diff --git "a/checkpoint-12000/trainer_state.json" "b/checkpoint-12000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-12000/trainer_state.json"
@@ -0,0 +1,8421 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8828560392870938,
+  "eval_steps": 500,
+  "global_step": 12000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007357133660725781,
+      "grad_norm": 223.60693359375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 20.4521,
+      "step": 10
+    },
+    {
+      "epoch": 0.0014714267321451562,
+      "grad_norm": 115.68708038330078,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 17.0724,
+      "step": 20
+    },
+    {
+      "epoch": 0.0022071400982177345,
+      "grad_norm": 9.73855972290039,
+      "learning_rate": 3e-06,
+      "loss": 11.1765,
+      "step": 30
+    },
+    {
+      "epoch": 0.0029428534642903124,
+      "grad_norm": 4.239206314086914,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 9.9334,
+      "step": 40
+    },
+    {
+      "epoch": 0.0036785668303628907,
+      "grad_norm": 5.939801216125488,
+      "learning_rate": 5e-06,
+      "loss": 9.6538,
+      "step": 50
+    },
+    {
+      "epoch": 0.004414280196435469,
+      "grad_norm": 7.083678245544434,
+      "learning_rate": 6e-06,
+      "loss": 9.4877,
+      "step": 60
+    },
+    {
+      "epoch": 0.0051499935625080465,
+      "grad_norm": 4.155125617980957,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 9.2998,
+      "step": 70
+    },
+    {
+      "epoch": 0.005885706928580625,
+      "grad_norm": 8.080842018127441,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 9.1653,
+      "step": 80
+    },
+    {
+      "epoch": 0.006621420294653203,
+      "grad_norm": 5.2444353103637695,
+      "learning_rate": 9e-06,
+      "loss": 9.0503,
+      "step": 90
+    },
+    {
+      "epoch": 0.0073571336607257815,
+      "grad_norm": 5.591372966766357,
+      "learning_rate": 1e-05,
+      "loss": 8.9182,
+      "step": 100
+    },
+    {
+      "epoch": 0.008092847026798359,
+      "grad_norm": 6.641251564025879,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 8.7308,
+      "step": 110
+    },
+    {
+      "epoch": 0.008828560392870938,
+      "grad_norm": 8.2617769241333,
+      "learning_rate": 1.2e-05,
+      "loss": 8.4667,
+      "step": 120
+    },
+    {
+      "epoch": 0.009564273758943516,
+      "grad_norm": 8.181135177612305,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 8.1982,
+      "step": 130
+    },
+    {
+      "epoch": 0.010299987125016093,
+      "grad_norm": 5.930114269256592,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 7.8692,
+      "step": 140
+    },
+    {
+      "epoch": 0.011035700491088672,
+      "grad_norm": 4.546728610992432,
+      "learning_rate": 1.5e-05,
+      "loss": 7.5773,
+      "step": 150
+    },
+    {
+      "epoch": 0.01177141385716125,
+      "grad_norm": 4.127258777618408,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 7.3816,
+      "step": 160
+    },
+    {
+      "epoch": 0.012507127223233829,
+      "grad_norm": 4.3951849937438965,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 7.2603,
+      "step": 170
+    },
+    {
+      "epoch": 0.013242840589306406,
+      "grad_norm": 2.085538625717163,
+      "learning_rate": 1.8e-05,
+      "loss": 7.16,
+      "step": 180
+    },
+    {
+      "epoch": 0.013978553955378984,
+      "grad_norm": 3.098052501678467,
+      "learning_rate": 1.9e-05,
+      "loss": 7.0236,
+      "step": 190
+    },
+    {
+      "epoch": 0.014714267321451563,
+      "grad_norm": 4.054460525512695,
+      "learning_rate": 2e-05,
+      "loss": 7.0169,
+      "step": 200
+    },
+    {
+      "epoch": 0.01544998068752414,
+      "grad_norm": 1.6470402479171753,
+      "learning_rate": 2.1e-05,
+      "loss": 6.9933,
+      "step": 210
+    },
+    {
+      "epoch": 0.016185694053596718,
+      "grad_norm": 1.3456400632858276,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 6.9151,
+      "step": 220
+    },
+    {
+      "epoch": 0.016921407419669295,
+      "grad_norm": 1.3457653522491455,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 6.9246,
+      "step": 230
+    },
+    {
+      "epoch": 0.017657120785741876,
+      "grad_norm": 2.5111804008483887,
+      "learning_rate": 2.4e-05,
+      "loss": 6.8615,
+      "step": 240
+    },
+    {
+      "epoch": 0.018392834151814454,
+      "grad_norm": 2.082517147064209,
+      "learning_rate": 2.5e-05,
+      "loss": 6.8109,
+      "step": 250
+    },
+    {
+      "epoch": 0.01912854751788703,
+      "grad_norm": 2.9603984355926514,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 6.7036,
+      "step": 260
+    },
+    {
+      "epoch": 0.01986426088395961,
+      "grad_norm": 1.7493945360183716,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 6.6403,
+      "step": 270
+    },
+    {
+      "epoch": 0.020599974250032186,
+      "grad_norm": 3.83150315284729,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 6.6216,
+      "step": 280
+    },
+    {
+      "epoch": 0.021335687616104767,
+      "grad_norm": 2.9999842643737793,
+      "learning_rate": 2.9e-05,
+      "loss": 6.6013,
+      "step": 290
+    },
+    {
+      "epoch": 0.022071400982177344,
+      "grad_norm": 2.799868583679199,
+      "learning_rate": 3e-05,
+      "loss": 6.4751,
+      "step": 300
+    },
+    {
+      "epoch": 0.022807114348249922,
+      "grad_norm": 3.548393487930298,
+      "learning_rate": 3.1e-05,
+      "loss": 6.5082,
+      "step": 310
+    },
+    {
+      "epoch": 0.0235428277143225,
+      "grad_norm": 3.523336410522461,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 6.4566,
+      "step": 320
+    },
+    {
+      "epoch": 0.024278541080395077,
+      "grad_norm": 2.0755739212036133,
+      "learning_rate": 3.3e-05,
+      "loss": 6.3776,
+      "step": 330
+    },
+    {
+      "epoch": 0.025014254446467658,
+      "grad_norm": 3.288764238357544,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 6.3559,
+      "step": 340
+    },
+    {
+      "epoch": 0.025749967812540235,
+      "grad_norm": 3.3932132720947266,
+      "learning_rate": 3.5e-05,
+      "loss": 6.3234,
+      "step": 350
+    },
+    {
+      "epoch": 0.026485681178612813,
+      "grad_norm": 2.7775251865386963,
+      "learning_rate": 3.6e-05,
+      "loss": 6.2551,
+      "step": 360
+    },
+    {
+      "epoch": 0.02722139454468539,
+      "grad_norm": 2.6773834228515625,
+      "learning_rate": 3.7e-05,
+      "loss": 6.206,
+      "step": 370
+    },
+    {
+      "epoch": 0.027957107910757967,
+      "grad_norm": 2.556666612625122,
+      "learning_rate": 3.8e-05,
+      "loss": 6.1396,
+      "step": 380
+    },
+    {
+      "epoch": 0.02869282127683055,
+      "grad_norm": 3.2724132537841797,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 6.106,
+      "step": 390
+    },
+    {
+      "epoch": 0.029428534642903126,
+      "grad_norm": 2.354327440261841,
+      "learning_rate": 4e-05,
+      "loss": 6.1258,
+      "step": 400
+    },
+    {
+      "epoch": 0.030164248008975703,
+      "grad_norm": 1.804317593574524,
+      "learning_rate": 4.1e-05,
+      "loss": 6.0516,
+      "step": 410
+    },
+    {
+      "epoch": 0.03089996137504828,
+      "grad_norm": 1.7134747505187988,
+      "learning_rate": 4.2e-05,
+      "loss": 5.9582,
+      "step": 420
+    },
+    {
+      "epoch": 0.03163567474112086,
+      "grad_norm": 1.4939428567886353,
+      "learning_rate": 4.3e-05,
+      "loss": 5.9328,
+      "step": 430
+    },
+    {
+      "epoch": 0.032371388107193436,
+      "grad_norm": 1.9662060737609863,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 5.8578,
+      "step": 440
+    },
+    {
+      "epoch": 0.03310710147326602,
+      "grad_norm": 1.748353123664856,
+      "learning_rate": 4.5e-05,
+      "loss": 5.7874,
+      "step": 450
+    },
+    {
+      "epoch": 0.03384281483933859,
+      "grad_norm": 2.1945955753326416,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 5.7787,
+      "step": 460
+    },
+    {
+      "epoch": 0.03457852820541117,
+      "grad_norm": 2.1684770584106445,
+      "learning_rate": 4.7e-05,
+      "loss": 5.7138,
+      "step": 470
+    },
+    {
+      "epoch": 0.03531424157148375,
+      "grad_norm": 1.728945255279541,
+      "learning_rate": 4.8e-05,
+      "loss": 5.6774,
+      "step": 480
+    },
+    {
+      "epoch": 0.036049954937556326,
+      "grad_norm": 1.8206098079681396,
+      "learning_rate": 4.9e-05,
+      "loss": 5.6214,
+      "step": 490
+    },
+    {
+      "epoch": 0.03678566830362891,
+      "grad_norm": 1.5937659740447998,
+      "learning_rate": 5e-05,
+      "loss": 5.5198,
+      "step": 500
+    },
+    {
+      "epoch": 0.03752138166970148,
+      "grad_norm": 1.617659568786621,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 5.4818,
+      "step": 510
+    },
+    {
+      "epoch": 0.03825709503577406,
+      "grad_norm": 1.7942918539047241,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 5.5156,
+      "step": 520
+    },
+    {
+      "epoch": 0.03899280840184664,
+      "grad_norm": 1.830945611000061,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 5.3836,
+      "step": 530
+    },
+    {
+      "epoch": 0.03972852176791922,
+      "grad_norm": 1.5295898914337158,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 5.3989,
+      "step": 540
+    },
+    {
+      "epoch": 0.0404642351339918,
+      "grad_norm": 1.527603268623352,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 5.3724,
+      "step": 550
+    },
+    {
+      "epoch": 0.04119994850006437,
+      "grad_norm": 1.6225807666778564,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 5.2788,
+      "step": 560
+    },
+    {
+      "epoch": 0.04193566186613695,
+      "grad_norm": 2.0076022148132324,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 5.2227,
+      "step": 570
+    },
+    {
+      "epoch": 0.042671375232209534,
+      "grad_norm": 1.5891458988189697,
+      "learning_rate": 5.8e-05,
+      "loss": 5.2265,
+      "step": 580
+    },
+    {
+      "epoch": 0.04340708859828211,
+      "grad_norm": 1.5814419984817505,
+      "learning_rate": 5.9e-05,
+      "loss": 5.1948,
+      "step": 590
+    },
+    {
+      "epoch": 0.04414280196435469,
+      "grad_norm": 1.7139537334442139,
+      "learning_rate": 6e-05,
+      "loss": 5.1036,
+      "step": 600
+    },
+    {
+      "epoch": 0.04487851533042726,
+      "grad_norm": 1.2224972248077393,
+      "learning_rate": 6.1e-05,
+      "loss": 5.1932,
+      "step": 610
+    },
+    {
+      "epoch": 0.045614228696499844,
+      "grad_norm": 1.6268278360366821,
+      "learning_rate": 6.2e-05,
+      "loss": 5.068,
+      "step": 620
+    },
+    {
+      "epoch": 0.046349942062572425,
+      "grad_norm": 1.4944841861724854,
+      "learning_rate": 6.3e-05,
+      "loss": 5.032,
+      "step": 630
+    },
+    {
+      "epoch": 0.047085655428645,
+      "grad_norm": 1.442824363708496,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 5.0077,
+      "step": 640
+    },
+    {
+      "epoch": 0.04782136879471758,
+      "grad_norm": 1.6031944751739502,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 4.9808,
+      "step": 650
+    },
+    {
+      "epoch": 0.04855708216079015,
+      "grad_norm": 1.6006144285202026,
+      "learning_rate": 6.6e-05,
+      "loss": 4.9239,
+      "step": 660
+    },
+    {
+      "epoch": 0.049292795526862734,
+      "grad_norm": 1.3961182832717896,
+      "learning_rate": 6.7e-05,
+      "loss": 4.9158,
+      "step": 670
+    },
+    {
+      "epoch": 0.050028508892935315,
+      "grad_norm": 1.4314249753952026,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 4.8126,
+      "step": 680
+    },
+    {
+      "epoch": 0.05076422225900789,
+      "grad_norm": 1.4509848356246948,
+      "learning_rate": 6.9e-05,
+      "loss": 4.8713,
+      "step": 690
+    },
+    {
+      "epoch": 0.05149993562508047,
+      "grad_norm": 1.5618196725845337,
+      "learning_rate": 7e-05,
+      "loss": 4.7814,
+      "step": 700
+    },
+    {
+      "epoch": 0.052235648991153044,
+      "grad_norm": 1.5423110723495483,
+      "learning_rate": 7.1e-05,
+      "loss": 4.7409,
+      "step": 710
+    },
+    {
+      "epoch": 0.052971362357225625,
+      "grad_norm": 1.2559527158737183,
+      "learning_rate": 7.2e-05,
+      "loss": 4.73,
+      "step": 720
+    },
+    {
+      "epoch": 0.053707075723298206,
+      "grad_norm": 1.1119533777236938,
+      "learning_rate": 7.3e-05,
+      "loss": 4.7055,
+      "step": 730
+    },
+    {
+      "epoch": 0.05444278908937078,
+      "grad_norm": 1.2400840520858765,
+      "learning_rate": 7.4e-05,
+      "loss": 4.6501,
+      "step": 740
+    },
+    {
+      "epoch": 0.05517850245544336,
+      "grad_norm": 1.2191784381866455,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 4.6239,
+      "step": 750
+    },
+    {
+      "epoch": 0.055914215821515935,
+      "grad_norm": 1.269572377204895,
+      "learning_rate": 7.6e-05,
+      "loss": 4.6498,
+      "step": 760
+    },
+    {
+      "epoch": 0.056649929187588516,
+      "grad_norm": 1.275608777999878,
+      "learning_rate": 7.7e-05,
+      "loss": 4.6326,
+      "step": 770
+    },
+    {
+      "epoch": 0.0573856425536611,
+      "grad_norm": 1.3013856410980225,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 4.5869,
+      "step": 780
+    },
+    {
+      "epoch": 0.05812135591973367,
+      "grad_norm": 1.37381112575531,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 4.5785,
+      "step": 790
+    },
+    {
+      "epoch": 0.05885706928580625,
+      "grad_norm": 1.2166635990142822,
+      "learning_rate": 8e-05,
+      "loss": 4.5126,
+      "step": 800
+    },
+    {
+      "epoch": 0.059592782651878826,
+      "grad_norm": 1.4644134044647217,
+      "learning_rate": 8.1e-05,
+      "loss": 4.5228,
+      "step": 810
+    },
+    {
+      "epoch": 0.06032849601795141,
+      "grad_norm": 1.206525206565857,
+      "learning_rate": 8.2e-05,
+      "loss": 4.4874,
+      "step": 820
+    },
+    {
+      "epoch": 0.06106420938402399,
+      "grad_norm": 1.1188849210739136,
+      "learning_rate": 8.3e-05,
+      "loss": 4.4619,
+      "step": 830
+    },
+    {
+      "epoch": 0.06179992275009656,
+      "grad_norm": 1.2260874509811401,
+      "learning_rate": 8.4e-05,
+      "loss": 4.4118,
+      "step": 840
+    },
+    {
+      "epoch": 0.06253563611616914,
+      "grad_norm": 1.19864022731781,
+      "learning_rate": 8.5e-05,
+      "loss": 4.3522,
+      "step": 850
+    },
+    {
+      "epoch": 0.06327134948224172,
+      "grad_norm": 1.1710054874420166,
+      "learning_rate": 8.6e-05,
+      "loss": 4.4048,
+      "step": 860
+    },
+    {
+      "epoch": 0.0640070628483143,
+      "grad_norm": 0.8916666507720947,
+      "learning_rate": 8.7e-05,
+      "loss": 4.3865,
+      "step": 870
+    },
+    {
+      "epoch": 0.06474277621438687,
+      "grad_norm": 1.135651707649231,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 4.3554,
+      "step": 880
+    },
+    {
+      "epoch": 0.06547848958045946,
+      "grad_norm": 0.9037818312644958,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 4.4055,
+      "step": 890
+    },
+    {
+      "epoch": 0.06621420294653203,
+      "grad_norm": 1.1230660676956177,
+      "learning_rate": 9e-05,
+      "loss": 4.3131,
+      "step": 900
+    },
+    {
+      "epoch": 0.06694991631260461,
+      "grad_norm": 1.149318814277649,
+      "learning_rate": 9.1e-05,
+      "loss": 4.2923,
+      "step": 910
+    },
+    {
+      "epoch": 0.06768562967867718,
+      "grad_norm": 1.1154922246932983,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 4.2774,
+      "step": 920
+    },
+    {
+      "epoch": 0.06842134304474977,
+      "grad_norm": 1.0399177074432373,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 4.2766,
+      "step": 930
+    },
+    {
+      "epoch": 0.06915705641082234,
+      "grad_norm": 1.0165122747421265,
+      "learning_rate": 9.4e-05,
+      "loss": 4.2479,
+      "step": 940
+    },
+    {
+      "epoch": 0.06989276977689492,
+      "grad_norm": 1.3305314779281616,
+      "learning_rate": 9.5e-05,
+      "loss": 4.2443,
+      "step": 950
+    },
+    {
+      "epoch": 0.0706284831429675,
+      "grad_norm": 0.8728047609329224,
+      "learning_rate": 9.6e-05,
+      "loss": 4.3267,
+      "step": 960
+    },
+    {
+      "epoch": 0.07136419650904008,
+      "grad_norm": 1.117966890335083,
+      "learning_rate": 9.7e-05,
+      "loss": 4.2435,
+      "step": 970
+    },
+    {
+      "epoch": 0.07209990987511265,
+      "grad_norm": 0.934165894985199,
+      "learning_rate": 9.8e-05,
+      "loss": 4.2125,
+      "step": 980
+    },
+    {
+      "epoch": 0.07283562324118524,
+      "grad_norm": 1.0023953914642334,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 4.1705,
+      "step": 990
+    },
+    {
+      "epoch": 0.07357133660725781,
+      "grad_norm": 0.9707255959510803,
+      "learning_rate": 0.0001,
+      "loss": 4.202,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07430704997333039,
+      "grad_norm": 0.988510012626648,
+      "learning_rate": 9.999998440456413e-05,
+      "loss": 4.1781,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07504276333940296,
+      "grad_norm": 1.0348843336105347,
+      "learning_rate": 9.99999376182662e-05,
+      "loss": 4.1589,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07577847670547555,
+      "grad_norm": 1.0255194902420044,
+      "learning_rate": 9.999985964113542e-05,
+      "loss": 4.1653,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07651419007154812,
+      "grad_norm": 0.831902801990509,
+      "learning_rate": 9.999975047322044e-05,
+      "loss": 4.1595,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0772499034376207,
+      "grad_norm": 0.964509904384613,
+      "learning_rate": 9.999961011458933e-05,
+      "loss": 4.1414,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07798561680369329,
+      "grad_norm": 0.796290397644043,
+      "learning_rate": 9.99994385653297e-05,
+      "loss": 4.0971,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07872133016976586,
+      "grad_norm": 0.8830526471138,
+      "learning_rate": 9.99992358255485e-05,
+      "loss": 4.1182,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07945704353583843,
+      "grad_norm": 1.0227144956588745,
+      "learning_rate": 9.999900189537226e-05,
+      "loss": 4.0917,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08019275690191102,
+      "grad_norm": 1.0185173749923706,
+      "learning_rate": 9.999873677494689e-05,
+      "loss": 4.1066,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0809284702679836,
+      "grad_norm": 0.8834981322288513,
+      "learning_rate": 9.999844046443776e-05,
+      "loss": 4.0413,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08166418363405617,
+      "grad_norm": 0.9037182331085205,
+      "learning_rate": 9.999811296402975e-05,
+      "loss": 4.1012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08239989700012874,
+      "grad_norm": 0.9970824122428894,
+      "learning_rate": 9.99977542739271e-05,
+      "loss": 4.0548,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08313561036620133,
+      "grad_norm": 0.9786319136619568,
+      "learning_rate": 9.999736439435364e-05,
+      "loss": 4.0613,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0838713237322739,
+      "grad_norm": 0.8991515636444092,
+      "learning_rate": 9.999694332555253e-05,
+      "loss": 4.07,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08460703709834648,
+      "grad_norm": 0.8191851377487183,
+      "learning_rate": 9.999649106778646e-05,
+      "loss": 4.0169,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08534275046441907,
+      "grad_norm": 0.7964187860488892,
+      "learning_rate": 9.999600762133756e-05,
+      "loss": 4.0185,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08607846383049164,
+      "grad_norm": 0.8330606818199158,
+      "learning_rate": 9.99954929865074e-05,
+      "loss": 4.0695,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08681417719656422,
+      "grad_norm": 0.8456650376319885,
+      "learning_rate": 9.999494716361703e-05,
+      "loss": 3.971,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0875498905626368,
+      "grad_norm": 0.7506271600723267,
+      "learning_rate": 9.999437015300694e-05,
+      "loss": 4.0053,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08828560392870938,
+      "grad_norm": 0.7525317668914795,
+      "learning_rate": 9.999376195503709e-05,
+      "loss": 3.9699,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08902131729478195,
+      "grad_norm": 0.7871759533882141,
+      "learning_rate": 9.999312257008685e-05,
+      "loss": 3.9137,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08975703066085453,
+      "grad_norm": 0.8398749828338623,
+      "learning_rate": 9.999245199855512e-05,
+      "loss": 3.9667,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09049274402692711,
+      "grad_norm": 0.7666105031967163,
+      "learning_rate": 9.99917502408602e-05,
+      "loss": 3.9273,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09122845739299969,
+      "grad_norm": 0.8470506072044373,
+      "learning_rate": 9.999101729743985e-05,
+      "loss": 3.9694,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09196417075907226,
+      "grad_norm": 0.7060391306877136,
+      "learning_rate": 9.999025316875129e-05,
+      "loss": 3.9618,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09269988412514485,
+      "grad_norm": 0.7745850682258606,
+      "learning_rate": 9.99894578552712e-05,
+      "loss": 3.9252,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09343559749121742,
+      "grad_norm": 0.8749181032180786,
+      "learning_rate": 9.998863135749575e-05,
+      "loss": 3.9104,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09417131085729,
+      "grad_norm": 0.906620979309082,
+      "learning_rate": 9.998777367594046e-05,
+      "loss": 3.9174,
+      "step": 1280
+    },
+    {
+      "epoch": 0.09490702422336257,
+      "grad_norm": 0.7657135725021362,
+      "learning_rate": 9.998688481114039e-05,
+      "loss": 3.9238,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09564273758943516,
+      "grad_norm": 0.6726455688476562,
+      "learning_rate": 9.998596476365006e-05,
+      "loss": 3.8952,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09637845095550773,
+      "grad_norm": 0.7561966776847839,
+      "learning_rate": 9.998501353404336e-05,
+      "loss": 3.9123,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0971141643215803,
+      "grad_norm": 0.8008501529693604,
+      "learning_rate": 9.998403112291372e-05,
+      "loss": 3.8617,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0978498776876529,
+      "grad_norm": 0.7718631029129028,
+      "learning_rate": 9.998301753087398e-05,
+      "loss": 3.8859,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09858559105372547,
+      "grad_norm": 0.8178690671920776,
+      "learning_rate": 9.998197275855644e-05,
+      "loss": 3.8473,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09932130441979804,
+      "grad_norm": 0.7664726376533508,
+      "learning_rate": 9.998089680661282e-05,
+      "loss": 3.8305,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10005701778587063,
+      "grad_norm": 0.8279010653495789,
+      "learning_rate": 9.997978967571436e-05,
+      "loss": 3.8802,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1007927311519432,
+      "grad_norm": 0.7610680460929871,
+      "learning_rate": 9.997865136655166e-05,
+      "loss": 3.8869,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10152844451801578,
+      "grad_norm": 0.7211320400238037,
+      "learning_rate": 9.997748187983487e-05,
+      "loss": 3.9122,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10226415788408835,
+      "grad_norm": 0.7668148279190063,
+      "learning_rate": 9.997628121629349e-05,
+      "loss": 3.8238,
+      "step": 1390
+    },
+    {
+      "epoch": 0.10299987125016094,
+      "grad_norm": 0.6940723061561584,
+      "learning_rate": 9.997504937667654e-05,
+      "loss": 3.8464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.10373558461623351,
+      "grad_norm": 0.681667149066925,
+      "learning_rate": 9.997378636175245e-05,
+      "loss": 3.8182,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10447129798230609,
+      "grad_norm": 0.6430841684341431,
+      "learning_rate": 9.99724921723091e-05,
+      "loss": 3.8379,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10520701134837868,
+      "grad_norm": 0.6482489705085754,
+      "learning_rate": 9.997116680915389e-05,
+      "loss": 3.8117,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10594272471445125,
+      "grad_norm": 0.6649356484413147,
+      "learning_rate": 9.996981027311352e-05,
+      "loss": 3.8452,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10667843808052382,
+      "grad_norm": 0.757146954536438,
+      "learning_rate": 9.996842256503428e-05,
+      "loss": 3.7886,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10741415144659641,
+      "grad_norm": 0.6373863816261292,
+      "learning_rate": 9.99670036857818e-05,
+      "loss": 3.7975,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10814986481266899,
+      "grad_norm": 0.6672409176826477,
+      "learning_rate": 9.996555363624124e-05,
+      "loss": 3.843,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10888557817874156,
+      "grad_norm": 0.6595909595489502,
+      "learning_rate": 9.996407241731717e-05,
+      "loss": 3.8397,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10962129154481413,
+      "grad_norm": 0.725779116153717,
+      "learning_rate": 9.996256002993358e-05,
+      "loss": 3.8264,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11035700491088672,
+      "grad_norm": 0.6590442061424255,
+      "learning_rate": 9.99610164750339e-05,
+      "loss": 3.8375,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1110927182769593,
+      "grad_norm": 0.707144558429718,
+      "learning_rate": 9.99594417535811e-05,
+      "loss": 3.7805,
+      "step": 1510
+    },
+    {
+      "epoch": 0.11182843164303187,
+      "grad_norm": 0.7135934233665466,
+      "learning_rate": 9.995783586655745e-05,
+      "loss": 3.7696,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11256414500910446,
+      "grad_norm": 0.7143461108207703,
+      "learning_rate": 9.995619881496474e-05,
+      "loss": 3.7558,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11329985837517703,
+      "grad_norm": 0.6768718957901001,
+      "learning_rate": 9.995453059982422e-05,
+      "loss": 3.8221,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1140355717412496,
+      "grad_norm": 0.6788365244865417,
