7345
+    },
+    {
+      "epoch": 0.7616924482693815,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016989572827905134,
+      "loss": 4.2966,
+      "step": 7346
+    },
+    {
+      "epoch": 0.7617961363238697,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016988795955143852,
+      "loss": 4.275,
+      "step": 7347
+    },
+    {
+      "epoch": 0.7618998243783577,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016988018999922167,
+      "loss": 4.256,
+      "step": 7348
+    },
+    {
+      "epoch": 0.7620035124328458,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001698724196224924,
+      "loss": 4.3285,
+      "step": 7349
+    },
+    {
+      "epoch": 0.7621072004873338,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016986464842134246,
+      "loss": 4.3024,
+      "step": 7350
+    },
+    {
+      "epoch": 0.762210888541822,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016985687639586354,
+      "loss": 4.3106,
+      "step": 7351
+    },
+    {
+      "epoch": 0.76231457659631,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016984910354614732,
+      "loss": 4.3155,
+      "step": 7352
+    },
+    {
+      "epoch": 0.7624182646507981,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016984132987228547,
+      "loss": 4.3108,
+      "step": 7353
+    },
+    {
+      "epoch": 0.7625219527052861,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016983355537436977,
+      "loss": 4.3093,
+      "step": 7354
+    },
+    {
+      "epoch": 0.7626256407597742,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016982578005249197,
+      "loss": 4.3132,
+      "step": 7355
+    },
+    {
+      "epoch": 0.7627293288142623,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001698180039067437,
+      "loss": 4.2788,
+      "step": 7356
+    },
+    {
+      "epoch": 0.7628330168687504,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016981022693721688,
+      "loss": 4.299,
+      "step": 7357
+    },
+    {
+      "epoch": 0.7629367049232384,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001698024491440031,
+      "loss": 4.2963,
+      "step": 7358
+    },
+    {
+      "epoch": 0.7630403929777265,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016979467052719423,
+      "loss": 4.2889,
+      "step": 7359
+    },
+    {
+      "epoch": 0.7631440810322145,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016978689108688202,
+      "loss": 4.2818,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7632477690867027,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016977911082315827,
+      "loss": 4.2491,
+      "step": 7361
+    },
+    {
+      "epoch": 0.7633514571411907,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016977132973611475,
+      "loss": 4.2968,
+      "step": 7362
+    },
+    {
+      "epoch": 0.7634551451956788,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016976354782584333,
+      "loss": 4.3111,
+      "step": 7363
+    },
+    {
+      "epoch": 0.7635588332501668,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016975576509243578,
+      "loss": 4.296,
+      "step": 7364
+    },
+    {
+      "epoch": 0.763662521304655,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016974798153598393,
+      "loss": 4.289,
+      "step": 7365
+    },
+    {
+      "epoch": 0.763766209359143,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001697401971565796,
+      "loss": 4.3073,
+      "step": 7366
+    },
+    {
+      "epoch": 0.7638698974136311,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016973241195431468,
+      "loss": 4.3217,
+      "step": 7367
+    },
+    {
+      "epoch": 0.7639735854681191,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000169724625929281,
+      "loss": 4.3015,
+      "step": 7368
+    },
+    {
+      "epoch": 0.7640772735226072,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016971683908157046,
+      "loss": 4.2788,
+      "step": 7369
+    },
+    {
+      "epoch": 0.7641809615770954,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001697090514112749,
+      "loss": 4.2878,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7642846496315834,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001697012629184862,
+      "loss": 4.2664,
+      "step": 7371
+    },
+    {
+      "epoch": 0.7643883376860715,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001696934736032963,
+      "loss": 4.3435,
+      "step": 7372
+    },
+    {
+      "epoch": 0.7644920257405595,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016968568346579707,
+      "loss": 4.2861,
+      "step": 7373
+    },
+    {
+      "epoch": 0.7645957137950476,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016967789250608046,
+      "loss": 4.2991,
+      "step": 7374
+    },
+    {
+      "epoch": 0.7646994018495357,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016967010072423835,
+      "loss": 4.2715,
+      "step": 7375
+    },
+    {
+      "epoch": 0.7648030899040238,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016966230812036267,
+      "loss": 4.3264,
+      "step": 7376
+    },
+    {
+      "epoch": 0.7649067779585118,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016965451469454546,
+      "loss": 4.2816,
+      "step": 7377
+    },
+    {
+      "epoch": 0.7650104660129999,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016964672044687853,
+      "loss": 4.2958,
+      "step": 7378
+    },
+    {
+      "epoch": 0.765114154067488,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016963892537745395,
+      "loss": 4.2681,
+      "step": 7379
+    },
+    {
+      "epoch": 0.7652178421219761,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001696311294863637,
+      "loss": 4.2735,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7653215301764641,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016962333277369967,
+      "loss": 4.2949,
+      "step": 7381
+    },
+    {
+      "epoch": 0.7654252182309522,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016961553523955393,
+      "loss": 4.2902,
+      "step": 7382
+    },
+    {
+      "epoch": 0.7655289062854402,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016960773688401846,
+      "loss": 4.2752,
+      "step": 7383
+    },
+    {
+      "epoch": 0.7656325943399284,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016959993770718524,
+      "loss": 4.3044,
+      "step": 7384
+    },
+    {
+      "epoch": 0.7657362823944164,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016959213770914637,
+      "loss": 4.3009,
+      "step": 7385
+    },
+    {
+      "epoch": 0.7658399704489045,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016958433688999381,
+      "loss": 4.2998,
+      "step": 7386
+    },
+    {
+      "epoch": 0.7659436585033925,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016957653524981968,
+      "loss": 4.2673,
+      "step": 7387
+    },
+    {
+      "epoch": 0.7660473465578806,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016956873278871592,
+      "loss": 4.3262,
+      "step": 7388
+    },
+    {
+      "epoch": 0.7661510346123687,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001695609295067747,
+      "loss": 4.3042,
+      "step": 7389
+    },
+    {
+      "epoch": 0.7662547226668568,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.000169553125404088,
+      "loss": 4.2929,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7663584107213448,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016954532048074793,
+      "loss": 4.287,
+      "step": 7391
+    },
+    {
+      "epoch": 0.7664620987758329,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016953751473684664,
+      "loss": 4.276,
+      "step": 7392
+    },
+    {
+      "epoch": 0.766565786830321,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016952970817247614,
+      "loss": 4.2847,
+      "step": 7393
+    },
+    {
+      "epoch": 0.7666694748848091,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001695219007877286,
+      "loss": 4.2912,
+      "step": 7394
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016951409258269612,
+      "loss": 4.2992,
+      "step": 7395
+    },
+    {
+      "epoch": 0.7668768509937852,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001695062835574708,
+      "loss": 4.3023,
+      "step": 7396
+    },
+    {
+      "epoch": 0.7669805390482732,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016949847371214484,
+      "loss": 4.2982,
+      "step": 7397
+    },
+    {
+      "epoch": 0.7670842271027614,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016949066304681033,
+      "loss": 4.2983,
+      "step": 7398
+    },
+    {
+      "epoch": 0.7671879151572494,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016948285156155945,
+      "loss": 4.2468,
+      "step": 7399
+    },
+    {
+      "epoch": 0.7672916032117375,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001694750392564844,
+      "loss": 4.3253,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7673952912662255,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001694672261316773,
+      "loss": 4.3069,
+      "step": 7401
+    },
+    {
+      "epoch": 0.7674989793207136,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016945941218723037,
+      "loss": 4.3027,
+      "step": 7402
+    },
+    {
+      "epoch": 0.7676026673752017,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001694515974232358,
+      "loss": 4.3055,
+      "step": 7403
+    },
+    {
+      "epoch": 0.7677063554296898,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016944378183978578,
+      "loss": 4.3045,
+      "step": 7404
+    },
+    {
+      "epoch": 0.7678100434841778,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016943596543697256,
+      "loss": 4.3241,
+      "step": 7405
+    },
+    {
+      "epoch": 0.7679137315386659,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016942814821488834,
+      "loss": 4.2945,
+      "step": 7406
+    },
+    {
+      "epoch": 0.768017419593154,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016942033017362533,
+      "loss": 4.2977,
+      "step": 7407
+    },
+    {
+      "epoch": 0.7681211076476421,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016941251131327581,
+      "loss": 4.2918,
+      "step": 7408
+    },
+    {
+      "epoch": 0.7682247957021301,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016940469163393207,
+      "loss": 4.3023,
+      "step": 7409
+    },
+    {
+      "epoch": 0.7683284837566182,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001693968711356863,
+      "loss": 4.2654,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7684321718111062,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001693890498186308,
+      "loss": 4.3465,
+      "step": 7411
+    },
+    {
+      "epoch": 0.7685358598655944,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016938122768285786,
+      "loss": 4.2624,
+      "step": 7412
+    },
+    {
+      "epoch": 0.7686395479200824,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001693734047284598,
+      "loss": 4.3084,
+      "step": 7413
+    },
+    {
+      "epoch": 0.7687432359745705,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016936558095552887,
+      "loss": 4.3405,
+      "step": 7414
+    },
+    {
+      "epoch": 0.7688469240290586,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016935775636415742,
+      "loss": 4.2755,
+      "step": 7415
+    },
+    {
+      "epoch": 0.7689506120835466,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016934993095443776,
+      "loss": 4.3349,
+      "step": 7416
+    },
+    {
+      "epoch": 0.7690543001380348,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001693421047264622,
+      "loss": 4.2602,
+      "step": 7417
+    },
+    {
+      "epoch": 0.7691579881925228,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016933427768032314,
+      "loss": 4.2741,
+      "step": 7418
+    },
+    {
+      "epoch": 0.7692616762470109,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016932644981611288,
+      "loss": 4.312,
+      "step": 7419
+    },
+    {
+      "epoch": 0.7693653643014989,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001693186211339238,
+      "loss": 4.2978,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7694690523559871,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016931079163384823,
+      "loss": 4.2716,
+      "step": 7421
+    },
+    {
+      "epoch": 0.7695727404104751,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016930296131597862,
+      "loss": 4.3112,
+      "step": 7422
+    },
+    {
+      "epoch": 0.7696764284649632,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016929513018040728,
+      "loss": 4.2339,
+      "step": 7423
+    },
+    {
+      "epoch": 0.7697801165194512,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016928729822722667,
+      "loss": 4.2802,
+      "step": 7424
+    },
+    {
+      "epoch": 0.7698838045739393,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016927946545652918,
+      "loss": 4.2905,
+      "step": 7425
+    },
+    {
+      "epoch": 0.7699874926284274,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016927163186840726,
+      "loss": 4.2913,
+      "step": 7426
+    },
+    {
+      "epoch": 0.7700911806829155,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001692637974629533,
+      "loss": 4.2782,
+      "step": 7427
+    },
+    {
+      "epoch": 0.7701948687374035,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001692559622402597,
+      "loss": 4.2655,
+      "step": 7428
+    },
+    {
+      "epoch": 0.7702985567918916,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000169248126200419,
+      "loss": 4.2924,
+      "step": 7429
+    },
+    {
+      "epoch": 0.7704022448463796,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001692402893435236,
+      "loss": 4.307,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7705059329008678,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016923245166966595,
+      "loss": 4.2712,
+      "step": 7431
+    },
+    {
+      "epoch": 0.7706096209553558,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016922461317893855,
+      "loss": 4.3062,
+      "step": 7432
+    },
+    {
+      "epoch": 0.7707133090098439,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016921677387143392,
+      "loss": 4.2883,
+      "step": 7433
+    },
+    {
+      "epoch": 0.7708169970643319,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016920893374724455,
+      "loss": 4.2437,
+      "step": 7434
+    },
+    {
+      "epoch": 0.7709206851188201,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016920109280646285,
+      "loss": 4.2196,
+      "step": 7435
+    },
+    {
+      "epoch": 0.7710243731733081,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016919325104918143,
+      "loss": 4.3101,
+      "step": 7436
+    },
+    {
+      "epoch": 0.7711280612277962,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016918540847549277,
+      "loss": 4.3135,
+      "step": 7437
+    },
+    {
+      "epoch": 0.7712317492822842,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016917756508548948,
+      "loss": 4.3182,
+      "step": 7438
+    },
+    {
+      "epoch": 0.7713354373367723,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.000169169720879264,
+      "loss": 4.2481,
+      "step": 7439
+    },
+    {
+      "epoch": 0.7714391253912604,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016916187585690895,
+      "loss": 4.3098,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7715428134457485,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016915403001851683,
+      "loss": 4.2693,
+      "step": 7441
+    },
+    {
+      "epoch": 0.7716465015002365,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016914618336418032,
+      "loss": 4.2724,
+      "step": 7442
+    },
+    {
+      "epoch": 0.7717501895547246,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001691383358939919,
+      "loss": 4.3085,
+      "step": 7443
+    },
+    {
+      "epoch": 0.7718538776092126,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016913048760804417,
+      "loss": 4.2871,
+      "step": 7444
+    },
+    {
+      "epoch": 0.7719575656637008,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016912263850642981,
+      "loss": 4.3041,
+      "step": 7445
+    },
+    {
+      "epoch": 0.7720612537181888,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016911478858924137,
+      "loss": 4.2777,
+      "step": 7446
+    },
+    {
+      "epoch": 0.7721649417726769,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016910693785657146,
+      "loss": 4.2965,
+      "step": 7447
+    },
+    {
+      "epoch": 0.7722686298271649,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016909908630851277,
+      "loss": 4.3142,
+      "step": 7448
+    },
+    {
+      "epoch": 0.7723723178816531,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016909123394515785,
+      "loss": 4.3168,
+      "step": 7449
+    },
+    {
+      "epoch": 0.7724760059361411,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016908338076659945,
+      "loss": 4.3266,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7725796939906292,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016907552677293018,
+      "loss": 4.3167,
+      "step": 7451
+    },
+    {
+      "epoch": 0.7726833820451172,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001690676719642427,
+      "loss": 4.2969,
+      "step": 7452
+    },
+    {
+      "epoch": 0.7727870700996053,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016905981634062967,
+      "loss": 4.2735,
+      "step": 7453
+    },
+    {
+      "epoch": 0.7728907581540934,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001690519599021838,
+      "loss": 4.3345,
+      "step": 7454
+    },
+    {
+      "epoch": 0.7729944462085815,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016904410264899787,
+      "loss": 4.3112,
+      "step": 7455
+    },
+    {
+      "epoch": 0.7730981342630695,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016903624458116444,
+      "loss": 4.2882,
+      "step": 7456
+    },
+    {
+      "epoch": 0.7732018223175576,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001690283856987763,
+      "loss": 4.271,
+      "step": 7457
+    },
+    {
+      "epoch": 0.7733055103720456,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001690205260019262,
+      "loss": 4.3036,
+      "step": 7458
+    },
+    {
+      "epoch": 0.7734091984265338,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016901266549070688,
+      "loss": 4.2912,
+      "step": 7459
+    },
+    {
+      "epoch": 0.7735128864810219,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016900480416521103,
+      "loss": 4.2563,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7736165745355099,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016899694202553143,
+      "loss": 4.2761,
+      "step": 7461
+    },
+    {
+      "epoch": 0.773720262589998,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016898907907176084,
+      "loss": 4.2598,
+      "step": 7462
+    },
+    {
+      "epoch": 0.7738239506444861,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016898121530399202,
+      "loss": 4.2726,
+      "step": 7463
+    },
+    {
+      "epoch": 0.7739276386989742,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001689733507223178,
+      "loss": 4.2995,
+      "step": 7464
+    },
+    {
+      "epoch": 0.7740313267534622,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016896548532683098,
+      "loss": 4.31,
+      "step": 7465
+    },
+    {
+      "epoch": 0.7741350148079503,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001689576191176243,
+      "loss": 4.2907,
+      "step": 7466
+    },
+    {
+      "epoch": 0.7742387028624383,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001689497520947906,
+      "loss": 4.3051,
+      "step": 7467
+    },
+    {
+      "epoch": 0.7743423909169265,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016894188425842273,
+      "loss": 4.3089,
+      "step": 7468
+    },
+    {
+      "epoch": 0.7744460789714145,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016893401560861347,
+      "loss": 4.2659,
+      "step": 7469
+    },
+    {
+      "epoch": 0.7745497670259026,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016892614614545572,
+      "loss": 4.281,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7746534550803906,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016891827586904226,
+      "loss": 4.2811,
+      "step": 7471
+    },
+    {
+      "epoch": 0.7747571431348788,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016891040477946606,
+      "loss": 4.2619,
+      "step": 7472
+    },
+    {
+      "epoch": 0.7748608311893668,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001689025328768199,
+      "loss": 4.2999,
+      "step": 7473
+    },
+    {
+      "epoch": 0.7749645192438549,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016889466016119665,
+      "loss": 4.3273,
+      "step": 7474
+    },
+    {
+      "epoch": 0.7750682072983429,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016888678663268925,
+      "loss": 4.2815,
+      "step": 7475
+    },
+    {
+      "epoch": 0.775171895352831,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00016887891229139064,
+      "loss": 4.2958,
+      "step": 7476
+    },
+    {
+      "epoch": 0.7752755834073191,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016887103713739363,
+      "loss": 4.2667,
+      "step": 7477
+    },
+    {
+      "epoch": 0.7753792714618072,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001688631611707912,
+      "loss": 4.3166,
+      "step": 7478
+    },
+    {
+      "epoch": 0.7754829595162952,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001688552843916762,
+      "loss": 4.3385,
+      "step": 7479
+    },
+    {
+      "epoch": 0.7755866475707833,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016884740680014169,
+      "loss": 4.2935,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7756903356252713,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016883952839628053,
+      "loss": 4.2766,
+      "step": 7481
+    },
+    {
+      "epoch": 0.7757940236797595,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001688316491801857,
+      "loss": 4.2529,
+      "step": 7482
+    },
+    {
+      "epoch": 0.7758977117342475,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016882376915195019,
+      "loss": 4.3134,
+      "step": 7483
+    },
+    {
+      "epoch": 0.7760013997887356,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016881588831166696,
+      "loss": 4.2798,
+      "step": 7484
+    },
+    {
+      "epoch": 0.7761050878432236,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016880800665942895,
+      "loss": 4.297,
+      "step": 7485
+    },
+    {
+      "epoch": 0.7762087758977118,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001688001241953292,
+      "loss": 4.3174,
+      "step": 7486
+    },
+    {
+      "epoch": 0.7763124639521998,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001687922409194607,
+      "loss": 4.2795,
+      "step": 7487
+    },
+    {
+      "epoch": 0.7764161520066879,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001687843568319165,
+      "loss": 4.3163,
+      "step": 7488
+    },
+    {
+      "epoch": 0.7765198400611759,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001687764719327896,
+      "loss": 4.2761,
+      "step": 7489
+    },
+    {
+      "epoch": 0.776623528115664,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00016876858622217304,
+      "loss": 4.299,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7767272161701521,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016876069970015985,
+      "loss": 4.2744,
+      "step": 7491
+    },
+    {
+      "epoch": 0.7768309042246402,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016875281236684305,
+      "loss": 4.2877,
+      "step": 7492
+    },
+    {
+      "epoch": 0.7769345922791282,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016874492422231578,
+      "loss": 4.313,
+      "step": 7493
+    },
+    {
+      "epoch": 0.7770382803336163,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016873703526667106,
+      "loss": 4.2781,
+      "step": 7494
+    },
+    {
+      "epoch": 0.7771419683881043,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016872914550000197,
+      "loss": 4.2608,
+      "step": 7495
+    },
+    {
+      "epoch": 0.7772456564425925,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016872125492240161,
+      "loss": 4.2584,
+      "step": 7496
+    },
+    {
+      "epoch": 0.7773493444970805,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001687133635339631,
+      "loss": 4.2925,
+      "step": 7497
+    },
+    {
+      "epoch": 0.7774530325515686,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001687054713347795,
+      "loss": 4.3133,
+      "step": 7498
+    },
+    {
+      "epoch": 0.7775567206060566,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.000168697578324944,
+      "loss": 4.315,
+      "step": 7499
+    },
+    {
+      "epoch": 0.7776604086605448,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016868968450454966,
+      "loss": 4.2717,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7777640967150328,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001686817898736897,
+      "loss": 4.2838,
+      "step": 7501
+    },
+    {
+      "epoch": 0.7778677847695209,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001686738944324572,
+      "loss": 4.2746,
+      "step": 7502
+    },
+    {
+      "epoch": 0.7779714728240089,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016866599818094531,
+      "loss": 4.3302,
+      "step": 7503
+    },
+    {
+      "epoch": 0.778075160878497,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016865810111924728,
+      "loss": 4.2894,
+      "step": 7504
+    },
+    {
+      "epoch": 0.7781788489329852,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016865020324745617,
+      "loss": 4.2608,
+      "step": 7505
+    },
+    {
+      "epoch": 0.7782825369874732,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016864230456566527,
+      "loss": 4.2797,
+      "step": 7506
+    },
+    {
+      "epoch": 0.7783862250419613,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001686344050739677,
+      "loss": 4.2473,
+      "step": 7507
+    },
+    {
+      "epoch": 0.7784899130964493,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016862650477245672,
+      "loss": 4.289,
+      "step": 7508
+    },
+    {
+      "epoch": 0.7785936011509375,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016861860366122552,
+      "loss": 4.3078,
+      "step": 7509
+    },
+    {
+      "epoch": 0.7786972892054255,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016861070174036734,
+      "loss": 4.31,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7788009772599136,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001686027990099754,
+      "loss": 4.2991,
+      "step": 7511
+    },
+    {
+      "epoch": 0.7789046653144016,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016859489547014293,
+      "loss": 4.2701,
+      "step": 7512
+    },
+    {
+      "epoch": 0.7790083533688897,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016858699112096322,
+      "loss": 4.2792,
+      "step": 7513
+    },
+    {
+      "epoch": 0.7791120414233778,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016857908596252952,
+      "loss": 4.3019,
+      "step": 7514
+    },
+    {
+      "epoch": 0.7792157294778659,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001685711799949351,
+      "loss": 4.3028,
+      "step": 7515
+    },
+    {
+      "epoch": 0.7793194175323539,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001685632732182732,
+      "loss": 4.3022,
+      "step": 7516
+    },
+    {
+      "epoch": 0.779423105586842,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016855536563263715,
+      "loss": 4.2835,
+      "step": 7517
+    },
+    {
+      "epoch": 0.77952679364133,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016854745723812032,
+      "loss": 4.332,
+      "step": 7518
+    },
+    {
+      "epoch": 0.7796304816958182,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001685395480348159,
+      "loss": 4.3164,
+      "step": 7519
+    },
+    {
+      "epoch": 0.7797341697503062,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016853163802281728,
+      "loss": 4.2839,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7798378578047943,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016852372720221775,
+      "loss": 4.3423,
+      "step": 7521
+    },
+    {
+      "epoch": 0.7799415458592823,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016851581557311072,
+      "loss": 4.3084,
+      "step": 7522
+    },
+    {
+      "epoch": 0.7800452339137705,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016850790313558943,
+      "loss": 4.3181,
+      "step": 7523
+    },
+    {
+      "epoch": 0.7801489219682585,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016849998988974733,
+      "loss": 4.2593,
+      "step": 7524
+    },
+    {
+      "epoch": 0.7802526100227466,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016849207583567776,
+      "loss": 4.2949,
+      "step": 7525
+    },
+    {
+      "epoch": 0.7803562980772346,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016848416097347407,
+      "loss": 4.2945,
+      "step": 7526
+    },
+    {
+      "epoch": 0.7804599861317227,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001684762453032297,
+      "loss": 4.3089,
+      "step": 7527
+    },
+    {
+      "epoch": 0.7805636741862108,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.000168468328825038,
+      "loss": 4.2943,
+      "step": 7528
+    },
+    {
+      "epoch": 0.7806673622406989,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016846041153899242,
+      "loss": 4.333,
+      "step": 7529
+    },
+    {
+      "epoch": 0.7807710502951869,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016845249344518633,
+      "loss": 4.3041,
+      "step": 7530
+    },
+    {
+      "epoch": 0.780874738349675,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016844457454371316,
+      "loss": 4.2783,
+      "step": 7531
+    },
+    {
+      "epoch": 0.780978426404163,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016843665483466636,
+      "loss": 4.3132,
+      "step": 7532
+    },
+    {
+      "epoch": 0.7810821144586512,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001684287343181394,
+      "loss": 4.2628,
+      "step": 7533
+    },
+    {
+      "epoch": 0.7811858025131392,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001684208129942257,
+      "loss": 4.2853,
+      "step": 7534
+    },
+    {
+      "epoch": 0.7812894905676273,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016841289086301874,
+      "loss": 4.3297,
+      "step": 7535
+    },
+    {
+      "epoch": 0.7813931786221153,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016840496792461194,
+      "loss": 4.2628,
+      "step": 7536
+    },
+    {
+      "epoch": 0.7814968666766035,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016839704417909886,
+      "loss": 4.2701,
+      "step": 7537
+    },
+    {
+      "epoch": 0.7816005547310915,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016838911962657296,
+      "loss": 4.3012,
+      "step": 7538
+    },
+    {
+      "epoch": 0.7817042427855796,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016838119426712775,
+      "loss": 4.3009,
+      "step": 7539
+    },
+    {
+      "epoch": 0.7818079308400676,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016837326810085668,
+      "loss": 4.2807,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7819116188945557,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016836534112785336,
+      "loss": 4.3013,
+      "step": 7541
+    },
+    {
+      "epoch": 0.7820153069490438,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016835741334821126,
+      "loss": 4.2881,
+      "step": 7542
+    },
+    {
+      "epoch": 0.7821189950035319,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.000168349484762024,
+      "loss": 4.2784,
+      "step": 7543
+    },
+    {
+      "epoch": 0.7822226830580199,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.000168341555369385,
+      "loss": 4.2688,
+      "step": 7544
+    },
+    {
+      "epoch": 0.782326371112508,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001683336251703879,
+      "loss": 4.3094,
+      "step": 7545
+    },
+    {
+      "epoch": 0.782430059166996,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016832569416512624,
+      "loss": 4.2929,
+      "step": 7546
+    },
+    {
+      "epoch": 0.7825337472214842,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016831776235369365,
+      "loss": 4.2823,
+      "step": 7547
+    },
+    {
+      "epoch": 0.7826374352759723,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016830982973618368,
+      "loss": 4.2762,
+      "step": 7548
+    },
+    {
+      "epoch": 0.7827411233304603,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016830189631268988,
+      "loss": 4.3096,
+      "step": 7549
+    },
+    {
+      "epoch": 0.7828448113849484,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016829396208330595,
+      "loss": 4.2435,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7829484994394365,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016828602704812543,
+      "loss": 4.2706,
+      "step": 7551
+    },
+    {
+      "epoch": 0.7830521874939246,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016827809120724199,
+      "loss": 4.3124,
+      "step": 7552
+    },
+    {
+      "epoch": 0.7831558755484126,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016827015456074925,
+      "loss": 4.3115,
+      "step": 7553
+    },
+    {
+      "epoch": 0.7832595636029007,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001682622171087408,
+      "loss": 4.2603,
+      "step": 7554
+    },
+    {
+      "epoch": 0.7833632516573887,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001682542788513104,
+      "loss": 4.2819,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7834669397118769,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016824633978855164,
+      "loss": 4.3383,
+      "step": 7556
+    },
+    {
+      "epoch": 0.7835706277663649,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001682383999205582,
+      "loss": 4.2508,
+      "step": 7557
+    },
+    {
+      "epoch": 0.783674315820853,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001682304592474238,
+      "loss": 4.2703,
+      "step": 7558
+    },
+    {
+      "epoch": 0.783778003875341,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016822251776924208,
+      "loss": 4.2566,
+      "step": 7559
+    },
+    {
+      "epoch": 0.7838816919298291,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016821457548610677,
+      "loss": 4.3067,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7839853799843172,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00016820663239811156,
+      "loss": 4.3048,
+      "step": 7561
+    },
+    {
+      "epoch": 0.7840890680388053,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001681986885053502,
+      "loss": 4.2958,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7841927560932933,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016819074380791638,
+      "loss": 4.3104,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7842964441477814,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016818279830590392,
+      "loss": 4.3193,
+      "step": 7564
+    },
+    {
+      "epoch": 0.7844001322022695,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016817485199940648,
+      "loss": 4.2839,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7845038202567576,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016816690488851785,
+      "loss": 4.2268,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7846075083112456,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016815895697333178,
+      "loss": 4.2412,
+      "step": 7567
+    },
+    {
+      "epoch": 0.7847111963657337,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001681510082539421,
+      "loss": 4.2996,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7848148844202217,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016814305873044254,
+      "loss": 4.2629,
+      "step": 7569
+    },
+    {
+      "epoch": 0.7849185724747099,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001681351084029269,
+      "loss": 4.2047,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7850222605291979,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016812715727148906,
+      "loss": 4.2992,
+      "step": 7571
+    },
+    {
+      "epoch": 0.785125948583686,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016811920533622274,
+      "loss": 4.31,
+      "step": 7572
+    },
+    {
+      "epoch": 0.785229636638174,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001681112525972218,
+      "loss": 4.2708,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7853333246926621,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016810329905458008,
+      "loss": 4.2971,
+      "step": 7574
+    },
+    {
+      "epoch": 0.7854370127471502,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001680953447083914,
+      "loss": 4.2911,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7855407008016383,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016808738955874968,
+      "loss": 4.3078,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7856443888561263,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001680794336057487,
+      "loss": 4.266,
+      "step": 7577
+    },
+    {
+      "epoch": 0.7857480769106144,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016807147684948237,
+      "loss": 4.2838,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7858517649651025,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016806351929004457,
+      "loss": 4.3373,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7859554530195906,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016805556092752921,
+      "loss": 4.2938,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7860591410740786,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016804760176203015,
+      "loss": 4.2999,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7861628291285667,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016803964179364132,
+      "loss": 4.3438,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7862665171830547,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001680316810224566,
+      "loss": 4.3247,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7863702052375429,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016802371944857,
+      "loss": 4.3308,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7864738932920309,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016801575707207542,
+      "loss": 4.2712,
+      "step": 7585
+    },
+    {
+      "epoch": 0.786577581346519,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00016800779389306678,
+      "loss": 4.3176,
+      "step": 7586
+    },
+    {
+      "epoch": 0.786681269401007,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016799982991163808,
+      "loss": 4.2888,
+      "step": 7587
+    },
+    {
+      "epoch": 0.7867849574554951,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016799186512788324,
+      "loss": 4.2784,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7868886455099832,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001679838995418963,
+      "loss": 4.2813,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7869923335644713,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001679759331537712,
+      "loss": 4.3295,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7870960216189593,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001679679659636019,
+      "loss": 4.2676,
+      "step": 7591
+    },
+    {
+      "epoch": 0.7871997096734474,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016795999797148244,
+      "loss": 4.2144,
+      "step": 7592
+    },
+    {
+      "epoch": 0.7873033977279356,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016795202917750692,
+      "loss": 4.2928,
+      "step": 7593
+    },
+    {
+      "epoch": 0.7874070857824236,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001679440595817692,
+      "loss": 4.2746,
+      "step": 7594
+    },
+    {
+      "epoch": 0.7875107738369117,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016793608918436347,
+      "loss": 4.308,
+      "step": 7595
+    },
+    {
+      "epoch": 0.7876144618913997,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016792811798538364,
+      "loss": 4.2943,
+      "step": 7596
+    },
+    {
+      "epoch": 0.7877181499458878,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016792014598492386,
+      "loss": 4.2543,
+      "step": 7597
+    },
+    {
+      "epoch": 0.7878218380003759,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016791217318307814,
+      "loss": 4.2815,
+      "step": 7598
+    },
+    {
+      "epoch": 0.787925526054864,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016790419957994054,
+      "loss": 4.2256,
+      "step": 7599
+    },
+    {
+      "epoch": 0.788029214109352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001678962251756052,
+      "loss": 4.2607,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7881329021638401,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016788824997016616,
+      "loss": 4.2988,
+      "step": 7601
+    },
+    {
+      "epoch": 0.7882365902183281,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016788027396371751,
+      "loss": 4.2565,
+      "step": 7602
+    },
+    {
+      "epoch": 0.7883402782728163,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016787229715635343,
+      "loss": 4.3026,
+      "step": 7603
+    },
+    {
+      "epoch": 0.7884439663273043,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016786431954816792,
+      "loss": 4.3119,
+      "step": 7604
+    },
+    {
+      "epoch": 0.7885476543817924,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016785634113925522,
+      "loss": 4.232,
+      "step": 7605
+    },
+    {
+      "epoch": 0.7886513424362804,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001678483619297094,
+      "loss": 4.3448,
+      "step": 7606
+    },
+    {
+      "epoch": 0.7887550304907686,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016784038191962463,
+      "loss": 4.272,
+      "step": 7607
+    },
+    {
+      "epoch": 0.7888587185452566,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00016783240110909508,
+      "loss": 4.3144,
+      "step": 7608
+    },
+    {
+      "epoch": 0.7889624065997447,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001678244194982149,
+      "loss": 4.2749,
+      "step": 7609
+    },
+    {
+      "epoch": 0.7890660946542327,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016781643708707828,
+      "loss": 4.2794,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7891697827087208,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016780845387577936,
+      "loss": 4.295,
+      "step": 7611
+    },
+    {
+      "epoch": 0.7892734707632089,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016780046986441236,
+      "loss": 4.3052,
+      "step": 7612
+    },
+    {
+      "epoch": 0.789377158817697,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001677924850530715,
+      "loss": 4.3186,
+      "step": 7613
+    },
+    {
+      "epoch": 0.789480846872185,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016778449944185092,
+      "loss": 4.3061,
+      "step": 7614
+    },
+    {
+      "epoch": 0.7895845349266731,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016777651303084495,
+      "loss": 4.2802,
+      "step": 7615
+    },
+    {
+      "epoch": 0.7896882229811611,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00016776852582014778,
+      "loss": 4.2582,
+      "step": 7616
+    },
+    {
+      "epoch": 0.7897919110356493,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001677605378098536,
+      "loss": 4.3021,
+      "step": 7617
+    },
+    {
+      "epoch": 0.7898955990901373,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016775254900005675,
+      "loss": 4.2635,
+      "step": 7618
+    },
+    {
+      "epoch": 0.7899992871446254,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016774455939085137,
+      "loss": 4.3136,
+      "step": 7619
+    },
+    {
+      "epoch": 0.7901029751991134,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016773656898233185,
+      "loss": 4.2791,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7902066632536016,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001677285777745924,
+      "loss": 4.2588,
+      "step": 7621
+    },
+    {
+      "epoch": 0.7903103513080896,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00016772058576772735,
+      "loss": 4.2855,
+      "step": 7622
+    },
+    {
+      "epoch": 0.7904140393625777,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016771259296183093,
+      "loss": 4.2701,
+      "step": 7623
+    },
+    {
+      "epoch": 0.7905177274170657,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001677045993569975,
+      "loss": 4.2737,
+      "step": 7624
+    },
+    {
+      "epoch": 0.7906214154715538,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016769660495332138,
+      "loss": 4.2514,
+      "step": 7625
+    },
+    {
+      "epoch": 0.7907251035260419,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001676886097508969,
+      "loss": 4.3126,
+      "step": 7626
+    },
+    {
+      "epoch": 0.79082879158053,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016768061374981832,
+      "loss": 4.3186,
+      "step": 7627
+    },
+    {
+      "epoch": 0.790932479635018,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016767261695018007,
+      "loss": 4.2789,
+      "step": 7628
+    },
+    {
+      "epoch": 0.7910361676895061,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016766461935207647,
+      "loss": 4.2943,
+      "step": 7629
+    },
+    {
+      "epoch": 0.7911398557439941,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001676566209556019,
+      "loss": 4.3317,
+      "step": 7630
+    },
+    {
+      "epoch": 0.7912435437984823,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001676486217608507,
+      "loss": 4.2931,
+      "step": 7631
+    },
+    {
+      "epoch": 0.7913472318529703,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016764062176791725,
+      "loss": 4.2701,
+      "step": 7632
+    },
+    {
+      "epoch": 0.7914509199074584,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.000167632620976896,
