diff --git "a/codet5+/q90useseq/checkpoint-16161/trainer_state.json" "b/codet5+/q90useseq/checkpoint-16161/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/codet5+/q90useseq/checkpoint-16161/trainer_state.json"
@@ -0,0 +1,11340 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998607824028958,
+  "eval_steps": 500,
+  "global_step": 16161,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.563289642333984,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 8.4336,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 62.98283386230469,
+      "learning_rate": 2.5e-06,
+      "loss": 8.5697,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 54.93251419067383,
+      "learning_rate": 5e-06,
+      "loss": 7.4634,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.467167854309082,
+      "learning_rate": 7.5e-06,
+      "loss": 5.4256,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.084922790527344,
+      "learning_rate": 1e-05,
+      "loss": 4.2997,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.685922622680664,
+      "learning_rate": 1.25e-05,
+      "loss": 3.9332,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.632925987243652,
+      "learning_rate": 1.5e-05,
+      "loss": 3.7494,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.258584499359131,
+      "learning_rate": 1.75e-05,
+      "loss": 3.4965,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.852085113525391,
+      "learning_rate": 2e-05,
+      "loss": 3.5823,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.215030670166016,
+      "learning_rate": 2.25e-05,
+      "loss": 3.3422,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.968944072723389,
+      "learning_rate": 2.5e-05,
+      "loss": 3.4434,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.891924858093262,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 3.3898,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.870503902435303,
+      "learning_rate": 3e-05,
+      "loss": 3.3386,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.607165336608887,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 3.2335,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.651923179626465,
+      "learning_rate": 3.5e-05,
+      "loss": 3.216,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.639860153198242,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 3.2893,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.180785179138184,
+      "learning_rate": 4e-05,
+      "loss": 3.1823,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.757967472076416,
+      "learning_rate": 4.25e-05,
+      "loss": 3.1802,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.358530044555664,
+      "learning_rate": 4.5e-05,
+      "loss": 3.2204,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.176595687866211,
+      "learning_rate": 4.75e-05,
+      "loss": 3.1479,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.044156551361084,
+      "learning_rate": 5e-05,
+      "loss": 3.2019,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.459234714508057,
+      "learning_rate": 4.996867364200238e-05,
+      "loss": 3.1785,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.484199047088623,
+      "learning_rate": 4.993734728400476e-05,
+      "loss": 2.9384,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.390307426452637,
+      "learning_rate": 4.990602092600715e-05,
+      "loss": 3.0402,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.658885478973389,
+      "learning_rate": 4.987469456800953e-05,
+      "loss": 3.1198,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.551351070404053,
+      "learning_rate": 4.98433682100119e-05,
+      "loss": 3.0451,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.482519149780273,
+      "learning_rate": 4.981204185201428e-05,
+      "loss": 3.165,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.959535598754883,
+      "learning_rate": 4.978071549401667e-05,
+      "loss": 3.156,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.811633586883545,
+      "learning_rate": 4.974938913601905e-05,
+      "loss": 3.0748,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.658722400665283,
+      "learning_rate": 4.971806277802143e-05,
+      "loss": 2.9398,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.098793983459473,
+      "learning_rate": 4.968673642002381e-05,
+      "loss": 2.9919,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.794303894042969,
+      "learning_rate": 4.965541006202619e-05,
+      "loss": 3.0925,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.6707611083984375,
+      "learning_rate": 4.9624083704028576e-05,
+      "loss": 3.0318,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.9659223556518555,
+      "learning_rate": 4.9592757346030956e-05,
+      "loss": 3.0405,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.642695426940918,
+      "learning_rate": 4.9561430988033336e-05,
+      "loss": 2.9802,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.881353378295898,
+      "learning_rate": 4.953010463003571e-05,
+      "loss": 2.9958,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.710669040679932,
+      "learning_rate": 4.9498778272038096e-05,
+      "loss": 2.853,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.929610013961792,
+      "learning_rate": 4.9467451914040476e-05,
+      "loss": 2.8538,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2416839599609375,
+      "learning_rate": 4.9436125556042856e-05,
+      "loss": 2.9904,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.125613689422607,
+      "learning_rate": 4.9404799198045237e-05,
+      "loss": 3.0157,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.323487758636475,
+      "learning_rate": 4.9373472840047617e-05,
+      "loss": 3.0734,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.283641338348389,
+      "learning_rate": 4.934214648205e-05,
+      "loss": 3.0373,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.504553318023682,
+      "learning_rate": 4.931082012405238e-05,
+      "loss": 3.1041,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.953495740890503,
+      "learning_rate": 4.9279493766054763e-05,
+      "loss": 2.9039,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.179076671600342,
+      "learning_rate": 4.9248167408057143e-05,
+      "loss": 3.0008,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.270627498626709,
+      "learning_rate": 4.921684105005952e-05,
+      "loss": 2.9601,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.364182472229004,
+      "learning_rate": 4.9185514692061904e-05,
+      "loss": 2.8768,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.999711751937866,
+      "learning_rate": 4.9154188334064284e-05,
+      "loss": 3.0547,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.382208347320557,
+      "learning_rate": 4.9122861976066664e-05,
+      "loss": 2.9902,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.151576995849609,
+      "learning_rate": 4.9091535618069044e-05,
+      "loss": 2.8,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.85528039932251,
+      "learning_rate": 4.9060209260071424e-05,
+      "loss": 2.8966,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.874793767929077,
+      "learning_rate": 4.902888290207381e-05,
+      "loss": 3.0284,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.303886890411377,
+      "learning_rate": 4.899755654407619e-05,
+      "loss": 2.7964,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.484199047088623,
+      "learning_rate": 4.896623018607857e-05,
+      "loss": 2.9817,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.022371768951416,
+      "learning_rate": 4.893490382808095e-05,
+      "loss": 2.9067,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7025790214538574,
+      "learning_rate": 4.890357747008333e-05,
+      "loss": 2.9264,
+      "step": 550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.232612133026123,
+      "learning_rate": 4.887225111208571e-05,
+      "loss": 2.8737,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.300458908081055,
+      "learning_rate": 4.884092475408809e-05,
+      "loss": 2.9234,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4513115882873535,
+      "learning_rate": 4.880959839609047e-05,
+      "loss": 2.8182,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.555144786834717,
+      "learning_rate": 4.877827203809285e-05,
+      "loss": 2.9132,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.582031726837158,
+      "learning_rate": 4.874694568009524e-05,
+      "loss": 2.8415,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.070923328399658,
+      "learning_rate": 4.871561932209762e-05,
+      "loss": 2.9713,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.049088954925537,
+      "learning_rate": 4.86842929641e-05,
+      "loss": 3.0673,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.457193374633789,
+      "learning_rate": 4.865296660610238e-05,
+      "loss": 2.9815,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.2220258712768555,
+      "learning_rate": 4.862164024810476e-05,
+      "loss": 2.969,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.977060556411743,
+      "learning_rate": 4.859031389010714e-05,
+      "loss": 3.0564,
+      "step": 650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.071286678314209,
+      "learning_rate": 4.855898753210952e-05,
+      "loss": 2.9144,
+      "step": 660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.329357147216797,
+      "learning_rate": 4.85276611741119e-05,
+      "loss": 2.9175,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.701284646987915,
+      "learning_rate": 4.849633481611428e-05,
+      "loss": 2.911,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.07001256942749,
+      "learning_rate": 4.8465008458116665e-05,
+      "loss": 2.8067,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.72738790512085,
+      "learning_rate": 4.8433682100119045e-05,
+      "loss": 3.0779,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.782459259033203,
+      "learning_rate": 4.8402355742121425e-05,
+      "loss": 2.9219,
+      "step": 710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.410059452056885,
+      "learning_rate": 4.8371029384123805e-05,
+      "loss": 3.0018,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.083688259124756,
+      "learning_rate": 4.8339703026126185e-05,
+      "loss": 2.9102,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7951762676239014,
+      "learning_rate": 4.8308376668128565e-05,
+      "loss": 2.8857,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.722205400466919,
+      "learning_rate": 4.8277050310130945e-05,
+      "loss": 2.8893,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.130901336669922,
+      "learning_rate": 4.8245723952133325e-05,
+      "loss": 2.943,
+      "step": 760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.320456027984619,
+      "learning_rate": 4.8214397594135705e-05,
+      "loss": 2.8967,
+      "step": 770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.467789649963379,
+      "learning_rate": 4.818307123613809e-05,
+      "loss": 2.721,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.661444902420044,
+      "learning_rate": 4.815174487814047e-05,
+      "loss": 2.8507,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.432613372802734,
+      "learning_rate": 4.812041852014285e-05,
+      "loss": 2.818,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8605380058288574,
+      "learning_rate": 4.808909216214523e-05,
+      "loss": 2.8162,
+      "step": 810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.229771137237549,
+      "learning_rate": 4.805776580414761e-05,
+      "loss": 2.9675,
+      "step": 820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.812516212463379,
+      "learning_rate": 4.802643944615e-05,
+      "loss": 2.8635,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.7779951095581055,
+      "learning_rate": 4.799511308815237e-05,
+      "loss": 2.7532,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.997580051422119,
+      "learning_rate": 4.796378673015475e-05,
+      "loss": 2.8165,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.363816738128662,
+      "learning_rate": 4.793246037215713e-05,
+      "loss": 2.7508,
+      "step": 860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.727748155593872,
+      "learning_rate": 4.790113401415952e-05,
+      "loss": 2.9275,
+      "step": 870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7258036136627197,
+      "learning_rate": 4.78698076561619e-05,
+      "loss": 2.7293,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.236034393310547,
+      "learning_rate": 4.783848129816428e-05,
+      "loss": 2.7062,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.967883348464966,
+      "learning_rate": 4.780715494016666e-05,
+      "loss": 2.8896,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3881027698516846,
+      "learning_rate": 4.777582858216904e-05,
+      "loss": 2.8254,
+      "step": 910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.047328472137451,
+      "learning_rate": 4.7744502224171426e-05,
+      "loss": 2.8455,
+      "step": 920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7306952476501465,
+      "learning_rate": 4.77131758661738e-05,
+      "loss": 2.6811,
+      "step": 930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.450077056884766,
+      "learning_rate": 4.768184950817618e-05,
+      "loss": 2.7895,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.613541603088379,
+      "learning_rate": 4.765052315017856e-05,
+      "loss": 2.8591,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.935161828994751,
+      "learning_rate": 4.7619196792180946e-05,
+      "loss": 2.8168,
+      "step": 960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0404486656188965,
+      "learning_rate": 4.7587870434183326e-05,
+      "loss": 2.9274,
+      "step": 970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5758628845214844,
+      "learning_rate": 4.7556544076185706e-05,
+      "loss": 2.8127,
+      "step": 980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.76188588142395,
+      "learning_rate": 4.7525217718188086e-05,
+      "loss": 2.8306,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9346506595611572,
+      "learning_rate": 4.7493891360190466e-05,
+      "loss": 2.8627,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9877500534057617,
+      "learning_rate": 4.7462565002192846e-05,
+      "loss": 2.7633,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7940495014190674,
+      "learning_rate": 4.743123864419523e-05,
+      "loss": 2.9002,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6916394233703613,
+      "learning_rate": 4.7399912286197606e-05,
+      "loss": 2.7947,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.910442352294922,
+      "learning_rate": 4.7368585928199986e-05,
+      "loss": 2.7819,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.798110246658325,
+      "learning_rate": 4.7337259570202366e-05,
+      "loss": 2.781,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.843815803527832,
+      "learning_rate": 4.730593321220475e-05,
+      "loss": 2.8417,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2749738693237305,
+      "learning_rate": 4.727460685420713e-05,
+      "loss": 2.8834,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.073596477508545,
+      "learning_rate": 4.724328049620951e-05,
+      "loss": 2.7237,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9949352741241455,
+      "learning_rate": 4.721195413821189e-05,
+      "loss": 2.8605,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.504049062728882,
+      "learning_rate": 4.718062778021427e-05,
+      "loss": 2.7565,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.345983982086182,
+      "learning_rate": 4.714930142221666e-05,
+      "loss": 2.8596,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.165377616882324,
+      "learning_rate": 4.711797506421904e-05,
+      "loss": 2.8525,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8506295680999756,
+      "learning_rate": 4.708664870622141e-05,
+      "loss": 2.8325,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.528014659881592,
+      "learning_rate": 4.705532234822379e-05,
+      "loss": 2.7329,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.269871234893799,
+      "learning_rate": 4.702399599022618e-05,
+      "loss": 2.7769,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8873980045318604,
+      "learning_rate": 4.699266963222856e-05,
+      "loss": 2.8734,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6617813110351562,
+      "learning_rate": 4.696134327423094e-05,
+      "loss": 2.9458,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9084973335266113,
+      "learning_rate": 4.693001691623332e-05,
+      "loss": 2.9346,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.360800266265869,
+      "learning_rate": 4.68986905582357e-05,
+      "loss": 2.8122,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2493598461151123,
+      "learning_rate": 4.686736420023809e-05,
+      "loss": 2.7762,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.779627561569214,
+      "learning_rate": 4.683603784224047e-05,
+      "loss": 2.9664,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.850245475769043,
+      "learning_rate": 4.680471148424285e-05,
+      "loss": 2.9419,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.702455759048462,
+      "learning_rate": 4.677338512624522e-05,
+      "loss": 2.7791,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.315182685852051,
+      "learning_rate": 4.674205876824761e-05,
+      "loss": 2.9742,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8250505924224854,
+      "learning_rate": 4.671073241024999e-05,
+      "loss": 2.6065,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.961651563644409,
+      "learning_rate": 4.667940605225237e-05,
+      "loss": 2.7905,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.800668239593506,
+      "learning_rate": 4.664807969425475e-05,
+      "loss": 2.6825,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.948024272918701,
+      "learning_rate": 4.661675333625713e-05,
+      "loss": 2.6316,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6402764320373535,
+      "learning_rate": 4.6585426978259514e-05,
+      "loss": 2.8085,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.029736042022705,
+      "learning_rate": 4.6554100620261894e-05,
+      "loss": 2.6869,
+      "step": 1300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9908854961395264,
+      "learning_rate": 4.6522774262264274e-05,
+      "loss": 2.7812,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9148330688476562,
+      "learning_rate": 4.649144790426665e-05,
+      "loss": 2.8214,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.979130744934082,
+      "learning_rate": 4.6460121546269034e-05,
+      "loss": 2.9188,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.001101970672607,
+      "learning_rate": 4.6428795188271414e-05,
+      "loss": 2.7542,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8721792697906494,
+      "learning_rate": 4.6397468830273794e-05,
+      "loss": 2.8949,
+      "step": 1350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9208767414093018,
+      "learning_rate": 4.6366142472276174e-05,
+      "loss": 2.6644,
+      "step": 1360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7345130443573,
+      "learning_rate": 4.6334816114278554e-05,
+      "loss": 2.7879,
+      "step": 1370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.162068843841553,
+      "learning_rate": 4.630348975628094e-05,
+      "loss": 2.8323,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.239302635192871,
+      "learning_rate": 4.627216339828332e-05,
+      "loss": 2.8012,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.081932544708252,
+      "learning_rate": 4.62408370402857e-05,
+      "loss": 2.7515,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6898136138916016,
+      "learning_rate": 4.620951068228808e-05,
+      "loss": 2.7073,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.468146562576294,
+      "learning_rate": 4.617818432429046e-05,
+      "loss": 2.9264,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6380133628845215,
+      "learning_rate": 4.614685796629284e-05,
+      "loss": 2.8197,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.102017402648926,
+      "learning_rate": 4.611553160829522e-05,
+      "loss": 2.7289,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9504294395446777,
+      "learning_rate": 4.60842052502976e-05,
+      "loss": 2.9459,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.055880069732666,
+      "learning_rate": 4.605287889229998e-05,
+      "loss": 2.776,
+      "step": 1460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9048073291778564,
+      "learning_rate": 4.602155253430237e-05,
+      "loss": 2.7488,
+      "step": 1470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5023839473724365,
+      "learning_rate": 4.599022617630475e-05,
+      "loss": 2.7139,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.604358434677124,
+      "learning_rate": 4.595889981830713e-05,
+      "loss": 2.7338,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7570292949676514,
+      "learning_rate": 4.592757346030951e-05,
+      "loss": 2.7606,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.767991065979004,
+      "learning_rate": 4.589624710231189e-05,
+      "loss": 2.8058,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.012033939361572,
+      "learning_rate": 4.586492074431427e-05,
+      "loss": 2.8896,
+      "step": 1520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8234939575195312,
+      "learning_rate": 4.583359438631665e-05,
+      "loss": 2.7363,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.762657880783081,
+      "learning_rate": 4.580226802831903e-05,
+      "loss": 2.9173,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2961838245391846,
+      "learning_rate": 4.577094167032141e-05,
+      "loss": 2.7286,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9149203300476074,
+      "learning_rate": 4.573961531232379e-05,
+      "loss": 2.9862,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.118852138519287,
+      "learning_rate": 4.5708288954326175e-05,
+      "loss": 2.8026,
+      "step": 1570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7272684574127197,
+      "learning_rate": 4.5676962596328555e-05,
+      "loss": 2.8653,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5148510932922363,
+      "learning_rate": 4.5645636238330935e-05,
+      "loss": 2.8848,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6071255207061768,
+      "learning_rate": 4.5614309880333315e-05,
+      "loss": 2.8441,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5002853870391846,
+      "learning_rate": 4.5582983522335695e-05,
+      "loss": 2.8536,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4483225345611572,
+      "learning_rate": 4.5551657164338075e-05,
+      "loss": 2.6594,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.165606737136841,
+      "learning_rate": 4.5520330806340455e-05,
+      "loss": 2.6541,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3406383991241455,
+      "learning_rate": 4.5489004448342835e-05,
+      "loss": 2.8958,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9873170852661133,
+      "learning_rate": 4.5457678090345215e-05,
+      "loss": 2.7329,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.1082024574279785,
+      "learning_rate": 4.54263517323476e-05,
+      "loss": 2.6916,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.557987689971924,
+      "learning_rate": 4.539502537434998e-05,
+      "loss": 2.8576,
+      "step": 1670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5951321125030518,
+      "learning_rate": 4.536369901635236e-05,
+      "loss": 2.9174,
+      "step": 1680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.360711097717285,
+      "learning_rate": 4.533237265835474e-05,
+      "loss": 2.677,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6791539192199707,
+      "learning_rate": 4.530104630035712e-05,
+      "loss": 2.8051,
+      "step": 1700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4669108390808105,
+      "learning_rate": 4.52697199423595e-05,
+      "loss": 2.751,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8815035820007324,
+      "learning_rate": 4.523839358436188e-05,
+      "loss": 2.832,
+      "step": 1720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.085338115692139,
+      "learning_rate": 4.520706722636426e-05,
+      "loss": 2.8625,
+      "step": 1730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5768237113952637,
+      "learning_rate": 4.517574086836664e-05,
+      "loss": 2.8064,
+      "step": 1740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7768394947052,
+      "learning_rate": 4.514441451036903e-05,
+      "loss": 2.8454,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4390206336975098,
+      "learning_rate": 4.511308815237141e-05,
+      "loss": 2.8693,
+      "step": 1760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.720719575881958,
+      "learning_rate": 4.508176179437379e-05,
+      "loss": 2.7908,
+      "step": 1770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6335737705230713,
+      "learning_rate": 4.505043543637617e-05,
+      "loss": 2.7125,
+      "step": 1780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9518234729766846,
+      "learning_rate": 4.501910907837855e-05,
+      "loss": 2.6874,
+      "step": 1790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.64328670501709,
+      "learning_rate": 4.4987782720380936e-05,
+      "loss": 2.814,
+      "step": 1800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.018370151519775,
+      "learning_rate": 4.495645636238331e-05,
+      "loss": 2.8589,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9419972896575928,
+      "learning_rate": 4.492513000438569e-05,
+      "loss": 2.8111,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3817763328552246,
+      "learning_rate": 4.489380364638807e-05,
+      "loss": 2.6824,
+      "step": 1830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3612987995147705,
+      "learning_rate": 4.4862477288390456e-05,
+      "loss": 2.6577,
+      "step": 1840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4325666427612305,
+      "learning_rate": 4.4831150930392836e-05,
+      "loss": 2.7635,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.85490345954895,
+      "learning_rate": 4.4799824572395216e-05,
+      "loss": 2.5752,
+      "step": 1860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.271368026733398,
+      "learning_rate": 4.4768498214397596e-05,
+      "loss": 2.9514,
+      "step": 1870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.287911891937256,
+      "learning_rate": 4.4737171856399976e-05,
+      "loss": 2.7836,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.003190517425537,
+      "learning_rate": 4.470584549840236e-05,
+      "loss": 2.845,
+      "step": 1890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.682821750640869,
+      "learning_rate": 4.467451914040474e-05,
+      "loss": 2.8079,
+      "step": 1900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5387868881225586,
+      "learning_rate": 4.4643192782407116e-05,
+      "loss": 2.7532,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.777780055999756,
+      "learning_rate": 4.4611866424409496e-05,
+      "loss": 2.699,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1882541179656982,
+      "learning_rate": 4.458054006641188e-05,
+      "loss": 2.8145,
+      "step": 1930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4730749130249023,
+      "learning_rate": 4.454921370841426e-05,
+      "loss": 2.7746,
+      "step": 1940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5697054862976074,
+      "learning_rate": 4.451788735041664e-05,
+      "loss": 2.721,
+      "step": 1950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4670157432556152,
+      "learning_rate": 4.448656099241902e-05,
+      "loss": 2.7096,
+      "step": 1960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8159096240997314,
+      "learning_rate": 4.44552346344214e-05,
+      "loss": 2.7708,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4364335536956787,
+      "learning_rate": 4.442390827642379e-05,
+      "loss": 2.7625,
+      "step": 1980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3506743907928467,
+      "learning_rate": 4.439258191842617e-05,
+      "loss": 2.7489,
+      "step": 1990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.412740230560303,
+      "learning_rate": 4.436125556042854e-05,
+      "loss": 2.7105,
+      "step": 2000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.519390821456909,
+      "learning_rate": 4.432992920243092e-05,
+      "loss": 2.7684,
+      "step": 2010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9247891902923584,
+      "learning_rate": 4.429860284443331e-05,
+      "loss": 2.7872,
+      "step": 2020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.020471572875977,
+      "learning_rate": 4.426727648643569e-05,
+      "loss": 2.8781,
+      "step": 2030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6361448764801025,
+      "learning_rate": 4.423595012843807e-05,
+      "loss": 2.8656,
+      "step": 2040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7794973850250244,
+      "learning_rate": 4.420462377044045e-05,
+      "loss": 2.7562,
+      "step": 2050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0659725666046143,
+      "learning_rate": 4.417329741244283e-05,
+      "loss": 2.7344,
+      "step": 2060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.266765594482422,
+      "learning_rate": 4.414197105444521e-05,
+      "loss": 2.6772,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.694528102874756,
+      "learning_rate": 4.41106446964476e-05,
+      "loss": 2.8256,
+      "step": 2080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6741867065429688,
+      "learning_rate": 4.407931833844998e-05,
+      "loss": 2.6166,
+      "step": 2090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6495282649993896,
+      "learning_rate": 4.404799198045235e-05,
+      "loss": 2.8,
+      "step": 2100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7345871925354004,
+      "learning_rate": 4.401666562245473e-05,
+      "loss": 2.6533,
+      "step": 2110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.817913770675659,
+      "learning_rate": 4.398533926445712e-05,
+      "loss": 2.8853,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.097714424133301,
+      "learning_rate": 4.39540129064595e-05,
+      "loss": 2.7138,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.885183095932007,
+      "learning_rate": 4.392268654846188e-05,
+      "loss": 2.6833,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.780548095703125,
+      "learning_rate": 4.389136019046426e-05,
+      "loss": 2.5998,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.925668716430664,
+      "learning_rate": 4.386003383246664e-05,
+      "loss": 2.8803,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.053170680999756,
+      "learning_rate": 4.3828707474469024e-05,
+      "loss": 2.722,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3865647315979004,
+      "learning_rate": 4.3797381116471404e-05,
+      "loss": 2.7116,
+      "step": 2180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8181633949279785,
+      "learning_rate": 4.3766054758473784e-05,
+      "loss": 2.7195,
