diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.100037750094375,
+  "eval_steps": 500,
+  "global_step": 40000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003775009437523594,
+      "grad_norm": 12.114919662475586,
+      "learning_rate": 4.5e-07,
+      "loss": 4.3735,
+      "step": 10
+    },
+    {
+      "epoch": 0.007550018875047188,
+      "grad_norm": 9.25322437286377,
+      "learning_rate": 9.5e-07,
+      "loss": 4.2737,
+      "step": 20
+    },
+    {
+      "epoch": 0.011325028312570781,
+      "grad_norm": 11.06075668334961,
+      "learning_rate": 1.45e-06,
+      "loss": 4.3709,
+      "step": 30
+    },
+    {
+      "epoch": 0.015100037750094376,
+      "grad_norm": 11.528477668762207,
+      "learning_rate": 1.95e-06,
+      "loss": 4.5857,
+      "step": 40
+    },
+    {
+      "epoch": 0.01887504718761797,
+      "grad_norm": 9.225029945373535,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 4.1594,
+      "step": 50
+    },
+    {
+      "epoch": 0.022650056625141562,
+      "grad_norm": 9.194082260131836,
+      "learning_rate": 2.95e-06,
+      "loss": 3.9774,
+      "step": 60
+    },
+    {
+      "epoch": 0.02642506606266516,
+      "grad_norm": 9.647115707397461,
+      "learning_rate": 3.4500000000000004e-06,
+      "loss": 3.7629,
+      "step": 70
+    },
+    {
+      "epoch": 0.03020007550018875,
+      "grad_norm": 7.582672595977783,
+      "learning_rate": 3.95e-06,
+      "loss": 3.3272,
+      "step": 80
+    },
+    {
+      "epoch": 0.03397508493771234,
+      "grad_norm": 6.46481466293335,
+      "learning_rate": 4.45e-06,
+      "loss": 2.9437,
+      "step": 90
+    },
+    {
+      "epoch": 0.03775009437523594,
+      "grad_norm": 4.374029159545898,
+      "learning_rate": 4.950000000000001e-06,
+      "loss": 2.6371,
+      "step": 100
+    },
+    {
+      "epoch": 0.041525103812759534,
+      "grad_norm": 2.939470052719116,
+      "learning_rate": 5.45e-06,
+      "loss": 2.2299,
+      "step": 110
+    },
+    {
+      "epoch": 0.045300113250283124,
+      "grad_norm": 2.47430682182312,
+      "learning_rate": 5.95e-06,
+      "loss": 2.0141,
+      "step": 120
+    },
+    {
+      "epoch": 0.04907512268780672,
+      "grad_norm": 1.9557963609695435,
+      "learning_rate": 6.45e-06,
+      "loss": 1.8435,
+      "step": 130
+    },
+    {
+      "epoch": 0.05285013212533032,
+      "grad_norm": 1.204042673110962,
+      "learning_rate": 6.950000000000001e-06,
+      "loss": 1.5451,
+      "step": 140
+    },
+    {
+      "epoch": 0.056625141562853906,
+      "grad_norm": 0.9715531468391418,
+      "learning_rate": 7.45e-06,
+      "loss": 1.5027,
+      "step": 150
+    },
+    {
+      "epoch": 0.0604001510003775,
+      "grad_norm": 1.261393427848816,
+      "learning_rate": 7.95e-06,
+      "loss": 1.3894,
+      "step": 160
+    },
+    {
+      "epoch": 0.0641751604379011,
+      "grad_norm": 1.091807246208191,
+      "learning_rate": 8.45e-06,
+      "loss": 1.3591,
+      "step": 170
+    },
+    {
+      "epoch": 0.06795016987542468,
+      "grad_norm": 0.8383681178092957,
+      "learning_rate": 8.95e-06,
+      "loss": 1.2757,
+      "step": 180
+    },
+    {
+      "epoch": 0.07172517931294828,
+      "grad_norm": 1.3401917219161987,
+      "learning_rate": 9.450000000000001e-06,
+      "loss": 1.2084,
+      "step": 190
+    },
+    {
+      "epoch": 0.07550018875047187,
+      "grad_norm": 0.8086752891540527,
+      "learning_rate": 9.950000000000001e-06,
+      "loss": 1.3253,
+      "step": 200
+    },
+    {
+      "epoch": 0.07927519818799547,
+      "grad_norm": 0.9537988901138306,
+      "learning_rate": 1.045e-05,
+      "loss": 1.171,
+      "step": 210
+    },
+    {
+      "epoch": 0.08305020762551907,
+      "grad_norm": 0.8162275552749634,
+      "learning_rate": 1.095e-05,
+      "loss": 1.1737,
+      "step": 220
+    },
+    {
+      "epoch": 0.08682521706304266,
+      "grad_norm": 1.144768476486206,
+      "learning_rate": 1.145e-05,
+      "loss": 1.1406,
+      "step": 230
+    },
+    {
+      "epoch": 0.09060022650056625,
+      "grad_norm": 1.2052839994430542,
+      "learning_rate": 1.195e-05,
+      "loss": 1.0841,
+      "step": 240
+    },
+    {
+      "epoch": 0.09437523593808984,
+      "grad_norm": 1.1619216203689575,
+      "learning_rate": 1.2450000000000001e-05,
+      "loss": 1.2028,
+      "step": 250
+    },
+    {
+      "epoch": 0.09815024537561344,
+      "grad_norm": 0.971839189529419,
+      "learning_rate": 1.2950000000000001e-05,
+      "loss": 1.0536,
+      "step": 260
+    },
+    {
+      "epoch": 0.10192525481313704,
+      "grad_norm": 1.1455223560333252,
+      "learning_rate": 1.3450000000000002e-05,
+      "loss": 1.0539,
+      "step": 270
+    },
+    {
+      "epoch": 0.10570026425066063,
+      "grad_norm": 0.9137556552886963,
+      "learning_rate": 1.3950000000000002e-05,
+      "loss": 1.0439,
+      "step": 280
+    },
+    {
+      "epoch": 0.10947527368818422,
+      "grad_norm": 1.4055426120758057,
+      "learning_rate": 1.4449999999999999e-05,
+      "loss": 1.147,
+      "step": 290
+    },
+    {
+      "epoch": 0.11325028312570781,
+      "grad_norm": 1.3591498136520386,
+      "learning_rate": 1.4950000000000001e-05,
+      "loss": 1.0227,
+      "step": 300
+    },
+    {
+      "epoch": 0.11702529256323141,
+      "grad_norm": 1.004621148109436,
+      "learning_rate": 1.545e-05,
+      "loss": 1.0075,
+      "step": 310
+    },
+    {
+      "epoch": 0.120800302000755,
+      "grad_norm": 1.2812256813049316,
+      "learning_rate": 1.595e-05,
+      "loss": 1.0799,
+      "step": 320
+    },
+    {
+      "epoch": 0.1245753114382786,
+      "grad_norm": 1.5927451848983765,
+      "learning_rate": 1.645e-05,
+      "loss": 0.9892,
+      "step": 330
+    },
+    {
+      "epoch": 0.1283503208758022,
+      "grad_norm": 1.2106682062149048,
+      "learning_rate": 1.6950000000000002e-05,
+      "loss": 0.9776,
+      "step": 340
+    },
+    {
+      "epoch": 0.1321253303133258,
+      "grad_norm": 1.2619518041610718,
+      "learning_rate": 1.745e-05,
+      "loss": 0.9786,
+      "step": 350
+    },
+    {
+      "epoch": 0.13590033975084936,
+      "grad_norm": 1.5045936107635498,
+      "learning_rate": 1.795e-05,
+      "loss": 0.9763,
+      "step": 360
+    },
+    {
+      "epoch": 0.13967534918837296,
+      "grad_norm": 1.4551500082015991,
+      "learning_rate": 1.845e-05,
+      "loss": 0.9857,
+      "step": 370
+    },
+    {
+      "epoch": 0.14345035862589656,
+      "grad_norm": 1.3843104839324951,
+      "learning_rate": 1.895e-05,
+      "loss": 0.9163,
+      "step": 380
+    },
+    {
+      "epoch": 0.14722536806342015,
+      "grad_norm": 1.4003466367721558,
+      "learning_rate": 1.9450000000000002e-05,
+      "loss": 0.9637,
+      "step": 390
+    },
+    {
+      "epoch": 0.15100037750094375,
+      "grad_norm": 1.3285315036773682,
+      "learning_rate": 1.995e-05,
+      "loss": 0.9353,
+      "step": 400
+    },
+    {
+      "epoch": 0.15477538693846735,
+      "grad_norm": 1.0920555591583252,
+      "learning_rate": 2.045e-05,
+      "loss": 1.0432,
+      "step": 410
+    },
+    {
+      "epoch": 0.15855039637599094,
+      "grad_norm": 1.4697812795639038,
+      "learning_rate": 2.095e-05,
+      "loss": 1.0497,
+      "step": 420
+    },
+    {
+      "epoch": 0.16232540581351454,
+      "grad_norm": 1.444948434829712,
+      "learning_rate": 2.145e-05,
+      "loss": 0.9317,
+      "step": 430
+    },
+    {
+      "epoch": 0.16610041525103814,
+      "grad_norm": 1.8489030599594116,
+      "learning_rate": 2.195e-05,
+      "loss": 0.9296,
+      "step": 440
+    },
+    {
+      "epoch": 0.16987542468856173,
+      "grad_norm": 1.27092444896698,
+      "learning_rate": 2.245e-05,
+      "loss": 0.8998,
+      "step": 450
+    },
+    {
+      "epoch": 0.17365043412608533,
+      "grad_norm": 1.3666588068008423,
+      "learning_rate": 2.2950000000000002e-05,
+      "loss": 1.0234,
+      "step": 460
+    },
+    {
+      "epoch": 0.1774254435636089,
+      "grad_norm": 0.9460456967353821,
+      "learning_rate": 2.345e-05,
+      "loss": 0.9337,
+      "step": 470
+    },
+    {
+      "epoch": 0.1812004530011325,
+      "grad_norm": 1.4684996604919434,
+      "learning_rate": 2.395e-05,
+      "loss": 0.9308,
+      "step": 480
+    },
+    {
+      "epoch": 0.1849754624386561,
+      "grad_norm": 1.284343957901001,
+      "learning_rate": 2.445e-05,
+      "loss": 0.908,
+      "step": 490
+    },
+    {
+      "epoch": 0.1887504718761797,
+      "grad_norm": 0.9511118531227112,
+      "learning_rate": 2.495e-05,
+      "loss": 0.8959,
+      "step": 500
+    },
+    {
+      "epoch": 0.19252548131370328,
+      "grad_norm": 1.701767086982727,
+      "learning_rate": 2.5450000000000002e-05,
+      "loss": 0.9207,
+      "step": 510
+    },
+    {
+      "epoch": 0.19630049075122688,
+      "grad_norm": 1.341230034828186,
+      "learning_rate": 2.595e-05,
+      "loss": 0.9837,
+      "step": 520
+    },
+    {
+      "epoch": 0.20007550018875048,
+      "grad_norm": 1.1268380880355835,
+      "learning_rate": 2.6450000000000003e-05,
+      "loss": 0.8893,
+      "step": 530
+    },
+    {
+      "epoch": 0.20385050962627407,
+      "grad_norm": 1.5450079441070557,
+      "learning_rate": 2.6950000000000005e-05,
+      "loss": 0.8875,
+      "step": 540
+    },
+    {
+      "epoch": 0.20762551906379767,
+      "grad_norm": 1.827506184577942,
+      "learning_rate": 2.7450000000000003e-05,
+      "loss": 0.9069,
+      "step": 550
+    },
+    {
+      "epoch": 0.21140052850132127,
+      "grad_norm": 1.225759506225586,
+      "learning_rate": 2.7950000000000005e-05,
+      "loss": 0.8977,
+      "step": 560
+    },
+    {
+      "epoch": 0.21517553793884484,
+      "grad_norm": 1.1317689418792725,
+      "learning_rate": 2.845e-05,
+      "loss": 0.8673,
+      "step": 570
+    },
+    {
+      "epoch": 0.21895054737636843,
+      "grad_norm": 1.3549116849899292,
+      "learning_rate": 2.895e-05,
+      "loss": 0.8613,
+      "step": 580
+    },
+    {
+      "epoch": 0.22272555681389203,
+      "grad_norm": 1.5646511316299438,
+      "learning_rate": 2.945e-05,
+      "loss": 0.8847,
+      "step": 590
+    },
+    {
+      "epoch": 0.22650056625141562,
+      "grad_norm": 1.453460693359375,
+      "learning_rate": 2.995e-05,
+      "loss": 0.8523,
+      "step": 600
+    },
+    {
+      "epoch": 0.23027557568893922,
+      "grad_norm": 1.3521499633789062,
+      "learning_rate": 3.045e-05,
+      "loss": 0.8773,
+      "step": 610
+    },
+    {
+      "epoch": 0.23405058512646282,
+      "grad_norm": 1.041245698928833,
+      "learning_rate": 3.095e-05,
+      "loss": 0.857,
+      "step": 620
+    },
+    {
+      "epoch": 0.23782559456398641,
+      "grad_norm": 1.2298246622085571,
+      "learning_rate": 3.145e-05,
+      "loss": 0.8364,
+      "step": 630
+    },
+    {
+      "epoch": 0.24160060400151,
+      "grad_norm": 1.2487760782241821,
+      "learning_rate": 3.1950000000000004e-05,
+      "loss": 0.8697,
+      "step": 640
+    },
+    {
+      "epoch": 0.2453756134390336,
+      "grad_norm": 1.3753890991210938,
+      "learning_rate": 3.245e-05,
+      "loss": 0.8579,
+      "step": 650
+    },
+    {
+      "epoch": 0.2491506228765572,
+      "grad_norm": 1.354738712310791,
+      "learning_rate": 3.295e-05,
+      "loss": 0.8574,
+      "step": 660
+    },
+    {
+      "epoch": 0.2529256323140808,
+      "grad_norm": 1.3853284120559692,
+      "learning_rate": 3.345000000000001e-05,
+      "loss": 0.8346,
+      "step": 670
+    },
+    {
+      "epoch": 0.2567006417516044,
+      "grad_norm": 1.4447784423828125,
+      "learning_rate": 3.3950000000000005e-05,
+      "loss": 0.9262,
+      "step": 680
+    },
+    {
+      "epoch": 0.26047565118912797,
+      "grad_norm": 1.2130733728408813,
+      "learning_rate": 3.445e-05,
+      "loss": 0.8176,
+      "step": 690
+    },
+    {
+      "epoch": 0.2642506606266516,
+      "grad_norm": 1.698705792427063,
+      "learning_rate": 3.495e-05,
+      "loss": 0.8424,
+      "step": 700
+    },
+    {
+      "epoch": 0.26802567006417516,
+      "grad_norm": 1.724214792251587,
+      "learning_rate": 3.545e-05,
+      "loss": 0.9077,
+      "step": 710
+    },
+    {
+      "epoch": 0.2718006795016987,
+      "grad_norm": 1.296976923942566,
+      "learning_rate": 3.595e-05,
+      "loss": 0.8253,
+      "step": 720
+    },
+    {
+      "epoch": 0.27557568893922235,
+      "grad_norm": 1.2953417301177979,
+      "learning_rate": 3.645e-05,
+      "loss": 0.8949,
+      "step": 730
+    },
+    {
+      "epoch": 0.2793506983767459,
+      "grad_norm": 1.267232894897461,
+      "learning_rate": 3.6950000000000004e-05,
+      "loss": 0.8476,
+      "step": 740
+    },
+    {
+      "epoch": 0.28312570781426954,
+      "grad_norm": 1.76276695728302,
+      "learning_rate": 3.745e-05,
+      "loss": 0.8981,
+      "step": 750
+    },
+    {
+      "epoch": 0.2869007172517931,
+      "grad_norm": 1.5421119928359985,
+      "learning_rate": 3.795e-05,
+      "loss": 0.8343,
+      "step": 760
+    },
+    {
+      "epoch": 0.29067572668931674,
+      "grad_norm": 1.1343638896942139,
+      "learning_rate": 3.845e-05,
+      "loss": 0.8316,
+      "step": 770
+    },
+    {
+      "epoch": 0.2944507361268403,
+      "grad_norm": 0.9002248048782349,
+      "learning_rate": 3.8950000000000005e-05,
+      "loss": 0.8207,
+      "step": 780
+    },
+    {
+      "epoch": 0.29822574556436393,
+      "grad_norm": 1.273121953010559,
+      "learning_rate": 3.9450000000000003e-05,
+      "loss": 0.8782,
+      "step": 790
+    },
+    {
+      "epoch": 0.3020007550018875,
+      "grad_norm": 1.6263741254806519,
+      "learning_rate": 3.995e-05,
+      "loss": 0.8359,
+      "step": 800
+    },
+    {
+      "epoch": 0.3057757644394111,
+      "grad_norm": 1.3746614456176758,
+      "learning_rate": 4.045000000000001e-05,
+      "loss": 0.7946,
+      "step": 810
+    },
+    {
+      "epoch": 0.3095507738769347,
+      "grad_norm": 1.2415761947631836,
+      "learning_rate": 4.095e-05,
+      "loss": 0.7973,
+      "step": 820
+    },
+    {
+      "epoch": 0.31332578331445826,
+      "grad_norm": 1.372182846069336,
+      "learning_rate": 4.145e-05,
+      "loss": 0.9107,
+      "step": 830
+    },
+    {
+      "epoch": 0.3171007927519819,
+      "grad_norm": 1.5173795223236084,
+      "learning_rate": 4.195e-05,
+      "loss": 0.7988,
+      "step": 840
+    },
+    {
+      "epoch": 0.32087580218950545,
+      "grad_norm": 1.5358010530471802,
+      "learning_rate": 4.245e-05,
+      "loss": 0.7724,
+      "step": 850
+    },
+    {
+      "epoch": 0.3246508116270291,
+      "grad_norm": 0.867175817489624,
+      "learning_rate": 4.295e-05,
+      "loss": 0.7816,
+      "step": 860
+    },
+    {
+      "epoch": 0.32842582106455265,
+      "grad_norm": 0.9137406349182129,
+      "learning_rate": 4.345e-05,
+      "loss": 0.7855,
+      "step": 870
+    },
+    {
+      "epoch": 0.33220083050207627,
+      "grad_norm": 0.9933797717094421,
+      "learning_rate": 4.3950000000000004e-05,
+      "loss": 0.8015,
+      "step": 880
+    },
+    {
+      "epoch": 0.33597583993959984,
+      "grad_norm": 1.1218225955963135,
+      "learning_rate": 4.445e-05,
+      "loss": 0.7944,
+      "step": 890
+    },
+    {
+      "epoch": 0.33975084937712347,
+      "grad_norm": 0.9928627610206604,
+      "learning_rate": 4.495e-05,
+      "loss": 0.785,
+      "step": 900
+    },
+    {
+      "epoch": 0.34352585881464703,
+      "grad_norm": 1.0522204637527466,
+      "learning_rate": 4.545000000000001e-05,
+      "loss": 0.7819,
+      "step": 910
+    },
+    {
+      "epoch": 0.34730086825217066,
+      "grad_norm": 1.162044644355774,
+      "learning_rate": 4.5950000000000006e-05,
+      "loss": 0.7721,
+      "step": 920
+    },
+    {
+      "epoch": 0.3510758776896942,
+      "grad_norm": 0.9964906573295593,
+      "learning_rate": 4.6450000000000004e-05,
+      "loss": 0.7716,
+      "step": 930
+    },
+    {
+      "epoch": 0.3548508871272178,
+      "grad_norm": 1.3030009269714355,
+      "learning_rate": 4.695e-05,
+      "loss": 0.7756,
+      "step": 940
+    },
+    {
+      "epoch": 0.3586258965647414,
+      "grad_norm": 1.413189172744751,
+      "learning_rate": 4.745e-05,
+      "loss": 0.7709,
+      "step": 950
+    },
+    {
+      "epoch": 0.362400906002265,
+      "grad_norm": 1.1650564670562744,
+      "learning_rate": 4.795e-05,
+      "loss": 0.7647,
+      "step": 960
+    },
+    {
+      "epoch": 0.3661759154397886,
+      "grad_norm": 1.0290861129760742,
+      "learning_rate": 4.845e-05,
+      "loss": 0.7601,
+      "step": 970
+    },
+    {
+      "epoch": 0.3699509248773122,
+      "grad_norm": 1.1548311710357666,
+      "learning_rate": 4.8950000000000004e-05,
+      "loss": 0.8407,
+      "step": 980
+    },
+    {
+      "epoch": 0.3737259343148358,
+      "grad_norm": 1.037048578262329,
+      "learning_rate": 4.945e-05,
+      "loss": 0.7576,
+      "step": 990
+    },
+    {
+      "epoch": 0.3775009437523594,
+      "grad_norm": 1.5112618207931519,
+      "learning_rate": 4.995e-05,
+      "loss": 0.7462,
+      "step": 1000
+    },
+    {
+      "epoch": 0.381275953189883,
+      "grad_norm": 1.65474534034729,
+      "learning_rate": 5.045e-05,
+      "loss": 0.7722,
+      "step": 1010
+    },
+    {
+      "epoch": 0.38505096262740657,
+      "grad_norm": 1.3295589685440063,
+      "learning_rate": 5.095e-05,
+      "loss": 0.7551,
+      "step": 1020
+    },
+    {
+      "epoch": 0.38882597206493014,
+      "grad_norm": 0.9878939390182495,
+      "learning_rate": 5.145e-05,
+      "loss": 0.7522,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39260098150245376,
+      "grad_norm": 1.3945156335830688,
+      "learning_rate": 5.1949999999999996e-05,
+      "loss": 0.7827,
+      "step": 1040
+    },
+    {
+      "epoch": 0.39637599093997733,
+      "grad_norm": 1.2583463191986084,
+      "learning_rate": 5.245e-05,
+      "loss": 0.7658,
+      "step": 1050
+    },
+    {
+      "epoch": 0.40015100037750095,
+      "grad_norm": 1.0991382598876953,
+      "learning_rate": 5.295e-05,
+      "loss": 0.8056,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4039260098150245,
+      "grad_norm": 1.0522119998931885,
+      "learning_rate": 5.345e-05,
+      "loss": 0.8069,
+      "step": 1070
+    },
+    {
+      "epoch": 0.40770101925254815,
+      "grad_norm": 0.8524134755134583,
+      "learning_rate": 5.3950000000000004e-05,
+      "loss": 0.8186,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4114760286900717,
+      "grad_norm": 1.1254100799560547,
+      "learning_rate": 5.445e-05,
+      "loss": 0.7362,
+      "step": 1090
+    },
+    {
+      "epoch": 0.41525103812759534,
+      "grad_norm": 0.7959914207458496,
+      "learning_rate": 5.495e-05,
+      "loss": 0.7387,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4190260475651189,
+      "grad_norm": 1.2034642696380615,
+      "learning_rate": 5.545e-05,
+      "loss": 0.7458,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42280105700264253,
+      "grad_norm": 1.2874923944473267,
+      "learning_rate": 5.5950000000000005e-05,
+      "loss": 0.7442,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4265760664401661,
+      "grad_norm": 1.2025848627090454,
+      "learning_rate": 5.645e-05,
+      "loss": 0.7624,
+      "step": 1130
+    },
+    {
+      "epoch": 0.43035107587768967,
+      "grad_norm": 1.1338868141174316,
+      "learning_rate": 5.695e-05,
+      "loss": 0.7311,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4341260853152133,
+      "grad_norm": 1.1724085807800293,
+      "learning_rate": 5.745e-05,
+      "loss": 0.7383,
+      "step": 1150
+    },
+    {
+      "epoch": 0.43790109475273686,
+      "grad_norm": 1.3248220682144165,
+      "learning_rate": 5.7950000000000006e-05,
+      "loss": 0.7379,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4416761041902605,
+      "grad_norm": 1.4836875200271606,
+      "learning_rate": 5.8450000000000005e-05,
+      "loss": 0.7369,
+      "step": 1170
+    },
+    {
+      "epoch": 0.44545111362778406,
+      "grad_norm": 1.1852290630340576,
+      "learning_rate": 5.895e-05,
+      "loss": 0.8117,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4492261230653077,
+      "grad_norm": 2.466386079788208,
+      "learning_rate": 5.945000000000001e-05,
+      "loss": 0.7997,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45300113250283125,
+      "grad_norm": 1.0344855785369873,
+      "learning_rate": 5.995000000000001e-05,
+      "loss": 0.7181,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4567761419403549,
+      "grad_norm": 1.1200584173202515,
+      "learning_rate": 6.0450000000000006e-05,
+      "loss": 0.7204,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46055115137787844,
+      "grad_norm": 1.0157345533370972,
+      "learning_rate": 6.0950000000000004e-05,
+      "loss": 0.7261,
+      "step": 1220
+    },
+    {
+      "epoch": 0.464326160815402,
+      "grad_norm": 0.9601739048957825,
+      "learning_rate": 6.145e-05,
+      "loss": 0.7099,
+      "step": 1230
+    },
+    {
+      "epoch": 0.46810117025292564,
+      "grad_norm": 0.8642465472221375,
+      "learning_rate": 6.195e-05,
+      "loss": 0.7242,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4718761796904492,
+      "grad_norm": 0.9395157098770142,
+      "learning_rate": 6.245000000000001e-05,
+      "loss": 0.716,
+      "step": 1250
+    },
+    {
+      "epoch": 0.47565118912797283,
+      "grad_norm": 1.0848698616027832,
+      "learning_rate": 6.295e-05,
+      "loss": 0.714,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4794261985654964,
+      "grad_norm": 1.1483030319213867,
+      "learning_rate": 6.345e-05,
+      "loss": 0.7224,
+      "step": 1270
+    },
+    {
+      "epoch": 0.48320120800302,
+      "grad_norm": 0.8236299753189087,
+      "learning_rate": 6.395e-05,
+      "loss": 0.7639,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4869762174405436,
+      "grad_norm": 0.9821550846099854,
+      "learning_rate": 6.445e-05,
+      "loss": 0.7136,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4907512268780672,
+      "grad_norm": 0.94017493724823,
+      "learning_rate": 6.494999999999999e-05,
+      "loss": 0.7056,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4945262363155908,
+      "grad_norm": 0.8587389588356018,
+      "learning_rate": 6.545e-05,
+      "loss": 0.7518,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4983012457531144,
+      "grad_norm": 0.9293227791786194,
+      "learning_rate": 6.595e-05,
+      "loss": 0.7058,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5020762551906379,
+      "grad_norm": 1.033216118812561,
+      "learning_rate": 6.645e-05,
+      "loss": 0.7106,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5058512646281615,
+      "grad_norm": 0.8532618880271912,
+      "learning_rate": 6.695e-05,
+      "loss": 0.7153,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5096262740656852,
+      "grad_norm": 0.7834755182266235,
+      "learning_rate": 6.745e-05,
+      "loss": 0.705,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5134012835032088,
+      "grad_norm": 0.7160763740539551,
+      "learning_rate": 6.795e-05,
+      "loss": 0.6986,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5171762929407323,
+      "grad_norm": 0.9718685150146484,
+      "learning_rate": 6.845e-05,
+      "loss": 0.7002,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5209513023782559,
+      "grad_norm": 0.9730781316757202,
+      "learning_rate": 6.895000000000001e-05,
+      "loss": 0.7107,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5247263118157796,
+      "grad_norm": 1.1289321184158325,
+      "learning_rate": 6.945000000000001e-05,
+      "loss": 0.6985,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5285013212533032,
+      "grad_norm": 0.6597205400466919,
+      "learning_rate": 6.995e-05,
+      "loss": 0.7051,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5322763306908267,
+      "grad_norm": 1.0954090356826782,
+      "learning_rate": 7.045e-05,
+      "loss": 0.6965,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5360513401283503,
+      "grad_norm": 0.9379609823226929,
+      "learning_rate": 7.095e-05,
+      "loss": 0.6937,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5398263495658739,
+      "grad_norm": 1.048135757446289,
+      "learning_rate": 7.145e-05,
+      "loss": 0.6957,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5436013590033975,
+      "grad_norm": 1.2906863689422607,
+      "learning_rate": 7.195e-05,
+      "loss": 0.6994,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5473763684409211,
+      "grad_norm": 1.1085039377212524,
+      "learning_rate": 7.245000000000001e-05,
+      "loss": 0.6932,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5511513778784447,
+      "grad_norm": 1.1647526025772095,
+      "learning_rate": 7.295000000000001e-05,
+      "loss": 0.7641,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5549263873159683,
+      "grad_norm": 0.9016187191009521,
+      "learning_rate": 7.345000000000001e-05,
+      "loss": 0.7047,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5587013967534918,
+      "grad_norm": 0.7149607539176941,
+      "learning_rate": 7.395000000000001e-05,
+      "loss": 0.6924,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5624764061910155,
+      "grad_norm": 0.9485162496566772,
+      "learning_rate": 7.445000000000001e-05,
+      "loss": 0.6896,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5662514156285391,
+      "grad_norm": 0.9566022157669067,
+      "learning_rate": 7.495e-05,
+      "loss": 0.6893,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5700264250660627,
+      "grad_norm": 0.8616087436676025,
+      "learning_rate": 7.545e-05,
+      "loss": 0.7356,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5738014345035862,
+      "grad_norm": 0.7664632201194763,
+      "learning_rate": 7.595e-05,
+      "loss": 0.6909,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5775764439411099,
+      "grad_norm": 0.673774003982544,
+      "learning_rate": 7.645e-05,
+      "loss": 0.6804,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5813514533786335,
+      "grad_norm": 1.0145224332809448,
+      "learning_rate": 7.695e-05,
+      "loss": 0.696,
+      "step": 1540
+    },
+    {
+      "epoch": 0.585126462816157,
+      "grad_norm": 0.805166482925415,
+      "learning_rate": 7.745e-05,
+      "loss": 0.7003,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5889014722536806,
+      "grad_norm": 0.7893358469009399,
+      "learning_rate": 7.795e-05,
+      "loss": 0.6933,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5926764816912042,
+      "grad_norm": 0.9435643553733826,
+      "learning_rate": 7.845e-05,
+      "loss": 0.6926,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5964514911287279,
+      "grad_norm": 0.9065937399864197,
+      "learning_rate": 7.895000000000001e-05,
+      "loss": 0.685,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6002265005662514,
+      "grad_norm": 1.084681749343872,
+      "learning_rate": 7.945e-05,
+      "loss": 0.7046,
+      "step": 1590
+    },
+    {
+      "epoch": 0.604001510003775,
+      "grad_norm": 0.6447507739067078,
+      "learning_rate": 7.995e-05,
+      "loss": 0.7189,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6077765194412986,
+      "grad_norm": 0.855850100517273,
+      "learning_rate": 8.045e-05,
+      "loss": 0.6817,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6115515288788222,
+      "grad_norm": 1.1896417140960693,
+      "learning_rate": 8.095e-05,
+      "loss": 0.6844,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6153265383163458,
+      "grad_norm": 0.8419579267501831,
+      "learning_rate": 8.145e-05,
+      "loss": 0.6771,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6191015477538694,
+      "grad_norm": 0.8358510732650757,
+      "learning_rate": 8.195e-05,
+      "loss": 0.6855,
+      "step": 1640
+    },
+    {
+      "epoch": 0.622876557191393,
+      "grad_norm": 0.896302342414856,
+      "learning_rate": 8.245e-05,
+      "loss": 0.6955,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6266515666289165,
+      "grad_norm": 1.024187684059143,
+      "learning_rate": 8.295000000000001e-05,
+      "loss": 0.6811,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6304265760664401,
+      "grad_norm": 0.8998435735702515,
+      "learning_rate": 8.345000000000001e-05,
+      "loss": 0.6807,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6342015855039638,
+      "grad_norm": 0.6220159530639648,
+      "learning_rate": 8.395000000000001e-05,
+      "loss": 0.6795,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6379765949414874,
+      "grad_norm": 0.9150254130363464,
+      "learning_rate": 8.445e-05,
+      "loss": 0.6695,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6417516043790109,
+      "grad_norm": 0.7432071566581726,
+      "learning_rate": 8.495e-05,
+      "loss": 0.6792,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6455266138165345,
+      "grad_norm": 0.808611273765564,
+      "learning_rate": 8.545e-05,
+      "loss": 0.6734,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6493016232540582,
+      "grad_norm": 0.6958522200584412,
+      "learning_rate": 8.595e-05,
+      "loss": 0.6867,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6530766326915818,
+      "grad_norm": 0.7854651808738708,
+      "learning_rate": 8.645000000000001e-05,
+      "loss": 0.6837,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6568516421291053,
+      "grad_norm": 0.7910404205322266,
+      "learning_rate": 8.695000000000001e-05,
+      "loss": 0.6746,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6606266515666289,
+      "grad_norm": 0.7932810187339783,
+      "learning_rate": 8.745000000000001e-05,
+      "loss": 0.6642,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6644016610041525,
+      "grad_norm": 0.7843324542045593,
+      "learning_rate": 8.795e-05,
+      "loss": 0.6641,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6681766704416761,
+      "grad_norm": 0.728706955909729,
+      "learning_rate": 8.845e-05,
+      "loss": 0.6745,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6719516798791997,
+      "grad_norm": 0.8016153573989868,
+      "learning_rate": 8.895e-05,
+      "loss": 0.6671,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6757266893167233,
+      "grad_norm": 0.695740282535553,
+      "learning_rate": 8.945e-05,
+      "loss": 0.6634,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6795016987542469,
+      "grad_norm": 0.8751599788665771,
+      "learning_rate": 8.995e-05,
+      "loss": 0.6747,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6832767081917704,
+      "grad_norm": 0.46772128343582153,
+      "learning_rate": 9.045e-05,
+      "loss": 0.6721,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6870517176292941,
+      "grad_norm": 0.7569857239723206,
+      "learning_rate": 9.095e-05,
+      "loss": 0.6675,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6908267270668177,
+      "grad_norm": 0.853040337562561,
+      "learning_rate": 9.145e-05,
+      "loss": 0.6685,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6946017365043413,
+      "grad_norm": 0.6927493214607239,
+      "learning_rate": 9.195e-05,
+      "loss": 0.654,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6983767459418648,
+      "grad_norm": 0.6520769000053406,
+      "learning_rate": 9.245e-05,
+      "loss": 0.6641,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7021517553793885,
+      "grad_norm": 0.9358047246932983,
+      "learning_rate": 9.295000000000001e-05,
+      "loss": 0.6696,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7059267648169121,
+      "grad_norm": 0.7453898787498474,
+      "learning_rate": 9.345000000000001e-05,
+      "loss": 0.6549,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7097017742544356,
+      "grad_norm": 0.6966333389282227,
+      "learning_rate": 9.395000000000001e-05,
+      "loss": 0.6653,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7134767836919592,
+      "grad_norm": 0.7843179702758789,
+      "learning_rate": 9.445e-05,
+      "loss": 0.6715,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7172517931294828,
+      "grad_norm": 0.6947644948959351,
+      "learning_rate": 9.495e-05,
+      "loss": 0.6584,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7210268025670065,
+      "grad_norm": 0.8173465132713318,
+      "learning_rate": 9.545e-05,
+      "loss": 0.6649,
+      "step": 1910
+    },
+    {
+      "epoch": 0.72480181200453,
+      "grad_norm": 0.9682592749595642,
+      "learning_rate": 9.595e-05,
+      "loss": 0.6549,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7285768214420536,
+      "grad_norm": 1.0117123126983643,
+      "learning_rate": 9.645000000000001e-05,
+      "loss": 0.6536,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7323518308795772,
+      "grad_norm": 0.9777953028678894,
+      "learning_rate": 9.695000000000001e-05,
+      "loss": 0.6578,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7361268403171007,
+      "grad_norm": 0.7957656383514404,
+      "learning_rate": 9.745000000000001e-05,
+      "loss": 0.6566,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7399018497546244,
+      "grad_norm": 0.7392054796218872,
+      "learning_rate": 9.795000000000001e-05,
+      "loss": 0.66,
+      "step": 1960
+    },
+    {
+      "epoch": 0.743676859192148,
+      "grad_norm": 0.811776340007782,
+      "learning_rate": 9.845000000000001e-05,
+      "loss": 0.6535,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7474518686296716,
+      "grad_norm": 0.750037670135498,
+      "learning_rate": 9.895e-05,
+      "loss": 0.651,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7512268780671951,
+      "grad_norm": 0.6936712861061096,
+      "learning_rate": 9.945e-05,
+      "loss": 0.6594,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7550018875047187,
+      "grad_norm": 0.9359191060066223,
+      "learning_rate": 9.995e-05,
+      "loss": 0.6546,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7587768969422424,
+      "grad_norm": 0.8320378065109253,
+      "learning_rate": 9.99999861593158e-05,
+      "loss": 0.6427,
+      "step": 2010
+    },
+    {
+      "epoch": 0.762551906379766,
+      "grad_norm": 0.6402168869972229,
+      "learning_rate": 9.999993831498517e-05,
+      "loss": 0.6711,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7663269158172895,
+      "grad_norm": 0.6975911259651184,
+      "learning_rate": 9.99998562961682e-05,
+      "loss": 0.6551,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7701019252548131,
+      "grad_norm": 0.7861037254333496,
+      "learning_rate": 9.99997401029209e-05,
+      "loss": 0.6548,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7738769346923368,
+      "grad_norm": 0.7647865414619446,
+      "learning_rate": 9.999958973532271e-05,
+      "loss": 0.6553,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7776519441298603,
+      "grad_norm": 1.0054905414581299,
+      "learning_rate": 9.999940519347642e-05,
+      "loss": 0.6523,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7814269535673839,
+      "grad_norm": 0.9435162544250488,
+      "learning_rate": 9.999918647750812e-05,
+      "loss": 0.6573,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7852019630049075,
+      "grad_norm": 0.7330409288406372,
+      "learning_rate": 9.999893358756736e-05,
+      "loss": 0.6601,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7889769724424311,
+      "grad_norm": 1.0650360584259033,
+      "learning_rate": 9.999864652382692e-05,
+      "loss": 0.6481,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7927519818799547,
+      "grad_norm": 0.6721388697624207,
+      "learning_rate": 9.999832528648307e-05,
+      "loss": 0.6395,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7965269913174783,
+      "grad_norm": 0.8571961522102356,
+      "learning_rate": 9.999796987575532e-05,
+      "loss": 0.6459,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8003020007550019,
+      "grad_norm": 0.575950026512146,
+      "learning_rate": 9.999758029188662e-05,
+      "loss": 0.6459,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8040770101925255,
+      "grad_norm": 0.5099409818649292,
+      "learning_rate": 9.999715653514324e-05,
+      "loss": 0.6484,
+      "step": 2130
+    },
+    {
+      "epoch": 0.807852019630049,
+      "grad_norm": 0.6396839618682861,
+      "learning_rate": 9.999669860581482e-05,
+      "loss": 0.6504,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8116270290675727,
+      "grad_norm": 0.6424649357795715,
+      "learning_rate": 9.999620650421433e-05,
+      "loss": 0.6396,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8154020385050963,
+      "grad_norm": 0.7561330199241638,
+      "learning_rate": 9.999568023067813e-05,
+      "loss": 0.6481,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8191770479426198,
+      "grad_norm": 0.8530335426330566,
+      "learning_rate": 9.999511978556595e-05,
+      "loss": 0.6392,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8229520573801434,
+      "grad_norm": 0.8621054887771606,
+      "learning_rate": 9.99945251692608e-05,
+      "loss": 0.652,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8267270668176671,
+      "grad_norm": 0.7959499359130859,
+      "learning_rate": 9.999389638216912e-05,
+      "loss": 0.6384,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8305020762551907,
+      "grad_norm": 0.6927378177642822,
+      "learning_rate": 9.999323342472068e-05,
+      "loss": 0.6432,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8342770856927142,
+      "grad_norm": 0.5878588557243347,
+      "learning_rate": 9.99925362973686e-05,
+      "loss": 0.6464,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8380520951302378,
+      "grad_norm": 0.6220201253890991,
+      "learning_rate": 9.999180500058935e-05,
+      "loss": 0.6978,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8418271045677614,
+      "grad_norm": 0.6637720465660095,
+      "learning_rate": 9.99910395348828e-05,
+      "loss": 0.6336,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8456021140052851,
+      "grad_norm": 0.7113637924194336,
+      "learning_rate": 9.99902399007721e-05,
+      "loss": 0.633,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8493771234428086,
+      "grad_norm": 0.7006509900093079,
+      "learning_rate": 9.99894060988038e-05,
+      "loss": 0.634,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8531521328803322,
+      "grad_norm": 0.7758764624595642,
+      "learning_rate": 9.998853812954783e-05,
+      "loss": 0.6327,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8569271423178558,
+      "grad_norm": 0.7731608748435974,
+      "learning_rate": 9.998763599359739e-05,
+      "loss": 0.6301,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8607021517553793,
+      "grad_norm": 0.9600358009338379,
+      "learning_rate": 9.998669969156912e-05,
+      "loss": 0.6473,
+      "step": 2280
+    },
+    {
+      "epoch": 0.864477161192903,
+      "grad_norm": 0.6508148312568665,
+      "learning_rate": 9.998572922410294e-05,
+      "loss": 0.6401,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8682521706304266,
+      "grad_norm": 0.6029082536697388,
+      "learning_rate": 9.998472459186219e-05,
+      "loss": 0.6355,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8720271800679502,
+      "grad_norm": 0.5723931193351746,
+      "learning_rate": 9.99836857955335e-05,
+      "loss": 0.6352,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8758021895054737,
+      "grad_norm": 0.5845490097999573,
+      "learning_rate": 9.998261283582688e-05,
+      "loss": 0.6348,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8795771989429974,
+      "grad_norm": 0.702069103717804,
+      "learning_rate": 9.998150571347571e-05,
+      "loss": 0.7144,
+      "step": 2330
+    },
+    {
+      "epoch": 0.883352208380521,
+      "grad_norm": 0.7790460586547852,
+      "learning_rate": 9.998036442923667e-05,
+      "loss": 0.6357,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8871272178180446,
+      "grad_norm": 0.6690578460693359,
+      "learning_rate": 9.997918898388983e-05,
+      "loss": 0.6327,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8909022272555681,
+      "grad_norm": 0.7188066244125366,
+      "learning_rate": 9.997797937823859e-05,
+      "loss": 0.626,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8946772366930917,
+      "grad_norm": 0.6960968971252441,
+      "learning_rate": 9.997673561310971e-05,
+      "loss": 0.6359,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8984522461306154,
+      "grad_norm": 0.5999962091445923,
+      "learning_rate": 9.997545768935331e-05,
+      "loss": 0.642,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9022272555681389,
+      "grad_norm": 0.6590844988822937,
+      "learning_rate": 9.997414560784282e-05,
+      "loss": 0.6334,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9060022650056625,
+      "grad_norm": 0.5802249908447266,
+      "learning_rate": 9.997279936947502e-05,
+      "loss": 0.6319,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9097772744431861,
+      "grad_norm": 0.5891656279563904,
+      "learning_rate": 9.997141897517006e-05,
+      "loss": 0.6261,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9135522838807097,
+      "grad_norm": 0.6824517846107483,
+      "learning_rate": 9.997000442587145e-05,
+      "loss": 0.6261,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9173272933182333,
+      "grad_norm": 0.7084891200065613,
+      "learning_rate": 9.9968555722546e-05,
+      "loss": 0.6266,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9211023027557569,
+      "grad_norm": 0.5537872314453125,
+      "learning_rate": 9.996707286618387e-05,
+      "loss": 0.6271,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9248773121932805,
+      "grad_norm": 0.5806447863578796,
+      "learning_rate": 9.996555585779861e-05,
+      "loss": 0.6356,
+      "step": 2450
+    },
+    {
+      "epoch": 0.928652321630804,
+      "grad_norm": 0.6291959285736084,
+      "learning_rate": 9.996400469842708e-05,
+      "loss": 0.6344,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9324273310683276,
+      "grad_norm": 0.5299859046936035,
+      "learning_rate": 9.996241938912945e-05,
+      "loss": 0.6256,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9362023405058513,
+      "grad_norm": 0.5358120203018188,
+      "learning_rate": 9.99607999309893e-05,
+      "loss": 0.6164,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9399773499433749,
+      "grad_norm": 0.4257400929927826,
+      "learning_rate": 9.99591463251135e-05,
+      "loss": 0.6238,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9437523593808984,
+      "grad_norm": 0.5488512516021729,
+      "learning_rate": 9.995745857263226e-05,
+      "loss": 0.632,
+      "step": 2500
+    },
+    {
+      "epoch": 0.947527368818422,
+      "grad_norm": 0.45664718747138977,
+      "learning_rate": 9.995573667469918e-05,
+      "loss": 0.6276,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9513023782559457,
+      "grad_norm": 0.7003333568572998,
+      "learning_rate": 9.995398063249111e-05,
+      "loss": 0.6248,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9550773876934693,
+      "grad_norm": 0.5362398028373718,
+      "learning_rate": 9.995219044720833e-05,
+      "loss": 0.6277,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9588523971309928,
+      "grad_norm": 0.8197181224822998,
+      "learning_rate": 9.99503661200744e-05,
+      "loss": 0.6219,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9626274065685164,
+      "grad_norm": 0.671510636806488,
+      "learning_rate": 9.994850765233624e-05,
+      "loss": 0.6262,
+      "step": 2550
+    },
+    {
+      "epoch": 0.96640241600604,
+      "grad_norm": 0.6340009570121765,
+      "learning_rate": 9.994661504526406e-05,
+      "loss": 0.6167,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9701774254435636,
+      "grad_norm": 0.5801326632499695,
+      "learning_rate": 9.994468830015148e-05,
+      "loss": 0.6319,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9739524348810872,
+      "grad_norm": 0.6170343160629272,
+      "learning_rate": 9.99427274183154e-05,
+      "loss": 0.6233,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9777274443186108,
+      "grad_norm": 0.5810487866401672,
+      "learning_rate": 9.994073240109606e-05,
+      "loss": 0.6281,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9815024537561344,
+      "grad_norm": 0.46137765049934387,
+      "learning_rate": 9.993870324985703e-05,
+      "loss": 0.6144,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9852774631936579,
+      "grad_norm": 0.6433082222938538,
+      "learning_rate": 9.993663996598523e-05,
+      "loss": 0.6129,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9890524726311816,
+      "grad_norm": 0.6449345350265503,
+      "learning_rate": 9.993454255089089e-05,
+      "loss": 0.6149,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9928274820687052,
+      "grad_norm": 0.6302064061164856,
+      "learning_rate": 9.993241100600757e-05,
+      "loss": 0.6192,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9966024915062288,
+      "grad_norm": 0.7228666543960571,
+      "learning_rate": 9.993024533279215e-05,
+      "loss": 0.6179,
+      "step": 2640
+    },
+    {
+      "epoch": 1.0003775009437523,
+      "grad_norm": 0.5953015685081482,
+      "learning_rate": 9.992804553272487e-05,
+      "loss": 0.6176,
+      "step": 2650
+    },
+    {
+      "epoch": 1.0041525103812758,
+      "grad_norm": 0.586175799369812,
+      "learning_rate": 9.992581160730926e-05,
+      "loss": 0.6095,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0079275198187996,
+      "grad_norm": 0.6601322293281555,
+      "learning_rate": 9.992354355807218e-05,
+      "loss": 0.608,
+      "step": 2670
+    },
+    {
+      "epoch": 1.011702529256323,
+      "grad_norm": 0.7241320013999939,
+      "learning_rate": 9.992124138656385e-05,
+      "loss": 0.6204,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0154775386938468,
+      "grad_norm": 0.7130767107009888,
+      "learning_rate": 9.991890509435774e-05,
+      "loss": 0.6191,
+      "step": 2690
+    },
+    {
+      "epoch": 1.0192525481313703,
+      "grad_norm": 0.5906509757041931,
+      "learning_rate": 9.991653468305071e-05,
+      "loss": 0.6216,
+      "step": 2700
+    },
+    {
+      "epoch": 1.0230275575688939,
+      "grad_norm": 0.4647708535194397,
+      "learning_rate": 9.991413015426292e-05,
+      "loss": 0.6226,
+      "step": 2710
+    },
+    {
+      "epoch": 1.0268025670064176,
+      "grad_norm": 0.5289870500564575,
+      "learning_rate": 9.991169150963781e-05,
+      "loss": 0.6204,
+      "step": 2720
+    },
+    {
+      "epoch": 1.030577576443941,
+      "grad_norm": 0.5250043272972107,
+      "learning_rate": 9.990921875084221e-05,
+      "loss": 0.6101,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0343525858814646,
+      "grad_norm": 0.553863525390625,
+      "learning_rate": 9.990671187956622e-05,
+      "loss": 0.6236,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0381275953189883,
+      "grad_norm": 0.6878931522369385,
+      "learning_rate": 9.990417089752324e-05,
+      "loss": 0.6211,
+      "step": 2750
+    },
+    {
+      "epoch": 1.0419026047565119,
+      "grad_norm": 0.5037298798561096,
+      "learning_rate": 9.990159580645001e-05,
+      "loss": 0.6128,
+      "step": 2760
+    },
+    {
+      "epoch": 1.0456776141940356,
+      "grad_norm": 0.5697045922279358,
+      "learning_rate": 9.98989866081066e-05,
+      "loss": 0.6088,
+      "step": 2770
+    },
+    {
+      "epoch": 1.0494526236315591,
+      "grad_norm": 0.6046783328056335,
+      "learning_rate": 9.989634330427636e-05,
+      "loss": 0.624,
+      "step": 2780
+    },
+    {
+      "epoch": 1.0532276330690826,
+      "grad_norm": 0.6193393468856812,
+      "learning_rate": 9.989366589676596e-05,
+      "loss": 0.6133,
+      "step": 2790
+    },
+    {
+      "epoch": 1.0570026425066064,
+      "grad_norm": 0.6406804323196411,
+      "learning_rate": 9.98909543874054e-05,
+      "loss": 0.6164,
+      "step": 2800
+    },
+    {
+      "epoch": 1.0607776519441299,
+      "grad_norm": 0.6087714433670044,
+      "learning_rate": 9.988820877804795e-05,
+      "loss": 0.6113,
+      "step": 2810
+    },
+    {
+      "epoch": 1.0645526613816534,
+      "grad_norm": 0.4003846347332001,
+      "learning_rate": 9.988542907057022e-05,
+      "loss": 0.6065,
+      "step": 2820
+    },
+    {
+      "epoch": 1.0683276708191771,
+      "grad_norm": 0.5162433385848999,
+      "learning_rate": 9.988261526687212e-05,
+      "loss": 0.6105,
+      "step": 2830
+    },
+    {
+      "epoch": 1.0721026802567006,
+      "grad_norm": 0.5106794238090515,
+      "learning_rate": 9.987976736887685e-05,
+      "loss": 0.6105,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0758776896942241,
+      "grad_norm": 0.5757883787155151,
+      "learning_rate": 9.987688537853091e-05,
+      "loss": 0.6105,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0796526991317479,
+      "grad_norm": 0.5995069742202759,
+      "learning_rate": 9.987396929780413e-05,
+      "loss": 0.6119,
+      "step": 2860
+    },
+    {
+      "epoch": 1.0834277085692714,
+      "grad_norm": 0.5145584940910339,
+      "learning_rate": 9.987101912868962e-05,
+      "loss": 0.6108,
+      "step": 2870
+    },
+    {
+      "epoch": 1.087202718006795,
+      "grad_norm": 0.418322890996933,
+      "learning_rate": 9.98680348732038e-05,
+      "loss": 0.6076,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0909777274443186,
+      "grad_norm": 0.526392936706543,
+      "learning_rate": 9.986501653338636e-05,
+      "loss": 0.61,
+      "step": 2890
+    },
+    {
+      "epoch": 1.0947527368818422,
+      "grad_norm": 0.6863415837287903,
+      "learning_rate": 9.986196411130031e-05,
+      "loss": 0.6483,
+      "step": 2900
+    },
+    {
+      "epoch": 1.098527746319366,
+      "grad_norm": 0.8329252600669861,
+      "learning_rate": 9.985887760903197e-05,
+      "loss": 0.6162,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1023027557568894,
+      "grad_norm": 0.6749866008758545,
+      "learning_rate": 9.985575702869093e-05,
+      "loss": 0.6157,
+      "step": 2920
+    },
+    {
+      "epoch": 1.106077765194413,
+      "grad_norm": 0.6485674381256104,
+      "learning_rate": 9.985260237241008e-05,
+      "loss": 0.6161,
+      "step": 2930
+    },
+    {
+      "epoch": 1.1098527746319367,
+      "grad_norm": 0.5751580595970154,
+      "learning_rate": 9.984941364234557e-05,
+      "loss": 0.6196,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1136277840694602,
+      "grad_norm": 0.539655864238739,
+      "learning_rate": 9.98461908406769e-05,
+      "loss": 0.6143,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1174027935069837,
+      "grad_norm": 0.5958253145217896,
+      "learning_rate": 9.98429339696068e-05,
+      "loss": 0.6111,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1211778029445074,
+      "grad_norm": 0.6310438513755798,
+      "learning_rate": 9.983964303136133e-05,
+      "loss": 0.6088,
+      "step": 2970
+    },
+    {
+      "epoch": 1.124952812382031,
+      "grad_norm": 0.5665124654769897,
+      "learning_rate": 9.983631802818981e-05,
+      "loss": 0.6046,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1287278218195547,
+      "grad_norm": 0.4563884437084198,
+      "learning_rate": 9.983295896236484e-05,
+      "loss": 0.5985,
+      "step": 2990
+    },
+    {
+      "epoch": 1.1325028312570782,
+      "grad_norm": 0.5045254230499268,
+      "learning_rate": 9.982956583618232e-05,
+      "loss": 0.6024,
+      "step": 3000
+    },
+    {
+      "epoch": 1.1362778406946017,
+      "grad_norm": 0.5779048800468445,
+      "learning_rate": 9.982613865196142e-05,
+      "loss": 0.6088,
+      "step": 3010
+    },
+    {
+      "epoch": 1.1400528501321254,
+      "grad_norm": 0.5393741726875305,
+      "learning_rate": 9.982267741204458e-05,
+      "loss": 0.6145,
+      "step": 3020
+    },
+    {
+      "epoch": 1.143827859569649,
+      "grad_norm": 0.5488253831863403,
+      "learning_rate": 9.981918211879753e-05,
+      "loss": 0.6019,
+      "step": 3030
+    },
+    {
+      "epoch": 1.1476028690071725,
+      "grad_norm": 0.5438612103462219,
+      "learning_rate": 9.981565277460927e-05,
+      "loss": 0.6165,
+      "step": 3040
+    },
+    {
+      "epoch": 1.1513778784446962,
+      "grad_norm": 0.5116326808929443,
+      "learning_rate": 9.981208938189206e-05,
+      "loss": 0.6062,
+      "step": 3050
+    },
+    {
+      "epoch": 1.1551528878822197,
+      "grad_norm": 0.48625659942626953,
+      "learning_rate": 9.980849194308149e-05,
+      "loss": 0.6113,
+      "step": 3060
+    },
+    {
+      "epoch": 1.1589278973197432,
+      "grad_norm": 0.5895056128501892,
+      "learning_rate": 9.98048604606363e-05,
+      "loss": 0.6002,
+      "step": 3070
+    },
+    {
+      "epoch": 1.162702906757267,
+      "grad_norm": 0.7021406292915344,
+      "learning_rate": 9.980119493703864e-05,
+      "loss": 0.6139,
+      "step": 3080
+    },
+    {
+      "epoch": 1.1664779161947905,
+      "grad_norm": 0.7085283398628235,
+      "learning_rate": 9.979749537479383e-05,
+      "loss": 0.6058,
+      "step": 3090
+    },
+    {
+      "epoch": 1.170252925632314,
+      "grad_norm": 0.6100856065750122,
+      "learning_rate": 9.979376177643051e-05,
+      "loss": 0.6003,
+      "step": 3100
+    },
+    {
+      "epoch": 1.1740279350698377,
+      "grad_norm": 0.6801294684410095,
+      "learning_rate": 9.978999414450052e-05,
+      "loss": 0.6171,
+      "step": 3110
+    },
+    {
+      "epoch": 1.1778029445073612,
+      "grad_norm": 0.47147953510284424,
+      "learning_rate": 9.978619248157904e-05,
+      "loss": 0.6057,
+      "step": 3120
+    },
+    {
+      "epoch": 1.1815779539448847,
+      "grad_norm": 0.5377033352851868,
+      "learning_rate": 9.978235679026445e-05,
+      "loss": 0.603,
+      "step": 3130
+    },
+    {
+      "epoch": 1.1853529633824085,
+      "grad_norm": 0.6004739999771118,
+      "learning_rate": 9.97784870731784e-05,
+      "loss": 0.5997,
+      "step": 3140
+    },
+    {
+      "epoch": 1.189127972819932,
+      "grad_norm": 0.42120230197906494,
+      "learning_rate": 9.977458333296584e-05,
+      "loss": 0.6065,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1929029822574557,
+      "grad_norm": 0.4445311427116394,
+      "learning_rate": 9.977064557229492e-05,
+      "loss": 0.6049,
+      "step": 3160
+    },
+    {
+      "epoch": 1.1966779916949792,
+      "grad_norm": 0.5501314997673035,
+      "learning_rate": 9.976667379385705e-05,
+      "loss": 0.6177,
+      "step": 3170
+    },
+    {
+      "epoch": 1.2004530011325028,
+      "grad_norm": 0.6366961598396301,
+      "learning_rate": 9.97626680003669e-05,
+      "loss": 0.6134,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2042280105700265,
+      "grad_norm": 0.5312647223472595,
+      "learning_rate": 9.975862819456242e-05,
+      "loss": 0.5979,
+      "step": 3190
+    },
+    {
+      "epoch": 1.20800302000755,
+      "grad_norm": 0.5397608876228333,
+      "learning_rate": 9.975455437920477e-05,
+      "loss": 0.5947,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2117780294450737,
+      "grad_norm": 0.4828698933124542,
+      "learning_rate": 9.975044655707834e-05,
+      "loss": 0.6012,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2155530388825972,
+      "grad_norm": 0.5150812864303589,
+      "learning_rate": 9.974630473099082e-05,
+      "loss": 0.604,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2193280483201208,
+      "grad_norm": 0.6184629201889038,
+      "learning_rate": 9.974212890377311e-05,
+      "loss": 0.5917,
+      "step": 3230
+    },
+    {
+      "epoch": 1.2231030577576445,
+      "grad_norm": 0.48113495111465454,
+      "learning_rate": 9.973791907827931e-05,
+      "loss": 0.5953,
+      "step": 3240
+    },
+    {
+      "epoch": 1.226878067195168,
+      "grad_norm": 0.4420939087867737,
+      "learning_rate": 9.973367525738683e-05,
+      "loss": 0.6002,
+      "step": 3250
+    },
+    {
+      "epoch": 1.2306530766326915,
+      "grad_norm": 0.610998272895813,
+      "learning_rate": 9.972939744399627e-05,
+      "loss": 0.5944,
+      "step": 3260
+    },
+    {
+      "epoch": 1.2344280860702153,
+      "grad_norm": 0.5040599703788757,
+      "learning_rate": 9.97250856410315e-05,
+      "loss": 0.5988,
+      "step": 3270
+    },
+    {
+      "epoch": 1.2382030955077388,
+      "grad_norm": 0.5534839630126953,
+      "learning_rate": 9.972073985143955e-05,
+      "loss": 0.5982,
+      "step": 3280
+    },
+    {
+      "epoch": 1.2419781049452623,
+      "grad_norm": 0.5562023520469666,
+      "learning_rate": 9.971636007819074e-05,
+      "loss": 0.6003,
+      "step": 3290
+    },
+    {
+      "epoch": 1.245753114382786,
+      "grad_norm": 0.6595808863639832,
+      "learning_rate": 9.971194632427863e-05,
+      "loss": 0.602,
+      "step": 3300
+    },
+    {
+      "epoch": 1.2495281238203095,
+      "grad_norm": 0.7071061134338379,
+      "learning_rate": 9.970749859271997e-05,
+      "loss": 0.6124,
+      "step": 3310
+    },
+    {
+      "epoch": 1.253303133257833,
+      "grad_norm": 0.7782847285270691,
+      "learning_rate": 9.970301688655473e-05,
+      "loss": 0.5991,
+      "step": 3320
+    },
+    {
+      "epoch": 1.2570781426953568,
+      "grad_norm": 0.7165196537971497,
+      "learning_rate": 9.969850120884612e-05,
+      "loss": 0.5957,
+      "step": 3330
+    },
+    {
+      "epoch": 1.2608531521328803,
+      "grad_norm": 0.6089136004447937,
+      "learning_rate": 9.969395156268055e-05,
+      "loss": 0.5891,
+      "step": 3340
+    },
+    {
+      "epoch": 1.2646281615704038,
+      "grad_norm": 0.586713969707489,
+      "learning_rate": 9.968936795116768e-05,
+      "loss": 0.594,
+      "step": 3350
+    },
+    {
+      "epoch": 1.2684031710079275,
+      "grad_norm": 0.6137309074401855,
+      "learning_rate": 9.968475037744036e-05,
+      "loss": 0.6046,
+      "step": 3360
+    },
+    {
+      "epoch": 1.272178180445451,
+      "grad_norm": 0.5665665864944458,
+      "learning_rate": 9.968009884465465e-05,
+      "loss": 0.5905,
+      "step": 3370
+    },
+    {
+      "epoch": 1.2759531898829746,
+      "grad_norm": 0.5130990743637085,
+      "learning_rate": 9.967541335598984e-05,
+      "loss": 0.5971,
+      "step": 3380
+    },
+    {
+      "epoch": 1.2797281993204983,
+      "grad_norm": 0.6189398765563965,
+      "learning_rate": 9.967069391464841e-05,
+      "loss": 0.5988,
+      "step": 3390
+    },
+    {
+      "epoch": 1.2835032087580218,
+      "grad_norm": 0.5427692532539368,
+      "learning_rate": 9.966594052385608e-05,
+      "loss": 0.5979,
+      "step": 3400
+    },
+    {
+      "epoch": 1.2872782181955456,
+      "grad_norm": 0.5762437582015991,
+      "learning_rate": 9.96611531868617e-05,
+      "loss": 0.5932,
+      "step": 3410
+    },
+    {
+      "epoch": 1.291053227633069,
+      "grad_norm": 0.5316555500030518,
+      "learning_rate": 9.96563319069374e-05,
+      "loss": 0.5962,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2948282370705928,
+      "grad_norm": 0.52821284532547,
+      "learning_rate": 9.965147668737847e-05,
+      "loss": 0.5899,
+      "step": 3430
+    },
+    {
+      "epoch": 1.2986032465081163,
+      "grad_norm": 0.6211134195327759,
+      "learning_rate": 9.96465875315034e-05,
+      "loss": 0.595,
+      "step": 3440
+    },
+    {
+      "epoch": 1.3023782559456398,
+      "grad_norm": 0.4804578125476837,
+      "learning_rate": 9.96416644426539e-05,
+      "loss": 0.5943,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3061532653831636,
+      "grad_norm": 0.520340621471405,
+      "learning_rate": 9.963670742419485e-05,
+      "loss": 0.585,
+      "step": 3460
+    },
+    {
+      "epoch": 1.309928274820687,
+      "grad_norm": 0.5734260678291321,
+      "learning_rate": 9.96317164795143e-05,
+      "loss": 0.5853,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3137032842582106,
+      "grad_norm": 0.6706385612487793,
+      "learning_rate": 9.962669161202356e-05,
+      "loss": 0.6087,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3174782936957343,
+      "grad_norm": 0.5032151341438293,
+      "learning_rate": 9.962163282515705e-05,
+      "loss": 0.5877,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3212533031332578,
+      "grad_norm": 0.5707389116287231,
+      "learning_rate": 9.961654012237241e-05,
+      "loss": 0.5801,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3250283125707814,
+      "grad_norm": 0.5376269817352295,
+      "learning_rate": 9.961141350715044e-05,
+      "loss": 0.5817,
+      "step": 3510
+    },
+    {
+      "epoch": 1.328803322008305,
+      "grad_norm": 0.4683854579925537,
+      "learning_rate": 9.960625298299516e-05,
+      "loss": 0.5847,
+      "step": 3520
+    },
+    {
+      "epoch": 1.3325783314458286,
+      "grad_norm": 0.4962872266769409,
+      "learning_rate": 9.960105855343372e-05,
+      "loss": 0.5854,
+      "step": 3530
+    },
+    {
+      "epoch": 1.3363533408833521,
+      "grad_norm": 0.4632238745689392,
+      "learning_rate": 9.959583022201647e-05,
+      "loss": 0.5844,
+      "step": 3540
+    },
+    {
+      "epoch": 1.3401283503208758,
+      "grad_norm": 0.49450498819351196,
+      "learning_rate": 9.959056799231692e-05,
+      "loss": 0.5877,
+      "step": 3550
+    },
+    {
+      "epoch": 1.3439033597583994,
+      "grad_norm": 0.45359042286872864,
+      "learning_rate": 9.958527186793176e-05,
+      "loss": 0.5798,
+      "step": 3560
+    },
+    {
+      "epoch": 1.3476783691959229,
+      "grad_norm": 0.48816585540771484,
+      "learning_rate": 9.957994185248086e-05,
+      "loss": 0.587,
+      "step": 3570
+    },
+    {
+      "epoch": 1.3514533786334466,
+      "grad_norm": 0.5069116353988647,
+      "learning_rate": 9.957457794960718e-05,
+      "loss": 0.5972,
+      "step": 3580
+    },
+    {
+      "epoch": 1.3552283880709701,
+      "grad_norm": 0.4354189336299896,
+      "learning_rate": 9.956918016297694e-05,
+      "loss": 0.5872,
+      "step": 3590
+    },
+    {
+      "epoch": 1.3590033975084936,
+      "grad_norm": 0.5266293883323669,
+      "learning_rate": 9.956374849627948e-05,
+      "loss": 0.5869,
+      "step": 3600
+    },
+    {
+      "epoch": 1.3627784069460174,
+      "grad_norm": 0.4389692544937134,
+      "learning_rate": 9.955828295322728e-05,
+      "loss": 0.5792,
+      "step": 3610
+    },
+    {
+      "epoch": 1.3665534163835409,
+      "grad_norm": 0.5600664615631104,
+      "learning_rate": 9.955278353755598e-05,
+      "loss": 0.582,
+      "step": 3620
+    },
+    {
+      "epoch": 1.3703284258210646,
+      "grad_norm": 0.6405794024467468,
+      "learning_rate": 9.954725025302439e-05,
+      "loss": 0.5898,
+      "step": 3630
+    },
+    {
+      "epoch": 1.3741034352585881,
+      "grad_norm": 0.5375349521636963,
+      "learning_rate": 9.954168310341445e-05,
+      "loss": 0.5863,
+      "step": 3640
+    },
+    {
+      "epoch": 1.3778784446961119,
+      "grad_norm": 0.6790446043014526,
+      "learning_rate": 9.953608209253126e-05,
+      "loss": 0.5904,
+      "step": 3650
+    },
+    {
+      "epoch": 1.3816534541336354,
+      "grad_norm": 0.4370107352733612,
+      "learning_rate": 9.953044722420307e-05,
+      "loss": 0.5874,
+      "step": 3660
+    },
+    {
+      "epoch": 1.385428463571159,
+      "grad_norm": 0.524257242679596,
+      "learning_rate": 9.952477850228124e-05,
+      "loss": 0.5879,
+      "step": 3670
+    },
+    {
+      "epoch": 1.3892034730086826,
+      "grad_norm": 0.4980759918689728,
+      "learning_rate": 9.95190759306403e-05,
+      "loss": 0.5817,
+      "step": 3680
+    },
+    {
+      "epoch": 1.3929784824462061,
+      "grad_norm": 0.5062684416770935,
+      "learning_rate": 9.951333951317789e-05,
+      "loss": 0.5827,
+      "step": 3690
+    },
+    {
+      "epoch": 1.3967534918837297,
+      "grad_norm": 0.7371021509170532,
+      "learning_rate": 9.950756925381479e-05,
+      "loss": 0.5829,
+      "step": 3700
+    },
+    {
+      "epoch": 1.4005285013212534,
+      "grad_norm": 0.5569843649864197,
+      "learning_rate": 9.950176515649496e-05,
+      "loss": 0.5779,
+      "step": 3710
+    },
+    {
+      "epoch": 1.404303510758777,
+      "grad_norm": 0.5766958594322205,
+      "learning_rate": 9.949592722518542e-05,
+      "loss": 0.5861,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4080785201963004,
+      "grad_norm": 0.5430606603622437,
+      "learning_rate": 9.949005546387631e-05,
+      "loss": 0.579,
+      "step": 3730
+    },
+    {
+      "epoch": 1.4118535296338242,
+      "grad_norm": 0.49899691343307495,
+      "learning_rate": 9.948414987658098e-05,
+      "loss": 0.5759,
+      "step": 3740
+    },
+    {
+      "epoch": 1.4156285390713477,
+      "grad_norm": 0.54237300157547,
+      "learning_rate": 9.94782104673358e-05,
+      "loss": 0.5809,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4194035485088712,
+      "grad_norm": 0.6004769802093506,
+      "learning_rate": 9.947223724020034e-05,
+      "loss": 0.5743,
+      "step": 3760
+    },
+    {
+      "epoch": 1.423178557946395,
+      "grad_norm": 0.5146884322166443,
+      "learning_rate": 9.94662301992572e-05,
+      "loss": 0.5782,
+      "step": 3770
+    },
+    {
+      "epoch": 1.4269535673839184,
+      "grad_norm": 0.5596660375595093,
+      "learning_rate": 9.946018934861216e-05,
+      "loss": 0.5787,
+      "step": 3780
+    },
+    {
+      "epoch": 1.430728576821442,
+      "grad_norm": 0.5646491050720215,
+      "learning_rate": 9.945411469239409e-05,
+      "loss": 0.5871,
+      "step": 3790
+    },
+    {
+      "epoch": 1.4345035862589657,
+      "grad_norm": 0.6031035780906677,
+      "learning_rate": 9.944800623475497e-05,
+      "loss": 0.5735,
+      "step": 3800
+    },
+    {
+      "epoch": 1.4382785956964892,
+      "grad_norm": 0.6148284077644348,
+      "learning_rate": 9.944186397986984e-05,
+      "loss": 0.5807,
+      "step": 3810
+    },
+    {
+      "epoch": 1.4420536051340127,
+      "grad_norm": 0.541500449180603,
+      "learning_rate": 9.943568793193687e-05,
+      "loss": 0.5806,
+      "step": 3820
+    },
+    {
+      "epoch": 1.4458286145715364,
+      "grad_norm": 0.5454701781272888,
+      "learning_rate": 9.942947809517737e-05,
+      "loss": 0.584,
+      "step": 3830
+    },
+    {
+      "epoch": 1.44960362400906,
+      "grad_norm": 0.6133937835693359,
+      "learning_rate": 9.942323447383568e-05,
+      "loss": 0.5811,
+      "step": 3840
+    },
+    {
+      "epoch": 1.4533786334465837,
+      "grad_norm": 0.5456222295761108,
+      "learning_rate": 9.941695707217925e-05,
+      "loss": 0.5747,
+      "step": 3850
+    },
+    {
+      "epoch": 1.4571536428841072,
+      "grad_norm": 0.5243522524833679,
+      "learning_rate": 9.941064589449865e-05,
+      "loss": 0.5792,
+      "step": 3860
+    },
+    {
+      "epoch": 1.460928652321631,
+      "grad_norm": 0.5548741221427917,
+      "learning_rate": 9.940430094510748e-05,
+      "loss": 0.5756,
+      "step": 3870
+    },
+    {
+      "epoch": 1.4647036617591545,
+      "grad_norm": 0.65179044008255,
+      "learning_rate": 9.939792222834248e-05,
+      "loss": 0.5788,
+      "step": 3880
+    },
+    {
+      "epoch": 1.468478671196678,
+      "grad_norm": 0.6072384715080261,
+      "learning_rate": 9.939150974856343e-05,
+      "loss": 0.5779,
+      "step": 3890
+    },
+    {
+      "epoch": 1.4722536806342017,
+      "grad_norm": 0.6061842441558838,
+      "learning_rate": 9.93850635101532e-05,
+      "loss": 0.5735,
+      "step": 3900
+    },
+    {
+      "epoch": 1.4760286900717252,
+      "grad_norm": 0.5786122679710388,
+      "learning_rate": 9.937858351751775e-05,
+      "loss": 0.5794,
+      "step": 3910
+    },
+    {
+      "epoch": 1.4798036995092487,
+      "grad_norm": 0.5088226199150085,
+      "learning_rate": 9.937206977508604e-05,
+      "loss": 0.5737,
+      "step": 3920
+    },
+    {
+      "epoch": 1.4835787089467725,
+      "grad_norm": 0.5596620440483093,
+      "learning_rate": 9.936552228731022e-05,
+      "loss": 0.5912,
+      "step": 3930
+    },
+    {
+      "epoch": 1.487353718384296,
+      "grad_norm": 0.4660915732383728,
+      "learning_rate": 9.93589410586654e-05,
+      "loss": 0.5735,
+      "step": 3940
+    },
+    {
+      "epoch": 1.4911287278218195,
+      "grad_norm": 0.4994395971298218,
+      "learning_rate": 9.93523260936498e-05,
+      "loss": 0.5807,
+      "step": 3950
+    },
+    {
+      "epoch": 1.4949037372593432,
+      "grad_norm": 0.4547739624977112,
+      "learning_rate": 9.934567739678467e-05,
+      "loss": 0.5723,
+      "step": 3960
+    },
+    {
+      "epoch": 1.4986787466968667,
+      "grad_norm": 0.4519752264022827,
+      "learning_rate": 9.933899497261433e-05,
+      "loss": 0.5749,
+      "step": 3970
+    },
+    {
+      "epoch": 1.5024537561343903,
+      "grad_norm": 0.49861350655555725,
+      "learning_rate": 9.933227882570617e-05,
+      "loss": 0.5825,
+      "step": 3980
+    },
+    {
+      "epoch": 1.506228765571914,
+      "grad_norm": 0.5241252183914185,
+      "learning_rate": 9.932552896065062e-05,
+      "loss": 0.582,
+      "step": 3990
+    },
+    {
+      "epoch": 1.5100037750094375,
+      "grad_norm": 0.6853564977645874,
+      "learning_rate": 9.931874538206114e-05,
+      "loss": 0.5793,
+      "step": 4000
+    },
+    {
+      "epoch": 1.513778784446961,
+      "grad_norm": 0.47244563698768616,
+      "learning_rate": 9.931192809457423e-05,
+      "loss": 0.5737,
+      "step": 4010
+    },
+    {
+      "epoch": 1.5175537938844847,
+      "grad_norm": 0.5106404423713684,
+      "learning_rate": 9.930507710284944e-05,
+      "loss": 0.5773,
+      "step": 4020
+    },
+    {
+      "epoch": 1.5213288033220083,
+      "grad_norm": 0.5653221011161804,
+      "learning_rate": 9.92981924115694e-05,
+      "loss": 0.5729,
+      "step": 4030
+    },
+    {
+      "epoch": 1.5251038127595318,
+      "grad_norm": 0.7388667464256287,
+      "learning_rate": 9.929127402543968e-05,
+      "loss": 0.5695,
+      "step": 4040
+    },
+    {
+      "epoch": 1.5288788221970555,
+      "grad_norm": 0.55259770154953,
+      "learning_rate": 9.928432194918895e-05,
+      "loss": 0.5715,
+      "step": 4050
+    },
+    {
+      "epoch": 1.5326538316345792,
+      "grad_norm": 0.5290758609771729,
+      "learning_rate": 9.927733618756889e-05,
+      "loss": 0.5753,
+      "step": 4060
+    },
+    {
+      "epoch": 1.5364288410721025,
+      "grad_norm": 0.48387640714645386,
+      "learning_rate": 9.92703167453542e-05,
+      "loss": 0.5744,
+      "step": 4070
+    },
+    {
+      "epoch": 1.5402038505096263,
+      "grad_norm": 0.5005270838737488,
+      "learning_rate": 9.926326362734263e-05,
+      "loss": 0.5767,
+      "step": 4080
+    },
+    {
+      "epoch": 1.54397885994715,
+      "grad_norm": 0.5427155494689941,
+      "learning_rate": 9.925617683835489e-05,
+      "loss": 0.5726,
+      "step": 4090
+    },
+    {
+      "epoch": 1.5477538693846733,
+      "grad_norm": 0.5388875603675842,
+      "learning_rate": 9.924905638323472e-05,
+      "loss": 0.5673,
+      "step": 4100
+    },
+    {
+      "epoch": 1.551528878822197,
+      "grad_norm": 0.48597651720046997,
+      "learning_rate": 9.92419022668489e-05,
+      "loss": 0.5684,
+      "step": 4110
+    },
+    {
+      "epoch": 1.5553038882597208,
+      "grad_norm": 0.5301367044448853,
+      "learning_rate": 9.923471449408723e-05,
+      "loss": 0.5752,
+      "step": 4120
+    },
+    {
+      "epoch": 1.5590788976972443,
+      "grad_norm": 0.5001996159553528,
+      "learning_rate": 9.922749306986244e-05,
+      "loss": 0.5629,
+      "step": 4130
+    },
+    {
+      "epoch": 1.5628539071347678,
+      "grad_norm": 0.5677825212478638,
+      "learning_rate": 9.922023799911033e-05,
+      "loss": 0.5742,
+      "step": 4140
+    },
+    {
+      "epoch": 1.5666289165722915,
+      "grad_norm": 0.5715368390083313,
+      "learning_rate": 9.921294928678968e-05,
+      "loss": 0.5751,
+      "step": 4150
+    },
+    {
+      "epoch": 1.570403926009815,
+      "grad_norm": 0.632077157497406,
+      "learning_rate": 9.920562693788222e-05,
+      "loss": 0.5702,
+      "step": 4160
+    },
+    {
+      "epoch": 1.5741789354473386,
+      "grad_norm": 0.5607465505599976,
+      "learning_rate": 9.919827095739273e-05,
+      "loss": 0.5713,
+      "step": 4170
+    },
+    {
+      "epoch": 1.5779539448848623,
+      "grad_norm": 0.5231792330741882,
+      "learning_rate": 9.919088135034898e-05,
+      "loss": 0.5696,
+      "step": 4180
+    },
+    {
+      "epoch": 1.5817289543223858,
+      "grad_norm": 0.46094974875450134,
+      "learning_rate": 9.918345812180165e-05,
+      "loss": 0.5599,
+      "step": 4190
+    },
+    {
+      "epoch": 1.5855039637599093,
+      "grad_norm": 0.5033226609230042,
+      "learning_rate": 9.917600127682446e-05,
+      "loss": 0.5706,
+      "step": 4200
+    },
+    {
+      "epoch": 1.589278973197433,
+      "grad_norm": 0.5279197096824646,
+      "learning_rate": 9.91685108205141e-05,
+      "loss": 0.5674,
+      "step": 4210
+    },
+    {
+      "epoch": 1.5930539826349566,
+      "grad_norm": 0.49496957659721375,
+      "learning_rate": 9.916098675799024e-05,
+      "loss": 0.5644,
+      "step": 4220
+    },
+    {
+      "epoch": 1.59682899207248,
+      "grad_norm": 0.5188071131706238,
+      "learning_rate": 9.915342909439548e-05,
+      "loss": 0.5699,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6006040015100038,
+      "grad_norm": 0.6602573394775391,
+      "learning_rate": 9.914583783489543e-05,
+      "loss": 0.5699,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6043790109475273,
+      "grad_norm": 0.44954216480255127,
+      "learning_rate": 9.913821298467863e-05,
+      "loss": 0.5622,
+      "step": 4250
+    },
+    {
+      "epoch": 1.6081540203850508,
+      "grad_norm": 0.40007320046424866,
+      "learning_rate": 9.913055454895661e-05,
+      "loss": 0.5649,
+      "step": 4260
+    },
+    {
+      "epoch": 1.6119290298225746,
+      "grad_norm": 0.5066043734550476,
+      "learning_rate": 9.912286253296381e-05,
+      "loss": 0.5689,
+      "step": 4270
+    },
+    {
+      "epoch": 1.6157040392600983,
+      "grad_norm": 0.5492740273475647,
+      "learning_rate": 9.911513694195768e-05,
+      "loss": 0.5592,
+      "step": 4280
+    },
+    {
+      "epoch": 1.6194790486976216,
+      "grad_norm": 0.40257328748703003,
+      "learning_rate": 9.910737778121859e-05,
+      "loss": 0.5671,
+      "step": 4290
+    },
+    {
+      "epoch": 1.6232540581351453,
+      "grad_norm": 0.42792394757270813,
+      "learning_rate": 9.909958505604984e-05,
+      "loss": 0.5624,
+      "step": 4300
+    },
+    {
+      "epoch": 1.627029067572669,
+      "grad_norm": 0.5710054039955139,
+      "learning_rate": 9.909175877177767e-05,
+      "loss": 0.5697,
+      "step": 4310
+    },
+    {
+      "epoch": 1.6308040770101924,
+      "grad_norm": 0.47160592675209045,
+      "learning_rate": 9.908389893375129e-05,
+      "loss": 0.5684,
+      "step": 4320
+    },
+    {
+      "epoch": 1.634579086447716,
+      "grad_norm": 0.6177746653556824,
+      "learning_rate": 9.907600554734283e-05,
+      "loss": 0.5724,
+      "step": 4330
+    },
+    {
+      "epoch": 1.6383540958852398,
+      "grad_norm": 0.56391841173172,
+      "learning_rate": 9.906807861794734e-05,
+      "loss": 0.5675,
+      "step": 4340
+    },
+    {
+      "epoch": 1.6421291053227633,
+      "grad_norm": 0.6061177849769592,
+      "learning_rate": 9.906011815098279e-05,
+      "loss": 0.5607,
+      "step": 4350
+    },
+    {
+      "epoch": 1.6459041147602869,
+      "grad_norm": 0.580389678478241,
+      "learning_rate": 9.90521241518901e-05,
+      "loss": 0.563,
+      "step": 4360
+    },
+    {
+      "epoch": 1.6496791241978106,
+      "grad_norm": 0.6616781949996948,
+      "learning_rate": 9.904409662613308e-05,
+      "loss": 0.5664,
+      "step": 4370
+    },
+    {
+      "epoch": 1.6534541336353341,
+      "grad_norm": 0.5005329847335815,
+      "learning_rate": 9.903603557919849e-05,
+      "loss": 0.5592,
+      "step": 4380
+    },
+    {
+      "epoch": 1.6572291430728576,
+      "grad_norm": 0.4905034601688385,
+      "learning_rate": 9.902794101659594e-05,
+      "loss": 0.5678,
+      "step": 4390
+    },
+    {
+      "epoch": 1.6610041525103814,
+      "grad_norm": 0.44465407729148865,
+      "learning_rate": 9.901981294385803e-05,
+      "loss": 0.5684,
+      "step": 4400
+    },
+    {
+      "epoch": 1.6647791619479049,
+      "grad_norm": 0.5083572864532471,
+      "learning_rate": 9.901165136654018e-05,
+      "loss": 0.561,
+      "step": 4410
+    },
+    {
+      "epoch": 1.6685541713854284,
+      "grad_norm": 0.5734196305274963,
+      "learning_rate": 9.900345629022079e-05,
+      "loss": 0.57,
+      "step": 4420
+    },
+    {
+      "epoch": 1.6723291808229521,
+      "grad_norm": 0.568246066570282,
+      "learning_rate": 9.89952277205011e-05,
+      "loss": 0.5666,
+      "step": 4430
+    },
+    {
+      "epoch": 1.6761041902604756,
+      "grad_norm": 0.5339452624320984,
+      "learning_rate": 9.898696566300527e-05,
+      "loss": 0.5599,
+      "step": 4440
+    },
+    {
+      "epoch": 1.6798791996979991,
+      "grad_norm": 0.5378457307815552,
+      "learning_rate": 9.897867012338032e-05,
+      "loss": 0.5605,
+      "step": 4450
+    },
+    {
+      "epoch": 1.6836542091355229,
+      "grad_norm": 0.5373885035514832,
+      "learning_rate": 9.897034110729617e-05,
+      "loss": 0.5641,
+      "step": 4460
+    },
+    {
+      "epoch": 1.6874292185730464,
+      "grad_norm": 1.103470802307129,
+      "learning_rate": 9.896197862044564e-05,
+      "loss": 0.5773,
+      "step": 4470
+    },
+    {
+      "epoch": 1.69120422801057,
+      "grad_norm": 0.6465436220169067,
+      "learning_rate": 9.895358266854442e-05,
+      "loss": 0.5764,
+      "step": 4480
+    },
+    {
+      "epoch": 1.6949792374480936,
+      "grad_norm": 0.5247374176979065,
+      "learning_rate": 9.894515325733103e-05,
+      "loss": 0.5593,
+      "step": 4490
+    },
+    {
+      "epoch": 1.6987542468856174,
+      "grad_norm": 0.589596688747406,
+      "learning_rate": 9.893669039256693e-05,
+      "loss": 0.5637,
+      "step": 4500
+    },
+    {
+      "epoch": 1.7025292563231407,
+      "grad_norm": 0.4953310489654541,
+      "learning_rate": 9.89281940800364e-05,
+      "loss": 0.5622,
+      "step": 4510
+    },
+    {
+      "epoch": 1.7063042657606644,
+      "grad_norm": 0.45567750930786133,
+      "learning_rate": 9.891966432554655e-05,
+      "loss": 0.5594,
+      "step": 4520
+    },
+    {
+      "epoch": 1.7100792751981881,
+      "grad_norm": 0.4500698149204254,
+      "learning_rate": 9.891110113492745e-05,
+      "loss": 0.5619,
+      "step": 4530
+    },
+    {
+      "epoch": 1.7138542846357114,
+      "grad_norm": 0.4617537260055542,
+      "learning_rate": 9.89025045140319e-05,
+      "loss": 0.5623,
+      "step": 4540
+    },
+    {
+      "epoch": 1.7176292940732352,
+      "grad_norm": 0.5609983801841736,
+      "learning_rate": 9.889387446873567e-05,
+      "loss": 0.5629,
+      "step": 4550
+    },
+    {
+      "epoch": 1.721404303510759,
+      "grad_norm": 0.6847432851791382,
+      "learning_rate": 9.888521100493726e-05,
+      "loss": 0.57,
+      "step": 4560
+    },
+    {
+      "epoch": 1.7251793129482824,
+      "grad_norm": 0.50513756275177,
+      "learning_rate": 9.887651412855809e-05,
+      "loss": 0.5586,
+      "step": 4570
+    },
+    {
+      "epoch": 1.728954322385806,
+      "grad_norm": 0.45217183232307434,
+      "learning_rate": 9.88677838455424e-05,
+      "loss": 0.5493,
+      "step": 4580
+    },
+    {
+      "epoch": 1.7327293318233297,
+      "grad_norm": 0.445190966129303,
+      "learning_rate": 9.885902016185725e-05,
+      "loss": 0.5597,
+      "step": 4590
+    },
+    {
+      "epoch": 1.7365043412608532,
+      "grad_norm": 0.49329379200935364,
+      "learning_rate": 9.885022308349252e-05,
+      "loss": 0.5608,
+      "step": 4600
+    },
+    {
+      "epoch": 1.7402793506983767,
+      "grad_norm": 0.531954824924469,
+      "learning_rate": 9.884139261646095e-05,
+      "loss": 0.5646,
+      "step": 4610
+    },
+    {
+      "epoch": 1.7440543601359004,
+      "grad_norm": 0.5033367276191711,
+      "learning_rate": 9.883252876679807e-05,
+      "loss": 0.564,
+      "step": 4620
+    },
+    {
+      "epoch": 1.747829369573424,
+      "grad_norm": 0.8055509328842163,
+      "learning_rate": 9.882363154056225e-05,
+      "loss": 0.5519,
+      "step": 4630
+    },
+    {
+      "epoch": 1.7516043790109475,
+      "grad_norm": 0.592330813407898,
+      "learning_rate": 9.881470094383465e-05,
+      "loss": 0.5565,
+      "step": 4640
+    },
+    {
+      "epoch": 1.7553793884484712,
+      "grad_norm": 0.605402946472168,
+      "learning_rate": 9.880573698271924e-05,
+      "loss": 0.5737,
+      "step": 4650
+    },
+    {
+      "epoch": 1.7591543978859947,
+      "grad_norm": 0.6107518076896667,
+      "learning_rate": 9.879673966334282e-05,
+      "loss": 0.5569,
+      "step": 4660
+    },
+    {
+      "epoch": 1.7629294073235182,
+      "grad_norm": 0.5053714513778687,
+      "learning_rate": 9.878770899185496e-05,
+      "loss": 0.5599,
+      "step": 4670
+    },
+    {
+      "epoch": 1.766704416761042,
+      "grad_norm": 0.4542829394340515,
+      "learning_rate": 9.877864497442804e-05,
+      "loss": 0.5606,
+      "step": 4680
+    },
+    {
+      "epoch": 1.7704794261985655,
+      "grad_norm": 0.6275748610496521,
+      "learning_rate": 9.876954761725723e-05,
+      "loss": 0.5576,
+      "step": 4690
+    },
+    {
+      "epoch": 1.774254435636089,
+      "grad_norm": 0.5331715941429138,
+      "learning_rate": 9.876041692656052e-05,
+      "loss": 0.5567,
+      "step": 4700
+    },
+    {
+      "epoch": 1.7780294450736127,
+      "grad_norm": 0.49501025676727295,
+      "learning_rate": 9.87512529085786e-05,
+      "loss": 0.5617,
+      "step": 4710
+    },
+    {
+      "epoch": 1.7818044545111362,
+      "grad_norm": 0.4898965060710907,
+      "learning_rate": 9.874205556957503e-05,
+      "loss": 0.5482,
+      "step": 4720
+    },
+    {
+      "epoch": 1.7855794639486597,
+      "grad_norm": 0.47128167748451233,
+      "learning_rate": 9.873282491583608e-05,
+      "loss": 0.5579,
+      "step": 4730
+    },
+    {
+      "epoch": 1.7893544733861835,
+      "grad_norm": 0.4385862946510315,
+      "learning_rate": 9.872356095367084e-05,
+      "loss": 0.5551,
+      "step": 4740
+    },
+    {
+      "epoch": 1.7931294828237072,
+      "grad_norm": 0.5858024954795837,
+      "learning_rate": 9.87142636894111e-05,
+      "loss": 0.5474,
+      "step": 4750
+    },
+    {
+      "epoch": 1.7969044922612305,
+      "grad_norm": 0.4746013283729553,
+      "learning_rate": 9.870493312941148e-05,
+      "loss": 0.5596,
+      "step": 4760
+    },
+    {
+      "epoch": 1.8006795016987542,
+      "grad_norm": 0.5050995945930481,
+      "learning_rate": 9.869556928004933e-05,
+      "loss": 0.5451,
+      "step": 4770
+    },
+    {
+      "epoch": 1.804454511136278,
+      "grad_norm": 0.6011149287223816,
+      "learning_rate": 9.868617214772476e-05,
+      "loss": 0.5543,
+      "step": 4780
+    },
+    {
+      "epoch": 1.8082295205738015,
+      "grad_norm": 0.5801728367805481,
+      "learning_rate": 9.867674173886059e-05,
+      "loss": 0.549,
+      "step": 4790
+    },
+    {
+      "epoch": 1.812004530011325,
+      "grad_norm": 0.5891419053077698,
+      "learning_rate": 9.866727805990245e-05,
+      "loss": 0.5539,
+      "step": 4800
+    },
+    {
+      "epoch": 1.8157795394488487,
+      "grad_norm": 0.538972795009613,
+      "learning_rate": 9.865778111731865e-05,
+      "loss": 0.5506,
+      "step": 4810
+    },
+    {
+      "epoch": 1.8195545488863722,
+      "grad_norm": 0.5291750431060791,
+      "learning_rate": 9.864825091760028e-05,
+      "loss": 0.5576,
+      "step": 4820
+    },
+    {
+      "epoch": 1.8233295583238958,
+      "grad_norm": 0.5145063400268555,
+      "learning_rate": 9.863868746726112e-05,
+      "loss": 0.5523,
+      "step": 4830
+    },
+    {
+      "epoch": 1.8271045677614195,
+      "grad_norm": 0.4349626898765564,
+      "learning_rate": 9.862909077283773e-05,
+      "loss": 0.5525,
+      "step": 4840
+    },
+    {
+      "epoch": 1.830879577198943,
+      "grad_norm": 0.5479471683502197,
+      "learning_rate": 9.861946084088933e-05,
+      "loss": 0.5528,
+      "step": 4850
+    },
+    {
+      "epoch": 1.8346545866364665,
+      "grad_norm": 0.7466859817504883,
+      "learning_rate": 9.860979767799792e-05,
+      "loss": 0.5539,
+      "step": 4860
+    },
+    {
+      "epoch": 1.8384295960739903,
+      "grad_norm": 0.5633376836776733,
+      "learning_rate": 9.860010129076813e-05,
+      "loss": 0.5538,
+      "step": 4870
+    },
+    {
+      "epoch": 1.8422046055115138,
+      "grad_norm": 0.49453553557395935,
+      "learning_rate": 9.859037168582741e-05,
+      "loss": 0.5486,
+      "step": 4880
+    },
+    {
+      "epoch": 1.8459796149490373,
+      "grad_norm": 0.5297577381134033,
+      "learning_rate": 9.85806088698258e-05,
+      "loss": 0.5394,
+      "step": 4890
+    },
+    {
+      "epoch": 1.849754624386561,
+      "grad_norm": 0.5536269545555115,
+      "learning_rate": 9.857081284943612e-05,
+      "loss": 0.5494,
+      "step": 4900
+    },
+    {
+      "epoch": 1.8535296338240845,
+      "grad_norm": 0.4998897612094879,
+      "learning_rate": 9.856098363135385e-05,
+      "loss": 0.5459,
+      "step": 4910
+    },
+    {
+      "epoch": 1.857304643261608,
+      "grad_norm": 0.48842960596084595,
+      "learning_rate": 9.855112122229717e-05,
+      "loss": 0.5523,
+      "step": 4920
+    },
+    {
+      "epoch": 1.8610796526991318,
+      "grad_norm": 0.4732976257801056,
+      "learning_rate": 9.854122562900696e-05,
+      "loss": 0.568,
+      "step": 4930
+    },
+    {
+      "epoch": 1.8648546621366553,
+      "grad_norm": 0.47209542989730835,
+      "learning_rate": 9.853129685824673e-05,
+      "loss": 0.5526,
+      "step": 4940
+    },
+    {
+      "epoch": 1.8686296715741788,
+      "grad_norm": 0.4802228510379791,
+      "learning_rate": 9.852133491680273e-05,
+      "loss": 0.5431,
+      "step": 4950
+    },
+    {
+      "epoch": 1.8724046810117025,
+      "grad_norm": 0.507754385471344,
+      "learning_rate": 9.851133981148385e-05,
+      "loss": 0.5486,
+      "step": 4960
+    },
+    {
+      "epoch": 1.8761796904492263,
+      "grad_norm": 0.5926216244697571,
+      "learning_rate": 9.850131154912164e-05,
+      "loss": 0.549,
+      "step": 4970
+    },
+    {
+      "epoch": 1.8799546998867496,
+      "grad_norm": 0.49967560172080994,
+      "learning_rate": 9.849125013657031e-05,
+      "loss": 0.5459,
+      "step": 4980
+    },
+    {
+      "epoch": 1.8837297093242733,
+      "grad_norm": 0.5526815056800842,
+      "learning_rate": 9.848115558070676e-05,
+      "loss": 0.5439,
+      "step": 4990
+    },
+    {
+      "epoch": 1.887504718761797,
+      "grad_norm": 0.5583565831184387,
+      "learning_rate": 9.847102788843055e-05,
+      "loss": 0.5418,
+      "step": 5000
+    },
+    {
+      "epoch": 1.8912797281993206,
+      "grad_norm": 0.5497617125511169,
+      "learning_rate": 9.84608670666638e-05,
+      "loss": 0.5473,
+      "step": 5010
+    },
+    {
+      "epoch": 1.895054737636844,
+      "grad_norm": 0.6256689429283142,
+      "learning_rate": 9.845067312235138e-05,
+      "loss": 0.5485,
+      "step": 5020
+    },
+    {
+      "epoch": 1.8988297470743678,
+      "grad_norm": 0.5378864407539368,
+      "learning_rate": 9.844044606246074e-05,
+      "loss": 0.5432,
+      "step": 5030
+    },
+    {
+      "epoch": 1.9026047565118913,
+      "grad_norm": 0.5322272777557373,
+      "learning_rate": 9.843018589398199e-05,
+      "loss": 0.5605,
+      "step": 5040
+    },
+    {
+      "epoch": 1.9063797659494148,
+      "grad_norm": 0.47538843750953674,
+      "learning_rate": 9.841989262392785e-05,
+      "loss": 0.5452,
+      "step": 5050
+    },
+    {
+      "epoch": 1.9101547753869386,
+      "grad_norm": 0.5338093042373657,
+      "learning_rate": 9.840956625933367e-05,
+      "loss": 0.5443,
+      "step": 5060
+    },
+    {
+      "epoch": 1.913929784824462,
+      "grad_norm": 0.5679603815078735,
+      "learning_rate": 9.839920680725743e-05,
+      "loss": 0.5463,
+      "step": 5070
+    },
+    {
+      "epoch": 1.9177047942619856,
+      "grad_norm": 0.5604372024536133,
+      "learning_rate": 9.83888142747797e-05,
+      "loss": 0.5508,
+      "step": 5080
+    },
+    {
+      "epoch": 1.9214798036995093,
+      "grad_norm": 0.6041125655174255,
+      "learning_rate": 9.83783886690037e-05,
+      "loss": 0.5441,
+      "step": 5090
+    },
+    {
+      "epoch": 1.9252548131370328,
+      "grad_norm": 0.6924186944961548,
+      "learning_rate": 9.836792999705524e-05,