+      "loss": 4.2703,
+      "step": 7633
+    },
+    {
+      "epoch": 0.7915546079619464,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016762461938788132,
+      "loss": 4.2488,
+      "step": 7634
+    },
+    {
+      "epoch": 0.7916582960164346,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001676166170009676,
+      "loss": 4.2902,
+      "step": 7635
+    },
+    {
+      "epoch": 0.7917619840709226,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016760861381624926,
+      "loss": 4.312,
+      "step": 7636
+    },
+    {
+      "epoch": 0.7918656721254107,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016760060983382077,
+      "loss": 4.2857,
+      "step": 7637
+    },
+    {
+      "epoch": 0.7919693601798988,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016759260505377652,
+      "loss": 4.2868,
+      "step": 7638
+    },
+    {
+      "epoch": 0.7920730482343868,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016758459947621097,
+      "loss": 4.2934,
+      "step": 7639
+    },
+    {
+      "epoch": 0.792176736288875,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001675765931012186,
+      "loss": 4.3025,
+      "step": 7640
+    },
+    {
+      "epoch": 0.792280424343363,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016756858592889383,
+      "loss": 4.3105,
+      "step": 7641
+    },
+    {
+      "epoch": 0.7923841123978511,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016756057795933122,
+      "loss": 4.2634,
+      "step": 7642
+    },
+    {
+      "epoch": 0.7924878004523391,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016755256919262517,
+      "loss": 4.2992,
+      "step": 7643
+    },
+    {
+      "epoch": 0.7925914885068273,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016754455962887023,
+      "loss": 4.2771,
+      "step": 7644
+    },
+    {
+      "epoch": 0.7926951765613153,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016753654926816088,
+      "loss": 4.3138,
+      "step": 7645
+    },
+    {
+      "epoch": 0.7927988646158034,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016752853811059163,
+      "loss": 4.2902,
+      "step": 7646
+    },
+    {
+      "epoch": 0.7929025526702914,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016752052615625704,
+      "loss": 4.3219,
+      "step": 7647
+    },
+    {
+      "epoch": 0.7930062407247795,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016751251340525159,
+      "loss": 4.3061,
+      "step": 7648
+    },
+    {
+      "epoch": 0.7931099287792676,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016750449985766984,
+      "loss": 4.3073,
+      "step": 7649
+    },
+    {
+      "epoch": 0.7932136168337557,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016749648551360634,
+      "loss": 4.2367,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7933173048882437,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016748847037315566,
+      "loss": 4.284,
+      "step": 7651
+    },
+    {
+      "epoch": 0.7934209929427318,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001674804544364124,
+      "loss": 4.3115,
+      "step": 7652
+    },
+    {
+      "epoch": 0.7935246809972198,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001674724377034711,
+      "loss": 4.2805,
+      "step": 7653
+    },
+    {
+      "epoch": 0.793628369051708,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016746442017442632,
+      "loss": 4.3198,
+      "step": 7654
+    },
+    {
+      "epoch": 0.793732057106196,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016745640184937272,
+      "loss": 4.3037,
+      "step": 7655
+    },
+    {
+      "epoch": 0.7938357451606841,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00016744838272840488,
+      "loss": 4.2644,
+      "step": 7656
+    },
+    {
+      "epoch": 0.7939394332151721,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001674403628116174,
+      "loss": 4.2905,
+      "step": 7657
+    },
+    {
+      "epoch": 0.7940431212696603,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00016743234209910498,
+      "loss": 4.3112,
+      "step": 7658
+    },
+    {
+      "epoch": 0.7941468093241483,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016742432059096214,
+      "loss": 4.2794,
+      "step": 7659
+    },
+    {
+      "epoch": 0.7942504973786364,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016741629828728364,
+      "loss": 4.2744,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7943541854331244,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016740827518816405,
+      "loss": 4.2451,
+      "step": 7661
+    },
+    {
+      "epoch": 0.7944578734876125,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016740025129369807,
+      "loss": 4.3041,
+      "step": 7662
+    },
+    {
+      "epoch": 0.7945615615421006,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016739222660398038,
+      "loss": 4.3188,
+      "step": 7663
+    },
+    {
+      "epoch": 0.7946652495965887,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016738420111910566,
+      "loss": 4.2473,
+      "step": 7664
+    },
+    {
+      "epoch": 0.7947689376510767,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001673761748391686,
+      "loss": 4.2878,
+      "step": 7665
+    },
+    {
+      "epoch": 0.7948726257055648,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00016736814776426387,
+      "loss": 4.2988,
+      "step": 7666
+    },
+    {
+      "epoch": 0.7949763137600528,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001673601198944862,
+      "loss": 4.2549,
+      "step": 7667
+    },
+    {
+      "epoch": 0.795080001814541,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016735209122993033,
+      "loss": 4.3131,
+      "step": 7668
+    },
+    {
+      "epoch": 0.795183689869029,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000167344061770691,
+      "loss": 4.3129,
+      "step": 7669
+    },
+    {
+      "epoch": 0.7952873779235171,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016733603151686288,
+      "loss": 4.2754,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7953910659780051,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00016732800046854082,
+      "loss": 4.3063,
+      "step": 7671
+    },
+    {
+      "epoch": 0.7954947540324933,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001673199686258195,
+      "loss": 4.2615,
+      "step": 7672
+    },
+    {
+      "epoch": 0.7955984420869813,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001673119359887937,
+      "loss": 4.3246,
+      "step": 7673
+    },
+    {
+      "epoch": 0.7957021301414694,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001673039025575582,
+      "loss": 4.2785,
+      "step": 7674
+    },
+    {
+      "epoch": 0.7958058181959574,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016729586833220782,
+      "loss": 4.2913,
+      "step": 7675
+    },
+    {
+      "epoch": 0.7959095062504455,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00016728783331283734,
+      "loss": 4.3111,
+      "step": 7676
+    },
+    {
+      "epoch": 0.7960131943049336,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016727979749954153,
+      "loss": 4.3188,
+      "step": 7677
+    },
+    {
+      "epoch": 0.7961168823594217,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016727176089241521,
+      "loss": 4.2945,
+      "step": 7678
+    },
+    {
+      "epoch": 0.7962205704139097,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016726372349155325,
+      "loss": 4.2384,
+      "step": 7679
+    },
+    {
+      "epoch": 0.7963242584683978,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001672556852970504,
+      "loss": 4.2974,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7964279465228858,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016724764630900163,
+      "loss": 4.2904,
+      "step": 7681
+    },
+    {
+      "epoch": 0.796531634577374,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001672396065275017,
+      "loss": 4.2817,
+      "step": 7682
+    },
+    {
+      "epoch": 0.7966353226318621,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016723156595264545,
+      "loss": 4.2707,
+      "step": 7683
+    },
+    {
+      "epoch": 0.7967390106863501,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016722352458452782,
+      "loss": 4.3052,
+      "step": 7684
+    },
+    {
+      "epoch": 0.7968426987408382,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016721548242324365,
+      "loss": 4.3118,
+      "step": 7685
+    },
+    {
+      "epoch": 0.7969463867953263,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016720743946888785,
+      "loss": 4.2763,
+      "step": 7686
+    },
+    {
+      "epoch": 0.7970500748498144,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016719939572155528,
+      "loss": 4.2655,
+      "step": 7687
+    },
+    {
+      "epoch": 0.7971537629043024,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016719135118134092,
+      "loss": 4.3199,
+      "step": 7688
+    },
+    {
+      "epoch": 0.7972574509587905,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016718330584833958,
+      "loss": 4.3064,
+      "step": 7689
+    },
+    {
+      "epoch": 0.7973611390132785,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016717525972264626,
+      "loss": 4.2913,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7974648270677667,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001671672128043559,
+      "loss": 4.3011,
+      "step": 7691
+    },
+    {
+      "epoch": 0.7975685151222547,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016715916509356344,
+      "loss": 4.2835,
+      "step": 7692
+    },
+    {
+      "epoch": 0.7976722031767428,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001671511165903638,
+      "loss": 4.272,
+      "step": 7693
+    },
+    {
+      "epoch": 0.7977758912312308,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016714306729485195,
+      "loss": 4.2619,
+      "step": 7694
+    },
+    {
+      "epoch": 0.797879579285719,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001671350172071229,
+      "loss": 4.2933,
+      "step": 7695
+    },
+    {
+      "epoch": 0.797983267340207,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00016712696632727164,
+      "loss": 4.3161,
+      "step": 7696
+    },
+    {
+      "epoch": 0.7980869553946951,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001671189146553931,
+      "loss": 4.2422,
+      "step": 7697
+    },
+    {
+      "epoch": 0.7981906434491831,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001671108621915823,
+      "loss": 4.2794,
+      "step": 7698
+    },
+    {
+      "epoch": 0.7982943315036712,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001671028089359343,
+      "loss": 4.2902,
+      "step": 7699
+    },
+    {
+      "epoch": 0.7983980195581593,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016709475488854407,
+      "loss": 4.3202,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7985017076126474,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016708670004950666,
+      "loss": 4.3195,
+      "step": 7701
+    },
+    {
+      "epoch": 0.7986053956671354,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001670786444189171,
+      "loss": 4.3011,
+      "step": 7702
+    },
+    {
+      "epoch": 0.7987090837216235,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016707058799687044,
+      "loss": 4.3031,
+      "step": 7703
+    },
+    {
+      "epoch": 0.7988127717761115,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00016706253078346175,
+      "loss": 4.3169,
+      "step": 7704
+    },
+    {
+      "epoch": 0.7989164598305997,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016705447277878607,
+      "loss": 4.2572,
+      "step": 7705
+    },
+    {
+      "epoch": 0.7990201478850877,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001670464139829385,
+      "loss": 4.294,
+      "step": 7706
+    },
+    {
+      "epoch": 0.7991238359395758,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016703835439601413,
+      "loss": 4.3035,
+      "step": 7707
+    },
+    {
+      "epoch": 0.7992275239940638,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016703029401810802,
+      "loss": 4.2962,
+      "step": 7708
+    },
+    {
+      "epoch": 0.799331212048552,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016702223284931528,
+      "loss": 4.3065,
+      "step": 7709
+    },
+    {
+      "epoch": 0.79943490010304,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016701417088973107,
+      "loss": 4.3254,
+      "step": 7710
+    },
+    {
+      "epoch": 0.7995385881575281,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016700610813945044,
+      "loss": 4.3224,
+      "step": 7711
+    },
+    {
+      "epoch": 0.7996422762120161,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016699804459856862,
+      "loss": 4.3194,
+      "step": 7712
+    },
+    {
+      "epoch": 0.7997459642665042,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016698998026718064,
+      "loss": 4.2604,
+      "step": 7713
+    },
+    {
+      "epoch": 0.7998496523209923,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00016698191514538172,
+      "loss": 4.3033,
+      "step": 7714
+    },
+    {
+      "epoch": 0.7999533403754804,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016697384923326704,
+      "loss": 4.3104,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8000570284299684,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001669657825309317,
+      "loss": 4.315,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8001607164844565,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016695771503847092,
+      "loss": 4.3228,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8002644045389445,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016694964675597986,
+      "loss": 4.2683,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8003680925934327,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016694157768355376,
+      "loss": 4.298,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8004717806479207,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016693350782128778,
+      "loss": 4.2719,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8005754687024088,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016692543716927718,
+      "loss": 4.2916,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8006791567568968,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00016691736572761715,
+      "loss": 4.3065,
+      "step": 7722
+    },
+    {
+      "epoch": 0.800782844811385,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016690929349640296,
+      "loss": 4.3154,
+      "step": 7723
+    },
+    {
+      "epoch": 0.800886532865873,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016690122047572983,
+      "loss": 4.3292,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8009902209203611,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.000166893146665693,
+      "loss": 4.3223,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8010939089748491,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016688507206638777,
+      "loss": 4.2767,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8011975970293372,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016687699667790936,
+      "loss": 4.3026,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8013012850838254,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001668689205003531,
+      "loss": 4.2776,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8014049731383134,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016686084353381426,
+      "loss": 4.2744,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8015086611928015,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00016685276577838815,
+      "loss": 4.2787,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8016123492472895,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016684468723417005,
+      "loss": 4.2959,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8017160373017777,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016683660790125533,
+      "loss": 4.3154,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8018197253562657,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001668285277797393,
+      "loss": 4.2811,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8019234134107538,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001668204468697172,
+      "loss": 4.2388,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8020271014652418,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001668123651712845,
+      "loss": 4.3022,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8021307895197299,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016680428268453653,
+      "loss": 4.2492,
+      "step": 7736
+    },
+    {
+      "epoch": 0.802234477574218,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016679619940956864,
+      "loss": 4.2857,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8023381656287061,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001667881153464762,
+      "loss": 4.3029,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8024418536831941,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016678003049535459,
+      "loss": 4.2936,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8025455417376822,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001667719448562992,
+      "loss": 4.3072,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8026492297921702,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016676385842940547,
+      "loss": 4.2703,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8027529178466584,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016675577121476876,
+      "loss": 4.2924,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8028566059011464,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016674768321248452,
+      "loss": 4.3096,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8029602939556345,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001667395944226482,
+      "loss": 4.3014,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8030639820101225,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016673150484535518,
+      "loss": 4.2482,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8031676700646107,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016672341448070095,
+      "loss": 4.2823,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8032713581190987,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016671532332878094,
+      "loss": 4.3267,
+      "step": 7747
+    },
+    {
+      "epoch": 0.8033750461735868,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016670723138969065,
+      "loss": 4.3252,
+      "step": 7748
+    },
+    {
+      "epoch": 0.8034787342280748,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016669913866352556,
+      "loss": 4.2556,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8035824222825629,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001666910451503811,
+      "loss": 4.2379,
+      "step": 7750
+    },
+    {
+      "epoch": 0.803686110337051,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016668295085035286,
+      "loss": 4.2545,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8037897983915391,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016667485576353624,
+      "loss": 4.3009,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8038934864460271,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001666667598900268,
+      "loss": 4.3247,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8039971745005152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016665866322992007,
+      "loss": 4.2759,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8041008625550032,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001666505657833116,
+      "loss": 4.2795,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8042045506094914,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001666424675502969,
+      "loss": 4.2908,
+      "step": 7756
+    },
+    {
+      "epoch": 0.8043082386639794,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001666343685309715,
+      "loss": 4.2934,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8044119267184675,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.000166626268725431,
+      "loss": 4.2682,
+      "step": 7758
+    },
+    {
+      "epoch": 0.8045156147729555,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00016661816813377095,
+      "loss": 4.2481,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8046193028274436,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00016661006675608694,
+      "loss": 4.2937,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8047229908819317,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016660196459247458,
+      "loss": 4.3113,
+      "step": 7761
+    },
+    {
+      "epoch": 0.8048266789364198,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001665938616430294,
+      "loss": 4.3276,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8049303669909078,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016658575790784704,
+      "loss": 4.2966,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8050340550453959,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016657765338702317,
+      "loss": 4.3491,
+      "step": 7764
+    },
+    {
+      "epoch": 0.805137743099884,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016656954808065333,
+      "loss": 4.3034,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8052414311543721,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001665614419888332,
+      "loss": 4.3113,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8053451192088601,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001665533351116584,
+      "loss": 4.3144,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8054488072633482,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016654522744922461,
+      "loss": 4.32,
+      "step": 7768
+    },
+    {
+      "epoch": 0.8055524953178362,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00016653711900162748,
+      "loss": 4.293,
+      "step": 7769
+    },
+    {
+      "epoch": 0.8056561833723244,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016652900976896263,
+      "loss": 4.3441,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8057598714268124,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016652089975132585,
+      "loss": 4.2961,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8058635594813005,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016651278894881278,
+      "loss": 4.2946,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8059672475357886,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016650467736151905,
+      "loss": 4.3085,
+      "step": 7773
+    },
+    {
+      "epoch": 0.8060709355902766,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016649656498954042,
+      "loss": 4.2685,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8061746236447648,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016648845183297266,
+      "loss": 4.2989,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8062783116992528,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016648033789191144,
+      "loss": 4.3006,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8063819997537409,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001664722231664525,
+      "loss": 4.2543,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8064856878082289,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001664641076566916,
+      "loss": 4.2665,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8065893758627171,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016645599136272447,
+      "loss": 4.3282,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8066930639172051,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001664478742846469,
+      "loss": 4.3053,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8067967519716932,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016643975642255466,
+      "loss": 4.293,
+      "step": 7781
+    },
+    {
+      "epoch": 0.8069004400261812,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001664316377765435,
+      "loss": 4.2947,
+      "step": 7782
+    },
+    {
+      "epoch": 0.8070041280806693,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016642351834670924,
+      "loss": 4.2965,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8071078161351574,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016641539813314768,
+      "loss": 4.2896,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8072115041896455,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001664072771359546,
+      "loss": 4.2974,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8073151922441335,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001663991553552259,
+      "loss": 4.2654,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8074188802986216,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001663910327910573,
+      "loss": 4.2828,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8075225683531096,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016638290944354474,
+      "loss": 4.2936,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8076262564075978,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000166374785312784,
+      "loss": 4.3182,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8077299444620858,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016636666039887094,
+      "loss": 4.2962,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8078336325165739,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016635853470190146,
+      "loss": 4.316,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8079373205710619,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016635040822197142,
+      "loss": 4.2575,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8080410086255501,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016634228095917667,
+      "loss": 4.2889,
+      "step": 7793
+    },
+    {
+      "epoch": 0.8081446966800381,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016633415291361314,
+      "loss": 4.2599,
+      "step": 7794
+    },
+    {
+      "epoch": 0.8082483847345262,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001663260240853767,
+      "loss": 4.2845,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8083520727890142,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016631789447456337,
+      "loss": 4.2388,
+      "step": 7796
+    },
+    {
+      "epoch": 0.8084557608435023,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016630976408126891,
+      "loss": 4.2798,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8085594488979904,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016630163290558937,
+      "loss": 4.2705,
+      "step": 7798
+    },
+    {
+      "epoch": 0.8086631369524785,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001662935009476206,
+      "loss": 4.2865,
+      "step": 7799
+    },
+    {
+      "epoch": 0.8087668250069665,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001662853682074586,
+      "loss": 4.2664,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8088705130614546,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016627723468519935,
+      "loss": 4.2632,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8089742011159426,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016626910038093878,
+      "loss": 4.296,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8090778891704308,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016626096529477288,
+      "loss": 4.3453,
+      "step": 7803
+    },
+    {
+      "epoch": 0.8091815772249188,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001662528294267976,
+      "loss": 4.2582,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8092852652794069,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.000166244692777109,
+      "loss": 4.2668,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8093889533338949,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016623655534580303,
+      "loss": 4.2773,
+      "step": 7806
+    },
+    {
+      "epoch": 0.8094926413883831,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001662284171329757,
+      "loss": 4.3475,
+      "step": 7807
+    },
+    {
+      "epoch": 0.8095963294428711,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016622027813872312,
+      "loss": 4.3152,
+      "step": 7808
+    },
+    {
+      "epoch": 0.8097000174973592,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016621213836314123,
+      "loss": 4.2213,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8098037055518472,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016620399780632608,
+      "loss": 4.2859,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8099073936063353,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016619585646837374,
+      "loss": 4.2515,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8100110816608234,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016618771434938025,
+      "loss": 4.2569,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8101147697153115,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00016617957144944172,
+      "loss": 4.2405,
+      "step": 7813
+    },
+    {
+      "epoch": 0.8102184577697995,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001661714277686542,
+      "loss": 4.2934,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8103221458242876,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016616328330711376,
+      "loss": 4.3069,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8104258338787758,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016615513806491654,
+      "loss": 4.2408,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8105295219332638,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001661469920421586,
+      "loss": 4.334,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8106332099877519,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016613884523893608,
+      "loss": 4.2613,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8107368980422399,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001661306976553451,
+      "loss": 4.3069,
+      "step": 7819
+    },
+    {
+      "epoch": 0.810840586096728,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016612254929148179,
+      "loss": 4.3041,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8109442741512161,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001661144001474423,
+      "loss": 4.3017,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8110479622057042,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016610625022332276,
+      "loss": 4.2651,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8111516502601922,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016609809951921936,
+      "loss": 4.3191,
+      "step": 7823
+    },
+    {
+      "epoch": 0.8112553383146803,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016608994803522824,
+      "loss": 4.2134,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8113590263691683,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001660817957714456,
+      "loss": 4.3125,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8114627144236565,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016607364272796762,
+      "loss": 4.2822,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8115664024781445,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001660654889048905,
+      "loss": 4.2994,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8116700905326326,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016605733430231044,
+      "loss": 4.2685,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8117737785871206,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016604917892032366,
+      "loss": 4.2875,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8118774666416088,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001660410227590264,
+      "loss": 4.2983,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8119811546960968,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016603286581851488,
+      "loss": 4.2994,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8120848427505849,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001660247080988853,
+      "loss": 4.2526,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8121885308050729,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016601654960023398,
+      "loss": 4.2908,
+      "step": 7833
+    },
+    {
+      "epoch": 0.812292218859561,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001660083903226572,
+      "loss": 4.2445,
+      "step": 7834
+    },
+    {
+      "epoch": 0.8123959069140491,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016600023026625116,
+      "loss": 4.2771,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8124995949685372,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016599206943111215,
+      "loss": 4.2637,
+      "step": 7836
+    },
+    {
+      "epoch": 0.8126032830230252,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001659839078173365,
+      "loss": 4.2666,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8127069710775133,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016597574542502047,
+      "loss": 4.2428,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8128106591320013,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001659675822542604,
+      "loss": 4.28,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8129143471864895,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016595941830515256,
+      "loss": 4.2637,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8130180352409775,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016595125357779332,
+      "loss": 4.272,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8131217232954656,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016594308807227904,
+      "loss": 4.265,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8132254113499536,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016593492178870598,
+      "loss": 4.2538,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8133290994044418,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016592675472717054,
+      "loss": 4.2471,
+      "step": 7844
+    },
+    {
+      "epoch": 0.8134327874589298,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001659185868877691,
+      "loss": 4.2553,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8135364755134179,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016591041827059802,
+      "loss": 4.3116,
+      "step": 7846
+    },
+    {
+      "epoch": 0.8136401635679059,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016590224887575366,
+      "loss": 4.2769,
+      "step": 7847
+    },
+    {
+      "epoch": 0.813743851622394,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016589407870333243,
+      "loss": 4.3086,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8138475396768821,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016588590775343072,
+      "loss": 4.2312,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8139512277313702,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016587773602614494,
+      "loss": 4.3059,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8140549157858582,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001658695635215715,
+      "loss": 4.2705,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8141586038403463,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016586139023980686,
+      "loss": 4.2674,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8142622918948343,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016585321618094742,
+      "loss": 4.3153,
+      "step": 7853
+    },
+    {
+      "epoch": 0.8143659799493225,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016584504134508965,
+      "loss": 4.3233,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8144696680038105,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016583686573233,
+      "loss": 4.304,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8145733560582986,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001658286893427649,
+      "loss": 4.2762,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8146770441127866,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016582051217649087,
+      "loss": 4.3054,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8147807321672748,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016581233423360433,
+      "loss": 4.2704,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8148844202217628,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016580415551420189,
+      "loss": 4.3082,
+      "step": 7859
+    },
+    {
+      "epoch": 0.8149881082762509,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016579597601837993,
+      "loss": 4.2897,
+      "step": 7860
+    },
+    {
+      "epoch": 0.815091796330739,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.000165787795746235,
+      "loss": 4.3225,
+      "step": 7861
+    },
+    {
+      "epoch": 0.815195484385227,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016577961469786364,
+      "loss": 4.3007,
+      "step": 7862
+    },
+    {
+      "epoch": 0.8152991724397152,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016577143287336234,
+      "loss": 4.3154,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8154028604942032,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016576325027282764,
+      "loss": 4.2933,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8155065485486913,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016575506689635612,
+      "loss": 4.2656,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8156102366031793,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016574688274404432,
+      "loss": 4.239,
+      "step": 7866
+    },
+    {
+      "epoch": 0.8157139246576675,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016573869781598882,
+      "loss": 4.2593,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8158176127121555,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016573051211228614,
+      "loss": 4.2934,
+      "step": 7868
+    },
+    {
+      "epoch": 0.8159213007666436,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016572232563303292,
+      "loss": 4.2886,
+      "step": 7869
+    },
+    {
+      "epoch": 0.8160249888211316,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001657141383783257,
+      "loss": 4.2693,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8161286768756197,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016570595034826115,
+      "loss": 4.3098,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8162323649301078,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001656977615429358,
+      "loss": 4.2957,
+      "step": 7872
+    },
+    {
+      "epoch": 0.8163360529845959,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016568957196244632,
+      "loss": 4.2775,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8164397410390839,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00016568138160688936,
+      "loss": 4.2855,
+      "step": 7874
+    },
+    {
+      "epoch": 0.816543429093572,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001656731904763615,
+      "loss": 4.2679,
+      "step": 7875
+    },
+    {
+      "epoch": 0.81664711714806,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016566499857095942,
+      "loss": 4.3152,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8167508052025482,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016565680589077976,
+      "loss": 4.2852,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8168544932570362,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016564861243591918,
+      "loss": 4.2883,
+      "step": 7878
+    },
+    {
+      "epoch": 0.8169581813115243,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016564041820647438,
+      "loss": 4.2654,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8170618693660123,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016563222320254206,
+      "loss": 4.3057,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8171655574205005,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016562402742421883,
+      "loss": 4.2914,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8172692454749885,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016561583087160145,
+      "loss": 4.2848,
+      "step": 7882
+    },
+    {
+      "epoch": 0.8173729335294766,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016560763354478666,
+      "loss": 4.2953,
+      "step": 7883
+    },
+    {
+      "epoch": 0.8174766215839646,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016559943544387114,
+      "loss": 4.2588,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8175803096384527,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016559123656895158,
+      "loss": 4.2727,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8176839976929408,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016558303692012482,
+      "loss": 4.2575,
+      "step": 7886
+    },
+    {
+      "epoch": 0.8177876857474289,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001655748364974875,
+      "loss": 4.293,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016556663530113648,
+      "loss": 4.267,
+      "step": 7888
+    },
+    {
+      "epoch": 0.817995061856405,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001655584333311684,
+      "loss": 4.3019,
+      "step": 7889
+    },
+    {
+      "epoch": 0.818098749910893,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016555023058768015,
+      "loss": 4.2912,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8182024379653812,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016554202707076847,
+      "loss": 4.3066,
+      "step": 7891
+    },
+    {
+      "epoch": 0.8183061260198692,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016553382278053014,
+      "loss": 4.2105,
+      "step": 7892
+    },
+    {
+      "epoch": 0.8184098140743573,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016552561771706196,
+      "loss": 4.2934,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8185135021288453,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00016551741188046076,
+      "loss": 4.2425,
+      "step": 7894
+    },
+    {
+      "epoch": 0.8186171901833335,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016550920527082336,
+      "loss": 4.2912,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8187208782378215,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001655009978882466,
+      "loss": 4.2229,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8188245662923096,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001654927897328273,
+      "loss": 4.3013,
+      "step": 7897
+    },
+    {
+      "epoch": 0.8189282543467976,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001654845808046623,
+      "loss": 4.2932,
+      "step": 7898
+    },
+    {
+      "epoch": 0.8190319424012857,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016547637110384846,
+      "loss": 4.3038,
+      "step": 7899
+    },
+    {
+      "epoch": 0.8191356304557738,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016546816063048268,
+      "loss": 4.2764,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8192393185102619,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016545994938466175,
+      "loss": 4.2641,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8193430065647499,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016545173736648265,
+      "loss": 4.2777,
+      "step": 7902
+    },
+    {
+      "epoch": 0.819446694619238,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00016544352457604225,
+      "loss": 4.3514,
+      "step": 7903
+    },
+    {
+      "epoch": 0.819550382673726,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001654353110134374,
+      "loss": 4.266,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8196540707282142,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001654270966787651,
+      "loss": 4.2791,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8197577587827023,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00016541888157212218,
+      "loss": 4.2612,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8198614468371903,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001654106656936056,
+      "loss": 4.3454,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8199651348916784,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001654024490433123,
+      "loss": 4.2956,
+      "step": 7908
+    },
+    {
+      "epoch": 0.8200688229461665,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016539423162133926,
+      "loss": 4.2367,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8201725110006546,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016538601342778344,
+      "loss": 4.2403,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8202761990551426,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00016537779446274174,
+      "loss": 4.3065,
+      "step": 7911
+    },
+    {
+      "epoch": 0.8203798871096307,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00016536957472631115,
+      "loss": 4.2632,
+      "step": 7912
+    },
+    {
+      "epoch": 0.8204835751641187,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00016536135421858875,
+      "loss": 4.2896,
+      "step": 7913
+    },
+    {
+      "epoch": 0.8205872632186069,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001653531329396714,
+      "loss": 4.2405,
+      "step": 7914
+    },
+    {
+      "epoch": 0.8206909512730949,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00016534491088965615,
+      "loss": 4.275,
+      "step": 7915
+    },
+    {
+      "epoch": 0.820794639327583,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00016533668806864007,
+      "loss": 4.2577,
+      "step": 7916
+    },
+    {
+      "epoch": 0.820898327382071,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00016532846447672012,
+      "loss": 4.2587,
+      "step": 7917
+    },
+    {
+      "epoch": 0.8210020154365592,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016532024011399335,
+      "loss": 4.3051,
+      "step": 7918
+    },
+    {
+      "epoch": 0.8211057034910472,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00016531201498055676,
+      "loss": 4.3112,
+      "step": 7919
+    },
+    {
+      "epoch": 0.8212093915455353,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016530378907650745,
+      "loss": 4.28,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8213130796000233,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001652955624019425,
+      "loss": 4.2912,
+      "step": 7921
+    },
+    {