+      "step": 2190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9811198711395264,
+      "learning_rate": 4.373472840047616e-05,
+      "loss": 2.7277,
+      "step": 2200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.141151428222656,
+      "learning_rate": 4.3703402042478544e-05,
+      "loss": 2.7285,
+      "step": 2210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5377368927001953,
+      "learning_rate": 4.3672075684480924e-05,
+      "loss": 2.8002,
+      "step": 2220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.696154832839966,
+      "learning_rate": 4.3640749326483304e-05,
+      "loss": 2.6738,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.39733624458313,
+      "learning_rate": 4.3609422968485684e-05,
+      "loss": 2.7906,
+      "step": 2240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4818873405456543,
+      "learning_rate": 4.3578096610488064e-05,
+      "loss": 2.754,
+      "step": 2250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6994543075561523,
+      "learning_rate": 4.354677025249045e-05,
+      "loss": 2.7822,
+      "step": 2260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5926380157470703,
+      "learning_rate": 4.351544389449283e-05,
+      "loss": 2.8609,
+      "step": 2270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7803244590759277,
+      "learning_rate": 4.348411753649521e-05,
+      "loss": 2.691,
+      "step": 2280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8215737342834473,
+      "learning_rate": 4.3452791178497584e-05,
+      "loss": 2.793,
+      "step": 2290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6718356609344482,
+      "learning_rate": 4.342146482049997e-05,
+      "loss": 2.7768,
+      "step": 2300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3875210285186768,
+      "learning_rate": 4.339013846250235e-05,
+      "loss": 2.8338,
+      "step": 2310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8644275665283203,
+      "learning_rate": 4.335881210450473e-05,
+      "loss": 2.6768,
+      "step": 2320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.872359037399292,
+      "learning_rate": 4.332748574650711e-05,
+      "loss": 2.7054,
+      "step": 2330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.703859567642212,
+      "learning_rate": 4.329615938850949e-05,
+      "loss": 2.7998,
+      "step": 2340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.350034236907959,
+      "learning_rate": 4.326483303051188e-05,
+      "loss": 2.7704,
+      "step": 2350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.220928192138672,
+      "learning_rate": 4.323350667251426e-05,
+      "loss": 2.8345,
+      "step": 2360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6803479194641113,
+      "learning_rate": 4.320218031451664e-05,
+      "loss": 2.8015,
+      "step": 2370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.12473464012146,
+      "learning_rate": 4.317085395651902e-05,
+      "loss": 2.6426,
+      "step": 2380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.914316177368164,
+      "learning_rate": 4.31395275985214e-05,
+      "loss": 2.784,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1893820762634277,
+      "learning_rate": 4.310820124052378e-05,
+      "loss": 2.6237,
+      "step": 2400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.594568967819214,
+      "learning_rate": 4.307687488252616e-05,
+      "loss": 2.6872,
+      "step": 2410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.504798412322998,
+      "learning_rate": 4.304554852452854e-05,
+      "loss": 2.6686,
+      "step": 2420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.518606662750244,
+      "learning_rate": 4.301422216653092e-05,
+      "loss": 2.7731,
+      "step": 2430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7136247158050537,
+      "learning_rate": 4.2982895808533305e-05,
+      "loss": 2.7606,
+      "step": 2440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.783796787261963,
+      "learning_rate": 4.2951569450535685e-05,
+      "loss": 2.722,
+      "step": 2450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.313835382461548,
+      "learning_rate": 4.2920243092538065e-05,
+      "loss": 2.7654,
+      "step": 2460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9201815128326416,
+      "learning_rate": 4.2888916734540445e-05,
+      "loss": 2.7922,
+      "step": 2470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.019741058349609,
+      "learning_rate": 4.2857590376542825e-05,
+      "loss": 2.7475,
+      "step": 2480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6343600749969482,
+      "learning_rate": 4.2826264018545205e-05,
+      "loss": 2.7395,
+      "step": 2490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6151719093322754,
+      "learning_rate": 4.2794937660547585e-05,
+      "loss": 2.6704,
+      "step": 2500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.841431140899658,
+      "learning_rate": 4.2763611302549965e-05,
+      "loss": 2.7389,
+      "step": 2510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.631673574447632,
+      "learning_rate": 4.2732284944552346e-05,
+      "loss": 2.7652,
+      "step": 2520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.611374616622925,
+      "learning_rate": 4.270095858655473e-05,
+      "loss": 2.8215,
+      "step": 2530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.66530179977417,
+      "learning_rate": 4.266963222855711e-05,
+      "loss": 2.8931,
+      "step": 2540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8743066787719727,
+      "learning_rate": 4.263830587055949e-05,
+      "loss": 2.6281,
+      "step": 2550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.602386951446533,
+      "learning_rate": 4.260697951256187e-05,
+      "loss": 2.5779,
+      "step": 2560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.621814489364624,
+      "learning_rate": 4.257565315456425e-05,
+      "loss": 2.7075,
+      "step": 2570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.740424871444702,
+      "learning_rate": 4.254432679656664e-05,
+      "loss": 2.8009,
+      "step": 2580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.92523193359375,
+      "learning_rate": 4.251300043856901e-05,
+      "loss": 2.7026,
+      "step": 2590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.489698886871338,
+      "learning_rate": 4.248167408057139e-05,
+      "loss": 2.6533,
+      "step": 2600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.64864444732666,
+      "learning_rate": 4.245034772257377e-05,
+      "loss": 2.836,
+      "step": 2610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7577192783355713,
+      "learning_rate": 4.241902136457615e-05,
+      "loss": 2.6508,
+      "step": 2620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5754554271698,
+      "learning_rate": 4.238769500657854e-05,
+      "loss": 2.6652,
+      "step": 2630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6247026920318604,
+      "learning_rate": 4.235636864858092e-05,
+      "loss": 2.7284,
+      "step": 2640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.587557315826416,
+      "learning_rate": 4.23250422905833e-05,
+      "loss": 2.8586,
+      "step": 2650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.577183723449707,
+      "learning_rate": 4.229371593258568e-05,
+      "loss": 2.7014,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8624074459075928,
+      "learning_rate": 4.226238957458806e-05,
+      "loss": 2.7655,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9416675567626953,
+      "learning_rate": 4.223106321659044e-05,
+      "loss": 2.6485,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.667386293411255,
+      "learning_rate": 4.219973685859282e-05,
+      "loss": 2.6066,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4557063579559326,
+      "learning_rate": 4.21684105005952e-05,
+      "loss": 2.7029,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.884753704071045,
+      "learning_rate": 4.213708414259758e-05,
+      "loss": 2.657,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.569491147994995,
+      "learning_rate": 4.2105757784599966e-05,
+      "loss": 2.7833,
+      "step": 2720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7412760257720947,
+      "learning_rate": 4.2074431426602346e-05,
+      "loss": 2.7846,
+      "step": 2730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8474833965301514,
+      "learning_rate": 4.2043105068604726e-05,
+      "loss": 2.7825,
+      "step": 2740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9280788898468018,
+      "learning_rate": 4.2011778710607107e-05,
+      "loss": 2.8284,
+      "step": 2750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.080498695373535,
+      "learning_rate": 4.1980452352609487e-05,
+      "loss": 2.7753,
+      "step": 2760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6435611248016357,
+      "learning_rate": 4.194912599461187e-05,
+      "loss": 2.6859,
+      "step": 2770
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6379106044769287,
+      "learning_rate": 4.191779963661425e-05,
+      "loss": 2.6579,
+      "step": 2780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.765613079071045,
+      "learning_rate": 4.188647327861663e-05,
+      "loss": 2.6523,
+      "step": 2790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.204902648925781,
+      "learning_rate": 4.185514692061901e-05,
+      "loss": 2.5839,
+      "step": 2800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.621143102645874,
+      "learning_rate": 4.1823820562621393e-05,
+      "loss": 2.638,
+      "step": 2810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0735650062561035,
+      "learning_rate": 4.1792494204623774e-05,
+      "loss": 2.8022,
+      "step": 2820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.949160099029541,
+      "learning_rate": 4.1761167846626154e-05,
+      "loss": 2.647,
+      "step": 2830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7632877826690674,
+      "learning_rate": 4.1729841488628534e-05,
+      "loss": 2.6991,
+      "step": 2840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5531980991363525,
+      "learning_rate": 4.1698515130630914e-05,
+      "loss": 2.7372,
+      "step": 2850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4067349433898926,
+      "learning_rate": 4.16671887726333e-05,
+      "loss": 2.6866,
+      "step": 2860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5877177715301514,
+      "learning_rate": 4.163586241463568e-05,
+      "loss": 2.7729,
+      "step": 2870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.848222255706787,
+      "learning_rate": 4.1604536056638054e-05,
+      "loss": 2.658,
+      "step": 2880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.431466817855835,
+      "learning_rate": 4.1573209698640434e-05,
+      "loss": 2.8442,
+      "step": 2890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4475386142730713,
+      "learning_rate": 4.154188334064282e-05,
+      "loss": 2.6148,
+      "step": 2900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8094379901885986,
+      "learning_rate": 4.15105569826452e-05,
+      "loss": 2.5793,
+      "step": 2910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5097644329071045,
+      "learning_rate": 4.147923062464758e-05,
+      "loss": 2.6795,
+      "step": 2920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5722837448120117,
+      "learning_rate": 4.144790426664996e-05,
+      "loss": 2.6313,
+      "step": 2930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.841233491897583,
+      "learning_rate": 4.141657790865234e-05,
+      "loss": 2.6852,
+      "step": 2940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.675516128540039,
+      "learning_rate": 4.138525155065473e-05,
+      "loss": 2.7584,
+      "step": 2950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3295745849609375,
+      "learning_rate": 4.135392519265711e-05,
+      "loss": 2.6703,
+      "step": 2960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6419780254364014,
+      "learning_rate": 4.132259883465949e-05,
+      "loss": 2.7687,
+      "step": 2970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3836734294891357,
+      "learning_rate": 4.129127247666186e-05,
+      "loss": 2.754,
+      "step": 2980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.0170369148254395,
+      "learning_rate": 4.125994611866425e-05,
+      "loss": 2.8139,
+      "step": 2990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.121138572692871,
+      "learning_rate": 4.122861976066663e-05,
+      "loss": 2.8392,
+      "step": 3000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.2617597579956055,
+      "learning_rate": 4.119729340266901e-05,
+      "loss": 2.6012,
+      "step": 3010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2569031715393066,
+      "learning_rate": 4.116596704467139e-05,
+      "loss": 2.7925,
+      "step": 3020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.606506824493408,
+      "learning_rate": 4.113464068667377e-05,
+      "loss": 2.7402,
+      "step": 3030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7875235080718994,
+      "learning_rate": 4.1103314328676154e-05,
+      "loss": 2.683,
+      "step": 3040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.055811643600464,
+      "learning_rate": 4.1071987970678535e-05,
+      "loss": 2.7721,
+      "step": 3050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.042735576629639,
+      "learning_rate": 4.1040661612680915e-05,
+      "loss": 2.6119,
+      "step": 3060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.30137038230896,
+      "learning_rate": 4.100933525468329e-05,
+      "loss": 2.7351,
+      "step": 3070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.567495107650757,
+      "learning_rate": 4.097800889668567e-05,
+      "loss": 2.6587,
+      "step": 3080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2337441444396973,
+      "learning_rate": 4.0946682538688055e-05,
+      "loss": 2.7335,
+      "step": 3090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5480425357818604,
+      "learning_rate": 4.0915356180690435e-05,
+      "loss": 2.6796,
+      "step": 3100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.805624008178711,
+      "learning_rate": 4.0884029822692815e-05,
+      "loss": 2.8585,
+      "step": 3110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9506654739379883,
+      "learning_rate": 4.0852703464695195e-05,
+      "loss": 2.7387,
+      "step": 3120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.039269924163818,
+      "learning_rate": 4.0821377106697575e-05,
+      "loss": 2.7157,
+      "step": 3130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0865540504455566,
+      "learning_rate": 4.079005074869996e-05,
+      "loss": 2.8107,
+      "step": 3140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5151524543762207,
+      "learning_rate": 4.075872439070234e-05,
+      "loss": 2.768,
+      "step": 3150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7833340167999268,
+      "learning_rate": 4.072739803270472e-05,
+      "loss": 2.6107,
+      "step": 3160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.864769697189331,
+      "learning_rate": 4.0696071674707095e-05,
+      "loss": 2.8684,
+      "step": 3170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.961406707763672,
+      "learning_rate": 4.066474531670948e-05,
+      "loss": 2.8007,
+      "step": 3180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.2316365242004395,
+      "learning_rate": 4.063341895871186e-05,
+      "loss": 2.8068,
+      "step": 3190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.583223581314087,
+      "learning_rate": 4.060209260071424e-05,
+      "loss": 2.6651,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6090996265411377,
+      "learning_rate": 4.057076624271662e-05,
+      "loss": 2.7669,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1495776176452637,
+      "learning_rate": 4.0539439884719e-05,
+      "loss": 2.696,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.555459499359131,
+      "learning_rate": 4.050811352672139e-05,
+      "loss": 2.6938,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.707810163497925,
+      "learning_rate": 4.047678716872377e-05,
+      "loss": 2.8262,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3268215656280518,
+      "learning_rate": 4.044546081072615e-05,
+      "loss": 2.6679,
+      "step": 3250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.708392858505249,
+      "learning_rate": 4.041413445272853e-05,
+      "loss": 2.5789,
+      "step": 3260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.439485788345337,
+      "learning_rate": 4.038280809473091e-05,
+      "loss": 2.642,
+      "step": 3270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.641021490097046,
+      "learning_rate": 4.035148173673329e-05,
+      "loss": 2.7528,
+      "step": 3280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.777803897857666,
+      "learning_rate": 4.032015537873567e-05,
+      "loss": 2.7106,
+      "step": 3290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6226184368133545,
+      "learning_rate": 4.028882902073805e-05,
+      "loss": 2.5178,
+      "step": 3300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5166571140289307,
+      "learning_rate": 4.025750266274043e-05,
+      "loss": 2.7359,
+      "step": 3310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.474928617477417,
+      "learning_rate": 4.0226176304742816e-05,
+      "loss": 2.6186,
+      "step": 3320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6442630290985107,
+      "learning_rate": 4.0194849946745196e-05,
+      "loss": 2.8285,
+      "step": 3330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.43820858001709,
+      "learning_rate": 4.0163523588747576e-05,
+      "loss": 2.7525,
+      "step": 3340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.47707462310791,
+      "learning_rate": 4.0132197230749956e-05,
+      "loss": 2.7491,
+      "step": 3350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5458624362945557,
+      "learning_rate": 4.0100870872752336e-05,
+      "loss": 2.575,
+      "step": 3360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.445703983306885,
+      "learning_rate": 4.0069544514754716e-05,
+      "loss": 2.6054,
+      "step": 3370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1514811515808105,
+      "learning_rate": 4.0038218156757096e-05,
+      "loss": 2.5804,
+      "step": 3380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6534276008605957,
+      "learning_rate": 4.0006891798759476e-05,
+      "loss": 2.6048,
+      "step": 3390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.285447120666504,
+      "learning_rate": 3.9975565440761856e-05,
+      "loss": 2.7041,
+      "step": 3400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4054813385009766,
+      "learning_rate": 3.994423908276424e-05,
+      "loss": 2.7286,
+      "step": 3410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6479828357696533,
+      "learning_rate": 3.991291272476662e-05,
+      "loss": 2.7448,
+      "step": 3420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1932740211486816,
+      "learning_rate": 3.9881586366769e-05,
+      "loss": 2.6782,
+      "step": 3430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.734436511993408,
+      "learning_rate": 3.985026000877138e-05,
+      "loss": 2.5687,
+      "step": 3440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.552328586578369,
+      "learning_rate": 3.981893365077376e-05,
+      "loss": 2.6521,
+      "step": 3450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3161375522613525,
+      "learning_rate": 3.978760729277614e-05,
+      "loss": 2.7503,
+      "step": 3460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.588078260421753,
+      "learning_rate": 3.975628093477852e-05,
+      "loss": 2.6705,
+      "step": 3470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.963134765625,
+      "learning_rate": 3.97249545767809e-05,
+      "loss": 2.6418,
+      "step": 3480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.536984443664551,
+      "learning_rate": 3.969362821878328e-05,
+      "loss": 2.6098,
+      "step": 3490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.218698978424072,
+      "learning_rate": 3.966230186078567e-05,
+      "loss": 2.6781,
+      "step": 3500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5511672496795654,
+      "learning_rate": 3.963097550278805e-05,
+      "loss": 2.6131,
+      "step": 3510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.55438494682312,
+      "learning_rate": 3.959964914479043e-05,
+      "loss": 2.5787,
+      "step": 3520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8446435928344727,
+      "learning_rate": 3.956832278679281e-05,
+      "loss": 2.6975,
+      "step": 3530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2814626693725586,
+      "learning_rate": 3.953699642879519e-05,
+      "loss": 2.5755,
+      "step": 3540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7546284198760986,
+      "learning_rate": 3.950567007079758e-05,
+      "loss": 2.7988,
+      "step": 3550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.575570583343506,
+      "learning_rate": 3.947434371279995e-05,
+      "loss": 2.8421,
+      "step": 3560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8108553886413574,
+      "learning_rate": 3.944301735480233e-05,
+      "loss": 2.7151,
+      "step": 3570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.531538963317871,
+      "learning_rate": 3.941169099680471e-05,
+      "loss": 2.6352,
+      "step": 3580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4760665893554688,
+      "learning_rate": 3.93803646388071e-05,
+      "loss": 2.6097,
+      "step": 3590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.213020086288452,
+      "learning_rate": 3.934903828080948e-05,
+      "loss": 2.6864,
+      "step": 3600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4527084827423096,
+      "learning_rate": 3.931771192281186e-05,
+      "loss": 2.7612,
+      "step": 3610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.189217567443848,
+      "learning_rate": 3.928638556481424e-05,
+      "loss": 2.7369,
+      "step": 3620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6353416442871094,
+      "learning_rate": 3.925505920681662e-05,
+      "loss": 2.6292,
+      "step": 3630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.04370641708374,
+      "learning_rate": 3.9223732848819004e-05,
+      "loss": 2.6481,
+      "step": 3640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7645108699798584,
+      "learning_rate": 3.9192406490821384e-05,
+      "loss": 2.8236,
+      "step": 3650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6773457527160645,
+      "learning_rate": 3.916108013282376e-05,
+      "loss": 2.8449,
+      "step": 3660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6739068031311035,
+      "learning_rate": 3.912975377482614e-05,
+      "loss": 2.5603,
+      "step": 3670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2916781902313232,
+      "learning_rate": 3.909842741682852e-05,
+      "loss": 2.7602,
+      "step": 3680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.625528573989868,
+      "learning_rate": 3.9067101058830904e-05,
+      "loss": 2.6513,
+      "step": 3690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4369454383850098,
+      "learning_rate": 3.9035774700833284e-05,
+      "loss": 2.7553,
+      "step": 3700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1812028884887695,
+      "learning_rate": 3.9004448342835664e-05,
+      "loss": 2.6196,
+      "step": 3710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.635446071624756,
+      "learning_rate": 3.8973121984838044e-05,
+      "loss": 2.6613,
+      "step": 3720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7549827098846436,
+      "learning_rate": 3.8941795626840424e-05,
+      "loss": 2.522,
+      "step": 3730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.777796983718872,
+      "learning_rate": 3.891046926884281e-05,
+      "loss": 2.8314,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.964322328567505,
+      "learning_rate": 3.8879142910845184e-05,
+      "loss": 2.7249,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.494810104370117,
+      "learning_rate": 3.8847816552847564e-05,
+      "loss": 2.7195,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.552868127822876,
+      "learning_rate": 3.8816490194849944e-05,
+      "loss": 2.5469,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6302225589752197,
+      "learning_rate": 3.878516383685233e-05,
+      "loss": 2.6063,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.940479040145874,
+      "learning_rate": 3.875383747885471e-05,
+      "loss": 2.6185,
+      "step": 3790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.674966812133789,
+      "learning_rate": 3.872251112085709e-05,
+      "loss": 2.8044,
+      "step": 3800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3397276401519775,
+      "learning_rate": 3.869118476285947e-05,
+      "loss": 2.6704,
+      "step": 3810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9168059825897217,
+      "learning_rate": 3.865985840486185e-05,
+      "loss": 2.7444,
+      "step": 3820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8766160011291504,
+      "learning_rate": 3.862853204686424e-05,
+      "loss": 2.6179,
+      "step": 3830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8253989219665527,
+      "learning_rate": 3.859720568886662e-05,
+      "loss": 2.6412,
+      "step": 3840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.872537612915039,
+      "learning_rate": 3.856587933086899e-05,
+      "loss": 2.688,
+      "step": 3850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.495595932006836,
+      "learning_rate": 3.853455297287137e-05,
+      "loss": 2.6828,
+      "step": 3860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5899336338043213,
+      "learning_rate": 3.850322661487376e-05,
+      "loss": 2.7323,
+      "step": 3870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.072638034820557,
+      "learning_rate": 3.847190025687614e-05,
+      "loss": 2.7366,
+      "step": 3880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4157538414001465,
+      "learning_rate": 3.844057389887852e-05,
+      "loss": 2.6253,
+      "step": 3890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.646038055419922,
+      "learning_rate": 3.84092475408809e-05,
+      "loss": 2.6964,
+      "step": 3900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.447880744934082,
+      "learning_rate": 3.837792118288328e-05,
+      "loss": 2.6015,
+      "step": 3910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5513362884521484,
+      "learning_rate": 3.8346594824885665e-05,
+      "loss": 2.6274,
+      "step": 3920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7466373443603516,
+      "learning_rate": 3.8315268466888045e-05,
+      "loss": 2.7064,
+      "step": 3930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1654741764068604,
+      "learning_rate": 3.8283942108890425e-05,
+      "loss": 2.7,
+      "step": 3940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.263967514038086,
+      "learning_rate": 3.82526157508928e-05,
+      "loss": 2.728,
+      "step": 3950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.407287836074829,
+      "learning_rate": 3.8221289392895185e-05,
+      "loss": 2.7616,
+      "step": 3960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7538676261901855,
+      "learning_rate": 3.8189963034897565e-05,
+      "loss": 2.7607,
+      "step": 3970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.369481325149536,
+      "learning_rate": 3.8158636676899945e-05,
+      "loss": 2.708,
+      "step": 3980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2874176502227783,
+      "learning_rate": 3.8127310318902325e-05,
+      "loss": 2.5958,
+      "step": 3990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8853976726531982,
+      "learning_rate": 3.8095983960904705e-05,
+      "loss": 2.7041,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.389012336730957,
+      "learning_rate": 3.806465760290709e-05,
+      "loss": 2.5207,
+      "step": 4010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7182796001434326,
+      "learning_rate": 3.803333124490947e-05,
+      "loss": 2.6887,
+      "step": 4020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6546928882598877,
+      "learning_rate": 3.800200488691185e-05,
+      "loss": 2.7441,
+      "step": 4030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5602946281433105,
+      "learning_rate": 3.7970678528914225e-05,
+      "loss": 2.6461,
+      "step": 4040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9851481914520264,
+      "learning_rate": 3.793935217091661e-05,
+      "loss": 2.6847,
+      "step": 4050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.431507110595703,
+      "learning_rate": 3.790802581291899e-05,
+      "loss": 2.8197,
+      "step": 4060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4529969692230225,
+      "learning_rate": 3.787669945492137e-05,
+      "loss": 2.6065,
+      "step": 4070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.545023202896118,
+      "learning_rate": 3.784537309692375e-05,
+      "loss": 2.6978,
+      "step": 4080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2153308391571045,
+      "learning_rate": 3.781404673892613e-05,
+      "loss": 2.5534,
+      "step": 4090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.68514347076416,
+      "learning_rate": 3.778272038092852e-05,
+      "loss": 2.6236,
+      "step": 4100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.956491231918335,
+      "learning_rate": 3.77513940229309e-05,
+      "loss": 2.8223,
+      "step": 4110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5437729358673096,
+      "learning_rate": 3.772006766493328e-05,
+      "loss": 2.6469,
+      "step": 4120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1809537410736084,
+      "learning_rate": 3.768874130693566e-05,
+      "loss": 2.7147,
+      "step": 4130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4994957447052,
+      "learning_rate": 3.765741494893803e-05,
+      "loss": 2.685,
+      "step": 4140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3174326419830322,
+      "learning_rate": 3.762608859094042e-05,
+      "loss": 2.6743,
+      "step": 4150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.003337383270264,
+      "learning_rate": 3.75947622329428e-05,
+      "loss": 2.6788,
+      "step": 4160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.902500867843628,
+      "learning_rate": 3.756343587494518e-05,
+      "loss": 2.7302,
+      "step": 4170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3858609199523926,
+      "learning_rate": 3.753210951694756e-05,
+      "loss": 2.6986,
+      "step": 4180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.005223751068115,
+      "learning_rate": 3.750078315894994e-05,
+      "loss": 2.6507,
+      "step": 4190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3084933757781982,
+      "learning_rate": 3.7469456800952326e-05,
+      "loss": 2.7117,
+      "step": 4200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4372692108154297,
+      "learning_rate": 3.7438130442954706e-05,
+      "loss": 2.6912,
+      "step": 4210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3670456409454346,
+      "learning_rate": 3.7406804084957086e-05,