+      "learning_rate": 9.995283122217653e-05,
+      "loss": 3.7792,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1147712851073222,
+      "grad_norm": 0.6880599856376648,
+      "learning_rate": 9.995110068308179e-05,
+      "loss": 3.7191,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11550699847339477,
+      "grad_norm": 0.6511822938919067,
+      "learning_rate": 9.994933898361951e-05,
+      "loss": 3.7585,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11624271183946734,
+      "grad_norm": 0.6109036803245544,
+      "learning_rate": 9.99475461248887e-05,
+      "loss": 3.7555,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11697842520553992,
+      "grad_norm": 0.6387907862663269,
+      "learning_rate": 9.994572210800777e-05,
+      "loss": 3.7283,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1177141385716125,
+      "grad_norm": 0.6535074710845947,
+      "learning_rate": 9.994386693411455e-05,
+      "loss": 3.7781,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11844985193768508,
+      "grad_norm": 0.6718204617500305,
+      "learning_rate": 9.994198060436637e-05,
+      "loss": 3.7622,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11918556530375765,
+      "grad_norm": 0.75942063331604,
+      "learning_rate": 9.994006311993991e-05,
+      "loss": 3.6595,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11992127866983024,
+      "grad_norm": 0.6169456243515015,
+      "learning_rate": 9.993811448203136e-05,
+      "loss": 3.6921,
+      "step": 1630
+    },
+    {
+      "epoch": 0.12065699203590281,
+      "grad_norm": 0.6598173379898071,
+      "learning_rate": 9.993613469185631e-05,
+      "loss": 3.6957,
+      "step": 1640
+    },
+    {
+      "epoch": 0.12139270540197539,
+      "grad_norm": 0.6640283465385437,
+      "learning_rate": 9.993412375064979e-05,
+      "loss": 3.7507,
+      "step": 1650
+    },
+    {
+      "epoch": 0.12212841876804797,
+      "grad_norm": 0.720782458782196,
+      "learning_rate": 9.993208165966623e-05,
+      "loss": 3.741,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12286413213412055,
+      "grad_norm": 0.6343806385993958,
+      "learning_rate": 9.993000842017957e-05,
+      "loss": 3.7112,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12359984550019312,
+      "grad_norm": 0.72925865650177,
+      "learning_rate": 9.99279040334831e-05,
+      "loss": 3.7455,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1243355588662657,
+      "grad_norm": 0.7056775093078613,
+      "learning_rate": 9.99257685008896e-05,
+      "loss": 3.7007,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12507127223233827,
+      "grad_norm": 0.6219266057014465,
+      "learning_rate": 9.992360182373122e-05,
+      "loss": 3.6704,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12580698559841086,
+      "grad_norm": 0.8017805814743042,
+      "learning_rate": 9.992140400335958e-05,
+      "loss": 3.7098,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12654269896448345,
+      "grad_norm": 0.6517687439918518,
+      "learning_rate": 9.991917504114574e-05,
+      "loss": 3.7145,
+      "step": 1720
+    },
+    {
+      "epoch": 0.127278412330556,
+      "grad_norm": 0.6560338735580444,
+      "learning_rate": 9.991691493848015e-05,
+      "loss": 3.7046,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1280141256966286,
+      "grad_norm": 0.7055456638336182,
+      "learning_rate": 9.991462369677267e-05,
+      "loss": 3.6855,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12874983906270118,
+      "grad_norm": 0.6441034078598022,
+      "learning_rate": 9.991230131745268e-05,
+      "loss": 3.7338,
+      "step": 1750
+    },
+    {
+      "epoch": 0.12948555242877374,
+      "grad_norm": 0.6746709942817688,
+      "learning_rate": 9.990994780196889e-05,
+      "loss": 3.7237,
+      "step": 1760
+    },
+    {
+      "epoch": 0.13022126579484633,
+      "grad_norm": 0.6609126925468445,
+      "learning_rate": 9.990756315178945e-05,
+      "loss": 3.7393,
+      "step": 1770
+    },
+    {
+      "epoch": 0.13095697916091892,
+      "grad_norm": 0.648073673248291,
+      "learning_rate": 9.990514736840197e-05,
+      "loss": 3.7057,
+      "step": 1780
+    },
+    {
+      "epoch": 0.13169269252699148,
+      "grad_norm": 0.6735105514526367,
+      "learning_rate": 9.990270045331344e-05,
+      "loss": 3.7574,
+      "step": 1790
+    },
+    {
+      "epoch": 0.13242840589306407,
+      "grad_norm": 0.6125868558883667,
+      "learning_rate": 9.99002224080503e-05,
+      "loss": 3.6548,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13316411925913665,
+      "grad_norm": 0.5899696946144104,
+      "learning_rate": 9.98977132341584e-05,
+      "loss": 3.624,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13389983262520921,
+      "grad_norm": 0.6936771273612976,
+      "learning_rate": 9.989517293320299e-05,
+      "loss": 3.7022,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1346355459912818,
+      "grad_norm": 0.6176466345787048,
+      "learning_rate": 9.989260150676876e-05,
+      "loss": 3.697,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13537125935735436,
+      "grad_norm": 0.6583747863769531,
+      "learning_rate": 9.988999895645981e-05,
+      "loss": 3.7048,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13610697272342695,
+      "grad_norm": 0.6942030191421509,
+      "learning_rate": 9.988736528389969e-05,
+      "loss": 3.6784,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13684268608949954,
+      "grad_norm": 0.6578043103218079,
+      "learning_rate": 9.988470049073126e-05,
+      "loss": 3.6411,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1375783994555721,
+      "grad_norm": 0.648202657699585,
+      "learning_rate": 9.988200457861694e-05,
+      "loss": 3.6472,
+      "step": 1870
+    },
+    {
+      "epoch": 0.13831411282164469,
+      "grad_norm": 0.6567854285240173,
+      "learning_rate": 9.987927754923843e-05,
+      "loss": 3.6734,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13904982618771727,
+      "grad_norm": 0.6515029668807983,
+      "learning_rate": 9.987651940429695e-05,
+      "loss": 3.6463,
+      "step": 1890
+    },
+    {
+      "epoch": 0.13978553955378983,
+      "grad_norm": 0.597316324710846,
+      "learning_rate": 9.987373014551302e-05,
+      "loss": 3.5951,
+      "step": 1900
+    },
+    {
+      "epoch": 0.14052125291986242,
+      "grad_norm": 0.6288499236106873,
+      "learning_rate": 9.987090977462668e-05,
+      "loss": 3.675,
+      "step": 1910
+    },
+    {
+      "epoch": 0.141256966285935,
+      "grad_norm": 0.6451112627983093,
+      "learning_rate": 9.986805829339729e-05,
+      "loss": 3.6552,
+      "step": 1920
+    },
+    {
+      "epoch": 0.14199267965200757,
+      "grad_norm": 0.6517444252967834,
+      "learning_rate": 9.986517570360368e-05,
+      "loss": 3.6206,
+      "step": 1930
+    },
+    {
+      "epoch": 0.14272839301808016,
+      "grad_norm": 0.6413518786430359,
+      "learning_rate": 9.986226200704404e-05,
+      "loss": 3.6848,
+      "step": 1940
+    },
+    {
+      "epoch": 0.14346410638415275,
+      "grad_norm": 0.6431183815002441,
+      "learning_rate": 9.9859317205536e-05,
+      "loss": 3.6339,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1441998197502253,
+      "grad_norm": 0.7160407900810242,
+      "learning_rate": 9.985634130091657e-05,
+      "loss": 3.6306,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1449355331162979,
+      "grad_norm": 0.602085292339325,
+      "learning_rate": 9.985333429504216e-05,
+      "loss": 3.6556,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14567124648237048,
+      "grad_norm": 0.6392213702201843,
+      "learning_rate": 9.985029618978863e-05,
+      "loss": 3.6195,
+      "step": 1980
+    },
+    {
+      "epoch": 0.14640695984844304,
+      "grad_norm": 0.5774185657501221,
+      "learning_rate": 9.984722698705115e-05,
+      "loss": 3.6463,
+      "step": 1990
+    },
+    {
+      "epoch": 0.14714267321451563,
+      "grad_norm": 0.6097477674484253,
+      "learning_rate": 9.984412668874442e-05,
+      "loss": 3.6759,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1478783865805882,
+      "grad_norm": 0.5511993765830994,
+      "learning_rate": 9.984099529680237e-05,
+      "loss": 3.6254,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14861409994666078,
+      "grad_norm": 0.6293554306030273,
+      "learning_rate": 9.983783281317847e-05,
+      "loss": 3.6285,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14934981331273336,
+      "grad_norm": 0.5784502625465393,
+      "learning_rate": 9.983463923984551e-05,
+      "loss": 3.5952,
+      "step": 2030
+    },
+    {
+      "epoch": 0.15008552667880593,
+      "grad_norm": 0.5587512254714966,
+      "learning_rate": 9.983141457879572e-05,
+      "loss": 3.5877,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1508212400448785,
+      "grad_norm": 0.5835023522377014,
+      "learning_rate": 9.982815883204068e-05,
+      "loss": 3.6139,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1515569534109511,
+      "grad_norm": 0.6096065044403076,
+      "learning_rate": 9.982487200161139e-05,
+      "loss": 3.6261,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15229266677702366,
+      "grad_norm": 0.6385859847068787,
+      "learning_rate": 9.982155408955822e-05,
+      "loss": 3.6228,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15302838014309625,
+      "grad_norm": 0.6276549100875854,
+      "learning_rate": 9.981820509795096e-05,
+      "loss": 3.6525,
+      "step": 2080
+    },
+    {
+      "epoch": 0.15376409350916884,
+      "grad_norm": 0.5835126638412476,
+      "learning_rate": 9.981482502887878e-05,
+      "loss": 3.6081,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1544998068752414,
+      "grad_norm": 0.6056969165802002,
+      "learning_rate": 9.981141388445019e-05,
+      "loss": 3.6241,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15523552024131398,
+      "grad_norm": 0.6270286440849304,
+      "learning_rate": 9.980797166679314e-05,
+      "loss": 3.6231,
+      "step": 2110
+    },
+    {
+      "epoch": 0.15597123360738657,
+      "grad_norm": 0.6744200587272644,
+      "learning_rate": 9.980449837805494e-05,
+      "loss": 3.5789,
+      "step": 2120
+    },
+    {
+      "epoch": 0.15670694697345913,
+      "grad_norm": 0.5492793321609497,
+      "learning_rate": 9.98009940204023e-05,
+      "loss": 3.58,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15744266033953172,
+      "grad_norm": 0.601390540599823,
+      "learning_rate": 9.979745859602129e-05,
+      "loss": 3.6059,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1581783737056043,
+      "grad_norm": 0.6107606887817383,
+      "learning_rate": 9.979389210711737e-05,
+      "loss": 3.5928,
+      "step": 2150
+    },
+    {
+      "epoch": 0.15891408707167687,
+      "grad_norm": 0.6118071675300598,
+      "learning_rate": 9.979029455591538e-05,
+      "loss": 3.5771,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15964980043774946,
+      "grad_norm": 0.6007488369941711,
+      "learning_rate": 9.978666594465953e-05,
+      "loss": 3.5729,
+      "step": 2170
+    },
+    {
+      "epoch": 0.16038551380382204,
+      "grad_norm": 0.5910535454750061,
+      "learning_rate": 9.978300627561343e-05,
+      "loss": 3.5953,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1611212271698946,
+      "grad_norm": 0.5529370903968811,
+      "learning_rate": 9.977931555106002e-05,
+      "loss": 3.6167,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1618569405359672,
+      "grad_norm": 0.6961104273796082,
+      "learning_rate": 9.977559377330163e-05,
+      "loss": 3.5541,
+      "step": 2200
+    },
+    {
+      "epoch": 0.16259265390203975,
+      "grad_norm": 0.5334804654121399,
+      "learning_rate": 9.977184094466001e-05,
+      "loss": 3.6368,
+      "step": 2210
+    },
+    {
+      "epoch": 0.16332836726811234,
+      "grad_norm": 0.6472899913787842,
+      "learning_rate": 9.976805706747622e-05,
+      "loss": 3.5896,
+      "step": 2220
+    },
+    {
+      "epoch": 0.16406408063418493,
+      "grad_norm": 0.6276665329933167,
+      "learning_rate": 9.976424214411071e-05,
+      "loss": 3.5907,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1647997940002575,
+      "grad_norm": 0.5503706932067871,
+      "learning_rate": 9.976039617694328e-05,
+      "loss": 3.6037,
+      "step": 2240
+    },
+    {
+      "epoch": 0.16553550736633008,
+      "grad_norm": 0.5655608773231506,
+      "learning_rate": 9.975651916837313e-05,
+      "loss": 3.6157,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16627122073240266,
+      "grad_norm": 0.5357344746589661,
+      "learning_rate": 9.975261112081879e-05,
+      "loss": 3.583,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16700693409847522,
+      "grad_norm": 0.5251188278198242,
+      "learning_rate": 9.974867203671819e-05,
+      "loss": 3.5705,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1677426474645478,
+      "grad_norm": 0.5895013809204102,
+      "learning_rate": 9.974470191852858e-05,
+      "loss": 3.5648,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1684783608306204,
+      "grad_norm": 0.5837428569793701,
+      "learning_rate": 9.974070076872662e-05,
+      "loss": 3.5607,
+      "step": 2290
+    },
+    {
+      "epoch": 0.16921407419669296,
+      "grad_norm": 0.5866957306861877,
+      "learning_rate": 9.973666858980824e-05,
+      "loss": 3.5689,
+      "step": 2300
+    },
+    {
+      "epoch": 0.16994978756276555,
+      "grad_norm": 0.5927156209945679,
+      "learning_rate": 9.973260538428884e-05,
+      "loss": 3.5806,
+      "step": 2310
+    },
+    {
+      "epoch": 0.17068550092883814,
+      "grad_norm": 0.5668119788169861,
+      "learning_rate": 9.972851115470307e-05,
+      "loss": 3.5247,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1714212142949107,
+      "grad_norm": 0.5907779335975647,
+      "learning_rate": 9.972438590360503e-05,
+      "loss": 3.5757,
+      "step": 2330
+    },
+    {
+      "epoch": 0.17215692766098328,
+      "grad_norm": 0.5702281594276428,
+      "learning_rate": 9.97202296335681e-05,
+      "loss": 3.5698,
+      "step": 2340
+    },
+    {
+      "epoch": 0.17289264102705587,
+      "grad_norm": 0.5785143375396729,
+      "learning_rate": 9.971604234718504e-05,
+      "loss": 3.5547,
+      "step": 2350
+    },
+    {
+      "epoch": 0.17362835439312843,
+      "grad_norm": 0.5959532856941223,
+      "learning_rate": 9.971182404706792e-05,
+      "loss": 3.4787,
+      "step": 2360
+    },
+    {
+      "epoch": 0.17436406775920102,
+      "grad_norm": 0.5923709869384766,
+      "learning_rate": 9.970757473584824e-05,
+      "loss": 3.5127,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1750997811252736,
+      "grad_norm": 0.6314843893051147,
+      "learning_rate": 9.970329441617676e-05,
+      "loss": 3.5123,
+      "step": 2380
+    },
+    {
+      "epoch": 0.17583549449134617,
+      "grad_norm": 0.6579912304878235,
+      "learning_rate": 9.969898309072363e-05,
+      "loss": 3.5295,
+      "step": 2390
+    },
+    {
+      "epoch": 0.17657120785741875,
+      "grad_norm": 0.5852422118186951,
+      "learning_rate": 9.969464076217831e-05,
+      "loss": 3.5761,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17730692122349132,
+      "grad_norm": 0.5590935945510864,
+      "learning_rate": 9.969026743324967e-05,
+      "loss": 3.5697,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1780426345895639,
+      "grad_norm": 0.5512566566467285,
+      "learning_rate": 9.968586310666583e-05,
+      "loss": 3.5069,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1787783479556365,
+      "grad_norm": 0.6014286279678345,
+      "learning_rate": 9.968142778517429e-05,
+      "loss": 3.5985,
+      "step": 2430
+    },
+    {
+      "epoch": 0.17951406132170905,
+      "grad_norm": 0.5852993726730347,
+      "learning_rate": 9.967696147154187e-05,
+      "loss": 3.4978,
+      "step": 2440
+    },
+    {
+      "epoch": 0.18024977468778164,
+      "grad_norm": 0.5915689468383789,
+      "learning_rate": 9.967246416855475e-05,
+      "loss": 3.5926,
+      "step": 2450
+    },
+    {
+      "epoch": 0.18098548805385423,
+      "grad_norm": 0.5336611866950989,
+      "learning_rate": 9.966793587901844e-05,
+      "loss": 3.5764,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1817212014199268,
+      "grad_norm": 0.6106781959533691,
+      "learning_rate": 9.966337660575775e-05,
+      "loss": 3.5095,
+      "step": 2470
+    },
+    {
+      "epoch": 0.18245691478599937,
+      "grad_norm": 0.5323026180267334,
+      "learning_rate": 9.965878635161682e-05,
+      "loss": 3.4732,
+      "step": 2480
+    },
+    {
+      "epoch": 0.18319262815207196,
+      "grad_norm": 0.5882947444915771,
+      "learning_rate": 9.965416511945915e-05,
+      "loss": 3.56,
+      "step": 2490
+    },
+    {
+      "epoch": 0.18392834151814452,
+      "grad_norm": 0.5978416800498962,
+      "learning_rate": 9.964951291216755e-05,
+      "loss": 3.4982,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1846640548842171,
+      "grad_norm": 0.5574873089790344,
+      "learning_rate": 9.964482973264413e-05,
+      "loss": 3.5024,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1853997682502897,
+      "grad_norm": 0.6445927023887634,
+      "learning_rate": 9.964011558381035e-05,
+      "loss": 3.4935,
+      "step": 2520
+    },
+    {
+      "epoch": 0.18613548161636226,
+      "grad_norm": 0.5893427729606628,
+      "learning_rate": 9.963537046860697e-05,
+      "loss": 3.5344,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18687119498243485,
+      "grad_norm": 0.5860798954963684,
+      "learning_rate": 9.963059438999409e-05,
+      "loss": 3.5498,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18760690834850743,
+      "grad_norm": 0.5362831950187683,
+      "learning_rate": 9.962578735095109e-05,
+      "loss": 3.4655,
+      "step": 2550
+    },
+    {
+      "epoch": 0.18834262171458,
+      "grad_norm": 0.5293241143226624,
+      "learning_rate": 9.96209493544767e-05,
+      "loss": 3.5236,
+      "step": 2560
+    },
+    {
+      "epoch": 0.18907833508065258,
+      "grad_norm": 0.6069090366363525,
+      "learning_rate": 9.961608040358894e-05,
+      "loss": 3.5332,
+      "step": 2570
+    },
+    {
+      "epoch": 0.18981404844672514,
+      "grad_norm": 0.5620616674423218,
+      "learning_rate": 9.961118050132517e-05,
+      "loss": 3.4838,
+      "step": 2580
+    },
+    {
+      "epoch": 0.19054976181279773,
+      "grad_norm": 0.5729257464408875,
+      "learning_rate": 9.9606249650742e-05,
+      "loss": 3.4809,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19128547517887032,
+      "grad_norm": 0.5876139998435974,
+      "learning_rate": 9.96012878549154e-05,
+      "loss": 3.4935,
+      "step": 2600
+    },
+    {
+      "epoch": 0.19202118854494288,
+      "grad_norm": 0.5324752330780029,
+      "learning_rate": 9.959629511694061e-05,
+      "loss": 3.5621,
+      "step": 2610
+    },
+    {
+      "epoch": 0.19275690191101547,
+      "grad_norm": 0.5787373781204224,
+      "learning_rate": 9.959127143993219e-05,
+      "loss": 3.4896,
+      "step": 2620
+    },
+    {
+      "epoch": 0.19349261527708805,
+      "grad_norm": 0.5609232783317566,
+      "learning_rate": 9.958621682702403e-05,
+      "loss": 3.4973,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1942283286431606,
+      "grad_norm": 0.5869426131248474,
+      "learning_rate": 9.958113128136923e-05,
+      "loss": 3.5203,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1949640420092332,
+      "grad_norm": 0.609747052192688,
+      "learning_rate": 9.957601480614029e-05,
+      "loss": 3.5297,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1956997553753058,
+      "grad_norm": 0.5092261433601379,
+      "learning_rate": 9.957086740452894e-05,
+      "loss": 3.4951,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19643546874137835,
+      "grad_norm": 0.5409265756607056,
+      "learning_rate": 9.956568907974621e-05,
+      "loss": 3.528,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19717118210745094,
+      "grad_norm": 0.5979306101799011,
+      "learning_rate": 9.956047983502245e-05,
+      "loss": 3.5356,
+      "step": 2680
+    },
+    {
+      "epoch": 0.19790689547352353,
+      "grad_norm": 0.5668174624443054,
+      "learning_rate": 9.955523967360725e-05,
+      "loss": 3.5797,
+      "step": 2690
+    },
+    {
+      "epoch": 0.19864260883959609,
+      "grad_norm": 0.5992650389671326,
+      "learning_rate": 9.954996859876953e-05,
+      "loss": 3.5107,
+      "step": 2700
+    },
+    {
+      "epoch": 0.19937832220566867,
+      "grad_norm": 0.5725685358047485,
+      "learning_rate": 9.954466661379748e-05,
+      "loss": 3.5161,
+      "step": 2710
+    },
+    {
+      "epoch": 0.20011403557174126,
+      "grad_norm": 0.5398104786872864,
+      "learning_rate": 9.953933372199855e-05,
+      "loss": 3.5191,
+      "step": 2720
+    },
+    {
+      "epoch": 0.20084974893781382,
+      "grad_norm": 0.5120781064033508,
+      "learning_rate": 9.953396992669953e-05,
+      "loss": 3.4936,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2015854623038864,
+      "grad_norm": 0.5185467004776001,
+      "learning_rate": 9.952857523124641e-05,
+      "loss": 3.4757,
+      "step": 2740
+    },
+    {
+      "epoch": 0.202321175669959,
+      "grad_norm": 0.6107959747314453,
+      "learning_rate": 9.952314963900453e-05,
+      "loss": 3.4972,
+      "step": 2750
+    },
+    {
+      "epoch": 0.20305688903603156,
+      "grad_norm": 0.5769765377044678,
+      "learning_rate": 9.951769315335844e-05,
+      "loss": 3.4983,
+      "step": 2760
+    },
+    {
+      "epoch": 0.20379260240210414,
+      "grad_norm": 0.5694507956504822,
+      "learning_rate": 9.951220577771201e-05,
+      "loss": 3.5063,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2045283157681767,
+      "grad_norm": 0.5544769763946533,
+      "learning_rate": 9.950668751548835e-05,
+      "loss": 3.507,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2052640291342493,
+      "grad_norm": 0.5395680665969849,
+      "learning_rate": 9.950113837012984e-05,
+      "loss": 3.5315,
+      "step": 2790
+    },
+    {
+      "epoch": 0.20599974250032188,
+      "grad_norm": 0.575812816619873,
+      "learning_rate": 9.949555834509817e-05,
+      "loss": 3.4873,
+      "step": 2800
+    },
+    {
+      "epoch": 0.20673545586639444,
+      "grad_norm": 0.5248579978942871,
+      "learning_rate": 9.948994744387419e-05,
+      "loss": 3.4849,
+      "step": 2810
+    },
+    {
+      "epoch": 0.20747116923246703,
+      "grad_norm": 0.5287348031997681,
+      "learning_rate": 9.948430566995816e-05,
+      "loss": 3.5028,
+      "step": 2820
+    },
+    {
+      "epoch": 0.20820688259853962,
+      "grad_norm": 0.5448064804077148,
+      "learning_rate": 9.947863302686945e-05,
+      "loss": 3.4801,
+      "step": 2830
+    },
+    {
+      "epoch": 0.20894259596461218,
+      "grad_norm": 0.583583652973175,
+      "learning_rate": 9.947292951814679e-05,
+      "loss": 3.4899,
+      "step": 2840
+    },
+    {
+      "epoch": 0.20967830933068476,
+      "grad_norm": 0.5897386074066162,
+      "learning_rate": 9.946719514734813e-05,
+      "loss": 3.5659,
+      "step": 2850
+    },
+    {
+      "epoch": 0.21041402269675735,
+      "grad_norm": 0.5261275172233582,
+      "learning_rate": 9.946142991805062e-05,
+      "loss": 3.4662,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2111497360628299,
+      "grad_norm": 0.5870264768600464,
+      "learning_rate": 9.945563383385079e-05,
+      "loss": 3.5123,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2118854494289025,
+      "grad_norm": 0.5746486186981201,
+      "learning_rate": 9.944980689836429e-05,
+      "loss": 3.4351,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2126211627949751,
+      "grad_norm": 0.6177893877029419,
+      "learning_rate": 9.944394911522606e-05,
+      "loss": 3.5206,
+      "step": 2890
+    },
+    {
+      "epoch": 0.21335687616104765,
+      "grad_norm": 0.5619845390319824,
+      "learning_rate": 9.94380604880903e-05,
+      "loss": 3.4801,
+      "step": 2900
+    },
+    {
+      "epoch": 0.21409258952712024,
+      "grad_norm": 0.5628277063369751,
+      "learning_rate": 9.943214102063043e-05,
+      "loss": 3.5564,
+      "step": 2910
+    },
+    {
+      "epoch": 0.21482830289319282,
+      "grad_norm": 0.5306517481803894,
+      "learning_rate": 9.942619071653914e-05,
+      "loss": 3.5094,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21556401625926538,
+      "grad_norm": 0.6212943196296692,
+      "learning_rate": 9.942020957952831e-05,
+      "loss": 3.5182,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21629972962533797,
+      "grad_norm": 0.5883834362030029,
+      "learning_rate": 9.941419761332908e-05,
+      "loss": 3.4715,
+      "step": 2940
+    },
+    {
+      "epoch": 0.21703544299141056,
+      "grad_norm": 0.5104158520698547,
+      "learning_rate": 9.940815482169184e-05,
+      "loss": 3.4712,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21777115635748312,
+      "grad_norm": 0.5349141955375671,
+      "learning_rate": 9.940208120838616e-05,
+      "loss": 3.4517,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2185068697235557,
+      "grad_norm": 0.555456280708313,
+      "learning_rate": 9.939597677720089e-05,
+      "loss": 3.4984,
+      "step": 2970
+    },
+    {
+      "epoch": 0.21924258308962827,
+      "grad_norm": 0.5867215991020203,
+      "learning_rate": 9.938984153194406e-05,
+      "loss": 3.4821,
+      "step": 2980
+    },
+    {
+      "epoch": 0.21997829645570086,
+      "grad_norm": 0.5781378149986267,
+      "learning_rate": 9.938367547644296e-05,
+      "loss": 3.4437,
+      "step": 2990
+    },
+    {
+      "epoch": 0.22071400982177344,
+      "grad_norm": 0.5571138858795166,
+      "learning_rate": 9.937747861454407e-05,
+      "loss": 3.4935,
+      "step": 3000
+    },
+    {
+      "epoch": 0.221449723187846,
+      "grad_norm": 0.47858932614326477,
+      "learning_rate": 9.93712509501131e-05,
+      "loss": 3.499,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2221854365539186,
+      "grad_norm": 0.500126838684082,
+      "learning_rate": 9.936499248703499e-05,
+      "loss": 3.4822,
+      "step": 3020
+    },
+    {
+      "epoch": 0.22292114991999118,
+      "grad_norm": 0.5339503288269043,
+      "learning_rate": 9.935870322921387e-05,
+      "loss": 3.5028,
+      "step": 3030
+    },
+    {
+      "epoch": 0.22365686328606374,
+      "grad_norm": 0.5465017557144165,
+      "learning_rate": 9.93523831805731e-05,