+      "epoch": 0.8214167676545114,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016528733495695886,
+      "loss": 4.2139,
+      "step": 7922
+    },
+    {
+      "epoch": 0.8215204557089995,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00016527910674165374,
+      "loss": 4.289,
+      "step": 7923
+    },
+    {
+      "epoch": 0.8216241437634876,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016527087775612413,
+      "loss": 4.2447,
+      "step": 7924
+    },
+    {
+      "epoch": 0.8217278318179756,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00016526264800046717,
+      "loss": 4.305,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8218315198724637,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016525441747477994,
+      "loss": 4.3099,
+      "step": 7926
+    },
+    {
+      "epoch": 0.8219352079269517,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00016524618617915957,
+      "loss": 4.2869,
+      "step": 7927
+    },
+    {
+      "epoch": 0.8220388959814399,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016523795411370316,
+      "loss": 4.2937,
+      "step": 7928
+    },
+    {
+      "epoch": 0.8221425840359279,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00016522972127850784,
+      "loss": 4.2678,
+      "step": 7929
+    },
+    {
+      "epoch": 0.822246272090416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001652214876736708,
+      "loss": 4.2553,
+      "step": 7930
+    },
+    {
+      "epoch": 0.822349960144904,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016521325329928911,
+      "loss": 4.3384,
+      "step": 7931
+    },
+    {
+      "epoch": 0.8224536481993922,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00016520501815546,
+      "loss": 4.3044,
+      "step": 7932
+    },
+    {
+      "epoch": 0.8225573362538802,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016519678224228055,
+      "loss": 4.2717,
+      "step": 7933
+    },
+    {
+      "epoch": 0.8226610243083683,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016518854555984803,
+      "loss": 4.2566,
+      "step": 7934
+    },
+    {
+      "epoch": 0.8227647123628563,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016518030810825957,
+      "loss": 4.3309,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8228684004173444,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016517206988761238,
+      "loss": 4.3035,
+      "step": 7936
+    },
+    {
+      "epoch": 0.8229720884718325,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016516383089800363,
+      "loss": 4.2951,
+      "step": 7937
+    },
+    {
+      "epoch": 0.8230757765263206,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016515559113953056,
+      "loss": 4.2759,
+      "step": 7938
+    },
+    {
+      "epoch": 0.8231794645808086,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001651473506122904,
+      "loss": 4.2353,
+      "step": 7939
+    },
+    {
+      "epoch": 0.8232831526352967,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016513910931638038,
+      "loss": 4.3049,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8233868406897847,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001651308672518977,
+      "loss": 4.2876,
+      "step": 7941
+    },
+    {
+      "epoch": 0.8234905287442729,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016512262441893967,
+      "loss": 4.2671,
+      "step": 7942
+    },
+    {
+      "epoch": 0.8235942167987609,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016511438081760348,
+      "loss": 4.2758,
+      "step": 7943
+    },
+    {
+      "epoch": 0.823697904853249,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016510613644798642,
+      "loss": 4.3016,
+      "step": 7944
+    },
+    {
+      "epoch": 0.823801592907737,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016509789131018577,
+      "loss": 4.2983,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8239052809622252,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016508964540429882,
+      "loss": 4.2565,
+      "step": 7946
+    },
+    {
+      "epoch": 0.8240089690167132,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016508139873042287,
+      "loss": 4.3073,
+      "step": 7947
+    },
+    {
+      "epoch": 0.8241126570712013,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016507315128865517,
+      "loss": 4.3019,
+      "step": 7948
+    },
+    {
+      "epoch": 0.8242163451256893,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001650649030790931,
+      "loss": 4.2778,
+      "step": 7949
+    },
+    {
+      "epoch": 0.8243200331801774,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016505665410183396,
+      "loss": 4.2676,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8244237212346656,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016504840435697504,
+      "loss": 4.2584,
+      "step": 7951
+    },
+    {
+      "epoch": 0.8245274092891536,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001650401538446137,
+      "loss": 4.2756,
+      "step": 7952
+    },
+    {
+      "epoch": 0.8246310973436417,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016503190256484732,
+      "loss": 4.3243,
+      "step": 7953
+    },
+    {
+      "epoch": 0.8247347853981297,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016502365051777326,
+      "loss": 4.3209,
+      "step": 7954
+    },
+    {
+      "epoch": 0.8248384734526178,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001650153977034888,
+      "loss": 4.3155,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8249421615071059,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001650071441220914,
+      "loss": 4.3073,
+      "step": 7956
+    },
+    {
+      "epoch": 0.825045849561594,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016499888977367842,
+      "loss": 4.2505,
+      "step": 7957
+    },
+    {
+      "epoch": 0.825149537616082,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016499063465834723,
+      "loss": 4.2585,
+      "step": 7958
+    },
+    {
+      "epoch": 0.8252532256705701,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016498237877619526,
+      "loss": 4.2695,
+      "step": 7959
+    },
+    {
+      "epoch": 0.8253569137250581,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016497412212731992,
+      "loss": 4.2895,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8254606017795463,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016496586471181863,
+      "loss": 4.3072,
+      "step": 7961
+    },
+    {
+      "epoch": 0.8255642898340343,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016495760652978877,
+      "loss": 4.2812,
+      "step": 7962
+    },
+    {
+      "epoch": 0.8256679778885224,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016494934758132782,
+      "loss": 4.2581,
+      "step": 7963
+    },
+    {
+      "epoch": 0.8257716659430104,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016494108786653327,
+      "loss": 4.2804,
+      "step": 7964
+    },
+    {
+      "epoch": 0.8258753539974986,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016493282738550246,
+      "loss": 4.2712,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8259790420519866,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016492456613833299,
+      "loss": 4.3037,
+      "step": 7966
+    },
+    {
+      "epoch": 0.8260827301064747,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016491630412512223,
+      "loss": 4.311,
+      "step": 7967
+    },
+    {
+      "epoch": 0.8261864181609627,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001649080413459677,
+      "loss": 4.2494,
+      "step": 7968
+    },
+    {
+      "epoch": 0.8262901062154508,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001648997778009669,
+      "loss": 4.2881,
+      "step": 7969
+    },
+    {
+      "epoch": 0.8263937942699389,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016489151349021732,
+      "loss": 4.3125,
+      "step": 7970
+    },
+    {
+      "epoch": 0.826497482324427,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016488324841381648,
+      "loss": 4.2862,
+      "step": 7971
+    },
+    {
+      "epoch": 0.826601170378915,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001648749825718619,
+      "loss": 4.286,
+      "step": 7972
+    },
+    {
+      "epoch": 0.8267048584334031,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016486671596445109,
+      "loss": 4.2535,
+      "step": 7973
+    },
+    {
+      "epoch": 0.8268085464878911,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001648584485916816,
+      "loss": 4.2822,
+      "step": 7974
+    },
+    {
+      "epoch": 0.8269122345423793,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.000164850180453651,
+      "loss": 4.2299,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8270159225968673,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001648419115504568,
+      "loss": 4.3058,
+      "step": 7976
+    },
+    {
+      "epoch": 0.8271196106513554,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016483364188219656,
+      "loss": 4.3092,
+      "step": 7977
+    },
+    {
+      "epoch": 0.8272232987058434,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001648253714489679,
+      "loss": 4.3129,
+      "step": 7978
+    },
+    {
+      "epoch": 0.8273269867603316,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001648171002508684,
+      "loss": 4.3436,
+      "step": 7979
+    },
+    {
+      "epoch": 0.8274306748148196,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001648088282879956,
+      "loss": 4.2833,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8275343628693077,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016480055556044714,
+      "loss": 4.2834,
+      "step": 7981
+    },
+    {
+      "epoch": 0.8276380509237957,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016479228206832064,
+      "loss": 4.2804,
+      "step": 7982
+    },
+    {
+      "epoch": 0.8277417389782838,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016478400781171366,
+      "loss": 4.2894,
+      "step": 7983
+    },
+    {
+      "epoch": 0.8278454270327719,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016477573279072391,
+      "loss": 4.2667,
+      "step": 7984
+    },
+    {
+      "epoch": 0.82794911508726,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016476745700544896,
+      "loss": 4.3018,
+      "step": 7985
+    },
+    {
+      "epoch": 0.828052803141748,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016475918045598647,
+      "loss": 4.2352,
+      "step": 7986
+    },
+    {
+      "epoch": 0.8281564911962361,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016475090314243412,
+      "loss": 4.2554,
+      "step": 7987
+    },
+    {
+      "epoch": 0.8282601792507241,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016474262506488958,
+      "loss": 4.3044,
+      "step": 7988
+    },
+    {
+      "epoch": 0.8283638673052123,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016473434622345047,
+      "loss": 4.2943,
+      "step": 7989
+    },
+    {
+      "epoch": 0.8284675553597003,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001647260666182145,
+      "loss": 4.3148,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8285712434141884,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016471778624927938,
+      "loss": 4.2165,
+      "step": 7991
+    },
+    {
+      "epoch": 0.8286749314686764,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016470950511674278,
+      "loss": 4.2778,
+      "step": 7992
+    },
+    {
+      "epoch": 0.8287786195231646,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016470122322070245,
+      "loss": 4.274,
+      "step": 7993
+    },
+    {
+      "epoch": 0.8288823075776526,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016469294056125602,
+      "loss": 4.2821,
+      "step": 7994
+    },
+    {
+      "epoch": 0.8289859956321407,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001646846571385013,
+      "loss": 4.2289,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8290896836866288,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016467637295253603,
+      "loss": 4.3043,
+      "step": 7996
+    },
+    {
+      "epoch": 0.8291933717411168,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001646680880034579,
+      "loss": 4.2625,
+      "step": 7997
+    },
+    {
+      "epoch": 0.829297059795605,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016465980229136471,
+      "loss": 4.274,
+      "step": 7998
+    },
+    {
+      "epoch": 0.829400747850093,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016465151581635415,
+      "loss": 4.2753,
+      "step": 7999
+    },
+    {
+      "epoch": 0.8295044359045811,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001646432285785241,
+      "loss": 4.2849,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8296081239590691,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016463494057797226,
+      "loss": 4.2604,
+      "step": 8001
+    },
+    {
+      "epoch": 0.8297118120135573,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016462665181479644,
+      "loss": 4.2643,
+      "step": 8002
+    },
+    {
+      "epoch": 0.8298155000680453,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016461836228909445,
+      "loss": 4.3256,
+      "step": 8003
+    },
+    {
+      "epoch": 0.8299191881225334,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016461007200096407,
+      "loss": 4.309,
+      "step": 8004
+    },
+    {
+      "epoch": 0.8300228761770214,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016460178095050316,
+      "loss": 4.2878,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8301265642315095,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001645934891378095,
+      "loss": 4.3001,
+      "step": 8006
+    },
+    {
+      "epoch": 0.8302302522859976,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016458519656298095,
+      "loss": 4.3005,
+      "step": 8007
+    },
+    {
+      "epoch": 0.8303339403404857,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016457690322611537,
+      "loss": 4.2724,
+      "step": 8008
+    },
+    {
+      "epoch": 0.8304376283949737,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001645686091273106,
+      "loss": 4.2699,
+      "step": 8009
+    },
+    {
+      "epoch": 0.8305413164494618,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016456031426666446,
+      "loss": 4.3013,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8306450045039498,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016455201864427486,
+      "loss": 4.2032,
+      "step": 8011
+    },
+    {
+      "epoch": 0.830748692558438,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001645437222602397,
+      "loss": 4.276,
+      "step": 8012
+    },
+    {
+      "epoch": 0.830852380612926,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016453542511465682,
+      "loss": 4.2788,
+      "step": 8013
+    },
+    {
+      "epoch": 0.8309560686674141,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016452712720762416,
+      "loss": 4.2797,
+      "step": 8014
+    },
+    {
+      "epoch": 0.8310597567219021,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016451882853923958,
+      "loss": 4.2979,
+      "step": 8015
+    },
+    {
+      "epoch": 0.8311634447763903,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016451052910960106,
+      "loss": 4.3467,
+      "step": 8016
+    },
+    {
+      "epoch": 0.8312671328308783,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016450222891880646,
+      "loss": 4.2835,
+      "step": 8017
+    },
+    {
+      "epoch": 0.8313708208853664,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016449392796695373,
+      "loss": 4.284,
+      "step": 8018
+    },
+    {
+      "epoch": 0.8314745089398544,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016448562625414085,
+      "loss": 4.32,
+      "step": 8019
+    },
+    {
+      "epoch": 0.8315781969943425,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001644773237804657,
+      "loss": 4.2855,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8316818850488306,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016446902054602635,
+      "loss": 4.294,
+      "step": 8021
+    },
+    {
+      "epoch": 0.8317855731033187,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016446071655092068,
+      "loss": 4.2968,
+      "step": 8022
+    },
+    {
+      "epoch": 0.8318892611578067,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016445241179524668,
+      "loss": 4.2859,
+      "step": 8023
+    },
+    {
+      "epoch": 0.8319929492122948,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016444410627910235,
+      "loss": 4.2405,
+      "step": 8024
+    },
+    {
+      "epoch": 0.8320966372667828,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001644358000025857,
+      "loss": 4.2623,
+      "step": 8025
+    },
+    {
+      "epoch": 0.832200325321271,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001644274929657947,
+      "loss": 4.2707,
+      "step": 8026
+    },
+    {
+      "epoch": 0.832304013375759,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001644191851688274,
+      "loss": 4.2848,
+      "step": 8027
+    },
+    {
+      "epoch": 0.8324077014302471,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016441087661178181,
+      "loss": 4.3119,
+      "step": 8028
+    },
+    {
+      "epoch": 0.8325113894847351,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016440256729475597,
+      "loss": 4.3303,
+      "step": 8029
+    },
+    {
+      "epoch": 0.8326150775392233,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001643942572178479,
+      "loss": 4.3078,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8327187655937113,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016438594638115567,
+      "loss": 4.3261,
+      "step": 8031
+    },
+    {
+      "epoch": 0.8328224536481994,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016437763478477732,
+      "loss": 4.2575,
+      "step": 8032
+    },
+    {
+      "epoch": 0.8329261417026874,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016436932242881093,
+      "loss": 4.2534,
+      "step": 8033
+    },
+    {
+      "epoch": 0.8330298297571755,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016436100931335458,
+      "loss": 4.2907,
+      "step": 8034
+    },
+    {
+      "epoch": 0.8331335178116636,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016435269543850636,
+      "loss": 4.2756,
+      "step": 8035
+    },
+    {
+      "epoch": 0.8332372058661517,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016434438080436436,
+      "loss": 4.267,
+      "step": 8036
+    },
+    {
+      "epoch": 0.8333408939206397,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016433606541102667,
+      "loss": 4.2986,
+      "step": 8037
+    },
+    {
+      "epoch": 0.8334445819751278,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016432774925859142,
+      "loss": 4.2979,
+      "step": 8038
+    },
+    {
+      "epoch": 0.8335482700296158,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016431943234715673,
+      "loss": 4.2951,
+      "step": 8039
+    },
+    {
+      "epoch": 0.833651958084104,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001643111146768207,
+      "loss": 4.2659,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8337556461385921,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016430279624768152,
+      "loss": 4.2604,
+      "step": 8041
+    },
+    {
+      "epoch": 0.8338593341930801,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016429447705983732,
+      "loss": 4.3355,
+      "step": 8042
+    },
+    {
+      "epoch": 0.8339630222475682,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016428615711338624,
+      "loss": 4.2892,
+      "step": 8043
+    },
+    {
+      "epoch": 0.8340667103020563,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016427783640842646,
+      "loss": 4.282,
+      "step": 8044
+    },
+    {
+      "epoch": 0.8341703983565444,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016426951494505617,
+      "loss": 4.3052,
+      "step": 8045
+    },
+    {
+      "epoch": 0.8342740864110324,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016426119272337352,
+      "loss": 4.2935,
+      "step": 8046
+    },
+    {
+      "epoch": 0.8343777744655205,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016425286974347674,
+      "loss": 4.2614,
+      "step": 8047
+    },
+    {
+      "epoch": 0.8344814625200085,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000164244546005464,
+      "loss": 4.2344,
+      "step": 8048
+    },
+    {
+      "epoch": 0.8345851505744967,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016423622150943355,
+      "loss": 4.2243,
+      "step": 8049
+    },
+    {
+      "epoch": 0.8346888386289847,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016422789625548356,
+      "loss": 4.2501,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8347925266834728,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001642195702437123,
+      "loss": 4.2824,
+      "step": 8051
+    },
+    {
+      "epoch": 0.8348962147379608,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.000164211243474218,
+      "loss": 4.2817,
+      "step": 8052
+    },
+    {
+      "epoch": 0.834999902792449,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00016420291594709889,
+      "loss": 4.3298,
+      "step": 8053
+    },
+    {
+      "epoch": 0.835103590846937,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016419458766245323,
+      "loss": 4.283,
+      "step": 8054
+    },
+    {
+      "epoch": 0.8352072789014251,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00016418625862037932,
+      "loss": 4.334,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8353109669559131,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016417792882097537,
+      "loss": 4.2788,
+      "step": 8056
+    },
+    {
+      "epoch": 0.8354146550104012,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016416959826433974,
+      "loss": 4.3003,
+      "step": 8057
+    },
+    {
+      "epoch": 0.8355183430648893,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016416126695057063,
+      "loss": 4.269,
+      "step": 8058
+    },
+    {
+      "epoch": 0.8356220311193774,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00016415293487976644,
+      "loss": 4.2525,
+      "step": 8059
+    },
+    {
+      "epoch": 0.8357257191738654,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016414460205202539,
+      "loss": 4.2904,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8358294072283535,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00016413626846744584,
+      "loss": 4.2732,
+      "step": 8061
+    },
+    {
+      "epoch": 0.8359330952828415,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016412793412612614,
+      "loss": 4.2605,
+      "step": 8062
+    },
+    {
+      "epoch": 0.8360367833373297,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00016411959902816462,
+      "loss": 4.2769,
+      "step": 8063
+    },
+    {
+      "epoch": 0.8361404713918177,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016411126317365958,
+      "loss": 4.2547,
+      "step": 8064
+    },
+    {
+      "epoch": 0.8362441594463058,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00016410292656270943,
+      "loss": 4.2837,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8363478475007938,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016409458919541248,
+      "loss": 4.2996,
+      "step": 8066
+    },
+    {
+      "epoch": 0.836451535555282,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00016408625107186713,
+      "loss": 4.2966,
+      "step": 8067
+    },
+    {
+      "epoch": 0.83655522360977,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016407791219217178,
+      "loss": 4.2681,
+      "step": 8068
+    },
+    {
+      "epoch": 0.8366589116642581,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001640695725564248,
+      "loss": 4.2706,
+      "step": 8069
+    },
+    {
+      "epoch": 0.8367625997187461,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016406123216472452,
+      "loss": 4.2618,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8368662877732342,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016405289101716953,
+      "loss": 4.3088,
+      "step": 8071
+    },
+    {
+      "epoch": 0.8369699758277223,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016404454911385805,
+      "loss": 4.2826,
+      "step": 8072
+    },
+    {
+      "epoch": 0.8370736638822104,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016403620645488858,
+      "loss": 4.2745,
+      "step": 8073
+    },
+    {
+      "epoch": 0.8371773519366984,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016402786304035963,
+      "loss": 4.3094,
+      "step": 8074
+    },
+    {
+      "epoch": 0.8372810399911865,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016401951887036952,
+      "loss": 4.2542,
+      "step": 8075
+    },
+    {
+      "epoch": 0.8373847280456745,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001640111739450168,
+      "loss": 4.3313,
+      "step": 8076
+    },
+    {
+      "epoch": 0.8374884161001627,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016400282826439986,
+      "loss": 4.31,
+      "step": 8077
+    },
+    {
+      "epoch": 0.8375921041546507,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001639944818286172,
+      "loss": 4.2806,
+      "step": 8078
+    },
+    {
+      "epoch": 0.8376957922091388,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001639861346377673,
+      "loss": 4.2812,
+      "step": 8079
+    },
+    {
+      "epoch": 0.8377994802636268,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001639777866919487,
+      "loss": 4.2328,
+      "step": 8080
+    },
+    {
+      "epoch": 0.837903168318115,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001639694379912598,
+      "loss": 4.2796,
+      "step": 8081
+    },
+    {
+      "epoch": 0.838006856372603,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016396108853579912,
+      "loss": 4.2733,
+      "step": 8082
+    },
+    {
+      "epoch": 0.8381105444270911,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016395273832566523,
+      "loss": 4.292,
+      "step": 8083
+    },
+    {
+      "epoch": 0.8382142324815792,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00016394438736095667,
+      "loss": 4.304,
+      "step": 8084
+    },
+    {
+      "epoch": 0.8383179205360672,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001639360356417719,
+      "loss": 4.2959,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8384216085905554,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00016392768316820947,
+      "loss": 4.2942,
+      "step": 8086
+    },
+    {
+      "epoch": 0.8385252966450434,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000163919329940368,
+      "loss": 4.2869,
+      "step": 8087
+    },
+    {
+      "epoch": 0.8386289846995315,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00016391097595834598,
+      "loss": 4.2924,
+      "step": 8088
+    },
+    {
+      "epoch": 0.8387326727540195,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016390262122224195,
+      "loss": 4.3011,
+      "step": 8089
+    },
+    {
+      "epoch": 0.8388363608085077,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00016389426573215458,
+      "loss": 4.2705,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8389400488629957,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016388590948818245,
+      "loss": 4.2856,
+      "step": 8091
+    },
+    {
+      "epoch": 0.8390437369174838,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00016387755249042406,
+      "loss": 4.2641,
+      "step": 8092
+    },
+    {
+      "epoch": 0.8391474249719718,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016386919473897812,
+      "loss": 4.3175,
+      "step": 8093
+    },
+    {
+      "epoch": 0.8392511130264599,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016386083623394314,
+      "loss": 4.2739,
+      "step": 8094
+    },
+    {
+      "epoch": 0.839354801080948,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016385247697541785,
+      "loss": 4.2844,
+      "step": 8095
+    },
+    {
+      "epoch": 0.8394584891354361,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016384411696350083,
+      "loss": 4.2292,
+      "step": 8096
+    },
+    {
+      "epoch": 0.8395621771899241,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016383575619829065,
+      "loss": 4.2973,
+      "step": 8097
+    },
+    {
+      "epoch": 0.8396658652444122,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001638273946798861,
+      "loss": 4.2837,
+      "step": 8098
+    },
+    {
+      "epoch": 0.8397695532989002,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001638190324083857,
+      "loss": 4.2605,
+      "step": 8099
+    },
+    {
+      "epoch": 0.8398732413533884,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001638106693838882,
+      "loss": 4.2699,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8399769294078764,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016380230560649226,
+      "loss": 4.2393,
+      "step": 8101
+    },
+    {
+      "epoch": 0.8400806174623645,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016379394107629656,
+      "loss": 4.314,
+      "step": 8102
+    },
+    {
+      "epoch": 0.8401843055168525,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016378557579339978,
+      "loss": 4.2962,
+      "step": 8103
+    },
+    {
+      "epoch": 0.8402879935713407,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016377720975790062,
+      "loss": 4.301,
+      "step": 8104
+    },
+    {
+      "epoch": 0.8403916816258287,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016376884296989783,
+      "loss": 4.2674,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8404953696803168,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016376047542949007,
+      "loss": 4.3346,
+      "step": 8106
+    },
+    {
+      "epoch": 0.8405990577348048,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016375210713677612,
+      "loss": 4.28,
+      "step": 8107
+    },
+    {
+      "epoch": 0.8407027457892929,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001637437380918547,
+      "loss": 4.3053,
+      "step": 8108
+    },
+    {
+      "epoch": 0.840806433843781,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001637353682948245,
+      "loss": 4.3224,
+      "step": 8109
+    },
+    {
+      "epoch": 0.8409101218982691,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016372699774578435,
+      "loss": 4.3216,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8410138099527571,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.000163718626444833,
+      "loss": 4.2668,
+      "step": 8111
+    },
+    {
+      "epoch": 0.8411174980072452,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016371025439206922,
+      "loss": 4.2696,
+      "step": 8112
+    },
+    {
+      "epoch": 0.8412211860617332,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016370188158759176,
+      "loss": 4.2908,
+      "step": 8113
+    },
+    {
+      "epoch": 0.8413248741162214,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016369350803149946,
+      "loss": 4.3051,
+      "step": 8114
+    },
+    {
+      "epoch": 0.8414285621707094,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016368513372389107,
+      "loss": 4.2693,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8415322502251975,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001636767586648654,
+      "loss": 4.3012,
+      "step": 8116
+    },
+    {
+      "epoch": 0.8416359382796855,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016366838285452134,
+      "loss": 4.2654,
+      "step": 8117
+    },
+    {
+      "epoch": 0.8417396263341737,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016366000629295763,
+      "loss": 4.2762,
+      "step": 8118
+    },
+    {
+      "epoch": 0.8418433143886617,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016365162898027315,
+      "loss": 4.2977,
+      "step": 8119
+    },
+    {
+      "epoch": 0.8419470024431498,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001636432509165667,
+      "loss": 4.2502,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8420506904976378,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016363487210193723,
+      "loss": 4.3008,
+      "step": 8121
+    },
+    {
+      "epoch": 0.8421543785521259,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001636264925364835,
+      "loss": 4.2472,
+      "step": 8122
+    },
+    {
+      "epoch": 0.842258066606614,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001636181122203044,
+      "loss": 4.2992,
+      "step": 8123
+    },
+    {
+      "epoch": 0.8423617546611021,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001636097311534988,
+      "loss": 4.2512,
+      "step": 8124
+    },
+    {
+      "epoch": 0.8424654427155901,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016360134933616566,
+      "loss": 4.288,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8425691307700782,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016359296676840378,
+      "loss": 4.3224,
+      "step": 8126
+    },
+    {
+      "epoch": 0.8426728188245662,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016358458345031216,
+      "loss": 4.2729,
+      "step": 8127
+    },
+    {
+      "epoch": 0.8427765068790544,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016357619938198962,
+      "loss": 4.3161,
+      "step": 8128
+    },
+    {
+      "epoch": 0.8428801949335425,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016356781456353518,
+      "loss": 4.3149,
+      "step": 8129
+    },
+    {
+      "epoch": 0.8429838829880305,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001635594289950477,
+      "loss": 4.2801,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8430875710425186,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016355104267662613,
+      "loss": 4.2477,
+      "step": 8131
+    },
+    {
+      "epoch": 0.8431912590970067,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016354265560836946,
+      "loss": 4.2534,
+      "step": 8132
+    },
+    {
+      "epoch": 0.8432949471514948,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001635342677903766,
+      "loss": 4.2756,
+      "step": 8133
+    },
+    {
+      "epoch": 0.8433986352059828,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016352587922274653,
+      "loss": 4.2894,
+      "step": 8134
+    },
+    {
+      "epoch": 0.8435023232604709,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016351748990557825,
+      "loss": 4.2701,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8436060113149589,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016350909983897075,
+      "loss": 4.2833,
+      "step": 8136
+    },
+    {
+      "epoch": 0.8437096993694471,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016350070902302298,
+      "loss": 4.301,
+      "step": 8137
+    },
+    {
+      "epoch": 0.8438133874239351,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016349231745783397,
+      "loss": 4.3139,
+      "step": 8138
+    },
+    {
+      "epoch": 0.8439170754784232,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016348392514350273,
+      "loss": 4.3215,
+      "step": 8139
+    },
+    {
+      "epoch": 0.8440207635329112,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016347553208012828,
+      "loss": 4.2747,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8441244515873993,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016346713826780965,
+      "loss": 4.277,
+      "step": 8141
+    },
+    {
+      "epoch": 0.8442281396418874,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016345874370664583,
+      "loss": 4.2962,
+      "step": 8142
+    },
+    {
+      "epoch": 0.8443318276963755,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016345034839673598,
+      "loss": 4.2753,
+      "step": 8143
+    },
+    {
+      "epoch": 0.8444355157508635,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016344195233817908,
+      "loss": 4.2507,
+      "step": 8144
+    },
+    {
+      "epoch": 0.8445392038053516,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016343355553107415,
+      "loss": 4.2846,
+      "step": 8145
+    },
+    {
+      "epoch": 0.8446428918598397,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001634251579755204,
+      "loss": 4.2818,
+      "step": 8146
+    },
+    {
+      "epoch": 0.8447465799143278,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00016341675967161676,
+      "loss": 4.3059,
+      "step": 8147
+    },
+    {
+      "epoch": 0.8448502679688158,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016340836061946244,
+      "loss": 4.2929,
+      "step": 8148
+    },
+    {
+      "epoch": 0.8449539560233039,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00016339996081915644,
+      "loss": 4.3118,
+      "step": 8149
+    },
+    {
+      "epoch": 0.8450576440777919,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016339156027079797,
+      "loss": 4.2649,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8451613321322801,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016338315897448605,
+      "loss": 4.2986,
+      "step": 8151
+    },
+    {
+      "epoch": 0.8452650201867681,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016337475693031984,
+      "loss": 4.3008,
+      "step": 8152
+    },
+    {
+      "epoch": 0.8453687082412562,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016336635413839853,
+      "loss": 4.2551,
+      "step": 8153
+    },
+    {
+      "epoch": 0.8454723962957442,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016335795059882125,
+      "loss": 4.3054,
+      "step": 8154
+    },
+    {
+      "epoch": 0.8455760843502323,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016334954631168708,
+      "loss": 4.2829,
+      "step": 8155
+    },
+    {
+      "epoch": 0.8456797724047204,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016334114127709523,
+      "loss": 4.2798,
+      "step": 8156
+    },
+    {
+      "epoch": 0.8457834604592085,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016333273549514488,
+      "loss": 4.3232,
+      "step": 8157
+    },
+    {
+      "epoch": 0.8458871485136965,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001633243289659352,
+      "loss": 4.2421,
+      "step": 8158
+    },
+    {
+      "epoch": 0.8459908365681846,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016331592168956533,
+      "loss": 4.2662,
+      "step": 8159
+    },
+    {
+      "epoch": 0.8460945246226727,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016330751366613454,
+      "loss": 4.2964,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8461982126771608,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016329910489574202,
+      "loss": 4.3097,
+      "step": 8161
+    },
+    {
+      "epoch": 0.8463019007316488,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016329069537848696,
+      "loss": 4.249,
+      "step": 8162
+    },
+    {
+      "epoch": 0.8464055887861369,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001632822851144686,
+      "loss": 4.308,
+      "step": 8163
+    },
+    {
+      "epoch": 0.8465092768406249,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016327387410378617,
+      "loss": 4.2608,
+      "step": 8164
+    },
+    {
+      "epoch": 0.8466129648951131,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016326546234653893,
+      "loss": 4.3032,
+      "step": 8165
+    },
+    {
+      "epoch": 0.8467166529496011,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001632570498428261,
+      "loss": 4.2588,
+      "step": 8166
+    },
+    {
+      "epoch": 0.8468203410040892,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016324863659274692,
+      "loss": 4.2746,
+      "step": 8167
+    },
+    {
+      "epoch": 0.8469240290585772,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001632402225964007,
+      "loss": 4.2657,
+      "step": 8168
+    },
+    {
+      "epoch": 0.8470277171130653,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001632318078538867,
+      "loss": 4.2658,
+      "step": 8169
+    },
+    {
+      "epoch": 0.8471314051675534,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001632233923653042,
+      "loss": 4.2676,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8472350932220415,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016321497613075252,
+      "loss": 4.2574,
+      "step": 8171
+    },
+    {
+      "epoch": 0.8473387812765295,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001632065591503309,
+      "loss": 4.2547,
+      "step": 8172
+    },
+    {
+      "epoch": 0.8474424693310176,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016319814142413874,
+      "loss": 4.2806,
+      "step": 8173
+    },
+    {
+      "epoch": 0.8475461573855058,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016318972295227527,
+      "loss": 4.2752,
+      "step": 8174
+    },
+    {
+      "epoch": 0.8476498454399938,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016318130373483994,
+      "loss": 4.2638,
+      "step": 8175
+    },
+    {
+      "epoch": 0.8477535334944819,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016317288377193197,
+      "loss": 4.2541,
+      "step": 8176
+    },
+    {
+      "epoch": 0.8478572215489699,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016316446306365072,
+      "loss": 4.3027,
+      "step": 8177
+    },
+    {
+      "epoch": 0.847960909603458,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001631560416100956,
+      "loss": 4.2608,
+      "step": 8178
+    },
+    {
+      "epoch": 0.8480645976579461,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016314761941136594,
+      "loss": 4.2515,
+      "step": 8179
+    },
+    {
+      "epoch": 0.8481682857124342,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016313919646756113,
+      "loss": 4.2605,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8482719737669222,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016313077277878052,
+      "loss": 4.3121,
+      "step": 8181
+    },
+    {
+      "epoch": 0.8483756618214103,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016312234834512355,
+      "loss": 4.2779,
+      "step": 8182
+    },
+    {
+      "epoch": 0.8484793498758983,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001631139231666896,
+      "loss": 4.2677,
+      "step": 8183
+    },
+    {
+      "epoch": 0.8485830379303865,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016310549724357802,
+      "loss": 4.2662,
+      "step": 8184
+    },
+    {
+      "epoch": 0.8486867259848745,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016309707057588833,
+      "loss": 4.305,
+      "step": 8185
+    },
+    {
+      "epoch": 0.8487904140393626,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016308864316371988,
+      "loss": 4.2841,
+      "step": 8186
+    },
+    {
+      "epoch": 0.8488941020938506,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016308021500717213,
+      "loss": 4.2788,
+      "step": 8187
+    },
+    {
+      "epoch": 0.8489977901483388,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016307178610634453,
+      "loss": 4.307,
+      "step": 8188
+    },
+    {
+      "epoch": 0.8491014782028268,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016306335646133652,
+      "loss": 4.2861,
+      "step": 8189
+    },
+    {
+      "epoch": 0.8492051662573149,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016305492607224755,
+      "loss": 4.2639,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8493088543118029,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016304649493917715,
+      "loss": 4.324,
+      "step": 8191
+    },
+    {
+      "epoch": 0.849412542366291,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016303806306222472,
+      "loss": 4.2783,
+      "step": 8192
+    },
+    {
+      "epoch": 0.8495162304207791,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016302963044148978,
+      "loss": 4.2854,
+      "step": 8193
+    },
+    {
+      "epoch": 0.8496199184752672,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016302119707707182,
+      "loss": 4.2542,
+      "step": 8194
+    },
+    {
+      "epoch": 0.8497236065297552,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016301276296907033,
+      "loss": 4.2758,
+      "step": 8195
+    },
+    {
+      "epoch": 0.8498272945842433,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001630043281175849,
+      "loss": 4.248,
+      "step": 8196
+    },
+    {
+      "epoch": 0.8499309826387313,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016299589252271495,
+      "loss": 4.2591,
+      "step": 8197
+    },
+    {
+      "epoch": 0.8500346706932195,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001629874561845601,
+      "loss": 4.2892,
+      "step": 8198
+    },
+    {
+      "epoch": 0.8501383587477075,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001629790191032198,
+      "loss": 4.3016,
+      "step": 8199
+    },
+    {
+      "epoch": 0.8502420468021956,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016297058127879367,
+      "loss": 4.3234,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8503457348566836,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016296214271138126,
+      "loss": 4.3115,
+      "step": 8201
+    },
+    {
+      "epoch": 0.8504494229111718,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001629537034010821,
+      "loss": 4.2885,
+      "step": 8202
+    },
+    {
+      "epoch": 0.8505531109656598,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016294526334799576,
+      "loss": 4.2767,
+      "step": 8203
+    },
+    {
+      "epoch": 0.8506567990201479,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016293682255222192,
+      "loss": 4.2572,
+      "step": 8204
+    },
+    {
+      "epoch": 0.8507604870746359,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016292838101386004,
+      "loss": 4.2418,
+      "step": 8205
+    },
+    {
+      "epoch": 0.850864175129124,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016291993873300979,
+      "loss": 4.2631,
+      "step": 8206
+    },
+    {
+      "epoch": 0.8509678631836121,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016291149570977077,
+      "loss": 4.2422,
+      "step": 8207
+    },
+    {
+      "epoch": 0.8510715512381002,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016290305194424263,
+      "loss": 4.3361,
+      "step": 8208
+    },
+    {
+      "epoch": 0.8511752392925882,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016289460743652495,