+      "loss": 2.6367,
+      "step": 4220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.919041156768799,
+      "learning_rate": 3.7375477726959466e-05,
+      "loss": 2.7021,
+      "step": 4230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.321124076843262,
+      "learning_rate": 3.7344151368961846e-05,
+      "loss": 2.5745,
+      "step": 4240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7045040130615234,
+      "learning_rate": 3.7312825010964226e-05,
+      "loss": 2.7285,
+      "step": 4250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5482301712036133,
+      "learning_rate": 3.7281498652966606e-05,
+      "loss": 2.6938,
+      "step": 4260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5309770107269287,
+      "learning_rate": 3.7250172294968986e-05,
+      "loss": 2.6954,
+      "step": 4270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.313293218612671,
+      "learning_rate": 3.7218845936971366e-05,
+      "loss": 2.7527,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3939263820648193,
+      "learning_rate": 3.718751957897375e-05,
+      "loss": 2.6962,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4195632934570312,
+      "learning_rate": 3.715619322097613e-05,
+      "loss": 2.6265,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.189661741256714,
+      "learning_rate": 3.712486686297851e-05,
+      "loss": 2.7842,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6959714889526367,
+      "learning_rate": 3.709354050498089e-05,
+      "loss": 2.6563,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.067373275756836,
+      "learning_rate": 3.706221414698327e-05,
+      "loss": 2.7157,
+      "step": 4330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.105192184448242,
+      "learning_rate": 3.703088778898565e-05,
+      "loss": 2.7298,
+      "step": 4340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.968775510787964,
+      "learning_rate": 3.699956143098803e-05,
+      "loss": 2.6267,
+      "step": 4350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.360661268234253,
+      "learning_rate": 3.696823507299041e-05,
+      "loss": 2.7032,
+      "step": 4360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.23892068862915,
+      "learning_rate": 3.693690871499279e-05,
+      "loss": 2.6217,
+      "step": 4370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.726921558380127,
+      "learning_rate": 3.690558235699518e-05,
+      "loss": 2.634,
+      "step": 4380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8494482040405273,
+      "learning_rate": 3.687425599899756e-05,
+      "loss": 2.6996,
+      "step": 4390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6485435962677,
+      "learning_rate": 3.684292964099994e-05,
+      "loss": 2.6526,
+      "step": 4400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7774016857147217,
+      "learning_rate": 3.681160328300232e-05,
+      "loss": 2.5145,
+      "step": 4410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6008520126342773,
+      "learning_rate": 3.67802769250047e-05,
+      "loss": 2.6462,
+      "step": 4420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4602198600769043,
+      "learning_rate": 3.674895056700708e-05,
+      "loss": 2.6813,
+      "step": 4430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7059261798858643,
+      "learning_rate": 3.671762420900946e-05,
+      "loss": 2.503,
+      "step": 4440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0723626613616943,
+      "learning_rate": 3.668629785101184e-05,
+      "loss": 2.615,
+      "step": 4450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.386141538619995,
+      "learning_rate": 3.665497149301422e-05,
+      "loss": 2.4334,
+      "step": 4460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4753592014312744,
+      "learning_rate": 3.662364513501661e-05,
+      "loss": 2.7665,
+      "step": 4470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.745521306991577,
+      "learning_rate": 3.659231877701899e-05,
+      "loss": 2.7753,
+      "step": 4480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5241260528564453,
+      "learning_rate": 3.656099241902137e-05,
+      "loss": 2.6306,
+      "step": 4490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.418544292449951,
+      "learning_rate": 3.652966606102375e-05,
+      "loss": 2.6334,
+      "step": 4500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.802028179168701,
+      "learning_rate": 3.649833970302613e-05,
+      "loss": 2.6841,
+      "step": 4510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.664736747741699,
+      "learning_rate": 3.6467013345028514e-05,
+      "loss": 2.4587,
+      "step": 4520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5815389156341553,
+      "learning_rate": 3.643568698703089e-05,
+      "loss": 2.7365,
+      "step": 4530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.934508800506592,
+      "learning_rate": 3.640436062903327e-05,
+      "loss": 2.6904,
+      "step": 4540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8538053035736084,
+      "learning_rate": 3.637303427103565e-05,
+      "loss": 2.5488,
+      "step": 4550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1405739784240723,
+      "learning_rate": 3.6341707913038034e-05,
+      "loss": 2.6419,
+      "step": 4560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7020983695983887,
+      "learning_rate": 3.6310381555040414e-05,
+      "loss": 2.7255,
+      "step": 4570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.042298793792725,
+      "learning_rate": 3.6279055197042794e-05,
+      "loss": 2.6505,
+      "step": 4580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.720578670501709,
+      "learning_rate": 3.6247728839045174e-05,
+      "loss": 2.596,
+      "step": 4590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2827565670013428,
+      "learning_rate": 3.6216402481047554e-05,
+      "loss": 2.7212,
+      "step": 4600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7220571041107178,
+      "learning_rate": 3.618507612304994e-05,
+      "loss": 2.7044,
+      "step": 4610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.455204725265503,
+      "learning_rate": 3.615374976505232e-05,
+      "loss": 2.7255,
+      "step": 4620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.557081460952759,
+      "learning_rate": 3.6122423407054694e-05,
+      "loss": 2.6022,
+      "step": 4630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3913400173187256,
+      "learning_rate": 3.6091097049057074e-05,
+      "loss": 2.583,
+      "step": 4640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.968672752380371,
+      "learning_rate": 3.605977069105946e-05,
+      "loss": 2.6592,
+      "step": 4650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2826192378997803,
+      "learning_rate": 3.602844433306184e-05,
+      "loss": 2.7529,
+      "step": 4660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4522457122802734,
+      "learning_rate": 3.599711797506422e-05,
+      "loss": 2.6546,
+      "step": 4670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4398117065429688,
+      "learning_rate": 3.59657916170666e-05,
+      "loss": 2.7162,
+      "step": 4680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.898144483566284,
+      "learning_rate": 3.593446525906898e-05,
+      "loss": 2.7552,
+      "step": 4690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4324584007263184,
+      "learning_rate": 3.590313890107136e-05,
+      "loss": 2.7634,
+      "step": 4700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.674661636352539,
+      "learning_rate": 3.587181254307375e-05,
+      "loss": 2.5938,
+      "step": 4710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.56758975982666,
+      "learning_rate": 3.584048618507613e-05,
+      "loss": 2.7581,
+      "step": 4720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8987762928009033,
+      "learning_rate": 3.58091598270785e-05,
+      "loss": 2.7237,
+      "step": 4730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.58803653717041,
+      "learning_rate": 3.577783346908088e-05,
+      "loss": 2.6871,
+      "step": 4740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4172046184539795,
+      "learning_rate": 3.574650711108327e-05,
+      "loss": 2.5975,
+      "step": 4750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9828925132751465,
+      "learning_rate": 3.571518075308565e-05,
+      "loss": 2.8031,
+      "step": 4760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6485350131988525,
+      "learning_rate": 3.568385439508803e-05,
+      "loss": 2.6183,
+      "step": 4770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6514647006988525,
+      "learning_rate": 3.565252803709041e-05,
+      "loss": 2.7714,
+      "step": 4780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.168116569519043,
+      "learning_rate": 3.562120167909279e-05,
+      "loss": 2.6368,
+      "step": 4790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9774527549743652,
+      "learning_rate": 3.5589875321095175e-05,
+      "loss": 2.5899,
+      "step": 4800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.036356210708618,
+      "learning_rate": 3.5558548963097555e-05,
+      "loss": 2.5359,
+      "step": 4810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.019016742706299,
+      "learning_rate": 3.552722260509993e-05,
+      "loss": 2.5671,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3434457778930664,
+      "learning_rate": 3.549589624710231e-05,
+      "loss": 2.5714,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.958373785018921,
+      "learning_rate": 3.5464569889104695e-05,
+      "loss": 2.7045,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5077250003814697,
+      "learning_rate": 3.5433243531107075e-05,
+      "loss": 2.5583,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.2609686851501465,
+      "learning_rate": 3.5401917173109455e-05,
+      "loss": 2.5628,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3074331283569336,
+      "learning_rate": 3.5370590815111835e-05,
+      "loss": 2.6041,
+      "step": 4870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6179749965667725,
+      "learning_rate": 3.5339264457114216e-05,
+      "loss": 2.6826,
+      "step": 4880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8938159942626953,
+      "learning_rate": 3.53079380991166e-05,
+      "loss": 2.642,
+      "step": 4890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.054030418395996,
+      "learning_rate": 3.527661174111898e-05,
+      "loss": 2.5999,
+      "step": 4900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.450873374938965,
+      "learning_rate": 3.524528538312136e-05,
+      "loss": 2.757,
+      "step": 4910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.52093243598938,
+      "learning_rate": 3.5213959025123736e-05,
+      "loss": 2.6478,
+      "step": 4920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4646759033203125,
+      "learning_rate": 3.518263266712612e-05,
+      "loss": 2.6925,
+      "step": 4930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8345937728881836,
+      "learning_rate": 3.51513063091285e-05,
+      "loss": 2.5773,
+      "step": 4940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.86470890045166,
+      "learning_rate": 3.511997995113088e-05,
+      "loss": 2.7242,
+      "step": 4950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.446380853652954,
+      "learning_rate": 3.508865359313326e-05,
+      "loss": 2.6121,
+      "step": 4960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9185121059417725,
+      "learning_rate": 3.505732723513564e-05,
+      "loss": 2.8018,
+      "step": 4970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.135132789611816,
+      "learning_rate": 3.502600087713803e-05,
+      "loss": 2.6571,
+      "step": 4980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8625378608703613,
+      "learning_rate": 3.499467451914041e-05,
+      "loss": 2.6498,
+      "step": 4990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.788400650024414,
+      "learning_rate": 3.496334816114279e-05,
+      "loss": 2.5997,
+      "step": 5000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.110319137573242,
+      "learning_rate": 3.493202180314517e-05,
+      "loss": 2.7054,
+      "step": 5010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1062545776367188,
+      "learning_rate": 3.490069544514755e-05,
+      "loss": 2.6907,
+      "step": 5020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9589688777923584,
+      "learning_rate": 3.486936908714993e-05,
+      "loss": 2.6445,
+      "step": 5030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5143542289733887,
+      "learning_rate": 3.483804272915231e-05,
+      "loss": 2.6587,
+      "step": 5040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.875641822814941,
+      "learning_rate": 3.480671637115469e-05,
+      "loss": 2.5477,
+      "step": 5050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7877542972564697,
+      "learning_rate": 3.477539001315707e-05,
+      "loss": 2.7636,
+      "step": 5060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.071690559387207,
+      "learning_rate": 3.4744063655159456e-05,
+      "loss": 2.6171,
+      "step": 5070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.912278413772583,
+      "learning_rate": 3.4712737297161836e-05,
+      "loss": 2.7926,
+      "step": 5080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.779947519302368,
+      "learning_rate": 3.4681410939164216e-05,
+      "loss": 2.7714,
+      "step": 5090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9601123332977295,
+      "learning_rate": 3.4650084581166597e-05,
+      "loss": 2.7388,
+      "step": 5100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5037424564361572,
+      "learning_rate": 3.4618758223168977e-05,
+      "loss": 2.5711,
+      "step": 5110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4753992557525635,
+      "learning_rate": 3.4587431865171357e-05,
+      "loss": 2.6326,
+      "step": 5120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.041739463806152,
+      "learning_rate": 3.4556105507173737e-05,
+      "loss": 2.5689,
+      "step": 5130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.938133478164673,
+      "learning_rate": 3.452477914917612e-05,
+      "loss": 2.5507,
+      "step": 5140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.775325059890747,
+      "learning_rate": 3.44934527911785e-05,
+      "loss": 2.6409,
+      "step": 5150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3595011234283447,
+      "learning_rate": 3.4462126433180883e-05,
+      "loss": 2.6732,
+      "step": 5160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6374683380126953,
+      "learning_rate": 3.4430800075183263e-05,
+      "loss": 2.5409,
+      "step": 5170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4543371200561523,
+      "learning_rate": 3.4399473717185644e-05,
+      "loss": 2.6854,
+      "step": 5180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.120486259460449,
+      "learning_rate": 3.4368147359188024e-05,
+      "loss": 2.5458,
+      "step": 5190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.764808177947998,
+      "learning_rate": 3.4336821001190404e-05,
+      "loss": 2.6321,
+      "step": 5200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5418665409088135,
+      "learning_rate": 3.4305494643192784e-05,
+      "loss": 2.6508,
+      "step": 5210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4328222274780273,
+      "learning_rate": 3.4274168285195164e-05,
+      "loss": 2.6223,
+      "step": 5220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.281841516494751,
+      "learning_rate": 3.4242841927197544e-05,
+      "loss": 2.6083,
+      "step": 5230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.453336238861084,
+      "learning_rate": 3.4211515569199924e-05,
+      "loss": 2.7111,
+      "step": 5240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7157461643218994,
+      "learning_rate": 3.4180189211202304e-05,
+      "loss": 2.5758,
+      "step": 5250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.228655815124512,
+      "learning_rate": 3.414886285320469e-05,
+      "loss": 2.3715,
+      "step": 5260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6937756538391113,
+      "learning_rate": 3.411753649520707e-05,
+      "loss": 2.68,
+      "step": 5270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7848565578460693,
+      "learning_rate": 3.408621013720945e-05,
+      "loss": 2.5128,
+      "step": 5280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.687631607055664,
+      "learning_rate": 3.405488377921183e-05,
+      "loss": 2.6281,
+      "step": 5290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8058853149414062,
+      "learning_rate": 3.402355742121421e-05,
+      "loss": 2.6807,
+      "step": 5300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.656662940979004,
+      "learning_rate": 3.399223106321659e-05,
+      "loss": 2.5632,
+      "step": 5310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.496729850769043,
+      "learning_rate": 3.396090470521897e-05,
+      "loss": 2.5288,
+      "step": 5320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4001872539520264,
+      "learning_rate": 3.392957834722135e-05,
+      "loss": 2.5592,
+      "step": 5330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.141587495803833,
+      "learning_rate": 3.389825198922373e-05,
+      "loss": 2.6705,
+      "step": 5340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4357247352600098,
+      "learning_rate": 3.386692563122612e-05,
+      "loss": 2.6251,
+      "step": 5350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3765175342559814,
+      "learning_rate": 3.38355992732285e-05,
+      "loss": 2.7846,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.75774884223938,
+      "learning_rate": 3.380427291523088e-05,
+      "loss": 2.7333,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.415477991104126,
+      "learning_rate": 3.377294655723326e-05,
+      "loss": 2.7039,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3807625770568848,
+      "learning_rate": 3.374162019923564e-05,
+      "loss": 2.634,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7736899852752686,
+      "learning_rate": 3.3710293841238025e-05,
+      "loss": 2.3838,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7280807495117188,
+      "learning_rate": 3.36789674832404e-05,
+      "loss": 2.5235,
+      "step": 5410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.045708179473877,
+      "learning_rate": 3.364764112524278e-05,
+      "loss": 2.4621,
+      "step": 5420
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.7396812438964844,
+      "learning_rate": 3.361631476724516e-05,
+      "loss": 2.4619,
+      "step": 5430
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.0114898681640625,
+      "learning_rate": 3.3584988409247545e-05,
+      "loss": 2.4536,
+      "step": 5440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.227764844894409,
+      "learning_rate": 3.3553662051249925e-05,
+      "loss": 2.5467,
+      "step": 5450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.8109536170959473,
+      "learning_rate": 3.3522335693252305e-05,
+      "loss": 2.5285,
+      "step": 5460
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.8032093048095703,
+      "learning_rate": 3.3491009335254685e-05,
+      "loss": 2.5643,
+      "step": 5470
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.6691324710845947,
+      "learning_rate": 3.3459682977257065e-05,
+      "loss": 2.4793,
+      "step": 5480
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.8392367362976074,
+      "learning_rate": 3.342835661925945e-05,
+      "loss": 2.4332,
+      "step": 5490
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.776864528656006,
+      "learning_rate": 3.3397030261261825e-05,
+      "loss": 2.3964,
+      "step": 5500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.805351734161377,
+      "learning_rate": 3.3365703903264205e-05,
+      "loss": 2.4288,
+      "step": 5510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.900991439819336,
+      "learning_rate": 3.3334377545266585e-05,
+      "loss": 2.535,
+      "step": 5520
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.3974769115448,
+      "learning_rate": 3.330305118726897e-05,
+      "loss": 2.4969,
+      "step": 5530
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.656438112258911,
+      "learning_rate": 3.327172482927135e-05,
+      "loss": 2.3034,
+      "step": 5540
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.124125957489014,
+      "learning_rate": 3.324039847127373e-05,
+      "loss": 2.3805,
+      "step": 5550
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.620126247406006,
+      "learning_rate": 3.320907211327611e-05,
+      "loss": 2.5268,
+      "step": 5560
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.6136839389801025,
+      "learning_rate": 3.317774575527849e-05,
+      "loss": 2.478,
+      "step": 5570
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.039578914642334,
+      "learning_rate": 3.314641939728088e-05,
+      "loss": 2.5355,
+      "step": 5580
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.641757011413574,
+      "learning_rate": 3.311509303928326e-05,
+      "loss": 2.4235,
+      "step": 5590
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.343202590942383,
+      "learning_rate": 3.308376668128563e-05,
+      "loss": 2.4022,
+      "step": 5600
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.0208258628845215,
+      "learning_rate": 3.305244032328801e-05,
+      "loss": 2.5053,
+      "step": 5610
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.351205348968506,
+      "learning_rate": 3.30211139652904e-05,
+      "loss": 2.5123,
+      "step": 5620
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.987982988357544,
+      "learning_rate": 3.298978760729278e-05,
+      "loss": 2.5238,
+      "step": 5630
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.0602312088012695,
+      "learning_rate": 3.295846124929516e-05,
+      "loss": 2.4661,
+      "step": 5640
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.855670928955078,
+      "learning_rate": 3.292713489129754e-05,
+      "loss": 2.5684,
+      "step": 5650
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.313904762268066,
+      "learning_rate": 3.289580853329992e-05,
+      "loss": 2.59,
+      "step": 5660
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.257472515106201,
+      "learning_rate": 3.2864482175302306e-05,
+      "loss": 2.4152,
+      "step": 5670
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.372264385223389,
+      "learning_rate": 3.2833155817304686e-05,
+      "loss": 2.6152,
+      "step": 5680
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.014118194580078,
+      "learning_rate": 3.2801829459307066e-05,
+      "loss": 2.4703,
+      "step": 5690
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.317049503326416,
+      "learning_rate": 3.277050310130944e-05,
+      "loss": 2.4456,
+      "step": 5700
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.42697811126709,
+      "learning_rate": 3.2739176743311826e-05,
+      "loss": 2.4755,
+      "step": 5710
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7758564949035645,
+      "learning_rate": 3.2707850385314206e-05,
+      "loss": 2.4876,
+      "step": 5720
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.636563539505005,
+      "learning_rate": 3.2676524027316586e-05,
+      "loss": 2.3994,
+      "step": 5730
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.024879455566406,
+      "learning_rate": 3.2645197669318966e-05,
+      "loss": 2.3539,
+      "step": 5740
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.519440650939941,
+      "learning_rate": 3.2613871311321346e-05,
+      "loss": 2.4389,
+      "step": 5750
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.379619598388672,
+      "learning_rate": 3.2582544953323726e-05,
+      "loss": 2.3534,
+      "step": 5760
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.5998525619506836,
+      "learning_rate": 3.255121859532611e-05,
+      "loss": 2.4591,
+      "step": 5770
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.6781418323516846,
+      "learning_rate": 3.251989223732849e-05,
+      "loss": 2.4257,
+      "step": 5780
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.661477088928223,
+      "learning_rate": 3.2488565879330866e-05,
+      "loss": 2.5522,
+      "step": 5790
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.9432196617126465,
+      "learning_rate": 3.2457239521333246e-05,
+      "loss": 2.414,
+      "step": 5800
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.805156707763672,
+      "learning_rate": 3.242591316333563e-05,
+      "loss": 2.5464,
+      "step": 5810
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.142496109008789,
+      "learning_rate": 3.239458680533801e-05,
+      "loss": 2.611,
+      "step": 5820
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.8504600524902344,
+      "learning_rate": 3.236326044734039e-05,
+      "loss": 2.4074,
+      "step": 5830
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.016257286071777,
+      "learning_rate": 3.233193408934277e-05,
+      "loss": 2.5023,
+      "step": 5840
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.938784599304199,
+      "learning_rate": 3.230060773134515e-05,
+      "loss": 2.3188,
+      "step": 5850
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.088259696960449,
+      "learning_rate": 3.226928137334754e-05,
+      "loss": 2.5935,
+      "step": 5860
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8898637294769287,
+      "learning_rate": 3.223795501534992e-05,
+      "loss": 2.4305,
+      "step": 5870
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.501204013824463,
+      "learning_rate": 3.22066286573523e-05,
+      "loss": 2.4884,
+      "step": 5880
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9667720794677734,
+      "learning_rate": 3.217530229935467e-05,
+      "loss": 2.433,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.108957290649414,
+      "learning_rate": 3.214397594135706e-05,
+      "loss": 2.604,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.968045234680176,
+      "learning_rate": 3.211264958335944e-05,
+      "loss": 2.4587,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.952162742614746,
+      "learning_rate": 3.208132322536182e-05,
+      "loss": 2.4285,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.0700178146362305,
+      "learning_rate": 3.20499968673642e-05,
+      "loss": 2.6214,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.97859787940979,
+      "learning_rate": 3.201867050936658e-05,
+      "loss": 2.5462,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.449424743652344,
+      "learning_rate": 3.198734415136897e-05,
+      "loss": 2.4818,
+      "step": 5950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.160023212432861,
+      "learning_rate": 3.195601779337135e-05,
+      "loss": 2.4427,
+      "step": 5960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.9818923473358154,
+      "learning_rate": 3.192469143537373e-05,
+      "loss": 2.4902,
+      "step": 5970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.702857494354248,
+      "learning_rate": 3.189336507737611e-05,
+      "loss": 2.3095,
+      "step": 5980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.913214683532715,
+      "learning_rate": 3.186203871937849e-05,
+      "loss": 2.5468,
+      "step": 5990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.280374526977539,
+      "learning_rate": 3.183071236138087e-05,
+      "loss": 2.4706,
+      "step": 6000
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.013718128204346,
+      "learning_rate": 3.179938600338325e-05,
+      "loss": 2.3338,
+      "step": 6010
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.405649662017822,
+      "learning_rate": 3.176805964538563e-05,
+      "loss": 2.5071,
+      "step": 6020
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.047056198120117,
+      "learning_rate": 3.173673328738801e-05,
+      "loss": 2.3017,
+      "step": 6030
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.8888051509857178,
+      "learning_rate": 3.1705406929390394e-05,
+      "loss": 2.4258,
+      "step": 6040
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.3164544105529785,
+      "learning_rate": 3.1674080571392774e-05,
+      "loss": 2.3265,
+      "step": 6050
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.50031042098999,
+      "learning_rate": 3.1642754213395154e-05,
+      "loss": 2.4742,
+      "step": 6060
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.77244234085083,
+      "learning_rate": 3.1611427855397534e-05,
+      "loss": 2.5043,
+      "step": 6070
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.641749382019043,
+      "learning_rate": 3.1580101497399914e-05,
+      "loss": 2.4812,
+      "step": 6080
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.318567752838135,
+      "learning_rate": 3.1548775139402294e-05,
+      "loss": 2.513,
+      "step": 6090
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.369333744049072,
+      "learning_rate": 3.1517448781404674e-05,
+      "loss": 2.3888,
+      "step": 6100
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.169533729553223,
+      "learning_rate": 3.1486122423407054e-05,
+      "loss": 2.5043,
+      "step": 6110
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.253003120422363,
+      "learning_rate": 3.1454796065409434e-05,
+      "loss": 2.4792,
+      "step": 6120
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.288143634796143,
+      "learning_rate": 3.142346970741182e-05,
+      "loss": 2.4291,
+      "step": 6130
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.162696361541748,
+      "learning_rate": 3.13921433494142e-05,
+      "loss": 2.4592,
+      "step": 6140
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.5401482582092285,
+      "learning_rate": 3.136081699141658e-05,
+      "loss": 2.5767,
+      "step": 6150
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.044467926025391,
+      "learning_rate": 3.132949063341896e-05,
+      "loss": 2.4809,
+      "step": 6160
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.036025524139404,
+      "learning_rate": 3.129816427542134e-05,
+      "loss": 2.608,
+      "step": 6170
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.690577983856201,
+      "learning_rate": 3.126683791742372e-05,
+      "loss": 2.5302,
+      "step": 6180
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.768272399902344,
+      "learning_rate": 3.12355115594261e-05,
+      "loss": 2.3404,
+      "step": 6190
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.854149580001831,
+      "learning_rate": 3.120418520142848e-05,
+      "loss": 2.4891,
+      "step": 6200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.387482643127441,
+      "learning_rate": 3.117285884343086e-05,
+      "loss": 2.4742,
+      "step": 6210
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.609062671661377,
+      "learning_rate": 3.114153248543325e-05,
+      "loss": 2.3332,
+      "step": 6220
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.023780822753906,
+      "learning_rate": 3.111020612743563e-05,
+      "loss": 2.3789,
+      "step": 6230
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.633157253265381,
+      "learning_rate": 3.107887976943801e-05,
+      "loss": 2.5869,