+      "loss": 3.4338,
+      "step": 3040
+    },
+    {
+      "epoch": 0.22439257665213633,
+      "grad_norm": 0.5432729125022888,
+      "learning_rate": 9.934603234505519e-05,
+      "loss": 3.4831,
+      "step": 3050
+    },
+    {
+      "epoch": 0.22512829001820892,
+      "grad_norm": 0.5051527619361877,
+      "learning_rate": 9.933965072662197e-05,
+      "loss": 3.4473,
+      "step": 3060
+    },
+    {
+      "epoch": 0.22586400338428148,
+      "grad_norm": 0.5150954127311707,
+      "learning_rate": 9.933323832925437e-05,
+      "loss": 3.4755,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22659971675035406,
+      "grad_norm": 0.5967200398445129,
+      "learning_rate": 9.932679515695253e-05,
+      "loss": 3.4938,
+      "step": 3080
+    },
+    {
+      "epoch": 0.22733543011642665,
+      "grad_norm": 0.5499495267868042,
+      "learning_rate": 9.932032121373587e-05,
+      "loss": 3.5033,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2280711434824992,
+      "grad_norm": 0.502854585647583,
+      "learning_rate": 9.931381650364291e-05,
+      "loss": 3.3716,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2288068568485718,
+      "grad_norm": 0.5126801133155823,
+      "learning_rate": 9.93072810307314e-05,
+      "loss": 3.4727,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2295425702146444,
+      "grad_norm": 0.48045796155929565,
+      "learning_rate": 9.930071479907831e-05,
+      "loss": 3.4312,
+      "step": 3120
+    },
+    {
+      "epoch": 0.23027828358071695,
+      "grad_norm": 0.5664313435554504,
+      "learning_rate": 9.929411781277974e-05,
+      "loss": 3.4225,
+      "step": 3130
+    },
+    {
+      "epoch": 0.23101399694678953,
+      "grad_norm": 0.5236510634422302,
+      "learning_rate": 9.928749007595101e-05,
+      "loss": 3.4255,
+      "step": 3140
+    },
+    {
+      "epoch": 0.23174971031286212,
+      "grad_norm": 0.5875115990638733,
+      "learning_rate": 9.928083159272666e-05,
+      "loss": 3.4699,
+      "step": 3150
+    },
+    {
+      "epoch": 0.23248542367893468,
+      "grad_norm": 0.5420896410942078,
+      "learning_rate": 9.92741423672603e-05,
+      "loss": 3.3887,
+      "step": 3160
+    },
+    {
+      "epoch": 0.23322113704500727,
+      "grad_norm": 0.5325738191604614,
+      "learning_rate": 9.926742240372483e-05,
+      "loss": 3.4807,
+      "step": 3170
+    },
+    {
+      "epoch": 0.23395685041107983,
+      "grad_norm": 0.5513830780982971,
+      "learning_rate": 9.926067170631227e-05,
+      "loss": 3.4647,
+      "step": 3180
+    },
+    {
+      "epoch": 0.23469256377715242,
+      "grad_norm": 0.5124702453613281,
+      "learning_rate": 9.925389027923382e-05,
+      "loss": 3.4392,
+      "step": 3190
+    },
+    {
+      "epoch": 0.235428277143225,
+      "grad_norm": 0.532706081867218,
+      "learning_rate": 9.924707812671985e-05,
+      "loss": 3.4456,
+      "step": 3200
+    },
+    {
+      "epoch": 0.23616399050929757,
+      "grad_norm": 0.5108603835105896,
+      "learning_rate": 9.924023525301991e-05,
+      "loss": 3.4621,
+      "step": 3210
+    },
+    {
+      "epoch": 0.23689970387537015,
+      "grad_norm": 0.5478849411010742,
+      "learning_rate": 9.92333616624027e-05,
+      "loss": 3.4602,
+      "step": 3220
+    },
+    {
+      "epoch": 0.23763541724144274,
+      "grad_norm": 0.481143057346344,
+      "learning_rate": 9.922645735915608e-05,
+      "loss": 3.4647,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2383711306075153,
+      "grad_norm": 0.5544288754463196,
+      "learning_rate": 9.921952234758709e-05,
+      "loss": 3.4312,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2391068439735879,
+      "grad_norm": 0.5568479299545288,
+      "learning_rate": 9.921255663202189e-05,
+      "loss": 3.4306,
+      "step": 3250
+    },
+    {
+      "epoch": 0.23984255733966048,
+      "grad_norm": 0.5197471380233765,
+      "learning_rate": 9.92055602168058e-05,
+      "loss": 3.4488,
+      "step": 3260
+    },
+    {
+      "epoch": 0.24057827070573304,
+      "grad_norm": 0.5361900329589844,
+      "learning_rate": 9.919853310630336e-05,
+      "loss": 3.395,
+      "step": 3270
+    },
+    {
+      "epoch": 0.24131398407180563,
+      "grad_norm": 0.5017499327659607,
+      "learning_rate": 9.919147530489816e-05,
+      "loss": 3.4654,
+      "step": 3280
+    },
+    {
+      "epoch": 0.24204969743787821,
+      "grad_norm": 0.5076097846031189,
+      "learning_rate": 9.9184386816993e-05,
+      "loss": 3.3991,
+      "step": 3290
+    },
+    {
+      "epoch": 0.24278541080395077,
+      "grad_norm": 0.5834246873855591,
+      "learning_rate": 9.917726764700981e-05,
+      "loss": 3.4058,
+      "step": 3300
+    },
+    {
+      "epoch": 0.24352112417002336,
+      "grad_norm": 0.5632571578025818,
+      "learning_rate": 9.917011779938961e-05,
+      "loss": 3.4442,
+      "step": 3310
+    },
+    {
+      "epoch": 0.24425683753609595,
+      "grad_norm": 0.5473082065582275,
+      "learning_rate": 9.916293727859265e-05,
+      "loss": 3.4536,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2449925509021685,
+      "grad_norm": 0.5029418468475342,
+      "learning_rate": 9.915572608909824e-05,
+      "loss": 3.4263,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2457282642682411,
+      "grad_norm": 0.5429378151893616,
+      "learning_rate": 9.914848423540483e-05,
+      "loss": 3.4652,
+      "step": 3340
+    },
+    {
+      "epoch": 0.24646397763431366,
+      "grad_norm": 0.5347001552581787,
+      "learning_rate": 9.914121172203005e-05,
+      "loss": 3.4009,
+      "step": 3350
+    },
+    {
+      "epoch": 0.24719969100038625,
+      "grad_norm": 0.538618266582489,
+      "learning_rate": 9.913390855351058e-05,
+      "loss": 3.457,
+      "step": 3360
+    },
+    {
+      "epoch": 0.24793540436645883,
+      "grad_norm": 0.5618345141410828,
+      "learning_rate": 9.912657473440232e-05,
+      "loss": 3.4588,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2486711177325314,
+      "grad_norm": 0.5444225072860718,
+      "learning_rate": 9.911921026928019e-05,
+      "loss": 3.4401,
+      "step": 3380
+    },
+    {
+      "epoch": 0.24940683109860398,
+      "grad_norm": 0.5804809927940369,
+      "learning_rate": 9.911181516273826e-05,
+      "loss": 3.3904,
+      "step": 3390
+    },
+    {
+      "epoch": 0.25014254446467654,
+      "grad_norm": 0.5047810673713684,
+      "learning_rate": 9.910438941938978e-05,
+      "loss": 3.42,
+      "step": 3400
+    },
+    {
+      "epoch": 0.25087825783074913,
+      "grad_norm": 0.5213153958320618,
+      "learning_rate": 9.9096933043867e-05,
+      "loss": 3.4572,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2516139711968217,
+      "grad_norm": 0.49462226033210754,
+      "learning_rate": 9.908944604082138e-05,
+      "loss": 3.432,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2523496845628943,
+      "grad_norm": 0.5353688597679138,
+      "learning_rate": 9.908192841492343e-05,
+      "loss": 3.4174,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2530853979289669,
+      "grad_norm": 0.4931166470050812,
+      "learning_rate": 9.907438017086277e-05,
+      "loss": 3.4175,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2538211112950394,
+      "grad_norm": 0.5079531669616699,
+      "learning_rate": 9.906680131334813e-05,
+      "loss": 3.478,
+      "step": 3450
+    },
+    {
+      "epoch": 0.254556824661112,
+      "grad_norm": 0.5377930402755737,
+      "learning_rate": 9.905919184710733e-05,
+      "loss": 3.4353,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2552925380271846,
+      "grad_norm": 0.5525156855583191,
+      "learning_rate": 9.90515517768873e-05,
+      "loss": 3.4084,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2560282513932572,
+      "grad_norm": 0.5057874321937561,
+      "learning_rate": 9.904388110745403e-05,
+      "loss": 3.4503,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2567639647593298,
+      "grad_norm": 0.5415982007980347,
+      "learning_rate": 9.903617984359263e-05,
+      "loss": 3.412,
+      "step": 3490
+    },
+    {
+      "epoch": 0.25749967812540236,
+      "grad_norm": 0.5116530656814575,
+      "learning_rate": 9.902844799010729e-05,
+      "loss": 3.3947,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2582353914914749,
+      "grad_norm": 0.49354326725006104,
+      "learning_rate": 9.902068555182124e-05,
+      "loss": 3.4183,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2589711048575475,
+      "grad_norm": 0.5207719206809998,
+      "learning_rate": 9.901289253357688e-05,
+      "loss": 3.4225,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2597068182236201,
+      "grad_norm": 0.5140487551689148,
+      "learning_rate": 9.900506894023558e-05,
+      "loss": 3.4375,
+      "step": 3530
+    },
+    {
+      "epoch": 0.26044253158969266,
+      "grad_norm": 0.5068781971931458,
+      "learning_rate": 9.899721477667785e-05,
+      "loss": 3.3917,
+      "step": 3540
+    },
+    {
+      "epoch": 0.26117824495576525,
+      "grad_norm": 0.5903786420822144,
+      "learning_rate": 9.898933004780328e-05,
+      "loss": 3.4376,
+      "step": 3550
+    },
+    {
+      "epoch": 0.26191395832183784,
+      "grad_norm": 0.5345304012298584,
+      "learning_rate": 9.898141475853046e-05,
+      "loss": 3.4211,
+      "step": 3560
+    },
+    {
+      "epoch": 0.26264967168791037,
+      "grad_norm": 0.5153301954269409,
+      "learning_rate": 9.89734689137971e-05,
+      "loss": 3.4378,
+      "step": 3570
+    },
+    {
+      "epoch": 0.26338538505398296,
+      "grad_norm": 0.48121339082717896,
+      "learning_rate": 9.896549251855998e-05,
+      "loss": 3.4369,
+      "step": 3580
+    },
+    {
+      "epoch": 0.26412109842005554,
+      "grad_norm": 0.49092957377433777,
+      "learning_rate": 9.89574855777949e-05,
+      "loss": 3.3808,
+      "step": 3590
+    },
+    {
+      "epoch": 0.26485681178612813,
+      "grad_norm": 0.49230051040649414,
+      "learning_rate": 9.894944809649671e-05,
+      "loss": 3.4153,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2655925251522007,
+      "grad_norm": 0.542644739151001,
+      "learning_rate": 9.894138007967935e-05,
+      "loss": 3.4127,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2663282385182733,
+      "grad_norm": 0.49647256731987,
+      "learning_rate": 9.893328153237578e-05,
+      "loss": 3.3944,
+      "step": 3620
+    },
+    {
+      "epoch": 0.26706395188434584,
+      "grad_norm": 0.4804913103580475,
+      "learning_rate": 9.892515245963803e-05,
+      "loss": 3.4299,
+      "step": 3630
+    },
+    {
+      "epoch": 0.26779966525041843,
+      "grad_norm": 0.5026209354400635,
+      "learning_rate": 9.891699286653714e-05,
+      "loss": 3.4199,
+      "step": 3640
+    },
+    {
+      "epoch": 0.268535378616491,
+      "grad_norm": 0.5007442235946655,
+      "learning_rate": 9.890880275816322e-05,
+      "loss": 3.4058,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2692710919825636,
+      "grad_norm": 0.49342918395996094,
+      "learning_rate": 9.890058213962538e-05,
+      "loss": 3.408,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2700068053486362,
+      "grad_norm": 0.48717033863067627,
+      "learning_rate": 9.889233101605184e-05,
+      "loss": 3.3731,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2707425187147087,
+      "grad_norm": 0.5098682045936584,
+      "learning_rate": 9.888404939258973e-05,
+      "loss": 3.3768,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2714782320807813,
+      "grad_norm": 0.48432040214538574,
+      "learning_rate": 9.88757372744053e-05,
+      "loss": 3.4547,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2722139454468539,
+      "grad_norm": 0.487289160490036,
+      "learning_rate": 9.886739466668379e-05,
+      "loss": 3.4094,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2729496588129265,
+      "grad_norm": 0.5253422260284424,
+      "learning_rate": 9.885902157462948e-05,
+      "loss": 3.445,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2736853721789991,
+      "grad_norm": 0.5656958222389221,
+      "learning_rate": 9.885061800346563e-05,
+      "loss": 3.3925,
+      "step": 3720
+    },
+    {
+      "epoch": 0.27442108554507166,
+      "grad_norm": 0.5607571601867676,
+      "learning_rate": 9.884218395843452e-05,
+      "loss": 3.386,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2751567989111442,
+      "grad_norm": 0.48992305994033813,
+      "learning_rate": 9.883371944479749e-05,
+      "loss": 3.4261,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2758925122772168,
+      "grad_norm": 0.5179588794708252,
+      "learning_rate": 9.882522446783484e-05,
+      "loss": 3.3555,
+      "step": 3750
+    },
+    {
+      "epoch": 0.27662822564328937,
+      "grad_norm": 0.5124242305755615,
+      "learning_rate": 9.88166990328459e-05,
+      "loss": 3.4095,
+      "step": 3760
+    },
+    {
+      "epoch": 0.27736393900936196,
+      "grad_norm": 0.5088645815849304,
+      "learning_rate": 9.880814314514894e-05,
+      "loss": 3.4152,
+      "step": 3770
+    },
+    {
+      "epoch": 0.27809965237543455,
+      "grad_norm": 0.5194174647331238,
+      "learning_rate": 9.87995568100813e-05,
+      "loss": 3.3998,
+      "step": 3780
+    },
+    {
+      "epoch": 0.27883536574150714,
+      "grad_norm": 0.5403995513916016,
+      "learning_rate": 9.879094003299928e-05,
+      "loss": 3.3553,
+      "step": 3790
+    },
+    {
+      "epoch": 0.27957107910757967,
+      "grad_norm": 0.5247926712036133,
+      "learning_rate": 9.87822928192782e-05,
+      "loss": 3.4321,
+      "step": 3800
+    },
+    {
+      "epoch": 0.28030679247365226,
+      "grad_norm": 0.5156255960464478,
+      "learning_rate": 9.877361517431231e-05,
+      "loss": 3.3851,
+      "step": 3810
+    },
+    {
+      "epoch": 0.28104250583972484,
+      "grad_norm": 0.4857509732246399,
+      "learning_rate": 9.876490710351489e-05,
+      "loss": 3.3663,
+      "step": 3820
+    },
+    {
+      "epoch": 0.28177821920579743,
+      "grad_norm": 0.5125151872634888,
+      "learning_rate": 9.875616861231819e-05,
+      "loss": 3.3637,
+      "step": 3830
+    },
+    {
+      "epoch": 0.28251393257187,
+      "grad_norm": 0.47020360827445984,
+      "learning_rate": 9.874739970617341e-05,
+      "loss": 3.4038,
+      "step": 3840
+    },
+    {
+      "epoch": 0.28324964593794255,
+      "grad_norm": 0.48255953192710876,
+      "learning_rate": 9.87386003905508e-05,
+      "loss": 3.3459,
+      "step": 3850
+    },
+    {
+      "epoch": 0.28398535930401514,
+      "grad_norm": 0.47846198081970215,
+      "learning_rate": 9.872977067093947e-05,
+      "loss": 3.376,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2847210726700877,
+      "grad_norm": 0.5100752115249634,
+      "learning_rate": 9.872091055284756e-05,
+      "loss": 3.3609,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2854567860361603,
+      "grad_norm": 0.5000672936439514,
+      "learning_rate": 9.87120200418022e-05,
+      "loss": 3.4512,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2861924994022329,
+      "grad_norm": 0.5057271718978882,
+      "learning_rate": 9.870309914334942e-05,
+      "loss": 3.3253,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2869282127683055,
+      "grad_norm": 0.5031580328941345,
+      "learning_rate": 9.869414786305424e-05,
+      "loss": 3.3809,
+      "step": 3900
+    },
+    {
+      "epoch": 0.287663926134378,
+      "grad_norm": 0.49000057578086853,
+      "learning_rate": 9.868516620650062e-05,
+      "loss": 3.4302,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2883996395004506,
+      "grad_norm": 0.4677979052066803,
+      "learning_rate": 9.867615417929147e-05,
+      "loss": 3.3782,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2891353528665232,
+      "grad_norm": 0.4826011061668396,
+      "learning_rate": 9.866711178704869e-05,
+      "loss": 3.3509,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2898710662325958,
+      "grad_norm": 0.5133295655250549,
+      "learning_rate": 9.865803903541301e-05,
+      "loss": 3.4039,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2906067795986684,
+      "grad_norm": 0.5011733174324036,
+      "learning_rate": 9.864893593004422e-05,
+      "loss": 3.3433,
+      "step": 3950
+    },
+    {
+      "epoch": 0.29134249296474096,
+      "grad_norm": 0.49504631757736206,
+      "learning_rate": 9.863980247662099e-05,
+      "loss": 3.3795,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2920782063308135,
+      "grad_norm": 0.5216391682624817,
+      "learning_rate": 9.863063868084093e-05,
+      "loss": 3.4249,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2928139196968861,
+      "grad_norm": 0.5909170508384705,
+      "learning_rate": 9.862144454842055e-05,
+      "loss": 3.341,
+      "step": 3980
+    },
+    {
+      "epoch": 0.29354963306295867,
+      "grad_norm": 0.5585739016532898,
+      "learning_rate": 9.861222008509534e-05,
+      "loss": 3.4334,
+      "step": 3990
+    },
+    {
+      "epoch": 0.29428534642903126,
+      "grad_norm": 0.5350683927536011,
+      "learning_rate": 9.860296529661966e-05,
+      "loss": 3.4152,
+      "step": 4000
+    },
+    {
+      "epoch": 0.29502105979510385,
+      "grad_norm": 0.4655880630016327,
+      "learning_rate": 9.859368018876682e-05,
+      "loss": 3.4511,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2957567731611764,
+      "grad_norm": 0.483699232339859,
+      "learning_rate": 9.858436476732904e-05,
+      "loss": 3.3366,
+      "step": 4020
+    },
+    {
+      "epoch": 0.29649248652724897,
+      "grad_norm": 0.49257728457450867,
+      "learning_rate": 9.857501903811742e-05,
+      "loss": 3.3481,
+      "step": 4030
+    },
+    {
+      "epoch": 0.29722819989332155,
+      "grad_norm": 0.46038171648979187,
+      "learning_rate": 9.856564300696201e-05,
+      "loss": 3.4007,
+      "step": 4040
+    },
+    {
+      "epoch": 0.29796391325939414,
+      "grad_norm": 0.5034530758857727,
+      "learning_rate": 9.855623667971173e-05,
+      "loss": 3.3658,
+      "step": 4050
+    },
+    {
+      "epoch": 0.29869962662546673,
+      "grad_norm": 0.466950386762619,
+      "learning_rate": 9.854680006223441e-05,
+      "loss": 3.397,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2994353399915393,
+      "grad_norm": 0.5001063346862793,
+      "learning_rate": 9.853733316041678e-05,
+      "loss": 3.3742,
+      "step": 4070
+    },
+    {
+      "epoch": 0.30017105335761185,
+      "grad_norm": 0.48969554901123047,
+      "learning_rate": 9.852783598016446e-05,
+      "loss": 3.3504,
+      "step": 4080
+    },
+    {
+      "epoch": 0.30090676672368444,
+      "grad_norm": 0.4851774573326111,
+      "learning_rate": 9.851830852740196e-05,
+      "loss": 3.3486,
+      "step": 4090
+    },
+    {
+      "epoch": 0.301642480089757,
+      "grad_norm": 0.5046120882034302,
+      "learning_rate": 9.850875080807267e-05,
+      "loss": 3.3826,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3023781934558296,
+      "grad_norm": 0.5258597135543823,
+      "learning_rate": 9.849916282813885e-05,
+      "loss": 3.4172,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3031139068219022,
+      "grad_norm": 0.5025837421417236,
+      "learning_rate": 9.848954459358167e-05,
+      "loss": 3.3501,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3038496201879748,
+      "grad_norm": 0.47234296798706055,
+      "learning_rate": 9.847989611040111e-05,
+      "loss": 3.3812,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3045853335540473,
+      "grad_norm": 0.46150118112564087,
+      "learning_rate": 9.847021738461611e-05,
+      "loss": 3.3761,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3053210469201199,
+      "grad_norm": 0.4842502474784851,
+      "learning_rate": 9.846050842226442e-05,
+      "loss": 3.3183,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3060567602861925,
+      "grad_norm": 0.48497274518013,
+      "learning_rate": 9.845076922940264e-05,
+      "loss": 3.3959,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3067924736522651,
+      "grad_norm": 0.4537655711174011,
+      "learning_rate": 9.844099981210625e-05,
+      "loss": 3.3764,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3075281870183377,
+      "grad_norm": 0.46067747473716736,
+      "learning_rate": 9.843120017646959e-05,
+      "loss": 3.3569,
+      "step": 4180
+    },
+    {
+      "epoch": 0.30826390038441026,
+      "grad_norm": 0.4854758083820343,
+      "learning_rate": 9.842137032860586e-05,
+      "loss": 3.4087,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3089996137504828,
+      "grad_norm": 0.4684829115867615,
+      "learning_rate": 9.841151027464707e-05,
+      "loss": 3.4011,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3097353271165554,
+      "grad_norm": 0.46631598472595215,
+      "learning_rate": 9.840162002074409e-05,
+      "loss": 3.4003,
+      "step": 4210
+    },
+    {
+      "epoch": 0.31047104048262797,
+      "grad_norm": 0.4732069671154022,
+      "learning_rate": 9.839169957306664e-05,
+      "loss": 3.3787,
+      "step": 4220
+    },
+    {
+      "epoch": 0.31120675384870056,
+      "grad_norm": 0.5108915567398071,
+      "learning_rate": 9.838174893780329e-05,
+      "loss": 3.3448,
+      "step": 4230
+    },
+    {
+      "epoch": 0.31194246721477314,
+      "grad_norm": 0.48219287395477295,
+      "learning_rate": 9.837176812116137e-05,
+      "loss": 3.4146,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3126781805808457,
+      "grad_norm": 0.48349088430404663,
+      "learning_rate": 9.836175712936715e-05,
+      "loss": 3.3261,
+      "step": 4250
+    },
+    {
+      "epoch": 0.31341389394691826,
+      "grad_norm": 0.46656161546707153,
+      "learning_rate": 9.83517159686656e-05,
+      "loss": 3.3588,
+      "step": 4260
+    },
+    {
+      "epoch": 0.31414960731299085,
+      "grad_norm": 0.5095760822296143,
+      "learning_rate": 9.83416446453206e-05,
+      "loss": 3.3572,
+      "step": 4270
+    },
+    {
+      "epoch": 0.31488532067906344,
+      "grad_norm": 0.4995076358318329,
+      "learning_rate": 9.833154316561484e-05,
+      "loss": 3.3413,
+      "step": 4280
+    },
+    {
+      "epoch": 0.31562103404513603,
+      "grad_norm": 0.5134519338607788,
+      "learning_rate": 9.832141153584976e-05,
+      "loss": 3.3933,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3163567474112086,
+      "grad_norm": 0.49631309509277344,
+      "learning_rate": 9.831124976234567e-05,
+      "loss": 3.391,
+      "step": 4300
+    },
+    {
+      "epoch": 0.31709246077728115,
+      "grad_norm": 0.5069088339805603,
+      "learning_rate": 9.830105785144163e-05,
+      "loss": 3.4392,
+      "step": 4310
+    },
+    {
+      "epoch": 0.31782817414335374,
+      "grad_norm": 0.46999242901802063,
+      "learning_rate": 9.829083580949558e-05,
+      "loss": 3.3529,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3185638875094263,
+      "grad_norm": 0.5165208578109741,
+      "learning_rate": 9.828058364288419e-05,
+      "loss": 3.3442,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3192996008754989,
+      "grad_norm": 0.5024532675743103,
+      "learning_rate": 9.827030135800292e-05,
+      "loss": 3.3333,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3200353142415715,
+      "grad_norm": 0.4881153106689453,
+      "learning_rate": 9.825998896126605e-05,
+      "loss": 3.3771,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3207710276076441,
+      "grad_norm": 0.4938250780105591,
+      "learning_rate": 9.824964645910664e-05,
+      "loss": 3.3848,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3215067409737166,
+      "grad_norm": 0.512721836566925,
+      "learning_rate": 9.82392738579765e-05,
+      "loss": 3.3882,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3222424543397892,
+      "grad_norm": 0.46891894936561584,
+      "learning_rate": 9.822887116434629e-05,
+      "loss": 3.4072,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3229781677058618,
+      "grad_norm": 0.5028683543205261,
+      "learning_rate": 9.821843838470534e-05,
+      "loss": 3.3296,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3237138810719344,
+      "grad_norm": 0.5311598777770996,
+      "learning_rate": 9.820797552556184e-05,
+      "loss": 3.3729,
+      "step": 4400
+    },
+    {
+      "epoch": 0.32444959443800697,
+      "grad_norm": 0.4591037631034851,
+      "learning_rate": 9.819748259344266e-05,
+      "loss": 3.3593,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3251853078040795,
+      "grad_norm": 0.46094292402267456,
+      "learning_rate": 9.818695959489351e-05,
+      "loss": 3.3822,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3259210211701521,
+      "grad_norm": 0.5148822665214539,
+      "learning_rate": 9.81764065364788e-05,
+      "loss": 3.3965,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3266567345362247,
+      "grad_norm": 0.5029377937316895,
+      "learning_rate": 9.816582342478173e-05,
+      "loss": 3.3036,
+      "step": 4440
+    },
+    {
+      "epoch": 0.32739244790229727,
+      "grad_norm": 0.46696317195892334,
+      "learning_rate": 9.815521026640422e-05,
+      "loss": 3.3496,
+      "step": 4450
+    },
+    {
+      "epoch": 0.32812816126836986,
+      "grad_norm": 0.49031320214271545,
+      "learning_rate": 9.814456706796693e-05,
+      "loss": 3.4073,
+      "step": 4460
+    },
+    {
+      "epoch": 0.32886387463444244,
+      "grad_norm": 0.49924179911613464,
+      "learning_rate": 9.81338938361093e-05,
+      "loss": 3.3481,
+      "step": 4470
+    },
+    {
+      "epoch": 0.329599588000515,
+      "grad_norm": 0.505862295627594,
+      "learning_rate": 9.812319057748944e-05,
+      "loss": 3.3617,
+      "step": 4480
+    },
+    {
+      "epoch": 0.33033530136658756,
+      "grad_norm": 0.5051801204681396,
+      "learning_rate": 9.811245729878426e-05,
+      "loss": 3.3493,
+      "step": 4490
+    },
+    {
+      "epoch": 0.33107101473266015,
+      "grad_norm": 0.4463536739349365,
+      "learning_rate": 9.810169400668938e-05,
+      "loss": 3.3342,
+      "step": 4500
+    },
+    {
+      "epoch": 0.33180672809873274,
+      "grad_norm": 0.4612107276916504,
+      "learning_rate": 9.809090070791909e-05,
+      "loss": 3.391,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3325424414648053,
+      "grad_norm": 0.48259028792381287,
+      "learning_rate": 9.808007740920646e-05,
+      "loss": 3.3383,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3332781548308779,
+      "grad_norm": 0.4375728666782379,
+      "learning_rate": 9.806922411730323e-05,
+      "loss": 3.361,
+      "step": 4530
+    },