+      "loss": 0.546,
+      "step": 5100
+    },
+    {
+      "epoch": 1.9290298225745564,
+      "grad_norm": 0.4839281737804413,
+      "learning_rate": 9.835743826608271e-05,
+      "loss": 0.5505,
+      "step": 5110
+    },
+    {
+      "epoch": 1.93280483201208,
+      "grad_norm": 0.49724170565605164,
+      "learning_rate": 9.834691348325709e-05,
+      "loss": 0.5384,
+      "step": 5120
+    },
+    {
+      "epoch": 1.9365798414496036,
+      "grad_norm": 0.5565958619117737,
+      "learning_rate": 9.833635565577199e-05,
+      "loss": 0.5432,
+      "step": 5130
+    },
+    {
+      "epoch": 1.9403548508871271,
+      "grad_norm": 0.5241300463676453,
+      "learning_rate": 9.832576479084359e-05,
+      "loss": 0.5466,
+      "step": 5140
+    },
+    {
+      "epoch": 1.9441298603246508,
+      "grad_norm": 0.4891330599784851,
+      "learning_rate": 9.831514089571064e-05,
+      "loss": 0.5466,
+      "step": 5150
+    },
+    {
+      "epoch": 1.9479048697621744,
+      "grad_norm": 0.582870364189148,
+      "learning_rate": 9.830448397763447e-05,
+      "loss": 0.5394,
+      "step": 5160
+    },
+    {
+      "epoch": 1.9516798791996979,
+      "grad_norm": 0.5557212829589844,
+      "learning_rate": 9.8293794043899e-05,
+      "loss": 0.5426,
+      "step": 5170
+    },
+    {
+      "epoch": 1.9554548886372216,
+      "grad_norm": 0.5366427302360535,
+      "learning_rate": 9.828307110181066e-05,
+      "loss": 0.5488,
+      "step": 5180
+    },
+    {
+      "epoch": 1.9592298980747453,
+      "grad_norm": 0.512102484703064,
+      "learning_rate": 9.827231515869852e-05,
+      "loss": 0.5463,
+      "step": 5190
+    },
+    {
+      "epoch": 1.9630049075122686,
+      "grad_norm": 0.4691350758075714,
+      "learning_rate": 9.826152622191413e-05,
+      "loss": 0.5435,
+      "step": 5200
+    },
+    {
+      "epoch": 1.9667799169497924,
+      "grad_norm": 0.5275509357452393,
+      "learning_rate": 9.825070429883167e-05,
+      "loss": 0.5343,
+      "step": 5210
+    },
+    {
+      "epoch": 1.970554926387316,
+      "grad_norm": 0.4976699948310852,
+      "learning_rate": 9.82398493968478e-05,
+      "loss": 0.5423,
+      "step": 5220
+    },
+    {
+      "epoch": 1.9743299358248394,
+      "grad_norm": 0.5303046107292175,
+      "learning_rate": 9.822896152338173e-05,
+      "loss": 0.5428,
+      "step": 5230
+    },
+    {
+      "epoch": 1.9781049452623631,
+      "grad_norm": 0.47112399339675903,
+      "learning_rate": 9.821804068587523e-05,
+      "loss": 0.5357,
+      "step": 5240
+    },
+    {
+      "epoch": 1.9818799546998869,
+      "grad_norm": 0.4589209258556366,
+      "learning_rate": 9.820708689179259e-05,
+      "loss": 0.5445,
+      "step": 5250
+    },
+    {
+      "epoch": 1.9856549641374104,
+      "grad_norm": 0.47271859645843506,
+      "learning_rate": 9.819610014862063e-05,
+      "loss": 0.5339,
+      "step": 5260
+    },
+    {
+      "epoch": 1.989429973574934,
+      "grad_norm": 0.45445945858955383,
+      "learning_rate": 9.818508046386868e-05,
+      "loss": 0.5405,
+      "step": 5270
+    },
+    {
+      "epoch": 1.9932049830124576,
+      "grad_norm": 0.5319023728370667,
+      "learning_rate": 9.817402784506859e-05,
+      "loss": 0.5421,
+      "step": 5280
+    },
+    {
+      "epoch": 1.9969799924499811,
+      "grad_norm": 0.5033361315727234,
+      "learning_rate": 9.816294229977472e-05,
+      "loss": 0.5405,
+      "step": 5290
+    },
+    {
+      "epoch": 2.0007550018875047,
+      "grad_norm": 0.5742616057395935,
+      "learning_rate": 9.815182383556394e-05,
+      "loss": 0.5346,
+      "step": 5300
+    },
+    {
+      "epoch": 2.0045300113250284,
+      "grad_norm": 0.591316819190979,
+      "learning_rate": 9.81406724600356e-05,
+      "loss": 0.5396,
+      "step": 5310
+    },
+    {
+      "epoch": 2.0083050207625517,
+      "grad_norm": 0.4500562846660614,
+      "learning_rate": 9.812948818081161e-05,
+      "loss": 0.5375,
+      "step": 5320
+    },
+    {
+      "epoch": 2.0120800302000754,
+      "grad_norm": 0.5774345993995667,
+      "learning_rate": 9.811827100553623e-05,
+      "loss": 0.5345,
+      "step": 5330
+    },
+    {
+      "epoch": 2.015855039637599,
+      "grad_norm": 0.627068042755127,
+      "learning_rate": 9.810702094187638e-05,
+      "loss": 0.5432,
+      "step": 5340
+    },
+    {
+      "epoch": 2.019630049075123,
+      "grad_norm": 0.5072972178459167,
+      "learning_rate": 9.809573799752135e-05,
+      "loss": 0.5488,
+      "step": 5350
+    },
+    {
+      "epoch": 2.023405058512646,
+      "grad_norm": 0.5516794919967651,
+      "learning_rate": 9.808442218018288e-05,
+      "loss": 0.5383,
+      "step": 5360
+    },
+    {
+      "epoch": 2.02718006795017,
+      "grad_norm": 0.5118292570114136,
+      "learning_rate": 9.807307349759527e-05,
+      "loss": 0.539,
+      "step": 5370
+    },
+    {
+      "epoch": 2.0309550773876937,
+      "grad_norm": 0.5869454145431519,
+      "learning_rate": 9.806169195751525e-05,
+      "loss": 0.5374,
+      "step": 5380
+    },
+    {
+      "epoch": 2.034730086825217,
+      "grad_norm": 0.46649426221847534,
+      "learning_rate": 9.805027756772194e-05,
+      "loss": 0.5426,
+      "step": 5390
+    },
+    {
+      "epoch": 2.0385050962627407,
+      "grad_norm": 0.5441772937774658,
+      "learning_rate": 9.803883033601702e-05,
+      "loss": 0.5428,
+      "step": 5400
+    },
+    {
+      "epoch": 2.0422801057002644,
+      "grad_norm": 0.4856852889060974,
+      "learning_rate": 9.802735027022453e-05,
+      "loss": 0.5407,
+      "step": 5410
+    },
+    {
+      "epoch": 2.0460551151377877,
+      "grad_norm": 0.9828983545303345,
+      "learning_rate": 9.8015837378191e-05,
+      "loss": 0.5436,
+      "step": 5420
+    },
+    {
+      "epoch": 2.0498301245753114,
+      "grad_norm": 0.5788944959640503,
+      "learning_rate": 9.80042916677854e-05,
+      "loss": 0.54,
+      "step": 5430
+    },
+    {
+      "epoch": 2.053605134012835,
+      "grad_norm": 0.5199455618858337,
+      "learning_rate": 9.799271314689908e-05,
+      "loss": 0.5301,
+      "step": 5440
+    },
+    {
+      "epoch": 2.0573801434503585,
+      "grad_norm": 0.5322635173797607,
+      "learning_rate": 9.798110182344588e-05,
+      "loss": 0.5441,
+      "step": 5450
+    },
+    {
+      "epoch": 2.061155152887882,
+      "grad_norm": 0.4913989305496216,
+      "learning_rate": 9.796945770536204e-05,
+      "loss": 0.5466,
+      "step": 5460
+    },
+    {
+      "epoch": 2.064930162325406,
+      "grad_norm": 0.5005329847335815,
+      "learning_rate": 9.795778080060615e-05,
+      "loss": 0.5479,
+      "step": 5470
+    },
+    {
+      "epoch": 2.0687051717629292,
+      "grad_norm": 0.45034486055374146,
+      "learning_rate": 9.79460711171593e-05,
+      "loss": 0.5365,
+      "step": 5480
+    },
+    {
+      "epoch": 2.072480181200453,
+      "grad_norm": 0.530417799949646,
+      "learning_rate": 9.793432866302497e-05,
+      "loss": 0.5343,
+      "step": 5490
+    },
+    {
+      "epoch": 2.0762551906379767,
+      "grad_norm": 0.5080109238624573,
+      "learning_rate": 9.792255344622897e-05,
+      "loss": 0.5389,
+      "step": 5500
+    },
+    {
+      "epoch": 2.0800302000755,
+      "grad_norm": 0.5648706555366516,
+      "learning_rate": 9.791074547481957e-05,
+      "loss": 0.5375,
+      "step": 5510
+    },
+    {
+      "epoch": 2.0838052095130237,
+      "grad_norm": 0.4704912602901459,
+      "learning_rate": 9.789890475686742e-05,
+      "loss": 0.5314,
+      "step": 5520
+    },
+    {
+      "epoch": 2.0875802189505475,
+      "grad_norm": 0.47333037853240967,
+      "learning_rate": 9.788703130046552e-05,
+      "loss": 0.5357,
+      "step": 5530
+    },
+    {
+      "epoch": 2.091355228388071,
+      "grad_norm": 0.5688567161560059,
+      "learning_rate": 9.78751251137293e-05,
+      "loss": 0.5406,
+      "step": 5540
+    },
+    {
+      "epoch": 2.0951302378255945,
+      "grad_norm": 0.5980736613273621,
+      "learning_rate": 9.786318620479646e-05,
+      "loss": 0.541,
+      "step": 5550
+    },
+    {
+      "epoch": 2.0989052472631182,
+      "grad_norm": 0.5263574719429016,
+      "learning_rate": 9.78512145818272e-05,
+      "loss": 0.5337,
+      "step": 5560
+    },
+    {
+      "epoch": 2.102680256700642,
+      "grad_norm": 0.6168349981307983,
+      "learning_rate": 9.783921025300398e-05,
+      "loss": 0.534,
+      "step": 5570
+    },
+    {
+      "epoch": 2.1064552661381652,
+      "grad_norm": 0.5928676128387451,
+      "learning_rate": 9.782717322653164e-05,
+      "loss": 0.5298,
+      "step": 5580
+    },
+    {
+      "epoch": 2.110230275575689,
+      "grad_norm": 0.5435881018638611,
+      "learning_rate": 9.781510351063738e-05,
+      "loss": 0.5395,
+      "step": 5590
+    },
+    {
+      "epoch": 2.1140052850132127,
+      "grad_norm": 0.5674425363540649,
+      "learning_rate": 9.780300111357075e-05,
+      "loss": 0.5384,
+      "step": 5600
+    },
+    {
+      "epoch": 2.117780294450736,
+      "grad_norm": 0.60903400182724,
+      "learning_rate": 9.779086604360361e-05,
+      "loss": 0.5377,
+      "step": 5610
+    },
+    {
+      "epoch": 2.1215553038882597,
+      "grad_norm": 0.6011250615119934,
+      "learning_rate": 9.777869830903019e-05,
+      "loss": 0.5371,
+      "step": 5620
+    },
+    {
+      "epoch": 2.1253303133257835,
+      "grad_norm": 0.5067744851112366,
+      "learning_rate": 9.776649791816698e-05,
+      "loss": 0.529,
+      "step": 5630
+    },
+    {
+      "epoch": 2.1291053227633068,
+      "grad_norm": 0.5736885070800781,
+      "learning_rate": 9.775426487935285e-05,
+      "loss": 0.5298,
+      "step": 5640
+    },
+    {
+      "epoch": 2.1328803322008305,
+      "grad_norm": 0.5602228045463562,
+      "learning_rate": 9.774199920094898e-05,
+      "loss": 0.5312,
+      "step": 5650
+    },
+    {
+      "epoch": 2.1366553416383542,
+      "grad_norm": 0.5565608739852905,
+      "learning_rate": 9.772970089133884e-05,
+      "loss": 0.531,
+      "step": 5660
+    },
+    {
+      "epoch": 2.1404303510758775,
+      "grad_norm": 0.6767264604568481,
+      "learning_rate": 9.77173699589282e-05,
+      "loss": 0.53,
+      "step": 5670
+    },
+    {
+      "epoch": 2.1442053605134013,
+      "grad_norm": 0.5440645813941956,
+      "learning_rate": 9.770500641214513e-05,
+      "loss": 0.5317,
+      "step": 5680
+    },
+    {
+      "epoch": 2.147980369950925,
+      "grad_norm": 0.6627269387245178,
+      "learning_rate": 9.769261025944003e-05,
+      "loss": 0.5445,
+      "step": 5690
+    },
+    {
+      "epoch": 2.1517553793884483,
+      "grad_norm": 1.5312882661819458,
+      "learning_rate": 9.768018150928552e-05,
+      "loss": 0.5425,
+      "step": 5700
+    },
+    {
+      "epoch": 2.155530388825972,
+      "grad_norm": 0.621485710144043,
+      "learning_rate": 9.766772017017654e-05,
+      "loss": 0.5347,
+      "step": 5710
+    },
+    {
+      "epoch": 2.1593053982634958,
+      "grad_norm": 0.5858162641525269,
+      "learning_rate": 9.765522625063032e-05,
+      "loss": 0.5268,
+      "step": 5720
+    },
+    {
+      "epoch": 2.163080407701019,
+      "grad_norm": 0.4868731200695038,
+      "learning_rate": 9.76426997591863e-05,
+      "loss": 0.5361,
+      "step": 5730
+    },
+    {
+      "epoch": 2.166855417138543,
+      "grad_norm": 0.5163987874984741,
+      "learning_rate": 9.763014070440625e-05,
+      "loss": 0.5292,
+      "step": 5740
+    },
+    {
+      "epoch": 2.1706304265760665,
+      "grad_norm": 0.5586041808128357,
+      "learning_rate": 9.761754909487415e-05,
+      "loss": 0.5264,
+      "step": 5750
+    },
+    {
+      "epoch": 2.17440543601359,
+      "grad_norm": 0.45366203784942627,
+      "learning_rate": 9.760492493919626e-05,
+      "loss": 0.5405,
+      "step": 5760
+    },
+    {
+      "epoch": 2.1781804454511136,
+      "grad_norm": 0.5266132354736328,
+      "learning_rate": 9.759226824600105e-05,
+      "loss": 0.5403,
+      "step": 5770
+    },
+    {
+      "epoch": 2.1819554548886373,
+      "grad_norm": 0.516366183757782,
+      "learning_rate": 9.757957902393928e-05,
+      "loss": 0.5345,
+      "step": 5780
+    },
+    {
+      "epoch": 2.185730464326161,
+      "grad_norm": 0.4906608760356903,
+      "learning_rate": 9.756685728168387e-05,
+      "loss": 0.5437,
+      "step": 5790
+    },
+    {
+      "epoch": 2.1895054737636843,
+      "grad_norm": 0.5202317237854004,
+      "learning_rate": 9.755410302793004e-05,
+      "loss": 0.5299,
+      "step": 5800
+    },
+    {
+      "epoch": 2.193280483201208,
+      "grad_norm": 0.5337672829627991,
+      "learning_rate": 9.754131627139522e-05,
+      "loss": 0.5359,
+      "step": 5810
+    },
+    {
+      "epoch": 2.197055492638732,
+      "grad_norm": 0.5173420906066895,
+      "learning_rate": 9.752849702081901e-05,
+      "loss": 0.5345,
+      "step": 5820
+    },
+    {
+      "epoch": 2.200830502076255,
+      "grad_norm": 0.4675118625164032,
+      "learning_rate": 9.751564528496324e-05,
+      "loss": 0.531,
+      "step": 5830
+    },
+    {
+      "epoch": 2.204605511513779,
+      "grad_norm": 0.48218834400177,
+      "learning_rate": 9.750276107261197e-05,
+      "loss": 0.5279,
+      "step": 5840
+    },
+    {
+      "epoch": 2.2083805209513026,
+      "grad_norm": 0.5200138688087463,
+      "learning_rate": 9.748984439257142e-05,
+      "loss": 0.5289,
+      "step": 5850
+    },
+    {
+      "epoch": 2.212155530388826,
+      "grad_norm": 0.6025936603546143,
+      "learning_rate": 9.747689525367005e-05,
+      "loss": 0.526,
+      "step": 5860
+    },
+    {
+      "epoch": 2.2159305398263496,
+      "grad_norm": 0.4721621572971344,
+      "learning_rate": 9.746391366475845e-05,
+      "loss": 0.535,
+      "step": 5870
+    },
+    {
+      "epoch": 2.2197055492638733,
+      "grad_norm": 0.5410889983177185,
+      "learning_rate": 9.745089963470942e-05,
+      "loss": 0.5316,
+      "step": 5880
+    },
+    {
+      "epoch": 2.2234805587013966,
+      "grad_norm": 0.5121446251869202,
+      "learning_rate": 9.743785317241791e-05,
+      "loss": 0.5283,
+      "step": 5890
+    },
+    {
+      "epoch": 2.2272555681389203,
+      "grad_norm": 0.48920613527297974,
+      "learning_rate": 9.742477428680108e-05,
+      "loss": 0.5278,
+      "step": 5900
+    },
+    {
+      "epoch": 2.231030577576444,
+      "grad_norm": 0.47083306312561035,
+      "learning_rate": 9.741166298679821e-05,
+      "loss": 0.5301,
+      "step": 5910
+    },
+    {
+      "epoch": 2.2348055870139674,
+      "grad_norm": 0.45593857765197754,
+      "learning_rate": 9.739851928137076e-05,
+      "loss": 0.5278,
+      "step": 5920
+    },
+    {
+      "epoch": 2.238580596451491,
+      "grad_norm": 0.5642737746238708,
+      "learning_rate": 9.738534317950232e-05,
+      "loss": 0.5226,
+      "step": 5930
+    },
+    {
+      "epoch": 2.242355605889015,
+      "grad_norm": 0.5394318103790283,
+      "learning_rate": 9.737213469019864e-05,
+      "loss": 0.5278,
+      "step": 5940
+    },
+    {
+      "epoch": 2.246130615326538,
+      "grad_norm": 0.5335193872451782,
+      "learning_rate": 9.735889382248757e-05,
+      "loss": 0.5239,
+      "step": 5950
+    },
+    {
+      "epoch": 2.249905624764062,
+      "grad_norm": 0.5988762974739075,
+      "learning_rate": 9.734562058541916e-05,
+      "loss": 0.5418,
+      "step": 5960
+    },
+    {
+      "epoch": 2.2536806342015856,
+      "grad_norm": 0.5146274566650391,
+      "learning_rate": 9.733231498806552e-05,
+      "loss": 0.5384,
+      "step": 5970
+    },
+    {
+      "epoch": 2.2574556436391093,
+      "grad_norm": 0.5223647356033325,
+      "learning_rate": 9.731897703952088e-05,
+      "loss": 0.5321,
+      "step": 5980
+    },
+    {
+      "epoch": 2.2612306530766326,
+      "grad_norm": 0.5377148985862732,
+      "learning_rate": 9.730560674890165e-05,
+      "loss": 0.5249,
+      "step": 5990
+    },
+    {
+      "epoch": 2.2650056625141564,
+      "grad_norm": 0.6117594838142395,
+      "learning_rate": 9.729220412534628e-05,
+      "loss": 0.524,
+      "step": 6000
+    },
+    {
+      "epoch": 2.2687806719516797,
+      "grad_norm": 0.612998366355896,
+      "learning_rate": 9.72787691780153e-05,
+      "loss": 0.5314,
+      "step": 6010
+    },
+    {
+      "epoch": 2.2725556813892034,
+      "grad_norm": 0.5014058947563171,
+      "learning_rate": 9.726530191609142e-05,
+      "loss": 0.5267,
+      "step": 6020
+    },
+    {
+      "epoch": 2.276330690826727,
+      "grad_norm": 0.5796371698379517,
+      "learning_rate": 9.725180234877937e-05,
+      "loss": 0.5323,
+      "step": 6030
+    },
+    {
+      "epoch": 2.280105700264251,
+      "grad_norm": 0.6417374610900879,
+      "learning_rate": 9.723827048530595e-05,
+      "loss": 0.5308,
+      "step": 6040
+    },
+    {
+      "epoch": 2.283880709701774,
+      "grad_norm": 0.4562576711177826,
+      "learning_rate": 9.722470633492011e-05,
+      "loss": 0.5298,
+      "step": 6050
+    },
+    {
+      "epoch": 2.287655719139298,
+      "grad_norm": 0.5561509728431702,
+      "learning_rate": 9.721110990689278e-05,
+      "loss": 0.5265,
+      "step": 6060
+    },
+    {
+      "epoch": 2.2914307285768216,
+      "grad_norm": 0.5673072338104248,
+      "learning_rate": 9.719748121051699e-05,
+      "loss": 0.5195,
+      "step": 6070
+    },
+    {
+      "epoch": 2.295205738014345,
+      "grad_norm": 0.5953671336174011,
+      "learning_rate": 9.718382025510785e-05,
+      "loss": 0.5248,
+      "step": 6080
+    },
+    {
+      "epoch": 2.2989807474518686,
+      "grad_norm": 0.6203503608703613,
+      "learning_rate": 9.717012705000248e-05,
+      "loss": 0.5264,
+      "step": 6090
+    },
+    {
+      "epoch": 2.3027557568893924,
+      "grad_norm": 0.5094435811042786,
+      "learning_rate": 9.715640160456006e-05,
+      "loss": 0.5325,
+      "step": 6100
+    },
+    {
+      "epoch": 2.3065307663269157,
+      "grad_norm": 0.5434731245040894,
+      "learning_rate": 9.714264392816181e-05,
+      "loss": 0.5243,
+      "step": 6110
+    },
+    {
+      "epoch": 2.3103057757644394,
+      "grad_norm": 0.5409674048423767,
+      "learning_rate": 9.712885403021095e-05,
+      "loss": 0.523,
+      "step": 6120
+    },
+    {
+      "epoch": 2.314080785201963,
+      "grad_norm": 0.5891657471656799,
+      "learning_rate": 9.711503192013276e-05,
+      "loss": 0.529,
+      "step": 6130
+    },
+    {
+      "epoch": 2.3178557946394864,
+      "grad_norm": 0.6148802042007446,
+      "learning_rate": 9.710117760737454e-05,
+      "loss": 0.5234,
+      "step": 6140
+    },
+    {
+      "epoch": 2.32163080407701,
+      "grad_norm": 0.5786886215209961,
+      "learning_rate": 9.708729110140554e-05,
+      "loss": 0.524,
+      "step": 6150
+    },
+    {
+      "epoch": 2.325405813514534,
+      "grad_norm": 0.5547001361846924,
+      "learning_rate": 9.707337241171709e-05,
+      "loss": 0.5212,
+      "step": 6160
+    },
+    {
+      "epoch": 2.329180822952057,
+      "grad_norm": 0.6747357845306396,
+      "learning_rate": 9.705942154782245e-05,
+      "loss": 0.5274,
+      "step": 6170
+    },
+    {
+      "epoch": 2.332955832389581,
+      "grad_norm": 0.4640480875968933,
+      "learning_rate": 9.704543851925696e-05,
+      "loss": 0.5254,
+      "step": 6180
+    },
+    {
+      "epoch": 2.3367308418271047,
+      "grad_norm": 0.5004093647003174,
+      "learning_rate": 9.703142333557784e-05,
+      "loss": 0.5208,
+      "step": 6190
+    },
+    {
+      "epoch": 2.340505851264628,
+      "grad_norm": 0.5276467204093933,
+      "learning_rate": 9.701737600636436e-05,
+      "loss": 0.5205,
+      "step": 6200
+    },
+    {
+      "epoch": 2.3442808607021517,
+      "grad_norm": 0.5766600370407104,
+      "learning_rate": 9.700329654121771e-05,
+      "loss": 0.5145,
+      "step": 6210
+    },
+    {
+      "epoch": 2.3480558701396754,
+      "grad_norm": 0.4839555025100708,
+      "learning_rate": 9.698918494976109e-05,
+      "loss": 0.5234,
+      "step": 6220
+    },
+    {
+      "epoch": 2.351830879577199,
+      "grad_norm": 0.5838026404380798,
+      "learning_rate": 9.697504124163965e-05,
+      "loss": 0.5232,
+      "step": 6230
+    },
+    {
+      "epoch": 2.3556058890147225,
+      "grad_norm": 0.553741991519928,
+      "learning_rate": 9.696086542652045e-05,
+      "loss": 0.5223,
+      "step": 6240
+    },
+    {
+      "epoch": 2.359380898452246,
+      "grad_norm": 0.48992642760276794,
+      "learning_rate": 9.694665751409256e-05,
+      "loss": 0.521,
+      "step": 6250
+    },
+    {
+      "epoch": 2.3631559078897695,
+      "grad_norm": 0.511161208152771,
+      "learning_rate": 9.693241751406694e-05,
+      "loss": 0.5186,
+      "step": 6260
+    },
+    {
+      "epoch": 2.366930917327293,
+      "grad_norm": 0.480486124753952,
+      "learning_rate": 9.69181454361765e-05,
+      "loss": 0.517,
+      "step": 6270
+    },
+    {
+      "epoch": 2.370705926764817,
+      "grad_norm": 0.49377861618995667,
+      "learning_rate": 9.690384129017605e-05,
+      "loss": 0.5187,
+      "step": 6280
+    },
+    {
+      "epoch": 2.3744809362023407,
+      "grad_norm": 0.5575188398361206,
+      "learning_rate": 9.688950508584237e-05,
+      "loss": 0.5194,
+      "step": 6290
+    },
+    {
+      "epoch": 2.378255945639864,
+      "grad_norm": 0.501025378704071,
+      "learning_rate": 9.68751368329741e-05,
+      "loss": 0.519,
+      "step": 6300
+    },
+    {
+      "epoch": 2.3820309550773877,
+      "grad_norm": 0.5015434622764587,
+      "learning_rate": 9.68607365413918e-05,
+      "loss": 0.5213,
+      "step": 6310
+    },
+    {
+      "epoch": 2.3858059645149114,
+      "grad_norm": 0.4865148663520813,
+      "learning_rate": 9.684630422093797e-05,
+      "loss": 0.5212,
+      "step": 6320
+    },
+    {
+      "epoch": 2.3895809739524347,
+      "grad_norm": 0.545136034488678,
+      "learning_rate": 9.683183988147693e-05,
+      "loss": 0.5199,
+      "step": 6330
+    },
+    {
+      "epoch": 2.3933559833899585,
+      "grad_norm": 0.4947832226753235,
+      "learning_rate": 9.681734353289491e-05,
+      "loss": 0.5217,
+      "step": 6340
+    },
+    {
+      "epoch": 2.397130992827482,
+      "grad_norm": 0.554629921913147,
+      "learning_rate": 9.680281518510006e-05,
+      "loss": 0.5204,
+      "step": 6350
+    },
+    {
+      "epoch": 2.4009060022650055,
+      "grad_norm": 0.49678510427474976,
+      "learning_rate": 9.678825484802236e-05,
+      "loss": 0.5212,
+      "step": 6360
+    },
+    {
+      "epoch": 2.4046810117025292,
+      "grad_norm": 0.5372890830039978,
+      "learning_rate": 9.677366253161365e-05,
+      "loss": 0.5252,
+      "step": 6370
+    },
+    {
+      "epoch": 2.408456021140053,
+      "grad_norm": 0.4868631660938263,
+      "learning_rate": 9.675903824584765e-05,
+      "loss": 0.5309,
+      "step": 6380
+    },
+    {
+      "epoch": 2.4122310305775763,
+      "grad_norm": 0.5607171654701233,
+      "learning_rate": 9.674438200071991e-05,
+      "loss": 0.5197,
+      "step": 6390
+    },
+    {
+      "epoch": 2.4160060400151,
+      "grad_norm": 0.6623253226280212,
+      "learning_rate": 9.672969380624786e-05,
+      "loss": 0.5247,
+      "step": 6400
+    },
+    {
+      "epoch": 2.4197810494526237,
+      "grad_norm": 0.5090137720108032,
+      "learning_rate": 9.671497367247073e-05,
+      "loss": 0.5169,
+      "step": 6410
+    },
+    {
+      "epoch": 2.4235560588901475,
+      "grad_norm": 0.4954577684402466,
+      "learning_rate": 9.670022160944959e-05,
+      "loss": 0.5109,
+      "step": 6420
+    },
+    {
+      "epoch": 2.4273310683276708,
+      "grad_norm": 0.471021831035614,
+      "learning_rate": 9.668543762726734e-05,
+      "loss": 0.5147,
+      "step": 6430
+    },
+    {
+      "epoch": 2.4311060777651945,
+      "grad_norm": 0.5328172445297241,
+      "learning_rate": 9.667062173602868e-05,
+      "loss": 0.5144,
+      "step": 6440
+    },
+    {
+      "epoch": 2.434881087202718,
+      "grad_norm": 0.596956193447113,
+      "learning_rate": 9.665577394586013e-05,
+      "loss": 0.5231,
+      "step": 6450
+    },
+    {
+      "epoch": 2.4386560966402415,
+      "grad_norm": 0.5546612739562988,
+      "learning_rate": 9.664089426691006e-05,
+      "loss": 0.5233,
+      "step": 6460
+    },
+    {
+      "epoch": 2.4424311060777653,
+      "grad_norm": 0.5410290360450745,
+      "learning_rate": 9.662598270934852e-05,
+      "loss": 0.5234,
+      "step": 6470
+    },
+    {
+      "epoch": 2.446206115515289,
+      "grad_norm": 0.5477890968322754,
+      "learning_rate": 9.661103928336748e-05,
+      "loss": 0.5153,
+      "step": 6480
+    },
+    {
+      "epoch": 2.4499811249528123,
+      "grad_norm": 0.5000752210617065,
+      "learning_rate": 9.659606399918058e-05,
+      "loss": 0.5104,
+      "step": 6490
+    },
+    {
+      "epoch": 2.453756134390336,
+      "grad_norm": 0.5385310053825378,
+      "learning_rate": 9.658105686702334e-05,
+      "loss": 0.5143,
+      "step": 6500
+    },
+    {
+      "epoch": 2.4575311438278598,
+      "grad_norm": 0.48516371846199036,
+      "learning_rate": 9.656601789715294e-05,
+      "loss": 0.5117,
+      "step": 6510
+    },
+    {
+      "epoch": 2.461306153265383,
+      "grad_norm": 0.574826717376709,
+      "learning_rate": 9.655094709984842e-05,
+      "loss": 0.5114,
+      "step": 6520
+    },
+    {
+      "epoch": 2.465081162702907,
+      "grad_norm": 0.5119508504867554,
+      "learning_rate": 9.653584448541048e-05,
+      "loss": 0.519,
+      "step": 6530
+    },
+    {
+      "epoch": 2.4688561721404305,
+      "grad_norm": 0.5045502781867981,
+      "learning_rate": 9.652071006416166e-05,
+      "loss": 0.5134,
+      "step": 6540
+    },
+    {
+      "epoch": 2.472631181577954,
+      "grad_norm": 0.4496314227581024,
+      "learning_rate": 9.650554384644617e-05,
+      "loss": 0.513,
+      "step": 6550
+    },
+    {
+      "epoch": 2.4764061910154775,
+      "grad_norm": 0.45445775985717773,
+      "learning_rate": 9.649034584263e-05,
+      "loss": 0.5129,
+      "step": 6560
+    },
+    {
+      "epoch": 2.4801812004530013,
+      "grad_norm": 0.5979011058807373,
+      "learning_rate": 9.64751160631008e-05,
+      "loss": 0.5236,
+      "step": 6570
+    },
+    {
+      "epoch": 2.4839562098905246,
+      "grad_norm": 0.595025897026062,
+      "learning_rate": 9.645985451826803e-05,
+      "loss": 0.5233,
+      "step": 6580
+    },
+    {
+      "epoch": 2.4877312193280483,
+      "grad_norm": 0.5131351351737976,
+      "learning_rate": 9.644456121856275e-05,
+      "loss": 0.5186,
+      "step": 6590
+    },
+    {
+      "epoch": 2.491506228765572,
+      "grad_norm": 0.5269580483436584,
+      "learning_rate": 9.642923617443784e-05,
+      "loss": 0.527,
+      "step": 6600
+    },
+    {
+      "epoch": 2.4952812382030953,
+      "grad_norm": 0.43365490436553955,
+      "learning_rate": 9.641387939636782e-05,
+      "loss": 0.5164,
+      "step": 6610
+    },
+    {
+      "epoch": 2.499056247640619,
+      "grad_norm": 0.5464170575141907,
+      "learning_rate": 9.639849089484888e-05,
+      "loss": 0.5193,
+      "step": 6620
+    },
+    {
+      "epoch": 2.502831257078143,
+      "grad_norm": 0.5362778306007385,
+      "learning_rate": 9.638307068039893e-05,
+      "loss": 0.5204,
+      "step": 6630
+    },
+    {
+      "epoch": 2.506606266515666,
+      "grad_norm": 0.5616941452026367,
+      "learning_rate": 9.636761876355753e-05,
+      "loss": 0.5144,
+      "step": 6640
+    },
+    {
+      "epoch": 2.51038127595319,
+      "grad_norm": 0.4664197862148285,
+      "learning_rate": 9.635213515488593e-05,
+      "loss": 0.5116,
+      "step": 6650
+    },
+    {
+      "epoch": 2.5141562853907136,
+      "grad_norm": 0.5451635122299194,
+      "learning_rate": 9.633661986496702e-05,
+      "loss": 0.5167,
+      "step": 6660
+    },
+    {
+      "epoch": 2.5179312948282373,
+      "grad_norm": 0.47441262006759644,
+      "learning_rate": 9.632107290440535e-05,
+      "loss": 0.5114,
+      "step": 6670
+    },
+    {
+      "epoch": 2.5217063042657606,
+      "grad_norm": 0.5057079195976257,
+      "learning_rate": 9.630549428382715e-05,
+      "loss": 0.518,
+      "step": 6680
+    },
+    {
+      "epoch": 2.5254813137032843,
+      "grad_norm": 0.5143377184867859,
+      "learning_rate": 9.628988401388024e-05,
+      "loss": 0.5096,
+      "step": 6690
+    },
+    {
+      "epoch": 2.5292563231408076,
+      "grad_norm": 0.485270231962204,
+      "learning_rate": 9.627424210523407e-05,
+      "loss": 0.5115,
+      "step": 6700
+    },
+    {
+      "epoch": 2.5330313325783314,
+      "grad_norm": 0.5043366551399231,
+      "learning_rate": 9.62585685685798e-05,
+      "loss": 0.5142,
+      "step": 6710
+    },
+    {
+      "epoch": 2.536806342015855,
+      "grad_norm": 0.5317198038101196,
+      "learning_rate": 9.624286341463008e-05,
+      "loss": 0.5193,
+      "step": 6720
+    },
+    {
+      "epoch": 2.540581351453379,
+      "grad_norm": 0.4412330985069275,
+      "learning_rate": 9.622712665411927e-05,
+      "loss": 0.5119,
+      "step": 6730
+    },
+    {
+      "epoch": 2.544356360890902,
+      "grad_norm": 0.5754643082618713,
+      "learning_rate": 9.621135829780328e-05,
+      "loss": 0.515,
+      "step": 6740
+    },
+    {
+      "epoch": 2.548131370328426,
+      "grad_norm": 0.5370513200759888,
+      "learning_rate": 9.619555835645964e-05,
+      "loss": 0.5132,
+      "step": 6750
+    },
+    {
+      "epoch": 2.551906379765949,
+      "grad_norm": 0.5530955791473389,
+      "learning_rate": 9.617972684088747e-05,
+      "loss": 0.5167,
+      "step": 6760
+    },
+    {
+      "epoch": 2.555681389203473,
+      "grad_norm": 0.5820709466934204,
+      "learning_rate": 9.616386376190745e-05,
+      "loss": 0.5118,
+      "step": 6770
+    },
+    {
+      "epoch": 2.5594563986409966,
+      "grad_norm": 0.48913341760635376,
+      "learning_rate": 9.614796913036184e-05,
+      "loss": 0.5152,
+      "step": 6780
+    },
+    {
+      "epoch": 2.5632314080785203,
+      "grad_norm": 0.48742911219596863,
+      "learning_rate": 9.61320429571145e-05,
+      "loss": 0.5074,
+      "step": 6790
+    },
+    {
+      "epoch": 2.5670064175160436,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 9.611608525305078e-05,
+      "loss": 0.5215,
+      "step": 6800
+    },
+    {
+      "epoch": 2.5707814269535674,
+      "grad_norm": 0.4997836947441101,
+      "learning_rate": 9.610009602907764e-05,
+      "loss": 0.5082,
+      "step": 6810
+    },
+    {
+      "epoch": 2.574556436391091,
+      "grad_norm": 0.47966811060905457,
+      "learning_rate": 9.608407529612357e-05,
+      "loss": 0.5044,
+      "step": 6820
+    },
+    {
+      "epoch": 2.5783314458286144,
+      "grad_norm": 0.5587261915206909,
+      "learning_rate": 9.606802306513857e-05,
+      "loss": 0.506,
+      "step": 6830
+    },
+    {
+      "epoch": 2.582106455266138,
+      "grad_norm": 0.4997664988040924,
+      "learning_rate": 9.605193934709421e-05,
+      "loss": 0.5145,
+      "step": 6840
+    },
+    {
+      "epoch": 2.585881464703662,
+      "grad_norm": 0.5600175261497498,
+      "learning_rate": 9.603582415298354e-05,
+      "loss": 0.5111,
+      "step": 6850
+    },
+    {
+      "epoch": 2.5896564741411856,
+      "grad_norm": 0.49723485112190247,
+      "learning_rate": 9.601967749382117e-05,
+      "loss": 0.5114,
+      "step": 6860
+    },
+    {
+      "epoch": 2.593431483578709,
+      "grad_norm": 0.4990619421005249,
+      "learning_rate": 9.600349938064316e-05,
+      "loss": 0.5086,
+      "step": 6870
+    },
+    {
+      "epoch": 2.5972064930162326,
+      "grad_norm": 0.4644363522529602,
+      "learning_rate": 9.598728982450711e-05,
+      "loss": 0.5134,
+      "step": 6880
+    },
+    {
+      "epoch": 2.600981502453756,
+      "grad_norm": 0.4740346670150757,
+      "learning_rate": 9.597104883649211e-05,
+      "loss": 0.5176,
+      "step": 6890
+    },
+    {
+      "epoch": 2.6047565118912797,
+      "grad_norm": 0.5966189503669739,
+      "learning_rate": 9.59547764276987e-05,
+      "loss": 0.5044,
+      "step": 6900
+    },
+    {
+      "epoch": 2.6085315213288034,
+      "grad_norm": 0.5578002333641052,
+      "learning_rate": 9.593847260924892e-05,
+      "loss": 0.5072,
+      "step": 6910
+    },
+    {
+      "epoch": 2.612306530766327,
+      "grad_norm": 0.5736666321754456,
+      "learning_rate": 9.59221373922863e-05,
+      "loss": 0.5144,
+      "step": 6920
+    },
+    {
+      "epoch": 2.6160815402038504,
+      "grad_norm": 0.6220727562904358,
+      "learning_rate": 9.59057707879758e-05,
+      "loss": 0.5031,
+      "step": 6930
+    },
+    {
+      "epoch": 2.619856549641374,
+      "grad_norm": 0.565088152885437,
+      "learning_rate": 9.588937280750382e-05,
+      "loss": 0.5083,
+      "step": 6940
+    },
+    {
+      "epoch": 2.6236315590788974,
+      "grad_norm": 0.5720441937446594,
+      "learning_rate": 9.587294346207824e-05,
+      "loss": 0.5303,
+      "step": 6950
+    },
+    {
+      "epoch": 2.627406568516421,
+      "grad_norm": 0.6338963508605957,
+      "learning_rate": 9.585648276292836e-05,
+      "loss": 0.5149,
+      "step": 6960
+    },
+    {
+      "epoch": 2.631181577953945,
+      "grad_norm": 0.504148542881012,
+      "learning_rate": 9.583999072130492e-05,
+      "loss": 0.5045,
+      "step": 6970
+    },
+    {
+      "epoch": 2.6349565873914687,
+      "grad_norm": 0.5418031215667725,
+      "learning_rate": 9.582346734848005e-05,
+      "loss": 0.5128,
+      "step": 6980
+    },
+    {
+      "epoch": 2.638731596828992,
+      "grad_norm": 0.5693923234939575,
+      "learning_rate": 9.580691265574735e-05,
+      "loss": 0.5106,
+      "step": 6990
+    },
+    {
+      "epoch": 2.6425066062665157,
+      "grad_norm": 0.4889356195926666,
+      "learning_rate": 9.579032665442178e-05,
+      "loss": 0.5053,
+      "step": 7000
+    },
+    {
+      "epoch": 2.6462816157040394,
+      "grad_norm": 0.5362417697906494,
+      "learning_rate": 9.577370935583971e-05,
+      "loss": 0.5196,
+      "step": 7010
+    },
+    {
+      "epoch": 2.6500566251415627,
+      "grad_norm": 0.572538435459137,
+      "learning_rate": 9.575706077135893e-05,
+      "loss": 0.5022,
+      "step": 7020
+    },
+    {
+      "epoch": 2.6538316345790864,
+      "grad_norm": 0.5328758955001831,
+      "learning_rate": 9.574038091235857e-05,
+      "loss": 0.5186,
+      "step": 7030
+    },
+    {
+      "epoch": 2.65760664401661,
+      "grad_norm": 0.4795249104499817,
+      "learning_rate": 9.572366979023916e-05,
+      "loss": 0.5051,
+      "step": 7040
+    },
+    {
+      "epoch": 2.661381653454134,
+      "grad_norm": 1.024570345878601,
+      "learning_rate": 9.57069274164226e-05,
+      "loss": 0.5087,
+      "step": 7050
+    },
+    {
+      "epoch": 2.665156662891657,
+      "grad_norm": 0.641263484954834,
+      "learning_rate": 9.569015380235213e-05,
+      "loss": 0.5094,
+      "step": 7060
+    },
+    {
+      "epoch": 2.668931672329181,
+      "grad_norm": 0.5658965706825256,
+      "learning_rate": 9.567334895949238e-05,
+      "loss": 0.5089,
+      "step": 7070
+    },
+    {
+      "epoch": 2.6727066817667042,
+      "grad_norm": 0.6503294706344604,
+      "learning_rate": 9.565651289932928e-05,
+      "loss": 0.5071,
+      "step": 7080
+    },
+    {
+      "epoch": 2.676481691204228,
+      "grad_norm": 0.5400107502937317,
+      "learning_rate": 9.563964563337015e-05,
+      "loss": 0.5072,
+      "step": 7090
+    },
+    {
+      "epoch": 2.6802567006417517,
+      "grad_norm": 0.5809869766235352,
+      "learning_rate": 9.562274717314357e-05,
+      "loss": 0.5055,
+      "step": 7100
+    },
+    {
+      "epoch": 2.6840317100792754,
+      "grad_norm": 0.5175840854644775,
+      "learning_rate": 9.56058175301995e-05,
+      "loss": 0.5084,
+      "step": 7110
+    },
+    {
+      "epoch": 2.6878067195167987,
+      "grad_norm": 0.510823130607605,
+      "learning_rate": 9.558885671610918e-05,
+      "loss": 0.5142,
+      "step": 7120
+    },
+    {
+      "epoch": 2.6915817289543225,
+      "grad_norm": 0.5192266702651978,
+      "learning_rate": 9.557186474246517e-05,
+      "loss": 0.51,
+      "step": 7130
+    },
+    {
+      "epoch": 2.6953567383918458,
+      "grad_norm": 0.4871031939983368,
+      "learning_rate": 9.555484162088133e-05,
+      "loss": 0.5029,
+      "step": 7140
+    },
+    {
+      "epoch": 2.6991317478293695,
+      "grad_norm": 0.6060387492179871,
+      "learning_rate": 9.553778736299279e-05,
+      "loss": 0.5032,
+      "step": 7150
+    },
+    {
+      "epoch": 2.7029067572668932,
+      "grad_norm": 0.5182626247406006,
+      "learning_rate": 9.552070198045599e-05,
+      "loss": 0.5086,
+      "step": 7160
+    },
+    {
+      "epoch": 2.706681766704417,
+      "grad_norm": 0.509029746055603,
+      "learning_rate": 9.550358548494863e-05,
+      "loss": 0.5076,
+      "step": 7170
+    },
+    {
+      "epoch": 2.7104567761419402,
+      "grad_norm": 0.5284494757652283,
+      "learning_rate": 9.548643788816965e-05,
+      "loss": 0.4986,
+      "step": 7180
+    },
+    {
+      "epoch": 2.714231785579464,
+      "grad_norm": 0.5049721598625183,
+      "learning_rate": 9.546925920183929e-05,
+      "loss": 0.5059,
+      "step": 7190
+    },
+    {
+      "epoch": 2.7180067950169873,
+      "grad_norm": 0.5326721668243408,
+      "learning_rate": 9.545204943769902e-05,
+      "loss": 0.5078,
+      "step": 7200
+    },
+    {
+      "epoch": 2.721781804454511,
+      "grad_norm": 0.5484986305236816,
+      "learning_rate": 9.543480860751155e-05,
+      "loss": 0.5036,
+      "step": 7210
+    },
+    {
+      "epoch": 2.7255568138920347,
+      "grad_norm": 0.5126481652259827,
+      "learning_rate": 9.541753672306081e-05,
+      "loss": 0.5037,
+      "step": 7220
+    },
+    {
+      "epoch": 2.7293318233295585,
+      "grad_norm": 0.4992707669734955,
+      "learning_rate": 9.540023379615198e-05,
+      "loss": 0.499,
+      "step": 7230
+    },
+    {
+      "epoch": 2.7331068327670818,
+      "grad_norm": 0.46070438623428345,
+      "learning_rate": 9.538289983861146e-05,
+      "loss": 0.504,
+      "step": 7240
+    },
+    {
+      "epoch": 2.7368818422046055,
+      "grad_norm": 0.6228576898574829,
+      "learning_rate": 9.536553486228683e-05,
+      "loss": 0.4989,
+      "step": 7250
+    },
+    {
+      "epoch": 2.7406568516421292,
+      "grad_norm": 0.5347064137458801,
+      "learning_rate": 9.534813887904689e-05,
+      "loss": 0.5057,
+      "step": 7260
+    },
+    {
+      "epoch": 2.7444318610796525,
+      "grad_norm": 0.5296137928962708,
+      "learning_rate": 9.533071190078163e-05,
+      "loss": 0.5062,
+      "step": 7270
+    },
+    {
+      "epoch": 2.7482068705171763,
+      "grad_norm": 0.521789014339447,
+      "learning_rate": 9.531325393940225e-05,
+      "loss": 0.5067,
+      "step": 7280
+    },
+    {
+      "epoch": 2.7519818799547,
+      "grad_norm": 0.5194920897483826,
+      "learning_rate": 9.529576500684105e-05,
+      "loss": 0.4991,
+      "step": 7290
+    },
+    {
+      "epoch": 2.7557568893922237,
+      "grad_norm": 0.6369245648384094,
+      "learning_rate": 9.52782451150516e-05,
+      "loss": 0.5752,
+      "step": 7300
+    },
+    {
+      "epoch": 2.759531898829747,
+      "grad_norm": 0.6341930031776428,
+      "learning_rate": 9.526069427600857e-05,
+      "loss": 0.5042,
+      "step": 7310
+    },
+    {
+      "epoch": 2.7633069082672708,
+      "grad_norm": 0.5371559262275696,
+      "learning_rate": 9.524311250170776e-05,
+      "loss": 0.5055,
+      "step": 7320
+    },
+    {
+      "epoch": 2.767081917704794,
+      "grad_norm": 0.6049063205718994,
+      "learning_rate": 9.522549980416619e-05,
+      "loss": 0.5062,
+      "step": 7330
+    },
+    {
+      "epoch": 2.770856927142318,
+      "grad_norm": 0.5581164956092834,
+      "learning_rate": 9.520785619542196e-05,
+      "loss": 0.5072,
+      "step": 7340
+    },
+    {
+      "epoch": 2.7746319365798415,
+      "grad_norm": 0.5462380051612854,
+      "learning_rate": 9.519018168753428e-05,
+      "loss": 0.5026,
+      "step": 7350
+    },
+    {
+      "epoch": 2.7784069460173653,
+      "grad_norm": 0.5068135857582092,
+      "learning_rate": 9.517247629258357e-05,
+      "loss": 0.5122,
+      "step": 7360
+    },
+    {
+      "epoch": 2.7821819554548886,
+      "grad_norm": 0.5248288512229919,
+      "learning_rate": 9.515474002267124e-05,
+      "loss": 0.5,
+      "step": 7370
+    },
+    {
+      "epoch": 2.7859569648924123,
+      "grad_norm": 0.5446317791938782,
+      "learning_rate": 9.513697288991989e-05,
+      "loss": 0.5096,
+      "step": 7380
+    },
+    {
+      "epoch": 2.7897319743299356,
+      "grad_norm": 0.5539740324020386,
+      "learning_rate": 9.51191749064732e-05,
+      "loss": 0.5037,
+      "step": 7390
+    },
+    {
+      "epoch": 2.7935069837674593,
+      "grad_norm": 0.48345574736595154,
+      "learning_rate": 9.51013460844959e-05,
+      "loss": 0.4938,
+      "step": 7400
+    },
+    {
+      "epoch": 2.797281993204983,
+      "grad_norm": 0.538747251033783,
+      "learning_rate": 9.508348643617382e-05,
+      "loss": 0.501,
+      "step": 7410
+    },
+    {
+      "epoch": 2.801057002642507,
+      "grad_norm": 0.4825133681297302,
+      "learning_rate": 9.50655959737139e-05,
+      "loss": 0.4987,
+      "step": 7420
+    },
+    {
+      "epoch": 2.80483201208003,
+      "grad_norm": 0.5534849762916565,
+      "learning_rate": 9.504767470934405e-05,
+      "loss": 0.4992,
+      "step": 7430
+    },
+    {
+      "epoch": 2.808607021517554,
+      "grad_norm": 0.5716150999069214,
+      "learning_rate": 9.502972265531332e-05,
+      "loss": 0.5085,
+      "step": 7440
+    },
+    {
+      "epoch": 2.8123820309550775,
+      "grad_norm": 0.47607553005218506,
+      "learning_rate": 9.501173982389175e-05,
+      "loss": 0.501,
+      "step": 7450
+    },
+    {
+      "epoch": 2.816157040392601,
+      "grad_norm": 0.5296433568000793,
+      "learning_rate": 9.499372622737047e-05,
+      "loss": 0.5004,
+      "step": 7460
+    },
+    {
+      "epoch": 2.8199320498301246,
+      "grad_norm": 0.5382222533226013,
+      "learning_rate": 9.497568187806156e-05,
+      "loss": 0.5099,
+      "step": 7470
+    },
+    {
+      "epoch": 2.8237070592676483,
+      "grad_norm": 0.480925977230072,
+      "learning_rate": 9.495760678829819e-05,
+      "loss": 0.5031,
+      "step": 7480
+    },
+    {
+      "epoch": 2.8274820687051716,
+      "grad_norm": 0.5002157092094421,
+      "learning_rate": 9.49395009704345e-05,
+      "loss": 0.5087,
+      "step": 7490
+    },
+    {
+      "epoch": 2.8312570781426953,
+      "grad_norm": 0.517042338848114,
+      "learning_rate": 9.492136443684565e-05,
+      "loss": 0.5033,
+      "step": 7500
+    },
+    {
+      "epoch": 2.835032087580219,
+      "grad_norm": 0.5893968343734741,
+      "learning_rate": 9.49031971999278e-05,
+      "loss": 0.5025,
+      "step": 7510
+    },
+    {
+      "epoch": 2.8388070970177424,
+      "grad_norm": 0.47264131903648376,
+      "learning_rate": 9.488499927209806e-05,
+      "loss": 0.4947,
+      "step": 7520
+    },
+    {
+      "epoch": 2.842582106455266,
+      "grad_norm": 0.5126714706420898,
+      "learning_rate": 9.486677066579456e-05,
+      "loss": 0.4967,
+      "step": 7530
+    },
+    {
+      "epoch": 2.84635711589279,
+      "grad_norm": 0.5384365916252136,
+      "learning_rate": 9.484851139347639e-05,
+      "loss": 0.4961,
+      "step": 7540
+    },
+    {
+      "epoch": 2.8501321253303136,
+      "grad_norm": 0.5498970746994019,
+      "learning_rate": 9.483022146762358e-05,
+      "loss": 0.4979,
+      "step": 7550
+    },
+    {
+      "epoch": 2.853907134767837,
+      "grad_norm": 0.5383314490318298,
+      "learning_rate": 9.48119009007371e-05,
+      "loss": 0.4993,
+      "step": 7560
+    },
+    {
+      "epoch": 2.8576821442053606,
+      "grad_norm": 0.46135392785072327,
+      "learning_rate": 9.479354970533892e-05,
+      "loss": 0.4989,
+      "step": 7570
+    },
+    {
+      "epoch": 2.861457153642884,
+      "grad_norm": 0.5344179272651672,
+      "learning_rate": 9.477516789397189e-05,
+      "loss": 0.4975,
+      "step": 7580
+    },
+    {
+      "epoch": 2.8652321630804076,
+      "grad_norm": 0.47662675380706787,
+      "learning_rate": 9.475675547919983e-05,
+      "loss": 0.5002,
+      "step": 7590
+    },
+    {
+      "epoch": 2.8690071725179314,
+      "grad_norm": 0.5206663012504578,
+      "learning_rate": 9.47383124736074e-05,
+      "loss": 0.4969,
+      "step": 7600
+    },
+    {
+      "epoch": 2.872782181955455,
+      "grad_norm": 0.5600720643997192,
+      "learning_rate": 9.471983888980028e-05,
+      "loss": 0.4965,
+      "step": 7610
+    },
+    {
+      "epoch": 2.8765571913929784,
+      "grad_norm": 0.5266720056533813,
+      "learning_rate": 9.470133474040494e-05,
+      "loss": 0.5027,
+      "step": 7620
+    },
+    {
+      "epoch": 2.880332200830502,
+      "grad_norm": 2.0006062984466553,
+      "learning_rate": 9.468280003806882e-05,
+      "loss": 0.5084,
+      "step": 7630
+    },
+    {
+      "epoch": 2.8841072102680254,
+      "grad_norm": 0.5050598978996277,
+      "learning_rate": 9.46642347954602e-05,
+      "loss": 0.5037,
+      "step": 7640
+    },
+    {
+      "epoch": 2.887882219705549,
+      "grad_norm": 0.47348907589912415,
+      "learning_rate": 9.464563902526826e-05,
+      "loss": 0.494,
+      "step": 7650
+    },
+    {
+      "epoch": 2.891657229143073,
+      "grad_norm": 0.5870567560195923,
+      "learning_rate": 9.462701274020303e-05,
+      "loss": 0.4916,
+      "step": 7660
+    },
+    {
+      "epoch": 2.8954322385805966,
+      "grad_norm": 0.580295979976654,
+      "learning_rate": 9.46083559529954e-05,
+      "loss": 0.4945,
+      "step": 7670
+    },
+    {
+      "epoch": 2.89920724801812,
+      "grad_norm": 0.5126973390579224,
+      "learning_rate": 9.458966867639712e-05,
+      "loss": 0.4977,
+      "step": 7680
+    },
+    {
+      "epoch": 2.9029822574556436,
+      "grad_norm": 0.5632660984992981,
+      "learning_rate": 9.457095092318074e-05,
+      "loss": 0.5005,
+      "step": 7690
+    },
+    {
+      "epoch": 2.9067572668931674,
+      "grad_norm": 0.5285568833351135,
+      "learning_rate": 9.455220270613969e-05,
+      "loss": 0.4945,
+      "step": 7700
+    },
+    {
+      "epoch": 2.9105322763306907,
+      "grad_norm": 0.8250537514686584,
+      "learning_rate": 9.45334240380882e-05,
+      "loss": 0.49,
+      "step": 7710
+    },
+    {
+      "epoch": 2.9143072857682144,
+      "grad_norm": 0.5369483232498169,
+      "learning_rate": 9.451461493186129e-05,
+      "loss": 0.4984,
+      "step": 7720
+    },
+    {
+      "epoch": 2.918082295205738,
+      "grad_norm": 0.49857333302497864,
+      "learning_rate": 9.449577540031482e-05,
+      "loss": 0.4919,
+      "step": 7730
+    },
+    {
+      "epoch": 2.921857304643262,
+      "grad_norm": 0.5103145241737366,
+      "learning_rate": 9.447690545632544e-05,
+      "loss": 0.4908,
+      "step": 7740
+    },
+    {
+      "epoch": 2.925632314080785,
+      "grad_norm": 0.5372836589813232,
+      "learning_rate": 9.445800511279058e-05,
+      "loss": 0.496,
+      "step": 7750
+    },
+    {
+      "epoch": 2.929407323518309,
+      "grad_norm": 0.5468769669532776,
+      "learning_rate": 9.443907438262844e-05,
+      "loss": 0.5104,
+      "step": 7760
+    },
+    {
+      "epoch": 2.933182332955832,
+      "grad_norm": 0.6511932015419006,
+      "learning_rate": 9.442011327877798e-05,
+      "loss": 0.4951,
+      "step": 7770
+    },
+    {
+      "epoch": 2.936957342393356,
+      "grad_norm": 0.5281775593757629,
+      "learning_rate": 9.440112181419895e-05,
+      "loss": 0.498,
+      "step": 7780
+    },
+    {
+      "epoch": 2.9407323518308797,
+      "grad_norm": 0.6145192384719849,
+      "learning_rate": 9.438210000187186e-05,
+      "loss": 0.4998,
+      "step": 7790
+    },
+    {
+      "epoch": 2.9445073612684034,
+      "grad_norm": 0.5092405676841736,
+      "learning_rate": 9.43630478547979e-05,
+      "loss": 0.4879,
+      "step": 7800
+    },
+    {
+      "epoch": 2.9482823707059267,
+      "grad_norm": 0.5056966543197632,
+      "learning_rate": 9.434396538599906e-05,
+      "loss": 0.49,
+      "step": 7810
+    },
+    {
+      "epoch": 2.9520573801434504,
+      "grad_norm": 0.49092942476272583,
+      "learning_rate": 9.4324852608518e-05,
+      "loss": 0.4925,
+      "step": 7820
+    },
+    {
+      "epoch": 2.9558323895809737,
+      "grad_norm": 0.48439252376556396,
+      "learning_rate": 9.430570953541816e-05,
+      "loss": 0.4888,
+      "step": 7830
+    },
+    {
+      "epoch": 2.9596073990184975,
+      "grad_norm": 0.553947389125824,
+      "learning_rate": 9.428653617978362e-05,
+      "loss": 0.4942,
+      "step": 7840
+    },
+    {
+      "epoch": 2.963382408456021,
+      "grad_norm": 0.5610196590423584,
+      "learning_rate": 9.42673325547192e-05,
+      "loss": 0.4906,
+      "step": 7850
+    },
+    {
+      "epoch": 2.967157417893545,
+      "grad_norm": 0.6619260907173157,
+      "learning_rate": 9.42480986733504e-05,
+      "loss": 0.4923,
+      "step": 7860
+    },
+    {
+      "epoch": 2.970932427331068,
+      "grad_norm": 0.5648078918457031,
+      "learning_rate": 9.422883454882338e-05,
+      "loss": 0.4951,
+      "step": 7870
+    },
+    {
+      "epoch": 2.974707436768592,
+      "grad_norm": 0.505842387676239,
+      "learning_rate": 9.4209540194305e-05,
+      "loss": 0.5008,
+      "step": 7880
+    },
+    {
+      "epoch": 2.9784824462061152,
+      "grad_norm": 0.5517747402191162,
+      "learning_rate": 9.419021562298278e-05,
+      "loss": 0.4977,
+      "step": 7890
+    },
+    {
+      "epoch": 2.982257455643639,
+      "grad_norm": 0.5466206073760986,
+      "learning_rate": 9.417086084806486e-05,
+      "loss": 0.5011,
+      "step": 7900
+    },
+    {
+      "epoch": 2.9860324650811627,
+      "grad_norm": 0.5874969363212585,
+      "learning_rate": 9.415147588278005e-05,
+      "loss": 0.4834,
+      "step": 7910
+    },
+    {
+      "epoch": 2.9898074745186864,
+      "grad_norm": 0.6627646684646606,
+      "learning_rate": 9.413206074037781e-05,
+      "loss": 0.4895,
+      "step": 7920
+    },
+    {
+      "epoch": 2.9935824839562097,
+      "grad_norm": 0.5905627608299255,
+      "learning_rate": 9.411261543412819e-05,
+      "loss": 0.4971,
+      "step": 7930
+    },
+    {
+      "epoch": 2.9973574933937335,
+      "grad_norm": 0.5630371570587158,
+      "learning_rate": 9.409313997732188e-05,
+      "loss": 0.4933,
+      "step": 7940
+    },
+    {
+      "epoch": 3.001132502831257,
+      "grad_norm": 0.5096926093101501,
+      "learning_rate": 9.407363438327018e-05,
+      "loss": 0.5077,
+      "step": 7950
+    },
+    {
+      "epoch": 3.0049075122687805,
+      "grad_norm": 0.545479416847229,
+      "learning_rate": 9.405409866530496e-05,
+      "loss": 0.4925,
+      "step": 7960
+    },
+    {
+      "epoch": 3.0086825217063042,
+      "grad_norm": 0.4964505434036255,
+      "learning_rate": 9.40345328367787e-05,
+      "loss": 0.4872,
+      "step": 7970
+    },
+    {
+      "epoch": 3.012457531143828,
+      "grad_norm": 0.5932561159133911,
+      "learning_rate": 9.401493691106446e-05,
+      "loss": 0.4889,
+      "step": 7980
+    },
+    {
+      "epoch": 3.0162325405813513,
+      "grad_norm": 0.4858354926109314,
+      "learning_rate": 9.39953109015559e-05,
+      "loss": 0.4962,
+      "step": 7990
+    },
+    {
+      "epoch": 3.020007550018875,
+      "grad_norm": 0.4681301712989807,
+      "learning_rate": 9.397565482166718e-05,
+      "loss": 0.493,
+      "step": 8000
+    },
+    {
+      "epoch": 3.0237825594563987,
+      "grad_norm": 0.5294594168663025,
+      "learning_rate": 9.395596868483305e-05,
+      "loss": 0.4941,
+      "step": 8010
+    },
+    {
+      "epoch": 3.027557568893922,
+      "grad_norm": 0.5004336833953857,
+      "learning_rate": 9.393625250450881e-05,
+      "loss": 0.4886,
+      "step": 8020
+    },
+    {
+      "epoch": 3.0313325783314458,
+      "grad_norm": 0.5321143269538879,
+      "learning_rate": 9.391650629417028e-05,
+      "loss": 0.497,
+      "step": 8030
+    },
+    {
+      "epoch": 3.0351075877689695,
+      "grad_norm": 0.6064821481704712,
+      "learning_rate": 9.38967300673138e-05,
+      "loss": 0.4979,
+      "step": 8040
+    },
+    {
+      "epoch": 3.0388825972064932,
+      "grad_norm": 0.5963532328605652,
+      "learning_rate": 9.387692383745624e-05,
+      "loss": 0.4883,
+      "step": 8050
+    },
+    {
+      "epoch": 3.0426576066440165,
+      "grad_norm": 0.510930597782135,
+      "learning_rate": 9.385708761813496e-05,
+      "loss": 0.4906,
+      "step": 8060
+    },
+    {
+      "epoch": 3.0464326160815403,
+      "grad_norm": 0.627373456954956,
+      "learning_rate": 9.383722142290787e-05,
+      "loss": 0.4881,
+      "step": 8070
+    },
+    {
+      "epoch": 3.050207625519064,
+      "grad_norm": 0.5580819249153137,
+      "learning_rate": 9.381732526535328e-05,
+      "loss": 0.4941,
+      "step": 8080
+    },
+    {
+      "epoch": 3.0539826349565873,
+      "grad_norm": 0.4533594250679016,
+      "learning_rate": 9.379739915907004e-05,
+      "loss": 0.4924,
+      "step": 8090
+    },
+    {
+      "epoch": 3.057757644394111,
+      "grad_norm": 0.47666409611701965,
+      "learning_rate": 9.377744311767746e-05,
+      "loss": 0.4957,
+      "step": 8100
+    },
+    {
+      "epoch": 3.0615326538316348,
+      "grad_norm": 0.5064053535461426,
+      "learning_rate": 9.375745715481532e-05,
+      "loss": 0.4978,
+      "step": 8110
+    },
+    {
+      "epoch": 3.065307663269158,
+      "grad_norm": 0.5491259694099426,
+      "learning_rate": 9.373744128414382e-05,
+      "loss": 0.4879,
+      "step": 8120
+    },
+    {
+      "epoch": 3.069082672706682,
+      "grad_norm": 0.5816760063171387,
+      "learning_rate": 9.371739551934359e-05,
+      "loss": 0.4947,
+      "step": 8130
+    },
+    {
+      "epoch": 3.0728576821442055,
+      "grad_norm": 0.63339763879776,
+      "learning_rate": 9.36973198741158e-05,
+      "loss": 0.4921,
+      "step": 8140
+    },
+    {
+      "epoch": 3.076632691581729,
+      "grad_norm": 0.48095303773880005,
+      "learning_rate": 9.367721436218189e-05,
+      "loss": 0.5039,
+      "step": 8150
+    },
+    {
+      "epoch": 3.0804077010192525,
+      "grad_norm": 1.313503623008728,
+      "learning_rate": 9.365707899728383e-05,
+      "loss": 0.4874,
+      "step": 8160
+    },
+    {
+      "epoch": 3.0841827104567763,
+      "grad_norm": 0.5337556004524231,
+      "learning_rate": 9.363691379318391e-05,
+      "loss": 0.4949,
+      "step": 8170
+    },
+    {
+      "epoch": 3.0879577198942996,
+      "grad_norm": 0.5201692581176758,
+      "learning_rate": 9.361671876366491e-05,
+      "loss": 0.4903,
+      "step": 8180
+    },
+    {
+      "epoch": 3.0917327293318233,
+      "grad_norm": 0.48047223687171936,
+      "learning_rate": 9.35964939225299e-05,
+      "loss": 0.4895,
+      "step": 8190
+    },
+    {
+      "epoch": 3.095507738769347,
+      "grad_norm": 0.5520181655883789,
+      "learning_rate": 9.35762392836024e-05,
+      "loss": 0.5046,
+      "step": 8200
+    },
+    {
+      "epoch": 3.0992827482068703,
+      "grad_norm": 0.5537580847740173,
+      "learning_rate": 9.355595486072624e-05,
+      "loss": 0.4904,
+      "step": 8210
+    },
+    {
+      "epoch": 3.103057757644394,
+      "grad_norm": 0.5307925343513489,
+      "learning_rate": 9.353564066776563e-05,
+      "loss": 0.4851,
+      "step": 8220
+    },
+    {
+      "epoch": 3.106832767081918,
+      "grad_norm": 0.5735664367675781,
+      "learning_rate": 9.351529671860516e-05,
+      "loss": 0.4852,
+      "step": 8230
+    },
+    {
+      "epoch": 3.110607776519441,
+      "grad_norm": 0.553432047367096,
+      "learning_rate": 9.349492302714969e-05,
+      "loss": 0.5036,
+      "step": 8240
+    },
+    {
+      "epoch": 3.114382785956965,
+      "grad_norm": 0.6025165319442749,
+      "learning_rate": 9.347451960732447e-05,
+      "loss": 0.4889,
+      "step": 8250
+    },
+    {
+      "epoch": 3.1181577953944886,
+      "grad_norm": 10.570920944213867,
+      "learning_rate": 9.345408647307506e-05,
+      "loss": 0.5668,
+      "step": 8260
+    },
+    {
+      "epoch": 3.121932804832012,
+      "grad_norm": 0.6134909391403198,
+      "learning_rate": 9.343362363836726e-05,
+      "loss": 0.488,
+      "step": 8270
+    },
+    {
+      "epoch": 3.1257078142695356,
+      "grad_norm": 0.5113944411277771,
+      "learning_rate": 9.341313111718728e-05,
+      "loss": 0.4943,
+      "step": 8280
+    },
+    {
+      "epoch": 3.1294828237070593,
+      "grad_norm": 0.586234986782074,
+      "learning_rate": 9.339260892354153e-05,
+      "loss": 0.4888,
+      "step": 8290
+    },
+    {
+      "epoch": 3.133257833144583,
+      "grad_norm": 0.5822952389717102,
+      "learning_rate": 9.337205707145676e-05,
+      "loss": 0.5001,
+      "step": 8300
+    },
+    {
+      "epoch": 3.1370328425821064,
+      "grad_norm": 0.6024602651596069,
+      "learning_rate": 9.335147557497994e-05,
+      "loss": 0.4862,
+      "step": 8310
+    },
+    {
+      "epoch": 3.14080785201963,
+      "grad_norm": 0.5428383946418762,
+      "learning_rate": 9.333086444817835e-05,
+      "loss": 0.4889,
+      "step": 8320
+    },
+    {
+      "epoch": 3.144582861457154,
+      "grad_norm": 0.5743405818939209,
+      "learning_rate": 9.331022370513951e-05,
+      "loss": 0.4873,
+      "step": 8330
+    },
+    {
+      "epoch": 3.148357870894677,
+      "grad_norm": 0.497543066740036,
+      "learning_rate": 9.328955335997113e-05,
+      "loss": 0.4962,
+      "step": 8340
+    },
+    {
+      "epoch": 3.152132880332201,
+      "grad_norm": 0.4729820489883423,
+      "learning_rate": 9.326885342680125e-05,
+      "loss": 0.4873,
+      "step": 8350
+    },
+    {
+      "epoch": 3.1559078897697246,
+      "grad_norm": 0.5927625298500061,
+      "learning_rate": 9.324812391977806e-05,
+      "loss": 0.4924,
+      "step": 8360
+    },
+    {
+      "epoch": 3.159682899207248,
+      "grad_norm": 0.5837076902389526,
+      "learning_rate": 9.322736485306995e-05,
+      "loss": 0.4886,
+      "step": 8370
+    },
+    {
+      "epoch": 3.1634579086447716,
+      "grad_norm": 0.5531527996063232,
+      "learning_rate": 9.32065762408656e-05,
+      "loss": 0.4897,
+      "step": 8380
+    },
+    {
+      "epoch": 3.1672329180822953,
+      "grad_norm": 0.5851390361785889,
+      "learning_rate": 9.318575809737377e-05,
+      "loss": 0.4864,
+      "step": 8390
+    },
+    {
+      "epoch": 3.1710079275198186,
+      "grad_norm": 0.5682461261749268,
+      "learning_rate": 9.316491043682351e-05,
+      "loss": 0.4849,
+      "step": 8400
+    },
+    {
+      "epoch": 3.1747829369573424,
+      "grad_norm": 0.623104989528656,
+      "learning_rate": 9.3144033273464e-05,
+      "loss": 0.5003,
+      "step": 8410
+    },
+    {
+      "epoch": 3.178557946394866,
+      "grad_norm": 0.6800477504730225,
+      "learning_rate": 9.312312662156457e-05,
+      "loss": 0.4887,
+      "step": 8420
+    },
+    {
+      "epoch": 3.1823329558323894,
+      "grad_norm": 0.5521610975265503,
+      "learning_rate": 9.310219049541471e-05,
+      "loss": 0.5015,
+      "step": 8430
+    },
+    {
+      "epoch": 3.186107965269913,
+      "grad_norm": 0.619483232498169,
+      "learning_rate": 9.308122490932409e-05,
+      "loss": 0.4849,
+      "step": 8440
+    },
+    {
+      "epoch": 3.189882974707437,
+      "grad_norm": 0.5283095240592957,
+      "learning_rate": 9.306022987762246e-05,
+      "loss": 0.4877,
+      "step": 8450
+    },
+    {
+      "epoch": 3.19365798414496,
+      "grad_norm": 0.569065511226654,
+      "learning_rate": 9.303920541465974e-05,
+      "loss": 0.4844,
+      "step": 8460
+    },
+    {
+      "epoch": 3.197432993582484,
+      "grad_norm": 0.4644882082939148,
+      "learning_rate": 9.301815153480598e-05,
+      "loss": 0.4875,
+      "step": 8470
+    },
+    {
+      "epoch": 3.2012080030200076,
+      "grad_norm": 0.5540872812271118,
+      "learning_rate": 9.299706825245126e-05,
+      "loss": 0.4936,
+      "step": 8480
+    },
+    {
+      "epoch": 3.2049830124575314,
+      "grad_norm": 0.5254048705101013,
+      "learning_rate": 9.297595558200583e-05,
+      "loss": 0.4812,
+      "step": 8490
+    },
+    {
+      "epoch": 3.2087580218950547,
+      "grad_norm": 0.4865260422229767,
+      "learning_rate": 9.295481353790001e-05,
+      "loss": 0.4874,
+      "step": 8500
+    },
+    {
+      "epoch": 3.2125330313325784,
+      "grad_norm": 0.6707580089569092,
+      "learning_rate": 9.293364213458417e-05,
+      "loss": 0.4893,
+      "step": 8510
+    },
+    {
+      "epoch": 3.216308040770102,
+      "grad_norm": 0.5525599718093872,
+      "learning_rate": 9.291244138652877e-05,
+      "loss": 0.4847,
+      "step": 8520
+    },
+    {
+      "epoch": 3.2200830502076254,
+      "grad_norm": 0.5599222183227539,
+      "learning_rate": 9.289121130822431e-05,
+      "loss": 0.4823,
+      "step": 8530
+    },
+    {
+      "epoch": 3.223858059645149,
+      "grad_norm": 0.6073276996612549,
+      "learning_rate": 9.286995191418137e-05,
+      "loss": 0.481,
+      "step": 8540
+    },
+    {
+      "epoch": 3.227633069082673,
+      "grad_norm": 0.5141738057136536,
+      "learning_rate": 9.284866321893056e-05,
+      "loss": 0.4807,
+      "step": 8550
+    },
+    {
+      "epoch": 3.231408078520196,
+      "grad_norm": 0.5618724822998047,
+      "learning_rate": 9.282734523702244e-05,
+      "loss": 0.4911,
+      "step": 8560
+    },
+    {
+      "epoch": 3.23518308795772,
+      "grad_norm": 0.5364367961883545,
+      "learning_rate": 9.280599798302771e-05,
+      "loss": 0.4817,
+      "step": 8570
+    },
+    {
+      "epoch": 3.2389580973952437,
+      "grad_norm": 0.5237820744514465,
+      "learning_rate": 9.278462147153699e-05,
+      "loss": 0.4798,
+      "step": 8580
+    },
+    {
+      "epoch": 3.242733106832767,
+      "grad_norm": 0.5670002698898315,
+      "learning_rate": 9.276321571716094e-05,
+      "loss": 0.4839,
+      "step": 8590
+    },
+    {
+      "epoch": 3.2465081162702907,
+      "grad_norm": 0.5286555290222168,
+      "learning_rate": 9.274178073453017e-05,
+      "loss": 0.4827,
+      "step": 8600
+    },
+    {
+      "epoch": 3.2502831257078144,
+      "grad_norm": 0.7837060689926147,
+      "learning_rate": 9.27203165382953e-05,
+      "loss": 0.4848,
+      "step": 8610
+    },
+    {
+      "epoch": 3.2540581351453377,
+      "grad_norm": 0.5733458995819092,
+      "learning_rate": 9.26988231431269e-05,
+      "loss": 0.4874,
+      "step": 8620
+    },
+    {
+      "epoch": 3.2578331445828614,
+      "grad_norm": 0.5995279550552368,
+      "learning_rate": 9.267730056371551e-05,
+      "loss": 0.4856,
+      "step": 8630
+    },
+    {
+      "epoch": 3.261608154020385,
+      "grad_norm": 0.585941731929779,
+      "learning_rate": 9.265574881477161e-05,
+      "loss": 0.495,
+      "step": 8640
+    },
+    {
+      "epoch": 3.2653831634579085,
+      "grad_norm": 0.592987596988678,
+      "learning_rate": 9.263416791102561e-05,
+      "loss": 0.4823,
+      "step": 8650
+    },
+    {
+      "epoch": 3.269158172895432,
+      "grad_norm": 0.47990015149116516,
+      "learning_rate": 9.261255786722786e-05,
+      "loss": 0.487,
+      "step": 8660
+    },
+    {
+      "epoch": 3.272933182332956,
+      "grad_norm": 0.49177876114845276,
+      "learning_rate": 9.259091869814864e-05,
+      "loss": 0.4745,
+      "step": 8670
+    },
+    {
+      "epoch": 3.2767081917704797,
+      "grad_norm": 0.5345214009284973,
+      "learning_rate": 9.256925041857807e-05,
+      "loss": 0.4883,
+      "step": 8680
+    },
+    {
+      "epoch": 3.280483201208003,
+      "grad_norm": 0.5280712842941284,
+      "learning_rate": 9.254755304332626e-05,
+      "loss": 0.4837,
+      "step": 8690
+    },
+    {
+      "epoch": 3.2842582106455267,
+      "grad_norm": 0.5407823920249939,
+      "learning_rate": 9.252582658722313e-05,
+      "loss": 0.4824,
+      "step": 8700
+    },
+    {
+      "epoch": 3.28803322008305,
+      "grad_norm": 0.510508120059967,
+      "learning_rate": 9.250407106511853e-05,
+      "loss": 0.4763,
+      "step": 8710
+    },
+    {
+      "epoch": 3.2918082295205737,
+      "grad_norm": 0.5562644004821777,
+      "learning_rate": 9.248228649188215e-05,
+      "loss": 0.4756,
+      "step": 8720
+    },
+    {
+      "epoch": 3.2955832389580975,
+      "grad_norm": 0.5866154432296753,
+      "learning_rate": 9.246047288240354e-05,
+      "loss": 0.4792,
+      "step": 8730
+    },
+    {
+      "epoch": 3.299358248395621,
+      "grad_norm": 0.5227904319763184,
+      "learning_rate": 9.243863025159208e-05,
+      "loss": 0.48,
+      "step": 8740
+    },
+    {
+      "epoch": 3.3031332578331445,
+      "grad_norm": 0.6132093667984009,
+      "learning_rate": 9.241675861437704e-05,
+      "loss": 0.4762,
+      "step": 8750
+    },
+    {
+      "epoch": 3.3069082672706682,
+      "grad_norm": 0.5366108417510986,
+      "learning_rate": 9.239485798570742e-05,
+      "loss": 0.4813,
+      "step": 8760
+    },
+    {
+      "epoch": 3.310683276708192,
+      "grad_norm": 0.5920717716217041,
+      "learning_rate": 9.237292838055215e-05,
+      "loss": 0.4834,
+      "step": 8770
+    },
+    {
+      "epoch": 3.3144582861457152,
+      "grad_norm": 0.5619731545448303,
+      "learning_rate": 9.235096981389988e-05,
+      "loss": 0.4846,
+      "step": 8780
+    },
+    {
+      "epoch": 3.318233295583239,
+      "grad_norm": 0.45175281167030334,
+      "learning_rate": 9.232898230075909e-05,
+      "loss": 0.4807,
+      "step": 8790
+    },
+    {
+      "epoch": 3.3220083050207627,
+      "grad_norm": 0.613134503364563,
+      "learning_rate": 9.230696585615801e-05,
+      "loss": 0.4724,
+      "step": 8800
+    },
+    {
+      "epoch": 3.325783314458286,
+      "grad_norm": 0.5235288739204407,
+      "learning_rate": 9.228492049514473e-05,
+      "loss": 0.4767,
+      "step": 8810
+    },
+    {
+      "epoch": 3.3295583238958097,
+      "grad_norm": 0.5533393025398254,
+      "learning_rate": 9.2262846232787e-05,
+      "loss": 0.4836,
+      "step": 8820
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.5668758749961853,
+      "learning_rate": 9.224074308417238e-05,
+      "loss": 0.485,
+      "step": 8830
+    },
+    {
+      "epoch": 3.3371083427708568,
+      "grad_norm": 0.568915069103241,
+      "learning_rate": 9.221861106440818e-05,
+      "loss": 0.4767,
+      "step": 8840
+    },
+    {
+      "epoch": 3.3408833522083805,
+      "grad_norm": 0.6423974633216858,
+      "learning_rate": 9.219645018862142e-05,
+      "loss": 0.4766,
+      "step": 8850
+    },
+    {
+      "epoch": 3.3446583616459042,
+      "grad_norm": 0.8083536028862,
+      "learning_rate": 9.217426047195882e-05,
+      "loss": 0.4734,
+      "step": 8860
+    },
+    {
+      "epoch": 3.3484333710834275,
+      "grad_norm": 0.5543457865715027,
+      "learning_rate": 9.21520419295869e-05,
+      "loss": 0.4806,
+      "step": 8870
+    },
+    {
+      "epoch": 3.3522083805209513,
+      "grad_norm": 0.5718511939048767,
+      "learning_rate": 9.212979457669178e-05,
+      "loss": 0.4761,
+      "step": 8880
+    },
+    {
+      "epoch": 3.355983389958475,
+      "grad_norm": 0.6094539165496826,
+      "learning_rate": 9.210751842847931e-05,
+      "loss": 0.4789,
+      "step": 8890
+    },
+    {
+      "epoch": 3.3597583993959983,
+      "grad_norm": 0.5675216913223267,
+      "learning_rate": 9.208521350017506e-05,
+      "loss": 0.4796,
+      "step": 8900
+    },
+    {
+      "epoch": 3.363533408833522,
+      "grad_norm": 0.5290115475654602,
+      "learning_rate": 9.206287980702422e-05,
+      "loss": 0.477,
+      "step": 8910
+    },
+    {
+      "epoch": 3.3673084182710458,
+      "grad_norm": 0.5056905150413513,
+      "learning_rate": 9.204051736429166e-05,
+      "loss": 0.4738,
+      "step": 8920
+    },
+    {
+      "epoch": 3.3710834277085695,
+      "grad_norm": 0.8471300601959229,
+      "learning_rate": 9.20181261872619e-05,
+      "loss": 0.4825,
+      "step": 8930
+    },
+    {
+      "epoch": 3.374858437146093,
+      "grad_norm": 0.516025185585022,
+      "learning_rate": 9.19957062912391e-05,
+      "loss": 0.4708,
+      "step": 8940
+    },
+    {
+      "epoch": 3.3786334465836165,
+      "grad_norm": 0.5982151627540588,
+      "learning_rate": 9.197325769154704e-05,
+      "loss": 0.4698,
+      "step": 8950
+    },
+    {
+      "epoch": 3.38240845602114,
+      "grad_norm": 0.6391962766647339,
+      "learning_rate": 9.195078040352914e-05,
+      "loss": 0.477,
+      "step": 8960
+    },
+    {
+      "epoch": 3.3861834654586636,
+      "grad_norm": 0.6175360679626465,
+      "learning_rate": 9.192827444254839e-05,
+      "loss": 0.4771,
+      "step": 8970
+    },
+    {
+      "epoch": 3.3899584748961873,
+      "grad_norm": 0.5909090042114258,
+      "learning_rate": 9.190573982398744e-05,
+      "loss": 0.4789,
+      "step": 8980
+    },
+    {
+      "epoch": 3.393733484333711,
+      "grad_norm": 0.6406884789466858,
+      "learning_rate": 9.188317656324845e-05,
+      "loss": 0.4735,
+      "step": 8990
+    },
+    {
+      "epoch": 3.3975084937712343,
+      "grad_norm": 0.509844183921814,
+      "learning_rate": 9.186058467575322e-05,
+      "loss": 0.4786,
+      "step": 9000
+    },
+    {
+      "epoch": 3.401283503208758,
+      "grad_norm": 0.524660587310791,
+      "learning_rate": 9.183796417694309e-05,
+      "loss": 0.4789,
+      "step": 9010
+    },
+    {
+      "epoch": 3.405058512646282,
+      "grad_norm": 0.5184488892555237,
+      "learning_rate": 9.181531508227897e-05,
+      "loss": 0.4794,
+      "step": 9020
+    },
+    {
+      "epoch": 3.408833522083805,
+      "grad_norm": 0.4693617522716522,
+      "learning_rate": 9.179263740724125e-05,
+      "loss": 0.483,
+      "step": 9030
+    },
+    {
+      "epoch": 3.412608531521329,
+      "grad_norm": 0.5032320618629456,
+      "learning_rate": 9.176993116732996e-05,
+      "loss": 0.4748,
+      "step": 9040
+    },
+    {
+      "epoch": 3.4163835409588525,
+      "grad_norm": 0.5809365510940552,
+      "learning_rate": 9.17471963780646e-05,
+      "loss": 0.4671,
+      "step": 9050
+    },
+    {
+      "epoch": 3.420158550396376,
+      "grad_norm": 0.547820508480072,
+      "learning_rate": 9.172443305498414e-05,
+      "loss": 0.4729,
+      "step": 9060
+    },
+    {
+      "epoch": 3.4239335598338996,
+      "grad_norm": 0.48663732409477234,
+      "learning_rate": 9.170164121364714e-05,
+      "loss": 0.475,
+      "step": 9070
+    },
+    {
+      "epoch": 3.4277085692714233,
+      "grad_norm": 0.6125722527503967,
+      "learning_rate": 9.167882086963158e-05,
+      "loss": 0.4802,
+      "step": 9080
+    },
+    {
+      "epoch": 3.4314835787089466,
+      "grad_norm": 0.5475627183914185,
+      "learning_rate": 9.165597203853499e-05,
+      "loss": 0.4822,
+      "step": 9090
+    },
+    {
+      "epoch": 3.4352585881464703,
+      "grad_norm": 1.4172425270080566,
+      "learning_rate": 9.163309473597428e-05,
+      "loss": 0.4796,
+      "step": 9100
+    },
+    {
+      "epoch": 3.439033597583994,
+      "grad_norm": 0.5351138114929199,
+      "learning_rate": 9.161018897758591e-05,
+      "loss": 0.4766,
+      "step": 9110
+    },
+    {
+      "epoch": 3.4428086070215174,
+      "grad_norm": 0.6007091999053955,
+      "learning_rate": 9.158725477902572e-05,
+      "loss": 0.4755,
+      "step": 9120
+    },
+    {
+      "epoch": 3.446583616459041,
+      "grad_norm": 0.5067334175109863,
+      "learning_rate": 9.156429215596905e-05,
+      "loss": 0.4923,
+      "step": 9130
+    },
+    {
+      "epoch": 3.450358625896565,
+      "grad_norm": 0.7098965644836426,
+      "learning_rate": 9.154130112411061e-05,
+      "loss": 0.4741,
+      "step": 9140
+    },
+    {
+      "epoch": 3.454133635334088,
+      "grad_norm": 0.555362343788147,
+      "learning_rate": 9.151828169916455e-05,
+      "loss": 0.4741,
+      "step": 9150
+    },
+    {
+      "epoch": 3.457908644771612,
+      "grad_norm": 0.5457745790481567,
+      "learning_rate": 9.149523389686445e-05,
+      "loss": 0.4723,
+      "step": 9160
+    },
+    {
+      "epoch": 3.4616836542091356,
+      "grad_norm": 0.5627350807189941,
+      "learning_rate": 9.147215773296327e-05,
+      "loss": 0.4732,
+      "step": 9170
+    },
+    {
+      "epoch": 3.4654586636466593,
+      "grad_norm": 0.5835473537445068,
+      "learning_rate": 9.144905322323334e-05,
+      "loss": 0.4693,
+      "step": 9180
+    },
+    {
+      "epoch": 3.4692336730841826,
+      "grad_norm": 0.5265465974807739,
+      "learning_rate": 9.142592038346635e-05,
+      "loss": 0.477,
+      "step": 9190
+    },
+    {
+      "epoch": 3.4730086825217064,
+      "grad_norm": 0.5562682747840881,
+      "learning_rate": 9.140275922947343e-05,
+      "loss": 0.4757,
+      "step": 9200
+    },
+    {
+      "epoch": 3.4767836919592296,
+      "grad_norm": 0.6338258981704712,
+      "learning_rate": 9.137956977708496e-05,
+      "loss": 0.4843,
+      "step": 9210
+    },
+    {
+      "epoch": 3.4805587013967534,
+      "grad_norm": 0.5517134666442871,
+      "learning_rate": 9.135635204215075e-05,
+      "loss": 0.47,
+      "step": 9220
+    },
+    {
+      "epoch": 3.484333710834277,
+      "grad_norm": 0.6941575407981873,
+      "learning_rate": 9.133310604053986e-05,
+      "loss": 0.4773,
+      "step": 9230
+    },
+    {
+      "epoch": 3.488108720271801,
+      "grad_norm": 0.532227098941803,
+      "learning_rate": 9.130983178814077e-05,
+      "loss": 0.4677,
+      "step": 9240
+    },
+    {
+      "epoch": 3.491883729709324,
+      "grad_norm": 0.5691418051719666,
+      "learning_rate": 9.128652930086116e-05,
+      "loss": 0.476,
+      "step": 9250
+    },
+    {
+      "epoch": 3.495658739146848,
+      "grad_norm": 0.5184800028800964,
+      "learning_rate": 9.126319859462808e-05,
+      "loss": 0.4822,
+      "step": 9260
+    },
+    {
+      "epoch": 3.4994337485843716,
+      "grad_norm": 0.55794358253479,
+      "learning_rate": 9.123983968538787e-05,
+      "loss": 0.4786,
+      "step": 9270
+    },
+    {
+      "epoch": 3.503208758021895,
+      "grad_norm": 0.5698232650756836,
+      "learning_rate": 9.121645258910609e-05,
+      "loss": 0.4716,
+      "step": 9280
+    },
+    {
+      "epoch": 3.5069837674594186,
+      "grad_norm": 0.53619784116745,
+      "learning_rate": 9.119303732176762e-05,
+      "loss": 0.4761,
+      "step": 9290
+    },
+    {
+      "epoch": 3.5107587768969424,
+      "grad_norm": 0.5619392395019531,
+      "learning_rate": 9.116959389937656e-05,
+      "loss": 0.4741,
+      "step": 9300
+    },
+    {
+      "epoch": 3.5145337863344657,
+      "grad_norm": 0.5324605107307434,
+      "learning_rate": 9.114612233795628e-05,
+      "loss": 0.4683,
+      "step": 9310
+    },
+    {
+      "epoch": 3.5183087957719894,
+      "grad_norm": 0.5709056258201599,
+      "learning_rate": 9.112262265354937e-05,
+      "loss": 0.4778,
+      "step": 9320
+    },
+    {
+      "epoch": 3.522083805209513,
+      "grad_norm": 0.5106092691421509,
+      "learning_rate": 9.109909486221763e-05,
+      "loss": 0.4673,
+      "step": 9330
+    },
+    {
+      "epoch": 3.5258588146470364,
+      "grad_norm": 0.5639387369155884,
+      "learning_rate": 9.107553898004208e-05,
+      "loss": 0.4741,
+      "step": 9340
+    },
+    {
+      "epoch": 3.52963382408456,
+      "grad_norm": 0.5696660280227661,
+      "learning_rate": 9.105195502312298e-05,
+      "loss": 0.4716,
+      "step": 9350
+    },
+    {
+      "epoch": 3.533408833522084,
+      "grad_norm": 0.5708215236663818,
+      "learning_rate": 9.102834300757972e-05,
+      "loss": 0.4703,
+      "step": 9360
+    },
+    {
+      "epoch": 3.5371838429596076,
+      "grad_norm": 0.5708835124969482,
+      "learning_rate": 9.100470294955087e-05,
+      "loss": 0.4707,
+      "step": 9370
+    },
+    {
+      "epoch": 3.540958852397131,
+      "grad_norm": 0.49455785751342773,
+      "learning_rate": 9.098103486519423e-05,
+      "loss": 0.4693,
+      "step": 9380
+    },
+    {
+      "epoch": 3.5447338618346547,
+      "grad_norm": 0.5507346987724304,
+      "learning_rate": 9.095733877068666e-05,
+      "loss": 0.468,
+      "step": 9390
+    },
+    {
+      "epoch": 3.548508871272178,
+      "grad_norm": 0.5424711108207703,
+      "learning_rate": 9.093361468222426e-05,
+      "loss": 0.4784,
+      "step": 9400
+    },
+    {
+      "epoch": 3.5522838807097017,
+      "grad_norm": 0.5007194876670837,
+      "learning_rate": 9.090986261602221e-05,
+      "loss": 0.4659,
+      "step": 9410
+    },
+    {
+      "epoch": 3.5560588901472254,
+      "grad_norm": 0.6124007701873779,
+      "learning_rate": 9.088608258831482e-05,
+      "loss": 0.4768,
+      "step": 9420
+    },
+    {
+      "epoch": 3.559833899584749,
+      "grad_norm": 0.5798971056938171,
+      "learning_rate": 9.08622746153555e-05,
+      "loss": 0.4814,
+      "step": 9430
+    },
+    {
+      "epoch": 3.5636089090222725,
+      "grad_norm": 0.5349039435386658,
+      "learning_rate": 9.083843871341679e-05,
+      "loss": 0.4676,
+      "step": 9440
+    },
+    {
+      "epoch": 3.567383918459796,
+      "grad_norm": 0.557998776435852,
+      "learning_rate": 9.08145748987903e-05,
+      "loss": 0.4758,
+      "step": 9450
+    },
+    {
+      "epoch": 3.5711589278973195,
+      "grad_norm": 0.5525006055831909,
+      "learning_rate": 9.079068318778674e-05,
+      "loss": 0.4692,
+      "step": 9460
+    },
+    {
+      "epoch": 3.574933937334843,
+      "grad_norm": 0.6234112977981567,
+      "learning_rate": 9.076676359673586e-05,
+      "loss": 0.4658,
+      "step": 9470
+    },
+    {
+      "epoch": 3.578708946772367,
+      "grad_norm": 0.6095515489578247,
+      "learning_rate": 9.074281614198645e-05,
+      "loss": 0.4685,
+      "step": 9480
+    },
+    {
+      "epoch": 3.5824839562098907,
+      "grad_norm": 0.5489786267280579,
+      "learning_rate": 9.071884083990638e-05,
+      "loss": 0.4732,
+      "step": 9490
+    },
+    {
+      "epoch": 3.586258965647414,
+      "grad_norm": 0.5583869814872742,
+      "learning_rate": 9.069483770688259e-05,
+      "loss": 0.4704,
+      "step": 9500
+    },
+    {
+      "epoch": 3.5900339750849377,
+      "grad_norm": 0.5421421527862549,
+      "learning_rate": 9.067080675932093e-05,
+      "loss": 0.4735,
+      "step": 9510
+    },
+    {
+      "epoch": 3.5938089845224614,
+      "grad_norm": 0.6016372442245483,
+      "learning_rate": 9.064674801364634e-05,
+      "loss": 0.4794,
+      "step": 9520
+    },
+    {
+      "epoch": 3.5975839939599847,
+      "grad_norm": 0.5686151385307312,
+      "learning_rate": 9.062266148630279e-05,
+      "loss": 0.4758,
+      "step": 9530
+    },
+    {
+      "epoch": 3.6013590033975085,
+      "grad_norm": 0.5975391864776611,
+      "learning_rate": 9.059854719375313e-05,
+      "loss": 0.4748,
+      "step": 9540
+    },
+    {
+      "epoch": 3.605134012835032,
+      "grad_norm": 0.5515162944793701,
+      "learning_rate": 9.057440515247932e-05,
+      "loss": 0.4738,
+      "step": 9550
+    },
+    {
+      "epoch": 3.608909022272556,
+      "grad_norm": 0.5409767627716064,
+      "learning_rate": 9.055023537898215e-05,
+      "loss": 0.4786,
+      "step": 9560
+    },
+    {
+      "epoch": 3.6126840317100792,
+      "grad_norm": 0.4877788722515106,
+      "learning_rate": 9.052603788978149e-05,
+      "loss": 0.4849,
+      "step": 9570
+    },
+    {
+      "epoch": 3.616459041147603,
+      "grad_norm": 0.6089911460876465,
+      "learning_rate": 9.050181270141606e-05,
+      "loss": 0.4728,
+      "step": 9580
+    },
+    {
+      "epoch": 3.6202340505851263,
+      "grad_norm": 0.64964359998703,
+      "learning_rate": 9.047755983044355e-05,
+      "loss": 0.4684,
+      "step": 9590
+    },
+    {
+      "epoch": 3.62400906002265,
+      "grad_norm": 0.5478320121765137,
+      "learning_rate": 9.045327929344058e-05,
+      "loss": 0.4765,
+      "step": 9600
+    },
+    {
+      "epoch": 3.6277840694601737,
+      "grad_norm": 0.5902840495109558,
+      "learning_rate": 9.042897110700268e-05,
+      "loss": 0.4705,
+      "step": 9610
+    },
+    {
+      "epoch": 3.6315590788976975,
+      "grad_norm": 0.5538387894630432,
+      "learning_rate": 9.040463528774423e-05,
+      "loss": 0.473,
+      "step": 9620
+    },
+    {
+      "epoch": 3.6353340883352208,
+      "grad_norm": 0.5468869209289551,
+      "learning_rate": 9.038027185229856e-05,
+      "loss": 0.4745,
+      "step": 9630
+    },
+    {
+      "epoch": 3.6391090977727445,
+      "grad_norm": 0.6634562015533447,
+      "learning_rate": 9.035588081731784e-05,
+      "loss": 0.4749,
+      "step": 9640
+    },
+    {
+      "epoch": 3.642884107210268,
+      "grad_norm": 0.5892646908760071,
+      "learning_rate": 9.033146219947313e-05,
+      "loss": 0.4674,
+      "step": 9650
+    },
+    {
+      "epoch": 3.6466591166477915,
+      "grad_norm": 0.5538533926010132,
+      "learning_rate": 9.030701601545429e-05,
+      "loss": 0.4665,
+      "step": 9660
+    },
+    {
+      "epoch": 3.6504341260853153,
+      "grad_norm": 0.5980792045593262,
+      "learning_rate": 9.028254228197009e-05,
+      "loss": 0.4635,
+      "step": 9670
+    },
+    {
+      "epoch": 3.654209135522839,
+      "grad_norm": 0.5423790216445923,
+      "learning_rate": 9.025804101574808e-05,
+      "loss": 0.4637,
+      "step": 9680
+    },
+    {
+      "epoch": 3.6579841449603623,
+      "grad_norm": 0.5625879168510437,
+      "learning_rate": 9.023351223353463e-05,
+      "loss": 0.4744,
+      "step": 9690
+    },
+    {
+      "epoch": 3.661759154397886,
+      "grad_norm": 0.5283533930778503,
+      "learning_rate": 9.020895595209498e-05,
+      "loss": 0.4697,
+      "step": 9700
+    },
+    {
+      "epoch": 3.6655341638354098,
+      "grad_norm": 0.569664478302002,
+      "learning_rate": 9.018437218821305e-05,
+      "loss": 0.4591,
+      "step": 9710
+    },
+    {
+      "epoch": 3.669309173272933,
+      "grad_norm": 0.5049920678138733,
+      "learning_rate": 9.015976095869166e-05,
+      "loss": 0.4713,
+      "step": 9720
+    },
+    {
+      "epoch": 3.673084182710457,
+      "grad_norm": 0.5221184492111206,
+      "learning_rate": 9.013512228035235e-05,
+      "loss": 0.4634,
+      "step": 9730
+    },
+    {
+      "epoch": 3.6768591921479805,
+      "grad_norm": 0.6969385743141174,
+      "learning_rate": 9.01104561700354e-05,
+      "loss": 0.4678,
+      "step": 9740
+    },
+    {
+      "epoch": 3.680634201585504,
+      "grad_norm": 0.6279971599578857,
+      "learning_rate": 9.008576264459987e-05,
+      "loss": 0.4684,
+      "step": 9750
+    },
+    {
+      "epoch": 3.6844092110230275,
+      "grad_norm": 0.7082238793373108,
+      "learning_rate": 9.006104172092357e-05,
+      "loss": 0.4734,
+      "step": 9760
+    },
+    {
+      "epoch": 3.6881842204605513,
+      "grad_norm": 0.5783301591873169,
+      "learning_rate": 9.0036293415903e-05,
+      "loss": 0.494,
+      "step": 9770
+    },
+    {
+      "epoch": 3.6919592298980746,
+      "grad_norm": 0.5902776122093201,
+      "learning_rate": 9.00115177464534e-05,
+      "loss": 0.5231,
+      "step": 9780
+    },
+    {
+      "epoch": 3.6957342393355983,
+      "grad_norm": 0.6361207365989685,
+      "learning_rate": 8.998671472950872e-05,
+      "loss": 0.4692,
+      "step": 9790
+    },
+    {
+      "epoch": 3.699509248773122,
+      "grad_norm": 0.5117760300636292,
+      "learning_rate": 8.996188438202156e-05,
+      "loss": 0.4679,
+      "step": 9800
+    },
+    {
+      "epoch": 3.7032842582106458,
+      "grad_norm": 0.6711943745613098,
+      "learning_rate": 8.993702672096324e-05,
+      "loss": 0.4669,
+      "step": 9810
+    },
+    {
+      "epoch": 3.707059267648169,
+      "grad_norm": 0.5595587491989136,
+      "learning_rate": 8.991214176332377e-05,
+      "loss": 0.4706,
+      "step": 9820
+    },
+    {
+      "epoch": 3.710834277085693,
+      "grad_norm": 0.5421229004859924,
+      "learning_rate": 8.988722952611171e-05,
+      "loss": 0.47,
+      "step": 9830
+    },
+    {
+      "epoch": 3.714609286523216,
+      "grad_norm": 0.5640469789505005,
+      "learning_rate": 8.986229002635441e-05,
+      "loss": 0.4672,
+      "step": 9840
+    },
+    {
+      "epoch": 3.71838429596074,
+      "grad_norm": 0.5714800357818604,
+      "learning_rate": 8.983732328109775e-05,
+      "loss": 0.4649,
+      "step": 9850
+    },
+    {
+      "epoch": 3.7221593053982636,
+      "grad_norm": 0.5369188189506531,
+      "learning_rate": 8.981232930740627e-05,
+      "loss": 0.4677,
+      "step": 9860
+    },
+    {
+      "epoch": 3.7259343148357873,
+      "grad_norm": 0.8434441089630127,
+      "learning_rate": 8.978730812236313e-05,
+      "loss": 0.4664,
+      "step": 9870
+    },
+    {
+      "epoch": 3.7297093242733106,
+      "grad_norm": 0.5156511664390564,
+      "learning_rate": 8.976225974307004e-05,
+      "loss": 0.4658,
+      "step": 9880
+    },
+    {
+      "epoch": 3.7334843337108343,
+      "grad_norm": 0.5229775905609131,
+      "learning_rate": 8.973718418664736e-05,
+      "loss": 0.4692,
+      "step": 9890
+    },
+    {
+      "epoch": 3.7372593431483576,
+      "grad_norm": 0.5336432456970215,
+      "learning_rate": 8.971208147023397e-05,
+      "loss": 0.4629,
+      "step": 9900
+    },
+    {
+      "epoch": 3.7410343525858814,
+      "grad_norm": 0.5237134695053101,
+      "learning_rate": 8.968695161098739e-05,
+      "loss": 0.4709,
+      "step": 9910
+    },
+    {
+      "epoch": 3.744809362023405,
+      "grad_norm": 0.5551522970199585,
+      "learning_rate": 8.96617946260836e-05,
+      "loss": 0.4639,
+      "step": 9920
+    },
+    {
+      "epoch": 3.748584371460929,
+      "grad_norm": 0.5532174706459045,
+      "learning_rate": 8.963661053271716e-05,
+      "loss": 0.4654,
+      "step": 9930
+    },
+    {
+      "epoch": 3.752359380898452,
+      "grad_norm": 0.5492597818374634,
+      "learning_rate": 8.961139934810117e-05,
+      "loss": 0.4636,
+      "step": 9940
+    },
+    {
+      "epoch": 3.756134390335976,
+      "grad_norm": 0.7447717785835266,
+      "learning_rate": 8.958616108946725e-05,
+      "loss": 0.4703,
+      "step": 9950
+    },
+    {
+      "epoch": 3.7599093997734996,
+      "grad_norm": 0.5668697357177734,
+      "learning_rate": 8.95608957740655e-05,
+      "loss": 0.4647,
+      "step": 9960
+    },
+    {
+      "epoch": 3.763684409211023,
+      "grad_norm": 0.6193100214004517,
+      "learning_rate": 8.953560341916452e-05,
+      "loss": 0.4757,
+      "step": 9970
+    },
+    {
+      "epoch": 3.7674594186485466,
+      "grad_norm": 0.522419273853302,
+      "learning_rate": 8.95102840420514e-05,
+      "loss": 0.4662,
+      "step": 9980
+    },
+    {
+      "epoch": 3.7712344280860703,
+      "grad_norm": 0.5857412219047546,
+      "learning_rate": 8.94849376600317e-05,
+      "loss": 0.4657,
+      "step": 9990
+    },
+    {
+      "epoch": 3.775009437523594,
+      "grad_norm": 0.557212769985199,
+      "learning_rate": 8.945956429042943e-05,
+      "loss": 0.4557,
+      "step": 10000
+    },
+    {
+      "epoch": 3.7787844469611174,
+      "grad_norm": 0.600222647190094,
+      "learning_rate": 8.943416395058705e-05,
+      "loss": 0.4661,
+      "step": 10010
+    },
+    {
+      "epoch": 3.782559456398641,
+      "grad_norm": 0.6195318102836609,
+      "learning_rate": 8.940873665786544e-05,
+      "loss": 0.467,
+      "step": 10020
+    },
+    {
+      "epoch": 3.7863344658361644,
+      "grad_norm": 0.8778566122055054,
+      "learning_rate": 8.938328242964394e-05,
+      "loss": 0.4602,
+      "step": 10030
+    },
+    {
+      "epoch": 3.790109475273688,
+      "grad_norm": 0.6205072999000549,
+      "learning_rate": 8.935780128332026e-05,
+      "loss": 0.4604,
+      "step": 10040
+    },
+    {
+      "epoch": 3.793884484711212,
+      "grad_norm": 0.5959478616714478,
+      "learning_rate": 8.933229323631052e-05,
+      "loss": 0.4582,
+      "step": 10050
+    },
+    {
+      "epoch": 3.7976594941487356,
+      "grad_norm": 0.5634093880653381,
+      "learning_rate": 8.930675830604925e-05,
+      "loss": 0.4649,
+      "step": 10060
+    },
+    {
+      "epoch": 3.801434503586259,
+      "grad_norm": 0.5344751477241516,
+      "learning_rate": 8.92811965099893e-05,
+      "loss": 0.465,
+      "step": 10070
+    },
+    {
+      "epoch": 3.8052095130237826,
+      "grad_norm": 0.5962951183319092,
+      "learning_rate": 8.925560786560194e-05,
+      "loss": 0.4676,
+      "step": 10080
+    },
+    {
+      "epoch": 3.808984522461306,
+      "grad_norm": 0.5603546500205994,
+      "learning_rate": 8.922999239037677e-05,
+      "loss": 0.4597,
+      "step": 10090
+    },
+    {
+      "epoch": 3.8127595318988297,
+      "grad_norm": 0.5416293144226074,
+      "learning_rate": 8.920435010182171e-05,
+      "loss": 0.4627,
+      "step": 10100
+    },
+    {
+      "epoch": 3.8165345413363534,
+      "grad_norm": 0.5338233709335327,
+      "learning_rate": 8.917868101746302e-05,
+      "loss": 0.4745,
+      "step": 10110
+    },
+    {
+      "epoch": 3.820309550773877,