+      "step": 6240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.554311275482178,
+      "learning_rate": 3.104755341144039e-05,
+      "loss": 2.4892,
+      "step": 6250
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.6120660305023193,
+      "learning_rate": 3.101622705344277e-05,
+      "loss": 2.3765,
+      "step": 6260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.1312336921691895,
+      "learning_rate": 3.0984900695445155e-05,
+      "loss": 2.3906,
+      "step": 6270
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6260271072387695,
+      "learning_rate": 3.095357433744753e-05,
+      "loss": 2.4732,
+      "step": 6280
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.927614212036133,
+      "learning_rate": 3.092224797944991e-05,
+      "loss": 2.4285,
+      "step": 6290
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.142879009246826,
+      "learning_rate": 3.089092162145229e-05,
+      "loss": 2.5053,
+      "step": 6300
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.045203685760498,
+      "learning_rate": 3.085959526345467e-05,
+      "loss": 2.3809,
+      "step": 6310
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.602118492126465,
+      "learning_rate": 3.0828268905457055e-05,
+      "loss": 2.4972,
+      "step": 6320
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.50063419342041,
+      "learning_rate": 3.0796942547459435e-05,
+      "loss": 2.5565,
+      "step": 6330
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.114358901977539,
+      "learning_rate": 3.0765616189461815e-05,
+      "loss": 2.3903,
+      "step": 6340
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.6293530464172363,
+      "learning_rate": 3.0734289831464195e-05,
+      "loss": 2.4091,
+      "step": 6350
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8283889293670654,
+      "learning_rate": 3.0702963473466575e-05,
+      "loss": 2.6608,
+      "step": 6360
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8779544830322266,
+      "learning_rate": 3.067163711546896e-05,
+      "loss": 2.3576,
+      "step": 6370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8340346813201904,
+      "learning_rate": 3.0640310757471335e-05,
+      "loss": 2.4389,
+      "step": 6380
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.959573268890381,
+      "learning_rate": 3.0608984399473715e-05,
+      "loss": 2.5481,
+      "step": 6390
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6378843784332275,
+      "learning_rate": 3.0577658041476095e-05,
+      "loss": 2.3975,
+      "step": 6400
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.129283428192139,
+      "learning_rate": 3.054633168347848e-05,
+      "loss": 2.4949,
+      "step": 6410
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.536279201507568,
+      "learning_rate": 3.0515005325480862e-05,
+      "loss": 2.4757,
+      "step": 6420
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.183682918548584,
+      "learning_rate": 3.0483678967483242e-05,
+      "loss": 2.4207,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.7908713817596436,
+      "learning_rate": 3.0452352609485622e-05,
+      "loss": 2.4127,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6448137760162354,
+      "learning_rate": 3.0421026251488006e-05,
+      "loss": 2.4096,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.8170764446258545,
+      "learning_rate": 3.0389699893490386e-05,
+      "loss": 2.6154,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.10953426361084,
+      "learning_rate": 3.0358373535492762e-05,
+      "loss": 2.4478,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.442537307739258,
+      "learning_rate": 3.0327047177495142e-05,
+      "loss": 2.6607,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.768803834915161,
+      "learning_rate": 3.0295720819497526e-05,
+      "loss": 2.5009,
+      "step": 6490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.953761339187622,
+      "learning_rate": 3.0264394461499906e-05,
+      "loss": 2.4567,
+      "step": 6500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.973198413848877,
+      "learning_rate": 3.023306810350229e-05,
+      "loss": 2.5144,
+      "step": 6510
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.184175491333008,
+      "learning_rate": 3.020174174550467e-05,
+      "loss": 2.3374,
+      "step": 6520
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.105236530303955,
+      "learning_rate": 3.017041538750705e-05,
+      "loss": 2.4755,
+      "step": 6530
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.367380619049072,
+      "learning_rate": 3.0139089029509433e-05,
+      "loss": 2.4868,
+      "step": 6540
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.8320703506469727,
+      "learning_rate": 3.0107762671511813e-05,
+      "loss": 2.4634,
+      "step": 6550
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.322927474975586,
+      "learning_rate": 3.0076436313514196e-05,
+      "loss": 2.5089,
+      "step": 6560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.073949813842773,
+      "learning_rate": 3.004510995551657e-05,
+      "loss": 2.3301,
+      "step": 6570
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.9154913425445557,
+      "learning_rate": 3.0013783597518953e-05,
+      "loss": 2.5612,
+      "step": 6580
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.1147332191467285,
+      "learning_rate": 2.9982457239521333e-05,
+      "loss": 2.4923,
+      "step": 6590
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.687198162078857,
+      "learning_rate": 2.9951130881523716e-05,
+      "loss": 2.5775,
+      "step": 6600
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.9940314292907715,
+      "learning_rate": 2.9919804523526096e-05,
+      "loss": 2.489,
+      "step": 6610
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.8620872497558594,
+      "learning_rate": 2.9888478165528476e-05,
+      "loss": 2.337,
+      "step": 6620
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.9539003372192383,
+      "learning_rate": 2.985715180753086e-05,
+      "loss": 2.4726,
+      "step": 6630
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.662155628204346,
+      "learning_rate": 2.982582544953324e-05,
+      "loss": 2.4849,
+      "step": 6640
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.394937038421631,
+      "learning_rate": 2.9794499091535623e-05,
+      "loss": 2.4574,
+      "step": 6650
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.711225509643555,
+      "learning_rate": 2.9763172733538003e-05,
+      "loss": 2.5811,
+      "step": 6660
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.110074996948242,
+      "learning_rate": 2.973184637554038e-05,
+      "loss": 2.4939,
+      "step": 6670
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.297110557556152,
+      "learning_rate": 2.970052001754276e-05,
+      "loss": 2.4746,
+      "step": 6680
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.9598119258880615,
+      "learning_rate": 2.966919365954514e-05,
+      "loss": 2.4667,
+      "step": 6690
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.400740146636963,
+      "learning_rate": 2.9637867301547523e-05,
+      "loss": 2.4556,
+      "step": 6700
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.5050714015960693,
+      "learning_rate": 2.9606540943549903e-05,
+      "loss": 2.4305,
+      "step": 6710
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.708401679992676,
+      "learning_rate": 2.9575214585552287e-05,
+      "loss": 2.4872,
+      "step": 6720
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.355490207672119,
+      "learning_rate": 2.9543888227554667e-05,
+      "loss": 2.4645,
+      "step": 6730
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.8205530643463135,
+      "learning_rate": 2.9512561869557047e-05,
+      "loss": 2.5601,
+      "step": 6740
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.833449363708496,
+      "learning_rate": 2.948123551155943e-05,
+      "loss": 2.4713,
+      "step": 6750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.864391803741455,
+      "learning_rate": 2.944990915356181e-05,
+      "loss": 2.594,
+      "step": 6760
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.996546506881714,
+      "learning_rate": 2.9418582795564187e-05,
+      "loss": 2.5044,
+      "step": 6770
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.637711048126221,
+      "learning_rate": 2.9387256437566567e-05,
+      "loss": 2.4977,
+      "step": 6780
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.310656547546387,
+      "learning_rate": 2.935593007956895e-05,
+      "loss": 2.4352,
+      "step": 6790
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.202852725982666,
+      "learning_rate": 2.932460372157133e-05,
+      "loss": 2.4125,
+      "step": 6800
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.8896448612213135,
+      "learning_rate": 2.9293277363573714e-05,
+      "loss": 2.3875,
+      "step": 6810
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.873081684112549,
+      "learning_rate": 2.9261951005576094e-05,
+      "loss": 2.4571,
+      "step": 6820
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.322564601898193,
+      "learning_rate": 2.9230624647578474e-05,
+      "loss": 2.3468,
+      "step": 6830
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.952484130859375,
+      "learning_rate": 2.9199298289580857e-05,
+      "loss": 2.4347,
+      "step": 6840
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.476901531219482,
+      "learning_rate": 2.9167971931583237e-05,
+      "loss": 2.6337,
+      "step": 6850
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.968850612640381,
+      "learning_rate": 2.9136645573585614e-05,
+      "loss": 2.466,
+      "step": 6860
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.549422264099121,
+      "learning_rate": 2.9105319215587994e-05,
+      "loss": 2.5071,
+      "step": 6870
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.74324107170105,
+      "learning_rate": 2.9073992857590377e-05,
+      "loss": 2.4761,
+      "step": 6880
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.241300106048584,
+      "learning_rate": 2.9042666499592757e-05,
+      "loss": 2.403,
+      "step": 6890
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.1481428146362305,
+      "learning_rate": 2.901134014159514e-05,
+      "loss": 2.3425,
+      "step": 6900
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.6267499923706055,
+      "learning_rate": 2.898001378359752e-05,
+      "loss": 2.4772,
+      "step": 6910
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.275341033935547,
+      "learning_rate": 2.89486874255999e-05,
+      "loss": 2.4042,
+      "step": 6920
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.284965991973877,
+      "learning_rate": 2.8917361067602284e-05,
+      "loss": 2.52,
+      "step": 6930
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.403405666351318,
+      "learning_rate": 2.8886034709604664e-05,
+      "loss": 2.4763,
+      "step": 6940
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.16456413269043,
+      "learning_rate": 2.8854708351607044e-05,
+      "loss": 2.485,
+      "step": 6950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.8839504718780518,
+      "learning_rate": 2.882338199360942e-05,
+      "loss": 2.4893,
+      "step": 6960
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.768056392669678,
+      "learning_rate": 2.8792055635611804e-05,
+      "loss": 2.4582,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.094647407531738,
+      "learning_rate": 2.8760729277614184e-05,
+      "loss": 2.4824,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.018396377563477,
+      "learning_rate": 2.8729402919616564e-05,
+      "loss": 2.5067,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.8706161975860596,
+      "learning_rate": 2.8698076561618948e-05,
+      "loss": 2.5126,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.7700846195220947,
+      "learning_rate": 2.8666750203621328e-05,
+      "loss": 2.5013,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.8400371074676514,
+      "learning_rate": 2.863542384562371e-05,
+      "loss": 2.5477,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.360779285430908,
+      "learning_rate": 2.860409748762609e-05,
+      "loss": 2.5295,
+      "step": 7030
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.506185531616211,
+      "learning_rate": 2.857277112962847e-05,
+      "loss": 2.5114,
+      "step": 7040
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.395011901855469,
+      "learning_rate": 2.8541444771630855e-05,
+      "loss": 2.508,
+      "step": 7050
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.438186168670654,
+      "learning_rate": 2.851011841363323e-05,
+      "loss": 2.345,
+      "step": 7060
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.920180082321167,
+      "learning_rate": 2.847879205563561e-05,
+      "loss": 2.4989,
+      "step": 7070
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.980530023574829,
+      "learning_rate": 2.844746569763799e-05,
+      "loss": 2.5219,
+      "step": 7080
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.988844871520996,
+      "learning_rate": 2.8416139339640375e-05,
+      "loss": 2.5288,
+      "step": 7090
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.9802753925323486,
+      "learning_rate": 2.8384812981642755e-05,
+      "loss": 2.3894,
+      "step": 7100
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.512762069702148,
+      "learning_rate": 2.835348662364514e-05,
+      "loss": 2.4125,
+      "step": 7110
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.419162273406982,
+      "learning_rate": 2.832216026564752e-05,
+      "loss": 2.4308,
+      "step": 7120
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.053534984588623,
+      "learning_rate": 2.82908339076499e-05,
+      "loss": 2.4747,
+      "step": 7130
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.382496356964111,
+      "learning_rate": 2.8259507549652282e-05,
+      "loss": 2.4205,
+      "step": 7140
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.047353267669678,
+      "learning_rate": 2.8228181191654662e-05,
+      "loss": 2.4959,
+      "step": 7150
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.391164302825928,
+      "learning_rate": 2.819685483365704e-05,
+      "loss": 2.4672,
+      "step": 7160
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.1165618896484375,
+      "learning_rate": 2.816552847565942e-05,
+      "loss": 2.3797,
+      "step": 7170
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.1894307136535645,
+      "learning_rate": 2.8134202117661802e-05,
+      "loss": 2.5119,
+      "step": 7180
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.277898788452148,
+      "learning_rate": 2.8102875759664182e-05,
+      "loss": 2.398,
+      "step": 7190
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.254609107971191,
+      "learning_rate": 2.8071549401666562e-05,
+      "loss": 2.4263,
+      "step": 7200
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.395539283752441,
+      "learning_rate": 2.8040223043668945e-05,
+      "loss": 2.5494,
+      "step": 7210
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.832606792449951,
+      "learning_rate": 2.8008896685671325e-05,
+      "loss": 2.4564,
+      "step": 7220
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.776379108428955,
+      "learning_rate": 2.797757032767371e-05,
+      "loss": 2.4843,
+      "step": 7230
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.442601442337036,
+      "learning_rate": 2.794624396967609e-05,
+      "loss": 2.3466,
+      "step": 7240
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.7504169940948486,
+      "learning_rate": 2.7914917611678466e-05,
+      "loss": 2.479,
+      "step": 7250
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.110808849334717,
+      "learning_rate": 2.7883591253680846e-05,
+      "loss": 2.4306,
+      "step": 7260
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.7822816371917725,
+      "learning_rate": 2.785226489568323e-05,
+      "loss": 2.4555,
+      "step": 7270
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.296886444091797,
+      "learning_rate": 2.782093853768561e-05,
+      "loss": 2.4778,
+      "step": 7280
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.053720951080322,
+      "learning_rate": 2.778961217968799e-05,
+      "loss": 2.4027,
+      "step": 7290
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.153025150299072,
+      "learning_rate": 2.7758285821690372e-05,
+      "loss": 2.5618,
+      "step": 7300
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.224099159240723,
+      "learning_rate": 2.7726959463692753e-05,
+      "loss": 2.4304,
+      "step": 7310
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.171821594238281,
+      "learning_rate": 2.7695633105695136e-05,
+      "loss": 2.4833,
+      "step": 7320
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.8762876987457275,
+      "learning_rate": 2.7664306747697516e-05,
+      "loss": 2.407,
+      "step": 7330
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.1427693367004395,
+      "learning_rate": 2.7632980389699896e-05,
+      "loss": 2.4689,
+      "step": 7340
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.639638423919678,
+      "learning_rate": 2.7601654031702273e-05,
+      "loss": 2.5578,
+      "step": 7350
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.923847198486328,
+      "learning_rate": 2.7570327673704656e-05,
+      "loss": 2.4735,
+      "step": 7360
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.8497817516326904,
+      "learning_rate": 2.7539001315707036e-05,
+      "loss": 2.3363,
+      "step": 7370
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.234973907470703,
+      "learning_rate": 2.7507674957709416e-05,
+      "loss": 2.3606,
+      "step": 7380
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.026918411254883,
+      "learning_rate": 2.74763485997118e-05,
+      "loss": 2.3884,
+      "step": 7390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.6976168155670166,
+      "learning_rate": 2.744502224171418e-05,
+      "loss": 2.508,
+      "step": 7400
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.062636852264404,
+      "learning_rate": 2.7413695883716563e-05,
+      "loss": 2.5993,
+      "step": 7410
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.24471664428711,
+      "learning_rate": 2.7382369525718943e-05,
+      "loss": 2.5709,
+      "step": 7420
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.350656509399414,
+      "learning_rate": 2.7351043167721323e-05,
+      "loss": 2.4863,
+      "step": 7430
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.904040813446045,
+      "learning_rate": 2.7319716809723706e-05,
+      "loss": 2.4859,
+      "step": 7440
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.8008460998535156,
+      "learning_rate": 2.728839045172608e-05,
+      "loss": 2.5323,
+      "step": 7450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.462713241577148,
+      "learning_rate": 2.7257064093728463e-05,
+      "loss": 2.4299,
+      "step": 7460
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.533032417297363,
+      "learning_rate": 2.7225737735730843e-05,
+      "loss": 2.4732,
+      "step": 7470
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.7745182514190674,
+      "learning_rate": 2.7194411377733227e-05,
+      "loss": 2.2655,
+      "step": 7480
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.192812919616699,
+      "learning_rate": 2.7163085019735607e-05,
+      "loss": 2.415,
+      "step": 7490
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.448957443237305,
+      "learning_rate": 2.7131758661737987e-05,
+      "loss": 2.4624,
+      "step": 7500
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.633139610290527,
+      "learning_rate": 2.710043230374037e-05,
+      "loss": 2.3715,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.003988742828369,
+      "learning_rate": 2.706910594574275e-05,
+      "loss": 2.5053,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.706925392150879,
+      "learning_rate": 2.7037779587745133e-05,
+      "loss": 2.463,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.274069309234619,
+      "learning_rate": 2.7006453229747507e-05,
+      "loss": 2.368,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.7519917488098145,
+      "learning_rate": 2.697512687174989e-05,
+      "loss": 2.4108,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.982792615890503,
+      "learning_rate": 2.694380051375227e-05,
+      "loss": 2.4604,
+      "step": 7560
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.7987496852874756,
+      "learning_rate": 2.6912474155754654e-05,
+      "loss": 2.3429,
+      "step": 7570
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.573733806610107,
+      "learning_rate": 2.6881147797757034e-05,
+      "loss": 2.3859,
+      "step": 7580
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.3863019943237305,
+      "learning_rate": 2.6849821439759414e-05,
+      "loss": 2.5498,
+      "step": 7590
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.614287376403809,
+      "learning_rate": 2.6818495081761797e-05,
+      "loss": 2.5396,
+      "step": 7600
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.737358570098877,
+      "learning_rate": 2.6787168723764177e-05,
+      "loss": 2.4409,
+      "step": 7610
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.048407554626465,
+      "learning_rate": 2.675584236576656e-05,
+      "loss": 2.4081,
+      "step": 7620
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.848973274230957,
+      "learning_rate": 2.672451600776894e-05,
+      "loss": 2.343,
+      "step": 7630
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.802037715911865,
+      "learning_rate": 2.6693189649771317e-05,
+      "loss": 2.3406,
+      "step": 7640
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.5075483322143555,
+      "learning_rate": 2.6661863291773697e-05,
+      "loss": 2.4503,
+      "step": 7650
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.523159503936768,
+      "learning_rate": 2.663053693377608e-05,
+      "loss": 2.4643,
+      "step": 7660
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.8925390243530273,
+      "learning_rate": 2.659921057577846e-05,
+      "loss": 2.3056,
+      "step": 7670
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.5294060707092285,
+      "learning_rate": 2.656788421778084e-05,
+      "loss": 2.4761,
+      "step": 7680
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.151029586791992,
+      "learning_rate": 2.6536557859783224e-05,
+      "loss": 2.6412,
+      "step": 7690
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9961705207824707,
+      "learning_rate": 2.6505231501785604e-05,
+      "loss": 2.4344,
+      "step": 7700
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.595808982849121,
+      "learning_rate": 2.6473905143787988e-05,
+      "loss": 2.3549,
+      "step": 7710
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.139496803283691,
+      "learning_rate": 2.6442578785790368e-05,
+      "loss": 2.5274,
+      "step": 7720
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9915213584899902,
+      "learning_rate": 2.6411252427792748e-05,
+      "loss": 2.3668,
+      "step": 7730
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.448828220367432,
+      "learning_rate": 2.6379926069795124e-05,
+      "loss": 2.2979,
+      "step": 7740
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.070132255554199,
+      "learning_rate": 2.6348599711797504e-05,
+      "loss": 2.3898,
+      "step": 7750
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.449703216552734,
+      "learning_rate": 2.6317273353799888e-05,
+      "loss": 2.4488,
+      "step": 7760
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.869356393814087,
+      "learning_rate": 2.6285946995802268e-05,
+      "loss": 2.3731,
+      "step": 7770
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.151821136474609,
+      "learning_rate": 2.625462063780465e-05,
+      "loss": 2.5609,
+      "step": 7780
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.265530109405518,
+      "learning_rate": 2.622329427980703e-05,
+      "loss": 2.4425,
+      "step": 7790
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.279960632324219,
+      "learning_rate": 2.619196792180941e-05,
+      "loss": 2.429,
+      "step": 7800
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.096279144287109,
+      "learning_rate": 2.6160641563811795e-05,
+      "loss": 2.4357,
+      "step": 7810
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.8605334758758545,
+      "learning_rate": 2.6129315205814175e-05,
+      "loss": 2.4386,
+      "step": 7820
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.625364065170288,
+      "learning_rate": 2.6097988847816558e-05,
+      "loss": 2.3049,
+      "step": 7830
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.029632091522217,
+      "learning_rate": 2.606666248981893e-05,
+      "loss": 2.4797,
+      "step": 7840
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.971041202545166,
+      "learning_rate": 2.6035336131821315e-05,
+      "loss": 2.2626,
+      "step": 7850
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.361209392547607,
+      "learning_rate": 2.6004009773823695e-05,
+      "loss": 2.4605,
+      "step": 7860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.867220163345337,
+      "learning_rate": 2.5972683415826078e-05,
+      "loss": 2.4137,
+      "step": 7870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.512621879577637,
+      "learning_rate": 2.5941357057828458e-05,
+      "loss": 2.3925,
+      "step": 7880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.05349588394165,
+      "learning_rate": 2.5910030699830838e-05,
+      "loss": 2.4041,
+      "step": 7890
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.112064361572266,
+      "learning_rate": 2.5878704341833222e-05,
+      "loss": 2.3245,
+      "step": 7900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.162067413330078,
+      "learning_rate": 2.5847377983835602e-05,
+      "loss": 2.3881,
+      "step": 7910
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.984915018081665,
+      "learning_rate": 2.5816051625837985e-05,
+      "loss": 2.5021,
+      "step": 7920
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.4609036445617676,
+      "learning_rate": 2.578472526784036e-05,
+      "loss": 2.3664,
+      "step": 7930
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.9925968647003174,
+      "learning_rate": 2.5753398909842742e-05,
+      "loss": 2.3428,
+      "step": 7940
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.392812252044678,
+      "learning_rate": 2.5722072551845122e-05,
+      "loss": 2.4966,
+      "step": 7950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.691169261932373,
+      "learning_rate": 2.5690746193847505e-05,
+      "loss": 2.4003,
+      "step": 7960
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.080898284912109,
+      "learning_rate": 2.5659419835849885e-05,
+      "loss": 2.4125,
+      "step": 7970
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.430429935455322,
+      "learning_rate": 2.5628093477852265e-05,
+      "loss": 2.4005,
+      "step": 7980
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.048476696014404,
+      "learning_rate": 2.559676711985465e-05,
+      "loss": 2.4376,
+      "step": 7990
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.243685722351074,
+      "learning_rate": 2.556544076185703e-05,
+      "loss": 2.5779,
+      "step": 8000
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.010811805725098,
+      "learning_rate": 2.553411440385941e-05,
+      "loss": 2.4601,
+      "step": 8010
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.074519634246826,
+      "learning_rate": 2.5502788045861792e-05,
+      "loss": 2.4129,
+      "step": 8020
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.494532108306885,
+      "learning_rate": 2.547146168786417e-05,
+      "loss": 2.4148,
+      "step": 8030
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.979499340057373,
+      "learning_rate": 2.544013532986655e-05,
+      "loss": 2.4027,
+      "step": 8040
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.250908851623535,
+      "learning_rate": 2.540880897186893e-05,
+      "loss": 2.3471,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.406290531158447,
+      "learning_rate": 2.5377482613871312e-05,
+      "loss": 2.5687,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.947685480117798,
+      "learning_rate": 2.5346156255873692e-05,
+      "loss": 2.3511,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.436914920806885,
+      "learning_rate": 2.5314829897876076e-05,
+      "loss": 2.4439,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.601666212081909,
+      "learning_rate": 2.5283503539878456e-05,
+      "loss": 2.396,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.406430721282959,
+      "learning_rate": 2.5252177181880836e-05,
+      "loss": 2.3042,
+      "step": 8100
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.84181809425354,
+      "learning_rate": 2.522085082388322e-05,
+      "loss": 2.4797,
+      "step": 8110
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.066086769104004,
+      "learning_rate": 2.51895244658856e-05,
+      "loss": 2.595,
+      "step": 8120
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.644285678863525,
+      "learning_rate": 2.5158198107887976e-05,
+      "loss": 2.5138,
+      "step": 8130
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.4279303550720215,
+      "learning_rate": 2.5126871749890356e-05,
+      "loss": 2.4851,
+      "step": 8140
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.7944159507751465,
+      "learning_rate": 2.509554539189274e-05,
+      "loss": 2.3989,
+      "step": 8150
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.593160152435303,
+      "learning_rate": 2.506421903389512e-05,
+      "loss": 2.5637,
+      "step": 8160
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.316919326782227,
+      "learning_rate": 2.5032892675897503e-05,
+      "loss": 2.3603,
+      "step": 8170
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.496823787689209,
+      "learning_rate": 2.5001566317899883e-05,
+      "loss": 2.4756,
+      "step": 8180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.07798957824707,
+      "learning_rate": 2.4970239959902263e-05,
+      "loss": 2.4217,
+      "step": 8190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.062137126922607,
+      "learning_rate": 2.4938913601904643e-05,
+      "loss": 2.2771,
+      "step": 8200
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.37224817276001,
+      "learning_rate": 2.4907587243907023e-05,
+      "loss": 2.4003,
+      "step": 8210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.260566234588623,
+      "learning_rate": 2.4876260885909406e-05,
+      "loss": 2.3725,
+      "step": 8220
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.098892688751221,
+      "learning_rate": 2.4844934527911786e-05,
+      "loss": 2.3704,
+      "step": 8230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.625776767730713,
+      "learning_rate": 2.481360816991417e-05,
+      "loss": 2.4508,
+      "step": 8240
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.206084728240967,
+      "learning_rate": 2.4782281811916546e-05,
+      "loss": 2.4504,
+      "step": 8250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.482570171356201,