+    {
+      "epoch": 0.33401386819695045,
+      "grad_norm": 0.5056487917900085,
+      "learning_rate": 9.805834083897993e-05,
+      "loss": 3.3882,
+      "step": 4540
+    },
+    {
+      "epoch": 0.33474958156302304,
+      "grad_norm": 0.5022489428520203,
+      "learning_rate": 9.804742758102567e-05,
+      "loss": 3.3858,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3354852949290956,
+      "grad_norm": 0.5145370364189148,
+      "learning_rate": 9.803648435024837e-05,
+      "loss": 3.3201,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3362210082951682,
+      "grad_norm": 0.905788004398346,
+      "learning_rate": 9.80255111534746e-05,
+      "loss": 3.3431,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3369567216612408,
+      "grad_norm": 0.456967294216156,
+      "learning_rate": 9.801450799754963e-05,
+      "loss": 3.3356,
+      "step": 4580
+    },
+    {
+      "epoch": 0.33769243502731333,
+      "grad_norm": 0.47754788398742676,
+      "learning_rate": 9.800347488933743e-05,
+      "loss": 3.3502,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3384281483933859,
+      "grad_norm": 0.474417507648468,
+      "learning_rate": 9.799241183572062e-05,
+      "loss": 3.38,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3391638617594585,
+      "grad_norm": 0.453109472990036,
+      "learning_rate": 9.798131884360054e-05,
+      "loss": 3.3487,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3398995751255311,
+      "grad_norm": 0.4781600832939148,
+      "learning_rate": 9.797019591989721e-05,
+      "loss": 3.3442,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3406352884916037,
+      "grad_norm": 0.4538155794143677,
+      "learning_rate": 9.795904307154927e-05,
+      "loss": 3.3541,
+      "step": 4630
+    },
+    {
+      "epoch": 0.34137100185767627,
+      "grad_norm": 0.45867714285850525,
+      "learning_rate": 9.794786030551409e-05,
+      "loss": 3.3293,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3421067152237488,
+      "grad_norm": 0.455285906791687,
+      "learning_rate": 9.793664762876766e-05,
+      "loss": 3.3302,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3428424285898214,
+      "grad_norm": 0.4696787893772125,
+      "learning_rate": 9.792540504830465e-05,
+      "loss": 3.337,
+      "step": 4660
+    },
+    {
+      "epoch": 0.343578141955894,
+      "grad_norm": 0.4765075147151947,
+      "learning_rate": 9.791413257113838e-05,
+      "loss": 3.3768,
+      "step": 4670
+    },
+    {
+      "epoch": 0.34431385532196657,
+      "grad_norm": 0.4819774031639099,
+      "learning_rate": 9.790283020430079e-05,
+      "loss": 3.3431,
+      "step": 4680
+    },
+    {
+      "epoch": 0.34504956868803915,
+      "grad_norm": 0.47870171070098877,
+      "learning_rate": 9.789149795484252e-05,
+      "loss": 3.34,
+      "step": 4690
+    },
+    {
+      "epoch": 0.34578528205411174,
+      "grad_norm": 0.4422255754470825,
+      "learning_rate": 9.788013582983285e-05,
+      "loss": 3.317,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3465209954201843,
+      "grad_norm": 0.4485225975513458,
+      "learning_rate": 9.78687438363596e-05,
+      "loss": 3.3314,
+      "step": 4710
+    },
+    {
+      "epoch": 0.34725670878625686,
+      "grad_norm": 0.4411897659301758,
+      "learning_rate": 9.785732198152936e-05,
+      "loss": 3.3204,
+      "step": 4720
+    },
+    {
+      "epoch": 0.34799242215232945,
+      "grad_norm": 0.49034199118614197,
+      "learning_rate": 9.784587027246726e-05,
+      "loss": 3.3287,
+      "step": 4730
+    },
+    {
+      "epoch": 0.34872813551840204,
+      "grad_norm": 0.4629896581172943,
+      "learning_rate": 9.783438871631707e-05,
+      "loss": 3.3663,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3494638488844746,
+      "grad_norm": 0.4768849313259125,
+      "learning_rate": 9.782287732024118e-05,
+      "loss": 3.2822,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3501995622505472,
+      "grad_norm": 0.45183610916137695,
+      "learning_rate": 9.781133609142062e-05,
+      "loss": 3.3665,
+      "step": 4760
+    },
+    {
+      "epoch": 0.35093527561661975,
+      "grad_norm": 0.46104103326797485,
+      "learning_rate": 9.779976503705499e-05,
+      "loss": 3.3192,
+      "step": 4770
+    },
+    {
+      "epoch": 0.35167098898269233,
+      "grad_norm": 0.47722548246383667,
+      "learning_rate": 9.778816416436253e-05,
+      "loss": 3.3519,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3524067023487649,
+      "grad_norm": 0.45925357937812805,
+      "learning_rate": 9.777653348058004e-05,
+      "loss": 3.3195,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3531424157148375,
+      "grad_norm": 0.46248316764831543,
+      "learning_rate": 9.7764872992963e-05,
+      "loss": 3.3221,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3538781290809101,
+      "grad_norm": 0.45539459586143494,
+      "learning_rate": 9.775318270878537e-05,
+      "loss": 3.3492,
+      "step": 4810
+    },
+    {
+      "epoch": 0.35461384244698263,
+      "grad_norm": 0.47933024168014526,
+      "learning_rate": 9.774146263533976e-05,
+      "loss": 3.3284,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3553495558130552,
+      "grad_norm": 0.5591832399368286,
+      "learning_rate": 9.77297127799374e-05,
+      "loss": 3.3435,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3560852691791278,
+      "grad_norm": 0.48164159059524536,
+      "learning_rate": 9.7717933149908e-05,
+      "loss": 3.3433,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3568209825452004,
+      "grad_norm": 0.48207688331604004,
+      "learning_rate": 9.770612375259991e-05,
+      "loss": 3.3348,
+      "step": 4850
+    },
+    {
+      "epoch": 0.357556695911273,
+      "grad_norm": 0.49394533038139343,
+      "learning_rate": 9.769428459538008e-05,
+      "loss": 3.3,
+      "step": 4860
+    },
+    {
+      "epoch": 0.35829240927734557,
+      "grad_norm": 0.45821693539619446,
+      "learning_rate": 9.768241568563392e-05,
+      "loss": 3.3498,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3590281226434181,
+      "grad_norm": 0.4812382459640503,
+      "learning_rate": 9.767051703076552e-05,
+      "loss": 3.2435,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3597638360094907,
+      "grad_norm": 0.44694334268569946,
+      "learning_rate": 9.765858863819744e-05,
+      "loss": 3.3879,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3604995493755633,
+      "grad_norm": 0.4820118546485901,
+      "learning_rate": 9.764663051537082e-05,
+      "loss": 3.3189,
+      "step": 4900
+    },
+    {
+      "epoch": 0.36123526274163587,
+      "grad_norm": 0.464145302772522,
+      "learning_rate": 9.763464266974534e-05,
+      "loss": 3.3726,
+      "step": 4910
+    },
+    {
+      "epoch": 0.36197097610770845,
+      "grad_norm": 0.4518837630748749,
+      "learning_rate": 9.762262510879925e-05,
+      "loss": 3.3453,
+      "step": 4920
+    },
+    {
+      "epoch": 0.36270668947378104,
+      "grad_norm": 0.4643986225128174,
+      "learning_rate": 9.761057784002929e-05,
+      "loss": 3.3463,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3634424028398536,
+      "grad_norm": 0.4880530834197998,
+      "learning_rate": 9.759850087095076e-05,
+      "loss": 3.3116,
+      "step": 4940
+    },
+    {
+      "epoch": 0.36417811620592616,
+      "grad_norm": 0.4663744270801544,
+      "learning_rate": 9.75863942090975e-05,
+      "loss": 3.2909,
+      "step": 4950
+    },
+    {
+      "epoch": 0.36491382957199875,
+      "grad_norm": 0.5150174498558044,
+      "learning_rate": 9.757425786202184e-05,
+      "loss": 3.2635,
+      "step": 4960
+    },
+    {
+      "epoch": 0.36564954293807134,
+      "grad_norm": 0.5246181488037109,
+      "learning_rate": 9.756209183729466e-05,
+      "loss": 3.2857,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3663852563041439,
+      "grad_norm": 0.48053282499313354,
+      "learning_rate": 9.754989614250532e-05,
+      "loss": 3.3379,
+      "step": 4980
+    },
+    {
+      "epoch": 0.36712096967021646,
+      "grad_norm": 0.44896063208580017,
+      "learning_rate": 9.753767078526173e-05,
+      "loss": 3.2669,
+      "step": 4990
+    },
+    {
+      "epoch": 0.36785668303628904,
+      "grad_norm": 0.47481274604797363,
+      "learning_rate": 9.752541577319025e-05,
+      "loss": 3.3064,
+      "step": 5000
+    },
+    {
+      "epoch": 0.36859239640236163,
+      "grad_norm": 0.4941543638706207,
+      "learning_rate": 9.75131311139358e-05,
+      "loss": 3.3203,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3693281097684342,
+      "grad_norm": 0.41711509227752686,
+      "learning_rate": 9.750081681516174e-05,
+      "loss": 3.2739,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3700638231345068,
+      "grad_norm": 0.48057687282562256,
+      "learning_rate": 9.748847288454996e-05,
+      "loss": 3.3552,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3707995365005794,
+      "grad_norm": 0.454571396112442,
+      "learning_rate": 9.747609932980083e-05,
+      "loss": 3.3066,
+      "step": 5040
+    },
+    {
+      "epoch": 0.37153524986665193,
+      "grad_norm": 0.4346379041671753,
+      "learning_rate": 9.746369615863317e-05,
+      "loss": 3.3061,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3722709632327245,
+      "grad_norm": 0.42454394698143005,
+      "learning_rate": 9.74512633787843e-05,
+      "loss": 3.2971,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3730066765987971,
+      "grad_norm": 0.4378200173377991,
+      "learning_rate": 9.743880099801e-05,
+      "loss": 3.3169,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3737423899648697,
+      "grad_norm": 0.47664156556129456,
+      "learning_rate": 9.742630902408452e-05,
+      "loss": 3.3459,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3744781033309423,
+      "grad_norm": 0.48526403307914734,
+      "learning_rate": 9.741378746480057e-05,
+      "loss": 3.3237,
+      "step": 5090
+    },
+    {
+      "epoch": 0.37521381669701487,
+      "grad_norm": 0.48349133133888245,
+      "learning_rate": 9.740123632796933e-05,
+      "loss": 3.3074,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3759495300630874,
+      "grad_norm": 0.46515288949012756,
+      "learning_rate": 9.738865562142043e-05,
+      "loss": 3.286,
+      "step": 5110
+    },
+    {
+      "epoch": 0.37668524342916,
+      "grad_norm": 0.45699048042297363,
+      "learning_rate": 9.737604535300188e-05,
+      "loss": 3.3676,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3774209567952326,
+      "grad_norm": 0.48287123441696167,
+      "learning_rate": 9.736340553058025e-05,
+      "loss": 3.2953,
+      "step": 5130
+    },
+    {
+      "epoch": 0.37815667016130516,
+      "grad_norm": 0.4803648293018341,
+      "learning_rate": 9.735073616204045e-05,
+      "loss": 3.3047,
+      "step": 5140
+    },
+    {
+      "epoch": 0.37889238352737775,
+      "grad_norm": 0.4733999967575073,
+      "learning_rate": 9.733803725528585e-05,
+      "loss": 3.3511,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3796280968934503,
+      "grad_norm": 0.48089954257011414,
+      "learning_rate": 9.732530881823824e-05,
+      "loss": 3.3259,
+      "step": 5160
+    },
+    {
+      "epoch": 0.38036381025952287,
+      "grad_norm": 0.47004076838493347,
+      "learning_rate": 9.731255085883787e-05,
+      "loss": 3.3201,
+      "step": 5170
+    },
+    {
+      "epoch": 0.38109952362559546,
+      "grad_norm": 0.8502194881439209,
+      "learning_rate": 9.729976338504336e-05,
+      "loss": 3.3501,
+      "step": 5180
+    },
+    {
+      "epoch": 0.38183523699166805,
+      "grad_norm": 0.48328205943107605,
+      "learning_rate": 9.728694640483176e-05,
+      "loss": 3.2916,
+      "step": 5190
+    },
+    {
+      "epoch": 0.38257095035774064,
+      "grad_norm": 0.43932202458381653,
+      "learning_rate": 9.727409992619853e-05,
+      "loss": 3.3287,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3833066637238132,
+      "grad_norm": 0.4448724091053009,
+      "learning_rate": 9.726122395715751e-05,
+      "loss": 3.3104,
+      "step": 5210
+    },
+    {
+      "epoch": 0.38404237708988576,
+      "grad_norm": 0.454637736082077,
+      "learning_rate": 9.724831850574099e-05,
+      "loss": 3.2872,
+      "step": 5220
+    },
+    {
+      "epoch": 0.38477809045595834,
+      "grad_norm": 0.5293341875076294,
+      "learning_rate": 9.723538357999958e-05,
+      "loss": 3.369,
+      "step": 5230
+    },
+    {
+      "epoch": 0.38551380382203093,
+      "grad_norm": 0.46058687567710876,
+      "learning_rate": 9.722241918800234e-05,
+      "loss": 3.3044,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3862495171881035,
+      "grad_norm": 0.45656171441078186,
+      "learning_rate": 9.720942533783666e-05,
+      "loss": 3.329,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3869852305541761,
+      "grad_norm": 0.4675185978412628,
+      "learning_rate": 9.719640203760835e-05,
+      "loss": 3.2812,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3877209439202487,
+      "grad_norm": 0.47009697556495667,
+      "learning_rate": 9.718334929544155e-05,
+      "loss": 3.3665,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3884566572863212,
+      "grad_norm": 0.4447770118713379,
+      "learning_rate": 9.71702671194788e-05,
+      "loss": 3.3148,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3891923706523938,
+      "grad_norm": 0.4634477198123932,
+      "learning_rate": 9.715715551788101e-05,
+      "loss": 3.3076,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3899280840184664,
+      "grad_norm": 0.4768572151660919,
+      "learning_rate": 9.71440144988274e-05,
+      "loss": 3.283,
+      "step": 5300
+    },
+    {
+      "epoch": 0.390663797384539,
+      "grad_norm": 0.46009397506713867,
+      "learning_rate": 9.713084407051556e-05,
+      "loss": 3.3147,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3913995107506116,
+      "grad_norm": 0.48648467659950256,
+      "learning_rate": 9.711764424116145e-05,
+      "loss": 3.2939,
+      "step": 5320
+    },
+    {
+      "epoch": 0.39213522411668417,
+      "grad_norm": 0.49335330724716187,
+      "learning_rate": 9.710441501899936e-05,
+      "loss": 3.3232,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3928709374827567,
+      "grad_norm": 0.4541822373867035,
+      "learning_rate": 9.709115641228189e-05,
+      "loss": 3.3027,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3936066508488293,
+      "grad_norm": 0.4860965311527252,
+      "learning_rate": 9.707786842927999e-05,
+      "loss": 3.3016,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3943423642149019,
+      "grad_norm": 0.46567508578300476,
+      "learning_rate": 9.706455107828296e-05,
+      "loss": 3.3336,
+      "step": 5360
+    },
+    {
+      "epoch": 0.39507807758097446,
+      "grad_norm": 0.5198284387588501,
+      "learning_rate": 9.705120436759837e-05,
+      "loss": 3.3211,
+      "step": 5370
+    },
+    {
+      "epoch": 0.39581379094704705,
+      "grad_norm": 0.44841471314430237,
+      "learning_rate": 9.703782830555214e-05,
+      "loss": 3.2929,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3965495043131196,
+      "grad_norm": 0.46957024931907654,
+      "learning_rate": 9.70244229004885e-05,
+      "loss": 3.32,
+      "step": 5390
+    },
+    {
+      "epoch": 0.39728521767919217,
+      "grad_norm": 0.43909287452697754,
+      "learning_rate": 9.701098816076996e-05,
+      "loss": 3.2571,
+      "step": 5400
+    },
+    {
+      "epoch": 0.39802093104526476,
+      "grad_norm": 0.4354236125946045,
+      "learning_rate": 9.699752409477735e-05,
+      "loss": 3.2885,
+      "step": 5410
+    },
+    {
+      "epoch": 0.39875664441133735,
+      "grad_norm": 0.4583248496055603,
+      "learning_rate": 9.69840307109098e-05,
+      "loss": 3.2627,
+      "step": 5420
+    },
+    {
+      "epoch": 0.39949235777740993,
+      "grad_norm": 0.4307055175304413,
+      "learning_rate": 9.697050801758469e-05,
+      "loss": 3.2989,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4002280711434825,
+      "grad_norm": 0.47710663080215454,
+      "learning_rate": 9.695695602323776e-05,
+      "loss": 3.2879,
+      "step": 5440
+    },
+    {
+      "epoch": 0.40096378450955505,
+      "grad_norm": 0.4449078142642975,
+      "learning_rate": 9.694337473632293e-05,
+      "loss": 3.2946,
+      "step": 5450
+    },
+    {
+      "epoch": 0.40169949787562764,
+      "grad_norm": 0.4442616403102875,
+      "learning_rate": 9.692976416531247e-05,
+      "loss": 3.3026,
+      "step": 5460
+    },
+    {
+      "epoch": 0.40243521124170023,
+      "grad_norm": 0.4815976321697235,
+      "learning_rate": 9.691612431869688e-05,
+      "loss": 3.3186,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4031709246077728,
+      "grad_norm": 0.4745250344276428,
+      "learning_rate": 9.690245520498495e-05,
+      "loss": 3.2948,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4039066379738454,
+      "grad_norm": 0.4966549277305603,
+      "learning_rate": 9.68887568327037e-05,
+      "loss": 3.2993,
+      "step": 5490
+    },
+    {
+      "epoch": 0.404642351339918,
+      "grad_norm": 0.43987634778022766,
+      "learning_rate": 9.687502921039842e-05,
+      "loss": 3.3187,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4053780647059905,
+      "grad_norm": 0.4542040228843689,
+      "learning_rate": 9.686127234663261e-05,
+      "loss": 3.2944,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4061137780720631,
+      "grad_norm": 0.44791513681411743,
+      "learning_rate": 9.68474862499881e-05,
+      "loss": 3.3427,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4068494914381357,
+      "grad_norm": 0.44339093565940857,
+      "learning_rate": 9.683367092906484e-05,
+      "loss": 3.3479,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4075852048042083,
+      "grad_norm": 0.41371455788612366,
+      "learning_rate": 9.681982639248111e-05,
+      "loss": 3.2453,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4083209181702809,
+      "grad_norm": 0.4320749342441559,
+      "learning_rate": 9.680595264887334e-05,
+      "loss": 3.335,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4090566315363534,
+      "grad_norm": 0.45809799432754517,
+      "learning_rate": 9.679204970689624e-05,
+      "loss": 3.3114,
+      "step": 5560
+    },
+    {
+      "epoch": 0.409792344902426,
+      "grad_norm": 0.43507814407348633,
+      "learning_rate": 9.67781175752227e-05,
+      "loss": 3.2928,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4105280582684986,
+      "grad_norm": 0.474032998085022,
+      "learning_rate": 9.676415626254381e-05,
+      "loss": 3.3059,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4112637716345712,
+      "grad_norm": 0.4753536581993103,
+      "learning_rate": 9.67501657775689e-05,
+      "loss": 3.2696,
+      "step": 5590
+    },
+    {
+      "epoch": 0.41199948500064376,
+      "grad_norm": 0.4632064700126648,
+      "learning_rate": 9.673614612902547e-05,
+      "loss": 3.2998,
+      "step": 5600
+    },
+    {
+      "epoch": 0.41273519836671635,
+      "grad_norm": 0.46428951621055603,
+      "learning_rate": 9.672209732565922e-05,
+      "loss": 3.2535,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4134709117327889,
+      "grad_norm": 0.44999969005584717,
+      "learning_rate": 9.670801937623405e-05,
+      "loss": 3.2638,
+      "step": 5620
+    },
+    {
+      "epoch": 0.41420662509886147,
+      "grad_norm": 0.4350806176662445,
+      "learning_rate": 9.669391228953202e-05,
+      "loss": 3.3014,
+      "step": 5630
+    },
+    {
+      "epoch": 0.41494233846493406,
+      "grad_norm": 0.45323678851127625,
+      "learning_rate": 9.667977607435337e-05,
+      "loss": 3.2623,
+      "step": 5640
+    },
+    {
+      "epoch": 0.41567805183100665,
+      "grad_norm": 0.47555407881736755,
+      "learning_rate": 9.666561073951655e-05,
+      "loss": 3.3212,
+      "step": 5650
+    },
+    {
+      "epoch": 0.41641376519707923,
+      "grad_norm": 0.46264752745628357,
+      "learning_rate": 9.665141629385809e-05,
+      "loss": 3.3034,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4171494785631518,
+      "grad_norm": 0.45676395297050476,
+      "learning_rate": 9.663719274623277e-05,
+      "loss": 3.323,
+      "step": 5670
+    },
+    {
+      "epoch": 0.41788519192922435,
+      "grad_norm": 0.4800090789794922,
+      "learning_rate": 9.662294010551348e-05,
+      "loss": 3.2806,
+      "step": 5680
+    },
+    {
+      "epoch": 0.41862090529529694,
+      "grad_norm": 0.4712800085544586,
+      "learning_rate": 9.660865838059128e-05,
+      "loss": 3.3134,
+      "step": 5690
+    },
+    {
+      "epoch": 0.41935661866136953,
+      "grad_norm": 0.4544120132923126,
+      "learning_rate": 9.659434758037532e-05,
+      "loss": 3.2902,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4200923320274421,
+      "grad_norm": 0.4356352388858795,
+      "learning_rate": 9.658000771379297e-05,
+      "loss": 3.3005,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4208280453935147,
+      "grad_norm": 0.46152356266975403,
+      "learning_rate": 9.656563878978965e-05,
+      "loss": 3.3022,
+      "step": 5720
+    },
+    {
+      "epoch": 0.42156375875958724,
+      "grad_norm": 0.44487205147743225,
+      "learning_rate": 9.655124081732897e-05,
+      "loss": 3.253,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4222994721256598,
+      "grad_norm": 0.4438563287258148,
+      "learning_rate": 9.653681380539263e-05,
+      "loss": 3.2019,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4230351854917324,
+      "grad_norm": 0.44498592615127563,
+      "learning_rate": 9.652235776298046e-05,
+      "loss": 3.3301,
+      "step": 5750
+    },
+    {
+      "epoch": 0.423770898857805,
+      "grad_norm": 0.4550943672657013,
+      "learning_rate": 9.650787269911037e-05,
+      "loss": 3.2636,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4245066122238776,
+      "grad_norm": 0.44825536012649536,
+      "learning_rate": 9.649335862281843e-05,
+      "loss": 3.2851,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4252423255899502,
+      "grad_norm": 0.43271371722221375,
+      "learning_rate": 9.647881554315873e-05,
+      "loss": 3.2943,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4259780389560227,
+      "grad_norm": 0.46533486247062683,
+      "learning_rate": 9.646424346920353e-05,
+      "loss": 3.276,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4267137523220953,
+      "grad_norm": 0.42707332968711853,
+      "learning_rate": 9.644964241004313e-05,
+      "loss": 3.3076,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4274494656881679,
+      "grad_norm": 0.43310198187828064,
+      "learning_rate": 9.643501237478592e-05,
+      "loss": 3.3004,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4281851790542405,
+      "grad_norm": 0.4489288926124573,
+      "learning_rate": 9.642035337255839e-05,
+      "loss": 3.3036,
+      "step": 5820
+    },
+    {
+      "epoch": 0.42892089242031306,
+      "grad_norm": 0.43780744075775146,
+      "learning_rate": 9.640566541250508e-05,
+      "loss": 3.2972,
+      "step": 5830
+    },
+    {
+      "epoch": 0.42965660578638565,
+      "grad_norm": 0.4519082307815552,
+      "learning_rate": 9.639094850378858e-05,
+      "loss": 3.3251,
+      "step": 5840
+    },
+    {
+      "epoch": 0.4303923191524582,
+      "grad_norm": 0.46353888511657715,
+      "learning_rate": 9.637620265558955e-05,
+      "loss": 3.2342,
+      "step": 5850
+    },
+    {
+      "epoch": 0.43112803251853077,
+      "grad_norm": 0.45390161871910095,
+      "learning_rate": 9.636142787710672e-05,
+      "loss": 3.2985,
+      "step": 5860
+    },
+    {
+      "epoch": 0.43186374588460336,
+      "grad_norm": 0.43484824895858765,
+      "learning_rate": 9.634662417755685e-05,
+      "loss": 3.2721,
+      "step": 5870
+    },
+    {
+      "epoch": 0.43259945925067594,
+      "grad_norm": 0.48000025749206543,
+      "learning_rate": 9.633179156617476e-05,
+      "loss": 3.2685,
+      "step": 5880
+    },
+    {
+      "epoch": 0.43333517261674853,
+      "grad_norm": 0.4266999065876007,
+      "learning_rate": 9.631693005221327e-05,
+      "loss": 3.2389,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4340708859828211,
+      "grad_norm": 0.4476128816604614,
+      "learning_rate": 9.630203964494327e-05,
+      "loss": 3.2802,
+      "step": 5900
+    },
+    {
+      "epoch": 0.43480659934889365,
+      "grad_norm": 0.436440646648407,
+      "learning_rate": 9.628712035365363e-05,
+      "loss": 3.2765,
+      "step": 5910
+    },
+    {
+      "epoch": 0.43554231271496624,
+      "grad_norm": 0.4390304386615753,
+      "learning_rate": 9.62721721876513e-05,
+      "loss": 3.2969,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4362780260810388,
+      "grad_norm": 0.45296919345855713,
+      "learning_rate": 9.625719515626117e-05,
+      "loss": 3.2646,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4370137394471114,
+      "grad_norm": 0.4439026117324829,
+      "learning_rate": 9.62421892688262e-05,
+      "loss": 3.3088,
+      "step": 5940
+    },
+    {
+      "epoch": 0.437749452813184,
+      "grad_norm": 0.43667683005332947,
+      "learning_rate": 9.622715453470732e-05,
+      "loss": 3.285,
+      "step": 5950
+    },
+    {
+      "epoch": 0.43848516617925654,
+      "grad_norm": 0.4610852599143982,
+      "learning_rate": 9.621209096328343e-05,
+      "loss": 3.2502,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4392208795453291,
+      "grad_norm": 0.43769463896751404,
+      "learning_rate": 9.61969985639515e-05,
+      "loss": 3.3171,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4399565929114017,
+      "grad_norm": 0.4450635313987732,
+      "learning_rate": 9.61818773461264e-05,
+      "loss": 3.2985,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4406923062774743,
+      "grad_norm": 0.423850417137146,
+      "learning_rate": 9.616672731924101e-05,
+      "loss": 3.3041,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4414280196435469,
+      "grad_norm": 0.4177245497703552,
+      "learning_rate": 9.615154849274618e-05,
+      "loss": 3.2825,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4421637330096195,
+      "grad_norm": 0.44956111907958984,
+      "learning_rate": 9.613634087611073e-05,
+      "loss": 3.2986,
+      "step": 6010
+    },
+    {
+      "epoch": 0.442899446375692,
+      "grad_norm": 0.4504786431789398,
+      "learning_rate": 9.612110447882144e-05,
+      "loss": 3.3064,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4436351597417646,