+      "grad_norm": 0.6542847752571106,
+      "learning_rate": 8.91529851548453e-05,
+      "loss": 0.4676,
+      "step": 10120
+    },
+    {
+      "epoch": 3.8240845602114004,
+      "grad_norm": 0.6008440256118774,
+      "learning_rate": 8.912726253153142e-05,
+      "loss": 0.4685,
+      "step": 10130
+    },
+    {
+      "epoch": 3.827859569648924,
+      "grad_norm": 0.6180557012557983,
+      "learning_rate": 8.910151316510255e-05,
+      "loss": 0.4589,
+      "step": 10140
+    },
+    {
+      "epoch": 3.8316345790864474,
+      "grad_norm": 0.6951040625572205,
+      "learning_rate": 8.907573707315813e-05,
+      "loss": 0.4624,
+      "step": 10150
+    },
+    {
+      "epoch": 3.835409588523971,
+      "grad_norm": 0.5660848021507263,
+      "learning_rate": 8.904993427331588e-05,
+      "loss": 0.4647,
+      "step": 10160
+    },
+    {
+      "epoch": 3.839184597961495,
+      "grad_norm": 0.5535010099411011,
+      "learning_rate": 8.902410478321176e-05,
+      "loss": 0.4602,
+      "step": 10170
+    },
+    {
+      "epoch": 3.8429596073990187,
+      "grad_norm": 0.5885663628578186,
+      "learning_rate": 8.899824862050002e-05,
+      "loss": 0.4686,
+      "step": 10180
+    },
+    {
+      "epoch": 3.846734616836542,
+      "grad_norm": 0.6347377300262451,
+      "learning_rate": 8.897236580285308e-05,
+      "loss": 0.4584,
+      "step": 10190
+    },
+    {
+      "epoch": 3.8505096262740657,
+      "grad_norm": 0.6041572093963623,
+      "learning_rate": 8.894645634796159e-05,
+      "loss": 0.4584,
+      "step": 10200
+    },
+    {
+      "epoch": 3.8542846357115894,
+      "grad_norm": 0.5319682359695435,
+      "learning_rate": 8.892052027353444e-05,
+      "loss": 0.4608,
+      "step": 10210
+    },
+    {
+      "epoch": 3.8580596451491127,
+      "grad_norm": 0.5556275248527527,
+      "learning_rate": 8.889455759729866e-05,
+      "loss": 0.465,
+      "step": 10220
+    },
+    {
+      "epoch": 3.8618346545866364,
+      "grad_norm": 0.6249091029167175,
+      "learning_rate": 8.886856833699955e-05,
+      "loss": 0.4611,
+      "step": 10230
+    },
+    {
+      "epoch": 3.86560966402416,
+      "grad_norm": 0.5192009806632996,
+      "learning_rate": 8.884255251040046e-05,
+      "loss": 0.4656,
+      "step": 10240
+    },
+    {
+      "epoch": 3.869384673461684,
+      "grad_norm": 0.5966506600379944,
+      "learning_rate": 8.8816510135283e-05,
+      "loss": 0.4614,
+      "step": 10250
+    },
+    {
+      "epoch": 3.873159682899207,
+      "grad_norm": 0.5522090196609497,
+      "learning_rate": 8.879044122944688e-05,
+      "loss": 0.4583,
+      "step": 10260
+    },
+    {
+      "epoch": 3.876934692336731,
+      "grad_norm": 0.8987534046173096,
+      "learning_rate": 8.876434581070996e-05,
+      "loss": 0.4639,
+      "step": 10270
+    },
+    {
+      "epoch": 3.8807097017742542,
+      "grad_norm": 0.5576398372650146,
+      "learning_rate": 8.87382238969082e-05,
+      "loss": 0.4597,
+      "step": 10280
+    },
+    {
+      "epoch": 3.884484711211778,
+      "grad_norm": 0.5064070820808411,
+      "learning_rate": 8.871207550589568e-05,
+      "loss": 0.4612,
+      "step": 10290
+    },
+    {
+      "epoch": 3.8882597206493017,
+      "grad_norm": 0.536137580871582,
+      "learning_rate": 8.868590065554458e-05,
+      "loss": 0.4681,
+      "step": 10300
+    },
+    {
+      "epoch": 3.8920347300868254,
+      "grad_norm": 0.5211617350578308,
+      "learning_rate": 8.865969936374519e-05,
+      "loss": 0.458,
+      "step": 10310
+    },
+    {
+      "epoch": 3.8958097395243487,
+      "grad_norm": 0.5458037853240967,
+      "learning_rate": 8.863347164840581e-05,
+      "loss": 0.469,
+      "step": 10320
+    },
+    {
+      "epoch": 3.8995847489618725,
+      "grad_norm": 0.5937955379486084,
+      "learning_rate": 8.860721752745285e-05,
+      "loss": 0.4657,
+      "step": 10330
+    },
+    {
+      "epoch": 3.9033597583993958,
+      "grad_norm": 0.5774049758911133,
+      "learning_rate": 8.858093701883077e-05,
+      "loss": 0.454,
+      "step": 10340
+    },
+    {
+      "epoch": 3.9071347678369195,
+      "grad_norm": 0.5954073667526245,
+      "learning_rate": 8.8554630140502e-05,
+      "loss": 0.4606,
+      "step": 10350
+    },
+    {
+      "epoch": 3.910909777274443,
+      "grad_norm": 0.568393349647522,
+      "learning_rate": 8.85282969104471e-05,
+      "loss": 0.4559,
+      "step": 10360
+    },
+    {
+      "epoch": 3.914684786711967,
+      "grad_norm": 0.7279831767082214,
+      "learning_rate": 8.850193734666456e-05,
+      "loss": 0.462,
+      "step": 10370
+    },
+    {
+      "epoch": 3.9184597961494902,
+      "grad_norm": 0.610234260559082,
+      "learning_rate": 8.84755514671709e-05,
+      "loss": 0.4536,
+      "step": 10380
+    },
+    {
+      "epoch": 3.922234805587014,
+      "grad_norm": 0.6353870630264282,
+      "learning_rate": 8.84491392900006e-05,
+      "loss": 0.456,
+      "step": 10390
+    },
+    {
+      "epoch": 3.9260098150245377,
+      "grad_norm": 0.6264570951461792,
+      "learning_rate": 8.842270083320617e-05,
+      "loss": 0.4607,
+      "step": 10400
+    },
+    {
+      "epoch": 3.929784824462061,
+      "grad_norm": 0.5441379547119141,
+      "learning_rate": 8.839623611485801e-05,
+      "loss": 0.4638,
+      "step": 10410
+    },
+    {
+      "epoch": 3.9335598338995847,
+      "grad_norm": 0.5710632801055908,
+      "learning_rate": 8.836974515304453e-05,
+      "loss": 0.4552,
+      "step": 10420
+    },
+    {
+      "epoch": 3.9373348433371085,
+      "grad_norm": 0.6466676592826843,
+      "learning_rate": 8.834322796587204e-05,
+      "loss": 0.4643,
+      "step": 10430
+    },
+    {
+      "epoch": 3.9411098527746318,
+      "grad_norm": 0.5665135383605957,
+      "learning_rate": 8.831668457146478e-05,
+      "loss": 0.4565,
+      "step": 10440
+    },
+    {
+      "epoch": 3.9448848622121555,
+      "grad_norm": 0.48944804072380066,
+      "learning_rate": 8.829011498796493e-05,
+      "loss": 0.4624,
+      "step": 10450
+    },
+    {
+      "epoch": 3.9486598716496792,
+      "grad_norm": 0.4873522222042084,
+      "learning_rate": 8.826351923353253e-05,
+      "loss": 0.4627,
+      "step": 10460
+    },
+    {
+      "epoch": 3.9524348810872025,
+      "grad_norm": 0.7126550674438477,
+      "learning_rate": 8.823689732634555e-05,
+      "loss": 0.4545,
+      "step": 10470
+    },
+    {
+      "epoch": 3.9562098905247263,
+      "grad_norm": 0.5510223507881165,
+      "learning_rate": 8.82102492845998e-05,
+      "loss": 0.466,
+      "step": 10480
+    },
+    {
+      "epoch": 3.95998489996225,
+      "grad_norm": 0.5616976022720337,
+      "learning_rate": 8.818357512650896e-05,
+      "loss": 0.4593,
+      "step": 10490
+    },
+    {
+      "epoch": 3.9637599093997737,
+      "grad_norm": 0.523526668548584,
+      "learning_rate": 8.815687487030458e-05,
+      "loss": 0.4557,
+      "step": 10500
+    },
+    {
+      "epoch": 3.967534918837297,
+      "grad_norm": 0.5704669952392578,
+      "learning_rate": 8.8130148534236e-05,
+      "loss": 0.4687,
+      "step": 10510
+    },
+    {
+      "epoch": 3.9713099282748208,
+      "grad_norm": 0.6883910894393921,
+      "learning_rate": 8.810339613657047e-05,
+      "loss": 0.4598,
+      "step": 10520
+    },
+    {
+      "epoch": 3.975084937712344,
+      "grad_norm": 0.9575225114822388,
+      "learning_rate": 8.807661769559295e-05,
+      "loss": 0.4557,
+      "step": 10530
+    },
+    {
+      "epoch": 3.978859947149868,
+      "grad_norm": 0.5738497376441956,
+      "learning_rate": 8.804981322960628e-05,
+      "loss": 0.4612,
+      "step": 10540
+    },
+    {
+      "epoch": 3.9826349565873915,
+      "grad_norm": 0.530934751033783,
+      "learning_rate": 8.802298275693106e-05,
+      "loss": 0.4532,
+      "step": 10550
+    },
+    {
+      "epoch": 3.9864099660249153,
+      "grad_norm": 0.5506056547164917,
+      "learning_rate": 8.799612629590568e-05,
+      "loss": 0.4542,
+      "step": 10560
+    },
+    {
+      "epoch": 3.9901849754624386,
+      "grad_norm": 0.6080657243728638,
+      "learning_rate": 8.796924386488624e-05,
+      "loss": 0.4601,
+      "step": 10570
+    },
+    {
+      "epoch": 3.9939599848999623,
+      "grad_norm": 0.5610547661781311,
+      "learning_rate": 8.794233548224666e-05,
+      "loss": 0.4545,
+      "step": 10580
+    },
+    {
+      "epoch": 3.9977349943374856,
+      "grad_norm": 0.5351642966270447,
+      "learning_rate": 8.791540116637853e-05,
+      "loss": 0.4598,
+      "step": 10590
+    },
+    {
+      "epoch": 4.001510003775009,
+      "grad_norm": 0.5863658785820007,
+      "learning_rate": 8.788844093569124e-05,
+      "loss": 0.4532,
+      "step": 10600
+    },
+    {
+      "epoch": 4.005285013212533,
+      "grad_norm": 0.530657172203064,
+      "learning_rate": 8.786145480861184e-05,
+      "loss": 0.4474,
+      "step": 10610
+    },
+    {
+      "epoch": 4.009060022650057,
+      "grad_norm": 0.6831036806106567,
+      "learning_rate": 8.783444280358507e-05,
+      "loss": 0.4602,
+      "step": 10620
+    },
+    {
+      "epoch": 4.0128350320875805,
+      "grad_norm": 0.579365074634552,
+      "learning_rate": 8.780740493907342e-05,
+      "loss": 0.4698,
+      "step": 10630
+    },
+    {
+      "epoch": 4.016610041525103,
+      "grad_norm": 0.5720619559288025,
+      "learning_rate": 8.778034123355698e-05,
+      "loss": 0.4594,
+      "step": 10640
+    },
+    {
+      "epoch": 4.020385050962627,
+      "grad_norm": 0.6572504043579102,
+      "learning_rate": 8.775325170553357e-05,
+      "loss": 0.4755,
+      "step": 10650
+    },
+    {
+      "epoch": 4.024160060400151,
+      "grad_norm": 0.584198534488678,
+      "learning_rate": 8.77261363735186e-05,
+      "loss": 0.4579,
+      "step": 10660
+    },
+    {
+      "epoch": 4.027935069837675,
+      "grad_norm": 0.5127620100975037,
+      "learning_rate": 8.769899525604517e-05,
+      "loss": 0.4559,
+      "step": 10670
+    },
+    {
+      "epoch": 4.031710079275198,
+      "grad_norm": 0.5613308548927307,
+      "learning_rate": 8.767182837166397e-05,
+      "loss": 0.4607,
+      "step": 10680
+    },
+    {
+      "epoch": 4.035485088712722,
+      "grad_norm": 0.49144065380096436,
+      "learning_rate": 8.764463573894328e-05,
+      "loss": 0.4517,
+      "step": 10690
+    },
+    {
+      "epoch": 4.039260098150246,
+      "grad_norm": 0.533999502658844,
+      "learning_rate": 8.761741737646902e-05,
+      "loss": 0.4584,
+      "step": 10700
+    },
+    {
+      "epoch": 4.043035107587769,
+      "grad_norm": 0.521372377872467,
+      "learning_rate": 8.759017330284471e-05,
+      "loss": 0.4653,
+      "step": 10710
+    },
+    {
+      "epoch": 4.046810117025292,
+      "grad_norm": 0.6814863681793213,
+      "learning_rate": 8.756290353669142e-05,
+      "loss": 0.4665,
+      "step": 10720
+    },
+    {
+      "epoch": 4.050585126462816,
+      "grad_norm": 0.5394108891487122,
+      "learning_rate": 8.753560809664774e-05,
+      "loss": 0.4549,
+      "step": 10730
+    },
+    {
+      "epoch": 4.05436013590034,
+      "grad_norm": 0.624742329120636,
+      "learning_rate": 8.750828700136986e-05,
+      "loss": 0.4544,
+      "step": 10740
+    },
+    {
+      "epoch": 4.058135145337864,
+      "grad_norm": 0.6260218620300293,
+      "learning_rate": 8.74809402695315e-05,
+      "loss": 0.4608,
+      "step": 10750
+    },
+    {
+      "epoch": 4.061910154775387,
+      "grad_norm": 0.5796566009521484,
+      "learning_rate": 8.745356791982391e-05,
+      "loss": 0.4623,
+      "step": 10760
+    },
+    {
+      "epoch": 4.06568516421291,
+      "grad_norm": 0.5631840825080872,
+      "learning_rate": 8.742616997095578e-05,
+      "loss": 0.4543,
+      "step": 10770
+    },
+    {
+      "epoch": 4.069460173650434,
+      "grad_norm": 0.6347076892852783,
+      "learning_rate": 8.739874644165341e-05,
+      "loss": 0.4613,
+      "step": 10780
+    },
+    {
+      "epoch": 4.073235183087958,
+      "grad_norm": 0.5708393454551697,
+      "learning_rate": 8.737129735066048e-05,
+      "loss": 0.4609,
+      "step": 10790
+    },
+    {
+      "epoch": 4.077010192525481,
+      "grad_norm": 0.5556269288063049,
+      "learning_rate": 8.734382271673821e-05,
+      "loss": 0.4542,
+      "step": 10800
+    },
+    {
+      "epoch": 4.080785201963005,
+      "grad_norm": 0.587650716304779,
+      "learning_rate": 8.731632255866525e-05,
+      "loss": 0.4546,
+      "step": 10810
+    },
+    {
+      "epoch": 4.084560211400529,
+      "grad_norm": 0.5899844765663147,
+      "learning_rate": 8.728879689523767e-05,
+      "loss": 0.4493,
+      "step": 10820
+    },
+    {
+      "epoch": 4.088335220838052,
+      "grad_norm": 0.5399772524833679,
+      "learning_rate": 8.726124574526905e-05,
+      "loss": 0.4597,
+      "step": 10830
+    },
+    {
+      "epoch": 4.092110230275575,
+      "grad_norm": 0.534979522228241,
+      "learning_rate": 8.72336691275903e-05,
+      "loss": 0.4519,
+      "step": 10840
+    },
+    {
+      "epoch": 4.095885239713099,
+      "grad_norm": 0.5515267848968506,
+      "learning_rate": 8.720606706104979e-05,
+      "loss": 0.453,
+      "step": 10850
+    },
+    {
+      "epoch": 4.099660249150623,
+      "grad_norm": 0.5918541550636292,
+      "learning_rate": 8.71784395645133e-05,
+      "loss": 0.4586,
+      "step": 10860
+    },
+    {
+      "epoch": 4.103435258588147,
+      "grad_norm": 0.5579758286476135,
+      "learning_rate": 8.715078665686392e-05,
+      "loss": 0.4613,
+      "step": 10870
+    },
+    {
+      "epoch": 4.10721026802567,
+      "grad_norm": 0.5700817108154297,
+      "learning_rate": 8.712310835700218e-05,
+      "loss": 0.4591,
+      "step": 10880
+    },
+    {
+      "epoch": 4.110985277463194,
+      "grad_norm": 0.5911275148391724,
+      "learning_rate": 8.709540468384591e-05,
+      "loss": 0.4544,
+      "step": 10890
+    },
+    {
+      "epoch": 4.114760286900717,
+      "grad_norm": 0.6004135012626648,
+      "learning_rate": 8.706767565633033e-05,
+      "loss": 0.4484,
+      "step": 10900
+    },
+    {
+      "epoch": 4.118535296338241,
+      "grad_norm": 0.6015802621841431,
+      "learning_rate": 8.7039921293408e-05,
+      "loss": 0.4672,
+      "step": 10910
+    },
+    {
+      "epoch": 4.122310305775764,
+      "grad_norm": 0.6620537638664246,
+      "learning_rate": 8.70121416140487e-05,
+      "loss": 0.4506,
+      "step": 10920
+    },
+    {
+      "epoch": 4.126085315213288,
+      "grad_norm": 0.5017173886299133,
+      "learning_rate": 8.698433663723962e-05,
+      "loss": 0.4554,
+      "step": 10930
+    },
+    {
+      "epoch": 4.129860324650812,
+      "grad_norm": 0.5419957637786865,
+      "learning_rate": 8.695650638198518e-05,
+      "loss": 0.456,
+      "step": 10940
+    },
+    {
+      "epoch": 4.133635334088336,
+      "grad_norm": 0.6066375374794006,
+      "learning_rate": 8.692865086730713e-05,
+      "loss": 0.4519,
+      "step": 10950
+    },
+    {
+      "epoch": 4.1374103435258585,
+      "grad_norm": 0.5887925624847412,
+      "learning_rate": 8.69007701122444e-05,
+      "loss": 0.4557,
+      "step": 10960
+    },
+    {
+      "epoch": 4.141185352963382,
+      "grad_norm": 0.5485467910766602,
+      "learning_rate": 8.687286413585328e-05,
+      "loss": 0.4606,
+      "step": 10970
+    },
+    {
+      "epoch": 4.144960362400906,
+      "grad_norm": 0.5847924947738647,
+      "learning_rate": 8.684493295720719e-05,
+      "loss": 0.4481,
+      "step": 10980
+    },
+    {
+      "epoch": 4.14873537183843,
+      "grad_norm": 0.6698928475379944,
+      "learning_rate": 8.681697659539685e-05,
+      "loss": 0.4543,
+      "step": 10990
+    },
+    {
+      "epoch": 4.152510381275953,
+      "grad_norm": 0.578162431716919,
+      "learning_rate": 8.678899506953019e-05,
+      "loss": 0.4537,
+      "step": 11000
+    },
+    {
+      "epoch": 4.156285390713477,
+      "grad_norm": 0.5858516097068787,
+      "learning_rate": 8.676098839873227e-05,
+      "loss": 0.4531,
+      "step": 11010
+    },
+    {
+      "epoch": 4.160060400151,
+      "grad_norm": 0.6587435603141785,
+      "learning_rate": 8.673295660214545e-05,
+      "loss": 0.4573,
+      "step": 11020
+    },
+    {
+      "epoch": 4.163835409588524,
+      "grad_norm": 0.5828598141670227,
+      "learning_rate": 8.670489969892914e-05,
+      "loss": 0.4543,
+      "step": 11030
+    },
+    {
+      "epoch": 4.1676104190260475,
+      "grad_norm": 0.6038519740104675,
+      "learning_rate": 8.667681770826e-05,
+      "loss": 0.448,
+      "step": 11040
+    },
+    {
+      "epoch": 4.171385428463571,
+      "grad_norm": 0.5508007407188416,
+      "learning_rate": 8.66487106493318e-05,
+      "loss": 0.4513,
+      "step": 11050
+    },
+    {
+      "epoch": 4.175160437901095,
+      "grad_norm": 0.5896838903427124,
+      "learning_rate": 8.662057854135544e-05,
+      "loss": 0.4543,
+      "step": 11060
+    },
+    {
+      "epoch": 4.178935447338619,
+      "grad_norm": 0.5867491960525513,
+      "learning_rate": 8.659242140355897e-05,
+      "loss": 0.449,
+      "step": 11070
+    },
+    {
+      "epoch": 4.182710456776142,
+      "grad_norm": 0.6184877157211304,
+      "learning_rate": 8.65642392551875e-05,
+      "loss": 0.4619,
+      "step": 11080
+    },
+    {
+      "epoch": 4.186485466213665,
+      "grad_norm": 0.6027196049690247,
+      "learning_rate": 8.65360321155033e-05,
+      "loss": 0.4591,
+      "step": 11090
+    },
+    {
+      "epoch": 4.190260475651189,
+      "grad_norm": 0.588248610496521,
+      "learning_rate": 8.650780000378566e-05,
+      "loss": 0.4567,
+      "step": 11100
+    },
+    {
+      "epoch": 4.194035485088713,
+      "grad_norm": 0.5748762488365173,
+      "learning_rate": 8.647954293933096e-05,
+      "loss": 0.4568,
+      "step": 11110
+    },
+    {
+      "epoch": 4.1978104945262364,
+      "grad_norm": 0.5934643149375916,
+      "learning_rate": 8.645126094145264e-05,
+      "loss": 0.4538,
+      "step": 11120
+    },
+    {
+      "epoch": 4.20158550396376,
+      "grad_norm": 0.6221973896026611,
+      "learning_rate": 8.642295402948117e-05,
+      "loss": 0.4602,
+      "step": 11130
+    },
+    {
+      "epoch": 4.205360513401284,
+      "grad_norm": 0.5034570097923279,
+      "learning_rate": 8.639462222276409e-05,
+      "loss": 0.4543,
+      "step": 11140
+    },
+    {
+      "epoch": 4.209135522838807,
+      "grad_norm": 0.6960034966468811,
+      "learning_rate": 8.636626554066589e-05,
+      "loss": 0.4548,
+      "step": 11150
+    },
+    {
+      "epoch": 4.2129105322763305,
+      "grad_norm": 0.5671536326408386,
+      "learning_rate": 8.633788400256811e-05,
+      "loss": 0.4484,
+      "step": 11160
+    },
+    {
+      "epoch": 4.216685541713854,
+      "grad_norm": 0.48511791229248047,
+      "learning_rate": 8.630947762786927e-05,
+      "loss": 0.4558,
+      "step": 11170
+    },
+    {
+      "epoch": 4.220460551151378,
+      "grad_norm": 0.6003875732421875,
+      "learning_rate": 8.628104643598483e-05,
+      "loss": 0.4512,
+      "step": 11180
+    },
+    {
+      "epoch": 4.224235560588902,
+      "grad_norm": 0.544647753238678,
+      "learning_rate": 8.625259044634726e-05,
+      "loss": 0.4552,
+      "step": 11190
+    },
+    {
+      "epoch": 4.228010570026425,
+      "grad_norm": 0.5275722146034241,
+      "learning_rate": 8.622410967840597e-05,
+      "loss": 0.4498,
+      "step": 11200
+    },
+    {
+      "epoch": 4.231785579463948,
+      "grad_norm": 0.5396122932434082,
+      "learning_rate": 8.619560415162731e-05,
+      "loss": 0.4526,
+      "step": 11210
+    },
+    {
+      "epoch": 4.235560588901472,
+      "grad_norm": 0.5510291457176208,
+      "learning_rate": 8.616707388549447e-05,
+      "loss": 0.4465,
+      "step": 11220
+    },
+    {
+      "epoch": 4.239335598338996,
+      "grad_norm": 0.6948357820510864,
+      "learning_rate": 8.613851889950771e-05,
+      "loss": 0.4498,
+      "step": 11230
+    },
+    {
+      "epoch": 4.2431106077765195,
+      "grad_norm": 0.6219735145568848,
+      "learning_rate": 8.610993921318402e-05,
+      "loss": 0.4567,
+      "step": 11240
+    },
+    {
+      "epoch": 4.246885617214043,
+      "grad_norm": 0.5520322322845459,
+      "learning_rate": 8.608133484605738e-05,
+      "loss": 0.4523,
+      "step": 11250
+    },
+    {
+      "epoch": 4.250660626651567,
+      "grad_norm": 0.5849624872207642,
+      "learning_rate": 8.605270581767859e-05,
+      "loss": 0.4526,
+      "step": 11260
+    },
+    {
+      "epoch": 4.25443563608909,
+      "grad_norm": 0.5052843689918518,
+      "learning_rate": 8.602405214761536e-05,
+      "loss": 0.4525,
+      "step": 11270
+    },
+    {
+      "epoch": 4.2582106455266135,
+      "grad_norm": 0.518661618232727,
+      "learning_rate": 8.599537385545215e-05,
+      "loss": 0.4503,
+      "step": 11280
+    },
+    {
+      "epoch": 4.261985654964137,
+      "grad_norm": 0.5502333045005798,
+      "learning_rate": 8.596667096079032e-05,
+      "loss": 0.4477,
+      "step": 11290
+    },
+    {
+      "epoch": 4.265760664401661,
+      "grad_norm": 0.6355045437812805,
+      "learning_rate": 8.593794348324806e-05,
+      "loss": 0.4492,
+      "step": 11300
+    },
+    {
+      "epoch": 4.269535673839185,
+      "grad_norm": 0.5576472878456116,
+      "learning_rate": 8.590919144246028e-05,
+      "loss": 0.4517,
+      "step": 11310
+    },
+    {
+      "epoch": 4.2733106832767085,
+      "grad_norm": 0.5880599021911621,
+      "learning_rate": 8.588041485807876e-05,
+      "loss": 0.4478,
+      "step": 11320
+    },
+    {
+      "epoch": 4.277085692714232,
+      "grad_norm": 0.5388513207435608,
+      "learning_rate": 8.585161374977202e-05,
+      "loss": 0.4523,
+      "step": 11330
+    },
+    {
+      "epoch": 4.280860702151755,
+      "grad_norm": 0.5540359616279602,
+      "learning_rate": 8.582278813722533e-05,
+      "loss": 0.4484,
+      "step": 11340
+    },
+    {
+      "epoch": 4.284635711589279,
+      "grad_norm": 0.6409743428230286,
+      "learning_rate": 8.579393804014076e-05,
+      "loss": 0.4489,
+      "step": 11350
+    },
+    {
+      "epoch": 4.2884107210268025,
+      "grad_norm": 0.6246625185012817,
+      "learning_rate": 8.576506347823703e-05,
+      "loss": 0.4488,
+      "step": 11360
+    },
+    {
+      "epoch": 4.292185730464326,
+      "grad_norm": 0.7034008502960205,
+      "learning_rate": 8.573616447124968e-05,
+      "loss": 0.4492,
+      "step": 11370
+    },
+    {
+      "epoch": 4.29596073990185,
+      "grad_norm": 0.6897478103637695,
+      "learning_rate": 8.570724103893086e-05,
+      "loss": 0.4477,
+      "step": 11380
+    },
+    {
+      "epoch": 4.299735749339374,
+      "grad_norm": 0.5276645421981812,
+      "learning_rate": 8.567829320104951e-05,
+      "loss": 0.4424,
+      "step": 11390
+    },
+    {
+      "epoch": 4.303510758776897,
+      "grad_norm": 0.5786343812942505,
+      "learning_rate": 8.564932097739118e-05,
+      "loss": 0.4503,
+      "step": 11400
+    },
+    {
+      "epoch": 4.30728576821442,
+      "grad_norm": 0.5901648998260498,
+      "learning_rate": 8.562032438775811e-05,
+      "loss": 0.4426,
+      "step": 11410
+    },
+    {
+      "epoch": 4.311060777651944,
+      "grad_norm": 0.60638427734375,
+      "learning_rate": 8.559130345196921e-05,
+      "loss": 0.4519,
+      "step": 11420
+    },
+    {
+      "epoch": 4.314835787089468,
+      "grad_norm": 0.6317540407180786,
+      "learning_rate": 8.556225818986e-05,
+      "loss": 0.4567,
+      "step": 11430
+    },
+    {
+      "epoch": 4.3186107965269915,
+      "grad_norm": 0.5992687344551086,
+      "learning_rate": 8.553318862128265e-05,
+      "loss": 0.4547,
+      "step": 11440
+    },
+    {
+      "epoch": 4.322385805964515,
+      "grad_norm": 0.6526298522949219,
+      "learning_rate": 8.550409476610593e-05,
+      "loss": 0.4417,
+      "step": 11450
+    },
+    {
+      "epoch": 4.326160815402038,
+      "grad_norm": 0.5577118992805481,
+      "learning_rate": 8.547497664421522e-05,
+      "loss": 0.4517,
+      "step": 11460
+    },
+    {
+      "epoch": 4.329935824839562,
+      "grad_norm": 0.6398622989654541,
+      "learning_rate": 8.544583427551249e-05,
+      "loss": 0.4474,
+      "step": 11470
+    },
+    {
+      "epoch": 4.333710834277086,
+      "grad_norm": 0.5577737092971802,
+      "learning_rate": 8.541666767991628e-05,
+      "loss": 0.4508,
+      "step": 11480
+    },
+    {
+      "epoch": 4.337485843714609,
+      "grad_norm": 0.5385574698448181,
+      "learning_rate": 8.538747687736166e-05,
+      "loss": 0.4417,
+      "step": 11490
+    },
+    {
+      "epoch": 4.341260853152133,
+      "grad_norm": 0.595480740070343,
+      "learning_rate": 8.535826188780026e-05,
+      "loss": 0.4463,
+      "step": 11500
+    },
+    {
+      "epoch": 4.345035862589657,
+      "grad_norm": 0.5585368871688843,
+      "learning_rate": 8.532902273120029e-05,
+      "loss": 0.4421,
+      "step": 11510
+    },
+    {
+      "epoch": 4.34881087202718,
+      "grad_norm": 0.6641288995742798,
+      "learning_rate": 8.52997594275464e-05,
+      "loss": 0.444,
+      "step": 11520
+    },
+    {
+      "epoch": 4.352585881464703,
+      "grad_norm": 0.5479051470756531,
+      "learning_rate": 8.52704719968398e-05,
+      "loss": 0.4611,
+      "step": 11530
+    },
+    {
+      "epoch": 4.356360890902227,
+      "grad_norm": 0.5787330269813538,
+      "learning_rate": 8.524116045909818e-05,
+      "loss": 0.4524,
+      "step": 11540
+    },
+    {
+      "epoch": 4.360135900339751,
+      "grad_norm": 0.5680203437805176,
+      "learning_rate": 8.521182483435569e-05,
+      "loss": 0.4497,
+      "step": 11550
+    },
+    {
+      "epoch": 4.363910909777275,
+      "grad_norm": 0.6969168186187744,
+      "learning_rate": 8.518246514266295e-05,
+      "loss": 0.4511,
+      "step": 11560
+    },
+    {
+      "epoch": 4.367685919214798,
+      "grad_norm": 0.6396414041519165,
+      "learning_rate": 8.515308140408703e-05,
+      "loss": 0.4435,
+      "step": 11570
+    },
+    {
+      "epoch": 4.371460928652322,
+      "grad_norm": 0.6805716753005981,
+      "learning_rate": 8.512367363871145e-05,
+      "loss": 0.4603,
+      "step": 11580
+    },
+    {
+      "epoch": 4.375235938089845,
+      "grad_norm": 0.712642252445221,
+      "learning_rate": 8.509424186663614e-05,
+      "loss": 0.4488,
+      "step": 11590
+    },
+    {
+      "epoch": 4.379010947527369,
+      "grad_norm": 0.6053329110145569,
+      "learning_rate": 8.506478610797743e-05,
+      "loss": 0.4522,
+      "step": 11600
+    },
+    {
+      "epoch": 4.382785956964892,
+      "grad_norm": 0.5076550245285034,
+      "learning_rate": 8.503530638286805e-05,
+      "loss": 0.4405,
+      "step": 11610
+    },
+    {
+      "epoch": 4.386560966402416,
+      "grad_norm": 0.6560049057006836,
+      "learning_rate": 8.500580271145712e-05,
+      "loss": 0.4475,
+      "step": 11620
+    },
+    {
+      "epoch": 4.39033597583994,
+      "grad_norm": 0.5975435972213745,
+      "learning_rate": 8.497627511391014e-05,
+      "loss": 0.4492,
+      "step": 11630
+    },
+    {
+      "epoch": 4.394110985277464,
+      "grad_norm": 0.6190789341926575,
+      "learning_rate": 8.494672361040891e-05,
+      "loss": 0.4587,
+      "step": 11640
+    },
+    {
+      "epoch": 4.397885994714986,
+      "grad_norm": 0.6516814827919006,
+      "learning_rate": 8.491714822115162e-05,
+      "loss": 0.4436,
+      "step": 11650
+    },
+    {
+      "epoch": 4.40166100415251,
+      "grad_norm": 0.5875669121742249,
+      "learning_rate": 8.488754896635277e-05,
+      "loss": 0.445,
+      "step": 11660
+    },
+    {
+      "epoch": 4.405436013590034,
+      "grad_norm": 0.49789679050445557,
+      "learning_rate": 8.485792586624317e-05,
+      "loss": 0.4551,
+      "step": 11670
+    },
+    {
+      "epoch": 4.409211023027558,
+      "grad_norm": 0.6263242959976196,
+      "learning_rate": 8.482827894106993e-05,
+      "loss": 0.4457,
+      "step": 11680
+    },
+    {
+      "epoch": 4.412986032465081,
+      "grad_norm": 0.5868949294090271,
+      "learning_rate": 8.479860821109646e-05,
+      "loss": 0.4527,
+      "step": 11690
+    },
+    {
+      "epoch": 4.416761041902605,
+      "grad_norm": 0.6368350386619568,
+      "learning_rate": 8.476891369660239e-05,
+      "loss": 0.4464,
+      "step": 11700
+    },
+    {
+      "epoch": 4.420536051340128,
+      "grad_norm": 0.5921139717102051,
+      "learning_rate": 8.473919541788366e-05,
+      "loss": 0.4423,
+      "step": 11710
+    },
+    {
+      "epoch": 4.424311060777652,
+      "grad_norm": 0.5540608763694763,
+      "learning_rate": 8.470945339525245e-05,
+      "loss": 0.4407,
+      "step": 11720
+    },
+    {
+      "epoch": 4.428086070215175,
+      "grad_norm": 0.5866498351097107,
+      "learning_rate": 8.467968764903713e-05,
+      "loss": 0.4408,
+      "step": 11730
+    },
+    {
+      "epoch": 4.431861079652699,
+      "grad_norm": 0.6014940738677979,
+      "learning_rate": 8.46498981995823e-05,
+      "loss": 0.4541,
+      "step": 11740
+    },
+    {
+      "epoch": 4.435636089090223,
+      "grad_norm": 0.5633965134620667,
+      "learning_rate": 8.462008506724879e-05,
+      "loss": 0.4428,
+      "step": 11750
+    },
+    {
+      "epoch": 4.439411098527747,
+      "grad_norm": 0.5375277400016785,
+      "learning_rate": 8.459024827241359e-05,
+      "loss": 0.4449,
+      "step": 11760
+    },
+    {
+      "epoch": 4.4431861079652695,
+      "grad_norm": 0.566908597946167,
+      "learning_rate": 8.456038783546985e-05,
+      "loss": 0.4425,
+      "step": 11770
+    },
+    {
+      "epoch": 4.446961117402793,
+      "grad_norm": 0.5475563406944275,
+      "learning_rate": 8.453050377682691e-05,
+      "loss": 0.4469,
+      "step": 11780
+    },
+    {
+      "epoch": 4.450736126840317,
+      "grad_norm": 0.5368761420249939,
+      "learning_rate": 8.450059611691026e-05,
+      "loss": 0.441,
+      "step": 11790
+    },
+    {
+      "epoch": 4.454511136277841,
+      "grad_norm": 0.561191976070404,
+      "learning_rate": 8.447066487616146e-05,
+      "loss": 0.4444,
+      "step": 11800
+    },
+    {
+      "epoch": 4.458286145715364,
+      "grad_norm": 0.6211469173431396,
+      "learning_rate": 8.444071007503826e-05,
+      "loss": 0.443,
+      "step": 11810
+    },
+    {
+      "epoch": 4.462061155152888,
+      "grad_norm": 0.6266103982925415,
+      "learning_rate": 8.441073173401449e-05,
+      "loss": 0.4392,
+      "step": 11820
+    },
+    {
+      "epoch": 4.465836164590412,
+      "grad_norm": 0.5911449193954468,
+      "learning_rate": 8.438072987358006e-05,
+      "loss": 0.4502,
+      "step": 11830
+    },
+    {
+      "epoch": 4.469611174027935,
+      "grad_norm": 0.7175887227058411,
+      "learning_rate": 8.435070451424094e-05,
+      "loss": 0.4447,
+      "step": 11840
+    },
+    {
+      "epoch": 4.4733861834654585,
+      "grad_norm": 0.5663455724716187,
+      "learning_rate": 8.432065567651919e-05,
+      "loss": 0.4387,
+      "step": 11850
+    },
+    {
+      "epoch": 4.477161192902982,
+      "grad_norm": 0.8334063291549683,
+      "learning_rate": 8.429058338095291e-05,
+      "loss": 0.4453,
+      "step": 11860
+    },
+    {
+      "epoch": 4.480936202340506,
+      "grad_norm": 0.5345563888549805,
+      "learning_rate": 8.426048764809624e-05,
+      "loss": 0.446,
+      "step": 11870
+    },
+    {
+      "epoch": 4.48471121177803,
+      "grad_norm": 0.5361315608024597,
+      "learning_rate": 8.423036849851932e-05,
+      "loss": 0.4468,
+      "step": 11880
+    },
+    {
+      "epoch": 4.488486221215553,
+      "grad_norm": 0.624729573726654,
+      "learning_rate": 8.42002259528083e-05,
+      "loss": 0.4496,
+      "step": 11890
+    },
+    {
+      "epoch": 4.492261230653076,
+      "grad_norm": 0.5892227292060852,
+      "learning_rate": 8.417006003156532e-05,
+      "loss": 0.4537,
+      "step": 11900
+    },
+    {
+      "epoch": 4.4960362400906,
+      "grad_norm": 0.5822799205780029,
+      "learning_rate": 8.413987075540852e-05,
+      "loss": 0.4537,
+      "step": 11910
+    },
+    {
+      "epoch": 4.499811249528124,
+      "grad_norm": 0.656360924243927,
+      "learning_rate": 8.4109658144972e-05,
+      "loss": 0.4475,
+      "step": 11920
+    },
+    {
+      "epoch": 4.5035862589656475,
+      "grad_norm": 0.6090147495269775,
+      "learning_rate": 8.407942222090573e-05,
+      "loss": 0.4528,
+      "step": 11930
+    },
+    {
+      "epoch": 4.507361268403171,
+      "grad_norm": 0.6417983174324036,
+      "learning_rate": 8.404916300387576e-05,
+      "loss": 0.4479,
+      "step": 11940
+    },
+    {
+      "epoch": 4.511136277840695,
+      "grad_norm": 0.6105488538742065,
+      "learning_rate": 8.401888051456391e-05,
+      "loss": 0.4563,
+      "step": 11950
+    },
+    {
+      "epoch": 4.514911287278219,
+      "grad_norm": 0.6055814027786255,
+      "learning_rate": 8.398857477366803e-05,
+      "loss": 0.4442,
+      "step": 11960
+    },
+    {
+      "epoch": 4.5186862967157415,
+      "grad_norm": 0.6824854612350464,
+      "learning_rate": 8.395824580190178e-05,
+      "loss": 0.4473,
+      "step": 11970
+    },
+    {
+      "epoch": 4.522461306153265,
+      "grad_norm": 0.5766848921775818,
+      "learning_rate": 8.392789361999473e-05,
+      "loss": 0.4444,
+      "step": 11980
+    },
+    {
+      "epoch": 4.526236315590789,
+      "grad_norm": 0.5585546493530273,
+      "learning_rate": 8.38975182486923e-05,
+      "loss": 0.4419,
+      "step": 11990
+    },
+    {
+      "epoch": 4.530011325028313,
+      "grad_norm": 0.5684893727302551,
+      "learning_rate": 8.386711970875581e-05,
+      "loss": 0.4381,
+      "step": 12000
+    },
+    {
+      "epoch": 4.5337863344658365,
+      "grad_norm": 0.5678910613059998,
+      "learning_rate": 8.383669802096232e-05,
+      "loss": 0.4394,
+      "step": 12010
+    },
+    {
+      "epoch": 4.537561343903359,
+      "grad_norm": 0.5642674565315247,
+      "learning_rate": 8.38062532061048e-05,
+      "loss": 0.4401,
+      "step": 12020
+    },
+    {
+      "epoch": 4.541336353340883,
+      "grad_norm": 0.5675539374351501,
+      "learning_rate": 8.3775785284992e-05,
+      "loss": 0.4403,
+      "step": 12030
+    },
+    {
+      "epoch": 4.545111362778407,
+      "grad_norm": 0.6496412754058838,
+      "learning_rate": 8.374529427844843e-05,
+      "loss": 0.4428,
+      "step": 12040
+    },
+    {
+      "epoch": 4.5488863722159305,
+      "grad_norm": 0.6716994643211365,
+      "learning_rate": 8.371478020731442e-05,
+      "loss": 0.4419,
+      "step": 12050
+    },
+    {
+      "epoch": 4.552661381653454,
+      "grad_norm": 0.5777934193611145,
+      "learning_rate": 8.368424309244607e-05,
+      "loss": 0.4414,
+      "step": 12060
+    },
+    {
+      "epoch": 4.556436391090978,
+      "grad_norm": 0.6052347421646118,
+      "learning_rate": 8.365368295471517e-05,
+      "loss": 0.4444,
+      "step": 12070
+    },
+    {
+      "epoch": 4.560211400528502,
+      "grad_norm": 0.6728206276893616,
+      "learning_rate": 8.362309981500931e-05,
+      "loss": 0.4366,
+      "step": 12080
+    },
+    {
+      "epoch": 4.563986409966025,
+      "grad_norm": 0.576066792011261,
+      "learning_rate": 8.359249369423177e-05,
+      "loss": 0.4433,
+      "step": 12090
+    },
+    {
+      "epoch": 4.567761419403548,
+      "grad_norm": 0.566498875617981,
+      "learning_rate": 8.356186461330155e-05,
+      "loss": 0.444,
+      "step": 12100
+    },
+    {
+      "epoch": 4.571536428841072,
+      "grad_norm": 0.5168943405151367,
+      "learning_rate": 8.353121259315334e-05,
+      "loss": 0.4345,
+      "step": 12110
+    },
+    {
+      "epoch": 4.575311438278596,
+      "grad_norm": 0.6019729375839233,
+      "learning_rate": 8.350053765473751e-05,
+      "loss": 0.4342,
+      "step": 12120
+    },
+    {
+      "epoch": 4.5790864477161195,
+      "grad_norm": 0.6395158767700195,
+      "learning_rate": 8.346983981902005e-05,
+      "loss": 0.4388,
+      "step": 12130
+    },
+    {
+      "epoch": 4.582861457153643,
+      "grad_norm": 0.6709539890289307,
+      "learning_rate": 8.343911910698271e-05,
+      "loss": 0.4448,
+      "step": 12140
+    },
+    {
+      "epoch": 4.586636466591166,
+      "grad_norm": 0.6900480389595032,
+      "learning_rate": 8.340837553962278e-05,
+      "loss": 0.4453,
+      "step": 12150
+    },
+    {
+      "epoch": 4.59041147602869,
+      "grad_norm": 0.6444860696792603,
+      "learning_rate": 8.337760913795316e-05,
+      "loss": 0.4453,
+      "step": 12160
+    },
+    {
+      "epoch": 4.5941864854662136,
+      "grad_norm": 0.5633404850959778,
+      "learning_rate": 8.334681992300244e-05,
+      "loss": 0.4447,
+      "step": 12170
+    },
+    {
+      "epoch": 4.597961494903737,
+      "grad_norm": 0.5962954163551331,
+      "learning_rate": 8.331600791581475e-05,
+      "loss": 0.4448,
+      "step": 12180
+    },
+    {
+      "epoch": 4.601736504341261,
+      "grad_norm": 0.5366525053977966,
+      "learning_rate": 8.328517313744978e-05,
+      "loss": 0.4433,
+      "step": 12190
+    },
+    {
+      "epoch": 4.605511513778785,
+      "grad_norm": 0.5605337619781494,
+      "learning_rate": 8.325431560898286e-05,
+      "loss": 0.4383,
+      "step": 12200
+    },
+    {
+      "epoch": 4.6092865232163085,
+      "grad_norm": 0.7788808941841125,
+      "learning_rate": 8.322343535150478e-05,
+      "loss": 0.4409,
+      "step": 12210
+    },
+    {
+      "epoch": 4.613061532653831,
+      "grad_norm": 0.6166380643844604,
+      "learning_rate": 8.319253238612191e-05,
+      "loss": 0.4415,
+      "step": 12220
+    },
+    {
+      "epoch": 4.616836542091355,
+      "grad_norm": 0.5688962936401367,
+      "learning_rate": 8.316160673395614e-05,
+      "loss": 0.4401,
+      "step": 12230
+    },
+    {
+      "epoch": 4.620611551528879,
+      "grad_norm": 0.6172645688056946,
+      "learning_rate": 8.313065841614487e-05,
+      "loss": 0.4439,
+      "step": 12240
+    },
+    {
+      "epoch": 4.6243865609664025,
+      "grad_norm": 0.7282325029373169,
+      "learning_rate": 8.309968745384096e-05,
+      "loss": 0.4383,
+      "step": 12250
+    },
+    {
+      "epoch": 4.628161570403926,
+      "grad_norm": 0.519595742225647,
+      "learning_rate": 8.306869386821282e-05,
+      "loss": 0.4384,
+      "step": 12260
+    },
+    {
+      "epoch": 4.631936579841449,
+      "grad_norm": 0.5448688268661499,
+      "learning_rate": 8.30376776804442e-05,
+      "loss": 0.4418,
+      "step": 12270
+    },
+    {
+      "epoch": 4.635711589278973,
+      "grad_norm": 0.5774092674255371,
+      "learning_rate": 8.300663891173443e-05,
+      "loss": 0.4396,
+      "step": 12280
+    },
+    {
+      "epoch": 4.639486598716497,
+      "grad_norm": 0.6279200315475464,
+      "learning_rate": 8.297557758329822e-05,
+      "loss": 0.4434,
+      "step": 12290
+    },
+    {
+      "epoch": 4.64326160815402,
+      "grad_norm": 0.6159646511077881,
+      "learning_rate": 8.294449371636564e-05,
+      "loss": 0.4406,
+      "step": 12300
+    },
+    {
+      "epoch": 4.647036617591544,
+      "grad_norm": 0.6391906142234802,
+      "learning_rate": 8.291338733218226e-05,
+      "loss": 0.4379,
+      "step": 12310
+    },
+    {
+      "epoch": 4.650811627029068,
+      "grad_norm": 0.6429669260978699,
+      "learning_rate": 8.2882258452009e-05,
+      "loss": 0.4328,
+      "step": 12320
+    },
+    {
+      "epoch": 4.6545866364665915,
+      "grad_norm": 0.5893004536628723,
+      "learning_rate": 8.285110709712214e-05,
+      "loss": 0.4385,
+      "step": 12330
+    },
+    {
+      "epoch": 4.658361645904114,
+      "grad_norm": 0.5372092127799988,
+      "learning_rate": 8.281993328881337e-05,
+      "loss": 0.434,
+      "step": 12340
+    },
+    {
+      "epoch": 4.662136655341638,
+      "grad_norm": 0.5648928284645081,
+      "learning_rate": 8.278873704838964e-05,
+      "loss": 0.4438,
+      "step": 12350
+    },
+    {
+      "epoch": 4.665911664779162,
+      "grad_norm": 0.5447239875793457,
+      "learning_rate": 8.275751839717334e-05,
+      "loss": 0.4314,
+      "step": 12360
+    },
+    {
+      "epoch": 4.669686674216686,
+      "grad_norm": 0.5694062113761902,
+      "learning_rate": 8.272627735650208e-05,
+      "loss": 0.443,
+      "step": 12370
+    },
+    {
+      "epoch": 4.673461683654209,
+      "grad_norm": 0.6510103940963745,
+      "learning_rate": 8.269501394772884e-05,
+      "loss": 0.4352,
+      "step": 12380
+    },
+    {
+      "epoch": 4.677236693091733,
+      "grad_norm": 0.5314978361129761,
+      "learning_rate": 8.266372819222189e-05,
+      "loss": 0.4385,
+      "step": 12390
+    },
+    {
+      "epoch": 4.681011702529256,
+      "grad_norm": 0.6842690110206604,
+      "learning_rate": 8.26324201113647e-05,
+      "loss": 0.4378,
+      "step": 12400
+    },
+    {
+      "epoch": 4.68478671196678,
+      "grad_norm": 0.6124461889266968,
+      "learning_rate": 8.260108972655606e-05,
+      "loss": 0.4416,
+      "step": 12410
+    },
+    {
+      "epoch": 4.688561721404303,
+      "grad_norm": 0.568488359451294,
+      "learning_rate": 8.256973705921e-05,
+      "loss": 0.4378,
+      "step": 12420
+    },
+    {
+      "epoch": 4.692336730841827,
+      "grad_norm": 0.5795332193374634,
+      "learning_rate": 8.25383621307558e-05,
+      "loss": 0.4411,
+      "step": 12430
+    },
+    {
+      "epoch": 4.696111740279351,
+      "grad_norm": 0.5848597884178162,
+      "learning_rate": 8.25069649626379e-05,
+      "loss": 0.4318,
+      "step": 12440
+    },
+    {
+      "epoch": 4.699886749716875,
+      "grad_norm": 0.5752859115600586,
+      "learning_rate": 8.247554557631596e-05,
+      "loss": 0.4373,
+      "step": 12450
+    },
+    {
+      "epoch": 4.703661759154398,
+      "grad_norm": 0.5697815418243408,
+      "learning_rate": 8.244410399326483e-05,
+      "loss": 0.4434,
+      "step": 12460
+    },
+    {
+      "epoch": 4.707436768591921,
+      "grad_norm": 0.5662826895713806,
+      "learning_rate": 8.241264023497457e-05,
+      "loss": 0.4412,
+      "step": 12470
+    },
+    {
+      "epoch": 4.711211778029445,
+      "grad_norm": 0.5779763460159302,
+      "learning_rate": 8.238115432295034e-05,
+      "loss": 0.4347,
+      "step": 12480
+    },
+    {
+      "epoch": 4.714986787466969,
+      "grad_norm": 0.5694817304611206,
+      "learning_rate": 8.234964627871247e-05,
+      "loss": 0.44,
+      "step": 12490
+    },
+    {
+      "epoch": 4.718761796904492,
+      "grad_norm": 0.6271215677261353,
+      "learning_rate": 8.231811612379639e-05,
+      "loss": 0.4388,
+      "step": 12500
+    },
+    {
+      "epoch": 4.722536806342016,
+      "grad_norm": 0.647670567035675,
+      "learning_rate": 8.228656387975268e-05,
+      "loss": 0.4424,
+      "step": 12510
+    },
+    {
+      "epoch": 4.726311815779539,
+      "grad_norm": 0.58333820104599,
+      "learning_rate": 8.225498956814702e-05,
+      "loss": 0.4432,
+      "step": 12520
+    },
+    {
+      "epoch": 4.730086825217063,
+      "grad_norm": 0.5770929455757141,
+      "learning_rate": 8.222339321056014e-05,
+      "loss": 0.4366,
+      "step": 12530
+    },
+    {
+      "epoch": 4.733861834654586,
+      "grad_norm": 0.5523325800895691,
+      "learning_rate": 8.219177482858785e-05,
+      "loss": 0.4332,
+      "step": 12540
+    },
+    {
+      "epoch": 4.73763684409211,
+      "grad_norm": 0.8121219873428345,
+      "learning_rate": 8.216013444384099e-05,
+      "loss": 0.4383,
+      "step": 12550
+    },
+    {
+      "epoch": 4.741411853529634,
+      "grad_norm": 0.5820472240447998,
+      "learning_rate": 8.21284720779455e-05,
+      "loss": 0.4418,
+      "step": 12560
+    },
+    {
+      "epoch": 4.745186862967158,
+      "grad_norm": 0.6363267302513123,
+      "learning_rate": 8.209678775254231e-05,
+      "loss": 0.4436,
+      "step": 12570
+    },
+    {
+      "epoch": 4.748961872404681,
+      "grad_norm": 0.5754694938659668,
+      "learning_rate": 8.206508148928733e-05,
+      "loss": 0.448,
+      "step": 12580
+    },
+    {
+      "epoch": 4.752736881842204,
+      "grad_norm": 0.6008169054985046,
+      "learning_rate": 8.203335330985151e-05,
+      "loss": 0.4448,
+      "step": 12590
+    },
+    {
+      "epoch": 4.756511891279728,
+      "grad_norm": 0.6524280905723572,
+      "learning_rate": 8.200160323592076e-05,
+      "loss": 0.4572,
+      "step": 12600
+    },
+    {
+      "epoch": 4.760286900717252,
+      "grad_norm": 0.5874181389808655,
+      "learning_rate": 8.196983128919598e-05,
+      "loss": 0.4404,
+      "step": 12610
+    },
+    {
+      "epoch": 4.764061910154775,
+      "grad_norm": 0.8897712826728821,
+      "learning_rate": 8.193803749139295e-05,
+      "loss": 0.4319,
+      "step": 12620
+    },
+    {
+      "epoch": 4.767836919592299,
+      "grad_norm": 0.5824182033538818,
+      "learning_rate": 8.190622186424244e-05,
+      "loss": 0.4436,
+      "step": 12630
+    },
+    {
+      "epoch": 4.771611929029823,
+      "grad_norm": 0.6077965497970581,
+      "learning_rate": 8.187438442949016e-05,
+      "loss": 0.4422,
+      "step": 12640
+    },
+    {
+      "epoch": 4.775386938467346,
+      "grad_norm": 0.7708337903022766,
+      "learning_rate": 8.184252520889668e-05,
+      "loss": 0.4312,
+      "step": 12650
+    },
+    {
+      "epoch": 4.7791619479048695,
+      "grad_norm": 0.5734137296676636,
+      "learning_rate": 8.181064422423748e-05,
+      "loss": 0.4382,
+      "step": 12660
+    },
+    {
+      "epoch": 4.782936957342393,
+      "grad_norm": 0.5938032865524292,
+      "learning_rate": 8.177874149730289e-05,
+      "loss": 0.4417,
+      "step": 12670
+    },
+    {
+      "epoch": 4.786711966779917,
+      "grad_norm": 0.6791886687278748,
+      "learning_rate": 8.174681704989816e-05,
+      "loss": 0.4365,
+      "step": 12680
+    },
+    {
+      "epoch": 4.790486976217441,
+      "grad_norm": 0.6398690342903137,
+      "learning_rate": 8.171487090384333e-05,
+      "loss": 0.4407,
+      "step": 12690
+    },
+    {
+      "epoch": 4.794261985654964,
+      "grad_norm": 0.6101489067077637,
+      "learning_rate": 8.168290308097328e-05,
+      "loss": 0.4296,
+      "step": 12700
+    },
+    {
+      "epoch": 4.798036995092488,
+      "grad_norm": 0.6080114841461182,
+      "learning_rate": 8.165091360313774e-05,
+      "loss": 0.4463,
+      "step": 12710
+    },
+    {
+      "epoch": 4.801812004530011,
+      "grad_norm": 0.5545832514762878,
+      "learning_rate": 8.161890249220119e-05,
+      "loss": 0.4354,
+      "step": 12720
+    },
+    {
+      "epoch": 4.805587013967535,
+      "grad_norm": 0.6806719899177551,
+      "learning_rate": 8.158686977004295e-05,
+      "loss": 0.4329,
+      "step": 12730
+    },
+    {
+      "epoch": 4.8093620234050585,
+      "grad_norm": 0.6922208070755005,
+      "learning_rate": 8.155481545855706e-05,
+      "loss": 0.4361,
+      "step": 12740
+    },
+    {
+      "epoch": 4.813137032842582,
+      "grad_norm": 0.6132349371910095,
+      "learning_rate": 8.152273957965233e-05,
+      "loss": 0.4392,
+      "step": 12750
+    },
+    {
+      "epoch": 4.816912042280106,
+      "grad_norm": 0.5297583341598511,
+      "learning_rate": 8.149064215525237e-05,
+      "loss": 0.4371,
+      "step": 12760
+    },
+    {
+      "epoch": 4.82068705171763,
+      "grad_norm": 0.5206873416900635,
+      "learning_rate": 8.14585232072954e-05,
+      "loss": 0.4358,
+      "step": 12770
+    },
+    {
+      "epoch": 4.8244620611551525,
+      "grad_norm": 0.5716885328292847,
+      "learning_rate": 8.142638275773449e-05,
+      "loss": 0.4391,
+      "step": 12780
+    },
+    {
+      "epoch": 4.828237070592676,
+      "grad_norm": 0.5915391445159912,
+      "learning_rate": 8.139422082853729e-05,
+      "loss": 0.4369,
+      "step": 12790
+    },
+    {
+      "epoch": 4.8320120800302,
+      "grad_norm": 0.5491467118263245,
+      "learning_rate": 8.136203744168618e-05,
+      "loss": 0.4377,
+      "step": 12800
+    },
+    {
+      "epoch": 4.835787089467724,
+      "grad_norm": 0.5877269506454468,
+      "learning_rate": 8.132983261917819e-05,
+      "loss": 0.4327,
+      "step": 12810
+    },
+    {
+      "epoch": 4.8395620989052475,
+      "grad_norm": 0.8957927227020264,
+      "learning_rate": 8.129760638302504e-05,
+      "loss": 0.4363,
+      "step": 12820
+    },
+    {
+      "epoch": 4.843337108342771,
+      "grad_norm": 0.6469703912734985,
+      "learning_rate": 8.126535875525305e-05,
+      "loss": 0.4354,
+      "step": 12830
+    },
+    {
+      "epoch": 4.847112117780295,
+      "grad_norm": 0.5585578680038452,
+      "learning_rate": 8.123308975790316e-05,
+      "loss": 0.4417,
+      "step": 12840
+    },
+    {
+      "epoch": 4.850887127217818,
+      "grad_norm": 0.858641505241394,
+      "learning_rate": 8.120079941303094e-05,
+      "loss": 0.4365,
+      "step": 12850
+    },
+    {
+      "epoch": 4.8546621366553415,
+      "grad_norm": 0.6256632804870605,
+      "learning_rate": 8.116848774270651e-05,
+      "loss": 0.44,
+      "step": 12860
+    },
+    {
+      "epoch": 4.858437146092865,
+      "grad_norm": 0.6245275735855103,
+      "learning_rate": 8.113615476901461e-05,
+      "loss": 0.4331,
+      "step": 12870
+    },
+    {
+      "epoch": 4.862212155530389,
+      "grad_norm": 0.6392505168914795,
+      "learning_rate": 8.110380051405454e-05,
+      "loss": 0.4316,
+      "step": 12880
+    },
+    {
+      "epoch": 4.865987164967913,
+      "grad_norm": 0.6834761500358582,
+      "learning_rate": 8.107142499994009e-05,
+      "loss": 0.4369,
+      "step": 12890
+    },
+    {
+      "epoch": 4.869762174405436,
+      "grad_norm": 0.6249173879623413,
+      "learning_rate": 8.103902824879966e-05,
+      "loss": 0.4378,
+      "step": 12900
+    },
+    {
+      "epoch": 4.873537183842959,
+      "grad_norm": 0.7045850157737732,
+      "learning_rate": 8.10066102827761e-05,
+      "loss": 0.4355,
+      "step": 12910
+    },
+    {
+      "epoch": 4.877312193280483,
+      "grad_norm": 0.6210693120956421,
+      "learning_rate": 8.097417112402676e-05,
+      "loss": 0.4387,
+      "step": 12920
+    },
+    {
+      "epoch": 4.881087202718007,
+      "grad_norm": 0.6081416010856628,
+      "learning_rate": 8.094171079472355e-05,
+      "loss": 0.4362,
+      "step": 12930
+    },
+    {
+      "epoch": 4.8848622121555305,
+      "grad_norm": 0.651164710521698,
+      "learning_rate": 8.090922931705277e-05,
+      "loss": 0.4349,
+      "step": 12940
+    },
+    {
+      "epoch": 4.888637221593054,
+      "grad_norm": 0.6259034276008606,
+      "learning_rate": 8.08767267132152e-05,
+      "loss": 0.4379,
+      "step": 12950
+    },
+    {
+      "epoch": 4.892412231030578,
+      "grad_norm": 0.5768272876739502,
+      "learning_rate": 8.084420300542608e-05,
+      "loss": 0.4296,
+      "step": 12960
+    },
+    {
+      "epoch": 4.896187240468101,
+      "grad_norm": 0.6301991939544678,
+      "learning_rate": 8.081165821591505e-05,
+      "loss": 0.434,
+      "step": 12970
+    },
+    {
+      "epoch": 4.899962249905625,
+      "grad_norm": 0.6101955771446228,
+      "learning_rate": 8.077909236692615e-05,
+      "loss": 0.4301,
+      "step": 12980
+    },
+    {
+      "epoch": 4.903737259343148,
+      "grad_norm": 0.6158044338226318,
+      "learning_rate": 8.074650548071787e-05,
+      "loss": 0.4378,
+      "step": 12990
+    },
+    {
+      "epoch": 4.907512268780672,
+      "grad_norm": 0.5665403604507446,
+      "learning_rate": 8.071389757956301e-05,
+      "loss": 0.436,
+      "step": 13000
+    },
+    {
+      "epoch": 4.911287278218196,
+      "grad_norm": 0.6046426892280579,
+      "learning_rate": 8.068126868574876e-05,
+      "loss": 0.4314,
+      "step": 13010
+    },
+    {
+      "epoch": 4.9150622876557195,
+      "grad_norm": 0.5915471315383911,
+      "learning_rate": 8.064861882157668e-05,
+      "loss": 0.435,
+      "step": 13020
+    },
+    {
+      "epoch": 4.918837297093242,
+      "grad_norm": 0.6827958226203918,
+      "learning_rate": 8.061594800936263e-05,
+      "loss": 0.4365,
+      "step": 13030
+    },
+    {
+      "epoch": 4.922612306530766,
+      "grad_norm": 0.5622991323471069,
+      "learning_rate": 8.058325627143681e-05,
+      "loss": 0.4344,
+      "step": 13040
+    },
+    {
+      "epoch": 4.92638731596829,
+      "grad_norm": 0.5483256578445435,
+      "learning_rate": 8.055054363014372e-05,
+      "loss": 0.4383,
+      "step": 13050
+    },
+    {
+      "epoch": 4.930162325405814,
+      "grad_norm": 0.6701416969299316,
+      "learning_rate": 8.051781010784211e-05,
+      "loss": 0.4397,
+      "step": 13060
+    },
+    {
+      "epoch": 4.933937334843337,
+      "grad_norm": 0.5472255945205688,
+      "learning_rate": 8.048505572690506e-05,
+      "loss": 0.4302,
+      "step": 13070
+    },
+    {
+      "epoch": 4.937712344280861,
+      "grad_norm": 0.655945897102356,
+      "learning_rate": 8.045228050971988e-05,
+      "loss": 0.4367,
+      "step": 13080
+    },
+    {
+      "epoch": 4.941487353718385,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 8.041948447868814e-05,
+      "loss": 0.4292,
+      "step": 13090
+    },
+    {
+      "epoch": 4.945262363155908,
+      "grad_norm": 0.7679628133773804,
+      "learning_rate": 8.038666765622558e-05,
+      "loss": 0.4333,
+      "step": 13100
+    },
+    {
+      "epoch": 4.949037372593431,
+      "grad_norm": 0.5666245818138123,
+      "learning_rate": 8.03538300647622e-05,
+      "loss": 0.443,
+      "step": 13110
+    },
+    {
+      "epoch": 4.952812382030955,
+      "grad_norm": 0.6425677537918091,
+      "learning_rate": 8.03209717267422e-05,
+      "loss": 0.4288,
+      "step": 13120
+    },
+    {
+      "epoch": 4.956587391468479,
+      "grad_norm": 0.585444450378418,
+      "learning_rate": 8.028809266462395e-05,
+      "loss": 0.4317,
+      "step": 13130
+    },
+    {
+      "epoch": 4.9603624009060026,
+      "grad_norm": 0.6069588661193848,
+      "learning_rate": 8.025519290087994e-05,
+      "loss": 0.4352,
+      "step": 13140
+    },
+    {
+      "epoch": 4.964137410343525,
+      "grad_norm": 0.5473458170890808,
+      "learning_rate": 8.022227245799688e-05,
+      "loss": 0.4407,
+      "step": 13150
+    },
+    {
+      "epoch": 4.967912419781049,
+      "grad_norm": 0.5702902674674988,
+      "learning_rate": 8.018933135847557e-05,
+      "loss": 0.4293,
+      "step": 13160
+    },
+    {
+      "epoch": 4.971687429218573,
+      "grad_norm": 0.6425644755363464,
+      "learning_rate": 8.015636962483096e-05,
+      "loss": 0.4394,
+      "step": 13170
+    },
+    {
+      "epoch": 4.975462438656097,
+      "grad_norm": 0.5478054285049438,
+      "learning_rate": 8.012338727959205e-05,
+      "loss": 0.4267,
+      "step": 13180
+    },
+    {
+      "epoch": 4.97923744809362,
+      "grad_norm": 0.6049672961235046,
+      "learning_rate": 8.009038434530198e-05,
+      "loss": 0.4381,
+      "step": 13190
+    },
+    {
+      "epoch": 4.983012457531144,
+      "grad_norm": 0.5757096409797668,
+      "learning_rate": 8.005736084451796e-05,
+      "loss": 0.4292,
+      "step": 13200
+    },
+    {
+      "epoch": 4.986787466968668,
+      "grad_norm": 0.6315653920173645,
+      "learning_rate": 8.002431679981122e-05,
+      "loss": 0.4249,
+      "step": 13210
+    },
+    {
+      "epoch": 4.990562476406191,
+      "grad_norm": 0.6188434958457947,
+      "learning_rate": 7.999125223376706e-05,
+      "loss": 0.4312,
+      "step": 13220
+    },
+    {
+      "epoch": 4.994337485843714,
+      "grad_norm": 0.5495557188987732,
+      "learning_rate": 7.99581671689848e-05,
+      "loss": 0.4272,
+      "step": 13230
+    },
+    {
+      "epoch": 4.998112495281238,
+      "grad_norm": 0.6032848954200745,
+      "learning_rate": 7.992506162807775e-05,
+      "loss": 0.4378,
+      "step": 13240
+    },
+    {
+      "epoch": 5.001887504718762,
+      "grad_norm": 0.563420832157135,
+      "learning_rate": 7.989193563367328e-05,
+      "loss": 0.4333,
+      "step": 13250
+    },
+    {
+      "epoch": 5.005662514156286,
+      "grad_norm": 0.5446894764900208,
+      "learning_rate": 7.985878920841266e-05,
+      "loss": 0.4252,
+      "step": 13260
+    },
+    {
+      "epoch": 5.009437523593809,
+      "grad_norm": 0.6635614633560181,
+      "learning_rate": 7.982562237495117e-05,
+      "loss": 0.4289,
+      "step": 13270
+    },
+    {
+      "epoch": 5.013212533031332,
+      "grad_norm": 0.6846327781677246,
+      "learning_rate": 7.979243515595802e-05,
+      "loss": 0.4408,
+      "step": 13280
+    },
+    {
+      "epoch": 5.016987542468856,
+      "grad_norm": 0.6627950668334961,
+      "learning_rate": 7.975922757411636e-05,
+      "loss": 0.4443,
+      "step": 13290
+    },
+    {
+      "epoch": 5.02076255190638,
+      "grad_norm": 0.5397953987121582,
+      "learning_rate": 7.972599965212329e-05,
+      "loss": 0.4393,
+      "step": 13300
+    },
+    {
+      "epoch": 5.024537561343903,
+      "grad_norm": 0.5803020000457764,
+      "learning_rate": 7.969275141268973e-05,
+      "loss": 0.4316,
+      "step": 13310
+    },
+    {
+      "epoch": 5.028312570781427,
+      "grad_norm": 0.625580370426178,
+      "learning_rate": 7.96594828785406e-05,
+      "loss": 0.4343,
+      "step": 13320
+    },
+    {
+      "epoch": 5.032087580218951,
+      "grad_norm": 0.5824500918388367,
+      "learning_rate": 7.962619407241456e-05,
+      "loss": 0.4314,
+      "step": 13330
+    },
+    {
+      "epoch": 5.035862589656474,
+      "grad_norm": 0.6776422262191772,
+      "learning_rate": 7.959288501706424e-05,
+      "loss": 0.444,
+      "step": 13340
+    },
+    {
+      "epoch": 5.0396375990939974,
+      "grad_norm": 0.5621914267539978,
+      "learning_rate": 7.955955573525605e-05,
+      "loss": 0.435,
+      "step": 13350
+    },
+    {
+      "epoch": 5.043412608531521,
+      "grad_norm": 0.6005207300186157,
+      "learning_rate": 7.952620624977026e-05,
+      "loss": 0.4282,
+      "step": 13360
+    },
+    {
+      "epoch": 5.047187617969045,
+      "grad_norm": 0.5816413760185242,
+      "learning_rate": 7.949283658340089e-05,
+      "loss": 0.4396,
+      "step": 13370
+    },
+    {
+      "epoch": 5.050962627406569,
+      "grad_norm": 0.6071826219558716,
+      "learning_rate": 7.945944675895585e-05,
+      "loss": 0.4344,
+      "step": 13380
+    },
+    {
+      "epoch": 5.054737636844092,
+      "grad_norm": 0.5671345591545105,
+      "learning_rate": 7.942603679925671e-05,
+      "loss": 0.4225,
+      "step": 13390
+    },
+    {
+      "epoch": 5.058512646281616,
+      "grad_norm": 0.5134578943252563,
+      "learning_rate": 7.93926067271389e-05,
+      "loss": 0.4412,
+      "step": 13400
+    },
+    {
+      "epoch": 5.062287655719139,
+      "grad_norm": 0.5748825669288635,
+      "learning_rate": 7.935915656545155e-05,
+      "loss": 0.4246,
+      "step": 13410
+    },
+    {
+      "epoch": 5.066062665156663,
+      "grad_norm": 0.6098251938819885,
+      "learning_rate": 7.932568633705752e-05,
+      "loss": 0.4316,
+      "step": 13420
+    },
+    {
+      "epoch": 5.069837674594186,
+      "grad_norm": 0.5781272649765015,
+      "learning_rate": 7.929219606483341e-05,
+      "loss": 0.4375,
+      "step": 13430
+    },
+    {
+      "epoch": 5.07361268403171,
+      "grad_norm": 0.5721918940544128,
+      "learning_rate": 7.925868577166948e-05,
+      "loss": 0.4328,
+      "step": 13440
+    },
+    {
+      "epoch": 5.077387693469234,
+      "grad_norm": 0.6408784985542297,
+      "learning_rate": 7.922515548046974e-05,
+      "loss": 0.4351,
+      "step": 13450
+    },
+    {
+      "epoch": 5.081162702906758,
+      "grad_norm": 0.5710509419441223,
+      "learning_rate": 7.919160521415179e-05,
+      "loss": 0.4305,
+      "step": 13460
+    },
+    {
+      "epoch": 5.0849377123442805,
+      "grad_norm": 0.6249564290046692,
+      "learning_rate": 7.915803499564694e-05,
+      "loss": 0.4246,
+      "step": 13470
+    },
+    {
+      "epoch": 5.088712721781804,
+      "grad_norm": 0.6229961514472961,
+      "learning_rate": 7.912444484790013e-05,
+      "loss": 0.4307,
+      "step": 13480
+    },
+    {
+      "epoch": 5.092487731219328,
+      "grad_norm": 0.5835480093955994,
+      "learning_rate": 7.909083479386987e-05,
+      "loss": 0.4262,
+      "step": 13490
+    },
+    {
+      "epoch": 5.096262740656852,
+      "grad_norm": 0.6598435044288635,
+      "learning_rate": 7.905720485652836e-05,
+      "loss": 0.4326,
+      "step": 13500
+    },
+    {
+      "epoch": 5.100037750094375,
+      "grad_norm": 0.6232707500457764,
+      "learning_rate": 7.902355505886132e-05,
+      "loss": 0.4312,
+      "step": 13510
+    },
+    {
+      "epoch": 5.103812759531899,
+      "grad_norm": 0.5751008987426758,
+      "learning_rate": 7.898988542386805e-05,
+      "loss": 0.4318,
+      "step": 13520
+    },
+    {
+      "epoch": 5.107587768969422,
+      "grad_norm": 0.7185839414596558,
+      "learning_rate": 7.895619597456147e-05,
+      "loss": 0.431,
+      "step": 13530
+    },
+    {
+      "epoch": 5.111362778406946,
+      "grad_norm": 0.6064713597297668,
+      "learning_rate": 7.892248673396798e-05,
+      "loss": 0.4428,
+      "step": 13540
+    },
+    {
+      "epoch": 5.1151377878444695,
+      "grad_norm": 0.5585970282554626,
+      "learning_rate": 7.888875772512754e-05,
+      "loss": 0.4285,
+      "step": 13550
+    },
+    {
+      "epoch": 5.118912797281993,
+      "grad_norm": 0.590541660785675,
+      "learning_rate": 7.885500897109359e-05,
+      "loss": 0.4327,
+      "step": 13560
+    },
+    {
+      "epoch": 5.122687806719517,
+      "grad_norm": 0.6079216003417969,
+      "learning_rate": 7.882124049493309e-05,
+      "loss": 0.4219,
+      "step": 13570
+    },
+    {
+      "epoch": 5.126462816157041,
+      "grad_norm": 0.6529505848884583,
+      "learning_rate": 7.878745231972649e-05,
+      "loss": 0.4195,
+      "step": 13580
+    },
+    {
+      "epoch": 5.130237825594564,
+      "grad_norm": 0.6022513508796692,
+      "learning_rate": 7.875364446856766e-05,
+      "loss": 0.4318,
+      "step": 13590
+    },
+    {
+      "epoch": 5.134012835032087,
+      "grad_norm": 0.6363440752029419,
+      "learning_rate": 7.871981696456398e-05,
+      "loss": 0.4323,
+      "step": 13600
+    },
+    {
+      "epoch": 5.137787844469611,
+      "grad_norm": 0.6193153262138367,
+      "learning_rate": 7.868596983083623e-05,
+      "loss": 0.4285,
+      "step": 13610
+    },
+    {
+      "epoch": 5.141562853907135,
+      "grad_norm": 0.7710440158843994,
+      "learning_rate": 7.865210309051858e-05,
+      "loss": 0.4266,
+      "step": 13620
+    },
+    {
+      "epoch": 5.1453378633446585,
+      "grad_norm": 0.5872613787651062,
+      "learning_rate": 7.861821676675863e-05,
+      "loss": 0.4347,
+      "step": 13630
+    },
+    {
+      "epoch": 5.149112872782182,
+      "grad_norm": 0.536893367767334,
+      "learning_rate": 7.858431088271739e-05,
+      "loss": 0.4304,
+      "step": 13640
+    },
+    {
+      "epoch": 5.152887882219706,
+      "grad_norm": 0.5925413966178894,
+      "learning_rate": 7.855038546156918e-05,
+      "loss": 0.4284,
+      "step": 13650
+    },
+    {
+      "epoch": 5.156662891657229,
+      "grad_norm": 0.562650203704834,
+      "learning_rate": 7.851644052650173e-05,
+      "loss": 0.4268,
+      "step": 13660
+    },
+    {
+      "epoch": 5.1604379010947525,
+      "grad_norm": 0.5925480723381042,
+      "learning_rate": 7.848247610071609e-05,
+      "loss": 0.4251,
+      "step": 13670
+    },
+    {
+      "epoch": 5.164212910532276,
+      "grad_norm": 0.6896264553070068,
+      "learning_rate": 7.844849220742658e-05,
+      "loss": 0.43,
+      "step": 13680
+    },
+    {
+      "epoch": 5.1679879199698,
+      "grad_norm": 0.643406867980957,
+      "learning_rate": 7.841448886986092e-05,
+      "loss": 0.4298,
+      "step": 13690
+    },
+    {
+      "epoch": 5.171762929407324,
+      "grad_norm": 0.5790614485740662,
+      "learning_rate": 7.838046611126004e-05,
+      "loss": 0.4333,
+      "step": 13700
+    },
+    {
+      "epoch": 5.1755379388448475,
+      "grad_norm": 0.6230219602584839,
+      "learning_rate": 7.834642395487819e-05,
+      "loss": 0.432,
+      "step": 13710
+    },
+    {
+      "epoch": 5.17931294828237,
+      "grad_norm": 0.6495296955108643,
+      "learning_rate": 7.831236242398285e-05,
+      "loss": 0.4336,
+      "step": 13720
+    },
+    {
+      "epoch": 5.183087957719894,
+      "grad_norm": 0.5876701474189758,
+      "learning_rate": 7.827828154185477e-05,
+      "loss": 0.4362,
+      "step": 13730
+    },
+    {
+      "epoch": 5.186862967157418,
+      "grad_norm": 0.5457055568695068,
+      "learning_rate": 7.82441813317879e-05,
+      "loss": 0.43,
+      "step": 13740
+    },
+    {
+      "epoch": 5.1906379765949415,
+      "grad_norm": 0.618160605430603,
+      "learning_rate": 7.821006181708944e-05,
+      "loss": 0.4281,
+      "step": 13750
+    },
+    {
+      "epoch": 5.194412986032465,
+      "grad_norm": 0.5848748087882996,
+      "learning_rate": 7.81759230210797e-05,
+      "loss": 0.4397,
+      "step": 13760
+    },
+    {
+      "epoch": 5.198187995469989,
+      "grad_norm": 0.5916726589202881,
+      "learning_rate": 7.814176496709227e-05,
+      "loss": 0.4268,
+      "step": 13770
+    },
+    {
+      "epoch": 5.201963004907512,
+      "grad_norm": 0.5741629004478455,
+      "learning_rate": 7.810758767847385e-05,
+      "loss": 0.4286,
+      "step": 13780
+    },
+    {
+      "epoch": 5.205738014345036,
+      "grad_norm": 0.548261284828186,
+      "learning_rate": 7.807339117858427e-05,
+      "loss": 0.4242,
+      "step": 13790
+    },
+    {
+      "epoch": 5.209513023782559,
+      "grad_norm": 0.5949342846870422,
+      "learning_rate": 7.803917549079655e-05,
+      "loss": 0.4291,
+      "step": 13800
+    },
+    {
+      "epoch": 5.213288033220083,
+      "grad_norm": 0.611446738243103,
+      "learning_rate": 7.800494063849679e-05,
+      "loss": 0.4322,
+      "step": 13810
+    },
+    {
+      "epoch": 5.217063042657607,
+      "grad_norm": 0.5855134129524231,
+      "learning_rate": 7.797068664508416e-05,
+      "loss": 0.4329,
+      "step": 13820
+    },
+    {
+      "epoch": 5.2208380520951305,
+      "grad_norm": 0.615392804145813,
+      "learning_rate": 7.793641353397096e-05,
+      "loss": 0.4321,
+      "step": 13830
+    },
+    {
+      "epoch": 5.224613061532654,
+      "grad_norm": 0.6607218980789185,
+      "learning_rate": 7.790212132858253e-05,
+      "loss": 0.4314,
+      "step": 13840
+    },
+    {
+      "epoch": 5.228388070970177,
+      "grad_norm": 0.6020424365997314,
+      "learning_rate": 7.786781005235728e-05,
+      "loss": 0.4282,
+      "step": 13850
+    },
+    {
+      "epoch": 5.232163080407701,
+      "grad_norm": 0.5580222606658936,
+      "learning_rate": 7.783347972874662e-05,
+      "loss": 0.4303,
+      "step": 13860
+    },
+    {
+      "epoch": 5.235938089845225,
+      "grad_norm": 0.5938263535499573,
+      "learning_rate": 7.779913038121504e-05,
+      "loss": 0.4318,
+      "step": 13870
+    },
+    {
+      "epoch": 5.239713099282748,
+      "grad_norm": 0.5820572376251221,
+      "learning_rate": 7.776476203323997e-05,
+      "loss": 0.4244,
+      "step": 13880
+    },
+    {
+      "epoch": 5.243488108720272,
+      "grad_norm": 0.5716726779937744,
+      "learning_rate": 7.773037470831185e-05,
+      "loss": 0.4268,
+      "step": 13890
+    },
+    {
+      "epoch": 5.247263118157796,
+      "grad_norm": 0.6699029803276062,
+      "learning_rate": 7.76959684299341e-05,
+      "loss": 0.4448,
+      "step": 13900
+    },
+    {
+      "epoch": 5.251038127595319,
+      "grad_norm": 0.6544981002807617,
+      "learning_rate": 7.76615432216231e-05,
+      "loss": 0.4324,
+      "step": 13910
+    },
+    {
+      "epoch": 5.254813137032842,
+      "grad_norm": 0.6434966325759888,
+      "learning_rate": 7.762709910690811e-05,
+      "loss": 0.431,
+      "step": 13920
+    },
+    {
+      "epoch": 5.258588146470366,
+      "grad_norm": 0.6706520915031433,
+      "learning_rate": 7.759263610933141e-05,
+      "loss": 0.4239,
+      "step": 13930
+    },
+    {
+      "epoch": 5.26236315590789,
+      "grad_norm": 0.7404385209083557,
+      "learning_rate": 7.755815425244811e-05,
+      "loss": 0.4211,
+      "step": 13940
+    },
+    {
+      "epoch": 5.266138165345414,
+      "grad_norm": 0.750700831413269,
+      "learning_rate": 7.752365355982624e-05,
+      "loss": 0.4297,
+      "step": 13950
+    },
+    {
+      "epoch": 5.269913174782937,
+      "grad_norm": 0.6595298051834106,
+      "learning_rate": 7.748913405504668e-05,
+      "loss": 0.4395,
+      "step": 13960
+    },
+    {
+      "epoch": 5.27368818422046,
+      "grad_norm": 0.7262568473815918,
+      "learning_rate": 7.745459576170322e-05,
+      "loss": 0.4189,
+      "step": 13970
+    },
+    {
+      "epoch": 5.277463193657984,
+      "grad_norm": 0.6093210577964783,
+      "learning_rate": 7.742003870340242e-05,
+      "loss": 0.4403,
+      "step": 13980
+    },
+    {
+      "epoch": 5.281238203095508,
+      "grad_norm": 0.6888540387153625,
+      "learning_rate": 7.738546290376373e-05,
+      "loss": 0.4357,
+      "step": 13990
+    },
+    {
+      "epoch": 5.285013212533031,
+      "grad_norm": 0.9906258583068848,
+      "learning_rate": 7.735086838641937e-05,
+      "loss": 0.4655,
+      "step": 14000
+    },
+    {
+      "epoch": 5.288788221970555,
+      "grad_norm": 0.6373776197433472,
+      "learning_rate": 7.731625517501437e-05,
+      "loss": 0.4296,
+      "step": 14010
+    },
+    {
+      "epoch": 5.292563231408079,
+      "grad_norm": 0.5870838761329651,
+      "learning_rate": 7.728162329320655e-05,
+      "loss": 0.4318,
+      "step": 14020
+    },
+    {
+      "epoch": 5.296338240845602,
+      "grad_norm": 0.5722277164459229,
+      "learning_rate": 7.724697276466645e-05,
+      "loss": 0.426,
+      "step": 14030
+    },
+    {
+      "epoch": 5.300113250283125,
+      "grad_norm": 0.595984697341919,
+      "learning_rate": 7.721230361307738e-05,
+      "loss": 0.4287,
+      "step": 14040
+    },
+    {
+      "epoch": 5.303888259720649,
+      "grad_norm": 2.9559850692749023,
+      "learning_rate": 7.71776158621354e-05,
+      "loss": 0.4298,
+      "step": 14050
+    },
+    {
+      "epoch": 5.307663269158173,
+      "grad_norm": 0.5847383737564087,
+      "learning_rate": 7.714290953554925e-05,
+      "loss": 0.4304,
+      "step": 14060
+    },
+    {
+      "epoch": 5.311438278595697,
+      "grad_norm": 0.5809390544891357,
+      "learning_rate": 7.710818465704037e-05,
+      "loss": 0.4305,
+      "step": 14070
+    },
+    {
+      "epoch": 5.31521328803322,
+      "grad_norm": 0.5706303119659424,
+      "learning_rate": 7.707344125034288e-05,
+      "loss": 0.4256,
+      "step": 14080
+    },
+    {
+      "epoch": 5.318988297470744,
+      "grad_norm": 0.5360699892044067,
+      "learning_rate": 7.703867933920359e-05,
+      "loss": 0.4197,
+      "step": 14090
+    },
+    {
+      "epoch": 5.322763306908267,
+      "grad_norm": 0.5971339344978333,
+      "learning_rate": 7.700389894738194e-05,
+      "loss": 0.4181,
+      "step": 14100
+    },
+    {
+      "epoch": 5.326538316345791,
+      "grad_norm": 0.6393430233001709,
+      "learning_rate": 7.696910009864999e-05,
+      "loss": 0.4214,
+      "step": 14110
+    },
+    {
+      "epoch": 5.330313325783314,
+      "grad_norm": 0.5459288358688354,
+      "learning_rate": 7.693428281679241e-05,
+      "loss": 0.422,
+      "step": 14120
+    },
+    {
+      "epoch": 5.334088335220838,
+      "grad_norm": 0.516596794128418,
+      "learning_rate": 7.689944712560652e-05,
+      "loss": 0.4297,
+      "step": 14130
+    },
+    {
+      "epoch": 5.337863344658362,
+      "grad_norm": 0.5618916153907776,
+      "learning_rate": 7.686459304890214e-05,
+      "loss": 0.426,
+      "step": 14140
+    },
+    {
+      "epoch": 5.341638354095886,
+      "grad_norm": 0.6064931750297546,
+      "learning_rate": 7.682972061050175e-05,
+      "loss": 0.424,
+      "step": 14150
+    },
+    {
+      "epoch": 5.3454133635334085,
+      "grad_norm": 0.6051250696182251,
+      "learning_rate": 7.679482983424032e-05,
+      "loss": 0.4242,
+      "step": 14160
+    },
+    {
+      "epoch": 5.349188372970932,
+      "grad_norm": 0.6006349921226501,
+      "learning_rate": 7.675992074396534e-05,
+      "loss": 0.4295,
+      "step": 14170
+    },
+    {
+      "epoch": 5.352963382408456,
+      "grad_norm": 1.0653759241104126,
+      "learning_rate": 7.672499336353687e-05,
+      "loss": 0.4233,
+      "step": 14180
+    },
+    {
+      "epoch": 5.35673839184598,
+      "grad_norm": 0.6030228734016418,
+      "learning_rate": 7.669004771682744e-05,
+      "loss": 0.4259,
+      "step": 14190
+    },
+    {
+      "epoch": 5.360513401283503,
+      "grad_norm": 0.6684322357177734,
+      "learning_rate": 7.665508382772206e-05,
+      "loss": 0.4259,
+      "step": 14200
+    },
+    {
+      "epoch": 5.364288410721027,
+      "grad_norm": 0.6103203892707825,
+      "learning_rate": 7.662010172011824e-05,
+      "loss": 0.4218,
+      "step": 14210
+    },
+    {
+      "epoch": 5.36806342015855,
+      "grad_norm": 0.5854779481887817,
+      "learning_rate": 7.658510141792588e-05,
+      "loss": 0.4223,
+      "step": 14220
+    },
+    {
+      "epoch": 5.371838429596074,
+      "grad_norm": 0.6130976676940918,
+      "learning_rate": 7.65500829450674e-05,
+      "loss": 0.4235,
+      "step": 14230
+    },
+    {
+      "epoch": 5.3756134390335975,
+      "grad_norm": 0.5876962542533875,
+      "learning_rate": 7.651504632547759e-05,
+      "loss": 0.4262,
+      "step": 14240
+    },
+    {
+      "epoch": 5.379388448471121,
+      "grad_norm": 0.5816875100135803,
+      "learning_rate": 7.647999158310364e-05,
+      "loss": 0.4238,
+      "step": 14250
+    },
+    {
+      "epoch": 5.383163457908645,
+      "grad_norm": 0.6344910860061646,
+      "learning_rate": 7.644491874190512e-05,
+      "loss": 0.4277,
+      "step": 14260
+    },
+    {
+      "epoch": 5.386938467346169,
+      "grad_norm": 0.5912450551986694,
+      "learning_rate": 7.6409827825854e-05,
+      "loss": 0.4217,
+      "step": 14270
+    },
+    {
+      "epoch": 5.3907134767836915,
+      "grad_norm": 0.5558596849441528,
+      "learning_rate": 7.637471885893459e-05,
+      "loss": 0.4291,
+      "step": 14280
+    },
+    {
+      "epoch": 5.394488486221215,
+      "grad_norm": 0.6248513460159302,
+      "learning_rate": 7.633959186514354e-05,
+      "loss": 0.4281,
+      "step": 14290
+    },
+    {
+      "epoch": 5.398263495658739,
+      "grad_norm": 0.623848557472229,
+      "learning_rate": 7.630444686848984e-05,
+      "loss": 0.4189,
+      "step": 14300
+    },
+    {
+      "epoch": 5.402038505096263,
+      "grad_norm": 0.5933666825294495,
+      "learning_rate": 7.626928389299471e-05,
+      "loss": 0.4238,
+      "step": 14310
+    },
+    {
+      "epoch": 5.4058135145337864,
+      "grad_norm": 0.6335635781288147,
+      "learning_rate": 7.623410296269175e-05,
+      "loss": 0.4352,
+      "step": 14320
+    },
+    {
+      "epoch": 5.40958852397131,
+      "grad_norm": 0.5906968116760254,
+      "learning_rate": 7.61989041016268e-05,
+      "loss": 0.4217,
+      "step": 14330
+    },
+    {
+      "epoch": 5.413363533408834,
+      "grad_norm": 1.1786829233169556,
+      "learning_rate": 7.616368733385793e-05,
+      "loss": 0.4292,
+      "step": 14340
+    },
+    {
+      "epoch": 5.417138542846357,
+      "grad_norm": 0.6028378009796143,
+      "learning_rate": 7.612845268345547e-05,
+      "loss": 0.4245,
+      "step": 14350
+    },
+    {
+      "epoch": 5.4209135522838805,
+      "grad_norm": 1.1031107902526855,
+      "learning_rate": 7.609320017450199e-05,
+      "loss": 0.4235,
+      "step": 14360
+    },
+    {
+      "epoch": 5.424688561721404,
+      "grad_norm": 0.6411410570144653,
+      "learning_rate": 7.605792983109222e-05,
+      "loss": 0.4187,
+      "step": 14370
+    },
+    {
+      "epoch": 5.428463571158928,
+      "grad_norm": 0.547097384929657,
+      "learning_rate": 7.602264167733313e-05,
+      "loss": 0.4282,
+      "step": 14380
+    },
+    {
+      "epoch": 5.432238580596452,
+      "grad_norm": 0.6103999018669128,
+      "learning_rate": 7.598733573734384e-05,
+      "loss": 0.4256,
+      "step": 14390
+    },
+    {
+      "epoch": 5.436013590033975,
+      "grad_norm": 0.5741267800331116,
+      "learning_rate": 7.595201203525561e-05,
+      "loss": 0.4254,
+      "step": 14400
+    },
+    {
+      "epoch": 5.439788599471498,
+      "grad_norm": 0.5826230645179749,
+      "learning_rate": 7.591667059521187e-05,
+      "loss": 0.4279,
+      "step": 14410
+    },
+    {
+      "epoch": 5.443563608909022,
+      "grad_norm": 0.5484496355056763,
+      "learning_rate": 7.588131144136815e-05,
+      "loss": 0.429,
+      "step": 14420
+    },
+    {
+      "epoch": 5.447338618346546,
+      "grad_norm": 0.566040575504303,
+      "learning_rate": 7.584593459789212e-05,
+      "loss": 0.4213,
+      "step": 14430
+    },
+    {
+      "epoch": 5.4511136277840695,
+      "grad_norm": 0.5821641087532043,
+      "learning_rate": 7.58105400889635e-05,
+      "loss": 0.4327,
+      "step": 14440
+    },
+    {
+      "epoch": 5.454888637221593,
+      "grad_norm": 0.6450483798980713,
+      "learning_rate": 7.57751279387741e-05,
+      "loss": 0.4227,
+      "step": 14450
+    },
+    {
+      "epoch": 5.458663646659117,
+      "grad_norm": 0.5825949907302856,
+      "learning_rate": 7.573969817152782e-05,
+      "loss": 0.423,
+      "step": 14460
+    },
+    {
+      "epoch": 5.462438656096641,
+      "grad_norm": 0.5413886308670044,
+      "learning_rate": 7.570425081144052e-05,
+      "loss": 0.425,
+      "step": 14470
+    },
+    {
+      "epoch": 5.4662136655341635,
+      "grad_norm": 0.5597252249717712,
+      "learning_rate": 7.56687858827402e-05,
+      "loss": 0.4192,
+      "step": 14480
+    },
+    {
+      "epoch": 5.469988674971687,
+      "grad_norm": 0.5918856263160706,
+      "learning_rate": 7.563330340966675e-05,
+      "loss": 0.421,
+      "step": 14490
+    },
+    {
+      "epoch": 5.473763684409211,
+      "grad_norm": 0.574697732925415,
+      "learning_rate": 7.559780341647212e-05,
+      "loss": 0.4286,
+      "step": 14500
+    },
+    {
+      "epoch": 5.477538693846735,
+      "grad_norm": 0.7420601844787598,
+      "learning_rate": 7.556228592742026e-05,
+      "loss": 0.4226,
+      "step": 14510
+    },
+    {
+      "epoch": 5.4813137032842585,
+      "grad_norm": 0.8182733058929443,
+      "learning_rate": 7.552675096678696e-05,
+      "loss": 0.4199,
+      "step": 14520
+    },
+    {
+      "epoch": 5.485088712721781,
+      "grad_norm": 0.6243227124214172,
+      "learning_rate": 7.549119855886012e-05,
+      "loss": 0.4234,
+      "step": 14530
+    },
+    {
+      "epoch": 5.488863722159305,
+      "grad_norm": 0.5414767265319824,
+      "learning_rate": 7.545562872793941e-05,
+      "loss": 0.4243,
+      "step": 14540
+    },
+    {
+      "epoch": 5.492638731596829,
+      "grad_norm": 0.6131302714347839,
+      "learning_rate": 7.542004149833648e-05,
+      "loss": 0.4277,
+      "step": 14550
+    },
+    {
+      "epoch": 5.4964137410343525,
+      "grad_norm": 0.6857340335845947,
+      "learning_rate": 7.538443689437492e-05,
+      "loss": 0.4259,
+      "step": 14560
+    },
+    {
+      "epoch": 5.500188750471876,
+      "grad_norm": 0.5744028091430664,
+      "learning_rate": 7.53488149403901e-05,
+      "loss": 0.4208,
+      "step": 14570
+    },
+    {
+      "epoch": 5.5039637599094,
+      "grad_norm": 0.602278470993042,
+      "learning_rate": 7.531317566072929e-05,
+      "loss": 0.4211,
+      "step": 14580
+    },
+    {
+      "epoch": 5.507738769346924,
+      "grad_norm": 0.6976267695426941,
+      "learning_rate": 7.527751907975158e-05,
+      "loss": 0.4207,
+      "step": 14590
+    },
+    {
+      "epoch": 5.511513778784447,
+      "grad_norm": 0.5509871244430542,
+      "learning_rate": 7.524184522182793e-05,
+      "loss": 0.4165,
+      "step": 14600
+    },
+    {
+      "epoch": 5.51528878822197,
+      "grad_norm": 0.6053850650787354,
+      "learning_rate": 7.520615411134112e-05,
+      "loss": 0.4168,
+      "step": 14610
+    },
+    {
+      "epoch": 5.519063797659494,
+      "grad_norm": 0.592635989189148,
+      "learning_rate": 7.517044577268564e-05,
+      "loss": 0.4263,
+      "step": 14620
+    },
+    {
+      "epoch": 5.522838807097018,
+      "grad_norm": 0.6511247158050537,
+      "learning_rate": 7.513472023026782e-05,
+      "loss": 0.4141,
+      "step": 14630
+    },
+    {
+      "epoch": 5.5266138165345415,
+      "grad_norm": 0.5748429298400879,
+      "learning_rate": 7.509897750850572e-05,
+      "loss": 0.4176,
+      "step": 14640
+    },
+    {
+      "epoch": 5.530388825972065,
+      "grad_norm": 0.6025387048721313,
+      "learning_rate": 7.506321763182918e-05,
+      "loss": 0.4144,
+      "step": 14650
+    },
+    {
+      "epoch": 5.534163835409588,
+      "grad_norm": 0.6291730999946594,
+      "learning_rate": 7.50274406246797e-05,
+      "loss": 0.4273,
+      "step": 14660
+    },
+    {
+      "epoch": 5.537938844847112,
+      "grad_norm": 0.8842790722846985,
+      "learning_rate": 7.499164651151056e-05,
+      "loss": 0.4211,
+      "step": 14670
+    },
+    {
+      "epoch": 5.541713854284636,
+      "grad_norm": 0.6468051075935364,
+      "learning_rate": 7.495583531678669e-05,
+      "loss": 0.4204,
+      "step": 14680
+    },
+    {
+      "epoch": 5.545488863722159,
+      "grad_norm": 0.649259090423584,
+      "learning_rate": 7.492000706498469e-05,
+      "loss": 0.4179,
+      "step": 14690
+    },
+    {
+      "epoch": 5.549263873159683,
+      "grad_norm": 0.5695837140083313,
+      "learning_rate": 7.488416178059284e-05,
+      "loss": 0.4255,
+      "step": 14700
+    },
+    {
+      "epoch": 5.553038882597207,
+      "grad_norm": 0.66133713722229,
+      "learning_rate": 7.484829948811107e-05,
+      "loss": 0.4263,
+      "step": 14710
+    },
+    {
+      "epoch": 5.5568138920347305,
+      "grad_norm": 0.6037610173225403,
+      "learning_rate": 7.48124202120509e-05,
+      "loss": 0.4221,
+      "step": 14720
+    },
+    {
+      "epoch": 5.560588901472253,
+      "grad_norm": 0.6156513094902039,
+      "learning_rate": 7.477652397693549e-05,
+      "loss": 0.4173,
+      "step": 14730
+    },
+    {
+      "epoch": 5.564363910909777,
+      "grad_norm": 0.5722167491912842,
+      "learning_rate": 7.474061080729955e-05,
+      "loss": 0.4218,
+      "step": 14740
+    },
+    {
+      "epoch": 5.568138920347301,
+      "grad_norm": 1.3760823011398315,
+      "learning_rate": 7.470468072768941e-05,
+      "loss": 0.4296,
+      "step": 14750
+    },
+    {
+      "epoch": 5.571913929784825,
+      "grad_norm": 0.7008207440376282,
+      "learning_rate": 7.466873376266297e-05,
+      "loss": 0.4214,
+      "step": 14760
+    },
+    {
+      "epoch": 5.575688939222348,
+      "grad_norm": 0.5777148604393005,
+      "learning_rate": 7.46327699367896e-05,
+      "loss": 0.4208,
+      "step": 14770
+    },
+    {
+      "epoch": 5.579463948659871,
+      "grad_norm": 0.6137166023254395,
+      "learning_rate": 7.459678927465026e-05,
+      "loss": 0.4207,
+      "step": 14780
+    },
+    {
+      "epoch": 5.583238958097395,
+      "grad_norm": 0.6016892194747925,
+      "learning_rate": 7.456079180083737e-05,
+      "loss": 0.42,
+      "step": 14790
+    },
+    {
+      "epoch": 5.587013967534919,
+      "grad_norm": 0.6223365068435669,
+      "learning_rate": 7.452477753995489e-05,
+      "loss": 0.4188,
+      "step": 14800
+    },
+    {
+      "epoch": 5.590788976972442,
+      "grad_norm": 0.5763020515441895,
+      "learning_rate": 7.448874651661823e-05,
+      "loss": 0.4192,
+      "step": 14810
+    },
+    {
+      "epoch": 5.594563986409966,
+      "grad_norm": 0.5202614665031433,
+      "learning_rate": 7.445269875545423e-05,
+      "loss": 0.4213,
+      "step": 14820
+    },
+    {
+      "epoch": 5.59833899584749,
+      "grad_norm": 0.6435422301292419,
+      "learning_rate": 7.44166342811012e-05,
+      "loss": 0.4166,
+      "step": 14830
+    },
+    {
+      "epoch": 5.602114005285014,
+      "grad_norm": 0.6565425992012024,
+      "learning_rate": 7.438055311820886e-05,
+      "loss": 0.42,
+      "step": 14840
+    },
+    {
+      "epoch": 5.605889014722536,
+      "grad_norm": 0.5886382460594177,
+      "learning_rate": 7.434445529143837e-05,
+      "loss": 0.4143,
+      "step": 14850
+    },
+    {
+      "epoch": 5.60966402416006,
+      "grad_norm": 0.5876275897026062,
+      "learning_rate": 7.430834082546225e-05,
+      "loss": 0.4178,
+      "step": 14860
+    },
+    {
+      "epoch": 5.613439033597584,
+      "grad_norm": 0.5644301772117615,
+      "learning_rate": 7.427220974496438e-05,
+      "loss": 0.4271,
+      "step": 14870
+    },
+    {
+      "epoch": 5.617214043035108,
+      "grad_norm": 0.6446945667266846,
+      "learning_rate": 7.423606207464005e-05,
+      "loss": 0.4173,
+      "step": 14880
+    },
+    {
+      "epoch": 5.620989052472631,
+      "grad_norm": 0.6406450271606445,
+      "learning_rate": 7.419989783919578e-05,
+      "loss": 0.4194,
+      "step": 14890
+    },
+    {
+      "epoch": 5.624764061910155,
+      "grad_norm": 0.6164763569831848,
+      "learning_rate": 7.416371706334956e-05,
+      "loss": 0.4221,
+      "step": 14900
+    },
+    {
+      "epoch": 5.628539071347678,
+      "grad_norm": 0.6049771904945374,
+      "learning_rate": 7.412751977183056e-05,
+      "loss": 0.4181,
+      "step": 14910
+    },
+    {
+      "epoch": 5.632314080785202,
+      "grad_norm": 0.6096076965332031,
+      "learning_rate": 7.409130598937932e-05,
+      "loss": 0.4198,
+      "step": 14920
+    },
+    {
+      "epoch": 5.636089090222725,
+      "grad_norm": 0.5442625880241394,
+      "learning_rate": 7.40550757407476e-05,
+      "loss": 0.4373,
+      "step": 14930
+    },
+    {
+      "epoch": 5.639864099660249,
+      "grad_norm": 0.6001285910606384,
+      "learning_rate": 7.401882905069843e-05,
+      "loss": 0.4248,
+      "step": 14940
+    },
+    {
+      "epoch": 5.643639109097773,
+      "grad_norm": 0.60331791639328,
+      "learning_rate": 7.39825659440061e-05,
+      "loss": 0.4246,
+      "step": 14950
+    },
+    {
+      "epoch": 5.647414118535297,
+      "grad_norm": 0.5998172760009766,
+      "learning_rate": 7.394628644545609e-05,
+      "loss": 0.4225,
+      "step": 14960
+    },
+    {
+      "epoch": 5.65118912797282,
+      "grad_norm": 0.6505115628242493,
+      "learning_rate": 7.390999057984507e-05,
+      "loss": 0.4245,
+      "step": 14970
+    },
+    {
+      "epoch": 5.654964137410343,
+      "grad_norm": 0.6152286529541016,
+      "learning_rate": 7.387367837198097e-05,
+      "loss": 0.4314,
+      "step": 14980
+    },
+    {
+      "epoch": 5.658739146847867,
+      "grad_norm": 0.595940113067627,
+      "learning_rate": 7.383734984668281e-05,
+      "loss": 0.4192,
+      "step": 14990
+    },
+    {
+      "epoch": 5.662514156285391,
+      "grad_norm": 0.6145548820495605,
+      "learning_rate": 7.38010050287808e-05,
+      "loss": 0.414,
+      "step": 15000
+    },
+    {
+      "epoch": 5.666289165722914,
+      "grad_norm": 1.0223385095596313,
+      "learning_rate": 7.376464394311628e-05,
+      "loss": 0.4289,
+      "step": 15010
+    },
+    {
+      "epoch": 5.670064175160438,
+      "grad_norm": 0.6325558423995972,
+      "learning_rate": 7.372826661454172e-05,
+      "loss": 0.4227,
+      "step": 15020
+    },
+    {
+      "epoch": 5.673839184597962,
+      "grad_norm": 0.6463314294815063,
+      "learning_rate": 7.369187306792068e-05,
+      "loss": 0.4183,
+      "step": 15030
+    },
+    {
+      "epoch": 5.677614194035485,
+      "grad_norm": 0.6034806370735168,
+      "learning_rate": 7.365546332812779e-05,
+      "loss": 0.4133,
+      "step": 15040
+    },
+    {
+      "epoch": 5.6813892034730085,
+      "grad_norm": 0.5699142813682556,
+      "learning_rate": 7.361903742004876e-05,
+      "loss": 0.4185,
+      "step": 15050
+    },
+    {
+      "epoch": 5.685164212910532,
+      "grad_norm": 0.6562982797622681,
+      "learning_rate": 7.358259536858039e-05,
+      "loss": 0.4136,
+      "step": 15060
+    },
+    {
+      "epoch": 5.688939222348056,
+      "grad_norm": 0.691800594329834,
+      "learning_rate": 7.354613719863044e-05,
+      "loss": 0.4263,
+      "step": 15070
+    },
+    {
+      "epoch": 5.69271423178558,
+      "grad_norm": 0.6193991303443909,
+      "learning_rate": 7.350966293511776e-05,
+      "loss": 0.422,
+      "step": 15080
+    },
+    {
+      "epoch": 5.696489241223103,
+      "grad_norm": 0.6109247803688049,
+      "learning_rate": 7.347317260297212e-05,
+      "loss": 0.4272,
+      "step": 15090
+    },
+    {
+      "epoch": 5.700264250660627,
+      "grad_norm": 0.5521445274353027,
+      "learning_rate": 7.343666622713437e-05,
+      "loss": 0.4142,
+      "step": 15100
+    },
+    {
+      "epoch": 5.70403926009815,
+      "grad_norm": 0.5884120464324951,
+      "learning_rate": 7.340014383255624e-05,
+      "loss": 0.4103,
+      "step": 15110
+    },
+    {
+      "epoch": 5.707814269535674,