+      "learning_rate": 2.4750955453918927e-05,
+      "loss": 2.3864,
+      "step": 8260
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.5410237312316895,
+      "learning_rate": 2.471962909592131e-05,
+      "loss": 2.5781,
+      "step": 8270
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.9175031185150146,
+      "learning_rate": 2.468830273792369e-05,
+      "loss": 2.3522,
+      "step": 8280
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.778284549713135,
+      "learning_rate": 2.4656976379926073e-05,
+      "loss": 2.4545,
+      "step": 8290
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.8703789710998535,
+      "learning_rate": 2.462565002192845e-05,
+      "loss": 2.4581,
+      "step": 8300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.1762847900390625,
+      "learning_rate": 2.4594323663930833e-05,
+      "loss": 2.3582,
+      "step": 8310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.193213939666748,
+      "learning_rate": 2.4562997305933213e-05,
+      "loss": 2.2968,
+      "step": 8320
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.130049705505371,
+      "learning_rate": 2.4531670947935593e-05,
+      "loss": 2.5317,
+      "step": 8330
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.380124092102051,
+      "learning_rate": 2.4500344589937977e-05,
+      "loss": 2.4211,
+      "step": 8340
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.307916164398193,
+      "learning_rate": 2.4469018231940354e-05,
+      "loss": 2.3788,
+      "step": 8350
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.4227166175842285,
+      "learning_rate": 2.4437691873942737e-05,
+      "loss": 2.3199,
+      "step": 8360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.434095859527588,
+      "learning_rate": 2.4406365515945117e-05,
+      "loss": 2.4138,
+      "step": 8370
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.016805171966553,
+      "learning_rate": 2.43750391579475e-05,
+      "loss": 2.3639,
+      "step": 8380
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.9864070415496826,
+      "learning_rate": 2.4343712799949877e-05,
+      "loss": 2.4281,
+      "step": 8390
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.868408441543579,
+      "learning_rate": 2.431238644195226e-05,
+      "loss": 2.4809,
+      "step": 8400
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.65122127532959,
+      "learning_rate": 2.428106008395464e-05,
+      "loss": 2.4681,
+      "step": 8410
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.935698986053467,
+      "learning_rate": 2.424973372595702e-05,
+      "loss": 2.3455,
+      "step": 8420
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.314015865325928,
+      "learning_rate": 2.4218407367959404e-05,
+      "loss": 2.4527,
+      "step": 8430
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.124387264251709,
+      "learning_rate": 2.418708100996178e-05,
+      "loss": 2.4122,
+      "step": 8440
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.7196121215820312,
+      "learning_rate": 2.4155754651964164e-05,
+      "loss": 2.5219,
+      "step": 8450
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.566164016723633,
+      "learning_rate": 2.4124428293966544e-05,
+      "loss": 2.424,
+      "step": 8460
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.448875904083252,
+      "learning_rate": 2.4093101935968927e-05,
+      "loss": 2.4276,
+      "step": 8470
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.657510280609131,
+      "learning_rate": 2.4061775577971307e-05,
+      "loss": 2.4654,
+      "step": 8480
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.8274145126342773,
+      "learning_rate": 2.4030449219973684e-05,
+      "loss": 2.4654,
+      "step": 8490
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.3402228355407715,
+      "learning_rate": 2.3999122861976068e-05,
+      "loss": 2.3597,
+      "step": 8500
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.7467360496520996,
+      "learning_rate": 2.3967796503978448e-05,
+      "loss": 2.4173,
+      "step": 8510
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.307443618774414,
+      "learning_rate": 2.393647014598083e-05,
+      "loss": 2.4235,
+      "step": 8520
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.339209079742432,
+      "learning_rate": 2.390514378798321e-05,
+      "loss": 2.5007,
+      "step": 8530
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.1376118659973145,
+      "learning_rate": 2.387381742998559e-05,
+      "loss": 2.3295,
+      "step": 8540
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.090506076812744,
+      "learning_rate": 2.384249107198797e-05,
+      "loss": 2.4239,
+      "step": 8550
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.034234046936035,
+      "learning_rate": 2.381116471399035e-05,
+      "loss": 2.419,
+      "step": 8560
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.6677346229553223,
+      "learning_rate": 2.3779838355992735e-05,
+      "loss": 2.5574,
+      "step": 8570
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.422360897064209,
+      "learning_rate": 2.3748511997995115e-05,
+      "loss": 2.3951,
+      "step": 8580
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.14889669418335,
+      "learning_rate": 2.3717185639997495e-05,
+      "loss": 2.4181,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.343019962310791,
+      "learning_rate": 2.3685859281999875e-05,
+      "loss": 2.4356,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.155813217163086,
+      "learning_rate": 2.3654532924002258e-05,
+      "loss": 2.4328,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.325819969177246,
+      "learning_rate": 2.3623206566004638e-05,
+      "loss": 2.4753,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.4720540046691895,
+      "learning_rate": 2.3591880208007018e-05,
+      "loss": 2.4074,
+      "step": 8630
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.326481342315674,
+      "learning_rate": 2.3560553850009398e-05,
+      "loss": 2.2994,
+      "step": 8640
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.1999592781066895,
+      "learning_rate": 2.3529227492011778e-05,
+      "loss": 2.276,
+      "step": 8650
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.166476726531982,
+      "learning_rate": 2.349790113401416e-05,
+      "loss": 2.4736,
+      "step": 8660
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.361721515655518,
+      "learning_rate": 2.346657477601654e-05,
+      "loss": 2.4374,
+      "step": 8670
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.636282920837402,
+      "learning_rate": 2.3435248418018925e-05,
+      "loss": 2.3267,
+      "step": 8680
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.051875591278076,
+      "learning_rate": 2.34039220600213e-05,
+      "loss": 2.3327,
+      "step": 8690
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.308687686920166,
+      "learning_rate": 2.3372595702023685e-05,
+      "loss": 2.3804,
+      "step": 8700
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.162394046783447,
+      "learning_rate": 2.3341269344026065e-05,
+      "loss": 2.487,
+      "step": 8710
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.165082931518555,
+      "learning_rate": 2.3309942986028445e-05,
+      "loss": 2.5037,
+      "step": 8720
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.127786636352539,
+      "learning_rate": 2.3278616628030825e-05,
+      "loss": 2.3867,
+      "step": 8730
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.055369853973389,
+      "learning_rate": 2.3247290270033205e-05,
+      "loss": 2.4696,
+      "step": 8740
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.151759624481201,
+      "learning_rate": 2.321596391203559e-05,
+      "loss": 2.4747,
+      "step": 8750
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.65687894821167,
+      "learning_rate": 2.318463755403797e-05,
+      "loss": 2.3878,
+      "step": 8760
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.260522365570068,
+      "learning_rate": 2.315331119604035e-05,
+      "loss": 2.5407,
+      "step": 8770
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.139651775360107,
+      "learning_rate": 2.312198483804273e-05,
+      "loss": 2.4235,
+      "step": 8780
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.276207447052002,
+      "learning_rate": 2.309065848004511e-05,
+      "loss": 2.2987,
+      "step": 8790
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.2793731689453125,
+      "learning_rate": 2.3059332122047492e-05,
+      "loss": 2.4383,
+      "step": 8800
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.7700252532958984,
+      "learning_rate": 2.3028005764049872e-05,
+      "loss": 2.4058,
+      "step": 8810
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.340311527252197,
+      "learning_rate": 2.2996679406052256e-05,
+      "loss": 2.4854,
+      "step": 8820
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.079639434814453,
+      "learning_rate": 2.2965353048054632e-05,
+      "loss": 2.4113,
+      "step": 8830
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.728581666946411,
+      "learning_rate": 2.2934026690057016e-05,
+      "loss": 2.33,
+      "step": 8840
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.095951080322266,
+      "learning_rate": 2.2902700332059396e-05,
+      "loss": 2.4925,
+      "step": 8850
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.893881320953369,
+      "learning_rate": 2.2871373974061776e-05,
+      "loss": 2.5037,
+      "step": 8860
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.797602653503418,
+      "learning_rate": 2.284004761606416e-05,
+      "loss": 2.4481,
+      "step": 8870
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.614184856414795,
+      "learning_rate": 2.2808721258066536e-05,
+      "loss": 2.4042,
+      "step": 8880
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.491405487060547,
+      "learning_rate": 2.277739490006892e-05,
+      "loss": 2.3421,
+      "step": 8890
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.135850429534912,
+      "learning_rate": 2.27460685420713e-05,
+      "loss": 2.3792,
+      "step": 8900
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.752901554107666,
+      "learning_rate": 2.2714742184073683e-05,
+      "loss": 2.4197,
+      "step": 8910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.097724914550781,
+      "learning_rate": 2.2683415826076063e-05,
+      "loss": 2.5074,
+      "step": 8920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.123813629150391,
+      "learning_rate": 2.2652089468078443e-05,
+      "loss": 2.4965,
+      "step": 8930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.163042068481445,
+      "learning_rate": 2.2620763110080823e-05,
+      "loss": 2.4533,
+      "step": 8940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.805618762969971,
+      "learning_rate": 2.2589436752083203e-05,
+      "loss": 2.3576,
+      "step": 8950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.58022403717041,
+      "learning_rate": 2.2558110394085586e-05,
+      "loss": 2.4423,
+      "step": 8960
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.180849552154541,
+      "learning_rate": 2.2526784036087966e-05,
+      "loss": 2.4075,
+      "step": 8970
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.48970365524292,
+      "learning_rate": 2.2495457678090346e-05,
+      "loss": 2.4105,
+      "step": 8980
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.819067478179932,
+      "learning_rate": 2.2464131320092726e-05,
+      "loss": 2.3278,
+      "step": 8990
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.588673114776611,
+      "learning_rate": 2.243280496209511e-05,
+      "loss": 2.5266,
+      "step": 9000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.151445388793945,
+      "learning_rate": 2.240147860409749e-05,
+      "loss": 2.6258,
+      "step": 9010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.227921485900879,
+      "learning_rate": 2.237015224609987e-05,
+      "loss": 2.2707,
+      "step": 9020
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.091616153717041,
+      "learning_rate": 2.233882588810225e-05,
+      "loss": 2.3745,
+      "step": 9030
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.434237480163574,
+      "learning_rate": 2.230749953010463e-05,
+      "loss": 2.5214,
+      "step": 9040
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.7377941608428955,
+      "learning_rate": 2.2276173172107013e-05,
+      "loss": 2.2664,
+      "step": 9050
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.9352807998657227,
+      "learning_rate": 2.2244846814109393e-05,
+      "loss": 2.4196,
+      "step": 9060
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.058120250701904,
+      "learning_rate": 2.2213520456111773e-05,
+      "loss": 2.3696,
+      "step": 9070
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.753314733505249,
+      "learning_rate": 2.2182194098114153e-05,
+      "loss": 2.4226,
+      "step": 9080
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.035268783569336,
+      "learning_rate": 2.2150867740116533e-05,
+      "loss": 2.3935,
+      "step": 9090
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.0359649658203125,
+      "learning_rate": 2.2119541382118917e-05,
+      "loss": 2.5171,
+      "step": 9100
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.143360614776611,
+      "learning_rate": 2.2088215024121297e-05,
+      "loss": 2.5488,
+      "step": 9110
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.005861759185791,
+      "learning_rate": 2.2056888666123677e-05,
+      "loss": 2.5157,
+      "step": 9120
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.021326541900635,
+      "learning_rate": 2.2025562308126057e-05,
+      "loss": 2.3483,
+      "step": 9130
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.683558225631714,
+      "learning_rate": 2.199423595012844e-05,
+      "loss": 2.5039,
+      "step": 9140
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.9170446395874023,
+      "learning_rate": 2.196290959213082e-05,
+      "loss": 2.3853,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.003520965576172,
+      "learning_rate": 2.19315832341332e-05,
+      "loss": 2.5109,
+      "step": 9160
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.8243813514709473,
+      "learning_rate": 2.190025687613558e-05,
+      "loss": 2.4832,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.661709308624268,
+      "learning_rate": 2.186893051813796e-05,
+      "loss": 2.3179,
+      "step": 9180
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.192775249481201,
+      "learning_rate": 2.1837604160140344e-05,
+      "loss": 2.4681,
+      "step": 9190
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.422582149505615,
+      "learning_rate": 2.1806277802142724e-05,
+      "loss": 2.3914,
+      "step": 9200
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.069379806518555,
+      "learning_rate": 2.1774951444145107e-05,
+      "loss": 2.4434,
+      "step": 9210
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.300215721130371,
+      "learning_rate": 2.1743625086147484e-05,
+      "loss": 2.3971,
+      "step": 9220
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.8108959197998047,
+      "learning_rate": 2.1712298728149867e-05,
+      "loss": 2.4546,
+      "step": 9230
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.656400442123413,
+      "learning_rate": 2.1680972370152247e-05,
+      "loss": 2.4079,
+      "step": 9240
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.267639636993408,
+      "learning_rate": 2.1649646012154627e-05,
+      "loss": 2.4183,
+      "step": 9250
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.378291130065918,
+      "learning_rate": 2.161831965415701e-05,
+      "loss": 2.3521,
+      "step": 9260
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.273284435272217,
+      "learning_rate": 2.1586993296159387e-05,
+      "loss": 2.3965,
+      "step": 9270
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.253837585449219,
+      "learning_rate": 2.155566693816177e-05,
+      "loss": 2.4458,
+      "step": 9280
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.412284851074219,
+      "learning_rate": 2.152434058016415e-05,
+      "loss": 2.5076,
+      "step": 9290
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.642493724822998,
+      "learning_rate": 2.149301422216653e-05,
+      "loss": 2.4012,
+      "step": 9300
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.4505698680877686,
+      "learning_rate": 2.1461687864168914e-05,
+      "loss": 2.3663,
+      "step": 9310
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.132535934448242,
+      "learning_rate": 2.143036150617129e-05,
+      "loss": 2.3603,
+      "step": 9320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.732874870300293,
+      "learning_rate": 2.1399035148173674e-05,
+      "loss": 2.2906,
+      "step": 9330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.370763301849365,
+      "learning_rate": 2.1367708790176054e-05,
+      "loss": 2.4297,
+      "step": 9340
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.205658435821533,
+      "learning_rate": 2.1336382432178438e-05,
+      "loss": 2.4445,
+      "step": 9350
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.8486382961273193,
+      "learning_rate": 2.1305056074180818e-05,
+      "loss": 2.4696,
+      "step": 9360
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.221676349639893,
+      "learning_rate": 2.1273729716183198e-05,
+      "loss": 2.3604,
+      "step": 9370
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.008775234222412,
+      "learning_rate": 2.1242403358185578e-05,
+      "loss": 2.5529,
+      "step": 9380
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.488894939422607,
+      "learning_rate": 2.1211077000187958e-05,
+      "loss": 2.3793,
+      "step": 9390
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.996288776397705,
+      "learning_rate": 2.117975064219034e-05,
+      "loss": 2.4756,
+      "step": 9400
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.614337921142578,
+      "learning_rate": 2.114842428419272e-05,
+      "loss": 2.3483,
+      "step": 9410
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.302055835723877,
+      "learning_rate": 2.11170979261951e-05,
+      "loss": 2.4792,
+      "step": 9420
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.433431148529053,
+      "learning_rate": 2.108577156819748e-05,
+      "loss": 2.4674,
+      "step": 9430
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.374996662139893,
+      "learning_rate": 2.1054445210199865e-05,
+      "loss": 2.3494,
+      "step": 9440
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.1850361824035645,
+      "learning_rate": 2.1023118852202245e-05,
+      "loss": 2.4677,
+      "step": 9450
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.117648601531982,
+      "learning_rate": 2.0991792494204625e-05,
+      "loss": 2.3398,
+      "step": 9460
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.606243848800659,
+      "learning_rate": 2.0960466136207005e-05,
+      "loss": 2.3634,
+      "step": 9470
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.0432658195495605,
+      "learning_rate": 2.0929139778209385e-05,
+      "loss": 2.5315,
+      "step": 9480
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.853435516357422,
+      "learning_rate": 2.089781342021177e-05,
+      "loss": 2.4228,
+      "step": 9490
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.112980365753174,
+      "learning_rate": 2.086648706221415e-05,
+      "loss": 2.5621,
+      "step": 9500
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.1322808265686035,
+      "learning_rate": 2.083516070421653e-05,
+      "loss": 2.3932,
+      "step": 9510
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.9664454460144043,
+      "learning_rate": 2.080383434621891e-05,
+      "loss": 2.4334,
+      "step": 9520
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.619282245635986,
+      "learning_rate": 2.0772507988221292e-05,
+      "loss": 2.4343,
+      "step": 9530
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.839085578918457,
+      "learning_rate": 2.0741181630223672e-05,
+      "loss": 2.4187,
+      "step": 9540
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.345643520355225,
+      "learning_rate": 2.0709855272226052e-05,
+      "loss": 2.4387,
+      "step": 9550
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.0575642585754395,
+      "learning_rate": 2.0678528914228432e-05,
+      "loss": 2.4043,
+      "step": 9560
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.259551048278809,
+      "learning_rate": 2.0647202556230812e-05,
+      "loss": 2.3613,
+      "step": 9570
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.213980674743652,
+      "learning_rate": 2.0615876198233195e-05,
+      "loss": 2.4696,
+      "step": 9580
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.4075446128845215,
+      "learning_rate": 2.0584549840235576e-05,
+      "loss": 2.4833,
+      "step": 9590
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.569279193878174,
+      "learning_rate": 2.0553223482237956e-05,
+      "loss": 2.389,
+      "step": 9600
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.837458372116089,
+      "learning_rate": 2.0521897124240336e-05,
+      "loss": 2.3882,
+      "step": 9610
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.220689296722412,
+      "learning_rate": 2.0490570766242716e-05,
+      "loss": 2.5108,
+      "step": 9620
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.240020751953125,
+      "learning_rate": 2.04592444082451e-05,
+      "loss": 2.5004,
+      "step": 9630
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.560367584228516,
+      "learning_rate": 2.042791805024748e-05,
+      "loss": 2.5024,
+      "step": 9640
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.0892181396484375,
+      "learning_rate": 2.0396591692249862e-05,
+      "loss": 2.3545,
+      "step": 9650
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.34207010269165,
+      "learning_rate": 2.036526533425224e-05,
+      "loss": 2.4782,
+      "step": 9660
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.363237380981445,
+      "learning_rate": 2.0333938976254623e-05,
+      "loss": 2.206,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.8488094806671143,
+      "learning_rate": 2.0302612618257003e-05,
+      "loss": 2.3918,
+      "step": 9680
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.453420639038086,
+      "learning_rate": 2.0271286260259383e-05,
+      "loss": 2.3648,
+      "step": 9690
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.7727625370025635,
+      "learning_rate": 2.0239959902261766e-05,
+      "loss": 2.4054,
+      "step": 9700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.444394588470459,
+      "learning_rate": 2.0208633544264143e-05,
+      "loss": 2.4831,
+      "step": 9710
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.399291038513184,
+      "learning_rate": 2.0177307186266526e-05,
+      "loss": 2.4292,
+      "step": 9720
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.340801239013672,
+      "learning_rate": 2.0145980828268906e-05,
+      "loss": 2.2941,
+      "step": 9730
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.915050029754639,
+      "learning_rate": 2.011465447027129e-05,
+      "loss": 2.446,
+      "step": 9740
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.6448071002960205,
+      "learning_rate": 2.008332811227367e-05,
+      "loss": 2.3937,
+      "step": 9750
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.50324010848999,
+      "learning_rate": 2.005200175427605e-05,
+      "loss": 2.3799,
+      "step": 9760
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.91646146774292,
+      "learning_rate": 2.002067539627843e-05,
+      "loss": 2.3239,
+      "step": 9770
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.287485599517822,
+      "learning_rate": 1.998934903828081e-05,
+      "loss": 2.3565,
+      "step": 9780
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.101718425750732,
+      "learning_rate": 1.9958022680283193e-05,
+      "loss": 2.3881,
+      "step": 9790
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.803049087524414,
+      "learning_rate": 1.992669632228557e-05,
+      "loss": 2.4375,
+      "step": 9800
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.953117370605469,
+      "learning_rate": 1.9895369964287953e-05,
+      "loss": 2.3769,
+      "step": 9810
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.480542182922363,
+      "learning_rate": 1.9864043606290333e-05,
+      "loss": 2.3742,
+      "step": 9820
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.368782997131348,
+      "learning_rate": 1.9832717248292713e-05,
+      "loss": 2.5032,
+      "step": 9830
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.5232367515563965,
+      "learning_rate": 1.9801390890295097e-05,
+      "loss": 2.3914,
+      "step": 9840
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.795730113983154,
+      "learning_rate": 1.9770064532297473e-05,
+      "loss": 2.4551,
+      "step": 9850
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.36059045791626,
+      "learning_rate": 1.9738738174299857e-05,
+      "loss": 2.426,
+      "step": 9860
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.606714725494385,
+      "learning_rate": 1.9707411816302237e-05,
+      "loss": 2.3973,
+      "step": 9870
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.512914180755615,
+      "learning_rate": 1.967608545830462e-05,
+      "loss": 2.4423,
+      "step": 9880
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.6559979915618896,
+      "learning_rate": 1.9644759100307e-05,
+      "loss": 2.2358,
+      "step": 9890
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.248470306396484,
+      "learning_rate": 1.961343274230938e-05,
+      "loss": 2.2735,
+      "step": 9900
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.629327774047852,
+      "learning_rate": 1.958210638431176e-05,
+      "loss": 2.4155,
+      "step": 9910
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.973908424377441,
+      "learning_rate": 1.955078002631414e-05,
+      "loss": 2.5043,
+      "step": 9920
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.580214977264404,
+      "learning_rate": 1.9519453668316524e-05,
+      "loss": 2.5079,
+      "step": 9930
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.727945804595947,
+      "learning_rate": 1.9488127310318904e-05,
+      "loss": 2.3416,
+      "step": 9940
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.331240177154541,
+      "learning_rate": 1.9456800952321284e-05,
+      "loss": 2.424,
+      "step": 9950
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.183077812194824,
+      "learning_rate": 1.9425474594323664e-05,
+      "loss": 2.4149,
+      "step": 9960
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.234660625457764,
+      "learning_rate": 1.9394148236326047e-05,
+      "loss": 2.4433,
+      "step": 9970
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.186022758483887,
+      "learning_rate": 1.9362821878328427e-05,
+      "loss": 2.559,
+      "step": 9980
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.2144856452941895,
+      "learning_rate": 1.9331495520330807e-05,
+      "loss": 2.4598,
+      "step": 9990
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.981110095977783,
+      "learning_rate": 1.9300169162333187e-05,
+      "loss": 2.4261,
+      "step": 10000
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.171727180480957,
+      "learning_rate": 1.9268842804335567e-05,
+      "loss": 2.3655,
+      "step": 10010
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.13902473449707,
+      "learning_rate": 1.923751644633795e-05,
+      "loss": 2.5224,
+      "step": 10020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.296168804168701,
+      "learning_rate": 1.920619008834033e-05,
+      "loss": 2.3685,
+      "step": 10030
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.683605194091797,
+      "learning_rate": 1.9174863730342714e-05,
+      "loss": 2.4102,
+      "step": 10040
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.063732147216797,
+      "learning_rate": 1.914353737234509e-05,
+      "loss": 2.2953,
+      "step": 10050
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.8384222984313965,
+      "learning_rate": 1.9112211014347474e-05,
+      "loss": 2.501,
+      "step": 10060
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.450084209442139,
+      "learning_rate": 1.9080884656349854e-05,
+      "loss": 2.4856,
+      "step": 10070
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.239566802978516,
+      "learning_rate": 1.9049558298352234e-05,
+      "loss": 2.448,
+      "step": 10080
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.171853065490723,
+      "learning_rate": 1.9018231940354618e-05,
+      "loss": 2.4451,
+      "step": 10090
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.816874980926514,
+      "learning_rate": 1.8986905582356994e-05,
+      "loss": 2.4613,
+      "step": 10100
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.647317886352539,
+      "learning_rate": 1.8955579224359378e-05,
+      "loss": 2.5905,
+      "step": 10110
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.123816967010498,
+      "learning_rate": 1.8924252866361758e-05,
+      "loss": 2.4489,
+      "step": 10120
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.358800411224365,
+      "learning_rate": 1.8892926508364138e-05,
+      "loss": 2.4604,
+      "step": 10130
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.496860980987549,
+      "learning_rate": 1.8861600150366518e-05,
+      "loss": 2.376,
+      "step": 10140
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.996311902999878,
+      "learning_rate": 1.8830273792368898e-05,
+      "loss": 2.5195,
+      "step": 10150
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.059746742248535,
+      "learning_rate": 1.879894743437128e-05,
+      "loss": 2.346,
+      "step": 10160
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.107177257537842,
+      "learning_rate": 1.876762107637366e-05,
+      "loss": 2.4174,
+      "step": 10170
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.9779505729675293,
+      "learning_rate": 1.8736294718376045e-05,
+      "loss": 2.5275,
+      "step": 10180
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.001068592071533,
+      "learning_rate": 1.870496836037842e-05,
+      "loss": 2.5245,
+      "step": 10190
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.435389995574951,
+      "learning_rate": 1.8673642002380805e-05,
+      "loss": 2.4051,
+      "step": 10200
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.331353664398193,
+      "learning_rate": 1.8642315644383185e-05,
+      "loss": 2.4511,
+      "step": 10210
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.031054496765137,
+      "learning_rate": 1.8610989286385565e-05,
+      "loss": 2.4381,
+      "step": 10220
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.013234615325928,
+      "learning_rate": 1.8579662928387948e-05,
+      "loss": 2.4026,
+      "step": 10230
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.422504901885986,
+      "learning_rate": 1.8548336570390325e-05,
+      "loss": 2.5228,
+      "step": 10240
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.050678730010986,