+      "grad_norm": 0.4433842897415161,
+      "learning_rate": 9.610583931038304e-05,
+      "loss": 3.2896,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4443708731078372,
+      "grad_norm": 0.4560666084289551,
+      "learning_rate": 9.609054538031821e-05,
+      "loss": 3.2007,
+      "step": 6040
+    },
+    {
+      "epoch": 0.44510658647390977,
+      "grad_norm": 0.4344310760498047,
+      "learning_rate": 9.607522269816756e-05,
+      "loss": 3.2874,
+      "step": 6050
+    },
+    {
+      "epoch": 0.44584229983998236,
+      "grad_norm": 0.4632498025894165,
+      "learning_rate": 9.605987127348966e-05,
+      "loss": 3.2885,
+      "step": 6060
+    },
+    {
+      "epoch": 0.44657801320605495,
+      "grad_norm": 0.4609225392341614,
+      "learning_rate": 9.604449111586098e-05,
+      "loss": 3.2744,
+      "step": 6070
+    },
+    {
+      "epoch": 0.4473137265721275,
+      "grad_norm": 0.43838661909103394,
+      "learning_rate": 9.602908223487594e-05,
+      "loss": 3.2586,
+      "step": 6080
+    },
+    {
+      "epoch": 0.44804943993820007,
+      "grad_norm": 0.4453481435775757,
+      "learning_rate": 9.601364464014688e-05,
+      "loss": 3.2933,
+      "step": 6090
+    },
+    {
+      "epoch": 0.44878515330427265,
+      "grad_norm": 0.4828120172023773,
+      "learning_rate": 9.599817834130402e-05,
+      "loss": 3.2679,
+      "step": 6100
+    },
+    {
+      "epoch": 0.44952086667034524,
+      "grad_norm": 0.5243907570838928,
+      "learning_rate": 9.598268334799552e-05,
+      "loss": 3.2942,
+      "step": 6110
+    },
+    {
+      "epoch": 0.45025658003641783,
+      "grad_norm": 0.45438796281814575,
+      "learning_rate": 9.59671596698874e-05,
+      "loss": 3.2831,
+      "step": 6120
+    },
+    {
+      "epoch": 0.45099229340249036,
+      "grad_norm": 0.4482550323009491,
+      "learning_rate": 9.595160731666366e-05,
+      "loss": 3.2718,
+      "step": 6130
+    },
+    {
+      "epoch": 0.45172800676856295,
+      "grad_norm": 0.43046483397483826,
+      "learning_rate": 9.593602629802608e-05,
+      "loss": 3.2916,
+      "step": 6140
+    },
+    {
+      "epoch": 0.45246372013463554,
+      "grad_norm": 0.499222993850708,
+      "learning_rate": 9.592041662369438e-05,
+      "loss": 3.2908,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4531994335007081,
+      "grad_norm": 0.429533988237381,
+      "learning_rate": 9.590477830340613e-05,
+      "loss": 3.2711,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4539351468667807,
+      "grad_norm": 0.47045251727104187,
+      "learning_rate": 9.588911134691684e-05,
+      "loss": 3.293,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4546708602328533,
+      "grad_norm": 0.42130205035209656,
+      "learning_rate": 9.587341576399979e-05,
+      "loss": 3.26,
+      "step": 6180
+    },
+    {
+      "epoch": 0.45540657359892583,
+      "grad_norm": 0.4418286681175232,
+      "learning_rate": 9.585769156444614e-05,
+      "loss": 3.2765,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4561422869649984,
+      "grad_norm": 0.43507617712020874,
+      "learning_rate": 9.584193875806498e-05,
+      "loss": 3.3047,
+      "step": 6200
+    },
+    {
+      "epoch": 0.456878000331071,
+      "grad_norm": 0.42862462997436523,
+      "learning_rate": 9.582615735468314e-05,
+      "loss": 3.2935,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4576137136971436,
+      "grad_norm": 0.44380027055740356,
+      "learning_rate": 9.581034736414536e-05,
+      "loss": 3.2585,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4583494270632162,
+      "grad_norm": 0.42932000756263733,
+      "learning_rate": 9.579450879631416e-05,
+      "loss": 3.2444,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4590851404292888,
+      "grad_norm": 0.4627602696418762,
+      "learning_rate": 9.577864166106993e-05,
+      "loss": 3.2885,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4598208537953613,
+      "grad_norm": 0.4066958427429199,
+      "learning_rate": 9.576274596831087e-05,
+      "loss": 3.3016,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4605565671614339,
+      "grad_norm": 0.4360782504081726,
+      "learning_rate": 9.574682172795299e-05,
+      "loss": 3.2607,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4612922805275065,
+      "grad_norm": 0.44138815999031067,
+      "learning_rate": 9.573086894993008e-05,
+      "loss": 3.288,
+      "step": 6270
+    },
+    {
+      "epoch": 0.46202799389357907,
+      "grad_norm": 0.4372709095478058,
+      "learning_rate": 9.571488764419381e-05,
+      "loss": 3.2366,
+      "step": 6280
+    },
+    {
+      "epoch": 0.46276370725965166,
+      "grad_norm": 0.4353005886077881,
+      "learning_rate": 9.569887782071356e-05,
+      "loss": 3.3092,
+      "step": 6290
+    },
+    {
+      "epoch": 0.46349942062572425,
+      "grad_norm": 0.4514767527580261,
+      "learning_rate": 9.568283948947655e-05,
+      "loss": 3.3014,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4642351339917968,
+      "grad_norm": 0.4290522336959839,
+      "learning_rate": 9.566677266048778e-05,
+      "loss": 3.2763,
+      "step": 6310
+    },
+    {
+      "epoch": 0.46497084735786937,
+      "grad_norm": 0.43439796566963196,
+      "learning_rate": 9.565067734376999e-05,
+      "loss": 3.2469,
+      "step": 6320
+    },
+    {
+      "epoch": 0.46570656072394195,
+      "grad_norm": 0.47817927598953247,
+      "learning_rate": 9.563455354936375e-05,
+      "loss": 3.2636,
+      "step": 6330
+    },
+    {
+      "epoch": 0.46644227409001454,
+      "grad_norm": 0.44186869263648987,
+      "learning_rate": 9.561840128732735e-05,
+      "loss": 3.2496,
+      "step": 6340
+    },
+    {
+      "epoch": 0.46717798745608713,
+      "grad_norm": 0.4209322929382324,
+      "learning_rate": 9.560222056773684e-05,
+      "loss": 3.2629,
+      "step": 6350
+    },
+    {
+      "epoch": 0.46791370082215966,
+      "grad_norm": 0.42116010189056396,
+      "learning_rate": 9.558601140068606e-05,
+      "loss": 3.272,
+      "step": 6360
+    },
+    {
+      "epoch": 0.46864941418823225,
+      "grad_norm": 0.4619746804237366,
+      "learning_rate": 9.556977379628656e-05,
+      "loss": 3.2602,
+      "step": 6370
+    },
+    {
+      "epoch": 0.46938512755430484,
+      "grad_norm": 0.44723060727119446,
+      "learning_rate": 9.555350776466763e-05,
+      "loss": 3.2318,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4701208409203774,
+      "grad_norm": 0.4265129864215851,
+      "learning_rate": 9.553721331597632e-05,
+      "loss": 3.2932,
+      "step": 6390
+    },
+    {
+      "epoch": 0.47085655428645,
+      "grad_norm": 0.43263575434684753,
+      "learning_rate": 9.55208904603774e-05,
+      "loss": 3.2246,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4715922676525226,
+      "grad_norm": 0.4542834460735321,
+      "learning_rate": 9.550453920805333e-05,
+      "loss": 3.2558,
+      "step": 6410
+    },
+    {
+      "epoch": 0.47232798101859513,
+      "grad_norm": 0.4443358778953552,
+      "learning_rate": 9.548815956920429e-05,
+      "loss": 3.2722,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4730636943846677,
+      "grad_norm": 0.5469712018966675,
+      "learning_rate": 9.547175155404824e-05,
+      "loss": 3.2956,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4737994077507403,
+      "grad_norm": 0.45458346605300903,
+      "learning_rate": 9.545531517282073e-05,
+      "loss": 3.3005,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4745351211168129,
+      "grad_norm": 0.4310155212879181,
+      "learning_rate": 9.543885043577511e-05,
+      "loss": 3.301,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4752708344828855,
+      "grad_norm": 0.44873273372650146,
+      "learning_rate": 9.542235735318232e-05,
+      "loss": 3.244,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4760065478489581,
+      "grad_norm": 0.42980292439460754,
+      "learning_rate": 9.540583593533107e-05,
+      "loss": 3.2755,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4767422612150306,
+      "grad_norm": 0.41110992431640625,
+      "learning_rate": 9.538928619252772e-05,
+      "loss": 3.2582,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4774779745811032,
+      "grad_norm": 0.4515661597251892,
+      "learning_rate": 9.537270813509622e-05,
+      "loss": 3.2538,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4782136879471758,
+      "grad_norm": 0.4295632541179657,
+      "learning_rate": 9.535610177337833e-05,
+      "loss": 3.1912,
+      "step": 6500
+    },
+    {
+      "epoch": 0.47894940131324837,
+      "grad_norm": 0.43053045868873596,
+      "learning_rate": 9.533946711773335e-05,
+      "loss": 3.2977,
+      "step": 6510
+    },
+    {
+      "epoch": 0.47968511467932096,
+      "grad_norm": 0.4485345482826233,
+      "learning_rate": 9.532280417853825e-05,
+      "loss": 3.3098,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4804208280453935,
+      "grad_norm": 0.45671671628952026,
+      "learning_rate": 9.53061129661877e-05,
+      "loss": 3.1899,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4811565414114661,
+      "grad_norm": 0.45602527260780334,
+      "learning_rate": 9.528939349109393e-05,
+      "loss": 3.2406,
+      "step": 6540
+    },
+    {
+      "epoch": 0.48189225477753866,
+      "grad_norm": 0.43597233295440674,
+      "learning_rate": 9.52726457636869e-05,
+      "loss": 3.2476,
+      "step": 6550
+    },
+    {
+      "epoch": 0.48262796814361125,
+      "grad_norm": 0.4405997097492218,
+      "learning_rate": 9.525586979441407e-05,
+      "loss": 3.2843,
+      "step": 6560
+    },
+    {
+      "epoch": 0.48336368150968384,
+      "grad_norm": 0.4386025071144104,
+      "learning_rate": 9.523906559374063e-05,
+      "loss": 3.2067,
+      "step": 6570
+    },
+    {
+      "epoch": 0.48409939487575643,
+      "grad_norm": 0.45547547936439514,
+      "learning_rate": 9.52222331721493e-05,
+      "loss": 3.3049,
+      "step": 6580
+    },
+    {
+      "epoch": 0.48483510824182896,
+      "grad_norm": 0.4486294090747833,
+      "learning_rate": 9.520537254014045e-05,
+      "loss": 3.2632,
+      "step": 6590
+    },
+    {
+      "epoch": 0.48557082160790155,
+      "grad_norm": 0.47906458377838135,
+      "learning_rate": 9.518848370823205e-05,
+      "loss": 3.2809,
+      "step": 6600
+    },
+    {
+      "epoch": 0.48630653497397414,
+      "grad_norm": 0.4234841763973236,
+      "learning_rate": 9.517156668695962e-05,
+      "loss": 3.2427,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4870422483400467,
+      "grad_norm": 0.44342437386512756,
+      "learning_rate": 9.515462148687633e-05,
+      "loss": 3.246,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4877779617061193,
+      "grad_norm": 0.4205591082572937,
+      "learning_rate": 9.513764811855284e-05,
+      "loss": 3.2596,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4885136750721919,
+      "grad_norm": 0.4499843418598175,
+      "learning_rate": 9.512064659257749e-05,
+      "loss": 3.2535,
+      "step": 6640
+    },
+    {
+      "epoch": 0.48924938843826443,
+      "grad_norm": 0.4281398355960846,
+      "learning_rate": 9.510361691955608e-05,
+      "loss": 3.2301,
+      "step": 6650
+    },
+    {
+      "epoch": 0.489985101804337,
+      "grad_norm": 0.44669631123542786,
+      "learning_rate": 9.508655911011204e-05,
+      "loss": 3.2901,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4907208151704096,
+      "grad_norm": 0.4267096221446991,
+      "learning_rate": 9.506947317488634e-05,
+      "loss": 3.2702,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4914565285364822,
+      "grad_norm": 0.4454759657382965,
+      "learning_rate": 9.505235912453746e-05,
+      "loss": 3.2507,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4921922419025548,
+      "grad_norm": 0.42352691292762756,
+      "learning_rate": 9.503521696974146e-05,
+      "loss": 3.2663,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4929279552686273,
+      "grad_norm": 0.42943668365478516,
+      "learning_rate": 9.50180467211919e-05,
+      "loss": 3.2458,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4936636686346999,
+      "grad_norm": 0.4593370854854584,
+      "learning_rate": 9.50008483895999e-05,
+      "loss": 3.2092,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4943993820007725,
+      "grad_norm": 0.45478084683418274,
+      "learning_rate": 9.498362198569406e-05,
+      "loss": 3.3157,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4951350953668451,
+      "grad_norm": 0.4786463975906372,
+      "learning_rate": 9.496636752022053e-05,
+      "loss": 3.274,
+      "step": 6730
+    },
+    {
+      "epoch": 0.49587080873291767,
+      "grad_norm": 0.4654693901538849,
+      "learning_rate": 9.494908500394293e-05,
+      "loss": 3.2673,
+      "step": 6740
+    },
+    {
+      "epoch": 0.49660652209899026,
+      "grad_norm": 0.4555895924568176,
+      "learning_rate": 9.493177444764238e-05,
+      "loss": 3.2796,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4973422354650628,
+      "grad_norm": 0.4198709726333618,
+      "learning_rate": 9.491443586211756e-05,
+      "loss": 3.2293,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4980779488311354,
+      "grad_norm": 0.416696161031723,
+      "learning_rate": 9.489706925818454e-05,
+      "loss": 3.2122,
+      "step": 6770
+    },
+    {
+      "epoch": 0.49881366219720796,
+      "grad_norm": 0.43891972303390503,
+      "learning_rate": 9.487967464667691e-05,
+      "loss": 3.2953,
+      "step": 6780
+    },
+    {
+      "epoch": 0.49954937556328055,
+      "grad_norm": 0.42918211221694946,
+      "learning_rate": 9.486225203844575e-05,
+      "loss": 3.2869,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5002850889293531,
+      "grad_norm": 0.4077829420566559,
+      "learning_rate": 9.484480144435958e-05,
+      "loss": 3.2129,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5010208022954257,
+      "grad_norm": 0.4325737953186035,
+      "learning_rate": 9.482732287530438e-05,
+      "loss": 3.2208,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5017565156614983,
+      "grad_norm": 0.4620782732963562,
+      "learning_rate": 9.48098163421836e-05,
+      "loss": 3.2308,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5024922290275708,
+      "grad_norm": 0.4308145046234131,
+      "learning_rate": 9.479228185591809e-05,
+      "loss": 3.2112,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5032279423936434,
+      "grad_norm": 0.44704335927963257,
+      "learning_rate": 9.47747194274462e-05,
+      "loss": 3.2742,
+      "step": 6840
+    },
+    {
+      "epoch": 0.503963655759716,
+      "grad_norm": 0.4341581165790558,
+      "learning_rate": 9.475712906772367e-05,
+      "loss": 3.2398,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5046993691257886,
+      "grad_norm": 0.4126861095428467,
+      "learning_rate": 9.473951078772367e-05,
+      "loss": 3.2395,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5054350824918612,
+      "grad_norm": 0.4292146861553192,
+      "learning_rate": 9.47218645984368e-05,
+      "loss": 3.2677,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5061707958579338,
+      "grad_norm": 0.38945844769477844,
+      "learning_rate": 9.470419051087104e-05,
+      "loss": 3.3027,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5069065092240064,
+      "grad_norm": 0.42329898476600647,
+      "learning_rate": 9.46864885360518e-05,
+      "loss": 3.2205,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5076422225900789,
+      "grad_norm": 0.4216727614402771,
+      "learning_rate": 9.466875868502189e-05,
+      "loss": 3.2631,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5083779359561514,
+      "grad_norm": 0.45686569809913635,
+      "learning_rate": 9.46510009688415e-05,
+      "loss": 3.256,
+      "step": 6910
+    },
+    {
+      "epoch": 0.509113649322224,
+      "grad_norm": 0.4636845588684082,
+      "learning_rate": 9.463321539858821e-05,
+      "loss": 3.2798,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5098493626882966,
+      "grad_norm": 0.42751073837280273,
+      "learning_rate": 9.461540198535695e-05,
+      "loss": 3.2488,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5105850760543692,
+      "grad_norm": 0.44353827834129333,
+      "learning_rate": 9.459756074026003e-05,
+      "loss": 3.3067,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5113207894204418,
+      "grad_norm": 0.43093836307525635,
+      "learning_rate": 9.457969167442716e-05,
+      "loss": 3.2447,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5120565027865144,
+      "grad_norm": 0.4117187261581421,
+      "learning_rate": 9.456179479900536e-05,
+      "loss": 3.2757,
+      "step": 6960
+    },
+    {
+      "epoch": 0.512792216152587,
+      "grad_norm": 0.4384092688560486,
+      "learning_rate": 9.454387012515902e-05,
+      "loss": 3.244,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5135279295186596,
+      "grad_norm": 0.44672396779060364,
+      "learning_rate": 9.452591766406983e-05,
+      "loss": 3.2527,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5142636428847321,
+      "grad_norm": 0.41822147369384766,
+      "learning_rate": 9.45079374269369e-05,
+      "loss": 3.2535,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5149993562508047,
+      "grad_norm": 0.4356013834476471,
+      "learning_rate": 9.448992942497659e-05,
+      "loss": 3.2886,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5157350696168773,
+      "grad_norm": 0.4153957664966583,
+      "learning_rate": 9.44718936694226e-05,
+      "loss": 3.2152,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5164707829829498,
+      "grad_norm": 0.4125785827636719,
+      "learning_rate": 9.445383017152596e-05,
+      "loss": 3.2183,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5172064963490224,
+      "grad_norm": 0.41547253727912903,
+      "learning_rate": 9.4435738942555e-05,
+      "loss": 3.2421,
+      "step": 7030
+    },
+    {
+      "epoch": 0.517942209715095,
+      "grad_norm": 0.44014525413513184,
+      "learning_rate": 9.441761999379533e-05,
+      "loss": 3.2572,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5186779230811676,
+      "grad_norm": 0.4082861542701721,
+      "learning_rate": 9.439947333654985e-05,
+      "loss": 3.2241,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5194136364472401,
+      "grad_norm": 0.4412677586078644,
+      "learning_rate": 9.438129898213879e-05,
+      "loss": 3.2687,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5201493498133127,
+      "grad_norm": 0.42670226097106934,
+      "learning_rate": 9.436309694189963e-05,
+      "loss": 3.2835,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5208850631793853,
+      "grad_norm": 0.4251077473163605,
+      "learning_rate": 9.434486722718712e-05,
+      "loss": 3.2057,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5216207765454579,
+      "grad_norm": 0.4343535602092743,
+      "learning_rate": 9.432660984937325e-05,
+      "loss": 3.2419,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5223564899115305,
+      "grad_norm": 0.43313470482826233,
+      "learning_rate": 9.430832481984731e-05,
+      "loss": 3.2737,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5230922032776031,
+      "grad_norm": 0.4118582308292389,
+      "learning_rate": 9.429001215001581e-05,
+      "loss": 3.2394,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5238279166436757,
+      "grad_norm": 0.41588348150253296,
+      "learning_rate": 9.427167185130252e-05,
+      "loss": 3.232,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5245636300097481,
+      "grad_norm": 0.40552860498428345,
+      "learning_rate": 9.425330393514843e-05,
+      "loss": 3.2939,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5252993433758207,
+      "grad_norm": 0.45274195075035095,
+      "learning_rate": 9.423490841301178e-05,
+      "loss": 3.2761,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5260350567418933,
+      "grad_norm": 0.43010491132736206,
+      "learning_rate": 9.421648529636801e-05,
+      "loss": 3.2386,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5267707701079659,
+      "grad_norm": 0.4486023783683777,
+      "learning_rate": 9.41980345967098e-05,
+      "loss": 3.2547,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5275064834740385,
+      "grad_norm": 0.43645188212394714,
+      "learning_rate": 9.417955632554697e-05,
+      "loss": 3.2605,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5282421968401111,
+      "grad_norm": 0.4431368410587311,
+      "learning_rate": 9.416105049440662e-05,
+      "loss": 3.2333,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5289779102061837,
+      "grad_norm": 0.447656512260437,
+      "learning_rate": 9.414251711483303e-05,
+      "loss": 3.267,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5297136235722563,
+      "grad_norm": 0.4308032989501953,
+      "learning_rate": 9.41239561983876e-05,
+      "loss": 3.2702,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5304493369383289,
+      "grad_norm": 0.42887935042381287,
+      "learning_rate": 9.4105367756649e-05,
+      "loss": 3.2607,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5311850503044014,
+      "grad_norm": 0.44105619192123413,
+      "learning_rate": 9.408675180121298e-05,
+      "loss": 3.2511,
+      "step": 7220
+    },
+    {
+      "epoch": 0.531920763670474,
+      "grad_norm": 0.41123899817466736,
+      "learning_rate": 9.406810834369253e-05,
+      "loss": 3.2257,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5326564770365466,
+      "grad_norm": 0.4103626310825348,
+      "learning_rate": 9.404943739571774e-05,
+      "loss": 3.2482,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5333921904026191,
+      "grad_norm": 0.4202837646007538,
+      "learning_rate": 9.40307389689359e-05,
+      "loss": 3.2751,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5341279037686917,
+      "grad_norm": 0.4100697934627533,
+      "learning_rate": 9.401201307501139e-05,
+      "loss": 3.2473,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5348636171347643,
+      "grad_norm": 0.41449519991874695,
+      "learning_rate": 9.399325972562576e-05,
+      "loss": 3.1886,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5355993305008369,
+      "grad_norm": 0.42789408564567566,
+      "learning_rate": 9.397447893247767e-05,
+      "loss": 3.2318,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5363350438669094,
+      "grad_norm": 0.4444492757320404,
+      "learning_rate": 9.395567070728292e-05,
+      "loss": 3.2582,
+      "step": 7290
+    },
+    {
+      "epoch": 0.537070757232982,
+      "grad_norm": 0.4582795798778534,
+      "learning_rate": 9.39368350617744e-05,
+      "loss": 3.296,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5378064705990546,
+      "grad_norm": 0.41776713728904724,
+      "learning_rate": 9.39179720077021e-05,
+      "loss": 3.2169,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5385421839651272,
+      "grad_norm": 0.4293855130672455,
+      "learning_rate": 9.389908155683315e-05,
+      "loss": 3.2569,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5392778973311998,
+      "grad_norm": 0.4292116165161133,
+      "learning_rate": 9.388016372095172e-05,
+      "loss": 3.2732,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5400136106972724,
+      "grad_norm": 0.40844982862472534,
+      "learning_rate": 9.38612185118591e-05,
+      "loss": 3.2535,
+      "step": 7340
+    },
+    {
+      "epoch": 0.540749324063345,
+      "grad_norm": 0.4502781331539154,
+      "learning_rate": 9.384224594137363e-05,
+      "loss": 3.3008,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5414850374294174,
+      "grad_norm": 0.41482973098754883,
+      "learning_rate": 9.382324602133075e-05,
+      "loss": 3.2708,
+      "step": 7360
+    },
+    {
+      "epoch": 0.54222075079549,
+      "grad_norm": 0.407712459564209,
+      "learning_rate": 9.380421876358289e-05,
+      "loss": 3.2193,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5429564641615626,
+      "grad_norm": 0.40948861837387085,
+      "learning_rate": 9.378516417999966e-05,
+      "loss": 3.2599,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5436921775276352,
+      "grad_norm": 0.4116959273815155,
+      "learning_rate": 9.376608228246759e-05,
+      "loss": 3.212,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5444278908937078,
+      "grad_norm": 0.40750423073768616,
+      "learning_rate": 9.374697308289033e-05,
+      "loss": 3.228,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5451636042597804,
+      "grad_norm": 0.4093017280101776,
+      "learning_rate": 9.372783659318852e-05,
+      "loss": 3.2442,
+      "step": 7410
+    },
+    {
+      "epoch": 0.545899317625853,
+      "grad_norm": 0.4156046509742737,
+      "learning_rate": 9.37086728252998e-05,
+      "loss": 3.2938,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5466350309919256,
+      "grad_norm": 0.4261894226074219,
+      "learning_rate": 9.368948179117891e-05,
+      "loss": 3.2177,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5473707443579982,
+      "grad_norm": 0.4255836606025696,
+      "learning_rate": 9.367026350279755e-05,
+      "loss": 3.2701,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5481064577240707,
+      "grad_norm": 0.43402138352394104,
+      "learning_rate": 9.365101797214439e-05,
+      "loss": 3.2173,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5488421710901433,
+      "grad_norm": 0.4054631292819977,
+      "learning_rate": 9.363174521122516e-05,
+      "loss": 3.2659,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5495778844562158,
+      "grad_norm": 0.4579883813858032,
+      "learning_rate": 9.361244523206255e-05,
+      "loss": 3.2249,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5503135978222884,
+      "grad_norm": 0.40640729665756226,
+      "learning_rate": 9.359311804669617e-05,
+      "loss": 3.1534,
+      "step": 7480
+    },
+    {
+      "epoch": 0.551049311188361,
+      "grad_norm": 0.4273945093154907,
+      "learning_rate": 9.35737636671827e-05,
+      "loss": 3.271,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5517850245544336,
+      "grad_norm": 0.4102952778339386,
+      "learning_rate": 9.355438210559575e-05,
+      "loss": 3.2139,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5525207379205062,
+      "grad_norm": 0.43881139159202576,
+      "learning_rate": 9.353497337402583e-05,
+      "loss": 3.2115,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5532564512865787,
+      "grad_norm": 0.4113704264163971,
+      "learning_rate": 9.351553748458049e-05,