+      "loss": 4.3048,
+      "step": 8209
+    },
+    {
+      "epoch": 0.8512789273470763,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001628861621867174,
+      "loss": 4.2793,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8513826154015643,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016287771619491958,
+      "loss": 4.2686,
+      "step": 8211
+    },
+    {
+      "epoch": 0.8514863034560525,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001628692694612312,
+      "loss": 4.2783,
+      "step": 8212
+    },
+    {
+      "epoch": 0.8515899915105405,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00016286082198575187,
+      "loss": 4.2605,
+      "step": 8213
+    },
+    {
+      "epoch": 0.8516936795650286,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001628523737685813,
+      "loss": 4.3076,
+      "step": 8214
+    },
+    {
+      "epoch": 0.8517973676195166,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016284392480981916,
+      "loss": 4.2711,
+      "step": 8215
+    },
+    {
+      "epoch": 0.8519010556740048,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016283547510956514,
+      "loss": 4.2939,
+      "step": 8216
+    },
+    {
+      "epoch": 0.8520047437284928,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016282702466791893,
+      "loss": 4.2644,
+      "step": 8217
+    },
+    {
+      "epoch": 0.8521084317829809,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016281857348498027,
+      "loss": 4.3043,
+      "step": 8218
+    },
+    {
+      "epoch": 0.852212119837469,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001628101215608488,
+      "loss": 4.2732,
+      "step": 8219
+    },
+    {
+      "epoch": 0.852315807891957,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001628016688956243,
+      "loss": 4.248,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8524194959464452,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001627932154894065,
+      "loss": 4.26,
+      "step": 8221
+    },
+    {
+      "epoch": 0.8525231840009332,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016278476134229514,
+      "loss": 4.2633,
+      "step": 8222
+    },
+    {
+      "epoch": 0.8526268720554213,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00016277630645438991,
+      "loss": 4.2904,
+      "step": 8223
+    },
+    {
+      "epoch": 0.8527305601099093,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001627678508257907,
+      "loss": 4.2674,
+      "step": 8224
+    },
+    {
+      "epoch": 0.8528342481643975,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016275939445659714,
+      "loss": 4.2886,
+      "step": 8225
+    },
+    {
+      "epoch": 0.8529379362188855,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001627509373469091,
+      "loss": 4.2872,
+      "step": 8226
+    },
+    {
+      "epoch": 0.8530416242733736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001627424794968263,
+      "loss": 4.26,
+      "step": 8227
+    },
+    {
+      "epoch": 0.8531453123278616,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00016273402090644855,
+      "loss": 4.2756,
+      "step": 8228
+    },
+    {
+      "epoch": 0.8532490003823497,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016272556157587574,
+      "loss": 4.264,
+      "step": 8229
+    },
+    {
+      "epoch": 0.8533526884368378,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016271710150520753,
+      "loss": 4.2563,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8534563764913259,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016270864069454385,
+      "loss": 4.2632,
+      "step": 8231
+    },
+    {
+      "epoch": 0.8535600645458139,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016270017914398453,
+      "loss": 4.2908,
+      "step": 8232
+    },
+    {
+      "epoch": 0.853663752600302,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016269171685362932,
+      "loss": 4.3089,
+      "step": 8233
+    },
+    {
+      "epoch": 0.85376744065479,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016268325382357815,
+      "loss": 4.3015,
+      "step": 8234
+    },
+    {
+      "epoch": 0.8538711287092782,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001626747900539308,
+      "loss": 4.271,
+      "step": 8235
+    },
+    {
+      "epoch": 0.8539748167637662,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016266632554478723,
+      "loss": 4.2738,
+      "step": 8236
+    },
+    {
+      "epoch": 0.8540785048182543,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00016265786029624724,
+      "loss": 4.2862,
+      "step": 8237
+    },
+    {
+      "epoch": 0.8541821928727423,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016264939430841072,
+      "loss": 4.2891,
+      "step": 8238
+    },
+    {
+      "epoch": 0.8542858809272305,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00016264092758137757,
+      "loss": 4.2754,
+      "step": 8239
+    },
+    {
+      "epoch": 0.8543895689817185,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001626324601152477,
+      "loss": 4.2848,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8544932570362066,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016262399191012102,
+      "loss": 4.2629,
+      "step": 8241
+    },
+    {
+      "epoch": 0.8545969450906946,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016261552296609742,
+      "loss": 4.2785,
+      "step": 8242
+    },
+    {
+      "epoch": 0.8547006331451827,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016260705328327682,
+      "loss": 4.2961,
+      "step": 8243
+    },
+    {
+      "epoch": 0.8548043211996708,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001625985828617592,
+      "loss": 4.2976,
+      "step": 8244
+    },
+    {
+      "epoch": 0.8549080092541589,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016259011170164443,
+      "loss": 4.3077,
+      "step": 8245
+    },
+    {
+      "epoch": 0.8550116973086469,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016258163980303254,
+      "loss": 4.264,
+      "step": 8246
+    },
+    {
+      "epoch": 0.855115385363135,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016257316716602343,
+      "loss": 4.3042,
+      "step": 8247
+    },
+    {
+      "epoch": 0.855219073417623,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001625646937907171,
+      "loss": 4.2716,
+      "step": 8248
+    },
+    {
+      "epoch": 0.8553227614721112,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001625562196772135,
+      "loss": 4.199,
+      "step": 8249
+    },
+    {
+      "epoch": 0.8554264495265992,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016254774482561267,
+      "loss": 4.315,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8555301375810873,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016253926923601454,
+      "loss": 4.2734,
+      "step": 8251
+    },
+    {
+      "epoch": 0.8556338256355753,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016253079290851915,
+      "loss": 4.2881,
+      "step": 8252
+    },
+    {
+      "epoch": 0.8557375136900635,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016252231584322648,
+      "loss": 4.3141,
+      "step": 8253
+    },
+    {
+      "epoch": 0.8558412017445515,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001625138380402366,
+      "loss": 4.3066,
+      "step": 8254
+    },
+    {
+      "epoch": 0.8559448897990396,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016250535949964952,
+      "loss": 4.3038,
+      "step": 8255
+    },
+    {
+      "epoch": 0.8560485778535276,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016249688022156524,
+      "loss": 4.2631,
+      "step": 8256
+    },
+    {
+      "epoch": 0.8561522659080157,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016248840020608384,
+      "loss": 4.2473,
+      "step": 8257
+    },
+    {
+      "epoch": 0.8562559539625038,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001624799194533054,
+      "loss": 4.2545,
+      "step": 8258
+    },
+    {
+      "epoch": 0.8563596420169919,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001624714379633299,
+      "loss": 4.2926,
+      "step": 8259
+    },
+    {
+      "epoch": 0.8564633300714799,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001624629557362575,
+      "loss": 4.2821,
+      "step": 8260
+    },
+    {
+      "epoch": 0.856567018125968,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016245447277218823,
+      "loss": 4.2727,
+      "step": 8261
+    },
+    {
+      "epoch": 0.856670706180456,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016244598907122222,
+      "loss": 4.2537,
+      "step": 8262
+    },
+    {
+      "epoch": 0.8567743942349442,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016243750463345953,
+      "loss": 4.274,
+      "step": 8263
+    },
+    {
+      "epoch": 0.8568780822894323,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016242901945900031,
+      "loss": 4.2731,
+      "step": 8264
+    },
+    {
+      "epoch": 0.8569817703439203,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016242053354794463,
+      "loss": 4.2518,
+      "step": 8265
+    },
+    {
+      "epoch": 0.8570854583984084,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016241204690039266,
+      "loss": 4.2789,
+      "step": 8266
+    },
+    {
+      "epoch": 0.8571891464528965,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016240355951644447,
+      "loss": 4.2943,
+      "step": 8267
+    },
+    {
+      "epoch": 0.8572928345073846,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016239507139620028,
+      "loss": 4.2876,
+      "step": 8268
+    },
+    {
+      "epoch": 0.8573965225618726,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016238658253976018,
+      "loss": 4.2763,
+      "step": 8269
+    },
+    {
+      "epoch": 0.8575002106163607,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016237809294722435,
+      "loss": 4.2659,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8576038986708487,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016236960261869297,
+      "loss": 4.2733,
+      "step": 8271
+    },
+    {
+      "epoch": 0.8577075867253369,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001623611115542662,
+      "loss": 4.2487,
+      "step": 8272
+    },
+    {
+      "epoch": 0.8578112747798249,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016235261975404422,
+      "loss": 4.2883,
+      "step": 8273
+    },
+    {
+      "epoch": 0.857914962834313,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016234412721812726,
+      "loss": 4.3047,
+      "step": 8274
+    },
+    {
+      "epoch": 0.858018650888801,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001623356339466155,
+      "loss": 4.28,
+      "step": 8275
+    },
+    {
+      "epoch": 0.8581223389432892,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016232713993960914,
+      "loss": 4.255,
+      "step": 8276
+    },
+    {
+      "epoch": 0.8582260269977772,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001623186451972084,
+      "loss": 4.2602,
+      "step": 8277
+    },
+    {
+      "epoch": 0.8583297150522653,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016231014971951352,
+      "loss": 4.2697,
+      "step": 8278
+    },
+    {
+      "epoch": 0.8584334031067533,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016230165350662477,
+      "loss": 4.2785,
+      "step": 8279
+    },
+    {
+      "epoch": 0.8585370911612414,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016229315655864234,
+      "loss": 4.3177,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8586407792157295,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016228465887566652,
+      "loss": 4.2407,
+      "step": 8281
+    },
+    {
+      "epoch": 0.8587444672702176,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016227616045779754,
+      "loss": 4.2322,
+      "step": 8282
+    },
+    {
+      "epoch": 0.8588481553247056,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016226766130513574,
+      "loss": 4.2513,
+      "step": 8283
+    },
+    {
+      "epoch": 0.8589518433791937,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016225916141778132,
+      "loss": 4.274,
+      "step": 8284
+    },
+    {
+      "epoch": 0.8590555314336817,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001622506607958346,
+      "loss": 4.2683,
+      "step": 8285
+    },
+    {
+      "epoch": 0.8591592194881699,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016224215943939593,
+      "loss": 4.2584,
+      "step": 8286
+    },
+    {
+      "epoch": 0.8592629075426579,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016223365734856552,
+      "loss": 4.2968,
+      "step": 8287
+    },
+    {
+      "epoch": 0.859366595597146,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016222515452344376,
+      "loss": 4.32,
+      "step": 8288
+    },
+    {
+      "epoch": 0.859470283651634,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016221665096413095,
+      "loss": 4.2641,
+      "step": 8289
+    },
+    {
+      "epoch": 0.8595739717061222,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001622081466707274,
+      "loss": 4.2229,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8596776597606102,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016219964164333351,
+      "loss": 4.2653,
+      "step": 8291
+    },
+    {
+      "epoch": 0.8597813478150983,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001621911358820496,
+      "loss": 4.2972,
+      "step": 8292
+    },
+    {
+      "epoch": 0.8598850358695863,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016218262938697602,
+      "loss": 4.3156,
+      "step": 8293
+    },
+    {
+      "epoch": 0.8599887239240744,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016217412215821313,
+      "loss": 4.29,
+      "step": 8294
+    },
+    {
+      "epoch": 0.8600924119785625,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001621656141958613,
+      "loss": 4.2752,
+      "step": 8295
+    },
+    {
+      "epoch": 0.8601961000330506,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016215710550002098,
+      "loss": 4.2789,
+      "step": 8296
+    },
+    {
+      "epoch": 0.8602997880875386,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001621485960707925,
+      "loss": 4.2865,
+      "step": 8297
+    },
+    {
+      "epoch": 0.8604034761420267,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016214008590827628,
+      "loss": 4.2618,
+      "step": 8298
+    },
+    {
+      "epoch": 0.8605071641965147,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001621315750125727,
+      "loss": 4.2809,
+      "step": 8299
+    },
+    {
+      "epoch": 0.8606108522510029,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016212306338378226,
+      "loss": 4.286,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8607145403054909,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016211455102200533,
+      "loss": 4.2798,
+      "step": 8301
+    },
+    {
+      "epoch": 0.860818228359979,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016210603792734233,
+      "loss": 4.2832,
+      "step": 8302
+    },
+    {
+      "epoch": 0.860921916414467,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016209752409989374,
+      "loss": 4.271,
+      "step": 8303
+    },
+    {
+      "epoch": 0.8610256044689552,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016208900953976004,
+      "loss": 4.2956,
+      "step": 8304
+    },
+    {
+      "epoch": 0.8611292925234432,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016208049424704162,
+      "loss": 4.2594,
+      "step": 8305
+    },
+    {
+      "epoch": 0.8612329805779313,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.000162071978221839,
+      "loss": 4.2507,
+      "step": 8306
+    },
+    {
+      "epoch": 0.8613366686324193,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016206346146425263,
+      "loss": 4.2584,
+      "step": 8307
+    },
+    {
+      "epoch": 0.8614403566869074,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016205494397438303,
+      "loss": 4.2283,
+      "step": 8308
+    },
+    {
+      "epoch": 0.8615440447413956,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016204642575233072,
+      "loss": 4.3046,
+      "step": 8309
+    },
+    {
+      "epoch": 0.8616477327958836,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016203790679819612,
+      "loss": 4.2764,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8617514208503717,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016202938711207984,
+      "loss": 4.2966,
+      "step": 8311
+    },
+    {
+      "epoch": 0.8618551089048597,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016202086669408233,
+      "loss": 4.3013,
+      "step": 8312
+    },
+    {
+      "epoch": 0.8619587969593479,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016201234554430417,
+      "loss": 4.2819,
+      "step": 8313
+    },
+    {
+      "epoch": 0.8620624850138359,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016200382366284588,
+      "loss": 4.2945,
+      "step": 8314
+    },
+    {
+      "epoch": 0.862166173068324,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.000161995301049808,
+      "loss": 4.2843,
+      "step": 8315
+    },
+    {
+      "epoch": 0.862269861122812,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016198677770529112,
+      "loss": 4.2686,
+      "step": 8316
+    },
+    {
+      "epoch": 0.8623735491773001,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016197825362939577,
+      "loss": 4.3057,
+      "step": 8317
+    },
+    {
+      "epoch": 0.8624772372317882,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016196972882222255,
+      "loss": 4.3086,
+      "step": 8318
+    },
+    {
+      "epoch": 0.8625809252862763,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016196120328387204,
+      "loss": 4.262,
+      "step": 8319
+    },
+    {
+      "epoch": 0.8626846133407643,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001619526770144448,
+      "loss": 4.3007,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8627883013952524,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016194415001404147,
+      "loss": 4.2464,
+      "step": 8321
+    },
+    {
+      "epoch": 0.8628919894497404,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016193562228276263,
+      "loss": 4.273,
+      "step": 8322
+    },
+    {
+      "epoch": 0.8629956775042286,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016192709382070896,
+      "loss": 4.2781,
+      "step": 8323
+    },
+    {
+      "epoch": 0.8630993655587166,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.000161918564627981,
+      "loss": 4.2916,
+      "step": 8324
+    },
+    {
+      "epoch": 0.8632030536132047,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016191003470467948,
+      "loss": 4.3182,
+      "step": 8325
+    },
+    {
+      "epoch": 0.8633067416676927,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016190150405090496,
+      "loss": 4.2624,
+      "step": 8326
+    },
+    {
+      "epoch": 0.8634104297221808,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001618929726667581,
+      "loss": 4.2549,
+      "step": 8327
+    },
+    {
+      "epoch": 0.8635141177766689,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016188444055233961,
+      "loss": 4.253,
+      "step": 8328
+    },
+    {
+      "epoch": 0.863617805831157,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016187590770775016,
+      "loss": 4.2926,
+      "step": 8329
+    },
+    {
+      "epoch": 0.863721493885645,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016186737413309037,
+      "loss": 4.2453,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8638251819401331,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016185883982846097,
+      "loss": 4.2331,
+      "step": 8331
+    },
+    {
+      "epoch": 0.8639288699946212,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016185030479396267,
+      "loss": 4.2789,
+      "step": 8332
+    },
+    {
+      "epoch": 0.8640325580491093,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016184176902969615,
+      "loss": 4.2669,
+      "step": 8333
+    },
+    {
+      "epoch": 0.8641362461035973,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001618332325357621,
+      "loss": 4.2263,
+      "step": 8334
+    },
+    {
+      "epoch": 0.8642399341580854,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016182469531226125,
+      "loss": 4.2698,
+      "step": 8335
+    },
+    {
+      "epoch": 0.8643436222125734,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001618161573592944,
+      "loss": 4.314,
+      "step": 8336
+    },
+    {
+      "epoch": 0.8644473102670616,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001618076186769622,
+      "loss": 4.282,
+      "step": 8337
+    },
+    {
+      "epoch": 0.8645509983215496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016179907926536547,
+      "loss": 4.2735,
+      "step": 8338
+    },
+    {
+      "epoch": 0.8646546863760377,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00016179053912460486,
+      "loss": 4.3129,
+      "step": 8339
+    },
+    {
+      "epoch": 0.8647583744305257,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016178199825478125,
+      "loss": 4.3089,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8648620624850138,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016177345665599536,
+      "loss": 4.2849,
+      "step": 8341
+    },
+    {
+      "epoch": 0.8649657505395019,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00016176491432834798,
+      "loss": 4.2581,
+      "step": 8342
+    },
+    {
+      "epoch": 0.86506943859399,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001617563712719399,
+      "loss": 4.2763,
+      "step": 8343
+    },
+    {
+      "epoch": 0.865173126648478,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00016174782748687192,
+      "loss": 4.2793,
+      "step": 8344
+    },
+    {
+      "epoch": 0.8652768147029661,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016173928297324484,
+      "loss": 4.2949,
+      "step": 8345
+    },
+    {
+      "epoch": 0.8653805027574542,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001617307377311595,
+      "loss": 4.2183,
+      "step": 8346
+    },
+    {
+      "epoch": 0.8654841908119423,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001617221917607167,
+      "loss": 4.2548,
+      "step": 8347
+    },
+    {
+      "epoch": 0.8655878788664303,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016171364506201727,
+      "loss": 4.2516,
+      "step": 8348
+    },
+    {
+      "epoch": 0.8656915669209184,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00016170509763516205,
+      "loss": 4.2631,
+      "step": 8349
+    },
+    {
+      "epoch": 0.8657952549754064,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016169654948025193,
+      "loss": 4.2793,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8658989430298946,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016168800059738773,
+      "loss": 4.2804,
+      "step": 8351
+    },
+    {
+      "epoch": 0.8660026310843827,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001616794509866703,
+      "loss": 4.2818,
+      "step": 8352
+    },
+    {
+      "epoch": 0.8661063191388707,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001616709006482006,
+      "loss": 4.232,
+      "step": 8353
+    },
+    {
+      "epoch": 0.8662100071933588,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00016166234958207946,
+      "loss": 4.2797,
+      "step": 8354
+    },
+    {
+      "epoch": 0.8663136952478468,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016165379778840776,
+      "loss": 4.2438,
+      "step": 8355
+    },
+    {
+      "epoch": 0.866417383302335,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016164524526728638,
+      "loss": 4.2721,
+      "step": 8356
+    },
+    {
+      "epoch": 0.866521071356823,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001616366920188163,
+      "loss": 4.3177,
+      "step": 8357
+    },
+    {
+      "epoch": 0.8666247594113111,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016162813804309841,
+      "loss": 4.2867,
+      "step": 8358
+    },
+    {
+      "epoch": 0.8667284474657991,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00016161958334023365,
+      "loss": 4.2791,
+      "step": 8359
+    },
+    {
+      "epoch": 0.8668321355202873,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001616110279103229,
+      "loss": 4.2265,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8669358235747753,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016160247175346716,
+      "loss": 4.282,
+      "step": 8361
+    },
+    {
+      "epoch": 0.8670395116292634,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016159391486976737,
+      "loss": 4.2605,
+      "step": 8362
+    },
+    {
+      "epoch": 0.8671431996837514,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016158535725932453,
+      "loss": 4.2556,
+      "step": 8363
+    },
+    {
+      "epoch": 0.8672468877382395,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016157679892223953,
+      "loss": 4.2746,
+      "step": 8364
+    },
+    {
+      "epoch": 0.8673505757927276,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001615682398586134,
+      "loss": 4.2715,
+      "step": 8365
+    },
+    {
+      "epoch": 0.8674542638472157,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016155968006854713,
+      "loss": 4.3402,
+      "step": 8366
+    },
+    {
+      "epoch": 0.8675579519017037,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016155111955214172,
+      "loss": 4.2307,
+      "step": 8367
+    },
+    {
+      "epoch": 0.8676616399561918,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016154255830949814,
+      "loss": 4.3315,
+      "step": 8368
+    },
+    {
+      "epoch": 0.8677653280106798,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016153399634071742,
+      "loss": 4.3183,
+      "step": 8369
+    },
+    {
+      "epoch": 0.867869016065168,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016152543364590058,
+      "loss": 4.2694,
+      "step": 8370
+    },
+    {
+      "epoch": 0.867972704119656,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001615168702251487,
+      "loss": 4.2861,
+      "step": 8371
+    },
+    {
+      "epoch": 0.8680763921741441,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016150830607856276,
+      "loss": 4.2575,
+      "step": 8372
+    },
+    {
+      "epoch": 0.8681800802286321,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001614997412062438,
+      "loss": 4.2933,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8682837682831203,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001614911756082929,
+      "loss": 4.3062,
+      "step": 8374
+    },
+    {
+      "epoch": 0.8683874563376083,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016148260928481117,
+      "loss": 4.3121,
+      "step": 8375
+    },
+    {
+      "epoch": 0.8684911443920964,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001614740422358996,
+      "loss": 4.3087,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8685948324465844,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016146547446165935,
+      "loss": 4.2649,
+      "step": 8377
+    },
+    {
+      "epoch": 0.8686985205010725,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016145690596219142,
+      "loss": 4.2673,
+      "step": 8378
+    },
+    {
+      "epoch": 0.8688022085555606,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.000161448336737597,
+      "loss": 4.2901,
+      "step": 8379
+    },
+    {
+      "epoch": 0.8689058966100487,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016143976678797717,
+      "loss": 4.2934,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016143119611343302,
+      "loss": 4.2689,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8691132727190248,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016142262471406568,
+      "loss": 4.2696,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8692169607735128,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016141405258997632,
+      "loss": 4.2788,
+      "step": 8383
+    },
+    {
+      "epoch": 0.869320648828001,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016140547974126602,
+      "loss": 4.2743,
+      "step": 8384
+    },
+    {
+      "epoch": 0.869424336882489,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.000161396906168036,
+      "loss": 4.2375,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8695280249369771,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016138833187038737,
+      "loss": 4.2954,
+      "step": 8386
+    },
+    {
+      "epoch": 0.8696317129914651,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016137975684842135,
+      "loss": 4.2602,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8697354010459533,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016137118110223905,
+      "loss": 4.2751,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8698390891004413,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016136260463194167,
+      "loss": 4.2718,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8699427771549294,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016135402743763043,
+      "loss": 4.2387,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8700464652094174,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016134544951940656,
+      "loss": 4.2696,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8701501532639055,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016133687087737118,
+      "loss": 4.2762,
+      "step": 8392
+    },
+    {
+      "epoch": 0.8702538413183936,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016132829151162557,
+      "loss": 4.299,
+      "step": 8393
+    },
+    {
+      "epoch": 0.8703575293728817,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016131971142227097,
+      "loss": 4.2447,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8704612174273697,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016131113060940857,
+      "loss": 4.2158,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8705649054818578,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016130254907313965,
+      "loss": 4.2522,
+      "step": 8396
+    },
+    {
+      "epoch": 0.870668593536346,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016129396681356545,
+      "loss": 4.2729,
+      "step": 8397
+    },
+    {
+      "epoch": 0.870772281590834,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001612853838307872,
+      "loss": 4.303,
+      "step": 8398
+    },
+    {
+      "epoch": 0.8708759696453221,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016127680012490624,
+      "loss": 4.2124,
+      "step": 8399
+    },
+    {
+      "epoch": 0.8709796576998101,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016126821569602381,
+      "loss": 4.2845,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8710833457542982,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001612596305442412,
+      "loss": 4.3315,
+      "step": 8401
+    },
+    {
+      "epoch": 0.8711870338087863,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00016125104466965966,
+      "loss": 4.3241,
+      "step": 8402
+    },
+    {
+      "epoch": 0.8712907218632744,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001612424580723806,
+      "loss": 4.2597,
+      "step": 8403
+    },
+    {
+      "epoch": 0.8713944099177624,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00016123387075250522,
+      "loss": 4.2711,
+      "step": 8404
+    },
+    {
+      "epoch": 0.8714980979722505,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001612252827101349,
+      "loss": 4.3072,
+      "step": 8405
+    },
+    {
+      "epoch": 0.8716017860267385,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016121669394537097,
+      "loss": 4.2542,
+      "step": 8406
+    },
+    {
+      "epoch": 0.8717054740812267,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016120810445831478,
+      "loss": 4.231,
+      "step": 8407
+    },
+    {
+      "epoch": 0.8718091621357147,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00016119951424906764,
+      "loss": 4.3089,
+      "step": 8408
+    },
+    {
+      "epoch": 0.8719128501902028,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016119092331773094,
+      "loss": 4.2775,
+      "step": 8409
+    },
+    {
+      "epoch": 0.8720165382446908,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.000161182331664406,
+      "loss": 4.2568,
+      "step": 8410
+    },
+    {
+      "epoch": 0.872120226299179,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016117373928919423,
+      "loss": 4.248,
+      "step": 8411
+    },
+    {
+      "epoch": 0.872223914353667,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.000161165146192197,
+      "loss": 4.237,
+      "step": 8412
+    },
+    {
+      "epoch": 0.8723276024081551,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016115655237351572,
+      "loss": 4.2702,
+      "step": 8413
+    },
+    {
+      "epoch": 0.8724312904626431,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00016114795783325173,
+      "loss": 4.2676,
+      "step": 8414
+    },
+    {
+      "epoch": 0.8725349785171312,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016113936257150649,
+      "loss": 4.2378,
+      "step": 8415
+    },
+    {
+      "epoch": 0.8726386665716193,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001611307665883814,
+      "loss": 4.3031,
+      "step": 8416
+    },
+    {
+      "epoch": 0.8727423546261074,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00016112216988397788,
+      "loss": 4.276,
+      "step": 8417
+    },
+    {
+      "epoch": 0.8728460426805954,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0001611135724583974,
+      "loss": 4.2516,
+      "step": 8418
+    },
+    {
+      "epoch": 0.8729497307350835,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001611049743117413,
+      "loss": 4.3003,
+      "step": 8419
+    },
+    {
+      "epoch": 0.8730534187895715,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00016109637544411113,
+      "loss": 4.2494,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8731571068440597,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001610877758556083,
+      "loss": 4.3208,
+      "step": 8421
+    },
+    {
+      "epoch": 0.8732607948985477,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00016107917554633429,
+      "loss": 4.2399,
+      "step": 8422
+    },
+    {
+      "epoch": 0.8733644829530358,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016107057451639057,
+      "loss": 4.2707,
+      "step": 8423
+    },
+    {
+      "epoch": 0.8734681710075238,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00016106197276587864,
+      "loss": 4.274,
+      "step": 8424
+    },
+    {
+      "epoch": 0.873571859062012,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016105337029489997,
+      "loss": 4.256,
+      "step": 8425
+    },
+    {
+      "epoch": 0.8736755471165,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00016104476710355608,
+      "loss": 4.2639,
+      "step": 8426
+    },
+    {
+      "epoch": 0.8737792351709881,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001610361631919484,
+      "loss": 4.2571,
+      "step": 8427
+    },
+    {
+      "epoch": 0.8738829232254761,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00016102755856017858,
+      "loss": 4.3151,
+      "step": 8428
+    },
+    {
+      "epoch": 0.8739866112799642,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016101895320834805,
+      "loss": 4.3228,
+      "step": 8429
+    },
+    {
+      "epoch": 0.8740902993344523,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001610103471365584,
+      "loss": 4.2533,
+      "step": 8430
+    },
+    {
+      "epoch": 0.8741939873889404,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001610017403449111,
+      "loss": 4.2683,
+      "step": 8431
+    },
+    {
+      "epoch": 0.8742976754434284,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001609931328335078,
+      "loss": 4.2586,
+      "step": 8432
+    },
+    {
+      "epoch": 0.8744013634979165,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016098452460245,
+      "loss": 4.3005,
+      "step": 8433
+    },
+    {
+      "epoch": 0.8745050515524045,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001609759156518392,
+      "loss": 4.3092,
+      "step": 8434
+    },
+    {
+      "epoch": 0.8746087396068927,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001609673059817771,
+      "loss": 4.2785,
+      "step": 8435
+    },
+    {
+      "epoch": 0.8747124276613807,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00016095869559236526,
+      "loss": 4.2505,
+      "step": 8436
+    },
+    {
+      "epoch": 0.8748161157158688,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001609500844837052,
+      "loss": 4.2825,
+      "step": 8437
+    },
+    {
+      "epoch": 0.8749198037703568,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001609414726558986,
+      "loss": 4.2787,
+      "step": 8438
+    },
+    {
+      "epoch": 0.875023491824845,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016093286010904705,
+      "loss": 4.2771,
+      "step": 8439
+    },
+    {
+      "epoch": 0.875127179879333,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016092424684325214,
+      "loss": 4.2597,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8752308679338211,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016091563285861553,
+      "loss": 4.2585,
+      "step": 8441
+    },
+    {
+      "epoch": 0.8753345559883092,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001609070181552388,
+      "loss": 4.2992,
+      "step": 8442
+    },
+    {
+      "epoch": 0.8754382440427972,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016089840273322367,
+      "loss": 4.262,
+      "step": 8443
+    },
+    {
+      "epoch": 0.8755419320972854,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016088978659267177,
+      "loss": 4.2614,
+      "step": 8444
+    },
+    {
+      "epoch": 0.8756456201517734,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016088116973368477,
+      "loss": 4.2499,
+      "step": 8445
+    },
+    {
+      "epoch": 0.8757493082062615,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00016087255215636428,
+      "loss": 4.2869,
+      "step": 8446
+    },
+    {
+      "epoch": 0.8758529962607495,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000160863933860812,
+      "loss": 4.2898,
+      "step": 8447
+    },
+    {
+      "epoch": 0.8759566843152377,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016085531484712968,
+      "loss": 4.2726,
+      "step": 8448
+    },
+    {
+      "epoch": 0.8760603723697257,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016084669511541897,
+      "loss": 4.2572,
+      "step": 8449
+    },
+    {
+      "epoch": 0.8761640604242138,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016083807466578157,
+      "loss": 4.2457,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8762677484787018,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001608294534983192,
+      "loss": 4.2877,
+      "step": 8451
+    },
+    {
+      "epoch": 0.8763714365331899,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016082083161313355,
+      "loss": 4.3083,
+      "step": 8452
+    },
+    {
+      "epoch": 0.876475124587678,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016081220901032638,
+      "loss": 4.3249,
+      "step": 8453
+    },
+    {
+      "epoch": 0.8765788126421661,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016080358568999948,
+      "loss": 4.2733,
+      "step": 8454
+    },
+    {
+      "epoch": 0.8766825006966541,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016079496165225448,
+      "loss": 4.2688,
+      "step": 8455
+    },
+    {
+      "epoch": 0.8767861887511422,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016078633689719322,
+      "loss": 4.3026,
+      "step": 8456
+    },
+    {
+      "epoch": 0.8768898768056302,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016077771142491745,
+      "loss": 4.2397,
+      "step": 8457
+    },
+    {
+      "epoch": 0.8769935648601184,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016076908523552893,
+      "loss": 4.2921,
+      "step": 8458
+    },
+    {
+      "epoch": 0.8770972529146064,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016076045832912942,
+      "loss": 4.2783,
+      "step": 8459
+    },
+    {
+      "epoch": 0.8772009409690945,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016075183070582072,
+      "loss": 4.2605,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8773046290235825,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016074320236570463,
+      "loss": 4.2831,
+      "step": 8461
+    },
+    {
+      "epoch": 0.8774083170780707,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016073457330888303,
+      "loss": 4.2776,
+      "step": 8462
+    },
+    {
+      "epoch": 0.8775120051325587,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001607259435354576,
+      "loss": 4.2924,
+      "step": 8463
+    },
+    {
+      "epoch": 0.8776156931870468,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016071731304553025,
+      "loss": 4.2821,
+      "step": 8464
+    },
+    {
+      "epoch": 0.8777193812415348,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016070868183920277,
+      "loss": 4.3323,
+      "step": 8465
+    },
+    {
+      "epoch": 0.8778230692960229,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016070004991657703,
+      "loss": 4.2644,
+      "step": 8466
+    },
+    {
+      "epoch": 0.877926757350511,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016069141727775485,
+      "loss": 4.2329,
+      "step": 8467
+    },
+    {
+      "epoch": 0.8780304454049991,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016068278392283812,
+      "loss": 4.2735,
+      "step": 8468
+    },
+    {
+      "epoch": 0.8781341334594871,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001606741498519287,
+      "loss": 4.2749,
+      "step": 8469
+    },
+    {
+      "epoch": 0.8782378215139752,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001606655150651284,
+      "loss": 4.2961,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8783415095684632,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016065687956253916,
+      "loss": 4.2416,
+      "step": 8471
+    },
+    {
+      "epoch": 0.8784451976229514,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016064824334426287,
+      "loss": 4.3016,
+      "step": 8472
+    },
+    {
+      "epoch": 0.8785488856774394,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001606396064104014,
+      "loss": 4.2524,
+      "step": 8473
+    },
+    {
+      "epoch": 0.8786525737319275,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001606309687610567,
+      "loss": 4.2683,
+      "step": 8474
+    },
+    {
+      "epoch": 0.8787562617864155,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016062233039633065,
+      "loss": 4.2676,
+      "step": 8475
+    },
+    {
+      "epoch": 0.8788599498409037,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016061369131632516,
+      "loss": 4.2734,
+      "step": 8476
+    },
+    {
+      "epoch": 0.8789636378953917,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016060505152114222,
+      "loss": 4.277,
+      "step": 8477
+    },
+    {
+      "epoch": 0.8790673259498798,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016059641101088369,
+      "loss": 4.2818,
+      "step": 8478
+    },
+    {
+      "epoch": 0.8791710140043678,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016058776978565163,
+      "loss": 4.3112,
+      "step": 8479
+    },
+    {
+      "epoch": 0.8792747020588559,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016057912784554786,
+      "loss": 4.2279,
+      "step": 8480
+    },
+    {
+      "epoch": 0.879378390113344,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016057048519067446,
+      "loss": 4.2666,
+      "step": 8481
+    },
+    {
+      "epoch": 0.8794820781678321,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016056184182113335,
+      "loss": 4.2834,
+      "step": 8482
+    },
+    {
+      "epoch": 0.8795857662223201,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016055319773702653,
+      "loss": 4.2535,
+      "step": 8483
+    },
+    {
+      "epoch": 0.8796894542768082,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016054455293845597,
+      "loss": 4.2957,
+      "step": 8484
+    },
+    {
+      "epoch": 0.8797931423312962,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016053590742552367,
+      "loss": 4.2143,
+      "step": 8485
+    },
+    {
+      "epoch": 0.8798968303857844,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016052726119833167,
+      "loss": 4.3376,
+      "step": 8486
+    },
+    {
+      "epoch": 0.8800005184402725,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.000160518614256982,
+      "loss": 4.227,
+      "step": 8487
+    },
+    {
+      "epoch": 0.8801042064947605,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001605099666015766,
+      "loss": 4.2465,
+      "step": 8488
+    },
+    {
+      "epoch": 0.8802078945492486,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016050131823221756,
+      "loss": 4.2975,
+      "step": 8489
+    },
+    {
+      "epoch": 0.8803115826037367,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016049266914900694,
+      "loss": 4.2436,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8804152706582248,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016048401935204676,
+      "loss": 4.3068,
+      "step": 8491
+    },
+    {
+      "epoch": 0.8805189587127128,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001604753688414391,
+      "loss": 4.2671,
+      "step": 8492
+    },
+    {
+      "epoch": 0.8806226467672009,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016046671761728597,
+      "loss": 4.2261,
+      "step": 8493
+    },
+    {
+      "epoch": 0.8807263348216889,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016045806567968954,
+      "loss": 4.3031,
+      "step": 8494
+    },
+    {
+      "epoch": 0.8808300228761771,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001604494130287518,
+      "loss": 4.2663,
+      "step": 8495
+    },
+    {
+      "epoch": 0.8809337109306651,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001604407596645749,
+      "loss": 4.2643,
+      "step": 8496
+    },
+    {