+      "grad_norm": 0.7118320465087891,
+      "learning_rate": 7.336360544420044e-05,
+      "loss": 0.4234,
+      "step": 15120
+    },
+    {
+      "epoch": 5.7115892789731975,
+      "grad_norm": 0.6134358644485474,
+      "learning_rate": 7.332705108704064e-05,
+      "loss": 0.4077,
+      "step": 15130
+    },
+    {
+      "epoch": 5.715364288410721,
+      "grad_norm": 0.5904216170310974,
+      "learning_rate": 7.329048078606138e-05,
+      "loss": 0.4202,
+      "step": 15140
+    },
+    {
+      "epoch": 5.719139297848245,
+      "grad_norm": 0.6160265803337097,
+      "learning_rate": 7.32538945662581e-05,
+      "loss": 0.4229,
+      "step": 15150
+    },
+    {
+      "epoch": 5.722914307285768,
+      "grad_norm": 0.6015536189079285,
+      "learning_rate": 7.321729245263718e-05,
+      "loss": 0.458,
+      "step": 15160
+    },
+    {
+      "epoch": 5.7266893167232915,
+      "grad_norm": 0.5921651721000671,
+      "learning_rate": 7.318067447021578e-05,
+      "loss": 0.42,
+      "step": 15170
+    },
+    {
+      "epoch": 5.730464326160815,
+      "grad_norm": 0.5869306921958923,
+      "learning_rate": 7.314404064402198e-05,
+      "loss": 0.4119,
+      "step": 15180
+    },
+    {
+      "epoch": 5.734239335598339,
+      "grad_norm": 0.5936274528503418,
+      "learning_rate": 7.310739099909461e-05,
+      "loss": 0.4105,
+      "step": 15190
+    },
+    {
+      "epoch": 5.738014345035863,
+      "grad_norm": 0.6002879738807678,
+      "learning_rate": 7.307072556048339e-05,
+      "loss": 0.4178,
+      "step": 15200
+    },
+    {
+      "epoch": 5.7417893544733865,
+      "grad_norm": 0.6081198453903198,
+      "learning_rate": 7.30340443532488e-05,
+      "loss": 0.4202,
+      "step": 15210
+    },
+    {
+      "epoch": 5.74556436391091,
+      "grad_norm": 0.6480271816253662,
+      "learning_rate": 7.299734740246208e-05,
+      "loss": 0.4105,
+      "step": 15220
+    },
+    {
+      "epoch": 5.749339373348433,
+      "grad_norm": 0.5739256739616394,
+      "learning_rate": 7.296063473320528e-05,
+      "loss": 0.4184,
+      "step": 15230
+    },
+    {
+      "epoch": 5.753114382785957,
+      "grad_norm": 0.749381422996521,
+      "learning_rate": 7.292390637057113e-05,
+      "loss": 0.4158,
+      "step": 15240
+    },
+    {
+      "epoch": 5.7568893922234805,
+      "grad_norm": 0.6204841732978821,
+      "learning_rate": 7.288716233966314e-05,
+      "loss": 0.4133,
+      "step": 15250
+    },
+    {
+      "epoch": 5.760664401661004,
+      "grad_norm": 1.603737235069275,
+      "learning_rate": 7.285040266559551e-05,
+      "loss": 0.4199,
+      "step": 15260
+    },
+    {
+      "epoch": 5.764439411098528,
+      "grad_norm": 0.6340755820274353,
+      "learning_rate": 7.281362737349312e-05,
+      "loss": 0.4139,
+      "step": 15270
+    },
+    {
+      "epoch": 5.768214420536052,
+      "grad_norm": 0.6750447154045105,
+      "learning_rate": 7.277683648849153e-05,
+      "loss": 0.4176,
+      "step": 15280
+    },
+    {
+      "epoch": 5.771989429973575,
+      "grad_norm": 0.5785810351371765,
+      "learning_rate": 7.2740030035737e-05,
+      "loss": 0.4137,
+      "step": 15290
+    },
+    {
+      "epoch": 5.775764439411098,
+      "grad_norm": 0.5631486773490906,
+      "learning_rate": 7.270320804038634e-05,
+      "loss": 0.4183,
+      "step": 15300
+    },
+    {
+      "epoch": 5.779539448848622,
+      "grad_norm": 0.6204792857170105,
+      "learning_rate": 7.266637052760708e-05,
+      "loss": 0.4185,
+      "step": 15310
+    },
+    {
+      "epoch": 5.783314458286146,
+      "grad_norm": 0.6543823480606079,
+      "learning_rate": 7.262951752257728e-05,
+      "loss": 0.4132,
+      "step": 15320
+    },
+    {
+      "epoch": 5.7870894677236695,
+      "grad_norm": 0.6588866114616394,
+      "learning_rate": 7.259264905048564e-05,
+      "loss": 0.4141,
+      "step": 15330
+    },
+    {
+      "epoch": 5.790864477161193,
+      "grad_norm": 0.5547715425491333,
+      "learning_rate": 7.255576513653142e-05,
+      "loss": 0.4132,
+      "step": 15340
+    },
+    {
+      "epoch": 5.794639486598717,
+      "grad_norm": 0.642467200756073,
+      "learning_rate": 7.251886580592439e-05,
+      "loss": 0.4144,
+      "step": 15350
+    },
+    {
+      "epoch": 5.79841449603624,
+      "grad_norm": 0.6495168209075928,
+      "learning_rate": 7.248195108388496e-05,
+      "loss": 0.4233,
+      "step": 15360
+    },
+    {
+      "epoch": 5.8021895054737636,
+      "grad_norm": 0.6559742093086243,
+      "learning_rate": 7.244502099564395e-05,
+      "loss": 0.415,
+      "step": 15370
+    },
+    {
+      "epoch": 5.805964514911287,
+      "grad_norm": 0.6655861139297485,
+      "learning_rate": 7.240807556644271e-05,
+      "loss": 0.4143,
+      "step": 15380
+    },
+    {
+      "epoch": 5.809739524348811,
+      "grad_norm": 0.5971380472183228,
+      "learning_rate": 7.237111482153314e-05,
+      "loss": 0.4144,
+      "step": 15390
+    },
+    {
+      "epoch": 5.813514533786335,
+      "grad_norm": 0.6410373449325562,
+      "learning_rate": 7.233413878617751e-05,
+      "loss": 0.4212,
+      "step": 15400
+    },
+    {
+      "epoch": 5.817289543223858,
+      "grad_norm": 0.5639101266860962,
+      "learning_rate": 7.229714748564864e-05,
+      "loss": 0.4118,
+      "step": 15410
+    },
+    {
+      "epoch": 5.821064552661381,
+      "grad_norm": 0.631568193435669,
+      "learning_rate": 7.22601409452297e-05,
+      "loss": 0.4152,
+      "step": 15420
+    },
+    {
+      "epoch": 5.824839562098905,
+      "grad_norm": 0.6407223343849182,
+      "learning_rate": 7.222311919021433e-05,
+      "loss": 0.4209,
+      "step": 15430
+    },
+    {
+      "epoch": 5.828614571536429,
+      "grad_norm": 0.8518263101577759,
+      "learning_rate": 7.218608224590655e-05,
+      "loss": 0.4134,
+      "step": 15440
+    },
+    {
+      "epoch": 5.8323895809739525,
+      "grad_norm": 0.6463886499404907,
+      "learning_rate": 7.214903013762074e-05,
+      "loss": 0.4219,
+      "step": 15450
+    },
+    {
+      "epoch": 5.836164590411476,
+      "grad_norm": 0.6677048206329346,
+      "learning_rate": 7.21119628906817e-05,
+      "loss": 0.4104,
+      "step": 15460
+    },
+    {
+      "epoch": 5.839939599849,
+      "grad_norm": 0.6562634706497192,
+      "learning_rate": 7.207488053042454e-05,
+      "loss": 0.4146,
+      "step": 15470
+    },
+    {
+      "epoch": 5.843714609286523,
+      "grad_norm": 0.6001513600349426,
+      "learning_rate": 7.203778308219467e-05,
+      "loss": 0.4196,
+      "step": 15480
+    },
+    {
+      "epoch": 5.847489618724047,
+      "grad_norm": 0.571530818939209,
+      "learning_rate": 7.200067057134787e-05,
+      "loss": 0.4142,
+      "step": 15490
+    },
+    {
+      "epoch": 5.85126462816157,
+      "grad_norm": 0.6734705567359924,
+      "learning_rate": 7.196354302325019e-05,
+      "loss": 0.4137,
+      "step": 15500
+    },
+    {
+      "epoch": 5.855039637599094,
+      "grad_norm": 0.7626957297325134,
+      "learning_rate": 7.192640046327795e-05,
+      "loss": 0.4193,
+      "step": 15510
+    },
+    {
+      "epoch": 5.858814647036618,
+      "grad_norm": 0.589971125125885,
+      "learning_rate": 7.188924291681777e-05,
+      "loss": 0.4144,
+      "step": 15520
+    },
+    {
+      "epoch": 5.8625896564741415,
+      "grad_norm": 0.6109906435012817,
+      "learning_rate": 7.185207040926643e-05,
+      "loss": 0.4154,
+      "step": 15530
+    },
+    {
+      "epoch": 5.866364665911664,
+      "grad_norm": 0.6139832139015198,
+      "learning_rate": 7.181488296603103e-05,
+      "loss": 0.415,
+      "step": 15540
+    },
+    {
+      "epoch": 5.870139675349188,
+      "grad_norm": 0.5642750263214111,
+      "learning_rate": 7.177768061252885e-05,
+      "loss": 0.4155,
+      "step": 15550
+    },
+    {
+      "epoch": 5.873914684786712,
+      "grad_norm": 0.6058228015899658,
+      "learning_rate": 7.174046337418729e-05,
+      "loss": 0.4193,
+      "step": 15560
+    },
+    {
+      "epoch": 5.877689694224236,
+      "grad_norm": 0.6035988926887512,
+      "learning_rate": 7.170323127644403e-05,
+      "loss": 0.4152,
+      "step": 15570
+    },
+    {
+      "epoch": 5.881464703661759,
+      "grad_norm": 0.5951765179634094,
+      "learning_rate": 7.166598434474683e-05,
+      "loss": 0.4125,
+      "step": 15580
+    },
+    {
+      "epoch": 5.885239713099283,
+      "grad_norm": 0.6099646091461182,
+      "learning_rate": 7.162872260455364e-05,
+      "loss": 0.4121,
+      "step": 15590
+    },
+    {
+      "epoch": 5.889014722536807,
+      "grad_norm": 0.6386427879333496,
+      "learning_rate": 7.159144608133248e-05,
+      "loss": 0.4079,
+      "step": 15600
+    },
+    {
+      "epoch": 5.89278973197433,
+      "grad_norm": 0.6007016897201538,
+      "learning_rate": 7.155415480056153e-05,
+      "loss": 0.411,
+      "step": 15610
+    },
+    {
+      "epoch": 5.896564741411853,
+      "grad_norm": 0.6118659973144531,
+      "learning_rate": 7.151684878772902e-05,
+      "loss": 0.4271,
+      "step": 15620
+    },
+    {
+      "epoch": 5.900339750849377,
+      "grad_norm": 0.6127933859825134,
+      "learning_rate": 7.147952806833324e-05,
+      "loss": 0.4097,
+      "step": 15630
+    },
+    {
+      "epoch": 5.904114760286901,
+      "grad_norm": 0.6609582304954529,
+      "learning_rate": 7.14421926678826e-05,
+      "loss": 0.4102,
+      "step": 15640
+    },
+    {
+      "epoch": 5.907889769724425,
+      "grad_norm": 0.6080712080001831,
+      "learning_rate": 7.140484261189543e-05,
+      "loss": 0.4141,
+      "step": 15650
+    },
+    {
+      "epoch": 5.911664779161947,
+      "grad_norm": 0.621368408203125,
+      "learning_rate": 7.136747792590017e-05,
+      "loss": 0.4098,
+      "step": 15660
+    },
+    {
+      "epoch": 5.915439788599471,
+      "grad_norm": 0.6989858746528625,
+      "learning_rate": 7.133009863543524e-05,
+      "loss": 0.4273,
+      "step": 15670
+    },
+    {
+      "epoch": 5.919214798036995,
+      "grad_norm": 0.5902137160301208,
+      "learning_rate": 7.129270476604901e-05,
+      "loss": 0.4121,
+      "step": 15680
+    },
+    {
+      "epoch": 5.922989807474519,
+      "grad_norm": 0.6309996843338013,
+      "learning_rate": 7.125529634329988e-05,
+      "loss": 0.4136,
+      "step": 15690
+    },
+    {
+      "epoch": 5.926764816912042,
+      "grad_norm": 0.6389601230621338,
+      "learning_rate": 7.12178733927561e-05,
+      "loss": 0.4177,
+      "step": 15700
+    },
+    {
+      "epoch": 5.930539826349566,
+      "grad_norm": 0.5543851852416992,
+      "learning_rate": 7.118043593999593e-05,
+      "loss": 0.4172,
+      "step": 15710
+    },
+    {
+      "epoch": 5.93431483578709,
+      "grad_norm": 0.5611990094184875,
+      "learning_rate": 7.114298401060752e-05,
+      "loss": 0.4153,
+      "step": 15720
+    },
+    {
+      "epoch": 5.938089845224613,
+      "grad_norm": 0.6465316414833069,
+      "learning_rate": 7.11055176301889e-05,
+      "loss": 0.4102,
+      "step": 15730
+    },
+    {
+      "epoch": 5.941864854662136,
+      "grad_norm": 0.6097255349159241,
+      "learning_rate": 7.1068036824348e-05,
+      "loss": 0.4169,
+      "step": 15740
+    },
+    {
+      "epoch": 5.94563986409966,
+      "grad_norm": 0.5505424737930298,
+      "learning_rate": 7.10305416187026e-05,
+      "loss": 0.4134,
+      "step": 15750
+    },
+    {
+      "epoch": 5.949414873537184,
+      "grad_norm": 0.554263710975647,
+      "learning_rate": 7.099303203888029e-05,
+      "loss": 0.4127,
+      "step": 15760
+    },
+    {
+      "epoch": 5.953189882974708,
+      "grad_norm": 0.6733128428459167,
+      "learning_rate": 7.095550811051855e-05,
+      "loss": 0.4059,
+      "step": 15770
+    },
+    {
+      "epoch": 5.956964892412231,
+      "grad_norm": 0.6691094040870667,
+      "learning_rate": 7.09179698592646e-05,
+      "loss": 0.416,
+      "step": 15780
+    },
+    {
+      "epoch": 5.960739901849754,
+      "grad_norm": 0.581087052822113,
+      "learning_rate": 7.088041731077551e-05,
+      "loss": 0.4081,
+      "step": 15790
+    },
+    {
+      "epoch": 5.964514911287278,
+      "grad_norm": 0.6719875931739807,
+      "learning_rate": 7.084285049071806e-05,
+      "loss": 0.4086,
+      "step": 15800
+    },
+    {
+      "epoch": 5.968289920724802,
+      "grad_norm": 0.6516507267951965,
+      "learning_rate": 7.080526942476886e-05,
+      "loss": 0.4166,
+      "step": 15810
+    },
+    {
+      "epoch": 5.972064930162325,
+      "grad_norm": 0.6685953140258789,
+      "learning_rate": 7.076767413861418e-05,
+      "loss": 0.4253,
+      "step": 15820
+    },
+    {
+      "epoch": 5.975839939599849,
+      "grad_norm": 0.6204284429550171,
+      "learning_rate": 7.073006465795005e-05,
+      "loss": 0.4134,
+      "step": 15830
+    },
+    {
+      "epoch": 5.979614949037373,
+      "grad_norm": 0.5573307275772095,
+      "learning_rate": 7.06924410084822e-05,
+      "loss": 0.4112,
+      "step": 15840
+    },
+    {
+      "epoch": 5.983389958474897,
+      "grad_norm": 0.5689889192581177,
+      "learning_rate": 7.065480321592604e-05,
+      "loss": 0.4107,
+      "step": 15850
+    },
+    {
+      "epoch": 5.9871649679124195,
+      "grad_norm": 0.6300421953201294,
+      "learning_rate": 7.061715130600663e-05,
+      "loss": 0.4019,
+      "step": 15860
+    },
+    {
+      "epoch": 5.990939977349943,
+      "grad_norm": 0.5939746499061584,
+      "learning_rate": 7.057948530445873e-05,
+      "loss": 0.4093,
+      "step": 15870
+    },
+    {
+      "epoch": 5.994714986787467,
+      "grad_norm": 0.576476514339447,
+      "learning_rate": 7.054180523702668e-05,
+      "loss": 0.4084,
+      "step": 15880
+    },
+    {
+      "epoch": 5.998489996224991,
+      "grad_norm": 0.6696039438247681,
+      "learning_rate": 7.050411112946442e-05,
+      "loss": 0.412,
+      "step": 15890
+    },
+    {
+      "epoch": 6.002265005662514,
+      "grad_norm": 0.6292182803153992,
+      "learning_rate": 7.046640300753557e-05,
+      "loss": 0.4151,
+      "step": 15900
+    },
+    {
+      "epoch": 6.006040015100038,
+      "grad_norm": 0.5818023085594177,
+      "learning_rate": 7.042868089701325e-05,
+      "loss": 0.409,
+      "step": 15910
+    },
+    {
+      "epoch": 6.009815024537561,
+      "grad_norm": 0.601872444152832,
+      "learning_rate": 7.039094482368016e-05,
+      "loss": 0.4074,
+      "step": 15920
+    },
+    {
+      "epoch": 6.013590033975085,
+      "grad_norm": 0.629568874835968,
+      "learning_rate": 7.035319481332858e-05,
+      "loss": 0.4241,
+      "step": 15930
+    },
+    {
+      "epoch": 6.0173650434126085,
+      "grad_norm": 0.6180824637413025,
+      "learning_rate": 7.031543089176023e-05,
+      "loss": 0.413,
+      "step": 15940
+    },
+    {
+      "epoch": 6.021140052850132,
+      "grad_norm": 0.7090234160423279,
+      "learning_rate": 7.027765308478644e-05,
+      "loss": 0.4159,
+      "step": 15950
+    },
+    {
+      "epoch": 6.024915062287656,
+      "grad_norm": 0.8287419676780701,
+      "learning_rate": 7.023986141822798e-05,
+      "loss": 0.4117,
+      "step": 15960
+    },
+    {
+      "epoch": 6.02869007172518,
+      "grad_norm": 0.6538956761360168,
+      "learning_rate": 7.02020559179151e-05,
+      "loss": 0.4134,
+      "step": 15970
+    },
+    {
+      "epoch": 6.0324650811627025,
+      "grad_norm": 0.6877662539482117,
+      "learning_rate": 7.016423660968748e-05,
+      "loss": 0.4132,
+      "step": 15980
+    },
+    {
+      "epoch": 6.036240090600226,
+      "grad_norm": 0.9112296104431152,
+      "learning_rate": 7.012640351939428e-05,
+      "loss": 0.4225,
+      "step": 15990
+    },
+    {
+      "epoch": 6.04001510003775,
+      "grad_norm": 0.6331153512001038,
+      "learning_rate": 7.008855667289404e-05,
+      "loss": 0.4109,
+      "step": 16000
+    },
+    {
+      "epoch": 6.043790109475274,
+      "grad_norm": 0.6433661580085754,
+      "learning_rate": 7.005069609605476e-05,
+      "loss": 0.4129,
+      "step": 16010
+    },
+    {
+      "epoch": 6.0475651189127975,
+      "grad_norm": 0.5703040361404419,
+      "learning_rate": 7.001282181475377e-05,
+      "loss": 0.422,
+      "step": 16020
+    },
+    {
+      "epoch": 6.051340128350321,
+      "grad_norm": 0.5761269330978394,
+      "learning_rate": 6.997493385487775e-05,
+      "loss": 0.4079,
+      "step": 16030
+    },
+    {
+      "epoch": 6.055115137787844,
+      "grad_norm": 0.6320202350616455,
+      "learning_rate": 6.99370322423228e-05,
+      "loss": 0.405,
+      "step": 16040
+    },
+    {
+      "epoch": 6.058890147225368,
+      "grad_norm": 0.5953516960144043,
+      "learning_rate": 6.989911700299433e-05,
+      "loss": 0.4178,
+      "step": 16050
+    },
+    {
+      "epoch": 6.0626651566628915,
+      "grad_norm": 0.6841408014297485,
+      "learning_rate": 6.9861188162807e-05,
+      "loss": 0.4176,
+      "step": 16060
+    },
+    {
+      "epoch": 6.066440166100415,
+      "grad_norm": 0.6144746541976929,
+      "learning_rate": 6.982324574768487e-05,
+      "loss": 0.4184,
+      "step": 16070
+    },
+    {
+      "epoch": 6.070215175537939,
+      "grad_norm": 0.6368954181671143,
+      "learning_rate": 6.978528978356117e-05,
+      "loss": 0.4169,
+      "step": 16080
+    },
+    {
+      "epoch": 6.073990184975463,
+      "grad_norm": 0.6184800863265991,
+      "learning_rate": 6.974732029637846e-05,
+      "loss": 0.4123,
+      "step": 16090
+    },
+    {
+      "epoch": 6.0777651944129865,
+      "grad_norm": 0.6910690665245056,
+      "learning_rate": 6.970933731208855e-05,
+      "loss": 0.413,
+      "step": 16100
+    },
+    {
+      "epoch": 6.081540203850509,
+      "grad_norm": 0.5969312787055969,
+      "learning_rate": 6.967134085665244e-05,
+      "loss": 0.4145,
+      "step": 16110
+    },
+    {
+      "epoch": 6.085315213288033,
+      "grad_norm": 0.5405552983283997,
+      "learning_rate": 6.963333095604034e-05,
+      "loss": 0.4136,
+      "step": 16120
+    },
+    {
+      "epoch": 6.089090222725557,
+      "grad_norm": 0.5933461785316467,
+      "learning_rate": 6.959530763623166e-05,
+      "loss": 0.4144,
+      "step": 16130
+    },
+    {
+      "epoch": 6.0928652321630805,
+      "grad_norm": 0.6700628995895386,
+      "learning_rate": 6.955727092321497e-05,
+      "loss": 0.4092,
+      "step": 16140
+    },
+    {
+      "epoch": 6.096640241600604,
+      "grad_norm": 0.6101819276809692,
+      "learning_rate": 6.951922084298803e-05,
+      "loss": 0.418,
+      "step": 16150
+    },
+    {
+      "epoch": 6.100415251038128,
+      "grad_norm": 0.5824779272079468,
+      "learning_rate": 6.948115742155769e-05,
+      "loss": 0.4169,
+      "step": 16160
+    },
+    {
+      "epoch": 6.104190260475651,
+      "grad_norm": 0.5644822716712952,
+      "learning_rate": 6.944308068493996e-05,
+      "loss": 0.42,
+      "step": 16170
+    },
+    {
+      "epoch": 6.107965269913175,
+      "grad_norm": 0.6036456823348999,
+      "learning_rate": 6.940499065915992e-05,
+      "loss": 0.4083,
+      "step": 16180
+    },
+    {
+      "epoch": 6.111740279350698,
+      "grad_norm": 0.5669860243797302,
+      "learning_rate": 6.936688737025173e-05,
+      "loss": 0.4195,
+      "step": 16190
+    },
+    {
+      "epoch": 6.115515288788222,
+      "grad_norm": 0.5517399311065674,
+      "learning_rate": 6.932877084425867e-05,
+      "loss": 0.414,
+      "step": 16200
+    },
+    {
+      "epoch": 6.119290298225746,
+      "grad_norm": 0.5780723094940186,
+      "learning_rate": 6.929064110723297e-05,
+      "loss": 0.4133,
+      "step": 16210
+    },
+    {
+      "epoch": 6.1230653076632695,
+      "grad_norm": 0.5744929909706116,
+      "learning_rate": 6.925249818523598e-05,
+      "loss": 0.4055,
+      "step": 16220
+    },
+    {
+      "epoch": 6.126840317100792,
+      "grad_norm": 0.5747778415679932,
+      "learning_rate": 6.921434210433801e-05,
+      "loss": 0.4068,
+      "step": 16230
+    },
+    {
+      "epoch": 6.130615326538316,
+      "grad_norm": 0.6638509631156921,
+      "learning_rate": 6.917617289061841e-05,
+      "loss": 0.4097,
+      "step": 16240
+    },
+    {
+      "epoch": 6.13439033597584,
+      "grad_norm": 0.6439339518547058,
+      "learning_rate": 6.913799057016547e-05,
+      "loss": 0.4181,
+      "step": 16250
+    },
+    {
+      "epoch": 6.138165345413364,
+      "grad_norm": 1.5434315204620361,
+      "learning_rate": 6.909979516907641e-05,
+      "loss": 0.4165,
+      "step": 16260
+    },
+    {
+      "epoch": 6.141940354850887,
+      "grad_norm": 0.7095974683761597,
+      "learning_rate": 6.906158671345746e-05,
+      "loss": 0.4082,
+      "step": 16270
+    },
+    {
+      "epoch": 6.145715364288411,
+      "grad_norm": 0.5394503474235535,
+      "learning_rate": 6.902336522942374e-05,
+      "loss": 0.4141,
+      "step": 16280
+    },
+    {
+      "epoch": 6.149490373725934,
+      "grad_norm": 1.054152011871338,
+      "learning_rate": 6.898513074309924e-05,
+      "loss": 0.4109,
+      "step": 16290
+    },
+    {
+      "epoch": 6.153265383163458,
+      "grad_norm": 0.6516214609146118,
+      "learning_rate": 6.894688328061693e-05,
+      "loss": 0.4187,
+      "step": 16300
+    },
+    {
+      "epoch": 6.157040392600981,
+      "grad_norm": 0.5728663802146912,
+      "learning_rate": 6.890862286811853e-05,
+      "loss": 0.4231,
+      "step": 16310
+    },
+    {
+      "epoch": 6.160815402038505,
+      "grad_norm": 0.6336321830749512,
+      "learning_rate": 6.88703495317547e-05,
+      "loss": 0.4066,
+      "step": 16320
+    },
+    {
+      "epoch": 6.164590411476029,
+      "grad_norm": 0.6610395908355713,
+      "learning_rate": 6.883206329768492e-05,
+      "loss": 0.4141,
+      "step": 16330
+    },
+    {
+      "epoch": 6.1683654209135526,
+      "grad_norm": 0.6659107208251953,
+      "learning_rate": 6.879376419207743e-05,
+      "loss": 0.4055,
+      "step": 16340
+    },
+    {
+      "epoch": 6.172140430351076,
+      "grad_norm": 0.6790131330490112,
+      "learning_rate": 6.875545224110935e-05,
+      "loss": 0.4178,
+      "step": 16350
+    },
+    {
+      "epoch": 6.175915439788599,
+      "grad_norm": 0.5676288604736328,
+      "learning_rate": 6.871712747096651e-05,
+      "loss": 0.4282,
+      "step": 16360
+    },
+    {
+      "epoch": 6.179690449226123,
+      "grad_norm": 0.6402314901351929,
+      "learning_rate": 6.867878990784353e-05,
+      "loss": 0.4146,
+      "step": 16370
+    },
+    {
+      "epoch": 6.183465458663647,
+      "grad_norm": 0.6174314022064209,
+      "learning_rate": 6.864043957794377e-05,
+      "loss": 0.414,
+      "step": 16380
+    },
+    {
+      "epoch": 6.18724046810117,
+      "grad_norm": 0.6669917702674866,
+      "learning_rate": 6.860207650747934e-05,
+      "loss": 0.4154,
+      "step": 16390
+    },
+    {
+      "epoch": 6.191015477538694,
+      "grad_norm": 0.6218419671058655,
+      "learning_rate": 6.856370072267104e-05,
+      "loss": 0.4125,
+      "step": 16400
+    },
+    {
+      "epoch": 6.194790486976218,
+      "grad_norm": 0.5821741223335266,
+      "learning_rate": 6.852531224974831e-05,
+      "loss": 0.4081,
+      "step": 16410
+    },
+    {
+      "epoch": 6.198565496413741,
+      "grad_norm": 0.6421241164207458,
+      "learning_rate": 6.848691111494936e-05,
+      "loss": 0.4067,
+      "step": 16420
+    },
+    {
+      "epoch": 6.202340505851264,
+      "grad_norm": 0.7413673996925354,
+      "learning_rate": 6.844849734452097e-05,
+      "loss": 0.4059,
+      "step": 16430
+    },
+    {
+      "epoch": 6.206115515288788,
+      "grad_norm": 0.715653657913208,
+      "learning_rate": 6.841007096471862e-05,
+      "loss": 0.414,
+      "step": 16440
+    },
+    {
+      "epoch": 6.209890524726312,
+      "grad_norm": 0.9158740639686584,
+      "learning_rate": 6.837163200180636e-05,
+      "loss": 0.4127,
+      "step": 16450
+    },
+    {
+      "epoch": 6.213665534163836,
+      "grad_norm": 0.6491902470588684,
+      "learning_rate": 6.833318048205684e-05,
+      "loss": 0.4046,
+      "step": 16460
+    },
+    {
+      "epoch": 6.217440543601359,
+      "grad_norm": 0.5672116875648499,
+      "learning_rate": 6.829471643175136e-05,
+      "loss": 0.4076,
+      "step": 16470
+    },
+    {
+      "epoch": 6.221215553038882,
+      "grad_norm": 0.7057968378067017,
+      "learning_rate": 6.825623987717969e-05,
+      "loss": 0.4077,
+      "step": 16480
+    },
+    {
+      "epoch": 6.224990562476406,
+      "grad_norm": 0.6522260308265686,
+      "learning_rate": 6.821775084464022e-05,
+      "loss": 0.4096,
+      "step": 16490
+    },
+    {
+      "epoch": 6.22876557191393,
+      "grad_norm": 0.6316922903060913,
+      "learning_rate": 6.817924936043982e-05,
+      "loss": 0.4026,
+      "step": 16500
+    },
+    {
+      "epoch": 6.232540581351453,
+      "grad_norm": 0.6496508717536926,
+      "learning_rate": 6.81407354508939e-05,
+      "loss": 0.4087,
+      "step": 16510
+    },
+    {
+      "epoch": 6.236315590788977,
+      "grad_norm": 0.5875167846679688,
+      "learning_rate": 6.810220914232636e-05,
+      "loss": 0.4075,
+      "step": 16520
+    },
+    {
+      "epoch": 6.240090600226501,
+      "grad_norm": 0.6535125970840454,
+      "learning_rate": 6.806367046106959e-05,
+      "loss": 0.4127,
+      "step": 16530
+    },
+    {
+      "epoch": 6.243865609664024,
+      "grad_norm": 0.6569979786872864,
+      "learning_rate": 6.802511943346435e-05,
+      "loss": 0.4154,
+      "step": 16540
+    },
+    {
+      "epoch": 6.247640619101547,
+      "grad_norm": 0.6199260354042053,
+      "learning_rate": 6.798655608585997e-05,
+      "loss": 0.4116,
+      "step": 16550
+    },
+    {
+      "epoch": 6.251415628539071,
+      "grad_norm": 0.6537362337112427,
+      "learning_rate": 6.79479804446141e-05,
+      "loss": 0.4057,
+      "step": 16560
+    },
+    {
+      "epoch": 6.255190637976595,
+      "grad_norm": 0.6324570775032043,
+      "learning_rate": 6.790939253609284e-05,
+      "loss": 0.4028,
+      "step": 16570
+    },
+    {
+      "epoch": 6.258965647414119,
+      "grad_norm": 0.6022115349769592,
+      "learning_rate": 6.787079238667065e-05,
+      "loss": 0.4087,
+      "step": 16580
+    },
+    {
+      "epoch": 6.262740656851642,
+      "grad_norm": 0.6702650189399719,
+      "learning_rate": 6.783218002273039e-05,
+      "loss": 0.4071,
+      "step": 16590
+    },
+    {
+      "epoch": 6.266515666289166,
+      "grad_norm": 0.6910784244537354,
+      "learning_rate": 6.779355547066322e-05,
+      "loss": 0.4038,
+      "step": 16600
+    },
+    {
+      "epoch": 6.270290675726689,
+      "grad_norm": 0.6253206729888916,
+      "learning_rate": 6.775491875686865e-05,
+      "loss": 0.4126,
+      "step": 16610
+    },
+    {
+      "epoch": 6.274065685164213,
+      "grad_norm": 0.5710755586624146,
+      "learning_rate": 6.771626990775457e-05,
+      "loss": 0.4108,
+      "step": 16620
+    },
+    {
+      "epoch": 6.277840694601736,
+      "grad_norm": 0.6698145866394043,
+      "learning_rate": 6.767760894973704e-05,
+      "loss": 0.4424,
+      "step": 16630
+    },
+    {
+      "epoch": 6.28161570403926,
+      "grad_norm": 1.365607500076294,
+      "learning_rate": 6.763893590924048e-05,
+      "loss": 0.4173,
+      "step": 16640
+    },
+    {
+      "epoch": 6.285390713476784,
+      "grad_norm": 0.6131840944290161,
+      "learning_rate": 6.760025081269756e-05,
+      "loss": 0.4056,
+      "step": 16650
+    },
+    {
+      "epoch": 6.289165722914308,
+      "grad_norm": 0.5999181866645813,
+      "learning_rate": 6.756155368654915e-05,
+      "loss": 0.4133,
+      "step": 16660
+    },
+    {
+      "epoch": 6.2929407323518305,
+      "grad_norm": 0.6875386238098145,
+      "learning_rate": 6.752284455724442e-05,
+      "loss": 0.4055,
+      "step": 16670
+    },
+    {
+      "epoch": 6.296715741789354,
+      "grad_norm": 0.5545741319656372,
+      "learning_rate": 6.748412345124065e-05,
+      "loss": 0.4132,
+      "step": 16680
+    },
+    {
+      "epoch": 6.300490751226878,
+      "grad_norm": 0.623226523399353,
+      "learning_rate": 6.744539039500335e-05,
+      "loss": 0.4087,
+      "step": 16690
+    },
+    {
+      "epoch": 6.304265760664402,
+      "grad_norm": 0.6600329279899597,
+      "learning_rate": 6.740664541500625e-05,
+      "loss": 0.4003,
+      "step": 16700
+    },
+    {
+      "epoch": 6.308040770101925,
+      "grad_norm": 0.5790383219718933,
+      "learning_rate": 6.736788853773112e-05,
+      "loss": 0.4076,
+      "step": 16710
+    },
+    {
+      "epoch": 6.311815779539449,
+      "grad_norm": 0.6672446131706238,
+      "learning_rate": 6.732911978966796e-05,
+      "loss": 0.4089,
+      "step": 16720
+    },
+    {
+      "epoch": 6.315590788976973,
+      "grad_norm": 0.5762278437614441,
+      "learning_rate": 6.729033919731482e-05,
+      "loss": 0.4116,
+      "step": 16730
+    },
+    {
+      "epoch": 6.319365798414496,
+      "grad_norm": 0.5791032314300537,
+      "learning_rate": 6.725154678717787e-05,
+      "loss": 0.407,
+      "step": 16740
+    },
+    {
+      "epoch": 6.3231408078520195,
+      "grad_norm": 0.6549326777458191,
+      "learning_rate": 6.721274258577138e-05,
+      "loss": 0.4114,
+      "step": 16750
+    },
+    {
+      "epoch": 6.326915817289543,
+      "grad_norm": 0.6120374202728271,
+      "learning_rate": 6.717392661961763e-05,
+      "loss": 0.4089,
+      "step": 16760
+    },
+    {
+      "epoch": 6.330690826727067,
+      "grad_norm": 0.6165977716445923,
+      "learning_rate": 6.713509891524697e-05,
+      "loss": 0.4128,
+      "step": 16770
+    },
+    {
+      "epoch": 6.334465836164591,
+      "grad_norm": 0.6072388887405396,
+      "learning_rate": 6.709625949919777e-05,
+      "loss": 0.4091,
+      "step": 16780
+    },
+    {
+      "epoch": 6.3382408456021135,
+      "grad_norm": 0.6050112843513489,
+      "learning_rate": 6.705740839801642e-05,
+      "loss": 0.4031,
+      "step": 16790
+    },
+    {
+      "epoch": 6.342015855039637,
+      "grad_norm": 0.7295091152191162,
+      "learning_rate": 6.701854563825727e-05,
+      "loss": 0.4105,
+      "step": 16800
+    },
+    {
+      "epoch": 6.345790864477161,
+      "grad_norm": 0.6450115442276001,
+      "learning_rate": 6.697967124648266e-05,
+      "loss": 0.4026,
+      "step": 16810
+    },
+    {
+      "epoch": 6.349565873914685,
+      "grad_norm": 0.6749234199523926,
+      "learning_rate": 6.694078524926285e-05,
+      "loss": 0.4106,
+      "step": 16820
+    },
+    {
+      "epoch": 6.3533408833522085,
+      "grad_norm": 0.6108365058898926,
+      "learning_rate": 6.690188767317607e-05,
+      "loss": 0.403,
+      "step": 16830
+    },
+    {
+      "epoch": 6.357115892789732,
+      "grad_norm": 0.6435592174530029,
+      "learning_rate": 6.686297854480843e-05,
+      "loss": 0.4091,
+      "step": 16840
+    },
+    {
+      "epoch": 6.360890902227256,
+      "grad_norm": 0.5970647931098938,
+      "learning_rate": 6.682405789075398e-05,
+      "loss": 0.404,
+      "step": 16850
+    },
+    {
+      "epoch": 6.364665911664779,
+      "grad_norm": 0.5717617869377136,
+      "learning_rate": 6.67851257376146e-05,
+      "loss": 0.3988,
+      "step": 16860
+    },
+    {
+      "epoch": 6.3684409211023025,
+      "grad_norm": 0.6971402168273926,
+      "learning_rate": 6.674618211200004e-05,
+      "loss": 0.4016,
+      "step": 16870
+    },
+    {
+      "epoch": 6.372215930539826,
+      "grad_norm": 0.5539186000823975,
+      "learning_rate": 6.670722704052792e-05,
+      "loss": 0.4109,
+      "step": 16880
+    },
+    {
+      "epoch": 6.37599093997735,
+      "grad_norm": 0.5876161456108093,
+      "learning_rate": 6.666826054982365e-05,
+      "loss": 0.406,
+      "step": 16890
+    },
+    {
+      "epoch": 6.379765949414874,
+      "grad_norm": 0.6357724666595459,
+      "learning_rate": 6.662928266652048e-05,
+      "loss": 0.4059,
+      "step": 16900
+    },
+    {
+      "epoch": 6.3835409588523975,
+      "grad_norm": 0.7432243227958679,
+      "learning_rate": 6.659029341725941e-05,
+      "loss": 0.4126,
+      "step": 16910
+    },
+    {
+      "epoch": 6.38731596828992,
+      "grad_norm": 0.5893529057502747,
+      "learning_rate": 6.655129282868923e-05,
+      "loss": 0.4053,
+      "step": 16920
+    },
+    {
+      "epoch": 6.391090977727444,
+      "grad_norm": 0.5809387564659119,
+      "learning_rate": 6.651228092746646e-05,
+      "loss": 0.41,
+      "step": 16930
+    },
+    {
+      "epoch": 6.394865987164968,
+      "grad_norm": 0.6010847687721252,
+      "learning_rate": 6.647325774025539e-05,
+      "loss": 0.42,
+      "step": 16940
+    },
+    {
+      "epoch": 6.3986409966024915,
+      "grad_norm": 0.6078194975852966,
+      "learning_rate": 6.643422329372798e-05,
+      "loss": 0.4069,
+      "step": 16950
+    },
+    {
+      "epoch": 6.402416006040015,
+      "grad_norm": 0.637675940990448,
+      "learning_rate": 6.639517761456392e-05,
+      "loss": 0.407,
+      "step": 16960
+    },
+    {
+      "epoch": 6.406191015477539,
+      "grad_norm": 0.5883272886276245,
+      "learning_rate": 6.635612072945054e-05,
+      "loss": 0.4043,
+      "step": 16970
+    },
+    {
+      "epoch": 6.409966024915063,
+      "grad_norm": 0.5617126226425171,
+      "learning_rate": 6.631705266508289e-05,
+      "loss": 0.4079,
+      "step": 16980
+    },
+    {
+      "epoch": 6.413741034352586,
+      "grad_norm": 0.6175063252449036,
+      "learning_rate": 6.62779734481636e-05,
+      "loss": 0.4003,
+      "step": 16990
+    },
+    {
+      "epoch": 6.417516043790109,
+      "grad_norm": 0.6425361037254333,
+      "learning_rate": 6.623888310540294e-05,
+      "loss": 0.4033,
+      "step": 17000
+    },
+    {
+      "epoch": 6.421291053227633,
+      "grad_norm": 0.5959002375602722,
+      "learning_rate": 6.619978166351882e-05,
+      "loss": 0.4037,
+      "step": 17010
+    },
+    {
+      "epoch": 6.425066062665157,
+      "grad_norm": 0.6390359997749329,
+      "learning_rate": 6.616066914923666e-05,
+      "loss": 0.4058,
+      "step": 17020
+    },
+    {
+      "epoch": 6.4288410721026805,
+      "grad_norm": 0.5752782225608826,
+      "learning_rate": 6.612154558928955e-05,
+      "loss": 0.4081,
+      "step": 17030
+    },
+    {
+      "epoch": 6.432616081540204,
+      "grad_norm": 0.684485137462616,
+      "learning_rate": 6.608241101041804e-05,
+      "loss": 0.4,
+      "step": 17040
+    },
+    {
+      "epoch": 6.436391090977727,
+      "grad_norm": 0.6156774759292603,
+      "learning_rate": 6.604326543937025e-05,
+      "loss": 0.4071,
+      "step": 17050
+    },
+    {
+      "epoch": 6.440166100415251,
+      "grad_norm": 0.6333221197128296,
+      "learning_rate": 6.60041089029018e-05,
+      "loss": 0.4074,
+      "step": 17060
+    },
+    {
+      "epoch": 6.443941109852775,
+      "grad_norm": 0.6324211955070496,
+      "learning_rate": 6.596494142777583e-05,
+      "loss": 0.4126,
+      "step": 17070
+    },
+    {
+      "epoch": 6.447716119290298,
+      "grad_norm": 0.6180362701416016,
+      "learning_rate": 6.592576304076294e-05,
+      "loss": 0.4113,
+      "step": 17080
+    },
+    {
+      "epoch": 6.451491128727822,
+      "grad_norm": 0.5131887197494507,
+      "learning_rate": 6.588657376864119e-05,
+      "loss": 0.4074,
+      "step": 17090
+    },
+    {
+      "epoch": 6.455266138165346,
+      "grad_norm": 0.570022702217102,
+      "learning_rate": 6.584737363819605e-05,
+      "loss": 0.4051,
+      "step": 17100
+    },
+    {
+      "epoch": 6.459041147602869,
+      "grad_norm": 0.5947955846786499,
+      "learning_rate": 6.580816267622048e-05,
+      "loss": 0.3996,
+      "step": 17110
+    },
+    {
+      "epoch": 6.462816157040392,
+      "grad_norm": 0.5993560552597046,
+      "learning_rate": 6.576894090951478e-05,
+      "loss": 0.4017,
+      "step": 17120
+    },
+    {
+      "epoch": 6.466591166477916,
+      "grad_norm": 0.5770085453987122,
+      "learning_rate": 6.572970836488665e-05,
+      "loss": 0.4095,
+      "step": 17130
+    },
+    {
+      "epoch": 6.47036617591544,
+      "grad_norm": 0.5673043727874756,
+      "learning_rate": 6.569046506915119e-05,
+      "loss": 0.3987,
+      "step": 17140
+    },
+    {
+      "epoch": 6.474141185352964,
+      "grad_norm": 0.6231826543807983,
+      "learning_rate": 6.56512110491308e-05,
+      "loss": 0.3943,
+      "step": 17150
+    },
+    {
+      "epoch": 6.477916194790487,
+      "grad_norm": 0.6213942766189575,
+      "learning_rate": 6.561194633165523e-05,
+      "loss": 0.4093,
+      "step": 17160
+    },
+    {
+      "epoch": 6.48169120422801,
+      "grad_norm": 0.6126642823219299,
+      "learning_rate": 6.557267094356155e-05,
+      "loss": 0.4164,
+      "step": 17170
+    },
+    {
+      "epoch": 6.485466213665534,
+      "grad_norm": 0.5821068286895752,
+      "learning_rate": 6.553338491169414e-05,
+      "loss": 0.4137,
+      "step": 17180
+    },
+    {
+      "epoch": 6.489241223103058,
+      "grad_norm": 0.5541254878044128,
+      "learning_rate": 6.54940882629046e-05,
+      "loss": 0.4062,
+      "step": 17190
+    },
+    {
+      "epoch": 6.493016232540581,
+      "grad_norm": 6.614664554595947,
+      "learning_rate": 6.545478102405184e-05,
+      "loss": 0.4263,
+      "step": 17200
+    },
+    {
+      "epoch": 6.496791241978105,
+      "grad_norm": 0.6379892230033875,
+      "learning_rate": 6.541546322200199e-05,
+      "loss": 0.4085,
+      "step": 17210
+    },
+    {
+      "epoch": 6.500566251415629,
+      "grad_norm": 0.5639785528182983,
+      "learning_rate": 6.537613488362837e-05,
+      "loss": 0.3997,
+      "step": 17220
+    },
+    {
+      "epoch": 6.504341260853153,
+      "grad_norm": 0.6088283061981201,
+      "learning_rate": 6.533679603581155e-05,
+      "loss": 0.4141,
+      "step": 17230
+    },
+    {
+      "epoch": 6.508116270290675,
+      "grad_norm": 0.7054754495620728,
+      "learning_rate": 6.529744670543926e-05,
+      "loss": 0.4144,
+      "step": 17240
+    },
+    {
+      "epoch": 6.511891279728199,
+      "grad_norm": 0.6340017914772034,
+      "learning_rate": 6.52580869194064e-05,
+      "loss": 0.4131,
+      "step": 17250
+    },
+    {
+      "epoch": 6.515666289165723,
+      "grad_norm": 0.6918211579322815,
+      "learning_rate": 6.521871670461499e-05,
+      "loss": 0.411,
+      "step": 17260
+    },
+    {
+      "epoch": 6.519441298603247,
+      "grad_norm": 0.6365815997123718,
+      "learning_rate": 6.517933608797422e-05,
+      "loss": 0.4104,
+      "step": 17270
+    },
+    {
+      "epoch": 6.52321630804077,
+      "grad_norm": 0.599440336227417,
+      "learning_rate": 6.513994509640038e-05,
+      "loss": 0.4075,
+      "step": 17280
+    },
+    {
+      "epoch": 6.526991317478293,
+      "grad_norm": 0.6258647441864014,
+      "learning_rate": 6.510054375681682e-05,
+      "loss": 0.4004,
+      "step": 17290
+    },
+    {
+      "epoch": 6.530766326915817,
+      "grad_norm": 0.7459936141967773,
+      "learning_rate": 6.506113209615398e-05,
+      "loss": 0.4078,
+      "step": 17300
+    },
+    {
+      "epoch": 6.534541336353341,
+      "grad_norm": 0.5855075120925903,
+      "learning_rate": 6.502171014134938e-05,
+      "loss": 0.4061,
+      "step": 17310
+    },
+    {
+      "epoch": 6.538316345790864,
+      "grad_norm": 0.6313618421554565,
+      "learning_rate": 6.498227791934755e-05,
+      "loss": 0.4085,
+      "step": 17320
+    },
+    {
+      "epoch": 6.542091355228388,
+      "grad_norm": 0.714318573474884,
+      "learning_rate": 6.494283545710003e-05,
+      "loss": 0.4015,
+      "step": 17330
+    },
+    {
+      "epoch": 6.545866364665912,
+      "grad_norm": 0.6481465697288513,
+      "learning_rate": 6.490338278156538e-05,
+      "loss": 0.4065,
+      "step": 17340
+    },
+    {
+      "epoch": 6.549641374103436,
+      "grad_norm": 0.7039319276809692,
+      "learning_rate": 6.486391991970913e-05,
+      "loss": 0.4136,
+      "step": 17350
+    },
+    {
+      "epoch": 6.553416383540959,
+      "grad_norm": 0.642465353012085,
+      "learning_rate": 6.482444689850377e-05,
+      "loss": 0.4026,
+      "step": 17360
+    },
+    {
+      "epoch": 6.557191392978482,
+      "grad_norm": 0.5987950563430786,
+      "learning_rate": 6.478496374492875e-05,
+      "loss": 0.4091,
+      "step": 17370
+    },
+    {
+      "epoch": 6.560966402416006,
+      "grad_norm": 0.6216861009597778,
+      "learning_rate": 6.474547048597042e-05,
+      "loss": 0.4006,
+      "step": 17380
+    },
+    {
+      "epoch": 6.56474141185353,
+      "grad_norm": 0.6449426412582397,
+      "learning_rate": 6.470596714862205e-05,
+      "loss": 0.4062,
+      "step": 17390
+    },
+    {
+      "epoch": 6.568516421291053,
+      "grad_norm": 0.5876918435096741,
+      "learning_rate": 6.46664537598838e-05,
+      "loss": 0.4061,
+      "step": 17400
+    },
+    {
+      "epoch": 6.572291430728577,
+      "grad_norm": 0.5914142727851868,
+      "learning_rate": 6.462693034676271e-05,
+      "loss": 0.4012,
+      "step": 17410
+    },
+    {
+      "epoch": 6.5760664401661,
+      "grad_norm": 0.561425507068634,
+      "learning_rate": 6.458739693627265e-05,
+      "loss": 0.3989,
+      "step": 17420
+    },
+    {
+      "epoch": 6.579841449603624,
+      "grad_norm": 0.6401806473731995,
+      "learning_rate": 6.454785355543432e-05,
+      "loss": 0.3995,
+      "step": 17430
+    },
+    {
+      "epoch": 6.5836164590411475,
+      "grad_norm": 0.660929262638092,
+      "learning_rate": 6.450830023127528e-05,
+      "loss": 0.4131,
+      "step": 17440
+    },
+    {
+      "epoch": 6.587391468478671,
+      "grad_norm": 0.6242609620094299,
+      "learning_rate": 6.446873699082982e-05,
+      "loss": 0.403,
+      "step": 17450
+    },
+    {
+      "epoch": 6.591166477916195,
+      "grad_norm": 0.5753269195556641,
+      "learning_rate": 6.44291638611391e-05,
+      "loss": 0.4042,
+      "step": 17460
+    },
+    {
+      "epoch": 6.594941487353719,
+      "grad_norm": 0.6394173502922058,
+      "learning_rate": 6.43895808692509e-05,
+      "loss": 0.4081,
+      "step": 17470
+    },
+    {
+      "epoch": 6.598716496791242,
+      "grad_norm": 0.6524138450622559,
+      "learning_rate": 6.434998804221986e-05,
+      "loss": 0.4026,
+      "step": 17480
+    },
+    {
+      "epoch": 6.602491506228765,
+      "grad_norm": 0.5924917459487915,
+      "learning_rate": 6.431038540710732e-05,
+      "loss": 0.4033,
+      "step": 17490
+    },
+    {
+      "epoch": 6.606266515666289,
+      "grad_norm": 2.0175888538360596,
+      "learning_rate": 6.427077299098129e-05,
+      "loss": 0.4034,
+      "step": 17500
+    },
+    {
+      "epoch": 6.610041525103813,
+      "grad_norm": 0.6277138590812683,
+      "learning_rate": 6.423115082091651e-05,
+      "loss": 0.3949,
+      "step": 17510
+    },
+    {
+      "epoch": 6.6138165345413364,
+      "grad_norm": 0.5728242993354797,
+      "learning_rate": 6.419151892399429e-05,
+      "loss": 0.3917,
+      "step": 17520
+    },
+    {
+      "epoch": 6.61759154397886,
+      "grad_norm": 0.6258454918861389,
+      "learning_rate": 6.415187732730273e-05,
+      "loss": 0.4026,
+      "step": 17530
+    },
+    {
+      "epoch": 6.621366553416384,
+      "grad_norm": 0.574672281742096,
+      "learning_rate": 6.411222605793645e-05,
+      "loss": 0.4017,
+      "step": 17540
+    },
+    {
+      "epoch": 6.625141562853907,
+      "grad_norm": 0.6237361431121826,
+      "learning_rate": 6.407256514299674e-05,
+      "loss": 0.3992,
+      "step": 17550
+    },
+    {
+      "epoch": 6.6289165722914305,
+      "grad_norm": 0.6078923940658569,
+      "learning_rate": 6.403289460959147e-05,
+      "loss": 0.4027,
+      "step": 17560
+    },
+    {
+      "epoch": 6.632691581728954,
+      "grad_norm": 0.6238769888877869,
+      "learning_rate": 6.399321448483501e-05,
+      "loss": 0.3993,
+      "step": 17570
+    },
+    {
+      "epoch": 6.636466591166478,
+      "grad_norm": 0.732795238494873,
+      "learning_rate": 6.395352479584844e-05,
+      "loss": 0.4017,
+      "step": 17580
+    },
+    {
+      "epoch": 6.640241600604002,
+      "grad_norm": 0.6137060523033142,
+      "learning_rate": 6.391382556975923e-05,
+      "loss": 0.3989,
+      "step": 17590
+    },
+    {
+      "epoch": 6.644016610041525,
+      "grad_norm": 0.6174196004867554,
+      "learning_rate": 6.387411683370144e-05,
+      "loss": 0.4061,
+      "step": 17600
+    },
+    {
+      "epoch": 6.647791619479049,
+      "grad_norm": 0.6093177199363708,
+      "learning_rate": 6.383439861481562e-05,
+      "loss": 0.3993,
+      "step": 17610
+    },
+    {
+      "epoch": 6.651566628916572,
+      "grad_norm": 0.624611496925354,
+      "learning_rate": 6.379467094024879e-05,
+      "loss": 0.4061,
+      "step": 17620
+    },
+    {
+      "epoch": 6.655341638354096,
+      "grad_norm": 0.7298497557640076,
+      "learning_rate": 6.375493383715445e-05,
+      "loss": 0.3959,
+      "step": 17630
+    },
+    {
+      "epoch": 6.6591166477916195,
+      "grad_norm": 0.5991964936256409,
+      "learning_rate": 6.371518733269254e-05,
+      "loss": 0.3992,
+      "step": 17640
+    },
+    {
+      "epoch": 6.662891657229143,
+      "grad_norm": 0.6233243942260742,
+      "learning_rate": 6.367543145402942e-05,
+      "loss": 0.4132,
+      "step": 17650
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 0.6464060544967651,
+      "learning_rate": 6.363566622833785e-05,
+      "loss": 0.399,
+      "step": 17660
+    },
+    {
+      "epoch": 6.67044167610419,
+      "grad_norm": 0.7238352298736572,
+      "learning_rate": 6.359589168279698e-05,
+      "loss": 0.3989,
+      "step": 17670
+    },
+    {
+      "epoch": 6.6742166855417135,
+      "grad_norm": 0.7250881791114807,
+      "learning_rate": 6.355610784459235e-05,
+      "loss": 0.4039,
+      "step": 17680
+    },
+    {
+      "epoch": 6.677991694979237,
+      "grad_norm": 0.5912888646125793,
+      "learning_rate": 6.351631474091585e-05,
+      "loss": 0.4017,
+      "step": 17690
+    },
+    {
+      "epoch": 6.681766704416761,
+      "grad_norm": 0.625900149345398,
+      "learning_rate": 6.347651239896566e-05,
+      "loss": 0.4034,
+      "step": 17700
+    },
+    {
+      "epoch": 6.685541713854285,
+      "grad_norm": 0.6302257180213928,
+      "learning_rate": 6.343670084594633e-05,
+      "loss": 0.4045,
+      "step": 17710
+    },
+    {
+      "epoch": 6.6893167232918085,
+      "grad_norm": 0.6661359667778015,
+      "learning_rate": 6.339688010906866e-05,
+      "loss": 0.4092,
+      "step": 17720
+    },
+    {
+      "epoch": 6.693091732729332,
+      "grad_norm": 0.5921412706375122,
+      "learning_rate": 6.335705021554975e-05,
+      "loss": 0.3951,
+      "step": 17730
+    },
+    {
+      "epoch": 6.696866742166855,
+      "grad_norm": 0.6364487409591675,
+      "learning_rate": 6.3317211192613e-05,
+      "loss": 0.4093,
+      "step": 17740
+    },
+    {
+      "epoch": 6.700641751604379,
+      "grad_norm": 0.6267909407615662,
+      "learning_rate": 6.327736306748795e-05,
+      "loss": 0.4035,
+      "step": 17750
+    },
+    {
+      "epoch": 6.7044167610419025,
+      "grad_norm": 0.5861899852752686,
+      "learning_rate": 6.323750586741047e-05,
+      "loss": 0.4027,
+      "step": 17760
+    },
+    {
+      "epoch": 6.708191770479426,
+      "grad_norm": 0.647732138633728,
+      "learning_rate": 6.319763961962252e-05,
+      "loss": 0.4064,
+      "step": 17770
+    },
+    {
+      "epoch": 6.71196677991695,
+      "grad_norm": 0.618825376033783,
+      "learning_rate": 6.315776435137233e-05,
+      "loss": 0.4008,
+      "step": 17780
+    },
+    {
+      "epoch": 6.715741789354474,
+      "grad_norm": 0.7474274039268494,
+      "learning_rate": 6.311788008991432e-05,
+      "loss": 0.4089,
+      "step": 17790
+    },
+    {
+      "epoch": 6.719516798791997,
+      "grad_norm": 0.6482340693473816,
+      "learning_rate": 6.307798686250891e-05,
+      "loss": 0.4037,
+      "step": 17800
+    },
+    {
+      "epoch": 6.72329180822952,
+      "grad_norm": 0.6014053821563721,
+      "learning_rate": 6.303808469642284e-05,
+      "loss": 0.4031,
+      "step": 17810
+    },
+    {
+      "epoch": 6.727066817667044,
+      "grad_norm": 0.6458423137664795,
+      "learning_rate": 6.29981736189288e-05,
+      "loss": 0.3998,
+      "step": 17820
+    },
+    {
+      "epoch": 6.730841827104568,
+      "grad_norm": 0.6253558397293091,
+      "learning_rate": 6.295825365730567e-05,
+      "loss": 0.4044,
+      "step": 17830
+    },
+    {
+      "epoch": 6.7346168365420915,
+      "grad_norm": 0.6395803093910217,
+      "learning_rate": 6.291832483883835e-05,
+      "loss": 0.4026,
+      "step": 17840
+    },
+    {
+      "epoch": 6.738391845979615,
+      "grad_norm": 0.6076302528381348,
+      "learning_rate": 6.28783871908178e-05,
+      "loss": 0.3979,
+      "step": 17850
+    },
+    {
+      "epoch": 6.742166855417139,
+      "grad_norm": 0.591390073299408,
+      "learning_rate": 6.283844074054107e-05,
+      "loss": 0.3967,
+      "step": 17860
+    },
+    {
+      "epoch": 6.745941864854662,
+      "grad_norm": 0.5707950592041016,
+      "learning_rate": 6.279848551531112e-05,
+      "loss": 0.3941,
+      "step": 17870
+    },
+    {
+      "epoch": 6.749716874292186,
+      "grad_norm": 0.6005119681358337,
+      "learning_rate": 6.275852154243702e-05,
+      "loss": 0.407,
+      "step": 17880
+    },
+    {
+      "epoch": 6.753491883729709,
+      "grad_norm": 0.5978226065635681,
+      "learning_rate": 6.271854884923377e-05,
+      "loss": 0.3962,
+      "step": 17890
+    },
+    {
+      "epoch": 6.757266893167233,
+      "grad_norm": 0.637346088886261,
+      "learning_rate": 6.267856746302228e-05,
+      "loss": 0.3962,
+      "step": 17900
+    },
+    {
+      "epoch": 6.761041902604757,
+      "grad_norm": 0.7648394107818604,
+      "learning_rate": 6.263857741112948e-05,
+      "loss": 0.4004,
+      "step": 17910
+    },
+    {
+      "epoch": 6.76481691204228,
+      "grad_norm": 0.6177964806556702,
+      "learning_rate": 6.259857872088821e-05,
+      "loss": 0.3901,
+      "step": 17920
+    },
+    {
+      "epoch": 6.768591921479803,
+      "grad_norm": 0.6134040355682373,
+      "learning_rate": 6.255857141963719e-05,
+      "loss": 0.3985,
+      "step": 17930
+    },
+    {
+      "epoch": 6.772366930917327,
+      "grad_norm": 0.5989758968353271,
+      "learning_rate": 6.251855553472101e-05,
+      "loss": 0.398,
+      "step": 17940
+    },
+    {
+      "epoch": 6.776141940354851,
+      "grad_norm": 0.6749169230461121,
+      "learning_rate": 6.247853109349016e-05,
+      "loss": 0.3976,
+      "step": 17950
+    },
+    {
+      "epoch": 6.779916949792375,
+      "grad_norm": 0.6461236476898193,
+      "learning_rate": 6.243849812330098e-05,
+      "loss": 0.3982,
+      "step": 17960
+    },
+    {
+      "epoch": 6.783691959229898,
+      "grad_norm": 1.9516565799713135,
+      "learning_rate": 6.239845665151563e-05,
+      "loss": 0.3939,
+      "step": 17970
+    },
+    {
+      "epoch": 6.787466968667422,
+      "grad_norm": 0.6713792681694031,
+      "learning_rate": 6.235840670550204e-05,
+      "loss": 0.4018,
+      "step": 17980
+    },
+    {
+      "epoch": 6.791241978104945,
+      "grad_norm": 0.6273426413536072,
+      "learning_rate": 6.231834831263403e-05,
+      "loss": 0.401,
+      "step": 17990
+    },
+    {
+      "epoch": 6.795016987542469,
+      "grad_norm": 0.6532526612281799,
+      "learning_rate": 6.22782815002911e-05,
+      "loss": 0.4034,
+      "step": 18000
+    },
+    {
+      "epoch": 6.798791996979992,
+      "grad_norm": 0.6998550891876221,
+      "learning_rate": 6.223820629585852e-05,
+      "loss": 0.4019,
+      "step": 18010
+    },
+    {
+      "epoch": 6.802567006417516,
+      "grad_norm": 0.6373661756515503,
+      "learning_rate": 6.219812272672737e-05,
+      "loss": 0.4007,
+      "step": 18020
+    },
+    {
+      "epoch": 6.80634201585504,
+      "grad_norm": 0.6695569157600403,
+      "learning_rate": 6.215803082029434e-05,
+      "loss": 0.3977,
+      "step": 18030
+    },
+    {
+      "epoch": 6.810117025292564,
+      "grad_norm": 0.7880366444587708,
+      "learning_rate": 6.211793060396188e-05,
+      "loss": 0.3985,
+      "step": 18040
+    },
+    {
+      "epoch": 6.813892034730086,
+      "grad_norm": 0.6441872119903564,
+      "learning_rate": 6.207782210513811e-05,
+      "loss": 0.4013,
+      "step": 18050
+    },
+    {
+      "epoch": 6.81766704416761,
+      "grad_norm": 0.6251243948936462,
+      "learning_rate": 6.203770535123683e-05,
+      "loss": 0.4084,
+      "step": 18060
+    },
+    {
+      "epoch": 6.821442053605134,
+      "grad_norm": 0.710367739200592,
+      "learning_rate": 6.199758036967747e-05,
+      "loss": 0.4,
+      "step": 18070
+    },
+    {
+      "epoch": 6.825217063042658,
+      "grad_norm": 0.568421483039856,
+      "learning_rate": 6.195744718788503e-05,
+      "loss": 0.3971,
+      "step": 18080
+    },
+    {
+      "epoch": 6.828992072480181,
+      "grad_norm": 0.6238227486610413,
+      "learning_rate": 6.191730583329021e-05,
+      "loss": 0.3973,
+      "step": 18090
+    },
+    {
+      "epoch": 6.832767081917705,
+      "grad_norm": 0.6105861067771912,
+      "learning_rate": 6.187715633332921e-05,
+      "loss": 0.3994,
+      "step": 18100
+    },
+    {
+      "epoch": 6.836542091355229,
+      "grad_norm": 0.5954229235649109,
+      "learning_rate": 6.183699871544386e-05,
+      "loss": 0.4027,
+      "step": 18110
+    },
+    {
+      "epoch": 6.840317100792752,
+      "grad_norm": 0.5816265344619751,
+      "learning_rate": 6.179683300708152e-05,
+      "loss": 0.3998,
+      "step": 18120
+    },
+    {
+      "epoch": 6.844092110230275,
+      "grad_norm": 0.609210729598999,
+      "learning_rate": 6.175665923569503e-05,
+      "loss": 0.3993,
+      "step": 18130
+    },
+    {
+      "epoch": 6.847867119667799,
+      "grad_norm": 0.6947378516197205,
+      "learning_rate": 6.171647742874281e-05,
+      "loss": 0.3964,
+      "step": 18140
+    },
+    {
+      "epoch": 6.851642129105323,
+      "grad_norm": 0.6516598463058472,
+      "learning_rate": 6.167628761368875e-05,
+      "loss": 0.4027,
+      "step": 18150
+    },
+    {
+      "epoch": 6.855417138542847,
+      "grad_norm": 0.5862056612968445,
+      "learning_rate": 6.163608981800222e-05,
+      "loss": 0.4079,
+      "step": 18160
+    },
+    {
+      "epoch": 6.8591921479803695,
+      "grad_norm": 0.626376211643219,
+      "learning_rate": 6.159588406915803e-05,
+      "loss": 0.4003,
+      "step": 18170
+    },
+    {
+      "epoch": 6.862967157417893,
+      "grad_norm": 0.6769831776618958,
+      "learning_rate": 6.155567039463639e-05,
+      "loss": 0.4041,
+      "step": 18180
+    },
+    {
+      "epoch": 6.866742166855417,
+      "grad_norm": 0.6997120380401611,
+      "learning_rate": 6.151544882192302e-05,
+      "loss": 0.4028,
+      "step": 18190
+    },
+    {
+      "epoch": 6.870517176292941,
+      "grad_norm": 0.5860679745674133,
+      "learning_rate": 6.147521937850895e-05,
+      "loss": 0.4132,
+      "step": 18200
+    },
+    {
+      "epoch": 6.874292185730464,
+      "grad_norm": 0.5797774791717529,
+      "learning_rate": 6.143498209189066e-05,
+      "loss": 0.4019,
+      "step": 18210
+    },
+    {
+      "epoch": 6.878067195167988,
+      "grad_norm": 0.6230421662330627,
+      "learning_rate": 6.139473698956993e-05,
+      "loss": 0.4041,
+      "step": 18220
+    },
+    {
+      "epoch": 6.881842204605512,
+      "grad_norm": 0.5917826890945435,
+      "learning_rate": 6.13544840990539e-05,
+      "loss": 0.402,
+      "step": 18230
+    },
+    {
+      "epoch": 6.885617214043035,
+      "grad_norm": 1.265276312828064,
+      "learning_rate": 6.131422344785507e-05,
+      "loss": 0.3984,
+      "step": 18240
+    },
+    {
+      "epoch": 6.8893922234805585,
+      "grad_norm": 0.5542703866958618,
+      "learning_rate": 6.127395506349119e-05,
+      "loss": 0.3988,
+      "step": 18250
+    },
+    {
+      "epoch": 6.893167232918082,
+      "grad_norm": 0.736250102519989,
+      "learning_rate": 6.123367897348533e-05,
+      "loss": 0.3968,
+      "step": 18260
+    },
+    {
+      "epoch": 6.896942242355606,
+      "grad_norm": 0.8266544342041016,
+      "learning_rate": 6.119339520536584e-05,
+      "loss": 0.4013,
+      "step": 18270
+    },
+    {
+      "epoch": 6.90071725179313,
+      "grad_norm": 0.6568412780761719,
+      "learning_rate": 6.115310378666625e-05,
+      "loss": 0.4019,
+      "step": 18280
+    },
+    {
+      "epoch": 6.904492261230653,
+      "grad_norm": 0.5549454689025879,
+      "learning_rate": 6.11128047449254e-05,
+      "loss": 0.3982,
+      "step": 18290
+    },
+    {
+      "epoch": 6.908267270668176,
+      "grad_norm": 0.6285567283630371,
+      "learning_rate": 6.107249810768729e-05,
+      "loss": 0.4012,
+      "step": 18300
+    },
+    {
+      "epoch": 6.9120422801057,
+      "grad_norm": 0.6208999156951904,
+      "learning_rate": 6.1032183902501125e-05,
+      "loss": 0.3999,
+      "step": 18310
+    },
+    {
+      "epoch": 6.915817289543224,
+      "grad_norm": 0.5961741805076599,
+      "learning_rate": 6.099186215692131e-05,
+      "loss": 0.4093,
+      "step": 18320
+    },
+    {
+      "epoch": 6.9195922989807475,
+      "grad_norm": 0.5774527192115784,
+      "learning_rate": 6.095153289850734e-05,
+      "loss": 0.401,
+      "step": 18330
+    },
+    {
+      "epoch": 6.923367308418271,
+      "grad_norm": 0.6928045749664307,
+      "learning_rate": 6.0911196154823904e-05,
+      "loss": 0.3907,
+      "step": 18340
+    },
+    {
+      "epoch": 6.927142317855795,
+      "grad_norm": 0.7921097874641418,
+      "learning_rate": 6.087085195344079e-05,
+      "loss": 0.395,
+      "step": 18350
+    },
+    {
+      "epoch": 6.930917327293319,
+      "grad_norm": 0.6023475527763367,
+      "learning_rate": 6.083050032193286e-05,
+      "loss": 0.4031,
+      "step": 18360
+    },
+    {
+      "epoch": 6.9346923367308415,
+      "grad_norm": 1.3727141618728638,
+      "learning_rate": 6.0790141287880097e-05,
+      "loss": 0.4059,
+      "step": 18370
+    },
+    {
+      "epoch": 6.938467346168365,
+      "grad_norm": 0.606326699256897,
+      "learning_rate": 6.0749774878867496e-05,
+      "loss": 0.3957,
+      "step": 18380
+    },
+    {
+      "epoch": 6.942242355605889,
+      "grad_norm": 0.6082478165626526,
+      "learning_rate": 6.0709401122485146e-05,
+      "loss": 0.3975,
+      "step": 18390
+    },
+    {
+      "epoch": 6.946017365043413,
+      "grad_norm": 0.7824208736419678,
+      "learning_rate": 6.066902004632811e-05,
+      "loss": 0.3967,
+      "step": 18400
+    },
+    {
+      "epoch": 6.9497923744809365,
+      "grad_norm": 0.6232104301452637,
+      "learning_rate": 6.062863167799646e-05,
+      "loss": 0.4024,
+      "step": 18410
+    },
+    {
+      "epoch": 6.953567383918459,
+      "grad_norm": 0.6417713165283203,
+      "learning_rate": 6.058823604509529e-05,
+      "loss": 0.3967,
+      "step": 18420
+    },
+    {
+      "epoch": 6.957342393355983,
+      "grad_norm": 0.6295148730278015,
+      "learning_rate": 6.054783317523462e-05,
+      "loss": 0.3991,
+      "step": 18430
+    },
+    {
+      "epoch": 6.961117402793507,
+      "grad_norm": 0.6527547836303711,
+      "learning_rate": 6.050742309602944e-05,
+      "loss": 0.3988,
+      "step": 18440
+    },
+    {
+      "epoch": 6.9648924122310305,
+      "grad_norm": 0.6303055882453918,
+      "learning_rate": 6.046700583509965e-05,
+      "loss": 0.3933,
+      "step": 18450
+    },
+    {
+      "epoch": 6.968667421668554,
+      "grad_norm": 0.637154757976532,
+      "learning_rate": 6.042658142007007e-05,
+      "loss": 0.4024,
+      "step": 18460
+    },
+    {
+      "epoch": 6.972442431106078,
+      "grad_norm": 0.6554228663444519,
+      "learning_rate": 6.038614987857041e-05,
+      "loss": 0.3999,
+      "step": 18470
+    },
+    {
+      "epoch": 6.976217440543602,
+      "grad_norm": 0.6131272315979004,
+      "learning_rate": 6.0345711238235224e-05,
+      "loss": 0.3946,
+      "step": 18480
+    },
+    {
+      "epoch": 6.9799924499811254,
+      "grad_norm": 0.5949199199676514,
+      "learning_rate": 6.030526552670399e-05,
+      "loss": 0.4006,
+      "step": 18490
+    },
+    {
+      "epoch": 6.983767459418648,
+      "grad_norm": 0.5946569442749023,
+      "learning_rate": 6.0264812771620925e-05,
+      "loss": 0.3971,
+      "step": 18500
+    },
+    {
+      "epoch": 6.987542468856172,
+      "grad_norm": 0.7052402496337891,
+      "learning_rate": 6.022435300063512e-05,
+      "loss": 0.3864,
+      "step": 18510
+    },
+    {
+      "epoch": 6.991317478293696,
+      "grad_norm": 0.5748345255851746,
+      "learning_rate": 6.0183886241400466e-05,
+      "loss": 0.4049,
+      "step": 18520
+    },
+    {
+      "epoch": 6.9950924877312195,
+      "grad_norm": 0.6450919508934021,
+      "learning_rate": 6.0143412521575584e-05,
+      "loss": 0.4042,
+      "step": 18530
+    },
+    {
+      "epoch": 6.998867497168743,
+      "grad_norm": 0.7220432162284851,
+      "learning_rate": 6.010293186882389e-05,
+      "loss": 0.4025,
+      "step": 18540
+    },
+    {
+      "epoch": 7.002642506606266,
+      "grad_norm": 0.6318272352218628,
+      "learning_rate": 6.0062444310813525e-05,
+      "loss": 0.4026,
+      "step": 18550
+    },
+    {
+      "epoch": 7.00641751604379,
+      "grad_norm": 0.5739586353302002,
+      "learning_rate": 6.0021949875217355e-05,
+      "loss": 0.3905,
+      "step": 18560
+    },
+    {
+      "epoch": 7.0101925254813136,
+      "grad_norm": 0.6182407140731812,
+      "learning_rate": 5.998144858971295e-05,
+      "loss": 0.3936,
+      "step": 18570
+    },
+    {
+      "epoch": 7.013967534918837,
+      "grad_norm": 0.6267920732498169,
+      "learning_rate": 5.994094048198257e-05,
+      "loss": 0.4063,
+      "step": 18580
+    },
+    {
+      "epoch": 7.017742544356361,
+      "grad_norm": 0.7207063436508179,
+      "learning_rate": 5.990042557971307e-05,
+      "loss": 0.3992,
+      "step": 18590
+    },
+    {
+      "epoch": 7.021517553793885,
+      "grad_norm": 0.6681888103485107,
+      "learning_rate": 5.985990391059607e-05,
+      "loss": 0.3976,
+      "step": 18600
+    },
+    {
+      "epoch": 7.0252925632314085,
+      "grad_norm": 0.6886135935783386,
+      "learning_rate": 5.981937550232771e-05,
+      "loss": 0.3898,
+      "step": 18610
+    },
+    {
+      "epoch": 7.029067572668931,
+      "grad_norm": 0.7061811685562134,
+      "learning_rate": 5.9778840382608794e-05,
+      "loss": 0.4119,
+      "step": 18620
+    },
+    {
+      "epoch": 7.032842582106455,
+      "grad_norm": 0.5984750986099243,
+      "learning_rate": 5.9738298579144695e-05,
+      "loss": 0.3965,
+      "step": 18630
+    },
+    {
+      "epoch": 7.036617591543979,
+      "grad_norm": 0.6792024374008179,
+      "learning_rate": 5.9697750119645314e-05,
+      "loss": 0.4087,
+      "step": 18640
+    },
+    {
+      "epoch": 7.0403926009815025,
+      "grad_norm": 0.5802191495895386,
+      "learning_rate": 5.96571950318252e-05,
+      "loss": 0.4003,
+      "step": 18650
+    },
+    {
+      "epoch": 7.044167610419026,
+      "grad_norm": 0.6558259129524231,
+      "learning_rate": 5.9616633343403316e-05,
+      "loss": 0.408,
+      "step": 18660
+    },
+    {
+      "epoch": 7.04794261985655,
+      "grad_norm": 0.7390403151512146,
+      "learning_rate": 5.957606508210324e-05,
+      "loss": 0.404,
+      "step": 18670
+    },
+    {
+      "epoch": 7.051717629294073,
+      "grad_norm": 0.6561256051063538,
+      "learning_rate": 5.953549027565297e-05,
+      "loss": 0.3936,
+      "step": 18680
+    },
+    {
+      "epoch": 7.055492638731597,
+      "grad_norm": 0.6152408123016357,
+      "learning_rate": 5.949490895178501e-05,
+      "loss": 0.3959,
+      "step": 18690
+    },
+    {
+      "epoch": 7.05926764816912,
+      "grad_norm": 1.134881854057312,
+      "learning_rate": 5.945432113823632e-05,
+      "loss": 0.3955,
+      "step": 18700
+    },
+    {
+      "epoch": 7.063042657606644,
+      "grad_norm": 0.611450731754303,
+      "learning_rate": 5.9413726862748276e-05,
+      "loss": 0.3999,
+      "step": 18710
+    },
+    {
+      "epoch": 7.066817667044168,
+      "grad_norm": 1.1153512001037598,
+      "learning_rate": 5.9373126153066694e-05,
+      "loss": 0.3965,
+      "step": 18720
+    },
+    {
+      "epoch": 7.0705926764816915,
+      "grad_norm": 0.6631627678871155,
+      "learning_rate": 5.933251903694177e-05,
+      "loss": 0.4053,
+      "step": 18730
+    },
+    {
+      "epoch": 7.074367685919214,
+      "grad_norm": 0.6014063954353333,
+      "learning_rate": 5.929190554212807e-05,
+      "loss": 0.3936,
+      "step": 18740
+    },
+    {
+      "epoch": 7.078142695356738,
+      "grad_norm": 0.6215736269950867,
+      "learning_rate": 5.9251285696384565e-05,
+      "loss": 0.404,
+      "step": 18750
+    },
+    {
+      "epoch": 7.081917704794262,
+      "grad_norm": 0.791002094745636,
+      "learning_rate": 5.921065952747451e-05,
+      "loss": 0.3992,
+      "step": 18760
+    },
+    {
+      "epoch": 7.085692714231786,
+      "grad_norm": 0.9129757881164551,
+      "learning_rate": 5.917002706316552e-05,
+      "loss": 0.3984,
+      "step": 18770
+    },
+    {
+      "epoch": 7.089467723669309,
+      "grad_norm": 0.5969563722610474,
+      "learning_rate": 5.912938833122952e-05,
+      "loss": 0.3988,
+      "step": 18780
+    },
+    {
+      "epoch": 7.093242733106833,
+      "grad_norm": 1.1693058013916016,
+      "learning_rate": 5.908874335944265e-05,
+      "loss": 0.3969,
+      "step": 18790
+    },
+    {
+      "epoch": 7.097017742544356,
+      "grad_norm": 0.6244138479232788,
+      "learning_rate": 5.904809217558542e-05,
+      "loss": 0.3992,
+      "step": 18800
+    },
+    {
+      "epoch": 7.10079275198188,
+      "grad_norm": 0.6120061874389648,
+      "learning_rate": 5.90074348074425e-05,
+      "loss": 0.3984,
+      "step": 18810
+    },
+    {
+      "epoch": 7.104567761419403,
+      "grad_norm": 0.6857101917266846,
+      "learning_rate": 5.8966771282802814e-05,
+      "loss": 0.3949,
+      "step": 18820
+    },
+    {
+      "epoch": 7.108342770856927,
+      "grad_norm": 0.6439679265022278,
+      "learning_rate": 5.892610162945952e-05,
+      "loss": 0.401,
+      "step": 18830
+    },
+    {
+      "epoch": 7.112117780294451,
+      "grad_norm": 0.686291515827179,
+      "learning_rate": 5.8885425875209924e-05,
+      "loss": 0.3991,
+      "step": 18840
+    },
+    {
+      "epoch": 7.115892789731975,
+      "grad_norm": 0.893528938293457,
+      "learning_rate": 5.884474404785553e-05,
+      "loss": 0.3903,
+      "step": 18850
+    },
+    {
+      "epoch": 7.119667799169498,
+      "grad_norm": 0.6380524635314941,
+      "learning_rate": 5.8804056175201983e-05,
+      "loss": 0.3936,
+      "step": 18860
+    },
+    {
+      "epoch": 7.123442808607021,
+      "grad_norm": 0.5932635068893433,
+      "learning_rate": 5.876336228505904e-05,
+      "loss": 0.3988,
+      "step": 18870
+    },
+    {
+      "epoch": 7.127217818044545,
+      "grad_norm": 0.6427797079086304,
+      "learning_rate": 5.872266240524062e-05,
+      "loss": 0.4023,
+      "step": 18880
+    },
+    {
+      "epoch": 7.130992827482069,
+      "grad_norm": 0.625927209854126,
+      "learning_rate": 5.86819565635647e-05,
+      "loss": 0.3963,
+      "step": 18890
+    },
+    {
+      "epoch": 7.134767836919592,
+      "grad_norm": 0.5989848971366882,
+      "learning_rate": 5.8641244787853334e-05,
+      "loss": 0.3958,
+      "step": 18900
+    },
+    {
+      "epoch": 7.138542846357116,
+      "grad_norm": 0.6157662868499756,
+      "learning_rate": 5.860052710593265e-05,
+      "loss": 0.3974,
+      "step": 18910
+    },
+    {
+      "epoch": 7.14231785579464,
+      "grad_norm": 0.6159812211990356,
+      "learning_rate": 5.855980354563276e-05,
+      "loss": 0.3868,
+      "step": 18920
+    },
+    {
+      "epoch": 7.146092865232163,
+      "grad_norm": 0.5764108300209045,
+      "learning_rate": 5.8519074134787874e-05,
+      "loss": 0.3937,
+      "step": 18930
+    },
+    {
+      "epoch": 7.149867874669686,
+      "grad_norm": 0.657748281955719,
+      "learning_rate": 5.847833890123614e-05,
+      "loss": 0.3937,
+      "step": 18940
+    },
+    {
+      "epoch": 7.15364288410721,
+      "grad_norm": 0.6909071207046509,
+      "learning_rate": 5.8437597872819737e-05,
+      "loss": 0.397,
+      "step": 18950
+    },
+    {
+      "epoch": 7.157417893544734,
+      "grad_norm": 0.6924566030502319,
+      "learning_rate": 5.839685107738473e-05,
+      "loss": 0.397,
+      "step": 18960
+    },
+    {
+      "epoch": 7.161192902982258,
+      "grad_norm": 0.6301664710044861,
+      "learning_rate": 5.835609854278118e-05,
+      "loss": 0.3921,
+      "step": 18970
+    },
+    {
+      "epoch": 7.164967912419781,
+      "grad_norm": 0.6503030061721802,
+      "learning_rate": 5.831534029686308e-05,
+      "loss": 0.3958,
+      "step": 18980
+    },
+    {
+      "epoch": 7.168742921857304,
+      "grad_norm": 0.6070290207862854,
+      "learning_rate": 5.82745763674883e-05,
+      "loss": 0.3943,
+      "step": 18990
+    },
+    {
+      "epoch": 7.172517931294828,
+      "grad_norm": 0.6799952387809753,
+      "learning_rate": 5.823380678251861e-05,
+      "loss": 0.3899,
+      "step": 19000
+    },
+    {
+      "epoch": 7.176292940732352,
+      "grad_norm": 0.6328525543212891,
+      "learning_rate": 5.81930315698196e-05,
+      "loss": 0.3935,
+      "step": 19010
+    },
+    {
+      "epoch": 7.180067950169875,
+      "grad_norm": 0.6666650176048279,
+      "learning_rate": 5.815225075726076e-05,
+      "loss": 0.3931,
+      "step": 19020
+    },
+    {
+      "epoch": 7.183842959607399,
+      "grad_norm": 0.6060150861740112,
+      "learning_rate": 5.811146437271543e-05,
+      "loss": 0.3976,
+      "step": 19030
+    },
+    {
+      "epoch": 7.187617969044923,
+      "grad_norm": 0.6627577543258667,
+      "learning_rate": 5.807067244406066e-05,
+      "loss": 0.4025,
+      "step": 19040
+    },
+    {
+      "epoch": 7.191392978482447,
+      "grad_norm": 0.5842311978340149,
+      "learning_rate": 5.8029874999177405e-05,
+      "loss": 0.3941,
+      "step": 19050
+    },
+    {
+      "epoch": 7.1951679879199695,
+      "grad_norm": 1.3617955446243286,
+      "learning_rate": 5.798907206595029e-05,
+      "loss": 0.3934,
+      "step": 19060
+    },
+    {
+      "epoch": 7.198942997357493,
+      "grad_norm": 0.5990311503410339,
+      "learning_rate": 5.794826367226773e-05,
+      "loss": 0.3947,
+      "step": 19070
+    },
+    {
+      "epoch": 7.202718006795017,
+      "grad_norm": 0.5806323885917664,
+      "learning_rate": 5.790744984602193e-05,
+      "loss": 0.4021,
+      "step": 19080
+    },
+    {
+      "epoch": 7.206493016232541,
+      "grad_norm": 0.6318025588989258,
+      "learning_rate": 5.786663061510872e-05,
+      "loss": 0.3978,
+      "step": 19090
+    },
+    {
+      "epoch": 7.210268025670064,
+      "grad_norm": 0.6155949234962463,
+      "learning_rate": 5.782580600742765e-05,
+      "loss": 0.395,
+      "step": 19100
+    },
+    {
+      "epoch": 7.214043035107588,
+      "grad_norm": 0.6419726014137268,
+      "learning_rate": 5.7784976050881965e-05,
+      "loss": 0.395,
+      "step": 19110
+    },
+    {
+      "epoch": 7.217818044545111,
+      "grad_norm": 0.6743911504745483,
+      "learning_rate": 5.774414077337855e-05,
+      "loss": 0.3907,
+      "step": 19120
+    },
+    {
+      "epoch": 7.221593053982635,
+      "grad_norm": 0.6767613887786865,
+      "learning_rate": 5.770330020282796e-05,
+      "loss": 0.3911,
+      "step": 19130
+    },
+    {
+      "epoch": 7.2253680634201585,
+      "grad_norm": 0.6418374180793762,
+      "learning_rate": 5.7662454367144317e-05,
+      "loss": 0.3978,
+      "step": 19140
+    },
+    {
+      "epoch": 7.229143072857682,
+      "grad_norm": 0.5886910557746887,
+      "learning_rate": 5.762160329424536e-05,
+      "loss": 0.398,
+      "step": 19150
+    },
+    {
+      "epoch": 7.232918082295206,
+      "grad_norm": 0.56149822473526,
+      "learning_rate": 5.7580747012052416e-05,
+      "loss": 0.3917,
+      "step": 19160
+    },
+    {
+      "epoch": 7.23669309173273,
+      "grad_norm": 0.8316841125488281,
+      "learning_rate": 5.753988554849037e-05,
+      "loss": 0.3967,
+      "step": 19170
+    },
+    {
+      "epoch": 7.2404681011702525,
+      "grad_norm": 0.5947608351707458,
+      "learning_rate": 5.749901893148766e-05,
+      "loss": 0.395,
+      "step": 19180
+    },
+    {
+      "epoch": 7.244243110607776,
+      "grad_norm": 0.6548729538917542,
+      "learning_rate": 5.745814718897621e-05,
+      "loss": 0.3925,
+      "step": 19190
+    },
+    {
+      "epoch": 7.2480181200453,
+      "grad_norm": 0.5887550115585327,
+      "learning_rate": 5.74172703488915e-05,
+      "loss": 0.3922,
+      "step": 19200
+    },
+    {
+      "epoch": 7.251793129482824,
+      "grad_norm": 0.6358749270439148,
+      "learning_rate": 5.737638843917242e-05,
+      "loss": 0.3876,
+      "step": 19210
+    },
+    {
+      "epoch": 7.2555681389203475,
+      "grad_norm": 0.7427713871002197,
+      "learning_rate": 5.73355014877614e-05,
+      "loss": 0.3955,
+      "step": 19220
+    },
+    {
+      "epoch": 7.259343148357871,
+      "grad_norm": 0.6359094977378845,
+      "learning_rate": 5.7294609522604316e-05,
+      "loss": 0.3964,
+      "step": 19230
+    },
+    {
+      "epoch": 7.263118157795395,
+      "grad_norm": 0.5932873487472534,
+      "learning_rate": 5.7253712571650376e-05,
+      "loss": 0.3966,
+      "step": 19240
+    },
+    {
+      "epoch": 7.266893167232918,
+      "grad_norm": 0.6501133441925049,
+      "learning_rate": 5.721281066285229e-05,
+      "loss": 0.3909,
+      "step": 19250
+    },
+    {
+      "epoch": 7.2706681766704415,
+      "grad_norm": 0.6515145301818848,
+      "learning_rate": 5.717190382416615e-05,
+      "loss": 0.3967,
+      "step": 19260
+    },
+    {
+      "epoch": 7.274443186107965,
+      "grad_norm": 0.6237747073173523,
+      "learning_rate": 5.713099208355135e-05,
+      "loss": 0.3959,
+      "step": 19270
+    },
+    {
+      "epoch": 7.278218195545489,
+      "grad_norm": 0.5545824766159058,
+      "learning_rate": 5.709007546897074e-05,
+      "loss": 0.396,
+      "step": 19280
+    },
+    {
+      "epoch": 7.281993204983013,
+      "grad_norm": 0.6651487350463867,
+      "learning_rate": 5.704915400839037e-05,
+      "loss": 0.3994,
+      "step": 19290
+    },
+    {
+      "epoch": 7.2857682144205365,
+      "grad_norm": 0.701859176158905,
+      "learning_rate": 5.700822772977971e-05,
+      "loss": 0.4039,
+      "step": 19300
+    },
+    {
+      "epoch": 7.289543223858059,
+      "grad_norm": 0.9425735473632812,
+      "learning_rate": 5.696729666111148e-05,
+      "loss": 0.4047,
+      "step": 19310
+    },
+    {
+      "epoch": 7.293318233295583,
+      "grad_norm": 0.6198208928108215,
+      "learning_rate": 5.692636083036168e-05,
+      "loss": 0.4123,
+      "step": 19320
+    },
+    {
+      "epoch": 7.297093242733107,
+      "grad_norm": 0.7142148613929749,
+      "learning_rate": 5.688542026550958e-05,
+      "loss": 0.3903,
+      "step": 19330
+    },
+    {
+      "epoch": 7.3008682521706305,
+      "grad_norm": 0.6616962552070618,
+      "learning_rate": 5.684447499453763e-05,
+      "loss": 0.3973,
+      "step": 19340
+    },
+    {
+      "epoch": 7.304643261608154,
+      "grad_norm": 0.6994644403457642,
+      "learning_rate": 5.680352504543156e-05,
+      "loss": 0.3936,
+      "step": 19350
+    },
+    {
+      "epoch": 7.308418271045678,
+      "grad_norm": 0.6935665011405945,
+      "learning_rate": 5.67625704461803e-05,
+      "loss": 0.3882,
+      "step": 19360
+    },
+    {
+      "epoch": 7.312193280483201,
+      "grad_norm": 0.6593263745307922,
+      "learning_rate": 5.672161122477589e-05,
+      "loss": 0.4025,
+      "step": 19370
+    },
+    {
+      "epoch": 7.315968289920725,
+      "grad_norm": 1.2950445413589478,
+      "learning_rate": 5.668064740921359e-05,
+      "loss": 0.3984,
+      "step": 19380
+    },
+    {
+      "epoch": 7.319743299358248,
+      "grad_norm": 0.6627549529075623,
+      "learning_rate": 5.663967902749179e-05,
+      "loss": 0.3958,
+      "step": 19390
+    },
+    {
+      "epoch": 7.323518308795772,
+      "grad_norm": 0.7319017648696899,
+      "learning_rate": 5.6598706107611965e-05,
+      "loss": 0.3883,
+      "step": 19400
+    },
+    {
+      "epoch": 7.327293318233296,
+      "grad_norm": 0.6958197951316833,
+      "learning_rate": 5.655772867757876e-05,
+      "loss": 0.3962,
+      "step": 19410
+    },
+    {
+      "epoch": 7.3310683276708195,
+      "grad_norm": 0.6247055530548096,
+      "learning_rate": 5.651674676539982e-05,
+      "loss": 0.3911,
+      "step": 19420
+    },
+    {
+      "epoch": 7.334843337108342,
+      "grad_norm": 0.6021857857704163,
+      "learning_rate": 5.647576039908593e-05,
+      "loss": 0.3909,
+      "step": 19430
+    },
+    {
+      "epoch": 7.338618346545866,
+      "grad_norm": 0.6767669916152954,
+      "learning_rate": 5.6434769606650864e-05,
+      "loss": 0.3908,
+      "step": 19440
+    },
+    {
+      "epoch": 7.34239335598339,
+      "grad_norm": 0.8901523351669312,
+      "learning_rate": 5.639377441611143e-05,
+      "loss": 0.3988,
+      "step": 19450
+    },
+    {
+      "epoch": 7.346168365420914,
+      "grad_norm": 0.6947163939476013,
+      "learning_rate": 5.635277485548751e-05,
+      "loss": 0.3917,
+      "step": 19460
+    },
+    {
+      "epoch": 7.349943374858437,
+      "grad_norm": 0.7375092506408691,
+      "learning_rate": 5.631177095280186e-05,
+      "loss": 0.3965,
+      "step": 19470
+    },
+    {
+      "epoch": 7.353718384295961,
+      "grad_norm": 1.315040111541748,
+      "learning_rate": 5.627076273608027e-05,
+      "loss": 0.3952,
+      "step": 19480
+    },
+    {
+      "epoch": 7.357493393733485,
+      "grad_norm": 0.6428912878036499,
+      "learning_rate": 5.622975023335148e-05,
+      "loss": 0.3936,
+      "step": 19490
+    },
+    {
+      "epoch": 7.361268403171008,
+      "grad_norm": 0.7309068441390991,
+      "learning_rate": 5.618873347264716e-05,
+      "loss": 0.3972,
+      "step": 19500
+    },
+    {
+      "epoch": 7.365043412608531,
+      "grad_norm": 0.6258814930915833,
+      "learning_rate": 5.614771248200188e-05,
+      "loss": 0.3858,
+      "step": 19510
+    },
+    {
+      "epoch": 7.368818422046055,
+      "grad_norm": 0.64792400598526,
+      "learning_rate": 5.6106687289453066e-05,
+      "loss": 0.398,
+      "step": 19520
+    },
+    {
+      "epoch": 7.372593431483579,
+      "grad_norm": 0.6821390986442566,
+      "learning_rate": 5.606565792304108e-05,
+      "loss": 0.4027,
+      "step": 19530
+    },
+    {
+      "epoch": 7.3763684409211026,
+      "grad_norm": 0.7018711566925049,
+      "learning_rate": 5.602462441080909e-05,
+      "loss": 0.3907,
+      "step": 19540
+    },
+    {
+      "epoch": 7.380143450358626,
+      "grad_norm": 0.6835519671440125,
+      "learning_rate": 5.5983586780803135e-05,
+      "loss": 0.3912,
+      "step": 19550
+    },
+    {
+      "epoch": 7.383918459796149,
+      "grad_norm": 0.6249481439590454,
+      "learning_rate": 5.594254506107205e-05,
+      "loss": 0.3903,
+      "step": 19560
+    },
+    {
+      "epoch": 7.387693469233673,
+      "grad_norm": 0.6832369565963745,
+      "learning_rate": 5.590149927966743e-05,
+      "loss": 0.3882,
+      "step": 19570
+    },
+    {
+      "epoch": 7.391468478671197,
+      "grad_norm": 0.6031800508499146,
+      "learning_rate": 5.58604494646437e-05,
+      "loss": 0.3954,
+      "step": 19580
+    },
+    {
+      "epoch": 7.39524348810872,
+      "grad_norm": 0.6349557638168335,
+      "learning_rate": 5.5819395644058025e-05,
+      "loss": 0.3894,
+      "step": 19590
+    },
+    {
+      "epoch": 7.399018497546244,
+      "grad_norm": 0.6155382990837097,
+      "learning_rate": 5.577833784597031e-05,
+      "loss": 0.3928,
+      "step": 19600
+    },
+    {
+      "epoch": 7.402793506983768,
+      "grad_norm": 0.6620678305625916,
+      "learning_rate": 5.573727609844316e-05,
+      "loss": 0.3906,
+      "step": 19610
+    },
+    {
+      "epoch": 7.406568516421291,
+      "grad_norm": 0.6608960628509521,
+      "learning_rate": 5.5696210429541884e-05,
+      "loss": 0.3975,
+      "step": 19620
+    },
+    {
+      "epoch": 7.410343525858814,
+      "grad_norm": 0.6048488020896912,
+      "learning_rate": 5.565514086733451e-05,
+      "loss": 0.3969,
+      "step": 19630
+    },
+    {
+      "epoch": 7.414118535296338,
+      "grad_norm": 0.7584668397903442,
+      "learning_rate": 5.5614067439891657e-05,
+      "loss": 0.3942,
+      "step": 19640
+    },
+    {
+      "epoch": 7.417893544733862,
+      "grad_norm": 0.6703759431838989,
+      "learning_rate": 5.557299017528666e-05,
+      "loss": 0.3879,
+      "step": 19650
+    },
+    {
+      "epoch": 7.421668554171386,
+      "grad_norm": 0.6314090490341187,
+      "learning_rate": 5.5531909101595436e-05,
+      "loss": 0.391,
+      "step": 19660
+    },
+    {
+      "epoch": 7.425443563608909,
+      "grad_norm": 0.6210800409317017,
+      "learning_rate": 5.549082424689649e-05,
+      "loss": 0.4007,
+      "step": 19670
+    },
+    {
+      "epoch": 7.429218573046432,
+      "grad_norm": 0.6025400161743164,
+      "learning_rate": 5.544973563927095e-05,
+      "loss": 0.3935,
+      "step": 19680
+    },
+    {
+      "epoch": 7.432993582483956,
+      "grad_norm": 0.7302088141441345,
+      "learning_rate": 5.540864330680249e-05,
+      "loss": 0.4046,
+      "step": 19690
+    },
+    {
+      "epoch": 7.43676859192148,
+      "grad_norm": 0.6146891117095947,
+      "learning_rate": 5.536754727757733e-05,
+      "loss": 0.396,
+      "step": 19700
+    },
+    {
+      "epoch": 7.440543601359003,
+      "grad_norm": 0.6504469513893127,
+      "learning_rate": 5.532644757968422e-05,
+      "loss": 0.3957,
+      "step": 19710
+    },
+    {
+      "epoch": 7.444318610796527,
+      "grad_norm": 0.6892192363739014,
+      "learning_rate": 5.528534424121441e-05,
+      "loss": 0.5172,
+      "step": 19720
+    },
+    {
+      "epoch": 7.448093620234051,
+      "grad_norm": 0.673173189163208,
+      "learning_rate": 5.524423729026165e-05,
+      "loss": 0.3959,
+      "step": 19730
+    },
+    {
+      "epoch": 7.451868629671575,
+      "grad_norm": 0.689117431640625,
+      "learning_rate": 5.5203126754922164e-05,
+      "loss": 0.3885,
+      "step": 19740
+    },
+    {
+      "epoch": 7.455643639109097,
+      "grad_norm": 0.6626030802726746,
+      "learning_rate": 5.5162012663294585e-05,
+      "loss": 0.3877,
+      "step": 19750
+    },
+    {
+      "epoch": 7.459418648546621,
+      "grad_norm": 0.7295408248901367,
+      "learning_rate": 5.512089504348003e-05,
+      "loss": 0.392,
+      "step": 19760
+    },
+    {
+      "epoch": 7.463193657984145,
+      "grad_norm": 0.6033636331558228,
+      "learning_rate": 5.5079773923582e-05,
+      "loss": 0.3939,
+      "step": 19770
+    },
+    {
+      "epoch": 7.466968667421669,
+      "grad_norm": 0.6206973791122437,
+      "learning_rate": 5.50386493317064e-05,
+      "loss": 0.3915,
+      "step": 19780
+    },
+    {
+      "epoch": 7.470743676859192,
+      "grad_norm": 0.6662564277648926,
+      "learning_rate": 5.49975212959615e-05,
+      "loss": 0.3923,
+      "step": 19790
+    },
+    {
+      "epoch": 7.474518686296716,
+      "grad_norm": 0.6757304072380066,
+      "learning_rate": 5.4956389844457904e-05,
+      "loss": 0.3936,
+      "step": 19800
+    },
+    {
+      "epoch": 7.478293695734239,
+      "grad_norm": 0.717755913734436,
+      "learning_rate": 5.491525500530859e-05,
+      "loss": 0.3932,
+      "step": 19810
+    },
+    {
+      "epoch": 7.482068705171763,
+      "grad_norm": 0.6577935218811035,
+      "learning_rate": 5.487411680662882e-05,
+      "loss": 0.3985,
+      "step": 19820
+    },
+    {
+      "epoch": 7.485843714609286,
+      "grad_norm": 0.6445197463035583,
+      "learning_rate": 5.483297527653618e-05,
+      "loss": 0.3979,
+      "step": 19830
+    },
+    {
+      "epoch": 7.48961872404681,
+      "grad_norm": 0.6646354794502258,
+      "learning_rate": 5.4791830443150516e-05,
+      "loss": 0.3896,
+      "step": 19840
+    },
+    {
+      "epoch": 7.493393733484334,
+      "grad_norm": 0.6168741583824158,
+      "learning_rate": 5.475068233459392e-05,
+      "loss": 0.398,
+      "step": 19850
+    },
+    {
+      "epoch": 7.497168742921858,
+      "grad_norm": 0.7372630834579468,
+      "learning_rate": 5.470953097899075e-05,
+      "loss": 0.3972,
+      "step": 19860
+    },
+    {
+      "epoch": 7.500943752359381,
+      "grad_norm": 0.7275758981704712,
+      "learning_rate": 5.466837640446756e-05,
+      "loss": 0.3923,
+      "step": 19870
+    },
+    {
+      "epoch": 7.504718761796904,
+      "grad_norm": 0.6699514389038086,
+      "learning_rate": 5.462721863915312e-05,
+      "loss": 0.3958,
+      "step": 19880
+    },
+    {
+      "epoch": 7.508493771234428,
+      "grad_norm": 0.6782812476158142,
+      "learning_rate": 5.4586057711178374e-05,
+      "loss": 0.3939,
+      "step": 19890
+    },
+    {
+      "epoch": 7.512268780671952,
+      "grad_norm": 0.6210011839866638,
+      "learning_rate": 5.454489364867642e-05,
+      "loss": 0.3935,
+      "step": 19900
+    },
+    {
+      "epoch": 7.516043790109475,
+      "grad_norm": 0.6767253279685974,
+      "learning_rate": 5.4503726479782523e-05,
+      "loss": 0.3922,
+      "step": 19910
+    },
+    {
+      "epoch": 7.519818799546999,
+      "grad_norm": 0.5947399735450745,
+      "learning_rate": 5.446255623263403e-05,
+      "loss": 0.3867,
+      "step": 19920
+    },
+    {
+      "epoch": 7.523593808984522,
+      "grad_norm": 0.6279717683792114,
+      "learning_rate": 5.4421382935370445e-05,
+      "loss": 0.3933,
+      "step": 19930
+    },
+    {
+      "epoch": 7.527368818422046,
+      "grad_norm": 0.6436152458190918,
+      "learning_rate": 5.438020661613331e-05,
+      "loss": 0.3879,
+      "step": 19940
+    },
+    {
+      "epoch": 7.5311438278595695,
+      "grad_norm": 0.6227508187294006,
+      "learning_rate": 5.433902730306625e-05,
+      "loss": 0.3899,
+      "step": 19950
+    },
+    {
+      "epoch": 7.534918837297093,
+      "grad_norm": 0.7720093131065369,
+      "learning_rate": 5.429784502431495e-05,
+      "loss": 0.3946,
+      "step": 19960
+    },
+    {
+      "epoch": 7.538693846734617,
+      "grad_norm": 0.7744880318641663,
+      "learning_rate": 5.42566598080271e-05,
+      "loss": 0.3946,
+      "step": 19970
+    },
+    {
+      "epoch": 7.542468856172141,
+      "grad_norm": 1.112042784690857,
+      "learning_rate": 5.421547168235241e-05,
+      "loss": 0.392,
+      "step": 19980
+    },
+    {
+      "epoch": 7.546243865609664,
+      "grad_norm": 0.6036012172698975,
+      "learning_rate": 5.417428067544258e-05,
+      "loss": 0.3993,
+      "step": 19990
+    },
+    {
+      "epoch": 7.550018875047187,
+      "grad_norm": 0.6224648952484131,
+      "learning_rate": 5.413308681545126e-05,
+      "loss": 0.3947,
+      "step": 20000
+    },
+    {
+      "epoch": 7.553793884484711,
+      "grad_norm": 0.6193528771400452,
+      "learning_rate": 5.409189013053408e-05,
+      "loss": 0.3918,
+      "step": 20010
+    },
+    {
+      "epoch": 7.557568893922235,
+      "grad_norm": 0.5908026099205017,
+      "learning_rate": 5.4050690648848576e-05,
+      "loss": 0.3911,
+      "step": 20020
+    },
+    {
+      "epoch": 7.5613439033597585,
+      "grad_norm": 0.6350475549697876,
+      "learning_rate": 5.400948839855421e-05,
+      "loss": 0.3936,
+      "step": 20030
+    },
+    {
+      "epoch": 7.565118912797282,
+      "grad_norm": 0.6277138590812683,
+      "learning_rate": 5.396828340781234e-05,
+      "loss": 0.3937,
+      "step": 20040
+    },
+    {
+      "epoch": 7.568893922234806,
+      "grad_norm": 0.8242809176445007,
+      "learning_rate": 5.392707570478617e-05,
+      "loss": 0.407,
+      "step": 20050
+    },
+    {
+      "epoch": 7.572668931672329,
+      "grad_norm": 0.7098134756088257,
+      "learning_rate": 5.388586531764078e-05,
+      "loss": 0.3781,
+      "step": 20060
+    },
+    {
+      "epoch": 7.5764439411098525,
+      "grad_norm": 0.9367830753326416,
+      "learning_rate": 5.384465227454311e-05,
+      "loss": 0.3971,
+      "step": 20070
+    },
+    {
+      "epoch": 7.580218950547376,
+      "grad_norm": 0.5797680616378784,
+      "learning_rate": 5.380343660366184e-05,
+      "loss": 0.3923,
+      "step": 20080
+    },
+    {
+      "epoch": 7.5839939599849,
+      "grad_norm": 0.6200445890426636,
+      "learning_rate": 5.376221833316752e-05,
+      "loss": 0.3958,
+      "step": 20090
+    },
+    {
+      "epoch": 7.587768969422424,
+      "grad_norm": 0.651992678642273,
+      "learning_rate": 5.3720997491232436e-05,
+      "loss": 0.3893,
+      "step": 20100
+    },
+    {
+      "epoch": 7.5915439788599475,
+      "grad_norm": 0.6731122732162476,
+      "learning_rate": 5.367977410603068e-05,
+      "loss": 0.3918,
+      "step": 20110
+    },
+    {
+      "epoch": 7.595318988297471,