+      "loss": 3.2322,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5539921646526513,
+      "grad_norm": 0.4157336950302124,
+      "learning_rate": 9.349607444938416e-05,
+      "loss": 3.2485,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5547278780187239,
+      "grad_norm": 0.4389907717704773,
+      "learning_rate": 9.34765842805782e-05,
+      "loss": 3.2542,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5554635913847965,
+      "grad_norm": 0.45203500986099243,
+      "learning_rate": 9.345706699032093e-05,
+      "loss": 3.2511,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5561993047508691,
+      "grad_norm": 0.42221277952194214,
+      "learning_rate": 9.343752259078761e-05,
+      "loss": 3.2598,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5569350181169417,
+      "grad_norm": 0.4307563304901123,
+      "learning_rate": 9.341795109417033e-05,
+      "loss": 3.2495,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5576707314830143,
+      "grad_norm": 0.40989741683006287,
+      "learning_rate": 9.339835251267816e-05,
+      "loss": 3.2723,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5584064448490867,
+      "grad_norm": 0.4154078960418701,
+      "learning_rate": 9.337872685853702e-05,
+      "loss": 3.1988,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5591421582151593,
+      "grad_norm": 0.4232662618160248,
+      "learning_rate": 9.335907414398975e-05,
+      "loss": 3.2036,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5598778715812319,
+      "grad_norm": 0.4037014842033386,
+      "learning_rate": 9.333939438129603e-05,
+      "loss": 3.2554,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5606135849473045,
+      "grad_norm": 0.4122985005378723,
+      "learning_rate": 9.331968758273248e-05,
+      "loss": 3.2355,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5613492983133771,
+      "grad_norm": 0.40354782342910767,
+      "learning_rate": 9.32999537605925e-05,
+      "loss": 3.2135,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5620850116794497,
+      "grad_norm": 0.43905869126319885,
+      "learning_rate": 9.328019292718643e-05,
+      "loss": 3.139,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5628207250455223,
+      "grad_norm": 0.4707077443599701,
+      "learning_rate": 9.326040509484139e-05,
+      "loss": 3.2188,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5635564384115949,
+      "grad_norm": 0.4245714247226715,
+      "learning_rate": 9.324059027590142e-05,
+      "loss": 3.2614,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5642921517776675,
+      "grad_norm": 0.4019360840320587,
+      "learning_rate": 9.32207484827273e-05,
+      "loss": 3.2512,
+      "step": 7670
+    },
+    {
+      "epoch": 0.56502786514374,
+      "grad_norm": 0.4229870140552521,
+      "learning_rate": 9.320087972769671e-05,
+      "loss": 3.2514,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5657635785098126,
+      "grad_norm": 0.401112824678421,
+      "learning_rate": 9.318098402320412e-05,
+      "loss": 3.2666,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5664992918758851,
+      "grad_norm": 0.396757036447525,
+      "learning_rate": 9.316106138166082e-05,
+      "loss": 3.2273,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5672350052419577,
+      "grad_norm": 0.44432303309440613,
+      "learning_rate": 9.31411118154949e-05,
+      "loss": 3.2244,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5679707186080303,
+      "grad_norm": 0.4479587972164154,
+      "learning_rate": 9.312113533715125e-05,
+      "loss": 3.2359,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5687064319741029,
+      "grad_norm": 0.4089357554912567,
+      "learning_rate": 9.310113195909155e-05,
+      "loss": 3.2186,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5694421453401755,
+      "grad_norm": 0.3963668644428253,
+      "learning_rate": 9.308110169379425e-05,
+      "loss": 3.2326,
+      "step": 7740
+    },
+    {
+      "epoch": 0.570177858706248,
+      "grad_norm": 0.4088764190673828,
+      "learning_rate": 9.306104455375457e-05,
+      "loss": 3.2348,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5709135720723206,
+      "grad_norm": 0.41223156452178955,
+      "learning_rate": 9.304096055148452e-05,
+      "loss": 3.2306,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5716492854383932,
+      "grad_norm": 0.4141092896461487,
+      "learning_rate": 9.302084969951283e-05,
+      "loss": 3.1796,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5723849988044658,
+      "grad_norm": 0.4188261330127716,
+      "learning_rate": 9.300071201038503e-05,
+      "loss": 3.2539,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5731207121705384,
+      "grad_norm": 0.42103490233421326,
+      "learning_rate": 9.298054749666334e-05,
+      "loss": 3.2387,
+      "step": 7790
+    },
+    {
+      "epoch": 0.573856425536611,
+      "grad_norm": 0.41063937544822693,
+      "learning_rate": 9.296035617092673e-05,
+      "loss": 3.2052,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5745921389026836,
+      "grad_norm": 0.41358301043510437,
+      "learning_rate": 9.294013804577093e-05,
+      "loss": 3.2474,
+      "step": 7810
+    },
+    {
+      "epoch": 0.575327852268756,
+      "grad_norm": 0.4047495722770691,
+      "learning_rate": 9.291989313380833e-05,
+      "loss": 3.1907,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5760635656348286,
+      "grad_norm": 0.4267469644546509,
+      "learning_rate": 9.289962144766806e-05,
+      "loss": 3.2109,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5767992790009012,
+      "grad_norm": 0.42642202973365784,
+      "learning_rate": 9.287932299999597e-05,
+      "loss": 3.2168,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5775349923669738,
+      "grad_norm": 0.4209352433681488,
+      "learning_rate": 9.285899780345458e-05,
+      "loss": 3.2363,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5782707057330464,
+      "grad_norm": 0.39841902256011963,
+      "learning_rate": 9.283864587072311e-05,
+      "loss": 3.2399,
+      "step": 7860
+    },
+    {
+      "epoch": 0.579006419099119,
+      "grad_norm": 0.42725417017936707,
+      "learning_rate": 9.281826721449741e-05,
+      "loss": 3.224,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5797421324651916,
+      "grad_norm": 0.4100605845451355,
+      "learning_rate": 9.279786184749009e-05,
+      "loss": 3.184,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5804778458312642,
+      "grad_norm": 0.4014204144477844,
+      "learning_rate": 9.277742978243033e-05,
+      "loss": 3.2129,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5812135591973367,
+      "grad_norm": 0.4076472222805023,
+      "learning_rate": 9.275697103206406e-05,
+      "loss": 3.2587,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5819492725634093,
+      "grad_norm": 0.4247482120990753,
+      "learning_rate": 9.273648560915376e-05,
+      "loss": 3.2307,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5826849859294819,
+      "grad_norm": 0.3832865059375763,
+      "learning_rate": 9.271597352647861e-05,
+      "loss": 3.2394,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5834206992955544,
+      "grad_norm": 0.38227957487106323,
+      "learning_rate": 9.269543479683441e-05,
+      "loss": 3.2031,
+      "step": 7930
+    },
+    {
+      "epoch": 0.584156412661627,
+      "grad_norm": 0.4235737919807434,
+      "learning_rate": 9.267486943303358e-05,
+      "loss": 3.2341,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5848921260276996,
+      "grad_norm": 0.40164512395858765,
+      "learning_rate": 9.265427744790513e-05,
+      "loss": 3.1821,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5856278393937722,
+      "grad_norm": 0.40419676899909973,
+      "learning_rate": 9.263365885429474e-05,
+      "loss": 3.2159,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5863635527598448,
+      "grad_norm": 0.38921478390693665,
+      "learning_rate": 9.26130136650646e-05,
+      "loss": 3.2628,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5870992661259173,
+      "grad_norm": 0.43494653701782227,
+      "learning_rate": 9.259234189309359e-05,
+      "loss": 3.1855,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5878349794919899,
+      "grad_norm": 0.4748268127441406,
+      "learning_rate": 9.257164355127705e-05,
+      "loss": 3.2326,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5885706928580625,
+      "grad_norm": 0.40674278140068054,
+      "learning_rate": 9.255091865252704e-05,
+      "loss": 3.2202,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5893064062241351,
+      "grad_norm": 0.4126112759113312,
+      "learning_rate": 9.253016720977208e-05,
+      "loss": 3.1978,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5900421195902077,
+      "grad_norm": 0.40558794140815735,
+      "learning_rate": 9.250938923595729e-05,
+      "loss": 3.2041,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5907778329562803,
+      "grad_norm": 0.41431719064712524,
+      "learning_rate": 9.248858474404431e-05,
+      "loss": 3.2145,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5915135463223528,
+      "grad_norm": 0.4020199477672577,
+      "learning_rate": 9.246775374701139e-05,
+      "loss": 3.1712,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5922492596884253,
+      "grad_norm": 0.4117688238620758,
+      "learning_rate": 9.244689625785323e-05,
+      "loss": 3.2534,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5929849730544979,
+      "grad_norm": 0.4227346181869507,
+      "learning_rate": 9.24260122895811e-05,
+      "loss": 3.2094,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5937206864205705,
+      "grad_norm": 0.3910035490989685,
+      "learning_rate": 9.24051018552228e-05,
+      "loss": 3.2119,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5944563997866431,
+      "grad_norm": 0.41752034425735474,
+      "learning_rate": 9.238416496782261e-05,
+      "loss": 3.1458,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5951921131527157,
+      "grad_norm": 0.40074622631073,
+      "learning_rate": 9.236320164044132e-05,
+      "loss": 3.2612,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5959278265187883,
+      "grad_norm": 0.4130959212779999,
+      "learning_rate": 9.234221188615624e-05,
+      "loss": 3.1766,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5966635398848609,
+      "grad_norm": 0.4178720712661743,
+      "learning_rate": 9.232119571806114e-05,
+      "loss": 3.2147,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5973992532509335,
+      "grad_norm": 0.408380925655365,
+      "learning_rate": 9.230015314926625e-05,
+      "loss": 3.2038,
+      "step": 8120
+    },
+    {
+      "epoch": 0.598134966617006,
+      "grad_norm": 0.40057528018951416,
+      "learning_rate": 9.227908419289833e-05,
+      "loss": 3.203,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5988706799830786,
+      "grad_norm": 0.42866384983062744,
+      "learning_rate": 9.225798886210052e-05,
+      "loss": 3.217,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5996063933491512,
+      "grad_norm": 0.4330470860004425,
+      "learning_rate": 9.223686717003248e-05,
+      "loss": 3.1826,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6003421067152237,
+      "grad_norm": 0.389643132686615,
+      "learning_rate": 9.221571912987029e-05,
+      "loss": 3.2339,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6010778200812963,
+      "grad_norm": 0.412850558757782,
+      "learning_rate": 9.219454475480644e-05,
+      "loss": 3.1956,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6018135334473689,
+      "grad_norm": 0.4014042317867279,
+      "learning_rate": 9.217334405804991e-05,
+      "loss": 3.1871,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6025492468134415,
+      "grad_norm": 0.4323343336582184,
+      "learning_rate": 9.215211705282604e-05,
+      "loss": 3.2368,
+      "step": 8190
+    },
+    {
+      "epoch": 0.603284960179514,
+      "grad_norm": 0.4045550227165222,
+      "learning_rate": 9.213086375237663e-05,
+      "loss": 3.186,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6040206735455866,
+      "grad_norm": 0.39329978823661804,
+      "learning_rate": 9.210958416995984e-05,
+      "loss": 3.2015,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6047563869116592,
+      "grad_norm": 0.40735575556755066,
+      "learning_rate": 9.208827831885026e-05,
+      "loss": 3.2184,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6054921002777318,
+      "grad_norm": 0.4116288125514984,
+      "learning_rate": 9.206694621233882e-05,
+      "loss": 3.2264,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6062278136438044,
+      "grad_norm": 0.4052058458328247,
+      "learning_rate": 9.204558786373288e-05,
+      "loss": 3.2455,
+      "step": 8240
+    },
+    {
+      "epoch": 0.606963527009877,
+      "grad_norm": 0.41139885783195496,
+      "learning_rate": 9.202420328635616e-05,
+      "loss": 3.1959,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6076992403759496,
+      "grad_norm": 0.4093032777309418,
+      "learning_rate": 9.200279249354872e-05,
+      "loss": 3.1881,
+      "step": 8260
+    },
+    {
+      "epoch": 0.608434953742022,
+      "grad_norm": 0.443782240152359,
+      "learning_rate": 9.198135549866699e-05,
+      "loss": 3.2383,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6091706671080946,
+      "grad_norm": 0.42610660195350647,
+      "learning_rate": 9.195989231508373e-05,
+      "loss": 3.2359,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6099063804741672,
+      "grad_norm": 0.4072706401348114,
+      "learning_rate": 9.193840295618807e-05,
+      "loss": 3.2154,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6106420938402398,
+      "grad_norm": 0.4002029299736023,
+      "learning_rate": 9.191688743538542e-05,
+      "loss": 3.1847,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6113778072063124,
+      "grad_norm": 0.4040053188800812,
+      "learning_rate": 9.189534576609757e-05,
+      "loss": 3.2215,
+      "step": 8310
+    },
+    {
+      "epoch": 0.612113520572385,
+      "grad_norm": 0.4218162000179291,
+      "learning_rate": 9.187377796176256e-05,
+      "loss": 3.2408,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6128492339384576,
+      "grad_norm": 0.42694374918937683,
+      "learning_rate": 9.185218403583478e-05,
+      "loss": 3.2176,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6135849473045302,
+      "grad_norm": 0.40308234095573425,
+      "learning_rate": 9.18305640017849e-05,
+      "loss": 3.1799,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6143206606706028,
+      "grad_norm": 0.3952566981315613,
+      "learning_rate": 9.180891787309985e-05,
+      "loss": 3.244,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6150563740366753,
+      "grad_norm": 0.39796093106269836,
+      "learning_rate": 9.178724566328287e-05,
+      "loss": 3.1809,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6157920874027479,
+      "grad_norm": 0.41804805397987366,
+      "learning_rate": 9.176554738585348e-05,
+      "loss": 3.228,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6165278007688205,
+      "grad_norm": 0.4276120066642761,
+      "learning_rate": 9.174382305434742e-05,
+      "loss": 3.2218,
+      "step": 8380
+    },
+    {
+      "epoch": 0.617263514134893,
+      "grad_norm": 0.44311702251434326,
+      "learning_rate": 9.172207268231674e-05,
+      "loss": 3.2191,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6179992275009656,
+      "grad_norm": 0.43954578042030334,
+      "learning_rate": 9.170029628332966e-05,
+      "loss": 3.2388,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6187349408670382,
+      "grad_norm": 0.38372981548309326,
+      "learning_rate": 9.167849387097072e-05,
+      "loss": 3.2059,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6194706542331108,
+      "grad_norm": 0.3934963643550873,
+      "learning_rate": 9.165666545884062e-05,
+      "loss": 3.2022,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6202063675991834,
+      "grad_norm": 0.41195252537727356,
+      "learning_rate": 9.163481106055629e-05,
+      "loss": 3.1958,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6209420809652559,
+      "grad_norm": 0.42152824997901917,
+      "learning_rate": 9.161293068975092e-05,
+      "loss": 3.2529,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6216777943313285,
+      "grad_norm": 0.41656965017318726,
+      "learning_rate": 9.159102436007385e-05,
+      "loss": 3.2559,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6224135076974011,
+      "grad_norm": 0.3994307518005371,
+      "learning_rate": 9.156909208519062e-05,
+      "loss": 3.2101,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6231492210634737,
+      "grad_norm": 0.40668097138404846,
+      "learning_rate": 9.154713387878297e-05,
+      "loss": 3.1927,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6238849344295463,
+      "grad_norm": 0.42031413316726685,
+      "learning_rate": 9.152514975454884e-05,
+      "loss": 3.2174,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6246206477956189,
+      "grad_norm": 0.3852512836456299,
+      "learning_rate": 9.150313972620227e-05,
+      "loss": 3.1662,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6253563611616914,
+      "grad_norm": 0.3993820250034332,
+      "learning_rate": 9.148110380747352e-05,
+      "loss": 3.1738,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6260920745277639,
+      "grad_norm": 0.4049205482006073,
+      "learning_rate": 9.145904201210898e-05,
+      "loss": 3.1861,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6268277878938365,
+      "grad_norm": 0.40449580550193787,
+      "learning_rate": 9.143695435387117e-05,
+      "loss": 3.2229,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6275635012599091,
+      "grad_norm": 0.39980998635292053,
+      "learning_rate": 9.141484084653878e-05,
+      "loss": 3.2294,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6282992146259817,
+      "grad_norm": 0.42464175820350647,
+      "learning_rate": 9.139270150390658e-05,
+      "loss": 3.2189,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6290349279920543,
+      "grad_norm": 0.3991568684577942,
+      "learning_rate": 9.137053633978548e-05,
+      "loss": 3.2124,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6297706413581269,
+      "grad_norm": 0.40061402320861816,
+      "learning_rate": 9.13483453680025e-05,
+      "loss": 3.1887,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6305063547241995,
+      "grad_norm": 0.3983337879180908,
+      "learning_rate": 9.132612860240075e-05,
+      "loss": 3.1858,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6312420680902721,
+      "grad_norm": 0.38647231459617615,
+      "learning_rate": 9.130388605683946e-05,
+      "loss": 3.2476,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6319777814563446,
+      "grad_norm": 0.40448683500289917,
+      "learning_rate": 9.128161774519389e-05,
+      "loss": 3.1775,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6327134948224172,
+      "grad_norm": 0.4101662039756775,
+      "learning_rate": 9.125932368135542e-05,
+      "loss": 3.2317,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6334492081884897,
+      "grad_norm": 0.4105026125907898,
+      "learning_rate": 9.123700387923144e-05,
+      "loss": 3.2358,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6341849215545623,
+      "grad_norm": 0.42156288027763367,
+      "learning_rate": 9.121465835274549e-05,
+      "loss": 3.2202,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.4276079833507538,
+      "learning_rate": 9.119228711583705e-05,
+      "loss": 3.2128,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6356563482867075,
+      "grad_norm": 0.3834649324417114,
+      "learning_rate": 9.116989018246171e-05,
+      "loss": 3.1966,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6363920616527801,
+      "grad_norm": 0.39798030257225037,
+      "learning_rate": 9.114746756659105e-05,
+      "loss": 3.1433,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6371277750188526,
+      "grad_norm": 0.3975779414176941,
+      "learning_rate": 9.112501928221271e-05,
+      "loss": 3.2573,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6378634883849252,
+      "grad_norm": 0.3947142958641052,
+      "learning_rate": 9.110254534333033e-05,
+      "loss": 3.191,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6385992017509978,
+      "grad_norm": 0.4371863901615143,
+      "learning_rate": 9.10800457639635e-05,
+      "loss": 3.2229,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6393349151170704,
+      "grad_norm": 0.39925292134284973,
+      "learning_rate": 9.105752055814789e-05,
+      "loss": 3.2405,
+      "step": 8690
+    },
+    {
+      "epoch": 0.640070628483143,
+      "grad_norm": 0.40571677684783936,
+      "learning_rate": 9.10349697399351e-05,
+      "loss": 3.2267,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6408063418492156,
+      "grad_norm": 0.3942791223526001,
+      "learning_rate": 9.101239332339272e-05,
+      "loss": 3.2074,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6415420552152882,
+      "grad_norm": 0.42931607365608215,
+      "learning_rate": 9.098979132260435e-05,
+      "loss": 3.1736,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6422777685813607,
+      "grad_norm": 0.4185764789581299,
+      "learning_rate": 9.096716375166945e-05,
+      "loss": 3.1654,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6430134819474332,
+      "grad_norm": 0.4023076295852661,
+      "learning_rate": 9.094451062470353e-05,
+      "loss": 3.1562,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6437491953135058,
+      "grad_norm": 0.40013644099235535,
+      "learning_rate": 9.0921831955838e-05,
+      "loss": 3.209,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6444849086795784,
+      "grad_norm": 0.4097345471382141,
+      "learning_rate": 9.089912775922021e-05,
+      "loss": 3.1925,
+      "step": 8760
+    },
+    {
+      "epoch": 0.645220622045651,
+      "grad_norm": 0.4017821252346039,
+      "learning_rate": 9.087639804901343e-05,
+      "loss": 3.1849,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6459563354117236,
+      "grad_norm": 0.40877318382263184,
+      "learning_rate": 9.085364283939686e-05,
+      "loss": 3.2203,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6466920487777962,
+      "grad_norm": 0.4291410744190216,
+      "learning_rate": 9.083086214456557e-05,
+      "loss": 3.1789,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6474277621438688,
+      "grad_norm": 0.4058206379413605,
+      "learning_rate": 9.08080559787306e-05,
+      "loss": 3.1903,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6481634755099414,
+      "grad_norm": 0.3808108866214752,
+      "learning_rate": 9.078522435611879e-05,
+      "loss": 3.1841,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6488991888760139,
+      "grad_norm": 0.43283572793006897,
+      "learning_rate": 9.076236729097294e-05,
+      "loss": 3.2212,
+      "step": 8820
+    },
+    {
+      "epoch": 0.6496349022420865,
+      "grad_norm": 0.39330145716667175,
+      "learning_rate": 9.073948479755164e-05,
+      "loss": 3.2158,
+      "step": 8830
+    },
+    {
+      "epoch": 0.650370615608159,
+      "grad_norm": 0.40737244486808777,
+      "learning_rate": 9.071657689012943e-05,
+      "loss": 3.2306,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6511063289742316,
+      "grad_norm": 0.3892798125743866,
+      "learning_rate": 9.069364358299666e-05,
+      "loss": 3.2162,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6518420423403042,
+      "grad_norm": 0.400855153799057,
+      "learning_rate": 9.067068489045949e-05,
+      "loss": 3.1358,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6525777557063768,
+      "grad_norm": 0.39898017048835754,
+      "learning_rate": 9.064770082684e-05,
+      "loss": 3.1866,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6533134690724494,
+      "grad_norm": 0.41870006918907166,
+      "learning_rate": 9.062469140647601e-05,
+      "loss": 3.1855,
+      "step": 8880
+    },
+    {
+      "epoch": 0.654049182438522,
+      "grad_norm": 0.4006561040878296,
+      "learning_rate": 9.060165664372122e-05,
+      "loss": 3.2137,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6547848958045945,
+      "grad_norm": 0.4057448208332062,
+      "learning_rate": 9.05785965529451e-05,
+      "loss": 3.1655,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6555206091706671,
+      "grad_norm": 0.399264931678772,
+      "learning_rate": 9.055551114853295e-05,
+      "loss": 3.214,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6562563225367397,
+      "grad_norm": 0.399167001247406,
+      "learning_rate": 9.053240044488586e-05,
+      "loss": 3.1959,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6569920359028123,
+      "grad_norm": 0.39689528942108154,
+      "learning_rate": 9.050926445642067e-05,
+      "loss": 3.2065,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6577277492688849,
+      "grad_norm": 0.4583052098751068,
+      "learning_rate": 9.048610319757002e-05,
+      "loss": 3.2269,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6584634626349575,
+      "grad_norm": 0.4117734134197235,
+      "learning_rate": 9.046291668278231e-05,
+      "loss": 3.1955,
+      "step": 8950
+    },
+    {
+      "epoch": 0.65919917600103,
+      "grad_norm": 0.3891928195953369,
+      "learning_rate": 9.043970492652168e-05,
+      "loss": 3.2248,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6599348893671025,
+      "grad_norm": 0.4023609757423401,
+      "learning_rate": 9.041646794326803e-05,
+      "loss": 3.2241,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6606706027331751,
+      "grad_norm": 0.39141857624053955,
+      "learning_rate": 9.039320574751702e-05,
+      "loss": 3.1826,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6614063160992477,
+      "grad_norm": 0.38678160309791565,
+      "learning_rate": 9.036991835377997e-05,
+      "loss": 3.1664,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6621420294653203,
+      "grad_norm": 0.3716258704662323,
+      "learning_rate": 9.0346605776584e-05,
+      "loss": 3.1835,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6628777428313929,
+      "grad_norm": 0.38958126306533813,
+      "learning_rate": 9.032326803047189e-05,
+      "loss": 3.2211,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6636134561974655,
+      "grad_norm": 0.3944793939590454,
+      "learning_rate": 9.029990513000213e-05,