+      "epoch": 0.8810373989851532,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016043210558726095,
+      "loss": 4.2479,
+      "step": 8497
+    },
+    {
+      "epoch": 0.8811410870396412,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000160423450796912,
+      "loss": 4.2624,
+      "step": 8498
+    },
+    {
+      "epoch": 0.8812447750941294,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001604147952936302,
+      "loss": 4.2587,
+      "step": 8499
+    },
+    {
+      "epoch": 0.8813484631486174,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016040613907751769,
+      "loss": 4.2823,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8814521512031055,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016039748214867662,
+      "loss": 4.2809,
+      "step": 8501
+    },
+    {
+      "epoch": 0.8815558392575935,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016038882450720906,
+      "loss": 4.2661,
+      "step": 8502
+    },
+    {
+      "epoch": 0.8816595273120816,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001603801661532172,
+      "loss": 4.2395,
+      "step": 8503
+    },
+    {
+      "epoch": 0.8817632153665697,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016037150708680324,
+      "loss": 4.2503,
+      "step": 8504
+    },
+    {
+      "epoch": 0.8818669034210578,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001603628473080693,
+      "loss": 4.3186,
+      "step": 8505
+    },
+    {
+      "epoch": 0.8819705914755458,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016035418681711755,
+      "loss": 4.2744,
+      "step": 8506
+    },
+    {
+      "epoch": 0.8820742795300339,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016034552561405022,
+      "loss": 4.2869,
+      "step": 8507
+    },
+    {
+      "epoch": 0.8821779675845219,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016033686369896946,
+      "loss": 4.2335,
+      "step": 8508
+    },
+    {
+      "epoch": 0.8822816556390101,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001603282010719775,
+      "loss": 4.3039,
+      "step": 8509
+    },
+    {
+      "epoch": 0.8823853436934981,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001603195377331765,
+      "loss": 4.3127,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8824890317479862,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016031087368266876,
+      "loss": 4.2854,
+      "step": 8511
+    },
+    {
+      "epoch": 0.8825927198024742,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016030220892055642,
+      "loss": 4.2955,
+      "step": 8512
+    },
+    {
+      "epoch": 0.8826964078569624,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001602935434469418,
+      "loss": 4.2926,
+      "step": 8513
+    },
+    {
+      "epoch": 0.8828000959114504,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016028487726192707,
+      "loss": 4.3063,
+      "step": 8514
+    },
+    {
+      "epoch": 0.8829037839659385,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016027621036561454,
+      "loss": 4.3108,
+      "step": 8515
+    },
+    {
+      "epoch": 0.8830074720204265,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001602675427581064,
+      "loss": 4.2623,
+      "step": 8516
+    },
+    {
+      "epoch": 0.8831111600749146,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.000160258874439505,
+      "loss": 4.2864,
+      "step": 8517
+    },
+    {
+      "epoch": 0.8832148481294027,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016025020540991257,
+      "loss": 4.2584,
+      "step": 8518
+    },
+    {
+      "epoch": 0.8833185361838908,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016024153566943138,
+      "loss": 4.2823,
+      "step": 8519
+    },
+    {
+      "epoch": 0.8834222242383788,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016023286521816375,
+      "loss": 4.2925,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8835259122928669,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.000160224194056212,
+      "loss": 4.3236,
+      "step": 8521
+    },
+    {
+      "epoch": 0.8836296003473549,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016021552218367844,
+      "loss": 4.3257,
+      "step": 8522
+    },
+    {
+      "epoch": 0.8837332884018431,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016020684960066532,
+      "loss": 4.2686,
+      "step": 8523
+    },
+    {
+      "epoch": 0.8838369764563311,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016019817630727505,
+      "loss": 4.2799,
+      "step": 8524
+    },
+    {
+      "epoch": 0.8839406645108192,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016018950230360993,
+      "loss": 4.3084,
+      "step": 8525
+    },
+    {
+      "epoch": 0.8840443525653072,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016018082758977227,
+      "loss": 4.3094,
+      "step": 8526
+    },
+    {
+      "epoch": 0.8841480406197954,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016017215216586447,
+      "loss": 4.262,
+      "step": 8527
+    },
+    {
+      "epoch": 0.8842517286742834,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016016347603198892,
+      "loss": 4.3118,
+      "step": 8528
+    },
+    {
+      "epoch": 0.8843554167287715,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001601547991882479,
+      "loss": 4.2696,
+      "step": 8529
+    },
+    {
+      "epoch": 0.8844591047832595,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016014612163474387,
+      "loss": 4.2951,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8845627928377476,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00016013744337157917,
+      "loss": 4.2564,
+      "step": 8531
+    },
+    {
+      "epoch": 0.8846664808922358,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001601287643988562,
+      "loss": 4.2667,
+      "step": 8532
+    },
+    {
+      "epoch": 0.8847701689467238,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001601200847166774,
+      "loss": 4.2845,
+      "step": 8533
+    },
+    {
+      "epoch": 0.8848738570012119,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016011140432514511,
+      "loss": 4.271,
+      "step": 8534
+    },
+    {
+      "epoch": 0.8849775450556999,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001601027232243618,
+      "loss": 4.2858,
+      "step": 8535
+    },
+    {
+      "epoch": 0.885081233110188,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016009404141442988,
+      "loss": 4.2776,
+      "step": 8536
+    },
+    {
+      "epoch": 0.8851849211646761,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016008535889545182,
+      "loss": 4.2547,
+      "step": 8537
+    },
+    {
+      "epoch": 0.8852886092191642,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016007667566753,
+      "loss": 4.2769,
+      "step": 8538
+    },
+    {
+      "epoch": 0.8853922972736522,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016006799173076694,
+      "loss": 4.2438,
+      "step": 8539
+    },
+    {
+      "epoch": 0.8854959853281403,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016005930708526506,
+      "loss": 4.2793,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8855996733826283,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016005062173112688,
+      "loss": 4.243,
+      "step": 8541
+    },
+    {
+      "epoch": 0.8857033614371165,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016004193566845478,
+      "loss": 4.2848,
+      "step": 8542
+    },
+    {
+      "epoch": 0.8858070494916045,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016003324889735134,
+      "loss": 4.2932,
+      "step": 8543
+    },
+    {
+      "epoch": 0.8859107375460926,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016002456141791903,
+      "loss": 4.2527,
+      "step": 8544
+    },
+    {
+      "epoch": 0.8860144256005806,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001600158732302603,
+      "loss": 4.2389,
+      "step": 8545
+    },
+    {
+      "epoch": 0.8861181136550688,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016000718433447774,
+      "loss": 4.2734,
+      "step": 8546
+    },
+    {
+      "epoch": 0.8862218017095568,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015999849473067386,
+      "loss": 4.236,
+      "step": 8547
+    },
+    {
+      "epoch": 0.8863254897640449,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001599898044189511,
+      "loss": 4.2682,
+      "step": 8548
+    },
+    {
+      "epoch": 0.8864291778185329,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001599811133994121,
+      "loss": 4.2913,
+      "step": 8549
+    },
+    {
+      "epoch": 0.886532865873021,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015997242167215935,
+      "loss": 4.2738,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8866365539275091,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015996372923729544,
+      "loss": 4.2653,
+      "step": 8551
+    },
+    {
+      "epoch": 0.8867402419819972,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015995503609492288,
+      "loss": 4.2559,
+      "step": 8552
+    },
+    {
+      "epoch": 0.8868439300364852,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015994634224514428,
+      "loss": 4.2434,
+      "step": 8553
+    },
+    {
+      "epoch": 0.8869476180909733,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015993764768806222,
+      "loss": 4.2631,
+      "step": 8554
+    },
+    {
+      "epoch": 0.8870513061454613,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015992895242377927,
+      "loss": 4.2876,
+      "step": 8555
+    },
+    {
+      "epoch": 0.8871549941999495,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.000159920256452398,
+      "loss": 4.2772,
+      "step": 8556
+    },
+    {
+      "epoch": 0.8872586822544375,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015991155977402108,
+      "loss": 4.29,
+      "step": 8557
+    },
+    {
+      "epoch": 0.8873623703089256,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015990286238875108,
+      "loss": 4.2843,
+      "step": 8558
+    },
+    {
+      "epoch": 0.8874660583634136,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015989416429669062,
+      "loss": 4.3156,
+      "step": 8559
+    },
+    {
+      "epoch": 0.8875697464179018,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001598854654979423,
+      "loss": 4.2318,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8876734344723898,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015987676599260882,
+      "loss": 4.2924,
+      "step": 8561
+    },
+    {
+      "epoch": 0.8877771225268779,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001598680657807928,
+      "loss": 4.2523,
+      "step": 8562
+    },
+    {
+      "epoch": 0.8878808105813659,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00015985936486259688,
+      "loss": 4.253,
+      "step": 8563
+    },
+    {
+      "epoch": 0.887984498635854,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015985066323812372,
+      "loss": 4.2691,
+      "step": 8564
+    },
+    {
+      "epoch": 0.8880881866903421,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.000159841960907476,
+      "loss": 4.3306,
+      "step": 8565
+    },
+    {
+      "epoch": 0.8881918747448302,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015983325787075642,
+      "loss": 4.2717,
+      "step": 8566
+    },
+    {
+      "epoch": 0.8882955627993182,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015982455412806763,
+      "loss": 4.271,
+      "step": 8567
+    },
+    {
+      "epoch": 0.8883992508538063,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001598158496795123,
+      "loss": 4.2797,
+      "step": 8568
+    },
+    {
+      "epoch": 0.8885029389082943,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015980714452519323,
+      "loss": 4.2774,
+      "step": 8569
+    },
+    {
+      "epoch": 0.8886066269627825,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00015979843866521302,
+      "loss": 4.2678,
+      "step": 8570
+    },
+    {
+      "epoch": 0.8887103150172705,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001597897320996745,
+      "loss": 4.2739,
+      "step": 8571
+    },
+    {
+      "epoch": 0.8888140030717586,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00015978102482868032,
+      "loss": 4.2616,
+      "step": 8572
+    },
+    {
+      "epoch": 0.8889176911262466,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001597723168523332,
+      "loss": 4.2571,
+      "step": 8573
+    },
+    {
+      "epoch": 0.8890213791807348,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015976360817073596,
+      "loss": 4.3135,
+      "step": 8574
+    },
+    {
+      "epoch": 0.8891250672352228,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001597548987839913,
+      "loss": 4.3112,
+      "step": 8575
+    },
+    {
+      "epoch": 0.8892287552897109,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015974618869220203,
+      "loss": 4.273,
+      "step": 8576
+    },
+    {
+      "epoch": 0.889332443344199,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015973747789547086,
+      "loss": 4.2647,
+      "step": 8577
+    },
+    {
+      "epoch": 0.889436131398687,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015972876639390058,
+      "loss": 4.2989,
+      "step": 8578
+    },
+    {
+      "epoch": 0.8895398194531752,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015972005418759405,
+      "loss": 4.256,
+      "step": 8579
+    },
+    {
+      "epoch": 0.8896435075076632,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015971134127665395,
+      "loss": 4.3002,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8897471955621513,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015970262766118318,
+      "loss": 4.2529,
+      "step": 8581
+    },
+    {
+      "epoch": 0.8898508836166393,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015969391334128447,
+      "loss": 4.2892,
+      "step": 8582
+    },
+    {
+      "epoch": 0.8899545716711275,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015968519831706073,
+      "loss": 4.2553,
+      "step": 8583
+    },
+    {
+      "epoch": 0.8900582597256155,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015967648258861472,
+      "loss": 4.2492,
+      "step": 8584
+    },
+    {
+      "epoch": 0.8901619477801036,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015966776615604928,
+      "loss": 4.2728,
+      "step": 8585
+    },
+    {
+      "epoch": 0.8902656358345916,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001596590490194673,
+      "loss": 4.2634,
+      "step": 8586
+    },
+    {
+      "epoch": 0.8903693238890797,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001596503311789716,
+      "loss": 4.2454,
+      "step": 8587
+    },
+    {
+      "epoch": 0.8904730119435678,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015964161263466503,
+      "loss": 4.3347,
+      "step": 8588
+    },
+    {
+      "epoch": 0.8905766999980559,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015963289338665046,
+      "loss": 4.294,
+      "step": 8589
+    },
+    {
+      "epoch": 0.8906803880525439,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001596241734350308,
+      "loss": 4.3227,
+      "step": 8590
+    },
+    {
+      "epoch": 0.890784076107032,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001596154527799089,
+      "loss": 4.2758,
+      "step": 8591
+    },
+    {
+      "epoch": 0.89088776416152,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015960673142138775,
+      "loss": 4.3,
+      "step": 8592
+    },
+    {
+      "epoch": 0.8909914522160082,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001595980093595701,
+      "loss": 4.278,
+      "step": 8593
+    },
+    {
+      "epoch": 0.8910951402704962,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015958928659455895,
+      "loss": 4.2745,
+      "step": 8594
+    },
+    {
+      "epoch": 0.8911988283249843,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001595805631264572,
+      "loss": 4.2686,
+      "step": 8595
+    },
+    {
+      "epoch": 0.8913025163794723,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001595718389553678,
+      "loss": 4.2547,
+      "step": 8596
+    },
+    {
+      "epoch": 0.8914062044339605,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015956311408139365,
+      "loss": 4.2825,
+      "step": 8597
+    },
+    {
+      "epoch": 0.8915098924884485,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015955438850463776,
+      "loss": 4.2491,
+      "step": 8598
+    },
+    {
+      "epoch": 0.8916135805429366,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.000159545662225203,
+      "loss": 4.2302,
+      "step": 8599
+    },
+    {
+      "epoch": 0.8917172685974246,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001595369352431924,
+      "loss": 4.2561,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8918209566519127,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015952820755870886,
+      "loss": 4.2302,
+      "step": 8601
+    },
+    {
+      "epoch": 0.8919246447064008,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001595194791718554,
+      "loss": 4.2646,
+      "step": 8602
+    },
+    {
+      "epoch": 0.8920283327608889,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015951075008273504,
+      "loss": 4.2478,
+      "step": 8603
+    },
+    {
+      "epoch": 0.8921320208153769,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015950202029145068,
+      "loss": 4.305,
+      "step": 8604
+    },
+    {
+      "epoch": 0.892235708869865,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001594932897981054,
+      "loss": 4.3149,
+      "step": 8605
+    },
+    {
+      "epoch": 0.892339396924353,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001594845586028022,
+      "loss": 4.2881,
+      "step": 8606
+    },
+    {
+      "epoch": 0.8924430849788412,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015947582670564406,
+      "loss": 4.222,
+      "step": 8607
+    },
+    {
+      "epoch": 0.8925467730333292,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001594670941067341,
+      "loss": 4.2469,
+      "step": 8608
+    },
+    {
+      "epoch": 0.8926504610878173,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001594583608061752,
+      "loss": 4.244,
+      "step": 8609
+    },
+    {
+      "epoch": 0.8927541491423053,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015944962680407057,
+      "loss": 4.2382,
+      "step": 8610
+    },
+    {
+      "epoch": 0.8928578371967935,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015944089210052312,
+      "loss": 4.2642,
+      "step": 8611
+    },
+    {
+      "epoch": 0.8929615252512815,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000159432156695636,
+      "loss": 4.2569,
+      "step": 8612
+    },
+    {
+      "epoch": 0.8930652133057696,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015942342058951225,
+      "loss": 4.2415,
+      "step": 8613
+    },
+    {
+      "epoch": 0.8931689013602576,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015941468378225497,
+      "loss": 4.2533,
+      "step": 8614
+    },
+    {
+      "epoch": 0.8932725894147457,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015940594627396722,
+      "loss": 4.2619,
+      "step": 8615
+    },
+    {
+      "epoch": 0.8933762774692338,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001593972080647521,
+      "loss": 4.2925,
+      "step": 8616
+    },
+    {
+      "epoch": 0.8934799655237219,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001593884691547127,
+      "loss": 4.2796,
+      "step": 8617
+    },
+    {
+      "epoch": 0.8935836535782099,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015937972954395213,
+      "loss": 4.3201,
+      "step": 8618
+    },
+    {
+      "epoch": 0.893687341632698,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015937098923257353,
+      "loss": 4.2301,
+      "step": 8619
+    },
+    {
+      "epoch": 0.8937910296871862,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015936224822068,
+      "loss": 4.3146,
+      "step": 8620
+    },
+    {
+      "epoch": 0.8938947177416742,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001593535065083747,
+      "loss": 4.2895,
+      "step": 8621
+    },
+    {
+      "epoch": 0.8939984057961623,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015934476409576077,
+      "loss": 4.2772,
+      "step": 8622
+    },
+    {
+      "epoch": 0.8941020938506503,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015933602098294137,
+      "loss": 4.2764,
+      "step": 8623
+    },
+    {
+      "epoch": 0.8942057819051384,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015932727717001962,
+      "loss": 4.2412,
+      "step": 8624
+    },
+    {
+      "epoch": 0.8943094699596265,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001593185326570987,
+      "loss": 4.2683,
+      "step": 8625
+    },
+    {
+      "epoch": 0.8944131580141146,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015930978744428182,
+      "loss": 4.2657,
+      "step": 8626
+    },
+    {
+      "epoch": 0.8945168460686026,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015930104153167213,
+      "loss": 4.2425,
+      "step": 8627
+    },
+    {
+      "epoch": 0.8946205341230907,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015929229491937286,
+      "loss": 4.2625,
+      "step": 8628
+    },
+    {
+      "epoch": 0.8947242221775787,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015928354760748716,
+      "loss": 4.2919,
+      "step": 8629
+    },
+    {
+      "epoch": 0.8948279102320669,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015927479959611826,
+      "loss": 4.2587,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8949315982865549,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015926605088536938,
+      "loss": 4.2266,
+      "step": 8631
+    },
+    {
+      "epoch": 0.895035286341043,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015925730147534374,
+      "loss": 4.301,
+      "step": 8632
+    },
+    {
+      "epoch": 0.895138974395531,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001592485513661446,
+      "loss": 4.289,
+      "step": 8633
+    },
+    {
+      "epoch": 0.8952426624500192,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015923980055787518,
+      "loss": 4.2844,
+      "step": 8634
+    },
+    {
+      "epoch": 0.8953463505045072,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001592310490506387,
+      "loss": 4.2731,
+      "step": 8635
+    },
+    {
+      "epoch": 0.8954500385589953,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001592222968445385,
+      "loss": 4.2905,
+      "step": 8636
+    },
+    {
+      "epoch": 0.8955537266134833,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015921354393967779,
+      "loss": 4.3063,
+      "step": 8637
+    },
+    {
+      "epoch": 0.8956574146679714,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015920479033615983,
+      "loss": 4.2803,
+      "step": 8638
+    },
+    {
+      "epoch": 0.8957611027224595,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015919603603408792,
+      "loss": 4.3242,
+      "step": 8639
+    },
+    {
+      "epoch": 0.8958647907769476,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015918728103356538,
+      "loss": 4.2296,
+      "step": 8640
+    },
+    {
+      "epoch": 0.8959684788314356,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015917852533469548,
+      "loss": 4.3023,
+      "step": 8641
+    },
+    {
+      "epoch": 0.8960721668859237,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015916976893758154,
+      "loss": 4.2442,
+      "step": 8642
+    },
+    {
+      "epoch": 0.8961758549404117,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015916101184232687,
+      "loss": 4.278,
+      "step": 8643
+    },
+    {
+      "epoch": 0.8962795429948999,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001591522540490348,
+      "loss": 4.2436,
+      "step": 8644
+    },
+    {
+      "epoch": 0.8963832310493879,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015914349555780865,
+      "loss": 4.2421,
+      "step": 8645
+    },
+    {
+      "epoch": 0.896486919103876,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001591347363687518,
+      "loss": 4.3135,
+      "step": 8646
+    },
+    {
+      "epoch": 0.896590607158364,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015912597648196753,
+      "loss": 4.2696,
+      "step": 8647
+    },
+    {
+      "epoch": 0.8966942952128522,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001591172158975592,
+      "loss": 4.2535,
+      "step": 8648
+    },
+    {
+      "epoch": 0.8967979832673402,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015910845461563028,
+      "loss": 4.2705,
+      "step": 8649
+    },
+    {
+      "epoch": 0.8969016713218283,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015909969263628408,
+      "loss": 4.2643,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8970053593763163,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015909092995962393,
+      "loss": 4.2681,
+      "step": 8651
+    },
+    {
+      "epoch": 0.8971090474308044,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001590821665857533,
+      "loss": 4.302,
+      "step": 8652
+    },
+    {
+      "epoch": 0.8972127354852925,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015907340251477558,
+      "loss": 4.2141,
+      "step": 8653
+    },
+    {
+      "epoch": 0.8973164235397806,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001590646377467941,
+      "loss": 4.324,
+      "step": 8654
+    },
+    {
+      "epoch": 0.8974201115942686,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015905587228191235,
+      "loss": 4.2819,
+      "step": 8655
+    },
+    {
+      "epoch": 0.8975237996487567,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015904710612023372,
+      "loss": 4.284,
+      "step": 8656
+    },
+    {
+      "epoch": 0.8976274877032447,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015903833926186168,
+      "loss": 4.228,
+      "step": 8657
+    },
+    {
+      "epoch": 0.8977311757577329,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015902957170689966,
+      "loss": 4.2526,
+      "step": 8658
+    },
+    {
+      "epoch": 0.8978348638122209,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015902080345545102,
+      "loss": 4.2731,
+      "step": 8659
+    },
+    {
+      "epoch": 0.897938551866709,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015901203450761932,
+      "loss": 4.303,
+      "step": 8660
+    },
+    {
+      "epoch": 0.898042239921197,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.000159003264863508,
+      "loss": 4.3071,
+      "step": 8661
+    },
+    {
+      "epoch": 0.8981459279756852,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015899449452322055,
+      "loss": 4.2771,
+      "step": 8662
+    },
+    {
+      "epoch": 0.8982496160301732,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015898572348686037,
+      "loss": 4.3085,
+      "step": 8663
+    },
+    {
+      "epoch": 0.8983533040846613,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015897695175453105,
+      "loss": 4.2636,
+      "step": 8664
+    },
+    {
+      "epoch": 0.8984569921391494,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.000158968179326336,
+      "loss": 4.2446,
+      "step": 8665
+    },
+    {
+      "epoch": 0.8985606801936374,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001589594062023788,
+      "loss": 4.2981,
+      "step": 8666
+    },
+    {
+      "epoch": 0.8986643682481256,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015895063238276292,
+      "loss": 4.2697,
+      "step": 8667
+    },
+    {
+      "epoch": 0.8987680563026136,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015894185786759189,
+      "loss": 4.2807,
+      "step": 8668
+    },
+    {
+      "epoch": 0.8988717443571017,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015893308265696923,
+      "loss": 4.2924,
+      "step": 8669
+    },
+    {
+      "epoch": 0.8989754324115897,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001589243067509985,
+      "loss": 4.2421,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8990791204660779,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015891553014978324,
+      "loss": 4.2539,
+      "step": 8671
+    },
+    {
+      "epoch": 0.8991828085205659,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.000158906752853427,
+      "loss": 4.2523,
+      "step": 8672
+    },
+    {
+      "epoch": 0.899286496575054,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015889797486203332,
+      "loss": 4.2962,
+      "step": 8673
+    },
+    {
+      "epoch": 0.899390184629542,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015888919617570584,
+      "loss": 4.2683,
+      "step": 8674
+    },
+    {
+      "epoch": 0.8994938726840301,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015888041679454806,
+      "loss": 4.2513,
+      "step": 8675
+    },
+    {
+      "epoch": 0.8995975607385182,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001588716367186636,
+      "loss": 4.2272,
+      "step": 8676
+    },
+    {
+      "epoch": 0.8997012487930063,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015886285594815606,
+      "loss": 4.2718,
+      "step": 8677
+    },
+    {
+      "epoch": 0.8998049368474943,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015885407448312905,
+      "loss": 4.2413,
+      "step": 8678
+    },
+    {
+      "epoch": 0.8999086249019824,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001588452923236862,
+      "loss": 4.304,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9000123129564704,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015883650946993104,
+      "loss": 4.2845,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9001160010109586,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001588277259219673,
+      "loss": 4.2831,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9002196890654466,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015881894167989855,
+      "loss": 4.2684,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9003233771199347,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001588101567438285,
+      "loss": 4.2956,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9004270651744227,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015880137111386075,
+      "loss": 4.2602,
+      "step": 8684
+    },
+    {
+      "epoch": 0.9005307532289109,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015879258479009896,
+      "loss": 4.257,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9006344412833989,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001587837977726468,
+      "loss": 4.2761,
+      "step": 8686
+    },
+    {
+      "epoch": 0.900738129337887,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015877501006160804,
+      "loss": 4.2571,
+      "step": 8687
+    },
+    {
+      "epoch": 0.900841817392375,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015876622165708623,
+      "loss": 4.2385,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9009455054468631,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015875743255918512,
+      "loss": 4.2775,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9010491935013512,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015874864276800845,
+      "loss": 4.27,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9011528815558393,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015873985228365988,
+      "loss": 4.2789,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9012565696103273,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015873106110624315,
+      "loss": 4.3142,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9013602576648154,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015872226923586196,
+      "loss": 4.2239,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9014639457193034,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015871347667262007,
+      "loss": 4.2556,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9015676337737916,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015870468341662124,
+      "loss": 4.3006,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9016713218282796,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015869588946796918,
+      "loss": 4.2797,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9017750098827677,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00015868709482676766,
+      "loss": 4.3279,
+      "step": 8697
+    },
+    {
+      "epoch": 0.9018786979372557,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015867829949312045,
+      "loss": 4.3145,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9019823859917439,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015866950346713136,
+      "loss": 4.3113,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9020860740462319,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015866070674890414,
+      "loss": 4.2536,
+      "step": 8700
+    },
+    {
+      "epoch": 0.90218976210072,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015865190933854259,
+      "loss": 4.2619,
+      "step": 8701
+    },
+    {
+      "epoch": 0.902293450155208,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001586431112361505,
+      "loss": 4.2475,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9023971382096961,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015863431244183168,
+      "loss": 4.2817,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9025008262641842,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015862551295568996,
+      "loss": 4.2685,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9026045143186723,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015861671277782918,
+      "loss": 4.2415,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9027082023731603,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001586079119083531,
+      "loss": 4.2506,
+      "step": 8706
+    },
+    {
+      "epoch": 0.9028118904276484,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015859911034736562,
+      "loss": 4.2706,
+      "step": 8707
+    },
+    {
+      "epoch": 0.9029155784821364,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001585903080949706,
+      "loss": 4.2659,
+      "step": 8708
+    },
+    {
+      "epoch": 0.9030192665366246,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001585815051512719,
+      "loss": 4.2601,
+      "step": 8709
+    },
+    {
+      "epoch": 0.9031229545911127,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001585727015163733,
+      "loss": 4.2822,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9032266426456007,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015856389719037876,
+      "loss": 4.2895,
+      "step": 8711
+    },
+    {
+      "epoch": 0.9033303307000888,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015855509217339219,
+      "loss": 4.2873,
+      "step": 8712
+    },
+    {
+      "epoch": 0.9034340187545769,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015854628646551737,
+      "loss": 4.2613,
+      "step": 8713
+    },
+    {
+      "epoch": 0.903537706809065,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001585374800668583,
+      "loss": 4.2723,
+      "step": 8714
+    },
+    {
+      "epoch": 0.903641394863553,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015852867297751877,
+      "loss": 4.2495,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9037450829180411,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015851986519760286,
+      "loss": 4.2634,
+      "step": 8716
+    },
+    {
+      "epoch": 0.9038487709725291,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015851105672721433,
+      "loss": 4.2665,
+      "step": 8717
+    },
+    {
+      "epoch": 0.9039524590270173,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015850224756645717,
+      "loss": 4.3059,
+      "step": 8718
+    },
+    {
+      "epoch": 0.9040561470815053,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015849343771543538,
+      "loss": 4.2917,
+      "step": 8719
+    },
+    {
+      "epoch": 0.9041598351359934,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015848462717425282,
+      "loss": 4.2838,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9042635231904814,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001584758159430135,
+      "loss": 4.2943,
+      "step": 8721
+    },
+    {
+      "epoch": 0.9043672112449695,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015846700402182134,
+      "loss": 4.267,
+      "step": 8722
+    },
+    {
+      "epoch": 0.9044708992994576,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015845819141078037,
+      "loss": 4.2886,
+      "step": 8723
+    },
+    {
+      "epoch": 0.9045745873539457,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015844937810999453,
+      "loss": 4.2663,
+      "step": 8724
+    },
+    {
+      "epoch": 0.9046782754084337,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015844056411956777,
+      "loss": 4.2901,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9047819634629218,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015843174943960418,
+      "loss": 4.3164,
+      "step": 8726
+    },
+    {
+      "epoch": 0.9048856515174099,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015842293407020765,
+      "loss": 4.2537,
+      "step": 8727
+    },
+    {
+      "epoch": 0.904989339571898,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015841411801148232,
+      "loss": 4.27,
+      "step": 8728
+    },
+    {
+      "epoch": 0.905093027626386,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015840530126353213,
+      "loss": 4.2543,
+      "step": 8729
+    },
+    {
+      "epoch": 0.9051967156808741,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001583964838264611,
+      "loss": 4.2713,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9053004037353621,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015838766570037326,
+      "loss": 4.2698,
+      "step": 8731
+    },
+    {
+      "epoch": 0.9054040917898503,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001583788468853727,
+      "loss": 4.2757,
+      "step": 8732
+    },
+    {
+      "epoch": 0.9055077798443383,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001583700273815635,
+      "loss": 4.291,
+      "step": 8733
+    },
+    {
+      "epoch": 0.9056114678988264,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015836120718904967,
+      "loss": 4.2981,
+      "step": 8734
+    },
+    {
+      "epoch": 0.9057151559533144,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015835238630793524,
+      "loss": 4.3023,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9058188440078025,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015834356473832438,
+      "loss": 4.2939,
+      "step": 8736
+    },
+    {
+      "epoch": 0.9059225320622906,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015833474248032111,
+      "loss": 4.2536,
+      "step": 8737
+    },
+    {
+      "epoch": 0.9060262201167787,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001583259195340295,
+      "loss": 4.2877,
+      "step": 8738
+    },
+    {
+      "epoch": 0.9061299081712667,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015831709589955376,
+      "loss": 4.2692,
+      "step": 8739
+    },
+    {
+      "epoch": 0.9062335962257548,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001583082715769979,
+      "loss": 4.2686,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9063372842802428,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015829944656646608,
+      "loss": 4.2784,
+      "step": 8741
+    },
+    {
+      "epoch": 0.906440972334731,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001582906208680624,
+      "loss": 4.2617,
+      "step": 8742
+    },
+    {
+      "epoch": 0.906544660389219,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015828179448189103,
+      "loss": 4.2634,
+      "step": 8743
+    },
+    {
+      "epoch": 0.9066483484437071,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015827296740805608,
+      "loss": 4.2801,
+      "step": 8744
+    },
+    {
+      "epoch": 0.9067520364981951,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015826413964666172,
+      "loss": 4.2969,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9068557245526833,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001582553111978121,
+      "loss": 4.2976,
+      "step": 8746
+    },
+    {
+      "epoch": 0.9069594126071713,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015824648206161138,
+      "loss": 4.2384,
+      "step": 8747
+    },
+    {
+      "epoch": 0.9070631006616594,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015823765223816372,
+      "loss": 4.2754,
+      "step": 8748
+    },
+    {
+      "epoch": 0.9071667887161474,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015822882172757333,
+      "loss": 4.3017,
+      "step": 8749
+    },
+    {
+      "epoch": 0.9072704767706355,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015821999052994441,
+      "loss": 4.2288,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9073741648251236,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015821115864538113,
+      "loss": 4.2528,
+      "step": 8751
+    },
+    {
+      "epoch": 0.9074778528796117,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015820232607398772,
+      "loss": 4.2473,
+      "step": 8752
+    },
+    {
+      "epoch": 0.9075815409340997,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001581934928158684,
+      "loss": 4.2968,
+      "step": 8753
+    },
+    {
+      "epoch": 0.9076852289885878,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015818465887112738,
+      "loss": 4.24,
+      "step": 8754
+    },
+    {
+      "epoch": 0.907788917043076,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015817582423986886,
+      "loss": 4.2909,
+      "step": 8755
+    },
+    {
+      "epoch": 0.907892605097564,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001581669889221971,
+      "loss": 4.2621,
+      "step": 8756
+    },
+    {
+      "epoch": 0.9079962931520521,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015815815291821638,
+      "loss": 4.3012,
+      "step": 8757
+    },
+    {
+      "epoch": 0.9080999812065401,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015814931622803094,
+      "loss": 4.2564,
+      "step": 8758
+    },
+    {
+      "epoch": 0.9082036692610282,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.000158140478851745,
+      "loss": 4.299,
+      "step": 8759
+    },
+    {
+      "epoch": 0.9083073573155163,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001581316407894629,
+      "loss": 4.2498,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9084110453700044,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015812280204128887,
+      "loss": 4.2857,
+      "step": 8761
+    },
+    {
+      "epoch": 0.9085147334244924,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015811396260732722,
+      "loss": 4.2647,
+      "step": 8762
+    },
+    {
+      "epoch": 0.9086184214789805,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015810512248768226,
+      "loss": 4.2407,
+      "step": 8763
+    },
+    {
+      "epoch": 0.9087221095334685,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015809628168245821,
+      "loss": 4.2563,
+      "step": 8764
+    },
+    {
+      "epoch": 0.9088257975879567,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015808744019175951,
+      "loss": 4.2673,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9089294856424447,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015807859801569043,
+      "loss": 4.2512,
+      "step": 8766
+    },
+    {
+      "epoch": 0.9090331736969328,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015806975515435522,
+      "loss": 4.3002,
+      "step": 8767
+    },
+    {
+      "epoch": 0.9091368617514208,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001580609116078583,
+      "loss": 4.2859,
+      "step": 8768
+    },
+    {
+      "epoch": 0.909240549805909,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015805206737630402,
+      "loss": 4.2534,
+      "step": 8769
+    },
+    {
+      "epoch": 0.909344237860397,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001580432224597967,
+      "loss": 4.3108,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9094479259148851,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015803437685844073,
+      "loss": 4.2701,
+      "step": 8771
+    },
+    {
+      "epoch": 0.9095516139693731,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001580255305723404,
+      "loss": 4.2773,
+      "step": 8772
+    },
+    {
+      "epoch": 0.9096553020238612,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001580166836016002,
+      "loss": 4.2535,
+      "step": 8773
+    },
+    {
+      "epoch": 0.9097589900783493,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015800783594632444,
+      "loss": 4.2504,
+      "step": 8774
+    },
+    {
+      "epoch": 0.9098626781328374,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015799898760661757,
+      "loss": 4.2086,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9099663661873254,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015799013858258388,
+      "loss": 4.2892,
+      "step": 8776
+    },
+    {
+      "epoch": 0.9100700542418135,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001579812888743279,
+      "loss": 4.2683,
+      "step": 8777
+    },
+    {
+      "epoch": 0.9101737422963015,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.000157972438481954,
+      "loss": 4.2804,
+      "step": 8778
+    },
+    {
+      "epoch": 0.9102774303507897,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015796358740556658,
+      "loss": 4.2349,
+      "step": 8779
+    },
+    {
+      "epoch": 0.9103811184052777,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015795473564527012,
+      "loss": 4.2952,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9104848064597658,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015794588320116904,
+      "loss": 4.3084,
+      "step": 8781
+    },
+    {
+      "epoch": 0.9105884945142538,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001579370300733678,
+      "loss": 4.2992,
+      "step": 8782
+    },
+    {
+      "epoch": 0.910692182568742,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015792817626197084,
+      "loss": 4.2066,
+      "step": 8783
+    },
+    {
+      "epoch": 0.91079587062323,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001579193217670826,
+      "loss": 4.2721,
+      "step": 8784