+      "learning_rate": 1.8517010212392708e-05,
+      "loss": 2.3733,
+      "step": 10250
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.986072301864624,
+      "learning_rate": 1.848568385439509e-05,
+      "loss": 2.4298,
+      "step": 10260
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.056432247161865,
+      "learning_rate": 1.8454357496397472e-05,
+      "loss": 2.3867,
+      "step": 10270
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.96640682220459,
+      "learning_rate": 1.8423031138399852e-05,
+      "loss": 2.3661,
+      "step": 10280
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.190069198608398,
+      "learning_rate": 1.8391704780402232e-05,
+      "loss": 2.3079,
+      "step": 10290
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.016003131866455,
+      "learning_rate": 1.8360378422404612e-05,
+      "loss": 2.3572,
+      "step": 10300
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.608148097991943,
+      "learning_rate": 1.8329052064406992e-05,
+      "loss": 2.4708,
+      "step": 10310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.779339790344238,
+      "learning_rate": 1.8297725706409375e-05,
+      "loss": 2.4984,
+      "step": 10320
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.29642391204834,
+      "learning_rate": 1.8266399348411755e-05,
+      "loss": 2.4389,
+      "step": 10330
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.987062692642212,
+      "learning_rate": 1.8235072990414135e-05,
+      "loss": 2.5029,
+      "step": 10340
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.135943412780762,
+      "learning_rate": 1.8203746632416515e-05,
+      "loss": 2.392,
+      "step": 10350
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.60791540145874,
+      "learning_rate": 1.8172420274418895e-05,
+      "loss": 2.4186,
+      "step": 10360
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.076789855957031,
+      "learning_rate": 1.814109391642128e-05,
+      "loss": 2.449,
+      "step": 10370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.475488185882568,
+      "learning_rate": 1.810976755842366e-05,
+      "loss": 2.4329,
+      "step": 10380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.552157878875732,
+      "learning_rate": 1.807844120042604e-05,
+      "loss": 2.4157,
+      "step": 10390
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.7248082160949707,
+      "learning_rate": 1.804711484242842e-05,
+      "loss": 2.3301,
+      "step": 10400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.07235050201416,
+      "learning_rate": 1.8015788484430802e-05,
+      "loss": 2.4807,
+      "step": 10410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.504077911376953,
+      "learning_rate": 1.7984462126433182e-05,
+      "loss": 2.371,
+      "step": 10420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.039147853851318,
+      "learning_rate": 1.7953135768435562e-05,
+      "loss": 2.4404,
+      "step": 10430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.518662452697754,
+      "learning_rate": 1.7921809410437942e-05,
+      "loss": 2.3984,
+      "step": 10440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.8982303142547607,
+      "learning_rate": 1.7890483052440322e-05,
+      "loss": 2.4115,
+      "step": 10450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.247034549713135,
+      "learning_rate": 1.7859156694442706e-05,
+      "loss": 2.4675,
+      "step": 10460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.116578102111816,
+      "learning_rate": 1.7827830336445086e-05,
+      "loss": 2.3421,
+      "step": 10470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.851707696914673,
+      "learning_rate": 1.7796503978447466e-05,
+      "loss": 2.4673,
+      "step": 10480
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.440132141113281,
+      "learning_rate": 1.7765177620449846e-05,
+      "loss": 2.2985,
+      "step": 10490
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.295937538146973,
+      "learning_rate": 1.773385126245223e-05,
+      "loss": 2.3367,
+      "step": 10500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.4065423011779785,
+      "learning_rate": 1.770252490445461e-05,
+      "loss": 2.4196,
+      "step": 10510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.549808502197266,
+      "learning_rate": 1.767119854645699e-05,
+      "loss": 2.4587,
+      "step": 10520
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.632240056991577,
+      "learning_rate": 1.763987218845937e-05,
+      "loss": 2.3999,
+      "step": 10530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.433363914489746,
+      "learning_rate": 1.760854583046175e-05,
+      "loss": 2.354,
+      "step": 10540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.315452575683594,
+      "learning_rate": 1.7577219472464133e-05,
+      "loss": 2.4668,
+      "step": 10550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.167226791381836,
+      "learning_rate": 1.7545893114466513e-05,
+      "loss": 2.3308,
+      "step": 10560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.258768081665039,
+      "learning_rate": 1.7514566756468896e-05,
+      "loss": 2.3729,
+      "step": 10570
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.293150901794434,
+      "learning_rate": 1.7483240398471273e-05,
+      "loss": 2.3616,
+      "step": 10580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.8591740131378174,
+      "learning_rate": 1.7451914040473656e-05,
+      "loss": 2.3923,
+      "step": 10590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.937702178955078,
+      "learning_rate": 1.7420587682476036e-05,
+      "loss": 2.5261,
+      "step": 10600
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.620584964752197,
+      "learning_rate": 1.7389261324478416e-05,
+      "loss": 2.3608,
+      "step": 10610
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.822798013687134,
+      "learning_rate": 1.73579349664808e-05,
+      "loss": 2.3959,
+      "step": 10620
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.2800164222717285,
+      "learning_rate": 1.7326608608483177e-05,
+      "loss": 2.5203,
+      "step": 10630
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.448559284210205,
+      "learning_rate": 1.729528225048556e-05,
+      "loss": 2.4006,
+      "step": 10640
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.095371723175049,
+      "learning_rate": 1.726395589248794e-05,
+      "loss": 2.4556,
+      "step": 10650
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.553467750549316,
+      "learning_rate": 1.723262953449032e-05,
+      "loss": 2.3279,
+      "step": 10660
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.532577991485596,
+      "learning_rate": 1.7201303176492703e-05,
+      "loss": 2.4905,
+      "step": 10670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.202332496643066,
+      "learning_rate": 1.716997681849508e-05,
+      "loss": 2.3601,
+      "step": 10680
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.557814121246338,
+      "learning_rate": 1.7138650460497463e-05,
+      "loss": 2.4099,
+      "step": 10690
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.7724504470825195,
+      "learning_rate": 1.7107324102499844e-05,
+      "loss": 2.4612,
+      "step": 10700
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.376640319824219,
+      "learning_rate": 1.7075997744502227e-05,
+      "loss": 2.3714,
+      "step": 10710
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.8816592693328857,
+      "learning_rate": 1.7044671386504607e-05,
+      "loss": 2.383,
+      "step": 10720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.241684913635254,
+      "learning_rate": 1.7013345028506987e-05,
+      "loss": 2.3658,
+      "step": 10730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.058201313018799,
+      "learning_rate": 1.6982018670509367e-05,
+      "loss": 2.5489,
+      "step": 10740
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.433839797973633,
+      "learning_rate": 1.6950692312511747e-05,
+      "loss": 2.3589,
+      "step": 10750
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.016234397888184,
+      "learning_rate": 1.691936595451413e-05,
+      "loss": 2.3439,
+      "step": 10760
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.006540298461914,
+      "learning_rate": 1.688803959651651e-05,
+      "loss": 2.2962,
+      "step": 10770
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.735622406005859,
+      "learning_rate": 1.685671323851889e-05,
+      "loss": 2.3675,
+      "step": 10780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.5275349617004395,
+      "learning_rate": 1.682538688052127e-05,
+      "loss": 2.2321,
+      "step": 10790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.588350296020508,
+      "learning_rate": 1.6794060522523654e-05,
+      "loss": 2.3445,
+      "step": 10800
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.578807830810547,
+      "learning_rate": 1.6762734164526034e-05,
+      "loss": 2.2022,
+      "step": 10810
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.442037582397461,
+      "learning_rate": 1.6731407806528414e-05,
+      "loss": 2.1944,
+      "step": 10820
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.329452991485596,
+      "learning_rate": 1.6700081448530794e-05,
+      "loss": 2.2167,
+      "step": 10830
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.357962608337402,
+      "learning_rate": 1.6668755090533174e-05,
+      "loss": 2.3061,
+      "step": 10840
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.071457862854004,
+      "learning_rate": 1.6637428732535558e-05,
+      "loss": 2.1617,
+      "step": 10850
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.430005073547363,
+      "learning_rate": 1.6606102374537938e-05,
+      "loss": 2.2533,
+      "step": 10860
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.683684825897217,
+      "learning_rate": 1.6574776016540318e-05,
+      "loss": 2.1427,
+      "step": 10870
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.369844436645508,
+      "learning_rate": 1.6543449658542698e-05,
+      "loss": 2.2505,
+      "step": 10880
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.296201229095459,
+      "learning_rate": 1.6512123300545078e-05,
+      "loss": 2.2814,
+      "step": 10890
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.347317218780518,
+      "learning_rate": 1.648079694254746e-05,
+      "loss": 2.2257,
+      "step": 10900
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.395873069763184,
+      "learning_rate": 1.644947058454984e-05,
+      "loss": 2.3565,
+      "step": 10910
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.883645057678223,
+      "learning_rate": 1.641814422655222e-05,
+      "loss": 2.2033,
+      "step": 10920
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.499231338500977,
+      "learning_rate": 1.63868178685546e-05,
+      "loss": 2.1711,
+      "step": 10930
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.7956953048706055,
+      "learning_rate": 1.6355491510556985e-05,
+      "loss": 2.141,
+      "step": 10940
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.847490310668945,
+      "learning_rate": 1.6324165152559365e-05,
+      "loss": 2.3622,
+      "step": 10950
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.346663951873779,
+      "learning_rate": 1.6292838794561745e-05,
+      "loss": 2.2275,
+      "step": 10960
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.818025588989258,
+      "learning_rate": 1.6261512436564125e-05,
+      "loss": 2.2773,
+      "step": 10970
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.506927967071533,
+      "learning_rate": 1.6230186078566505e-05,
+      "loss": 2.2266,
+      "step": 10980
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.6733622550964355,
+      "learning_rate": 1.6198859720568888e-05,
+      "loss": 2.2539,
+      "step": 10990
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.735600471496582,
+      "learning_rate": 1.6167533362571268e-05,
+      "loss": 2.2552,
+      "step": 11000
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.016603946685791,
+      "learning_rate": 1.613620700457365e-05,
+      "loss": 2.252,
+      "step": 11010
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.684229850769043,
+      "learning_rate": 1.6104880646576028e-05,
+      "loss": 2.2791,
+      "step": 11020
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.259072780609131,
+      "learning_rate": 1.607355428857841e-05,
+      "loss": 2.2869,
+      "step": 11030
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.278809547424316,
+      "learning_rate": 1.604222793058079e-05,
+      "loss": 2.3341,
+      "step": 11040
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.164812088012695,
+      "learning_rate": 1.601090157258317e-05,
+      "loss": 2.2361,
+      "step": 11050
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.670204162597656,
+      "learning_rate": 1.5979575214585555e-05,
+      "loss": 2.2125,
+      "step": 11060
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.744375228881836,
+      "learning_rate": 1.5948248856587932e-05,
+      "loss": 2.1317,
+      "step": 11070
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.084680080413818,
+      "learning_rate": 1.5916922498590315e-05,
+      "loss": 2.1588,
+      "step": 11080
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.742985725402832,
+      "learning_rate": 1.5885596140592695e-05,
+      "loss": 2.2572,
+      "step": 11090
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.529914379119873,
+      "learning_rate": 1.585426978259508e-05,
+      "loss": 2.2142,
+      "step": 11100
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.66815185546875,
+      "learning_rate": 1.582294342459746e-05,
+      "loss": 2.2082,
+      "step": 11110
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.613138675689697,
+      "learning_rate": 1.579161706659984e-05,
+      "loss": 2.196,
+      "step": 11120
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.429781913757324,
+      "learning_rate": 1.576029070860222e-05,
+      "loss": 2.3071,
+      "step": 11130
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.774982929229736,
+      "learning_rate": 1.57289643506046e-05,
+      "loss": 2.2391,
+      "step": 11140
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.520074367523193,
+      "learning_rate": 1.5697637992606982e-05,
+      "loss": 2.3476,
+      "step": 11150
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.699487209320068,
+      "learning_rate": 1.566631163460936e-05,
+      "loss": 2.2987,
+      "step": 11160
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.9261250495910645,
+      "learning_rate": 1.5634985276611742e-05,
+      "loss": 2.2451,
+      "step": 11170
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.742948055267334,
+      "learning_rate": 1.5603658918614122e-05,
+      "loss": 2.2451,
+      "step": 11180
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.750619411468506,
+      "learning_rate": 1.5572332560616502e-05,
+      "loss": 2.236,
+      "step": 11190
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.201536178588867,
+      "learning_rate": 1.5541006202618886e-05,
+      "loss": 2.3287,
+      "step": 11200
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.444113731384277,
+      "learning_rate": 1.5509679844621262e-05,
+      "loss": 2.1338,
+      "step": 11210
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.299532890319824,
+      "learning_rate": 1.5478353486623646e-05,
+      "loss": 2.2989,
+      "step": 11220
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.273916721343994,
+      "learning_rate": 1.5447027128626026e-05,
+      "loss": 2.2331,
+      "step": 11230
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.77936315536499,
+      "learning_rate": 1.541570077062841e-05,
+      "loss": 2.2602,
+      "step": 11240
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.429702281951904,
+      "learning_rate": 1.538437441263079e-05,
+      "loss": 2.3563,
+      "step": 11250
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.403205871582031,
+      "learning_rate": 1.535304805463317e-05,
+      "loss": 2.2532,
+      "step": 11260
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.882217884063721,
+      "learning_rate": 1.532172169663555e-05,
+      "loss": 2.3309,
+      "step": 11270
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.7882232666015625,
+      "learning_rate": 1.529039533863793e-05,
+      "loss": 2.1901,
+      "step": 11280
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.28114652633667,
+      "learning_rate": 1.5259068980640313e-05,
+      "loss": 2.2094,
+      "step": 11290
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.084491729736328,
+      "learning_rate": 1.5227742622642693e-05,
+      "loss": 2.1454,
+      "step": 11300
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.040525913238525,
+      "learning_rate": 1.5196416264645073e-05,
+      "loss": 2.1998,
+      "step": 11310
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.516851425170898,
+      "learning_rate": 1.5165089906647453e-05,
+      "loss": 2.1945,
+      "step": 11320
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.610063076019287,
+      "learning_rate": 1.5133763548649835e-05,
+      "loss": 2.2231,
+      "step": 11330
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.8471598625183105,
+      "learning_rate": 1.5102437190652216e-05,
+      "loss": 2.345,
+      "step": 11340
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.626014709472656,
+      "learning_rate": 1.5071110832654598e-05,
+      "loss": 2.285,
+      "step": 11350
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.075950622558594,
+      "learning_rate": 1.5039784474656976e-05,
+      "loss": 2.1407,
+      "step": 11360
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.894886016845703,
+      "learning_rate": 1.5008458116659358e-05,
+      "loss": 2.181,
+      "step": 11370
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.974881172180176,
+      "learning_rate": 1.4977131758661738e-05,
+      "loss": 2.3302,
+      "step": 11380
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.369019508361816,
+      "learning_rate": 1.494580540066412e-05,
+      "loss": 2.1679,
+      "step": 11390
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.445016860961914,
+      "learning_rate": 1.4914479042666502e-05,
+      "loss": 2.2998,
+      "step": 11400
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.570034503936768,
+      "learning_rate": 1.488315268466888e-05,
+      "loss": 2.2268,
+      "step": 11410
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.77443265914917,
+      "learning_rate": 1.4851826326671262e-05,
+      "loss": 2.2892,
+      "step": 11420
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.94240665435791,
+      "learning_rate": 1.4820499968673643e-05,
+      "loss": 2.1972,
+      "step": 11430
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.679396152496338,
+      "learning_rate": 1.4789173610676025e-05,
+      "loss": 2.2014,
+      "step": 11440
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.996157169342041,
+      "learning_rate": 1.4757847252678405e-05,
+      "loss": 2.1538,
+      "step": 11450
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.940011501312256,
+      "learning_rate": 1.4726520894680783e-05,
+      "loss": 2.2863,
+      "step": 11460
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.545008659362793,
+      "learning_rate": 1.4695194536683165e-05,
+      "loss": 2.2417,
+      "step": 11470
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.11240291595459,
+      "learning_rate": 1.4663868178685547e-05,
+      "loss": 2.2372,
+      "step": 11480
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.07687520980835,
+      "learning_rate": 1.4632541820687929e-05,
+      "loss": 2.2018,
+      "step": 11490
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.014510631561279,
+      "learning_rate": 1.460121546269031e-05,
+      "loss": 2.1412,
+      "step": 11500
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.126745223999023,
+      "learning_rate": 1.4569889104692689e-05,
+      "loss": 2.3769,
+      "step": 11510
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.206657409667969,
+      "learning_rate": 1.453856274669507e-05,
+      "loss": 2.2612,
+      "step": 11520
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.919262886047363,
+      "learning_rate": 1.450723638869745e-05,
+      "loss": 2.4109,
+      "step": 11530
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.906914710998535,
+      "learning_rate": 1.4475910030699832e-05,
+      "loss": 2.1306,
+      "step": 11540
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.850189685821533,
+      "learning_rate": 1.444458367270221e-05,
+      "loss": 2.3328,
+      "step": 11550
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.900529861450195,
+      "learning_rate": 1.4413257314704592e-05,
+      "loss": 2.1918,
+      "step": 11560
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.0364990234375,
+      "learning_rate": 1.4381930956706974e-05,
+      "loss": 2.3654,
+      "step": 11570
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.807575225830078,
+      "learning_rate": 1.4350604598709356e-05,
+      "loss": 2.3238,
+      "step": 11580
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.882495403289795,
+      "learning_rate": 1.4319278240711737e-05,
+      "loss": 2.3126,
+      "step": 11590
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.3398118019104,
+      "learning_rate": 1.4287951882714116e-05,
+      "loss": 2.3554,
+      "step": 11600
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.0508599281311035,
+      "learning_rate": 1.4256625524716496e-05,
+      "loss": 2.1841,
+      "step": 11610
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.476123809814453,
+      "learning_rate": 1.4225299166718877e-05,
+      "loss": 2.1732,
+      "step": 11620
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.1260085105896,
+      "learning_rate": 1.4193972808721259e-05,
+      "loss": 2.3234,
+      "step": 11630
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.232734680175781,
+      "learning_rate": 1.4162646450723641e-05,
+      "loss": 2.3772,
+      "step": 11640
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.351359844207764,
+      "learning_rate": 1.413132009272602e-05,
+      "loss": 2.1584,
+      "step": 11650
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.6979146003723145,
+      "learning_rate": 1.4099993734728401e-05,
+      "loss": 2.2411,
+      "step": 11660
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.614090919494629,
+      "learning_rate": 1.4068667376730783e-05,
+      "loss": 2.246,
+      "step": 11670
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.733226299285889,
+      "learning_rate": 1.4037341018733163e-05,
+      "loss": 2.1933,
+      "step": 11680
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.780957221984863,
+      "learning_rate": 1.4006014660735544e-05,
+      "loss": 2.1053,
+      "step": 11690
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.091309547424316,
+      "learning_rate": 1.3974688302737923e-05,
+      "loss": 2.328,
+      "step": 11700
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.841605186462402,
+      "learning_rate": 1.3943361944740304e-05,
+      "loss": 2.3678,
+      "step": 11710
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.011669635772705,
+      "learning_rate": 1.3912035586742686e-05,
+      "loss": 2.1706,
+      "step": 11720
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.761959552764893,
+      "learning_rate": 1.3880709228745068e-05,
+      "loss": 2.3205,
+      "step": 11730
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.310210704803467,
+      "learning_rate": 1.3849382870747448e-05,
+      "loss": 2.1968,
+      "step": 11740
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.438429832458496,
+      "learning_rate": 1.3818056512749828e-05,
+      "loss": 2.3005,
+      "step": 11750
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.3258771896362305,
+      "learning_rate": 1.3786730154752208e-05,
+      "loss": 2.3267,
+      "step": 11760
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.707846641540527,
+      "learning_rate": 1.375540379675459e-05,
+      "loss": 2.1814,
+      "step": 11770
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.156281471252441,
+      "learning_rate": 1.3724077438756971e-05,
+      "loss": 2.297,
+      "step": 11780
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.802826404571533,
+      "learning_rate": 1.3692751080759353e-05,
+      "loss": 2.2419,
+      "step": 11790
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.247189044952393,
+      "learning_rate": 1.3661424722761732e-05,
+      "loss": 2.2894,
+      "step": 11800
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.583905220031738,
+      "learning_rate": 1.3630098364764113e-05,
+      "loss": 2.0925,
+      "step": 11810
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.071223258972168,
+      "learning_rate": 1.3598772006766495e-05,
+      "loss": 2.3123,
+      "step": 11820
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.113489627838135,
+      "learning_rate": 1.3567445648768875e-05,
+      "loss": 2.2259,
+      "step": 11830
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.095146179199219,
+      "learning_rate": 1.3536119290771257e-05,
+      "loss": 2.3375,
+      "step": 11840
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.298186302185059,
+      "learning_rate": 1.3504792932773635e-05,
+      "loss": 2.1844,
+      "step": 11850
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.170765399932861,
+      "learning_rate": 1.3473466574776017e-05,
+      "loss": 2.138,
+      "step": 11860
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.8907246589660645,
+      "learning_rate": 1.3442140216778399e-05,
+      "loss": 2.3352,
+      "step": 11870
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.390583515167236,
+      "learning_rate": 1.341081385878078e-05,
+      "loss": 2.2547,
+      "step": 11880
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.003589630126953,
+      "learning_rate": 1.3379487500783159e-05,
+      "loss": 2.2183,
+      "step": 11890
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.281210899353027,
+      "learning_rate": 1.334816114278554e-05,
+      "loss": 2.3186,
+      "step": 11900
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.8342461585998535,
+      "learning_rate": 1.331683478478792e-05,
+      "loss": 2.2049,
+      "step": 11910
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.283405780792236,
+      "learning_rate": 1.3285508426790302e-05,
+      "loss": 2.3101,
+      "step": 11920
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.727782726287842,
+      "learning_rate": 1.3254182068792684e-05,
+      "loss": 2.1859,
+      "step": 11930
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.096275329589844,
+      "learning_rate": 1.3222855710795062e-05,
+      "loss": 2.2318,
+      "step": 11940
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.112122058868408,
+      "learning_rate": 1.3191529352797444e-05,
+      "loss": 2.2752,
+      "step": 11950
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.975126266479492,
+      "learning_rate": 1.3160202994799826e-05,
+      "loss": 2.2632,
+      "step": 11960
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.01999044418335,
+      "learning_rate": 1.3128876636802207e-05,
+      "loss": 2.2912,
+      "step": 11970
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.252249240875244,
+      "learning_rate": 1.3097550278804587e-05,
+      "loss": 2.1791,
+      "step": 11980
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.292074680328369,
+      "learning_rate": 1.3066223920806966e-05,
+      "loss": 2.3141,
+      "step": 11990
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.5826921463012695,
+      "learning_rate": 1.3034897562809347e-05,
+      "loss": 2.3514,
+      "step": 12000
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.747256278991699,
+      "learning_rate": 1.3003571204811729e-05,
+      "loss": 2.2335,
+      "step": 12010
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.5181779861450195,
+      "learning_rate": 1.297224484681411e-05,
+      "loss": 2.3244,
+      "step": 12020
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.566039085388184,
+      "learning_rate": 1.2940918488816493e-05,
+      "loss": 2.1537,
+      "step": 12030
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.884830474853516,
+      "learning_rate": 1.2909592130818871e-05,
+      "loss": 2.2966,
+      "step": 12040
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.978501319885254,
+      "learning_rate": 1.2878265772821253e-05,
+      "loss": 2.2039,
+      "step": 12050
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.066789627075195,
+      "learning_rate": 1.2846939414823633e-05,
+      "loss": 2.3125,
+      "step": 12060
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.925536632537842,
+      "learning_rate": 1.2815613056826014e-05,
+      "loss": 2.2723,
+      "step": 12070
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.975927352905273,
+      "learning_rate": 1.2784286698828396e-05,
+      "loss": 2.3694,
+      "step": 12080
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.9011335372924805,
+      "learning_rate": 1.2752960340830774e-05,
+      "loss": 2.1171,
+      "step": 12090
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.841973304748535,
+      "learning_rate": 1.2721633982833156e-05,
+      "loss": 2.3144,
+      "step": 12100
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.277383804321289,
+      "learning_rate": 1.2690307624835538e-05,
+      "loss": 2.2665,
+      "step": 12110
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.377317428588867,
+      "learning_rate": 1.265898126683792e-05,
+      "loss": 2.1452,
+      "step": 12120
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.7158074378967285,
+      "learning_rate": 1.26276549088403e-05,
+      "loss": 2.1965,
+      "step": 12130
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.411214351654053,
+      "learning_rate": 1.2596328550842678e-05,
+      "loss": 2.2097,
+      "step": 12140
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.748459339141846,
+      "learning_rate": 1.256500219284506e-05,
+      "loss": 2.303,
+      "step": 12150
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.961619853973389,
+      "learning_rate": 1.2533675834847441e-05,
+      "loss": 2.2179,
+      "step": 12160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.2982001304626465,
+      "learning_rate": 1.2502349476849823e-05,
+      "loss": 2.2276,
+      "step": 12170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.200771808624268,
+      "learning_rate": 1.2471023118852203e-05,
+      "loss": 2.2831,
+      "step": 12180
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.932064533233643,
+      "learning_rate": 1.2439696760854585e-05,
+      "loss": 2.2783,
+      "step": 12190
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.722069263458252,
+      "learning_rate": 1.2408370402856965e-05,
+      "loss": 2.2223,
+      "step": 12200
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.084744453430176,
+      "learning_rate": 1.2377044044859345e-05,
+      "loss": 2.3635,
+      "step": 12210
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.1085357666015625,
+      "learning_rate": 1.2345717686861725e-05,