+      "grad_norm": 0.5849369764328003,
+      "learning_rate": 5.3638548205738004e-05,
+      "loss": 0.3909,
+      "step": 20120
+    },
+    {
+      "epoch": 7.599093997734994,
+      "grad_norm": 0.6614940762519836,
+      "learning_rate": 5.359731981853194e-05,
+      "loss": 0.3949,
+      "step": 20130
+    },
+    {
+      "epoch": 7.602869007172518,
+      "grad_norm": 3.908480644226074,
+      "learning_rate": 5.35560889725917e-05,
+      "loss": 0.3875,
+      "step": 20140
+    },
+    {
+      "epoch": 7.6066440166100415,
+      "grad_norm": 0.6389253735542297,
+      "learning_rate": 5.3514855696098176e-05,
+      "loss": 0.3883,
+      "step": 20150
+    },
+    {
+      "epoch": 7.610419026047565,
+      "grad_norm": 0.6248882412910461,
+      "learning_rate": 5.347362001723394e-05,
+      "loss": 0.3848,
+      "step": 20160
+    },
+    {
+      "epoch": 7.614194035485089,
+      "grad_norm": 0.6055360436439514,
+      "learning_rate": 5.3432381964183176e-05,
+      "loss": 0.3936,
+      "step": 20170
+    },
+    {
+      "epoch": 7.617969044922612,
+      "grad_norm": 0.7338094115257263,
+      "learning_rate": 5.3391141565131685e-05,
+      "loss": 0.3904,
+      "step": 20180
+    },
+    {
+      "epoch": 7.621744054360136,
+      "grad_norm": 0.8716061115264893,
+      "learning_rate": 5.3349898848266935e-05,
+      "loss": 0.3902,
+      "step": 20190
+    },
+    {
+      "epoch": 7.625519063797659,
+      "grad_norm": 1.380915641784668,
+      "learning_rate": 5.330865384177789e-05,
+      "loss": 0.3927,
+      "step": 20200
+    },
+    {
+      "epoch": 7.629294073235183,
+      "grad_norm": 0.6246406435966492,
+      "learning_rate": 5.326740657385515e-05,
+      "loss": 0.389,
+      "step": 20210
+    },
+    {
+      "epoch": 7.633069082672707,
+      "grad_norm": 0.6647712588310242,
+      "learning_rate": 5.322615707269083e-05,
+      "loss": 0.3956,
+      "step": 20220
+    },
+    {
+      "epoch": 7.6368440921102305,
+      "grad_norm": 0.6063628792762756,
+      "learning_rate": 5.318490536647856e-05,
+      "loss": 0.3887,
+      "step": 20230
+    },
+    {
+      "epoch": 7.640619101547754,
+      "grad_norm": 0.625741720199585,
+      "learning_rate": 5.3143651483413524e-05,
+      "loss": 0.3936,
+      "step": 20240
+    },
+    {
+      "epoch": 7.644394110985277,
+      "grad_norm": 0.676576554775238,
+      "learning_rate": 5.310239545169232e-05,
+      "loss": 0.3916,
+      "step": 20250
+    },
+    {
+      "epoch": 7.648169120422801,
+      "grad_norm": 0.5982394218444824,
+      "learning_rate": 5.30611372995131e-05,
+      "loss": 0.3868,
+      "step": 20260
+    },
+    {
+      "epoch": 7.651944129860325,
+      "grad_norm": 0.6712155342102051,
+      "learning_rate": 5.30198770550754e-05,
+      "loss": 0.3903,
+      "step": 20270
+    },
+    {
+      "epoch": 7.655719139297848,
+      "grad_norm": 0.5862340331077576,
+      "learning_rate": 5.297861474658019e-05,
+      "loss": 0.389,
+      "step": 20280
+    },
+    {
+      "epoch": 7.659494148735372,
+      "grad_norm": 0.6452902555465698,
+      "learning_rate": 5.29373504022299e-05,
+      "loss": 0.3847,
+      "step": 20290
+    },
+    {
+      "epoch": 7.663269158172896,
+      "grad_norm": 0.6566249132156372,
+      "learning_rate": 5.28960840502283e-05,
+      "loss": 0.3969,
+      "step": 20300
+    },
+    {
+      "epoch": 7.667044167610419,
+      "grad_norm": 0.6458132863044739,
+      "learning_rate": 5.285481571878056e-05,
+      "loss": 0.3942,
+      "step": 20310
+    },
+    {
+      "epoch": 7.670819177047942,
+      "grad_norm": 0.6279521584510803,
+      "learning_rate": 5.281354543609321e-05,
+      "loss": 0.3903,
+      "step": 20320
+    },
+    {
+      "epoch": 7.674594186485466,
+      "grad_norm": 0.5799058079719543,
+      "learning_rate": 5.277227323037406e-05,
+      "loss": 0.3871,
+      "step": 20330
+    },
+    {
+      "epoch": 7.67836919592299,
+      "grad_norm": 0.5981761813163757,
+      "learning_rate": 5.273099912983233e-05,
+      "loss": 0.3854,
+      "step": 20340
+    },
+    {
+      "epoch": 7.682144205360514,
+      "grad_norm": 0.6565678119659424,
+      "learning_rate": 5.268972316267843e-05,
+      "loss": 0.3878,
+      "step": 20350
+    },
+    {
+      "epoch": 7.685919214798037,
+      "grad_norm": 0.6143926382064819,
+      "learning_rate": 5.26484453571241e-05,
+      "loss": 0.3997,
+      "step": 20360
+    },
+    {
+      "epoch": 7.689694224235561,
+      "grad_norm": 0.630825936794281,
+      "learning_rate": 5.260716574138235e-05,
+      "loss": 0.3865,
+      "step": 20370
+    },
+    {
+      "epoch": 7.693469233673084,
+      "grad_norm": 0.636688768863678,
+      "learning_rate": 5.256588434366739e-05,
+      "loss": 0.3895,
+      "step": 20380
+    },
+    {
+      "epoch": 7.697244243110608,
+      "grad_norm": 0.675250768661499,
+      "learning_rate": 5.25246011921947e-05,
+      "loss": 0.3901,
+      "step": 20390
+    },
+    {
+      "epoch": 7.701019252548131,
+      "grad_norm": 0.6246991157531738,
+      "learning_rate": 5.248331631518089e-05,
+      "loss": 0.385,
+      "step": 20400
+    },
+    {
+      "epoch": 7.704794261985655,
+      "grad_norm": 0.6083090901374817,
+      "learning_rate": 5.244202974084379e-05,
+      "loss": 0.3906,
+      "step": 20410
+    },
+    {
+      "epoch": 7.708569271423179,
+      "grad_norm": 0.63591068983078,
+      "learning_rate": 5.240074149740239e-05,
+      "loss": 0.3842,
+      "step": 20420
+    },
+    {
+      "epoch": 7.712344280860702,
+      "grad_norm": 0.7097567319869995,
+      "learning_rate": 5.2359451613076814e-05,
+      "loss": 0.3777,
+      "step": 20430
+    },
+    {
+      "epoch": 7.716119290298225,
+      "grad_norm": 0.6984173059463501,
+      "learning_rate": 5.231816011608832e-05,
+      "loss": 0.3889,
+      "step": 20440
+    },
+    {
+      "epoch": 7.719894299735749,
+      "grad_norm": 0.6347232460975647,
+      "learning_rate": 5.227686703465924e-05,
+      "loss": 0.3888,
+      "step": 20450
+    },
+    {
+      "epoch": 7.723669309173273,
+      "grad_norm": 0.5677266120910645,
+      "learning_rate": 5.2235572397013e-05,
+      "loss": 0.3899,
+      "step": 20460
+    },
+    {
+      "epoch": 7.727444318610797,
+      "grad_norm": 0.579582929611206,
+      "learning_rate": 5.2194276231374114e-05,
+      "loss": 0.3896,
+      "step": 20470
+    },
+    {
+      "epoch": 7.73121932804832,
+      "grad_norm": 0.5820940732955933,
+      "learning_rate": 5.21529785659681e-05,
+      "loss": 0.3921,
+      "step": 20480
+    },
+    {
+      "epoch": 7.734994337485844,
+      "grad_norm": 0.6141514778137207,
+      "learning_rate": 5.2111679429021565e-05,
+      "loss": 0.3986,
+      "step": 20490
+    },
+    {
+      "epoch": 7.738769346923367,
+      "grad_norm": 0.6320546269416809,
+      "learning_rate": 5.207037884876205e-05,
+      "loss": 0.3882,
+      "step": 20500
+    },
+    {
+      "epoch": 7.742544356360891,
+      "grad_norm": 0.6534587740898132,
+      "learning_rate": 5.202907685341809e-05,
+      "loss": 0.3885,
+      "step": 20510
+    },
+    {
+      "epoch": 7.746319365798414,
+      "grad_norm": 0.8131640553474426,
+      "learning_rate": 5.198777347121926e-05,
+      "loss": 0.3791,
+      "step": 20520
+    },
+    {
+      "epoch": 7.750094375235938,
+      "grad_norm": 0.5980712175369263,
+      "learning_rate": 5.194646873039598e-05,
+      "loss": 0.3823,
+      "step": 20530
+    },
+    {
+      "epoch": 7.753869384673462,
+      "grad_norm": 0.6374219059944153,
+      "learning_rate": 5.1905162659179696e-05,
+      "loss": 0.3967,
+      "step": 20540
+    },
+    {
+      "epoch": 7.757644394110986,
+      "grad_norm": 0.6131422519683838,
+      "learning_rate": 5.18638552858027e-05,
+      "loss": 0.3865,
+      "step": 20550
+    },
+    {
+      "epoch": 7.7614194035485085,
+      "grad_norm": 1.034074306488037,
+      "learning_rate": 5.182254663849818e-05,
+      "loss": 0.3907,
+      "step": 20560
+    },
+    {
+      "epoch": 7.765194412986032,
+      "grad_norm": 0.6826948523521423,
+      "learning_rate": 5.178123674550023e-05,
+      "loss": 0.3857,
+      "step": 20570
+    },
+    {
+      "epoch": 7.768969422423556,
+      "grad_norm": 0.5983694195747375,
+      "learning_rate": 5.173992563504375e-05,
+      "loss": 0.3914,
+      "step": 20580
+    },
+    {
+      "epoch": 7.77274443186108,
+      "grad_norm": 0.5786053538322449,
+      "learning_rate": 5.169861333536451e-05,
+      "loss": 0.3863,
+      "step": 20590
+    },
+    {
+      "epoch": 7.776519441298603,
+      "grad_norm": 0.607843816280365,
+      "learning_rate": 5.165729987469907e-05,
+      "loss": 0.3928,
+      "step": 20600
+    },
+    {
+      "epoch": 7.780294450736127,
+      "grad_norm": 0.597486138343811,
+      "learning_rate": 5.161598528128478e-05,
+      "loss": 0.3845,
+      "step": 20610
+    },
+    {
+      "epoch": 7.784069460173651,
+      "grad_norm": 0.6441140174865723,
+      "learning_rate": 5.157466958335981e-05,
+      "loss": 0.3856,
+      "step": 20620
+    },
+    {
+      "epoch": 7.787844469611174,
+      "grad_norm": 0.8996423482894897,
+      "learning_rate": 5.1533352809163025e-05,
+      "loss": 0.3818,
+      "step": 20630
+    },
+    {
+      "epoch": 7.7916194790486974,
+      "grad_norm": 0.6202008724212646,
+      "learning_rate": 5.1492034986934046e-05,
+      "loss": 0.3865,
+      "step": 20640
+    },
+    {
+      "epoch": 7.795394488486221,
+      "grad_norm": 0.6499298810958862,
+      "learning_rate": 5.1450716144913225e-05,
+      "loss": 0.3887,
+      "step": 20650
+    },
+    {
+      "epoch": 7.799169497923745,
+      "grad_norm": 0.6905611753463745,
+      "learning_rate": 5.1409396311341595e-05,
+      "loss": 0.3841,
+      "step": 20660
+    },
+    {
+      "epoch": 7.802944507361269,
+      "grad_norm": 0.6405202150344849,
+      "learning_rate": 5.136807551446089e-05,
+      "loss": 0.3883,
+      "step": 20670
+    },
+    {
+      "epoch": 7.8067195167987915,
+      "grad_norm": 0.6478602290153503,
+      "learning_rate": 5.132675378251346e-05,
+      "loss": 0.3856,
+      "step": 20680
+    },
+    {
+      "epoch": 7.810494526236315,
+      "grad_norm": 0.6184169054031372,
+      "learning_rate": 5.1285431143742325e-05,
+      "loss": 0.3901,
+      "step": 20690
+    },
+    {
+      "epoch": 7.814269535673839,
+      "grad_norm": 0.6250163316726685,
+      "learning_rate": 5.1244107626391136e-05,
+      "loss": 0.4039,
+      "step": 20700
+    },
+    {
+      "epoch": 7.818044545111363,
+      "grad_norm": 0.6539636254310608,
+      "learning_rate": 5.12027832587041e-05,
+      "loss": 0.3866,
+      "step": 20710
+    },
+    {
+      "epoch": 7.821819554548886,
+      "grad_norm": 0.6491449475288391,
+      "learning_rate": 5.116145806892607e-05,
+      "loss": 0.3965,
+      "step": 20720
+    },
+    {
+      "epoch": 7.82559456398641,
+      "grad_norm": 0.653968870639801,
+      "learning_rate": 5.1120132085302384e-05,
+      "loss": 0.3825,
+      "step": 20730
+    },
+    {
+      "epoch": 7.829369573423934,
+      "grad_norm": 0.7389346957206726,
+      "learning_rate": 5.107880533607898e-05,
+      "loss": 0.3935,
+      "step": 20740
+    },
+    {
+      "epoch": 7.833144582861458,
+      "grad_norm": 0.6781933903694153,
+      "learning_rate": 5.103747784950231e-05,
+      "loss": 0.3852,
+      "step": 20750
+    },
+    {
+      "epoch": 7.8369195922989805,
+      "grad_norm": 0.6142136454582214,
+      "learning_rate": 5.09961496538193e-05,
+      "loss": 0.3956,
+      "step": 20760
+    },
+    {
+      "epoch": 7.840694601736504,
+      "grad_norm": 0.5901447534561157,
+      "learning_rate": 5.095482077727742e-05,
+      "loss": 0.3849,
+      "step": 20770
+    },
+    {
+      "epoch": 7.844469611174028,
+      "grad_norm": 0.6106001138687134,
+      "learning_rate": 5.091349124812452e-05,
+      "loss": 0.3852,
+      "step": 20780
+    },
+    {
+      "epoch": 7.848244620611552,
+      "grad_norm": 0.7389256954193115,
+      "learning_rate": 5.087216109460897e-05,
+      "loss": 0.3912,
+      "step": 20790
+    },
+    {
+      "epoch": 7.852019630049075,
+      "grad_norm": 0.7008233070373535,
+      "learning_rate": 5.083083034497954e-05,
+      "loss": 0.3856,
+      "step": 20800
+    },
+    {
+      "epoch": 7.855794639486598,
+      "grad_norm": 0.5946593880653381,
+      "learning_rate": 5.07894990274854e-05,
+      "loss": 0.3878,
+      "step": 20810
+    },
+    {
+      "epoch": 7.859569648924122,
+      "grad_norm": 0.5986919403076172,
+      "learning_rate": 5.074816717037614e-05,
+      "loss": 0.3842,
+      "step": 20820
+    },
+    {
+      "epoch": 7.863344658361646,
+      "grad_norm": 0.7096017003059387,
+      "learning_rate": 5.070683480190165e-05,
+      "loss": 0.3923,
+      "step": 20830
+    },
+    {
+      "epoch": 7.8671196677991695,
+      "grad_norm": 0.6921922564506531,
+      "learning_rate": 5.066550195031223e-05,
+      "loss": 0.3864,
+      "step": 20840
+    },
+    {
+      "epoch": 7.870894677236693,
+      "grad_norm": 0.5950612425804138,
+      "learning_rate": 5.062416864385852e-05,
+      "loss": 0.384,
+      "step": 20850
+    },
+    {
+      "epoch": 7.874669686674217,
+      "grad_norm": 0.6273109912872314,
+      "learning_rate": 5.058283491079142e-05,
+      "loss": 0.3778,
+      "step": 20860
+    },
+    {
+      "epoch": 7.878444696111741,
+      "grad_norm": 0.7356699109077454,
+      "learning_rate": 5.054150077936216e-05,
+      "loss": 0.3905,
+      "step": 20870
+    },
+    {
+      "epoch": 7.8822197055492635,
+      "grad_norm": 0.6300787329673767,
+      "learning_rate": 5.0500166277822214e-05,
+      "loss": 0.387,
+      "step": 20880
+    },
+    {
+      "epoch": 7.885994714986787,
+      "grad_norm": 0.6487736105918884,
+      "learning_rate": 5.0458831434423334e-05,
+      "loss": 0.3935,
+      "step": 20890
+    },
+    {
+      "epoch": 7.889769724424311,
+      "grad_norm": 0.6694430708885193,
+      "learning_rate": 5.0417496277417506e-05,
+      "loss": 0.3982,
+      "step": 20900
+    },
+    {
+      "epoch": 7.893544733861835,
+      "grad_norm": 0.6112816333770752,
+      "learning_rate": 5.037616083505691e-05,
+      "loss": 0.3769,
+      "step": 20910
+    },
+    {
+      "epoch": 7.8973197432993585,
+      "grad_norm": 0.5769428014755249,
+      "learning_rate": 5.0334825135593935e-05,
+      "loss": 0.387,
+      "step": 20920
+    },
+    {
+      "epoch": 7.901094752736882,
+      "grad_norm": 0.623289167881012,
+      "learning_rate": 5.029348920728111e-05,
+      "loss": 0.3874,
+      "step": 20930
+    },
+    {
+      "epoch": 7.904869762174405,
+      "grad_norm": 0.6469810605049133,
+      "learning_rate": 5.0252153078371186e-05,
+      "loss": 0.3863,
+      "step": 20940
+    },
+    {
+      "epoch": 7.908644771611929,
+      "grad_norm": 0.6396733522415161,
+      "learning_rate": 5.021081677711704e-05,
+      "loss": 0.3933,
+      "step": 20950
+    },
+    {
+      "epoch": 7.9124197810494525,
+      "grad_norm": 0.6472081542015076,
+      "learning_rate": 5.016948033177159e-05,
+      "loss": 0.3804,
+      "step": 20960
+    },
+    {
+      "epoch": 7.916194790486976,
+      "grad_norm": 0.670315146446228,
+      "learning_rate": 5.012814377058793e-05,
+      "loss": 0.3922,
+      "step": 20970
+    },
+    {
+      "epoch": 7.9199697999245,
+      "grad_norm": 0.6452091932296753,
+      "learning_rate": 5.008680712181921e-05,
+      "loss": 0.3906,
+      "step": 20980
+    },
+    {
+      "epoch": 7.923744809362024,
+      "grad_norm": 0.6105216145515442,
+      "learning_rate": 5.0045470413718645e-05,
+      "loss": 0.3864,
+      "step": 20990
+    },
+    {
+      "epoch": 7.9275198187995475,
+      "grad_norm": 0.6681371331214905,
+      "learning_rate": 5.00041336745395e-05,
+      "loss": 0.3806,
+      "step": 21000
+    },
+    {
+      "epoch": 7.93129482823707,
+      "grad_norm": 0.6200874447822571,
+      "learning_rate": 4.996279693253499e-05,
+      "loss": 0.3895,
+      "step": 21010
+    },
+    {
+      "epoch": 7.935069837674594,
+      "grad_norm": 0.620955765247345,
+      "learning_rate": 4.992146021595847e-05,
+      "loss": 0.3872,
+      "step": 21020
+    },
+    {
+      "epoch": 7.938844847112118,
+      "grad_norm": 0.6408780217170715,
+      "learning_rate": 4.988012355306313e-05,
+      "loss": 0.3821,
+      "step": 21030
+    },
+    {
+      "epoch": 7.9426198565496415,
+      "grad_norm": 0.6567278504371643,
+      "learning_rate": 4.98387869721022e-05,
+      "loss": 0.3987,
+      "step": 21040
+    },
+    {
+      "epoch": 7.946394865987165,
+      "grad_norm": 0.6321324706077576,
+      "learning_rate": 4.9797450501328866e-05,
+      "loss": 0.382,
+      "step": 21050
+    },
+    {
+      "epoch": 7.950169875424688,
+      "grad_norm": 0.6207996606826782,
+      "learning_rate": 4.97561141689962e-05,
+      "loss": 0.3921,
+      "step": 21060
+    },
+    {
+      "epoch": 7.953944884862212,
+      "grad_norm": 0.6274701952934265,
+      "learning_rate": 4.971477800335721e-05,
+      "loss": 0.3803,
+      "step": 21070
+    },
+    {
+      "epoch": 7.957719894299736,
+      "grad_norm": 0.6380013227462769,
+      "learning_rate": 4.967344203266475e-05,
+      "loss": 0.3857,
+      "step": 21080
+    },
+    {
+      "epoch": 7.961494903737259,
+      "grad_norm": 0.6619207859039307,
+      "learning_rate": 4.9632106285171584e-05,
+      "loss": 0.3828,
+      "step": 21090
+    },
+    {
+      "epoch": 7.965269913174783,
+      "grad_norm": 0.6286654472351074,
+      "learning_rate": 4.959077078913031e-05,
+      "loss": 0.3857,
+      "step": 21100
+    },
+    {
+      "epoch": 7.969044922612307,
+      "grad_norm": 0.7368075847625732,
+      "learning_rate": 4.954943557279333e-05,
+      "loss": 0.3843,
+      "step": 21110
+    },
+    {
+      "epoch": 7.9728199320498305,
+      "grad_norm": 0.6597092747688293,
+      "learning_rate": 4.9508100664412916e-05,
+      "loss": 0.3883,
+      "step": 21120
+    },
+    {
+      "epoch": 7.976594941487353,
+      "grad_norm": 0.673981249332428,
+      "learning_rate": 4.946676609224105e-05,
+      "loss": 0.3855,
+      "step": 21130
+    },
+    {
+      "epoch": 7.980369950924877,
+      "grad_norm": 0.6356333494186401,
+      "learning_rate": 4.942543188452952e-05,
+      "loss": 0.3922,
+      "step": 21140
+    },
+    {
+      "epoch": 7.984144960362401,
+      "grad_norm": 0.6984882354736328,
+      "learning_rate": 4.938409806952988e-05,
+      "loss": 0.3834,
+      "step": 21150
+    },
+    {
+      "epoch": 7.987919969799925,
+      "grad_norm": 0.6634735465049744,
+      "learning_rate": 4.93427646754934e-05,
+      "loss": 0.3838,
+      "step": 21160
+    },
+    {
+      "epoch": 7.991694979237448,
+      "grad_norm": 0.6481113433837891,
+      "learning_rate": 4.930143173067108e-05,
+      "loss": 0.3863,
+      "step": 21170
+    },
+    {
+      "epoch": 7.995469988674972,
+      "grad_norm": 0.684773325920105,
+      "learning_rate": 4.9260099263313565e-05,
+      "loss": 0.379,
+      "step": 21180
+    },
+    {
+      "epoch": 7.999244998112495,
+      "grad_norm": 0.6397860646247864,
+      "learning_rate": 4.921876730167123e-05,
+      "loss": 0.3962,
+      "step": 21190
+    },
+    {
+      "epoch": 8.003020007550019,
+      "grad_norm": 0.6320487260818481,
+      "learning_rate": 4.917743587399409e-05,
+      "loss": 0.3839,
+      "step": 21200
+    },
+    {
+      "epoch": 8.006795016987542,
+      "grad_norm": 0.6578797101974487,
+      "learning_rate": 4.913610500853178e-05,
+      "loss": 0.3841,
+      "step": 21210
+    },
+    {
+      "epoch": 8.010570026425066,
+      "grad_norm": 0.6178431510925293,
+      "learning_rate": 4.909477473353354e-05,
+      "loss": 0.3848,
+      "step": 21220
+    },
+    {
+      "epoch": 8.01434503586259,
+      "grad_norm": 0.6298217177391052,
+      "learning_rate": 4.9053445077248236e-05,
+      "loss": 0.3876,
+      "step": 21230
+    },
+    {
+      "epoch": 8.018120045300114,
+      "grad_norm": 0.6131578087806702,
+      "learning_rate": 4.901211606792429e-05,
+      "loss": 0.3914,
+      "step": 21240
+    },
+    {
+      "epoch": 8.021895054737637,
+      "grad_norm": 0.6626631021499634,
+      "learning_rate": 4.89707877338097e-05,
+      "loss": 0.3864,
+      "step": 21250
+    },
+    {
+      "epoch": 8.025670064175161,
+      "grad_norm": 0.6320914626121521,
+      "learning_rate": 4.892946010315199e-05,
+      "loss": 0.3853,
+      "step": 21260
+    },
+    {
+      "epoch": 8.029445073612685,
+      "grad_norm": 0.657503068447113,
+      "learning_rate": 4.8888133204198204e-05,
+      "loss": 0.3864,
+      "step": 21270
+    },
+    {
+      "epoch": 8.033220083050207,
+      "grad_norm": 0.6045080423355103,
+      "learning_rate": 4.8846807065194886e-05,
+      "loss": 0.3859,
+      "step": 21280
+    },
+    {
+      "epoch": 8.03699509248773,
+      "grad_norm": 0.6179243326187134,
+      "learning_rate": 4.880548171438806e-05,
+      "loss": 0.39,
+      "step": 21290
+    },
+    {
+      "epoch": 8.040770101925254,
+      "grad_norm": 0.7276541590690613,
+      "learning_rate": 4.8764157180023245e-05,
+      "loss": 0.3869,
+      "step": 21300
+    },
+    {
+      "epoch": 8.044545111362778,
+      "grad_norm": 0.5852580666542053,
+      "learning_rate": 4.872283349034533e-05,
+      "loss": 0.3847,
+      "step": 21310
+    },
+    {
+      "epoch": 8.048320120800302,
+      "grad_norm": 0.8796269297599792,
+      "learning_rate": 4.8681510673598674e-05,
+      "loss": 0.3854,
+      "step": 21320
+    },
+    {
+      "epoch": 8.052095130237825,
+      "grad_norm": 0.5888720750808716,
+      "learning_rate": 4.8640188758027046e-05,
+      "loss": 0.3767,
+      "step": 21330
+    },
+    {
+      "epoch": 8.05587013967535,
+      "grad_norm": 0.6211332678794861,
+      "learning_rate": 4.859886777187357e-05,
+      "loss": 0.392,
+      "step": 21340
+    },
+    {
+      "epoch": 8.059645149112873,
+      "grad_norm": 0.6416304111480713,
+      "learning_rate": 4.855754774338077e-05,
+      "loss": 0.3906,
+      "step": 21350
+    },
+    {
+      "epoch": 8.063420158550397,
+      "grad_norm": 0.6397213339805603,
+      "learning_rate": 4.851622870079048e-05,
+      "loss": 0.3865,
+      "step": 21360
+    },
+    {
+      "epoch": 8.06719516798792,
+      "grad_norm": 0.655972957611084,
+      "learning_rate": 4.847491067234389e-05,
+      "loss": 0.3845,
+      "step": 21370
+    },
+    {
+      "epoch": 8.070970177425444,
+      "grad_norm": 0.7151166796684265,
+      "learning_rate": 4.843359368628146e-05,
+      "loss": 0.379,
+      "step": 21380
+    },
+    {
+      "epoch": 8.074745186862968,
+      "grad_norm": 0.7403479814529419,
+      "learning_rate": 4.8392277770842975e-05,
+      "loss": 0.3857,
+      "step": 21390
+    },
+    {
+      "epoch": 8.078520196300492,
+      "grad_norm": 0.8259317278862,
+      "learning_rate": 4.83509629542675e-05,
+      "loss": 0.3833,
+      "step": 21400
+    },
+    {
+      "epoch": 8.082295205738014,
+      "grad_norm": 0.5347198843955994,
+      "learning_rate": 4.830964926479329e-05,
+      "loss": 0.3884,
+      "step": 21410
+    },
+    {
+      "epoch": 8.086070215175537,
+      "grad_norm": 0.6426698565483093,
+      "learning_rate": 4.826833673065785e-05,
+      "loss": 0.3822,
+      "step": 21420
+    },
+    {
+      "epoch": 8.089845224613061,
+      "grad_norm": 0.6192202568054199,
+      "learning_rate": 4.822702538009794e-05,
+      "loss": 0.386,
+      "step": 21430
+    },
+    {
+      "epoch": 8.093620234050585,
+      "grad_norm": 0.622089684009552,
+      "learning_rate": 4.818571524134945e-05,
+      "loss": 0.3854,
+      "step": 21440
+    },
+    {
+      "epoch": 8.097395243488108,
+      "grad_norm": 0.6114779114723206,
+      "learning_rate": 4.8144406342647496e-05,
+      "loss": 0.391,
+      "step": 21450
+    },
+    {
+      "epoch": 8.101170252925632,
+      "grad_norm": 0.6789860129356384,
+      "learning_rate": 4.81030987122263e-05,
+      "loss": 0.3818,
+      "step": 21460
+    },
+    {
+      "epoch": 8.104945262363156,
+      "grad_norm": 0.6163697242736816,
+      "learning_rate": 4.806179237831926e-05,
+      "loss": 0.3791,
+      "step": 21470
+    },
+    {
+      "epoch": 8.10872027180068,
+      "grad_norm": 0.7106571793556213,
+      "learning_rate": 4.802048736915884e-05,
+      "loss": 0.3811,
+      "step": 21480
+    },
+    {
+      "epoch": 8.112495281238203,
+      "grad_norm": 0.6782732009887695,
+      "learning_rate": 4.797918371297666e-05,
+      "loss": 0.3845,
+      "step": 21490
+    },
+    {
+      "epoch": 8.116270290675727,
+      "grad_norm": 0.662436842918396,
+      "learning_rate": 4.793788143800334e-05,
+      "loss": 0.39,
+      "step": 21500
+    },
+    {
+      "epoch": 8.12004530011325,
+      "grad_norm": 0.6009316444396973,
+      "learning_rate": 4.789658057246862e-05,
+      "loss": 0.3929,
+      "step": 21510
+    },
+    {
+      "epoch": 8.123820309550775,
+      "grad_norm": 0.628048300743103,
+      "learning_rate": 4.7855281144601227e-05,
+      "loss": 0.386,
+      "step": 21520
+    },
+    {
+      "epoch": 8.127595318988298,
+      "grad_norm": 0.6932970881462097,
+      "learning_rate": 4.781398318262897e-05,
+      "loss": 0.3889,
+      "step": 21530
+    },
+    {
+      "epoch": 8.13137032842582,
+      "grad_norm": 0.6603420376777649,
+      "learning_rate": 4.777268671477858e-05,
+      "loss": 0.3847,
+      "step": 21540
+    },
+    {
+      "epoch": 8.135145337863344,
+      "grad_norm": 0.7748346328735352,
+      "learning_rate": 4.773139176927582e-05,
+      "loss": 0.3916,
+      "step": 21550
+    },
+    {
+      "epoch": 8.138920347300868,
+      "grad_norm": 0.7710997462272644,
+      "learning_rate": 4.769009837434539e-05,
+      "loss": 0.3834,
+      "step": 21560
+    },
+    {
+      "epoch": 8.142695356738392,
+      "grad_norm": 0.6665760278701782,
+      "learning_rate": 4.764880655821095e-05,
+      "loss": 0.3885,
+      "step": 21570
+    },
+    {
+      "epoch": 8.146470366175915,
+      "grad_norm": 0.6791707873344421,
+      "learning_rate": 4.760751634909508e-05,
+      "loss": 0.3798,
+      "step": 21580
+    },
+    {
+      "epoch": 8.150245375613439,
+      "grad_norm": 0.6167095899581909,
+      "learning_rate": 4.756622777521919e-05,
+      "loss": 0.3799,
+      "step": 21590
+    },
+    {
+      "epoch": 8.154020385050963,
+      "grad_norm": 0.7474955320358276,
+      "learning_rate": 4.752494086480368e-05,
+      "loss": 0.3873,
+      "step": 21600
+    },
+    {
+      "epoch": 8.157795394488486,
+      "grad_norm": 0.6586377024650574,
+      "learning_rate": 4.7483655646067744e-05,
+      "loss": 0.386,
+      "step": 21610
+    },
+    {
+      "epoch": 8.16157040392601,
+      "grad_norm": 0.6137120127677917,
+      "learning_rate": 4.744237214722944e-05,
+      "loss": 0.3842,
+      "step": 21620
+    },
+    {
+      "epoch": 8.165345413363534,
+      "grad_norm": 0.613304615020752,
+      "learning_rate": 4.740109039650567e-05,
+      "loss": 0.3886,
+      "step": 21630
+    },
+    {
+      "epoch": 8.169120422801058,
+      "grad_norm": 0.6361426115036011,
+      "learning_rate": 4.73598104221121e-05,
+      "loss": 0.3856,
+      "step": 21640
+    },
+    {
+      "epoch": 8.172895432238581,
+      "grad_norm": 1.1138379573822021,
+      "learning_rate": 4.731853225226322e-05,
+      "loss": 0.3922,
+      "step": 21650
+    },
+    {
+      "epoch": 8.176670441676103,
+      "grad_norm": 0.666292667388916,
+      "learning_rate": 4.727725591517225e-05,
+      "loss": 0.3847,
+      "step": 21660
+    },
+    {
+      "epoch": 8.180445451113627,
+      "grad_norm": 0.6527111530303955,
+      "learning_rate": 4.723598143905119e-05,
+      "loss": 0.3905,
+      "step": 21670
+    },
+    {
+      "epoch": 8.18422046055115,
+      "grad_norm": 0.634685754776001,
+      "learning_rate": 4.719470885211077e-05,
+      "loss": 0.3734,
+      "step": 21680
+    },
+    {
+      "epoch": 8.187995469988675,
+      "grad_norm": 0.7132300734519958,
+      "learning_rate": 4.7153438182560387e-05,
+      "loss": 0.3891,
+      "step": 21690
+    },
+    {
+      "epoch": 8.191770479426198,
+      "grad_norm": 0.6142542958259583,
+      "learning_rate": 4.711216945860815e-05,
+      "loss": 0.3857,
+      "step": 21700
+    },
+    {
+      "epoch": 8.195545488863722,
+      "grad_norm": 0.6071975827217102,
+      "learning_rate": 4.707090270846088e-05,
+      "loss": 0.3857,
+      "step": 21710
+    },
+    {
+      "epoch": 8.199320498301246,
+      "grad_norm": 0.7059184312820435,
+      "learning_rate": 4.702963796032397e-05,
+      "loss": 0.3856,
+      "step": 21720
+    },
+    {
+      "epoch": 8.20309550773877,
+      "grad_norm": 0.6421445608139038,
+      "learning_rate": 4.6988375242401514e-05,
+      "loss": 0.3778,
+      "step": 21730
+    },
+    {
+      "epoch": 8.206870517176293,
+      "grad_norm": 0.7011498808860779,
+      "learning_rate": 4.694711458289618e-05,
+      "loss": 0.3898,
+      "step": 21740
+    },
+    {
+      "epoch": 8.210645526613817,
+      "grad_norm": 0.7598146796226501,
+      "learning_rate": 4.690585601000925e-05,
+      "loss": 0.3778,
+      "step": 21750
+    },
+    {
+      "epoch": 8.21442053605134,
+      "grad_norm": 0.6215041875839233,
+      "learning_rate": 4.686459955194055e-05,
+      "loss": 0.3854,
+      "step": 21760
+    },
+    {
+      "epoch": 8.218195545488864,
+      "grad_norm": 0.7370675206184387,
+      "learning_rate": 4.6823345236888504e-05,
+      "loss": 0.3823,
+      "step": 21770
+    },
+    {
+      "epoch": 8.221970554926388,
+      "grad_norm": 0.6150240898132324,
+      "learning_rate": 4.678209309305002e-05,
+      "loss": 0.3955,
+      "step": 21780
+    },
+    {
+      "epoch": 8.22574556436391,
+      "grad_norm": 0.5834974050521851,
+      "learning_rate": 4.674084314862057e-05,
+      "loss": 0.3852,
+      "step": 21790
+    },
+    {
+      "epoch": 8.229520573801434,
+      "grad_norm": 0.6348870992660522,
+      "learning_rate": 4.669959543179409e-05,
+      "loss": 0.3903,
+      "step": 21800
+    },
+    {
+      "epoch": 8.233295583238958,
+      "grad_norm": 0.7599389553070068,
+      "learning_rate": 4.665834997076303e-05,
+      "loss": 0.3849,
+      "step": 21810
+    },
+    {
+      "epoch": 8.237070592676481,
+      "grad_norm": 0.7290443181991577,
+      "learning_rate": 4.661710679371823e-05,
+      "loss": 0.3846,
+      "step": 21820
+    },
+    {
+      "epoch": 8.240845602114005,
+      "grad_norm": 0.6612056493759155,
+      "learning_rate": 4.657586592884905e-05,
+      "loss": 0.3849,
+      "step": 21830
+    },
+    {
+      "epoch": 8.244620611551529,
+      "grad_norm": 0.5885748863220215,
+      "learning_rate": 4.653462740434322e-05,
+      "loss": 0.3801,
+      "step": 21840
+    },
+    {
+      "epoch": 8.248395620989053,
+      "grad_norm": 0.7608574032783508,
+      "learning_rate": 4.649339124838689e-05,
+      "loss": 0.3873,
+      "step": 21850
+    },
+    {
+      "epoch": 8.252170630426576,
+      "grad_norm": 0.5708411335945129,
+      "learning_rate": 4.6452157489164574e-05,
+      "loss": 0.3858,
+      "step": 21860
+    },
+    {
+      "epoch": 8.2559456398641,
+      "grad_norm": 0.6522685289382935,
+      "learning_rate": 4.6410926154859155e-05,
+      "loss": 0.3869,
+      "step": 21870
+    },
+    {
+      "epoch": 8.259720649301624,
+      "grad_norm": 0.688291609287262,
+      "learning_rate": 4.636969727365186e-05,
+      "loss": 0.3839,
+      "step": 21880
+    },
+    {
+      "epoch": 8.263495658739147,
+      "grad_norm": 0.7088874578475952,
+      "learning_rate": 4.632847087372226e-05,
+      "loss": 0.3872,
+      "step": 21890
+    },
+    {
+      "epoch": 8.267270668176671,
+      "grad_norm": 0.6327393054962158,
+      "learning_rate": 4.628724698324818e-05,
+      "loss": 0.3819,
+      "step": 21900
+    },
+    {
+      "epoch": 8.271045677614193,
+      "grad_norm": 0.6789700388908386,
+      "learning_rate": 4.6246025630405795e-05,
+      "loss": 0.384,
+      "step": 21910
+    },
+    {
+      "epoch": 8.274820687051717,
+      "grad_norm": 0.6304154396057129,
+      "learning_rate": 4.6204806843369474e-05,
+      "loss": 0.3883,
+      "step": 21920
+    },
+    {
+      "epoch": 8.27859569648924,
+      "grad_norm": 0.6723567843437195,
+      "learning_rate": 4.616359065031191e-05,
+      "loss": 0.386,
+      "step": 21930
+    },
+    {
+      "epoch": 8.282370705926764,
+      "grad_norm": 0.801613450050354,
+      "learning_rate": 4.6122377079403946e-05,
+      "loss": 0.3893,
+      "step": 21940
+    },
+    {
+      "epoch": 8.286145715364288,
+      "grad_norm": 0.6835589408874512,
+      "learning_rate": 4.6081166158814695e-05,
+      "loss": 0.3801,
+      "step": 21950
+    },
+    {
+      "epoch": 8.289920724801812,
+      "grad_norm": 0.6579275131225586,
+      "learning_rate": 4.603995791671144e-05,
+      "loss": 0.3877,
+      "step": 21960
+    },
+    {
+      "epoch": 8.293695734239336,
+      "grad_norm": 0.6721914410591125,
+      "learning_rate": 4.599875238125957e-05,
+      "loss": 0.382,
+      "step": 21970
+    },
+    {
+      "epoch": 8.29747074367686,
+      "grad_norm": 0.6746684312820435,
+      "learning_rate": 4.595754958062273e-05,
+      "loss": 0.3833,
+      "step": 21980
+    },
+    {
+      "epoch": 8.301245753114383,
+      "grad_norm": 0.6919302940368652,
+      "learning_rate": 4.591634954296265e-05,
+      "loss": 0.3862,
+      "step": 21990
+    },
+    {
+      "epoch": 8.305020762551907,
+      "grad_norm": 0.6918922662734985,
+      "learning_rate": 4.587515229643913e-05,
+      "loss": 0.3834,
+      "step": 22000
+    },
+    {
+      "epoch": 8.30879577198943,
+      "grad_norm": 0.7155516743659973,
+      "learning_rate": 4.583395786921013e-05,
+      "loss": 0.3745,
+      "step": 22010
+    },
+    {
+      "epoch": 8.312570781426954,
+      "grad_norm": 0.607598602771759,
+      "learning_rate": 4.579276628943164e-05,
+      "loss": 0.3794,
+      "step": 22020
+    },
+    {
+      "epoch": 8.316345790864478,
+      "grad_norm": 0.673291027545929,
+      "learning_rate": 4.575157758525772e-05,
+      "loss": 0.3824,
+      "step": 22030
+    },
+    {
+      "epoch": 8.320120800302,
+      "grad_norm": 0.7992842197418213,
+      "learning_rate": 4.571039178484046e-05,
+      "loss": 0.3756,
+      "step": 22040
+    },
+    {
+      "epoch": 8.323895809739524,
+      "grad_norm": 0.7000048756599426,
+      "learning_rate": 4.566920891632998e-05,
+      "loss": 0.38,
+      "step": 22050
+    },
+    {
+      "epoch": 8.327670819177047,
+      "grad_norm": 0.6638155579566956,
+      "learning_rate": 4.562802900787436e-05,
+      "loss": 0.3841,
+      "step": 22060
+    },
+    {
+      "epoch": 8.331445828614571,
+      "grad_norm": 0.6273589730262756,
+      "learning_rate": 4.558685208761968e-05,
+      "loss": 0.3826,
+      "step": 22070
+    },
+    {
+      "epoch": 8.335220838052095,
+      "grad_norm": 0.6618062853813171,
+      "learning_rate": 4.554567818370998e-05,
+      "loss": 0.3874,
+      "step": 22080
+    },
+    {
+      "epoch": 8.338995847489619,
+      "grad_norm": 0.6083613038063049,
+      "learning_rate": 4.550450732428726e-05,
+      "loss": 0.3755,
+      "step": 22090
+    },
+    {
+      "epoch": 8.342770856927142,
+      "grad_norm": 0.6382494568824768,
+      "learning_rate": 4.546333953749137e-05,
+      "loss": 0.3794,
+      "step": 22100
+    },
+    {
+      "epoch": 8.346545866364666,
+      "grad_norm": 0.6135038733482361,
+      "learning_rate": 4.5422174851460154e-05,
+      "loss": 0.3803,
+      "step": 22110
+    },
+    {
+      "epoch": 8.35032087580219,
+      "grad_norm": 0.6180480718612671,
+      "learning_rate": 4.538101329432924e-05,
+      "loss": 0.3844,
+      "step": 22120
+    },
+    {
+      "epoch": 8.354095885239714,
+      "grad_norm": 0.6306928396224976,
+      "learning_rate": 4.5339854894232195e-05,
+      "loss": 0.3843,
+      "step": 22130
+    },
+    {
+      "epoch": 8.357870894677237,
+      "grad_norm": 0.594245433807373,
+      "learning_rate": 4.52986996793004e-05,
+      "loss": 0.3936,
+      "step": 22140
+    },
+    {
+      "epoch": 8.361645904114761,
+      "grad_norm": 0.6511313319206238,
+      "learning_rate": 4.5257547677663024e-05,
+      "loss": 0.3795,
+      "step": 22150
+    },
+    {
+      "epoch": 8.365420913552285,
+      "grad_norm": 0.6905748248100281,
+      "learning_rate": 4.52163989174471e-05,
+      "loss": 0.378,
+      "step": 22160
+    },
+    {
+      "epoch": 8.369195922989807,
+      "grad_norm": 0.6889414191246033,
+      "learning_rate": 4.51752534267774e-05,
+      "loss": 0.3779,
+      "step": 22170
+    },
+    {
+      "epoch": 8.37297093242733,
+      "grad_norm": 0.7038361430168152,
+      "learning_rate": 4.513411123377649e-05,
+      "loss": 0.3789,
+      "step": 22180
+    },
+    {
+      "epoch": 8.376745941864854,
+      "grad_norm": 0.8612080216407776,
+      "learning_rate": 4.5092972366564675e-05,
+      "loss": 0.3814,
+      "step": 22190
+    },
+    {
+      "epoch": 8.380520951302378,
+      "grad_norm": 0.6390385627746582,
+      "learning_rate": 4.505183685325997e-05,
+      "loss": 0.3787,
+      "step": 22200
+    },
+    {
+      "epoch": 8.384295960739902,
+      "grad_norm": 0.7001070976257324,
+      "learning_rate": 4.5010704721978125e-05,
+      "loss": 0.3821,
+      "step": 22210
+    },
+    {
+      "epoch": 8.388070970177425,
+      "grad_norm": 0.6782472729682922,
+      "learning_rate": 4.496957600083255e-05,
+      "loss": 0.3777,
+      "step": 22220
+    },
+    {
+      "epoch": 8.39184597961495,
+      "grad_norm": 0.6302990317344666,
+      "learning_rate": 4.4928450717934343e-05,
+      "loss": 0.379,
+      "step": 22230
+    },
+    {
+      "epoch": 8.395620989052473,
+      "grad_norm": 0.6895720958709717,
+      "learning_rate": 4.488732890139227e-05,
+      "loss": 0.3809,
+      "step": 22240
+    },
+    {
+      "epoch": 8.399395998489997,
+      "grad_norm": 0.6838264465332031,
+      "learning_rate": 4.4846210579312665e-05,
+      "loss": 0.3814,
+      "step": 22250
+    },
+    {
+      "epoch": 8.40317100792752,
+      "grad_norm": 0.6539376974105835,
+      "learning_rate": 4.480509577979953e-05,
+      "loss": 0.3828,
+      "step": 22260
+    },
+    {
+      "epoch": 8.406946017365044,
+      "grad_norm": 0.6490859985351562,
+      "learning_rate": 4.476398453095445e-05,
+      "loss": 0.387,
+      "step": 22270
+    },
+    {
+      "epoch": 8.410721026802568,
+      "grad_norm": 0.6515709161758423,
+      "learning_rate": 4.472287686087656e-05,
+      "loss": 0.389,
+      "step": 22280
+    },
+    {
+      "epoch": 8.41449603624009,
+      "grad_norm": 0.6394408941268921,
+      "learning_rate": 4.468177279766259e-05,
+      "loss": 0.381,
+      "step": 22290
+    },
+    {
+      "epoch": 8.418271045677614,
+      "grad_norm": 0.6975342631340027,
+      "learning_rate": 4.4640672369406746e-05,
+      "loss": 0.373,
+      "step": 22300
+    },
+    {
+      "epoch": 8.422046055115137,
+      "grad_norm": 0.6658486723899841,
+      "learning_rate": 4.459957560420082e-05,
+      "loss": 0.3773,
+      "step": 22310
+    },
+    {
+      "epoch": 8.425821064552661,
+      "grad_norm": 0.649869441986084,
+      "learning_rate": 4.455848253013403e-05,
+      "loss": 0.3816,
+      "step": 22320
+    },
+    {
+      "epoch": 8.429596073990185,
+      "grad_norm": 0.7822151184082031,
+      "learning_rate": 4.4517393175293146e-05,
+      "loss": 0.3872,
+      "step": 22330
+    },
+    {
+      "epoch": 8.433371083427708,
+      "grad_norm": 0.6239002346992493,
+      "learning_rate": 4.447630756776232e-05,
+      "loss": 0.3834,
+      "step": 22340
+    },
+    {
+      "epoch": 8.437146092865232,
+      "grad_norm": 0.6562296748161316,
+      "learning_rate": 4.443522573562318e-05,
+      "loss": 0.3933,
+      "step": 22350
+    },
+    {
+      "epoch": 8.440921102302756,
+      "grad_norm": 0.6250916123390198,
+      "learning_rate": 4.4394147706954776e-05,
+      "loss": 0.383,
+      "step": 22360
+    },
+    {
+      "epoch": 8.44469611174028,
+      "grad_norm": 0.6783263683319092,
+      "learning_rate": 4.435307350983355e-05,
+      "loss": 0.382,
+      "step": 22370
+    },
+    {
+      "epoch": 8.448471121177803,
+      "grad_norm": 0.5855938792228699,
+      "learning_rate": 4.4312003172333326e-05,
+      "loss": 0.3753,
+      "step": 22380
+    },
+    {
+      "epoch": 8.452246130615327,
+      "grad_norm": 0.6959243416786194,
+      "learning_rate": 4.427093672252531e-05,
+      "loss": 0.3769,
+      "step": 22390
+    },
+    {
+      "epoch": 8.45602114005285,
+      "grad_norm": 0.750425398349762,
+      "learning_rate": 4.422987418847802e-05,
+      "loss": 0.3819,
+      "step": 22400
+    },
+    {
+      "epoch": 8.459796149490373,
+      "grad_norm": 0.6782823204994202,
+      "learning_rate": 4.4188815598257325e-05,
+      "loss": 0.3751,
+      "step": 22410
+    },
+    {
+      "epoch": 8.463571158927897,
+      "grad_norm": 0.7516863346099854,
+      "learning_rate": 4.414776097992638e-05,
+      "loss": 0.3794,
+      "step": 22420
+    },
+    {
+      "epoch": 8.46734616836542,
+      "grad_norm": 0.694094717502594,
+      "learning_rate": 4.4106710361545595e-05,
+      "loss": 0.3835,
+      "step": 22430
+    },
+    {
+      "epoch": 8.471121177802944,
+      "grad_norm": 0.7622425556182861,
+      "learning_rate": 4.406566377117272e-05,
+      "loss": 0.3817,
+      "step": 22440
+    },
+    {
+      "epoch": 8.474896187240468,
+      "grad_norm": 0.6150171160697937,
+      "learning_rate": 4.40246212368627e-05,
+      "loss": 0.3821,
+      "step": 22450
+    },
+    {
+      "epoch": 8.478671196677992,
+      "grad_norm": 0.5998985171318054,
+      "learning_rate": 4.3983582786667715e-05,
+      "loss": 0.3769,
+      "step": 22460
+    },
+    {
+      "epoch": 8.482446206115515,
+      "grad_norm": 0.6367303133010864,
+      "learning_rate": 4.394254844863716e-05,
+      "loss": 0.3853,
+      "step": 22470
+    },
+    {
+      "epoch": 8.486221215553039,
+      "grad_norm": 0.7072235941886902,
+      "learning_rate": 4.390151825081762e-05,
+      "loss": 0.3844,
+      "step": 22480
+    },
+    {
+      "epoch": 8.489996224990563,
+      "grad_norm": 0.678936779499054,
+      "learning_rate": 4.386049222125286e-05,
+      "loss": 0.3695,
+      "step": 22490
+    },
+    {
+      "epoch": 8.493771234428086,
+      "grad_norm": 0.6695368885993958,
+      "learning_rate": 4.3819470387983774e-05,
+      "loss": 0.3838,
+      "step": 22500
+    },
+    {
+      "epoch": 8.49754624386561,
+      "grad_norm": 0.6475124359130859,
+      "learning_rate": 4.377845277904841e-05,
+      "loss": 0.3847,
+      "step": 22510
+    },
+    {
+      "epoch": 8.501321253303134,
+      "grad_norm": 0.6073307991027832,
+      "learning_rate": 4.37374394224819e-05,
+      "loss": 0.3805,
+      "step": 22520
+    },
+    {
+      "epoch": 8.505096262740658,
+      "grad_norm": 0.6134095191955566,
+      "learning_rate": 4.369643034631648e-05,
+      "loss": 0.388,
+      "step": 22530
+    },
+    {
+      "epoch": 8.50887127217818,
+      "grad_norm": 1.0386673212051392,
+      "learning_rate": 4.365542557858149e-05,
+      "loss": 0.3853,
+      "step": 22540
+    },
+    {
+      "epoch": 8.512646281615703,
+      "grad_norm": 0.6191480755805969,
+      "learning_rate": 4.361442514730329e-05,
+      "loss": 0.3866,
+      "step": 22550
+    },
+    {
+      "epoch": 8.516421291053227,
+      "grad_norm": 0.6350696682929993,
+      "learning_rate": 4.357342908050528e-05,
+      "loss": 0.3817,
+      "step": 22560
+    },
+    {
+      "epoch": 8.52019630049075,
+      "grad_norm": 0.6684324145317078,
+      "learning_rate": 4.3532437406207895e-05,
+      "loss": 0.382,
+      "step": 22570
+    },
+    {
+      "epoch": 8.523971309928275,
+      "grad_norm": 0.6637814044952393,
+      "learning_rate": 4.349145015242856e-05,
+      "loss": 0.3775,
+      "step": 22580
+    },
+    {
+      "epoch": 8.527746319365798,
+      "grad_norm": 0.7314398884773254,
+      "learning_rate": 4.345046734718168e-05,
+      "loss": 0.3755,
+      "step": 22590
+    },
+    {
+      "epoch": 8.531521328803322,
+      "grad_norm": 0.6954400539398193,
+      "learning_rate": 4.34094890184786e-05,
+      "loss": 0.3816,
+      "step": 22600
+    },
+    {
+      "epoch": 8.535296338240846,
+      "grad_norm": 0.6581496596336365,
+      "learning_rate": 4.336851519432765e-05,
+      "loss": 0.3869,
+      "step": 22610
+    },
+    {
+      "epoch": 8.53907134767837,
+      "grad_norm": 0.6038832068443298,
+      "learning_rate": 4.332754590273403e-05,
+      "loss": 0.3821,
+      "step": 22620
+    },
+    {
+      "epoch": 8.542846357115893,
+      "grad_norm": 0.6651731133460999,
+      "learning_rate": 4.3286581171699855e-05,
+      "loss": 0.3916,
+      "step": 22630
+    },
+    {
+      "epoch": 8.546621366553417,
+      "grad_norm": 0.6608867645263672,
+      "learning_rate": 4.324562102922416e-05,
+      "loss": 0.3764,
+      "step": 22640
+    },
+    {
+      "epoch": 8.55039637599094,
+      "grad_norm": 0.6170617938041687,
+      "learning_rate": 4.320466550330278e-05,
+      "loss": 0.3805,
+      "step": 22650
+    },
+    {
+      "epoch": 8.554171385428464,
+      "grad_norm": 0.6135375499725342,
+      "learning_rate": 4.3163714621928466e-05,
+      "loss": 0.3792,
+      "step": 22660
+    },
+    {
+      "epoch": 8.557946394865986,
+      "grad_norm": 0.6618521809577942,
+      "learning_rate": 4.312276841309074e-05,
+      "loss": 0.3768,
+      "step": 22670
+    },
+    {
+      "epoch": 8.56172140430351,
+      "grad_norm": 0.6322848200798035,
+      "learning_rate": 4.3081826904775945e-05,
+      "loss": 0.3733,
+      "step": 22680
+    },
+    {
+      "epoch": 8.565496413741034,
+      "grad_norm": 0.6854828000068665,
+      "learning_rate": 4.3040890124967246e-05,
+      "loss": 0.3895,
+      "step": 22690
+    },
+    {
+      "epoch": 8.569271423178558,
+      "grad_norm": 0.6372854113578796,
+      "learning_rate": 4.2999958101644537e-05,
+      "loss": 0.3766,
+      "step": 22700
+    },
+    {
+      "epoch": 8.573046432616081,
+      "grad_norm": 0.6355828642845154,
+      "learning_rate": 4.2959030862784435e-05,
+      "loss": 0.3769,
+      "step": 22710
+    },
+    {
+      "epoch": 8.576821442053605,
+      "grad_norm": 0.6458565592765808,
+      "learning_rate": 4.291810843636036e-05,
+      "loss": 0.3925,
+      "step": 22720
+    },
+    {
+      "epoch": 8.580596451491129,
+      "grad_norm": 0.6839278936386108,
+      "learning_rate": 4.2877190850342375e-05,
+      "loss": 0.388,
+      "step": 22730
+    },
+    {
+      "epoch": 8.584371460928653,
+      "grad_norm": 0.6284090876579285,
+      "learning_rate": 4.2836278132697294e-05,
+      "loss": 0.3827,
+      "step": 22740
+    },
+    {
+      "epoch": 8.588146470366176,
+      "grad_norm": 0.6752241253852844,
+      "learning_rate": 4.279537031138855e-05,
+      "loss": 0.3969,
+      "step": 22750
+    },
+    {
+      "epoch": 8.5919214798037,
+      "grad_norm": 0.6706361770629883,
+      "learning_rate": 4.275446741437625e-05,
+      "loss": 0.3845,
+      "step": 22760
+    },
+    {
+      "epoch": 8.595696489241224,
+      "grad_norm": 0.6637896299362183,
+      "learning_rate": 4.2713569469617176e-05,
+      "loss": 0.3749,
+      "step": 22770
+    },
+    {
+      "epoch": 8.599471498678747,
+      "grad_norm": 0.7166405916213989,
+      "learning_rate": 4.267267650506465e-05,
+      "loss": 0.3803,
+      "step": 22780
+    },
+    {
+      "epoch": 8.603246508116271,
+      "grad_norm": 1.1399887800216675,
+      "learning_rate": 4.263178854866866e-05,
+      "loss": 0.3818,
+      "step": 22790
+    },
+    {
+      "epoch": 8.607021517553793,
+      "grad_norm": 0.5608764886856079,
+      "learning_rate": 4.259090562837571e-05,
+      "loss": 0.3855,
+      "step": 22800
+    },
+    {
+      "epoch": 8.610796526991317,
+      "grad_norm": 0.6282104253768921,
+      "learning_rate": 4.255002777212888e-05,
+      "loss": 0.3775,
+      "step": 22810
+    },
+    {
+      "epoch": 8.61457153642884,
+      "grad_norm": 0.7265238165855408,
+      "learning_rate": 4.250915500786783e-05,
+      "loss": 0.38,
+      "step": 22820
+    },
+    {
+      "epoch": 8.618346545866364,
+      "grad_norm": 0.6828112006187439,
+      "learning_rate": 4.24682873635287e-05,
+      "loss": 0.3821,
+      "step": 22830
+    },
+    {
+      "epoch": 8.622121555303888,
+      "grad_norm": 0.6247879862785339,
+      "learning_rate": 4.242742486704414e-05,
+      "loss": 0.388,
+      "step": 22840
+    },
+    {
+      "epoch": 8.625896564741412,
+      "grad_norm": 0.6014358997344971,
+      "learning_rate": 4.238656754634327e-05,
+      "loss": 0.3832,
+      "step": 22850
+    },
+    {
+      "epoch": 8.629671574178936,
+      "grad_norm": 0.6423870921134949,
+      "learning_rate": 4.234571542935168e-05,
+      "loss": 0.3951,
+      "step": 22860
+    },
+    {
+      "epoch": 8.63344658361646,
+      "grad_norm": 3.2498388290405273,
+      "learning_rate": 4.230486854399144e-05,
+      "loss": 0.3945,
+      "step": 22870
+    },
+    {
+      "epoch": 8.637221593053983,
+      "grad_norm": 0.9906812906265259,
+      "learning_rate": 4.226402691818098e-05,
+      "loss": 0.3884,
+      "step": 22880
+    },
+    {
+      "epoch": 8.640996602491507,
+      "grad_norm": 0.6815230250358582,
+      "learning_rate": 4.2223190579835196e-05,
+      "loss": 0.3842,
+      "step": 22890
+    },
+    {
+      "epoch": 8.64477161192903,
+      "grad_norm": 0.6938421726226807,
+      "learning_rate": 4.218235955686531e-05,
+      "loss": 0.3797,
+      "step": 22900
+    },
+    {
+      "epoch": 8.648546621366553,
+      "grad_norm": 1.4999902248382568,
+      "learning_rate": 4.214153387717894e-05,
+      "loss": 0.3921,
+      "step": 22910
+    },
+    {
+      "epoch": 8.652321630804076,
+      "grad_norm": 1.1413220167160034,
+      "learning_rate": 4.210071356868007e-05,
+      "loss": 0.3825,
+      "step": 22920
+    },
+    {
+      "epoch": 8.6560966402416,
+      "grad_norm": 0.633385419845581,
+      "learning_rate": 4.205989865926898e-05,
+      "loss": 0.3819,
+      "step": 22930
+    },
+    {
+      "epoch": 8.659871649679124,
+      "grad_norm": 0.680814802646637,
+      "learning_rate": 4.2019089176842294e-05,
+      "loss": 0.3818,
+      "step": 22940
+    },
+    {
+      "epoch": 8.663646659116647,
+      "grad_norm": 0.5858637094497681,
+      "learning_rate": 4.1978285149292894e-05,
+      "loss": 0.3804,
+      "step": 22950
+    },
+    {
+      "epoch": 8.667421668554171,
+      "grad_norm": 0.6605408191680908,
+      "learning_rate": 4.193748660450996e-05,
+      "loss": 0.389,
+      "step": 22960
+    },
+    {
+      "epoch": 8.671196677991695,
+      "grad_norm": 0.6238226890563965,
+      "learning_rate": 4.189669357037891e-05,
+      "loss": 0.378,
+      "step": 22970
+    },
+    {
+      "epoch": 8.674971687429219,
+      "grad_norm": 0.7179723978042603,
+      "learning_rate": 4.1855906074781405e-05,
+      "loss": 0.3778,
+      "step": 22980
+    },
+    {
+      "epoch": 8.678746696866742,
+      "grad_norm": 0.6245715618133545,
+      "learning_rate": 4.1815124145595285e-05,
+      "loss": 0.3823,
+      "step": 22990
+    },
+    {
+      "epoch": 8.682521706304266,
+      "grad_norm": 0.6498119831085205,
+      "learning_rate": 4.1774347810694644e-05,
+      "loss": 0.3801,
+      "step": 23000
+    },
+    {
+      "epoch": 8.68629671574179,
+      "grad_norm": 0.8200873136520386,
+      "learning_rate": 4.17335770979497e-05,
+      "loss": 0.3793,
+      "step": 23010
+    },
+    {
+      "epoch": 8.690071725179314,
+      "grad_norm": 0.6485256552696228,
+      "learning_rate": 4.169281203522687e-05,
+      "loss": 0.3811,
+      "step": 23020
+    },
+    {
+      "epoch": 8.693846734616837,
+      "grad_norm": 0.5570192933082581,
+      "learning_rate": 4.1652052650388674e-05,
+      "loss": 0.3783,
+      "step": 23030
+    },
+    {
+      "epoch": 8.69762174405436,
+      "grad_norm": 0.7071910500526428,
+      "learning_rate": 4.1611298971293786e-05,
+      "loss": 0.3839,
+      "step": 23040
+    },
+    {
+      "epoch": 8.701396753491883,
+      "grad_norm": 0.6622625589370728,
+      "learning_rate": 4.1570551025796935e-05,
+      "loss": 0.3754,
+      "step": 23050
+    },
+    {
+      "epoch": 8.705171762929407,
+      "grad_norm": 0.680098831653595,
+      "learning_rate": 4.152980884174897e-05,
+      "loss": 0.3802,
+      "step": 23060
+    },
+    {
+      "epoch": 8.70894677236693,
+      "grad_norm": 0.6106149554252625,
+      "learning_rate": 4.148907244699682e-05,
+      "loss": 0.3755,
+      "step": 23070
+    },
+    {
+      "epoch": 8.712721781804454,
+      "grad_norm": 0.5729265809059143,
+      "learning_rate": 4.1448341869383395e-05,
+      "loss": 0.38,
+      "step": 23080
+    },
+    {
+      "epoch": 8.716496791241978,
+      "grad_norm": 0.7992091774940491,
+      "learning_rate": 4.140761713674765e-05,
+      "loss": 0.3833,
+      "step": 23090
+    },
+    {
+      "epoch": 8.720271800679502,
+      "grad_norm": 0.6829302906990051,
+      "learning_rate": 4.1366898276924574e-05,
+      "loss": 0.3739,
+      "step": 23100
+    },
+    {
+      "epoch": 8.724046810117025,
+      "grad_norm": 0.6297142505645752,
+      "learning_rate": 4.132618531774512e-05,
+      "loss": 0.3754,
+      "step": 23110
+    },
+    {
+      "epoch": 8.72782181955455,
+      "grad_norm": 0.6362572908401489,
+      "learning_rate": 4.128547828703622e-05,
+      "loss": 0.3764,
+      "step": 23120
+    },
+    {
+      "epoch": 8.731596828992073,
+      "grad_norm": 0.7810057997703552,
+      "learning_rate": 4.1244777212620725e-05,
+      "loss": 0.3822,
+      "step": 23130
+    },
+    {
+      "epoch": 8.735371838429597,
+      "grad_norm": 0.6120405197143555,
+      "learning_rate": 4.120408212231746e-05,
+      "loss": 0.3755,
+      "step": 23140
+    },
+    {
+      "epoch": 8.73914684786712,
+      "grad_norm": 0.6071982979774475,
+      "learning_rate": 4.116339304394111e-05,
+      "loss": 0.3795,
+      "step": 23150
+    },
+    {
+      "epoch": 8.742921857304644,
+      "grad_norm": 0.7162799835205078,
+      "learning_rate": 4.112271000530229e-05,
+      "loss": 0.3815,
+      "step": 23160
+    },
+    {
+      "epoch": 8.746696866742166,
+      "grad_norm": 0.5881337523460388,
+      "learning_rate": 4.10820330342075e-05,
+      "loss": 0.3798,
+      "step": 23170
+    },
+    {
+      "epoch": 8.75047187617969,
+      "grad_norm": 0.6906888484954834,
+      "learning_rate": 4.1041362158459027e-05,
+      "loss": 0.3732,
+      "step": 23180
+    },
+    {
+      "epoch": 8.754246885617214,
+      "grad_norm": 0.6561579704284668,
+      "learning_rate": 4.1000697405855024e-05,
+      "loss": 0.3825,
+      "step": 23190
+    },
+    {
+      "epoch": 8.758021895054737,
+      "grad_norm": 0.5920860171318054,
+      "learning_rate": 4.096003880418951e-05,
+      "loss": 0.3819,
+      "step": 23200
+    },
+    {
+      "epoch": 8.761796904492261,
+      "grad_norm": 1.4881322383880615,
+      "learning_rate": 4.0919386381252215e-05,
+      "loss": 0.3763,
+      "step": 23210
+    },
+    {
+      "epoch": 8.765571913929785,
+      "grad_norm": 0.7468271255493164,
+      "learning_rate": 4.087874016482872e-05,
+      "loss": 0.3806,
+      "step": 23220
+    },
+    {
+      "epoch": 8.769346923367308,
+      "grad_norm": 0.5837022066116333,
+      "learning_rate": 4.0838100182700295e-05,
+      "loss": 0.382,
+      "step": 23230
+    },
+    {
+      "epoch": 8.773121932804832,
+      "grad_norm": 0.590907633304596,
+      "learning_rate": 4.079746646264402e-05,
+      "loss": 0.3886,
+      "step": 23240
+    },
+    {
+      "epoch": 8.776896942242356,
+      "grad_norm": 0.9434071779251099,
+      "learning_rate": 4.075683903243262e-05,
+      "loss": 0.3737,
+      "step": 23250
+    },
+    {
+      "epoch": 8.78067195167988,
+      "grad_norm": 0.7620961666107178,
+      "learning_rate": 4.071621791983462e-05,
+      "loss": 0.3776,
+      "step": 23260
+    },
+    {
+      "epoch": 8.784446961117403,
+      "grad_norm": 0.5980284214019775,
+      "learning_rate": 4.06756031526141e-05,
+      "loss": 0.3804,
+      "step": 23270
+    },
+    {
+      "epoch": 8.788221970554927,
+      "grad_norm": 0.6051507592201233,
+      "learning_rate": 4.063499475853092e-05,
+      "loss": 0.3804,
+      "step": 23280
+    },
+    {
+      "epoch": 8.791996979992451,
+      "grad_norm": 0.6261473894119263,
+      "learning_rate": 4.0594392765340506e-05,
+      "loss": 0.3658,
+      "step": 23290
+    },
+    {
+      "epoch": 8.795771989429973,
+      "grad_norm": 0.6253399848937988,
+      "learning_rate": 4.0553797200793954e-05,
+      "loss": 0.3819,
+      "step": 23300
+    },
+    {
+      "epoch": 8.799546998867497,
+      "grad_norm": 0.6369916200637817,
+      "learning_rate": 4.0513208092637926e-05,
+      "loss": 0.3783,
+      "step": 23310
+    },
+    {
+      "epoch": 8.80332200830502,
+      "grad_norm": 0.6382907032966614,
+      "learning_rate": 4.0472625468614735e-05,
+      "loss": 0.3774,
+      "step": 23320
+    },
+    {
+      "epoch": 8.807097017742544,
+      "grad_norm": 1.013975977897644,
+      "learning_rate": 4.043204935646218e-05,
+      "loss": 0.3976,
+      "step": 23330
+    },
+    {
+      "epoch": 8.810872027180068,
+      "grad_norm": 0.7413721084594727,
+      "learning_rate": 4.0391479783913675e-05,
+      "loss": 0.3813,
+      "step": 23340
+    },
+    {
+      "epoch": 8.814647036617592,
+      "grad_norm": 0.7551056146621704,
+      "learning_rate": 4.0350916778698155e-05,
+      "loss": 0.3754,
+      "step": 23350
+    },
+    {
+      "epoch": 8.818422046055115,
+      "grad_norm": 0.6768625378608704,
+      "learning_rate": 4.031036036854001e-05,
+      "loss": 0.3672,
+      "step": 23360
+    },
+    {
+      "epoch": 8.822197055492639,
+      "grad_norm": 0.7512798309326172,
+      "learning_rate": 4.026981058115918e-05,
+      "loss": 0.3806,
+      "step": 23370
+    },
+    {
+      "epoch": 8.825972064930163,
+      "grad_norm": 0.6166073679924011,
+      "learning_rate": 4.022926744427108e-05,
+      "loss": 0.3818,
+      "step": 23380
+    },
+    {
+      "epoch": 8.829747074367686,
+      "grad_norm": 0.6651231050491333,
+      "learning_rate": 4.018873098558654e-05,
+      "loss": 0.379,
+      "step": 23390
+    },
+    {
+      "epoch": 8.83352208380521,
+      "grad_norm": 0.6436861157417297,
+      "learning_rate": 4.014820123281186e-05,
+      "loss": 0.3816,
+      "step": 23400
+    },
+    {
+      "epoch": 8.837297093242734,
+      "grad_norm": 0.6689978837966919,
+      "learning_rate": 4.0107678213648735e-05,
+      "loss": 0.3756,
+      "step": 23410
+    },
+    {
+      "epoch": 8.841072102680256,
+      "grad_norm": 0.6367429494857788,
+      "learning_rate": 4.006716195579428e-05,
+      "loss": 0.3759,
+      "step": 23420
+    },
+    {
+      "epoch": 8.84484711211778,
+      "grad_norm": 0.649011492729187,
+      "learning_rate": 4.002665248694096e-05,
+      "loss": 0.3727,
+      "step": 23430
+    },
+    {
+      "epoch": 8.848622121555303,
+      "grad_norm": 0.7108133435249329,
+      "learning_rate": 3.998614983477664e-05,
+      "loss": 0.3741,
+      "step": 23440
+    },
+    {
+      "epoch": 8.852397130992827,
+      "grad_norm": 0.6738606691360474,
+      "learning_rate": 3.994565402698448e-05,
+      "loss": 0.3754,
+      "step": 23450
+    },
+    {
+      "epoch": 8.85617214043035,
+      "grad_norm": 0.6250699758529663,
+      "learning_rate": 3.9905165091242975e-05,
+      "loss": 0.3832,
+      "step": 23460
+    },
+    {
+      "epoch": 8.859947149867875,
+      "grad_norm": 0.6175537705421448,
+      "learning_rate": 3.9864683055225936e-05,
+      "loss": 0.3705,
+      "step": 23470
+    },
+    {
+      "epoch": 8.863722159305398,
+      "grad_norm": 0.6735917925834656,
+      "learning_rate": 3.982420794660247e-05,
+      "loss": 0.3799,
+      "step": 23480
+    },
+    {
+      "epoch": 8.867497168742922,
+      "grad_norm": 0.7157145738601685,
+      "learning_rate": 3.978373979303691e-05,
+      "loss": 0.3749,
+      "step": 23490
+    },
+    {
+      "epoch": 8.871272178180446,
+      "grad_norm": 0.6760987639427185,
+      "learning_rate": 3.974327862218888e-05,
+      "loss": 0.3851,
+      "step": 23500
+    },
+    {
+      "epoch": 8.87504718761797,
+      "grad_norm": 0.6828089356422424,
+      "learning_rate": 3.970282446171318e-05,
+      "loss": 0.3787,
+      "step": 23510
+    },
+    {
+      "epoch": 8.878822197055493,
+      "grad_norm": 0.6380763053894043,
+      "learning_rate": 3.966237733925988e-05,
+      "loss": 0.3748,
+      "step": 23520
+    },
+    {
+      "epoch": 8.882597206493017,
+      "grad_norm": 0.6100180745124817,
+      "learning_rate": 3.962193728247418e-05,
+      "loss": 0.3767,
+      "step": 23530
+    },
+    {
+      "epoch": 8.886372215930539,
+      "grad_norm": 0.6232542991638184,
+      "learning_rate": 3.958150431899651e-05,
+      "loss": 0.381,
+      "step": 23540
+    },
+    {
+      "epoch": 8.890147225368063,
+      "grad_norm": 0.636439323425293,
+      "learning_rate": 3.954107847646238e-05,
+      "loss": 0.3755,
+      "step": 23550
+    },
+    {
+      "epoch": 8.893922234805586,
+      "grad_norm": 0.611513078212738,
+      "learning_rate": 3.950065978250249e-05,
+      "loss": 0.3761,
+      "step": 23560
+    },
+    {
+      "epoch": 8.89769724424311,
+      "grad_norm": 0.6682894825935364,
+      "learning_rate": 3.9460248264742624e-05,
+      "loss": 0.3789,
+      "step": 23570
+    },
+    {
+      "epoch": 8.901472253680634,
+      "grad_norm": 0.6221024990081787,
+      "learning_rate": 3.941984395080371e-05,
+      "loss": 0.3734,
+      "step": 23580
+    },
+    {
+      "epoch": 8.905247263118158,
+      "grad_norm": 0.6244845390319824,
+      "learning_rate": 3.937944686830167e-05,
+      "loss": 0.3826,
+      "step": 23590
+    },
+    {
+      "epoch": 8.909022272555681,
+      "grad_norm": 0.6587332487106323,
+      "learning_rate": 3.933905704484756e-05,
+      "loss": 0.3759,
+      "step": 23600
+    },
+    {
+      "epoch": 8.912797281993205,
+      "grad_norm": 0.6589406728744507,
+      "learning_rate": 3.929867450804743e-05,
+      "loss": 0.3756,
+      "step": 23610
+    },
+    {
+      "epoch": 8.916572291430729,
+      "grad_norm": 0.6549832224845886,
+      "learning_rate": 3.925829928550237e-05,
+      "loss": 0.3883,
+      "step": 23620
+    },
+    {
+      "epoch": 8.920347300868253,
+      "grad_norm": 0.7647480964660645,
+      "learning_rate": 3.921793140480847e-05,
+      "loss": 0.3709,
+      "step": 23630
+    },
+    {
+      "epoch": 8.924122310305776,
+      "grad_norm": 0.6407207250595093,
+      "learning_rate": 3.917757089355677e-05,
+      "loss": 0.3736,
+      "step": 23640
+    },
+    {
+      "epoch": 8.9278973197433,
+      "grad_norm": 1.7529853582382202,
+      "learning_rate": 3.9137217779333326e-05,
+      "loss": 0.3909,
+      "step": 23650
+    },
+    {
+      "epoch": 8.931672329180824,
+      "grad_norm": 0.7098379135131836,
+      "learning_rate": 3.9096872089719083e-05,
+      "loss": 0.3768,
+      "step": 23660
+    },
+    {
+      "epoch": 8.935447338618346,
+      "grad_norm": 0.5911573767662048,
+      "learning_rate": 3.905653385228996e-05,
+      "loss": 0.3764,
+      "step": 23670
+    },
+    {
+      "epoch": 8.93922234805587,
+      "grad_norm": 0.652247965335846,
+      "learning_rate": 3.901620309461677e-05,
+      "loss": 0.3774,
+      "step": 23680
+    },
+    {
+      "epoch": 8.942997357493393,
+      "grad_norm": 0.6572662591934204,
+      "learning_rate": 3.897587984426518e-05,
+      "loss": 0.3836,
+      "step": 23690
+    },
+    {
+      "epoch": 8.946772366930917,
+      "grad_norm": 0.6210899949073792,
+      "learning_rate": 3.893556412879577e-05,
+      "loss": 0.3724,
+      "step": 23700
+    },
+    {
+      "epoch": 8.95054737636844,
+      "grad_norm": 0.6767255067825317,
+      "learning_rate": 3.889525597576395e-05,
+      "loss": 0.3752,
+      "step": 23710
+    },
+    {
+      "epoch": 8.954322385805964,
+      "grad_norm": 0.6062052249908447,
+      "learning_rate": 3.8854955412719965e-05,
+      "loss": 0.3703,
+      "step": 23720
+    },
+    {
+      "epoch": 8.958097395243488,
+      "grad_norm": 0.7351146936416626,
+      "learning_rate": 3.881466246720887e-05,
+      "loss": 0.3779,
+      "step": 23730
+    },
+    {
+      "epoch": 8.961872404681012,
+      "grad_norm": 0.6535313725471497,
+      "learning_rate": 3.8774377166770484e-05,
+      "loss": 0.3723,
+      "step": 23740
+    },
+    {
+      "epoch": 8.965647414118536,
+      "grad_norm": 0.6908565163612366,
+      "learning_rate": 3.8734099538939474e-05,
+      "loss": 0.374,
+      "step": 23750
+    },
+    {
+      "epoch": 8.96942242355606,
+      "grad_norm": 0.6240267157554626,
+      "learning_rate": 3.869382961124518e-05,
+      "loss": 0.3791,
+      "step": 23760
+    },
+    {
+      "epoch": 8.973197432993583,
+      "grad_norm": 0.6270217895507812,
+      "learning_rate": 3.8653567411211736e-05,
+      "loss": 0.3766,
+      "step": 23770
+    },
+    {
+      "epoch": 8.976972442431107,
+      "grad_norm": 0.6301257014274597,
+      "learning_rate": 3.8613312966357987e-05,
+      "loss": 0.3816,
+      "step": 23780
+    },
+    {
+      "epoch": 8.98074745186863,
+      "grad_norm": 0.6038647890090942,
+      "learning_rate": 3.857306630419745e-05,
+      "loss": 0.3798,
+      "step": 23790
+    },
+    {
+      "epoch": 8.984522461306153,
+      "grad_norm": 0.6066597104072571,
+      "learning_rate": 3.853282745223834e-05,
+      "loss": 0.3679,
+      "step": 23800
+    },
+    {
+      "epoch": 8.988297470743676,
+      "grad_norm": 0.7371004819869995,
+      "learning_rate": 3.8492596437983546e-05,
+      "loss": 0.3713,
+      "step": 23810
+    },
+    {
+      "epoch": 8.9920724801812,
+      "grad_norm": 0.7040576934814453,
+      "learning_rate": 3.8452373288930586e-05,
+      "loss": 0.3827,
+      "step": 23820
+    },
+    {
+      "epoch": 8.995847489618724,
+      "grad_norm": 0.6456139087677002,
+      "learning_rate": 3.841215803257159e-05,
+      "loss": 0.3755,
+      "step": 23830
+    },
+    {
+      "epoch": 8.999622499056247,
+      "grad_norm": 0.7062108516693115,
+      "learning_rate": 3.83719506963933e-05,
+      "loss": 0.374,
+      "step": 23840
+    },
+    {
+      "epoch": 9.003397508493771,
+      "grad_norm": 0.6744314432144165,
+      "learning_rate": 3.8331751307877087e-05,
+      "loss": 0.372,
+      "step": 23850
+    },
+    {
+      "epoch": 9.007172517931295,
+      "grad_norm": 0.6453410983085632,
+      "learning_rate": 3.82915598944988e-05,
+      "loss": 0.3676,
+      "step": 23860
+    },
+    {
+      "epoch": 9.010947527368819,
+      "grad_norm": 0.7047562599182129,
+      "learning_rate": 3.825137648372893e-05,
+      "loss": 0.3786,
+      "step": 23870
+    },
+    {
+      "epoch": 9.014722536806342,
+      "grad_norm": 0.6693297624588013,
+      "learning_rate": 3.8211201103032465e-05,
+      "loss": 0.3704,
+      "step": 23880
+    },
+    {
+      "epoch": 9.018497546243866,
+      "grad_norm": 0.6609267592430115,
+      "learning_rate": 3.817103377986887e-05,
+      "loss": 0.3851,
+      "step": 23890
+    },
+    {
+      "epoch": 9.02227255568139,
+      "grad_norm": 0.6702944040298462,
+      "learning_rate": 3.813087454169215e-05,
+      "loss": 0.3761,
+      "step": 23900
+    },
+    {
+      "epoch": 9.026047565118914,
+      "grad_norm": 0.61748206615448,
+      "learning_rate": 3.809072341595078e-05,
+      "loss": 0.3738,
+      "step": 23910
+    },
+    {
+      "epoch": 9.029822574556436,
+      "grad_norm": 0.6437124609947205,
+      "learning_rate": 3.8050580430087636e-05,
+      "loss": 0.3763,
+      "step": 23920
+    },
+    {
+      "epoch": 9.03359758399396,
+      "grad_norm": 0.7444436550140381,
+      "learning_rate": 3.8010445611540096e-05,
+      "loss": 0.3817,
+      "step": 23930
+    },
+    {
+      "epoch": 9.037372593431483,
+      "grad_norm": 0.6082374453544617,
+      "learning_rate": 3.797031898773992e-05,
+      "loss": 0.3766,
+      "step": 23940
+    },
+    {
+      "epoch": 9.041147602869007,
+      "grad_norm": 0.6564713716506958,
+      "learning_rate": 3.793020058611329e-05,
+      "loss": 0.3716,
+      "step": 23950
+    },
+    {
+      "epoch": 9.04492261230653,
+      "grad_norm": 0.5503236651420593,
+      "learning_rate": 3.789009043408074e-05,
+      "loss": 0.3771,
+      "step": 23960
+    },
+    {
+      "epoch": 9.048697621744054,
+      "grad_norm": 0.6116209030151367,
+      "learning_rate": 3.7849988559057194e-05,
+      "loss": 0.3796,
+      "step": 23970
+    },
+    {
+      "epoch": 9.052472631181578,
+      "grad_norm": 0.6737613081932068,
+      "learning_rate": 3.78098949884519e-05,
+      "loss": 0.3765,
+      "step": 23980
+    },
+    {
+      "epoch": 9.056247640619102,
+      "grad_norm": 0.6291811466217041,
+      "learning_rate": 3.776980974966843e-05,
+      "loss": 0.3685,
+      "step": 23990
+    },
+    {
+      "epoch": 9.060022650056625,
+      "grad_norm": 0.6767562031745911,
+      "learning_rate": 3.772973287010468e-05,
+      "loss": 0.3759,
+      "step": 24000
+    },
+    {
+      "epoch": 9.06379765949415,
+      "grad_norm": 0.6803905963897705,
+      "learning_rate": 3.768966437715283e-05,
+      "loss": 0.372,
+      "step": 24010
+    },
+    {
+      "epoch": 9.067572668931673,
+      "grad_norm": 0.6342119574546814,
+      "learning_rate": 3.7649604298199274e-05,
+      "loss": 0.3688,
+      "step": 24020
+    },
+    {
+      "epoch": 9.071347678369197,
+      "grad_norm": 0.6667430996894836,
+      "learning_rate": 3.760955266062473e-05,
+      "loss": 0.3792,
+      "step": 24030
+    },
+    {
+      "epoch": 9.07512268780672,
+      "grad_norm": 0.6684991717338562,
+      "learning_rate": 3.75695094918041e-05,
+      "loss": 0.3715,
+      "step": 24040
+    },
+    {
+      "epoch": 9.078897697244242,
+      "grad_norm": 0.6000959277153015,
+      "learning_rate": 3.752947481910652e-05,
+      "loss": 0.3751,
+      "step": 24050
+    },
+    {
+      "epoch": 9.082672706681766,
+      "grad_norm": 0.6405602693557739,
+      "learning_rate": 3.7489448669895324e-05,
+      "loss": 0.3863,
+      "step": 24060
+    },
+    {
+      "epoch": 9.08644771611929,
+      "grad_norm": 0.6103823781013489,
+      "learning_rate": 3.744943107152798e-05,
+      "loss": 0.3703,
+      "step": 24070
+    },
+    {
+      "epoch": 9.090222725556814,
+      "grad_norm": 0.6388285756111145,
+      "learning_rate": 3.7409422051356165e-05,
+      "loss": 0.3784,
+      "step": 24080
+    },
+    {
+      "epoch": 9.093997734994337,
+      "grad_norm": 0.7293974161148071,
+      "learning_rate": 3.736942163672564e-05,
+      "loss": 0.393,
+      "step": 24090
+    },
+    {
+      "epoch": 9.097772744431861,
+      "grad_norm": 0.7579730749130249,
+      "learning_rate": 3.732942985497636e-05,
+      "loss": 0.3702,
+      "step": 24100
+    },
+    {
+      "epoch": 9.101547753869385,
+      "grad_norm": 0.6169874668121338,
+      "learning_rate": 3.728944673344228e-05,
+      "loss": 0.3748,
+      "step": 24110
+    },
+    {
+      "epoch": 9.105322763306908,
+      "grad_norm": 0.6163791418075562,
+      "learning_rate": 3.72494722994515e-05,
+      "loss": 0.3786,
+      "step": 24120
+    },
+    {
+      "epoch": 9.109097772744432,
+      "grad_norm": 0.7307747602462769,
+      "learning_rate": 3.720950658032617e-05,
+      "loss": 0.3765,
+      "step": 24130
+    },
+    {
+      "epoch": 9.112872782181956,
+      "grad_norm": 0.6685255765914917,
+      "learning_rate": 3.716954960338249e-05,
+      "loss": 0.3714,
+      "step": 24140
+    },
+    {
+      "epoch": 9.11664779161948,
+      "grad_norm": 0.6877933144569397,
+      "learning_rate": 3.712960139593066e-05,
+      "loss": 0.3769,
+      "step": 24150
+    },
+    {
+      "epoch": 9.120422801057003,
+      "grad_norm": 0.6554901599884033,
+      "learning_rate": 3.708966198527493e-05,
+      "loss": 0.3784,
+      "step": 24160
+    },
+    {
+      "epoch": 9.124197810494525,
+      "grad_norm": 0.7071563005447388,
+      "learning_rate": 3.704973139871349e-05,
+      "loss": 0.3673,
+      "step": 24170
+    },
+    {
+      "epoch": 9.12797281993205,
+      "grad_norm": 0.5478484630584717,
+      "learning_rate": 3.700980966353853e-05,
+      "loss": 0.3713,
+      "step": 24180
+    },
+    {
+      "epoch": 9.131747829369573,
+      "grad_norm": 0.6506239175796509,
+      "learning_rate": 3.696989680703619e-05,
+      "loss": 0.3786,
+      "step": 24190
+    },
+    {
+      "epoch": 9.135522838807097,
+      "grad_norm": 0.6592217087745667,
+      "learning_rate": 3.69299928564865e-05,
+      "loss": 0.3782,
+      "step": 24200
+    },
+    {
+      "epoch": 9.13929784824462,
+      "grad_norm": 0.6553889513015747,
+      "learning_rate": 3.689009783916345e-05,
+      "loss": 0.3705,
+      "step": 24210
+    },
+    {
+      "epoch": 9.143072857682144,
+      "grad_norm": 0.6616779565811157,
+      "learning_rate": 3.6850211782334895e-05,
+      "loss": 0.3685,
+      "step": 24220
+    },
+    {
+      "epoch": 9.146847867119668,
+      "grad_norm": 0.6655992269515991,
+      "learning_rate": 3.681033471326261e-05,
+      "loss": 0.3741,
+      "step": 24230
+    },
+    {
+      "epoch": 9.150622876557192,
+      "grad_norm": 0.6539639830589294,
+      "learning_rate": 3.677046665920216e-05,
+      "loss": 0.3741,
+      "step": 24240
+    },
+    {
+      "epoch": 9.154397885994715,
+      "grad_norm": 0.6297611594200134,
+      "learning_rate": 3.6730607647403005e-05,
+      "loss": 0.3766,
+      "step": 24250
+    },
+    {
+      "epoch": 9.158172895432239,
+      "grad_norm": 0.6319004893302917,
+      "learning_rate": 3.6690757705108416e-05,
+      "loss": 0.3773,
+      "step": 24260
+    },
+    {
+      "epoch": 9.161947904869763,
+      "grad_norm": 1.1053649187088013,
+      "learning_rate": 3.665091685955542e-05,
+      "loss": 0.3735,
+      "step": 24270
+    },
+    {
+      "epoch": 9.165722914307286,
+      "grad_norm": 0.6567557454109192,
+      "learning_rate": 3.6611085137974896e-05,
+      "loss": 0.3817,
+      "step": 24280
+    },
+    {
+      "epoch": 9.16949792374481,
+      "grad_norm": 0.6980771422386169,
+      "learning_rate": 3.657126256759143e-05,
+      "loss": 0.3696,
+      "step": 24290
+    },
+    {
+      "epoch": 9.173272933182332,
+      "grad_norm": 0.6510254144668579,
+      "learning_rate": 3.653144917562335e-05,
+      "loss": 0.3991,
+      "step": 24300
+    },
+    {
+      "epoch": 9.177047942619856,
+      "grad_norm": 0.5957704782485962,
+      "learning_rate": 3.649164498928277e-05,
+      "loss": 0.3752,
+      "step": 24310
+    },
+    {
+      "epoch": 9.18082295205738,
+      "grad_norm": 0.7079484462738037,
+      "learning_rate": 3.645185003577546e-05,
+      "loss": 0.3757,
+      "step": 24320
+    },
+    {
+      "epoch": 9.184597961494903,
+      "grad_norm": 0.6439237594604492,
+      "learning_rate": 3.6412064342300906e-05,
+      "loss": 0.3793,
+      "step": 24330
+    },
+    {
+      "epoch": 9.188372970932427,
+      "grad_norm": 0.6560051441192627,
+      "learning_rate": 3.637228793605224e-05,
+      "loss": 0.3781,
+      "step": 24340
+    },
+    {
+      "epoch": 9.19214798036995,
+      "grad_norm": 0.6205071210861206,
+      "learning_rate": 3.6332520844216264e-05,
+      "loss": 0.373,
+      "step": 24350
+    },
+    {
+      "epoch": 9.195922989807475,
+      "grad_norm": 0.8601105213165283,
+      "learning_rate": 3.6292763093973425e-05,
+      "loss": 0.3697,
+      "step": 24360
+    },
+    {
+      "epoch": 9.199697999244998,
+      "grad_norm": 0.6811267733573914,
+      "learning_rate": 3.6253014712497754e-05,
+      "loss": 0.3826,
+      "step": 24370
+    },
+    {
+      "epoch": 9.203473008682522,
+      "grad_norm": 0.6935433745384216,
+      "learning_rate": 3.621327572695692e-05,
+      "loss": 0.3782,
+      "step": 24380
+    },
+    {
+      "epoch": 9.207248018120046,
+      "grad_norm": 0.5961577892303467,
+      "learning_rate": 3.617354616451211e-05,
+      "loss": 0.375,
+      "step": 24390
+    },
+    {
+      "epoch": 9.21102302755757,
+      "grad_norm": 0.5627548694610596,
+      "learning_rate": 3.6133826052318116e-05,
+      "loss": 0.3731,
+      "step": 24400
+    },
+    {
+      "epoch": 9.214798036995093,
+      "grad_norm": 0.6557216048240662,
+      "learning_rate": 3.609411541752327e-05,
+      "loss": 0.3741,
+      "step": 24410
+    },
+    {
+      "epoch": 9.218573046432615,
+      "grad_norm": 0.7413145899772644,
+      "learning_rate": 3.6054414287269405e-05,
+      "loss": 0.3712,
+      "step": 24420
+    },
+    {
+      "epoch": 9.222348055870139,
+      "grad_norm": 0.6858183145523071,
+      "learning_rate": 3.601472268869188e-05,
+      "loss": 0.3814,
+      "step": 24430
+    },
+    {
+      "epoch": 9.226123065307663,
+      "grad_norm": 0.6973282098770142,
+      "learning_rate": 3.597504064891952e-05,
+      "loss": 0.377,
+      "step": 24440
+    },
+    {
+      "epoch": 9.229898074745186,
+      "grad_norm": 0.6426398158073425,
+      "learning_rate": 3.5935368195074636e-05,
+      "loss": 0.3757,
+      "step": 24450
+    },
+    {
+      "epoch": 9.23367308418271,
+      "grad_norm": 0.6414214372634888,
+      "learning_rate": 3.589570535427297e-05,
+      "loss": 0.3642,
+      "step": 24460
+    },
+    {
+      "epoch": 9.237448093620234,
+      "grad_norm": 0.6278491020202637,
+      "learning_rate": 3.585605215362371e-05,
+      "loss": 0.3778,
+      "step": 24470
+    },
+    {
+      "epoch": 9.241223103057758,
+      "grad_norm": 0.695353090763092,
+      "learning_rate": 3.581640862022941e-05,
+      "loss": 0.3763,
+      "step": 24480
+    },
+    {
+      "epoch": 9.244998112495281,
+      "grad_norm": 0.6020937561988831,
+      "learning_rate": 3.57767747811861e-05,
+      "loss": 0.3761,
+      "step": 24490
+    },
+    {
+      "epoch": 9.248773121932805,
+      "grad_norm": 0.7220026254653931,
+      "learning_rate": 3.573715066358308e-05,
+      "loss": 0.3751,
+      "step": 24500
+    },
+    {
+      "epoch": 9.252548131370329,
+      "grad_norm": 0.6463865041732788,
+      "learning_rate": 3.569753629450311e-05,
+      "loss": 0.3783,
+      "step": 24510
+    },
+    {
+      "epoch": 9.256323140807853,
+      "grad_norm": 0.6139450073242188,
+      "learning_rate": 3.565793170102221e-05,
+      "loss": 0.3754,
+      "step": 24520
+    },
+    {
+      "epoch": 9.260098150245376,
+      "grad_norm": 0.6605015993118286,
+      "learning_rate": 3.561833691020976e-05,
+      "loss": 0.3685,
+      "step": 24530
+    },
+    {
+      "epoch": 9.2638731596829,
+      "grad_norm": 0.6926522254943848,
+      "learning_rate": 3.5578751949128415e-05,
+      "loss": 0.374,
+      "step": 24540
+    },
+    {
+      "epoch": 9.267648169120422,
+      "grad_norm": 0.6264640092849731,
+      "learning_rate": 3.5539176844834125e-05,
+      "loss": 0.3783,
+      "step": 24550
+    },
+    {
+      "epoch": 9.271423178557946,
+      "grad_norm": 0.9688851237297058,
+      "learning_rate": 3.5499611624376125e-05,
+      "loss": 0.3795,
+      "step": 24560
+    },
+    {
+      "epoch": 9.27519818799547,
+      "grad_norm": 0.5947738885879517,
+      "learning_rate": 3.546005631479684e-05,
+      "loss": 0.3662,
+      "step": 24570
+    },
+    {
+      "epoch": 9.278973197432993,
+      "grad_norm": 0.6355003714561462,
+      "learning_rate": 3.542051094313196e-05,
+      "loss": 0.3785,
+      "step": 24580
+    },
+    {
+      "epoch": 9.282748206870517,
+      "grad_norm": 0.6188831329345703,
+      "learning_rate": 3.5380975536410364e-05,
+      "loss": 0.3804,
+      "step": 24590
+    },
+    {
+      "epoch": 9.28652321630804,
+      "grad_norm": 0.6349472403526306,
+      "learning_rate": 3.534145012165415e-05,
+      "loss": 0.3757,
+      "step": 24600
+    },
+    {
+      "epoch": 9.290298225745564,
+      "grad_norm": 0.6269373297691345,
+      "learning_rate": 3.5301934725878546e-05,
+      "loss": 0.3698,
+      "step": 24610
+    },
+    {
+      "epoch": 9.294073235183088,
+      "grad_norm": 0.6995264291763306,
+      "learning_rate": 3.526242937609197e-05,
+      "loss": 0.3798,
+      "step": 24620
+    },
+    {
+      "epoch": 9.297848244620612,
+      "grad_norm": 0.6309302449226379,
+      "learning_rate": 3.522293409929595e-05,
+      "loss": 0.3652,
+      "step": 24630
+    },
+    {
+      "epoch": 9.301623254058136,
+      "grad_norm": 0.8019225597381592,
+      "learning_rate": 3.518344892248513e-05,
+      "loss": 0.3808,
+      "step": 24640
+    },
+    {
+      "epoch": 9.30539826349566,
+      "grad_norm": 0.6818584203720093,
+      "learning_rate": 3.514397387264725e-05,
+      "loss": 0.3726,
+      "step": 24650
+    },
+    {
+      "epoch": 9.309173272933183,
+      "grad_norm": 0.6697105169296265,
+      "learning_rate": 3.5104508976763176e-05,
+      "loss": 0.3678,
+      "step": 24660
+    },
+    {
+      "epoch": 9.312948282370705,
+      "grad_norm": 0.5741595029830933,
+      "learning_rate": 3.506505426180674e-05,
+      "loss": 0.38,
+      "step": 24670
+    },
+    {
+      "epoch": 9.316723291808229,
+      "grad_norm": 0.6436684131622314,
+      "learning_rate": 3.502560975474488e-05,
+      "loss": 0.3747,
+      "step": 24680
+    },
+    {
+      "epoch": 9.320498301245753,
+      "grad_norm": 0.6101210117340088,
+      "learning_rate": 3.4986175482537566e-05,
+      "loss": 0.3678,
+      "step": 24690
+    },
+    {
+      "epoch": 9.324273310683276,
+      "grad_norm": 0.8269137144088745,
+      "learning_rate": 3.4946751472137725e-05,
+      "loss": 0.3721,
+      "step": 24700
+    },
+    {
+      "epoch": 9.3280483201208,
+      "grad_norm": 0.6154971718788147,
+      "learning_rate": 3.490733775049132e-05,
+      "loss": 0.379,
+      "step": 24710
+    },
+    {
+      "epoch": 9.331823329558324,
+      "grad_norm": 0.642996072769165,
+      "learning_rate": 3.4867934344537236e-05,
+      "loss": 0.3823,
+      "step": 24720
+    },
+    {
+      "epoch": 9.335598338995847,
+      "grad_norm": 0.6304564476013184,
+      "learning_rate": 3.482854128120735e-05,
+      "loss": 0.3742,
+      "step": 24730
+    },
+    {
+      "epoch": 9.339373348433371,
+      "grad_norm": 0.6240081787109375,
+      "learning_rate": 3.478915858742643e-05,
+      "loss": 0.3698,
+      "step": 24740
+    },
+    {
+      "epoch": 9.343148357870895,
+      "grad_norm": 0.6386808156967163,
+      "learning_rate": 3.4749786290112205e-05,
+      "loss": 0.3709,
+      "step": 24750
+    },
+    {
+      "epoch": 9.346923367308419,
+      "grad_norm": 0.6619960069656372,
+      "learning_rate": 3.471042441617524e-05,
+      "loss": 0.3689,
+      "step": 24760
+    },
+    {
+      "epoch": 9.350698376745942,
+      "grad_norm": 0.8115395903587341,
+      "learning_rate": 3.467107299251902e-05,
+      "loss": 0.3721,
+      "step": 24770
+    },
+    {
+      "epoch": 9.354473386183466,
+      "grad_norm": 0.6916590929031372,
+      "learning_rate": 3.463173204603984e-05,
+      "loss": 0.3736,
+      "step": 24780
+    },
+    {
+      "epoch": 9.35824839562099,
+      "grad_norm": 0.6382922530174255,
+      "learning_rate": 3.4592401603626924e-05,
+      "loss": 0.3715,
+      "step": 24790
+    },
+    {
+      "epoch": 9.362023405058512,
+      "grad_norm": 0.6073492765426636,
+      "learning_rate": 3.45530816921622e-05,
+      "loss": 0.3714,
+      "step": 24800
+    },
+    {
+      "epoch": 9.365798414496036,
+      "grad_norm": 0.6381345987319946,
+      "learning_rate": 3.451377233852051e-05,
+      "loss": 0.3758,
+      "step": 24810
+    },
+    {
+      "epoch": 9.36957342393356,
+      "grad_norm": 0.5805862545967102,
+      "learning_rate": 3.4474473569569385e-05,
+      "loss": 0.37,
+      "step": 24820
+    },
+    {
+      "epoch": 9.373348433371083,
+      "grad_norm": 0.7061961889266968,
+      "learning_rate": 3.443518541216918e-05,
+      "loss": 0.3721,
+      "step": 24830
+    },
+    {
+      "epoch": 9.377123442808607,
+      "grad_norm": 0.6575196981430054,
+      "learning_rate": 3.439590789317299e-05,
+      "loss": 0.3731,
+      "step": 24840
+    },
+    {
+      "epoch": 9.38089845224613,
+      "grad_norm": 0.6737225651741028,
+      "learning_rate": 3.4356641039426607e-05,
+      "loss": 0.3712,
+      "step": 24850
+    },
+    {
+      "epoch": 9.384673461683654,
+      "grad_norm": 0.7008658051490784,
+      "learning_rate": 3.431738487776857e-05,
+      "loss": 0.3713,
+      "step": 24860
+    },
+    {
+      "epoch": 9.388448471121178,
+      "grad_norm": 0.6766116619110107,
+      "learning_rate": 3.4278139435030084e-05,
+      "loss": 0.3772,
+      "step": 24870
+    },
+    {
+      "epoch": 9.392223480558702,
+      "grad_norm": 0.6456305384635925,
+      "learning_rate": 3.423890473803504e-05,
+      "loss": 0.3804,
+      "step": 24880
+    },
+    {
+      "epoch": 9.395998489996225,
+      "grad_norm": 0.7182512879371643,
+      "learning_rate": 3.41996808136e-05,
+      "loss": 0.37,
+      "step": 24890
+    },
+    {
+      "epoch": 9.39977349943375,
+      "grad_norm": 0.6886579990386963,
+      "learning_rate": 3.416046768853413e-05,
+      "loss": 0.3696,
+      "step": 24900
+    },
+    {
+      "epoch": 9.403548508871273,
+      "grad_norm": 0.6319290399551392,
+      "learning_rate": 3.412126538963925e-05,
+      "loss": 0.3771,
+      "step": 24910
+    },
+    {
+      "epoch": 9.407323518308797,
+      "grad_norm": 0.5859635472297668,
+      "learning_rate": 3.4082073943709727e-05,
+      "loss": 0.3729,
+      "step": 24920
+    },
+    {
+      "epoch": 9.411098527746319,
+      "grad_norm": 0.7094352841377258,
+      "learning_rate": 3.404289337753258e-05,
+      "loss": 0.374,
+      "step": 24930
+    },
+    {
+      "epoch": 9.414873537183842,
+      "grad_norm": 0.6192387938499451,
+      "learning_rate": 3.400372371788736e-05,
+      "loss": 0.3749,
+      "step": 24940
+    },
+    {
+      "epoch": 9.418648546621366,
+      "grad_norm": 0.5935202836990356,
+      "learning_rate": 3.3964564991546124e-05,
+      "loss": 0.3707,
+      "step": 24950
+    },
+    {
+      "epoch": 9.42242355605889,
+      "grad_norm": 0.6603958606719971,
+      "learning_rate": 3.392541722527351e-05,
+      "loss": 0.3825,
+      "step": 24960
+    },
+    {
+      "epoch": 9.426198565496414,
+      "grad_norm": 0.6225345134735107,
+      "learning_rate": 3.3886280445826644e-05,
+      "loss": 0.3669,
+      "step": 24970
+    },
+    {
+      "epoch": 9.429973574933937,
+      "grad_norm": 0.6600792407989502,
+      "learning_rate": 3.3847154679955154e-05,
+      "loss": 0.373,
+      "step": 24980
+    },
+    {
+      "epoch": 9.433748584371461,
+      "grad_norm": 0.6622351408004761,
+      "learning_rate": 3.380803995440113e-05,
+      "loss": 0.3696,
+      "step": 24990
+    },
+    {
+      "epoch": 9.437523593808985,
+      "grad_norm": 0.6164969801902771,
+      "learning_rate": 3.3768936295899115e-05,
+      "loss": 0.3682,
+      "step": 25000
+    },
+    {
+      "epoch": 9.441298603246508,
+      "grad_norm": 0.704637348651886,
+      "learning_rate": 3.3729843731176094e-05,
+      "loss": 0.3734,
+      "step": 25010
+    },
+    {
+      "epoch": 9.445073612684032,
+      "grad_norm": 0.6271415948867798,
+      "learning_rate": 3.369076228695146e-05,
+      "loss": 0.3691,
+      "step": 25020
+    },
+    {
+      "epoch": 9.448848622121556,
+      "grad_norm": 0.6947619318962097,
+      "learning_rate": 3.365169198993703e-05,
+      "loss": 0.3781,
+      "step": 25030
+    },
+    {
+      "epoch": 9.45262363155908,
+      "grad_norm": 0.6423031091690063,
+      "learning_rate": 3.361263286683697e-05,
+      "loss": 0.3713,
+      "step": 25040
+    },
+    {
+      "epoch": 9.456398640996602,
+      "grad_norm": 0.6936047673225403,
+      "learning_rate": 3.35735849443478e-05,
+      "loss": 0.3714,
+      "step": 25050
+    },
+    {
+      "epoch": 9.460173650434125,
+      "grad_norm": 0.6276016235351562,
+      "learning_rate": 3.3534548249158435e-05,
+      "loss": 0.3695,
+      "step": 25060
+    },
+    {
+      "epoch": 9.46394865987165,
+      "grad_norm": 0.6831814646720886,
+      "learning_rate": 3.3495522807950086e-05,
+      "loss": 0.3706,
+      "step": 25070
+    },
+    {
+      "epoch": 9.467723669309173,
+      "grad_norm": 0.6282784342765808,
+      "learning_rate": 3.345650864739627e-05,
+      "loss": 0.3717,
+      "step": 25080
+    },
+    {
+      "epoch": 9.471498678746697,
+      "grad_norm": 0.6343705654144287,
+      "learning_rate": 3.3417505794162794e-05,
+      "loss": 0.3685,
+      "step": 25090
+    },
+    {
+      "epoch": 9.47527368818422,
+      "grad_norm": 0.6878871917724609,
+      "learning_rate": 3.3378514274907745e-05,
+      "loss": 0.3773,
+      "step": 25100
+    },
+    {
+      "epoch": 9.479048697621744,
+      "grad_norm": 0.7008819580078125,
+      "learning_rate": 3.333953411628147e-05,
+      "loss": 0.3681,
+      "step": 25110