+    },
+    {
+      "epoch": 0.9108995586777181,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015791046658880759,
+      "loss": 4.3195,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9110032467322061,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015790161072725033,
+      "loss": 4.2495,
+      "step": 8786
+    },
+    {
+      "epoch": 0.9111069347866942,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015789275418251527,
+      "loss": 4.2673,
+      "step": 8787
+    },
+    {
+      "epoch": 0.9112106228411823,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015788389695470687,
+      "loss": 4.2598,
+      "step": 8788
+    },
+    {
+      "epoch": 0.9113143108956704,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015787503904392969,
+      "loss": 4.2678,
+      "step": 8789
+    },
+    {
+      "epoch": 0.9114179989501584,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001578661804502882,
+      "loss": 4.2586,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9115216870046465,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015785732117388698,
+      "loss": 4.2912,
+      "step": 8791
+    },
+    {
+      "epoch": 0.9116253750591345,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015784846121483053,
+      "loss": 4.3045,
+      "step": 8792
+    },
+    {
+      "epoch": 0.9117290631136227,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015783960057322335,
+      "loss": 4.2654,
+      "step": 8793
+    },
+    {
+      "epoch": 0.9118327511681107,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015783073924917004,
+      "loss": 4.2305,
+      "step": 8794
+    },
+    {
+      "epoch": 0.9119364392225988,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015782187724277519,
+      "loss": 4.2843,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9120401272770868,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015781301455414324,
+      "loss": 4.2485,
+      "step": 8796
+    },
+    {
+      "epoch": 0.912143815331575,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015780415118337885,
+      "loss": 4.2562,
+      "step": 8797
+    },
+    {
+      "epoch": 0.912247503386063,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015779528713058655,
+      "loss": 4.2176,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9123511914405511,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000157786422395871,
+      "loss": 4.3066,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9124548794950392,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015777755697933673,
+      "loss": 4.2327,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9125585675495272,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015776869088108834,
+      "loss": 4.2274,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9126622556040154,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001577598241012305,
+      "loss": 4.2839,
+      "step": 8802
+    },
+    {
+      "epoch": 0.9127659436585034,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015775095663986778,
+      "loss": 4.2804,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9128696317129915,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015774208849710482,
+      "loss": 4.299,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9129733197674795,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015773321967304624,
+      "loss": 4.272,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9130770078219677,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001577243501677967,
+      "loss": 4.2984,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9131806958764557,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015771547998146086,
+      "loss": 4.2805,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9132843839309438,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015770660911414332,
+      "loss": 4.2923,
+      "step": 8808
+    },
+    {
+      "epoch": 0.9133880719854318,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015769773756594885,
+      "loss": 4.2588,
+      "step": 8809
+    },
+    {
+      "epoch": 0.9134917600399199,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000157688865336982,
+      "loss": 4.2737,
+      "step": 8810
+    },
+    {
+      "epoch": 0.913595448094408,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015767999242734756,
+      "loss": 4.2774,
+      "step": 8811
+    },
+    {
+      "epoch": 0.9136991361488961,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015767111883715018,
+      "loss": 4.2914,
+      "step": 8812
+    },
+    {
+      "epoch": 0.9138028242033841,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015766224456649453,
+      "loss": 4.2476,
+      "step": 8813
+    },
+    {
+      "epoch": 0.9139065122578722,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015765336961548538,
+      "loss": 4.2929,
+      "step": 8814
+    },
+    {
+      "epoch": 0.9140102003123602,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015764449398422738,
+      "loss": 4.2975,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9141138883668484,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015763561767282534,
+      "loss": 4.2574,
+      "step": 8816
+    },
+    {
+      "epoch": 0.9142175764213364,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001576267406813839,
+      "loss": 4.2037,
+      "step": 8817
+    },
+    {
+      "epoch": 0.9143212644758245,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001576178630100078,
+      "loss": 4.2451,
+      "step": 8818
+    },
+    {
+      "epoch": 0.9144249525303125,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015760898465880184,
+      "loss": 4.2579,
+      "step": 8819
+    },
+    {
+      "epoch": 0.9145286405848007,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015760010562787077,
+      "loss": 4.289,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9146323286392887,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015759122591731936,
+      "loss": 4.2257,
+      "step": 8821
+    },
+    {
+      "epoch": 0.9147360166937768,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015758234552725234,
+      "loss": 4.224,
+      "step": 8822
+    },
+    {
+      "epoch": 0.9148397047482648,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015757346445777448,
+      "loss": 4.2634,
+      "step": 8823
+    },
+    {
+      "epoch": 0.9149433928027529,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015756458270899066,
+      "loss": 4.3201,
+      "step": 8824
+    },
+    {
+      "epoch": 0.915047080857241,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015755570028100558,
+      "loss": 4.2593,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9151507689117291,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015754681717392407,
+      "loss": 4.2888,
+      "step": 8826
+    },
+    {
+      "epoch": 0.9152544569662171,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015753793338785092,
+      "loss": 4.2838,
+      "step": 8827
+    },
+    {
+      "epoch": 0.9153581450207052,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015752904892289102,
+      "loss": 4.2517,
+      "step": 8828
+    },
+    {
+      "epoch": 0.9154618330751932,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015752016377914915,
+      "loss": 4.3019,
+      "step": 8829
+    },
+    {
+      "epoch": 0.9155655211296814,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015751127795673013,
+      "loss": 4.283,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9156692091841694,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015750239145573885,
+      "loss": 4.2745,
+      "step": 8831
+    },
+    {
+      "epoch": 0.9157728972386575,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015749350427628008,
+      "loss": 4.3084,
+      "step": 8832
+    },
+    {
+      "epoch": 0.9158765852931455,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015748461641845878,
+      "loss": 4.2416,
+      "step": 8833
+    },
+    {
+      "epoch": 0.9159802733476337,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015747572788237977,
+      "loss": 4.2328,
+      "step": 8834
+    },
+    {
+      "epoch": 0.9160839614021217,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001574668386681479,
+      "loss": 4.2722,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9161876494566098,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001574579487758681,
+      "loss": 4.2735,
+      "step": 8836
+    },
+    {
+      "epoch": 0.9162913375110978,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015744905820564524,
+      "loss": 4.2824,
+      "step": 8837
+    },
+    {
+      "epoch": 0.9163950255655859,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001574401669575842,
+      "loss": 4.2671,
+      "step": 8838
+    },
+    {
+      "epoch": 0.916498713620074,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015743127503178994,
+      "loss": 4.2662,
+      "step": 8839
+    },
+    {
+      "epoch": 0.9166024016745621,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001574223824283673,
+      "loss": 4.2068,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9167060897290501,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015741348914742126,
+      "loss": 4.2622,
+      "step": 8841
+    },
+    {
+      "epoch": 0.9168097777835382,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015740459518905677,
+      "loss": 4.3,
+      "step": 8842
+    },
+    {
+      "epoch": 0.9169134658380262,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001573957005533787,
+      "loss": 4.2615,
+      "step": 8843
+    },
+    {
+      "epoch": 0.9170171538925144,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015738680524049202,
+      "loss": 4.2809,
+      "step": 8844
+    },
+    {
+      "epoch": 0.9171208419470025,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015737790925050173,
+      "loss": 4.2593,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9172245300014905,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015736901258351277,
+      "loss": 4.271,
+      "step": 8846
+    },
+    {
+      "epoch": 0.9173282180559786,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015736011523963006,
+      "loss": 4.2784,
+      "step": 8847
+    },
+    {
+      "epoch": 0.9174319061104667,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015735121721895866,
+      "loss": 4.2922,
+      "step": 8848
+    },
+    {
+      "epoch": 0.9175355941649548,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001573423185216035,
+      "loss": 4.236,
+      "step": 8849
+    },
+    {
+      "epoch": 0.9176392822194428,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015733341914766962,
+      "loss": 4.2735,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9177429702739309,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015732451909726195,
+      "loss": 4.2984,
+      "step": 8851
+    },
+    {
+      "epoch": 0.9178466583284189,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001573156183704856,
+      "loss": 4.2864,
+      "step": 8852
+    },
+    {
+      "epoch": 0.9179503463829071,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015730671696744555,
+      "loss": 4.2688,
+      "step": 8853
+    },
+    {
+      "epoch": 0.9180540344373951,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001572978148882468,
+      "loss": 4.3005,
+      "step": 8854
+    },
+    {
+      "epoch": 0.9181577224918832,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015728891213299435,
+      "loss": 4.2281,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9182614105463712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001572800087017933,
+      "loss": 4.2397,
+      "step": 8856
+    },
+    {
+      "epoch": 0.9183650986008594,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015727110459474878,
+      "loss": 4.1826,
+      "step": 8857
+    },
+    {
+      "epoch": 0.9184687866553474,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015726219981196567,
+      "loss": 4.2488,
+      "step": 8858
+    },
+    {
+      "epoch": 0.9185724747098355,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015725329435354918,
+      "loss": 4.24,
+      "step": 8859
+    },
+    {
+      "epoch": 0.9186761627643235,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015724438821960432,
+      "loss": 4.2979,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9187798508188116,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015723548141023617,
+      "loss": 4.3016,
+      "step": 8861
+    },
+    {
+      "epoch": 0.9188835388732997,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015722657392554988,
+      "loss": 4.274,
+      "step": 8862
+    },
+    {
+      "epoch": 0.9189872269277878,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015721766576565048,
+      "loss": 4.2504,
+      "step": 8863
+    },
+    {
+      "epoch": 0.9190909149822758,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015720875693064313,
+      "loss": 4.2652,
+      "step": 8864
+    },
+    {
+      "epoch": 0.9191946030367639,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015719984742063292,
+      "loss": 4.2431,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9192982910912519,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015719093723572496,
+      "loss": 4.2611,
+      "step": 8866
+    },
+    {
+      "epoch": 0.9194019791457401,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001571820263760244,
+      "loss": 4.2941,
+      "step": 8867
+    },
+    {
+      "epoch": 0.9195056672002281,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015717311484163634,
+      "loss": 4.2652,
+      "step": 8868
+    },
+    {
+      "epoch": 0.9196093552547162,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.000157164202632666,
+      "loss": 4.2666,
+      "step": 8869
+    },
+    {
+      "epoch": 0.9197130433092042,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015715528974921848,
+      "loss": 4.2376,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9198167313636924,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015714637619139898,
+      "loss": 4.3034,
+      "step": 8871
+    },
+    {
+      "epoch": 0.9199204194181804,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015713746195931264,
+      "loss": 4.2802,
+      "step": 8872
+    },
+    {
+      "epoch": 0.9200241074726685,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001571285470530646,
+      "loss": 4.2629,
+      "step": 8873
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015711963147276015,
+      "loss": 4.2792,
+      "step": 8874
+    },
+    {
+      "epoch": 0.9202314835816446,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015711071521850437,
+      "loss": 4.2617,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9203351716361327,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015710179829040257,
+      "loss": 4.3119,
+      "step": 8876
+    },
+    {
+      "epoch": 0.9204388596906208,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015709288068855987,
+      "loss": 4.3074,
+      "step": 8877
+    },
+    {
+      "epoch": 0.9205425477451088,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015708396241308155,
+      "loss": 4.2577,
+      "step": 8878
+    },
+    {
+      "epoch": 0.9206462357995969,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00015707504346407282,
+      "loss": 4.2658,
+      "step": 8879
+    },
+    {
+      "epoch": 0.9207499238540849,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015706612384163888,
+      "loss": 4.2884,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9208536119085731,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015705720354588503,
+      "loss": 4.2734,
+      "step": 8881
+    },
+    {
+      "epoch": 0.9209572999630611,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015704828257691647,
+      "loss": 4.2624,
+      "step": 8882
+    },
+    {
+      "epoch": 0.9210609880175492,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015703936093483848,
+      "loss": 4.2541,
+      "step": 8883
+    },
+    {
+      "epoch": 0.9211646760720372,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015703043861975635,
+      "loss": 4.2992,
+      "step": 8884
+    },
+    {
+      "epoch": 0.9212683641265254,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015702151563177531,
+      "loss": 4.2204,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9213720521810134,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015701259197100067,
+      "loss": 4.2905,
+      "step": 8886
+    },
+    {
+      "epoch": 0.9214757402355015,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015700366763753772,
+      "loss": 4.2781,
+      "step": 8887
+    },
+    {
+      "epoch": 0.9215794282899896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001569947426314917,
+      "loss": 4.2797,
+      "step": 8888
+    },
+    {
+      "epoch": 0.9216831163444776,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015698581695296798,
+      "loss": 4.2928,
+      "step": 8889
+    },
+    {
+      "epoch": 0.9217868043989658,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015697689060207188,
+      "loss": 4.2258,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9218904924534538,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015696796357890868,
+      "loss": 4.2968,
+      "step": 8891
+    },
+    {
+      "epoch": 0.9219941805079419,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015695903588358373,
+      "loss": 4.2534,
+      "step": 8892
+    },
+    {
+      "epoch": 0.9220978685624299,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015695010751620237,
+      "loss": 4.2481,
+      "step": 8893
+    },
+    {
+      "epoch": 0.922201556616918,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001569411784768699,
+      "loss": 4.2957,
+      "step": 8894
+    },
+    {
+      "epoch": 0.9223052446714061,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015693224876569178,
+      "loss": 4.2965,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9224089327258942,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015692331838277327,
+      "loss": 4.2929,
+      "step": 8896
+    },
+    {
+      "epoch": 0.9225126207803822,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015691438732821979,
+      "loss": 4.3019,
+      "step": 8897
+    },
+    {
+      "epoch": 0.9226163088348703,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015690545560213668,
+      "loss": 4.2573,
+      "step": 8898
+    },
+    {
+      "epoch": 0.9227199968893584,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015689652320462937,
+      "loss": 4.2578,
+      "step": 8899
+    },
+    {
+      "epoch": 0.9228236849438465,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001568875901358032,
+      "loss": 4.2717,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9229273729983345,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015687865639576361,
+      "loss": 4.1983,
+      "step": 8901
+    },
+    {
+      "epoch": 0.9230310610528226,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.000156869721984616,
+      "loss": 4.2582,
+      "step": 8902
+    },
+    {
+      "epoch": 0.9231347491073106,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001568607869024658,
+      "loss": 4.295,
+      "step": 8903
+    },
+    {
+      "epoch": 0.9232384371617988,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001568518511494184,
+      "loss": 4.3088,
+      "step": 8904
+    },
+    {
+      "epoch": 0.9233421252162868,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015684291472557927,
+      "loss": 4.2766,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9234458132707749,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001568339776310538,
+      "loss": 4.2754,
+      "step": 8906
+    },
+    {
+      "epoch": 0.9235495013252629,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001568250398659475,
+      "loss": 4.3055,
+      "step": 8907
+    },
+    {
+      "epoch": 0.923653189379751,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015681610143036579,
+      "loss": 4.2631,
+      "step": 8908
+    },
+    {
+      "epoch": 0.9237568774342391,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001568071623244141,
+      "loss": 4.2663,
+      "step": 8909
+    },
+    {
+      "epoch": 0.9238605654887272,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015679822254819798,
+      "loss": 4.2789,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9239642535432152,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001567892821018229,
+      "loss": 4.2898,
+      "step": 8911
+    },
+    {
+      "epoch": 0.9240679415977033,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015678034098539427,
+      "loss": 4.267,
+      "step": 8912
+    },
+    {
+      "epoch": 0.9241716296521914,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015677139919901764,
+      "loss": 4.2639,
+      "step": 8913
+    },
+    {
+      "epoch": 0.9242753177066795,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015676245674279855,
+      "loss": 4.2804,
+      "step": 8914
+    },
+    {
+      "epoch": 0.9243790057611675,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015675351361684242,
+      "loss": 4.2755,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9244826938156556,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015674456982125485,
+      "loss": 4.2354,
+      "step": 8916
+    },
+    {
+      "epoch": 0.9245863818701436,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015673562535614135,
+      "loss": 4.2736,
+      "step": 8917
+    },
+    {
+      "epoch": 0.9246900699246318,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001567266802216074,
+      "loss": 4.2275,
+      "step": 8918
+    },
+    {
+      "epoch": 0.9247937579791198,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001567177344177586,
+      "loss": 4.2584,
+      "step": 8919
+    },
+    {
+      "epoch": 0.9248974460336079,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001567087879447005,
+      "loss": 4.2091,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9250011340880959,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015669984080253865,
+      "loss": 4.2761,
+      "step": 8921
+    },
+    {
+      "epoch": 0.925104822142584,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015669089299137862,
+      "loss": 4.3291,
+      "step": 8922
+    },
+    {
+      "epoch": 0.9252085101970721,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015668194451132596,
+      "loss": 4.2769,
+      "step": 8923
+    },
+    {
+      "epoch": 0.9253121982515602,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015667299536248626,
+      "loss": 4.3015,
+      "step": 8924
+    },
+    {
+      "epoch": 0.9254158863060482,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015666404554496514,
+      "loss": 4.275,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9255195743605363,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001566550950588682,
+      "loss": 4.2754,
+      "step": 8926
+    },
+    {
+      "epoch": 0.9256232624150244,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000156646143904301,
+      "loss": 4.298,
+      "step": 8927
+    },
+    {
+      "epoch": 0.9257269504695125,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015663719208136917,
+      "loss": 4.2431,
+      "step": 8928
+    },
+    {
+      "epoch": 0.9258306385240005,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015662823959017836,
+      "loss": 4.1848,
+      "step": 8929
+    },
+    {
+      "epoch": 0.9259343265784886,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015661928643083417,
+      "loss": 4.2754,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9260380146329766,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015661033260344224,
+      "loss": 4.2398,
+      "step": 8931
+    },
+    {
+      "epoch": 0.9261417026874648,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015660137810810825,
+      "loss": 4.2746,
+      "step": 8932
+    },
+    {
+      "epoch": 0.9262453907419529,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001565924229449378,
+      "loss": 4.2448,
+      "step": 8933
+    },
+    {
+      "epoch": 0.9263490787964409,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015658346711403662,
+      "loss": 4.2526,
+      "step": 8934
+    },
+    {
+      "epoch": 0.926452766850929,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001565745106155103,
+      "loss": 4.3112,
+      "step": 8935
+    },
+    {
+      "epoch": 0.926556454905417,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001565655534494646,
+      "loss": 4.2805,
+      "step": 8936
+    },
+    {
+      "epoch": 0.9266601429599052,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001565565956160051,
+      "loss": 4.2638,
+      "step": 8937
+    },
+    {
+      "epoch": 0.9267638310143932,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001565476371152376,
+      "loss": 4.2515,
+      "step": 8938
+    },
+    {
+      "epoch": 0.9268675190688813,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015653867794726773,
+      "loss": 4.2495,
+      "step": 8939
+    },
+    {
+      "epoch": 0.9269712071233693,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015652971811220125,
+      "loss": 4.2629,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9270748951778575,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015652075761014384,
+      "loss": 4.2958,
+      "step": 8941
+    },
+    {
+      "epoch": 0.9271785832323455,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015651179644120123,
+      "loss": 4.1946,
+      "step": 8942
+    },
+    {
+      "epoch": 0.9272822712868336,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015650283460547914,
+      "loss": 4.259,
+      "step": 8943
+    },
+    {
+      "epoch": 0.9273859593413216,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015649387210308336,
+      "loss": 4.2503,
+      "step": 8944
+    },
+    {
+      "epoch": 0.9274896473958097,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001564849089341196,
+      "loss": 4.2597,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9275933354502978,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001564759450986936,
+      "loss": 4.2625,
+      "step": 8946
+    },
+    {
+      "epoch": 0.9276970235047859,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001564669805969112,
+      "loss": 4.3084,
+      "step": 8947
+    },
+    {
+      "epoch": 0.9278007115592739,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015645801542887806,
+      "loss": 4.2666,
+      "step": 8948
+    },
+    {
+      "epoch": 0.927904399613762,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015644904959470003,
+      "loss": 4.2831,
+      "step": 8949
+    },
+    {
+      "epoch": 0.92800808766825,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015644008309448291,
+      "loss": 4.2836,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9281117757227382,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015643111592833245,
+      "loss": 4.2428,
+      "step": 8951
+    },
+    {
+      "epoch": 0.9282154637772262,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015642214809635444,
+      "loss": 4.2951,
+      "step": 8952
+    },
+    {
+      "epoch": 0.9283191518317143,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015641317959865476,
+      "loss": 4.2916,
+      "step": 8953
+    },
+    {
+      "epoch": 0.9284228398862023,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001564042104353392,
+      "loss": 4.3157,
+      "step": 8954
+    },
+    {
+      "epoch": 0.9285265279406905,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015639524060651356,
+      "loss": 4.2512,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9286302159951785,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001563862701122837,
+      "loss": 4.2779,
+      "step": 8956
+    },
+    {
+      "epoch": 0.9287339040496666,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015637729895275545,
+      "loss": 4.2798,
+      "step": 8957
+    },
+    {
+      "epoch": 0.9288375921041546,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001563683271280347,
+      "loss": 4.2495,
+      "step": 8958
+    },
+    {
+      "epoch": 0.9289412801586427,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015635935463822722,
+      "loss": 4.2622,
+      "step": 8959
+    },
+    {
+      "epoch": 0.9290449682131308,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015635038148343895,
+      "loss": 4.2685,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9291486562676189,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015634140766377573,
+      "loss": 4.264,
+      "step": 8961
+    },
+    {
+      "epoch": 0.9292523443221069,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001563324331793435,
+      "loss": 4.31,
+      "step": 8962
+    },
+    {
+      "epoch": 0.929356032376595,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00015632345803024805,
+      "loss": 4.2839,
+      "step": 8963
+    },
+    {
+      "epoch": 0.929459720431083,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015631448221659536,
+      "loss": 4.3095,
+      "step": 8964
+    },
+    {
+      "epoch": 0.9295634084855712,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00015630550573849132,
+      "loss": 4.2685,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9296670965400592,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015629652859604182,
+      "loss": 4.2873,
+      "step": 8966
+    },
+    {
+      "epoch": 0.9297707845945473,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015628755078935278,
+      "loss": 4.2562,
+      "step": 8967
+    },
+    {
+      "epoch": 0.9298744726490353,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00015627857231853014,
+      "loss": 4.2917,
+      "step": 8968
+    },
+    {
+      "epoch": 0.9299781607035235,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001562695931836798,
+      "loss": 4.2779,
+      "step": 8969
+    },
+    {
+      "epoch": 0.9300818487580115,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015626061338490783,
+      "loss": 4.2989,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9301855368124996,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015625163292232004,
+      "loss": 4.2377,
+      "step": 8971
+    },
+    {
+      "epoch": 0.9302892248669876,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015624265179602244,
+      "loss": 4.2382,
+      "step": 8972
+    },
+    {
+      "epoch": 0.9303929129214757,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015623367000612098,
+      "loss": 4.2848,
+      "step": 8973
+    },
+    {
+      "epoch": 0.9304966009759638,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001562246875527217,
+      "loss": 4.2626,
+      "step": 8974
+    },
+    {
+      "epoch": 0.9306002890304519,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015621570443593053,
+      "loss": 4.2344,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9307039770849399,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015620672065585345,
+      "loss": 4.2367,
+      "step": 8976
+    },
+    {
+      "epoch": 0.930807665139428,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015619773621259648,
+      "loss": 4.2997,
+      "step": 8977
+    },
+    {
+      "epoch": 0.9309113531939162,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015618875110626562,
+      "loss": 4.2668,
+      "step": 8978
+    },
+    {
+      "epoch": 0.9310150412484042,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015617976533696692,
+      "loss": 4.2893,
+      "step": 8979
+    },
+    {
+      "epoch": 0.9311187293028923,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015617077890480637,
+      "loss": 4.2655,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9312224173573803,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015616179180988996,
+      "loss": 4.2781,
+      "step": 8981
+    },
+    {
+      "epoch": 0.9313261054118684,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015615280405232383,
+      "loss": 4.2794,
+      "step": 8982
+    },
+    {
+      "epoch": 0.9314297934663565,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015614381563221395,
+      "loss": 4.307,
+      "step": 8983
+    },
+    {
+      "epoch": 0.9315334815208446,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015613482654966638,
+      "loss": 4.286,
+      "step": 8984
+    },
+    {
+      "epoch": 0.9316371695753326,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015612583680478718,
+      "loss": 4.23,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9317408576298207,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001561168463976825,
+      "loss": 4.2303,
+      "step": 8986
+    },
+    {
+      "epoch": 0.9318445456843087,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001561078553284583,
+      "loss": 4.33,
+      "step": 8987
+    },
+    {
+      "epoch": 0.9319482337387969,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001560988635972207,
+      "loss": 4.2243,
+      "step": 8988
+    },
+    {
+      "epoch": 0.9320519217932849,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001560898712040758,
+      "loss": 4.2901,
+      "step": 8989
+    },
+    {
+      "epoch": 0.932155609847773,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015608087814912978,
+      "loss": 4.2306,
+      "step": 8990
+    },
+    {
+      "epoch": 0.932259297902261,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015607188443248862,
+      "loss": 4.2759,
+      "step": 8991
+    },
+    {
+      "epoch": 0.9323629859567492,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015606289005425848,
+      "loss": 4.2388,
+      "step": 8992
+    },
+    {
+      "epoch": 0.9324666740112372,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015605389501454554,
+      "loss": 4.2662,
+      "step": 8993
+    },
+    {
+      "epoch": 0.9325703620657253,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015604489931345588,
+      "loss": 4.246,
+      "step": 8994
+    },
+    {
+      "epoch": 0.9326740501202133,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015603590295109565,
+      "loss": 4.2991,
+      "step": 8995
+    },
+    {
+      "epoch": 0.9327777381747014,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.000156026905927571,
+      "loss": 4.2641,
+      "step": 8996
+    },
+    {
+      "epoch": 0.9328814262291895,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00015601790824298808,
+      "loss": 4.2938,
+      "step": 8997
+    },
+    {
+      "epoch": 0.9329851142836776,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015600890989745305,
+      "loss": 4.2682,
+      "step": 8998
+    },
+    {
+      "epoch": 0.9330888023381656,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00015599991089107208,
+      "loss": 4.2853,
+      "step": 8999
+    },
+    {
+      "epoch": 0.9331924903926537,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015599091122395139,
+      "loss": 4.2797,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9332961784471417,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001559819108961971,
+      "loss": 4.2417,
+      "step": 9001
+    },
+    {
+      "epoch": 0.9333998665016299,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015597290990791544,
+      "loss": 4.2504,
+      "step": 9002
+    },
+    {
+      "epoch": 0.9335035545561179,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00015596390825921264,
+      "loss": 4.2459,
+      "step": 9003
+    },
+    {
+      "epoch": 0.933607242610606,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015595490595019483,
+      "loss": 4.2283,
+      "step": 9004
+    },
+    {
+      "epoch": 0.933710930665094,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00015594590298096832,
+      "loss": 4.2622,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9338146187195822,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015593689935163924,
+      "loss": 4.2009,
+      "step": 9006
+    },
+    {
+      "epoch": 0.9339183067740702,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015592789506231394,
+      "loss": 4.2585,
+      "step": 9007
+    },
+    {
+      "epoch": 0.9340219948285583,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015591889011309858,
+      "loss": 4.2724,
+      "step": 9008
+    },
+    {
+      "epoch": 0.9341256828830463,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015590988450409939,
+      "loss": 4.301,
+      "step": 9009
+    },
+    {
+      "epoch": 0.9342293709375344,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015590087823542267,
+      "loss": 4.2327,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9343330589920225,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001558918713071747,
+      "loss": 4.2382,
+      "step": 9011
+    },
+    {
+      "epoch": 0.9344367470465106,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001558828637194617,
+      "loss": 4.2565,
+      "step": 9012
+    },
+    {
+      "epoch": 0.9345404351009986,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015587385547239,
+      "loss": 4.2808,
+      "step": 9013
+    },
+    {
+      "epoch": 0.9346441231554867,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001558648465660659,
+      "loss": 4.2589,
+      "step": 9014
+    },
+    {
+      "epoch": 0.9347478112099747,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015585583700059564,
+      "loss": 4.2889,
+      "step": 9015
+    },
+    {
+      "epoch": 0.9348514992644629,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001558468267760855,
+      "loss": 4.2689,
+      "step": 9016
+    },
+    {
+      "epoch": 0.9349551873189509,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015583781589264186,
+      "loss": 4.2566,
+      "step": 9017
+    },
+    {
+      "epoch": 0.935058875373439,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015582880435037103,
+      "loss": 4.2746,
+      "step": 9018
+    },
+    {
+      "epoch": 0.935162563427927,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015581979214937932,
+      "loss": 4.2494,
+      "step": 9019
+    },
+    {
+      "epoch": 0.9352662514824152,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015581077928977308,
+      "loss": 4.3052,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9353699395369032,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001558017657716586,
+      "loss": 4.2869,
+      "step": 9021
+    },
+    {
+      "epoch": 0.9354736275913913,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001557927515951423,
+      "loss": 4.1821,
+      "step": 9022
+    },
+    {
+      "epoch": 0.9355773156458794,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001557837367603305,
+      "loss": 4.2827,
+      "step": 9023
+    },
+    {
+      "epoch": 0.9356810037003674,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001557747212673296,
+      "loss": 4.2785,
+      "step": 9024
+    },
+    {
+      "epoch": 0.9357846917548556,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001557657051162459,
+      "loss": 4.2271,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9358883798093436,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015575668830718583,
+      "loss": 4.26,
+      "step": 9026
+    },
+    {
+      "epoch": 0.9359920678638317,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001557476708402558,
+      "loss": 4.2341,
+      "step": 9027
+    },
+    {
+      "epoch": 0.9360957559183197,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015573865271556217,
+      "loss": 4.2817,
+      "step": 9028
+    },
+    {
+      "epoch": 0.9361994439728079,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015572963393321136,
+      "loss": 4.294,
+      "step": 9029
+    },
+    {
+      "epoch": 0.9363031320272959,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015572061449330976,
+      "loss": 4.2312,
+      "step": 9030
+    },
+    {
+      "epoch": 0.936406820081784,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015571159439596382,
+      "loss": 4.2447,
+      "step": 9031
+    },
+    {
+      "epoch": 0.936510508136272,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015570257364127995,
+      "loss": 4.2631,
+      "step": 9032
+    },
+    {
+      "epoch": 0.9366141961907601,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001556935522293646,
+      "loss": 4.2454,
+      "step": 9033
+    },
+    {
+      "epoch": 0.9367178842452482,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001556845301603242,
+      "loss": 4.2791,
+      "step": 9034
+    },
+    {
+      "epoch": 0.9368215722997363,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001556755074342652,
+      "loss": 4.2614,
+      "step": 9035
+    },
+    {
+      "epoch": 0.9369252603542243,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001556664840512941,
+      "loss": 4.2768,
+      "step": 9036
+    },
+    {
+      "epoch": 0.9370289484087124,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001556574600115173,
+      "loss": 4.2986,
+      "step": 9037
+    },
+    {
+      "epoch": 0.9371326364632004,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001556484353150413,
+      "loss": 4.3204,
+      "step": 9038
+    },
+    {
+      "epoch": 0.9372363245176886,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001556394099619726,
+      "loss": 4.262,
+      "step": 9039
+    },
+    {
+      "epoch": 0.9373400125721766,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015563038395241766,
+      "loss": 4.3063,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9374437006266647,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015562135728648303,
+      "loss": 4.2485,
+      "step": 9041
+    },
+    {
+      "epoch": 0.9375473886811527,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015561232996427513,
+      "loss": 4.2451,
+      "step": 9042
+    },
+    {
+      "epoch": 0.9376510767356409,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015560330198590054,
+      "loss": 4.2971,
+      "step": 9043
+    },
+    {
+      "epoch": 0.9377547647901289,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015559427335146579,
+      "loss": 4.2891,
+      "step": 9044
+    },
+    {
+      "epoch": 0.937858452844617,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015558524406107733,
+      "loss": 4.27,
+      "step": 9045
+    },
+    {
+      "epoch": 0.937962140899105,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015557621411484176,
+      "loss": 4.2278,
+      "step": 9046
+    },
+    {
+      "epoch": 0.9380658289535931,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015556718351286564,
+      "loss": 4.2598,
+      "step": 9047
+    },
+    {
+      "epoch": 0.9381695170080812,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015555815225525547,
+      "loss": 4.2699,
+      "step": 9048
+    },
+    {
+      "epoch": 0.9382732050625693,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015554912034211783,
+      "loss": 4.2374,
+      "step": 9049
+    },
+    {
+      "epoch": 0.9383768931170573,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015554008777355927,
+      "loss": 4.2409,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9384805811715454,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015553105454968642,
+      "loss": 4.2517,
+      "step": 9051
+    },
+    {
+      "epoch": 0.9385842692260334,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001555220206706058,
+      "loss": 4.2676,
+      "step": 9052
+    },
+    {
+      "epoch": 0.9386879572805216,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015551298613642405,
+      "loss": 4.2632,
+      "step": 9053
+    },
+    {
+      "epoch": 0.9387916453350096,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001555039509472477,
+      "loss": 4.2829,
+      "step": 9054
+    },
+    {
+      "epoch": 0.9388953333894977,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015549491510318344,
+      "loss": 4.2757,
+      "step": 9055
+    },
+    {
+      "epoch": 0.9389990214439857,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015548587860433782,
+      "loss": 4.2739,
+      "step": 9056
+    },
+    {
+      "epoch": 0.9391027094984739,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001554768414508175,
+      "loss": 4.3261,
+      "step": 9057
+    },
+    {
+      "epoch": 0.9392063975529619,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001554678036427291,
+      "loss": 4.2985,
+      "step": 9058
+    },
+    {
+      "epoch": 0.93931008560745,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015545876518017923,
+      "loss": 4.2876,
+      "step": 9059
+    },
+    {
+      "epoch": 0.939413773661938,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015544972606327458,
+      "loss": 4.256,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9395174617164261,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015544068629212173,
+      "loss": 4.2775,
+      "step": 9061
+    },
+    {
+      "epoch": 0.9396211497709142,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015543164586682742,
+      "loss": 4.28,
+      "step": 9062
+    },
+    {
+      "epoch": 0.9397248378254023,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015542260478749827,
+      "loss": 4.2796,
+      "step": 9063
+    },
+    {
+      "epoch": 0.9398285258798903,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015541356305424095,
+      "loss": 4.2735,
+      "step": 9064
+    },
+    {
+      "epoch": 0.9399322139343784,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001554045206671622,
+      "loss": 4.2218,
+      "step": 9065
+    },
+    {
+      "epoch": 0.9400359019888664,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015539547762636864,
+      "loss": 4.2663,
+      "step": 9066
+    },
+    {
+      "epoch": 0.9401395900433546,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015538643393196703,
+      "loss": 4.2504,
+      "step": 9067
+    },
+    {
+      "epoch": 0.9402432780978427,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.000155377389584064,
+      "loss": 4.1995,
+      "step": 9068
+    },
+    {
+      "epoch": 0.9403469661523307,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015536834458276634,
+      "loss": 4.2034,
+      "step": 9069
+    },
+    {
+      "epoch": 0.9404506542068188,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015535929892818073,
+      "loss": 4.2722,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9405543422613069,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001553502526204139,
+      "loss": 4.2463,
+      "step": 9071
+    },
+    {
+      "epoch": 0.940658030315795,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015534120565957263,
+      "loss": 4.2895,
+      "step": 9072