+      "loss": 2.2819,
+      "step": 12220
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.719175338745117,
+      "learning_rate": 1.2314391328864107e-05,
+      "loss": 2.3169,
+      "step": 12230
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.763276100158691,
+      "learning_rate": 1.2283064970866488e-05,
+      "loss": 2.3517,
+      "step": 12240
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.76812744140625,
+      "learning_rate": 1.2251738612868868e-05,
+      "loss": 2.3041,
+      "step": 12250
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.920300006866455,
+      "learning_rate": 1.222041225487125e-05,
+      "loss": 2.2292,
+      "step": 12260
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.785378932952881,
+      "learning_rate": 1.218908589687363e-05,
+      "loss": 2.3244,
+      "step": 12270
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.033608913421631,
+      "learning_rate": 1.215775953887601e-05,
+      "loss": 2.2818,
+      "step": 12280
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.9999775886535645,
+      "learning_rate": 1.212643318087839e-05,
+      "loss": 2.1768,
+      "step": 12290
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.320832252502441,
+      "learning_rate": 1.2095106822880772e-05,
+      "loss": 2.2857,
+      "step": 12300
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.435542106628418,
+      "learning_rate": 1.2063780464883154e-05,
+      "loss": 2.2643,
+      "step": 12310
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.2425336837768555,
+      "learning_rate": 1.2032454106885534e-05,
+      "loss": 2.2234,
+      "step": 12320
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.80018949508667,
+      "learning_rate": 1.2001127748887915e-05,
+      "loss": 2.372,
+      "step": 12330
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.5564045906066895,
+      "learning_rate": 1.1969801390890295e-05,
+      "loss": 2.239,
+      "step": 12340
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.771701335906982,
+      "learning_rate": 1.1938475032892677e-05,
+      "loss": 2.3006,
+      "step": 12350
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.302359104156494,
+      "learning_rate": 1.1907148674895057e-05,
+      "loss": 2.3294,
+      "step": 12360
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.32417106628418,
+      "learning_rate": 1.1875822316897437e-05,
+      "loss": 2.1954,
+      "step": 12370
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.4360456466674805,
+      "learning_rate": 1.1844495958899819e-05,
+      "loss": 2.2502,
+      "step": 12380
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.270229816436768,
+      "learning_rate": 1.1813169600902199e-05,
+      "loss": 2.2166,
+      "step": 12390
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.748353004455566,
+      "learning_rate": 1.178184324290458e-05,
+      "loss": 2.171,
+      "step": 12400
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.804108619689941,
+      "learning_rate": 1.1750516884906962e-05,
+      "loss": 2.3239,
+      "step": 12410
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.346190452575684,
+      "learning_rate": 1.1719190526909342e-05,
+      "loss": 2.1899,
+      "step": 12420
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.100567817687988,
+      "learning_rate": 1.1687864168911723e-05,
+      "loss": 2.3432,
+      "step": 12430
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.597306728363037,
+      "learning_rate": 1.1656537810914103e-05,
+      "loss": 2.2212,
+      "step": 12440
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.855820655822754,
+      "learning_rate": 1.1625211452916484e-05,
+      "loss": 2.2674,
+      "step": 12450
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.747831344604492,
+      "learning_rate": 1.1593885094918864e-05,
+      "loss": 2.298,
+      "step": 12460
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.8098673820495605,
+      "learning_rate": 1.1562558736921246e-05,
+      "loss": 2.2259,
+      "step": 12470
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.240844249725342,
+      "learning_rate": 1.1531232378923628e-05,
+      "loss": 2.2065,
+      "step": 12480
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.017106056213379,
+      "learning_rate": 1.1499906020926008e-05,
+      "loss": 2.1382,
+      "step": 12490
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.257959365844727,
+      "learning_rate": 1.146857966292839e-05,
+      "loss": 2.3176,
+      "step": 12500
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.328380107879639,
+      "learning_rate": 1.1437253304930768e-05,
+      "loss": 2.2397,
+      "step": 12510
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.653416156768799,
+      "learning_rate": 1.140592694693315e-05,
+      "loss": 2.1877,
+      "step": 12520
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.2487592697143555,
+      "learning_rate": 1.1374600588935531e-05,
+      "loss": 2.2622,
+      "step": 12530
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.720800399780273,
+      "learning_rate": 1.1343274230937911e-05,
+      "loss": 2.1956,
+      "step": 12540
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.962193489074707,
+      "learning_rate": 1.1311947872940293e-05,
+      "loss": 2.1048,
+      "step": 12550
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.106245994567871,
+      "learning_rate": 1.1280621514942673e-05,
+      "loss": 2.2074,
+      "step": 12560
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.294451713562012,
+      "learning_rate": 1.1249295156945055e-05,
+      "loss": 2.3578,
+      "step": 12570
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.8226847648620605,
+      "learning_rate": 1.1217968798947435e-05,
+      "loss": 2.2542,
+      "step": 12580
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.236142158508301,
+      "learning_rate": 1.1186642440949815e-05,
+      "loss": 2.1268,
+      "step": 12590
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.430288314819336,
+      "learning_rate": 1.1155316082952197e-05,
+      "loss": 2.1378,
+      "step": 12600
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.315874099731445,
+      "learning_rate": 1.1123989724954577e-05,
+      "loss": 2.1105,
+      "step": 12610
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.95720911026001,
+      "learning_rate": 1.1092663366956958e-05,
+      "loss": 2.2494,
+      "step": 12620
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.5179667472839355,
+      "learning_rate": 1.106133700895934e-05,
+      "loss": 2.2131,
+      "step": 12630
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.903148174285889,
+      "learning_rate": 1.103001065096172e-05,
+      "loss": 2.1935,
+      "step": 12640
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.403289794921875,
+      "learning_rate": 1.09986842929641e-05,
+      "loss": 2.1902,
+      "step": 12650
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.413578987121582,
+      "learning_rate": 1.096735793496648e-05,
+      "loss": 2.3298,
+      "step": 12660
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.452661991119385,
+      "learning_rate": 1.0936031576968862e-05,
+      "loss": 2.2673,
+      "step": 12670
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.752969264984131,
+      "learning_rate": 1.0904705218971242e-05,
+      "loss": 2.1987,
+      "step": 12680
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.89524507522583,
+      "learning_rate": 1.0873378860973624e-05,
+      "loss": 2.2216,
+      "step": 12690
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.80157995223999,
+      "learning_rate": 1.0842052502976005e-05,
+      "loss": 2.2769,
+      "step": 12700
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.679454326629639,
+      "learning_rate": 1.0810726144978385e-05,
+      "loss": 2.2917,
+      "step": 12710
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.759067058563232,
+      "learning_rate": 1.0779399786980767e-05,
+      "loss": 2.2885,
+      "step": 12720
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.863231658935547,
+      "learning_rate": 1.0748073428983147e-05,
+      "loss": 2.2205,
+      "step": 12730
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.668412208557129,
+      "learning_rate": 1.0716747070985527e-05,
+      "loss": 2.184,
+      "step": 12740
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.075872421264648,
+      "learning_rate": 1.0685420712987909e-05,
+      "loss": 2.2732,
+      "step": 12750
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.573457717895508,
+      "learning_rate": 1.0654094354990289e-05,
+      "loss": 2.1865,
+      "step": 12760
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.100333213806152,
+      "learning_rate": 1.062276799699267e-05,
+      "loss": 2.1851,
+      "step": 12770
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.814767360687256,
+      "learning_rate": 1.059144163899505e-05,
+      "loss": 2.2246,
+      "step": 12780
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.054508686065674,
+      "learning_rate": 1.0560115280997432e-05,
+      "loss": 2.3433,
+      "step": 12790
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.447021007537842,
+      "learning_rate": 1.0528788922999812e-05,
+      "loss": 2.2906,
+      "step": 12800
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.723243713378906,
+      "learning_rate": 1.0497462565002192e-05,
+      "loss": 2.1602,
+      "step": 12810
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.797729969024658,
+      "learning_rate": 1.0466136207004574e-05,
+      "loss": 2.213,
+      "step": 12820
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.9997735023498535,
+      "learning_rate": 1.0434809849006954e-05,
+      "loss": 2.2403,
+      "step": 12830
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.520061492919922,
+      "learning_rate": 1.0403483491009336e-05,
+      "loss": 2.2191,
+      "step": 12840
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.711124420166016,
+      "learning_rate": 1.0372157133011716e-05,
+      "loss": 2.2004,
+      "step": 12850
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.090340614318848,
+      "learning_rate": 1.0340830775014098e-05,
+      "loss": 2.2455,
+      "step": 12860
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.26514196395874,
+      "learning_rate": 1.030950441701648e-05,
+      "loss": 2.1776,
+      "step": 12870
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.399611473083496,
+      "learning_rate": 1.027817805901886e-05,
+      "loss": 2.171,
+      "step": 12880
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.001101016998291,
+      "learning_rate": 1.024685170102124e-05,
+      "loss": 2.2674,
+      "step": 12890
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.122973442077637,
+      "learning_rate": 1.021552534302362e-05,
+      "loss": 2.1839,
+      "step": 12900
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.108747959136963,
+      "learning_rate": 1.0184198985026001e-05,
+      "loss": 2.2297,
+      "step": 12910
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.05750036239624,
+      "learning_rate": 1.0152872627028383e-05,
+      "loss": 2.3335,
+      "step": 12920
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.032607555389404,
+      "learning_rate": 1.0121546269030763e-05,
+      "loss": 2.3326,
+      "step": 12930
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.749469757080078,
+      "learning_rate": 1.0090219911033145e-05,
+      "loss": 2.2638,
+      "step": 12940
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.402693748474121,
+      "learning_rate": 1.0058893553035525e-05,
+      "loss": 2.2938,
+      "step": 12950
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.914982318878174,
+      "learning_rate": 1.0027567195037905e-05,
+      "loss": 2.2562,
+      "step": 12960
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.372003078460693,
+      "learning_rate": 9.996240837040286e-06,
+      "loss": 2.1931,
+      "step": 12970
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.706940174102783,
+      "learning_rate": 9.964914479042667e-06,
+      "loss": 2.2326,
+      "step": 12980
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.209222316741943,
+      "learning_rate": 9.933588121045048e-06,
+      "loss": 2.144,
+      "step": 12990
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.133508205413818,
+      "learning_rate": 9.902261763047428e-06,
+      "loss": 2.3475,
+      "step": 13000
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.845363616943359,
+      "learning_rate": 9.87093540504981e-06,
+      "loss": 2.2409,
+      "step": 13010
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.653754234313965,
+      "learning_rate": 9.83960904705219e-06,
+      "loss": 2.3025,
+      "step": 13020
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.139420509338379,
+      "learning_rate": 9.808282689054572e-06,
+      "loss": 2.2286,
+      "step": 13030
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.228572845458984,
+      "learning_rate": 9.776956331056952e-06,
+      "loss": 2.4346,
+      "step": 13040
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.307019233703613,
+      "learning_rate": 9.745629973059332e-06,
+      "loss": 2.2245,
+      "step": 13050
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.422297954559326,
+      "learning_rate": 9.714303615061714e-06,
+      "loss": 2.1797,
+      "step": 13060
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.005941390991211,
+      "learning_rate": 9.682977257064094e-06,
+      "loss": 2.2727,
+      "step": 13070
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.4763689041137695,
+      "learning_rate": 9.651650899066475e-06,
+      "loss": 2.1005,
+      "step": 13080
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.4908576011657715,
+      "learning_rate": 9.620324541068857e-06,
+      "loss": 2.162,
+      "step": 13090
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.450204849243164,
+      "learning_rate": 9.588998183071237e-06,
+      "loss": 2.2548,
+      "step": 13100
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.938765048980713,
+      "learning_rate": 9.557671825073617e-06,
+      "loss": 2.1718,
+      "step": 13110
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.608719825744629,
+      "learning_rate": 9.526345467075997e-06,
+      "loss": 2.2955,
+      "step": 13120
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.325197696685791,
+      "learning_rate": 9.495019109078379e-06,
+      "loss": 2.167,
+      "step": 13130
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.6594767570495605,
+      "learning_rate": 9.46369275108076e-06,
+      "loss": 2.2667,
+      "step": 13140
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.628880977630615,
+      "learning_rate": 9.43236639308314e-06,
+      "loss": 2.162,
+      "step": 13150
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.200460433959961,
+      "learning_rate": 9.401040035085522e-06,
+      "loss": 2.2934,
+      "step": 13160
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.961476802825928,
+      "learning_rate": 9.369713677087902e-06,
+      "loss": 2.3302,
+      "step": 13170
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.322465896606445,
+      "learning_rate": 9.338387319090282e-06,
+      "loss": 2.1746,
+      "step": 13180
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.763502597808838,
+      "learning_rate": 9.307060961092662e-06,
+      "loss": 2.2987,
+      "step": 13190
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.322182655334473,
+      "learning_rate": 9.275734603095044e-06,
+      "loss": 2.326,
+      "step": 13200
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.088197708129883,
+      "learning_rate": 9.244408245097426e-06,
+      "loss": 2.1486,
+      "step": 13210
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.4456787109375,
+      "learning_rate": 9.213081887099806e-06,
+      "loss": 2.2645,
+      "step": 13220
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.5985260009765625,
+      "learning_rate": 9.181755529102188e-06,
+      "loss": 2.3224,
+      "step": 13230
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.466270923614502,
+      "learning_rate": 9.150429171104568e-06,
+      "loss": 2.3032,
+      "step": 13240
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.2544732093811035,
+      "learning_rate": 9.11910281310695e-06,
+      "loss": 2.1505,
+      "step": 13250
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.213293552398682,
+      "learning_rate": 9.08777645510933e-06,
+      "loss": 2.2228,
+      "step": 13260
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.244242191314697,
+      "learning_rate": 9.05645009711171e-06,
+      "loss": 2.28,
+      "step": 13270
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.124194145202637,
+      "learning_rate": 9.025123739114091e-06,
+      "loss": 2.3366,
+      "step": 13280
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.310089111328125,
+      "learning_rate": 8.993797381116471e-06,
+      "loss": 2.2799,
+      "step": 13290
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.166079998016357,
+      "learning_rate": 8.962471023118853e-06,
+      "loss": 2.2094,
+      "step": 13300
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.7795329093933105,
+      "learning_rate": 8.931144665121235e-06,
+      "loss": 2.2734,
+      "step": 13310
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.956170558929443,
+      "learning_rate": 8.899818307123615e-06,
+      "loss": 2.1813,
+      "step": 13320
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.007333278656006,
+      "learning_rate": 8.868491949125995e-06,
+      "loss": 2.2364,
+      "step": 13330
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.24399995803833,
+      "learning_rate": 8.837165591128375e-06,
+      "loss": 2.2734,
+      "step": 13340
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.286114692687988,
+      "learning_rate": 8.805839233130756e-06,
+      "loss": 2.2163,
+      "step": 13350
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.692686557769775,
+      "learning_rate": 8.774512875133136e-06,
+      "loss": 2.201,
+      "step": 13360
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.9664483070373535,
+      "learning_rate": 8.743186517135518e-06,
+      "loss": 2.2684,
+      "step": 13370
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.770281791687012,
+      "learning_rate": 8.7118601591379e-06,
+      "loss": 2.2949,
+      "step": 13380
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.857741832733154,
+      "learning_rate": 8.68053380114028e-06,
+      "loss": 2.2934,
+      "step": 13390
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.026750564575195,
+      "learning_rate": 8.649207443142662e-06,
+      "loss": 2.2011,
+      "step": 13400
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.276968955993652,
+      "learning_rate": 8.617881085145042e-06,
+      "loss": 2.2554,
+      "step": 13410
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.570988655090332,
+      "learning_rate": 8.586554727147422e-06,
+      "loss": 2.1372,
+      "step": 13420
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.223836421966553,
+      "learning_rate": 8.555228369149803e-06,
+      "loss": 2.3167,
+      "step": 13430
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.428679943084717,
+      "learning_rate": 8.523902011152183e-06,
+      "loss": 2.2213,
+      "step": 13440
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.6929402351379395,
+      "learning_rate": 8.492575653154565e-06,
+      "loss": 2.3149,
+      "step": 13450
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.652947902679443,
+      "learning_rate": 8.461249295156945e-06,
+      "loss": 2.3203,
+      "step": 13460
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.486786842346191,
+      "learning_rate": 8.429922937159327e-06,
+      "loss": 2.2709,
+      "step": 13470
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.114050388336182,
+      "learning_rate": 8.398596579161707e-06,
+      "loss": 2.3454,
+      "step": 13480
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.2786173820495605,
+      "learning_rate": 8.367270221164087e-06,
+      "loss": 2.2613,
+      "step": 13490
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.374701023101807,
+      "learning_rate": 8.335943863166469e-06,
+      "loss": 2.1595,
+      "step": 13500
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.902259349822998,
+      "learning_rate": 8.304617505168849e-06,
+      "loss": 2.3199,
+      "step": 13510
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.804900646209717,
+      "learning_rate": 8.27329114717123e-06,
+      "loss": 2.2834,
+      "step": 13520
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.4029645919799805,
+      "learning_rate": 8.24196478917361e-06,
+      "loss": 2.3464,
+      "step": 13530
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.877260684967041,
+      "learning_rate": 8.210638431175992e-06,
+      "loss": 2.3503,
+      "step": 13540
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.730605602264404,
+      "learning_rate": 8.179312073178374e-06,
+      "loss": 2.344,
+      "step": 13550
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.457608699798584,
+      "learning_rate": 8.147985715180754e-06,
+      "loss": 2.1494,
+      "step": 13560
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.026783466339111,
+      "learning_rate": 8.116659357183134e-06,
+      "loss": 2.3395,
+      "step": 13570
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.831313133239746,
+      "learning_rate": 8.085332999185514e-06,
+      "loss": 2.1817,
+      "step": 13580
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.2338337898254395,
+      "learning_rate": 8.054006641187896e-06,
+      "loss": 2.2464,
+      "step": 13590
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.4740519523620605,
+      "learning_rate": 8.022680283190277e-06,
+      "loss": 2.1953,
+      "step": 13600
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.073474884033203,
+      "learning_rate": 7.991353925192658e-06,
+      "loss": 2.1982,
+      "step": 13610
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.817521572113037,
+      "learning_rate": 7.96002756719504e-06,
+      "loss": 2.2207,
+      "step": 13620
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.21115779876709,
+      "learning_rate": 7.92870120919742e-06,
+      "loss": 2.1872,
+      "step": 13630
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.0212931632995605,
+      "learning_rate": 7.8973748511998e-06,
+      "loss": 2.2251,
+      "step": 13640
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.8617939949035645,
+      "learning_rate": 7.866048493202181e-06,
+      "loss": 2.3814,
+      "step": 13650
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.384790420532227,
+      "learning_rate": 7.834722135204561e-06,
+      "loss": 2.2883,
+      "step": 13660
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.272597789764404,
+      "learning_rate": 7.803395777206943e-06,
+      "loss": 2.2633,
+      "step": 13670
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.676297187805176,
+      "learning_rate": 7.772069419209323e-06,
+      "loss": 2.2315,
+      "step": 13680
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.78164005279541,
+      "learning_rate": 7.740743061211705e-06,
+      "loss": 2.2953,
+      "step": 13690
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.913782119750977,
+      "learning_rate": 7.709416703214085e-06,
+      "loss": 2.3002,
+      "step": 13700
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.161614894866943,
+      "learning_rate": 7.678090345216465e-06,
+      "loss": 2.2274,
+      "step": 13710
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.582902908325195,
+      "learning_rate": 7.646763987218846e-06,
+      "loss": 2.3175,
+      "step": 13720
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.013784885406494,
+      "learning_rate": 7.615437629221226e-06,
+      "loss": 2.0668,
+      "step": 13730
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.877525806427002,
+      "learning_rate": 7.584111271223608e-06,
+      "loss": 2.2382,
+      "step": 13740
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.978527545928955,
+      "learning_rate": 7.552784913225988e-06,
+      "loss": 2.3498,
+      "step": 13750
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.610579967498779,
+      "learning_rate": 7.52145855522837e-06,
+      "loss": 2.1325,
+      "step": 13760
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.137538433074951,
+      "learning_rate": 7.490132197230751e-06,
+      "loss": 2.2978,
+      "step": 13770
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.440046787261963,
+      "learning_rate": 7.458805839233131e-06,
+      "loss": 2.2039,
+      "step": 13780
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.702132225036621,
+      "learning_rate": 7.4274794812355125e-06,
+      "loss": 2.2908,
+      "step": 13790
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.557126998901367,
+      "learning_rate": 7.3961531232378925e-06,
+      "loss": 2.2488,
+      "step": 13800
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.8882222175598145,
+      "learning_rate": 7.364826765240273e-06,
+      "loss": 2.2059,
+      "step": 13810
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.6189093589782715,
+      "learning_rate": 7.333500407242655e-06,
+      "loss": 2.1869,
+      "step": 13820
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.6229400634765625,
+      "learning_rate": 7.302174049245035e-06,
+      "loss": 2.2872,
+      "step": 13830
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.87022590637207,
+      "learning_rate": 7.270847691247416e-06,
+      "loss": 2.2808,
+      "step": 13840
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.438892841339111,
+      "learning_rate": 7.239521333249796e-06,
+      "loss": 2.3188,
+      "step": 13850
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.652724266052246,
+      "learning_rate": 7.208194975252178e-06,
+      "loss": 2.2262,
+      "step": 13860
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.241449356079102,
+      "learning_rate": 7.176868617254558e-06,
+      "loss": 2.1773,
+      "step": 13870
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.65318489074707,
+      "learning_rate": 7.145542259256939e-06,
+      "loss": 2.2564,
+      "step": 13880
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.262519836425781,
+      "learning_rate": 7.11421590125932e-06,
+      "loss": 2.3014,
+      "step": 13890
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.8109025955200195,
+      "learning_rate": 7.0828895432617e-06,
+      "loss": 2.1783,
+      "step": 13900
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.613912582397461,
+      "learning_rate": 7.051563185264082e-06,
+      "loss": 2.3027,
+      "step": 13910
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.484175682067871,
+      "learning_rate": 7.020236827266462e-06,
+      "loss": 2.2926,
+      "step": 13920
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.171184539794922,
+      "learning_rate": 6.988910469268843e-06,
+      "loss": 2.179,
+      "step": 13930
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.096653938293457,
+      "learning_rate": 6.957584111271225e-06,
+      "loss": 2.1388,
+      "step": 13940
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.9391069412231445,
+      "learning_rate": 6.926257753273605e-06,
+      "loss": 2.1041,
+      "step": 13950
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.7677459716796875,
+      "learning_rate": 6.894931395275986e-06,
+      "loss": 2.2272,
+      "step": 13960
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.2970147132873535,
+      "learning_rate": 6.863605037278366e-06,
+      "loss": 2.1962,
+      "step": 13970
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.634983539581299,
+      "learning_rate": 6.8322786792807474e-06,
+      "loss": 2.307,
+      "step": 13980
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.822270393371582,
+      "learning_rate": 6.800952321283128e-06,
+      "loss": 2.0941,
+      "step": 13990
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.998510360717773,
+      "learning_rate": 6.769625963285508e-06,
+      "loss": 2.2717,
+      "step": 14000
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.330641269683838,
+      "learning_rate": 6.73829960528789e-06,
+      "loss": 2.2546,
+      "step": 14010
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.172261714935303,
+      "learning_rate": 6.70697324729027e-06,
+      "loss": 2.2181,
+      "step": 14020
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.40900993347168,
+      "learning_rate": 6.675646889292651e-06,
+      "loss": 2.3187,
+      "step": 14030
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.111061096191406,
+      "learning_rate": 6.644320531295031e-06,
+      "loss": 2.1499,
+      "step": 14040
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.557284355163574,
+      "learning_rate": 6.612994173297413e-06,
+      "loss": 2.3055,
+      "step": 14050
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.03141450881958,
+      "learning_rate": 6.5816678152997944e-06,
+      "loss": 2.2504,
+      "step": 14060
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.888171195983887,
+      "learning_rate": 6.550341457302174e-06,
+      "loss": 2.0959,
+      "step": 14070
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.806394100189209,
+      "learning_rate": 6.519015099304555e-06,
+      "loss": 2.2443,
+      "step": 14080
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.673512935638428,
+      "learning_rate": 6.487688741306935e-06,
+      "loss": 2.2093,
+      "step": 14090
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.377676486968994,
+      "learning_rate": 6.456362383309317e-06,
+      "loss": 2.2957,
+      "step": 14100
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.2512359619140625,
+      "learning_rate": 6.425036025311698e-06,
+      "loss": 2.1859,
+      "step": 14110
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.803738594055176,
+      "learning_rate": 6.393709667314078e-06,
+      "loss": 2.2449,
+      "step": 14120
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.135036945343018,
+      "learning_rate": 6.36238330931646e-06,
+      "loss": 2.2144,
+      "step": 14130
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.972255706787109,
+      "learning_rate": 6.33105695131884e-06,
+      "loss": 2.1646,
+      "step": 14140
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.952043533325195,
+      "learning_rate": 6.299730593321221e-06,
+      "loss": 2.2383,
+      "step": 14150
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.106963157653809,
+      "learning_rate": 6.268404235323602e-06,
+      "loss": 2.2635,
+      "step": 14160
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.5615129470825195,
+      "learning_rate": 6.237077877325982e-06,
+      "loss": 2.3088,
+      "step": 14170
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.770695209503174,
+      "learning_rate": 6.205751519328363e-06,
+      "loss": 2.2693,
+      "step": 14180
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.178565502166748,
+      "learning_rate": 6.174425161330744e-06,
+      "loss": 2.3658,
+      "step": 14190
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.672639846801758,