+    },
+    {
+      "epoch": 9.482823707059268,
+      "grad_norm": 0.653729259967804,
+      "learning_rate": 3.330056534492653e-05,
+      "loss": 0.3794,
+      "step": 25120
+    },
+    {
+      "epoch": 9.486598716496792,
+      "grad_norm": 0.7341142296791077,
+      "learning_rate": 3.32616079874777e-05,
+      "loss": 0.3691,
+      "step": 25130
+    },
+    {
+      "epoch": 9.490373725934315,
+      "grad_norm": 0.9911676049232483,
+      "learning_rate": 3.322266207056197e-05,
+      "loss": 0.37,
+      "step": 25140
+    },
+    {
+      "epoch": 9.494148735371839,
+      "grad_norm": 0.6838816404342651,
+      "learning_rate": 3.318372762079852e-05,
+      "loss": 0.3757,
+      "step": 25150
+    },
+    {
+      "epoch": 9.497923744809363,
+      "grad_norm": 0.7315078973770142,
+      "learning_rate": 3.3144804664798666e-05,
+      "loss": 0.3732,
+      "step": 25160
+    },
+    {
+      "epoch": 9.501698754246885,
+      "grad_norm": 0.6944950222969055,
+      "learning_rate": 3.3105893229165894e-05,
+      "loss": 0.3777,
+      "step": 25170
+    },
+    {
+      "epoch": 9.505473763684408,
+      "grad_norm": 0.6807419657707214,
+      "learning_rate": 3.30669933404958e-05,
+      "loss": 0.3688,
+      "step": 25180
+    },
+    {
+      "epoch": 9.509248773121932,
+      "grad_norm": 0.7115993499755859,
+      "learning_rate": 3.302810502537609e-05,
+      "loss": 0.3691,
+      "step": 25190
+    },
+    {
+      "epoch": 9.513023782559456,
+      "grad_norm": 0.8097192645072937,
+      "learning_rate": 3.298922831038655e-05,
+      "loss": 0.3776,
+      "step": 25200
+    },
+    {
+      "epoch": 9.51679879199698,
+      "grad_norm": 0.7089084386825562,
+      "learning_rate": 3.2950363222099073e-05,
+      "loss": 0.3708,
+      "step": 25210
+    },
+    {
+      "epoch": 9.520573801434503,
+      "grad_norm": 0.6845264434814453,
+      "learning_rate": 3.291150978707758e-05,
+      "loss": 0.3715,
+      "step": 25220
+    },
+    {
+      "epoch": 9.524348810872027,
+      "grad_norm": 0.6236247420310974,
+      "learning_rate": 3.287266803187798e-05,
+      "loss": 0.3671,
+      "step": 25230
+    },
+    {
+      "epoch": 9.52812382030955,
+      "grad_norm": 0.7404671907424927,
+      "learning_rate": 3.283383798304829e-05,
+      "loss": 0.3735,
+      "step": 25240
+    },
+    {
+      "epoch": 9.531898829747075,
+      "grad_norm": 0.6793319582939148,
+      "learning_rate": 3.279501966712847e-05,
+      "loss": 0.3674,
+      "step": 25250
+    },
+    {
+      "epoch": 9.535673839184598,
+      "grad_norm": 0.7164490222930908,
+      "learning_rate": 3.275621311065047e-05,
+      "loss": 0.3726,
+      "step": 25260
+    },
+    {
+      "epoch": 9.539448848622122,
+      "grad_norm": 0.8006794452667236,
+      "learning_rate": 3.271741834013822e-05,
+      "loss": 0.3671,
+      "step": 25270
+    },
+    {
+      "epoch": 9.543223858059646,
+      "grad_norm": 0.6206616759300232,
+      "learning_rate": 3.267863538210756e-05,
+      "loss": 0.3736,
+      "step": 25280
+    },
+    {
+      "epoch": 9.54699886749717,
+      "grad_norm": 0.6397393941879272,
+      "learning_rate": 3.2639864263066296e-05,
+      "loss": 0.3697,
+      "step": 25290
+    },
+    {
+      "epoch": 9.550773876934691,
+      "grad_norm": 1.0796784162521362,
+      "learning_rate": 3.26011050095141e-05,
+      "loss": 0.3907,
+      "step": 25300
+    },
+    {
+      "epoch": 9.554548886372215,
+      "grad_norm": 2.355830192565918,
+      "learning_rate": 3.256235764794259e-05,
+      "loss": 0.3707,
+      "step": 25310
+    },
+    {
+      "epoch": 9.558323895809739,
+      "grad_norm": 0.6870741844177246,
+      "learning_rate": 3.2523622204835194e-05,
+      "loss": 0.3682,
+      "step": 25320
+    },
+    {
+      "epoch": 9.562098905247263,
+      "grad_norm": 0.7428625822067261,
+      "learning_rate": 3.2484898706667214e-05,
+      "loss": 0.369,
+      "step": 25330
+    },
+    {
+      "epoch": 9.565873914684786,
+      "grad_norm": 0.7070319652557373,
+      "learning_rate": 3.2446187179905806e-05,
+      "loss": 0.3685,
+      "step": 25340
+    },
+    {
+      "epoch": 9.56964892412231,
+      "grad_norm": 0.6212445497512817,
+      "learning_rate": 3.240748765100995e-05,
+      "loss": 0.3802,
+      "step": 25350
+    },
+    {
+      "epoch": 9.573423933559834,
+      "grad_norm": 0.6794653534889221,
+      "learning_rate": 3.236880014643039e-05,
+      "loss": 0.3665,
+      "step": 25360
+    },
+    {
+      "epoch": 9.577198942997358,
+      "grad_norm": 0.6495829224586487,
+      "learning_rate": 3.233012469260969e-05,
+      "loss": 0.3645,
+      "step": 25370
+    },
+    {
+      "epoch": 9.580973952434881,
+      "grad_norm": 0.6042360067367554,
+      "learning_rate": 3.229146131598213e-05,
+      "loss": 0.3745,
+      "step": 25380
+    },
+    {
+      "epoch": 9.584748961872405,
+      "grad_norm": 2.035146713256836,
+      "learning_rate": 3.2252810042973794e-05,
+      "loss": 0.3827,
+      "step": 25390
+    },
+    {
+      "epoch": 9.588523971309929,
+      "grad_norm": 0.5512405633926392,
+      "learning_rate": 3.2214170900002456e-05,
+      "loss": 0.3873,
+      "step": 25400
+    },
+    {
+      "epoch": 9.592298980747453,
+      "grad_norm": 0.7587074041366577,
+      "learning_rate": 3.217554391347758e-05,
+      "loss": 0.3671,
+      "step": 25410
+    },
+    {
+      "epoch": 9.596073990184976,
+      "grad_norm": 0.7429296970367432,
+      "learning_rate": 3.213692910980037e-05,
+      "loss": 0.3708,
+      "step": 25420
+    },
+    {
+      "epoch": 9.599848999622498,
+      "grad_norm": 0.6271414160728455,
+      "learning_rate": 3.2098326515363666e-05,
+      "loss": 0.3715,
+      "step": 25430
+    },
+    {
+      "epoch": 9.603624009060022,
+      "grad_norm": 0.6165705919265747,
+      "learning_rate": 3.205973615655199e-05,
+      "loss": 0.3669,
+      "step": 25440
+    },
+    {
+      "epoch": 9.607399018497546,
+      "grad_norm": 0.5890395641326904,
+      "learning_rate": 3.202115805974149e-05,
+      "loss": 0.3681,
+      "step": 25450
+    },
+    {
+      "epoch": 9.61117402793507,
+      "grad_norm": 0.6870768070220947,
+      "learning_rate": 3.1982592251299916e-05,
+      "loss": 0.3736,
+      "step": 25460
+    },
+    {
+      "epoch": 9.614949037372593,
+      "grad_norm": 0.7155221700668335,
+      "learning_rate": 3.1944038757586656e-05,
+      "loss": 0.3716,
+      "step": 25470
+    },
+    {
+      "epoch": 9.618724046810117,
+      "grad_norm": 0.6171225905418396,
+      "learning_rate": 3.190549760495263e-05,
+      "loss": 0.3712,
+      "step": 25480
+    },
+    {
+      "epoch": 9.62249905624764,
+      "grad_norm": 0.7019498944282532,
+      "learning_rate": 3.186696881974039e-05,
+      "loss": 0.3641,
+      "step": 25490
+    },
+    {
+      "epoch": 9.626274065685164,
+      "grad_norm": 0.7143846154212952,
+      "learning_rate": 3.1828452428283986e-05,
+      "loss": 0.3769,
+      "step": 25500
+    },
+    {
+      "epoch": 9.630049075122688,
+      "grad_norm": 0.6226781606674194,
+      "learning_rate": 3.178994845690898e-05,
+      "loss": 0.3636,
+      "step": 25510
+    },
+    {
+      "epoch": 9.633824084560212,
+      "grad_norm": 0.6864441633224487,
+      "learning_rate": 3.17514569319325e-05,
+      "loss": 0.3741,
+      "step": 25520
+    },
+    {
+      "epoch": 9.637599093997736,
+      "grad_norm": 0.6476948261260986,
+      "learning_rate": 3.171297787966312e-05,
+      "loss": 0.3763,
+      "step": 25530
+    },
+    {
+      "epoch": 9.64137410343526,
+      "grad_norm": 0.6583799123764038,
+      "learning_rate": 3.167451132640093e-05,
+      "loss": 0.3715,
+      "step": 25540
+    },
+    {
+      "epoch": 9.645149112872783,
+      "grad_norm": 0.8483378291130066,
+      "learning_rate": 3.163605729843746e-05,
+      "loss": 0.3742,
+      "step": 25550
+    },
+    {
+      "epoch": 9.648924122310305,
+      "grad_norm": 0.7014634609222412,
+      "learning_rate": 3.159761582205565e-05,
+      "loss": 0.3693,
+      "step": 25560
+    },
+    {
+      "epoch": 9.652699131747829,
+      "grad_norm": 0.8034195899963379,
+      "learning_rate": 3.155918692352992e-05,
+      "loss": 0.3697,
+      "step": 25570
+    },
+    {
+      "epoch": 9.656474141185353,
+      "grad_norm": 0.7062366008758545,
+      "learning_rate": 3.152077062912602e-05,
+      "loss": 0.3776,
+      "step": 25580
+    },
+    {
+      "epoch": 9.660249150622876,
+      "grad_norm": 0.6510414481163025,
+      "learning_rate": 3.148236696510117e-05,
+      "loss": 0.3764,
+      "step": 25590
+    },
+    {
+      "epoch": 9.6640241600604,
+      "grad_norm": 0.7279618978500366,
+      "learning_rate": 3.144397595770388e-05,
+      "loss": 0.3773,
+      "step": 25600
+    },
+    {
+      "epoch": 9.667799169497924,
+      "grad_norm": 0.6099146008491516,
+      "learning_rate": 3.1405597633174036e-05,
+      "loss": 0.376,
+      "step": 25610
+    },
+    {
+      "epoch": 9.671574178935447,
+      "grad_norm": 0.6417626142501831,
+      "learning_rate": 3.136723201774289e-05,
+      "loss": 0.3717,
+      "step": 25620
+    },
+    {
+      "epoch": 9.675349188372971,
+      "grad_norm": 0.7016961574554443,
+      "learning_rate": 3.132887913763295e-05,
+      "loss": 0.3698,
+      "step": 25630
+    },
+    {
+      "epoch": 9.679124197810495,
+      "grad_norm": 0.7685357928276062,
+      "learning_rate": 3.129053901905806e-05,
+      "loss": 0.3648,
+      "step": 25640
+    },
+    {
+      "epoch": 9.682899207248019,
+      "grad_norm": 0.7496814131736755,
+      "learning_rate": 3.125221168822335e-05,
+      "loss": 0.3781,
+      "step": 25650
+    },
+    {
+      "epoch": 9.686674216685542,
+      "grad_norm": 0.6221609711647034,
+      "learning_rate": 3.1213897171325154e-05,
+      "loss": 0.3694,
+      "step": 25660
+    },
+    {
+      "epoch": 9.690449226123066,
+      "grad_norm": 0.6706365942955017,
+      "learning_rate": 3.1175595494551116e-05,
+      "loss": 0.3709,
+      "step": 25670
+    },
+    {
+      "epoch": 9.694224235560588,
+      "grad_norm": 0.5938177704811096,
+      "learning_rate": 3.1137306684080045e-05,
+      "loss": 0.3659,
+      "step": 25680
+    },
+    {
+      "epoch": 9.697999244998112,
+      "grad_norm": 0.69608074426651,
+      "learning_rate": 3.1099030766081985e-05,
+      "loss": 0.3712,
+      "step": 25690
+    },
+    {
+      "epoch": 9.701774254435636,
+      "grad_norm": 0.6536232829093933,
+      "learning_rate": 3.106076776671818e-05,
+      "loss": 0.3678,
+      "step": 25700
+    },
+    {
+      "epoch": 9.70554926387316,
+      "grad_norm": 0.6710280179977417,
+      "learning_rate": 3.102251771214101e-05,
+      "loss": 0.3667,
+      "step": 25710
+    },
+    {
+      "epoch": 9.709324273310683,
+      "grad_norm": 0.6235374808311462,
+      "learning_rate": 3.098428062849404e-05,
+      "loss": 0.3684,
+      "step": 25720
+    },
+    {
+      "epoch": 9.713099282748207,
+      "grad_norm": 0.6490816473960876,
+      "learning_rate": 3.094605654191195e-05,
+      "loss": 0.3715,
+      "step": 25730
+    },
+    {
+      "epoch": 9.71687429218573,
+      "grad_norm": 0.6581624746322632,
+      "learning_rate": 3.090784547852055e-05,
+      "loss": 0.3714,
+      "step": 25740
+    },
+    {
+      "epoch": 9.720649301623254,
+      "grad_norm": 0.7839778065681458,
+      "learning_rate": 3.0869647464436746e-05,
+      "loss": 0.373,
+      "step": 25750
+    },
+    {
+      "epoch": 9.724424311060778,
+      "grad_norm": 0.6403361558914185,
+      "learning_rate": 3.0831462525768496e-05,
+      "loss": 0.3684,
+      "step": 25760
+    },
+    {
+      "epoch": 9.728199320498302,
+      "grad_norm": 0.6090444326400757,
+      "learning_rate": 3.079329068861488e-05,
+      "loss": 0.3671,
+      "step": 25770
+    },
+    {
+      "epoch": 9.731974329935825,
+      "grad_norm": 0.6840675473213196,
+      "learning_rate": 3.075513197906597e-05,
+      "loss": 0.3682,
+      "step": 25780
+    },
+    {
+      "epoch": 9.73574933937335,
+      "grad_norm": 0.6021028757095337,
+      "learning_rate": 3.071698642320286e-05,
+      "loss": 0.3723,
+      "step": 25790
+    },
+    {
+      "epoch": 9.739524348810871,
+      "grad_norm": 0.6243388652801514,
+      "learning_rate": 3.067885404709772e-05,
+      "loss": 0.371,
+      "step": 25800
+    },
+    {
+      "epoch": 9.743299358248395,
+      "grad_norm": 0.6200324296951294,
+      "learning_rate": 3.0640734876813636e-05,
+      "loss": 0.3729,
+      "step": 25810
+    },
+    {
+      "epoch": 9.747074367685919,
+      "grad_norm": 1.0078961849212646,
+      "learning_rate": 3.060262893840473e-05,
+      "loss": 0.3695,
+      "step": 25820
+    },
+    {
+      "epoch": 9.750849377123442,
+      "grad_norm": 0.6697998642921448,
+      "learning_rate": 3.056453625791603e-05,
+      "loss": 0.3641,
+      "step": 25830
+    },
+    {
+      "epoch": 9.754624386560966,
+      "grad_norm": 0.8702282905578613,
+      "learning_rate": 3.052645686138353e-05,
+      "loss": 0.3734,
+      "step": 25840
+    },
+    {
+      "epoch": 9.75839939599849,
+      "grad_norm": 0.6085755825042725,
+      "learning_rate": 3.0488390774834153e-05,
+      "loss": 0.3702,
+      "step": 25850
+    },
+    {
+      "epoch": 9.762174405436014,
+      "grad_norm": 0.7779916524887085,
+      "learning_rate": 3.0450338024285684e-05,
+      "loss": 0.3718,
+      "step": 25860
+    },
+    {
+      "epoch": 9.765949414873537,
+      "grad_norm": 0.6143949627876282,
+      "learning_rate": 3.0412298635746855e-05,
+      "loss": 0.3663,
+      "step": 25870
+    },
+    {
+      "epoch": 9.769724424311061,
+      "grad_norm": 0.7032211422920227,
+      "learning_rate": 3.03742726352172e-05,
+      "loss": 0.3688,
+      "step": 25880
+    },
+    {
+      "epoch": 9.773499433748585,
+      "grad_norm": 0.6979837417602539,
+      "learning_rate": 3.0336260048687125e-05,
+      "loss": 0.3648,
+      "step": 25890
+    },
+    {
+      "epoch": 9.777274443186108,
+      "grad_norm": 0.6320093274116516,
+      "learning_rate": 3.0298260902137897e-05,
+      "loss": 0.3666,
+      "step": 25900
+    },
+    {
+      "epoch": 9.781049452623632,
+      "grad_norm": 0.6914851665496826,
+      "learning_rate": 3.0260275221541566e-05,
+      "loss": 0.3743,
+      "step": 25910
+    },
+    {
+      "epoch": 9.784824462061156,
+      "grad_norm": 0.6460880041122437,
+      "learning_rate": 3.0222303032860987e-05,
+      "loss": 0.376,
+      "step": 25920
+    },
+    {
+      "epoch": 9.788599471498678,
+      "grad_norm": 0.7050313353538513,
+      "learning_rate": 3.018434436204979e-05,
+      "loss": 0.3655,
+      "step": 25930
+    },
+    {
+      "epoch": 9.792374480936202,
+      "grad_norm": 0.7031101584434509,
+      "learning_rate": 3.014639923505237e-05,
+      "loss": 0.367,
+      "step": 25940
+    },
+    {
+      "epoch": 9.796149490373725,
+      "grad_norm": 0.6252807378768921,
+      "learning_rate": 3.0108467677803863e-05,
+      "loss": 0.3669,
+      "step": 25950
+    },
+    {
+      "epoch": 9.79992449981125,
+      "grad_norm": 0.6728451251983643,
+      "learning_rate": 3.0070549716230156e-05,
+      "loss": 0.3685,
+      "step": 25960
+    },
+    {
+      "epoch": 9.803699509248773,
+      "grad_norm": 0.6584720015525818,
+      "learning_rate": 3.003264537624777e-05,
+      "loss": 0.3704,
+      "step": 25970
+    },
+    {
+      "epoch": 9.807474518686297,
+      "grad_norm": 0.5711626410484314,
+      "learning_rate": 2.9994754683764e-05,
+      "loss": 0.3833,
+      "step": 25980
+    },
+    {
+      "epoch": 9.81124952812382,
+      "grad_norm": 0.7549001574516296,
+      "learning_rate": 2.9956877664676754e-05,
+      "loss": 0.3661,
+      "step": 25990
+    },
+    {
+      "epoch": 9.815024537561344,
+      "grad_norm": 0.6147065162658691,
+      "learning_rate": 2.9919014344874636e-05,
+      "loss": 0.3647,
+      "step": 26000
+    },
+    {
+      "epoch": 9.818799546998868,
+      "grad_norm": 0.7019761800765991,
+      "learning_rate": 2.9881164750236857e-05,
+      "loss": 0.3668,
+      "step": 26010
+    },
+    {
+      "epoch": 9.822574556436392,
+      "grad_norm": 0.6753397583961487,
+      "learning_rate": 2.984332890663326e-05,
+      "loss": 0.3729,
+      "step": 26020
+    },
+    {
+      "epoch": 9.826349565873915,
+      "grad_norm": 0.6772192120552063,
+      "learning_rate": 2.9805506839924292e-05,
+      "loss": 0.3735,
+      "step": 26030
+    },
+    {
+      "epoch": 9.830124575311439,
+      "grad_norm": 0.6899954080581665,
+      "learning_rate": 2.9767698575960968e-05,
+      "loss": 0.3628,
+      "step": 26040
+    },
+    {
+      "epoch": 9.833899584748963,
+      "grad_norm": 0.6952143907546997,
+      "learning_rate": 2.9729904140584913e-05,
+      "loss": 0.3724,
+      "step": 26050
+    },
+    {
+      "epoch": 9.837674594186485,
+      "grad_norm": 0.6387107968330383,
+      "learning_rate": 2.9692123559628234e-05,
+      "loss": 0.368,
+      "step": 26060
+    },
+    {
+      "epoch": 9.841449603624008,
+      "grad_norm": 0.6607224345207214,
+      "learning_rate": 2.9654356858913596e-05,
+      "loss": 0.3642,
+      "step": 26070
+    },
+    {
+      "epoch": 9.845224613061532,
+      "grad_norm": 0.6394532322883606,
+      "learning_rate": 2.9616604064254206e-05,
+      "loss": 0.3641,
+      "step": 26080
+    },
+    {
+      "epoch": 9.848999622499056,
+      "grad_norm": 0.6913762092590332,
+      "learning_rate": 2.9578865201453732e-05,
+      "loss": 0.3614,
+      "step": 26090
+    },
+    {
+      "epoch": 9.85277463193658,
+      "grad_norm": 0.7300879955291748,
+      "learning_rate": 2.9541140296306335e-05,
+      "loss": 0.3701,
+      "step": 26100
+    },
+    {
+      "epoch": 9.856549641374103,
+      "grad_norm": 0.6649528741836548,
+      "learning_rate": 2.9503429374596627e-05,
+      "loss": 0.3746,
+      "step": 26110
+    },
+    {
+      "epoch": 9.860324650811627,
+      "grad_norm": 0.693498432636261,
+      "learning_rate": 2.946573246209967e-05,
+      "loss": 0.3665,
+      "step": 26120
+    },
+    {
+      "epoch": 9.86409966024915,
+      "grad_norm": 0.8667472004890442,
+      "learning_rate": 2.942804958458094e-05,
+      "loss": 0.3788,
+      "step": 26130
+    },
+    {
+      "epoch": 9.867874669686675,
+      "grad_norm": 0.6629686951637268,
+      "learning_rate": 2.9390380767796343e-05,
+      "loss": 0.3673,
+      "step": 26140
+    },
+    {
+      "epoch": 9.871649679124198,
+      "grad_norm": 0.6008899211883545,
+      "learning_rate": 2.9352726037492174e-05,
+      "loss": 0.3666,
+      "step": 26150
+    },
+    {
+      "epoch": 9.875424688561722,
+      "grad_norm": 0.7541015148162842,
+      "learning_rate": 2.9315085419405052e-05,
+      "loss": 0.364,
+      "step": 26160
+    },
+    {
+      "epoch": 9.879199697999246,
+      "grad_norm": 0.6057034730911255,
+      "learning_rate": 2.927745893926199e-05,
+      "loss": 0.3711,
+      "step": 26170
+    },
+    {
+      "epoch": 9.88297470743677,
+      "grad_norm": 0.6768243908882141,
+      "learning_rate": 2.9239846622780358e-05,
+      "loss": 0.3707,
+      "step": 26180
+    },
+    {
+      "epoch": 9.886749716874291,
+      "grad_norm": 0.7239327430725098,
+      "learning_rate": 2.9202248495667788e-05,
+      "loss": 0.3791,
+      "step": 26190
+    },
+    {
+      "epoch": 9.890524726311815,
+      "grad_norm": 0.5899133086204529,
+      "learning_rate": 2.916466458362227e-05,
+      "loss": 0.3661,
+      "step": 26200
+    },
+    {
+      "epoch": 9.894299735749339,
+      "grad_norm": 0.5926426649093628,
+      "learning_rate": 2.9127094912332033e-05,
+      "loss": 0.3753,
+      "step": 26210
+    },
+    {
+      "epoch": 9.898074745186863,
+      "grad_norm": 2.781858205795288,
+      "learning_rate": 2.9089539507475606e-05,
+      "loss": 0.3749,
+      "step": 26220
+    },
+    {
+      "epoch": 9.901849754624386,
+      "grad_norm": 0.6369478106498718,
+      "learning_rate": 2.9051998394721748e-05,
+      "loss": 0.3652,
+      "step": 26230
+    },
+    {
+      "epoch": 9.90562476406191,
+      "grad_norm": 0.6876001954078674,
+      "learning_rate": 2.901447159972948e-05,
+      "loss": 0.3613,
+      "step": 26240
+    },
+    {
+      "epoch": 9.909399773499434,
+      "grad_norm": 0.6942611932754517,
+      "learning_rate": 2.8976959148148e-05,
+      "loss": 0.372,
+      "step": 26250
+    },
+    {
+      "epoch": 9.913174782936958,
+      "grad_norm": 0.7924140095710754,
+      "learning_rate": 2.8939461065616674e-05,
+      "loss": 0.3695,
+      "step": 26260
+    },
+    {
+      "epoch": 9.916949792374481,
+      "grad_norm": 0.6020333766937256,
+      "learning_rate": 2.8901977377765127e-05,
+      "loss": 0.3688,
+      "step": 26270
+    },
+    {
+      "epoch": 9.920724801812005,
+      "grad_norm": 0.6307262778282166,
+      "learning_rate": 2.8864508110213094e-05,
+      "loss": 0.3649,
+      "step": 26280
+    },
+    {
+      "epoch": 9.924499811249529,
+      "grad_norm": 0.6264989972114563,
+      "learning_rate": 2.8827053288570503e-05,
+      "loss": 0.3745,
+      "step": 26290
+    },
+    {
+      "epoch": 9.92827482068705,
+      "grad_norm": 0.7633573412895203,
+      "learning_rate": 2.8789612938437315e-05,
+      "loss": 0.3692,
+      "step": 26300
+    },
+    {
+      "epoch": 9.932049830124575,
+      "grad_norm": 0.9851648211479187,
+      "learning_rate": 2.8752187085403683e-05,
+      "loss": 0.3746,
+      "step": 26310
+    },
+    {
+      "epoch": 9.935824839562098,
+      "grad_norm": 0.6989134550094604,
+      "learning_rate": 2.8714775755049818e-05,
+      "loss": 0.3723,
+      "step": 26320
+    },
+    {
+      "epoch": 9.939599848999622,
+      "grad_norm": 0.6544426679611206,
+      "learning_rate": 2.867737897294604e-05,
+      "loss": 0.3744,
+      "step": 26330
+    },
+    {
+      "epoch": 9.943374858437146,
+      "grad_norm": 0.6202349662780762,
+      "learning_rate": 2.8639996764652653e-05,
+      "loss": 0.3753,
+      "step": 26340
+    },
+    {
+      "epoch": 9.94714986787467,
+      "grad_norm": 0.6523613929748535,
+      "learning_rate": 2.8602629155720084e-05,
+      "loss": 0.3651,
+      "step": 26350
+    },
+    {
+      "epoch": 9.950924877312193,
+      "grad_norm": 0.6727064251899719,
+      "learning_rate": 2.8565276171688703e-05,
+      "loss": 0.3732,
+      "step": 26360
+    },
+    {
+      "epoch": 9.954699886749717,
+      "grad_norm": 0.7078951001167297,
+      "learning_rate": 2.8527937838088943e-05,
+      "loss": 0.3648,
+      "step": 26370
+    },
+    {
+      "epoch": 9.95847489618724,
+      "grad_norm": 0.7242457270622253,
+      "learning_rate": 2.84906141804412e-05,
+      "loss": 0.3718,
+      "step": 26380
+    },
+    {
+      "epoch": 9.962249905624764,
+      "grad_norm": 0.6599700450897217,
+      "learning_rate": 2.8453305224255867e-05,
+      "loss": 0.364,
+      "step": 26390
+    },
+    {
+      "epoch": 9.966024915062288,
+      "grad_norm": 0.6147913336753845,
+      "learning_rate": 2.8416010995033216e-05,
+      "loss": 0.3598,
+      "step": 26400
+    },
+    {
+      "epoch": 9.969799924499812,
+      "grad_norm": 0.6449320316314697,
+      "learning_rate": 2.8378731518263524e-05,
+      "loss": 0.3718,
+      "step": 26410
+    },
+    {
+      "epoch": 9.973574933937336,
+      "grad_norm": 0.6980818510055542,
+      "learning_rate": 2.834146681942696e-05,
+      "loss": 0.3695,
+      "step": 26420
+    },
+    {
+      "epoch": 9.977349943374858,
+      "grad_norm": 0.6812340617179871,
+      "learning_rate": 2.8304216923993622e-05,
+      "loss": 0.3734,
+      "step": 26430
+    },
+    {
+      "epoch": 9.981124952812381,
+      "grad_norm": 0.644469678401947,
+      "learning_rate": 2.8266981857423413e-05,
+      "loss": 0.3704,
+      "step": 26440
+    },
+    {
+      "epoch": 9.984899962249905,
+      "grad_norm": 0.6593320369720459,
+      "learning_rate": 2.8229761645166197e-05,
+      "loss": 0.3713,
+      "step": 26450
+    },
+    {
+      "epoch": 9.988674971687429,
+      "grad_norm": 0.5889549255371094,
+      "learning_rate": 2.81925563126616e-05,
+      "loss": 0.3609,
+      "step": 26460
+    },
+    {
+      "epoch": 9.992449981124953,
+      "grad_norm": 0.6138685345649719,
+      "learning_rate": 2.8155365885339124e-05,
+      "loss": 0.3712,
+      "step": 26470
+    },
+    {
+      "epoch": 9.996224990562476,
+      "grad_norm": 0.6867060661315918,
+      "learning_rate": 2.8118190388618093e-05,
+      "loss": 0.3694,
+      "step": 26480
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.6916232705116272,
+      "learning_rate": 2.8081029847907614e-05,
+      "loss": 0.375,
+      "step": 26490
+    },
+    {
+      "epoch": 10.003775009437524,
+      "grad_norm": 0.6438423991203308,
+      "learning_rate": 2.8043884288606525e-05,
+      "loss": 0.3668,
+      "step": 26500
+    },
+    {
+      "epoch": 10.007550018875047,
+      "grad_norm": 0.5959725975990295,
+      "learning_rate": 2.8006753736103496e-05,
+      "loss": 0.3637,
+      "step": 26510
+    },
+    {
+      "epoch": 10.011325028312571,
+      "grad_norm": 0.5706154108047485,
+      "learning_rate": 2.7969638215776918e-05,
+      "loss": 0.3712,
+      "step": 26520
+    },
+    {
+      "epoch": 10.015100037750095,
+      "grad_norm": 0.6247308254241943,
+      "learning_rate": 2.793253775299487e-05,
+      "loss": 0.3686,
+      "step": 26530
+    },
+    {
+      "epoch": 10.018875047187619,
+      "grad_norm": 0.6258418560028076,
+      "learning_rate": 2.7895452373115184e-05,
+      "loss": 0.3733,
+      "step": 26540
+    },
+    {
+      "epoch": 10.022650056625142,
+      "grad_norm": 0.6243855357170105,
+      "learning_rate": 2.785838210148539e-05,
+      "loss": 0.3663,
+      "step": 26550
+    },
+    {
+      "epoch": 10.026425066062664,
+      "grad_norm": 0.6902307868003845,
+      "learning_rate": 2.782132696344263e-05,
+      "loss": 0.3783,
+      "step": 26560
+    },
+    {
+      "epoch": 10.030200075500188,
+      "grad_norm": 0.650276243686676,
+      "learning_rate": 2.7784286984313745e-05,
+      "loss": 0.368,
+      "step": 26570
+    },
+    {
+      "epoch": 10.033975084937712,
+      "grad_norm": 0.6885844469070435,
+      "learning_rate": 2.7747262189415236e-05,
+      "loss": 0.3603,
+      "step": 26580
+    },
+    {
+      "epoch": 10.037750094375236,
+      "grad_norm": 0.6648432612419128,
+      "learning_rate": 2.7710252604053205e-05,
+      "loss": 0.3738,
+      "step": 26590
+    },
+    {
+      "epoch": 10.04152510381276,
+      "grad_norm": 0.6929357051849365,
+      "learning_rate": 2.767325825352332e-05,
+      "loss": 0.3692,
+      "step": 26600
+    },
+    {
+      "epoch": 10.045300113250283,
+      "grad_norm": 0.6674989461898804,
+      "learning_rate": 2.7636279163110913e-05,
+      "loss": 0.3668,
+      "step": 26610
+    },
+    {
+      "epoch": 10.049075122687807,
+      "grad_norm": 0.6698505282402039,
+      "learning_rate": 2.7599315358090795e-05,
+      "loss": 0.3727,
+      "step": 26620
+    },
+    {
+      "epoch": 10.05285013212533,
+      "grad_norm": 0.6277247071266174,
+      "learning_rate": 2.7562366863727407e-05,
+      "loss": 0.3664,
+      "step": 26630
+    },
+    {
+      "epoch": 10.056625141562854,
+      "grad_norm": 0.6725890636444092,
+      "learning_rate": 2.7525433705274695e-05,
+      "loss": 0.3661,
+      "step": 26640
+    },
+    {
+      "epoch": 10.060400151000378,
+      "grad_norm": 0.6743810772895813,
+      "learning_rate": 2.748851590797614e-05,
+      "loss": 0.3725,
+      "step": 26650
+    },
+    {
+      "epoch": 10.064175160437902,
+      "grad_norm": 0.7060372829437256,
+      "learning_rate": 2.7451613497064675e-05,
+      "loss": 0.3743,
+      "step": 26660
+    },
+    {
+      "epoch": 10.067950169875425,
+      "grad_norm": 0.6378623843193054,
+      "learning_rate": 2.7414726497762765e-05,
+      "loss": 0.3688,
+      "step": 26670
+    },
+    {
+      "epoch": 10.071725179312947,
+      "grad_norm": 0.6196885704994202,
+      "learning_rate": 2.737785493528232e-05,
+      "loss": 0.3732,
+      "step": 26680
+    },
+    {
+      "epoch": 10.075500188750471,
+      "grad_norm": 0.770141065120697,
+      "learning_rate": 2.7340998834824745e-05,
+      "loss": 0.3697,
+      "step": 26690
+    },
+    {
+      "epoch": 10.079275198187995,
+      "grad_norm": 0.6975536942481995,
+      "learning_rate": 2.7304158221580777e-05,
+      "loss": 0.3688,
+      "step": 26700
+    },
+    {
+      "epoch": 10.083050207625519,
+      "grad_norm": 0.6799348592758179,
+      "learning_rate": 2.7267333120730675e-05,
+      "loss": 0.3716,
+      "step": 26710
+    },
+    {
+      "epoch": 10.086825217063042,
+      "grad_norm": 0.6629582643508911,
+      "learning_rate": 2.7230523557444017e-05,
+      "loss": 0.3653,
+      "step": 26720
+    },
+    {
+      "epoch": 10.090600226500566,
+      "grad_norm": 0.612046480178833,
+      "learning_rate": 2.7193729556879798e-05,
+      "loss": 0.3646,
+      "step": 26730
+    },
+    {
+      "epoch": 10.09437523593809,
+      "grad_norm": 0.8744679093360901,
+      "learning_rate": 2.715695114418637e-05,
+      "loss": 0.3704,
+      "step": 26740
+    },
+    {
+      "epoch": 10.098150245375614,
+      "grad_norm": 0.6572837829589844,
+      "learning_rate": 2.7120188344501475e-05,
+      "loss": 0.368,
+      "step": 26750
+    },
+    {
+      "epoch": 10.101925254813137,
+      "grad_norm": 0.680645763874054,
+      "learning_rate": 2.7083441182952067e-05,
+      "loss": 0.3746,
+      "step": 26760
+    },
+    {
+      "epoch": 10.105700264250661,
+      "grad_norm": 0.6419723629951477,
+      "learning_rate": 2.7046709684654527e-05,
+      "loss": 0.366,
+      "step": 26770
+    },
+    {
+      "epoch": 10.109475273688185,
+      "grad_norm": 0.6527169346809387,
+      "learning_rate": 2.700999387471448e-05,
+      "loss": 0.3715,
+      "step": 26780
+    },
+    {
+      "epoch": 10.113250283125709,
+      "grad_norm": 0.6019080281257629,
+      "learning_rate": 2.6973293778226854e-05,
+      "loss": 0.3676,
+      "step": 26790
+    },
+    {
+      "epoch": 10.117025292563232,
+      "grad_norm": 0.6170861721038818,
+      "learning_rate": 2.6936609420275804e-05,
+      "loss": 0.381,
+      "step": 26800
+    },
+    {
+      "epoch": 10.120800302000754,
+      "grad_norm": 0.8691985607147217,
+      "learning_rate": 2.689994082593472e-05,
+      "loss": 0.3656,
+      "step": 26810
+    },
+    {
+      "epoch": 10.124575311438278,
+      "grad_norm": 0.6585285067558289,
+      "learning_rate": 2.6863288020266264e-05,
+      "loss": 0.368,
+      "step": 26820
+    },
+    {
+      "epoch": 10.128350320875802,
+      "grad_norm": 0.6726837158203125,
+      "learning_rate": 2.682665102832228e-05,
+      "loss": 0.3652,
+      "step": 26830
+    },
+    {
+      "epoch": 10.132125330313325,
+      "grad_norm": 0.7205042839050293,
+      "learning_rate": 2.67900298751438e-05,
+      "loss": 0.3722,
+      "step": 26840
+    },
+    {
+      "epoch": 10.13590033975085,
+      "grad_norm": 0.6744375228881836,
+      "learning_rate": 2.6753424585761067e-05,
+      "loss": 0.3751,
+      "step": 26850
+    },
+    {
+      "epoch": 10.139675349188373,
+      "grad_norm": 3.0684523582458496,
+      "learning_rate": 2.671683518519341e-05,
+      "loss": 0.3647,
+      "step": 26860
+    },
+    {
+      "epoch": 10.143450358625897,
+      "grad_norm": 0.585762083530426,
+      "learning_rate": 2.668026169844936e-05,
+      "loss": 0.3621,
+      "step": 26870
+    },
+    {
+      "epoch": 10.14722536806342,
+      "grad_norm": 0.6915805339813232,
+      "learning_rate": 2.6643704150526538e-05,
+      "loss": 0.3648,
+      "step": 26880
+    },
+    {
+      "epoch": 10.151000377500944,
+      "grad_norm": 0.7378256916999817,
+      "learning_rate": 2.6607162566411716e-05,
+      "loss": 0.3619,
+      "step": 26890
+    },
+    {
+      "epoch": 10.154775386938468,
+      "grad_norm": 0.750008761882782,
+      "learning_rate": 2.6570636971080697e-05,
+      "loss": 0.3755,
+      "step": 26900
+    },
+    {
+      "epoch": 10.158550396375992,
+      "grad_norm": 0.6996826529502869,
+      "learning_rate": 2.6534127389498364e-05,
+      "loss": 0.3635,
+      "step": 26910
+    },
+    {
+      "epoch": 10.162325405813515,
+      "grad_norm": 0.6714365482330322,
+      "learning_rate": 2.6497633846618696e-05,
+      "loss": 0.3672,
+      "step": 26920
+    },
+    {
+      "epoch": 10.166100415251037,
+      "grad_norm": 3.8461427688598633,
+      "learning_rate": 2.6461156367384677e-05,
+      "loss": 0.3705,
+      "step": 26930
+    },
+    {
+      "epoch": 10.169875424688561,
+      "grad_norm": 0.6795220375061035,
+      "learning_rate": 2.6424694976728316e-05,
+      "loss": 0.3727,
+      "step": 26940
+    },
+    {
+      "epoch": 10.173650434126085,
+      "grad_norm": 0.7694344520568848,
+      "learning_rate": 2.6388249699570667e-05,
+      "loss": 0.3705,
+      "step": 26950
+    },
+    {
+      "epoch": 10.177425443563608,
+      "grad_norm": 0.6592543721199036,
+      "learning_rate": 2.6351820560821672e-05,
+      "loss": 0.3717,
+      "step": 26960
+    },
+    {
+      "epoch": 10.181200453001132,
+      "grad_norm": 0.7184135317802429,
+      "learning_rate": 2.631540758538034e-05,
+      "loss": 0.3713,
+      "step": 26970
+    },
+    {
+      "epoch": 10.184975462438656,
+      "grad_norm": 0.6492671966552734,
+      "learning_rate": 2.6279010798134597e-05,
+      "loss": 0.3637,
+      "step": 26980
+    },
+    {
+      "epoch": 10.18875047187618,
+      "grad_norm": 0.6501529812812805,
+      "learning_rate": 2.6242630223961305e-05,
+      "loss": 0.3774,
+      "step": 26990
+    },
+    {
+      "epoch": 10.192525481313703,
+      "grad_norm": 0.8947603106498718,
+      "learning_rate": 2.6206265887726244e-05,
+      "loss": 0.369,
+      "step": 27000
+    },
+    {
+      "epoch": 10.196300490751227,
+      "grad_norm": 0.7419344186782837,
+      "learning_rate": 2.6169917814284066e-05,
+      "loss": 0.3672,
+      "step": 27010
+    },
+    {
+      "epoch": 10.20007550018875,
+      "grad_norm": 0.703224241733551,
+      "learning_rate": 2.6133586028478364e-05,
+      "loss": 0.3683,
+      "step": 27020
+    },
+    {
+      "epoch": 10.203850509626275,
+      "grad_norm": 0.6347489356994629,
+      "learning_rate": 2.609727055514155e-05,
+      "loss": 0.367,
+      "step": 27030
+    },
+    {
+      "epoch": 10.207625519063798,
+      "grad_norm": 0.6552822589874268,
+      "learning_rate": 2.606097141909494e-05,
+      "loss": 0.3647,
+      "step": 27040
+    },
+    {
+      "epoch": 10.211400528501322,
+      "grad_norm": 0.6825968623161316,
+      "learning_rate": 2.6024688645148644e-05,
+      "loss": 0.3713,
+      "step": 27050
+    },
+    {
+      "epoch": 10.215175537938844,
+      "grad_norm": 0.7292717695236206,
+      "learning_rate": 2.5988422258101564e-05,
+      "loss": 0.3673,
+      "step": 27060
+    },
+    {
+      "epoch": 10.218950547376368,
+      "grad_norm": 1.0935133695602417,
+      "learning_rate": 2.5952172282741453e-05,
+      "loss": 0.3688,
+      "step": 27070
+    },
+    {
+      "epoch": 10.222725556813892,
+      "grad_norm": 0.6263819932937622,
+      "learning_rate": 2.5915938743844853e-05,
+      "loss": 0.3764,
+      "step": 27080
+    },
+    {
+      "epoch": 10.226500566251415,
+      "grad_norm": 0.6874544024467468,
+      "learning_rate": 2.5879721666177003e-05,
+      "loss": 0.3675,
+      "step": 27090
+    },
+    {
+      "epoch": 10.230275575688939,
+      "grad_norm": 0.640953004360199,
+      "learning_rate": 2.5843521074491972e-05,
+      "loss": 0.3691,
+      "step": 27100
+    },
+    {
+      "epoch": 10.234050585126463,
+      "grad_norm": 0.6374151110649109,
+      "learning_rate": 2.5807336993532487e-05,
+      "loss": 0.3623,
+      "step": 27110
+    },
+    {
+      "epoch": 10.237825594563986,
+      "grad_norm": 0.6693562269210815,
+      "learning_rate": 2.577116944803004e-05,
+      "loss": 0.3702,
+      "step": 27120
+    },
+    {
+      "epoch": 10.24160060400151,
+      "grad_norm": 0.6881791949272156,
+      "learning_rate": 2.5735018462704818e-05,
+      "loss": 0.3695,
+      "step": 27130
+    },
+    {
+      "epoch": 10.245375613439034,
+      "grad_norm": 0.6473117470741272,
+      "learning_rate": 2.5698884062265665e-05,
+      "loss": 0.3652,
+      "step": 27140
+    },
+    {
+      "epoch": 10.249150622876558,
+      "grad_norm": 0.7239118218421936,
+      "learning_rate": 2.5662766271410134e-05,
+      "loss": 0.3715,
+      "step": 27150
+    },
+    {
+      "epoch": 10.252925632314081,
+      "grad_norm": 0.6931194067001343,
+      "learning_rate": 2.5626665114824343e-05,
+      "loss": 0.3679,
+      "step": 27160
+    },
+    {
+      "epoch": 10.256700641751605,
+      "grad_norm": 0.7126404643058777,
+      "learning_rate": 2.5590580617183148e-05,
+      "loss": 0.3729,
+      "step": 27170
+    },
+    {
+      "epoch": 10.260475651189129,
+      "grad_norm": 0.612730085849762,
+      "learning_rate": 2.5554512803149912e-05,
+      "loss": 0.3626,
+      "step": 27180
+    },
+    {
+      "epoch": 10.26425066062665,
+      "grad_norm": 0.631647527217865,
+      "learning_rate": 2.5518461697376662e-05,
+      "loss": 0.3656,
+      "step": 27190
+    },
+    {
+      "epoch": 10.268025670064175,
+      "grad_norm": 0.692253053188324,
+      "learning_rate": 2.548242732450402e-05,
+      "loss": 0.3658,
+      "step": 27200
+    },
+    {
+      "epoch": 10.271800679501698,
+      "grad_norm": 0.8286940455436707,
+      "learning_rate": 2.5446409709161095e-05,
+      "loss": 0.3708,
+      "step": 27210
+    },
+    {
+      "epoch": 10.275575688939222,
+      "grad_norm": 1.055858850479126,
+      "learning_rate": 2.541040887596561e-05,
+      "loss": 0.362,
+      "step": 27220
+    },
+    {
+      "epoch": 10.279350698376746,
+      "grad_norm": 0.6639923453330994,
+      "learning_rate": 2.537442484952378e-05,
+      "loss": 0.3684,
+      "step": 27230
+    },
+    {
+      "epoch": 10.28312570781427,
+      "grad_norm": 0.6410587430000305,
+      "learning_rate": 2.533845765443037e-05,
+      "loss": 0.3736,
+      "step": 27240
+    },
+    {
+      "epoch": 10.286900717251793,
+      "grad_norm": 0.6842541694641113,
+      "learning_rate": 2.530250731526863e-05,
+      "loss": 0.3624,
+      "step": 27250
+    },
+    {
+      "epoch": 10.290675726689317,
+      "grad_norm": 0.6937667727470398,
+      "learning_rate": 2.5266573856610253e-05,
+      "loss": 0.3588,
+      "step": 27260
+    },
+    {
+      "epoch": 10.29445073612684,
+      "grad_norm": 0.6150068044662476,
+      "learning_rate": 2.5230657303015403e-05,
+      "loss": 0.3604,
+      "step": 27270
+    },
+    {
+      "epoch": 10.298225745564364,
+      "grad_norm": 0.6743102669715881,
+      "learning_rate": 2.5194757679032728e-05,
+      "loss": 0.3679,
+      "step": 27280
+    },
+    {
+      "epoch": 10.302000755001888,
+      "grad_norm": 0.6566803455352783,
+      "learning_rate": 2.5158875009199278e-05,
+      "loss": 0.3657,
+      "step": 27290
+    },
+    {
+      "epoch": 10.305775764439412,
+      "grad_norm": 0.6627777218818665,
+      "learning_rate": 2.5123009318040537e-05,
+      "loss": 0.3677,
+      "step": 27300
+    },
+    {
+      "epoch": 10.309550773876934,
+      "grad_norm": 0.7576314806938171,
+      "learning_rate": 2.508716063007034e-05,
+      "loss": 0.3617,
+      "step": 27310
+    },
+    {
+      "epoch": 10.313325783314458,
+      "grad_norm": 1.1635149717330933,
+      "learning_rate": 2.5051328969790934e-05,
+      "loss": 0.367,
+      "step": 27320
+    },
+    {
+      "epoch": 10.317100792751981,
+      "grad_norm": 0.6241893768310547,
+      "learning_rate": 2.501551436169292e-05,
+      "loss": 0.3721,
+      "step": 27330
+    },
+    {
+      "epoch": 10.320875802189505,
+      "grad_norm": 0.5901798605918884,
+      "learning_rate": 2.4979716830255255e-05,
+      "loss": 0.3629,
+      "step": 27340
+    },
+    {
+      "epoch": 10.324650811627029,
+      "grad_norm": 0.6239721179008484,
+      "learning_rate": 2.4943936399945233e-05,
+      "loss": 0.3647,
+      "step": 27350
+    },
+    {
+      "epoch": 10.328425821064553,
+      "grad_norm": 0.6076030135154724,
+      "learning_rate": 2.4908173095218412e-05,
+      "loss": 0.3602,
+      "step": 27360
+    },
+    {
+      "epoch": 10.332200830502076,
+      "grad_norm": 0.6556723117828369,
+      "learning_rate": 2.4872426940518663e-05,
+      "loss": 0.369,
+      "step": 27370
+    },
+    {
+      "epoch": 10.3359758399396,
+      "grad_norm": 0.6872203350067139,
+      "learning_rate": 2.4836697960278156e-05,
+      "loss": 0.3669,
+      "step": 27380
+    },
+    {
+      "epoch": 10.339750849377124,
+      "grad_norm": 0.6859211921691895,
+      "learning_rate": 2.480098617891732e-05,
+      "loss": 0.3678,
+      "step": 27390
+    },
+    {
+      "epoch": 10.343525858814647,
+      "grad_norm": 0.6686061024665833,
+      "learning_rate": 2.4765291620844837e-05,
+      "loss": 0.3558,
+      "step": 27400
+    },
+    {
+      "epoch": 10.347300868252171,
+      "grad_norm": 0.6350464820861816,
+      "learning_rate": 2.472961431045756e-05,
+      "loss": 0.3808,
+      "step": 27410
+    },
+    {
+      "epoch": 10.351075877689695,
+      "grad_norm": 0.6436896324157715,
+      "learning_rate": 2.4693954272140622e-05,
+      "loss": 0.3644,
+      "step": 27420
+    },
+    {
+      "epoch": 10.354850887127217,
+      "grad_norm": 0.7389644980430603,
+      "learning_rate": 2.4658311530267315e-05,
+      "loss": 0.368,
+      "step": 27430
+    },
+    {
+      "epoch": 10.35862589656474,
+      "grad_norm": 0.7080246210098267,
+      "learning_rate": 2.4622686109199124e-05,
+      "loss": 0.368,
+      "step": 27440
+    },
+    {
+      "epoch": 10.362400906002264,
+      "grad_norm": 0.6405688524246216,
+      "learning_rate": 2.4587078033285695e-05,
+      "loss": 0.3731,
+      "step": 27450
+    },
+    {
+      "epoch": 10.366175915439788,
+      "grad_norm": 0.6456767320632935,
+      "learning_rate": 2.45514873268648e-05,
+      "loss": 0.3615,
+      "step": 27460
+    },
+    {
+      "epoch": 10.369950924877312,
+      "grad_norm": 0.6688347458839417,
+      "learning_rate": 2.4515914014262336e-05,
+      "loss": 0.3618,
+      "step": 27470
+    },
+    {
+      "epoch": 10.373725934314836,
+      "grad_norm": 0.7001455426216125,
+      "learning_rate": 2.4480358119792345e-05,
+      "loss": 0.3669,
+      "step": 27480
+    },
+    {
+      "epoch": 10.37750094375236,
+      "grad_norm": 0.6570714712142944,
+      "learning_rate": 2.4444819667756942e-05,
+      "loss": 0.3574,
+      "step": 27490
+    },
+    {
+      "epoch": 10.381275953189883,
+      "grad_norm": 0.6306593418121338,
+      "learning_rate": 2.4409298682446346e-05,
+      "loss": 0.3646,
+      "step": 27500
+    },
+    {
+      "epoch": 10.385050962627407,
+      "grad_norm": 0.6927255988121033,
+      "learning_rate": 2.437379518813877e-05,
+      "loss": 0.3671,
+      "step": 27510
+    },
+    {
+      "epoch": 10.38882597206493,
+      "grad_norm": 0.7374937534332275,
+      "learning_rate": 2.4338309209100547e-05,
+      "loss": 0.3626,
+      "step": 27520
+    },
+    {
+      "epoch": 10.392600981502454,
+      "grad_norm": 1.0554702281951904,
+      "learning_rate": 2.4302840769586004e-05,
+      "loss": 0.3635,
+      "step": 27530
+    },
+    {
+      "epoch": 10.396375990939978,
+      "grad_norm": 0.6217535734176636,
+      "learning_rate": 2.42673898938375e-05,
+      "loss": 0.3728,
+      "step": 27540
+    },
+    {
+      "epoch": 10.400151000377502,
+      "grad_norm": 0.654321551322937,
+      "learning_rate": 2.4231956606085343e-05,
+      "loss": 0.3616,
+      "step": 27550
+    },
+    {
+      "epoch": 10.403926009815024,
+      "grad_norm": 0.7654052972793579,
+      "learning_rate": 2.419654093054789e-05,
+      "loss": 0.3723,
+      "step": 27560
+    },
+    {
+      "epoch": 10.407701019252547,
+      "grad_norm": 0.6208662390708923,
+      "learning_rate": 2.4161142891431375e-05,
+      "loss": 0.3639,
+      "step": 27570
+    },
+    {
+      "epoch": 10.411476028690071,
+      "grad_norm": 0.7695934772491455,
+      "learning_rate": 2.412576251293005e-05,
+      "loss": 0.3657,
+      "step": 27580
+    },
+    {
+      "epoch": 10.415251038127595,
+      "grad_norm": 0.6474165916442871,
+      "learning_rate": 2.4090399819226068e-05,
+      "loss": 0.3664,
+      "step": 27590
+    },
+    {
+      "epoch": 10.419026047565119,
+      "grad_norm": 0.716998815536499,
+      "learning_rate": 2.4055054834489514e-05,
+      "loss": 0.3695,
+      "step": 27600
+    },
+    {
+      "epoch": 10.422801057002642,
+      "grad_norm": 0.6260504722595215,
+      "learning_rate": 2.401972758287832e-05,
+      "loss": 0.3659,
+      "step": 27610
+    },
+    {
+      "epoch": 10.426576066440166,
+      "grad_norm": 0.668073296546936,
+      "learning_rate": 2.398441808853834e-05,
+      "loss": 0.3555,
+      "step": 27620
+    },
+    {
+      "epoch": 10.43035107587769,
+      "grad_norm": 0.6870536208152771,
+      "learning_rate": 2.3949126375603288e-05,
+      "loss": 0.3675,
+      "step": 27630
+    },
+    {
+      "epoch": 10.434126085315214,
+      "grad_norm": 0.6690881848335266,
+      "learning_rate": 2.3913852468194724e-05,
+      "loss": 0.3678,
+      "step": 27640
+    },
+    {
+      "epoch": 10.437901094752737,
+      "grad_norm": 0.6636070013046265,
+      "learning_rate": 2.387859639042201e-05,
+      "loss": 0.3693,
+      "step": 27650
+    },
+    {
+      "epoch": 10.441676104190261,
+      "grad_norm": 0.7105987668037415,
+      "learning_rate": 2.3843358166382368e-05,
+      "loss": 0.3618,
+      "step": 27660
+    },
+    {
+      "epoch": 10.445451113627785,
+      "grad_norm": 0.6671977639198303,
+      "learning_rate": 2.3808137820160757e-05,
+      "loss": 0.3639,
+      "step": 27670
+    },
+    {
+      "epoch": 10.449226123065309,
+      "grad_norm": 0.6227991580963135,
+      "learning_rate": 2.3772935375829975e-05,
+      "loss": 0.3587,
+      "step": 27680
+    },
+    {
+      "epoch": 10.45300113250283,
+      "grad_norm": 0.5635626316070557,
+      "learning_rate": 2.3737750857450553e-05,
+      "loss": 0.3723,
+      "step": 27690
+    },
+    {
+      "epoch": 10.456776141940354,
+      "grad_norm": 0.6364959478378296,
+      "learning_rate": 2.3702584289070805e-05,
+      "loss": 0.3624,
+      "step": 27700
+    },
+    {
+      "epoch": 10.460551151377878,
+      "grad_norm": 0.6209636330604553,
+      "learning_rate": 2.36674356947267e-05,
+      "loss": 0.3678,
+      "step": 27710
+    },
+    {
+      "epoch": 10.464326160815402,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 2.3632305098442004e-05,
+      "loss": 0.3583,
+      "step": 27720
+    },
+    {
+      "epoch": 10.468101170252925,
+      "grad_norm": 0.7123914361000061,
+      "learning_rate": 2.3597192524228156e-05,
+      "loss": 0.3636,
+      "step": 27730
+    },
+    {
+      "epoch": 10.47187617969045,
+      "grad_norm": 0.6567232012748718,
+      "learning_rate": 2.356209799608424e-05,
+      "loss": 0.3654,
+      "step": 27740
+    },
+    {
+      "epoch": 10.475651189127973,
+      "grad_norm": 0.6329348087310791,
+      "learning_rate": 2.352702153799704e-05,
+      "loss": 0.3665,
+      "step": 27750
+    },
+    {
+      "epoch": 10.479426198565497,
+      "grad_norm": 0.6293632984161377,
+      "learning_rate": 2.3491963173941018e-05,
+      "loss": 0.3626,
+      "step": 27760
+    },
+    {
+      "epoch": 10.48320120800302,
+      "grad_norm": 0.652357816696167,
+      "learning_rate": 2.3456922927878196e-05,
+      "loss": 0.3759,
+      "step": 27770
+    },
+    {
+      "epoch": 10.486976217440544,
+      "grad_norm": 0.6584979295730591,
+      "learning_rate": 2.3421900823758257e-05,
+      "loss": 0.3652,
+      "step": 27780
+    },
+    {
+      "epoch": 10.490751226878068,
+      "grad_norm": 0.6862519383430481,
+      "learning_rate": 2.3386896885518496e-05,
+      "loss": 0.3648,
+      "step": 27790
+    },
+    {
+      "epoch": 10.494526236315592,
+      "grad_norm": 0.6221929788589478,
+      "learning_rate": 2.335191113708378e-05,
+      "loss": 0.3721,
+      "step": 27800
+    },
+    {
+      "epoch": 10.498301245753115,
+      "grad_norm": 0.6476244330406189,
+      "learning_rate": 2.331694360236651e-05,
+      "loss": 0.3746,
+      "step": 27810
+    },
+    {
+      "epoch": 10.502076255190637,
+      "grad_norm": 0.626964807510376,
+      "learning_rate": 2.3281994305266702e-05,
+      "loss": 0.3616,
+      "step": 27820
+    },
+    {
+      "epoch": 10.505851264628161,
+      "grad_norm": 0.7498915791511536,
+      "learning_rate": 2.3247063269671826e-05,
+      "loss": 0.3689,
+      "step": 27830
+    },
+    {
+      "epoch": 10.509626274065685,
+      "grad_norm": 0.6770737767219543,
+      "learning_rate": 2.321215051945695e-05,
+      "loss": 0.3672,
+      "step": 27840
+    },
+    {
+      "epoch": 10.513401283503208,
+      "grad_norm": 0.7251414656639099,
+      "learning_rate": 2.3177256078484588e-05,
+      "loss": 0.3664,
+      "step": 27850
+    },
+    {
+      "epoch": 10.517176292940732,
+      "grad_norm": 0.7046555876731873,
+      "learning_rate": 2.3142379970604798e-05,
+      "loss": 0.361,
+      "step": 27860
+    },
+    {
+      "epoch": 10.520951302378256,
+      "grad_norm": 0.6266787648200989,
+      "learning_rate": 2.3107522219655025e-05,
+      "loss": 0.3649,
+      "step": 27870
+    },
+    {
+      "epoch": 10.52472631181578,
+      "grad_norm": 0.6274697780609131,
+      "learning_rate": 2.3072682849460236e-05,
+      "loss": 0.3601,
+      "step": 27880
+    },
+    {
+      "epoch": 10.528501321253303,
+      "grad_norm": 0.6082983613014221,
+      "learning_rate": 2.303786188383281e-05,
+      "loss": 0.3688,
+      "step": 27890
+    },
+    {
+      "epoch": 10.532276330690827,
+      "grad_norm": 0.6668601632118225,
+      "learning_rate": 2.300305934657257e-05,
+      "loss": 0.361,
+      "step": 27900
+    },
+    {
+      "epoch": 10.53605134012835,
+      "grad_norm": 0.6825070977210999,
+      "learning_rate": 2.2968275261466677e-05,
+      "loss": 0.3637,
+      "step": 27910
+    },
+    {
+      "epoch": 10.539826349565875,
+      "grad_norm": 0.6331124901771545,
+      "learning_rate": 2.293350965228977e-05,
+      "loss": 0.3639,
+      "step": 27920
+    },
+    {
+      "epoch": 10.543601359003397,
+      "grad_norm": 0.6340346932411194,
+      "learning_rate": 2.2898762542803776e-05,
+      "loss": 0.3625,
+      "step": 27930
+    },
+    {
+      "epoch": 10.54737636844092,
+      "grad_norm": 0.7296791672706604,
+      "learning_rate": 2.286403395675803e-05,
+      "loss": 0.3667,
+      "step": 27940
+    },
+    {
+      "epoch": 10.551151377878444,
+      "grad_norm": 0.6600122451782227,
+      "learning_rate": 2.28293239178892e-05,
+      "loss": 0.3671,
+      "step": 27950
+    },
+    {
+      "epoch": 10.554926387315968,
+      "grad_norm": 0.7184299230575562,
+      "learning_rate": 2.2794632449921287e-05,
+      "loss": 0.3702,
+      "step": 27960
+    },
+    {
+      "epoch": 10.558701396753492,
+      "grad_norm": 0.6096726655960083,
+      "learning_rate": 2.275995957656555e-05,
+      "loss": 0.3645,
+      "step": 27970
+    },
+    {
+      "epoch": 10.562476406191015,
+      "grad_norm": 0.6670506000518799,
+      "learning_rate": 2.272530532152058e-05,
+      "loss": 0.3646,
+      "step": 27980
+    },
+    {
+      "epoch": 10.566251415628539,
+      "grad_norm": 0.6431169509887695,
+      "learning_rate": 2.2690669708472233e-05,
+      "loss": 0.361,
+      "step": 27990
+    },
+    {
+      "epoch": 10.570026425066063,
+      "grad_norm": 0.6406874060630798,
+      "learning_rate": 2.2656052761093655e-05,
+      "loss": 0.3677,
+      "step": 28000
+    },
+    {
+      "epoch": 10.573801434503586,
+      "grad_norm": 0.6442010402679443,
+      "learning_rate": 2.262145450304517e-05,
+      "loss": 0.3661,
+      "step": 28010
+    },
+    {
+      "epoch": 10.57757644394111,
+      "grad_norm": 0.6408098340034485,
+      "learning_rate": 2.2586874957974352e-05,
+      "loss": 0.3727,
+      "step": 28020
+    },
+    {
+      "epoch": 10.581351453378634,
+      "grad_norm": 0.648128092288971,
+      "learning_rate": 2.2552314149516012e-05,
+      "loss": 0.3656,
+      "step": 28030
+    },
+    {
+      "epoch": 10.585126462816158,
+      "grad_norm": 0.7829309105873108,
+      "learning_rate": 2.2517772101292133e-05,
+      "loss": 0.3665,
+      "step": 28040
+    },
+    {
+      "epoch": 10.588901472253681,
+      "grad_norm": 0.768104612827301,
+      "learning_rate": 2.248324883691188e-05,
+      "loss": 0.3687,
+      "step": 28050
+    },
+    {
+      "epoch": 10.592676481691203,
+      "grad_norm": 0.631535530090332,
+      "learning_rate": 2.24487443799716e-05,
+      "loss": 0.3683,
+      "step": 28060
+    },
+    {
+      "epoch": 10.596451491128727,
+      "grad_norm": 0.7682189345359802,
+      "learning_rate": 2.241425875405472e-05,
+      "loss": 0.3705,
+      "step": 28070
+    },
+    {
+      "epoch": 10.60022650056625,
+      "grad_norm": 0.6943038702011108,
+      "learning_rate": 2.2379791982731868e-05,
+      "loss": 0.4653,
+      "step": 28080
+    },
+    {
+      "epoch": 10.604001510003775,
+      "grad_norm": 0.634013295173645,
+      "learning_rate": 2.2345344089560756e-05,
+      "loss": 0.3621,
+      "step": 28090
+    },
+    {
+      "epoch": 10.607776519441298,
+      "grad_norm": 0.7249836325645447,
+      "learning_rate": 2.2310915098086206e-05,
+      "loss": 0.3655,
+      "step": 28100
+    },
+    {
+      "epoch": 10.611551528878822,
+      "grad_norm": 0.8200324773788452,
+      "learning_rate": 2.227650503184009e-05,
+      "loss": 0.373,
+      "step": 28110
+    },
+    {
+      "epoch": 10.615326538316346,
+      "grad_norm": 0.8548815250396729,
+      "learning_rate": 2.2242113914341357e-05,
+      "loss": 0.3581,
+      "step": 28120
+    },
+    {
+      "epoch": 10.61910154775387,
+      "grad_norm": 0.6571340560913086,
+      "learning_rate": 2.220774176909602e-05,
+      "loss": 0.3629,
+      "step": 28130
+    },
+    {
+      "epoch": 10.622876557191393,
+      "grad_norm": 0.6676787734031677,
+      "learning_rate": 2.2173388619597114e-05,
+      "loss": 0.3682,
+      "step": 28140
+    },
+    {
+      "epoch": 10.626651566628917,
+      "grad_norm": 0.7401626706123352,
+      "learning_rate": 2.21390544893247e-05,
+      "loss": 0.3667,
+      "step": 28150
+    },
+    {
+      "epoch": 10.63042657606644,
+      "grad_norm": 0.6297852396965027,
+      "learning_rate": 2.210473940174585e-05,
+      "loss": 0.3739,
+      "step": 28160
+    },
+    {
+      "epoch": 10.634201585503964,
+      "grad_norm": 0.6574814319610596,
+      "learning_rate": 2.207044338031456e-05,
+      "loss": 0.3613,
+      "step": 28170
+    },
+    {
+      "epoch": 10.637976594941488,
+      "grad_norm": 0.5910388827323914,
+      "learning_rate": 2.203616644847186e-05,
+      "loss": 0.3778,
+      "step": 28180
+    },
+    {
+      "epoch": 10.64175160437901,
+      "grad_norm": 0.6458864808082581,
+      "learning_rate": 2.200190862964571e-05,
+      "loss": 0.3678,
+      "step": 28190
+    },
+    {
+      "epoch": 10.645526613816534,
+      "grad_norm": 0.7278522849082947,
+      "learning_rate": 2.1967669947251024e-05,
+      "loss": 0.364,
+      "step": 28200
+    },
+    {
+      "epoch": 10.649301623254058,
+      "grad_norm": 0.6762776970863342,
+      "learning_rate": 2.1933450424689583e-05,
+      "loss": 0.3594,
+      "step": 28210
+    },
+    {
+      "epoch": 10.653076632691581,
+      "grad_norm": 1.226162075996399,
+      "learning_rate": 2.1899250085350142e-05,
+      "loss": 0.3733,
+      "step": 28220
+    },
+    {
+      "epoch": 10.656851642129105,
+      "grad_norm": 0.5935925841331482,
+      "learning_rate": 2.1865068952608277e-05,
+      "loss": 0.3628,
+      "step": 28230
+    },
+    {
+      "epoch": 10.660626651566629,
+      "grad_norm": 0.7492917776107788,
+      "learning_rate": 2.1830907049826487e-05,
+      "loss": 0.3611,
+      "step": 28240
+    },
+    {
+      "epoch": 10.664401661004153,
+      "grad_norm": 0.7051345705986023,
+      "learning_rate": 2.179676440035411e-05,
+      "loss": 0.3602,
+      "step": 28250
+    },
+    {
+      "epoch": 10.668176670441676,
+      "grad_norm": 0.6424268484115601,
+      "learning_rate": 2.1762641027527337e-05,
+      "loss": 0.366,
+      "step": 28260
+    },
+    {
+      "epoch": 10.6719516798792,
+      "grad_norm": 0.8368542790412903,
+      "learning_rate": 2.1728536954669143e-05,
+      "loss": 0.3646,
+      "step": 28270
+    },
+    {
+      "epoch": 10.675726689316724,
+      "grad_norm": 0.7216016054153442,
+      "learning_rate": 2.169445220508936e-05,
+      "loss": 0.36,
+      "step": 28280
+    },
+    {
+      "epoch": 10.679501698754247,
+      "grad_norm": 0.6066749691963196,
+      "learning_rate": 2.166038680208461e-05,
+      "loss": 0.3587,
+      "step": 28290
+    },
+    {
+      "epoch": 10.683276708191771,
+      "grad_norm": 0.5810279250144958,
+      "learning_rate": 2.162634076893823e-05,
+      "loss": 0.3697,
+      "step": 28300
+    },
+    {
+      "epoch": 10.687051717629295,
+      "grad_norm": 0.6657140254974365,
+      "learning_rate": 2.1592314128920388e-05,
+      "loss": 0.3678,
+      "step": 28310
+    },
+    {
+      "epoch": 10.690826727066817,
+      "grad_norm": 0.636150598526001,
+      "learning_rate": 2.155830690528799e-05,
+      "loss": 0.3654,
+      "step": 28320
+    },
+    {
+      "epoch": 10.69460173650434,
+      "grad_norm": 0.6530934572219849,
+      "learning_rate": 2.1524319121284613e-05,
+      "loss": 0.3573,
+      "step": 28330
+    },
+    {
+      "epoch": 10.698376745941864,
+      "grad_norm": 0.6528118848800659,
+      "learning_rate": 2.1490350800140607e-05,
+      "loss": 0.3712,
+      "step": 28340
+    },
+    {
+      "epoch": 10.702151755379388,
+      "grad_norm": 0.6614232659339905,
+      "learning_rate": 2.1456401965073002e-05,
+      "loss": 0.3647,
+      "step": 28350
+    },
+    {
+      "epoch": 10.705926764816912,
+      "grad_norm": 0.6931934356689453,
+      "learning_rate": 2.1422472639285524e-05,
+      "loss": 0.3641,
+      "step": 28360
+    },
+    {
+      "epoch": 10.709701774254436,
+      "grad_norm": 0.6107961535453796,
+      "learning_rate": 2.13885628459685e-05,
+      "loss": 0.3576,
+      "step": 28370
+    },
+    {
+      "epoch": 10.71347678369196,
+      "grad_norm": 0.6835388541221619,
+      "learning_rate": 2.135467260829901e-05,
+      "loss": 0.364,
+      "step": 28380
+    },
+    {
+      "epoch": 10.717251793129483,
+      "grad_norm": 0.655625581741333,
+      "learning_rate": 2.1320801949440654e-05,
+      "loss": 0.3629,
+      "step": 28390
+    },
+    {
+      "epoch": 10.721026802567007,
+      "grad_norm": 0.7267212867736816,
+      "learning_rate": 2.1286950892543744e-05,
+      "loss": 0.3641,
+      "step": 28400
+    },
+    {
+      "epoch": 10.72480181200453,
+      "grad_norm": 0.6693907380104065,
+      "learning_rate": 2.125311946074515e-05,
+      "loss": 0.3657,
+      "step": 28410
+    },
+    {
+      "epoch": 10.728576821442054,
+      "grad_norm": 0.6558740735054016,
+      "learning_rate": 2.1219307677168355e-05,
+      "loss": 0.3646,
+      "step": 28420
+    },
+    {
+      "epoch": 10.732351830879578,
+      "grad_norm": 0.7080862522125244,
+      "learning_rate": 2.118551556492336e-05,
+      "loss": 0.361,
+      "step": 28430
+    },
+    {
+      "epoch": 10.7361268403171,
+      "grad_norm": 0.6530542969703674,
+      "learning_rate": 2.1151743147106774e-05,
+      "loss": 0.3648,
+      "step": 28440
+    },
+    {
+      "epoch": 10.739901849754624,
+      "grad_norm": 0.705801784992218,
+      "learning_rate": 2.111799044680172e-05,
+      "loss": 0.3622,
+      "step": 28450
+    },
+    {
+      "epoch": 10.743676859192147,
+      "grad_norm": 0.6428266763687134,
+      "learning_rate": 2.1084257487077873e-05,
+      "loss": 0.3669,
+      "step": 28460
+    },
+    {
+      "epoch": 10.747451868629671,
+      "grad_norm": 0.7444311380386353,
+      "learning_rate": 2.1050544290991357e-05,
+      "loss": 0.3596,
+      "step": 28470
+    },
+    {
+      "epoch": 10.751226878067195,
+      "grad_norm": 0.6238864660263062,
+      "learning_rate": 2.101685088158486e-05,
+      "loss": 0.3694,
+      "step": 28480
+    },
+    {
+      "epoch": 10.755001887504719,
+      "grad_norm": 0.681848406791687,
+      "learning_rate": 2.0983177281887472e-05,
+      "loss": 0.369,
+      "step": 28490
+    },
+    {
+      "epoch": 10.758776896942242,
+      "grad_norm": 0.6995017528533936,
+      "learning_rate": 2.0949523514914798e-05,
+      "loss": 0.3644,
+      "step": 28500
+    },
+    {
+      "epoch": 10.762551906379766,
+      "grad_norm": 0.6024167537689209,
+      "learning_rate": 2.0915889603668876e-05,
+      "loss": 0.357,
+      "step": 28510
+    },
+    {
+      "epoch": 10.76632691581729,
+      "grad_norm": 0.6610404849052429,
+      "learning_rate": 2.0882275571138175e-05,
+      "loss": 0.3634,
+      "step": 28520
+    },
+    {
+      "epoch": 10.770101925254814,
+      "grad_norm": 0.629758894443512,
+      "learning_rate": 2.0848681440297545e-05,
+      "loss": 0.3644,
+      "step": 28530
+    },
+    {
+      "epoch": 10.773876934692337,
+      "grad_norm": 0.6765365600585938,
+      "learning_rate": 2.081510723410827e-05,
+      "loss": 0.3606,
+      "step": 28540
+    },
+    {
+      "epoch": 10.777651944129861,
+      "grad_norm": 0.7556784152984619,
+      "learning_rate": 2.0781552975518003e-05,
+      "loss": 0.3672,
+      "step": 28550
+    },
+    {
+      "epoch": 10.781426953567383,
+      "grad_norm": 0.7298506498336792,
+      "learning_rate": 2.074801868746078e-05,
+      "loss": 0.368,
+      "step": 28560
+    },
+    {
+      "epoch": 10.785201963004907,
+      "grad_norm": 0.7526068091392517,
+      "learning_rate": 2.0714504392856955e-05,
+      "loss": 0.368,
+      "step": 28570
+    },
+    {
+      "epoch": 10.78897697244243,
+      "grad_norm": 0.6001387238502502,
+      "learning_rate": 2.0681010114613215e-05,
+      "loss": 0.3641,
+      "step": 28580
+    },
+    {
+      "epoch": 10.792751981879954,
+      "grad_norm": 0.662828803062439,
+      "learning_rate": 2.0647535875622597e-05,
+      "loss": 0.362,
+      "step": 28590
+    },
+    {
+      "epoch": 10.796526991317478,
+      "grad_norm": 2.932863712310791,
+      "learning_rate": 2.0614081698764432e-05,
+      "loss": 0.3642,
+      "step": 28600
+    },
+    {
+      "epoch": 10.800302000755002,
+      "grad_norm": 0.7236543297767639,
+      "learning_rate": 2.0580647606904334e-05,
+      "loss": 0.3654,
+      "step": 28610
+    },
+    {
+      "epoch": 10.804077010192525,
+      "grad_norm": 0.6587764620780945,
+      "learning_rate": 2.0547233622894208e-05,
+      "loss": 0.3629,
+      "step": 28620
+    },
+    {
+      "epoch": 10.80785201963005,
+      "grad_norm": 0.6333833932876587,
+      "learning_rate": 2.0513839769572157e-05,
+      "loss": 0.3685,
+      "step": 28630
+    },
+    {
+      "epoch": 10.811627029067573,
+      "grad_norm": 0.596711277961731,
+      "learning_rate": 2.0480466069762584e-05,
+      "loss": 0.3737,
+      "step": 28640
+    },
+    {
+      "epoch": 10.815402038505097,
+      "grad_norm": 0.7744527459144592,
+      "learning_rate": 2.0447112546276104e-05,
+      "loss": 0.3628,
+      "step": 28650
+    },
+    {
+      "epoch": 10.81917704794262,
+      "grad_norm": 0.6719459295272827,
+      "learning_rate": 2.0413779221909547e-05,
+      "loss": 0.3714,
+      "step": 28660
+    },
+    {
+      "epoch": 10.822952057380144,
+      "grad_norm": 0.6692177653312683,
+      "learning_rate": 2.0380466119445912e-05,
+      "loss": 0.3652,
+      "step": 28670
+    },
+    {
+      "epoch": 10.826727066817668,
+      "grad_norm": 0.6114712953567505,
+      "learning_rate": 2.0347173261654373e-05,
+      "loss": 0.367,
+      "step": 28680
+    },
+    {
+      "epoch": 10.83050207625519,
+      "grad_norm": 0.6642469763755798,
+      "learning_rate": 2.03139006712903e-05,
+      "loss": 0.3674,
+      "step": 28690
+    },
+    {
+      "epoch": 10.834277085692714,
+      "grad_norm": 0.6511169075965881,
+      "learning_rate": 2.028064837109519e-05,
+      "loss": 0.3648,
+      "step": 28700
+    },
+    {
+      "epoch": 10.838052095130237,
+      "grad_norm": 0.7235956788063049,
+      "learning_rate": 2.0247416383796685e-05,
+      "loss": 0.3712,
+      "step": 28710
+    },
+    {
+      "epoch": 10.841827104567761,
+      "grad_norm": 0.6243189573287964,
+      "learning_rate": 2.0214204732108548e-05,
+      "loss": 0.3573,
+      "step": 28720
+    },
+    {
+      "epoch": 10.845602114005285,
+      "grad_norm": 0.6611545085906982,
+      "learning_rate": 2.0181013438730596e-05,
+      "loss": 0.3603,
+      "step": 28730
+    },
+    {
+      "epoch": 10.849377123442808,
+      "grad_norm": 0.6944743990898132,
+      "learning_rate": 2.0147842526348783e-05,
+      "loss": 0.3589,
+      "step": 28740
+    },
+    {
+      "epoch": 10.853152132880332,
+      "grad_norm": 0.6612570881843567,
+      "learning_rate": 2.011469201763511e-05,
+      "loss": 0.3635,
+      "step": 28750
+    },
+    {
+      "epoch": 10.856927142317856,
+      "grad_norm": 0.6813220977783203,
+      "learning_rate": 2.0081561935247665e-05,
+      "loss": 0.3619,
+      "step": 28760
+    },
+    {
+      "epoch": 10.86070215175538,
+      "grad_norm": 0.638427734375,
+      "learning_rate": 2.0048452301830523e-05,
+      "loss": 0.3605,
+      "step": 28770
+    },
+    {
+      "epoch": 10.864477161192903,
+      "grad_norm": 0.607269823551178,
+      "learning_rate": 2.0015363140013788e-05,
+      "loss": 0.3669,
+      "step": 28780
+    },
+    {
+      "epoch": 10.868252170630427,
+      "grad_norm": 0.6158018708229065,
+      "learning_rate": 1.9982294472413606e-05,
+      "loss": 0.3642,
+      "step": 28790
+    },
+    {
+      "epoch": 10.87202718006795,
+      "grad_norm": 0.650993287563324,
+      "learning_rate": 1.9949246321632103e-05,
+      "loss": 0.3563,
+      "step": 28800
+    },
+    {
+      "epoch": 10.875802189505475,
+      "grad_norm": 0.6607373952865601,
+      "learning_rate": 1.9916218710257377e-05,
+      "loss": 0.3608,
+      "step": 28810
+    },
+    {
+      "epoch": 10.879577198942997,
+      "grad_norm": 0.7220985889434814,
+      "learning_rate": 1.988321166086351e-05,
+      "loss": 0.3646,
+      "step": 28820
+    },
+    {
+      "epoch": 10.88335220838052,
+      "grad_norm": 0.7125725150108337,
+      "learning_rate": 1.9850225196010468e-05,
+      "loss": 0.3605,
+      "step": 28830
+    },
+    {
+      "epoch": 10.887127217818044,
+      "grad_norm": 0.6737269759178162,
+      "learning_rate": 1.981725933824421e-05,
+      "loss": 0.363,
+      "step": 28840
+    },
+    {
+      "epoch": 10.890902227255568,
+      "grad_norm": 0.6922343373298645,
+      "learning_rate": 1.978431411009661e-05,
+      "loss": 0.3687,
+      "step": 28850
+    },
+    {
+      "epoch": 10.894677236693092,
+      "grad_norm": 4.779869079589844,
+      "learning_rate": 1.9751389534085375e-05,
+      "loss": 0.3591,
+      "step": 28860
+    },
+    {
+      "epoch": 10.898452246130615,
+      "grad_norm": 0.6346884369850159,
+      "learning_rate": 1.9718485632714184e-05,
+      "loss": 0.364,
+      "step": 28870
+    },
+    {
+      "epoch": 10.902227255568139,
+      "grad_norm": 0.6880037188529968,
+      "learning_rate": 1.968560242847251e-05,
+      "loss": 0.3693,
+      "step": 28880
+    },
+    {
+      "epoch": 10.906002265005663,
+      "grad_norm": 0.7235897183418274,
+      "learning_rate": 1.965273994383573e-05,
+      "loss": 0.3605,
+      "step": 28890
+    },
+    {
+      "epoch": 10.909777274443186,
+      "grad_norm": 0.6026926636695862,
+      "learning_rate": 1.961989820126504e-05,
+      "loss": 0.3659,
+      "step": 28900
+    },
+    {
+      "epoch": 10.91355228388071,
+      "grad_norm": 0.6524337530136108,
+      "learning_rate": 1.958707722320746e-05,
+      "loss": 0.3559,
+      "step": 28910
+    },
+    {
+      "epoch": 10.917327293318234,
+      "grad_norm": 0.6593090295791626,
+      "learning_rate": 1.955427703209584e-05,
+      "loss": 0.366,
+      "step": 28920
+    },
+    {
+      "epoch": 10.921102302755758,
+      "grad_norm": 0.6755902767181396,
+      "learning_rate": 1.9521497650348764e-05,
+      "loss": 0.3599,
+      "step": 28930
+    },
+    {
+      "epoch": 10.924877312193281,
+      "grad_norm": 0.7313739061355591,
+      "learning_rate": 1.948873910037067e-05,
+      "loss": 0.363,
+      "step": 28940
+    },
+    {
+      "epoch": 10.928652321630803,
+      "grad_norm": 0.686374306678772,
+      "learning_rate": 1.9456001404551678e-05,
+      "loss": 0.377,
+      "step": 28950
+    },
+    {
+      "epoch": 10.932427331068327,
+      "grad_norm": 0.789500892162323,
+      "learning_rate": 1.942328458526771e-05,
+      "loss": 0.3672,
+      "step": 28960
+    },
+    {
+      "epoch": 10.93620234050585,
+      "grad_norm": 0.6305020451545715,
+      "learning_rate": 1.9390588664880427e-05,
+      "loss": 0.3595,
+      "step": 28970
+    },
+    {
+      "epoch": 10.939977349943375,
+      "grad_norm": 0.6587271690368652,
+      "learning_rate": 1.9357913665737145e-05,
+      "loss": 0.3596,
+      "step": 28980
+    },
+    {
+      "epoch": 10.943752359380898,
+      "grad_norm": 0.6776351928710938,
+      "learning_rate": 1.932525961017093e-05,
+      "loss": 0.3646,
+      "step": 28990
+    },
+    {
+      "epoch": 10.947527368818422,
+      "grad_norm": 0.6554774641990662,
+      "learning_rate": 1.9292626520500533e-05,
+      "loss": 0.3653,
+      "step": 29000
+    },
+    {
+      "epoch": 10.951302378255946,
+      "grad_norm": 0.702714204788208,
+      "learning_rate": 1.9260014419030354e-05,
+      "loss": 0.3693,
+      "step": 29010
+    },
+    {
+      "epoch": 10.95507738769347,
+      "grad_norm": 0.692285418510437,
+      "learning_rate": 1.9227423328050475e-05,
+      "loss": 0.3628,
+      "step": 29020
+    },
+    {
+      "epoch": 10.958852397130993,
+      "grad_norm": 0.6136350631713867,
+      "learning_rate": 1.9194853269836582e-05,
+      "loss": 0.3628,
+      "step": 29030
+    },
+    {
+      "epoch": 10.962627406568517,
+      "grad_norm": 0.8171948790550232,
+      "learning_rate": 1.916230426664999e-05,
+      "loss": 0.3799,
+      "step": 29040
+    },
+    {
+      "epoch": 10.96640241600604,
+      "grad_norm": 0.6025060415267944,
+      "learning_rate": 1.912977634073765e-05,
+      "loss": 0.3664,
+      "step": 29050
+    },
+    {
+      "epoch": 10.970177425443563,
+      "grad_norm": 0.678777277469635,
+      "learning_rate": 1.9097269514332083e-05,
+      "loss": 0.3663,
+      "step": 29060
+    },
+    {
+      "epoch": 10.973952434881086,
+      "grad_norm": 0.6317446827888489,
+      "learning_rate": 1.9064783809651433e-05,
+      "loss": 0.3644,
+      "step": 29070
+    },
+    {
+      "epoch": 10.97772744431861,
+      "grad_norm": 0.6572142243385315,
+      "learning_rate": 1.9032319248899333e-05,
+      "loss": 0.3561,
+      "step": 29080
+    },
+    {
+      "epoch": 10.981502453756134,
+      "grad_norm": 0.5952383279800415,
+      "learning_rate": 1.8999875854265015e-05,
+      "loss": 0.3578,
+      "step": 29090
+    },
+    {
+      "epoch": 10.985277463193658,
+      "grad_norm": 0.6633411645889282,
+      "learning_rate": 1.8967453647923232e-05,
+      "loss": 0.362,
+      "step": 29100
+    },
+    {
+      "epoch": 10.989052472631181,
+      "grad_norm": 0.6560747027397156,
+      "learning_rate": 1.893505265203427e-05,
+      "loss": 0.3602,
+      "step": 29110
+    },
+    {
+      "epoch": 10.992827482068705,
+      "grad_norm": 0.6242331862449646,
+      "learning_rate": 1.8902672888743907e-05,
+      "loss": 0.3589,
+      "step": 29120
+    },
+    {
+      "epoch": 10.996602491506229,
+      "grad_norm": 0.6771829724311829,
+      "learning_rate": 1.8870314380183396e-05,
+      "loss": 0.3618,
+      "step": 29130
+    },
+    {
+      "epoch": 11.000377500943753,
+      "grad_norm": 1.5967614650726318,
+      "learning_rate": 1.8837977148469448e-05,
+      "loss": 0.3566,
+      "step": 29140
+    },
+    {
+      "epoch": 11.004152510381276,
+      "grad_norm": 0.6474411487579346,
+      "learning_rate": 1.880566121570429e-05,
+      "loss": 0.3596,
+      "step": 29150
+    },
+    {
+      "epoch": 11.0079275198188,
+      "grad_norm": 0.6497718691825867,
+      "learning_rate": 1.877336660397554e-05,
+      "loss": 0.3564,
+      "step": 29160
+    },
+    {
+      "epoch": 11.011702529256324,
+      "grad_norm": 0.6360942125320435,
+      "learning_rate": 1.874109333535628e-05,
+      "loss": 0.3681,
+      "step": 29170
+    },
+    {
+      "epoch": 11.015477538693847,
+      "grad_norm": 0.7708438634872437,
+      "learning_rate": 1.870884143190496e-05,
+      "loss": 0.3651,
+      "step": 29180
+    },
+    {
+      "epoch": 11.01925254813137,
+      "grad_norm": 0.8585805892944336,
+      "learning_rate": 1.867661091566546e-05,
+      "loss": 0.3656,
+      "step": 29190
+    },
+    {
+      "epoch": 11.023027557568893,
+      "grad_norm": 2.161421537399292,
+      "learning_rate": 1.864440180866704e-05,
+      "loss": 0.3588,
+      "step": 29200
+    },
+    {
+      "epoch": 11.026802567006417,
+      "grad_norm": 0.6585597991943359,
+      "learning_rate": 1.8612214132924317e-05,
+      "loss": 0.3567,
+      "step": 29210
+    },
+    {
+      "epoch": 11.03057757644394,
+      "grad_norm": 0.6775062084197998,
+      "learning_rate": 1.858004791043728e-05,
+      "loss": 0.3581,
+      "step": 29220
+    },
+    {
+      "epoch": 11.034352585881464,
+      "grad_norm": 0.8132985234260559,
+      "learning_rate": 1.854790316319123e-05,
+      "loss": 0.3758,
+      "step": 29230
+    },
+    {
+      "epoch": 11.038127595318988,
+      "grad_norm": 0.6738925576210022,
+      "learning_rate": 1.8515779913156766e-05,
+      "loss": 0.3583,
+      "step": 29240
+    },
+    {
+      "epoch": 11.041902604756512,
+      "grad_norm": 0.6279427409172058,
+      "learning_rate": 1.848367818228986e-05,
+      "loss": 0.3697,
+      "step": 29250
+    },
+    {
+      "epoch": 11.045677614194036,
+      "grad_norm": 0.6961294412612915,
+      "learning_rate": 1.8451597992531733e-05,
+      "loss": 0.3619,
+      "step": 29260
+    },
+    {
+      "epoch": 11.04945262363156,
+      "grad_norm": 0.8088012337684631,
+      "learning_rate": 1.8419539365808914e-05,
+      "loss": 0.3755,
+      "step": 29270
+    },
+    {
+      "epoch": 11.053227633069083,
+      "grad_norm": 0.6103382706642151,
+      "learning_rate": 1.838750232403313e-05,
+      "loss": 0.3592,
+      "step": 29280
+    },
+    {
+      "epoch": 11.057002642506607,
+      "grad_norm": 0.6636168956756592,
+      "learning_rate": 1.835548688910142e-05,
+      "loss": 0.3726,
+      "step": 29290
+    },
+    {
+      "epoch": 11.06077765194413,
+      "grad_norm": 0.6306051015853882,
+      "learning_rate": 1.8323493082896037e-05,
+      "loss": 0.3665,
+      "step": 29300
+    },
+    {
+      "epoch": 11.064552661381654,
+      "grad_norm": 0.6422659754753113,
+      "learning_rate": 1.8291520927284454e-05,
+      "loss": 0.3643,
+      "step": 29310
+    },
+    {
+      "epoch": 11.068327670819176,
+      "grad_norm": 0.6881163716316223,
+      "learning_rate": 1.8259570444119305e-05,
+      "loss": 0.363,
+      "step": 29320
+    },
+    {
+      "epoch": 11.0721026802567,
+      "grad_norm": 0.5860868096351624,
+      "learning_rate": 1.8227641655238488e-05,
+      "loss": 0.3639,
+      "step": 29330
+    },
+    {
+      "epoch": 11.075877689694224,
+      "grad_norm": 0.6115444302558899,
+      "learning_rate": 1.819573458246498e-05,
+      "loss": 0.3687,
+      "step": 29340
+    },
+    {
+      "epoch": 11.079652699131747,
+      "grad_norm": 0.6907765865325928,
+      "learning_rate": 1.816384924760699e-05,
+      "loss": 0.367,
+      "step": 29350
+    },
+    {
+      "epoch": 11.083427708569271,
+      "grad_norm": 0.7525956034660339,
+      "learning_rate": 1.813198567245784e-05,
+      "loss": 0.3615,
+      "step": 29360
+    },
+    {
+      "epoch": 11.087202718006795,
+      "grad_norm": 0.625262975692749,
+      "learning_rate": 1.8100143878796006e-05,
+      "loss": 0.3565,
+      "step": 29370
+    },
+    {
+      "epoch": 11.090977727444319,
+      "grad_norm": 0.6794587969779968,
+      "learning_rate": 1.8068323888385015e-05,
+      "loss": 0.363,
+      "step": 29380
+    },
+    {
+      "epoch": 11.094752736881842,
+      "grad_norm": 0.6202002167701721,
+      "learning_rate": 1.803652572297355e-05,
+      "loss": 0.3639,
+      "step": 29390
+    },
+    {
+      "epoch": 11.098527746319366,
+      "grad_norm": 0.679071307182312,
+      "learning_rate": 1.8004749404295353e-05,
+      "loss": 0.3658,
+      "step": 29400
+    },
+    {
+      "epoch": 11.10230275575689,
+      "grad_norm": 0.6545715928077698,
+      "learning_rate": 1.797299495406926e-05,
+      "loss": 0.3692,
+      "step": 29410
+    },
+    {
+      "epoch": 11.106077765194414,
+      "grad_norm": 0.6335902810096741,
+      "learning_rate": 1.7941262393999103e-05,
+      "loss": 0.3591,
+      "step": 29420
+    },
+    {
+      "epoch": 11.109852774631937,
+      "grad_norm": 1.0289889574050903,
+      "learning_rate": 1.7909551745773816e-05,
+      "loss": 0.3531,
+      "step": 29430
+    },
+    {
+      "epoch": 11.113627784069461,
+      "grad_norm": 0.6615308523178101,
+      "learning_rate": 1.7877863031067304e-05,
+      "loss": 0.3552,
+      "step": 29440
+    },
+    {
+      "epoch": 11.117402793506983,
+      "grad_norm": 0.6820453405380249,
+      "learning_rate": 1.7846196271538516e-05,
+      "loss": 0.3688,
+      "step": 29450
+    },
+    {
+      "epoch": 11.121177802944507,
+      "grad_norm": 0.6518820524215698,
+      "learning_rate": 1.7814551488831384e-05,
+      "loss": 0.3598,
+      "step": 29460
+    },
+    {
+      "epoch": 11.12495281238203,
+      "grad_norm": 0.725908637046814,
+      "learning_rate": 1.7782928704574835e-05,
+      "loss": 0.3621,
+      "step": 29470
+    },
+    {
+      "epoch": 11.128727821819554,
+      "grad_norm": 2.0189263820648193,
+      "learning_rate": 1.775132794038271e-05,
+      "loss": 0.3616,
+      "step": 29480
+    },
+    {
+      "epoch": 11.132502831257078,
+      "grad_norm": 0.6810202598571777,
+      "learning_rate": 1.7719749217853855e-05,
+      "loss": 0.3604,
+      "step": 29490
+    },
+    {
+      "epoch": 11.136277840694602,
+      "grad_norm": 0.6146163940429688,
+      "learning_rate": 1.7688192558572038e-05,
+      "loss": 0.3593,
+      "step": 29500
+    },
+    {
+      "epoch": 11.140052850132125,
+      "grad_norm": 0.7058233022689819,
+      "learning_rate": 1.7656657984105906e-05,
+      "loss": 0.3563,
+      "step": 29510
+    },
+    {
+      "epoch": 11.14382785956965,
+      "grad_norm": 0.6673465967178345,
+      "learning_rate": 1.7625145516009068e-05,
+      "loss": 0.3613,
+      "step": 29520
+    },
+    {
+      "epoch": 11.147602869007173,
+      "grad_norm": 0.6134123802185059,
+      "learning_rate": 1.7593655175820005e-05,
+      "loss": 0.3697,
+      "step": 29530
+    },
+    {
+      "epoch": 11.151377878444697,
+      "grad_norm": 0.7190696597099304,
+      "learning_rate": 1.7562186985062046e-05,
+      "loss": 0.3666,
+      "step": 29540
+    },
+    {
+      "epoch": 11.15515288788222,
+      "grad_norm": 0.7087527513504028,
+      "learning_rate": 1.7530740965243403e-05,
+      "loss": 0.3696,
+      "step": 29550
+    },
+    {
+      "epoch": 11.158927897319744,
+      "grad_norm": 1.1012113094329834,
+      "learning_rate": 1.7499317137857153e-05,
+      "loss": 0.3582,
+      "step": 29560
+    },
+    {
+      "epoch": 11.162702906757266,
+      "grad_norm": 0.6906222701072693,
+      "learning_rate": 1.7467915524381184e-05,
+      "loss": 0.3591,
+      "step": 29570
+    },
+    {
+      "epoch": 11.16647791619479,
+      "grad_norm": 0.6418552398681641,
+      "learning_rate": 1.7436536146278182e-05,
+      "loss": 0.3702,
+      "step": 29580
+    },
+    {
+      "epoch": 11.170252925632314,
+      "grad_norm": 1.7984684705734253,
+      "learning_rate": 1.7405179024995688e-05,
+      "loss": 0.3667,
+      "step": 29590
+    },
+    {
+      "epoch": 11.174027935069837,
+      "grad_norm": 1.6302399635314941,
+      "learning_rate": 1.737384418196596e-05,
+      "loss": 0.3828,
+      "step": 29600
+    },
+    {
+      "epoch": 11.177802944507361,
+      "grad_norm": 0.7185459136962891,
+      "learning_rate": 1.734253163860609e-05,
+      "loss": 0.3602,
+      "step": 29610
+    },
+    {
+      "epoch": 11.181577953944885,
+      "grad_norm": 0.6619534492492676,
+      "learning_rate": 1.7311241416317896e-05,
+      "loss": 0.3618,
+      "step": 29620
+    },
+    {
+      "epoch": 11.185352963382408,
+      "grad_norm": 0.6776392459869385,
+      "learning_rate": 1.7279973536487982e-05,
+      "loss": 0.3626,
+      "step": 29630
+    },
+    {
+      "epoch": 11.189127972819932,
+      "grad_norm": 0.8006789088249207,
+      "learning_rate": 1.724872802048761e-05,
+      "loss": 0.3607,
+      "step": 29640
+    },
+    {
+      "epoch": 11.192902982257456,
+      "grad_norm": 0.6797590255737305,
+      "learning_rate": 1.7217504889672803e-05,
+      "loss": 0.3616,
+      "step": 29650
+    },
+    {
+      "epoch": 11.19667799169498,
+      "grad_norm": 0.6782128810882568,
+      "learning_rate": 1.7186304165384287e-05,
+      "loss": 0.3651,
+      "step": 29660
+    },
+    {
+      "epoch": 11.200453001132503,
+      "grad_norm": 0.7670999765396118,
+      "learning_rate": 1.7155125868947475e-05,
+      "loss": 0.3615,
+      "step": 29670
+    },
+    {
+      "epoch": 11.204228010570027,
+      "grad_norm": 0.6768984198570251,
+      "learning_rate": 1.7123970021672404e-05,
+      "loss": 0.3597,
+      "step": 29680
+    },
+    {
+      "epoch": 11.20800302000755,
+      "grad_norm": 0.6669506430625916,
+      "learning_rate": 1.709283664485384e-05,
+      "loss": 0.3612,
+      "step": 29690
+    },
+    {
+      "epoch": 11.211778029445073,
+      "grad_norm": 0.6505656838417053,
+      "learning_rate": 1.7061725759771113e-05,
+      "loss": 0.3583,
+      "step": 29700
+    },
+    {
+      "epoch": 11.215553038882597,
+      "grad_norm": 0.6229871511459351,
+      "learning_rate": 1.7030637387688248e-05,
+      "loss": 0.364,
+      "step": 29710
+    },
+    {
+      "epoch": 11.21932804832012,
+      "grad_norm": 2.578798770904541,
+      "learning_rate": 1.6999571549853836e-05,
+      "loss": 0.361,
+      "step": 29720
+    },
+    {
+      "epoch": 11.223103057757644,
+      "grad_norm": 0.6974425911903381,
+      "learning_rate": 1.696852826750112e-05,
+      "loss": 0.3677,
+      "step": 29730
+    },
+    {
+      "epoch": 11.226878067195168,
+      "grad_norm": 0.6888813376426697,
+      "learning_rate": 1.6937507561847844e-05,
+      "loss": 0.3687,
+      "step": 29740
+    },
+    {
+      "epoch": 11.230653076632692,
+      "grad_norm": 0.6524753570556641,
+      "learning_rate": 1.6906509454096385e-05,
+      "loss": 0.3633,
+      "step": 29750
+    },
+    {
+      "epoch": 11.234428086070215,
+      "grad_norm": 0.7103590965270996,
+      "learning_rate": 1.687553396543367e-05,
+      "loss": 0.3671,
+      "step": 29760
+    },
+    {
+      "epoch": 11.238203095507739,
+      "grad_norm": 0.6661110520362854,
+      "learning_rate": 1.6844581117031154e-05,
+      "loss": 0.3597,
+      "step": 29770
+    },
+    {
+      "epoch": 11.241978104945263,
+      "grad_norm": 1.3560032844543457,
+      "learning_rate": 1.681365093004481e-05,
+      "loss": 0.3731,
+      "step": 29780
+    },
+    {
+      "epoch": 11.245753114382786,
+      "grad_norm": 0.6432333588600159,
+      "learning_rate": 1.678274342561511e-05,
+      "loss": 0.3623,
+      "step": 29790
+    },
+    {
+      "epoch": 11.24952812382031,
+      "grad_norm": 0.6939373016357422,
+      "learning_rate": 1.675185862486706e-05,
+      "loss": 0.367,
+      "step": 29800
+    },
+    {
+      "epoch": 11.253303133257834,
+      "grad_norm": 1.7712675333023071,
+      "learning_rate": 1.6720996548910127e-05,
+      "loss": 0.3675,
+      "step": 29810
+    },
+    {
+      "epoch": 11.257078142695356,
+      "grad_norm": 0.7062354683876038,
+      "learning_rate": 1.6690157218838247e-05,
+      "loss": 0.3667,
+      "step": 29820
+    },
+    {
+      "epoch": 11.26085315213288,
+      "grad_norm": 0.6343713402748108,
+      "learning_rate": 1.665934065572984e-05,
+      "loss": 0.3529,
+      "step": 29830
+    },
+    {
+      "epoch": 11.264628161570403,
+      "grad_norm": 0.6509836316108704,
+      "learning_rate": 1.6628546880647688e-05,
+      "loss": 0.3628,
+      "step": 29840
+    },
+    {
+      "epoch": 11.268403171007927,
+      "grad_norm": 0.6391473412513733,
+      "learning_rate": 1.6597775914639076e-05,
+      "loss": 0.3665,
+      "step": 29850
+    },
+    {
+      "epoch": 11.27217818044545,
+      "grad_norm": 0.7817753553390503,
+      "learning_rate": 1.6567027778735654e-05,
+      "loss": 0.3619,
+      "step": 29860
+    },
+    {
+      "epoch": 11.275953189882975,
+      "grad_norm": 0.5986528992652893,
+      "learning_rate": 1.653630249395351e-05,
+      "loss": 0.3592,
+      "step": 29870
+    },
+    {
+      "epoch": 11.279728199320498,
+      "grad_norm": 0.6571201682090759,
+      "learning_rate": 1.6505600081293072e-05,
+      "loss": 0.3613,
+      "step": 29880
+    },
+    {
+      "epoch": 11.283503208758022,
+      "grad_norm": 0.7287798523902893,
+      "learning_rate": 1.647492056173912e-05,
+      "loss": 0.3705,
+      "step": 29890
+    },
+    {
+      "epoch": 11.287278218195546,
+      "grad_norm": 0.6623769998550415,
+      "learning_rate": 1.6444263956260848e-05,
+      "loss": 0.3559,
+      "step": 29900
+    },
+    {
+      "epoch": 11.29105322763307,
+      "grad_norm": 0.6253150105476379,
+      "learning_rate": 1.641363028581175e-05,
+      "loss": 0.363,
+      "step": 29910
+    },
+    {
+      "epoch": 11.294828237070593,
+      "grad_norm": 0.6631579399108887,
+      "learning_rate": 1.638301957132965e-05,
+      "loss": 0.3649,
+      "step": 29920
+    },
+    {
+      "epoch": 11.298603246508117,
+      "grad_norm": 0.7632430791854858,
+      "learning_rate": 1.6352431833736703e-05,
+      "loss": 0.3576,
+      "step": 29930
+    },
+    {
+      "epoch": 11.30237825594564,
+      "grad_norm": 0.8125953674316406,
+      "learning_rate": 1.6321867093939298e-05,
+      "loss": 0.3624,
+      "step": 29940
+    },
+    {
+      "epoch": 11.306153265383163,
+      "grad_norm": 0.6205570101737976,
+      "learning_rate": 1.629132537282817e-05,
+      "loss": 0.3575,
+      "step": 29950
+    },
+    {
+      "epoch": 11.309928274820686,
+      "grad_norm": 0.6771644949913025,
+      "learning_rate": 1.62608066912783e-05,
+      "loss": 0.3572,
+      "step": 29960
+    },
+    {
+      "epoch": 11.31370328425821,
+      "grad_norm": 0.6554526090621948,
+      "learning_rate": 1.623031107014893e-05,
+      "loss": 0.3662,
+      "step": 29970
+    },
+    {
+      "epoch": 11.317478293695734,
+      "grad_norm": 0.6397532820701599,
+      "learning_rate": 1.619983853028351e-05,
+      "loss": 0.3668,
+      "step": 29980
+    },
+    {
+      "epoch": 11.321253303133258,
+      "grad_norm": 0.6637176871299744,
+      "learning_rate": 1.6169389092509724e-05,
+      "loss": 0.3611,
+      "step": 29990
+    },
+    {
+      "epoch": 11.325028312570781,
+      "grad_norm": 0.6467320322990417,
+      "learning_rate": 1.6138962777639494e-05,
+      "loss": 0.3625,
+      "step": 30000
+    },
+    {
+      "epoch": 11.328803322008305,
+      "grad_norm": 0.6978408694267273,
+      "learning_rate": 1.610855960646891e-05,
+      "loss": 0.3677,
+      "step": 30010
+    },
+    {
+      "epoch": 11.332578331445829,
+      "grad_norm": 0.6771886348724365,
+      "learning_rate": 1.607817959977826e-05,
+      "loss": 0.3654,
+      "step": 30020
+    },
+    {
+      "epoch": 11.336353340883353,
+      "grad_norm": 0.691331684589386,
+      "learning_rate": 1.6047822778332005e-05,
+      "loss": 0.363,
+      "step": 30030
+    },
+    {
+      "epoch": 11.340128350320876,
+      "grad_norm": 0.6371560096740723,
+      "learning_rate": 1.6017489162878713e-05,
+      "loss": 0.3671,
+      "step": 30040
+    },
+    {
+      "epoch": 11.3439033597584,
+      "grad_norm": 0.6647712588310242,
+      "learning_rate": 1.5987178774151147e-05,
+      "loss": 0.3557,
+      "step": 30050
+    },
+    {
+      "epoch": 11.347678369195924,
+      "grad_norm": 0.6369796991348267,
+      "learning_rate": 1.5956891632866184e-05,
+      "loss": 0.3607,
+      "step": 30060
+    },
+    {
+      "epoch": 11.351453378633446,
+      "grad_norm": 0.6717673540115356,
+      "learning_rate": 1.592662775972476e-05,
+      "loss": 0.3559,
+      "step": 30070
+    },
+    {
+      "epoch": 11.35522838807097,
+      "grad_norm": 0.6065929532051086,
+      "learning_rate": 1.5896387175411986e-05,
+      "loss": 0.3572,