+    },
+    {
+      "epoch": 0.940761718370283,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015533215804576362,
+      "loss": 4.2403,
+      "step": 9073
+    },
+    {
+      "epoch": 0.9408654064247711,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001553231097790936,
+      "loss": 4.2882,
+      "step": 9074
+    },
+    {
+      "epoch": 0.9409690944792591,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001553140608596694,
+      "loss": 4.2911,
+      "step": 9075
+    },
+    {
+      "epoch": 0.9410727825337473,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015530501128759773,
+      "loss": 4.2612,
+      "step": 9076
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015529596106298542,
+      "loss": 4.2227,
+      "step": 9077
+    },
+    {
+      "epoch": 0.9412801586427234,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015528691018593918,
+      "loss": 4.2334,
+      "step": 9078
+    },
+    {
+      "epoch": 0.9413838466972114,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015527785865656588,
+      "loss": 4.2482,
+      "step": 9079
+    },
+    {
+      "epoch": 0.9414875347516996,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015526880647497224,
+      "loss": 4.2654,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9415912228061876,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015525975364126513,
+      "loss": 4.3063,
+      "step": 9081
+    },
+    {
+      "epoch": 0.9416949108606757,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015525070015555135,
+      "loss": 4.2706,
+      "step": 9082
+    },
+    {
+      "epoch": 0.9417985989151637,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015524164601793768,
+      "loss": 4.2785,
+      "step": 9083
+    },
+    {
+      "epoch": 0.9419022869696518,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.000155232591228531,
+      "loss": 4.234,
+      "step": 9084
+    },
+    {
+      "epoch": 0.9420059750241399,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001552235357874381,
+      "loss": 4.2799,
+      "step": 9085
+    },
+    {
+      "epoch": 0.942109663078628,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001552144796947659,
+      "loss": 4.2756,
+      "step": 9086
+    },
+    {
+      "epoch": 0.942213351133116,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015520542295062117,
+      "loss": 4.2696,
+      "step": 9087
+    },
+    {
+      "epoch": 0.9423170391876041,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015519636555511078,
+      "loss": 4.2314,
+      "step": 9088
+    },
+    {
+      "epoch": 0.9424207272420921,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015518730750834167,
+      "loss": 4.2627,
+      "step": 9089
+    },
+    {
+      "epoch": 0.9425244152965803,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015517824881042066,
+      "loss": 4.2633,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9426281033510683,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001551691894614546,
+      "loss": 4.2876,
+      "step": 9091
+    },
+    {
+      "epoch": 0.9427317914055564,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015516012946155047,
+      "loss": 4.231,
+      "step": 9092
+    },
+    {
+      "epoch": 0.9428354794600444,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015515106881081508,
+      "loss": 4.3029,
+      "step": 9093
+    },
+    {
+      "epoch": 0.9429391675145326,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015514200750935543,
+      "loss": 4.28,
+      "step": 9094
+    },
+    {
+      "epoch": 0.9430428555690206,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015513294555727833,
+      "loss": 4.2345,
+      "step": 9095
+    },
+    {
+      "epoch": 0.9431465436235087,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015512388295469077,
+      "loss": 4.2922,
+      "step": 9096
+    },
+    {
+      "epoch": 0.9432502316779967,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015511481970169964,
+      "loss": 4.2515,
+      "step": 9097
+    },
+    {
+      "epoch": 0.9433539197324848,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015510575579841192,
+      "loss": 4.3032,
+      "step": 9098
+    },
+    {
+      "epoch": 0.9434576077869729,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00015509669124493452,
+      "loss": 4.2944,
+      "step": 9099
+    },
+    {
+      "epoch": 0.943561295841461,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015508762604137438,
+      "loss": 4.2555,
+      "step": 9100
+    },
+    {
+      "epoch": 0.943664983895949,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001550785601878385,
+      "loss": 4.2433,
+      "step": 9101
+    },
+    {
+      "epoch": 0.9437686719504371,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015506949368443386,
+      "loss": 4.2553,
+      "step": 9102
+    },
+    {
+      "epoch": 0.9438723600049251,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015506042653126736,
+      "loss": 4.2619,
+      "step": 9103
+    },
+    {
+      "epoch": 0.9439760480594133,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015505135872844608,
+      "loss": 4.2375,
+      "step": 9104
+    },
+    {
+      "epoch": 0.9440797361139013,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001550422902760769,
+      "loss": 4.322,
+      "step": 9105
+    },
+    {
+      "epoch": 0.9441834241683894,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001550332211742669,
+      "loss": 4.2436,
+      "step": 9106
+    },
+    {
+      "epoch": 0.9442871122228774,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001550241514231231,
+      "loss": 4.2417,
+      "step": 9107
+    },
+    {
+      "epoch": 0.9443908002773655,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001550150810227524,
+      "loss": 4.3237,
+      "step": 9108
+    },
+    {
+      "epoch": 0.9444944883318536,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00015500600997326195,
+      "loss": 4.2396,
+      "step": 9109
+    },
+    {
+      "epoch": 0.9445981763863417,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015499693827475874,
+      "loss": 4.2631,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9447018644408297,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00015498786592734977,
+      "loss": 4.2638,
+      "step": 9111
+    },
+    {
+      "epoch": 0.9448055524953178,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001549787929311421,
+      "loss": 4.2598,
+      "step": 9112
+    },
+    {
+      "epoch": 0.944909240549806,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001549697192862428,
+      "loss": 4.2533,
+      "step": 9113
+    },
+    {
+      "epoch": 0.945012928604294,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001549606449927589,
+      "loss": 4.2861,
+      "step": 9114
+    },
+    {
+      "epoch": 0.9451166166587821,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015495157005079751,
+      "loss": 4.2727,
+      "step": 9115
+    },
+    {
+      "epoch": 0.9452203047132701,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015494249446046566,
+      "loss": 4.3136,
+      "step": 9116
+    },
+    {
+      "epoch": 0.9453239927677582,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015493341822187047,
+      "loss": 4.2814,
+      "step": 9117
+    },
+    {
+      "epoch": 0.9454276808222463,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000154924341335119,
+      "loss": 4.2213,
+      "step": 9118
+    },
+    {
+      "epoch": 0.9455313688767344,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015491526380031836,
+      "loss": 4.257,
+      "step": 9119
+    },
+    {
+      "epoch": 0.9456350569312224,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015490618561757566,
+      "loss": 4.3009,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9457387449857105,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.000154897106786998,
+      "loss": 4.2634,
+      "step": 9121
+    },
+    {
+      "epoch": 0.9458424330401985,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015488802730869252,
+      "loss": 4.3353,
+      "step": 9122
+    },
+    {
+      "epoch": 0.9459461210946867,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015487894718276633,
+      "loss": 4.2733,
+      "step": 9123
+    },
+    {
+      "epoch": 0.9460498091491747,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015486986640932658,
+      "loss": 4.2482,
+      "step": 9124
+    },
+    {
+      "epoch": 0.9461534972036628,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001548607849884804,
+      "loss": 4.2753,
+      "step": 9125
+    },
+    {
+      "epoch": 0.9462571852581508,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015485170292033496,
+      "loss": 4.2713,
+      "step": 9126
+    },
+    {
+      "epoch": 0.946360873312639,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001548426202049974,
+      "loss": 4.2871,
+      "step": 9127
+    },
+    {
+      "epoch": 0.946464561367127,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015483353684257487,
+      "loss": 4.2852,
+      "step": 9128
+    },
+    {
+      "epoch": 0.9465682494216151,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001548244528331746,
+      "loss": 4.2846,
+      "step": 9129
+    },
+    {
+      "epoch": 0.9466719374761031,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001548153681769037,
+      "loss": 4.2558,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9467756255305912,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001548062828738694,
+      "loss": 4.2772,
+      "step": 9131
+    },
+    {
+      "epoch": 0.9468793135850793,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015479719692417892,
+      "loss": 4.2572,
+      "step": 9132
+    },
+    {
+      "epoch": 0.9469830016395674,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00015478811032793946,
+      "loss": 4.288,
+      "step": 9133
+    },
+    {
+      "epoch": 0.9470866896940554,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015477902308525817,
+      "loss": 4.2884,
+      "step": 9134
+    },
+    {
+      "epoch": 0.9471903777485435,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015476993519624233,
+      "loss": 4.2595,
+      "step": 9135
+    },
+    {
+      "epoch": 0.9472940658030315,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015476084666099914,
+      "loss": 4.2354,
+      "step": 9136
+    },
+    {
+      "epoch": 0.9473977538575197,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015475175747963586,
+      "loss": 4.1977,
+      "step": 9137
+    },
+    {
+      "epoch": 0.9475014419120077,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015474266765225968,
+      "loss": 4.294,
+      "step": 9138
+    },
+    {
+      "epoch": 0.9476051299664958,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00015473357717897792,
+      "loss": 4.2882,
+      "step": 9139
+    },
+    {
+      "epoch": 0.9477088180209838,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015472448605989783,
+      "loss": 4.2827,
+      "step": 9140
+    },
+    {
+      "epoch": 0.947812506075472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001547153942951266,
+      "loss": 4.2591,
+      "step": 9141
+    },
+    {
+      "epoch": 0.94791619412996,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001547063018847716,
+      "loss": 4.2223,
+      "step": 9142
+    },
+    {
+      "epoch": 0.9480198821844481,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015469720882894,
+      "loss": 4.2614,
+      "step": 9143
+    },
+    {
+      "epoch": 0.9481235702389361,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015468811512773923,
+      "loss": 4.2703,
+      "step": 9144
+    },
+    {
+      "epoch": 0.9482272582934242,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015467902078127645,
+      "loss": 4.2755,
+      "step": 9145
+    },
+    {
+      "epoch": 0.9483309463479123,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015466992578965907,
+      "loss": 4.2887,
+      "step": 9146
+    },
+    {
+      "epoch": 0.9484346344024004,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015466083015299432,
+      "loss": 4.2957,
+      "step": 9147
+    },
+    {
+      "epoch": 0.9485383224568884,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00015465173387138955,
+      "loss": 4.2571,
+      "step": 9148
+    },
+    {
+      "epoch": 0.9486420105113765,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001546426369449521,
+      "loss": 4.2698,
+      "step": 9149
+    },
+    {
+      "epoch": 0.9487456985658645,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001546335393737893,
+      "loss": 4.2932,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9488493866203527,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001546244411580085,
+      "loss": 4.2408,
+      "step": 9151
+    },
+    {
+      "epoch": 0.9489530746748407,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015461534229771698,
+      "loss": 4.2668,
+      "step": 9152
+    },
+    {
+      "epoch": 0.9490567627293288,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001546062427930222,
+      "loss": 4.2251,
+      "step": 9153
+    },
+    {
+      "epoch": 0.9491604507838168,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015459714264403149,
+      "loss": 4.2717,
+      "step": 9154
+    },
+    {
+      "epoch": 0.949264138838305,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015458804185085215,
+      "loss": 4.2636,
+      "step": 9155
+    },
+    {
+      "epoch": 0.9493678268927931,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00015457894041359164,
+      "loss": 4.2856,
+      "step": 9156
+    },
+    {
+      "epoch": 0.9494715149472811,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015456983833235735,
+      "loss": 4.2362,
+      "step": 9157
+    },
+    {
+      "epoch": 0.9495752030017692,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015456073560725665,
+      "loss": 4.2561,
+      "step": 9158
+    },
+    {
+      "epoch": 0.9496788910562572,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015455163223839692,
+      "loss": 4.2903,
+      "step": 9159
+    },
+    {
+      "epoch": 0.9497825791107454,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001545425282258856,
+      "loss": 4.3087,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9498862671652334,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015453342356983014,
+      "loss": 4.3171,
+      "step": 9161
+    },
+    {
+      "epoch": 0.9499899552197215,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001545243182703379,
+      "loss": 4.2506,
+      "step": 9162
+    },
+    {
+      "epoch": 0.9500936432742095,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015451521232751628,
+      "loss": 4.2497,
+      "step": 9163
+    },
+    {
+      "epoch": 0.9501973313286977,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015450610574147284,
+      "loss": 4.2478,
+      "step": 9164
+    },
+    {
+      "epoch": 0.9503010193831857,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.000154496998512315,
+      "loss": 4.2293,
+      "step": 9165
+    },
+    {
+      "epoch": 0.9504047074376738,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001544878906401501,
+      "loss": 4.2779,
+      "step": 9166
+    },
+    {
+      "epoch": 0.9505083954921618,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015447878212508572,
+      "loss": 4.226,
+      "step": 9167
+    },
+    {
+      "epoch": 0.9506120835466499,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001544696729672293,
+      "loss": 4.2892,
+      "step": 9168
+    },
+    {
+      "epoch": 0.950715771601138,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015446056316668837,
+      "loss": 4.2678,
+      "step": 9169
+    },
+    {
+      "epoch": 0.9508194596556261,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00015445145272357028,
+      "loss": 4.2994,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9509231477101141,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00015444234163798263,
+      "loss": 4.2739,
+      "step": 9171
+    },
+    {
+      "epoch": 0.9510268357646022,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001544332299100329,
+      "loss": 4.2484,
+      "step": 9172
+    },
+    {
+      "epoch": 0.9511305238190902,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015442411753982863,
+      "loss": 4.2454,
+      "step": 9173
+    },
+    {
+      "epoch": 0.9512342118735784,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015441500452747725,
+      "loss": 4.2962,
+      "step": 9174
+    },
+    {
+      "epoch": 0.9513378999280664,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015440589087308636,
+      "loss": 4.2961,
+      "step": 9175
+    },
+    {
+      "epoch": 0.9514415879825545,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015439677657676346,
+      "loss": 4.2194,
+      "step": 9176
+    },
+    {
+      "epoch": 0.9515452760370425,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001543876616386161,
+      "loss": 4.3136,
+      "step": 9177
+    },
+    {
+      "epoch": 0.9516489640915307,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001543785460587518,
+      "loss": 4.2778,
+      "step": 9178
+    },
+    {
+      "epoch": 0.9517526521460187,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001543694298372782,
+      "loss": 4.2829,
+      "step": 9179
+    },
+    {
+      "epoch": 0.9518563402005068,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015436031297430276,
+      "loss": 4.2372,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9519600282549948,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00015435119546993307,
+      "loss": 4.2937,
+      "step": 9181
+    },
+    {
+      "epoch": 0.9520637163094829,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015434207732427675,
+      "loss": 4.2077,
+      "step": 9182
+    },
+    {
+      "epoch": 0.952167404363971,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00015433295853744134,
+      "loss": 4.2331,
+      "step": 9183
+    },
+    {
+      "epoch": 0.9522710924184591,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015432383910953448,
+      "loss": 4.2744,
+      "step": 9184
+    },
+    {
+      "epoch": 0.9523747804729471,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00015431471904066372,
+      "loss": 4.2522,
+      "step": 9185
+    },
+    {
+      "epoch": 0.9524784685274352,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015430559833093672,
+      "loss": 4.2428,
+      "step": 9186
+    },
+    {
+      "epoch": 0.9525821565819232,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015429647698046104,
+      "loss": 4.2723,
+      "step": 9187
+    },
+    {
+      "epoch": 0.9526858446364114,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001542873549893443,
+      "loss": 4.2884,
+      "step": 9188
+    },
+    {
+      "epoch": 0.9527895326908994,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00015427823235769416,
+      "loss": 4.2651,
+      "step": 9189
+    },
+    {
+      "epoch": 0.9528932207453875,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001542691090856183,
+      "loss": 4.2941,
+      "step": 9190
+    },
+    {
+      "epoch": 0.9529969087998755,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015425998517322426,
+      "loss": 4.2693,
+      "step": 9191
+    },
+    {
+      "epoch": 0.9531005968543637,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001542508606206198,
+      "loss": 4.2732,
+      "step": 9192
+    },
+    {
+      "epoch": 0.9532042849088517,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00015424173542791253,
+      "loss": 4.2487,
+      "step": 9193
+    },
+    {
+      "epoch": 0.9533079729633398,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015423260959521008,
+      "loss": 4.2602,
+      "step": 9194
+    },
+    {
+      "epoch": 0.9534116610178278,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001542234831226202,
+      "loss": 4.325,
+      "step": 9195
+    },
+    {
+      "epoch": 0.9535153490723159,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00015421435601025048,
+      "loss": 4.2692,
+      "step": 9196
+    },
+    {
+      "epoch": 0.953619037126804,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001542052282582087,
+      "loss": 4.251,
+      "step": 9197
+    },
+    {
+      "epoch": 0.9537227251812921,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00015419609986660256,
+      "loss": 4.2267,
+      "step": 9198
+    },
+    {
+      "epoch": 0.9538264132357801,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015418697083553968,
+      "loss": 4.2704,
+      "step": 9199
+    },
+    {
+      "epoch": 0.9539301012902682,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015417784116512784,
+      "loss": 4.2562,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9540337893447564,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015416871085547474,
+      "loss": 4.2652,
+      "step": 9201
+    },
+    {
+      "epoch": 0.9541374773992444,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015415957990668808,
+      "loss": 4.2949,
+      "step": 9202
+    },
+    {
+      "epoch": 0.9542411654537325,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00015415044831887568,
+      "loss": 4.2267,
+      "step": 9203
+    },
+    {
+      "epoch": 0.9543448535082205,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00015414131609214522,
+      "loss": 4.2524,
+      "step": 9204
+    },
+    {
+      "epoch": 0.9544485415627086,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015413218322660443,
+      "loss": 4.2644,
+      "step": 9205
+    },
+    {
+      "epoch": 0.9545522296171967,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015412304972236112,
+      "loss": 4.2714,
+      "step": 9206
+    },
+    {
+      "epoch": 0.9546559176716848,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015411391557952303,
+      "loss": 4.3073,
+      "step": 9207
+    },
+    {
+      "epoch": 0.9547596057261728,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015410478079819795,
+      "loss": 4.2954,
+      "step": 9208
+    },
+    {
+      "epoch": 0.9548632937806609,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001540956453784936,
+      "loss": 4.2384,
+      "step": 9209
+    },
+    {
+      "epoch": 0.9549669818351489,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015408650932051786,
+      "loss": 4.2725,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9550706698896371,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015407737262437848,
+      "loss": 4.2239,
+      "step": 9211
+    },
+    {
+      "epoch": 0.9551743579441251,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015406823529018327,
+      "loss": 4.247,
+      "step": 9212
+    },
+    {
+      "epoch": 0.9552780459986132,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015405909731804003,
+      "loss": 4.2726,
+      "step": 9213
+    },
+    {
+      "epoch": 0.9553817340531012,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015404995870805656,
+      "loss": 4.25,
+      "step": 9214
+    },
+    {
+      "epoch": 0.9554854221075894,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015404081946034074,
+      "loss": 4.2884,
+      "step": 9215
+    },
+    {
+      "epoch": 0.9555891101620774,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015403167957500036,
+      "loss": 4.2743,
+      "step": 9216
+    },
+    {
+      "epoch": 0.9556927982165655,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015402253905214325,
+      "loss": 4.2711,
+      "step": 9217
+    },
+    {
+      "epoch": 0.9557964862710535,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015401339789187731,
+      "loss": 4.2759,
+      "step": 9218
+    },
+    {
+      "epoch": 0.9559001743255416,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001540042560943104,
+      "loss": 4.2727,
+      "step": 9219
+    },
+    {
+      "epoch": 0.9560038623800297,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015399511365955034,
+      "loss": 4.2178,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9561075504345178,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015398597058770497,
+      "loss": 4.2876,
+      "step": 9221
+    },
+    {
+      "epoch": 0.9562112384890058,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015397682687888224,
+      "loss": 4.274,
+      "step": 9222
+    },
+    {
+      "epoch": 0.9563149265434939,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015396768253319004,
+      "loss": 4.2447,
+      "step": 9223
+    },
+    {
+      "epoch": 0.9564186145979819,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015395853755073617,
+      "loss": 4.2406,
+      "step": 9224
+    },
+    {
+      "epoch": 0.9565223026524701,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001539493919316286,
+      "loss": 4.2687,
+      "step": 9225
+    },
+    {
+      "epoch": 0.9566259907069581,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015394024567597528,
+      "loss": 4.2737,
+      "step": 9226
+    },
+    {
+      "epoch": 0.9567296787614462,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015393109878388404,
+      "loss": 4.2817,
+      "step": 9227
+    },
+    {
+      "epoch": 0.9568333668159342,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015392195125546286,
+      "loss": 4.2752,
+      "step": 9228
+    },
+    {
+      "epoch": 0.9569370548704224,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015391280309081968,
+      "loss": 4.2593,
+      "step": 9229
+    },
+    {
+      "epoch": 0.9570407429249104,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015390365429006236,
+      "loss": 4.2702,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9571444309793985,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015389450485329894,
+      "loss": 4.2631,
+      "step": 9231
+    },
+    {
+      "epoch": 0.9572481190338865,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001538853547806373,
+      "loss": 4.2499,
+      "step": 9232
+    },
+    {
+      "epoch": 0.9573518070883746,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015387620407218545,
+      "loss": 4.2652,
+      "step": 9233
+    },
+    {
+      "epoch": 0.9574554951428627,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015386705272805133,
+      "loss": 4.2738,
+      "step": 9234
+    },
+    {
+      "epoch": 0.9575591831973508,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001538579007483429,
+      "loss": 4.2667,
+      "step": 9235
+    },
+    {
+      "epoch": 0.9576628712518388,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015384874813316826,
+      "loss": 4.2577,
+      "step": 9236
+    },
+    {
+      "epoch": 0.9577665593063269,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015383959488263525,
+      "loss": 4.2821,
+      "step": 9237
+    },
+    {
+      "epoch": 0.9578702473608149,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015383044099685192,
+      "loss": 4.3023,
+      "step": 9238
+    },
+    {
+      "epoch": 0.9579739354153031,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00015382128647592632,
+      "loss": 4.2862,
+      "step": 9239
+    },
+    {
+      "epoch": 0.9580776234697911,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00015381213131996643,
+      "loss": 4.2631,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9581813115242792,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00015380297552908026,
+      "loss": 4.2545,
+      "step": 9241
+    },
+    {
+      "epoch": 0.9582849995787672,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015379381910337583,
+      "loss": 4.2514,
+      "step": 9242
+    },
+    {
+      "epoch": 0.9583886876332554,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00015378466204296121,
+      "loss": 4.2767,
+      "step": 9243
+    },
+    {
+      "epoch": 0.9584923756877434,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015377550434794447,
+      "loss": 4.29,
+      "step": 9244
+    },
+    {
+      "epoch": 0.9585960637422315,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00015376634601843358,
+      "loss": 4.2716,
+      "step": 9245
+    },
+    {
+      "epoch": 0.9586997517967196,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015375718705453663,
+      "loss": 4.2961,
+      "step": 9246
+    },
+    {
+      "epoch": 0.9588034398512076,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00015374802745636172,
+      "loss": 4.2452,
+      "step": 9247
+    },
+    {
+      "epoch": 0.9589071279056958,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015373886722401693,
+      "loss": 4.2632,
+      "step": 9248
+    },
+    {
+      "epoch": 0.9590108159601838,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015372970635761025,
+      "loss": 4.2766,
+      "step": 9249
+    },
+    {
+      "epoch": 0.9591145040146719,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015372054485724988,
+      "loss": 4.2929,
+      "step": 9250
+    },
+    {
+      "epoch": 0.9592181920691599,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015371138272304383,
+      "loss": 4.2799,
+      "step": 9251
+    },
+    {
+      "epoch": 0.959321880123648,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015370221995510025,
+      "loss": 4.2487,
+      "step": 9252
+    },
+    {
+      "epoch": 0.9594255681781361,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001536930565535273,
+      "loss": 4.2844,
+      "step": 9253
+    },
+    {
+      "epoch": 0.9595292562326242,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015368389251843297,
+      "loss": 4.258,
+      "step": 9254
+    },
+    {
+      "epoch": 0.9596329442871122,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015367472784992545,
+      "loss": 4.2792,
+      "step": 9255
+    },
+    {
+      "epoch": 0.9597366323416003,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001536655625481129,
+      "loss": 4.2807,
+      "step": 9256
+    },
+    {
+      "epoch": 0.9598403203960884,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015365639661310342,
+      "loss": 4.2627,
+      "step": 9257
+    },
+    {
+      "epoch": 0.9599440084505765,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015364723004500518,
+      "loss": 4.2875,
+      "step": 9258
+    },
+    {
+      "epoch": 0.9600476965050645,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015363806284392637,
+      "loss": 4.2563,
+      "step": 9259
+    },
+    {
+      "epoch": 0.9601513845595526,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001536288950099751,
+      "loss": 4.266,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9602550726140406,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015361972654325952,
+      "loss": 4.2698,
+      "step": 9261
+    },
+    {
+      "epoch": 0.9603587606685288,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015361055744388789,
+      "loss": 4.2405,
+      "step": 9262
+    },
+    {
+      "epoch": 0.9604624487230168,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015360138771196833,
+      "loss": 4.2959,
+      "step": 9263
+    },
+    {
+      "epoch": 0.9605661367775049,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00015359221734760907,
+      "loss": 4.2977,
+      "step": 9264
+    },
+    {
+      "epoch": 0.9606698248319929,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001535830463509183,
+      "loss": 4.2779,
+      "step": 9265
+    },
+    {
+      "epoch": 0.960773512886481,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015357387472200417,
+      "loss": 4.2732,
+      "step": 9266
+    },
+    {
+      "epoch": 0.9608772009409691,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000153564702460975,
+      "loss": 4.266,
+      "step": 9267
+    },
+    {
+      "epoch": 0.9609808889954572,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015355552956793891,
+      "loss": 4.2521,
+      "step": 9268
+    },
+    {
+      "epoch": 0.9610845770499452,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015354635604300425,
+      "loss": 4.2541,
+      "step": 9269
+    },
+    {
+      "epoch": 0.9611882651044333,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015353718188627915,
+      "loss": 4.2916,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9612919531589214,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001535280070978719,
+      "loss": 4.2589,
+      "step": 9271
+    },
+    {
+      "epoch": 0.9613956412134095,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015351883167789073,
+      "loss": 4.2507,
+      "step": 9272
+    },
+    {
+      "epoch": 0.9614993292678975,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015350965562644397,
+      "loss": 4.2826,
+      "step": 9273
+    },
+    {
+      "epoch": 0.9616030173223856,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015350047894363978,
+      "loss": 4.2858,
+      "step": 9274
+    },
+    {
+      "epoch": 0.9617067053768736,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015349130162958653,
+      "loss": 4.3188,
+      "step": 9275
+    },
+    {
+      "epoch": 0.9618103934313618,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015348212368439246,
+      "loss": 4.2959,
+      "step": 9276
+    },
+    {
+      "epoch": 0.9619140814858498,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015347294510816584,
+      "loss": 4.2796,
+      "step": 9277
+    },
+    {
+      "epoch": 0.9620177695403379,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000153463765901015,
+      "loss": 4.2404,
+      "step": 9278
+    },
+    {
+      "epoch": 0.9621214575948259,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015345458606304827,
+      "loss": 4.2171,
+      "step": 9279
+    },
+    {
+      "epoch": 0.962225145649314,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001534454055943739,
+      "loss": 4.3025,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9623288337038021,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001534362244951002,
+      "loss": 4.2821,
+      "step": 9281
+    },
+    {
+      "epoch": 0.9624325217582902,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015342704276533558,
+      "loss": 4.2664,
+      "step": 9282
+    },
+    {
+      "epoch": 0.9625362098127782,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001534178604051883,
+      "loss": 4.2356,
+      "step": 9283
+    },
+    {
+      "epoch": 0.9626398978672663,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015340867741476676,
+      "loss": 4.2559,
+      "step": 9284
+    },
+    {
+      "epoch": 0.9627435859217544,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015339949379417927,
+      "loss": 4.2702,
+      "step": 9285
+    },
+    {
+      "epoch": 0.9628472739762425,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015339030954353424,
+      "loss": 4.2583,
+      "step": 9286
+    },
+    {
+      "epoch": 0.9629509620307305,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015338112466293992,
+      "loss": 4.2732,
+      "step": 9287
+    },
+    {
+      "epoch": 0.9630546500852186,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001533719391525048,
+      "loss": 4.2976,
+      "step": 9288
+    },
+    {
+      "epoch": 0.9631583381397066,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015336275301233723,
+      "loss": 4.2614,
+      "step": 9289
+    },
+    {
+      "epoch": 0.9632620261941948,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015335356624254556,
+      "loss": 4.2133,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9633657142486829,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001533443788432382,
+      "loss": 4.2687,
+      "step": 9291
+    },
+    {
+      "epoch": 0.9634694023031709,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015333519081452357,
+      "loss": 4.3103,
+      "step": 9292
+    },
+    {
+      "epoch": 0.963573090357659,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015332600215651004,
+      "loss": 4.2408,
+      "step": 9293
+    },
+    {
+      "epoch": 0.963676778412147,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001533168128693061,
+      "loss": 4.2903,
+      "step": 9294
+    },
+    {
+      "epoch": 0.9637804664666352,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015330762295302008,
+      "loss": 4.2581,
+      "step": 9295
+    },
+    {
+      "epoch": 0.9638841545211232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015329843240776049,
+      "loss": 4.2235,
+      "step": 9296
+    },
+    {
+      "epoch": 0.9639878425756113,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001532892412336357,
+      "loss": 4.2837,
+      "step": 9297
+    },
+    {
+      "epoch": 0.9640915306300993,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001532800494307542,
+      "loss": 4.261,
+      "step": 9298
+    },
+    {
+      "epoch": 0.9641952186845875,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015327085699922446,
+      "loss": 4.2868,
+      "step": 9299
+    },
+    {
+      "epoch": 0.9642989067390755,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001532616639391549,
+      "loss": 4.2561,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9644025947935636,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015325247025065403,
+      "loss": 4.2627,
+      "step": 9301
+    },
+    {
+      "epoch": 0.9645062828480516,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015324327593383027,
+      "loss": 4.2958,
+      "step": 9302
+    },
+    {
+      "epoch": 0.9646099709025397,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015323408098879217,
+      "loss": 4.2615,
+      "step": 9303
+    },
+    {
+      "epoch": 0.9647136589570278,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00015322488541564817,
+      "loss": 4.2625,
+      "step": 9304
+    },
+    {
+      "epoch": 0.9648173470115159,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015321568921450676,
+      "loss": 4.2513,
+      "step": 9305
+    },
+    {
+      "epoch": 0.9649210350660039,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001532064923854765,
+      "loss": 4.2893,
+      "step": 9306
+    },
+    {
+      "epoch": 0.965024723120492,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015319729492866584,
+      "loss": 4.2635,
+      "step": 9307
+    },
+    {
+      "epoch": 0.96512841117498,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00015318809684418337,
+      "loss": 4.2596,
+      "step": 9308
+    },
+    {
+      "epoch": 0.9652320992294682,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015317889813213753,
+      "loss": 4.2766,
+      "step": 9309
+    },
+    {
+      "epoch": 0.9653357872839562,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00015316969879263694,
+      "loss": 4.2442,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9654394753384443,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001531604988257901,
+      "loss": 4.2975,
+      "step": 9311
+    },
+    {
+      "epoch": 0.9655431633929323,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00015315129823170555,
+      "loss": 4.2573,
+      "step": 9312
+    },
+    {
+      "epoch": 0.9656468514474205,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001531420970104919,
+      "loss": 4.2364,
+      "step": 9313
+    },
+    {
+      "epoch": 0.9657505395019085,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00015313289516225766,
+      "loss": 4.2463,
+      "step": 9314
+    },
+    {
+      "epoch": 0.9658542275563966,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015312369268711144,
+      "loss": 4.2555,
+      "step": 9315
+    },
+    {
+      "epoch": 0.9659579156108846,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00015311448958516176,
+      "loss": 4.2938,
+      "step": 9316
+    },
+    {
+      "epoch": 0.9660616036653727,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015310528585651725,
+      "loss": 4.2862,
+      "step": 9317
+    },
+    {
+      "epoch": 0.9661652917198608,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00015309608150128654,
+      "loss": 4.2608,
+      "step": 9318
+    },
+    {
+      "epoch": 0.9662689797743489,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015308687651957817,
+      "loss": 4.2827,
+      "step": 9319
+    },
+    {
+      "epoch": 0.9663726678288369,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015307767091150078,
+      "loss": 4.31,
+      "step": 9320
+    },
+    {
+      "epoch": 0.966476355883325,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015306846467716295,
+      "loss": 4.2828,
+      "step": 9321
+    },
+    {
+      "epoch": 0.966580043937813,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015305925781667335,
+      "loss": 4.2935,
+      "step": 9322
+    },
+    {
+      "epoch": 0.9666837319923012,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015305005033014064,
+      "loss": 4.2706,
+      "step": 9323
+    },
+    {
+      "epoch": 0.9667874200467892,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015304084221767335,
+      "loss": 4.2371,
+      "step": 9324
+    },
+    {
+      "epoch": 0.9668911081012773,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001530316334793802,
+      "loss": 4.2946,
+      "step": 9325
+    },
+    {
+      "epoch": 0.9669947961557653,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015302242411536988,
+      "loss": 4.2853,
+      "step": 9326
+    },
+    {
+      "epoch": 0.9670984842102535,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015301321412575095,
+      "loss": 4.273,
+      "step": 9327
+    },
+    {
+      "epoch": 0.9672021722647415,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015300400351063215,
+      "loss": 4.2475,
+      "step": 9328
+    },
+    {
+      "epoch": 0.9673058603192296,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015299479227012214,
+      "loss": 4.3119,
+      "step": 9329
+    },
+    {
+      "epoch": 0.9674095483737176,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001529855804043296,
+      "loss": 4.2729,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9675132364282057,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001529763679133632,
+      "loss": 4.2845,
+      "step": 9331
+    },
+    {
+      "epoch": 0.9676169244826938,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001529671547973317,
+      "loss": 4.2634,
+      "step": 9332
+    },
+    {
+      "epoch": 0.9677206125371819,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00015295794105634372,
+      "loss": 4.2497,
+      "step": 9333
+    },
+    {
+      "epoch": 0.9678243005916699,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015294872669050798,
+      "loss": 4.2846,
+      "step": 9334
+    },
+    {
+      "epoch": 0.967927988646158,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015293951169993331,
+      "loss": 4.272,
+      "step": 9335
+    },
+    {
+      "epoch": 0.9680316767006462,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015293029608472834,
+      "loss": 4.2757,
+      "step": 9336
+    },
+    {
+      "epoch": 0.9681353647551342,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015292107984500182,
+      "loss": 4.2667,
+      "step": 9337
+    },
+    {
+      "epoch": 0.9682390528096223,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015291186298086248,
+      "loss": 4.2901,
+      "step": 9338
+    },
+    {
+      "epoch": 0.9683427408641103,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001529026454924191,
+      "loss": 4.284,
+      "step": 9339
+    },
+    {
+      "epoch": 0.9684464289185984,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015289342737978044,
+      "loss": 4.2652,
+      "step": 9340
+    },
+    {
+      "epoch": 0.9685501169730865,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015288420864305522,
+      "loss": 4.2741,
+      "step": 9341
+    },
+    {
+      "epoch": 0.9686538050275746,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015287498928235227,
+      "loss": 4.2619,
+      "step": 9342
+    },
+    {
+      "epoch": 0.9687574930820626,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001528657692977803,
+      "loss": 4.2337,
+      "step": 9343
+    },
+    {
+      "epoch": 0.9688611811365507,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015285654868944817,
+      "loss": 4.2751,
+      "step": 9344
+    },
+    {
+      "epoch": 0.9689648691910387,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001528473274574646,
+      "loss": 4.2606,
+      "step": 9345
+    },
+    {
+      "epoch": 0.9690685572455269,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015283810560193846,
+      "loss": 4.2429,
+      "step": 9346
+    },
+    {
+      "epoch": 0.9691722453000149,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001528288831229785,
+      "loss": 4.2399,
+      "step": 9347
+    },
+    {
+      "epoch": 0.969275933354503,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015281966002069362,
+      "loss": 4.284,
+      "step": 9348
+    },
+    {
+      "epoch": 0.969379621408991,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015281043629519252,
+      "loss": 4.2841,
+      "step": 9349
+    },
+    {
+      "epoch": 0.9694833094634792,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001528012119465841,
+      "loss": 4.292,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9695869975179672,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015279198697497722,
+      "loss": 4.2625,
+      "step": 9351
+    },
+    {
+      "epoch": 0.9696906855724553,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015278276138048068,
+      "loss": 4.2883,
+      "step": 9352
+    },
+    {
+      "epoch": 0.9697943736269433,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015277353516320337,
+      "loss": 4.2763,
+      "step": 9353
+    },
+    {
+      "epoch": 0.9698980616814314,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001527643083232541,
+      "loss": 4.2784,
+      "step": 9354
+    },
+    {
+      "epoch": 0.9700017497359195,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015275508086074176,
+      "loss": 4.235,
+      "step": 9355
+    },
+    {
+      "epoch": 0.9701054377904076,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001527458527757753,
+      "loss": 4.3079,
+      "step": 9356
+    },
+    {
+      "epoch": 0.9702091258448956,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001527366240684635,
+      "loss": 4.2871,
+      "step": 9357
+    },
+    {
+      "epoch": 0.9703128138993837,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001527273947389152,
+      "loss": 4.2377,
+      "step": 9358
+    },
+    {
+      "epoch": 0.9704165019538717,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015271816478723945,
+      "loss": 4.3026,
+      "step": 9359
+    },
+    {