+      "learning_rate": 6.143098803333125e-06,
+      "loss": 2.206,
+      "step": 14200
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.218943119049072,
+      "learning_rate": 6.111772445335506e-06,
+      "loss": 2.156,
+      "step": 14210
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.135859489440918,
+      "learning_rate": 6.080446087337886e-06,
+      "loss": 2.1683,
+      "step": 14220
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.083263397216797,
+      "learning_rate": 6.049119729340267e-06,
+      "loss": 2.1906,
+      "step": 14230
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.857705593109131,
+      "learning_rate": 6.017793371342648e-06,
+      "loss": 2.1322,
+      "step": 14240
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.133806228637695,
+      "learning_rate": 5.986467013345029e-06,
+      "loss": 2.1363,
+      "step": 14250
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.316963195800781,
+      "learning_rate": 5.9551406553474094e-06,
+      "loss": 2.1679,
+      "step": 14260
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.094076156616211,
+      "learning_rate": 5.92381429734979e-06,
+      "loss": 2.2329,
+      "step": 14270
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.19959020614624,
+      "learning_rate": 5.892487939352171e-06,
+      "loss": 2.2172,
+      "step": 14280
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.317006587982178,
+      "learning_rate": 5.861161581354552e-06,
+      "loss": 2.3854,
+      "step": 14290
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.754021644592285,
+      "learning_rate": 5.829835223356933e-06,
+      "loss": 2.1969,
+      "step": 14300
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.336154937744141,
+      "learning_rate": 5.798508865359314e-06,
+      "loss": 2.3282,
+      "step": 14310
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.85452938079834,
+      "learning_rate": 5.767182507361695e-06,
+      "loss": 2.35,
+      "step": 14320
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.174147605895996,
+      "learning_rate": 5.735856149364075e-06,
+      "loss": 2.1795,
+      "step": 14330
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.9505228996276855,
+      "learning_rate": 5.704529791366456e-06,
+      "loss": 2.0833,
+      "step": 14340
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.462031364440918,
+      "learning_rate": 5.6732034333688365e-06,
+      "loss": 2.2265,
+      "step": 14350
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.277505397796631,
+      "learning_rate": 5.641877075371218e-06,
+      "loss": 2.3298,
+      "step": 14360
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.46626615524292,
+      "learning_rate": 5.610550717373598e-06,
+      "loss": 2.2778,
+      "step": 14370
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.012364387512207,
+      "learning_rate": 5.579224359375979e-06,
+      "loss": 2.2084,
+      "step": 14380
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.736668586730957,
+      "learning_rate": 5.54789800137836e-06,
+      "loss": 2.2649,
+      "step": 14390
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.878015995025635,
+      "learning_rate": 5.516571643380741e-06,
+      "loss": 2.199,
+      "step": 14400
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.050587177276611,
+      "learning_rate": 5.485245285383122e-06,
+      "loss": 2.2238,
+      "step": 14410
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.755616664886475,
+      "learning_rate": 5.453918927385503e-06,
+      "loss": 2.1919,
+      "step": 14420
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.753908157348633,
+      "learning_rate": 5.4225925693878835e-06,
+      "loss": 2.2535,
+      "step": 14430
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.909355640411377,
+      "learning_rate": 5.391266211390264e-06,
+      "loss": 2.1856,
+      "step": 14440
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.696736812591553,
+      "learning_rate": 5.359939853392644e-06,
+      "loss": 2.1645,
+      "step": 14450
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.7360520362854,
+      "learning_rate": 5.328613495395025e-06,
+      "loss": 2.2176,
+      "step": 14460
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.935107707977295,
+      "learning_rate": 5.297287137397407e-06,
+      "loss": 2.3128,
+      "step": 14470
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.986132621765137,
+      "learning_rate": 5.265960779399787e-06,
+      "loss": 2.0601,
+      "step": 14480
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.014797210693359,
+      "learning_rate": 5.234634421402168e-06,
+      "loss": 2.1279,
+      "step": 14490
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.099488258361816,
+      "learning_rate": 5.203308063404549e-06,
+      "loss": 2.157,
+      "step": 14500
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.824801445007324,
+      "learning_rate": 5.17198170540693e-06,
+      "loss": 2.2426,
+      "step": 14510
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.455517768859863,
+      "learning_rate": 5.14065534740931e-06,
+      "loss": 2.3237,
+      "step": 14520
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.147612571716309,
+      "learning_rate": 5.1093289894116914e-06,
+      "loss": 2.157,
+      "step": 14530
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.479028224945068,
+      "learning_rate": 5.078002631414072e-06,
+      "loss": 2.2894,
+      "step": 14540
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.406881809234619,
+      "learning_rate": 5.046676273416453e-06,
+      "loss": 2.1629,
+      "step": 14550
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.455532073974609,
+      "learning_rate": 5.015349915418833e-06,
+      "loss": 2.0312,
+      "step": 14560
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.509167194366455,
+      "learning_rate": 4.984023557421214e-06,
+      "loss": 2.3442,
+      "step": 14570
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.4643330574035645,
+      "learning_rate": 4.952697199423596e-06,
+      "loss": 2.1014,
+      "step": 14580
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.990197658538818,
+      "learning_rate": 4.921370841425976e-06,
+      "loss": 2.2279,
+      "step": 14590
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.77827262878418,
+      "learning_rate": 4.890044483428357e-06,
+      "loss": 2.1901,
+      "step": 14600
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.1118083000183105,
+      "learning_rate": 4.858718125430738e-06,
+      "loss": 2.3313,
+      "step": 14610
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.493002891540527,
+      "learning_rate": 4.8273917674331185e-06,
+      "loss": 2.2621,
+      "step": 14620
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.995043754577637,
+      "learning_rate": 4.796065409435499e-06,
+      "loss": 2.0892,
+      "step": 14630
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.6254119873046875,
+      "learning_rate": 4.76473905143788e-06,
+      "loss": 2.1689,
+      "step": 14640
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.596071243286133,
+      "learning_rate": 4.733412693440261e-06,
+      "loss": 2.2291,
+      "step": 14650
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.493969917297363,
+      "learning_rate": 4.702086335442642e-06,
+      "loss": 2.0971,
+      "step": 14660
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.211254119873047,
+      "learning_rate": 4.670759977445022e-06,
+      "loss": 2.3204,
+      "step": 14670
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.054335594177246,
+      "learning_rate": 4.639433619447403e-06,
+      "loss": 2.2843,
+      "step": 14680
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.881555557250977,
+      "learning_rate": 4.608107261449784e-06,
+      "loss": 2.212,
+      "step": 14690
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.238095760345459,
+      "learning_rate": 4.5767809034521655e-06,
+      "loss": 2.2209,
+      "step": 14700
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.044524192810059,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 2.3081,
+      "step": 14710
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.570669651031494,
+      "learning_rate": 4.514128187456926e-06,
+      "loss": 2.0992,
+      "step": 14720
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.995473861694336,
+      "learning_rate": 4.482801829459307e-06,
+      "loss": 2.3147,
+      "step": 14730
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.873288631439209,
+      "learning_rate": 4.451475471461688e-06,
+      "loss": 2.2422,
+      "step": 14740
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.107931613922119,
+      "learning_rate": 4.420149113464069e-06,
+      "loss": 2.2725,
+      "step": 14750
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.3016533851623535,
+      "learning_rate": 4.38882275546645e-06,
+      "loss": 2.1693,
+      "step": 14760
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.348349571228027,
+      "learning_rate": 4.357496397468831e-06,
+      "loss": 2.2864,
+      "step": 14770
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.596982002258301,
+      "learning_rate": 4.326170039471212e-06,
+      "loss": 2.2719,
+      "step": 14780
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.691074371337891,
+      "learning_rate": 4.294843681473592e-06,
+      "loss": 2.3477,
+      "step": 14790
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.885664463043213,
+      "learning_rate": 4.2635173234759726e-06,
+      "loss": 2.079,
+      "step": 14800
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.616001129150391,
+      "learning_rate": 4.232190965478354e-06,
+      "loss": 2.1574,
+      "step": 14810
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.573058605194092,
+      "learning_rate": 4.200864607480734e-06,
+      "loss": 2.2285,
+      "step": 14820
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.251656532287598,
+      "learning_rate": 4.169538249483115e-06,
+      "loss": 2.2224,
+      "step": 14830
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.339073657989502,
+      "learning_rate": 4.138211891485496e-06,
+      "loss": 2.3068,
+      "step": 14840
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.788910865783691,
+      "learning_rate": 4.106885533487877e-06,
+      "loss": 2.2796,
+      "step": 14850
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.806098461151123,
+      "learning_rate": 4.075559175490257e-06,
+      "loss": 2.2256,
+      "step": 14860
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.479308128356934,
+      "learning_rate": 4.044232817492639e-06,
+      "loss": 2.2174,
+      "step": 14870
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.946530342102051,
+      "learning_rate": 4.01290645949502e-06,
+      "loss": 2.2028,
+      "step": 14880
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.5223822593688965,
+      "learning_rate": 3.9815801014974005e-06,
+      "loss": 2.1426,
+      "step": 14890
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.463576316833496,
+      "learning_rate": 3.9502537434997805e-06,
+      "loss": 2.1199,
+      "step": 14900
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.110754013061523,
+      "learning_rate": 3.918927385502161e-06,
+      "loss": 2.1892,
+      "step": 14910
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.9819512367248535,
+      "learning_rate": 3.887601027504543e-06,
+      "loss": 2.1054,
+      "step": 14920
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.723372459411621,
+      "learning_rate": 3.856274669506923e-06,
+      "loss": 2.2199,
+      "step": 14930
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.602808475494385,
+      "learning_rate": 3.824948311509304e-06,
+      "loss": 2.1164,
+      "step": 14940
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.397231101989746,
+      "learning_rate": 3.793621953511685e-06,
+      "loss": 2.3285,
+      "step": 14950
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.740065097808838,
+      "learning_rate": 3.7622955955140657e-06,
+      "loss": 2.2537,
+      "step": 14960
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.875923156738281,
+      "learning_rate": 3.730969237516446e-06,
+      "loss": 2.2326,
+      "step": 14970
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.001502990722656,
+      "learning_rate": 3.6996428795188275e-06,
+      "loss": 2.17,
+      "step": 14980
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.4369282722473145,
+      "learning_rate": 3.6683165215212084e-06,
+      "loss": 2.1858,
+      "step": 14990
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.982198715209961,
+      "learning_rate": 3.636990163523589e-06,
+      "loss": 2.1899,
+      "step": 15000
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.815186977386475,
+      "learning_rate": 3.6056638055259697e-06,
+      "loss": 2.1004,
+      "step": 15010
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.177938938140869,
+      "learning_rate": 3.57433744752835e-06,
+      "loss": 2.2429,
+      "step": 15020
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.127625942230225,
+      "learning_rate": 3.543011089530731e-06,
+      "loss": 2.1429,
+      "step": 15030
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.616495132446289,
+      "learning_rate": 3.5116847315331123e-06,
+      "loss": 2.2848,
+      "step": 15040
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.018442153930664,
+      "learning_rate": 3.4803583735354932e-06,
+      "loss": 2.2801,
+      "step": 15050
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.80119514465332,
+      "learning_rate": 3.4490320155378737e-06,
+      "loss": 2.2253,
+      "step": 15060
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.015815734863281,
+      "learning_rate": 3.4177056575402545e-06,
+      "loss": 2.2014,
+      "step": 15070
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.846337795257568,
+      "learning_rate": 3.386379299542635e-06,
+      "loss": 2.2732,
+      "step": 15080
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.304662227630615,
+      "learning_rate": 3.3550529415450163e-06,
+      "loss": 2.3228,
+      "step": 15090
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.77760124206543,
+      "learning_rate": 3.323726583547397e-06,
+      "loss": 2.2546,
+      "step": 15100
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.972886562347412,
+      "learning_rate": 3.2924002255497776e-06,
+      "loss": 2.2863,
+      "step": 15110
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.197108745574951,
+      "learning_rate": 3.2610738675521585e-06,
+      "loss": 2.314,
+      "step": 15120
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.3886799812316895,
+      "learning_rate": 3.229747509554539e-06,
+      "loss": 2.2824,
+      "step": 15130
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.588632583618164,
+      "learning_rate": 3.19842115155692e-06,
+      "loss": 2.1507,
+      "step": 15140
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.631302356719971,
+      "learning_rate": 3.167094793559301e-06,
+      "loss": 2.3639,
+      "step": 15150
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.998614311218262,
+      "learning_rate": 3.135768435561682e-06,
+      "loss": 2.2462,
+      "step": 15160
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.030011177062988,
+      "learning_rate": 3.1044420775640625e-06,
+      "loss": 2.3389,
+      "step": 15170
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.784780502319336,
+      "learning_rate": 3.0731157195664433e-06,
+      "loss": 2.2283,
+      "step": 15180
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.889308452606201,
+      "learning_rate": 3.0417893615688242e-06,
+      "loss": 2.1219,
+      "step": 15190
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.779045581817627,
+      "learning_rate": 3.010463003571205e-06,
+      "loss": 2.2276,
+      "step": 15200
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.917059898376465,
+      "learning_rate": 2.9791366455735856e-06,
+      "loss": 2.1144,
+      "step": 15210
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.68520450592041,
+      "learning_rate": 2.947810287575967e-06,
+      "loss": 2.4605,
+      "step": 15220
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.577776908874512,
+      "learning_rate": 2.9164839295783473e-06,
+      "loss": 2.2559,
+      "step": 15230
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.267956733703613,
+      "learning_rate": 2.885157571580728e-06,
+      "loss": 2.2218,
+      "step": 15240
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.275062561035156,
+      "learning_rate": 2.853831213583109e-06,
+      "loss": 2.2437,
+      "step": 15250
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.182082176208496,
+      "learning_rate": 2.82250485558549e-06,
+      "loss": 2.2602,
+      "step": 15260
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.157698631286621,
+      "learning_rate": 2.7911784975878704e-06,
+      "loss": 2.2095,
+      "step": 15270
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.279921054840088,
+      "learning_rate": 2.7598521395902513e-06,
+      "loss": 2.1501,
+      "step": 15280
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.391173362731934,
+      "learning_rate": 2.728525781592632e-06,
+      "loss": 2.1421,
+      "step": 15290
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.1854023933410645,
+      "learning_rate": 2.697199423595013e-06,
+      "loss": 2.3726,
+      "step": 15300
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.282604217529297,
+      "learning_rate": 2.665873065597394e-06,
+      "loss": 2.1056,
+      "step": 15310
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.295992851257324,
+      "learning_rate": 2.6345467075997744e-06,
+      "loss": 2.3241,
+      "step": 15320
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.822473049163818,
+      "learning_rate": 2.6032203496021557e-06,
+      "loss": 2.2288,
+      "step": 15330
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.813709259033203,
+      "learning_rate": 2.571893991604536e-06,
+      "loss": 2.2335,
+      "step": 15340
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.577134132385254,
+      "learning_rate": 2.540567633606917e-06,
+      "loss": 2.2846,
+      "step": 15350
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.539644241333008,
+      "learning_rate": 2.509241275609298e-06,
+      "loss": 2.1281,
+      "step": 15360
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.731322288513184,
+      "learning_rate": 2.4779149176116787e-06,
+      "loss": 2.2189,
+      "step": 15370
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.15498161315918,
+      "learning_rate": 2.446588559614059e-06,
+      "loss": 2.1157,
+      "step": 15380
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.277298450469971,
+      "learning_rate": 2.4152622016164405e-06,
+      "loss": 2.3078,
+      "step": 15390
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.953650951385498,
+      "learning_rate": 2.383935843618821e-06,
+      "loss": 2.2351,
+      "step": 15400
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.887104034423828,
+      "learning_rate": 2.352609485621202e-06,
+      "loss": 2.108,
+      "step": 15410
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.2219767570495605,
+      "learning_rate": 2.3212831276235827e-06,
+      "loss": 2.3245,
+      "step": 15420
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.635054588317871,
+      "learning_rate": 2.289956769625963e-06,
+      "loss": 2.2183,
+      "step": 15430
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.352700710296631,
+      "learning_rate": 2.258630411628344e-06,
+      "loss": 2.2135,
+      "step": 15440
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.863667011260986,
+      "learning_rate": 2.227304053630725e-06,
+      "loss": 2.1377,
+      "step": 15450
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.384456157684326,
+      "learning_rate": 2.1959776956331058e-06,
+      "loss": 2.2731,
+      "step": 15460
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.075336456298828,
+      "learning_rate": 2.1646513376354867e-06,
+      "loss": 2.12,
+      "step": 15470
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.152316570281982,
+      "learning_rate": 2.1333249796378675e-06,
+      "loss": 2.1699,
+      "step": 15480
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.9257917404174805,
+      "learning_rate": 2.101998621640248e-06,
+      "loss": 2.1559,
+      "step": 15490
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.926279544830322,
+      "learning_rate": 2.0706722636426293e-06,
+      "loss": 2.2984,
+      "step": 15500
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.468809604644775,
+      "learning_rate": 2.0393459056450097e-06,
+      "loss": 2.0854,
+      "step": 15510
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 14.988544464111328,
+      "learning_rate": 2.0080195476473906e-06,
+      "loss": 2.2182,
+      "step": 15520
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.264973163604736,
+      "learning_rate": 1.9766931896497715e-06,
+      "loss": 2.3079,
+      "step": 15530
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.474333763122559,
+      "learning_rate": 1.9453668316521524e-06,
+      "loss": 2.1352,
+      "step": 15540
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.057586193084717,
+      "learning_rate": 1.914040473654533e-06,
+      "loss": 2.1288,
+      "step": 15550
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.180942535400391,
+      "learning_rate": 1.882714115656914e-06,
+      "loss": 2.3115,
+      "step": 15560
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.402101516723633,
+      "learning_rate": 1.8513877576592946e-06,
+      "loss": 2.1244,
+      "step": 15570
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.428703308105469,
+      "learning_rate": 1.8200613996616752e-06,
+      "loss": 2.1784,
+      "step": 15580
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.959710597991943,
+      "learning_rate": 1.7887350416640563e-06,
+      "loss": 2.3097,
+      "step": 15590
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.586377143859863,
+      "learning_rate": 1.757408683666437e-06,
+      "loss": 2.1855,
+      "step": 15600
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.836302280426025,
+      "learning_rate": 1.7260823256688177e-06,
+      "loss": 2.2558,
+      "step": 15610
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.129261493682861,
+      "learning_rate": 1.6947559676711988e-06,
+      "loss": 2.1964,
+      "step": 15620
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.32321310043335,
+      "learning_rate": 1.6634296096735794e-06,
+      "loss": 2.1325,
+      "step": 15630
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.156566619873047,
+      "learning_rate": 1.6321032516759603e-06,
+      "loss": 2.25,
+      "step": 15640
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.7280755043029785,
+      "learning_rate": 1.600776893678341e-06,
+      "loss": 2.2077,
+      "step": 15650
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.000637531280518,
+      "learning_rate": 1.5694505356807218e-06,
+      "loss": 2.2315,
+      "step": 15660
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.943790435791016,
+      "learning_rate": 1.5381241776831025e-06,
+      "loss": 2.1645,
+      "step": 15670
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.3277668952941895,
+      "learning_rate": 1.5067978196854834e-06,
+      "loss": 2.2574,
+      "step": 15680
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.879887580871582,
+      "learning_rate": 1.4754714616878643e-06,
+      "loss": 2.1678,
+      "step": 15690
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.744934558868408,
+      "learning_rate": 1.444145103690245e-06,
+      "loss": 2.1535,
+      "step": 15700
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.511887550354004,
+      "learning_rate": 1.4128187456926258e-06,
+      "loss": 2.1764,
+      "step": 15710
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.7104811668396,
+      "learning_rate": 1.3814923876950067e-06,
+      "loss": 2.1938,
+      "step": 15720
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.853920936584473,
+      "learning_rate": 1.3501660296973873e-06,
+      "loss": 2.1526,
+      "step": 15730
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.676470756530762,
+      "learning_rate": 1.3188396716997682e-06,
+      "loss": 2.1985,
+      "step": 15740
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.885119915008545,
+      "learning_rate": 1.287513313702149e-06,
+      "loss": 2.2122,
+      "step": 15750
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.216524600982666,
+      "learning_rate": 1.25618695570453e-06,
+      "loss": 2.3254,
+      "step": 15760
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.545653343200684,
+      "learning_rate": 1.2248605977069106e-06,
+      "loss": 2.2905,
+      "step": 15770
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.003011703491211,
+      "learning_rate": 1.1935342397092915e-06,
+      "loss": 2.2472,
+      "step": 15780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.019803524017334,
+      "learning_rate": 1.1622078817116724e-06,
+      "loss": 2.3646,
+      "step": 15790
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.200807571411133,
+      "learning_rate": 1.130881523714053e-06,
+      "loss": 2.4219,
+      "step": 15800
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.575010299682617,
+      "learning_rate": 1.099555165716434e-06,
+      "loss": 2.3115,
+      "step": 15810
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.7481842041015625,
+      "learning_rate": 1.0682288077188146e-06,
+      "loss": 2.3336,
+      "step": 15820
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.135469436645508,
+      "learning_rate": 1.0369024497211955e-06,
+      "loss": 2.1925,
+      "step": 15830
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.409132957458496,
+      "learning_rate": 1.0055760917235761e-06,
+      "loss": 2.2156,
+      "step": 15840
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.531829357147217,
+      "learning_rate": 9.74249733725957e-07,
+      "loss": 2.2851,
+      "step": 15850
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.880288124084473,
+      "learning_rate": 9.429233757283379e-07,
+      "loss": 2.2134,
+      "step": 15860
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.735778331756592,
+      "learning_rate": 9.115970177307186e-07,
+      "loss": 2.1477,
+      "step": 15870
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.089371681213379,
+      "learning_rate": 8.802706597330994e-07,
+      "loss": 2.1704,
+      "step": 15880
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.147428512573242,
+      "learning_rate": 8.489443017354803e-07,
+      "loss": 2.2521,
+      "step": 15890
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.981338977813721,
+      "learning_rate": 8.176179437378612e-07,
+      "loss": 2.2188,
+      "step": 15900
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 6.076031684875488,
+      "learning_rate": 7.862915857402419e-07,
+      "loss": 2.1894,
+      "step": 15910
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.003175258636475,
+      "learning_rate": 7.549652277426227e-07,
+      "loss": 2.2015,
+      "step": 15920
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.125662803649902,
+      "learning_rate": 7.236388697450035e-07,
+      "loss": 2.1842,
+      "step": 15930
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.404399871826172,
+      "learning_rate": 6.923125117473843e-07,
+      "loss": 2.2261,
+      "step": 15940
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.939316272735596,
+      "learning_rate": 6.60986153749765e-07,
+      "loss": 2.2224,
+      "step": 15950
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 6.108341693878174,
+      "learning_rate": 6.296597957521459e-07,
+      "loss": 2.1957,
+      "step": 15960
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.7472333908081055,
+      "learning_rate": 5.983334377545267e-07,
+      "loss": 2.1746,
+      "step": 15970
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.0396599769592285,
+      "learning_rate": 5.670070797569075e-07,
+      "loss": 2.2503,
+      "step": 15980
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 6.954887866973877,
+      "learning_rate": 5.356807217592883e-07,
+      "loss": 2.1326,
+      "step": 15990
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.1434736251831055,
+      "learning_rate": 5.043543637616691e-07,
+      "loss": 2.1736,
+      "step": 16000
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.424656391143799,
+      "learning_rate": 4.7302800576404993e-07,
+      "loss": 2.297,
+      "step": 16010
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.53824520111084,
+      "learning_rate": 4.417016477664307e-07,
+      "loss": 2.1963,
+      "step": 16020
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.093719005584717,
+      "learning_rate": 4.103752897688115e-07,
+      "loss": 2.2149,
+      "step": 16030
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.9572062492370605,
+      "learning_rate": 3.790489317711923e-07,
+      "loss": 2.2097,
+      "step": 16040
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.130666255950928,
+      "learning_rate": 3.477225737735731e-07,
+      "loss": 2.2865,
+      "step": 16050
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.0656633377075195,
+      "learning_rate": 3.163962157759539e-07,
+      "loss": 2.2507,
+      "step": 16060
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.007623195648193,
+      "learning_rate": 2.850698577783347e-07,
+      "loss": 2.196,
+      "step": 16070
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.159407138824463,
+      "learning_rate": 2.537434997807155e-07,
+      "loss": 2.3414,
+      "step": 16080
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.766927719116211,
+      "learning_rate": 2.2241714178309631e-07,
+      "loss": 2.252,
+      "step": 16090
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.673045635223389,
+      "learning_rate": 1.910907837854771e-07,
+      "loss": 2.1263,
+      "step": 16100
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.0451507568359375,
+      "learning_rate": 1.597644257878579e-07,
+      "loss": 2.1164,
+      "step": 16110
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.848662853240967,
+      "learning_rate": 1.284380677902387e-07,
+      "loss": 2.1926,
+      "step": 16120
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.478034973144531,
+      "learning_rate": 9.711170979261952e-08,
+      "loss": 2.2998,
+      "step": 16130
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.586289405822754,
+      "learning_rate": 6.578535179500031e-08,
+      "loss": 2.2523,
+      "step": 16140
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.570580005645752,
+      "learning_rate": 3.445899379738112e-08,
+      "loss": 2.2318,
+      "step": 16150
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.062375068664551,
+      "learning_rate": 3.13263579976192e-09,
+      "loss": 2.234,
+      "step": 16160
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 16161,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 6.298475820377702e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}