+      "step": 30080
+    },
+    {
+      "epoch": 11.359003397508493,
+      "grad_norm": 0.7193185091018677,
+      "learning_rate": 1.586616990059699e-05,
+      "loss": 0.3617,
+      "step": 30090
+    },
+    {
+      "epoch": 11.362778406946017,
+      "grad_norm": 0.7065017223358154,
+      "learning_rate": 1.5835975955933002e-05,
+      "loss": 0.3536,
+      "step": 30100
+    },
+    {
+      "epoch": 11.36655341638354,
+      "grad_norm": 0.7551389336585999,
+      "learning_rate": 1.5805805362057295e-05,
+      "loss": 0.3586,
+      "step": 30110
+    },
+    {
+      "epoch": 11.370328425821064,
+      "grad_norm": 0.6497236490249634,
+      "learning_rate": 1.57756581395912e-05,
+      "loss": 0.3634,
+      "step": 30120
+    },
+    {
+      "epoch": 11.374103435258588,
+      "grad_norm": 0.6430761814117432,
+      "learning_rate": 1.574553430914006e-05,
+      "loss": 0.3627,
+      "step": 30130
+    },
+    {
+      "epoch": 11.377878444696112,
+      "grad_norm": 0.6875909566879272,
+      "learning_rate": 1.5715433891293206e-05,
+      "loss": 0.3609,
+      "step": 30140
+    },
+    {
+      "epoch": 11.381653454133636,
+      "grad_norm": 0.61098313331604,
+      "learning_rate": 1.5685356906624016e-05,
+      "loss": 0.3585,
+      "step": 30150
+    },
+    {
+      "epoch": 11.38542846357116,
+      "grad_norm": 0.6427933573722839,
+      "learning_rate": 1.56553033756898e-05,
+      "loss": 0.3601,
+      "step": 30160
+    },
+    {
+      "epoch": 11.389203473008683,
+      "grad_norm": 0.662592351436615,
+      "learning_rate": 1.5625273319031873e-05,
+      "loss": 0.364,
+      "step": 30170
+    },
+    {
+      "epoch": 11.392978482446207,
+      "grad_norm": 0.7252653241157532,
+      "learning_rate": 1.5595266757175518e-05,
+      "loss": 0.3614,
+      "step": 30180
+    },
+    {
+      "epoch": 11.39675349188373,
+      "grad_norm": 0.6587193608283997,
+      "learning_rate": 1.55652837106299e-05,
+      "loss": 0.3618,
+      "step": 30190
+    },
+    {
+      "epoch": 11.400528501321253,
+      "grad_norm": 0.7090880274772644,
+      "learning_rate": 1.5535324199888164e-05,
+      "loss": 0.3572,
+      "step": 30200
+    },
+    {
+      "epoch": 11.404303510758776,
+      "grad_norm": 0.6730210781097412,
+      "learning_rate": 1.5505388245427355e-05,
+      "loss": 0.3621,
+      "step": 30210
+    },
+    {
+      "epoch": 11.4080785201963,
+      "grad_norm": 0.6814711689949036,
+      "learning_rate": 1.5475475867708418e-05,
+      "loss": 0.3647,
+      "step": 30220
+    },
+    {
+      "epoch": 11.411853529633824,
+      "grad_norm": 0.6800294518470764,
+      "learning_rate": 1.5445587087176195e-05,
+      "loss": 0.3635,
+      "step": 30230
+    },
+    {
+      "epoch": 11.415628539071347,
+      "grad_norm": 0.6638886332511902,
+      "learning_rate": 1.5415721924259346e-05,
+      "loss": 0.3634,
+      "step": 30240
+    },
+    {
+      "epoch": 11.419403548508871,
+      "grad_norm": 0.6355138421058655,
+      "learning_rate": 1.538588039937047e-05,
+      "loss": 0.36,
+      "step": 30250
+    },
+    {
+      "epoch": 11.423178557946395,
+      "grad_norm": 0.6450951099395752,
+      "learning_rate": 1.535606253290594e-05,
+      "loss": 0.3644,
+      "step": 30260
+    },
+    {
+      "epoch": 11.426953567383919,
+      "grad_norm": 0.640677273273468,
+      "learning_rate": 1.5326268345245993e-05,
+      "loss": 0.3594,
+      "step": 30270
+    },
+    {
+      "epoch": 11.430728576821442,
+      "grad_norm": 0.7016587257385254,
+      "learning_rate": 1.5296497856754693e-05,
+      "loss": 0.3645,
+      "step": 30280
+    },
+    {
+      "epoch": 11.434503586258966,
+      "grad_norm": 0.6003674864768982,
+      "learning_rate": 1.5266751087779858e-05,
+      "loss": 0.3645,
+      "step": 30290
+    },
+    {
+      "epoch": 11.43827859569649,
+      "grad_norm": 0.6319372057914734,
+      "learning_rate": 1.5237028058653142e-05,
+      "loss": 0.3662,
+      "step": 30300
+    },
+    {
+      "epoch": 11.442053605134014,
+      "grad_norm": 0.6442000865936279,
+      "learning_rate": 1.5207328789689946e-05,
+      "loss": 0.3606,
+      "step": 30310
+    },
+    {
+      "epoch": 11.445828614571536,
+      "grad_norm": 0.6055095791816711,
+      "learning_rate": 1.5177653301189454e-05,
+      "loss": 0.3652,
+      "step": 30320
+    },
+    {
+      "epoch": 11.44960362400906,
+      "grad_norm": 0.7226874232292175,
+      "learning_rate": 1.5148001613434587e-05,
+      "loss": 0.3553,
+      "step": 30330
+    },
+    {
+      "epoch": 11.453378633446583,
+      "grad_norm": 0.7227387428283691,
+      "learning_rate": 1.5118373746691983e-05,
+      "loss": 0.3628,
+      "step": 30340
+    },
+    {
+      "epoch": 11.457153642884107,
+      "grad_norm": 0.6881961226463318,
+      "learning_rate": 1.5088769721211992e-05,
+      "loss": 0.359,
+      "step": 30350
+    },
+    {
+      "epoch": 11.46092865232163,
+      "grad_norm": 0.7442403435707092,
+      "learning_rate": 1.505918955722871e-05,
+      "loss": 0.3632,
+      "step": 30360
+    },
+    {
+      "epoch": 11.464703661759154,
+      "grad_norm": 0.6394659876823425,
+      "learning_rate": 1.5029633274959886e-05,
+      "loss": 0.3559,
+      "step": 30370
+    },
+    {
+      "epoch": 11.468478671196678,
+      "grad_norm": 0.6404333114624023,
+      "learning_rate": 1.5000100894606989e-05,
+      "loss": 0.3605,
+      "step": 30380
+    },
+    {
+      "epoch": 11.472253680634202,
+      "grad_norm": 0.7091559171676636,
+      "learning_rate": 1.4970592436355079e-05,
+      "loss": 0.3546,
+      "step": 30390
+    },
+    {
+      "epoch": 11.476028690071725,
+      "grad_norm": 0.6807871460914612,
+      "learning_rate": 1.4941107920372938e-05,
+      "loss": 0.3703,
+      "step": 30400
+    },
+    {
+      "epoch": 11.47980369950925,
+      "grad_norm": 0.6545217633247375,
+      "learning_rate": 1.4911647366812947e-05,
+      "loss": 0.3574,
+      "step": 30410
+    },
+    {
+      "epoch": 11.483578708946773,
+      "grad_norm": 0.6775568723678589,
+      "learning_rate": 1.4882210795811114e-05,
+      "loss": 0.3653,
+      "step": 30420
+    },
+    {
+      "epoch": 11.487353718384297,
+      "grad_norm": 0.7269086837768555,
+      "learning_rate": 1.485279822748708e-05,
+      "loss": 0.3608,
+      "step": 30430
+    },
+    {
+      "epoch": 11.49112872782182,
+      "grad_norm": 0.7882838249206543,
+      "learning_rate": 1.4823409681944034e-05,
+      "loss": 0.3708,
+      "step": 30440
+    },
+    {
+      "epoch": 11.494903737259342,
+      "grad_norm": 0.6561617255210876,
+      "learning_rate": 1.4794045179268763e-05,
+      "loss": 0.3586,
+      "step": 30450
+    },
+    {
+      "epoch": 11.498678746696866,
+      "grad_norm": 0.6408365368843079,
+      "learning_rate": 1.4764704739531643e-05,
+      "loss": 0.3661,
+      "step": 30460
+    },
+    {
+      "epoch": 11.50245375613439,
+      "grad_norm": 0.6767753958702087,
+      "learning_rate": 1.473538838278658e-05,
+      "loss": 0.3576,
+      "step": 30470
+    },
+    {
+      "epoch": 11.506228765571914,
+      "grad_norm": 0.6685946583747864,
+      "learning_rate": 1.4706096129071045e-05,
+      "loss": 0.3657,
+      "step": 30480
+    },
+    {
+      "epoch": 11.510003775009437,
+      "grad_norm": 0.7614819407463074,
+      "learning_rate": 1.4676827998405996e-05,
+      "loss": 0.3657,
+      "step": 30490
+    },
+    {
+      "epoch": 11.513778784446961,
+      "grad_norm": 0.6719235181808472,
+      "learning_rate": 1.4647584010795928e-05,
+      "loss": 0.3587,
+      "step": 30500
+    },
+    {
+      "epoch": 11.517553793884485,
+      "grad_norm": 0.6133652925491333,
+      "learning_rate": 1.4618364186228834e-05,
+      "loss": 0.3656,
+      "step": 30510
+    },
+    {
+      "epoch": 11.521328803322008,
+      "grad_norm": 0.6366606950759888,
+      "learning_rate": 1.4589168544676207e-05,
+      "loss": 0.3592,
+      "step": 30520
+    },
+    {
+      "epoch": 11.525103812759532,
+      "grad_norm": 0.6164776086807251,
+      "learning_rate": 1.455999710609296e-05,
+      "loss": 0.3579,
+      "step": 30530
+    },
+    {
+      "epoch": 11.528878822197056,
+      "grad_norm": 0.5788912177085876,
+      "learning_rate": 1.453084989041753e-05,
+      "loss": 0.3612,
+      "step": 30540
+    },
+    {
+      "epoch": 11.53265383163458,
+      "grad_norm": 0.6258981227874756,
+      "learning_rate": 1.4501726917571728e-05,
+      "loss": 0.3599,
+      "step": 30550
+    },
+    {
+      "epoch": 11.536428841072103,
+      "grad_norm": 0.8489433526992798,
+      "learning_rate": 1.4472628207460848e-05,
+      "loss": 0.3561,
+      "step": 30560
+    },
+    {
+      "epoch": 11.540203850509627,
+      "grad_norm": 0.7995439767837524,
+      "learning_rate": 1.4443553779973584e-05,
+      "loss": 0.3608,
+      "step": 30570
+    },
+    {
+      "epoch": 11.54397885994715,
+      "grad_norm": 0.642866313457489,
+      "learning_rate": 1.4414503654982037e-05,
+      "loss": 0.3593,
+      "step": 30580
+    },
+    {
+      "epoch": 11.547753869384673,
+      "grad_norm": 0.688869059085846,
+      "learning_rate": 1.4385477852341706e-05,
+      "loss": 0.3612,
+      "step": 30590
+    },
+    {
+      "epoch": 11.551528878822197,
+      "grad_norm": 0.6555159091949463,
+      "learning_rate": 1.4356476391891421e-05,
+      "loss": 0.3577,
+      "step": 30600
+    },
+    {
+      "epoch": 11.55530388825972,
+      "grad_norm": 0.8551262021064758,
+      "learning_rate": 1.4327499293453423e-05,
+      "loss": 0.371,
+      "step": 30610
+    },
+    {
+      "epoch": 11.559078897697244,
+      "grad_norm": 0.6704701781272888,
+      "learning_rate": 1.4298546576833304e-05,
+      "loss": 0.3564,
+      "step": 30620
+    },
+    {
+      "epoch": 11.562853907134768,
+      "grad_norm": 0.7134976387023926,
+      "learning_rate": 1.426961826181994e-05,
+      "loss": 0.3619,
+      "step": 30630
+    },
+    {
+      "epoch": 11.566628916572292,
+      "grad_norm": 0.6610779166221619,
+      "learning_rate": 1.4240714368185592e-05,
+      "loss": 0.3642,
+      "step": 30640
+    },
+    {
+      "epoch": 11.570403926009815,
+      "grad_norm": 0.6555269956588745,
+      "learning_rate": 1.4211834915685767e-05,
+      "loss": 0.3627,
+      "step": 30650
+    },
+    {
+      "epoch": 11.574178935447339,
+      "grad_norm": 0.677355170249939,
+      "learning_rate": 1.4182979924059313e-05,
+      "loss": 0.3527,
+      "step": 30660
+    },
+    {
+      "epoch": 11.577953944884863,
+      "grad_norm": 0.6868191361427307,
+      "learning_rate": 1.415414941302835e-05,
+      "loss": 0.3604,
+      "step": 30670
+    },
+    {
+      "epoch": 11.581728954322386,
+      "grad_norm": 0.7704634666442871,
+      "learning_rate": 1.4125343402298257e-05,
+      "loss": 0.3605,
+      "step": 30680
+    },
+    {
+      "epoch": 11.58550396375991,
+      "grad_norm": 0.6135473251342773,
+      "learning_rate": 1.4096561911557688e-05,
+      "loss": 0.36,
+      "step": 30690
+    },
+    {
+      "epoch": 11.589278973197432,
+      "grad_norm": 1.2090810537338257,
+      "learning_rate": 1.4067804960478481e-05,
+      "loss": 0.3648,
+      "step": 30700
+    },
+    {
+      "epoch": 11.593053982634956,
+      "grad_norm": 0.6949913501739502,
+      "learning_rate": 1.4039072568715784e-05,
+      "loss": 0.3637,
+      "step": 30710
+    },
+    {
+      "epoch": 11.59682899207248,
+      "grad_norm": 0.6498703360557556,
+      "learning_rate": 1.401036475590788e-05,
+      "loss": 0.3651,
+      "step": 30720
+    },
+    {
+      "epoch": 11.600604001510003,
+      "grad_norm": 0.7282187342643738,
+      "learning_rate": 1.3981681541676301e-05,
+      "loss": 0.3629,
+      "step": 30730
+    },
+    {
+      "epoch": 11.604379010947527,
+      "grad_norm": 0.6524527668952942,
+      "learning_rate": 1.3953022945625772e-05,
+      "loss": 0.3581,
+      "step": 30740
+    },
+    {
+      "epoch": 11.60815402038505,
+      "grad_norm": 0.6055677533149719,
+      "learning_rate": 1.3924388987344138e-05,
+      "loss": 0.3575,
+      "step": 30750
+    },
+    {
+      "epoch": 11.611929029822575,
+      "grad_norm": 0.717749834060669,
+      "learning_rate": 1.3895779686402465e-05,
+      "loss": 0.3602,
+      "step": 30760
+    },
+    {
+      "epoch": 11.615704039260098,
+      "grad_norm": 0.7012677788734436,
+      "learning_rate": 1.3867195062354931e-05,
+      "loss": 0.3574,
+      "step": 30770
+    },
+    {
+      "epoch": 11.619479048697622,
+      "grad_norm": 0.7111396193504333,
+      "learning_rate": 1.3838635134738858e-05,
+      "loss": 0.3537,
+      "step": 30780
+    },
+    {
+      "epoch": 11.623254058135146,
+      "grad_norm": 0.7217731475830078,
+      "learning_rate": 1.3810099923074704e-05,
+      "loss": 0.3599,
+      "step": 30790
+    },
+    {
+      "epoch": 11.62702906757267,
+      "grad_norm": 0.609724760055542,
+      "learning_rate": 1.3781589446866006e-05,
+      "loss": 0.3743,
+      "step": 30800
+    },
+    {
+      "epoch": 11.630804077010193,
+      "grad_norm": 0.6935325860977173,
+      "learning_rate": 1.3753103725599387e-05,
+      "loss": 0.3577,
+      "step": 30810
+    },
+    {
+      "epoch": 11.634579086447715,
+      "grad_norm": 0.7039898037910461,
+      "learning_rate": 1.3724642778744579e-05,
+      "loss": 0.3615,
+      "step": 30820
+    },
+    {
+      "epoch": 11.638354095885239,
+      "grad_norm": 0.6811739802360535,
+      "learning_rate": 1.3696206625754376e-05,
+      "loss": 0.3589,
+      "step": 30830
+    },
+    {
+      "epoch": 11.642129105322763,
+      "grad_norm": 0.6018708348274231,
+      "learning_rate": 1.3667795286064639e-05,
+      "loss": 0.3626,
+      "step": 30840
+    },
+    {
+      "epoch": 11.645904114760286,
+      "grad_norm": 0.6382670998573303,
+      "learning_rate": 1.3639408779094215e-05,
+      "loss": 0.3649,
+      "step": 30850
+    },
+    {
+      "epoch": 11.64967912419781,
+      "grad_norm": 0.6437296867370605,
+      "learning_rate": 1.361104712424503e-05,
+      "loss": 0.364,
+      "step": 30860
+    },
+    {
+      "epoch": 11.653454133635334,
+      "grad_norm": 0.6563718318939209,
+      "learning_rate": 1.358271034090201e-05,
+      "loss": 0.3597,
+      "step": 30870
+    },
+    {
+      "epoch": 11.657229143072858,
+      "grad_norm": 0.6579415798187256,
+      "learning_rate": 1.3554398448433076e-05,
+      "loss": 0.3694,
+      "step": 30880
+    },
+    {
+      "epoch": 11.661004152510381,
+      "grad_norm": 0.7728137373924255,
+      "learning_rate": 1.3526111466189156e-05,
+      "loss": 0.3608,
+      "step": 30890
+    },
+    {
+      "epoch": 11.664779161947905,
+      "grad_norm": 0.649925708770752,
+      "learning_rate": 1.3497849413504121e-05,
+      "loss": 0.3571,
+      "step": 30900
+    },
+    {
+      "epoch": 11.668554171385429,
+      "grad_norm": 0.6554239392280579,
+      "learning_rate": 1.3469612309694802e-05,
+      "loss": 0.3653,
+      "step": 30910
+    },
+    {
+      "epoch": 11.672329180822953,
+      "grad_norm": 0.6548380851745605,
+      "learning_rate": 1.3441400174061008e-05,
+      "loss": 0.367,
+      "step": 30920
+    },
+    {
+      "epoch": 11.676104190260476,
+      "grad_norm": 0.7009982466697693,
+      "learning_rate": 1.3413213025885469e-05,
+      "loss": 0.362,
+      "step": 30930
+    },
+    {
+      "epoch": 11.679879199698,
+      "grad_norm": 0.6359342336654663,
+      "learning_rate": 1.3385050884433853e-05,
+      "loss": 0.3625,
+      "step": 30940
+    },
+    {
+      "epoch": 11.683654209135522,
+      "grad_norm": 0.6045933365821838,
+      "learning_rate": 1.3356913768954676e-05,
+      "loss": 0.3534,
+      "step": 30950
+    },
+    {
+      "epoch": 11.687429218573046,
+      "grad_norm": 0.6957305669784546,
+      "learning_rate": 1.3328801698679415e-05,
+      "loss": 0.3594,
+      "step": 30960
+    },
+    {
+      "epoch": 11.69120422801057,
+      "grad_norm": 1.19200599193573,
+      "learning_rate": 1.3300714692822413e-05,
+      "loss": 0.3681,
+      "step": 30970
+    },
+    {
+      "epoch": 11.694979237448093,
+      "grad_norm": 0.6715503931045532,
+      "learning_rate": 1.3272652770580857e-05,
+      "loss": 0.3561,
+      "step": 30980
+    },
+    {
+      "epoch": 11.698754246885617,
+      "grad_norm": 0.679587721824646,
+      "learning_rate": 1.3244615951134831e-05,
+      "loss": 0.3577,
+      "step": 30990
+    },
+    {
+      "epoch": 11.70252925632314,
+      "grad_norm": 0.624066948890686,
+      "learning_rate": 1.3216604253647219e-05,
+      "loss": 0.3624,
+      "step": 31000
+    },
+    {
+      "epoch": 11.706304265760664,
+      "grad_norm": 0.7792338728904724,
+      "learning_rate": 1.3188617697263738e-05,
+      "loss": 0.3621,
+      "step": 31010
+    },
+    {
+      "epoch": 11.710079275198188,
+      "grad_norm": 0.6704654693603516,
+      "learning_rate": 1.3160656301112956e-05,
+      "loss": 0.3506,
+      "step": 31020
+    },
+    {
+      "epoch": 11.713854284635712,
+      "grad_norm": 0.6014306545257568,
+      "learning_rate": 1.3132720084306211e-05,
+      "loss": 0.3666,
+      "step": 31030
+    },
+    {
+      "epoch": 11.717629294073236,
+      "grad_norm": 0.6405402421951294,
+      "learning_rate": 1.3104809065937668e-05,
+      "loss": 0.3574,
+      "step": 31040
+    },
+    {
+      "epoch": 11.72140430351076,
+      "grad_norm": 0.6713264584541321,
+      "learning_rate": 1.3076923265084212e-05,
+      "loss": 0.3619,
+      "step": 31050
+    },
+    {
+      "epoch": 11.725179312948283,
+      "grad_norm": 0.6099828481674194,
+      "learning_rate": 1.3049062700805543e-05,
+      "loss": 0.3555,
+      "step": 31060
+    },
+    {
+      "epoch": 11.728954322385807,
+      "grad_norm": 0.6921255588531494,
+      "learning_rate": 1.3021227392144092e-05,
+      "loss": 0.3596,
+      "step": 31070
+    },
+    {
+      "epoch": 11.732729331823329,
+      "grad_norm": 0.6106016635894775,
+      "learning_rate": 1.2993417358125037e-05,
+      "loss": 0.3541,
+      "step": 31080
+    },
+    {
+      "epoch": 11.736504341260853,
+      "grad_norm": 0.5807941555976868,
+      "learning_rate": 1.2965632617756257e-05,
+      "loss": 0.3613,
+      "step": 31090
+    },
+    {
+      "epoch": 11.740279350698376,
+      "grad_norm": 0.5835058093070984,
+      "learning_rate": 1.2937873190028387e-05,
+      "loss": 0.3577,
+      "step": 31100
+    },
+    {
+      "epoch": 11.7440543601359,
+      "grad_norm": 0.5961214303970337,
+      "learning_rate": 1.2910139093914697e-05,
+      "loss": 0.3567,
+      "step": 31110
+    },
+    {
+      "epoch": 11.747829369573424,
+      "grad_norm": 0.696262538433075,
+      "learning_rate": 1.2882430348371205e-05,
+      "loss": 0.3555,
+      "step": 31120
+    },
+    {
+      "epoch": 11.751604379010947,
+      "grad_norm": 0.6330527663230896,
+      "learning_rate": 1.2854746972336567e-05,
+      "loss": 0.3644,
+      "step": 31130
+    },
+    {
+      "epoch": 11.755379388448471,
+      "grad_norm": 0.6530078649520874,
+      "learning_rate": 1.2827088984732132e-05,
+      "loss": 0.3728,
+      "step": 31140
+    },
+    {
+      "epoch": 11.759154397885995,
+      "grad_norm": 0.596555769443512,
+      "learning_rate": 1.279945640446184e-05,
+      "loss": 0.3646,
+      "step": 31150
+    },
+    {
+      "epoch": 11.762929407323519,
+      "grad_norm": 0.6525673270225525,
+      "learning_rate": 1.2771849250412315e-05,
+      "loss": 0.3603,
+      "step": 31160
+    },
+    {
+      "epoch": 11.766704416761042,
+      "grad_norm": 0.716068685054779,
+      "learning_rate": 1.2744267541452792e-05,
+      "loss": 0.3579,
+      "step": 31170
+    },
+    {
+      "epoch": 11.770479426198566,
+      "grad_norm": 0.5986026525497437,
+      "learning_rate": 1.2716711296435112e-05,
+      "loss": 0.3612,
+      "step": 31180
+    },
+    {
+      "epoch": 11.77425443563609,
+      "grad_norm": 0.6966813206672668,
+      "learning_rate": 1.268918053419369e-05,
+      "loss": 0.3645,
+      "step": 31190
+    },
+    {
+      "epoch": 11.778029445073614,
+      "grad_norm": 0.6633387207984924,
+      "learning_rate": 1.2661675273545564e-05,
+      "loss": 0.3607,
+      "step": 31200
+    },
+    {
+      "epoch": 11.781804454511136,
+      "grad_norm": 0.7241559028625488,
+      "learning_rate": 1.26341955332903e-05,
+      "loss": 0.3615,
+      "step": 31210
+    },
+    {
+      "epoch": 11.78557946394866,
+      "grad_norm": 0.6759303212165833,
+      "learning_rate": 1.2606741332210048e-05,
+      "loss": 0.3611,
+      "step": 31220
+    },
+    {
+      "epoch": 11.789354473386183,
+      "grad_norm": 0.642245888710022,
+      "learning_rate": 1.2579312689069495e-05,
+      "loss": 0.3591,
+      "step": 31230
+    },
+    {
+      "epoch": 11.793129482823707,
+      "grad_norm": 0.6876919865608215,
+      "learning_rate": 1.2551909622615881e-05,
+      "loss": 0.3537,
+      "step": 31240
+    },
+    {
+      "epoch": 11.79690449226123,
+      "grad_norm": 0.6188132762908936,
+      "learning_rate": 1.252453215157891e-05,
+      "loss": 0.3683,
+      "step": 31250
+    },
+    {
+      "epoch": 11.800679501698754,
+      "grad_norm": 0.6617509722709656,
+      "learning_rate": 1.2497180294670841e-05,
+      "loss": 0.3575,
+      "step": 31260
+    },
+    {
+      "epoch": 11.804454511136278,
+      "grad_norm": 0.6628468036651611,
+      "learning_rate": 1.2469854070586423e-05,
+      "loss": 0.3568,
+      "step": 31270
+    },
+    {
+      "epoch": 11.808229520573802,
+      "grad_norm": 0.6406071186065674,
+      "learning_rate": 1.2442553498002845e-05,
+      "loss": 0.3705,
+      "step": 31280
+    },
+    {
+      "epoch": 11.812004530011325,
+      "grad_norm": 0.704318106174469,
+      "learning_rate": 1.241527859557981e-05,
+      "loss": 0.3602,
+      "step": 31290
+    },
+    {
+      "epoch": 11.81577953944885,
+      "grad_norm": 0.6852730512619019,
+      "learning_rate": 1.238802938195946e-05,
+      "loss": 0.3549,
+      "step": 31300
+    },
+    {
+      "epoch": 11.819554548886373,
+      "grad_norm": 0.7002149820327759,
+      "learning_rate": 1.2360805875766357e-05,
+      "loss": 0.3582,
+      "step": 31310
+    },
+    {
+      "epoch": 11.823329558323895,
+      "grad_norm": 0.6709305047988892,
+      "learning_rate": 1.2333608095607524e-05,
+      "loss": 0.3598,
+      "step": 31320
+    },
+    {
+      "epoch": 11.827104567761419,
+      "grad_norm": 0.6867847442626953,
+      "learning_rate": 1.230643606007238e-05,
+      "loss": 0.3582,
+      "step": 31330
+    },
+    {
+      "epoch": 11.830879577198942,
+      "grad_norm": 0.7019827961921692,
+      "learning_rate": 1.2279289787732778e-05,
+      "loss": 0.3513,
+      "step": 31340
+    },
+    {
+      "epoch": 11.834654586636466,
+      "grad_norm": 0.5999962687492371,
+      "learning_rate": 1.2252169297142906e-05,
+      "loss": 0.3575,
+      "step": 31350
+    },
+    {
+      "epoch": 11.83842959607399,
+      "grad_norm": 0.6360159516334534,
+      "learning_rate": 1.2225074606839392e-05,
+      "loss": 0.3611,
+      "step": 31360
+    },
+    {
+      "epoch": 11.842204605511514,
+      "grad_norm": 0.5966977477073669,
+      "learning_rate": 1.2198005735341173e-05,
+      "loss": 0.364,
+      "step": 31370
+    },
+    {
+      "epoch": 11.845979614949037,
+      "grad_norm": 0.6092012524604797,
+      "learning_rate": 1.2170962701149584e-05,
+      "loss": 0.3562,
+      "step": 31380
+    },
+    {
+      "epoch": 11.849754624386561,
+      "grad_norm": 0.7129442095756531,
+      "learning_rate": 1.2143945522748285e-05,
+      "loss": 0.3621,
+      "step": 31390
+    },
+    {
+      "epoch": 11.853529633824085,
+      "grad_norm": 0.6315842270851135,
+      "learning_rate": 1.2116954218603266e-05,
+      "loss": 0.3636,
+      "step": 31400
+    },
+    {
+      "epoch": 11.857304643261608,
+      "grad_norm": 0.7170038223266602,
+      "learning_rate": 1.2089988807162817e-05,
+      "loss": 0.3601,
+      "step": 31410
+    },
+    {
+      "epoch": 11.861079652699132,
+      "grad_norm": 0.5954352021217346,
+      "learning_rate": 1.2063049306857543e-05,
+      "loss": 0.3585,
+      "step": 31420
+    },
+    {
+      "epoch": 11.864854662136656,
+      "grad_norm": 0.6874489784240723,
+      "learning_rate": 1.2036135736100346e-05,
+      "loss": 0.3589,
+      "step": 31430
+    },
+    {
+      "epoch": 11.86862967157418,
+      "grad_norm": 0.6595900058746338,
+      "learning_rate": 1.200924811328641e-05,
+      "loss": 0.3526,
+      "step": 31440
+    },
+    {
+      "epoch": 11.872404681011702,
+      "grad_norm": 0.6593577861785889,
+      "learning_rate": 1.1982386456793143e-05,
+      "loss": 0.3656,
+      "step": 31450
+    },
+    {
+      "epoch": 11.876179690449225,
+      "grad_norm": 0.6922234296798706,
+      "learning_rate": 1.1955550784980257e-05,
+      "loss": 0.3593,
+      "step": 31460
+    },
+    {
+      "epoch": 11.87995469988675,
+      "grad_norm": 0.6410038471221924,
+      "learning_rate": 1.1928741116189663e-05,
+      "loss": 0.3544,
+      "step": 31470
+    },
+    {
+      "epoch": 11.883729709324273,
+      "grad_norm": 0.6991416811943054,
+      "learning_rate": 1.1901957468745529e-05,
+      "loss": 0.3648,
+      "step": 31480
+    },
+    {
+      "epoch": 11.887504718761797,
+      "grad_norm": 0.6780281662940979,
+      "learning_rate": 1.1875199860954217e-05,
+      "loss": 0.3662,
+      "step": 31490
+    },
+    {
+      "epoch": 11.89127972819932,
+      "grad_norm": 0.8736029863357544,
+      "learning_rate": 1.1848468311104316e-05,
+      "loss": 0.3568,
+      "step": 31500
+    },
+    {
+      "epoch": 11.895054737636844,
+      "grad_norm": 0.5960100889205933,
+      "learning_rate": 1.1821762837466566e-05,
+      "loss": 0.3573,
+      "step": 31510
+    },
+    {
+      "epoch": 11.898829747074368,
+      "grad_norm": 0.6374369263648987,
+      "learning_rate": 1.179508345829392e-05,
+      "loss": 0.3708,
+      "step": 31520
+    },
+    {
+      "epoch": 11.902604756511892,
+      "grad_norm": 0.7211053967475891,
+      "learning_rate": 1.1768430191821478e-05,
+      "loss": 0.3584,
+      "step": 31530
+    },
+    {
+      "epoch": 11.906379765949415,
+      "grad_norm": 0.8091753721237183,
+      "learning_rate": 1.174180305626651e-05,
+      "loss": 0.3602,
+      "step": 31540
+    },
+    {
+      "epoch": 11.910154775386939,
+      "grad_norm": 0.6501530408859253,
+      "learning_rate": 1.1715202069828401e-05,
+      "loss": 0.3593,
+      "step": 31550
+    },
+    {
+      "epoch": 11.913929784824463,
+      "grad_norm": 0.7378227710723877,
+      "learning_rate": 1.1688627250688661e-05,
+      "loss": 0.3638,
+      "step": 31560
+    },
+    {
+      "epoch": 11.917704794261986,
+      "grad_norm": 0.6729764342308044,
+      "learning_rate": 1.1662078617010935e-05,
+      "loss": 0.3598,
+      "step": 31570
+    },
+    {
+      "epoch": 11.921479803699508,
+      "grad_norm": 0.6561091542243958,
+      "learning_rate": 1.1635556186940972e-05,
+      "loss": 0.3649,
+      "step": 31580
+    },
+    {
+      "epoch": 11.925254813137032,
+      "grad_norm": 0.6765795946121216,
+      "learning_rate": 1.1609059978606602e-05,
+      "loss": 0.3709,
+      "step": 31590
+    },
+    {
+      "epoch": 11.929029822574556,
+      "grad_norm": 0.641923189163208,
+      "learning_rate": 1.1582590010117745e-05,
+      "loss": 0.3595,
+      "step": 31600
+    },
+    {
+      "epoch": 11.93280483201208,
+      "grad_norm": 0.6592920422554016,
+      "learning_rate": 1.1556146299566345e-05,
+      "loss": 0.3543,
+      "step": 31610
+    },
+    {
+      "epoch": 11.936579841449603,
+      "grad_norm": 0.6286612153053284,
+      "learning_rate": 1.1529728865026452e-05,
+      "loss": 0.3685,
+      "step": 31620
+    },
+    {
+      "epoch": 11.940354850887127,
+      "grad_norm": 1.297844648361206,
+      "learning_rate": 1.150333772455412e-05,
+      "loss": 0.3626,
+      "step": 31630
+    },
+    {
+      "epoch": 11.94412986032465,
+      "grad_norm": 0.6182207465171814,
+      "learning_rate": 1.1476972896187466e-05,
+      "loss": 0.355,
+      "step": 31640
+    },
+    {
+      "epoch": 11.947904869762175,
+      "grad_norm": 0.7070791125297546,
+      "learning_rate": 1.1450634397946592e-05,
+      "loss": 0.359,
+      "step": 31650
+    },
+    {
+      "epoch": 11.951679879199698,
+      "grad_norm": 0.6540588140487671,
+      "learning_rate": 1.142432224783359e-05,
+      "loss": 0.358,
+      "step": 31660
+    },
+    {
+      "epoch": 11.955454888637222,
+      "grad_norm": 0.6664989590644836,
+      "learning_rate": 1.1398036463832584e-05,
+      "loss": 0.3555,
+      "step": 31670
+    },
+    {
+      "epoch": 11.959229898074746,
+      "grad_norm": 0.6335260272026062,
+      "learning_rate": 1.1371777063909661e-05,
+      "loss": 0.3611,
+      "step": 31680
+    },
+    {
+      "epoch": 11.96300490751227,
+      "grad_norm": 0.614297091960907,
+      "learning_rate": 1.1345544066012874e-05,
+      "loss": 0.3628,
+      "step": 31690
+    },
+    {
+      "epoch": 11.966779916949793,
+      "grad_norm": 0.746405839920044,
+      "learning_rate": 1.131933748807224e-05,
+      "loss": 0.3556,
+      "step": 31700
+    },
+    {
+      "epoch": 11.970554926387315,
+      "grad_norm": 0.6299577951431274,
+      "learning_rate": 1.1293157347999683e-05,
+      "loss": 0.3573,
+      "step": 31710
+    },
+    {
+      "epoch": 11.974329935824839,
+      "grad_norm": 0.643743634223938,
+      "learning_rate": 1.1267003663689097e-05,
+      "loss": 0.3597,
+      "step": 31720
+    },
+    {
+      "epoch": 11.978104945262363,
+      "grad_norm": 0.7774217128753662,
+      "learning_rate": 1.1240876453016275e-05,
+      "loss": 0.3816,
+      "step": 31730
+    },
+    {
+      "epoch": 11.981879954699886,
+      "grad_norm": 0.6441808342933655,
+      "learning_rate": 1.1214775733838933e-05,
+      "loss": 0.3589,
+      "step": 31740
+    },
+    {
+      "epoch": 11.98565496413741,
+      "grad_norm": 0.6050808429718018,
+      "learning_rate": 1.118870152399666e-05,
+      "loss": 0.3601,
+      "step": 31750
+    },
+    {
+      "epoch": 11.989429973574934,
+      "grad_norm": 0.6663985252380371,
+      "learning_rate": 1.1162653841310905e-05,
+      "loss": 0.3602,
+      "step": 31760
+    },
+    {
+      "epoch": 11.993204983012458,
+      "grad_norm": 0.6749792695045471,
+      "learning_rate": 1.1136632703585043e-05,
+      "loss": 0.3576,
+      "step": 31770
+    },
+    {
+      "epoch": 11.996979992449981,
+      "grad_norm": 0.731533408164978,
+      "learning_rate": 1.111063812860426e-05,
+      "loss": 0.3565,
+      "step": 31780
+    },
+    {
+      "epoch": 12.000755001887505,
+      "grad_norm": 0.6132811903953552,
+      "learning_rate": 1.1084670134135616e-05,
+      "loss": 0.3547,
+      "step": 31790
+    },
+    {
+      "epoch": 12.004530011325029,
+      "grad_norm": 0.6406649947166443,
+      "learning_rate": 1.1058728737927998e-05,
+      "loss": 0.3593,
+      "step": 31800
+    },
+    {
+      "epoch": 12.008305020762553,
+      "grad_norm": 0.6679116487503052,
+      "learning_rate": 1.1032813957712086e-05,
+      "loss": 0.3594,
+      "step": 31810
+    },
+    {
+      "epoch": 12.012080030200076,
+      "grad_norm": 0.6831429600715637,
+      "learning_rate": 1.1006925811200385e-05,
+      "loss": 0.355,
+      "step": 31820
+    },
+    {
+      "epoch": 12.015855039637598,
+      "grad_norm": 0.6503385305404663,
+      "learning_rate": 1.0981064316087231e-05,
+      "loss": 0.3616,
+      "step": 31830
+    },
+    {
+      "epoch": 12.019630049075122,
+      "grad_norm": 0.6955201625823975,
+      "learning_rate": 1.0955229490048675e-05,
+      "loss": 0.3606,
+      "step": 31840
+    },
+    {
+      "epoch": 12.023405058512646,
+      "grad_norm": 0.6384215950965881,
+      "learning_rate": 1.0929421350742603e-05,
+      "loss": 0.3632,
+      "step": 31850
+    },
+    {
+      "epoch": 12.02718006795017,
+      "grad_norm": 0.6311360001564026,
+      "learning_rate": 1.090363991580861e-05,
+      "loss": 0.3628,
+      "step": 31860
+    },
+    {
+      "epoch": 12.030955077387693,
+      "grad_norm": 0.8894734382629395,
+      "learning_rate": 1.0877885202868083e-05,
+      "loss": 0.3631,
+      "step": 31870
+    },
+    {
+      "epoch": 12.034730086825217,
+      "grad_norm": 0.6360461115837097,
+      "learning_rate": 1.0852157229524118e-05,
+      "loss": 0.3779,
+      "step": 31880
+    },
+    {
+      "epoch": 12.03850509626274,
+      "grad_norm": 0.7593185901641846,
+      "learning_rate": 1.0826456013361542e-05,
+      "loss": 0.3516,
+      "step": 31890
+    },
+    {
+      "epoch": 12.042280105700264,
+      "grad_norm": 0.681585431098938,
+      "learning_rate": 1.080078157194691e-05,
+      "loss": 0.3682,
+      "step": 31900
+    },
+    {
+      "epoch": 12.046055115137788,
+      "grad_norm": 0.7083263397216797,
+      "learning_rate": 1.077513392282844e-05,
+      "loss": 0.3588,
+      "step": 31910
+    },
+    {
+      "epoch": 12.049830124575312,
+      "grad_norm": 0.7412534952163696,
+      "learning_rate": 1.0749513083536073e-05,
+      "loss": 0.3593,
+      "step": 31920
+    },
+    {
+      "epoch": 12.053605134012836,
+      "grad_norm": 0.6712055802345276,
+      "learning_rate": 1.072391907158139e-05,
+      "loss": 0.3601,
+      "step": 31930
+    },
+    {
+      "epoch": 12.05738014345036,
+      "grad_norm": 0.6988897323608398,
+      "learning_rate": 1.0698351904457676e-05,
+      "loss": 0.3553,
+      "step": 31940
+    },
+    {
+      "epoch": 12.061155152887883,
+      "grad_norm": 0.7310452461242676,
+      "learning_rate": 1.0672811599639849e-05,
+      "loss": 0.3612,
+      "step": 31950
+    },
+    {
+      "epoch": 12.064930162325405,
+      "grad_norm": 0.6422081589698792,
+      "learning_rate": 1.0647298174584453e-05,
+      "loss": 0.3648,
+      "step": 31960
+    },
+    {
+      "epoch": 12.068705171762929,
+      "grad_norm": 0.6467820405960083,
+      "learning_rate": 1.0621811646729673e-05,
+      "loss": 0.3573,
+      "step": 31970
+    },
+    {
+      "epoch": 12.072480181200453,
+      "grad_norm": 0.5941371917724609,
+      "learning_rate": 1.0596352033495328e-05,
+      "loss": 0.3646,
+      "step": 31980
+    },
+    {
+      "epoch": 12.076255190637976,
+      "grad_norm": 0.6265065670013428,
+      "learning_rate": 1.057091935228281e-05,
+      "loss": 0.3674,
+      "step": 31990
+    },
+    {
+      "epoch": 12.0800302000755,
+      "grad_norm": 0.7404178380966187,
+      "learning_rate": 1.0545513620475134e-05,
+      "loss": 0.3639,
+      "step": 32000
+    },
+    {
+      "epoch": 12.083805209513024,
+      "grad_norm": 0.6679591536521912,
+      "learning_rate": 1.0520134855436863e-05,
+      "loss": 0.3705,
+      "step": 32010
+    },
+    {
+      "epoch": 12.087580218950547,
+      "grad_norm": 0.6543688774108887,
+      "learning_rate": 1.049478307451413e-05,
+      "loss": 0.3586,
+      "step": 32020
+    },
+    {
+      "epoch": 12.091355228388071,
+      "grad_norm": 0.6841951012611389,
+      "learning_rate": 1.0469458295034656e-05,
+      "loss": 0.3585,
+      "step": 32030
+    },
+    {
+      "epoch": 12.095130237825595,
+      "grad_norm": 0.7727581858634949,
+      "learning_rate": 1.044416053430769e-05,
+      "loss": 0.3591,
+      "step": 32040
+    },
+    {
+      "epoch": 12.098905247263119,
+      "grad_norm": 0.6680077910423279,
+      "learning_rate": 1.0418889809624021e-05,
+      "loss": 0.3614,
+      "step": 32050
+    },
+    {
+      "epoch": 12.102680256700642,
+      "grad_norm": 0.5810276865959167,
+      "learning_rate": 1.0393646138255925e-05,
+      "loss": 0.3641,
+      "step": 32060
+    },
+    {
+      "epoch": 12.106455266138166,
+      "grad_norm": 0.6576191782951355,
+      "learning_rate": 1.036842953745723e-05,
+      "loss": 0.3583,
+      "step": 32070
+    },
+    {
+      "epoch": 12.110230275575688,
+      "grad_norm": 0.6342170834541321,
+      "learning_rate": 1.034324002446324e-05,
+      "loss": 0.355,
+      "step": 32080
+    },
+    {
+      "epoch": 12.114005285013212,
+      "grad_norm": 0.6298282146453857,
+      "learning_rate": 1.0318077616490751e-05,
+      "loss": 0.3631,
+      "step": 32090
+    },
+    {
+      "epoch": 12.117780294450736,
+      "grad_norm": 0.6232679486274719,
+      "learning_rate": 1.0292942330738054e-05,
+      "loss": 0.3566,
+      "step": 32100
+    },
+    {
+      "epoch": 12.12155530388826,
+      "grad_norm": 0.663240373134613,
+      "learning_rate": 1.026783418438485e-05,
+      "loss": 0.3568,
+      "step": 32110
+    },
+    {
+      "epoch": 12.125330313325783,
+      "grad_norm": 0.6801791191101074,
+      "learning_rate": 1.0242753194592319e-05,
+      "loss": 0.3593,
+      "step": 32120
+    },
+    {
+      "epoch": 12.129105322763307,
+      "grad_norm": 0.6788033246994019,
+      "learning_rate": 1.0217699378503092e-05,
+      "loss": 0.3492,
+      "step": 32130
+    },
+    {
+      "epoch": 12.13288033220083,
+      "grad_norm": 0.6879256963729858,
+      "learning_rate": 1.0192672753241218e-05,
+      "loss": 0.3575,
+      "step": 32140
+    },
+    {
+      "epoch": 12.136655341638354,
+      "grad_norm": 0.623914897441864,
+      "learning_rate": 1.0167673335912164e-05,
+      "loss": 0.3612,
+      "step": 32150
+    },
+    {
+      "epoch": 12.140430351075878,
+      "grad_norm": 0.6156293749809265,
+      "learning_rate": 1.0142701143602779e-05,
+      "loss": 0.3612,
+      "step": 32160
+    },
+    {
+      "epoch": 12.144205360513402,
+      "grad_norm": 0.6880207657814026,
+      "learning_rate": 1.0117756193381328e-05,
+      "loss": 0.3544,
+      "step": 32170
+    },
+    {
+      "epoch": 12.147980369950925,
+      "grad_norm": 0.6038033962249756,
+      "learning_rate": 1.0092838502297453e-05,
+      "loss": 0.3633,
+      "step": 32180
+    },
+    {
+      "epoch": 12.15175537938845,
+      "grad_norm": 0.6807112693786621,
+      "learning_rate": 1.0067948087382162e-05,
+      "loss": 0.3554,
+      "step": 32190
+    },
+    {
+      "epoch": 12.155530388825973,
+      "grad_norm": 0.6536206007003784,
+      "learning_rate": 1.0043084965647826e-05,
+      "loss": 0.3642,
+      "step": 32200
+    },
+    {
+      "epoch": 12.159305398263495,
+      "grad_norm": 0.685413122177124,
+      "learning_rate": 1.001824915408815e-05,
+      "loss": 0.4406,
+      "step": 32210
+    },
+    {
+      "epoch": 12.163080407701019,
+      "grad_norm": 0.6319077014923096,
+      "learning_rate": 9.99344066967815e-06,
+      "loss": 0.3569,
+      "step": 32220
+    },
+    {
+      "epoch": 12.166855417138542,
+      "grad_norm": 0.6181957125663757,
+      "learning_rate": 9.968659529374219e-06,
+      "loss": 0.3557,
+      "step": 32230
+    },
+    {
+      "epoch": 12.170630426576066,
+      "grad_norm": 0.6865522861480713,
+      "learning_rate": 9.943905750114025e-06,
+      "loss": 0.358,
+      "step": 32240
+    },
+    {
+      "epoch": 12.17440543601359,
+      "grad_norm": 0.7125778794288635,
+      "learning_rate": 9.919179348816559e-06,
+      "loss": 0.3643,
+      "step": 32250
+    },
+    {
+      "epoch": 12.178180445451114,
+      "grad_norm": 0.654378354549408,
+      "learning_rate": 9.894480342382056e-06,
+      "loss": 0.3583,
+      "step": 32260
+    },
+    {
+      "epoch": 12.181955454888637,
+      "grad_norm": 0.6028674244880676,
+      "learning_rate": 9.869808747692061e-06,
+      "loss": 0.3586,
+      "step": 32270
+    },
+    {
+      "epoch": 12.185730464326161,
+      "grad_norm": 0.6447907090187073,
+      "learning_rate": 9.84516458160939e-06,
+      "loss": 0.3629,
+      "step": 32280
+    },
+    {
+      "epoch": 12.189505473763685,
+      "grad_norm": 0.687196671962738,
+      "learning_rate": 9.820547860978097e-06,
+      "loss": 0.3575,
+      "step": 32290
+    },
+    {
+      "epoch": 12.193280483201208,
+      "grad_norm": 0.6808565258979797,
+      "learning_rate": 9.795958602623461e-06,
+      "loss": 0.3607,
+      "step": 32300
+    },
+    {
+      "epoch": 12.197055492638732,
+      "grad_norm": 0.6441514492034912,
+      "learning_rate": 9.77139682335203e-06,
+      "loss": 0.3544,
+      "step": 32310
+    },
+    {
+      "epoch": 12.200830502076256,
+      "grad_norm": 0.5570544600486755,
+      "learning_rate": 9.74686253995153e-06,
+      "loss": 0.3567,
+      "step": 32320
+    },
+    {
+      "epoch": 12.204605511513778,
+      "grad_norm": 0.639428436756134,
+      "learning_rate": 9.722355769190911e-06,
+      "loss": 0.3592,
+      "step": 32330
+    },
+    {
+      "epoch": 12.208380520951302,
+      "grad_norm": 0.636038064956665,
+      "learning_rate": 9.697876527820333e-06,
+      "loss": 0.3634,
+      "step": 32340
+    },
+    {
+      "epoch": 12.212155530388825,
+      "grad_norm": 0.6335992813110352,
+      "learning_rate": 9.673424832571132e-06,
+      "loss": 0.3599,
+      "step": 32350
+    },
+    {
+      "epoch": 12.21593053982635,
+      "grad_norm": 0.6807220578193665,
+      "learning_rate": 9.649000700155791e-06,
+      "loss": 0.3591,
+      "step": 32360
+    },
+    {
+      "epoch": 12.219705549263873,
+      "grad_norm": 0.7252523303031921,
+      "learning_rate": 9.624604147267984e-06,
+      "loss": 0.3608,
+      "step": 32370
+    },
+    {
+      "epoch": 12.223480558701397,
+      "grad_norm": 0.6216152906417847,
+      "learning_rate": 9.600235190582524e-06,
+      "loss": 0.358,
+      "step": 32380
+    },
+    {
+      "epoch": 12.22725556813892,
+      "grad_norm": 0.6576191186904907,
+      "learning_rate": 9.57589384675538e-06,
+      "loss": 0.3652,
+      "step": 32390
+    },
+    {
+      "epoch": 12.231030577576444,
+      "grad_norm": 0.626547634601593,
+      "learning_rate": 9.551580132423598e-06,
+      "loss": 0.3683,
+      "step": 32400
+    },
+    {
+      "epoch": 12.234805587013968,
+      "grad_norm": 0.6060603857040405,
+      "learning_rate": 9.527294064205405e-06,
+      "loss": 0.363,
+      "step": 32410
+    },
+    {
+      "epoch": 12.238580596451492,
+      "grad_norm": 0.6533463001251221,
+      "learning_rate": 9.503035658700072e-06,
+      "loss": 0.3593,
+      "step": 32420
+    },
+    {
+      "epoch": 12.242355605889015,
+      "grad_norm": 0.6849529147148132,
+      "learning_rate": 9.478804932488e-06,
+      "loss": 0.3578,
+      "step": 32430
+    },
+    {
+      "epoch": 12.246130615326539,
+      "grad_norm": 0.6339350938796997,
+      "learning_rate": 9.454601902130667e-06,
+      "loss": 0.3487,
+      "step": 32440
+    },
+    {
+      "epoch": 12.249905624764063,
+      "grad_norm": 0.7915511727333069,
+      "learning_rate": 9.430426584170626e-06,
+      "loss": 0.3567,
+      "step": 32450
+    },
+    {
+      "epoch": 12.253680634201585,
+      "grad_norm": 0.6535844206809998,
+      "learning_rate": 9.40627899513145e-06,
+      "loss": 0.3514,
+      "step": 32460
+    },
+    {
+      "epoch": 12.257455643639108,
+      "grad_norm": 0.6630949378013611,
+      "learning_rate": 9.38215915151781e-06,
+      "loss": 0.3596,
+      "step": 32470
+    },
+    {
+      "epoch": 12.261230653076632,
+      "grad_norm": 0.9704565405845642,
+      "learning_rate": 9.358067069815396e-06,
+      "loss": 0.3575,
+      "step": 32480
+    },
+    {
+      "epoch": 12.265005662514156,
+      "grad_norm": 0.6713346838951111,
+      "learning_rate": 9.334002766490901e-06,
+      "loss": 0.3604,
+      "step": 32490
+    },
+    {
+      "epoch": 12.26878067195168,
+      "grad_norm": 0.6596808433532715,
+      "learning_rate": 9.309966257992054e-06,
+      "loss": 0.3557,
+      "step": 32500
+    },
+    {
+      "epoch": 12.272555681389203,
+      "grad_norm": 0.6223334670066833,
+      "learning_rate": 9.2859575607476e-06,
+      "loss": 0.3603,
+      "step": 32510
+    },
+    {
+      "epoch": 12.276330690826727,
+      "grad_norm": 0.638768196105957,
+      "learning_rate": 9.26197669116723e-06,
+      "loss": 0.3637,
+      "step": 32520
+    },
+    {
+      "epoch": 12.28010570026425,
+      "grad_norm": 0.6910896897315979,
+      "learning_rate": 9.238023665641655e-06,
+      "loss": 0.352,
+      "step": 32530
+    },
+    {
+      "epoch": 12.283880709701775,
+      "grad_norm": 0.7031621932983398,
+      "learning_rate": 9.21409850054254e-06,
+      "loss": 0.3612,
+      "step": 32540
+    },
+    {
+      "epoch": 12.287655719139298,
+      "grad_norm": 0.6571682095527649,
+      "learning_rate": 9.190201212222526e-06,
+      "loss": 0.3605,
+      "step": 32550
+    },
+    {
+      "epoch": 12.291430728576822,
+      "grad_norm": 0.670776903629303,
+      "learning_rate": 9.166331817015155e-06,
+      "loss": 0.3516,
+      "step": 32560
+    },
+    {
+      "epoch": 12.295205738014346,
+      "grad_norm": 0.5901675820350647,
+      "learning_rate": 9.142490331234965e-06,
+      "loss": 0.358,
+      "step": 32570
+    },
+    {
+      "epoch": 12.298980747451868,
+      "grad_norm": 0.7036232948303223,
+      "learning_rate": 9.11867677117736e-06,
+      "loss": 0.3604,
+      "step": 32580
+    },
+    {
+      "epoch": 12.302755756889391,
+      "grad_norm": 0.6272095441818237,
+      "learning_rate": 9.09489115311869e-06,
+      "loss": 0.3574,
+      "step": 32590
+    },
+    {
+      "epoch": 12.306530766326915,
+      "grad_norm": 0.7180802226066589,
+      "learning_rate": 9.071133493316202e-06,
+      "loss": 0.3626,
+      "step": 32600
+    },
+    {
+      "epoch": 12.310305775764439,
+      "grad_norm": 0.6323314309120178,
+      "learning_rate": 9.047403808008053e-06,
+      "loss": 0.3604,
+      "step": 32610
+    },
+    {
+      "epoch": 12.314080785201963,
+      "grad_norm": 0.6044765114784241,
+      "learning_rate": 9.023702113413224e-06,
+      "loss": 0.3679,
+      "step": 32620
+    },
+    {
+      "epoch": 12.317855794639486,
+      "grad_norm": 0.7297821044921875,
+      "learning_rate": 9.00002842573162e-06,
+      "loss": 0.3595,
+      "step": 32630
+    },
+    {
+      "epoch": 12.32163080407701,
+      "grad_norm": 0.8278955221176147,
+      "learning_rate": 8.976382761143981e-06,
+      "loss": 0.3556,
+      "step": 32640
+    },
+    {
+      "epoch": 12.325405813514534,
+      "grad_norm": 0.6725118160247803,
+      "learning_rate": 8.952765135811909e-06,
+      "loss": 0.3635,
+      "step": 32650
+    },
+    {
+      "epoch": 12.329180822952058,
+      "grad_norm": 0.6396539807319641,
+      "learning_rate": 8.929175565877795e-06,
+      "loss": 0.3559,
+      "step": 32660
+    },
+    {
+      "epoch": 12.332955832389581,
+      "grad_norm": 1.543081521987915,
+      "learning_rate": 8.905614067464924e-06,
+      "loss": 0.3599,
+      "step": 32670
+    },
+    {
+      "epoch": 12.336730841827105,
+      "grad_norm": 0.6082138419151306,
+      "learning_rate": 8.882080656677333e-06,
+      "loss": 0.3559,
+      "step": 32680
+    },
+    {
+      "epoch": 12.340505851264629,
+      "grad_norm": 0.6150057911872864,
+      "learning_rate": 8.85857534959989e-06,
+      "loss": 0.3602,
+      "step": 32690
+    },
+    {
+      "epoch": 12.344280860702153,
+      "grad_norm": 0.8048917651176453,
+      "learning_rate": 8.835098162298256e-06,
+      "loss": 0.3605,
+      "step": 32700
+    },
+    {
+      "epoch": 12.348055870139675,
+      "grad_norm": 0.735430896282196,
+      "learning_rate": 8.811649110818875e-06,
+      "loss": 0.353,
+      "step": 32710
+    },
+    {
+      "epoch": 12.351830879577198,
+      "grad_norm": 0.6606729030609131,
+      "learning_rate": 8.78822821118893e-06,
+      "loss": 0.3521,
+      "step": 32720
+    },
+    {
+      "epoch": 12.355605889014722,
+      "grad_norm": 0.6922546029090881,
+      "learning_rate": 8.764835479416405e-06,
+      "loss": 0.359,
+      "step": 32730
+    },
+    {
+      "epoch": 12.359380898452246,
+      "grad_norm": 0.6572328805923462,
+      "learning_rate": 8.741470931490003e-06,
+      "loss": 0.3557,
+      "step": 32740
+    },
+    {
+      "epoch": 12.36315590788977,
+      "grad_norm": 0.6792584657669067,
+      "learning_rate": 8.718134583379189e-06,
+      "loss": 0.354,
+      "step": 32750
+    },
+    {
+      "epoch": 12.366930917327293,
+      "grad_norm": 0.6346094012260437,
+      "learning_rate": 8.69482645103412e-06,
+      "loss": 0.3549,
+      "step": 32760
+    },
+    {
+      "epoch": 12.370705926764817,
+      "grad_norm": 0.6417624950408936,
+      "learning_rate": 8.671546550385679e-06,
+      "loss": 0.3598,
+      "step": 32770
+    },
+    {
+      "epoch": 12.37448093620234,
+      "grad_norm": 0.7251724004745483,
+      "learning_rate": 8.648294897345466e-06,
+      "loss": 0.352,
+      "step": 32780
+    },
+    {
+      "epoch": 12.378255945639864,
+      "grad_norm": 0.628129243850708,
+      "learning_rate": 8.62507150780576e-06,
+      "loss": 0.3574,
+      "step": 32790
+    },
+    {
+      "epoch": 12.382030955077388,
+      "grad_norm": 0.6672044992446899,
+      "learning_rate": 8.601876397639542e-06,
+      "loss": 0.3662,
+      "step": 32800
+    },
+    {
+      "epoch": 12.385805964514912,
+      "grad_norm": 0.614966630935669,
+      "learning_rate": 8.578709582700456e-06,
+      "loss": 0.3499,
+      "step": 32810
+    },
+    {
+      "epoch": 12.389580973952436,
+      "grad_norm": 0.629789412021637,
+      "learning_rate": 8.555571078822783e-06,
+      "loss": 0.358,
+      "step": 32820
+    },
+    {
+      "epoch": 12.39335598338996,
+      "grad_norm": 0.6185657978057861,
+      "learning_rate": 8.532460901821477e-06,
+      "loss": 0.3626,
+      "step": 32830
+    },
+    {
+      "epoch": 12.397130992827481,
+      "grad_norm": 0.6272408366203308,
+      "learning_rate": 8.509379067492128e-06,
+      "loss": 0.3543,
+      "step": 32840
+    },
+    {
+      "epoch": 12.400906002265005,
+      "grad_norm": 0.7130616903305054,
+      "learning_rate": 8.486325591610967e-06,
+      "loss": 0.3542,
+      "step": 32850
+    },
+    {
+      "epoch": 12.404681011702529,
+      "grad_norm": 0.6526590585708618,
+      "learning_rate": 8.463300489934811e-06,
+      "loss": 0.3627,
+      "step": 32860
+    },
+    {
+      "epoch": 12.408456021140053,
+      "grad_norm": 0.6703090071678162,
+      "learning_rate": 8.440303778201093e-06,
+      "loss": 0.3571,
+      "step": 32870
+    },
+    {
+      "epoch": 12.412231030577576,
+      "grad_norm": 0.6281063556671143,
+      "learning_rate": 8.417335472127858e-06,
+      "loss": 0.3667,
+      "step": 32880
+    },
+    {
+      "epoch": 12.4160060400151,
+      "grad_norm": 0.6609539985656738,
+      "learning_rate": 8.394395587413722e-06,
+      "loss": 0.3607,
+      "step": 32890
+    },
+    {
+      "epoch": 12.419781049452624,
+      "grad_norm": 0.6523950099945068,
+      "learning_rate": 8.371484139737889e-06,
+      "loss": 0.3728,
+      "step": 32900
+    },
+    {
+      "epoch": 12.423556058890147,
+      "grad_norm": 0.6022361516952515,
+      "learning_rate": 8.348601144760127e-06,
+      "loss": 0.3586,
+      "step": 32910
+    },
+    {
+      "epoch": 12.427331068327671,
+      "grad_norm": 0.6702349185943604,
+      "learning_rate": 8.325746618120716e-06,
+      "loss": 0.3576,
+      "step": 32920
+    },
+    {
+      "epoch": 12.431106077765195,
+      "grad_norm": 0.6322970986366272,
+      "learning_rate": 8.302920575440531e-06,
+      "loss": 0.356,
+      "step": 32930
+    },
+    {
+      "epoch": 12.434881087202719,
+      "grad_norm": 0.6621061563491821,
+      "learning_rate": 8.280123032320953e-06,
+      "loss": 0.3598,
+      "step": 32940
+    },
+    {
+      "epoch": 12.438656096640242,
+      "grad_norm": 2.3764612674713135,
+      "learning_rate": 8.257354004343892e-06,
+      "loss": 0.3598,
+      "step": 32950
+    },
+    {
+      "epoch": 12.442431106077764,
+      "grad_norm": 0.6495422720909119,
+      "learning_rate": 8.234613507071754e-06,
+      "loss": 0.3675,
+      "step": 32960
+    },
+    {
+      "epoch": 12.446206115515288,
+      "grad_norm": 0.5902871489524841,
+      "learning_rate": 8.211901556047475e-06,
+      "loss": 0.3548,
+      "step": 32970
+    },
+    {
+      "epoch": 12.449981124952812,
+      "grad_norm": 0.597322404384613,
+      "learning_rate": 8.189218166794427e-06,
+      "loss": 0.35,
+      "step": 32980
+    },
+    {
+      "epoch": 12.453756134390336,
+      "grad_norm": 0.656929075717926,
+      "learning_rate": 8.166563354816509e-06,
+      "loss": 0.3605,
+      "step": 32990
+    },
+    {
+      "epoch": 12.45753114382786,
+      "grad_norm": 0.6605633497238159,
+      "learning_rate": 8.143937135598074e-06,
+      "loss": 0.3512,
+      "step": 33000
+    },
+    {
+      "epoch": 12.461306153265383,
+      "grad_norm": 0.6512786149978638,
+      "learning_rate": 8.121339524603943e-06,
+      "loss": 0.3524,
+      "step": 33010
+    },
+    {
+      "epoch": 12.465081162702907,
+      "grad_norm": 0.8367393016815186,
+      "learning_rate": 8.098770537279338e-06,
+      "loss": 0.3565,
+      "step": 33020
+    },
+    {
+      "epoch": 12.46885617214043,
+      "grad_norm": 0.602468729019165,
+      "learning_rate": 8.076230189049966e-06,
+      "loss": 0.3513,
+      "step": 33030
+    },
+    {
+      "epoch": 12.472631181577954,
+      "grad_norm": 0.7088872790336609,
+      "learning_rate": 8.05371849532196e-06,
+      "loss": 0.3584,
+      "step": 33040
+    },
+    {
+      "epoch": 12.476406191015478,
+      "grad_norm": 0.7384507060050964,
+      "learning_rate": 8.031235471481813e-06,
+      "loss": 0.3559,
+      "step": 33050
+    },
+    {
+      "epoch": 12.480181200453002,
+      "grad_norm": 0.714253842830658,
+      "learning_rate": 8.008781132896486e-06,
+      "loss": 0.3569,
+      "step": 33060
+    },
+    {
+      "epoch": 12.483956209890525,
+      "grad_norm": 0.6598066091537476,
+      "learning_rate": 7.986355494913312e-06,
+      "loss": 0.3623,
+      "step": 33070
+    },
+    {
+      "epoch": 12.487731219328047,
+      "grad_norm": 0.7399073839187622,
+      "learning_rate": 7.963958572859975e-06,
+      "loss": 0.3538,
+      "step": 33080
+    },
+    {
+      "epoch": 12.491506228765571,
+      "grad_norm": 0.6438301205635071,
+      "learning_rate": 7.941590382044583e-06,
+      "loss": 0.3642,
+      "step": 33090
+    },
+    {
+      "epoch": 12.495281238203095,
+      "grad_norm": 0.6220930814743042,
+      "learning_rate": 7.919250937755574e-06,
+      "loss": 0.3514,
+      "step": 33100
+    },
+    {
+      "epoch": 12.499056247640619,
+      "grad_norm": 0.5921124815940857,
+      "learning_rate": 7.896940255261764e-06,
+      "loss": 0.3619,
+      "step": 33110
+    },
+    {
+      "epoch": 12.502831257078142,
+      "grad_norm": 0.6638910174369812,
+      "learning_rate": 7.874658349812269e-06,
+      "loss": 0.3648,
+      "step": 33120
+    },
+    {
+      "epoch": 12.506606266515666,
+      "grad_norm": 0.6646683812141418,
+      "learning_rate": 7.85240523663659e-06,
+      "loss": 0.3603,
+      "step": 33130
+    },
+    {
+      "epoch": 12.51038127595319,
+      "grad_norm": 0.6426159739494324,
+      "learning_rate": 7.830180930944487e-06,
+      "loss": 0.3619,
+      "step": 33140
+    },
+    {
+      "epoch": 12.514156285390714,
+      "grad_norm": 0.6459843516349792,
+      "learning_rate": 7.80798544792608e-06,
+      "loss": 0.3581,
+      "step": 33150
+    },
+    {
+      "epoch": 12.517931294828237,
+      "grad_norm": 0.7059284448623657,
+      "learning_rate": 7.785818802751776e-06,
+      "loss": 0.3593,
+      "step": 33160
+    },
+    {
+      "epoch": 12.521706304265761,
+      "grad_norm": 0.5886619687080383,
+      "learning_rate": 7.763681010572277e-06,
+      "loss": 0.3711,
+      "step": 33170
+    },
+    {
+      "epoch": 12.525481313703285,
+      "grad_norm": 0.6966323852539062,
+      "learning_rate": 7.741572086518528e-06,
+      "loss": 0.3547,
+      "step": 33180
+    },
+    {
+      "epoch": 12.529256323140809,
+      "grad_norm": 0.6798071265220642,
+      "learning_rate": 7.719492045701782e-06,
+      "loss": 0.358,
+      "step": 33190
+    },
+    {
+      "epoch": 12.533031332578332,
+      "grad_norm": 0.7177457213401794,
+      "learning_rate": 7.697440903213549e-06,
+      "loss": 0.3652,
+      "step": 33200
+    },
+    {
+      "epoch": 12.536806342015854,
+      "grad_norm": 0.7038207054138184,
+      "learning_rate": 7.675418674125573e-06,
+      "loss": 0.3558,
+      "step": 33210
+    },
+    {
+      "epoch": 12.540581351453378,
+      "grad_norm": 0.7092708945274353,
+      "learning_rate": 7.65342537348983e-06,
+      "loss": 0.3569,
+      "step": 33220
+    },
+    {
+      "epoch": 12.544356360890902,
+      "grad_norm": 0.6563994884490967,
+      "learning_rate": 7.631461016338549e-06,
+      "loss": 0.3582,
+      "step": 33230
+    },
+    {
+      "epoch": 12.548131370328425,
+      "grad_norm": 0.6491038799285889,
+      "learning_rate": 7.609525617684138e-06,
+      "loss": 0.3556,
+      "step": 33240
+    },
+    {
+      "epoch": 12.55190637976595,
+      "grad_norm": 0.6468068957328796,
+      "learning_rate": 7.587619192519252e-06,
+      "loss": 0.3535,
+      "step": 33250
+    },
+    {
+      "epoch": 12.555681389203473,
+      "grad_norm": 0.6689091920852661,
+      "learning_rate": 7.565741755816725e-06,
+      "loss": 0.363,
+      "step": 33260
+    },
+    {
+      "epoch": 12.559456398640997,
+      "grad_norm": 0.6048831939697266,
+      "learning_rate": 7.543893322529588e-06,
+      "loss": 0.358,
+      "step": 33270
+    },
+    {
+      "epoch": 12.56323140807852,
+      "grad_norm": 0.6832661032676697,
+      "learning_rate": 7.522073907591015e-06,
+      "loss": 0.3602,
+      "step": 33280
+    },
+    {
+      "epoch": 12.567006417516044,
+      "grad_norm": 0.6341314911842346,
+      "learning_rate": 7.500283525914381e-06,
+      "loss": 0.358,
+      "step": 33290
+    },
+    {
+      "epoch": 12.570781426953568,
+      "grad_norm": 0.6825064420700073,
+      "learning_rate": 7.478522192393211e-06,
+      "loss": 0.3489,
+      "step": 33300
+    },
+    {
+      "epoch": 12.574556436391092,
+      "grad_norm": 0.5636059045791626,
+      "learning_rate": 7.456789921901181e-06,
+      "loss": 0.3566,
+      "step": 33310
+    },
+    {
+      "epoch": 12.578331445828615,
+      "grad_norm": 0.7039142847061157,
+      "learning_rate": 7.435086729292079e-06,
+      "loss": 0.357,
+      "step": 33320
+    },
+    {
+      "epoch": 12.582106455266139,
+      "grad_norm": 0.6117770075798035,
+      "learning_rate": 7.413412629399824e-06,
+      "loss": 0.3609,
+      "step": 33330
+    },
+    {
+      "epoch": 12.585881464703661,
+      "grad_norm": 0.6217195391654968,
+      "learning_rate": 7.391767637038466e-06,
+      "loss": 0.3557,
+      "step": 33340
+    },
+    {
+      "epoch": 12.589656474141185,
+      "grad_norm": 0.7603038549423218,
+      "learning_rate": 7.370151767002159e-06,
+      "loss": 0.3557,
+      "step": 33350
+    },
+    {
+      "epoch": 12.593431483578708,
+      "grad_norm": 0.660520613193512,
+      "learning_rate": 7.348565034065147e-06,
+      "loss": 0.3648,
+      "step": 33360
+    },
+    {
+      "epoch": 12.597206493016232,
+      "grad_norm": 0.6169815063476562,
+      "learning_rate": 7.327007452981765e-06,
+      "loss": 0.3573,
+      "step": 33370
+    },
+    {
+      "epoch": 12.600981502453756,
+      "grad_norm": 0.59542316198349,
+      "learning_rate": 7.305479038486396e-06,
+      "loss": 0.3576,
+      "step": 33380
+    },
+    {
+      "epoch": 12.60475651189128,
+      "grad_norm": 0.6083225011825562,
+      "learning_rate": 7.283979805293517e-06,
+      "loss": 0.353,
+      "step": 33390
+    },
+    {
+      "epoch": 12.608531521328803,
+      "grad_norm": 0.6213321089744568,
+      "learning_rate": 7.262509768097653e-06,
+      "loss": 0.3552,
+      "step": 33400
+    },
+    {
+      "epoch": 12.612306530766327,
+      "grad_norm": 0.6928298473358154,
+      "learning_rate": 7.241068941573381e-06,
+      "loss": 0.3615,
+      "step": 33410
+    },
+    {
+      "epoch": 12.61608154020385,
+      "grad_norm": 0.6654199957847595,
+      "learning_rate": 7.219657340375296e-06,
+      "loss": 0.3566,
+      "step": 33420
+    },
+    {
+      "epoch": 12.619856549641375,
+      "grad_norm": 0.6036897301673889,
+      "learning_rate": 7.198274979138014e-06,
+      "loss": 0.3679,
+      "step": 33430
+    },
+    {
+      "epoch": 12.623631559078898,
+      "grad_norm": 0.685053825378418,
+      "learning_rate": 7.176921872476189e-06,
+      "loss": 0.3523,
+      "step": 33440
+    },
+    {
+      "epoch": 12.627406568516422,
+      "grad_norm": 0.6758593320846558,
+      "learning_rate": 7.1555980349844674e-06,
+      "loss": 0.3555,
+      "step": 33450
+    },
+    {
+      "epoch": 12.631181577953946,
+      "grad_norm": 1.1203943490982056,
+      "learning_rate": 7.134303481237492e-06,
+      "loss": 0.3602,
+      "step": 33460
+    },
+    {
+      "epoch": 12.634956587391468,
+      "grad_norm": 0.6802670955657959,
+      "learning_rate": 7.113038225789892e-06,
+      "loss": 0.3648,
+      "step": 33470
+    },
+    {
+      "epoch": 12.638731596828991,
+      "grad_norm": 0.6041070222854614,
+      "learning_rate": 7.091802283176258e-06,
+      "loss": 0.3545,
+      "step": 33480
+    },
+    {
+      "epoch": 12.642506606266515,
+      "grad_norm": 0.6177239418029785,
+      "learning_rate": 7.070595667911162e-06,
+      "loss": 0.3596,
+      "step": 33490
+    },
+    {
+      "epoch": 12.646281615704039,
+      "grad_norm": 0.6666644215583801,
+      "learning_rate": 7.049418394489127e-06,
+      "loss": 0.3587,
+      "step": 33500
+    },
+    {
+      "epoch": 12.650056625141563,
+      "grad_norm": 0.7733400464057922,
+      "learning_rate": 7.028270477384636e-06,
+      "loss": 0.3585,
+      "step": 33510
+    },
+    {
+      "epoch": 12.653831634579086,
+      "grad_norm": 0.8025512099266052,
+      "learning_rate": 7.0071519310520695e-06,
+      "loss": 0.3509,
+      "step": 33520
+    },
+    {
+      "epoch": 12.65760664401661,
+      "grad_norm": 0.6963528990745544,
+      "learning_rate": 6.98606276992575e-06,
+      "loss": 0.3549,
+      "step": 33530
+    },
+    {
+      "epoch": 12.661381653454134,
+      "grad_norm": 0.705504298210144,
+      "learning_rate": 6.965003008419929e-06,
+      "loss": 0.3559,
+      "step": 33540
+    },
+    {
+      "epoch": 12.665156662891658,
+      "grad_norm": 0.7005383968353271,
+      "learning_rate": 6.943972660928755e-06,
+      "loss": 0.3532,
+      "step": 33550
+    },
+    {
+      "epoch": 12.668931672329181,
+      "grad_norm": 0.7059851884841919,
+      "learning_rate": 6.922971741826273e-06,
+      "loss": 0.3553,
+      "step": 33560
+    },
+    {
+      "epoch": 12.672706681766705,
+      "grad_norm": 0.6162865161895752,
+      "learning_rate": 6.902000265466424e-06,
+      "loss": 0.3571,
+      "step": 33570
+    },
+    {
+      "epoch": 12.676481691204227,
+      "grad_norm": 0.6487360596656799,
+      "learning_rate": 6.881058246182998e-06,
+      "loss": 0.36,
+      "step": 33580
+    },
+    {
+      "epoch": 12.68025670064175,
+      "grad_norm": 0.63123619556427,
+      "learning_rate": 6.860145698289666e-06,
+      "loss": 0.3617,
+      "step": 33590
+    },
+    {
+      "epoch": 12.684031710079275,
+      "grad_norm": 0.6987262964248657,
+      "learning_rate": 6.839262636079985e-06,
+      "loss": 0.3586,
+      "step": 33600
+    },
+    {
+      "epoch": 12.687806719516798,
+      "grad_norm": 0.6346328258514404,
+      "learning_rate": 6.818409073827292e-06,
+      "loss": 0.3589,
+      "step": 33610
+    },
+    {
+      "epoch": 12.691581728954322,
+      "grad_norm": 0.6731489896774292,
+      "learning_rate": 6.79758502578483e-06,
+      "loss": 0.361,
+      "step": 33620
+    },
+    {
+      "epoch": 12.695356738391846,
+      "grad_norm": 0.6909154653549194,
+      "learning_rate": 6.776790506185621e-06,
+      "loss": 0.3558,
+      "step": 33630
+    },
+    {
+      "epoch": 12.69913174782937,
+      "grad_norm": 0.6457541584968567,
+      "learning_rate": 6.756025529242533e-06,
+      "loss": 0.3564,
+      "step": 33640
+    },
+    {
+      "epoch": 12.702906757266893,
+      "grad_norm": 0.6108506321907043,
+      "learning_rate": 6.735290109148223e-06,
+      "loss": 0.3597,
+      "step": 33650
+    },
+    {
+      "epoch": 12.706681766704417,
+      "grad_norm": 0.699229896068573,
+      "learning_rate": 6.714584260075163e-06,
+      "loss": 0.3595,
+      "step": 33660
+    },
+    {
+      "epoch": 12.71045677614194,
+      "grad_norm": 0.6555954813957214,
+      "learning_rate": 6.693907996175619e-06,
+      "loss": 0.3443,
+      "step": 33670
+    },
+    {
+      "epoch": 12.714231785579464,
+      "grad_norm": 0.6641535758972168,
+      "learning_rate": 6.673261331581593e-06,
+      "loss": 0.3564,
+      "step": 33680
+    },
+    {
+      "epoch": 12.718006795016988,
+      "grad_norm": 0.6798548698425293,
+      "learning_rate": 6.652644280404912e-06,
+      "loss": 0.3535,
+      "step": 33690
+    },
+    {
+      "epoch": 12.721781804454512,
+      "grad_norm": 0.6146652698516846,
+      "learning_rate": 6.6320568567371175e-06,
+      "loss": 0.3541,
+      "step": 33700
+    },
+    {
+      "epoch": 12.725556813892034,
+      "grad_norm": 0.6196102499961853,
+      "learning_rate": 6.611499074649518e-06,
+      "loss": 0.3546,
+      "step": 33710
+    },
+    {
+      "epoch": 12.729331823329558,
+      "grad_norm": 0.7105995416641235,
+      "learning_rate": 6.590970948193187e-06,
+      "loss": 0.3663,
+      "step": 33720
+    },
+    {
+      "epoch": 12.733106832767081,
+      "grad_norm": 0.6891978979110718,
+      "learning_rate": 6.5704724913988825e-06,
+      "loss": 0.3672,
+      "step": 33730
+    },
+    {
+      "epoch": 12.736881842204605,
+      "grad_norm": 0.6553974151611328,
+      "learning_rate": 6.550003718277109e-06,
+      "loss": 0.356,
+      "step": 33740
+    },
+    {
+      "epoch": 12.740656851642129,
+      "grad_norm": 0.699963390827179,
+      "learning_rate": 6.529564642818087e-06,
+      "loss": 0.3572,
+      "step": 33750
+    },
+    {
+      "epoch": 12.744431861079653,
+      "grad_norm": 0.6198428869247437,
+      "learning_rate": 6.50915527899173e-06,
+      "loss": 0.3634,
+      "step": 33760
+    },
+    {
+      "epoch": 12.748206870517176,
+      "grad_norm": 0.6034395098686218,
+      "learning_rate": 6.4887756407476575e-06,
+      "loss": 0.3631,
+      "step": 33770
+    },
+    {
+      "epoch": 12.7519818799547,
+      "grad_norm": 0.6118060946464539,
+      "learning_rate": 6.468425742015149e-06,
+      "loss": 0.3553,
+      "step": 33780
+    },
+    {
+      "epoch": 12.755756889392224,
+      "grad_norm": 0.6215893626213074,
+      "learning_rate": 6.448105596703152e-06,
+      "loss": 0.3587,
+      "step": 33790
+    },
+    {
+      "epoch": 12.759531898829747,
+      "grad_norm": 0.6398429274559021,
+      "learning_rate": 6.427815218700317e-06,
+      "loss": 0.3582,
+      "step": 33800
+    },
+    {
+      "epoch": 12.763306908267271,
+      "grad_norm": 0.6768361330032349,
+      "learning_rate": 6.407554621874911e-06,
+      "loss": 0.3592,
+      "step": 33810
+    },
+    {
+      "epoch": 12.767081917704795,
+      "grad_norm": 0.6989614963531494,
+      "learning_rate": 6.3873238200748864e-06,
+      "loss": 0.3553,
+      "step": 33820
+    },
+    {
+      "epoch": 12.770856927142319,
+      "grad_norm": 0.7120910882949829,
+      "learning_rate": 6.367122827127775e-06,
+      "loss": 0.3587,
+      "step": 33830
+    },
+    {
+      "epoch": 12.77463193657984,
+      "grad_norm": 0.7099307179450989,
+      "learning_rate": 6.346951656840777e-06,
+      "loss": 0.3575,
+      "step": 33840
+    },
+    {
+      "epoch": 12.778406946017364,
+      "grad_norm": 0.6789335012435913,
+      "learning_rate": 6.326810323000698e-06,
+      "loss": 0.3565,
+      "step": 33850
+    },
+    {
+      "epoch": 12.782181955454888,
+      "grad_norm": 0.6464335322380066,
+      "learning_rate": 6.306698839373948e-06,
+      "loss": 0.3583,
+      "step": 33860
+    },
+    {
+      "epoch": 12.785956964892412,
+      "grad_norm": 0.7024797201156616,
+      "learning_rate": 6.286617219706547e-06,
+      "loss": 0.3623,
+      "step": 33870
+    },
+    {
+      "epoch": 12.789731974329936,
+      "grad_norm": 0.6883578896522522,
+      "learning_rate": 6.26656547772409e-06,
+      "loss": 0.3593,
+      "step": 33880
+    },
+    {
+      "epoch": 12.79350698376746,
+      "grad_norm": 0.656173050403595,
+      "learning_rate": 6.246543627131734e-06,
+      "loss": 0.3568,
+      "step": 33890
+    },
+    {
+      "epoch": 12.797281993204983,
+      "grad_norm": 0.593108594417572,
+      "learning_rate": 6.22655168161424e-06,
+      "loss": 0.3583,
+      "step": 33900
+    },
+    {
+      "epoch": 12.801057002642507,
+      "grad_norm": 0.7652211785316467,
+      "learning_rate": 6.20658965483591e-06,
+      "loss": 0.3587,
+      "step": 33910
+    },
+    {
+      "epoch": 12.80483201208003,
+      "grad_norm": 0.6598030924797058,
+      "learning_rate": 6.186657560440612e-06,
+      "loss": 0.3575,
+      "step": 33920
+    },
+    {
+      "epoch": 12.808607021517554,
+      "grad_norm": 0.6479567289352417,
+      "learning_rate": 6.166755412051734e-06,
+      "loss": 0.3528,
+      "step": 33930
+    },
+    {
+      "epoch": 12.812382030955078,
+      "grad_norm": 0.7282853722572327,
+      "learning_rate": 6.1468832232721985e-06,
+      "loss": 0.3615,
+      "step": 33940
+    },
+    {
+      "epoch": 12.816157040392602,
+      "grad_norm": 0.670918345451355,
+      "learning_rate": 6.127041007684476e-06,
+      "loss": 0.364,
+      "step": 33950
+    },
+    {
+      "epoch": 12.819932049830125,
+      "grad_norm": 0.6073424220085144,
+      "learning_rate": 6.107228778850527e-06,
+      "loss": 0.3621,
+      "step": 33960
+    },
+    {
+      "epoch": 12.823707059267647,
+      "grad_norm": 0.6667119264602661,
+      "learning_rate": 6.087446550311837e-06,
+      "loss": 0.3633,
+      "step": 33970
+    },
+    {
+      "epoch": 12.827482068705171,
+      "grad_norm": 0.6431853175163269,
+      "learning_rate": 6.067694335589353e-06,
+      "loss": 0.3569,
+      "step": 33980
+    },
+    {
+      "epoch": 12.831257078142695,
+      "grad_norm": 0.74583899974823,
+      "learning_rate": 6.047972148183528e-06,
+      "loss": 0.3595,
+      "step": 33990
+    },
+    {
+      "epoch": 12.835032087580219,
+      "grad_norm": 0.7069852948188782,
+      "learning_rate": 6.0282800015742936e-06,
+      "loss": 0.3536,
+      "step": 34000
+    },
+    {
+      "epoch": 12.838807097017742,
+      "grad_norm": 0.6726012825965881,
+      "learning_rate": 6.008617909221054e-06,
+      "loss": 0.3586,
+      "step": 34010
+    },
+    {
+      "epoch": 12.842582106455266,
+      "grad_norm": 0.704440712928772,
+      "learning_rate": 5.988985884562664e-06,
+      "loss": 0.357,
+      "step": 34020
+    },
+    {
+      "epoch": 12.84635711589279,
+      "grad_norm": 0.6290958523750305,
+      "learning_rate": 5.969383941017414e-06,
+      "loss": 0.3569,
+      "step": 34030
+    },
+    {
+      "epoch": 12.850132125330314,
+      "grad_norm": 0.6034786105155945,
+      "learning_rate": 5.949812091983059e-06,
+      "loss": 0.3541,
+      "step": 34040
+    },
+    {
+      "epoch": 12.853907134767837,
+      "grad_norm": 0.6244705319404602,
+      "learning_rate": 5.9302703508367695e-06,
+      "loss": 0.3569,
+      "step": 34050
+    },
+    {
+      "epoch": 12.857682144205361,
+      "grad_norm": 0.6468330025672913,
+      "learning_rate": 5.910758730935157e-06,
+      "loss": 0.3611,
+      "step": 34060
+    },
+    {
+      "epoch": 12.861457153642885,
+      "grad_norm": 0.6652206182479858,
+      "learning_rate": 5.8912772456142016e-06,
+      "loss": 0.3511,
+      "step": 34070
+    },
+    {
+      "epoch": 12.865232163080409,
+      "grad_norm": 0.6235213875770569,
+      "learning_rate": 5.871825908189338e-06,
+      "loss": 0.3538,
+      "step": 34080
+    },
+    {
+      "epoch": 12.86900717251793,
+      "grad_norm": 0.6621882319450378,
+      "learning_rate": 5.852404731955352e-06,
+      "loss": 0.3609,
+      "step": 34090
+    },
+    {
+      "epoch": 12.872782181955454,
+      "grad_norm": 0.6129846572875977,
+      "learning_rate": 5.833013730186443e-06,
+      "loss": 0.3619,
+      "step": 34100
+    },
+    {
+      "epoch": 12.876557191392978,
+      "grad_norm": 0.6356860399246216,
+      "learning_rate": 5.813652916136181e-06,
+      "loss": 0.3535,
+      "step": 34110
+    },
+    {
+      "epoch": 12.880332200830502,
+      "grad_norm": 0.6019640564918518,
+      "learning_rate": 5.794322303037508e-06,
+      "loss": 0.3631,
+      "step": 34120
+    },
+    {
+      "epoch": 12.884107210268025,
+      "grad_norm": 0.7462950348854065,
+      "learning_rate": 5.775021904102701e-06,
+      "loss": 0.3548,
+      "step": 34130
+    },
+    {
+      "epoch": 12.88788221970555,
+      "grad_norm": 0.8151988983154297,
+      "learning_rate": 5.755751732523401e-06,
+      "loss": 0.3544,
+      "step": 34140
+    },
+    {
+      "epoch": 12.891657229143073,
+      "grad_norm": 0.6986622214317322,
+      "learning_rate": 5.736511801470595e-06,
+      "loss": 0.3494,
+      "step": 34150
+    },
+    {
+      "epoch": 12.895432238580597,
+      "grad_norm": 0.6462591290473938,
+      "learning_rate": 5.717302124094603e-06,
+      "loss": 0.368,
+      "step": 34160
+    },
+    {
+      "epoch": 12.89920724801812,
+      "grad_norm": 0.6205435991287231,
+      "learning_rate": 5.6981227135250396e-06,
+      "loss": 0.3582,
+      "step": 34170
+    },
+    {
+      "epoch": 12.902982257455644,
+      "grad_norm": 4.8062543869018555,
+      "learning_rate": 5.678973582870867e-06,
+      "loss": 0.3628,
+      "step": 34180
+    },
+    {
+      "epoch": 12.906757266893168,
+      "grad_norm": 0.6735921502113342,
+      "learning_rate": 5.659854745220317e-06,
+      "loss": 0.3619,
+      "step": 34190
+    },
+    {
+      "epoch": 12.910532276330692,
+      "grad_norm": 0.5726591944694519,
+      "learning_rate": 5.640766213640947e-06,
+      "loss": 0.3549,
+      "step": 34200
+    },
+    {
+      "epoch": 12.914307285768214,
+      "grad_norm": 0.6301986575126648,
+      "learning_rate": 5.6217080011795934e-06,
+      "loss": 0.3588,
+      "step": 34210
+    },
+    {
+      "epoch": 12.918082295205737,
+      "grad_norm": 0.6712170243263245,
+      "learning_rate": 5.602680120862364e-06,
+      "loss": 0.3573,
+      "step": 34220
+    },
+    {
+      "epoch": 12.921857304643261,
+      "grad_norm": 0.7288546562194824,
+      "learning_rate": 5.583682585694616e-06,
+      "loss": 0.3583,
+      "step": 34230
+    },
+    {
+      "epoch": 12.925632314080785,
+      "grad_norm": 0.6534035801887512,
+      "learning_rate": 5.564715408661009e-06,
+      "loss": 0.3538,
+      "step": 34240
+    },
+    {
+      "epoch": 12.929407323518308,
+      "grad_norm": 0.6488263607025146,
+      "learning_rate": 5.5457786027254265e-06,
+      "loss": 0.3587,
+      "step": 34250
+    },
+    {
+      "epoch": 12.933182332955832,
+      "grad_norm": 0.6462862491607666,
+      "learning_rate": 5.526872180830977e-06,
+      "loss": 0.3626,
+      "step": 34260
+    },
+    {
+      "epoch": 12.936957342393356,
+      "grad_norm": 0.6145511269569397,
+      "learning_rate": 5.507996155900041e-06,
+      "loss": 0.3568,
+      "step": 34270
+    },
+    {
+      "epoch": 12.94073235183088,
+      "grad_norm": 0.6510082483291626,
+      "learning_rate": 5.489150540834203e-06,
+      "loss": 0.3537,
+      "step": 34280
+    },
+    {
+      "epoch": 12.944507361268403,
+      "grad_norm": 0.7098113298416138,
+      "learning_rate": 5.47033534851425e-06,
+      "loss": 0.3498,
+      "step": 34290
+    },
+    {
+      "epoch": 12.948282370705927,
+      "grad_norm": 0.6072222590446472,
+      "learning_rate": 5.451550591800198e-06,
+      "loss": 0.3544,
+      "step": 34300
+    },
+    {
+      "epoch": 12.95205738014345,
+      "grad_norm": 0.6960251331329346,
+      "learning_rate": 5.43279628353125e-06,
+      "loss": 0.3611,
+      "step": 34310
+    },
+    {
+      "epoch": 12.955832389580975,
+      "grad_norm": 0.6805595755577087,
+      "learning_rate": 5.4140724365258075e-06,
+      "loss": 0.3595,
+      "step": 34320
+    },
+    {
+      "epoch": 12.959607399018498,
+      "grad_norm": 0.651698648929596,
+      "learning_rate": 5.395379063581424e-06,
+      "loss": 0.3566,
+      "step": 34330
+    },
+    {
+      "epoch": 12.96338240845602,
+      "grad_norm": 0.6734188199043274,
+      "learning_rate": 5.376716177474866e-06,
+      "loss": 0.3475,
+      "step": 34340
+    },
+    {
+      "epoch": 12.967157417893544,
+      "grad_norm": 0.6215294599533081,
+      "learning_rate": 5.358083790962015e-06,
+      "loss": 0.3579,
+      "step": 34350
+    },
+    {
+      "epoch": 12.970932427331068,
+      "grad_norm": 0.7030113339424133,
+      "learning_rate": 5.33948191677795e-06,
+      "loss": 0.356,
+      "step": 34360
+    },
+    {
+      "epoch": 12.974707436768592,
+      "grad_norm": 0.6129488945007324,
+      "learning_rate": 5.320910567636866e-06,
+      "loss": 0.3583,
+      "step": 34370
+    },
+    {
+      "epoch": 12.978482446206115,
+      "grad_norm": 2.6713273525238037,
+      "learning_rate": 5.302369756232123e-06,
+      "loss": 0.3658,
+      "step": 34380
+    },
+    {
+      "epoch": 12.982257455643639,
+      "grad_norm": 0.6619277596473694,
+      "learning_rate": 5.283859495236165e-06,
+      "loss": 0.3596,
+      "step": 34390
+    },
+    {
+      "epoch": 12.986032465081163,
+      "grad_norm": 0.6460644006729126,
+      "learning_rate": 5.265379797300602e-06,
+      "loss": 0.352,
+      "step": 34400
+    },
+    {
+      "epoch": 12.989807474518686,
+      "grad_norm": 0.6166976690292358,
+      "learning_rate": 5.246930675056117e-06,
+      "loss": 0.364,
+      "step": 34410
+    },
+    {
+      "epoch": 12.99358248395621,
+      "grad_norm": 0.5918395519256592,
+      "learning_rate": 5.228512141112535e-06,
+      "loss": 0.3565,
+      "step": 34420
+    },
+    {
+      "epoch": 12.997357493393734,
+      "grad_norm": 3.736072063446045,
+      "learning_rate": 5.21012420805872e-06,
+      "loss": 0.4422,
+      "step": 34430
+    },
+    {
+      "epoch": 13.001132502831258,
+      "grad_norm": 0.6360698342323303,
+      "learning_rate": 5.191766888462679e-06,
+      "loss": 0.3608,
+      "step": 34440
+    },
+    {
+      "epoch": 13.004907512268781,
+      "grad_norm": 1.15024733543396,
+      "learning_rate": 5.173440194871437e-06,
+      "loss": 0.3568,
+      "step": 34450
+    },
+    {
+      "epoch": 13.008682521706305,
+      "grad_norm": 0.6446855068206787,
+      "learning_rate": 5.155144139811141e-06,
+      "loss": 0.3572,
+      "step": 34460
+    },
+    {
+      "epoch": 13.012457531143827,
+      "grad_norm": 0.6619588732719421,
+      "learning_rate": 5.13687873578696e-06,
+      "loss": 0.3568,
+      "step": 34470
+    },
+    {
+      "epoch": 13.01623254058135,
+      "grad_norm": 0.7428315281867981,
+      "learning_rate": 5.11864399528314e-06,
+      "loss": 0.3547,
+      "step": 34480
+    },
+    {
+      "epoch": 13.020007550018875,
+      "grad_norm": 0.6570870876312256,
+      "learning_rate": 5.1004399307629346e-06,
+      "loss": 0.3606,
+      "step": 34490
+    },
+    {
+      "epoch": 13.023782559456398,
+      "grad_norm": 0.660350501537323,
+      "learning_rate": 5.082266554668652e-06,
+      "loss": 0.3647,
+      "step": 34500
+    },
+    {
+      "epoch": 13.027557568893922,
+      "grad_norm": 0.8094748258590698,
+      "learning_rate": 5.06412387942164e-06,
+      "loss": 0.3574,
+      "step": 34510
+    },
+    {
+      "epoch": 13.031332578331446,
+      "grad_norm": 0.669248640537262,
+      "learning_rate": 5.046011917422239e-06,
+      "loss": 0.3524,
+      "step": 34520
+    },
+    {
+      "epoch": 13.03510758776897,
+      "grad_norm": 0.6466529369354248,
+      "learning_rate": 5.0279306810498026e-06,
+      "loss": 0.3646,
+      "step": 34530
+    },
+    {
+      "epoch": 13.038882597206493,
+      "grad_norm": 0.6580777168273926,
+      "learning_rate": 5.009880182662679e-06,
+      "loss": 0.3482,
+      "step": 34540
+    },
+    {
+      "epoch": 13.042657606644017,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 4.991860434598217e-06,
+      "loss": 0.3527,
+      "step": 34550
+    },
+    {
+      "epoch": 13.04643261608154,
+      "grad_norm": 0.6849576830863953,
+      "learning_rate": 4.973871449172746e-06,
+      "loss": 0.3616,
+      "step": 34560
+    },
+    {
+      "epoch": 13.050207625519064,
+      "grad_norm": 0.6503666043281555,
+      "learning_rate": 4.9559132386815695e-06,
+      "loss": 0.3614,
+      "step": 34570
+    },
+    {
+      "epoch": 13.053982634956588,
+      "grad_norm": 0.6452770233154297,
+      "learning_rate": 4.93798581539896e-06,
+      "loss": 0.3598,
+      "step": 34580
+    },
+    {
+      "epoch": 13.05775764439411,
+      "grad_norm": 0.6467193961143494,
+      "learning_rate": 4.920089191578131e-06,
+      "loss": 0.36,
+      "step": 34590
+    },
+    {
+      "epoch": 13.061532653831634,
+      "grad_norm": 0.6155425906181335,
+      "learning_rate": 4.90222337945126e-06,
+      "loss": 0.3565,
+      "step": 34600
+    },
+    {
+      "epoch": 13.065307663269158,
+      "grad_norm": 0.687741219997406,
+      "learning_rate": 4.8843883912294605e-06,
+      "loss": 0.3541,
+      "step": 34610
+    },
+    {
+      "epoch": 13.069082672706681,
+      "grad_norm": 0.6692128777503967,
+      "learning_rate": 4.866584239102795e-06,
+      "loss": 0.3546,
+      "step": 34620
+    },
+    {
+      "epoch": 13.072857682144205,
+      "grad_norm": 1.0037407875061035,
+      "learning_rate": 4.8488109352402164e-06,
+      "loss": 0.3644,
+      "step": 34630
+    },
+    {
+      "epoch": 13.076632691581729,
+      "grad_norm": 0.6341879963874817,
+      "learning_rate": 4.831068491789603e-06,
+      "loss": 0.3578,
+      "step": 34640
+    },
+    {
+      "epoch": 13.080407701019253,
+      "grad_norm": 0.6491614580154419,
+      "learning_rate": 4.81335692087776e-06,
+      "loss": 0.3579,
+      "step": 34650
+    },
+    {
+      "epoch": 13.084182710456776,
+      "grad_norm": 0.6379111409187317,
+      "learning_rate": 4.795676234610374e-06,
+      "loss": 0.3643,
+      "step": 34660
+    },
+    {
+      "epoch": 13.0879577198943,
+      "grad_norm": 0.678970992565155,
+      "learning_rate": 4.778026445072037e-06,
+      "loss": 0.3638,
+      "step": 34670
+    },
+    {
+      "epoch": 13.091732729331824,
+      "grad_norm": 0.6357844471931458,
+      "learning_rate": 4.760407564326208e-06,
+      "loss": 0.3521,
+      "step": 34680
+    },
+    {
+      "epoch": 13.095507738769347,
+      "grad_norm": 0.5853003859519958,
+      "learning_rate": 4.742819604415222e-06,
+      "loss": 0.3604,
+      "step": 34690
+    },
+    {
+      "epoch": 13.099282748206871,
+      "grad_norm": 0.7209833264350891,
+      "learning_rate": 4.725262577360284e-06,
+      "loss": 0.3599,
+      "step": 34700
+    },
+    {
+      "epoch": 13.103057757644395,
+      "grad_norm": 0.6747405529022217,
+      "learning_rate": 4.707736495161458e-06,
+      "loss": 0.3538,
+      "step": 34710
+    },
+    {
+      "epoch": 13.106832767081917,
+      "grad_norm": 0.6171832084655762,
+      "learning_rate": 4.690241369797671e-06,
+      "loss": 0.3574,
+      "step": 34720
+    },
+    {
+      "epoch": 13.11060777651944,
+      "grad_norm": 0.9365268349647522,
+      "learning_rate": 4.672777213226659e-06,
+      "loss": 0.3562,
+      "step": 34730
+    },
+    {
+      "epoch": 13.114382785956964,
+      "grad_norm": 0.6216146945953369,
+      "learning_rate": 4.655344037385001e-06,
+      "loss": 0.3577,
+      "step": 34740
+    },
+    {
+      "epoch": 13.118157795394488,
+      "grad_norm": 0.6404061317443848,
+      "learning_rate": 4.6379418541881215e-06,
+      "loss": 0.3617,
+      "step": 34750
+    },
+    {
+      "epoch": 13.121932804832012,
+      "grad_norm": 0.638793408870697,
+      "learning_rate": 4.620570675530245e-06,
+      "loss": 0.3593,
+      "step": 34760
+    },
+    {
+      "epoch": 13.125707814269536,
+      "grad_norm": 0.6440224647521973,
+      "learning_rate": 4.603230513284407e-06,
+      "loss": 0.3637,
+      "step": 34770
+    },
+    {
+      "epoch": 13.12948282370706,
+      "grad_norm": 0.7139766216278076,
+      "learning_rate": 4.58592137930246e-06,
+      "loss": 0.3537,
+      "step": 34780
+    },
+    {
+      "epoch": 13.133257833144583,
+      "grad_norm": 0.752666711807251,
+      "learning_rate": 4.568643285415009e-06,
+      "loss": 0.3647,
+      "step": 34790
+    },
+    {
+      "epoch": 13.137032842582107,
+      "grad_norm": 0.7453562617301941,
+      "learning_rate": 4.551396243431483e-06,
+      "loss": 0.3593,
+      "step": 34800
+    },
+    {
+      "epoch": 13.14080785201963,
+      "grad_norm": 0.6269282698631287,
+      "learning_rate": 4.534180265140075e-06,
+      "loss": 0.3643,
+      "step": 34810
+    },
+    {
+      "epoch": 13.144582861457154,
+      "grad_norm": 0.7206990122795105,
+      "learning_rate": 4.5169953623077356e-06,
+      "loss": 0.3554,
+      "step": 34820
+    },
+    {
+      "epoch": 13.148357870894678,
+      "grad_norm": 0.6602156758308411,
+      "learning_rate": 4.499841546680189e-06,
+      "loss": 0.3563,
+      "step": 34830
+    },
+    {
+      "epoch": 13.1521328803322,
+      "grad_norm": 0.6244385838508606,
+      "learning_rate": 4.4827188299819025e-06,
+      "loss": 0.348,
+      "step": 34840
+    },
+    {
+      "epoch": 13.155907889769724,
+      "grad_norm": 0.6173793077468872,
+      "learning_rate": 4.465627223916086e-06,
+      "loss": 0.359,
+      "step": 34850
+    },
+    {
+      "epoch": 13.159682899207247,
+      "grad_norm": 0.6792297959327698,
+      "learning_rate": 4.448566740164706e-06,
+      "loss": 0.3597,
+      "step": 34860
+    },
+    {
+      "epoch": 13.163457908644771,
+      "grad_norm": 0.6277993321418762,
+      "learning_rate": 4.431537390388429e-06,
+      "loss": 0.3617,
+      "step": 34870
+    },
+    {
+      "epoch": 13.167232918082295,
+      "grad_norm": 0.6655922532081604,
+      "learning_rate": 4.414539186226674e-06,
+      "loss": 0.353,
+      "step": 34880
+    },
+    {
+      "epoch": 13.171007927519819,
+      "grad_norm": 0.670093834400177,
+      "learning_rate": 4.397572139297529e-06,
+      "loss": 0.3579,
+      "step": 34890
+    },
+    {
+      "epoch": 13.174782936957342,
+      "grad_norm": 0.6460064649581909,
+      "learning_rate": 4.380636261197829e-06,
+      "loss": 0.3587,
+      "step": 34900
+    },
+    {
+      "epoch": 13.178557946394866,
+      "grad_norm": 0.5938677191734314,
+      "learning_rate": 4.36373156350307e-06,
+      "loss": 0.3583,
+      "step": 34910
+    },
+    {
+      "epoch": 13.18233295583239,
+      "grad_norm": 0.6573899388313293,
+      "learning_rate": 4.3468580577674665e-06,
+      "loss": 0.3568,
+      "step": 34920
+    },
+    {
+      "epoch": 13.186107965269914,
+      "grad_norm": 0.6724887490272522,
+      "learning_rate": 4.330015755523908e-06,
+      "loss": 0.3636,
+      "step": 34930
+    },
+    {
+      "epoch": 13.189882974707437,
+      "grad_norm": 0.8144491910934448,
+      "learning_rate": 4.313204668283926e-06,
+      "loss": 0.3654,
+      "step": 34940
+    },
+    {
+      "epoch": 13.193657984144961,
+      "grad_norm": 0.5686742663383484,
+      "learning_rate": 4.296424807537752e-06,
+      "loss": 0.3628,
+      "step": 34950
+    },
+    {
+      "epoch": 13.197432993582485,
+      "grad_norm": 0.9146296977996826,
+      "learning_rate": 4.279676184754267e-06,
+      "loss": 0.3596,
+      "step": 34960
+    },
+    {
+      "epoch": 13.201208003020007,
+      "grad_norm": 0.639041006565094,
+      "learning_rate": 4.262958811380985e-06,
+      "loss": 0.3575,
+      "step": 34970
+    },
+    {
+      "epoch": 13.20498301245753,
+      "grad_norm": 0.7431603074073792,
+      "learning_rate": 4.246272698844095e-06,
+      "loss": 0.3555,
+      "step": 34980
+    },
+    {
+      "epoch": 13.208758021895054,
+      "grad_norm": 0.6070488691329956,
+      "learning_rate": 4.229617858548374e-06,
+      "loss": 0.3574,
+      "step": 34990
+    },
+    {
+      "epoch": 13.212533031332578,
+      "grad_norm": 1.7812660932540894,
+      "learning_rate": 4.212994301877271e-06,
+      "loss": 0.3628,
+      "step": 35000
+    },
+    {
+      "epoch": 13.216308040770102,
+      "grad_norm": 0.763363242149353,
+      "learning_rate": 4.196402040192804e-06,
+      "loss": 0.3554,
+      "step": 35010
+    },
+    {
+      "epoch": 13.220083050207625,
+      "grad_norm": 1.0130776166915894,
+      "learning_rate": 4.179841084835639e-06,
+      "loss": 0.3557,
+      "step": 35020
+    },
+    {
+      "epoch": 13.22385805964515,
+      "grad_norm": 0.64314204454422,
+      "learning_rate": 4.163311447125046e-06,
+      "loss": 0.3546,
+      "step": 35030
+    },
+    {
+      "epoch": 13.227633069082673,
+      "grad_norm": 0.6824278831481934,
+      "learning_rate": 4.146813138358851e-06,
+      "loss": 0.3567,
+      "step": 35040
+    },
+    {
+      "epoch": 13.231408078520197,
+      "grad_norm": 0.6351913213729858,
+      "learning_rate": 4.130346169813509e-06,
+      "loss": 0.3514,