+      "epoch": 0.9705201900083599,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015270893421354508,
+      "loss": 4.2523,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9706238780628479,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015269970301794102,
+      "loss": 4.2463,
+      "step": 9361
+    },
+    {
+      "epoch": 0.970727566117336,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001526904712005361,
+      "loss": 4.2646,
+      "step": 9362
+    },
+    {
+      "epoch": 0.970831254171824,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015268123876143938,
+      "loss": 4.2801,
+      "step": 9363
+    },
+    {
+      "epoch": 0.9709349422263122,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015267200570075973,
+      "loss": 4.2408,
+      "step": 9364
+    },
+    {
+      "epoch": 0.9710386302808002,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015266277201860608,
+      "loss": 4.1902,
+      "step": 9365
+    },
+    {
+      "epoch": 0.9711423183352883,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015265353771508737,
+      "loss": 4.2369,
+      "step": 9366
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015264430279031256,
+      "loss": 4.2613,
+      "step": 9367
+    },
+    {
+      "epoch": 0.9713496944442644,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015263506724439064,
+      "loss": 4.2568,
+      "step": 9368
+    },
+    {
+      "epoch": 0.9714533824987525,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015262583107743057,
+      "loss": 4.2521,
+      "step": 9369
+    },
+    {
+      "epoch": 0.9715570705532406,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00015261659428954133,
+      "loss": 4.2344,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9716607586077286,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001526073568808319,
+      "loss": 4.2492,
+      "step": 9371
+    },
+    {
+      "epoch": 0.9717644466622167,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001525981188514112,
+      "loss": 4.2453,
+      "step": 9372
+    },
+    {
+      "epoch": 0.9718681347167047,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001525888802013884,
+      "loss": 4.2665,
+      "step": 9373
+    },
+    {
+      "epoch": 0.9719718227711929,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015257964093087233,
+      "loss": 4.2541,
+      "step": 9374
+    },
+    {
+      "epoch": 0.9720755108256809,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015257040103997208,
+      "loss": 4.2422,
+      "step": 9375
+    },
+    {
+      "epoch": 0.972179198880169,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001525611605287967,
+      "loss": 4.2632,
+      "step": 9376
+    },
+    {
+      "epoch": 0.972282886934657,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015255191939745518,
+      "loss": 4.2566,
+      "step": 9377
+    },
+    {
+      "epoch": 0.9723865749891452,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015254267764605657,
+      "loss": 4.2434,
+      "step": 9378
+    },
+    {
+      "epoch": 0.9724902630436332,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015253343527470986,
+      "loss": 4.287,
+      "step": 9379
+    },
+    {
+      "epoch": 0.9725939510981213,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015252419228352414,
+      "loss": 4.2895,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9726976391526094,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015251494867260854,
+      "loss": 4.2692,
+      "step": 9381
+    },
+    {
+      "epoch": 0.9728013272070974,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.000152505704442072,
+      "loss": 4.3054,
+      "step": 9382
+    },
+    {
+      "epoch": 0.9729050152615856,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015249645959202366,
+      "loss": 4.2397,
+      "step": 9383
+    },
+    {
+      "epoch": 0.9730087033160736,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001524872141225726,
+      "loss": 4.2267,
+      "step": 9384
+    },
+    {
+      "epoch": 0.9731123913705617,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015247796803382789,
+      "loss": 4.2755,
+      "step": 9385
+    },
+    {
+      "epoch": 0.9732160794250497,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015246872132589862,
+      "loss": 4.2202,
+      "step": 9386
+    },
+    {
+      "epoch": 0.9733197674795379,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001524594739988939,
+      "loss": 4.2612,
+      "step": 9387
+    },
+    {
+      "epoch": 0.9734234555340259,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015245022605292285,
+      "loss": 4.2896,
+      "step": 9388
+    },
+    {
+      "epoch": 0.973527143588514,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015244097748809456,
+      "loss": 4.2922,
+      "step": 9389
+    },
+    {
+      "epoch": 0.973630831643002,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001524317283045182,
+      "loss": 4.2702,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9737345196974901,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015242247850230282,
+      "loss": 4.2298,
+      "step": 9391
+    },
+    {
+      "epoch": 0.9738382077519782,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015241322808155763,
+      "loss": 4.2585,
+      "step": 9392
+    },
+    {
+      "epoch": 0.9739418958064663,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015240397704239177,
+      "loss": 4.3028,
+      "step": 9393
+    },
+    {
+      "epoch": 0.9740455838609543,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015239472538491439,
+      "loss": 4.2511,
+      "step": 9394
+    },
+    {
+      "epoch": 0.9741492719154424,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015238547310923457,
+      "loss": 4.2455,
+      "step": 9395
+    },
+    {
+      "epoch": 0.9742529599699304,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015237622021546158,
+      "loss": 4.2601,
+      "step": 9396
+    },
+    {
+      "epoch": 0.9743566480244186,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015236696670370455,
+      "loss": 4.2497,
+      "step": 9397
+    },
+    {
+      "epoch": 0.9744603360789066,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015235771257407268,
+      "loss": 4.1968,
+      "step": 9398
+    },
+    {
+      "epoch": 0.9745640241333947,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001523484578266751,
+      "loss": 4.2809,
+      "step": 9399
+    },
+    {
+      "epoch": 0.9746677121878827,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015233920246162107,
+      "loss": 4.2493,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9747714002423709,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015232994647901982,
+      "loss": 4.2857,
+      "step": 9401
+    },
+    {
+      "epoch": 0.9748750882968589,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001523206898789805,
+      "loss": 4.2667,
+      "step": 9402
+    },
+    {
+      "epoch": 0.974978776351347,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015231143266161232,
+      "loss": 4.2546,
+      "step": 9403
+    },
+    {
+      "epoch": 0.975082464405835,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015230217482702454,
+      "loss": 4.2398,
+      "step": 9404
+    },
+    {
+      "epoch": 0.9751861524603231,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001522929163753264,
+      "loss": 4.2702,
+      "step": 9405
+    },
+    {
+      "epoch": 0.9752898405148112,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015228365730662712,
+      "loss": 4.2692,
+      "step": 9406
+    },
+    {
+      "epoch": 0.9753935285692993,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015227439762103594,
+      "loss": 4.3059,
+      "step": 9407
+    },
+    {
+      "epoch": 0.9754972166237873,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001522651373186621,
+      "loss": 4.2772,
+      "step": 9408
+    },
+    {
+      "epoch": 0.9756009046782754,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015225587639961492,
+      "loss": 4.2697,
+      "step": 9409
+    },
+    {
+      "epoch": 0.9757045927327634,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001522466148640036,
+      "loss": 4.2494,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9758082807872516,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015223735271193748,
+      "loss": 4.2737,
+      "step": 9411
+    },
+    {
+      "epoch": 0.9759119688417396,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015222808994352582,
+      "loss": 4.2676,
+      "step": 9412
+    },
+    {
+      "epoch": 0.9760156568962277,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001522188265588779,
+      "loss": 4.2648,
+      "step": 9413
+    },
+    {
+      "epoch": 0.9761193449507157,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015220956255810304,
+      "loss": 4.2446,
+      "step": 9414
+    },
+    {
+      "epoch": 0.9762230330052039,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001522002979413105,
+      "loss": 4.2817,
+      "step": 9415
+    },
+    {
+      "epoch": 0.9763267210596919,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015219103270860965,
+      "loss": 4.2879,
+      "step": 9416
+    },
+    {
+      "epoch": 0.97643040911418,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015218176686010974,
+      "loss": 4.2554,
+      "step": 9417
+    },
+    {
+      "epoch": 0.976534097168668,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015217250039592018,
+      "loss": 4.2662,
+      "step": 9418
+    },
+    {
+      "epoch": 0.9766377852231561,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015216323331615023,
+      "loss": 4.271,
+      "step": 9419
+    },
+    {
+      "epoch": 0.9767414732776442,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001521539656209093,
+      "loss": 4.261,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9768451613321323,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015214469731030666,
+      "loss": 4.3136,
+      "step": 9421
+    },
+    {
+      "epoch": 0.9769488493866203,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015213542838445176,
+      "loss": 4.2493,
+      "step": 9422
+    },
+    {
+      "epoch": 0.9770525374411084,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001521261588434539,
+      "loss": 4.2347,
+      "step": 9423
+    },
+    {
+      "epoch": 0.9771562254955966,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015211688868742247,
+      "loss": 4.2623,
+      "step": 9424
+    },
+    {
+      "epoch": 0.9772599135500846,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015210761791646684,
+      "loss": 4.2435,
+      "step": 9425
+    },
+    {
+      "epoch": 0.9773636016045727,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001520983465306964,
+      "loss": 4.2485,
+      "step": 9426
+    },
+    {
+      "epoch": 0.9774672896590607,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001520890745302205,
+      "loss": 4.2038,
+      "step": 9427
+    },
+    {
+      "epoch": 0.9775709777135488,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015207980191514863,
+      "loss": 4.2652,
+      "step": 9428
+    },
+    {
+      "epoch": 0.9776746657680369,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015207052868559012,
+      "loss": 4.2598,
+      "step": 9429
+    },
+    {
+      "epoch": 0.977778353822525,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015206125484165445,
+      "loss": 4.2875,
+      "step": 9430
+    },
+    {
+      "epoch": 0.977882041877013,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.000152051980383451,
+      "loss": 4.2917,
+      "step": 9431
+    },
+    {
+      "epoch": 0.9779857299315011,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015204270531108915,
+      "loss": 4.2378,
+      "step": 9432
+    },
+    {
+      "epoch": 0.9780894179859891,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015203342962467843,
+      "loss": 4.2538,
+      "step": 9433
+    },
+    {
+      "epoch": 0.9781931060404773,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015202415332432826,
+      "loss": 4.2516,
+      "step": 9434
+    },
+    {
+      "epoch": 0.9782967940949653,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015201487641014803,
+      "loss": 4.2188,
+      "step": 9435
+    },
+    {
+      "epoch": 0.9784004821494534,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015200559888224727,
+      "loss": 4.2706,
+      "step": 9436
+    },
+    {
+      "epoch": 0.9785041702039414,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015199632074073538,
+      "loss": 4.294,
+      "step": 9437
+    },
+    {
+      "epoch": 0.9786078582584296,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015198704198572189,
+      "loss": 4.2657,
+      "step": 9438
+    },
+    {
+      "epoch": 0.9787115463129176,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015197776261731627,
+      "loss": 4.3236,
+      "step": 9439
+    },
+    {
+      "epoch": 0.9788152343674057,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.000151968482635628,
+      "loss": 4.2368,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9789189224218937,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015195920204076654,
+      "loss": 4.2963,
+      "step": 9441
+    },
+    {
+      "epoch": 0.9790226104763818,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015194992083284142,
+      "loss": 4.2503,
+      "step": 9442
+    },
+    {
+      "epoch": 0.9791262985308699,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015194063901196217,
+      "loss": 4.2708,
+      "step": 9443
+    },
+    {
+      "epoch": 0.979229986585358,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015193135657823827,
+      "loss": 4.2709,
+      "step": 9444
+    },
+    {
+      "epoch": 0.979333674639846,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015192207353177922,
+      "loss": 4.2549,
+      "step": 9445
+    },
+    {
+      "epoch": 0.9794373626943341,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015191278987269463,
+      "loss": 4.2305,
+      "step": 9446
+    },
+    {
+      "epoch": 0.9795410507488221,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00015190350560109398,
+      "loss": 4.2507,
+      "step": 9447
+    },
+    {
+      "epoch": 0.9796447388033103,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015189422071708685,
+      "loss": 4.2778,
+      "step": 9448
+    },
+    {
+      "epoch": 0.9797484268577983,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00015188493522078276,
+      "loss": 4.2954,
+      "step": 9449
+    },
+    {
+      "epoch": 0.9798521149122864,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015187564911229125,
+      "loss": 4.298,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9799558029667744,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00015186636239172198,
+      "loss": 4.2874,
+      "step": 9451
+    },
+    {
+      "epoch": 0.9800594910212626,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015185707505918438,
+      "loss": 4.2759,
+      "step": 9452
+    },
+    {
+      "epoch": 0.9801631790757506,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00015184778711478818,
+      "loss": 4.3086,
+      "step": 9453
+    },
+    {
+      "epoch": 0.9802668671302387,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015183849855864286,
+      "loss": 4.2788,
+      "step": 9454
+    },
+    {
+      "epoch": 0.9803705551847267,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00015182920939085806,
+      "loss": 4.2358,
+      "step": 9455
+    },
+    {
+      "epoch": 0.9804742432392148,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001518199196115434,
+      "loss": 4.2496,
+      "step": 9456
+    },
+    {
+      "epoch": 0.9805779312937029,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001518106292208084,
+      "loss": 4.2617,
+      "step": 9457
+    },
+    {
+      "epoch": 0.980681619348191,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015180133821876278,
+      "loss": 4.2754,
+      "step": 9458
+    },
+    {
+      "epoch": 0.980785307402679,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015179204660551614,
+      "loss": 4.2655,
+      "step": 9459
+    },
+    {
+      "epoch": 0.9808889954571671,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015178275438117807,
+      "loss": 4.2661,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9809926835116551,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015177346154585824,
+      "loss": 4.2846,
+      "step": 9461
+    },
+    {
+      "epoch": 0.9810963715661433,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001517641680996663,
+      "loss": 4.2724,
+      "step": 9462
+    },
+    {
+      "epoch": 0.9812000596206313,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001517548740427119,
+      "loss": 4.2302,
+      "step": 9463
+    },
+    {
+      "epoch": 0.9813037476751194,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015174557937510467,
+      "loss": 4.2434,
+      "step": 9464
+    },
+    {
+      "epoch": 0.9814074357296074,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015173628409695432,
+      "loss": 4.257,
+      "step": 9465
+    },
+    {
+      "epoch": 0.9815111237840956,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001517269882083705,
+      "loss": 4.2477,
+      "step": 9466
+    },
+    {
+      "epoch": 0.9816148118385836,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015171769170946287,
+      "loss": 4.2774,
+      "step": 9467
+    },
+    {
+      "epoch": 0.9817184998930717,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00015170839460034122,
+      "loss": 4.2568,
+      "step": 9468
+    },
+    {
+      "epoch": 0.9818221879475598,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015169909688111512,
+      "loss": 4.2727,
+      "step": 9469
+    },
+    {
+      "epoch": 0.9819258760020478,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001516897985518943,
+      "loss": 4.2676,
+      "step": 9470
+    },
+    {
+      "epoch": 0.982029564056536,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015168049961278854,
+      "loss": 4.2741,
+      "step": 9471
+    },
+    {
+      "epoch": 0.982133252111024,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001516712000639075,
+      "loss": 4.2287,
+      "step": 9472
+    },
+    {
+      "epoch": 0.9822369401655121,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001516618999053609,
+      "loss": 4.2734,
+      "step": 9473
+    },
+    {
+      "epoch": 0.9823406282200001,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001516525991372585,
+      "loss": 4.2906,
+      "step": 9474
+    },
+    {
+      "epoch": 0.9824443162744882,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015164329775971003,
+      "loss": 4.2236,
+      "step": 9475
+    },
+    {
+      "epoch": 0.9825480043289763,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015163399577282526,
+      "loss": 4.2571,
+      "step": 9476
+    },
+    {
+      "epoch": 0.9826516923834644,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015162469317671392,
+      "loss": 4.2449,
+      "step": 9477
+    },
+    {
+      "epoch": 0.9827553804379524,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015161538997148573,
+      "loss": 4.2457,
+      "step": 9478
+    },
+    {
+      "epoch": 0.9828590684924405,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015160608615725054,
+      "loss": 4.2468,
+      "step": 9479
+    },
+    {
+      "epoch": 0.9829627565469286,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015159678173411805,
+      "loss": 4.2502,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9830664446014167,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001515874767021981,
+      "loss": 4.2675,
+      "step": 9481
+    },
+    {
+      "epoch": 0.9831701326559047,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00015157817106160044,
+      "loss": 4.3162,
+      "step": 9482
+    },
+    {
+      "epoch": 0.9832738207103928,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001515688648124349,
+      "loss": 4.2683,
+      "step": 9483
+    },
+    {
+      "epoch": 0.9833775087648808,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015155955795481124,
+      "loss": 4.2951,
+      "step": 9484
+    },
+    {
+      "epoch": 0.983481196819369,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001515502504888393,
+      "loss": 4.2818,
+      "step": 9485
+    },
+    {
+      "epoch": 0.983584884873857,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001515409424146289,
+      "loss": 4.3042,
+      "step": 9486
+    },
+    {
+      "epoch": 0.9836885729283451,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015153163373228987,
+      "loss": 4.2414,
+      "step": 9487
+    },
+    {
+      "epoch": 0.9837922609828331,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015152232444193202,
+      "loss": 4.2496,
+      "step": 9488
+    },
+    {
+      "epoch": 0.9838959490373212,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015151301454366522,
+      "loss": 4.2784,
+      "step": 9489
+    },
+    {
+      "epoch": 0.9839996370918093,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015150370403759929,
+      "loss": 4.2508,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9841033251462974,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001514943929238441,
+      "loss": 4.2928,
+      "step": 9491
+    },
+    {
+      "epoch": 0.9842070132007854,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015148508120250948,
+      "loss": 4.2469,
+      "step": 9492
+    },
+    {
+      "epoch": 0.9843107012552735,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015147576887370535,
+      "loss": 4.2814,
+      "step": 9493
+    },
+    {
+      "epoch": 0.9844143893097616,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015146645593754155,
+      "loss": 4.2694,
+      "step": 9494
+    },
+    {
+      "epoch": 0.9845180773642497,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.000151457142394128,
+      "loss": 4.2363,
+      "step": 9495
+    },
+    {
+      "epoch": 0.9846217654187377,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001514478282435745,
+      "loss": 4.2998,
+      "step": 9496
+    },
+    {
+      "epoch": 0.9847254534732258,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015143851348599108,
+      "loss": 4.242,
+      "step": 9497
+    },
+    {
+      "epoch": 0.9848291415277138,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015142919812148752,
+      "loss": 4.2483,
+      "step": 9498
+    },
+    {
+      "epoch": 0.984932829582202,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001514198821501738,
+      "loss": 4.2775,
+      "step": 9499
+    },
+    {
+      "epoch": 0.98503651763669,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015141056557215984,
+      "loss": 4.2735,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9851402056911781,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015140124838755554,
+      "loss": 4.2642,
+      "step": 9501
+    },
+    {
+      "epoch": 0.9852438937456661,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015139193059647086,
+      "loss": 4.2407,
+      "step": 9502
+    },
+    {
+      "epoch": 0.9853475818001542,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001513826121990157,
+      "loss": 4.3053,
+      "step": 9503
+    },
+    {
+      "epoch": 0.9854512698546423,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015137329319530002,
+      "loss": 4.2636,
+      "step": 9504
+    },
+    {
+      "epoch": 0.9855549579091304,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015136397358543382,
+      "loss": 4.2455,
+      "step": 9505
+    },
+    {
+      "epoch": 0.9856586459636184,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015135465336952702,
+      "loss": 4.2578,
+      "step": 9506
+    },
+    {
+      "epoch": 0.9857623340181065,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015134533254768958,
+      "loss": 4.2418,
+      "step": 9507
+    },
+    {
+      "epoch": 0.9858660220725945,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015133601112003145,
+      "loss": 4.2535,
+      "step": 9508
+    },
+    {
+      "epoch": 0.9859697101270827,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001513266890866627,
+      "loss": 4.2542,
+      "step": 9509
+    },
+    {
+      "epoch": 0.9860733981815707,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015131736644769328,
+      "loss": 4.308,
+      "step": 9510
+    },
+    {
+      "epoch": 0.9861770862360588,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015130804320323318,
+      "loss": 4.2822,
+      "step": 9511
+    },
+    {
+      "epoch": 0.9862807742905468,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015129871935339238,
+      "loss": 4.2731,
+      "step": 9512
+    },
+    {
+      "epoch": 0.986384462345035,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001512893948982809,
+      "loss": 4.2935,
+      "step": 9513
+    },
+    {
+      "epoch": 0.9864881503995231,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015128006983800883,
+      "loss": 4.2167,
+      "step": 9514
+    },
+    {
+      "epoch": 0.9865918384540111,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015127074417268613,
+      "loss": 4.2594,
+      "step": 9515
+    },
+    {
+      "epoch": 0.9866955265084992,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015126141790242282,
+      "loss": 4.2714,
+      "step": 9516
+    },
+    {
+      "epoch": 0.9867992145629872,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015125209102732895,
+      "loss": 4.2274,
+      "step": 9517
+    },
+    {
+      "epoch": 0.9869029026174754,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001512427635475146,
+      "loss": 4.2527,
+      "step": 9518
+    },
+    {
+      "epoch": 0.9870065906719634,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015123343546308984,
+      "loss": 4.2661,
+      "step": 9519
+    },
+    {
+      "epoch": 0.9871102787264515,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015122410677416467,
+      "loss": 4.2873,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9872139667809395,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00015121477748084917,
+      "loss": 4.259,
+      "step": 9521
+    },
+    {
+      "epoch": 0.9873176548354277,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015120544758325346,
+      "loss": 4.2133,
+      "step": 9522
+    },
+    {
+      "epoch": 0.9874213428899157,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001511961170814876,
+      "loss": 4.252,
+      "step": 9523
+    },
+    {
+      "epoch": 0.9875250309444038,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015118678597566163,
+      "loss": 4.2915,
+      "step": 9524
+    },
+    {
+      "epoch": 0.9876287189988918,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001511774542658857,
+      "loss": 4.2505,
+      "step": 9525
+    },
+    {
+      "epoch": 0.98773240705338,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015116812195226995,
+      "loss": 4.2576,
+      "step": 9526
+    },
+    {
+      "epoch": 0.987836095107868,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015115878903492443,
+      "loss": 4.3013,
+      "step": 9527
+    },
+    {
+      "epoch": 0.9879397831623561,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001511494555139593,
+      "loss": 4.2584,
+      "step": 9528
+    },
+    {
+      "epoch": 0.9880434712168441,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001511401213894846,
+      "loss": 4.2957,
+      "step": 9529
+    },
+    {
+      "epoch": 0.9881471592713322,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015113078666161055,
+      "loss": 4.2904,
+      "step": 9530
+    },
+    {
+      "epoch": 0.9882508473258202,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015112145133044732,
+      "loss": 4.2956,
+      "step": 9531
+    },
+    {
+      "epoch": 0.9883545353803084,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015111211539610496,
+      "loss": 4.2502,
+      "step": 9532
+    },
+    {
+      "epoch": 0.9884582234347964,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015110277885869362,
+      "loss": 4.2664,
+      "step": 9533
+    },
+    {
+      "epoch": 0.9885619114892845,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015109344171832355,
+      "loss": 4.2533,
+      "step": 9534
+    },
+    {
+      "epoch": 0.9886655995437725,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015108410397510487,
+      "loss": 4.2616,
+      "step": 9535
+    },
+    {
+      "epoch": 0.9887692875982607,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001510747656291478,
+      "loss": 4.2467,
+      "step": 9536
+    },
+    {
+      "epoch": 0.9888729756527487,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015106542668056241,
+      "loss": 4.2971,
+      "step": 9537
+    },
+    {
+      "epoch": 0.9889766637072368,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015105608712945903,
+      "loss": 4.2468,
+      "step": 9538
+    },
+    {
+      "epoch": 0.9890803517617248,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015104674697594776,
+      "loss": 4.2745,
+      "step": 9539
+    },
+    {
+      "epoch": 0.989184039816213,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00015103740622013882,
+      "loss": 4.2288,
+      "step": 9540
+    },
+    {
+      "epoch": 0.989287727870701,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001510280648621425,
+      "loss": 4.3209,
+      "step": 9541
+    },
+    {
+      "epoch": 0.9893914159251891,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015101872290206888,
+      "loss": 4.286,
+      "step": 9542
+    },
+    {
+      "epoch": 0.9894951039796771,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015100938034002826,
+      "loss": 4.2625,
+      "step": 9543
+    },
+    {
+      "epoch": 0.9895987920341652,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015100003717613095,
+      "loss": 4.249,
+      "step": 9544
+    },
+    {
+      "epoch": 0.9897024800886532,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015099069341048703,
+      "loss": 4.2947,
+      "step": 9545
+    },
+    {
+      "epoch": 0.9898061681431414,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015098134904320686,
+      "loss": 4.2678,
+      "step": 9546
+    },
+    {
+      "epoch": 0.9899098561976294,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015097200407440065,
+      "loss": 4.2481,
+      "step": 9547
+    },
+    {
+      "epoch": 0.9900135442521175,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001509626585041787,
+      "loss": 4.2628,
+      "step": 9548
+    },
+    {
+      "epoch": 0.9901172323066055,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015095331233265123,
+      "loss": 4.3012,
+      "step": 9549
+    },
+    {
+      "epoch": 0.9902209203610937,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015094396555992852,
+      "loss": 4.2533,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9903246084155817,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015093461818612087,
+      "loss": 4.263,
+      "step": 9551
+    },
+    {
+      "epoch": 0.9904282964700698,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015092527021133857,
+      "loss": 4.2751,
+      "step": 9552
+    },
+    {
+      "epoch": 0.9905319845245578,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015091592163569195,
+      "loss": 4.2127,
+      "step": 9553
+    },
+    {
+      "epoch": 0.990635672579046,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015090657245929123,
+      "loss": 4.2407,
+      "step": 9554
+    },
+    {
+      "epoch": 0.990739360633534,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015089722268224676,
+      "loss": 4.247,
+      "step": 9555
+    },
+    {
+      "epoch": 0.9908430486880221,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001508878723046689,
+      "loss": 4.2134,
+      "step": 9556
+    },
+    {
+      "epoch": 0.9909467367425101,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001508785213266679,
+      "loss": 4.2577,
+      "step": 9557
+    },
+    {
+      "epoch": 0.9910504247969982,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015086916974835413,
+      "loss": 4.2612,
+      "step": 9558
+    },
+    {
+      "epoch": 0.9911541128514864,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015085981756983796,
+      "loss": 4.259,
+      "step": 9559
+    },
+    {
+      "epoch": 0.9912578009059744,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001508504647912297,
+      "loss": 4.2905,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9913614889604625,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001508411114126397,
+      "loss": 4.2423,
+      "step": 9561
+    },
+    {
+      "epoch": 0.9914651770149505,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015083175743417832,
+      "loss": 4.2527,
+      "step": 9562
+    },
+    {
+      "epoch": 0.9915688650694386,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015082240285595592,
+      "loss": 4.2274,
+      "step": 9563
+    },
+    {
+      "epoch": 0.9916725531239267,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015081304767808294,
+      "loss": 4.2877,
+      "step": 9564
+    },
+    {
+      "epoch": 0.9917762411784148,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015080369190066965,
+      "loss": 4.2442,
+      "step": 9565
+    },
+    {
+      "epoch": 0.9918799292329028,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015079433552382654,
+      "loss": 4.2374,
+      "step": 9566
+    },
+    {
+      "epoch": 0.9919836172873909,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015078497854766393,
+      "loss": 4.2583,
+      "step": 9567
+    },
+    {
+      "epoch": 0.9920873053418789,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015077562097229227,
+      "loss": 4.2587,
+      "step": 9568
+    },
+    {
+      "epoch": 0.9921909933963671,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015076626279782196,
+      "loss": 4.2521,
+      "step": 9569
+    },
+    {
+      "epoch": 0.9922946814508551,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001507569040243634,
+      "loss": 4.2325,
+      "step": 9570
+    },
+    {
+      "epoch": 0.9923983695053432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015074754465202702,
+      "loss": 4.2582,
+      "step": 9571
+    },
+    {
+      "epoch": 0.9925020575598312,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015073818468092326,
+      "loss": 4.2995,
+      "step": 9572
+    },
+    {
+      "epoch": 0.9926057456143194,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015072882411116257,
+      "loss": 4.2835,
+      "step": 9573
+    },
+    {
+      "epoch": 0.9927094336688074,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015071946294285536,
+      "loss": 4.2457,
+      "step": 9574
+    },
+    {
+      "epoch": 0.9928131217232955,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001507101011761121,
+      "loss": 4.3043,
+      "step": 9575
+    },
+    {
+      "epoch": 0.9929168097777835,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015070073881104325,
+      "loss": 4.3035,
+      "step": 9576
+    },
+    {
+      "epoch": 0.9930204978322716,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015069137584775933,
+      "loss": 4.2266,
+      "step": 9577
+    },
+    {
+      "epoch": 0.9931241858867597,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015068201228637072,
+      "loss": 4.2521,
+      "step": 9578
+    },
+    {
+      "epoch": 0.9932278739412478,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001506726481269879,
+      "loss": 4.2694,
+      "step": 9579
+    },
+    {
+      "epoch": 0.9933315619957358,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015066328336972142,
+      "loss": 4.2715,
+      "step": 9580
+    },
+    {
+      "epoch": 0.9934352500502239,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015065391801468176,
+      "loss": 4.2333,
+      "step": 9581
+    },
+    {
+      "epoch": 0.9935389381047119,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015064455206197937,
+      "loss": 4.285,
+      "step": 9582
+    },
+    {
+      "epoch": 0.9936426261592001,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015063518551172486,
+      "loss": 4.2574,
+      "step": 9583
+    },
+    {
+      "epoch": 0.9937463142136881,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015062581836402866,
+      "loss": 4.2694,
+      "step": 9584
+    },
+    {
+      "epoch": 0.9938500022681762,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015061645061900132,
+      "loss": 4.2823,
+      "step": 9585
+    },
+    {
+      "epoch": 0.9939536903226642,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015060708227675337,
+      "loss": 4.2766,
+      "step": 9586
+    },
+    {
+      "epoch": 0.9940573783771524,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001505977133373953,
+      "loss": 4.2578,
+      "step": 9587
+    },
+    {
+      "epoch": 0.9941610664316404,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015058834380103772,
+      "loss": 4.2511,
+      "step": 9588
+    },
+    {
+      "epoch": 0.9942647544861285,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001505789736677912,
+      "loss": 4.2933,
+      "step": 9589
+    },
+    {
+      "epoch": 0.9943684425406165,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001505696029377662,
+      "loss": 4.248,
+      "step": 9590
+    },
+    {
+      "epoch": 0.9944721305951046,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001505602316110734,
+      "loss": 4.2504,
+      "step": 9591
+    },
+    {
+      "epoch": 0.9945758186495927,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015055085968782326,
+      "loss": 4.2362,
+      "step": 9592
+    },
+    {
+      "epoch": 0.9946795067040808,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015054148716812642,
+      "loss": 4.2888,
+      "step": 9593
+    },
+    {
+      "epoch": 0.9947831947585688,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015053211405209347,
+      "loss": 4.2504,
+      "step": 9594
+    },
+    {
+      "epoch": 0.9948868828130569,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015052274033983495,
+      "loss": 4.2843,
+      "step": 9595
+    },
+    {
+      "epoch": 0.9949905708675449,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015051336603146153,
+      "loss": 4.277,
+      "step": 9596
+    },
+    {
+      "epoch": 0.9950942589220331,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001505039911270838,
+      "loss": 4.2381,
+      "step": 9597
+    },
+    {
+      "epoch": 0.9951979469765211,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015049461562681232,
+      "loss": 4.2435,
+      "step": 9598
+    },
+    {
+      "epoch": 0.9953016350310092,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015048523953075778,
+      "loss": 4.2282,
+      "step": 9599
+    },
+    {
+      "epoch": 0.9954053230854972,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015047586283903076,
+      "loss": 4.2414,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9955090111399854,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015046648555174194,
+      "loss": 4.3018,
+      "step": 9601
+    },
+    {
+      "epoch": 0.9956126991944734,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015045710766900194,
+      "loss": 4.2932,
+      "step": 9602
+    },
+    {
+      "epoch": 0.9957163872489615,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015044772919092138,
+      "loss": 4.266,
+      "step": 9603
+    },
+    {
+      "epoch": 0.9958200753034496,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015043835011761095,
+      "loss": 4.3192,
+      "step": 9604
+    },
+    {
+      "epoch": 0.9959237633579376,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001504289704491813,
+      "loss": 4.316,
+      "step": 9605
+    },
+    {
+      "epoch": 0.9960274514124258,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015041959018574313,
+      "loss": 4.2825,
+      "step": 9606
+    },
+    {
+      "epoch": 0.9961311394669138,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015041020932740707,
+      "loss": 4.2815,
+      "step": 9607
+    },
+    {
+      "epoch": 0.9962348275214019,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001504008278742838,
+      "loss": 4.261,
+      "step": 9608
+    },
+    {
+      "epoch": 0.9963385155758899,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001503914458264841,
+      "loss": 4.2402,
+      "step": 9609
+    },
+    {
+      "epoch": 0.996442203630378,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015038206318411854,
+      "loss": 4.2206,
+      "step": 9610
+    },
+    {
+      "epoch": 0.9965458916848661,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015037267994729795,
+      "loss": 4.295,
+      "step": 9611
+    },
+    {
+      "epoch": 0.9966495797393542,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001503632961161329,
+      "loss": 4.2348,
+      "step": 9612
+    },
+    {
+      "epoch": 0.9967532677938422,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015035391169073426,
+      "loss": 4.2728,
+      "step": 9613
+    },
+    {
+      "epoch": 0.9968569558483303,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00015034452667121267,
+      "loss": 4.2732,
+      "step": 9614
+    },
+    {
+      "epoch": 0.9969606439028184,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001503351410576789,
+      "loss": 4.2078,
+      "step": 9615
+    },
+    {
+      "epoch": 0.9970643319573065,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015032575485024365,
+      "loss": 4.2402,
+      "step": 9616
+    },
+    {
+      "epoch": 0.9971680200117945,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001503163680490177,
+      "loss": 4.2576,
+      "step": 9617
+    },
+    {
+      "epoch": 0.9972717080662826,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015030698065411177,
+      "loss": 4.2684,
+      "step": 9618
+    },
+    {
+      "epoch": 0.9973753961207706,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015029759266563667,
+      "loss": 4.2699,
+      "step": 9619
+    },
+    {
+      "epoch": 0.9974790841752588,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015028820408370314,
+      "loss": 4.2519,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9975827722297468,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015027881490842194,
+      "loss": 4.2527,
+      "step": 9621
+    },
+    {
+      "epoch": 0.9976864602842349,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001502694251399039,
+      "loss": 4.2643,
+      "step": 9622
+    },
+    {
+      "epoch": 0.9977901483387229,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015026003477825977,
+      "loss": 4.2781,
+      "step": 9623
+    },
+    {
+      "epoch": 0.997893836393211,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015025064382360036,
+      "loss": 4.2317,
+      "step": 9624
+    },
+    {
+      "epoch": 0.9979975244476991,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015024125227603645,
+      "loss": 4.2581,
+      "step": 9625
+    },
+    {
+      "epoch": 0.9981012125021872,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015023186013567887,
+      "loss": 4.2255,
+      "step": 9626
+    },
+    {
+      "epoch": 0.9982049005566752,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015022246740263844,
+      "loss": 4.225,
+      "step": 9627
+    },
+    {
+      "epoch": 0.9983085886111633,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015021307407702604,
+      "loss": 4.2266,
+      "step": 9628
+    },
+    {
+      "epoch": 0.9984122766656514,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001502036801589524,
+      "loss": 4.2333,
+      "step": 9629
+    },
+    {
+      "epoch": 0.9985159647201395,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015019428564852838,
+      "loss": 4.248,
+      "step": 9630
+    },
+    {
+      "epoch": 0.9986196527746275,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001501848905458649,
+      "loss": 4.2952,
+      "step": 9631
+    },
+    {
+      "epoch": 0.9987233408291156,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001501754948510727,
+      "loss": 4.2362,
+      "step": 9632
+    },
+    {
+      "epoch": 0.9988270288836036,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001501660985642627,
+      "loss": 4.2598,
+      "step": 9633
+    },
+    {
+      "epoch": 0.9989307169380918,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001501567016855458,
+      "loss": 4.2798,
+      "step": 9634
+    },
+    {
+      "epoch": 0.9990344049925798,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015014730421503286,
+      "loss": 4.2741,
+      "step": 9635
+    },
+    {
+      "epoch": 0.9991380930470679,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015013790615283468,
+      "loss": 4.2794,
+      "step": 9636
+    },
+    {
+      "epoch": 0.9992417811015559,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015012850749906222,
+      "loss": 4.2997,
+      "step": 9637
+    },
+    {
+      "epoch": 0.999345469156044,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015011910825382638,
+      "loss": 4.2481,
+      "step": 9638
+    },
+    {
+      "epoch": 0.9994491572105321,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015010970841723803,
+      "loss": 4.2476,
+      "step": 9639
+    },
+    {
+      "epoch": 0.9995528452650202,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015010030798940808,
+      "loss": 4.2513,
+      "step": 9640
+    },
+    {
+      "epoch": 0.9996565333195082,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015009090697044747,
+      "loss": 4.2688,
+      "step": 9641
+    },
+    {
+      "epoch": 0.9997602213739963,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001500815053604671,
+      "loss": 4.2505,
+      "step": 9642
+    },
+    {
+      "epoch": 0.9998639094284844,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015007210315957792,
+      "loss": 4.2664,
+      "step": 9643
+    },
+    {
+      "epoch": 0.9999675974829725,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015006270036789084,
+      "loss": 4.2492,
+      "step": 9644
+    },
+    {
+      "epoch": 0.9999675974829725,
+      "eval_loss": 4.279594898223877,
+      "eval_runtime": 0.4479,
+      "eval_samples_per_second": 332.645,
+      "eval_steps_per_second": 15.628,
+      "step": 9644
     }
   ],
   "logging_steps": 1,
@@ -16907,7 +67562,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 2411,
-  "total_flos": 3.4729528980175585e+18,
+  "total_flos": 1.3911233195677843e+19,
   "train_batch_size": 3,
   "trial_name": null,
   "trial_params": null