+      "loss": 3.2145,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6643491695635381,
+      "grad_norm": 0.41081830859184265,
+      "learning_rate": 9.027651708974891e-05,
+      "loss": 3.2223,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6650848829296107,
+      "grad_norm": 0.3970061242580414,
+      "learning_rate": 9.025310392430208e-05,
+      "loss": 3.2183,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6658205962956832,
+      "grad_norm": 0.4093196988105774,
+      "learning_rate": 9.022966564826721e-05,
+      "loss": 3.2203,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6665563096617558,
+      "grad_norm": 0.4082465171813965,
+      "learning_rate": 9.020620227626548e-05,
+      "loss": 3.1892,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6672920230278283,
+      "grad_norm": 0.40329813957214355,
+      "learning_rate": 9.018271382293377e-05,
+      "loss": 3.1848,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6680277363939009,
+      "grad_norm": 0.3946298062801361,
+      "learning_rate": 9.015920030292457e-05,
+      "loss": 3.2053,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6687634497599735,
+      "grad_norm": 0.39222556352615356,
+      "learning_rate": 9.013566173090604e-05,
+      "loss": 3.2543,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6694991631260461,
+      "grad_norm": 0.39037418365478516,
+      "learning_rate": 9.011209812156194e-05,
+      "loss": 3.1635,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6702348764921187,
+      "grad_norm": 0.4084848463535309,
+      "learning_rate": 9.008850948959164e-05,
+      "loss": 3.1981,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6709705898581912,
+      "grad_norm": 0.4072742760181427,
+      "learning_rate": 9.00648958497102e-05,
+      "loss": 3.197,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6717063032242638,
+      "grad_norm": 0.396036833524704,
+      "learning_rate": 9.004125721664816e-05,
+      "loss": 3.2572,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6724420165903364,
+      "grad_norm": 0.3960723578929901,
+      "learning_rate": 9.001759360515174e-05,
+      "loss": 3.2029,
+      "step": 9140
+    },
+    {
+      "epoch": 0.673177729956409,
+      "grad_norm": 0.4176108241081238,
+      "learning_rate": 8.999390502998271e-05,
+      "loss": 3.2187,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6739134433224816,
+      "grad_norm": 0.39103132486343384,
+      "learning_rate": 8.99701915059184e-05,
+      "loss": 3.2403,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6746491566885542,
+      "grad_norm": 0.40875765681266785,
+      "learning_rate": 8.994645304775175e-05,
+      "loss": 3.196,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6753848700546267,
+      "grad_norm": 0.3849492073059082,
+      "learning_rate": 8.992268967029121e-05,
+      "loss": 3.2015,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6761205834206993,
+      "grad_norm": 0.3874247968196869,
+      "learning_rate": 8.989890138836077e-05,
+      "loss": 3.2458,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6768562967867718,
+      "grad_norm": 0.4249545931816101,
+      "learning_rate": 8.98750882168e-05,
+      "loss": 3.2442,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6775920101528444,
+      "grad_norm": 0.37782514095306396,
+      "learning_rate": 8.985125017046397e-05,
+      "loss": 3.2012,
+      "step": 9210
+    },
+    {
+      "epoch": 0.678327723518917,
+      "grad_norm": 0.3978985846042633,
+      "learning_rate": 8.982738726422325e-05,
+      "loss": 3.2127,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6790634368849896,
+      "grad_norm": 0.38309648633003235,
+      "learning_rate": 8.980349951296396e-05,
+      "loss": 3.1858,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6797991502510622,
+      "grad_norm": 0.42622271180152893,
+      "learning_rate": 8.977958693158768e-05,
+      "loss": 3.177,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6805348636171348,
+      "grad_norm": 0.3749631345272064,
+      "learning_rate": 8.97556495350115e-05,
+      "loss": 3.1836,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6812705769832074,
+      "grad_norm": 0.4103826582431793,
+      "learning_rate": 8.973168733816799e-05,
+      "loss": 3.188,
+      "step": 9260
+    },
+    {
+      "epoch": 0.68200629034928,
+      "grad_norm": 0.3920148015022278,
+      "learning_rate": 8.970770035600518e-05,
+      "loss": 3.2252,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6827420037153525,
+      "grad_norm": 0.3953123688697815,
+      "learning_rate": 8.968368860348657e-05,
+      "loss": 3.1857,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6834777170814251,
+      "grad_norm": 0.39929309487342834,
+      "learning_rate": 8.965965209559111e-05,
+      "loss": 3.178,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6842134304474976,
+      "grad_norm": 0.4072224795818329,
+      "learning_rate": 8.963559084731319e-05,
+      "loss": 3.1687,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6849491438135702,
+      "grad_norm": 0.421752005815506,
+      "learning_rate": 8.961150487366265e-05,
+      "loss": 3.1959,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6856848571796428,
+      "grad_norm": 0.3856801986694336,
+      "learning_rate": 8.958739418966472e-05,
+      "loss": 3.1962,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6864205705457154,
+      "grad_norm": 0.4316472113132477,
+      "learning_rate": 8.956325881036008e-05,
+      "loss": 3.1759,
+      "step": 9330
+    },
+    {
+      "epoch": 0.687156283911788,
+      "grad_norm": 0.39210107922554016,
+      "learning_rate": 8.953909875080477e-05,
+      "loss": 3.1787,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6878919972778605,
+      "grad_norm": 0.39624738693237305,
+      "learning_rate": 8.951491402607031e-05,
+      "loss": 3.2155,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6886277106439331,
+      "grad_norm": 0.4137401878833771,
+      "learning_rate": 8.949070465124352e-05,
+      "loss": 3.2029,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6893634240100057,
+      "grad_norm": 0.37497127056121826,
+      "learning_rate": 8.946647064142662e-05,
+      "loss": 3.213,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6900991373760783,
+      "grad_norm": 0.4009439945220947,
+      "learning_rate": 8.944221201173723e-05,
+      "loss": 3.1978,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6908348507421509,
+      "grad_norm": 0.390789270401001,
+      "learning_rate": 8.941792877730829e-05,
+      "loss": 3.1502,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6915705641082235,
+      "grad_norm": 0.3750687837600708,
+      "learning_rate": 8.939362095328811e-05,
+      "loss": 3.2011,
+      "step": 9400
+    },
+    {
+      "epoch": 0.692306277474296,
+      "grad_norm": 0.3848758041858673,
+      "learning_rate": 8.936928855484035e-05,
+      "loss": 3.2202,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6930419908403685,
+      "grad_norm": 0.406440794467926,
+      "learning_rate": 8.934493159714397e-05,
+      "loss": 3.1547,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6937777042064411,
+      "grad_norm": 0.37811365723609924,
+      "learning_rate": 8.932055009539326e-05,
+      "loss": 3.1863,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6945134175725137,
+      "grad_norm": 0.3805864155292511,
+      "learning_rate": 8.929614406479785e-05,
+      "loss": 3.1736,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6952491309385863,
+      "grad_norm": 0.3834070563316345,
+      "learning_rate": 8.927171352058262e-05,
+      "loss": 3.1659,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6959848443046589,
+      "grad_norm": 0.37774044275283813,
+      "learning_rate": 8.924725847798776e-05,
+      "loss": 3.1884,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6967205576707315,
+      "grad_norm": 0.3782852590084076,
+      "learning_rate": 8.92227789522688e-05,
+      "loss": 3.1593,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6974562710368041,
+      "grad_norm": 0.4085491895675659,
+      "learning_rate": 8.919827495869645e-05,
+      "loss": 3.2332,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6981919844028767,
+      "grad_norm": 0.38499242067337036,
+      "learning_rate": 8.917374651255675e-05,
+      "loss": 3.2051,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6989276977689493,
+      "grad_norm": 0.41638848185539246,
+      "learning_rate": 8.914919362915096e-05,
+      "loss": 3.1209,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6996634111350218,
+      "grad_norm": 0.3790314197540283,
+      "learning_rate": 8.912461632379561e-05,
+      "loss": 3.1852,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7003991245010944,
+      "grad_norm": 0.3726657032966614,
+      "learning_rate": 8.910001461182245e-05,
+      "loss": 3.1979,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7011348378671669,
+      "grad_norm": 0.4013272523880005,
+      "learning_rate": 8.907538850857844e-05,
+      "loss": 3.1678,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7018705512332395,
+      "grad_norm": 0.4136090576648712,
+      "learning_rate": 8.905073802942579e-05,
+      "loss": 3.1486,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7026062645993121,
+      "grad_norm": 0.4203248918056488,
+      "learning_rate": 8.902606318974189e-05,
+      "loss": 3.2102,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7033419779653847,
+      "grad_norm": 0.41708293557167053,
+      "learning_rate": 8.900136400491934e-05,
+      "loss": 3.2344,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7040776913314573,
+      "grad_norm": 0.40059348940849304,
+      "learning_rate": 8.897664049036594e-05,
+      "loss": 3.1628,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7048134046975298,
+      "grad_norm": 0.41126495599746704,
+      "learning_rate": 8.89518926615046e-05,
+      "loss": 3.2079,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7055491180636024,
+      "grad_norm": 0.39519059658050537,
+      "learning_rate": 8.89271205337735e-05,
+      "loss": 3.1738,
+      "step": 9590
+    },
+    {
+      "epoch": 0.706284831429675,
+      "grad_norm": 0.3954789936542511,
+      "learning_rate": 8.890232412262588e-05,
+      "loss": 3.2131,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7070205447957476,
+      "grad_norm": 0.37941592931747437,
+      "learning_rate": 8.887750344353023e-05,
+      "loss": 3.1606,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7077562581618202,
+      "grad_norm": 0.37760573625564575,
+      "learning_rate": 8.885265851197006e-05,
+      "loss": 3.1641,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7084919715278928,
+      "grad_norm": 0.3869093358516693,
+      "learning_rate": 8.882778934344412e-05,
+      "loss": 3.1779,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7092276848939653,
+      "grad_norm": 0.37767353653907776,
+      "learning_rate": 8.88028959534662e-05,
+      "loss": 3.1538,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7099633982600378,
+      "grad_norm": 0.3897157907485962,
+      "learning_rate": 8.877797835756523e-05,
+      "loss": 3.1598,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7106991116261104,
+      "grad_norm": 0.4006219804286957,
+      "learning_rate": 8.875303657128526e-05,
+      "loss": 3.1951,
+      "step": 9660
+    },
+    {
+      "epoch": 0.711434824992183,
+      "grad_norm": 0.3921482563018799,
+      "learning_rate": 8.872807061018542e-05,
+      "loss": 3.177,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7121705383582556,
+      "grad_norm": 0.3957517147064209,
+      "learning_rate": 8.870308048983989e-05,
+      "loss": 3.2062,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7129062517243282,
+      "grad_norm": 0.4034371078014374,
+      "learning_rate": 8.867806622583793e-05,
+      "loss": 3.1455,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7136419650904008,
+      "grad_norm": 0.38453301787376404,
+      "learning_rate": 8.865302783378391e-05,
+      "loss": 3.1829,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7143776784564734,
+      "grad_norm": 0.40622347593307495,
+      "learning_rate": 8.862796532929717e-05,
+      "loss": 3.1663,
+      "step": 9710
+    },
+    {
+      "epoch": 0.715113391822546,
+      "grad_norm": 0.4114111065864563,
+      "learning_rate": 8.86028787280122e-05,
+      "loss": 3.2128,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7158491051886186,
+      "grad_norm": 0.3917147219181061,
+      "learning_rate": 8.857776804557842e-05,
+      "loss": 3.1601,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7165848185546911,
+      "grad_norm": 0.3908882141113281,
+      "learning_rate": 8.85526332976603e-05,
+      "loss": 3.2072,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7173205319207636,
+      "grad_norm": 0.3869069814682007,
+      "learning_rate": 8.852747449993735e-05,
+      "loss": 3.1435,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7180562452868362,
+      "grad_norm": 0.37940549850463867,
+      "learning_rate": 8.850229166810407e-05,
+      "loss": 3.1857,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7187919586529088,
+      "grad_norm": 0.40026140213012695,
+      "learning_rate": 8.847708481786996e-05,
+      "loss": 3.1549,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7195276720189814,
+      "grad_norm": 0.39495396614074707,
+      "learning_rate": 8.845185396495947e-05,
+      "loss": 3.1916,
+      "step": 9780
+    },
+    {
+      "epoch": 0.720263385385054,
+      "grad_norm": 0.3766854405403137,
+      "learning_rate": 8.842659912511204e-05,
+      "loss": 3.1937,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7209990987511266,
+      "grad_norm": 0.39338842034339905,
+      "learning_rate": 8.84013203140821e-05,
+      "loss": 3.1683,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7217348121171991,
+      "grad_norm": 0.3664061725139618,
+      "learning_rate": 8.837601754763902e-05,
+      "loss": 3.1868,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7224705254832717,
+      "grad_norm": 0.37891605496406555,
+      "learning_rate": 8.835069084156709e-05,
+      "loss": 3.1325,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7232062388493443,
+      "grad_norm": 0.4005630910396576,
+      "learning_rate": 8.832534021166555e-05,
+      "loss": 3.2046,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7239419522154169,
+      "grad_norm": 0.3800733983516693,
+      "learning_rate": 8.829996567374857e-05,
+      "loss": 3.2048,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7246776655814895,
+      "grad_norm": 0.396013468503952,
+      "learning_rate": 8.827456724364525e-05,
+      "loss": 3.1523,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7254133789475621,
+      "grad_norm": 0.37606799602508545,
+      "learning_rate": 8.824914493719952e-05,
+      "loss": 3.1769,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7261490923136346,
+      "grad_norm": 0.3814736604690552,
+      "learning_rate": 8.82236987702703e-05,
+      "loss": 3.1618,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7268848056797071,
+      "grad_norm": 0.42829668521881104,
+      "learning_rate": 8.819822875873136e-05,
+      "loss": 3.1892,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7276205190457797,
+      "grad_norm": 0.3849022686481476,
+      "learning_rate": 8.817273491847129e-05,
+      "loss": 3.2523,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7283562324118523,
+      "grad_norm": 0.3935648202896118,
+      "learning_rate": 8.814721726539366e-05,
+      "loss": 3.1743,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7290919457779249,
+      "grad_norm": 0.3935147523880005,
+      "learning_rate": 8.812167581541675e-05,
+      "loss": 3.1683,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7298276591439975,
+      "grad_norm": 0.41291266679763794,
+      "learning_rate": 8.809611058447382e-05,
+      "loss": 3.202,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7305633725100701,
+      "grad_norm": 0.39814290404319763,
+      "learning_rate": 8.807052158851289e-05,
+      "loss": 3.2389,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7312990858761427,
+      "grad_norm": 0.38150617480278015,
+      "learning_rate": 8.804490884349683e-05,
+      "loss": 3.1738,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7320347992422153,
+      "grad_norm": 0.41697201132774353,
+      "learning_rate": 8.801927236540329e-05,
+      "loss": 3.2173,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7327705126082878,
+      "grad_norm": 0.40556851029396057,
+      "learning_rate": 8.799361217022477e-05,
+      "loss": 3.115,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7335062259743604,
+      "grad_norm": 0.3922794759273529,
+      "learning_rate": 8.796792827396855e-05,
+      "loss": 3.1743,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7342419393404329,
+      "grad_norm": 0.3818107545375824,
+      "learning_rate": 8.794222069265669e-05,
+      "loss": 3.1707,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7349776527065055,
+      "grad_norm": 0.380044549703598,
+      "learning_rate": 8.791648944232603e-05,
+      "loss": 3.1717,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7357133660725781,
+      "grad_norm": 0.41100969910621643,
+      "learning_rate": 8.789073453902814e-05,
+      "loss": 3.1353,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7364490794386507,
+      "grad_norm": 0.39325809478759766,
+      "learning_rate": 8.786495599882943e-05,
+      "loss": 3.1665,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7371847928047233,
+      "grad_norm": 0.39705923199653625,
+      "learning_rate": 8.783915383781099e-05,
+      "loss": 3.181,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7379205061707959,
+      "grad_norm": 0.3803333640098572,
+      "learning_rate": 8.781332807206861e-05,
+      "loss": 3.1477,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7386562195368684,
+      "grad_norm": 0.38700178265571594,
+      "learning_rate": 8.778747871771292e-05,
+      "loss": 3.2384,
+      "step": 10040
+    },
+    {
+      "epoch": 0.739391932902941,
+      "grad_norm": 0.4298153519630432,
+      "learning_rate": 8.776160579086916e-05,
+      "loss": 3.2115,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7401276462690136,
+      "grad_norm": 0.3944063186645508,
+      "learning_rate": 8.773570930767732e-05,
+      "loss": 3.1313,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7408633596350862,
+      "grad_norm": 0.3921501636505127,
+      "learning_rate": 8.770978928429208e-05,
+      "loss": 3.1643,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7415990730011588,
+      "grad_norm": 0.3809838593006134,
+      "learning_rate": 8.76838457368828e-05,
+      "loss": 3.2096,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7423347863672314,
+      "grad_norm": 0.38074982166290283,
+      "learning_rate": 8.765787868163351e-05,
+      "loss": 3.1887,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7430704997333039,
+      "grad_norm": 0.401103138923645,
+      "learning_rate": 8.763188813474293e-05,
+      "loss": 3.1852,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7438062130993764,
+      "grad_norm": 0.38959553837776184,
+      "learning_rate": 8.76058741124244e-05,
+      "loss": 3.1682,
+      "step": 10110
+    },
+    {
+      "epoch": 0.744541926465449,
+      "grad_norm": 0.41069236397743225,
+      "learning_rate": 8.757983663090592e-05,
+      "loss": 3.1528,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7452776398315216,
+      "grad_norm": 0.4105203449726105,
+      "learning_rate": 8.755377570643014e-05,
+      "loss": 3.177,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7460133531975942,
+      "grad_norm": 0.4029715657234192,
+      "learning_rate": 8.752769135525431e-05,
+      "loss": 3.1583,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7467490665636668,
+      "grad_norm": 0.37962058186531067,
+      "learning_rate": 8.750158359365031e-05,
+      "loss": 3.1393,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7474847799297394,
+      "grad_norm": 0.39237701892852783,
+      "learning_rate": 8.74754524379046e-05,
+      "loss": 3.192,
+      "step": 10160
+    },
+    {
+      "epoch": 0.748220493295812,
+      "grad_norm": 0.40408843755722046,
+      "learning_rate": 8.744929790431826e-05,
+      "loss": 3.1452,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7489562066618846,
+      "grad_norm": 0.38078784942626953,
+      "learning_rate": 8.742312000920696e-05,
+      "loss": 3.1608,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7496919200279571,
+      "grad_norm": 0.37361815571784973,
+      "learning_rate": 8.73969187689009e-05,
+      "loss": 3.1642,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7504276333940297,
+      "grad_norm": 0.3917367160320282,
+      "learning_rate": 8.737069419974489e-05,
+      "loss": 3.1804,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7511633467601022,
+      "grad_norm": 0.3755326271057129,
+      "learning_rate": 8.734444631809827e-05,
+      "loss": 3.1408,
+      "step": 10210
+    },
+    {
+      "epoch": 0.7518990601261748,
+      "grad_norm": 0.38932299613952637,
+      "learning_rate": 8.73181751403349e-05,
+      "loss": 3.1216,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7526347734922474,
+      "grad_norm": 0.3907861113548279,
+      "learning_rate": 8.729188068284324e-05,
+      "loss": 3.1688,
+      "step": 10230
+    },
+    {
+      "epoch": 0.75337048685832,
+      "grad_norm": 0.4114513099193573,
+      "learning_rate": 8.726556296202621e-05,
+      "loss": 3.2136,
+      "step": 10240
+    },
+    {
+      "epoch": 0.7541062002243926,
+      "grad_norm": 0.38536152243614197,
+      "learning_rate": 8.723922199430127e-05,
+      "loss": 3.1645,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7548419135904652,
+      "grad_norm": 0.39056259393692017,
+      "learning_rate": 8.721285779610035e-05,
+      "loss": 3.1345,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7555776269565377,
+      "grad_norm": 0.37613552808761597,
+      "learning_rate": 8.718647038386995e-05,
+      "loss": 3.123,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7563133403226103,
+      "grad_norm": 0.4041471481323242,
+      "learning_rate": 8.716005977407093e-05,
+      "loss": 3.2016,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7570490536886829,
+      "grad_norm": 0.3961995244026184,
+      "learning_rate": 8.713362598317875e-05,
+      "loss": 3.1718,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7577847670547555,
+      "grad_norm": 0.3890473246574402,
+      "learning_rate": 8.710716902768324e-05,
+      "loss": 3.1566,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7585204804208281,
+      "grad_norm": 0.38676023483276367,
+      "learning_rate": 8.708068892408872e-05,
+      "loss": 3.1446,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7592561937869006,
+      "grad_norm": 0.4547286331653595,
+      "learning_rate": 8.705418568891391e-05,
+      "loss": 3.1901,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7599919071529732,
+      "grad_norm": 0.37865954637527466,
+      "learning_rate": 8.702765933869203e-05,
+      "loss": 3.1168,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7607276205190457,
+      "grad_norm": 0.3800100088119507,
+      "learning_rate": 8.700110988997066e-05,
+      "loss": 3.1179,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7614633338851183,
+      "grad_norm": 0.38291603326797485,
+      "learning_rate": 8.697453735931182e-05,
+      "loss": 3.1596,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7621990472511909,
+      "grad_norm": 0.3799181282520294,
+      "learning_rate": 8.69479417632919e-05,
+      "loss": 3.222,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7629347606172635,
+      "grad_norm": 0.3791024684906006,
+      "learning_rate": 8.692132311850171e-05,
+      "loss": 3.1126,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7636704739833361,
+      "grad_norm": 0.37659361958503723,
+      "learning_rate": 8.689468144154642e-05,
+      "loss": 3.1594,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7644061873494087,
+      "grad_norm": 0.3841865062713623,
+      "learning_rate": 8.686801674904557e-05,
+      "loss": 3.1806,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7651419007154813,
+      "grad_norm": 0.3803307116031647,
+      "learning_rate": 8.684132905763309e-05,
+      "loss": 3.1807,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7658776140815539,
+      "grad_norm": 0.38950908184051514,
+      "learning_rate": 8.681461838395719e-05,
+      "loss": 3.1673,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7666133274476264,
+      "grad_norm": 0.3641705811023712,
+      "learning_rate": 8.678788474468045e-05,
+      "loss": 3.1618,
+      "step": 10420
+    },
+    {
+      "epoch": 0.767349040813699,
+      "grad_norm": 0.3801932632923126,
+      "learning_rate": 8.67611281564798e-05,
+      "loss": 3.1935,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7680847541797715,
+      "grad_norm": 0.36981040239334106,
+      "learning_rate": 8.673434863604648e-05,
+      "loss": 3.1661,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7688204675458441,
+      "grad_norm": 0.3908541202545166,
+      "learning_rate": 8.670754620008598e-05,
+      "loss": 3.1673,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7695561809119167,
+      "grad_norm": 0.3636099398136139,
+      "learning_rate": 8.668072086531818e-05,
+      "loss": 3.1264,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7702918942779893,
+      "grad_norm": 0.37544921040534973,
+      "learning_rate": 8.665387264847715e-05,
+      "loss": 3.1016,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7710276076440619,
+      "grad_norm": 0.39099445939064026,
+      "learning_rate": 8.662700156631127e-05,
+      "loss": 3.2002,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7717633210101345,
+      "grad_norm": 0.3761730492115021,
+      "learning_rate": 8.660010763558324e-05,
+      "loss": 3.1221,
+      "step": 10490
+    },
+    {
+      "epoch": 0.772499034376207,
+      "grad_norm": 0.4066612124443054,
+      "learning_rate": 8.65731908730699e-05,
+      "loss": 3.1845,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7732347477422796,
+      "grad_norm": 0.3759371340274811,
+      "learning_rate": 8.654625129556244e-05,
+      "loss": 3.1465,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7739704611083522,
+      "grad_norm": 0.3886657655239105,