+      "step": 35050
+    },
+    {
+      "epoch": 13.23518308795772,
+      "grad_norm": 0.6978806853294373,
+      "learning_rate": 4.113910552744038e-06,
+      "loss": 0.3568,
+      "step": 35060
+    },
+    {
+      "epoch": 13.238958097395244,
+      "grad_norm": 0.6495206356048584,
+      "learning_rate": 4.097506298384024e-06,
+      "loss": 0.3525,
+      "step": 35070
+    },
+    {
+      "epoch": 13.242733106832768,
+      "grad_norm": 0.6304375529289246,
+      "learning_rate": 4.081133417945626e-06,
+      "loss": 0.3564,
+      "step": 35080
+    },
+    {
+      "epoch": 13.246508116270292,
+      "grad_norm": 0.6751596927642822,
+      "learning_rate": 4.0647919226195484e-06,
+      "loss": 0.3574,
+      "step": 35090
+    },
+    {
+      "epoch": 13.250283125707814,
+      "grad_norm": 0.6097880005836487,
+      "learning_rate": 4.04848182357504e-06,
+      "loss": 0.3571,
+      "step": 35100
+    },
+    {
+      "epoch": 13.254058135145337,
+      "grad_norm": 0.6437689661979675,
+      "learning_rate": 4.032203131959911e-06,
+      "loss": 0.3664,
+      "step": 35110
+    },
+    {
+      "epoch": 13.257833144582861,
+      "grad_norm": 0.6554880142211914,
+      "learning_rate": 4.015955858900483e-06,
+      "loss": 0.3576,
+      "step": 35120
+    },
+    {
+      "epoch": 13.261608154020385,
+      "grad_norm": 0.6221573352813721,
+      "learning_rate": 3.999740015501641e-06,
+      "loss": 0.3552,
+      "step": 35130
+    },
+    {
+      "epoch": 13.265383163457908,
+      "grad_norm": 0.8590520024299622,
+      "learning_rate": 3.9835556128467236e-06,
+      "loss": 0.3618,
+      "step": 35140
+    },
+    {
+      "epoch": 13.269158172895432,
+      "grad_norm": 0.6513279676437378,
+      "learning_rate": 3.967402661997644e-06,
+      "loss": 0.3588,
+      "step": 35150
+    },
+    {
+      "epoch": 13.272933182332956,
+      "grad_norm": 0.7151180505752563,
+      "learning_rate": 3.951281173994781e-06,
+      "loss": 0.3559,
+      "step": 35160
+    },
+    {
+      "epoch": 13.27670819177048,
+      "grad_norm": 0.6879009008407593,
+      "learning_rate": 3.935191159857016e-06,
+      "loss": 0.3538,
+      "step": 35170
+    },
+    {
+      "epoch": 13.280483201208003,
+      "grad_norm": 0.6725070476531982,
+      "learning_rate": 3.919132630581746e-06,
+      "loss": 0.364,
+      "step": 35180
+    },
+    {
+      "epoch": 13.284258210645527,
+      "grad_norm": 0.6446594595909119,
+      "learning_rate": 3.903105597144802e-06,
+      "loss": 0.3523,
+      "step": 35190
+    },
+    {
+      "epoch": 13.28803322008305,
+      "grad_norm": 0.6886351704597473,
+      "learning_rate": 3.887110070500505e-06,
+      "loss": 0.3607,
+      "step": 35200
+    },
+    {
+      "epoch": 13.291808229520575,
+      "grad_norm": 0.6238877177238464,
+      "learning_rate": 3.871146061581654e-06,
+      "loss": 0.3507,
+      "step": 35210
+    },
+    {
+      "epoch": 13.295583238958097,
+      "grad_norm": 0.640650749206543,
+      "learning_rate": 3.855213581299505e-06,
+      "loss": 0.3612,
+      "step": 35220
+    },
+    {
+      "epoch": 13.29935824839562,
+      "grad_norm": 0.665745198726654,
+      "learning_rate": 3.839312640543763e-06,
+      "loss": 0.359,
+      "step": 35230
+    },
+    {
+      "epoch": 13.303133257833144,
+      "grad_norm": 0.6206352114677429,
+      "learning_rate": 3.823443250182551e-06,
+      "loss": 0.3547,
+      "step": 35240
+    },
+    {
+      "epoch": 13.306908267270668,
+      "grad_norm": 1.118415117263794,
+      "learning_rate": 3.80760542106246e-06,
+      "loss": 0.3534,
+      "step": 35250
+    },
+    {
+      "epoch": 13.310683276708192,
+      "grad_norm": 0.6065353751182556,
+      "learning_rate": 3.7917991640084972e-06,
+      "loss": 0.3532,
+      "step": 35260
+    },
+    {
+      "epoch": 13.314458286145715,
+      "grad_norm": 0.6253830194473267,
+      "learning_rate": 3.7760244898240925e-06,
+      "loss": 0.4729,
+      "step": 35270
+    },
+    {
+      "epoch": 13.318233295583239,
+      "grad_norm": 0.5790948271751404,
+      "learning_rate": 3.760281409291078e-06,
+      "loss": 0.3582,
+      "step": 35280
+    },
+    {
+      "epoch": 13.322008305020763,
+      "grad_norm": 0.6971233487129211,
+      "learning_rate": 3.7445699331697113e-06,
+      "loss": 0.3591,
+      "step": 35290
+    },
+    {
+      "epoch": 13.325783314458286,
+      "grad_norm": 0.7019782066345215,
+      "learning_rate": 3.7288900721986262e-06,
+      "loss": 0.3583,
+      "step": 35300
+    },
+    {
+      "epoch": 13.32955832389581,
+      "grad_norm": 0.682455837726593,
+      "learning_rate": 3.7132418370948663e-06,
+      "loss": 0.3568,
+      "step": 35310
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 0.6666685938835144,
+      "learning_rate": 3.6976252385538535e-06,
+      "loss": 0.358,
+      "step": 35320
+    },
+    {
+      "epoch": 13.337108342770858,
+      "grad_norm": 0.634299099445343,
+      "learning_rate": 3.6820402872493832e-06,
+      "loss": 0.3542,
+      "step": 35330
+    },
+    {
+      "epoch": 13.34088335220838,
+      "grad_norm": 0.6098415851593018,
+      "learning_rate": 3.6664869938336343e-06,
+      "loss": 0.3598,
+      "step": 35340
+    },
+    {
+      "epoch": 13.344658361645903,
+      "grad_norm": 0.6019866466522217,
+      "learning_rate": 3.650965368937115e-06,
+      "loss": 0.3575,
+      "step": 35350
+    },
+    {
+      "epoch": 13.348433371083427,
+      "grad_norm": 0.6554697155952454,
+      "learning_rate": 3.6354754231687214e-06,
+      "loss": 0.3604,
+      "step": 35360
+    },
+    {
+      "epoch": 13.35220838052095,
+      "grad_norm": 1.0949931144714355,
+      "learning_rate": 3.620017167115697e-06,
+      "loss": 0.3586,
+      "step": 35370
+    },
+    {
+      "epoch": 13.355983389958475,
+      "grad_norm": 0.663237988948822,
+      "learning_rate": 3.604590611343589e-06,
+      "loss": 0.3582,
+      "step": 35380
+    },
+    {
+      "epoch": 13.359758399395998,
+      "grad_norm": 0.6626077890396118,
+      "learning_rate": 3.589195766396325e-06,
+      "loss": 0.3581,
+      "step": 35390
+    },
+    {
+      "epoch": 13.363533408833522,
+      "grad_norm": 0.5932807326316833,
+      "learning_rate": 3.573832642796121e-06,
+      "loss": 0.3545,
+      "step": 35400
+    },
+    {
+      "epoch": 13.367308418271046,
+      "grad_norm": 0.6529201865196228,
+      "learning_rate": 3.5585012510435336e-06,
+      "loss": 0.3567,
+      "step": 35410
+    },
+    {
+      "epoch": 13.37108342770857,
+      "grad_norm": 0.6512988805770874,
+      "learning_rate": 3.5432016016174184e-06,
+      "loss": 0.3592,
+      "step": 35420
+    },
+    {
+      "epoch": 13.374858437146093,
+      "grad_norm": 0.6883192658424377,
+      "learning_rate": 3.5279337049749506e-06,
+      "loss": 0.3495,
+      "step": 35430
+    },
+    {
+      "epoch": 13.378633446583617,
+      "grad_norm": 0.625953733921051,
+      "learning_rate": 3.5126975715515985e-06,
+      "loss": 0.3524,
+      "step": 35440
+    },
+    {
+      "epoch": 13.38240845602114,
+      "grad_norm": 0.7120693325996399,
+      "learning_rate": 3.4974932117610994e-06,
+      "loss": 0.3546,
+      "step": 35450
+    },
+    {
+      "epoch": 13.386183465458664,
+      "grad_norm": 0.6227659583091736,
+      "learning_rate": 3.4823206359955065e-06,
+      "loss": 0.3585,
+      "step": 35460
+    },
+    {
+      "epoch": 13.389958474896186,
+      "grad_norm": 0.6581076979637146,
+      "learning_rate": 3.4671798546251133e-06,
+      "loss": 0.3584,
+      "step": 35470
+    },
+    {
+      "epoch": 13.39373348433371,
+      "grad_norm": 0.6351528167724609,
+      "learning_rate": 3.4520708779985088e-06,
+      "loss": 0.3486,
+      "step": 35480
+    },
+    {
+      "epoch": 13.397508493771234,
+      "grad_norm": 0.6142309308052063,
+      "learning_rate": 3.4369937164425493e-06,
+      "loss": 0.3578,
+      "step": 35490
+    },
+    {
+      "epoch": 13.401283503208758,
+      "grad_norm": 0.6407093405723572,
+      "learning_rate": 3.421948380262313e-06,
+      "loss": 0.3559,
+      "step": 35500
+    },
+    {
+      "epoch": 13.405058512646281,
+      "grad_norm": 0.6450949311256409,
+      "learning_rate": 3.406934879741158e-06,
+      "loss": 0.3575,
+      "step": 35510
+    },
+    {
+      "epoch": 13.408833522083805,
+      "grad_norm": 0.7444202899932861,
+      "learning_rate": 3.391953225140665e-06,
+      "loss": 0.3688,
+      "step": 35520
+    },
+    {
+      "epoch": 13.412608531521329,
+      "grad_norm": 0.706688642501831,
+      "learning_rate": 3.3770034267006533e-06,
+      "loss": 0.3587,
+      "step": 35530
+    },
+    {
+      "epoch": 13.416383540958853,
+      "grad_norm": 0.6597870588302612,
+      "learning_rate": 3.362085494639178e-06,
+      "loss": 0.3541,
+      "step": 35540
+    },
+    {
+      "epoch": 13.420158550396376,
+      "grad_norm": 0.5902731418609619,
+      "learning_rate": 3.3471994391524986e-06,
+      "loss": 0.3547,
+      "step": 35550
+    },
+    {
+      "epoch": 13.4239335598339,
+      "grad_norm": 0.6152647733688354,
+      "learning_rate": 3.3323452704150825e-06,
+      "loss": 0.3527,
+      "step": 35560
+    },
+    {
+      "epoch": 13.427708569271424,
+      "grad_norm": 0.6410410404205322,
+      "learning_rate": 3.3175229985796253e-06,
+      "loss": 0.3597,
+      "step": 35570
+    },
+    {
+      "epoch": 13.431483578708947,
+      "grad_norm": 0.6362301111221313,
+      "learning_rate": 3.3027326337770115e-06,
+      "loss": 0.3563,
+      "step": 35580
+    },
+    {
+      "epoch": 13.435258588146471,
+      "grad_norm": 0.6383989453315735,
+      "learning_rate": 3.2879741861163216e-06,
+      "loss": 0.3579,
+      "step": 35590
+    },
+    {
+      "epoch": 13.439033597583993,
+      "grad_norm": 0.6493092179298401,
+      "learning_rate": 3.2732476656847976e-06,
+      "loss": 0.3608,
+      "step": 35600
+    },
+    {
+      "epoch": 13.442808607021517,
+      "grad_norm": 0.6834067106246948,
+      "learning_rate": 3.2585530825478883e-06,
+      "loss": 0.3529,
+      "step": 35610
+    },
+    {
+      "epoch": 13.44658361645904,
+      "grad_norm": 0.7077740430831909,
+      "learning_rate": 3.243890446749204e-06,
+      "loss": 0.3575,
+      "step": 35620
+    },
+    {
+      "epoch": 13.450358625896564,
+      "grad_norm": 0.6362556219100952,
+      "learning_rate": 3.2292597683105107e-06,
+      "loss": 0.3594,
+      "step": 35630
+    },
+    {
+      "epoch": 13.454133635334088,
+      "grad_norm": 0.6774179339408875,
+      "learning_rate": 3.214661057231755e-06,
+      "loss": 0.3668,
+      "step": 35640
+    },
+    {
+      "epoch": 13.457908644771612,
+      "grad_norm": 0.7050675749778748,
+      "learning_rate": 3.2000943234910097e-06,
+      "loss": 0.3517,
+      "step": 35650
+    },
+    {
+      "epoch": 13.461683654209136,
+      "grad_norm": 0.8577159643173218,
+      "learning_rate": 3.1855595770444835e-06,
+      "loss": 0.3551,
+      "step": 35660
+    },
+    {
+      "epoch": 13.46545866364666,
+      "grad_norm": 0.6533629298210144,
+      "learning_rate": 3.1710568278265574e-06,
+      "loss": 0.347,
+      "step": 35670
+    },
+    {
+      "epoch": 13.469233673084183,
+      "grad_norm": 0.63999342918396,
+      "learning_rate": 3.156586085749719e-06,
+      "loss": 0.3484,
+      "step": 35680
+    },
+    {
+      "epoch": 13.473008682521707,
+      "grad_norm": 0.6630726456642151,
+      "learning_rate": 3.1421473607045893e-06,
+      "loss": 0.3556,
+      "step": 35690
+    },
+    {
+      "epoch": 13.47678369195923,
+      "grad_norm": 1.4107182025909424,
+      "learning_rate": 3.127740662559886e-06,
+      "loss": 0.355,
+      "step": 35700
+    },
+    {
+      "epoch": 13.480558701396754,
+      "grad_norm": 0.7354855537414551,
+      "learning_rate": 3.1133660011624655e-06,
+      "loss": 0.3602,
+      "step": 35710
+    },
+    {
+      "epoch": 13.484333710834276,
+      "grad_norm": 0.6842527389526367,
+      "learning_rate": 3.099023386337263e-06,
+      "loss": 0.3602,
+      "step": 35720
+    },
+    {
+      "epoch": 13.4881087202718,
+      "grad_norm": 0.6937289237976074,
+      "learning_rate": 3.0847128278873317e-06,
+      "loss": 0.3585,
+      "step": 35730
+    },
+    {
+      "epoch": 13.491883729709324,
+      "grad_norm": 0.7109084725379944,
+      "learning_rate": 3.0704343355938027e-06,
+      "loss": 0.3568,
+      "step": 35740
+    },
+    {
+      "epoch": 13.495658739146847,
+      "grad_norm": 0.7426275610923767,
+      "learning_rate": 3.056187919215886e-06,
+      "loss": 0.3545,
+      "step": 35750
+    },
+    {
+      "epoch": 13.499433748584371,
+      "grad_norm": 0.7047889828681946,
+      "learning_rate": 3.041973588490865e-06,
+      "loss": 0.3583,
+      "step": 35760
+    },
+    {
+      "epoch": 13.503208758021895,
+      "grad_norm": 1.050673007965088,
+      "learning_rate": 3.0277913531341128e-06,
+      "loss": 0.3619,
+      "step": 35770
+    },
+    {
+      "epoch": 13.506983767459419,
+      "grad_norm": 0.6489849090576172,
+      "learning_rate": 3.0136412228390476e-06,
+      "loss": 0.3544,
+      "step": 35780
+    },
+    {
+      "epoch": 13.510758776896942,
+      "grad_norm": 0.635733425617218,
+      "learning_rate": 2.999523207277155e-06,
+      "loss": 0.3575,
+      "step": 35790
+    },
+    {
+      "epoch": 13.514533786334466,
+      "grad_norm": 0.6572074890136719,
+      "learning_rate": 2.9854373160979552e-06,
+      "loss": 0.3506,
+      "step": 35800
+    },
+    {
+      "epoch": 13.51830879577199,
+      "grad_norm": 0.6802811622619629,
+      "learning_rate": 2.9713835589290362e-06,
+      "loss": 0.3496,
+      "step": 35810
+    },
+    {
+      "epoch": 13.522083805209514,
+      "grad_norm": 0.6372085213661194,
+      "learning_rate": 2.9573619453759913e-06,
+      "loss": 0.3557,
+      "step": 35820
+    },
+    {
+      "epoch": 13.525858814647037,
+      "grad_norm": 0.6938003897666931,
+      "learning_rate": 2.943372485022489e-06,
+      "loss": 0.3472,
+      "step": 35830
+    },
+    {
+      "epoch": 13.52963382408456,
+      "grad_norm": 0.6606130599975586,
+      "learning_rate": 2.9294151874301623e-06,
+      "loss": 0.3722,
+      "step": 35840
+    },
+    {
+      "epoch": 13.533408833522083,
+      "grad_norm": 0.6342771649360657,
+      "learning_rate": 2.9154900621387093e-06,
+      "loss": 0.355,
+      "step": 35850
+    },
+    {
+      "epoch": 13.537183842959607,
+      "grad_norm": 0.691672682762146,
+      "learning_rate": 2.9015971186658164e-06,
+      "loss": 0.3586,
+      "step": 35860
+    },
+    {
+      "epoch": 13.54095885239713,
+      "grad_norm": 0.5848689675331116,
+      "learning_rate": 2.887736366507182e-06,
+      "loss": 0.3591,
+      "step": 35870
+    },
+    {
+      "epoch": 13.544733861834654,
+      "grad_norm": 0.5893767476081848,
+      "learning_rate": 2.873907815136501e-06,
+      "loss": 0.3476,
+      "step": 35880
+    },
+    {
+      "epoch": 13.548508871272178,
+      "grad_norm": 0.6813375949859619,
+      "learning_rate": 2.8601114740054556e-06,
+      "loss": 0.3556,
+      "step": 35890
+    },
+    {
+      "epoch": 13.552283880709702,
+      "grad_norm": 0.629166841506958,
+      "learning_rate": 2.8463473525437155e-06,
+      "loss": 0.357,
+      "step": 35900
+    },
+    {
+      "epoch": 13.556058890147225,
+      "grad_norm": 0.6545192003250122,
+      "learning_rate": 2.8326154601589216e-06,
+      "loss": 0.3553,
+      "step": 35910
+    },
+    {
+      "epoch": 13.55983389958475,
+      "grad_norm": 0.639487087726593,
+      "learning_rate": 2.818915806236705e-06,
+      "loss": 0.3558,
+      "step": 35920
+    },
+    {
+      "epoch": 13.563608909022273,
+      "grad_norm": 0.6719843149185181,
+      "learning_rate": 2.8052484001406478e-06,
+      "loss": 0.3527,
+      "step": 35930
+    },
+    {
+      "epoch": 13.567383918459797,
+      "grad_norm": 0.598436713218689,
+      "learning_rate": 2.791613251212277e-06,
+      "loss": 0.3582,
+      "step": 35940
+    },
+    {
+      "epoch": 13.57115892789732,
+      "grad_norm": 0.6548103094100952,
+      "learning_rate": 2.7780103687711166e-06,
+      "loss": 0.3549,
+      "step": 35950
+    },
+    {
+      "epoch": 13.574933937334844,
+      "grad_norm": 0.6896668076515198,
+      "learning_rate": 2.764439762114579e-06,
+      "loss": 0.3508,
+      "step": 35960
+    },
+    {
+      "epoch": 13.578708946772366,
+      "grad_norm": 0.6703106164932251,
+      "learning_rate": 2.750901440518061e-06,
+      "loss": 0.3556,
+      "step": 35970
+    },
+    {
+      "epoch": 13.58248395620989,
+      "grad_norm": 0.664481520652771,
+      "learning_rate": 2.7373954132348776e-06,
+      "loss": 0.356,
+      "step": 35980
+    },
+    {
+      "epoch": 13.586258965647414,
+      "grad_norm": 0.6433864831924438,
+      "learning_rate": 2.7239216894962784e-06,
+      "loss": 0.3526,
+      "step": 35990
+    },
+    {
+      "epoch": 13.590033975084937,
+      "grad_norm": 0.6484426259994507,
+      "learning_rate": 2.710480278511407e-06,
+      "loss": 0.3545,
+      "step": 36000
+    },
+    {
+      "epoch": 13.593808984522461,
+      "grad_norm": 5.11317777633667,
+      "learning_rate": 2.6970711894673483e-06,
+      "loss": 0.3684,
+      "step": 36010
+    },
+    {
+      "epoch": 13.597583993959985,
+      "grad_norm": 0.7673658132553101,
+      "learning_rate": 2.6836944315291047e-06,
+      "loss": 0.3563,
+      "step": 36020
+    },
+    {
+      "epoch": 13.601359003397508,
+      "grad_norm": 0.6486364603042603,
+      "learning_rate": 2.67035001383954e-06,
+      "loss": 0.3549,
+      "step": 36030
+    },
+    {
+      "epoch": 13.605134012835032,
+      "grad_norm": 0.5933120250701904,
+      "learning_rate": 2.6570379455194527e-06,
+      "loss": 0.3573,
+      "step": 36040
+    },
+    {
+      "epoch": 13.608909022272556,
+      "grad_norm": 0.7512244582176208,
+      "learning_rate": 2.6437582356675206e-06,
+      "loss": 0.3507,
+      "step": 36050
+    },
+    {
+      "epoch": 13.61268403171008,
+      "grad_norm": 0.7129798531532288,
+      "learning_rate": 2.6305108933602775e-06,
+      "loss": 0.362,
+      "step": 36060
+    },
+    {
+      "epoch": 13.616459041147603,
+      "grad_norm": 0.5950183868408203,
+      "learning_rate": 2.6172959276521803e-06,
+      "loss": 0.3546,
+      "step": 36070
+    },
+    {
+      "epoch": 13.620234050585127,
+      "grad_norm": 0.6639828085899353,
+      "learning_rate": 2.6041133475755153e-06,
+      "loss": 0.3559,
+      "step": 36080
+    },
+    {
+      "epoch": 13.62400906002265,
+      "grad_norm": 0.654403567314148,
+      "learning_rate": 2.590963162140475e-06,
+      "loss": 0.3544,
+      "step": 36090
+    },
+    {
+      "epoch": 13.627784069460173,
+      "grad_norm": 0.6647920608520508,
+      "learning_rate": 2.5778453803350587e-06,
+      "loss": 0.3572,
+      "step": 36100
+    },
+    {
+      "epoch": 13.631559078897697,
+      "grad_norm": 0.6501339077949524,
+      "learning_rate": 2.5647600111251723e-06,
+      "loss": 0.3574,
+      "step": 36110
+    },
+    {
+      "epoch": 13.63533408833522,
+      "grad_norm": 0.6624252200126648,
+      "learning_rate": 2.551707063454528e-06,
+      "loss": 0.3604,
+      "step": 36120
+    },
+    {
+      "epoch": 13.639109097772744,
+      "grad_norm": 1.6550086736679077,
+      "learning_rate": 2.5386865462446897e-06,
+      "loss": 0.3647,
+      "step": 36130
+    },
+    {
+      "epoch": 13.642884107210268,
+      "grad_norm": 0.6511169672012329,
+      "learning_rate": 2.52569846839506e-06,
+      "loss": 0.3546,
+      "step": 36140
+    },
+    {
+      "epoch": 13.646659116647792,
+      "grad_norm": 0.6163169741630554,
+      "learning_rate": 2.512742838782883e-06,
+      "loss": 0.3556,
+      "step": 36150
+    },
+    {
+      "epoch": 13.650434126085315,
+      "grad_norm": 0.6587780117988586,
+      "learning_rate": 2.4998196662631924e-06,
+      "loss": 0.3611,
+      "step": 36160
+    },
+    {
+      "epoch": 13.654209135522839,
+      "grad_norm": 0.6356183290481567,
+      "learning_rate": 2.4869289596688604e-06,
+      "loss": 0.35,
+      "step": 36170
+    },
+    {
+      "epoch": 13.657984144960363,
+      "grad_norm": 0.6465098857879639,
+      "learning_rate": 2.4740707278105567e-06,
+      "loss": 0.3539,
+      "step": 36180
+    },
+    {
+      "epoch": 13.661759154397886,
+      "grad_norm": 0.6922428011894226,
+      "learning_rate": 2.461244979476779e-06,
+      "loss": 0.3567,
+      "step": 36190
+    },
+    {
+      "epoch": 13.66553416383541,
+      "grad_norm": 0.6384284496307373,
+      "learning_rate": 2.4484517234337866e-06,
+      "loss": 0.3517,
+      "step": 36200
+    },
+    {
+      "epoch": 13.669309173272934,
+      "grad_norm": 0.6639804244041443,
+      "learning_rate": 2.4356909684256634e-06,
+      "loss": 0.3593,
+      "step": 36210
+    },
+    {
+      "epoch": 13.673084182710458,
+      "grad_norm": 0.6959599852561951,
+      "learning_rate": 2.4229627231742545e-06,
+      "loss": 0.3552,
+      "step": 36220
+    },
+    {
+      "epoch": 13.67685919214798,
+      "grad_norm": 0.6804720163345337,
+      "learning_rate": 2.4102669963791956e-06,
+      "loss": 0.3578,
+      "step": 36230
+    },
+    {
+      "epoch": 13.680634201585503,
+      "grad_norm": 0.6466209888458252,
+      "learning_rate": 2.3976037967179065e-06,
+      "loss": 0.3557,
+      "step": 36240
+    },
+    {
+      "epoch": 13.684409211023027,
+      "grad_norm": 0.6484111547470093,
+      "learning_rate": 2.3849731328455582e-06,
+      "loss": 0.3582,
+      "step": 36250
+    },
+    {
+      "epoch": 13.68818422046055,
+      "grad_norm": 0.637500524520874,
+      "learning_rate": 2.3723750133950894e-06,
+      "loss": 0.3571,
+      "step": 36260
+    },
+    {
+      "epoch": 13.691959229898075,
+      "grad_norm": 3.0127627849578857,
+      "learning_rate": 2.3598094469771957e-06,
+      "loss": 0.3572,
+      "step": 36270
+    },
+    {
+      "epoch": 13.695734239335598,
+      "grad_norm": 0.6502493619918823,
+      "learning_rate": 2.347276442180324e-06,
+      "loss": 0.3556,
+      "step": 36280
+    },
+    {
+      "epoch": 13.699509248773122,
+      "grad_norm": 0.6446804404258728,
+      "learning_rate": 2.3347760075706715e-06,
+      "loss": 0.3613,
+      "step": 36290
+    },
+    {
+      "epoch": 13.703284258210646,
+      "grad_norm": 0.6583272218704224,
+      "learning_rate": 2.3223081516921595e-06,
+      "loss": 0.3598,
+      "step": 36300
+    },
+    {
+      "epoch": 13.70705926764817,
+      "grad_norm": 0.5799362063407898,
+      "learning_rate": 2.3098728830664494e-06,
+      "loss": 0.3619,
+      "step": 36310
+    },
+    {
+      "epoch": 13.710834277085693,
+      "grad_norm": 0.7087305188179016,
+      "learning_rate": 2.2974702101929314e-06,
+      "loss": 0.3569,
+      "step": 36320
+    },
+    {
+      "epoch": 13.714609286523217,
+      "grad_norm": 0.6843621730804443,
+      "learning_rate": 2.2851001415487185e-06,
+      "loss": 0.3589,
+      "step": 36330
+    },
+    {
+      "epoch": 13.71838429596074,
+      "grad_norm": 0.71535325050354,
+      "learning_rate": 2.272762685588631e-06,
+      "loss": 0.3547,
+      "step": 36340
+    },
+    {
+      "epoch": 13.722159305398263,
+      "grad_norm": 0.6820167303085327,
+      "learning_rate": 2.260457850745218e-06,
+      "loss": 0.3498,
+      "step": 36350
+    },
+    {
+      "epoch": 13.725934314835786,
+      "grad_norm": 0.6012173295021057,
+      "learning_rate": 2.2481856454286963e-06,
+      "loss": 0.3575,
+      "step": 36360
+    },
+    {
+      "epoch": 13.72970932427331,
+      "grad_norm": 0.687364399433136,
+      "learning_rate": 2.235946078027018e-06,
+      "loss": 0.3519,
+      "step": 36370
+    },
+    {
+      "epoch": 13.733484333710834,
+      "grad_norm": 0.6643616557121277,
+      "learning_rate": 2.223739156905802e-06,
+      "loss": 0.3559,
+      "step": 36380
+    },
+    {
+      "epoch": 13.737259343148358,
+      "grad_norm": 0.6354239583015442,
+      "learning_rate": 2.2115648904083796e-06,
+      "loss": 0.3561,
+      "step": 36390
+    },
+    {
+      "epoch": 13.741034352585881,
+      "grad_norm": 0.6376845836639404,
+      "learning_rate": 2.199423286855734e-06,
+      "loss": 0.3633,
+      "step": 36400
+    },
+    {
+      "epoch": 13.744809362023405,
+      "grad_norm": 0.6230825185775757,
+      "learning_rate": 2.1873143545465323e-06,
+      "loss": 0.3594,
+      "step": 36410
+    },
+    {
+      "epoch": 13.748584371460929,
+      "grad_norm": 0.644271969795227,
+      "learning_rate": 2.1752381017571267e-06,
+      "loss": 0.3556,
+      "step": 36420
+    },
+    {
+      "epoch": 13.752359380898453,
+      "grad_norm": 0.6420625448226929,
+      "learning_rate": 2.1631945367415142e-06,
+      "loss": 0.3586,
+      "step": 36430
+    },
+    {
+      "epoch": 13.756134390335976,
+      "grad_norm": 0.8240686655044556,
+      "learning_rate": 2.15118366773136e-06,
+      "loss": 0.3611,
+      "step": 36440
+    },
+    {
+      "epoch": 13.7599093997735,
+      "grad_norm": 0.6971853971481323,
+      "learning_rate": 2.1392055029359926e-06,
+      "loss": 0.3599,
+      "step": 36450
+    },
+    {
+      "epoch": 13.763684409211024,
+      "grad_norm": 0.6388399004936218,
+      "learning_rate": 2.127260050542357e-06,
+      "loss": 0.3603,
+      "step": 36460
+    },
+    {
+      "epoch": 13.767459418648546,
+      "grad_norm": 0.6157034039497375,
+      "learning_rate": 2.1153473187150608e-06,
+      "loss": 0.3599,
+      "step": 36470
+    },
+    {
+      "epoch": 13.77123442808607,
+      "grad_norm": 0.6063554286956787,
+      "learning_rate": 2.1034673155963467e-06,
+      "loss": 0.3592,
+      "step": 36480
+    },
+    {
+      "epoch": 13.775009437523593,
+      "grad_norm": 0.6081372499465942,
+      "learning_rate": 2.0916200493060976e-06,
+      "loss": 0.3581,
+      "step": 36490
+    },
+    {
+      "epoch": 13.778784446961117,
+      "grad_norm": 0.6699353456497192,
+      "learning_rate": 2.0798055279417916e-06,
+      "loss": 0.3602,
+      "step": 36500
+    },
+    {
+      "epoch": 13.78255945639864,
+      "grad_norm": 0.6695933938026428,
+      "learning_rate": 2.0680237595785356e-06,
+      "loss": 0.3552,
+      "step": 36510
+    },
+    {
+      "epoch": 13.786334465836164,
+      "grad_norm": 0.5835472941398621,
+      "learning_rate": 2.0562747522690774e-06,
+      "loss": 0.3573,
+      "step": 36520
+    },
+    {
+      "epoch": 13.790109475273688,
+      "grad_norm": 0.7515608072280884,
+      "learning_rate": 2.0445585140437373e-06,
+      "loss": 0.3574,
+      "step": 36530
+    },
+    {
+      "epoch": 13.793884484711212,
+      "grad_norm": 0.6449537873268127,
+      "learning_rate": 2.0328750529104602e-06,
+      "loss": 0.3573,
+      "step": 36540
+    },
+    {
+      "epoch": 13.797659494148736,
+      "grad_norm": 0.6437288522720337,
+      "learning_rate": 2.021224376854791e-06,
+      "loss": 0.3579,
+      "step": 36550
+    },
+    {
+      "epoch": 13.80143450358626,
+      "grad_norm": 0.6661025285720825,
+      "learning_rate": 2.0096064938398386e-06,
+      "loss": 0.3663,
+      "step": 36560
+    },
+    {
+      "epoch": 13.805209513023783,
+      "grad_norm": 0.6766248941421509,
+      "learning_rate": 1.998021411806328e-06,
+      "loss": 0.3595,
+      "step": 36570
+    },
+    {
+      "epoch": 13.808984522461307,
+      "grad_norm": 0.6111851930618286,
+      "learning_rate": 1.9864691386725586e-06,
+      "loss": 0.3578,
+      "step": 36580
+    },
+    {
+      "epoch": 13.81275953189883,
+      "grad_norm": 0.7111583948135376,
+      "learning_rate": 1.9749496823343814e-06,
+      "loss": 0.3598,
+      "step": 36590
+    },
+    {
+      "epoch": 13.816534541336353,
+      "grad_norm": 0.7276173830032349,
+      "learning_rate": 1.963463050665254e-06,
+      "loss": 0.349,
+      "step": 36600
+    },
+    {
+      "epoch": 13.820309550773876,
+      "grad_norm": 0.6221917271614075,
+      "learning_rate": 1.9520092515161682e-06,
+      "loss": 0.3696,
+      "step": 36610
+    },
+    {
+      "epoch": 13.8240845602114,
+      "grad_norm": 0.7171021699905396,
+      "learning_rate": 1.940588292715695e-06,
+      "loss": 0.355,
+      "step": 36620
+    },
+    {
+      "epoch": 13.827859569648924,
+      "grad_norm": 0.7668521404266357,
+      "learning_rate": 1.929200182069951e-06,
+      "loss": 0.3491,
+      "step": 36630
+    },
+    {
+      "epoch": 13.831634579086447,
+      "grad_norm": 0.9479874968528748,
+      "learning_rate": 1.9178449273625997e-06,
+      "loss": 0.3594,
+      "step": 36640
+    },
+    {
+      "epoch": 13.835409588523971,
+      "grad_norm": 0.6809915900230408,
+      "learning_rate": 1.90652253635486e-06,
+      "loss": 0.3518,
+      "step": 36650
+    },
+    {
+      "epoch": 13.839184597961495,
+      "grad_norm": 0.6257880330085754,
+      "learning_rate": 1.8952330167854693e-06,
+      "loss": 0.358,
+      "step": 36660
+    },
+    {
+      "epoch": 13.842959607399019,
+      "grad_norm": 0.6460468769073486,
+      "learning_rate": 1.8839763763707107e-06,
+      "loss": 0.3555,
+      "step": 36670
+    },
+    {
+      "epoch": 13.846734616836542,
+      "grad_norm": 0.7003082036972046,
+      "learning_rate": 1.8727526228043912e-06,
+      "loss": 0.3521,
+      "step": 36680
+    },
+    {
+      "epoch": 13.850509626274066,
+      "grad_norm": 0.6699946522712708,
+      "learning_rate": 1.8615617637578408e-06,
+      "loss": 0.3566,
+      "step": 36690
+    },
+    {
+      "epoch": 13.85428463571159,
+      "grad_norm": 0.6396785378456116,
+      "learning_rate": 1.850403806879908e-06,
+      "loss": 0.3517,
+      "step": 36700
+    },
+    {
+      "epoch": 13.858059645149114,
+      "grad_norm": 0.6421851515769958,
+      "learning_rate": 1.8392787597969475e-06,
+      "loss": 0.3532,
+      "step": 36710
+    },
+    {
+      "epoch": 13.861834654586637,
+      "grad_norm": 0.6195537447929382,
+      "learning_rate": 1.828186630112827e-06,
+      "loss": 0.3673,
+      "step": 36720
+    },
+    {
+      "epoch": 13.86560966402416,
+      "grad_norm": 0.5870434045791626,
+      "learning_rate": 1.8171274254089098e-06,
+      "loss": 0.3594,
+      "step": 36730
+    },
+    {
+      "epoch": 13.869384673461683,
+      "grad_norm": 0.7309495806694031,
+      "learning_rate": 1.806101153244061e-06,
+      "loss": 0.36,
+      "step": 36740
+    },
+    {
+      "epoch": 13.873159682899207,
+      "grad_norm": 0.6230713129043579,
+      "learning_rate": 1.7951078211546401e-06,
+      "loss": 0.3546,
+      "step": 36750
+    },
+    {
+      "epoch": 13.87693469233673,
+      "grad_norm": 0.6043184399604797,
+      "learning_rate": 1.7841474366544764e-06,
+      "loss": 0.3515,
+      "step": 36760
+    },
+    {
+      "epoch": 13.880709701774254,
+      "grad_norm": 0.6820389032363892,
+      "learning_rate": 1.7732200072348825e-06,
+      "loss": 0.3657,
+      "step": 36770
+    },
+    {
+      "epoch": 13.884484711211778,
+      "grad_norm": 0.6838846206665039,
+      "learning_rate": 1.7623255403646677e-06,
+      "loss": 0.3519,
+      "step": 36780
+    },
+    {
+      "epoch": 13.888259720649302,
+      "grad_norm": 0.6790589094161987,
+      "learning_rate": 1.7514640434900865e-06,
+      "loss": 0.3544,
+      "step": 36790
+    },
+    {
+      "epoch": 13.892034730086825,
+      "grad_norm": 0.6140289902687073,
+      "learning_rate": 1.7406355240348838e-06,
+      "loss": 0.353,
+      "step": 36800
+    },
+    {
+      "epoch": 13.89580973952435,
+      "grad_norm": 0.6230329871177673,
+      "learning_rate": 1.7298399894002392e-06,
+      "loss": 0.3547,
+      "step": 36810
+    },
+    {
+      "epoch": 13.899584748961873,
+      "grad_norm": 0.674197793006897,
+      "learning_rate": 1.7190774469647952e-06,
+      "loss": 0.3633,
+      "step": 36820
+    },
+    {
+      "epoch": 13.903359758399397,
+      "grad_norm": 0.8990030884742737,
+      "learning_rate": 1.7083479040846617e-06,
+      "loss": 0.3618,
+      "step": 36830
+    },
+    {
+      "epoch": 13.90713476783692,
+      "grad_norm": 0.6309157609939575,
+      "learning_rate": 1.697651368093367e-06,
+      "loss": 0.3521,
+      "step": 36840
+    },
+    {
+      "epoch": 13.910909777274444,
+      "grad_norm": 0.6408490538597107,
+      "learning_rate": 1.6869878463019017e-06,
+      "loss": 0.357,
+      "step": 36850
+    },
+    {
+      "epoch": 13.914684786711966,
+      "grad_norm": 0.6487820148468018,
+      "learning_rate": 1.6763573459986858e-06,
+      "loss": 0.3681,
+      "step": 36860
+    },
+    {
+      "epoch": 13.91845979614949,
+      "grad_norm": 0.663537859916687,
+      "learning_rate": 1.6657598744495516e-06,
+      "loss": 0.3592,
+      "step": 36870
+    },
+    {
+      "epoch": 13.922234805587014,
+      "grad_norm": 0.629589319229126,
+      "learning_rate": 1.6551954388977774e-06,
+      "loss": 0.3481,
+      "step": 36880
+    },
+    {
+      "epoch": 13.926009815024537,
+      "grad_norm": 0.6090066432952881,
+      "learning_rate": 1.644664046564054e-06,
+      "loss": 0.3581,
+      "step": 36890
+    },
+    {
+      "epoch": 13.929784824462061,
+      "grad_norm": 0.765868067741394,
+      "learning_rate": 1.634165704646501e-06,
+      "loss": 0.3518,
+      "step": 36900
+    },
+    {
+      "epoch": 13.933559833899585,
+      "grad_norm": 0.6522102355957031,
+      "learning_rate": 1.6237004203206175e-06,
+      "loss": 0.3544,
+      "step": 36910
+    },
+    {
+      "epoch": 13.937334843337108,
+      "grad_norm": 0.6589536666870117,
+      "learning_rate": 1.6132682007393373e-06,
+      "loss": 0.3544,
+      "step": 36920
+    },
+    {
+      "epoch": 13.941109852774632,
+      "grad_norm": 0.6735085248947144,
+      "learning_rate": 1.6028690530329847e-06,
+      "loss": 0.3427,
+      "step": 36930
+    },
+    {
+      "epoch": 13.944884862212156,
+      "grad_norm": 0.6307739615440369,
+      "learning_rate": 1.5925029843092742e-06,
+      "loss": 0.3557,
+      "step": 36940
+    },
+    {
+      "epoch": 13.94865987164968,
+      "grad_norm": 0.6922662854194641,
+      "learning_rate": 1.5821700016533214e-06,
+      "loss": 0.3629,
+      "step": 36950
+    },
+    {
+      "epoch": 13.952434881087203,
+      "grad_norm": 0.6589555740356445,
+      "learning_rate": 1.5718701121276268e-06,
+      "loss": 0.3487,
+      "step": 36960
+    },
+    {
+      "epoch": 13.956209890524725,
+      "grad_norm": 0.7357824444770813,
+      "learning_rate": 1.5616033227720483e-06,
+      "loss": 0.3495,
+      "step": 36970
+    },
+    {
+      "epoch": 13.95998489996225,
+      "grad_norm": 0.6229332089424133,
+      "learning_rate": 1.5513696406038558e-06,
+      "loss": 0.3577,
+      "step": 36980
+    },
+    {
+      "epoch": 13.963759909399773,
+      "grad_norm": 0.5904605388641357,
+      "learning_rate": 1.5411690726176653e-06,
+      "loss": 0.3531,
+      "step": 36990
+    },
+    {
+      "epoch": 13.967534918837297,
+      "grad_norm": 0.7504417300224304,
+      "learning_rate": 1.5310016257854832e-06,
+      "loss": 0.3604,
+      "step": 37000
+    },
+    {
+      "epoch": 13.97130992827482,
+      "grad_norm": 0.6638484001159668,
+      "learning_rate": 1.5208673070566504e-06,
+      "loss": 0.3546,
+      "step": 37010
+    },
+    {
+      "epoch": 13.975084937712344,
+      "grad_norm": 0.6932189464569092,
+      "learning_rate": 1.5107661233578818e-06,
+      "loss": 0.353,
+      "step": 37020
+    },
+    {
+      "epoch": 13.978859947149868,
+      "grad_norm": 0.871865451335907,
+      "learning_rate": 1.5006980815932437e-06,
+      "loss": 0.3684,
+      "step": 37030
+    },
+    {
+      "epoch": 13.982634956587392,
+      "grad_norm": 0.6513112783432007,
+      "learning_rate": 1.490663188644148e-06,
+      "loss": 0.3602,
+      "step": 37040
+    },
+    {
+      "epoch": 13.986409966024915,
+      "grad_norm": 0.6713985204696655,
+      "learning_rate": 1.4806614513693418e-06,
+      "loss": 0.3541,
+      "step": 37050
+    },
+    {
+      "epoch": 13.990184975462439,
+      "grad_norm": 0.681988537311554,
+      "learning_rate": 1.4706928766049289e-06,
+      "loss": 0.3581,
+      "step": 37060
+    },
+    {
+      "epoch": 13.993959984899963,
+      "grad_norm": 0.948577344417572,
+      "learning_rate": 1.4607574711643257e-06,
+      "loss": 0.3532,
+      "step": 37070
+    },
+    {
+      "epoch": 13.997734994337486,
+      "grad_norm": 0.7015256881713867,
+      "learning_rate": 1.4508552418382949e-06,
+      "loss": 0.355,
+      "step": 37080
+    },
+    {
+      "epoch": 14.00151000377501,
+      "grad_norm": 0.6600831151008606,
+      "learning_rate": 1.4409861953949112e-06,
+      "loss": 0.3522,
+      "step": 37090
+    },
+    {
+      "epoch": 14.005285013212532,
+      "grad_norm": 0.6127675771713257,
+      "learning_rate": 1.43115033857959e-06,
+      "loss": 0.3566,
+      "step": 37100
+    },
+    {
+      "epoch": 14.009060022650056,
+      "grad_norm": 0.6839843988418579,
+      "learning_rate": 1.4213476781150258e-06,
+      "loss": 0.3503,
+      "step": 37110
+    },
+    {
+      "epoch": 14.01283503208758,
+      "grad_norm": 0.6490364074707031,
+      "learning_rate": 1.411578220701254e-06,
+      "loss": 0.3628,
+      "step": 37120
+    },
+    {
+      "epoch": 14.016610041525103,
+      "grad_norm": 0.6181484460830688,
+      "learning_rate": 1.4018419730156108e-06,
+      "loss": 0.3498,
+      "step": 37130
+    },
+    {
+      "epoch": 14.020385050962627,
+      "grad_norm": 0.638804018497467,
+      "learning_rate": 1.3921389417127283e-06,
+      "loss": 0.3595,
+      "step": 37140
+    },
+    {
+      "epoch": 14.02416006040015,
+      "grad_norm": 0.6494945287704468,
+      "learning_rate": 1.3824691334245298e-06,
+      "loss": 0.36,
+      "step": 37150
+    },
+    {
+      "epoch": 14.027935069837675,
+      "grad_norm": 0.6143277287483215,
+      "learning_rate": 1.3728325547602505e-06,
+      "loss": 0.3568,
+      "step": 37160
+    },
+    {
+      "epoch": 14.031710079275198,
+      "grad_norm": 0.6732072234153748,
+      "learning_rate": 1.3632292123063828e-06,
+      "loss": 0.3524,
+      "step": 37170
+    },
+    {
+      "epoch": 14.035485088712722,
+      "grad_norm": 0.6264946460723877,
+      "learning_rate": 1.3536591126267317e-06,
+      "loss": 0.3535,
+      "step": 37180
+    },
+    {
+      "epoch": 14.039260098150246,
+      "grad_norm": 0.6948634386062622,
+      "learning_rate": 1.3441222622623761e-06,
+      "loss": 0.3541,
+      "step": 37190
+    },
+    {
+      "epoch": 14.04303510758777,
+      "grad_norm": 0.6478399038314819,
+      "learning_rate": 1.3346186677316575e-06,
+      "loss": 0.3537,
+      "step": 37200
+    },
+    {
+      "epoch": 14.046810117025293,
+      "grad_norm": 0.727102518081665,
+      "learning_rate": 1.3251483355301852e-06,
+      "loss": 0.3499,
+      "step": 37210
+    },
+    {
+      "epoch": 14.050585126462817,
+      "grad_norm": 0.5733349919319153,
+      "learning_rate": 1.315711272130854e-06,
+      "loss": 0.3598,
+      "step": 37220
+    },
+    {
+      "epoch": 14.054360135900339,
+      "grad_norm": 0.6830695867538452,
+      "learning_rate": 1.306307483983804e-06,
+      "loss": 0.3582,
+      "step": 37230
+    },
+    {
+      "epoch": 14.058135145337863,
+      "grad_norm": 0.6603965163230896,
+      "learning_rate": 1.2969369775164275e-06,
+      "loss": 0.3528,
+      "step": 37240
+    },
+    {
+      "epoch": 14.061910154775386,
+      "grad_norm": 0.672484278678894,
+      "learning_rate": 1.28759975913339e-06,
+      "loss": 0.353,
+      "step": 37250
+    },
+    {
+      "epoch": 14.06568516421291,
+      "grad_norm": 0.6874468922615051,
+      "learning_rate": 1.2782958352165863e-06,
+      "loss": 0.3559,
+      "step": 37260
+    },
+    {
+      "epoch": 14.069460173650434,
+      "grad_norm": 0.6835435032844543,
+      "learning_rate": 1.2690252121251633e-06,
+      "loss": 0.359,
+      "step": 37270
+    },
+    {
+      "epoch": 14.073235183087958,
+      "grad_norm": 0.6037148237228394,
+      "learning_rate": 1.2597878961954968e-06,
+      "loss": 0.3595,
+      "step": 37280
+    },
+    {
+      "epoch": 14.077010192525481,
+      "grad_norm": 0.7436287999153137,
+      "learning_rate": 1.250583893741214e-06,
+      "loss": 0.3606,
+      "step": 37290
+    },
+    {
+      "epoch": 14.080785201963005,
+      "grad_norm": 0.6740867495536804,
+      "learning_rate": 1.2414132110531607e-06,
+      "loss": 0.3607,
+      "step": 37300
+    },
+    {
+      "epoch": 14.084560211400529,
+      "grad_norm": 0.7421841025352478,
+      "learning_rate": 1.2322758543994062e-06,
+      "loss": 0.3554,
+      "step": 37310
+    },
+    {
+      "epoch": 14.088335220838053,
+      "grad_norm": 0.7035681009292603,
+      "learning_rate": 1.2231718300252603e-06,
+      "loss": 0.3539,
+      "step": 37320
+    },
+    {
+      "epoch": 14.092110230275576,
+      "grad_norm": 0.6054702401161194,
+      "learning_rate": 1.214101144153229e-06,
+      "loss": 0.3691,
+      "step": 37330
+    },
+    {
+      "epoch": 14.0958852397131,
+      "grad_norm": 0.6276698112487793,
+      "learning_rate": 1.205063802983042e-06,
+      "loss": 0.3569,
+      "step": 37340
+    },
+    {
+      "epoch": 14.099660249150624,
+      "grad_norm": 0.652378261089325,
+      "learning_rate": 1.1960598126916301e-06,
+      "loss": 0.3559,
+      "step": 37350
+    },
+    {
+      "epoch": 14.103435258588146,
+      "grad_norm": 0.6906107664108276,
+      "learning_rate": 1.187089179433154e-06,
+      "loss": 0.3614,
+      "step": 37360
+    },
+    {
+      "epoch": 14.10721026802567,
+      "grad_norm": 0.6413514018058777,
+      "learning_rate": 1.178151909338937e-06,
+      "loss": 0.3563,
+      "step": 37370
+    },
+    {
+      "epoch": 14.110985277463193,
+      "grad_norm": 0.6062707901000977,
+      "learning_rate": 1.1692480085175317e-06,
+      "loss": 0.3571,
+      "step": 37380
+    },
+    {
+      "epoch": 14.114760286900717,
+      "grad_norm": 0.6939738392829895,
+      "learning_rate": 1.1603774830546587e-06,
+      "loss": 0.3524,
+      "step": 37390
+    },
+    {
+      "epoch": 14.11853529633824,
+      "grad_norm": 0.6625202894210815,
+      "learning_rate": 1.1515403390132573e-06,
+      "loss": 0.3564,
+      "step": 37400
+    },
+    {
+      "epoch": 14.122310305775764,
+      "grad_norm": 0.7371194362640381,
+      "learning_rate": 1.1427365824334124e-06,
+      "loss": 0.3535,
+      "step": 37410
+    },
+    {
+      "epoch": 14.126085315213288,
+      "grad_norm": 0.5587255954742432,
+      "learning_rate": 1.133966219332422e-06,
+      "loss": 0.3549,
+      "step": 37420
+    },
+    {
+      "epoch": 14.129860324650812,
+      "grad_norm": 0.6779747605323792,
+      "learning_rate": 1.1252292557047405e-06,
+      "loss": 0.3618,
+      "step": 37430
+    },
+    {
+      "epoch": 14.133635334088336,
+      "grad_norm": 0.6948127150535583,
+      "learning_rate": 1.1165256975219974e-06,
+      "loss": 0.3502,
+      "step": 37440
+    },
+    {
+      "epoch": 14.13741034352586,
+      "grad_norm": 0.796514630317688,
+      "learning_rate": 1.1078555507330002e-06,
+      "loss": 0.3551,
+      "step": 37450
+    },
+    {
+      "epoch": 14.141185352963383,
+      "grad_norm": 0.6783875226974487,
+      "learning_rate": 1.0992188212637089e-06,
+      "loss": 0.355,
+      "step": 37460
+    },
+    {
+      "epoch": 14.144960362400907,
+      "grad_norm": 0.6731642484664917,
+      "learning_rate": 1.0906155150172458e-06,
+      "loss": 0.351,
+      "step": 37470
+    },
+    {
+      "epoch": 14.148735371838429,
+      "grad_norm": 0.670313835144043,
+      "learning_rate": 1.0820456378738852e-06,
+      "loss": 0.3541,
+      "step": 37480
+    },
+    {
+      "epoch": 14.152510381275953,
+      "grad_norm": 0.768552303314209,
+      "learning_rate": 1.0735091956910692e-06,
+      "loss": 0.3524,
+      "step": 37490
+    },
+    {
+      "epoch": 14.156285390713476,
+      "grad_norm": 0.6623304486274719,
+      "learning_rate": 1.0650061943033695e-06,
+      "loss": 0.3575,
+      "step": 37500
+    },
+    {
+      "epoch": 14.160060400151,
+      "grad_norm": 0.632932186126709,
+      "learning_rate": 1.0565366395225041e-06,
+      "loss": 0.3482,
+      "step": 37510
+    },
+    {
+      "epoch": 14.163835409588524,
+      "grad_norm": 0.633401095867157,
+      "learning_rate": 1.0481005371373366e-06,
+      "loss": 0.3536,
+      "step": 37520
+    },
+    {
+      "epoch": 14.167610419026047,
+      "grad_norm": 0.6039905548095703,
+      "learning_rate": 1.0396978929138602e-06,
+      "loss": 0.3505,
+      "step": 37530
+    },
+    {
+      "epoch": 14.171385428463571,
+      "grad_norm": 0.6414741277694702,
+      "learning_rate": 1.0313287125952032e-06,
+      "loss": 0.3634,
+      "step": 37540
+    },
+    {
+      "epoch": 14.175160437901095,
+      "grad_norm": 0.6331318020820618,
+      "learning_rate": 1.0229930019016288e-06,
+      "loss": 0.3648,
+      "step": 37550
+    },
+    {
+      "epoch": 14.178935447338619,
+      "grad_norm": 0.6243364214897156,
+      "learning_rate": 1.0146907665305183e-06,
+      "loss": 0.3647,
+      "step": 37560
+    },
+    {
+      "epoch": 14.182710456776142,
+      "grad_norm": 0.6351826786994934,
+      "learning_rate": 1.0064220121563551e-06,
+      "loss": 0.358,
+      "step": 37570
+    },
+    {
+      "epoch": 14.186485466213666,
+      "grad_norm": 0.6927258372306824,
+      "learning_rate": 9.98186744430768e-07,
+      "loss": 0.3517,
+      "step": 37580
+    },
+    {
+      "epoch": 14.19026047565119,
+      "grad_norm": 0.6401153206825256,
+      "learning_rate": 9.899849689824826e-07,
+      "loss": 0.3611,
+      "step": 37590
+    },
+    {
+      "epoch": 14.194035485088712,
+      "grad_norm": 0.7551748752593994,
+      "learning_rate": 9.818166914173365e-07,
+      "loss": 0.3524,
+      "step": 37600
+    },
+    {
+      "epoch": 14.197810494526236,
+      "grad_norm": 0.6238972544670105,
+      "learning_rate": 9.736819173182642e-07,
+      "loss": 0.3578,
+      "step": 37610
+    },
+    {
+      "epoch": 14.20158550396376,
+      "grad_norm": 0.795691728591919,
+      "learning_rate": 9.655806522453126e-07,
+      "loss": 0.3544,
+      "step": 37620
+    },
+    {
+      "epoch": 14.205360513401283,
+      "grad_norm": 0.558307945728302,
+      "learning_rate": 9.575129017356187e-07,
+      "loss": 0.3547,
+      "step": 37630
+    },
+    {
+      "epoch": 14.209135522838807,
+      "grad_norm": 0.6478992104530334,
+      "learning_rate": 9.494786713034109e-07,
+      "loss": 0.3494,
+      "step": 37640
+    },
+    {
+      "epoch": 14.21291053227633,
+      "grad_norm": 0.646564245223999,
+      "learning_rate": 9.414779664400075e-07,
+      "loss": 0.3572,
+      "step": 37650
+    },
+    {
+      "epoch": 14.216685541713854,
+      "grad_norm": 0.6614131927490234,
+      "learning_rate": 9.335107926138342e-07,
+      "loss": 0.3513,
+      "step": 37660
+    },
+    {
+      "epoch": 14.220460551151378,
+      "grad_norm": 0.7705824375152588,
+      "learning_rate": 9.255771552703519e-07,
+      "loss": 0.3504,
+      "step": 37670
+    },
+    {
+      "epoch": 14.224235560588902,
+      "grad_norm": 0.6299923658370972,
+      "learning_rate": 9.176770598321449e-07,
+      "loss": 0.3535,
+      "step": 37680
+    },
+    {
+      "epoch": 14.228010570026425,
+      "grad_norm": 0.6774992346763611,
+      "learning_rate": 9.098105116988498e-07,
+      "loss": 0.3546,
+      "step": 37690
+    },
+    {
+      "epoch": 14.23178557946395,
+      "grad_norm": 0.6203901171684265,
+      "learning_rate": 9.019775162471767e-07,
+      "loss": 0.3477,
+      "step": 37700
+    },
+    {
+      "epoch": 14.235560588901473,
+      "grad_norm": 0.590056300163269,
+      "learning_rate": 8.94178078830904e-07,
+      "loss": 0.3571,
+      "step": 37710
+    },
+    {
+      "epoch": 14.239335598338997,
+      "grad_norm": 0.6835026741027832,
+      "learning_rate": 8.864122047808731e-07,
+      "loss": 0.362,
+      "step": 37720
+    },
+    {
+      "epoch": 14.243110607776519,
+      "grad_norm": 0.7431061863899231,
+      "learning_rate": 8.786798994049882e-07,
+      "loss": 0.3595,
+      "step": 37730
+    },
+    {
+      "epoch": 14.246885617214042,
+      "grad_norm": 0.6774893999099731,
+      "learning_rate": 8.709811679882052e-07,
+      "loss": 0.3546,
+      "step": 37740
+    },
+    {
+      "epoch": 14.250660626651566,
+      "grad_norm": 0.6696170568466187,
+      "learning_rate": 8.633160157925313e-07,
+      "loss": 0.3597,
+      "step": 37750
+    },
+    {
+      "epoch": 14.25443563608909,
+      "grad_norm": 0.592789351940155,
+      "learning_rate": 8.556844480570369e-07,
+      "loss": 0.3576,
+      "step": 37760
+    },
+    {
+      "epoch": 14.258210645526614,
+      "grad_norm": 0.6565343141555786,
+      "learning_rate": 8.480864699978164e-07,
+      "loss": 0.3634,
+      "step": 37770
+    },
+    {
+      "epoch": 14.261985654964137,
+      "grad_norm": 0.5975145697593689,
+      "learning_rate": 8.405220868080154e-07,
+      "loss": 0.3502,
+      "step": 37780
+    },
+    {
+      "epoch": 14.265760664401661,
+      "grad_norm": 0.9825413227081299,
+      "learning_rate": 8.329913036578263e-07,
+      "loss": 0.3514,
+      "step": 37790
+    },
+    {
+      "epoch": 14.269535673839185,
+      "grad_norm": 0.6201307773590088,
+      "learning_rate": 8.254941256944593e-07,
+      "loss": 0.358,
+      "step": 37800
+    },
+    {
+      "epoch": 14.273310683276708,
+      "grad_norm": 0.6925731897354126,
+      "learning_rate": 8.18030558042171e-07,
+      "loss": 0.3512,
+      "step": 37810
+    },
+    {
+      "epoch": 14.277085692714232,
+      "grad_norm": 0.6344226002693176,
+      "learning_rate": 8.106006058022419e-07,
+      "loss": 0.3585,
+      "step": 37820
+    },
+    {
+      "epoch": 14.280860702151756,
+      "grad_norm": 0.6603698134422302,
+      "learning_rate": 8.03204274052971e-07,
+      "loss": 0.3566,
+      "step": 37830
+    },
+    {
+      "epoch": 14.28463571158928,
+      "grad_norm": 0.7116165161132812,
+      "learning_rate": 7.958415678496756e-07,
+      "loss": 0.359,
+      "step": 37840
+    },
+    {
+      "epoch": 14.288410721026803,
+      "grad_norm": 0.6361221075057983,
+      "learning_rate": 7.885124922247022e-07,
+      "loss": 0.3539,
+      "step": 37850
+    },
+    {
+      "epoch": 14.292185730464325,
+      "grad_norm": 0.6573535203933716,
+      "learning_rate": 7.812170521874051e-07,
+      "loss": 0.3601,
+      "step": 37860
+    },
+    {
+      "epoch": 14.29596073990185,
+      "grad_norm": 0.6462699770927429,
+      "learning_rate": 7.739552527241456e-07,
+      "loss": 0.3573,
+      "step": 37870
+    },
+    {
+      "epoch": 14.299735749339373,
+      "grad_norm": 0.587371826171875,
+      "learning_rate": 7.667270987983033e-07,
+      "loss": 0.3561,
+      "step": 37880
+    },
+    {
+      "epoch": 14.303510758776897,
+      "grad_norm": 0.6839852929115295,
+      "learning_rate": 7.595325953502375e-07,
+      "loss": 0.3578,
+      "step": 37890
+    },
+    {
+      "epoch": 14.30728576821442,
+      "grad_norm": 0.5843148231506348,
+      "learning_rate": 7.523717472973368e-07,
+      "loss": 0.3582,
+      "step": 37900
+    },
+    {
+      "epoch": 14.311060777651944,
+      "grad_norm": 0.6995810866355896,
+      "learning_rate": 7.452445595339696e-07,
+      "loss": 0.3567,
+      "step": 37910
+    },
+    {
+      "epoch": 14.314835787089468,
+      "grad_norm": 0.6964356899261475,
+      "learning_rate": 7.381510369315059e-07,
+      "loss": 0.3596,
+      "step": 37920
+    },
+    {
+      "epoch": 14.318610796526992,
+      "grad_norm": 1.5723952054977417,
+      "learning_rate": 7.310911843382895e-07,
+      "loss": 0.3567,
+      "step": 37930
+    },
+    {
+      "epoch": 14.322385805964515,
+      "grad_norm": 0.7851320505142212,
+      "learning_rate": 7.240650065796716e-07,
+      "loss": 0.3499,
+      "step": 37940
+    },
+    {
+      "epoch": 14.326160815402039,
+      "grad_norm": 0.6869962811470032,
+      "learning_rate": 7.170725084579777e-07,
+      "loss": 0.3588,
+      "step": 37950
+    },
+    {
+      "epoch": 14.329935824839563,
+      "grad_norm": 0.6377101540565491,
+      "learning_rate": 7.101136947525233e-07,
+      "loss": 0.3529,
+      "step": 37960
+    },
+    {
+      "epoch": 14.333710834277086,
+      "grad_norm": 0.713228166103363,
+      "learning_rate": 7.031885702195707e-07,
+      "loss": 0.357,
+      "step": 37970
+    },
+    {
+      "epoch": 14.337485843714608,
+      "grad_norm": 0.6053051352500916,
+      "learning_rate": 6.962971395924e-07,
+      "loss": 0.3571,
+      "step": 37980
+    },
+    {
+      "epoch": 14.341260853152132,
+      "grad_norm": 0.6226997971534729,
+      "learning_rate": 6.894394075812217e-07,
+      "loss": 0.3558,
+      "step": 37990
+    },
+    {
+      "epoch": 14.345035862589656,
+      "grad_norm": 0.6585350036621094,
+      "learning_rate": 6.826153788732359e-07,
+      "loss": 0.3541,
+      "step": 38000
+    },
+    {
+      "epoch": 14.34881087202718,
+      "grad_norm": 0.7358768582344055,
+      "learning_rate": 6.758250581326009e-07,
+      "loss": 0.3588,
+      "step": 38010
+    },
+    {
+      "epoch": 14.352585881464703,
+      "grad_norm": 0.7400320172309875,
+      "learning_rate": 6.690684500004485e-07,
+      "loss": 0.3553,
+      "step": 38020
+    },
+    {
+      "epoch": 14.356360890902227,
+      "grad_norm": 0.6965455412864685,
+      "learning_rate": 6.623455590948458e-07,
+      "loss": 0.3536,
+      "step": 38030
+    },
+    {
+      "epoch": 14.36013590033975,
+      "grad_norm": 0.6424039006233215,
+      "learning_rate": 6.556563900108226e-07,
+      "loss": 0.3559,
+      "step": 38040
+    },
+    {
+      "epoch": 14.363910909777275,
+      "grad_norm": 0.664570152759552,
+      "learning_rate": 6.490009473203662e-07,
+      "loss": 0.3596,
+      "step": 38050
+    },
+    {
+      "epoch": 14.367685919214798,
+      "grad_norm": 0.6233497262001038,
+      "learning_rate": 6.423792355724156e-07,
+      "loss": 0.3518,
+      "step": 38060
+    },
+    {
+      "epoch": 14.371460928652322,
+      "grad_norm": 0.6261218190193176,
+      "learning_rate": 6.35791259292845e-07,
+      "loss": 0.3594,
+      "step": 38070
+    },
+    {
+      "epoch": 14.375235938089846,
+      "grad_norm": 0.7287961840629578,
+      "learning_rate": 6.292370229844691e-07,
+      "loss": 0.3546,
+      "step": 38080
+    },
+    {
+      "epoch": 14.37901094752737,
+      "grad_norm": 0.5852963328361511,
+      "learning_rate": 6.227165311270488e-07,
+      "loss": 0.3556,
+      "step": 38090
+    },
+    {
+      "epoch": 14.382785956964893,
+      "grad_norm": 0.6326406002044678,
+      "learning_rate": 6.16229788177275e-07,
+      "loss": 0.3512,
+      "step": 38100
+    },
+    {
+      "epoch": 14.386560966402415,
+      "grad_norm": 0.6186873912811279,
+      "learning_rate": 6.097767985687841e-07,
+      "loss": 0.3541,
+      "step": 38110
+    },
+    {
+      "epoch": 14.390335975839939,
+      "grad_norm": 1.111286997795105,
+      "learning_rate": 6.033575667121317e-07,
+      "loss": 0.3576,
+      "step": 38120
+    },
+    {
+      "epoch": 14.394110985277463,
+      "grad_norm": 0.6550930738449097,
+      "learning_rate": 5.969720969947967e-07,
+      "loss": 0.355,
+      "step": 38130
+    },
+    {
+      "epoch": 14.397885994714986,
+      "grad_norm": 0.6559245586395264,
+      "learning_rate": 5.906203937811883e-07,
+      "loss": 0.3628,
+      "step": 38140
+    },
+    {
+      "epoch": 14.40166100415251,
+      "grad_norm": 0.6878376007080078,
+      "learning_rate": 5.843024614126447e-07,
+      "loss": 0.3554,
+      "step": 38150
+    },
+    {
+      "epoch": 14.405436013590034,
+      "grad_norm": 0.644935667514801,
+      "learning_rate": 5.780183042074061e-07,
+      "loss": 0.3574,
+      "step": 38160
+    },
+    {
+      "epoch": 14.409211023027558,
+      "grad_norm": 0.6619318723678589,
+      "learning_rate": 5.717679264606312e-07,
+      "loss": 0.3569,
+      "step": 38170
+    },
+    {
+      "epoch": 14.412986032465081,
+      "grad_norm": 0.6411536931991577,
+      "learning_rate": 5.655513324444028e-07,
+      "loss": 0.3592,
+      "step": 38180
+    },
+    {
+      "epoch": 14.416761041902605,
+      "grad_norm": 0.6617749929428101,
+      "learning_rate": 5.593685264076942e-07,
+      "loss": 0.3463,
+      "step": 38190
+    },
+    {
+      "epoch": 14.420536051340129,
+      "grad_norm": 0.6915404796600342,
+      "learning_rate": 5.532195125764029e-07,
+      "loss": 0.3586,
+      "step": 38200
+    },
+    {
+      "epoch": 14.424311060777653,
+      "grad_norm": 0.6411628723144531,
+      "learning_rate": 5.471042951533167e-07,
+      "loss": 0.3425,
+      "step": 38210
+    },
+    {
+      "epoch": 14.428086070215176,
+      "grad_norm": 0.645287275314331,
+      "learning_rate": 5.410228783181371e-07,
+      "loss": 0.3582,
+      "step": 38220
+    },
+    {
+      "epoch": 14.431861079652698,
+      "grad_norm": 0.6621134281158447,
+      "learning_rate": 5.349752662274444e-07,
+      "loss": 0.3567,
+      "step": 38230
+    },
+    {
+      "epoch": 14.435636089090222,
+      "grad_norm": 0.6822801232337952,
+      "learning_rate": 5.289614630147321e-07,
+      "loss": 0.3463,
+      "step": 38240
+    },
+    {
+      "epoch": 14.439411098527746,
+      "grad_norm": 0.6010865569114685,
+      "learning_rate": 5.229814727903737e-07,
+      "loss": 0.3603,
+      "step": 38250
+    },
+    {
+      "epoch": 14.44318610796527,
+      "grad_norm": 0.7438214421272278,
+      "learning_rate": 5.170352996416439e-07,
+      "loss": 0.3529,
+      "step": 38260
+    },
+    {
+      "epoch": 14.446961117402793,
+      "grad_norm": 0.6098829507827759,
+      "learning_rate": 5.111229476326918e-07,
+      "loss": 0.3516,
+      "step": 38270
+    },
+    {
+      "epoch": 14.450736126840317,
+      "grad_norm": 0.6266598105430603,
+      "learning_rate": 5.052444208045459e-07,
+      "loss": 0.3569,
+      "step": 38280
+    },
+    {
+      "epoch": 14.45451113627784,
+      "grad_norm": 0.6430060267448425,
+      "learning_rate": 4.993997231751312e-07,
+      "loss": 0.3495,
+      "step": 38290
+    },
+    {
+      "epoch": 14.458286145715364,
+      "grad_norm": 0.6154651045799255,
+      "learning_rate": 4.935888587392468e-07,
+      "loss": 0.3546,
+      "step": 38300
+    },
+    {
+      "epoch": 14.462061155152888,
+      "grad_norm": 0.6602094173431396,
+      "learning_rate": 4.878118314685598e-07,
+      "loss": 0.3539,
+      "step": 38310
+    },
+    {
+      "epoch": 14.465836164590412,
+      "grad_norm": 0.7772398591041565,
+      "learning_rate": 4.820686453116119e-07,
+      "loss": 0.3564,
+      "step": 38320
+    },
+    {
+      "epoch": 14.469611174027936,
+      "grad_norm": 0.629912257194519,
+      "learning_rate": 4.763593041938186e-07,
+      "loss": 0.355,
+      "step": 38330
+    },
+    {
+      "epoch": 14.47338618346546,
+      "grad_norm": 0.6255464553833008,
+      "learning_rate": 4.706838120174584e-07,
+      "loss": 0.3516,
+      "step": 38340
+    },
+    {
+      "epoch": 14.477161192902983,
+      "grad_norm": 0.6643255949020386,
+      "learning_rate": 4.650421726616838e-07,
+      "loss": 0.3522,
+      "step": 38350
+    },
+    {
+      "epoch": 14.480936202340505,
+      "grad_norm": 0.6950973868370056,
+      "learning_rate": 4.594343899824993e-07,
+      "loss": 0.3541,
+      "step": 38360
+    },
+    {
+      "epoch": 14.484711211778029,
+      "grad_norm": 0.7115216851234436,
+      "learning_rate": 4.538604678127667e-07,
+      "loss": 0.3539,
+      "step": 38370
+    },
+    {
+      "epoch": 14.488486221215553,
+      "grad_norm": 0.6523749232292175,
+      "learning_rate": 4.483204099622107e-07,
+      "loss": 0.3555,
+      "step": 38380
+    },
+    {
+      "epoch": 14.492261230653076,
+      "grad_norm": 0.6457560658454895,
+      "learning_rate": 4.4281422021740793e-07,
+      "loss": 0.3583,
+      "step": 38390
+    },
+    {
+      "epoch": 14.4960362400906,
+      "grad_norm": 0.6862673163414001,
+      "learning_rate": 4.373419023417924e-07,
+      "loss": 0.3614,
+      "step": 38400
+    },
+    {
+      "epoch": 14.499811249528124,
+      "grad_norm": 0.623674750328064,
+      "learning_rate": 4.319034600756389e-07,
+      "loss": 0.3476,
+      "step": 38410
+    },
+    {
+      "epoch": 14.503586258965647,
+      "grad_norm": 0.6614101529121399,
+      "learning_rate": 4.264988971360739e-07,
+      "loss": 0.3646,
+      "step": 38420
+    },
+    {
+      "epoch": 14.507361268403171,
+      "grad_norm": 0.650454044342041,
+      "learning_rate": 4.211282172170594e-07,
+      "loss": 0.3581,
+      "step": 38430
+    },
+    {
+      "epoch": 14.511136277840695,
+      "grad_norm": 2.998237133026123,
+      "learning_rate": 4.1579142398940894e-07,
+      "loss": 0.3612,
+      "step": 38440
+    },
+    {
+      "epoch": 14.514911287278219,
+      "grad_norm": 0.6962543725967407,
+      "learning_rate": 4.1048852110076585e-07,
+      "loss": 0.351,
+      "step": 38450
+    },
+    {
+      "epoch": 14.518686296715742,
+      "grad_norm": 0.6083894371986389,
+      "learning_rate": 4.052195121756197e-07,
+      "loss": 0.3617,
+      "step": 38460
+    },
+    {
+      "epoch": 14.522461306153266,
+      "grad_norm": 0.6192116737365723,
+      "learning_rate": 3.9998440081528974e-07,
+      "loss": 0.3561,
+      "step": 38470
+    },
+    {
+      "epoch": 14.52623631559079,
+      "grad_norm": 0.755118727684021,
+      "learning_rate": 3.9478319059791936e-07,
+      "loss": 0.3502,
+      "step": 38480
+    },
+    {
+      "epoch": 14.530011325028312,
+      "grad_norm": 0.606930673122406,
+      "learning_rate": 3.8961588507848143e-07,
+      "loss": 0.3569,
+      "step": 38490
+    },
+    {
+      "epoch": 14.533786334465836,
+      "grad_norm": 0.6492167115211487,
+      "learning_rate": 3.844824877887954e-07,
+      "loss": 0.3622,
+      "step": 38500
+    },
+    {
+      "epoch": 14.53756134390336,
+      "grad_norm": 0.771045982837677,
+      "learning_rate": 3.7938300223748224e-07,
+      "loss": 0.3567,
+      "step": 38510
+    },
+    {
+      "epoch": 14.541336353340883,
+      "grad_norm": 0.6617088317871094,
+      "learning_rate": 3.743174319099929e-07,
+      "loss": 0.3589,
+      "step": 38520
+    },
+    {
+      "epoch": 14.545111362778407,
+      "grad_norm": 0.7514125108718872,
+      "learning_rate": 3.6928578026859673e-07,
+      "loss": 0.3504,
+      "step": 38530
+    },
+    {
+      "epoch": 14.54888637221593,
+      "grad_norm": 0.797964870929718,
+      "learning_rate": 3.642880507523816e-07,
+      "loss": 0.3609,
+      "step": 38540
+    },
+    {
+      "epoch": 14.552661381653454,
+      "grad_norm": 0.6243863105773926,
+      "learning_rate": 3.5932424677724285e-07,
+      "loss": 0.3564,
+      "step": 38550
+    },
+    {
+      "epoch": 14.556436391090978,
+      "grad_norm": 0.6841285824775696,
+      "learning_rate": 3.5439437173589996e-07,
+      "loss": 0.3466,
+      "step": 38560
+    },
+    {
+      "epoch": 14.560211400528502,
+      "grad_norm": 0.6735568642616272,
+      "learning_rate": 3.4949842899787423e-07,
+      "loss": 0.3542,
+      "step": 38570
+    },
+    {
+      "epoch": 14.563986409966025,
+      "grad_norm": 0.6594800353050232,
+      "learning_rate": 3.446364219094944e-07,
+      "loss": 0.3551,
+      "step": 38580
+    },
+    {
+      "epoch": 14.56776141940355,
+      "grad_norm": 0.6403582692146301,
+      "learning_rate": 3.3980835379389675e-07,
+      "loss": 0.3547,
+      "step": 38590
+    },
+    {
+      "epoch": 14.571536428841073,
+      "grad_norm": 0.7487730979919434,
+      "learning_rate": 3.350142279510249e-07,
+      "loss": 0.3494,
+      "step": 38600
+    },
+    {
+      "epoch": 14.575311438278595,
+      "grad_norm": 0.6844361424446106,
+      "learning_rate": 3.3025404765761324e-07,
+      "loss": 0.3558,
+      "step": 38610
+    },
+    {
+      "epoch": 14.579086447716119,
+      "grad_norm": 0.6371558904647827,
+      "learning_rate": 3.255278161672037e-07,
+      "loss": 0.354,
+      "step": 38620
+    },
+    {
+      "epoch": 14.582861457153642,
+      "grad_norm": 0.6130198836326599,
+      "learning_rate": 3.20835536710129e-07,
+      "loss": 0.3504,
+      "step": 38630
+    },
+    {
+      "epoch": 14.586636466591166,
+      "grad_norm": 0.6150995492935181,
+      "learning_rate": 3.1617721249351804e-07,
+      "loss": 0.3579,
+      "step": 38640
+    },
+    {
+      "epoch": 14.59041147602869,
+      "grad_norm": 0.6839898824691772,
+      "learning_rate": 3.115528467012963e-07,
+      "loss": 0.3568,
+      "step": 38650
+    },
+    {
+      "epoch": 14.594186485466214,
+      "grad_norm": 0.6125872731208801,
+      "learning_rate": 3.0696244249416305e-07,
+      "loss": 0.3575,
+      "step": 38660
+    },
+    {
+      "epoch": 14.597961494903737,
+      "grad_norm": 0.6067168712615967,
+      "learning_rate": 3.024060030096254e-07,
+      "loss": 0.3578,
+      "step": 38670
+    },
+    {
+      "epoch": 14.601736504341261,
+      "grad_norm": 0.6161952614784241,
+      "learning_rate": 2.9788353136196435e-07,
+      "loss": 0.3523,
+      "step": 38680
+    },
+    {
+      "epoch": 14.605511513778785,
+      "grad_norm": 0.6272019147872925,
+      "learning_rate": 2.9339503064224615e-07,
+      "loss": 0.3504,
+      "step": 38690
+    },
+    {
+      "epoch": 14.609286523216308,
+      "grad_norm": 1.410046935081482,
+      "learning_rate": 2.889405039183224e-07,
+      "loss": 0.3514,
+      "step": 38700
+    },
+    {
+      "epoch": 14.613061532653832,
+      "grad_norm": 0.6313207745552063,
+      "learning_rate": 2.8451995423481315e-07,
+      "loss": 0.3526,
+      "step": 38710
+    },
+    {
+      "epoch": 14.616836542091356,
+      "grad_norm": 0.5658713579177856,
+      "learning_rate": 2.801333846131293e-07,
+      "loss": 0.3529,
+      "step": 38720
+    },
+    {
+      "epoch": 14.620611551528878,
+      "grad_norm": 0.6897971034049988,
+      "learning_rate": 2.757807980514504e-07,
+      "loss": 0.3529,
+      "step": 38730
+    },
+    {
+      "epoch": 14.624386560966402,
+      "grad_norm": 0.6845794320106506,
+      "learning_rate": 2.714621975247189e-07,
+      "loss": 0.3556,
+      "step": 38740
+    },
+    {
+      "epoch": 14.628161570403925,
+      "grad_norm": 0.6772444844245911,
+      "learning_rate": 2.671775859846626e-07,
+      "loss": 0.3499,
+      "step": 38750
+    },
+    {
+      "epoch": 14.63193657984145,
+      "grad_norm": 0.5830336213111877,
+      "learning_rate": 2.629269663597722e-07,
+      "loss": 0.3561,
+      "step": 38760
+    },
+    {
+      "epoch": 14.635711589278973,
+      "grad_norm": 1.1887177228927612,
+      "learning_rate": 2.587103415553127e-07,
+      "loss": 0.3635,
+      "step": 38770
+    },
+    {
+      "epoch": 14.639486598716497,
+      "grad_norm": 0.6681121587753296,
+      "learning_rate": 2.545277144533009e-07,
+      "loss": 0.3584,
+      "step": 38780
+    },
+    {
+      "epoch": 14.64326160815402,
+      "grad_norm": 0.6925423741340637,
+      "learning_rate": 2.5037908791252227e-07,
+      "loss": 0.352,
+      "step": 38790
+    },
+    {
+      "epoch": 14.647036617591544,
+      "grad_norm": 0.6650035977363586,
+      "learning_rate": 2.462644647685308e-07,
+      "loss": 0.357,
+      "step": 38800
+    },
+    {
+      "epoch": 14.650811627029068,
+      "grad_norm": 0.7153099775314331,
+      "learning_rate": 2.421838478336269e-07,
+      "loss": 0.3473,
+      "step": 38810
+    },
+    {
+      "epoch": 14.654586636466592,
+      "grad_norm": 0.738753616809845,
+      "learning_rate": 2.3813723989687398e-07,
+      "loss": 0.358,
+      "step": 38820
+    },
+    {
+      "epoch": 14.658361645904115,
+      "grad_norm": 0.6535517573356628,
+      "learning_rate": 2.3412464372409293e-07,
+      "loss": 0.36,
+      "step": 38830
+    },
+    {
+      "epoch": 14.662136655341639,
+      "grad_norm": 0.6161585450172424,
+      "learning_rate": 2.301460620578455e-07,
+      "loss": 0.3573,
+      "step": 38840
+    },
+    {
+      "epoch": 14.665911664779163,
+      "grad_norm": 0.6572152972221375,
+      "learning_rate": 2.262014976174731e-07,
+      "loss": 0.349,
+      "step": 38850
+    },
+    {
+      "epoch": 14.669686674216685,
+      "grad_norm": 0.6994696855545044,
+      "learning_rate": 2.2229095309903025e-07,
+      "loss": 0.364,
+      "step": 38860
+    },
+    {
+      "epoch": 14.673461683654208,
+      "grad_norm": 0.6057741641998291,
+      "learning_rate": 2.184144311753511e-07,
+      "loss": 0.3513,
+      "step": 38870
+    },
+    {
+      "epoch": 14.677236693091732,
+      "grad_norm": 0.6180564165115356,
+      "learning_rate": 2.14571934495994e-07,
+      "loss": 0.3535,
+      "step": 38880
+    },
+    {
+      "epoch": 14.681011702529256,
+      "grad_norm": 0.7175959944725037,
+      "learning_rate": 2.1076346568726923e-07,
+      "loss": 0.3528,
+      "step": 38890
+    },
+    {
+      "epoch": 14.68478671196678,
+      "grad_norm": 0.7563894391059875,
+      "learning_rate": 2.0698902735223346e-07,
+      "loss": 0.3621,
+      "step": 38900
+    },
+    {
+      "epoch": 14.688561721404303,
+      "grad_norm": 0.6452277302742004,
+      "learning_rate": 2.0324862207067863e-07,
+      "loss": 0.3579,
+      "step": 38910
+    },
+    {
+      "epoch": 14.692336730841827,
+      "grad_norm": 0.613015353679657,
+      "learning_rate": 1.9954225239913195e-07,
+      "loss": 0.353,
+      "step": 38920
+    },
+    {
+      "epoch": 14.69611174027935,
+      "grad_norm": 0.5917133688926697,
+      "learning_rate": 1.9586992087087252e-07,
+      "loss": 0.3599,
+      "step": 38930
+    },
+    {
+      "epoch": 14.699886749716875,
+      "grad_norm": 0.61355060338974,
+      "learning_rate": 1.9223162999588705e-07,
+      "loss": 0.3512,
+      "step": 38940
+    },
+    {
+      "epoch": 14.703661759154398,
+      "grad_norm": 0.6601895689964294,
+      "learning_rate": 1.886273822609308e-07,
+      "loss": 0.3498,
+      "step": 38950
+    },
+    {
+      "epoch": 14.707436768591922,
+      "grad_norm": 0.6723988056182861,
+      "learning_rate": 1.8505718012946094e-07,
+      "loss": 0.3566,
+      "step": 38960
+    },
+    {
+      "epoch": 14.711211778029446,
+      "grad_norm": 0.5990282893180847,
+      "learning_rate": 1.8152102604168664e-07,
+      "loss": 0.3544,
+      "step": 38970
+    },
+    {
+      "epoch": 14.71498678746697,
+      "grad_norm": 0.7967637181282043,
+      "learning_rate": 1.7801892241452457e-07,
+      "loss": 0.3581,
+      "step": 38980
+    },
+    {
+      "epoch": 14.718761796904491,
+      "grad_norm": 0.7710434198379517,
+      "learning_rate": 1.745508716416433e-07,
+      "loss": 0.3552,
+      "step": 38990
+    },
+    {
+      "epoch": 14.722536806342015,
+      "grad_norm": 0.7385621666908264,
+      "learning_rate": 1.7111687609341342e-07,
+      "loss": 0.3571,
+      "step": 39000
+    },
+    {
+      "epoch": 14.726311815779539,
+      "grad_norm": 0.6296396851539612,
+      "learning_rate": 1.6771693811693523e-07,
+      "loss": 0.358,
+      "step": 39010
+    },
+    {
+      "epoch": 14.730086825217063,
+      "grad_norm": 0.6770869493484497,
+      "learning_rate": 1.643510600360443e-07,
+      "loss": 0.3515,
+      "step": 39020
+    },
+    {
+      "epoch": 14.733861834654586,
+      "grad_norm": 0.6164085865020752,
+      "learning_rate": 1.610192441512781e-07,
+      "loss": 0.3659,
+      "step": 39030
+    },
+    {
+      "epoch": 14.73763684409211,
+      "grad_norm": 1.3167994022369385,
+      "learning_rate": 1.5772149273990956e-07,
+      "loss": 0.3633,
+      "step": 39040
+    },
+    {
+      "epoch": 14.741411853529634,
+      "grad_norm": 0.6471686959266663,
+      "learning_rate": 1.5445780805591337e-07,
+      "loss": 0.3429,
+      "step": 39050
+    },
+    {
+      "epoch": 14.745186862967158,
+      "grad_norm": 0.6361897587776184,
+      "learning_rate": 1.5122819232998852e-07,
+      "loss": 0.3591,
+      "step": 39060
+    },
+    {
+      "epoch": 14.748961872404681,
+      "grad_norm": 0.6572710275650024,
+      "learning_rate": 1.4803264776955262e-07,
+      "loss": 0.3654,
+      "step": 39070
+    },
+    {
+      "epoch": 14.752736881842205,
+      "grad_norm": 0.6715047359466553,
+      "learning_rate": 1.448711765587196e-07,
+      "loss": 0.3587,
+      "step": 39080
+    },
+    {
+      "epoch": 14.756511891279729,
+      "grad_norm": 0.6003199219703674,
+      "learning_rate": 1.4174378085833885e-07,
+      "loss": 0.3599,
+      "step": 39090
+    },
+    {
+      "epoch": 14.760286900717253,
+      "grad_norm": 0.637787938117981,
+      "learning_rate": 1.3865046280594486e-07,
+      "loss": 0.3602,
+      "step": 39100
+    },
+    {
+      "epoch": 14.764061910154776,
+      "grad_norm": 0.6328327655792236,
+      "learning_rate": 1.3559122451579643e-07,
+      "loss": 0.354,
+      "step": 39110
+    },
+    {
+      "epoch": 14.767836919592298,
+      "grad_norm": 0.6023069024085999,
+      "learning_rate": 1.3256606807884874e-07,
+      "loss": 0.3549,
+      "step": 39120
+    },
+    {
+      "epoch": 14.771611929029822,
+      "grad_norm": 0.631928563117981,
+      "learning_rate": 1.2957499556277563e-07,
+      "loss": 0.3608,
+      "step": 39130
+    },
+    {
+      "epoch": 14.775386938467346,
+      "grad_norm": 0.6594300866127014,
+      "learning_rate": 1.266180090119473e-07,
+      "loss": 0.3556,
+      "step": 39140
+    },
+    {
+      "epoch": 14.77916194790487,
+      "grad_norm": 0.6091174483299255,
+      "learning_rate": 1.2369511044742488e-07,
+      "loss": 0.3608,
+      "step": 39150
+    },
+    {
+      "epoch": 14.782936957342393,
+      "grad_norm": 0.7767238020896912,
+      "learning_rate": 1.2080630186698805e-07,
+      "loss": 0.3599,
+      "step": 39160
+    },
+    {
+      "epoch": 14.786711966779917,
+      "grad_norm": 0.6567763090133667,
+      "learning_rate": 1.1795158524511851e-07,
+      "loss": 0.3513,
+      "step": 39170
+    },
+    {
+      "epoch": 14.79048697621744,
+      "grad_norm": 0.6258882880210876,
+      "learning_rate": 1.1513096253297773e-07,
+      "loss": 0.3553,
+      "step": 39180
+    },
+    {
+      "epoch": 14.794261985654964,
+      "grad_norm": 0.6142001152038574,
+      "learning_rate": 1.1234443565843466e-07,
+      "loss": 0.3534,
+      "step": 39190
+    },
+    {
+      "epoch": 14.798036995092488,
+      "grad_norm": 0.6969514489173889,
+      "learning_rate": 1.0959200652606028e-07,
+      "loss": 0.3575,
+      "step": 39200
+    },
+    {
+      "epoch": 14.801812004530012,
+      "grad_norm": 0.729709804058075,
+      "learning_rate": 1.0687367701711082e-07,
+      "loss": 0.347,
+      "step": 39210
+    },
+    {
+      "epoch": 14.805587013967536,
+      "grad_norm": 0.6622199416160583,
+      "learning_rate": 1.0418944898953898e-07,
+      "loss": 0.3567,
+      "step": 39220
+    },
+    {
+      "epoch": 14.809362023405058,
+      "grad_norm": 0.7699572443962097,
+      "learning_rate": 1.0153932427798829e-07,
+      "loss": 0.3591,
+      "step": 39230
+    },
+    {
+      "epoch": 14.813137032842581,
+      "grad_norm": 0.6490731239318848,
+      "learning_rate": 9.892330469379874e-08,
+      "loss": 0.359,
+      "step": 39240
+    },
+    {
+      "epoch": 14.816912042280105,
+      "grad_norm": 0.6460186839103699,
+      "learning_rate": 9.634139202498449e-08,
+      "loss": 0.3545,
+      "step": 39250
+    },
+    {
+      "epoch": 14.820687051717629,
+      "grad_norm": 0.789674699306488,
+      "learning_rate": 9.379358803627281e-08,
+      "loss": 0.3515,
+      "step": 39260
+    },
+    {
+      "epoch": 14.824462061155153,
+      "grad_norm": 0.6353222131729126,
+      "learning_rate": 9.12798944690485e-08,
+      "loss": 0.354,
+      "step": 39270
+    },
+    {
+      "epoch": 14.828237070592676,
+      "grad_norm": 0.6650903820991516,
+      "learning_rate": 8.880031304141501e-08,
+      "loss": 0.349,
+      "step": 39280
+    },
+    {
+      "epoch": 14.8320120800302,
+      "grad_norm": 0.7060394287109375,
+      "learning_rate": 8.635484544812777e-08,
+      "loss": 0.3572,
+      "step": 39290
+    },
+    {
+      "epoch": 14.835787089467724,
+      "grad_norm": 2.0902068614959717,
+      "learning_rate": 8.394349336064422e-08,
+      "loss": 0.3551,
+      "step": 39300
+    },
+    {
+      "epoch": 14.839562098905247,
+      "grad_norm": 0.7247877717018127,
+      "learning_rate": 8.156625842710153e-08,
+      "loss": 0.3518,
+      "step": 39310
+    },
+    {
+      "epoch": 14.843337108342771,
+      "grad_norm": 0.6578109264373779,
+      "learning_rate": 7.92231422723222e-08,
+      "loss": 0.3499,
+      "step": 39320
+    },
+    {
+      "epoch": 14.847112117780295,
+      "grad_norm": 0.630346953868866,
+      "learning_rate": 7.691414649780293e-08,
+      "loss": 0.351,
+      "step": 39330
+    },
+    {
+      "epoch": 14.850887127217819,
+      "grad_norm": 0.6793875098228455,
+      "learning_rate": 7.463927268171467e-08,
+      "loss": 0.3486,
+      "step": 39340
+    },
+    {
+      "epoch": 14.854662136655342,
+      "grad_norm": 0.6113439798355103,
+      "learning_rate": 7.239852237891365e-08,
+      "loss": 0.3605,
+      "step": 39350
+    },
+    {
+      "epoch": 14.858437146092864,
+      "grad_norm": 0.7049407958984375,
+      "learning_rate": 7.019189712093588e-08,
+      "loss": 0.3501,
+      "step": 39360
+    },
+    {
+      "epoch": 14.862212155530388,
+      "grad_norm": 0.7017509937286377,
+      "learning_rate": 6.801939841598603e-08,
+      "loss": 0.3599,
+      "step": 39370
+    },
+    {
+      "epoch": 14.865987164967912,
+      "grad_norm": 0.6390900015830994,
+      "learning_rate": 6.588102774894856e-08,
+      "loss": 0.3557,
+      "step": 39380
+    },
+    {
+      "epoch": 14.869762174405436,
+      "grad_norm": 0.6362788081169128,
+      "learning_rate": 6.377678658137653e-08,
+      "loss": 0.3548,
+      "step": 39390
+    },
+    {
+      "epoch": 14.87353718384296,
+      "grad_norm": 0.6280719041824341,
+      "learning_rate": 6.17066763514973e-08,
+      "loss": 0.3493,
+      "step": 39400
+    },
+    {
+      "epoch": 14.877312193280483,
+      "grad_norm": 0.6526157855987549,
+      "learning_rate": 5.967069847421792e-08,
+      "loss": 0.3532,
+      "step": 39410
+    },
+    {
+      "epoch": 14.881087202718007,
+      "grad_norm": 0.7428958415985107,
+      "learning_rate": 5.7668854341108583e-08,
+      "loss": 0.3563,
+      "step": 39420
+    },
+    {
+      "epoch": 14.88486221215553,
+      "grad_norm": 0.6039178371429443,
+      "learning_rate": 5.570114532040815e-08,
+      "loss": 0.3567,
+      "step": 39430
+    },
+    {
+      "epoch": 14.888637221593054,
+      "grad_norm": 1.0244649648666382,
+      "learning_rate": 5.3767572757029706e-08,
+      "loss": 0.3521,
+      "step": 39440
+    },
+    {
+      "epoch": 14.892412231030578,
+      "grad_norm": 0.6426689028739929,
+      "learning_rate": 5.186813797254386e-08,
+      "loss": 0.3525,
+      "step": 39450
+    },
+    {
+      "epoch": 14.896187240468102,
+      "grad_norm": 0.7617814540863037,
+      "learning_rate": 5.000284226521213e-08,
+      "loss": 0.3529,
+      "step": 39460
+    },
+    {
+      "epoch": 14.899962249905625,
+      "grad_norm": 0.7546370625495911,
+      "learning_rate": 4.8171686909931346e-08,
+      "loss": 0.3519,
+      "step": 39470
+    },
+    {
+      "epoch": 14.90373725934315,
+      "grad_norm": 0.6495998501777649,
+      "learning_rate": 4.637467315828925e-08,
+      "loss": 0.358,
+      "step": 39480
+    },
+    {
+      "epoch": 14.907512268780671,
+      "grad_norm": 0.6548603177070618,
+      "learning_rate": 4.461180223852002e-08,
+      "loss": 0.3483,
+      "step": 39490
+    },
+    {
+      "epoch": 14.911287278218195,
+      "grad_norm": 0.6339406371116638,
+      "learning_rate": 4.288307535553204e-08,
+      "loss": 0.3604,
+      "step": 39500
+    },
+    {
+      "epoch": 14.915062287655719,
+      "grad_norm": 0.6532941460609436,
+      "learning_rate": 4.1188493690896836e-08,
+      "loss": 0.3598,
+      "step": 39510
+    },
+    {
+      "epoch": 14.918837297093242,
+      "grad_norm": 0.6393380761146545,
+      "learning_rate": 3.952805840284346e-08,
+      "loss": 0.3521,
+      "step": 39520
+    },
+    {
+      "epoch": 14.922612306530766,
+      "grad_norm": 0.6625932455062866,
+      "learning_rate": 3.790177062625855e-08,
+      "loss": 0.3567,
+      "step": 39530
+    },
+    {
+      "epoch": 14.92638731596829,
+      "grad_norm": 0.6500564217567444,
+      "learning_rate": 3.6309631472702946e-08,
+      "loss": 0.3549,
+      "step": 39540
+    },
+    {
+      "epoch": 14.930162325405814,
+      "grad_norm": 0.6394986510276794,
+      "learning_rate": 3.475164203038395e-08,
+      "loss": 0.3602,
+      "step": 39550
+    },
+    {
+      "epoch": 14.933937334843337,
+      "grad_norm": 0.6065540313720703,
+      "learning_rate": 3.3227803364171974e-08,
+      "loss": 0.3435,
+      "step": 39560
+    },
+    {
+      "epoch": 14.937712344280861,
+      "grad_norm": 0.6980146765708923,
+      "learning_rate": 3.173811651560055e-08,
+      "loss": 0.3484,
+      "step": 39570
+    },
+    {
+      "epoch": 14.941487353718385,
+      "grad_norm": 0.6110873222351074,
+      "learning_rate": 3.0282582502855204e-08,
+      "loss": 0.3528,
+      "step": 39580
+    },
+    {
+      "epoch": 14.945262363155909,
+      "grad_norm": 0.74271559715271,
+      "learning_rate": 2.8861202320773496e-08,
+      "loss": 0.3547,
+      "step": 39590
+    },
+    {
+      "epoch": 14.949037372593432,
+      "grad_norm": 0.5941160321235657,
+      "learning_rate": 2.7473976940867175e-08,
+      "loss": 0.3606,
+      "step": 39600
+    },
+    {
+      "epoch": 14.952812382030956,
+      "grad_norm": 0.6520451903343201,
+      "learning_rate": 2.6120907311277808e-08,
+      "loss": 0.351,
+      "step": 39610
+    },
+    {
+      "epoch": 14.956587391468478,
+      "grad_norm": 0.6158818006515503,
+      "learning_rate": 2.480199435682673e-08,
+      "loss": 0.3502,
+      "step": 39620
+    },
+    {
+      "epoch": 14.960362400906002,
+      "grad_norm": 0.6889585852622986,
+      "learning_rate": 2.3517238978976174e-08,
+      "loss": 0.3592,
+      "step": 39630
+    },
+    {
+      "epoch": 14.964137410343525,
+      "grad_norm": 0.5970800518989563,
+      "learning_rate": 2.2266642055834842e-08,
+      "loss": 0.359,
+      "step": 39640
+    },
+    {
+      "epoch": 14.96791241978105,
+      "grad_norm": 0.65416419506073,
+      "learning_rate": 2.105020444219119e-08,
+      "loss": 0.3527,
+      "step": 39650
+    },
+    {
+      "epoch": 14.971687429218573,
+      "grad_norm": 0.664504885673523,
+      "learning_rate": 1.9867926969452388e-08,
+      "loss": 0.364,
+      "step": 39660
+    },
+    {
+      "epoch": 14.975462438656097,
+      "grad_norm": 0.6672757863998413,
+      "learning_rate": 1.8719810445705366e-08,
+      "loss": 0.3531,
+      "step": 39670
+    },
+    {
+      "epoch": 14.97923744809362,
+      "grad_norm": 0.6902869343757629,
+      "learning_rate": 1.760585565567241e-08,
+      "loss": 0.3564,
+      "step": 39680
+    },
+    {
+      "epoch": 14.983012457531144,
+      "grad_norm": 0.7175024151802063,
+      "learning_rate": 1.6526063360727816e-08,
+      "loss": 0.3559,
+      "step": 39690
+    },
+    {
+      "epoch": 14.986787466968668,
+      "grad_norm": 0.6465026140213013,
+      "learning_rate": 1.5480434298908997e-08,
+      "loss": 0.3539,
+      "step": 39700
+    },
+    {
+      "epoch": 14.990562476406192,
+      "grad_norm": 0.6712453961372375,
+      "learning_rate": 1.4468969184883164e-08,
+      "loss": 0.3535,
+      "step": 39710
+    },
+    {
+      "epoch": 14.994337485843715,
+      "grad_norm": 0.620090126991272,
+      "learning_rate": 1.3491668709980643e-08,
+      "loss": 0.3544,
+      "step": 39720
+    },
+    {
+      "epoch": 14.998112495281239,
+      "grad_norm": 0.6814067363739014,
+      "learning_rate": 1.254853354217822e-08,
+      "loss": 0.3551,
+      "step": 39730
+    },
+    {
+      "epoch": 15.001887504718761,
+      "grad_norm": 0.8120419979095459,
+      "learning_rate": 1.1639564326099139e-08,
+      "loss": 0.3553,
+      "step": 39740
+    },
+    {
+      "epoch": 15.005662514156285,
+      "grad_norm": 0.6603171825408936,
+      "learning_rate": 1.0764761683018653e-08,
+      "loss": 0.3537,
+      "step": 39750
+    },
+    {
+      "epoch": 15.009437523593808,
+      "grad_norm": 0.6315841674804688,
+      "learning_rate": 9.924126210852925e-09,
+      "loss": 0.3503,
+      "step": 39760
+    },
+    {
+      "epoch": 15.013212533031332,
+      "grad_norm": 0.6496405601501465,
+      "learning_rate": 9.117658484170122e-09,
+      "loss": 0.3577,
+      "step": 39770
+    },
+    {
+      "epoch": 15.016987542468856,
+      "grad_norm": 0.7022120356559753,
+      "learning_rate": 8.345359054179324e-09,
+      "loss": 0.3678,
+      "step": 39780
+    },
+    {
+      "epoch": 15.02076255190638,
+      "grad_norm": 0.7122491598129272,
+      "learning_rate": 7.60722844874162e-09,
+      "loss": 0.3561,
+      "step": 39790
+    },
+    {
+      "epoch": 15.024537561343903,
+      "grad_norm": 0.625100314617157,
+      "learning_rate": 6.903267172364558e-09,
+      "loss": 0.347,
+      "step": 39800
+    },
+    {
+      "epoch": 15.028312570781427,
+      "grad_norm": 0.6396900415420532,
+      "learning_rate": 6.23347570619659e-09,
+      "loss": 0.3605,
+      "step": 39810
+    },
+    {
+      "epoch": 15.03208758021895,
+      "grad_norm": 0.5994987487792969,
+      "learning_rate": 5.59785450803818e-09,
+      "loss": 0.3573,
+      "step": 39820
+    },
+    {
+      "epoch": 15.035862589656475,
+      "grad_norm": 0.7558580636978149,
+      "learning_rate": 4.996404012325151e-09,
+      "loss": 0.3536,
+      "step": 39830
+    },
+    {
+      "epoch": 15.039637599093998,
+      "grad_norm": 0.6943445801734924,
+      "learning_rate": 4.429124630145331e-09,
+      "loss": 0.3561,
+      "step": 39840
+    },
+    {
+      "epoch": 15.043412608531522,
+      "grad_norm": 0.9574018716812134,
+      "learning_rate": 3.89601674922746e-09,
+      "loss": 0.362,
+      "step": 39850
+    },
+    {
+      "epoch": 15.047187617969046,
+      "grad_norm": 0.6449063420295715,
+      "learning_rate": 3.3970807339467336e-09,
+      "loss": 0.3537,
+      "step": 39860
+    },
+    {
+      "epoch": 15.050962627406568,
+      "grad_norm": 0.5976826548576355,
+      "learning_rate": 2.9323169253248073e-09,
+      "loss": 0.3513,
+      "step": 39870
+    },
+    {
+      "epoch": 15.054737636844091,
+      "grad_norm": 0.6099610328674316,
+      "learning_rate": 2.501725641018693e-09,
+      "loss": 0.3545,
+      "step": 39880
+    },
+    {
+      "epoch": 15.058512646281615,
+      "grad_norm": 0.7549223899841309,
+      "learning_rate": 2.105307175331861e-09,
+      "loss": 0.3567,
+      "step": 39890
+    },
+    {
+      "epoch": 15.062287655719139,
+      "grad_norm": 0.6407955288887024,
+      "learning_rate": 1.7430617992197918e-09,
+      "loss": 0.3558,
+      "step": 39900
+    },
+    {
+      "epoch": 15.066062665156663,
+      "grad_norm": 0.6533637046813965,
+      "learning_rate": 1.4149897602677708e-09,
+      "loss": 0.3484,
+      "step": 39910
+    },
+    {
+      "epoch": 15.069837674594186,
+      "grad_norm": 0.604796826839447,
+      "learning_rate": 1.1210912827130937e-09,
+      "loss": 0.3582,
+      "step": 39920
+    },
+    {
+      "epoch": 15.07361268403171,
+      "grad_norm": 0.5980611443519592,
+      "learning_rate": 8.613665674284121e-10,
+      "loss": 0.3548,
+      "step": 39930
+    },
+    {
+      "epoch": 15.077387693469234,
+      "grad_norm": 0.6383752226829529,
+      "learning_rate": 6.358157919383878e-10,
+      "loss": 0.3544,
+      "step": 39940
+    },
+    {
+      "epoch": 15.081162702906758,
+      "grad_norm": 0.7199327349662781,
+      "learning_rate": 4.4443911040303877e-10,
+      "loss": 0.3526,
+      "step": 39950
+    },
+    {
+      "epoch": 15.084937712344281,
+      "grad_norm": 0.7080827951431274,
+      "learning_rate": 2.8723665362329066e-10,
+      "loss": 0.3549,
+      "step": 39960
+    },
+    {
+      "epoch": 15.088712721781805,
+      "grad_norm": 0.6483982801437378,
+      "learning_rate": 1.6420852905207895e-10,
+      "loss": 0.3583,
+      "step": 39970
+    },
+    {
+      "epoch": 15.092487731219329,
+      "grad_norm": 0.6438269019126892,
+      "learning_rate": 7.535482076659328e-11,
+      "loss": 0.3451,
+      "step": 39980
+    },
+    {
+      "epoch": 15.09626274065685,
+      "grad_norm": 0.6255228519439697,
+      "learning_rate": 2.0675589512686443e-11,
+      "loss": 0.3557,
+      "step": 39990
+    },
+    {
+      "epoch": 15.100037750094375,
+      "grad_norm": 0.8365612626075745,
+      "learning_rate": 1.7087265491433358e-13,
+      "loss": 0.353,
+      "step": 40000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 40000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 16,
+  "save_steps": 4000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0739727642197072e+20,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}