+      "learning_rate": 8.651928891986622e-05,
+      "loss": 3.1517,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7747061744744248,
+      "grad_norm": 0.38135573267936707,
+      "learning_rate": 8.649230376280087e-05,
+      "loss": 3.1349,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7754418878404974,
+      "grad_norm": 0.4106856882572174,
+      "learning_rate": 8.646529584120015e-05,
+      "loss": 3.2045,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7761776012065699,
+      "grad_norm": 0.40560615062713623,
+      "learning_rate": 8.64382651719121e-05,
+      "loss": 3.1467,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7769133145726425,
+      "grad_norm": 0.3781823515892029,
+      "learning_rate": 8.641121177179893e-05,
+      "loss": 3.1531,
+      "step": 10560
+    },
+    {
+      "epoch": 0.777649027938715,
+      "grad_norm": 0.37491971254348755,
+      "learning_rate": 8.6384135657737e-05,
+      "loss": 3.1884,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7783847413047876,
+      "grad_norm": 0.380998432636261,
+      "learning_rate": 8.635703684661689e-05,
+      "loss": 3.1635,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7791204546708602,
+      "grad_norm": 0.4219949245452881,
+      "learning_rate": 8.632991535534329e-05,
+      "loss": 3.1436,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7798561680369328,
+      "grad_norm": 0.4060940742492676,
+      "learning_rate": 8.630277120083507e-05,
+      "loss": 3.2004,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7805918814030054,
+      "grad_norm": 0.36930394172668457,
+      "learning_rate": 8.627560440002524e-05,
+      "loss": 3.1599,
+      "step": 10610
+    },
+    {
+      "epoch": 0.781327594769078,
+      "grad_norm": 0.3666263222694397,
+      "learning_rate": 8.624841496986089e-05,
+      "loss": 3.1352,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7820633081351506,
+      "grad_norm": 0.4501335322856903,
+      "learning_rate": 8.622120292730327e-05,
+      "loss": 3.1673,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7827990215012232,
+      "grad_norm": 0.4114534258842468,
+      "learning_rate": 8.619396828932774e-05,
+      "loss": 3.1506,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7835347348672957,
+      "grad_norm": 0.37227872014045715,
+      "learning_rate": 8.616671107292373e-05,
+      "loss": 3.2114,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7842704482333683,
+      "grad_norm": 0.39157432317733765,
+      "learning_rate": 8.613943129509478e-05,
+      "loss": 3.2101,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7850061615994408,
+      "grad_norm": 0.41202887892723083,
+      "learning_rate": 8.61121289728585e-05,
+      "loss": 3.1828,
+      "step": 10670
+    },
+    {
+      "epoch": 0.7857418749655134,
+      "grad_norm": 0.3904684782028198,
+      "learning_rate": 8.608480412324652e-05,
+      "loss": 3.2128,
+      "step": 10680
+    },
+    {
+      "epoch": 0.786477588331586,
+      "grad_norm": 0.39074650406837463,
+      "learning_rate": 8.605745676330457e-05,
+      "loss": 3.1899,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7872133016976586,
+      "grad_norm": 0.4131028950214386,
+      "learning_rate": 8.603008691009243e-05,
+      "loss": 3.1664,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7879490150637312,
+      "grad_norm": 0.3683393597602844,
+      "learning_rate": 8.600269458068387e-05,
+      "loss": 3.1694,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7886847284298037,
+      "grad_norm": 0.38590583205223083,
+      "learning_rate": 8.597527979216672e-05,
+      "loss": 3.1582,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7894204417958763,
+      "grad_norm": 0.38443443179130554,
+      "learning_rate": 8.594784256164279e-05,
+      "loss": 3.1427,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7901561551619489,
+      "grad_norm": 0.38856041431427,
+      "learning_rate": 8.59203829062279e-05,
+      "loss": 3.1903,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7908918685280215,
+      "grad_norm": 0.3836011290550232,
+      "learning_rate": 8.589290084305187e-05,
+      "loss": 3.159,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7916275818940941,
+      "grad_norm": 0.3806968629360199,
+      "learning_rate": 8.58653963892585e-05,
+      "loss": 3.1668,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7923632952601667,
+      "grad_norm": 0.41352030634880066,
+      "learning_rate": 8.583786956200553e-05,
+      "loss": 3.1994,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7930990086262392,
+      "grad_norm": 0.3863872289657593,
+      "learning_rate": 8.581032037846466e-05,
+      "loss": 3.174,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7938347219923118,
+      "grad_norm": 0.4021219313144684,
+      "learning_rate": 8.578274885582159e-05,
+      "loss": 3.1763,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7945704353583843,
+      "grad_norm": 0.3936547338962555,
+      "learning_rate": 8.575515501127592e-05,
+      "loss": 3.1346,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7953061487244569,
+      "grad_norm": 0.38866353034973145,
+      "learning_rate": 8.57275388620411e-05,
+      "loss": 3.1529,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7960418620905295,
+      "grad_norm": 0.3719068467617035,
+      "learning_rate": 8.569990042534466e-05,
+      "loss": 3.0875,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7967775754566021,
+      "grad_norm": 0.37235334515571594,
+      "learning_rate": 8.567223971842788e-05,
+      "loss": 3.1509,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7975132888226747,
+      "grad_norm": 0.3882967531681061,
+      "learning_rate": 8.564455675854599e-05,
+      "loss": 3.1392,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7982490021887473,
+      "grad_norm": 0.3814670145511627,
+      "learning_rate": 8.561685156296813e-05,
+      "loss": 3.1725,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7989847155548199,
+      "grad_norm": 0.38729429244995117,
+      "learning_rate": 8.558912414897728e-05,
+      "loss": 3.1396,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7997204289208925,
+      "grad_norm": 0.3847314715385437,
+      "learning_rate": 8.556137453387027e-05,
+      "loss": 3.0893,
+      "step": 10870
+    },
+    {
+      "epoch": 0.800456142286965,
+      "grad_norm": 0.3961573541164398,
+      "learning_rate": 8.553360273495779e-05,
+      "loss": 3.146,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8011918556530375,
+      "grad_norm": 0.3685840666294098,
+      "learning_rate": 8.55058087695644e-05,
+      "loss": 3.1371,
+      "step": 10890
+    },
+    {
+      "epoch": 0.8019275690191101,
+      "grad_norm": 0.37563103437423706,
+      "learning_rate": 8.547799265502844e-05,
+      "loss": 3.1751,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8026632823851827,
+      "grad_norm": 0.38960275053977966,
+      "learning_rate": 8.545015440870208e-05,
+      "loss": 3.1834,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8033989957512553,
+      "grad_norm": 0.3829275369644165,
+      "learning_rate": 8.542229404795132e-05,
+      "loss": 3.1301,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8041347091173279,
+      "grad_norm": 0.37760934233665466,
+      "learning_rate": 8.539441159015594e-05,
+      "loss": 3.1482,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8048704224834005,
+      "grad_norm": 0.3774561285972595,
+      "learning_rate": 8.536650705270948e-05,
+      "loss": 3.1495,
+      "step": 10940
+    },
+    {
+      "epoch": 0.805606135849473,
+      "grad_norm": 0.36744073033332825,
+      "learning_rate": 8.53385804530193e-05,
+      "loss": 3.1253,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8063418492155456,
+      "grad_norm": 0.3932728171348572,
+      "learning_rate": 8.53106318085065e-05,
+      "loss": 3.1845,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8070775625816182,
+      "grad_norm": 0.38730084896087646,
+      "learning_rate": 8.528266113660591e-05,
+      "loss": 3.1891,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8078132759476908,
+      "grad_norm": 0.382874995470047,
+      "learning_rate": 8.525466845476616e-05,
+      "loss": 3.1655,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8085489893137634,
+      "grad_norm": 0.36904412508010864,
+      "learning_rate": 8.522665378044952e-05,
+      "loss": 3.1939,
+      "step": 10990
+    },
+    {
+      "epoch": 0.809284702679836,
+      "grad_norm": 0.36776015162467957,
+      "learning_rate": 8.519861713113206e-05,
+      "loss": 3.1426,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8100204160459085,
+      "grad_norm": 0.3698399066925049,
+      "learning_rate": 8.517055852430356e-05,
+      "loss": 3.1427,
+      "step": 11010
+    },
+    {
+      "epoch": 0.810756129411981,
+      "grad_norm": 0.3883744478225708,
+      "learning_rate": 8.514247797746744e-05,
+      "loss": 3.1611,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8114918427780536,
+      "grad_norm": 0.37801653146743774,
+      "learning_rate": 8.51143755081408e-05,
+      "loss": 3.1424,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8122275561441262,
+      "grad_norm": 0.3822377026081085,
+      "learning_rate": 8.508625113385453e-05,
+      "loss": 3.0966,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8129632695101988,
+      "grad_norm": 0.3712664246559143,
+      "learning_rate": 8.505810487215305e-05,
+      "loss": 3.1409,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8136989828762714,
+      "grad_norm": 0.37806233763694763,
+      "learning_rate": 8.502993674059449e-05,
+      "loss": 3.1397,
+      "step": 11060
+    },
+    {
+      "epoch": 0.814434696242344,
+      "grad_norm": 0.3708866238594055,
+      "learning_rate": 8.500174675675063e-05,
+      "loss": 3.1526,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8151704096084166,
+      "grad_norm": 0.3839915990829468,
+      "learning_rate": 8.497353493820688e-05,
+      "loss": 3.123,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8159061229744892,
+      "grad_norm": 0.4151860475540161,
+      "learning_rate": 8.494530130256227e-05,
+      "loss": 3.1412,
+      "step": 11090
+    },
+    {
+      "epoch": 0.8166418363405618,
+      "grad_norm": 0.3969019949436188,
+      "learning_rate": 8.49170458674294e-05,
+      "loss": 3.1789,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8173775497066343,
+      "grad_norm": 0.37846624851226807,
+      "learning_rate": 8.488876865043454e-05,
+      "loss": 3.1684,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8181132630727068,
+      "grad_norm": 0.37555307149887085,
+      "learning_rate": 8.48604696692175e-05,
+      "loss": 3.1397,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8188489764387794,
+      "grad_norm": 0.39476168155670166,
+      "learning_rate": 8.483214894143166e-05,
+      "loss": 3.1518,
+      "step": 11130
+    },
+    {
+      "epoch": 0.819584689804852,
+      "grad_norm": 0.3914696276187897,
+      "learning_rate": 8.480380648474403e-05,
+      "loss": 3.1596,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8203204031709246,
+      "grad_norm": 0.3912908136844635,
+      "learning_rate": 8.477544231683507e-05,
+      "loss": 3.1871,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8210561165369972,
+      "grad_norm": 0.37355631589889526,
+      "learning_rate": 8.474705645539886e-05,
+      "loss": 3.1434,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8217918299030698,
+      "grad_norm": 0.3698188066482544,
+      "learning_rate": 8.471864891814304e-05,
+      "loss": 3.0801,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8225275432691423,
+      "grad_norm": 0.3645183742046356,
+      "learning_rate": 8.469021972278867e-05,
+      "loss": 3.1388,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8232632566352149,
+      "grad_norm": 0.3691493272781372,
+      "learning_rate": 8.466176888707041e-05,
+      "loss": 3.1758,
+      "step": 11190
+    },
+    {
+      "epoch": 0.8239989700012875,
+      "grad_norm": 0.37398234009742737,
+      "learning_rate": 8.463329642873638e-05,
+      "loss": 3.1055,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8247346833673601,
+      "grad_norm": 0.38691118359565735,
+      "learning_rate": 8.46048023655482e-05,
+      "loss": 3.1475,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8254703967334327,
+      "grad_norm": 0.36699140071868896,
+      "learning_rate": 8.457628671528094e-05,
+      "loss": 3.1344,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8262061100995053,
+      "grad_norm": 0.3784330487251282,
+      "learning_rate": 8.454774949572318e-05,
+      "loss": 3.1172,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8269418234655778,
+      "grad_norm": 0.3638308644294739,
+      "learning_rate": 8.451919072467693e-05,
+      "loss": 3.1731,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8276775368316504,
+      "grad_norm": 0.3659362196922302,
+      "learning_rate": 8.449061041995766e-05,
+      "loss": 3.1357,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8284132501977229,
+      "grad_norm": 0.37726885080337524,
+      "learning_rate": 8.446200859939425e-05,
+      "loss": 3.109,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8291489635637955,
+      "grad_norm": 0.3847838044166565,
+      "learning_rate": 8.443338528082902e-05,
+      "loss": 3.1779,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8298846769298681,
+      "grad_norm": 0.3812064230442047,
+      "learning_rate": 8.440474048211768e-05,
+      "loss": 3.158,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8306203902959407,
+      "grad_norm": 0.40196287631988525,
+      "learning_rate": 8.437607422112938e-05,
+      "loss": 3.173,
+      "step": 11290
+    },
+    {
+      "epoch": 0.8313561036620133,
+      "grad_norm": 0.3687420189380646,
+      "learning_rate": 8.434738651574662e-05,
+      "loss": 3.1256,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8320918170280859,
+      "grad_norm": 0.36912721395492554,
+      "learning_rate": 8.431867738386531e-05,
+      "loss": 3.1726,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8328275303941585,
+      "grad_norm": 0.374905526638031,
+      "learning_rate": 8.428994684339466e-05,
+      "loss": 3.1554,
+      "step": 11320
+    },
+    {
+      "epoch": 0.833563243760231,
+      "grad_norm": 0.37698689103126526,
+      "learning_rate": 8.426119491225733e-05,
+      "loss": 3.1715,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8342989571263036,
+      "grad_norm": 0.3949843645095825,
+      "learning_rate": 8.423242160838926e-05,
+      "loss": 3.145,
+      "step": 11340
+    },
+    {
+      "epoch": 0.8350346704923761,
+      "grad_norm": 0.38596290349960327,
+      "learning_rate": 8.420362694973972e-05,
+      "loss": 3.1545,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8357703838584487,
+      "grad_norm": 0.361113041639328,
+      "learning_rate": 8.417481095427134e-05,
+      "loss": 3.1668,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8365060972245213,
+      "grad_norm": 0.38437777757644653,
+      "learning_rate": 8.414597363996004e-05,
+      "loss": 3.1842,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8372418105905939,
+      "grad_norm": 0.38113656640052795,
+      "learning_rate": 8.411711502479505e-05,
+      "loss": 3.1228,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8379775239566665,
+      "grad_norm": 0.37318846583366394,
+      "learning_rate": 8.408823512677883e-05,
+      "loss": 3.1408,
+      "step": 11390
+    },
+    {
+      "epoch": 0.8387132373227391,
+      "grad_norm": 0.3935425579547882,
+      "learning_rate": 8.405933396392722e-05,
+      "loss": 3.1441,
+      "step": 11400
+    },
+    {
+      "epoch": 0.8394489506888116,
+      "grad_norm": 0.37884998321533203,
+      "learning_rate": 8.403041155426925e-05,
+      "loss": 3.1926,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8401846640548842,
+      "grad_norm": 0.38540568947792053,
+      "learning_rate": 8.400146791584722e-05,
+      "loss": 3.1961,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8409203774209568,
+      "grad_norm": 0.392183393239975,
+      "learning_rate": 8.397250306671666e-05,
+      "loss": 3.0998,
+      "step": 11430
+    },
+    {
+      "epoch": 0.8416560907870294,
+      "grad_norm": 0.3929624557495117,
+      "learning_rate": 8.394351702494637e-05,
+      "loss": 3.1336,
+      "step": 11440
+    },
+    {
+      "epoch": 0.842391804153102,
+      "grad_norm": 0.38060176372528076,
+      "learning_rate": 8.391450980861835e-05,
+      "loss": 3.1406,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8431275175191745,
+      "grad_norm": 0.38198819756507874,
+      "learning_rate": 8.38854814358278e-05,
+      "loss": 3.1256,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8438632308852471,
+      "grad_norm": 0.3775307536125183,
+      "learning_rate": 8.385643192468312e-05,
+      "loss": 3.1567,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8445989442513196,
+      "grad_norm": 0.3741263449192047,
+      "learning_rate": 8.382736129330591e-05,
+      "loss": 3.1545,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8453346576173922,
+      "grad_norm": 0.3674101233482361,
+      "learning_rate": 8.379826955983092e-05,
+      "loss": 3.1613,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8460703709834648,
+      "grad_norm": 0.38992542028427124,
+      "learning_rate": 8.376915674240611e-05,
+      "loss": 3.1323,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8468060843495374,
+      "grad_norm": 0.3849126994609833,
+      "learning_rate": 8.374002285919253e-05,
+      "loss": 3.2039,
+      "step": 11510
+    },
+    {
+      "epoch": 0.84754179771561,
+      "grad_norm": 0.3764283061027527,
+      "learning_rate": 8.371086792836443e-05,
+      "loss": 3.125,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8482775110816826,
+      "grad_norm": 0.37651705741882324,
+      "learning_rate": 8.368169196810916e-05,
+      "loss": 3.1228,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8490132244477552,
+      "grad_norm": 0.3919517397880554,
+      "learning_rate": 8.365249499662719e-05,
+      "loss": 3.1619,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8497489378138278,
+      "grad_norm": 0.377399206161499,
+      "learning_rate": 8.362327703213208e-05,
+      "loss": 3.1564,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8504846511799004,
+      "grad_norm": 0.3962908983230591,
+      "learning_rate": 8.359403809285053e-05,
+      "loss": 3.1482,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8512203645459729,
+      "grad_norm": 0.399509996175766,
+      "learning_rate": 8.356477819702231e-05,
+      "loss": 3.0999,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8519560779120454,
+      "grad_norm": 0.4016306698322296,
+      "learning_rate": 8.353549736290023e-05,
+      "loss": 3.1327,
+      "step": 11580
+    },
+    {
+      "epoch": 0.852691791278118,
+      "grad_norm": 0.38924241065979004,
+      "learning_rate": 8.350619560875017e-05,
+      "loss": 3.1378,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8534275046441906,
+      "grad_norm": 0.39699849486351013,
+      "learning_rate": 8.347687295285112e-05,
+      "loss": 3.131,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8541632180102632,
+      "grad_norm": 0.36073020100593567,
+      "learning_rate": 8.344752941349503e-05,
+      "loss": 3.1618,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8548989313763358,
+      "grad_norm": 0.38798099756240845,
+      "learning_rate": 8.341816500898692e-05,
+      "loss": 3.1768,
+      "step": 11620
+    },
+    {
+      "epoch": 0.8556346447424084,
+      "grad_norm": 0.37226933240890503,
+      "learning_rate": 8.338877975764483e-05,
+      "loss": 3.1506,
+      "step": 11630
+    },
+    {
+      "epoch": 0.856370358108481,
+      "grad_norm": 0.3917200565338135,
+      "learning_rate": 8.335937367779976e-05,
+      "loss": 3.1574,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8571060714745535,
+      "grad_norm": 0.37370383739471436,
+      "learning_rate": 8.332994678779579e-05,
+      "loss": 3.1438,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8578417848406261,
+      "grad_norm": 0.3972039222717285,
+      "learning_rate": 8.330049910598987e-05,
+      "loss": 3.1557,
+      "step": 11660
+    },
+    {
+      "epoch": 0.8585774982066987,
+      "grad_norm": 0.38387760519981384,
+      "learning_rate": 8.327103065075202e-05,
+      "loss": 3.1509,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8593132115727713,
+      "grad_norm": 0.36181387305259705,
+      "learning_rate": 8.324154144046515e-05,
+      "loss": 3.1726,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8600489249388438,
+      "grad_norm": 0.3810345530509949,
+      "learning_rate": 8.321203149352513e-05,
+      "loss": 3.1901,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8607846383049164,
+      "grad_norm": 0.3864314556121826,
+      "learning_rate": 8.318250082834083e-05,
+      "loss": 3.1788,
+      "step": 11700
+    },
+    {
+      "epoch": 0.861520351670989,
+      "grad_norm": 0.39438217878341675,
+      "learning_rate": 8.315294946333396e-05,
+      "loss": 3.1094,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8622560650370615,
+      "grad_norm": 0.3909660279750824,
+      "learning_rate": 8.312337741693917e-05,
+      "loss": 3.1339,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8629917784031341,
+      "grad_norm": 0.37569791078567505,
+      "learning_rate": 8.309378470760402e-05,
+      "loss": 3.2063,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8637274917692067,
+      "grad_norm": 0.3675804138183594,
+      "learning_rate": 8.306417135378899e-05,
+      "loss": 3.1175,
+      "step": 11740
+    },
+    {
+      "epoch": 0.8644632051352793,
+      "grad_norm": 0.3829854726791382,
+      "learning_rate": 8.303453737396736e-05,
+      "loss": 3.1524,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8651989185013519,
+      "grad_norm": 0.3902895152568817,
+      "learning_rate": 8.300488278662537e-05,
+      "loss": 3.1543,
+      "step": 11760
+    },
+    {
+      "epoch": 0.8659346318674245,
+      "grad_norm": 0.3615528345108032,
+      "learning_rate": 8.297520761026203e-05,
+      "loss": 3.1835,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8666703452334971,
+      "grad_norm": 0.3822636008262634,
+      "learning_rate": 8.294551186338925e-05,
+      "loss": 3.1482,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8674060585995697,
+      "grad_norm": 0.36693650484085083,
+      "learning_rate": 8.291579556453176e-05,
+      "loss": 3.1794,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8681417719656422,
+      "grad_norm": 0.36783790588378906,
+      "learning_rate": 8.28860587322271e-05,
+      "loss": 3.1473,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8688774853317147,
+      "grad_norm": 0.36205679178237915,
+      "learning_rate": 8.28563013850256e-05,
+      "loss": 3.1309,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8696131986977873,
+      "grad_norm": 0.37128111720085144,
+      "learning_rate": 8.282652354149047e-05,
+      "loss": 3.1223,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8703489120638599,
+      "grad_norm": 0.3675532341003418,
+      "learning_rate": 8.279672522019759e-05,
+      "loss": 3.1302,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8710846254299325,
+      "grad_norm": 0.36570990085601807,
+      "learning_rate": 8.27669064397357e-05,
+      "loss": 3.0489,
+      "step": 11840
+    },
+    {
+      "epoch": 0.8718203387960051,
+      "grad_norm": 0.3862569034099579,
+      "learning_rate": 8.273706721870626e-05,
+      "loss": 3.0992,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8725560521620777,
+      "grad_norm": 0.35596564412117004,
+      "learning_rate": 8.270720757572352e-05,
+      "loss": 3.173,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8732917655281502,
+      "grad_norm": 0.38876011967658997,
+      "learning_rate": 8.267732752941443e-05,
+      "loss": 3.1282,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8740274788942228,
+      "grad_norm": 0.3655962646007538,
+      "learning_rate": 8.264742709841868e-05,
+      "loss": 3.1591,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8747631922602954,
+      "grad_norm": 0.378047376871109,
+      "learning_rate": 8.26175063013887e-05,
+      "loss": 3.1106,
+      "step": 11890
+    },
+    {
+      "epoch": 0.875498905626368,
+      "grad_norm": 0.41224589943885803,
+      "learning_rate": 8.258756515698958e-05,
+      "loss": 3.1519,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8762346189924406,
+      "grad_norm": 0.3781079351902008,
+      "learning_rate": 8.255760368389914e-05,
+      "loss": 3.1649,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8769703323585131,
+      "grad_norm": 0.3772408366203308,
+      "learning_rate": 8.252762190080787e-05,
+      "loss": 3.1815,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8777060457245857,
+      "grad_norm": 0.36887019872665405,
+      "learning_rate": 8.249761982641893e-05,
+      "loss": 3.1422,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8784417590906582,
+      "grad_norm": 0.37302523851394653,
+      "learning_rate": 8.246759747944813e-05,
+      "loss": 3.1111,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8791774724567308,
+      "grad_norm": 0.3750922977924347,
+      "learning_rate": 8.243755487862392e-05,
+      "loss": 3.1001,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8799131858228034,
+      "grad_norm": 0.3755902647972107,
+      "learning_rate": 8.240749204268745e-05,
+      "loss": 3.134,
+      "step": 11960
+    },
+    {
+      "epoch": 0.880648899188876,
+      "grad_norm": 0.3703187108039856,
+      "learning_rate": 8.237740899039238e-05,
+      "loss": 3.1396,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8813846125549486,
+      "grad_norm": 0.37014931440353394,
+      "learning_rate": 8.234730574050509e-05,
+      "loss": 3.1177,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8821203259210212,
+      "grad_norm": 0.39811164140701294,
+      "learning_rate": 8.231718231180448e-05,
+      "loss": 3.1516,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8828560392870938,
+      "grad_norm": 0.37592417001724243,
+      "learning_rate": 8.22870387230821e-05,
+      "loss": 3.1192,
+      "step": 12000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 40776,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 4000,
+  "total_flos": 1.087952394387456e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}